| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
* Department of Dermatology,
Department of Immunology,
Department of Otolaryngology, and
Department of Cell Biology and Physiology and Center for Biologic Imaging, University of Pittsburgh School of Medicine, Pittsburgh, PA 15217
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
In the presence of inflammation or "danger" (10), such as bacterial infection (LPS) (11), viral components, cell death products (TNF-
) (12), cell displacement (6, 12, 13), or CD4+ T cell interactions (CD40L) (14), DCs undergo a complex process of morphologic and functional activation to rapidly progress from intermediate to fully mature/activated DCs. During the intermediate maturation stage, class II MHC molecules are no longer sequestered in MIIC and instead associate with lgp/LAMP-negative multilamellar class II vesicle structures (15, 16). Class II vesicles may be important in accumulating MHC class I Ag presentation ligands and accessory molecules such as B7.2, which together are shuttled to the plasma membrane as a microdomain upon final maturation (17). Activated or "mature" DCs become potent APCs, with augmented cell surface patterning of molecules involved in Ag presentation and costimulation throughout extensive veiled membrane protrusions, whereas MIIC lysosomes become relegated to tight perinuclear clusters in these cells (15, 18, 19). These changes are concurrent with DC transit to draining secondary lymphoid organs, where priming of naive T cell populations can occur through TCR recognition of antigenic peptides in the context of MHC and costimulation (20, 21). Along with these extensive phenotypic and morphologic changes, terminally mature DCs have been thought to lose much of their macropinocytic and phagocytic capacity.
Phagocytosis is a fundamental process that links the innate and acquired arms of the immune system (22). Phagocytosis of particulates proceeds through the regulated action of actin polymerization through ligation of cell surface receptors that recruit the action of RhoGTPases such as Rac1, cdc42, and Rho (23). Recent work suggests that transfection of DCs with Rho and cdc42 stimulates enhanced T cell presentation and solute endocytosis (23) and may be important in phagosome formation (24). Little is defined regarding the requirements of immature or mature DC to ingest particulates, although studies indicate that the ionic characteristics of particulates (25, 26) and specific DC cell surface molecules (27) appear to contribute to Ag acquisition. Many groups are also investigating the role of DCs in accessing apoptotic vs necrotic bodies as "particulates" that lead to immunostimulatory and/or tolerogenic responses (24, 28, 29). With regard to DCs, immature DCs are known to effectively internalize and process a variety of particulate Ags (30). However, many reports investigating Ag acquisition by DCs document that phagocytosis of particulates is significantly curtailed but not eliminated in activated DC populations (2, 12, 31), although macropinocytosis of soluble tracers is effectively terminated (32, 33). These findings suggest that some DCs in activated DC populations may retain a unique window for particulate Ag sampling that has been under-appreciated.
Particulates include >500-nm diameter Ags that can be synthetic (iron oxide beads, latex microspheres) or biologic (zymosan, bacteria, cellular debris) structures. These Ags, upon phagocytosis by APCs, can be efficiently processed through the class II presentation pathway for CD4+ Th cell priming, and, interestingly, can be cross-presented through the MHC class I-processing pathway for presentation to CD8+ T cells (34, 35, 36). Recent work has shown that newly formed phagosomes within DCs can merge with the endoplasmic reticulum (ER) to acquire many of the molecular constituents necessary for class I-mediated presentation (37, 38). Functionally, we have previously shown that in vivo administration of iron oxide particulates carrying a model tumor Ag effectively protected mice from a lethal challenge of Ag-expressing tumor cells, and that Ag-specific lytic activity can be achieved in equine animal models following delivery of particulate vaccines (39, 40). The capacity of Ag-loaded DCs to induce effective tumor immunity is well established (41, 42). We also demonstrated that DCs have a remarkable capacity to take up cell-associated Ags directly from tumor cells through a cell contact-dependent mechanism, and can generate therapeutically effective Ag-specific antitumor immune responses against these Ags (43, 44). As yet, few DC Ag-loading strategies have been developed to use the potentially advantageous properties of activated DC populations. Effects of specific activation signals such as CD40-cross-linking (14, 45, 46, 47), LPS (48, 49), or reculture (6, 13) on subsequent particulate Ag uptake and presentation have not been defined. In this study, we demonstrate that subsets of DCs in activated DC populations can actively phagocytose particulate Ags. Internalization of particulate Ags by activated DCs initiates a remodeling process that results in dispersion of clustered lysosomes. Interestingly, although the phagocytic capacity of mature DCs appears to be maintained by DCs activated in response to a variety of activation stimuli, presentation of class I and II ligands from phagocytosed particulates is achieved when activation is induced in the context of DC reculturing or CD40 ligation, but not when DCs are activated through exposure to LPS. These results suggest the existence of distinct regulatory pathways that maintain phagocytic, but not endocytic, pathways of Ag entry in activated DCs and that Ag presentation of particulates varies with the maturation signal. Taken together, these results provide new insight into DC biology and contribute to the ongoing development of particle-based immunization strategies, including novel approaches designed to simultaneously deliver Ag and DC-activating adjuvants in vivo.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
C57BL/6 (H-2Kb) female mice, 8–16 wk old, were purchased from The Jackson Laboratory and housed in the pathogen-free Central Animal Facility at the University of Pittsburgh. Mice were used in experiments according to institutional guidelines.
