HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) A correction has been published Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ratzinger, G. Articles by Young, J. W. Search for Related Content PubMed PubMed Citation Articles by Ratzinger, G. Articles by Young, J. W.

Mature Human Langerhans Cells Derived from CD34⁺ Hematopoietic Progenitors Stimulate Greater Cytolytic T Lymphocyte Activity in the Absence of Bioactive IL-12p70, by Either Single Peptide Presentation or Cross-Priming, Than Do Dermal-Interstitial or Monocyte-Derived Dendritic Cells¹

Gudrun Ratzinger^2,3,*,, Jan Baggers^2,*,, Maria A. de Cos^4,*,, Jianda Yuan^*,, Tao Dao^*,, John L. Reagan^*,, Christian Münz^||, Glenn Heller^,¶ and James W. Young^5,*,,,¶

^* Laboratory of Cellular Immunobiology, Allogeneic Bone Marrow Transplantation and Clinical Immunology Services, Division of Hematologic Oncology, Department of Medicine, and Biostatistics Service, Department of Biostatistics and Epidemiology, Memorial Sloan-Kettering Cancer Center, ^¶ Weill Medical College of Cornell University, and ^|| Laboratory of Viral Immunobiology, The Rockefeller University, New York, NY 10021

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The emerging heterogeneity of dendritic cells (DCs) mirrors their increasingly recognized division of labor at myriad control points in innate and acquired cellular immunity. We separately generated blood monocyte-derived DCs (moDCs), as well as Langerhans cells (LCs) and dermal-interstitial DCs (DDC-IDCs) from CD34⁺ hematopoietic progenitor cells. Differential expression of CD11b, CD52, CD91, and the CD1 isoforms proved useful in distinguishing these three DC types. All mature DCs uniformly expressed comparable levels of HLA-DR, CD83, CD80, and CD86, and were potent stimulators of allogeneic T cells after exposure either to recombinant human CD40L trimer or a combination of inflammatory cytokines with PGE₂. moDCs, however, required 0.5–1 log greater numbers than LCs or DDC-IDCs to stimulate comparable T cell proliferation. Only moDCs secreted the bioactive heterodimer IL-12p70, and moDCs phagocytosed significantly more dying tumor cells than did either LCs or DDC-IDCs. LCs nevertheless proved superior to moDCs and DDC-IDCs in stimulating CTL against a recall viral Ag by presenting passively loaded peptide or against tumor Ag by cross-priming autologous CD8⁺ T cells. LCs also secreted significantly more IL-15 than did either moDCs or DDC-IDCs, which is especially important to the generation of CTL. These findings merit further comparisons in clinical trials designed to determine the physiologic relevance of these distinctions in activity between LCs and other DCs.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Progress in the generation of dendritic cells (DCs)⁶ from defined hematopoietic precursors has revealed the heterogeneity of these APCs, which are critical to the onset and control of T cell immune responses (1, 2, 3, 4, 5). Human DCs comprise at least four cell types, which are CD14⁺ blood monocyte-derived DCs (moDCs), CD34⁺ hematopoietic progenitor cell (HPC)-derived dermal or interstitial DCs (DDC-IDCs), CD34⁺ HPC-derived Langerhans cells (LCs), and plasmacytoid DCs (4, 6, 7, 8, 9, 10, 11, 12, 13).

moDCs, LCs, and DDC-IDCs represent the classic immunogenic DCs, often called "myeloid" DCs. DC1 formally refers to moDCs, and DC2 denotes plasmacytoid DCs, because of their respective capacities to polarize Th1- or Th2-type responses (14). The term DC1 did not originally apply to LCs and DDC-IDCs, but these share many phenotypic (lineage^neg, HLA-DR^bright, CD11c⁺, CD14^neg, CD33⁺, CD83⁺, CD123^low) and functional properties with DC1 or moDCs, which are distinct from those of DC2 (lineage^neg, HLA-DR⁺⁺, CD11c^neg, CD33^neg, CD83⁺, CD123^bright). Both LCs and DDC-IDCs also express typical myeloid-type DC markers with the exception that human LCs lack CD11b (7). LCs also express e-cadherin (15, 16) and Langerin (CD207), the latter of which is pertinent to the formation of Birbeck granules (17, 18).

DCs that develop from CD34⁺ HPCs in the presence of GM-CSF and other cytokines differentiate along two pathways (6, 7, 8): LCs develop via a CD1a⁺CD14^– intermediate, whereas CD34⁺ HPC-derived CD1a^–CD14⁺ intermediate precursors develop into DDC-IDCs or alternatively into macrophages (6, 7, 8). TGF1 supports LC differentiation (9, 10), a requirement first discovered using TGF1 null mice (19).

Because DDC-IDCs and blood moDCs both develop from a CD14⁺ intermediate or precursor (6, 7, 8), investigators have assumed that DDC-IDCs and moDCs are homologous, if not the same. However, pursuant to our recent finding that monocyte precursors and moDCs express CD52 targeted by alemtuzumab, whereas LCs and DDC-IDCs never express this epitope either in vitro or in vivo (20), we have begun to find other phenotypic and functional distinctions between these three conventional or myeloid DC types. This area merits investigation because of the distinct cytokine profiles, chemokine responsiveness, and functional segregation of at least LC and DDC-IDC émigrés to discrete T cell or B cell predominant areas of secondary lymphoid organs (21, 22, 23). A bias has also emerged in some vaccine trials that inclusion of CD34⁺ HPC-derived DCs is beneficial, presumably because these include LCs (24). The basis for such benefit, however, has not been established; and in practice, the blood monocyte precursors of moDCs are more readily available and more widely used than are CD34⁺ HPCs.

Taking advantage of established approaches for generating these three types of DCs in vitro with defined cytokines in the absence of FCS, we undertook a detailed characterization of phenotype and function in standard assays of T cell proliferation and CTL development. We further explored the cytokine secretory profile of these DCs with respect to supporting Ag-specific CTL stimulation. Apropos the use of DCs for cross-presentation, we also quantified the phagocytic uptake of dying tumor cells by these three different DC types. More importantly, we compared the capacity of LCs with moDCs to prime Ag-specific T cell responses against cross-presented Ag from dying tumor cells, as an approximation of MHC-restricted Ag presentation relative to phagocytic uptake.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Media, sera/plasma, and noncytokine supplements

Complete RPMI 1640 and complete IMDM were supplemented with 10 mM HEPES, 1% penicillin/streptomycin (Media Laboratory, Memorial Sloan-Kettering Cancer Center (MSKCC), New York, NY), 50 µM 2-ME (Invitrogen Life Technologies, Carlsbad, CA), 1% L-glutamine (Invitrogen Life Technologies), and heat-inactivated, autologous, or single donor human plasma or serum (1, 10, or 20% v/v as specified for a particular experiment). X-VIVO 15 (BioWhittaker, Walkersville, Maryland) was used as manufactured without additives. All media and reagents were endotoxin-free.

Cytokines

Sterile recombinant, endotoxin-, pyrogen-, and mycoplasma-free human cytokines were used to support generation of moDCs, LCs, and DDC-IDCs in vitro, with exact doses specified within those procedures below. These included GM-CSF (sargramostim, leukine; Immunex, Seattle, WA; now Berlex, Wayne/Montville, NJ); FLT-3 ligand (FL), IL-4, TNF-, TGF1, c-kit ligand (KL) or stem cell factor, IL-1-, IL-6 (all from R&D Systems, Minneapolis, MN); PGE₂ (Calbiochem, San Diego, CA; or Pharmacia-Upjohn Pharmaceuticals, now Pfizer, Vienna, Austria); and recombinant human (rhu) CD40L trimer (kind gift of Immunex; no longer available from this source). All cytokines were supplied carrier-free by the manufacturer. All R&D Systems cytokines were reconstituted in PBS with HSA (1% v/v final HSA; from stock 25% HSA, NDC 63546-251-05, pharmaceutical grade, manufactured by Swiss Red Cross, distributed by Alpine Biologics, Orangeburg, NY) in PBS; other cytokines were reconstituted and stored according to manufacturer’s directions.

