Neutrophils Are Required During Immunization With the Pneumococcal Conjugate Vaccine for Protective Antibody Responses and Host Defense Against Infection

Abstract

Neutrophils can shape adaptive immunity; however, their role in vaccine-induced protection against infections in vivo remains unclear. Here, we tested their role in the clinically relevant polysaccharide conjugate vaccine against Streptococcus pneumoniae (pneumococcus). We antibody depleted neutrophils during vaccination, allowed them to recover, and 4 weeks later challenged mice with pneumococci. We found that while isotype-treated vaccinated controls were protected against an otherwise lethal infection in naive mice, full protection was lost upon neutrophil depletion. Compared to vaccinated controls, neutrophil-depleted mice had higher lung bacterial burdens, increased incidence of bacteremia, and lower survival rates. Sera from neutrophil-depleted mice had less antipneumococcal IgG2c and IgG3, were less efficient at inducing opsonophagocytic killing of bacteria by neutrophils in vitro, and were worse at protecting naive mice against pneumococcal pneumonia. In summary, neutrophils are required during vaccination for optimal host protection, which has important implications for future vaccine design against pneumococci and other pathogens.

Streptococcus pneumoniae is a gram-positive bacterium with > 90 serotypes based on capsular polysaccharides [1]. These bacteria can cause pneumonia, meningitis, and bacteremia [2] and remain a serious cause of mortality and morbidity worldwide, particularly in the elderly [3]. Currently, 2 vaccines covering common disease-causing bacterial serotypes are available [4]. The pneumococcal polysaccharide vaccine (PPSV) consists of polysaccharides that directly cross-link B-cell receptors on mature B cells, leading to antibody (Ab) production independent of T cells [5]. PPSV is recommended for elderly individuals > 65 years old and adults with medical conditions [6]. As children < 2 years old lack mature B cells, they fail to produce T-independent Abs [7]. Therefore, the pneumococcal conjugate vaccine (PCV) was introduced for use in children. PCV contains polysaccharides linked to a carrier protein that triggers a T-dependent Ab response [4]. PCV has had great efficacy in children and is currently recommended for use in immunocompromised adults and elderly individuals with underlying conditions [6]. As PCV is recommended for adults with compromised immunity including B- and T-cell responses [4], it is important to elucidate novel players in vaccines that could be potential targets to boost protection.

Abs against capsular polysaccharides following vaccination bind to S. pneumoniae and protect the host against infection [4]. The functionality of Abs is determined by their antigen affinity. Affinity to antigens is mediated via the variable regions that make up the Fab or antigen binding portions and is optimized by somatic hypermutation (SHM) [8]. The Fc or constant region of Ab, which determines their class, also shapes their function, with the different classes of Abs having distinct immune modulating activities [8]. Abs against T-independent antigens such as bacterial polysaccharides are typically produced by marginal zone B cells in the spleen [9]. In contrast, T cell–dependent Ab production occurs in germinal centers, where a specialized subset of CD4⁺ T-follicular helper (TFH) cells [10] induces B cells to undergo class-switch recombination and SHM, resulting in Abs with improved function [11]. PCV significantly boosts class switching to immunoglobulin G (IgG) as compared to PPSV [12] and further induces TFH cells, which correlate with enhanced Ab function [13].

