key: cord-0946532-ks5k9ti3
authors: Carmen, Joshua M.; Shrivastava, Shikha; Lu, Zhongyan; Anderson, Alexander; Morrison, Elaine B.; Sankhala, Rajeshwer S.; Chen, Wei-Hung; Chang, William C.; Bolton, Jessica S.; Matyas, Gary R.; Michael, Nelson L.; Joyce, M. Gordon; Modjarrad, Kayvon; Currier, Jeffrey R.; Bergmann-Leitner, Elke; Malloy, Allison M.W.; Rao, Mangala
title: A spike-ferritin nanoparticle vaccine induces robust innate immune activity and drives polyfunctional SARS-CoV-2-specific T cells
date: 2021-04-28
journal: bioRxiv
DOI: 10.1101/2021.04.28.441763
sha: a4c8f44acb91c5c7acf722c59f372bd6309d379b
doc_id: 946532
cord_uid: ks5k9ti3

Potent cellular responses to viral infections are pivotal for long -lived protection. Evidence is growing that these responses are critical in SARS -CoV-2 immunity. Assessment of a SARS -CoV-2 spike ferritin nanoparticle (SpFN) immunogen paired with two distinct adjuvants, Alhydrogel® (AH) or Army Liposome Formulation containing QS-21 (ALFQ) demonstrated unique vaccine evoked immune signatures. SpFN+ALFQ enhanced recruitment of highly activated classical and non -classical antigen presenting cells (APCs) to the vaccine-draining lymph nodes of mice. The multifaceted APC response of SpFN+ALFQ vaccinated mice was associated with an increased frequency of polyfunctional spike -specific T cells with a bias towards TH1 responses and more robust SARS-CoV-2 spike-specific recall response. In addition, SpFN+ALFQ induced Kb spike(539-546)-specific memory CD8+ T cells with effective cytolytic function and distribution to the lungs. This epitope is also present in SARS-CoV, thus suggesting that generation of cross-reactive T cells may provide protection against other coronavirus strains. Our study reveals that a nanoparticle vaccine, combined with a potent adjuvant, generates effective SARS-CoV-2 specific innate and adaptive immune T cell responses that are key components to inducing long-lived immunity. One Sentence Summary SpFN vaccine generates multifactorial cellular immune responses.

Coronaviruses (CoV) are positive-sense, single stranded RNA viruses that cause varying disease pathologies ranging from common cold symptoms to acute respiratory distress syndrome, as well as, gastrointestinal symptoms (1- 3) . SARS-CoV-2, the causative agent of COVID-19, represents the seventh CoV to be isolated from humans, and is the third to cause severe disease (4) . The rapid and unparalleled spread of SARS-CoV-2 into a global pandemic has driven an urgent need for rapidly deployable and scalable vaccine platforms. A myriad of vaccine platforms and approaches have been adopted and developed by government, industry, academic, and nongovernmental organizations. Two messenger RNA-based vaccines and an adenovirus vector-based vaccine have been approved for emergency use in the United States (5) (https://www.raps.org/news-and-articles/newsarticles/2020/3/covid-19-vaccine-tracker). While messenger RNA-based vaccines (6, 7) and recombinant adenovirus vectored vaccines (8) (9) (10) have demonstrated potent efficacy and unprecedented deployment speed, there exists a need for precision vaccine design that may offer improved efficacy to different demographic groups, induce durable immune responses, and provide a multivalent strategy to protect against multiple CoV species and the recent emergence of rapidly evolving immunological variants (11) (12) (13) .

We have recently developed a SARS-CoV-2 sub-unit vaccine based on the ferritin nanoparticle platform (14) that displays a pre-fusion stabilized spike on its surface (15) and has the potential to fill these current gaps.

The stabilized prefusion-spike protein of the Wuhan-Hu-1 strain of SARS-CoV-2 was genetically linked to form a ferritin-fusion recombinant protein, which naturally forms a Spike-Ferritin nanoparticle (SpFN). Ferritin is a naturally occurring, ubiquitous, iron-carrying protein that self-oligomerizes into a 24-unit spherical particle and is currently being evaluated as a vaccine platform for influenza in two phase 1 clinical trials (NCT03186781, NCT03814720) with two further trials in the recruitment phase for Epstein Barr virus (NCT04645147) and

Influenza H10 (NCT04579250).

The spike protein was modified in the following ways to generate a stable spike trimer formation on the ferritin molecule. Two proline residues (K986P, V987P) were introduced to the spike ectodomain, the furin cleavage site was mutated (RRAS to GSAS) (16) , and the coil-coil interactions were stabilized by mutating the heptad repeat between hinge 1 and 2 (residues 1140-1161). In addition, a short linker to the ferritin molecule was used to exploit the natural three-fold axis, for display of eight spikes. SpFN was then formulated with either of the two distinct adjuvants to evaluate innate and adaptive T cell responses induced by each of the adjuvants. Our prior experience with Army Liposome Formulation containing the saponin, QS-21(ALFQ) made it a top candidate to compare with one of the most commonly used adjuvants, Aluminum Hydroxide gel (Alhydrogel ® ) (AH). ALFQ contains two immunostimulants, synthetic MPLA (3D-PHAD ® ) and QS-21 (17) , the same two immunostimulants that are also present in AS01B, the adjuvant in the highly efficacious licensed herpes-zoster vaccine, Shingrix TM (18) . However, the liposomes in ALFQ fundamentally differ from those in AS01B, both in the phospholipid type and cholesterol content in addition to differences in the amount of 3D-PHAD ® and QS-21. Previously, it has been shown in several different models that ALFQ generates well-balanced TH1/TH2 immunity and protective efficacy (19) (20) (21) .

