Direct Cytoplasmic Transcription and Trimeric RBD Design Synergize to Enhance DNA Vaccine Potency Against SARS-CoV-2

Abstract

Background/Objectives: The emergence of immune-evasive SARS-CoV-2 variants highlights the need for adaptable vaccine strategies. Trimeric receptor-binding domain (tRBD) antigens offer structural and immunological advantages over monomeric RBDs, but DNA vaccine efficacy has been limited by inefficient antigen expression, particularly in non-dividing antigen-presenting cells. Although cytoplasmic transcription–based DNA platforms have been developed to overcome nuclear entry barriers, their utility for antigen structure–function optimization remains underexplored. This study evaluated whether integrating a rationally designed trimeric RBD with a T7-driven cytoplasmic transcription system could enhance immunogenic performance. Methods: A DNA vaccine encoding a tandem trimeric SARS-CoV-2 RBD was delivered using a T7 RNA polymerase-driven cytoplasmic transcription system. In vitro antigen expression was assessed following Lipofectamine 3000-mediated transfection. In vivo, mice were immunized with the SM-102-based Rpol/tRBD/LNP formulation, and immunogenicity was assessed by antigen-specific antibody titers, serum neutralizing activity, and T-cell response profiling, together with basic safety/tolerability evaluations. Results: The T7-driven cytoplasmic transcription system markedly increased antigen mRNA and protein expression compared with conventional plasmid delivery. Rpol/tRBD vaccination induced higher anti-RBD IgG titers, enhanced neutralizing antibody activity, and robust CD8⁺ T cell responses relative to monomeric RBD and plasmid-based trimeric RBD vaccines. Immune responses were Th1-skewed and accompanied by germinal center activation without excessive inflammatory cytokine induction, body-weight loss, or hepatic and renal toxicity. Conclusions: This study demonstrates that integrating rational trimeric antigen engineering with direct cytoplasmic transcription enables balanced and well-tolerated immune activation in a DNA vaccine context. The T7 autogene-based platform provides a flexible framework for antigen structure–function optimization and supports the development of next-generation DNA vaccines targeting rapidly evolving viral pathogens.

1. Introduction

The persistent emergence of more transmissible and immune-evasive SARS-CoV-2 variants threatens the effectiveness of current vaccination programs worldwide. While current COVID-19 vaccines have markedly reduced the rates of severe disease and hospitalization, breakthrough infections remain prevalent, especially with highly transmissible variants such as Omicron and its sublineages [1]. Given the antigenic drift of SARS-CoV-2, there is a pressing need for next-generation vaccine strategies that not only provide broader protection but are also rapidly adaptable to emerging mutations.

The spike (S) glycoprotein, essential for viral entry, binds to the human angiotensin-converting enzyme 2 (hACE2) receptor. Naturally forming a membrane-anchored homotrimer, the S protein contains an immunodominant receptor-binding domain (RBD) within the S1 subunit, which engages hACE2 and harbors key neutralizing epitopes [2,3]. RBD-based vaccines have induced potent neutralizing activity in vitro and conferred protection in vivo [4,5]. However, as the RBD is also the primary site of mutations in emerging SARS-CoV-2 variants, such changes have been a major contributor to the reduced efficacy of existing vaccines. Moreover, the relatively small molecular size of monomeric RBD limits its intrinsic immunogenicity, likely due to suboptimal B cell receptor (BCR) cross-linking and inefficient antigen presentation [6,7].

To enhance the immunogenicity and epitope presentation of RBD-based vaccines, multimeric constructs such as trimeric RBDs (tRBDs) have gained increasing attention. Trimeric formats aim to recapitulate the quaternary structure of the native spike, thereby improving structural stability, valency, and BCR cross-linking [8,9]. Structural analyses indicate that the flexible termini of RBD permit tandem linkage within a single polypeptide chain. Accordingly, various trimerization strategies have been developed, including tandem homologous repeats, heterotrimeric designs incorporating variant-specific mutations, and mosaic formats to broaden immunological coverage [9,10,11].

Nevertheless, the immunogenic potential of trimeric RBDs is critically dependent on the delivery platform. Among diverse vaccine platforms—including viral vectors, protein subunits, and nucleic acid–based approaches—nucleic acid vaccines offer unique advantages in manufacturability, modularity, and scalability [12]. mRNA vaccines have demonstrated high clinical efficacy and rapid development timelines; however, their intrinsic instability necessitates ultracold-chain storage and poses logistical challenges in global deployment [13,14]. DNA vaccines are comparatively thermostable and cost-effective for large-scale manufacturing, yet conventional plasmid DNA vaccines typically suffer from inefficient nuclear entry and limited antigen expression, particularly in non-dividing cells such as antigen-presenting cells (APCs) [15]. In many experimental settings, less than 1% of cytoplasm-delivered plasmid DNA reaches the nucleus, highlighting a fundamental mechanistic bottleneck [16]. In addition, bacterial backbone elements within plasmids introduce regulatory and safety concerns and may further reduce transcriptional efficiency [17,18].

To overcome these inherent limitations, cytoplasmic transcription-based DNA vaccine systems have recently been described by our group, in which T7 autogene-driven transcription enables direct cytoplasmic RNA synthesis from a DNA template, thereby bypassing the nuclear barrier that constrains conventional plasmid DNA vaccines [19]. This system supports efficient antigen expression in both dividing and non-dividing cells, including professional APCs. Moreover, cytoplasmic transcription may reduce exposure to exogenous double-stranded RNA (dsRNA) impurities associated with in vitro-transcribed mRNA, while retaining the chemical stability and scalable manufacturability of DNA.

In IVT mRNA vaccines, in vitro transcription frequently generates double-stranded RNA (dsRNA) by-products, which strongly activate innate RNA sensors and can both increase reactogenicity and suppress antigen translation [20,21,22]. In contrast, the T7 autogene-driven DNA platform synthesizes RNA intracellularly without introducing exogenous dsRNA contaminants, thereby mitigating dsRNA-associated inflammatory and translational penalties. While this platform has previously been validated using the full-length SARS-CoV-2 spike as a proof-of-concept for transcriptional efficiency, its potential to support systematic antigen structure–function optimization—particularly with respect to multivalent antigen architectures—has not yet been explored.

In contrast to prior studies that emphasized platform validation, the present study is specifically designed to isolate and define the immunological impact of antigen architecture under conditions in which transcriptional constraints are substantially alleviated. By embedding a rationally engineered tandem trimeric RBD (tRBD) immunogen within a highly efficient T7-driven cytoplasmic transcription framework, we investigate how multivalent antigen organization governs humoral potency, T helper cell polarization, cytotoxic lymphocyte activation, and germinal center dynamics under equivalent expression conditions. Direct comparisons with monomeric RBD (mRBD) constructs and conventional plasmid DNA vaccines encoding identical immunogens enable a controlled dissection of structure–immunity relationships independent of transcriptional efficiency.

Here, we report a structure-guided DNA vaccine strategy in which trimeric RBD antigen design is functionally integrated with cytoplasmic transcription to improve the quality and balance of adaptive immune responses. We demonstrate that this platform supports robust antigen expression in non-dividing cells, induces robust cellular and humoral immune responses with a favorable Th1 bias, and exhibits a favorable short-term safety profile characterized by minimal systemic inflammation and no detectable hepatic, renal, or cellular toxicity. We used the ancestral Wuhan-Hu-1 RBD as a well-characterized reference antigen to establish a proof-of-concept evaluation of a T7 autogene-driven cytoplasmic transcription system coupled with rational trimeric antigen engineering. Collectively, this work establishes a synergistic design principle in which rational trimeric antigen engineering and direct cytoplasmic transcription cooperate to achieve superior vaccine performance. This strategy provides a flexible and scalable framework for next-generation DNA vaccines targeting rapidly evolving viral pathogens.

2. Materials and Methods

2.1. Construction of tRBD-Producing Cassettes

The construction of T7 autogene cassettes, which are linear double-stranded DNA sequences, followed a previously established protocol [19]. These cassettes incorporate key regulatory components such as Nuclear Import Sequences (NIS), a CMV enhancer/promoter, intervening sequences (IVS), a T7 promoter, an Internal Ribosome Entry Site (IRES) derived from encephalomyocarditis virus, the gene encoding T7 RNA polymerase, a 3′ untranslated region (UTR) from human α-globin, an SV40 polyadenylation signal, and a T7 transcription terminator. T7 autogene cassettes were inserted into the pET32a vector using the Golden Gate Assembly method at the BsaI site. The resulting pET32a vector served as a template for the subsequent amplification of T7 autogene cassettes.