Polymorphonuclear leukocytes (PMNs), or neutrophils, play a crucial role in innate immunity to infections [14]. It is now appreciated that PMNs can also regulate adaptive immunity. PMNs can directly induce Ab production by B cells [11]. In the spleen, a subset of PMNs termed B-helper neutrophils was described to produce a proliferation inducing ligand (APRIL), B cell activating factor (BAFF), and interleukin 21 [9] that triggered Ab production by marginal-zone B cells [9, 15]. This was described for T-independent antigens including bacterial polysaccharides [9]. Although less established, PMNs may also affect T cell–dependent Ab responses [15]. PMNs are thought to both activate and suppress T cells [16]. PMNs produce a repertoire of chemokines that recruit T cells and also produce cytokines that drive T-cell subset differentiation [16]. PMN-derived products can prime T cells to more efficiently respond to antigens [17]. PMNs also activate T cells via recruiting and activating antigen-presenting cells [18] or presenting antigens themselves [16, 19–23]. In contrast, PMNs produce compounds that inhibit T-cell activation [24] including ROS [25], arginase 1 [26], and serine proteases [27]. Therefore, it is unclear whether, upon in vivo vaccination, PMNs would suppress or induce T-dependent Ab responses. Furthermore, although there have been elegant studies characterizing mechanisms of PMN interactions with B and T cells, most of the work has been done either in vitro or in vivo using model antigens [11, 18, 24]. Thus, studies examining the role of PMNs in clinically relevant vaccinations and how that shapes protection against in vivo infections are needed.

PMNs are required to control bacterial numbers following S. pneumoniae infection [28, 29]. PMNs also play a role in Ab responses against pneumococci. When compared to healthy controls, patients with neutropenic disorders had lower levels of Abs to some pneumococcal polysaccharides [9]. In mice, splenic PMNs localized with marginal zone B cells and were required for production of T-independent Abs during pneumococcal infection [15, 30]. However, whether PMNs shape responses to the pneumococcal conjugate vaccine and if they impair or promote Ab production remains unexplored. Here we tested the role of PMNs in response to PCV and found they were required at the time of vaccination for optimal Ab responses as well as host protection against pneumococcal infection. This study highlights the link between PMNs and Ab responses in the context of a clinically relevant immunization, which has far-reaching implications for vaccine design against S. pneumoniae and other pathogens.

MATERIALS AND METHODS

Mice

Female C57BL/6 mice (6–8 weeks) were purchased from Jackson Laboratories (Bar Harbor, Maine) for all experiments. Mice were housed in a specific-pathogen–free facility at the University at Buffalo and experiments were conducted in accordance with Institutional Animal Care and Use Committee guidelines.

Bacteria

Wild-type S. pneumoniae TIGR4 and capsule-deletion mutant (∆cps) S. pneumoniae were gifts from Andrew Camilli. Bacteria were grown to midexponential phase in Todd–Hewitt broth (BD Biosciences) supplemented with Oxyrase (Oxyrase) and 0.5% yeast extract at 37°C/5% carbon dioxide_. Aliquots were frozen at –80°C in growth media with 20% glycerol. Prior to use, aliquots were thawed, washed, and diluted in phosphate-buffered saline. Titers were confirmed by plating on tryptic soy agar plates supplemented with 5% sheep blood agar (Northeast Laboratory Services).

Immunization

Mice were immunized via intramuscular injection of 50 µL of the PCV Prevnar-13 (Wyeth Pharmaceuticals) into the caudal thigh muscle. Sera were collected from all mice before and at weeks 2 and 4 postimmunization and saved at –80°C for subsequent assays.

Neutrophil Depletion

Mice were treated intraperitoneally with 50 µ g of the Ly6G-depleting antibody IA8 or isotype IgG control (BioXCell) following the timeline in Figure 1A.

Adoptive Transfer of Sera

Five weeks following immunization, vaccinated, vaccinated PMN-depleted, and naive mice were euthanized and blood was harvested via cardiac puncture. Sera were obtained from the blood, pooled for each group, and transferred intraperitoneally (250 µ L) into naive recipients. Recipients were then infected 1 hour later [31].

Animal Infections

Mice were intratracheally challenged with 10⁷ colony-forming units (CFU) of S. pneumoniae [31]. Following infection, 1 set of mice was monitored daily over 1 week for bacteremia and clinical signs of disease including weight loss, activity level, posture, and breathing and blindly scored from 0 (healthy) to 21 (severely sick). Twenty-four hours postinfection, another set of mice was euthanized, and lung and blood were assessed for CFU.

