Minor Changes in the Hemagglutinin of Influenza A(H1N1)2009 Virus Alter Its Antigenic Properties

Background The influenza A(H1N1)2009 virus has been the dominant type of influenza A virus in Finland during the 2009–2010 and 2010–2011 epidemic seasons. We analyzed the antigenic characteristics of several influenza A(H1N1)2009 viruses isolated during the two influenza seasons by analyzing the amino acid sequences of the hemagglutinin (HA), modeling the amino acid changes in the HA structure and measuring antibody responses induced by natural infection or influenza vaccination. Methods/Results Based on the HA sequences of influenza A(H1N1)2009 viruses we selected 13 different strains for antigenic characterization. The analysis included the vaccine virus, A/California/07/2009 and multiple California-like isolates from 2009–2010 and 2010–2011 epidemic seasons. These viruses had two to five amino acid changes in their HA1 molecule. The mutation(s) were located in antigenic sites Sa, Ca1, Ca2 and Cb region. Analysis of the antibody levels by hemagglutination inhibition test (HI) indicated that vaccinated individuals and people who had experienced a natural influenza A(H1N1)2009 virus infection showed good immune responses against the vaccine virus and most of the wild-type viruses. However, one to two amino acid changes in the antigenic site Sa dramatically affected the ability of antibodies to recognize these viruses. In contrast, the tested viruses were indistinguishable in regard to antibody recognition by the sera from elderly individuals who had been exposed to the Spanish influenza or its descendant viruses during the early 20th century. Conclusions According to our results, one to two amino acid changes (N125D and/or N156K) in the major antigenic sites of the hemagglutinin of influenza A(H1N1)2009 virus may lead to significant reduction in the ability of patient and vaccine sera to recognize A(H1N1)2009 viruses.


