• Open Access

Discrimination of influenza A subtype by antibodies recognizing host-specific amino acids in the viral nucleoprotein

Authors


Tohru Miyoshi-Akiyama, Department of Infectious Diseases, Research Institute, National Center for Global Health and Medicine, 1-21-1, Toyama, Shinjuku-ku, Tokyo 162-8655, Japan. E-mail: takiyam@ri.ncgm.go.jp

Abstract

Please cite this paper as: Miyoshi-Akiyama et al. (2012) Discrimination of influenza A subtype by antibodies recognizing host-specific amino acids in the viral nucleoprotein. Influenza and Other Respiratory Viruses 6(6), 434–441.

Background  Nucleoprotein (NP) of influenza viruses is utilized to differentiate between the A, B, and C viral serotypes. The availability of influenza genome sequence data has allowed us to identify specific amino acids at particular positions in viral proteins, including NP, known as “signature residues,” which can be used to discriminate human influenza A viruses from H5N1 highly pathogenic avian influenza in human cases (HPAI) and pandemic H1N1(2009) (H1N1/2009) viruses.

Methods  Screening and epitope mapping of monoclonal antibodies (mAb) against NP of influenza A, which reacted differently with NP from human influenza A virus from HPAI and H1N1/2009 A virus. To identify the epitope(s) responsible for the discrimination of viral NP by mAbs, we prepared mutant NP proteins in the 293 cell expression system because some of the mAbs reacted with non-linear epitopes.

Results and Conclusions  In the present study, we identified 3 mAbs. The results of epitope mapping showed that the epitopes were located at the signature residues. These results indicated that signature residues of NP could discriminate influenza A viruses from different origin.

Introduction

The recent pandemic of swine-origin H1N1 2009 influenza A (H1N1/2009) virus1,2 and outbreaks of human cases of H5N1 highly pathogenic avian influenza (HPAI) infection3 (http://gamapserver.who.int/mapLibrary/app/searchResults.aspx) have prompted the development of diagnosis methods, which specifically detects new-type influenza. In clinical practice, rapid diagnostic kits (RDKs) based on immunochromatography utilizing antibodies against nucleoprotein (NP) of influenza virus are used to diagnose influenza, allowing the immediate initiation of antiviral drug administration.4,5 We developed an RDK capable of distinguishing H1N1/2009 viruses from seasonal influenza viruses 6 and evaluated its diagnostic efficacy in a prospective multicenter clinical trial.7 During the course of these studies, we obtained a unique monoclonal antibody (mAb) that reacted with the NP proteins from H1N1/2009 virus and HPAI, but not those from human seasonal H1N1 and H3N2 viruses. Epitope mapping experiments showed that the mAb recognizes a specific sequence of amino acids found in the NP proteins from H1N1/2009 virus and HPAI viruses located at residues 16–18 of these NPs.6 These findings prompted us to identify amino acids that distinguish human influenza viruses from HPAI and H1N1/2009 viruses at specific positions in the NP proteins. Such residues are known as “signature” residues.8–11

In the present study, we identified 5 and 9 signature residues in the NP proteins of HPAI viruses from human cases and H1N1/2009 influenza viruses, respectively. During the screening of monoclonal antibodies (mAbs) against the NP proteins, we identified 3 mAbs that reacted differently to NP from human influenza A virus compared to those of HPAI and H1N1/2009. Epitope mapping indicated that these mAbs recognized residues identified as signature amino acids in each NP. These results indicated that host-specific amino acids of NP could discriminate influenza A viruses from different origin.

Materials and methods

Identification of signature residues in influenza A virus nucleoprotein (NP) and prediction of accessibility for antibody binding

A total of 1182 of NP sequences of H1N1/2009 viruses in addition to HPAI and human viruses based on the data from January 1, 2007 to September 11, 2009 registered as human cases were retrieved from the Influenza Virus Resources in National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/genomes/FLU/) and analyzed to identify signature amino acids that distinguish human influenza viruses from HPAI and H1N1/2009 viruses based on an alignment obtained using the blast program (http://blast.ncbi.nlm.nih.gov/Blast.cgi). To predict the accessibility of the signature residues for antibodies, the locations of the signature residues based on the crystal structure of the NP proteins12,13in silico using Cn3D (http://www.ncbi.nlm.nih.gov/Structure/CN3D/cn3d.shtml), a crystal structure viewing software.

