Skip to main content
PMC home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2009 May 13;106(21):8701–8706. doi: 10.1073/pnas.0903427106

Differential neutralization efficiency of hemagglutinin epitopes, antibody interference, and the design of influenza vaccines

Wilfred Ndifon a,1, Ned S Wingreen b,1, Simon A Levin a,1
PMCID: PMC2688967  PMID: 19439657

Abstract

It is generally assumed that amino acid mutations in the surface protein, hemagglutinin (HA), of influenza viruses allow these viruses to circumvent neutralization by antibodies induced during infection. However, empirical data on circulating influenza viruses show that certain amino acid changes to HA actually increase the efficiency of neutralization of the mutated virus by antibodies raised against the parent virus. Here, we suggest that this surprising increase in neutralization efficiency after HA mutation could reflect steric interference between antibodies. Specifically, if there is a steric competition for binding to HA by antibodies with different neutralization efficiencies, then a mutation that reduces the binding of antibodies with low neutralization efficiencies could increase overall viral neutralization. We use a mathematical model of virus–antibody interaction to elucidate the conditions under which amino acid mutations to HA could lead to an increase in viral neutralization. Using insights gained from the model, together with genetic and structural data, we predict that amino acid mutations to epitopes C and E of the HA of influenza A/H3N2 viruses could lead on average to an increase in the neutralization of the mutated viruses. We present data supporting this prediction and discuss the implications for the design of more effective vaccines against influenza viruses and other pathogens.

Keywords: antigenic distance, epidemic, epitope vaccine, evolution

Influenza viruses infect ≈5–15% of the world population each year (1). Infection leads to the production of antibodies that preferentially recognize the influenza viral hemagglutinin (HA) protein (2, 3). Most of these antibodies neutralize influenza viruses and, hence, limit infection by binding to specific regions of HA called (functional) epitopes, which are located within presumed topologically distinct sites called antigenic sites (denoted simply by epitopes) (Fig. 1A). Amino acid changes to HA have complex effects on viral neutralization by antibodies (4–6). For example, the antigenic similarity (a measure of the degree to which antibodies raised against one virus neutralize another virus) between certain pairs of influenza viruses actually increases after the introduction of additional amino acid differences between the HAs of the two viruses [see, e.g., ref. 6]. This observation could be explained by positing that the additional amino acid changes compensate for preexisting amino acid differences between the viruses. However, because some of the changes in question occur in entirely distinct HA epitopes from the preexisting differences, it is possible that there is another mechanism at play. Here, we propose such a mechanism based on steric interference between antibodies (Fig. 1B), and we discuss the implications for improving the effectiveness of influenza vaccines.

Fig. 1.

Fig. 1.

Open in a new tab

HA and antibody interference. (A) Globular head of a monomer of HA (Protein Data Bank ID code 1hgf), showing five antibody-binding sites (or epitopes) and the receptor-binding site. The figure was drawn by using the PyMOL molecular graphics system. (B) Interference between antibodies that bind to two different HA epitopes. Illustrated are cross-sections of an IgG molecule and an HA trimer. The molecules were drawn approximately to scale. IgG is a Y-shaped molecule that can be separated into three fragments (two Fab fragments and one Fc fragment) of approximately the same size. A Fab fragment has approximate dimensions 80 × 50 × 40 Å (7). In comparison, an HA trimer has a length of ≈135 Å and a diameter of ≈55 Å (8), approximately equal to the width of the 40- × 50-Å distal surface of a Fab fragment, which contains the Fab-binding pocket.

The ability of an antibody to neutralize a virus depends on the strength of the virus–antibody bond (i.e., the affinity of the antibody for the virus) and on the neutralization efficiency of the viral epitope bound by the antibody (4, 9–11).* In the case of influenza virus, there is a simple physical explanation for such epitope dependence of viral neutralization. A large body of experimental work (11–15) suggests that occlusion of the receptor-binding site by antibodies bound to HA constitutes the dominant mechanism of influenza viral neutralization. Antibodies that bind to HA epitopes located at a distance from the receptor-binding site may therefore fail to occlude the site efficiently, thereby leading to a low degree of viral neutralization (4, 11, 12). Moreover, it has been shown that antibodies that bind to a given HA epitope can prevent further binding of antibodies to other epitopes of the same HA protein (Fig. 1B) and even to epitopes found on adjacent HA proteins (13, 14). [Note that some antibodies may neutralize influenza virus by inhibiting fusion of the viral envelope with the cellular membrane (11, 16).]

