The Neutralizing Monoclonal Antibodies against SFTS Group Bandaviruses Suggest New Targets of Specific or Broad-Spectrum Antivirals

ABSTRACT.

Severe fever with thrombocytopenia syndrome virus (SFTSV), Heartland virus (HRTV) and Guertu virus (GTV) belong to the severe fever with thrombocytopenia syndrome/Heartland group of genus Bandavirus in the family Phenuiviridae of order Bunyavirales. Severe fever with thrombocytopenia syndrome virus and HRTV, identified from ticks from Asia and America, respectively, are important pathogens causing severe febrile diseases in humans. Guertu virus, closely related to these two viruses, is a potential pathogen, but no confirmed infection has been identified. So far, human-derived neutralizing monoclonal antibodies (mAbs) against SFTSV have been identified as having a great potential to be developed as antivirals; however, there is still a lack of neutralizing mAbs to GTV and HRTV. In this study, five neutralizing the mAbs against GTV and HRTV were obtained by hybridoma screening technology, four of which (14B4, 14D8, and 20D4 derived from GTV, and 27C8 derived from HRTV) showed cross reactivity and neutralization to all three viruses, and one derived from HRTV (10D6) neutralized HRTV specifically. The possible mechanisms of mAbs cross neutralization among the three viruses are discussed by analyzing their glycoprotein (GP) sequences and structures. Generating these neutralizing mAbs provides important antiviral candidates against GTV, HRTV, and SFTSV despite their differential activities, and their protective effect could be further evaluated in virus-infected mice. Their differential neutralizing efficiency and specificity further suggested that the three viruses share common mechanisms on the basis of GP functioning, and that HRTV poses a unique mechanism that differs from the other viruses. These findings shed light on developing broad-spectrum antiviral strategies against bandaviruses and promoting an understanding of the bandavirus infection process.

INTRODUCTION

Emerging pathogenic tick-borne viruses (TBVs) that can infect animals and humans have attracted considerable attention, as evidenced by the increasing incidence of tick-borne viral diseases and their significant impacts on the human health-care system.¹ Attributed to the rapid development of metagenomics, novel arboviruses have been identified and isolated from various hosts,² indicating increasing potential threats from emerging arbovirus spillover to different hosts. In recent years, the genus Bandavirus in family Phenuiviridae of Bunyavirales were found containing TBVs associated with human diseases, including severe fever with thrombocytopenia syndrome (SFTS)/Heartland group and Bhanja group. The Bhanja group is comprised of the Bhanja virus, the Palma virus, and the Lone Star virus. The SFTS/Heartland group is comprised of the Severe fever with thrombocytopenia syndrome virus (SFTSV), the Guertu virus (GTV), the Heartland virus (HRTV), and the Hunter Island group virus.³ Severe fever with thrombocytopenia syndrome virus was first isolated from serum samples of patients with SFTS in 2011 in China,⁴ which affected more than 23 provinces in China and resulted in a high mortality rate of up to 30%.⁵ Later, SFTSV was reported in Japan and South Korea.⁶ Serological evidence suggests that SFTSV could be widely distributed in Asian countries, including Vietnam and Pakistan.^7,8 Subsequently, in United States, HRTV was identified in patients with clinical symptoms similar to those observed for SFTS.⁹ To date, there have been more than 50 cases of HRTV, including three deaths.¹⁰ Guertu virus was isolated from Dermacentor nuttalli ticks in Xinjiang, China. Although there have not been any reports of human cases with GTV infection, it is suggested that GTV is a potential pathogen of human diseases according to the results of a serological survey among humans and its close phylogenetic relation with SFTSV and HRTV.¹ Hunter Island group virus was isolated from ticks carried by the albatross in Australia, and there was no evidence suggesting it was associated with human disease.^11,12 So far, there is still a lack of approved, specific antiviral drugs against the viruses belonging to the SFTS/Heartland group.

