Unglycosylated Soluble SARS-CoV-2 Receptor Binding Domain (RBD) Produced in E. coli Combined with the Army Liposomal Formulation Containing QS21 (ALFQ) Elicits Neutralizing Antibodies against Mismatched Variants

The emergence of novel potentially pandemic pathogens necessitates the rapid manufacture and deployment of effective, stable, and locally manufacturable vaccines on a global scale. In this study, the ability of the Escherichia coli expression system to produce the receptor binding domain (RBD) of the SARS-CoV-2 spike protein was evaluated. The RBD of the original Wuhan-Hu1 variant and of the Alpha and Beta variants of concern (VoC) were expressed in E. coli, and their biochemical and immunological profiles were compared to RBD produced in mammalian cells. The E. coli-produced RBD variants recapitulated the structural character of mammalian-expressed RBD and bound to human angiotensin converting enzyme (ACE2) receptor and a panel of neutralizing SARS-CoV-2 monoclonal antibodies. A pilot vaccination in mice with bacterial RBDs formulated with a novel liposomal adjuvant, Army Liposomal Formulation containing QS21 (ALFQ), induced polyclonal antibodies that inhibited RBD association to ACE2 in vitro and potently neutralized homologous and heterologous SARS-CoV-2 pseudoviruses. Although all vaccines induced neutralization of the non-vaccine Delta variant, only the Beta RBD vaccine produced in E. coli and mammalian cells effectively neutralized the Omicron BA.1 pseudovirus. These outcomes warrant further exploration of E. coli as an expression platform for non-glycosylated, soluble immunogens for future rapid response to emerging pandemic pathogens.


Introduction
The zoonotic transmission of betacoronaviruses such as severe acute respiratory syndrome coronavirus 1 (SARS-CoV-1) [1], Middle East respiratory syndrome [2], and SARS-CoV-2 have caused significant global morbidity and mortality [3]. Three years after the first documentation of SARS-CoV-2 transmission in humans, more than 600 million cases and 6.4 million mortalities have been attributed to COVID-19 [4]. Typical of its subfamily, SARS-CoV-2 utilizes a homotrimeric spike (S) protein to interact with the human angiotensin

E. coli expression of RBD
The nucleotide sequences corresponding to a.a. 319-542 of S protein from the original Wuhan-Hu1, Alpha (B.

E. coli Expression of RBD
The nucleotide sequences corresponding to a.a. 319-542 of S protein from the original Wuhan-Hu1, Alpha (B.1.1.7), and Beta (B.1.351) variants of SARS-CoV-2 with a N-terminal 6x His tag were codon optimized for high level bacterial expression, commercially synthesized, and cloned into the pD451-SR vector (Atum, Newark, CA, USA). Genes were expressed in BL21(DE) 3 E. coli cells (Novagen, Madison, WI, USA). Expression was initiated by a 100-mL starter lysogeny broth culture that was grown overnight at 37 • C. The culture was expanded the following day to a 1 L-scale using terrific broth; both cultures contained 50 µg/mL kanamycin. The culture was grown at 37 • C to an OD 600 = 0.8 and induced with 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) for 3 h. Biomass was recovered by centrifugation at 4000× g for 20 min at 4 • C. This paste was frozen at −80 • C until purification.

