Antigenic commonality and divergence of hemagglutinin-esterase-fusion protein among influenza D virus lineages revealed using epitope mapping

ABSTRACT

Influenza D virus (IDV) is one of the causative agents of bovine respiratory disease complex, which is the most common and economically burdensome disease affecting the cattle industry, and the need for an IDV vaccine has been proposed to enhance disease control. IDVs are classified into five genetic lineages based on the coding sequences of the hemagglutinin-esterase-fusion (HEF) protein, an envelope glycoprotein, which is the main target of protective antibodies against IDV infection. Herein, we prepared a panel of monoclonal antibodies (mAbs) against the HEF protein of viruses of various lineages to investigate the antigenic characteristics of IDVs and found that the mAbs could be largely separated into three groups. The first, second, and third groups demonstrated lineage-specific reactivity, cross-reactivity to viruses of multiple but not all lineages, and cross-reactivity to viruses of all lineages, respectively. Analyzing the escape mutant viruses from virus-neutralizing mAbs revealed that the receptor-binding region of the HEF molecule harbors virus-neutralizing epitopes that are conserved across multiple lineage viruses. In contrast, the apex region of the molecule possessed epitopes unique to each lineage virus. Furthermore, reverse genetics-generated recombinant viruses with point mutations revealed that amino acids within positions 210–214 of the HEF protein determined the antigenic specificity of each lineage virus. Taken together, this study reveals considerable antigenic variation among IDV lineages, although they are presumed to form a single serotype in terms of HEF antigenicity. Characterization of the antigenic epitope structure of HEF may contribute to selecting and creating effective vaccine viruses against IDV.

IMPORTANCE

Influenza D viruses (IDVs) are suggested to create cross-reactive single serotypes in hemagglutinin-esterase-fusion (HEF) antigenicity, as indicated by serological analyses among distinct HEF lineage viruses. This is supported by the high identities of HEF gene sequences among strains, unlike the hemagglutinin (HA) genes of the influenza A virus that exhibit HA subtypes. Herein, we analyzed HEF antigenicity using a monoclonal antibody panel prepared from several virus lineages and found the existence of lineage-conserved and lineage-specific epitopes in HEF molecules. These findings confirm the HEF commonality and divergence among IDVs and provide useful information for constructing a vaccine containing a recombinant IDV virus with an engineered HEF gene, thereby leading to broad immunogenicity.

INTRODUCTION

In 2011, influenza D virus (IDV), a member of the family Orthomyxoviridae, was first isolated from pig nasal fluid with respiratory symptoms in Oklahoma, USA (1). Epidemiological studies have identified cattle as the primary IDV host, causing respiratory symptoms in them (2–4). Cattle and pigs that are sero- or virus-positive for IDVs have been confirmed in several countries worldwide. Additionally, antibody-positive individuals have been detected in sheep (5, 6), goats (5, 6), dromedary camels (7, 8), horses (9), and humans (10). This suggests the potential widespread transmission of IDV among livestock and wild animals worldwide.

IDV has been implicated as a causative virus of the bovine respiratory disease complex (BRDC) (11–13), contributing to substantial economic losses in the cattle industry (14). Mixed vaccines against several viruses and bacteria that can cause BRDC have been administered in several countries (14, 15). Despite their usage, these vaccines have not demonstrated complete efficacy in controlling BRDC. This could be because IDV and other agents not included in the vaccines may be responsible for this disease. Metagenomic analyses of cattle have revealed IDVs, rather than known causative viruses, to be substantially involved in BRDC progression (11, 13). Therefore, controlling BRDC more effectively by developing and adding an IDV vaccine to the current vaccines may be a promising option.

The IDV genome comprises seven segments of negative-sense single-stranded RNA. Among these, the hemagglutinin-esterase-fusion (HEF) segment encodes the surface glycoprotein HEF, a key protective antigen against IDV infection. Phylogenetic analysis using HEF gene sequences categorizes IDVs into five genetic lineages: D/OK (1), D/660 (16), D/Yama2016 (17), D/Yama2019 (18), and D/CA2019 (19) (Fig. 1). Previous studies have suggested potential antigenic heterogeneity of HEFs among the viral lineages, as reported via serological tests using rabbit immune sera or field bovine sera (16, 20, 21). Herein, we dissected HEF antigenic structures via epitope mapping using monoclonal antibodies (mAbs) and a reverse genetics approach.

Fig 1.

Phylogenetic analysis of the IDV based on the nucleotide sequence of the HEF gene coding region. The phylogenetic tree was constructed using the maximum likelihood method with a bootstrap of 1,000 replicates using MEGAX. Percentages of bootstrap values are indicated at the nodes. Bar indicates an evolutionary distance of the substitutions per site. The tested viruses are enclosed in a red frame.

