Comparison of Lung Dendritic Cells and B Cells in Stimulating Naive Antigen-Specific T Cells

Mice

BALB/c (H-2^d) and heterozygous mice transgenic for the DO11.10 αβ-TCR on a BALB/c background were bred in specific pathogen-free facilities in the University of New Mexico (UNM) Animal Resource Facility (Albuquerque, NM). Heterozygous DO11.10 male breeders, a kind gift of Dr. Dennis Loh (Hoffman LaRoche), were bred to BALB/c females to generate heterozygous offspring. Offspring were tested for expression of the transgene by obtaining tail vein blood, staining leukocytes with mAb KJ1-26, and analyzing by flow cytometry. All mice were kept in a specific pathogen-free environment in sterilized cages and provided sterilized food and water. A sentinel animal kept in this facility was monitored each quarter for Ab conversion to Sendai and mouse hepatitis viruses and Mycoplasma pulmonis. Both males and females were used at 10–12 wk of age.

Reagents

Ags used in this study were chicken OVA (Sigma, St. Louis, MO) and synthesized OVA peptide, representing amino acids 323–339 of chicken OVA. The UNM Protein Chemistry Laboratory synthesized the antigenic OVA_323–339 peptide using an Applied Biosystems 430A peptide synthesizer (PE Applied Biosystems, Foster City, CA). Cells were cultured in RPMI 1640 supplemented with 2 mM l-glutamine, 50 μM 2-ME, 1 mM sodium pyruvate, 1 mM nonessential amino acids, 100 U/ml penicillin, 100 μg/ml streptomycin, 20 μg/ml gentamicin, and 10% FBS (Life Technologies, Grand Island, NY), and referred to as complete RPMI (cRPMI). For the lymphoproliferation assays and Ag pulsing of APCs, cRPMI was additionally supplemented with 250 U/ml catalase and 1 μg/ml indomethacin (Sigma).

Preparation of APCs

Lung cells were prepared as described (18, 19). Briefly, mice were injected i.p. with 150 U heparin (ICN Biomedicals, Aurora, OH) and 1 ml of the anesthetic, Avertin (44 μM 2, 2, 2-tribromoethanol, 71 μM tert-amyl alcohol (Aldrich Chemical, Milwaukee, WI) in distilled water), and exsanguinated. Mice lung airways were lavaged using an intratracheal cannula with cold PBS containing 0.6 mM EDTA. The pulmonary vasculature was perfused with sterile saline to remove peripheral blood cells. The lavaged, perfused lungs were removed, minced, and enzyme treated at 37°C for 90 min in cRPMI containing collagenase A (0.7 mg/ml; Boehringer Mannheim, Indianapolis, IN) and type IV bovine pancreatic DNase I (30 μg/ml; Sigma). Digested lung tissue was tapped through a wire screen, particulate matter was removed by rapid filtration through a nylon wool plug, and the filtered cells were washed twice in HBSS (Life Technologies). Lung cells were resuspended in high density Percoll (Pharmacia, Piscataway, NJ; density = 1.075 g/ml), overlaid with equal volume of lower density Percoll (density = 1.030 g/ml), and centrifuged at 400 × g for 20 min. Cells at the 1.075/1.030 Percoll interface, which are enriched for mononuclear cells, recovered, and washed with HBSS, are referred to as freshly isolated lung cells. The isolation of freshly isolated lung cells takes approximately 3 h post lung tissue harvest. Three preparations of cells were used in experiments. First, for whole lung cell experiments, overnight Ag-pulsing experiments of whole lung cells, and for some phenotyping experiments, cells at the interface were recovered, washed with HBSS, resuspended (2 × 10⁶ cells/ml) in cRPMI, and incubated overnight (18–24 h) in 50 ml polypropylene conical centrifuge tubes. Second, for some phenotyping experiments, lung cells at the Percoll interface were recovered, washed with HBSS, resuspended (2 × 10⁶ cells/ml) in cRPMI, and incubated in tissue culture dishes overnight (18–24 h). Thus, these cells comprised both nonadherent and loosely adherent cells. Third, for experiments in which B cells and DCs were compared functionally, lung cells at the Percoll interface were recovered, washed with HBSS, resuspended (2 × 10⁶ cells/ml) in cRPMI, and incubated in tissue culture dishes for 2 h at 37°C in a 5% CO₂ incubator to immediately separate B cell and DC populations. Nonadherent cells were transferred to a new tissue culture dish and labeled enriched B cells. Adherent cells were cultured overnight in cRPMI. The cells that spontaneously detached during the overnight culture are referred to as loosely adherent cells (LAd) and were even further enriched for DCs. DCs represent approximately 11 ± 3% of the total stained LAd population (18). A similar protocol was followed for the isolation of spleen cells, with the exception that after removal from the mice, spleens were perfused, until white, with a solution of collagenase A (0.7 mg/ml; Boehringer Mannheim) and type IV bovine pancreatic Dnase I (30 μg/ml; Sigma) in cRPMI before the enzymatic digestion step. In addition, a 1000 g/ml Percoll/RPMI gradient was used to enrich for mononuclear cells. Spleen cells at and above the interface were collected.

