Neutrophils influence the level of antigen presentation during the immune response to protein antigens in adjuvants

Abstract

Neutrophils modulated antigen presentation following immunization with antigens in complete or incomplete Freund's adjuvant or alum. The neutrophils had an important negative role in the CD4 T cell and B cell responses to three protein antigens, hen-egg white lysozyme (HEL), ovalbumin (OVA) and listeriolysin O (LLO). In their absence (by depleting with antibodies for only the first twenty-four hours, or using genetically neutropenic mice), the cellular responses increased several fold. The CD8 response was not affected or slightly decreased. Competition for antigen between the presenting cells and the neutrophils, as well as an effect on the response to antigen-bearing dendritic cells (DC) was documented. Neutrophils entered the draining lymph nodes, rapidly and for a brief period of several hours, localizing mainly to the marginal sinus and superficial cortex. There they established brief contacts with DC and macrophages. Moreover, neutrophils imprinted on the quality of the subsequent DC-T cell interactions, despite no physical contacts with them: by intravital microscopy, the clustering of antigen-specific T cells and DC was improved in neutropenic mice. Thus, neutrophils are obligate cells that briefly enter sites of immunization and set the level of antigen presentation: a brief depletion may have a considerably positive impact on vaccination.

Introduction

Neutrophils are essential effector cells in acute infections associated with a wide number of microrganisms. They are well established to be the major effectors in infections with extracellular bacteria, but also play a role in controlling infections with intracellular bacteria, viruses, fungi and parasites (1–3). Neutrophils kill or inhibit the growth of many of these organisms through their generation of reactive oxygen species and/or microbicidal peptides (4). Whether neutrophils during infection influence the T cell responses independent of their microbicidal effect has been difficult to firmly establish because of the ensuing infection. Although to most infectious diseases neutrophil depletion results in their exacerbation, a few instances have shown an improvement (5), suggesting a regulatory role (6–9). For one example, in the Listeria monocytogenes model where neutrophils have a major effect on the course of the infections, there is evidence that their depletion affected the degree of CD8 T cell priming (10). It is known, mostly by ex vivo experiments, that neutrophils migrate to the lymph nodes (11, 12), release cytokines and intracellular enzymes (13–16), degrade internalized proteins, and interact with DC (17, 18), all relevant issues when examining their function during in vivo immune reactions (19, 20).

We examined here whether neutrophils had an effect on the lymphocyte response to protein antigens in in vivo situations. Mice were immunized with proteins in adjuvant and the cellular and antibody response was examined in the absence of neutrophils and compared to that of untreated mice. We found a consistent and strikingly negative effect of neutrophils in the CD4 T and B cell responses when antigens were given in any of various adjuvants.

Materials and Methods

Mice

All mice were bred and maintained under pathogen-free conditions at Washington University in St. Louis in accordance with institutional animal care guidelines. C57.BL/6 (B6) B10.BR and C.B-17 mice (H-2^d haplotype) were obtained from The Jackson Laboratory. B10.BR mice having NOX2^−/− or iNOS^−/− were generated from B6 mutant mice (21). C57.BL/6, G-CSFR^−/− mice were generated and obtained from the laboratory of Daniel Link (Washington University in St. Louis, MO). LysM-eGFP were generated by Faust et al. (22) and CD11c-eYFP by Lindquist et al. (23). All mice were crossed to B10.BR background. LysM-eGFP/CD11c-eYFP double reporter mice were the F1 heterozygote. B10.BR mice bearing a membrane form of HEL, mHEL, were generated in our laboratory (24). C.B-17 mice were originally purchased from Taconic and maintained at the animal facility of Washington University.

Antigen

HEL was obtained from Sigma and purified by affinity chromatography to remove about 3% of contaminant proteins. Purified HEL contained <0.1 EU/μg LPS. Recombinant LLO protein was generated and purified as previously described (25). OVA protein was from Worthington Biochemical. Peptides were synthesized by Fmoc techniques and verified by mass spectrometry: HEL:48–62 (DGSTDYGILQINSRWW), HEL:20–35 (YRGYSLGNWVCAAKFE), HEL:31–47 (AAKFESNFNTQATNRNT), HEL:114–129 (RCKGTDVQAWIRGCRL), LLO:91–99 (GYKDGNEYI), LLO:188–201 (RWNEKYAQAYPNVS), OVA:323–339 (ISQAVHAAHAEINEAGR) and OVA:257–264 (SIINFEKL).