Cell purification and generation of DCs (Fig. 1)

Healthy donors provided all cells after signing informed consent using institutional review board-reviewed protocols. moDC precursors were tissue culture plastic-adherent (no. 35-3003; Falcon; BD Labware, Franklin Lakes, NJ) PBMCs, obtained by standard centrifugation over Ficoll-Paque PLUS (endotoxin-free, no. 17-1440-03; Amersham Pharmacia Biotech, Uppsala, Sweden) from either whole blood diluted 1/1 with buffered saline or leukocyte concentrates (MSKCC Blood Bank). Bone marrow or G-CSF-elicited peripheral blood stem cell products from healthy donors already undergoing collection for allogeneic transplantation were similarly centrifuged over Ficoll-Paque PLUS to yield mononuclear cells (MNCs). These MNCs then underwent positive immunomagnetic selection according to the manufacturer’s instructions (CD34⁺ isolation kit and LS separation columns; Miltenyi Biotec, Bergisch Gladbach, Germany) for purification of CD34⁺ HPCs to generate LCs and DDC-IDCs.

View larger version (23K): [in this window] [in a new window]   FIGURE 1. moDCs and CD34+ HPC-derived LCs and DDC-IDCs develop from distinct cytokine-supported pathways. The cytokine-supported development of the three myeloid-type DCs under study is summarized here. KL, c-kit ligand or stem cell factor; FL, FLT-3 ligand; GM, GM-CSF; TNF, TNF-; TGF, TGF-.

Purified CD34⁺ HPCs were cultured initially at 2 x 10⁵/3 ml/well (six-well cell culture plates, no. 3516; Costar, Corning Life Sciences, Acton, MA) in X-VIVO 15 for generating LCs or in complete IMDM-20% plasma for generating DDC-IDCs. This percentage of human serum or plasma was selected in pilot experiments because of feasibility and greater CD34⁺ HPC expansion than with 5 or 10% human plasma. Specific cytokine supplements used in common for generating both LCs and DDC-IDCs included GM-CSF (1000 IU/ml), TNF- (5 ng/ml), c-kit ligand (20 ng/ml), and FLT-3 ligand (50 µg/ml). Cultures were replenished with cytokines and media on day 3. On days 5–6, the developing progeny were thoroughly washed and then recultured at 2 x 10⁶/3 ml/well in the same cytokines, but without c-kit ligand and FLT-3 ligand from days 5–6 onward. The DDC-IDC cultures were also transferred to X-VIVO 15 at this time. Thereafter, cytokines and medium were replenished every other day for the duration of the culture.

The following additions were made for the respective generation of LCs vs DDC-IDCs (6, 7, 8, 9, 10, 25). To support LC development, TGF1 (10 ng/ml) was added to the common cytokines throughout the entire culture period (9, 19, 25). For the specific generation of DDC-IDCs, IL-4 (500 IU/ml) was added to suppress macrophage differentiation (25, 26) when the cells were recultured on days 5–6 in X-VIVO 15 with GM-CSF and TNF-, but without c-kit ligand and FLT-3 ligand. The DDC-IDC cultures were never exposed to TGF1.

Maturation and activation of DCs (Fig. 1)

Immature DCs (days 5–6 moDCs; days 11–12 LCs or DDC-IDCs) were used for assessment of dying tumor cell uptake and as flu-infected target cells for CTL assays where indicated. Otherwise, terminal maturation was essential to ensure optimal activation of DC progeny (1, 13, 27). This was accomplished from days 6–8 for moDCs and from days 12 to 14 for LCs and DDC-IDCs by exposure either to CD40L (28, 29) (soluble rhuCD40-L trimer, 0.5 µg/ml; Immunex) or to a combination of inflammatory cytokines (IL-1 (2 ng/ml), IL-6 (1000 IU/ml), TNF- (10 ng/ml)) and PGE₂ (5 mM/ml) (27). Either method resulted in mature, large forward scatter (FSC), CD14^neg, HLA-DR^++/+++, CD83⁺, CD80⁺⁺⁺, CD86⁺⁺ DCs of all three types, which would not revert to less mature forms or alternatively differentiate into adherent macrophages upon removal of cytokines.

Optional DC enrichment

Certain experiments required high purity achieved by one or both of two methods to ensure equivalent numbers of different DC types in functional assays. Developing DC clusters were sedimented by gravity over HSA columns (10), but at the expense of overall cell yield. Cluster formation was most apparent in the LC cultures (10, 16); but to the extent that macrophage differentiation was sufficiently inhibited by IL-4, developing DDC-IDC and moDC clusters were also separable over HSA columns. The bulk cultures or the recultured cluster populations could also be sorted by flow cytometry for large FSC, HLA-DR^bright, CD83⁺ cells, as needed for a particular experiment.

T lymphocytes

T cells were obtained from tissue culture plastic nonadherent PBMCs, then further purified by nonadherence and elution from nylon wool columns (Polysciences, Warrington, PA). Purity was >90–95% based on CD3 expression.

Cryopreservation

DC precursors, DC progeny, and T cells could be cryopreserved at 5–10 x 10⁶ cells/ml in autologous serum (90% final v/v) or HSA (12.5% final v/v in RPMI 1640; Swiss Red Cross, distributed by Alpine Biologics) plus DMSO (10% final v/v; Sigma-Aldrich, St. Louis, MO), by controlled rate freezing at –1 to –3°C/minute (Model 9000; Gordinier Electronic, Roseville, MI). Frozen cells were next transferred to a –70°C freezer overnight and then to liquid N₂ the following day. Cells were thawed once at the time of use by rapid exposure of the vial to 37°C with gentle agitation in a water bath, then immediately transferred into 10–20 ml of medium supplemented with 10% autologous serum and centrifuged. Cells were plated after one additional wash. Recovered cells had 90–95% viability. Nevertheless, we avoided cryopreservation of immature DCs, due to the possibility that thawed viable immature DCs would phagocytose dying immature DCs, present autologous Ags to T cells, and increase the background responses in functional assays.

Cell cycle analysis

Cells were fixed at different time points in their development in 0.4% paraformaldehyde (Sigma-Aldrich) and permeabilized in 50% v/v PBS (MSKCC Media Laboratory) + 50% v/v 0.2% Triton X-100 (Sigma-Aldrich). Cells were stained first by anti-Ki-67 (no. IM 0606; Immunotech, Marseilles, France), and the DNA was subsequently stained with 7-aminoactinomycin D (no. 9400, Sigma-Aldrich). Cells were analyzed by flow cytometry for cell cycle entry into S/G₂/M phase as a measure of proliferation (30).

Phenotypic analyses by flow cytometry

Direct fluorescein (FITC) and PE-conjugated mouse anti-human mAbs included anti-CD3, anti-CD11b, anti-CD14, anti-CD25, anti-CD45RO, anti-CD80, anti-CD86, anti-CD83(-FITC), anti-CD91, anti-HLA-DR (BD Pharmingen, Franklin Lakes, NJ); anti-CD83(-PE) (Immunotech, Beckman Coulter, Marseille, France); and rat anti-CD52-FITC (Serotec, Raleigh, NC). Unconjugated mAbs included anti-e-cadherin (clone HECD-1; R&D Systems); and anti-CD1a/b/c/d (generous gifts of Dr. S. A. Porcelli, Albert Einstein College of Medicine, Bronx, NY). Isotype controls included the appropriate fluorochrome conjugated or unconjugated mouse IgG1 or IgG2a (DAKO, Carpinteria, CA) or rat IgG2b-FITC (Serotec). Unconjugated primary Igs were secondarily stained with biotinylated goat anti-mouse Ig followed by streptavidin-FITC or -PE (BioSource International, Camarillo, CA). Flow cytometry studies used a FACScan (BD Immunocytometry Systems, San Jose, CA), gating for collection and analysis of live events. For analysis of specific epitope expression by DCs, candidate cells were gated for viable, large FSC, HLA-DR bright cells, and 10,000 events were collected.