In this study, we show for the first time the effect of adjuvant design on antigen presenting cell (APC) recruitment to the lymph nodes draining the vaccination site, and its impact on a SARS-CoV-2 vaccine platform.

We demonstrate that SpFN formulated with ALFQ (SpFN+ALFQ) compared to SpFN formulated with AH (SpFN+AH) significantly increased recruitment and activation of classical and non-classical APCs. Intriguingly, this was associated with a potent TH1-biased cellular response and highly functional spike-specific memory T cells. The APC response to SpFN+ALFQ was characterized by conventional type 1 and type 2 dendritic cells (cDC1 and cDC2) with upregulated costimulatory molecules, in contrast to SpFN+AH. The effective APC response induced by SpFN+ALFQ correlated with differentiation of spike-specific CD4 + T cells expressing markers of TH1 phenotype. Strikingly, vaccination with SpFN+ALFQ resulted in spike-specific CD8 + T cells that established a memory pool. We identified eleven SARS-CoV-2 T cell epitopes in C57BL/6 mice vaccinated with SpFN+ALFQ that mapped to the spike protein. Several of these epitopes are corroborated by recently described findings using an overlapping SARS-CoV-2 peptide library (22) following SARS-CoV-2 infection in H2 b restricted mice. Using an MHC class I tetramer, we identified murine K b restricted SARS-CoV-2 specific memory CD8 + T cells recognizing an eight amino acid sequence (amino acids 539-546; VNFNFNGL) of the SARS-CoV-2 spike protein that is conserved in the SARS-CoV spike protein. The K b spike(539-546)-specific memory CD8 + T cells generated by vaccination with SpFN+ALFQ exhibited significantly more cytotoxic activity compared to those identified in the SpFN+AH vaccinated mice. Together these finding demonstrate a novel vaccine platform for SAR-CoV-2 that leverages the innate immune response to induce potent memory-specific antiviral T cells.

Adjuvants are immunostimulants that activate the innate immune system and facilitate vaccine antigen presentation (23, 24) . To define the ability of distinct adjuvants to engage the innate immune response, we compared our novel vaccine formulations, SpFN+ALFQ with SpFN+AH. Mice were vaccinated at the specified time points and innate immune cells were analyzed from the inguinal and popliteal lymph nodes that drain the vaccinated muscle (dLNs) at days 3 and 5 post vaccination (Fig. 1A) . A multiparameter spectral flow cytometry panel (table S1) was used to discriminate APC subsets. T-Distributed Stochastic Neighbor Embedding (tSNE) was used to visualize the high-dimensional datasets (Fig. 1B) . The flow cytometric gating strategy is shown in fig. S1 , A. Significant differences in the composition of APCs present in the dLNs were seen in mice vaccinated with SpFN+ALFQ compared to SpFN+AH (Fig. 1C, and fig. S1 , B and C). The overall number of APCs present in the dLNs was increased 3-7-fold in SpFN+ALFQ vaccinated mice compared to SpFN+AH vaccinated mice, as shown by the pie graphs in Fig. 1C . Interestingly, the number and diversity of APCs began to contract 5 days post vaccination in the SpFN+AH vaccinated mice to levels exhibited in naïve mice. In contrast, mice vaccinated with SpFN+ALFQ exhibited continued expansion of APCs in the dLNs at this time point, demonstrating a more sustained response to the adjuvant formulation, ALFQ (Fig. 1C) .

Furthermore, not only were the numbers of APCs increased after SpFN+ALFQ vaccination at days 3 and 5, but classical APCs, such as migratory DCs and lymph node resident APCs, predominated (Fig. 1D ). Migratory DCs, which include cDC1 and cDC2s, migrate from tissue sites to dLNs to present antigen and support T cell activation and differentiation and were increased in number in the SpFN+ALFQ vaccinated mice (Fig. 1D ). In addition, plasmacytoid DCs (pDCs), that produce and provide type I interferon for DC and T cell activation, were also increased in the dLNs after the priming vaccination with SpFN+ALFQ compared to SpFN+AH (Fig. 1D) .

Lymph node resident macrophages play a role in acquisition of antigens through pinocytosis and phagocytosis, and participate in immune regulation. Medullary sinus macrophages (MSM) defined as CD169 + CD11b + F4/80 + MHCII int CD11c lo and medullary cord macrophages (MCM) defined as CD169 -CD11b + F4/80 + MHCII int CD11c lo CD64 +/- (25, 26) were the most significantly divergent (p <0.0001) between the two vaccinated groups; higher numbers of MCM were measured in the SpFN+ALFQ vaccinated mice, whereas MSM numbers were significantly higher in the SpFN+AH vaccinated mice, (Fig. 1E ). CD64and CD64 + MCM were infrequent in numbers on both days 3 and 5 in the SpFN+AH vaccinated mice, while CD64 + MCM were present in higher numbers on both days 3 and 5 post-vaccination with SpFN+ALFQ. Interestingly, in the SpFN+ALFQ vaccinated mice, CD64 -MCM, which were infrequent in number on day 3 post-vaccination increased by ~10,000-fold on day 5 post-vaccination.

