key: cord-0286766-cospko9t
authors: Le Nouën, Cyril; Nelson, Christine E.; Liu, Xueqiao; Park, Hong-Su; Matsuoka, Yumiko; Luongo, Cindy; Santos, Celia; Yang, Lijuan; Herbert, Richard; Castens, Ashley; Moore, Ian N.; Wilder-Kofie, Temeri; Moore, Rashida; Walker, April; Zhang, Peng; Lusso, Paolo; Johnson, Reed F.; Garza, Nicole L.; Via, Laura E.; Munir, Shirin; Barber, Daniel; Buchholz, Ursula J.
title: Intranasal pediatric parainfluenza virus-vectored SARS-CoV-2 vaccine candidate is protective in macaques
date: 2022-05-23
journal: bioRxiv
DOI: 10.1101/2022.05.21.492923
sha: 2ba0e63e0f06313fff858c1973bd51114377de8c
doc_id: 286766
cord_uid: cospko9t

Pediatric SARS-CoV-2 vaccines are needed that elicit immunity directly in the airways, as well as systemically. Building on pediatric parainfluenza virus vaccines in clinical development, we generated a live-attenuated parainfluenza virus-vectored vaccine candidate expressing SARS-CoV-2 prefusion-stabilized spike (S) protein (B/HPIV3/S-6P) and evaluated its immunogenicity and protective efficacy in rhesus macaques. A single intranasal/intratracheal dose of B/HPIV3/S-6P induced strong S-specific airway mucosal IgA and IgG responses. High levels of S-specific antibodies were also induced in serum, which efficiently neutralized SARS-CoV-2 variants of concern. Furthermore, B/HPIV3/S-6P induced robust systemic and pulmonary S-specific CD4+ and CD8+ T-cell responses, including tissue-resident memory cells in lungs. Following challenge, SARS-CoV-2 replication was undetectable in airways and lung tissues of immunized macaques. B/HPIV3/S-6P will be evaluated clinically as pediatric intranasal SARS-CoV-2/parainfluenza virus type 3 vaccine. One-Sentence Summary Intranasal parainfluenza virus-vectored COVID-19 vaccine induces anti-S antibodies, T-cell memory and protection in macaques.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infects and causes disease in all age groups. Although COVID-19 is generally milder in young children than in adults, thousands of children have been hospitalized due to COVID-19 in the US alone, including about one-third without preexisting medical conditions (Funk et al., 2022; Rankin et al., 2021) . More than 800 children 0 to 11 years of age have died from COVID-19 in the US (https://covid.cdc.gov/covid-data-tracker/#demographics, accessed on April 22, 2022), and during the fall/winter COVID-19 wave of 2021/2022, children accounted for over 25% of US cases (Gerber and Offit, 2021) . In rare cases (∼0.03% of infected children), COVID-19 can cause a multisystem inflammatory syndrome in children [MIS-C; (Riphagen et al., 2020; Verdoni et al., 2020) ], arising within about 4 weeks after SARS-CoV-2 exposure. While mRNAbased vaccines are available for children 5 years of age and older, to date, no vaccine has been authorized or recommended for children under 5 years of age. Furthermore, interim results from an ongoing Phase 1/2/3 clinical study suggested that a third 3-µg dose of the BNT162b2 mRNA vaccine may be needed to elicit robust immune responses in children >2 to <5 years of age; accordingly, a 3-dose regimen is being evaluated in children >6 months to <5 years of age (NCT04816643). One limitation of the current mRNA and other injectable SARS-CoV-2 vaccines is that they do not directly stimulate immunity in the respiratory tract, the major site of SARS-CoV-2 entry, replication, disease, and egress (DiPiazza et al., 2021) . Therefore, it is important to evaluate additional vaccine approaches that are suitable for pediatric use and stimulate both mucosal and systemic immunity. Ideally, a vaccine should be effective at a single dose, and could be administered topically to the respiratory tract to induce robust systemic and respiratory mucosal immunity that restricts SARS-CoV-2 infection and shedding.

Here, we describe the pre-clinical evaluation of the safety, immunogenicity, and protective efficacy of a live intranasal SARS-CoV-2 vaccine candidate, B/HPIV3/S-6P, in rhesus macaques. B/HPIV3/S-6P consists of a live-attenuated chimeric bovine/human parainfluenza virus type-3 (B/HPIV3) that was modified to express the SARS-CoV-2 S-protein trimer stabilized in the prefusion form. The B/HPIV3 vector originally was developed as a pediatric vaccine candidate against human PIV3 (HPIV3), a single-stranded negative-sense RNA virus which is an important cause of respiratory illness, especially in infants and young children under 5 years of age (DeGroote et al., 2020; Howard et al., 2021) . Previously, B/HPIV3 was welltolerated in a Phase 1 study in infants and young children (Karron et al., 2012) . B/HPIV3 also has been used to express the fusion (F) glycoprotein of another human respiratory pathogen, human respiratory syncytial virus (RSV), providing a bivalent HPIV3/RSV vaccine candidate which was well-tolerated in children >2 months of age [(Bernstein et al., 2012) , Clinicaltrials.gov NCT00686075] . We recently reported that B/HPIV3 expressing a stabilized prefusion form of S efficiently protected hamsters against infection with a vaccine antigenmatched SARS-CoV-2 isolate, preventing weight loss, lung inflammation and efficiently reducing SARS-CoV-2 replication in the upper and lower respiratory tract . In the present study, we evaluated the safety and efficacy of a single intranasal/intratracheal (IN/IT) dose of B/HPIV3/S-6P in rhesus macaques (RM). Immunogenicity evaluations included Sspecific mucosal and systemic antibody and T-cell responses as well as neutralizing antibody responses to the vaccine-matched SARS-CoV-2 strain and four major variants of concern. In addition, we assessed the protective efficacy of B/HPIV3/S-6P against SARS-CoV-2 challenge prior to advancing this candidate to a Phase 1 clinical study.

