Characterization of the humoral immune response of experimentally infected and vaccinated pigs to swine influenza viral proteins

Transcription

1 Arch Virol (2005) DOI /s Characterization of the humoral immune response of experimentally infected and vaccinated pigs to swine influenza viral proteins W.-I. Kim 1, W.-H. Wu 2,B.Janke 2, and K.-J. Yoon 1,2 1 Department of Veterinary Microbiology and Preventive Medicine, College of Veterinary Medicine, Iowa State University, Ames, Iowa, USA 2 Department of Veterinary Diagnostic and Production Animal Medicine, College of Veterinary Medicine, Iowa State University, Ames, Iowa, USA Received January 25, 2005; accepted July 6, 2005 Published online August 23, 2005 c Springer-Verlag 2005 Summary. The value of serologic tests for diagnosis of swine influenza virus (SIV) infection has been diminished by the emergence of new subtypes and by antigenic drift within subtype. The intensive use of vaccination also has complicated interpretation of serology results. Serologic assays are needed that can detect infection regardless of subtype or antigenic variation and that can differentiate antibody induced by infection from that induced by vaccination. In this study, the antibody responses to specific viral proteins in pigs infected by or vaccinated for SIV were characterized by Western immunoblot. Both IgM and IgG against hemagglutinin, nucleoprotein, NS1 and NS2 were detected in experimentally infected pigs by 7 days post inoculation (DPI). IgG against these proteins was still detectable at the end of the study (28 DPI). In contrast, IgG against neuraminidase and M1 was not detected until 14 DPI and no IgM against these proteins was detected. In vaccinated pigs, no antibody against NS1 was detected while antibody responses to other proteins were identical to those in exposed pigs. In conclusion, nucleoprotein may be a suitable antigen for use in a subtype-unrestricted serologic assay. NS1 protein may be suitable for a serologic assay that differentiates between infected and vaccinated pigs. Introduction Swine influenza, initially recognized in the early 1900 s, has persisted as an economically significant respiratory disease that affects pigs of all ages throughout the world [28]. Pregnant animals may also abort due to the high fever induced by infection. Because of the detrimental economic effect of the disease, vaccination of

2 W.-I. Kim et al. gilts and sows with inactivated vaccines has been commonly practiced to prevent reproductive loss in breeding animals and respiratory disease in younger pigs. The virus responsible for swine influenza is a member of the genus Influenzavirus A in the family Orthomyxoviridae [8]. Swine influenza virus (SIV) contains 8 RNA segments which encode 4 major structural proteins [hemagglutinin (H), neuraminidase (N), nucleoprotein (NP), and matrix (M1 and M2) proteins], 3 subunits of RNA-dependent RNA polymerases (PA, PB1 and PB2) and non-structural proteins (NS1 and NS2). The NP and M1 proteins are internal structural proteins and serve as highly conserved antigenic determinants for virus type (A, B, C). The NP protein coats RNA segments to form ribonucleoproteins (RNPs). The M1 protein functions as the scaffold of the virus and plays a role in the nuclear export of the RNPs [11, 21]. M2 protein is transcribed from the M segment through gene splicing and acts as an ion channel that favors virus entry [9]. The H and N proteins are surface proteins that play a critical role in virus entry into and release from target cells, respectively. These proteins determine virus subtype. Although 16 H and 9 N subtypes are known to exist among avian and mammalian type A influenza viruses [8], only 3 subtypes of influenza A viruses, H1N1, H1N2 and H3N2, have been consistently implicated in swine influenza. The NS1 protein is expressed only during infection and has been shown to inhibit interferon production by the host [23, 29, 32]. NS2 protein, which is expressed from the NS1 gene through gene splicing, is a part of the virion and functions as Nuclear Export Protein (NEP) for translocation of RNPs [11, 21, 25, 29]. A definitive diagnosis of swine influenza requires detection of virus or viral antigen in the tissues or secretions of clinically affected animals. However, because the disease has a very short course and the causative agent becomes undetectable in infected animals quickly after infection [37], serological assays, such as hemagglutination inhibition (HI) test, serum-virus neutralization (SVN) test, indirect fluorescent antibody (IFA) test or enzyme-linked immunosorbent assay (ELISA), are often employed to identify animals that have been exposed to the virus. Serology also is used to assess the immune status of pigs at various stages of production to determine optimal time of vaccination. However, interpretation of serologic data is frequently confounded by factors such as antigenic differences between the virus that induced the antibody and the virus used in the serologic test [17] and the difficulty in differentiating whether detected antibody is of maternal origin or was induced by infection or vaccination [24]. Until the late 1990 s, only classical H1N1 circulated in US swine populations, but since 1998 new reassortants (H3N2, H1N2) have appeared and become endemic [34]. Furthermore, antigenic variants occur within the same subtype [34, 35]. These observations emphasize the need for serologic assays that can detect infection regardless of subtype or antigenic variation and that can differentiate between infection and vaccination. As a first step toward this goal, antibody response of pigs to SIV infection and vaccination were assessed over time. In the following study, the humoral immune response of young swine to SIV infection and vaccination was characterized using Western immunoblot analysis to assess relative immunogenicity of influenza viral proteins and to identify possible antigen(s) for new serodiagnostic assays for SIV.