The above observations suggest that antibodies that bind to low-neutralization efficiency epitopes of HA might interfere with the binding of antibodies to high-neutralization efficiency epitopes, thereby impeding the neutralization of influenza viruses (Fig. 2). This raises the intriguing possibility that the influenza virus may have evolved to decrease the probability of its own neutralization by allowing itself to be bound preferentially by antibodies that recognize low-neutralization efficiency epitopes of HA. Indeed, steric interference by antibodies that bind low-neutralization efficiency epitopes could greatly reduce the extent of mutation required for a virus to evade neutralization by host antibodies. For example, if all epitopes had high neutralization efficiencies, then, for a virus to evade neutralization, viral mutations would have to reduce the affinities of antibodies for all epitopes to such a level that only a small fraction of epitopes would be bound by antibodies (Fig. 3A). By comparison, if some epitopes had low neutralization efficiencies, then, to evade neutralization, viral mutations would only need to decrease the affinities of antibodies for high-neutralization efficiency epitopes enough for the bound antibodies to consist predominantly of those for low-neutralization efficiency epitopes (Fig. 3B). A corollary to this observation is that a decrease in the affinities of antibodies for epitopes with low neutralization efficiencies could lead to an increase in viral neutralization (Fig. 2). This in turn suggests a possible approach to designing “low-interference” influenza vaccines that could greatly reduce the impact of antibody interference (Fig. 3 C–E). Rather different approaches to vaccine design are suggested by the idea of “deceptive imprinting” (e.g., Ref. 17), which posits that vaccine effectiveness could be improved by limiting the antibody-induction potential of immunodominant epitopes. However, these approaches have not yet proven fruitful (18).

Fig. 2.

Fig. 2.

Open in a new tab

Effects of antibody interference on viral neutralization. (A) Wild-type HA [denoted HA(a)] containing two epitopes, a high-neutralization efficiency epitope (denoted E1) located close to the receptor-binding site, and a low-neutralization efficiency epitope (denoted E2) located farther from the receptor-binding site. A variant of HA(a) [denoted HA(b)] contains mutations to E1, whereas another variant [denoted HA(c)] also contains mutations to E2. (B) Antibodies raised against HA(a) bind to E1 of HA(c) much more readily than they bind to E1 of HA(b) because the additional mutations to E2 remove antibody interference. Viruses carrying the more mutated HA(c) are therefore neutralized more efficiently than viruses carrying the less mutated HA(b) by wild-type antibodies. Lines emanating from the receptor-binding site indicate that the site is not completely occluded.

Fig. 3.

Fig. 3.

Open in a new tab

Deleterious effects of antibody interference on the host and proposed strategy for influenza vaccine design. (A) If viral HA contained only epitopes with high-neutralization efficiences, then only viruses with large epitopic changes could escape from antibodies. (B) Antibody interference from low-neutralization efficiency epitopes enables viruses with small epitopic changes also to escape from antibodies. (C) Proposed low-interference vaccine strain is genetically modified from viral target at low-neutralization efficiency epitopes of HA. Vaccine-induced antibodies only recognize high-neutralization efficiency epitopes of target. (D) Antibodies induced by low-interference vaccine strain have low affinity for low-neutralization efficiency epitopes of the target and therefore do not interfere with antibodies to high-neutralization efficiency epitopes, implying better neutralization. (E) Without antibody interference the target virus cannot escape from vaccine-induced antibodies via small epitopic changes. Lines emanating from the receptor-binding site indicate that the site is not completely occluded.

Results and Discussion

Simple Model of Interference Between Antibodies, and Implications for Viral Neutralization.

