Reprogramming of monocytes by GM-CSF contributes to regulatory immune functions during intestinal inflammation.

Reprogramming of Monocytes by GM-CSF Contributes to Regulatory Immune Functions during Intestinal Inflammation Jan Däbritz, Toni Weinhage, Georg Varga, Timo Wirth, Karoline Walscheid, Anne Brockhausen, David Schwarzmaier, Markus Brückner, Matthias Ross, Dominik Bettenworth, Johannes Roth, Jan M. Ehrchen and Dirk Foell J Immunol published online 4 February 2015 http://www.jimmunol.org/content/early/2015/02/04/jimmun ol.1401482

Supplementary Material

http://www.jimmunol.org/content/suppl/2015/02/04/jimmunol.140148 2.DCSupplemental.html

Subscriptions

Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscriptions

Permissions Email Alerts

Submit copyright permission requests at: http://www.aai.org/ji/copyright.html Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/cgi/alerts/etoc

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 9650 Rockville Pike, Bethesda, MD 20814-3994. Copyright © 2015 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606.

Downloaded from http://www.jimmunol.org/ at Univ of Sussex Lib/Periodicals Sec on February 11, 2015

This information is current as of February 11, 2015.

Published February 4, 2015, doi:10.4049/jimmunol.1401482 The Journal of Immunology

Reprogramming of Monocytes by GM-CSF Contributes to Regulatory Immune Functions during Intestinal Inflammation

Human and murine studies showed that GM-CSF exerts beneficial effects in intestinal inflammation. To explore whether GM-CSF mediates its effects via monocytes, we analyzed effects of GM-CSF on monocytes in vitro and assessed the immunomodulatory potential of GM-CSF–activated monocytes (GMaMs) in vivo. We used microarray technology and functional assays to characterize GMaMs in vitro and used a mouse model of colitis to study GMaM functions in vivo. GM-CSF activates monocytes to increase adherence, migration, chemotaxis, and oxidative burst in vitro, and primes monocyte response to secondary microbial stimuli. In addition, GMaMs accelerate epithelial healing in vitro. Most important, in a mouse model of experimental T cell– induced colitis, GMaMs show therapeutic activity and protect mice from colitis. This is accompanied by increased production of IL-4, IL-10, and IL-13, and decreased production of IFN-g in lamina propria mononuclear cells in vivo. Confirming this finding, GMaMs attract T cells and shape their differentiation toward Th2 by upregulating IL-4, IL-10, and IL-13 in T cells in vitro. Beneficial effects of GM-CSF in Crohn’s disease may possibly be mediated through reprogramming of monocytes to simultaneously improved bacterial clearance and induction of wound healing, as well as regulation of adaptive immunity to limit excessive inflammation. The Journal of Immunology, 2015, 194: 000–000. ur concepts of immunology have changed dramatically over the past decades. The postulates of primary functions assigned to innate or adaptive immunity have been challenged by the recognition of a complex interplay between the different cellular and humoral factors that all together constitute our immune system. This helped in understanding how we are protected from infections, but it also enabled discovering key aspects of autoimmunity and chronic inflammation including regulatory mechanisms that counteract a perpetuated immune activation. Although different functions of adaptive immune cells, including regulatory T cells (Tregs), are already consolidated, our understanding of different functions of innate immune cells has only recently been enriched. As an example, phagocytes were

O

traditionally seen solely as effector cells of innate immunity promoting host defense and driving chronic inflammation. It is now accepted that monocytes can differentiate into macrophages with various activation patterns ranging from classically activated proinflammatory to anti-inflammatory phenotypes. These cells (often referred to as M1 and M2 macrophages) represent the outer margins of a broad spectrum of numerous activation and differentiation patterns of heterogeneous monocyte-derived cells (1–3). As the concepts of immunity evolve, the pathophysiology of chronic inflammatory diseases is also being revisited. As a striking example, our view of Crohn’s disease (CD) is constantly challenged. Traditionally, CD has been associated with a Th1 cytokine profile. In addition, because CD is a chronic granulomatous dis-

*Department of Pediatric Rheumatology and Immunology, University Children’s Hospital M€ unster, M€ unster 48149, Germany; †Interdisciplinary Center of Clinical Research, University Hospital M€unster, M€unster 48149, Germany; ‡Gastrointestinal Research in Inflammation & Pathology, Murdoch Children’s Research Institute, The Royal Children’s Hospital Melbourne, Parkville 3052, Victoria, Australia; x Department of Pediatrics, University of Melbourne, Melbourne Medical School, Parkville 3052, Victoria, Australia; { Department of Dermatology, University Hospital M€ unster, M€ unster 48149, Germany; ‖Institute of Immunology, University Hospital M€ unster, M€ unster 48149, Germany; and #Department of Medicine B, University Hospital M€ unster, M€unster 48149, Germany

J.D. and D.F. developed the concept, designed the experiments, and supervised the experiments; J.M.E., G.V., and J.R. gave technical support and conceptual advice; M.R., G.V., and J.D. obtained ethical approval from the competent animal welfare authorities; J.D., T. Weinhage, T. Wirth, K.W., A.B., and D.S. performed the experiments and collected data; M.R., G.V., M.B., D.B., and T. Wirth helped with animal models of experimental colitis; J.D., T. Weinhage, G.V., and D.F. analyzed the data and interpreted results; J.D. wrote the manuscript; and each author has approved the final version of the report and takes full responsibility for the manuscript.

1

J.D. and T. Weinhage contributed equally and should be considered cofirst authors.