Phenotypic analyses by confocal microscopy

Cells for confocal microscopy were centrifuged onto glass slides (15,000 cells/slide) at 900 rpm x 5 min (Cytospin 3; Shandon, Pittsburgh, PA), air-dried overnight, placed in slide boxes with desiccant (Drierite, anhydrous calcium sulfate; W. A. Hammond Drierite, Xenia, OH), sealed with Parafilm (American National Can, Neenah, WI) to prevent condensation, and stored at –70°C until use. Just before staining, slides were brought to room temperature while still sealed in the slide box with desiccant. The cytocentrifuged cells were fixed and permeabilized by exposure to cold 90% acetone for 10 min. Slides were air-dried again and then stained at room temperature using unconjugated primary mAbs against CD207 (Langerin; kindly provided by Schering-Plough, Dardilly, France) or CD208 (DC-LAMP; Schering-Plough) for 30 min, followed by an additional 30-min incubation with Texas Red-conjugated goat-anti-mouse Ig (Jackson ImmunoResearch Laboratories, West Grove, PA) without exposure to ambient light. For two-color staining, the cells were then quenched with 10% normal mouse serum in PBS, followed by the addition of anti-HLA-DR FITC (clone L243; BD Pharmingen). Cell staining was evaluated using a confocal laser scanning microscope (Zeiss LSM510; Oberkochen, Germany).

Cytokine assays

Direct hemacytometer counts and flow cytometry quantification of HLA-DR^bright, CD14^neg, CD86⁺, large FSC cells confirmed equivalent immature DC numbers for reculture at 10⁶ DCs/ml in fresh respective media and cytokines (days 5–6 for moDCs and days 11–12 for LCs or DDC-IDCs). Either soluble rhuCD40L trimer or the inflammatory cytokine combination with PGE₂ matured the DCs. Supernatants were collected after 18–20 h of maturation, immediately frozen, and thawed once for assay. The correct maturation phenotype (HLA-DR^bright, CD14^–, CD83⁺, CD86^++/+++) was confirmed at the conclusion of each culture.

IL-12 and IL-15 were measured using commercial ELISA according to the manufacturer’s protocol (IL-12p40, Human IL-12 US Immunoassay kit; BioSource International; IL-12p70, Human IL-12p70 Immunoassay kit, BioSource International; IL-15, Quantikine; R&D Systems, Minneapolis, MN). Measurement of secreted, rather than membrane-bound IL-15 first required 10-fold concentration by ultrafiltration with Vivaspin concentrators (10-kDa exclusion size; Vivascience, Goettingen, Germany). A number of other cytokines were evaluated by a flow cytometry-based assay (no. 551811, Human Inflammation Cytokine Bead Array; BD Biosciences, San Diego, CA).

Mixed leukocyte reactions (MLRs)

DCs were cocultured with 10⁵ purified allogeneic T cells (allo-MLRs) in triplicate round-bottom microwells (Costar 3799; Corning) at variable ratios from 30:1 to 3000:1 (T-DC), in complete RPMI 1640 supplemented with 10% autologous or single donor serum or plasma. DCs were extensively washed to remove cytokines and irradiated 1500 r ¹³⁷Cs before addition to T cells. The resulting proliferation of responder T cells was based on the incorporation of methyl-[³H]thymidine ([³H]TdR, 1 uCi/well; New England Nuclear, Division of PerkinElmer Life Sciences, Boston, MA) during the last 8–12 h of a 4–5 day culture, as measured in a beta scintillation counter (Betaplate, Wallac, Division of PerkinElmer Life Sciences, Wellesley, MA).

Induction and measurement of influenza virus-specific CTL

Mature DCs from HLA-A*0201 donors were washed and resuspended in cytokine and serum-free RPMI 1640, then either loaded directly with influenza matrix peptide (fluMP) or infected with live influenza virus. For peptide pulsing, mature DCs were separately cultured with 10 µM fluMP (HLA-A2.1-restricted 9-mer fluMP, GILGFVFTL; ResGen, Division of Invitrogen, Carlsbad, CA) for 1 h at room temperature. For direct infection, mature DCs were exposed to influenza virus strain PR/8/34 (SPAFAS, Preston, CT) at a dose of 1000 hemagglutinin U/ml per 5–10 x 10⁶ DCs/ml serum-free RPMI 1640 for 60 min at 37°C (31). DCs were washed three times after either peptide pulsing or infection.

fIuMP-pulsed or influenza-infected mature DCs of each type were cocultured with 10⁵ purified autologous T cells in triplicate round-bottom microwells (Costar no. 3799 96-well plates; Corning) at variable ratios from 30:1 to 1000:1 (T:DC) in complete RPMI 1640–10% autologous or single donor serum or plasma. After 6–7 days, T cells were assayed for the amount of cytolytic activity generated per primary culture by direct addition of ⁵¹Cr-labeled (75 uCi of Na⁵¹CrO₄/2 x 10⁶ targets at 37°C for 1 h followed by four washes; New England Nuclear) target cells, which were influenza-virus infected or control uninfected, immature autologous moDCs. Some experiments also used T2 cells or A*0201-transduced K562 cells (a kind gift of T. Wolfel, University of Mainz, Mainz, Germany) as targets after peptide pulsing, although the T2 cells tended to release variably higher background ⁵¹Cr.

After a 4–6 h incubation with the radiolabeled targets, 30 µl of supernatant were removed from each well and transferred in the same triplicates to a LumaPlateTM-96 (Yttrium Silicate Scintillator-coated white microplate; Packard Bioscience, Groningen, The Netherlands), which was dried for 12 h at room temperature, sealed (TopSealTM-A; Packard Bioscience), and counted in a gamma counter (TopCount NXTTM, Microplate Scintillation and Luminescence Counter; Packard Bioscience). Cytolytic activity was calculated using triplicate means to obtain the quotient of the experimental less the spontaneous ⁵¹Cr release, divided by the total less the spontaneous ⁵¹Cr release, multiplied by 100 to give a percentage. Adjusted specific lysis was determined by subtracting the percent specific lysis of control uninfected targets (autologous immature moDCs, irrelevant peptide-pulsed (melanoma gp100) T2 or A*0201 transfected K562 cells) from that of infected targets.

Uptake of dying cells by DCs

A melanoma cell line (SKMel29; MSKCC) that expressed HLA-A*0201 was proven mycoplasma-free by PCR (mAb Core Facility, MSKCC). Viable, growing tumor cells were adherent to tissue culture plastic but were released by gentle rinsing after a 3–5 min incubation at 37°C in prewarmed trypsin (trypsin 0.25%; ICN Biomedical, Aurora, OH) diluted 1/1 with prewarmed RPMI 1640. After two washes, the cells were recultured in RPMI 1640–10% FCS (FCS; Gemini BioProducts, Woodland, CA) at 3 x 10⁶ cells/3 ml/35-mm tissue culture well of a six-well plate and induced to undergo apoptosis by 20-h exposure to mitomycin C 10 µg/ml final (Bristol Laboratories, Bristol-Myers Squibb, Princeton, NJ). This method consistently yielded 80% Annexin V⁺, propidium iodide^neg, apoptotic cells (early apoptosis detection kit; Kamiya Biomedical, Seattle, WA).

To assess actual uptake, tumor cells were first stained with PKH67 green fluorescent vital membrane dye (Green Fluorescent Cell Linker kit; Sigma-Aldrich), rendered apoptotic by mitomycin exposure, washed, then recultured in complete RPMI 1640–10% FCS with immature DCs at a 1:1 ratio. After 24 h, the cell mixture was collected and stained with anti-HLA-DR-PE. Flow cytometric analysis showed three populations. Single HLA-DR⁺ cells were DCs, single PKH67⁺ cells were tumor cells, and double-positive cells were DCs that had taken up apoptotic tumor cells. Intracellular uptake rather than adherence was confirmed by fluorescent microscopy. By gating on the double-positive cells, the amount of phagocytosed tumor cells could be quantified based on the mean fluorescent intensity (MFI) for PKH67 in the FITC channel.