Bone marrow-derived DC (BMDC) differentiation and maturation in vitro
BMDCs were prepared as described previously (50). Briefly, bone marrow progenitor cells were collected and depleted of T/B lymphocytes using mAbs to CD4, CD8, and B220 markers, and rabbit complement. Cells were resuspended to 0.5–0.7 x 106 cells/ml in complete RPMI 1640 containing murine (m)GM-CSF/mIL-4 (1000 U/ml each) and cultured in T150 flasks at 37°C in 5% CO2 over 5 days. To produce activated DC populations, primary BMDCs after 5 days of culture were harvested from flasks leading to cluster disruption, washed, and further replated in flasks for an additional 2 days in the presence of the following: 1) cytokine-supplemented RPMI 1640 alone (reculture); 2) anti-CD40 IgM (2.5 µg/ml; BD Pharmingen); and 3) LPS (from Salmonella enterica serovar Typhi S. typhi, 50–100 µg/ml; Sigma-Aldrich). For the immature DC group (day 5 DC), 2 days after the above cells were already in culture, fresh bone marrow was harvested and plated in the presence of mGM-CSF/mIL-4 as described. This staggered, metachronous BMDC preparation allowed all four cell populations under study to be collected and used for head-to-head comparison experiments on the same day. Cytokines were a gift from Shearing-Plough or were generated in our laboratory from supernatants of CHO cells transiently transfected with expression vectors encoding for GM-CSF and IL-4 (National Gene Vector Library, University of Michigan).
Phenotype of BMDCs by flow cytometry
Immature (day 5) or mature (day 7 ± maturation signal) BMDCs were stained for FACS analysis using primarily I-Ab-FITC, B7.2-PE, CD11c-biotin (BD Pharmingen), and streptavidin-Tricolor (Caltag Laboratories) as an indirect label. Mouse IgG1-FITC, rat IgG2b-PE, and biotinylated hamster IgG served as isotype staining controls in all studies. BMDC preparations were deficient in cells expressing markers for B220, CD4, CD8, and Mac-3 in agreement with previous reports. All cell surface labeling was performed in PBS at 4°C, 30 min, and cells fixed in buffered 1% paraformaldehyde before FACS acquisition (BD Biosciences).
Mature DC enrichment over metrizamide gradients
Mature BMDC populations were routinely enriched to 
90% purity using repeated 14.5% metrizamide gradients. Grade 1 metrizamide powder (Sigma-Aldrich) was resuspended to 14.5% w/v in PBS devoid of Ca2+ and Mg2+ and sterile filtered through a 0.22-µm filter. For enrichment, 10–20 million day 7 BMDCs were resuspended in 25 ml of RPMI 1640 and carefully layered atop 12.5 ml of 14.5% metrizamide in 50 ml conicals using Pasteur pipettes. Gradients were centrifuged at 1.9–2K rpm, 20 min, room temperature (RT), with mature DCs collected from the interface layer. Typically, single-pass metrizamide gradients yielded a purity of 75–85% for mature BMDC. Centrifugation through metrizamide gradients was repeated two additional times to further enrich mature DC subpopulations to >90% purity as demonstrated in the text.
Uptake assays for soluble fluoresceinated ligands
Mannosylated polyacrylamide (man-PAA) (51) is an inert, small m.w. Ag internalized by endocytosis (51), and was a gift from Dr. X. Dong (University of Pittsburgh, Pittsburgh, PA). DQ-OVA (Molecular Probes) is a self-quenching macromolecule containing a BODIPY-fluorophore that will fluoresce if the marker Ag is internalized into low pH endosomes and/or is cleaved through proteolysis. Tracers were combined with immature or mature BMDC populations for 2 h in one of three conditions: 1) 4°C (minimal endocytosis), 2) 37°C (maximal endocytosis), and 3) 37°C plus cytochalasin D (CCD; inhibitor of microtubule polymerization). A total of 2 x 105 DCs/well was preincubated at 4°C or 37°C in the presence or absence of CCD (25 µg/ml) for 45 min before addition of man-PAA or DQ-OVA (10 µg/ml final concentration). Internalization was allowed to proceed for 2–3 h at the specified temperatures and conditions with occasional agitation, followed by four consecutive washes in excess ice-cold PBS containing 0.01% azide to remove uninternalized ligand. DC populations were then kept on ice and labeled using CD11c-biotin primary Ab followed by streptavidin-Tricolor and B7.2-PE secondary reagents. Samples were maintained at RT for an additional hour, fixed, and acquired using three-color flow cytometry on a FACSCalibur instrument (BD Biosciences). Events contained in a typical forward scatter/side scatter DC profile to eliminate debris/dead cells were then additionally gated for B7.2dimCD11c+ (immature) or B7.2brightCD11c+ (mature) expression patterns.
