Repeated vaccination with homologous influenza hemagglutinin broadens human antibody responses to unmatched flu viruses
- Ragon Institute of Mass General, MIT, and Harvard, United States
- Department of Physics, Massachusetts Institute of Technology, United States
- Center for Vaccines and Immunology, University of Georgia, United States
- Department of Infectious Diseases, University of Georgia, United States
- Department of Microbiology, Harvard Medical School, United States
- Department of Chemical Engineering, Massachusetts Institute of Technology, United States
- Department of Chemistry, Massachusetts Institute of Technology, United States
- Institute for Medical Engineering and Science, Massachusetts Institute of Technology, United States
Figures
Figure 1 with 1 supplement
Divergent amino acid relatedness in the ectodomain and receptor-binding site (RBS) patch of the pandemic influenza HA.
(A) The hemagglutinin (HA) ectodomain, where relatedness is calculated using the formula ‘N_matched/N_total’; N_matched is the number of amino acids that match between the compared sequences and N_total is the total number of amino acids in the aligned sequence. (B) Heat map of HA ectodomain relatedness values for influenza A (H3N2, H1N1) and B viruses spanning almost 100 years (HA ectodomain sequences analyzed). (C) The RBS patch was structurally identified by four human bnAbs whose paratopes engage the RBS by mimicking cell surface sialic acid (CH67, CH67, H2526, 641 I-9) (Schmidt et al., 2015). We defined the RBS patch as the viral sialic acid binding residues (black) + the surrounding antibody epitope ‘ring’, collectively identified by the peripheral contacts made by the four bnAbs. Amino relatedness within the RBS patch is then calculated using the same formula except that the residues are now restricted to patch. (D) Heat map of HA RBS patch relatedness values for influenza A (H3N2, H1N1) and B viruses spanning almost 100 years (RBS patch sequences from the same 38 HA sequences as in B). See also Figure 1—figure supplement 1 for extended resolution on the heat map scale.
Figure 1—figure supplement 1
Relatedness heat maps with extended resolution.
The scale of the amino acid relatedness heat maps was expanded to visually resolve differences amongst non-pandemic influenza virus strains (compare to Figure 1). Amino acid relatedness values in the full-length hemagglutinin (HA) ectodomain (A) and receptor-binding site (RBS) patch (B) for the 38 influenza viral strains covering influenza A (H3N2, H1N1) and B viruses spanning >100 years are directly taken from Figure 1.
Figure 2
Sequential immunization with homologous pHA also elicits hemagglutination inhibition (HAI) against highly unrelated H1N1 strains.
(A) Four-year influenza vaccine trial (Nuñez et al., 2017). We analyzed HAI elicited from subjects that were longitudinally followed and immunized each year with the vaccine strains indicated. Notably, these individuals received the same H1N1 component (A/California/07/2009 = pHA) in each of the 4 years. (B) Fold change in HAI titer (pre vs 20 days post-vaccination) elicited each year and graphed as a function of hemagglutinin (HA) ectodomain relatedness between the vaccine strain and the viruses within the HAI panels. Each dot is a single subject at the relatedness value: white dots are fold changes for strains from the virus panel; the colored dots indicate the vaccine-matched viral strain (relatedness = 1.00). (C) Same data as in (B) only now graphed as a function of receptor-binding site (RBS) patch relatedness between the vaccine strain and the viruses within the panels.
-
Figure 2—source data 1
Hemagglutination inhibition (HAI) values for the influenza virus strains, measured across longitudinal vaccine study (2013–2016) for n = 136 de-identified subjects >50 years of age and <38 years of age.
This is the same source data used for Figure 3, Figure 3—figure supplement 1.
- https://cdn.elifesciences.org/articles/107042/elife-107042-fig2-data1-v1.xlsx
Figure 3 with 1 supplement
Sequential immunization with homologous pHA gradually broadens the response within individuals with no initial immune memory/recall to historical strains.
Responders (green) versus non-responders (red) within each year is graphed for each H1N1 strain in the hemagglutination inhibition (HAI) panel. Responders are defined by having non-decreasing fold changes in HAI titers (post-vaccination HAI titer/pre-vaccination HAI titer; i.e. fold change >1). Non-responders are defined by having decreasing fold changes of HAI titers (post-vaccination HAI titer/pre-vaccination HAI titer; fold change <1). Because non-responders (red) do not back-boost against historical strains in the panel, they, by definition, lack imprinted immunity to these viruses that is recalled by pHA. In the regression analyses, each white dot denotes the proportion of non-responders for each viral strain. (A) Yearly response for all longitudinally analyzed individuals; at right is a linear regression of the proportion of non-responders over the 4-year vaccine data (p = 6e−04). (B) Data for subjects >50 years in age (p = 6e−04, linear regression). (C) Data for subjects <38 years in age (p = 0.0164, linear regression). See also Figure 3—figure supplement 1 for linear regression of the proportion of responders in each age group.
