Influenza vaccine–induced human bone marrow plasma cells decline within a year after vaccination

See allHide authors and affiliations

Science  09 Oct 2020:
Vol. 370, Issue 6513, pp. 237-241
DOI: 10.1126/science.aaz8432

Immunity after the flu shot

The seasonal flu shot is currently recommended each year because the influenza viral strains in circulation are continuously changing and because the antibody responses produced by the vaccine decline over time. In a human study of healthy volunteers, Davis et al. tracked antibody responses after flu vaccination. They investigated whether the vaccine led to the generation of antibody-secreting plasma cells in the bone marrow, a lymphoid organ that supports the survival of these cells for years. Although vaccination did generate influenza-specific cells, most were short-lived and lost within 1 year. The fact that a small number did persist over 1 year raises prospects that the longevity of flu vaccines can be improved and provides key information for the development of universal vaccines against influenza.

Science, this issue p. 237


A universal vaccine against influenza would ideally generate protective immune responses that are not only broadly reactive against multiple influenza strains but also long-lasting. Because long-term serum antibody levels are maintained by bone marrow plasma cells (BMPCs), we investigated the production and maintenance of these cells after influenza vaccination. We found increased numbers of influenza-specific BMPCs 4 weeks after immunization with the seasonal inactivated influenza vaccine, but numbers returned to near their prevaccination levels after 1 year. This decline was driven by the loss of BMPCs induced by the vaccine, whereas preexisting BMPCs were maintained. Our results suggest that most BMPCs generated by influenza vaccination in adults are short-lived. Designing strategies to enhance their persistence will be a key challenge for the next generation of influenza vaccines.

The goal of vaccination is to generate long-lasting protection against infection. Most vaccines in clinical use achieve this protection at least in part through the generation of pathogen-specific antibody responses. Antibody levels peak in the months after vaccination, followed by a decline to a plateau level that may be maintained for decades with minimal decline (13). Animal models have shown that these plateau antibody levels are maintained by nondividing, bone marrow–resident long-lived plasma cells (4, 5). Studies of antibody synthesis rates have suggested that bone marrow plasma cells (BMPCs) are likely to produce most serum immunoglobulin G (IgG) in humans as well (6, 7). Consistent with this, total and antigen-specific serum antibody levels correlate closely with BMPC numbers in humans (8, 9). Antibody titers and protective immunity decline rapidly after seasonal influenza vaccination (10, 11), suggesting either that the vaccine may fail to elicit BMPCs or that these BMPCs fail to become long-lived. To investigate these possibilities, we designed a clinical study of healthy adults (20 to 45 years of age) from whom we were able to obtain paired bone marrow samples before and 1 month after influenza vaccination, in conjunction with a full schedule of blood samples. In addition, we obtained a third bone marrow draw at ~1 year after vaccination from some of the subjects, allowing us to assess the long-term maintenance of vaccine responses in the bone marrow.

BMPC responses were measured in 53 volunteers who received the inactivated influenza vaccine between 2009 and 2018. Some of these 53 donors enrolled in multiple study years (see the materials and methods) and were treated as separate enrollments, for a total of 75 vaccine responses studied. In addition to providing blood samples on the day of vaccination (day 0) and on days 7, 14, 28, and 90 after vaccination, the volunteers also underwent bone marrow aspiration procedures on days 0 and 28. In addition, some donors provided blood and bone marrow samples ~1 year after vaccination, either at their reenrollment for the following year’s study or as their final study time point.

We first examined influenza-specific immunity at baseline before vaccination. Using plasma cells enriched from the bone marrow aspirates, we performed enzyme-linked immunosorbent spot (ELISpot) assays to quantify influenza-specific BMPCs (Fig. 1A). Cells secreting influenza-specific antibodies (both IgG and IgA) were readily detectable in study volunteers before vaccination and were markedly more common than cells producing antibodies specific for tetanus toxin. Influenza-specific BMPC numbers were also five- to 10-fold higher than we have observed for varicella zoster virus (12). The percentage of influenza-specific IgG BMPCs on day 0 followed an approximately log-normal distribution (Fig. 1B). The geometric mean percentage of IgG-secreting BMPCs producing influenza-specific antibodies was 0.8%, with an interquartile range (IQR) of 0.52 to 1.3%. Across the cohort, the percentage of influenza-specific IgG+ BMPCs correlated significantly with influenza-specific IgG titers in the blood (Fig. 1C). These data are consistent with a key role for BMPCs in maintaining humoral immunity to influenza. We also measured influenza-specific memory B cells (MBCs) in the blood of the donors. Influenza-specific cells accounted for 0.93% (IQR: 0.58 to 1.5%) of IgG MBCs in the blood (Fig. 1D). This MBC frequency correlated with blood antibody titers (Fig. 1E). Furthermore, there was a clear trend toward higher numbers of influenza-specific MBCs in those donors with the most influenza-specific BMPCs (Fig. 1F).

