Open Access

Human XCR1+ Dendritic Cells Derived In Vitro from CD34+ Progenitors Closely Resemble Blood Dendritic Cells, Including Their Adjuvant Responsiveness, Contrary to Monocyte-Derived Dendritic Cells

Sreekumar Balan, Vincent Ollion, Nicholas Colletti, Rabie Chelbi, Frédéric Montanana-Sanchis, Hong Liu, Thien-Phong Vu Manh, Cindy Sanchez, Juliette Savoret, Ivan Perrot, Anne-Claire Doffin, Even Fossum, Didier Bechlian, Christian Chabannon, Bjarne Bogen, Carine Asselin-Paturel, Michael Shaw, Timothy Soos, Christophe Caux, Jenny Valladeau-Guilemond and Marc Dalod
J Immunol August 15, 2014, 193 (4) 1622-1635; DOI: https://doi.org/10.4049/jimmunol.1401243

Abstract

Human monocyte-derived dendritic cell (MoDC) have been used in the clinic with moderately encouraging results. Mouse XCR1+ DC excel at cross-presentation, can be targeted in vivo to induce protective immunity, and share characteristics with XCR1+ human DC. Assessment of the immunoactivation potential of XCR1+ human DC is hindered by their paucity in vivo and by their lack of a well-defined in vitro counterpart. We report in this study a protocol generating both XCR1+ and XCR1− human DC in CD34+ progenitor cultures (CD34-DC). Gene expression profiling, phenotypic characterization, and functional studies demonstrated that XCR1− CD34-DC are similar to canonical MoDC, whereas XCR1+ CD34-DC resemble XCR1+ blood DC (bDC). XCR1+ DC were strongly activated by polyinosinic-polycytidylic acid but not LPS, and conversely for MoDC. XCR1+ DC and MoDC expressed strikingly different patterns of molecules involved in inflammation and in cross-talk with NK or T cells. XCR1+ CD34-DC but not MoDC efficiently cross-presented a cell-associated Ag upon stimulation by polyinosinic-polycytidylic acid or R848, likewise to what was reported for XCR1+ bDC. Hence, it is feasible to generate high numbers of bona fide XCR1+ human DC in vitro as a model to decipher the functions of XCR1+ bDC and as a potential source of XCR1+ DC for clinical use.

This article is featured in In This Issue, p.1517

Introduction

Dendritic cells (DC) are a heterogeneous family of rare leukocytes that sense danger signals and convey them to lymphocytes for the orchestration of adaptive immune defenses. Clinical trials used monocyte-derived DC (MoDC) to attempt to promote protective immunity in patients suffering from infections or cancer. These immunotherapies showed limited efficacy, owing to the poor recirculation of MoDC to lymph nodes (1, 2) and likely to other yet uncharacterized functional differences between MoDC and lymphoid tissue–resident DC (LT-DC). Hence, major efforts are being made to better characterize human LT-DC and to evaluate their immunoactivation potential.

Steady-state human blood and secondary lymphoid organs contain three major DC subsets, CD141(BDCA3)+CLEC9A+ DC, CD1c(BDCA1)+ DC, and CLEC4C(BDCA2)+ plasmacytoid DC (pDC) (3). Homologies exist between mouse and human LT-DC subsets (4, 5). Comparative transcriptomics (4–7) and functional studies (7–12) showed that human CD141+CLEC9A+ DC are homologous to mouse spleen CD8α+ DC, which are specialized in cross-presentation. Mouse CD8α+ DC and human CD141+CLEC9A+ DC specifically express the XCR1 chemokine receptor (4, 8, 9, 13, 14) and can therefore be coined XCR1+ DC. The ligands of XCR1 are selectively expressed in NK and CD8 T cells, promoting their interactions with XCR1+ DC (13). Human XCR1+ DC have been described in many tissues (6, 7, 15). Human and mouse XCR1+ DC specifically express high levels of TLR-3 (16) and respond to its triggering with hepatitis C virus or with the synthetic ligand polyinosinic-polycytidylic acid (PolyI:C) by IFN-λ production (12, 15, 17) and by enhanced cross-presentation (7–11). The extent to which human XCR1+ DC are more efficient for cross-presentation than other human DC subsets is debated. It depends on the tissue origin of the DC subsets, on their activation status, and on the mode of Ag delivery (7–10, 18–23). However, several independent studies showed that human XCR1+ blood DC (bDC) excel at cross-presentation of cell-associated Ags (8–10, 18) and of particulate Ags delivered through FcγRs, through lysosomes (19, 20) or upon polyI:C stimulation (8, 10, 23). Because they share unique characteristics with mouse XCR1+ DC, human XCR1+ bDC constitute a distinct human DC subset that may have potential clinical applications (24–26). To determine whether and how human XCR1+ DC could be harnessed in the clinic, it is necessary to better characterize them. This includes determining their global responses to adjuvants. A major bottleneck for such studies is the paucity of XCR1+ DC in human tissues. Generating high numbers of XCR1+ DC in vitro from CD34+ hematopoietic progenitors will help overcome this problem.

Human DC derived in vitro from monocytes or CD34+ progenitors have been widely used to investigate DC biology. Only MoDC (27) and Langerhans cells (LC) (28) have been generated reproducibly in vitro by many teams. Gene expression profiling and/or ontogenic studies showed that MoDC, Langerhans cell, and DC lines strikingly differ from LT-DC subsets and are more similar to monocytes or macrophages (4, 29–32). Cultures of CD34+ progenitors with FLT3-ligand (FLT3-L) or thrombopoietin (TPO) have been reported to yield pDC, CD141+CLEC9A+ DC, or CD1c+ DC (11, 33, 34). The extent to which the CD141+CLEC9A+, the CD1c+, or the other HLA-DR+ cells in these cultures resembled human bDC subsets was, however, not thoroughly examined. In this study, we report a protocol generating high numbers of both XCR1+ DC and XCR1− DC in CD34+ progenitor cultures (CD34-DC), with a high-throughput characterization of these DC subsets by gene expression and secreted molecule profiling under steady-state and activation conditions and by comparison with XCR1+ bDC and MoDC.

Materials and Methods

Isolation of cord blood CD34+ cells

Human nontherapeutic-grade cord blood (CB) samples were collected from informed and consenting mothers at the Etablissement Français du Sang (EFS) Lyon in the frame of an EFS–INSERM agreement or at the Marseille University Hospital de la Conception birth clinic and obtained through the Marseille CB bank. Marseille CB bank is a not for-profit allogeneic CB bank affiliated with the French consortium of CB bank (RFSP, Reseau Francais du Sang Placentaire) and the “Agence de Bioledecine.” CB units that were disqualified for therapeutic use upon reception at the Institut Paoli-Calmettes cell therapy facility, because of insufficient volume or total cell number, were used for research, as indicated to pregnant mothers in the information document. The protocol was reviewed by the Institut Paoli-Calmettes Comité d’Orientation Stratégique (COS or IRB). CD34+ cells were enriched magnetically by using the Dynal (Invitrogen) or Miltenyi Biotec CD34+ cell isolation kit. CD34+ cells from adult donor bone marrow or mobilized peripheral blood were purchased from properly licensed commercial companies (AllCells and StemCells).

In vitro generation of MoDC from enriched monocytes derived from CB CD34+ cells

MoDC were differentiated from enriched monocytes derived from CB CD34+ cells, as adapted from a published protocol (35). CD34+ cells were expanded for 21 d in expansion#1 medium (stemspan from StemCell Technologies; 10% FBS, 25 ng/ml Flt-3L, 20 ng/ml stem cell factor [SCF], and 10 ng/ml TPO). The cells were replated in weekly intervals at 5 × 104 cells/ml, 2 ml/well, in 6-well plates, in fresh expansion#1 medium. On day 14, the cells were plated in 100-mm dishes at 2 × 105 cells/ml, 10 ml/dish, in fresh expansion#1 medium. On day 21, CD14+ monocytes were enriched by positive selection through MACS (Miltenyi Biotec) and differentiated for 7 d in differentiation#1 medium (RPMI 1640 medium (Sigma-Aldrich), 10% FBS, 50 ng/ml GM-CSF, and 25 ng/ml IL-4) (GM4 protocol). The growth factors and cytokines used were from PeproTech unless specified otherwise.

