key: cord-0703394-vdmnerbv
authors: Ortega-Rivera, Oscar A.; Shukla, Sourabh; Shin, Matthew D.; Chen, Angela; Beiss, Veronique; Moreno-Gonzalez, Miguel A.; Zheng, Yi; Clark, Alex E.; Carlin, Aaron F.; Pokorski, Jonathan K.; Steinmetz, Nicole F.
title: Cowpea Mosaic Virus Nanoparticle Vaccine Candidates Displaying Peptide Epitopes Can Neutralize the Severe Acute Respiratory Syndrome Coronavirus
date: 2021-10-21
journal: ACS Infect Dis
DOI: 10.1021/acsinfecdis.1c00410
sha: b30f8cd99c3511edb4e742a179819329996f3182
doc_id: 703394
cord_uid: vdmnerbv

[Image: see text] The development of vaccines against coronaviruses has focused on the spike (S) protein, which is required for the recognition of host-cell receptors and thus elicits neutralizing antibodies. Targeting conserved epitopes on the S protein offers the potential for pan-beta-coronavirus vaccines that could prevent future pandemics. We displayed five B-cell epitopes, originally identified in the convalescent sera from recovered severe acute respiratory syndrome (SARS) patients, on the surface of the cowpea mosaic virus (CPMV) and evaluated these formulations as vaccines. Prime-boost immunization of mice with three of these candidate vaccines, CPMV-988, CPMV-1173, and CPMV-1209, elicited high antibody titers that neutralized the severe acute respiratory syndrome coronavirus (SARS-CoV) in vitro and showed an early Th1-biased profile (2–4 weeks) transitioning to a slightly Th2-biased profile just after the second boost (6 weeks). A pentavalent slow-release implant comprising all five peptides displayed on the CPMV elicited anti-S protein and epitope-specific antibody titers, albeit at a lower magnitude compared to the soluble formulations. While the CPMV remained intact when released from the PLGA implants, processing results in loss of RNA, which acts as an adjuvant. Loss of RNA may be a reason for the lower efficacy of the implants. Finally, although the three epitopes (988, 1173, and 1209) that were found to be neutralizing the SARS-CoV were 100% identical to the SARS-CoV-2, none of the vaccine candidates neutralized the SARS-CoV-2 in vitro suggesting differences in the natural epitope perhaps caused by conformational changes or the presence of N-linked glycans. While a cross-protective vaccine candidate was not developed, a multivalent SARS vaccine was developed. The technology discussed here is a versatile vaccination platform that can be pivoted toward other diseases and applications that are not limited to infectious diseases.

the remainder using the conventional approach of virus inactivation. 8 More than 300 further vaccine candidates have entered preclinical development, including several produced in plants. 9 The development of vaccines against all three viruses has focused on the spike (S) protein, which projects from the viral envelope and is required for the recognition of host-cell receptors that mediate viral entry. 10−13 In the case of the SARS-CoV and SARS-CoV-2, this receptor is human angiotensin converting enzyme 2 (hACE2), 14, 15 whereas the MERS-CoV receptor is dipeptidyl peptidase 4 (CD26/ hDPP4). 16 The most potent neutralizing antibodies developed thus far recognize the receptor-binding domain (RBD) of the S protein and directly interfere with receptor−ligand binding. 17 The RBD contains multiple neutralizing antibody epitopes but few immunodominant epitopes that could induce harmful immune responses such as antibody-dependent enhancement of viral infection and therefore appears to be a better vaccine candidate than the entire S protein. 18 Indeed, the RBD induces a higher titer of neutralizing antibodies than the S1 ectodomain, 19 but the RBD alone has low immunogenicity. This can be addressed by the simultaneous administration of adjuvants 20 combined with alternative strategies such as RBD dimers 21 or the presentation of multiple epitopes on virus-like particles (VLPs) and nanoparticles. 22 Given that the S proteins from the SARS-CoV and SARS-CoV-2 share 77% sequence identity and recognize the same receptor 23 and the nucleocapsid (N) and S proteins of the viruses share 27 T-cell epitopes and 49 B-cell epitopes, 24 groups developing vaccines have investigated the potential for cross-protection. 25 Several of these studies have shown that sera or purified IgG from animals immunized with the S protein or the RBD of one virus can neutralize the other. 26−31 This suggests that vaccine candidates targeting the S protein, and more specifically the RBD, have the potential to be developed as pan-beta-coronavirus vaccines that elicit broadly neutralizing antibodies, which can maintain efficacy against coronavirus divergence, thus offering a potential strategy to prevent/combat future pandemics caused by this virus group. 32 , 33 We set out to engineer a broadly effective vaccine against beta-coronaviruses by selecting five B-cell epitopes, four with 100% sequence identity between the SARS-CoV and SARS-CoV-2, 24 which were identified by immunoinformatic mapping 34−36 and validated as immunogenic or reactive to sera from convalescent SARS-CoV patients. 37 −39 These peptide epitopes were conjugated to the surface of a plant virus (cowpea mosaic virus, CPMV), a strategy that has proven to be effective in previous studies for the development of potent vaccine candidates based on plant viral nanoparticles (VNPs) displaying multiple copies of the epitope to elicit an enhanced immune response. 40 Such VNP-based vaccines can be manufactured via propagation in host plants with relatively lower technological and financial burden. Plants have been used to produce vaccine candidates for 30 years, and several companies are developing COVID-19 vaccines in plants including one based on self-assembling S-protein virus-like particles (VLPs) that has completed phase I clinical trials 22 and is now beginning phase II/III, with millions of doses already ordered by the Canadian Government. 41 Here, we report how plant VNPs based on the CPMV conjugated with B-cell epitopes can be developed as a vaccine candidate for beta-coronavirus neutralization. Soluble CPMV vaccine candidates were evaluated using prime-boost immunization. The ongoing COVID-19 pandemic highlights the need for vaccine candidates and delivery devices that overcome cold chain requirements and are effective after a single administration. Therefore, we also formulated the CPMV vaccine candidates as a slow-release polymer implant prepared by hotmelt extrusion, a highly scalable process technology suited for epidemic or pandemic response. Our previous research showed that VNPs can withstand hot-melt extrusion, yielding slowrelease polymer melts that release structurally intact and biologically active VNPs; in previous work, virus-like particles (VLPs) derived from a bacteriophage were considered; 42−46 here, we apply these technologies to the plant virus CPMV.

