TLR-Stimulated CD34 Stem Cell-Derived Human Skin-Like and Monocyte-Derived Dendritic Cells Fail to Induce Th17 Polarization of Naive T Cells but Do This information is current as Stimulate Th1 and Th17 Memory Responses of September 24, 2021. Sai Suda Duraisingham, Julia Hornig, Frances Gotch and Steven Patterson J Immunol 2009; 183:2242-2251; Prepublished online 22 July 2009; Downloaded from doi: 10.4049/jimmunol.0900474 http://www.jimmunol.org/content/183/4/2242 http://www.jimmunol.org/ Supplementary http://www.jimmunol.org/content/suppl/2009/07/23/jimmunol.090047 Material 4.DC1 References This article cites 40 articles, 14 of which you can access for free at: http://www.jimmunol.org/content/183/4/2242.full#ref-list-1

Why The JI? Submit online. by guest on September 24, 2021

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2009 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

TLR-Stimulated CD34 Stem Cell-Derived Human Skin-Like and Monocyte-Derived Dendritic Cells Fail to Induce Th17 Polarization of Naive T Cells but Do Stimulate Th1 and Th17 Memory Responses1

Sai Suda Duraisingham, Julia Hornig, Frances Gotch, and Steven Patterson2

Dendritic cells (DCs) are important in linking innate and adaptive immune responses by priming and polarizing naive CD4؉ Th cells, but little is known about the effect of different human DC subsets on Th cells, particularly Th17 cells. We have investigated the ability of TLR-stimulated human Langerhans cells (LC), dermal DCs (dDC), and monocyte-derived DCs (moDC) to affect naive and memory Th17 and Th1 responses. MoDCs stimulated greater memory proliferation while LCs and dDCs more Downloaded from potently stimulated proliferation, indicating functionally distinct subsets of DCs. TLR stimulation of all three DC types was unable to induce Th17 polarization from naive T cell precursors, despite inducing Th1 polarization. Dectin stimulation of DCs in IMDM was however able to produce Th17 cells. TLR-stimulated DCs were capable of inducing IL-17A and IFN-␥ production from memory T cells, although the mechanism used by each DC subset differed. MoDCs partially mediated this effect on memory Th1 and Th17 cells by the production of soluble factors, which correlated with their ability to secrete IL-12p70 and IL-23. In contrast, LCs and dDCs were able to elicit a similar memory response to moDCs, but in a contact dependent manner. Additionally, http://www.jimmunol.org/ the influence of microbial stimulation was demonstrated with TLR3 and TLR7/8 agonists inducing a Th1 response, whereas TLR2 or dectin stimulation of moDCs enhanced the IL-17 response. This study emphasizes the differences between human DC subsets and demonstrates that both the DC subset and the microbial stimulus influence the Th cell response. The Journal of Immunology, 2009, 183: 2242–2251.

he CD4ϩ Th cell response comprises several subsets of at later stages on IL-23 (4). In humans, the polarizing cells characterized by the secretion of different cytokines, that are necessary have not been definitively characterized, with including the classical IFN-␥-producing Th1 cells and the studies variably citing the importance of IL-1, IL-6 (5, 6), and

T by guest on September 24, 2021 IL-4-producing Th2 cells, which are involved in cellular or hu- TGF-␤ (7–9). Additionally, IL-23 and IL-1␤ have been shown to moral immunity, respectively (1). Recently described IL-17-pro- induce the production of IL-17A from human memory T cells (10, ducing Th17 cells have been implicated in the immune response to 11). Alternatively, DC production of the IL-12p70 , which extracellular bacteria and fungi as well as in several autoimmune shares a common subunit with IL-23, is involved in the polariza- diseases (2). Dendritic cells (DCs)3 play a key role in linking the tion of Th1 cells and the induction of IFN-␥ from memory cells innate and adaptive immune systems by priming naive T cells that (12). Many of the studies regarding Th17 cell development have can become polarized toward a particular Th cell phenotype. The been conducted using APC-free cultures with anti-CD3/CD28 stimuli, triggering of TLRs by pathogenic molecules results in the up-reg- exogenous cytokines, and IFN-␥/IL-4 neutralization. However, little ulation of MHC and costimulatory molecules and the secretion of is known about the physiological conditions needed for human Th17 cytokines by DCs, which can in turn skew effector Th cell differ- development and maintenance, particularly the role of DCs. The vari- entiation (3). able production of key cytokines and other factors by DCs after TLR In mice, the polarization of Th17 cells from naive T cells is stimulation may therefore influence the differentiation and reactiva- initially dependent on the presence of TGF-␤, IL-21, and IL-6, and tion of Th17 and Th1 cells. In humans, several distinct DC populations exist that can be broadly categorized into myeloid DC and plasmacytoid DC types Department of Immunology, Faculty of Medicine, Imperial College, Chelsea and Westminster Hospital, London, United Kingdom (13). In human skin, there are two major myeloid DC subsets: epidermal Langerhans cells (LCs), which contain Birbeck granules Received for publication February 11, 2009. Accepted for publication June 6, 2009. and are langerinϩ, and the dermal DCs (dDCs), which express 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 DC-specific intracellular adhesion molecule 3-grabbing noninte- with 18 U.S.C. Section 1734 solely to indicate this fact. grin (DC-SIGN) and CD11b (14). DC subsets have been reported 1 This work was supported by United Kingdom Medical Research Council and Eu- to express different arrays of TLR molecules, which may hint at a roprise Consortium. role for functional specialization in the response to pathogens (15, 2 Address correspondence and reprint requests to Dr. Steven Patterson, Department of 16). Hence, the triggering of TLRs on each DC population may Immunology, Faculty of Medicine, Imperial College, Chelsea and Westminster Hos- pital, 369 Fulham Road, London SW10 9NH, United Kingdom. E-mail address: produce maturation signals that have differential effects on the re- [email protected] sultant Th cell response. The study of human DC biology has been 3 Abbreviations used in this paper: DC, ; LC, Langerhans cell; dDC, greatly assisted by the generation of DCs in vitro; the most com- dermal DC; moDC, monocyte-derived DC; DC-SIGN, DC-specific intracellular ad- monly used model of human DCs is monocyte-derived DCs hesion molecule 3-grabbing nonintegrin. (moDCs) generated from blood monocytes (17). For LCs and Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00 dDCs, phenotypic equivalents can be generated in vitro from www.jimmunol.org/cgi/doi/10.4049/jimmunol.0900474 The Journal of Immunology 2243