In vitro generation of XCR1+ and XCR1− DC from CB CD34+ cells

XCR1+ CD34-DC cultures were performed accordingly to the FS36 protocol (11) and modifications thereof as described below. For the FS36 protocol, CD34+ cells were cultured for 7 d at 5 × 104 cells/ml, 2 ml/well, in 6-well plates, with expansion#2 medium (stemspan; 100 ng/ml Flt-3L, 100 ng/ml SCF, 20 ng/ml IL-3, and 20 ng/ml IL-6) in the absence of serum supplementation. Expanded cells were harvested on day 7. They were used for DC generation, either directly or after cryopreservation, through culture in U-bottom 96-well plates in differentiation#2 medium (RPMI 1640 medium; 10% FCS, 100 ng/ml Flt-3L, 20 ng/ml SCF, 20 ng/ml GM-CSF, and 20 ng/ml IL-4). On day 6, half of the medium was replaced, and fresh cytokines were added. Cultures were harvested at days 8–11. In the adapted version FS36A, 5% heat-inactivated autologous plasma was added to the expansion medium. In the FS3TA protocol, for the expansion phase, 50 ng/ml TPO was substituted to IL-6 (expansion#3 medium), and 5% heat-inactivated autologous plasma was added. The FS3T protocol was based on the FS3TA with substitution of autologous plasma by 10% FBS. The optimized FS3T* protocol was based on the FS3T protocol, with reduced cell density and culture in 96-well plates during the expansion phase and decreased concentration of GM-CSF and IL-4 during the differentiation phase. In detail, for the FS3T* protocol, CD34+ cells were cultured for 7 d at 2.5 × 104 cells/ml, 200 μl/well, in round-bottom 96-well plates, with expansion#3 medium (stemspan; 100 ng/ml Flt-3L, 100 ng/ml SCF, 20 ng/ml IL-3, and 50 ng/ml TPO) with 10% FCS. Expanded cells were differentiated in RPMI 1640 medium plus 10% FCS, with 100 ng/ml Flt-3L, 20 ng/ml SCF, 2.5 ng/ml GM-CSF, and 2.5 ng/ml IL-4 in U-bottom 96-well plates (6.25 × 104 cells/ml and 200 μl well). On day 6, half of the medium was replaced, and fresh cytokines were added. Cultures were harvested on days 8–13, depending on the experiments.

Surface staining and sorting of in vitro–generated DC

Cells were stained with Abs (Table I) for 30 min at 4°C, washed with 1× PBS, and acquired on a LSR II or Canto II flow cytometer (BD Biosciences). Human XCR1 was stained with a fluorescently coupled recombinant human chemokine ligand vaccibody (rhXCL1-mCherry) (36). FS36 and FS3T cultures were sorted into >97% pure DC subsets with a FACSAria cytometer (BD Biosciences), based on viability (negativity for Sytox blue; Invitrogen) and expression of CD141 and CD11c.

Purification of peripheral blood DC

PBMC were prepared by Ficoll-Hypaque (GE Healthcare) gradient centrifugation from buffy coats received from EFS (Marseille) and depleted of monocytes, B cells, and T cells by negative selection using LD columns and magnetic beads coupled to Abs against CD3, CD14, and CD19 (Miltenyi Biotec). Enriched cells were stained for 30 min at 4°C in 1× PBS and 0.5% BSA. Viable (Sytox blue-negative) DC subsets were sorted with a FACSAria cytometer (BD Biosciences) to a purity > 96%, as Lin−HLA-DR+CD11c+CD1c−CLEC4C−CD141high cells for XCR1+ DC, Lin−HLA-DR+CD11chighCD1c+CLEC4C−CD141−/low cells for CD1c+ DC, and Lin−HLA-DR+CD11c−CD1c−CLEC4C+CD141−/low cells for pDC.

Cytospin and Wright–Giemsa stainings

Cells were diluted at 10–15 × 103 cells/50 μl and used for the cytospin at 400 rpm for 4 min. The slides were air-dried for 15 min, fixed in methanol for 7 min, and stained with Wright–Geimsa solution (Sigma-Aldrich). Images were taken with a Nikon eclipse (camera “DS-Fi2) microscope.

TLR stimulation of DC subsets

Optimal doses of adjuvants were defined by titration curves in assays assessing the phenotypic maturation of CD34-DC cultures (data not shown). Sorted DC subsets were stimulated for 16 h with PolyI:C (5 μg/ml), R848 (10 μg/ml), or LPS (1 μg/ml) (all from InvivoGen). Cells were then examined for phenotypic maturation or used for RNA extraction. Sorted bDC were maintained in Flt-3L (100 ng/ml) plus GM-CSF (20 ng/ml). MoDC were stimulated on day 6.

Cytokine production assays

Cytokines were titrated in culture supernatants with a custom procarta immune assay kit following the instructions from the manufacturer. Data were acquired with the Magpix reader using the xPONENT 4.2 software and analyzed with the Milliplex Analyst 5.1 Software.

RNA extraction and microarray experiments

Total RNA was prepared with the RNeasy Micro kit (Qiagen). The quantity and integrity of mRNA were measured using the 2100 Bioanalyzer system with picochips (Agilent Technologies). The RNA integrity numbers were generally above 8 and never below 7. Microarray experiments were performed at the Plate-forme Biopuces (Strasbourg, France) (http://www-microarrays.u-strasbg.fr). In brief, cDNA was generated and labeled using the Ribo-SPIATM Whole Transcript Amplification Process (NuGen) using as a starting material 20 ng total RNA for most of the in vitro–derived DC and 1 ng for ex vivo–isolated blood DC. GeneChip Human Gene 1.0 ST DAT files were processed using Affymetrix Gene Chip Operating System to generate .CEL files. Quality control of the array hybridization and normalization of the raw Affymetrix expression data with Robust Multi-chip Analysis 2 (37) were performed through Bioconductor in the R statistical environment (version 2.14.1) using the oligo package. Density plots of the normalized expression data were generated to estimate the background noise. For TLR stimulations of CD34-DC subsets, differentially expressed genes between conditions were identified by General Linear Model with paired analysis comparing PolyI:C, R848, or LPS-treated samples to unstimulated controls for each DC subset of each individual CB culture tested. For hierarchical clustering and heat maps, each gene was represented by only one ProbeSet selected as the one giving the highest signal. Principal component analysis (PCA) was computed with the MeV software and represented with the R software (scatterplot3D package). Hierarchical clustering was made with the pvclust package in R to evaluate the robustness of the branches of the tree through multiscale bootstrap resampling of the genes (http://www.is.titech.ac.jp/∼shimo/prog/pvclust/). Heat-map representations were made with the GeneE software (http://www.broadinstitute.org/cancer/software/GENE-E/). Genes were clustered based on coregulation analysis using BioLayout Express3D for computing Pearson’s correlation coefficients and for graphical representation (38).

Allogeneic CD4+ T cell stimulation

CD4+ T cells were isolated from buffy coat with the RosetteSep human CD4+ T cell enrichment mixture (StemCell Technologies). CD4+ T cells were labeled with CFSE (2 μM) and cocultured with DC at 20:1 ratio for 6 d. T cell proliferation and CFSE dilution was assessed by FACS after input of a same number of latex beads in each well and staining with PE-CD3 and allophycocyanin-CD4. Absolute cell counts were determined through normalization by acquisition of the same number of latex beads per sample.

Generation of K562-N9V or GFP expressing cell lines

We used lentiviral- based vectors to generate the K562 cell line expressing the N9V epitope of the CMV pp65 fused in frame at the 5′-end to cytoplasmic OVA. The HIV lentiviral backbone is based on self-inactivating 3′-long terminal repeat vectors. This vector was engineered to carry first the human CMV early enhancer/chicken β-actin hybrid promoter driving the N9V-OVA protein, followed by the spleen focus-forming virus promoter driving the enhanced GFP reporter gene. To increase the level of transcription the woodchuck hepatitis virus posttranscriptional regulatory element was inserted downstream of N9V-OVA. Amplification of the fusion protein was done using a vector encoding a full OVA as template and the following primers: forward N9V-CytOVA, 5′-GCCTCGAGTGAGCCACCATGATCCTGGCCCGCAACCTGGTGCCCATGGTGGCTACGGTTCAGGGTCAGAATACCAGGACACAAATAAATAAGGTTGTTCG-3′, and RVS-OVA, 5′-CTGCTCAAGGCCTGAGACTTC-3′. Viruses were pseudotyped with vesicular stomatitis virus glycoprotein, infections at MOI 5 were performed, and 5 d posttransduction, GFP-positive K562 cells were cell sorted, expanded in culture, and used as a stable bulk population. GFP expression was checked before each experiment. Negative controls for cross-presentation experiments always included K562 cells infected with a lentivirus encoding GFP alone, which never induced IFN-γ production by the CD8 T cell clone (data not shown).

Cross-presentation assay

A total of 105 K562 cells expressing the N9V (pp65493-503) epitope from the major CMV structural protein and 0.5 × 105 DC were cocultured in round-bottom 96-well plates in complete medium for 24–36 h with or without PolyI:C (5 μg/ml) or R848 (10 μg/ml) (InvivoGen). As positive control, 10 nM soluble N9V peptide was added to DC for 1 h before extensive washes. An N9V-specific T cell clone (0.5 × 105cells/well) (39) was then added in the presence of GolgiStop (BD Biosciences) for an additional 12 h (ratio DC:T = 1:2). After fixation and permeabilization, T cell activation was determined as the percentage of IFN-γ–expressing cells within CD3+GFP− cells.