Fig. 1 Analysis of influenza-specific BMPCs before vaccination.

(A) Example ELISpot data using magnetically enriched BMPCs with detection of cells secreting total or antigen-specific IgG or IgA. (B) Number of BMPCs producing influenza-specific IgG before vaccination expressed as a percentage of the total number of IgG-secreting BMPCs. (C) Correlation between prevaccination influenza-specific IgG BMPCs and influenza-specific IgG levels in the blood as measured by Enzyme-linked immunosorbent assay (ELISA). (D) Influenza-specific IgG MBCs from the peripheral blood are shown as a percentage of total IgG MBCs before vaccination. (E) Correlation between influenza-specific IgG MBCs and IgG titers in the blood before vaccination. (F) Correlation between the percentages of influenza-specific IgG BMPCs and MBCs. The data shown in (B) to (F) are for all study donors for which day 0 influenza-specific BMPC and MBC percentages were measured (n = 59 including repeat study participants; n = 43 unique donors). The means and standard deviations of the log-transformed data are indicated by red lines in (B) and (D). In (C), (E), and (F), Pearson correlation analysis was performed on the log-transformed BMPC, MBC, and ELISA data. The green dots represent a donor with evidence of recent influenza infection who was excluded from statistical calculations.

We next investigated whether seasonal influenza vaccination resulted in the production of new influenza-specific BMPCs. Influenza-specific BMPCs increased significantly to 1.9% (IQR: 1.2 to 2.9%) of IgG BMPCs by day 28 after vaccination (Fig. 2A and fig. S1A). The seasonal influenza vaccine contains hemagglutinin (HA) proteins from multiple viral strains: one from an H1N1 influenza A strain, one from an H3N2 strain, and either one or two HA proteins from influenza B strains. An HA protein similar or identical to that of the pandemic H1N1 strain A/California/7/2009 was included in the vaccine from the 2010–2011 season until the 2016–2017 season, allowing us to monitor responses against a single, consistent antigen in multiple seasons. IgG BMPCs specific for this HA increased from 0.11% (IQR: 0.07 to 0.18%) to 0.36% (IQR: 0.18 to 0.76%) on day 28 (Fig. 2A and fig. S1B). We also measured influenza-specific IgA BMPCs when plasma cell yields were high enough to run multiple antigen-specific ELISpot assays from each sample. Although we only obtained paired IgA data for six donors, there was a trend toward increasing numbers of IgA BMPCs (Fig. 2A). This trend was significant in the unpaired analysis (fig. S1C), which included additional donors for whom IgA BMPCs could be measured at day 0 or day 28 but not both. These increases were specific to plasma cells recognizing influenza vaccine antigens; no changes were observed in the percentage of cells secreting IgG specific for tetanus toxin (Fig. 2A and fig. S1C). Thus, we found no evidence of bystander effects of the vaccine on the frequencies of BMPCs specific for unrelated antigens.

Fig. 2 Increase in influenza-specific plasma cells in the bone marrow after vaccination.

(A) The percentage of IgG- or IgA-secreting BMPCs specific for the indicated antigens was measured by ELISpot on the day of vaccination (d0) and 28 days later (d28). Influenza, trivalent or quadrivalent inactivated influenza vaccine (n = 46 IgG responses measured from 34 unique donors; n = 6 IgA responses from six unique donors); pdm H1 HA, recombinant HA protein from pandemic H1N1 strain A/California/07/2009 (n = 20 responses from 14 unique donors); tetanus, tetanus toxin C fragment (n = 6 responses from six unique donors). P values shown are for paired, two-tailed t tests using the log-transformed data. The means and standard deviations of the log-transformed data are indicated in red. (B) Blood IgG ASC response to the influenza vaccine measured by ELISpot. Data are shown for n = 45 vaccine responses from 35 unique donors. The dotted line represents the lower limit of detection for the assay. (C) Pearson correlation between the magnitude of the IgG ASC response at day 7 and the increase in influenza-specific BMPCs, calculated as the percentage of influenza-specific IgG BMPCs at day 28 minus this percentage at day 0. Green symbols and lines in (A) to (C) indicate a donor with evidence of recent influenza infection who was excluded from the statistical analyses.

