Advances in SARS-CoV-2 receptor-binding domain-based COVID-19 vaccines

ABSTRACT

Introduction

The Coronavirus Disease 2019 (COVID-19) pandemic has caused devastating human and economic costs. Vaccination is an important step in controlling the pandemic. Severe acute respiratory coronavirus-2 (SARS-CoV-2), the causative agent of COVID-19, infects cells by binding a cellular receptor through the receptor-binding domain (RBD) within the S1 subunit of the spike (S) protein. Viral entry and membrane fusion are mediated by the S2 subunit.

Areas covered

SARS-CoV-2 S protein, particularly RBD, serves as an important target for vaccines. Here we review the structure and function of SARS-CoV-2 S protein and its RBD, summarize current COVID-19 vaccines targeting the RBD, and outline potential strategies for improving RBD-based vaccines. Overall, this review provides important information that will facilitate rational design and development of safer and more effective COVID-19 vaccines.

Expert opinion

The S protein of SARS-CoV-2 harbors numerous mutations, mostly in the RBD, resulting in multiple variant strains. Although many COVID-19 vaccines targeting the RBD of original virus strain (and previous variants) can prevent infection of these strains, their ability against recent dominant variants, particularly Omicron and its offspring, is significantly reduced. Collective efforts are needed to develop effective broad-spectrum vaccines to control current and future variants that have pandemic potential.

KEYWORDS:

• Coronavirus
• SARS-CoV-2
• COVID-19
• spike
• receptor-binding domain
• vaccines

1. Introduction

The Coronavirus Disease 2019 (COVID-19), first reported in December 2019 [1], has resulted in more than 763 million confirmed cases globally, with at least 6.9 million deaths, as of 19 April 2023 [2]. This COVID-19 pandemic has had unprecedented and disruptive impacts, both socially and economically. COVID-19 is caused by a novel human coronavirus (CoV) called severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) [1].

Currently, there are seven human CoVs. Among them, HCoV-229E, HCoV-OC43, HCoV-NL63, and HCoV-HKU1 normally lead to seasonal and mild respiratory tract infection, accompanied by symptoms of the common cold [3]. By contrast, the other three human CoVs (SARS-CoV-2, Middle East respiratory syndrome coronavirus (MERS-CoV), and SARS-CoV), all of which emerged over the past 20 years, are highly pathogenic and often cause severe or life-threatening respiratory tract infection and lung injury [4–6]. Like SARS-CoV and MERS-CoV, SARS-CoV-2 is an enveloped, single-stranded RNA virus belonging to the genus Betacoronavirus (family, Coronaviridae; order, Nidovirales) [3,6].

The genome of SARS-CoV-2 encodes at least six non-structural proteins and four major structural proteins. The non-structural proteins 3a, 6, 7a, 7b, 8 and 10 are responsible for virus processing and replication [3]. The structural proteins, spike (S), envelope (E), membrane (M), and nucleocapsid (N) (Figure 1a), are associated with viral pathogenesis, and participate in the assembly and release of new viral particles [7–10]. For example, the M protein interacts with other structural proteins to assemble the viral envelope. The E protein contributes mainly to viral assembly and budding. The N protein interacts with the viral RNA strand to form a helical ribonucleocapsid complex. Notably, the E and M proteins modulate the S protein, thereby promoting viral particle assembly [9].

Figure 1. SARS-CoV-2 virion and spike protein RBD. (a) Schematic structure of SARS-CoV-2 virion. (b) Structure of SARS-CoV-2 spike (S) protein. (left) Cryo-EM structure of SARS-CoV-2 S protein trimer (PDB 6VXX). The three subunits are colored in salmon, green and blue, respectively. (right) Close-up view of SARS-CoV-2 S receptor-binding domain (RBD). (c) Schematic structure of SARS-CoV-2 S protein. It contains S1 and S2 subunits. SP, signal peptide. NTD, N-terminal domain. RBM, receptor-binding motif. FP, fusion peptide. HR1 and HR2, heptad repeat 1 and 2. TM, transmembrane. CP, cytoplasmic tail.

Different from SARS-CoV and MERS-CoV, SARS-CoV-2 has undergone continuous mutations, leading to the emergence of multiple variants of concern (VOCs), including Alpha, Beta, Gamma, Delta, and Omicron; Omicron has mutated into subvariants, such as BA1, BA2, BA3, BA4, BA5, with the dominant subvariants covering XBB.1.5, XBB.1.16., and BQ.1.1 [11–13]. These VOCs, including BA5 and the more recent Omicron subvariants, transmit rapidly among humans and are resistant to current vaccines and therapeutic antibodies, which target the original strain or previous VOCs [14–19]. Thus, there is an urgent need to develop new vaccines and therapeutic agents to control the current COVID-19 pandemic and prevent potential future pandemics. Vaccines with sufficient efficacy should prevent future infection by SARS-CoV-2 and its variants.

The remainder of this review will discuss the spike (S) protein receptor-binding domain (RBD) of SARS-CoV-2, which is a critical target for development of COVID-19 vaccines, as well as vaccine-induced immune responses and RBD-based COVID-19 vaccines under preclinical and clinical development. We also discuss the limitations of these vaccines, and propose potential approaches to improve their efficacy against infection by emerging variants.

2. SARS-CoV-2 S protein and RBD

The S protein of SARS-CoV-2 presents as a homotrimer, each monomer containing S1 and S2 subunits (Figure 1b) [20,21]. The S1 subunit comprises an N-terminal domain (NTD) and an RBD (Figure 1b, c). The S2 subunit comprises four structural regions: fusion peptide, two heptad repeats (HR1 and HR2), and a transmembrane region (Figure 1c). The S protein is a critical component that mediates viral attachment and entry into host cells via interaction between its RBD fragment and a cellular receptor, angiotensin-converting enzyme 2 (ACE2) [22]. Among the three RBD molecules within an S trimer, only the RBD with an ‘up’ conformation binds to the ACE2 receptor [20,23]. The S protein and RBD of SARS-CoV-2 have higher affinity for human (h)ACE2 than those of SARS-CoV; the ACE2 receptor is widely expressed in various human organs, including lung, heart, brain and kidney, which explains the tropism of SARS-CoV-2 [20,24,25]. In addition, SARS-CoV-2 S protein and its RBD also bind to ACE2 receptors from other species, such as mice, hamsters, horses, civets, and bats, with an overall binding affinity similar to, or different from (weaker or stronger), that for hACE2 [11,26–29].

The S protein of SARS-CoV-2 plays a critical role in viral infection and pathogenesis. SARS-CoV-2 enters into host cells through multiple coordinated and critical steps, including ACE2 receptor binding, S protein cleavage, virus and cell membrane fusion, and release of the viral genome [30,31]. During SARS-CoV-2 infection, the S1 subunit of S protein first binds to the ACE2 receptor via the RBD, which triggers S1 shedding. After cleavage by host proteases, the S2 subunit of S protein changes its conformation from the unstable pre-fusion state to the stable post-fusion state [32,33]. A second cleavage releases the fusion peptide, which is responsible for priming the host cell membrane for fusion [34]. The HR1 and HR2 regions adjacent to the fusion peptide are exposed and undergo conformational changes, thereby forming an irreversible six-α-helical (6-HB) bundle structure to promote membrane fusion [35]. Overall, binding between the S protein and the ACE2 receptor via the RBD triggers a cascade of events that results in fusion of the viral membrane and host cells. Therefore, the S protein, particularly the RBD, is a critical target for development of vaccines and therapeutic agents against SARS-CoV-2 [7,11,30,36,37]. A number of crystal and Cryo-EM structures of the RBD in complex with the ACE2 receptor have been resolved, which include the RBD of SARS-CoV-2 isolated from wildtype or variant strains (e.g. Beta, Delta, or Omicron), as well as the ACE2 from species such as bats, horses, mice, and civets (Figure 2) [26,38,39]. These structures provide important information for the rational design of effective COVID-19 vaccines.