CD34ϩ hematopoetic stem cells (18). Understanding how human DC stimulation and cytokine production DCs respond to innate immune stimuli and the effect that this has The following TLR agonists from Invivogen were used: 1 ␮g/ml on the adaptive immune response is important in generating vac- Pam3Csk4 (TLR1/2), 1 ␮g/ml P. gingivalis LPS (TLR2), 25 ␮g/ml polyI:C cine adjuvants that can more efficiently induce an appropriate re- (TLR3), 1 ␮g/ml E. coli LPS (TLR4), 5 ␮g/ml CL097 (TLR7/8), 0.1 ␮g/ml sponse. The study of human skin DCs may be particularly impor- zymosan (TLR2/dectin-1), and 1 ␮g/ml curdlan (dectin-1) (Wako). CD40L ␮ tant due to the interest surrounding transcutaneous vaccination (5 g/ml) was also added for some experiments (R&D Systems). DCs (106/ml) were stimulated for 24 h with the TLR agonists and the superna- strategies (19). tants were assayed for IL-12p40, IL-12p70, and IL-23 by ELISA. IL-12p70 In this study, we have examined the ability of TLR-stimulated and IL-23 were assayed using available ELISA kits (eBiosciences). For LC, dDC, and moDC subsets to induce cytokine production from IL-12/23p40, goat anti-mouse IgG precoated ELISA plates (Cayman naive and memory Th cells. Human DCs stimulated through their Chemicals) were coated with capture Ab (1 ␮g/ml, clone 31052, R&D Systems) for 1 h. Supernatants and standards were added for 2 h at room TLRs were found to be unable to polarize Th17 cells from naive T temperature. The plates were washed and biotinylated-detection Ab was cells but were able to induce Th1 and Th2 differentiation. Dectin- added (0.5 ␮g/ml, R&D Systems) for 1 h. The plates were developed with stimulated DCs were however able to polarize naive Th17 cells. HRP avidin enzyme (1 ␮g/ml, Vector Laboratories), 3,3Ј,5,5Ј-tetrameth- All DC types were also capable of inducing IL-17A production ylbenzidine substrate (Becton Dickinson), and stop solution (R&D Sys- from memory T cells. The production of the key cytokines IL- tems). Plates were read on an Anthos ELISA plate reader (ASYS Hitech) at 450 nm. 12p70 and IL-23 by moDCs in response to TLR or dectin stimu- lation was associated with the corresponding Th1 or Th17 memory Mixed reactions T cell response. In contrast, LCs and dDCs appeared to use an DCs (104/well) in complete RPMI 1640 with 10% human AB serum IL-12/IL-23-independent mechanism to elicit similar responses. (Sigma-Aldrich) were stimulated overnight with TLR or dectin agonists Downloaded from This study highlights the differences between DC subsets and the before culturing with different allogeneic T cell populations (105/well) for importance of both the DC type and microbial stimulus in deter- 5 days (proliferation assays) or 7 days (cytokine staining). In one experi- ment, complete IMDM supplemented with 10% AB or serum-free StemX- mining Th cell responses. Vivo Dendritic cell base medium (R&D Systems) was used for DC-T co- cultures. T cells alone were used as control wells. Naive CD4ϩ T cells were isolated from umbilical cord blood using the naive T cell isolation kit Materials and Methods ϩ (Miltenyi Biotec) whereas memory CD4 T cells were isolated from adult http://www.jimmunol.org/ Generation of DC subsets peripheral blood (Miltenyi Biotec), T cells were Ͼ98% CD45ROϪ or CD45RAϪ, respectively. Where total were used, PBMCs Umbilical cord blood samples were obtained from consented full-term were separated over a 50% Percoll gradient (Sigma-Aldrich) to yield en- pregnant women in accordance with the local Ethical Approval Committee. Ͼ ϩ ϩ riched lymphocytes which were 75% CD3 . CD34 cells were isolated from cord blood by immunomagnetic beads For proliferation assays, T cells were cultured with graded numbers of (Miltenyi Biotec) and cultured in complete RPMI 1640 (100 IU/ml peni- DCs in triplicate. Cultures were pulsed on day 5 with 1␮Ci/well [3H]thy- cillin, 0.1 mg/ml streptomycin, 2 mM L-glutamine; Sigma-Aldrich) sup- midine (PerkinElmer) for 16 h before harvesting and measurement of the plemented with 10% FCS (Invitrogen), 100 ng/ml GM-CSF (Immunex), 25 cpm. For intracellular cytokine staining, cocultures were restimulated on ng/ml stem cell factor, 25 ng/ml Flt3 ligand, and 5 ng/ml TNF-␣ (R&D ␮ ␮ ϩ day 7 with PMA (0.1 g/ml) and ionomycin (1 M) for 5 h, with Brefeldin Systems). After 6 days, the CD14 cells were isolated with immunomag- A (10 ␮g/ml; Sigma-Aldrich) added for the last 4 h. Cells were surface netic beads and cultured for a further 7 days with 100 ng/ml GM-CSF and stained with CD3-APCCy7 and CD4-PECy7 Abs (Becton Dickinson) be- by guest on September 24, 2021 1000 U/ml IL-4 (R&D Systems) to generate DC-SIGNϩ CD11bϩ dermal- Ϫ fore fixing with 2% paraformaldehyde. Cells were incubated in 0.1% sa- type DCs. The remaining CD14 cells were cultured for 7 days with 100 ponin buffer with IFN-␥-FITC, IL-17A-PE, and IL-4-allophycocyanin or ␤ ϩ ng/ml GM-CSF and 5 ng/ml TGF- (R&D Systems) to generate langerin IL-10-APC Abs (Becton Dickinson) and acquired on an LSRII. Langerhans-type cells. Both DC types were finally isolated by MACS based on CD1a expression with a purity of Ͼ95%. For quantitative-PCR T cell cultures experiments, cells were passed over two MACS columns to yield a purity 6 of 99%. Monocyte-derived DCs were generated from CD14ϩ peripheral Cord blood naive or adult blood memory T cells (10 /ml) in complete blood monocytes isolated with immunomagnetic beads (Miltenyi Biotec) RPMI 1640 medium with 10% human AB serum were stimulated with and cultured with 100 ng/ml GM-CSF and 1000 U/ml IL-4 for 5–6 days. anti-CD2/CD3/CD28 T cell activation beads (Miltenyi Biotec) with a bead ␤ Human skin samples were obtained from healthy individuals under- to cell ratio of 1:2. The IL-17-inducing cytokines IL-1 , IL-6, IL-23 (all at ␤ going abdominoplasty surgery. Specimens were cut into 25 mm2 pieces 10 ng/ml; eBiosciences), and TGF- (5 ng/ml; R&D Systems) were added and incubated overnight at 4°C in 0.2% dispase (Roche Applied Sci- to half of the wells. Control wells were cultured without these cytokines. ences). The next day, the epidermis was carefully removed, rinsed with On days 3 and 6, IL-2 (20 U/ml, R&D Systems) was added to all wells and ␤ PBS, and incubated in complete RPMI 1640 medium at 37°C for 48 h. IL-1 , IL-6, and IL-23 (10 ng/ml) were added to the cytokine wells. On LCs were purified from the emigrated cell population using CD1a im- day 7, cultures were restimulated and intracellular cytokine staining was munomagnetic beads (Miltenyi Biotec). LCs isolated in this manner are conducted. 6 phenotypically mature, in that they are CD83ϩ, CD86high, CD40high, Cord blood naive or adult blood memory T cells (10 /ml) with T cell 5 CCR7ϩ MHCIIhigh (20). activation beads were added to supernatants taken from DCs (10 /ml) stim- ulated for 24 h with TLR or dectin agonists. On days 3 and 6 of culture, T cell cultures were gently pipetted, before being harvested and assessed for RNA isolation and quantitative PCR intracellular cytokines on day 7. RNA was isolated and cDNA synthesized from purified DC populations Statistical analysis using the One-step cDNA kit according to manufacturer’s instructions (Miltenyi Biotec). Quantitative real-time PCR for TLR expression were Data were analyzed using GraphPad Prism 5.0 software by ANOVA with prepared using SYBR Green I Mastermix (Roche Applied Systems) with Bonferroni’s post test for grouped comparisons and Mann-Whitney U test the forward and reverse primers in supplemental Table I4 (Operon Bio- for comparison within groups. A p-value of Ͻ0.05 was considered

