Dose-Dependent Efficacy of a Riboflavin and Ultraviolet Light-Inactivated Whole-Virion SARS-CoV-2 Vaccine in a Hamster Infection Challenge Model

Abstract

Background: A novel platform to produce whole-virion vaccines using riboflavin and ultraviolet (UV) light for photochemical inactivation has been developed. We previously reported on the potential for this platform to produce a safe and effective inactivated whole-virion SARS-CoV-2 vaccine. Feasibility studies used a hamster infection challenge model to explore the effects of route of administration and adjuvants on immune responses elicited by the vaccine candidate. Here, we utilized the same animal model to evaluate the dose response to the vaccine candidate in combination with the adjuvant CpG1018. Methods: A pilot batch of the vaccine candidate was produced at a contract development and manufacturing organization (CDMO) for use in this study. A two-dose intramuscular regimen at three antigen concentrations formulated with CpG1018 adjuvant was assessed against a live SARS-CoV-2 (USA-WA-1/2020) challenge. Results: The vaccine elicited dose-dependent neutralizing antibody responses, with peak PRNT₅₀ titers exceeding 1:5120. Vaccination significantly reduced lung viral burden and mitigated pulmonary pathology compared to controls. Antibodies persisted up to 154 days post-vaccination and neutralized Delta and Omicron (Jn.1) variants but showed limited activity against XBB.1.5. Flow cytometry revealed enhanced CD4⁺ Th1-biased responses in higher-dose groups. Conclusions: These findings demonstrate the protective efficacy of the SolaVAX SARS-CoV-2 vaccine candidate and support further evaluation of this vaccine production platform.

1. Introduction

For decades, infectious disease experts have warned of the threat posed by emerging infectious diseases (EID) and the potential for a devastating pandemic [1,2,3,4,5,6]. Prophetically, in September 2019, the Global Preparedness Monitoring Board warned “… there is a very real threat of a rapidly moving, highly lethal pandemic of a respiratory pathogen killing 50 to 80 million people and wiping out nearly 5% of the world’s economy” [7]. By December of that year, pneumonia cases of unknown etiology began to be reported in Wuhan, China [8]. The first draft genome of what is now known as Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) was published in January 2020. On 11 March 2020, the resulting disease, Coronavirus Disease 2019 (COVID-19), was declared a pandemic by the World Health Organization (WHO) [9].

Early estimates of the economic impact of the pandemic on the US economy ranged into the trillions of dollars [10,11,12]. Thus, the rapid development of COVID-19 vaccines became a national priority, resulting in the May 2020 formation of Operation Warp Speed (OWS), a public–private partnership among the Department of Health and Human Services (HHS), the Department of Defense (DOD), and selected pharmaceutical companies [13]. Of note, OWS was not focused on scientific discovery but rather optimizing time-to-market for vaccines based on already-developed vaccine platforms (mRNA, viral vector, and protein-based vaccines) and on managing the logistics for mass-vaccination campaigns [14]. The first OWS-supported vaccine received United States (US) Food and Drug Administration (FDA) Emergency Use Authorization (EUA) on 11 December 2020—an unprecedented 11 months from the identification of the virus to availability of a vaccine [9,13,14]. Readiness of vaccine platforms is a crucial tool for rapid response to the next pandemic.

A novel platform (SolaVAX^TM) to produce inactivated whole-virion vaccines via photochemical inactivation with riboflavin and UV light has been developed at Colorado State University (CSU). Inactivated whole-virion vaccines have a long history of contribution to public health, going back to the development of the Salk vaccine against poliomyelitis [15,16,17,18]. Conventional inactivation methods primarily rely on chemical compounds such as formalin and beta-propiolactone, which, while effective at inhibiting viral replication, have several drawbacks. These include inadvertent damage to antigens caused by the non-specific chemical action of these agents that may reduce effective immunization [17,19]. In addition, these agents pose safety risks, including carcinogenicity, mutagenicity, and combustibility [20,21]. Riboflavin, on the other hand, is a non-toxic essential vitamin, and when it is in proximity to nucleic acids and excited by UV light, an electron-transfer reaction alters guanine bases. At the same time, free riboflavin is photoconverted to lumichrome and other normal riboflavin metabolites [22]. This photochemistry has previously been shown to be effective in inactivating the SARS-CoV-2 virus in blood products to prevent transfusion-transmitted disease [23,24]. Additional work demonstrated that while the virus is successfully inactivated, antibodies present in plasma collected from COVID-19 convalescent donors remain intact [25]. The selectivity of this photochemical process enables complete virus inactivation while preserving the structural integrity of viral antigens, potentially enabling the vaccine to induce a vigorous immune response at relatively low doses.

