Vaccine 37 (2019) 2232–2243 Contents lists available at ScienceDirect Vaccine journal homepage: www.elsevier.com/locate/vaccine Efficacy of novel recombinant fowlpox vaccine against recent Mexican H7N3 highly pathogenic avian influenza virus Miria Ferreira Criado a, Kateri Bertran a,d, Dong-Hun Lee a,e, Lindsay Killmaster a, Christopher B. Stephens a, Erica Spackman a, Mariana Sa e Silva b, Emily Atkins b, Teshome Mebatsion b, Justin Widener b, Nikki Pritchard c, Hallie King b, David E. Swayne a,⇑ a Exotic and Emerging Avian Viral Diseases Research Unit, Southeast Poultry Research Laboratory, U.S. National Poultry Research Center, Agricultural Research Service, U.S. Department of Agriculture, Athens, GA, USA b Boehringer Ingelheim Animal Health, Athens, GA 30601, USA c Boehringer Ingelheim Animal Health, Gainesville, GA 30503, USA d Institute of Agrifood Research and Technology, Centre de Recerca en Sanitat Animal, Campus Universitat Autonoma de Barcelona, 08193 Bellaterra, Spain1 e Department of Pathobiology & Veterinary Science, University of Connecticut, Storrs, CT 06269, USA1 a r t i c l e i n f o a b s t r a c t Article history: Since 2012, H7N3 highly pathogenic avian influenza (HPAI) has produced negative economic and animal Received 18 October 2018 welfare impacts on poultry in central Mexico. In the present study, chickens were vaccinated with two Received in revised form 1 March 2019 different recombinant fowlpox virus vaccines (rFPV-H7/3002 with 2015 H7 hemagglutinin [HA] gene Accepted 3 March 2019 insert, and rFPV-H7/2155 with 2002 H7 HA gene insert), and were then challenged three weeks later with Available online 15 March 2019 H7N3 HPAI virus (A/chicken/Jalisco/CPA-37905/2015). The rFPV-H7/3002 vaccine conferred 100% protec- tion against mortality and morbidity, and significantly reduced virus shed titers from the respiratory and Keywords: gastrointestinal tracts. In contrast, 100% of sham and rFPV-H7/2155 vaccinated birds shed virus at higher Chickens High pathogenicity avian influenza titers and died within 4 days. Pre- (15/20) and post- (20/20) challenge serum of birds vaccinated with H7N3 rFPV-H7/3002 had antibodies detectable by hemagglutination inhibition (HI) assay using challenge virus Immunity antigen. However, only a few birds (3/20) in the rFPV-H7/2155 vaccinated group had antibodies that Recombinant fowlpox virus vaccine reacted against the challenge strain but all birds had antibodies that reacted against the homologous vac- Vaccine cine antigen (A/turkey/Virginia/SEP-66/2002) (20/20). One possible explanation for differences in vacci- nes efficacy is the antigenic drift between circulating viruses and vaccines. Molecular analysis demonstrated that the Mexican H7N3 strains have continued to rapidly evolve since 2012. In addition, we identified in silico three potential new N-glycosylation sites on the globular head of the H7 HA of A/chicken/Jalisco/CPA-37905/2015 challenge virus, which were absent in 2012 H7N3 outbreak virus. Our results suggested that mutations in the HA antigenic sites including increased glycosylation sites, accumulated in the new circulating Mexican H7 HPAIV strains, altered the recognition of neutralizing antibodies from the older vaccine strain rFPV-H7/2155. Therefore, the protective efficacy of novel rFPV-H7/3002 against recent outbreak Mexican H7N3 HPAIV confirms the importance of frequent updat- ing of vaccines seed strains for long-term effective control of H7 HPAI virus. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons. org/licenses/by/4.0/). Abbreviations: Jalisco/12283/2012, A/chicken/Jalisco/CPA-12283/2012 [H7N3]; Jalisco/37905/2015, A/chicken/Jalisco/CPA-37905/2015 [H7N3]; AP, Alkaline phosphatase; ABSL-2, Animal biosafety level 2; ABSL-3E, Animal biosafety level 3 enhanced; AIV, Avian influenza viruses; CL, Cloacal swabs; dpc, Days post-challenge; DI, Defective interfering; ECE, Embryonating chicken eggs; EID50, 50 percent embryo infectious doses; GMT, Geometric mean titers; HI, Hemagglutination inhibition; HA, Hemagglutinin; HPAI, Highly pathogenic avian influenza; LPAI, Low pathogenicity avian influenza; MDT, Mean death time; NA, Neuraminidase; OP, Oropharyngeal swabs; qRRT-PCR, Quantitative real-time RT-PCR; SDS-PAGE, SDS-polyacrylamide gel electrophoresis; SEPRL, Southeast Poultry Research Laboratory; SPF, Specific-pathogen-free; SQ, Subcutaneous; TBST, Tris-buffered saline with Tween 20; WL, White Leghorn. ⇑ Corresponding author. E-mail addresses: Miria.Criado@ars.usda.gov (M.F. Criado), kateri.bertran@gmail.com (K. Bertran), dong-hun.lee@uconn.edu (D.-H. Lee), Lindsay.Killmaster@ars.usda.gov (L. Killmaster), Chris.Stephens@ars.usda.gov (C.B. Stephens), Erica.Spackman@ars.usda.gov (E. Spackman), mariana.saesilva@boehringer-ingelheim.com (M. Sa e Silva), Teshome.Mebatsion@boehringer-ingelheim.com (T. Mebatsion), Justin.Widener@boehringer-ingelheim.com (J. Widener), Nikki.Pritchard@boehringer-ingelheim.com (N. Pritchard), Hallie.King@boehringer-ingelheim.com (H. King), David.Swayne@ars.usda.gov (D.E. Swayne). 1 Present address. https://doi.org/10.1016/j.vaccine.2019.03.009 0264-410X/Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 2233 1. Introduction protect chickens against challenge with H7N3 HPAIV isolated from a recent outbreak in Mexico, A/chicken/Jalisco/CPA-37905/2015 The H5 and H7 highly pathogenic (HP) avian influenza (AI) (Jalisco/37905/2015). Protective efficacy was determined by ana- viruses (HPAIV) have been one of the most frequent causes of sev- lyzing morbidity and mortality rates, titer of challenge virus shed, ere poultry outbreaks around the world [1,2]. Recently, the H7N3 and pre-challenge serum antibody levels. Phylogenetic analysis HPAI in Mexico has caused dramatic negative economic and social was accomplished to understand the antigenic variation in the impacts due to a high number of poultry losses from deaths and H7 HA from Mexico in comparison to historical North American culling. First reported in June 2012, the H7N3 HPAI has affected lineage H7 AIV. Molecular analysis of H7 HA, as amino acid substi- over 22 million chickens in Jalisco State, Mexico’s most important tutions, protein structure, and N-glycosylation were evaluated to table egg producing area [2,3]. Simultaneously, an immunization understand the possibilities of how the HA antigenicity could have campaign using H7 inactivated virus [A/cinnamon teal/Mex- contributed to field virus escape from vaccine-induced immunity. ico/2817/2006 (H7N3) vaccine strain], and several measures were established to control the disease including quarantine of affected 2. Materials and methods premises, depopulation of infected poultry, and enhanced surveil- lance for infection in birds [3]. The initial vaccine seed strain 2.1. Animals (A/cinnamon teal/Mexico/2817/2006 [H7N3] vaccine strain) was protective against the June 2012 outbreak virus (A/chicken/Jalisco/ One-day-old SPF White leghorn chickens were housed sepa- CPA-12283/2012 [H7N3], Jalisco/12283/2012) [4–6]. Since 2013, rately by group in negative pressure isolators with HEPA-filtered additional H7N3 HPAIV outbreaks have