Immunization and cellular assays

Mice were injected i.p. with 250μg of RB6-8C5 (2), or 1A8 (BioXCell) monoclonal antibodies or with isotype control rat IgG (Sigma) before immunization. Protein antigen was emulsified with Freund's complete or incomplete adjuvant or alum and injected into the footpad. T cell responses were measured, usually at day 7 of immunization using ELISPOT analysis in which popliteal lymph nodes were harvested for single cell suspension and challenged with individual proteins or peptides in vitro.

For antigen presentation assays, popliteal lymph nodes were harvested and digested with Liberase TL (Roche) at 37°C. Single cell suspension was negative selected with anti-CD90.2 magnetic beads (Miltenyi Biotec) and used as a base for the following treatment: DC was separated with anti-CD11c magnetic beads; B cells were separated by anti-CD19 magnetic beads; macrophages were obtained by excluding CD3⁺, CD11c⁺ and CD19⁺ cells. Each single fraction was fixed with 1% paraformaldehyde and tested on T cell hybridoma 3A9 reactive to HEL:48–62.

In experiments transferring APC, B10.BR or mHEL mice were i.p. injected with 10 μg Flt3-ligand for 3 consecutive days. At day 8, spleens were harvested and digested with collagenase D (Sigma) for single cell suspension. DC was enriched with anti-CD11c magnetic beads. For macrophage, splenic cells were negative selected with anti-CD11c magnetic beads, followed by positive selection with anti-CD11b magnetic beads. Cells were stimulated with 10ng/ml LPS for 1 hour at 37°C, followed by extensive wash. Cells (5×10⁵-1×10⁶) were transferred into the footpad of neutrophil-depleted or -undepleted B10.BR recipient mice. After 3 hours of cell transfer, incomplete Freund's adjuvant was injected into the footpad to induce neutrophil migration. T cell response was measured by ELISPOT at day 7 of cell transfer.

Serum antibody assay

The serum antibodies were detected by ELISA. ELISA plates (Nunc, Roskilde, Denmark) were precoated with 10 μg/ml HEL or peptides in sodium bicarbonate buffer (pH=8.8) overnight at 4°C. Plates were blocked with 1% bovine serum albumin for 1 hour at room temperature and washed with 0.05% Tween 20. Serial diluted serum was added and incubated for 2 hours. After wash, secondary antibody (goat anti-mouse IgG-peroxidase 1:5000) was incubated for 1 hour and the plates were developed with 1 mM 2-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) in citrate buffer with 0.03% H₂O₂ (Roche, Basel, Switzerland). The absorbency was measured at 405 nm.

Microscopy techniques

Popliteal lymph nodes were embedded with Tissue-tek OCT compound (Sakura Finetek) on dry ice and 5 μm frozen sections were obtained using cryostat (Microm). Immunofluorescence staining was performed by applying fluorescent dye-conjugated antibodies onto the sections for 30–60 minutes at room temperature. Images were captured at 4× or 10× magnification using an Olympus B×51 microscope (Olympus). For confocal microscopy, images were captured using a Zeiss 510 laser scanning confocal microscope. Histological sections of lymph nodes stained by hematoxylin and eosin were prepared by standard histological techniques.

For autoradiography studies, HEL was labeled with ¹²⁵I (Amersham Biosciences) using the chloramine T method to the activity of 5.7×10⁶ cpm/μg. B10.BR mice were injected with 130μg I¹²⁵-HEL (7.4×10⁸ cpm) in complete Freund's adjuvant into the footpad for 16 hours. Individual cell subsets from popliteal lymph nodes were prepared as described above. The radioactivity of each cell subset was determined using automatic gamma counter (PerkinElmer). For autoradiography analysis, cells were spun onto slides which were then coated with NTB liquid emulsion (Eastman Kodak), dried, and exposed at 4°C for variable periods of time. Development was performed with Kodak D19 developer followed by fixation (Eastman Kodak). Cells were counterstained with the Hema3 stain set (Fisher Scientific). In different experiments the localization of HEL in the lymph nodes was carried out using HEL labeled with HiLyte Fluor 555 using a protein labeling kit (AnaSpec).