RESULTS

Antigenic diversity among IDVs using immune antisera

We conducted hemagglutination (HA) inhibition (HI) and viral neutralization (VN) assays to compare HEF antigenicity among IDVs using mouse immune antiserum (polyclonal antibodies) prepared for each virus of a separate lineage (OK strain of D/OK lineage, NE strain of D/660 lineage, Y16 strain of D/Yama2016 lineage, or Y19 strain of D/Yama2019 lineage) (Fig. 1) against homologous and heterologous virus strains. The recombinant virus rCA/OK that possesses the HEF of D/bovine/California/0363/2019 of the D/CA2019 lineage in the background of the OK strain was used for the tests, as we did not obtain any intact virus strain of this lineage. In both assays, all immune antisera reacted positively to both homologous and heterologous viral strains, supporting the presence of a single IDV serotype. However, their reactivities to heterologous viruses became diverse, exhibiting various lower titers than those to homologous viruses, depending on the combination of the virus strain and antiserum used (Table 1). Mouse antisera exhibited reactivities at the highest HI titers to homologous viruses, but at a maximum eight-time lower HI titers to heterologous viruses [anti-OK serum (αOK) for Y19 or rCA/OK, αNE for OK or rCA/OK, and αY16 for Y19 or rCA/OK] compared to each homologous virus. The parallel results were obtained in the VN test; however, >16-time lower titers were observed in some combinations (αNE for OK or rCA/OK and αY16 for rCA/OK). These data suggest the possibility of antigenic diversity among the IDV HEF lineages.

TABLE 1.

Reactivities of the mouse antisera against viral strains^a

Antiserum	Test	Titer to
		OK	NE	Y16	Y19	rCA/OK
αOK	HI	5,120	1,280	1,280	640	640
	VN	12,800	6,400	3,200	1,600	1,600
αNE	HI	80	640	320	160	80
	VN	100	3,200	6,400	800	200
αY16	HI	320	320	640	80	80
	VN	800	800	3,200	800	200
αY19	HI	640	160	320	640	320
	VN	800	800	1,600	1,600	800

^a Titers are obtained using HI or VN tests.

Antigenic diversity among the IDV HEFs using mAbs

As tools for analyzing HEF antigenicity in detail, we generated mouse hybridomas that secrete mAbs against purified virus. From these hybridomas, we selected HI-positive hybridoma clones. Finally, we established two hybridoma clones for OK (αOK/1F2 and αOK/4C7), five for NE (αNE/2G3, αNE/5E2, αNE/5F1, αNE/7B3, and αNE/7D7), four for Y16 (αY16/B4, αY16/G22, αY16/G27, and αY16/R36), and five for Y19 (αY19/1A8, αY19/1C10, αY19/1E12, αY19/2D11, and αY19/4D6), and obtained mouse ascites containing high mAb concentrations. We examined the reactivity of each mAb to homologous and heterologous viral strains using the HI and VN tests (Table 2). mAbs were largely divided into three groups; one essentially reacted specifically to each homologous virus (αOK/1F2, αOK/4C7, αNE/7D7, αY16/G27, αY19/1C10, and αY19/4D6), another did to multiple but not all heterologous virus strains (αNE/2G3, αNE/5E2, αNE/5F1, αNE/7B3, αY16/G22, and αY19/2D11), and the other did to all heterologous virus strains (αY16/B4, αY16/R36, αY19/1A8, and αY19/1E12). These data indicated the presence of strain-specific epitopes and cross-reactive epitopes among the strains in IDV HEFs.

TABLE 2.

Reactivities of the monoclonal antibodies against viral strains^a

mAb		Test	Titer to
			OK	NE	Y16	Y19	rCA/OK
αOK/	1F2	HI	1,280	<40	<40	<40	<40
		VN	400	<100	<100	<100	<100
	4C7	HI	2,560	40	<40	<40	<40
		VN	400	<100	<100	<100	<100
αNE/	2G3	HI	<40	<40	<40	<40	<40
		VN	400	3,200	400	25,600	<100
	5E2	HI	<40	2,560	5,120	320	40
		VN	<100	204,800	204,800	25,600	25,600
	5F1	HI	5,120	10,240	10,240	<40	2,560
		VN	25,600	204,800	102,400	800	12,800
	7B3	HI	160	160	2560	80	<40
		VN	100	800	<100	400	<100
	7D7	HI	<40	320	<40	40	<40
		VN	<100	25,600	200	<100	<100
αY16/	B4	HI	1,280	5,120	2,560	2,560	5,120
		VN	12,800	25,600	12,800	12,800	6,400
	G22	HI	640	640	20,480	<40	<40
		VN	3,200	6,400	409,600	<100	100
	G27	HI	<40	<40	81,920	<40	<40
		VN	<100	<100	819,200	<100	<100
	R36	HI	40,960	1,280	40,960	40,960	5,120
		VN	204,800	204,800	819,200	51,200	25,600
αY19/	1A8	HI	204,800	40	160	10,240	20,480
		VN	819,200	6,400	6,400	204,800	819,200
	1C10	HI	<40	<40	<40	320	<40
		VN	<100	<100	<100	3,200	<100
	1E12	HI	10,240	10,240	20,480	10,240	20,480
		VN	204,800	102,400	204,800	409,600	819,200
	2D11	HI	40,960	40	160	10,240	5,120
		VN	819,200	<100	400	204,800	51,200
	4D6	HI	<40	<40	<40	2,560	<40
		VN	<100	<100	<100	12,800	<100

^a Titers are obtained using HI or VN tests.