Induction and measurement of tumor-specific CTL by ELISPOT after cross-priming

moDCs, which are more actively phagocytic than either LCs or DDC-IDCs, were compared with LCs as the more immunogenic of the two DC types generated from CD34⁺ HPCs. These DCs were matured by the inflammatory cytokine mixture with PGE₂ for 48 h, beginning 24 h after the start of unlabeled apoptotic tumor cell uptake as above. The cell suspension was then collected and replated with autologous T cells at a DC:T ratio of 1:30 in RPMI 1640–10% autologous or single donor serum or plasma. After 7 days, the T cells were harvested, washed, counted, and restimulated for 5 days with fresh autologous, mature DCs that had also phagocytosed SKMel29 cells. No exogenous cytokines, e.g., IL-2, were ever added to these cultures. Control wells for the afferent stimulation included autologous T cells stimulated by mock-loaded DCs handled exactly as above.

After these two rounds of 7 and 5 days of stimulation, CD8⁺ T cells were positively selected using anti-CD8 microbeads (Miltenyi Biotec) and analyzed for IFN- production by an enzyme-linked immunosorbent spot assay (ELISPOT for human IFN-; Mabtech, DiaPharma Group, West Chester, OH). A total of 10⁵ purified CD8⁺ T cells followed by 10⁴ target cells were added to each well in triplicate. Target cells comprised either HLA-A*0201 SKMel29 tumor cells if the donor also expressed HLA-A*0201, or immature autologous DCs loaded with dying SKMel29 tumor cells if the donor did not express HLA-A*0201. Nonmelanoma tumor cells (prostate cancer cell line, HLA-A*0201-positive LnCAP) or immature DCs that had not taken up dying tumor cells were added as control targets to separate wells. Wells were read by an automated ELISPOT reader (Automated ELISPOT Reader System, Carl Zeiss Vision) to give the number of spot-forming cells (SFC) per 10⁵ input cells. Mean values were calculated from triplicate wells. Adjusted numbers of SFCs were calculated by subtracting the IFN- release from both of the two types of control wells (T cells stimulated by mock-loaded DCs and tested against tumor Ag-expressing targets; T cells stimulated by loaded DCs and tested against empty DC or LnCAP targets), from that of the experimental wells.

Statistics

A permutation test was used to compute differences in DC stimulatory capacity for allogeneic T cell proliferation or influenza-specific, autologous CTL generation. The null hypothesis for this test was that any two types of DCs being compared exerted no differential stimulation of responder T cells. The statistic used to test this hypothesis was the sum of the squared differences between mean cpm (T cell proliferation in the allo-MLRs) or mean percent-specific lysis values (CTL generation in autologous, flu-specific MLRs), summed over all dose levels. This generated a p value corresponding to the proportion of all test statistics from the permutation distribution that were equal to or more extreme than the test statistic observed. Pairwise comparisons of multiple single readouts from independent experiments used either the Student t test or the nonparametric Wilcoxon rank sum test as indicated.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
moDCs and CD34⁺ HPC-derived LCs and DDC-IDCs are phenotypically distinct

All three DCs achieved comparable expression of HLA-DR, CD86, CD45RO (not shown), CD80 (not shown), and CD83 after maturation (Fig. 2A). The combination of IL-1-, IL-6, TNF-, and PGE₂ (27) achieved phenotypic maturation of all DC types comparable to that effected by rhuCD40L-trimer (Fig. 2B). Maturation of both DDC-IDCs and LCs increased DC-LAMP/CD208 (32), although some was already detectable at day 12 because of physical manipulation and activation of the cells (Fig. 2C). The distribution of CD208 merges with that of HLA-DR in mature DDC-IDCs but not LCs, the basis and implications of which remain unknown.

View larger version (53K): [in this window] [in a new window]   FIGURE 2. Distinct epitope expression differentiates moDCs and CD34+ HPC-derived LCs and DDC-IDCs. A, Blood monocytes and CD34+ HPCs were isolated and cultured with cytokines as described. Terminal maturation and activation was accomplished by exposure to inflammatory cytokines, and the resulting progeny were stained with mAbs for the indicated epitopes on days 7–8 for moDCs and days 13–14 for LCs and DDC-IDCs. The exception was e-cadherin (see *e-cad), which is lost with maturation, and was therefore stained on days 11–12 before final exposure to the inflammatory cytokine mixture. Fluorescent intensity of single epitopes (thick line histograms) compared with an isotype control (thin line histograms) is indicated along the x-axes (log10), and cell frequency is depicted against the y-axes. B, Maturation by either rhuCD40L or an inflammatory cytokine mixture including IL-1-, TGF-, IL-6, and PGE2, achieved comparable phenotypic up-regulation of maturation/activation markers CD83 and class II MHC. Furthermore, the expression of CD83 and HLA-DR (as well as CD86 and CD40, not shown) was similar between LCs, DDC-IDCs, and moDCs. Correlation coefficients were >0.9 for each single type of DC (LCs, DDC-IDCs, moDCs), when comparing that single DC type matured with either rhuCD40L or the inflammatory cytokines in the same assay of immunostimulatory function (allo-MLR or induction of fluMP-specific CTL; not shown). C, Cytocentrifuged LCs and DDC-IDCs were stained for CD208 (DC-LAMP; Texas Red) and HLA-DR (FITC) on days 12 and 14 of culture. DC-LAMP expression increased with maturation of both populations and merged with the HLA-DR in the DDC-IDCs. D, Cytocentrifuged specimens were stained for CD207 (Langerin; Texas Red) and HLA-DR (FITC) at serial time points during culture and assessed by confocal microscopy. CD207 expression decreased with maturation of LCs. DDC-IDCs (see inset on day 12) never expressed this LC molecule, which supports the formation of Birbeck granules. The overlapping and distinct epitopes that aid phenotypic distinction of these three DC types are summarized in Table I.

Finally, we examined expression of the CD1 isoforms (Fig. 2A). CD1a was originally considered an LC-specific epitope, but it has long been known that GM-CSF exposure will induce CD1a/b/c on monocytes (37). All three DC types actually expressed CD1a, and LCs down-regulated this epitope with maturation more than did the other two. CD34⁺ HPC-derived LCs expressed little to none of the other three isoforms, whereas CD1b, CD1c, and notably CD1d expression by moDCs paralleled that of DDC-IDCs (38).

moDCs, having arisen from a more uniform starting population of committed precursors, generally achieved a final purity of 80–90%, the principal contaminants being B cells from the initial plastic adherence of PBMCs. Without intermediate enrichment steps, LCs and DDC-IDCs derived from pluripotent CD34⁺ HPCs achieved a final purity of 65–70%. The remaining cells comprised immature granulocytic cells, especially immature eosinophils due to the high concentrations of GM-CSF.