Cytometric particle phagocytosis (CPP) assay
To determine specific particulate Ag internalization, we devised a method to assess internalized from cell surface-associated beads for quantitative flow cytometric analysis. A total of 5 x 105 APC was added to each of 3–15 ml of conical tubes and exposed to one of three conditions in incomplete RPMI 1640 at 37°C: no inhibitor, CCD (stock resuspended in DMSO) at 25 µg/ml, or an equal volume of DMSO alone to serve as mock inhibitor. After a 30-min preincubation, 3 µl (100 beads/cell) of 1-µm diameter, carboxylated FITC-laden latex beads (Polysciences) were introduced to each cell system and briefly pelleted for 5 min, 1200 rpm, to adsorb cells with lower density particles. After an additional 20-min incubation at 37°C, the cell-bead mixture was vortexed, underlaid (1/1 v/v) with FCS, and centrifuged 10 min, 37°C, which separates 90% of free latex particles (interface layer) from those more strongly associated with BMDCs (cell pellet). Supernatant was then carefully aspirated, and cells were replaced at 37°C and resuspended in their starting buffer (±inhibitors) to allow continued phagocytosis of the remaining beads for 2–3 h. After this time, all cells were brought to 4°C, stained for BMDC surface markers (see above), and fixed in 1% overnight. Before FACS, vortexed cells were centrifuged through an FCS protein gradient to remove residual uninternalized latex beads. Latex microspheres exhibit discrete peaking or banding fluorescence when cell-associated (52, 53). Nonphagocytic tumor cell lines such as EL4 thymoma and B16 melanoma were found to be incapable of particulate phagocytosis using this assay (data not shown). In some experiments, Toxin B (1 ng/ml), wortmannin (100 nM), bradykinin (1 µM), and PMA (60 mg/ml) were incubated with DCs before particle introduction to assess PI3K and RhoGTPase contributions to phagocytosis (data not shown).
Timed latex bead phagocytosis
CPP was performed essentially as described with all samples at 37°C with or without inhibitor in 15-ml tubes pulsed with 100 beads/cell at time zero. One and 3 h after the start of the assay, tubes were transferred to ice to inhibit additional internalization. Cells at time t = 0 were immediately placed on ice before bead addition to other samples. Staining using CD11c-biotin followed by B7-2-PE and avidin-Tricolor was then simultaneously performed on all samples following the 3-h time point indicated. Cells were fixed in 1% paraformaldehyde following cell surface Ab labeling and FACS acquisition.
Quantitation of DC surface-adherent and intracellular particles using a membrane marker and imaging software
Metrizamide-separated DCs, with or without CCD pretreatment, were allowed to phagocytose 1 µm of FITC-latex particulates, and later were adhered onto glass coverslips coated with poly-L-lysine (total time of phagocytosis between 6 and 9 h). Coverslips were then counterstained with rhodamine phalloidin as a membrane reference, and fixed onto glass slides using Gelvatol mounting medium. Images (magnification, x60) of cells with beads were collected using an Olympus Fluoview 1000 scanning confocal microscope (Olympus America). An algorithm to count, in an automated fashion, the number of beads both within and on the membrane of DCs was then programmed into the Metamorph imaging data analysis software application (Molecular Devices). Phalloidin staining allows segmentation of the cell periphery such that individual cells may be automatically delineated. Subsequently, the number of internalized beads was counted automatically using the segmented cells as defined regions. To define individual beads, the program was trained to recognized the beads as thresholded objects of defined size and shape. To quantify the number of beads at the cell surface, the periphery of the cell was defined automatically using the actin cytoskeleton, and a binary mask was generated extending from the periphery of each cell by 1.2 µm. The bead image was thresholded and a binary mask generated. Using a Boolean AND statement, the beads that were found within the 1.2-µm periphery of the cell were defined. A similar algorithm as defined was implemented for quantifying beads within a given cell. Using this algorithm, we analyzed bead-associated DCs to assess intracellular internalization of beads. An average of 99 DCs per treatment group were assessed for particle internalization at a magnification of x60. For each cell analyzed, the number of internalized and membrane- associated beads was determined. For each condition, we determined the percentage of bead-associated cells that internalized beads ((number of cells with internalized beads/total number of cells associated with beads) x 100), and the percentage of cell-associated beads that were internalized ((number of internalized beads/(number of internalized beads + membrane bound beads)) x 100) per cell.