We used B/HPIV3 to express a prefusion-stabilized version of the SARS-CoV-2 S protein.

B/HPIV3 is a cDNA-derived version of bovine PIV3 (BPIV3) strain Kansas in which the BPIV3 hemagglutinin-neuraminidase (HN) and fusion (F) glycoproteins (the two PIV3 neutralization antigens) have been replaced by those of the human PIV3 strain JS (Karron et al., 2012; Schmidt et al., 2000) ( Figure 1A ). The bovine PIV3 backbone provides stable attenuation due to hostrange restriction in humans (Bernstein et al., 2012; Karron et al., 2012) . B/HPIV3/S-6P expresses a full-length prefusion-stabilized version (S-6P) of the SARS-CoV-2 S protein (1,273 aa) from a supernumerary gene, inserted between the N and P genes, an insertion site previously shown to be optimal for efficient expression and genetic stability (Liang et al., 2014) ( Figure 1A ). The S-6P version of the S-protein contains 6 proline substitutions that stabilize S in its trimeric prefusion form and increase expression and immunogenicity. The S1/S2 polybasic furin cleavage motif "RRAR" was ablated by amino acid substitutions (RRAR-to-GSAS) (Figure 1A ), rendering S-6P non-functional for virus entry, which eliminates the possibility of S altering the tissue tropism of the B/HPIV3 vector.

To evaluate vaccine replication and immunogenicity, we immunized 2 groups of RMs (n=4 per group) with a single dose of 6.3 log10 plaque-forming units (PFU) of B/HPIV3/S-6P or the B/HPIV3 vector control, respectively, administered by the combined intranasal and intratracheal route (IN/IT) ( Figure S1 ). Nasopharyngeal swabs (NS) and tracheal lavages (TL) were performed daily and every other day, respectively, from day 0 to 12 post-immunization (pi) to evaluate vaccine replication in the upper and lower airways (UA and LA, respectively; Figure   1B -C, Figure S1 ). Replication of B/HPIV3/S-6P and the B/HPIV3 control was detectable through days 8 or 9 pi in the UA and LA. In the UA, peak replication of B/HPIV3/S-6P and B/HPIV3 control was detected between study days 4 and 6 (medians independent of study day: 4.9 log10 PFU/mL vs 5.9 log10 PFU/mL, respectively; P=0.1429 by two-tailed Mann-Whitney test); replication of B/HPIV3/S-6P was delayed by 1-2 days compared to that of the empty vector (P<0.0001 on days 2 to 4) ( Figure 1B ). In the LA, B/HPIV3/S-6P replicated with similar kinetics as B/HPIV3, reaching median peak titers of 4.5 log10 PFU/mL and 4.0 log10 PFU/mL, respectively, on day 6 pi ( Figure 1C ).

To evaluate the stability of S expression during vector replication, NS and TL specimens positive for B/HPIV3/S-6P were evaluated by a dual-staining immunoplaque assay, which detects the expression of S and vector proteins. On average, 89% of the B/HPIV3/S-6P plaques recovered between days 5 and 7 from NS were positive for S expression ( Figure S2 ), suggesting stable S-6P expression in the UA. In TL specimens collected on day 6 pi, S expression was stable in 3 of 4 RM, with on average 88% of the plaques positive for S expression. In TL samples from one B/HPIV3/S-6P-immunized RM (B/HPIV3/S-6P #4), plaques were negative for S expression on day 6 pi. Sanger sequencing of the S gene revealed 13 cytidine-to-thymidine mutations in a 430-nucleotide region, suggestive of deaminase activity in the LA of this animal.

Eleven were missense mutations resulting in amino acid substitutions, including 7 proline substitutions which might affect S protein folding.

No changes in body weight, rectal temperature, respiration, oxygen saturation or pulse were detected following immunization of RMs with B/HPIV3 or B/HPIV3/S-6P ( Figure S3 ).

Thus, B/HPIV3/S-6P replicates efficiently in the UA and LA of RMs, causes no apparent symptoms, and is cleared in approximately ten days.

