Plasmablast antibody repertoires in elderly influenza vaccine responders exhibit restricted diversity but increased breadth of binding across influenza strains

Abstract

Seasonal influenza vaccines elicit antibody responses that can prevent infection, but their efficacy is reduced in the elderly. While a subset of elderly individuals can still mount sufficient vaccine-induced antibody responses, little is known about the properties of the vaccine-induced antibody repertoires in elderly as compared to young responders. To gain insights into the effects of aging on influenza vaccine-induced antibody responses, we used flow cytometry and a cell-barcoding method to sequence antibody heavy and light chain gene pairs expressed by individual blood plasmablasts generated in response to influenza vaccination in elderly (aged 70-89) and young (aged 20-29) responders. We found similar blood plasmablast levels in the elderly and young responders seven days post vaccination. Informatics analysis revealed increased clonality, but similar heavy chain V(D)J gene usage in the elderly as compared to young vaccine responders. Although the elderly responders exhibited decreased antibody sequence diversity and fewer consequential mutations relative to young responders, recombinant antibodies from elderly responders bound a broader range of influenza strain HAs. Thus elderly influenza vaccine responders mount plasmablast responses with restricted diversity but with an increased breadth of binding across influenza strains. Our results suggest that the ability to generate plasmablast responses encoding cross-strain binding antibodies likely represents a mechanism important to vaccine responses in the elderly.

Trial registration: ClinicalTrials.gov NCT01827462.

Conflict of interest statement

Potential conflicts of interest

W.H.R. owns equity in, is a consultant to, and serves on the Board of Directors of Atreca, Inc. All other authors declare no potential conflicts of interest.

Published by Elsevier Inc.

Figures

Figure 1

Analysis of the blood plasmablast response in young and elderly responders to the seasonal influenza vaccine. Samples from 14 young (age 20-29) or elderly (age 70-89) responders enrolled in a seasonal trivalent inactivated influenza vaccine study (TIV, 2010/2011) were selected for analysis. (A) Seroconversion rates of the selected young or elderly vaccine responders to the 2010 seasonal influenza vaccine. All analyzed individuals were defined as influenza vaccine responders based on a ⩾ 4 fold increase in HAI titers against 2 or 3 vaccine strains 28 days following vaccination. (B) Percent B cells in the young and elderly responders 7 days following vaccination. (C) Relative plasmablast numbers in the young and elderly responders 7 days following vaccination. Data are the mean ± SEM. *, p < 0.05 by t test.

Figure 2

Analysis of the IgG plasmablast repertoire in young and elderly influenza vaccine responders. (A) Clustering of paired heavy chain and light chain dendrograms based on the heavy chain V region usage of individual plasmablasts. Each node at the end of a branch of a phylogenetic tree represents one plasmablast. Antibodies in clonal families are colored red. (B) Clonal vs non-clonal plasmablasts in the antibody repertoires of the young and elderly responders. (C) Increased clonal family size in the elderly responders. Each blue circle denotes the size of a clonal family in a young subject, and each red square denotes the size of a clonal family in an elderly subject. The mean ± SEM of each group are indicated. *, p <0.05 by t test.

Figure 3

Similar V and J gene usage in the heavy chains of plasmablast antibodies from the young and elderly responders. The germline V gene usage plots for the entire antibody repertoire (A) and for antibodies that are in clonal families (B) from young and elderly responders. (C) Chord diagram illustrates the heavy chain usage of the major V and J gene combination. Each chord is scaled to the frequency of antibodies by the corresponding gene, and each arc is scaled to the number of antibodies by gene segment combination.

Figure 3

Similar V and J gene usage in the heavy chains of plasmablast antibodies from the young and elderly responders. The germline V gene usage plots for the entire antibody repertoire (A) and for antibodies that are in clonal families (B) from young and elderly responders. (C) Chord diagram illustrates the heavy chain usage of the major V and J gene combination. Each chord is scaled to the frequency of antibodies by the corresponding gene, and each arc is scaled to the number of antibodies by gene segment combination.

Figure 3

Similar V and J gene usage in the heavy chains of plasmablast antibodies from the young and elderly responders. The germline V gene usage plots for the entire antibody repertoire (A) and for antibodies that are in clonal families (B) from young and elderly responders. (C) Chord diagram illustrates the heavy chain usage of the major V and J gene combination. Each chord is scaled to the frequency of antibodies by the corresponding gene, and each arc is scaled to the number of antibodies by gene segment combination.

Figure 4

Analysis of somatic hypermutations and CDR3 lengths of the plasmablast antibody repertoires in young and elderly influenza vaccine responders. Mutation rates in the heavy chain are higher (A), and are predominantly non-silent (B), in the young group. (C) CMV seropositive individuals exhibited higher mutation frequencies in their heavy chains of clonally expanded antibodies as compared to CMV seronegative individuals. (D) The length of the CDR3 region in the heavy (left) and light (right) chains of clonal families is longer in the young responders. Each dot represents data from an individual. The mean ± SEM of each group are also indicated. *, p <0.05 by t test.

Figure 5

Recombinant antibodies expressed from plasmablast clonal families in the elderly responders bound to HA1 peptides derived from a broader range of influenza strains. Heatmap of median fluorescence intensity (MFI) is displayed for recombinant antibody binding to HA1 peptides from influenza strains prevalent in previous influenza seasons. Antibodies cloned from plasmablasts in the young vs. elderly responders are listed along the top of the heatmap, and the different HA1 peptides in various seasonal influenza strains are listed to the right. Blue boxes demarcate no reactivity, yellow moderate reactivity, and orange-red high reactivity.