Confocal microscopy of BMDC populations after intracellular staining for CD107b/LAMP-2
Particulate Ags were initially prepared by covalent coupling of DQ-OVA protein onto 4',6'-diamidino-2-phenylindole (DAPI) fluorescent 1-µm latex beads (Polysciences) following the manufacturer’s instructions for carbodiimide coupling of carboxylated microspheres. This linking procedure unfolds the DQ-OVA molecule to yield a particle of strong, pH-insensitive BODIPY fluorescence. Latex particles were coincubated with day 5 or day 7 metrizamide-enriched, LPS-activated DCs as described under CPP above, for 2 or 14 h, respectively, in the absence or continual presence of CCD. At the indicated time points, BMDCs were passed through a pure FCS gradient, and 2–5 x 106 cells were "grown" on poly-L-lysine-treated coverslips in 100 µl of PBS for 20' and RPMI 1640 for an additional 40 min, followed by fixation of cells overnight in 1% paraformaldehyde. For intracellular LAMP-2 staining, cells were permeabilized with 0.5% saponin in PBS for 30 min, RT, and labeled in a three-step sequence in saponin-containing medium: 1) purified rat anti-LAMP-2 (1:2500; BD Pharmingen); 2) anti-rat-biotin (1/400; Vector Laboratories); and 3) streptavidin-Cy3 (1/10,000; Vector Laboratories). Purified rat anti-TNP served as the isotype control Ab. After removal of saponin using FACS buffer, coverslips were mounted onto glass slides using Gelvatol mounting medium and imaged using a Leica DM IRBE confocal microscope equipped with TCS NT software. Images were acquired using a x100 objective in differential interference contrast and fluorescence and analyzed using Adobe Photoshop image editing and analysis software.
Digital live cell image acquisition of particle capture
Metrizamide-enriched, day 7 recultured BMDCs (5–10 x 105 cells) were resuspended in 1 ml of RPMI 1640 containing 50 nM Lysotracker reagent and incubated on 40-mm glass coverslips (Fisher Scientific) at 37°C for 45 min to allow initial DC adherence. Coverslips were mounted in a FCS2 Closed System live cell chamber (Bioptechs) and subjected to a perfusion rate of 0.5 ml/h complete RPMI 1640 prewarmed to 37°C. Five independent stage positions on a Multimode Inverted Nikon TE300 microscope (Nikon) were identified for Lysotracker+ DC before addition of 10 x 106 1-µm latex beads loaded with DAPI (BB latex beads; Polysciences). Multiparametric sequential fluorescent images at each stage position were acquired at 5-min intervals following bead infusion. Images were then analyzed using Metamorph software (Molecular Devices).
Confocal microscopy of iron oxide particle phagocytosis
Activated iron beads were labeled with Alexa Fluor 546 carboxylic acid succinimidyl ester (Molecular Probes) by suspending 500 µl of beads in 5 ml of PBS and adding Alexa Fluor 546 to a final concentration of 8 µM. The dye was allowed to conjugate to the beads for 1 h on a rocker at RT. Labeled beads were washed in PBS, and remaining free amines were blocked with glycine buffer. Beads were then washed extensively in PBS (more than five times) and suspended in 1 ml of PBS for use in experiments. An aliquot of labeled beads was examined under fluorescence microscopy to verify labeling. Day 5 DC and mature DC groups (after metrizamide purification) were suspended in RPMI 1640 containing 1000 U of GM-CSF/IL-4 at a concentration of 106 cells/ml and incubated with Alexa-labeled iron beads (1 µl of beads/200 µl of RPMI 1640) and incubated overnight at 37°C. Cells were collected and stained with 5-chloromethylfluorescein diacetate (CMFDA) (Molecular Probes), and bead-containing cells were purified using magnetic columns (Miltenyi Biotec). Cells were examined by confocal microscopy at the Center for Biologic Imaging at the University of Pittsburgh. Images were analyzed and processed with Metamorph software (Molecular Devices) to generate a montage of cells demonstrating typical iron-bead phagocytosis in a single intracellular plane.