E. coli RBD Purification and Refolding
Cell paste in~5 g batches was resuspended in resuspension buffer (RB) containing 20 mM Tris, 150 mM NaCl, 5 mM β-mercaptoethanol (pH 7.0). Cells were lysed by microfluidization using the M-110Y Microfluidizer (Microfluidics, Newton, MA) and centrifuged at 15,000× g for 20 min at 4 • C. The inclusion body pellet was washed with RB containing 1.5% Triton X-100 for endotoxin reduction, centrifuged at 15,000× g for 20 min at 4 • C, and solubilized in 25 to 50 mL equilibration buffer (EqB), which was RB that contained 7 M urea (pH 7.0). The protein solution was centrifuged again as above and the resultant supernatant was applied to a 5-mL Ni-NTA Superflow column (Qiagen, Germantown, MD, USA) mounted on an AKTA Purifier 100 (Cytiva, Marlborough, MA, USA). The column was pre-equilibrated with 5 column volumes (CV) of distilled water and 10 CV of EqB. After protein loading, the Ni-NTA column was washed with 10 CV of wash buffer containing 20 mM Tris, 300 mM NaCl, 40 mM imidazole, 5 mM β-mercaptoethanol, 7 M urea, 1% CHAPS (pH 7.0), then washed again with 10 CV of EqB. The protein was eluted using 5 CV of elution buffer (EB) containing 20 mM Tris, 150 mM NaCl, 250 mM imidazole, 5 mM β-mercaptoethanol, 7 M urea (pH 7.0). The eluate was diluted 4-fold in EqB to~60 mM imidazole, then passed over a 5-mL Q Sepharose ® Fast Flow column (Cytiva) to reduce endotoxin. The flow-through was collected and dialyzed overnight at 4 • C in 20 mM Tris, 150 mM NaCl, 5% glycerol (pH 7.0), in a Slide-A-Lyzer™ G2 Dialysis Cassette, 10K MWCO (Thermo Fisher Scientific, Waltham, MA, USA). The purity of the final product was assessed by SDS-PAGE with Coomassie Blue staining and densitometric analysis using ImageJ [36]. The buffer-exchanged protein was centrifuged at 10,000× g for 15 min and concentrated to 1 mg/mL before aliquoting and storage at −80 • C.

Mammalian Expression and Purification
The Beta variant of RBD with a N-terminal 6x His tag was codon optimized and cloned into the proprietary pD2610-v5 vector (Atum) for high-level mammalian expression. For expression, transient transfection of the RBD-encoding plasmid into Expi293F™ cells was achieved using the ExpiFectamine™ 293 Transfection Kit per the manufacturer's protocols (Thermo Fisher Scientific). Briefly, plasmid DNA purified using the EndoFree Plasmid Maxi Kit (Qiagen) was diluted in Opti-MEM I medium (1 µg/mL of final culture volume), sterile filtered, and combined with ExpiFectamine™ 293 reagent diluted in Opti-MEM I medium; this complex was incubated for 10 to 20 min at room temperature, then added to the cells diluted to a density of 3 × 10 6 cells/mL. The culture was incubated in a 37 • C incubator with 85% humidity, 8% CO 2 on an orbital shaker at 125 rpm for 18 to 22 h. Enhancer mixture was added according to manufacturer's recommended ratio, and the culture was incubated for another 5 to 6 days. The culture supernatant was harvested and purified over a 2-mL Ni-NTA Superflow bed (Qiagen) in an Econo-Pac ® chromatography column (Bio-Rad, Hercules, CA, USA) equilibrated twice with 5 CV of 20 mM Tris, 150 mM NaCl, 2.5 mM β-mercaptoethanol (pH 7.0), washed with 5 CV of 20 mM Tris, 300 mM NaCl, 20 mM imidazole, 2.5 mM β-mercaptoethanol (pH 7.0), and eluted using 20 mM Tris, 150 mM NaCl, 250 mM imidazole, 2.5 mM β-mercaptoethanol (pH 7.0). The protein was finally dialyzed into phosphate-buffered saline (pH 7.4).

Endotoxin Content Quantitation
The endotoxin content of RBDs purified from E. coli was determined using a kinetic Limulus amebocyte lysate assay (Associates of Cape Cod, East Falmouth, MA, USA). Samples were diluted 1:10 in endotoxin-free water and serially diluted 11-fold down a certified endotoxin-free 96-well microplate (Associates of Cape Cod). Endotoxin standards ranging from 2 EU/mL to 0.125 EU/mL were prepared in endotoxin-free water using the Control Standard Endotoxin reagent (Associates of Cape Cod). Pyrochrome reagent was added to the sample in a 1:1 ratio and plates were read on a Synergy 4 microplate reader (BioTek, Winooski, VT, USA) at 30 • C for 1 h. The OD 405 value was recorded every 2 min and subtracted from a blank well containing 0 EU/mL endotoxin. Endotoxin content (EU/mL and EU/µg of protein) was calculated per the manufacturer's instructions.

