Mesenchymal Stem/Stromal Cells Induce the Generation of Novel IL-10 −Dependent Regulatory Dendritic Cells by SOCS3 Activation This information is current as of October 3, 2021. Xingxia Liu, Xuebin Qu, Yuan Chen, Lianming Liao, Kai Cheng, Changshun Shao, Martin Zenke, Armand Keating and Robert C. H. Zhao J Immunol 2012; 189:1182-1192; Prepublished online 2 July

2012; Downloaded from doi: 10.4049/jimmunol.1102996 http://www.jimmunol.org/content/189/3/1182

Supplementary http://www.jimmunol.org/content/suppl/2012/07/02/jimmunol.110299 http://www.jimmunol.org/ Material 6.DC1 References This article cites 50 articles, 18 of which you can access for free at: http://www.jimmunol.org/content/189/3/1182.full#ref-list-1

Why The JI? Submit online.

by guest on October 3, 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 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. The Journal of Immunology

Mesenchymal Stem/Stromal Cells Induce the Generation of Novel IL-10–Dependent Regulatory Dendritic Cells by SOCS3 Activation

Xingxia Liu,*,1 Xuebin Qu,*,1 Yuan Chen,* Lianming Liao,† Kai Cheng,* Changshun Shao,‡ Martin Zenke,x Armand Keating,{,‖,# and Robert C. H. Zhao*

Suppression of immune response by mesenchymal stem/stromal cells (MSCs) is well documented. However, their regulatory effects on immune cells, especially regulatory dendritic cells, are not fully understood. We have identified a novel Sca-1+Lin2CD1172 MSC population isolated from mouse embryonic fibroblasts (MEF) that suppressed proliferation in vitro. Moreover, the Sca-1+Lin2CD1172 MEF-MSCs induced hematopoietic stem/progenitor cells to differentiate into novel regulatory dendritic cells (DCs) (Sca-1+Lin2CD1172 MEF-MSC–induced DCs) when cocultured in the absence of exogenous . Small interfering RNA silencing showed that Sca-1+Lin2CD1172 MEF-MSCs induced the generation of Sca-1+Lin2CD1172 MEF-MSC–induced Downloaded from DCs via IL-10–activated SOCS3, whose expression was regulated by the JAK–STAT pathway. We observed a high degree of H3K4me3 modification mediated by MLL1 and a relatively low degree of H3K27me3 modification regulated by SUZ12 on the promoter of SOCS3 during SOCS3 activation. Importantly, infusion of Sca-1+CD1172Lin2 MEF-MSCs suppressed the inflam- matory response by increasing DCs with a regulatory phenotype. Thus, our results shed new light on the role of MSCs in modulating regulatory DC production and support the clinical application of MSCs to reduce the inflammatory response in numerous disease states. The Journal of Immunology, 2012, 189: 1182–1192. http://www.jimmunol.org/

esenchymal stem/stromal cells (MSCs), capable of DCs, especially regulatory DCs (regDCs), are key regulators of multilineage differentiation (1, 2), are weakly immu- immune responses. Several types of regDCs, with similar functions M nogenic and possess immune regulatory properties (3– but different phenotypes (13–18), can be induced by a 5). Previous studies have shown that bone marrow-derived MSCs mixture that includes GM-CSF, IL-10, and TGF-b in vitro; how- (BM-MSCs) can secrete inhibitory cytokines, suppress ever, these induction conditions do not accurately mimic the mi-

activation and proliferation, and regulate dendritic cell (DC) croenvironment in vivo, which involves a complex orchestration by guest on October 3, 2021 functions (6–9). In vivo, MSCs can prevent and treat graft-versus- and regulation of cytokine production and cell-cell interaction. host disease by suppressing Th1 lymphocyte function (10), at- Previous reports showed that BM-MSCs could influence DC dif- tenuate the pathological manifestations of experimental autoim- ferentiation (8, 9), and Cao’s group (19) showed that endothelial mune encephalomyelitis via regulatory T cell expansion and splenic stroma could induce the differentiation of CD11bhighIalow proinflammatory Th17 cell depression (11), and inhibit B cell regDCs from bone marrow hematopoietic stem/progenitor cells activation, proliferation, and IgG secretion in systemic lupus (BM-HPCs), but the induction of regDCs in vivo remains con- erythematosus (12). Although MSCs can regulate T and B immune troversial. cells via many mechanisms, relatively little is known about the Recent data support the concept that specific expression effects of MSCs on DC function. patterns that determine the behavior of individual cells are under

*Institute of Basic Medical Sciences and School of Basic Medicine, Center of Excel- and Project of Institute of Basic Medical Sciences of Chinese Academy of Medical lence in Tissue Engineering, Chinese Academy of Medical Sciences and Peking Union Sciences Grant 2008PY09. Medical College, Beijing 100005, People’s Republic of China; †Academy of Integra- Address correspondence and reprint requests to Prof. Robert C.H. Zhao or Dr. tive Medicine, Fujian University of Traditional Chinese Medicine, Fuzhou 350108, Armand Keating, Institute of Basic Medical Sciences and School of Basic Medicine, Fujian, People’s Republic of China; ‡Department of Genetics, Rutgers, The State x Center of Excellence in Tissue Engineering, Chinese Academy of Medical Sciences University of New Jersey, Piscataway, NJ 08854; Department of Cell Biology, Insti- and Peking Union Medical College, 5 Dongdansantiao, Beijing 100005, People’s tute for Biomedical Engineering, Rhenish-Westphalian Technical University, Aachen { Republic of China (R.C.H.Z.) or Cell Therapy Program, Princess Margaret Hospital, University Medical School, 52074 Aachen, Germany; Cell Therapy Program, Prin- ‖ 610 University Avenue, Toronto, Ontario M5G 2M9, Canada (A.K.). E-mail cess Margaret Hospital, Toronto, Ontario M5G 2M9, Canada; Institute of Biomate- addresses: [email protected] (R.C.H.Z.) and [email protected] (A.K.) rials and Biomedical Engineering, University of Toronto, Toronto, Ontario M5G 2M9, Canada; and #Institute of Medical Science, University of Toronto, Toronto, The online version of this article contains supplemental material. Ontario M5G 2M9, Canada Abbreviations used in this article: BM-HPC, bone marrow hematopoietic stem/ 1X.L. and X.Q. contributed equally to this work. progenitor cell; BM-MSC, bone marrow-derived mesenchymal stem/stromal cell; ChIP, chromatin immunoprecipitation; DC, dendritic cell; HPC, hematopoietic Received for publication October 17, 2011. Accepted for publication May 23, 2012. stem/progenitor cell; imDC, immature DC; maDC, mature DC; MEF, mouse embry- This work was supported by “863 Projects” of Ministry of Science and Technology of onic fibroblast; MSC, mesenchymal stem/stromal cell; NA, neutralizing Ab; NS, People’s Republic of China Grant 2006AA02A109; National Natural Science Foun- normal saline; PCA, principal components analysis; qRT-PCR, quantitative RT- 2 2 dation of China Grants 30830052, 30700321, 30800429, and 30911130363; Beijing PCR; regDC, regulatory DC; sDC, Sca-1+Lin CD117 MEF-MSC–induced DC; Ministry of Science and Technology Grant D07050701350701; Major National siRNA, small interfering RNA. Science and Technology Project Grants 2008ZX09101-044 and 2009ZX09503-025; National Key Scientific Program of China Grant 2011CB964900; Program for Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 Changjiang Scholars and Innovative Research Team in University Grant IRT0909; www.jimmunol.org/cgi/doi/10.4049/jimmunol.1102996 The Journal of Immunology 1183 the control of epigenetic alterations, including histone mod- Coculture experiment ifications. Methylation of the histone lysines, H3K4 and H3K27, HPCs were enriched from bone marrow cells by depleting lineage-specific mediated by distinct histone methyltransferase complexes, is cells. In brief, an EasySep mouse hematopoietic progenitor enrichment kit highly correlated with transcriptional activation or repression (20, (StemCell Technologies) was used to deplete lineage-specific cells. After + 2 2 21). A growing body of evidence has shown that epigenetic HPCs were purified, they were seeded onto Sca-1 CD117 Lin mono- 3 5 alterations are involved in the regulation of various layers at a density of 1 10 in 2 ml per well in 6-well plates. The ratio of MSC/HPC is 1:10. LPS was added on day 4. In some experiments, expressed during normal embryonic development, stem cell dif- a neutralizing Ab (NA) for IL-10 (10 mg/ml) was added every 2 d; 100 mM ferentiation, cancer growth, and autoimmune disease progress (22, NSC 74859 (a phospho-STAT3 inhibitor; Merck) or 10 mM CGP 41251 (a 23). Histone methylation has also been implicated in the devel- phospho-STAT5 inhibitor; Merck) was used for 24 or 48 h, respectively; opment of immune responses, especially in the production of IL- serum replacement (StemCell Technologies) was used to substitute for serum. After coculture on Sca-1+CD1172Lin2 feeder layers for 9 d, the 12 by DCs (24, 25); however, little is known about the epigenetic nonadherent cells were gently washed out, and the remaining loosely ad- alterations during the generation of regDCs. herent cell clusters were collected and purified using DC Isolation Kit Although many studies have demonstrated that MSCs can (Miltenyi Biotec) for the following experiment (the purity of cells is modulate the function of DCs in vitro, few have addressed the .97%). Mouse BM-derived immature DCs (imDCs) and mature DCs underlying mechanisms and none has reported whether MSCs can (maDCs) were generated according to an established protocol (9). 2 modulate regDCs in vivo. In this study, we show that Sca-1+Lin Flow cytometric analysis CD1172 MSCs from mouse embryonic fibroblasts (MEF-MSCs) Flow cytometric analysis was performed, as previously described (9). The are capable of immune modulation and can induce the production fluorescent Abs used in the study included FITC-conjugated anti-mouse + 2 2 of a unique regDC subset (Sca-1 Lin CD117 MEF-MSC–in- Sca-1, CD4, CD9, CD90, CD29, CD31, CD34, CD44, CD45, CD117, duced DCs [sDCs]) in the absence of exogenous cytokines. Fur- CD14, CD19, H-2Kd, I-Ad, CD11c, CD80, CD86, CD40, IL-10, and PE- Downloaded from ther study shows that MEF-MSCs secrete IL-10 and regulate sDC conjugated anti-mouse CD25, CD73, Flk-1, CD11b, and Ia (BD Bio- sciences). For each Ab, IgG of the same isotype from the same species was production by activating SOCS3 through the JAK–STAT signaling used as isotype control (BD Biosciences). Cells were analyzed by flow pathway. Moreover, SOCS3 expression is regulated by stable cytometry using CellQuest software (BD Biosciences). reciprocal changes between H3K4me3 and H3K27me3 on the promoter regions mediated by MLL1 and SUZ12, respectively. Endocytosis assay and MLC