Here, we consider an in vitro assay of the neutralization of an influenza virus by antibody-containing antiserum. We assume that the neutralization reaction reaches the equilibrium state and is therefore amenable to analysis by means of standard mass-action models of antibody–antigen binding (e.g., ref. 19). Let Ai, i = 1, …, m, denote antibody that binds to the ith HA epitope of the virus under consideration. Using the induced-fit model of antibody–antigen binding (20, 21) and assuming that the binding of antibody to an HA epitope interferes with the binding of antibodies to other epitopes of the same HA protein (13, 14), it can be readily shown that the probability Pi that the ith epitope is bound at equilibrium is given by:

graphic file with name zpq02109-7987-m01.jpg

where [Ai] denotes the free concentration of antibodies that bind to the ith epitope and Ki is the equilibrium dissociation constant (or the reciprocal of the affinity) of those antibodies.

Let A0 denote the total concentration of antibodies found in the antiserum and fi the fraction of those antibodies specific to the ith epitope. Because the concentration of antibodies produced during influenza virus infection tends to be substantially greater than the estimated concentration of viral HA (SI Methods), we can accurately approximate the free concentration of antibodies as [Ai] = fiA0, such that Eq. 1 becomes:

graphic file with name zpq02109-7987-m02.jpg

Previous experimental results suggest that the fraction of neutralized influenza viruses is described well by a sigmoidal Hill function of the average number of bound antibodies (proportional to Pi for a monoclonal antibody) and that the efficiency of viral neutralization depends on the epitopes bound by the antibodies (SI Methods and Fig. S1). In the following, we show that antibody interference combined with such epitope-dependent neutralization efficiency could explain certain puzzling aspects of the antigenic effects of amino acid changes to the HA of influenza viruses.

The standard experimental method of quantifying the antigenic effects of amino acid changes to the HA of influenza viruses is the hemagglutination-inhibition (HI) assay (22). The HI assay measures the maximum amount h (also called the HI titer) by which antiserum raised against one virus (the “homologous” virus) can be diluted without losing its capacity to neutralize a second virus (the “heterologous” virus). Neutralization is indicated by the inability of the heterologous virus to agglutinate a standardized amount of red blood cells. The normalized HI titer is given by the ratio of the HI titer obtained with the heterologous virus to the HI titer obtained with the homologous virus. We define the antigenic similarity between two influenza viruses as the geometric mean of the normalized HI titer of each virus relative to antiserum raised against the other virus; we call this the Archetti–Horsfall measure (AHM) of antigenic similarity (6, 23, 24). Empirical data show that the heterologous HI titer is occasionally greater than the homologous HI titer by more than a factor of 2 (the dilution factor used in the HI assay) and that the AHM of the antigenic similarity between two influenza viruses sometimes increases after additional amino acid changes to the HA of one of those viruses (ref. 6 and Appendix A of ref. 24).

We propose a mechanistic explanation for the above counterintuitive observations: Consider a virus containing an HA [denoted HA(a)] against which antibodies have been raised. Let HA(a) contain two epitopes, one (denoted E1) having high neutralization efficiency and the other (denoted E2) having much lower neutralization efficiency. Also, let the antibodies that bind to E1 and E2 be denoted by A1 and A2, respectively, and let the affinities of these antibodies for their respective epitopes be given by 1/K1 and 1/K2. If [A2]/K2 is greater than [A1]/K1, then at equilibrium P2 > P1 (see Eq. 2), and, depending on the neutralization efficiency of E2, a substantial fraction of the virus could escape neutralization. Now, if a variant of HA(a) [denoted HA(b)] differs from HA(a) by an antigenically important mutation to E1, then, all else being equal, the affinity of A1 antibodies for E1 of HA(b) would be lower than the affinity of A1 antibodies for E1 of HA(a), and the fraction of antibodies bound to E1 of HA(b) at equilibrium would decrease, resulting in lower neutralization of virus carrying HA(b) compared with the parent virus (Fig. 2).