Figure 3—figure supplement 1
H1N1 responders regressed over the vaccine.
Responders (green) versus non-responders (red) within each year are graphed for each H1N1 strain in the hemagglutination inhibition (HAI) panel, as in Figure 3. (A) Yearly response for all longitudinally analyzed individuals; at right is a linear regression of the proportion of non-responders against over the 4-year vaccine data (p = 6e−04). (B) Data for subjects >50 years in age (p = 6e−04, linear regression). (C) Data for subjects <38 years in age (p = 0.0164, linear regression).
Figure 4
The influenza hemagglutinin (HA) head is coarse-grained into three epitopes that are perceived with different germline-endowed B cell affinities.
(A) Diagram of epitope differences. In the right panel, the level of conservation of the three epitopes is depicted using different shapes (not very conserved) or similar shapes (relatively conserved). Epitope 1 (dominant epitope on pHA) is not conserved between the three variants. Epitope 2 (subdominant epitope) is relatively conserved between strain 1 (vaccine strain) and strain 2, but not between strains 1 and 3. Epitope 3 (another subdominant epitope) is conserved between strains 1 and 3, but not between strains 2 and 3. (B) Germline-endowed affinity distribution of naive B cells. Germline B cells targeting more dominant epitopes are more numerous and exhibit a longer high-affinity tail. Epitope 1 is more dominant than epitope 2, and epitope 2 is more dominant than epitope 3. Here, the fractions of naive B cells targeting epitope i are .
Figure 5 with 4 supplements
Antibody broadening via feedback regulation of the humoral response.
(A) The antibody titers against both the vaccine strain and historical strains (strains 2 and 3) increase over four immunizations. The antibodies are produced by plasma cells from both the germinal centers (GCs) and the extra germinal centers (EGCs). Antibody coverage increases first for strain 2 (after the second immunization), and strain 3 is engaged after the third immunization. In this simulation, the initial fractions of B cells that target epitope i are . The conservation of epitope 2 between strains 1 and 2 and the conservation of epitope 3 between strains 1 and 3 are both equal to 0.95. (B) The expansion of pathogen-specific memory B cells from the first immunization and differentiation into plasma cells that produce antibodies significantly increases the antigen concentration on follicular dendritic cells (FDCs) in the second immunization. This allows lower-affinity B cells that target subdominant epitopes to enter GCs and undergo affinity maturation. The antigen concentration on FDCs slightly increases from the second to the third immunization, allowing more B cells that target the subdominant epitopes to enter GCs and undergo affinity maturation. (C) The distribution of memory cells produced in the GCs during the first three immunizations. Upon subsequent antigen exposure, these memory cells are selected and expanded in EGCs. Thus, they contribute significantly to circulating antibodies and increased titers during subsequent immunizations. The first immunization primarily produces memory cells that target the dominant epitope (epitope 1), along with some memory cells targeting epitope 2. The second and third vaccinations produce an overall greater number of memory cells bearing generally higher affinity for the subdominant epitopes (epitopes 2 and 3) than the first immunization. (D) Strain 3 is engaged less potently when the initial fractions of B cells that target epitope i are . (E) The titers against strain 2 are lower than titers against strain 3 when the conservation of epitope 2 is decreased. Here the conservation of epitope 2 between strains 1 and 2 is 0.7 while the conservation of epitope 3 between strains 1 and 3 is kept at 0.95. Other values of are explored in Figure 5—figure supplement 3. The fractions of B cells that target epitope i are the same as those in A.
Figure 5—figure supplement 1
Booster shots of homologous pHA increase the number of high-affinity memory cells that target subdominant epitopes.
The distribution of memory cells produced by the germinal centers (GCs) against each epitope in the first three immunizations of the computational simulations, as in Figure 5C. Upon subsequent antigen exposure, these memory cells are selected and expanded in extra germinal centers (EGCs). Thus, they contribute significantly to circulating antibodies and increased titers during subsequent immunizations. The effects of epitope masking are not considered here. Among the high-affinity memory cells produced in the first immunization, relatively few target the subdominant epitopes (epitopes 2 and 3) compared to the dominant epitope (epitope 1). After the second and third immunizations, the number of memory cells increases for all epitopes but especially for epitopes 2 and 3.