Antibody Enzyme-Linked Immunosorbent Assay

Sera Ab levels were measured by enzyme-linked immunosorbent assay (ELISA) as previously described [31]. Nunc maxisorp plates were coated overnight at 4°C with type 4 pneumococcal polysaccharide (ATCC) at 2 μg/well. The sera were preabsorbed with a pneumococcal cell wall polysaccharide mixture (CWP-multi, Cederlane) to neutralize noncapsular Abs and added to the plate. After 3 hours of incubation and washing, pneumococcal-specific Abs were detected using horseradish peroxidase–conjugated goat antimouse immunoglobulin M (IgM; Invitrogen), IgG (Millipore Sigma), or IgG1, IgG2b, IgG2c, or IgG3 (Southern Biotech) followed by TMB substrate (Thermo Scientific), and readings were done at optical density 650 (OD₆₅₀) using a BioTek reader. Kinetic ELISAs were performed with readings every minute for 10 minutes. Ab units were calculated as percentages of a control hyperimmune serum included in each ELISA. Hyperimmune sera were pooled from mice that were intranasally inoculated with S. pneumoniae TIGR4 over 4 weeks as previously described [31], immunized with PCV at week 4, and injected intraperitoneally with heat-killed bacteria at week 5.

Myeloperoxidase ELISA

Myeloperoxidase (MPO) levels were measured in the lung, spleen, and blood using the Mouse Myeloperoxidase ELISA kit from Invitrogen as per the manufacturer’s instructions. The number of PMNs was then calculated using purified bone-marrow PMNs from naive mice to generate standard curves [28, 29].

Isolation of PMNs

PMNs were isolated from the bone marrow using density centrifugation with Histopaque 1119 and Histopaque 1077 (Sigma) as previously described [28]. PMNs were resuspended in Hanks’ balanced salt solution (HBSS without Ca²⁺ and Mg²⁺) supplemented with 0.1% gelatin and used in subsequent experiments. Purity was confirmed by flow cytometry and the isolated cells were 85%–90% Ly6G⁺.

Opsonophagocytic Killing Assay

The ability of PMNs to kill pneumococci was assessed as previously described [28]. In brief, 1 × 10⁵ PMNs were infected with 1 × 10³ bacteria preopsonized with 3% mouse sera and rotated at 37°C for 40 minutes. Reactions were stopped on ice and plated for CFU. The percentage of bacteria killed was calculated using no PMN controls.

Flow Cytometry

One day following the last depletion, mice were euthanized and blood, vaccine draining popliteal lymph nodes, and spleen were harvested. Single-cell suspensions of splenocytes and lymph nodes were prepared by mashing the organs. Red blood cells were lysed with a hypotonic buffer and the cells were stained for Ly6G (IA8 or RB6, Biolegend), CD11b (M1/70, Invitrogen,), CD11c (N418, BD Bioscience), F4/80 (BM8, BD Bioscience), and Ly6C (AL-21, BD Bioscience) in the presence of Fc-block (BD Bioscience). Live cells were identified using a dead cell stain kit (Life Technologies). Fluorescence intensities were measured on a BD Fortessa and at least 20 000 events were analyzed using FlowJo software.

Statistical Analysis

All statistical analysis was done using GraphPad Prism version 8 software. Significant differences were determined by Fisher exact test, 1-way analysis of variance followed by Dunnett test, Student t test, or 1-sample t test. Survival was analyzed using the log-rank (Mantel-Cox) test. P values < .05 were considered significant.