Selection of B-Cell Epitopes. We selected five B-cell epitopes that were previously identified as targets of antibodies from sera of convalescent patients with SARS 37, 38 or that were shown to neutralize the SARS-CoV when used as a source of antigens in a vaccine candidate. 39 Four of the epitopes shared 100% sequence identity between the SARS-CoV and SARS-CoV-2 S proteins, whereas epitope 362 showed 76% sequence identity between the viruses (Table 1 ). 24 Two of the epitopes (317 and 362) are located within or immediately adjacent to the RBD, which binds the receptor ACE2, whereas the other three are located in the HR1/HR2 stalk region of the S2 subunit, which is required for cell fusion ( Figure 1 ). 47, 48 Preparation of CPMV-Based Vaccine Candidates. Each of the five B-cell epitopes was synthesized as a synthetic peptide with N-terminal cysteine followed by a GGG linker for conjugation to CPMV particles using a two-step procedure (Figure 2A,B) . The vaccine candidates were purified by ultracentrifugation and characterized by DLS, TEM, and SDS-PAGE under reducing conditions to confirm the structural integrity and determine the degree of antigen incorporation ( Figure 2C −E). SDS-PAGE analysis indicated that each VNP presented 46−52 peptides ( Figure 2C ). Efficiency of the conjugations varied by peptideit was interesting to note the band laddering effect for the CPMV-1173 and CPMV-1209 ( Figure 2C lanes 2 and 3) , which indicates that more than one peptide was conjugated per coat protein; this is possible because the S and L proteins are both present with multiple lysine side chains (2 per S and 3 per L). 49 Nevertheless, the overall peptide loading per CPMV was comparable among the five vaccine candidates. DLS and TEM confirmed that the VNPs were monodisperse before and after conjugation ( Figure  2D ,E) and that their hydrodynamic diameter increased only marginally from 30.4 nm for the wild-type particles to 32.7 nm for each vaccine candidate ( Figure 2D ,E). We also observed no clouding of the suspension; therefore, data indicate stable formulations free of protein aggregation. Immunogenicity of the CPMV-Based Vaccine Candidates. All five vaccine candidates were injected into mice with a prime-boost-boost schedule, and antibodies against all five peptides were recovered ( Figure 3A ,B). Endpoint IgG titers against CPMV-317 increased steadily, from 1:560 (week 2) to 1:6160 (week 4), 1:100,480 (week 6), and 1:105,600 (week 10). For CPMV-988, IgG titers were already high after 2 weeks (1:11,520) and then gradually declined to 1:10,240 (week 4), 1:7680 (week 6), and 1:8320 (week 10). For CPMV-1173, IgG titers initially increased from 1:3840 (week 2) to 1:18,560 (week 4) and 1:92,160 (week 6) but then declined to 1:81,920 (week 10). Similarly, the IgG titers against CPMV-1209 increased from 1:2560 (week 2) to 1:14,080 (week 4) and 1:35,840 (week 6) before falling to 1:19,200 (week 10). Finally, endpoint IgG titers against CPMV-362 remained relatively constant at 1:12,800 (week 2), 1:25,600 (week 4), 1:11,520 (week 6), and 1:12,800 (week 10), albeit with a transient increase at week 4. As expected, free peptides were not immunogenic, and no IgG titers were detected after priming and two boosts ( Figure 3C ).

Ig isotypes ( Figure S1 ) and IgG subclasses ( Figure 3D ) were investigated to determine whether the candidates induced a Th cell response based on the IgG1/IgG2a ratio (values <1 defined as Th1-biased and values >1 defined as Th2-biased). After priming (week 2), all candidates except CPMV-362 showed a Th1-biased profile, but this switched to a Th2-biased profile after the second boost ( Figure 3D) .

Immunoreactivity of the Serum against the SARS-CoV-2 S Protein. The plasma samples used to determine IgG titers and isotype profiles were also used to provide a preliminary indication of immunoreactivity against the SARS-CoV-2 S protein. ELISA experiments revealed that IgG from the plasma samples taken on weeks 2, 4, and 6 from all mice injected with the vaccine candidates was able to recognize the SARS-CoV-2 S protein ( Figure 4A ). Four of the candidates showed week-2 titers of 1:25,000, the exception being CPMV-362 with a titer of only 1:5000. By week 6, three of the candidates achieved titers of 1:125,000, with CPMV-1173 slightly lower at 1:105,000 and CPMV-1209 the lowest at 1:85,000. Plasma from mice injected with the free peptides did not bind to the S protein ( Figure 4B ).

Analysis of T-Cell Responses to the Vaccine Candidates. The T-cell responses to the vaccine candidates were evaluated using an ELISpot assay following the vaccination of mice with a single dose (prime) or the complete schedule (prime and two boosters). Splenocytes were collected from immunized animals and stimulated with each peptide (317, 988, 1173, 1209, or 362), with a normal cell culture medium as a negative control or with the CPMV plus PMA and ionomycin as a positive control. ELISpot analysis of splenocytes from animals receiving the vaccine candidates (CPMV-317, CPMV-988, CPMV-1173, CPMV-1209, or CPMV-362) did not show the presence of IFN-γ spot-forming colonies (SFCs) when stimulated with the matching peptide regardless of whether the VNPs were delivered as a single dose ( Figure 5A ) or three doses ( Figure 5B ), which supports evidence of a Th2-biased profile provided by the IgG subclass ratios ( Figure 3D ). However, we also observed no significant increase in the abundance of IL-4 SFCs following stimulation with the matching peptides ( Figure 5A ,B), indicating that basal levels of IL-4 are sufficient for a Th2-biased response. Splenocytes stimulated with the CPMV resulted in the appearance of more IFN-γ SFCs ( Figure 5A ,B), demonstrating that the CPMV itself triggers a Th1 response as previously reported. 50 As expected, stimulation with PMA and ionomycin triggered an increase in both IFN-γ and IL-4 SFCs. Representative images from each vaccine group and the various stimulants are shown in Figure 5C .