Statistics

Statistics for gene chips data analyses are detailed in Materials and Methods and the figure legends. Significance for other experimental data was evaluated with Graph pad Prism 6, using the tests indicated in the figure legends.

Accession numbers

The microarray data have been submitted to the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/gds) under accession number GSE57671.

Results

Combined use of FLT3-L and TPO in CD34+ progenitor cultures yields high numbers of XCR1+ DC

We reproduced what to our knowledge is the first published protocol for in vitro generation of XCR1+ DC from CB CD34+ progenitors (11). It consisted in an initial 7-d expansion phase with FLT3-L, SCF, IL-3, and IL-6 (“FS36” cultures), followed by a 13-d differentiation phase with FLT3-L, SCF, GM-CSF, and IL-4 (Supplemental Fig. 1A). We tested modifications of this protocol by varying the cytokines used for the expansion phase (data not shown), in particular IL-6 replacement by TPO (“FS3T” cultures). FS36 and FS3T cultures were both composed in majority of CD11chighCD141+/−HLA-DR+ cells (Fig. 1A, Table I, Supplemental Fig. 1B), which lacked expression of XCR1, CLEC9A, and CADM1 (Fig. 1B, Supplemental Fig. 1C) and were therefore named XCR1− CD34-DC. The cultures also consistently encompassed CD11clowCD141highHLA-DR+ cells (Fig. 1A, Supplemental Fig. 1B), which shared with XCR1+ bDC (36) specific expression of XCR1, CLEC9A, and CADM1 (Fig. 1B, Supplemental Fig. 1C) and were thus named XCR1+ CD34-DC. MoDC were derived from the same CB CD34+ cells (Supplemental Fig. 1A) (35). The morphology of XCR1+ CD34-DC was similar to that of XCR1+ bDC, and XCR1− CD34-DC resembled MoDC (Fig. 1C). FS3T cultures yielded higher percentages and absolute numbers of XCR1+ CD34-DC (Supplemental Fig. 1D, 1E). The yields of XCR1+ CD34-DC were improved by lowering cell density and adding 10% FBS during the expansion phase. During the differentiation phase, lower concentrations of GM-CSF and IL-4 were sufficient to yield maximal numbers of XCR1+ CD34-DC (data not shown). XCR1+ CD34-DC were not detected at the end of the expansion phase; their development during the differentiation phase required both GM-CSF and IL-4; they reached maximal frequencies ∼11 d after the initiation of differentiation (data not shown). Hence, unless specified otherwise, XCR1+ CD34-DC were generated with an optimized protocol coined FS3T*, by culturing cells during expansion at 2.5 × 104 cells/ml, in U-bottom 96-well plates, with 10% FBS and, during differentiation, by decreasing GM-CSF and IL-4 concentrations to 2.5 ng/ml and harvesting cultures at day 11. FS3T* cultures yielded 1.8 ± 1.5 × 106 XCR1+ CD34-DC from 105 CD34+ CB progenitors (Supplemental Fig. 1E).

FIGURE 1.
FIGURE 1.

Characterization of in vitro–generated CD34-DC subsets. (A) Identification of cell subsets in cultures of CD34+ CB progenitors based on the expression of CD11c and CD141 for FS3T cultures (left contour plot, double-negative cells in violet, XCR1− CD34-DC in green and XCR1+ CD34-DC in red) and G4 cultures (right contour plot, MoDC in black). (B) Expression of XCR1, CLEC9A, and CADM1 on each of the cell subsets identified in (A), using the same color code. Isoptype control stainings are shown as gray curve on each histogram. For (A) and (B), one representative result of at least three independent cultures is shown. (C) Microscopy analysis of the morphology of Giemsa/May–Grünwald-stained DC subsets sorted from the cultures of CD34+ CB progenitors or from the blood of adult healthy donors. (D–G) Gene expression profiling of XCR1+ CD34-DC, XCR1− CD34-DC, and MoDC. Microarrays were performed on total RNA extracted from DC subsets FACS purified from FS3T culture (XCR1+ CD34-DC and XCR1− CD34-DC), from G4 cultures (MoDC), or from the blood of adult healthy donors (bpDC, CD1c+ bDC, and XCR1+ bDC). (D) PCA performed on all ProbeSets. (E) Hierarchical clustering performed on the 3934 genes showing a fold change ≥ 2 between at least two of the cell types studied, using Pearson and average as distance metric/linkage parameters. The numbers above edges indicate the robustness of the corresponding node, calculated as the percentage of occurrence of this node among 1000 independent trees generated by multiscale bootstrap resampling. (F) BioLayoutExpress clustering of the same set of genes as used in (E). On the left, the size of the different clusters and their relationships are shown as tridimensional clouds with each gene represented as a node colored accordingly to the gene clusters to which it belongs. The calculation parameters used were a cutoff for Pearson correlation coefficient at 0.8 and an inflation value at 2. Only clusters encompassing ≥50 genes are represented. On the right, the average gene expression pattern across all cell types examined is shown for each cluster. (G) Heat maps showing the expression patterns for 20 individual genes selected from each cluster mostly based on prior knowledge of their selective expression in ex vivo–isolated bDC subsets (bold red for XCR1+ bDC-specific genes) or monocyte-derived inflammatory DC (bold green) and/or based on prior knowledge of their functions.

View this table:
Table I. Abs and other reagents used for flow cytometry stainings

XCR1+ and XCR1− CD34-DC respectively expressed XCR1+ bDC and MoDC transcriptomic fingerprints

To determine in the broadest and most unbiased way possible to which extent XCR1+ CD34-DC resembled XCR1+ bDC, we profiled their gene expression. PCA (Fig. 1D) demonstrated a clear proximity between XCR1+ CD34-DC and XCR1+ bDC, and between MoDC and XCR1− CD34-DC, along the PC1 axis representing the majority of gene expression variability in the dataset. Cell type distribution along PC1 was explained by their differential expression of many genes previously identified as being selectively expressed either in XCR1+ bDC or in MoDC (4, 40) (Supplemental Table I). On the PC2 axis, XCR1+ CD34-DC clearly segregated away from XCR1+ bDC, which resulted in part from their higher expression of mitotic genes and lower expression of lysosomal genes (Supplemental Table I). Hence, XCR1+ CD34-DC are a good model of human XCR1+ bDC, contrary to MoDC, but their differentiation might be stopped at a preterminal stage as it has been reported for in vitro generated mouse XCR1+ DC (41). Hierarchical clustering confirmed the strong similarity between XCR1+ CD34-DC and XCR1+ bDC and between MoDC and XCR1− CD34-DC (Fig. 1E). Coregulation analysis using BioLayoutExpress (Fig. 1F) clustered together genes with a high selective expression in MoDC and XCR1− CD34-DC (C1, n = 1228) or in XCR1+ CD34-DC and XCR1+ bDC (C2, n = 747 and C4, n = 188), in consistency with the transcriptomic fingerprints previously reported for in vitro–derived (4) or ex vivo–isolated (40) human MoDC versus for human blood, skin, and intestine XCR1+ DC (Fig. 1G) (4, 6, 7). CD34-DC subsets from FS36 and FS3T cultures harbored similar gene expression programs (Supplemental Fig. 2). The expression patterns of several genes were confirmed at the protein level (Fig. 2). Bona fide XCR1+ DC also could be efficiently generated from CD34+ progenitors isolated from bone marrow (Fig. 3) or blood (data not shown) of adult donors. In conclusion, to our knowledge, we demonstrated for the first time that a culture protocol of CB CD34+ cells generated simultaneously different DC subsets along two independent ontogenic pathways, XCR1− CD34-DC related to the monocyte lineage and similar to MoDC as well as XCR1+ CD34-DC resembling XCR1+ bDC.

FIGURE 2.
FIGURE 2.

Phenotypic analyses confirm microarray data regarding the identity of CD34-DC subsets. DC cultures were stained with Abs against TLR3 or a variety of cell surface markers selected based on available Abs and differential expression of the corresponding genes between XCR1+ versus XCR1− CD34-DC as shown in Fig. 1G. Data were analyzed as in Fig. 1A and 1B. One representative result of at least three independent cultures is shown.

FIGURE 3.
FIGURE 3.

Efficient differentiation of XCR1+ DC in FS3T cultures of CD34+ hematopoietic progenitors from the bone marrow of adult donors. (A) Identification of cell subsets in cultures of adult bone marrow CD34+ progenitors based on the expression of CD11c and CD141, for FS3T cultures, similarly to what is shown for CB cultures on Fig. 1A. (B) Expression of selected markers on each of the cell subsets identified in (A), similarly to what is shown for CB cultures cultures on Figs. 1B and 2. One representative result of at least two independent cultures is shown.

