ABSTRACT
Dendritic cells (DCs) are potent antigen-presenting cells involved in the initiation and modulation of immune responses after immunization via their ability to process and present antigen to naive T cells. We wanted to examine the role of DCs in the development of protective immunity against nontypeable Haemophilus influenzae (NTHI)-induced experimental otitis media (OM) after intranasal immunization of chinchillas with an NTHI P5-derived synthetic peptide immunogen called LB1. As chinchilla DCs have not been described, we adapted well-established protocols to induce the differentiation of chinchilla bone marrow precursor cells into DCs, which resulted in cells that were morphologically and phenotypically similar to DCs of other species. In vitro, chinchilla DCs readily internalized LB1, upregulated expression of the maturation markers CD80 and major histocompatibility complex class II, and presented processed LB1 to primed CD3+ T cells, which resulted in antigen-specific T-cell proliferation. In vivo, LB1-activated DCs trafficked from the chinchilla nasal cavity primarily to the nasal-associated lymphoid tissues and were detected in close proximity to CD3+ T cells within this lymphoid aggregate. These data are the first to characterize chinchilla DCs and their functional properties. Furthermore, they suggest an important role for chinchilla DCs in the development of protective immunity against experimental NTHI-induced OM after intranasal immunization.
Nontypeable Haemophilus influenzae (NTHI) is a commensal of the human nasopharyngeal flora yet is also responsible for diseases of the upper and lower airway, including otitis media (OM), sinusitis, bronchitis, and exacerbations of chronic obstructive pulmonary disease (17, 50). Adherence of the bacterium at mucosal sites is a critical first step toward NTHI pathogenesis, and as a result, there is great interest in the development of vaccines that target NTHI surface-exposed adhesins (13, 21, 51, 56). The outer membrane protein (OMP) P5-homologous adhesin (OMP P5) is one of several adhesins expressed by NTHI and has been shown to facilitate the adherence of the bacterium to human oropharyngeal cells (9) and nasopharyngeal mucin (57), Chinchilla Eustachian tube mucin (47), respiratory syncytial virus-infected A549 cells (35), carcinoembryonic antigen-related cell adhesion molecule 1-transfected HeLa (11) or CHO (26) cells, and human intercellular adhesion molecule 1 (4).
A synthetic chimeric peptide immunogen, called LB1, was derived from a surface-exposed region of OMP P5 (8). In rat and chinchilla models of OM, LB1 has demonstrated significant protective efficacy against both homologous and heterologous NTHI challenge when delivered parenterally (39, 42). Furthermore, significant efficacy was observed when LB1 was administered intranasally (i.n.) to chinchillas, despite the induction of relatively low levels of serum and mucosal antibodies compared to those induced by immunization via a parenteral route (28, 53). It has been shown that antigen processing by dendritic cells (DCs) can influence vaccine efficacy (14, 16), as upon migration to lymphoid tissues, DCs present antigen in the context of major histocompatibility complex (MHC) molecules to naive T cells, serving to either initiate an immune response or induce T-cell tolerance (10, 59, 67, 73). Therefore, characteristics of the antigen itself and its processing by the DCs could influence the outcome of this interaction (immune induction versus tolerance).
Recently, it was reported that recombinant OmpA proteins from Escherichia coli O157:H7, Klebsiella pneumoniae, and Acinetobacter baumannii interact with and induce maturation of DCs, an observation that resulted in the designation of the OmpA family of bacterial proteins as a new pathogen-associated molecular pattern (PAMP) (32, 33, 43, 71). As OMP P5 is a member of the OmpA family of proteins (64), we also wanted to determine whether LB1, which is derived from OMP P5, showed properties similar to those of a PAMP (i.e., the ability to interact with and induce the maturation of DCs), among others.
Here, we have identified the optimal conditions under which to culture chinchilla bone marrow-derived precursor cells to promote their differentiation into DCs and have characterized the differential uptake and presentation of LB1 by these cells in vitro. Additionally, we began to identify sites of immune induction by tracking the migration of DCs to lymphoid tissues in vivo. To do so, we labeled DCs resident within the chinchilla nasal cavity by i.n. delivery of the dye carboxyfluoroscein succinimidyl ester (CFSE), followed by i.n. administration of LB1. We subsequently confirmed our observations by specifically tracking LB1-activated, bone marrow-derived DCs that had been labeled with CFSE in vitro prior to their i.n. administration. These data are the first to characterize chinchilla DCs. Moreover, our data demonstrated a role for DCs in the processing and presentation of the peptide immunogen LB1 when delivered to the chinchilla nasal cavity, which likely underlies the observed preclinical vaccine efficacy obtained upon i.n. immunization.
MATERIALS AND METHODS
Animals.Adult chinchillas (Chinchilla lanigera) were obtained from Rauscher's Chinchilla Ranch (LaRue, OH). Animal care and all related procedures were performed in accordance with institutional and federal guidelines and were conducted under an IACUC-approved protocol.
