ABSTRACT
The recent association of certain influenza A virus subtypes with clinically relevant phenotypes has led to the increasing importance of subtyping by clinical virology laboratories. To provide clinical laboratories with a definitive immunofluorescence assay for the subtyping of influenza A virus isolates, we generated a panel of monoclonal antibodies (MAbs) against the major circulating influenza A virus subtypes using multiple inactivated H1N1, H3N2, and 2009 H1N1 strains individually as immunogens. Eleven MAbs that target hemagglutinin (HA) of H1N1 and H3N2 subtypes were selected. These MAbs were combined into three subtype-specific reagents, one each for pan-H1 (seasonal and 2009 strains), H3, and 2009 H1, for the subtyping of influenza A virus-positive specimens by indirect immunofluorescence assay (IFA). Each subtype-specific reagent was tested on 21 prototype influenza A virus strains and confirmed to be specific for its intended subtype. In addition, the subtyping reagents did not cross-react with any of 40 other viruses. The clinical performance of the subtyping reagents was evaluated with 75 archived clinical samples collected between 2006 and 2009 using the D3 Ultra DFA influenza A virus identification reagent (Diagnostic Hybrids, Inc., Athens, OH) and the influenza A virus subtyping reagents by IFA simultaneously. Sixty-four samples grew virus and were subtyped as follows: 30 as H3N2, 9 as seasonal H1N1, and 25 as 2009 H1N1. RT-PCR was used to confirm the influenza A virus subtyping of these samples, and there was 100% agreement with IFA. This subtyping IFA provides clinical laboratories with a cost-effective diagnostic tool for better management of influenza virus infection and surveillance of influenza virus activity.
INTRODUCTION
Influenza A virus is subtyped based on the antigenicity of its two surface glycoproteins, hemagglutinin and neuraminidase. Although 16 hemagglutinin (H1 to H16) and 9 neuraminidase (N1 to N9) variants have been identified, only three combinations (H1N1, H2N2, and H3N2) have been responsible for human epidemics (2). Today there are three influenza A virus subtypes circulating in humans: H3N2, seasonal H1N1, and 2009 H1N1. Subtyping influenza A virus in the clinical laboratory has become more important, especially since the emergence of viral mutations that confer drug resistance to antiviral treatment. According to the Centers for Disease Control and Prevention (CDC), nearly all influenza A virus H1N1 isolates tested during the 2008–2009 influenza season were resistant to oseltamivir, while 100% of influenza A virus H3N2 isolates tested were resistant to amantadine (3). The outbreak of pandemic H1N1 virus in 2009 also drew a great deal of attention to the identification of this influenza A virus subtype, which was found to be almost uniformly susceptible to oseltamivir (3). The fact that seasonal influenza A viruses H1N1, H3N2, and 2009 H1N1 and influenza B viruses with variable antiviral drug susceptibilities are cocirculating has driven the need for a rapid and accurate test for both typing and subtyping influenza virus isolates to establish a definitive laboratory diagnosis.
Although subtyping was historically based on antigenicity, today the most commonly used methods for subtyping influenza A virus are based on nucleic acid sequence-specific tests such as reverse transcription-PCR (RT-PCR) (1, 7, 9, 10, 15–17). However, many laboratories do not have the capability to perform molecular tests and might prefer to have a less technically demanding assay to identify influenza A virus subtypes. The traditional hemagglutination inhibition assay for subtyping is a time-consuming and tedious method. A number of reports have described the development of subtype-specific MAbs, but they have not resulted in reagents for routine use (11, 13, 14, 18). Recently we developed MAbs that exhibited specificity for the 2009 H1N1 pandemic influenza A virus, which formed the basis for a commercially available reagent for the identification of this virus (6). To extend this approach to other clinically relevant subtypes, we generated a series of subtype-specific MAbs for H3N2 and seasonal H1N1 subtypes as well and blended multiple subtype-specific MAbs to create indirect immunofluorescence assay (IFA) subtyping reagents. When used with a type-specific direct fluorescence assay (DFA), the subtyping IFA could enable the simultaneous typing and subtyping of currently circulating influenza A viruses from viral cultures.