Analytical Size Exclusion Chromatography
Analytical size exclusion chromatography was performed using the Acclaim™ SEC-1000 Size Exclusion Chromatography HPLC Column (Thermo Fisher Scientific) mounted on the Waters™ e2695 Separations Module (Waters, Milford, MA, USA) per the manufacturer's operating guidelines. The running buffer consisted of 20 mM Tris, 150 mM NaCl, 5% glycerol (pH 7.0), and the flow rate was maintained at 0.5 mL/min. The elution time was calibrated using a protein molecular size standard. Approximately 50 µg of RBD was injected per run.

Mouse Immunizations
BALB/c mice (n = 4 per group) were intramuscularly immunized three times at 2-week intervals with RBD antigen formulated in 50 µL of ALFQ containing 20 µg 3D=PHAD ® and 10 µg QS-21 [39]. Three groups of mice received 10 µg of one of the three RBD variants (E. coli), a fourth group was given a mixture of all three proteins (~3.3 µg each), and a fifth group 10 µg of glycosylated Beta variant RBD produced in mammalian cells. All formulations were rotated on a rocker for 1 h and administered intramuscularly into alternating thigh muscles. Blood was collected by nicking the tail vein two weeks after the third vaccination (2WP3), at week 6 of the study, and six weeks after the third dose (6WP3) at week 10.

Human ACE2-Fc and Monoclonal Antibody CR3022 Binding Assay
Biosensors used for biolayer interferometry (BLI) were hydrated in 1x PBS prior to the assay run. All assay steps were performed at 30 • C with agitation set at 1000 rpm in the Octet RED96 instrument (FortéBio, Fremont, CA, USA). Pre-hydrated AHC biosensors (FortéBio) were loaded with 30 µg/mL of ACE2-Fc for 120 s. Varying dilutions of RBD were associated for 180 s and dissociated for 300 s in 1x PBS to obtain binding kinetics. For mAb CR3022 binding, pre-hydrated AHC biosensors were loaded with 30 µg/mL of CR3022 for 120 s and varying dilutions of RBD were associated for 180 s and dissociated for 300 s in 1x PBS to obtain binding kinetics.

ACE2 Binding Inhibition Assay
HIS1K sensors (FortéBio) pre-hydrated in AB were loaded with 5 µg/mL of individual RBD variants for 480 s. The 6WP3 sera pool at 1:50, 1:500, 1:5000, 1:10,000 and 1:20,000 dilutions were loaded for 120 s, then associated with 400 nM ACE2 for 120 s. Dissociation was monitored for 300 s. Percent inhibition of ACE2 binding by immune sera was calculated with respect to a naïve serum control pool (Supplementary Figure S1).

ELISA
Antigenicity ELISAs were carried out using 100 ng/well RBD variant coated onto a 96-well 4HBX Immulon (Thermo Fisher Scientific) plate at 4 • C overnight. Plates were washed thrice with 1x PBS containing 0.05% Tween-20, blocked for 1.5 h with 5% casein, and incubated with a 1:3 serial dilution of the monoclonal antibodies starting at 1 µg/mL for 2 h. The plates were washed, then incubated for 1 h with horseradish peroxidase-conjugated goat anti-human IgG diluted 1:4000 (Southern Biotech). Plates were washed and developed using 100 µL/well KPL substrate (SeraCare, Gaithersburg, MD, USA). After 1 h, 10 µL of 20% SDS stop solution was added and plates were read at 415 nm. The OD = 1 titers were determined from the readings on the Synergy 4 microplate reader (BioTek) using 4-parameter curve fitting. For immunogenicity ELISAs, 50 ng of each RBD variant was coated, and 1:3 serial dilutions of mouse sera starting at 1:300 dilution was incubated for 1 h. Goat anti-mouse IgG (H+L) horseradish peroxidase-conjugated (Southern Biotech) was used as the secondary antibody and the plates were developed as above.