Generation of chinchilla DCs.Bone marrow cells were isolated from the femurs of adult chinchillas and cultured in RPMI 1640 medium containing 2 mM l-glutamine, 100 U penicillin/ml, 100 μg streptomycin/ml, 1 mM pyruvate (Mediatech), 5 mM HEPES (Acros), 0.75 g NaHCO3/ml (Fisher Scientific), and 0.05 mM β-mercaptoethanol (Sigma). The DCs were induced by culturing 2 × 106 bone marrow cells in 8 ml medium supplemented with 40 ng recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF)/ml and 40 ng recombinant human interleukin 4 (IL-4)/ml (R&D Systems) in 100-mm tissue culture dishes. The medium was also supplemented with either 5% fetal bovine serum (FBS) (Mediatech) or 5% naive chinchilla serum. On day 4 of culture, an additional 4 ml of fresh medium was added to the cultures, and on day 6, nonadherent cells were collected for assay as immature cells.
Antibodies and antigens.We used the following fluorochrome-conjugated, commercially available antisera to detect cell surface molecules expressed by chinchilla bone marrow-derived DCs: mouse anti-rat CD11c-fluorescein isothiocyanate (FITC), mouse anti-rat CD80-phycoerythrin, and mouse anti-guinea pig MHC class II revealed with goat anti-mouse immunoglobulin G (IgG)-phycoerythrin (AbD Serotec). Appropriate isotype-matched antibody controls were included in all assays. To induce the maturation of chinchilla DC-like cells in vitro, the cells were stimulated with either 10 ng E. coli lipopolysaccharide (LPS)/ml (Sigma) or 10 ng recombinant human tumor necrosis factor alpha (TNF-α)/ml (Sigma) for 24 h prior to labeling the cells for flow cytometry.
Native P5 protein is an approximately 36-kDa OMP and was isolated and purified from NTHI strain 86-028NP, as previously described (64). LB1 is a 4.6-kDa 40-mer synthetic chimeric peptide vaccinogen derived from a surface-exposed loop of OMP P5 and has also been described (8). To manage potential differences in the binding and uptake of antigens by the chinchilla DCs due to the substantial difference in the relative molecular masses of OMP P5 and LB1, a 26-kDa NTHI protein (41) called “control protein” and a 3.5-kDa 34-mer synthetic peptide called “TFPQ3” (1, 2, 5) were used as size control antigens. These “control” antigens were selected based exclusively on molecular mass and not on potential known immunoreactivity or lack thereof. Endotoxin contamination of the proteins and synthetic peptides was determined to be ≤0.25 endotoxin units/ml by Pyrosate agglutination assay (Associates of Cape Cod, Inc.). Peptides and proteins were coupled to FITC with a fluorescein protein-labeling kit according to the manufacturer's instructions (Pierce Biotechnology). Based on the coupling chemistries and amino acid compositions of LB1 and TFPQ3, the peptides were estimated to be comparably labeled with FITC.
Chemotaxis.To demonstrate the ability of immature LB1-activated or TFPQ3-activated chinchilla DCs to sense and respond to a stimulus, a chemotaxis assay was performed. Immature DCs were induced to mature by incubation overnight with 1 μg of either TFPQ3 or LB1/ml culture medium supplemented with 5% naive chinchilla serum, but without cytokines. The cells were washed and adjusted to a concentration of 106 DCs/ml, and 12 μl of unstimulated or stimulated DCs was added to one of two horizontally positioned compartments that were separated by a 260-μm gradient channel in a TAXIScan instrument (Effector Cell Institute, Inc.) (38, 52). To align the DCs with the edge of the channel, buffer was aspirated from the second compartment. A volume of 15 μl of 1 μM recombinant human C5a (Calbiochem) or buffer was then added to the second compartment. After 60 min, the cumulative number of DCs that had entered the chemotactic gradient in the channel toward C5a or buffer was calculated for each of three independent experiments.
Antigen binding and internalization by chinchilla DCs.Immature DCs (1 × 105) were incubated with medium containing 1 μg FITC-labeled peptide or protein/ml for 1 h at 4°C (to assay surface binding of antigen) prior to shifting the cultures to 37°C (to assay internalization of the bound antigen). The cells were washed, fixed with 2% paraformaldehyde in 0.1 M phosphate buffer, and assayed immediately by flow cytometry with a BD FACSCalibur. A total of 10,000 viable events were collected for three independent assays, and the data were analyzed with FloJo software (Tree Star, Inc.).
Antigen endocytosis.Immature DCs were mixed with 1 μg FITC-labeled peptide or protein/ml medium and allowed to adhere to poly-l-lysine (Sigma)-coated coverslips for 30 min at 37°C. Alternatively, to block endocytosis, immature DCs were preincubated with 10 nM cytochalasin D (Sigma) for 20 min prior to and during incubation with FITC-labeled peptides. The cells were then fixed, and nonspecific binding sites were blocked with either 2% bovine serum albumin in Dulbecco's phosphate-buffered saline (DPBS) (Mediatech) or 10% normal goat serum in DPBS (Zymed). To label F-actin, cells were incubated with rhodamine-phalloidin (Molecular Probes). To label MHC class II molecules, DCs were incubated with mouse anti-guinea pig MHC class II (AbD Serotec) and detected with goat anti-mouse IgG conjugated to Texas Red (Invitrogen). To inhibit endosome acidification and MHC class II antigen presentation, cells were treated with 10 nM choloroquine (MP Biologicals) for 20 min prior to and during incubation with FITC-labeled peptides. An appropriate isotype-matched antibody control was included in each assay. To inhibit fluorescence quenching, Prolong AntiFade Gold (Invitrogen) was added to each slide prior to sealing it. The slides were examined with a Zeiss LSM 510 Meta confocal system attached to a Zeiss Axiovert 200 inverted microscope (Carl Zeiss, Inc.).