Identification of antigenic epitopes by mAbs

We generated VN escape mutants against homologous viruses to analyze the antigenic epitopes on the HEF that the mAbs recognized (two for αOK, three for αNE, three for αY16, and four for αY19 mAbs) (Table 3). If mutants could not be selected against the homologous virus, they could be generated against the heterologous OK (with αNE/2G3, αNE/5F1, αY16/B4, and αY19/1E12) and NE strains (with αY19/1E12). We confirmed that all viruses selected were VN escape mutants with dramatically reduced VN titers (Table 3), suggesting that VN epitopes that the mAbs recognized were lost or altered in the mutant viruses. Therefore, we compared the HEF sequences of the escape mutants with those of their wild-type parent viruses. Before comparison, we checked the genome sequences of the parent viruses used in this study, which indicated that the HEF amino acid sequence of OK was identical to that listed in the National Center for Biotechnology Information (NCBI) database, whereas the others possessed different residues from the database in HEF: K251T/G290R with NE, E290K with Y16, and D203A/M276V/A473T with Y19. Sequence comparison revealed amino acid changes in the escape mutants against αOK mAbs (K212R for 1F2 and K212N for 4C7), αNE mAbs (D373N for 2G3, S190F for 5E2, T187I for 5F1, A252V for 7B3, and K177E and F464S for 7D7), αY16 mAbs (S341L and S342R for B4, N208D and S306R for G22, G210R and A211T for G27, and S190F and S306R for R36), and αY19 mAbs (T187I for 1A8 and 2D11, Q214K for 1C10, S189F for 1E12, and L33H and Q214R for 4D6), respectively (Table 3). Notably, two mutations at relatively distant positions on the HEF molecule were observed in the escape mutants with αNE/7D7, αY16/G22, αY16/R36, and αY19/4D6. We created recombinant viruses with every single mutation by reverse genetics and examined their HI and VN titers against the parent viruses because we cannot determine which mutation is responsible for VN escape with these mutants except αY16/R36 (Table S1). K177E, N208D, and Q214R were responsible for VN escape with αNE/7D7, αY16/G22, and αY19/4D6 mutants, respectively. Thus, the VN epitopes of OK HEF included 187T, 189S, 212K, 341S, 342R, and 373D; those of NE HEF included 177K, 189S, 190S, and 252A; those of Y16 HEF included 190S, 208N, 210G, and 211A; and those of Y19 HEF included 187T and 214Q.

TABLE 3.

Reactivities of the mAbs against the parent and mutant viruses

mAb		Virus	VN titer to		Mutation
			Parent virus	Mutant virus	Nucleotide	Amino acid
αOK/	1F2	OK	800	<100	A659G	K212R
	4C7	OK	800	<100	A660C	K212N
αNE/	2G3	OK	3,200	<100	G1141A	D373N
	5E2	NE	204,800	800	C593T	S190F
	5F1	OK	25,600	1,600	C584T	T187I
	7B3	NE	800	<100	C779T	A252V
	7D7	NE	25,600	<100	A553G/T1415C	K177E/F464S
αY16/	B4	OK	12,800	<100	C1045T/G1047T	S341L/R342L
	G22	Y16	819,200	400	A646G/C942A	N208D/S306R
	G27	Y16	1,638,400	<100	G652A/G655A	G210R/A211T
	R36	Y16	819,200	<100	C593T/C942A	S190F/S306R
αY19/	1A8	Y19	51,200	<100	C584T	T187I
	1C10	Y19	1,600	<100	C664A	Q214K
	1E12	OK	102,400	<100	C590T	S189F
		NE	409,600	<100	C590T	S189F
	2D11	Y19	51,200	100	C584T	T187I
	4D6	Y19	12,800	100	T122A/A665G	L33H/Q214R

Epitope mapping on the HEF molecule

We plotted the amino acid positions observed in the VN escape mutants, which indicated that these positions were included in the VN epitopes on the HEF monomer (Fig. 2). Positions 187, 189, and 190 are in the vicinity of the binding region for the 9-O-acetylated sialic acid receptor of the target cells. Positions 208, 210, 212, and 214 are located in the apex region of the HEF molecule. Positions 341, 342, and 373 are in the esterase domain of the HEF. Positions 177 and 252 in the HEF head region are not extremely distant from the receptor-binding region or the top region, respectively.

Fig 2.

The positions of the amino acids with mutations detected in the VN escape viruses on HEF monomer [Protein Data Bank (PDB) ID: 5E65] (22). The mAb corresponding to each mutation is designated in parentheses. The receptor analog of 9-O-acetylated sialic acid is indicated in magenta. Positions of amino acids 187–190 (indicated in green) are a part of the receptor-binding domain. Amino acid 214 is indicated in yellow, while the structure containing amino acids 210–212 remains ambiguous. Variations in amino acids across lineages suggest no functional concern regarding this HEF apex region, although antibodies targeting this region could affect the receptor-binding abilities of HEF. The positions of other amino acids with mutations 177, 208, and 252 are indicated in red and those of amino acids 341, 342, and 373 located on the esterase domain are indicated in blue.