RESULTS

PMNs Are Required at the Time of Immunization With PCV for Host Protection Against Pneumococcal Infection

To test if PMNs were required for protection at the time of vaccination with PCV, we used isotype controls or the anti-Ly6G Ab IA8 to deplete PMNs 1 day prior to and every 2 days throughout the first week following vaccination (see timeline in Figure 1A). One day after the final treatment, we verified depletion in the blood and found an approximately 99% reduction in the number of circulating PMNs (Supplementary Figure 1A and 1B). Splenic PMNs also have a role in Ab production [9, 15]. There was an approximately 2-fold increase in splenic PMNs following vaccination, and anti-Ly6G treatment resulted in approximately 98% depletion of those cells (Supplementary Figure 1C). We verified that Ab treatment was specific to PMNs and did not result in changes in the number of circulating and splenic monocytes, dendritic cells, or macrophages (Supplementary Figure 3). We further tracked PMNs over time using MPO ELISA (Supplementary Figure 2A). In the circulation, by day 14, the amounts of PMNs in the depleted groups had recovered to approximately 73% of isotype-treated controls, and the amounts in the 2 groups were indistinguishable by day 28 (Supplementary Figure 2B). Similarly, we saw a gradual recovery of splenic PMNs following depletion (Supplementary Figure 2C).

Four weeks following vaccination, we challenged mice intratracheally with S. pneumoniae TIGR4 strain. Invasive S. pneumoniae infection results in pneumonia primarily, but up to 30% of patients with pneumococcal pneumonia also develop bacteremia and have a worse prognosis [32]. As S. pneumoniae strains can differ considerably [33], we chose the well-characterized serotype 4 isolate TIGR4, originally isolated from a bacteremic patient, as a model of a highly invasive infection modeling pneumonia that results in bacteremia [34] and that is covered by PCV. We then monitored the disease course and found that while all naive mice rapidly succumbed to infection, 100% of vaccinated mice survived (Figure 1B). However, unlike vaccinated controls, full protection was lost in PMN-depleted mice (Figure 1). The majority of PMN-depleted mice displayed severe clinical signs of disease where 77.8% got sick as compared to only 12.5% of vaccinated controls (Figure 1C). PMN-depleted mice had between 10- and 100-fold higher pulmonary bacterial numbers (Figure 2A) and systemic spread (Figure 2B), culminating in significantly reduced survival (Figure 1B) as compared to vaccinated controls. This reduced protection in PMN-depleted mice was not due to the continued absence of PMNs at the time of challenge, as we verified that there was no difference in PMNs in the lungs, spleen, or blood between the anti-Ly6G and isotype treated groups postinfection (Supplementary Figure 2D). Rather, our findings suggest that PMNs are required at the time of vaccination for full protection against subsequent pneumococcal infection.

PMNs Are Required for Optimal Antibody Isotype Switching in Response to PCV Immunization

Next, we wanted to explore the mechanisms by which PMNs contributed to vaccine-induced protection. We first examined Ab production and, as expected, observed isotype switching to IgG by week 4 postvaccination (Figure 3). We found that PMN depletion did not alter IgM or total IgG levels against capsular polysaccharide type 4 (Figure 3A and 3B) or heat-killed S. pneumoniae (not shown). However, when we examined IgG subtypes, we found that PMN depletion during vaccination resulted in slightly reduced levels of IgG2b (Figure 4A) and significantly lower levels of IgG2c (Figure 4B) and IgG3 (Figure 4C) at week 4 postvaccination. Interestingly, IgG1 levels (Figure 4D) were slightly, but not significantly elevated in the PMN-depleted group as compared to vaccinated controls. These data suggest that PMNs play a role in class switching to certain IgG subtypes.

PMNs Are Required for Optimal Antibody Function Following PCV Immunization

Apart from Ab levels, Ab function is key for vaccine efficacy [35]. We next explored if PMNs affected Ab affinity to bacterial surfaces. We tested the ability of IgG in the sera of the different mouse groups to bind the surface of S. pneumoniae by flow cytometry. Very little IgG bound to bacteria upon incubation with naive sera. However, we observed a 30-fold increase in the amount of IgG bound to bacteria when sera from vaccinated mice were used (Figure 5A). As expected, in immune sera, the bound IgG was specific to capsular polysaccharides as very little IgG bound to acapsular bacteria (∆cps S. pneumoniae). Interestingly, we observed a significant decrease in the amount of IgG bound to S. pneumoniae opsonized with sera from PMN-depleted mice as compared sera from vaccinated controls (Figure 5A).

