An engineered recombinant protein containing three structural domains in SARS-CoV-2 S2 protein has potential to act as a pan-human coronavirus entry inhibitor or vaccine antigen

ABSTRACT

The threat to global health caused by three highly pathogenic human coronaviruses (HCoV), SARS-CoV-2, MERS-CoV and SARS-CoV, calls for the development of pan-HCoV therapeutics and vaccines. This study reports the design and engineering of a recombinant protein designated HR1LS. It contains three linked molecules, each consisting of three structural domains, including a heptad repeat 1 (HR1), a central helix (CH), and a stem helix (SH) region, in the S2 subunit of SARS-CoV-2 spike (S) protein. It was found that HR1LS protein automatically formed a trimer able to bind with heptad repeat 2 (HR2) region in the SARS-CoV-2 S2 subunit, thus potently inhibiting HCoV fusion and entry into host cells. Furthermore, immunization of mice with HR1LS, when combined with CF501 adjuvant, resulted in the production of neutralizing antibodies against infection of SARS-CoV-2 and its variants, as well as SARS-CoV, MERS-CoV, HCoV-229E, HCoV-NL63 and MjHKU4r-CoV-1. These results suggest that HR1LS is a promising candidate for further development as a novel HR1-trimer-based pan-HCoV entry inhibitor or vaccine for the treatment and prevention of infection by SARS-CoV-2 and its variants, but also other HCoVs with the potential to cause future emerging and re-emerging infectious coronavirus diseases.

Introduction

Over the last three years, the global COVID-19 pandemic caused by SARS-CoV-2 and its variants has continued to pose a serious threat to public health, economic development and social stability around the world. Although several drugs and vaccines have been approved, the continuous emergence of new variants has caused varying degrees of escape from first-generation vaccines [1]. Along with this threat, Middle East respiratory syndrome coronavirus (MERS-CoV) infection cases are still being reported in Qatar and Saudi Arabia [2]. Although MERS-CoV has a much lower transmission rate than that of SARS-CoV-2, its case-fatality rate (∼35%) is much higher than that of SARS-CoV-2. It is most alarming that coinfection of SARS-CoV-2 and MERS-CoV in some patients could result in the emergence of a new β-CoV clade, e.g. SARS-CoV-3 or MERS-CoV-2. This was hypothesized to occur from genetic recombination between SARS-CoV-2 and MERS-CoV, resulting in a high SARS-CoV-2-like transmission rate and a high MERS-CoV-like case-fatality rate [3]. Manis javanica HKU4-related coronavirus (MjHKU4r-CoV), a bat MERS-like coronavirus, uses human DPP4 as a receptor, and it also presents the risk of spilling over into humans [4]. Thus, when coupled with endemic HCoVs, such as HCoV-229E, HCoV-NL63, and HCoV-OC43, which typically cause respiratory illnesses with varying severity [5], the expanded and collective risk calls for the urgent development of pan-CoV therapeutics, prophylactics and vaccines.

Spike (S) protein of HCoV consists of S1 and S2 subunits, which play important roles in receptor-binding and viral fusion, respectively. The receptor-binding domain (RBD) in S1 subunit contains the main neutralizing epitopes. As such, the RBD was a key target for the development of first-generation COVID-19 vaccines for eliciting neutralizing antibody (nAb) response. However, researchers have since learned that the highly variable amino acid sequence of RBD in HCoV S protein is not a good target for the design and engineering of pan-CoV vaccines [6–12]. Meanwhile, the S2 subunit in S protein of HCoV contains fusion peptide (FP), heptad repeat 1 (HR1), central helix (CH), connector domain (CD), stem helix (SH), heptad repeat 2 (HR2), and transmembrane anchor (TM) [13] (Figure 1A). Since it is relatively more conserved than S1 subunit [14], albeit with weaker capacity to elicit nAb responses, it still makes a better target for the development of pan-CoV vaccines [15].

Figure 1.

Pattern diagram of SARS-CoV-2 S2 subunit and conserved sequence analysis and design of trimeric proteins. (A) Schema showing SARS-CoV-2 S2 subunit and amino acids of heptad repeat 1 (HR1), central helix (CH) and stem helix (SH). (B) Amino acid alignment of HCoV CH region. (C) Amino acid alignment of HCoV SH region. (D) Amino acid alignment of HCoV HR1 region. (E) 3D structure of SARS-CoV-2 spike protein in prefusion conformation with the HR1, CH, and SH segments coloured blue, red, and orange, respectively (PDB: 6XR8). (F) 3D structure of SARS-CoV-2 S2 subunit in the postfusion conformation. Structures are indicated as palegreen cartoon. The HR1, CH, and SH segments were coloured blue, red, and orange, respectively. The HR2 in the postfusion state of the SARS-CoV-2 S2 subunit was coloured magenta (PDB: 6XRA). (G) Schema of trimeric proteins HR1, HR1S, and HR1LS. Blue, red and green bars represent HR1, CH, and SH region, respectively. His-tag was fused at C-terminus of protein for purification. White arrow indicates amino acid sequence from N-terminus to C-terminus.