Commensurate with increased numbers of classical APCs, the functional activation of these cells was also enhanced by SpFN+ALFQ compared to SpFN+AH. In addition to antigen presentation, costimulatory molecules, such as CD80 and CD86, are expressed by activated DCs and engage T cells, inducing activation rather than anergy in response to ligation with their peptide-MHC complex (27, 28) . The percentage of cDC1 and cDC2 expressing CD80 and CD86 three days post vaccination was 3 times higher in the SpFN+ALFQ group compared to SpFN+AH ( Fig. 1F and fig. S1B ). We also measured the activation of macrophage subsets. While the expression of CD80 and CD40 were similarly low in CD64 -MCMs, the CD64 + MCMs from SpFN+ALFQ vaccinated mice significantly upregulated CD80 and CD40 by 3-fold at day 3, and 2-fold at day 5 post priming vaccination compared to the SpFN+AH group (Fig. 1G ). In addition, although the MSM numbers were higher with AH as the adjuvant, the expression of costimulatory molecules, CD80 and CD40, was increased in the SpFN+ALFQ vaccinated mice suggesting a higher activation status (Fig. 1G) .

Neutrophils, eosinophils, and monocytes have been shown to function as APCs or support antigen-specific T and B cell differentiation in humans and mice (29) (30) (31) . Increased numbers of eosinophils, neutrophils, and monocytes were observed in the dLNs of SpFN+ALFQ vaccinated mice at days 3 and 5 post vaccination (figs. S1, D and E). In addition, a notable proportion of CD11b + MHCII + monocyte-like cells were CD11c intermediate, indicative of monocyte-derived DCs (moDCs), and were further differentiated by expression of the chemokine receptor XCR1 (Fig. 1B; fig. 1SA ). The XCR1+moDCs were distinct from cDC1, as XCR1+moDCs were MHC-II low, CD11c low, and CD103 negative, and displayed two separate populations by t-SNE. Few XCR1+moDCs were detected in the dLNs of either vaccine group on day 3 post priming vaccination. However, by day 5 post vaccination this moDC subset had significantly increased in the SpFN+ALFQ group suggesting that these cells were involved in acquisition and presentation of the antigen ( fig. S1E ).

Next, we analyzed the recruitment of total and spike-specific T cells to the dLNs following vaccination with SpFN+ALFQ or SpFN+AH. T cells from the dLNs were analyzed at days 7 and 10 post priming vaccination for phenotype and cytokine expression, respectively. At day 7 post priming vaccination, the total T cells, as well as the CD4 + and CD8 + T cell subsets, in the dLNs of SpFN+ALFQ vaccinated mice were significantly higher than those in the SpFN+AH vaccinated mice ( Fig. 2A) . T cell memory differentiation potential was characterized by the expression of CD62L and CD44 ( fig. S2A ). In addition to the numeric differences of the T cell subsets, SpFN+ALFQ vaccination resulted in proportionally more effector memory-like (CD62L -CD44 + ) CD4 + and CD8 + T cells. In contrast, SpFN+AH vaccination induced more naïve T cells (CD62L + CD44 -) to the dLNs, although this subset was seen after vaccination in both groups ( fig. S2B ).

At day 10 post priming vaccination, SARS-CoV-2 spike-specific T cells were identified by intracellular cytokine staining (ICS) following stimulation with a spike peptide pool (S1) containing the epitopes in the receptor binding domain (RBD). Spike-specific CD4 + T cells were readily detected by expression of IL-2 and TNF- upon peptide stimulation in the dLNs of mice vaccinated with SpFN+ALFQ, and to a much lower extent (p<0.0001) in those of SpFN+AH vaccinated mice (Fig. 2B ). IL-2 and TNF- were co-expressed in a subset of the CD4 + T cells (Fig 2C) , which were primarily CD44 + CD62L -CCR7effector memory CD4 + T cells ( fig. S2C ).

The intracellular cytokine responses measured from the spike-specific T cells indicate that the CD4 + T cell responses were predominantly directed toward TH1 (TNF- and IL-2) rather than a TH2 profile. To further confirm our findings, multiplex cytokine analysis was performed with the sera of mice after priming vaccination on days 3, 5 and 10. The TH1 cytokines (IFN- and TNF-), as well as IL-10 were significantly higher in the peripheral blood of the SpFN+ALFQ vaccinated mice (Fig. 2D) indicating that the cytokine profile was skewed towards a TH1 profile, which was consistent with the measurement of intracellular cytokine expression in the SpFN+ALFQ vaccinated mice.

Early expansion of SARS-CoV-2 spike-specific CD4 + and CD8 + T cells following priming vaccination with

Maturation and activation of APCs, especially DCs, in the dLNs is a prerequisite for the induction of antigenspecific T cell responses and maintenance of long-term memory responses. Since SpFN+ALFQ triggered more robust APC activation along with significantly higher percentages of functional antigen-specific CD4 + T cells in the dLNs, we next sought to quantify and characterize the kinetics of vaccine-induced T cell expansion. We therefore evaluated the phenotype and functional CD4 + and CD8 + T cell responses in the spleen at days 5 and 10 post vaccination. Cells were stimulated with SARS-CoV-2 spike-specific peptides from the S1 and S2 peptide pools from JPT, followed by surface and ICS and measurement by flow cytometry. Although there were no significant differences in the frequency of total T cells, CD4 + and CD8 + T cells at day 5 and day 10 post vaccination ( fig. S3A ), we did observe significant differences in the SARS-CoV-2 spike-specific cytokine expressing CD4 + and CD8 + T cells. CD4 + T cells expressing IL-2 (p=0.008) and IFN-γ (p=0.03) were significantly higher in the ALFQ group compared to the AH group at day 10 ( post-vaccination (Fig. 3 , A to C). The frequency of cytokine-positive cells was generally higher in the CD8 + T cell population than the CD4 + T cell population at day 10 compared to day 5 post-vaccination (Fig. 3 , A to C).