Introduction
During the recent two years, the pandemic influenza A virus of swine origin, influenza A(H1N1)2009 virus, has been the predominant circulating influenza virus in most parts of the world. The virus has infected millions of people and the infection has lead to the death of at least 18 400 individuals. In Finland the first cases of the influenza A(H1N1)2009 were identified in May 2009. During September the first local outbreaks occurred in garrisons and schools, after which the virus spread rapidly in the general population. The peak pandemic activity was observed during weeks 43-49 and by the end of the year the epidemic was over in Finland [1,2]. In serosurveys elderly individuals were found to have preexisting, cross-reactive antibodies against the novel 2009 pandemic virus that were likely originating from previous infections with antigenically related viruses such as the 1918 influenza virus and its immediate descendants that were circulating during the early decades of the 20 th century [3] [4][5][6][7][8]. Except for the elderly, large segments of the human population throughout the world lacked protective immunity against the novel influenza A(H1N1)2009 virus and were thus susceptible to the virus infection. Until now, likely due to limited immunological pressure in the general population, the virus has not yet undergone significant genetic or antigenic changes.
Through the hemagglutinin (HA) the influenza virus binds to sialic-acid receptors on the host cell surface, after which the virus is internalized and the viral genome enters the nucleus in order to initiate viral RNA synthesis. Since the HA is situated on the surface of the viral particles it is also a target for immune response, especially antibodies. The major antigenic epitopes in the HA molecule mutate frequently enabling the virus to escape immune responses. Recent reports on the evolution of influenza A(H1N1)2009 describe mutations S183P (amino acid numbering throughout the text starts from the mature HA0 without signal peptide) and I191L in HA that enhance viral replication in cell culture and in embryonated hens' eggs [9]. Other mutations, such as D127E, S183P and D222G have been shown to be associated with a more virulent phenotype in humans or mice [2,[10][11][12]. The D127E and S183P mutations have lead to antigenic changes and impaired recognition by ferret antisera raised against the A/California/07/2009 virus [11]. Recently, a new frequently observed mutation, E374K has been found [13]. This mutation locates in the HA oligomerization interface and is also part of a known antigenic site. This mutation is not unambiguously associated with severe disease, but interestingly it has been detected in pandemic vaccine breakthrough infections. These viruses typically had N125D substitution in their HA molecule [14]. During the year 2010 viruses with double mutations N125D and E374K have been found with increased frequency in the southern hemisphere [14]. These viruses have been associated with several vaccine breakthrough infections and were identified in a number of fatal cases [14]. The N125D mutation is located in the Sa epitope of A(H1N1)2009 HA [2] and in an avian influenza strain (A/ Mallard/Pennsylvania/10218/84(H5N2)) the corresponding amino acid was reported to cause antigenic drift as an immune escape mutant [15]. However, antigenic analysis with ferret anti-sera showed no remarkable differences in antibody titers against the N125D virus as compared to the A/California/07/2009 vaccine virus [14].
Genetic characterization of the influenza A(H1N1)2009 viruses that circulated in Finland and elsewhere during 2009 show that these viruses were closely related to A/California/07/2009 vaccine virus [2,6,16]. At present, genetic analyses of the HA or neuraminidase (NA) genes do not reveal any changes that may have led to the selection of a virus with increased epidemic potential or exceptionally high virulence. However, the antigenic characteristics of these viruses have not been systematically analyzed in humans. Here we report that several amino acid changes, namely N125D, N156K, S162R, and E374K, are associated with reduced recognition by sera collected from patients having experienced natural A(H1N1)2009 infection or having received vaccination as analyzed by the hemagglutination inhibition (HI) test. Especially, double mutations N125D and N156K (2009 virus isolate) or N125D and E374K (2010 virus isolates) were associated with declined antibody recognition in vaccinated individuals.   Figure S1). Compared to the vaccine strain, the Finnish A(H1N1)2009 viruses have two to five amino acid changes in their HA1 region, which are distributed to all the major antigenic epitopes. One conserved change in all Finnish A(H1N1)2009 virus strains, except A/Finland/554/2009, was S203T which is located in the antigenic epitope Ca1 of HA, close to the receptor binding pocket (Figure 2). This mutation has also been detected globally in majority of influenza A(H1N1)2009 viruses [18]. In addition to this common change, most of the Finnish A(H1N1)2009 viruses showed one or two additional amino acid changes in other antigenic epitopes. Of potential interest in respect to antibody recognition is the group I/2009-2010 virus, A/Finland/694/2009, which has N125D and N156K amino acid substitutions on the distal end of the HA molecule locating to the antigenic site Sa (Figure 2 and Figure S2). In addition, novel double mutations (N125D and E374K) were also found in the group II/2010-2011 viruses, namely in A/Finland/19/2010 and A/Finland/22/2010 viruses. The amino acid change E374K has been identified in some A(H1N1)2009 viruses isolated in Finland ( Figure 1) and elsewhere [2,13,14]. Additional mutations at potential antigenic sites, in relation to the California vaccine virus, were also identified at positions 73, 74, 137, 162, 185, and 222 ( Figure 2 and Figure S2).

Finnish influenza A(H1N1)2009 virus strains show considerable antigenic heterogeneity
Sequence analysis of the HA gene revealed several amino acid changes that have occurred in the viruses that circulated in Finland during the 2009-2010 and 2010-2011 epidemic seasons. By analyzing the nucleotide sequence and the molecular structure of HA it is, however, difficult to estimate whether these changes have any importance in antibody recognition. To analyze the impact of amino acid changes on the antigenic characteristics of circulating epidemic viruses we selected two representative viruses from each of the four epidemic groups of the 2009-2010 viruses and one to two viruses from the three epidemic groups of the 2010-2011 viruses ( Figure 1). We determined antibody levels against the thirteen Finnish influenza strains and the vaccine A/ California/07/2009 virus by HI test in a set of 120 serum samples ( Table 1). The 2010 (epidemic season 2010-2011) viruses were tested in separate sets of analyses with the same panel of sera and the data is therefore tabled separately (  Table 1). Some of the lowest antibody titers were also observed against A/Finland/698/2009, which had amino acid change in the antigenic epitope Cb (Figure 2 and Figure S2).

