Characterization of lentiviral pseudotypes with influenza H5N1 hemagglutinin and their performance in neutralization assays

Journal of Virological Methods 165 (2010) 305–310

Contents lists available at ScienceDirect

Journal of Virological Methods journal homepage: www.elsevier.com/locate/jviromet

Short communication

Characterization of lentiviral pseudotypes with influenza H5N1 hemagglutinin and their performance in neutralization assays夽 Wei Wang, Hang Xie, Zhiping Ye, Russell Vassell, Carol D. Weiss ∗ Center for Biologics Evaluation and Research, US Food and Drug Administration, 29 Lincoln Drive, Bethesda, MD 20892, United States

a b s t r a c t Article history: Received 2 December 2009 Received in revised form 1 February 2010 Accepted 4 February 2010 Available online 11 February 2010 Keywords: Hemagglutinin Influenza Neutralization Pseudotype

Pseudotype reporter viruses are being used as safe, quantitative, and high-throughput tools for assessing antibody neutralization for many viruses, including influenza. However, characterization of pseudotypes containing influenza hemagglutinin (HA-pseudotypes) is needed before this system is widely adopted for evaluating neutralizing antibodies in sera following vaccination or infection. In this report HA-pseudotype stocks were analyzed for HA content, stability, and performance in neutralization assays under various conditions. HA-pseudotypes produced with HA genes of H5 strains representing clades 1, 2.2, and 2.3.4 consistently contain similar HA contents, and infectivity was not greatly affected by the purity of the HA-pseudotype preparations or variations in storage conditions. HA-pseudotype neutralization titers using a reference serum panel were also consistent across a wide range of dilutions of HA-pseudotype stocks and correlated well with results from microneutralization assays involving replicating influenza. Concentrated HA-pseudotypes were further shown to work well in hemagglutination inhibition assays. Finally, antisera elicited by genetically modified HA, with changes in the polybasic cleavage site that have been used in some H5 vaccines and reduce pathogenicity, gave identical neutralization titers against HA-pseudotypes with wild type or modified HA. These findings support continued development of HApseudotypes as a robust tool for analyzing sera in vaccine and serologic studies. Published by Elsevier B.V.

Influenza virus surface glycoprotein hemagglutinin (HA) mediates virus entry and is the most important target of antibodymediated protection (reviewed in (Skehel and Wiley, 2000)). During infection and vaccination, HA elicits neutralizing and hemagglutination inhibiting antibodies that protect against infection and aid selection of vaccine strains. The classical methods for measuring protective antibody responses are hemagglutination inhibition (HI) and virus microneutralization (MN). However, the use of live viruses in these assays requires virus isolation and propagation, which require high-level biocontainment facilities when highly pathogenic strains, such as H5 and H7, are used. Because pseudotypes infect

Abbreviations: HA, hemagglutinin; NA, neuraminidase; HI, hemagglutination inhibition; MN, microneutralization; PV-MN, HA-pseudotype microneutralization; Vietnam, A/Vietnam/1203/2004; Laos, A/duck/Laos/3295/2006; Anhui, A/Anhui/1/2005; Turkey, A/turkey/Turkey/1/2005; Luc, luciferase; HAT, human airway trypsin-like protease; NIH, National Institutes of Health; FDA, Food and Drug Administration; CDC, Centers for Disease Control and Prevention. 夽 The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Center for Biologics Evaluation and Research. ∗ Corresponding author at: FDA/CBER, HFM-466, Bldg. 29, Room 532, 29 Lincoln Dr., Bethesda, MD 20892, United States. Tel.: +1 301 402 3190; fax: +1 301 496 1810. E-mail address: [email protected] (C.D. Weiss). 0166-0934/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.jviromet.2010.02.009

