Cutting Edge: 4-1BB Controls Regulatory Activity in Dendritic Cells through Promoting Optimal Expression of Dehydrogenase This information is current as of October 1, 2021. Seung-Woo Lee, Yunji Park, So-Young Eun, Shravan Madireddi, Hilde Cheroutre and Michael Croft J Immunol 2012; 189:2697-2701; Prepublished online 15 August 2012;

doi: 10.4049/jimmunol.1201248 Downloaded from http://www.jimmunol.org/content/189/6/2697

Supplementary http://www.jimmunol.org/content/suppl/2012/08/15/jimmunol.120124 Material 8.DC1 http://www.jimmunol.org/ References This article cites 22 articles, 8 of which you can access for free at: http://www.jimmunol.org/content/189/6/2697.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on October 1, 2021

• 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 Immunology 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 © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Cutting Edge: 4-1BB Controls Regulatory Activity in Dendritic Cells through Promoting Optimal Expression of Retinal Dehydrogenase Seung-Woo Lee,*,† Yunji Park,†,‡ So-Young Eun,* Shravan Madireddi,* Hilde Cheroutre,‡ and Michael Croft* Dendritic cells (DC) in the gut promote immune tol- activity is specifically associated with expression of retinal erance by expressing retinal dehydrogenase (RALDH), dehydrogenases (RALDH or ) that di- an that promotes , which aids dif- rect retinoic acid (RA) production (5, 6). RA is made by DC in ferentiation of Foxp3+ inducible regulatory T cells both Peyer’s patches (PP) and mesenteric lymph nodes (MLN) (iTreg) in the intestinal mucosa. How RALDH expres- and can direct T cells to have gut-homing potential by inducing a4b7 and CCR9 (7, 8). Furthermore, RA synergizes with Downloaded from sion is regulated is unclear. We found that 4-1BB + (CD137), a member of the TNFR family, together TGF-b in promoting the differentiation of Foxp3 inducible with CD103, marked mesenteric lymph node DC with Treg (iTreg) while suppressing Th17 development (9–11). the highest level of RALDH activity, and ligation of Although the RA-producing GALT DC have been well char- 4-1BB maintained RALDH expression in these gut acterized in terms of activity, the factors that control the ca- DC. Moreover, 4-1BB signals synergized with those pacity of these DC to express RALDH and, therefore exhibit suppressive function, are not clear. http://www.jimmunol.org/ through TLR2 or GM-CSFR to promote RALDH ac- 4-1BB (CD137, TNFRSF9) (12) is a receptor in the TNFR tivity in undifferentiated DC. Correspondingly, 4-1BB– superfamily that is mostly known as a cosignaling molecule deficient mice were impaired in their ability to generate promoting the activation and expansion of peripheral effector iTreg in the GALT when exposed to oral Ag, and T cells (13, 14). Although the majority of research on 4-1BB has 4-1BB–deficient mesenteric lymph node DC displayed been directed to T cells, it is becoming clear that this molecule weak RALDH activity and were poor at promoting iTreg can influence other lineages of cells in both positive and negative development. Thus, our data demonstrate a novel activity manners, as exemplified by our recent data showing that 4-1BB

of 4-1BB in controlling RALDH expression and the reg- expressed on bone marrow progenitors limits myelopoiesis and by guest on October 1, 2021 ulatory activity of DC. The Journal of Immunology, DC development (15). In further pursuing this latter observa- 2012, 189: 2697–2701. tion, we found unusual constitutive expression of 4-1BB on MLN DC that overlapped to a great extent with CD103. We now show that these 4-1BB+ gut DC expressed the highest levels he mucosa of the gut is the largest surface in the body of RALDH. 4-1BB was also induced on undifferentiated DC and is continuously exposed to exogenous Ags that through TLR2 and GM-CSFR, molecules that have been im- T may be innocuous (food, commensal bacteria) or plicated in driving the development of RALDH+ regulatory pathogenic (invasive bacteria, food allergens) (1, 2). As DC. Moreover, signaling through 4-1BB promoted optimal such, lymphocytes distributed throughout the GALT must RALDH in these DC and imparted regulatory activity in the remain immunologically hyporesponsive to many Ags. Mi- cells. Correlating with these observations, Foxp3+ iTreg con- gratory intestinal dendritic cells (DC), marked by CD103 ex- version to oral Ag was dampened in the GALT of 4-1BB–de- pression, control immune homeostasis in the GALT through ficient mice, and MLN DC from 4-1BB–deficient mice the promotion of high frequencies of Foxp3+ regulatory T cells displayed an attenuated ability to promote iTreg differentiation (Treg) that suppress Th1 and Th17 populations in effector as the result of reduced expression of RALDH. These data sites, such as the lamina propria (3, 4). This regulatory DC define a new and novel role for 4-1BB in DC biology.