A feasibility study using this inactivation platform was performed in a small-animal model, evaluating the immune response to inactivated SARS-CoV-2 whole-virion alone and in combination with adjuvants [26]. The vaccine was well tolerated and elicited a robust protective response against viral challenge in hamsters, with the best response observed with CpG1018 adjuvant and intramuscular administration. CpG1018 is a Th1-promoting adjuvant used in HEPLISAV-B, an FDA-approved hepatitis B vaccine, and has been shown to induce a higher antibody response over a shorter interval than a comparable alum-adjuvanted vaccine [27]. The current study was undertaken to evaluate dose-dependent responses and to establish neutralizing antibody (nAb) levels in the in vivo hamster infection challenge model. The study assessed three antigen doses combined with CpG1018 in a two-dose intramuscular (IM) administration regimen. Response parameters used to determine vaccine efficacy included viral tissue burden, histopathology of lung tissue, flow cytometric evaluation of immune cell activation markers, ELISA for antibody levels against target viral antigens, and clinical parameters post-vaccination. Furthermore, this study was used to establish the duration of antibody persistence post-vaccination and the in vitro neutralization effectiveness on the USA-WA-1/2020 (WA-1) strain and other SARS-CoV-2 variants. We also describe technology transfer of the SolaVAX platform to a contract development and manufacturing organization (CDMO; BioMARC, Fort Collins, CO, USA) to produce a pilot batch of inactivated virus material under conditions similar to those required for FDA-regulated vaccine manufacturing.

2. Materials and Methods

2.1. Study Design

The objective of this study was to evaluate the dose-dependent response of a whole-virion SARS-CoV-2 vaccine (SolaVAX-CoV-2) manufactured using riboflavin and UV inactivation technology (SolaVAX) in a hamster infection challenge model. The hamsters were randomly assigned into groups of 10 (5 male/5 female) hamsters per cohort. The untreated control group was further divided into two sub-cohorts, one challenged with live virus (Group 1a) and the other unchallenged (Group 1b). Both untreated control cohorts received sham injections of vaccine buffer. Three vaccine treatment groups (Groups 3–5) received vaccine doses ranging from 1.09 to 4.36 ng combined with CpG1018 adjuvant (Dynavax Technologies, Emeryville, CA, USA). An Adjuvant-Only control group (Group 2) and the Low-, Medium-, and High-Dose vaccine treatment groups were also challenged with live virus. Study endpoints were predetermined to evaluate the immune response to vaccination (Groups 1–5) and the persistence of the antibody response over time (Long-Term, Group 6). Due to deaths occurring at varying time points, unequal numbers of hamsters were evaluated in each group, as indicated in

Table 1. Vaccination groups.

Group Number	Group Name	SolaVAX-CoV-2 Dose (ng)	CpG1018 Adjuvant Dose (μg)	End Point (Day)	Challenge Strain	N 1 (Sex M/F)
1a	Control (Challenged)	0	None	45	WA-1	5 (3M/2F)
1b	Control (Unchallenged)	0	None	45	N/A	4 (2M/2F)
2	Adjuvant-Only	0	200	45	WA-1	10 (5M/5F)
3	Low-Dose	1.09	200	45	WA-1	9 (4M/5F)
4	Medium-Dose	2.18	200	45	WA-1	7 (2M/5F)
5 2	High-Dose	4.36	400	45	WA-1	10 (5M/5F)
6	LongTerm	2.18	200	182	N/A	7 (3M/4F)

¹ Number of animals surviving to the planned endpoint. ² The inactivated virus was provided at the medium dose concentration; therefore, Group 5 animals received two injections, one each in both flanks, at both injection time points to achieve the desired doubling of the vaccine dose.

.

The timeline for study treatments and sample collection is shown in Figure 1. All animals received vaccination or sham treatment by IM injection in the right or left flank on Days 0 and 21 of the study. Blood samples were collected under anesthesia before all injections. Blood samples and oropharyngeal swabs were collected from animals receiving live virus challenge prior to inoculation on Day 42. Oropharyngeal swabs were collected daily until Day 45. On Day 45, an additional blood sample was collected, the animals were humanely euthanized, and necropsies were performed. The animals in the Long-Term group were maintained without infection challenge and with periodic blood sampling (Days 43, 63, 91, 126, 154, and 182) until Day 182, at which time they were humanely euthanized and necropsied.

2.2. Vaccine Material Production

A pilot batch of vaccine candidate material (Lot #N18-22-19B-I) was produced at a CDMO (BioMARC, Fort Collins, CO, USA) under BSL-3 manufacturing conditions using the process flow shown in Figure 2. Briefly, Vero cells were propagated in OptiPRO Seum-Free Media (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with GlutaMAX™ (Thermo Fisher Scientific, Waltham, MA, USA) under serum-free conditions. Cell cultures were started in static culture flasks and then transferred to 3 L spinner flasks for expansion on Cytodex 1 microcarriers (Cytiva Life Sciences, Marlborough, MA, USA). The cells were inoculated with WA-1, and viral propagation was allowed to proceed at 37 °C with 5% CO₂ for approximately 72 h.

The viral supernatant was harvested and treated with Benzonase (EMD Millipore, Burlington, MA, USA) to a final concentration of 50 U/mL and incubated at 37 °C and 5% CO₂ for approximately 1 h with mixing. The Benzonase-treated supernatant was clarified using a Sartopore 2 XLG 0.8/0.2 µm filter (Sartorius AG, Bohemia, NY, USA). The clarified culture filtrate (CCF) formulation was adjusted to 10 mM phosphate and 4% sucrose. The CCF material was further processed by tangential flow filtration (TFF) using a 100 kD hollow fiber (HF) filter (Repligen Corp, Boston, MA, USA). The material was ultrafiltered (UF) to a concentration factor (CF) of 5.

The UF material was dispensed into an illumination bag (Mirasol Illumination bag, Terumo BCT, Lakewood, CO, USA) and combined with 35 mL of 500 μM riboflavin solution (Terumo BCT, Lakewood, CO, USA). The solution was gently mixed and residual air was expressed from the bag. The bag was placed into an illuminator (Mirasol PRT System, Terumo BCT, Lakewood, CO, USA) and exposed to 150 J UV light. A second round of inactivation was similarly performed.