cells only once and lack infectious genomes, they cannot replicate and are therefore safe to use in most lab settings. In addition, HA genes used to make HA-pseudotypes can be easily manipulated without the need to reassort or grow virus to sufficient titers needed for HI or MN studies, facilitating assessments of how different HA genes or particular mutations affect neutralization. Recently, several groups have reported the use of HA-pseudotypes to detect and identify HA neutralizing antibody (Alberini et al., 2009; Kong et al., 2006; Nefkens et al., 2007; Oh et al., 2009; Temperton et al., 2007; Wang et al., 2008). Given the potential broad utility of HA-pseudotypes for evaluating neutralizing antibodies in a variety of settings, the performance of HA-pseudotypes under various experimental conditions was studied further. Studies focused on H5 subtype HA in part to take advantage of a reference serum panel consisting of human sera obtained after vaccination against Vietnam H5 HA that was used in a prior study to assess reproducibility of MN and HI assays across 15 labs (Stephenson et al., 2009). HA-pseudotypes were produced as described previously (Wang et al., 2008). To make different H5 HA-pseudotypes, full-length wild type HA ORF from A/Vietnam/1203/2004 (Vietnam) (GenBank EF541403) and A/duck/Laos/3295/2006 (Laos) (GenBank ABG67978) was amplified with PCR, and HA ORF of A/Anhui/1/2005 (Anhui) (GenBank ABD28180.1) and A/turkey/Turkey/1/2005 (Turkey)

306

W. Wang et al. / Journal of Virological Methods 165 (2010) 305–310

Table 1 Comparison of MN and PV-MN titers. Serum sample

Viruses Vietnam

Anhui

MN Titer (range) International standard 07/150 Sheep serum O P Postvaccination serum Low titer C D High titer A E F G H I L False positive serum K Negative serum M Prevaccination serum B N J

PV-MN titer

Turkey

MN titer (range)

PV-MN titer

MN titer (range)

PV-MN titer

518 (127–2032)

568

299 (80–806)

106

291 (52–1810)

188

216 (40–1280) 7317 (1280–40,960)

692 21,189

49 (BD-80) 3806 (640–20,480)

<80 3270

148 (10–1280) 732 (10–2560)

<80 2232

63 (BD-640) 19 (BD-320)

265 97

23 (BD-80) 12 (BD-80)

244 211

27 (BD-320) 15 (BD-160)

<80 <80

1389 (320–5120) 148 (40–640) 83 (10–640) 504 (80–2560) 130 (20–640) 379 (20–1280) 1453 (320–5120)

1593 700 440 1148 556 1352 2791

2893 (640–5120) 20 (BD-80) 35 (BD-80) 140 (40–640) 34 (BD-80) 161 (20–640) 3097 (1280–10,240)

1588 <80 406 761 231 908 1387

1313 (160–5120) 87 (10–640) 18 (BD-80) 274 (10–1280) 91 (10–640) 106 (BD-640) 1520 (320–5120)

901 <80 <80 209 <80 148 1033

52 (BD-320)

202

13 (BD-80)

<80

44 (BD-320)

<80

8

<80

7

<80

7

<80

12 (BD-320) 14 (BD-160) 19 (BD-320)

<80 <80 <80

10 (BD-40) 10 (BD-80) 11 (BD-80)

<80 <80 <80

12 (BD-80) 13 (BD-80) 18 (BD-320)

<80 <80 <80

PV-MN titers (IC95) of Vietnam HA antisera from a sera reference panel for standardizing MN assay were evaluated with Vietnam, Anhui and Turkey HA-pseudotypes. Data reported were from at least duplicate testing of serum samples. PV-MN titers were compared with average MN titers reported previously (Stephenson et al., 2009). BD: below detection limit.

(GenBank ABQ58921.1) was chemically synthesized by using human-preferred codons by GenScript (Piscataway, NJ). These HA ORFs were then placed into CMV/R 8␬B expression plasmid obtained from Dr. Gary J. Nabel (NIH, Bethesda, MD). In addition, the wild type polybasic amino acid sequences RERRRKKR in Vietnam HA and RERRRKR in Laos HA, which contribute to pathogenicity, were replaced with the low pathogenic RETR amino acid sequences that have been used in vaccine strains and are referred to as Vietnam–RETR and Laos–RETR, respectively. Human airway trypsin-like protease gene (HAT) (GenBank NP 004253) was synthesized by using human-preferred codons by GenScript (Piscataway, NJ), and cloned into the pCAGGS expression plasmid (Niwa et al., 1991). This HAT plasmid performed similarly to the