If in addition to the difference at E1, another variant of HA(a) [denoted HA(c)] also differs from HA(a) at E2, then, all else being equal, the affinity of A2 antibodies for E2 of HA(c) would decrease, meaning that the fraction of A1 antibodies bound at equilibrium would increase, resulting in greater neutralization of virus carrying HA(c) compared with virus carrying HA(b) (Fig. 2). Consequently, the AHM measure of the antigenic similarity between the parent virus [carrying HA(a)] and a virus carrying the more mutated HA(c) would be greater than between the parent virus and a virus carrying the less mutated HA(b). Following a similar line of reasoning, the antigenically important difference at E2 would naturally cause the HI titer of an HA(c) virus relative to antiserum raised against an HA(b) virus to be greater than the corresponding homologous HI titer, as seen in empirical data. Therefore, antibody interference could provide a mechanistic explanation for some of the puzzling empirical observations regarding antigenic similarity.

Practical Applications of the Theory of Antibody Interference.

Antigenic effects of amino acid changes to A/H3N2 viruses.

Here, we combine our theoretical model for antibody interference with genetic and structural data to make experimentally testable predictions for the antigenic effects of amino acid changes to the HA epitopes of influenza A/H3N2 viruses. We chose to consider A/H3N2 viruses because of the wealth of available genetic and antigenic data on the evolution of these viruses and because of their epidemiological importance. The HA of A/H3N2 viruses contains five epitopes, denoted A–E (the amino acid positions found in each of these epitopes are given in Table S1). Several lines of evidence suggest that epitopes A and B contain functional epitopes with high neutralization efficiencies. First, the physical locations of these epitopes on the HA protein are much closer to the receptor-binding site compared with the locations of the other three epitopes (Fig. 1A and Table 1). This fact, in light of the aforementioned occlusion mechanism of antibody-mediated neutralization (11, 15), suggests that functional epitopes found within epitopes A and B could indeed have high neutralization efficiencies.

Table 1.

Spatial distance between epitopes of HA and the receptor-binding site

Epitope Spatial distance, Å
A 10.7
B 13.0
C 44.5
D 17.7
E 25.5
O 35.8
Open in a new tab

Shown are the distances between the centroid of atomic positions of amino acids found in each epitope and the centroid of atomic positions of amino acids found in the receptor-binding site. The distances are based on chain A of the HA 3D structure (PDB ID: 1hgf) of A/Aichi/2/68 (A/H3N2). A, B, C, D, E, and O are defined in the legend of Fig. 4.

Moreover, the vast majority of HA codons predicted to be under positive selection occurs in epitopes A, B, and, to a lesser extent, D (25). One expects the selective pressure to acquire antigenically consequential amino acid changes to be greater for epitopes with high neutralization efficiencies compared with epitopes with lower neutralization efficiencies because changes to the high-neutralization efficiency epitopes have a greater probability of effecting viral escape from neutralization. In contrast to epitopes A, B, and D, epitopes C and E, which are the furthest of the five epitopes from the receptor-binding site (Table 1), accrue amino acid changes at a comparatively lower rate (25), suggesting that these epitopes could have lower neutralization efficiencies. (Note that the above-mentioned pattern of positive selection could still occur if all epitopes had the same neutralization efficiencies, provided epitopes A, B, and D were much more immunogenic than epitopes C and E.)

It follows from the above observations and from a consideration of the effects of antibody interference that amino acid changes to epitopes A, B, and D could lead on average to a decrease in neutralization and, hence, an increase in the antigenic distances between influenza A/H3N2 viruses, whereas amino acid changes to epitopes A and C could lead on average to an increase in neutralization (cf. Fig. 2, with epitope A corresponding to E1 and epitope C corresponding to E2). To test these predictions, we estimated the contribution of amino acid changes to individual HA epitopes to antigenic change (Materials and Methods). We quantified antigenic change by the reciprocal of the AHM measure of antigenic similarity, whereas amino acid change was quantified by the metric of Atchley et al. (26), denoted the A metric, and by the Hamming metric, denoted the H metric (Materials and Methods). The results (Fig. 4 and Table S2) suggest that the hypothesis that amino acid changes to (dominant) epitopes A, B, and D could lead on average to an increase in antigenic distance, whereas amino acid changes to (subdominant) epitopes C and E could lead on average to a decrease in antigenic distance cannot be rejected at the 95% confidence level. Thus, changes to epitopes A, B, and D could be highly favored by natural selection, whereas changes to epitopes C and E could be disadvantageous to influenza viruses.