occurred and affected intake air in an ABSL-2 facility (Merial/Boehringer-Ingelheim) for primarily layers, but also some broilers, breeders, and backyard first 3 weeks of the study, and subsequently transferred to ABSL- poultry in the Mexico States of Jalisco, Aguascalientes, Guanajuato 3 enhanced facility (Southeast Poultry Research Laboratory) for and Puebla [2]. Recently, two outbreaks of H7N3 HPAI in Mexico HPAIV challenge. Birds had ad libitum access to feed and water were reported by OIE (World Organization for Animal Health) throughout the experiment. All procedures were performed in demonstrating the outbreaks are continuing [7]. Previous studies accordance with the protocol approved by the Institutional Labora- have characterized the pathobiology of H7N3 HPAIV, and effective- tory Animal Care and Use Committee. ness of different vaccine seeds strains on protection [4–6]. The vac- cine trials have shown variable efficacy, with older H7 vaccine seed strains providing poor protection due to a combination of genetic 2.2. Viruses and vaccines and antigenic distance and low immunogenicity [4–6]. Outbreaks of H7N3 HPAI in vaccinated flocks have been reported in Mexico The H7N3 HPAIV, A/chicken/Jalisco/CPA-37905/2015 since 2016 [8]. (Jalisco/37905/2015), (GenBank accession number MH342039) A recombinant fowlpox vaccine (rFPV) expressing HA from H5 was isolated from a backyard chicken during 2015 in the state of LPAIV (A/turkey/Ireland/1378/1983) has been licensed and suc- Jalisco, Mexico (courtesy of MVZ Joaquín B. D. Álvarez and Mario cessfully used during the last twenty years in Mexico and other Solís Hernández, Servicio Nacional de Sanidad, Inocuidad y Calidad countries in Central America for immunization at 1-day-of-age Agroalimentaria [SENASICA], Mexico). The virus was propagated [1], but there is no similarly licensed rFPV-H7 AIV available in and titrated by allantoic sac inoculation of 9–10 day-old embry- Mexico for hatchery vaccination and induction of early immunity. onating chicken eggs (ECE) by standard methods [19] and used An experimental vaccine, rFPV-H7/2155 was previously tested by as a challenge virus in this study. our group and found to be protective when used alone or can Two constructs of recombinant fowlpox virus vector each con- enhance the licensed inactivated H7N3 AIV vaccine in a prime- taining a single H7 LPAIV gene insert (rFPV-H7-AIV) from: (1) boost vaccination regime [5]. North American strain A/turkey/Virginia/SEP-66/2002 (H7N2, Efficacious AIV vaccines have demonstrated protection against GenBank accession number AY240913) (rFPV-H7/2155) or (2) the morbidity and mortality from HPAIV challenge and have reduced Mexican strain A/chicken/Guanajuato/07437-15/2015 (H7N3, viral shedding from the respiratory and gastrointestinal tract of GenBank accession number AKL91078.1) (rFPV-H7/3002) were birds [1,9]. In developing countries, poultry vaccination has been constructed the same way as the licensed TROVAC-H5 and tested an important control tool in combination with depopulation of in this experiment. The fowlpox vector for TROVAC-H5 was infected poultry, biosecurity measures, education, rapid diagnostic, derived from the vaccine strain contained in the DIFTOSEC fowlpox and surveillance to control HPAIV [1,9]. However, mutations of vaccine. TROVAC-H5 received license in the United States in 1998 influenza A virus, especially in the HA gene, have resulted in loss and has since been used in Mexico, Guatemala, El Salvador, and of vaccine optimal efficacy, hindering HPAIV control [9–11]. Muta- Vietnam [20]. tions in the AIV genomes have generated changes responsible for antigenic drift and loss of protection by some vaccine seed strains 2.3. Experimental design and sampling [2,12]. Furthermore, within influenza A viruses, the acquisition of N-glycosylation has prevented neutralization by antibodies Experiment design and sampling were done using animals, vac- directed against previous strains and escaped vaccine-induced cination, and challenge protocols following identical conditions as immunity elicited by heterologous vaccines [13–16]. Therefore, previously published from our group [5]. In summary, sixty chick- the N-glycosylation of HA also can influence antigenicity and thus ens were distributed into three groups of twenty birds each and play an important role in candidate influenza A vaccine strain were given an individual identification number. All birds were vac- selection and vaccination effectiveness against newly emerging cinated at one-day of age by the subcutaneous (SQ) route, at a dose field viruses [15,17,18]. of 103.5 50 percent embryo infectious doses (EID50) in 0.2 ml per Our goal was to analyze the protective efficacy of two different bird. Groups 1 and 2 were inoculated with the rFPV-H7-AIV vacci- constructs of recombinant fowlpox virus vaccine (rFPV) containing nes: rFPV-H7/3002 and rFPV-H7/2155, respectively. Group 3 a 2002 (rFPV-H7/2155) and 2015 (rFPV-H7/3002) H7 low (Sham) received the rFPV-H7-AIV vaccine diluent. Three-weeks pathogenicity avian influenza (LPAI) virus (LPAIV) HA inserts to post-vaccination (21-days-of-age), birds were bled to evaluate 2234 M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 antibody titers in serum, and challenged by the intranasal route titers below 3 log2 GMT were considered negative, and then with 106 EID50 of HPAIV H7N3, Jalisco/37905/2015, in a volume assigned as 2 log2 GMT for statistical purpose. of 0.1 ml. The inoculum titer was verified as 105.9 EID50/0.1 ml by back titration in ECE. 2.6. Phylogenetic analysis for complete HA gene of H7 avian influenza Oropharyngeal (OP), which includes swabbing the choanal cleft, viruses and cloacal (CL) swabs were collected at 2- and 4-days post- challenge (dpc), placed in Becton-Dickinson BBL brain heart infu- A phylogenetic tree was created by the Maximum-likelihood sion (BHI) medium with 2X antibiotics (Penicillin/Streptomycin/F (ML) algorithm for the HA gene of H7 avian influenza viruses. All ungiezone; Hyclone, Logan, UT, USA), and stored at 80 °C until available complete HA gene sequences identified from North tested to determine virus shed titers. All chickens were observed America, South America, Asia, and Europe (n = 2135) were daily for clinical signs and mortality from 0 to 14 dpc. Chickens, retrieved from the Influenza Virus Resource database (http:// which were euthanized for humane reasons when clinical symp- www.ncbi.nlm.nih.gov/genomes/FLU/FLU.html) on October 18, toms as severe listlessness, neurological signs, respiratory distress, 2017. The sequences were pruned by using the software cd-hit or others HPAI infection signs, were counted as dead the next day [26] at 97% homology to filter-out 92 representative sequences. for mean death time calculations. At the end of the experiment The nucleotide sequences of HA segment from 2015 to 2016 (14 dpc, 5-weeks-of-age), surviving birds were bled to evaluate sequenced in this study (n = 5) and all Mexican H7 viruses avail- antibody titers and euthanized according to the approved IACUC able in the GenBank (n = 17) were added to the dataset. A total