To determine whether an intramuscular vaccination could prime SARS-CoV-2-specific memory T cells to home to tissue sites relevant to potential infection, the T cell response in the mediastinal lymph nodes were measured three weeks following a booster vaccination. The gating strategy for the identification of SARS-CoV-2 spikespecific CD4 + and CD8 + T cells is shown in fig. S2D . Spike-specific CD4 + and CD8 + T cells expressed IFN-, TNF-, IL-17A (Fig. 4A ) and IL-10 ( Fig. 4B ) in the SpFN+ALFQ vaccine group. The IFN- expressing CD4 + and CD8 + T cells exhibited activated effector memory markers (CD69 + CD44 + and CD62L -CD103 -; fig. S2 , E and F). Furthermore, IFN- and TNF- were dominant and co-expressed by both spike-specific CD4 + and CD8 + T cell in the SpFN+ALFQ vaccine group (Fig. 4C ).

The profile of pro-inflammatory cytokines was also compared in the spleen at 3 weeks following booster vaccination to track the differentiation of antiviral T cell responses. Splenocytes were stimulated with pooled SARS-CoV-2 spike peptides spanning the S1 and S2 subunits and the cytokine profiles were assessed. Assessing the two vaccination groups, IFN-γ, TNF-, IL-2 and IL-6 levels were significantly increased in SpFN+ALFQ vaccinated mice compared to mice that received SpFN+AH vaccination (Fig. 4D) . The IL-4 and IL-10 levels were also elevated in the SpFN+ALFQ vaccinated group, although the magnitude of increase was much lower in comparison to the TH1 cytokines (Fig. 4D ).

To assess the durability of T cell responses, we further characterized the T cell phenotypes and functional responses in the spleen at three weeks post booster vaccination. Upon evaluating the total T cell response, we observed an expansion of CD8 + T cells compared to CD4 + T cells as measured by a decrease in the total CD4 + (p=0.002) and an increase in the total CD8 + T cells (p=0.0006) in the SpFN+ALFQ compared to SpFN+AH vaccination group (Fig. 5, A and B) . To assess the frequency of cytokine secreting CD4 + and CD8 + T cells, splenocytes were stimulated ex-vivo with SARS-CoV-2 spike peptides spanning the S1 or S2 subunits. The flow gating strategy used to determine the frequency of cytokine secreting CD4 + and CD8 + T cells is represented in fig. S4 . Our data revealed that the vaccine boost dramatically increased the frequency of SARS-CoV-2 spike S1specific IL-2 (CD4 + : p= 0.004; CD8 + : p=0.0005, Fig. 5C ), IFN-γ (CD4 + : p= 0.005; CD8 + : p=0.0005, Fig. 5D) and TNF- (CD4 + : p= 0.0007; CD8 + : p=0.0005, Fig. 5E ) secreting CD4 + and CD8 + T cells, while no differences were observed in frequency of IL-4 secreting cells (data not shown). The booster dose of SpFN+ALFQ resulted in a further expansion of cytokine producing T cells above the priming response at day 5 and/or day 10. This suggested that the two-dose regimen of the SpFN+ALFQ vaccine formulation improved the generation and differentiation of SARS-CoV-2-specific T cell memory responses. Importantly, the S1 peptide pool containing the RBD and NTD portions of the spike protein produced significantly greater cytokine expression compared to the minimal responses generated by the S2 peptide pool (data not shown). Therefore, we focused further functional assays on this portion of spike that contains both conserved peptide residues and peptides unique to SARS-CoV-2.

Next, we sought to characterize the peptides of the spike protein targeted by the vaccine induced T cell response.

Our data demonstrated that the cytokine response to SARS-CoV-2 was dominated by IFN-γ, therefore an IFN-γspecific ELISpot assay was used to screen 52 peptide pools composed of 315 peptides as this methodology requires fewer cells and has a higher sensitivity than other measurement assays. Peptides comprised the full length of the spike protein, represented with its major sub-units and domains (S-1 domain containing the NTD, RBD, S-2 domain containing FP, HR1 and HR2) in Fig. 6A . Peptides were pooled in a matrix scheme to allow the highthroughput screening and identification of potential epitopes (Fig. 6B, and fig. S5 ). Most of the measured reactivity to the peptide pools was focused to the NTD and RBD within S-1 domain of the spike protein with the highest responses in pools 4 and 5, and 19 (Fig. 6C) . The results revealed one epitope, VNFNFNGL (aa 539-546) that was not predicted by the NetMHCPan 4.1 EL prediction tool (iedb.org), but was previously reported (22) in mice transfected with human ACE2 mRNA and subsequent infected with SARS-CoV-2, as well as defined in SARS-CoV (32). This epitope is conserved between the two viruses. Comparing the results from the NetMHCPan 4.1EL results with the IFN-γ ELISpot results suggest that all epitopes are either H2K b or H2D b restricted and none of them represented CD4 + epitopes.