To prepare an anti-NP mAb, recombinant NPs of influenza A virus (A/Viet Nam/VL-020/2005(H5N1)) (accession number: AAZ72762), a virus isolated from a patient infected with HPAI, and H1N1/2009 (A/California/04/2009) (accession number: ACP44151.1) were prepared from Escherichia coli BL21 (DE3) CodonPlus-RIPL (Stratagene, La Jolla, CA, USA) and used to immunize 7–9-week-old female WKY rats (Oriental Yeast Co. Ltd., Tsukuba, Japan), and rat mAbs were prepared as described.6 The mAb 3G2 was prepared by immunization of GANP Mice™ (TransGenic Inc., Kumamoto, Japan).

ELISA analysis of mAbs

Reactivity of the mAbs with NPs derived from seasonal influenza, H1N1/2009, and H5N1 was analyzed by conventional ELISA using microplates coated with NPs or by sandwich ELISA using microplates coated with polyclonal antibodies prepared from rabbits immunized with recombinant NPs as described previously.6

Sources of NP proteins for the sandwich ELISA included cultured human A/New York/55/2004(H3N2) and A/New Caledonia/20/1999 (H1N1) viruses in tissue culture, and recombinant NPs from HEK293 cells transfected with cytomegalovirus (CMV) promoter-driven plasmids14,15 encoding an NP gene with the sequence of H1N1/2009 (A/California/04/2009(H1N1)) and that of HPAI (A/Viet Nam/VL-020/2005(H5N1).

The concentration of each NP was normalized by conventional Western blotting with rabbit anti-NP polyclonal antibody. To perform sandwich ELISA, 250 ng of rabbit anti-NP polyclonal Ab dissolved in 50 mm sodium carbonate buffer (pH 9·0) was fixed to each well of a 96-well microtiter plate (Corning Inc., Corning, NY, USA) at room temperature for 1 h. After washing with phosphate-buffered saline containing 0·02% Tween-20 (PBS-T) and blocking with SuperBlock (Pierce, Rockford, IL, USA), 10 ng of the NP proteins dissolved in PBS-T was added to each well. Following incubation and washing, the wells were incubated with 50 ng/well of mAbs indicated. In conventional ELISA, 50 ng/wells of antigens were fixed onto the plates directly. Binding of mAbs was detected with the HRP-goat anti-rat IgG (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK) and TMB (Bio-Rad, Hercules, CA, USA).

Epitope mapping of mAb using recombinant NP fragments or synthetic peptides

Eight NP fragments (Table S1) derived from NP (A/California/04/2009(H1N1) Accession No. ACP44151) were prepared in E. coli as described above and used for epitope mapping of mAbs based on the ELISA results. Synthetic peptides prepared by a commercial service (500 ng each, > 70% purity; Invitrogen, Carlsbad, CA, USA) were fixed to the plates by incubation in 50 mm carbonate buffer (pH 9·0) containing 1 mm of the chemical cross-linker disuccinimidyl suberate (DSS; Pierce) at room temperature for 1 h, followed by epitope mapping using 500 ng/well of mAb indicated.

Epitope mapping of mAb using mutant NP proteins expressed in HEK293 cells

The NP proteins containing signature amino acids in the background of HPAI or H1N1/2009 viruses were produced by conventional PCR using mega-primers to introduce codons corresponding to each amino acid. Primer sequences are available on request. The resultant NP constructs were expressed in HEK293 cell as described above and used for epitope mapping using 50 ng/well of M322211.

Phylogenetic analysis of NP

The full-length amino acid sequence data of NP registered at NCBI (http://www.ncbi.nlm.nih.gov/genomes/FLU/Database/nph-select.cgi?go=database) dated 2005 to 2011 were retrieved. Number of sequences retrieved was 1917. Sequence alignment and phylogenetic analysis were performed using MUSCLE16 and MEGA5,17 respectively, on our own server.