XCR1+ versus XCR1− CD34-DC show reciprocal responsiveness to TLR3 versus TLR4 triggering

Because we generated XCR1+ CD34-DC and MoDC-like XCR1− CD34-DC from the same CD34+ cells in the same cultures, it allowed a rigorous side-by-side comparison of the cell-intrinsic immunoactivation potential of these two DC subsets, without potential confounding effects linked to differentiation from distinct cell progenitors and with different cytokines. XCR1+ CD34-DC matured phenotypically in response to PolyI:C but not to LPS, and conversely for XCR1− DC and MoDC (Fig. 4A), in consistency with the opposite expression pattern in CD34-DC subsets of the respective receptors of these adjuvants, TLR3 and TLR4 (Fig. 2). TLR7/8 activation by R848 induced an intermediate response in all DC subsets (Fig. 4A). These response patterns were confirmed by gene expression profiling. In particular, both PCA (Fig. 4B) and hierarchical clustering (Fig. 4C) showed that PolyI:C stimulation of XCR1+ DC and LPS stimulation of XCR1− CD34-DC and MoDC strongly affected their gene expression programs, setting them apart from their unstimulated counterparts. LPS stimulation of XCR1+ CD34-DC and PolyI:C stimulation of XCR1− CD34-DC and MoDC did not strongly affect their gene expression programs. The highest number of differentially expressed genes (n = 2010) was observed for PolyI:C stimulation of XCR1+ DC (Fig. 4D, Supplemental Table I), followed by LPS stimulation of MoDC (n = 1447) and XCR1− CD34-DC (n = 1047).

FIGURE 4.
FIGURE 4.

XCR1+ CD34-DC are uniquely and very strongly activated by polyI:C but fail to respond to LPS, whereas XCR1− CD34-DC and MoDC show a reverse pattern. (A) Phenotypic maturation of CD34-DC subsets in response to stimulation with PolyI:C, R848, or LPS. MoDC generated from five different CB samples and seven independent cultures and XCR1− versus XCR1+ CD34-DC subsets sorted from nine independent CB cultures representing five different CB samples were incubated for 16 h in medium alone (control) or with PolyI:C, R848, or LPS. Cells were then stained with Abs for the maturation markers CD83, CD86, CD40, or isotype control Abs. Results are represented as fold change (FC) of mean fluorescence intensity (MFI) above control cultures (left graphs) or as percentage of positive cells (right graphs) as a function of stimuli for each individual CB samples (symbols). To determine significance, a Wilcoxon matched-pairs signed rank test was performed on 10 independent cultures from five different donors. *p < 0.05, **p < 0.01, ***p < 10−3. (B–D) Gene expression profiling of CD34-DC subsets exposed to PolyI:C, R848, or LPS. Genome-wide expression analysis was performed on XCR1+ versus XCR1− CD34-DC subsets sorted from three to six independent FS3T CB cultures and on three independent MoDC cultures. (B) PCA performed on all ProbeSets. (C) Unsupervised hierarchical clustering of stimulated DC subsets performed on the 5015 genes showing a FC ≥ 1.5 between at least two conditions. (D) Venn diagrams showing the number of genes induced (UP) or repressed (DOWN) in each DC subset by R848 (left), PolyI:C (middle), or LPS (right), as well as the overlap for a given stimulus between DC subset responses.

Hundreds of genes are similarly modulated in PolyI:C-stimulated XCR1+ CD34-DC and LPS-stimulated XCR1− CD34-DC or MoDC

PCA showed a parallel shift in the gene expression programs of each DC subset from unstimulated to optimally activated states (Fig. 4B). Indeed, 370 genes were similarly modulated in response to PolyI:C in XCR1+ CD34-DC and in response to LPS in XCR1− CD34-DC or MoDC (Fig. 5, Supplemental Table I). Ingenuity pathway analyses found many annotations to be significantly enriched for the genes commonly upregulated in mature DCs (Fig. 5), highlighting the contribution of this core maturation gene set to the canonical process of DC maturation, the activation of NK or T cells or more generally the immune defenses against bacteria or viruses, in a large part under the instruction of NFKB or type I IFN (IFN-I) (Fig. 5C).

FIGURE 5.
FIGURE 5.

Comparison of the genetic reprogramming of XCR1+ DC in response to PolyI:C and XCR1− DC or MoDC in response to LPS. Venn diagrams were performed on the data from Fig. 4E to compare the sets of genes induced (A) or repressed (B) in XCR1+ DC in response to PolyI:C and XCR1- DC or MoDC in response to LPS. Ingenuity pathway analyses were performed for each set of genes as defined by the different areas of the Venn diagrams to determine whether these lists where enriched for genes annotated for specific pathways. The results are shown as heat maps where the color intensity increases with statistical significance of the annotation enrichment. (C) A heat map showing the expression patterns of the genes responsible for the enrichment of the pathways NF-KB signaling and IFN signaling shown in (A). The red font identifies the genes previously reported to be part of the core transcriptomic signature commonly upregulated in mouse DC subsets and in human pDC or MoDC upon activation with a variety of microbial stimuli.

Many genes are modulated in specific combinations of DC subset/stimulus

Many genes were significantly modulated in specific combinations of DC subset/stimulus (Fig. 5A, 5B Supplemental Table I), likely determining the nature of the specific functions exerted by mature DC subsets. Fewer annotations were enriched for these gene sets and with lower statistical significance. Under steady-state conditions, XCR1− CD34-DC or MoDC harbored a higher expression of genes coding for inflammatory molecules than XCR1+ CD34-DC (Fig. 6). TLR4 or TLR8 triggering of XCR1− CD34-DC and MoDC further enhanced the expression of these inflammatory molecules and induced additional ones, including IL18 and IL12B (Fig. 6). IL-10 was only detected in XCR1− CD34-DC and MoDC in response to R848 or LPS (Fig. 6A, 6B), consistent with production of this cytokine only by CD14+ DC but not XCR1+ DC in skin (7, 42). As described for XCR1+ bDC in vitro (12), XCR1+ CD34-DC uniquely expressed IFN-λ upon TLR3 triggering (Fig. 6A, 6B). Many other genes were newly found to be selectively induced to high levels in XCR1+ CD34-DC upon PolyI:C stimulation (Fig. 6, Supplemental Table I), including genes encoding antiapoptotic, antiviral, or costimulation molecules as well as cytokines/chemokines acting on NK or T cells.

FIGURE 6.
FIGURE 6.

Cytokine production by CD34-DC subsets and MoDC. DC subsets were sorted and stimulated as described in Fig. 4. (A) Expression pattern of selected genes. Genes highlighted in bold green are 1) already expressed to higher levels in steady state XCR1− CD34-DC or MoDC, as compared with XCR1+ CD34-DC (except for IL12B and IL18, which are not detected at steady state), 2) further induced specifically in XCR1− CD34-DC or MoDC upon R848 or LPS stimulation, and 3) encoded for molecules with proinflammatory or matrix remodeling functions, except for IL-10, which is anti-inflammatory. Genes highlighted in bold red are induced to higher levels in PolyI:C-activated XCR1+ CD34-DC as compared with all the other conditions examined and encode for antiviral, antiapoptotic, or NK/T cell–stimulating molecules. (B) Secretion of selected cytokines. Culture supernatants were used for Luminex-based detection of 18 different analytes. Results from three to eight independent cultures for each DC subset are shown as individual points, with bar overlays indicating mean ± SD. Dotted lines in the graph show detection (bottom) or saturation (top) thresholds. Color codes for molecule names are the same as in (A).

The response pattern of XCR1+ CD34-DC and XCR1+ bDC to candidate adjuvants are strikingly similar

To examine to which extent the responses of XCR1+ CD34-DC to adjuvants were similar to those of XCR1+ bDC, we compared the gene expression profiling of CD1c+ bDC, XCR1+ bDC, XCR1+ CD34-DC, and XCR1− CD34-DC upon stimulation with R848, PolyI:C, or LPS. As observed for XCR1+ CD34-DC, XCR1+ bDC were not activated by LPS, moderately by R848, and strongly by PolyI:C (Fig. 7A, 7B). PolyI:C-stimulated XCR1+ CD34-DC, XCR1+ bDC, and CD1c+ bDC regrouped together, away from other conditions of stimulation of these cell types and from XCR1− CD34-DC (Fig. 7A, 7B) or MoDC (data not shown). Cell type distribution along PCA axes resulted in a large part from their differential expression of many genes already at steady state (Fig. 7C, clusters 1 and 5–8) as well as from selective induction of genes in 1) CD1c+ bDC and XCR1− CD34-DC by R848 and LPS (subcluster A), 2) bDC and XCR1+ CD34-DC by PolyI:C as well as XCR1− CD34-DC by both PolyI:C and LPS (cluster 2), and 3) bDC and XCR1+ CD34-DC by PolyI:C (cluster 4). IL12A was strongly and exclusively induced in bDC subsets by PolyI:C. Despite being based on a very stringent and constrained selection of genes bearing a heavy weight in the first three components of the PCA, this analysis highlighted striking similarities and differences between DC subset responses to PolyI:C. To extend this observation, PolyI:C-induced gene clusters harboring contrasting expression pattern across all conditions were identified through iterative self-organizing maps (Fig. 7D). This confirmed that DC subset responses to PolyI:C were characterized both by 1) the induction of many genes specifically in bDC subsets and XCR1+ CD34-DC (cluster 3, n = 388), with a higher induction of a number of these genes in XCR1+ bDC and/or XCR1+ CD34-DC as compared with CD1c+ bDC (Supplemental Fig. 5), and 2) the induction of many genes, in particular interferon stimulated genes, by PolyI:C in all the DC subsets studied as well as by LPS in XCR1− CD34-DC (cluster 5, n = 203). In addition to IL12A, several genes were induced exclusively by PolyI:C in bDC (cluster 4, n = 63), including IL23A and IL13. Thus, in addition to inducing interferon stimulated genes similar to MoDC and XCR1− CD34-DC, bDC mounted a strikingly specific and strong response to PolyI:C, which was very well modeled by XCR1+ CD34-DC.