Antigen presentation by chinchilla DCs.For immunization, three doses of 10 μg LB1 plus 10 μg monophosphoryl lipid A (MPL) (Corixa) or 10 μg MPL alone were administered to alert chinchillas in a volume of 100 μl via subcutaneous injection at weekly intervals. One week after receipt of the third dose, chinchilla splenocytes were isolated by crushing the spleen and passing the cells through a 40-μm cell strainer. Erythrocytes were removed, and the suspensions were enriched for viable cells by centrifugation with Ficoll Paque (GE Healthcare). The resultant cells were labeled with mouse anti-rat CD3 conjugated to biotin (AbD Serotec) for enrichment of CD3-positive (CD3+) T cells by streptavidin-conjugated magnetic-bead separation (Miltenyi). CD3+ T cells were then cultured in 100-mm2 petri dishes in culture medium supplemented with 5% naive chinchilla serum, but without cytokines. As chinchillas are outbred and thus are not genetically identical animals, we also cultured bone marrow cells from the same animal for 3 days to allow differentiation of precursor cells into DCs in order to demonstrate that the observed T-cell proliferation was due to the presentation of antigen by DCs and not to an allogeneic MHC molecule response. After 3 days in culture, the CD3+ T cells were seeded into wells of 96-well round-bottom plates at a density of 2 × 105 cells per well in triplicate. The autologous bone marrow-derived DCs were incubated with 1 μg or 10 μg LB1 or control peptide/ml for 2 h to allow the uptake of antigen, washed, and then added to wells that contained the CD3+ T cells at a density of 104 cells per well in medium without cytokines. Controls consisted of wells that contained 1 μg or 10 μg concanavalin A/ml (Sigma) plus CD3+ T cells and DCs, wells without an antigen (unstimulated), or wells with CD3+ T cells only. After 72 h, 20 μl of Alamar blue (AbD Serotec) was added to each well and the cultures were incubated for an additional 24 h. Proliferation was assessed as the chemical reduction of Alamar blue by the cells in culture and was detected as the fluorescence of each well at 530-nm excitation and 590-nm emission wavelengths with a Bio-Tek Synergy HT microplate reader. The statistical significance of three independent assays was determined by Tukey's honestly significant difference test for analysis of variance pairwise comparisons.
In vivo migration of chinchilla respiratory DCs following i.n. immunization.To track the migration of chinchilla nasal mucosa DCs, we adapted the technique described by Legge and Braciale (44) for use in the chinchilla host. CFSE (Invitrogen) was dissolved to 100 μM in dimethyl sulfoxide and further diluted in pyrogen-free saline to 8 μM prior to administration via microaerosol sprayer (Wolfe Tory Medical, Inc.) in a volume of 40 μl/naris to alert, prone chinchillas. Fluid administered via this regimen is retained exclusively within the chinchilla nasal cavity, with no dose loss to the respiratory or gastrointestinal tract (72). Six hours later, the chinchillas received either 10 μg LB1 plus 10 μg MPL or 10 μg MPL alone via microaerosol spray as before. One, 3, and 12 h later, animals were sacrificed and the cervical, brachial, axillary, and mediastinal lymph nodes were removed, in addition to the spleen. The nasal-associated lymphoid tissue (NALT) was also removed (37). Tissues were crushed into DPBS, and the cells were passed through a 40-μm cell strainer. The spleen was also crushed and strained, and the cells were centrifuged with Ficoll-Paque. DCs were labeled with mouse anti-rat CD11c-Alexa Fluor 647 (AbD Serotec), and CD11c+ CFSE+ cells were detected by flow cytometry. A total of 5,000 viable events were collected for each of three independent assays.
To investigate whether LB1-activated, bone marrow-derived DCs would similarly follow homing signals within the nasal cavity to the NALT, in vitro-derived DCs were labeled with 10 μM CFSE prior to incubation with either 10 μg LB1 plus 10 μg MPL or 10 μg MPL alone for 15 min. These treated DCs were washed to remove unbound peptide, adjusted to 106 cells/80 μl, and instilled dropwise by micropipette to alert, prone chinchillas in a volume of 40 μl per naris. The animals were sacrificed 1 h later, and the lymphoid tissues were processed as described above. The DCs were labeled with mouse anti-rat CD11c-Alexa Fluor 647, and CD11c+ CFSE+ cells were detected by flow cytometry. A total of 5,000 viable events were collected for each of three independent assays.