It has been previously reported that some vaccine candidates contain HR1 or SH domain, including a recombinant protein HR121 [16], a trimeric S2 protein [17], and a self-assembled RBD-HR1/2 trimeric protein [18]. Particular, the vaccine candidate HR121 consisting of HR1-HR2-HR1 domains that forms a unique fusion intermediate conformation could elicit potent cross-nAbs against infection of SARS-CoV-2 and its variants by mainly targeting the HR1 domain [16], indicating that the HR1 domain in SARS-CoV-2 S2 subunit may serve as an antigen of pan-CoV vaccines. Different from the HR1 domain of HIV-1 gp41, the HR1 and CH domains of SARS-CoV-2 form an unusually long, central, three-stranded coiled coil in the postfusion state [13]. Several nAbs, such as S2P6, targeting the SH region exhibit broad neutralization activity against betacoronaviruses, including SARS-CoV, MERS-CoV, SARS-CoV-2 and its variants [19–21], suggesting that the SH domain contains a conserved neutralizing epitope. Previously, a vaccine based on MERS-CoV S2 subunit could elicit SH-targeted nAbs in mice [22]. However, no one has been successful so far in developing a coronavirus vaccine that contains SH domain only, possibly because it is extremely difficult to keep the SH domain-containing vaccine in a proper conformation to elicit neutralizing antibodies. Sequence analysis showed that structural domains of HR1, CH, and SH in the S2 subunit of SARS-CoV-2, SARS-CoV, MERS-CoV, HCoV-OC43, HCoV-229E, and HCoV-NL63 are relatively conserved by sharing a number of identical amino acids [23] (Figure 1B–D). In addition, recombinant proteins contain HR1 domain can also be developed as viral fusion and entry inhibitors, such as HR1MFd and HR121, for prevention and treatment of SARS-CoV-2 infection [24,25].

In this study, we designed and engineered three recombinant proteins, HR1, HR1S, and HR1LS, containing 3 HR1, 3 HR1-SH, and 3 HR1-CH-SH, respectively (Figure 1). These recombinant proteins are expected to form trimers able to bind to HR2P, the peptide derived from the SARS-CoV-2 S protein HR2 region (aa 1168-1203), and they all exhibited potent inhibitory activity against SARS-CoV-2 pseudovirus infection. Among them, HR1LS had the most potent inhibitory activity against SARS-CoV-2 infection with an IC₅₀ (half maximal inhibitory concentration) of 0.17 µM. Even more striking, HR1LS showed broad inhibitory activity against infection of pseudotyped SARS-CoV, MERS-CoV, HCoV-229E, HCoV-NL63, and MjHKU4r-CoV-1. HR1LS was effective against authentic HCoV-OC43 infection in vitro and protect the neonatal mouse from HCoV-OC43 infection in vivo. Immunization of mice with HR1, HR1S, and HR1LS, when combined with Aluminium (Alum) adjuvant, respectively, could elicit a low level of neutralizing antibodies (nAbs) against SARS-CoV-2 infection. Again, HR1LS induced a titre of nAbs higher than that of HR1 or HR1S. However, it was also found that mice immunized with HR1LS, when combined with STING agonist-based adjuvant CF501 [26–28], induced the production of broad-spectrum nAbs against SARS-CoV-2, as well as SARS-CoV and MERS-CoV. To our surprise, mouse sera of the HR1LS/CF501 group could also neutralize infection of HCoV-229E and HCoV-NL63, both of which belong to the group of alphacoronaviruses, and MjHKU4r-CoV-1, a bat MERS-like coronavirus [4]. Taken together, these findings suggest that adjuvanted HR1LS is a promising candidate for development as a pan-HCoV vaccine, as well as entry inhibitor-based therapeutics and prophylactics.

Materials and methods

Cells, peptides, and plasmids

293 T, RD, and Caco2 cells were purchased from ATCC and stocked in our laboratory. 293 T and RD cells were cultured with DMEM containing 10% FBS. Caco2 cells were cultured in MEM containing 10% FBS. The peptide SARS-CoV-2 HR1P (aa 924-965) and HR2P (aa 1168-1203) were synthesize by Synpeptide Co., Ltd, with the sequence of ANQFNSAIGKIQDSLSSTASALGKLQDVVNQNAQALNTLVKQ and DISGINASVVNIQKEIDRLNEVAKNLNESLIDLQEL.

Protein expression and purification

The reference sequence of HR1 is TQNVLYENQKLIANQFNSAIGKIQDSLSSTASALGKLQDVVNQNAQALNTLVKQLSSNFGAISSVLNDILSRLDKVE (Genebank: YP_009724390.1). HR1 protein was constructed with three HR1 that was linked with two linkers, GKGNQ and GLIDG, respectively. For HR1S, based on recombinant protein HR1, SH region was inserted after each HR1 segment. For HR1LS, the CH segment and SH segment were inserted after HR1. His-tag were added to the C-terminus with GGGGS linker for protein purification.

For protein expression, E. coli BL21 (DE3) strain was transformed with the plasmids of pet30a-HR1-His and amplified to express HR1 recombinant proteins at 16^oC for 20 h under the induction of IPTG at a final concentration of 0.5 mM. Bacteria were centrifuged, resuspended in lysis buffer (50 mM Tris-HCl, 500 mM NaCl, pH 8.6) and disrupted by ultrasonication, followed by centrifugation at 10,000 rpm for 30 min. Supernatant was purified by Ni Sepharose (GE), washed with buffer containing 50 mM imidazole, and eluted with buffer containing 250 mM imidazole. Proteins were concentrated by ultrafiltration and redissolved in PBS. Protein integrity was analyzed by SDS-PAGE. Protein concentration was determined by the BCA Kit (Takara).