FIGURE 7.
FIGURE 7.

Comparison of the responses to adjuvants of CD34-DC and bDC subsets. XCR1+ bDC and CD1c+ bDC were FACS sorted from the blood of adult healthy donors, stimulated in vitro under the same conditions described in Fig. 4 for CD34-DC subsets, and processed for microarrays. Data were analyzed together with those of CD34-DC subsets shown in Fig. 4. (A) PCA performed on the 2978 genes showing a fold change ≥ 2 between at least two of the conditions studied. (B) Hierarchical clustering performed on the same set of genes as in (A), using Pearson/Ward as distance metric/linkage parameters. (C) A heat map showing the expression patterns of the genes bearing the most weight on cell type distribution along the PC1, PC2, and PC3 axes of the PCA illustrated in (A). (D) A heat map showing the expression patterns of PolyI:C-induced genes harboring contrasting expression patterns across conditions. Individual expression patterns of a few selected PolyI:C-induced genes are shown on Supplemental Fig. 3 as bar graphs of mean ± SD of relative linear expression values for each combination of DC subset/stimulus to better show the differential expression between conditions.

PolyI:C-activated XCR1+ CD34-DC increased their ability to stimulate allogeneic CD4 T cells and have the highest capacity for cell-associated Ag cross-presentation

PolyI:C and R848 stimulations significantly increased the ability of XCR1+ CD34-DC to induce allogeneic CD4+ T cell proliferation, consistent with previous observations on human blood or skin XCR1+ DC (7, 10), although this was not the case with LPS (Fig. 8A). Under steady-state conditions, no significant differences were observed between DC subsets for their ability to cross-present a cell-associated Ag (Fig. 8B). PolyI:C stimulation increased this function in XCR1+ CD34-DC significantly above the levels observed at steady state and in XCR1− CD34-DC and MoDC (Fig. 8B, 8C). Although R848 activated all DC subsets to the same extent (Fig. 4), it induced cross-presentation to significantly higher levels only in XCR1+ CD34-DC (Fig. 8B, 8C).

FIGURE 8.
FIGURE 8.

Activation of T lymphocytes by resting and stimulated CD34-DC subsets and MoDC. (A) Allogeneic CD4+ T cell activation. FS3T* CD34-DC subsets and MoDC were cocultured for 6 d with CFSE-labeled allogeneic CD4+ T cells. The percentages of CD4+ T cells showing CFSE dilution are shown on the left graphs and the fold changes in CD4+ T cell numbers on the right graphs as a function of stimuli. Results are shown as mean ± SD for three to four independent MLR cultures for each DC subset with three replicate wells per culture. To determine significance, a Wilcoxon matched-pairs signed rank test was performed. (B and C) Cross-presentation of a cell-associated Ag by FS36 CD34-DC subsets and by MoDC derived from adult peripheral blood monocytes. (B) Each DC subset was cocultured 24 h with N9V/OVA-expressing K562 cells with or without R848 or PolyI:C (ratio DC:tumor = 1:1) and tested for induction of IFN-γ production by an N9V-specific T cell clone as assessed by intracellular staining. (C) N9V peptide–pulsed DC subsets were used as a positive control for the activation of the N9V-specific T cell clone and to ensure that all DC subsets were viable and had a similar efficacy for direct presentation of an optimal MHC-I–restricted epitope. No cross-presentation was observed in absence of cognate Ag when using K562 cells transfected with a vector expressing GFP only (data not shown). Results are shown as individual percentages of IFN-γ–expressing T cells for four to six independent experiments for each DC subset, with mean values indicated by black horizontal bars. To determine significance, a Wilcoxon matched-pairs signed rank test was performed. *p < 0.05, **p < 0.01.

Discussion

In this paper, we report an optimized protocol for in vitro generation of high numbers of human XCR1+ DC from CD34+ hematopoietic progenitor cultures. We characterized these XCR1+ CD34-DC, as well as the XCR1− CD34-DC representing the major cell population in the same cultures, through an unbiased, high-throughput comparison with XCR1+ bDC and with MoDC. These results unambiguously demonstrated that XCR1− CD34-DC resemble MoDC, whereas XCR1+ CD34-DC are a good model of bDC and LT-DC, both under steady-state as well as activation conditions.

XCR1+ CD34-DC expressed the molecular signature characteristic of XCR1+ bDC, including the chemokine receptor XCR1, which likely promotes physical interactions with CD8+ T cells (8, 13), the endocytic receptor CLEC9A, which promotes cross-presentation of Ag derived from dying cells (43), the adhesion receptor CADM1, which may promote the induction of CTL responses (44), and the pattern recognition receptor TLR3, which allows sensing of dsRNA including responsiveness to the synthetic adjuvant PolyI:C. CD1c and CD141 are not reliable markers of human DC subsets, because they were expressed on all the three in vitro–derived DC subsets studied in this work, consistent with similar observations reported for ex vivo–isolated human DC subsets (7, 11, 42). Hence, flow cytometry assessment of cell surface expression of the discriminative markers XCR1 or CLEC9A and CADM1, or gene expression profiling, are critical to ensure the identity of HLA-DR+CD141+ cells in humans (7, 8, 11, 36).

To examine the immunoactivation potential of XCR1+ CD34-DC, XCR1− CD34-DC, and MoDC, we selected for their stimulation three adjuvants widely used for basic research on DC maturation and/or for the design of novel immunotherapeutic treatments, LPS, PolyI:C, and R848. Ag targeting to XCR1+ DC combined with PolyI:C adjuvant is one of the best vaccination strategies to induce protective immune responses in mouse models of infectious diseases and cancer (45, 46). It is also very efficient for the induction of antiviral human CD4+ T cell responses in humanized mice in experimental settings where XCR1+ DC are a major source of IFN-I (47). The unique potency of PolyI:C in eliciting strong adaptive cellular immune responses also extends to nonhuman primate vaccination against SIV, papillomavirus, or Plasmodium falciparum (48–50). As compared with many other stimuli tested including LPS and R848, PolyI:C induces a unique combination of cytokines and chemokines in human whole blood assays (51), highlighting the particular immunoactivation potential of this adjuvant. However, the in vivo primary targets and specific modes of action of PolyI:C are still largely unknown, which hampers efficient translation to human of preclinical animal vaccination trials using this adjuvant. XCR1+ and CD1c+ bDC were strongly and uniquely activated by PolyI:C, but not by LPS, and conversely for MoDC. The differential responses of bDC versus MoDC to candidate adjuvants were nicely reproduced by the contrasting responses of in vitro–derived XCR1+ versus XCR1− CD34-DC. Hence, XCR1+ CD34-DC are a good in vitro model for XCR1+ bDC not only under steady-state conditions but also in terms of responses to adjuvants. The differences in the responses to adjuvants of bDC versus MoDC are largely cell intrinsic rather than being imprinted by their development in distinct environments through exposure to different growth factors and cytokines. Of note, however, regulation of a core set of hundreds of genes was observed in all optimally activated human DC subsets irrespective of stimuli, including NFKB or IFN-I target genes. These results extend to human XCR1+ DC, our previous observations on mouse DC subsets and human MoDC and pDC (52), and emphasizes the relevance of MoDC as a practical model to study the cell biology and molecular regulation underlying this conserved maturation subprogram.