Antibody titers for cell culture-grown viruses correlate with egg-propagated virus in HI test
All Finnish strains were isolated and propagated in Madin-Darby canine kidney (MDCK) cells. For comparison we tested both MDCK-and egg-grown vaccine strain, A/California/07/2009, in the HI test. The geometric mean titers to batches of MDCKand egg-grown A/California/07/2009 were comparable although analyses with egg-grown virus stocks gave somewhat higher antibody levels ( Table 1). The correlation coefficient for antibody titers obtained by HI test varied between 0.55 and 0.96, which indicated a strong positive correlation between the vaccine virus and Finnish epidemic strains ( Figure 3A and B). Correlation coefficients for all tested viruses were significant, p,0.001, which suggested that MDCK-and egg-grown viruses were comparable in HI test. However, the scatter blots in Figure 3 C showed that the correlation between different viruses, especially Finland/694/2009, Finland/6/2010, Finland/19/2010, and Finland/22/2010, is better for sera with high antibody titers (.40) as compared to that observed with sera having lower titers, where the spots (HI values) seemed to be more dispersed.

The antibodies induced by natural infection with influenza A(H1N1)2009 show differential recognition of viruses with mutations in antigenic epitopes
The antibody titers against different influenza A(H1N1)2009 viruses differ notably. In order to analyze more carefully how the sera from the different donor groups recognize influenza

The elderly have broad-spectrum low-level cross-reactive antibodies against pandemic influenza viruses
We have previously reported that older Finns had cross-reactive antibodies against the pandemic influenza A(H1N1)2009 virus providing potential protection against the infection [4]. We tested a set of serum samples collected in 2005 from individuals whose age was at that time 80 years or over. Among these elderly individuals the antibody titers against the current pandemic viruses were generally low (range of . In addition, contrary to patient sera where the antibodies failed to recognize some of the tested viruses, in the elderly none of the tested viruses, except the vaccine strains, were distinguishable in HI test ( Figure 4B). This may indicate that infections of the elderly by the Spanish influenza and/or its immediate descendant viruses were able to induce quite broad-spectrum cross-reactive humoral immunity against the pandemic influenza A(H1N1)2009 virus. The significance of the differences was not calculated since none of the studied viruses can be set as a baseline. On the other hand, as compared to the A/ California/07/2009 vaccine virus the antibody titers against all the Finnish strains were higher.   these viruses and whether the viruses with mutations in antigenic sites can escape from vaccine-induced immunity. Analysis of serum specimens from twenty vaccinated healthy adults showed that 21 days after vaccination anti-influenza antibodies showed a five to fifteen-fold increase in geometric mean titers against all viruses, except that of A/Finland/694/2009 with only less than a three-fold increase ( Figure 5A). One year after the vaccination, the mean antibody levels against most studied viruses had clearly decreased but they were still three-to six-folds higher as compared to prevaccination antibody levels.   Figure 6A) indicates that the vaccine induces very high antibody levels but for most of the viruses the mean antibody levels reach their maximal or near-maximal levels already after two immunizations. Additional immunizations did not enhance the cross-reactivity of the antibodies against different A(H1N1)2009 strains, since the relative antibody ratios in comparison to vaccine virus were not increased during repeated immunizations ( Figure 6A and data not shown). The lowest cross-reactive antibody levels in Serum specimens from guinea pigs were collected only at day 0 before the first immunization and one week after the last vaccination. Mean antibody levels increased to very high levels after four doses of vaccine and antibody titers against most viruses were at a comparable level. The only exception was again the A/Finland/694/2009 virus, against which the antibodies remained at much lower level as compared to other viruses ( Figure 6B). This suggests that even multiple doses of the vaccine failed to induce highly cross-reactive antibodies against this virus which carries two mutations in antigenic site Sa (Figure 2 and Figure S2).   vaccine in rabbits and guinea pigs. Rabbits (A) and guinea pigs (B) were immunized four times at three weeks intervals with pandemic influenza A(H1N1) vaccine and the serum samples were collected on day 0 (rabbits and guinea pigs), before each immunization (rabbits), and one week after the last immunization (rabbits and guinea pigs). The immunizations (black) and sample collections (grey) are marked with arrows. The serum samples were analyzed by HI test and geometric mean titers for each virus were calculated. The means represent the geometric means of three rabbit and five guinea pig anti-sera. doi:10.1371/journal.pone.0025848.g006 vaccine candidate for the 2011-2012 vaccine to be used in the northern hemisphere. This virus is still assumed to give reasonable protection against infection with globally circulating viruses. However, even few amino acid changes in antigenically important sites of the HA, as demonstrated in the present study, may render the vaccine virus a poor match to the circulating virus strains.