technologies) and additional 1 mM MgCl2 (Promega) for TLRs 1 and 5. significant. The PCR were denatured for 8 min at 95°C and then amplified over 40 cycles of 95°C for 12 s, 56°C or 59°C for 8 s and 72°C for 10 s for data Results acquisition. Samples were run on a LightCycler 480 and analysis was con- Human DC subsets express different TLRs and have differential ducted using Advanced Relative Quantification software, where TLR rel- abilities to induce T cell proliferation ative expression was calculated by normalization to ␤-actin expression (Roche Applied Systems). Human Langerhans-type and dermal-type DCs were generated from CD34ϩ stem cells and their TLR expression levels were com- pared with conventional monocyte-derived DCs by quantitative 4 The online version of this article contains supplemental material. real-time PCR. All three DC types expressed TLRs 1, 2, 4, 5, and 2244 T CELL POLARIZATION INDUCED BY DCs Downloaded from http://www.jimmunol.org/

FIGURE 1. DC subsets have different TLR expression levels and different abilities to induce T cell proliferation. A, cDNA from in vitro-derived LCs, dDCs, and moDCs was assessed for TLR1–10 expression normalized to ␤-actin expression. Data shown are the mean Ϯ SEM of five to six samples. B, Proliferation of allogeneic naive (filled symbols) and memory (open symbols) CD4ϩ T cells stimulated with graded ratios of unstimulated in vitro-derived LCs, dDCs, or moDCs as measured by [3H]thymidine incorporation. Data shown are the mean Ϯ SEM of six to eight independent experiments. C, FACS by guest on September 24, 2021 plots showing up-regulation of DC maturation markers upon stimulation of in vitro-derived LCs with CL097. Up-regulation of these molecules was observed on LC, dDC, and moDC stimulated with other agonists for which the corresponding TLR was expressed (data not shown). Data shown are representative of six independent experiments for each DC type with each TLR agonist.

6 at varying levels (Fig. 1A). LCs expressed higher levels of TLR1, CD4ϩ T cell cytokine production was assessed. LCs, and to a TLR7, and TLR10 mRNA compared with dDCs and moDCs, but lesser extent dDCs, induced more IL-17Aϩ cells than moDCs re- they did not express TLR3 or TLR8. dDCs expressed TLR8 but gardless of the type of TLR stimulus (Fig. 2, A and D). LCs stim- not TLR7, whereas moDCs expressed both TLR7 and TLR8. Stim- ulated with Pam3Csk (TLR1/2), TLR4-LPS, and CL097 (TLR7/8) ulation of the DCs with agonists targeting each TLR resulted in the produced a significantly greater IL-17 response than moDCs. All up-regulation of HLA-DR, CD80, CD83 and CD86 (Fig. 1C)in three DCs induced strong Th1 IFN-␥ϩ responses, which were in- agreement with the TLR expression data for all DC types and creased after stimulation with polyIC (TLR3), TLR4-LPS or agonists tested. For example, LCs that did not express TLR3 did CL097 (TLR7/8) (Fig. 2B). Notably, the IFN-␥ response induced not mature after treatment with the TLR3 ligand polyIC, but did by LCs was increased after polyIC stimulation despite the lack of have a greater maturational response to the TLR1/2 agonist TLR3 expression on LCs and may be due to signaling through Pam3Csk compared with dDCs or moDCs (data not shown). cytoplasmic MDA5 that recognizes double-stranded RNA (21). To compare the intrinsic ability of each DC type to prime T Alternatively, there may be up-regulation of TLR3 on the LCs cells, unstimulated DCs were cultured in MLRs with naive or during the MLR culture or a role for other TLR3-expressing cell ϩ memory CD4 T cells. LCs and dDCs induced greater prolifera- types in the total lymphocyte population. A minor population of tion of naive T cells whereas moDCs were more efficient at in- IL-17AϩIFN-␥ϩ CD4ϩ T cells were consistently seen in all cul- ducing memory T cell expansion (Fig. 1B). There were, however, tures (Fig. 2D) as reported previously (6, 22). LCs were also able no apparent differences between the LCs and dDCs. Thus, each to induce a greater IL-4ϩ Th2 response than dDCs and moDCs, human DC type had a different TLR expression profile and differ- which was statistically significant with TLR4-LPS stimulation ential ability to induce the proliferation of naive or memory T (Fig. 2C). Hence, in the presence of a total lymphocyte population, cells, suggesting that there may be functional specialization of the LCs elicited a mixed Th cell cytokine profile but with significantly DC subsets. greater Th17 and Th2 responses, whereas dDCs and moDCs pref- erentially induced a Th1 response. The nature of the TLR stimulus DC subsets stimulated through their TLRs induce different also contributed to the type of response, with TLR3, TLR4, or T cell cytokine production profiles TLR7/8 stimulation augmenting the IFN-␥ response even in LCs. DCs stimulated with TLR agonists were cultured with total lym- Considering we had previously used total lymphocyte popula- phocytes enriched from peripheral blood (Ͼ75% CD3ϩ) and then tions, we next investigated whether the DCs were exerting their The Journal of Immunology 2245 Downloaded from http://www.jimmunol.org/