Fig. 4.

Fig. 4.

Open in a new tab

Antigenic effects of amino acid changes to individual epitopes of influenza virus HA. Antigenic effects (regression coefficients) were estimated as described in Materials and Methods. The 95% confidence limits for each estimated regression coefficient, obtained by bootstrap resampling, are shown. A, B, C, D, and E denote the five epitopes of HA. O are epitopic sites that do not belong to any of the five epitopes, and N are sites not known to be bound by antibodies (see Table S1 for additional details).

In addition, the results suggest that nonepitopic sites and sites that do not belong to any of the five epitopes of HA but are nevertheless experimentally confirmed targets of monoclonal antibodies (Table S1) do not make a significantly positive contribution to antigenic change. The latter sites are also located further from the receptor-binding site than epitope E (Table 1), suggesting that they could have lower neutralization efficiencies than the dominant epitopes. Similar results were obtained when the computed genetic distances were normalized to account for differences in the number of amino acid positions found in each epitope and when genetic distances were quantified by using other popular metrics (Table S2). Note that when we summed the number of amino acid changes to all low-neutralization efficiency epitopes (A, B, and D) and the number of amino acid changes to all high-neutralization efficiency epitopes (C and E), we found a statistically significant (P < 0.05) antagonistic interaction between changes to the low- and high-neutralization efficiency epitopes. Although this result is consistent with the above predictions, it relies on the assumption that the interaction between changes to low- and high-neutralization efficiency epitopes is multiplicative.

Amino acid changes responsible for transitions between viral antigenic clusters.

A previous empirical study (5) of the evolutionary dynamics of influenza A/H3N2 led to the intriguing observation that whereas the genetic evolution of the virus is continuous, its antigenic evolution is characterized by discontinuous transitions between clusters of antigenically similar viruses. The transition from one antigenic cluster to another (containing viruses capable of escaping from induced antibodies) was generally associated with amino acid changes to two or more HA epitopes but, occasionally, with a single amino acid change. The results of the preceding section suggest that cluster transition should be favored by amino acid changes to epitopes with high neutralization efficiencies, but disfavored by changes to epitopes with low neutralization efficiencies. To test this hypothesis, we quantified the correlation between amino acid changes to individual epitopes of HA and transitions between each pair of temporally adjacent antigenic clusters K and K′, where K′ is the cluster that replaced K (Materials and Methods).

The results (Fig. 5) show that amino acid differences at epitopes A and B correlate positively with transitions from K to K′. In contrast, transitions from K to K′ are negatively correlated with amino acid differences at epitopes D and E, at epitopic sites that do not belong to any of the five epitopes of HA, and at sites not known to be bound by antibodies. The effect of amino acid differences at epitope C on cluster transitions could not be quantified accurately. With the exception of epitope D, the observed effects of amino acid differences are consistent with results presented in the preceding section and with the predictions of our antibody-interference model. Similar results were obtained when the number of amino acid differences at each epitope was adjusted to account for differences in the lengths of individual HA epitopes (Fig. S2).

Fig. 5.

Fig. 5.

Open in a new tab

Correlation between amino acid differences at individual HA epitopes and transitions between 10 pairs of temporally adjacent influenza A/H3N2 viral antigenic clusters. The correlation between amino acid differences and the transition from antigenic cluster K to K′ was quantified as described in Materials and Methods. In the figure, a positive regression coefficient indicates that an increase in the number of amino acid differences at the associated epitope increases the probability that a given virus belongs to K′. Error bars indicate standard errors of estimated regression coefficients, and they were computed by assuming that the number of viruses belonging to either cluster K or K′ is binomially distributed. A, B, C, D, E, O, and N are defined in the legend of Fig. 4.

Conclusion

We have argued that antibodies that bind to influenza virus HA epitopes located at a distance from the receptor-binding site could be less efficient at occluding the site (and hence at effecting viral neutralization) than antibodies that bind closer to the receptor-binding site. Importantly, the antibodies that bind at a distance from the receptor-binding site could sterically interfere with the antibodies than bind closer to the receptor-binding site (13, 14), thereby impeding viral neutralization. We suggested that the deleterious effects of such antibody interference on viral neutralization could explain some puzzling aspects of the observed antigenic effects of amino acid changes to HA epitopes. As an example, we predicted that amino acid changes to epitopes C and E of the HA of influenza A/H3N2 viruses would lead on average to a decrease in the antigenic difference between A/H3N2 viruses. We showed that existing genetic and antigenic data on A/H3N2 virus evolution support this prediction.