3 Western immunoblot analysis of antibody responses to SIV Materials and methods Viruses and cell Two SIV field isolates, A/Swine/IA/40776/92 (IA92, classical H1N1) and A/Swine/IA/41305 /98 (IA98, cluster I H3N2) were used for this study. Both viruses were isolated originally from pigs in commercial swine operations undergoing severe respiratory disease. Initial virus isolation was conducted in 9- to 10-day-old embryonated chicken eggs [6]. The isolates induced clinical respiratory disease in experimentally inoculated pigs [14, 31]. Madin-Darby Canine Kidney (MDCK) cells were used to propagate the viruses [30]. Monolayers of MDCK cells were prepared under Minimum Essential Medium (MEM, Sigma, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS, Sigma), 20 mm L-glutamine, 100 IU/ml penicillin, 100 µg/ml streptomycin, 50 µg/ml gentamicin and 0.25 µg/ml amphotericin B (hereafter, MEM-GM) at 37 C in a water-jacketed CO 2 incubator. Before inoculation of virus, cell monolayers were rinsed with Earle s Balanced Salt Solution (EBSS, Sigma), containing 0.2 µg/ml of TPCK-treated trypsin (Worthington, Lakewood, NJ, USA), 100 IU/ml penicillin, 100 µg/ml streptomycin, 50 µg/ml gentamicin, and 0.25 µg/ml amphotericin B (EBSS-trypsin). The cells were inoculated with 1 hemagglutination (HA) unit of virus and incubated for 2 hrs at 37 C in the incubator. After incubation, the cells were rinsed three times with EBSS-trypsin and covered with MEM-GM supplemented with 0.5% bovine serum albumin (BSA, Sigma) instead of FBS. BSA was removed from the cell culture medium when preparing the viral proteins for Western immunoblot analysis because during preliminary study BSA was found to block antibody adherence to some viral proteins, particularly H and N proteins. After further incubation for 24 to 36 hrs at 37 C, each virus was harvested and the virus titer of each preparation was determined by HA assay. Serum samples and reference antibodies Antibody-positive serum samples (n = 18 per virus) were collected from 5-week-old conventional pigs experimentally inoculated with H1N1 (IA92) or H3N2 (IA98) SIV. Inoculum (2 ml) consisting of lung homogenate prepared from gnotobiotic pigs infected with the virus was administered via nebulization through a nose cone. All animals were bled on days 0, 7, 14, and 28 post inoculation (PI). Antibody-negative serum samples (n = 18) consisted of day 0 serum samples from the 12 pigs described above and from 6 sham-inoculated control pigs in the same study. Sera with vaccine-induced antibody were collected from 5 pigs vaccinated with a commercial bivalent (H1N1 and H3N2) SIV vaccine (End-FLUence, Intervet, Grand Island, NY, USA) intramuscularly twice at a 2-week interval as recommended by the manufacturer. The sera were collected 2 weeks after each vaccination. Positive control reference antisera from hyperimmunized gnotobiotic pigs against H1N1 (A/Sw/IA/73) and H3N2 (A/Sw/TX/98, cluster I) [33] were purchased from the National Veterinary Services Laboratories (NVSL), Ames, Iowa, USA. Murine monoclonal antibody (mab) specific for the M1 protein of an H3N2 human influenza virus (ViroStat, Portland, ME, USA) also was employed as a reference antibody. Preparation of crude viral antigen When cytopathic effect was evident in approximately 50% of the monolayer, the flasks were subjected to one cycle of freeze-thawing at 80 C/37 C, and the cells and supernatant pelleted by ultracentrifugation for 3 hrs at 100,000 g. The resulting pellet was resuspended in 500 µl of Tris lysis buffer (10 mm Tris, 0.04% CHAPS, 0.15 M NaCl, % aprotinin and 1 mm EDTA), ph 8.0, and incubated overnight at 4 C with gentle stirring, then stored at