Pseudovirus Neutralization Assay
The neutralization assay for SARS-CoV-2 variants (Wuhan-Hu1, B.1.1.7, B.1.351, B.1.617.2 and B.1.1.529) was performed as previously described [40][41][42]. Infectivity and neutralization titers were determined using ACE2-expressing HEK293 target cells (Integral Molecular, Philadelphia, PA, USA). Test sera were diluted 1:40 in growth medium and serially diluted; 25 µL/well was then added to a white 96-well plate. An equal volume of diluted pseudovirus was added to each well, and plates were incubated for 1 h at 37 • C. Target cells were added to each well (40,000 cells/well), then plates were incubated for an additional 48 h. Relative light units were measured with the EnVision Multimode Plate Reader (Perkin-Elmer, Waltham, MA, USA) using the Bright-Glo Luciferase Assay System (Promega, Madison, WI). Neutralization dose-response curves were fitted by nonlinear regression using the LabKey Server. Final titers are reported as the reciprocal of the serum dilution necessary to achieve 50% inhibition of SARS-CoV-2 infectivity (ID 50 ).

Statistical Analysis
A proof-of-concept vaccination study was conducted (n = 4). Both descriptive and summary statistics were applied. Differences between ELISA and neutralization titers of vaccine groups were compared using ANOVA followed by Tukey's comparison test. p values < 0.05 were considered significant. Individual data and geometric means were plotted with GraphPad Prism 9.1 software.

Wuhan-Hu1 and VoC RBD Expression in E. coli and Mammalian Cells
Wuhan-Hu1 RBD protein was expressed at a relatively high level following IPTG induction ( Figure 1B, lanes 1 & 2). The majority of protein was localized in the inclusion body pellet ( Figure 1B, lanes 3 & 4) and required 7 M urea for solubilization ( Figure 1B,  lane 7). The protein was purified to near homogeneity using Ni-NTA chromatography under reduced and denatured conditions ( Figure 1B, lane 10) and refolded by one-step dialysis ( Figure 1B, lane 11). The refolded Wuhan-Hu1 RBD was compared to mammalian cellexpressed Wuhan-Hu1 reference standard using a BLI-based binding assay (Supplementary Figures S2 and S3). The association of bacterial Wuhan-Hu1 to hACE2-Fc and to a panel of SARS-CoV-2 mAbs (P2B-2F6, CC12.1, CC12.156, [43,44] and WRAIR-2125 [8]) and the SARS-CoV-1 mAb CR3022 [37] was comparable to the reference standard (Supplementary Figure S2). The dissociation constants of E. coli-produced Wuhan-Hu1 RBD's interaction with hACE2-Fc and mAb CR3022 were 3.5 nM and 6.1 nM, respectively, which are close to previously reported values for eukaryotic-expressed RBDs (4.7 nM for ACE2 and  Figure S3). These data suggest that E. coli may be used to express a biologically active form of Wuhan-Hu1 RBD protein. In late 2020, as the Alpha and Beta VoCs of SARS-CoV-2 emerged [21][22][23], we applied this expression and purification protocol to express these two additional RBD variants. Alpha and Beta SARS-CoV-2 RBDs showed similar purification ( Figure 1C,D) and refolding profiles as Wuhan-Hu1. Product yields per gram wet cell paste were~0.5 mg,~0.2 mg, and~0.15 mg RBD for Wuhan-Hu1, Alpha, and Beta variants, respectively. The endotoxin content of products ranged from 0.4 to 4 EU/µg. For immunological comparison we expressed a fully glycosylated version of the Beta RBD variant in Expi293F™ mammalian cells [30]. The mammalian product was purified by one-step Ni-NTA chromatography ( Figure 2). cell-expressed Wuhan-Hu1 reference standard using a BLI-based binding assay (Supplementary Figures S2 and S3). The association of bacterial Wuhan-Hu1 to hACE2-Fc and to a panel of SARS-CoV-2 mAbs (P2B-2F6, CC12.1, CC12.156, [43,44] and WRAIR-2125 [8]) and the SARS-CoV-1 mAb CR3022 [37] was comparable to the reference standard (Supplementary Figure S2). The dissociation constants of E. coli-produced Wuhan-Hu1 RBD's interaction with hACE2-Fc and mAb CR3022 were 3.5 nM and 6.1 nM, respectively, which are close to previously reported values for eukaryotic-expressed RBDs (4.7 nM for ACE2 and 6.2-6.3 nM for CR3022) [45,46] (Supplementary Figure S3). These data suggest that E. coli may be used to express a biologically active form of Wuhan-Hu1 RBD protein. In late 2020, as the Alpha and Beta VoCs of SARS-CoV-2 emerged [21][22][23], we applied this expression and purification protocol to express these two additional RBD variants. Alpha and Beta SARS-CoV-2 RBDs showed similar purification ( Figure 1C,D) and refolding profiles as Wuhan-Hu1. Product yields per gram wet cell paste were ~0.5 mg, ~0.2 mg, and ~0.15 mg RBD for Wuhan-Hu1, Alpha, and Beta variants, respectively. The endotoxin content of products ranged from 0.4 to 4 EU/μg. For immunological comparison we expressed a fully glycosylated version of the Beta RBD variant in Expi293F™ mammalian cells [30]. The mammalian product was purified by one-step Ni-NTA chromatography (Figure 2).

