Human Regulatory T Cells Mediate Transcriptional Modulation of Dendritic Cell Function

This information is current as Emily Mavin, Lindsay Nicholson, Syed Rafez Ahmed, Fei of September 29, 2021. Gao, Anne Dickinson and Xiao-nong Wang J Immunol published online 28 November 2016 http://www.jimmunol.org/content/early/2016/11/26/jimmun ol.1502487 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2016/11/26/jimmunol.150248 Material 7.DCSupplemental http://www.jimmunol.org/ Why The JI? Submit online.

• 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

by guest on September 29, 2021 *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 © 2016 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published November 28, 2016, doi:10.4049/jimmunol.1502487 The Journal of Immunology

Human Regulatory T Cells Mediate Transcriptional Modulation of Dendritic Cell Function

Emily Mavin,*,1 Lindsay Nicholson,*,1 Syed Rafez Ahmed,* Fei Gao,† Anne Dickinson,* and Xiao-nong Wang*

Regulatory T cells (Treg) attenuate dendritic cell (DC) maturation and stimulatory function. Current knowledge on the functional impact of semimature DC is limited to CD4+ T cell proliferation and cytokine production. Little is known about the molecular basis underpinning the functional effects of Treg-treated DC (Treg-DC). We present novel evidence that Treg-DC skewed CD4+ naive T cell polarization toward a regulatory phenotype and impaired CD8+ T cell allo-reactive responses, including their ability to induce target tissue damage in a unique in vitro human graft-versus-host disease skin explant model. Microarray analysis clustered Treg-DC as a discrete population from mature-DC and immature-DC, with 51 and 93 genes that were significantly

over- or underexpressed, respectively, compared with mature-DC. Quantitative real-time PCR analysis revealed an intermediate Downloaded from expression level of CD38, CD83, CD80 and CD86 mRNA in Treg-DC, lower than mature-DC, higher than immature-DC. We also observed an attenuation of NF-kB pathway, an upstream regulator of the aforementioned genes, concomitant with reduced expression of two NF-kB-signaling related genes RELB and NFkBIZ, in the Treg-DC, together with an increased expression of Wnt5a, a negative regulator of DC differentiation. We further confirmed that the Treg-DC–mediated skewed CD4+ naive T cell polarization resulted from decreased IL-12 secretion by Treg-DC, which may be post-transcriptionally modulated by decreased

expression of microRNA-155 in Treg-DC. To our knowledge, this is the first study demonstrating a transcriptional modulation of http://www.jimmunol.org/ DC function by human Treg, partially via attenuation of the NF-kB signaling pathway and upregulation of Wnt5a, suggesting Treg may interfere with DC reprogramming during maturation, thereby modulating DC function. The Journal of Immunology, 2017, 198: 000–000.

egulatory T cells are of central importance for the main- However, previous studies have only reported a very narrow range tenance of peripheral tolerance and regulation of cellular of functional outcomes in terms of the effect of Treg-treated DC R immune responses. Despite originally being recognized for on target T cells, which were primarily limited to the ability of their suppressive function directly exerted on effector T cells (1–3), Treg-cultured DC to stimulate CD4+ T cell proliferation (4–6). by guest on September 29, 2021 it is now clear that regulatory T cells (Treg) can block effective Furthermore, little is known about the molecular basis underlying T cell priming by influencing dendritic cell (DC) functions. A the Treg-mediated modulation of DC function. Conflicting results number of studies have reported that DC can be arrested at a have been reported in murine studies about whether Treg modu- semimature status following coculture with Treg, expressing low late DC costimulatory molecule expression at the transcriptional levels of costimulatory molecules and high levels of HLA-DR, or posttranscriptional level (7, 8). Research on the modulation of rendering them deficient in initiating T cell proliferation (4, 5). DC function by human Treg is scarce. It is not known if human Treg can modulate DC phenotypic and functional properties at the transcriptional level. Our study aimed to extend limited knowl- *Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, NE2 edge in the functional impacts of Treg-treated DC on target T cells, 4HH, United Kingdom; and †Institute of Neuroscience, Newcastle University, New- castle upon Tyne, NE7 1RU, United Kingdom particularly in the context of allograft transplant complications 1E.M. and L.N. contributed equally to this work. where Treg have shown a promising therapeutic potential. We also ORCIDs: 0000-0002-2132-9011 (S.R.A.); 0000-0003-1797-0479 (F.G.); 0000-0002- sought to gain further insight into the molecular mechanisms by 7356-7636 (A.D.). which Treg modulate DC function. To address these research aims, Received for publication November 25, 2015. Accepted for publication October 21, we have generated monocyte-derived DCs with or without Treg 2016. coculture. The DCs were then purified from the coculture by FACS This work was supported by funding from CellEurope, the British Society for Hae- sorting following LPS challenge and used for subsequent functional matology, and the Wellcome Trust Institutional Strategic Support Fund. and molecular analysis. The functional impacts of Treg-treated DC The microarray data presented in this article have been submitted to the Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/) under accession number were investigated by examining their ability to uptake Ag, skew + + GSE72893. CD4 naive T cell polarization and elicit CD8 T cell alloreactive Address correspondence and reprint requests to Dr. Xiao-nong Wang, Haemato- responses including the induction of graft-versus-host target tissue logical Sciences, Newcastle University, Leech Building, The Medical School, damage using an in vitro human graft-versus-host disease (GvHD) Framlington Place, Newcastle upon Tyne NE2 4HH, U.K. E-mail address: X.N. [email protected] skin explant model. The Treg-treated DC were further subjected to The online version of this article contains supplemental material. gene expression profiling using the Illumina Human HT-12 micro- array. The transcript data analysis was focused on selected genes Abbreviations used in this article: CBA, cytometric bead array; DC, dendritic cell; GvHD, graft-versus-host disease; imm-DC, immature DC; mat-DC, mature DC; with known functions in regulating DC maturation, stimulatory func- miR-155, microRNA-155; qRT-PCR, quantitative real-time PCR; Treg, regulatory tion and cytokine production, as well as the genes involved in NF- T cell; Treg-DC, Treg-treated DC. kB signaling pathway activation to identify the molecular basis Copyright Ó 2016 by The American Association of Immunologists, Inc. 0022-1767/16/$30.00 underlying the observed functional impact.

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1502487 2 HUMAN Treg MODULATE DC FUNCTION AT THE TRANSCRIPTIONAL LEVEL

Materials and Methods IFN-g, IL-4, and FOXP3 as well as mRNA expression of TBX21 (T-bet), GATA-3 and FOXP3. Following restimulation with CD3/CD28 T cell ex- All samples were collected from healthy volunteers with informed consent pander beads at a 1:1 ratio for 24 h cytokine secretion of IFN-g, IL-4, and and approval from the local research ethics committee. Unless otherwise IL-10 in the culture supernatants was measured with CBA (BD Biosci- stated, all cell populations were isolated from leukocyte reduction system ences). For IL-12 neutralization experiments, human anti-IL-12 and cor- cones (National Blood Services, Newcastle, U.K.). responding polyclonal goat IgG (R&D Systems) were added at 5 mg/ml to + + Cell isolation and culture the coculture of CD4 naive T cells and mat-DC on day 0. The naive CD4 polarization was assessed following stimulation and resting as previously General culture medium (RF10) consisted of RPMI 1640 with 100 IU/ml mentioned. Following polarization of CD4+ naive T cells by Treg-DC, the penicillin, 100 mg/ml streptomycin, 2 mM L-glutamine, and 10% heat- FOXP3-expressing T cells were isolated by FACS sorting based on inactivated FCS (Invitrogen). Naturally occurring Treg were isolated by CD25highCD127low gating. The function of Treg-DC induced Treg was negative enrichment of CD4+ T cells (RosetteSep; StemCell Tech) fol- then assessed by their ability to suppress CD8+ T cell activation and pro- lowed by immuno-magnetic positive selection of CD25high cells (RoboSep; liferation in an allogeneic MLR setting as previously described (10). StemCell Tech). Treg were then activated and expanded for up to 3 wk by culturing with CD3/CD28 Dynabeads (Invitrogen), 500 IU/ml IL-2 Skin explant assay (Roche) and 100 nM rapamycin (Sigma) prior to use, as previously de- A well-established in vitro human GvHD skin explant model was used to scribed (9). All Treg were well characterized phenotypically and func- assess the ability of Treg-treated DC to induce cutaneous tissue damage tionally prior to use as described previously (10). Mature and immature mediated by alloreactive T cell responses (11). Punch skin biopsies and DCs (mat-DC and imm-DC) were generated from magnetically isolated + peripheral blood monocyte-derived DC were obtained from the same CD14 monocytes (Miltenyi) cultured with IL-4 and GM-CSF (50 ng/ml; healthy donor. Allogeneic CD8+ T cells were stimulated with imm-DC, ImmunoTools) for 7 d, with or without LPS maturation (100 ng/ml; Sigma) mat-DC, or Treg-DC at a T cell:DC ratio of 10:1 for 5 d in RF10 sup- respectively on day 6. Treg-treated DC (Treg-DC) were generated by +