The synthesis of tRBD-producing cassettes was performed using the same methodology. These constructs were designed with a T7 promoter, IVS, IRES, a trimeric RBD (tRBD) region, a human α-globin-derived 3′-UTR, an SV40 polyadenylation signal, and a T7 terminator. The tRBD construct was designed as a single polypeptide comprising three tandem repeats of homologous RBD domains (amino acids 331–525 of the SARS-CoV-2 spike protein, based on the Wuhan-Hu-1 reference strain; GenBank: MN908947.3). This single-chain configuration was selected to facilitate the co-assembly of the trimeric RBD structure [10]. Additionally, an N-terminal signal peptide (MGVKVLFALICIAVAEA) of 17 amino acids was incorporated into the tRBD region to enhance protein expression efficiency. This peptide corresponds to the endogenous signal peptide of Gaussia luciferase, a commonly utilized leader sequence for efficient protein secretion in mammalian cells [23]. A comparable approach was employed to construct mRBD-producing cassettes, which differed from the tRBD-producing cassettes by containing a monomeric RBD (mRBD) instead of the trimeric configuration. The monomeric RBD, comprising amino acids 331–525 of the SARS-CoV-2 spike protein, was incorporated while preserving all other sequence elements and following the same synthesis steps as those used for tRBD-producing cassette generation. Both tRBD- and mRBD-producing cassettes were inserted into the pET32a vector using the same method, serving as templates for subsequent amplification.

To provide a plasmid-based DNA vaccine control, the p_tRBD plasmid was prepared, inspired by the general structural framework of the ZyCoV-D DNA vaccine [24]. The p_tRBD construct utilized a CMV promoter for trimeric RBD expression and contained bacterial replication elements to ensure propagation in prokaryotic hosts. The p_tRBD plasmid was assembled within the pVAX-1 vector (Addgene, Watertown, MA, USA), incorporating an IgE signal peptide sequence, identical to that used in ZyCoV-D, along with the spike gene, a human α-globin-derived 3′-UTR, and an SV40 polyadenylation signal.

2.2. Amplification of DNA Cassettes and Incorporation of Modified Nucleotides

To amplify DNA cassettes by PCR, the reaction was prepared using 0.2 µL of template DNA (100 ng/µL), 2.5 µL each of forward and reverse primers (10 pmol/µL), 10 µL of 5× Phusion GC Enhancer (Thermo Fisher Scientific, Waltham, MA, USA), and 25 µL of the 2× Phusion Plus DNA PCR Master Mix (Thermo Fisher Scientific). The total reaction volume was adjusted to 50 µL by adding distilled water. Thermal cycling began with an initial denaturation at 98 °C for 35 s, followed by 40 cycles of 98 °C for 10 s, 65 °C for 35 s, and 72 °C for 160 s. A terminal extension was conducted at 72 °C for 3 min. Resulting PCR products were cleaned up using the FavorPrep GEL/PCR Purification Mini Kit (Favorgen, Ping Tung, Taiwan).

The T7 autogene cassettes (Rpol) were amplified using the forward primer 5′-G^§A^§G^§A^§T^§A^*T^*A^*C^*A^*TTTGAGATGCATGCTTTGCATACTTCTGC-3′ and the reverse primer 5′-C^§T^§C^§C^§T^§T^*T^*C^*A^*G^*CAAAAAACCCCTCAAGACCCGTTTAGAGG-3′. Similarly, for the amplification of tRBD- and mRBD-encoding cassettes, the forward primer 5′-T^§C^§C^§C^§G^§C^*G^*A^*A^*A^*TTAATACGACTCACTATAGGGATAATG-3′ and the reverse primer 5′-C^§T^§C^§C^§T^§T^*T^*C^*A^*G^*CAAAAAACCCCTCAAGACCCGTTTAGAGG-3′ were used. In these sequences, the symbol ‘§’ represents a 2′-methoxy nucleotide, while ‘*’ indicates a phosphorothioate-modified nucleotide. This terminal modification strategy was employed to improve nuclease resistance, as previously described, and does not alter the coding sequence or transcriptional unit.

2.3. Terminal Deoxynucleotidyl Transferase Reaction

To enhance nuclease resistance at the 3′ end of the DNA cassette, terminal deoxynucleotidyl transferase (TdT) was employed to add 2′-O-methyl guanosine triphosphates (Apexbio, Houston, TX, USA) to the 3′-OH ends. The reaction mixture was prepared in a final volume of 100 µL, containing 2 pmol of PCR-amplified DNA, 20 µL of 5× TdT buffer, 10 µL of 10× CoCl₂, 2 nmol of 2′-methoxy-guanosine-5′-triphosphates, 2 µL of 1 mM GTP, 20 U of TdT enzyme (Thermo Fisher Scientific), and nuclease-free water. Incubation was carried out at 37 °C for 1 h, and the reaction was subsequently stopped by adding 10 µL of 0.5 M EDTA (pH 8.0). The PCR product was resolved by agarose gel electrophoresis, and the band corresponding to the expected amplicon size was excised and purified using gel extraction (FavorPrep GEL/PCR Purification Mini Kit) to remove residual template DNA and nonspecific products. The purity of the gel-purified DNA was confirmed by agarose gel electrophoresis (Figure S1).

2.4. Concentration of DNA Cassettes

All DNA cassettes were concentrated using isopropanol precipitation. Sodium acetate (3.0 M, pH 5.2) was added to the DNA solution to a final concentration of 0.3 M, followed by isopropanol (0.6× total volume). Following centrifugation at 4 °C for 30 min, the supernatant was carefully discarded, and the resulting DNA pellet was rinsed with 500 µL of 70% ethanol. The pellet was left to air-dry at ambient temperature for 10–20 min to eliminate residual ethanol, then resuspended in 100 µL of nuclease-free water with gentle pipetting.

2.5. Preparation and Physicochemical Characterization of Rpol/tRBD/LNP Nanoparticles

Rpol/tRBD/LNP nanoparticles were formulated based on previously reported methods [25,26]. The lipid composition included SM-102, 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), cholesterol, and 1,2-dimyristoyl-rac-glycero-3-methoxypolyethylene glycol-2000 (DMG-PEG2000) (Avanti Polar Lipids, Alabaster, AL, USA), mixed in a molar ratio of 50:10:38.5:1.5. This lipid blend was dissolved in ethanol and subsequently combined with DNA cassettes (T7 autogene cassettes and tRBD-producing cassettes at a 0.44:1 molar ratio) in citrate buffer (pH 4.0). The mixture was prepared at a weight ratio of 100:6 and a volume ratio of 1:3. To fabricate the nanoparticles, the lipid–DNA formulation was extruded using a Mini-extruder (Avanti Polar Lipids) containing Whatman nuclepore track-etched filters (0.08 µm, Merck, Darmstadt, Germany) and a 10 mm polyester backing disc (Cytiva, Marlborough, MA, USA). This extrusion step was repeated ten times to ensure uniform particle formation. The prepared nanoparticles were subjected to dialysis against PBS (pH 7.4) at 4 °C for 12 h using Spectra/Por 3 membranes with a molecular weight cutoff of 3.5 kDa (Thermo Fisher Scientific). Extrusion-based formulation was selected to ensure reproducible nanoparticle assembly under small-scale experimental conditions, consistent with previously reported methods. Rpol/mRBD/LNP nanoparticles were synthesized following the same procedure. For the preparation of p_tRBD/LNP nanoparticles, a lipid mixture was combined with plasmid DNA in citrate buffer (pH 4.0) at a weight ratio of 100:9 and a volume ratio of 1:3. The subsequent steps mirrored those used for tRBD/LNP nanoparticles.

Particle size (Z-average diameter) and polydispersity index (PDI) were measured by dynamic light scattering (DLS) using a Zetasizer (Malvern Panalytical, Malvern, UK). In addition, the zeta potential of the LNPs was measured by electrophoretic light scattering using the same instrument at 25 °C. Zeta potential was measured using a Zetasizer after diluting LNP samples in 10-fold diluted PBS (0.1× PBS) to minimize ionic strength-dependent artifacts while maintaining a defined buffering environment. All measurements were performed in triplicate, and results are presented as mean ± standard deviation. Particle size distributions were reported as number-based distributions to represent the predominant LNP population, as intensity-based distributions may be biased by low-abundance larger particles or minor aggregates. DNA encapsulation efficiency was quantified using the Quanti-IT Picogreen dsDNA Assay Kit (Thermo Fisher Scientific).

2.6. Animal Experiments

Female C57BL/6 mice, aged between 5 and 6 weeks, were obtained from Orient Bio and housed under standard laboratory conditions before experimentation. For vaccination, a total dose of 18.4 µg of Rpol/tRBD/LNP was administered via intramuscular injection. Each mouse received two separate injections of 50 µL per site, targeting the rectus femoris muscles in both hind limbs to ensure uniform distribution, as per previously established protocols [27]. This immunization procedure followed a three-dose regimen, with each dose given at two-week intervals. Control groups were included for comparison, receiving either PBS (100 µL), Rpol/mRBD/LNP (14.0 µg/100 µL), or p_tRBD/LNP (18.4 µg/100 µL). The reported vaccine dose indicates the total mass of the LNP formulation (DNA + lipid components). Specifically, the Rpol/tRBD/LNP formulation (18.4 µg per dose) contained 1.04 µg of Rpol/tRBD DNA and 17.36 µg of lipids; the Rpol/mRBD/LNP formulation (14.0 µg per dose) contained 0.79 µg of Rpol/mRBD DNA and 13.21 µg of lipids; and the p_tRBD/LNP formulation (18.4 µg per dose) contained 1.52 µg of p_tRBD plasmid DNA and 16.88 µg of lipids. The DNA inputs were normalized to an equimolar basis (Rpol/tRBD:Rpol/mRBD:p_tRBD = 1:1:1). Blood was drawn from the submandibular vein and promptly mixed with a 4% citrate solution (at 10% of the total blood volume) to inhibit clot formation. Plasma was subsequently isolated by centrifugation at 2000× g for 10 min, ensuring the removal of cellular components before further analysis. Doses were normalized on a molar basis, and differences in DNA mass reflect the larger molecular size of the trimeric RBD cassette rather than increased functional dosing. PBS was used as the negative control condition in both in vitro and in vivo experiments unless otherwise stated.