protocol. of 114 nucleotide sequences were aligned using MAFFT and man- ual editing of alignments were performed in Geneious 8 software 2.4. Determination of virus shedding [27]. The ML tree was estimated by the RAxML software [28] using the general time-reversible model of nucleotide substitution. Boot- OP and CL swabs were processed for quantitative real-time RT- strap support values were generated by using 500 rapid bootstrap PCR (qRRT-PCR) to determine the HPAIV viral load from the replicates. The ML phylogenetic tree was visualized with MEGA 7 oropharynx and gastrointestinal tract on 2 and 4 dpc. This method- software (http://www.megasoftware.net). Bootstrap values >70% ology has been used as standard protocol among published veteri- are shown at the branch nodes. nary influenza vaccine studies. As published [21], the protocol demonstrated the high correlation between qRRT-PCR quantity 2.7. Molecular characterization of H7 hemagglutinin: residues, N- and infectious titers when: (a) using only influenza A challenge glycosylation sites, and protein structure viruses chicken adapted and propagated in embryonating chicken eggs (ECE), (b) using low MOI to inoculate ECE in propagating The molecular characterization of H7 hemagglutinin of the chal- the viruses to generate the challenge inoculum which minimizes lenge virus (Jalisco/37905/2015), vaccines (rFPV-H7/2155 and defective interfering (DI) RNAs, and (c) using the same specific rFPV-H7/3002) were performed using different methodologies. challenge virus stock to generate the standard curve with each The virus from 2012 outbreak in Mexico, Jalisco/12283/2012, was dilution point directly comparing qRRT-PCR CT values and infec- included in the molecular analyses for comparison with the chal- tious titer. lenge virus, Jalisco/37905/2015, isolated in a recent outbreak. Briefly, in this study the viral RNA from OP and CL swabs was The HA sequence were used for alignment and residue analysis extracted using MagMAXTM-96 AI/ND Viral RNA Isolation KitÒ in the Lasergene 12 using Clustal W, MegAlign software (DNA (ThermoFisher Scientific, Carlsbad, CA, USA) following the manu- STAR, Madison, WI). The potential N-glycosylation sites were pre- facturer’s instruction. Further, qRRT-PCR assay was performed as dicted using NetNGlyc server 1.0 [29,30]. The HA structure was previously described [22] using the AgPath-ID One-step RT-PCR modelling using the H7 HA template (PDB accession number, kit (ThermoFisher Scientific) with primers and probe targeting 3M5G) in the SWISS-MODEL server [31–33]. The 3D molecular the influenza matrix gene. Samples were run on the 7500 FAST Real HA structures were visualized using the PyMOL Molecular Graph- time PCR System (Applied Biosystems, Foster City, CA, USA). For ics System (Version 2.0 Schrödinger, LLC). viral quantification, a standard curve was established with viral RNA extracted from 10-fold dilutions of the challenge virus [23], 2.8. Deglycosylation of hemagglutinin Jalisco/37905/2015. Results were reported as log10 EID50/ml and the lower limit of detection was 1.1 log10 EID50/ml; therefore A reaction for removal of glycans from the HA protein (deglyco- qRRT-PCR negative samples were treated as 1.0 log10 EID50/ml for sylation) was conducted to identify the N-glycosylation sites in the statistical purposes. HA. Samples were digested with peptide N-Glycosidase F (PNGase F) kit (New England Biolabs, Ipswich, MA, USA) according to the 2.5. Serology manufactured instructions. Briefly, the glycoproteins were dena- tured with 1X of Glycoprotein Denaturing Buffer at 100 °C for The serum collected pre- and post-challenge was used in 10 min. Further, the mixture of 1X of GlycoBuffer, 1% NP40, water hemagglutination inhibition (HI) assay to determine the antibody and PNGase F were added to the reaction followed by incubation of levels in the vaccinated birds. The antigens were prepared as pre- 37 °C for 1 h. The reaction was heat inactivated at 75 °C for 15 min. viously described [24] and the HI assays were performed according Samples were loaded in the SDS-polyacrylamide gel electrophore- to standard protocol [25]. The challenge virus, Jalisco/37905/2015 sis (SDS-PAGE) and the molecular weight of AI HA analyzed by was used as antigen in the HI with each of the 3 groups of vacci- western blot using antisera against H7N3 [A/turkey/OR/1971 nated birds (rFPV-H7/3002, rFPV-H7/2155, and Sham). To confirm (H7N3)]. vaccination, the rFPV-H7/2155 vaccinated group was also tested with the homologous vaccine antigen, A/turkey/Virginia/SEP- 2.9. Western blot assay 66/2002 (H7N2). To observe differences in serological response we tested the sera from rFPV-H7/2155-vaccinated birds with the Sample homogenates were prepared in Laemmli buffer v/v challenge virus, Jalisco/37905/2015, and virus from the first (62.5 mM Tris–HCl pH 6.8; 5% b-mercaptoethanol; 2%SDS; 20% H7N3 outbreak in Mexico, Jalisco/12283/2012, as antigens. Titers glycerol; 0.01% bromophenol blue) (Bio-Rad, Hercules, CA, USA) were expressed as geometric mean titers (GMT-log2). Samples with and denatured 5 min at 95 °C. Samples were then separate on M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 2235 8–16% Mini-PROTEANÒ TGXTM Precast Protein Gels (Bio Rad) with natized for humane reasons. Similar clinical results were observed 1X Tris/Glycine/SDS buffer (Bio-Rad) and transferred to nitrocellu- for all rFPV-H7/2155 vaccinated birds (MDT = 3.15 days) (Table 1, lose membrane using liquid electrophoresis systems CriterionTM Fig. 1); i.e. birds were not protected against challenge, and conse- Vertical Electrophoresis Cell (Bio-Rad). The nitrocellulose mem- quently had severe clinical signs (7/20) or died (13/20) within brane was blocked with 5% nonfat dry milk in Tris-buffered saline 4 dpc. By contrast, all rFPV-H7/3002 vaccinated birds had complete with Tween 20 (TBST) buffer (20 mM Tris, pH 7.5, 150 mM NaCl, protection from clinical signs and death (Table 1, Fig. 1). 0.1% Tween 20) during 1 h at room temperature, and then incubated overnight at 4 °C with chicken antisera against H7N3 3.2. Evaluation of virus shedding (A/turkey/OR/1971) at a dilution of 1:2000. The membrane was washed with TBST and incubated with an alkaline phosphatase Virus shedding was evaluated by quantification of AIV RNA in (AP) labelled goat anti-chicken IgY (H + L) (Thermo Fisher Scien- OP and CL swabs from chickens at 2 and 4 dpc using qRRT-PCR tific, Rockford, IL, USA) diluted 1:10,000. The western blot was [21]. Viral RNA was detected at 2 dpc in OP and CL swabs from visualized with the substrate BCIP/NBT (Sigma-Aldrich) according all birds in the sham group, with mean titer of 4.9 and 6.4 log10 the manufacturers protocol. The molecular size of the HA pre- EID50/ml, respectively. Similar results were observed for birds in and post-digestion with PNGase F was analyzed with the marker the rFPV-H7/2155 group, with 4.8 log10 EID50/ml (OP) and 5.9 Precision Plus Protein Dual Xtra Standard (Biorad). log10 EID50/ml (CL) titers (Table 1, Fig. 2). At 4 dpc, only one bird survived in the sham group with a viral titer in the OP and CL sam- ples, 