For imaging of DC and T cell interactions in intact lymph nodes by two-photon microscopy, 10⁷ splenic T cells isolated from 3A9 mice (CD4 T cell isolation kit, Miltenyi Biotec) were labeled with 10mM red CMTPX (Invitrogen Molecular Probes) at 37°C for 30 minutes and transferred i.v. into LysM-eGFP/CD11c-eYFP mice 18 hours before 10nmole HEL/IFA immunization. Popliteal lymph nodes were harvested at 3–10 hours after immunization, stabilized on a plastic coverslip and placed in a flow chamber and maintained at 37°C by perfusion with warm media bubbled with a mixture of 95% O₂ and 5% CO₂. Images were acquired using an Olympus BX51WI fluorescence microscope equipped with a 20× objective (Olympus) controlled by Image Warp software (BitFlow). Chamelion Ti:Sapphire laser was used to excite the sample at 890nm. For detection, eGFP was collected below 510nm (viewed as cyan), eYFP was collected between 510 and 560nm (viewed as green), and CMTPX was collected above 560nm (viewed as red). Tissue volumes of 200×220×50μm were captured every 27 seconds by acquiring 21 sequential 2.5μm Z-steps, each with 0.5 second averaging to increase signal contrast. Typically, recordings lasted between 30–40 minutes. Multi-dimensional rendering and three-dimensional cell tracking were performed using Imaris software (Bitplane AG). Automated tracking was manually checked for accuracy. DC-T cell contact analysis was performed manually on 3–5 individual videos acquired from independent experiments. At least 100 cells in each single field were counted. DC-T cell contacts were observed in three-dimensional images in which two cells were tightly apposed, meaning there were no black pixels between the cells. Total number of DC and T cells as well as the number of DC-T cell contacts was counted every 5 minutes on the videos.

Statistical analysis

All measurements were presented as mean ± SD. Statistical analysis was processed by Mann-Whitney test using Prism 5.02 software (GraphPad, San Diego, CA). In all figures statistical significance of p values <0.05 are indicated as asterisks.

Results

Influx of neutrophils into draining lymph nodes after immunization

Administering the proteins in any of the three adjuvants (complete or incomplete Freund's, or alum) resulted in neutrophils appearing in the draining lymph node already by fifteen minutes after immunization, reaching their peak number within an hour, and then decreasing. By 24 hours their presence was barely over the normal number found in lymph nodes. The presence of neutrophils at the early times was readily documented by standard histological sections. The majority of neutrophils accumulated in the cortical sinus and in the superficial cortex of the draining lymph nodes, and not in the distal lymph nodes. Fig. 1A shows results of immunization with HEL in complete Freund's adjuvant examining the lymph nodes by immunofluorescence technique. HEL, the antigen most extensively examined, accumulated in the cortical sinus, mainly visible in the macrophages lining the sinus (Fig. 1B). The presence of neutrophils in the cortical sinuses indicated that they entered by way of afferent lymphatics, pointing to a pathway from blood to the site of adjuvant injection to the lymphatics and into the draining node.

Figure 1. Influx of neutrophils into draining lymph nodes after immunization.

B10.BR mice were immunized with 10nmole HEL (A, C–E) or Hilyte Fluor 555-labeled HEL (B) in complete Freund's adjuvant. Popliteal lymph nodes were harvested at the indicated time. (A, D) 5μm frozen sections were stained with anti-CD3 (red) and anti-Gr-1(green) for epifluorescent microscopy (A); anti-CD11c (blue) and anti-Gr-1 (green) for confocal microscopy (D). Shown are representative from 8 individual experiments using 2 mice for each time point. Histological analysis disclosed the presence of typical neutrophils as described above. (C) Flow cytometric analysis of CD11b⁺ Ly-6G⁺ cells in popliteal lymph nodes at indicated time points was done after 10nmole HEL immunization. Data is a representative result taken from four experiments. Statistical analysis was processed by Mann-Whitney test. (E) LysM-eGFP/CD11c-eYFP mice were immunized with 10nmole HEL/IFA in the footpad. Popliteal lymph nodes were removed for two-photon microscopy. Contacts between DC (green) and neutrophils (blue) were visualized in the cortical sinus at 3 hours post-immunization. Scale bars: 200μm (A, B); 30μm (D); 150μm (E, left); 30μm (E, right).