Like other typical bunyaviruses, the genome of bandaviruses contains three RNA segments, which are designated as L, M, and S segments, according to their length.¹³ The L segment encodes the RNA-dependent RNA polymerase; M segment encodes the glycoprotein (GP), which can be cleaved into the N-terminal fragment (Gn) and the C-terminal fragment (Gc); and the S segment contains an ambisense RNA encoding the nucleocapsid (NP) and nonstructural proteins. Because the bandavirus Gn and Gc are responsible for viral entry into host cells, they play an important role in the infection process and are considered the major target for developing antiviral antibodies and/or small molecules to prevent viral infection. Compared with small molecules, monoclonal antibodies (mAbs) recognize specific protein targets and have been widely used to treat cancers.¹⁴ The mAbs developed against viral GPs may pose neutralizing activity to prevent viral infection and could be developed into specific antiviral drugs.¹⁵

In this study, three murine mAbs against GTV and two mAbs against HRTV were generated using the hybridoma technique, which showed neutralizing activity to prevent viral infection. The cross reaction and neutralization of these mAbs to bandaviruses, including SFTSV, GTV, and HRTV, were evaluated and identified mAbs having broad-spectrum activity against all three viruses, and a mAb with specific activity to one of them (HRTV). The molecular mechanisms of mAb cross reaction with related viruses are discussed based on GP sequence alignments and prediction of three-dimensional structures. The finding of broad-spectrum mAbs sheds light on the common mechanisms of SFTSV, GTV, and HRTV infection, and suggests they share similar antigenicity of GPs, whereas the specific neutralizing mAb against HRTV suggests a manner of infection different from SFTSV and GTV. These neutralizing mAbs provide fundamental materials for investigating viral infection mechanisms further, and promote the development of broad-spectrum and specific antiviral strategies as for bandaviruses.

MATERIALS AND METHODS

Cells and viruses.

The Vero (American Type Culture Collection [ATCC] no. CCL-81), HeLa (ATCC no. CCL-2), Sf9 (ATCC no. CRL-1711), and mouse myeloma (ATCC no. PTA-9396) cell lines were obtained from ATCC (Manassas, VA). Vero cells were cultured by using high-glucose Dulbecco’s Modified Eagle Medium (NZK Biotech, Wuhan, China) containing 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY), Sf9 cells were cultured by using insect cell culture medium (NZK Biotech), and HeLa cells were cultured high-glucose Modified Eagle Medium (NZK Biotech) with 10% FBS. Severe fever with thrombocytopenia syndrome virus strain HBGS13 (GenBank accession no. KY440774), GTV strain DXM (GenBank accession no. KT328591), and HRTV strain patient 1 (GenBank accession no. NC024496) were preserved at the National Virus Resource Center in Wuhan, China. Infection assays using GTV, SFTSV, and HRTV were conducted in a biosafety level-2 laboratory.

Plasmids and antibodies.

The genes of GTV NP and HRTV NP were amplified from viral complementary DNA by polymerase chain reaction and cloned into the plasmid containing a beta-chicken actin promoter and a CMV-enhancer (pCAGGS), and fused with the 6×His tag to generate the recombinant plasmids pCAGGS-GTV NP and pCAGGS-HRTV NP, respectively, which were verified by Sanger sequencing. The rabbit polyclonal antibodies (pAbs) against GTV or HRTV Gn (α-GTV-Gn or α-HRTV-Gn) were prepared previously and used as the primary antibody to verify expression of GTV GP or HRTV GP (1:2,000 dilution). The anti-His antibody (α-His) (abcam, Shanghai, China) and the β-actin monoclonal antibody (α-β-actin) (ABclonal, Wuhan, China) were used to verify protein expression and served as the internal control. The pAb specific to Autographa caniformica multiple nucleopolyhedrovirus GP64 (α-vAc-GP64) and the rabbit polyclonal antibody against SFTSV NP were generated as described previously.^16,17 The horseradish peroxidase (HRP)-conjugated affinipure goat anti-rabbit/mouse IgG (H+L) (Proteintech, Wuhan, China) and the goat anti-rabbit/mouse IgG H&L (Alexa Fluor™ 488; abcam, Wuhan, China) were used as the secondary antibodies according to manufacturers’ instructions. Serum samples collected from healthy C57/BL6 mice were diluted and used as negative controls.

Production of neutralizing mAbs against GTV and HRTV.

Vero cells were infected by GTV and HRTV at a multiplicity of infection (MOI) of one, and supernatants were harvested on 3 days postinfection. The virus titer was determined using Vero cells by end point dilution assay and was calculated using the Reed-Muench method. Viruses in supernatants were inactivated with propiolactone (1:4,000 v/v) (SERVA Electrophoresis GmbH, Heidelberg, Germany), and then centrifuged and filtered to remove cell debris as described previously.¹⁸ Virions were purified by ultracentrifuge with a 10% sucrose cushion (4°C, 28,000 rpm, 3 hours) and resuspended in phosphate buffered saline (PBS).