After establishing the strong effector functions of T cells after the prime-boost vaccination strategy, we then enumerated the frequency of antigen specific CD8 + T cells with MHC class I tetramer staining. Epitope mapping data revealed SARS-CoV-2 spike aa 539-546 (VNFNFNGL) as the most immunogenic epitope present in peptide pools 4, 5, and 19 ( Fig. 6 , B and C). MHC class I, K b , tetramers presenting the VNFNFNGL epitope (aa 539-546) along-with two other H2-K b restricted epitopes, GNYNYLYRL (aa 433-441) and YNYLYRLF (aa 435-443) from the spike protein, were used to identify antigen-specific CD8 + T cells. A H2-K d restricted tetramer, containing CYGVSPTKL (aa 365-373) from the spike protein served as a control (table S5) Cytokine polyfunctionality is of major importance for the successful efficacy of a vaccine (33) . Several studies have shown a strong correlation between the protection level and the induction of high frequencies of polyfunctional T cells co-expressing IFN-γ, TNF-, and IL-2 (33) . In order to determine if SpFN+ALFQ generated antigen-specific polyfunctional CD8 + T cells, we assessed their frequency post booster vaccination. Functionality was measured based on stimulation with the S1 peptide pool, as our prior data showed minimal responses to the S2 peptide pool by ICS.

SpFN+ALFQ induced a strikingly higher percentage of polyfunctional spike-specific CD8 + T cells coexpressing either 3 (IFN-γ, IL-2, and TNF-α) or 2 (IFN-γ, TNF-α) cytokines compared to the SpFN+AH (Fig.   7F ). Polyfunctional IFN-γ and TNF-α producing T cells dominated the response from the SpFN+ALFQ vaccinated animals, in contrast to the single cytokine secreting T cell response exhibited by the SpFN+AH group.

These data demonstrate the differences in the landscape of T cell responses induced by the two different vaccine formulations.

In order to determine if the induction of a more robust CD8 + T cell response in the SpFN+ALFQ vaccinated mice compared to SpFN+AH vaccinated mice translated into improved effector function, we conducted an invitro killing assay. Target cells were generated by pulsing naïve splenocytes with K b -spike(539-546)-peptide or left unpulsed and labeled CFSE high or low, respectively. The gating strategy for target CFSE-high and low and 

Prime-boost with SpFN+ALFQ generated effector memory IFN-γ producing CD4 + and CD8 + T cells from the mediastinal lymph nodes (Fig. S2 , E and F). Therefore, we wanted to determine the distribution of the cytotoxic spike(539-546)-specific CD8 + T cells at a similar memory time point in the lungs of mice vaccinated with SpFN+ALFQ and SpFN+AH. At 3 weeks following booster vaccination, perfused lungs were processed from both groups of vaccinated mice and stained with K b -spike(539-546) tetramer. Tetramer positive CD8 + T cells were detected from the lungs of mice in both adjuvant groups (Fig. 8D ). Similar to the results from splenocytes, the lungs from the SpFN+ALFQ group exhibited a higher percentage of K b -spike(539-546)-specific CD8 + T cells (9.3%) compared to lungs from the SpFN+AH group (3.1%).

In the present study, we assessed cell-mediated immunity induced by a novel immunogen, SpFN, combined with a potent adjuvant, ALFQ. This candidate vaccine induced robust recruitment of APCs and polyfunctional SARS-CoV-2 spike-specific T cells, and spike epitope-specific cytolytic memory CD8 + T cells.

The vaccine, SpFN+ALFQ is currently in a phase 1 clinical trial in the United States, sponsored by the U.S. Army (ClinicalTrials.gov Identifier: NCT04784767).

Both humoral and cellular responses have been shown to provide protection against respiratory viral pathogens, such as influenza and RSV (35, 36) . Moreover, cross-reactive T cells have been shown to protect against a heterologous influenza virus infection in the absence of virus-specific antibodies (37) . While vaccine evaluation traditionally involves assessment of neutralizing antibodies, currently, the full spectrum of correlates of protection for COVID-19 remain unknown. There are several reports that both CD4 + and CD8 + T cells specific for SARS-CoV-2 arise during an active infection and during the convalescence period (38) and may also provide protection. Resident memory T cells were shown to protect against SARS-CoV in an intranasal mouse vaccine model (39) , suggesting perhaps a similar important role for T cells in the context of SARS-CoV-2.

Based on the breadth of the observed immune response with SpFN+ALFQ, and protection following challenge with USA-WA1/2020 strain in nonhuman primates (40) , the induction of broad SARS coronavirus immune responses in mice (unpublished data), a detailed study of the T cell responses induced by SpFN+ALFQ was warranted. The mouse system offers the advantage of defining the in vivo mechanisms of innate and adaptive immunity following vaccination with SpFN+ALFQ in comparison with a common and well-established adjuvant AH.