XCR1+ DC and MoDC expressed strikingly different patterns of molecules involved in the orchestration of inflammatory responses and in cross-talk with NK or T cells. Resting XCR1− CD34-DC and MoDC expressed a much broader array of inflammatory molecules than XCR1+ CD34-DC and XCR1+ bDC, a difference that was further increased upon stimulation by LPS or R848. Resting XCR1+ CD34-DC and XCR1+ bDC expressed higher levels of genes critical for Ag processing and presentation, homing to lymph nodes or cross-talk with NK and T cells. In PolyI:C-activated XCR1+ CD34-DC, XCR1+ bDC, and eventually CD1c+ bDC, a specific secretion of IFN-λ was measured as well as the selective induction of genes involved in cross-talk with NK or T cells. Likewise to what was reported for XCR1+ bDC (8, 10), XCR1+ CD34-DC efficiently cross-presented a cell-associated Ag upon PolyI:C or R848 stimulation. This was not the case for XCR1- CD34-DC and MoDC. Hence, MoDC may be poised for inducing inflammation, whereas XCR1+ CD34-DC and XCR1+ bDC might be endowed with a stronger potential for the activation of NK cells and T cells. However, some differences were observed between XCR1+ bDC and XCR1+ CD34-DC. XCR1+ bDC expressed higher levels of lysosomal genes at steady state (Supplemental Table I) and, in response to PolyI:C, induced more strongly many cytokine genes including IFNLs, IFNB1, IL12A, and IL23A (Supplemental Fig. 5). In contrast, XCR1+ CD34-DC expressed higher levels of mitotic genes (Supplemental Table I). This suggests that XCR1+ CD34-DC might not be completely differentiated.

In summary, our results demonstrated that XCR1+ CD34-DC and XCR1+ bDC are similar in their gene expression programs and functions not only under steady-state conditions but also upon stimulation. In particular, they mount strikingly broad but specific responses to TLR3 stimulation and are specialized in cross-presentation of cell-associated Ags. In contrast, XCR1− CD34-DC resemble MoDC and are specialized in the modulation of inflammation. Our study demonstrated the feasibility of generating high numbers of bona fide XCR1+ DC in vitro from clinically relevant sources of CD34+ progenitors and demonstrated their unique responsiveness to the PolyI:C adjuvant, which represents a significant advance toward the potential use of human XCR1+ DC in the clinic.

Disclosures

E.F. and B.B. are inventors on a patent application on XCL1-targeted vaccibodies. I.P. and C.A.-P. are employees of Innate Pharma and are inventors on a patent application on anti-human TLR3 Abs. H.L., M.S., N.C., and T.S. are employees of SANOFI. The other authors have no financial conflicts of interest.

Acknowledgments

We thank Pierre Grenot, Marie-Laure Thibult, Marc Barad, Atika Zouine, and Isabelle Durand for technical support on cell sorting, Violaine Alunni and Christelle Thibault-Carpentier from the Plate-forme Biopuces (Strasbourg, France) for performing the microarray experiments (www.microarrays.ustrasbg.fr), and Lise-Marie Daufresne and Julie Bernard for management of the CB bank.

Footnotes

  • S.B., J.V.-G., and M.D. conceived the project and designed the experiments; S.B., V.O., R.C., F.M.-S., H.L., A.-C.D., C.S., J.S., N.C., M.S., and I.P. conducted experiments; B.B. and E.F. provided XCL1 vaccibodies; D.B. and C. Chabannon provided cord blood samples; S.B., V.O., R.C., F.M.-S., H.L., C.S., C. Caux, J.V.-G., and M.D. conducted data analyses; T.-P.V.M. and M.D. trained and supervised R.C. and F.M.-S. for bioinformatics analyses; S.B. and M.D. wrote the manuscript; all authors contributed to the editing of the manuscript; C. Caux, J.V.-G., and M.D. gathered funding; and J.V.-G. and M.D. supervised the project.

  • 1 J.V.-G. and M.D. are cosenior authors.

  • This work was supported by the I2HD CIML-SANOFI project, INSERM, Agence Nationale de la Recherche (EMICIF, ANR-08-MIEN-008-02; to M.D.), Agence Nationale de Recherches sur le SIDA et les Hépatites Virales (to M.D.), Institut National du Cancer Grant 2011-155 (to C. Caux, J.V.-G., and M.D.), Ligue contre la Cancer (Comité de la Savoie), Lyon Biopole (France), Association pour la Recherche sur le Cancer (France), Fondation pour la Recherche Médicale (Equipe Labellisée to M.D.), and the European Research Council under the European Community’s Seventh Framework Programme (FP7/2007-2013 Grant Agreement 281225; to M.D.). V.O. is a laureate from a LabEx DEVweCAN Ph.D. fellowship. S.B. was supported through the Agence Nationale de la Recherche (EMICIF, Grant ANR-08-MIEN-008-02). S.B., F.M.-S., R.C., and C.S. were supported by the I2HD CIML-SANOFI project. T.-P.V.M. was supported through the Agence Nationale de la Recherche (PhyloGenDC, Grant ANR-09-BLAN-0073-02), and the European Research Council (FP7/2007-2013 Grant Agreement 281225).

  • The microarray data presented in this article have been submitted to the Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE57671.

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    bDC
    blood DC
    CB
    cord blood
    DC
    dendritic cell
    EFS
    Etablissement Français du Sang
    FLT3-L
    FLT3-ligand
    GSEA
    GeneSet Enrichment Analysis
    LT-DC
    lymphoid tissue–resident DC
    MoDC
    monocyte-derived DC
    PCA
    principal component analysis
    pDC
    plasmacytoid DC
    PolyI:C
    polyinosinic-polycytidylic acid
    SCF
    stem cell factor
    TPO
    thrombopoietin.

  • Received May 15, 2014.
  • Accepted June 11, 2014.

This article is distributed under The American Association of Immunologists, Inc., Reuse Terms and Conditions for Author Choice articles.