*Division of Immune Regulation, La Jolla Institute for Allergy and Immunology, The online version of this article contains supplemental material. La Jolla, CA 92037; †Division of Integrative Bioscience and Biotechnology, Pohang Abbreviations used in this article: DC, dendritic cell; iTreg, inducible regulatory T cell; University of Science and Technology, Pohang, Gyeongbuk 790-784, Korea; and MDC, mesenteric lymph node dendritic cell; MFI, mean fluorescence intensity; MLN, ‡Developmental Immunology, La Jolla Institute for Allergy and Immunology, La Jolla, mesenteric lymph node; PP, Peyer’s patches; RA, retinoic acid; RALDH, retinal dehy- CA 92037 drogenase; SDC, spleen dendritic cell; Treg, regulatory T cell; wt, wild-type. Received for publication June 5, 2012. Accepted for publication July 16, 2012. Ó This work was supported by National Institutes of Health Grants AI42944 and AI85291 Copyright 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 (to M.C.) and Grant AI050265 (to H.C.). This is manuscript no. 1149 from the La Jolla Institute for Allergy and Immunology. Address correspondence and reprint requests to Dr. Michael Croft, Division of Immune Regulation, La Jolla Institute for Allergy and Immunology, 9420 Athena Circle, La Jolla, CA 92037. E-mail address: [email protected] www.jimmunol.org/cgi/doi/10.4049/jimmunol.1201248 2698 CUTTING EDGE: 4-1BB CONTROLS RALDH IN DCs

Materials and Methods Mice 4-1BB–deficient (16) and OT-II mice on a BL/6 background were bred at La Jolla Institute for Allergy and Immunology. C57BL/6J (CD45.2+) and C57BL/6Pep3b/BoyJ (CD45.1+) mice were purchased from The Jackson Lab- oratory. All experiments were in compliance with the regulations of the La Jolla Institute for Allergy and Immunology animal care committee in ac- cordance with guidelines of the Association for the Assessment and Accred- itation of Laboratory Animal Care. Cells DC from spleens and mesenteric lymph nodes (SDC and MDC, respectively) were isolated by MACS with CD11c MicroBeads, as described previously (15). Naive OT-II CD4 T cells were further purified by cell sorting by gating cells 2 as CD4+CD25 CD44loCD62Lhi, which results in negligible contamination + of Foxp3 cells, as described previously (17). Lymphocytes were isolated from FIGURE 1. 4-1BB is expressed on MDC and correlates with RALDH spleens, lymph nodes, and PP by treatment with collagenase D (Roche), activity. (A) 4-1BB expression on gated CD11c+MHC class IIhi DC in MDC m followed by mashing through a 70- m cell strainer. Intraepithelial lym- and SDC isolated from wt mice (8 wk old). Shaded graphs represent isotype phocytes and lamina propria lymphocytes were isolated as described (11). control. (B) MDC stained for CD103 versus 4-1BB expression (top left panel). 2 RALDH induction in SDC MDC gated into CD103+4-1BB and CD103+4-1BB+ subsets (top right

panel). MDC gated based on levels of CD103 and 4-1BB (1, orange, Downloaded from 3 5 2 2 A total of 2–3 10 SDC was cultured with a cytokine, GM-CSF (10–1000 CD103lo4-1BB ;2,green,CD103hi4-1BB ;3,blue,CD103lo4-1BBlo; pg/ml), or TLR2 ligands zymosan (0.25–25 mg/ml; Sigma-Aldrich) and Pam- 4, red, CD103hi4-1BBhi)(bottom panels). RALDH activity was determined 3-cys (0.1–10 mg/ml; InvivoGen) for 48 h. Where indicated, rat IgG (KLH/ C G1-2-2; 25 mg/ml), anti–4-1BB (3H3; 25 mg/ml), and inhibitors of signaling by ALDEFLUOR staining (top right panel, bottom panels). ( )MDCwere pathways were added at the start of culture. For blockade of signaling, the stained for 4-1BB and separated into subsets based on 4-1BB levels (left following were used: ERK inhibitor U0126 (5 mM), PI3K inhibitor Ly294002 panel). RALDH activity was assayed by ALDEFLUOR staining (right panel), (5 mM), NF-kB inhibitor Bay 11-7082 (1 mM, all from Sigma-Aldrich), and with graphs representing gated subsets. Results are representative of three