plemented with exogenous IL-2 (25 IU/ml). Allogeneic CD8 T cells Downloaded from adding Treg into the aforementioned DC generation culture on day 3 at primed with mat-DC in the presence of Treg at a T cell:Treg ratio of 4:1 a Treg:DC ratio of 3:1, and remained in the culture till day 7 with LPS served as an additional control (9). The primed alloreactive CD8+ T cells maturation on day 6. Treg-treated immature DC were generated the same (1 3 106) were cocultured with small skin sections for a further 3 d. Skin way as Treg-DC but without exposure to LPS. Prior to gene expression and sections cocultured with unprimed CD8+ T cells or culture medium alone functional analysis, Treg-DC were purified from the coculture by cell served as negative controls, which have historically given rise to back- sorting using the FACS Fusion Sorter (BD Biosciences) based on for- ground grade I tissue damage (11). Skin sections were then routinely fixed, ward scatter/side scatter and CD3 negative gating and sorted DC were + + sectioned and H&E stained. The severity of cutaneous tissue damage was reanalyzed to ensure .90% purity. Naive CD4 T cells and CD8 T cells scored by two independent evaluators according to the Lerner grading http://www.jimmunol.org/ (.90% purity) were isolated via negative enrichment, according to the system (12) with grade 0, normal skin; grade I, mild vacuolization of basal manufacturer’s instructions (StemCell Tech). epidermal layer; grade II, diffuse vacuolization of basal cells with scattered Flow cytometry dyskeratotic bodies; grade III, sub-epidermal cleft formation; and grade IV, complete dermal epidermal separation. Flow cytometry data were acquired on BD FACS Canto II cytometer and analyzed with FlowJo software (Tree Star). Intracellular staining was Illumina BeadChip expression arrays performed according to manufacturer-recommended protocols after cell Total RNA from imm-DC, mat-DC, and Treg-DC was extracted using the fixation and permeabilization (Fix/Perm buffer kit; eBioscience or Cytofix/ RNeasy Micro kit (Qiagen, Sussex, U.K.) and quantified using a NanoDrop Cytoperm; BD Biosciences). Unless otherwise stated all Abs were supplied ND-1000 spectrophotometer (LabTech, U.K.). The RNA integrity was

by BD Biosciences: CD25 APC (M-A251), CD127 PE (HIL-7R-M21), CD8 by guest on September 29, 2021 confirmed using an Agilent 2100 bioanalyzer (Agilent, U.K.). To control for PerCpcy5.5 (SK1), CD40 APC (5C3), CD80 PE (L307.4), CD83 FITC any changes in gene expression caused by shear forces, all samples were (HB15e), CD86 PEcy7 (FUN-1), CD38 PE (HB-7), FOXP3 APC (PCH101; subjected to a FACS sorting procedure prior to RNA extraction. mRNA eBiosciences), HLA-DR PerCP (L243), IFN-g PE (25723.11), IL-4 APC profiling was carried out on the Illumina HT12 microarray platform by the (8D4-8), LAP-TGFb PE (27232; R&D), and phospho-p65 Alexa Fluor 488 High-Throughput Genomics group at the Wellcome Trust Centre for Human (Ser536) (Cell Signaling Technology, Danvers, MA). Appropriate isotypes Genetics, as per the manufacturer’s protocol (GEO accession number were used as controls. GSE72893, http://www.ncbi.nlm.nih.gov/geo/). Raw microarray data were Phagocytosis assay processed using well-established R/Bioconductor workflows. Data was normalized using Illumina variant stability transformation and robust The ability of DC to uptake Ag was assessed by incubation of DCs (5 3 spline normalization. Differential expression was detected using linear 10^4)withFITC-dextran(1mg/ml;Sigma-Aldrich)inRF10for1h models and empirical Bayesian techniques (Limma). Clustering analysis at 37˚C including a control incubated at 4˚C to exclude extracellular (hierarchical clustering) was used to show the relationships between dif- binding of FITC-dextran. Cells were then extensively washed in cold ferent DC populations. Genes with a log2-fold change and a p value #0.05 FACS buffer and analyzed by flow cytometry. (corrected for multiple testing) were considered significantly different between groups. Microscopy Quantitative real-time PCR Cytospins were prepared from imm-DC, mat-DC, and Treg-DC. The cells were fixed with 2% paraformaldehyde and stained with Giemsa staining kit Total RNA was extracted using a Qiagen RNeasy Mini Kit according to the (Clin-Tech, U.K.) according to the manufacturer’s instructions. Images manufacturer’s instructions (Qiagen, Crawley, U.K.). RNA quality and were captured using an Axioplan 2 microscope with Axiocam and Axio- concentration were assessed using NanoDrop ND-1000 (Thermo Scientific). vision software (Cal Zeiss). Copy DNA synthesis was carried out using the Applied Biosystem High- Capacity cDNA Reverse Transcriptase kit according to the manufacturer’s T cell activation, proliferation and cytokine production instructions. Real-time PCR was performed using Taqman primer and The activation and proliferation of allogeneic CD8+ T cells (1 3 106/ml) probe sets for CD80 (Hs00175478_m1), CD86 (Hs01567026_m1), CD38 3 5 (Hs01120071_m1), CD83 (Hs00188486_m1), RelB (Hs00232399_m1), primed with imm-DC, mat-DC or Treg-DC (1 10 /ml) were determined k by CD25 expression and CFSE dilution following MLC for 5 d. The level NF BIZ (Hs00230071_m1), WNT5a (Hs00998537_m1), TBX21 of cytokines secreted by each DC group was quantified in the culture (Hs00203436_m1), GATA-3 (Hs00231122_m1), FOXP3 (Hs01085834_m1), supernatants collected on day 7 of DC generation using the cytometric and GAPDH (Hs99999905_m1) using TaqMan Universal PCR MasterMix bead array (CBA) flex kit according to the manufacturer’s instructions (BD (all Applied Biosystems, Warrington, U.K.) according to the manufacturer’s Biosciences). Data were analyzed with FCAP software (BD Biosciences). instructions. For microRNA-155 (miR-155) expression, RNA was extracted using a Total RNA Purification kit (Norgen Biotek, Canada) and converted Naive T cell polarization into cDNA using specific stem-loop primers (miR-155: 002623 and the endogenous control, RNU48: 001006) and TaqMan MicroRNA Reverse CD4+ naive T cells were stimulated with either imm-DC, mat-DC or Treg- Transcription Kit (all Applied Biosystems). Quantitative real-time PCR DC for 6 d in RF10 at a ratio of 10:1. The stimulated T cells were sub- analysis (qRT-PCR) was performed using the microRNA-specific Taqman jected to a resting culture in RF10 supplemented with IL-2 (25 IU/ml) for a assay (Applied Biosystems) and SensiFAST (Bioline). All samples were further 4 d. The cells were examined for their intracellular expression of analyzed in triplicate from at least three independent occasions with the ABI The Journal of Immunology 3