Class I/II presentation using Ag-specific hybridomas
Day 5 or activated day 7 BMDCs were resuspended in RPMI 1640 and pulsed with 100 µg/ml (39) particulate chicken egg OVA or the respective OVA-derivative peptides SIINFEKL (OVA257–264, class I; 10 ng/ml) or ISQAVHAAHAEINEAGR (OVA323–339, class II; 1 µg/ml) each in the presence or absence of brefeldin A (10 µg/ml) or CCD (25 µg/ml). After overnight Ag processing (20 h), cells were washed, pelleted, fixed in 0.5% paraformaldehyde for 1 h to inhibit further Ag uptake or processing, and washed three times to remove fixative. Particle-pulsed DCs were then divided into two pools to examine Ag presentation in class I and II from the same cell population. DCs were plated at the indicated cell numbers over serial-2 dilutions in 96-well flat-bottom plates. OVA peptide-specific T-T hybridoma cell lines—H-2Kb-restricted 33.70 (MHC class I) or I-Ab/d-restricted DO11.10 (MHC class II)—were then individually paired at 0.8 x 106 cells/well with Ag-bearing APCs. IL-2 production from hybrids is strictly dependent upon presence of MHC-peptide complexes and independent of costimulatory molecule expression. After a 24- to 30-h incubation at 37°C, 100 µl of culture supernatant was collected and frozen at –80°C until assayed for IL-2 by [3H]thymidine incorporation (NEN-DuPont) with the HT-2 indicator cell line. All assays were performed in duplicate wells and repeated in at least two independent experiments unless otherwise indicated.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
To generate CD11c+ DCs from bone marrow precursors, we cultured lymphocyte-depleted bone marrow cells in medium supplemented with GM-CSF and IL-4. After 5 days (day 5 DC), loosely adherent clusters of immature cells with small veiled processes were observed as described previously (50). DCs at this point were removed by gentle pipetting and recultured in either fresh cytokine-supplemented medium (recultured DC) or medium containing the agonist anti-CD40-IgM cross-linking Ab (anti-CD40 DC) or LPS (LPS DC) for an additional 36- to 40-h period to promote terminal DC maturation/activation in vitro. Enrichment for activated/mature DCs using 14.5% metrizamide density gradients resulted in a DC population that is 90% B7.2brightCD11c+ (Fig. 1A). These DCs were extensively veiled, exhibited clustered MIIC/lysosomal structures typical of DC maturation, and were >99% viable as determined by trypan blue exclusion (data not shown).
|
Mature DC demonstrate reduced capacity to internalize soluble ligands but can internalize particulates
To compare immature and activated/mature DC populations for macropinocytic and phagocytic Ag uptake, we used fluoresceinated soluble ligands (54) and fluorochrome-loaded 1-µm diameter latex microspheres, respectively. As expected, immature DCs (day 5) demonstrated a marked shift in the FL1-fluorescence when cultured with labeled man-PAA short chain soluble polymers, indicative of rapid internalization of man-PAA (Fig. 2A, top panel). Similar results were obtained when DCs were exposed to soluble proteins as demonstrated by the internalization of DQ-OVA (Fig. 2B, top panel), a protein that becomes fluorescent upon denaturation and proteolysis. Uptake of both man-PAA and DQ-OVA was significantly inhibited by incubation at 4°C or the presence of CCD, confirming that the fluorescence shift is due to active internalization dependent on microtubule polarization rather than from membrane association alone. As anticipated, under the same experimental conditions, activated/mature DC populations did not significantly internalize either the polyacrylamide ligand (Fig. 2A, bottom panels) or DQ-OVA protein (Fig. 2B, bottom panels). For man-PAA, no significant difference in internalization was observed between cells incubated at 4°C or in the presence of CCD, and those incubated at 37°C. DQ-OVA fluorescence at 37°C was not inhibited following microtubule depolymerization, suggesting that the cell-associated fluorescence observed at 4°C was primarily due to the association of exogenously degraded or denatured DQ-OVA with the surface of the plasma membrane. Of note, we also observed a dramatic decrease in fluorescent macropinocytic vesicles in cytospins of Ag-exposed mature DCs compared with immature DC populations (data not shown).
|
|
To systematically "count" the number of internalized vs cell membrane-associated latex particles for three DC subpopulations, we developed a new approach using microscopy and the Metamorph software application (Table I). Consistent with flow cytometry data, all subpopulations have subsets of cells that participate in phagocytosis, and CCD exposure significantly diminishes the proportion of cells that internalize latex beads. Activated DCs were noted to be less efficient at bead internalization on an individual cell basis. Furthermore, we found that in each DC population, cells that were associated with beads internalized a similar proportion of associated beads (98, 96, and 91%, respectively), and CCD reduced internalization of cell-associated beads proportionately in each DC population (to 20, 32, and 32%, respectively) (Table I). CCD also reduced the proportion of DCs that internalized cell-associated beads. Interestingly, although CD40-stimulated DCs internalized similar proportions of cell-associated particles, substantially fewer bead-associated DCs internalized beads. This suggests that a smaller proportion of DCs activated in the presence of CD40 maintain phagocytic capacity than is evident from flow analysis.
|
DCs condense MIIC/lysosomes into perinuclear clusters following terminal maturation, and the formation of LAMP-1,2+ clusters has been used as an independent marker of DC activation (55). Using confocal microscopy, we observed as expected that day 5 immature DCs lacked lysosomal clusters (Fig. 4). Day 5 DC have punctate LAMP-2+ lysosomes broadly distributed throughout the cell that rapidly associate with microspheres upon internalization (t = 2 h; Fig. 4A) and after overnight culture (t = 14 h; Fig. 4C). Phagocytosis is prevented with CCD pretreatment, with particles limited to the bone marrow DC plasma membrane (Fig. 4, B and E).