5.8 and 7.7 log10 EID50/ml, respectively. In the rFPV-H7/2155 2.10. Statistical analysis vaccine group, only three birds were alive at 4 dpc and shed virus both orally, mean titer of 4.4 log10 EID50/ml, and cloacally, 6.2 log10 Statistical analyses were performed using Prism 7 (GraphPad EID50/ml (Table 1, Fig. 2). software, San Diego, CA, USA). The survival rate data was analyzed In contrast, rFPV-H7/3002 vaccinated birds had either no or using the Mantel-Cox Log-Rank test. Fisher’s exact test was used to very low viral shedding titers. The mean virus titer shed in the analyze statistical significance of virus shedding. Statistical differ- CL samples was 2.7 (15/20) and 1.9 (12/20) log10 EID50/ml at 2 ences in mean viral titers and antibody levels between groups were and 4 dpc, respectively. Interestingly, in the rFPV-H7/3002- analyzed using Mann–Whitney test and Tukey one-way ANOVA. A vaccinated group only 7/20 birds were positive at 2dpc and 5/20 P-value of <0.05 was considered to be significant. were positive at 4 dpc for OP with a mean titer of 1.5 log10 EID50/ml on both days (Table 1, Fig. 2). 3. Results Overall, rFPV-H7/3002 vaccinated birds had statistically signifi- cant lower OP and CL titers than sham and rFPV-H7/2155 vacci- 3.1. Morbidity and mortality in specific pathogen free (SPF) White nated groups, and significantly fewer birds shedding virus from Leghorn chickens challenged with HPAIV H7N3 the respiratory tract (p < 0.05) (Fig. 2). One-day-old chickens were vaccinated once and challenged 3.3. Antibody levels in immunized birds pre- and post-challenge with three weeks later with H7N3 HPAIV, Jalisco/37905/2015 (H7N3). HPAIV H7N3 After challenge, all the sham-vaccinated control birds developed acute severe clinical disease and died between 2 and 4 days post- Sera collected from all chickens pre- (0 dpc) and the survivors at challenge (dpc) (mean death time [MDT] = 3.20 days) (Table 1, termination (14 dpc) were processed for HI assay to measure Fig. 1). Most of the sham-vaccinated birds were found dead humoral immune response to the challenge virus antigen and vac- (11/20) but others (9/20) showed clinical signs like respiratory dis- cine seed strain (Table 1, Fig. 3). Sera from sham-vaccinated chick- tress, listlessness, lethargy, and prostration. The latter were eutha- ens were negative for anti-AIV antibodies on the day of challenge. Table 1 Summary data from one-day-old WL chickens vaccinated with one of two constructs of recombinant fowlpox virus vaccine, rFPV-H7/3002 and rFPV-H7/2155, and subsequently challenged with Jalisco/37905/2015 H7N3 HPAIV. Group Vaccinea N° of birds Morbidityb Mortalityb MDTc Oral sheddingd Cloacal sheddingd HI serologye Vaccine Jalisco/37905/2015 as strain as antigen antigen 2 dpc 4 dpc 2 dpc 4 dpc Pre- Pre- Post- challenge challenge challenge 1 rFPV-H7/3002 20 0/20 0/20 na 07/20 (1.5) 05/20 (1.5) 15/20 (2.7) 12/20 (1.9) nd * 15/20 (3.0) 20/20 (5.0) 2 rFPV-H7/2155 20 20/20 20/20 3.2 20/20 (4.8) 3/3 (4.4) 20/20 (5.9) 3/3 (6.2) 20/20 (6.6) 3/20 (2.2) nd 3 Sham 20 20/20 20/20 3.2 20/20 (4.9) 1/1 (5.8) 20/20 (6.4) 1/1 (7.7) nd 00/20 (2.0) nd a Constructs of recombinant fowlpox virus vaccine containing an H7 LPAIV gene insert. Group 1, rFPV-H7/3002 vaccine, contain the HA from the Mexican strain A/chicken/ Guanajuato/07437-15/2015 (H7N3), and group 2, rFPV-H7/2155 vaccine, the HA from North American strain A/turkey/VA/SEP-66/2002 (H7N2). Important, the HA gene had the cleavage site changed back to LPAIV. b Morbidity = no. with clinical disease/total; mortality = no. survivors/total. c MDT, mean death time, *#dead birds  dpc/total dead birds (expressed as dpc, days post-challenge). d The numbers represent no. virus positive/total in group followed by mean virus shed titer. Results were reported as log10 EID50 titer/ml and the lower limit of detection was 1.1 log10 EID50/ml; therefore qRRT-PCR negative samples were treated as 1.1 log10 EID50/ml. For the purposes of statistical calculations, all the negative samples were assigned the value of 1.0 log10 EID50/ml. Different superscript lowercase denotes statistical significance of number of birds shedding between vaccine and corresponding sham by Fisher Exact (p < 0.05). Different superscript uppercase denotes statistical significance of shedding titers between vaccine and corresponding sham by Mann-Whitney test and ANOVA (p < 0.05). §, number of positive birds was too low for statistical purposes. e The numbers represent no. serology positive/total in group (geometric mean titers (GMT-log2) only includes positive birds) followed by mean HI titers against vaccine virus or challenge virus. Samples with titers below 3 log2 GMT were considered negative, and then assigned as 2 log2 GMT for statistical purpose. nd = not determined. nd * = The challenge virus and vaccine rFPV-H7/3002 are nearly identical, then only challenge virus was used as HI antigen in this vaccinated group. 2236 M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 Fig. 1. Survival curve of vaccinated chickens challenged with Jalisco/37905/2015 (H7N3) HPAIV. Fig. 2. Scatter plot of oropharyngeal (OP) and cloacal (CL) shedding detected by qRRT-PCR after challenge with the Jalisco/37905/2015 (H7N3) HPAIV. Viral titers in swabs from vaccinated birds, rFPV-H7/3002, rFPV-H7/2155, and Sham, at 2 and 4 days post challenge (dpc) are expressed as log10 EID50/ml. The limit of detection was 1.1 log10 EID50/ml and negative samples were treated as 1.0 log10 EID50/ml. Statistical significance between mean titers was determined with ANOVA using the Tukey’s Multiple Comparison Test (p < 0.05). The p < 0.005 are indicated by ‘‘**”. The pre-challenge serum from rFPV-H7/2155 vaccinated group had 3.4. Molecular analysis of H7 avian influenza virus low levels of antibody (log2 GMT = 2.2) in a few birds (3/20) when using the challenge virus as the antigen (Table 1, Fig. 3a). However, To understand the role of HA in the protection of chickens when using the vaccine seed as antigen (A/turkey/VA/SEP- against challenge with H7N3 HPAIV, we compared the nucleotide 66/2002), all birds (20/20) had high titers of HI antibody (log2 identity for complete HA, specific amino acid sequences, and pro- GMT = 6.6) (Table 1, Fig. 3b), and when using Jalisco/12283/2012 tein structure between HA of vaccine viruses and challenge strain. (13/20), a high number of birds (13/20) had specific antibodies The ML phylogenetic tree indicates the initial Mexican H7N3 (Fig. 3c), confirming results previously published [5]. Therefore, HPAIV strain from June 2012 emerged from a North American H7 the r-FPV-H7/2155 vaccinated-birds have specific antibodies that LPAIV lineage. Since then, it evolved through genetic drift to create significantly (p < 0.05) recognize the virus from the first outbreak the three contemporary genetic subgroups (I, II, and III) during in 2012, as Jalisco/12283/2012, than Jalisco/37905/2015 (Fig. 3c). 