2 2 http://www.jimmunol.org/ Most importantly, Sca-1+CD117 Lin MEF-MSCs can alleviate Endocytosis and MLCs were performed, as previously described (9). an inflammatory immune response by upregulating regDCs Cytokine analysis in vivo. Our studies therefore show that MSCs can induce the production of regDCs both in vitro and in vivo to regulate immune Culture supernatants were collected on days 1, 3, 5, 7, and 9, respectively. b a responses. The concentrations of GM-CSF, IL-4, IL-10, TGF- , IL-12, and TNF- in supernatants were determined using ELISA kits (BD Technologies). GeneChip microarray and ProteinChip microarray Materials and Methods Mice and cell line Total RNA from each group (imDCs, maDCs, sDCs, and NA-IL-10–treated sDCs) and total from sDCs and NA-IL-10–treated sDCs were by guest on October 3, 2021 Five- to 6-wk-old BALB/c and C57BL/6 mice were purchased from the analyzed by GeneChip microarray (Agilent mouse gene expression 4 3 Laboratory Animal Center of Chinese Academy of Medical Sciences 44,000 microarrays) and by ProteinChip microarray (the JAK–STAT (Beijing, China). Pregnant mice between days 13 and 15 of gestation were pathway phospho Ab array; Shanghai Biochip), respectively. The raw data used. All mice were bred and maintained under specific pathogen-free of GeneChip microarray are deposited in the Gene Expression Omnibus conditions. Animal use and experimental procedures were approved by at http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE31273. Prin- the Animal Care and Use Committee of the Chinese Academy of Medical cipal components analysis (PCA) and cluster analyses were analyzed using Sciences. Macrophage line RAW264.7 was obtained from Cell Culture the SBC Analysis System (http://www.ebioservice.com/) provided by Center of Chinese Academy of Medical Sciences. Shanghai Biochip.

Culture of Sca-1+CD1172Lin2 cells RT-PCR MEF were isolated from fetal mice on days 13–15, as previously described Total RNA was extracted using TRIzol (Invitrogen) and then reverse (26). The isolated cells were then cultured in complete DMEM (Life transcribed using a Quantscript RT Kit (Tiangen). PCR products were Technologies) supplemented with 10% FBS (StemCell Technologies), 1 analyzed by electrophoresis on 2% agarose gel and photographed after mM glutamine, 0.1 mM 2-ME, 1% nonessential amino acids (Sigma- ethidium bromide staining. In some experiments, gene expression was Aldrich), and 1000 U/ml mouse LIF (Chemicon). Sca-1+CD1172 cells evaluated using an IQ5 System (Bio-Rad) and a SYBR Green real-time PCR 2 b were sorted using an EasySep CD117-positive selection kit for CD117 kit (Takara). Data were normalized to the reference gene -actin. The cells and an EasySep Sca-1–positive selection kit (StemCell Technologies) primers used are listed in Table I. + + 2 2 for Sca-1 cells. The purity of Sca-1 CD117 Lin cells was confirmed by Western blot analysis flow cytometric analysis. Cells were cultured in complete DMEM. Cell growth and cell cycle were analyzed, as previously described (9). Chro- Briefly, equal quantities of extracts (30 mg total protein) were subjected to mosome analyses were performed by staining cells with conventional SDS-PAGE electrophoresis, and the were transferred to polyvi- Giemsa. nylidene difluoride membranes. After blocking, the membranes were se- quentially incubated with primary and secondary Abs diluted in 3% BSA. Differentiation induction HRP-labeled IgG was used as the secondary Ab for detection. Abs were obtained from Cell Signaling Technology. In each type of differentiation culture, cells were seeded at 2 3 104 cells/ 2 + 2 2 cm . For adipogenic differentiation, Sca-1 CD117 Lin cells were in- Chromatin immunoprecipitation duced for 3 wk in DMEM supplemented with 10% FBS, 1 mM dexa- methasone, 0.5 mM 3-isobutyl-1-methylxanthine, and 0.1 mM ascorbic Chromatin immunoprecipitation (ChIP) was performed using an EZ-ChIP acid. The cells were then fixed in 10% formalin for 10 min and stained kit (Millipore), according to the manufacturer’s instructions. ChIP-grade with fresh oil-red-O solution (Sigma-Aldrich) to visualize lipid droplets in Abs specific for H3K4me3, H3K27me3, and EED were obtained from differentiated cells. For osteogenic differentiation, cells were induced in Millipore; anti-WDR5, anti-SUZ12, and anti-STAT5 were obtained from osteogenic medium (DMEM supplemented with 0.1 mM dexamethasone, Abcam; and anti-ASH2, anti-RbBP5, and anti-MLL1 were obtained from 10 mM b-glycerophosphate, 0.05 mM ascorbic acid, and 10% FBS) for 3 Bethyl. Primers for SOCS3 promoter are near 2720/2581 regions, as wk. The cells were then stained for alkaline phosphatase to assess osteo- follows: 59-CCACGTCTGTCACCGAAGAAC-39,59-GGGGATGCGCG- genic differentiation. ACATTTACCT-39. 1184 MSCs INDUCE NOVEL IL-10-regDC VIA SOCS3

Table I. List of gene-specific primers for PCR

Gene Name Forward Primer and Reverse Primer (59–39) Adipsin CCTGGGAGCGGCTGTATGT; GGGACCCAACGAGGCATT Osteocalcin GCCCAGACCTAGCAGACA; AGGACAGGGAGGATCAAG SOCS3 GTTGAGCGTCAAGACCCAGT; GGGTGGCAAAGAAAAGGAGG STAT3 TGGCACCTTGGATTGAGAGTC; GCAGGAATCGGCTATATTGCT STAT5a CACGTTTCACAGGGCTACCT; CTCTTACACGAGAGGCCTGG STAT5b GACTCCGTCCTTGATACCGA; GGGATCCACTGACTGTCCAT IL-10 ACTCCAGGACCTAGACAGA; GCCAGCCTTAGAATAGAA PU.1 TTACAGGCGTGCAAAATGGAA; AGTAACCAAGTCATCCGATGGA RelB GTTCCAGTGACCTCTCTTCCC; CCAAAGCCGTTCTCCTTAATGTA TCF4 GCCGTCTTCAGTCTACGCTC; TCCCTGTTGTAGTCGGCAGT SpiB GCCACTCCCAAACTGTTCAG; GGGTACGGAGCATAAGCCAAG c-Maf CTGGAGGATCTGTACTGGATGA; CGCACGGACATGGATACCA MafB TGGATGGCGAGCAACTACC; CCAGGTCATCGTGAGTCACA B-actin GAGACCTTCAACACCCCAGCC; AATGTCACGCACGATTTCCC

Small interfering RNA silencing assay RNAiMAX (Invitrogen), according to the instruction. The transfected HPCs were further purified by FACS.

Three pairs of FAM-labeled IL-10 small interfering RNAs (siRNAs) (59- Downloaded from TTCCACUUCUGAAAGAAAGUU-39,59-TTGUCCUGAAAUUCCCAA- In vivo allogeneic delayed-type hypersensitivity assay UGA-39,59-TTCGGUUCGGAAUAGCCUUUA-39), SOCS3 siRNAs (59- TTUCAUGGUCGACCACCACUU-39,59-TTCGAAGCUGACACAUGA- Sca-1+Lin2CD1172 MEF-MSCs, sDCs, or maDCs derived from BALB/c GUU-39,59-TTCGAAGCUGACACAUGAGUU-39), and negative control mice were i.p. injected (3 3 106 cells/mouse) into BALB/c recipient mice (59-UUCUCCGAACGUGUCACGUTT-39) were synthesized by Shanghai on days 26, 24, and 0. C57BL/6 splenocytes were then s.c. injected into GenePharma. siRNAs were transfected into cells with Lipofectamine the dorsal flank of BALB/c recipient mice (1 3 107 cells/mouse) on day 0. http://www.jimmunol.org/

+ 2 2

FIGURE 1. Characteristics of Sca-1 CD117 Lin by guest on October 3, 2021 MEF-MSCs. (A) Spindle-shaped morphology of Sca- 1+CD1172Lin2 MEF-MSCs under an inverted mi- croscope. Scale bar, 100 mm. (B) Immunophenotypic analysis of Sca-1+CD1172Lin2 MEF-MSCs by FACS. The Sca-1+CD1172Lin2 MSCs were prepared and stained using Abs specific for Sca-1, CD9, CD73, CD90, CD29, CD31, CD34, CD44, CD45, CD117, CD14, CD19, Flk-1, H-2Kd, and I-Ad, as described in Materials and Methods. Representative data from one of three independent experiments are shown. (C) Adi- pogenic differentiation of Sca-1+CD1172Lin2 MEF- MSCs. Left, Morphology of induced adipocytes; mid- dle, oil-red-O staining of differentiated adipocytes, red color displaying the lipid droplets in cytoplasm; right, RT-PCR revealed the expression of adipsin.(D) Oste- ogenic differentiation of Sca-1+CD1172Lin2 MEF- MSCs. Left, Morphology of induced osteoblasts; mid- dle, alkaline phosphatase staining of osteocytes; right, RT-PCR revealed the expression of osteocalcin. Scale bars, 100 mm. (E) CFSE-labeled splenocytes (sp) from C57BL/6 mice used as responders and inactivated sp from BALB/c mice used as stimulators were cultured in the presence or absence of Sca-1+CD1172Lin2 MEF-MSC monolayer. On day 3, the responder cells were harvested and analyzed by FACS. The Journal of Immunology 1185

On day 7, recipient mice were challenged by injection of C57BL/6 splenocytes into the right hind footpad (1 3 107 cells/mouse). As a con- trol, the left hind footpad of each recipient mouse was injected with normal saline (NS). Footpad thickness was measured 24 h after challenge using a microcaliper. The extent of swelling was determined by subtracting the baseline thickness of the left footpad from the thickness of the right footpad. For mice injected with MEF-MSCs, spleens were collected and DCs were isolated with DC Isolation Kit (Miltenyi Biotec) for FACS assay. Skin transplantation BALB/c mice were infused i.v. with 2 3 106 sDCs or maDCs on days 26, 24, and 0. On day 0, allogeneic skin transplantation was performed using C57BL/6 mice as donors and BALB/c mice as recipients, as previously described, with minor modifications (3). Briefly, square full-thickness skin grafts (1 cm2) were prepared from the trunk skin of donors. A graft bed (1 cm2) was prepared on the dorsal wall of each recipient mouse. Each graft was fixed to the graft bed and covered with protective surgical tape. Grafts were inspected daily from day 5. Grafts were considered rejected at the time of complete sloughing or when they formed dry scars. Survival is expressed as the mean graft survival time 6 SD. Statistical analysis Downloaded from All experiments were performed at least three times. The results are expressed as the mean 6 SD. Paired t tests were used for intergroup comparison, and p values ,0.05 were considered statistically significant.

Results Isolation of Sca-1+CD1172Lin2 MSCs from MEF http://www.jimmunol.org/ It is well known the MEF are a heterogenous cell population and routinely used in vitro as a feeder layer to support expansion of embryonic stem cells and maintain an undifferentiated state, which implies that some undiscovered cell populations, such as certain primitive stem cells, might exist in MEF for constituting and maintaining the microenvironment ESCs needed. To identify the possible stem cells within this complex population, we isolated Sca-1+CD1172Lin2 cell population from MEF. These cells ex-

hibited a spindle-shaped morphology (Fig. 1A) and a typical S- by guest on October 3, 2021 shaped growth curve (Supplemental Fig. 1A). Cell cycle analysis showed 82.9% were in G0/G1 phase, 10.1% in G2/M phase, and 7.0% in S phase (Supplemental Fig. 1B). The cultured Sca-1+ CD1172Lin2 cells were cytogenetically normal (Supplemental Fig. 1C). Immunophenotypic analysis showed high levels of Sca-1 (99.23%), CD9 (97.9%), CD90 (95.2%), and CD73 (49.5%), and low levels of CD29 (20.45%), CD44 (35.12%), and H-2Kd (6.24%); cells were negative for the lineage-specific markers CD14, CD19, Flk-1, CD31, CD34, CD45, I-Ad, and CD117 (Fig. 1B). Moreover, we demonstrated that these cells could differen- tiate into adipocytes and osteoblasts under specific differentia- tion induction conditions (Fig. 1C, 1D). Sca-1+CD1172Lin2 cells could suppress the proliferation of allogeneic in vitro + 2 2 + 2 2 (Fig. 1E). These results suggest that Sca-1 CD117 Lin cells FIGURE 2. Sca-1 CD117 Lin MEF-MSCs induce the differentiation from MEF are MSCs, as they meet the criteria set by International of regulatory sDCs. (A) Morphology of sDCs (upper) induced from BM- 2 2 Society for Cellular Therapy (27). HPCs by Sca-1+CD117 Lin MEF-MSCs compared with imDCs (middle) and maDCs (lower). Scale bars, 50 mm. (B) Expression of functional Sca-1+CD1172Lin2 MEF-MSCs induce the production of molecules on sDCs, imDCs, and maDCs. Gray lines represent cells stained novel regDCs with isotype-matched control Abs. (C) Phagocytic ability of sDCs and maDCs examined by flow cytometry analysis. Open histograms represent Our previous study has shown that BM-MSCs can promote the the controls. Representative data from one of three independent experi- differentiation of maDCs into regDCs to exert immune regulation ments are shown. (D) Different cytokine secretion of sDCs and imDCs in the presence of LPS or not. The supernatants were collected on day 9, and IL-10 and IL-12 levels were analyzed by ELISA. (E and F) The expression of CD4+ T cells was analyzed by FACS. CD4+ T cells cultured alone were of transcription factors PU.1, RelB, TCF4, SpiB, c-Maf, and MafB was used as control group. Data are expressed as the means 6 SD. (H) PCA detected by qRT-PCR (E) and Western blot analysis (F). sDCs were col- performed on the indicated cell populations according to Genearray data. lected at different time points during the coculture of HPCs and MEF- (I) Footpad swelling of BALB/c mice receiving C57BL/6 splenocytes was MSCs. maDCs and macrophage lineage RAW264.7 (RAW) were used as measured after i.p. infusion of sDCs, maDCs, or NS (n = 6). (J) The control. (G) CD4+ T cells from C57BL/6 mice were cocultured with survival days of skin allografts in the recipient mice infused with sDCs, BALB/c mice derived imDCs, maDCs, or sDCs for 3 d, and then the count maDCs, or NS (n = 6). Data show the means 6 SD. ★p , 0.05, qp , 0.01. 1186 MSCs INDUCE NOVEL IL-10-regDC VIA SOCS3

(9). To investigate whether MEF-MSCs could induce regDC We next investigated whether the sDCs produced in vitro pos- generation from HPCs in vitro, we cultured BM-HPCs on a Sca-1+ sessed immune regulatory capacity in vivo. The delayed-type CD1172Lin2 MEF-MSC monolayer without adding any exoge- hypersensitivity assay showed that footpad swelling of BALB/c nous cytokines at different ratios (HPC:MSC, 50:1, 20:1, 10:1, mice receiving C57BL/6 splenocytes was suppressed by i.p. in- 5:1). MEF-MSCs promoted the generation of CD11b+ cells and fusion of sDCs derived from BALB/c mice (p , 0.05), whereas had the strongest inducing effect at a ratio of 10:1 (Supplemental maDC infusion had no effect (Fig. 2I). The survival time of skin Table I). After 9 d of coculture, nonadherent DC-like cells with allografts was 10 d in control recipients, whereas the time was many long dendrites were frequently observed (Fig. 2A). These prolonged to 24 d when sDCs were infused (p , 0.05, Fig. 2J). Sca-1+CD1172Lin2 MEF-MSC–induced DC-like cells (sDCs) Taken together, these results demonstrate that Sca-1+CD1172 were quite different from maDCs, as they expressed a high level of Lin2 MEF-MSCs induce the generation of a unique regDC pop- myeloid lineage marker CD11b but low levels of functional ulation that can modulate immune responses both in vitro and markers CD11c, Ia, CD80, CD86, and CD40 (Fig. 2B). Also, these in vivo. sDCs exhibited phagocytic capacity (Fig. 2C). Moreover, LPS + 2 2 stimulation did not affect the secretion of IL-10 and IL-12 by Sca-1 CD117 Lin MEF-MSC–derived IL-10 plays an sDCs (Fig. 2D), indicating that sDCs are different from imDCs. To essential role in generation of sDCs further characterize sDCs, we determined the expression of key To identify the key factors responsible for sDC generation, we transcription factors, including PU.1, RelB, TCF4, SpiB, c-Maf, measured the soluble factors in the supernatant at different time and MafB, by quantitative RT-PCR (qRT-PCR) and Western blot points. During coculture of BM-HPCs and Sca-1+CD1172Lin2 analysis. In contrast to macrophages, sDCs expressed high levels MEF-MSCs, higher levels of GM-CSF, IL-4, and anti-inflammatory of transcription factors PU.1, RelB, TCF4, and SpiB, and low factor IL-10 were detected compared with MSCs or BM-HPCs Downloaded from levels of c-Maf and MafB (Fig. 2E, 2F), which function at several alone. Levels of another important anti-inflammatory factor TGF- branch points of DC versus macrophage choice upon differentia- b were similar among the three groups (Fig. 3A–F). To test the tion of HPC into sDC (28). This profile supports the notion effect of IL-10 on sDCs, we added neutralizing Ab against IL-10 that sDCs are DCs and not macrophages. When CD4+ T cells, (NA-IL-10) to the coculture system. Addition of NA-IL-10 in- used as responders, were cocultured with allogeneic imDCs, creased the secretion of IL-12 and TNF-a into the supernatant maDCs, or sDCs, respectively, sDCs had the weakest effect on by sDCs (Fig. 3E, 3F) and altered the phenotype of sDCs by in- http://www.jimmunol.org/ stimulated lymphocyte proliferation. Moreover, sDCs significantly creasing expression of Ia, CD11c, CD80, CD86, and CD40 and suppressed CD4+ T lymphocyte proliferation stimulated by decreasing expression of CD11b (Fig. 3G). Moreover, both PCA maDCs (Fig. 2G). and cluster analysis showed that sDCs treated with NA-IL-10 were The PCA of gene expression microarrays further revealed the different from sDCs (Fig. 2H, Supplemental Fig. 2). Differences differences between sDCs and conventional DCs (Fig. 2H). The in genes related to immune responses are listed in Supplemental differences in immune system processes between sDCs and Table II. These results demonstrate that generation of sDCs is IL-10 imDCs or maDCs were significant, and the genes related to im- dependent. mune responses were shown in Supplemental Table II. These We found that both IL-10–dependent sDCs and MEF-MSCs by guest on October 3, 2021 sDCs downregulated transcription of the costimulatory molecule could secrete IL-10 when they were cocultured (Fig. 4A, 4B). CD86 by 14.7-fold, DC maturation-related genes CCL22 by To explore whether MEF-MSC–derived IL-10 is key to the pro- 244.1-fold, and CCR7 by 118.6-fold. sDCs expressed lower levels duction of sDCs, we used RNA silencing technique to knockdown of cytokines of TNF, CCL, and CXCL families, but higher levels of IL-10 in MEF-MSCs. We also used serum replacement in the IL-10 and SOCS3 than maDCs (Supplemental Fig. 2). experiment to replace serum to exclude the interference of IL-10

FIGURE 3. IL-10 is essential for the production of sDCs. (A–F) Cytokine profiles measured by ELISA in the supernatants. Sca-1+CD1172Lin2 MEF-MSCs alone (MSCs), BM-HPCs without GM-CSF or IL- 4 (HPCs), BM-HPCs with GM-CSF, IL-4 and LPS (HPCs + GM-CSF + IL-4), and cocultured BM-HPCs and Sca-1+CD1172 Lin2 MEF-MSCs with (HPCs/MSCs + NA- IL-10) or without NA-IL-10 (HPCs/MSCs) were cultured for 9 d. The supernatants were collected and analyzed on days 1, 3, 5, 7, and 9 for the levels of TGF-b1(A), IL-10 (B), GM-CSF (C), IL-4 (D), IL-12 (E), and TNF-a (F) by ELISA. Data are expressed as the means 6 SD of triplicate wells. ★p , 0.05, qp , 0.01. (G) After HPCs were cocultured with MEF-MSCs in the presence or absence of NA-IL-10 for 9 d, the induced sDCs were collected, purified, and measured the expression of functional molecules (black lines) by FACS. Gray lines represent cells stained with isotype-matched control Abs. Representative data from one of three independent experiments are shown. The Journal of Immunology 1187

in the serum. siRNAs–IL-10 efficiently downregulated expression of IL-10 in MEF-MSCs (Fig. 4C–E), and silencing of IL-10 in MSCs markedly changed the phenotype of IL-10–dependent sDCs, as indicated by increased expression of Ia, CD11c, and CD80, and decreased expression of CD11b (Fig. 4F). These data suggest that the Sca+Lin2CD1172 MEF-MSC–derived IL-10 is responsible for the generation of sDCs. IL-10–activated SOCS3 is fundamental for sDC generation It is well known that IL-10 can activate the expression of SOCS3, a key physiological regulator of the immune system. Our results showed that expression of SOCS3 increased gradually, but de- creased significantly after NA-IL-10 was added during induction of sDCs (Fig. 5A, 5B). We knocked down expression of SOCS3 in HPCs, and then cocultured the cells with MEF-MSCs. Compared with the control group, the percentage of CD11b+Ialow sDCs dif- ferentiated from siRNA-SOCS3–transfected HPCs was reduced significantly (Fig. 5C, 5D). Our data are consistent with the central role of SOCS3 in sDC generation. IL-10 activates SOCS3 via JAK–STAT pathway and histone Downloaded from modifications As JAK–STAT signal pathway activates the expression of SOCS3, we found that phospho-STAT3 and phospho-STAT5 were grad- ually increased (Fig. 6A), consistent with the activation of SOCS3 during induction of sDCs. The results of protein chip and http://www.jimmunol.org/ Western blot analysis showed that expression of SOCS3 was also downregulated when the activation of JAK2 (phospho-Tyr1007), STAT5A (phospho-Tyr694), STAT5A (phospho-Ser780), and STAT3 (phospho-Tyr705) was decreased by neutralization of IL- 10 (Fig. 6B, 6C). These data suggest that IL-10 regulates ex- pression of SOCS3 via STAT3 and STAT5 activation. To further explore the underlying mechanisms, we added a signal pathway inhibitor to the coculture medium, and found that the level of

SOCS3 decreased significantly when CGP 41251, a phospho- by guest on October 3, 2021 STAT5 inhibitor, was added, whereas the addition of NSC 74859, a phospho-STAT3 inhibitor, showed no effect (Fig. 6D). Furthermore, less STAT5 was bound to the SOCS3 promoter in the presence of NA-IL-10 (Fig. 6E). Thus, IL-10–activated SOCS3 expression depends on STAT5 activation. Epigenetic alterations are often involved in the regulation of gene expression. H3K4me3 and H3K27me3 are mediated by the histone methyltransferase complexes MLL (include MLL1, WDR5, RbBP5, and ASH2L) and PRC2 (include SUZ12 and EED), re- spectively. To verify the hypothesis that IL-10 also regulates SOCS3 via epigenetic mechanisms in addition to JAK–STAT sig- naling, we tested the levels of H3K4me3 and H3K27me3 on the SOCS3 promoter regions by ChIP. During the production of sDCs, the H3K27me3 level at the SOCS3 promoter gradually decreased, whereas NA-IL-10–treated sDCs showed a relative increase in H3K27me3 and a decrease in H3K4me3 (Fig. 6F). These changes were reversed after NA-IL-10 was removed (shown as D7 in Fig. 6F), indicating that the chromatin structure of the SOCS3 is maintained in a repressive state in the presence of NA-IL-10. Our 2 2 FIGURE 4. IL-10 silencing of Sca-1+CD117 Lin MEF-MSCs affects data showed that MLL1 of the MLL complex and SUZ12 of the the production of sDCs. (A) RT-PCR analysis of IL-10 mRNA levels of the PRC2 complex are mainly responsible for the H3K4me3 and respective cells (sDCs and MSCs) in the coculture system on D0, D1, D3, H3K27me3 modifications of SOCS3 in sDCs, respectively (Fig. B D5, and D7. ( ) Flow cytometry analysis of IL-10 levels in sDCs and 6G). Taken together, IL-10 strongly upregulates SOCS3 in sDCs MSCs isolated from the coculture, respectively. Black lines represent for through STAT5 activation and changes of histone methylation. the isotype controls. (C–E) ELISA (C), RT-PCR analysis (D), and flow cytometry analysis (E) of IL-10 levels in MSCs cultured alone with serum replacement instead of serum in medium in different conditions. No, no transfection; NC, negative control of siRNA transfection; siRNAs, siR- MSCs (siRNAs) in the presence of serum replacement. The results shown NAs–IL-10 transfection. (F) Immunophenotype analysis of induced sDCs are representative of three independent experiments and are expressed as the cocultured with NC-transfected MSCs (NC) or siRNA-IL-10–transfected mean 6 SD. qp , 0.01. 1188 MSCs INDUCE NOVEL IL-10-regDC VIA SOCS3 Downloaded from http://www.jimmunol.org/ by guest on October 3, 2021

FIGURE 5. The production of sDCs depends on SOCS3. (A and B) The expression of SOCS3 was measured by qRT-PCR (A) and Western blot (B)on days 1, 3, and 5 during sDC differentiation in the presence or absence of NA-IL-10. (C and D) The transfected HPCs were cocultured with MEF-MSCs for 7 d, and then the DCs generated were assayed by FACS for immunophenotypic analysis. No, not transfected HPCs; NC, NC-transfected HPCs; siRNAs, siRNA-SOCS3–transfected HPCs. Data are expressed as the means 6 SD of triplicate assays. ★p , 0.05.

Sca-1+CD1172Lin2 MEF-MSCs are immunomodulatory by the peripheral blood when compared with control mice (Fig. 7B). stimulating regDCs in vivo Consistent with our finding that MEF-MSCs induce production of One possible mechanism for the immune regulatory properties of regulatory sDCs via SOCS3 in vitro, spleen mononuclear cells MSCs in vivo might be the increased generation of regDCs after expressed more SOCS3 in mice receiving MEF-MSCs (Fig. 7C). In addition, compared with NS controls, DCs expressed higher MSC infusion. To test this hypothesis, we performed the allogeneic levels of CD11b, but lower levels of Ia, CD11c, and CD86, delayed-type hypersensitivity assay in mice. Compared with NS + 2 2 characteristic of regDC (Fig. 7D). Collectively, these results control, infusion of Sca-1 Lin CD117 MEF-MSCs markedly demonstrate that Sca-1+CD117-Lin2 MEF-MSCs modulate the relieved footpad swelling of BALB/c recipient mice (Fig. 7A). We immune response by inducing generation of a regDC population + 2 2 found that Sca-1 Lin CD117 MEF-MSC recipient mice had in vivo, consistent with our observation that Sca-1+Lin2CD1172 higher levels of the anti-inflammatory factors, IL-10 and TGF-b, MEF-MSCs induce IL-10–dependent differentiation of regulatory and lower levels of proinflammatory factors IL-12 and TNF-a in sDCs in vitro (Fig. 8). The Journal of Immunology 1189 Downloaded from http://www.jimmunol.org/ by guest on October 3, 2021

FIGURE 6. IL-10 regulates the expression of SOCS3 through JAK–STAT signal pathway and histone H3 methylation modification. (A) The activation of STAT3 and STAT5 on days 1, 3, and 5 during sDC differentiation in the presence or absence of NA-IL-10 by Western blot. (B) Results of ProteinChip analysis show the different phosphorylation levels of JAK–STAT signaling pathway between sDCs in the presence or absence of NA-IL-10. (C) The induced sDCs in the presence or absence of NA-IL-10 were collected, isolated, and analyzed by Western blot for several signal molecules. Representative data from one of three independent experiments are shown. (D) Expression of SOCS3 in sDCs was measured by Western blot with (+) or without (2) a p-STAT3 inhibitor (NSC 74859) or a p-STAT5 inhibitor (CGP 41251). (E) The recruitment of STAT5 at the promoter region of SOCS3 in sDCs was tested by ChIP assay. (F) ChIP was performed to determine the histone H3 methylation status at the promoter regions of SOCS3 in sDCs in the presence or absence of NA- IL-10 on days 1, 3, and 5. On day 5, all of the culture media were replaced to remove NA-IL-10, and the sDCs were cultured for another 2 d (day 7). The values were normalized to the input for each sample. (G) ChIP was performed on untreated or NA-IL-10–treated sDCs on day 5 using Abs specific for the core components of the MLL and PRC2 complexes to examine the recruitment of these components to the promoter regions of SOCS3. The values were normalized to the input for each sample. The results shown are representatives of three independent experiments and are expressed as mean 6 SD. ★p , 0.05.

Discussion Sca+Lin2CD1172 MEF-MSC–derived IL-10 is responsible for MEF prepared from nonhematopoietic tissues are often used as sDC generation by upregulating SOCS3 through the JAK–STAT feeder layers to maintain ESCs in an undifferentiated state. MEF signaling pathway and by histone H3 modification (Fig. 8). Im- are a heterogeneous cell population. The isolation and identifica- portantly, Sca-1+CD1172Lin2 MEF-MSCs also can modulate the tion of MSCs from the MEF population have not been previously inflammatory response by increasing regDC population in vivo. 2 2 reported. In this study, we isolated Sca-1+CD117 Lin MSCs Our previous studies have shown that adult BM-MSCs regulate from MEF and evaluated their multilineage differentiation po- the differentiation, maturation, and function of human monocyte- tential and immune regulatory function. We demonstrated that derived DCs (6, 9). In the current study, we found that Sca-1+ the MEF-MSCs induce BM-HPCs to differentiate into a distinct CD1172Lin2 MEF-MSCs, from nonhematopoietic tissue, can regulatory sDC subset in the absence of exogenous cytokines. induce BM-HPCs to differentiate into regulatory sDCs in the 1190 MSCs INDUCE NOVEL IL-10-regDC VIA SOCS3

FIGURE 7. Infusion of Sca-1+ CD1172Lin2 MEF-MSCs can alleviate the immune reaction by allogeneic delayed-type hypersensitivity assay. (A) Footpad swelling of BALB/c mice receiving C57BL/6 splenocytes was measured after i.p. infusion of NS so- lution or MEF-MSCs (n = 6). (B) The levels of cytokines in the peripheral blood of recipient BALB/c mice were C tested by ELISA. ( ) Gene expression Downloaded from analysis of spleen mononuclear cells from recipient mice. ★p , 0.05, qp , 0.01. (D) Phenotypes of DCs isolated from spleens of recipient mice were analyzed by FACS. Representative data from one of three independent experi- ments are shown. http://www.jimmunol.org/ by guest on October 3, 2021

absence of an exogenous cytokine mixture. Upon differentiation many signaling pathways, such as JAK–STAT and NF-kB, is of HPCs into sDCs in the coculture, we found a marked increase different among these DC subsets. The microarray studies provide in the expression of PU.1. Also, the expression of RelB, TCF4, a comprehensive evaluation of the differences between the newly and SpiB was detected at high levels, and c-Maf and MafB at low identified sDCs and imDCs or maDCs. levels. The results are in accordance with a recent report showing We have shown that the contact effect, especially the Notch that PU.1 is a key selector gene that functions during macrophage pathway, is necessary for the differentiation of regDCs (9), con- versus DC choice. High expression of PU.1 is required to induce sistent with related reports (33, 34). However, transwell assay or a DC fate in monocytes and to antagonize the macrophage- Notch pathway blockage could not thoroughly inhibit the differ- inducing transcription factors, c-Maf and MafB (28). The sDCs, entiation of sDCs (data not shown), suggesting there are key which express CD11c and Ia at low levels, are phenotypically cytokines, besides cell-cell contact effect, also responsible for the different from the CD11chighIahigh regDCs induced by murine production of sDCs. Many results have shown that PGE2 from GM-CSF, murine IL-10, and human TGF-b (29). Because sDCs MEF or MSCs is a mechanism for the suppression of immune express low levels of CD80 and CD40 and are derived from BM- responses (35–37). Consistently, we also detected a high level of HPCs, they are also different from differentiated DCs and MSCs- PGE2 in the coculture medium (data not shown), suggesting the DCs, both of which are derived from maDCs (9, 30). Recently, possible roles of PGE2. However, Shiraishi et al. (37) pointed out it has been reported that CD11bhighIalow regDCs can be induced there were other suppressive factors besides PGE2. So, in our by splenic stroma (19). The present study defines the effects of work, we found IL-10 was another suppressive factor besides Sca-1+CD1172Lin2 MEF-MSCs on DC differentiation from BM- PGE2 for the immune suppression. IL-10 is an anti-inflammatory HPCs. To further characterize these sDCs, we detected the ex- cytokine that contributes to the downregulation of T cell prolif- pression of CD103 and CX3CR1 in sDCs (data not shown), and eration and other effector responses by suppressing the production the results showed sDCs were characteristic of CD11bhighCD1032 of proinflammatory cytokines known to be transcriptionally con- CX3CR1low DCs, a different subpopulation from previously de- trolled by NF-kB (38, 39). SOCS3 is a key physiological regulator scribed (31, 32). Moreover, these sDCs are very different from of the immune system, and has been shown to be induced by IL-10 imDCs and maDCs, not only in phenotype and function, but also (40). Moreover, accumulating data suggest that SOCS3 may in gene expression profile. We found the expression of genes in- suppress inflammatory responses in many diseases by interfering volved in immune processes, such as IL, CXC, CCL cytokine with proinflammatory factor-mediated signaling (41–44). MSC- families, and genes encoding transcription factors involved in derived IL-10 induces sDC differentiation from HPCs, and sDCs The Journal of Immunology 1191 Downloaded from http://www.jimmunol.org/

FIGURE 8. A proposed schema of the mechanism of action of the Sca-1+CD1172Lin2 MEF-MSC–mediated immune regulation of sDCs. Sca-1+ CD1172Lin2 MEF-MSCs play important roles in the generation and function of sDCs, a regDC subset distinct from both imDCs and maDCs. In the presence of IL-10, IL-10Rs on sDCs are activated and further phosphorylated through JAK2–STAT5 signaling pathway to upregulate SOCS3, which blocks the activation of NF-kB p65, thereby decreasing the secretion of IL-12. Moreover, SOCS3 expression is also regulated by the stable reciprocal changes in the levels of H3K4me3 and H3K27me3 mediated by the MLL and PRC2 complexes, respectively. MLL1 (red) of the MLL complex and SUZ12 (red) of the PRC2 complex are mainly responsible for the histone H3 modifications of the SOCS3 promoter in sDCs. by guest on October 3, 2021 themselves also secrete IL-10, which may further augment im- not yet been elucidated, our results shed new light on the role of mune suppression. Cao’s group (19) has pointed out that IL-10 MSCs in treating immune-related diseases. secreted by endothelial splenic stroma plays a critical role in the generation of the regulatory CD11bhighIalow DCs. Disclosures In this study, we demonstrate that epigenetic mechanisms and The authors have no financial conflicts of interest. transcriptional signaling pathways are both involved during the IL- 10–dependent production of regulatory sDCs (Fig. 8). Neutrali- References zation of IL-10 causes decreased H3K4me3 and increased 1. Prockop, D. J. 1997. Marrow stromal cells as stem cells for nonhematopoietic H3K27me3 at the promoter region of SOCS3 mediated by MLL1 tissues. Science 276: 71–74. and SUZ12, respectively, resulting in the downregulation of 2. Pittenger, M. F., A. M. Mackay, S. C. Beck, R. K. Jaiswal, R. Douglas, J. D. Mosca, M. A. Moorman, D. W. Simonetti, S. Craig, and D. R. Marshak. SOCS3 expression. Thus, these data support our hypothesis that 1999. Multilineage potential of adult human mesenchymal stem cells. Science SOCS3 expression is coregulated by JAK–STAT5 signaling and 284: 143–147. histone modifications. This is underscored by the observations that 3. Bartholomew, A., C. Sturgeon, M. Siatskas, K. Ferrer, K. McIntosh, S. Patil, W. Hardy, S. Devine, D. Ucker, R. Deans, et al. 2002. Mesenchymal stem cells STAT5 can bind the pSRE (272/264) region in the SOCS3 pro- suppress lymphocyte proliferation in vitro and prolong skin graft survival moter for transcriptional activation (45) and that the epigenetic in vivo. Exp. Hematol. 30: 42–48. 2 2 4. Uccelli, A., L. Moretta, and V. Pistoia. 2008. Mesenchymal stem cells in health modification mainly occurs at the sites near 720/ 581 regions and disease. Nat. Rev. Immunol. 8: 726–736. that we tested. 5. Zhao, R. C., L. Liao, and Q. Han. 2004. Mechanisms of and perspectives on the MSCs are a promising source of cells for tissue repair and re- mesenchymal stem cell in immunotherapy. J. Lab. Clin. Med. 143: 284–291. 6. Zhang, W., W. Ge, C. Li, S. You, L. Liao, Q. Han, W. Deng, and R. C. Zhao. generation (10, 46–48) and also dampen the inflammatory process 2004. Effects of mesenchymal stem cells on differentiation, maturation, and (49, 50). Sca-1+CD1172Lin2 MEF-MSCs can induce BM-HPCs function of human monocyte-derived dendritic cells. Stem Cells Dev. 13: 263– 271. to differentiate into regulatory sDCs to play immune regulatory 7. Ren, G., L. Zhang, X. Zhao, G. Xu, Y. Zhang, A. I. Roberts, R. C. Zhao, and functions; whether other adult MSCs also have this effect has not Y. Shi. 2008. Mesenchymal stem cell-mediated immunosuppression occurs via been mentioned in the current study but deserves further explo- concerted action of chemokines and nitric oxide. Cell Stem Cell 2: 141–150. 8. Chen, L., W. Zhang, H. Yue, Q. Han, B. Chen, M. Shi, J. Li, B. Li, S. You, Y. Shi, ration. Resident and remote MSCs may be mobilized and recruited and R. C. Zhao. 2007. Effects of human mesenchymal stem cells on the dif- to the injured sites to play a dual role, that is, differentiation into ferentiation of dendritic cells from CD34+ cells. Stem Cells Dev. 16: 719–731. 9. Zhang, B., R. Liu, D. Shi, X. Liu, Y. Chen, X. Dou, X. Zhu, C. Lu, W. Liang, special damaged cells and interaction with immune cells to reduce L. Liao, et al. 2009. Mesenchymal stem cells induce mature dendritic cells into inflammation, thus providing a normal microenvironment for tis- a novel Jagged-2-dependent regulatory dendritic cell population. Blood 113: 46– 57. sue regeneration and wound repair. Although the precise mecha- 10. Le Blanc, K., I. Rasmusson, B. Sundberg, C. Go¨therstro¨m, M. Hassan, nisms underlying MSC-induced generation of regDC in vivo have M. Uzunel, and O. Ringde´n. 2004. Treatment of severe acute graft-versus-host 1192 MSCs INDUCE NOVEL IL-10-regDC VIA SOCS3

disease with third party haploidentical mesenchymal stem cells. Lancet 363: propria dendritic cell subsets have different origin and functions. Immunity 31: 1439–1441. 502–512. 11. Rafei, M., P. M. Campeau, A. Aguilar-Mahecha, M. Buchanan, P. Williams, 33. Li, Y. P., S. Paczesny, E. Lauret, S. Poirault, P. Bordigoni, F. Mekhloufi, E. Birman, S. Yuan, Y. K. Young, M. N. Boivin, K. Forner, et al. 2009. Mes- O. Hequet, Y. Bertrand, J. P. Ou-Yang, J. F. Stoltz, et al. 2008. Human mesen- enchymal stromal cells ameliorate experimental autoimmune encephalomyelitis chymal stem cells license adult CD34+ hemopoietic progenitor cells to differ- by inhibiting CD4 Th17 T cells in a CC chemokine ligand 2-dependent manner. entiate into regulatory dendritic cells through activation of the Notch pathway. J. J. Immunol. 182: 5994–6002. Immunol. 180: 1598–1608. 12. Deng, W., Q. Han, L. Liao, S. You, H. Deng, and R. C. Zhao. 2005. Effects of 34. Lewis, K. L., M. L. Caton, M. Bogunovic, M. Greter, L. T. Grajkowska, D. Ng, allogeneic bone marrow-derived mesenchymal stem cells on T and A. Klinakis, I. F. Charo, S. Jung, J. L. Gommerman, et al. 2011. Notch2 receptor B lymphocytes from BXSB mice. DNA Cell Biol. 24: 458–463. signaling controls functional differentiation of dendritic cells in the spleen and 13. Cook, C. H., A. A. Bickerstaff, J. J. Wang, T. Nadasdy, P. Della Pelle, intestine. Immunity 35: 780–791. R. B. Colvin, and C. G. Orosz. 2008. Spontaneous renal allograft acceptance 35. Ne´meth, K., A. Leelahavanichkul, P. S. Yuen, B. Mayer, A. Parmelee, K. Doi, associated with “regulatory” dendritic cells and IDO. J. Immunol. 180: 3103– P. G. Robey, K. Leelahavanichkul, B. H. Koller, J. M. Brown, et al. 2009. Bone 3112. marrow stromal cells attenuate sepsis via prostaglandin E(2)-dependent 14. Novak, N., and T. Bieber. 2008. 2. Dendritic cells as regulators of immunity and reprogramming of host macrophages to increase their interleukin-10 produc- tolerance. J. Allergy Clin. Immunol. 121: S370-S374; quiz S413. tion. Nat. Med. 15: 42–49. 15. Li, M., X. Zhang, X. Zheng, D. Lian, Z. X. Zhang, H. Sun, M. Suzuki, 36. Maggini, J., G. Mirkin, I. Bognanni, J. Holmberg, I. M. Piazzo´n, C. Vladau, X. Huang, X. Xia, et al. 2008. Tolerogenic dendritic cells transferring I. Nepomnaschy, H. Costa, C. Can˜ones, S. Raiden, M. Vermeulen, and hyporesponsiveness and synergizing T regulatory cells in transplant tolerance. J. R. Geffner. 2010. Mouse bone marrow-derived mesenchymal stromal cells Int. Immunol. 20: 285–293. turn activated macrophages into a regulatory-like profile. PLoS One 5: e9252. 16. Ureta, G., F. Osorio, J. Morales, M. Rosemblatt, M. R. Bono, and J. A. Fierro. 37. Shiraishi, H., H. Yoshida, K. Saeki, Y. Miura, S. Watanabe, T. Ishizaki, 2007. Generation of dendritic cells with regulatory properties. Transplant. Proc. M. Hashimoto, G. Takaesu, T. Kobayashi, and A. Yoshimura. 2008. Prosta- 39: 633–637. glandin E2 is a major soluble factor produced by stromal cells for preventing 17. Steinman, R. M., D. Hawiger, and M. C. Nussenzweig. 2003. Tolerogenic inflammatory cytokine production from dendritic cells. Int. Immunol. 20: 1219– dendritic cells. Annu. Rev. Immunol. 21: 685–711. 1229. 38. Schottelius, A. J., M. W. Mayo, R. B. Sartor, and A. S. Baldwin, Jr. 1999.

18. Wakkach, A., N. Fournier, V. Brun, J. P. Breittmayer, F. Cottrez, and H. Groux. Downloaded from 2003. Characterization of dendritic cells that induce tolerance and T regulatory 1 Interleukin-10 signaling blocks inhibitor of kappaB kinase activity and nuclear cell differentiation in vivo. Immunity 18: 605–617. factor kappaB DNA binding. J. Biol. Chem. 274: 31868–31874. 19. Tang, H., Z. Guo, M. Zhang, J. Wang, G. Chen, and X. Cao. 2006. Endothelial 39. Folch-Puy, E., S. Granell, J. C. Dagorn, J. L. Iovanna, and D. Closa. 2006. stroma programs hematopoietic stem cells to differentiate into regulatory den- Pancreatitis-associated protein I suppresses NF-kappa B activation through a JAK/STAT-mediated mechanism in epithelial cells. J. Immunol. 176: 3774– dritic cells through IL-10. Blood 108: 1189–1197. 3779. 20. Dou, Y., T. A. Milne, A. J. Tackett, E. R. Smith, A. Fukuda, J. Wysocka, 40. Cassatella, M. A., S. Gasperini, C. Bovolenta, F. Calzetti, M. Vollebregt, C. D. Allis, B. T. Chait, J. L. Hess, and R. G. Roeder. 2005. Physical association P. Scapini, M. Marchi, R. Suzuki, A. Suzuki, and A. Yoshimura. 1999. and coordinate function of the H3 K4 methyltransferase MLL1 and the H4 K16

Interleukin-10 (IL-10) selectively enhances CIS3/SOCS3 mRNA expression in http://www.jimmunol.org/ acetyltransferase MOF. Cell 121: 873–885. human neutrophils: evidence for an IL-10-induced pathway that is independent 21. Cao, R., and Y. Zhang. 2004. The functions of E(Z)/EZH2-mediated methylation of STAT protein activation. Blood 94: 2880–2889. of lysine 27 in histone H3. Curr. Opin. Genet. Dev. 14: 155–164. 41. Kubo, M., T. Hanada, and A. Yoshimura. 2003. Suppressors of cytokine sig- 22. Reik, W. 2007. Stability and flexibility of epigenetic gene regulation in mam- naling and immunity. Nat. Immunol. 4: 1169–1176. malian development. Nature 447: 425–432. 42. Yasukawa, H., M. Ohishi, H. Mori, M. Murakami, T. Chinen, D. Aki, T. Hanada, 23. Ballestar, E., M. Esteller, and B. C. Richardson. 2006. The epigenetic face of K. Takeda, S. Akira, M. Hoshijima, et al. 2003. IL-6 induces an anti- systemic lupus erythematosus. J. Immunol. 176: 7143–7147. inflammatory response in the absence of SOCS3 in macrophages. Nat. Immu- 24. Wen, H., Y. Dou, C. M. Hogaboam, and S. L. Kunkel. 2008. Epigenetic regu- nol. 4: 551–556. lation of dendritic cell-derived interleukin-12 facilitates immunosuppression 43. Suzuki, A., T. Hanada, K. Mitsuyama, T. Yoshida, S. Kamizono, T. Hoshino, after a severe innate immune response. Blood 111: 1797–1804. M. Kubo, A. Yamashita, M. Okabe, K. Takeda, et al. 2001. CIS3/SOCS3/SSI3 25. Blu¨ml, S., G. Zupkovitz, S. Kirchberger, M. Seyerl, V. N. Bochkov, plays a negative regulatory role in STAT3 activation and intestinal inflammation. K. Stuhlmeier, O. Majdic, G. J. Zlabinger, C. Seiser, and J. Sto¨ckl. 2009. Epi- J. Exp. Med. 193: 471–481. by guest on October 3, 2021 genetic regulation of dendritic cell differentiation and function by oxidized 44. Shouda, T., T. Yoshida, T. Hanada, T. Wakioka, M. Oishi, K. Miyoshi, phospholipids. Blood 114: 5481–5489. S. Komiya, K. Kosai, Y. Hanakawa, K. Hashimoto, et al. 2001. Induction of the 26. Cheng, J., A. Dutra, A. Takesono, L. Garrett-Beal, and P. L. Schwartzberg. 2004. cytokine signal regulator SOCS3/CIS3 as a therapeutic strategy for treating in- Improved generation of C57BL/6J mouse embryonic stem cells in a defined flammatory arthritis. J. Clin. Invest. 108: 1781–1788. serum-free media. Genesis 39: 100–104. 45. Barclay, J. L., S. T. Anderson, M. J. Waters, and J. D. Curlewis. 2007. Regulation 27. Dominici, M., K. Le Blanc, I. Mueller, I. Slaper-Cortenbach, F. Marini, of suppressor of cytokine signaling 3 (SOC3) by growth hormone in pro-B cells. D. Krause, R. Deans, A. Keating, Dj. Prockop, and E. Horwitz. 2006. Minimal Mol. Endocrinol. 21: 2503–2515. criteria for defining multipotent mesenchymal stromal cells: the International 46. Constantin, G., S. Marconi, B. Rossi, S. Angiari, L. Calderan, E. Anghileri, Society for Cellular Therapy position statement. Cytotherapy 8: 315–317. B. Gini, S. D. Bach, M. Martinello, F. Bifari, et al. 2009. Adipose-derived 28. Geissmann, F., M. G. Manz, S. Jung, M. H. Sieweke, M. Merad, and K. Ley. mesenchymal stem cells ameliorate chronic experimental autoimmune enceph- 2010. Development of monocytes, macrophages, and dendritic cells. Science alomyelitis. Stem Cells 27: 2624–2635. 327: 656–661. 47. Lee, J. W., X. Fang, N. Gupta, V. Serikov, and M. A. Matthay. 2009. Allogeneic 29. Sato, K., N. Yamashita, N. Yamashita, M. Baba, and T. Matsuyama. 2003. human mesenchymal stem cells for treatment of E. coli endotoxin-induced acute Regulatory dendritic cells protect mice from murine acute graft-versus-host lung injury in the ex vivo perfused human lung. Proc. Natl. Acad. Sci. USA 106: disease and leukemia relapse. Immunity 18: 367–379. 16357–16362. 30. Zhang, M., H. Tang, Z. Guo, H. An, X. Zhu, W. Song, J. Guo, X. Huang, 48. Zappia, E., S. Casazza, E. Pedemonte, F. Benvenuto, I. Bonanni, E. Gerdoni, T. Chen, J. Wang, and X. Cao. 2004. Splenic stroma drives mature dendritic cells D. Giunti, A. Ceravolo, F. Cazzanti, F. Frassoni, et al. 2005. Mesenchymal stem to differentiate into regulatory dendritic cells. Nat. Immunol. 5: 1124–1133. cells ameliorate experimental autoimmune encephalomyelitis inducing T-cell 31. Auffray, C., D. K. Fogg, E. Narni-Mancinelli, B. Senechal, C. Trouillet, anergy. Blood 106: 1755–1761. N. Saederup, J. Leemput, K. Bigot, L. Campisi, M. Abitbol, et al. 2009. 49. Shi, Y., G. Hu, J. Su, W. Li, Q. Chen, P. Shou, C. Xu, X. Chen, Y. Huang, Z. Zhu, CX3CR1+ CD115+ CD135+ common macrophage/DC precursors and the role et al. 2010. Mesenchymal stem cells: a new strategy for immunosuppression and of CX3CR1 in their response to inflammation. J. Exp. Med. 206: 595–606. tissue repair. Cell Res. 20: 510–518. 32. Varol, C., A. Vallon-Eberhard, E. Elinav, T. Aychek, Y. Shapira, H. Luche, 50. Singer, N. G., and A. I. Caplan. 2011. Mesenchymal stem cells: mechanisms of H. J. Fehling, W. D. Hardt, G. Shakhar, and S. Jung. 2009. Intestinal lamina inflammation. Annu. Rev. Pathol. 6: 457–478.

Supplemental figure 1. Growth pattern, cell cycle and karyotype analysis of

Sca-1+CD117-Lin- MEF-MSCs

(A) Growth pattern of Sca-1+CD117-Lin- MSCs. Sca-1+CD117-Lin- MSCs were continuously cultured at 5×103 cells/well in a 24-well plate for 9 days. (B) Cell cycle analysis of Sca-1+CD117-Lin- MSCs. (C) Rearranged of Sca-1+CD117-Lin-

MSCs after conventional Giemsa staining.

Supplemental figure 2. Cluster analysis of sDC, imDC and maDC

Cluster analysis of immune related genes was based on the GeneChip data of the indicated cell populations. The color scale for relative expression values, obtained after log10 transformation and median centering of the values of cell samples for each gene, is shown in the top left panel.

Supplemental table 1. FACS analysis of DCs differentiated from HPCs cocultured with Sca-1+CD117-Lin- MEF-MSCs at different ratios.

HPC MSC MSC/HPC CD11b CD11c Ia CD40 CD80 CD86

2×106/well 0/well 0 16.0% 82.0% 85.9% 32.4% 40.2% 38.7%

2×106/well 4×104/well 1/50 73.1% 34.2% 15.0% 15.0% 21.7% 16.7%

2×106/well 1×105/well 1/20 74.2% 30.2% 15.7% 11.5% 23.0% 18.8%

2×106/well 2×105/well 1/10 90.9% 23.4% 11.2% 9.4% 9.0% 10.7%

2×106/well 4×105/well 1/5 90.7% 27.1% 12.3% 12.2% 4.3% 17.2%

Supplemental table 2. Differential expression of immune response-related genes between sDC (C) and imDC (A), sDC (C) and maDC (B), sDC (C) and sDC with

NA-IL-10 (D).

Fold_chan Regulatio Fold_chan Regulation Symbol GeneId Symbol GeneId ge(C vs A) n(C vs A) ge(C vs A) (C vs A)

Cd276 102657 6.7562 up Inppl1 16332 6.8217 up

Il1rl2 107527 6.9318 down Lbp 16803 8.4184 down

Cenpf 108000 10.9802 down Ltb 16994 13.6651 down

Olr1 108078 82.338 down Tlr7 170743 6.4108 up

Card11 108723 22.7313 down Cd180 17079 8.3329 up

Rab27a 11891 13.2127 down Il1rl1 17082 35.5839 up

Atp7a 11977 7.1571 up Ly86 17084 24.2756 up

Bcl6 12053 10.9731 up Ly96 17087 6.5641 up

Serping1 12258 88.4471 up Masp1 17174 38.7508 up

C1qa 12259 39.5129 up Cxcl9 17329 7.3743 down

C1qb 12260 9.444 up Clec4d 17474 19.9947 up

C1qc 12262 114.585 up Mpo 17523 174.0587 down

C2 12263 8.514 up Slc11a1 18173 14.1536 up

Ciita 12265 63.1898 down P2rx7 18439 25.3102 up

C3 12266 8.3854 down Enpp1 18605 15.7529 up

Cacnb3 12297 13.2766 down Enpp2 18606 11.1313 down

Cd14 12475 13.4459 up Prg2 19074 1276.4233 down

Cd1d2 12480 15.3902 down Ptx3 19288 181.8982 up

Cd24a 12484 48.2497 down Rbp4 19662 9.8158 down

Cd28 12487 9.9412 down Ccl1 20290 16.9474 down

Cd8b1 12526 10.7771 up Ccl11 20292 22.4273 up

Cfh 12628 35.7662 up Ccl12 20293 34.5558 up

Ccr7 12775 36.3329 down Ccl22 20299 244.1463 down

Cplx2 12890 6.9968 up Ccl7 20306 9.2685 up

Csf3 12985 28.4376 up Ccl8 20307 46.7509 up Ctsg 13035 103.2969 down Cxcl12 20315 27.8604 up

Cx3cr1 13051 45.954 up Serpina3g 20715 39.5778 down

Eomes 13813 6.7639 down Spn 20737 32.5105 down

Ereg 13874 10.1744 up Il20rb 213208 6.4439 down

4732429D AF251705 140497 27.7497 up 217305 16.5488 up 16Rik

Colec12 140792 92.0878 up Cd27 21940 6.6975 down

Fcgr1 14129 7.4868 up Tnfsf12 21944 9.4406 up

Fcnb 14134 81.8801 down Pglyrp1 21946 377.5163 down

Tlr3 142980 11.9368 up Tnfsf9 21950 21.3929 up

Cxcl1 14825 53.0168 up Tnfsf4 22164 24.2549 down

H2-Aa 14960 183.0994 down Igh-2 238447 13.1247 down

H2-Ab1 14961 57.7031 down Oasl2 23962 6.3058 up

Cfb 14962 6.3665 up Klhl6 239743 38.8052 down

H2-Bl 14963 6.573 up Sh2d1b1 26904 56.1905 down

H2-Ea 14968 169.5772 down Exo1 26909 7.7686 down

H2-Eb1 14969 121.0949 down Nlrx1 270151 7.1958 down

H2-DMb1 14999 20.857 down Ubash3a 328795 53.5022 down

H2-DMb2 15000 20.9606 down Pram1 378460 109.0825 down

H2-Q10 15007 6.6795 up Raet1e 379043 7.5993 up

H2-Q8 15019 7.7911 up Cd300c 387565 6.7706 down

Mr1 15064 7.3802 up Ela2 50701 417.9266 down

Ido1 15930 27.2473 down C1s 50908 36.2391 up

Cxcl10 15945 6.4651 down C1r 50909 10.8372 up

Igj 16069 14.3135 down Il27ra 50931 7.2005 down

Cd74 16149 80.4098 down Prg3 53856 81.2211 down

Il12a 16159 9.9655 down Cadm1 54725 29.0948 up

Il18 16173 27.8679 up Samhd1 56045 6.4322 up

Il18r1 16182 11.0308 down Cxcl11 56066 7.9947 down

Il2 16183 10.0448 up Ccl24 56221 32.1098 down

Il4 16189 10.168 down Clec4e 56619 12.5558 up Il7r 16197 12.9362 up Clec7a 56644 9.0744 down

Il9 16198 24.7672 down Ptms 69202 7.9481 up

Lat2 56743 6.811 up Klrb1b 80782 14.792 down

Cxcl14 57266 28.2528 up Tac4 93670 87.1126 down

Pglyrp2 57757 12.7487 down Il10 16153 6.7794 down

Rnf125 67664 49.8432 down

Fold Fold Regulatio Regulatio Symbol GeneId change (C Symbol GeneId change (C n (C vs B) n (C vs B) vs B) vs B)

Cd276 102657 8.9716 up Cd40 21939 27.9026 down

Olr1 108078 183.5113 down Pglyrp1 21946 9.7345 down

Card11 108723 14.6478 down Tnfsf8 21949 21.2732 down

Acp5 11433 20.1136 down Tnfsf9 21950 5.2171 up

Bcl6 12053 5.2809 up Tnfsf4 22164 98.9138 down

Serping1 12258 32.1423 up Clec12a 232413 10.3578 up

C1qa 12259 21.2268 up Igh-2 238447 20.9937 down

C1qb 12260 6.7511 up Klhl6 239743 13.4459 down

C1qc 12262 19.0413 up Malt1 240354 7.3634 down

Ciita 12265 51.7696 down Ccl19 24047 8.0216 down

Cacnb3 12297 36.8157 down Sigirr 24058 5.1777 down

Cd1d1 12479 9.4672 down Lax1 240754 5.1563 down

Cd1d2 12480 40.5076 down Clec4a2 26888 8.7879 up

Cd24a 12484 15.8176 down Sh2d1b1 26904 16.8487 down

Cd28 12487 20.9549 down Klrk1 27007 14.0557 down

Cd247 12503 24.5625 down Tnfsf15 326623 12.3792 down

Cd7 12516 10.3713 down Ubash3a 328795 7.6884 down

Cd86 12524 14.6694 down Cxcl3 330122 14.2525 down

Cd8b1 12526 13.3663 up Pram1 378460 5.8718 down

Cfh 12628 93.0202 up Raet1e 379043 6.8837 up

Cfi 12630 9.8508 down Ela2 50701 6.1856 down Ccr2 12772 6.5648 up Il27ra 50931 7.668 down

Ccr7 12775 118.5597 down Prg3 53856 18.6545 down

Cplx2 12890 8.3945 down Cadm1 54725 27.6602 up

Csf2 12981 214.3277 down Cxcl11 56066 92.7217 down

Csf3 12985 7.2398 down Ccl24 56221 58.1986 down

Ctsg 13035 26.2162 down Clec7a 56644 15.6708 down

Cx3cr1 13051 7.4361 up Pglyrp2 57757 54.7771 down

Cd55 13136 10.7006 down Rsad2 58185 18.0445 down

Daf2 13137 14.1241 down Rnf125 67664 38.1092 down

Eomes 13813 27.8243 down Susd2 71733 7.7161 down

AF251705 140497 5.3829 up Tnfrsf13c 72049 8.9421 down

Colec12 140792 30.182 up Tbkbp1 73174 5.4557 up

Fcer1a 14125 91.0838 down Nfam1 74039 8.4514 up

H2-Aa 14960 134.1204 down Rnf19b 75234 10.5875 down

H2-Ab1 14961 23.6483 down Trat1 77647 6.9578 down

Cfb 14962 13.8694 down Klrb1b 80782 71.2111 down

H2-D1 14964 5.2417 down Il23a 83430 44.4946 down

H2-Ea 14968 130.2667 down Mfsd6 98682 5.6213 up

H2-Eb1 14969 96.2583 down Mpo 17523 18.7261 down

H2-DMb1 14999 25.8959 down Mx1 17857 8.0794 down

H2-DMb2 15000 16.0802 down Mx2 17858 7.8392 down

H2-Ob 15002 5.836 down Nfkb2 18034 5.2593 down

H2-Q1 15006 25.6875 down Enpp1 18605 19.7449 up

Hc 15139 8.7601 down Enpp2 18606 496.2508 down

Ido1 15930 100.9183 down Prg2 19074 146.0881 down

Irgm1 15944 6.5 down Procr 19124 9.1462 down

Cxcl10 15945 64.0637 down Ripk2 192656 5.401 down

Ifng 15978 158.6541 down Ptx3 19288 83.7397 up

Cd74 16149 94.9037 down Raet1a 19368 8.6523 up

Il12a 16159 43.9226 down Rbp4 19662 107.5116 down

Il18 16173 13.395 up Ccl1 20290 156.479 down Il1a 16175 507.1441 down Ccl11 20292 22.3422 up

Il1b 16176 6.2033 down Ccl20 20297 11.0368 down

Il18r1 16182 12.233 down Ccl22 20299 2468.976 down

Il2 16183 9.0586 up Ccl5 20304 14.3582 down

Il4 16189 33.4534 down Cxcl5 20311 6.3077 down

Il6 16193 86.1418 down Cx3cl1 20312 67.484 down

Il9 16198 400.4299 down Cxcl12 20315 68.0731 up

Lag3 16768 12.7406 down Serpina3g 20715 49.8128 down

Lat 16797 12.9487 down Spn 20737 5.2747 down

Lbp 16803 5.0759 up Btla 208154 7.2999 down

Lck 16818 6.0577 down Stat5a 20850 5.6918 down

Xcl1 16963 21.8857 down Il20rb 213208 12.9651 down

4732429D Klrb1c 17059 6.1365 down 217305 8.5562 up 16Rik

Tlr8 170744 9.2409 up Tgtp 21822 12.0407 down

Il1rl1 17082 32.4142 down Cxcl9 17329 12.7902 down

Ly86 17084 12.3274 up

Fold Fold Regulation Regulation Symbol GeneId change (D Symbol GeneId change (D (D vs C) (D vs C) vs C) vs C)

Cfd 11537 19.4853 down Cd70 21948 3.0983 down

Aire 11634 6.7877 up Tnfsf4 22164 4.6365 down

Rab27a 11891 7.7976 up Gm106 226866 4.9926 down

C4bp 12269 2.5429 up Ela2 50701 2.3222 up

Cacnb4 12298 3.6582 down Il1rapl2 60367 12.6788 up

Cd1d2 12480 3.6649 up Nfam1 74039 3.105 up

Ctsg 13035 2.2564 up Trat1 77647 17.0559 up Daf2 13137 2.1073 up Sla2 77799 2.0164 down

Il1rl1 17082 2.1658 down Btla 208154 5.4323 down

Mpo 17523 2.0435 up Cd27 21940 3.9945 up

Prg2 19074 2.9721 up Tnfsf11 21943 2.0074 down

Rbp4 19662 2.3442 down Pglyrp1 21946 3.7317 up

Cx3cl1 20312 3.6207 up