7900 Fast Real-Time PCR System (Applied Biosystems). Data was analyzed decreased mRNA expression of the master transcription factor, using SDS version 2.4 software; relative gene expression was calculated TBX21 (T-bet) (p = 0.0231), which controls the expression of Th1 using the comparative Ct method (13). cytokine and initiates Th1 lineage development. Furthermore, Wnt5a inhibition Treg-DC stimulated CD4+ naive T cells exhibited elevated FOXP3 expression at both protein and mRNA levels (p = 0.0035 and The Wnt5a inhibitor, Box5, was purchased from Merck Millipore. One hundred micromolars of Box5 was added into the coculture of DC and Treg 0.0190 respectively), coupled with increased secretion of IL-10 at day 3, alongside Treg, or day 6, in combination with LPS. The Box5 (p = 0.0012) (Fig. 3A–C). No significant difference was observed concentration used has previously been shown to antagonize cellular ac- in Th2 lineage development indicated by similar levels of GATA3 tivities (14) and was further confirmed to be non-toxic to moDC or Treg in expression, together with comparable IL-4 at intracellular and our pilot study (data not shown). Treg-DC generated in the absence of Box5 served as a control. secreted levels, irrespective of stimulation with imm-DC, mat-DC, or Treg-DCs (data not shown). The induction in FOXP3 expres- Statistical analysis sion and IL-10 secretion was also observed in imm-DC stimulated + All statistical analysis was carried out using GraphPad Prism version 5 CD4 naive T cells (Fig. 3A–C). (GraphPad software, San Diego, CA). Mann-Whitney U tests or paired To investigate if this skewed polarization driven by Treg-DC is t tests were used to determine statistical significance. a direct effect of the reduced IL-12 secretion by Treg-DC, we cocultured CD4+ naive T cells with mat-DC in the absence or Results presence of IL-12 neutralizing Ab. Blockade of IL-12 completely Treg-treated DC exhibit semimature phenotype and mimicked the Treg-DC mediated suppression in Th1 polarization of morphology CD4+ naive T cells, showing decreased frequency of IFN-g Downloaded from We first examined the maturation status and stimulatory potential expressing cells, diminished secretion of IFN-g and reduced TBX21 of human Treg-treated DC by analyzing the cell-surface marker gene expression (p = 0.0036, 0.0072, and 0.00449, respectively) expression. As expected, mat-DC expressed the highest levels of (Fig. 3D). However, IL-12 blocking failed to mimic Treg-DC me- + CD83, HLA-DR, CD80, CD86, and CD40 whereas imm-DC diated suppression of CD8 T cell activation or proliferation (data expressed the lowest. Treg-DC expressed all these markers at in- not shown), suggesting that Treg-treated DC may use different + termediate levels, which were significantly lower than those of the mechanistic pathways to exert their function on modulating CD4 http://www.jimmunol.org/ naive T cell polarization and CD8+ effector T cell function. mat-DC but higher than the imm-DC, except HLA-DR, which was + comparable to the mat-DC (Fig. 1A). Interestingly, Treg-DC expressed To confirm the finding that Treg-DC drove naive CD4 T cell a markedly reduced level of CD38 compared to mat-DC (Fig. 1A). polarization toward the Treg phenotype, we evaluated the suppres- sive function of the FOXP3-expressing T cells following polarization The semimature status of the Treg-DC was also supported by their high low distinctive morphology (Fig. 1B). by FACS sorting these cells using a CD25 and CD127 gate. The presence of titrated numbers of the sorted cells in the coculture of Treg-treated DC show impaired Ag uptake, reduced cytokine mat-DC and CD8+ T cell gave rise to a dose-dependent suppression in production and increased surface TGFb expression the activation and proliferation of allogeneic CD8+ T cells (Fig. 3E). Immature DCs effectively capture Ags, which can be assessed by by guest on September 29, 2021 Treg-treated DC are less able to induce CD8+ T cell activation their uptake of FITC-dextran. We found that Treg-treated immature and proliferation DC had a very limited ability to uptake Ag with an over 6-fold reduction compared with the imm-DC that had no exposure to Treg To add to the current knowledge on the effect of Treg-treated DC on + (p = 0.047) (Fig. 2A), suggesting that Treg can exert their mod- CD4 T cell proliferation we evaluated the ability of Treg-DC to + ulatory effect on DC function by blocking the very first step of DC stimulate CD8 T cell responses. Treg-DC showed a significantly + maturation. reduced ability to stimulate CD8 T cell activation and prolifer- Modulation of cytokine production is a key mechanism by which ation compared with mat-DC (p = 0.009 and 0.0467 respectively) Treg can regulate DC function. We measured the cytokine secretion (Fig. 4D versus Fig. 4C) with no significant difference to imm-DC levels of three DC groups. Treg-DC secreted lower levels of IL-12 (Fig. 4D versus Fig. 4B). To assess if treating DC alone with Treg and IL-6 than mat-DC (p = 0.0079 and 0.02, respectively), with no could have a similar effect as adding Treg into the coculture of DC + + significant difference to imm-DC. Both Treg-DC and imm-DC and CD8 T cells, we compared the responses of CD8 T cells secreted very low levels of IL-10, ∼30-fold lower than that pro- stimulated with Treg-treated DC alone or with mat-DC in the duced by mat-DC (p = 0.0317) (Fig. 2B). Preactivated Treg alone presence of Treg. We found that the presence of Treg in the co- + served as a control. They secreted neither IL-12 nor IL-6 and little culture of DC and CD8 T cells was much more effective in + IL-10 (Fig. 2B). Furthermore, Treg-DC showed an upregulated suppressing CD8 T cell activation and proliferation than Treg- expression of surface-bound LAP-TGFb compared with the mat-DC treated DC alone (p = 0.0085) (Fig. 4D versus Fig. 4E), suggesting (p = 0.0396) (Fig. 2C), highlighting the fact that Treg-treated DC that multifactorial mechanisms are required for Treg to achieve could modulate target T cell function via both cell to cell contact their optimal effect on modulating target T cell responses. and soluble factor pathways. Treg-DC primed CD8+ T cells are impaired in their ability to Treg-treated DC skew CD4+ naive T cell polarization toward a induce graft-versus-host tissue damage regulatory phenotype We then further evaluated the functional impact of Treg-DC on Directing naive T cell polarization is one of the prime mechanisms CD8+ T cells in a human disease model by examining the ability whereby DC dictate the outcome of T cell responses. To date, there of Treg-DC primed allogeneic CD8+ T cells to mediate graft- is no evidence available in the literature with regard to how Treg- versus-host tissue damage using an in vitro human GvHD skin treated DC would drive CD4+ naive T cell polarization. We found explant model. This model allows detection of host tissue damage that, compared with mat-DC stimulation, CD4+ naive T cells stim- mediated by donor alloreactive T cells presensitized by host DC ulated with Treg-DC showed a significantly reduced capacity to (9, 11). In this experiment, small skin sections were incubated for produce the Th1 cytokine, IFN-g, at both intracellular and se- 3 d with effector alloreactive CD8+ T cells that were presensitized creted levels (p = 0.0029 and 0.0205, respectively). They also had for 5 d with imm-DC, mat-DC, and Treg-DC from the skin donor. 4 HUMAN Treg MODULATE DC FUNCTION AT THE TRANSCRIPTIONAL LEVEL Downloaded from http://www.jimmunol.org/ FIGURE 1. Treg-treated DC exhibit semimature phenotype and morphology. (A) Mean fluorescence intensity and representative histograms of indicated cell surface markers on imm-DC, mat-DC and Treg-DC. Error bars indicate the SEM of 4–24 independent experiments. Prior to plots shown, gates were applied on live cells. (B) Morphology of imm-DC, mat-DC, and Treg-DC following cytospin and Giemsa staining, representative of five independent experiments. *p , 0.5, **p , 0.01, ***p , 0.001.