Aliquots of the NALT cell suspension retrieved 1 h after i.n. administration of LB1 plus MPL were incubated for 1 hour at 37°C and 5% CO2 on poly-l-lysine-coated coverslips. The cells were fixed, blocked with 10% normal goat serum, and incubated with mouse anti-rat CD11c conjugated to Alexa Fluor 647, mouse anti-guinea pig MHC class II revealed with goat anti-mouse IgG-conjugated to Texas Red, and mouse anti-rat CD3 conjugated to biotin revealed with streptavidin conjugated to Alexa Fluor 405 (Invitrogen). The slides were sealed after the addition of Prolong AntiFade Gold, with DAPI (4′,6′-diamidino-2-phenylindole) or without DAPI, prior to being viewed via confocal microscopy.
RESULTS
The cytokine and serum source influenced the differentiation of chinchilla bone marrow-derived cells in culture.To establish in vitro cultures of chinchilla DCs, we isolated bone marrow cells from the femurs of chinchillas and cultured these cells in medium supplemented with recombinant human cytokines (40 ng GM-CSF/ml and 40 ng IL-4/ml), which resulted in viable cultures. We further assessed whether the serum source in the medium influenced the development of these cells. After 1 day, cells in the medium supplemented with a homologous serum source (naive chinchilla serum) remained in suspension as individual cells (Fig. 1A) that, in time, developed into a mixture of floating and adherent cells (Fig. 1C). After 6 days, we observed numerous floating aggregates of cells (Fig. 1E) within this culture. Chinchilla bone marrow cells cultured in medium supplemented with FBS, a serum source widely used in cell culture systems, formed aggregates suggestive of rapidly proliferating cells that had attached to the bottom of the culture dish (Fig. 1B), which appeared larger after 4 days in culture (Fig. 1D) and finally dispersed after 6 days (Fig. 1F).
Influence of serum source on the survival and differentiation of chinchilla bone marrow-derived cells in culture. (A to F) Light microscopy images of chinchilla bone marrow-derived cells cultured in medium supplemented with 5% naive chinchilla serum (A, C, and E) or fetal bovine serum (B, D, and F). The arrows in panels E and F indicate floating clusters of cells. (G and H) Representative dot plots demonstrating forward scatter and side scatter profiles of bone marrow-derived cells cultured for 6 days in medium supplemented with either naive chinchilla serum (G) or FBS (H).
By flow cytometry, the viable cells cultured in medium supplemented with naive chinchilla serum were comparable in size to those within the medium containing FBS, as demonstrated by the forward scatter profile (Fig. 1G and H, respectively). However, the cells cultured in medium supplemented with chinchilla serum appeared more complex (as demonstrated by the side scatter profile) than cells within the medium containing FBS. This observation suggested differences in the inner complexity of the cells (i.e., the shape of the nucleus, the amount and type of cytoplasmic granules, or the membrane roughness). As the development of cells within the medium supplemented with naive chinchilla serum closely followed that reported for murine and rat bone marrow cultures (31, 70), we concluded that optimal medium supplements required to culture chinchilla bone marrow cells included the use of recombinant human cytokines and a homologous serum source.
Chinchilla bone marrow-derived cells exhibited classic DC morphology and phenotype after culture with GM-CSF and IL-4.In order to examine the expression of classical DC surface markers by the chinchilla DC-like cells, we tested antibodies raised against molecules expressed by other species, including human, mouse, and rat, for reactivity with the chinchilla bone marrow-derived cells, as no chinchilla-specific reagents are available commercially. Of the panel tested, we observed the greatest positive result with mouse anti-rat CD11c (Fig. 2A). It was not unprecedented that a rat-specific reagent recognized a chinchilla homologue, as we and others routinely use rat-specific (and human-specific) reagents for immunodetection in the chinchilla model (18, 29, 40, 53, 62). Greater than 75% of cells within each culture were CD11c+, as determined by flow cytometry, also comparable to reports of cells cultured from other species. We further examined the morphology and phenotype of the DC-like cells by inducing the maturation of the bone marrow-derived cells with 10 ng E. coli LPS/ml. Immature (unstimulated) cells in culture expressed fine processes that extended from the cell surface (Fig. 2B) and had a large, indented nucleus, which is commonly described for DCs (Fig. 2B, inset), whereas mature or LPS-stimulated DC-like cells expressed much larger dendrites (Fig. 2C and inset). CD11c+ DC-like cells stimulated with either LPS or recombinant human TNF-α upregulated expression of both MHC class II molecules (Fig. 2D) and CD80 (Fig. 2E) compared to immature cells. Therefore, as the chinchilla DC-like cells were both morphologically and phenotypically comparable to described murine and rat bone marrow-derived DCs, we were confident we had identified conditions to promote the differentiation of precursor cells within the chinchilla bone marrow into DCs.
Chinchilla bone marrow-derived cells exhibited DC-like phenotype and morphology. (A) Representative histogram demonstrating expression of CD11c by chinchilla bone marrow-derived cells as determined by flow cytometry. Black histogram, isotype control; red histogram, anti-CD11c. (B and C) Light microscopic images of immature (B) or LPS-matured (C) chinchilla bone marrow-derived cells in culture. The arrows indicate surface expression of fine processes by immature cells or of larger dendrites by LPS-matured DCs. (Insets) Morphology of immature (B) or LPS-matured (C) cells after being stained with Giemsa. (D and E) Histograms demonstrating expression of MHC class II (D) and CD80 (E) by immature chinchilla bone marrow-derived cells (black histograms), TNF-α-matured cells (red histograms), and LPS-matured cells (blue histograms) as determined by flow cytometry.