Circular dichroism (CD) spectroscopy

The secondary structure of HR1 recombinant proteins was assessed by CD spectroscopy as described previously [29]. Briefly, proteins were dissolved in PBS (pH 7.4) with final concentration as 15 µM and then tested on a Jasco spectropolarimeter (Model J-815), using a 1 nm bandwidth with a 1 nm step resolution from 200 to 260 nm at 25^oC. The baseline curve was determined on PBS alone. The CD signal was converted into a molar ellipsometry curve by the number and molar concentration of amino acids in the sample. The a-helical content was calculated as described previously.

Bio-layer interferometry, BLI

The binding kinetics of recombination proteins to HR2P were measured by BLI on an Octet-RED96 (ForteBio) [30]. Biotinylated HR2P was synthesized by Synpeptide Co., Ltd and was loaded onto streptavidin-coated (SA) biosensors at 5 μg/mL. Recombination proteins were serially diluted 1:3 starting at 10 μM with PBST (PBS containing 0.02% Tween-20, 0.1% BSA). After association with recombination proteins for 600 s at 30°C, the antigen immobilized sensors were immersed into PBST for another 600 s. All the curves were fitted with 1:1 binding model and K_D values were determined only if R2 values of greater than 95% confidence level.

HCoV-OC43 viral challenge in neonatal mouse

Pregnant ICR mice (18 days) were separated into six groups and each group contained six newborn mice. For mouse in the prophylactic and therapeutic groups, HR1LS was administered through the intranasal route at 10 or 25 mg/kg before or after 0.5 h challenge with HCoV-OC43. After 5 d post-infection, the neonatal mouse was dissected. The relative viral RNA expression level in brain was tested through RT-PCR and adjusted with mouse housekeeping gene GAPDH as previously described [31].

Authentic HCoV-OC43 inhibition assay

RD cells were seeded into wells of a 96-well plate (10,000 cells per well) 12 h in advance. Recombinant proteins were serially diluted 1:3 in DMEM, followed by incubation with 100 TCID50 HCoV-OC43 at 37°C for 30 min. Then the mixture was transferred into RD cells. After 12 h, the mixture was replaced with fresh DMEM containing 2% FBS and cell viability was detected with Cell Counting Kit-8 (CCK-8).

Animal vaccination

In the first animal immunization, specific-pathogen-free (SPF) female Balb/c mice (six weeks old) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. Twenty Balb/c mice were randomly divided into 4 groups. HR1, HR1S, and HR1LS proteins were dissolved in PBS, and each mouse was intramuscularly immunized with 20 μg protein and an equal volume of injectable aluminium adjuvant (Thermo Scientific). Another group of mice was intramuscularly injected with an equal volume of sterile PBS as a negative control. Mice were immunized at day 0, 28, and 108. Blood samples were collected from the infraorbital vein 2 weeks after immunization to analyze antigen-specific Ig and neutralization activity of nAbs against pseudotyped coronavirus.

In the second animal immunization, 35 Balb/c mice were randomly divided into 5 groups. HR1 and HR1LS proteins were selected as immunogens, and each group of mice was intramuscularly immunized with 50 μg HR1 + 20 μg CF501, 50 μg HR1 + an equal volume of injectable aluminium adjuvant, 50 μg HR1LS + 20 μg CF501, 50 μg HR1 + an equal volume of injectable aluminium adjuvant, and an equal volume of sterile PBS, respectively. Mice were immunized at day 0, 21, 42, 70, and 119. Blood samples were collected from the infraorbital vein 2 weeks post-second immunization to analyze antigen-specific antibodies and neutralization activity of nAbs against pseudotyped coronavirus. After the last blood collection, mice were euthanized, and spleens were collected to perform an enzyme-linked immunospot (ELISpot) assay. All animal experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the Institutional Laboratory Animal Care of Fudan University (20201111-001).

Inhibition of pseudotyped coronavirus infection

The inhibitory activity of mouse serum and HR1 recombinant proteins against pseudotyped coronavirus infection was evaluated as described previously [31–33]. For pseudotyped coronavirus production, pNL4-3.Luc.R-E- was co-transfected with pcDNA3.1-SARS-CoV-2-S or plasmid encoding S protein of other coronaviruses into 293 T cells using Vigofect transfection reagent. After incubating at 37^oC for 6–8 h, medium was replaced with fresh DMEM containing 10% FBS. Culturing continued for another 48 h, and then the supernatants containing pseudotyped coronavirus were collected. For inhibitory activity detection, Caco2 cells were seeded into wells of a 96-well plate (8000 cells per well) 12 h in advance. Mouse serum (inactivation at 56^oC in advance) and HR1 recombinant proteins were serially diluted 1:3 or 1:4 in MEM, followed by incubation with pseudotyped coronavirus at 37^oC for 30 min. Then the mixture was transferred into Caco2 cells. After 12 h, the mixture was replaced with fresh MEM containing 10% FBS. After culturing for another 48 h, the cells were lysed with cell lysis buffer (Promega, USA), and luciferase activity was detected with the Luciferase Assay System (Promega, USA). The 50% neutralization titre (NT₅₀) was determined using the half-maximal inhibitory concentration values of serum samples from their serial dilution. The input dilution of serum was 1:90, samples that did not neutralize 50% infection of pseudotyped coronaviruses at 1:90 were plotted as 1:45, which was used for geometric mean calculation [26,32,33].