Commonality of the HEF epitopes among the virus lineages

The mAbs (αY19/1A8, αY19/1E12, and αY16/R36) recognized the epitopes in the receptor-binding region that cross-reacted with all heterologous strains (Table 2), suggesting that these epitopes are common among the virus lineages. To this end, we surveyed the amino acid conservation at positions 187, 189, and 190 observed in the VN escape mutants with these mAbs (Table 4). As expected, 187T and 189S were mostly conserved among all lineages, indicating that these residues were responsible for forming the lineage-conserved common epitope. Interestingly, 190S was conserved among the four lineages, except for the D/Yama2019 lineage, suggesting that 190T with the D/Yama2019 lineage could be permissible for a common epitope. Notably, the other mAbs (αNE/5F1, αY19/2D11, and αNE/5E2) recognized the epitopes in the receptor-binding region and cross-reacted to multiple, albeit not all, lineage viruses. These findings imply that several common epitopes, including 187T and 190S, are present in this region. Similarly, the mAbs (αY16/B4 and αNE/2G3) recognized the epitopes in the esterase domain region and cross-reacted to all or multiple heterologous strains (Table 2), suggesting that these epitopes are common among the virus lineages. Additionally, the mAb αNE/7B3, which recognized the epitope including 252A that is a bit far from the receptor-binding region, reacted to multiple lineage viruses other than rCA/OK. The commonality of 252A may be regulated by amino acid sequences other than 252A because it is evident in the D/CA2019 lineage viruses (Table 4).

TABLE 4.

Amino acid variations at the positions included in the VN epitope among the viral lineages

HEF region	Position	Amino acid	Percentage (%) of each lineage virus (no. of strains)
			OK (106)	660 (40)	Yama16 (8)	Yama19 (9)	CA19 (3)
Receptor-binding region	187	T	100	100	100	100	100
	189	S	99.1	100	87.5	100	100
		F	0.9	0	0	0	0
		A	0	0	12.5	0	0
	190	S	100	100	100	0	100
		T	0	0	0	100	0
Apex region	208	N	98.1	100	100	100	100
		K	1.9	0	0	0	0
	210	G	100	100	87.5	0	100
		R	0	0	12.5	77.8	0
		K	0	0	0	22.2	0
	211	A	86.8	100	100	100	0
		T	11.3	0	0	0	0
		V	1.9	0	0	0	100
	212	K	100	2.5	0	0	0
		R	0	97.5	0	0	0
		S	0	0	100	100	100
	213	P	100	100	100	0	100
		S	0	0	0	100	0
	214	Q	100	100	100	100	0
		K	0	0	0	0	100
	215	V	100	0	37.5	100	100
		A	0	100	0	0	0
		L	0	0	62.5	0	0
Esterase domain	341	R	100	100	100	100	100
	342	R	100	100	100	100	100
	373	D	100	100	100	100	100
Other	177	K	100	100	100	100	100
	252	A	63.2	100	100	100	100
		T	33.0	0	0	0	0
		K	3.8	0	0	0	0

HEF epitope divergence among the virus lineages

The mAbs (αY16/G27, αOK/1F2, αOK/4C7, αY19/1C10, and αY19/4D6) recognized the epitopes on the apex region of the HEF molecule (Fig. 2) and reacted only to each homologous strain in the HI and VN tests (Table 2). Therefore, the amino acids at positions 210, 211, 212, and 214 at the apex of the HEF molecule, whose mutations were observed in the VN escape viruses with these mAbs (Table 3), are strongly suggested to form lineage-specific epitopes. Interestingly, a lineage-specific amino acid conservation that was different from that of the other lineages was observed in this region: 210R/K for the D/Yama2019 lineage, 211V and 214K for the D/CA2019 lineage, and 212K for the D/OK lineage, suggesting that these specific residues may form lineage-specific epitopes (Table 4). Additionally, mAb αNE/7D7, which recognized the epitope including 177K in the top region that is slightly far from the apex (Fig. 2), reacted only with the homologous NE strain (Table 3), although the 177K is conserved among all virus lineages (Table 4). Thus, D/660 lineage specificity of this epitope was determined using amino acid sequences other than 177K.