4 W.-I. Kim et al. 80 C. Mock-infected cells were prepared similarly and used as cell control antigen. Protein concentration was determined with DC TM Protein Assay (Bio-Rad, Hercules, CA, USA). Preparation of recombinant SIV antigens To construct baculovirus expression vectors containing NS1, NP, N1 or N2 genes, the specific cdna was first amplified from SIV genomic RNA by a reverse transcription-polymerase chain reaction (RT-PCR). H1N1 SIV was the donor of NS1, NP and N1 genes and H3N2 SIV the donor of N2 gene. Each PCR product was sequenced for verification and then cloned into a plasmid vector pgem-t-easy (Promega, Madison, WI, USA). The sequences of the PCR primers were: NS1-forward 5 AAGCAAAAGCAGGGAAAATAA3 ; NS1-reverse 5 AGTAGAAACAAGGGTGTTTTT3 ; NP-forward 5 AGCAAAAGCAGGGTAGATAAT3 ; NP-reverse 5 TTCTTCTTTAATTGTCATACT3 ; N1 and 2-forward 5 AGCAAAAGCAGGAGTTTAAAAT3 ; and N1 and 2-reverse 5 AGTAGAAACAAGGAGTTTTTT-3. Each PCR product in pgem-t-easy vector and the baculovirus protein expression plasmid pfastbac1 were digested by restriction enzymes, electrophoresed on an agarose gel, purified and ligated to each other at corresponding restriction enzyme sites. NS1 fragment was digested with SalI and SphI, NP and N2 with EcoRI and SpeI, and N1 with EcoRI and XbaI, respectively. Each digested fragment was inserted into the corresponding site of pfastbac1 (Invitrogen, Carlsbad, CA, USA). After transformation, pfbns1, pfbnp, pfbn1 and pfbn2, which were used to express NS1, NP, N1 and N2 respectively in the baculovirus expression system, were cloned. DNA sequence analysis was used to confirm the authenticity of each plasmid construct. The pfbns1, pfbnp, pfbn1 and pfbn2 were then individually transformed into DH10Bac TM, and recombinant baculoviruses AcNS1, AcNP, AcN1, and AcN2 were constructed by transposition into AcMNPV genomic DNA according to the manufacturer s instructions ( BAC-TO-BAC TM manual, Invitrogen). Recombinant viruses were inoculated into Sf-9 insect cells (Invitrogen) and propagated for 4 days. The cells were then lysed in the Tris lysis buffer and used for SDS-PAGE after the protein concentration of each preparation had been determined. Hemagglutination inhibition test Antibody response of the animals was assessed by HI test using the standard protocol recommended by the Centers for Disease Control [4]. Each serum sample was tested for HI activity against H1 and H3 SIV. The HI antibody titer of each sample was expressed as the reciprocal of the highest dilution at which no HI was observed. Samples with no HI activity at 1:10 dilution were considered negative for HI antibody. SDS-PAGE and Western immunoblot Viral and cellular or recombinant proteins were electrophoretically separated on vertical 12% polyacrylamide slab gels under non-reducing conditions and electrotransferred to nitrocellulose membranes (Bio-Rad). In brief, each viral antigen, cellular antigen or recombinant protein preparation was mixed with 5 sample buffer containing 0.6 M Tris-HCl (ph 6.8), 25% glycerol, 2% SDS and 0.1% bromophenol blue, and incubated for 5 min at 65 C [7] or 5 min at 100 C for recombinant proteins. 100 µl of each sample were loaded on each gel and electrophoresed for 80 min at the constant voltage of 100 and pre-stained molecular weight markers (Bio-Rad). After electrophoresis, polypeptides and markers separated on each gel