Quantitation by flow cytometric analysis also confirmed an early wave of neutrophil migration in the draining lymph nodes (Fig. 1C). (The extent of infiltration was about 40–50% less when immunizing with incomplete Freund's adjuvant or alum, compared to complete Freund's adjuvant (data not shown).) By confocal microscopy, during 2 hours after immunization, many neutrophils contacted and surrounded DC (Fig. 1D).

The dynamics of neutrophil and DC in intact lymph nodes was examined in the LysM-eGFP/CD11c-eYFP dual mice (22, 23) by two-photon microscopy. In the immunized lymph node at the cortical sinus, but much less in the underlying superficial cortex, the neutrophils and DC were in contact during the first two hours post immunization (Fig. 1E and video 1). In these areas multiple neutrophils swarmed around a single DC and made primarily brief cell contacts, frequently moving from one DC to another within a limited time of contact (video 2), ranging from 1–3 minutes. (These findings of neutrophils swarming on DC are similar to those from Robey's group when examining lymph nodes after infection with Toxoplasma gondii (11).)

T cell and antibody responses after neutrophil depletion

The effects of the early neutrophil infiltrate were examined in mice depleted of neutrophils with antibodies, or in granulocyte colony-stimulating factor receptor (G-CSFR) gene deficient mice, which are neutropenic. The neutrophil depleting antibodies, either RB6-8C5 or 1A8, was administered to mice immunized with the proteins in adjuvants. Intraperitoneal injection with either of the antibodies resulted in ~90% depletion of neutrophils, but did not affect other cell types from peripheral blood or popliteal lymph nodes (data not shown). The monoclonal antibody RB6-8C5 reacts primarily with Ly6G⁺ cells, i.e. neutrophils, but also has weak reactivity with Ly6C, found in some monocytes (26, 27). The antibody 1A8 is Ly6G specific (28). Results with either antibody were identical. Most experimental manipulations compared both antibodies.

Many of the experiments examined the response to HEL: at different doses, depletion of neutrophils consistently enhanced by 100% or more the CD4 T cell responses measured by IL-2 or IFN-γ ELISPOT. (Fig. 2A summarizes the various experiment; most of the experiments show the IL-2 response.) The increase in the CD4 T cell response was found when HEL was incorporated in complete Freund's adjuvant, i.e. water in oil with dead Mycobacteria, as in Fig. 2B; or when given just in water in oil (incomplete Freund's adjuvant) or in alum (Figs. 2C and 2D, respectively). Enhancement was also found assaying the proliferation of lymphocytes to HEL (data not shown).

Figure 2. T cell responses after neutrophil depletion.

(A) Fold increase of HEL recalls response from neutrophil-depleted and control Ig-treated mice at different doses of HEL immunization in B10.BR mice. Data were obtained from a total of 31 independent experiments using 4 mice in each group. (B–D) T cell response of control Ig or neutrophil depleted Ab treated mice was measured by IL-2 ELISPOT at day 7 after 10nmole HEL immunization with complete Freund's adjuvant (HEL/CFA) (B), incomplete Freund's adjuvant (HEL/IFA) (C), or alum (D) (HEL/alum). Shown are representative results from at least 10 separate experiments. Data was presented as mean ± SD. (E–H) C57BL/6 or C.B-17 mice were immunized as indicated with OVA/CFA (E, F) or LLO/IFA (G, H), respectively, for 7 days in the presence or absence of neutrophils. Figures show the ratio of the recalls response by ELISPOT in neutrophil depleted and control mice from all experiments (E, G), or representative experiments (F, H). CD4 epitopes: 323–339 (OVA), 91–99 (LLO); CD8 epitopes: 257–264 (OVA), 188–201 (LLO). Data analysis was processed by Mann-Whitney test; indicated by an asterisk are the sets with values of *p<0.05; in the various comparisons the differences ranged between p<0.03 to p< 0.001.