2015 and 2016 (Fig. 4a). We observed that the HA gene of the chal- In contrast, the rFPV-H7/3002 vaccinated birds had HI antibod- lenge virus (Jalisco/37905/2015) and the H7 strain in the rFPV- ies in the pre- (15/20, log2 GMT = 3) and post- (20/20, log2 H7/2155 (A/turkey/Virginia/SEP-66/2002) had lower nucleotide GMT = 5) challenge serum when using the challenge virus identity (86.5%). The Jalisco/37905/2015 and the HA of rFPV- (Jalisco/37905/2015) as HI antigen (Fig. 3a), and when using H7/3002 vaccine [H7 from 2015 strain (A/chicken/Mexico/2015)] Jalisco/12283/2012 (6/20), a few number of birds (6/20) had speci- not only had a higher identity (99.2%) but also cluster together in fic antibodies (Fig. 3c). subgroup II. M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 2237 (a) (b) (c) Fig. 3. Scatter plot of HI titers for vaccinated birds in groups, rFPV-H7/3002, rFPV-H7/2155, and Sham pre- and post-challenge. The HI titers were analyzed using the challenge virus (A), or vaccine strain as antigen (B), or Jalisco/12283/2012 as antigen (C). Titers were expressed as geometric mean titers (GMT-log2), and samples with titers below 3 log2 GMT were considered negative. Dotted horizontal lines indicate limit of detection. Statistical significance between mean titers was determined with ANOVA using the Tukey’s multiple comparison and Mann–Whitney tests or t-test (p < 0.05). The p < 0.05 and p < 0.005 are indicated by ‘‘* ‘‘and ‘‘**”, respectively. ns = birds did not survive. Amino acid sequences were aligned to evaluate possible muta- H7N3 HPAIV outbreak strain (Jalisco/12283/2012). The tions in the HA of the challenge virus (Jalisco/37905/2015), vacci- Jalisco/12283/2012 strain was chosen for comparison because a nes (rFPV-H7/2155 and rFPV-H7/3002), and the initial Mexican previous study, the rFPV-H7/2155 vaccine provided 100% 2238 M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 protection from mortality against this lethal challenge strain [5]. (Fig. 4b). Some of these amino acid substitutions resulted in the The percentage of amino acid similarities among the HA sequences formation of potential new predicted glycosylation sites. ranged from 98.6 to 87.3%. The HA gene in the early strains (vac- Thus, we examined in silica the differences in the number of cine rFPV-H7/2155 and Jalisco/12283/2012) showed few amino potential of N-glycosylation sites in the HA sequence (Fig. 4c). Posi- acid differences. However, when we compared these strains to tions 30, 46, 249, and 501 had potential glycosylation sites in HA of the new vaccine (rFPV-H7/3002) and the Jalisco/37905/2015, Jalisco/12283/2012 virus, in which position 249 was the only one there was progressive accumulation of mutations from the early to localized in the globular head. However, between late 2012 and recent strains (Fig. 4b). The majority of the amino acid substitu- early 2016, H7N3 Mexican strains obtained additional glycosyla- tions are localized in the globular head of HA. Furthermore, these tion sites in the globular head of the HA protein at the changes occurred in regions close or inside the antigenic sites positions 141, 151 and 182: 1) subgroup I 141 and 151, 2) Fig. 4. Molecular analysis of H7 avian influenza virus hemagglutinin. (A) Phylogenetic tree created by the Maximum-likelihood algorithm for the HA gene of H7 avian influenza viruses. Available full-length HA gene sequence identified from North America, South America, Asia, and Europe (n = 2135) from the Influenza Virus Resource database were used. The vaccine strains and challenge virus used in this study are highlighted as black circles () and triangle (N), respectively. Bootstrap values >80% are shown at the branch nodes. (B) Amino acids residues of AIV H7 HA were aligned in Lasergene MegAlign (DNAStar). Antigenic sites of H3 HA (A, B, C, D, and E) are shown in the boxes. Only the amino acids different from those in the challenge virus (Jalisco/37905/2015) sequence are shown. The asterisks represent the position of predicted N- glycosylation sites are shown in blue (position 30, 46, 249, and 501) and red (position 141, 151, and 182). The N-glycosylation sites shown in blue were observed in all samples analyzed, and red are exclusive to the Jalisco/37905/2015 and vaccine rFPV-H7/3002. (C) Ribbon diagram and surface of H7 HA. Only monomers are shown. Predicted antigenic sites A (cyan), B (orange), C (magenta), D (green), and E (yellow) are labeled. Locations of predicted N-glycosylation sites are represented as spheres labeled with red (position 141, 151, and 182) and blue color (position 30, 46, 249, and 501). Antigenic sites A and E are highlighted. D) Western blot comparison of Jalisco/12283/2012 and Jalisco/37905/2015HA treated with (+) and without ( ) PNGase F. The HA was detected using chicken antiserum against H7N3 (A/turkey/OR/1971). The number of predicted N-glycans with and without PNGase treatment was shown as ‘‘HA N-Glyc”. Results are representative of two independent replicates. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 2239 (b) Fig. 4 (continued) subgroup II 141, 151, and 182, and 3) subgroup III: 141 (Fig. 4a). N-glycosylation sites were present in different quantities in the Specifically, the challenge virus (Jalisco/37905/2015) (Fig. 4b and HA proteins of the Mexican viruses from 2012 (Jalisco/12283/2012) c), which belongs to subgroup II, has additional potential N- and 2015 (Jalisco/37905/2015) outbreaks, as predicted in silico. glycosylation sites at the positions: 141, 151 and 182. Moreover, two of these sites, positions 151 and 182, were found within anti- genic sites A and E, respectively. Protein alignment and prediction 4. Discussion of N-glycosylation sites determine the structure of H7 HA from these strains affected the protein structure and possibly immune Both H5 and H7 HPAIV are the most frequent causes of severe recognition in the antigenic sites (Fig. 4c). disease outbreaks in poultry [2,34]. In developing countries, the To verify the in silico prediction of potential N-glycosylation use of vaccines is an important tool to control HPAI in combination sites in the Mexican H7 HA, we used pre-and post-digestion with with depopulation of infected poultry, strict epidemiological PNGase F followed by SDS-PAGE. Western blot analysis demon- surveillance, and enhanced biosecurity measures [35]. However, strated the HA of Jalisco/37905/2015 had a higher molecular continuous genetic evolution of H5 and H7 AIVs, as accumulation weight than HA of Jalisco/12283/2012 (Fig. 4d). Moreover, PNGAse of nucleotides and amino acids substitutions in the HA, has led F treatment, which removes the N-link glycans, resulted in a to significant antigenic diversity of these AIVs in poultry in many decrease of the molecular weight of all HA to a similar size parts of the world [36–39]. Such diversity has led to antigenic drift (Fig. 4d). Therefore, these results demonstrated that potential allowing AIVs to evade the herd immunity established from 2240 M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 (c) (d) Fig. 4 (continued) previous influenza infection or vaccination [9,35,40]. Therefore, AI should be replaced with updated vaccine seed strains if no longer vaccines must be periodically re-evaluated to determine the effec- protective. Outbreaks of H7N3 HPAI have been reported