Skin sections cultured with Treg-DC primed CD8+ T cells had the modulation of DC function. We first explored the global land- least graft-versus-host tissue damage (grade I–II) compared with scape of the gene expression profile of DCs using microarray those cultured with CD8+ T cells presensitized with mat-DC or analysis. We found that, relative to imm-DC, the levels of 1820

imm-DC (grade II–III) (Fig. 5D versus Fig. 5B, 5C, respectively) individual genes were differentially regulated in mat-DC by at least by guest on September 29, 2021 (p = 0.038). Skin sections incubated with unprimed CD8+ T cells 2-fold and 1315 genes were differentially modulated in the Treg- or culture medium alone exhibited grade I background changes DC (Fig. 6A). Hierarchical clustering analysis grouped Treg-DC (Fig. 5A, 5F). CD8+ T cells primed with mat-DC in the presence of as discrete populations from mat-DC and imm-DC (Fig. 6B). Treg also only gave rise to background level tissue damage (Fig. 5E). Principal component analysis displayed a tightly polarized dis- tribution for mat-DC and imm-DC, with Treg-DC positioned in Treg-treated DC show modified gene expression and attenuated between the two (Fig. 6C). Direct comparison between Treg-DC k NF- B activity versus mat-DC identified 51 and 93 genes that were significantly Our findings in the phenotypic changes and functional impairments over- or underexpressed in Treg-DC respectively (Supplemental of Treg-treated DC prompted us to search for the molecular evi- Tables I and II), indicative of the possibility that Treg may dence that could lead to further understanding of Treg-mediated modulate a wide range of DC functions at the transcriptional level.

FIGURE 2. Treg-treated DC show impaired Ag uptake, reduced cyto- kine production, and increased sur- face LAP-TGFb expression. (A) Uptake of FITC-dextran by DCs with a representative histogram shown on the left. Error bars indi- cate SEM of three to eight inde- pendent experiments. (B) Cytokine secretion measured by CBA flex and analyzed with FCAP array. Error bars indicate the SEM from five independent experiments. (C)Cell surface LAP-TGFb expression, as measured by flow cytometry. Live cells were selected prior to analy- sis. Data from four independent experiments. *p , 0.5, **p , 0.01. The Journal of Immunology 5 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 3. Treg-treated DC skew CD4+ naive T cell polarization toward a regulatory phenotype. (A) Frequency of IFN-g-secreting or FOXP3-expressing cells following CD4+ naive T cell polarization by imm-DC, mat-DC or Treg-DC; representative dot plots are shown on the left. Prior to plots shown, a live cell gate was applied. Error bars indicate SEM from 10 to 14 independent experiments. (B) Cytokine secretion by imm-DC, mat-DC and Treg-DC measured using CBA. Culture supernatants were collected following polarization, resting and restimulation of naive T cells with CD3/CD28 beads for a further 24 h. (C)QuantificationofmRNA expression of the master transcription factors for Th1 (TBX21)andTreg(FOXP3) following polarization of CD4+ naive T cells, normalized to GAPDH expression and error bars indicate the SEM for six independent experiments carried out in triplicate. (D) The frequency of IFN-g secreting cells, expression of TBX21 mRNA and levels of secreted IFN-g by mat-DC with or without IL-12 blocking. (E) Dose-dependent suppression of CD8+ T cell activation and proliferation by sorted FOXP3 expressing cells following CD4+ naive T cell polarization by Treg-DC; representative dot plots are shown on the left. *p , 0.5, **p , 0.01. 6 HUMAN Treg MODULATE DC FUNCTION AT THE TRANSCRIPTIONAL LEVEL

FIGURE 4. Treg-treated DC have a reduced capacity to induce CD8+ T cell activation and proliferation. Activation and proliferation of CD8+ T cells Downloaded from under the conditions of unstimulated (A), stimulated with imm-DC (B), mat-DC (C), Treg-DC (D) or mat-DC in the presence of Treg (E). Representative contour plots are shown on the left. CD25 expression and CFSE dilution was measured on day 5 of MLR. Prior to plots shown, gates were applied to live CD8+ T cells. Error bars indicate the SEM from six to eight independent experiments. *p , 0.5, **p , 0.01.

We next focused our evaluation on the transcript data for selected confirm the functional attenuation of NF-kB signaling in Treg-

genes with known involvement in DC maturation, stimulatory func- DC, we examined serine 536 phosphorylation of the p65 subunit http://www.jimmunol.org/ tion, cytokine production, and NF-kB signaling pathways (Fig. 6D). as a marker of NF-kB signaling pathway activity. In response to In comparison with imm-DC, the mat-DC group had the highest LPS stimulation, the Treg-treated DC showed significantly lower level of mRNA expression of CD80, CD83,andCD86 (12.89-, Ser536 phosphorylation compared with non–Treg-treated DC 5.69- and 3.94-fold increase, respectively) whereas Treg-DC had an (p = 0.0191), suggestive of attenuated NF-kB activation in Treg- intermediate level expression of these genes (4.46-, 3.90- and 1.92- DC, although the signaling was not completely abolished com- fold increase, respectively), conferring the semimature phenotype pared with DC without LPS exposure (Fig. 6G). of Treg-treated DC at the transcriptional level. CD38 mRNA ex- pression in Treg-DC was similar to that of imm-DC but over 7-fold Wnt5a expression in Treg-DC is partially responsible for their lower than the mat-DC. The mRNA expression of CD38, CD83, semimature status by guest on September 29, 2021 CD80, and CD86 was then validated independently by qRT-PCR Upregulation of Wnt5a expression plays an important role in at- and was fully concordant with the microarray data (Fig. 6E). tenuating DC maturation (15, 16). We found Wnt5a expression Given that the transcription of CD80, CD86, CD83, and CD38 was significantly upregulated in Treg-treated DC compared with can all be regulated by the transcription factor NF-kB and that this mat-DC (3.84-fold higher) (Fig. 6D). Treg mediated upregulation pathway is fundamental to DC maturation, we further analyzed of Wnt5a on DC was confirmed by qRT-PCR (Fig. 7A). To ex- NF-kB–signaling related genes. Microarray analysis and qRT- amine the functional significance of the upregulated Wnt5a in the PCR validation have both shown that Treg-DC expressed signifi- Treg-DC, we evaluated the expression of maturation and costim- cantly lower levels of the genes that encode the NF-kB proteins ulatory markers on the Treg-DC generated in the absence or RELB and NFKB1Z compared with mat-DC (Fig. 6D, 6F). To presence of a pharmacological Wnt5a inhibitor (Box5). In the

FIGURE 5. Treg-treated DC have an impaired ability to induce cutaneous GvH damage. Severity of GvH tissue damage in skin sections cocultured with unstimulated CD8+ T cells (A), CD8+ T cells stimulated with imm-DC (B), matDC (C), Treg-DC (D), mat-DC in the presence of Treg (E), and medium only (F). Scale bars on representative skin images indicate 50 mm. The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 6. Treg-treated DC show modified gene expression and attenuated NF-kB activity. (A–C) Venn diagram, hierarchical clustering and principle components plot representing the differences/similarities between the three DC populations. (D) Heatmap of selected genes with known involvement in DC maturation, modulatory function, and NF-kB signaling pathways. (E and F) Expression levels of CD38, CD80, CD86, CD83, RelB, and NFkBIZ mRNA, normalized against the expression of the housekeeping gene, GAPDH. Error bars represent the SEM of at least three independent experiments. (G). Phospho-p65 NF-kB activity in LPS-stimulated DCs with and without Treg coculture. Cells were stimulated with 100 ng/ml LPS for 15 mins prior to harvesting and stained with either phospho-p65 Ab or the corresponding isotype control. Data plotted is the ratio of mean fluorescence intensity between the phospho-p65 (Ser536) and isotype control. Error bars represent the SEM of seven independent experiments. *p , 0.5, **p , 0.01, ***p , 0.001. absence of Box5, Treg-treated DC showed a reduced surface ex- (Fig. 6D), indicating that a mechanism involving post- pression of CD38, CD83, CD80, and CD86 compared with mat-DC. transcriptional modulation could have contributed to the reduced The downregulated CD38, CD83, CD80, and CD86 were partially secretion of IL-12 and IL-10 by Treg-DC. Interestingly, the reversed in the presence of Box5 (Fig. 7B). Treatment of mat-DCs microarray and qRT-PCR analysis also showed a reduced mRNA with Box5 had a negligible effect on the levels of CD38, CD83 expression level of miR-155 (Figs. 6D, 7C), a key post- CD80, and CD86 (data not shown). transcriptional regulator of IL12 production, in Treg-treated DC.