Six BALB/c mice (6–8 weeks old) were immunized three times with 10⁸ median tissue culture infective dose (TCID₅₀) inactivated GTV or HRTV by both the intraperitoneal and subcutaneous routes, 2 weeks apart. On day 10 after the second immunization, serum samples were collected from each of the mice, and their antibody response to GTV or HRTV was evaluated primarily by indirect immunofluorescence assay (IFA). Mice that had an antibody reaction to GTV or HRTV were selected and spleen tissue was harvested. Hybridomas were prepared by fusion of SP2/0 myeloma cells with the mouse spleen cells using polyethylene glycol (GlpBio, Montclair, CA) as described previously.¹⁹ The hybridoma cells were grown in a selective medium of hypoxanthine aminopterin–thymidine (Sigma-Aldrich, St. Louis, MO) for 10 days and then assessed for mAbs using IFAs and microneutralization tests (MNTs). Cell clones containing antibody-positive hybridoma cells were generated by limiting dilution. Briefly, hybridoma cells were maintained in a medium of hypoxanthine–thymidine (Sigma-Aldrich) with 10% FBS for 10 days, then were diluted into 96-well microplates (approximately 1 cell/well) and maintained for another 10 days. Supernatants were collected and examined for an antibody response to GTV or HRTV by IFAs and MNTs. After four rounds of screening, antibody-positive cell clones were transferred to T-25 flasks and propagated further. Furthermore, cell clones with better growth and greater efficiency of neutralization were selected and infected intraperitoneally into mice to produce aseptic fluid. The mAbs were purified using protein G column chromatography (Agilent, Santa Clara, CA) according to the manufacturer’ s instructions.

Recombinant baculoviruses.

Full-length genes of GTV GP and HRTV GP and fragments of their Gn and Gc were amplified and inserted into the plasmid (pFastBac™ Dual; Invitrogen, Carlsbad, CA) under control of the polyhedrin promoter and fused with a 6× His tag. The Bac-to-Bac baculovirus expression system (Invitrogen, Montclair, NJ) was used to construct recombinant baculoviruses vAc-GTV-GP, vAc-GTV-Gn, vAc-GTV-Gc, vAc-HRTV-GP, vAc-HRTV-Gn, and vAc-HRTV-Gc, according to instructions. Protein expression in the baculovirus-infected Sf9 cells was validated by Western blot and IFAs using respective antibodies or the anti-His mAb.

Indirect immunofluorescence assay.

Vero cells infected with GTV or HRTV, or transfected with the GTV NP– or HRTV NP–expressing plasmids, and Sf9 cells infected with recombinant baculoviruses were fixed using PBS containing 4% paraformaldehyde and penetrated by 0.2% Triton X-100. After blocking by PBS containing 5% bovine serum albumin at 37°C for 2 hours, the cells were inoculated with α-His, mice serum, pAbs, or mAbs in serial dilutions as the primary antibody and Alexa Fluor mouse/rabbit 488 as the secondary antibody. Cell nuclei were stained using Hoechst 33258 (Beyotime, Shanghai, China). Antibody response levels were determined using the Operetta CLS High Content Analysis System (PerkinElmer, Norwalk, CT), which evaluated the fluorescence intensity of cells.

Western blot assays.

Vero cells were infected with GTV or HRTV and its concentrated cell supernatant, and Sf9 cells infected with recombinant baculoviruses were collected by 1× loading buffer to prepare for 12% sodium dodecyl-sulfate polyacrylamide gel electrophoresis. Subsequently, they were electro-transferred onto 0.2-μM polyvinylidene fluoride (PVDF) membrane for Western blotting. After blocking using PBS containing 5% skimmed milk at 37°C for 2 hours, the PVDF membranes were inoculated with α-His, α-vAc-GP64, α-β-actin, pAbs, or mAbs as the primary antibody and the goat anti-rabbit/mouse IgG H&L conjugated with HRP as the secondary antibody. The PVDF membranes were washed with Tris-buffered saline–Tween 20, and the blotted bands on it were developed by using an enhanced chemiluminescence color reagent. A hypersensitive chemiluminescence imager (Azure c500; Azure Biosystems, Dublin, CA) was used to visualize immunoblots.

Microneutralization tests.