Ag pulsing of APCs

Freshly isolated lung or spleen cells were resuspended (2–3 × 10⁶ cells/ml) in cRPMI containing no Ag, intact OVA protein, or OVA_323–339 peptide and 250 U/ml catalase (Sigma), and 1 μg/ml indomethacin (Sigma) in 50 ml polypropylene conical centrifuge tubes at 37°C, 5% CO₂ for 18–24 h. Freshly isolated lung and spleen cells were pulsed within 3 h after the lung tissue harvest. Catalase and indomethacin were added to inhibit macrophage-derived hydrogen peroxide and PGE₂ production, respectively.

Four-color immunophenotyping of DCs and B cells

To determine the phenotype of DCs and B cells from freshly isolated and cultured lung populations, cell preparations were first incubated for 1 h at 37°C in a 2.5% suspension of FITC-labeled 1.5-μm-diameter latex beads (Polyscience, Warrington, PA) in RPMI containing 20% FBS to identify phagocytic cells. Free latex beads were separated from the cells by centrifugation over FBS. Cells were then preincubated with purified anti-CD16/CD32 (2.4G2, rat IgG2b; PharMingen, San Diego, CA) to block both FcR-mediated binding of staining Ab and cell aggregation by FcγRII binding by macrophages of Ab-stained DCs and B cells. Subsequently, cells were stained with biotinylated anti-I-A^d/I-E^d (2G9, rat IgG2a; PharMingen), allophycocyanin-conjugated anti-CD45R/B220 (RA3-6B2, rat IgG2a; PharMingen), and a PE-conjugated mAb to the surface molecule of interest. I-A^d binding was identified by incubation with streptavidin-peridinin chlorophyll protein (PerCp) (Becton Dickinson, San Jose, CA). PE-conjugated anti-CD80 (1G10, rat IgG2a), anti-CD86 (GL1, rat IgG2a), anti-CD40 (3/23, rat IgG2a), anti-CD11a (2D7, rat IgG2a), anti-CD54 (3E2, hamster IgG), anti-CD25α (3C7, rat IgG2b), anti-CD24 (M1/69, rat IgG2b), anti-CD1d (1B1, rat IgG2b), anti-CD8α (53-6.7, rat IgG2a), anti-CD11b (M1/70, rat IgG2b), rat IgG2a (R35-95), rat IgG2b (R35-38), and hamster IgG (G235-2356) were from PharMingen. Cells were always stained at 4°C and washed with PBS containing 2% FBS and 40 μg/ml EDTA. A Becton Dickinson flow cytometry cell sorter FACSCalibur was used for data acquisition and Cell Quest software (Becton Dickinson) for analysis. Presently, no markers are available that clearly distinguish lung DCs from lung macrophages. However, eliminating macrophage contamination was an important concern for us in our comparison of lung DCs and B cells. Therefore, in our phenotype analysis of DCs, we knowingly excluded all cells that could ingest latex beads. For DC phenotyping, analysis was performed on the I-A^d-positive, FITC-bead-negative, B220-negative population for expression of the PE-labeled marker. For B cell phenotyping, analysis was performed on the I-A-positive, FITC-bead-negative, B220-positive population for expression of the PE-labeled marker.