Figure 5—figure supplement 2
Germinal centers (GCs) that form upon booster shots persist longer than those that form upon primary immunization.
Computational results show that the GCs formed after the first immunization last ~110 days, while GCs formed after the second immunization persist for ~165 days, and then subsequent immunizations last ~180 days. The secondary GCs are longer lasting because enhanced antigen presentation on follicular dendritic cells (FDCs) leads to increased and more sustained entry of B cells into the GCs.
Figure 5—figure supplement 3
Weaker conservation of subdominant epitopes between historical strains and the immunizing strain can outweigh more favorable germline frequencies of B cells that target these epitopes.
We vary the conservation of epitope 2 between the immunizing strain and strain 2 while fixing the fraction of naive B cells . Epitope 2 has a more favorable germline distribution than epitope 3 (Figure 4B). However, if epitope 2 is significantly less conserved between strains 1 and 2 than epitope 3 is between strains 1 and 3, the effects of weaker conservation outweigh the effects of a more favorable germline distribution, resulting in worse titers against strain 2 than against strain 3. The point at which the advantage of the immunodominance hierarchy is overcome by the effects of lower conservation depends on the germline B cell properties. The effects of epitope masking are not considered here. (A) Here, we fix the fraction of naive B cells targeting epitope i to . The conservation of epitope 2 between the immunizing strain and strain 2 is varied, and the titers after the second vaccination are shown. When is greater than the crossover point ~0.8, there are higher titers against strain 2 than strain 3. This is the same order as the order of the corresponding epitope in the immunodominance hierarchy; in particular, epitope 2 (which is conserved with strain 2) is more immunodominant than epitope 3 (which is conserved with strain 3). When is less than the crossover point ~0.8, there are higher titers against strain 3 than strain 2, which is not the same order as the immunodominance hierarchy for the corresponding epitopes. (B) When we alter the relative dominance of the epitopes by changing the fraction of naive B cells targeting each epitope to , the crossover point occurs at a lower conservation value .
Figure 5—figure supplement 4
Increasing selection stringency K (from 0.5 to 0.7) does not change the qualitative results.
(A) The antibody titers increase for both vaccine and non-vaccine strains over four immunizations, although higher stringency results in slightly lower titers against non-vaccine strains after the second vaccination. (B) Memory cells produced in the germinal centers (GCs) of the first immunization primarily target the dominant epitope (ep 1). The response to subdominant epitopes (ep 2 and 3) is muted for the higher stringency case. However, sufficient memory cells against the vaccine strain are generated to produce high-affinity antibodies to bind and present antigen on the follicular dendritic cell (FDC) in the second immunization. (C, D) The GCs in the second and third immunizations produce high-affinity memory cells that target subdominant epitopes, even after increasing stringency.
Figure 6
Regulation of antibody broadening through epitope masking.
Maximum antibody titers for the historical strains after the second vaccination, with and without epitope masking. Two cases are considered when there is epitope masking: (1) (Danecek et al., 2011) the epitopes are absolutely distinct; (2) the epitopes can overlap with each other. In the second case shown here, there is 30% overlap between epitope 1 (dominant) and epitope 2 (subdominant) and between epitope 1 (dominant) and epitope 3 (subdominant). Masking increases the titers against historical strains, even when there is some overlap between the dominant and subdominant epitopes. (B) Maximum antibody titers for the historical strains after the third vaccination, with and without masking. After the third vaccination, titers for Variant 2 with epitope masking are higher when there is epitope overlap than when the epitopes are distinct. (C) Relative number of memory cells produced with epitope masking. Epitopes are considered to be fully distinct. The epitope that is most targeted by the memory cells is also masked the most after the subsequent immunization. The dominant epitope is targeted most by Vax 1 memory cells and is masked the most in the second immunization. The orange subdominant epitope (epitope 2) and green subdominant epitope (epitope 3) are both relatively well targeted by Vax 2 memory cells. However, the subdominant epitopes are also masked during the third immunization, so the subdominant epitopes lose their advantage compared to the dominant epitope in the affinity maturation process after Vax 3. (D) Relative number of memory cells produced with epitope masking and overlap. The epitope that is most targeted by the memory cells is masked the most in the subsequent immunization. The dominant epitope is targeted most by Vax 1 memory cells and is masked the most in the second immunization. The orange subdominant epitope (epitope 2) is targeted most by Vax 2 memory cells, although more memory cells target the dominant epitope than when the epitopes are fully distinct. Due to the masking of epitope 2 in the third immunization, the dominant and green subdominant epitope (epitope 3) are both relatively well targeted by Vax 3 memory cells.