We next compared the opsonic capacity of Abs by comparing the ability of sera to induce opsonophagocytic killing of S. pneumoniae by primary PMNs isolated from naive mice. We found that sera from vaccinated controls significantly boosted bacterial killing by PMNs where 60% of the bacterial input was killed by PMNs in the presence of immune sera as compared to approximately 10% with naive sera (Figure 5B). Strikingly, sera from PMN-depleted mice failed to induce opsonophagocytic killing of S. pneumoniae by PMNs where only 3% of bacteria were killed (Figure 5B). Abs can also activate the complement pathway and directly kill bacteria [8], but we detected no differences in the ability of sera alone from any of the groups to kill pneumococci (Supplementary Figure 4).

Given the difference in the in vitro function we observed, we finally tested the protective activity of Abs generated upon vaccination in the absence of PMNs. Naive young mice were injected intraperitoneally with 5-week sera from either vaccinated controls, naive mock-immunized mice, or PMN-depleted vaccinated groups. Mice were then challenged intratracheally with S. pneumoniae TIGR4 1 hour following sera transfer. We found that while all of the mice receiving naive sera succumbed to infection, all of the mice receiving sera from vaccinated controls survived the challenge (Figure 5C). In contrast, only half of the mice receiving sera from the PMN-depleted vaccinated group survived (Figure 5C). These data indicate that Abs produced during vaccination in the absence of PMNs are not sufficient to provide protection against subsequent pneumococcal infection.

DISCUSSION

Traditionally, PMNs are viewed as effectors of vaccine responses where vaccination triggers Abs that bind to pathogens and promote their clearance via enhancing uptake and killing by PMNs [36]. However, the extent to which PMNs contribute to vaccine-mediated protection against infections in vivo has not been fully elucidated. In this study, we explored the role of PMNs in immunization with PCV. We found that PMNs were needed for production of functional Abs following vaccination. Importantly, PMNs were required at the time of immunization for full protection against subsequent invasive pneumococcal infection. Our findings highlight the in vivo role of PMNs as inducers of protective vaccine responses against S. pneumoniae infections.

The mechanisms by which PMNs mediate Ab production in response to PCV is unclear. In adults, polysaccharides can directly cross-link B-cell receptors and elicit Ab production independent of T cells [5]. PCV converts this T-independent response to one that involves T cells as it consists of polysaccharides linked to the carrier protein CRM₁₉₇ [5]. This generates T cells specific to the carrier protein [37, 38]. When B cells recognize polysaccharides, they are thought to bind and internalize the polysaccharide along with its protein carrier and then display peptides derived from the carrier on MHC-II. This allows these polysaccharide-specific B cells to interact with carrier-peptide specific T cells, which in turn help the B cells produce anti-polysaccharide Abs [39]. Therefore, in the context of PCV, PMNs could either be working on B cells, T cells, or both. In humans, a subset of splenic PMNs directly induces Ab production by marginal-zone B cells in response to T-independent antigens including bacterial polysaccharides [9]. In mice, splenic B helper PMNs were found to produce pantrexin3, which was important for IgM production following immunization with the unconjugated pneumococcal polysaccharide vaccine [15]. Pantrexin3 was also important for T cell–independent IgM and IgG production against polysaccharides following infection with S. pneumoniae [15]. The role of PMNs in T-dependent responses is less clear. In mice, PMNs impaired IgA but not IgG or IgM production in response to vaccination with the adjuvant Bacillus anthracis edema toxin [40]. Mouse PMNs were found to directly present ovalbumin peptides to CD4⁺ T cells triggering T-cell cytokine production and proliferation [20]. However, IgG2 responses to ovalbumin were not impaired in pantrexin3^–/– mice [15]. In contrast, Abs against influenza PR8 and the T-dependent antigen TNP-Ficoll required pantrexin3 production by murine PMNs [15]. Similarly, human PMNs were able to present influenza hemagglutinin to CD4⁺ T cells [22]. In rhesus macaques, PMNs presented human immunodeficiency virus envelope glycoproteins to CD4⁺ T cells [22] and induced Ab production against simian immunodeficiency virus when co-cultured with B cells [41]. Here, in the context of immunization with PCV, PMNs are clearly required for optimal Ab responses; however, whether they are acting on T cells remains to be determined.