Biochemical Characterization of RBD Variants Expressed in E. coli
Prior to immunological studies, a batch of E. coli-expressed RBD variants were compared to the mammalian-expressed RBD for biological and functional profiling. The 3 bacterial proteins showed ≥90% purity by densitometry, while the purity of the mammalian product exceeded 95%. A minor band corresponding to the~50 kDa dimer of RBD was observed ( Figure 1B-D and Figure 2A). Additionally, the mammalian control product migrated slower than the unglycosylated E. coli products due to its glycan decoration [30] ( Figure 2A). According to BLI analysis, E. coli RBD variants had equivalent association and dissociation kinetics to the hACE2-Fc receptor as those of the mammalian-produced Beta RBD ( Figure 2B). Dot blot of the E. coli-produced RBDs with a conformationally dependent SARS-CoV-1 mAb CR3022 provided additional confirmatory demonstration of proper folding ( Figure 2C) [38]. A negative control protein, the full-length circumsporozoite protein of Plasmodium falciparum, did not react with mAb CR3022. Only trace reactivity to the mammalian-expressed RBD was observed. Analytical size exclusion chromatography indicated that the mammalian-and E. coli-expressed RBDs were homogeneous and eluted as a monomer peak around 12 min ( Figure 2D).

Antigenicity of the RBD Variants
Recombinant RBDs from E. coli were next tested for the presence of epitopes corresponding to potently neutralizing Wuhan-Hu1-specific SARS-CoV-2 human mAbs by BLI and ELISA [8] (Figure 3). Wuhan-Hu1 RBD produced in E. coli reacted with all 7 mAbs specific to different sites on the RBD (Figure 3A,C). The Alpha variant reacted broadly with the queried mAbs apart from WRAIR-2123, to which it displayed attenuated reactivity due to mutations in the target epitope [8]. E. coliand mammalian-expressed Beta RBD reacted with mAbs WRAIR-2057, -2063, -2125, and -2151, but not mAbs WRAIR-2123, -2165 and -2173, which was also consistent with previous binding and neutralization analyses [8]. These observations suggest that E. coli and mammalian RBDs contain both variant-specific and cross-reactive mAb epitopes whose recognition is not impacted by protein glycosylation. The above-described trends in mAb reactivity observed by BLI were recapitulated by ELISA ( Figure 3B).