To identify antibody-positive hybridoma cell clones, GTV or HRTV (50 TCID₅₀ per test) was mixed and incubated with an equivalent volume of culture supernatants from hybridoma cells at 37°C for 2 hour. The Vero cells were then incubated with the virus–antibody mixtures and maintained at 37°C for another 96 hours. Infection was detected by IFA using the α-SFTSV NP polyclonal antibody because it showed efficient cross reaction with both GTV and HRTV NPs.¹

To determine the neutralizing efficiency of mAbs, GTV, HRTV, or SFTSV (100 TCID₅₀ per test) was incubated with equivalent volumes of purified mAbs, which were serially diluted 2-fold from 100 µg/mL to 0.0125 µg/mL. The mixtures were then incubated with Vero cells, and infection was inspected via IFA. Each test was performed in triplicate. Antibody neutralizing activities were determined using the Operetta CLS High Content Analysis System (PerkinElmer), and the number of cells with green fluorescence and cells stained by Hoechst 33258 was obtained.

Blocking ELISA.

First, IFAs were performed as described earlier to quantify viral infection rates of GTV or HRTV (MOI = 0.01 TCID₅₀/cell) in Vero cells at 24 and 48 hours postinfection using the Operetta CLS High Content Analysis System (PerkinElmer). When infection rates of both viruses in Vero cells were less than 10%, the cells were fixed, blocked, and used for blocking ELISA to ensure saturation of the blocking mAbs relative to the antigens. The infected cells were then incubated with 14B4, 20D4, or 27C8 (100 µg/mL) for 2 hours at 37°C, followed by washing with PBS with Tween 20 to remove any unbound antibodies. The cells were incubated with HRP-conjugated 14B4 or 27C8 (14B4-HRP or 27C8-HRP, 100 µg/mL) for an additional 2 hours at 37°C. After washing, TMB chromogen solution (Beyotime, Shanghai, China) was added and incubated at 37°C for 10 minutes, and sulfuric acid–free stop solution for TMB substrate (450 nm) was added to stop the reaction. The optical density (OD) value (OD_450-630) was measured using dual-wavelength detection at 450 and 630 nm. Each assay was performed in triplicate.

Bioinformatic and statistical analyses.

To analyze the sequence similarity of proteins, the Gn and Gc amino acid sequences of GTV, SFTSV, and HRTV were aligned using Megalign 7.1.0 and presented with ESPript 3.0. The tertiary structures of HRTV Gn, and GTV Gn and Gc were predicted using phyre² 2.0 and compared with the crystal structures of SFTSV Gn (5Y10),²⁰ SFTSV Gc (5G47),²¹ and HRTV Gc (5YOW)²² deposited in the RCSB Protein Data Bank. Similarities between these proteins and domains were analyzed with PyMOL™ software (version 2.7).

Half maximal inhibitory concentration (IC₅₀) values and 50% of maximal concentration (EC₅₀) values were calculated with nonlinear regression using GraphPad Prism 8.0 software. Significant differences in IC₅₀ values of mAb cross-neutralizing viruses, in the EC₅₀ values of mAb cross-recognizing viral GPs, and OD_450/630 values (OD₄₅₀ minus OD₆₃₀) in the blocking assays were analyzed using Student’s t-test.

RESULTS

Three GTV-derived and two HRTV-derived neutralizing mAbs were obtained.

As shown by IFA, serum samples from all three mice (mouse 1, mouse 2, and mouse 3) inoculated with inactivated GTV showed an antibody response to GTV-infected cells, whereas only two (mouse 1 and mouse 2) of the three mice inoculated with HRTV had an antibody response to HRTV-infected cells (Figure 1A). Subsequently, after the third inoculation, spleen cells from the mice (labeled GTV-1#, GTV-3#, HRTV-1#, and HRTV-2#) were harvested for preparing hybridoma cells. After four rounds of screening, supernatants from 16 clones of GTV-derived hybridoma cells and 21 clones of HRTV-derived hybridoma cells were identified that contained neutralizing antibodies of different efficiencies (36.73–91.97% for GTV, and 15.27–95.31% for HRTV) to neutralize the respective virus (Supplemental Table 1). Of these cell clones, three containing GTV-derived mAbs (14B4, 14D8, and 20D4) and two containing HRTV-derived mAbs (10D6 and 27C8) were selected for further examination. The GTV-derived mAbs and the HRTV-derived mAbs showed an efficient response to GTV- and HRTV-infected cells, respectively (Figure 1B). However, when using infected cells or purified viruses as antigens during Western blotting, we failed to detect any band by using these mAbs (Supplemental Figure 1A and B). This suggests that the mAbs recognize the spatial structures of the viral proteins rather than the linearized epitopes of viral proteins.

Figure 1.