MATERIALS AND METHODS
Cells and viruses.R-Mix cells in 24-well or 96-well formats and MRC-5 cells in conventional tube or 96-well formats (Diagnostic Hybrids, Inc. [DHI], Athens, OH) were used for virus infection for screening and selection of MAbs. Some IFAs and DFAs were performed with primary rhesus monkey kidney (rhMK) cells in shell vials (DHI, Athens, OH). The cells were cultured at 37°C in a CO2 incubator. Influenza A viruses and other prototype respiratory virus strains used in antibody development were from the DHI virus repository, and most of them were originally from the American Type Culture Collection (ATCC). The 2009 H1N1 viruses were obtained from the CDC. All virus stocks except those of human rhinovirus (HRV) strains were propagated in MDCK cells (ATCC, Manassas, VA) grown in flasks at 37°C in a CO2 incubator. HRV stocks were propagated in MRC-5 cells in slowly rotating tubes (DHI, Athens, OH) at 33°C. Both an enterovirus antigen control slide and an hMPV antigen control slide (DHI, Athens, OH) were used for evaluating the specificity and cross-reactivity of the influenza A virus subtyping reagents by IFA.
Clinical specimens.Nasopharyngeal aspirate specimens were obtained from University Hospitals, Case Medical Center (Cleveland OH). The specimens were collected from 2006 to 2009 under an institutional review board-approved protocol and had been stored at −70°C since collection. These samples were originally determined to be influenza A virus positive by DFA using the D3 Ultra influenza A virus reagent (DHI, Athens, OH) but had never been subtyped.
Generation of influenza A virus-subtyping MAbs.BALB/c mice were sequentially immunized with inactivated influenza A viruses representing four seasonal H1N1 strains (A/Solomon Islands/03/06, A/New Caledonia/20/99, A/Taiwan/1/86, and A/Malaya/302/1954), two H3N2 strains (A/Shangdong/9/93 and A/Brisbane/10/07) and one 2009 H1N1 strain (A/California/07/2009). The 2009 H1N1 strain A/California/07/2009 and the H1N1 strain A/Malaya/302/1954 were amplified in MDCK cells prior to concentration and inactivation with β-propiolactone. The H1N1 strains A/Solomon Islands/03/06, A/New Caledonia/20/99, and A/Taiwan/1/86 as well as the H3N2 strain A/Shangdong/9/93 were purchased from Meridian Life Science (Saco, ME), the H3N2 strain A/Brisbane/10/07 was purchased from Hy Test Ltd. (Turku, Finland). Hybridoma production was carried out using previously described methods (8). Hybridoma isolates and clones were screened on virus infected R-Mix cells by IFA against two 2009 H1N1 CDC strains and a panel of seasonal influenza viruses, including eight H3N2 strains and 10 H1N1 strains.
IFA and DFA.Standard protocols for IFA and DFA were followed as described previously for detecting influenza A viruses and other prototype virus strains (6). To perform IFA against HRV, 96-well plates seeded with MRC-5 cells were infected with the virus strains. The plates were centrifuged at 700 × g for 60 min and incubated at 33°C in a CO2 incubator for 20 h. The cell monolayers were fixed with 80% acetone for 10 min before IFA was carried out with the plates using the standard protocol. MAbs specific to the individual viruses were used as positive controls to confirm the infection level of each virus during experiments.
Hemagglutination inhibition assay (HIA).Several influenza A virus strains, including four seasonal H1N1 (A/Malaya/302/54, A/New Jersey/8/76, A/Texas/36/1991, and A/NWS/33), two H3N2 (A/Victoria/3/75 and A/Perth/16/2009), and one 2009 H1N1 (A/Mexico/4108/2009) strain were twofold serially diluted with phosphate-buffered saline (PBS) in a 96-well, round-bottom plate at 50 μl/well and mixed with an equal volume of 0.5% turkey red blood cells (LAMPIRE Biological Laboratories, Inc., Pipersville, PA) diluted in PBS. The lowest viral dilution displaying complete agglutination after a 30-min room temperature (RT) incubation was used for the HIA as follows. Each twofold serially diluted subtyping MAb with a starting concentration of 500 μg/ml was added to a 96-well, round-bottom plate in duplicate at 25 μl/well. Pairing the MAbs with their respective target viruses, an equal volume of PBS containing the appropriate amount of virus based on the hemagglutinin (HA) titer was added to the plate and incubated at RT for 30 min. Turkey red blood cells (50 μl) at 0.5% in PBS were added to each well, and the plate was incubated for 1 h at RT. The presence of agglutination in the wells was visually determined. An influenza A virus MAb targeting nucleoprotein was used as a negative control in addition to PBS.