Figure 2. Structures of SARS-CoV-2 RBD-receptor complexes. (a) Structures of SARS-CoV-2 receptor-binding domain (RBD) from wildtype (WT) strain (PDB 6M0J), Beta variant (PDB 7VX4), Delta variant (PDB 7TEW), Kappa variant (PDB 7TEZ), or Omicron variant (PDB 7×O9) in complex with human angiotensin-converting enzyme 2 (ACE2). (b) Structures of SARS-CoV-2 RBD in complex with ACE2 from non-human species. WT RBD-bat ACE2 complex (PDB 7C8J), WT RBD-horse ACE2 complex (PDB 7FC5), Omicron RBD-mouse ACE2 complex (PDB 7×O6), and Omicron RBD-civet ACE2 complex (PDB 7WSK) are shown.

3. RBD-based COVID-19 vaccines induce specific immune responses

The S protein and its RBD of SARS-CoV-2 contain many B cell and T cell epitopes, which induce specific antibody and cellular immune responses, respectively [40,41]. However, S protein peptides outside of the RBD might also be relevant with respect to induction of antibody-dependent enhancement, a phenomenon that can occur when binding of a virus to suboptimal antibodies enhances its entry into host cells [42]. The RBD is a promising vaccine target for induction of potent neutralizing antibodies against SARS-CoV-2; indeed, antibodies present in SARS-CoV-2-infected serum show strong neutralizing activity [43,44].

RBD-based vaccines confer protection against SARS-CoV-2 infection by inducing adaptive immune responses, including stimulating B cells to produce neutralizing antibodies. T cell-mediated cellular immune responses either enhance B cell responses or kill virus-infected target cells directly. RBD-based vaccines function by enabling antigen presenting cells (APCs) such as dendritic cells and macrophages to recognize the RBD protein. APCs, which are activated through pattern recognition receptors and migrate to the lymph nodes, present the peptides derived from the RBD fragment and activate CD4⁺ and CD8⁺ T cells via major histocompatibility complex-I (MHC-I) and MHC-II molecules. Activated CD4⁺ T cells drive B cell development in the lymph nodes, leading to maturation and secretion of antibodies specific for the RBD protein. The activated CD8⁺ effector T cells then kill and eliminate virus-infected cells. Immune memory cells, including memory B cells and CD8⁺ memory T cells, are also induced, which can trigger protection by activating and proliferating rapidly once they re-encounter pathogens. SARS-CoV-2 infection and RBD vaccine-induced immune responses are described in Figure 3 [8,45–49].

Figure 3. Schematic map of SARS-CoV-2 infection and RBD vaccine-induced immune responses. (a) SARS-CoV-2 infection. SARS-CoV-2 infects host cells by binding to cellular angiotensin converting enzyme 2 (ACE2) receptor via the receptor-binding domain (RBD) fragment in the spike (S) protein. (b) RBD vaccine-induced immune responses. RBD vaccines activate antigen presenting cells such as dendritic cells, and elicit specific CD4⁺ and CD8⁺ T cells, which either help B cells to produce antibodies, or directly kill virus-infected cells. The elicited neutralizing antibodies block the binding of RBD to the ACE2 receptor, thereby inhibiting subsequent viral entry process.

4. SARS-CoV-2 RBD-based COVID-19 vaccines in preclinical development

Most RBD-specific COVID-19 vaccines are in preclinical development; the majority of these induce effective immune responses and/or protection in vaccinated animals, including mice, rabbits, pigs, hamsters, and non-human primates (NHPs) [50–55] (Table 1). These vaccines can be categorized according to whether they are based on subunit proteins, nanoparticles, mRNAs, or other moieties (Table 1).

Table 1. Representative SARS-CoV-2 RBD-based COVID-19 vaccines in preclinical development^a.