2.7. Flow Cytometry Analysis for RBD Expression

HEK293T cells were seeded at a density of 7 × 10⁵ cells per well and co-transfected with 0.69 µg of the T7 autogene cassette and 1.00 µg of the tRBD-expressing cassette or 0.59 µg of the mRBD-producing cassette. For comparison, cells were transfected with 2.5 µg of the p_tRBD plasmid. Transfection was performed using Lipofectamine 3000 (Thermo Fisher Scientific), following the manufacturer’s protocol. Lipofectamine 3000 was used for in vitro transfection to minimize formulation-dependent variability and to allow standardized comparisons of antigen expression between constructs. Transfected cells were cultured for 48 h in a humidified incubator set to 37 °C and 5% CO₂. Following incubation, cells were rinsed with Dulbecco’s phosphate-buffered saline (DPBS; Welgene, Gyeongsan-si, Republic of Korea) and subsequently suspended in FACS buffer (DPBS containing 2.5% fetal bovine serum) for flow cytometry. To assess RBD expression, 1 × 10⁶ transfected cells were incubated with 30 µg/mL of anti-SARS-CoV-2 RBD primary antibodies (Thermo Fisher Scientific) for 30 min on ice. Antibody concentrations were selected based on preliminary titration experiments to ensure saturating detection of surface-expressed RBD. After incubation with the primary antibody, cells were rinsed using FACS buffer and subsequently incubated on ice for 30 min with 8 µg/mL FITC-labeled goat anti-rabbit IgG (BD Biosciences, San Jose, CA, USA). Prior to data acquisition, the cells underwent two washes with FACS buffer. Flow cytometric data were collected using a CytoFLEX instrument (Beckman Coulter, Brea, CA, USA) and processed with CytExpert v2.4 analysis software (Beckman Coulter).

2.8. Transfection of Arrested HEK293T Cells with Rpol/tRBD

To induce cell cycle arrest and a non-dividing state, HEK293T cells were treated with 5 µM Roscovitine (Sigma-Aldrich, St. Louis, MO, USA) and incubated for 48–72 h [28]. The effectiveness of proliferation inhibition was assessed using a Cell Counting Kit-8 (CCK-8) assay, while cell cycle arrest was verified by Propidium Iodide (PI) staining coupled with flow cytometry analysis. HEK293T cells were then plated into 24-well plates at a density of 3 × 10⁴ cells per well. After attachment, the cells were co-transfected with T7 autogene cassettes (0.069 µg) and tRBD-producing cassettes (0.1 µg) or mRBD-producing cassettes (0.059 µg) and incubated for 48 h. As a control, 0.25 µg of p_tRBD plasmid was transfected for comparison. The cells were then harvested for qPCR and flow cytometry analysis to evaluate RBD expression and cellular status. Cell cycle-arrested HEK293T cells were used as a reductionist model to evaluate transcriptional dependence on cell division status, independent of antigen-presenting cell-specific pathways.

2.9. Transfection of Non-Dividing BMDMs with Rpol/tRBD

Bone marrow-derived macrophages (BMDMs) were derived from the bone marrow of femurs and tibias obtained from 6-week-old C57BL/6 mice. Bone marrow cells were flushed with PBS and mononuclear cells were obtained by layering over Histopaque-1077 (Sigma-Aldrich) and centrifuging under density gradient conditions. The cells were cultured for 5 days in Iscove’s modified Dulbecco’s medium (IMDM) containing 10% FBS and 10 ng/mL mouse M-CSF (PeproTech, Cranbury, NJ, USA), with the medium replaced every 2 days. On day 5, adherent macrophages were used for downstream experiments. Trypan blue exclusion and cell counting over 4 days confirmed the cells remained non-dividing. Cell cycle quiescence was confirmed by stable cell counts over time and the absence of proliferative expansion under M-CSF-maintained conditions. Non-dividing BMDMs were co-transfected in 96-well plates with the T7 autogene cassette (0.069 µg) and the tRBD-expressing cassette (0.1 µg) per well using Lipofectamine 3000. For comparison, 0.25 µg of plasmid DNA (p_tRBD) was transfected under identical conditions. Cells were harvested at 24, 48, 72, and 96 h post-transfection, and total RNA was extracted with the CellAmp Direct RNA Prep Kit (Takara, Shiga, Japan). RT-qPCR was performed using the TB Green One Step RT-PCR Kit (Takara) with the following primers: GAPDH (forward 5′-TGAGCAAGAGAGGCCCTATC-3′ and reverse 5′-AGGCCCCTCCTGTTATTATG-3′), RBD (forward 5′-GCCGGTAGCACACCTTGTA-3′ and reverse 5′-ACAGTTGCTGGTGCATGTAG-3′). Relative gene expression levels were calculated using the ΔCt method. For protein expression analysis, BMDMs (1 × 10⁶ cells/well) were seeded in 6-well plates and co-transfected with the T7 autogene cassette (1.3 µg) and tRBD-expressing cassette (1.89 µg), or transfected with p_tRBD plasmid DNA (4.73 µg). Cells were harvested on days 1 to 4 post-transfection, washed with DPBS, and stained with an anti-SARS-CoV-2 RBD primary antibody (Sino Biological, Beijing, China) followed by a FITC-conjugated goat anti-rabbit IgG secondary antibody (BD Biosciences). Samples were analyzed on a CytoFLEX flow cytometer (Beckman Coulter), and fluorescence data were processed using CytExpert software to determine both the proportion of RBD-expressing cells and their mean fluorescence intensity (MFI).

2.10. FACS Analysis of T and B Cell Subsets Following Splenocyte Re-Stimulation

Murine splenocytes were isolated by mechanically dissociating spleens through a 70 µm cell strainer (SPL Life Science, Pocheon-si, Republic of Korea). Red blood cells were removed using eBioscience RBC lysis buffer (Invitrogen, Waltham, MA, USA), followed by washing with PBS. The isolated splenocytes were then maintained in T cell culture medium (RPMI 1640 supplemented with 10% FBS, 1% Penicillin-Streptomycin, and 10 mM β-mercaptoethanol). To prevent nonspecific Fc receptor binding, cells were pre-incubated with TruStain FcX Plus (BioLegend, San Diego, CA, USA) at 4 °C for 20 min before flow cytometry.

To assess antigen-specific immune responses, splenocytes were cultured under standard conditions (37 °C, 5% CO₂, humidified atmosphere) and stimulated for 5 h with recombinant RBD protein (4 μg/mL; Miltenyi Biotec, Bergisch Gladbach, Germany). During the stimulation, BD GolgiPlug (Brefeldin A; BD Biosciences) was added to inhibit protein secretion and enable intracellular cytokine accumulation for flow cytometric analysis. After stimulation, cells were stained at 4 °C for 30 min with fluorophore-labeled monoclonal antibodies against CD3 (clone 17A2), CD4 (RM4-5), CD8a (53-6.7), CD44 (IM7), and CD62L (MEL-14) (all from BioLegend), each diluted to 2 μg/mL in FACS buffer. For intracellular cytokine analysis, cells were fixed using Cyto-Fast Fix/Perm Solution and permeabilized with 10× Cyto-Fast Perm/Wash Solution (BioLegend), followed by intracellular staining with anti-IFN-γ (XMG1.2), anti-IL-4 (11B11), and anti-Granzyme B (GB11) antibodies (BioLegend).

For lymph node cell analysis, the collected inguinal LNs were excised, mechanically disrupted, and passed through a 70 μm cell strainer to generate a single-cell suspension consisting of LN-derived mononuclear cells. For GC B cell analysis, 5 × 10⁵ cells per tube were stained with fluorophore-conjugated antibodies. Fc receptors were blocked with TruStain FcX Plus at 4 °C for 20 min prior to staining. GC B cells were labeled with anti-B220 (RA3-6B2), anti-GL7 (GL7), and anti-CD95 (SA367A8) antibodies (BioLegend), while Tfh cells were stained with anti-CD4 (RM4-5), anti-CXCR5 (L138D7), and anti-PD-1 (29F.1A12) antibodies (BioLegend). Surface staining for all samples was carried out at 4 °C for 30 min. Samples were analyzed on a CytoFLEX flow cytometer (Beckman Coulter), and the resulting data were interpreted using CytExpert software.