Amino acids responsible for the lineage-specific HEF epitope

We constructed mutant viruses with altered amino acid sequences at positions 210, 212, and 213 using reverse genetics to investigate whether the apex region of HEF is responsible for antigenic variations among viral lineages. We generated a series of OK-based mutant viruses in which the amino acids at positions 212 and 213 were changed to the intended positions: recombinant OK (rOK)-KP (retaining OK-type sequence), rOK-RP (NE-type sequence), rOK-SP (Y16 and rCA/OK-type sequence), rOK-AP (alanine control), rOK-SS (Y19-type sequence), rOK-KS (OK and Y19-mixed type sequence), and rOK-RS (NE and Y19-mixed type sequence). Subsequently, we generated the Y19 R210G mutant (rY19-G), in which the R at position 210 was changed to G, which is common among the other four lineages (Table 5). Using these recombinant viruses, the HI assay was performed with five lineage-specific mAbs (αOK/1F2, αOK/4C7, αY16/G27, αY19/1C10, and αY19/4D6) (Table 5). The αOK/1F2 and αOK/4C7 reacted to the wild-type OK and rOK-KP, but not to all the other viruses, indicating that the OK lineage-conserved 212-KP-213 (Table 4) are responsible for defining the lineage-specific epitope and determining the D/OK lineage-specific antigenicity. The Y16-specific αY16/G27 reacted to rOK-SP, rOK-SS, and rY19-G, which possessed 210G and 212S, indicating that these residues are required for D/Yama2016 lineage-specific epitope (Table 5). However, αY16/G27 did not react to rCA/OK (Table 2), which also possessed these residues (Table 4). These findings suggest that the D/Yama2016 lineage-conserved 210-GAS-212 and 214Q (Table 4) are important for defining this epitope and for determining D/Yama2016 lineage-specific antigenicity. The Y19-specific αY19/1C10 and αY19/4D6 reacted to rOK-SS, rOK-KS, rOK-RS, and rY19-G, indicating that they reacted to the viruses with 213S, regardless of the amino acid types at positions 210 and 212 (Table 5). Additionally, the escape mutant viruses from αY19/1C10 and αY19/4D6 possessed Q214K/R (Table 3). Therefore, D/Yama2019 lineage-conserved 213-SQ-214 (Table 4) is responsible for defining the lineage-specific epitope and determining D/Yama2019 lineage-specific antigenicity.

TABLE 5.

Reactivity of the recombinant mutants against monoclonal antibodies^a

Virus	Amino acid sequence at positions 210–215						HI titer to
	210	211	212	213	214	215	αOK/		αY16/	αY19/
							1F2	4C7	G27	1C10	4D6
OK	G	A	K	P	Q	V	1,280	2,560	<40	<40	<40
NE	G	A	R	P	Q	A	<40	0	<40	<40	<40
Y16	G	A	S	P	Q	L	<40	<40	81,920	<40	<40
Y19	R	A	S	S	Q	V	<40	<40	<40	320	2,560
rCA/OK	G	V	S	P	K	V	<40	<40	<40	<40	<40
rOK-KP	G	A	K	P	Q	V	2,560	1,280	<40	<40	<40
-RP	G	A	R	P	Q	V	<40	<40	<40	<40	<40
-SP	G	A	S	P	Q	V	<40	<40	20,480	<40	<40
-AP	G	A	A	P	Q	V	<40	<40	<40	<40	<40
-SS	G	A	S	S	Q	V	<40	<40	2,560	160	2,560
-KS	G	A	K	S	Q	V	<40	<40	<40	320	1,280
-RS	G	A	R	S	Q	V	<40	<40	<40	320	1,280
rY19-G	G	A	S	S	Q	V	<40	<40	320	160	1,280

^a Amino acid substitutions in the recombinant mutants of the wild-type virus are indicated in boldface and underlined.

DISCUSSION

After being recognized as a new genus of the influenza virus by the International Committee of Taxonomy of Viruses in 2016 (23), several IDV strains have been detected worldwide and classified into five genetic lineages based on the HEF coding sequences. These HEF amino acid sequences are highly conserved and are presumed to represent a single IDV serotype, unlike influenza A viruses, which are classified into multiple subtypes. Herein, we investigated the antigenic homogeneity and heterogeneity among virus lineages by preparing an mAb panel against the HEF proteins of various virus lineages. Analyses of their reactivities to homologous and heterologous lineage strains, along with the VN escape mutants, revealed that the receptor-binding region and esterase domain of HEF possessed VN epitopes that are conserved in multiple lineage viruses. These findings are likely attributable to the indispensable structural and functional commonality of both the regions, leading to single serotype characteristics of IDVs. Contrastingly, the apex region of the HEF molecule possesses epitopes unique to each viral lineage, providing antigenic specificity for each viral lineage. Furthermore, a reverse genetics study with recombinant mutants indicated that the amino acids within positions 210–214 of the HEF are essentially responsible for lineage-specific antigenicity. These findings demonstrated that considerable antigenic variation exists in HEF among the IDV lineages.

We created an antigenic cartography of the IDV strains using the HI assay data from this study, together with previous studies (16, 20, 21), to confirm the antigenic variations in HEF among IDV lineages (Fig. S1). Except for one strain, D/bovine/Texas/3-13/2011 of the D/660 lineage, which possessed a unique 212K resembling those of the D/OK lineage, an obvious antigenic distance was observed between the D/OK and D/660 lineage viruses. Additionally, the D/CA lineage virus, albeit the only recombinant virus tested, appeared to possess antigenicity different from those of the D/OK and D/660 lineages. Furthermore, antigenic distances were observed between the D/Yama2016 and D/Yama2019 lineage viruses, although the antigenicity of D/Yama2016 and D/Yama2019 seemed close to those of the D/OK and D/CA lineages, respectively. These data suggest that multiple IDV strains with antigenic heterogeneity are circulating in countries such as the USA and Japan, requiring proper vaccine strain selection when considering IDV vaccine development.

