Please cite this article in press as: Grund C, et al. Highly pathogenic avian influenza virus H5N1 from Egypt escapes vaccine-induced immunity but confers clinical protection against a heterologous clade 2.2.1 Egyptian isolate. Vaccine (2011), doi:10.1016/j.vaccine.2011.01.006 G Model JVAC 11350 1���7 Vaccine xxx (2011) xxx���xxx 1 Contents lists available at ScienceDirect Vaccine j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / v a c c i n e Highly pathogenic avian influenza virus H5N1 from Egypt escapes Q1 vaccine-induced immunity but confers clinical protection against a heterologous clade 2.2.1 Egyptian isolate 1 2 3 Christian Grund a,1, El-Sayed M. Abdelwhab a,b,c,1, Abdel-Satar Arafa c, Mario Ziller a, Mohamed K. Hassan c, Mona M. Aly c, Hafez M. Hafez b, Timm C. Harder a, Martin Beer a,��� 4 5 6 a Friedrich-Loeffler Institute, Greifswald-Insel Riems, Germany 7 b Instiute for Poultry Diseases, Free University Berlin, Germany 8 c National laboratory for veterinary quality control on poultry production, Animal Health Research Institute, Dokki, Giza, Egypt 9 10 a r t i c l e i n f o 11 12 Article history: 13 Received 3 November 2010 14 Received in revised form 27 December 2010 15 16 Accepted 5 January 2011 17 Available online xxx 18 Keywords: 19 Highly pathogenic avian influenza 20 H5N1 21 Antigenic drift 22 Vaccination 23 Egypt 24 a b s t r a c t The poultry populations of Egypt are endemically infected by highly pathogenic avian influenza viruses (HPAIV) of subtype H5N1. Vaccination was chosen as an auxiliary tool to control HPAIV in poultry. Potency of commercial vaccines regarding emerging variants is under discussion. In the current study efficacy of four different inactivated whole H5 virus vaccines representing different sublineages of HPAIV H5N1 were tested in chickens against challenge viruses currently co-circulating in Egypt and representing two antigenically widely distinct HPAIV H5N1 lineages, i.e., ���variant��� (clade 2.2.1var) and ���proper��� (2.2.1pro) viruses. All vaccines induced clinical protection against challenge with classic 2.2.1pro Egyptian strains. In contrast, when challenged with a variant strain, only chickens vaccinated with the homologous Egyptian clade 2.2.1var virus or an inactivated re-assorted H5N1 strain (Re-5, clade 2.3) were protected. However, only the homologous virus induced sterile immunity whereas chickens clinically protected after Re- 5 vaccination shed virus at day two after infection indistinguishable to H5N2 vaccines. In conclusion, monitoring vaccine-driven evolution of HPAIV H5N1 by surveillance, antigenic characterization, and challenge studies is essential to assess efficacy of AIV vaccination campaigns. �� 2011 Published by Elsevier Ltd. 1. Introduction 25 Highly pathogenic avian influenza viruses (HPAIV) pose a severe 26 threat to poultry industry worldwide. Due to a certain zoonotic 27 potential of these viruses human populations may be at risk [1]. 28 An HPAIV lineage of subtype H5N1 which originated in South- 29 east Asia in 1996/1997 has spread across Eurasia since 2003 and 30 entered Africa in 2005 [2]. Despite intense attempts to eradicate 31 the virus endemic status is now reported from Indonesia and Egypt 32 where continuous viral circulation is likely associated with silently 33 infected domestic ducks [3]. Endemic infection of poultry increases 34 risks of sporadic human infections, the majority of which had a fatal 35 outcome [4]. 36 In Egypt, poultry possesses considerable importance as a source 37 of animal protein for human consumption. Before the incursion of 38 ��� Corresponding author at: Friedrich-Loeffler Institute, Suedufer 10, D-17493 Greifswald-Insel Riems, Germany. Tel.: +49 38351 7200 fax: +49 38351 7174. E-mail address: martin.beer@fli.bund.de (M. Beer). 1 Both authors contributed equally to this work. HPAIV H5N1 to Egypt in 2006, a huge poultry population of about 1 39 billion heads was reared for home consumption and trade to other 40 Middle East and African countries [5]. About 75% of the poultry 41 was raised in commercial farms. Almost all poultry enterprises are 42 located in the Nile valley, the majority in the Nile delta, paralleling 43 the human population density. Keeping of mixed backyard poultry 44 flocks amounts to about 25% of the countries��� poultry population 45 and is intimately intertwined with human rural and urban life. 46 Due to cultural preferences in consuming fresh meat the major- 47 ity of poultry is traded via life bird markets slaughterhouses cover 48 only about 30% of the total trading loads. Marketing is associated 49 with frequent and, outside the Nile delta, long-distance transport 50 of poultry. 