5 Western immunoblot analysis of antibody responses to SIV were electrotransferred onto a 0.2 µm nitrocellulose membrane overnight at 30 volts under chilled conditions. Each membrane was blocked with a 1% skim milk solution by submerging the membrane in the solution for 30 min at ambient temperature. Each membrane was then cut into 0.5 cm-wide strips, dried at ambient temperature and stored at 20 C until use. For Western immunoblot analysis, a pair of nitrocellulose membrane strips (one with viral antigen and one with cellular antigen) was incubated with each serum sample diluted 1:50 in 20 mm Tris-buffered saline solution with 0.05% Tween 20 (TBST, ph 7.5) for 90 min at ambient temperature with slow rocking, and then rinsed at least 3 times using TBST. Each pair of strips was incubated with goat anti-swine IgG or IgM labeled with peroxidase (Kirkegaard- Perry Laboratories, Gaithersburg, MD, USA) for 60 min at ambient temperature. Antigenantibody reactions were visualized by adding TMB substrate (Kirkegaard-Perry Laboratories) to each strip and incubated for 5 min at ambient temperature. Colorimetric reaction was stopped by rinsing the strips with distilled water. Results HI antibody responses to infection and vaccination All inoculated animals developed HI antibody only to the subtype with which they had been inoculated. Antibody titers at day 7 PI ranged from 1:320 to 1:640. All pigs were still seropositive (1:160 1:320) at the termination of the study (4 weeks PI) although titers had gradually declined after 7 days PI. All vaccinated animals were positive for HI antibody against SIV of both subtypes. HI titers induced by vaccination were not detected at 2 weeks after the first vaccination but were present 2 weeks after the second vaccination, with titers ranging from 1:80 to 1:320, similar to the previous reports [10, 13]. Protein profile of SIV Three distinct viral polypeptides were visualized from SIV-infected cell lysate using the hyperimmune polyclonal anti-siv reference sera. These polypeptides were determined to be H, N or NP, and M1, based on their observed molecular weight, electrophoretic pattern similar to that of recombinant NP and N proteins, and/or reactivity to mab specific for M1 protein (Fig. 1). The molecular mass of H protein of both H1N1 and H3N2 SIV was estimated to be 82 kda under non-reducing conditions. Each of the hyperimmune H1N1 and H3N2 reference antisera recognized the corresponding H protein (Fig. 1A). The molecular mass of both N and NP proteins was estimated to be 58 kda although the size of N and NP proteins was expected to be 50 and 55 kda, respectively, based on their deduced amino acid sequences. The bands were indistinguishable from each other by Western immunoblot even when baculovirus recombinant N and NP proteins were used as antigen (Fig. 1B). The subtypespecific hyperimmune reference antisera recognized the corresponding recombinant N protein but reacted with the recombinant NP in a subtype-unrestricted manner (Fig. 1A). The molecular mass of M1 protein was estimated to be 30 kda (Fig. 1A). The H1N1 hyperimmune reference antiserum reacted well with M1 protein from

6 W.-I. Kim et al. Fig. 1. Protein profile of swine influenza virus (SIV) as determined by Western immunoblot. Labels at the top and bottom of each panel represent antibody and antigen used, respectively. A Lysates of cells infected with SIV of each subtype (H1N1 or H3N2) and baculovirusexpressed recombinant (r) N1 and N2 proteins were separated by SDS-PAGE, electrotransferred to a nitrocellulose membrane, and subjected to immunoblot with H1N1 or H3N2 reference antiserum and matrix (M1) protein-specific monoclonal antibody. B Recombinant N1, N2, NP and NS proteins were separated by SDS-PAGE, electrotransferred to a nitrocellulose membrane, and subjected to immunoblot with mixed reference antiserum (H1N1 and H3N2) both subtypes. In contrast, the H3N2 hyperimmune reference antiserum appeared to have a minimal titer of antibody specific for M1 protein based on the band intensity in Western immunoblot. Since both the H1N1 reference antiserum and M1-specific mab demonstrated the presence of M1 protein in antigen preparations of both subtypes (Fig. 1A), the minimal reactivity with M1 protein by the H3N2 reference antiserum was not attributed to the difference in the amount of M1 protein between the 2 viral antigen preparations. The cause of this unexpected result remains to be further investigated. Two proteins were expressed from the same NS gene, due to internal splicing of mrna (Fig. 1). The smaller protein was considered to be NS2 rather than degraded NS1 product since only NS1 protein was expressed from the NS gene cloned into a prokaryotic system (data not shown) and the molecular size of the smaller protein was identical to that calculated from the amino acid sequence of NS2. The molecular mass of NS1 and NS2 proteins was estimated to be 28 kda

7 Western immunoblot analysis of antibody responses to SIV and 17 kda, respectively. The amount of NS1 and NS2 proteins present in the crude viral antigen preparation appeared to be lower than the detection limit of Western immunoblot as the reference antisera recognized the proteins only when recombinant NS protein was employed. Western immunoblot analysis of antibody responses to SIV infection The viral protein specificity of antibody responses in experimentally infected pigs as determined by Western immunoblot analysis are summarized and illustrated in Table 1 and Fig. 2, respectively. The antibody response of the infected pigs to the viral surface proteins H and N was subtype-specific. IgM antibody against H protein was detected in all pigs on day 7 PI. Based on the intensity of the band, the level of IgM antibody to H protein started to decrease after 14 days PI. IgG antibody specific for H protein was not detected in all pigs until day 14 PI, although weak IgG antibody response to H protein initially appeared by day 7 PI in some of the pigs. IgG antibody against H protein was still present on the last sampling day (28 days PI). In contrast, no IgM antibody against N protein was detected in any Table 1. Appearance of swine influenza viral protein-specific antibodies in pigs experimentally inoculated with H1N1 or H3N2 SIV as determined by Western immunoblot analysis Antibody SIV Days post infection isotype protein IgM H1 0/0 a 6/0 2/0 0/0 H3 0/0 0/6 0/6 0/4 N1 b 0/0 0/0 0/0 0/0 N2 b 0/0 0/0 0/0 0/0 NP b 0/0 6/6 2/6 2/6 M1 0/0 0/0 0/0 0/0 NS1 b 0/0 6/6 6/6 2/6 NS2 b 0/0 6/6 6/6 2/6 IgG H1 0/0 2/0 6/0 6/0 H3 0/0 0/2 0/6 0/6 N1 b 0/0 0/0 6/0 6/0 N2 b 0/0 0/0 0/6 0/6 NP b 0/0 6/6 6/6 6/6 M1 0/0 0/0 6/4 6/6 NS1 b 0/0 6/6 6/6 6/6 NS2 b 0/0 6/6 6/6 6/6 a The number of pigs (H1N1/H3N2 infected) seropositive for antibody against the given viral protein at the given day post inoculation b Antibody response was assessed with recombinant protein