HAT plasmid (Bottcher et al., 2006) used previously (Wang et al., 2008). The following sera were used in this study: (1) hyper-immune sheep Vietnam HA antiserum produced via immunization with bromelain-cleaved HA from A/Vietnam/1203/2004 virus (prepared by FDA); (2) ferret Laos HA antiserum generated from infection with A/duck/Laos/3295/2006 virus (prepared by FDA); (3) ferret Vietnam antiserum generated from infection with A/Vietnam/1203/2004 virus (prepared by FDA); (4) rabbit H5 HA1 antiserum produced via immunization with HA1 protein from A/Vietnam/1203/2004 (eENZYME, Montgomery Village, MD) at FDA; and (5) a reference serum panel (provided by Iain Stephenson, University of Leicester, UK) that includes human and sheep

Fig. 1. Impact of the amount of HA-pseudotypes on infectivity and neutralization. (A) Infectivity in 293T cells inoculated with different amount of Vietnam HA-pseudotypes containing HA from 0.08 to 8 ng and p24 from 0.111 to 11.1 ng (1:500 to 1:5 dilution). Results are expressed as the mean ± S.D. of three HA-pseudotype stocks. Neutralization titers of (B) sheep Vietnam HA antiserum and (C) human Vietnam H5N1 antiserum (sample C from Table 1) with 50% (IC50), 90% (IC90) and 95% (IC95) inhibition of serially diluted Vietnam HA-pseudotypes (in (A)). Results are expressed as the mean ± S.D. of experiments of three HA-pseudotype stocks.

W. Wang et al. / Journal of Virological Methods 165 (2010) 305–310

307

Fig. 2. Impact of impurities on HA-pseudotype infectivity and neutralization. (A) Western blot of supernatants containing Vietnam HA-pseudotypes or Vietnam HA alone, probed with rabbit H5 HA1 antiserum. (B) Infectivity of Vietnam HA-pseudotypes with or without supernatants containing an equivalent amount of HA alone. Results are expressed as the mean ± S.D. of three HA-pseudotype stocks. (C) Neutralization of Vietnam HA-pseudotypes by 2-fold, serially diluted ferret Vietnam HA antiserum, in the presence or absence of an equivalent of HA alone (+HA protein). Data reported are the mean ± S.D. of three HA-pseudotype stocks. (D) Western blot of Vietnam HA-pseudotypes produced in the presence or absence of fetal calf serum and probed with rabbit H5 HA1 antiserum (left). Infectivity of Vietnam HA-pseudotypes with or without fetal calf serum (right). Results are expressed as the mean ± S.D. of three HA-pseudotypes stocks.

Vietnam H5N1 vaccine antisera (Table 1) (Stephenson et al., 2009). H5 antigens used to generate the human, ferret, and sheep antisera that were used in the neutralization assays contained the modified RETR-HA cleavage site. The rabbit serum, specific for the HA1 subunit of HA, was used for Western blots. The amount of HA in pseudotype stocks that would be appropriate for neutralization studies was evaluated first. HA and gag in pseudotypes were measured by immunoblot analysis and p24 ELISA, respectively, as described previously (Wang et al., 2008). Using a constant ratio of plasmids for transfections (5 ␮g gag, 5.5 ␮g Luc, 0.5 ␮g HA), HA-pseudotype stocks for several H5 strains were generated. These HA-pseudotypes consistently contained around 500 ng/ml of p24 gag and 400 ng/ml of HA, with high infectivity that measured around 105 RLU/ml. HA-pseudotype stocks containing 400 ng/ml HA were diluted 5–500-fold, and 0.1 ml pseudotype was inoculated onto 96-well plates that were seeded with 2 × 104 293T cells/well 1 day prior to infection. Luciferase activity was measured 48 h later and showed a proportional decrease of infectivity