The increase in CD4 T cell responses was found among the major HEL peptides previously identified (Fig. 2): in mice expressing the class II MHC molecule I-A^k, the CD4 T cell response to HEL is equally focused, mainly on four peptides expressed in very different amounts by the antigen presenting cells (APC) (24, 29, 30). For example, the chemically dominant peptide centered on the 52–60 sequence was highly expressed, about 200 fold more than a minor peptide centered on the 20–35 sequence, yet the number of T cells was the same for each. In seeking an explanation to this result, we considered whether neutrophils could affect the relative response to the different epitopes. This was not the case, as all responses were increased about the same extent after neutrophil depletion. Analysis of a pool of 25 experiments showed the increase in the T cell response to each epitope after neutrophil depletion to be: 2.9, 2.7, 3.2, and 2.3 fold increases for HEL peptides 48–62, 20–35, 114–129 and 31–47, respectively. Peptides HEL:70–88 and HEL:97–112, which are very weak epitopes, increased two fold after neutrophil depletion (data not shown). In two experiments mice immunized with the HEL:48–61 peptide also had a 2 and 2.5 fold increase in the ELISPOT response after neutrophil depletion. This result indicates that the enhancement of the HEL response in neutropenic mice was not related to an effect of neutrophils on HEL processing. The involvement of neutrophils in CD4 T cell priming also reflected negatively on the memory response: mice were treated with the neutrophil depleting antibody, immunized with HEL, and rested for two weeks, after which a boost was given of HEL in incomplete Freund's adjuvant and the ELISPOT response was evaluated at the third week of immunization. In two experiments the response of the antibody-treated mice was increased by 150% and 200% over the control mice.

The impact of neutrophils was evaluated in the response to OVA and LLO, tested by ELISPOTS on CD4 or CD8 epitopes. In OVA immunization, in vitro re-stimulation with the OT-II epitope OVA:323–339 resulted in enhanced CD4 T cell responses after neutrophil depletion in all six experiments. However, the extent of recall responses against the OT-I epitope OVA:257–264 (SIINFEKL) remained unaffected (in 3 of 6 experiments) or reduced slightly by ~20–30% (in 3 of 6 experiments) compared to the control Ig treated mice (Fig. 2E, F). Similar results were shown in LLO immunization in C.B-17 mice. Both IL-2 and IFN-γ production by CD4 T cells were significantly enhanced after neutrophil depletion in the six experiments; however, the CD8 T cell response was only enhanced in one, while the remaining five showed a slight reduction (Fig. 2G, H).

In addition to examining mice depleted of neutrophils by antibodies, G-CSFR-deficient mice also were tested in their response to HEL. G-CSF is critical in stimulating granulopoiesis (31, 32) mediated through interaction with its receptor (G-CSFR) (33). Decreased levels of circulating neutrophils are found in G-CSFR-deficient mice (34). A higher CD4 T cell response to HEL was found in G-CSFR-deficient mice, compared to wild-type mice treated with control Ig or neutrophil depleting antibody. As would be expected, neutrophil depleting antibody to G-CSFR-deficient mice did not result in further enhancement, an important result indicating that in their absence, the neutrophil depleting antibody had no effect (Fig. 3A). Thus, examining G-CSFR deficient mice confirmed the findings of an enhanced CD4 T cell response by neutrophil depletion with antibody.

Figure 3. Neutrophil effect in mice deficient in G-CSFR, ROS, NO^˙ or B cells.

(A) G-CSFR^−/−, (D) NOX2^−/−, iNOS^−/− or (E) μMT mice were immunized with 10nmole HEL in complete Freund's adjuvant, groups were treated with either control Ig or neutrophil- depleting antibody, as indicated. (B) B10.BR mice were injected with either control IgG or with neutrophil-depleting antibody one day before immunization with 10nmoles of HEL in complete Freund's adjuvant. Popliteal lymph nodes were harvested for single cell suspension and the amount of neutrophils (CD11b⁺ Ly-6G⁺) was determined by FACS analysis. (n=4 in each group) (C) B10.BR mice were injected with neutrophil-depleting antibody at the indicated day. The number of cells producing IL-2 was determined by ELISPOT at day 7 after immunization. The data represents mean ± SD for a representative experiment from five individual experiments. Data analysis was processed by Mann-Whitney test; *p<0.05.

In addition to the first early wave of neutrophil influx in the draining lymph nodes after immunization with adjuvant, there was a second wave of neutrophils at day 3 of immunization (Fig. 3B). However, the enhancing effect of neutrophil depletion was restricted to the first twenty-four hours: their depletion by antibodies after the first day of immunization did not change the magnitude of the response (Fig. 3C). Thus the negative effect of neutrophils was limited to the initial period of time after immunization when antigen capture and presentation events are prominent.