Isolation of DCs and B cells

To prepare DCs and B cells for sorting, whole lung cells, LAd cells, or enriched B cells were first incubated with FITC-latex beads, followed by preincubation with purified anti-CD16/CD32 to block FcR and staining with biotinylated anti-I-A^d/I-E^d and PE-conjugated anti-CD45R/B220. I-A^d binding was identified by incubation with streptavidin-allophycocyanin (Becton Dickinson). To obtain lung DCs in high purity from stained whole lung cells or LAd, lung DCs were identified using an EPICS Elite flow cytometer, and sorted as FITC-bead-negative, B220-negative, I-A^d-expressing cells. To obtain lung B cells in high purity from stained whole lung cells or enriched B cells, lung B cells were identified using an EPICS Elite flow cytometer, and sorted as FITC-bead-negative, B220-expressing, I-A^d-expressing cells. The forward scatter/side scatter (FS/SSC) profile of stained cells was used to set a gate to exclude debris, free latex beads, some macrophages, and possibly some NK cells. To optimize sorting, this gate was further refined by back-gating from a gate set on I-A^d-positive, FITC-bead-negative cells to make I-A^d-positive, FITC-bead-negative cells the predominate population within the FS/SSC gate. To exclude extremely bright FITC fluorescent cells with fluorescence signals that wrapped around from channel 1024 to channel 1 due to the electronics of the Elite, a second gate was set starting at channel 1 and encompassed the FITC-negative population using a linear scale FITC histogram. For DC sorting, an additional gate was set using a log scale PE fluorescence histogram to exclude CD45R/B220-positive cells (B cells) from the sort. Thus, four gates were used to sort I-A^d-positive, nonphagocytic, B220-negative DCs: a FS/SSC gate; a histogram-based gate excluding FITC-bead-positive cells, a gate on I-A^d-positive, FITC-bead-negative cells; and a gate on B220-negative cells. For sorting B cells, I-A^d-positive, nonphagocytic, B220-positive cells were collected using the FS/SSC, linear scale FITC histogram, and I-A^d/FITC gates similar to those used for DCs, and a fourth gate set on CD45R/B220-positive cells. Sorted DC populations (>96% purity) were examined for latex bead uptake, I-A^d expression, and elongated cell processes, using an epifluorescent photomicroscope. Sorted B cell populations (>97% purity) were examined for latex bead uptake, I-A^d expression, and lack of elongated cell processes. Sorted cell populations were 99% viable, determined by trypan blue dye exclusion. On average, this protocol yields 2.5 × 10³ lung DCs/mouse. It was shown previously that this technique did not interfere with functional activity, as anti-I-A Abs must be continuously present in cultures to interfere with I-A-dependent lymphoproliferation (18). We always sorted for lung DCs after an overnight culture; a timepoint at which we believe the majority of DCs did not take up latex beads, and a timepoint at which DCs were still heterogeneous for the expression of I-A^d and accessory molecules. Thus, a significant portion of these nonphagocytic lung DCs studied was immature, i.e., still had a moderate density of I-A^d. A similar protocol was followed for staining and sorting spleen DCs.

Enrichment of naive (CD62L⁺) CD4⁺ DO11.10 T cells

To prepare T cells for flow cytometry cell sorting on a Coulter Epics Elite (Coulter, Hialeah, FL), spleen cells from DO11.10 mice were passed on nylon wool, stained with FITC-conjugated anti-CD4 (RM4-5, rat IgG2a; PharMingen), PE-conjugated anti-CD62L (Mel-14, rat IgG2a; PharMingen), and biotinylated anti-I-A^d/I-E^d (2G9, rat IgG2a; PharMingen), followed by secondary staining with streptavidin-allophycocyanin (Becton Dickinson). To obtain naive CD4⁺ T cells in high purity, cells expressing both CD4 and CD62L, but negative for I-A^d, were collected in a tube. The FS/SSC profile of stained cells was used to set a gate on the lymphocyte population. A second gate was set on cells expressing high levels of CD4 and CD62L. A third gate was set on cells expressing high levels of CD62L, but lacking expression of I-A^d to exclude APCs. All three gates were used together to sort naive (CD62L⁺) CD4⁺ T cells. Sorted cells were 99% viable, as determined by trypan blue dye exclusion.