(Fig. 1A). When the HA-pseudotype stocks were diluted serially and used in the neutralization assay, neutralization titers of antisera (including 50, 90, and 95% inhibitory concentrations (IC50, IC90 and IC95)) were not significantly affected using either high titer or low titer sera (Fig. 1B and C, respectively). These findings are consistent with the percentage law of virus neutralization, in which neutralization reflects the fraction of the initial infectivity that survives antibody treatment over a wide range of virion concentrations, as long as antibody is in excess (Andrewes and Elford, 1933; Burnet et al., 1937; Klasse and Sattentau, 2002). Thus, HApseudotype stocks containing 0.8–80 ng of HA per ml are suitable for use in neutralization assays. Previously it was reported that HA can bud off transfected cells independently of influenza matrix proteins and be released into the supernatant (Chen et al., 2007). Therefore, it was important to assess whether HA that is released into culture supernatants without being incorporated into the retroviral pseudotypes could compete with HA-pseudotypes to interfere with infectivity or neu-

Table 2 Comparison of HI titer. Sera

Anti-Vietnam H5N1 Anti-Laos H5N1

HI Titer Vietnam live virus

Vietnam HA-pseudotype

Laos live virus

Laos HA-pseudotype

1:40 ND

1:40 ND

1:5 1:320

1:5 1:320

For pseudotype HI assay, HA-pseudotype of Vietnam and Laos H5N1 were tested against individual anti-HA serum. For live virus HI assay, anti-Vietnam H5 HA serum was tested at FDA, and the HI titers of anti-Laos H5 HA serum was provided by CDC. ND: not done.

308

W. Wang et al. / Journal of Virological Methods 165 (2010) 305–310

Fig. 3. Stability of HA-pseudotypes. (A) Infectivity and HA content of Vietnam HA-pseudotypes stocks after several cycles of freezing at 80 ◦ C and thawing to room temperature (left) or prolonged storage at 4 ◦ C for the defined number of days (right). HA protein stability was detected by Western blot using rabbit Vietnam HA1 antiserum (top panels). Infectivity is expressed as the mean ± S.D. of three HA-pseudotype stocks run in triplicate (bottom panels). (B) Fresh Vietnam HA-pseudotypes and Vietnam HA-pseudotypes stored at 4 ◦ C for 14 days were tested against ferret Vietnam HA antiserum (2-fold serially dilutions). Vietnam HA-pseudotypes use the wild type A/Vietnam/1203/2004 sequences. There was no difference in stability between wild type Vietnam and modified Vietnam pseudotypes that have the RETR cleavage sequence (data not shown). Data reported are expressed as the mean ± S.D. from triplicate testing of serum samples. Con represents the control culture without pseudotype infection.

tralization. For this analysis, culture supernatants containing HA without pseudotypes (free HA) were generated and mixed with supernatants containing HA-pseudotype stocks (Fig. 2A) at a ratio that provided an equal amount of HA from both supernatants. Despite the presence of a substantial amount of free HA, there was no effect on HA-pseudotype infectivity (Fig. 2B) or neutralization titers involving serial dilutions of antiserum (Fig. 2C), consistent with the observations described above indicating that neutralization titers are not affected by virus input over a wide range. The presence of fetal calf serum during production of HA-pseudotype stocks also did not impair HA maturation or HA-pseudotype infectivity (Fig. 2D) and neutralization (data not shown). Thus, HA-pseudotypes from transfected culture supernatants without purification can be directly used for neutralization assays. HA-pseudotypes were also tested for suitability in hemagglutination inhibition (HI) assays. Briefly, chicken red blood cells from CBT Farms (Chestertown, MD) were used in an HI assay that was performed according to procedures in the WHO Animal Influenza Training Manual (WHO/CDS/CSR/NCS/2002.5). In this application, HA-pseudotypes direct from transfected culture supernatants did not have enough HA units to agglutinate red blood cells. However, this problem was easily overcome by concentrating the

HA-pseudotype stocks by centrifugation at 20,000 × g for 2 h in a microcentrifuge and resuspending the pellet in phosphate buffered saline. When tested in this manner using sheep Vietnam and ferret Laos H5 HA antisera, the HI titers matched the standard assay using live virus (Table 2). The HI titer of Laos HA antiserum using horse (Lampire Biologicals, Pipersville, PA) and chicken red blood cells were also similar, but the HI titer of Vietnam HA antiserum was very low when chicken red blood cells were used. Previous studies suggested that HI titer for H5 subtype viruses are relatively low when turkey or horse red blood cells were not used (Stephenson et al., 2004; WHO, 2003). These results suggest that the optimal type of red blood cells used in HI assays for H5 subtype HA may vary depending on the strain. The stability of HA-pseudotypes under various storage conditions was studied next. HA-pseudotypes corresponding to wild type A/Vietnam/1203/2004 were stored for different time periods at 4 ◦ C (Fig. 3A, right panel) or subjected to repeated freeze–thaw cycles (Fig. 3A, left panel), and then compared for infectivity. After four cycles alternating between −80 ◦ C and room temperature over a period of 3 h, infectivity of HA-pseudotypes decreased only slightly (Fig. 3A). When HA-pseudotypes were stored at 4 ◦ C, the slow decrease of infectivity was also observed (Fig. 3A). The gradual loss