Received for publication June 11, 2014. Accepted for publication January 4, 2015. This work was supported by the Broad Medical Research Program of the Eli and Edythe Broad Foundation (Grant IBD0201 to D.F., J.D., and J.M.E.), the German Research Foundation (Grant DFG DA1161/4-1 to J.D. and D.F., Grant DFG SU195/ 3-2 to G.V., Grant DFG SF1009B08 to M.B.), the Innovative Medical Research € ¨ 120904 and DA ¨ 3U21003 Program of the University of M€unster (Grants IMF DA to J.D. and D.F.), the Interdisciplinary Center for Clinical Research of the University of M€unster (Grant IZKF Eh2/019/11 to J.M.E.), the European Union’s Seventh Framework Programme (Grant EC-GA305266 ‘MIAMI’ to D.F.), and a research fellowship from the German Research Foundation (Grant DFG DA1161/5-1 to J.D.). Portions of this work were presented at the 50th Digestive Disease Week Annual Meeting, May 30–June 4, 2009, Chicago, IL and the 51st Digestive Disease Week Annual Meeting, May 1–5, 2010, New Orleans, LA. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1401482

The sequences presented in this article have been submitted to the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo) under accession number GSE63662. Address correspondence and reprint requests to Dr. Jan Da¨britz, Department of Pediatric Rheumatology and Immunology, University Children’s Hospital M€unster, Ro¨ntgenstrasse 21, M€unster 48149, Germany. E-mail address: [email protected] The online version of this article contains supplemental material. Abbreviations used in this article: CD, Crohn’s disease; DSS, dextran sulfate sodium; GCsM, glucocorticoid-stimulated monocyte; GMaM, GM-CSF–activated monocyte; LPMC, lamina propria mononuclear cell; LTB4, leukotriene B4; MEICS, murine endoscopic score of colitis severity; MFI, mean fluorescence intensity; MLN, mesenteric lymph node; qRT-PCR, quantitative real-time RT-PCR; ROS, reactive oxygen species; Treg, regulatory T cell. Copyright Ó 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00


Jan Da¨britz,*,†,‡,x,1 Toni Weinhage,*,1 Georg Varga,* Timo Wirth,* Karoline Walscheid,* Anne Brockhausen,{,‖ David Schwarzmaier,* Markus Br€ uckner,# Matthias Ross,# # †,‖ Dominik Bettenworth, Johannes Roth, Jan M. Ehrchen,†,{ and Dirk Foell*,†

2

Materials and Methods Human monocytes Blood samples from individual healthy donors were purchased from the Department of Transfusion Medicine at the University Hospital M€unster, M€ unster, Germany. Peripheral blood monocytes were obtained from donors by leukapheresis and isolated to .90% purity as previously described (24). Monocytes were cultured (1 3 106 cells/ml) in hydrophobic Teflon bags (Heraeus, Hanau, Germany) in McCoy’s 5a medium supplemented with 5% human AB serum, 2 mM L-glutamine, 200 IU/ml penicillin, 100 mg/ml streptomycin, and 13 nonessential amino acids (all from Biochrom, Berlin, Germany). Monocytes were allowed to rest for 16 h

before stimulation. Monocytes from at least three different individuals were assessed with each experiment.

Patients Clinical and demographic characteristics of the study subjects and methods have been reported in detail previously (25). Ethical approval was obtained from the Ethics Committee of the University of M€unster (reference no. 2006-267-f-S, obtained by Jan Da¨britz), and fully written informed consent was obtained from all patients or legal guardians.

DNA microarray hybridization Human monocytes were exposed to GM-CSF (10 ng/ml; MP Biomedicals, Santa Ana, CA) for 16 h or left untreated in three independent sets of experiments to analyze changes in gene expression patterns induced by GM-CSF. Using high-density microarrays with .22,000 oligonucleotide sets, we obtained the expression levels of at least 13,000 independent transcripts. RNA preparation, sample preparation, and hybridization to Affymetrix (Santa Clara, CA) Human Genome 133 A Gene Chip arrays for microarray analysis were performed as described previously (26).

Statistical analysis of microarray data For analysis of data from individual donors, raw data of GM-CSF–treated samples were processed by MicroArray Suite Software (Affymetrix) using data from corresponding control samples as baseline. Signals were scaled to a target intensity of 500 and log-transformed. Detection and change calls using perfect match and mismatching probes were assigned using a signed rank test as described previously (26–28). Data were submitted to the Gene Expression Omnibus database under accession number GSE63662 (http:// www.ncbi.nlm.nih.gov/geo). We retained only genes that were significantly regulated in every single experiment (change p , 0.05, fold-change $ 2.0, expression over background). The data of the complete set of experiments were further studied applying the Expressionist Suite software package (GeneData), which allows identification of genes that are significantly regulated in multiple independent experiments as described previously (26). Being aware of the low significance at low-intensity levels, we filtered for genes with an expression over background in at least one of the two experimental groups (GMaM versus monocytes). We finally retained only genes that were significantly regulated in every single experiment (change p , 0.05, fold-change $ 2.0, expression over background), as well as in the complete set of experiments (expression over background, fold-change $ 2.0, p , 0.05, paired t test). Reproducibility of the results was confirmed using RT-PCR for selected genes and three new independent experiments.

Quantitative real-time PCR Expression of selected genes in human and mouse (C57BL/6) monocytes was analyzed by quantitative real-time RT-PCR (qRT-PCR) as described previously (29). PCRs were performed and measured on a CFX384 Touch real-time PCR detection system (Bio-Rad, Munich, Germany). The relative expression was calculated using ribosomal protein L13a as endogenous housekeeping control gene. The primers used for PCR analysis are given in Supplemental Table III.