Considering the enormous evolutionary potential of influenza viruses it is possible that the antibody-interference mechanism described above has been exploited to destructive effect by these viruses, particularly as a means to escape from antibodies induced by prior infection and/or vaccination. Experimental studies of antibody interference aimed at the design of more effective, “low-interference” influenza vaccines may therefore be of value. For example, the HA of vaccine strains could be mutated in all regions except those that contain epitopes with high neutralization efficiencies (e.g., epitopes A and B of A/H3N2 HA). Such a vaccine would still raise antibodies against low-neutralization efficiency epitopes (e.g., epitopes C and E of A/H3N2 HA), but these low-neutralization efficiency antibodies would not interact with the target wild-type virus strains (Fig. 3 C–E). Alternatively, influenza vaccines could be designed to include only those regions of HA that correspond to epitopes with high neutralization efficiencies. Peptide-based influenza vaccines were proposed (27), but there has been little focus on peptides that correspond to epitopes with high neutralization efficiencies. Furthermore, antiinfluenza viral drugs could be designed to include HA proteins carrying modifications to high-neutralization efficiency epitopes; these modified HA proteins would compete with virus for binding to low-neutralization efficiency antibodies in a manner akin to the role played by neuraminidase inhibitors (28).

Because the antibody-interference mechanism described here is general in nature it could have implications for the evolution, neutralization, and management of other pathogens, including HIV type 1, for which there is evidence of competition between antibodies with different neutralization efficiencies (29). It is worth noting that our analyses of influenza viral neutralization were based on in vitro experimental data. Research needs to be done to determine precisely how much information these data provide about the neutralization of influenza viruses in vivo. Also, note that we did not account specifically for amino acid changes to HA resulting from interactions between HA and other viral proteins. The most important of these interactions, involving the viral neuraminidase protein (3), would most likely result in amino acid changes occurring primarily at the receptor-binding site of HA, and not at the epitopic sites that were of interest to us. (Note that HA must not bind to its receptor so strongly that the HA–receptor bond cannot be cleaved efficiently by the influenza viral neuraminidase protein.)

Materials and Methods

Quantifying the Antigenic Effects of Amino Acid Changes.

We quantified the antigenic effects of amino acid changes to individual HA epitopes by using the linear model log2(rAHMab) = Σiβidiab, where rAHMab denotes the reciprocal of the AHM measure of the antigenic similarity between viruses a and b (a measure of antigenic distance) (23), diab is the genetic distance between the ith epitope of the HAs of a and b, βi is the antigenic effect of genetic changes to the ith epitope, and ε is the residual. We used 72 pairs of influenza A/H3N2 viruses (Tables S3 and S4) to compute rAHMab and diab, with diab given by the metric of Atchley et al. (26), which accounts for antigenically relevant physicochemical properties of amino acids, and by the Hamming metric (see SI Methods for additional details). We computed 95% confidence limits for βi by (i) resampling (with replacement) the 72 combinations of rAHMab and diab 104 times, and (ii) assuming that log2(rAHMab) = N(Σiβidiab, σ), where σ denotes the standard deviation of antigenic distances and N denotes the Gaussian distribution.

Quantifying the Effects of Amino Acid Changes on Transitions Between Antigenic Clusters.

We compiled the HA amino acid sequences of influenza A/H3N2 viruses belonging to 11 antigenic clusters (ref. 5 and Table S5). For each pair of temporally adjacent antigenic clusters K and K′, with K′ the cluster that replaced K, we computed the number of amino acid differences, epitope by epitope, between the HA sequence of each virus and the consensus sequence of K. For each virus, we normalized the computed number of amino acid differences by the total number of amino acid differences between the viral HA sequence and the consensus sequence of K. We then used logistic regression to quantify the effect of amino acid differences at each epitope on the probability that the virus belongs to either K or K′.