W. Wang et al. / Journal of Virological Methods 165 (2010) 305–310

309

Fig. 4. Correlation between MN and PV-MN titers. PV-MN titers of Vietnam HA antisera from an NIBSC study panel were measured with Vietnam, Anhui and Turkey HA-pseudotypes and reported as the titers with 95% inhibition (IC95) of HApseudotypes infection of 293T cells. The PV-MN titers were compared with average MN titers (Stephenson et al., 2009).

Fig. 5. Impact of HA protease cleavage sequence on HA-pseudotype neutralization. (A) Infectivity of Vietnam and Laos HA-pseudotypes with wild type or modified HA containing RETR cleavage site. Results are expressed as the mean ± S.D. of three HA-pseudotype stocks. (B) The neutralization of Vietnam (left) or Laos (right) HApseudotypes with the wild type or RETR cleavage site by 2-fold, serially diluted ferret Vietnam and ferret Laos antiserum. Data reported were expressed as the mean ± S.D. of three HA-pseudotypes stocks.

of infectivity of HA-pseudotypes is probably due to pseudotype particle instability and aggregation rather than protein degradation, because HA protein did not show any change by Western blot (Fig. 3A). Significantly, fresh HA-pseudotypes showed the same neutralization pattern as HA-pseudotypes after storage for 14 days at 4 ◦ C (Fig. 3B). The infectivity of 2-year-old HA-pseudotype stocks stored at −80 ◦ C were still similar to that of their fresh stocks (data not shown). Thus, the stability of HA-pseudotypes is advantageous for neutralization assessments under various laboratory conditions. MN titers involving replicating virus were then compared with HA-pseudotype neutralization (PV-MN) titers using a reference serum panel that included human sera from subjects who had received vaccine derived from reassortant clade 1 virus A/Vietnam/1194/2004 or A/Vietnam/1203/2004. This serum panel was previously used for assessing reproducibility of serologic assays for H5N1 influenza viruses in a study of 15 laboratories from 9 countries (Stephenson et al., 2009). For neutralization assays with HA-pseudotypes, various end points have been used (Alberini et al., 2009; Oh et al., 2009; Temperton et al., 2007; Wang et al., 2008). To determine an end point for PV-MN that correlates best with MN, PVMN titers of all sera in the reference serum panel were calculated in a blinded manner using three different end points. After decoding the serum samples, the 95% inhibitory concentration (IC95) for PVMN using the established WHO standard antiserum 07/150 against Vietnam H5 HA most closely matched the corresponding MN titer

reported earlier (Stephenson et al., 2009) (Table 1). IC95 titers of PVMN of all other sera against Vietnam H5 HA in the study panel was also in good agreement with average MN titers in the previously reported study (Stephenson et al., 2009) (Table 1 and Fig. 4). The serum cross-neutralization to Anhui and Turkey H5 HA detected by MN reported in the Stephenson study (Stephenson et al., 2009) was also demonstrated in the PV-MN IC95 titers (Table 1). Compared to the titers against Vietnam and Anhui HA, the average MN titers and PV-MN IC95 titers against Turkey HA correlated less well (Fig. 4), but the PV-MN IC95 titers against Turkey HA still agreed with MN titers (Table 1). Therefore, IC95 neutralization titers from the PV-MN assay may be optimal for comparing with MN assays. Finally, whether changes in the polybasic amino acid sequence in the highly pathogenic H5 strains would affect antigenicity was addressed by mutating the HA gene in the pseudotype system. The polybasic amino acid sequence of highly pathogenic H5 strains is typically modified to a sequence that reduces viral pathogenicity, such as RETR, which has been suggested for some vaccine strains (Li et al., 1999; Subbarao et al., 2003). Such modifications provide extra safety for vaccines and allow the viruses to be used under less stringent biosafety conditions. To address the issue of whether such a change in the polybasic cleavage site affects infectivity or neutralization, neutralization of the HA-pseudotypes containing wild type polybasic amino acid sequence to those containing the RETR sequence was compared. In these studies, there was no difference in HA-pseudotype infectivity between wild type- and RETR-HA-