The escape mutants of the mAbs that recognized the epitopes in the receptor-binding region possessed mutations at positions 187, 189, and 190 (Table 2; Fig. 2). A crystallography study indicated that T187 and S189 are hydrogen-bonded to an analog of the receptor 9-O-acetylated sialic acid (22), suggesting that these two residues are important for receptor binding of the IDVs. Interestingly, the plaque sizes of the recombinant viruses with a single mutation at positions 187, 189, or 190 were slightly smaller than those of the wild-type viruses, suggesting that these mutations may not drastically affect the receptor-binding properties and cytopathology of the virus. However, the T187I, S189F, and S190F mutations detected in the VN escape mutants were not observed in any of the IDV isolates, except for one (Table 4), indicating that these mutations would not be introduced under natural settings. This suggests that the common epitopes in the receptor-binding region of the HEF are robustly conserved, explaining the single-serotype characteristics of IDVs. Similarly, the common epitope containing the 341-RR-342 in esterase domain was robustly conserved, contributing to these characteristics.

The mAbs recognizing the epitopes in the apex region of the HEF demonstrated VN activities only to each homologous virus (Table 2; Fig. 2). The amino acid sequences at positions 210–215 included in this region exhibited great variation among the viral lineages (Table 5), suggesting that they may provide lineage-specific epitopes. This hypothesis was verified by analyzing a series of recombinant viruses with amino acid substitutions in their sequences. We demonstrated that 212-KP-213 might determine the antigenic specificity of OK, 210-GASxQ-214 for Y16, and 213-SQ-214 for Y19 strains, suggesting that this definition may apply to viruses belonging to each lineage because of the conservation of the indicated amino acid sequences (Table 4). A D/660 lineage virus possessing 212-KP-213 exhibited high cross-reactivity with the anti-OK serum, unlike other D/660 lineage viruses (16), supporting our hypothesis. Taken together, we conclude that IDVs can tolerate mutations in the apex region of the HEF, possibly leading to antigenic drift, which should be considered for effective vaccine development. The uppermost region of the HEF of influenza C virus has been reported to form lineage-specific antigenic epitopes (24).

A β-propiolactone-inactivated construct has been assessed in calves as a prototype vaccine against influenza D, demonstrating that it was immunogenic but provided limited protection against a homologous virus challenge (25). Additionally, a DNA vaccine, in which consensus HEF sequences mixed with D/OK and D/660-lineage HEF gene sequences were designed, protected the guinea pigs from infection with both lineages of viruses (26). Although the efficacy of this DNA vaccine needs to be assessed in cattle, the data suggest that HEF gene manipulation can be applied to the preparation of other vaccine modalities with cross-reactive antigenicity. Therefore, dissecting HEF antigenicity, such as the characterization of the antigenic epitope structure, is important for selecting and creating influenza D vaccine candidates. Based on this information, we will be able to construct vaccine viruses with artificially manipulated antigenicity. One possible approach is to utilize the D/OK-based reverse genetics system to generate vaccine viruses targeting the D/Yama2016 and D/Yama2019 lineages via the introduction of HEF K212S/P213S mutations. Such tailored virus designs can be expected to make a significant practical contribution to future influenza D vaccine strategies.

MATERIALS AND METHODS

Cells and viruses

Human embryonic kidney 293T cells (obtained from Riken BRC, RCB2202), human rectal tumor 18G (HRT-18G) cells (obtained from ATCC, CRL-11663), and swine testis (ST) cells (obtained from ATCC, CRL-1746) were maintained in Dulbecco’s modified Eagle’s medium (Fujifilm Wako Pure Chemical, Osaka, Japan) supplemented with 10% fetal bovine serum (FBS) at 37°C. Mouse myeloma P3X63Ag8U.1 (P3U1) cells (obtained from NIBIOHN, JCRB0708) were maintained in Roswell Park Memorial Institute 1640 medium (Fujifilm Wako Pure Chemical, Osaka, Japan) supplemented with 10% FBS at 37°C. D/swine/Oklahoma/1334/2011 (OK) (GenBank accession no. JQ922308), a virus belonging to the D/OK lineage, and D/bovine/Nebraska/9-5/2012 (NE) (accession no. KM392471), a virus belonging to the D/660 lineage, were kindly provided by B. Hause (Kansas State University). D/bovine/Yamagata/10710/2016 (Y16) (accession no. LC318668) belonging to the D/Yama2016 lineage and D/bovine/Yamagata/1/2019 (Y19) (accession no. LC494108) belonging to the D/Yama2019 lineage were used (Fig. 1). These IDVs were propagated in ST cells in Eagle’s minimum essential medium (MEM; Life Technologies/Gibco, Paisley, UK) containing 0.3% bovine serum albumin (MEM/BSA) and supplemented with 0.5 µg/mL L-1-tosylamide-2-phenylmethyl chloromethyl ketone (TPCK)-trypsin (Worthington, Lakewood, NJ). The supernatant was centrifuged at 3,500 rpm for 10 min at 4°C and stored at −80°C. We also used a recombinant reassortant virus (rCA/OK) possessing the HEF segment of D/bovine/California/0363/2019 (CA) (accession no. MW020308), belonging to the D/CA2019 lineage, and all other segments of OK, which were generated using reverse genetics with the synthesized CA HEF gene.