8 W.-I. Kim et al.

9 Western immunoblot analysis of antibody responses to SIV of the inoculated pigs during the study period (28 days PI) whereas IgG antibody specific for N protein was detected by day 14 PI in all animals and thereafter until the termination of the study. Antibody responses to the internal proteins NP and M1 were similar for both subtypes of SIV inoculum. No pigs developed detectable IgM antibody specific for M1 protein during the study period. IgG antibody response to M1 protein was not detected until 14 days PI. In contrast, both IgM and IgG antibodies against NP protein were detected in all inoculated pigs on day 7 PI. NP-specific IgM antibody persisted in several pigs and IgG antibody in all pigs until the end of the study. Table 2. Appearance of swine influenza viral protein-specific antibodies in pigs vaccinated with a commercial inactivated bivalent SIV vaccine as determined by Western immunoblot analysis Antibody SIV Days post the first vaccination isotype protein IgM H1 & H3 a 0/5 b 0/5 0/5 N1&N2 a,c 0/5 0/5 0/5 NP c 0/5 5/5 5/5 M1 0/5 0/5 0/5 NS1 c 0/5 0/5 0/5 NS2 c 0/5 0/5 0/5 IgG H1 & H3 a 0/5 0/5 5/5 N1&N2 a,c 0/5 0/5 0/5 NP c 0/5 5/5 5/5 M1 0/5 0/5 0/5 NS1 c 0/5 0/5 0/5 NS2 c 0/5 0/5 0/5 a Antibody responses to both subtypes were identical b The number of pigs seropositive for antibody against the given viral protein/the total number of pigs tested at the given day post inoculation c Antibody response was assessed with recombinant protein Fig. 2. Representative Western immunoblot analysis of antibody response of pigs to SIV using sera collected weekly from animals experimentally infected with SIV H1N1 (A, C, D and E) or H3N2 (B and F). Protein specificity of antibody response of the pigs was determined using 5 antigen preparations: lysate of cells infected with H1N1 (A) or H3N2 SIV (B), baculovirusexpressed recombinant neuraminidase (rn1, C), nucleoprotein (rnp, D) and nonstructural protein (rns, E and F) of H1N1 SIV. Antibody responses of the pig to each viral protein were differentiated for isotypes (i.e., IgM and IgG) using isotype-specific secondary antibody. Hyperimmune pig sera raised against reference H1N1 or H3N2 SIV strain or a commercially available murine monoclonal antibody specific for M protein of a human influenza A virus were included as homologous positive control (A and B) or for differential purposes (B and C). Only IgG response is shown for rn1 since no IgM response was detected (C)

10 W.-I. Kim et al. Antibody responses to NS1 and NS2 proteins also were similar for both subtypes of SIV inoculum. Both IgM and IgG antibodies specific for NS1 and NS2 recombinant proteins were detected in all inoculated pigs on day 7 PI. IgG antibody continued to be detected until 28 day PI while IgM response decreased with time and was very weak on day 28 PI. Western immunoblot analysis of antibody responses to SIV vaccination The viral protein specificity of antibody response in pigs vaccinated with a bivalent killed product is summarized and illustrated in Table 2 and Fig. 3, respectively. The antibody response of the vaccinated animals to the surface proteins (H and N) of SIV after 2 doses of the vaccine was directed mainly against H protein. While no IgM antibody specific for H protein was detected in any of the sera collected after a 2-dose vaccination, H protein-specific IgG antibody was detected in the animals at 2 weeks after the second dose. In contrast, no IgM or IgG antibodies specific for N protein were detected in the sera with either native or recombinant N proteins used as antigen. Weak reactivity of antibody to the internal proteins (NP and M1) was observed by Western immunoblot when crude viral antigen was used. When recombinant NP was employed, both IgM and IgG antibodies specific for NP protein were detected Fig. 3. Representative Western immunoblot analysis of antibody response of pigs to SIV using sera collected biweekly from pigs vaccinated with a commercial inactivated bivalent vaccine. Protein specificity of antibody response of the pigs was determined using two antigens: lysate of cells infected with H1N1 SIV (A) and baculovirus-expressed recombinant nucleoprotein (B). Antibody responses to each viral protein were differentiated for isotype (IgM and IgG) using isotype-specific secondary antibody. Hyperimmune pig sera raised against reference H1N1 was included as an homologous positive control (A)