Vaccine	Adjuvant	Route	Animal model	Antibody response	Cellular immune response	Protection	Reference
SARS-CoV-2 RBD-based subunit vaccines
Glycosylated MU-RBD	Alum + MPL	I.M.	B6 hACE2-Tg mice	Elicited similar IgG Abs as wildtype-RBD specific to the RBDs of wildtype, Delta, Omicron-BA.1 and BA.2 variants, but higher neutralizing Abs against pseudotyped (wildtype, Alpha, Beta, Delta, Gamma, Epsilon and Omicron variants) and live SARS-CoV-2 (wildtype, Delta, and Omicron variants)	N/A	Completely protected hACE2-Tg mice against lethal challenge of SARS2-CoV-2 (wildtype and Delta variant) without significant weight change; protected B6 mice against SARS-CoV-2 (mouse-adapted N501YMA30 variant) with significantly lower viral titers in lung	[51]
SARS-CoV-2 RBD-His	AddaVax Imject	I.P.	B6 mice	Elicited RBD-specific (1:3.6 × 10^5) and S-specific (>1:10^6) IgG and neutralizing Abs against live SARS-CoV-2 (wildtype U.S.A-WAI/2020: PRNT90 1:640–1,280)	RBD-His adjuvanted with Imject induced RBD-specific CD4^+ T cells and CD8^+ CTLs with secretion of IFN-γ	N/A	[56]
Tag-free and His-tagged SARS-CoV-2 RBD	Alum	S.C.	B6 mice	Tag-free RBD elicited higher RBD-specific IgG Abs (1:1.1 × 10^4) than His-tagged RBD (1:1,995), and higher neutralizing Abs against pseudotyped (NT50 1:9,537) and live SARS-CoV-2 (Delta variant: NT50 1:1,845) than His-tagged RBD (NT50 1:895 for pseudovirus; NT50 1:355 for live Delta variant)	N/A	N/A	[57]
RBD203-N1 RBD219-N1C1	Alum + CpG1826	I. M.	BALB/c mice	RBD203-N1 and RBD219-N1C1 with Alum plus CpG1826 adjuvants elicited 1000-fold increase of RBD-specific IgG Abs than with Alum alone, and 2.5-fold higher pseudovirus neutralizing Abs than human convalescent plasma	RBD203-N1 or RBD219-N1C1 with Alum plus CpG1826 adjuvants induced stronger and more balanced Th1/Th2 response, and increased secretion of IL-2, IL-4, IL-6 and IFN- γ than with alum alone	N/A	[58]
RBD5m	Alum + CpG	I. M.	BALB/c mice	Elicited similar RBD-specific Abs as wildtype RBD and broader neutralizing Abs against SARS-CoV-2 (wildtype, Beta, Delta, Omicron-BA.1 and BA.2.12.1 variants: NT50 1:250, 1:226, 1:1,195, 1:2,543, and 1:1,102, respectively)	Induced RBD-specific T cells with secretion of IFN-γ, IL-2, IL-4 and IL-5	N/A	[59]
3Ro-NC	KFD Alum	I. N. I. M.	BALB/c hACE2-Tg mice	Elicited RBD-specific IgG Abs via I.M. route against SARS-CoV-2 (wildtype, Delta, Gamma and Omicron-BA.1 variants) and neutralizing Abs against pseudotyped SARS-CoV-2 (wildtype, Alpha, Beta, Gamma, Delta, Omicron-BA.1, BA.2, BA.4 and BA.5 variants); induced RBD-specific IgA Abs via I.N. route against Delta and Omicron-BA.1 variants (~1:10^3)	N/A	Protected hACE2-Tg mice against SARS-CoV-2 (Omicron-BA.1 variant) after I.N. immunization, with lower inflammation and virus load in airway and lung than I.M. immunization	[60]
RBD-CD	Alum	I.M.	BALB/c mice NHPs	Elicited RBD-specific IgG (>1:10^4 in mice; 1:4.2 × 10^4 in NHPs) and neutralizing Abs against live SARS-CoV-2 (CUT2010–2025/Cuba/2020: NT50 1:50–100 in NHPs)	Induced RBD-specific T cells with secretion of IFN-γ and IL-2	N/A	[52]
Baiya SARS-CoV-2 Vax 1	Alum	I.M.	hACE2-Tg mice NHPs	Elicited RBD-specific IgG (>1:10^4 in NHPs) and neutralizing Abs against live SARS-CoV-2 (wildtype: NT50 1:419 in mice and 1:3,880 in NHPs)	N/A	Protected hACE2-Tg mice against SARS-CoV-2 (wildtype: 2 × 10^4 PFU) with reduced viral RNA levels in brain and lung, and no vaccine-related pathological changes in NHPs	[53]
SARS-CoV-2 RBD-based nanoparticle vaccines
RBD-SpyVLP	AddaVax	I.M.	B6 BALB/c mice Pigs	Elicited RBD and S-specific IgG (>1:10^3 in mice) and neutralizing Abs against pseudotyped (NT50 1:500–4,000 in pigs) and live SARS-CoV-2 (hCoV-19/England/02/2020: NT50 1:450–2,095 in B6 mice; 1:230–1,405 in BALB/c mice; 1:6,000–1.1 × 10^4 in pigs)	N/A	N/A	[61]
VLP-RBD	Alum + MPL	I. M.	BALB/c mice	Elicited higher and long-lasting RBD and S-specific IgG (>1:10^5) and neutralizing Abs (NT50>1:1,400) than RBD alone against pseudotyped (wildtype, Alpha, Beta and Gamma variants) and live SARS-CoV-2 (wildtype US-WA-1: NT50>1:10^3), pseudotyped SARS-CoV and SARS-related bat CoVs	N/A	Protected mice against mouse-adapted SARS-CoV-2 (10^5 PFU), with no significant weight loss or detectable virus in lung	[55]
Ferritin-NP-RBD	CpG1826	S.C.	B6 mice	Elicited RBD-specific IgG (1:10^6) and neutralizing Abs against live SARS-CoV-2 (C-Tan-nCoV strain 04), which were maintained for over 7 months	N/A	N/A	[62]
RBD-ferritin nanoparticle	ALFQ	I.M.	B6 BALB/c hACE2-Tg mice NHPs	Elicited RBD and S-specific IgG (>1:10^4-10^5) and neutralizing Abs against pseudotyped (wildype, Alpha and Beta variants: NT50>1:10^4) and live SARS-CoV-2 (wildtype WA1/2020, Alpha and Beta variants: NT50>1:10^3)	Induced CD4^+ T cells with secretion of TNF-α, IL-2 and IFN-γ	Protected hACE2-Tg mice against SARS-CoV-2 (wildtype WA1/2020 strain) via passive transfer of immune sera; protected NHPs against SARS-CoV-2 (WA1/2020: 10^6 TCID50) with reduced viral replication and rapid clearance in upper and lower airways	[63,64]
RBD-I53-50	AddaVax AS03	I.M.	BALB/c Darwin mice NHPs	Elicited S-specific serum IgG (1:4.9 × 10^5 for wildtype; 1:5.2 × 10^4 for Omicron variant), mucosal IgG in BAL fluid and nasal swab, and neutralizing Abs in NHPs against pseudotyped (NT50 1:1,978) and live SARS-CoV-2 (wildtype: NT50 1:1,331) and other CoVs (i.e. SARS-CoV, bat-CoVs, etc.)	Induced Beta and Omicron S-specific CD4^+ T cells in PBMCs with secretion of IL-2, IL-4, IFN-γ with or without TNF-α	Protected NHPs against SARS-CoV-2 (Omicron-BA.1 variant: 2 × 10^6 PFU) with undetectable viral load in lung or BAL fluid	[65,66]
RBD-CoPoP	MPLA MPLA + QS‐21	I.M.	CD‐1 mice Rabbits	Elicited RBD‐specific IgG (>1:10^4) and neutralizing Abs against pseudotyped (NT50>1:10^4) and live virus SARS-CoV-2 (wildtype U.S.A‐WA1/2020), preventing pathogenic cellular infection at 1:1,280 diluted sera	Induced CD4^+ and CD8^+ T cells with secretion of IFN-γ, IL‐2 and TNF-α	N/A	[67]
RBD-TMC NPs	N/A	I.N.	BALB/c mice	Elicited RBD-specific IgG (>1:10^4), IgA (~1:10^2 in lung and BAL fluid; ~1:10^3 in sera) and neutralizing Abs against live SARS-CoV-2 (NT50 1:366).	Induced RBD-specific CD4^+ T cells with secretion of IFN-γ and CD8^+ T cells with secretion of IFN-γ	N/A	[68]
SARS-CoV-2 RBD-based mRNA vaccines
RBD-mRNA −LNP	N/A	I.D.	BALB/c mice	Elicited RBD-specific IgG (>1:10^5) and neutralizing Abs against multiple pseudotyped (NT50>1:10^4 for wildtype; slightly reduced titers against Alpha, Beta, Gamma, Delta, Omicron and other variants) and live SARS-CoV-2 (NT50>1:10^4 for wildtype; >1:10^3 for Delta; >1:10^2 for Omicron), which maintained for at least 8 months	Induced RBD-specific CD4^+ and CD8^+ T cells with secretion of IFN-γ, TNF-α and IL-4	Completely protected mice against SARS-CoV-2 (mouse-adapted N501YMA30 variant: 5 × 10^3 PFU) with increased body weight	[45,69]
mRNA-1283	N/A	I.M.	BALB/c hACE2-Tg mice	Elicited similar RBD and S-specific IgG (>1:10^4) and higher neutralizing Abs (>1:10^4) than clinically approved S-based mRNA-1273 against pseudotyped (D614G and Beta variant: NT50~41.2-fold increase in BALB/c mice) and live SARS-CoV-2 (Omicron-BA.1 variant in hACE2-Tg mice)	Induced RBD or S1-specific CD4^+ or CD8^+ T cells with secretion of IFN-γ and TNF-α	Protected hACE2-Tg mice against SARS-CoV-2 (WA1/2020 D614G or Omicron-BA.1 variant: 10^4 PFU), with significantly lower viral load in lung, nasal turbinate, and nasal washes than mice with mRNA-1273	[70]
RBD-hFc-mRNA	N/A	I.M. I.D.	BALB/c mice	Elicited S-specific IgG (>1:10^4) and neutralizing Abs against pseudotyped SARS-CoV-2 (NT50>1:10^3)	Induced S-specific T cells with secretion of IFN-γ and IL-2	N/A	[71]
RBD-trimer-mRNA	N/A	I.M.	BALB/c hACE2-Tg mice	Elicited S-specific IgG (>1:10^5) and neutralizing Abs against pseudotyped SARS-CoV-2 (wildtype: NT50>1:10^4; Alpha, Beta, Gamma, Delta and other variants: NT50 1:6.1 × 10^4) in BALB/c mice	Induced S-specific CD4^+ T cells with secretion of IFN-γ, TNF-α and IL-2	Protected hACE2-Tg mice against SARS-CoV-2 (wildtype D614, Beta and Delta variants: 10^3 PFU) with significantly lower viral load in lung	[72]
T-RBD mRNA TF-RBD mRNA TF-RBD-triCoV mRNA	N/A	I.M.	B6 mice	TF-RBD mRNA elicited higher RBD-specific IgG (>1:10^4) and neutralizing Abs than T-RBD mRNA against pseudotyped (NT50>1:10^4) and live SARS-CoV-2 (NT50>1:10^3); a TF-RBD-triCoV mRNA improved neutralizing Abs against pseudotyped SARS-CoV-2 (wildtype, Alpha and Beta variants)	TF-RBD mRNA induced Th1-biased response, with increased secretion of IFN-γ and TNF-α in CD4^+ and CD8^+ T cells	TF-RBD mRNA protected adenovirus-hACE2-transduced mice against SARS-CoV-2 (hCoV-19/China/CAS-B001/2020: 10^5 PFU) with no detectable viral titer in lungs	[73]
Other SARS-CoV-2 RBD-based vaccine types (including viral vectors)
RBD/LVP-K1-RBD19	N/A	I.M.	BALB/c mice	Elicited higher RBD-specific IgG and neutralizing Abs against pseudotyped SARS-CoV-2 than RBD protein with Alum adjuvant	N/A	N/A	[74]
AAV-3×RBD	N/A	I.M.	BALB/c mice Dogs	Elicited S-specific IgG (>1:10^5) and neutralizing Abs in mice against pseudotyped SARS-CoV-2 (NT50~1:2,750)	N/A	N/A	[75]
TOH-VAC1	N/A	I.P. I.N. I.M.	BALB/c mice NHPs	Elicited RBD-specific IgG (>1:10^4) and neutralizing Abs in NHPs against pseudotyped SARS-CoV-2 (NT50>1:128)	Induced RBD-specific T cells in PBMCs	N/A	[76]
VV−tRBD	N/A	I.N.	BALB/c mice Rabbits	Elicited RBD-specific IgG (~1:10^4), IgA/sIgA (~1:1,024 in sera, 1:67 in nasal washes and 1:29 in BALs) and neutralizing Abs in mice against pseudotyped SARS-CoV-2 (NT50 1:356 for WA1; 1:112 for Omicron in sera; 1:146 for nasal washes; 1:58 for BALs)	Induced RBD-specific T cells with secretion of IFN-γ, IL4 and IL-17a	N/A	[77]

Note: ^aAbs, antibodies; BAL, bronchoalveolar lavage; CoVs, coronaviruses; C57BL/6, B6; CTLs, cytotoxic T cells; hACE2-Tg, human angiotensin converting enzyme 2; I.M., intramuscular; I.N., intranasal; I.P., Intraperitoneal; N/A, not available; NHPs, non-human primates; NT₅₀, 50% neutralizing antibody titer; PBMCs, peripheral blood mononuclear cells; PFU, plaque forming unit; PRNT₉₀, 90% plaque reduction neutralization test Ab titer; S.C., subcutaneous.