Clinical-specimen testing.Archived clinical samples were individually inoculated into one well of R-Mix cells in 24-well plates with 100 μl inoculum. After centrifugation for 30 min at 900 × g, the plates were incubated at 34°C for 48 h. The culture supernatants were collected and frozen at −80°C for RT-PCR analysis. The infected cells were trypsinized, used to spot four wells on slides, and fixed with acetone. The slides were tested by DFA with the D3 Ultra kit and by IFA with the pan-H1 identification (ID) reagent, the 2009 H1 ID reagent, and the influenza A virus H3 ID reagent concurrently. The slides were examined for the presence of green-fluorescence-positive cells by an experienced professional without knowledge of the RT-PCR results.
RT-PCR.Nucleic acid from each influenza A virus culture specimen was extracted with a MagMAX-96 total nucleic acid isolation kit (Applied Biosystems, Foster City, CA) according to the manufacturer's protocol. Primers used to subtype influenza virus culture specimens originated from published reports involved in screening respiratory samples (4, 12). RT-PCR was performed on each sample using the three influenza primer sets in separate reactions with a Verso one-step RT-PCR kit (Thermo Scientific) according to the manufacturer's specification. The working primer concentration used for all primers in all reactions was 0.6 μM. The amplification protocol for all reactions was as follows: 50° for 15 min; 95°C for 2 min; 40 cycles of 94°C for 1 min, 60°C for 1 min, and 72°C for 30 s; and a final extension cycle of 72°C for 10 min. Each set of reverse transcription and PCR mixtures contained appropriate positive and negative controls. The PCR amplicons were run on 2.5% TAE (Tris-acetate-EDTA) agarose gels by electrophoresis, and after staining with ethidium bromide, gels were photographed under UV light and visually scored as positive or negative based on the presence of the correct-size band.
RESULTS
Generation of influenza A virus subtyping MAbs.Two 2009 H1N1 subtyping MAbs (9F8C7 and 16E11B3) were developed previously using inactivated A/California/07/2009 virus. These two MAbs were proven to have 100% specificity for the 2009 H1N1 subtype (6). Numerous H1N1 hybridomas were generated from five fusions using mice immunized with different inactivated H1N1 viruses, including A/Solomon Islands/03/2006, A/New Caledonia/20/99, A/Taiwan/1/86, A/Malaya/302/1954, and A/California/07/2009. These hybridomas were further screened several times by IFA against multiple strains of 2009 H1N1, seasonal H1N1, and H3N2 viruses to confirm the specificity and cross-reactivity of each MAb. Five H1N1 MAbs were selected after this diligent screening process. Of these five anti-H1N1 MAbs, three MAbs—6B6C10, 11D10D6, and 21F8G8—generated with A/Solomon Islands/03/2006, A/New Caledonia/20/99, and A/Taiwan/1/86, respectively, recognized either four or six seasonal H1N1 strains, while one MAb (23E6E6) generated with A/Malaya/302/1954 was specific to its target virus. The fifth anti-H1N1 MAb (9F11G7), generated with A/California/07/2009, recognized both the target virus and the A/New Jersey/8/76 strain. To develop MAbs for subtyping H3N2 viruses, two inactivated H3N2 strains, A/Shangdong/9/93 and A/Brisbane/10/07, were employed as antigens. A total of four H3N2 MAbs were selected after several rounds of hybridoma screening by IFA against numerous seasonal H3N2, H1N1, and 2009 H1N1 virus strains. Three MAbs (7H8C6, 21F6G4, and 21H5A9), generated with A/Shangdong/9/93, reacted with seven H3N2 strains, and one MAb (5D2), generated with A/Brisbane/10/07, recognized A/Brisbane/10/07, A/Uruguay/716/2007, and A/Perth/16/2009.
The target antigen of each MAb was confirmed to be hemagglutinin (HA) by a hemagglutination inhibition assay (HIA) using four seasonal H1N1 strains, two H3N2 strains, and one 2009 H1N1 strain. The HI titers varied from 1:2 to 1:12,800 among these subtyping MAbs, indicating different regions of each HA protein were recognized by the MAbs (Table 1). The antigen recognition sites of the two 2009 H1N1 MAbs were also analyzed by Western blot using full-length 2009 H1N1 HA protein expressed in Sf9 insect cells and the HA1 subunit protein expressed in Escherichia coli as targets, showing that the two 2009 H1N1 MAbs bound to either HA1 or HA2 subunits of 2009 H1N1 virus (6).
HA target protein confirmation by hemagglutination inhibition
Three groups of MAbs thus were identified as specific to each of the three circulating influenza A virus subtypes. These MAbs were blended to create three influenza A virus subtyping reagents: a pan-H1 ID reagent (i.e., identifies both seasonal H1 and 2009 H1), a 2009 H1 ID reagent, and a H3 ID reagent.