|
To verify cytoskeletal plasticity and lysosome remodeling events in terminally mature DCs, we used live cell imaging to view particulate Ag phagocytosis in real time. Lysotracker reagent (Fisher Scientific) was used to detect low-pH MIIC/lysosomes within DCs, followed by infusion of 1 µm of DAPI-loaded latex particles. In a mature DC that phagocytoses particulates (Fig. 5A), within 85 min, six beads occupy a polarized position to that of the eccentric lysosome cluster with the mature DC appearing to flatten or "tether" to the coverslip. Phagocytosis from this point was rapid, with the first beads captured within 20–30 min by mature DCs having MIIC clusters. Capture and internalization of all particulates by this cell occurred for 2 h after the initiation of the assay, at which time lysosomes are dispersed throughout the cytoplasm, indicated by diffuse and diminished Lysotracker fluorescence intensity. As suggested by our previous results, we also observed mature DCs in the same population that did not appear to be highly phagocytic during the time observed. In mature DCs that did not internalize particles over the same time course, the perinuclear clustered morphology of intensely stained lysosomes was retained (Fig. 5B). These results provide additional support to the interpretation that a subset of activated DCs can internalize particulate Ags and that the action of particulate phagocytosis leads to diffusion of LAMP-2+ lysosomal clusters.
|
In addition to latex particulates, we covalently labeled the surface of iron oxide particles with Alexa Fluor reagent (Alexa-iron) and DC populations with CMFDA, a fluorophore taken up and retained within the cytosol. We exposed day 5 and metrizamide-enriched day 7 DC populations to Alexa-iron beads overnight under conditions similar to those used in future functional assays, and examined cell populations under confocal microscopy at an intracellular plane. All four DC populations shown, whether immature or activated, readily internalized iron particulates, and multiple cells from each population are seen in each image (Fig. 6).
|
Based on the above findings, to better evaluate processing and presentation of particulate Ags, four populations of DC (day 5 DC, LPS DC, CD40 DC, and recultured DC) were pulsed overnight with 1 µm of iron oxide particles covalently bonded to OVA protein (OVA-iron) or soluble SIINFEKL peptide, fixed, and incubated with OVA MHC class II (Fig. 7) or class I (Fig. 8)-restricted T-T hybridomas to determine formation and presentation of peptide-MHC ligands. Day 5 DCs efficiently presented particulate OVA through the class II-restricted processing pathway (Fig. 7). Interestingly, both recultured and CD40-ligated mature DCs processed and presented particulate OVA through the class II pathway, but DC populations matured in the presence of LPS did not. Coincubation of DCs with CCD in the presence of particulate Ag blocked Ag-specific hybridoma stimulation, suggesting that presentation required internalization and processing of the particulates. All DC groups pulsed with OVA323–339 peptide strongly stimulated the hybridoma in the presence or absence of CCD (Fig. 7), excluding the possibility of bystander toxicity to the hybridoma detector cell line by CCD.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Our results with DCs concur with previous reports that immature, but not mature, DCs efficiently acquire soluble protein Ags (4, 33, 56, 57). Indeed, our activated DC subpopulations were found to be functionally mature with respect to Ag uptake of soluble ligands using man-PAA and DQ-OVA (Fig. 2). The formation of solute-bearing vesicles characteristic of macropinocytosis has been shown to be largely dependent on cdc42 GTPase, which is nearly undetectable in mature DCs (32). However, we find that some DCs that are exposed to maturation signals and appear phenotypically, morphologically, and functionally activated, retain an ability to phagocytose and process particulates, suggesting discrete regulation of macropinocytosis and phagocytosis in DCs. Molecular events contributing to the process of Ag internalization are best understood in macrophages, where CR3 - and FcR-mediated mechanisms of particle uptake are well characterized. Phagocytosis through CR3 is associated with the formation of paxillin- and vinculin-sequestering podosomes at sites of particle attachment, and invagination of plasma membranes to "passively sink" particles into phagocytes (58, 59). FcR-mediated phagocytosis instead couples focal phagocytic actin cups with filopodial and lamellipodial membrane processes to extend the outer membrane around a particle (60, 61, 62, 63). Both types of receptor-mediated phagocytosis are finely regulated by distinct Rho subfamily GTPases such as Rho, Rac1, and cdc42 (59), although it is not apparent whether either pathway is involved in phagocytic signaling within DC populations or upon internalization of inert latex and iron-oxide particles (59, 60). Once internalized, particulates within DCs become degraded by proteases within phagosomes, which can also fuse with ER organelles to incorporate molecules such as TAP, tapasin, and SEC61 (37, 38). We have performed preliminary studies to investigate the contributions of the PI3K and RhoGTPase pathways in the phagocytic activity observed. We evaluated the effects of Toxin B, a RhoGTPase family inhibitor, wortmannin, an inhibitor of PI3K, and PMA, a RhoGTPase activator (23, 60, 64) on phagocytosis in these DC populations. In these experiments, exposure to either wortmannin or to Toxin B led to a significant inhibition of particle uptake of all DC populations, suggesting that there are components of both the PI3K-dependent and RhoGTP-dependent mechanisms involved in the phagocytic activity of DCs exposed to activation signals (data not shown). Clearly, further studies will be necessary to understand the distinct pathways available for particle phagocytosis in these DC subsets, and the efficient cross-presentation of Ags associated with internalized particulates.