The final inactivated virus material was characterized against selected release criteria established for use of these products in human clinical trials, as negotiated with the FDA in a pre-Investigational New Drug (IND) submission meeting. Characterization of the inactivated material includes assays for identity, strength/potency, purity, quality, and safety.

The absence of replication-competent virus (RCV) was confirmed by a cytopathic effect (CPE) assay, with two blind passages on Vero cell monolayers. Vero monolayers were prepared in MEM (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with glucose (Teknova, Hollister, CA, USA), GlutaMAX™, non-essential amino acids (Thermo Fisher Scientific, Waltham, MA, USA), sodium pyruvate (Thermo Fisher Scientific, Waltham, MA, USA), and 10% fetal bovine serum (FBS; Atlas Biologicals, Fort Collins, CO, USA), and the inactivated material incubated with the Vero monolayers for approximately 72 h. The monolayers were examined for signs of CPE. The culture supernatants were harvested and incubated with a fresh Vero cell monolayer for an additional 72 h, and the monolayers were again examined for CPE. If no CPE was observed for both incubations, the samples were considered free of RCV. The inactivated virus material was diafiltered (DF) into vaccine buffer (0.2 M NaCl, 10 mM phosphate, and 4% sucrose) using a 100 kD HF filter (Repligen Inc., Boston, MA, USA).

The concentration of functional spike protein was determined using the Coronavirus Spike Protein Assay (InDevR Inc., Boulder, CO, USA) through comparison of the relative binding of the sample to two capture proteins, an anti-Receptor Binding Domain (RBD) antibody and the angiotensin-converting enzyme 2 (ACE2) receptor bound to an anti-his-tag antibody with the Coronavirus S1 Recombinant Spike Protein Standard (Sino Biological, Paoli, PA, USA) as standard.

2.3. Vaccine Preparation

The vaccine was prepared by combining inactivated virus material (BioMARC Lot #N18-22-19B-I) with vaccine buffer (BioMARC Lot #N18-IS-22-14) and CpG1018 adjuvant (Lot #QX303-01). The appropriate concentration of vaccine was loaded into 1 mL syringes coupled to 26G needles, kept on ice, and used within one hour of preparation. The vaccination was administered intramuscularly.

2.4. Challenge Virus

Challenge virus propagation took place in a BSL-3 laboratory. WA-1 was acquired from BEI Resources (product NR-52281, Lot #70036318, Manassas, VA, USA) and cultured in Vero cells (ATCC CCL-81). The Vero cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with glucose, L-glutamine, sodium pyruvate, and 5% fetal bovine serum (FBS). Infection of Vero cells with WA-1 was conducted directly in DMEM containing 1% FBS in a 175 cm² tissue culture flask. The flask was incubated for 3–4 days at 37 °C, 5% CO₂. The medium was harvested from the infected cells after the incubation period and clarified by centrifugation. The supernatant was collected, supplemented with 10% FBS, and frozen in aliquots to −80 °C. Viral titer was determined using a standard plaque assay. The challenge virus was previously sequenced using next-generation sequencing (NGS) as described in Ragan et al. [26]. NGS analysis was not repeated since the same batch of challenge virus inoculum was used in this study.

2.5. Animals

All hamsters were housed in animal facilities at Colorado State University (CSU), which are accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) International. The study received ethical approval from the CSU Institutional Animal Care and Use Committee (IACUC, protocol #6121). A total of 60 Syrian Golden hamsters (Mesocricetus auratus) were acquired from Charles River Laboratories (Wilmington, MA, USA) at 6 weeks of age. The hamsters were maintained in a Biosafety Level-2 (BSL-2) animal facility for acclimation and during the study vaccination period. They were group-housed in cages containing 2 or 3 animals each and fed a commercial diet with access to water ad libitum. A microchip with biothermal capabilities was administered subcutaneously to each animal dorsal to the scapula for identification.

As previously described, the hamsters were randomly assigned to six groups with 10 hamsters per group (5M/5F). They were treated according to their assigned group with an IM injection on Day 0 and Day 21 of the study. The Control hamsters received 100 μL of vaccine buffer per injection. The Adjuvant-Only hamsters received a 200 μg dose of CpG1018 administered as 100 μL of a stock preparation containing 1500 μL of vaccine buffer and 250 μL of CpG1018. The Low-Dose, Medium-Dose, and Long-Term groups were injected with 100 μL of vaccine formulation as specified in Table 1. High-Dose hamsters were injected twice with 100 μL volumes of the 2.18 ng SolaVAX-CoV-2/200 μg CpG1018 formulation to allow for doubling of the dose as specified.

Hamsters were maintained for an additional 19 days in the BSL-2 facility; then on Day 40, they were transferred to a BSL-3 facility in preparation for live virus challenge on Day 42. Hamsters in Groups 1–5 were anesthetized, and live WA-1 virus was administered by pipette into the nares (50 μL/nare) for a total volume of 100 μL per hamster. Back titration of the challenge virus stock was performed on the same day to verify the titer. Hamsters receiving live virus challenge were maintained thereafter for 3 days. Animals in the Long-Term group were maintained without challenge until Day 182.