Results

During the development of a rapid diagnostic kit specific for H1N1/2009 viruses using mAbs against the NP proteins of influenza A viruses, we obtained a mAb that reacted with NPs from H1N1/2009 and H5N1 HPAI viruses but not with those from seasonal H1N1 and seasonal H3N2 viruses. The epitope of the mAb, designated 6G6, was located at residues 16–18 of NP.6 The corresponding region of NPs from H1N1/2009 and H5N1 HPAI had the sequence GGE, while those of seasonal H1N1 and H3N2 were DGE and DGD, respectively. These findings prompted us to analyze the reactivity of mAbs against NPs of H1N1/2009, H5N1 HPAI, seasonal H1N1, or seasonal H3N2 prepared in our laboratory and from commercial sources.

Signature amino acids found in NPs from H1N1/2009 and HPAI viruses

Large-scale sequence analyses revealed signature amino acids at specific positions in viral proteins including NPs that distinguish human influenza viruses from avian viruses.9–11 We analyzed the 1182 of NP sequences of H1N1/2009 viruses in addition to HPAI and human viruses based on the data from January 1, 2007 to September 11, 2009 registered as human cases at Influenza Virus Resources in National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/genomes/FLU/) and found 5 and 9 signature residues specific for HPAI viruses and H1N1/2009 viruses, respectively (Table 1). These residues may be the keys to distinguish human seasonal viruses from HPAI and H1N1/2009 viruses using mAbs recognizing the viral NP proteins.

Table 1.   Signature amino acids specific for HPAI and H1N1/2009 viruses
Signature amino acidsHumanHPAIH1N1/2009Accessibility prediction
PositionResiduesH1N1H3N2
  1. A total of 1182 of NP sequence (372, 252, 14, and 542 of human H1N1, H3N2, HPAI in human cases, and H1N1/2001 registered as human cases at Influenza Virus Resources at NCBI (http://www.ncbi.nlm.nih.gov/genomes/FLU/) were analyzed to identify signature amino acids that distinguish human influenza viruses from HPAI and H1N1/2009 viruses. Prediction of the signature residues is based on the quatemary structure of the NP proteins.

 Specific for human HPAI
33V0/3720/25214/140/542Poor
100R0/3720/25214/140/542Good
136L0/3720/25214/140/542Poor
305R0/3720/25214/140/542Good
357Q0/3720/25214/140/542Good
 Specific for H1N1/2009
21D0/3721/2520/14542/542Good
53D0/3720/2520/14542/542Good
190A0/3721/2520/14542/542Poor
313V0/3720/2520/14542/542Good
316M0/3720/2520/14542/542Good
350K0/3721/2520/14542/542Poor
371V0/3721/2520/14542/542Good
433N0/3721/2520/14542/542Good
456L0/3721/2520/14542/542Poor

Characterization of mAbs that reacted differently with NP in an origin-dependent manner

3A4

The mAb 3A4 is one of the mAbs that was generated by immunization of recombinant NP from H1N1/2009 viruses (A/California/04/2009) and screening of mAbs that reacted with H1N1/2009 NP but not human or HPAI NPs. The results indicated that 3A4 reacted with NP from H1N1/2009 virus only (Figure 1A). To further narrow down the epitope of 3A4, we prepared 7 overlapping fragments of H1N1/2009 virus NP (Table S1). The mAb 3A4 reacted with fragments containing residues 1 – 56 (Full, F1, F1-1, and F1-1-1)(Figure 1B). Then, we used a 15-mer overlapping synthetic peptide corresponding to residues 1–71 of H1N1/2009 NP. Although a peptide corresponding to residues 31–45 reacted with 3A4, it also reacted to the positive control mAb and negative control IgG indicating that this peptide recognition was non-specific and 3A4 reacted specifically with a peptide corresponding to residues 11–25 (Figure 1C). Comparison of the corresponding sequences in human, HPAI, and H1N1/2009 viruses indicated that only N21D varied specifically in NP from H1N1/2009 viruses, and this was identified as a signature residue specific for NPs from H1N1/2009 viruses (Table 1). Thus, we tested the reactivity of a peptide corresponding to residues 11–25 of HPAI against 3A4. The results showed that 3A4 reacted the H1N1/2009 version of peptide but not the one of HPAI version (Figure 1D). These results indicated that variation of 21D is responsible for the specific recognition of NP from H1N1/2009 viruses by 3A4.