11 Western immunoblot analysis of antibody responses to SIV in all vaccinated pigs at 2 weeks after the first vaccination. IgG response to NP protein was rather weak at 2 weeks after the first vaccination, but bands increased in intensity after the second vaccination, suggesting increase in antibody titer against NP protein. Neither IgM nor IgG antibody against NS1 or NS2 protein was detected in any of the sera collected from vaccinated pigs during the study period. Discussion The hemagglutination inhibition test has been used routinely for serodiagnosis of SIV infection in veterinary diagnostic laboratories in North America because the assay is relatively inexpensive and the subtype and antigenicity of SIV in the U.S. swine population has been very stable. Recently, the dependability of this subtypespecific assay has been diminished because of the emergence of new subtypes and new reassortants in U.S. pig populations and because of an apparently increased rate of antigenic drift within the subtypes [34, 35]. Consequently, the diagnostic value of serologic test results has been diminished and the need for better serodiagnostic assays has become apparent [17]. In this study, we characterized the protein specificity of the antibody response of pigs to infection and vaccination over time and attempted to identify viral antigen candidates for use in serologic assays that are not restricted by virus subtype and that can differentiate vaccinated animals from those exposed to natural infection. Western immunoblot analysis of sera from experimentally infected pigs demonstrated that in addition to antibody against H protein, pigs developed a consistent antibody response to NP protein during 4 weeks after inoculation regardless of the subtype of the challenge virus (Table 1). Because NP protein is antigenically conserved among influenza A viruses, these results suggest that NP protein may be an excellent candidate for use as an antigen in a non-subtype-specific serologic assay. Nucleoprotein also may be an antigen that can be used in a serologic assay that overcomes the host species specificity of such assays. Several previous studies have reported that antibody response against NP protein is consistent across host species regardless of subtype [2, 27, 38]. The lack or delay of antibody response to M1 protein in exposed or vaccinated pigs was an unexpected observation since M1 protein is known to be abundant in the influenza virion and also antigenically conserved. Antisera against M1 protein are commonly used for serotyping of influenza viruses [8]. Results of the present study suggest that an assay using M1 protein as antigen may not be as sensitive in detecting recent infections as an assay using NP protein. Furthermore, seroconversion to vaccination might not be detected by such an assay (Table 2). Several previous studies also have reported a lower level of antibody production against M1 protein compared to NP protein [3, 16]. Antibody response stimulated to M1 protein may vary between strains. In this study, the H3N2 reference hyperimmune serum appeared to have a lower antibody titer to M1 protein than did the H1N1 reference hyperimmune serum based on the intensity of the band, even though the same antigen was used for immunoblot (Fig. 1). Some previous studies have also shown that the level of antibody response to M1 protein depends on the degree of viral replication in infection and the type of adjuvant in vaccination [15]. A