FIGURE 2. LCs are able to induce greater IL-17A and IL-4 responses from total lymphocyte populations whereas dDCs and moDCs preferentially induce an IFN-␥ response. A–C, TLR-stimulated in vitro-derived DC subtypes were cultured with allogeneic total lymphocytes (Ͼ75% CD3ϩ) for 7 days before restimulation and intracellular cytokine staining. D, Representative FACS plots of IFN-␥, IL-17A, and IL-4 staining. The percentage positive cells by guest on September 24, 2021 p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء .are shown in each quadrant (gated on the CD3ϩCD4ϩ cells). Data are the mean Ϯ SEM of five to six independent experiments by two-way ANOVA with Bonferroni’s post test (LCs vs moDCs). effect on the differentiation of naive T cells or memory T cells. The Given the apparent inability of in vitro-derived DCs to fully ability of each DC type to polarize naive T cells purified from polarize Th17 cells, we examined whether LCs isolated from nor- umbilical cord blood was compared. Cord blood was used to en- mal human skin could mediate this effect. Skin epidermal LCs sure a genuinely naive population, as IL-17Aϩ cells may be pro- were similarly unable to induce IL-17A secretion from naive T duced by the reactivation of contaminating memory T cells (23). cells, but could still polarize Th1 responses which were amplified Unexpectedly, none of the three DC types were able to induce the upon TLR stimulation (Fig. 5A). Skin LCs could however generate differentiation of Th17 cells, even after TLR1/2, TLR2, TLR3, strong IL-17A, IFN-␥ (Fig. 5B), and IL-4 (data not shown) re- TLR4, or TLR7/8 stimulation (Fig. 3A and data not shown). The sponses from memory CD4ϩ T cells. IL-17A production was in- DCs were however still able to efficiently polarize Th1 responses, creased with TLR stimulation of the LCs, particularly Pam3Csk particularly dDCs and moDCs stimulated with polyIC (TLR3) or (TLR1/2), TLR4-LPS, and CL097 (TLR7/8), but TLR stimulation CL097 (TLR7/8) (Fig. 3B). LCs induced lower levels of Th1 po- had only a minimal effect on IFN-␥ secretion. The percentages of larization, which could not be further augmented by TLR stimu- cytokine-secreting Th cells induced by the skin-derived LCs were lation. All DC types also generated small percentages of Th2 and comparable to that of the in vitro-derived LCs, demonstrating sim- IL-10ϩ T cells which were not significantly altered by TLR stim- ilarity between the in vitro model and ex vivo tissue DCs. ulation (Fig. 3, C and D, Fig. S1). Our inability to induce a Th17 response in naive T cells by stimulating with different DC types does not reflect an inherent incapacity of the naive T cells to pro- Polarization of Th17 cells from naive T cells duce IL-17 as demonstrated in later experiments (see Fig. 6). Because we were unable to induce Th17 polarization with any DCs Despite the poor ability of the DCs to polarize Th17 cells from tested, we wanted to confirm the ability of cord blood naive T cells naive precursors, all DC types could effect the secretion of IL-17A to produce IL-17A. Naive T cells were stimulated with anti-CD2/ from memory CD4ϩ T cells, which was significantly increased CD3/CD28 beads and cultured in the presence of IL-1␤, IL-6, upon stimulation with TLR2 or zymosan of the moDCs (Fig. 4A). IL-23, and TGF-␤ for 7 days. This combination of cytokines re- All DCs also induced strong IFN-␥ and moderate IL-4 memory sulted in increased percentages of IL-17Aϩ cells from both naive responses that could not be further amplified by TLR stimulation and memory T cells with no observable effect on IFN-␥ secretion (Fig. 4, B and C). The moDCs induced a greater percentage of (Fig. 6A), suggesting that the TLR-stimulated DCs may not be memory IL-10ϩ, which did not coproduce IFN-␥ or IL-17A (data producing all of these factors or in sufficient quantities for naive not shown), than the LCs or dDCs (Fig. 4D, Fig. S1). Th17 polarization. These experiments were performed in RPMI 2246 T CELL POLARIZATION INDUCED BY DCs Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021 FIGURE 3. DCs can polarise Th1, Th2, and IL-10ϩ cells but not Th17 cells from naive T cells. A–D, TLR-stimulated in vitro-derived DC types were cultured with allogeneic cord blood naive T cells for 7 days before re-stimulation and intracellular cytokine staining (N.D. indicates not determined). Data p Ͻ 0.001 (compared with unstimulated DC-T cells) by one-way ANOVA ,ءءء ;p Ͻ 0.05 ,ء .are the mean Ϯ SEM of five to eight independent experiments with Bonferroni’s post test.

1640 medium containing AB serum but similar results were ob- TLR-treated DCs were capable of inducing IL-17A secretion from tained using StemX-vivo medium (not shown). Because DCs stim- memory but not naive T cells, the mechanisms by which DCs mediate ulated with TLR agonists were unable to induce Th17 cells, we their effect on memory T cells may differ to naive T cells. tested whether the dectin-1 ligands zymosan or curdlan were able to polarise Th17 cells, as previously reported (24). In our culture MoDCs but not LCs or dDCs partially mediate memory T cell conditions, dectin-stimulated moDCs were unable to induce Th17 cytokine production via soluble factors cells (Fig. 6B). The presence of serum in culture medium has been To ascertain whether DCs were inducing differential memory T reported to inhibit Th17 differentiation (7), however we were still cell cytokine production via soluble factors, memory T cells were unable to induce Th17 cells with TLR-treated DCs using serum- cultured with anti-CD2/CD3/CD28 beads and supernatants taken free X-Vivo medium (Fig. 6B). A recent report demonstrated that from DCs stimulated with TLR-2, TLR-3, TL7/8 or dectin ligands. the use of IMDM enhanced Th17 differentiation compared with Supernatants from LCs and dDCs had no significant amplifying RPMI 1640 medium (25). Similar to the data with RPMI 1640 effect on memory T cell cytokine production, above that induced medium, TLR-treated DCs were unable to polarize Th17 cells even by the activation beads alone (Fig. 7). In many cases the addition in the presence of IMDM. However, stimulation of the skin LCs of LC or dDC supernatants led to a decrease in the percentage of and in vitro-derived LCs and dDCs with the dectin ligands zymo- cytokine-producing memory T cells. In contrast, supernatants from san or curdlan was able to induce 0.3–0.4% Th17 cells in IMDM zymosan-stimulated moDCs increased the percentage of IL-17A but not RPMI 1640 medium (Fig. 6C). Interestingly, the moDCs memory T cells and supernatants from moDCs stimulated with induced the lowest percentage of Th17 cells in IMDM compared polyIC (TLR3) and CL097 (TLR7/8) marginally increased IFN-␥ with the LCs and dDCs. The differences in responses induced by (Fig. 7). DC supernatants were unable to induce any cytokine pro- the skin LCs and the in vitro-derived LCs were more apparent in duction from naive T cells, with the exception of moDCs stimu- IMDM, compared with our previous experiments in RPMI 1640. Fur- lated with polyIC (TLR3), which could produce robust Th1 polar- thermore, the addition of exogenous IL-1␤, IL-6, IL-23, and TGF-␤ ization (data not shown). Thus despite all three DC types having a cytokines to the unstimulated DC-T cell cultures was able to restore similar ability to induce memory T cell IL-17A, IFN-␥, and IL-4 Th17 polarization of naive T cells (Fig. 6C), suggesting that these production, the mechanism by which this is mediated differs be- cytokines are not induced by TLR agonists. However, because the tween the DCs. MoDCs but not LCs or dDCs appear to produce a The Journal of Immunology 2247 Downloaded from http://www.jimmunol.org/