Wnt/b-catenin inhibitor XAV939 (5 mM; sc-296704; Santa Cruz Bio- experiments. http://www.jimmunol.org/ technology). The activity of RALDH was determined by ALDEFLUOR staining, as described (18). Most interestingly, the subset of MDC positive for both In vitro lineage differentiation of CD4+ T cells CD103 and 4-1BB displayed a higher level of activity of 5 4 A total of 1 3 10 naive OT-II T cells and 2 3 10 DC (MDC or SDC) was RALDH than did CD103+ cells that were negative for 4-1BB cultured in the presence of OVA323–339 peptide (0.1–10 mM) for 4 d. In + + + m (Fig. 1B, top right panel; CD103 4-1BB versus CD103 4- some cultures, SDC were preactivated with zymosan (2.5 g/ml) together 2 + with IgG or anti–4-1BB (25 mg/ml) for 48 h, washed three times with PBS, 1BB , 57.1 versus 38.9% ALDEFLUOR , and 19,128 versus and then cocultured with OT-II cells. Where indicated, exogenous cytokines 11,441 mean fluorescence intensity [MFI] of ALDEFLUOR). and reagents were added at the start of culture. Th cell lineages were deter- To control for RALDH activity, the assay was performed in by guest on October 1, 2021 mined by intracellular staining for IFN-g (Th1) and Foxp3 (Treg) after restimulating cells with PMA and ionomycin (Sigma-Aldrich) for 5 h in the the presence of diethylaminobenzaldehyde, a RALDH in- + presence of GolgiPlug (BD). hibitor (data not shown). When CD103 MDC were further subdivided based on the levels of CD103 and 4-1BB, the Oral tolerance and in vivo conversion of Treg activity of RALDH was found to be markedly stronger in 4- We exploited an oral tolerance model to observe the conversion of naive OT-II + 2/2 1BB high or low cells compared with 4-1BB negative cells cells into Foxp3 Treg in vivo (6). Briefly, wild-type (wt) or 4-1BB mice hi 2 (CD45.2+) received naive OT-II cells (1 3 106, CD45.1+) and 1 d later were (Fig. 1B, bottom panel, quadrant gates 1–4; CD103 4-1BB hi hi + challenged with OVA (1%; Sigma-Aldrich) in the drinking water for 5 d. versus CD103 4-1BB , 46.2 versus 71.9% ALDEFLUOR and 2 Water containing OVA was changed approximately every 36 h. The con- 12,944 versus 28,097 MFI of ALDEFLUOR; CD103lo4-1BB version of Treg was analyzed by intracellular staining of Foxp3 from donor lo lo + + + versus CD103 4-1BB , 35.7 versus 45.1% ALDEFLUOR and (CD45.1 ) and recipient (CD45.2 ) CD4 T cells. 10,804 versus 12,222 MFI of ALDEFLUOR). Similarly, gating Statistics only on the level of expression of 4-1BB revealed MDC with 2 All statistical analyses were done with a two-tailed Student t test. progressively increasing RALDH activity (Fig. 1C; 4-1BB versus 4-1BBlo versus 4-1BBhi, 20.2 versus 46.8 versus 62% Results and Discussion ALDEFLUOR+ and 6,339 versus 15,265 versus 20,879 MFI 4-1BB is expressed on CD103+ regulatory MDC of ALDEFLUOR). Thus, 4-1BB is a marker of RALDH Although SDC do not constitutively express 4-1BB, we found activity in migratory GALT DC. that approximately half of MDC expressed 4-1BB when an- alyzed immediately ex vivo (Fig. 1A). Moreover, the 4-1BB+ 4-1BB signaling maintains and augments RALDH DC population overlapped with the CD103+ population to To assess whether 4-1BB signaling might control RALDH a great extent, implying that 4-1BB expression might correlate expression or activity, MDC were cultured in vitro in the with RALDH activity and production