To assess the kinetics of airway mucosal antibody responses to the SARS-CoV-2 S protein in the UA and LA, we collected nasal washes (NW) 3 days before immunization and on days 14, 21, and 28 after immunization, and bronchoalveolar lavage (BAL) fluid on days 9, 21, and 28 pi ( Figure S1 ). We evaluated IgA and IgG binding antibodies using a soluble prefusion stabilized S-2P version of the vaccine-matched S protein or its receptor binding domain (RBD) peaked on day 21 pi in 2 RMs and remained steady, while they continued to rise until day 28 pi in the other 2 RMs (peak titers between 4.3 and 4.9 log10, P<0.01). Serum anti-RBD IgA titers peaked on day 21 in all 4 RMs (titers between 4.8 and 5.3 log10, P<0.01) and modestly declined by day 28 pi. High levels of serum anti-S and anti-RBD IgG were also measured in all B/HPIV3/S-6P-immunized RMs on day 14 pi, continuing to rise in all RMs until day 28 pi (GMTs between 5.8 to 6.4 log10 on day 28 pi, P<0.0001). These levels of anti-S and anti-RBD IgG antibodies were 16-fold and 180-fold higher than the mean anti-S and anti-RBD IgG titers, respectively, detected in the plasma obtained from 23 SARS-CoV-2-convalescent humans. As expected, 0/4 RMs immunized with empty B/HPIV3 control had serum anti-S or anti-RBD IgM, IgA or IgG antibodies detectable at any time.

We also evaluated the kinetics and breadth of the serum neutralizing antibody response to vaccine-matched SARS-CoV-2 and to 4 VoCs (B.1.1.7/Alpha, B.1.351/Beta, B.1.617.2/Delta, and B.1.1.529/Omicron BA.1 sublineage) using a lentivirus-based pseudotype neutralization assay (Corbett et al., 2020a) ( Figure 3B ). The sera efficiently and similarly neutralized lentivirus pseudotyped with vaccine-matched Wuhan-1 S protein (IC50 on day 28 between 2.7 and 3.5 log10) or with S from the Alpha lineage (IC50 between 3.0 and 3.5 log10). The sera also neutralized the Beta S-pseudotyped lentivirus, although the titer was reduced compared to the vaccine match (IC50 between 1.6 and 2.4 log10). Day 14 sera from all 4 RMs efficiently neutralized the Delta S-pseudotyped lentivirus; titers further increased, but, on day 28, were about 5-fold reduced compared to the vaccine match (IC50 between 2.4 and 2.8 log10). A low neutralizing activity against Omicron BA.1 was detected in day 28 sera from 3 of 4 RMs (IC50 between 1.4 and 1.8 log10) that was 59-fold reduced compared to the vaccine match.

The serum neutralizing antibody titers were also assessed by a live-virus SARS-CoV-2 neutralization assay using the vaccine-matched WA1/2020 isolate or an isolate of the Alpha or Beta lineages ( Figure 3C ). Results were overall comparable with those of the pseudotyped lentivirus neutralization assays, although, as expected, the sensitivity and the dynamic range of the live virus neutralization assays were lower than those of the pseudotype neutralization assays.

As expected, neutralizing antibodies against the various SARS-CoV-2 lineages were undetectable in sera from B/HPIV3-control immunized RMs by pseudotype or SARS-CoV-2 neutralization assay; additionally, all 8 RMs developed neutralizing serum antibodies against the HPIV3 vector (PRNT60 titers between 1.6 and 2.4 log10, Figure 3D ).

SARS-CoV-2 S-specific CD4 + and CD8 + T-cell responses were evaluated using peripheral blood mononuclear cells (PBMCs) and cells recovered from the LA by BAL (see Figure S4 for the gating strategy) at the indicated time points following immunization with B/HPIV3/S-6P (Figs. 4 and 5, figs. S5 and S6) and challenge with SARS-CoV-2 ( Figure 4 , Figure S7 ). SARS-CoV-2 S and N-specific CD4 + and CD8 + T-cells were identified as IFNg + /TNFa + double-positive cells after stimulation with pools of overlapping 15-mer peptides covering the entire length of these proteins. S-specific CD4 + T-cells were present in the blood of all B/HPIV3/S-6P-immunized RMs by day 9 pi ( Figure 4A , left panels; kinetics are shown in Figure 4B ); frequencies peaked on day 9 (2 RMs) or day 14 pi (2 RMs; average peak % of S-specific CD4 + T-cells irrespective of the peak day of 0.6%), and then steadily declined until day 28 pi. S-specific CD8 + T-cells were also detectable in the blood of B/HPIV3/S-6P-immunized RMs on days 9 pi ( Figure 4A , right panels, and Figure 4C ), and their frequencies peaked on day 14 pi in 3 of 4 RMs ( Figure   4C ; average peak % of S-specific CD8 + T-cells irrespective of the peak day of 1.1%).

In the LA of B/HPIV3/S-6P-immunized animals, S-specific CD4 + and CD8 + T-cells were abundant by day 9 pi or 14 ( Figure 4 , D to F). Remarkably, the average peak percentage of Sspecific CD4 + T-cells recovered from BAL irrespective of day pi reached 14.3% ( Figure 4E ). In 3 of 4 animals, their frequency declined between day 14 and 28 pi. S-specific CD8 + T-cells in BAL also peaked on day 14 pi in 3 of 4 RMs ( Figure 4F ; average peak % of S-specific IFNg + /TNFa + CD8 + T-cells irrespective of the peak day of 11.1%). No S-specific CD4 + or CD8 + T-cells were detected in the blood or BAL of RMs immunized with B/HPIV3 ( Figure 4 , A to F).