Body weights and temperatures were measured and recorded 5–7 days prior to vaccination during the acclimation period, at the time of prime vaccination (Day 0), weekly thereafter until the time of the live virus challenge (Day 42), and then daily until euthanasia on Day 45. Clinical observations were recorded weekly post-vaccination until Day 42 and then daily after the live virus challenge. The clinical evaluation tracked temperament, ocular discharge, nasal discharge, weight loss, coughing/sneezing, dyspnea, lethargy, anorexia, and moribundity. Hamsters were humanely euthanized at their predetermined endpoints. Necropsies were performed to assess gross pathological changes, and tissues were collected for virus quantification, histopathology, and flow cytometry.

2.6. Neutralizing Antibodies

Serum samples collected on Days 0, 21, 42, and 45 were evaluated for nAb titers by the plaque reduction neutralization test with a 50% cutoff (PRNT₅₀). Briefly, serum was heat-inactivated and then diluted two-fold in DMEM starting at a 1:5 dilution. An equal volume of SARS-CoV-2 viral stock containing approximately 200 plaque-forming units (pfu) per 0.1 mL was added to the serum dilutions as the challenge agent for the assay. Agents tested included the WA-1/2020 strain and the Delta, Omicron (Jn.1), and XBB.1.5 variants. The serum and virus were gently mixed, incubated for 60 min at 37 °C, and inoculated using 0.1 mL of the appropriate serum dilution and virus mixture into 6-well cell culture plates containing a confluent layer of Vero CCL-81 cells. A double overlay of 0.5% agarose in media containing neutral red dye was applied to each well, and plaques were counted 3 days after the second overlay. The nAb titers are reported as the reciprocal of the highest dilution at which 50% of the virus was neutralized.

2.7. Tissue Virus Titers

For virus quantification, portions of the cranial lung lobe, caudal lung lobe, and nasal turbinates from each hamster were collected into DMEM media containing antibiotics, homogenized, and frozen at −80 °C until the time of analysis. Tissue homogenates were centrifuged and viral titers in the clarified supernatant were determined by standard plaque assay. Oral swabs collected after challenge were assessed similarly.

2.8. ELISA for Detection of Anti-S1, Anti-S2, and Anti-RBD Antibodies

To assess antibody binding to specific regions of the SARS-CoV-2 spike protein—S1 (amino acids 16–685), S2 (amino acids 686–1213), and RBD (amino acids 319–541)—enzyme-linked immunosorbent assays (ELISA) were performed using recombinant proteins. The assay protocol was adapted from a previously published protocol [28]. Briefly, high-binding 96-well plates were coated with 50 ng of S1, S2, or RBD protein diluted in phosphate-buffered saline (PBS) and incubated overnight at 4 °C. Plates were then washed five times with PBS containing 0.05% Tween-20 and blocked for 2 h at room temperature with a buffer containing PBS, 2% bovine serum albumin (BSA), 2% normal goat serum, and 0.05% Tween-20. Serum samples from unvaccinated and vaccinated hamsters were diluted (1:100, 1:1000, 1:10,000, and 1:100,000) in the blocking buffer and added to the wells for 1 h at room temperature. After washing, plates were incubated with a 1:5000 dilution of horseradish peroxidase (HRP)-conjugated anti-hamster IgG (H+L) secondary antibody (Jackson ImmunoResearch, West Grove, PA, USA) for 1 h. Following another wash, tetramethylbenzidine (TMB) substrate (Thermo Fisher Scientific, Waltham, MA, USA) was added, and the reaction was stopped after 10 min by adding 100 μL of stopping buffer. Optical density was measured at 450 nm using a plate reader (BioTek Synergy 2, BioTek Instruments, Winooski, VT, USA).

2.9. Flow Cytometry

Flow cytometry was performed at 3 days post-WA-1 challenge to evaluate the type of immune response elicited by the vaccine post-infection. The flow cytometry panel was designed to distinguish between vaccine-induced Th1 or Th2 immunity. For this, cells from lungs and spleens were isolated, prepared as a single-cell suspension, and stained with T-helper cell marker (CD4), T-cytotoxic cell marker (CD8), Th1 markers (IFNγ, TNFα, IL-6, CXCR3, Tbet), and Th2 markers (IL-4, IL-10, CXCR4, GATA-3). All antibodies were obtained commercially (Supplementary Materials). Cells were read using a 4-laser spectral flow cytometer (Cytek Aurora, Cytek Biosciences, Fremont, CA, USA) and analyzed using FlowJo (Beckton Dickinson and Company, Ashland, OR, USA) and cytotypr (an R-based pipeline developed in Henao-Lab). Data was then plotted using R version 4.4.2.

2.10. Histopathology

Tissues collected during necropsy were placed in 10% neutral buffered formalin for 7 days, then embedded in paraffin and stained with hematoxylin and eosin (H&E). A board-certified veterinary pathologist interpreted histopathology blindly. Semi-quantitative grading was performed on H&E-stained lung tissue to assess bronchi infiltrative inflammation, bronchial hyperplasia, consolidating interstitial pneumonia, and thickening of alveolar septa [29]. Two sections of lung per animal were evaluated, and data were expressed as an averaged composite score of all graded parameters to demonstrate overall pulmonary pathology.