Decreased cytokine secretion by Treg-DC could be controlled Discussion at both transcriptional and post-transcriptional levels This report presents novel evidence that Treg-treated DC skewed Guided by our finding that Treg-DC secreted lower levels of IL-12, CD4+ naive T cell polarization toward a regulatory phenotype and IL-6, and IL-10 than mat-DC (Fig. 3B), we then sought to evaluate impaired CD8+ T cell allo-reactive responses including their transcriptional changes that control DC cytokine production. ability to induce GvHD target tissue damage. These functional Microarray analysis revealed that the mRNA expression level of outcomes could be attributed to the reduced IL-12 secretion and IL-6 was 4.4-fold lower in Treg-DC than in mat-DC (Fig. 6D), semimaturation phenotype expression by Treg-DC. The dissocia- suggestive of transcriptional modulation as a key mechanism tion between reduced expression of CD80/CD86 and high ex- leading to reduced IL-6 secretion by Treg-DC. On the contrary, pression of HLA-DR is critical for Treg-DC to switch their microarray analysis found no significant difference in mRNA function from stimulatory to tolerogenic. It has been well docu- expression levels of IL-12 or IL-10 between Treg-DC and mat-DC mented that DC can be arrested at a semimature status following 8 HUMAN Treg MODULATE DC FUNCTION AT THE TRANSCRIPTIONAL LEVEL Downloaded from

FIGURE 7. Upregulated Wnt5a expression in Treg-DC is partially responsible for their semimature status. (A) Expression of Wnt5a mRNA, normalized against the expression of the housekeeping gene, GAPDH. Error bars represent the SEM of $4 independent experiments. (B) Expression of surface markers. http://www.jimmunol.org/ Error bars represent the SEM of $3 independent experiments. Prior to plots shown, gates were applied to live cells. (C) Expression of miR-155 mRNA, normalized against the expression of the endogenous gene, RNU 48. Error bars represent the SEM of three independent experiments. *p , 0.5, **p , 0.01. treatment with Treg, rendering them deficient in stimulating CD4+ p65 subunit and decreased expression of two NF-kB signaling re- T cell proliferation (4, 5). Little is known about the molecular lated genes, RelB and NFkBIZ, indicative of attenuated NF-kB ac- basis underlying the Treg-mediated semimature status of DC. tivation. RelB belongs to the Rel protein family and is directly Conflicting results have been reported in murine studies with associated with DC differentiation and function (23). Silencing of regard to whether Treg modulate DC costimulatory molecule ex- RelB resulted in maturation arrest in murine DCs and failure to by guest on September 29, 2021 pression at the mRNA level or post-transcriptionally (7, 8). A stimulate a T cell response (24). NFkBIZ encodes the protein IkBz, recent murine study has shown that Ag-specific inducible Treg are an NF belonging to the Bcl-3 family. Very little is known of its role not able to suppress CD80 and CD86 expression in DC at the in DC, however, it is induced by TLR agonists in macrophages and mRNA level. Instead they downregulate these molecules post- interacts with another subunit of NF-kB, p50, to positively regulate a translationally via an IL-10/MARCH-1–dependent mechanism set of genes including IL-6 and IL-12b (25). Collectively our findings (7). We have demonstrated that human Treg suppress the ex- suggest that Treg may act as NF-kB inhibitors to block DC Ag pression of CD83, CD80, and CD86 in DC not only at the presentation, inhibiting the downstream immune responses. How- protein level as previously reported (4, 5), but also at the tran- ever, the mechanism of how Treg attenuate the NF-kB pathway in scriptional level. To our knowledge, Treg-mediated transcriptional DC remains to be elucidated. modulation of DC maturation has not been reported previously in To our knowledge this study also reported for the first time, that humans. Wnt5a expression is increased in Treg-DC compared with mat- Another novel finding of this study is that human Treg down- DC. Wnt5a is one of the most extensively studied Wnt family regulate CD38 expression and reduce NF-kB pathway activity in members and is known to play an important role in many devel- monocyte-derived DC. This may govern the upstream signaling opmental processes, including differentiation of human monocytes events leading to transcriptional modulation of DC maturation to DC (26). Growing evidence points to Wnt5a as a negative observed in this study. CD38 is a type II transmembrane glyco- regulator of DC differentiation (16, 27) demonstrating that Wnt5a protein that synthesizes and hydrolyzes cyclic ADP-ribose, which expression can skew differentiation to a more tolerogenic func- has previously been shown to act as a receptor mediating signaling tional state in both DCs (15, 28) and macrophages (16). Increased events upstream of CD83 expression and IL-12 secretion (17). It is Wnt5a signaling in monocyte-derived DCs has led to their reduced plausible to confer that Treg-mediated reduction of CD38 ex- ability in dextran uptake as well as reduced surface expression of pression on DC could, at least partially, arrest the molecular costimulatory molecules with compromised functional capacity process of DC maturation leading to a semimature DC phenotype (15), results similar to our findings of Treg-modulation of DCs. and impaired functionality. We found that pharmacological inhibition of Wnt5a activity can The transcription factor NF-kB plays a fundamental role in the partially restore the expression levels of maturation markers and complex multistep process of DC maturation, including regulation of costimulatory molecules on Treg-DC, suggesting an active role for CD38, CD83, CD80, and CD86 expression as well as IL-12 induc- upregulated Wnt5a in shaping the functional features of Treg- tion (17–21). Inhibition of NF-kB activation using pharmacological treated DC. Further research is warranted to define the precise agents results in a failure to upregulate the expression of CD80, role of Wnt5a in this process. CD86, and MHC class II in DCs (22). We have found that Treg- We further identified that human Treg use both transcriptional treated DC have inhibited Ser536 phosphorylation of the NF-kB and post-transcriptional mechanisms to modulate the production of The Journal of Immunology 9 different cytokines by DC. Although Treg-DC and mat-DC express 6. Houot, R., I. Perrot, E. Garcia, I. Durand, and S. Lebecque. 2006. Human CD4+CD25high regulatory T cells modulate myeloid but not plasmacytoid similar mRNA levels of IL-12 and IL-10, there is a significant dendritic cells activation. J. Immunol. 176: 5293–5298. reduction in the secreted levels of IL-12 and IL-10 by the Treg- 7. Chattopadhyay, G., and E. M. Shevach. 2013. Antigen-specific induced T reg- DC, whereas IL-6 was reduced at both the mRNA and secreted ulatory cells impair dendritic cell function via an IL-10/MARCH1-dependent mechanism. J. Immunol. 191: 5875–5884. levels in the Treg-DC. This suggests that human Treg modulate 8. Cederbom, L., H. Hall, and F. Ivars. 2000. CD4+CD25+ regulatory T cells down- DC IL-10 and IL-12 production post-transcriptionally, whereas regulate co-stimulatory molecules on antigen-presenting cells. Eur. J. Immunol. IL-6 is controlled transcriptionally. Post-transcriptional modifica- 30: 1538–1543. 9. Mavin, E., S. S. Ahmed, G. O’Boyle, B. Turner, S. Douglass, J. Norden, tions of cytokine mRNA provide a rapid and flexible mechanism M. Collin, S. Ali, A. Dickinson, and X. N. Wang. 2012. Regulatory T cells in- for controlling protein expression. MicroRNAs, small non-coding hibit CD8(+) T-cell tissue invasion in human skin graft-versus-host reactions. Transplantation 94: 456–464. RNA molecules, are one of the key factors that play an important 10. Mavin, E., A. Dickinson, and X. N. Wang. 2013. Do cryopreserved regulatory role in regulation of gene expression at the post-transcriptional T cells retain their suppressive potency? Transplantation 95: e68–e70. level and have been shown to regulate the secretion of cytokines 11. Wang, X. N., M. A. Haniffa, U. Holtick, M. P. Collin, G. Jackson, C. M. Hilkens, E. Holler, M. Edinger, P. Hoffmann, and A. M. Dickinson. 2009. Regulatory by DC. A reduced expression of miR-155 in Treg-treated DC has T-cell suppression of CD8+ T-cell-mediated graft-versus-host reaction re- been observed in our study, concomitant with their decreased quires their presence during priming. Transplantation 88: 188–197. protein secretion of IL-12p70 and IL-10. Increased miRNA-155 12. Lerner, K. G., G. F. Kao, R. Storb, C. D. Buckner, R. A. Clift, and E. D. Thomas. 1974. Histopathology of graft-vs-host reaction (GvHR) in human recipients of expression has been considered a defining feature of DC matu- marrow from HL-A–matched sibling donors. Transplant. Proc. 6: 367–371. ration and is necessary for the optimal function of mature DC (29). 13. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. A previous study has demonstrated that IL-12p70 and IL-10 Methods 25: 402–408. production by DC is significantly reduced after miR-155 silenc- 14. Jenei, V., V. Sherwood, J. Howlin, R. Linnskog, A. Sa¨fholm, L. Axelsson, and Downloaded from ing during DC maturation and that miR-155 regulates IL-12p70 T. Andersson. 2009. A t-butyloxycarbonyl–modified Wnt5a-derived hexapeptide functions as a potent antagonist of Wnt5a-dependent melanoma cell invasion. production through targeting suppressor of cytokine signaling 1 Proc. Natl. Acad. Sci. USA 106: 19473–19478. (30). Taken together, the reduction of IL-12 secretion by Treg-DC 15. Valencia, J., C. Herna´ndez-Lo´pez, V. G. Martı´nez, L. Hidalgo, A. G. Zapata, observed in this study may be secondary to Treg-mediated inhi- A´ . Vicente, A. Varas, and R. Sacedo´n. 2011. Wnt5a skews dendritic cell dif- ferentiation to an unconventional phenotype with tolerogenic features. J. bition of miR-155 expression. Immunol. 187: 4129–4139.