Lastly, stimulation with SARS-CoV-2 N peptides of CD4 + or CD8 + T-cells isolated from BAL, which was included as negative control, did not reveal IFNg + /TNFa + positive cells above the background present in unstimulated cells ( Figure 4D ).

On day 9 pi, close to 100% of the S-specific CD4 + and CD8 + T-cells in the blood ( 

). An additional subset of presumably tissue-resident S-specific CD4 + and CD8 + T-cells was identified as CD69 -CD103 + and has been previously detected in SARS-CoV-2-infected RMs (Nelson et al., 2022) . Circulating CD69 -CD103 -S-specific CD4 + and CD8 + Tcells were detectable in BAL on day 9 pi and were prominent until day 14, representing about 60% of the S-specific T-cells on this day ( Following antigen stimulation, the S-specific circulating and tissue-resident CD4 + or CD8 + T-cells recovered from the airways on days 9, 14, or 28 were phenotypically comparable with respect to strong expression of CD107ab and granzyme B ( Figure S6 ), suggesting that all S specific CD69/CD103 subsets present in the airways were highly functional.

To assess protective efficacy of intranasal/intratracheal immunization with B/HPIV3/S-6P, we challenged RMs from both groups intranasally and intratracheally with 5.8 log10 TCID50

of SARS-CoV-2 WA1/2020 on day 30 or 31 after immunization ( Figure S1 ). NS and BAL specimens were collected before challenge and on days 2, 4, and 6 post-challenge (pc). Viral Quantification of sgE and gN RNA was also performed for lung tissues from different areas obtained on day 6 pc ( Figure 6C ). In all 4 B/HPIV3 empty-vector immunized RMs, gN RNA and sgE mRNA were detected, mostly in the right upper and right lower lobes of the lungs.

Neither gN RNA nor sgE mRNA were detected in the lungs of the B/HPIV3/S-6P-immunized RMs, confirming robust protection against SARS-CoV-2 infection induced by B/HPIV3/S-6P.

Furthermore, no active SARS-CoV-2 replication was detected from rectal swab samples ( Figure   S8 ).

We finally assessed the CD4 + and CD8 + T-cell response in the blood (Figure 4 , B and C) and lower airways (Figure 4 , E and F) of immunized RMs, 4 days after challenge with SARS-CoV-2. In the blood, an increase of S-specific IFNg + /TNFa + CD4 + and CD8 + T-cells was detected in 3 and 1 of 4 B/HPIV3/S-6P-immunized RMs, respectively, that correlated well with the increased expression of Ki-67 by the S-specific CD4 + T-cells ( Figure S7C ). A modest increase of S-specific IFNg + /TNFa + CD4 + T-cells was also detected in 1 B/HPIV3-immunized RM ( Figure 4B ). However, in the lower airways, a decrease rather than an increase of the Sspecific IFNg + /TNFa + CD4 + and CD8 + T-cells was detected in the B/HPIV3/S6-P-immunized

RMs and no active T-cell proliferation was detected ( Figure S7D ).

SARS-CoV-2 vaccines for infants and young children are critically needed. Equally needed are topical SARS-CoV-2 vaccines that directly stimulate local respiratory tract immunity in addition to systemic immunity, which might effectively reduce infection and transmission. In the present study, we evaluated a chimeric B/HPIV3 virus as a live topical viral vector to express the SARS-CoV-2 S protein, stabilized in its prefusion form, expected to be non-functional for virus entry . We found that RMs were fully protected from SARS-CoV-2 challenge 1 month after immunization. No SARS-CoV-2 challenge virus replication was detectable in the UA or LA or in lung tissues of immunized RMs, suggesting sterilizing immunity under these experimental conditions. However, the duration of immunity is currently unknown and will be evaluated in a separate study. Following challenge, we detected a recall response of S-specific CD4 + T-cells in the blood of B/HPIV3/S-6P-immunized RMs, but not in the LA. The reasons for the apparent absence of a detectable recall response are still unknown. It is possible that the S-specific T-cells were not present in the airways and inaccessible through BAL, or that the BAL cell collection on day 4 pc was too early to detect any recall response, or that in absence of challenge virus replication, a recall response was not triggered in the LA. Later time points following challenge will be included in future studies. Even though we did not evaluate protection against HPIV3 challenge in this study, we showed that B/HPIV3 induced strong serum HPIV3 neutralizing antibodies, comparable to levels seen in previous studies (Schmidt et al., 2001) ; thus, B/HPIV3/S-6P represents a promising candidate for intranasal immunization against both SARS-CoV-2 and HPIV3, an important pediatric cause of lower respiratory tract disease.

In conclusion, a single topical immunization with B/HPIV3/S-6P was highly immunogenic and protective against SARS-CoV-2 in RMs. Our data support the further development of this vaccine candidate for potential use as a stand-alone vaccine and/or in a prime/boost combination with other SARS-CoV-2 vaccines for infants and young children. The B/HPIV3/S-6P vaccine candidate will be evaluated in a Phase 1 study.