Flow cytometry FACS measurements were performed using a Cyflow space equipped with FlowMax 2.8 (both Partec, M€unster, Germany), and analysis was performed using FlowJo software (TreeStar, Ashland, OR). Ab staining of cells was routinely done with 1 mg/ml of the according Ab. For detection of cell-surface molecules, flow cytometry was performed as described earlier (26). All intracellular stains were performed using the transcription factor staining buffer set (eBioscience, San Diego, CA). mAbs used are given in Supplemental Table IV.

Chemokine production of GMaMs Chemokine concentrations of CCL18 and CCL23 were determined in cell culture supernatants of monocytes treated for 24 h with GM-CSF (10 ng/ml) or untreated control cells by an ELISA system according to the manufacturer’s instructions (CCL18; Sigma-Aldrich, Steinheim, Germany; CCL23; Raybiotech, Norcross, GA).

Monocyte/macrophage polarization Human monocytes were stimulated for 4 and 16 h with IL-4 (100 mg/ml), IFN-g (100 mg/ml), or left untreated. Alternatively, human monocytes were polarized with GM-CSF (10 ng/ml) 6 IFN-g (100 mg/ml) or left untreated as a negative control. Expression of selected genes was analyzed by qRT-PCR as described earlier.


ease and anti-inflammatory therapies targeting innate immunity have proved effective, it was a paradigm that overactive phagocytes are involved. More recently, however, emerging evidence has consolidated the view of CD as a form of innate immunodeficiency (4–6). Central to this hypothesis were the observations of diminished neutrophil accumulation in patients with CD with impaired clearance of bacteria from tissues (7, 8). The underlying problem appeared to be a primary immunodeficiency of macrophages, which secreted insufficient concentrations of proinflammatory cytokines and chemokines upon bacterial challenge (9). The view of defective macrophage functions in CD is further supported by an inappropriate mucosal healing (10, 11). Resolution of inflammation and healing relies on the infiltration of monocytes as crucial regulators of tissue repair processes (12–14). Given the changing concepts on immunity and inflammation, changes in therapeutic strategies appear as a logical consequence. As a therapy that could help in overcoming insufficient macrophage functions, GM-CSF has been shown to alleviate acute dextran sulfate sodium (DSS)-induced colitis in mice (15, 16). Even more important, it is conceivable that GM-CSF–driven modulation of innate immune cells involved in mucosal repair and/or dampening of inflammatory reactions might contribute to the benefits of GM-CSF therapy observed in some CD patients (17). These findings may link the novel concepts of monocyte biology with that of CD pathogenesis, because recent developments in immunology and genetics suggest that monocytes and their derivative cells play an important role in the pathophysiology of CD. It is noteworthy that blood monocytes are the exclusive source of macrophages in inflamed intestinal mucosa (18). Undoubtedly, monocytes carry out specific effector functions during inflammation (19). Recent studies underpin the dual function of monocytes: on one hand, the impaired monocyte function initiating CD, and on the other hand, the overactivation of monocytes and adaptive immunity maintaining the disease (20). Cells of the monocyte/macrophage lineage are characterized by considerable diversity and plasticity (21). Furthermore, monocytes can drive modulation of adaptive immunity by regulating T cell responses (22). GM-CSF functions both as a growth factor for myeloid progenitors and as a cytokine acting directly on maturing cells. Data from animal models indicate an important role in inflammation and autoimmunity, with varying consequences that likely depend on the disease-specific context (23). We thus hypothesized that GM-CSF might activate monocytes in a way that modulates their function during intestinal inflammation. To this end, we chose an unbiased but comprehensive approach taking all potential functions of GM-CSF–activated monocytes (GMaMs) into account (gene expression, innate immune functions, interplay with adaptive immunity, wound healing) rather than focusing on polarizing edges. We show in this article that beneficial effects of GM-CSF in CD could be explained by a complex reprogramming of altered monocyte/macrophage functions. These findings suggest the exploration of stimulating, rather than suppressive, therapies with the potential to more specifically reprogram monocytes to modulate immune functions.

GM-CSF–ACTIVATED MONOCYTES REGULATE COLITIS

The Journal of Immunology In additional experiments, human monocytes were polarized for 24 h with IFN-g (M1; 50 ng/ml) or IL-4 (M2; 50 ng/ml). After polarization, monocytes were stimulated with GM-CSF (10 ng/ml) or left untreated for an additional 24 h. IL-1b, TNF-a, IL-10, and CD206 expression were measured by flow cytometry as described earlier. Finally, human monocytes were stimulated 6 GM-CSF (10 ng/ml) for 0, 30, 60, and 120 min, and expression of IFN-a, IFN-b, IFN-g, and IL-4 was analyzed by qRT-PCR as described earlier.

Migration, chemotaxis, monocyte trafficking, and adherence

GM-CSF priming For priming experiments, human monocytes were stimulated for 24 h with GM-CSF (10 ng/ml) or left untreated. After pretreatment, monocytes were stimulated with medium containing LPS (10 ng/ml) or left untreated for an additional 4 h. TNF-a and IL-1b content were measured in culture supernatants by ELISA (OptEIA ELISA kits; BD Pharmingen, Heidelberg, Germany). Expression of selected genes was confirmed by qRT-PCR as described earlier.