Enzyme-linked immunosorbent assay (ELISA)

ELISA was used to quantify the affinity of HR1 recombinant proteins to the HR2P peptide. Briefly, ELISA plates were coated with 5 µg/mL HR2P [34] overnight at 4^oC. The coated plates were blocked using PBST containing 5% BSA at 37^oC for 2 h. Proteins were serially diluted 1:4 in PBST, followed by incubation with the coated ELISA plates at 37^oC for 1 h. The plates were washed with PBST 5 times, before incubation with HRP-conjugated 6×His, His-Tag Monoclonal Antibody (Proteintech) at 37^oC for 30 min. TMB (3,3’, 5,5'-tetramethylbenzidine) was used to visualize the enzyme-linked reaction, and H₂SO₄ was used to stop the reaction. Absorbance at 450 nm was measured using a SpectraMax i3x (Molecular Devices).

ELISA was also used to quantify antigen-specific immunoglobulin titre in mouse sera as described previously [35]. Briefly, wells of ELISA plates were coated with 2 µg/mL of each recombinant protein tested overnight at 4^oC. The coated plates were blocked using PBST containing 5% BSA at 37^oC for 2 h. Mouse serum was serially diluted 1:4 in PBST before addition to the wells of coated ELISA plates. After incubation at 37^oC for 1 h, plates were washed 5 times with PBST before incubation with HRP-conjugated Rabbit anti-Mouse IgG (Dako, Denmark) at 37^oC for 30 min. The enzyme-linked reaction was activated by TMB and terminated by H₂SO₄ as previously described. ELISA endpoint titre was expressed as the highest reciprocal serum dilution exhibiting OD450 > 2.1-fold over the background values. The input dilution of serum was 1:100, and the titre was plotted as 1:50, if less than 1:100.

SH-specific immunoglobulin titre in mouse serum was also quantified by ELISA. Plates were coated with 2 µg/mL streptavidin overnight at 4^oC and then blocked with PBST containing 5% BSA, followed by incubation with 5 µg/mL biotin-SH peptide (Biotin-GGGLDSFKEELDKYFKNHTSP-NH2, synthesized by Synpeptide Co., Ltd) at 37^oC for 1 h. Mouse serum was serially diluted three-fold in PBST before addition to the wells of coated ELISA plates. After incubation at 37^oC for 1 h, plates were washed 5 times with PBST before incubation with HRP-conjugated Rabbit anti-Mouse IgG at 37^oC for 30 min. Enzyme-linked reaction was activated by TMB and terminated by H₂SO₄ as previously described.

Enzyme-linked immunospot (ELISpot)

IFN-γ ELISpot assays were conducted as previously described [26]. Briefly, ELISpot plates were blocked using 1640 containing 10% FBS at 37^oC for 2 h. Spleen cell suspension was prepared, and red blood cells were lysed by red blood cell lysate (Beyotime Biotechnology). Cells were counted and added into the plates (1 × 10⁶ per well) together with 1 µg/mL SARS-CoV-2 S2 full-length scan peptide library (synthesized by GL Biochem) and incubated at 37^oC for 48 h. Plates were washed 5 times with sterile PBS before incubation with 1 µg/mL biotin-anti-IFN-γ antibody for 2 h at room temperature. After washing with PBS, streptavidin-ALP at 100 µg/well was added and incubated for 1 h at room temperature. Substrate solution (BCIP/NBT) was added to visualize the enzyme-linked reaction, and the antigen-specific spots were counted using an AID ELISpot Reader System (AID, Strasburg, Germany).

Cytotoxicity assay

Cytotoxicity of HR1 recombinant proteins to Caco2 cells was tested by CCK-8 as previously described [36]. Briefly, HR1 recombinant proteins were added to cells cultured in a 96-well cell culture plate (8,000 cells per well) after serial dilution. After incubating at 37^oC for 12 h, medium was replaced with fresh MEM containing 10% FBS. Culturing continued for 48 h, and then CCK-8 solution was added to Caco2 cells (10 μL per well). Absorbance was measured at 450 nm.

Statistical analysis

Student’s t-test and Analysis of Variance (ANOVA) were used to compare the difference by GraphPad Prism 8. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Results

Design and characterization of recombinant trimeric HR1, HR1S, and HR1LS proteins