Neutralization of the SARS-CoV and SARS-CoV-2. Having confirmed the presence of IgG recognizing the peptide epitopes and the S protein following immunization with all five vaccine candidates, we tested the plasma samples for the presence of neutralizing antibodies against the SARS-CoV and SARS-CoV-2. Pooled plasma from each vaccinated group was preincubated separately with each virus and then added to Vero 76 cells to evaluate the impact on virus cytopathicity. The EC 50 and CC 50 were determined visually and from neutral red absorbance readings and were used to calculate the SI 50 with values ≥10 indicating neutralizing activity. The neutralization titers are reported in Table 2 . Only the plasma sample from mice injected with CPMV-988, CPMV-1173, and CPMV-1209 was able to neutralize the SARS-CoV, with similar neutralization titers but a slightly higher SI 50 value in the case of CPMV-1209 ( Table 2) . None of the samples were able to neutralize the SARS-CoV-2 in the same assay ( Table 2) .

Analysis of Pentavalent Vaccine Implants. The five CPMV-based vaccine candidates were formulated as a cylindrical implant by hot-melt extrusion using a mixture of the pentavalent VNPs, PLGA, and PEG8000 (10:75:15, w/w/ w). The dose loaded for each pentavalent implant (300 μg of each CPMV vaccine) was equivalent to the three-dose (100 μg each dose; prime-boost-boost) soluble injection schedule. The same approach was used to formulate CPMV-Cy5 VNPs (the characterization of CPMV-Cy5 is reported in the Supporting Information, Figure S2 ) to trace their slow-release profiles post-implantation and their fate during lymph node drainage. Lymph nodes were collected every week for 5 weeks from animals carrying the CPMV-Cy5 implant ( Figure 6A) , and blood was taken every 2 weeks for 6 weeks from animals carrying the pentavalent implant ( Figure 6B ). We collected cervical, axillary, and inguinal lymph nodes, as shown in Figure  6C . The gradual release of CPMV-Cy5 from the implant over 5 weeks was confirmed by the progressive loss of fluorescence from the implantation zone ( Figure 6D ).

In addition to the fluorescence imaging studies of CPMV-Cy5/PLGA implants, we also determined whether the CPMV was intact when released from the implants in vitro. The CPMV recovered from implants was subjected to DLS sizing experiments and TEM imaging. DLS and TEM are consistent with intact and monodisperse nanoparticles being released from the implants with sizes measuring ∼30 nm and indistinguishable features compared to the CPMV ( Figure S3 ). . For CPMV-317, titers were 1:1600, 1:7200, and 1:7200 (weeks 2, 4, and 6, respectively); for CPMV-362, titers were 1:1600, 1:3200, and 1:4160 (weeks 2, 4, and 6, respectively); for CPMV-988, titers were 1:1120, 1:2400, and 1:1440 (weeks 2, 4, and 6, respectively); for CPMV-1173, titers were 1:1040, 1:6080, and 1:10880 (weeks 2, 4, and 6, respectively); and for CPMV-1209, titers were 1:2080, 1:4000, and 1:3680 (weeks 2, 4, and 6, respectively). The antibody titer at week 6 against the SARS-CoV-2 S protein was 1:25000, which corresponds to all the IgG generated against the five epitopes used in the CPMV vaccines that can bind the S protein.

The fate of CPMV-Cy5 particles was traced in more detail by analyzing sections of cervical, axillary, and inguinal lymph nodes on days 0, 7, 14, and 28 using antibodies against the markers CD4 (T cells), CD45R (B cells), and CD11c (dendritic cells) for colocalization ( Figure 7) . On day 0, we observed the presence of B-cell-rich zones (CD45R, red panel) surrounded by T cells (CD4, blue panel) and a random distribution of dendritic cells (CD11c, green panel) within the T-cell-rich zone in all lymph nodes. No background was observed from the Cy5 channel (CPMV-Cy5, gray panel). On day 7, we detected the presence of CPMV-Cy5 in the cervical lymph nodes, including the formation of germinal centers (Figure 7 , merged panel). A weak signal was also detected in the axillary lymph node, but the presence of germinal centers was not clear. On day 14, CPMV-Cy5 was present in all lymph nodes, and germinal centers were clearly present in axillary ACS Infectious Diseases pubs.acs.org/journal/aidcbc Article lymph nodes. The distribution of CPMV-Cy5 predominantly within the T-cell-rich zone was more evident in the cervical and inguinal lymph nodes, but the CPMV-Cy5 signal in all lymph nodes was similar. Finally, on day 28, the CPMV-Cy5 signal was not detected in the axillary lymph nodes, but a residual signal remained in the cervical lymph nodes and a slightly stronger signal in the inguinal lymph nodes. These observations suggest that CPMV-Cy5 drains initially to the lymph nodes closest to the implant zone, in this case the cervical lymph nodes (Figure 7, day 7) , followed by a broader distribution to all lymph nodes by day 14. Ultimately, the last VNPs from the implant drain to the closest (cervical) lymph nodes and therefore show a stronger signal than the farthest (inguinal) lymph nodes. This corresponds to the degradation profile of the implant, with ∼50% of the CPMV-Cy5 released by day 14 and almost all released by day 28 ( Figure 6D ).