of RA that were re- absence of stimulation or were stimulated with an agonist Ab ported to associate with CD103 expression (Fig. 1B, top left to 4-1BB. Although MDC strongly lost RALDH activity over panel). To assess this, we used a sensitive flow technique with a 48-h period without further stimulation, triggering 4-1BB a fluorescent substrate for aldehyde dehydrogenase (ALDE- allowed a large proportion of the cells to maintain high levels FLUOR). Fluorescence strongly corresponds with both the of active RALDH (Fig. 2A). level of mRNA for Aldh1a2 (RALDH2) that is expressed in Regulatory GALT DC are thought to develop or mature in 2 DC and the ability of DC to promote generation of Foxp3+ the periphery from RALDH precursors in response to GM- iTreg (18, 19). CSFR and/or TLR2 signals (18, 20). To potentially mimic The Journal of Immunology 2699

isolated ex vivo, indicating that other factors from the gut environment are also likely to contribute to RALDH ex- pression in GALT DC. Next, we wanted to confirm that RALDH induced through 4- 1BB signaling was functional. SDC, pretreated with zymosan together with agonist anti-4-1BB, were cultured with naive OT- II T cells and Ag. Under neutral conditions, naive Th cells stimulated with zymosan-treated SDC predominantly differ- entiated into IFN-g–producing Th1 cells, whereas Foxp3 in- duction and weaker Th1 differentiation were seen in cultures supplemented with exogenous TGF-b (Fig. 2D). SDC stim- ulated through TLR2 and 4-1BB promoted fewer Th1 cells in either culture condition (Fig. 2D) and more Foxp3+ T cells in the presence of TGF-b (Fig. 2D, 2E). Most importantly, these activities of 4-1BB were dependent on the production of RA from the DC, because LE540, a pan antagonist of the RA re- ceptor (6, 9), abolished iTreg development and reversed the suppression of Th1 differentiation (Fig. 2D, 2E). 4-1BB can trigger several signaling pathways in T cells, Downloaded from notably activation of NF-kB, PI3K/Akt, and ERK1/2 (13, 14). Of potential significance, these intracellular mediators were reported to be induced by TLR2 and GM-CSFR. Therefore, we added inhibitors to SDC cultures to determine whether these pathways drove RALDH through 4-1BB. Blocking acti- vation of ERK, PI3K, or NF-kB strongly suppressed RALDH http://www.jimmunol.org/ activity induced by 4-1BB, as well as that induced by TLR2 alone or GM-CSF alone (Fig. 2F, data not shown). In con- trast, manipulation of the Wnt/b-catenin pathway revealed no significant role in RALDH induction. FIGURE 2. 4-1BB is a signaling receptor for RALDH induction in DC. (A) MDC, isolated by MACS, were cultured for 1 or 2 d in the absence of Defective Foxp3 iTreg generation to oral Ag in 4-1BB–deficient mice stimulation with control IgG or were stimulated with an agonist Ab to 4-1BB (25 mg/ml). Activity of RALDH was assayed by ALDEFLUOR staining. The expression of RALDH in GALT DC is thought to be crucial Percentages of ALDEFLUOR+ cells are indicated. (B–F) MACS-isolated for inducing tolerance against orally encountered Ags by pro- by guest on October 1, 2021 SDC were stimulated with GM-CSF and zymosan for 48 h. Where indicated, moting RA production that contributes to the development m agonist anti–4-1BB (25 g/ml) or rat IgG was added at the start of culture. of Foxp3+ iTreg (5, 6). We then performed oral-tolerance (B) 4-1BB expression in SDC stimulated with GM-CSF (100 pg/ml) and m C experiments in 4-1BB–deficient mice. To assess induction zymosan (25 g/ml). ( ) Activity of RALDH in SDC was determined by + ALDEFLUOR staining when treated with GM-CSF (10 