Figure 1.

 Comparison of the reactivity of 3A4 with NPs from human, H1N1/2009, and HPAI viruses and the results of epitope mapping. mAb reactive with NP of H1N1/2009 and HPAI viruses6 was used as a positive control. (A) Reactivity of 3A4 with NPs of human, H1N1/2009, and HPAI viruses were analyzed by sandwich ELISA using polyclonal anti-NP antibodies as the capture antibodies. (B) Dissection of the 3A4 epitope using recombinant NP fragments by direct ELISA. Corresponding amino acids for each fragment are shown in Table S1. (C) Mapping of the epitope of 3A4 using 15-mer overlapping peptides with direct ELISA. (D) Dissection of the critical residues to discriminate the NP protein from viruses of different origin by 3A4 using synthetic peptides shown in the insert. Data are presented as means ± SD.

3G2

The mAb 3G2 was also raised against recombinant NP from H1N1/2009 viruses by a method similar to that used for 3A4. 3G2 reacted with NP from H1N1/2009 virus and to a lesser extent with NP from HPAI virus, but did not react with those from human seasonal viruses (Figure 2A). Reactivity against NP fragments showed that the epitope of 3G2 was located in residues 383 – 498 (Full and F3) (Figure 2B). Dissection of the epitope using 15-mer overlapping synthetic peptide corresponding to residues 415–469 of H1N1/2009 NP showed that 3G2 reacted specifically with a peptide corresponding to residues 455–469 (Figure 2C). We also analyzed the reactivity of 3G2 against synthetic peptides corresponding to residues 455–469 of NPs from human and HPAI viruses. The results showed that 3G2 reacted with H1N1/2009-type peptide and to a lesser extent with HPAI-type peptide but not with the human-type peptide (Figure 2D). Comparison of the corresponding sequences in human, HPAI, and H1N1/2009 viruses indicated that the only residue that varied specifically in both NPs from H1N1/2009 and HPAI viruses was E455D. This residue, E455D, is not a signature residue specific for NPs from H1N1/2009 virus, although it is adjacent to 456L, which is a signature residue specific for NPs from H1N1/2009 viruses (Table 1). These results indicated that it is likely that variation of 455D is responsible for the discrimination of NPs from H1N1/2009 and HPAI viruses by 3G2, and that the signature residue 456L also contributes to the discrimination between NP from H1N1/2009 and HPAI viruses.

Figure 2.

 Comparison of the reactivity of 3G2 with NPs from human, H1N1/2009, and HPIA viruses and the results of epitope mapping. mAb 3A4 (mentioned above), which reacts only with H1N1/2009 NP, was used as a positive control. (A) Reactivity of 3G2 with NPs of human, H1N1/2009, and HPAI viruses were analyzed by sandwich ELISA using polyclonal anti-NP antibodies as the capture antibodies. (B) Dissection of the 3G2 epitope using recombinant NP fragments by direct ELISA. (C) Mapping of the epitope of 3G2 using 15-mer overlapping peptides by direct ELISA. (D) Dissection of the critical residues to discriminate the NP protein from viruses of different origin using synthetic peptides shown in the insert. A peptide corresponding to residues 445–459 of the NP protein of H1N1/2009 virus was used as a negative control. Data are presented as means ± SD.

M322211

M322211 is a commercial mAb that was previously found to react with NPs from human and H1N1/2009 viruses but not with that from HPAI virus.6 M322211 did not react with recombinant NP fragments (data not shown), suggesting that the mAb recognizes a structural rather than a linear epitope. In a previous study, we used a series of chimeras consisting of NPs from human and avian viruses to locate the epitope of M322211, and the results indicated that the epitope was located within residues 1–188.6 In this study, we expressed recombinant avian NP carrying R100V mutation, which reverted a signature mutation in NP of HPAI virus to that of H1N1/2009 virus, in HEK293 cells and performed sandwich ELISA. The results showed that M322211 gained reactivity against avian NP by the introduction of R100V mutation (Figure 3B). R100 was shown to be responsible for the discrimination between NPs from human and H1N1/2009 viruses and that from HPAI viruses by M322211.