Chinchilla DCs exhibited chemotaxis.One of many functions of DCs is the ability to sense a chemokine gradient and to respond by directional movement. We therefore assessed chinchilla DC chemotaxis toward a classic chemotactic agonist, C5a, in vitro (48). As demonstrated by the pseudocolored images captured with the TAXIScan system after 30 min, a greater number of LB1-activated DCs were present within the channel compared to the DCs that had been activated with another synthetic peptide of similar mass (TFPQ3) (Fig. 3A and B). Furthermore, the chemotactic response was specifically directed toward compartments that contained C5a, as minimal spontaneous migration of cells was observed toward compartments that contained only buffer (Fig. 3C), regardless of the maturation status. When plotted as the cumulative number of DCs that entered the chemotactic gradient in the channel and were thus moving toward C5a, the number of TFPQ3-activated DCs was marginally greater than that of immature DCs. In contrast, LB1-stimulated DCs continually entered the chamber when C5a was present. These data demonstrated that DCs stimulated with LB1 were induced to mature and responded to the presence of a chemoattractant.
Chemotaxis of chinchilla DCs toward C5a. (A and B) Representative images captured with the TAXIScan instrument after 30 min of chemotaxis toward C5a. The pseudocolored images show TFPQ3-activated DCs (A) and LB1-activated DCs (B). (C) Plot demonstrating the cumulative number of DCs that entered the channel over time. The solid lines depict DC chemotaxis toward C5a, while the dashed lines demonstrate spontaneous migration of DCs in buffer. Red lines, LB1-activated DCs; green lines, TFPQ3-activated DCs; gray lines, immature/unstimulated DCs.
Chinchilla DCs preferentially bound and internalized LB1.To examine the relative abilities of immature chinchilla DCs to bind and internalize NTHI OMP P5 and LB1, we utilized a panel of both native proteins and synthetic peptides. FITC-labeled proteins or peptides were incubated with immature chinchilla DCs at 4°C to examine antigen surface binding, and then the cultures were shifted to 37°C to promote internalization of the bound antigen. At 4°C, all cultures demonstrated binding of the protein or peptide, although at various magnitudes (Fig. 4A). Whereas DCs incubated with FITC-labeled OMP P5 demonstrated greater binding than the 26-kDa control peptide, the culture of DCs incubated with FITC-labeled LB1 was the most fluorescent of all cultures, with approximately eightfold-greater relative fluorescence than DCs incubated with a peptide of similar molecular mass (TFPQ3).
Chinchilla DCs differentially bound and internalized FITC-labeled proteins and peptides. (A and B) Representative fluorescence-activated cell sorter histograms demonstrating the mean fluorescence of chinchilla DC cultures incubated at 4°C to assess cell surface binding of antigens (A), followed by a temperature shift to 37°C to assess internalization of the bound antigens (B). Representative histograms demonstrate the relative fluorescence of immature DCs (black histograms) and DCs incubated with FITC-labeled antigens: control protein (blue histograms), OMP P5 (purple histograms), TFPQ3 (green histograms), or LB1 (red histograms). (C to G) Confocal microscopy images demonstrate the localization of FITC-labeled proteins and peptides within chinchilla DCs (green fluorescence) after the temperature shift to 37°C. DCs were incubated with either control protein (C), OMP P5 (D), TFPQ3 (E), or LB1 (F). F-actin was labeled with rhodamine-phalloidin (red fluorescence). Note the yellow fluorescence in panel F, which demonstrated colocalization of FITC-labeled LB1 with actin, which was abrogated upon treatment with cytochalasin D (G).
Upon temperature shift to 37°C, the trend of relative fluorescence remained the same as that observed at 4°C; however, the mean fluorescence of each culture increased (Fig. 4B). At either temperature, DC cultures incubated with the peptides were more fluorescent than cultures incubated with either protein, which suggested that the smaller molecular size of the peptides than of the native proteins may facilitate antigen binding. Furthermore, the greater binding and uptake of LB1 than of the peptide of comparable size suggested a unique characteristic specific to that immunogen, for example, the presence of a PAMP-like motif.
To further investigate this theory and to assess whether the preferential binding of LB1 by the chinchilla DCs was also receptor mediated, we examined the localization of the FITC-labeled proteins and peptides within DCs by confocal microscopy after the temperature shift to 37°C. In agreement with the flow cytometry data, immature chinchilla DCs internalized the FITC-labeled proteins and peptides, although to differing magnitudes. Whereas DCs incubated with the FITC-labeled 26-kDa control protein (Fig. 4C) or FITC-labeled OMP P5 (Fig. 4D) contained a few areas of punctate green fluorescence, greater internal fluorescence was observed in DCs incubated with either FITC-labeled TFPQ3 (Fig. 4E) or FITC-labeled LB1 (Fig. 4F). The greatest internal fluorescence was clearly visible within the cells incubated with LB1, as numerous punctate signals were observed (Fig. 4F). Furthermore, colocalization of actin and LB1 was observed (Fig. 4F, yellow signal), which suggested that uptake of the FITC-labeled LB1 likely occurred via endocytosis. To confirm this, DCs were treated with cytochalasin D prior to and during incubation with FITC-labeled LB1 to inhibit actin polymerization and formation of endocytic vesicles. Whereas cytochalasin D treatment did not prevent binding of LB1 by the chinchilla DCs, as demonstrated by the punctate green signals in Fig. 4G, internalization of the peptide was inhibited, as evidenced by the fact that no colocalization of actin and LB1 was observed. These data suggested that binding and internalization of LB1 were potentially mediated by receptor-mediated endocytosis and, further, that a motif might be present within the peptide immunogen LB1 itself which facilitated this interaction.