In summary, we present novel data on the gene expression 16. Bergenfelz, C., C. Medrek, E. Ekstro¨m, K. Jirstro¨m, H. Janols, M. Wullt, http://www.jimmunol.org/ A. Bredberg, and K. Leandersson. 2012. Wnt5a induces a tolerogenic phenotype profile of Treg-treated DCs and demonstrate that Treg-treated DC of macrophages in sepsis and breast cancer patients. J. Immunol. 188: 5448– are significantly different from their counterpart DCs, which have 5458. no exposure to Treg, both transcriptionally and functionally. We 17. Fedele, G., L. Frasca, R. Palazzo, E. Ferrero, F. Malavasi, and C. M. Ausiello. 2004. CD38 is expressed on human mature monocyte-derived dendritic cells and also present the novel findings that human Treg modulate DC is functionally involved in CD83 expression and IL-12 induction. Eur. J. function, in part, by attenuation of the NF-kB signaling pathway Immunol. 34: 1342–1350. and upregulation of Wnt5a. Treg-mediated transcriptional control 18. Hsing, Y., and G. A. Bishop. 1999. Requirement for nuclear factor-kappaB activation by a distinct subset of CD40-mediated effector functions in of DC maturation process introduces a new level of modulatory B lymphocytes. J. Immunol. 162: 2804–2811. mechanism. The large number of differentially expressed genes in 19. McKinsey, T. A., Z. Chu, T. F. Tedder, and D. W. Ballard. 2000. Transcription

factor NF-kappaB regulates inducible CD83 gene expression in activated by guest on September 29, 2021 the Treg-treated DC suggests that human Treg may interfere with T lymphocytes. Mol. Immunol. 37: 783–788. the DC reprogramming that occurs during maturation, thereby 20. Zou, G. M., and W. Y. Hu. 2005. LIGHT regulates CD86 expression on dendritic modulating a wide range of biological characteristics and func- cells through NF-kappaB, but not JNK/AP-1 signal transduction pathway. J. Cell. Physiol. 205: 437–443. tional properties of DC. 21. Ouaaz, F., J. Arron, Y. Zheng, Y. Choi, and A. A. Beg. 2002. Dendritic cell development and survival require distinct NF-kappaB subunits. Immunity 16: 257–270. Acknowledgments 22. Ardeshna, K. M., A. R. Pizzey, S. Devereux, and A. Khwaja. 2000. The PI3 We thank David MacDonald, Andrew Fuller, and Andrew Filby at the Flow kinase, p38 SAP kinase, and NF-kappaB signal transduction pathways are in- Cytometry Facility, Newcastle University for their assistance in cell sorting. volved in the survival and maturation of lipopolysaccharide-stimulated human monocyte-derived dendritic cells. Blood 96: 1039–1046. We thank the High-Throughput Genomics group at the Wellcome Trust 23. Zanetti, M., P. Castiglioni, S. Schoenberger, and M. Gerloni. 2003. The role of Centre for Human Genetics for the generation of the Gene Expression Data RelB in regulating the adaptive immune response. Ann. N. Y. Acad. Sci. 987: and John Casement at the Bioinformatics Unit, Newcastle University for the 249–257. microarray data analysis. 24. Luo, L., Z. Sun, Q. Fang, S. Huang, X. Bai, and G. Luo. 2013. Effects of tol- erogenic dendritic cells generated by siRNA-mediated RelB silencing on im- mune defense and surveillance functions of T cells. Cell. Immunol. 282: 28–37. Disclosures 25. Yamamoto, M., S. Yamazaki, S. Uematsu, S. Sato, H. Hemmi, K. Hoshino, The authors have no financial conflicts of interest. T. Kaisho, H. Kuwata, O. Takeuchi, K. Takeshige, et al. 2004. Regulation of Toll/ IL-1-receptor-mediated gene expression by the inducible nuclear protein IkappaBzeta. Nature 430: 218–222. 26. Lehtonen, A., H. Ahlfors, V. Veckman, M. Miettinen, R. Lahesmaa, and References I. Julkunen. 2007. Gene expression profiling during differentiation of human 1. Dieckmann, D., H. Plottner, S. Berchtold, T. Berger, and G. Schuler. 2001. Ex monocytes to macrophages or dendritic cells. J. Leukoc. Biol. 82: 710–720. vivo isolation and characterization of CD4(+)CD25(+) T cells with regulatory 27. Xiao, J., H. Zhou, N. Wu, and L. Wu. 2015. The non-canonical Wnt pathway properties from human blood. J. Exp. Med. 193: 1303–1310. negatively regulates dendritic cell differentiation by inhibiting the expansion of 2. Thornton, A. M., and E. M. Shevach. 2000. Suppressor effector function of Flt3+ lymphocyte-primed multipotent precursors. Cell. Mol. Immunol. 13: 593– CD4+CD25+ immunoregulatory T cells is antigen nonspecific. J. Immunol. 164: 604. 183–190. 28. Oderup, C., M. LaJevic, and E. C. Butcher. 2013. Canonical and noncanonical 3. von Boehmer, H. 2005. Mechanisms of suppression by suppressor T cells. Nat. Wnt proteins program dendritic cell responses for tolerance. J. Immunol. 190: Immunol. 6: 338–344. 6126–6134. 4. Misra, N., J. Bayry, S. Lacroix-Desmazes, M. D. Kazatchkine, and S. V. Kaveri. 29. Turner, M. L., F. M. Schnorfeil, and T. Brocker. 2011. MicroRNAs regulate 2004. Cutting edge: human CD4+CD25+ T cells restrain the maturation and dendritic cell differentiation and function. J. Immunol. 187: 3911–3917. antigen-presenting function of dendritic cells. J. Immunol. 172: 4676–4680. 30. Lu, C., X. Huang, X. Zhang, K. Roensch, Q. Cao, K. I. Nakayama, B. R. Blazar, 5. Bayry, J., F. Triebel, S. V. Kaveri, and D. F. Tough. 2007. Human dendritic cells Y. Zeng, and X. Zhou. 2011. miR-221 and miR-155 regulate human dendritic acquire a semimature phenotype and lymph node homing potential through in- cell development, apoptosis, and IL-12 production through targeting of p27kip1, teraction with CD4+CD25+ regulatory T cells. J. Immunol. 178: 4184–4193. KPC1, and SOCS-1. Blood 117: 4293–4303. Supplementary Table 1. List of genes upregulated in Treg-DC compared to mat-DC