Whether the immunosuppression by neutrophils resulted from the release of reactive oxygen species (ROS) or nitric oxide (NO^˙) by the neutrophils was evaluated. ROS and NO^˙ might diffuse into the neighboring T cells or APC and affect their response (21, 35). NADPH oxidase- and iNOS-deficient mice depleted of neutrophils still showed an enhanced T cell response (Fig. 3D), implying that neither ROS nor NO^˙ were involved in neutrophil suppression of T cell responses. Enhancement of the response was also found in μMT mice, indicating that B cells were not involved (Fig. 3E). In experiments not shown we discarded the involvement of IL-10 since the enhanced HEL response after neutrophil depletion was not affected in IL-10 receptor-deficient mice or by neutralization by antibodies to IL-10 receptor.

The antibody response to HEL was examined after neutrophil depletion. The peak time of the response was at the second week after immunization. An increase in titer was found after neutrophil depletion in all of the three experiments (Fig. 4A, B and C). The isotypes of serum antibodies were IgG1, IgG2b and IgG3 in either treated or control untreated mice. In unpublished experiments we had found that most of the anti-HEL response was directed to conformational determinants of HEL, while the antibody response to HEL peptides was weak or not detectable. In the case of HEL immunized mice not treated with the anti-neutrophil antibodies, there was a very weak response to the chemically- dominant 48–62 peptide, although consistently all mice responded to the 114–129 peptide (Fig. 4D, E, F). The responses to the 48–62 and 114–129 peptides were enhanced in the neutrophil-depleted mice. In only one of three mice did we find a small response to the 20–35 and 31–47 peptides in neutrophil-depleted mice (Fig. 4F).

Figure 4. Antibody responses after neutrophil depletion.

Control Ig or neutrophil-depleting Ab treated B10.BR mice were immunized with 10nmole of HEL in the various adjuvants as indicated in panels A, B and C. At day 14 of immunization, sera were collected for ELISA analysis of antibody response. D, E, F show anti peptide responses from 3 individual experiments (n=8–10). Asterisks indicate that the data analysis processed by Mann-Whitney test was significant at values of *p<0.05. Sera from the various groups were pooled and titrated against an anti-HEL standard; the titer in μg/ml of sera in control vs. anti-neutrophil antibody treated mice were: in panel A: 9.9 vs. 57.6; in panel B: 16.3 vs. 64.2; and in panel C: 3.7 vs. 55.5.

Antigen presentation after neutrophil depletion

Competition for HEL between neutrophils and the APC could explain the enhancement of T cell responses by neutrophil depletion. Antigen presentation was examined in lymph node APC from untreated and neutrophil-depleted mice at different time points after immunization with HEL. The cells were isolated, fixed and then tested for their content of HEL peptide-MHC complexes by the response of a T cell hybridoma (3A9) to the dominant epitope 48–62. The response of this T cell hybridoma depended on the amounts of 48–62 −I-A^k complexes in the APC, independent of costimulatory signals (36). Antigen presentation from whole lymph node cells harvested early, at 3 hours after immunization, was low in both neutrophil-depleted and undepleted lymph nodes (Fig. 5A). At 16 and 24 hours, presentation by APC was evident: it was markedly increased in the neutrophil depleted mice compared to the control mice (Fig. 5B, C).

Figure 5. Antigen presentation after neutrophil depletion.

B10.BR mice were immunized with 10nmole HEL in complete Freund's adjuvant after injection of control Ig or neutrophil depleting antibody. Popliteal lymph nodes were harvested at 3 hours (A), 16 hours (B), and 24 hours (C–F) after immunization. Single cell suspension from whole lymph nodes (A–C) or indicated cell subsets separated by MACS (D–F) were fixed with 1% paraformaldehyde and tested on the T cell response of the 3A9 hybridoma. (G, H) Mice were immunized with ¹²⁵I-HEL in complete Freund's adjuvant for 16 hours. Indicated cell populations were separated by MACS for autoradiography analysis. Cells containing HEL from each are shown. Scale bars: 50μm. Data were obtained from 3–6 individual experiments, and shown is one representative result presented as mean ± SD (n=5–10). Statistical analysis *p<0.05.

The increase in presentation after neutrophil depletion occurred in both CD11c⁺ and CD11c⁻ cells. In control Ig treated mice, the major APC for HEL were CD11c⁺ cells (Fig. 5D), in agreement with our previous findings (37). The relative enhancement was particularly evident among the CD11c⁻ cells (Fig. 5D) and, specifically, by the macrophages (CD11c⁻ CD19⁻) (Fig. 5E). Antigen presentation in B cells was low in both neutrophil-depleted and -undepleted mice (Fig. 5F). Together these results indicate that there was a reduction in the content of the HEL peptide-MHC complex by APC when neutrophils infiltrate the lymph nodes.