Detailed methods are provided in the online version of this paper and include the following: HN; red) . The full-length SARS-CoV-2 S ORF (codons 1-1,273) was codon-optimized and inserted as additional gene (orange) between the N and P ORFs. The S sequence includes 6 stabilizing proline substitutions (S-6P) and RRAR-to-GSAS substitutions to ablate the S1/S2 cleavage site. Each gene begins and ends with PIV3 gene-start and gene-end transcription signals (light and dark bars). (B and C) Replication of B/HPIV3/S-6P and B/HPIV3 in upper (B) and lower (C) airways of rhesus macaques (RM). Two groups of 4 RMs were immunized intranasally and intratracheally with 6.3 log10 PFU of B/HPIV3/S-6P (blue) or B/HPIV3 (green). Nasopharyngeal swabs and tracheal lavages were performed daily and every other day, respectively, on days 0 to 12 post-immunization (pi). Vaccine virus titers were determined by immunoplaque assay (Materials and Methods); expressed as log10 PFU/ml [Limit of detection: 0.7 log10 PFU/mL for nasopharyngeal swabs; 0.7 log10 PFU/mL for tracheal lavages (dotted line)]. Each RM is indicated by a symbol; lines represent medians (*P<0.05, **P<0.01, ****P<0.0001; two-way ANOVA, Sidak multiple comparison test).

Rhesus macaques (n=4 per group) were immunized with B/HPIV3/S-6P or B/HPIV3 (control) by the intranasal/intratracheal route ( Figure S1 ). To determine the mucosal antibody response in the upper airways, nasal washes (NW) were performed before immunization and on days 14, 21, and 28. To analyze the antibody response in the lower airways, bronchoalveolar lavages (BAL) were collected before immunization and on days 9, 21, and 28 pi. (A and B) S-and receptor binding domain (RBD)-specific mucosal IgA and IgG titers on indicated days post-immunization (pi) in the upper (A) and lower (B) airways, determined by time-resolved dissociation-enhance lanthanide fluorescence (DELFIA-TRF) immunoassay. Endpoint titers are expressed in log10 for mucosal IgA and IgG to a secreted prefusion-stabilized form (aa 1-1,208; S-2P) of the S protein (left panels) or to a fragment of the S protein (aa 328-531) containing SARS-CoV-2 RBD (right panels). The limit of detection is 1.6 log10 (dotted line). B/HPIV3/S-6P-immunized RMs are shown in blue, while B/HPIV3-immunized RMs are in green, with each RM represented by a symbol. *P<0.05 (two-way ANOVA, Sidak multiple comparison test). 

Requests for resources, reagents and further information regarding this manuscript should be addressed and fulfilled by the lead contact, Ursula Buchholz (ubuchholz@niaid.nih.gov) .

Plasmids and viruses newly generated in this study are available under material transfer request upon request to the lead contact. All data are included in the manuscript.

Baby hamster kidney cells expressing T7 RNA polymerase (BSR T7/5) were grown in Glasgow minimum essential medium (MEM) (Thermo Fisher Scientific) with 10% Fetal bovine serum (FBS), 1% L-glutamine (Thermo Fisher Scientific), and 2% MEM Amino Acids (Thermo Fisher Scientific). African green monkey kidney Vero (ATCC CCL-81) and Vero E6 (ATCC CRL-1586) cells were cultured in Dulbecco's MEM with GlutaMAX (Thermo Fisher Scientific) with 5% FBS and 1% L-glutamine. Vero E6 cells that express high levels of ACE2 Ren et al., 2006) and Vero E6 cells that stably express human TMPRSS2 were used to expand SARS-CoV-2.

The SARS-CoV-2 USA-WA1/2020 isolate (lineage A; GenBank MN985325; GISAID: EPI_ISL_404895; obtained from Dr. Natalie Thornburg et al., Centers for Disease Control and Prevention (CDC)) was passaged on Vero E6 cells. We used WA1/2020 as challenge virus in this study because its S sequence is homologous to the sequence from which S-6P was derived. The USA/CA_CDC_5574/2020 isolate [lineage B.1.1.7 (Alpha); GISAID: EPI_ISL_751801; obtained from CDC] and the USA/MD-HP01542/2021 isolate [lineage B.1.351 (Beta); GISAID: EPI_ISL_890360; obtained from Dr. Andrew Pekosz, Johns Hopkins University] were grown on TMPRSS2-expressing Vero E6 cells . The SARS-CoV-2 stocks were titrated in Vero E6 cells by determination of the 50% tissue culture infectious dose (TCID50) (Subbarao et al., 2004) . All experiments with SARS-CoV-2 were performed in biosafety level 3 (BSL3) containment laboratories approved by the USDA and CDC.