Following vaccination with SpFN+ALFQ, we observed robust and sustained recruitment and activation of classical and nonclassical APCs in the dLNs compared to SpFN+AH. Furthermore, we show that cDC1 and cDC2, in particular, exhibited enhanced upregulation of costimulatory molecules necessary for T cell engagement and differentiation. Interestingly, by thoroughly analyzing the immune response within the dLN of the vaccination site, we discovered that SpFN+ALFQ also recruited more non-classical APCs and monocytes that support recruitment and activation of APCs and T cells that enhance vaccine efficacy (41) . Moreover, we show that lymph node resident macrophages (42, 43) increased in number and functional activation in response to ALFQ+SpFN and may support and optimize plasma cell (25, 44) and T cell development (44) in response to this vaccine strategy.

Although both vaccine-adjuvant formulations (SpFN+ALFQ and SpFN+AH) recruited naïve T cells, the response elicited by ALFQ was superior and faster at generating or differentiating effector memory phenotype cells than seen with AH. Hence, the adjuvant determines the quality and the quantity of early innate responses which then sets the stage for downstream adaptive immune responses. Furthermore, our results suggest that the effective APC responses induced by SpFN+ALFQ resulted in differentiation of virus-specific CD4 + T cells that were biased towards a TH1 phenotype (TNF- and IL-2). Multiplex cytokine data analysis in the sera of SpFN+ALFQ vaccinated mice, post priming, also showed a similar skewing towards a TH1 response (IL-2, TNF-, and IFN-). We also noted a significant increase in the secretion of IFN-γ, IL-2, and TNF- from both CD4 + and CD8 + T cells from the spleen of SpFN+ALFQ vaccinated mice compared to SpFN+AH in response to spike peptide pools.

Following a booster vaccination, SARS-CoV-2 spike-specific T cell responses were also present in the mediastinal lymph nodes that drain the lungs. A 3-fold increase in MHC class I restricted K b -spike(539-546)-specific CD8 + T cells was also observed in the lungs of mice vaccinated with SpFN+ALFQ. The importance of T cell surveillance in the lungs cannot be overstated as they are the primary target of SARS-CoV-2 infection.

SpFN+ALFQ vaccinated mice exhibited a higher frequency of spike-specific CD4 + and CD8 + T cells secreting 

The major aim of the study was to utilize a unique liposomal adjuvant formulation and a well-known and widely 

SARS-CoV-2 prefusion antigen construct, derived from the Wuhan-Hu-1 genome sequence was designed as ferritin-fusion recombinant protein for expression as a nanoparticle. The Helicobacter pylori ferritin molecule was linked to the C-terminal region of pre-fusion stabilized ectodomain (residues 12-1158). Several modifications were introduced. The spike ectodomain was modified to introduce two prolines residues (K986P, V987P) as previously described (16) 

Spectral flow cytometry was performed to analyze APCs and T cells in the lymph nodes as indicated. Single cell suspensions were treated with Fc Block (BD Biosciences) for 5 minutes at 4C before further staining. Live dead blue dye (Invitrogen) was used to stain dead cells. Analysis of APC and T cell phenotypes were performed using the fluorochrome-conjugated antibodies in table S1. Intracellular cytokine analysis was performed to identify SARS-CoV2 spike-specific T cells. Single cell suspensions from the draining lymph node were plated at 0.5x10 6 cells per well in a 96-well plate and stimulated with the spike (S) 1 or S2 peptide pools from JPT at a final concentration of 1 g/mL for 6 hours at 37C in RPMI and 10% FBS. To prevent the secretion of cytokines, GolgiStop (BD Biosciences) was added 1 hour after the addition of peptides. Spike-specific T cells were measured by surface staining followed by fixation and permeabilization using BD fixation/permeabilization solution kit (BD Biosciences,) and intracellular cytokine staining (table S1). As negative and positive controls, cells were cultured in stimulation media without peptide stimulation or with phorbol 12-myristate 13-acetate (PMA) and

ionomycin (BioLegend), respectively. T cells were considered to be responsive to peptide stimulation if the frequency of cytokine-expressing T cells was greater than 0.01% of the parent population after background subtraction. Samples were analyzed on a 5-laser Cytek Aurora flow cytometer (Cytek Biosciences). The data were analyzed using FlowJo software v10 (Tree Star, Inc.). Gating strategy for APC subsets, T cell phenotyping and ICS are shown in fig. 1SA and fig. 2SC , respectively.

Cryopreserved splenocytes were quickly thawed and added to 10 mL of complete RPMI 1640 media supplemented with 5% FBS and 1% Pen-strep followed by viability assessment by trypan blue exclusion method.

Approximately, 1x10 6 splenocytes were cultured in the presence of peptide pools directed towards SARS CoV-2 spike protein S1 or S2 (JPT) (1µg/mL) at 37°C, 5% CO2. After 1 hour of incubation with peptides, protein transport inhibitor (BD Golgi Plug™ containing Brefeldin A, 1 µg/mL and BD Golgi Stop™ containing monensin, 1 µg/mL, BD Biosciences) was added for an additional 5 hours at 37°C, 5% CO2. 

To detect the epitope specific CD8 + T cell cytokine production, we combined antigen stimulation with tetramer labelling and intracellular cytokine staining. Splenocytes (2 × 10 6 /well) were cultured in a 96-well v-bottom plate in the presence of peptide pools directed towards SARS CoV-2 spike protein S1 (1µg/mL) at 37°C, 5% CO2.