References

    1. Adema G. J.,
    2. I. J. de Vries,
    3. C. J. Punt,
    4. C. G. Figdor
    . 2005. Migration of dendritic cell based cancer vaccines: in vivo veritas? Curr. Opin. Immunol. 17: 170–174.
    1. Plantinga M.,
    2. M. Guilliams,
    3. M. Vanheerswynghels,
    4. K. Deswarte,
    5. F. Branco-Madeira,
    6. W. Toussaint,
    7. L. Vanhoutte,
    8. K. Neyt,
    9. N. Killeen,
    10. B. Malissen,
    11. H. Hammad,
    12. B. N. Lambrecht
    . 2013. Conventional and monocyte-derived CD11b+ dendritic cells initiate and maintain T helper 2 cell-mediated immunity to house dust mite allergen. Immunity 38: 322‑335.
    1. Ziegler-Heitbrock L.,
    2. P. Ancuta,
    3. S. Crowe,
    4. M. Dalod,
    5. V. Grau,
    6. D. N. Hart,
    7. P. J. Leenen,
    8. Y. J. Liu,
    9. G. MacPherson,
    10. G. J. Randolph,
    11. et al
    . 2010. Nomenclature of monocytes and dendritic cells in blood. Blood 116: e74–e80.
    1. Robbins S. H.,
    2. T. Walzer,
    3. D. Dembélé,
    4. C. Thibault,
    5. A. Defays,
    6. G. Bessou,
    7. H. Xu,
    8. E. Vivier,
    9. M. Sellars,
    10. P. Pierre,
    11. et al
    . 2008. Novel insights into the relationships between dendritic cell subsets in human and mouse revealed by genome-wide expression profiling. Genome Biol. 9: R17.
    1. Crozat K.,
    2. R. Guiton,
    3. M. Guilliams,
    4. S. Henri,
    5. T. Baranek,
    6. I. Schwartz-Cornil,
    7. B. Malissen,
    8. M. Dalod
    . 2010. Comparative genomics as a tool to reveal functional equivalences between human and mouse dendritic cell subsets. Immunol. Rev. 234: 177–198.
    1. Watchmaker P. B.,
    2. K. Lahl,
    3. M. Lee,
    4. D. Baumjohann,
    5. J. Morton,
    6. S. J. Kim,
    7. R. Zeng,
    8. A. Dent,
    9. K. M. Ansel,
    10. B. Diamond,
    11. et al
    . 2014. Comparative transcriptional and functional profiling defines conserved programs of intestinal DC differentiation in humans and mice. Nat. Immunol. 15: 98–108.
    1. Haniffa M.,
    2. A. Shin,
    3. V. Bigley,
    4. N. McGovern,
    5. P. Teo,
    6. P. See,
    7. P. S. Wasan,
    8. X. N. Wang,
    9. F. Malinarich,
    10. B. Malleret,
    11. et al
    . 2012. Human tissues contain CD141hi cross-presenting dendritic cells with functional homology to mouse CD103+ nonlymphoid dendritic cells. Immunity 37: 60–73.
    1. Crozat K.,
    2. R. Guiton,
    3. V. Contreras,
    4. V. Feuillet,
    5. C. A. Dutertre,
    6. E. Ventre,
    7. T. P. Vu Manh,
    8. T. Baranek,
    9. A. K. Storset,
    10. J. Marvel,
    11. et al
    . 2010. The XC chemokine receptor 1 is a conserved selective marker of mammalian cells homologous to mouse CD8α+ dendritic cells. J. Exp. Med. 207: 1283–1292.
    1. Bachem A.,
    2. S. Güttler,
    3. E. Hartung,
    4. F. Ebstein,
    5. M. Schaefer,
    6. A. Tannert,
    7. A. Salama,
    8. K. Movassaghi,
    9. C. Opitz,
    10. H. W. Mages,
    11. et al
    . 2010. Superior antigen cross-presentation and XCR1 expression define human CD11c+CD141+ cells as homologues of mouse CD8+ dendritic cells. J. Exp. Med. 207: 1273–1281.
    1. Jongbloed S. L.,
    2. A. J. Kassianos,
    3. K. J. McDonald,
    4. G. J. Clark,
    5. X. Ju,
    6. C. E. Angel,
    7. C. J. Chen,
    8. P. R. Dunbar,
    9. R. B. Wadley,
    10. V. Jeet,
    11. et al
    . 2010. Human CD141+ (BDCA-3)+ dendritic cells (DCs) represent a unique myeloid DC subset that cross-presents necrotic cell antigens. J. Exp. Med. 207: 1247–1260.
    1. Poulin L. F.,
    2. M. Salio,
    3. E. Griessinger,
    4. F. Anjos-Afonso,
    5. L. Craciun,
    6. J. L. Chen,
    7. A. M. Keller,
    8. O. Joffre,
    9. S. Zelenay,
    10. E. Nye,
    11. et al
    . 2010. Characterization of human DNGR-1+BDCA3+ leukocytes as putative equivalents of mouse CD8α+ dendritic cells. J. Exp. Med. 207: 1261–1271.
    1. Lauterbach H.,
    2. B. Bathke,
    3. S. Gilles,
    4. C. Traidl-Hoffmann,
    5. C. A. Luber,
    6. G. Fejer,
    7. M. A. Freudenberg,
    8. G. M. Davey,
    9. D. Vremec,
    10. A. Kallies,
    11. et al
    . 2010. Mouse CD8α+ DCs and human BDCA3+ DCs are major producers of IFN-λ in response to poly IC. J. Exp. Med. 207: 2703–2717.
    1. Dorner B. G.,
    2. M. B. Dorner,
    3. X. Zhou,
    4. C. Opitz,
    5. A. Mora,
    6. S. Güttler,
    7. A. Hutloff,
    8. H. W. Mages,
    9. K. Ranke,
    10. M. Schaefer,
    11. et al
    . 2009. Selective expression of the chemokine receptor XCR1 on cross-presenting dendritic cells determines cooperation with CD8+ T cells. Immunity 31: 823–833.
    1. Crozat K.,
    2. S. Tamoutounour,
    3. T. P. Vu Manh,
    4. E. Fossum,
    5. H. Luche,
    6. L. Ardouin,
    7. M. Guilliams,
    8. H. Azukizawa,
    9. B. Bogen,
    10. B. Malissen,
    11. et al
    . 2011. Cutting edge: expression of XCR1 defines mouse lymphoid-tissue resident and migratory dendritic cells of the CD8α+ type. J. Immunol. 187: 4411–4415.
    1. Yoshio S.,
    2. T. Kanto,
    3. S. Kuroda,
    4. T. Matsubara,
    5. K. Higashitani,
    6. N. Kakita,
    7. H. Ishida,
    8. N. Hiramatsu,
    9. H. Nagano,
    10. M. Sugiyama,
    11. et al
    . 2013. Human blood dendritic cell antigen 3 (BDCA3)+ dendritic cells are a potent producer of interferon-λ in response to hepatitis C virus. Hepatology 57: 1705–1715.
    1. Crozat K.,
    2. E. Vivier,
    3. M. Dalod
    . 2009. Crosstalk between components of the innate immune system: promoting anti-microbial defenses and avoiding immunopathologies. Immunol. Rev. 227: 129–149.
    1. Zhang S.,
    2. K. Kodys,
    3. K. Li,
    4. G. Szabo
    . 2013. Human type 2 myeloid dendritic cells produce interferon-λ and amplify interferon-α in response to hepatitis C virus infection. Gastroenterology 144: 414‑425 e417.
    1. Segura E.,
    2. M. Durand,
    3. S. Amigorena
    . 2013. Similar antigen cross-presentation capacity and phagocytic functions in all freshly isolated human lymphoid organ-resident dendritic cells. J. Exp. Med. 210: 1035–1047.
    1. Cohn L.,
    2. B. Chatterjee,
    3. F. Esselborn,
    4. A. Smed-Sörensen,
    5. N. Nakamura,
    6. C. Chalouni,
    7. B. C. Lee,
    8. R. Vandlen,
    9. T. Keler,
    10. P. Lauer,
    11. et al
    . 2013. Antigen delivery to early endosomes eliminates the superiority of human blood BDCA3+ dendritic cells at cross presentation. J. Exp. Med. 210: 1049–1063.
    1. Flinsenberg T. W.,
    2. E. B. Compeer,
    3. D. Koning,
    4. M. Klein,
    5. F. J. Amelung,
    6. D. van Baarle,
    7. J. J. Boelens,
    8. M. Boes
    . 2012. Fcγ receptor antigen targeting potentiates cross-presentation by human blood and lymphoid tissue BDCA-3+ dendritic cells. Blood 120: 5163–5172.
    1. Schreibelt G.,
    2. L. J. Klinkenberg,
    3. L. J. Cruz,
    4. P. J. Tacken,
    5. J. Tel,
    6. M. Kreutz,
    7. G. J. Adema,
    8. G. D. Brown,
    9. C. G. Figdor,
    10. I. J. de Vries
    . 2012. The C-type lectin receptor CLEC9A mediates antigen uptake and (cross-)presentation by human blood BDCA3+ myeloid dendritic cells. Blood 119: 2284–2292.
    1. Mittag D.,
    2. A. I. Proietto,
    3. T. Loudovaris,
    4. S. I. Mannering,
    5. D. Vremec,
    6. K. Shortman,
    7. L. Wu,
    8. L. C. Harrison
    . 2011. Human dendritic cell subsets from spleen and blood are similar in phenotype and function but modified by donor health status. J. Immunol. 186: 6207–6217.
    1. Nizzoli G.,
    2. J. Krietsch,
    3. A. Weick,
    4. S. Steinfelder,
    5. F. Facciotti,
    6. P. Gruarin,
    7. A. Bianco,
    8. B. Steckel,
    9. M. Moro,
    10. M. Crosti,
    11. et al
    . 2013. Human CD1c+ dendritic cells secrete high levels of IL-12 and potently prime cytotoxic T-cell responses. Blood 122: 932–942.
    1. Gallois A.,
    2. N. Bhardwaj
    . 2010. A needle in the ‘cancer vaccine’ haystack. Nat. Med. 16: 854–856.
    1. Radford K. J.,
    2. I. Caminschi
    . 2013. New generation of dendritic cell vaccines. Hum. Vaccin. Immunother. 9: 259–264.
    1. Tacken P. J.,
    2. C. G. Figdor
    . 2011. Targeted antigen delivery and activation of dendritic cells in vivo: steps towards cost effective vaccines. Semin. Immunol. 23: 12–20.
    1. Sallusto F.,
    2. 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 α. J. Exp. Med. 179: 1109–1118.
    1. Caux C.,