Virus purification

The IDVs were purified and used as antigens for immunization in mice, and an enzyme-linked immunosorbent assay (ELISA) was used to detect specific antibodies. The virus-containing supernatants from the ST cells were clarified using low-speed centrifugation to remove cell debris and concentrated using ultracentrifugation at 16,000 rpm for 2.5 h at 4°C using RP19-262 rotor (Himac, Tokyo, Japan). The concentrated viruses were resuspended in phosphate-buffered saline (PBS), layered onto a 20%, 30%, 50%, and 60% discontinuous sucrose gradient, and ultracentrifuged at 30,000 rpm for 2.5 h at 4°C using P32ST rotor (Himac). The virus-containing interface between the 30% and 50% sucrose phases was collected, diluted in PBS, and ultracentrifuged once more at 28,000 rpm for 1.5 h at 4°C using the P32ST rotor. Purified viral pellets were suspended in a small volume of PBS. The protein concentration of the purified viral solution was determined using a BCA Protein Assay Kit (TAKARA BIO, Tokyo, Japan).

HA assay

The HA assay was performed in U-bottom 96-well microplates. Briefly, serial twofold dilutions of the virus-containing supernatants in 50 µL of PBS were mixed with 50 µL of 0.7% turkey red blood cells (Nippon Bio-test Laboratories, Saitama, Japan) and incubated for 30 min at 23°C before reading. The HA titers were determined as the reciprocal of the highest viral dilution showing complete HA.

Plaque assay

Confluent ST cells on a 12-well plate were washed two times with PBS, inoculated with 0.1 mL each of serially 10-fold diluted viruses in MEM/BSA, and incubated for 1 h at 37°C. After washing with MEM/BSA, the cells were covered with 1 mL of MEM/BSA containing 1% Seakem GTG agarose (Lonza Japan, Chiba, Japan) and 0.5 µg/mL TPCK-trypsin, and incubated at 37°C for 3 days. Subsequently, 0.5 mL of 30% formalin in PBS was added to each well for fixation at 4°C overnight. After the formalin and agarose were removed, the cells were washed with water. Plaques were visualized using amido black staining, and the plaque-forming units (pfu) were determined.

ELISA

The viral solution was diluted to 5 µg/mL using bicarbonate buffer [15 mM Na₂CO₃ and 35 mM NaHCO₃ (pH 9.6)]. They were subsequently added to ELISA plates (Nunc immunoplate MaxiSorp; Thermo Scientific, Waltham, MA, USA) at 0.5 µg/well and incubated at 4°C overnight. The plate was washed three times using PBS containing 0.05% Tween-20 (PBST) and blocked with BlockAce (MEGMILK SNOW BRAND, Tokyo, Japan) for 30 min at 23°C. After washing three times with PBST, the plates were incubated with hybridoma culture supernatant or diluted anti-IDV mouse immune serum or for 1 h at 23°C, followed by incubation with horseradish peroxidase-labeled goat anti-mouse IgG (Cytiva, Tokyo, Japan) for 30 min and 3,3´,5,5´-tetramethylbenzidine for 5 min (TMB Substrate Kit; Vector Laboratories, Burlingame, CA, USA). The enzymatic reaction was stopped by adding 25 µL of 1 N H₂SO₄, and the absorbance (A450-A620) of each well was measured using a plate reader (Bio-Rad Laboratories).

Generating mAbs against HEF

Four-week-old female BALB/c mice (Japan SLC, Shizuoka, Japan) were subcutaneously immunized three times with 50 µg purified virus with 100 µL adjuvant TiterMaxGold (Titer Max, Norcross, GA, USA) at 2-week intervals. Subsequently, a small amount of blood was collected to measure the antibody titer using ELISA. The mice that demonstrated high antibody titers were selected and administered with 0.5 µg of purified virus into their tail vein as booster immunization. The mice were euthanized 1 week later by cardiac blood sampling under deep anesthesia. The collected blood was centrifuged at 5,000 rpm for 5 min at 4°C, obtaining the polyclonal serum antibody against each IDV strain. The plasma cells were collected from the spleen and fused with mouse myeloma P3U1 cells in the presence of polyethylene glycol to produce hybridomas. Each hybridoma was screened for antibody secretion using ELISA. The hybridomas were cloned using three time-limiting dilutions. The culture supernatants of the hybridomas were subjected to HI assay for the selection of the hybridomas that produced anti-HEF antibodies. Positive hybridomas were administered into the abdominal cavity of mice pretreated with pristane, and ascites containing mAbs were collected. The collected ascites were clarified by centrifugation at 10,000 rpm for 30 min at 4°C and stored at −80°C. Finally, two clones of anti-HEF mAbs were obtained for OK (αOK/1F2 and αOK/4C7), five clones for NE (αNE/2G3, αNE/5E2, αNE/5F1, αNE/7B3, and αNE/7D7), four clones for the Y16 (αY16/B4, αY16/G22, αY16/G27, and αY16/R36), and five clones for the Y19 (αY19/1A8, αY19/1C10, αY19/1E12, αY19/2D11, and αY19/4D6) and used in subsequent experiments.