DCs upregulated MHC class II molecules and induced T-cell proliferation upon activation.To more closely examine the specific interaction between the chinchilla DCs and LB1, we examined whether immature chinchilla DCs stimulated with either LB1 or a peptide of similar size subsequently upregulated the expression of MHC class II molecules. As detected by flow cytometry (Fig. 5A) and confocal microscopy (Fig. 5B), chinchilla DCs incubated with TFPQ3 were labeled with an MHC class II-specific antibody, although the DCs exhibited only relatively moderate upregulation of MHC class II molecules over unstimulated DCs (Fig. 5A). Upon incubation of DCs with LB1, however, notably greater MHC class II-specific fluorescence was observed (Fig. 5A and C). Treatment of the DCs with chloroquine to inhibit endosome acidification and MHC class II antigen presentation abrogated the MHC class II-specific fluorescent signal (Fig. 5D). These data further demonstrated that internalization of LB1 induced DC maturation, as evidenced here by upregulated expression of MHC class II molecules.
Chinchilla DCs expressed MHC class II molecules and effectively presented antigen to CD3+ T cells in vitro. (A) Upregulation of MHC class II expression as demonstrated by flow cytometry. Black histogram, no stimulus; green histogram, DCs incubated with TFPQ3; red histogram, DCs incubated with LB1. (B and C) Confocal microscopy images demonstrating MHC class II molecules (red fluorescence) on the surfaces of chinchilla DCs incubated with either TFPQ3 (B) or LB1 (C). (D) Chloroquine treatment of DCs incubated with LB1 abrogated the MHC class II-specific fluorescent signal. (E) Chinchilla CD3+ T cells primed in vivo with LB1 proliferated upon DC presentation of LB1 in vitro. The results are shown as the mean ± standard deviation of three independent experiments. *, P ≤ 0.05 compared to cells with no stimulus; †, P ≤ 0.05 compared to CD3+ T cells isolated from chinchillas immunized with MPL only; °, P ≤ 0.05 compared to cells stimulated with TFPQ3. Significance was determined by the Tukey honestly significant difference test for analysis of variance pairwise comparisons.
We next examined the consequence of LB1 internalization and upregulation of MHC class II molecules by measuring the induction of CD3+ T-cell proliferation in vitro. We first confirmed that chinchilla T cells could be induced to proliferate, as there have been no reports, to our knowledge, of chinchilla T-cell proliferation assays. Using a nonspecific stimulus, we observed dose-dependent proliferation upon in vitro stimulation of chinchilla CD3+ T cells with 10 μg or 1 μg concanavalin A/ml in the presence of DCs (Fig. 5E). Furthermore, CD3+ T cells isolated from chinchillas immunized with MPL exhibited comparable proliferation at the 10-μg concanavalin A/ml dose. These responses were statistically significant compared to cultures with no stimulus (P ≤ 0.05) and thus confirmed the utility of this assay system to measure chinchilla T-cell proliferation.
To examine the proliferation induced by a specific stimulus, DCs were incubated with LB1 prior to culture with CD3+ T cells isolated from chinchillas immunized with LB1 plus MPL. Dose-dependent proliferation resulted, which was significantly greater than the proliferation induced by DCs incubated with cells with no stimulus or by CD3+ T cells isolated from animals immunized with MPL only (P ≤ 0.05). These data indicated that upon uptake of LB1, chinchilla DCs effectively processed the immunogen and presented peptide fragments, likely in the context of MHC class II and costimulatory molecules, to T cells, thereby inducing their proliferation. Moreover, our data suggested that T cells recovered from chinchillas immunized with LB1 demonstrated a memory response, as significantly greater proliferation of T cells isolated from chinchillas immunized with LB1 was induced by DCs presenting the same antigen than by DC presentation of an unrelated peptide of comparable size (P ≤ 0.05).