For gating, lymphocytes were first selected based on forward and side scatter, followed by exclusion of doublets using FSC-A versus FSC-H. T cells were defined as CD3⁺ cells and subsequently separated into CD4⁺ and CD8⁺ subsets. Within the CD4⁺ or CD8⁺ compartments, CD44⁺ CD62L⁻ cells were defined as an effector/effector-memory-like population. For intracellular staining analyses, cytokine-positive gates were set using negative baseline controls, and GzmB-positive populations were defined using fluorescence-minus-one (FMO) controls and PBS-treated samples as negative references. For lymph node analyses, GC B cells were identified as B220⁺ GL7⁺ CD95⁺ cells, while Tfh cells were defined as CD4⁺ CXCR5⁺ PD-1⁺ cells.

2.11. Cytokine Analysis by ELISA Following Splenocyte Re-Stimulation

Splenocytes (1 × 10⁶ cells/well) from immunized mice were seeded into 96-well plates and re-stimulated with recombinant RBD protein under the same conditions as described above. Additionally, Concanavalin A (5 µg/mL, Sigma-Aldrich) served as a positive control. The levels of secreted cytokines (TNF-α, IFN-γ, IL-6, IL-4, IL-2, and IL-12p40) in culture supernatants were determined using the TNF-alpha ELISA Kit, IFN-gamma ELISA Kit, IL-6 ELISA Kit, IL-4 ELISA Kit (Elabscience, Houston, TX, USA), IL-2 ELISA Kit (BD Biosciences), and IL12 p40/70 ELISA Kit (RayBio, Peachtree Corners, GA, USA) following the manufacturer’s guidelines.

2.12. Determination of Endpoint Antibody Titers Using ELISA

Nunc MaxiSorp ELISA plates (Thermo Fisher Scientific) were coated with SARS-CoV-2 RBD protein (2 μg/mL in PBS) and incubated overnight at 4 °C to allow antigen adsorption. On the following day, unbound antigen was removed, and the wells were blocked with 2.5% (w/v) skim milk prepared in PBS for 1 h at room temperature to minimize nonspecific binding. Plasma samples were subsequently added and incubated for an additional hour. After incubation, wells were washed several times with PBS-T (PBS supplemented with 0.05% Tween-20) to remove residual unbound components. Horseradish peroxidase (HRP)-conjugated anti-mouse IgG secondary antibodies (Thermo Fisher Scientific) diluted 1:10,000 in PBS were then applied to each well and incubated for 1 h at room temperature. Colorimetric signal was developed using 1-Step Ultra TMB-ELISA Substrate Solution (Thermo Fisher Scientific), and the reaction was terminated by adding 2 M sulfuric acid. Absorbance at 450 nm was measured using a microplate spectrophotometer. Endpoint antibody titers were determined from serial dilution curves using a five-parameter logistic (5PL) regression model, with the threshold for positivity defined as the mean of negative control values plus three standard deviations (Mean + 3SD), as calculated with Origin 2022 Software (OriginLab, Northampton, MA, USA).

2.13. Pseudovirus Neutralization Assay

To generate SARS-CoV-2 spike-pseudotyped viruses, HEK293T cells were co-transfected with three plasmids in equal proportions: a lentiviral packaging construct (psPAX2, Addgene), a reporter plasmid encoding NanoLuc luciferase and RFP (pLenti-SFFV-NanoLuc-PGK-RFP-T2A-PURO, Alstem, Richmond, CA, USA), and an envelope plasmid carrying the SARS-CoV-2 spike gene (Addgene), based on the Wuhan-Hu-1 reference strain. After a 48-h incubation, the culture supernatant was collected, concentrated using a Lenti-X concentrator (Takara), and subsequently stored at −80 °C until use. Serial dilutions of heat-inactivated plasma were incubated with pseudotyped viruses at 37 °C for 1 h, followed by addition to HEK293T cells pre-seeded at a density of 1 × 10⁴ cells per well. After 48 h of infection, luminescence signals were quantified using the Nano-Glo luciferase assay system, and measurements were performed with the GloMax Discover System (Promega, Madison, WI, USA).

2.14. Quantification of RBD mRNA

HEK293T cells (3 × 10⁴ per well) were plated in 24-well plates and transfected with a combination of T7 autogene cassettes (0.069 µg) and either tRBD-producing cassettes (0.1 µg) or mRBD-producing cassettes (0.059 µg). As a control, cells were transfected with 0.25 µg of p_tRBD. Lipofectamine 3000 (Thermo Fisher Scientific) was employed for transfection, adhering to the manufacturer’s protocol. Following incubation, cells were harvested and lysed using the CellAmp Direct Prep Kit (Takara), and lysates were stored at −80 °C for subsequent analysis. Quantification of RBD mRNA was carried out using RT-qPCR. For each sample, a 1-µL lysate aliquot was mixed with a 20-µL reaction mixture containing 0.4 µM of specific primers and subjected to amplification using the One Step TB Green PrimeScript RT-PCR Kit II (Takara) on a StepOne Real-Time PCR system (Thermo Fisher Scientific). PCR amplification was performed for 40 cycles, including denaturation at 95 °C for 5 s, annealing at 58 °C for 25 s, and extension at 72 °C for 30 s. A subsequent melting curve analysis was conducted to verify the specificity of the amplification. The primer sequences used for gene expression analysis were as follows: β-actin forward (5′-CACCATTGGCAATGAGCGGTTC-3′), β-actin reverse (5′-AGGTCTTTGCGGATGTCCACGT-3′) RBD forward (5′-GCCGGTAGCACACCTTGTA-3′), and RBD reverse (5′-ACAGTTGCTGGTGCATGTAG-3′). Relative gene expression was determined using the ΔCt method, normalizing to β-actin. Ct values were obtained from triplicate measurements, and all qPCR experiments were performed with both technical and biological triplicates to ensure reproducibility.

2.15. Blood Urea Nitrogen and ALT Analysis

The blood urea nitrogen (BUN) levels were determined using a colorimetric urease–indophenol reaction-based assay. Briefly, 45 μL of phenol reagent containing 1% (w/v) phenol and 0.005% (w/v) sodium nitroprusside was mixed with 70 μL of alkaline reagent composed of 0.5% (w/v) sodium hydroxide and 0.1% active chlorine from sodium hypochlorite. The reaction mixtures were incubated for 50 min, and absorbance was recorded at 630 nm using a microplate reader (Spectra Max, Molecular Devices, San Jose, CA, USA). All measurements were conducted under the reaction conditions specified by the manufacturer. Serum alanine aminotransferase (ALT) activity was quantified using a transaminase-CII kit (Wako Pure Chemical Industries, Osaka, Japan). Absorbance was measured using a microplate reader, and ALT concentrations were calculated based on the supplied calibration standards.

2.16. Statistical Analysis

All data are expressed as the mean ± standard deviation (SD). Statistical analyses were performed using Origin 2022 software (OriginLab). Group comparisons were assessed using either one-way analysis of variance (ANOVA) or an unpaired t-test, depending on the experimental design. For comparisons involving more than two groups, one-way ANOVA was applied, whereas unpaired two-tailed Student’s t-tests were used for comparisons between two groups. Significance levels were defined as follows: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, while “ns” indicates no significant difference.

3. Results

3.1. Design and Construction of T7-Driven DNA Vaccine Cassettes Targeting Trimeric RBD

Self-amplifying DNA vaccine platforms utilizing T7 RNA polymerase have been explored as a strategy to enable direct cytoplasmic mRNA transcription, thereby bypassing the limitations of nuclear-dependent plasmid DNA vaccines [19]. This self-amplifying T7 transcription system, characterized by a rapid polymerase elongation rate (200–260 nt/s), offers enhanced mRNA output in prior studies while retaining the stability and manufacturing advantages of DNA-based delivery.

To apply this general platform to a distinct antigenic context, we designed and constructed DNA cassettes encoding a trimeric form of the receptor-binding domain (tRBD) of SARS-CoV-2 (Figure 1A). Given the critical role of the RBD in viral entry and neutralizing antibody induction, and its emergence as a hotspot for viral mutation and immune evasion, we designed a trimeric RBD configuration to evaluate the impact of multimeric antigen architecture on immunogenic performance.

Previous studies have reported that trimeric RBD constructs more closely mimic the native spike trimer conformation and can enhance B cell receptor cross-linking and germinal center formation [29]. Accordingly, our focus was to investigate how this optimized multimeric antigen structure would perform when delivered via a T7-driven cytoplasmic transcription system.

The DNA cassettes were designed using a dual-promoter configuration. The CMV promoter initiates nuclear transcription of T7 RNA polymerase, facilitated by a nuclear import sequence (NIS) at the 5′ end. The translated T7 polymerase then drives cytoplasmic transcription from the downstream T7 promoter, resulting in self-amplification. The cassette also contains intervening sequences (IVS) and an internal ribosome entry site (IRES) in the 5′ UTR to support efficient translation in the absence of a 5′ cap [30]. A human α-globin-derived 3′-UTR and an SV40 polyA signal were included to support transcript stability and translation [14].

For antigen expression, a tRBD-producing cassette was constructed under control of the pT7 promoter. This construct encoded a tandem trimer of the RBD (amino acids 331–525 from the Wuhan-Hu-1 strain), preceded by a Gaussia luciferase signal peptide and followed by stabilizing UTR elements and a T7 terminator. As a comparative control, a monomeric RBD (mRBD) version was also constructed using the same regulatory architecture.