Discussion
The HA protein on the surface of influenza virions is one of the major target molecules for host antibody responses. Therefore, amino acid changes occurring either in the antigenic sites or on the surface of the HA molecule may have an important effect on antibody recognition. Five classical antigenic sites (Sa, Sb, Ca1, Ca2, and Cb) have been described in the HA of influenza A/ H1N1 virus [20,21]. These antigenic sites are located at the globular head of the HA (Figures 1, 2, and S2) although recent studies have described antigenically important determinants also in the stem region of HA [22][23][24]. At present, there is no experimental proof that the antigenic sites found in the HA molecule of classical seasonal H1N1 HA molecules are equally important for the immunity against recent pandemic A(H1N1)2009 viruses. However, our results showed that only one amino acid change in the antigenic site Sa (N125D) compromised antibody recognition, which could be seen as reduced antibody titers in the HI test. This amino acid change is characteristic of group II/2010-11 viruses, which represents the majority of Finnish viruses (57%) from the epidemic season 2010-11. If there was another amino acid change (N156K) in the Sa epitope, as was the case in the A/Finland/694/2009 virus (Figure 2 and Figure (Figure 2 and S2) showed impaired recognition by patient convalescent sera. This suggests that relatively strong immune response induced by natural infection might be evaded by only a few mutations in critical antigenic regions. The same effect was also seen among vaccinated individuals, since antibody responses against the above described viruses (694, 618 and 698) were significantly reduced as compared to the homologous vaccine virus strain ( Figure 5A). In addition, post-vaccination anti sera showed reduced recognition of epidemic season 2010-2011 viruses A/ Finland/19/2010 and A/Finland/22/2010, which showed point mutations in antigenic sites Sa (N125D) and Ca1 (S203T). This further suggests that the specificity of humoral immune responses induced by natural infection and vaccination are alike. In addition, the mean HI titers seen among the vaccinees and those individuals who had undergone a natural infection are comparable ( Figure 4A vs. Figure 5) thus indicating that the pandemic influenza A(H1N1)2009 vaccine (Pandemrix) is capable of inducing very good humoral immunity. It is also noteworthy that portion of antibodies that neutralize or in other ways inhibit virus replication may not be detected in HI test. The presence of such antibodies and the possible alterations in antibody recognition of the viruses under discussion would be interesting to analyze in further studies.
An interesting observation in the present study was that three out of four 2010-2011 season viruses had an E374K mutation in the HA2 region (A/Finland/19/2010, A/Finland/20/10, and A/ Finland/22/10). Viruses with the same mutation predominated in Oceania during local winter 2010, e.g. the southern hemisphere influenza season 2010 [14]. Barr and coworkers also reported that viruses carrying the N125D and E374K mutations were found in vaccine breakthrough cases in teenagers and adults, who had been vaccinated with the monovalent pandemic influenza A(H1N1)2009 vaccine and the same mutant viruses were also found in a number of fatal cases. However, the two Finnish strains (A/Finland/19/2010 and A/Finland/22/2010) with this dual mutation, representing the epidemic season 2010-2011 group II/10 viruses, were isolated from unvaccinated individuals.
An important question is whether antibody levels and the broadness of cross-reactivity against different virus strains can be stimulated by repeated immunizations with pandemic influenza A(H1N1)2009 vaccine. Since the Finnish population received only one dose of the pandemic vaccine before or during the pandemic we used experimental animals to study the effect of repeated immunizations on antibody responses and cross-reactivity. In rabbits, it was found that already two doses of the pandemic influenza A(H1N1)2009 vaccine were sufficient to induce maximal or near-maximal antibody responses against different epidemic virus strains. Interestingly, repeated immunizations were not able to expand the cross-reactivity against different viruses since the antibody ratios against different virus isolates vs. the vaccine virus strain remained rather constant ( Figure 6A). This may indicate that the initial specificity of the antibody response is determined at early times of immunization and repeated immunizations with the same vaccine, such as the Pandemrix vaccine, does not lead to broadening of the specificity of the antibody response. These findings may indicate that also in humans, in order to induce better antibody cross-reactivity, the composition of the vaccine has to be changed rather than giving repeated immunizations with the same vaccine.
The present study demonstrates that circulating influenza A(H1N1)2009 viruses are genetically highly diverse and at present a specific globally circulating virus strain has not been observed. Genetic characterization of the Finnish circulating viruses, which originated from different parts of the world, indicates that the evolutionary speed of the virus is quite typical for influenza A viruses. However, there appears to be a tendency that many viruses have mutations in antigenically important epitopes of the HA molecule and these mutations may have quite dramatic effects on antibody recognition both among vaccinated individuals and those that had experienced a natural infection. Thus, continuous genetic and antigenic surveillance of circulating influenza viruses as well as rapid and open sharing of information is of utmost importance for WHO to make the best possible recommendation for the vaccine viruses. In addition, it is important to rapidly identify virus strains with elevated virulence or with the ability to escape vaccine-or natural infection-induced immunity.