Supplementary Material

Supporting Information
supp_106_21_8701__index.html (657B, html)

Acknowledgments.

We thank Nigel Dimmock, Lynn Enquist, Eddie Holmes, and Adel Mahmoud for very constructive comments on this manuscript and Sandi Milburn for excellent administrative assistance. This work was supported by a Burroughs–Wellcome graduate fellowship (to W.N.) and by Defense Advanced Research Projects Agency Grant HR0011-05-1-0057.

Footnotes

The authors declare no conflict of interest.

*

Neutralization efficiency refers to an intrinsic (e.g., independent of antibody affinity) capacity of an epitope to support viral neutralization by bound antibodies. Note that in cases when antibodies bind to influenza viruses multivalently, viral neutralization depends on both antibody avidity and the neutralization efficiency of the bound epitopes.

This article contains supporting information online at www.pnas.org/cgi/content/full/0903427106/DCSupplemental.

References

  • 1.World Health Organization. Influenza Factsheet. [accessed September 14, 2008];2003 www.who.int/mediacentre/factsheets/fs211/en/
  • 2.Flint SJ, Enquist LW, Racaniello VR, Skalka AM. Principles of Virology: Molecular Biology, Pathogenesis, and Control. Washington DC: Am Soc Microbiol; 2004. [Google Scholar]
  • 3.Palese P, Shaw ML. In: Field's Virology. Knipe DM, et al., editors. Philadelphia: Lippincott; 2006. [Google Scholar]
  • 4.Brown LE, Murray JM, White DO, Jackson DC. An analysis of the properties of monoclonal antibodies directed to epitopes of influenza virus hemagglutinin. Arch Virol. 1990;114:1–26. doi: 10.1007/BF01311008. [DOI] [PubMed] [Google Scholar]
  • 5.Smith DJ, et al. Mapping the antigenic and genetic evolution of influenza virus. Science. 2004;305:371–376. doi: 10.1126/science.1097211. [DOI] [PubMed] [Google Scholar]
  • 6.Lee MS, Chen JS. Predicting antigenic variants of influenza A/H3N2 viruses. Emerg Infect Dis. 2004;10:1385–1390. doi: 10.3201/eid1008.040107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Poljak RJ, et al. Three-dimensional structure of the Fab′ fragment of a human immunoglobulin at 2.8 Å resolution. Proc Natl Acad Sci USA. 1973;70:3305–3310. doi: 10.1073/pnas.70.12.3305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Wilson IA, Skehel JJ, Wiley DC. Structure of the haemagglutinin membrane glycoprotein of influenza virus at 3 Å resolution. Nature. 1981;289:366–373. doi: 10.1038/289366a0. [DOI] [PubMed] [Google Scholar]
  • 9.Jackson DC, Murray JM, White DO, Gerhard WU. Enumeration of antigenic sites of influenza virus hemagglutinin. Infect Immun. 1982;37:912–918. doi: 10.1128/iai.37.3.912-918.1982. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Schofield DJ, Stephenson JR, Dimmock NJ. High and low efficiency neutralization epitopes on the haemagglutinin of type A influenza virus. J Gen Virol. 1997;78:2441–2446. doi: 10.1099/0022-1317-78-10-2441. [DOI] [PubMed] [Google Scholar]
  • 11.Knossow M, et al. Mechanism of neutralization of influenza virus infectivity by antibodies. Virology. 2002;302:294–298. doi: 10.1006/viro.2002.1625. [DOI] [PubMed] [Google Scholar]
  • 12.Fleury D, et al. A complex of influenza hemagglutinin with a neutralizing antibody that binds outside the virus receptor binding site. Nat Struct Biol. 1999;6:530–534. doi: 10.1038/9299. [DOI] [PubMed] [Google Scholar]
  • 13.Taylor HP, Dimmock NJ. Competitive binding of neutralizing monoclonal and polyclonal IgG to the HA of influenza A virions in solution: Only one IgG molecule is bound per HA trimer regardless of the specificity of the competitor. Virology. 1994;205:360–363. doi: 10.1006/viro.1994.1653. [DOI] [PubMed] [Google Scholar]