Screening of monoclonal antibodies against Guertu virus (GTV) and Heartland virus (HRTV). Antibody reaction to GTV or HRTV antigens in serum samples from virus-inoculated mice (A) and supernatants from hybridoma cells (B) were examined by indirect immunofluorescent assay. The Vero cells infected with GTV or HRTV were blotted with α-severe fever with thrombocytopenia syndrome virus-nucleocapsid as the positive control (PC), and the healthy cells were the negative control (NC). Bar = 200 nm.

The mAbs had a reaction with eukaryotic expression of GPs.

To identify which viral proteins recognized by the mAbs, recombinant baculoviruses vAc-GTV-GP, vAc-GTV-Gn, vAc-GTV-Gc, vAc-HRTV-GP, vAc-HRTV-Gn, and vAc-HRTV-Gc were generated (Figure 2A). Protein expression in baculovirus-infected Sf9 cells was confirmed by both Western blot and IFA (Figure 2B and C). As shown by the IFA results, three GTV mAbs reacted with the vAc-GTV-GP antigen, but not those infected with vAc-GTV-Gn or vAc-GTV-Gc, nor the cells transfected with the NP-expressing plasmid (pCAGGS-GTV-NP) (Figure 2D). This suggests the three neutralizing mAbs only recognize the full-length GTV GP instead of Gn, Gc, or NP. Similarly, the two HRTV neutralizing mAbs only recognize HRTV GP and not Gn, Gc, or NP (Figure 2D). However, we failed to detect any reaction of these mAbs to GPs of GTV or HRTV by Western blot (data not shown). This suggests that these mAbs recognize GTV or HRTV GPs by the spatial structures rather than linear epitopes.

Figure 2.

Identification of monoclonal antibody interaction with viral proteins by using a recombinant baculovirus expression system. (A) A schematic diagram showing the strategy for constructing recombinant baculoviruses expressing the full length or fragments of Guertu virus (GTV) and Heartland virus (HRTV) glycoproteins (GPs). (B, C) Western blot analysis (B) and indirect immunofluorescent assay analysis (C) of GPs and the N-terminal fragment (Gn) and C-terminal fragment (Gc) proteins in recombinant baculovirus-infected Sf9 cells. The expression of the Gn and Gc proteins was detected with α-His; GPs were detected with polyclonal antibodies specific for the Gn protein. The protein GP64 was detected as an internal control for baculovirus infection. (D) Validation of antigen recognition by monoclonal antibodies using baculovirus-infected Sf9 cells as the antigen. Bar = 200 nm.

Cross-neutralizing activity to related viruses were identified from GTV- and HRTV-derived mAbs.

Because SFTSV, GTV, and HRTV are phylogenetically related viruses,¹ which may have serological correlations, the cross-neutralizing activity of GTV-derived mAbs to SFTSV and HRTV, and of HRTV-derived mAbs to SFTSV and GTV were evaluated. As expected, the GTV-derived mAbs showed a greater neutralizing efficiency to GTV than to SFTSV and HRTV. Taking 14B4, for example, it had a high IC₅₀ value of 3.389 µg/mL to HRTV—more than two times greater than the IC₅₀ value of 1.281 µg/mL to SFTSV and significantly greater than the IC₅₀ value of 0.201 µg/mL to GTV (P < 0.05) (Table 1). Of the three mAbs, 14B4 showed neutralizing activity to three viruses, which is comparable to 14D8, as they derived from the same cell clone in the previous cell dilution. 14B4 and 14D8 are more efficient in neutralizing all three viruses than 20D4, as suggested by their lower IC₅₀ values (than 20D4) and the delayed reaction of 20D4 to viruses along with increasing mAb concentrations (Figure 3A). Although the HRTV-derived mAb 27C8 showed a greater neutralizing efficiency to HRTV than to GTV and SFTSV, the mAb 10D6 showed neutralization only to HRTV. The mAbs 27C8 was more efficient than 10D6 in preventing HRTV infection, as shown by the greater IC₅₀ value of 10D6 than 27C8, and the delayed reaction of 10D6 to HRTV along with increasing concentration (Table 1, Figure 3A).

Table 1.

IC₅₀ values of mAbs cross-neutralizing GTV, SFTSV, and HRTV

Virus	GTV-derived mAbs IC50, μg/mL (95% CI)			HRTV-derived mAbs IC50, μg/mL (95% CI)
	14B4	14D8	20D4	10D6	27C8
GTV	0.201 (0.024–0.099)	0.566 (0.199–0.568)	4.009 (2.743–5.787)	NR	4.430 (2.808–6.855)
SFTSV	1.281 (0.895–1.845)	1.168 (0.785–1.753)	5.888 (4.116–8.247)	NR	7.542 (5.346–10.81)
HRTV	3.389 (2.236–5.235)*	3.629 (2.378–5.629)	7.386 (8.116–28.65)	8.386 (6.398–11.08)	0.791 (0.492–1.267)

GTV = Guertu virus; HRTV = Heartland virus; IC₅₀ = half maximal inhibitory concentration; mAbs = monoclonal antibodies; NR = negative reaction detected; SFTSV = severe fever with thrombocytopenia syndrome virus.