Viruses
All viruses for serological analysis were selected based on their HA gene nucleotide sequence. We have previously reported that based on the HA gene, influenza A(H1N1)2009 viruses circulating in Finland during 2009-2010 season clustered in four genetic groups [2]. Two representatives of each of these four groups were selected for immunological analyses:

Serum specimens
In order to get a broader view of the antigenic characteristics of the selected viruses, we chose three different sets of serum specimens for serological analysis. The first set of sera was from 20 healthy adults who received pandemic influenza A(H1N1) vaccine Pandemrix in November 2009. Pandemrix is a monovalent split-virion vaccine generated from the A/California/07/2009-like strain (NYMC X-179A; New York Medical College) by GlaxoSmith-Kline (GSK) Biologicals [25]. A single dose of this vaccine contains 3.75 mg of HA and AS03 adjuvant. Serum samples from 20 volunteers (median age 42; range 27 to 62 years) were collected prior to vaccination on day 0, day 21, and one year after the vaccination. The second set of sera was collected from 50 unvaccinated military conscripts (median age 21; range [20][21][22][23][24][25][26][27][28] in the Dragsvik garrison (Raasepori), which is located in Southern Finland. An outbreak of influenza A(H1N1)2009 virus occurred in this garrison during September 2009 and 50% of the conscripts were infected (confirmed by PCR) with influenza A(H1N1)2009 [1]. For the present study we chose 40 A(H1N1)2009 virus seropositive and 10 seronegative randomly selected serum specimens. The third set of serum samples was obtained from the virus diagnostic unit at the Central Laboratory Services of the Helsinki University Hospital (HUSLAB). These sera had been collected prior to influenza A(H1N1)2009 pandemic in 2004-2005 and the specimens have been randomly selected from all sera sent to HUSLAB from different parts of the country. The samples were anonymous, only the age of the donor and sample collection date was known. For the present study, 50 serum samples from elderly individuals representing the age group of 80 years or older were chosen, since our previous observations indicated that these individuals have cross-reactive antibodies against influenza A(H1N1)2009 virus [4].
The collection of serum samples for the present study was approved by the Ethical Committee of the Helsinki-Uusimaa Health District (Permissions 382/E5/07 148/2008 and 1289/2010 and 199/13/03/00/2009 1164). All individuals participating in the studies (vaccinated and conscripts) gave written informed consent before enrolment in the study.