4.1. SARS-CoV-2 RBD-based COVID-19 subunit vaccines

Subunit vaccines refer to recombinant protein-based vaccines. They normally eliminate viral infectious components and thus are relatively safe, but their immunogenicity might be variant, depending on the conjugated adjuvants and other parameters described below. Several SARS-CoV-2 RBD-based subunit vaccines have been developed and tested in different animal models (Table 1).

4.1.1. Overview of current RBD-based subunit vaccines

The RBD of SARS-CoV-2 is an effective target for development of COVID-19 subunit vaccines. Indeed, the RBD protein-based subunit vaccines developed to date bind to ACE2, the receptor of SARS-CoV-2, generate SARS-CoV-2 or SARS-CoV RBD-specific neutralizing antibodies, and induce effective immune responses in immunized animals [25,51,55,78]. RBD proteins covering residues 330–526 or residues 331–524 of the SARS-CoV-2 S protein are stable, and bind effectively to human or bat ACE2 receptors to induce humoral and T cell responses, which neutralize against SARS-CoV-2 infection [25,56]. A tag-free RBD protein also binds strongly to the hACE2 receptor and induces more neutralizing antibodies than a His-tagged RBD protein [57]. Other RBD proteins such as RBD203-N1 and RBD5m are immunogenic, and induce effective neutralizing antibodies against original pseudotyped or authentic SARS-CoV-2, and against VOCs [58,59].

SARS-CoV-2 RBD-based subunit vaccines can be designed by linking two or more RBDs from the same or different virus strains, or RBD can be fused to other immunogens. Similar to the RBD- single chain dimers (sc-dimers) of MERS-CoV and SARS-CoV, a tandem repeat RBD-sc-dimer of SARS-CoV-2 elicits increased neutralizing antibodies than an RBD-monomer [79]. Indeed, a triple RBD vaccine (3Ro-NC) comprising one Delta variant RBD and two RBDs from the Omicron-BA.1 subvariant elicits mucosal immune responses and neutralizing antibodies against Omicron-BA.1 subvariant, thereby protecting mice from Omicron infection and virus-induced immunopathology [60]. Fusion of two SARS-CoV-2 RBDs with the PreS antigen of hepatitis B virus (HBV) induces RBD-specific IgG1 and IgG4 antibodies, which block the RBD-ACE2 interaction and neutralize SARS-CoV-2 infection [80]. Fusion of the RBD with an immune system modulator molecule, human CD154 extracellular domain, generates an RBD-CD vaccine that induces specific IgG antibodies in mice and NHPs, with NHP sera showing neutralizing activity against SARS-CoV-2 infection [52]. Fusing the RBD with human Fc (Baiya Vax 1) also induces antibodies with neutralizing activity in mice and NHPs, thereby providing protection from SARS-CoV-2 infection [53].

4.1.2. Characterization of RBD-based subunit vaccines and potential factors affecting their efficiency

4.1.2.1. Expression platforms for RBD-based subunit vaccines

Compared with the full-length S protein, the RBD fragment of SARS-CoV-2 is much easier to produce at high yields, suggesting that the cost of producing RBD-based subunit vaccines on a large scale would be low. Moreover, the production of the RBD protein can be scaled up easily due to its small size.

RBD-based subunit vaccines have been produced in different cell types, including yeasts [54,58,81,82], plant cells [83,84], E. coli [85,86], insect cells [87], and mammalian cells [88,89]. Yeast expression systems are commonly used to produce SARS-CoV-2 RBD-based subunit vaccines because they represent a low-cost production platform with high scale-up capacity, making them more attractive than other production systems, such as mammalian cell culture systems [90]. Moreover, yeast production platforms can be modified easily to increase productivity. For example, an engineered yeast strain (Komagataella phaffii) was used to produce a RBD protein without requirement for a methanol induction process, thereby enabling scale up of production by more than 5-fold [91]. Insect cells are another system with high-level expression of SARS-CoV-2 RBD proteins, and a Sf9 insect cell-expressed RBD protein driven under a chimeric promoter (polh-pSeL) resulted in the yield of 21.1 mg/L [92]. An insect cell-derived RBD protein from the Bac-to-Bac baculovirus system had high potency to induce neutralizing antibodies and protection against SARS-CoV-2 infection in animal models [87]. Several SARS-CoV-2 RBD proteins have been expressed in mammalian cells, including CHO, 293F and 293T cells [51,55]. Different from proteins expressed in E. coli and insect cells, mammalian cell-expressed RBD proteins maintain correct conformational structures, especially with respect to post-translational modifications such as glycosylation [51]. Efforts have been made to increase the productivity of RBD-based subunit vaccines in mammalian cells. For example, a RBD protein of SARS-CoV-2 transiently transfected in ExpiCHO cells at 32°C led to the production yield of 40 mg/L [93]. A CHO cell-expressed SARS-CoV-2 RBD-sc-dimer protein linked through a disulfide bond maintained at high yields (g/L level) with capability for clinical development [79]. Fusion of a human Fc tag to the RBD protein of SARS-CoV-2 improves productivity significantly when compared with of an RBD without the Fc tag. Adenovirus-infected 293SF cell expression systems have been used to increase the production of RBD proteins in culture supernatants by 7-fold compared with non-viral transfection systems [94]. Of note, the RBD proteins expressed in both yeast and mammalian cell culture systems are well folded and stable, although they may exhibit different glycosylation patterns [95]. In particular, the difference in glycosylation between yeast- and mammalian cell-produced RBD proteins does not affect their immunogenicity significantly [81].

4.1.2.2. Adjuvant choices for RBD-based subunit vaccines

Different from live-attenuated virus-based or mRNA vaccines, protein-based subunit vaccines often require adjuvants to stimulate an optimal host immune response [88]. Therefore, it is essential to select adjuvants that enable SARS-CoV-2 RBD-based subunit vaccines to induce optimal and maximal immune responses against SARS-CoV-2 infection.

Aluminum is a classical adjuvant that helps subunit vaccines to induce effective immune responses that normally show a Th2-bias [96,97]. Growing evidence suggests that addition of immunopotentiators increases the immunogenicity of SARS-CoV-2 RBD-based subunit vaccines [51,98]. Many immunopotentiators, when used as adjuvants, target toll-like receptors (TLRs); examples include the TLR9 agonist CpG [54,58], the TLR4 agonist monophosphoryl-lipid A (MPLA) [51,84], and the TLR-7/8 agonist 3 M–052 [98]. Conjugation of these adjuvants to aluminum-adjuvanted RBD proteins may trigger stronger neutralizing antibodies, and more Th1-biased or balanced Th1/Th2 responses, thereby promoting the RBD to elicit potent protection against SARS-CoV-2 infection [51,58]. For example, combination of 3 M–052 with an aluminum adjuvanted-RBD subunit vaccine leads to stronger RBD binding and higher neutralizing antibody titers, as well as improved Th1-biased CD4⁺ and CD8⁺ T cell responses, than the aluminum-adjuvanted RBD alone [98]. Other adjuvants with novel mechanisms of action have been studied. Examples include U-Omp19, a bacterial protease inhibitor from Brucella abortus [88]. Different from the aluminum adjuvant alone, conjugation of U-Omp19 adjuvant to aluminum significantly increases the ability of the RBD protein to induce mucosal and systemic neutralizing antibodies against the original and variant SARS-CoV-2 strains; it also improves generation of RBD-specific germinal center B cells and plasmablasts, and promotes Th1-biased responses in the lung and spleen [88].

