Induction of persistent protective immune responses is a key attribute of a successful vaccine formulation. MF59 adjuvant, an oil-in-water emulsion used in human vaccines, is known to induce persistent high-affinity functional Ab titers and memory B cells, but how it really shapes the Ag-specific B cell compartment is poorly documented. In this study, we characterized the Ab- and Ag-specific B cell compartment in wild-type mice immunized with HlaH35L, a Staphylococcus aureus Ag known to induce measurable functional Ab responses, formulated with MF59 or aluminum salts, focusing on germinal centers (GC) in secondary lymphoid organs. Taking advantage of single-cell flow cytometry analyses, HlaH35L-specific B cells were characterized for the expression of CD38 and GL-7, markers of memory and GC, respectively, and for CD80 and CD73 activation markers. We demonstrated that immunization with MF59-, but not aluminum salt–adjuvanted HlaH35L, induced expanded Ag-specific CD73+CD80 GC B cells in proximal- and distal-draining lymph nodes, and promoted the persistence of GC B cells, detected up to 4 mo after immunization. In addition to increasing GC B cells, MF59-adjuvanted HlaH35L also increased the frequency of T follicular helper cells. This work extends previous knowledge regarding adaptive immune responses to MF59-adjuvanted vaccines, and, to our knowledge, for the first time an adjuvant used in human licensed products is shown to promote strong and persistent Ag-specific GC responses that might benefit the rational design of new vaccination strategies.

Effective vaccines should generate specific and strong immune responses against target Ags with acceptable levels of inflammation induced by the immunization. In the past, live-attenuated or killed whole organisms were used, often causing side effects not acceptable any longer by a modern society. The next generation of vaccines mostly relies on purified subunit Ags that may be weakly immunogenic on their own and require adjuvants to enhance the strength and duration of the immune responses they induce. MF59 oil-in-water emulsion and aluminum salts (alum), including aluminum hydroxide and aluminum phosphate generally described as alum, are the two major adjuvants licensed in human vaccines, and have a long track record of safety and efficacy in humans.

Alum is empirically used in human vaccines since 1920 with the general goal to enhance Ab production against weak immunogens (1, 2). MF59 is a nanoemulsion originally licensed in 1997 to adjuvant human influenza vaccines; it consists of the naturally occurring squalene oil and of nonionic surfactants Tween 80 and Span 85 emulsified in uniform particles of ∼160 nm in size. In the past years, >150 million doses of MF59-adjuvanted vaccines have been administered worldwide, and a vast body of knowledge on MF59 immunogenicity has been collected from human clinical studies, in particular with influenza vaccines. MF59 adjuvant enhances both the quantity and the quality of Ag-specific immune responses, allowing for Ag dose sparing, improving the breadth of the Ab responses, and enhancing protective immunity, in particular target populations such as children and the elderly (35).

MF59 and alum were initially thought to exert their adjuvant effect by the formation of an Ag depot, but subsequent studies refuted this hypothesis for both adjuvants (68), revealing two separate mechanisms of action. Ag adsorption on alum directly delivers the Ag to resident dendritic cells (DC) and seems critical for alum adjuvanticity (9). The mechanisms by which the Ag uptake is facilitated are not yet clear, but a recent study suggested that this may occur in the absence of uptake of alum by APCs. In particular, alum was shown to bind lipids on the surface of APCs and trigger cellular activation cascades leading to initiation of an immune response, but without itself being internalized by the cells (9), suggesting an indirect role in delivering Ag into the Ag-processing pathway; however, other studies have shown that alum is internalized by APCs, resulting in endosomal destabilization (10). In contrast, MF59 primary mechanism of action resides in the generation of an immunocompetent environment at the injection site (muscle), characterized by the release of local extracellular ATP, expression of activation makers on resident DC, and secretion of chemokines that strongly increase phagocytic cells trafficking toward the injection site and the draining lymph nodes (LNs) (1115). Unlike alum, MF59 does not directly activate DC, but rather promotes cell migration, a more efficient Ag uptake, and more Ag-loaded APCs in draining LNs compared with alum or nonadjuvanted vaccines (12, 13, 15). Furthermore, the adjuvant effects of MF59 were maintained even when MF59 and the Ag were administered 24 h apart at the same site (16).

Although MF59 also induces strong and persistent T cell responses (17, 18), it is better known for the powerful adjuvant effect it exerts on vaccine-specific Ab and B cell responses. In particular, MF59 promotes high-affinity Abs, increased diversity of the epitope repertoire, and enhanced frequencies of Ag-specific memory B cells, resulting in higher frequencies of responding memory B cells and higher titers of protective neutralizing Abs after Ag re-exposure (1921). Remarkably, higher immune responses were observed with MF59 in younger (or Ag-naive) individuals compared with nonadjuvanted vaccine, suggesting that MF59 plays a major role in the activation and differentiation of Ag-specific naive B cells and priming of immune responses (5, 20).

Altogether, these findings suggest that MF59 strongly affects T-dependent B cell responses, in particular, shaping germinal centers (GC) where expansion, selection, and affinity maturation of Ag-specific B cells take place during primary responses. GC develop within secondary lymphoid organs from one or few B cell clones, called Ag-specific GC B cells that are defined by a switched IgMIgD phenotype, downregulation of CD38, and upregulation of GL-7 expression. Under the pressure of T follicular helper (TFH) cells and follicular DC, GC B cells extensively proliferate and increase their affinity for the Ag. Subsequently, GC B cells differentiate in the following: 1) long-lived bone marrow plasma cells that constantly feed the blood circulation with high-affinity Abs, and 2) high-affinity memory B cells that quickly respond to Ag recall in case of re-exposure.

Detoxified Staphylococcus aureus α-hemolysin (HlaH35L) was used as a model Ag to study the impact of MF59 on functional B cell responses. HlaH35L is a promising vaccine candidate against S. aureus infections (22, 23), and anti-Hla Abs are linked to reduced risk of severe diseases in humans (24). Moreover, the importance of protective neutralizing Abs against Hla has been demonstrated in several animal models of S. aureus infection (22, 23, 25).

In this study, the magnitude and quality of Ab responses were evaluated in three different vaccination schedules, as follows: 2 wk, 1 mo, and 4 mo between first and second dose. The 1-mo schedule was chosen to further characterize Hla-specific B cells by ELISPOT and flow cytometry. Frequencies of Ag-specific B cells were determined in PBMC, spleen, and LNs of immunized mice up to 4 mo after the last immunization. HlaH35L plus MF59 induced higher Hla-specific functional Ab titers compared with HlaH35L plus alum, regardless of the immunization schedule. Although both adjuvant formulations induced Hla-specific GC B cells in proximal-draining LNs, only HlaH35L plus MF59 induced GC B cells in axillary distal-draining LNs. Furthermore, HlaH35L plus MF59 increased Ag-specific GC persistence and B cell memory responses up to 4 mo after immunization. Finally, increased frequencies of TFH cells were also measured in HlaH35L plus MF59–immunized mice compared with mice in the unadjuvanted HlaH35L and HlaH35L plus alum groups.

Pathogen-free BALB/c female mice (Charles River) aged 6 wk were used in this study. All animal studies were carried out in compliance with current Italian legislation on the care and use of animals in experimentation (Legislative Decree 116/92) and with the Novartis Animal Welfare Policy and Standards. Protocols were approved by the Italian Ministry of Health (authorization 249/2011-B) and by the local Novartis Animal Welfare Body (authorization AWB 201106).

HlaH35L was amplified by PCR from S. aureus NCTC8325 strain and cloned as tagless constructs. MF59, a Novartis Vaccines proprietary oil-in-water emulsion consisting of 4.3% (v/v) squalene, 0.5% Tween 80, and 0.5% Span 85 in citrate buffer (10 mM), was prepared as described earlier (15). Alum [Al(OH)3] was from Novartis Vaccines.

Vaccine formulations were prepared by incubating 10 μg HlaH35L with 2 mg/ml alum or 50% MF59. All of the formulations, including the groups saline and the nonadjuvanted HlaH35L, were adjusted with PBS and NaCl to reach final pH 6.5–7 and osmolality 0.308 ± 0.060 Osm/Kg. Ag adsorption to alum and protein integrity were analyzed by SDS-PAGE (data not shown). BALB/c mice were immunized i.m. with 50 μl vaccine injected in each hind leg quadriceps (100 μl/mouse). The immunization schemes are reported in Figs. 1A and 6A. Briefly, BALB/c mice (at least five to six mice per group per experiment) were immunized twice, 2 wk, 1 mo, or 4 mo apart, and analyses of the immune response were performed 10 d or 4 mo after the second immunization.

FIGURE 1.