Considering the conservation of the S2 subunit, we used HR1, CH, and SH segments to construct several recombinant proteins. In the prefusion state, the HR1 and CH segment encapsulated in the S1 subunit, SH region exposed at the bottom (Figure 1E). While in the postfusion state, the HR1 and CH formed a trimer with the binding of HR2, and SH region formed a short helix close to the CH region (Figure 1F). According to previous studies [16,37,38], an antiparallel trimeric protein HR1 was designed to mimic the SARS-CoV-2 HR1 trimer. Three HR1 helical segments were linked with two short polypeptide linkers to form an HR1-linker-HR1-linker-HR1 structure. The first and third HR1 was sequenced as N-terminus to C-terminus, and the second HR1 was turned as C-terminus to N-terminus. For HR1S, SH fragment was inserted after each HR1 fragment to form a trimer structure. For the design of HR1LS, each CH fragment was inserted between the HR1 and SH fragments of HR1S trimer (Figure 1G). All proteins were fused with His-Tag at their C-terminus for purification. The amino acid sequences were shown in Table S1. SDS-PAGE analysis demonstrated that these proteins had been successfully expressed and purified. The bands of HR1, HR1S, and HR1LS in SDS-PAGE were consistent with the expected molecular weight (26.7, 31.4, and 39.2 KD, respectively) (Figure 2A). HR1, HR1S, and HR1LS had no significant cytotoxicity to Caco2 cells at the concentration as high as 20 µM (Figure 2B). Their secondary structure was then determined by CD spectroscopy. As shown in Figure 2C, all CD spectra showed peaks at 208 and 222 nm, indicating that these proteins had a typical α-helix structure. Then we used the peptide HR2P that derived from SARS-CoV-2 S protein HR2 region [34] to evaluate the ability of HR1, HR1S, and HR1LS to bind the HR2 domain of SARS-CoV-2 S protein. As expected, all proteins could bind the HR2P peptide derived from the HR2 domain of SARS-CoV-2 spike protein in a dose-dependent manner (Figure 2D), with the K_D of 10.32, 3.12, 7.03 nM (Figure 2E–G). AlphaFold2 [39] was used to predict 3D structures of HR1, HR1S, and HR1LS. As shown in Figure S1A, all these three recombinant proteins adopt regular chimeric trimers. This arrangement involved two forward helices and one reverse helix tightly packed together to form a triple helical bundle. The hydrophobic grooves responsible for viral HR2-peptide binding were fully exposed on the surface. Notably, in all the three structures, the majority of residues within the triple helical region exhibited pLDDT scores above 70, indicating a high level of model confidence in the predictions made by AlphaFold2 (Figure S1B). These results suggest that insertion of the CH and SH regions have no significant impact on the structure and function of HR1.

Figure 2.

Characterization of trimeric protein. (A) SDS-PAGE analysis of recombinant proteins HR1, HR1S, and HR1LS. (B) Cytotoxicity of HR1, HR1S, and HR1LS in Caco2 cells. (C) Secondary structure of HR1, HR1S, and HR1LS. The α-helicity of these proteins detected at the concentration of 15 µM by CD spectroscopy. (D) Binding effect of HR1, HR1S, and HR1LS to SARS-CoV-2 HR2P. (E-G) BLI sensorgrams and kinetic of HR1, HR1S, and HR1LS binding to HR2P, respectively. Five fitting curves are shown with the K_D values.

Inhibitory activity of HR1, HR1S, and HR1LS trimeric proteins against infection of HCoVs

To determine the antiviral activity of these trimeric proteins, pseudovirus inhibition assays were performed as previously reported [31]. As shown in Figure 2A, the trimeric HR1, HR1S, and HR1LS proteins efficiently inhibited SARS-CoV-2 pseudovirus infection with IC₅₀ values of 0.27, 0.24, and 0.17 µM, respectively (Figure 3Aa). Meanwhile, the HR1P peptide exhibited no detectable inhibitory activity against infection of pseudotyped SARS-CoV-2 at the concentration of 20 µM. The three recombinant trimeric proteins could also potently inhibit infection of pseudotyped SARS-CoV-2 Delta variant, as well as Omicron variant, such as BA.2.75, BQ.1, XBB, and CH.1.1. Among them, HR1LS had the best inhibitory activity against these viruses with IC₅₀ values ranging from 0.20 to 0.52 µM (Figure 3Ab–Af, Table S2). Since the HR1 region of coronaviruses was conserved, the inhibitory effect of these proteins was also tested on pseudotyped SARS-CoV, MERS-CoV, HCoV-229E, HCoV-NL63 and MjHKU4r-CoV-1. Consistent with the results of SARS-CoV-2, HR1LS showed the best inhibitory activity against these HCoVs with IC₅₀ values ranging from 0.19 to 1.15 µM (Figure 3B–F). All these results suggest that trimeric HR1LS possesses potent inhibitory activity with the potential to be developed as a pan-HCoV inhibitor-based therapeutic or prophylactic.

Figure 3.

Inhibitory activity of HR1, HR1S, and HR1LS against HCoVs and bat coronaviruses. Inhibitory activity of HR1, HR1S, and HR1LS against pseudotyped SARS-CoV-2 WT (Aa), B.1.617.2 (Ab), BA.2.75 (Ac), BQ.1 (Ad), XBB (Ae), CH.1.1 (Af), SARS-CoV (B), MERS-CoV (C), HCoV-229E (D), HCoV-NL63 (E), and MjHKU4r-CoV-1 (F) detected in Caco2 cells.

The protective effect of HR1LS in mice infected with HCoV-OC43

To evaluate the protective activity of recombinant proteins against live HCoV-OC43 infection, we primarily tested its inhibitory activity against HCoV-OC43 infection in RD cells. Figure 4A shows that the recombinant HR1, HR1S, and HR1LS proteins could effectively inhibit HCoV-OC43 infection with the IC₅₀ of 3.76, 4.40, and 1.83 µM, respectively. We then used the established HCoV-OC43-infected neonatal mouse model [31] to evaluate the inhibitory activity of HR1LS in vivo. For the prophylactic effect, the neonatal mice were administered with HR1LS intranasally 30 min before the HCoV-OC43 challenge. The viral RNA was detected at the 5-day post-virus challenge (Figure 4B). Compared with the PBS-treated group, the viral RNA levels in mouse brain were significantly decreased in mice treated with HR1LS at 10 and 25 mg/kg (Figure 4C). The results suggest that HR1LS has preventive effect in vivo against HCoV-OC43 infection. Meanwhile, we also evaluated the therapeutic effect of HR1LS. After the challenge with HCoV-OC43, the neonatal mice were continuously given HR1LS by intranasal route for three days (Figure 4B). The viral titres in mouse brain were also significantly decreased (Figure 4D), suggesting that HR1LS also has treatment effect against HCoV-OC43 infection in vivo. Altogether, the recombinant protein HR1LS is a potent fusion and entry inhibitor that can be used to prevent and treat HCoV-OC43 infection.