The B/HPIV3/S-6P vaccine candidate is an improved derivative of B/HPIV3/S-2P . To generate the B/HPIV3/S-2P cDNA , the ORF encoding the fulllength 1,273 aa SARS-CoV-2 S protein from the first available sequence (GenBank MN908947, Wuhan-Hu-1; amino acid sequence of the S protein identical to that of WA1/2020) was codonoptimized for human expression and synthesized commercially (BioBasic). Two proline substitutions (aa positions 986 and 987) and four aa substitutions (RRAR to GSAS, aa 682-685) that stabilize S in the prefusion conformation and ablate the furin cleavage site between S1 and S2 were introduced to generate the S-2P cDNA . This S-2P ORF was then inserted into a cDNA clone encoding the B/HPIV3 antigenome between the N and P ORFs to create the B/HPIV3/S-2P cDNA . To create the B/HPIV3/S-6P cDNA, the B/HPIV3/S-2P cDNA was modified to introduce 4 additional proline substitutions in the S ORF (aa position 817, 892, 899, and 942 for a total of 6 proline substitutions). The 4 additional proline substitutions had been shown to confer increased stability to a soluble version of the prefusion-stabilized S protein . The B/HPIV3/S-6P cDNA was used to transfect BHK21 cells (clone BSR T7/5, stably expressing T7 RNA polymerase (Buchholz et al., 1999) ), together with helper plasmids encoding the N, P and L proteins (Buchholz et al., 2004; Liu et al., 2021) , to produce the B/HPIV3/S-6P recombinant virus. The empty vector control virus B/HPIV3 was recovered in parallel. Virus stocks were grown in Vero cells, and viral genomes of recovered viruses were completely sequenced by Sanger sequencing using overlapping RT-PCR fragments, confirming the absence of any adventitious mutations.

All animal studies were approved by the NIAID Animal Care and Use Committee. The timeline of the experiment and sampling is summarized in Figure S1 . Eight juvenile to young adult male Indian-origin rhesus macaques (Macaca mulatta), confirmed to be seronegative for HPIV3 and SARS-CoV-2, were immunized intranasally (0.5 ml per nostril) and intratracheally (1 ml) with a total does of 6.3 log10 plaque-forming units (PFU) of B/HPIV3/S-6P or the empty vector control B/HPIV3. Animals were observed daily from day -3 until the end of the study. Each time they were sedated, animals were weighed, their rectal temperature was taken, as well as the pulse in beats per minute and the respiratory rate in breaths per minute. In addition, the blood oxygen levels were determined by pulse oximetry.

Blood for analysis of serum antibodies and peripheral blood mononuclear cells (PBMC) was collected on days -3, 4, 9, 14, 21, and 28 pi. Nasopharyngeal swabs (NS) for vaccine virus quantification in the upper airways (UA) were performed daily from day -3 to day 10 pi and on days 12 and 14 pi using cotton-tipped applicators. Swabs were placed in 2 ml Leibovitz (L-15) medium with 1x sucrose phosphate (SP) used as stabilizer, and vortexed for 10 seconds. Aliquots were then snap frozen in dry ice and stored at -80°C. Nasal washes (NWs) for analysis of mucosal IgA and IgG were performed using 1 ml of Lactated Ringer's solution per nostril (2 ml total) on days -3, 14, 21 and 28 pi and aliquots were snap frozen in dry ice and stored at -80°C until further analysis. Tracheal lavages (TL) for virus quantification were done every other day from day 2 to 8 pi and on day 12 pi using 3 ml PBS. The samples were mixed 1:1 with L-15 medium containing 2x SP and aliquots were snap frozen in dry ice and stored at -80°C for further analysis. Bronchoalveolar lavages (BALs) for analysis of mucosal IgA and IgG and airway immune cells from the lower airways (LA) were done on days -3, 9, 14 and 28 pi using 30 ml PBS (3 times 10 ml). For analysis of mononuclear cells, BAL was filtered through a 100 µm filter, and centrifuged at 544 x g for 15 min at 4°C. The cell pellet was resuspended at 2x10 7 cells/ml in X-VIVO 15 media supplemented with 10% FBS for subsequent analysis. The cellfree BAL was aliquoted, snap frozen in dry ice and stored at -80°C for further analysis. Rectal swabs were done on day -3 and then every other day from day 2 to 14 following the same procedure as NS.

Four weeks after immunization, animals were transferred to BSL3. On day 30 or 31 pi, animals were challenged intranasally and intratracheally on with 10 5.8 TCID50 of SARS-CoV-2, USA-WA1/2020, that was entirely sequenced and free of any prominent adventitious mutations.

RFJ, NLG Data analysis and visualization

Recalling the Future: Immunological Memory Toward Unpredictable Influenza Viruses. Front Immunol 10

Phase 1 study of the safety and immunogenicity of a live, attenuated respiratory syncytial virus and parainfluenza virus type 3 vaccine in seronegative children

Contributions of the structural proteins of severe acute respiratory syndrome coronavirus to protective immunity

Generation of bovine respiratory syncytial virus (BRSV) from cDNA: BRSV NS2 is not essential for virus replication in tissue culture, and the human RSV leader region acts as a functional BRSV genome promoter

SARS-CoV-2 infection protects against rechallenge in rhesus macaques

Vaccine Protection Against the SARS-CoV-2 Omicron Variant in Macaques

Comparative tropism, replication kinetics, and cell damage profiling of SARS-CoV-2 and SARS-CoV with implications for clinical manifestations, transmissibility, and laboratory studies of COVID-19: an observational study

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

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

Immune correlates of protection by mRNA-1273 vaccine against SARS-CoV-2 in nonhuman primates

mRNA-1273 protects against SARS-CoV-2 beta infection in nonhuman primates

Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR

Human parainfluenza virus circulation

T cell immunity to SARS-CoV-2 following natural infection and vaccination

Outcomes of SARS-CoV-2-Positive Youths Tested in Emergency Departments: The Global PERN-COVID-19 Study