HlaH35L-specific functional Ab titers. BALB/c mice (n = 8–10) were immunized twice, 2 wk, 1 mo, or 4 mo apart with HlaH35L, HlaH35L plus MF59, and HlaH35L plus alum. Humoral responses were measured in sera 10 d after the last immunization. (A) Treatment schedules. (B) Hla-specific IgG titers measured by ELISA. (C) Avidity of Hla-specific IgG expressed as the ratio between the LogEC50 values obtained by ELISA with and without thiocyanate treatments; high ratio indicates sera with Hla-specific high-avidity Abs. (D) Hla-neutralizing titers. (E and F) Correlation of Hla-neutralizing titers from HlaH35L plus MF59 (black dots) and HlaH35L plus alum (gray dots) groups with (E) Hla-specific IgG titers, or (F) IgG avidity ratio; linear regression curves are shown. Data show the merge of two independent experiments. Nonadjuvanted HlaH35L group reported only in (B). In (B)–(D), two-way ANOVA followed by Bonferroni’s posttest were used for statistical analyses: *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 1.

HlaH35L-specific functional Ab titers. BALB/c mice (n = 8–10) were immunized twice, 2 wk, 1 mo, or 4 mo apart with HlaH35L, HlaH35L plus MF59, and HlaH35L plus alum. Humoral responses were measured in sera 10 d after the last immunization. (A) Treatment schedules. (B) Hla-specific IgG titers measured by ELISA. (C) Avidity of Hla-specific IgG expressed as the ratio between the LogEC50 values obtained by ELISA with and without thiocyanate treatments; high ratio indicates sera with Hla-specific high-avidity Abs. (D) Hla-neutralizing titers. (E and F) Correlation of Hla-neutralizing titers from HlaH35L plus MF59 (black dots) and HlaH35L plus alum (gray dots) groups with (E) Hla-specific IgG titers, or (F) IgG avidity ratio; linear regression curves are shown. Data show the merge of two independent experiments. Nonadjuvanted HlaH35L group reported only in (B). In (B)–(D), two-way ANOVA followed by Bonferroni’s posttest were used for statistical analyses: *p < 0.05, **p < 0.01, ***p < 0.001.

Close modal
FIGURE 6.

Long-lasting Hla-specific functional humoral and B cell responses. BALB/c mice (n = 10) were immunized twice, 1 mo apart, with HlaH35L, HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). Humoral and B cell responses were measured 4 mo after the last immunization. (A) Treatment schedules; (B) Hla-specific IgG titers measured by ELISA; (C) avidity of Hla-specific IgG expressed as the ratio between the LogEC50 values obtained by ELISA with and without thiocyanate treatments; (D) Hla-neutralizing titers; (E) number of Hla-specific bone marrow plasma cells measured by ELISPOT; (F) frequencies of Hla-specific memory B cells in spleen measured by ELISPOT. Nonadjuvanted HlaH35L group reported only in (B). Data show the merge of two independent experiments; each dot in (E) and (F) represents one mouse. Unpaired Student t test (C and D) and one-way ANOVA followed by Tukey’s posttest (B, E, and F) were used for statistical analyses: *p < 0.05, **p < 0.01.

FIGURE 6.

Long-lasting Hla-specific functional humoral and B cell responses. BALB/c mice (n = 10) were immunized twice, 1 mo apart, with HlaH35L, HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). Humoral and B cell responses were measured 4 mo after the last immunization. (A) Treatment schedules; (B) Hla-specific IgG titers measured by ELISA; (C) avidity of Hla-specific IgG expressed as the ratio between the LogEC50 values obtained by ELISA with and without thiocyanate treatments; (D) Hla-neutralizing titers; (E) number of Hla-specific bone marrow plasma cells measured by ELISPOT; (F) frequencies of Hla-specific memory B cells in spleen measured by ELISPOT. Nonadjuvanted HlaH35L group reported only in (B). Data show the merge of two independent experiments; each dot in (E) and (F) represents one mouse. Unpaired Student t test (C and D) and one-way ANOVA followed by Tukey’s posttest (B, E, and F) were used for statistical analyses: *p < 0.05, **p < 0.01.

Close modal

Amine labeling of HlaH35L with Alexa Fluor 488 (A488) succinimidyl ester was performed according to the manufacturer’s protocol (Molecular Probes, Life Technologies). Briefly, 1 mg protein was incubated with A488 (ndye/nantigen = 10, where n is number of moles) for 1 h at room temperature in the presence of 100 mM NaHCO3. Unlabeled fluorochrome was removed using 10-kDa Amicon ultra-0.5 ml centrifugal filters (Millipore). The degree of labeling was 2.97 and calculated as follows: Amax × MW/[protein] × εdye, where Amax is the absorbance of the protein–fluorochrome conjugate (at 495 nm for A488), MW is the m.w. of the protein (Da), εdye is the extinction coefficient of the dye at its maximum absorbance (εA488: 71,000 cm−1 M−1), and the protein concentration is expressed in mg/ml. Protein concentration after labeling was determined by NanoDrop (Thermo Scientific), and its integrity was assessed by SDS-PAGE and size exclusion ultra pressure liquid chromatography (data not shown).

For the characterization of Ag-specific B cell responses, cell suspensions prepared from spleen, or pairs of axillary and inguinal LNs from each mouse, were stained with Yellow live/dead cell stain (Invitrogen) for 20 min at room temperature and then incubated with Fc block (BD Biosciences) in PBS plus 1% FBS (HyClone, Thermo Scientific) for 10 min at 4°C. Approximately 107 splenocytes or 2–3 × 106 LN cells were stained for 30 min at 4°C with the following mAbs: anti-CD19 allophycocyanin-H7 (clone 1D3), anti-IgM BV421 (clone R6-60.2), anti-IgD PE (clone 11-26c.2a), anti–GL-7 Alexa Fluor 647 (clone GL-7), and anti-CD80 HPE594 (clone 16-10A1) (all from BD Biosciences); anti-CD3 BV785 (clone 17A2; BioLegend); and anti-CD38 PE-Cy5 (clone 90) and anti-CD73 PE-Cy7 (clone TY/11.8) (eBioscience).

To identify Ag-specific B cells, samples were stained with 0.1 μg A488-labeled HlaH35L per 2–3 × 106 cells. The optimal amount of fluorescently labeled protein to be used was previously determined in titration experiments on cells from naive and immunized mice. Cells were analyzed on a special-order FACS LSRII or sorted using a FACS Aria II (BD Biosciences).

For the characterization of TFH cell responses, inguinal LN cell suspensions were stained with Near-Infra Red live/dead cell stain (Invitrogen) for 20 min at room temperature and then incubated with Fc block (BD Biosciences) in PBS plus 1% FBS (HyClone, Thermo Scientific) for 10 min at 4°C. Approximately 2 × 106 LN cells were stained for 30 min at 4°C with the following mAbs: anti-ICOS PE (clone 7E.17G9), anti–PD-1 PE-CF594 (clone J43), anti-CXCR5 PE-Cy7 (clone 2G8), anti-CD3 Brillant Violet 605 (clone 145-2C11) (all from BD Biosciences), and anti-CD4 Brillant Violet 510 (clone RM4-5; BioLegend).

Cells were analyzed on a special-order FACS LSRII or sorted using a FACS Aria II (BD Biosciences). Data analyses were performed using FlowJo software v9.6 (Tree Star) with a gating strategy to identify the Ag-specific B cells and TFH cells reported in Supplemental Figs. 2A and 4, respectively.

LNs were collected under dry conditions at the appropriate time points, immediately frozen using liquid nitrogen, and stored at −80°C until processing. Cryosections of LNs were obtained with the Leica CM1950 cryostat and stained using the following Abs: anti–CD21/35-Pacific Blue (BioLegend) to identify follicular DC; anti–GL-7–FITC (eBioscience) to identify GC B cells; and anti–IgD-PE (BD Pharmingen) to identify follicular B cells. The cryosections (8 μm thick) were cut along the entire organ to analyze all the planes of the organs. The cryosections were fixed using PBS plus 3% formaldehyde for 10 min at room temperature, washed twice with PBS, and permeabilized with PBS plus 3% BSA/1% saponin (permeabilization buffer) for 30 min at room temperature. Tissue sections were then incubated with the appropriate Abs diluted in permeabilization buffer for 1 h at room temperature in the dark. After washing three times with permeabilization buffer and once with PBS, stained tissue sections were sealed using Gold Antifade reagent (Invitrogen-Life Technologies) and a coverslip. Images were acquired with the Axio Observer LSM700 confocal microscope (Zeiss) at 20°C, using FLUAR 5× or Plan-Apochromat 40× objective lenses with 0.25 or 1.3 of numerical aperture, respectively. The 40× objective lens was used with the Zeiss Immersol 518F imaging medium. Images were processed with Zen 2008 software (Zeiss).