Figure 3.

 Comparison of the reactivity of M322111 with NPs of human, H1N1/2009, and HPAI viruses and the results of epitope mapping. mAb reactive with NP of H1N1/2009 and HPAI viruses6 was used as a positive control. (A) Reactivity of M322111 with NPs of human, H1N1/2009, and HPAI viruses were analyzed by sandwich ELISA using polyclonal anti-NP antibodies as the capture antibodies. (B) Reactivity of M322111 with NPs of wild-type avian virus and that with the introduction of R100V mutation. Wild-type and R100V mutants of avian NP proteins were expressed in HEK293 cells and used for the analysis by sandwich ELISA format. Data are presented as means ± SD.

Discussion

In the present study, we characterized three antibodies that can discriminate between the NP proteins from human influenza viruses and those of H1N1/2009 or HPAI origin. By dissection of the epitope of each antibody, we identified the mechanism of discrimination underlying the origin-dependent specific “signature residues” present in the amino acid sequence of NP.

As large amounts of NP protein sequence data are now available, we performed phylogenetic analysis of the NP proteins registered between 2006 and 2011 (Figure 4). The results indicated that the NP proteins of seasonal influenza, HPAI, and H1N1/2009 belong to different clades, even though all of these clades contain swine viral NP. Sequence identity among the NP proteins is around 90% (data not shown) and NP consists of about 500 amino acids, suggesting that about 50 amino acids could be different between two NP proteins selected at random. As these variations could be shared among the viruses, we compared the NP sequences and identified the specific residues in H1N1/2009 (nine residues) and HPAI isolated from human cases (five residues), which are different from the NP proteins of human seasonal viruses (Table 1). As mentioned above, the signature residues are responsible for the discrimination of NP among viruses of different origin by particular antibodies.

Figure 4.

 Phylogenetic analysis of NP based on their amino acid sequence. The NP sequences of human-, avian-, and swine-origin influenza viruses reported from 2005 to 2011 were included into the analysis. Full-length amino acid sequence data of NP registered in NCBI (http://www.ncbi.nlm.nih.gov/genomes/FLU/Database/nph-select.cgi?go=database) dated 2005 to 2011 were retrieved. Number of the sequences used in the analysis was 1917. Clades containing NPs of human-, avian-, and H1N1/2009 virus are indicated. All clades contained NP proteins of swine viruses.

To evaluate the accessibility of the signature residues identified by the antibodies, we predicted the locations of the signature residues based on the crystal structure of the NP proteins12,13in silico (Table 1). The NP proteins form oligomers in solution. Thus, in addition to residues not located on the surface of NPs, it would be difficult for antibodies to access those located at the interface of NP oligomers. The results of in silico analysis suggested that signature residues of HPAI at positions 33 and 136, and those of H1N1/2009 NP at positions 190, 350, and 456 are unlikely to be exposed to the external environment, while those of HPAI at positions 100, 305, and 357, and those of H1N1/2009 NP at positions 21, 53, 313, 316, 371, and 433 are likely candidates for recognition by antibodies. Indeed, the antibodies characterized in the present study only recognized the candidate signature residues, suggesting that these residues may be useful for the discrimination of influenza A virus based on the NP sequence.

In conclusion, we identified key amino acid residues in the NP protein, which could discriminate influenza A viruses based on the viral origin and verified the feasibility of their utilization to determine the origin of NP. As NP protein is the target of rapid influenza diagnostic kits, which have been implemented in clinical settings, the information presented here could be easily translated to clinical practice.

Acknowledgements

We thank Mrs. N. Saito for the preparation and analysis of the antibodies. This study was supported by a Grant for International Health Research (23A-301) from the Ministry of Health, Labor, and Welfare of Japan.

Ancillary