12 W.-I. Kim et al. previous study with human influenza virus reported that patients who showed severe clinical signs tended to develop a high-level antibody response to M1 protein while patients with no or mild symptoms did not respond [5]. M1 protein is believed to be associated more with stimulation of cell-mediated immunity than humoral immunity [36]. These observations suggest that M1 protein would be a less desirable antigen for a universal serologic assay for SIV infection. Antibodies to NS1 and NS2 proteins were detected only in sera from experimentally infected pigs and not in vaccinated animals. Similar observations have been made previously with equine influenza virus. Antibody response to recombinant viral NS1 protein was detected in experimentally infected horses but not in vaccinated horses by either immunoblot analysis [1] or by ELISA [22]. Nonstructural proteins of viruses are expressed only during viral replication and thus have been proposed as potential antigens for use in serologic assays that could differentiate animals inoculated with a killed vaccine from animals exposed to infection [12, 20]. NS1 protein also is known to be genetically and antigenically conserved among type A influenza viruses that infect mammals although some variants have been reported among avian influenza viruses [18, 19]. Such antigenic conservation is believed to be due to the protein s critical role in infection and pathogenesis [23, 32]. Although the duration of antibody against NS1 in infected pigs could not be determined due to the short length of the study, titers as determined by band intensity continued to increase to day 28 PI (Fig. 2E and F). Early and sustained immune response of infected pigs to NS1 protein favors this protein as a differentiation marker between naturally exposed and vaccinated pigs. In summary, results of this study suggest that NP protein may be the antigen of choice for a subtype-unrestricted universal serodiagnostic assay and NS1 protein may be used to differentiate animals vaccinated with current killed vaccines from infected animals. Further studies are in progress to develop an ELISA using these antigens and to evaluate performance of such an assay in the field. Antibody responses to conformational antigenic determinants remain to be further characterized as Western immunoblot techniques visualize only antibody recognition for linear antigenic determinants. Acknowledgement The authors thank Dr. Eileen Thacker at Iowa State University and Dr. Juergen Richt at the National Animal Disease Center, Ames, Iowa for kindly providing valuable serum samples for the study. References 1. Birch-Machin I, Rowan A, Pick J, Mumford J, Binns M (1997) Expression of the nonstructural protein NS1 of equine influenza A virus: detection of anti-ns1 antibody in post infection equine sera. J Virol Methods 65: De Boer GF, Back W, Osterhaus AD (1990) An ELISA for detection of antibodies against influenza A nucleoprotein in humans and various animal species. Arch Virol 115: Bucher DJ, Mikhail A, Popple S, Graves P, Meiklejohn G, Hodes DS, Johansson K, Halonen PE (1991) Rapid detection of type A influenza viruses with monoclonal

13 Western immunoblot analysis of antibody responses to SIV antibodies to the M protein (M1) by enzyme-linked immunosorbent assay and timeresolved fluoroimmunoassay. J Clin Microbiol 29: Centers for Disease Control (1982) Concepts and procedures for laboratory-based influenza surveillance. Centers for Disease Control, U.S. Department of Health and Human Services, Washington, DC 5. Cretescu L, Beare AS, Schild GC (1978) Formation of antibody to matrix protein in experimental human influenza A virus infections. Infect Immun 22: Dowdle WR, Schild GC (1975) Laboratory propagation of human influenza viruses, experimental host range, and isolation from clinical material. In: Kilbourne ED (ed) The influenza viruses and influenza. Academic Press, New York, pp Epand RM, Epand RF (2002) Thermal denaturation of influenza virus and its relationship to membrane fusion. Biochem J 365: Esterday BC, Van Reeth K (1999) Swine Influenza. In: Straw BE, D Allaire S, Mengeling WL, Taylor DJ (eds) Diseases of swine. Iowa State University Press, Ames, IA, pp Hay AJ (1992) The action of adamantanamines against influenza A viruses: inhibition of the M2 ion channel protein. Semin Virol 3: Hilbrands H, Kitikoon P, Erickson B, Thacker E (2004) SIV vaccine induced neutralizing and hemagglutination inhibition antibodies. Proceedings of the 35 th Annual Meeting of American Association of Swine Veterinarians, Des Moines, IA, pp Hilleman MR (2002) Realities and enigmas of human viral influenza: pathogenesis, epidemiology and control. Vaccine 20: InoueY, Suzuki R, MatsuuraY, Harada S, Chiba J, WatanabeY, Saito I, Miyamura T (1992) Expression of the amino-terminal half of the NS1 region of the hepatitis C virus genome and detection of an antibody to the expressed protein in patients with liver diseases. J Gen Virol 73: Jackson TA, Chandler-Conrey N, Prouty K (2004) Serologic responses to three commercial bivalent swine influenza virus vaccines in juvenile pigs. Proc, 35 th Annual Meeting of American Association of Swine Veterinarians, Des Moines, IA, pp Janke BH, Swalla RA, Yoon K-J (2001) Reciprocal cross-protection against heterologous challenge between H3N2 swine influenza viruses with minimal serologic cross-reactivity in the hemagglutination inhibition test. Proc, 44 th Annual Meeting of American Association of Veterinary Laboratory Diagnosticians, Hershey, PA, pp Kapaklis-Deliyannis GP, Drummer HE, Brown LE, Tannock GA, Jackson DC (1993) A study of the advantages and limitations of immunoblotting procedures for the detection of antibodies against influenza virus. Electrophoresis 14: Khristova ML, Egorenkova EM, Busse TL, Leonov SV, Demidova SA, Kharintonenkov IG (1988) Simultaneous determination of the level of antibodies to influenza virus surface and internal proteins by enzyme-linked immunosorbent assay. Acta Virol 32: Long BC, Goldberg TL, Swenson SL, Erickson G, Scherba G (2004) Adaptation and limitations of established hemagglutination inhibition assays for the detection of porcine anti-swine influenza virus H1N2 antibodies. J Vet Diagn Invest 16: Nakajima K, Nobusawa E, Ogawa T, Nakajima S (1990) Evolution of the NS genes of the influenza A viruses. I. The genetic relatedness of the NS genes of animal influenza viruses. Virus Genes 4: Nakajima K, Nobusawa E, Nakajima S (1990) Evolution of the NS genes of the influenza A viruses. II. Characteristics of the amino acid changes in the NS1 proteins of the influenza A viruses. Virus Genes 4: 15 26