ELISPOT analyses of HlaH35L-specific plasma cells were performed on bone marrow cells collected by washing the pair of femoral bone marrows. Cells were plated into ELISPOT plates (MultiScreen HTS-HA MSHAN45; Millipore) previously coated with 100 μl PBS containing 10 μg/ml HlaH35L or BSA, or 5 μg/ml goat anti-mouse Ig Ab (Southern Biotech), and blocked for 2 h at room temperature with PBS containing 10% FBS. Plates were washed after 4 h of incubation with cells, than incubated with biotin-conjugated goat anti-mouse IgG (1:2000; BD Pharmingen), followed by HRP-conjugated streptavidin (1:1000; Endogen), and developed with the 3-amino-9-ethylcarbazole substrate kit (Sigma-Aldrich). Spots corresponding to Ab-secreting cells were counted using the UV Spot ELISPOT plate Analyzer and the Immunospot software v5.1 Professional DC (Cellular Technology Limited). To determine the frequencies of HlaH35L-specific memory B cell, freshly collected splenocytes were harvested and cultured in vitro in complete medium with 5 μg/ml CpG (Primm) plus 1000 U/ml IL-2 (Novartis); after 5 d at 37°C 5% CO2, cells were collected, washed with medium plus 2.5 mM EDTA, and plated for ELISPOT analyses, as described above.

ELISPOT assays were also used to confirm the specificity of A488-HlaH35L+ B cells sorted by flow cytometry (Supplemental Fig. 2B). In particular, splenocytes from immune mice were stained with A488-HlaH35L, anti-CD3, anti-CD19, anti-IgM, and anti-IgD. HlaH35L-specific CD19+IgMIgD as well as CD19 splenocytes were sorted from immune mice, mixed at a ratio of 0.5:100, and cultured in vitro at 3 × 105 cells in 200 μl complete medium with 5 mg/ml CpG (Primm) and 1000 U/m IL-2. After 5 d, cells were harvested and plated for ELISPOT, as described above.

Titration of HlaH35L-specific serum IgG was performed on individual serum samples with ELISA plates (Nunc MaxiSorp, Thermo Scientific) coated overnight with HlaH35L (10 μg/ml). Plates were blocked with 2% BSA in PBS/Tween 20 0.05% for 1 h at 37°C, and then washed three times with 0.05% Tween 20 in PBS and incubated for 2 h at 37°C with individual mouse sera in 3-fold serial dilutions. Plates were washed, incubated for 2 h at 37°C with secondary anti-mouse total IgG (1:2000; Sigma-Aldrich) and washed, and p-nitrophenyl phosphate disodium (Sigma-Aldrich) was added in diethanolamine buffer at a concentration of 1 mg/ml. Color reaction was measured with SpectraMax (Molecular Devices) reader by determining the OD at 405 nm. A sigmoidal binding curve obtained from plotting the OD450 versus the relative dilution was used to calculate the LogEC50 for each serum. Avidity was measured as the overall strength of binding between Ab and Ag by ELISA, in which plates were incubated with ammonium thiocyanate 1 M for 15 min before the addition of the secondary Ab; avidity is expressed as the ratio between LogEC50 (+thiocyanate) and (−thiocyanate) treated sera, respectively.

Sera collected 10 d and 4 mo after immunization were characterized for their ability to inhibit Hla-induced hemolysis in an in vitro assay. Two-fold serial dilutions of sera were incubated with 12.5 nM Hla for 30 min at 37°C in agitation (350 rpm). Erythrocytes from defibrinated rabbit blood were added, and the incubation was prolonged for an additional 30 min at 37°C. Plates were then centrifuged for 5 min at 1000 × g, and the OD (OD540) of supernatants was determined with a SpectraMax 340PC384 Absorbance Microplate Reader (Molecular Devices). OD540 of supernatants derived from samples incubated with Hla without serum was considered as 100% hemolysis. Incubation with 1% Triton X-100 in water was used as positive control (maximal hemolysis), whereas incubation without Hla was the negative control. Percent hemolysis was calculated using as denominator the OD540 obtained following lysis of erythrocytes in absence of serum.

Data are presented as the mean ± SEM. Statistical analyses were performed using GraphPad Prism 5 software (GraphPad Software). Unpaired Student t test or one-way ANOVA statistical tests followed by Tukey’s or Bonferroni’s posttests were applied as reported in the legend of each figure. Hla-neutralizing titers were calculated using a log(agonist) versus normalized response nonlinear regression curve with Excel XLfit (Microsoft).

Hla-specific humoral responses were characterized in mice immunized with MF59- or alum-formulated HlaH35L. BALB/c mice were immunized twice, i.m., with 10 μg HlaH35L alone, HlaH35L plus MF59, or HlaH35L plus alum. Three different vaccination schedules, 2-wk, 1-mo, or 4-mo interval between first and second dose, were compared (Fig. 1A). Ten days after the second immunization, Hla-specific total IgG titers, Ab avidity, and Hla-neutralizing titers were determined (Fig. 1B–D). Both alum and MF59 formulations increased Hla-specific IgG titers compared with the unadjuvanted HlaH35L; furthermore, HlaH35L plus MF59 induced statistically significant higher Hla-specific IgG titers compared with HlaH35L plus alum (p < 0.001) (Fig. 1B) in all three schedules of immunization tested. Hla-specific titers in the MF59 group were higher in the 1-mo interval schedule, compared with the shorter 2-wk boosting interval (5.51 ± 0.05 LogEC50 and 4.98 ± 0.08 LogEC50, respectively; p < 0.001); no statistical differences were observed between the 1- and 4-mo boosting intervals. Avidity of Hla-specific IgG was further measured with a modified ELISA comprising a washing step with ammonium thiocyanate (NH4SCN) before the addition of the secondary Ab. IgG avidity was expressed as the ratio between LogEC50 (+thiocyanate) and LogEC50 (−thiocyanate) values. Hla-specific IgG with statistically significant higher avidity were induced by HlaH35L plus MF59 compared with HlaH35L plus alum in the 1-mo interval schedule (0.884 and 0.782, respectively; p < 0.05), with a positive trend to higher avidity Abs induced by the MF59 formulation also in the 2-wk and 4-mo schedules (Fig. 1C). Ab avidity increased with the increase of time between the two immunizations, with the 4-mo boosting interval inducing Hla-specific IgG with the highest avidity. Finally, we characterized the functionality of these Abs by measuring their ability to neutralize the activity of the Hla toxin. HlaH35L plus MF59 outperformed HlaH35L plus alum in inducing functional Hla-neutralizing Abs (Fig. 1D), and neutralizing titers were positively correlated with total IgG titers and avidity (Fig. 1E, 1F). The enhancement in serum-neutralizing activity observed in the MF59 group may reflect a greater magnitude of the Ab responses (Fig. 1B) or an intrinsic increase in Ab affinity. Based on the serological data, the 1-mo immunization schedule was chosen for further characterization of B cells at the single-cell level.

We first determined the number of Hla-specific plasma cells and the frequencies of Hla-specific memory B cells recovered in immunized mice from femoral bone marrows and spleens, respectively. HlaH35L plus MF59 induced the highest number of Ag-specific plasma cells (98.5 ± 14.1 per 106 cells; p < 0.01) and frequency of Hla-specific memory B cells (2.12 ± 0.41% of total IgG; p < 0.05) in the 1-mo interval schedule, as measured by ELISPOT (Fig. 2). In particular, there was a 2.23-fold increase in the number of Ag-specific bone marrow plasma cells and a 2.27-fold increase in the frequencies of Ag-specific memory B cells in the HlaH35L plus MF59 versus HlaH35L plus alum groups, hence confirming the enhanced adjuvant effect of MF59 on B cell responses. Higher responses in the HlaH35L plus MF59 group still persisted 1 mo postimmunization (Supplemental Fig. 1A, 1B).

FIGURE 2.

B cell responses to adjuvanted HlaH35L. BALB/c mice (n = 6–10) were immunized twice, 1 mo apart, with HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). B cell responses were measured 10 d after the last immunization by ELISPOT. (A) Number of bone marrow Hla-specific plasma cells. (B) Frequencies of Hla-specific memory B cells in spleen. Each dot represents one mouse, and data show the merge of two independent experiments. One-way ANOVA followed by Tukey’s posttest were used for statistical analyses: *p < 0.05, **p < 0.01.

FIGURE 2.