COVID-19 vaccines for children

Recovery from Acute SARS-CoV-2 Infection and Development of Anamnestic Immune Responses in T Cell-Depleted Rhesus Macaques

Parainfluenza Virus Types 1-3 Infections Among Children and Adults Hospitalized With Community-acquired Pneumonia

Structure-based design of prefusion-stabilized SARS-CoV-2 spikes

T Cell Memory: Understanding COVID-19

A SARS-CoV-2 ferritin nanoparticle vaccine elicits protective immune responses in nonhuman primates

Evaluation of two chimeric bovine-human parainfluenza virus type 3 vaccines in infants and young children

T cell responses to SARS-CoV-2 spike cross-recognize Omicron

Cutting edge: Antigen is not required for the activation and maintenance of virus-specific memory CD8+ T cells in the lung airways

Chimeric bovine/human parainfluenza virus type 3 expressing respiratory syncytial virus (RSV) F glycoprotein: effect of insert position on expression, replication, immunogenicity, stability, and protection against RSV infection

A single intranasal dose of a live-attenuated parainfluenza virus-vectored SARS-CoV-2 vaccine is protective in hamsters

Correlates of protection against SARS-CoV-2 in rhesus macaques

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

Mild SARS-CoV-2 infection in rhesus macaques is associated with viral control prior to antigen-specific T cell responses in tissues

Subacute SARS-CoV-2 replication can be controlled in the absence of CD8+ T cells in cynomolgus macaques

Intranasal priming induces local lung-resident B cell populations that secrete protective mucosal antiviral IgA

SARS-CoV-2 infection generates tissuelocalized immunological memory in humans

Epidemiologic trends and characteristics of SARS-CoV-2 infections among children in the United States

Analysis of ACE2 in polarized epithelial

Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein

Enhanced SARS-CoV-2 neutralization by dimeric IgA

Virological assessment of hospitalized patients with COVID-2019

Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation

Tissue resident memory T cells in the respiratory tract

2020) 5'-ATATTGCAGCAGTACGCACACA-3

2020) (pCMV DR8.2), and human transmembrane protease serine 2 (TMPRSS2) at a ratio of 1:20:20:0.3 into HEK293T/17 cells (ATCC) with transfection reagent LiFect293™. The pseudoviruses were harvested at 72 h post transfection. The supernatants were collected after centrifugation at 478 x g for 10 minutes to remove cell debris, then filtered through a 0.45 mm filter, aliquoted and titrated before neutralization assay. For the antibody neutralization assay, 6-point, 5-fold dilution series were

Fifty µl antibody dilution were mixed with 50 µl of diluted pseudoviruses in the 96-well plate and incubated for 30 min at 37°C. Ten thousand ACE2-expressing 293T-cells (293T-hACE2.MF stable cell line cells) were added in a final volume of 200 µl. Seventy-two h later, after carefully removing all the supernatants, cells were lysed with Bright-Glo™ luciferase assay substrate (Promega), and luciferase activity (relative light units, RLU) was measured. Percent neutralization was normalized relative to uninfected cells as 100% neutralization and cells infected with only pseudoviruses as 0% neutralization

Single cell suspensions of PBMCs that had been rested overnight or freshly collected BAL cells were plated at 2x10 7 cells/ml in 200 ml in 96 well plates with X-VIVO 15 media, with 10% FBS, 1000x Brefeldin (Thermo Fisher Cat#00-4506-51) and 1000x Monensin (Thermo Fisher Cat#00-4505-51), CD107a APC 1:50, CD107b APC 1:50, and the indicated peptide pools at 1mg/ml. Replica wells were not stimulated. Spike peptide pools consisted of Peptivator SARS-CoV-2 Prot_S1 (Miltenyi Cat#130-127-048), Peptivator SARS-CoV-2 Prot_S+ (Miltenyi Cat# 130-127-312), and Peptivator SARS-CoV-2 Prot_S (Miltenyi Cat#130-127-953) covering the whole spike protein. Nucleocapsid peptide pool consisted of Peptivator SARS-CoV-2 Prot_N (Miltenyi Cat# 130-126-699). Cells were stimulated for 6 h at 37°C with 5% CO2. After stimulation, cells were centrifuged at 544 x g for 5 min at 4°C, and further processed by surface staining. Cells were resuspended in 50 ml surface stain antibodies diluted in PBS with 1% FBS and

Cells were resuspended in 50 ml intracellular stains diluted in permeabilization buffer, and stained for 30 min at 4°C. The antibodies used for extracellular and intracellular staining were: CD69 (FITC, clone FN50, Biolegend), granzyme B (BV421, clone GB11, BD Biosciences), CD8a (eFluor 506, clone RPA-T8

We thank Alicia Wojcik and Amanda Havenner and the staff of the National Institute of Allergy and Infectious Diseases (NIAID) Comparative Medicine Branch for animal study support, Jeffrey I. Cohen for providing plasma from SARS-CoV-2 convalescent individuals, and Peter L.Collins for helpful discussions and for comments on the manuscript.