Chinchilla respiratory DCs migrated to local lymphoid tissues following i.n. administration of LB1.We next examined the migration of DCs resident within the chinchilla nasal cavity following i.n. administration of LB1 plus the adjuvant MPL. To identify potential sites of immune induction, we administered the dye CFSE via microaerosol spray to chinchillas prior to administration of the immunogen. CFSE labeling allowed discrimination between cells resident within lymphoid tissue and DCs that had migrated to these tissues from the nasopharyngeal mucosa and had been shown not to disturb DC homeostasis (44). By this method, at the 1- and 3-hour time points after i.n. administration, the greatest mean fluorescence of CD11c+ CFSE+ cells was detected within the NALTs of the animals that received LB1 plus MPL (Fig. 6A) compared to all other tissues examined (Fig. 6B to F), suggesting that the NALT could serve as a primary inductive site following i.n. immunization of the chinchilla host. Moreover, in addition to the NALT, greater mean CFSE+ fluorescence was detected from CD11c+ cells within the brachial and axillary lymph nodes from the animals that received LB1 plus MPL than from those from animals that received only MPL (Fig. 6B and C). These data demonstrated that the LB1 delivered with MPL influenced greater DC activation and subsequent migration to these lymphoid tissues than did the adjuvant alone. Over a 12-h period, a decrease in CFSE+ fluorescence was observed within the NALT and brachial and axillary lymph nodes, which indicated that i.n. delivery of LB1 plus MPL triggered rapid and transient influx of DCs from the nasal cavity. In contrast, the CFSE+ signal detected in the cervical and mediastinal lymph nodes, as well as the spleens, from the animals that received LB1 plus MPL was comparable to that from animals that received MPL alone (Fig. 6D to F), which suggested that these lymphoid tissues likely were not primary inductive sites following i.n. immunization of the host. Thus, DCs within the chinchilla nasal cavity were induced to mature following i.n. delivery of LB1 plus MPL, which resulted in migration of these activated cells, primarily to the NALT, but also to the axillary and brachial lymph nodes.
DCs resident within the nasal mucosa migrated to the local lymphoid tissues after i.n. administration of LB1 plus MPL. The scatter plots depict cells isolated 1 h after receipt of LB1 plus MPL (top plot) or MPL alone (bottom plot) for each lymphoid tissue examined. The shaded region represents CD11c+ CFSE+ cells. The bar graphs demonstrate the corresponding mean CFSE+ fluorescence of the CD11c+ CFSE+ DCs 1, 3, or 12 h after i.n. administration of LB1 plus MPL or MPL alone. Note the difference in scale between NALT and other lymphoid tissues.
To confirm that the positive signal described above was specifically due to an influx of DCs from the nasal cavity and not to diffusion of dye from that site, we labeled DCs derived from culture of bone marrow cells with CFSE and then activated these cells with LB1 plus MPL or MPL alone in vitro. These CFSE-labeled and activated cells were then instilled into the chinchilla nasal cavity. One hour later, a distinct population of CD11c+ CFSE+ cells was detected within the NALT after receipt of DCs activated with LB1 plus MPL (Fig. 7A) but not when activated with MPL alone (Fig. 7B). Further, 11.5% of the cells isolated from the NALT of the animal administered LB1-activated DCs were CD11c+ CFSE+, whereas populations of ≤1% were detected within the brachial, axillary, cervical, and mediastinal lymph nodes and the spleen (data not shown). Whereas homing of nasal mucosa DCs to the axillary and brachial lymph nodes is shown in Fig. 6B and C, few of the instilled bone marrow-derived DCs were detected within these tissues, suggesting subtle differences between the DC types (i.e., those derived from culture in vitro versus those cells which develop in vivo). Nevertheless, the migration of the instilled DCs from the nasal cavity primarily to the NALT correlated well with the study in which CFSE had been delivered directly to the nasal cavity (Fig. 6A). The similarity of the results obtained using these two distinct but complementary in vivo assays strongly indicated that local cytokine/chemokine signals likely guided the LB1-activated DCs from the chinchilla nasal cavity to the NALT.
CFSE-labeled, LB1-activated bone marrow-derived DCs migrate to the NALT upon i.n. instillation. The scatter plots represent cells isolated from the NALTs of animals that received DCs activated in vitro with LB1 plus MPL (A) or MPL (B) 1 h after i.n. administration.
By confocal microscopy, we further verified that DCs resident within the nasal cavity migrated to the NALT after i.n. administration of LB1 plus MPL. CFSE+ cells were easily distinguished from the other cells isolated from this lymphoid aggregate due to their intense green fluorescence and unique morphology (i.e., expression of extensive dendrites) (Fig. 8A). Furthermore, the CFSE+ cells were labeled with an anti-CD11c antibody, thus confirming that these cells were DCs (Fig. 8B, red fluorescence). We also observed CD3+ T cells in close proximity to the DCs within this sample (Fig. 8C, blue fluorescence), suggesting potential interaction between the two cell types. Upon closer examination, MHC class II molecule expression on the DCs was localized directly under the CD3+ T-cell fluorescent signal (Fig. 8D and inset; red and blue fluorescence, respectively). These data suggested that delivery of LB1 to the nasal cavities of chinchillas by microaerosol spray induced DCs resident within the nasal cavity to take up the antigen, mature, and migrate to local lymphoid tissues, where they appeared to interact with T cells within the NALT.
Interaction of respiratory CFSE+ DCs with CD3+ T cells upon migration to the NALT. (A) Confocal microscopy images of cells isolated from the NALT 1 hour after i.n. administration of LB1 plus MPL, demonstrating the presence of CFSE+ cells (green) within the cell suspension (nuclei stained blue). (B) CFSE+ cell expressing the DC marker CD11c (red). (C) Detection of CD3+ fluorescent signals (blue; see the arrows) in close proximity to a CFSE+ DC (green). (D) Expression of MHC class II (red) by the CFSE+ cell in panel C is localized directly under the CD3+ T cells (blue).
