|Home | About | Journals | Submit | Contact Us | Français|
Influenza virus is a highly contagious virus that causes yearly epidemics and occasional pandemics of great consequence. Influenza virus neutralizing antibodies (NAbs) are promising prophylactic and therapeutic reagents. Detection of NAbs in serum samples is critical to evaluate the prevalence and spread of new virus strains. Here we describe the development of a simple, sensitive, specific, and safe screening assay for the rapid detection of NAbs against highly pathogenic influenza viruses under biosafety level 2 (BSL-2) conditions. This assay is based on the use of influenza viruses in which the hemagglutinin (HA) gene is replaced by a gene expressing green fluorescent protein (GFP). These GFP-expressing influenza viruses replicate to high titers in HA-expressing cell lines, but in non-HA-expressing cells, their replication is restricted to a single cycle.
Influenza virus is a significant public health problem, causing an estimate of 200,000 hospitalizations and about 36,000 deaths yearly in the United States (19). Influenza virus evades population immunity by accumulating mutations in the hemagglutinin (HA) glycoprotein (antigenic drift) and allowing for reinfection of individuals with prior influenza immunity. On three occasions during the last century, a new influenza A virus subtype containing an antigenically unrelated HA (antigenic shift) appeared in humans, resulting in a global pandemic, as follows: the 1918-1919 “Spanish flu” (H1N1), the 1957-1958 “Asian flu” (H2N2), and the 1968-1969 “Hong Kong flu” (H3N2). The influenza pandemic of 1918 (H1N1) has been estimated to have caused approximately 50 million deaths worldwide and 500,000 deaths in the United States alone (4). These new HA subtypes were derived from influenza A virus strains present in animal reservoirs. Currently, only the H1N1 and H3N2 subtypes of influenza A viruses are circulating in humans. Concerns of a new human pandemic have been raised due to the widespread infections in poultry of highly pathogenic H5N1 viruses, which sporadically infect and cause severe disease in humans (9), and due to the recent outbreak in humans of a novel pandemic strain of antigenically distant H1N1 virus of swine origin with the capability of human-to-human transmission (3).
A simple assay for the detection of neutralizing antibodies (NAbs) against influenza viruses in humans and in animals will help to study the epidemiology and prevalence of these viruses and to evaluate humoral immunity of traditional and prototypic vaccines. Moreover, it will facilitate identification of monoclonal antibodies (MAbs) for potential use in passive immunization of people who might have been exposed to new influenza viruses as a result of an ongoing pandemic, an unfortunate laboratory accident, or an intentional bioterrorism release. Previous methods to detect the presence of influenza virus NAbs include hemagglutination inhibition (HI) and virus neutralization assays. It is generally accepted that virus neutralization assays are more sensitive and reliable than HI assays (14, 15). However, these laborious assays require manipulation of live viruses, sometimes under enhanced biosafety level 3 (BSL-3) containment conditions, and are subject to select agent clearance in the United States in the case of highly pathogenic influenza viruses. Pseudotyped retroviruses bearing influenza virus HA and neuraminidase (NA) glycoproteins represent a potential alternative (21). However, it remains to be determined whether pseudotyped retroviruses faithfully mimic all interactions required for influenza virus entry into its target cells. Here we show a new reporter gene-based assay for the detection and quantification of influenza virus NAbs that block influenza virus entry and replication.
In previous studies conducted to identify packaging signals within the influenza virus genes, a stable infectious influenza A/WSN/33 (WSN) virus with its HA gene replaced by a gene encoding green fluorescent protein (GFP) was generated (11). The viral GFP gene contained the GFP open reading frame (ORF) flanked by the replication, transcription, and packaging cis-acting sequences of the viral HA gene (Fig. 1A and B). This HA-deficient influenza virus can be passaged in a complementing cell line expressing HA of influenza A/WSN/33 virus (11). In order to complement the growth of the HA-deficient GFP-expressing influenza viruses with different subtypes of influenza HA, MDCK cells that constitutively express influenza virus HAs from the recent human H1 isolate influenza virus A/New Caledonia/1/99, the 1918 pandemic H1 virus A/Brevig Mission/1/18 (A/BM/1/18 or referred to as “1918”), or the highly pathogenic H5 virus A/Vietnam/1203/04 (16) were generated by cotransfection of the respective HA-expressing pCAGGs (13) plasmids with the hygromycin B-resistant vector pCB7 (3:1 ratio) using Lipofectamine 2000. Cells were cloned and screened for HA protein expression by immunofluorescence using specific HA monoclonal antibodies (Fig. (Fig.1C).1C). Using these stable cell lines, we generated HA-pseudotyped influenza viruses containing A/WSN/33, A/New Caledonia/1/99, A/BM/1/18, or A/Vietnam/1203/04 HAs (Fig. (Fig.1D1D).