 N/A B/HPIV3/S-2P N/A B/HPIV3/S-6P This paper N/A BPIV3 N, P and L helper plasmids Schmidt et al., 2000) N/A SARS-CoV-2 S GenBank MN908947 SARS-CoV-2 S-2P N/A SARS-CoV-2 S RBD N/A pHR' CMV Luc (Corbett et al., 2020b) N/A pCMV DR8.2 (Corbett et al., 2020b) N/A TMPRSS2 (Corbett et al., BioTek https://www.biotek.com/products/ software-robotics-software/gen5microplate-reader-and-imagersoftware/ Sample collections were done following the same procedures as during the immunization phase.Briefly, blood was collected before challenge and on day 6 post-challenge (pc). NS were performed every other day from day 0 to day 6 pc. NWs were done on day 6 pc, BAL on days 2, 4 and 6 pc, and rectal swabs on days 0, 2, 4 and 6 pc. Animals were necropsied on day 6 pc, and tissues were collected. 6 samples per animal from individual lung lobes were collected, and snap frozen in dry ice for further analysis.

Titers of B/HPIV3 and B/HPIV3/S-6P from NS and TLs were determined by dualstaining immunoplaque assay . Briefly, Vero cell monolayers in 24-well plates were infected in duplicate with 10-fold serially diluted samples. Infected monolayers were overlaid with culture medium containing 0.8% methylcellulose, and incubated at 32°C for 6 days, fixed with 80% methanol, and immunostained with a rabbit hyperimmune serum raised against purified HPIV3 virions to detect B/HPIV3 antigens, and a goat hyperimmune serum to the secreted SARS-CoV-2 S to detect co-expression of the S protein, followed by infrared-dye conjugated donkey anti-rabbit IRDye680 IgG and donkey anti-goat IRDye800 IgG secondary antibodies (LiCor). Plates were scanned with the Odyssey infrared imaging system (LiCor). Fluorescent staining for PIV3 proteins and SARS-CoV-2 S was visualized in green and red, respectively, providing for yellow plaque staining when merged.

Levels of anti-SARS-CoV-2 S antibodies elicited by B/HPIV3/S-6P were determined by DELFIA-TRF (Perkin Elmer) from NW or BAL following the supplier's protocol and from serum samples by ELISA using the recombinantly-expressed secreted version of S-2P , or a fragment (aa 328-531) containing the receptor binding domain (RBD) of the SARS-CoV-2 S protein (Walls et al., 2020) . The secondary antibodies used in both assays were goat anti-monkey IgG(H+L) horseradish peroxidase (HRP) (Thermo Fisher, Cat #PA1-84631), goat anti-monkey IgA (alpha chain)-biotin (Alpha Diagnostic International, Cat #70049), and goat anti-monkey IgM-biotin (Brookwood Biomedical, Cat#1152).HPIV3-specific neutralizing antibody titers were measured by a 60% plaque reduction neutralization test (PRNT 60 ) . The serum neutralizing antibody assays using live SARS-CoV-2 virus was performed in a BSL3 laboratory. Heat-inactivated sera were 2-fold serially diluted in Opti-MEM and mixed with an equal volume of SARS-CoV-2 (100 TCID50) and incubated at 37°C for 1 h. Mixtures were added to quadruplicate wells of Vero E6 cells in 96-well plates and incubated for four days. The 50% neutralizing dose (ND50) was defined as the highest dilution of serum that completely prevented cytopathic effect in 50% of the wells and was expressed as a log10 reciprocal value .

The SARS-CoV-2 pseudovirus neutralization assays were performed as previously reported (Corbett et al., 2020b) . Briefly, the single-round luciferase-expressing pseudoviruses were generated by co-transfection of plasmids encoding SARS-CoV-2 S of isolate Wuhan-Hu-1, GenBank accession number MN908947.3, or of lineages B.1.351/Beta, B.1.1.7/Alpha, B.1.617.2/Delta, B.1.1.529/Omicron), luciferase reporter (pHR' CMV Luc), lentivirus backbone

Hundred ml each of NS and BAL fluid collected on day 2, 4 and 6 pc and rectal swabs collected on day 6 pc were inactivated in a BSL3 laboratory using 400 ml buffer AVL (Qiagen) and 500 ml ethanol, and RNA was extracted using the QIAamp Viral RNA Mini Kit (Qiagen) according to the manufacturer's protocol. To extract total RNA from lung homogenates harvested on day 6 pc, 300 ml of each lung homogenate (at a concentration of 0.1 g of tissue/ml) was mixed with 900 ml TRIzol LS (Thermo Fisher) using Phasemaker Tubes (Thermo Fisher) and RNA was extracted using the PureLink RNA Mini Kit (Thermo Fisher) following the manufacturer's instructions. Then, the SARS-CoV-2 genomic N RNA and subgenomic E mRNA were quantified in triplicate using the TaqMan RNA-to-Ct 1-Step Kit (Thermo Fisher) using previously reported TaqMan primers/probes Corman et al., 2020; Wolfel et al., 2020) on the QuantStudio 7 Pro (ThermoFisher). Standard curves were generated using serially diluted pcDNA3.1 plasmids encoding gN, gE, or sgE sequences. The limit of detection was 2.57 log10 copies per ml of NS, BAL fluid, or rectal swabs and 3.32 log10 copies per g of lung tissue.

Data sets were assessed for significance using two-way ANOVA with Sidak's multiple comparison test using Prism 8 (GraphPad Software). Data were only considered significant at P < 0.05.