Phagocytosis and oxidative burst For detection of phagocytic capacity, cells were incubated with carboxyfluorescein diacetate (Invitrogen, Karlsruhe, Germany)–labeled Leishmania major parasites (ratio cells: L. major = 1:5) or FITC (MoBiTec, Go¨ttingen, Germany)-labeled latex beads (ratio cells: beads = 1:10) for 4 h (31). Rate of phagocytosis was determined by flow cytometry as described previously (30). Cells were incubated with or without PMA (50 nM; Sigma-Aldrich, Taufkirchen, Germany) in addition to 10 ng/ml GM-CSF to investigate the induction of oxidative burst. The extracellular chemiluminescence response was measured in the presence of isoluminol (50 mM; SigmaAldrich, Taufkirchen, Germany) as described previously (24).

In vitro scratch closure assay Cells of the Caco-2 human colon adenocarcinoma cell line (ATCC HTB-37) were cultured in DMEM supplemented with 10% FBS, penicillin (100 U/ml), streptomycin (100 mg/ml), 15 mM HEPES (pH 7.4), 2 mM L-glutamine, and 1% nonessential amino acids at 37˚C and 5% CO2 in a humidified incubator. For the scratch closure assay, cells were grown to confluence in 12-well plates and serum deprived (0.1% FBS) for 24 h before scratch wounding. Monolayers were scratched using a sterile pipette tip and washed twice. Thereafter the wounded monolayers were cultured in fresh serum-deprived medium in the presence or absence of 2.5 3 105 untreated monocytes or GMaMs. The initial wound size was determined by microscopy, and the area of the scratch was calculated with ImageJ software (Version 1.45s; National Institutes of Health). Additional photographs were taken using a reference line 24 h after wounding, and the rate of wound closure was

analyzed by measuring the scratch area relative to the initial wound area after each time point.

Influence of human GMaMs on T cell fate Human T cells were purified from donor-specific PBMCs using positive selection of CD2-expressing T cells by MACS technology according to the manufacturer’s protocol (Miltenyi Biotec, Bergisch-Gladbach, Germany). A total of 1 3 106 T cells were cocultured with 1 3 105 monocytes (Mo) for 7 d (ratio T/Mo = 10:1). Cells were harvested and stained using mAbs raised against CD4, CD25, and Foxp3 (Supplemental Table IV). Subsequent flow cytometry was performed as described earlier.

Mice Experiments were performed in accordance with approved protocols of the animal welfare committee of the North Rhine-Westphalia State Agency for Nature, Environment and Consumer Protection, Recklinghausen, Germany (LANUV NRW Reference No. 87-51.04.2010.A113). C57BL/6 and Rag12/2 mice were kept under specific pathogen-free conditions and according to federal regulations. Mice were purchased from Harlan (Paris, France) and used for experiments at the age of 10–12 wk.

Murine monocytes Freshly isolated monocytic bone marrow cells were prepared as described earlier (32). Cells were cultured for 48 h with 150 U/ml GM-CSF (Immunotools) or left untreated as control in 20% L929 cell supernatant (containing M-CSF) conditioned DMEM supplemented with 2 mM glutamine, 0.1 mM nonessential amino acids (all Invitrogen, Karlsruhe, Germany), 100 mg/ml penicillin/streptomycin, and 10% heat-inactivated FCS (both Biochrom, Berlin, Germany). After culture, cells were washed three times and subsequently used for analyses and coculture experiments.

T cell transfer colitis To induce colitis, we adoptively transferred 1 3 106 syngeneic CD4+CD252 T cells i.v. into Rag12/2 mice (on C57BL/6 background). Body weight of animals was monitored daily, and around day 40 animals that established colitis by weight loss on consecutive days received GMaMs or untreated monocytes (2 3 106 per mouse) i.v. Alternatively, 5 mg GM-CSF (Immunotools) diluted in PBS or PBS alone was administered i.p. on a daily basis. Body weight of mice was monitored for an additional 12 d. Finally, mice were euthanized by CO2 inhalation, and their colons were prepared, measured, and preserved for histology.

Isolation of murine T cells from spleen T cells were isolated from spleens as described previously (33). T cells used for induction of transfer colitis were further purified for CD4+ and depleted of CD25+ cells by MACS technology according to the manufacturer’s instructions (Miltenyi Biotech).

Histopathologic analysis For histopathologic analysis, tissue specimens from the proximal and distal colon were fixed in 10% buffered formalin phosphate and embedded in paraffin. Sections were cut at 3–5 mm and stained with H&E. Inflammation was graded from 0 to 4 in a blinded fashion: 0, no signs of inflammation; 1, low leukocyte infiltration; 2, moderate leukocyte infiltration; 3, high leukocyte infiltration, moderate fibrosis, high vascular density, thickening of the colon wall, moderate goblet cell loss, and focal loss of crypts; and 4, transmural infiltrations, massive loss of goblet cell, extensive fibrosis, and diffuse loss of crypts.

High-resolution colonoscopy Mice were anesthetized with isoflurane (100% v/v, 1.5 vol %, 1.5 L/min; Florene; Abbott, Wiesbaden, Germany) and administered an enema (FrekaClyss; Fresenius Kabi, Se`vres, France). High-resolution colonoscopy was performed using a veterinary endoscopy workstation (Coloview; Karl Storz, Tuttlingen, Germany) to assess colitis. Under visual control, the rigid miniature endoscope (1.9-mm outer diameter) was inserted ∼4 cm according to anatomic conditions. The modified murine endoscopic score of colitis severity (MEICS) observes thickening of the colon, changing of vascular pattern, presence of fibrin, granular mucosa surface, and stool consistence (0–3 points each, maximum of 15 points); it was used to evaluate colonic inflammation (34).