In summary (Table I), we find that among the large FSC, HLA-DR^bright, CD14^neg, CD83⁺ population, the presence of both CD11b and CD1d distinguishes CD34⁺ HPC-derived DDC-IDCs from LCs. The further presence of both CD52 and CD91 distinguishes moDCs from the CD34⁺ HPC-derived DDC-IDCs. Furthermore, either CD40L or the combination of inflammatory cytokines were equally effective in inducing the maturation or activation epitopes shared by the three DC types under study, e.g., class II MHC^bright, CD83⁺, CD86⁺⁺, CD80^++/+++, CD45RO⁺, CD208^++/+++. The single exception was that mature LCs expressed CD25 only after exposure to the inflammatory cytokines but not after exposure to CD40L, whereas mature moDCs and DDC-IDCs expressed this epitope after either maturation stimulus (data not shown). The function of the CD25 epitope on mature, activated DCs remains unknown.

moDC precursors, like all postmyelocyte cells, no longer divide or proliferate, but only differentiate. In contrast, serial cell cycle analyses documented similar entry into S/G₂/M phase for CD34⁺ HPCs developing into either LCs or DDC-IDCs (Fig. 3) in the absence of FCS (30). The first 5–6 days supported the greatest expansion of both populations, as previously reported for FCS-supplemented cultures (6). The absence of FCS compromised overall cell yields by 5-fold, but this was offset by increased purity of the DC progeny and reduced endogenous cytokine exposure in the presence of serum-containing media, especially those with FCS. LCs also had a distinct requirement for provision of exogenous TGF1 in the absence of FCS (9, 19, 25, 39, 40), which further compromised final LC yields by about half when compared with DDC-IDCs.

View larger version (27K): [in this window] [in a new window]   FIGURE 3. CD34+ HPC-derived LCs and DDC-IDCs achieve comparable cell cycle entry for expansion in the absence of FCS-supplemented media. CD34+ cells were cultured as described under Materials and Methods using conditions that supported the distinct generation of LCs or DDC-IDCs, but notably in the absence of FCS. Cells were stained for Ki-67 and 7-aminoactinomycin D at the indicated time points along the x-axis. The double-positive cells by cytofluorography represented the percent of the total population in S/G2/M phase, plotted along the y-axis. Cell cycle entry was similar for the two types of DCs, even though overall expansion was compromised compared with FCS-supplemented media and similar cytokines (3- to 10-fold vs 40- to 100-fold). Representative of three experiments assessed in this manner.

We compared moDCs with the CD34⁺ HPC-derived LCs and DDC-IDCs for their ability to stimulate resting allogeneic T cells in the MLR (allo-MLR), as this is a standard assay for DC function. DCs were added in serial 3-fold dilutions to a constant number of 10⁵ purified T cells (Fig. 4, x-axis). The proliferative response of the allogeneic T cells indicated that all three DC types were potent stimulators based on their mature, activated phenotypes (Fig. 4, note y-axis log₂ scale depicting T cell division and proliferation).

View larger version (25K): [in this window] [in a new window]   FIGURE 4. LCs, DDC-IDCs, and moDCs are all potent stimulators of resting allogeneic T cells, with LCs exerting somewhat greater activity despite comparable phenotypic maturation and activation among all three DC types. Mature moDCs, LCs, and DDC-IDCs were added in equivalent graded doses, based on mature phenotype (large FSC, HLA-DRbright, CD14neg, CD83+) and typical morphology (FACS corroborated by typical DC morphology on direct hemacytometer counts), to separate triplicate cultures of 105 allogeneic T cells per round-bottom microwell in a 96-well plate. DC doses ranged from 3000 to 100 cells per well, yielding starting DC-T ratios from 1:30 to 1:1000. [3H]TdR uptake by proliferating allogeneic T cells over the last 12 h of a 4–5 day culture was measured as an index of DC immunogenicity. The averaged triplicate means ± SEM for [3H]TdR incorporation by T cells stimulated in independent allo-MLRs by 16 pairs of LCs and DDC-IDCs from the same donor, 10 of which also included moDCs from the same donor, are depicted logarithmically (log2) against the y-axis. By the permutation test, p = 0.23 (NS) for LCs vs DDC-IDCs, but p < 0.01 for moDCs vs either LCs or DDC-IDCs.

LCs are more potent than moDCs in stimulating CTL against a viral recall Ag after presentation of passively loaded peptide, but similar after direct influenza infection

We explored the capacity of moDCs, LCs, and DDC-IDCs to stimulate flu-specific CTLs after a single round of stimulation without addition of exogenous cytokines. fluMP-pulsed LCs were more potent on a cell-for-cell basis than either fluMP-pulsed DDC-IDCs (Fig. 5, top, p < 0.01, n = 16 independent experiments) or moDCs (Fig. 5, top panel, p = 0.05, n = 4 of the 16 independent experiments). This distinction was abrogated after direct influenza infection of LCs, DDC-IDCs, and moDCs, with no significant differences between CTL stimulatory capacity (Fig. 5, bottom panel, p = NS, n = 10 independent experiments, 5 of which also included moDCs). Pilot experiments had revealed similar kinetics of CTL generation between the different DC types, so that all CTL assays were conducted after 6–7 days of primary stimulation by DCs. By the direct addition of targets to the primary cultures, we could directly assess the immunogenicity of the respective DC types. This would not have been possible had we harvested responder lymphocytes and compared equal E:T ratios between the different DC-stimulated cultures.

View larger version (20K): [in this window] [in a new window]   FIGURE 5. LCs are more potent on a cell-for-cell basis than DDC-IDCs or moDCs in stimulating CTL against an HLA-A*0201-restricted viral peptide (fluMP), a disparity that is abrogated by direct flu infection. LCs, DDC-IDCs, and moDCs were separately generated from the same HLA-A*0201 donor. CD34+ HPCs served as the starting population for LCs and DDC-IDCs, and adherent blood monocytes were used to generate moDCs, all with defined cytokines and media as described in Materials and Methods. The resulting HLA-A*0201 DCs were sorted cytofluorographically to select equivalent numbers of mature CD83+MHC IIbright progeny of each DC type. Separate cytofluorographic analyses confirmed the distinct phenotype of each DC type as shown in Fig. 2. Each population was pulsed separately with fluMP (top panel) or separately infected with influenza virus (bottom panel), then added in serial doses to triplicate microwells containing 1 x 105 autologous T cells. On days 6–7, the amount of influenza-specific cytolytic activity per primary culture was determined by adding 5 x 103 fluMP-pulsed, 51Cr-labeled target cells, and measuring 51Cr release 4–6 h later. The y-axes depict the adjusted mean percent-specific 51Cr release (after subtracting the lysis of control targets) ± SEM from 16 independent experiments using fluMP-pulsed LCs vs DDC-IDCs (p < 0.01), 4 of which also included moDCs (LCs vs moDCs, p = 0.05; DDC-IDCs vs moDCs, p = 0.40, NS). Ten independent experiments also compared flu-infected LCs vs DDC-IDCs (p = 0.54, NS), 5 of which also included moDCs (LCs vs moDCs, p = 0.49, NS; DDC-IDCs vs moDCs, p = 0.52, NS). The primary autologous stimulatory T-DC ratios (not E:T ratios) are depicted along the x-axes. 51Cr-labeled target cells were flu-infected vs uninfected autologous immature moDCs, or flu-MP vs gp100-pulsed K562-HLA*A201 transfectants or T2 cells.

Immature moDCs, LCs, and DDC-IDCs were densely replated at 10⁶ cells/ml for maturation using soluble rhuCD40L trimer. Supernatants were collected at serial time points during the 2 days of maturation, with peak cytokine detection at 18–20 h in pilot assays. We found a wide disparity between the secretion of IL-12p40 and the bioactive heterodimer IL-12p70 (Fig. 6). Although all three DCssecreted IL-12p40 (p = NS), we detected the bioactive IL-12p70 only in the supernatants of moDCs. Neither CD34⁺ HPC-derived LCs nor DDC-IDCs ever secreted measurable IL-12p70. We have measured almost identical patterns of IL-12p40 and p70 after maturation with inflammatory cytokines and PGE₂ (41), although these amounts are at least a full log lower than those achieved by maturation with CD40L in the absence of FCS (42).