All DNA cassettes were PCR-amplified as linear DNAs and were end-modified by TdT to improve nuclease resistance (Section 2). Additionally, we generated a plasmid-based control (p_tRBD) encoding the same trimeric RBD immunogen using a CMV promoter and pVAX-1 backbone, structurally analogous to previously approved DNA vaccine designs [24].

3.2. Formulation and Characterization of LNP-Encapsulated Rpol/tRBD

For in vivo immunogenicity assessment, we synthesized Rpol/tRBD/LNP nanoparticles using an SM-102 (ionizable lipid)-based lipid nanoparticle (LNP) platform, similar to the formulation employed in the mRNA-1273 vaccine (Moderna COVID-19 vaccine). The uptake of nanoparticles by antigen-presenting cells (APCs) is influenced by their size and surface properties, with virus-sized particles (20–200 nm) exhibiting enhanced endocytosis and intracellular delivery [31,32].

Dynamic light scattering revealed mean hydrodynamic diameters of 164.6 ± 2.2 nm for Rpol/tRBD/LNP and 158.4 ± 4.1 nm for Rpol/mRBD/LNP (Figure 1B), indicating only a modest size difference between the two formulations despite the longer coding sequence of the tRBD construct. Both formulations exhibited mildly to moderately positive surface charges (+6.2 to +24.6 mV), consistent with physicochemical properties favorable for cellular association and uptake. The polydispersity index (PDI) values of all DNA/LNP nanoparticles ranged from 0.1 to 0.2, indicating a uniform size distribution. The encapsulation efficiency of the Rpol/tRBD/LNP formulation, measured at an LNP-to-DNA weight ratio of 100:6, was 91.7%, suggesting highly efficient packaging of the DNA vaccine construct. The hydrodynamic diameter of blank LNPs (without DNA payload) was also measured and found to fall within a comparable nanoscale range, indicating that overall particle size is primarily determined by lipid composition and formulation parameters rather than by the absolute length of the encapsulated DNA. Collectively, these results demonstrate that both mRBD and tRBD constructs are packaged into LNPs with highly comparable physicochemical properties, minimizing the potential for particle size-dependent differences in cellular interactions.

3.3. Efficient Cytoplasmic Expression of RBD mRNA and Protein Using Rpol/tRBD DNA Vaccine Platform

Given the previously reported efficiency of T7 RNA polymerase-driven transcription, we next evaluated RBD mRNA synthesis in HEK293T cells following co-transfection with Rpol and tRBD cassettes at a molar ratio of 0.44 to 1. Over the four-day period post-transfection, RT-qPCR analysis demonstrated a marked increase in RBD mRNA levels following co-transfection with Rpol/tRBD (Figure 1C). Quantitatively, RBD mRNA levels were approximately 50.4-fold higher than those observed with the p_tRBD plasmid control, which depends on CMV promoter-driven transcription by the host cell’s nuclear RNA polymerase. Notably, the tRBD cassette—comprising a tandemly repeated RBD gene within a single transcript—demonstrated an approximately 5.9-fold higher mRNA level than its monomeric counterpart (mRBD) under identical transfection conditions. To ensure valid comparisons, molar inputs of all DNA constructs were carefully matched among the Rpol/tRBD, Rpol/mRBD, and p_tRBD groups. Because the tRBD transcript contains repeated RBD sequences, qPCR primers were deliberately designed to target a single, non-repetitive region within the RBD coding sequence that is present only once per transcript and identical between the mRBD and tRBD constructs. Accordingly, each mRNA molecule generated only a single amplification signal, confirming that the elevated mRNA levels observed for the tRBD construct reflect a true increase in transcript abundance rather than a primer-induced amplification artifact.

T7 RNA polymerase exhibits a uniquely high transcriptional speed (200–260 nt/s) with strict promoter specificity, operating largely independent of the complex regulatory constraints typical of eukaryotic RNA polymerase II [33,34]. Consistent with this property, robust mRNA synthesis was observed in both Rpol/tRBD and Rpol/mRBD constructs. In contrast, transcription from the p_tRBD plasmid—mediated by the host nuclear machinery—produced RBD mRNA at approximately 0.74-fold of β-actin expression on day 1, followed by a progressive decline from day 2 onward. These observations are consistent with reduced transcriptional output from the nuclear plasmid system over time under the tested conditions.

Flow cytometric quantification of RBD protein expression showed that expression in the Rpol/tRBD group peaked on day 3 and declined by day 4 (Figure 1D). Cumulative protein expression over four days, as measured by area under the curve (AUC), indicated that the Rpol/tRBD group exhibited 1.71-fold higher expression than the p_tRBD group, confirming enhanced overall gene expression output. Moreover, Rpol/tRBD showed a 1.31-fold increase in protein expression relative to Rpol/mRBD, further highlighting the benefits of multimeric antigen design within this cytoplasmic transcriptional framework. Together, these data indicate higher cumulative antigen expression from the Rpol/tRBD construct compared with monomeric and plasmid-based controls under identical conditions.

To verify the structural integrity of the expressed antigens and exclude the possibility of protein misfolding or intracellular degradation, Western blot analysis was performed using anti-RBD antibodies. As shown in Figure 1E, cells transfected with Rpol/mRBD exhibited a dominant band corresponding to the expected molecular weight of monomeric RBD (~30 kDa), whereas Rpol/tRBD-transfected cells displayed a distinct band at the higher molecular weight consistent with intact trimeric RBD. Importantly, no prominent lower-molecular weight degradation products were detected in either group, indicating that both monomeric and trimeric RBD proteins were properly translated and stably maintained within cells. These results confirm that rapid T7-driven cytoplasmic transcription does not induce abnormal antigen processing and that the trimeric RBD undergoes correct intracellular folding and stability.

3.4. Rpol/tRBD Enables Robust Gene Expression in Low-Mitotic Cellular Environments

Conventional DNA vaccines are commonly delivered by intramuscular, intradermal, or subcutaneous routes, targeting local cell populations such as keratinocytes, muscle cells, and antigen-presenting cells (APCs) in proximity to the injection site [31]. Among these target cells, muscle cells and APCs—including dendritic cells and macrophages—generally display low proliferative activity. After differentiation, these cells typically exit the cell cycle or exhibit limited proliferative capacity, particularly in tissues where their primary function is antigen uptake and presentation [35]. This inherently low mitotic rate limits the efficiency of plasmid-based DNA vaccines, which often depend on nuclear entry mechanisms that are more effective in dividing cells [36]. Therefore, bypassing the requirement for cell division is an important consideration in improving DNA vaccine delivery to physiologically relevant target cells.

To address this issue, we assessed whether the Rpol/tRBD vaccine platform could enable efficient RBD gene expression in non-dividing cells, thereby overcoming the transcriptional barriers encountered with conventional plasmid-based DNA vaccines. Accordingly, we assessed the expression performance of the trimeric RBD construct under mitotically inactive conditions. Human HEK293T cells were employed as a representative low-mitotic cell system by pharmacologically inducing cell cycle arrest with roscovitine, a cyclin-dependent kinase (CDK) inhibitor that suppresses mitotic activity [37]. Although HEK293T cells are not directly involved in vaccine-induced immune responses, they serve as a suitable system for evaluating transgene expression under mitotic inhibition. Flow cytometry analysis confirmed that roscovitine treatment resulted in an accumulation of cells in the G0/G1 phase, with a corresponding decrease in S and G2/M populations (Figure 2A). Annexin V/PI staining further confirmed that this arrest did not trigger significant apoptosis, indicating preserved cellular viability under non-dividing conditions (Figure 2B).

Under these non-dividing conditions, RT-qPCR analysis revealed that HEK293T cells co-transfected with Rpol/tRBD exhibited RBD mRNA levels approximately 19.6-fold relative to the β-actin-normalized baseline over a four-day period post-treatment (Figure 2C). Notably, RBD mRNA expression in the Rpol/tRBD group was approximately 150.9-fold higher than in cells transfected with the conventional p_tRBD plasmid, demonstrating the superior transcriptional output of the Rpol/tRBD system in non-dividing cells. Flow cytometry also showed that by day 2 post-transfection, about 45.2% of Rpol/tRBD-transfected cells expressed detectable RBD protein, compared to only 8.7% in the p_tRBD-transfected group (Figure 2D).

To further validate performance in a physiologically relevant low-mitotic system, we evaluated bone marrow-derived macrophages (BMDMs), which exhibit minimal proliferative activity. On day 1 post-transfection, RT-qPCR analysis showed that RBD mRNA expression in the Rpol/tRBD group reached levels comparable to endogenous GAPDH expression, while expression from p_tRBD-transfected cells remained below 10% of that level (Figure 2E). Although RBD mRNA levels gradually declined in both groups, the Rpol/tRBD group maintained substantially higher transcript levels through day 4. Flow cytometry further indicated that RBD protein expression in Rpol/tRBD-transfected BMDMs remained approximately 2.3-fold higher than that of the p_tRBD group throughout the experimental period (Figure 2F). BMDMs remained mitotically quiescent during this time, as indicated by stable cell counts and high viability assessed by Trypan blue exclusion. Together, these findings demonstrate that the Rpol/tRBD platform supports sustained RBD mRNA and protein expression in non-dividing and low-mitotic cells, achieving substantially higher expression than the plasmid-based DNA vaccine under mitotic constraint.