Tables
Appendix 1—table 1
HA sequence information for the influenza stains in this study.
| Strain | Genbank | Type | Year | Location | GISAID |
|---|---|---|---|---|---|
| H1N1 A_Beijing_262_1995 | AB304819.1 | H1N1 | 1995 | Beijing | |
| H1N1 A_Brazil_11_1978 | HQ008267.1 | H1N1 | 1978 | Brazil | |
| H1N1 A_Brisbane_59_2007 | JN899402.1 | H1N1 | 2007 | Brisbane | |
| H3N2 A_Brisbane_10_2007 | KM978061.1 | H3N2 | 2007 | Brisbane | |
| B_Brisbane_60_2008 | FJ766842.1 | B | 2008 | Brisbane | |
| H1N1 A_California_07_2009 | NC_026433.1 | H1N1 | 2009 | California | |
| H1N1 A_California_10_1978 | CY021717.1 | H1N1 | 1978 | California | |
| H1N1 A_Chile_1_1983 | CY121261.1 | H1N1 | 1983 | Chile | |
| B_Colorado_06_2017 | CY236607.1 | B | 2017 | Colorado | |
| H1N1 A_Denver_1957 | CY146793.1 | H1N1 | 1957 | Denver | |
| B_Florida_4_2006 | EU515992.1 | B | 2006 | Florida | |
| H1N1 A_Fort_Monmouth_1_1947 | AF494250.1 | H1N1 | 1947 | Fort_Monmouth | |
| H3N2 A_Fujian_411_2002 | EU501153.1 | H3N2 | 2002 | Fujian | |
| B_Harbin_7_1994 | CY040441.1 | B | 1994 | Harbin | |
| B_Hong_Kong_330_2001 | AF532549.1 | B | 2001 | Hong_Kong | |
| H3N2 A_Hong_Kong_1_1968 | AF348177.1 | H3N2 | 1968 | Hong_Kong | |
| H3N2 A_Hong_Kong_4801_2014 | H3N2 | 2014 | Hong_Kong | EPI1026711 | |
| H3N2 A_Kentucky_UR07-0028_2008 | CY037791.1 | H3N2 | 2008 | Kentucky_UR07-0028 | |
| B_Lee_1940 | K00423.1 | B | 1940 | Lee | |
| B_Malaysia_2506_2004 | EU124275.1 | B | 2004 | Malaysia | |
| B_Massachusetts_2_2012 | MT056027.1 | B | 2012 | Massachusetts | |
| H1N1 A_Michigan_45_2015 | KY117023.1 | H1N1 | 2015 | Michigan | |
| H3N2 A_Mississippi_1_1985 | L19003.1 | H3N2 | 1985 | Mississippi | |
| H3N2 A_Nanchang_933_1995 | CY108293.1 | H3N2 | 1995 | Nanchang | |
| H1N1 A_New_Caledonia_29_1999 | DQ508857.1 | H1N1 | 1999 | New_Caledonia | |
| H1N1 A_New_Jersey_1976 | CY147422.1 | H1N1 | 1976 | New_Jersey | |
| H3N2 A_New_York_55_2004 | KM821338.1 | H3N2 | 2004 | New_York | |
| H3N2 A_Panama_2007_1999 | EF626612.1 | H3N2 | 1999 | Panama | |
| H3N2 A_Perth_16_2009 | GQ293081.1 | H3N2 | 2009 | Perth | |
| B_Phuket_3073_2013 | B | 2013 | Phuket | EPI2195537 | |
| H3N2 A_Port_Chalmers_12_1973 | CY113109.1 | H3N2 | 1973 | Port_Chalmers | |
| H1N1 A_Puerto_Rico_8_1934 | EF467821.1 | H1N1 | 1934 | Puerto_Rico | |
| H3N2 A_Shangdong_9_1993 | Z46417.1 | H3N2 | 1993 | Shangdong | |
| B_Sichuan_379_1999 | EF566113.1 | B | 1999 | Sichuan | |
| H3N2 A_Sichuan_30_1989 | CY108211.1 | H3N2 | 1989 | Sichuan | |
| H1N1 A_Singapore_6_1986 | D00406.1 | H1N1 | 1986 | Singapore | |
| H1N1 A_Solomon_Islands_03_2006 | EU100724.1 | H1N1 | 2006 | Solomon_Islands | |
| H1N1 A_South_Carolina_1_1918 | AF117241.1 | H1N1 | 1918 | South_Carolina | |
| H3N2 A_Switzerland_9715293_2013 | H3N2 | 2013 | Switzerland | EPI814528 | |
| H3N2 A_Sydney_5_1997 | KM821316.1 | H3N2 | 1997 | Sydney | |
| H1N1 A_Texas_36_1991 | DQ508889.1 | H1N1 | 1991 | Texas | |
| H3N2 A_Texas_1_1977 | EF626623.1 | H3N2 | 1977 | Texas | |
| H3N2 A_Texas_50_2012 | KC892952.1 | H3N2 | 2012 | Texas | |
| B_Texas_06_2011 | KC813979.1 | B | 2011 | Texas | |
| H1N1 A_USSR_90_1977 | HQ008265.1 | H1N1 | 1977 | USSR | |
| H3N2 A_Victoria_361_2011 | KM821347.1 | H3N2 | 2011 | Victoria | |
| H1N1 A_Weiss_1943 | CY147366.1 | H1N1 | 1943 | Weiss | |
| H3N2 A_Wisconsin_67_2005 | CY163704.1 | H3N2 | 2005 | Wisconsin | |
| B_Wisconsin_1_2010 | CY115183.1 | B | 2010 | Wisconsin | |
| B_Yamagata_16_1988 | M36105.1 | B | 1988 | Yamagata |
Appendix 1—table 2
Simulation parameters.
Highlighted parameters were modified from the original model.