View larger version (12K): [in this window] [in a new window]   FIGURE 6. LCs, DDC-IDCs, and moDCs exhibit distinct cytokine secretion profiles during maturation, especially with respect to IL-12p70 and IL-15. Immature LCs, DDC-IDCs, and moDCs were separately generated from their respective precursors using defined cytokines and media as described in Materials and Methods, then densely recultured for maturation by either soluble rhuCD40L trimer (IL-12p40, n = 6 independent experiments, 4 of which also included moDCs; and IL-12p70, n = 3 independent experiments) or the inflammatory cytokine mixture (IL-15, n = 3 independent experiments). Supernatants were collected after 18–20 h of maturation for cytokine measurement by ELISAs. Supernatants for IL-15 were concentrated 10-fold before assaying. The mature, activated phenotype was confirmed for each DC type as in Fig. 2 at the end of a 48-h maturation. Cytokine levels (picograms per milliliter) are plotted against the y-axis. The Student t test was used to compare differences between the groups, and the p values are shown for each paired combination.

A number of other cytokines, e.g., TNF-, IL-1-, IL-6, IL-10, and IL-8, were evaluated by a cytokine bead array assay. Although IL-1- secreted by LCs achieved statistical significance over other DCs, this was largely because of low interassay variability as the absolute levels were similar (not shown). Flu infection did not exert a clear pattern of effect on the detectable differences between these DC-secreted cytokines, or IL-12 or IL-15, and is therefore not shown.

moDCs are more actively phagocytic of dying tumor cells than are either LCs or DDC-IDCs

Cross-presentation is an important mechanism by which DCs can obtain Ag from a third party cell and present that Ag in a self MHC-restricted fashion to autologous Ag-specific T cells (45, 46, 47, 48, 49, 50, 51, 52, 53, 54). An important first step is uptake of an Ag source, and one means is by phagocytosis.

We used a flow cytometry based method to compare the phagocytic capacity of immature day 5–6 moDCs, compared with immature day 11–12 LCs and DDC-IDCs, to take up apoptotic tumor cells. Each immature DC type was confirmed to lack the maturation phenotype, including absence of surface CD83. Tumor cells from the SKMel29 cell line labeled with a green fluorescent vital dye (PKH67) were rendered apoptotic by mitomycin C, and cocultured for 24 h at 37°C with each immature DC type by direct addition to the DC-containing wells. It was important not to physically manipulate the DCs excessively, especially the LCs, as this induced some maturation. After harvesting the cocultured cells and staining with anti-HLA-DR-PE, we gated the large FSC, HLA-DR^++/+++ cells and determined the fluorescence in the FITC channel to quantify the amount of PKH67-labeled, apoptotic tumor cells that each DC type had phagocytosed (Fig. 7A). Phagocytosis was also confirmed by direct inspection of the cells under a fluorescent microscope to rule out simple adhesion. Uptake also did not occur at 4°C as confirmed by flow cytometry (not shown).

View larger version (19K): [in this window] [in a new window]   FIGURE 7. moDCs are the more actively phagocytic by at least one log than are either LCs or DDC-IDCs. Immature moDCs, LCs, or DDC-IDCs, were cocultured 1:1 with SKMel29 tumor cells that had been membrane-labeled with PKH67 and rendered apoptotic (80% at onset of coculture) by mitomycin C exposure. At the end of a 24-h incubation at 37°C, cells were harvested, counterstained with anti-HLA-DR-PE and then gated for large FSC, HLA-DR-positive cells on a flow cytometer. The uptake by DCs in this gate of PKH67+ tumor cells, detected in the FITC channel, is shown along the x-axis (log10 fluorescence) for one representative experiment in A. Replicate MFIs for tumor cell uptake ± SD are shown in B (n = 10 experiments for moDCs, and n = 9 experiments each for LCs and DDC-IDCs), with p values for pairwise comparisons by the Wilcoxon rank sum statistic indicated on the figure.

Phagocytic uptake of Ag does not translate directly into Ag presentation and stimulation of T cells

Although Ag uptake is a critical first step to Ag processing and presentation, especially for cross-priming or cross-presentation on class I MHC molecules, the alternative fate of phagocytosed Ag is sequestration and degradation. Hence, the two processes are not necessarily directly correlated. We therefore compared the highly phagocytic moDCs with the more potent LCs of the two DC types generated from CD34⁺ HPCs, with respect to induction of CTLs after cross-presentation of tumor and tumor-associated Ags. Immature moDCs (days 5–6) and LCs (days 11–12) were cocultured with apoptotic tumor cells as in Fig. 7, but without PKH67 labeling, for 24 h. Thereafter, the DCs were exposed to the inflammatory cytokine mixture for terminal maturation over the next 48 h, then washed extensively before addition to autologous T cells. T cells underwent two rounds of stimulation, first over 7 days,followed by another 5 days with freshly produced, tumor-loaded, mature DCs at a DC-T ratio of 1:30. Exogenous cytokines were never added.

At the completion of these two stimulations, CD8⁺ T cells underwent positive selection by immunomagnetic beads and were then evaluated for IFN- secretion in ELISPOT assays after rechallenge with either the original SKMel29 cell line or immature moDCs loaded with apoptotic SKMel29. As negative controls, we included CD8⁺ T cells from cultures that had undergone two rounds of stimulation by the respective DC type, but which had never been loaded with apoptotic tumor cells. Target controls for tumor-loaded DCs were mock-loaded DCs, and controls for tumor cell targets were an HLA-A*0201-positive cell line bearing irrelevant tumor Ag (prostate cancer cell line, LnCAP). The SFCs induced by each of these control cultures were both subtracted from the SFCs generated by tumor-loaded DCs for plotting final results (Fig. 8).

View larger version (16K): [in this window] [in a new window]   FIGURE 8. LCs more effectively cross-prime autologous T cells to secrete IFN- in response to melanoma Ags than do moDCs. Immature moDC and LCs generated from healthy donors were cocultured as above with dying SKMel29 tumor cells for 24 h, followed by maturation with the inflammatory cytokine mixture. These mature, tumor-loaded DCs were then cocultured with autologous T cells at a DC-T ratio of 1:100 for 7 days. Thereafter, the cells were harvested, washed, and restimulated for 5 additional days with fresh tumor-loaded DCs of the same type as used for the first stimulation at a DC-T ratio of 1:30. Exogenous cytokines were never added to these cultures. At the end of 12 days total, CD8+ T cells underwent positive immunomagnetic selection and were tested for IFN- secretion by rechallenge in ELISPOT assays with SKMel29 cells (n = 1 for the single donor who was HLA-A*0201-positive, and n = 2 for DCs of the same type used for the two rounds of stimulation and loaded with apoptotic SKMel29). The number of SFCs per 105 input CD8+ T cells was determined in triplicate and averaged for each experiment by an automated, standardized ELISPOT plate reader. Nonspecific IFN- release from controls for both the afferent stimulation and efferent target recognition (see text) were subtracted from the experimental values to give an adjusted number of SFCs/105 input CD8+ T cells, and these values are plotted as averaged triplicate means ± SEM from three independent experiments (p = 0.01 by Wilcoxon rank sum statistic). Although the LCs exhibited a greater variability than the moDCs, the middle value was from rechallenge of the HLA-A*0201-positive donor by the HLA-A*0201-positive SKMel29 cells, whereas the flanking values were from rechallenge by autologous LCs that had taken up dying SKMel29 cells.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
We have generated phenotypically and functionally distinct populations of LCs, DDC-IDCs, and moDCs in vitro using defined precursors and cytokines in the absence of FCS. LCs stimulate the greatest CTL activity, even without the bioactive p70 heterodimeric form of IL-12, whether passively presenting a viral recall antigenic peptide or cross-presenting tumor and tumor-associated Ag to autologous CD8⁺ T cells. Among a variety of parameters investigated to account for the superior activity of mature LCs in vitro, secreted IL-15 protein is the greatest and most significantly detected difference between LCs and the other two DC types under study. Our phenotypic and functional data also demonstrate that DDC-IDCs and moDCs can no longer be considered interchangeable, even though both develop from a CD14⁺ intermediate or precursor.