310

W. Wang et al. / Journal of Virological Methods 165 (2010) 305–310

pseudotypes (Fig. 5A). The HA-pseudotypes with wild type and RETR-HA sequences also gave the same neutralization titers for each antiserum tested (Fig. 5B). These results are consistent with the location of major neutralizing epitopes being located at the membrane-distal heads of HA (Caton et al., 1982; Wiley and Skehel, 1987) and support the use of H5 HA with modified cleavage sites as vaccine strains and in neutralization studies. It may also be of interest to assess whether neutralizing antibodies targeting the HA2 transmembrane subunit of HA (Ekiert et al., 2009) could be affected by cleavage site mutations. In conclusion, HA-pseudotypes use convenient molecular biological methods for efficient construction of specific HAs and provide a robust and reliable tool for serological evaluation of influenza virus neutralization in MN and HI applications. The HApseudotype system not only complements the classical methods for assessing protective antibodies in sera, it allows assessments when classical methods cannot be done due to biosafety considerations or poor growth of virus. Conflicts of interest Authors do not have commercial or other associations that might pose a conflict of interest. Acknowledgements We thank the following persons for generously supplying key reagents for these studies: Dr. Gary Nabel (NIH, Bethesda, MD) for HIV-luciferase pseudotype system and expression vector CMV/R 8␬B; Galina Vodieko and Christine Anderson (FDA, Bethesda, MD) for H5 reference sheep antisera; Dr. Olga Zoueva (FDA, Bethesda, MD) for providing the plasmid for A/duck/Laos/3295/2006 HA gene; and Dr. Iain Stephenson (University of Leicester, UK) for the H5N1 reference serum panel. We also thank Dr. Xiyan Xu (CDC, Atlanta, GA) for providing HI titer of ferret anti-Laos H5N1 sera and Drs. Vladimir Lugovtsev and Matthew Sandbulte (FDA, Bethesda, MD) for critical reading of this manuscript. This work was support by institutional research funds from the FDA. Financial support: US Food and Drug Administration. References Alberini, I., Del Tordello, E., Fasolo, A., Temperton, N.J., Galli, G., Gentile, C., Montomoli, E., Hilbert, A.K., Banzhoff, A., Del Giudice, G., Donnelly, J.J., Rappuoli, R., Capecchi, B., 2009. Pseudoparticle neutralization is a reliable assay to measure immunity and cross-reactivity to H5N1 influenza viruses. Vaccine 27, 5998–6003.