14 W.-I. Kim et al.: Western immunoblot analysis of antibody responses to SIV 20. Neitzert E, Beck E, De Mello PA, Gomes I, Bergmann IE (1991) Expression of the aphthovirus RNA polymerase gene in Escherichia coli and its use together with other bioengineered nonstructural antigens in detection of late persistent infections. Virol 184: O Neill R, Talon EJ, Palese P (1998) The influenza virus NEP (NS2 protein) mediates the nuclear export of viral ribonucleoproteins. EMBO J 17: Ozaki H, Sugiura T, Sugita S, Imagawa H, Kida H (2001) Detection of antibodies to the nonstructural protein (NS1) of influenza A virus allows distinction between vaccinated and infected horses. Vet Microbiol 82: Palese P, Muster T, Zheng H, O Neill R, Garcia-Sastre A (1999) Learning from our foes: a novel vaccine concept for influenza virus. Arch Virol Suppl 15: Renshaw HW (1975) Influence of antibody-mediated immune suppression on clinical, viral, and immune responses to swine infection. Am J Vet Res 36: Richardson JC, Akkina RK (1991) NS2 protein of influenza virus is found in purified virus and phosphorylated in infected cells. Arch Virol 116: Schild GC, Oxford JS, De Jong JC, Webster RG (1983) Evidence for host-cell selection of influenza virus antigenic variants. Nature 303: Shafer AL, Katz JB, Eernisse KA (1998) Development and validation of a competitive enzyme-linked immunosorbent assay for detection of type A influenza antibodies in avian sera. Avian Dis 42: Shope RE (1931) Swine influenza III. Filtration experiments and etiology. J Exp Med 54: Skehel JJ (1972) Polypeptide synthesis in influenza virus-infected cells. Virol 49: Tobita K, Sugiura A, Enomote C, Furuyama M (1975) Plaque assay and primary isolation of influenzaa viruses in an established line of canine kidney cells (MDCK) in the presence of trypsin. Med Microbiol Immunol 162: Vincent LL, Janke BH (1997) A pathogenesis comparison study of two isolates of swine influenza virus. Proceedings of the 40 th Annual Meeting of American Association of Veterinary Laboratory Diagnosticians, Louisville, KY, pp Wang X, Li M, Zheng H, Muster T, Palese P, Beg AA, Garcia-Sastre A (2000) Influenza A virus NS1 protein prevents activation of NF-kappaB and induction of alpha/beta interferon. J Virol 74: Webby RJ, Swenson SL, Krauss SL, Gerrish PJ, Goyal SM, Webster RG (2000) Evolution of swine H3N2 influenza viruses in the United States. J Virol 74: Webby RJ, Rossow K, Erickson G, Sims Y, Webster RG (2004) Multiple lineages of antigenically and genetically diverse influenza A virus co-circulate in the United States swine population. Virus Res 103: Webster RG, Bean WJ, Gorman OT, Chambers TM, Kawaoka Y (1992) Evolution and ecology of influenza A viruses. Microbiol Rev 56: Webster RG, Hinshaw VS (1977) Matrix protein from influenza A virus and its role in cross-protection in mice. Infect Immunol 17: Yoon K-J, Janke BH (2002) Swine influenza virus: evolution, epidemiology and diagnosis. In: MorillaA,Yoon K-J, Zimmerman JJ (eds) Trends in emerging viral infections of swine. Iowa State University Press, Ames, IA, pp Zhou EM, Chan M, Heckert RA, Riva J, Cantin MF (1998) Evaluation of a competitive ELISA for detection of antibodies against avian influenza virus nucleoprotein. Avian Dis 42: Author s address: Kyoung-JinYoon, DVM, PhD, DACVM, Associate Professor and Head of Virology, Department of Veterinary Diagnostic and Production Animal Medicine, College of Veterinary Medicine, Iowa State University, 1600 South 16 th Street,Ames, IA 50011, USA; kyoon@iastate.edu