51 After the incursion of HPAIV H5N1 into Egypt in 2006, possi- 52 bly by wild migratory ducks [6], attempts to limit the spread and 53 eradicate H5N1 were concentrated in a test-and-cull control strat- 54 egy. Commercial flocks require a certificate stating freedom from 55 HPAIV after testing for viral RNA before general trading restrictions 56 are lifted. In parallel, and to limit mass depopulation of poultry, 57 blanket vaccination campaigns in commercial and then in back- 58 yard holdings (free of charge) were launched using several standard 59 0264-410X/$ ��� see front matter �� 2011 Published by Elsevier Ltd. doi:10.1016/j.vaccine.2011.01.006

Please cite this article in press as: Grund C, et al. Highly pathogenic avian influenza virus H5N1 from Egypt escapes vaccine-induced immunity but confers clinical protection against a heterologous clade 2.2.1 Egyptian isolate. Vaccine (2011), doi:10.1016/j.vaccine.2011.01.006 ARTICLE IN PRESS G Model JVAC 11350 1���7 2 C. Grund et al. / Vaccine xxx (2011) xxx���xxx H5 vaccines from different vendors and sources. Despite these 60 concerted efforts, the virus persisted in poultry and evolved into 61 phylogenetically distinguishable, co-circulating lineages [2,7,8]. As 62 a surrogate marker of virus circulation, human infections continue 63 to be registered [9]. 64 In poultry flocks with a high level of population immunity after 65 vaccination against subtype H5 influenza virus infection, transmis- 66 sion and spread of HPAI H5 virus is reduced below a reproduction 67 rate (R0) of 1 which eventually leads to eradication of the infection 68 [10]. At the same time, and especially in the case of incomplete 69 or waning population immunity, caused, e.g. by maternal derived 70 antibodies [11] or antigenically distantly related vaccine strains, a 71 strong selection pressure is imposed to which influenza A viruses 72 respond swiftly with the generation and expansion of antigenic 73 drift and neutralization escape mutants. Such variants are capa- 74 ble of circumnavigating vaccine-induced immunity [12,13]. This 75 dynamic process, the results of which can be depicted by anti- 76 genic cartography [14], is the basis for updating human seasonal 77 influenza virus vaccine in a bi-annual fashion [15]. Generation of 78 escape mutants in poultry vaccinated against subtype H5 influenza 79 A viruses has been first observed in the follow-up phase of HPAIV 80 H5N2 outbreaks in Mexico in the 1990s [16]. Also, among the HPAIV 81 H5N1 viruses of Asian origin, clades 2.3.4 (���Fujian���) and 7 (���Shanxi���) 82 are believed to represent such vaccine-escape strains [17]. In Egypt, 83 since 2007, recurrent outbreaks of HPAIV H5N1 in vaccinated flocks 84 were reported with increasing frequency and the efficacy of the 85 vaccine potency currently in use in Egypt was challenged [18]. 86 Sequence analysis of the circulating viruses revealed the emer- 87 gence of variant strains with considerable genetic (subclade 2.2.1) 88 and antigenic distances to the originally introduced strain of classic 89 clade 2.2 [7]. 90 Here we examined the potency of different inactivated com- 91 mercial and experimental whole H5 virus vaccines to protect 92 against representative challenge viruses derived from two anti- 93 genically widely distinct HPAIV H5N1 ���variant��� (clade 2.2.1 var) and 94 ���classic��� (2.2.1 pro) lineages which are currently co-circulating in 95 Egypt. 96 2. Material and methods 97 2.1. Origin of viruses 98 Egyptian HPAIV H5N1 isolated in embryonated chicken eggs 99 at the National laboratory for veterinary quality control on 100 poultry production (NLQP), Animal Health and Research Insti- 101 tute at Giza, Egypt, were obtained during routine surveillance 102 of poultry holdings in Egypt. Other viruses were taken from 103 the virus repository at the Friedrich-Loeffler Institute (FLI). All 104 viruses used in this study for antigen or vaccine production or 105 for challenge were grown in 9���11 day-old embryonated eggs 106 from specific pathogen free (SPF) chickens. Infectivity of HPAIV 107 H5N1 was titrated [19] in a mink lung cell line (Mv1Lu CCL64- 108 ATCC) based on development of cytopathic effects. The following 109 viruses were employed: A/chicken/Mexico/232/94 (Mexico/H5N2), 110 A/duck/Potsdam/1402-6/1986 (Potsdam/H5N2), rec A/Vietnam/ 111 1194/2004 (NIBRG-14/H5N1), A/chicken/Egypt/NLQP-0918/2009 112 (EGYpro/H5N1), A/chicken/Egypt/0879/2008 (EGYvar/H5N1) and 113 A/duck/Egypt/0827/08 (EGYext/H5N1). 