Andrewes, C.H., Elford, W.J., 1933. Observations on anti-phage sera. I. “The percentage law”. Br. J. Exp. Pathol. 14, 367–376. Bottcher, E., Matrosovich, T., Beyerle, M., Klenk, H.D., Garten, W., Matrosovich, M., 2006. Proteolytic activation of influenza viruses by serine proteases TMPRSS2 and HAT from human airway epithelium. J. Virol. 80, 9896–9898. Burnet, F.M., Keogh, E.V., Lush, D., 1937. The immunological reactions of the filterable viruses. Aust. J. Exp. Biol. Med. Sci. 15, 227–368. Caton, A.J., Brownlee, G.G., Yewdell, J.W., Gerhard, W., 1982. The antigenic structure of the influenza virus A/PR/8/34 hemagglutinin (H1 subtype). Cell 31, 417–427. Chen, B.J., Leser, G.P., Morita, E., Lamb, R.A., 2007. Influenza virus hemagglutinin and neuraminidase, but not the matrix protein, are required for assembly and budding of plasmid-derived virus-like particles. J. Virol. 81, 7111–7123. Ekiert, D.C., Bhabha, G., Elsliger, M.A., Friesen, R.H., Jongeneelen, M., Throsby, M., Goudsmit, J., Wilson, I.A., 2009. Antibody recognition of a highly conserved influenza virus epitope. Science 324, 246–251. Klasse, P.J., Sattentau, Q.J., 2002. Occupancy and mechanism in antibody-mediated neutralization of animal viruses. J. Gen. Virol. 83, 2091–2108. Kong, W.P., Hood, C., Yang, Z.Y., Wei, C.J., Xu, L., Garcia-Sastre, A., Tumpey, T.M., Nabel, G.J., 2006. Protective immunity to lethal challenge of the 1918 pandemic influenza virus by vaccination. Proc. Natl. Acad. Sci. U.S.A. 103, 15987–15991. Li, S., Liu, C., Klimov, A., Subbarao, K., Perdue, M.L., Mo, D., Ji, Y., Woods, L., Hietala, S., Bryant, M., 1999. Recombinant influenza A virus vaccines for the pathogenic human A/Hong Kong/97 (H5N1) viruses. J. Infect. Dis. 179, 1132–1138. Nefkens, I., Garcia, J.M., Ling, C.S., Lagarde, N., Nicholls, J., Tang, D.J., Peiris, M., Buchy, P., Altmeyer, R., 2007. Hemagglutinin pseudotyped lentiviral particles: characterization of a new method for avian H5N1 influenza sero-diagnosis. J. Clin. Virol. 39, 27–33. Niwa, H., Yamamura, K., Miyazaki, J., 1991. Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108, 193–199. Oh, S., Selleck, P., Temperton, N.J., Chan, P.K., Capecchi, B., Manavis, J., Higgins, G., Burrell, C.J., Kok, T., 2009. Neutralizing monoclonal antibodies to different clades of influenza A H5N1 viruses. J. Virol. Methods 157, 161–167. Skehel, J.J., Wiley, D.C., 2000. Receptor binding and membrane fusion in virus entry: the influenza hemagglutinin. Annu. Rev. Biochem. 69, 531–569. Stephenson, I., Heath, A., Major, D., Newman, R.W., Hoschler, K., Junzi, W., Katz, J.M., Weir, J.P., Zambon, M.C., Wood, J.M., 2009. Reproducibility of serologic assays for influenza virus A (H5N1). Emerg. Infect. Dis. 15, 1252–1259. Stephenson, I., Wood, J.M., Nicholson, K.G., Charlett, A., Zambon, M.C., 2004. Detection of anti-H5 responses in human sera by HI using horse erythrocytes following MF59-adjuvanted influenza A/Duck/Singapore/97 vaccine. Virus Res. 103, 91–95. Subbarao, K., Chen, H., Swayne, D., Mingay, L., Fodor, E., Brownlee, G., Xu, X., Lu, X., Katz, J., Cox, N., Matsuoka, Y., 2003. Evaluation of a genetically modified reassortant H5N1 influenza A virus vaccine candidate generated by plasmid-based reverse genetics. Virology 305, 192–200. Temperton, N.J., Hoschler, K., Major, D., Nicolson, C., Manvell, R., Hien, V.M., Ha, D.Q., Jong, M.D., Zambon, M., Takeuchi, Y., Weiss, R.A., 2007. A sensitive retroviral pseudotype assay for influenza H5N1-neutralizing antibodies. Influenza Other Respir. Viruses, doi:10.1111/j.1750-2659.2007.00016.x. Wang, W., Butler, E.N., Veguilla, V., Vassell, R., Thomas, J.T., Moos Jr., M., Ye, Z., Hancock, K., Weiss, C.D., 2008. Establishment of retroviral pseudotypes with influenza hemagglutinins from H1, H3, and H5 subtypes for sensitive and specific detection of neutralizing antibodies. J. Virol. Methods 153, 111–119. WHO, 2003. Assays for neutralizing antibody to influenza viruses. Report of an informal scientific workshop, Dresden, 18-19 March 2003. Wkly. Epidemiol. Rec. 78, 290–293. Wiley, D.C., Skehel, J.J., 1987. The structure and function of the hemagglutinin membrane glycoprotein of influenza virus. Annu. Rev. Biochem. 56, 365–394.