Probeset ID Gene symbol Description adj.P.Val Fold-change (mat-DC: Treg-DC) ILMN_1676663 TNFRSF11B tumor necrosis factor receptor superfamily, member 11b 0.027025 0.126219 ILMN_1815023 PIM1 Pim-1 proto-oncogene, serine/threonine kinase 0.030375 0.169854 ILMN_1713397 NCCRP1 non-specific cytotoxic cell receptor protein 1 homolog (zebrafish) 0.049059 0.208023 ILMN_1744381 SERPINE1 serpin peptidase inhibitor, clade E (nexin, plasminogen activator 0.030375 0.218185 inhibitor type 1), member 1 ILMN_1747271 ATP1B2 ATPase, Na+/K+ transporting, beta 2 polypeptide 0.029882 0.228563 ILMN_1686109 CCL23 chemokine (C-C motif) ligand 23 0.033803 0.230626 ILMN_2327860 MAL mal, T-cell differentiation protein 0.043188 0.241969 ILMN_2390853 CTSH cathepsin H 0.022457 0.247506 ILMN_1738742 PLAT plasminogen activator, tissue 0.044522 0.262598 ILMN_2376205 LTB lymphotoxin beta (TNF superfamily, member 3) 0.043188 0.263951 ILMN_1800317 WNT5A wingless-type MMTV integration site family, member 5A 0.031472 0.267937 ILMN_1790881 HNMT histamine N-methyltransferase 0.027025 0.270653 ILMN_2192072 MMP7 matrix metallopeptidase 7 (matrilysin, uterine) 0.030375 0.272007 ILMN_1796094 CD36 CD36 molecule (thrombospondin receptor) 0.042418 0.278712 ILMN_1662451 FCER2 Fc fragment of IgE, low affinity II, receptor for (CD23) 0.030375 0.279015 solute carrier family 4, sodium bicarbonate cotransporter, member ILMN_2200917 SLC4A7 7 0.029882 0.281025 ILMN_2181241 RPL23AP64 ribosomal protein L23a pseudogene 64 0.030375 0.296740 ILMN_1712985 C17orf58 chromosome 17 open reading frame 58 0.031622 0.299285 ILMN_1685403 MMP7 matrix metallopeptidase 7 (matrilysin, uterine) 0.0393 0.305544 ILMN_1794803 NDP Norrie disease (pseudoglioma) 0.029882 0.306483 ILMN_1784630 KBTBD11 kelch repeat and BTB (POZ) domain containing 11 0.030375 0.316027 ILMN_2079386 RPL22 ribosomal protein L22 0.030375 0.327396 ILMN_1812191 C12orf57 chromosome 12 open reading frame 57 0.029882 0.333240 ILMN_1785646 PMP22 peripheral myelin protein 22 0.048102 0.334843 ILMN_1795304 CRLF2 cytokine receptor-like factor 2 0.043036 0.335395 ILMN_2318725 EEF1B2 eukaryotic translation elongation factor 1 beta 2 0.030375 0.335963 ILMN_1798256 UPP1 uridine phosphorylase 1 0.043799 0.339828 ILMN_2261416 CD3D CD3d molecule, delta (CD3-TCR complex) 0.043036 0.341891 ILMN_1700515 C17orf58 chromosome 17 open reading frame 58 0.043188 0.342892 ILMN_3241051 RPL32P18 ribosomal protein L32 pseudogene 18 0.036577 0.358019 Supplementary Table 1 (contd). List of genes upregulated in Treg-DC compared to mat-DC

Probeset ID Gene symbol Description adj.P.Val Fold-change (mat-DC: Treg-DC) ILMN_1756541 MXD4 MAX dimerization protein 4 0.031277 0.358351 ILMN_1723035 OLR1 oxidized low density lipoprotein (lectin-like) receptor 1 0.030375 0.360042 ILMN_1699208 NAP1L1 nucleosome assembly protein 1-like 1 0.037541 0.361912 ILMN_2388975 CERK ceramide kinase 0.036577 0.368722 ILMN_1764030 CCL23 chemokine (C-C motif) ligand 23 0.036577 0.373796 ILMN_1663119 DSC2 desmocollin 2 0.037541 0.374818 ILMN_1671116 XXYLT1 xyloside xylosyltransferase 1 0.036577 0.392739 ILMN_1656537 SNRPN small nuclear ribonucleoprotein polypeptide N 0.043188 0.394753 ILMN_2083469 IRS2 insulin receptor substrate 2 0.037541 0.398079 ILMN_1719224 LRRC75A-AS1 LRRC75A antisense RNA 1 0.037541 0.403051 ILMN_1812321 FAM131C family with sequence similarity 131, member C 0.037541 0.411638 ILMN_2276904 ALOX15B arachidonate 15-lipoxygenase, type B 0.042858 0.419460 ILMN_2211065 TMEM91 transmembrane protein 91 0.043188 0.419538 ILMN_1795715 DPYD dihydropyrimidine dehydrogenase 0.04348 0.422688 ILMN_1729318 TOR1AIP1 torsin A interacting protein 1 0.042985 0.425158 ILMN_1906187 LOC283070 uncharacterized LOC283070 0.043188 0.425711 ILMN_1815292 RPL7 ribosomal protein L7 0.042826 0.429960 ILMN_1775708 SLC2A3 solute carrier family 2 (facilitated glucose transporter), member 3 0.042826 0.431491 ILMN_1761463 EFHD2 EF-hand domain family, member D2 0.042418 0.434978 ILMN_1763386 BID BH3 interacting domain death agonist 0.043188 0.450652 ILMN_2128967 C11orf1 chromosome 11 open reading frame 1 0.043188 0.453836

Supplementary Table 2. List of genes downregulated in Treg-DC compared to mat-DC

Probeset ID Gene symbol Description adj.P.Val Fold-change (mat-DC: Treg-DC) ILMN_1791759 CXCL10 chemokine (C-X-C motif) ligand 10 0.029882 47.240283 ILMN_2067890 CXCL11 chemokine (C-X-C motif) ligand 11 0.020554 20.515951 ILMN_1739428 IFIT2 interferon-induced protein with tetratricopeptide repeats 2 0.037541 18.536452 ILMN_1701789 IFIT3 interferon-induced protein with tetratricopeptide repeats 3 0.031277 16.638422 ILMN_1801307 TNFSF10 tumor necrosis factor (ligand) superfamily, member 10 0.005969 14.780260 ILMN_1657871 RSAD2 radical S-adenosyl methionine domain containing 2 0.030375 11.475962 ILMN_2058782 IFI27 interferon, alpha-inducible protein 27 0.049059 10.575324 ILMN_1772964 CCL8 chemokine (C-C motif) ligand 8 0.027025 10.236912 ILMN_1813338 LAG3 lymphocyte-activation gene 3 0.005969 9.841286 ILMN_1745356 CXCL9 chemokine (C-X-C motif) ligand 9 0.019928 9.740089 ILMN_1674811 OASL 2'-5'-oligoadenylate synthetase-like 0.031277 9.390935 ILMN_1796409 C1QB complement component 1, q subcomponent, B chain 0.019928 8.306914 ILMN_1653466 HES4 hes family bHLH transcription factor 4 0.029882 7.420680 ILMN_1680339 PDGFRL platelet-derived growth factor receptor-like 0.037541 6.795421 ILMN_1665865 IGFBP4 insulin-like growth factor binding protein 4 0.020554 6.710888 ILMN_2233783 CD38 CD38 molecule 0.029882 6.051429 ILMN_1785902 C1QC complement component 1, q subcomponent, C chain 0.027025 6.030831 ILMN_3240420 USP18 ubiquitin specific peptidase 18 0.036692 5.843263 ILMN_1730229 CGNL1 cingulin-like 1 0.030375 5.746096 ILMN_1760062 IFI44 interferon-induced protein 44 0.045646 5.659766 ILMN_1670305 SERPING1 serpin peptidase inhibitor, clade G (C1 inhibitor), member 1 0.030375 5.427818 ILMN_1792506 PLA1A phospholipase A1 member A 0.043036 5.276571 ILMN_2235851 NEURL3 neuralized E3 ubiquitin protein ligase 3 0.037455 5.074091 ILMN_1763523 HARS histidyl-tRNA synthetase 0.029882 4.999156 ILMN_1740466 FAM46A family with sequence similarity 46, member A 0.042826 4.744491 ILMN_1711702 CLEC2D C-type lectin domain family 2, member D 0.030375 4.678188 ILMN_1686664 MT2A metallothionein 2A 0.036577 4.640315 ILMN_1672024 ISCA1P1 iron-sulfur cluster assembly 1 pseudogene 1 0.049059 4.292038 ILMN_1700268 QPRT quinolinate phosphoribosyltransferase 0.030375 4.261784 ILMN_2038775 TUBB2A tubulin, beta 2A class IIa 0.030375 4.115890 ILMN_1752592 HLA-DRB4 major histocompatibility complex, class II, DR beta 4 0.036577 3.994790 ILMN_1769129 CCL19 chemokine (C-C motif) ligand 19 0.030375 3.989847 Supplementary Table 2 (contd). List of genes downregulated in Treg-DC compared to mat-DC