Immunohistochemistry (IHC) was performed on serial sections of the same lung portion via an automated Leica, Bond RXM stainer, and pre-validated staining methodologies. IHC was performed using a rabbit polyclonal antibody targeting the SARS-CoV-2 nucleocapsid protein. The antibody was developed by Dr. Brian Giess and optimized for IHC by the CSU Experimental Pathology Facility. IHC staining on samples was validated by a board-certified pathologist and quantified utilizing artificial intelligence (AI) analysis software (version 2025.08.1.18881 x64, Vsiopharm A/S, Hørsholm, Denmark). For each tissue section, a region of interest (ROI) identification algorithm was generated at low magnification with custom decision forest training and classification to differentiate tissue versus background based on color and area. Similar ROI generation was then utilized to isolate major airway spaces from the alveolar space. Cell counts, via identification of nuclei, were performed on whole tissue sections and counted as positive (yellow labeling) or negative (blue labelling) for SARS-CoV-2 antigen based on staining patterns. Data were expressed as percent positivity in both major airways and alveolar spaces.

2.11. Correlation Analysis

To quantitatively evaluate the relationships between indicators of protective immune response against lung injury and other parameters measured, a correlation analysis was performed. Measurements from individual animals (PRNT₅₀ at Day 45, lung IHC percent positivity, pfu/g in cranial and caudal lung lobes, and flow cytometry-derived frequencies of the indicated T cell populations) were compiled into a single dataset. Pairwise correlations between variables were calculated using Spearman’s rank correlation (two-tailed) to account for non-normal distributions. Correlation coefficients (r) were visualized as a heatmap with values overlaid in each cell, and the color scale represented the magnitude and direction of the association (−1 to +1).

2.12. Statistical Analysis

Unless otherwise noted, statistical analysis was performed using GraphPad Prism version 10.6.1 for Windows (GraphPad Software, Boston, MA, USA). For all analyses, adjusted p-values < 0.05 were considered statistically significant. An ordinary one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons test was used to analyze nAb and viral titers from collected tissues. Male versus female comparisons for nAbs used multiple Mann–Whitney tests with correction for multiple comparisons by the two-stage step-up method to estimate the false discovery rate (FDR). For the ELISA endpoint titers, values were log2-transformed prior to analysis to improve normality (ELISA-R pipeline). Group comparisons were performed using one-way ANOVA followed by Tukey’s multiple-comparisons test for all pairwise contrasts. Data are presented as individual animals with group summaries. Statistical analysis for flow cytometry was performed using ANOVA with the post hoc Tukey honestly significant difference (HSD) test in R using the stats package. Semi-quantitative and quantitative histopathology and immunohistochemistry data were analyzed using an ordinary one-way ANOVA with Dunnett’s test for multiple comparisons.

3. Results

3.1. Vaccine Materials

Characterization of the vaccine material used in this study (Lot #N18-22-19B-I) is summarized in

Table 2. Characterization of vaccine material lot #N18-22-19B-I.

Parameter	Test Method	Specification	Result
Identity	Identity by NGS	Confirm as SARS-CoV-2	Pass
Potency/Strength	Functional spike protein quantification by VaxArray	Report result	114.48 ng/mL
	Genomic copies by RT-qPCR	Report result	5.93 × 1010 gene copies/mL
Purity	Determination of residual HCP by ELISA	Report result	0.714 μg/mL (N18753)
	Detection and quantitation of residual Vero DNA	Report result	<0.500 ng/mL
	Determination of residual BE by ELISA	Report result	<30.85 ng/mL (N18697)
	Detection of residual riboflavin	<20.00 ng/mL (<LLOQ)	Pass
Quality	Visible particulates per USP <790>	Essentially free of visible particulates	Clear, colorless liquid with no visible particulates (N18815)
	pH determination per USP <791>	Report result	pH 7.4 (N18816)
	Total protein by microBCA assay	Report result	0.049 mg/mL (N18784)
	Osmolality per USP <785>	Report result	4.783 mOsm/kg
Safety	Detection of RCV through CPE Assay	Absence of RCV	Pass
	Bioburden per USP <61>	<10 CFU/mL	Pass No colonies observed (performed at 1:10 dilution)
	Endotoxin per USP <85>	500 EU/mL	9.0 EU/mL

BCA = bicinchoninic acid; BE = benzonase endonuclease; CFU = colony forming unit; CPE = cytopathic effect; ELISA = enzyme-linked immunosorbent assay; HCP = host cell protein; LLOQ = lower limit of quantitation; NGS = next-generation sequencing; RCV = replication-competent virus; RT-qPCR = quantitative reverse transcription polymerase chain reaction; USP = United States Pharmacopeia.

. Complete inactivation of the current lot was confirmed by the lack of CPE on indicator cells. The number of viral copies per volume was similar to previous lots. The quantity of spike protein as measured by ELISA was lower than in previous lots; this may be due to loss of free spike protein during further DF processing. Lot #N18-22-19B-I passed specifications established with the FDA through a pre-IND submission meeting and subsequent review.

3.2. Challenge Virus

Following administration of the live viral challenge to the treatment and control groups, the challenge virus material (BEI Resources product NR-52281, Lot# 70036318) was evaluated by plaque assay. The challenge dose was determined to be 2.3 × 10⁵ pfu/mL.

3.3. Clinical Measurements and Observations

All hamsters gained weight within normal limits before the challenge. The Long-Term monitoring group showed a steady increase in body weight from 116 ± 9 g on Day 0 to 187 ± 27 g on Day 63, after which body weights remained steady. No change in this trend was observed after the initial vaccination and boost.