114 2.2. Origin of vaccines 115 Vaccines Mexico/H5N2, Potsdam/H5N2, and the Chinese re- 116 assorted vaccine Re-5/H5N1 (based on the HA and NA genes 117 of A/Duck/Anhui/1/2006, clade 2.3, according to [20]) were 118 of commercial origin. The Egyptian virus EGYvar/H5N1 was 119 grown at the FLI and inactivated using -propriolactone before 120 used for producing a standard water-in-oil emulsion vaccine. 121 RecA/chicken/Anhui/1/2006 (Re-5/H5N1) was not available for 122 antigen production. 123 2.3. Animal experiments 124 Vaccination and challenge experiments were carried out in the 125 BSL3+ stable facilities of the FLI. All animal experiments were con- 126 ducted following official German animal welfare regulations (LALLF 127 M-V/TSD/7221.3-2.1-031/09). SPF chicks hatched at the FLI were 128 vaccinated at 3 weeks of age by the intramuscular route using 0.5 ml 129 of the four vaccine preparations (10���12 chicks/vaccine 9���10 non 130 vaccinated control chicks). Plasma samples obtained before vacci- 131 nation were tested negative for AIV-specific antibodies using the 132 ID Screen�� Influenza A NP Antibody Competition ELISA kit (ID.Vet, 133 Montpellier, France). 134 Three weeks after vaccination the animals were challenged by 135 inoculation via the oculo-oronasal route with 106 EID50 of HPAIV 136 H5N1 EGYpro or EGYvar diluted in cell culture medium containing 137 10% fetal calf serum (FCS). An aliquot of each diluted virus batch was 138 re-titrated to ensure proper dosage of the challenge virus. Clinical 139 signs were observed and scored (0 = healthy 1 = sick 2 = severely 140 sick 3 = dead) over 10 days. Combined oropharyngeal and cloa- 141 cal swabs were taken at days 2, 4, 7 and 10 post inoculation (dpi). 142 Heparinised blood samples were obtained from all animals before 143 vaccination, before inoculation and, from all surviving birds, at the 144 end of the observation period. 145 In addition, the intravenous pathogenicity index (IVPI) of 146 EGYpro/H5N1 and EGYvar/H5N1 was determined in SPF chickens 147 according to standard protocols. 148 2.4. Detection of H5-specific antibodies 149 Hemagglutination inhibition reactions according to standard 150 protocols of the OIE and EU [21] were used to detect and differ- 151 entiate serum antibodies against the different viruses and vaccine 152 antigens employed. 153 2.5. Detection of virus 154 Viral RNA was detected by real-time RT PCR using an M-specific 155 target as described elsewhere [21] and transformed to genome 156 equivalents (GEQ) using calibration curves of defined RNA stan- 157 dards that correlate to infectivity [22]. Standards were prepared as 158 RNA run-off transcripts of the cloned M-specific target fragment 159 were used as a copy-based standard. For in vitro transcription of 160 RNA the T7 RiboMAXTM Express Large Scale RNA Production System 161 (Promega GmbH, Mannheim, Germany) was used after linearizing 162 the plasmid by restriction digestion. RNA copies were calculated 163 according to the formula: RNA molecules (per l) = RNA concen- 164 tration [ g/ l]/transcript length [nucleotides] �� 182.5 �� 1013. 165 2.6. Sequencing and phylogenetic analysis 166 RNA extraction, RT-PCR assays and sequence analysis of full- 167 length HA genome segments were carried out as described 168 previously [23]. HA sequences established in this study, and a 169 comprehensive collection of all publicly available, close to full 170 length, HA sequences from GenBank data base, were aligned (316 171 sequences MUSCLE [24]) and analysed using TopAli v2.5 [25]. 172 A GTR model with rate heterogeneity was chosen (ModelSelec- 173 tion) for a PhyML analysis. A consensus tree after 100 bootstrap 174 runs was used for further analysis. Topologies established by a 175 neighbour-joining (Topali) or a parsimony approach (DNAPARS, 176