^*P < 0.05. A significant difference was determined by comparing the IC₅₀ values of neutralizing HRTV to those of neutralizing GTV using Student’s t-test.

Figure 3.

Efficiencies of monoclonal antibodies (mAbs) for cross neutralization and reaction to related bandaviruses. Three Guerto virus (GTV)–derived mAbs and two Heartland virus (HRTV)–derived mAbs in serial dilutions were tested for their efficiency in neutralizing severe fever with thrombocytopenia syndrome virus (SFTSV), GTV, and HRTV, respectively (A); and in recognizing SFTSV, GTV, and HRTV antigens, respectively (B). The inhibition and recognition efficiencies were analyzed with nonlinear regression using GraphPad Prism 8.0 software. Fifty percent inhibition is indicated by the dashed line in (A).

The efficiency of the five mAbs in cross reaction with SFTSV, GTV, and HRTV GPs was characterized further using Sf9 cells infected with the recombinant baculoviruses vAc-SFTSV-GP, vAc-GTV-GP, and vAc-HRTV-GP (Supplemental Figure 2). The GTV-derived mAbs 14B4 and 14D8 showed comparable recognition abilities to GTV, SFTSV, and HRTV, as each of them had comparable EC₅₀ values and similar reaction curves to the three viruses, along with the dilutions, without any significant difference (Table 2, Figure 3B). These two mAbs were much more efficient than 20D4, as 20D4 showed a delayed reaction curve to viruses along with increasing mAb concentrations, and its EC₅₀ values to three viruses were not applicable (Table 2, Figure 3B). Similar to the neutralization activity, mAb 27C8 derived from HRTV showed more efficient recognition to HRTV than to GTV and SFTSV, as the EC₅₀ values for SFTSV and GTV were almost two times that of HRTV (Table 2). As expected, mAb 10D6 only reacts with HRTV and not with SFTSV and GTV (Table 2, Figure 3B).

Table 2.

EC₅₀ values of mAbs cross-neutralizing GTV, SFTSV, and HRTV

Virus	GTV-derived mAbs EC50, μg/mL (95% CI)			HRTV-derived mAbs EC50, μg/mL (95% CI)
	14B4	14D8	20D4	10D6	27C8
GTV	1.477 (1.299–1.680)	1.191 (0.967–1.467)	N/A	NR	14.710 (12.580–17.590)
SFTSV	1.709 (1.381–2.127)	1.358 (1.080–1.723)	N/A	NR	15.580 (14.350–17.000)
HRTV	2.087 (1.749–2.501)	1.860 (1.284–2.770)	N/A	15.80 (14.110–17.910)	8.873 (7.553–10.530)

EC₅₀ = fifty percent of maximal concentration; GTV = Guertu virus; HRTV = Heartland virus; mAbs = monoclonal antibodies; N/A = not applicable; NR = negative reaction detected; SFTSV = severe fever with thrombocytopenia syndrome virus.

Therefore, for the mAbs derived from GTV, it is very likely that mAbs 14B4 and 14D8 recognize the same epitope, which is different from that recognized by 20D4, and that the epitopes are conserved in GPs of SFTSV, GTV, and HRTV. As for the mAbs derived from HRTV, 10D6 recognized an epitope specifically present in the HRTV GP, but not in SFTSV and GTV. This epitope is completely distinct from the one recognized by 27C8, which could be conserved among GPs of the three viruses.