Lymphoproliferation assay

For whole lung cell experiments, the lung cells were washed and irradiated (2000 rad) before use in the lymphoproliferation assay. For experiments using sorted APCs, B cells and DCs were purified from nonirradiated whole lung cells, LAd, or enriched B cells by FACS before use in the lymphoproliferation assay. Sorted naive CD4⁺ DO11.10 T cells (2.5 × 10⁴/well) were cultured with APCs in triplicate, unless noted, in 96-well flat-bottom plates at 37°C, 5% CO₂ for 4 days. Culture conditions were 200 μl cRPMI containing no Ag, OVA, or OVA_323–339 peptide, and 250 U/ml catalase (Sigma) and 1 μg/ml indomethacin (Sigma). Catalase and indomethacin were added to inhibit macrophage-derived hydrogen peroxide and PGE₂ production, respectively, which was needed for whole lung cells and was also used for purified B cell and DC cultures to assure valid comparisons. Eighteen hours before harvesting, 0.5 μCi methyl [³H]thymidine (Amersham, Arlington Heights, IL) was added to each well. To determine whether APCs were a contaminant in the sorted T cell population, sorted T cells (2.5 × 10⁴/well) were cultured with 10 μM OVA_323–339 peptide. Controls included each cell type with media only and media with OVA or OVA_323–339 peptide, where appropriate. Cultures were harvested with an automated cell harvester and analyzed in a Beckman LS 6500 liquid scintillation counter (Beckman Instruments, Fullerton, CA). Results were reported as cpm. For whole spleen cell experiments, the spleen cells were washed and irradiated (2000 rad) before use in the lymphoproliferation assay. For experiments using sorted spleen APCs, DCs were purified from nonirradiated spleen cells or LAd by FACS before use in the lymphoproliferation assay.

Statistics

All statistical analyses were performed with the Statview software (SAS Institute, Cary, NJ). For data in Figs. 3⇓, 4⇓, and 5⇓, an analysis of variance model and an unpaired t test were used for examining data. A p value <0.05 was considered significant.

Lung cells process and present Ag to naive T cells

Previous studies (18, 19) showed that freshly isolated lung cells were highly efficient in stimulating allogeneic T cells in a mixed lymphocyte response. To determine the capacity of lung cells to process and present soluble protein to naive Ag-specific T cell, freshly isolated irradiated lung cells were immediately cultured with naive CD4⁺ DO11.10 T cells and different concentrations of intact OVA protein or OVA_323–339 peptide. This experiment was biased toward detecting the APC activity of lung cells that were resistant to 2000 rad, i.e., DCs. Freshly isolated lung cells were capable of stimulating naive T cells to proliferate in an Ag-specific manner when OVA_323–339 peptide or OVA protein was included throughout the lymphoproliferation assay (Fig. 1⇓a). However, the molar concentration of protein required to stimulate T cell proliferation equal to that observed with peptide was at least 1000-fold more. The concentration of protein that showed optimal T cell proliferation was 100 μM OVA protein for all lung cell concentrations and was used for all subsequent assays, including experiments in which overnight pulsing was done. In a separate experiment, freshly isolated spleen cells were used as a comparative APC population. Freshly isolated spleen cells showed similar Ag dose and cell number responses (Fig. 1⇓b). However, on average, freshly isolated spleen cells were 40% more effective at stimulating naive DO11.10 T cells than freshly isolated lung cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Lung cells (LC; a) and spleen cells (Sc; b) stimulate naive CD4⁺ DO11.10 T cells to proliferate in the presence of OVA or OVA_323–339 peptide (pep). Freshly isolated lung cells were prepared as described in Materials and Methods and irradiated before coculture in flat-bottom 96-well plates with naive CD4⁺ DO11.10 T cells (2.5 × 10⁴ cells/well) and 0–100 μM of OVA protein or OVA_323–339 peptide. These cultures were initiated within 3 h of the lung tissue harvest. After 4 days, [³H]thymidine (0.5 μCi/well) was added for the final 18 h of culture to assess T cell proliferation, depicted as cpm. The data shown are the mean and SEM of triplicate wells. T only, 595 cpm; T + 100 μM OVA, 346 cpm; T + 100 μM peptide, 901 cpm.