When used to infect at a multiplicity of infection (MOI) of 0.001, the HA-deficient GFP-expressing influenza virus showed a GFP expression signal that increased in a time-dependent matter in the different MDCK-HA-expressing cell lines, indicative of their multicycle replication (Fig. (Fig.2A).2A). However, in the parental MDCK cells, the HA-deficient viruses showed only a few GFP-positive cells at early time points. To further confirm the productive viral replication of the HA-deficient GFP-expressing virus only in the MDCK-HA-expressing cell lines, virus titers in the supernatants were determined. As shown in Fig. Fig.2B,2B, wild-type WSN virus grew to similar titers in the parental and the HA-expressing MDCK cells. In contrast, the HA-deficient GFP-expressing virus replicated to high titers only in the HA-expressing cell lines and failed to replicate in wild-type non-HA-expressing MDCK cells. Similar results were obtained when replication of the virus was monitored by Western blotting against the viral protein NP and against GFP (Fig. (Fig.2C).2C). As expected, similar levels of GFP expression were reached when parental and MDCK-HA-expressing cells were infected at a high multiplicity of infection (MOI of 2) (Fig. (Fig.2D).2D). When traditional plaque assay methods were used (Fig. (Fig.2E),2E), the HA-pseudotyped GFP-expressing virus did not form plaques in MDCK parental cells but was able to form plaques in MDCK-HA-expressing cells. Overall, these results demonstrate multicycle replication and efficient GFP expression of the GFP-expressing influenza virus only in HA-expressing complementing MDCK cells.
We next characterized the morphology, RNA, and protein composition of HA-pseudotyped GFP-expressing virions. Analysis by polyacrylamide gel electrophoresis (PAGE) of viral RNA isolated from purified virions demonstrated the absence of a wild-type HA gene and the presence of a new viral RNA corresponding to the GFP-encoding gene. At the protein level, all pseudotyped viruses have a composition similar to that of the wild-type virus, with the exception of the levels of HA, which appear to be slightly reduced. Electron microscopy images of purified virions indicated that the HA-pseudotyped viruses have morphology and particle size similar to those of the parental WSN virus (data not shown).
In order to determine whether GFP expression by these HA-pseudotyped influenza viruses could be used as an indication of the neutralizing abilities of specific antibodies against HA, we investigated the neutralization capability of purified MAbs generated against A/BM/1/18 (Mount Sinai Hybridoma center facility) using a GFP neutralization assay based on A/BM/1/18 HA-pseudotyped virus (p1918). As illustrated in Fig. Fig.3A,3A, differences in the neutralization abilities of the A/BM/1/18 MAbs were observed. MAb 58F4 showed the most potent virus neutralization capacity. MAbs 5D3, 39E4, and 20D7 also neutralized p1918. Finally, the 9A2 and 12B12 MAbs did not show any neutralization capacity in the GFP neutralization assay. None of the A/BM/1/18 MAbs were able to inhibit viral infection by pWSN, demonstrating the specificity of this assay. To better understand the reactivity of these A/BM/1/18 MAbs, we performed immunofluorescence assays on parental, WSN HA-, and 1918 HA-expressing MDCK cells (Fig. (Fig.3B).3B). The two MAbs with no neutralization activity against p1918 (9A2 and 12B12) did not recognize viral 1918 HA in the stable cell line. To demonstrate the reproducibility of the assay, we measured the amount of GFP expression in triplicate samples using a microplate fluorescent reader. As shown in Fig. Fig.3C,3C, similar levels of neutralization for the 1918 HA-pseudotyped virus with the indicated MAbs were observed, while no neutralization capability for the A/WSN/33 HA-pseudotyped virus was observed (Fig. (Fig.3D3D).
To further validate the ability of the GFP neutralization assay to determine the MAb neutralization ability, we did a similar study using a panel of MAbs generated against A/Vietnam/1203/04 (Mount Sinai Hybridoma center facility) and the HA-pseudotyped (pViet) virus. As shown above, we detected differences in the neutralization abilities of the MAbs generated against A/Vietnam/1203/04 (Table (Table1).1). MAbs 4G4 and 1C10 showed the greatest neutralization capacities, with MAbs 6B4 and 3F6 being less effective at the same antibody concentrations. MAbs 22E7, 10C60, 23E6, 9C9, and 11F4 did not inhibit viral infection, as determined by GFP expression.