B cell responses to adjuvanted HlaH35L. BALB/c mice (n = 6–10) were immunized twice, 1 mo apart, with HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). B cell responses were measured 10 d after the last immunization by ELISPOT. (A) Number of bone marrow Hla-specific plasma cells. (B) Frequencies of Hla-specific memory B cells in spleen. Each dot represents one mouse, and data show the merge of two independent experiments. One-way ANOVA followed by Tukey’s posttest were used for statistical analyses: *p < 0.05, **p < 0.01.

Close modal

To further characterize Hla-specific B cells, HlaH35L was conjugated to A488 and used to stain Hla-specific BCR-bearing memory B cells ex vivo for flow cytometry analyses. Switched CD19+IgMIgDA488+ Hla-specific B cells were detected in mice immunized with HlaH35L plus MF59, but not in mice treated with MF59 alone (Supplemental Fig. 2A), and Hla-staining specificity was confirmed on sorted cells by memory B cell ELISPOT: 16 of 19 (84%) memory B cells selected by FACS were Hla positive also on the ELISPOT assay (Supplemental Fig. 2B). Taking advantage of this new staining protocol, frequencies of switched CD19+IgMIgDA488+ Hla-specific B cells were determined in PBMC and secondary lymphoid organs from mice immunized twice, 1 mo apart, with HlaH35L, HlaH35L plus MF59, HlaH35L plus alum, and saline formulation as negative control (Supplemental Fig. 2C).

The analyses of inguinal proximal-draining LNs from mice receiving HlaH35L plus MF59 showed a markedly expanded total CD19+IgMIgD compartment 10 d after the second immunization (Fig. 3A); Hla-specific B cells represented up to 7.3% of the total IgMIgD B cell subset, with no statistical differences observed among the three groups of immunization (Fig. 3B). We also identified a population of IgM+IgD Hla-specific B cells, likely to be Ag-specific IgM+ memory B cells, that was expanded in the HlaH35L plus MF59 group (Supplemental Fig. 3A, 3B). CD19+IgMIgD Hla-specific B cells were also detected in axillary distal LNs, with the highest frequency in the HlaH35L plus MF59 group (4.47 ± 0.65%; p < 0.05). Hla-specific B cells were also expanded in spleen and PBMC from mice receiving HlaH35L plus MF59, and represented the 0.81 and 0.76% of Ig-switched B cells, respectively. Lower frequencies were observed in mice immunized with HlaH35L plus alum (in spleen, p < 0.01) or HlaH35L alone, and this trend was in agreement with our previous observations by ELISPOT analyses in the same experimental settings.

FIGURE 3.

Frequencies of CD19+IgMIgD B cells. BALB/c mice were immunized twice, 1 mo apart, with HlaH35L, HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). B cell responses were measured 10 d after the last immunization by flow cytometry. (A) Frequencies of IgMIgD cells in the CD19+CD3 B cell compartment in inguinal proximal-, axillary distal-draining LN, spleen, and PBMC, as reported on top of each graph. (B) Frequencies of Hla-specific B cells gated in the CD19+IgMIgD compartment in inguinal proximal-, axillary distal-draining LN, spleen, and PBMC, as reported in (A). Each dot represents one mouse, and data are the merge of two to three independent experiments with the exception for PBMC, where each dot represents four to five pooled mice from one to two independent experiments. One-way ANOVA followed by Tukey’s posttest were used for statistical analyses: *p < 0.05, **p < 0.01.

FIGURE 3.

Frequencies of CD19+IgMIgD B cells. BALB/c mice were immunized twice, 1 mo apart, with HlaH35L, HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). B cell responses were measured 10 d after the last immunization by flow cytometry. (A) Frequencies of IgMIgD cells in the CD19+CD3 B cell compartment in inguinal proximal-, axillary distal-draining LN, spleen, and PBMC, as reported on top of each graph. (B) Frequencies of Hla-specific B cells gated in the CD19+IgMIgD compartment in inguinal proximal-, axillary distal-draining LN, spleen, and PBMC, as reported in (A). Each dot represents one mouse, and data are the merge of two to three independent experiments with the exception for PBMC, where each dot represents four to five pooled mice from one to two independent experiments. One-way ANOVA followed by Tukey’s posttest were used for statistical analyses: *p < 0.05, **p < 0.01.

Close modal

Overall, these results suggest that MF59 exerts its adjuvant effects beyond the injection site and proximal-draining LNs, and differs from alum by inducing higher frequencies of Ag-specific IgMIgD memory B cells in distal LNs, spleen, and blood; moreover, MF59 induces higher frequencies of Ag-specific IgM+ memory B cells in draining LNs compared with the other formulations.

To further determine the phenotype of CD19+IgMIgD Hla-specific B cells, we measured the expression of CD38 and GL-7, markers of memory and GC, respectively. CD38 is a cyclic ADP ribose hydrolase expressed by naive and memory B cells, and downregulated on GC B cells; GL-7 is commonly used as a marker for murine GC B cells (26). Combinatorial expression of these two markers in the IgMIgD compartment defines memory (CD38+GL-7) and GC B cells (CD38GL-7+), respectively, whereas naive B cells only express CD38 (Supplemental Fig. 3C).

In inguinal proximal-draining LN, most of the IgMIgD Hla-specific B cells were CD38GL-7+ (>75%) corresponding to a GC phenotype (Fig. 4A, 4B); no differences in the percentage were observed between MF59 and alum formulations, but the number of the IgMIgD Hla-spcific B cells/106 cells was higher in the MF59 group (Fig. 4C), possibly associated with the expansion of the total CD19+IgMIgD cells in this site (Fig. 3A). Axillary distal-draining LNs from HlaH35L plus MF59–immunized mice contained both GC CD38GL-7+ and memory CD38+GL-7 IgMIgD Hla-specific B cells (54 and 34%, respectively), whereas only memory CD38+GL-7 IgMIgD Hla-specific B cells were observed in HlaH35L plus alum–immunized animals (>85%) (Fig. 4A, 4B). In the spleen, IgMIgD Hla-specific B cells predominantly showed a CD38+GL-7 memory phenotype (>76%) without differences observed between MF59 and alum formulations (Fig. 4A, 4B).

FIGURE 4.

Characterization of CD19+IgMIgD Hla-specific B cells. BALB/c mice were immunized twice, 1 mo apart, with HlaH35L plus MF59 and HlaH35L plus alum. B cell responses were measured 10 d after the last immunization by flow cytometry. (A) Combinatorial expression of CD38 and GL-7 markers on CD19+IgMIgD Hla-specific B cells from proximal-, distal-draining LN, and spleen; numbers show the percentage of the gated populations. (B) Frequencies of HlaH35L-specific CD38GL-7+ GC B cells in inguinal proximal-, axillary distal-draining LN, and spleen, as gated in (A). (C) Number of HlaH35L-specific CD38GL-7+ GC B cells/106 cells in inguinal proximal-draining LN, as gated in (A). (D) Combinatorial expression of CD73 and CD80 on Hla-specific CD38+GL-7 memory and CD38GL-7+ GC B cells, as gated in (A); numbers show the percentage of the gated populations. Each organ from each mouse was processed and acquired separately. Fcs files from mice of the same group of immunization (four to five mice) were concatenated with the FlowJo software: (A) and (D) show one representative experiment, and (B) and (C) show three independent experiments, where each dot is one experiment. Unpaired Student t test was used for statistical analyses: *p < 0.05.

FIGURE 4.

Characterization of CD19+IgMIgD Hla-specific B cells. BALB/c mice were immunized twice, 1 mo apart, with HlaH35L plus MF59 and HlaH35L plus alum. B cell responses were measured 10 d after the last immunization by flow cytometry. (A) Combinatorial expression of CD38 and GL-7 markers on CD19+IgMIgD Hla-specific B cells from proximal-, distal-draining LN, and spleen; numbers show the percentage of the gated populations. (B) Frequencies of HlaH35L-specific CD38GL-7+ GC B cells in inguinal proximal-, axillary distal-draining LN, and spleen, as gated in (A). (C) Number of HlaH35L-specific CD38GL-7+ GC B cells/106 cells in inguinal proximal-draining LN, as gated in (A). (D) Combinatorial expression of CD73 and CD80 on Hla-specific CD38+GL-7 memory and CD38GL-7+ GC B cells, as gated in (A); numbers show the percentage of the gated populations. Each organ from each mouse was processed and acquired separately. Fcs files from mice of the same group of immunization (four to five mice) were concatenated with the FlowJo software: (A) and (D) show one representative experiment, and (B) and (C) show three independent experiments, where each dot is one experiment. Unpaired Student t test was used for statistical analyses: *p < 0.05.