After 1 hour of incubation with peptides, protein transport inhibitor (BD Golgi Plug™ containing Brefeldin A, 1 µg/mL and BD Golgi Stop™ containing monensin, 1 µg/mL, BD Biosciences) was added followed by an additional 5-hour incubation at 37°C, 5% CO2. At the end of the incubation period, cells were washed at 300 g at 4°C for 7 mins, then labeled with a dead cell discrimination dye (Aqua stain, 1:1000 dilution, Invitrogen) for 30 

Single cell suspensions from spleens of vaccinated mice were used as effector cells and naive spleens were used as target cells. Target cells were prepared by diluting cells to 20x10 6 /mL and making 2 x 500 µL aliquots (10x10 6 each). One aliquot was incubated in the presence of SARS CoV-2 spike protein specific peptides LVKNKCVNFNFNGLT and KCVNFNFNGLTGTGV at 1 µg/mL (JPT) and the other aliquot was incubated in media for 45 min in a 37C water bath. Both aliquots were washed twice with 3 mL PBS. CFSE high and low solutions were made at 0.5 and 0.05 µM respectively using ThermoFisher CellTrace™ CFSE Cell Proliferation

Kit (Catalog C34554). Peptide pulsed and non-pulsed cells were resuspended respectively in CFSE high and low solutions, incubated for 10 min in a 37C water bath, and then washed twice in media. Target cells were resuspended at 2x10 6 /mL and 50 µL of each CFSE high and low (1:1, CFSE high:CFSE low) was added to each well. Effector cells were resuspended at 40x10 6 /mL (10x10 6 cells in 250 µL) and 2-fold serial dilutions were made by adding 125 µL of cells to 125 µL of media. 100 µL of each dilution was added to a well containing the target cell mixture, bringing the total volume to 200 µL. Cells were co-incubated overnight at 37°C, 5% CO2.

Following incubation, cells were stained with Live/Dead Aqua stain (1:1000 dilution) (Invitrogen, Thermo Fisher Scientific) for 30 minutes at 4°C followed by washing twice with 1XPBS. Then, the cells were blocked with antimouse CD16/CD32 for 30 mins followed by further incubation with 1µL (1.2µg) of PE-conjugated H-2K b spike(539-546)-specific tetramer for 30 min at 4 °C in the dark. After washing, cells were further stained with BUV737-anti-mouse CD3, BUV395-anti-mouse CD4, BV711-anti-mouse CD8a and incubated for an additional 30 mins in the dark at room temperature. Flow cytometric acquisition and data analysis were performed as described above with appropriate controls.

Flow cytometric data were analyzed using FlowJo v.10.0.8 (BD). Data are displayed as dot plots, or bar graphs.

AH and ALFQ groups were compared by unpaired Student's t-test or Mann-Whitney U-test or Pearson correlation. Graphs were plotted using GraphPad Prism v.8.4.0. Statistical analyses were conducted using GraphPad Prism v.8.4.0 software. P values <0.05 were considered statistically significant. Fig. S1 . Flow cytometry analysis of innate immune cells in draining lymph nodes responding to SpFN vaccine adjuvanted with ALFQ or AH. Table S1 . List of antibodies used for Cytex flow cytometry based analysis of the APC phenotypes Table S2 . List of antibodies used for Cytex flow cytometry based analysis of the T cell phenotypes. Table S3 . List of antibodies used for Cytex flow cytometry based analysis of the T cell cytokine staining. Table S4 . List of antibodies used for intracellular cytokine staining by BD LSRII flow cytometry based analysis of the cytokine response in the splenocytes. ) and TNF- expressing (right panel) SARS-CoV-2 spike-specific CD4 + T cells in the dLNs 10 days post vaccination, following stimulation with spike S1 peptide pool. Bars indicate mean + SD. (C) IL-2 and TNF- are co-expressed in the SARS-CoV-2 spike-specific CD4 + T cells, indicating mainly a TH 1 type of response. (D) Cytokine profile in the sera of vaccinated mice as determined by multiplex ELISA using Meso Scale Discovery (MSD) platform on days 3, 5, and 10 post-vaccination. Dots represent data from individual mice. Experiments were repeated twice and differences between the two groups were analyzed by using non-parametric Mann-Whitney U-test with p ≤ 0.05 considered as statistically significant. Statistical significance between the groups was determined using non-parametric Mann-Whitney U-test with p ≤ 0.05 considered as statistically significant. Bar graph and representative flow plots depicting the significant differences in the frequency of the SARS-Cov-2 spike-specific CD8 + T cells concomitantly producing both IFN-γ and TNF-α. Bars represent the mean + SD. and significant differences between the groups were calculated using non-parametric two-tailed Mann-Whitney U test in Graph pad prism version 8. (F) CD8 + T cell polyfunctionality: Differences in the ability of CD8 + T cells stimulated with SARS-CoV-2 spike peptides to secrete more than one cytokine. Boolean gating was applied to identify all combinations of CD8 + T cell effector functions. The pie chart depicts the average proportion of spike-specific CD8 + T cells producing all three (IFN-γ, IL-2 or TNF-α), any two or any one cytokine. The graphs show cytokine-secreting peptide-dspecific CD8 + T cells for each individual mouse (n = 5 /group) in response to each adjuvant formulation. This analysis was performed using the SPICE software version 5.1 (54) . Table S1 . List of antibodies used for Cytex flow cytometry-based analysis of the APC phenotypes Table S2 . List of antibodies used for Cytex flow cytometry-based analysis of the T cell phenotypes. Table S3 . List of antibodies used for Cytex flow cytometry-based analysis of the T cell cytokine staining. Table S4 . List of antibodies used for intracellular cytokine staining by BD LSRII flow cytometry-based analysis of the cytokine response in the splenocytes. Highlighted peptides were part of ELISpot-reactive SARS-CoV-2 glycoprotein derived matrix peptide pools. Underlined peptides are sequences predicted by NETMHC prediction algorithm (iedb.org). Numbers indicate amino acid residues. 