We have previously shown that influenza-specific antibody-secreting cells (ASCs) appear in the blood at ~1 week after vaccination (13). Because bone marrow aspirate samples will unavoidably contain some level of contamination with peripheral blood, we wanted to confirm that influenza-specific ASCs were not in circulation at the time of bone marrow aspiration. Influenza-specific IgG ASCs were at very low levels or undetectable in the blood on day 0, with the exception of one donor with a recent history of an influenza-like illness (green symbols). ASC numbers peaked at day 7, declined sharply by day 14, and were again very rare or undetectable on day 28 (Fig. 2B). Similar results were seen for IgA and IgM ASCs (fig. S1, D and E). Thus, the percentages of influenza-specific BMPCs on days 0 and 28 would not have been inflated by contamination with circulating ASCs. The magnitude of the day 7 ASC response correlated with the increase in influenza-specific BMPCs after vaccination (Fig. 2C). This is consistent with models suggesting that BMPCs are derived from blood ASCs that migrated to the bone marrow (14).

To determine whether the increase in influenza-specific BMPCs was maintained, we obtained bone marrow samples from a subset of donors at a late time point (mean 325 days after vaccination; range: 220 to 457 days). The percentage of influenza-specific BMPCs declined significantly between day 28 and the late sample (Fig. 3A). In subjects who reenrolled in multiple seasons and were vaccinated a second time, influenza-specific BMPCs typically increased after both immunizations, although there was substantial heterogeneity among donors (fig. S2). As shown in Fig. 3B, in the subset of donors who provided a long-term follow-up sample, there was a geometric mean 1.9-fold change (IQR: 1.3- to 3.2-fold) in influenza-specific BMPCs between days 0 and 28, similar to the overall cohort. However, by 1 year, the percentage of influenza-specific BMPCs had declined to baseline. Analysis of serum antibody levels revealed similar trends (Fig. 3C). The level of decline in BMPCs did not appear to correlate with the timing of the 1-year bone marrow aspiration (fig. S3), suggesting that the loss of newly generated influenza-specific BMPCs happened within 7 months after vaccination.

Fig. 3 Decline of influenza-specific BMPCs at 1 year after vaccination.

(A) Percentage of influenza-specific IgG BMPCs shown over time in volunteers who returned for a third bone marrow aspiration ~1 year after vaccination. P values shown are for paired, two-tailed t tests of the log-transformed data. (B) Same data plotted as in (A) except the percentages of influenza-specific BMPCs at day 28 and 1 year are normalized to the day 0 percentage to visualize the fold change relative to prevaccination levels in each donor. Numbers above the points indicate the geometric means of the fold changes, with the IQRs in parentheses. (C) Influenza-specific blood IgG ELISA titers at day 28, day 90, and 1 year normalized to the day 0 titer and plotted as in (B). The same n = 18 responses from 15 unique donors are shown in all panels.

We next examined in detail the clonal composition of the B cell response to the vaccine by performing immunoglobulin heavy chain repertoire sequencing of BMPCs and ASCs from four donors. Three out of four donors (donors 35, 40, and 42) had a strong day 7 ASC response to the vaccine (fig. S4A), and all four donors showed serologic responses to each influenza strain in the vaccine (fig. S4B), particularly to the H1N1 component. To identify influenza-specific antibody families (or “clonotypes”) in the repertoire sequencing data, we used two strategies: production of monoclonal antibodies (mAbs) cloned from day 7 ASCs (all four donors) and a proteomics-based approach in which antibodies specific to the H1 HA protein were identified from the polyclonal serum (donors 35 and 42) (fig. S5A). We identified 15 clonotypes that were highly expanded (>0.5% of day 7 ASC sequences) and determined to be influenza specific by either mAb cloning or proteomics analysis (fig. S5B and table S5). These antibody families already exhibited extensive somatic hypermutation in their heavy chain variable region genes at day 7, indicating that they were likely derived from preexisting MBCs. All 15 clonotypes were represented among both IgA- and IgG-expressing ASCs, although some clonotypes were enriched for one isotype or the other. We hypothesize that multiple MBCs from each clonotype responded to the vaccine, with varying recruitment of IgA versus IgG memory cells. Of the 11 H1-specific clonotypes that we identified, only one of them (clonotype 8 from donor 42) contributed detectably to H1-binding antibodies found in the serum on day 0 (fig. S5B). The other 10 H1-specific clonotypes could only be detected in the serum starting at day 7 or day 14. This suggests either that there were few long-lived plasma cells belonging to these clonotypes before vaccination or that these preexisting plasma cells produced antibodies that bound poorly to the pandemic H1 HA protein.