Close modal

We further characterized the IgMIgD Hla-specific memory and GC B cells for the expression of CD80 and CD73 activation markers. CD80 is a ligand for CD28 and CTLA-4 expressed on T cells, and its expression on B cells is critical for a correct development of T-dependent B cell responses. A greater proportion of memory B cells upregulates CD80 expression (27, 28). CD73 is a nucleotidase that converts AMP in adenosine, is associated with high frequencies of mutation of the Ig genes, and plays a key role in Ig class switch recombination (2830). The combinatorial expression of CD80 and CD73 is not associated with a clear functional phenotype, but a double-positive population is considered more memory-like (hence, more differentiated) than a single-positive one (30, 31). In this study, we showed that IgMIgD Hla-specific CD38+GL-7 memory B cells upregulated the expression of both CD80 and CD73, whereas most of the IgMIgD Hla-specific CD38-GL-7+ GC B cells expressed mainly CD73, but not CD80 (Fig. 4D). No differences were observed between the two immunization groups. Similarly, IgM+IgD Hla-specific CD38+GL-7 memory B cells upregulated the expression of both CD80 and CD73 (Supplemental Fig. 3D), whereas naive IgM+IgD+ B cells expressed low levels of both CD80 and CD73 markers (Supplemental Fig. 3C).

To visualize GC structures in proximal-draining LNs of immune mice, we performed immunostaining of tissue sections with anti-CD21/35 (follicular DC), anti-IgD (naive B cells), and anti–GL-7 (GC) Abs, followed by confocal microscopy analyses. Mice that received HlaH35L plus MF59 or HlaH35L plus alum formulations showed characteristic GL-7+IgD GC structures 10 d after the last immunization (Fig. 5A, 5B). At the 1-mo time point, GL-7+ areas were reduced in size compared with day 10 in both vaccine formulation groups; however, LNs from mice that had received HlaH35L plus MF59 retained a higher number and larger GC structures (Fig. 5A, 5B).

FIGURE 5.

GC structures and Ag-specific TFH cells in immune mice. BALB/c mice were immunized twice, 1 mo apart, with HlaH35L, HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). (A) Confocal microscopy analyses of inguinal-draining LNs of mice immunized with HlaH35L plus MF59 or HlaH35L plus alum. LNs were collected 10 d or 1 mo after vaccination. LN sections were stained to detect follicular DC (green) and GC formation through upregulation of GL-7 (orange) and downregulation of IgD (blue). For each treatment at each time point are reported the images of IgD staining of whole LN sections (left, original magnification ×5) or the images of CD21/35, GL-7, and IgD stainings of selected follicles shown separately or merged (center and right, original magnification ×40). The image of one section is shown as example of consecutive sections of a whole LN, which is representative of three organs. Scale bar, 100 μm. (B) Ratio between the number of counted GC structures and the number of counted follicles, identified as in (A). (C) Frequencies of TFH cells (identified as PD-1+CXCR5+ICOS+ cells among all the CD4+ T cells) were measured in inguinal-draining LNs 10 d after the last immunization by flow cytometry. Data are the merge of two independent experiments with four to five mice/group. One-way ANOVA followed by Tukey’s posttest were used for statistical analyses: *p < 0.05.

FIGURE 5.

GC structures and Ag-specific TFH cells in immune mice. BALB/c mice were immunized twice, 1 mo apart, with HlaH35L, HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). (A) Confocal microscopy analyses of inguinal-draining LNs of mice immunized with HlaH35L plus MF59 or HlaH35L plus alum. LNs were collected 10 d or 1 mo after vaccination. LN sections were stained to detect follicular DC (green) and GC formation through upregulation of GL-7 (orange) and downregulation of IgD (blue). For each treatment at each time point are reported the images of IgD staining of whole LN sections (left, original magnification ×5) or the images of CD21/35, GL-7, and IgD stainings of selected follicles shown separately or merged (center and right, original magnification ×40). The image of one section is shown as example of consecutive sections of a whole LN, which is representative of three organs. Scale bar, 100 μm. (B) Ratio between the number of counted GC structures and the number of counted follicles, identified as in (A). (C) Frequencies of TFH cells (identified as PD-1+CXCR5+ICOS+ cells among all the CD4+ T cells) were measured in inguinal-draining LNs 10 d after the last immunization by flow cytometry. Data are the merge of two independent experiments with four to five mice/group. One-way ANOVA followed by Tukey’s posttest were used for statistical analyses: *p < 0.05.

Close modal

A hallmark of GC responses is the presence of TFH cells that sustain proliferation and differentiation of Ag-specific GC B cells; hence, we characterized the frequencies of TFH cells in inguinal-draining LNs of mice immunized with HlaH35L, HlaH35L plus alum, or HlaH35L plus MF59. LN cell suspensions were stained with PD-1, CXCR5, and ICOS markers to identify TFH cells in the CD3+CD4+ T cell compartment. Immune mice showed expanded frequencies of TFH cells compared with mice that received the saline formulation; furthermore, mice immunized with HlaH35L plus MF59 had higher frequencies of CD4+PD-1+CXCR5+ICOS+ T cells compared with the alum or nonadjuvanted group (Fig. 5C).

Overall, these data suggest that IgMIgD Hla-specific B cells are Ag-experienced B cells that, with the help of TFH cells, form strong GC responses in draining LN. Moreover, only MF59 induced Ag-specific GC B cells in proximal- and distal-draining LNs.

Induction of persistent immunity is desirable from each successful vaccine formulation; therefore, humoral and B cell responses were measured 4 mo after the second immunization in the 1-mo interval schedule (Fig. 6A). HlaH35L plus MF59 promoted higher Hla-specific IgG titers with higher avidity, and, most importantly, sera with higher functional titers compared with the HlaH35L plus alum group, 4 mo after the last immunization (Fig. 6B–D). In addition, HlaH35L plus MF59 were associated with higher number of Hla-specific bone marrow plasma cells (Fig. 6E) that could translate in the higher functional titers detected in this group compared with the HlaH35L plus alum group. Comparing these results with the 10-d time point in the 1-mo schedule (Fig. 2A), the number of Ag-specific bone marrow plasma cells increased by 2.3- and 2.6-fold for HlaH35L plus MF59 and HlaH35L plus alum, respectively. Although the overall frequencies of splenic Hla-specific memory B cells measured by ELISPOT decreased over time in all groups, HlaH35L plus MF59–immunized mice retained higher frequencies of Hla-specific memory B cells at all time points (Figs. 2B, 6F).

As memory B cells and bone marrow plasma cells mostly derive from GC, and both cell populations are maintained at higher levels in HlaH35L plus MF59–immunized mice, the persistence of Ag-specific GC B cells in these mice was analyzed by flow cytometry. HlaH35L plus MF59–immunized mice showed higher frequencies of IgMIgD Hla-specific B cells in inguinal-draining LNs compared with HlaH35L plus alum–treated mice (Fig. 7A, 7B). Most IgMIgD Hla-specific B cells showed a CD38+GL-7 memory phenotype in both immunization groups (67 and 89%, respectively). However, a persistent population of CD38GL-7+ Hla-specific GC B cells (22%) was still present in mice immunized with HlaH35L plus MF59 4 mo later (Fig. 7C, 7D). In agreement with the observations at the 10-d time point, splenic IgMIgD Hla-specific B cells were mainly CD38+GL-7 memory regardless of the adjuvant formulation (data not shown). Ag-specific memory and GC B cells were CD73+CD80+ and CD73+CD80, respectively (Fig. 7E), as already observed for those two populations at the 10-d time point.

FIGURE 7.

MF59 promotes persistent Ag-specific GC B cells. BALB/c mice were immunized twice, 1 mo apart, with HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). CD19+IgMIgD Hla-specific B cells were characterized by flow cytometry in inguinal-draining LNs 4 mo after the last immunization. (A and B) Frequencies of Hla-specific B cells; (C and D) combinatorial expression of CD38 and GL-7 memory and GC markers on Hla-specific B cells; (E) combinatorial expression of CD73 and CD80 activation markers on Hla-specific CD38+GL-7 memory and CD38GL-7+ GC populations, as gated in (C). (A, C, and E) Numbers show the percentage of the gated populations. Each pair of inguinal lymph nodes from each mouse was processed and acquired separately. Fcs files from mice of the same group of immunization (four to five mice) were concatenated with the FlowJo software: (A), (C), and (E) show one representative experiment, and (B)–(D) show three independent experiments where each dot is one experiment. Unpaired Student t test was used for statistical analyses: *p < 0.05. (F) Confocal microscopy analysis of inguinal-draining LNs of mice immunized with HlaH35L plus MF59 or HlaH35L plus alum, 4 mo after vaccination. LN sections were stained to detect follicular DC (green) and GC formation through upregulation of GL-7 (orange) and downregulation of IgD (blue). For each treatment are reported the images of IgD staining of whole LN sections (left, original magnification ×5) or the images of CD21/35, GL-7, and IgD stainings of selected follicles shown separately or merged (center and right, original magnification ×40). The image of one section is shown as example of consecutive sections of a whole LN, which is representative of three organs. Scale bar, 100 μm.