Figure 4.

The activity of HR1LS against HCoV-OC43 infection in vitro and in vivo. (A) Inhibitory activity of HR1, HR1S, and HR1LS against HCoV-OC43 infection in RD cells. (B) Sketch map for illustrating the prophylactic and treatment effect of HR1LS in HCoV-OC43-infected neonatal mice. The neonatal mice were challenged with HCoV-OC43 through the intranasal route. HR1LS was initially given to mice (5 d) intranally once 30 min before the virus challenge to evaluate the prevention effect. HR1LS was given to mice (5 d) intranasally once a day for three days continuously to assess the treatment effect. Then the mice were euthanized at 5 days post-infection to collect mouse brain. (C-D) The relative viral RNA level of HCoV-OC43 in mouse brain of each group.

Immunization of mice with trimeric HR1, HR1S, and HR1LS elicited specific immune responses

Since these recombinant proteins can form an HR1-containing trimer, they could be formulated with injectable aluminium adjuvant to immunize Balb/c mice intramuscularly three times at days 0, 28 and 108, accordingly, with sera collected at days 42 and 122, i.e. 2 weeks post-second and -third immunization (Figure 5A). After three immunizations, HR1, HR1S and HR1LS were found to elicit high-titre, HR1-specific IgG (Figure 5B–D), while HR1S and HR1LS also induced SH-specific antibody response (Figure 5E). Serum 50% neutralization titre (NT₅₀) was then measured against pseudotyped SARS-CoV-2 infection. At day 42, no significant difference was found in the level of nAbs induced by HR1, HR1S, and HR1LS formulated with Alum adjuvant (Figure 5F). However, at day 122, HR1S/Alum and HR1LS/Alum elicited substantially higher levels of serum nAbs with geometric mean (GM) NT₅₀ about 3-fold and 10.4-fold over that elicited by HR1/Alum (Figure 5G). This result suggests that trimeric HR1LS affords better neutralizing immunogenicity than either HR1S/Alum or HR1/Alum.

Figure 5.

Immunization with HR1LS elicited immune response in mice. (A) Illustration of immunization strategy. HR1, HR1S, and HR1LS formulated with adjuvant Alum immunized intramuscularly in mice at day 0, day 28, and day 108. Blood drawn at day 42 and day 122. (B-D) HR1-, HR1S-, and HR1LS-specific antibodies in mouse sera at day 122. (E) SH-specific antibodies in mouse sera at day 122. (F) and (G) Neutralization activity of mouse sera for inhibiting pseudotyped SARS-CoV-2 infection at day 42 and day 122. Each dot represents a mouse serum sample. Data are presented as geometric mean with geometric SD.

HR1LS combined with CF501 adjuvant elicited higher level of nAbs

It was previously shown that a CF501-adjuvanted RBD-Fc-based vaccine could elicit highly potent nAb and T cell responses in mice [26]. Here, mice were immunized with HR1LS formulated with CF501 adjuvant (HR1LS/CF501) with HR1/CF501, HR1/Alum and HR1LS/Alum as controls. Mice were vaccinated intramuscularly at days 0, 21, 42, 70, and 119, and sera were collected at days 35, 56, 84, and 133 post-first immunization, respectively (Figure 6A). CF501- and Alum-adjuvanted HR1 induced a similar level of immune responses at day 56 (Figure 6B), while HR1LS/CF501 elicited a significantly higher immune response compared to that of HR1LS/Alum (Figure 6C). The GM NT_50s of sera in mice immunized with HR1LS/CF501 for neutralizing SARS-CoV-2 were 2.7- and 2.4-fold over that of HR1/CF501 and HR1LS/Alum, respectively (Figure 6D). Compared with the results from the third immunization, HR1LS/CF501 and HR1LS/Alum induced similar immune responses at day 84 and day 133 (Figure 6E–G). However, the HR1LS/CF501 group produced more nAbs against SARS-CoV-2 with GM NT_50s of 535 and 987 at day 84 and day 133, which were 3.0- and 4.4-fold over that of the HR1LS/Alum group (Figure 6H and I). These results suggest that HR1LS/CF501 can induce stronger nAb responses compared to those of HR1LS/Alum.

Figure 6.