Immunogenicity of RBD Variants
Mice were vaccinated with three doses of RBD proteins formulated in a liposomal adjuvant containing QS21 (ALFQ). Sera were analyzed by ELISA and a pseudovirus neutralization assay ( Figure 4A). Vaccine groups included monovalent formulations of E. coli-derived Wuhan-Hu1, Alpha, or Beta RBD, a trivalent mixture of the three E. coli-derived RBDs (3.3 µg each), and the monovalent mammalian-expressed Beta RBD control. Sera collected at week 6 (2 weeks post-third vaccination, or 2WP3) and at week 10 (6WP3) were analyzed by ELISA ( Figure 4B,C). Pre-immune serum dilutions failed to achieve OD=1 even at the lowest dilution (not plotted). In contrast, the 2WP3 GM titers for the monovalent and trivalent RBD vaccines exceeded 10,000 ( Figure 4B). E. coli-derived Wuhan-Hu1 had the lowest immunogenicity, while the trivalent vaccine showed similarly elevated titers across plate antigens, indicative of the induction of broadly reactive antibodies. E. coli-derived Beta variant titers were comparable to those of the mammalian-expressed Beta variant, suggesting that E. coli was able to produce highly immunogenic viral antigens. By week 10, the titers showed a median 2-to 3-fold drop in magnitude ( Figure 4C). Sera from six weeks post-third immunization were analyzed for in vitro inhibition of RBD binding to ACE-2 receptor ( Figure 4D). Relative to the pre-immune controls (0% inhibition; not plotted), immune mouse serum pools from RBD-vaccinated mice inhibited the binding of ACE2 to the homologous RBD. Although inhibition by Wuhan-Hu1 was slightly lower, as expected due to its lower titers,~100% inhibition was achieved at 1:50 serum dilution for all other vaccines.

Pseudovirus Neutralization
Sera from mice vaccinated with E. coli-derived Wuhan-Hu1, Alpha, Beta, trivalent and mammalian-expressed Beta RBD were tested for inhibition of homologous and heterologous pseudoviruses ( Figure 5). Because 50% neutralization (ID 50 ) titers for pre-immune sera were <1:100 (not plotted), a geometric mean neutralization titer above 1:100 was considered positive. At 2WP3, E. coliand mammalian-expressed Beta RBDs were compared for immunological similarity. Although the E. coli product showed lower Alpha pseudovirus neutralization, there were no significant differences in ID 50 of these two vaccines across all five pseudoviruses ( Figure 5, top panels). Despite one mouse in the  1000). Surprisingly, the monovalent Beta RBDs also outperformed the trivalent formulation. Neutralization patterns at 6WP3 showed a similar trend as those of the 2WP3 timepoint, with a median 1.5 to 2-fold drop in individual ID 50 titers ( Figure 5, bottom panels).

Immunogenicity of RBD Variants
Mice were vaccinated with three doses of RBD proteins formulated in a liposomal adjuvant containing QS21 (ALFQ). Sera were analyzed by ELISA and a pseudovirus neutralization assay ( Figure 4A). Vaccine groups included monovalent formulations of E. coliderived Wuhan-Hu1, Alpha, or Beta RBD, a trivalent mixture of the three E. coli-derived RBDs (3.3 µg each), and the monovalent mammalian-expressed Beta RBD control. Sera collected at week 6 (2 weeks post-third vaccination, or 2WP3) and at week 10 (6WP3) were analyzed by ELISA ( Figure 4B,C). Pre-immune serum dilutions failed to achieve OD=1 even at the lowest dilution (not plotted). In contrast, the 2WP3 GM titers for the monovalent and trivalent RBD vaccines exceeded 10,000 ( Figure 4B). E. coli-derived Wuhan-Hu1 had the lowest immunogenicity, while the trivalent vaccine showed similarly elevated titers across plate antigens, indicative of the induction of broadly reactive antibodies. E. inhibition; not plotted), immune mouse serum pools from RBD-vaccinated mice inhibited the binding of ACE2 to the homologous RBD. Although inhibition by Wuhan-Hu1 was slightly lower, as expected due to its lower titers, ~100% inhibition was achieved at 1:50 serum dilution for all other vaccines.