4.1.2.3. Delivery routes and other approaches for RBD-based subunit vaccines

The injection routes may affect the immunogenicity and/or efficacy of RBD-based COVID-19 subunit vaccines. Traditionally, RBD-based protein vaccines are injected via the intramuscular route, which requires professional administration [51,98]. Other routes, such as the intradermal and subcutaneous routes, are also used [51,99]. For example, when BALB/c mice were injected intradermally with an RBD-based peptide vaccine, it was well tolerated and triggered a similar humoral antibody response (and better cellular immune response) than the intramuscular route [100]. Thus, this route may be a promising alternative to the intramuscular route.

Immunization doses, antigen dose, and other factors may also affect the immunogenicity and efficacy of RBD-based COVID-19 subunit vaccines. Generally, most RBD-based subunit vaccines require two to three doses to induce production of high and long‐lasting neutralizing antibodies. However, multiple vaccinations may be inconvenient for the individuals who are allergic to the vaccines or are afraid of needles. Additional approaches have been used to reduce the injection dose or increase the antigen dose of RBD proteins, resulting in similar or improved immune responses [99]. For instance, multimerizing a RBD protein on an injectable hydrogel scaffold boosts the immunogenicity of the RBD in mice [99]. Prolonged delivery of a single high dose of this RBD-hydrogel vaccine elicited similar serum IgG antibody titers, but more long‐lasting Th1-biased, IgG2b antibody titers than three individual bolus injections of low-dose RBD protein [99]. These vaccination strategies may serve as alternative approaches to either maintain or enhance the ability of the RBD to induce effective immune responses and long-lasting protection.

4.1.3. Approaches to improving RBD-based COVID-19 subunit vaccines

The RBD of SARS-CoV-2 is a safe and effective antigen for inducing protective immunity [51,101]. However, the RBD protein itself has relatively poor immunogenicity, and it triggers limited activity of germinal center B (GC B) and T follicular helper (Tfh) cells [102]. Hence, it is essential to increase the immunogenicity of RBD-based subunit vaccines, thereby expanding related applications used in vaccine development and manufacture. Different approaches have been applied to improve the immunogenicity and/or efficacy of RBD-based COVID-19 subunit vaccines. Some of these are described below.

4.1.3.1. Modification or engineering of RBD proteins

Glycostructure is associated with the immunogenicity of protein vaccines. For example, an N-linked glycosylation site can be attached to a non-neutralizing epitope of the SARS-CoV-2 RBD by introducing mutations at residues 519 and 521 of the S protein [51]. The resulting mutant RBD protein showed enhanced neutralizing activity against multiple SARS-CoV-2 variants and protected immunized mice from lethal challenge with a prototype SARS-CoV-2 strain and a Delta variant [51]. Four other N-glycosylation motifs can be generated on the RBD protein, which substantially improve expression and immunogenicity of the RBD subunit vaccine [103]. RBD proteins containing multiple mutations (as seen in the Alpha, Beta, or Gamma variants) have shown stronger immunogenicity than the RBD proteins containing a single mutation. Particularly, RBD proteins containing K417T-E484K-N501Y or K417N-L452R-E484K-N501Y mutations produced higher IgG antibody titers in mice, with more potent neutralizing antibodies against the wildtype and mutant pseudotyped SARS-CoV-2 [104]. Also, an engineered RBD protein containing L452K and F490W mutations, which maintains a structure similar to that of the original RBD, enhanced the ACE2 binding and immunogenicity, resulting in induction of higher neutralizing antibody titers against pseudotyped SARS-CoV-2 [54].

4.1.3.2. Dimerization or oligomerization of RBD proteins

Dimerization or oligomerization may enhance the immunogenicity, neutralizing activity, or protective efficacy of the RBD protein. Dimerization or oligomerization can be achieved by using a GGGGS linker to couple two RBDs together to form a dimer, or by using a carrier protein such as α-hemolysin to form a RBD heptamer [81,105]. An engineered dimeric RBD protein induces more neutralizing antibodies than a monomeric RBD protein [81]. Two chimeric RBD dimer subunit vaccines have been generated by fusing the original RBD with the Beta variant RBD, or by fusing the Delta variant RBD with the Omicron variant RBD [106]. Studies in mice or NHPs showed that these fusion proteins induced broader and better neutralizing antibodies and protective efficacy against SARS-CoV-2 variant strains, including Delta and Omicron. In rats, a trimeric RBD (mos-tri-RBD) vaccine comprising the RBDs of Omicron and other variants induced higher titer and broader-spectrum neutralizing antibodies than those induced by homo-tri-RBD consisting of three original RBDs against Omicron and other variants [107]. A RBD heptamer (mHla-RBD) also increases the ability of the RBD monomer to induce stronger RBD-specific antibodies and cellular immune responses, with broadly neutralizing activity against different SARS-CoV-2 variants [105].

4.1.3.3. Fc fusion of RBD proteins

Fc-tagged protein drugs and antibodies have been evaluated in clinical trials, and their safety has also been established in humans [108]. Some Fc-fused vaccines may modulate the immune system by binding and activating immune cells through their Fc receptors [109]. However, such phenomena have not been demonstrated in COVID-19 subunit vaccines. Fusion of the Fc domain of human IgG with the RBD protein of SARS-CoV-2 allows dimerization of the RBD. The Fc fragment, when conjugation to the RBD protein, may serve as an immunopotentiator to enhance the immunogenicity and/or protection of RBD subunit vaccines [110]. Heterologous boosting with Fc-fused RBD proteins or Fc-fused RBD proteins containing variants may increase neutralizing activity against variant strains. Indeed, heterologous boosting of NHPs with a Fc-fused RBD protein (RBD-Fc-Omicron) derived from the SARS-CoV-2 Omicron-BA1 subvariant increased the titers of neutralizing antibodies against several SARS-CoV-2 Omicron subvariants, such as BA1, BA2, BA3, and BA4, as well as other VOCs, such as Delta, to a greater extent than homologous boosting using a Fc-fused wildtype RBD protein (RBD-Fc-WT) [111]. In mice, RBD-Fc-Omicron elicited higher titers of Omicron-BA1- and BA2-specific neutralizing antibodies than the RBD-Fc-WT, thereby protecting them against infection with these variants [111].

4.2. SARS-CoV-2 RBD-based nanoparticle vaccines

Nanoparticles are defined as particulate substances with a diameter at 1–100 nm; they resemble the structure of a natural virus but do not contain any infectious components. Appropriately designed nanoparticles are a promising platform for developing vaccines that elicit strong and broad immune responses against variant viral pathogens [112].

4.2.1. Overview of current RBD-based COVID-19 nanoparticle vaccines

SARS-CoV-2 RBD-based nanoparticle vaccines have been designed with diverse formats, including virus-like particles (VLPs), protein nanoparticles and micelles, liposomes, and polymer nanoparticles (Table 1).

4.2.1.1. VLP-based nanoparticle RBD vaccines

This type of COVID-19 vaccines has been designed to display RBD with strong thermal stability, low-dose formulation, and strong induction of effective immune responses and protection against SARS-CoV-2 [54,55,61]. These vaccines are promising low-cost immunogens with the capacity for large-scale manufacturing. SpyCatcher003-mi3, a VLP platform based on an engineered aldolase from thermophilic bacteria, displays the SARS-CoV-2 RBD protein in the form of RBD-SpyVLP [61]. This vaccine is thermostable, and elicits potent neutralizing antibody responses in immunized mice and pigs; crucially, it maintains the immunogenicity of the RBD after lyophilization [61]. In addition, covalent conjugation of RBD-L452K-F490W to a VLP (i3–01) yields SpyCatcher-I3–01, each VLP containing 60 sites for RBD display [54]. This engineered RBD-VLP vaccine demonstrates high binding affinity for the ACE2 receptor and enhances immune responses in mice after a single, low dose injection; it also protects hamsters from SARS-CoV-2 challenge [54]. Another VLP-RBD-based vaccine is designed using lumazine synthase protein as a structural scaffold; this scaffold displays 120 copies of SARS-CoV-2 RBD on its surface. Compared with the soluble monomeric RBD protein, this VLP-RBD vaccine elicits higher-titer and longer-lasting neutralizing antibodies, which neutralize the SARS-CoV-2 original strain and multiple variants, as well as SARS-CoV and SARS-related bat coronaviruses, effectively protecting mice from challenge with SARS-CoV-2 [55].