All three DC types express comparable maturation phenotypes, while other epitopes distinguish one from the other (summarized in Table I). All three types of DCs are also potent stimulators in allogeneic MLRs, even though moDCs require half log higher numbers than either LCs or DDC-IDCs to stimulate the same degree of T cell proliferation. This is despite somewhat greater purity of moDCs than LCs or DDC-IDCs, which should have favored the activity of moDCs in case of any counting errors when the same numbers of CD83⁺, HLA-DR^bright DCs were added to allogeneic T cells.

The superior activity of LCs in stimulating autologous CD8⁺ CTL against a recall viral Ag was not the anticipated result. Direct influenza infection of the mature DCs, which is not as cytopathic as it is for immature DCs, enables all three to generate robust and equivalent CTL. This mitigates the differences observed when presenting a class I MHC-restricted peptide alone to a responder population containing both CD8⁺ and CD4⁺ T cells, perhaps by providing helper epitopes or by releasing type I IFNs in a paracrine fashion that could enhance activation of other DCs and equalize their activities (55). Although the starting population of T cells was >90–95% CD3⁺, NK cells were not specifically depleted at the outset. Given the emerging role of DCs in stimulating NK cells (41, 56), perhaps virally infected DCs could stimulate NK cells to secrete additional cytokines that would enhance their own immunogenicity and help polarize toward type I immune responses (57, 58).

As noted above, LCs secrete significantly more IL-15 than either DDC-IDCs or moDCs, as well as somewhat more IL-1-, both of which are pertinent to the initiation of primary immune responses and the generation of primary and memory CD8⁺ CTL (59, 60, 61, 62). Prior studies have shown that both LCs and DDC-IDCs express IL-15 mRNA (22), however, and a blocking mAb against IL-15 can inhibit some functions of moDCs (63). We therefore consider soluble IL-15 to be a surrogate measurement of bioactive IL-15, which functions at the membrane in picomolar concentrations in both cis and trans toward responder T cells (44), as a 10-fold concentration of the supernatants was required to detect IL-15 in our experiments (64). Our efforts using blocking mAbs to determine more precisely the contribution or hierarchy of individual cytokines did not yield clear results, perhaps reflecting the redundancy often found for factors supporting the immunogenicity of DCs. We also did not have convincing positive controls to confirm inhibition by the anti-cytokine mAbs. Similarly, attempts to ascribe normalization of virally infected LC and DDC-IDC activity to normalization of cytokine profiles did not yield a clear pattern, suggesting that other factors are operative.

Our data indicate that there is no absolute requirement for IL-12p70 in the generation of class I MHC-restricted CTL activity or IFN- secretion by CD8⁺ T cells. Only moDCs secrete measurable IL-12p70, while LCs and DDC-IDCs secrete little to none at any time point during maturation, even after an optimal stimulus delivered by rhuCD40L trimer (29, 42). We have corroborated similar patterns of IL-12p70 secretion by moDCs, but not LCs or DDC-IDCs, after maturation by inflammatory cytokines with PGE₂ in the absence of FCS, although the amounts are about a log lower than after maturation by rhuCD40L (41). Mature LC and DDC émigrés from human skin also do not secrete IL-12p70 (42). LCs are nevertheless more potent in eliciting CTL after presentation of a passively loaded peptide or after cross-presentation of Ag(s) taken up from dying tumor cells. These data lend further credence to recent commentary that the requirement for IL-12, while optimal for Th1 and CTL responses, may have been overemphasized by inadvertent detection of IL-12p40 rather than p70 (65).

Given the emerging role of DCs in linking innate and adaptive immunity, however, the rapid response of moDC precursors to inflammatory signals, including viral infection, can lead to maturation and early IL-12p70-mediated activation of NK cells (41, 56), which in turn secrete IFN- and TNF-. This could enhance the immunogenicity of moDCs as well as that of other DC types in mixed cell populations as would exist in vivo and further polarize toward type I adaptive immune responses (57, 66). Theoretically, this could also apply to early activation of NKT cells that would produce IFN- in response to glycolipid Ags presented by CD1d, which was solely expressed by moDCs and DDC-IDCs, but never by LCs (38).

Investigators are increasingly interested in cross-priming and cross-presentation by DCs for both immunity and tolerance (48, 49, 50, 51, 52, 67, 68, 69). This provides an important means for DCs to process and present exogenous Ags from the microenvironment, in addition to classic processing and presentation of endogenous Ags on class I MHC. Cross-priming and -presentation of Ag from a third party cell can also expand potential Ag sources, obviate the need to use defined peptides with known HLA-restrictions, and simultaneously provide epitopes for class I and II MHC presentation.

By quantifying the uptake of apoptotic tumor cells, we have shown that moDCs are efficiently phagocytic as expected, while LCs and DDC-IDCs are more indolent in this respect. LCs are on average 4-fold more potent than moDCs, however, in cross-priming autologous T cells to melanoma and associated Ags derived from apoptotic tumor cells without requiring exogenous cytokines. The mechanics of these types of experiments unfortunately cannot exclude the introduction of additional (e.g., allogeneic or even xenogeneic (cell lines cultured in the presence of FCS)) Ags from tumor cell lines. We assume these provided additional immunogenic Ags restricted to class I MHC, as well as helper epitopes presented on class II MHC to CD4⁺ T cells in the responder population. These experimental issues applied evenly to both moDCs and LCs, yet the LCs remained superior even after subtraction of controls for both the afferent stimulation and efferent target recognition.

Several points are therefore worth noting. Obviously, the balance between processing and presentation vs sequestration and degradation of endocytosed Ag is not entirely concordant, at least for moDCs (54). This may be one rationale underlying the benefit of targeting Ags to specific receptors or intracellular organelles for enhanced cross-presentation by DCs, especially human moDCs (51, 52, 53, 54, 67, 68). LCs, which are less phagocytic than moDCs, may nevertheless acquire sufficient Ag for cross-presentation by this means, or by another process altogether that LCs may or may not share with other conventional DCs. The outcome of phagocytosis for cross-presentation may also differ from that of pinocytosis of soluble, fluid phase Ags (54).

Although our finding that LCs (and DDC-IDCs) do not express CD91 is far from a comprehensive assessment of heat shock protein receptors, Ag uptake by LCs at least does not occur by acquisition of gp96 (33) expressed by dying cells. LCs, even around day 11 before maturation, also express very little if any CD36 (not shown), which is the thrombospondin receptor expressed by immature moDCs that is involved in their uptake of dying cells (49). Hence, LCs deserve specific further study of the cell biology of Ag processing that occurs between Ag uptake and presentation, as has been focused on mouse DCs and human moDCs (67, 70, 71, 72).

Our findings that LCs have a superior capacity to stimulate CTL against a single peptide or cross-presented tumor and tumor-associated Ags in vitro help explain a published report that CD34⁺-derived DCs are superior to moDCs in stimulating an Ag-specific CD8⁺ T cell line (73). Such CD34⁺ HPC-derived DCs would have included LCs (6, 7, 8), and our results indicate that the stimulatory properties specifically exerted by LCs are equally true for circulating T lymphocyte responders as for T cell lines. Lastly, our results also support the findings of some investigators (24) asserting advantages to the administration of DC vaccines that include LCs, as opposed to those that use only moDCs.

Published and ongoing clinical trials using peptide-pulsed moDCs, which are more readily obtained than CD34⁺ HPC-derived DC populations have proven the efficacy of moDCs in stimulating Ag-specific T cell responses in vivo (74, 75). The results reported here neither negate nor discount those important findings. Our results do, however, beg the question as to whether the differences observed in vitro between mature LCs and either DDC-IDCs or moDCs remain valid in vivo. Functional differences identified in this study in vitro could simply reflect a requirement for numerically more DDC-IDCs or moDCs than LCs to exert similar CTL stimulatory activity. Our data also suggest that methods of Ag loading, including those that provide helper epitopes and/or simultaneously stimulate NK cells, merit further attention. On balance, conventional, myeloid-type DCs have not only shared but also distinct phenotypes and functions, which may prove physiologically important to a multipronged activation of both innate and adaptive cellular immunity.