  • 14.Taylor HP, Armstrong SJ, Dimmock NJ. Quantitative relationships between an influenza virus and neutralizing antibody. Virology. 1987;159:288–298. doi: 10.1016/0042-6822(87)90466-1. [DOI] [PubMed] [Google Scholar]
  • 15.Bizebard T, et al. Structure of influenza virus hemagglutinin complexed with a neutralizing antibody. Nature. 2002;376:92–94. doi: 10.1038/376092a0. [DOI] [PubMed] [Google Scholar]
  • 16.Sui J, et al. Structural and functional bases for broad-spectrum neutralization of avian and human influenza A viruses. Nat Struct Biol. 2009;16:265–273. doi: 10.1038/nsmb.1566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Garrity RR, et al. Refocusing neutralizing antibody response by targeted dampening of an immunodominant epitope. J Immunol. 1997;159:279–289. [PubMed] [Google Scholar]
  • 18.Pantophlet R, Burton DR. GP120: Target for neutralizing HIV-1 antibodies. Annu Rev Immunol. 2006;24:739–769. doi: 10.1146/annurev.immunol.24.021605.090557. [DOI] [PubMed] [Google Scholar]
  • 19.Fazekas de St Groth S, Webster RG. The neutralization of animal viruses. IV. Parameters of the influenza virus–antibody system. J Immunol. 1963;90:151–164. [Google Scholar]
  • 20.Gibson RE, Levin SA. Distinctions between the two-state and sequential models for cooperative ligand binding. Proc Natl Acad Sci USA. 1976;74:139–143. doi: 10.1073/pnas.74.1.139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Rini JM, Schulze-Gahmen U, Wilson IA. Structural evidence for induced fit as a mechanism for antibody–antigen recognition. Science. 1992;255:959–965. doi: 10.1126/science.1546293. [DOI] [PubMed] [Google Scholar]
  • 22.Webster R, Cox N, Stohr K. WHO Manual on Animal Influenza Diagnosis and Surveillance. Geneva, Switzerland: WHO; 2002. WHO/CDS/CSR/NCS/2002.5 Rev. 1. [Google Scholar]
  • 23.Archetti I, Horsfall FL. Persistent antigenic variation of influenza A viruses after incomplete neutralization in ovo with heterologous immune serum. J Exp Med. 1950;92:441–462. doi: 10.1084/jem.92.5.441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Ndifon W, Dushoff J, Levin SA. On the use of hemagglutination-inhibition for influenza surveillance: Surveillance data are predictive of influenza vaccine effectiveness. Vaccine. 2009;27:2447–2452. doi: 10.1016/j.vaccine.2009.02.047. [DOI] [PubMed] [Google Scholar]
  • 25.Bush RM, Fitch WM, Bender CA, Cox NJ. Positive selection on the H3 hemagglutinin gene of human influenza virus A. Mol Biol Evol. 1999;16:1457–1465. doi: 10.1093/oxfordjournals.molbev.a026057. [DOI] [PubMed] [Google Scholar]
  • 26.Atchley WR, Zhao J, Fernandes AD, Drüke T. Solving the protein sequence metric problem. Proc Natl Acad Sci USA. 2005;102:6395–6400. doi: 10.1073/pnas.0408677102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Naruse H, et al. A potential peptide vaccine against two different strains of influenza virus isolated at intervals of about 10 years. Proc Natl Acad Sci USA. 1994;91:9588–9592. doi: 10.1073/pnas.91.20.9588. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Gubareva LV, Hayden FG. In: Influenza Virology: Current Topics. Kawaoka Y, editor. Norwich, UK: Caister Academic; 2006. [Google Scholar]
  • 29.Verrier F, Nadas A, Gorny MK, Zolla-Pazner S. Additive effects characterize the interaction of antibodies involved in neutralization of the primary dualtropic human immunodeficiency virus type 1 isolate 89.6. J Virol. 2001;75:9177–9186. doi: 10.1128/JVI.75.19.9177-9186.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
supp_106_21_8701__index.html (657B, html)
0903427106_0903427106SI.pdf (1.1MB, pdf)

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES