They expressed significant levels of integrins 1 and 4, and low levels of integrins 5, 6, and V (Fig. immobilized proteins, whereas immobilized anti-integrin 1 antibodies enhanced the GM-CSF/Am80-induced ALDH1A2 expression without inducing proinflammatory cytokines. Sequential stimulation of splenic pre-DCs with GM-CSF/Am80 and immobilized E-cadherin or anti-integrin 1 antibody also induced differentiation to mature DCs with high ALDH activity. The E-cadherin-treated BM-DCs induced PH-797804 gut-tropic Foxp3+T cells and alleviated DSSinduced colitis, whereas the TLR-L-treated DCs aggravated DSSinduced colitis. The results suggest that integrin 1-mediated signals contribute to the differentiation and maturation of RA-producing anti-inflammatory DCs. Retinoic acid (RA), an active vitamin A metabolite, plays important roles in gut immunity. Dendritic cells (DCs) in the gut-related lymphoid organs, mesenteric lymph nodes (MLNs) and Payers patches (PPs) produce RA and imprint gut-homing specificity on T and B cells upon antigenic stimulation1, 2 . RA also PH-797804 regulates the functional differentiation of T cells. RA suppresses the differentiation of Th1 and PH-797804 Th17 cells via RA receptors (RARs)3, 4, 5, 6, 7, although it has recently been reported that RA signaling is essential for Th1 cell lineage stability and preventing the conversion of Th1 to Th178. On the other hand, RA enhances the TGF–dependent differentiation of nave CD4+T cells into Foxp3+inducible regulatory T cells (iTregs)4, 5, 6, 7, 9, 10, 11, 12. Thus, RA has been implicated in the induction of oral tolerance. Indeed, vitamin A deficiency results in the failure to tolerate oral antigens13. RA appears to regulate the nature of MLN-DCs as well as that of T cells primed in MLNs13, 14, 15, 16. Retinal dehydrogenases [e. g., aldehyde dehydrogenase 1A (ALDH1A)] encoded by theAldh1agene family are key enzymes that produce RA, and they are expressed in limited cell types. RA-producing DCs in MLNs, PPs, and the small intestinal lamina propria (LP) express the ALDH1A2 isoform at different levels1, 14, 17. These ALDH1A2+DCs are included in the CD103+DC subset9, Rabbit Polyclonal to API-5 14. As most CD103+MLN-DCs migrate from the LP18, it is probable that ALDH1A2+MLN-DCs mostly originate from the LP. Gut-homing DC precursors, such as lineageCD11cintB220+47+CCR9cells, develop in the bone marrow (BM) and migrate into the intestinal LP19. The microenvironment of the LP appears to induce ALDH1A2 expression in immature DCs. We previously found that GM-CSF induces ALDH1A2 expression in FMS-like tyrosine kinase 3 (FLT3) ligand (L)generated immature BM-DCs14, 20. RAR-mediated signaling is required for GM-CSFinduced ALDH1A2 expression, although RA by itself only weakly inducesAldh1a2expression14. In the LP, RA may be produced by ALDH1A1+epithelial cells, ALDH1A2+DCs, and subpopulations of macrophages and stromal cells1, 21, 22, whereas GM-CSF may be produced by a variety of cell types, including macrophages and stromal cells, in an RA-dependent manner14, 21. ALDH1A2highMLN-DCs are mature DCs14. GM-CSF and RA are not sufficient to generate mature DCs with high ALDH1A2 activity from BM-DCsin vitro14, 23. We previously illustrated that the simultaneous stimulation of FLT3-Lgenerated BM-DCs with GM-CSF and Toll-like receptor (TLR)-L induced their maturation and high levels of ALDH1A2 expression14. However , we found here that they produced proinflammatory cytokines, unlike steady-state CD103+ALDH1A2+MLN-DCs in SPF mice. Thus, we searched for factors that are present in the intestinal tissue environment and enhance ALDH1A2 expression in BM-DCs without inducing proinflammatory cytokine production. We found that stimulation of FLT3-Lgenerated BM-DCs with GM-CSF/RA and subsequent stimulation with immobilized E-cadherin efficiently generate RAproducing anti-inflammatory DCs that can induce iTreg differentiation, and that integrin 1-mediated signals contribute to the E-cadherin effect. == Methods == == Mice == C57BL/6 mice and TCR-transgenic OT-II/Rag2/mice were obtained from CLEA Japan (Tokyo, Japan) and Taconic (Hudson, NY), respectively. GM-CSF-deficient (GM-CSF/) mice were kindly provided by Drs. Jeffrey A. Whitsett and Bruce C. Trapnell (Childrens Hospital Medical Center, Cincinnati, OH) with kind permission of Dr . Glenn Dranoff (Dana Farber Cancer Institute, Boston MA). All animal PH-797804 experiments were performed according to protocols approved by the Animal Care and Use Committee of Tokushima Bunri University (TokushimaKagawa, Japan; permit numbers: KP11-41-003a, KP12-41-001a, KP12-41-002, KP13-41-001, KP13-41-002, KP14-41-001, KP14-41-002, KP15-41-001, KP15-41-002, and KP-16-41-001). == Antibodies == The anti-integrin 1 monoclonal antibody (mAb), KMI6, was purchased from OriGene Technologies (Rockville, MD). Blocking mAbs to integrin 1 (clone HM1), 4 (R1-2), 5 (5H10-27), 1 (HM1-1), and 7 (FIB27) were from BioLegend (San Diego, CA); E/CD103 (M290) and 1 (9EG7) were from BD Biosciences (San Jose, CA). The following mAbs were used for flow cytometry and cell sorting: CD11c (N418), CD40 (1C10), CD86 (GL1), CD103 (2E7), integrin 1 (HM1), 2 (HM2), 4 (R1-2), 5 (5H10-27), 6 (GoH3), V (RMV-7), 1 (HM1-1),.