in vacci- tiveness against emergent variant circulating virus strains and nated flocks in Mexico since 2016, suggesting significant antigenic M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 2241 changes in the field viruses and a need to examine vaccine seed of the H7 HA. Recent advances in the interaction between AIV gly- strains for needed updates [8]. can and components of innate and adaptive immunity have been In the current study, immunization of one-day-old SPF White providing insights regarding the ability of N-glycans to target Leghorn chickens with two different constructs of recombinant specific cells and consequently stimulate cytokines production fowlpox virus vaccine (rFPV-H7/3002 and rFPV-H7/2155), contain- [13,16,41]. The effect of glycosylation on the HA tertiary structure ing a 2002 and 2015 LPAI H7 HA insert, respectively, were used to and possible immune recognition has been documented evaluate the protection against an H7N3 HPAI virus isolated in [13–15,42–44]. The progressive accumulation of N-glycan on the Mexico during 2015 (Jalisco/37905/2015). Interestingly, the closely HA can physically mask access to antigenic sites allowing escape related rFPV-H7/3002 vaccine conferred protection without mor- from antibody-mediated response produced against previous field bidity or mortality together with no or low virus shedding from strains and vaccine-induced immunity [13–15,44]. At the same the respiratory and gastrointestinal tracts. A different scenario time, additional glycosylation sites can interfere with the receptor was observed for birds vaccinated with the old vaccine, rFPV- binding properties, which consequently contributes to virus repli- H7/2155, which previously conferred complete protection result- cation and fitness [43]. Furthermore, previous studies have shown ing in no clinical signs or death against the 2012 original outbreak that changes in HA glycosylation in AIV can also affect the CMI, virus (Jalisco/122083/2012) [5]. Our results showed that rFPV- helping AIV to escape T cell recognition [45,46], and inhibit the H7/2155 and sham vaccinated birds after challenge with approach of clones bearing certain T cell receptors to the Jalisco/37905/2015 had 100% mortality within 4 dpc. glycopeptide-MHC complex [47]. Dendritic cells mediate the link Previous studies have demonstrated that systemic humoral between innate and adaptive immune response [48]. A recent immunity induced by AI vaccines, especially HA subtype specific, in vitro study showed that influenza virus HA with different contributed to protection against influenza [1,9]. Most of the N-glycan patterns could activate dendritic cells [48]. Although a rFPV-H7/3002-vaccinated birds had pre- and post-challenge HI few studies suggested that N-glycan could play a role in the antibodies titers against the challenge virus, which conferred pro- regulation of CMI, more experiments are still necessary to fully tection. Even though the rFPV-H7/2155 vaccine induced a strong understand how it can be modulated. immune response against the homologous vaccine antigen, these We found that the Jalisco/37905/2015 has 3 potential addi- HI antibodies were not sufficiently specific to confer any protection tional N-glycosylation sites on the globular head of HA at the against the challenge virus. Antigenic differences between the old amino acid position 141 (NGTT), 151 (NPSF), and 182 (NKSA), vaccine strain (rFPV-H7/2155) and the current challenge strain which were absent in previous strains, such as Jalisco/12283/2012. resulted in the recent H7N3 HPAI virus being poorly recognized Western blot analysis results demonstrated the higher molecular by antibodies from the rFPV-H7/2155 vaccine. This difference weight of the Jalisco/37905/2015 (challenge virus) compared to was probably caused by antigenic drift in field isolates of H7N3 the Jalisco/12238/2012 HA. Further, the glycosylation in the posi- virus. In addition to a correlation between HA antibodies response tion 151 and 182 are localized on the globular head of HA in the and protection against influenza, the cellular immune response antigenic sites A and E, respectively. Mutations in the antigenic also can contribute to protection after vaccination [1,35]. In our sites have altered recognition of neutralizing antibodies and con- study, the influence of cell-mediated immunity (CMI) perhaps con- tributed to escape mutant generation [13,17]. A previous report tributed to the post-challenge protection of a few rFPV-H7/3002- highlighted the importance of antigenic site A in HA of H7 AIV vaccinated birds (5/20) which lacked pre-challenge HI antibodies, for cross-reactive antibody response [49]. Even though residue but were protected from morbidity and mortality. Future studies 141 is not in the antigenic sites, N-glycosylation in this position are needed to better understand the mechanisms and benefits that also affects the neutralization by site A antibodies due to its loca- CMI could offer to vaccine effectiveness in poultry. tion in the protein structure [50]. Next to it, N-glycosylation at To better understand the difference observed in vaccine protec- position 142 was shown to be important to the HA and sialic acid tion, nucleotide and amino acid sequences of the HA were further interaction [18], playing a role in virus infectivity and host immune analyzed to identify possible antigenic changes in the evolving H7 response observed in different AIV strains [15,18,51]. Thus, these HPAIV viruses in Mexico. The HA phylogenetic tree suggested that modifications in the globular head of the H7 HA from Jalisco Jalisco/12283/2012 emerged from a North American H7 LPAIV lin- 2012 to 2015 viruses has potential to play a role in the changes eage and evolved to create multiple subgroups as early as 2015, observed in serological response in the rFPV-H7/2155-vaccinated which includes the challenge virus used in this study, birds, which antibodies produced poorly recognize the Jalisco/37905/2015. The amino acid sequence analysis showed that Jalisco/37905/2015. Future studies using reverse genetics should the challenge virus, Jalisco/37985/2015, and the original outbreak be done to better understand the mechanism of resistance to neu- virus, Jalisco/12283/2012, have 94.72% of HA identity. Similarity tralization conferred specific amino acid modifications and of the whole HA sequence is not predictive of protection because increases in glycosylation sites in the H7 HA from Mexico strains. a few specific changes in amino acid sequence or tertiary structure Overall, our data showed that vaccine containing the HA closely of HA protein can result in a lack of protection. In this case, the dif- related to recent outbreak H7 strains in Mexico, rFPV-H7/3002, ference in the HA identity was the result of amino acid substitu- completely protected chickens from morbidity and mortality fol- tions, especially in the globular head of HA, in regions close or lowing challenge with the 2015 Mexican H7 HPAIV strain, inside the antigenic sites. It is well known that changes at the anti- Jalisco/37905/2015. The vaccine significantly reduced