Within each activated DC group, we have observed that not all cells participate in phagocytosis after particle administration (Fig. 3), providing direct evidence of functional heterogeneity in these DC populations. Cells that merely attach particles may ostensibly form phagocytic cups with particles but lack the cellular machinery to complete phagocytosis. DCs refractory to particle association, by contrast, may be devoid of additional requisite elements that coordinate Ag phagocytosis. It is conceivable that the presence or activity of select RhoGTPases exist along a continuum, corresponding to the heterogeneity of the phagocytic proportion of a given matured DC population, and are recruited following strong particle adsorption and receptor engagement along the plasma membrane. This heterogeneity in phagocytosis within metrizamide-enriched DC populations is quite distinct from the uniform particulate Ag uptake activity seen in macrophages (35), and may explain why particulate Ag internalization by activated DCs has been difficult to appreciate.
Using imaging to quantify latex bead internalization, we found a generally similar pattern of particulate Ag uptake when comparing day 5, LPS-activated, and CD40-activated DC populations, which was consistent with our flow cytometry data. All three DC subpopulations participated in phagocytosis, and notably exposure to the inhibitor CCD reduced both the percentage of cells that were internalizing beads and the fraction of cell-associated beads internalized (Table I). Interestingly, we found that in each DC population, cells that were associated with beads phagocytosed a similar proportion of the beads (98, 96, and 91%, respectively), suggesting that under these conditions, flow cytometric analysis gave a reasonably accurate quantification of bead internalization. CCD exposure also limited uptake of cell-associated beads proportionally in each DC subset (to 20, 32, and 32%, respectively). CCD additionally reduced the proportion of DCs that internalized beads that were associated with cells, and specifically these data suggest that in the presence of cytochalasin, approximately two-thirds of bead-associated DCs are not aggressively phagocytic, perhaps accounting for the background observed for these groups in the flow analysis. Interestingly, substantially fewer particle-associated CD40 DCs internalized beads, indicating that a smaller proportion of CD40-stimulated DCs maintain phagocytic capacity than is evident from flow analysis.
Recent reports confirm Ag acquisition by terminal DCs residing in draining lymph nodes (65) and tonsils (66) as well, although the maturation state of DCs that internalized particles was not directly defined. Our observation of efficient Ag presentation in functional assays, despite the fact that only a subset of cells appear to be phagocytic by cytometric assays, implies that either a small proportion of DCs are quite efficient in Ag processing and MHC loading, as suggested previously (35, 67), or that other undefined mechanisms exist for presentation.
We used two types of particles for this study: latex particles in flow cytometry, confocal microscopy, and live cell experiments; and fluorochrome-labeled iron oxide (Alexa-iron) particles in both confocal experiments and functional presentation assays. There are advantages and limitations to both particulates: Fluorescent microspheres are well documented for "peaking" fluorescence with successive numbers of beads displaying brighter fluorescence emission. These latex particles are low in density, making irrelevant beads separable from cell-associated beads through density gradients like FCS. Alexa-iron particles, in contrast, do not have a uniform shape, and covalently attach (rather than encapsulate) fluorescent molecules. Also, there is no method that allows separation of cell-associated iron particles from free particles; iron is too dense for gradient separation, and magnetic fields will attract beads both contained within cells, and attached to cells, equally.
Both of the latter characteristics make iron particles unsuitable for quantitative analysis and FACS. In contrast, our experience has been that protein Ag amine conjugation onto iron oxide beads is highly efficient, with conjugation of 20–50 mg of OVA onto 1 ml of suspended iron beads. Carbodiimide conjugation onto latex particles has limited efficiency, with covalent linkage of only 0.1–0.3 mg of OVA using related protocols. For the above reasons, we felt that the most informative experiments would be achieved by using two distinct particulate Ag formulations for disparate purposes. To confirm that there were no overt differences in phagocytosis from particulate formulation, we exposed cells to both types of particles and analyzed single cells in detail (Fig. 5) and cell populations (Fig. 6).
Our findings suggest that particulates first enter activated DCs through an action that is initially independent from LAMP/lgp/MIIC endolysosomal compartments (Figs. 4 and 5). With time, clustered lysosomes found beside the nucleus disaggregate to colocalize with the internalized particulates. This remodeling event was unexpected because remnant MIIC are generally considered nonfunctional after terminal maturation. The cytoskeletal machinery involved in this reorganization is unknown. Remodeling activity occurred in the context of particle uptake and was dramatic given the terminal activation state of the DCs.