Of the 15 influenza-specific clonotypes that were expanded in the day 7 ASCs, 14 increased significantly among BMPC IgG sequences between days 0 and 28 after vaccination (P = 0.0003 by Wilcoxon signed-rank test; Fig. 4A). Thus, the same antibody families that contributed to the blood ASC response also contributed to the BMPC compartment, again consistent with the idea that blood ASCs give rise to BMPCs. Between day 28 and the late time point (~1 year), the percentage of BMPC IgG sequences belonging to these clonotypes decreased sharply (P = 0.0017). After 1 year, 11 of the 14 clonotypes still made up a higher percentage of BMPC IgG reads compared with day 0; however, this difference was not statistically significant (P = 0.068). There was a wide variation in how well the initial increase in frequency of each clonotype among BMPC sequences was sustained (Fig. 4B). The median increase in frequency for these 14 clonotypes after 1 year was only 0.15-fold as high as the initial increase seen at day 28 (IQR: 0.01- to 0.29-fold). In other words, for most clones, 70 to 99% of the newly generated BMPCs were lost over 1 year. As a separate strategy to measure clonotype abundances over time, we designed clone-specific quantitative polymerase chain reaction (qPCR) probe and primer sets (fig. S5C). Using this qPCR assay, we observed a similar trend where vaccine-specific antibody clonotypes expanded and then declined in BMPC RNA samples (median increase in frequency at 1 year was 0.09-fold of the initial increase at day 28; IQR: 0- to 0.38-fold). Both the initial increase in clonotype frequency among BMPCs and the long-term increase correlated with the representation of each clonotype among day 7 ASC sequences (Fig. 4, C and D), which is consistent with what we observed by ELISpot (Fig. 2C) and in keeping with the notion that blood ASCs give rise to BMPCs.

Fig. 4 Longitudinal tracking of influenza–specific antibody lineages (clonotypes) after vaccination.

Clonotypes were tracked over time in BMPC and ASC immunoglobulin sequences. Fifteen clonotypes comprising at least 0.5% of day 7 blood ASC IgG sequences were considered to have been “recruited” into the vaccine-specific response. Clonotypes are named by their frequency among ASC sequences (e.g., “35-2” is the clonotype with second-highest read count in IgG sequences from day 7 ASCs from donor 35). (A) Percentage of BMPC IgG sequences belonging to each recruited clonotype shown on day 0, day 28, and 1 year after vaccination. (B) For the 14 recruited clonotypes that increased in the bone marrow between days 0 and 28, the long-term increase after 1 year is shown as a proportion of the initial increase. A clonotype that neither expanded nor contracted between day 28 and 1 year would have a value of 1 on this scale, whereas a clonotype that declined to its starting frequency would have a value of 0. Values >1 indicate that the clonotype expanded among BMPC sequences between day 28 and 1 year, and values below 0 indicate clonotypes that comprised a lower percentage of IgG reads at 1 year compared with day 0. (C) Correlation between the percentage of day 7 ASC IgG sequences belonging to each clonotype and the initial increase in that clonotype among BMPC IgG transcripts between days 0 and 28. (D) Correlation between the percentage of day 7 ASC IgG sequences belonging to each clonotype and the long-term increase of these clonotypes among BMPC sequences between day 0 and 1 year. (E) Persistence of preexisting, nonrecruited H1 HA-specific clonotypes in donor 35. Clonotypes were defined as preexisting if they were found in the BMPC repertoire sequencing data at day 0 and confirmed to be influenza specific by serum proteomics analysis or mAb production. Clonotypes were defined as nonrecruited if they accounted for <0.1% of day 7 IgG, IgM, and IgA ASC sequences. (F) Persistence of preexisting, nonrecruited H1 HA-specific clonotypes in donor 42. P values shown in (A), (E), and (F) are for the Wilcoxon signed-rank test. In (C) and (D), Pearson correlation was performed using the nontransformed data, and the best-fit linear regression lines are shown.

Our clonotype-tracking results suggested that the contraction in influenza-specific BMPCs between day 28 and 1 year that we saw by ELISpot was due to the loss of newly generated cells. To determine whether preexisting influenza-specific BMPCs were more stable over time, we examined “nonrecruited” influenza-specific clonotypes in the two donors whose H1-specific responses were profiled by serum proteomics analysis. These clonotypes were defined as HA-specific antibody families that were detected in the serum and in IgG BMPC sequences on day 0, but which represented <0.1% of day 7 vaccination-induced ASCs by antibody repertoire sequencing (tables S4 and S5). Most of these clonotypes could be detected among both IgG and IgA BMPC sequences (see table S5). The nonrecruited clones showed no significant change in their abundances over 1 year in either donor (Fig. 4, E and F), indicating that they likely represented a stable population of long-lived plasma cells that had been generated by previous influenza infection or vaccination.