Examination of the number of cells in the lymph nodes bearing HEL confirmed the functional results. About 4% of isolated DC contained HEL by autoradiography examinations when I¹²⁵-HEL was administered to the mice with complete Freund's adjuvant (Fig. 5G, H). This number about doubled in the neutrophil depleted mice. The same findings were found in CD11b⁺ cells isolated from the nodes. However, the number of autoradiograph grains per APC were similar in both neutrophil-depleted and control groups. The auto radioactivity count of DC increased by 2.4 fold after neutrophil depletion (627 cpm/10⁶ cells in control Ig-treated group versus 1488 cpm/106 cells in neutrophil-depleted group). In macrophages the auto radioactivity count increased 4.2 fold (465 cpm/10⁶ cells in control Ig group versus 1953 cpm/10⁶ cells in the neutrophil-depleted group). In control mice, 10% of the neutrophils contained HEL (Fig. 5H).

Effect of neutrophils on antigen bearing APC

Aside from reducing the amount of HEL available to the APC, neutrophils could also interfere with the presentation by APC already bearing HEL. To functionally test if neutrophils were affecting antigen presentation by antigen-bearing cells, DC or macrophages from transgenic mice in which HEL was expressed on cell membrane (mHEL) (24), were transferred into neutrophil-depleted or -undepleted mice. The CD4 T cell response was evaluated in conditions where neutrophil migration into lymph node was induced by injection of incomplete Freund's adjuvant (without HEL). At day 7 of cell transfer, mHEL DC or macrophages elicited a CD4 T cell response in control Ig-treated recipient mice. In neutrophil-depleted mice, mHEL DC or macrophage transfer resulted in about 100% fold enhanced CD4 T cell response (Fig. 6A–B). These results demonstrated that neutrophils had a negative effect also by affecting presentation of antigen bearing DC and macrophages.

Figure 6. Effect of neutrophils on antigen bearing APC.

Splenic DC or macrophages were isolated from mHEL or B10.BR mice by MACS and transferred into footpad of neutrophil depleted- or undepleted-B10.BR recipients. At 3 hours after cell transfer, incomplete Freund's adjuvant was injected into the footpad to induce neutrophil migration. CD4 T cell response was measured by IL-2 ELISPOT analysis at day 7 of cell transfer. Shown is representative from seven individual experiments presented as mean ± SD (n=8). Data analysis was processed by Mann-Whitney test; *p<0.05.

Dynamic imaging of T cell-DC interactions in the absence of neutrophils

Examining lymph nodes by two-photon microscopy substantiated these results. In these experiments, LysM-eGFP/CD11c-eYFP dual mice were adoptively transferred with labeled transgenic CD4 T cells recognizing the dominant 48–62 epitope of HEL (3A9) and immunized with HEL. Pilot experiments established that 6–10 hrs later interactions between the T cells and the DC were prominent. This was a time subsequent to the fast influx of neutrophils. DC and T cell interaction was observed in the cortical areas of draining lymph nodes, just below the cortical sinus (Fig. 7A). At these times, in the control mice not given the neutrophil depleting antibodies, the neutrophils were rarely found next to or attached to the DC or to the DC-T cell clusters. The DC-T cell contacts were found in both treated and untreated mice; however, they were markedly improved in the neutrophil-depleted mice. T cell velocity was similar in both control Ig and neutrophil depleting antibody-treated mice (Fig. 7E); however, T cells showed more biased movement in the absence of neutrophils (Fig. 7F). The number of DC-T cell contacts increased ~150% after neutrophil depletion (Fig. 7C). The cell contact period also increased (Fig. 7B, D). Together these results indicated improved DC-T cell interactions after neutrophil depletion, surmising that in the normal mouse, neutrophils affect or imprint the DC-T cell interaction without having a direct physical blockade of it.

Figure 7. Improved DC-T cell interaction in the absence of neutrophils.