pg/ml), zymosan of Foxp3 Treg and focus on a potential role of 4-1BB on (2.5 mg/ml), and Pam-3-cys (1 mg/ml) for 48 h. (D and E) SDC, pretreated APC rather than T cells, we adoptively transferred wt naive 2 2 with zymosan in the presence of IgG or agonist anti-4-1BB for 48 h, were (Foxp3 CD25 ) OVA-specific OT-II T cells (CD45.1+) cultured with naive OT-II CD4 T cells and OVA peptide (0.01 mM) for 4 d. into 4-1BB–deficient mice and fed the animals OVA in their Where indicated, TGF-b (5 ng/ml) and LE540, the pan antagonist of RA drinking water. When transferred into wt mice, a significant receptor (2 mM), were added. Treg/Th1 development was determined by intracellular staining after restimulation of cells with PMA and ionomycin. number of T cells strongly upregulated Foxp3 in the GALT Representative flow plot (D) and the percentage of Foxp3+ T cells in indi- after oral Ag treatment (Fig. 3). In contrast, Treg induction vidual cultures (E) are shown. (F) Inhibitors for signaling pathways were was dramatically attenuated in the MLN, PP, and lamina added at the start of culture when SDC were stimulated with zymosan (25 propria lymphocytes when T cells were transferred into 4- 2 2 mg/ml). ERK, U0126 (5 mM); PI3K, Ly (5 mM); NF-kB(1mM); Wnt/ 1BB / mice. Defective Treg generation was specific to the b-catenin (5 mM). Results are representative of three experiments. **p . GALT because conversion was moderately increased in the 2 2 0.01. spleen of 4-1BB / recipients (Fig. 3). This implied that 2 2 4-1BB / mice possess an environment in the GALT that is this process and assess whether 4-1BB might participate in less favorable for inducible Treg development and that this such a differentiation event, we cultured SDC that do not is related to attenuated regulatory activity in mucosa-resident constitutively express RALDH or 4-1BB with GM-CSF or APC populations. zymosan (a TLR2 ligand). Importantly, 4-1BB was induced when SDC were stimulated through either pathway (Fig. 2B). 4-1BB–deficient MDC express low levels of RALDH and are poor at Moreover, when 4-1BB signaling was induced with an ago- inducing Treg nistic Ab. this led to significant upregulation of RALDH Although DC are believed to be the primary inducer of Treg activity compared with that promoted by either GM-CSF or induced by oral Ags in the GALT, some studies suggested that zymosan alone, although this was more pronounced with macrophages and intestinal epithelial cells may also contribute zymosan or another TLR2 ligand Pam-3-cys (Fig. 2C). The to the generation and/or maintenance of these iTreg (21, 22). level of RALDH activity induced in SDC with these com- To investigate whether the defective Treg induction that we binations of stimuli was lower than that observed in MDC observed in 4-1BB–deficient animals was likely related to al- 2700 CUTTING EDGE: 4-1BB CONTROLS RALDH IN DCs

FIGURE 3. Defective iTreg conversion to oral Ag in the GALT of 4-1BB– 2 deficient mice. Naive wt OT-II CD4 T cells (CD45.1+ and CD25-Foxp3 ) 2 2 were adoptively transferred into wt or 4-1BB / mice (CD45.2+) that were subsequently given OVA (1%) in the drinking water for 5 d. Lymphocytes from the indicated organs were analyzed on day 6 for the conversion of naive OT-II cells into Treg by gating on CD45.1 and staining for intracellular Foxp3. (A) Concatenate histograms of four individual samples shown in spleen (top panels) and MLN (bottom panels). Percentages of Foxp3+ cells in CD4+ populations positive for CD45.1 are shown. (B) Percentage of Foxp3+ cells within donor OT-II cells (CD45.1+) in each organ from individual mice.