Analytical performance of the influenza A virus subtyping reagents.Specificity of the influenza A virus subtyping reagents was evaluated by IFA performed on R-Mix cell monolayers infected with nine seasonal H1N1 strains, 10 seasonal H3N2 strains, and two 2009 H1N1 strains individually. As shown in Table 2, the pan-H1 ID reagent stained positive on 11 prototype H1N1 strains, including two 2009 H1N1 strains, without cross-reacting with any of 10 prototype H3N2 strains. The H3 ID reagent stained positive on the 10 prototype H3N2 strains without showing cross-reactivity to any of the H1N1 strains. Finally, the 2009 H1 ID reagent stained the two 2009 H1N1 strains specifically with no cross-reactivity to 19 non-2009 H1N1 and H3N2 influenza A viruses (Table 2). The subtyping reagents were further evaluated for use in shell vials on rhMK cells with selected 2009 H1N1, H1N1, and H3N2 influenza A virus strains following the standard IFA protocol as described previously (6). Results of this test were in 100% agreement with the previous tests on R-Mix monolayers (Fig. 1). In addition, each subtyping reagent was evaluated by IFA against a panel of seven major non-influenza A respiratory viruses, including six influenza B virus isolates, 11 adenovirus serotypes, three RSV strains, eight rhinovirus strains, and parainfluenza viruses 1, 2, 3, 4a and 4b. The subtyping reagents were also tested by IFA against three hMPV subtypes and four different enteroviruses using commercially prepared antigen control slides (DHI, Athens, OH) with specific MAbs to each individual virus as positive controls. (Millipore's Light Diagnostics panenterovirus reagent, which cross-reacts with rhinovirus, was used to confirm rhinovirus infection.) None of the three influenza A virus subtyping reagents showed cross-reactivity to any of these viruses (Table 3).
Specificity of influenza A virus subtyping reagents by IFA
Staining of influenza A virus-infected rhMK cells with subtyping IFA reagents and the D3 Ultra influenza A virus ID DFA reagent.
Other respiratory viruses used for cross-reactivity testing of subtyping reagents
Subtyping of influenza A virus clinical isolates by IFA and RT-PCR.Seventy-five archived influenza A virus-positive clinical specimens collected between 2006 and 2009 were cultured in R-Mix cells and then spotted on four-well slides for testing with both the D3 Ultra influenza A virus ID reagent and the influenza A virus subtyping reagents concurrently. Sixty-four samples were positive by DFA for influenza A virus. These influenza A virus-positive samples were subtyped by IFA to contain 30 H3N2 and 34 H1N1, of which 25 were 2009 H1N1 subtypes (the remaining nine were assumed to be seasonal H1N1 due to negative results with the 2009 H1N1 reagent). The same samples were tested by RT-PCR using subtype-specific primers, which resulted in 30 being classified as H3N2, nine as seasonal H1N1, and 25 as 2009 H1N1, a 100% match to the IFA subtyping results. One additional sample tested negative with the D3 Ultra influenza A virus ID reagent but was positive with the pan-H1 ID reagent (only three fluorescent cells were observed). This sample was confirmed to contain seasonal H1N1 virus by RT-PCR (Data not shown). Virus recovery was not successful with 10 specimens.
DISCUSSION
The high degree of genetic and antigenic variability of influenza A virus renders the development of MAbs that can recognize all the antigenically diverse epitopes of a particular influenza A virus subtype a challenging task. Our strategy was to immunize mice separately with multiple inactivated influenza A virus strains that are representative of recent and past circulating subtypes and to select subtype-specific MAbs based on results from several rounds of IFA screening with numerous seasonal H3N2, H1N1, and 2009 H1N1 laboratory strains and clinical isolates. Using this approach, we successfully produced 11 MAbs that displayed subtype-specific reactivity, and we blended these MAbs to make three influenza A virus subtype-specific reagents: an H3N2-specific reagent; a 2009 H1N1-specific reagent, and a pan-H1N1-specific reagent that detects both seasonal H1 and 2009 H1 subtypes. Considering the possibility that swine-derived H1N1 virus strains may circulate in future influenza seasons, we included one MAb (9F11G7) that recognizes both the California and Mexico 2009 H1N1 strains and the swine-derived A/New Jersey/8/76 strain in the influenza A virus pan-H1 ID reagent along with four other H1N1 subtype specific MAbs (Table 1). When these three reagents were tested on 21 prototype influenza A virus strains, their analytic performance proved to be excellent (Table 2; Fig. 1). In addition, the reagents did not cross-react with any of 40 non-influenza A viruses (Table 3).