LysM-eGFP/CD11c-eYFP mice were adoptively transferred with CMTPX-labeled 3A9 transgenic T cells and immunized with 10nmoles HEL/IFA in the footpad. Popliteal lymph nodes were removed for two-photon microscopy. (A, B) Contacts between DC (green) and T cells (red) were visualized in the cortex area 6–10 hours post-immunization. Arrows indicate the areas of DC-T cell contact. Scale bars: 20μm (A); 10μm (B). Statistical analysis of 3–5 videos acquired from individual mice (C–F). Non-draining lymph nodes were used as the no antigen control. Three dimensional tracking of T cells was performed using Imaris software (E, F). Cell contact analysis was performed manually (C, D). Shown was representative (A, B) or pooled data (C–F) from 3 individual experiments (n=6–10). Data analysis was processed by using the Mann-Whitney test and presented as mean ± SD (C–F).

Discussion

In brief, in a conventional in vivo immunization system a strong negative role of neutrophils was found in the adaptive CD4 T cell and B cell responses. The negative effect of neutrophils was shown by depleting with either of two antibodies or in neutropenic mice lacking the G-CSFR. Very rapidly after immunization, neutrophils migrated into the draining lymph nodes, by all indications, through the lymphatics draining the site of adjuvant inoculation. The very fast migration suggests the recruitment of neutrophils from blood and the marginal pool, but the mechanisms involved and the inciting molecules need identification. The point is that the neutrophil early response was an obligate component of the inflammation triggered by adjuvants, including not only the water in oil preparations but also those with alum.

The immunosuppressive effect of neutrophils was restricted to the first twenty-four hour entry, pointing to an effect on presentation of the antigen. First, it was clear that the level of HEL entering the APC system was in part controlled by neutrophils, pointing to a competition between them and the APC for HEL. In the absence of neutrophils, more APC contained HEL, correlating with enhanced presentation. The dampening effect of neutrophil was more marked in the macrophages, perhaps because of anatomical considerations in that the major site of accumulation of neutrophils was in the cortical sinus where macrophages took up the entering antigen (38, 39).

Aside from antigen competition, neutrophils affected presentation by both DC and macrophages already having peptide-MHC complexes. Notably, in different experimental setups, T cell contacts with DC in the superficial cortex were enhanced in the neutrophil-depleted mice, yet there was no evidence of a physical interaction with neutrophils that could have blocked the interaction. Neutrophils mostly surrounded DC in the cortical sinus and superficial cortex during the first hours following immunization (11). (In culture, evidence was provided of DC-neutrophil interactions mediated by SIGN-R1 (18).) These findings suggest that neutrophils leave an imprint on the DC-lymphocyte interaction, the nature of which we are now examining. This result could have one or more explanations. There could be a reduction in key antigen-presenting molecules resulting from the initial interactions between DC and neutrophils that preceded those between DC and T cells, a mechanism such as trogocytosis could be involved. In the cortical sinus the contacts of neutrophils likely involved migratory DC moving in from the immunization site. In the superficial cortex an involvement with resident DC should also be taken into account. Whether the resident DC or migratory DC is more affected by neutrophils was not evaluated. To note are the results in Fig. 6, of experiments transferring HEL-bearing APC which indicated a negative effect by neutrophils. We are also considering an effect on the biology of DC by neutrophil-released molecules, and have at this time eliminated ROS, nitric oxide and IL-10. Lipid mediators or other cytokines need to be evaluated. Whether the content of HEL per DC explains the result is unlikely based on the autoradiograph experiment, which indicated more antigen-bearing APC but about the same content of HEL per cell.

Other negative effects by neutrophils aside from those on antigen presentation are not evident from these results. The findings that there was a narrow time limit in which neutrophils participated, that is, the first hours, suggest that late effects such as those involving neutrophil apoptosis and their uptake by phagocytes are not major in the framework of these responses (40). Whether there is redistribution of other cells in the lymph nodes in the absence of neutrophils (41) is under evaluation.

Finally, to note is that the CD8 T cell response was not enhanced or was affected to a small degree, confirming that the presentation of soluble antigens in vivo for CD4 and CD8 T cells involve distinct cellular intermediates and pathways. Evidence has been presented that protein antigens in neutrophils can eventually enter the class I processing pathways, most likely via cross presentation following their apoptotic death (10, 42–45). Some of these studies include responses in vivo (10, 44, 45). In sum, placing all these findings in the context of vaccination suggest that a brief depletion of neutrophils could have a profound effect on improving CD4 and antibody responses, while having a small negative impact on CD8 T cells.