Characterization of influenza vaccine immunogenicity using influenza antigen microarrays

Abstract

Background: Existing methods to measure influenza vaccine immunogenicity prohibit detailed analysis of epitope determinants recognized by immunoglobulins. The development of highly multiplex proteomics platforms capable of capturing a high level of antibody binding information will enable researchers and clinicians to generate rapid and meaningful readouts of influenza-specific antibody reactivity.

Methods: We developed influenza hemagglutinin (HA) whole-protein and peptide microarrays and validated that the arrays allow detection of specific antibody reactivity across a broad dynamic range using commercially available antibodies targeted to linear and conformational HA epitopes. We derived serum from blood draws taken from 76 young and elderly subjects immediately before and 28±7 days post-vaccination with the 2008/2009 trivalent influenza vaccine and determined the antibody reactivity of these sera to influenza array antigens.

Results: Using linear regression and correcting for multiple hypothesis testing by the Benjamini and Hochberg method of permutations over 1000 resamplings, we identified antibody reactivity to influenza whole-protein and peptide array features that correlated significantly with age, H1N1, and B-strain post-vaccine titer as assessed through a standard microneutralization assay (p<0.05, q <0.2). Notably, we identified several peptide epitopes that were inversely correlated with regard to age and seasonal H1N1 and B-strain neutralization titer (p<0.05, q <0.2), implicating reactivity to these epitopes in age-related defects in response to H1N1 influenza. We also employed multivariate linear regression with cross-validation to build models based on age and pre-vaccine peptide reactivity that predicted vaccine-induced neutralization of seasonal H1N1 and H3N2 influenza strains with a high level of accuracy (84.7% and 74.0%, respectively).

Conclusion: Our methods provide powerful tools for rapid and accurate measurement of broad antibody-based immune responses to influenza, and may be useful in measuring response to other vaccines and infectious agents.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Hemagglutinin (HA) structure and synthetic peptides.

(a) Side and top view of 3D structure of H1N1 A/Solomon Islands/3/2006 (Research Collaboratory for Structural Bioinformatics, Protein Data Bank PDB#3SM5) hemagglutinin (HA) trimer. (b) Individual HA monomer with head and stalk regions (left), head domain alone (center), and head domain showing the region spanning amino acids 53–291, highlighted in red (right), which we selected for synthesis of overlapping peptides. (c) Individual H1N1 (A/Brisbane/59/2007) peptides (1–24) are shown in the context of the head domain 3D structure (above) and amino acid sequence (below), with colors corresponding to each unique synthesized peptide. For depiction of peptides in c, we aligned the sequence of H1N1 A/Brisbane/59/2007 to H1N1 A/Solomon Islands/3/2006 and highlighted aligned homologous peptides. Seasonal H3N2 and B strain peptides are not depicted.

Figure 2. Specificity and broad dynamic range of peptide arrays.

(a) Images displaying reactive peptide features from an array probed with HA-tag antibody (eBioscience 14-6756-81). Amino acid sequence of peptide targets is displayed in parentheses. Residues corresponding to HA-tag sequence (NH2-YPYDVPDYA-COOH) are underlined in seasonal H3N2 peptide sequences. (b) Graph displaying MFI of array peptide features depicted in a in a titration experiment using HA-tag antibody ranging in concentration from 2000 ng mL−1 to 0 ng mL−1. Error bars in b reflect mean ± SEM of triplicate array feature MFI for each titration point.

Figure 3. Peptide arrays allow for discrimination of unique peptide antibody targets.

(a) Images displaying reactive HA-tag, seasonal H3N2 H3–4-7 and FLAG-tag peptide features on arrays probed with HA-tag antibody (14-6756-81, eBioscience) or FLAG-tag antibody (A9594, Sigma) either not pre-cleared, or after three rounds of clearing with HA-tag peptide or FLAG-tag peptide conjugated to streptavidin beads. (b) Histogram displaying percent (%) of maximum MFI for peptide features shown in a. Error bars in b represent the mean ± SEM of % of maximum MFI of triplicate array features. For each peptide, % of maximum MFI of the highest MFI replicate feature (of three) in the non-pre-cleared condition was calculated.