Lung DCs process and present Ag to naive T cells

Our previous studies also showed that DCs sorted from cultured lung cells were potent stimulators of allogeneic T cell proliferative responses (18). To address whether isolated lung DCs have the capacity to stimulate naive Ag-specific T cells, FACS-purified lung DCs from nonirradiated lung cells were cultured withnaive CD4⁺ DO11.10 T cells and different concentrations of OVA_323–339 peptide. Lung DCs were very efficient at presenting peptide to naive Ag-specific T cells when the peptide was included throughout the lymphoproliferation assay (Fig. 2⇓). Optimal responses were obtained between 1 and 10 μM OVA_323–339 peptide; 10 μM of peptide was used in all subsequent experiments. A direct comparison of lung DCs with spleen DCs showed similar Ag dose responses and equivalent capacities to stimulate naive CD4⁺ DO11.10 T cells (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Purified lung DCs stimulate naive CD4⁺ DO11.10 T cells to proliferate in the presence of OVA peptide. Purified lung DCs were obtained from nonirradiated LAd, as described in Materials and Methods. Lung DCs (1 × 10³ cells/well) were cocultured with naive CD4⁺ DO11.10 T cells (2.5 × 10⁴ cells/well) and increasing concentrations of OVA_323–339 peptide. After 5 days, T cell proliferation was assessed as described in Fig. 1⇑. The data shown are the mean and SEM of triplicate wells and are representative of two independent experiments.

Naive T cells are stimulated equally by lung DCs and B cells with continuous OVA peptide in culture, but not with continuous OVA protein

DCs and B cells are two cell types in the murine lung with the potential to capture and display Ag to CD4⁺ T cells. To determine whether these two MHC class II (I-A^d)-expressing cell populations show a difference in Ag-specific APC activity, we compared the relative ability of lung B cells and DCs to process and present OVA protein to naive CD4⁺ DO11.10 T cells. DCs sorted from nonirradiated LAd and B cells sorted from nonirradiated nonadherent lung cells were cocultured with naive CD4⁺ DO11.10 T cells and OVA protein or OVA_323–339 peptide. On average, DCs were the only APCs able to efficiently stimulate T cell proliferation with intact protein (Fig. 3⇓: DC+T+ova vs B+T+ova, overall p ≤ 0.001). However, lymphoproliferation cultures in which OVA peptide was continuously present showed that B cells were equally efficient to DCs in presenting peptide to Ag-specific naive CD4⁺ T cells (DC+T+pep vs B+T+pep, overall p ≤ 0.3493).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Purified lung B cells and lung DCs are similar in stimulating naive CD4⁺ DO11.10 T cells to proliferate in the presence of continuous OVA_323–339 peptide, but only DCs are stimulatory in the presence of intact continuous OVA protein. Purified lung DCs and B cells were obtained from nonirradiated LAd and enriched B cells, respectively, as described in Materials and Methods. Lung DCs or lung B cells were cocultured with naive CD4⁺ DO11.10 T cells (2.5 × 10⁴ cells/well), and 100 μM OVA protein or 10 μM OVA_323–339 peptide. After 5 days, T cell proliferation was assessed as described in Fig. 1⇑. T only, 99 cpm; T + 100 μM OVA, 500 cpm; T + 10 μM peptide, 1796 cpm. The data shown are the mean and SEM of three independent experiments.

Pulsed lung cells can prime Ag-specific naive T cells

We next examined whether lung cells exposed to (pulsed) intact OVA protein before addition to the lymphoproliferation assay could stimulate naive T cells in an Ag-specific manner. Freshly isolated lung cells cultured with OVA protein for 18–24 h were compared with lung cells that were not pulsed for their capacity to stimulate naive CD4⁺ DO11.10 T cells. Both the pulsed and not pulsed lung cell populations were irradiated before use in the lymphoproliferation assay. Overnight pulsed lung cells showed specific priming in vitro, whereas not pulsed lung cells showed no APC activity (Fig. 4⇓a: Not pulsed vs Pulsed, p ≤ 0.001). In a separate experiment, not pulsed and overnight pulsed spleen cells showed similar results to not pulsed and overnight pulsed lung cells, respectively (Fig. 4⇓b).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

OVA-pulsed lung cells stimulate naive CD4⁺ DO11.10 T cells to proliferate. Freshly isolated lung cells (a) and spleen cells (b) were either pulsed (within 3 h after the lung tissue harvest) or not pulsed overnight with 100 μM OVA and irradiated before coculture with naive CD4⁺ DO11.10 T cells (2.5 × 10⁴ cells/well). After 5 days, T cell proliferation was assessed as described in Fig. 1⇑. The data shown are the mean and SEM of quadruplicate wells. T only, 338 cpm; T + 100 μM OVA, 800 cpm; T + 10 μM peptide, 5831 cpm.