HR1LS formulated with CF501 adjuvant elicited stronger nAb responses against SARS-CoV-2 infection. (A) Illustration of immunization strategy. HR1 and HR1LS formulated with adjuvant Alum and CF501, respectively, immunized intramuscularly in mice at day 0, day 21, and day 42. Sera collected at day 35 and day 56. HR1LS/CF501 and HR1LS/Alum injected into mice intramuscularly through days 70 and 119. Sera collected at day 84 and day 133. (B) and (C) Antibodies specific for HR1 and HR1LS in mouse sera at day 56. (D) Neutralization activity of immunized mouse sera for inhibiting pseudotyped SARS-CoV-2 infection at day 56. (E) Antibodies specific for HR1LS in mouse sera at day 84. (F) and (G) Antibodies specific for HR1LS and SH peptide in mouse sera at day 133. (H) and (I) Neutralization activity of mouse sera for inhibiting pseudotyped SARS-CoV-2 infection at day 84 and day 133. Each dot represents a mouse serum sample. Data are presented as geometric mean with geometric SD.

HR1LS/CF501-immunized mouse sera could broadly neutralize infection of SARS-CoV, MERS-CoV, HCoV-229E, HCoV-NL63 and MjHKU4r-CoV-1

Since some antibodies targeting the SH region exhibited broad-spectrum cross-nAb response against betacoronaviruses [20], the serum nAb response of mice immunized with HR1LS/CF501 was detected against other pseudotyped HCoVs. Based on the full-length genome, MERS-CoV and HCoV-OC43 were organized into betacoronavirus, while HCoV-229E and HCoV-NL63 belong to alphacoronaviruses (Figure 7A). As shown in Figure 7B and C, mouse sera could neutralize pseudotyped SARS-CoV and MERS-CoV infection. Surprisingly, mouse sera could also neutralize pseudotyped HCoV-229E and HCoV-NL63 with GM NT_50s of 427 and 661 (Figure 7D,E). These results suggest that HR1LS/CF501 immunization induced broad-spectrum anti-HCoV nAb response. More importantly, the sera of mice immunized with HR1LS/CF501 could also neutralize the novel MjHKU4r-CoV-1 with GM NT₅₀ of 435 (Figure 7F). HR1LS/CF501 immunization could also elicit an increased level of IFN-γ, indicating that it also induced strong T-cell response (Figure 7G,H). Together, these results suggest that HR1LS formulated with CF501 can induce cross-nAb responses against SARS-CoV-2 and its variants, as well as other HCoVs, in immunized mice.

Figure 7.

Neutralization activity of mouse sera against infection of pseudotyped HCoVs and bat coronavirus. (A) Phylogenetic tree based on the nucleotide sequences of the complete SARS-CoV, SARS-CoV-2, MERS-CoV, HCoV-OC43, HCoV-229E, and HCoV-NL63 genomes. (B-F) Cross-neutralization activity of sera from mice immunized with HR1LS stimulated with adjuvant CF501 against pseudotyped SARS-CoV (B), MERS-CoV (C), HCoV-229E (D), HCoV-NL63 (E), and MjHKU4r-CoV-1 (F) PsV at day 133. Each dot represents a mouse serum sample. (G) and (H). SARS-CoV-2 HR1-specific T cell immune response. Spleens collected at day 133. ELISpot used to determine the abundance of IFN-γ⁺ splenocytes.

Discussion

The continuous global COVID-19 pandemic caused by SARS-CoV-2 and its variants is now accompanied by MERS caused by the re-emerged MERS-CoV in Qatar and Saudi Arabia [2]. This pairing has raised the spectre of a potentially emerging novel recombinant coronavirus that may possess high SARS-CoV-2-like transmissibility and high MERS-CoV-like case-fatality rate, thus calling for the development of pan-HCoV vaccines and therapeutics [3,40]

HR1-trimer, formed by foldon-conjugated HR1 fragments of SARS-CoV-2 S protein, has been reported as a potent pan-CoV fusion inhibitor with potential for clinical development [24]. However, its future application may be limited because of the high immunogenicity of foldon (Fd) and the C-terminal domain of T4 fibritin able to induce strong anti-foldon antibody responses capable of attenuating the antiviral activity of HR1Fd-trimer. HR121 consisting of 2 HR1 domains and 1 HR1 domain formed a stable nonsymmetric 6-HB structure and exhibited broad-spectrum inhibitory activity against infection of SARS-CoV-2 and its variants, without showing its activity against other HCoVs [16,25]. Different with HR121, the recombinant protein HR1LS designed and developed in this study consists of HR1, CH, and SH domains, but contains no HR2 domain. HR1LS protein is expected to automatically form trimer by three linked HR1 domains of SARS-CoV-2 S protein in the absence of foldon (Figure 1G). This de novo HR1-trimer exhibited potent and broad-spectrum inhibitory activity against infection by SARS-CoV-2 and its variants, as well as other HCoVs, including SARS-CoV, MERS-CoV, HCoV-OC43, HCoV-229E, HCoV-NL63, and MjHKU4r-CoV-1. In particular, HR1LS-trimer was found to have more pan-HCoV inhibitory activity than HR1-trimer, suggesting that CH and SH domains do not attenuate, but rather enhance, the antiviral activity of HR1LS. Besides, HR1LS through intranasal administration could protect mice against HCoV-OC43 infection, suggesting that HR1LS has the potential to be formulated as topical or inhalation-based formulations to combat HCoV infection, similar to heparin that was previously reported [41].