Pseudovirus Neutralization
Sera from mice vaccinated with E. coli-derived Wuhan-Hu1, Alpha, Beta, trivalent and mammalian-expressed Beta RBD were tested for inhibition of homologous and heterologous pseudoviruses ( Figure 5). Because 50% neutralization (ID50) titers for pre-immune sera were <1:100 (not plotted), a geometric mean neutralization titer above 1:100 was considered positive. At 2WP3, E. coli-and mammalian-expressed Beta RBDs were compared for immunological similarity. Although the E. coli product showed lower Alpha pseudovirus neutralization, there were no significant differences in ID50 of these two vaccines across all five pseudoviruses ( Figure 5, top panels). Despite one mouse in the E. coli Wuhan-Hu1 group that failed to neutralize, the geometric mean ID50 of all vaccine formulations indicated positive neutralization of the vaccine-matched Wuhan-Hu1, Alpha and Beta pseudoviruses. Against the vaccine mismatched Delta peudovirus, all vaccines showed inhibitory responses (ID50 > 1000), with the E. coli-derived Wuhan-Hu1 exceeding neutralization by the Beta RBD expressed in mammalian cells. Against the Omicron BA.1 pseudovirus, only the E. coli and mammalian Beta RBD showed neutralization (ID50 > 1000). Surprisingly, the monovalent Beta RBDs also outperformed the trivalent Coating Antigen

Titer at OD=1
Alpha (E. coli)  formulation. Neutralization patterns at 6WP3 showed a similar trend as those of the 2WP3 timepoint, with a median 1.5 to 2-fold drop in individual ID50 titers ( Figure 5, bottom panels).

Discussion
Prokaryotic expression of SARS-  [53], and a.a. 519-541 [31]) further evinced the utility of this platform for coronaviral immunogen production. Across studies, the solubility of RBD in the E. coli cytosol was observed to be very poor, which necessitated extraction using a chaotropic agent followed by in vitro refolding. The reducing environment of the E. coli cytoplasm and inability of E. coli to chaperone correct disulfide bond formation [31] likely accounts for the relatively low yields observed across studies. Despite this, refolded RBDs have demonstrated binding to SARS-CoV-2 antibodies [49] and the human ACE2 receptor [31,50], and their overall structure was similar to mammalian-derived products [31,53]. E. coli RBD vaccination in small animals with multiple adjuvants, including Al(OH)3 + CpG ODN [51], Sigma Adjuvant System (Sigma-Aldrich) [47], TiterMax Gold (Sigma-Aldrich) [50], Alum [28], and Saponin [54], have shown high immunogenicity and the induction of neutralizing antibodies. Soluble proteins that are generally considered weak immunogens require formulation with potent adjuvants [55]. The ALFQ adjuvant used was recently employed in a soluble protein vaccine against malaria [33]; our data provides additional confirmation of the utility of this adjuvant for soluble immunogens. Overall, an accumulating body of evidence suggests that E. coli-produced viral antigens can match the vaccine potential of equivalent mammalian cell products. Additionally, the absence of surface glycans on