We observed decreases in body weight for all challenged groups in the days following viral infection. For the treatment groups, the lowest mean weights were observed on the day after (−4.33% for the Low-Dose group) or 2 days after (−4.29% for the Medum-Dose group and −3.14% for the High-Dose group) the challenge. In contrast, weights continued to decrease for both control groups through the 3rd day after the challenge (Supplementary Figure S1). The largest decrease on the day of euthanasia was observed in the Adjuvant-Only group with a recorded change in body weight of −4.21%. No statistically significant differences in body weight were observed among groups on the day of challenge nor at 3 days post-challenge.

Clinical scores and body temperatures for all hamsters were within normal limits. Average body temperatures for all groups ranged between 96.3 and 97.9 °F on the day of vaccination, between 96.3 and 98.8 °F on the day of live virus challenge, and between 95.5 and 101.4 °F at 3 days post-challenge. No abnormalities due to vaccination were noted.

3.4. Neutralizing Antibodies

The nAb titers against SARS-CoV-2 were below the limit of quantification, as expected at baseline for all groups prior to vaccination on Day 0 and for Control groups throughout the duration of the study. Animals in all vaccine dose groups had quantifiable nAb titers on Day 42, three weeks after administration of the second vaccine dose, and on Day 45 (Figure 3). It has been shown that nAb titers ≥160 correlate with high levels of protection in humans vaccinated against SARS-CoV-2 [30]. Generally, neutralizing antibody levels increased as the dose of antigen and/or adjuvant (High-Dose group) increased. No significant differences between male and female values were observed (Figure 4).

Group 6 (Medium-Dose; no challenge) maintained nAbs up to Day 154 (Figure 5). A statistically significant increase in nAbs was detected between Days 21 and 43 (p = 0.0072); likewise, a statistically significant decrease in nAbs was detected between Days 43 and 182, where nAbs returned to the Day 21 level. On day 182, the average PRNT₅₀ titer was at 1:160.

We observed that PRNT₅₀ values ranged from 640 to 5120 when challenged in vitro with WA-1 (Figure 6). The range of PRNT₅₀ values decreased successively when challenged by the Delta variant (160 to 1280), the Jn.1 variant (20 to 320), and XBB1.5 (<20). Neutralizing capacity should correlate with vaccine effectiveness, ranging from 75% for Jn.1 to >95% for WA-1 [31,32,33]. Testing with the variant XBB1.5 did not show effective neutralization capacity with this vaccine.

3.5. Tissue Virus Titers

Significant differences (p < 0.0001) were observed between the vaccinated groups and the control groups in the viral titers measured in cranial and caudal lung tissue (Figure 7a,b). Animals in the High-Dose vaccine group exhibited the lowest viral burden in lung tissue while the control groups had the highest levels of infectious virus. Differences in viral titers from nasal turbinates were observed, but except for the Medium-Dose group, compared with controls, they did not reach statistical significance (Figure 7c). Significant reductions in virus titers from oropharyngeal swabs were observed in all vaccinated animals compared to controls (Figure 7d). No detectable virus was observed in swabs from any of the vaccinated animal groups at 3 days post-infection with live virus (Day 45) despite attempts to concentrate samples to increase assay sensitivity.

At day 3 post-SARS-CoV-2 infection, lung T cell composition and effector polarization differed across groups in a way that separated vaccinated animals from sham Control and Adjuvant-Only controls (Figure 8A–D). Total CD4⁺ T cell frequencies were increased in vaccinated animals, reaching significance in the High-Dose group (4.36 ng + 400 µg) compared with controls (p < 0.05; Figure 8A). Within the CD4 compartment, the High-Dose group also exhibited the greatest enrichment of CD4⁺CXCR3⁺ cells (Figure 8B), consistent with preferential accumulation of Th1-associated effector cells in the lung early after infection. In contrast, CD4⁺TNFα⁺ cells were broadly present across groups without a clear dose-dependent separation (Figure 8B). Total CD8⁺ frequencies were comparatively stable across groups (Figure 8C); however, Control and Adjuvant-Only animals showed a distinctly more inflammatory CD8 phenotype, with elevated CD8⁺CXCR4⁺ and CD8⁺IFNγ⁺ frequencies that were significantly reduced in the Medium- and High-Dose vaccine groups (Figure 8D; p < 0.05 to p < 0.01). CD8⁺TNFα⁺ responses were also highest in the Control/Adjuvant-Only animals and were significantly lower in the Medium/High-Dose vaccinated animals (Figure 8D; p < 0.01), indicating attenuated early inflammatory CD8 activation following vaccination.

Group-level compositional summaries further supported a vaccine-associated shift away from inflammatory CD8 subsets (Supplementary Figures S2 and S3). When visualized as proportional stacked bars, Control (and to a lesser extent Adjuvant-Only) animals showed a larger contribution of activated/inflammatory CD8 compartments (e.g., CXCR4⁺ and IL-6⁺ fractions). In contrast, vaccinated groups displayed relatively reduced inflammatory CD8 contributions alongside a greater representation of non-CD4/CD8 (“CD4-CD8-”) and CD4⁺ compartments (Supplementary Figure S2). Sex-stratified views suggested these compositional patterns were broadly conserved in both females and males, indicating that the directional vaccine-associated remodeling was not driven by a single sex (Supplementary Figure S3).