We hypothesized that the target antigen of each subtyping MAb was hemagglutinin (HA), and thus we analyzed the MAbs by Western blotting with lysates from the viruses used for immunization. The majority of these influenza A virus subtyping MAbs did not recognize the denatured HA protein in this assay (data not shown). To determine if the MAbs recognized conformational epitopes on the target protein, we decided to test each MAb in a hemagglutination inhibition assay (HIA). The results showed that every subtyping MAb was able to inhibit the HA activity of its target virus. It is interesting that the three H3N2 subtyping MAbs have much lower HI titers than the H1N1 subtyping MAbs (Table 1). These three MAbs were generated using a single H3N2 strain (A/Shangdong/9/93) as the antigen, but each MAb is able to recognize seven H3N2 prototype strains with high specificity by IFA. Since HI measures the functional activity of the anti-HA antibodies and correlates well with antibody neutralizing activity (5), this could be an indication that the H3N2 subtyping MAbs recognize a more conserved region of the HA than the H1N1 MAbs.
There have been few studies published on the utility of immunoassays for subtyping influenza A virus. One recently published report (11) described the evaluation of numerous influenza A virus subtype-specific MAbs by IFA using archived respiratory specimens collected before 2007. Even though the specificities of the H1N1 and H3N2 MAbs were high, the data covered neither the specimens collected in the last 4 years nor the 2009 H1N1 virus. The study also showed that the H1N1- and H3N2-specific MAbs had high specificity for subtyping but poor sensitivity for detecting influenza A virus, especially when specimens were tested directly (11). Therefore, to avoid the problem of the sensitivity of our IFA test being negatively affected by the use of archived specimens which had been through freeze-thaw cycles, the clinical study was designed with the purpose of subtyping influenza A virus that already had been confirmed to be positive for influenza A virus with the D3 Ultra influenza A virus ID reagent after viral culture. Confirmation of the subtyping results was done by RT-PCR. Among 75 archived influenza A virus-positive clinical samples, 64 were successfully recovered from culture and subtyped. One additional sample that stained positive with the pan-H1 reagent showed a negative result with the D3 Ultra Influenza A virus ID reagent, which should stain all subtypes of influenza A viruses. We consider this to be a sample error caused by the small amount of virus in the specimen, since only three fluorescent cells were observed when cells were stained with the pan-H1 ID reagent. Furthermore, RT-PCR results showed that this sample contained seasonal H1N1 virus. In addition, two samples were identified by RT-PCR as containing the 2009 H1N1 virus but failed to show positive stains with either the D3 Ultra influenza A virus ID reagent or the subtyping reagents. This could be due to the presence in these samples of nonviable virus or possibly to the higher sensitivity of RT-PCR.
The excellent specificity exhibited by the MAbs on cultured clinical influenza A virus-positive specimens collected between 2006 and 2009 strongly supports the utility of these reagents as a subtyping tool for laboratories that lack molecular testing capability. Further studies are necessary to broaden the application of these subtyping reagents for direct specimens. As a cautionary note, it should be emphasized that these IFA subtyping reagents are focused on the subtypes of influenza A virus that are presently circulating among humans, and they are not intended for other subtyping applications, such as surveillance in nonhuman species. Finally, a human specimen that is influenza virus positive but is negative in the subtyping IFA should be designated nonsubtypeable. Such a negative result could occur because of a small amount of virus, the presence of another subtype in humans (e.g., avian H5 or H9), or, as occurred in 2009, a new variant of H1 or H3. Due to the possibility of polymorphisms in the HA generated during antigenic drift and of new variants as a result of antigenic shift, it will be necessary to monitor the performance of each influenza A virus subtyping reagent regularly to ensure continued recognition of circulating influenza A virus strains. Additional MAbs can be made to address any new variants that arise. Despite these caveats, these new reagents will be useful tools for both influenza diagnosis and surveillance.
ACKNOWLEDGMENTS
This study supported in part by a grant from Institute of Clinical Research of University Hospitals Case Medical Center.
We thank Ferda Yantiri-Wernimont of Quidel for providing us with H3 specific MAb 5D2.
FOOTNOTES
- Received 21 June 2011.
- Returned for modification 23 August 2011.
- Accepted 31 October 2011.
- Accepted manuscript posted online 9 November 2011.
- Copyright © 2012, American Society for Microbiology. All Rights Reserved.