Results are representative of three experiments. **p . 0.01, ***p . 0.001. Downloaded from LPL, Lamina propria lymphocytes. tered DC activity, we isolated DC from the MLN. This is the main location where naive CD4 T cells are converted into iTreg with oral Ag (5, 21). Significantly, the percentage of 2 2 MDC in 4-1BB / mice expressing RALDH activity was http://www.jimmunol.org/ strongly reduced and the level of activity was much lower than inwtMDC(Fig.4A,4B).Moreover, the level of active 2 2 RALDH in 4-1BB / MDC was comparable to that ob- 2 2/2 served in 4-1BB MDC from wt mice (compare with data in FIGURE 4. Defective regulatory function of MDC from 4-1BB mice. 2 2 A B 2/2 Fig. 1). In contrast, SDC from wt or 4-1BB / mice were ( and ) Activity of RALDH in MDC or SDC from wt or 4-1BB mice was determined by ALDEFLUOR staining. (A) Representative RALDH ac- indistinguishable and essentially negative for RALDH (Fig. + hi 2/2 tivity displayed as graph of ALDEFLUOR staining in gated CD11c class II 4A). The reduced activity of RALDH in 4-1BB MDC DC. (B) RALDH activity in MDC from individual mice. Left panel, was paralleled by a lower level of Aldh1a2 (RALDH2) mRNA percentage. Right panel,MFI.(C–F) Naive wt OT-II CD4 T cells by guest on October 1, 2021 2 2 (Supplemental Fig. 1A) and was not due to decreased num- (CD25 Foxp3 ) were cultured for 4 d with OVA peptide (1 mM) presented 2 2 bers of the CD103+ population (Supplemental Fig. 1B). on MDC isolated from wt or 4-1BB / mice. Cells were cultured without 2 2 C D F MDC from 4-1BB / mice did not show any obvious exogenous cytokines ( ) or with 5 ng/ml TGF-b ( – ) for 4 d to determine g surfacephenotypicdifferencescomparedwithwtMDC, default Th1 cell generation by intracellular staining of IFN- after restim- ulation of cells with PMA and ionomycin or Treg differentiation by intra- including expression of CD80, CD86, and MHC class II cellular Foxp3 staining in CD4+ T cells, respectively. Percentages of positive (Supplemental Fig. 1B). 4-1BB ligand was found on wt MDC cells are indicated. Where indicated, cells in (E) and (F) were cultured with 2/2 at low levels and at higher levels on 4-1BB MDC (Sup- RA (100 nM) or neutralizing Abs against IFN-g and IL-12 (10 mg/ml) added 2 2 plemental Fig. 1C), corresponding to our previous results that at the start of culture. (G) MDC from wt and 4-1BB / mice were stimulated showed that endogenously expressed 4-1BB antagonizes the as indicated for 24 h. IL-12p70 production was measured by ELISA in culture F G expression of 4-1BB ligand (15). To assess the functional ac- supernatants. Results are representative of two ( , ) and three experiments (A–E). **p . 0.01, ***p . 