HR121, as an immunogen, was previously found to elicit nAb response against SARS-CoV-2 infection in animals [16]. It was discovered that recombinant proteins containing SH fragment of MERS-CoV and SARS-CoV-2 S protein S2 subunits could also induce nAbs in immunized animals against SARS-CoV-2 and its variants, as well as other HCoVs [22,42]. Therefore, in this study, mice were immunized with HR1 and HR1LS formulated with CF501 adjuvant, or Alum adjuvant as a control. HR1LS/Alum could induce higher titre of nAbs against SARS-CoV-2 pseudovirus than HR1/Alum, but HR1LS/CF501 could elicit significantly stronger nAb response when compared to either HR1/CF501 or HR1LS/Alum. In addition, nAbs elicited by HR1LS/CF501 were effective against other HCoVs, including SARS-CoV, MERS-CoV, HCoV-229E, and HCoV-NL63. These results suggest that HR1LS/CF501 is a promising candidate for development as a pan-HCoV vaccine. Strikingly, the antisera induced by HR1LS/CF501 could also neutralize MjHKU4r-CoV-1, a bat MERS-like coronavirus that circulates in pangolins and is infectious and pathogenic in human airways and intestinal organs, as well as in human DPP4-transgenic mice [4]. These findings suggest that HR1LS/CF501 vaccine, if developed, may also be used to combat an outbreak caused by any newly emerged coronavirus, such as SARS-CoV-3 or MERS-CoV-2, which has been speculated to evolve from a zoonotic coronavirus presented in an intermediate host animal or a recombinant virus through genetic recombination between SARS-CoV-2 and MERS-CoV [3].

In summary, HR1LS as a protein inhibitor inhibits infection of the pseudotyped SARS-CoV-2 and its variants by interacting with HR2 region of viral S protein, while HR1LS as an immunogen induces neutralizing antibodies targeting HR1, CH, or SH region of viral S protein against SASR-CoV-2 and its variants. Although HR1LS protein and HR1LS-induced antibodies work on different targets in SASR-CoV-2 S protein, they have a similar mechanism of action, i.e. blocking SARS-CoV-2 fusion with and entry into the host cell for replication. The sequences of neutralizing epitopes in S2 subunit are relatively more conserved than those in S1 subunit of HCoVs [14], but the neutralizing immunogenicity of the neutralizing epitopes in S2 subunit are generally weaker than those in S1 subunit because the neutralizing epitopes in the S2 subunit are not exposed in the native conformation of S protein, but are, instead, instantaneously exposed at the fusion-intermediate stage. In contrast, neutralizing epitopes in the S1 subunit, particularly those in the receptor-binding motif (RBM) in RBD, are well exposed in the native conformation of S protein [43]. This, in turn, explains why the titres of nAbs induced by vaccines containing neutralizing epitopes in S2 subunit are generally lower than those induced by vaccines containing neutralizing epitopes in the S1 subunit [16,22,42,43]. Fortunately, the immunogenicity of a recombinant subunit vaccine can be significantly increased by using a potent adjuvant, like CF501, a STING agonist-based adjuvant, that was used in study. We have previously shown that CF501 could potently enhance the immunogenicity of RBD-Fc vaccine to elicit strong nAb and T cell responses in vaccinated animals [26–28]. In this study, we found that the NT₅₀ of sera of mice immunized with HR1LS/CF501 was about 3.4-fold higher than that of sera from the HR1LS/Alum group. We strongly believe that if we can use a stronger adjuvant formulated with HR1LS at proper ratio in a better delivery system, the level of nAbs elicited by this vaccine candidate can be further increased, even to the level of nAbs elicited by the RBD-based vaccines, thus highlighting the benefit of developing a pan-sarbecovirus, pan-β-coronavirus, or pan-HCoV vaccine rather than a SARS-CoV-2-specific vaccine.

Nevertheless, some limitations of the current study should be noted. First, although HR1LS-trimer-elicited immune responses were evaluated in vaccinated animals, the protective effect in vivo against challenge with highly pathogenic HCoVs was not assessed. Therefore, these experiments should be performed in the near future. Second, a previous study has shown that S2 nanoparticle vaccines elicit efficient ADCC activity in mice [17]. However, the present study did not undertake experiments to determine whether HR1- or HR1LS-specific antibodies could or could not mediate ADCC. Third, mRNA-, DNA-, and virus vector-based vaccines expressing HR1LS were not tested, suggesting that further studies on these kinds of vaccines should proceed.

In conclusion, a recombinant trimeric protein, HR1LS, has been designed and engineered. It contains HR1, CH, and SH regions in the S2 subunit of SARS-CoV-2 S protein and exhibits potent pan-CoV inhibitory activity and effective immunogenicity to elicit broad-spectrum nAb responses against infection by SARS-CoV-2 and its variants, as well as other HCoVs tested. These results suggest good potential for further development as pan-HCoV inhibitor-based therapeutics or prophylactics. HR1LS is also a promising candidate for development as a pan-HCoV vaccine for treatment and/or prevention of infection by SARS-CoV-2 and its variants responsible for the current of COVID-19 pandemic and by other emerging or reemerging HCoVs potentially causing future outbreaks of HCoV infectious diseases.

Funding Statement

This work was supported by grants from the National Key R&D Program of China (2022YFC2604102 and 2021YFC2300703 to L.L.), National Natural Science Foundation of China (grant numbers 92169112 to S.J.; 82202491 to X.W.), Shanghai Municipal Science and Technology Major Project (ZD2021CY001 to S.J, L.L and S.X.), and Program of Shanghai Academic/Technology Research (20XD1420300 to L.L.).

Disclosure statement

No potential conflict of interest was reported by the author(s).