Figure 4. Decreased reactivity of commercial HA antibodies to recombinant HA proteins and influenza vaccine after pre-clearing with linear HA peptides.

(a) Heat map displaying reactivity of commercial antibodies raised against purified virus or recombinant HA on a nitrocellulose array containing recombinant HA proteins, trivalent influenza vaccine (TIV – Fluzone seasonal 2008/2009 vaccine) and U1-A spliceosome protein printed at indicated concentration. Scale bar reflects median fluorescence intensity (MFI) of antibody-bound array features in a and b. H1mAb, ab66189; H1pAb, ab91531; H3 mAb, ab66187 (Abcam); H3 pAb, IA-PAN4-0100 (eEnzyme). (b) Heat maps displaying reactivity of indicated antibodies when incubated on the influenza peptide array. Antibodies are the same as those described in a. (c) Highest reactivity peptides bound by H1pAb and H3pAb. (d) Histogram displaying percent of maximum reactivity of HA antibodies H1pAb and H3pAb cleared and not cleared with selected peptides to recombinant H1, recombinant H3, and Fluzone vaccine antigens. Error bars represent the mean ± SEM of % of maximum MFI of triplicate array features. For each antigen, % of maximum MFI of the highest MFI replicate feature (of three) in the non-pre-cleared condition was calculated.

Figure 5. Correlation of age and post-vaccine neutralization titer with influenza whole-protein array reactivity.

(a) Table displaying correlation coefficient (Pearson r), p value and false discovery rate (FDR) q value of whole-protein influenza antigen array reactivity with age and corrected post-vaccine neutralization titer (cPost) for H1N1, H3N2 and B influenza strains. Correlations with a p value <0.05 are highlighted in gray. Color scale represents the direction and magnitude of the Pearson r correlation coefficient. (b) Comparison of age and cPost H1N1, H3N2 and B influenza strain neutralization titer. P values were generated using a two-tailed student’s T test. (c) Dot plots displaying relationship between cPost neutralization titer (x axes) and change in array reactivity (MFI) to Fluzone trivalent influenza vaccine (TIV Δ, y axes). The line in each graph displays the least-squares linear regression of cPost titer and TIV Δ MFI, and corresponds to the Pearson r values shown in a.

Figure 6. Influenza HA peptide reactivity correlates with age/post-vaccine neutralization titer.

(a–d) Volcano plots displaying significance of the Pearson r correlation coefficient “−log(p value Pearson r)”, corresponding to the relationship of (a) age; (b) H1N1 cPost; (c) H3N2 cPost; and (d) B cPost neutralization titer with pre- and post-vaccine reactivity to H1N1, H3N2 and B-strain peptides. The Pearson r correlation coefficient is plotted on the x-axis. Red and blue color represents positive or negative correlation, respectively, of associations with a q value determined to be less than 0.2. Gray color indicates peptide associations that were not significant. (e) Table and heatmap displaying correlation coefficient (Pearson r), p value and false discovery rate (FDR) q value of the significant peptide reactivity shown in a–d with age, H1N1 cPost, H3N2 cPost, and B cPost. Color scale represents the direction and magnitude of the Pearson r correlation coefficient. See tables S3, S4, S5, S6, for the statistical results of all comparisons.

Figure 7. Analysis of age and pre-vaccine peptide reactivity can predict effective neutralization outcome.

(a,b) Tables showing peptides selected in a multivariate prediction model that identified vaccine recipients as upper or lower quartile responders with respect to H1N1 (a) and H3N2 (b) cPost neutralization titer verified using leave-one-out (LOO) cross-validation. The accuracy of using age alone or the model presented in a and b for prediction of good (upper quartile cPost titer) or poor (lower quartile cPost titer) response to vaccine is shown below. Color scale indicates direction and magnitude of the regression coefficient (β) for each element in a and b. (c) Peptides selected in multivariate prediction models of H1N1 cPost-vaccine titer with no age penalty, or increasing age penalty (increasing α age value) using the elastic net method (see methods). Black bars indicate adjacent peptides (overlapping sequence) selected in models using age penalty.