ORIGINAL RESEARCH published: 16 September 2020 doi: 10.3389/fimmu.2020.02083 Identification of a Newly Conserved SLA-II Epitope in a Structural Protein of Swine Influenza Virus Massimiliano Baratelli 1† , Sophie Morgan 2 , Johanneke Dinie Hemmink 2† , Elizabeth Reid 2 , Brigid Veronica Carr 2 , Eric Lefevre 2 , Sergio Montaner-Tarbes 3,4 , Bryan Charleston 2 , Lorenzo Fraile 4 , Elma Tchilian 2 and Maria Montoya 1,2,5* 1 Centre de Recerca en Sanitat Animal (CReSA), UAB-IRTA, Universitat Autònoma de Barcelona, Bellaterra, Spain, 2 The Pirbright Institute, Surrey, United Kingdom, 3 Innovex Therapeutics S.L., Badalona, Spain, 4 Animal Health Department, Universidad de Lleida, Lleida, Spain, 5 Centro de Investigaciones Biológicas Margarita Salas (CIB-CSIC), Madrid, Spain Edited by: Jayne Hope, The University of Edinburgh, Despite the role of pigs as a source of new Influenza A Virus viruses (IAV) potentially United Kingdom capable of initiating human pandemics, immune responses to swine influenza virus Reviewed by: (SwIV) in pigs are not fully understood. Several SwIV epitopes presented by swine MHC Renukaradhya J. Gourapura, The Ohio State University, (SLA) class I have been identified using different approaches either in outbred pigs or United States in Babraham large white inbred pigs, which are 85% identical by genome wide SNP Sabine Hammer, University of Veterinary Medicine analysis. On the other hand, some class II SLA epitopes were recently described in Vienna, Austria outbred pigs. In this work, Babraham large white inbred pigs were selected to identify *Correspondence: SLA II epitopes from SwIV H1N1. PBMCs were screened for recognition of overlapping Maria Montoya peptides covering the NP and M1 proteins from heterologous IAV H1N1 in IFNγ ELISPOT. maria.montoya@cib.csic.es A novel SLA class II restricted epitope was identified in NP from swine H1N1. This † Present address: conserved novel epitope could be the base for further vaccine approaches against H1N1 Massimiliano Baratelli, HIPRA, Amer, Spain in pigs. Johanneke Dinie Hemmink, International Livestock Research Keywords: swine influenza, class II epitopes, swine, IFNγ, nucleoprotein Insitute, Nairobi, Kenya Specialty section: INTRODUCTION This article was submitted to Comparative Immunology, Swine Influenza is an important respiratory pig disease caused by Influenza A Virus (IAV) (1) which a section of the journal represents a significant problem to farming but also carries very substantial risks to human health. Frontiers in Immunology Domestic pigs cover an important role in the ecosystem of IAV; it is believed that they may serve Received: 18 April 2020 as “mixing vessels” for the generation of novel reassortant influenza A virus potentially capable of Accepted: 30 July 2020 initiating human pandemics (2, 3). Controlling the virus in pig populations is of great importance Published: 16 September 2020 not only for implementing and/or improving control and surveillance programs on farms but also Citation: for preventing zoonotic infections. Thus, increasing basic knowledge of immune responses of pigs Baratelli M, Morgan S, Hemmink JD, to IAV as a basis for developing novel and improved vaccines is of great importance. Reid E, Carr BV, Lefevre E, Selecting the optimal antigen is a cornerstone in rational vaccine design and depending on the Montaner-Tarbes S, Charleston B, desired response, proper B or T cell epitopes should be carefully selected. Besides the antibody Fraile L, Tchilian E and Montoya M response, the T cell response is increasingly gaining more attention in the research of influenza (2020) Identification of a Newly Conserved SLA-II Epitope in a vaccines due to its importance in protection and its ability to provide a wider breath of protection Structural Protein of Swine Influenza by targeting conserved domains (4). The T cell response against influenza is not fully understood Virus. Front. Immunol. 11:2083. in pigs (5) and even less is known about IAV T cells epitopes recognized by Swine Leucocytes doi: 10.3389/fimmu.2020.02083 Antigens (Mayor Histocompatibility Complex of pigs). T cells are considered actively involved Frontiers in Immunology | www.frontiersin.org 1 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV both in clearance of the primary infection and in the control of Immunizing Antigens heterologous reinfection of IAV (6). Specific CD4+ and CD8+ Three type of antigens were selected: SpH1N1, Gripovac3 T cells are produced locally and systemically in pigs after IAV and S-Flu. The strain A/Swine/Spain/SF11131/2007 (SpH1N1) infection and they show a multifunction profile, cross-reactivity belonged to the swine lineage of IAV and in particular to the and memory similar to viral infections in humans (6, 7). European “Avian-like” (13). This was propagated on Madin Reverse vaccinology is not fully developed and accurate for Darby Canine Kidney cells (MDCK) as described previously pigs due to the insufficient data on SLA properties on which (13) and inactivated by UV radiations. Inactivation of the virus to train and/or test algorithm for prediction of T cell epitopes. was confirmed as absence of formation of any cytopathic effect Nevertheless, pan methods have been successfully used to identify after 7 days incubation in a MDCKs monolayer. Integrity of SLA epitopes of both classes (8, 9); these methods can predict the viral proteins was indirectly confirmed as consistence of T cell epitopes that bind to MHC for which experimental data Hemagglutination Units (HAU) before and after UV inactivation. are limited or not available. Therefore, empirical strategies are Hemagglutination test was performed as described elsewhere still crucial to generate the basic knowledge to generate novel pig (WHO). One dose of the immunization antigen SpH1N1 was specific prediction methods. prepared by mixing the volume corresponding to 1–3.5 × This study aimed to identify novel SLA-II epitopes of IAV 107 TCID50 of the inactivated viral solution with 1 ml of in pigs by means of empirical methods. The Babraham is a the adjuvant Montanide ISA 206 VG (Seppic). The second large white inbred pig lineage that is 85% identical by genome immunizing antigen solution was Gripovac3 (IDTbiologika) wide SNP analysis (10). The matching of MHC class I and which is constituted by whole inactivated IAVs formulated II alleles between individual animals makes them instrumental with adjuvant. The vaccine contained Europeans strains of for immunological studies. Babraham’s pigs were immunized the swine lineage of H3N2, H1N1, and H1N2 subtypes. The in this study following the concept that inactivated vaccines third immunizing antigen was a recombinant live IAV (S-Flu) are generally designed to present and promote presentation of constituted by an H5 Hemagglutinin and the rest of proteins extracellular antigens. Thus, those type of antigens should be from the human IAV A/PR/8/34. This virus was generated and preferentially processed toward and presented by class II SLAs. attenuated by suppression of HA signal sequence by Dr. Alan It has been shown in humans that CD4+ T cell responses Townsend (University of Oxford, UK) as described in (14). One focused mainly on NP and M1 proteins of IAV and that class dose of S-Flu was containing 6 × 107 TCID50 . II HLA epitopes are most commonly found on internal proteins (11, 12). For this reason, the specificity of the recall response Pig Immunization and Sampling produced by overlapping peptides covering the NP and M1 The cohorts of pigs selected for the study were immunized with protein from heterologous IAV was tested using PBMC from three different strategies implying two dose of the IAV antigens immune Babraham’s pigs and an IFNγ ELISPOT assay. In this administered within a 28 days interval (Table 1). The inactivated way, we were able to identify two new SLA-II restricted epitopes antigens were administered intramuscularly in the front shoulder in a structural protein that was conserved between IAV and of the animals. The live antigens S-Flu was administered in could represent a potential epitope to induce protective responses the nasal cavities using a mucous administration device (Wolfe against several IAV subtypes. Tory Medical). Blood was collected in heparinised tubes by anterior vena cava venepuncture at 0, 7, 14, and or 21 days post boost (dpb). Peripheral Blood Mononuclear Cells (PBMCs) were MATERIALS AND METHODS obtained by density gradient centrifugation of blood, 1,200xg for 30 min over Histopaque 1.083 g/ml (Sigma). RBC were lysed Animals by 5 min incubation on ice with Ammonium Cloride buffer Babraham’s pigs were housed at the Greenfield farm (Compton) (Ammonium Cloride 155 mM, Potassium Bicarbonate 10 mM, belonging to The Pirbright Institute. Those are inbred large EDTA disodium salt 0.1 mM) and then washed several time with white pigs syngeneic at the MHC loci. Animals were Mycoplasma PBS. PBMCs samples were stored in liquid nitrogen until use. Hyopneumoniae and Porcine Reproductive and Respiratory Blood was also collected at the same time points for sera recovery. Syndrome Virus free. They were vaccinated against Glässer Finally, animals from herd 1 and 2 were culled at 28 days post disease (MSD Animal Health), Erysipelas (Porcillis Ery Injection, boost (dpb), wheras animals from herd 3 were culled at 14 dpb. MSD Animal Health), Parvo Virus (Suvaxyn Parvo/E Injection, Zoetis) and received treatment for Mites and Parasites with Culture Medium Panomec Injection (Merial). Cohort 1: animals were 25–30 weeks Culture medium was RPMI-1640 medium with glutamax- old and around 39–50 Kg of weight; they were labeled B557, I and 25 mM hepes (Life technologies) –emented with B558, B563, B564, B568, and B570. Cohort 2: animals were 8– penicillin/streptomycin (Life technologies), and 10% heat- 9 weeks old and around 5.5–9.5 Kg of weight; they were labeled inactivated pig serum (Life technologies) or fetal calf CTRL, 1 and 2. Cohort 3: animals were 2–3 years old and around serum (Gemini). 173–192 Kg of weight; they were labeled B625 and B650. All studies were carried out under UK Home Office License number Peptides 70/7505 and approved by the ethical review processes at The Pools 1–3 represented the entire NP (GenBank: AFG72805.1) Pirbright Institute. and M1 (GenBank: AFG72802.1) protein sequences of human Frontiers in Immunology | www.frontiersin.org 2 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV TABLE 1 | Babraham’s pigs immunization plan. TABLE 2 | Peptides representing the sequences of 43 and 44. Immunization Animals 1st dose 2nd dose Name Sequence Name Sequence Antigen Route Antigen Route 43a NQQRASAGQISVQPTFSVQR 43m ISVQPTFSVQRNL 43b RASAGQISVQPTFSVQR 43n QISVQPTFSVQRN Strategy 1 Cohort 1 43c AGQISVQPTFSVQR 43o ISVQPTFSVQRN B557, B563, B568 SpH1N1 IM SpH1N1 IM 43d QISVQPTFSVQR 44a SVQPTFSVQ B558, B564, B570 Gripovac3 IM Gripovac3 IM 43e SVQPTFSVQR 44b SVQPTFSVQRNL Strategy 2 Cohort 2 43f VQPTFSVQR 44c SVQPTFSVQRNLPF 1, 2 SpH1N1 IM SpH1N1 IM 43g ISVQPTFSVQR 44d SVQPTFSVQRNLPFEKS CTRL None None None None 43h QISVQPTFSVQRNLPF 44e SVQPTFSVQRNLPFEKSTVM Strategy 3 Cohort 3 43i ISVQPTFSVQRNLPF 44g SVQPTFSVQRNLP B625, B650 SpH1N1 IM S-Flu IN 43j QISVQPTFSVQRNLP Sp44 SVQPTFSVQRNLPFERATIM Two dose of the IAV were administered within a 28 days interval. 43k ISVQPTFSVQRNLP Irrelevant NIKNESKYSNTFINNAYNMS 43l QISVQPTFSVQRNL The underline text represents the overlapping region of the 2 peptides. IAV strain A/Panama/2007/1999 (H3N2) fused with the amino acid sequence GGGPGGG. Pool MSP represented the Malaria Merozoite Surface Protein-1 was used as irrelevant peptide IFNγ ELISPOT pool. Pools 1–3 and MSP were constituted by 80 peptides (numbered 1–80) with length between 17 and 20 amino acids Enzyme-linked immunosorbent spot (ELISPOT) plates were and overlapping 10 amino acids. Pools A-D were pools of prepared with the capture antibody mouse anti-pig IFN-γ (Clone 6–8 peptides from pool 2. The amino acid identity of NP P2G10) (Becton Dickinson) as previously described (16). PBMCs and M1 proteins between A/Panama/2007/1999 (H3N2) and were plated at 5 × 105 cells per well and stimulated with SpH1N1 was 90 and 92% respectively; thus they are different antigens as indicated above. Pokeweed mitogen (PWM) 1 µg/ml enough to be considered heterologous (<97% sequence identity). or Concanavalin A (ConA) 10 µg/ml (Sigma) were used to Peptides were synthesized by the laboratory of proteomics stimulate 105 PBMC as positive controls of the assay. Plates & protein chemistry, Department of experimental & health were cultured overnight at 37◦ C and 5% CO2 . Spots were sciences of the Pompeu Fabra University (Barcelona, Spain). revealed by using a biotinylated mouse anti-pig IFNγ (Clone All peptides were resuspended in DMSO and diluted in PTC11) (Becton Dickinson) antibody as previously described culture media to reach the desired concentration and stored (16). Finally, dark blue-colored immunospots were counted at −80◦ C until use. using the AID ELISPOT reader (AID Autoimmun Diagnostika). Stimulation Index (SI) was calculated as ratio between ISC values of peptides and background (Media). Antigens Testing by ex vivo Tissues Stimulation Prediction of SLA-II Epitopes MDCK supernatants corresponded to the vehicle of SpH1N1 and The NetMHCpan 3.2 (17) pan algorithm was used to predict it was used to determinate the background immune response SLA-II epitopes of the NP of A/Panama/2007/1999 (GenBank: produced as consequence of immunization with impure virus AFG72805.1). The Babraham’s pigs SLA-II alleles were used as preparation. This antigen was obtained from mock infected previously described: SLA-DRB1∗ 05:01 (IPD-MHC: SLA06022), MDCK cells by the same method used to propagate SpH1N1. SLA-DQA∗ 01:03 (IPD-MHC: SLA05927) and SLA-DQB1∗ 08:01 Culture media was used to determinate background of the (IPD-MHC: SLA05978) (18); their amino acid sequences were set assays (Media). These and other antigens were diluted in culture up for a customized prediction. media and when required DMSO. Where specified, DMSO was added to the stimulus diluted 500-fold to resemble the amount Conservation Analysis of Epitopes in IAV contained in peptide suspension. PBMCs were stimulated with Non identical and full length sequences of the NP protein of the following amount of antigens per 5 × 105 cells: 105 TCID50 IAV isolated in swine (n = 2,721), human (n = 3,482) and of SpH1N1, MDCK sup was diluted like virus, 2 × 105 TCID50 avian (n = 5,640) were retrieved from the Influenza Research of S-Flu, 1 µg of pools 1–3 and MSP and 0.2 µg of pools A-D Database (19) (through the web site at http://www.fludb.org) (as or individual peptides. The restriction of the recalled immune of January 13th, 2020). Sequences were aligned with MAFFT (20) responses was tested as previously described by (15). Anti SLA-I version 7 (https://mafft.cbrc.jp/alignment/server/large.html); the and SLA-II antibodies or isotype control antibodies were added alignment was inspected and curated by using Jalview (21) to the ex-vivo recall stimulation of PBMCs as indicated in version 2.11.0. The region spanning the identified epitope Table 3. The produced immune response was then quantified by sequences were used to generate a logo by Weblogo3 (22) version IFNγ ELISPOT. 3.7.4 (http://weblogo.threeplusone.com/). Frontiers in Immunology | www.frontiersin.org 3 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV TABLE 3 | Immunological reagents. Clone Subtype Fluorophore Brand Amount MHC BLOCKING 5 × 105 cells 1◦ Antibodies SLA-I 74-11-10 IgG2a – Kingfisher and in house 15 µg SLA-II MSA-3 IgG2b – Kingfisher and in house 15 µg Isotype controls W6/32 IgG2a – In house NA Data Analysis the NP protein whereas pool 3 represented the M protein of Data were transformed and graphed with Prism v5.01 IAV. The assay was initially performed by using the PBMCs (GraphPad). Descriptive statistics including distribution purified from Babraham’s pigs immunized with the heterologous images of results and tables were performed with SAS system antigens SpH1N1 and Gripovac3 (Cohort 1 and Cohort 2). The V.9.1.3 (SAS institute). Non-parametric statistics analysis immunodominant response was obtained by pool 2 being the was performed with Wilcoxon test implemented from SAS other pools not responding (Figures 1A,B,D). Further analysis software. Individual pigs were used as experimental unit. The of peptides included in pool 2 identified peptide 43 and 44 as significance level (p) was set at <0.05 with tendencies reported immunodominant (Figures 1B,D). when p < 0.1. To identify the shortest amino acid sequence inducing an IFNγ secretion response, a new set of peptides with sequential RESULTS deletion of the end terminals regions was designed from the peptides 43 and 44 sequences (Table 2). This set of peptides Ex-vivo IFNγ Recall Responses in was then tested by ELISPOT using PBMCs from Babraham’s Babraham’s Pigs pigs immunized with the heterologous antigens SpH1N1 and Babraham’s pigs (Cohorts 1–3) were sequentially and differently S-Flu (Cohort 3). The shortest identified immunogenic regions administered with attenuated live and/or inactivated IAV (Figures 1C,E) were represented by peptides 43d and 44c which antigens; the immunization strategy was set up and tested were then further dissected with a new set of peptides designed empirically with the aim of improving the ex vivo IFNγ secretion with the same strategy adopted before (Table 2). Moreover, response against IAV. The tissues collected from these pigs were IFNγ secretion responses induced by the region spanning then tested ex-vivo to evaluate the immunization performances the entire sequence of peptide 43d and 44c was also tested and also to identify and characterize T cell epitopes of IAV. (Figure 1F). Unfortunately, the potency to carry out a statistical The immunizations strategies primed an IFNγ secreting response analysis was extremely low. Thus, it was almost impossible against the live (S-Flu) or inactive (SpH1N1) whole virus when to find significant differences among peptides. Peptide 43g tested by ELISPOT (Figures 1A–D). Viruses used to immunize and 44g were the shortest immunogenic regions identified in the pigs were genetically divergent (13) but despite this, they were this study. Notably, the removal of I at the left terminal of able to produce a cross-reacting IFNγ response. In particular, S- peptide 43g and NLP in 44g was fundament to maintain the Flu belonged to a completely different lineage of IAV compared to immunogenic properties of each peptide. The effects of the SpH1N1 or Gripovac3 antigens; however, it was able to recall ex- right and left terminal end amino acids of peptides 43g and vivo an IFNγ response on PBMC collected from pigs immunized 44g were not further characterized. Moreover, the following with SpH1N1 (Cohort 2) (Figure 1B). effects have been observed: Peptide 43k and 43i fulfilled the hypothesis that the sum of 2 epitopes (44g and 43g) increased Identification of an Antigenic Regions in responses compared to those observed by stimulation with each the M1 and NP Proteins of IAV one alone. Peptide 43j showed that simultaneous addition of The strain A/Panama/2007/1999 (H3N2) is an IAV belonging Q and P amino acids enhance IFNγ response of peptide 43k. to a different subtype and lineage compared to the antigens Finally, peptide 43h suggested that F partially suppress responses used for the immunization of pigs. In particular, its M1 of 43J. and NP proteins shared between 92.06–95.63% and 89.76– Peptide 43g and 44g corresponded, respectively, to 91.97% of the amino acid sequences, respectively, with them. regions NP406−416 (ISVQPTFSVQR) and NP407−419 To identify the antigenic regions able to produce a cross (SVQPTFSVQRNLP) of A/Panama/2007/1999. Conservation reacting IFNγ response, a collection of overlapping peptides of these regions in the IAV used as immunization antigens in representing the two proteins of A/Panama/2007/1999 were this study was inspected. Region NP406−419 was mutated in tested ex vivo by ELISPOT. All peptides were initially tested in H3N2 virus contained in Gripovac3 (I406V) and in the S-Flu pools (pools 1–3) and subsequently dissected into smaller subsets (V408I) whereas it was completely conserved in the H1N1 and up to identify those individual peptides inducing a statistically H1N2 strains of Gripovac3 and also in SpH1N1. Therefore, the significant IFNγ secretion response. Pool 1 and 2 represented Babraham’s pigs were immunized at least once with an antigen Frontiers in Immunology | www.frontiersin.org 4 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV FIGURE 1 | Empirical identification of T cells epitopes. A collection of peptides spanning the NP and M1 proteins of A/Panama/2007/1999 was used to stimulate ex vivo the PBMCs collected from Babraham’s pigs. These were previously immunized using either (A) strategy 1 (i) SpH1N1 and (ii) Gripovac3 immunized pigs, (B) strategy 2 or (C,E,F) strategy 3. The recalled immune response was quantified by an IFNγ ELISPOT and calculated as number of interferon Gamma Secreting cells (ISC) per 106 PBMCs (A,B,F). The results obtained throughout the tests and cohorts were transformed as Stimulation Index (SI) and summarized in (D,E); each dot represents the average value from a single assessment in a pig. The peptides were tested as pools (pools 1–3, pools A–D) and individually (39–45, 43a−44e, 44g−43h). The letters on the top of the graph in (E) represent the amino acids sequentially deleted from peptide 43a NQQRASAGQISVQPTFSVQR or 44e SVQPTFSVQRNLPFEKSTVM. In contrast the letter on the top of the graph in (F) represent the amino acid sequentially added from the peptide 43–44 overlapping region (in bold). The UV-inactived SpH1N1 and live S-Flu (H5) IAVs were used to recall a specific immune response in the immunized pigs. The MSP pool, the irrelevant peptide, the media either supplemented or not with DMSO were used to evaluate the non-specific response of the testing system. The MDCK sup was used to evaluate the non-specific immune response against the whole virus and in particular the derived from the replication system adopted. Finally, the PWM either supplemented or not with DMSO and the ConA were used to check the responsiveness of the PBMCs used in the assay. Statistically significant differences in (A–C) are indicated by asterisk (Wilcoxon signed-rank test; p < 0.05). Figures are representative of three independent experiments. containing the same sequence of the IAV epitopes identified in greater magnitude compared to anti SLA-I antibody (Figure 2). this study. In particular, the first inhibited 84 and 86% of peptides 43 and 44 recall responses, respectively; in contrast, <25% of the responses Identification of the SLA Class Restriction were blocked by the anti SLA-I antibodies. The SLA class restriction of NP406−416 and NP407−419 (contained, The IEDB database was then inquired to investigate whether respectively, in peptides 43 and 44) was inferred by testing their other MHC-II epitopes were described falling in the region immunogenicity in presence of either an anti SLA-I or SLA-II NP406−419, regardless the species associated. For this purpose, a antibody. PBMCs from Babraham’s pigs immunized with IAV search for similar linear MHC-II restricted (blast 70%) epitopes were stimulated ex-vivo with the peptides in presence of either was set up. The database contained 17 sequences of MHC-II of the anti SLA antibodies which blocked the responses mediated restricted epitopes related to the NP407−419 of IAV deposited by those receptors; the produced immune response was then up to 2019 (Figure 3); notably, the human MHC-II HLA-DRB quantified by an IFNγ ELISPOT. Responses induced by the tested were the most related alleles, when this specific information antigens was inhibited by the anti SLA-II antibody with much was available. Frontiers in Immunology | www.frontiersin.org 5 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV FIGURE 2 | SLA class restriction of the identified immunogenic regions. The peptides 43 and 44 were used to recall ex-vivo an immune response in PBMC collected from pigs immunized with IAV (cohort 3) in presence of SLA-II and SLA-I specific antibodies. The response was then quantified by IFNγ ELISPOT and calculated as ISC/106 PBMCs. The irrelevant peptide was used to evaluate the specificity of the peptide induced response. The isotype control antibodies were used to check the specificity of the restriction of the anti SLA antibodies. (A) The response of each tested pigs was represented as mean with sd; the asterisks indicate a statistically significant difference (Wilcoxon signed-rank test; n = 3; p < 0.05). (B) Percentage of the ISC/106 PBMCs recalled in presence of anti SLA-antibodies compared to their absence. The asterisk indicates a statistically significant difference (Wilcoxon signed-rank test; inter assay n = 4; p < 0.05). The relationship of NP406−416 and NP407−419 with the 43i and 43j (Figure 1 SI). The assay had low power as it was Babraham’s pigs SLA-II alleles was tentatively inferred by in- performed just two times in a single pig and thus it deserved silico prediction. As there are no open available algorithm of further studies. prediction designed specifically for pigs a pan algorithm was used for such purpose The Babraham’s pigs SLA-II alleles (DRB1 and DQ) and the NP (A/Panama/2007/1999) sequences were set Conservation of NP406−416 and NP407−419 up in NetMHCpan 3.2. The resulting peptide sequences were Among IAV ranked depending on their predicted affinities and then those Conservation of the identified epitopes was inspected in the sharing part of their sequence with NP406−416 and NP407−419 IAV isolated from the three major host species (swine, human were retrieved. Notably, NP412−420 FSVQRNLPF and NP406−414 and avian specie). For this purpose, a collection of non- ISVQPTFSV were ranking, respectively, among the top 2% redundant full NP sequences of IAV was retrieved from a (strong binders) and 10% (weak binders) of the 9mer predicted public domain database and aligned; the generated database peptides for the SLA-DRB1. In contrast, no similar epitope was was therefore meant to contain unique variants of IAV. Then found for the SLA-DQ. the variability at the region NP407−419 were presented as logo The SLA-II allele restriction of the region NP406−419 was in Figure 4. SwIV showed a high conservation of the entire tentatively tested as described above but in presence of antibodies sequence; in particular, NP417 (98.09%) and NP406 (98.97%) specific for the alleles DR and DQ to give support to the were the least conserved positions, whereas the rest of the above predictions. The antibody specific for the SLA-II DR allele sequence was conserved in more than 99% of SwIV strains decreased up to a maximum of 80% the responses of peptides (Table 1 SI). Frontiers in Immunology | www.frontiersin.org 6 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV FIGURE 4 | Conservation of NP406−419 , ISVQPTFSVQRNLP. The logo was FIGURE 3 | Alignment of the MHC-II epitopes similar to NP406−419 described. generated using unique NP sequences of IAV isolated from (A) swine (B) The red square indicates the region shared by most of the epitopes. The code human or (C) avian species. Letters represent the amino acid code and their on the left represent the IEDB reference. height is proportional with the frequency. Human IAV showed an inconsistent conservation throughout was found in a higher proportion of strains (91.79%). the NP407−419 sequence. Position NP406 was conserved in just This position showed a high preference for the mutation 59.82% of the inspected strains whereas 38.71% of the rest of I406V (8.10%). Positions NP417 and NP408 were conserved, them showed the mutation I406T (Table 1 SI). Notably, 98.04% respectively, in 94.38 and 95.20%, whereas, the rest of (n = 1,327) of the strains bearing the mutation I406T were position was conserved in more than 99% of avian IAV belonging to the H3N2 subtype (Table 2 SI); this was not a (Table 1 SI). bias due to an over representation of the H3N2 subtype in the Based on the conservation findings, the cross-reactive database as they constituted the 48.7% of the strains, whereas the propriety of the identified region NP406−419 was tentatively H1N1 constituted the 43.5%. The number of strains containing tested in a completely different experiment. PBMCs samples Isoleucine or Threonine amino acids at position NP406 were were retrieved from a previous study conducted in Babraham’s distributed throughout the last 51 years (since the introduction pigs in which they were immunized with a non-adjuvated whole of H3N2 IAV in humans) depending on their isolation time. A(H1N1)pdm/09 influenza vaccines licensed for humans (16). Figure 5 showed that I406T mutation started to be detected This NP segment of this virus derived from the “classical” since the 2000 and its presence increased dramatically up to SwIV which is a different lineage compared to the immunizing 2017. Other least conserved positions were, respectively, NP408 antigens used above. Results showed that peptides 43i and 43j (91.67%), NP411 (96.41%) and NP418 (98.48%) whereas the rest were able to recall an IFNγ secretion responses in one of were conserved in more than 99% of the inspected human IAV the immunized pigs whereas the rest and the mock-vaccinated strains (Table 1 SI). pigs were not responding (Figure 2 SI). However, the assay had The avian IAV also showed the lowest conserved low power as it was repeated two times and thus it deserved position at NP406, however, unlike human IAV, Isoleucine further studies. Frontiers in Immunology | www.frontiersin.org 7 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV FIGURE 5 | Dynamic of the I406T amino acid mutation in sequenced H3N2 IAV variants since their introduction in humans. Dotted line represents the number of strains having I406 whereas the continuous line represents the number of strains having T406 in their NP protein. DISCUSSION a collection of peptides spanning the entire length of the NP and M1 proteins was tested ex-vivo by using an IFNγ ELISPOT The CD4+ T cell response is considered of great importance in assay on PBMC harvested from Babraham’s pigs previously immune response against IAV; it provides help to CD8+ T cell immunized with IAV. The adopted immunization and screening and B cell responses but it also directly controls virus spread by strategies permitted to successfully identify two overlapping means of cytotoxic effects on infected cells [Reviewed in (23)]. immunogenic regions: NP406−416 (ISVQPTFSVQR) and Novel strategies of rational design of IAV vaccines are therefore NP407−419 (SVQPTFSVQRNLP). Despite sharing most of their considering to modulate T cell responses including not only CD8 sequences, they are considered different epitopes because the but also the CD4 subsets (4). non-overlapping positions (NP406 and NP417−419 ) were critical In swine, few SLA-II epitopes of IAV have been identified, to maintain their respective immunogenic proprieties. NP406−416 as compared to HLA epitopes. In the present study, SLA-II (ISVQPTFSVQR) and NP407−419 (SVQPTFSVQRNLP are the epitopes of IAV were identified in pigs by using an empirical shortest epitopes tested in this study, the effect of the removal of approach. Babraham’s pigs were used as model to identify T the amino acid at the right terminal of NP406−416 or both end cells epitopes because their syngeneic MHC loci which aids terminals of NP407−419 were not tested. The MHC restriction identification of SLA restriction of the identified epitopes. of the identified regions was inferred by blocking IFNγ recall Previous studies demonstrated that immunization strategies responses with class specific antibodies, a strategy previously based on the administration of whole inactive virus (WIV) adopted in pigs by other authors (15). Peptides containing the formulated with adjuvant in pigs were able to produce an IFNγ or identified regions showed specificity for the SLA-II; therefore, no proliferative cell mediated response in peripheral blood involving SLA-I epitope was identified in this study. T helper subsets (8, 16, 24). Moreover, the generated immunity A previous study demonstrated that immunization with was also able to respond to heterologous IAV strains (24). inactive IAV antigens is able to produce a Cytotoxic T Similar results were obtained in this study by boost lymphocytes response against the whole virus in Babraham’s immunization of Babraham’s pigs with live and/or inactivated pigs, although this was detected as proliferative response (16); whole IAV antigens; in particular, the latter was constituted by moreover, SLA-1 of the same pig lineage were identified in an in-house UV-inactivated IAV formulated with adjuvant or the NP protein of IAV by means of a strategy involving long- a commercially available vaccine which was licensed by that term in vitro pig T-cell culture and cloning (25). Therefore, the time in pigs. The adopted immunization strategies produced missed identification of SLA-I epitopes might have been due to a specific IFNγ immune response in the Babraham’s pigs; this differences in the immunization and/or screening strategy. was recalled ex vivo by stimulating PBMC with homologous and The IDEB database contained just three M1 and one NP even heterologous IAV antigens and it was detected by means related epitopes restricted to the SLA-II. These were discovered of an IFNγ ELISPOT assay. This study is therefore supporting by previous authors using in silico prediction followed by ex- previous finding that influenza vaccines for pigs based on vivo IFNγ ELISPOT testing (8). In the present study, no SLA-II WIV and formulated with adjuvant are able to generate T cell epitope producing an IFNγ recall response was identified in the responses with IFNγ secreting effector functions; moreover, M1 protein whereas the epitopes found in the NP were located it supports the generation of T cell responses with cross in a different region compared to the one previously published reactivity against viruses of different linages For this purpose, (NP253−277 ). These differences might have been due to the allele Frontiers in Immunology | www.frontiersin.org 8 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV diversity between the cohorts of pigs of the two studies, the The analysis of swine IAV showed an impressive conservation SLA low resolution alleles found by the authors were different of the NP406−419 sequence which suggests this region might compared to Babraham’s pigs. be an optimal candidate for development of a broad reacting Despite the absence of similar epitopes in pigs, several vaccine for pigs. Nevertheless, conservation of the sequence in human MHC-II epitopes were previously identified by other other mayor species is also an important factor to take into authors. Notably, most of those were restricted to HLA-DRB1 account for the design of a broad protecting vaccine. IAV has alleles and raised the question whether this affinity might the ability to cross the specie barrier whereas pigs have the have been conserved in pigs as well. The pan algorithm ability to act as a “mixing vessel” and also to perpetrate other NetMHC II 3.2 was used to predict epitopes of the NP of species lineages. Human and avian species are the other two A/Panama/2007/1999 associated to the Babraham’s pig SLA-II major known host of IAV; these might act as reservoir of allele. Notably, two 9mer epitopes were predicted for the SLA- novel variants that might escape from the pig immune system DRB1, NP412−420 FSVQRNLPF and NP406−414 ISVQPTFSV; recognition of NP406−419 . Conservation analysis showed that these epitopes were partially contained in the sequences of this sequence was less conserved in human and avian IAV; in epitopes NP406−416 and NP407−419 identified in this study. A particular, the site NP406 showed the lowest conservation in proof of concept test conducted with low power showed that both lineages, especially in the human. Notably, results showed the immunogenic response produced by peptides containing the that the mutation I406T is highly frequent in human IAV region NP406−416 was not restricted to the DQ allele but instead lineage and it was particularly associated with H3N2 strains by the DRB allele. Therefore, these results support the idea that circulating since the ‘00s. During the same period a transition to those epitopes might be associated to the SLA-DRB1 allele in a novel antigenic cluster [A/Fujian/411/2002 (FU02)] linked to Babraham’s pigs, however, further study are required to validate a persistent reassortment event was observed (28). Considering the in silico prediction. the common ancestor between the avian IAV and other species Different functionally polarized subpopulations have been lineages it was not surprising to find the conserved amino acid established in studies with murine and human CD4+ T cells Isoleucine at the position NP406 ; however, results showed that and just a part of these were elucidated in pigs (26). This is a the divergent evolution of IAV in human hosts is apparently mayor limitation in