4.2.1.2. Self-assembling protein nanoparticle RBD vaccines

This type of COVID-19 vaccines has been developed to deliver multiple copies of the SARS-CoV-2 RBD to elicit stronger immune responses [62–65]. These vaccines can be manufactured on a large scale, and may be administered at low doses while inducing strong immune responses. For example, two or three doses of a ferritin-NP-RBD nanoparticle vaccine, generated by genetical fusion of a RBD protein to a gene encoding Helicobacter pylori ferritin using the SpyTag/SpyCatcher technique, induces more potent and durable neutralizing antibody responses and long-term immunological memory in mice than the soluble RBD-SpyTag protein vaccine alone [62]. Another study revealed that two doses of SARS-CoV-2 RBD-ferritin nanoparticles (RFN) elicit strong neutralizing antibody titers, and purified serum antibodies protect mice from a lethal SARS-CoV-2 challenge [63]. Additionally, two-doses of RFN vaccine adjuvanted with Army Liposomal Formulation QS-21 (ALFQ) lead to robust Th1-biased CD4⁺ T cell responses and neutralizing antibodies in rhesus macaques, thereby inhibiting viral replication in the upper and lower airways after challenge with a high-dose SARS-CoV-2 [64]. Moreover, two low-antigen doses of a structurally designed protein nanoparticle, I53–50, which displays 60 copies of the SARS-CoV-2 RBD on its exterior surface, induce more potent neutralizing antibody responses in vaccinated mice than a prefusion-stabilized S-2P trimeric protein [65]. In NHPs, this vaccine elicits durable neutralizing antibodies against the original virus strain and the Omicron variant after the second or third vaccination; it also protects against Omicron infection [66].

4.2.1.3. Liposome-based RBD vaccines

Liposomes offer another approach to delivering the SARS-CoV-2 RBD antigen in vaccine form. This type of vaccines can be delivered via the intranasal route, as well as the intramuscular and subcutaneous routes [67,113]. For example, a His-tagged RBD protein was attached to the surface of liposomes containing cobalt-porphyrin phospholipid to form an RBD-liposomal vaccine (RBD-CoPoP) [67]. The vaccine elicits effective neutralizing antibody titers in vaccinated mice when given via the intramuscular route, resulting in increased antigen uptake and immune cell recruitment to draining lymph nodes [67].

4.2.1.4. Polymer nanoparticle-based RBD vaccines

Polymers (such as chitosan and polyethylenimine) can be used to prepare RBD-based nanoparticle vaccines. Chitosan, a natural polysaccharide, is a useful substance for designing vaccines for mucosal delivery due to its mucoadhesive properties. Intranasal delivery of chitosan-loaded RBD nanoparticles (RBD-TMC NPs) into mice induced specific cellular immune responses and robust antibodies, both locally and systematically, neutralizing SARS-CoV-2 infection [68]. However, potential cytotoxicity and sides effects caused by these polymer carriers and vaccines need to be investigated before further development of related vaccines.

4.2.2. Approaches to improving RBD-based COVID-19 nanoparticle vaccines

SARS-CoV-2 is a mucosal pathogen; indeed, the mucosa of respiratory tract serves as the first barrier to infection by SARS-CoV-2. Thus, induction of protective mucosal immune responses is critical to protect vaccinated individuals from virus infection. The route of delivery of RBD-nanoparticle vaccines may affect their ability to elicit appropriate mucosal immune responses and/or protective efficacy. For instance, an inhalable nano-vaccine with a bionic virus-like structure was designed by assembling the RBD protein within the biomimetic pulmonary surfactant liposomes, which induced stronger mucosal immune responses in mice after intranasal administration than after intramuscular and subcutaneous routes [113].

Nanoparticle vaccines can be designed to form mosaic-type structures or display multiple SARS-CoV-2 RBDs from the same or different variants; the aim is to improve their broadly protective efficacy. A single dose of a multivalent SARS-CoV-2 RBD nanoparticle (RBD-NP) vaccine protected mice against SARS-CoV-2 infection [114]. In addition, a trivalent RBD-NP vaccine that displays the RBDs of the D614G strain, Beta, and Delta variants elicited cross-neutralizing antibodies in mice and NHPs; these antibodies are specific for multiple pseudotyped or authentic SARS-CoV-2 variants, including Alpha, Beta, Gamma, Delta, and Omicron [115]. Such approaches have also been used to design S-based mosaic nanoparticle vaccines that target different SARS-CoV-2 variants, as well as RBD-based universal mosaic nanoparticle vaccines against both SARS-CoV-2 and other Sarbecoviruses [114,116–118].

4.3. SARS-CoV-2 RBD-based mRNA vaccines

SARS-CoV-2 RBD-based mRNA vaccines use lipid nanoparticles (LNPs) to deliver mRNAs encoding SARS-CoV-2 RBD antigens. The mRNA, which encodes SARS-CoV-2 RBD protein encapsulated within LNPs to increase its stability, enters cells through endocytosis, and releases the mRNA into the cytoplasm; the mRNA is then translated into the target protein. Similar to LNP-encapsulated S-based mRNA vaccines, RBD-based mRNA-LNPs can be synthesized and manufactured in vitro in the absence of living cells, allowing for rapid production of vaccines on a large scale [8,36].

4.3.1. Overview of current SARS-CoV-2 RBD-based mRNA vaccines

Different from SARS-CoV-2 S-based COVID-19 mRNA vaccines, which have been approved for use in humans of various ages [119,120], SARS-CoV-2 RBD-based mRNA vaccines are less well developed. Indeed, the majority of these RBD-based mRNA vaccines are still at the preclinical stage, although they do induce effective immune responses in animals, protecting them against SARS-CoV-2 (Table 1).

RBD-based COVID-19 mRNA vaccines are delivered via conventional routes, including intradermal and intramuscular injection. Intradermal injection, or intradermal priming and intramuscular boosting, with a mRNA vaccine containing the original SARS-CoV-2 RBD sequence encapsulated within LNPs (RBD-mRNA-LNP), induced more potent neutralizing antibodies against pseudotyped or live SARS-CoV-2 than an LNP-encapsulated mRNA encoding SARS-CoV-2 S1 protein (S1-mRNA-LNP) [45]. RBD-mRNA-LNP effectively elicited RBD-specific cellular immune responses and production of CD4⁺ and CD8⁺ T cells. This mRNA vaccine also induced durable and broadly neutralizing antibodies against different SARS-CoV-2 variants, including Alpha, Beta, Gamma, Delta, and Omicron, thereby protecting mice against challenge with SARS-CoV-2 [69]. However, the titer of induced neutralizing antibodies against these variant strains was relatively lower than the titer of neutralizing antibodies against the original SARS-CoV-2 strain. Other studies showed that two doses of an LNP-encapsulated mRNA-RBD vaccine delivered to mice via the intramuscular route induced cellular immune responses and neutralizing antibodies against the original strain of SARS-CoV-2, whereas a third dose increased its ability to neutralize the Delta and Omicron variants [121].

Some SARS-CoV-2 RBD-based mRNA vaccines are designed to connect the RBD to other regions of the S protein, such as NTD, or with other components such as human Fc. For example, an NTD-RBD mRNA vaccine (mRNA-1283) has been designed to link the RBD to the NTD, thereby generating higher titers of neutralizing antibodies against SARS-CoV-2 original strain, Beta, Delta, or Omicron-BA1 than the clinically-approved S-based mRNA-1273 [70]. Like RBD-Fc-based subunit vaccines, an RBD-hFc-based mRNA vaccine is constructed by linking the RBD to the Fc fragment of human IgG, which elicits effective immune responses and neutralizing antibodies in mice [71]. However, it is not clear whether this hFc-fused RBD-mRNA increases immunogenicity to a greater extent than the RBD-mRNA without the hFc.