HI assay

The HI assay was conducted according to the WHO Manual on the Diagnosis and Investigation of Animal Influenza (27). The samples were treated with receptor-destroying enzyme [RDE (II); Denka Seiken, Tokyo, Japan] at 37°C for 16 h, followed by heat inactivation at 56°C for 30 min. Serially diluted samples (25 µL) were subsequently allowed to react to an equal volume of each IDV (4 HA units) for 30 min at 23°C, followed by incubation with 0.7% turkey red blood cells for 20 min at 23°C before reading. HI titers were determined as the reciprocal of the highest serum or ascites dilution demonstrating complete HA inhibition.

VN assay

The ascites was heat inactivated at 56°C for 30 min before the assay. Their serially twofold dilutions in BSA/MEM were mixed with the same volume of virus (100 pfu) and incubated at 37°C for 30 min. Confluent ST cells on a six-well plate were washed two times with PBS, inoculated with 0.1  mL of each mixture, and incubated for 1 h at 37°C. The subsequent steps were performed in the same manner followed for the plaque assay. The reciprocal of the highest ascites dilution that reduced plaque formation by >80% was indicated by VN titer.

Generating the VN escape mutant virus

The serially diluted mAb and viruses were mixed and incubated for 30 min at 37°C before infecting the ST cells. Viruses grown in wells with the highest mAb concentrations were collected, mixed again with diluted mAb, and used to infect the cells. Alternatively, the viruses were passed through a medium containing mAbs. These procedures were repeated, resulting in the generation of VN escape mutant viruses that could grow in the presence of mAbs. Sanger sequencing was performed on the HEF segment of the virus to identify the amino acid mutations responsible for VN escape.

Reverse genetics

We applied our reverse genetics system to IDV to generate recombinant viruses (28). We constructed pHH21-based plasmids (29) expressing viral RNA (vRNA) of the HEF segment of the OK, Y19, or CA strains with or without mutations by extracting the vRNA from each virus using Isogen-LS reagent (Nippon Gene, Tokyo, Japan). The extracted RNA was reverse transcribed using ReverTra Ace (Toyobo, Osaka, Japan). The cDNA was amplified via polymerase chain reaction (PCR) using KOD FX Neo (Toyobo), segment-specific primers containing a 15-nucleotide overlap sequence at the cloning site of pHH21 (29), and primers introducing the amino acid mutation of interest. The PCR products were purified using the FastGene gel/PCR extraction kit (Nippon Genetics, Tokyo, Japan) and cloned into BsmBI-digested pHH21 cells using a Gibson assembly master mix (New England BioLabs Japan, Tokyo, Japan). This resulted in the pPolI-HEF plasmids. Alternatively, the PCR products derived from NE or Y16, together with the DNA cassettes of the human RNA polymerase I promoter and murine RNA polymerase I terminator sequences from pHH21, were inserted into BamHI-digested pSMART BAC v2.0 (Lucigen, LGC Biosearch Technologies, Middleton, WI, USA), resulting in the BAC-polI-HEF. The HRT-18G cells were transfected with 0.1 µg of pPolI-HEF for OK, Y16, or CA or 0.2 µg of BAC-polI-HEF for NE or Y19, together with 0.1 µg each of pPolI-PB2, PB1, P3, NP, M, and NS of OK (29) and 0.25 µg each of the protein expression plasmid containing pCAGGS-PB2, PB1, P3, and NP of OK. After 2 days of incubation, the cells were washed with PBS and incubated with MEM/BSA with 0.5 µg/mL TPCK-trypsin for 4 days. The supernatant was transferred to the ST cells and incubated for an additional 2 to 3 days. Virus rescue was confirmed by the appearance of cytopathic effects and the presence of HA. Using these procedures, we generated rOK and a series of rOK mutants, NE, Y19, or CA reassortant viruses (rNE/OK, rY19/OK, or rCA/OK), and their mutants with HEF point mutations.

ETHICS APPROVAL

Our animal study protocols were conducted in accordance with the Regulations for Animal Care at the University of Tokyo and approved by the Animal Experiment Committee of the Graduate School of Agricultural and Life Sciences at the University of Tokyo (approval number P23-069) following the Fundamental Guidelines for Proper Conduct of Animal Experiment and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education, Culture, Sports, Science and Technology in Japan.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/jvi.01908-23.

Supplemental material. jvi.01908-23-s0001.pdf.

Tables S1 and S2; Fig. S1.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.