We also performed HI assays to compare the sensitivity and specificity with those of the GFP neutralization assay. Similar neutralization capabilities of the A/Vietnam/1203/04 and A/BM/1/18 MAbs were obtained with both assays (Table (Table1).1). These results demonstrate the feasibility of using the GFP neutralization assay to determine the neutralization abilities of MAbs. Importantly, these assays could be used to screen MAbs generated against highly pathogenic viruses under BSL-2 conditions.
We next wanted to determine if the GFP neutralization assay could be used for the detection of NAbs in human serum samples (Table (Table2).2). For that purpose, serum samples from women vaccinated with the 2006-2007 influenza trivalent inactivated vaccine were randomly selected and screened for the presence of A/New Caledonia/1/99 NAbs using the A/New Caledonia/1/99 HA-pseudotyped (pNC) virus. The 2006-2007 influenza vaccine contains the strains A/New Caledonia/1/99, A/Wisconsin/67/2005 (H3N2), and B/Malaysia/2506/2004. Undiluted and 1:2-fold serial dilutions of the human serum samples were incubated with the pNC virus prior to infection of HA-expressing MDCK cells. As shown in Table Table2,2, differences in the ability to neutralize viral infection were detected among 11 independent human serum samples. Similar inhibitory titers were obtained with the traditional HI assay, demonstrating the feasibility of this GFP neutralization assay to detect NAbs in human serum samples. In contrast, NAb titers against the 1933 H1N1 virus pWSN were almost negligible, demonstrating the specificity of the assay using substantially drifted H1N1 influenza virus strains.
In summary, replication of HA-deficient influenza viruses expressing GFP in complementing MDCK cell lines expressing specific HAs allowed us to establish a safe and sensitive assay to detect influenza virus subtype-specific NAbs, both monoclonal and polyclonal. It is therefore likely that this same system can be established for any HA subtype and strain of interest. This will facilitate measurements of antibody responses in vaccinated individuals, serological detection of exposed individuals, and screening of NAbs that could be used prophylactically. A few HA-specific MAbs with broad cross-neutralization activities have recently been described (5, 18, 20), and the use of HA-pseudotyped GFP-expressing influenza viruses should make possible rapid screenings for the identification of additional antibodies with such broad cross-neutralizing activities.
Although vaccination is the primary means for controlling influenza, antivirals provide an additional line of defense, particularly important for controlling a new rapidly spreading, potentially pandemic virus. Therapeutic options for the treatment and prevention of influenza infections include neuraminidase (NA) inhibitors (zanamivir and oseltamivir) and inhibitors for the viral M2 ion channel protein (amantadine and rimantadine). However, the emergence of drug-resistant influenza virus strains is an increasing problem (2, 6, 8). Therefore, there is a strong need for developing novel drugs against influenza viruses. Our approach is amenable to be optimized for high-throughput screening, and if translated to multiwell plates, mechanical readouts of wells can be used to monitor the suppression of GFP expression by small molecule compounds under BSL-2 conditions. In these assays, HA-pseudotyped influenza viruses have some intrinsic advantages over more traditional HA-pseudotyped retrovirus (10, 17) or lentivirus (1, 7, 12) systems in that (i) the virus particles are bona-fide influenza virus particles and (ii) the HA-pseudotyped influenza viruses undergo full-cycle replication in HA-expressing cells, making this system suitable for the screening of not only viral entry inhibitors but also postentry inhibitors, including influenza viral budding inhibitors.
We thank Erin Petersen at the Mount Sinai Hybridoma Center for the generation of the 1918 and H5 MAbs. We express our gratitude to Qinshan Gao and Glenn A. Marsh for technical advice and Osvaldo Martinez for technical assistance with the electron microscopy studies.
This work was supported in part by grants from the NIH, grant P01 AI058113 to A.G.-S., C.F.B., and P.P., grant U19 AI62623 (Center for Investigating Viral Immunity and Antagonism) to A.G.-S. and T.M.M., and grant U54 AI57158 (North East Biodefense Center) to A.G.-S. and P.P., from CRIP (Center for Research on Influenza Pathogenesis, NIAID contract HHSN266200700010C) to A.G.-S. and P.P., and from NIAID Special Populations (contract NO1-AI 50028, Immune Response to Virus Infection during Pregnancy) to T.M.M.
Published ahead of print on 25 November 2009.