the SLA-II epitope discovery strategies in pigs changing permanently its composition. NP406 was a critical whereby cytokines are biomarker of cells responsiveness. In the position to maintain the immunogenic properties of the epitope present study, IFNγ was used as functional marker to quantify NP406−416 and actually its removal disrupted the capacity of the the recall response to the overlapping IAV peptide collections T cell epitope to recall an IFNγ response in the immunized pigs. in peripheral blood. This cytokine is generally associated in pigs The effects of mutations at NP406 on the immune proprieties to the Th1 subset of the CD4+ T cells (26). Single and double of the T cell epitope are unknown; because of the variability producer of IFNγ and TNFα dominated the response of CD4+ and evolution of the IAV, this position deserve further studies to T cells in blood and other tissues during an IAV infection in evaluate its antigenic tolerance. pigs and thus they were suggested to be important in its control The overall high conservation of NP406−419 is probably (6). Therefore, the immunogenicity of NP406−416 and NP407−419 due to its functional importance in the NP protein. The during infection and their involvement in protection deserve region is localized into a flexible tail-loop (residues 402–428) further studies. The encounter SLA-II epitopes might be also which mediate the NP-NP oligomerization (29). In this region, helpful under other effector functions. The phenotype of CD4+ positions like the R416 can be found which is critical for T cells depends on the “cytokine milieu” (27) and thus this might the functionality of the protein (29); therefore, this region be rationally modulated to a specific phenotype involved in other is expected to be less permissive to mutations. Thus, region T helper functions. NP406−419 can be considered an optimal candidate for vaccines Sequence conservation of epitopes is an important aspect to research due to the presence of T cells epitopes and its take into account during the rational design of an IAV vaccine. high conservation. The use of conserved epitopes could contribute to the breadth of Considering the high conservation in the epitope, a protection conferred by the vaccine. The SLA-II T cell epitopes proof of concept test was conducted with low power to identified in this study were hailing from the human IAV strain evaluate the cross reactivity of the identified immunogenic A/Panama/2007/1999 (H3N2) and despite this, they showed to region NP406−416 . Results showed that this region showed be able to induce a cross reacting response in pigs immunized to be immunogenic in pigs even in a different immunity with IAV of completely different lineages. This fact was due to scenario involving the A(H1N1) pdm/09 influenza the complete conservation of their amino acid sequences among vaccines and thus support its validity as candidate for the antigens used in this study. NP and M1 proteins are codified further studies. by the internal cassette of genes of IAV which is by far more conserved compared to the HA and NA coding genes. Notably, DATA AVAILABILITY STATEMENT the internal cassette of the European swine lineages of the H3N2 and H1N2 subtypes is even more conserved because it was shared The raw data supporting the conclusions of this article will be by the H1N1 avian like lineage through reassortment events. made available by the authors, without undue reservation. Frontiers in Immunology | www.frontiersin.org 9 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV ETHICS STATEMENT FP7-228393 (a NADIR project); from the project AGL2010- 22200-C02-01 from the Spanish Ministry of Science and The animal study was reviewed and approved by The Pirbright Innovation; the UK’s BBSRC grant BBS/E/I/00002014. Institute ethics committee. AUTHOR CONTRIBUTIONS ACKNOWLEDGMENTS MM and BC: funding acquisition. MM: supervision and We acknowledge the Animal Services staff at The Pirbright conceptualization. MB, SM, JH, ER, BVC, EL, SM-T, LF, and ET: Institute for husbandry and sampling. The authors thanks Javier investigation and methodology. MB and MM: writing-original Dominguez for kindly providing monoclonal antibodies; S.C. draft. All authors contributed to the article and approved the Gilbert for providing the first collection of peptides; Alan submitted version. Townsend for providing S-Flu virus. FUNDING SUPPLEMENTARY MATERIAL The research leading to these results has received funding from The Supplementary Material for this article can be found the European Community’s Seventh Framework Program (FP7, online at: https://www.frontiersin.org/articles/10.3389/fimmu. 2007-2013), the Research Infrastructures Action under grant 2020.02083/full#supplementary-material REFERENCES 13. Baratelli M, Cordoba L, Perez LJ, Maldonado J, Fraile L, Nunez JI, et al. Genetic characterization of influenza A viruses circulating in pigs and isolated 1. Reeth KV, Vincent AL. Influenza viruses. In: Zimmerman JJ, Karriker LA, in north-east Spain during the period 2006-2007. Res Vet Sci. (2014) 96:380–8. Ramirez A, Schwartz KJ, Stevenson GW, Zhang J, editors. Diseases of Swine. doi: 10.1016/j.rvsc.2013.12.006 Eleventh ed. Hoboken, NJ: Wiley-Blackwel (2019). p. 576–93. 14. Powell TJ, Silk JD, Sharps J, Fodor E, Townsend AR. Pseudotyped influenza 2. Ito T, Couceiro JN, Kelm S, Baum LG, Krauss S, Castrucci MR, et al. Molecular A virus as a vaccine for the induction of heterotypic immunity. J Virol. (2012) basis for the generation in pigs of influenza A viruses with pandemic potential. 