Discussion
Prokaryotic expression of SARS-CoV-1 RBD (a.a. 318-510 [ [53], and a.a. 519-541 [31]) further evinced the utility of this platform for coronaviral immunogen production. Across studies, the solubility of RBD in the E. coli cytosol was observed to be very poor, which necessitated extraction using a chaotropic agent followed by in vitro refolding. The reducing environment of the E. coli cytoplasm and inability of E. coli to chaperone correct disulfide bond formation [31] likely accounts for the relatively low yields observed across studies. Despite this, refolded RBDs have demonstrated binding to SARS-CoV-2 antibodies [49] and the human ACE2 receptor [31,50], and their overall structure was similar to mammalian-derived products [31,53]. E. coli RBD vaccination in small animals with multiple adjuvants, including Al(OH) 3 + CpG ODN [51], Sigma Adjuvant System (Sigma-Aldrich) [47], TiterMax Gold (Sigma-Aldrich) [50], Alum [28], and Saponin [54], have shown high immunogenicity and the induction of neutralizing antibodies. Soluble proteins that are generally considered weak immunogens require formulation with potent adjuvants [55]. The ALFQ adjuvant used was recently employed in a soluble protein vaccine against malaria [33]; our data provides additional confirmation of the utility of this adjuvant for soluble immunogens. Overall, an accumulating body of evidence suggests that E. coli-produced viral antigens can match the vaccine potential of equivalent mammalian cell products. Additionally, the absence of surface glycans on E. coli-produced vaccines may reduce the risk of developing autoreactive antibodies against self-carbohydrate moieties [28]. As novel variants with the capacity to escape extant vaccines continually emerge, E. coli could provide a highly cost-effective platform for response against SARS-CoV-2 and other viral pathogens like influenza [56,57] and dengue fever [58][59][60].
ID 50 titers >100 have been generally associated with protective responses in humans [61,62]. Evaluation of approved human vaccines in mice have reported ID 50 titers in the 1000-10,000 range (Moderna's mRNA-1273 vaccine [63] and AstraZeneca's adenoviral vectored vaccine AZD1222 [64]). While only head-to-head comparison of vaccines can truly be definitive, we report neutralization titers similar to those reported for these licensed SARS-CoV-2 vaccines. Cross-neutralization of heterologous variants is another key requirement for next-generation coronavirus vaccines. The capacity of mismatched SARS-CoV-2 variants to escape monovalent vaccines has led to the development of heterologous prime-boost vaccine regimens [65,66]. In the present study, monovalent Wuhan Hu-1, Alpha, and the trivalent immunogens all failed to neutralize the Omicron BA.1 pseudovirus. However, E. coli-expressed and mammalian-expressed Beta variant antibodies cross-neutralized Omicron BA.1. This observation aligns with prior reports that infection or vaccination with the Beta variant induces potent, cross-neutralizing antibodies [25,67]. Possibly, the Beta variant RBD expressed in E. coli favorably presents conserved epitopes to the immune system due to the absence of glycosylation.
Trimeric S protein, virus-like particles [41], nanoparticles [68], and T-cell epitopeenriched bacterial carriers [52] have been shown to effective vaccine platforms against SARS-CoV-2. Our data suggests that soluble, prokaryotic-expressed proteins may also be effective in eliciting broad immunity. Along these lines, a soluble RBD protein-based SARS-CoV-2 vaccine is currently being evaluated for efficacy in India [69]. The purification and refolding methodology described here translated across three variants of concern; while the process needs to be further optimized, a similar method may be useful for the manufacture of future SARS-CoV-2 variant RBD vaccines.

Conclusions
A bacterial expressed viral immunogen was found to be equivalent to a mammalianexpressed immunogen across several evaluative criteria. The expression and purification protocol described here allowed for the production of several SARS-CoV-2 variants. Our data warrants further exploration of E. coli expression system for producing low-cost viral vaccines in resource-poor countries.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/vaccines11010042/s1. Supplementary Figure S1: Inhibition of recombinant RBD binding to ACE2 assessed by BLI using immune and pre-immune serum. Shown is a typical sensorgram where the binding of recombinant RBD (homologous variant) was inhibited using 1:50 dilution of immune serum. Full association to ACE2 by RBD was observed in the presence of pre-immune serum; Supplementary Figure S2: Association of E. coli-derived (A) and mammalianderived (B) Wuhan-Hu1 RBD to the ACE-2 receptor and a suite of anti-SARS-CoV antibodies as determined by BLI. Both manifest similar reactivity profiles. Supplementary Figure S3: Kinetics of E. coli-derived Wuhan-Hu1 RBD association to human ACE-2 receptor (A) and the anti-SARS-CoV-1 mAb CR3022 (B) as determined by BLI. The antigen produced from E. coli displays nanomolar affinity to both targets, which is in good agreement with the values reported for mammalian RBD [45,46].  Welfare Assurance and in compliance with the Animal Welfare Act and other federal statutes and regulations relating to laboratory animals. Material has been reviewed by the Walter Reed Army Institute of Research. There is no objection to its presentation and/or publication. The opinions or assertions contained herein are the private views of the author, and are not to be construed as official, or as reflecting true views of the Department of the Army or the Department of Defense.

Informed Consent Statement: Not applicable.
Data Availability Statement: Authors will provide the raw data to any publicly available datasets as request.
Conflicts of Interest: GM has a patent the ALFQ adjuvant. No other author has any conflict of interest. The funders had no role in the design the study or its interpretation.