4.3.2. Approaches to improving RBD-based COVID-19 mRNA vaccines

Several approaches have been taken to improve the efficiency of RBD-based mRNA vaccines; the aim here is to induce stronger and broader neutralizing antibodies against multiple SARS-CoV-2 variants. This can be achieved through a heterologous vaccination strategy, or by using trimeric/multivalent vaccine formulations.

4.3.2.1. Heterologous boost vaccination strategy

This approach increases the ability of the RBD-mRNA to induce broadly neutralizing antibodies against several SARS-CoV-2 variants. In mice, for example, one dose of a mRNA encoding the S protein of the Omicron-BA1 subvariant (BA1-S-mRNA), plus two doses of RBD-mRNA encoding the original SASR-CoV-2 stain, generated broad, and similarly high-titer, neutralizing antibodies against Alpha, Beta, Gamma, and Delta variants, as well as several Omicron subvariants (BA1, BA2, BA2.12.1, and BA5) [36]. Another study reported that an RBD-based mRNA vaccine encoding the Omicron RBD (RBD-O) only induced neutralizing antibodies with a narrow neutralization spectrum against the Omicron variant; however, a heterologous booster immunization with the RBD-O mRNA vaccine following two doses of the original RBD mRNA vaccine elicited moderate titers of neutralizing antibodies against the wildtype, Delta, and Omicron variants [122].

4.3.2.2. Trimeric or multivalent mRNA vaccine formulations

These mRNA vaccines have been constructed to achieve broadly neutralizing activity and provide protection against diverse SARS-CoV-2 variants. A trimerized RBD mRNA vaccine (RBD-trimer-mRNA), which is designed to encode a trimeric RBD via genetic fusion of the RBD to a foldon trimerization domain, induced more potent and durable neutralizing antibody responses than a monomeric RBD or a S mRNA vaccine [72]. This trimerized RBD mRNA vaccines also protected mice from challenge with the SARS-CoV-2 original strain and several variants, including the Beta and Delta variants [72]. In addition, an upgraded trimeric RBD mRNA vaccine (TF-RBD) is designed by displaying the trimeric RBD (T-RBD) on ferritin-formed nanoparticles [73]. Compared with T-RBD mRNA, TF-RBD mRNA elicited more potent and long-lasting antibody responses and Th1-biased cellular immune responses, thereby protecting hACE2-transduced mice from SARS-CoV-2 infection [73]. Moreover, a trivalent mRNA vaccine (TF-RBD-triCoV), which is formulated to contain the TF-RBD and two additional RBDs harboring N501Y or K417N-E484K-N501Y mutations, induced higher neutralizing antibody titers against the pseudotyped original strain, as well as the Alpha and Beta variants, than any of the three RBD-mRNAs [73].

4.4. Other SARS-CoV-2 RBD-based COVID-19 vaccine types

In addition to the vaccine types described above, SARS-CoV-2 RBD-targeting vaccines have been designed based on viral vectors. The viral vector-based vaccine platform uses replicating or non-replicating viral vectors to deliver target antigens. Multiple virus strains can serve as vectors to deliver the SARS-COV-2 RBD protein, including Newcastle disease virus, adenovirus-associated virus (AAV), and Vaccinia virus (VACV) [74,75,77]. For example, a viral vectored vaccine, which expresses SARS-CoV-2 RBD on the surface of Newcastle disease virus (LVP-K1-RBD19), generated higher RBD-specific and pseudovirus-neutralizing antibody titers than vaccines based on the RBD protein alone [74]. In addition, an AAV-vectored vaccine expressing one or three RBD proteins (AAV-3×RBD) elicited potent neutralizing antibodies against SARS-CoV-2 in mice and dogs [75]. Furthermore, a VACV (Tiantan strain)-vectored RBD (TOH-VAC1) vaccine induced more potent and long-lasting neutralizing antibodies and T cell responses in mice and NHPs, with no adverse side effects [76]. Also, a VACV-vectored vaccine (VV_−tRBD) expressing trimeric RBD elicited neutralizing IgG and IgA antibodies in mice and rabbits, which neutralized against infection by SARS-CoV-2 pseudovirus [77].

5. SARS-CoV-2 RBD-based COVID-19 vaccines under clinical trials

The majority of the currently approved or trialed COVID-19 vaccines target the SARS-CoV-2 S protein [119,123–126]. Below-mentioned COVID-19 vaccines targeting the SARS-CoV-2 RBD have also been proceeded to human clinical trials, or been approved for clinical use.

5.1. SARS-CoV-2 RBD-based subunit vaccines tested in adults

Several recombinant RBD proteins have been tested in human trials to assess their immunogenicity or efficacy against SARS-CoV-2 in adults [127,128]. Among these, a phase 1 trial showed that three doses of a RBD protein (Noora) elicited neutralizing antibody responses without severe adverse effects [129]. A phase 2 study revealed that two doses of another RBD protein induced high-titer RBD-specific antibodies with neutralizing activity [130]. A phase 2 trial of a yeast-expressed RBD protein (Abdala) adjuvanted with alumina induced effective IgG antibody responses after three doses, with minimal adverse reactions from the injection sites [131]. In phase 1 and 2 trials, an RBD dimer protein vaccine generated by linking two original RBD molecules (ZF2001) was immunogenic; a phase 3 trial demonstrated that it was generally safe after three doses, and was effective at preventing severe to critical COVID-19 [127,132]. Fc-fused RBD proteins have also been evaluated in clinical trials. For example, two doses of an RBD-Fc protein (Betuvax-CoV-2) elicited IgG and anti-SARS-CoV-2 neutralizing antibodies and CD4⁺ T cell responses without obvious adverse effects, in a phase 1/2 trial [133]. A third dose (given 7–9 months after the primary immunization) of a Fc-fused RBD vaccine (UB-612) incorporating helper and cytotoxic T lymphocyte (Th/CTL) epitope peptides elicited high-titer neutralizing antibodies against the Omicron-BA1 and BA2 subvariants, in a phase 1 trial [134]. The safety and immunogenicity of this vaccine (when used as a single dose or as a booster dose) are currently being evaluated in a phase 3 clinical trial (NCT05293665).

5.2. SARS-CoV-2 RBD-based subunit vaccines tested in children

Some RBD-based COVID-19 subunit vaccines are evaluated for their immunogenicity and/or safety in children. In a phase 1/2 trial, two doses of RBD-tetanus toxoid conjugate vaccine (FINLAY-FR-2) and a third dose of a RBD dimer protein (FINLAY-FR-1A) were immunogenic with some adverse events such as local pain in children (3–18 years old), eliciting neutralizing antibodies against several SARS-CoV-2 variants, such as Alpha, Beta, and Delta [135].

5.3. Other types of SARS-CoV-2 RBD-based vaccines for humans

Several RBD-based COVID-19 vaccines developed based on other vaccine types such as nanoparticles and mRNAs are in clinical trials. A phase 1/2 trial indicated that two doses of an RBD-nanoparticle vaccine (GBP510) were immunogenic, inducing RBD-specific antibody responses and neutralizing antibodies [136]. In a phase 2 trial, an RBD-nanoparticle vaccine (EuCorVac-19: ECV19) elicited a dose-dependent IgG antibody response with neutralizing activity, leading to low-level adverse events in the injection site with pain [137]. A SARS-CoV-2-RBD and HBsAg VLP (generated by fusing the RBD to the Hepatitis B surface antigen) is currently in phase 1/2 clinical trials (ACTRN12620000817943) [138]. A Beta variant RBD-mRNA vaccine (MIPSCo-mRNA-RBD-1) is in a phase 1 trial, and its safety and immunogenicity will be compared with a MF59-adjuvanted Beta variant RBD subunit vaccine (DoCo-Pro-RBD-1 + MF59®) (NCT05272605).