Pulsed lung DCs are more efficient than pulsed lung B cells at priming Ag-specific naive T cells

With APC activity evident in protein-pulsed populations of irradiated lung cells that contained 7.5 ± 2.4% B cells and 1.2 ± 0.6% DC, it was important to verify, under nonirradiating conditions, that DCs were the only APCs capable of inducing naive T cells to proliferate and confirm the lack of APC activity by B cells. B cells and DCs sorted from nonirradiated lung cells that were cultured overnight with intact protein or OVA_323–339 peptide were compared for their capacity to stimulate naive OVA-specific CD4⁺ T cells. In contrast to the similar levels of APC activities of lung B cells and DCs when peptide was present throughout the lymphoproliferation assay (refer to Fig. 3⇑), peptide-pulsed lung DCs were more efficient than peptide-pulsed B cells at stimulating naive T cells to proliferate (Fig. 5⇓a; Peptide-pulsed B cells vs Peptide-pulsed DCs, overall p ≤ 0.0001). Significant differences between peptide-pulsed B cells and DC cells were observed at each APC concentration: 1.25 × 10³ (p ≤ 0.0246); 2.5 × 10³ (p ≤ 0.0299); and 5 × 10³ (p ≤ 0.0173). As expected from the results of the experiments with OVA protein continuously present during the lymphoproliferation assay (refer to Fig. 4⇑), only OVA protein-pulsed DCs demonstrated any APC activity for naive T cells (Fig. 5⇓b; Protein-pulsed B cells vs Protein-pulsed DCs, overall p ≤ 0.0001). Significant differences between protein-pulsed B cells and DCs were observed at 2.5 × 10³ (p ≤ 0.0049) and 5 × 10³ (p ≤ 0.0005) concentrations of APCs.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Pulsed lung DCs are superior to pulsed lung B cells in stimulating naive CD4⁺ DO11.10 T cells to proliferate. Freshly isolated lung cells were pulsed overnight with 10 μM OVA_323–339 peptide (a) or 100 μM OVA (b). Lung DCs and B cells were sorted from these nonirradiated lung cell populations and cocultured with naive CD4⁺ DO11.10 T cells (2.5 × 10⁴ cells/well). After 5 days, T cell proliferation was assessed as described in Fig. 1⇑. T + 5K mock-pulsed DC, 311 cpm; T + 5K mock-pulsed B, 128 cpm. For a, the data shown are the mean and SEM of three independent experiments. For b, the data shown are the mean and SEM of triplicate wells and are representative of two independent experiments.

MHC class II expression on freshly isolated and cultured lung DCs and lung B cells

Activation of CD4⁺ T cells is initiated when their MHC-restricted, peptide-specific TCRs bind MHC class II-peptide complexes expressed on APCs. The observation that peptide-pulsed B cells and DCs differed in their ability to stimulate Ag-specific naive T cells prompted us to examine the levels of I-A^d expression on these two APC types to determine whether lung DCs have the capacity to present more peptide at their surface than B. In addition, the expression of I-A^d on lung APCs from both freshly isolated lung cells and lung cells cultured overnight in plastic culture dishes was compared to determine whether murine lung B cells and DCs up-regulate MHC class II expression during culture in a manner similar to that described for murine skin and splenic DCs (5, 6, 7, 21, 22, 23, 24). CD45R/B220 staining in combination with FITC-bead uptake and I-A^d staining was used to distinguish B cells from DCs so that the level of expression of I-A^d on both of these nonphagocytic, MHC class II-positive populations could be determined.