| Parameter | Value | Description | Note |
|---|---|---|---|
| Antigen and antibody dynamics | |||
| nM day–1 PC–1 | Rate of antibody production per plasma cell per day | Picked to match antibody titers at peak response to second SARS-CoV-2 vaccination (Goel et al., 2021; Muecksch et al., 2022) | |
| 0.025 day–1 | Antibody decay rate | Picked to give antibody a half-life of ~28 days (Goel et al., 2021) | |
| 3 day–1 | Antigen decay rate | Picked such that antigen decays rapidly in the first few days after vaccination (Aung et al., 2023; Bhagchandani et al., 2024; Martin et al., 2021; Tam et al., 2016) | |
| 1 hour–1 | Rate of immune complex transport to FDC | Picked such that antigen is rapidly transported to the FDC within ~48 hours after vaccination (Aung et al., 2023; Bhagchandani et al., 2024; Martin et al., 2021) | |
| 0.15 day–1 | Rate of decay of immune complex on FDC | Picked such that the secondary GCs last at least 3 months after SARS-CoV-2 vaccination as observed (Gaebler et al., 2021; Muecksch et al., 2022) | |
| 10 nM | Initial conditions | Picked within reasonable physiological ranges (Martin et al., 2021) | |
| 10–2 nM | Picked within reasonable physiological ranges (Demonbreun et al., 2021) | ||
| 0 nM | No initial IC exists | ||
| B cell affinities | |||
| 2000 cells/GC | Number of naïve B cells per GC | Based on the total number of naïve B cells (Boyd and Joshi, 2014; Rees, 2020) and frequency of naïve B cells that target the SARS-CoV-2 RBD (Feldman et al., 2021), divided across 200 GCs | |
| 0.8 | Fraction of naïve B cells that target epitope 1 | Tested for robustness in previous simulations (Yang et al., 2023). | |
| varied: 0.15, 0.18 | Fraction of naïve B cells that target epitope 2 | Varied in simulations | |
| varied: 0.05, 0.02 | Fraction of naïve B cells that target epitope 3 | Varied in simulations | |
| 7 | Parameter for the germline affinity distribution of epitope 1. Affinity at which there is on average one naïve B cell targeting epitope 1 available for each GC | Tested for robustness in previous simulations (Yang et al., 2023) | |
| 0.4 | Parameter for the germline affinity distribution of epitope 2. is the affinity at which there is on average one naïve B cell targeting epitope 2 available for each GC | Tested for robustness in previous simulations (Yang et al., 2023) | |
| 0.8 | Parameter for the germline affinity distribution of epitope 3. is the affinity at which there is on average one naïve B cell targeting epitope 2 available for each GC | Picked such that the germline affinity distribution for epitope 3 has a shorter high-affinity tail than epitope 2, i.e. epitope 3 is less immunodominant than epitope 2 | |
| 80 | Length of string representation of B cell residues | Upper range of sum of CDR lengths in light and heavy chain (Nowak et al., 2016) | |
| 3.1, 1.2, 3.08 | Parameters for shifted log-normal distribution that models the effect of affinity-changing mutations | Based on empirical distribution of affinity changes due to point-mutations in proteins (Zhang and Shakhnovich, 2010) | |
| epitope 1: 0.4 epitope 2: typically 0.95, varied: 0.6–0.95 epitope 3: 0.4 | Level of conservation between strain 1 and 2 | Picked such that a large portion of affinity-increasing mutations (~30 to 72%) in the conserved epitope (epitope 2) are beneficial for both strain 1 and 2;~19% for the other epitopes | |
| epitope 1: 0.4 epitope 2: 0.4 epitope 3: 0.95 | Level of conservation between strain 1 and 3 | Picked such that ~72% of affinity-increasing mutations in the conserved epitope (epitope 3) are beneficial for both strain 1 and 3;~19% for the other epitopes | |