In recent years, it has become clear that humans are capable of generating protective antibody responses directed against epitopes on the influenza HA and neuraminidase (NA) proteins that are broadly conserved across diverse viral strains. This has raised hopes that a “universal” influenza vaccine targeting these conserved regions could be developed. If such a vaccine elicited protective antibody responses that persisted for at least several years, it would eliminate the need for annual immunization, which would be a major improvement on current vaccines. The data presented here add to our understanding of why humoral responses to influenza vaccine are not well maintained. We show that influenza-specific titers in the serum correlate with BMPC numbers before vaccination, supporting a key role for these cells in maintaining antibody levels. Vaccination does lead to the generation of new BMPCs, and some of these persist for at least 1 year. However, most of the newly generated BMPCs are lost within 1 year, showing that localization to the bone marrow is not enough to determine longevity. Two steps are required for a BMPC to become a long-lived plasma cell: (i) The cell must reach the appropriate survival niche and successfully compete for space there and (ii) it must undergo changes in gene expression and metabolism that promote longevity (1520). Thus, our data suggest that most vaccine-induced BMPCs fail at one or both of these steps.

The decline in BMPC and antibody levels observed here was likely not specific to the influenza vaccine. One study of children receiving the hepatitis B vaccine showed declines in antibody titers of 85 to 90% from peak levels in the first year after vaccination (21). Conversely, in a study of adults given a first booster of the tetanus, diphtheria, and acellular pertussis vaccine (Tdap), IgG antibody levels to the vaccine antigens declined only 40 to 66% in the first year and another 15 to 25% in the following year (22). This is a considerably slower rate of decay than we observed, suggesting that there is room for improvement with influenza vaccines.

A few findings from this study may be helpful for designing next-generation influenza vaccines capable of eliciting longer-term immunity. First, the blood ASC response after vaccination was highly predictive of the overall BMPC response, both in the short term and in the long term, and thus may be a useful marker for screening vaccine candidates. Because ASC generation after influenza vaccination can be increased through the use of adjuvants (2325), it may be useful to assess whether adjuvants are able to improve long-term BMPC responses in humans. Second, most of the preexisting H1-specific antibody families detected in the serum on day 0 were not detected in the day 7 ASC response, and most of the H1-specific clonotypes that were recruited did not contribute appreciably to serum H1-specific titers before vaccination. This may reflect antibody feedback, in which preexisting serum antibodies binding to the vaccine block B cells recognizing the same epitopes from acquiring antigen (2630). Designing strategies to overcome this effect will be critical for boosting long-term numbers of BMPCs targeting conserved epitopes.

Supplementary Materials

Materials and Methods

Tables S1 to S5

Figs. S1 to S5

References (3142)

References and Notes

Acknowledgments: We thank the Emory Flow Cytometry Core for assistance with cell sorting, the Emory Integrated Genomics Core for assistance with MiSeq experiments, and R. Antia and V. Zarnitsyna for valuable discussions. Funding: NIAID Centers of Excellence for Influenza Research and Surveillance (CEIRS) contract HHSN272201400004C; Influenza Pathogenesis and Immunology Research Center (IPIRC, CEIRS) contract HHSN266200700006C; NIH grants 1R01AI127877 and 1R01AI130398 (SDB). Author contributions: Conceptualization: C.W.D., J.W., E.K.W., R.A.; Data curation: C.W.D., K.J.L.J., J.D., R.G., W.C.C.; Formal analysis: C.W.D., K.J.L.J.; Investigation: C.W.D., K.J.L.J., M.M.M., J.D., C.Che, C.Cha, S.L.L., W.C.C.; Methodology: C.W.D., K.J.L.J., M.M.M., J.D., J.W., W.C.C., S.D.B., E.K.W., R.A.; Project administration: E.K.W., R.A.; Funding acquisition: E.K.W., R.A.; Resources: K.J.L.J., J.D., R.G., W.C.C., A.K.M., S.D.B., E.K.W.; Software: K.J.L.J., S.D.B.; Supervision: W.C.C., S.D.B., E.K.W., R.A.; Visualization: C.W.D., K.J.L.J.; Writing – original draft: C.W.D., R.A.; Writing – Review & Editing: C.W.D., K.J.L.J., S.D.B., E.K.W., R.A. Competing interests: The authors declare no competing interests. Data and materials availability: Antibody high-throughput DNA-sequencing data have been deposited in the Sequence Read Archive (BioProject accession no. PRJNA649987).

Stay Connected to Science

Navigate This Article