Lung DCs in freshly isolated lung cell suspensions (I-A^d; 100%, 451 MFI) were comprised of DCs expressing moderate levels of I-A^d (83%, 286 MFI) and a smaller percentage expressing high levels of I-A^d (17%, 1325 MFI) (Fig. 6⇓). In contrast, DCs in lung cells cultured overnight (I-A^d; 100%, 1057 MFI) were comprised mainly of high I-A^d (54%, 1692 MFI) expressing DCs. The intensity of I-A^d expressed on B cells in freshly isolated lung cell suspensions was uniform and at moderate levels (100%, 271 MFI) and changed little after culture (100%, 340 MFI). When lung B cells and lung DCs were compared under freshly isolated and cultured conditions, it was apparent that lung DCs expressed higher intensities of I-A^d than lung B cells under either set of condition.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Lung DCs express higher levels of I-A^d than lung B cells. Lung cells were stained as described in Materials and Methods. The FS/SSC profile of stained cells was used to set gate A to exclude debris, free latex beads, some macrophages, and possibly some NK cells. To optimize analysis, this gate was further refined by back-gating from gate B set on I-A^d-positive, FITC-bead-negative cells to ensure that DCs and B cells were the predominate populations within gate A. For DC analysis, gate C was set on CD45R/B220-negative cells within gate A and analyzed for the expression of I-A^d. Note the majority of DCs in gate C are difficult to see because they fall into the lower B220 fluorescence channels that are not resolved from the baseline. However, the I-A^d staining on DCs in gate C is clearly shown in the I-A^d histogram for DCs. For B cell analysis, gate D was set on CD45R/B220-positive cells within gate A and analyzed for the expression of I-A^d.

In parallel experiments, we determined whether culture on plastic might affect the expression of I-A^d on lung APCs by comparing cells held overnight in suspension with lung APCs cultured overnight on plastic tissue culture dishes. The expression of I-A^d on B cells and DCs cultured in suspension was similar to that seen on B cells and DCs cultured on plastic dishes, respectively (data not shown). In addition, the expression of I-A^d on DCs and B cells pulsed with OVA protein or OVA peptide was similar to that seen on nonpulsed DCs and B cells, respectively (data not shown).

Accessory molecule expression on freshly isolated and cultured lung DCs and B cells

Previous studies showed that murine lung DCs utilized multiple accessory molecules in their role as initiators of allogeneic T cell responses, including CD2 ligand, CD11a, CD40, CD54, CD80, and CD86 (18). To further address possible mechanisms for the functional APC differences seen between peptide-pulsed lung B cells and DCs, the expression of accessory molecules was compared for these two I-A^d-expressing populations. In addition, the expression of accessory molecules on B cells and DCs from freshly isolated and cultured lung cells was compared to determine whether their accessory molecule expression increased in a manner similar to that described for murine skin and splenic DCs (5, 6, 7, 21, 22, 23, 24). Included in this present study were the lymphoid-related DC markers, CD8α and CD1d, and the myeloid-related DC marker, CD11b, to delineate lymphoid-derived and myeloid-derived DC subsets in the lung DCs (25).

Previous three-color phenotyping studies in our lab demonstrated that nonphagocytic, high I-A^d-expressing DCs in LAd generally expressed high levels of CD11a, CD54, CD86, CD25α, and CD24; and a subpopulation expressed moderate levels of CD80 (18, 19). A marker to clearly identify B cells was not included in these earlier lung studies so that accessory molecule expression on DCs expressing moderate levels of I-A^d might have also included some B cells. As discussed above, in the current study, CD45R/B220 staining was used to distinguish B cells from DCs.