FIGURE 4. DCs stimulated with TLRs are able to induce the production of IL-17A, IFN-␥, IL-4, and IL-10 from memory T cells. A–D, TLR- or

dectin-stimulated in vitro-derived DC types were cultured with allogeneic memory T cells for 7 days before restimulation and intracellular cytokine staining by guest on September 24, 2021 p Ͻ ,ء .N.D. indicates not determined). Data are mean Ϯ SEM of seven to eleven (LCs) or five to seven (dDCs and moDCs) independent experiments) 0.05 (compared with unstimulated DC-T cells) by one-way ANOVA with Bonferroni’s post test. soluble factor(s) which contribute to memory T cell cytokine Th17-inducing IL-23, and Th1-inducing IL-12p70 cytokines. production. CD40L costimulation was used for IL-12p70 and IL-23 to maxi- To explain this observation further, the TLR-stimulated DC su- mize DC cytokine secretion (26) and may be more representative pernatants were assayed for the IL-12/23p40 common subunit, of the conditions in DC-T cell cocultures. LCs and moDCs but not

FIGURE 5. Human skin LCs are unable to polarize Th17 cells from naive T cells, but can efficiently induce IL-17A secretion from memory T cells. Human epidermal LCs (isolated from cells that have migrated out over 48 h) were stimulated with TLR agonists and cultured with allogeneic naive (A) or memory T cells (B) for 7 days before restimulation and intracellular cytokine staining. The percentage positive cells are shown in each quadrant (gated on the CD3ϩCD4ϩ cells). Results are representative of three independent experiments. 2248 T CELL POLARIZATION INDUCED BY DCs Downloaded from

␤ ␤ FIGURE 6. DCs cultured with IL-1 , IL-23, IL-6, and TGF- or stimulated through dectin-1, but not TLRs, can induce Th17 differentiation in the http://www.jimmunol.org/ presence of IMDM. A, T cells were stimulated with anti-CD3/CD28/CD2 beads in medium alone (open symbols) or in the presence (filled symbols) of IL-1␤, IL-23, IL-6, and TGF-␤ for 7 days before restimulation and intracellular cytokine staining. Data of four independent experiments were analyzed ,p Ͻ 0.05. B, MoDCs were stimulated with TLR agonists in serum-free X-Vivo medium or with the dectin-1 ligands ,ء ;using the Mann-Whitney U test zymosan, and curdlan (in RPMI 1640 medium), before culturing with allogeneic naive T cells for 7 days. Data shown are representative of three independent experiments. C, TLR, dectin, or cytokine (IL-1␤, IL-23, IL-6, and TGF-␤) stimulated in vitro-derived DCs were cultured with naive T cells for 7 days in p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء .IMDM. Data shown are the mean Ϯ SEM of six to seven (LCs, DCs, moDCs) or three (skin LCs) independent experiments analyzed by the Mann-Whitney U test compared with unstimulated DCs. by guest on September 24, 2021 dDCs secreted large quantities of IL-12/23p40, even in the absence Discussion of CD40L costimulation, particularly after TLR2-LPS and CL097 DCs are important in driving T cell responses, but less is known about (TLR7/8) stimulation (Fig. 8A). MoDCs produced IL-12p70 after the influence of DC subsets on Th cell polarization, particularly Th17 stimulation with polyIC (TLR3) or CL097 (TLR7/8) and dDCs cells. Distinct human DC subsets exist including LCs, dDCs, and after stimulation with CL097 (TLR7/8) (Fig. 8B), although there moDCs, which were found to have different TLR expression patterns was donor variation in IL-12p70 secretion. The LCs were unable and thus have the potential to respond to different pathogen molecule to produce IL-12p70 even after TLR and CD40L stimulation. agonists. CD34-derived LCs are widely used as models for skin LC. MoDCs also secreted significant quantities of IL-23 after stimula- They express langerin and CD1a costimulatory molecules and have tion with polyIC (TLR3) or CL097 (TLR7/8). LCs and dDCs pro- duced lower amounts of IL-23 after TLR stimulation, however Birbeck granules but do not express markers associated with dDCs they were able to produce larger quantities when stimulated with such as CD11b, DC-SIGN, and the mannose receptor. The TLR pro- zymosan (TLR2/dectin-1), which is known to induce IL-23 (24) file of our in vitro-generated dDC are similar to that described for (Fig. 8C). None of the three DC populations were able to secrete dDC isolated from skin (28). There is lack of agreement between IL-1␤ following TLR stimulation as observed previously (5, 27), two independent studies of the TLR profile in LC isolated from with the exception of moDCs stimulated with the dectin-1 ligand skin. One reports expression of TLRs 1, 2, 3, 5, 6, and 10 (29) curdlan (data not shown). Thus, moDCs produced both IL-12p70 while a second observed expression of TLRs 1, 3, 6, and 7 (28). and IL-23 in response to TLR3 and TLR7/8 stimulation in accor- We find expression of TLRs 1, 2, 4, 5, 6, 7, and 10 in CD34- dance with their ability to induce IFN-␥ and IL-17A secretion from derived LC. Studies using LCs derived from skin have shown an memory T cells and Th1 polarization of naive precursors. All three absence of TLR4 and TLR7/8 (29) or TLR2, TLR4, and TLR5 DC types produced high levels of IL-23 after zymosan but not mRNA (28). Thus, there may be some differences between the TLR stimulation, which correlated with their naive Th17 polaris- CD34-derived in vitro model and ex vivo-isolated LCs. Interest- ing ability. However, only zymosan-stimulated moDCs or super- ingly, we show for the first time the three DC subsets tested had natants could increase the percentage of memory Th17 cells. De- differing abilities to induce T cell proliferation. LCs and dDCs spite the ability of LCs and dDCs to induce memory T cell were more proficient at inducing naive T cell proliferation whereas cytokine secretion and some degree of Th1 polarization, they did moDCs were better at inducing the proliferation of memory T not secrete as much IL-12p70 or IL-23 and their supernatants cells. Such differences may be partially due to the ability of could not induce memory T cell cytokine production, suggesting moDCs to produce cytokines, including IL-12p70 and IL-23, that their mechanism of action relies on cell-associated factors which can induce memory T cell proliferation (30). These intrinsic rather than soluble cytokines. differences between the DC subsets may indicate functional The Journal of Immunology 2249 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

FIGURE 7. MoDC supernatants induce IL-17A and IFN-␥ production by memory T cells.Memory CD4ϩ T cells were cultured with anti-CD2/ CD3/CD28 beads for 7 days with medium alone, supernatants from un- stimulated DCs or TLR-stimulated or zymosan-stimulated DCs (in vitro- derived) before re-stimulation and intracellular cytokine staining. Data p Ͻ ,ء .shown are the mean Ϯ SEM of four to six independent experiments 0.05 analyzed by the Mann-Whitney U test compared with medium control.