FIGURE 7.

MF59 promotes persistent Ag-specific GC B cells. BALB/c mice were immunized twice, 1 mo apart, with HlaH35L plus MF59, HlaH35L plus alum, or saline formulation (ctrl). CD19+IgMIgD Hla-specific B cells were characterized by flow cytometry in inguinal-draining LNs 4 mo after the last immunization. (A and B) Frequencies of Hla-specific B cells; (C and D) combinatorial expression of CD38 and GL-7 memory and GC markers on Hla-specific B cells; (E) combinatorial expression of CD73 and CD80 activation markers on Hla-specific CD38+GL-7 memory and CD38GL-7+ GC populations, as gated in (C). (A, C, and E) Numbers show the percentage of the gated populations. Each pair of inguinal lymph nodes from each mouse was processed and acquired separately. Fcs files from mice of the same group of immunization (four to five mice) were concatenated with the FlowJo software: (A), (C), and (E) show one representative experiment, and (B)–(D) show three independent experiments where each dot is one experiment. Unpaired Student t test was used for statistical analyses: *p < 0.05. (F) Confocal microscopy analysis of inguinal-draining LNs of mice immunized with HlaH35L plus MF59 or HlaH35L plus alum, 4 mo after vaccination. LN sections were stained to detect follicular DC (green) and GC formation through upregulation of GL-7 (orange) and downregulation of IgD (blue). For each treatment are reported the images of IgD staining of whole LN sections (left, original magnification ×5) or the images of CD21/35, GL-7, and IgD stainings of selected follicles shown separately or merged (center and right, original magnification ×40). The image of one section is shown as example of consecutive sections of a whole LN, which is representative of three organs. Scale bar, 100 μm.

Close modal

Finally, immunostaining of tissue sections from proximal-draining lymph nodes showed that mice that received HlaH35L plus MF59 had persistent GC structures detected up to 4 mo after the last immunization (3 GC structures identified on 27 follicles analyzed); instead, GC structures were not observed in the HlaH35L plus alum group at the same time point (0 GC structures identified on 38 follicles analyzed) (Fig. 7F). Altogether, these data suggest that MF59 promotes more persistent Ag-specific GC B cell populations with high expression of CD73 possibly associated with ongoing Ig somatic hypermutations (Fig. 8).

FIGURE 8.

Oil-in-water emulsion MF59 mechanisms of action. Altogether, we know that MF59 (I) enhances the recruitment of phagocytic cells at the site of injection 24 h after immunization (1113, 15); (II) induces cross-protective neutralizing Abs and memory B cells (1921); (III) expands Ag-specific T cell subsets that are predictive of protective levels of neutralizing Abs (17, 18); and in the present work, we learned that (IV) MF59 increases the persistence of Ag-specific GC B cells.

FIGURE 8.

Oil-in-water emulsion MF59 mechanisms of action. Altogether, we know that MF59 (I) enhances the recruitment of phagocytic cells at the site of injection 24 h after immunization (1113, 15); (II) induces cross-protective neutralizing Abs and memory B cells (1921); (III) expands Ag-specific T cell subsets that are predictive of protective levels of neutralizing Abs (17, 18); and in the present work, we learned that (IV) MF59 increases the persistence of Ag-specific GC B cells.

Close modal

Characterization of the Ag-specific B cell compartment in previously published studies has involved the use of BCR-transgenic mice or Ag/adjuvant models not immediately transferable to human applications (26, 28, 32). In this study, we characterized Ag-specific B cell responses in wild-type mice immunized with HlaH35L, a S. aureus Ag known to induce protective Ab responses (22, 23, 25), formulated with adjuvants licensed in human vaccines. Altogether, our findings demonstrate that MF59 induces higher and more persistent functional Ab titers compared with alum in three different vaccination schedules. ELISPOT analyses revealed that HlaH35L plus MF59 also induced higher Hla-specific bone marrow plasma cells and memory B cells compared with the alum formulation in the 1-mo schedule. We performed a comprehensive characterization of HlaH35L-specific B cells by flow cytometry and demonstrated that the oil-in-water emulsion MF59 not only induced increased number of Ag-specific GC B cells in proximal-draining LNs compared with the alum formulation 10 d after the immunization, but also promoted new GC formation in the distal-draining ones. In parallel to increasing the number of GC B cells in the draining LN, MF59 formulation also induced high frequencies of TFH cells. Most importantly, MF59 promoted persisting GC structures and Ag-specific GC B cells detected up to 4 mo after immunization by confocal microscopy and flow cytometry, respectively. In this study, we provide mechanistic insights of the B cell compartment associated with functional Ab responses following vaccination; we suggest that the same methodology can be extended to other vaccines or to response to infection.

In two seminal studies, Jenkins et al. (26, 33) reported that mice immunized with PE adjuvanted with CFA had Ag-specific GC B cells in distal-draining LN, long-lasting GC, and IgM+ memory B cells. CFA is a water-in-oil emulsion plus microbial products derived from Mycobacterium tuberculosis, whereas MF59 is an oil-in-water emulsion that may share some mechanisms of action, but still maintaining very different inflammatory profiles. Remarkably, MF59 also induced higher frequencies of Ag-specific IgM+ memory B cells compared with the other formulations; this population has the unique feature to further differentiate in Ag-specific GC B cells after Ag re-exposure, and could play a critical role in the context of MF59-aduvanted vaccines, as previous clinical evidence (1921, 34) and this work have demonstrated the strong impact that MF59 has on GC outcomes.

Differences between MF59 and alum in shaping the GC responses may reflect their different mechanisms of action to activate the innate immune system [extensively reviewed by De Gregorio et al. (11)]. Whereas alum efficiently delivers the Ag to conventional DC (9, 35), MF59 generates an immunocompetent environment at the injection site that favors the trafficking and activation of monocytes (14, 15, 36). Monocyte trafficking and differentiation toward a DC phenotype have already been associated with enhanced GC responses induced by TLR agonists (3741), and we speculate that monocyte-derived DC, rather than conventional DC, play a crucial role in MF59 adjuvanticity.

Previous publications have reported a correlation between the magnitude of GC B cells and TFH cells in response to increasing amount of Ag or to TLR agonists in the adjuvant formulation (39, 42, 43); in addition, a recent study from Siegrist and colleagues (44) has shown that MF59 promotes potent TFH and GC B cell responses in adult mice, whereas younger infant mice, which develop only few functional TFH cells, have poor GC and Ab responses. In the present work, we show higher frequencies of TFH cells in proximal-draining LNs of mice that received the MF59 formulation compared with the alum group, hence confirming the importance of the adjuvant in shaping TFH cell responses and providing a possible explanation to the sustained Ag-specific GC B cell responses observed with MF59 formulation.

To our knowledge, for the first time, we show that a human-licensed adjuvant promotes long-lasting Ag-specific GC B cells, suggesting that Ig-gene hypermutation and affinity maturation may still be ongoing up to 4 mo after immunization. Most importantly, our findings suggest that persisting GC reactions contribute to the generation of broadly neutralizing high-affinity Ab responses, typically associated with MF59-adjuvanted vaccines. Strategies to develop next-generation vaccines include Ag-adjuvant combinations that preferentially drive long-lasting and broad B cell and Ab responses, by promoting persistent GC reactions and improving the breadth of the Ig repertoire (45, 46). As example, structural vaccinology targets immunization strategies based on rationally designed Ags to drive B cell lineages toward protective epitopes. MF59 might be an ideal adjuvant candidate to meet this purpose as it is safe, is already licensed in humans, expands the Ab repertoire, and, as reported in this work, promotes strong and persistent GC B cells that can be targeted to enhance their affinity maturation fate.

Adjuvants have been traditionally used to increase the magnitude of licensed vaccines by enhancing the functional Ab responses. Today, we need that selected adjuvants guide the type of adaptive responses to produce the most effective forms of immunity against each specific pathogen. To our knowledge, this is the first time that an adjuvant used in human licensed vaccines is reported to induce strong and persistent GC B cell responses. Our work provides key insights on the mechanisms of action of MF59 (Fig. 8) and strongly supports the use of oil-in-water emulsions for the rational design of vaccine formulations to promote potent and persistent GC and high-avidity humoral responses.