Based on four-color FACS analysis (Fig. 7⇓), a small subset (26%) of freshly isolated lung B cells expressed very low levels of CD80, but this accessory molecule was not up-regulated in culture. In contrast, the majority (86%) of freshly isolated lung DCs expressed low levels of CD80 that increased strikingly during culture. CD86 was expressed on the majority of freshly isolated B cells (88%) and DCs (90%), although at a higher intensity on DCs. After culture, the expression of CD86 was appreciably up-regulated on both B cells and DCs, but to a higher intensity on DCs. The majority of freshly isolated B cells (99%) and a subset of freshly isolated lung DCs (64%) expressed CD40, but the intensity was up-regulated during culture only on DCs. A large percentage of freshly isolated B cells (87%) expressed uniform low levels of CD11a, but DCs (98%) were heterogeneous for the expression of CD11a with both low and moderate expressing populations. After culture, the expression of CD11a remained unchanged on B cells, but DCs became more homogeneous for the expression of moderate levels of CD11a. CD54 was expressed on all freshly isolated B cells (100%) and DCs (99%), although at a higher intensity on DCs. After culture, the expression of CD54 was up-regulated on both B cells and DCs, but to higher levels on DCs. The expression of CD25α on B cells and DCs before culture was low and only on a small subpopulation of B cells (5%) and DCs (9%) before culture. After culture, the expression of CD25α changed little on B cells, but was up-regulated markedly on the majority of DCs. Both freshly isolated lung B cells (100%) and lung DCs (94%) expressed high levels of CD24 that changed little after culture.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Lung DCs express higher levels of accessory molecules than lung B cells both when freshly isolated and after overnight culture. Lung cells were stained as described in Materials and Methods. For DC phenotyping, gates A, B, and C shown in Fig. 6⇑ were used to perform analysis on the I-A^d-positive, FITC-bead-negative, B220-negative population for the expression of a series of PE-labeled accessory molecule markers. For B cell phenotyping, gates A, B, and D shown in Fig. 6⇑ were used to perform analysis on the I-A^d-positive, FITC-bead-negative, B220-positive population for the expression of the PE-labeled accessory molecule markers.

Overall, the intensity of I-A^d, CD80, CD86, CD11a, CD54, and CD25α expressed on cultured B cells was less than on cultured lung DCs, as assessed by immunofluorescence and flow cytometry. The majority of fresh B cells expressed low levels of I-A^d, CD80, CD86, CD40, CD11a, CD54, and CD25α, and after culture only CD86, CD40, and CD54 showed a noteworthy increased expression, although never to the levels detected on cultured DCs. In contrast, the majority of freshly isolated DCs expressed moderate levels of I-A^d and CD54 and low levels of CD80, CD86, CD40, and CD11a, which after culture increased markedly. Four-color phenotyping (Fig. 8⇓) also showed that the expression of I-A^d on DCs coincided with the expression of several accessory molecules. The expression of CD80, CD86, CD40, CD11a, CD54, CD25α, and CD24 on DCs expressing high levels of I-A^d (DC^{high I-Ad}) was greater than that seen on DCs expressing moderate levels of I-A^d(DC^{mod I-Ad}).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Cultured DCs expressing high I-A^d show higher levels of accessory molecules than cultured DCs expressing moderate I-A^d. Freshly isolated and lung cells cultured overnight were stained as described in Materials and Methods, and gates A, B, and C shown in Fig. 6⇑ were used to identify the DC population. Two additional gates (not shown) were established to perform analysis on DCs expressing moderate levels of I-A^d (Fluorescence intensity, 40–750) and DCs expressing high levels of I-A^d (CDC^{high I-Ad}) (Fluorescence intensity, 751–10,000) for the expression of a series of PE-labeled accessory molecule markers.

As discussed above for I-A^d expression, the expression of accessory molecules on B cells and DCs cultured in suspension was similar to that seen on DCs and B cells cultured on plastic dishes, respectively (data not shown). In addition, the expression of accessory molecules on B cells and DCs pulsed with OVA protein or OVA peptide was similar to that seen on nonpulsed DCs and B cells, respectively (data not shown).

Distinct mature lymphoid- and myeloid-derived subpopulations of DCs have been described in the murine spleen (25, 26). We evaluated by four-color flow cytometry the expression of the lymphoid-related markers, CD1d and CD8α, and the myeloid-related marker, CD11b, to determine whether lung DCs could be divided into lymphoid and myeloid subpopulations, respectively. Lineage analysis revealed that the majority of freshly isolated lung B cells and DCs expressed low levels of CD1d, and a very small subpopulation expressed low levels of CD8α. CD1d expression on lung B cells did not change after culture. However, CD1d was up-regulated on cultured lung DCs, specifically lung DCs expressing high levels of I-A^d. CD8α expression on both lung B cells and lung DCs did not change after culture. A small subpopulation of freshly isolated and cultured lung B cells showed low levels of CD11b. The majority of freshly isolated lung DCs expressed moderate levels of CD11b that did not change after culture. However, the highest levels of I-A^d also expressed the highest level of CD11b.