DISCUSSION
Despite the great utility of the chinchilla for studies of viral and bacterial pathogenesis in OM (6, 12, 15, 19, 23, 30, 36, 46, 49, 54, 69), as well as for preclinical vaccine efficacy studies (7, 20, 22, 39, 56, 68), the cellular mechanisms for the induction of a protective immune response in this rodent host have yet to be fully characterized. One component of the innate immune system is the DC, a potent antigen-presenting cell found in many tissues of the body that is capable of initiation and modulation of primary immune responses (10, 55, 66). As we have observed significant efficacy against the development of NTHI-induced OM after i.n. immunization of the chinchilla with LB1, we were interested to determine whether processing and presentation of the peptide immunogen by nasal-cavity DCs likely participated in the observed protective response. We therefore examined the capture and presentation of LB1 by chinchilla bone marrow-derived DCs in vitro and by DCs within the chinchilla nasal cavity in vivo.
It is well established that precursor cells within the bone marrow of mice and rats can be induced to differentiate into DCs by the inclusion of GM-CSF and IL-4 in the culture medium (31, 45, 58, 60). We therefore sought to adapt this methodology to generate DCs from chinchilla bone marrow cells. Two factors impacted our ability to culture the chinchilla-derived cells: the cytokine specificity and the serum source. First, human-specific recombinant GM-CSF and IL-4, but not murine-specific cytokines, supported the viability of these cells in culture. This observation was not without precedent, as the use of human-specific (and rat-specific) reagents for immunodetection is more successful than when one tries to use mouse-specific reagents (27, 29, 63), which indirectly suggests that the chinchilla is closer to the rat and human than to mice in terms of similarity of immune molecules. Second, the serum source used to supplement the medium influenced the development of the culture, an observation similarly reported for murine DCs (24, 61). Therefore, whereas it is not surprising that chinchillas possess DCs, our data are the first to describe their culture.
It has been reported that DCs bind to OmpA proteins from K. pneumoniae, E. coli, and A. baumannii, inducing DC maturation and activation, including expression of costimulatory molecules, cytokine secretion, and stimulation of T-cell proliferation (34, 43, 71). These observations, among others, have led to the designation of this highly conserved family of bacterial proteins as a new PAMP (33). NTHI OMP P5 is an OmpA protein homologue (64). As shown by flow cytometry and confocal microscopy, chinchilla DCs bound and internalized OMP P5 in vitro. However, this result was overshadowed by far greater binding and internalization of LB1 by the DCs. This may have been due to the smaller molecular mass of LB1 than the native protein (4.6 kDa versus 36.4 kDa). Alternatively, a motif from OMP P5 that is included in the sequence of LB1 to which the DCs bind may be more optimally presented than in the native protein. It is interesting to speculate that this 19-mer motif (known to be a B-cell epitope) is a PAMP, which would allow DC binding to the immunogen via a pattern recognition receptor (3). We continue to investigate this hypothesis.
In vivo, by two techniques, we tracked the migration of DCs from the chinchilla nasal cavity to the NALT. These data demonstrated that LB1 induced DC maturation, which resulted in the ability of these cells to sense and respond to local homing signals within the nasal cavity, as similar migration was not detected when only MPL was administered. Our data further demonstrated that a single dose of LB1 delivered i.n. was sufficient to induce the maturation and migration of the DCs. Moreover, the NALT may serve as a primary inductive site following i.n. immunization in this host. Rodent NALT has been reported to have functional similarities to Waldeyer's lymphoid ring, the lymphatic tissue of the pharynx, palatine tonsils, and lingual tonsils, as well as other lymphoid tissue in the area that encircles the human nasopharynx and oropharynx (25, 65, 74). As such, the ability to target the lymphoid tissues proximal to the anatomical site of infection should be an important consideration when developing a vaccine and a vaccine delivery system.
In summary, our data are the first to describe the in vitro culture, functional characterization, and in vivo tracking of chinchilla DCs. We continue to examine the role of DCs in the development of a protective immune response following i.n. immunization with LB1 and several other NTHI adhesin-derived or surface-accessible antigens, for the prevention of NTHI-induced OM. Moreover, the chinchilla serves as a relevant animal in which to model multiple viral and bacterial diseases of the human respiratory tract; thus, the potential to apply the methods described here need not be limited to examination of vaccine efficacy for NTHI-induced OM only. Components of both innate and adaptive immunity can influence DC functions. Therefore, the data presented here will be useful as we continue to examine the roles of DCs in the initiation and modulation of the immune response in the chinchilla model.
ACKNOWLEDGMENTS
This work was supported by a grant from the NIH (NIDCD R01 DC03915) to L.O.B.
We thank Joseph Jurcisek, Harivadan Bhagat, and Adrianna Sumoza-Toledo for expert technical assistance, John Hayes for statistical analysis, and Jennifer Neelans for manuscript preparation. We are also grateful to Kevin Mason for helpful comments and suggestions.
FOOTNOTES
- Received 17 October 2007.
- Returned for modification 29 November 2007.
- Accepted 17 December 2007.
- Copyright © 2008 American Society for Microbiology