    2. C. Dezutter-Dambuyant,
    3. D. Schmitt,
    4. J. Banchereau
    . 1992. GM-CSF and TNF-α cooperate in the generation of dendritic Langerhans cells. Nature 360: 258–261.
    1. Miller J. C.,
    2. B. D. Brown,
    3. T. Shay,
    4. E. L. Gautier,
    5. V. Jojic,
    6. A. Cohain,
    7. G. Pandey,
    8. M. Leboeuf,
    9. K. G. Elpek,
    10. J. Helft,
    11. et al
    . 2012. Deciphering the transcriptional network of the dendritic cell lineage. Nat. Immunol. 13: 888–899.
    1. Hoeffel G.,
    2. Y. Wang,
    3. M. Greter,
    4. P. See,
    5. P. Teo,
    6. B. Malleret,
    7. M. Leboeuf,
    8. D. Low,
    9. G. Oller,
    10. F. Almeida,
    11. et al
    . 2012. Adult Langerhans cells derive predominantly from embryonic fetal liver monocytes with a minor contribution of yolk sac-derived macrophages. J. Exp. Med. 209: 1167–1181.
    1. Schulz C.,
    2. E. Gomez Perdiguero,
    3. L. Chorro,
    4. H. Szabo-Rogers,
    5. N. Cagnard,
    6. K. Kierdorf,
    7. M. Prinz,
    8. B. Wu,
    9. S. E. Jacobsen,
    10. J. W. Pollard,
    11. et al
    . 2012. A lineage of myeloid cells independent of Myb and hematopoietic stem cells. Science 336: 86–90.
    1. Lundberg K.,
    2. A. S. Albrekt,
    3. I. Nelissen,
    4. S. Santegoets,
    5. T. D. de Gruijl,
    6. S. Gibbs,
    7. M. Lindstedt
    . 2013. Transcriptional profiling of human dendritic cell populations and models—unique profiles of in vitro dendritic cells and implications on functionality and applicability. PLoS One 8: e52875.
    1. Chen W.,
    2. S. Antonenko,
    3. J. M. Sederstrom,
    4. X. Liang,
    5. A. S. Chan,
    6. H. Kanzler,
    7. B. Blom,
    8. B. R. Blazar,
    9. Y. J. Liu
    . 2004. Thrombopoietin cooperates with FLT3-ligand in the generation of plasmacytoid dendritic cell precursors from human hematopoietic progenitors. Blood 103: 2547–2553.
    1. Proietto A. I.,
    2. D. Mittag,
    3. A. W. Roberts,
    4. N. Sprigg,
    5. L. Wu
    . 2012. The equivalents of human blood and spleen dendritic cell subtypes can be generated in vitro from human CD34+ stem cells in the presence of fms-like tyrosine kinase 3 ligand and thrombopoietin. Cell. Mol. Immunol. 9: 446–454.
    1. Balan S.,
    2. V. P. Kale,
    3. L. S. Limaye
    . 2010. A large number of mature and functional dendritic cells can be efficiently generated from umbilical cord blood-derived mononuclear cells by a simple two-step culture method. Transfusion 50: 2413–2423.
    1. Dutertre C.-A.,
    2. J.-P. Jourdain,
    3. M. Rancez,
    4. S. Amraoui,
    5. E. Fossum,
    6. B. Bogen,
    7. C. Sanchez,
    8. A. Couëdel-Courteille,
    9. Y. Richard,
    10. M. Dalod,
    11. et al
    . 2014. TLR-3 responsive, XCR1+, CD141(BDCA3)+/CD8α+-equivalent dendritic cells uncovered in healthy and SIV-infected rhesus macaques. J. Immunol. 192: 4697‑4708.
    1. Irizarry R. A.,
    2. B. Hobbs,
    3. F. Collin,
    4. Y. D. Beazer-Barclay,
    5. K. J. Antonellis,
    6. U. Scherf,
    7. T. P. Speed
    . 2003. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4: 249–264.
    1. Theocharidis A.,
    2. S. van Dongen,
    3. A. J. Enright,
    4. T. C. Freeman
    . 2009. Network visualization and analysis of gene expression data using BioLayout Express(3D). Nat. Protoc. 4: 1535–1550.
    1. Retière C.,
    2. V. Prod’homme,
    3. B. M. Imbert-Marcille,
    4. M. Bonneville,
    5. H. Vié,
    6. M. M. Hallet
    . 2000. Generation of cytomegalovirus-specific human T-lymphocyte clones by using autologous B-lymphoblastoid cells with stable expression of pp65 or IE1 proteins: a tool to study the fine specificity of the antiviral response. J. Virol. 74: 3948–3952.
    1. Segura E.,
    2. M. Touzot,
    3. A. Bohineust,
    4. A. Cappuccio,
    5. G. Chiocchia,
    6. A. Hosmalin,
    7. M. Dalod,
    8. V. Soumelis,
    9. S. Amigorena
    . 2013. Human inflammatory dendritic cells induce Th17 cell differentiation. Immunity 38: 336–348.
    1. Naik S. H.,
    2. A. I. Proietto,
    3. N. S. Wilson,
    4. A. Dakic,
    5. P. Schnorrer,
    6. M. Fuchsberger,
    7. M. H. Lahoud,
    8. M. O’Keeffe,
    9. Q. X. Shao,
    10. W. F. Chen,
    11. et al
    . 2005. Cutting edge: generation of splenic CD8+ and CD8‑ dendritic cell equivalents in Fms-like tyrosine kinase 3 ligand bone marrow cultures. J. Immunol. 174: 6592–6597.
    1. Chu C. C.,
    2. N. Ali,
    3. P. Karagiannis,
    4. P. Di Meglio,
    5. A. Skowera,
    6. L. Napolitano,
    7. G. Barinaga,
    8. K. Grys,
    9. E. Sharif-Paghaleh,
    10. S. N. Karagiannis,
    11. et al
    . 2012. Resident CD141 (BDCA3)+ dendritic cells in human skin produce IL-10 and induce regulatory T cells that suppress skin inflammation. J. Exp. Med. 209: 935–945.
    1. Sancho D.,
    2. O. P. Joffre,
    3. A. M. Keller,
    4. N. C. Rogers,
    5. D. Martínez,
    6. P. Hernanz-Falcón,
    7. I. Rosewell,
    8. C. Reis e Sousa
    . 2009. Identification of a dendritic cell receptor that couples sensing of necrosis to immunity. Nature 458: 899–903.
    1. Takeuchi A.,
    2. Y. Itoh,
    3. A. Takumi,
    4. C. Ishihara,
    5. N. Arase,
    6. T. Yokosuka,
    7. H. Koseki,
    8. S. Yamasaki,
    9. Y. Takai,
    10. J. Miyoshi,
    11. et al
    . 2009. CRTAM confers late-stage activation of CD8+ T cells to regulate retention within lymph node. J. Immunol. 183: 4220–4228.
    1. Wang B.,
    2. N. Zaidi,
    3. L. Z. He,
    4. L. Zhang,
    5. J. M. Kuroiwa,
    6. T. Keler,
    7. R. M. Steinman
    . 2012. Targeting of the non-mutated tumor antigen HER2/neu to mature dendritic cells induces an integrated immune response that protects against breast cancer in mice. Breast Cancer Res. 14: R39.
    1. Do Y.,
    2. A. M. Didierlaurent,
    3. S. Ryu,
    4. H. Koh,
    5. C. G. Park,
    6. S. Park,
    7. D. S. Perlin,
    8. B. S. Powell,
    9. R. M. Steinman
    . 2012. Induction of pulmonary mucosal immune responses with a protein vaccine targeted to the DEC-205/CD205 receptor. Vaccine 30: 6359–6367.
    1. Meixlsperger S.,
    2. C. S. Leung,
    3. P. C. Rämer,
    4. M. Pack,
    5. L. D. Vanoaica,
    6. G. Breton,
    7. S. Pascolo,
    8. A. M. Salazar,
    9. A. Dzionek,
    10. J. Schmitz,
    11. et al
    . 2013. CD141+ dendritic cells produce prominent amounts of IFN-α after dsRNA recognition and can be targeted via DEC-205 in humanized mice. Blood 121: 5034–5044.
    1. Park H.,
    2. L. Adamson,
    3. T. Ha,
    4. K. Mullen,
    5. S. I. Hagen,
    6. A. Nogueron,
    7. A. W. Sylwester,
    8. M. K. Axthelm,
    9. A. Legasse,
    10. M. Piatak Jr..,
    11. et al
    . 2013. Polyinosinic-polycytidylic acid is the most effective TLR adjuvant for SIV Gag protein-induced T cell responses in nonhuman primates. J. Immunol. 190: 4103–4115.
    1. Tewari K.,
    2. B. J. Flynn,
    3. S. B. Boscardin,
    4. K. Kastenmueller,
    5. A. M. Salazar,
    6. C. A. Anderson,
    7. V. Soundarapandian,
    8. A. Ahumada,
    9. T. Keler,
    10. S. L. Hoffman,
    11. et al
    . 2010. Poly(I:C) is an effective adjuvant for antibody and multi-functional CD4+ T cell responses to Plasmodium falciparum circumsporozoite protein (CSP) and αDEC-CSP in non human primates. Vaccine 28: 7256–7266.
    1. Stahl-Hennig C.,
    2. M. Eisenblätter,
    3. E. Jasny,
    4. T. Rzehak,
    5. K. Tenner-Racz,
    6. C. Trumpfheller,
    7. A. M. Salazar,
    8. K. Uberla,
    9. K. Nieto,
    10. J. Kleinschmidt,
    11. et al
    . 2009. Synthetic double-stranded RNAs are adjuvants for the induction of T helper 1 and humoral immune responses to human papillomavirus in rhesus macaques. PLoS Pathog. 5: e1000373.
    1. Duffy D.,
    2. V. Rouilly,
    3. V. Libri,
    4. M. Hasan,
    5. B. Beitz,
    6. M. David,
    7. A. Urrutia,
    8. A. Bisiaux,
    9. S. T. Labrie,
    10. A. Dubois,
    11. et al
    . 2014. Functional analysis via standardized whole-blood stimulation systems defines the boundaries of a healthy immune response to complex stimuli. Immunity 40: 436–450.
    1. Manh T. P.,
    2. Y. Alexandre,
    3. T. Baranek,
    4. K. Crozat,
    5. M. Dalod
    . 2013. Plasmacytoid, conventional, and monocyte-derived dendritic cells undergo a profound and convergent genetic reprogramming during their maturation. Eur. J. Immunol. 43: 1706–1715.