3.5. T Effector and Cytotoxic T Cell Responses

Cell-mediated immunity is a critical determinant of protective efficacy for SARS-CoV-2 vaccines. To assess cellular immune responses induced by Rpol/tRBD vaccination, spleens were collected from C57BL/6 mice on day 17 after the second booster immunization. Splenocytes were re-stimulated ex vivo with recombinant RBD antigen and analyzed by flow cytometry.

The Rpol/tRBD group exhibited a significantly increased frequency of CD8⁺ CD44⁺ CD62L⁻ effector/effector-memory-like T cells compared with PBS-treated mice, as well as with the Rpol/mRBD and p_tRBD groups (Figure 3A). A parallel increase was observed in CD4⁺ CD44⁺ CD62L⁻ effector/effector-memory-like T cells, indicating that Rpol/tRBD vaccination enhances both CD8⁺ and CD4⁺ effector/effector-memory-like T cell differentiation (Figure 3B). Because all vaccine groups were administered at equivalent molar doses, these differences can be attributed to antigen architecture rather than DNA input. Consistent with prior reports demonstrating superior immunogenicity of trimeric RBD antigens relative to monomeric forms [9,10,11,29], endogenous expression of tRBD via the Rpol platform resulted in markedly enhanced effector/effector-memory-like T cell responses.

In addition to effector/effector-memory differentiation, Rpol/tRBD vaccination induced robust cytotoxic T lymphocyte (CTL) activity. Intracellular IFNγ staining revealed a significantly higher frequency of IFNγ⁺ CD8⁺ T cells in the Rpol/tRBD group than in the PBS, Rpol/mRBD, and p_tRBD groups (Figure 3C). Functional cytotoxicity was further evaluated by assessing CD107a surface mobilization and intracellular granzyme B (GzmB) expression. Analysis of GzmB⁺ CD107a⁺ CD8⁺ T cells demonstrated a pronounced increase in cytotoxic activity following Rpol/tRBD vaccination relative to all control groups (Figure 3D).

Notably, GzmB expression in CD8⁺ T cells displayed a continuous fluorescence intensity distribution rather than a discrete bimodal pattern, consistent with in vivo antigen-driven cytotoxic activation. GzmB-positive populations were therefore defined using fluorescence-minus-one (FMO) controls and PBS-treated samples as negative baselines. Using this standardized gating strategy, Rpol/tRBD induced significantly stronger antigen-specific CD8⁺ T cell cytotoxic responses than both Rpol/mRBD and p_tRBD. Together, these findings demonstrate that cytoplasmic expression of a structurally optimized trimeric RBD promotes robust T effector/effector-memory differentiation and functional CTL responses, underscoring the importance of multivalent antigen design for eliciting effective cellular immunity.

3.6. Promotion of Germinal Center B Cell and Tfh Cell Activation

Germinal center (GC) B cells are essential for affinity maturation, somatic hypermutation, and class switching, ultimately supporting the generation of high-quality antigen-specific antibody responses [38]. These cells further differentiate into memory B cells and long-lived plasma cells, underscoring the importance of GC reactions in effective humoral immunity. T follicular helper (Tfh) cells play a central role in regulating GC responses by providing key signals that promote B cell activation, selection, and maturation within germinal centers [39].

Compared with PBS-treated controls, Rpol/tRBD vaccination resulted in a significant increase in the frequency of GC B cells, accompanied by a modest expansion of Tfh cells (Figure 4A,B). Although the Rpol/tRBD group showed trends toward higher GC B cell and Tfh cell frequencies relative to the Rpol/mRBD and p_tRBD groups, these differences did not reach statistical significance. Nevertheless, the concurrent increase in GC B cells and Tfh cells is consistent with enhanced germinal center activity, given the established role of Tfh cells in supporting GC formation and antibody maturation.

3.7. Th1-Biased but Balanced Immune Response

Because excessive Th2-skewed immune responses have been associated with vaccine-associated enhanced respiratory disease (VAERD), evaluation of Th1/Th2 polarization represents an important safety consideration for SARS-CoV-2 vaccines [29,40]. To assess immune polarization induced by Rpol/tRBD vaccination, intracellular cytokine staining was performed to quantify IFNγ (Th1-associated cytokine) and IL-4 (Th2-associated cytokine) expression in CD4⁺ T cells.

The Rpol/tRBD vaccine group exhibited an IFNγ/IL-4 ratio greater than one, indicating a predominantly Th1-skewed immune response (Figure 4C–E). In contrast, both the Rpol/mRBD and p_tRBD groups displayed IFNγ/IL-4 ratios below one, consistent with a Th2-biased immune profile. Importantly, despite the Th1 dominance observed in the Rpol/tRBD group, IL-4-producing CD4⁺ T cells remained detectable, indicating preservation of Th2-associated responses. These results demonstrate that Rpol/tRBD vaccination induces a Th1-dominant yet balanced immune profile, rather than extreme polarization toward either axis.

3.8. Th1-Biased Cytokine Profile Without Excess Inflammatory Responses

To assess cytokine production following antigen recall, splenocytes harvested from vaccinated mice were re-stimulated in vitro with recombinant SARS-CoV-2 RBD protein, and cytokine concentrations in culture supernatants were measured by ELISA on day 5 after re-stimulation. As shown in Figure 5A, both Rpol/tRBD and Rpol/mRBD vaccination significantly increased the secretion of Th1-associated cytokine IL-2 compared with PBS-treated controls. In contrast, the levels of other canonical Th1 cytokines, including IFNγ and TNFα, remained largely unchanged in both groups.

Splenocytes from p_tRBD-vaccinated mice exhibited a distinct cytokine profile characterized by markedly elevated secretion of TNFα and IL-12p40, together with increased levels of IL-2, IL-4, and IL-6. This broader cytokine induction pattern contrasts with the more focused response observed following Rpol/tRBD vaccination. While transient cytokine induction can contribute to immune priming, elevated production of pro-inflammatory mediators such as TNFα and IL-6 is commonly associated with increased systemic inflammatory stress.

In comparison, Rpol/tRBD immunization induced robust IL-2 production without significantly altering the IL-4 or IL-6 levels, consistent with a Th1-skewed cytokine profile with limited Th2 or pro-inflammatory involvement. Although Rpol/tRBD and Rpol/mRBD elicited largely similar cytokine patterns, both differed substantially from the p_tRBD group, which displayed a broader and less regulated cytokine response.

To further evaluate systemic tolerability, body weight was monitored longitudinally following vaccination (Figure 5B). Mice immunized with p_tRBD showed a transient decrease in body weight before gradual recovery, whereas Rpol/tRBD-vaccinated mice maintained stable body weight throughout the observation period. This pattern is consistent with the differential cytokine profiles observed among the vaccine groups.

Hepatic safety was assessed by measuring serum alanine aminotransferase (ALT) levels at days 1, 3, and 6 after the final immunization (Figure 5C). ALT values in PBS- and LNP-only-treated mice remained within a low and stable range. Similarly, Rpol/tRBD-vaccinated mice exhibited only minor ALT fluctuations without significant differences from control groups. In contrast, the p_tRBD group showed a progressive increase in ALT levels at later time points, which was significantly higher than that observed in the Rpol/tRBD and LNP-only groups. Renal safety was evaluated by measuring the serum blood urea nitrogen (BUN) levels (Figure 5D), which remained within the normal physiological range across all treatment groups.

To examine early systemic inflammatory responses, expression of representative inflammatory and co-stimulatory genes was analyzed by RT-qPCR using whole blood collected 24 h after booster immunization (Figure 5E). At this early post-boost time point, Rpol/tRBD-vaccinated mice displayed reduced expression of multiple inflammation-associated genes, including Cd40, Cd86, Tgfbr1, Tgfb1, and Cd40lg, relative to PBS-treated controls, indicating limited activation of systemic inflammatory signaling.

Finally, potential cytotoxic effects associated with high-level cytoplasmic transcription were evaluated in vitro. Cell viability assays performed in Huh7, HepG2, and HEK293T cells following Rpol/tRBD/LNP treatment revealed no significant reduction in viability compared with p_tRBD/LNP or the PBS controls (Figure 5F). Cell viability remained above approximately 90% across all tested conditions, indicating that T7 autogene-driven cytoplasmic transcription does not induce detectable cytotoxicity under the formulation and dosing conditions used in this study.