Beta-coronaviruses share a small number of conserved B-cell and T-cell epitopes on the N and S proteins that suggest that it may be possible to develop pan-beta-coronavirus vaccines that protect not only against known species such as the SARS-CoV-2 but also new variants and even new species that could, if not tackled preemptively, lead to future pandemics. 32, 33 The S protein is considered the most suitable vaccine target for coronaviruses because it is responsible for interactions with host-cell receptors and cell fusion and thus elicits neutralizing antibodies. 17 We therefore selected five B-cell epitopes, 24 two within or adjacent to the RBD on the S1 subunit and three within the heptad repeats that define the stalk region of S2, which promotes membrane fusion ( Figure 1 ). Importantly, all but one of the epitopes showed 100% sequence identity between the SARS-CoV and SARS-CoV-2 and were previously found to be immunogenic or reactive to sera from convalescent SARS-CoV patients. 37−39 Each epitope was (separately) conjugated to CPMV particles to produce soluble vaccine candidates, 40 and we also prepared a pentavalent implant by mixing equimolar amounts of all five candidates with PLGA and PEG to form a slow-release formulation. 51−53 The display of peptides on the surface of plant viruses has been widely used as a strategy to develop vaccines, either by direct conjugation to virus particles or by the genetic engineering of coat proteins, but it is important to ensure that the peptides are compatible with virus assembly to avoid particle dissolution or aggregation. 54 DLS, TEM, and SDS-PAGE were used to confirm the structural integrity of the particles and also to determine the degree of antigen incorporation ( Figure 2 ). DLS and TEM confirmed that the particles were monodisperse both before and after conjugation, indicating that conjugation neither destabilized the particles (resulting in them breaking into individual coat protein subunits) nor caused them to aggregate and precipitate from solution. SDS-PAGE analysis showed that each VNP presented 46−52 peptides suggesting comparable bioconjugation of all five peptides using the SM-(PEG) 4 linker and highlighting the flexibility of the CPMV platform in accommodating peptides ranging from 8 to 22 amino acid residues and varying pIs of 3.7 to 11.8. ( Table 1) .

The five vaccine candidates were then injected into mice, and we found that plasma from the immunized animals was able to bind to the corresponding peptides in vitro, whereas plasma from animals injected with free peptides was not ( Figure 3) . These experiments confirmed that the CPMV vaccine candidates were able to induce significant antibody titers after priming and two boosts. Notably, some candidates elicited high antibody titers that remained consistent over the 2−10 week time frame (CPMV-362 and CPMV-988), whereas others elicited initially lower antibody titers that increased over time (CPMV-317, CPMV-1173, and CPMV-1209). This highlights the potency of the CPMV, which serves as an antigen display and delivery technology but also as a potent adjuvant that stimulates innate immune cells by signaling through the Toll-like receptors. We recently demonstrated that the CPMV signals through TLR-2, TLR-4, and TLR-7. 55 The temporal variability of antibody titers is likely influenced by the intrinsic properties of peptides (Table 1 ; hydrophobicity, length, or relative positioning of certain residues) and could be evaluated in future studies. For example, it would be useful to establish the attributes of peptides that elicit high antibody titers after a single dose when displayed on the CPMV and whether the administration of single doses of CPMV-362 and CPMV-988 would achieve lasting high antibody titers similar to those obtained after the prime-boost schedule. Likewise, the five vaccine candidates also differed in terms of the immune response based on the type of antibody produced. Interestingly, candidate CPMV-362 showed an immediate Th2-biased response, whereas all other candidates showed an initial Th1biased response that switched to a Th2-biased response after 2 weeks (CPMV-317) or 6 weeks (all others). VNP-based vaccine platforms displaying multimeric self-epitopes or heterologous epitopes (target vaccine antigen) on the surface promote cross-linking of B-cell receptors (BCRs) that could prime B cells to induce the production of antibodies even without the help of CD4 + T cells. 56, 57 The CPMV, like other sub-200 nm nanoparticles, can diffuse and drain to lymph nodes without presentation on APCs, reaching zones rich in T and B cells in the lymph node periphery. 56, 57 Early IgG2a (Th1-biased) responses may reflect fast and direct priming interactions between the CPMV vaccine candidates and B cells in the lymph node, whereas later IgG1 (Th2-biased) responses after two boosts may favor APC presentation. Accordingly, although the CPMV was thought to be a clear Th1 adjuvant for peptide vaccines, 50,58 every new peptide/epitope must be tested on a case-by-case basis. An early Th1 response was observed based on antibody isotypes, but the T-cell response did not mirror this response. We expected to see the production of IFN-γ, a signature cytokine for the Th1 profile, after coculturing the splenocytes from vaccinated mice with the same antigens used for the CPMV vaccines. The peptide alone may not be enough to stimulate the release of IFN-γ from T cells during the stimulation of splenocytes from vaccinated animals in vitro ( Figure 5 ), or the Th1-biased isotype may reflect a direct priming interaction between the CPMV vaccine candidates and B cells, without the input of T cells to control isotype switching. These observations should be investigated in more detail in the future.