86:13397–406. doi: 10.1128/JVI.01820-12 J Virol. (1998) 72:7367–73. doi: 10.1128/JVI.72.9.7367-7373.1998 15. Blanco E, Mccullough K, Summerfield A, Fiorini J, Andreu D, Chiva C, et al. 3. Smith GJ, Vijaykrishna D, Bahl J, Lycett SJ, Worobey M, Pybus OG, et al. Interspecies major histocompatibility complex-restricted Th cell epitope on Origins and evolutionary genomics of the 2009 swine-origin H1N1 influenza foot-and-mouth disease virus capsid protein VP4. J Virol. (2000) 74:4902–7. A epidemic. Nature. (2009) 459:1122–5. doi: 10.1038/nature08182 doi: 10.1128/JVI.74.10.4902-4907.2000 4. Fox A, Quinn KM, Subbarao K. Extending the breadth of influenza vaccines: 16. Lefevre EA, Carr BV, Inman CF, Prentice H, Brown IH, Brookes SM, et al. status and prospects for a universal vaccine. Drugs. (2018) 78:1297–308. Immune responses in pigs vaccinated with adjuvanted and non-adjuvanted doi: 10.1007/s40265-018-0958-7 A(H1N1)pdm/09 influenza vaccines used in human immunization 5. Crisci E, Mussa T, Fraile L, Montoya M. Review: influenza virus in pigs. Mol programmes. PLoS ONE. (2012) 7:e32400. doi: 10.1371/journal.pone.00 Immunol. (2013) 55:200–11. doi: 10.1016/j.molimm.2013.02.008 32400 6. Talker SC, Stadler M, Koinig HC, Mair KH, Rodriguez-Gomez IM, Graage 17. Jensen KK, Andreatta M, Marcatili P, Buus S, Greenbaum JA, Yan Z, et al. R, et al. Influenza A virus infection in pigs attracts multifunctional Improved methods for predicting peptide binding affinity to MHC class II and cross-reactive T cells to the lung. J Virol. (2016) 90:9364–82. molecules. Immunology. (2018) 154:394–406. doi: 10.1111/imm.12889 doi: 10.1128/JVI.01211-16 18. Schwartz JC, Hemmink JD, Graham SP, Tchilian E, Charleston B, Hammer SE, 7. Talker SC, Koinig HC, Stadler M, Graage R, Klingler E, Ladinig A, et al. et al. The major histocompatibility complex homozygous inbred Babraham Magnitude and kinetics of multifunctional CD4+ and CD8beta+ T cells pig as a resource for veterinary and translational medicine. HLA. (2018) in pigs infected with swine influenza A virus. Vet Res. (2015) 46:52. 92:40–3. doi: 10.1111/tan.13281 doi: 10.1186/s13567-015-0182-3 19. Zhang Y, Aevermann BD, Anderson TK, Burke DF, Dauphin G, Gu 8. Gutierrez AH, Loving C, Moise L, Terry FE, Brockmeier SL, Hughes Z, et al. Influenza Research Database: An integrated bioinformatics HR, et al. In vivo validation of predicted and conserved T cell resource for influenza virus research. Nucleic Acids Res. (2017) 45:D466–74. epitopes in a swine influenza model. PLoS ONE. (2016) 11:e0159237. doi: 10.1093/nar/gkw857 doi: 10.1371/journal.pone.0159237 20. Katoh K, Rozewicki J, Yamada KD. MAFFT online service: multiple sequence 9. Baratelli M, Pedersen LE, Trebbien R, Larsen LE, Jungersen G, Blanco E, et al. alignment, interactive sequence choice and visualization. Brief Bioinform. Identification of cross-reacting T-cell epitopes in structural and non-structural (2019) 20:1160–6. doi: 10.1093/bib/bbx108 proteins of swine and pandemic H1N1 influenza A virus strains in pigs. J Gen 21. Waterhouse AM, Procter JB, Martin DM, Clamp M, Barton GJ. Virol. (2017) 98:895–9. doi: 10.1099/jgv.0.000748 Jalview Version 2–a multiple sequence alignment editor and analysis 10. Nicholls S, Pong-Wong R, Mitchard L, Harley R, Archibald A, Dick A, workbench. Bioinformatics. (2009) 25:1189–91. doi: 10.1093/bioinformatics/b et al. Genome-wide analysis in swine associates corneal graft rejection with tp033 donor-recipient mismatches in three novel histocompatibility regions and one 22. Crooks GE, Hon G, Chandonia JM, Brenner SE. WebLogo: a sequence logo locus homologous to the mouse H-3 locus. PLoS ONE. (2016) 11:e0152155. generator. Genome Res. (2004) 14:1188–90. doi: 10.1101/gr.849004 doi: 10.1371/journal.pone.0152155 23. Sant AJ, Dipiazza AT, Nayak JL, Rattan A, Richards KA. CD4 T 11. Chen L, Zanker D, Xiao K, Wu C, Zou Q, Chen W. Immunodominant cells in protection from influenza virus: Viral antigen specificity and CD4+ T-cell responses to influenza A virus in healthy individuals focus on functional potential. Immunol Rev. (2018) 284:91–105. doi: 10.1111/imr. matrix 1 and nucleoprotein. J Virol. (2014) 88:11760–73. doi: 10.1128/JVI.0 12662 1631-14 24. Olson ZF, Sandbulte MR, Souza CK, Perez DR, Vincent AL, Loving CL. 12. Sridhar S. Heterosubtypic T-cell immunity to influenza in humans: challenges Factors affecting induction of peripheral IFN-gamma recall response to for universal T-cell influenza vaccines. Front Immunol. (2016) 7:195. influenza A virus vaccination in pigs. Vet Immunol Immunopathol. (2017) doi: 10.3389/fimmu.2016.00195 185:57–65. doi: 10.1016/j.vetimm.2017.01.009 Frontiers in Immunology | www.frontiersin.org 10 September 2020 | Volume 11 | Article 2083 Baratelli et al. New SLA-II Epitope of SwIV 25. Tungatt K, Dolton G, Morgan SB, Attaf M, Fuller A, Whalley T, et al. Induction antiviral drug target. Curr Top Med Chem. (2017) 17:2271–85. of influenza-specific local CD8 T-cells in the respiratory tract after aerosol doi: 10.2174/1568026617666170224122508 delivery of vaccine antigen or virus in the Babraham inbred pig. PLoS Pathog. (2018) 14:e1007017. doi: 10.1371/journal.ppat.1007017 Conflict of Interest: SM-T was employee of Innovex Therapeutics S.L. 26. Gerner W, Talker SC, Koinig HC, Sedlak C, Mair KH, Saalmuller A. Phenotypic and functional differentiation of porcine alphabeta T cells: The remaining authors declare that the research was conducted in the absence of current knowledge and available tools. Mol Immunol. (2015) 66:3–13. any commercial or financial relationships that could be construed as a potential doi: 10.1016/j.molimm.2014.10.025 conflict of interest. 27. Swain SL, Mckinstry KK, Strutt TM. Expanding roles for CD4(+) T cells in immunity to viruses. Nat Rev Immunol. (2012) 12:136–48. Copyright © 2020 Baratelli, Morgan, Hemmink, Reid, Carr, Lefevre, Montaner- doi: 10.1038/nri3152 Tarbes, Charleston, Fraile, Tchilian and Montoya. This is an open-access article 28. Westgeest KB, Russell CA, Lin X, Spronken MI, Bestebroer TM, Bahl J, et al. distributed under the terms of the Creative Commons Attribution License (CC BY). Genomewide analysis of reassortment and evolution of human influenza The use, distribution or reproduction in other forums is permitted, provided the A(H3N2) viruses circulating between 1968 and 2011. J Virol. (2014) 88:2844– original author(s) and the copyright owner(s) are credited and that the original 57. doi: 10.1128/JVI.02163-13 publication in this journal is cited, in accordance with accepted academic practice. 29. Hu Y, Sneyd H, Dekant R, Wang J. Influenza A virus nucleoprotein: No use, distribution or reproduction is permitted which does not comply with these a highly conserved multi-functional viral protein as a hot terms. Frontiers in Immunology | www.frontiersin.org 11 September 2020 | Volume 11 | Article 2083