3.9. Humoral and Neutralizing Antibody Responses

To evaluate the levels of IgG antibodies targeting the SARS-CoV-2 RBD, we conducted an ELISA assay using plasma samples from immunized mice. The ELISA plates were coated with RBD protein derived from the prototype strain of SARS-CoV-2. Administration of the Rpol/tRBD vaccine in a prime-boost regimen elicited a robust IgG response by day 12 after the second boost, significantly exceeding those observed in the PBS group (Figure 6A). The geometric mean titer (GMT) of anti-RBD IgG in mice vaccinated with Rpol/tRBD was significantly elevated to 8710, representing a 30.2-fold increase over the GMT induced by the p_tRBD vaccine. Furthermore, Rpol/tRBD immunization resulted in a notable enhancement of antibody titers, with a 24.5-fold higher GMT compared to that induced by Rpol/mRBD vaccination. Since all groups received equivalent molar doses of antigen, these findings highlight that the trimeric RBD vaccine exhibited significantly enhanced immunogenicity against the RBD protein compared to both the monomeric RBD vaccine and the conventional plasmid DNA vaccine (p_tRBD).

As the RBD constitutes the main epitope recognized by neutralizing antibodies in convalescent individuals [4,5], we assessed the neutralizing potential of vaccine-elicited antibodies using a SARS-CoV-2 pseudovirus assay conducted on day 12 following the second booster dose. Plasma from Rpol/tRBD-immunized mice exhibited potent inhibition of pseudovirus entry into hACE2-expressing cells, significantly outperforming the Rpol/mRBD and p_tRBD vaccine groups (Figure 6B). Specifically, Rpol/tRBD vaccination led to a significant increase in neutralizing antibody GMT against the prototype strain, reaching 645, which corresponded to a 31.8-fold and 35.0-fold increase compared to p_tRBD and Rpol/mRBD vaccination, respectively.

Together, these data demonstrate that the Rpol/tRBD vaccine induces antibodies with both strong RBD-binding capacity and effective neutralizing activity, indicating that trimeric antigen architecture combined with cytoplasmic expression enhances humoral immune responses compared with monomeric RBD and conventional plasmid DNA vaccine formats.

4. Discussion

In this study, we investigated the immunogenic properties of a DNA vaccine encoding a trimeric receptor-binding domain (tRBD) of SARS-CoV-2, delivered via a cytoplasmically active T7 autogene-based transcription platform. Previous studies have shown that trimeric or heterotrimeric RBD constructs used in protein subunit vaccine systems more closely resemble the native spike trimer, thereby improving antigen stability and promoting effective B cell receptor cross-linking [9,10,11,29]. These approaches have largely focused on maximizing neutralizing antibody responses, often through sequence diversification or mosaic antigen designs to broaden variant coverage. Despite the apparent similarity in antigen architecture, the present study addresses a distinct and vaccine-relevant question by examining how multivalent RBD design functions within a cytoplasmic transcription-based DNA vaccine platform, where antigen is expressed endogenously. This framework allows for the evaluation of how antigen architecture shapes not only humoral immunity but also qualitative features of adaptive immune responses, including CD8⁺ T cell activation and Th1/Th2 polarization, which are critical determinants of antiviral vaccine efficacy.

A key distinction between the present work and prior trimeric or heterotrimeric RBD vaccine studies lies in the vaccine platform itself. Whereas previous reports have relied on protein subunit-based vaccination strategies involving exogenous delivery of pre-assembled antigens, the DNA vaccine platform employed here enables intracellular antigen production within host cells. This feature has important implications for antigen processing and presentation pathways that shape downstream immune responses.

This platform distinction has direct immunological consequences. Protein subunit vaccines primarily depend on extracellular antigen uptake and MHC class II-restricted presentation, which can efficiently induce antibody responses but often require potent adjuvants to achieve robust cellular immunity [41]. In contrast, endogenous antigen expression achieved through DNA vaccination provides intrinsic access to both MHC class I and class II presentation pathways, supporting coordinated induction of cytotoxic CD8⁺ T cell responses alongside antibody production. Consistent with this platform property, trimeric RBD expression via cytoplasmic transcription in the present study was associated with enhanced CD8⁺ T effector/effector-memory and cytotoxic T lymphocyte activation, alongside robust humoral responses.

Although the present study does not include a direct head-to-head comparison with protein subunit vaccines, the observed immune profile highlights functional attributes that are particularly relevant for vaccine efficacy against viral infections. Within this context, trimeric RBD architecture expressed endogenously via cytoplasmic transcription appears to extend the immunological impact of trimeric antigen design beyond antibody avidity alone, supporting a more comprehensive antiviral immune response. Collectively, these findings indicate that integrating multivalent RBD design with a cytoplasmic transcription-based DNA vaccine platform provides functional advantages that are not readily captured by protein subunit-based trimeric RBD vaccination strategies.

Consistent with these platform- and architecture-dependent considerations, the immunological outcomes observed across Figure 3, Figure 4 and Figure 5 collectively indicate that Rpol/tRBD vaccination promotes a coordinated adaptive immune profile that evolves over the course of repeated immunization. Rather than reflecting a transient or isolated enhancement of a single immune component, the observed responses suggest progressive engagement and refinement across multiple arms of adaptive immunity. As shown in Figure 3, cytoplasmic expression of the trimeric RBD was associated with robust activation of CD8⁺ T effector/effector-memory cells and cytotoxic lymphocytes, consistent with effective cellular immune induction enabled by endogenous antigen expression and access to the MHC class I presentation pathway. In parallel, Figure 4 demonstrates that this cellular activation occurred within a regulated Th1-biased environment, as reflected by elevated IFNγ/IL-4 ratios while preserving measurable Th2-associated responses that support B cell function. Within this immunological context established through booster immunizations, a modest increase in T follicular helper (Tfh) cell frequencies was observed. Given the established role of Tfh cells in supporting germinal center reactions and B cell affinity maturation [39,42], the concurrent trend toward increased germinal center (GC) B cell populations suggests coordinated humoral organization accompanying cellular immune activation. Together with the cytokine and safety profiles presented in Figure 5, these findings indicate that repeated administration of trimeric RBD via a cytoplasmic transcription-based DNA vaccine platform supports structured and balanced immune engagement across cellular and humoral compartments, integrating cytotoxic T cell activity with regulated helper T cell polarization and germinal center-associated B cell responses, without evidence of excessive polarization or systemic inflammatory burden.

A limitation of this study is that a dedicated biodistribution analysis of the SM-102-based LNP formulation was not performed. While the systemic safety readouts presented in Figure 5 did not indicate overt adverse effects under the tested conditions, comprehensive organ-level evaluations—such as biodistribution and clearance profiling, as well as histopathological analyses of major organs—were not included in the current study. Accordingly, potential tissue-level toxicities cannot be fully ruled out at this stage, and future studies incorporating these assessments will be important to further validate the safety profile and support the translational development of the Rpol/tRBD/LNP vaccine platform.

In addition to its distinction from protein subunit-based vaccines, the cytoplasmic transcription-based DNA vaccine platform used in this study differs fundamentally from conventional IVT mRNA vaccine strategies. IVT mRNA vaccines enable rapid and high-level antigen expression shortly after delivery; however, their efficacy is constrained by the intrinsic chemical instability of RNA, susceptibility to degradation, and the generation of double-stranded RNA by-products during in vitro transcription, which can trigger excessive innate immune activation and necessitate extensive purification as well as stringent cold-chain storage during distribution [43]. In contrast, the T7 autogene-driven DNA platform relies on a chemically stable DNA template that supports intracellular RNA generation without dependence on exogenous IVT processes, thereby reducing concerns related to dsRNA contaminants, translational shutdown, and cold-chain logistics. Unlike conventional plasmid DNA vaccines that require nuclear entry for transcription, the T7 autogene system enables direct cytoplasmic RNA synthesis, circumventing a major efficiency barrier associated with nuclear trafficking. In the present study, cytoplasmic transcription from a DNA template facilitated efficient intracellular RNA production while avoiding several stability- and handling-related limitations inherent to mRNA-based vaccines. Beyond these expression-related features, the platform offers practical advantages in formulation robustness and storage under standard refrigeration conditions, which may improve accessibility in low- and middle-income countries [14,44,45]. Moreover, by minimizing nuclear translocation, the T7-driven DNA vaccine platform may reduce theoretical risks associated with genomic integration, potentially alleviating certain regulatory concerns linked to conventional plasmid DNA vaccination.

Importantly, DNA dosing in the in vivo studies was controlled on a molar basis. Although the absolute mass of Rpol/tRBD DNA administered was higher than that of Rpol/mRBD, this reflects the larger molecular size of the trimeric RBD cassette rather than an increased functional dose. All experimental groups received equivalent molar amounts of transcriptional units, ensuring that the observed immunological differences are attributable to antigen architecture rather than to DNA dose effects.

In the present study, neutralizing activity was evaluated only against the ancestral Wuhan-Hu-1 strain, and neutralization against variants of concern (VOCs), such as Omicron or Delta, was not assessed. Therefore, the breadth of cross-neutralization and translational relevance to currently circulating strains cannot be fully determined from the available data. Future studies employing VOC-matched antigen designs and cross-neutralization profiling will be important to validate the adaptability and protective potential of the Rpol/tRBD/LNP platform against rapidly evolving viral variants.