| 0 without overlap 0.3 with overlap | Fraction of antibodies that target epitope 1 that can mask epitope 2 and vice versa (i.e. the amount of spatial overlap between epitope 1 and 2) | Varied in simulations; effects are also studied in previous simulations (Yang et al., 2023) | |
| 0 without overlap 0.3 with overlap | Fraction of antibodies that target epitope 1 that can mask epitope 3 and vice versa (i.e. the amount of spatial overlap between epitope 1 and 3) | Varied in simulations; effects are also studied in previous simulations (Yang et al., 2023) | |
| 0 | Fraction of antibodies that target epitope 2 that can mask epitope 3 and vice versa (i.e. the amount of spatial overlap between epitope 2 and 3) | Picked such that the subdominant epitopes (epitopes 2 and 3) are distinct | |
| GC and EGC dynamics | |||
| nM | Reference antigen concentration | Tested for robustness in previous simulations (Yang et al., 2023) | |
| 6 | Reference binding affinity | Typical threshold for naïve B cell activation is (Batista and Neuberger, 1998) | |
| typically 0.5 varied: 0.5, 0.7 | Stringency of selection of naïve and GC B cells by helper T cells based on amount of captured antigen | Varied in simulations; effects are also studied in previous simulations (Yang et al., 2023) | |
| 10 day–1 | Approximately the maximum number of naïve B cells that can enter the GC per day | Based on experimental observation in mice for the number of GC B cells after 7 days (Tas et al., 2016) | |
| 2.5 day–1 | Maximum rate of positive selection for GC and EGC B cells | Maximum proliferation is about ~4 times / day (Tas et al., 2016) | |
| 0.5 day–1 | Death rate of GC B cells | Picked to allow B cells to survive for ~2 days before death if they are not selected | |
| 1200 | Maximum number of helper T cells involved in positive selection in GC and EGC | Picked such that GCs have a peak size of ~1,000 B cells for computational tractibility | |
| 14 days | Time at which number of helper T cells is maximal | Matches dynamics of T cell response to SARS-CoV-2 vaccination (Goel et al., 2021) | |
| 0.01 day–1 | Death rate of helper T cells after time | Matches dynamics of T cell response to SARS-CoV-2 vaccination (Goel et al., 2021) | |
| Plasma and memory cell dynamics | |||
| 0.05 | Probability that a positively selected GC B cell exits and differentiates | Tested for robustness in previous simulations (Yang et al., 2023) | |
| 0.1 | Probability that a differentiating GC B cell becomes a plasma cell | Tested for robustness in previous simulations (Yang et al., 2023) | |
| 0.6 | Probability that a proliferating memory cell in EGC differentiates into a plasma cell | Based on observation in mice that ~60% of reactivated B memory cells differentiate into short-lived plasma cells (Victora et al., 2010) | |
| 0.17 day–1 | Death rate of plasma cells | Short-lived plasma cells have a half-life of ~4 days (Moran et al., 2018) | |
Additional files
-
MDAR checklist
- https://cdn.elifesciences.org/articles/107042/elife-107042-mdarchecklist1-v1.docx
-
Source data 1
Hemagglutination inhibition (HAI) values for the influenza virus strains, measured across longitudinal vaccine study (2013–2016) for n = 136 de-identified subjects >50 years of age and <38 years of age.
- https://cdn.elifesciences.org/articles/107042/elife-107042-data1-v1.xlsx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Repeated vaccination with homologous influenza hemagglutinin broadens human antibody responses to unmatched flu viruses
eLife 14:e107042.
https://doi.org/10.7554/eLife.107042