Finally, we find that although these DC populations have somewhat comparable capacities for internalizing particulate Ags, the Ag presentation efficiency of activated DC populations varies depending on the activation signal (Figs. 7 and 8). Use of the peptide controls in these studies makes it unlikely that differences in Ag presentation could be due to a disparity in MHC molecule availability between DCs or toxic effects of inhibitors on either DC or T cell populations. Our knowledge of intracellular proteolytic activities and MHC class I/II loading capabilities in mature DCs is limited, although specific cathepsins are clearly involved in peptide generation (7, 16). Although activation stimuli typically induce rapid, coordinated changes in DCs, it should be appreciated that maturation signals have qualitatively different effects. For instance, CD40 ligation stimulates IL-12p40 production by DCs, and allows DCs to bypass the requirements for T cell help in priming CTLs (45, 46, 47, 68). LPS stimulates expression of the IL-12 p35 and p40 chains, and LPS exposure has been associated with both DC apoptosis in vivo (69) and NF-
B activation and DC survival in vitro (70).
There is evidence that T cell responses to specific Ags are diminished in the presence of LPS (71, 72, 73), but the mechanisms responsible for these findings have not been elucidated. LPS, a membrane component of Gram-negative bacteria, appears to evoke cellular responses following intracellular shuttling in association with soluble CD14 found in serum (74). This LPS-soluble CD14 complex was found to specifically localize to the Golgi apparatus rather than lysosomal or endoplasmic organelles (75). More recent work has also implicated LPS exposure with delayed receipt of proteases to phagosomes within DCs (7). We demonstrate that both MHC class I- and II-restricted presentation is impaired (Figs. 7 and 8) despite phagocytosis (Fig. 3) and lysosome colocalization (Figs. 4 and 5) in LPS-activated DCs. Because both of these Ag presentation pathways converge at the Golgi apparatus before downstream routing, it is tempting to speculate that LPS, through steric or idiopathic inhibitory effects, could be interfering with proper MHC class I and II molecule intracellular routing or processing by its presence at the common trafficking interface within the Golgi apparatus. Although the relevance of Golgi transport of LPS on Ag presentation is not directly addressed herein, other microenvironmental cues, such as IL-10, have been shown to impair presentation of phagocytosed Ags without influencing uptake even in the presence of CD40L (76), arguing that the context in which a maturation stimulus is encountered can influence presentation of particulate Ags.
By contrast, reculture or CD40 cross-linking promote loading of peptides onto newly synthesized or available MHC class I/II in activated DCs as we have shown. Our results with particulate Ag are in agreement with previous work on soluble Ag, where researchers found that CD40 cross-linking and DC cluster disruption induced cross-presentation of exogenous Ags (6). In contrast, we observed inhibition of MHC class II-restricted presentation of particulate Ags following LPS activation, possibly suggesting a different level of control over the mechanisms for presentation of soluble vs particulate Ags in the class II presentation pathway. Unlike DCs prepared from peripheral blood, BMDCs are labile cells, capable of undergoing spontaneous maturation in vitro (12, 13). Little is known about the reculture maturation signal, although others have speculated that this action may mimic changes associated with DC migration through the disruption of E-cadherin (6). A T cell-related signal through engagement of CD40 might be expected to maintain Ag presentation for future lymphocyte triggering as we have shown. These experiments, taken in whole, suggest that activated DC populations exhibit considerable heterogeneity with respect to uptake of large Ags, remodeling of cytoskeletal organelles, and Ag presentation. This may be an important consideration for the ongoing development of immunization strategies designed to both deliver Ag- and DC-activating "adjuvant" signals to DC populations simultaneously.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by grants from National Institutes of Health National Cancer Institute, National Institute of Arthritis and Musculoskeletal and Skin Diseases, and National Institute of Allergy and Infectious Diseases (to L.D.F.). 
2 Address correspondence and reprint requests to Dr. Louis D. Falo, Jr., Department of Dermatology, University of Pittsburgh School of Medicine, 145 Lothrop Hall, Pittsburgh, PA 15217. E-mail address: lof2{at}pitt.edu
3 Abbreviations used in this paper: DC, dendritic cell; MIIC, MHC class II compartment; LAMP, lysosome-associated membrane protein; ER, endoplasmic reticulum; BMDC, bone marrow-derived DC; m, murine; RT, room temperature; man-PAA, mannosylated polyacrylamide; CCD, cytochalasin D; CPP, cytometric particle phagocytosis; DAPI, 4',6'-diamidino-2-phenylindole; CMFDA, 5-chloromethylfluorescein diacetate; int, intermediate; MFI, mean fluorescence intensity; PCHO, paraformaldehyde.
Received for publication November 12, 2004. Accepted for publication September 29, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
v
5 integrin recruits the CrkII-Dock180-rac1 complex for phagocytosis of apoptotic cells. Nat. Cell Biol. 2: 899-905. [Medline]