6. Conclusion

SARS-CoV-2, particularly its current dominant variants or any future variant strains with pandemic potential, remains a serious threat to human health; thus there is an urgent need to develop safe and effective vaccines that show protective efficacy against infection by such variants. Among the four structural proteins of SARS-CoV-2, the S protein, particularly its RBD, plays an important role in viral infection and pathogenesis. This review article describes the structure and function of the SARS-CoV-2 S protein, including its RBD and other key regions, in mediating SARS-CoV-2 entry and membrane fusion, illustrates the function of RBD-based vaccines in inducing important adaptive immune responses in neutralizing SARS-CoV-2 or preventing the virus from further infection, and summarizes the currently developed COVID-19 vaccines based on the RBD protein of SARS-CoV-2. These vaccines are categorized as subunit, nanoparticle, mRNA, and other vaccine types in preclinical development, as well as several vaccines under clinical trials. Potential approaches on how to improve the immunogenicity and/or efficacy of these vaccines are also demonstrated in the related categories. Taken together, this review is expected to provide guidelines on the rational design and development of effective vaccines to prevent the current and future COVID-19 pandemics.

7. Expert opinion

Coronaviruses infect host cells by first binding to a cellular receptor through an RBD located in the S1 subunit of the S protein. Fusion between the virus and cell membranes is mediated by the S2 subunit. Therefore, the S protein, especially its RBD, plays a critical role in viral infection and pathogenesis. Among the three highly pathogenic coronaviruses, SARS-CoV-2 and SARS-CoV use the same receptor (ACE2) for viral binding, entry, and infection, whereas MERS-CoV utilizes a different receptor (dipeptidyl-peptidase 4: DPP4) [22,25,139].

Similar to the RBDs of pathogenic coronaviruses, such as SARS-CoV and MERS-CoV, the RBD of SARS-CoV-2 is a critical target for development of safe and effective vaccines and therapeutic antibodies to prevent and treat SARS-CoV-2 infection. Indeed, most neutralizing antibodies that show high potency and strong efficacy against SARS-CoV-2 infection target the RBD region [7,140–144]. As described in the previous paragraphs, COVID-19 vaccines targeting the RBD of the original SARS-CoV-2 strain, or previously dominant VOCs, induce potent neutralizing antibodies and/or cellular immune responses that protect humans and animals against infection by earlier SARS-CoV-2 strains or variants.

Different from SARS-CoV and MERS-CoV, the RBD of SARS-CoV-2 variants, particularly Omicron and its subvariants, contains multiple amino acid mutations [145]. Thus, the efficiency of RBD-based vaccines based on the original strain or other VOCs is compromised; these vaccines demonstrated reduced neutralizing activity and/or protection against the Omicron variant and its offspring subvariants [69,107]. In addition, additional SARS-CoV-2 VOC might cause future pandemics. Therefore, great efforts are required to develop RBD-based COVID-19 vaccines with improved efficacy against the current dominant strains, as well as future strains with pandemic potential.

Combined vaccination strategies, including priming or boosting with the same or different vaccine types, or combined immunizations, have the potential to generate alleviated immune responses against multiple variants. For instance, two doses of a RBD protein vaccine, followed by an original SARS-CoV-2 S-targeting mRNA vaccine, increased the titer of RBD-specific antibodies against the Delta variant, as well as neutralizing antibodies against the Omicron variant [146]. In addition, one dose of a variant S-targeting mRNA vaccine, followed by two doses of an original RBD-based mRNA vaccine, significantly enhanced broadly neutralizing activity against multiple SARS-CoV-2 variants, including several Omicron subvariants [36]. Due to significant variations in the S protein, especially the RBD, vaccines based on the RBD or S protein of the original SARS-CoV-2 strain or its variants normally generate a strong neutralizing antibody response and protection against the original or the respective variant strain, rather than against all strains (i.e. the original plus variant strains) [11,36]. Thus, cocktails combining two or more vaccines specific for the original strain and different variants are expected to elicit immune responses with broad-spectrum efficacy.

In addition to improved efficacy against various strains, particularly current and future VOCs with pandemic potential, COVID-19 vaccines need to be safe. It is of note that SARS-CoV-2 S-targeting mRNA or viral vectored vaccines may cause local and systemic reactions, allergic reactions, or other adverse events such as cardiac arrest, inflammation, or organ dysfunction [8]. By comparison, the RBD fragment is much shorter than the S protein, and optimized RBD-based vaccines may induce potent neutralizing antibodies without side effects; indeed, RBD-based subunit vaccines are much safer. It is worth noting that lipids, polymers, and other auxiliary components in mRNA or nanoparticle vaccines might also cause toxicity [147,148]. Steps must be taken to reduce the potential side effects caused by RBD-based COVID-19 vaccines.

Unlike SARS-CoV-2 S-targeting mRNA or other vaccines, most of the currently developed SARS-CoV-2 RBD-based vaccines are still under preclinical development or being evaluated in small animal models. Efforts are needed to rapidly evaluate the efficacy and safety of promising RBD-based vaccines, or their combinations with other vaccines, in large animal models, including NHPs, against multiple SARS-CoV-2 variants. Accordingly, sufficient funds need to be made available to support such vaccine development. In the next five years, we expect that vaccines targeting the RBD efficiently and safely will proceed to human clinical trials and be approved for clinical use. The potential threat from other coronaviruses, such as SARS-related coronaviruses that use the same ACE2 receptor as SARS-CoV-2 and SARS-CoV, or MERS-CoV or MERS-related coronaviruses that use a different receptor such as DPP4 [1,22,149–152], means that universal vaccines need to be developed to prevent future pandemics.

Article highlights

• Coronavirus Disease 2019 (COVID-19), first reported in December 2019, caused a global pandemic, leading to millions of deaths within 3 years. Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) is the causative agent of COVID-19.

• Like SARS-CoV and MERS-CoV, two other highly pathogenic coronaviruses reported in 2002 and 2012, respectively, SARS-CoV-2 is an enveloped and single-stranded RNA virus, belonging to the genus of Betacoronavirus (family Coronaviridae; and order, Nidovirales). The viral genome encodes four structural proteins, including spike (S), membrane (M), envelope (E), and nucleocapsid (N), as well as a number of non-structural proteins.

• The S protein of SARS-CoV-2 plays an important role in virus infection and pathogenesis. It comprises S1 and S2 subunits. SARS-CoV-2 infection is initiated by binding of the receptor-binding domain (RBD) in the S1 subunit to a host cellular receptor; viral entry and membrane fusion are mediated by the S2 subunit. SARS-CoV-2 and SARS-CoV use angiotensin-converting enzyme 2 (ACE2) for viral entry, whereas MERS-CoV utilizes dipeptidyl-peptidase 4 (DPP4) as its cellular receptor. Therefore, the S protein, particularly the RBD, of SARS-CoV-2 is a critical target for development of safe and effective COVID-19 vaccines.

• RBD-targeting COVID-19 vaccines prevent SARS-CoV-2 infection by inducing potent neutralizing antibodies against SARS-CoV-2, or by eliciting specific T cell responses to promote antibody production or direct killing of virus-infected cells.

• Most of the currently developed RBD-targeting COVID-19 vaccines are in preclinical development. They are categorized as recombinant subunit vaccines, nanoparticle vaccines, mRNA vaccines, and other vaccine types, such as those based on viral vectors.

• Several RBD-targeting COVID-19 vaccines have proceeded to clinical trials. It is expected that promising RBD-based vaccines can be approved for use in humans to prevent SARS-CoV-2 infection.

• Unlike SARS-CoV and MERS-CoV, SARS-CoV-2 mutates rapidly, resulting in multiple variants of concern (i.e. Alpha, Beta, Gamma, Delta, and Omicron); the Omicron variant has various subvariants that transmit rapidly and represent the current dominant strains. Most of these variants are highly resistant to the current vaccines, which target the original strain or previous variants. Novel approaches are needed urgently to develop safe and effective COVID-19 vaccines, including those based on the RBD, that show improved efficacy against the current pandemic variants, as well as future variants with pandemic potential.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contribution

Y.Y. prepared the structural figures. X.G. prepared the other figures and participated in the writing. L.D. designed, wrote, and revised the manuscript.

Additional information

Funding

This study was supported by the National Institutes of Health (NIH) grants (R01AI157975, R01AI139092, and R01AI137472)