A key finding is that PMNs are required for the production of functional Abs. Ab functionality is determined by the affinity and avidity to antigen [8]. Although we detected similar levels of IgM and IgG in sera from PMN-depleted and isotype-treated vaccinated mice, the ability of IgG to bind pneumococci significantly decreased when they were generated in the absence of PMNs. This suggests that Ab affinity is increased in the presence of PMNs. Ab affinity is improved by SHM of the Fab variable regions and typically occurs in germinal centers and requires T cells [42], although human splenic B helper PMNs may contribute to SHM in marginal-zone B cells [9].

Ab function is also influenced by their subclass, which is determined by their Fc region [8]. The Fc portion of Abs shape effector function since they determine binding to Fc receptors on PMNs and complement activation [8]. Here, PMNs were required for class switching to IgG2c and IgG3 but not IgG1 subtypes. The IgG subtype produced in response to PCV varies in humans based on age, with IgG2 being the predominant response in adults [12, 43]. We found that PMNs were crucial for the ability of Abs to elicit opsonophagocytic killing of bacteria by primary cells. This is in line with data from humans where IgG2 was reported to have the highest opsonophagocytic activity while IgG1 had the lowest activity against serotype 4 pneumococci [44]. We also found that PMNs were required for production of Abs that protect against infection. As IgG subtypes vary between human and mice, with human IgG2 and mouse IgG3 responding to bacterial polysaccharides [44, 45], our observation here is likely mediated by IgG3, which is protective in mice against pneumococcal infection [46].

How PMNs are inducing class-switching to IgG2c and IgG3 but not IgG1 in response to PCV is unclear. Efficient class switching from IgM to IgG requires T-cell help [39]. In adults, PCV significantly boosts class switching to IgG as compared to the unconjugated polysaccharide vaccine [12]. Cytokines produced by T cells further determine the subtype of IgG produced with interleukin 17 (IL-17) and interferon gamma (IFN-γ) enhancing switching to IgG3 and IgG2 more than IgG1 [47, 48]. As PMNs can both produce cytokines [16] and drive Th1 and Th17 cell differentiation [20], they may contribute to isotype switching by producing IFN- γ or IL-17 themselves or eliciting T cells to do so.

In summary, we demonstrate here that PMNs are required at the time of immunization with PCV for optimal protective Ab responses and host protection against subsequent S. pneumoniae infection. Our work has clinical implications for the timing of vaccine administration in neutropenic individuals, and those with known PMN defects. As serotype replacement by bacterial strains not covered by the current vaccines continue to emerge, novel serotype-independent vaccine formulations such as whole cell vaccines or common pneumococcal protein vaccines are being considered [49]. Therefore, future vaccine designs should take PMN responses into consideration, particularly in susceptible populations such as the elderly [2], in whom PMN responses are known to be dysregulated [50].

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Author contributions. E. Y. I. T. conducted research, analyzed data, and wrote the manuscript. M. B. conducted research and analyzed data. E. A. W. provided essential reagents and expertise. E. N. B. G. designed research, wrote the manuscript, and had responsibility for final content. All authors read and approved the final manuscript.

Acknowledgments. We acknowledge Shaunna Simmons for critical reading and discussion of the manuscript.

Financial support. This work was supported by the National Institutes of Health (grant number R00AG051784 to E. N. B. G.).

Potential conflicts of interest. All authors: No reported conflicts of interest.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References