Serum ELISA at day 45 demonstrated strong vaccine-induced spike-specific IgG responses (Figure 8E–G). Anti-RBD IgG endpoint titers were near baseline in the Control and Adjuvant-Only groups. Still, they were robustly elevated in all vaccinated groups, with the highest titers observed in the High-Dose SolaVAX-CoV-2 + CpG1018 cohort, and multiple significant pairwise differences were observed as indicated (Figure 8E). A similar pattern was observed for anti-S1 IgG, with clear induction across vaccine doses and the greatest magnitude again in the High-Dose group (Figure 8F). In contrast, anti-S2 IgG responses were lower overall and more heterogeneous, with significant induction detected in only select vaccine comparisons (Figure 8G), indicating that humoral responses were preferentially directed toward RBD/S1 relative to S2 in this dataset.

3.6. Histopathology

Tissue from necropsied hamsters was prepared for analysis of pathologic lesions. Semi-quantitative grading of all lung lobes was based on previously established grading criteria for assessing SARS-CoV-2 pulmonary pathology [34]. AI software provided objective quantification of SARS-CoV-2 antigen within major airways and alveolar spaces. All samples were blindly assessed until all analyses were completed. Semi-quantitative grading demonstrated a significant difference in composite pulmonary pathology between controls and higher-dose vaccine groups (Figure 9).

The largest range in pathologic scores is within group 3, possibly indicating varying or animal-specific effects of low-dose vaccine administration. Consistent with pulmonary pathologic findings, there is a significant difference in viral antigen burden (as shown via immunohistochemistry) between the control and vaccine groups, regardless of vaccine dosage (Figure 10 and Figure 11). In all groups, the highest viral antigen burden was appreciated in bronchioles with variable interstitial presence, with vaccinated groups having no alveolar/interstitial viral antigen. Cumulatively, the High-Dose group demonstrated the smallest degree of pulmonary pathology and viral antigen burden in the face of the infection challenge.

3.7. Correlation Analysis

In the correlation matrix (Supplementary Figure S4), nAb activity at day 45 (PRNT₅₀) aligned most strongly with CD4-associated lung responses and inversely with inflammatory CD8 features. PRNT₅₀ showed a moderate positive correlation with total CD4⁺ T cells (r = 0.47), whereas it was negatively correlated with CD8 inflammatory subsets, consistent with higher neutralization associating with a less inflammatory CD8-skewed lung profile. Measures of viral burden (pfu/g; cranial and caudal lung) and histopathology/antigen burden (IHC percent positivity) were primarily associated with inflammatory CD8 signatures and inversely associated with CD4 enrichment. IHC positivity correlated strongly with CD8⁺CXCR4⁺ (r = 0.71), CD8⁺IFNγ⁺ (r = 0.63), and CD8⁺TNFα⁺ (r = 0.69), indicating that higher viral load and tissue damage tracked with heightened inflammatory CD8 activation. Similarly, tissue viral burden levels correlated positively with activated CD8 subsets—most notably CD8⁺CXCR3⁺ (cranial r = 0.68; caudal r = 0.79), CD8⁺CXCR4⁺ (cranial r = 0.48; caudal r = 0.53), and CD8⁺TNFα⁺ (cranial r = 0.59; caudal r = 0.43). In contrast, both IHC positivity and tissue viral burden were negatively correlated with total CD4⁺ frequency (IHC r = −0.48; cranial r = −0.42; caudal r = −0.39) and with CD4⁺CXCR3⁺ (IHC r = −0.33; cranial r = −0.16; caudal r = −0.17). Collectively, these correlations link higher neutralization with greater CD4 representation (including CXCR3⁺ CD4 cells) and link higher viral and histologic burden with inflammatory CD8 activation, particularly CXCR4⁺, IFNγ⁺, and TNFα⁺ CD8 subsets.

4. Discussion

The vaccine material evaluated in this study was produced at BioMARC, a biologics CDMO with BSL-3 capabilities and experience with viral expression platforms for the production of attenuated or inactivated viral vaccines. Their typical manufacturing capacity is up to 200 L, indicating that technology transfer of the SolaVAX inactivated whole-virion vaccine platform to scaled-up manufacturing for Phase 1 clinical trials under current Good Manufacturing Practices (cGMP) has been achieved. This lot of vaccine material was similar to previous batches produced using simpler manual methods in a laboratory setting in that the number of viral copies per volume was consistent. However, the amount of spike protein was lower. The difference may be attributed to additional filtration steps built into the manufacturing process to ensure that specifications negotiated with the FDA are met. The ability to ramp manufacturing capacity is a key capability for a rapid response to EIDs with pandemic potential.

The first vaccine to be developed using the SolaVAX platform is targeted to prevent COVID-19. Testing in a hamster infection challenge model demonstrated protection against SARS-CoV-2 across all test groups as measured by antibody production, cellular immune responses, pathologic lesion burden, and viral burden in lung tissues. The vaccine elicited dose-dependent humoral and cellular immunity correlating with protection against SARS-CoV-2 infection challenge. Neutralizing antibody titers ≥1:160 observed in vaccinated hamsters align with human correlates of protection [30]. Even at the lowest dose utilized in this study, 1.09 ng of inactivated virus, neutralizing antibody titers exceeded the 1:160 threshold. The Long-Term survival group maintained antibody titers at or above the 1:160 threshold throughout the 6-month study. Since hamsters in the High-Dose group demonstrated higher initial neutralizing antibody titers, one might anticipate that protective neutralizing antibody levels would have been maintained for a longer period than the 6 months evaluated here.