ELISAs confirmed that all five vaccines generated high titers of antibodies that bound to the corresponding peptides and to Figure 5 . ELISpot assay with splenocytes from vaccinated mice. Isolated splenocytes (5 × 10 5 cells per well) from immunized mice (n = 3) were stimulated with the medium only (negative control), the matching free peptides (20 μg/mL), and the unmodified CPMV (10 μg/mL) or PMA (50 ng/mL) plus ionomycin (1 μg/mL) as a positive control. Cytokine-producing cells (IFN-γ or IL-4) were counted as spot-forming colonies (SFCs) from (A) week 2 (W2) or (B) W10 after the first immunization. (C) Representative images of blue (IL-4) and red (IFN-γ) spots formed by stimulated splenocytes from W2 or W10 after the first immunization. Statistical significance: *p < 0.001 vs the normal medium. the recombinant SARS-CoV-2 S protein ( Figure 4 ). However, in vitro neutralization assays against the SARS-CoV and SARS-CoV-2 showed that plasma samples from only three candidates, namely, CPMV-988, CPMV-1173, and CPMV-1209, were able to neutralize the SARS-CoV. Nevertheless, none of the candidates were able to neutralize the SARS-CoV-2 ( Table 2 ). Possible explanations include the inaccessibility of corresponding epitopes on the S protein to the antibodies due to the conformation of the protein or the presence of posttranslational modifications such as N-linked glycans or the inability of antibodies binding these epitopes to block interactions with hACE-2 or cause conformational changes that prevent receptor interactions. 59 The inability of the CPMV-317 and CPMV-362 candidates to neutralize either virus may reflect the unique properties of the displayed peptides: the small size of peptide 317 (eight residues) and/or the high pI of both peptides (9.5 and 11.8, respectively) may interfere with their ability to fold properly when displayed on the CPMV particle, thus eliciting antibodies that lack neutralizing efficacy. Notably, only CPMV-317 and CPMV-362 elicited an IgG1-predominant response, and these were also the only two candidates that did not neutralize the SARS-CoV. Interestingly, although the other epitopes 988, 1173, and 1209 show 100% sequence identity in the SARS-CoV and SARS-CoV-2 S proteins, the antibodies elicited by the CPMV-based vaccines were only able to neutralize the SARS-CoV. It is possible that the epitopes fold differently in the two viruses due to differences in the flanking residues that alter the overall conformation of the S protein. 60 However, a more likely explanation is the difference in glycosylation between the two viruses. There are 22 N-linked glycan sites on the SARS-CoV-2 S protein compared to 23 on the SARS-CoV, with 18 of the sites common to both viruses, and the glycan shield density is lower in the SARS-CoV than SARS-CoV-2 and MERS. 61 The COVID-19 pandemic highlights the need for innovation in vaccine design but also the need for more effective vaccine delivery strategies. The roll-out of mass vaccinations was burdened by the requirement of storage at ultra-low temperatures, delivery via injection thus requiring medical staff, and the requirement of a prime-boost vaccination schedule. Plant virus nanotechnologies hold potential to overcome the cold chain because these materials are stable under various environmental conditions. Toward overcoming the need for repeated injections, we evaluated the efficacy of a sustained release vaccine implant incorporating all five candidate vaccines. The five CPMV-based vaccine candidates were formulated as a pentavalent implant by hot-melt extrusion with PLGA and PEG8000. 44,51−53 Imaging studies revealed sustained release of the CPMV in vivo and trafficking of the CPMV to B-and T-cell-rich regions of the draining lymph nodes. Meanwhile, antibody titers against the CPMV carrier were 32-fold lower from the implant CPMV vs soluble injected CPMV vaccines (1:6400 vs 1:204,800, respectively; Figure  S4 ). We found that the IgG titers against the target epitopes produced using the CPMV/PLGA implants releasing the pentavalent vaccine candidates were not as high as those induced by the corresponding soluble vaccines ( Figure 5 for the implant vs Figure 3 for soluble vaccine candidates). The formulation process, which involves freezing, lyophilization, and heat extrusion, could be the underlying factor and influence the immunogenicity of the particles. DLS and TEM data were consistent with the intact CPMV being released from the polymer blends ( Figure S3) ; however, the lyophilization step resulted in loss of the RNA cargo ( Figure  S2 ), which is consistent with our previous findings. 63 CPMV genomic RNA has been identified as a TLR-7 agonist, and its absence during antigen processing can mitigate the selfadjuvant properties of CPMV vaccines. 55 The difference of the antibody titers of CPMV implants compared to prime-boost formulations requires further analysis and may reflect a 

We screened five B-cell epitopes originally identified in the convalescent sera from recovered SARS patients by displaying them on the surface of the CPMV. Three of these epitopes (peptides 988, 1173, and 1209) were found to be suitable for vaccine design. Immunization of mice using soluble formulation in a prime-boost-boost strategy elicited high antibody titers that neutralized the SARS-CoV in vitro. The neutralizing vaccine candidates (CPMV-988, CPMV-1173, and CPMV-1209) showed an early Th1-biased antibody profile (2−4 weeks) transitioning to a slightly Th2-biased profile after 6 weeks, just after the second boost. A pentavalent vaccine comprising all five peptides displayed on the CPMV was administered as a slow-release implant, and antibody titers were generated not as high as those elicited by the soluble formulation and maintained the specificity against the S protein. Sequence analysis revealed that the three epitopes (-988, -1173, and -1209) were 100% identical in the SARS- ACS Infectious Diseases pubs.acs.org/journal/aidcbc Article CoV and SARS-CoV-2, but none of the vaccine candidates were able to neutralize the SARS-CoV-2 suggesting differences in the structural context perhaps caused by conformational changes or the presence of N-linked glycans. Plant virus nanotechnologies offer high thermal stability, thus overcoming the need for cold chain storage and distribution. The technology presented here therefore offers a highly versatile vaccination platform that can be pivoted toward other diseases and applications that are not limited to infectious diseases.

■ METHODS CPMV Propagation. VNPs based on the CPMV were propagated and purified as previously described. 64 Purified VNPs were stored in 0.1 M potassium phosphate (KP) buffer (pH 7.0) at 4°C. VNP concentrations were determined by UV spectroscopy at 260 nm using the molar extinction coefficient

Antigen Characterization In Silico. We selected five Bcell epitopes 24 from the SARS-CoV-2 S protein (accession no. YP_009724390.1) as follows: 317 = KGIYQTSN, 362 = ATRFASVYAWNRKRISN, 988 = AISSVLNDILSRLDKVE, 1173 = KNHTSPDVDLGDISGIN, and 1209 = EIDRLNE-VAKNLNESLIDLQEL (Figure 1 ). We used an online peptide calculator (https://pepcalc.com/) to predict the molecular weights and isoelectric points. We determined the sequence identity compared to the SARS-CoV S protein (accession no. YP_009825051.1) using protein BLAST (https://blast.ncbi. nlm.nih.gov/).