the number genic sites affect the specificity of neutralizing antibodies, but the of birds shedding virus and the virus shed titers from the respira- trigger for when antigenic changes requires a vaccine change is not tory and gastrointestinal tracts. A reduction in virus shedding is an defined [1]. Overall, the amino acid substitutions at position 143 important outcome of vaccination, which reduces environmental (alanine to threonine), 151 (glycine to asparagine), 152 (serine to contamination and subsequent virus transmission from infected proline), and 184 (proline to serine) from early to recent H7 AIV vaccinated flock to uninfected flock [22]. Interesting, prior to June (Fig. 4b) suggest the formation of additional potential glycosylation of 2018, there was no licensed rFPV-H7 AIV in Mexico, although sites according to in silico analyses. the rFPV-H5 AIV has been successfully used in the last two decades Interestingly, we observed that H7 Mexican AIV strains have [1]. Additionally, rFPV vaccine choice, as opposed to the inactivated obtained potential additional glycosylation sites between June vaccine alone, has advantages as: (a) rapid insertion of any influ- 2012, early 2015, and 2016. In silico analysis showed that 2 to 3 enza HA gene due to constant antigenic shift and drift among influ- additional N-glycosylation sites were present on the globular head enza viruses [1,52,53], (b) stimulation of humoral and cellular 2242 M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 immunity when given parenterally [1], (c) priming the humoral substitutions immediately adjacent to the receptor binding site. Mbio 2014;5 (3):e01070–e1114. immune response to enhance a boost by inactivated vaccines [1], [12] Wiley DC, Skehel JJ. The structure and function of the hemagglutinin and (d) mass application and early immunity with hatchery appli- membrane glycoprotein of influenza virus. Annu Rev Biochem 1987. cation [1,54]. This vaccine was licensed in Mexico on 8 June 2018 56:365–394.10.1146. and such licensure was based upon these efficacy studies and will [13] Tate MD, Job ER, Deng YM, Guanalan V, Maurer-Stroh S, Reading PC. Playing hide and seek: how glycosylation of the influenza virus hemagglutinin can assist in reducing outbreaks of H7N3 HPAI. modulate the immune response to infection. Viruses 2014;6(3):1294–316. H7N3 HPAIV outbreaks in Mexico have been reported to OIE [14] Hervé PL, Lorin V, Jouvion G, Da Costa B, Escriou N. Addition of N-glycosylation since 2012 even though biosecurity measures and vaccination sites on the globular head of the H5 hemagglutinin induces the escape of highly pathogenic avian influenza A H5N1 viruses from vaccine-induced were already established. Molecular analysis showed that Mexican immunity. Virology 2015;486:134–45. H7N3 strains have continued to rapidly evolve and recent strains [15] Wei CJ, Boyington JC, Dai K, Houser KV, Pearce MB, Kong WP, et al. Cross- have multiple amino acid changes and potential additional of N- neutralization of 1918 and 2009 influenza viruses: role of glycans in viral evolution and vaccine design. Sci Transl Med 2010;2(24):24ra21. glycosylation on the antigenic sites of the HA. It is therefore essen- [16] Hütter J, Rödig JV, Höper D, Seeberger PH, Reichl U, Rapp E, et al. Toward tial to continue strict epidemiological surveillance to identify animal cell culture-based influenza vaccine design: viral hemagglutinin N- emerging variant viruses of H7 AIVs circulating in the field. Epi- glycosylation markedly impacts immunogenicity. J Immunol 2013;190 (1):220–30. demiological surveillance significantly contributes to the develop- [17] Wu CY, Lin CW, Tsai TI, Lee CD, Chuang HY, Chen JB, et al. Influenza A surface ment of strategies for disease prevention and control, as well as glycosylation and vaccine design. Proc Natl Acad Sci U S A 2017;114(2):280–5. more suitable vaccine strains for improving vaccination effective- [18] Wu NC, Wilson IA. A perspective on the structural and functional constraints for immune evasion: insights from influenza virus. J Mol Biol 2017;429 ness. Altogether, this study confirms the importance of updating (17):2694–709. vaccines seeds strains for long-term effective control of H7 HPAIV. [19] Swayne DE, Senne DA, Beard CW. Influenza. In: Swayne DE, Glisson JR, Jackwood MW, Pearson JE, Reed WM, editors. Isolation and identification of avian pathogens. Kennett Square, PA: American Association of Avian Declarations of interest Pathologists; 1998. p. 150–5. [20] Bublot M, Pritchard N, Swayne DE, Selleck P, Karaca K, Suarez DL, et al. Development and use of fowlpox vectored vaccines for avian influenza. Ann N None. Y Acad Sci 2006;1081:193–201. [21] Lee CW, Suarez DL. Application of real-time RT-PCR for the quantitation and competitive replication study of H5 and H7 subtype avian influenza virus. J Acknowledgments Virol Methods 2004;119(2):151–8. [22] Spackman E, Senne DA, Myers TJ, Bulaga LL, Garber LP, Perdue ML, et al. We thank Diane Smith, Scott Lee, Suzanne DeBlois, and Klaudia Development of a real-time reverse transcriptase PCR assay for type A influenza virus and the avian H5 and H7 hemagglutinin subtypes. J Clin Chrzastek for excellent technical assistance. In addition, we Microbiol 2002;40(9):3256–60. acknowledge Roger Brock and James Doster for animal care assis- [23] Swayne DE, Halvorson DA. Influenza. In: Saif YM, Fadly AM, Glisson JR, tance. We also thank MVZ Joaquín B.D. Álvarez and Mario Solís McDougald LR, Nolan LK, Swayne DE, editors. Diseases of poultry. Ames, Hernández, Servicio Nacional de Sanidad, Inocuidad y Calidad IA: Blackwell Publishing; 2013. p. 153–84. [24] Olsen B, Munster VJ, Wallensten A, Waldenstrom J, Osterhaus AD, Fouchier RA. Agroalimentaria [SENASICA], Mexico that provide the challenge Global patterns of influenza A virus in wild birds. Science 2006;312:384–8. virus strain used in this study. [25] Olsen CW, Karasin A, Erickson G. Characterization of a swine-like This research was supported by Agricutural Research Service reassortantH1N2 influenza virus isolated from a wild duck in the United States. Virus Res 2003;93:115–21. project 6040-32000-066-00D and Boehringer Ingelheim Animal [26] Huang Ying, Niu Beifang, Gao Ying, Fu Limin, Li Weizhong. CD-HIT Suite: a web Health, Inc., trust agreement 58-6040-7-002. server for clustering and comparing biological sequences. Bioinformatics 2010;26:680–2. [27] Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, et al. References Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics [1] Swayne DE, Kapczynski DR. 2016. Vaccines and vaccination for avian influenza 2012;28:1647–9. in poultry. In: Swayne DE, editor. Animal influenza. Ames, IA: Blackwell [28] Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post- Publishing; 2016. p. 378–438. analysis of large phylogenies. Bioinformatics 2014;30:1312–3. [2] OIE. Update on avian influenza in animals (types H5 and H7). Paris, France: [29] Blom N, Sicheritz-Pontén T, Gupta R, Gammeltoft S, Brunak S. Prediction of OIE, 2017 (http://www.oie.int/animal-health-in-the-world/update-on-avian- post-translational glycosylation and