In vivo cell tracking For in vivo cell tracking of GMaMs, cells were stained with a commercially available lipophilic tracer 1,1-dioctadecyl-3,3,3,3-tetramethylindotricarbocyanine


Monocyte assays in Transwell plates (Costar, New York, NY) were performed as described previously using MCP-1 (10 ng/ml; Immunotools, Friesoythe, Germany), IL-8 (25 ng/ml; Immunotools), and leukotriene B4 (LTB4; 100 nM; Biozol, Eching, Germany) (30). Cells were allowed to migrate for 4 h. T cell migration was analyzed using the Cultrex 96-well cell migration assay according to the manufacturer’s protocol (Trevigen, Gaithersburg, MD). T cells were isolated from fresh PBMCs using an EasySep human T cell enrichment kit according to the manufacturer’s protocol (STEMCELL Technologies, Vancouver, BC, Canada). T cells (5 3 104) were added to top-chamber and cell culture supernatants of untreated monocytes or GMaMs as well as CCL18 (1 ng/ml), CCL23 (8 ng/ml; both PeproTech, Rocky Hill, NJ), and medium (control) were added to the bottom chamber. T cells that had migrated into the lower compartment within 4 h were measured using an Infinite M200 Pro reader (TECAN, Crailsheim, Germany). Expression of intestinal-associated homing molecules in human monocytes treated 6 GM-CSF (10 ng/ml) for 24 h was analyzed by flow cytometry (gated on CD14+ cells) as described earlier. For determination of cell adhesion, monocytes (2 3 105) were stimulated with GM-CSF (10 ng/ml) for 24 h or left untreated. Monocytes were seeded in triplicates into 96-well flat-bottom plastic tissue-culture plates and incubated at 37˚C and 7% CO2 for 4 h. Nonadhering cells were removed by washing twice; remaining adherent cells were fixed with 2% glutaraldehyde (Sigma-Aldrich, Taufkirchen, Germany) for 10 min. Wells were washed two times with H2O and subsequently stained with 0.5% crystal violet (Merck, Darmstadt, Germany) in 2% EtOH (pH 6.0) for an additional 15 min at room temperature. Finally, wells were washed three times and cells were lysed. Ten percent acetic acid was added, and staining was quantified measuring the OD at 560 nm using an Asys Expert 96 Microplate ELISA reader (Anthos Mikrosysteme, Krefeld, Germany) (25).

3

4 iodide (Life Technologies, Darmstadt, Germany) with an emission maximum of 782 nm as described elsewhere (35). In vivo distribution of labeled cells across the intestine was studied 24 h after i.v. injection using a planar smallanimal FMT system (FMT 2500; VisEn Medical) as described previously (35).

Isolation of lamina propria mononuclear cells and cells from mesenteric lymph nodes

Coculture of naive T cells and restimulation of LPMCs and MLNs Naive T cells were isolated from splenocytes using a pan T cell kit II (Miltenyi Biotec) as described previously (37), and 1 3 105 T cells were cocultured for 4 d (ratio 10:1) with respective monocytes in triplicates for each condition in anti-CD3e and anti-CD28 Abs (5 mg/ml each; Supplemental Table IV) coated to 96-well, round-bottom plates in RPMI 1640 supplemented with 2 mM glutamine, 0.1 mM nonessential amino acids (all Invitrogen, Karlsruhe, Germany), 100 mg/ml penicillin/streptomycin, and 10% heat-inactivated FCS (both Biochrom, Berlin, Germany) at 37˚C and 5% CO2. To further test for cytokine production, we stimulated isolated LPMCs and MLN cells (2 3 105 per well) from mice used in transfer colitis experiments with anti-CD3e and anti-CD28 Abs (5 mg/ml each; Supplemental Table IV) coated to 96-well, round-bottom plates for 24 (LPMCs) or 96 h (MLNs). A total of 100 ml cell supernatants was stored until cytokine analysis was performed using a bead-based multiplex assay (mouse Th1/Th2 10plex FlowCytomix; eBioscience), according to manufacturer’s instructions.

Induction of Tregs by murine GMaMs LPMC and MLN single-cell suspensions from mice used in transfer colitis experiments were stained with Abs raised against CD4 and Foxp3 (Supplemental Table IV). Cells were measured by flow cytometry as described earlier. T cell coculture experiments were performed to evaluate the function of GMaM to induce Tregs. Therefore, 1 3 105 splenic T cells (described earlier) were cocultured with 1 3 104 monocytes (GM-CSF–activated or untreated) in 96-well, round-bottom plates for 7 d in RPMI 1640 supplemented with 2 mM glutamine, 0.1 mM nonessential amino acids (all Invitrogen, Karlsruhe, Germany), 100 mg/ml penicillin/streptomycin, and 10% heat-inactivated FCS (both Biochrom, Berlin, Germany) at 37˚C and 5% CO2. Cells were harvested and stained with 1 mg anti-CD4 and antiFoxp3. Used Ab clones are given in Supplemental Table IV. Flow cytometry was performed as described earlier.

Statistics Data are expressed as mean 6 SEM unless stated otherwise and were assessed using the Student t test. The p values ,0.05 were considered to be statistically significant. All calculations were performed using SPSS version 14 (SPSS, Chicago, IL).