Synthesis and Formulation of CPMV Vaccine Candidates. The five B-cell epitopes appended with an N-terminal cysteine residue a and triple glycine (GGG) linker were purchased from GenScript, with the following peptide sequences: 317 = C-GGG-KGIYQTSN, 362 = C-GGG-ATRFASVYAWNRKRISN, 988 = C-GGG-AISSVLN-DILSRLDKVE, 1173 = C-GGG-KNHTSPDVDLGDISGIN, and 1209 = C-GGG-EIDRLNEVAKNLNESLIDLQEL. Using our two-step protocol, 65 each peptide epitope was conjugated to the CPMV capsid via the heterobifunctional N-hydroxysuccinimide (NHS)-PEG 4 -maleimide linker SM-PEG 4 (Thermo Fisher Scientific) targeting the surface-exposed lysine residues. Briefly, wild-type CPMV particles (2 mg/mL in KP buffer) were reacted with a 3000-fold molar excess of the SM-PEG 4 linker at room temperature for 2 h followed by a 6000fold molar excess of each peptide overnight. The resulting vaccine candidates CPMV-317, CPMV-362, CPMV-988, CPMV-1173, and CPMV-1209 were purified using Amicon spin columns with a cutoff of 100 kDa (Sigma-Aldrich), resuspended in sterile KP buffer, and stored at 4°C.

Characterization of CPMV Vaccine Candidates. To verify peptide conjugation, 10 μg amounts of the unmodified CPMV and purified CPMV vaccine candidates (CPMV-317, CPMV-362, CPMV-988, CPMV-1173, and CPMV-1209) were compared by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) under reducing conditions on NuPAGE 4−12% Bis-Tris protein gels (Thermo Fisher Scientific) and stained with GelCode Blue Safe (Thermo Fisher Scientific). Gel images were acquired using the ProteinSimple FluorChem R imaging system, and lane density analysis with ImageJ 1.44o (http://imagej.nih.gov/ij) was used to determine the number of peptides conjugated per VNP. Particle integrity was confirmed by transmission electron microscopy (TEM) using a Tecnai Spirit G2 Bio TWIN (FEI Technologies) following 2% (w/v) uranyl acetate staining. Particle size was measured by dynamic light scattering (DLS) on a Zetasizer Nano (Malvern Instruments) at 25°C in plastic disposable cuvettes. Preparation of CPMV Implants by Hot-Melt Extrusion. Poly(lactic-co-glycolic acid) (PLGA) implants were prepared as previously described using our desktop meltprocessing system. 45,51−53 Briefly, PLGA powder with a 50:50 L/G ratio and a molecular weight of 10−15 kDa (Akina) was passed through a 45-mesh sieve (Sigma-Aldrich) for implant formulation. Lyophilized CPMV-317, CPMV-362 CPMV-988, CPMV-1173, and CPMV-1209 were mixed in equal amounts and then combined with PLGA and PEG8000 (Fisher Scientific) in the following ratio: 75% PLGA, 10% VNPs, and 15% PEG8000 (wt %). The components were mixed by vortexing, loaded into the hot melt-processing system, and heated to 70°C for 90 s. Implants were extruded at a pressure of 10 psi applied to the piston. Implants were dried and stored with desiccants until use. CPMV-Cy5 implants were processed in the same manner.

Immunization. All animal experiments were carried out in compliance with guidelines from the UC San Diego Institutional Animal Care and Use Committee. Eight-week-old male BALB/c mice (Jackson Laboratory) were kept under standard conditions with food and water provided ad libitum. Five mice were assigned per experimental group. For the subcutaneous injection of liquid formulations, each CPMV vaccine candidate was concentrated at 1 mg/mL in phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 , and 1.8 mM KH 2 PO 4 , pH 7.4), and 100 μL was injected three times (100 μg/dose) at 2-week intervals (prime + two boosts). For the implant, a single dose containing 300 μg of each vaccine candidate (mimicking the 100 μg per dose subcutaneous injection) was administered using an 18G needle (BD Sciences) behind the neck. We also administered 5 μg per dose of the free peptides (equivalent to peptides present in a 100 μg dose of the CPMV-peptide vaccine) to a control group. Retro-orbital blood was collected in lithium/heparin-treated tubes (Thomas Scientific) just before injection or implantation (week 0) and then at weeks 2, 4, 6, and 10. Plasma was collected by centrifugation at 2000g for 10 min at room temperature and was stored at −80°C.

IgG Titers against Peptides and the S Protein. Endpoint total IgG titers against each peptide epitope were determined by enzyme-linked immunosorbent assay (ELISA) in 96-well maleimide-activated plates (Thermo Fisher Scientific). Briefly, the plates were coated with 100 μL per well of each peptide (25 μg/mL in coating buffer: 0.1 M sodium phosphate, 0.15 M sodium chloride, and 10 mM EDTA, pH 7.2) overnight at 4°C. After three washes with 200 μL per well of PBS + 0.5% Tween-20 (PBST), the plates were blocked for 1 h at room temperature using 200 μL per well of 10 μg/mL L-cysteine (Sigma-Aldrich). After washing again as above, plasma from immunized animals was prepared as twofold serial dilutions in coating buffer and added to the plates. After incubation for 1 h at room temperature and another washing step, binding was detected using a horseradish peroxidase (HRP)-labeled goat antimouse IgG secondary antibody (Thermo Fisher Scientific) diluted 1:5000 in PBST (100 μL per well) for 1 h at room temperature. After a final washing step, we added 100 μL per well of a 1-Step Ultra TMB substrate (Thermo Fisher Scientific) and incubated it for 10 min before stopping the reaction with 100 μL per well of 2 M H 2 SO 4 .

The IgG titer against the SARS-CoV-2 S-protein was determined as described above but using 96-well nickelactivated plates (Thermo Fisher Scientific) coated with 200 ng of the His-tagged S protein per well (GenScript). Plasma samples were diluted five-fold in PBS, and the same secondary antibody dilution and substrate were used to develop the assay as described above. The absorbance was read at 450 nm on a Tecan microplate reader. The endpoint antibody titers were defined as the reciprocal serum dilution at which the absorbance exceeded twice the background value (blank wells without plasma samples).