0.001. tivity of MDC, we cultured naive wt OT-II CD4 T cells with 2 2 cells isolated from wt or 4-1BB / mice and determined their relative ability to support Th1 (medium-only culture) or Foxp3 was observed when both were inhibited, implying that IL-12 iTreg (TGF-b culture) development driven by OVA peptide. was aberrantly expressed in these cultures. In accordance with 2 2 MDC from 4-1BB / mice induced more Th1 cells than did this, more IL-12p70 was produced from 4-1BB–deficient MDC from wt mice (Fig. 4C), and most significantly pro- MDC when they were stimulated with CpG DNA (Fig. 4G). moted fewer Foxp3+ iTreg (Fig. 4D). In contrast, SDC isolated Interestingly, IL-12p70 production was augmented when 2 2 from 4-1BB / mice displayed normal activity for the devel- MDC were costimulated with IFN-g, implying that a feed- opment of Th1 cells and Treg (Supplemental Fig. 2A, 2B). forward mechanism from IFN-g produced from responding 2 2 The defective iTreg-promoting ability of 4-1BB / MDC T cells might also contribute to enhanced IL-12 production. was strongly recovered when RA was added exogenously into Thus, 4-1BB–deficient MDC display a primary defect in culture, suggesting that it was largely due to the reduced ac- RALDH expression associated with poor Treg-inducing activity, tivity of RALDH (Fig. 4E). However, Treg conversion in the as well as deregulated IL-12 activity. 2 2 presence of RA was moderately lower with 4-1BB / MDC Collectively, these data suggest that 4-1BB expression on than with wt MDC, indicating that other factors might also GALT DC is functionally relevant in terms of regulating the contribute to the aberrant regulatory activity. Neutralization expression of active RALDH. This is a novel activity of 4-1BB of IFN-g or IL-12 strongly enhanced the differentiation of distinct from its known effects as a costimulatory molecule for 2 2 iTreg induced by 4-1BB / MDC (Fig. 4F), but no synergy T cells. The Journal of Immunology 2701

12. Kwon, B. S., and S. M. Weissman. 1989. cDNA sequences of two inducible T-cell Disclosures genes. Proc. Natl. Acad. Sci. USA 86: 1963–1967. The authors have no financial conflicts of interest. 13. Wang, C., G. H. Lin, A. J. McPherson, and T. H. Watts. 2009. Immune regu- lation by 4-1BB and 4-1BBL: complexities and challenges. Immunol. Rev. 229: 192–215. 14. Croft, M. 2009. The role of TNF superfamily members in T-cell function and References diseases. Nat. Rev. Immunol. 9: 271–285. 1. Hooper, L. V., and A. J. Macpherson. 2010. Immune adaptations that maintain 15. Lee, S. W., Y. Park, T. So, B. S. Kwon, H. Cheroutre, R. S. Mittler, and M. Croft. homeostasis with the intestinal microbiota. Nat. Rev. Immunol. 10: 159–169. 2008. Identification of regulatory functions for 4-1BB and 4-1BBL in myelopoiesis 2. Cerf-Bensussan, N., and V. Gaboriau-Routhiau. 2010. The immune system and the and the development of dendritic cells. Nat. Immunol. 9: 917–926. gut microbiota: friends or foes? Nat. Rev. Immunol. 10: 735–744. 16. Kwon, B. S., J. C. Hurtado, Z. H. Lee, K. B. Kwack, S. K. Seo, B. K. Choi, 3. Coombes, J. L., and F. Powrie. 2008. Dendritic cells in intestinal immune regu- B. H. Koller, G. Wolisi, H. E. Broxmeyer, and D. S. Vinay. 2002. Immune lation. Nat. Rev. Immunol. 8: 435–446. responses in 4-1BB (CD137)-deficient mice. J. Immunol. 168: 5483–5490. 4. Belkaid, Y., and G. Oldenhove. 2008. Tuning microenvironments: induction of 17. Duan, W., T. So, A. K. Mehta, H. Choi, and M. Croft. 