	Acknowledgments

 
We gratefully acknowledge the various contributions of members of the Young Laboratory to the development of this work. We appreciate critical comments on the manuscript from Dr. Nikolaus Romani, University of Innsbruck (Innsbruck, Austria). We thank Dr. Steven A. Porcelli, Albert Einstein College of Medicine (Bronx, NY), for the generous gifts of mAbs to CD1 isoforms. We also thank the nurses and physicians of the Allogeneic Bone Marrow Transplantation Service, as well as the Allogeneic Transplant and Cytotherapy Laboratory staffs at Memorial Sloan-Kettering Cancer Center (New York, NY), for assistance with sample procurement and processing.

	Footnotes

 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ This work was supported by Grants R01 CA-83070 (to J.W.Y.), P01 CA-23766 (to J.W.Y. and G.H.), P01 CA-59350 (to J.W.Y. and G.H.), and R21 CA-97714 (to J.W.Y.) from the National Cancer Institute, National Institutes of Health; by Grant LLS 6124-99 (to J.W.Y.) from the Leukemia and Lymphoma Society; and by the Spanish Ministry of Education and Science (MAdeC). Equipment support was provided by William H. Goodwin and Alice Goodwin and the Commonwealth Cancer Foundation for Research and the Experimental Therapeutics Center of Memorial Sloan-Kettering Cancer Center.

² G.R. and J.B. contributed equally to this work.

³ Current address: Universitaetsklinik fuer Dermatologie und Venerologie Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria.

⁴ Current address: Servicio de Farmacologia Clinica, Hospital Universitario "Marques de Valdecilla", E-39008 Santander, Spain.

⁵ Address correspondence and reprint requests to Dr. James W. Young, Box 395, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021-6094. E-mail address: youngjw{at}mskcc.org

⁶ Abbreviations used in this paper: DC, dendritic cell; moDC, monocyte-derived DC; DDC-IDC, dermal-interstitial DC; HPC, hematopoietic progenitor cell; LC, Langerhans cell; rhu, recombinant human; FSC, forward scatter; fluMP, influenza matrix peptide; MFI, mean fluorescent intensity; SFC, spot-forming cell; MLR, mixed leukocyte reaction; MNC, mononuclear cell.

Received for publication February 25, 2004. Accepted for publication May 25, 2004.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y.-J. Liu, B. Pulendran, K. Palucka. 2000. Immunobiology of dendritic cells. Annu. Rev. Immunol. 18:767.[Medline]
2. Mellman, I., R. M. Steinman. 2001. Dendritic cells: specialized and regulated antigen processing machines. Cell 106:255.[Medline]
3. Lanzavecchia, A., F. Sallusto. 2001. Regulation of T cell immunity by dendritic cells. Cell 106:263.[Medline]
4. Liu, Y. J.. 2001. Dendritic cell subsets and lineages, and their functions in innate and adaptive immunity. Cell 106:259.[Medline]
5. Schuler, G., B. Schuler-Thurner, R. M. Steinman. 2003. The use of dendritic cells in cancer immunotherapy. Curr. Opin. Immunol. 15:138.[Medline]
6. Szabolcs, P., D. Avigan, S. Gezelter, D. H. Ciocon, M. A. S. Moore, R. M. Steinman, J. W. Young. 1996. Dendritic cells and macrophages can mature independently from a human bone marrow-derived, post-CFU intermediate. Blood 87:4520.[Abstract/Free Full Text]
7. Caux, C., B. Vanbervliet, C. Massacrier, C. Dezutter-Dambuyant, B. de Saint-Vis, C. Jacquet, K. Yoneda, S. Imamura, D. Schmitt, J. Banchereau. 1996. CD34⁺ hematopoietic progenitors from human cord blood differentiate along two independent dendritic cell pathways in response to GM-CSF + TNF . J. Exp. Med. 184:695.[Abstract/Free Full Text]
8. Caux, C., C. Massacrier, B. Vanbervliet, B. Dubois, I. Durand, M. Cella, A. Lanzavecchia, J. Banchereau. 1997. CD34⁺ hematopoietic progenitors from human cord blood differentiate along two independent dendritic cell pathways in response to granulocyte-macrophage colony-stimulating factor plus tumor necrosis factor : II. Functional analysis. Blood 90:1458.[Abstract/Free Full Text]
9. Strobl, H., E. Riedl, C. Scheinecker, C. Bello-Fernandez, W. F. Pickl, K. Rappersberger, O. Majdic, W. Knapp. 1996. TGF-1 promotes in vitro development of dendritic cells from CD34⁺ hemopoietic progenitors. J. Immunol. 157:1499.[Abstract]
10. Gatti, E., M. A. Velleca, B. C. Biedermann, W. Ma, J. Unternaehrer, M. W. Ebersold, R. Medzhitov, J. S. Pober, I. Mellman. 2000. Large-scale culture and selective maturation of human Langerhans cells from granulocyte colony-stimulating factor-mobilized CD34⁺ progenitors. J. Immunol. 164:3600.[Abstract/Free Full Text]
11. Sallusto, F., A. Lanzavecchia. 1994. Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregulated by tumor necrosis factor . J. Exp. Med. 179:1109.[Abstract/Free Full Text]
12. Romani, N., S. Gruner, D. Brang, E. Kampgen, A. Lenz, B. Trockenbacher, G. Konwalinka, P. O. Fritsch, R. M. Steinman, G. Schuler. 1994. Proliferating dendritic cell progenitors in human blood. J. Exp. Med. 180:83.[Abstract/Free Full Text]
13. Thurner, B., C. Roder, D. Dieckmann, M. Heuer, M. Kruse, A. Glaser, P. Keikavoussi, E. Kampgen, A. Bender, G. Schuler. 1999. Generation of large numbers of fully mature and stable dendritic cells from leukapheresis products for clinical application. J. Immunol. Methods 223:1.[Medline]
14. Rissoan, M. C., V. Soumelis, N. Kadowaki, G. Grouard, F. Briere, R. de Waal Malefyt, Y. J. Liu. 1999. Reciprocal control of T helper cell and dendritic cell differentiation. Science 283:1183.[Abstract/Free Full Text]
15. Tang, A., M. Amagai, L. G. Granger, J. R. Stanley, M. C. Udey. 1993. Adhesion of epidermal Langerhans cells to keratinocytes mediated by E-cadherin. Nature 361:82.[Medline]
16. Riedl, E., J. Stockl, O. Majdic, C. Scheinecker, K. Rappersberger, W. Knapp, H. Strobl. 2000. Functional involvement of E-cadherin in TGF-1-induced cell cluster formation of in vitro developing human Langerhans-type dendritic cells. J. Immunol. 165:1381.[Abstract/Free Full Text]
17. Valladeau, J., V. Duvert-Frances, J. J. Pin, C. Dezutter-Dambuyant, C. Vincent, C. Massacrier, J. Vincent, K. Yoneda, J. Banchereau, C. Caux, J. Davoust, S. Saeland. 1999. The monoclonal antibody DCGM4 recognizes Langerin, a protein specific of Langerhans cells, and is rapidly internalized from the cell surface. Eur. J. Immunol. 29:2695.[Medline]
18. Valladeau, J., O. Ravel, C. Dezutter-Dambuyant, K. Moore, M. Kleijmeer, Y. Liu, V. Duvert-Frances, C. Vincent, D. Schmitt, J. Davoust, et al 2000. Langerin, a novel C-type lectin specific to Langerhans cells, is an endocytic receptor that induces the formation of Birbeck granules. Immunity 12:71.[Medline]
19. Borkowski, T. A., J. J. Letterio, A. G. Farr, M. C. Udey. 1996. A role for endogenous transforming growth factor 1 in Langerhans cell biology: the skin of transforming growth factor 1 null mice is devoid of epidermal Langerhans cells. J. Exp. Med. 184:2417.