A novel coronavirus associated with severe acute respiratory syndrome

Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus-Infected Pneumonia in

Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia

A Novel Coronavirus Emerging in China -Key Questions for Impact Assessment

COVID-19 Vaccine: A comprehensive status report

m, An mRNA Vaccine against SARS-CoV-2 -Preliminary Report

Phase I/II study of COVID-19 RNA vaccine BNT162b1 in adults

Ad26 vector-based COVID-19 vaccine encoding a prefusion-stabilized SARS-CoV-2 Spike immunogen induces potent humoral and cellular immune responses

Safety and immunogenicity of an rAd26 and rAd5 vector-based heterologous prime-boost COVID-19 vaccine in two formulations: two open, non-randomised phase 1/2 studies from Russia

Resurgence of COVID-19 in Manaus, Brazil, despite high seroprevalence

Selfassembling influenza nanoparticle vaccines elicit broadly neutralizing H1N1 antibodies

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

Army Liposome Formulation (ALF) family of vaccine adjuvants

Adjuvant system AS01: helping to overcome the challenges of modern vaccines

Differential immune responses to HIV-1 envelope protein induced by liposomal adjuvant formulations containing monophosphoryl lipid A with or without QS21

Liposomes containing monophosphoryl lipid A and QS-21 serve as an effective adjuvant for soluble circumsporozoite protein malaria vaccine FMP013

Liposome Formulations as Adjuvants for Vaccines

mRNA induced expression of human angiotensin-converting enzyme 2 in mice for the study of the adaptive immune response to severe acute respiratory syndrome coronavirus 2

Innate immunity and vaccines

Adjuvants in the Driver's Seat: How Magnitude, Type, Fine Specificity and Longevity of Immune Responses Are Driven by Distinct Classes of Immune Potentiators

Lymph node macrophages

The role of lymph node sinus macrophages in host defense

T-cell unresponsiveness in vivo and in vitro: fine specificity of induction and molecular characterization of the unresponsive state

Molecular mechanisms of CD4+ T-cell anergy

Neutrophils acquire the capacity for antigen presentation to memory CD4 + T cells in vitro and ex vivo

Mouse neutrophils are professional antigen-presenting cells programmed to instruct Th1 and Th17 T-cell differentiation

Antigen presentation by monocytes and monocyte-derived cells

Identification of murine CD8 T cell epitopes in codon-optimized SARS-associated coronavirus spike protein

Pathogen-Specific T Cell Polyfunctionality Is a Correlate of T Cell Efficacy and Immune Protection

Dynamics of the cytotoxic T cell response to a model of acute viral infection

Lung airway-surveilling CXCR3(hi) memory CD8(+) T cells are critical for protection against influenza A virus

RSV-specific airway resident memory CD8+ T cells and differential disease severity after experimental human infection

Memory T Cell Dynamics in the Lung during Influenza Virus Infection

SARS-CoV-2 T cell immunity: Specificity, function, durability, and role in protection

Airway Memory CD4(+) T Cells Mediate Protective Immunity against Emerging Respiratory Coronaviruses

Neutrophils influence the level of antigen presentation during the immune response to protein antigens in adjuvants

Changes in the medulla of the parathymic lymph nodes of the rat during acute gastro-intestinal inflammation

Antigens in immunity. XIV. Electron microscopic radioautographic studies of antigen capture in the lymph node medulla

T Cell Zone Resident Macrophages Silently Dispose of Apoptotic Cells in the Lymph Node

Immunization with SARS coronavirus vaccines leads to pulmonary immunopathology on challenge with the SARS virus

An assessment of the use of Bayes' Theorem for forecasting in public health: the case of epidemic meningitis in China

The relationship between cell viability and changes in mitochondrial ultrastructure, cellular ATP, ion and water content following injury of Ehrlich ascites tumor cells

IFN-gamma production during initial infection determines the outcome of reinfection with respiratory syncytial virus

IFN-Gamma Inhibits JC Virus Replication in Glial Cells by Suppressing T-Antigen Expression

Interplay between SARS-CoV-2 and the type I interferon response

SARS-CoV-2-specific T cell immunity in cases of COVID-19 and SARS, and uninfected controls

Deep immune profiling of COVID-19 patients reveals distinct immunotypes with therapeutic implications

Immune response to antigen adsorbed to aluminum hydroxide particles: Effects of co-adsorption of ALF or ALFQ adjuvant to the aluminum-antigen complex

SPICE: exploration and analysis of post-cytometric complex multivariate datasets

Supplementary Materials SARS-CoV-2 spike-ferritin nanoparticle vaccine formulated with liposomal QS21 enhances functional antiviral cellular responses

Malloy 3 † , and Mangala Rao 1 ‡ ‡ Corresponding author. Email: mangala.rao.civ@mail.mil; mrao@hivresearch.org † Co-Corresponding Author

The tetramers were obtained from NIH Tetramer Core Facility at Emory University, Atlanta, GA.