Sequence and phylogenetic analysis of HA gene
The nucleotide sequences of the influenza A(H1N1)2009 viruses for this study (35 Finnish strains from the epidemic season 2010-2011) were sequenced from original clinical samples. RNA extraction, cDNA synthesis, DNA amplification and sequencing were performed as previously described [2]. The nucleotide sequences of the complete HA genes were assembled using Sequencer version 4.7 (Gene Codes Corporation, Ann Arbor, USA). Mega software version 4 [26] (Molecular Evolutionary Genetics Analysis) was used in nucleotide and amino acid sequence comparison and for the construction of the phylogenetic tree. The Neighbor-joining method [27] with the maximum composite likelihood model [28] was used to generate the phylogenetic tree. Bootstrapping was performed with 1000 replicates [29]. In the present study all HA amino acid residues are numbered without the signal peptide sequence. GenBank

Immunizations of rabbits and guinea pigs
In order to study the immunogenicity and the ability of the pandemic influenza A(H1N1)2009 vaccine to induce cross-reactive antibody responses in animals we immunized five guinea pigs and three rabbits four times in three weeks intervals. Each animal received one human dose (3,75 mg HA in AS03 adjuvant) of vaccine with each immunization. Serum samples from guinea pigs were collected before the first immunization (day 0) and one week after the last immunization (day 70). From the rabbits serum samples were collected on day 0, before each immunization on days 21, 42, and 63, and one week after the last immunization on day 70. Immunizations of the animals and the collection of serum samples were approved by the Ethical Committee of National Institute for Health and Welfare (permission KTL 2008-02). Serum samples were stored at 220uC and analyzed for influenza A virus specific antibodies by the HI test.

Serologic assays
All serum specimens were assayed by the HI test against nine influenza A(H1N1)2009 strains isolated in Finland during the 2009-2010 epidemic season, four Finnish influenza A(H1N1)2009 strains from season 2010-2011, and A/California/07/2009 vaccine strain. The HI tests were performed in accordance with WHO guidelines with established procedures [31,32] using turkey erythrocytes. In brief, serum samples were pretreated with receptor destroying enzyme (RDE) from Vibrio cholerae filtrate (Denka Seiken, Tokyo, Japan) at +37uC for 18 h and then with 100% turkey erythrocytes at +4uC for one hour to remove nonspecific inhibitors and agglutinins. Serial 1:2 serum dilutions starting from 1:10 initial dilution were made, live viral antigen (4 HA units) was added and incubated for one hour at ambient temperature. Turkey red blood cells (0.5%) were added to antigenserum solutions and incubated at room temperature for 30 minutes to determine the HI titers. Serum specimens with HI titers ,10 were assigned a titer value of 5 for statistical analyses.

Statistical analysis
Antibody titers obtained by the HI test were analyzed statistically by calculating geometric mean titers with 95% confidence intervals for each virus. Significance of the differences between mean titers for each virus compared to the vaccine strain was calculated using Student's t-test (paired, two-tailed) and the significance level was adjusted to p,0.01. The correlation coefficient and the significance of the correlation coefficient of the antibody titers for each virus were calculated against the A/ California/07/2009 egg-grown virus.