We thank M. Tortoli and L. Formoso from Novartis Animal Research Center and C. Pozzi from Novartis Research Center for technical assistance; E. Faenzi for special help with ELISPOT B cell protocol; G. Galli and M. Bardelli for useful discussion on Ag labeling; M. Pallaoro for Ag formulations; and E. Cartocci, S. Marchi, and C. Facciotti for HlaH35L expression and purification. We thank B. Galletti, N. Bechi, and D. Casini for useful discussion on ELISA protocols. We thank Prof. Rodolfo Negri and Prof. Maria Teresa Fiorillo from University of Rome “La Sapienza” for useful scientific discussion, and Prof. S. A. Costa Clemens and I. Ferlenghi from Novartis Vaccines Academy for general support.

The online version of this article contains supplemental material.

This work was supported by internal funding by Novartis Vaccines and Diagnostics Srl.

Abbreviations used in this article:

A488

Alexa Fluor 488

alum

aluminum salt

DC

dendritic cell

GC

germinal center

LN

lymph node

TFH

T follicular helper.

1
Gavin
A. L.
,
Hoebe
K.
,
Duong
B.
,
Ota
T.
,
Martin
C.
,
Beutler
B.
,
Nemazee
D.
.
2006
.
Adjuvant-enhanced antibody responses in the absence of Toll-like receptor signaling.
Science
314
:
1936
1938
.
2
Clements
C. J.
,
Griffiths
E.
.
2002
.
The global impact of vaccines containing aluminum adjuvants.
Vaccine
20
(
Suppl. 3
):
S24
S33
.
3
Durando
P.
,
Icardi
G.
,
Ansaldi
F.
.
2010
.
MF59-adjuvanted vaccine: a safe and useful tool to enhance and broaden protection against seasonal influenza viruses in subjects at risk.
Expert Opin. Biol. Ther.
10
:
639
651
.
4
O’Hagan
D. T.
,
Ott
G. S.
,
Nest
G. V.
,
Rappuoli
R.
,
Giudice
G. D.
.
2013
.
The history of MF59(®) adjuvant: a phoenix that arose from the ashes.
Expert Rev. Vaccines
12
:
13
30
.
5
Vesikari
T.
,
Groth
N.
,
Karvonen
A.
,
Borkowski
A.
,
Pellegrini
M.
.
2009
.
MF59-adjuvanted influenza vaccine (FLUAD) in children: safety and immunogenicity following a second year seasonal vaccination.
Vaccine
27
:
6291
6295
.
6
Hem
S. L.
,
Hogenesch
H.
.
2007
.
Relationship between physical and chemical properties of aluminum-containing adjuvants and immunopotentiation.
Expert Rev. Vaccines
6
:
685
698
.
7
Hutchison
S.
,
Benson
R. A.
,
Gibson
V. B.
,
Pollock
A. H.
,
Garside
P.
,
Brewer
J. M.
.
2012
.
Antigen depot is not required for alum adjuvanticity.
FASEB J.
26
:
1272
1279
.
8
Dupuis
M.
,
McDonald
D. M.
,
Ott
G.
.
1999
.
Distribution of adjuvant MF59 and antigen gD2 after intramuscular injection in mice.
Vaccine
18
:
434
439
.
9
Flach
T. L.
,
Ng
G.
,
Hari
A.
,
Desrosiers
M. D.
,
Zhang
P.
,
Ward
S. M.
,
Seamone
M. E.
,
Vilaysane
A.
,
Mucsi
A. D.
,
Fong
Y.
, et al
.
2011
.
Alum interaction with dendritic cell membrane lipids is essential for its adjuvanticity.
Nat. Med.
17
:
479
487
.
10
Hornung
V.
,
Bauernfeind
F.
,
Halle
A.
,
Samstad
E. O.
,
Kono
H.
,
Rock
K. L.
,
Fitzgerald
K. A.
,
Latz
E.
.
2008
.
Silica crystals and aluminum salts activate the NALP3 inflammasome through phagosomal destabilization.
Nat. Immunol.
9
:
847
856
.
11
De Gregorio
E.
,
Caproni
E.
,
Ulmer
J. B.
.
2013
.
Vaccine adjuvants: mode of action.
Front. Immunol.
4
:
214
.
12
Mosca
F.
,
Tritto
E.
,
Muzzi
A.
,
Monaci
E.
,
Bagnoli
F.
,
Iavarone
C.
,
O’Hagan
D.
,
Rappuoli
R.
,
De Gregorio
E.
.
2008
.
Molecular and cellular signatures of human vaccine adjuvants.
Proc. Natl. Acad. Sci. USA
105
:
10501
10506
.
13
Dupuis
M.
,
Denis-Mize
K.
,
LaBarbara
A.
,
Peters
W.
,
Charo
I. F.
,
McDonald
D. M.
,
Ott
G.
.
2001
.
Immunization with the adjuvant MF59 induces macrophage trafficking and apoptosis.
Eur. J. Immunol.
31
:
2910
2918
.
14
Vono
M.
,
Taccone
M.
,
Caccin
P.
,
Gallotta
M.
,
Donvito
G.
,
Falzoni
S.
,
Palmieri
E.
,
Pallaoro
M.
,
Rappuoli
R.
,
Di Virgilio
F.
, et al
.
2013
.
The adjuvant MF59 induces ATP release from muscle that potentiates response to vaccination.
Proc. Natl. Acad. Sci. USA
110
:
21095
21100
.
15
Calabro
S.
,
Tortoli
M.
,
Baudner
B. C.
,
Pacitto
A.
,
Cortese
M.
,
O’Hagan
D. T.
,
De Gregorio
E.
,
Seubert
A.
,
Wack
A.
.
2011
.
Vaccine adjuvants alum and MF59 induce rapid recruitment of neutrophils and monocytes that participate in antigen transport to draining lymph nodes.
Vaccine
29
:
1812
1823
.
16
Ott
G.
,
Barchfeld
G. L.
,
Chernoff
D.
,
Radhakrishnan
R.
,
van Hoogevest
P.
,
Van Nest
G.
.
1995
.
MF59 design and evaluation of a safe and potent adjuvant for human vaccines
. In
Vaccine Design, Pharmaceutical Biotechnology
,
Vol. 6
.
Powell
M. F.
,
Newman
M. J.
, eds.
Plenum Press
,
New York
, p.
277
296
.
17
Galli
G.
,
Medini
D.
,
Borgogni
E.
,
Zedda
L.
,
Bardelli
M.
,
Malzone
C.
,
Nuti
S.
,
Tavarini
S.
,
Sammicheli
C.
,
Hilbert
A. K.
, et al
.
2009
.
Adjuvanted H5N1 vaccine induces early CD4+ T cell response that predicts long-term persistence of protective antibody levels.
Proc. Natl. Acad. Sci. USA
106
:
3877
3882
.
18
Spensieri
F.
,
Borgogni
E.
,
Zedda
L.
,
Bardelli
M.
,
Buricchi
F.
,
Volpini
G.
,
Fragapane
E.
,
Tavarini
S.
,
Finco
O.
,
Rappuoli
R.
, et al
.
2013
.
Human circulating influenza-CD4+ ICOS1+IL-21+ T cells expand after vaccination, exert helper function, and predict antibody responses.
Proc. Natl. Acad. Sci. USA
110
:
14330
14335
.
19
Khurana, S., W. Chearwae, F. Castellino, J. Manischewitz, L. R. King, A. Honorkiewicz, M. T. Rock, K. M. Edwards, G. Del Giudice, R. Rappuoli, and H. Golding. 2010. Vaccines with MF59 adjuvant expand the antibody repertoire to target protective sites of pandemic avian H5N1 influenza virus. Sci. Transl. Med. 2: 15ra5
.
20
Khurana
S.
,
Verma
N.
,
Yewdell
J. W.
,
Hilbert
A. K.
,
Castellino
F.
,
Lattanzi
M.
,
Del Giudice
G.
,
Rappuoli
R.
,
Golding
H.
.
2011
.
MF59 adjuvant enhances diversity and affinity of antibody-mediated immune response to pandemic influenza vaccines.
Sci. Transl. Med.
3
:
85ra48
.
21
Galli
G.
,
Hancock
K.
,
Hoschler
K.
,
DeVos
J.
,
Praus
M.
,
Bardelli
M.
,
Malzone
C.
,
Castellino
F.
,
Gentile
C.
,
McNally
T.
, et al
.
2009
.
Fast rise of broadly cross-reactive antibodies after boosting long-lived human memory B cells primed by an MF59 adjuvanted prepandemic vaccine.