FIGURE 8. Cytokine production by DCs stimulated with TLR agonists. specialization in primary and secondary responses and thus differ- In vitro-derived DCs were stimulated for 24 h with TLR or dectin agonists ing roles in guiding adaptive immunity. and CD40L, after which the supernatants were assayed for IL-12/23p40 Ϯ Much work has been conducted on the differentiation of naive (A), IL-12p70 (B), and IL-23 (C) by ELISAs. Data shown are the mean -p Ͻ 0.05 analyzed by the Mann ,ء .SEM of five to eight experiments precursors into Th17 cells, however few studies have looked at the Whitney U test compared with unstimulated DCs. influence of DC subsets on naive T cell polarization. Studies have variably shown the importance of IL-1␤, IL-6, IL-23, and TGF-␤ in human Th17 polarization (5–9), however these studies have often included a strong TCR stimulus, high concentrations of cy- study by Mathers and colleagues (32), CD45RAϩ T cells were tokines, and IFN-␥/IL-4/IL-12 neutralization, which may not rep- isolated from adult peripheral blood to a purity of Ͼ90%, how- resent the conditions present in vivo. We were unable to induce the ever CD45ROϩ memory T cells can revert to expression of polarization of Th17 cells from naive T cells using TLR-stimulated CD45RA (23, 34), which means there may be contaminating DCs even in IMDM, in agreement with the findings of other stud- memory Th17 cells that are expanded upon culture, thus con- ies using moDCs (10) and monocytes (31). However, some studies founding the data of apparent de novo Th17 differentiation. In have seemingly been able to induce Th17 polarization from naive our system, we have not used anti-CD3 Ab or IL-2, and have T cells using DCs (5, 32). The differences between these data may confirmed data with serum-free or IMDM and used naive CD4ϩ be due to several technical factors including the addition of cells isolated from cord blood. anti-CD3 Ab and IL-2, the type of medium (33), the presence of In contrast to TLR stimulation, DCs stimulated with the dectin serum (7, 8), or the source and purity of naive T cells. In the ligands zymosan or curdlan were able to induce Th17 polarization 2250 T CELL POLARIZATION INDUCED BY DCs of naive T cells in IMDM, which correlated with the ability of all cytokines play a role. Several studies have highlighted the im- three DC types to secrete IL-23 in response to zymosan. Thus portance of IL-23 and IL-1 in memory IL-17 production (7, potentially pathogenic Th17 cells may only be induced in response 9–11), however we have shown that in the presence of low to a restricted range of pathogens that specifically stimulate dectin. amounts of IL-23, efficient memory IL-17 responses can still be The incapacity of TLR-stimulated DCs to induce Th17s appears to induced by LCs and dDCs. Because the addition of LC or dDC be due to the inability to produce key cytokines needed for Th17 supernatants had no enhancing effect on memory cell cytokine development, because the DCs crucially did not secrete IL-1␤ and production, it may be that a cell contact mechanism is operating only moDCs produced significant amounts of IL-23 after TLR- instead. Our findings are similar to that observed with TLR- stimulation. Additionally, the supplementation of MLR cultures stimulated monocytes, whose mode of action depended on cell with exogenous IL-1␤, IL-6, IL-23, and TGF-␤ cytokines was able contact and whose supernatants also had no amplifying effect on to re-establish Th17 development. This raises the question of the IL-17ϩ cells (31). whether in vitro T cell differentiation models using high concen- In contrast to LCs and dDCs, moDCs induced memory IFN-␥ trations of cytokines and strong TCR stimuli are comparable to the and IL-17 by a mechanism partially dependent on soluble factors. quality of signals offered by DCs in vivo. Nonetheless, it is also MoDCs produced high levels of both IL-12p70 and IL-23 with possible that in vivo, DCs are able to produce these cytokines in TLR3 or TLR7/8 stimulation. IL-12p70 induces IFN-␥ production sufficient quantities, with the help of other cells and factors present from both naive and memory T cells (30) in accordance with the in the lymph node environment. Other possible reasons for our strong Th1 polarization seen with TLR3 and TLR7/8 stimulation. inability to induce Th17 using TLR-stimulated DCs may include IL-23 induces memory IL-17 but also memory IFN-␥ production the kinetics of stimulation or the production of inhibitory factors (38), which may explain the particularly strong IFN-␥ memory Downloaded from by the DCs (including IL-12p70, IFN-␥, and IL-4) (4), although at responses. Stimulation of moDCs through dectin or TLR2 as well least some of these factors would presumably be present in the in as supernatants from dectin-stimulated moDCs significantly en- vivo situation. The presence of inhibitory factors such as IL-12p70 hanced memory IL-17 production. In contrast to dectin stimulation and IFN-␥ may also partially explain why the moDCs induced which induced high levels of IL-23, TLR2-LPS only induced lower levels of naive Th17sthan the LCs and dDCs in IMDM. lower levels of IL-23 which may indicate that only low amounts of