2011. Inducible CD4+LAP regulatory T cells by dendritic cells. Immunity 29: 362–371. +Foxp32 regulatory T cells suppress allergic inflammation. J. Immunol. 187: 6499– 5. Coombes, J. L., K. R. Siddiqui, C. V. Arancibia-Ca´rcamo, J. Hall, C. M. Sun, 6507. Y. Belkaid, and F. Powrie. 2007. A functionally specialized population of mucosal 18. Yokota, A., H. Takeuchi, N. Maeda, Y. Ohoka, C. Kato, S. Y. Song, and M. Iwata. CD103+ DCs induces Foxp3+ regulatory T cells via a TGF-beta and retinoic acid- 2009. GM-CSF and IL-4 synergistically trigger dendritic cells to acquire retinoic dependent mechanism. J. Exp. Med. 204: 1757–1764. acid-producing capacity. Int. Immunol. 21: 361–377. 6. Sun, C. M., J. A. Hall, R. B. Blank, N. Bouladoux, M. Oukka, J. R. Mora, and 19. Guilliams, M., K. Crozat, S. Henri, S. Tamoutounour, P. Grenot, E. Devilard, Y. Belkaid. 2007. Small intestine lamina propria dendritic cells promote de novo B. de Bovis, L. Alexopoulou, M. Dalod, and B. Malissen. 2010. Skin-draining generation of Foxp3 T reg cells via retinoic acid. J. Exp. Med. 204: 1775–1785. lymph nodes contain dermis-derived CD103(2) dendritic cells that constitutively 7. Iwata, M., A. Hirakiyama, Y. Eshima, H. Kagechika, C. Kato, and S. Y. Song. 2004. produce retinoic acid and induce Foxp3(+) regulatory T cells. Blood 115: 1958– Retinoic acid imprints gut-homing specificity on T cells. Immunity 21: 527–538. 1968. 8. Mora, J. R., M. Iwata, and U. H. von Andrian. 2008. Vitamin effects on the im- 20. Manicassamy, S., R. Ravindran, J. Deng, H. Oluoch, T. L. Denning, S. P. Kasturi, mune system: vitamins A and D take centre stage. Nat. Rev. Immunol. 8: 685–698. K. M. Rosenthal, B. D. Evavold, and B. Pulendran. 2009. Toll-like receptor 2- 9. Mucida, D., Y. Park, G. Kim, O. Turovskaya, I. Scott, M. Kronenberg, and dependent induction of -metabolizing in dendritic cells promotes Downloaded from H. Cheroutre. 2007. Reciprocal TH17 and regulatory T cell differentiation medi- T regulatory responses and inhibits autoimmunity. Nat. Med. 15: 401–409. ated by retinoic acid. Science 317: 256–260. 21. Hadis, U., B. Wahl, O. Schulz, M. Hardtke-Wolenski, A. Schippers, N. Wagner, 10. Benson, M. J., K. Pino-Lagos, M. Rosemblatt, and R. J. Noelle. 2007. All-trans W. Mu¨ller, T. Sparwasser, R. Fo¨rster, and O. Pabst. 2011. Intestinal tolerance retinoic acid mediates enhanced T reg cell growth, differentiation, and gut homing requires gut homing and expansion of FoxP3+ regulatory T cells in the lamina in the face of high levels of co-stimulation. J. Exp. Med. 204: 1765–1774. propria. Immunity 34: 237–246. 11. Denning, T. L., Y. C. Wang, S. R. Patel, I. R. Williams, and B. Pulendran. 2007. 22. Iliev, I. D., E. Mileti, G. Matteoli, M. Chieppa, and M. Rescigno. 2009. Intestinal Lamina propria macrophages and dendritic cells differentially induce regulatory and epithelial cells promote colitis-protective regulatory T-cell differentiation through

interleukin 17-producing T cell responses. Nat. Immunol. 8: 1086–1094. dendritic cell conditioning. Mucosal Immunol. 2: 340–350. http://www.jimmunol.org/ by guest on October 1, 2021