Proc. Natl. Acad. Sci. USA
106
:
7962
7967
.
22
Bubeck Wardenburg
J.
,
Schneewind
O.
.
2008
.
Vaccine protection against Staphylococcus aureus pneumonia.
J. Exp. Med.
205
:
287
294
.
23
Rauch
S.
,
DeDent
A. C.
,
Kim
H. K.
,
Bubeck Wardenburg
J.
,
Missiakas
D. M.
,
Schneewind
O.
.
2012
.
Abscess formation and alpha-hemolysin induced toxicity in a mouse model of Staphylococcus aureus peritoneal infection.
Infect. Immun.
80
:
3721
3732
.
24
Fritz
S. A.
,
Tiemann
K. M.
,
Hogan
P. G.
,
Epplin
E. K.
,
Rodriguez
M.
,
Al-Zubeidi
D. N.
,
Bubeck Wardenburg
J.
,
Hunstad
D. A.
.
2013
.
A serologic correlate of protective immunity against community-onset Staphylococcus aureus infection.
Clin. Infect. Dis.
56
:
1554
1561
.
25
Ragle
B. E.
,
Bubeck Wardenburg
J.
.
2009
.
Anti-alpha-hemolysin monoclonal antibodies mediate protection against Staphylococcus aureus pneumonia.
Infect. Immun.
77
:
2712
2718
.
26
Pape
K. A.
,
Taylor
J. J.
,
Maul
R. W.
,
Gearhart
P. J.
,
Jenkins
M. K.
.
2011
.
Different B cell populations mediate early and late memory during an endogenous immune response.
Science
331
:
1203
1207
.
27
Good-Jacobson
K. L.
,
Song
E.
,
Anderson
S.
,
Sharpe
A. H.
,
Shlomchik
M. J.
.
2012
.
CD80 expression on B cells regulates murine T follicular helper development, germinal center B cell survival, and plasma cell generation.
J. Immunol.
188
:
4217
4225
.
28
Anderson
S. M.
,
Tomayko
M. M.
,
Ahuja
A.
,
Haberman
A. M.
,
Shlomchik
M. J.
.
2007
.
New markers for murine memory B cells that define mutated and unmutated subsets.
J. Exp. Med.
204
:
2103
2114
.
29
Schena
F.
,
Volpi
S.
,
Faliti
C. E.
,
Penco
F.
,
Santi
S.
,
Proietti
M.
,
Schenk
U.
,
Damonte
G.
,
Salis
A.
,
Bellotti
M.
, et al
.
2013
.
Dependence of immunoglobulin class switch recombination in B cells on vesicular release of ATP and CD73 ectonucleotidase activity.
Cell Reports
3
:
1824
1831
.
30
Tomayko
M. M.
,
Steinel
N. C.
,
Anderson
S. M.
,
Shlomchik
M. J.
.
2010
.
Cutting edge: hierarchy of maturity of murine memory B cell subsets.
J. Immunol.
185
:
7146
7150
.
31
Bergmann
B.
,
Grimsholm
O.
,
Thorarinsdottir
K.
,
Ren
W.
,
Jirholt
P.
,
Gjertsson
I.
,
Mårtensson
I.-L.
.
2013
.
Memory B cells in mouse models.
Scand. J. Immunol.
78
:
149
156
.
32
Dogan
I.
,
Bertocci
B.
,
Vilmont
V.
,
Delbos
F.
,
Mégret
J.
,
Storck
S.
,
Reynaud
C.-A.
,
Weill
J.-C.
.
2009
.
Multiple layers of B cell memory with different effector functions.
Nat. Immunol.
10
:
1292
1299
.
33
Taylor
J. J.
,
Pape
K. A.
,
Jenkins
M. K.
.
2012
.
A germinal center-independent pathway generates unswitched memory B cells early in the primary response.
J. Exp. Med.
209
:
597
606
.
34
Khurana
S.
,
Coyle
E. M.
,
Dimitrova
M.
,
Castellino
F.
,
Nicholson
K.
,
Del Giudice
G.
,
Golding
H.
.
2014
.
Heterologous prime-boost vaccination with MF59-adjuvanted H5 vaccines promotes antibody affinity maturation towards the hemagglutinin HA1 domain and broad H5N1 cross-clade neutralization.
PLoS One
9
:
e95496
.
35
Kool
M.
,
Soullié
T.
,
van Nimwegen
M.
,
Willart
M. A.
,
Muskens
F.
,
Jung
S.
,
Hoogsteden
H. C.
,
Hammad
H.
,
Lambrecht
B. N.
.
2008
.
Alum adjuvant boosts adaptive immunity by inducing uric acid and activating inflammatory dendritic cells.
J. Exp. Med.
205
:
869
882
.
36
Seubert
A.
,
Monaci
E.
,
Pizza
M.
,
O’Hagan
D. T.
,
Wack
A.
.
2008
.
The adjuvants aluminum hydroxide and MF59 induce monocyte and granulocyte chemoattractants and enhance monocyte differentiation toward dendritic cells.
J. Immunol.
180
:
5402
5412
.
37
Qu
C.
,
Edwards
E. W.
,
Tacke
F.
,
Angeli
V.
,
Llodrá
J.
,
Sanchez-Schmitz
G.
,
Garin
A.
,
Haque
N. S.
,
Peters
W.
,
van Rooijen
N.
, et al
.
2004
.
Role of CCR8 and other chemokine pathways in the migration of monocyte-derived dendritic cells to lymph nodes.
J. Exp. Med.
200
:
1231
1241
.
38
Langlet
C.
,
Tamoutounour
S.
,
Henri
S.
,
Luche
H.
,
Ardouin
L.
,
Grégoire
C.
,
Malissen
B.
,
Guilliams
M.
.
2012
.
CD64 expression distinguishes monocyte-derived and conventional dendritic cells and reveals their distinct role during intramuscular immunization.
J. Immunol.
188
:
1751
1760
.
39
Chakarov
S.
,
Fazilleau
N.
.
2014
.
Monocyte-derived dendritic cells promote T follicular helper cell differentiation.
EMBO Mol. Med.
6
:
590
603
.
40
Du
J.
,
Wu
Z.
,
Ren
S.
,
Wei
Y.
,
Gao
M.
,
Randolph
G. J.
,
Qu
C.
.
2010
.
TLR8 agonists stimulate newly recruited monocyte-derived cells into potent APCs that enhance HBsAg immunogenicity.
Vaccine
28
:
6273
6281
.
41
Cheong
C.
,
Matos
I.
,
Choi
J.-H.
,
Dandamudi
D. B.
,
Shrestha
E.
,
Longhi
M. P.
,
Jeffrey
K. L.
,
Anthony
R. M.
,
Kluger
C.
,
Nchinda
G.
, et al
.
2010
.
Microbial stimulation fully differentiates monocytes to DC-SIGN/CD209(+) dendritic cells for immune T cell areas.
Cell
143
:
416
429
.
42
Baumjohann
D.
,
Preite
S.
,
Reboldi
A.
,
Ronchi
F.
,
Ansel
K. M.
,
Lanzavecchia
A.
,
Sallusto
F.
.
2013
.
Persistent antigen and germinal center B cells sustain T follicular helper cell responses and phenotype.
Immunity
38
:
596
605
.
43
Rookhuizen
D. C.
,
DeFranco
A. L.
.
2014
.
Toll-like receptor 9 signaling acts on multiple elements of the germinal center to enhance antibody responses.
Proc. Natl. Acad. Sci. USA
111
:
E3224
E3233
.
44
Gavillet
B. M.
,
Eberhardt
C. S.
,
Auderset
F.
,
Castellina
F.
,
Seubert
A.
,
Tregoning
J. S.
,
Lambert
P.-H.
,
De Gregorio
E.
,
Del Giudice
G.
,
Siegrist
C.-A.
.
2015
.
MF59 mediates its B cell adjuvanticity by promoting T follicular helper cells and thus germinal center responses in adult and early life.
J. Immunol.
194: 4836–4845.
45
Moody
M. A.
,
Yates
N. L.
,
Amos
J. D.
,
Drinker
M. S.
,
Eudailey
J. A.
,
Gurley
T. C.
,
Marshall
D. J.
,
Whitesides
J. F.
,
Chen
X.
,
Foulger
A.
, et al
.
2012
.
HIV-1 gp120 vaccine induces affinity maturation in both new and persistent antibody clonal lineages.
J. Virol.
86
:
7496
7507
.
46
Haynes
B. F.
,
Kelsoe
G.
,
Harrison
S. C.
,
Kepler
T. B.
.
2012
.
B-cell-lineage immunogen design in vaccine development with HIV-1 as a case study.
Nat. Biotechnol.
30
:
423
433
.

All authors are employees of Novartis Vaccines, except G.L., F.M., and G.S.

Supplementary data