To understand the mechanism more completely that underlies the efficient cross reaction of mAbs among the three viruses, pairwise comparison of sequence identities of GPs from SFTSV, GTV, and HRTV showed that GTV Gn and Gc share greater identities with SFTSV than HRTV. Heartland virus Gn had a lower identity with SFTSV than GTV, whereas its Gc showed the same identity as SFTSV and GTV (Supplemental Table 2, Supplemental Figure 3). Furthermore, GTV Gn and Gc, and HRTV Gn and Gc were predicted according to the division of structural domains based on the predicted structures of GTV and HRTV GPs according to SFTSV GP (data not shown) and were compared with SFTSV Gn and Gc (Supplemental Figure 4). The root-mean-square deviation (RMSD) values, which indicate the degree of atomic deviation from the ratio position, were compared between the full-length GP structures and each of the three domains of Gn and Gc from the three viruses. Although the Gn and Gc proteins seem to have similar structures among the three viruses, HRTV is more different from the other two viruses, as its Gn and Gc had greater RMSD values then SFTSV and GTV compared to the others (Supplemental Table 3). Therefore, these results may explain that the GTV-derived mAbs were more effective in neutralizing and recognizing SFTSV than HRTV. The sequence and structure divergence of HRTV from SFTSV more than GTV also explained that the HRTV-derived 27C8 was less effective for SFTSV than GTV.

The GTV-derived mAbs 14B4 and 20D4 recognized epitopes with spatial overlap in bandavirus GPs.

To understand the correlation of epitopes recognized by these mAbs with cross-neutralization activity to bandaviruses, antibody blocking assays were performed using 14B4, 20D4, and 27C8. 14D8 was not tested because it has the same origin as 14B4. The IFA results showed that the rate of GTV infection in Vero cells at 24 hours postinfection (5.10 ± 0.58%) was as comparable as that of HRTV infection in Vero cells at 48 hours postinfection (4.63 ± 0.56%) (data not shown). Therefore, the GTV-infected Vero cells at 24 hours postinfection and the HRTV-infected Vero cells at 48 hours postinfection were fixed for subsequent blocking ELISA.

As a result, compared with the OD values of cells incubated with 14B4-HRP alone, the incubation of 20D4 prior to 14B4-HRP decreased the tested OD values significantly of both the GTV- and HRTV-infected cells, whereas incubation with 27C8 affected 14B4 binding to antigens (Figure 4A). This result suggests that 14B4 binding to GTV and HRTV antigens was blocked by 20D4 rather than 27C8. In contrast, incubation with 14B4 or 20D4 did not result in a significant decrease in 27C8 recognition of GTV- and HRTV-infected cells (Figure 4B). This suggests that the GTV-derived mAbs 14B4 and 20D4 may recognize epitopes with spatial overlap in the GPs of GTV and HRTV, whereas HRTV-derived 27C8 recognizes an epitope in GTV and HRTV that is different from 14B4 and 20D4.

Figure 4.

Competitive binding assays of Guertu virus (GTV)– and Heartland virus (HRTV)–derived monoclonal antibodies. (A, B) Blocking ELISA was performed to determine the competitive effect of 14B4 with 20D4 and 27C8 (A), and of 27C8 with 14B4 and 20D4 (B) in recognizing GTV and HRTV antigens. The GTV- or HRTV-infected cells were first incubated with 20D4, 27C8, or medium, and later with 14B4–horseradish peroxidase (HRP); or were first incubated with 14B4, 20D4, or medium, and later with 27C8-HRP. The optical density (OD) at 450 and 630 nm was measured, and each assay was performed in triplicate. ns = nonsignificant at P > 0.05; *** P ≤ 0.001; **** P ≤ 0.0001.

DISCUSSION

Monoclonal antibodies have been used as an efficient tool to inhibit virus infection by interacting with viral GPs to block virus attachment to host cells. Because of the medical importance of SFTSV, it is important to generate mAbs against it that are of significance to be developed as antiviral candidates. So far, two neutralizing mAbs against SFTSV have been derived from patients with SFTS and identified. Monoclonal antibody Ab4-5 can react with a surface-exposed epitope on the N-terminal helix α6 in domain III of SFTSV Gn, resulting in an IC₅₀ value of 2.0 µg/mL.^23,24 The other mAb, Ab10, was reactive to SFTSV Gn, which can prevent virus infection in host cells and protect 80% of C57BL/6 IFNAR^–/– (A129) mice from SFTSV infection.²⁵ However, the cross reaction and neutralization activity of the two human-derived mAbs to GTV and HRTV were not addressed. Recently, a murine derived mAb, 40C10, against SFTSV was generated, which poses efficient neutralizing activity to SFTSV as well as to GTV and HRTV, and it could effectively protect the type I interferon receptor-deficient mice from SFTSV infection.²⁶ However, there was a lack of neutralizing mAbs against GTV and HRTV before this study.