Antibody Isotyping. The ELISA method set out above was adapted for antibody isotyping by testing samples from weeks 2, 4, 6, and 10 (diluted 1:1000 in coating buffer) with the following HRP-labeled secondary antibodies (Abcam) and dilutions: goat antimouse IgG1 (1:5000), IgG2a (1:1000), IgG2b (1:5000), IgG2c (1:5000), IgG3 (1:5000), IgM (1:5000), and IgE (1:1000). The IgG1/IgG2a ratio was reported for each group, and a ratio higher than 1 was considered to indicate a Th2 response.

ELISpot Assays. Briefly, 96-well ELISpot plates (Cellular Technology) were coated with a 1:166 dilution of the antimouse interferon gamma (IFN-γ) and interleukin-4 (IL-4) capture antibodies overnight at 4°C. Splenocyte suspensions collected from three mice, 2 or 10 weeks postimmunization with each CPMV vaccine candidate, were added to the plates (5 × 10 5 cells per well) following stimulation with 100 μL of the medium alone (negative control), free peptide epitopes (20 μg/mL), the unmodified CPMV (10 μg/mL), or 50 ng/mL phorbol 12-myristate 13-acetate (PMA) and 1 μg/ mL ionomycin (Sigma-Aldrich) (positive control) at 37°C and 5% CO 2 for 24 h. The plates were washed with PBST and then incubated with a 1:1000 dilution of FITC-labeled antimouse IFN-γ and a 1:666 dilution of biotin-labeled antimouse IL-4 antibodies at room temperature for 2 h. The plates were then washed with PBST and incubated at room temperature for 1 h with streptavidin-alkaline phosphatase (AP) and anti-FITC-HRP secondary antibodies (diluted 1:1000). Plates were washed with PBST and distilled water, then incubated with the AP substrate for 15 min at room temperature, washed with distilled water, and incubated with the HRP substrate for 10 min at room temperature. Plates were then rinsed with water and air-dried at room temperature overnight. Colored spots were quantified using an S6 ENTRY Analyzer (Immunospot). The splenocytes from each animal were tested in triplicate for each stimulant.

Neutralization Assays. The primary cytopathic effect assay 66 was carried out via the Preclinical Services offered by The National Institute of Allergy and Infectious Diseases (NIAID). The SARS-CoV strain Urbani and the SARS-CoV-2 strain USA_WA1/2020 were used to test neutralizing plasma. Briefly, confluent or near-confluent monolayers of Vero 76 cells were prepared in 96-well disposable microplates the day before testing. Cells were maintained in MEM (Sigma-Aldrich) supplemented with 5% fetal bovine serum (FBS) and were tested in the same medium with the FBS concentration reduced to 2% and supplemented with 50 μg/mL gentamicin. The pooled plasma samples (week 6 post-immunization) representing each CPMV vaccine candidate and the unmodified CPMV as a negative control were prepared as 10-fold serial dilutions. Five microwells were used per dilution: three for infected cultures and two for uninfected toxicity cultures. Controls consisted of six wells that were infected and not treated (virus controls) and six that were untreated and uninfected (cell controls) on every plate. Plasma samples were mixed with the virus (1:1 ratio) and incubated for 1 h at 37°C.

pubs.acs.org/journal/aidcbc Article

The growth medium was then removed from the cells, and the plasma/virus mixture was applied (0.1 mL per well). For the virus infection controls, the virus was added typically at ∼60 CCID 50 (50% cell culture infectious dose) in 0.1 mL of the medium. A medium without the virus was added to the toxicity control and cell control wells. Plates were incubated at 37°C in a 5% CO 2 incubator until a cytopathic effect CPE > 80% was observed in the virus control wells. The plates were then stained with 0.011% neutral red for ∼2 h at 37°C in a 5% CO 2 incubator. The neutral red medium was removed, and the cells were rinsed with PBS to remove the residual dye. PBS was completely removed, and the incorporated neutral red was eluted with 50% Sorensen's citrate buffer/50% ethanol for at least 30 min. The dye content in each well, proportional to the number of living cells, was quantified by spectrophotometry at 540 nm. The dye content in each set of wells was converted to a percentage of the dye present in untreated control wells and normalized against the virus control. The 50% effective concentration (EC 50 , virus inhibition) and 50% cytotoxic concentration (CC 50 , cell inhibition) were calculated by regression analysis. The CC 50 /EC 50 quotient was used to calculate the selectivity index (SI), and plasma samples with SI ≥ 10 were considered as neutralizing.

Immunofluorescence Imaging of Lymph Nodes. Cy5labeled CPMV nanoparticles were prepared for imaging studies. Fluorescent CPMV-Cy5 particles were synthesized by conjugating N-hydroxysuccinimide-activated esters of sulfo-Cy5 (NHS-sulfo-Cy5, Lumiprobe) to the CPMV capsid via the surface-exposed lysine residues. Briefly, a 1500 molar excess of NHS-sulfo-Cy5 was reacted overnight with the CPMV in 0.1 M KP buffer (pH 7.4) plus 10% (v/v) DMSO and a protein concentration of 2 mg/mL. Following the reaction, Cy5conjugated particles were purified from the excess unreacted dye by ultracentrifugation (112,000g, 1 h). The pellet was then resuspended in 0.1 M KP buffer. The CPMV concentration and the number of dye molecules per capsid were determined by UV/vis spectrophotometry using the CPMV extinction coefficient (ε CPMV ) = 8.1 mL mg −1 cm −1 at 260 nm and the NHS-sulfo-Cy5 specific molar extinction coefficient (ε sulfo-Cy5 ) = 27,1000 L mol −1 cm −1 at 647 nm. CPMV-Cy5 implants were prepared as described above and introduced subcutaneously via an 18G needle by pushing the implant out of the needle using a sterilized stainless-steel wire (0.51 mm diameter). Fluorescence images were acquired on an IVIS 200 imaging system at different time points and were analyzed using Living Image v3.0. To determine the fate of VNPs in the implant, cervical, axillary, and inguinal lymph nodes were collected from the mice in 10% formalin-buffered solution (Sigma-Aldrich) on days 0, 7, 14, 21, 28, and 35 after CPMV-Cy5 implant administration. Two mice were euthanized on each day. The lymph nodes were embedded in paraffin, and tissue sections were stained for the cell surface markers CD11c (LS Bio, 1:25), CD45R (Abcam, 1:50), and CD4 (Abcam, 1:100). A donkey antirabbit IgG secondary antibody conjugated to AlexaFluor 488 (Invitrogen, 1:500) was used to detect primary antibody binding. The sections were imaged on a Nikon A1R confocal microscope with an X20, 0.75 numerical aperture dry objective. Portions from two lymph node sections were imaged for each cell marker, and the corresponding channels were superimposed by aligning the DAPI signal to form composite images using Nikon Analysis software.