Despite the inability of TLR-stimulated DCs to generate Th17 IL-23 are needed for IL-17 production. Additionally, IL-12 inhib- http://www.jimmunol.org/ cells from naive precursors, DCs could efficiently polarize Th1 its IL-17 induction and the IL-17-enhancing effect of IL-23 (38), cells, particularly TLR3 or TLR7/8 stimulated dDCs and moDCs. suggesting that the outcome is dependent on the relative balance of MoDCs stimulated with these agonists induced high levels of IL- IL-12p70 vs IL-23. Because dectin and TLR2-stimulated moDCs 12p70, which is known to be a key Th1 polarising factor. Indeed, did not produce any IL-12p70 but still produced IL-23, this bal- TLR3-stimulated moDC supernatants were also able to induce na- ance may have been sufficient to induce IL-17. ive Th1 polarization, demonstrating the reliance on a soluble cy- In experiments where total lymphocytes were used in the MLRs, tokine for this effect. Interestingly, dDCs treated with the TLR3 LCs induced a significantly greater IL-17 and IL-4 response than agonist induced equivalent levels of Th1 polarization as the dDCs and moDCs as previously observed (32, 39). In our cultures, moDCs, but in the absence of detectable IL-12p70, as reported this may have been due to the other non-T cells present in the by guest on September 24, 2021 previously for LCs and dDCs (20). Thus, even though dDCs and lymphocyte-enriched population, which augmented the effect of moDCs have a similar surface phenotype (35) and similar ability to the LCs. Although the composition of these other cells may not be induce Th1 responses after TLR3 or TLR7/8 stimulation they have representative of cells in the skin tissue or lymph node, this result distinct modes of action. highlights how the presence of other cells may dramatically alter All DC types analyzed were able to induce memory Th1 and the responses induced by different DCs, which should be taken into Th17 responses with equal ability, although TLR stimulation of account when comparing in vitro and in vivo studies. LCs and dDCs had no significant effect on the magnitude of the In conclusion, we have shown that functionally distinct popu- memory response, as reported previously for moDCs (10). The lations of human DCs can be differentially stimulated by TLRs exception was dectin- or TLR2-stimulated moDCs which signifi- agonists, have distinct effects on T cell subsets and mediate these cantly increased the percentage of IL-17ϩ cells. At present it is effects by different mechanisms. TLR-stimulated DCs were unable unclear whether the DCs are causing the reactivation of precom- to induce Th17 polarization of naive T cells, despite strong Th1 mitted effector memory Th17 cells (11) which may explain why polarization, but could still efficiently produce memory Th1 and TLR stimulation only has a minimal effect, because memory re- Th17 responses. The effect of the type of TLR stimulus was also sponses are less dependent on costimulatory signals induced by demonstrated, with TLR3 and TLR7/8 stimulation skewing toward TLR stimulation (36). Alternatively, DC factors could be affecting a Th1 response, whereas TLR2 or dectin stimulation increased the un-polarised Ag-experienced cells, which may be why dectin and IL-17 response in moDCs. MoDCs appear to partially mediate TLR2-stimulated moDCs augment the proportion of IL-17ϩ cells. their effect via soluble factors, whereas LCs and dDCs can elicit a Even though all DC subsets induced a similar magnitude of Th1 similar memory response to moDCs, but this appears to be depen- and Th17 memory responses, the mechanism by which they me- dent on cell contact. diated their effect differed. MoDCs supernatants, containing IL- 12p70 and high levels of IL-23 were able to enhance the memory Disclosures Th1 and Th17 responses, whereas LC and dDC supernatants could The authors have no financial conflict of interest. not. In accordance with this observation, LC and dDC supernatants did not contain IL-12p70 (except TLR7/8-stimulated dDCs) and contained lower amounts of IL-23, despite the production of the References 1. Mosmann, T. R., and S. Sad. 1996. The expanding universe of T-cell subsets: IL-12/23p40 common subunit as observed previously (20, 27, 37). Th1, Th2 and more. Immunol. Today 17: 138–146. Interestingly, unlike moDCs, supernatants from zymosan-stimu- 2. Weaver, C. T., R. D. Hatton, P. R. Mangan, and L. E. Harrington. 2007. IL-17 lated LCs and dDCs which contained high levels of IL-23 were family cytokines and the expanding diversity of effector T cell lineages. Annu. Rev. Immunol. 25: 821–852. unable to enhance memory Th17 responses. Further studies that 3. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of neutralize IL-12p70 and IL-23 should clarify to what extent these immunity. Nature 392: 245–252. The Journal of Immunology 2251

4. Korn, T., E. Bettelli, M. Oukka, and V. K. Kuchroo. 2009. IL-17 and Th17 cells. 24. Gerosa, F., B. Baldani-Guerra, L. A. Lyakh, G. Batoni, S. Esin, Annu. Rev. Immunol. 27: 485–517. R. T. Winkler-Pickett, M. R. Consolaro, M. De Marchi, D. Giachino, 5. Acosta-Rodriguez, E. V., G. Napolitani, A. Lanzavecchia, and F. Sallusto. 2007. A. Robbiano, et al. 2008. Differential regulation of 12 and interleukin 1␤ and 6 but not transforming growth factor-␤ are essential for the 23 production in human dendritic cells. J. Exp. Med. 205: 1447–1461. differentiation of -producing human T helper cells. Nat Immunol. 8: 25. Veldhoen, M., K. Hirota, J. Christensen, A. O’Garra, and B. Stockinger. 2009. 942–949. Natural agonists for aryl hydrocarbon receptor in culture medium are essential for 6. Wilson, N. J., K. Boniface, J. R. Chan, B. S. McKenzie, W. M. Blumenschein, optimal differentiation of Th17 T cells. J. Exp. Med. 206: 43–49. J. D. Mattson, B. Basham, K. Smith, T. Chen, F. Morel, et al. 2007. Development, 26. Snijders, A., P. Kalinski, C. M. Hilkens, and M. L. Kapsenberg. 1998. High-level cytokine profile and function of human interleukin 17-producing helper T cells. IL-12 production by human dendritic cells requires two signals. Int. Immunol. 10: Nat. Immunol. 8: 950–957. 1593–1598. 7. Manel, N., D. Unutmaz, and D. R. Littman. 2008. The differentiation of human 27. Rozis, G., A. Benlahrech, S. Duraisingham, F. Gotch, and S. Patterson. 2008. TH-17 cells requires transforming growth factor-␤ and induction of the nuclear Human Langerhans’ cells and dermal-type dendritic cells generated from CD34 receptor ROR␥t. Nat. Immunol. 9: 641–649. stem cells express different toll-like receptors and secrete different cytokines in 8. Volpe, E., N. Servant, R. Zollinger, S. I. Bogiatzi, P. Hupe, E. Barillot, and response to toll-like receptor ligands. Immunology 124: 329–338. V. Soumelis. 2008. A critical function for transforming growth factor-␤, inter- 28. van der Aar, A. M., R. M. Sylva-Steenland, J. D. Bos, M. L. Kapsenberg, leukin 23 and proinflammatory cytokines in driving and modulating human E. C. de Jong, and M. B. Teunissen. 2007. Loss of TLR2, TLR4, and TLR5 on TH-17 responses. Nat. Immunol. 9: 650–657. Langerhans cells abolishes bacterial recognition. J. Immunol. 178: 1986–1990. 9. Yang, L., D. E. Anderson, C. Baecher-Allan, W. D. Hastings, E. Bettelli, 29. Flacher, V., M. Bouschbacher, E. Verronese, C. Massacrier, V. Sisirak, M. Oukka, V. K. Kuchroo, and D. A. Hafler. 2008. IL-21 and TGF-␤ are required O. Berthier-Vergnes, B. de Saint-Vis, C. Caux, C. Dezutter-Dambuyant, for differentiation of human TH17 cells. Nature 454: 350–352. S. Lebecque, and J. Valladeau. 2006. Human Langerhans cells express a specific 10. van Beelen, A. J., Z. Zelinkova, E. W. Taanman-Kueter, F. J. Muller, TLR profile and differentially respond to viruses and Gram-positive bacteria. D. W. Hommes, S. A. Zaat, M. L. Kapsenberg, and E. C. de Jong. 2007. Stim- J. Immunol. 177: 7959–7967. ulation of the intracellular bacterial sensor NOD2 programs dendritic cells to 30. Oppmann, B., R. Lesley, B. Blom, J. C. Timans, Y. Xu, B. Hunte, F. Vega, promote interleukin-17 production in human memory T cells. Immunity 27: N. Yu, J. Wang, K. Singh, et al. 2000. Novel p19 protein engages IL-12p40 to 660–669. form a cytokine, IL-23, with biological activities similar as well as distinct from 11. Liu, H., and C. Rohowsky-Kochan. 2008. Regulation of IL-17 in human CCR6ϩ IL-12. Immunity 13: 715–725. Downloaded from effector memory T cells. J. Immunol. 180: 7948–7957. 31. Evans, H. G., T. Suddason, I. Jackson, L. S. Taams, and G. M. Lord. 2007. 12. Trinchieri, G. 2003. Interleukin-12 and the regulation of innate resistance and Optimal induction of T helper 17 cells in humans requires T cell receptor ligation adaptive immunity. Nat. Rev. Immunol. 3: 133–146. in the context of Toll-like receptor-activated monocytes. Proc. Natl. Acad. Sci. 13. Shortman, K., and Y. J. Liu. 2002. Mouse and human dendritic cell subtypes. Nat. USA 104: 17034–17039. Rev. Immunol. 2: 151–161. 32. Mathers, A. R., B. M. Janelsins, J. P. Rubin, O. A. Tkacheva, W. J. Shufesky, 14. Valladeau, J., and S. Saeland. 2005. Cutaneous dendritic cells. Semin. Immunol. S. C. Watkins, A. E. Morelli, and A. T. Larregina. 2009. Differential capability 17: 273–283. of human cutaneous dendritic cell subsets to initiate Th17 responses. J. Immunol.