In our study, we obtained three neutralizing mAbs against GTV and two against HRTV from hybridoma cells, which were suggested to recognize spatial structures of GTV and HRTV GPs other than Gn or Gc separately. These neutralizing mAbs may provide important materials for developing antiviral candidate drugs against GTV and HRTV. We evaluated the efficiency of mAbs cross-neutralizing and cross-recognizing SFTSV, GTV, and HRTV. All three viruses could be cross-neutralized and -recognized by the three GTV-derived mAbs (14B4, 14D8, 20D4) and one HRTV-derived mAb (27C8). In contrast to these mAbs, the other HRTV-derived mAb 10D6 was only effective for HRTV. The comparison of GP sequences and structures of SFTSV, GTV, and HRTV revealed a close relationship between SFTSV and GTV more than HRTV, which is consistent with their revolutionary relationship that indicates that HRTV is supposed to diverge earlier than SFTSV and GTV.²⁷ Our results of mAbs cross-neutralization and -recognition supported the fact that HRTV has viral properties that differ from SFTSV and GTV. Because the GTV-derived mAbs showed comparable recognizing activities to GPs of SFTSV and GTV, and HRTV, and despite the fact that their neutralizing activity to HRTV was slightly less efficient than that of SFTSV and GTV, we speculate that the spatial epitopes recognized by the GTV-derived mAbs are conserved among the three viruses (Figure 5). Moreover, blocking ELISA showed that GTV-derived 14B4 and 20D4 recognize epitopes with spatial overlap in GTV and HRTV GPs (Figure 5). The HRTV-derived 27C8 showed greater neutralization activity and reaction to HRTV than to SFTSV and GTV; 10D6 only showed a reaction to HRTV. These results further suggest that HRTV GP has epitopes that are different from SFTSV and GTV, and the epitope reacting with 10D6 is distinct from SFTSV and GTV (Figure 5). Because these mAbs only react with spatial structures of GPs, by blocking the domains important for virus entry,²⁸ our results indicated that the domains responsible for virus entry that are constituted by Gn and Gc of SFTSV, GTV, or HRTV. Moreover, SFTSV and GTV may share common mechanisms of virus entry into host cells, whereas HRTV may adopt entry mechanism a bit different from the others.

Figure 5.

Proposed patterns of monoclonal antibodies (mAbs) interacting with Guertu virus (GTV) and severe fever with thrombocytopenia syndrome virus (SFTSV) (A), and Heartland virus (HRTV) (B) glycoproteins (GPs). Epitopes conserved in GTV, SFTSV, and HRTV GPs recognized by the mAbs with cross-neutralizing activity, including 20D4, 14B4, 14D8, and 27C8, are indicated by dashed orange ovals; specific epitope in HRTV GP recognized by 10D6 is indicated by the dashed blue oval.

Previous studies^29,30 have revealed that SFTSV can be divided into Chinese clades and Japanese clades. Although the clades were designated by alphabets, Roman numerals, or Cn and Jn (where C represents Chinese and J represents Japan, and n represents the clade number) in different studies,^29–32 these data showed that the division of the clades was associated with the geographic distribution of SFTSV. The SFTSV variants belonging to different genotypes exhibited different antigenicity and had different efficiency when reacting with neutralizing antibodies from other genotypes.²⁹ Our study investigated the mAbs cross-neutralizing SFTSV by strain HBGS13, which is a representative strain of the genotype found most commonly in mainland China.^29,30,33 The cross-neutralization to SFTSV variants of other genotypes was not characterized in our study, which might differ from the efficiency of the strain we tested. Nevertheless, our results still suggested that the mAbs generated in our study had effective cross-neutralizing activity against SFTSV, as the strain tested could represent the most common strains of SFTSV.

In summary, this study generated five neutralizing mAbs, three against GTV and two against HRTV. Except for one mAb that reacts only with HRTV, the other four mAbs showed cross-neutralizing activity to all three closely related viruses: SFTSV, GTV, and HRTV. All these mAbs recognize spatial structures of GPs. Although epitopes interacting with mAbs were not identified, by comparing the neutralization and reaction efficiency with GPs of the three viruses, more light may be shed on the mechanisms of the virus entry process. Our results suggest that SFTSV and GTV are more closely related and share similar mechanisms of virus entry, whereas HRTV is a bit different from them. The protective effects of these mAbs on experimental animals against virus infection could be further investigated, and could provide candidates for developing antiviral strategies to specific or broad-spectrum bandaviruses.

Financial Disclosure

This work was supported by the National Key Research and Development Project (Grant no. 2018YFE0200400), National Natural Science Foundation of China (Grant no. U20A20135); and the EVAg-Global Project (Grant no. 871029), Natural Science Foundation of China (Grant no. 72174159).

Note: Supplemental material appears at www.ajtmh.org.