Statistical Analysis. Data are presented as the mean ± SEM. Single comparisons based on an unpaired, two-tailed t-test were carried out using SPSS Statistics software or GraphPad Prism 6. Differences were considered significant at p < 0.05. The number of replicates is described for each experiment.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsinfecdis.1c00410.

Ig isotype profile ( Figure S1 ), characterization of CPMV-Cy5 ( Figure S2 ) and CPMV released from PLGA implants (Figure S3 ), and IgG titers against CPMV comparing soluble vs implant administration ( Figure S4) 

A case for the ancient origin of coronaviruses

Coronavirus infections in hospitalized pediatric patients with acute respiratory tract disease

Middle East respiratory syndrome

A novel coronavirus outbreak of global health concern

Online resource

London School of Hygiene & Tropical Medicine (2021) COVID-19 vaccine tracker

Potential applications of plant biotechnology against SARS-CoV-2

Bat-to-human: spike features determining 'host jump' of coronaviruses SARS-CoV, MERS-CoV, and beyond

Structure of SARS coronavirus spike receptor-binding domain complexed with receptor

Molecular basis of binding between novel human coronavirus MERS-CoV and its receptor CD26

Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2

Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus

SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor

Neutralizing antibodies against SARS-CoV-2 and other human coronaviruses

Learning from the past: development of safe and effective COVID-19 vaccines

Subunit vaccines against emerging pathogenic human coronaviruses. Front

A universal design of betacoronavirus vaccines against COVID-19, MERS, and SARS

Phase 1 randomized trial of a plant-derived virus-like particle vaccine for COVID-19

Characteristics of SARS-CoV-2 and COVID-19

Preliminary identification of potential vaccine targets for the COVID-19 coronavirus (SARS-CoV-2) based on SARS-CoV immunological studies

COVID-19 vaccine development and a potential nanomaterial path forward

Cross-reactive neutralization of SARS-CoV-2 by serum antibodies from recovered SARS patients and immunized animals

Immunization with the receptor-binding domain of SARS-CoV-2 elicits antibodies cross-neutralizing SARS-CoV-2 and SARS-CoV without antibody-dependent enhancement

XXXX, XXX, XXX−XXX binding domain (RBD)-based mRNA vaccine against SARS-CoV-2

Identification of SARS-CoV RBD-targeting monoclonal antibodies with cross-reactive or neutralizing activity against SARS-CoV-2

Cross-neutralization of SARS-CoV-2 by a human monoclonal SARS-CoV antibody

Broad-spectrum anti-coronavirus vaccines and therapeutics to combat the current COVID-19 pandemic and future coronavirus disease outbreaks

Vaccines that can protect against many coronaviruses could prevent another pandemic

Epitopes for a 2019-nCoV vaccine

A sequence homology and bioinformatic approach can predict candidate targets for immune responses to SARS-CoV-2

Immunoinformatics-aided identification of T cell and B cell epitopes in the surface glycoprotein of 2019-nCoV

SARS corona virus peptides recognized by antibodies in the sera of convalescent cases

Identification of immunodominant sites on the spike protein of severe acute respiratory syndrome (SARS) coronavirus: implication for developing SARS diagnostics and vaccines

Immunodominant SARS coronavirus epitopes in humans elicited both enhancing and neutralizing effects on infection in non-human primates

Structure-based engineering of an icosahedral virus for nanomedicine and nanotechnology

Poly(lactic-co-glycolic acid) devices: production and applications for sustained protein delivery

Milling solid proteins to enhance activity after melt-encapsulation

Virus-like particle laden biodegradable implants prepared via melt processing

Polymer structure and conformation alter immune recognition of virus-like particle-polymer conjugates

PEGylation to improve protein stability during melt processing

The spike glycoprotein of the new coronavirus 2019-NCoV contains a furin-like cleavage site absent in CoV of the same clade

Human SARS CoV-2 spike protein mutations

uniquely reactive Lys residues on cowpea mosaic virus

Heterologous prime-boost enhances the antitumor immune response elicited by plant-virus-based cancer vaccine

Design and fabrication of a lowcost pilot-scale melt-processing system

A scalable manufacturing approach to single dose vaccination against HPV

A singledose, implant-based, trivalent virus-like particle vaccine against "cholesterol checkpoint" proteins

Chemical addressability of potato virus X for its applications in bio/ nanotechnology

Cowpea mosaic virus stimulates antitumor immunity through recognition by multiple MYD88-dependent toll-like receptors

The 3Ds in virus-like particle based-vaccines: design, delivery and dynamics

Vaccine delivery: a matter of size, geometry, kinetics and molecular patterns

Plant viral nanoparticles-based HER2 vaccine: immune response influenced by differential transport, localization and cellular interactions of particulate carriers

Neutralization of SARS-CoV-2 by Destruction of the Prefusion Spike

The Presence of Prolines in the Flanking Region of an Immunodominant HIV-2 Gag Epitope Influences the Quality and Quantity of the Epitope Generated

Vulnerabilities in coronavirus glycan shields despite extensive glycosylation

Freeze-drying to produce efficacious CPMV virus-like particles

Viral nanoparticles for in vivo tumor imaging

A viral nanoparticle cancer vaccine delays tumor progression and prolongs survival in a HER2+ tumor mouse model

A simple method of estimating fifty percent endpoints