15. Kadowaki, N., S. Ho, S. Antonenko, R. W. Malefyt, R. A. Kastelein, F. Bazan, 182: 921–933. http://www.jimmunol.org/ and Y. J. Liu. 2001. Subsets of human dendritic cell precursors express different 33. Veldhoen, M., K. Hirota, J. Christensen, A. O’Garra, and B. Stockinger. 2008. toll-like receptors and respond to different microbial . J. Exp. Med. 194: Natural agonists for aryl hydrocarbon receptor in culture medium are essential for 863–869. optimal differentiation of Th17 T cells. J. Exp. Med. 206: 43–49. 16. Jarrossay, D., G. Napolitani, M. Colonna, F. Sallusto, and A. Lanzavecchia. 2001. 34. Michie, C. A., A. McLean, C. Alcock, and P. C. Beverley. 1992. Lifespan of Specialization and complementarity in microbial molecule recognition by human human lymphocyte subsets defined by CD45 isoforms. Nature 360: 264–265. myeloid and plasmacytoid dendritic cells. Eur. J. Immunol. 31: 3388–3393. 35. Grassi, F., C. Dezutter-Dambuyant, D. McIlroy, C. Jacquet, K. Yoneda, 17. Sallusto, F., and A. Lanzavecchia. 1994. Efficient presentation of soluble S. Imamura, L. Boumsell, D. Schmitt, B. Autran, P. Debre, and A. Hosmalin. by cultured human dendritic cells is maintained by granulocyte/macrophage col- 1998. Monocyte-derived dendritic cells have a phenotype comparable to that of ony-stimulating factor plus and downregulated by tumor necrosis dermal dendritic cells and display ultrastructural granules distinct from Birbeck factor ␣. J. Exp. Med. 179: 1109–1118. granules. J. Leukocyte Biol. 64: 484–493. 18. Caux, C., B. Vanbervliet, C. Massacrier, C. Dezutter-Dambuyant, B. de Saint-Vis, 36. Croft, M., L. M. Bradley, and S. L. Swain. 1994. Naive versus memory CD4 T ϩ C. Jacquet, K. Yoneda, S. Imamura, D. Schmitt, and J. Banchereau. 1996. CD34 cell response to antigen. Memory cells are less dependent on accessory cell co- by guest on September 24, 2021 hematopoietic progenitors from human cord blood differentiate along two inde- stimulation and can respond to many antigen-presenting cell types including rest- pendent dendritic cell pathways in response to GM-CSFϩTNF ␣. J. Exp. Med. ing B cells. J. Immunol. 152: 2675–2685. 184: 695–706. 37. Ratzinger, G., J. Baggers, M. A. de Cos, J. Yuan, T. Dao, J. L. Reagan, C. Munz, 19. Warger, T., H. Schild, and G. Rechtsteiner. 2007. Initiation of adaptive immune G. Heller, and J. W. Young. 2004. Mature human Langerhans cells derived from responses by transcutaneous . Immunol. Lett. 109: 13–20. CD34ϩ hematopoietic progenitors stimulate greater cytolytic T lymphocyte ac- 20. Morelli, A. E., J. P. Rubin, G. Erdos, O. A. Tkacheva, A. R. Mathers, tivity in the absence of bioactive IL-12p70, by either single peptide presentation A. F. Zahorchak, A. W. Thomson, L. D. Falo, Jr., and A. T. Larregina. 2005. or cross-priming, than do dermal-interstitial or monocyte-derived dendritic cells. CD4ϩ T cell responses elicited by different subsets of human skin migratory J. Immunol. 173: 2780–2791. dendritic cells. J. Immunol. 175: 7905–7915. 38. Hoeve, M. A., N. D. Savage, T. de Boer, D. M. Langenberg, R. de Waal Malefyt, 21. Kato, H., O. Takeuchi, S. Sato, M. Yoneyama, M. Yamamoto, K. Matsui, T. H. Ottenhoff, and F. A. Verreck. 2006. Divergent effects of IL-12 and IL-23 S. Uematsu, A. Jung, T. Kawai, K. J. Ishii, et al. 2006. Differential roles of MDA5 on the production of IL-17 by human T cells. Eur. J. Immunol. 36: 661–670. and RIG-I helicases in the recognition of RNA viruses. Nature 441: 101–105. 39. Klechevsky, E., R. Morita, M. Liu, Y. Cao, S. Coquery, L. Thompson-Snipes, 22. Annunziato, F., L. Cosmi, V. Santarlasci, L. Maggi, F. Liotta, B. Mazzinghi, F. Briere, D. Chaussabel, G. Zurawski, A. K. Palucka, et al. 2008. Functional E. Parente, L. Fili, S. Ferri, F. Frosali, et al. 2007. Phenotypic and functional specializations of human epidermal Langerhans cells and CD14ϩ dermal den- features of human Th17 cells. J. Exp. Med. 204: 1849–1861. dritic cells. Immunity 29: 497–510. 23. Stockinger, B., M. Veldhoen, and B. Martin. 2007. Th17 T cells: linking innate 40. Wang, X., and B. Seed. 2003. A PCR primer bank for quantitative gene expres- and adaptive immunity. Semin. Immunol. 19: 353–361. sion analysis. Nucleic Acids Res. 31: 1–8.