Results Gene expression and phenotype shift of GMaMs GM-CSF provokes nonclassical monocyte activation. To analyze monocyte activation comprehensively and unbiased, we performed a global RNA expression analysis. The microarray data were filtered using strict statistical criteria and revealed a significant regulation of genes involved in immune/inflammatory responses

(especially chemokines), cell motility, chemotaxis, regulation of cell growth, endocytosis, and Ag processing and presentation (Table I). Furthermore, we analyzed statistical overrepresentation of transcription factor binding sites in GM-CSF–regulated genes using CARRIE software (Supplemental Table I). Overall, we found that in human monocytes, 190 genes were significantly upregulated, whereas 212 were downregulated after 16-h stimulation with GM-CSF (Supplemental Table II). Raw data have been submitted to the Gene Expression Omnibus under accession number GSE63662 (http://www.ncbi.nlm.nih.gov/geo). We confirmed microarray expression data by qRT-PCR for selected genes (Fig. 1A). Analyses of expression levels at different time points of GM-CSF activation confirmed that gene regulation in GMaM is most relevant after 16–24 h (Fig. 1B). However, we also show that the expression of most of these genes is already significantly regulated after 4 h of GM-CSF stimulation. Genes that were not significantly regulated after 24 h of GM-CSF activation also showed no significant upregulation or downregulation of their expression after 4, 16, 48, or 120 h (Fig. 1B). In agreement with the microarray data, flow cytometry confirmed upregulation of CD80 and downregulation of CD9 upon GM-CSF activation (Fig. 1C, 1D). GM-CSF drives monocytes toward M2-like phenotype. Interestingly, the microarray expression data of GMaM chemokine ligands and receptors showed a gene expression pattern indicative of a GMCSF–induced shift toward a M2-like phenotype (Fig. 2A). This was confirmed by protein quantification (ELISA) for selected chemokines (Fig. 2B) and by gene expression analyses (qRT-PCR) for differentially expressed genes that were assigned to M1- and M2like monocytes based on currently accepted annotations (Table II) (2, 38, 39). Gene expression in GMaM is similar to IL-4–induced gene expression profiles. Our gene expression analyses (Fig. 1B) suggest that the observed gene expression in GMaM is similar to IL-4–induced gene expression profiles in monocytes. We stimulated human monocytes from healthy donors with or without GMCSF for 0, 30, 60, and 120 min and analyzed gene expression of IFN-a, IFN-b, IFN-g, and IL-4 by qRT-PCR. The expression of IFNs and IL-4 in monocytes was not upregulated at any time after GM-CSF stimulation when compared with untreated monocytes (data not shown). Interestingly, we found that the GM-CSF–induced gene expression pattern in human monocytes within the first 24 h is similar to the gene expression pattern of IL-4– but not IFN-g–stimulated monocytes (Fig. 2C, 2D). We assume that our transcriptomic data of human GMaMs rather reflect a primary effect, which, however, has similarities with the expression pattern of IL-4–induced M2a macrophages. It has been shown that endogenous type I IFN regulates the basal gene expression of bone marrow–derived macrophages grown in GM-CSF (40). Fifty GM-CSF–specific type I IFN–dependent regulated genes were identified, of which we found only five to be significantly regulated in GMaMs (IFIT3, CBX6, ISG20, ABCA9, COLEC12; Supplemental Table II). In human cells, it was found that the expression of 154 type I IFN–regulated genes were also different between monocyte-derived macrophages cultured in GM-CSF (GM-MDM) or M-CSF (MDM) (41). We found that only 7 of the top 50 type I IFN–dependent genes differentially expressed between human GM-MDM and MDM were significantly regulated in GMaMs (EDNRB, LGMN, P2RY14, CH25H, CDKN1C, GGTLA1, CD69; Supplemental Table II). In addition, we found no significant GM-CSF–dependent gene regulation of type I IFNs in GMaMs. Likewise, the gene expression of IFN-g in human monocytes was not


Lamina propria mononuclear cells (LPMCs) were isolated from the colon of colitogenic mice by a standard method (36). In brief, the colon was removed, opened longitudinally and cut into 5-mm pieces, and washed with cold Ca2+/Mg2+-free HBSS. The intestinal tissue specimens were transferred into HBSS with EDTA to remove intraepithelial lymphocytes. After 30 min of gentle shaking at 37˚C, the samples were vortexed and intraepithelial lymphocyte–containing supernatant was removed. This step was repeated twice. LPMC suspensions were prepared from the EDTAtreated de-epithelialized intestinal tissue by further incubation with 100 U/ml collagenase and 5 U/ml DNase for 30 min at 37˚C. LPMCs were washed, resuspended in 44% Percoll solution (Amersham Pharmacia Biotech, Piscataway, NJ), underlaid with 66% Percoll solution, and centrifuged for 30 min at 600 3 g. The LPMC fraction was harvested from the interface. For cell isolation from mesenteric lymph nodes (MLNs), the lymph nodes of treated mice and control mice were carefully isolated, pooled, and passed through a 40-mm cell strainer, and the resulting single-cell suspension was washed once with PBS.