Probeset ID Gene symbol Description adj.P.Val Fold-change (mat-DC: Treg-DC) ILMN_1710962 TMEM97 transmembrane protein 97 0.043188 3.910803 ILMN_1756374 SYCP1 synaptonemal complex protein 1 0.029882 3.894483 ILMN_1691156 MT1A metallothionein 1A 0.037455 3.894124 ILMN_1652065 KCNMB1 potassium large conductance calcium-activated channel, 0.029657 3.765643 subfamily M, beta member 1 ILMN_2053415 LDLR low density lipoprotein receptor 0.048102 3.740658 ILMN_1758323 ACPP acid phosphatase, prostate 0.030375 3.675483 ILMN_1808661 TOMM5 translocase of outer mitochondrial membrane 5 homolog 0.030375 3.574538 (yeast) ILMN_2287276 FAM177A1 family with sequence similarity 177, member A1 0.037455 3.505116 ILMN_1727271 WARS tryptophanyl-tRNA synthetase 0.043099 3.504966 ILMN_1790891 CKAP4 cytoskeleton-associated protein 4 0.038228 3.329990 ILMN_1769520 UBE2L6 ubiquitin-conjugating enzyme E2L 6 0.049059 3.317371 ILMN_1715931 ISCA1 iron-sulfur cluster assembly 1 0.043188 3.251217 ILMN_1720083 EHD4 EH-domain containing 4 0.029882 3.227800 ILMN_1680874 TUBB2B tubulin, beta 2B class IIb 0.036577 3.179411 ILMN_1753758 IL27 interleukin 27 0.042985 3.149064 ILMN_1698725 FRMD3 FERM domain containing 3 0.037455 3.072551 ILMN_1797728 HMGCS1 3-hydroxy-3-methylglutaryl-CoA synthase 1 (soluble) 0.030375 3.069864 ILMN_1663866 TGFBI transforming growth factor, beta-induced, 68kDa 0.037541 3.011580 ILMN_1660086 MYH11 myosin, heavy chain 11, smooth muscle 0.042985 2.987254 ILMN_2073604 EBP emopamil binding protein (sterol isomerase) 0.044845 2.981672 ILMN_1815626 DHCR7 7-dehydrocholesterol reductase 0.032789 2.946719 ILMN_1666594 IRF8 interferon regulatory factor 8 0.030375 2.880463 ILMN_1814726 SCARB2 scavenger receptor class B, member 2 0.031277 2.849468 ILMN_1677827 TLR7 toll-like receptor 7 0.037541 2.821970 ILMN_1789436 DENND1B DENN/MADD domain containing 1B 0.036577 2.802004 ILMN_3247560 BSPRY B-box and SPRY domain containing 0.030375 2.787414 ILMN_1674386 PITX1 paired-like homeodomain 1 0.045646 2.753182 ILMN_2129927 EXT1 exostosin glycosyltransferase 1 0.032789 2.737596 ILMN_1778599 SP140 SP140 nuclear body protein 0.049059 2.733013 ILMN_1760121 RRAGC Ras-related GTP binding C 0.043188 2.728264 Supplementary Table 2 (contd). List of genes downregulated in Treg-DC compared to mat-DC

Probeset ID Gene symbol Description adj.P.Val Fold-change (mat-DC: Treg-DC) ILMN_1664776 EFR3A EFR3 homolog A (S. cerevisiae) 0.043036 2.723558 ILMN_3234089 N4BP2L2 NEDD4 binding protein 2-like 2 0.031298 2.719300 ILMN_1669881 TSPAN13 tetraspanin 13 0.048216 2.718616 ILMN_1723211 L2HGDH L-2-hydroxyglutarate dehydrogenase 0.042418 2.697199 ILMN_1700831 SLC27A2 solute carrier family 27 (fatty acid transporter), member 2 0.037541 2.668341 ILMN_1802106 APOBEC3G apolipoprotein B mRNA editing enzyme, catalytic 0.038228 2.653553 polypeptide-like 3G ILMN_1682336 MASTL microtubule associated serine/threonine kinase-like 0.044845 2.652422 ILMN_2246882 SP140 SP140 nuclear body protein 0.049714 2.632900 ILMN_1725320 SIGLEC1 sialic acid binding Ig-like lectin 1, sialoadhesin 0.043188 2.610843 ILMN_1695058 SLC38A5 solute carrier family 38, member 5 0.038228 2.591680 ILMN_1758938 SLC31A2 solute carrier family 31 (copper transporter), member 2 0.037541 2.577898 ILMN_1805827 PPA1 pyrophosphatase (inorganic) 1 0.031472 2.577713 ILMN_2232478 APOBEC3G apolipoprotein B mRNA editing enzyme, catalytic 0.049059 2.551759 polypeptide-like 3G ILMN_1671554 LPIN1 lipin 1 0.036577 2.521246 ILMN_1681721 OASL 2'-5'-oligoadenylate synthetase-like 0.042826 2.474036 ILMN_1659688 LGALS3BP lectin, galactoside-binding, soluble, 3 binding protein 0.037541 2.459237 ILMN_2312719 EXOSC9 exosome component 9 0.040667 2.437670 ILMN_2135984 MASTL microtubule associated serine/threonine kinase-like 0.037541 2.402905 ILMN_1664920 C19orf12 chromosome 19 open reading frame 12 0.042826 2.398935 ILMN_2043126 CSAG3 CSAG family, member 3 0.038228 2.390559 ILMN_1745415 BBX bobby sox homolog (Drosophila) 0.038228 2.375174 ILMN_1670302 HK3 hexokinase 3 (white cell) 0.043188 2.357020 ILMN_1798061 ZFYVE26 zinc finger, FYVE domain containing 26 0.049059 2.341823 ILMN_1809259 HRASLS2 HRAS-like suppressor 2 0.042985 2.324214 ILMN_1724250 GRN granulin 0.049714 2.313150 ILMN_1684585 ACSL1 acyl-CoA synthetase long-chain family member 1 0.042418 2.284283 ILMN_1687403 MRPL40 mitochondrial ribosomal protein L40 0.043188 2.249974 ILMN_1728236 SLFN12 schlafen family member 12 0.048216 2.248819 ILMN_1703695 C19orf12 chromosome 19 open reading frame 12 0.042858 2.242688 ILMN_1773352 CCL5 chemokine (C-C motif) ligand 5 0.043188 2.202899 Supplementary Table 2 (contd). List of genes downregulated in Treg-DC compared to mat-DC

Probeset ID Gene symbol Description adj.P.Val Fold-change (mat-DC: Treg-DC) ILMN_2098126 CCL5 chemokine (C-C motif) ligand 5 0.043099 2.196082 ILMN_1669015 XPNPEP1 X-prolyl aminopeptidase (aminopeptidase P) 1, soluble 0.048216 2.105799