It is noteworthy that the inactivated whole-virion vaccines currently in use outside the United States utilize much higher doses of inactivated virus than were evaluated in this study. The major vaccines in this space—CoronaVac (Sinovac, Beijing, China), BBIBP-CorV (Sinopharm, Beijing, China), and Covaxin (Bharat Biotech, Hyderabad, India)—are manufactured using β-propionolactone for viral inactivation and use clinical doses ranging from 1 to 8 μg [35,36,37]. The roughly thousand-fold difference in antigen dose suggests that riboflavin and UV can fully inactivate viruses with markedly less damage to antigenic epitopes. In addition to lower vaccination doses, this enables more efficient manufacturing. The use of UV-C alone to inactivate SARS-CoV-2 has been evaluated by numerous investigators, but in addition to genomic damage, UV-C can induce structural damage through direct photolysis and oxidative stress mechanisms [38,39,40,41,42]. While it has been shown in a laboratory setting that UV-C irradiation at certain wavelengths can inactivate SARS-CoV-2 primarily through genomic damage, these laboratory set-ups have yet to be translated into robust processes or devices [39]. The SolaVAX platform integrates into conventional vaccine manufacturing processes relatively easily, making it a candidate for rapid implementation even in low- and middle-income countries, the areas that are least served by newer technologies such as mRNA vaccines, thereby improving global vaccine equity [37]. In addition to enabling local production of primary vaccines targeted at EIDs, the SolaVAX platform facilitates timely strain updates for rapidly mutating pathogens.

The persistence of antibody over time and the levels of neutralizing antibody against variants suggest that an inactivated whole-virion vaccine produced on this platform may have advantages over alternative approaches that target the spike protein or other single-protein targets alone. It is important to note that the conformation of the vaccine spike protein utilized in this study was confirmed by its ability to bind ACE-2, and that changes in spike protein conformation may destroy a key epitope needed to ensure adequate vaccine efficacy [43].

Building on the established manufacturing readiness and the protection observed in vivo, analysis of lung immune responses by flow cytometry at day three post-SARS-CoV-2 challenge in vaccinated versus control animals provides insight into how SolaVAX-CoV-2 plus CpG1018 mediates protection. Vaccination appears to shift the early lung response toward a coordinated, antiviral-leaning immune state while limiting excessive inflammatory effector activity that is commonly linked to pulmonary tissue injury during acute respiratory viral infection. This early immune “set point” is complemented by strong antibody immunity measured by ELISA at the same time point, where spike-binding IgG responses are robust and largely focused on RBD and S1, consistent with effective priming and maturation of class-switched B cell responses after whole-virion immunization. In contrast, S2-directed binding is comparatively weaker and more variable, suggesting domain-focused immunodominance under the conditions tested and highlighting a rational target for future optimization if broader, conserved-epitope coverage is desired, particularly in light of antigenic drift. In addition, the integrated outcome associations reinforce this framework: measures of viral and histologic burden correlate most significantly with inflammatory immune activation in the lung, whereas neutralizing activity trends with a less inflammatory immune profile, collectively supporting the idea that vaccine-mediated protection reflects not only the magnitude of antibody responses but also the quality and regulation of the pulmonary immune environment during early infection.

A limitation of the study is that the vaccine tested was based on the Wuhan-like USA-WA-1/2020 strain, which is phylogenetically and antigenically distant from current strains circulating worldwide. Choosing this strain allowed the use of the preclinical hamster model developed and characterized by Bednash et al. [34]. The model consistently results in acute SARS-CoV-2 infection, viral pneumonia, and systemic illness following intranasal inoculation. Infected hamsters exhibit key features of acute respiratory distress syndrome (ARDS), including histological evidence of lung injury, supporting its use to study COVID-19 ARDS. However, waning effectiveness was observed in vitro when hamster sera were challenged with the Delta, Omicron (Jn.1), and XBB.1.5 variants of SARS-CoV-2 in PRNT assays, reflecting antigenic drift as the evolving virus accumulated mutations. Conserved epitopes likely account for the partial neutralization of Delta and Jn.1. In contrast, the XBB lineage arose from a recombination event, merging numerous spike protein mutations with known immune evasion properties [44]. Mutations contributing to immune escape include K444T, which uses mechanisms such as steric hindrance and disruption of hydrogen bonds to reduce the binding affinity of antibodies induced against previous lineages [45,46]. Additionally, XBB.1.5 possesses the rare F486P mutation, which improves ACE2 receptor binding while interfering with class I neutralizing antibody recognition [44]. The minimal neutralization of XBB.1.5 observed in this study underscores the need for periodic updates to antigen formulations.

5. Conclusions

SolaVAX is a novel vaccine platform based upon riboflavin and UV light photochemical inactivation technology. It enables efficient vaccine manufacturing that is safe for operators and vaccine recipients alike. The first vaccine developed on this platform, SolaVAX-CoV-2, demonstrated dose-dependent efficacy in hamsters, reducing viral burden and pathology while inducing durable immunity at much lower doses than comparable inactivated whole-virion vaccines. These data support advancing this vaccine construct into clinical development.

6. Patents

Patents have been filed for the SolaVAX vaccine production platform and are pending approval.