The Journal of Immunology

5

Table I. Selected genes upregulated and downregulated by GM-CSF activation Gene

Description (NCBI Gene)

n-Fold

p

Upregulated by GM-CSF Activation

g-Glutamyltransferase-like activity 1 Complement factor H CD80 molecule Protein C receptor, endothelial (EPCR) CD69 molecule B-cell scaffold protein with ankyrin repeats 1 Signaling lymphocytic activation molecule IL 7 receptor C-type lectin domain family 5, member A IL 1 receptor accessory protein Arachidonate 5-lipoxygenase-activating protein Lymphotoxin b (TNF superfamily, member 3) PG-endoperoxide synthase 1

50.5 44.9 8.4 7.5 6.4 5.2 5.0 4.8 4.5 4.4 4.1 3.6 3.3

,0.001 0.013 ,0.001 ,0.001 0.002 0.011 0.041 0.001 0.011 0.008 0.003 0.002 0.003

Chemokine (C-C motif) ligand 13 Chemokine (C-C motif) ligand 23 Proplatelet basic protein (CXCL7) Chemokine (C-X-C motif) ligand 5 Chemokine (C-C motif) ligand 17 IL 8 receptor, b Chemokine (C-C motif) receptor 6 Sialophorin (leukosialin, CD43)

16.7 8.3 7.2 5.8 5.5 5.1 4.3 3.8

,0.001 0.002 0.001 0.025 0.003 0.002 ,0.001 0.007

18.4 15.5 9.8 6.4

,0.001 0.008 ,0.001 ,0.001

Transforming growth factor, b2 Fibroblast growth factor 13 Cytokine inducible SH2-containing protein

26.8 19.8 5.7

0.006 0.018 ,0.001

Fc fragment of IgG, high affinity Ib, receptor (CD64) Aquaporin 9 IFN-induced protein with tetratricopeptide repeats 1 Guanylate binding protein 5 Myxovirus (influenza virus) resistance 1 CD28 molecule 2’,59-oligoadenylate synthetase 1, 40/46kDa 2’-59-oligoadenylate synthetase 2, 69/71kDa Heparanase Guanylate binding protein 2, IFN-inducible Monoglyceride lipase

29.0 25.3 25.0 25.0 24.2 24.2 24.0 23.7 23.6 23.5 23.4

0.011 0.004 0.037 0.009 0.023 0.008 0.049 0.043 0.003 0.002 ,0.001

Chemokine (C-X-C motif) ligand 12 Chemokine (C-X-C motif) ligand 11 Chemokine (C-X-C motif) ligand 13 Chemokine (C-X-C motif) receptor 4 Chemokine (C-X-C motif) ligand 10

216.9 27.1 26.4 26.2 23.5

0.009 0.046 0.019 0.001 0.007

Fc fragment of IgG, high affinity Ia, receptor (CD64) Macrophage scavenger receptor 1

27.7 24.7

0.022 0.041

Cell adhesion molecule 1 CD9 molecule Insulin-like growth factor 1 (somatomedin C) Integrin, b 8 Nidogen 1 KIT ligand Leptin (obesity homolog, mouse) Cyclin-dependent kinase inhibitor 1C (p57, Kip2) Signal transducer and activator of transcription 1

28.5 26.4 25.0 25.0 24.7 23.1 29.1 23.5 23.2

0.001 0.012 ,0.001 0.006 ,0.001 0.002 ,0.001 0.041 0.034

CD1c CD1b CD1e CD1a

molecule molecule molecule molecule

Downregulated by GM-CSF Activation

Immune response FCGR1B AQP9 IFIT1 GBP5 MX1 CD28 OAS1 OAS2 HPSE GBP2 MGLL Chemokines CXCL12 CXCL11 CXCL13 CXCR4 CXCL10 Phagocytosis/Endocytosis FCGR1A MSR1 Other CADM1 CD9 IGF1 ITGB8 NID1 KITLG LEP CDKN1C STAT1

significantly regulated by GM-CSF after 16 h nor after 24 h (Table I, II). In addition, GM-CSF induced neither IRF4 (M2 polarization) nor IRF5 (M1 polarization) in human monocytes when activated with GM-CSF for 24 h (Table II). However,

activation of NF-kB, which promotes M1 macrophage polarization, was significantly downregulated by GM-CSF stimulation, whereas the activity of C/EBPb, which is crucial for expression of M2-regulated genes, was significantly upregulated


Inflammatory response GGTLA1 CFH CD80 PROCR CD69 BANK1 SLAMF1 IL7R CLEC5A IL1RAP ALOX5AP LTB PTGS1 Chemotaxis CCL13 CCL23 PPBP CXCL5 CCL17 IL8RB CCR6 SPN Ag processing and presentation CD1C CD1B CD1E CD1A Regulation of cell growth TGFB2 FGF13 CISH

6


by GM-CSF stimulation in the first 24 h (Supplemental Table I). In addition, further quantitative PCR analysis suggests that IFN-g does not contribute to our transcriptomic data provided for GMaMs. The gene expression of chemokines (CXCL10, CXCL11, CCL1, CCL13, CCL23, CXCL5), CD206, CD209, and IL-1b in human monocytes was specifically and significantly upregulated or downregulated by either GM-CSF or IFN-g after 24 h of stimulation (data not shown). Finally, we analyzed whether GM-CSF is affecting distinct monocyte subsets within the overall monocyte population differentially. Our results showed a homogenous shift of GM-CSF– induced cell-surface markers within the overall monocyte population using FACS (data not shown). Thus, we found no evidence that GM-CSF is affecting distinct monocyte subpopulations. Nevertheless, we have performed additional monocyte/macrophage polarization experiments. To this end, we stimulated human monocytes from healthy donors for 24 h with or without IFN-g (M1 macrophage polarization) or with or without IL-4 (M2 macrophage polarization), and cells were subsequently stimulated with or without GM-CSF for further 24 h. Expression of IL-1b, TNF-a, IL-10, and CD206 was analyzed by flow cytometry (Fig. 2E). The data suggest that GM-CSF is affecting both M1- and M2-polarized monocytes and that GM-CSF is not affecting distinct monocyte subsets within the overall population differentially.