To contextualize the magnitude of the humoral responses induced by the Rpol/tRBD/LNP platform, it is informative to compare our neutralizing antibody levels with published mouse studies using lipid nanoparticle (LNP)-formulated IVT-mRNA vaccines. In the present study, Rpol/tRBD/LNP induced a prototype-strain pseudovirus neutralizing ID50 GMT of 645 (Figure 6B). In contrast, an LNP-encapsulated SARS-CoV-2 RBD-mRNA vaccine was reported to elicit higher neutralizing titers in mice and broad cross-neutralization against multiple variants, including NT50 values of approximately 1:7500–10,000 for K417N/T–E484K–N501Y-containing variants [46]. Although such cross-study comparisons are inherently influenced by differences in antigen design, dose, route, boosting schedules, assay platforms, and sampling time points, these literature benchmarks help frame the current performance of the cytoplasmic transcription-driven DNA vaccine platform. Notably, the goal of this work was to establish a mechanistically grounded improvement over conventional plasmid DNA vaccination and evaluate the in vivo immunogenicity and tolerability of the Rpol/tRBD/LNP strategy. A controlled head-to-head comparison with an IVT-mRNA/LNP benchmark vaccine will be an important next step in future studies to more rigorously assess relative potency, durability, and breadth.

Despite the pronounced increase in intracellular RBD mRNA availability achieved by the T7 autogene-driven cytoplasmic transcription system, protein output did not increase proportionally, indicating a non-linear relationship between transcript abundance and antigen expression. Notably, in Figure 1, the Rpol/tRBD group yielded an ~50.4-fold increase in RBD mRNA relative to the plasmid control, whereas cumulative protein expression increased more modestly (~1.71-fold by AUC), highlighting a clear transcription–translation mismatch in this setting. A similar pattern was observed under low-mitotic conditions, where Rpol/tRBD sustained markedly higher RBD mRNA levels than p_tRBD across multiple time points, while protein expression remained elevated to a more moderate extent (Figure 2C–F). Together, these findings indicate that protein production in the cytoplasmic transcription system is constrained at the post-transcriptional level rather than by transcriptional efficiency alone. Compared with the highly optimized 5′ cap–dependent translation used in IVT mRNA vaccines, IRES-mediated translation is intrinsically less efficient, which may limit translational throughput despite abundant mRNA availability. In addition, excessive cytoplasmic mRNA accumulation may lead to translational saturation rather than proportional increases in protein synthesis. These observations further suggest that antigen yield in cytoplasmic transcription-based DNA vaccines may be improved through optimization of coding sequences, signal peptides, and intracellular expression contexts in future refinements [9,47,48]. To further optimize translational efficiency, delivery, and durability, future refinements may include improving IRES-mediated translation, minimizing immunogenic CpG motifs, and optimizing LNP formulations to enhance endosomal escape and intracellular stability [49]. In addition, heterologous prime-boost strategies may further amplify and sustain both humoral and cellular immunity [50].

An important safety consideration for RNA-generating vaccine platforms is the formation of double-stranded RNA (dsRNA), which can provoke excessive innate immune activation and inflammatory reactogenicity [20,22,51]. In IVT mRNA vaccines, dsRNA by-products generated during in vitro transcription are a well-recognized source of type I interferon induction, often necessitating additional purification steps to mitigate these effects [14,52,53]. In contrast, the T7 autogene-driven cytoplasmic transcription system produces antigen-encoding RNA intracellularly in a unidirectional and promoter-specific manner, thereby limiting exposure to exogenous dsRNA contaminants. Consistent with this mechanistic distinction, innate immune profiling in the present study indicates that Rpol/tRBD vaccination elicits a controlled innate immune response that supports antigen-presenting cell activation without evidence of excessive systemic inflammation, as reflected by restrained cytokine induction, stable body weight, and the absence of detectable hepatic or renal toxicity. These observations suggest that cytoplasmic transcription enables efficient antigen expression within a regulated immunostimulatory range, providing a platform context in which innate immune engagement is sufficiently activated to support adaptive immunity without incurring the inflammatory burden that can accompany IVT mRNA-based approaches [14,54,55]. This controlled innate immune environment provides an important foundation for interpreting the balanced adaptive immune responses observed following Rpol/tRBD vaccination.

Balanced Th1/Th2 immune polarization represents a critical safety consideration for vaccine-induced immunity, as excessive skewing toward either axis has been associated with immune-mediated pathology, including vaccine-associated enhanced respiratory disease (VAERD) and inflammatory tissue damage [56,57]. Th1 responses are essential for antiviral defense through cytotoxic T cell activation, whereas disproportionate Th2 responses have been implicated in adverse outcomes in prior coronavirus and RSV vaccine studies. In the present study, Rpol/tRBD vaccination elicited a predominantly Th1-skewed immune profile while maintaining measurable Th2-associated responses, indicative of a regulated adaptive immune environment rather than extreme polarization. This immune balance provided a rationale for subsequent evaluation of systemic inflammatory responses and short-term safety parameters. Although longer-term safety and durability were not assessed, the current data are consistent with a favorable short-term safety profile characterized by effective immunogenicity without excessive reactogenicity.

Although the present study includes an analysis of splenic T-cell phenotypes and demonstrates an increased CD44⁺ CD62L⁻ T-cell population consistent with an effector/effector-memory-like phenotype following vaccination, long-term immune memory was not directly evaluated at extended time points. Specifically, the T-cell phenotyping was performed at an early post-boost time point (Day 37; 17 days after the second boost), which primarily captures peak responses rather than established durable memory. Accordingly, the persistence of vaccine-induced immune memory, including central memory (CD44⁺ CD62L⁺) compartment formation and long-term recall potential, cannot be definitively determined from the current dataset. Future studies incorporating later longitudinal follow-up and expanded memory T-cell subset profiling (e.g., detailed characterization of effector memory and central memory populations using additional markers such as CD127) will be important to more rigorously assess immune durability.

In addition to systemic immune regulation, a potential concern associated with cytoplasmic T7-driven transcription is that high levels of intracellular RNA and protein synthesis could impose translational burden or cellular stress, potentially resulting in cytotoxicity. To address this possibility, in vitro cell viability was assessed following Rpol/tRBD/LNP treatment. No detectable cytotoxic effects were observed in Huh7, HepG2, or HEK293T cells, with cell viability remaining comparable to that of p_tRBD/LNP- and PBS-treated controls (Figure 5F). These findings indicate that under the formulation and dosing conditions tested, enhanced cytoplasmic transcription mediated by the T7 autogene system does not compromise cellular viability. When considered together with the absence of hepatotoxicity, nephrotoxicity, or excessive systemic inflammatory responses observed in vivo, the data support the conclusion that the cytoplasmic transcription-based DNA vaccine platform is well tolerated and does not induce detectable short-term cytotoxic effects across both in vitro and in vivo settings.

While Lipofectamine-mediated transfection was used in vitro to provide a robust and standardized comparison of construct-level expression, LNP-mediated delivery can introduce additional variability arising from formulation-dependent uptake and endosomal trafficking; thus, direct in vitro characterization of Rpol/tRBD/LNP remains an important area for future investigation. In addition, a limitation of this study is that vehicle-only controls (Lipofectamine-only in vitro and empty LNP in vivo) were not included. Therefore, while the observed responses reflect the overall performance of the formulated delivery and antigen expression system, the contribution of delivery vehicle-associated effects (e.g., transfection-related stress or innate immune activation) cannot be fully separated from antigen-specific effects in the current experimental design. Future studies incorporating Lipofectamine-only and empty LNP control groups will enable a more rigorous evaluation of vehicle-dependent effects and further strengthen the interpretation of antigen-specific immunogenicity.

In this study, neutralizing activity was evaluated against the prototype SARS-CoV-2 strain to assess the immunological impact of trimeric RBD architecture within a cytoplasmic transcription-based DNA vaccine platform. Although variant-specific neutralization was not examined, this was beyond the scope of the present mechanistic analysis. Importantly, the trimeric RBD design allows for modular substitution of variant sequences, and the T7 autogene system enables rapid antigen cassette exchange without modifying the core platform. Thus, while breadth and durability were not directly assessed here, the platform remains readily adaptable for future studies evaluating cross-variant neutralization and immune persistence.

Taken together, this study demonstrates that multivalent RBD architecture, when deployed within a cytoplasmic transcription-based DNA vaccine platform, supports coordinated induction of humoral and cellular immunity under conditions of controlled innate immune engagement. Rather than enhancing a single immune arm, endogenous expression of a structurally optimized trimeric RBD promoted integrated immune responses characterized by CD8⁺ T cell activation, regulated Th1-skewed polarization, and organized germinal center-associated B cell activity. These findings indicate that antigen architecture and expression modality function cooperatively to shape immune quality, highlighting the importance of considering immunogen design within its platform-specific cellular context.

5. Conclusions

In this study, we describe a DNA vaccine design framework in which rational trimeric antigen engineering is coupled with direct cytoplasmic transcription to support balanced and well-tolerated immunogenicity. By addressing key limitations associated with conventional plasmid DNA vaccines, the T7 autogene-driven platform may provide a flexible and mechanistically grounded strategy for antiviral vaccine development. In addition, this approach may offer potential advantages in terms of platform stability and dosing flexibility, which warrants further comparative investigation with current mRNA vaccine approaches. Collectively, our findings establish a foundation for advanced DNA vaccine approaches that integrate antigen structure with intracellular expression to promote effective and regulated adaptive immunity.