phosphorylation of proteins from the influenza). animo acid sequence. Proteomics 2004;4(6):1633–49. [3] FAO. Highly pathogenic avian influenza in Mexico (H7N3)- a significant threat [30] NetNGlyc 1.0 Server - DTU CBS. http://www.cbs.dtu.dk/services/NetNGlyc. to poultry production not to be underestimate: 2012, 2012, p. 26. [31] Biasini M, Bienert S, Waterhouse A, Arnold K, Studer G, Schmidt T, et al. SWISS- [4] Kapczynski DR, Pantin-Jackwood M, Guzman SG, Ricardez Y, Spackman E, MODEL: modelling protein tertiary and quaternary structure using Bertran K, et al. Characterization of the 2012 highly pathogenic avian influenza evolutionary information. Nucleic Acids Res 2014;42(W1):W252–8. H7N3 virus isolated from poultry in an outbreak in Mexico: pathobiology and [32] Bordoli L, Kiefer F, Arnold K, Benkert P, Battey J, Schwede T. Protein structure vaccine protection. J Virol 2013;87(16):9086–96. homology modelling using SWISS-MODEL Workspace. Nat Protoc 2009;4:1. [5] Bertran K, Sá E Silva M, Pantin-Jackwood MJ, Swayne DE. Protection against [33] Arnold K, Bordoli L, Kopp J, Schwede T. The SWISS-MODEL Workspace: A web- H7N3 high pathogenicity avian influenza in chickens immunized with a based environment for protein structure homology modelling. Bioinformatics recombinant fowlpox and an inactivated avian influenza vaccines. Vaccine 2006;22:195–201. 2013;31(35):3572–3576.. [34] Xu Y, Bailey E, Spackman E, Li T, Wang H, Long LP, et al. Limited Antigenic [6] Spackman E, Wan XF, Kapczynski D, Xu Y, Pantin-Jackwood M, Suarez DL, et al. diversity in contemporary H7 avian-origin influenza A Viruses from North Potency, efficacy, and antigenic mapping of H7 avian influenza virus vaccines America. Sci Rep 2016;6:20688. against the 2012 H7N3 highly pathogenic avian influenza virus from Mexico. [35] Swayne DE, Kapczynski DR. Vaccines and vaccination for avian influenza in Avian Dis 2014;58(3):359–66. poultry. In: Swayne DE, editor. Animal influenza. Ames, IA: Wiley-Blackwell; [7] OIE. Highly pathogenic avian influenza, Mexico, 2018 (http://www.oie.int/ 2016. p. 378–434. wahis_2/public/wahid.php/Reviewreport/Review?page_refer= [36] Lee DH, Bahl J, Torchetti MK, Killian ML, Ip HS, DeLiberto TJ, et al. Highly MapFullEventReport&reportid=23692). pathogenic avian influenza viruses and generation of novel Reassortants, [8] OIE. Highly pathogenic avian influenza, Mexico, 2017 (http://www.oie.int/ United States, 2014–2015. Emerg Infect Dis 2016;22(7):1283–5. wahis_2/public/wahid.php/Reviewreport/Review?page_refer= [37] de Vries E, Guo H, Dai M, Rottier PJ, van Kuppeveld FJ, de Haan CA. Rapid MapFullEventReport&reportid=23692). emergence of highly pathogenic avian influenza subtypes from a subtype [9] Swayne DE. The role of vaccines and vaccination in high pathogenicity avian H5N1 hemagglutinin variant. Emerg Infect Dis 2015;21(5):842–6. influenza control and eradication. Expert Rev Vaccines 2012;11:877–80. [38] Lee DH, Bertran K, Kwon JH, Swayne DE. Evolution, global spread, and [10] Swayne DE, Suarez DL, Spackman E, Jadhao S, Dauphin G, Kim-Torchetti M, pathogenicity of highly pathogenic avian influenza H5Nx clade 2.3.4.4. J Vet et al. Antibody titer has positive predictive value for vaccine protection against Sci 2017;18(S1):269–80. challenge with natural antigenic-drift variants of H5N1 high-pathogenicity [39] Krauss S, Stucker KM, Schobel SA, Danner A, Friedman K, Knowles JP, et al. avian influenza viruses from Indonesia. J Virol 2015;89(7):3746–62. Long-term surveillance of H7 influenza viruses in American wild aquatic birds: [11] Koel BF, van der Vliet S, Burke DF, Bestebroer TM, Bharoto EE, Yasa IWW, et al. are the H7N3 influenza viruses in wild birds the precursors of highly Antigenic variation of clade 2.1 H5N1 virus is determined by a few amino acid pathogenic strains in domestic poultry? Emerg Microbes Infect 2015;4:e35. M.F. Criado et al. / Vaccine 37 (2019) 2232–2243 2243 [40] Cattoli G, Fusaro A, Monne I, Coven F, Joannis T, El-Hamid HS, et al. Evidence hemagglutinin results in loss of recognition by CD4+ T-cell clones. Virology for differing evolutionary dynamics of A/H5N1 viruses among countries 1994;199:422–30. applying or not applying avian influenza vaccination in poultry. Vaccine. [48] Liu WC, Lin YL, Spearman M, Cheng PY, Butler M, Wu SC. Influenza virus 2011;29(50):9368–75. hemagglutinin glycoproteins with different n-glycan patterns activate [41] Johannssen T, Lepenies B. Glycan-based cell targeting to modulate immune dendritic cells in vitro. J Virol 2016;90(13):6085–96. responses. Trends Biotechnol. 2017;35(4):334–46. [49] Schmeisser F, Vasudevan A, Verma S, Wang W, Alvarado E, Weiss C, et al. [42] Hutter J, Rodig JV, Hoper D, Seeberger PH, Reichl U, Rapp E, et al. Toward Antibodies to antigenic site A of influenza H7 hemagglutinin provide animal cell culture-based influenza vaccine design: viral hemagglutinin N- protection against H7N9 challenge. PLoS One 2015;10(1):e0117108. glycosylation markedly impacts immunogenicity. J Immunol 2013;190 [50] Alvarado-Facundo E, Vassell R, Schmeisser F, Weir JP, Weiss CD, Wang W. (1):220–30. Glycosylation of residue 141 of subtype H7 influenza A Hemagglutinin (HA) [43] Das SR, Hensley SE, David A, Schmidt L, Gibbs JS, Puigbo P, et al. Fitness costs affects HA-pseudovirus infectivity and sensitivity to site a neutralizing limit influenza A virus hemagglutinin glycosylation as an immune evasion antibodies. PLoS One 2016;11(2):e0149149. strategy. Proc Natl Acad Sci U S A 2011;108(51):E1417–22. [51] Suzuki Y. Positive selection for gains of N-linked glycosylation sites in [44] Zost SJ, Parkhouse K, Gumina ME, Kim K, Diaz Perez S, Wilson PC, et al. hemagglutinin during evolution of H3N2 human influenza A virus. Genes Contemporary H3N2 influenza viruses have a glycosylation site that alters Genet Sys. 2011;86(5):287–94. binding of antibodies elicited by egg-adapted vaccine strains.. Proc Natl Acad [52] Bosworth B, Erdman MM, Stine DL, Harris I, Irwin C, Jens M, et al. Replicon Sci U S A 2017;114(47):12578–83. particle vaccine protects swine against influenza. Comp Immunol Microbiol [45] Thomas DB, Hodgson J, Riska PF, Graham CM. The role of the endoplasmic Infect Dis 2010;33(6):e99–e103. reticulum in antigen processing. N-glycosylation of influenza hemagglutinin [53] Vander Veen RL, Mogler MA, Russell BJ, Loynachan AT, Harris DL, Kamrud KI. abrogates cd4+ cytotoxic t cell recognition of endogenously processed antigen. Haemagglutinin and nucleoprotein replicon particle vaccination of swine J Immunol 1990;144:2789–94. protects against the pandemic H1N1 2009 virus. Vet Rec 2013;173(14):344. [46] Drummer HE, Jackson DC, Brown LE. Modulation of CD4+ T-cell recognition of [54] Spackman E, Swayne DE. Vaccination of gallinaceous poultry for H5N1 highly influenza hemagglutinin by carbohydrate side chains located outside a t-cell pathogenic avian influenza: Current questions and new technology. Virus Res determinant. Virology 1993;192:282–9. 2013;178(1):121–32. [47] Jackson DC, Drummer HE, Urge L, Otvos Jr L, Brown LE. Glycosylation of a synthetic peptide representing a T-cell determinant of influenza virus