Comprehensive characterization of GMaM innate immune functions GM-CSF promotes adherence and migration of monocytes. Major functions of monocytes include the capacity to adhere and migrate, which is crucial for their recruitment into tissue. Adherence of GMaMs to plastic surfaces was enhanced after 24 and 48 h compared with untreated control cells (Fig. 3A). We tested whether GM-CSF activation would affect migration and chemotaxis of monocytes in general and also specifically in response to MCP1/CCL2, IL-8/ CXCL8, and LTB4. By using a modified Boyden chamber assay, we detected that spontaneous migration of GMaM and migration toward MCP1 and LTB4 were significantly enhanced after 4 h (Fig. 3B). The chemotactic effect was specific because migration did not occur when MCP1 or LTB4 was added to the upper compartment of the Boyden chamber (data not shown). In addition, GM-CSF–induced increase in chemotaxis was also specific to the stimulus because we did not find increased migration toward IL-8 (Fig. 3B). Integrins and CC chemokine receptors play a potential role in monocyte trafficking into the mucosa in the context of mucosal homeostasis at the intestinal epithelial barrier. These molecules are also known to play a role in the pathogenesis of human inflammatory bowel diseases. Expression of integrins and CC chemokine receptors was analyzed by flow cytometry gated on


FIGURE 1. Confirmation of GM-CSF–regulated gene expression in monocytes (microarray data) by real-time PCR and flow cytometry. (A) Results obtained from microarray analysis of GM-CSF–dependent gene regulation in human monocytes (after 16 h) compared with unstimulated monocytes were confirmed by qRT-PCR. Genes analyzed were: lymphotoxin b (LTB); complement factor H (CFH); g-glutamyltransferase-like activity 1 (GGTLA); the chemokines CXCL10, CXCL12, CCL23, and CCL13; and the CD1c and CD80 molecule. Shown are the mean relative n-fold regulation (6 SEM) of three independent experiments. (B) GM-CSF–dependent gene expression analysis by qRT-PCR of selected genes at different time points. Shown is the mean relative n-fold regulation (6 SEM) of three independent experiments. (C and D) Expression of selected cell-surface molecules (C, CD80; D, CD9) found to be differentially expressed in GMaMs by microarray analysis, confirmed by flow cytometry. Specific profiles are shown by thick lines; isotype controls appear as thin lines. Numbers show the quotient of specific/isotype control MFI. The experiment was repeated three times with similar results, and the differences in MFI shifts between control and GMaMs were in accordance with microarray data. *p , 0.05, **p , 0.01, ***p , 0.001 compared with untreated monocytes.


7

CD14+ cells, and expression is stated as mean fluorescence intensity (MFI; geo-mean) 6 SEM of three independent experiments. Expression of CCR2 and CCR6 was significantly increased in GM-CSF–stimulated monocytes (CCR2 150.2 6 31.1; CCR6 24.5 6 4.4) compared with untreated cells (CCR2 88.4 6 22.5, p , 0.05; CCR6 10.0 6 0.4, p , 0.05), whereas expression of CCR7 was significantly reduced in GM-CSF–treated monocytes (22.2 6 1.6) compared with unstimulated cells (49.2 6 6.9, p , 0.05). Expression of b7, CCR1, CCR4, CCR9, and CX3CR1 was similar in GM-CSF–stimulated and untreated cells (data not shown). Production of reactive oxygen species is increased and phagocytosis is unimpaired in GMaMs. Another important function of monocytes after being recruited, for example, to sites of infection or defective barriers, is the phagocytosis and killing of pathogens (e.g., via production of reactive oxygen species [ROS]). Spontaneous and PMA-induced production of ROS was significantly enhanced in GMaMs (Fig. 3C). A few molecules involved in phagocytosis/endocytosis were significantly downregulated by GM-CSF stimulation (Table I). We therefore tested phagocytosis of latex beads and of complement opsonized living L. major parasites after activation of monocytes with GM-CSF. We detected

no significant difference in phagocytosis of latex beads by GMaMs, and also phagocytosis of L. major promastigotes was not altered compared with control monocytes (data not shown). GM-CSF primes the monocyte response to a secondary microbial stimulus. To test the hypothesis that GM-CSF specifically activates monocyte functions by augmenting anti-infectious/antimicrobial defense and bacterial clearance, we analyzed GMaMs for an increased response to a secondary microbial stimulus. Therefore, we analyzed the influence of GM-CSF stimulation (24 h) on cytokine production and gene expression in human monocytes after 4 h of costimulation with bacterial endotoxin (LPS). Compared with control monocytes, GMaMs exposed to LPS produced significantly more IL-1b and TNF-a (Fig. 3D). LPS stimulation of GMaMs also resulted in a much more pronounced expression of inflammatory genes as revealed by qRT-PCR (Fig. 3E). This confirms a GM-CSF–induced priming effect on monocytes, leading to an increase in vitro response to other stimuli (42). GMaM impact on epithelial healing and on adaptive immunity GMaMs accelerate epithelial healing. In addition to their innate phagocytic and killing activity in antimicrobial defense, monocytes are also involved in wound repair. Because we observed phenotypic


FIGURE 2. Polarization of GMaMs. (A) The gene expression of chemokine ligands and receptors in GMaMs (16 h) was analyzed by microarray analysis. The status of differentiation and polarization was classified according to characteristics of classically activated M1-like and alternatively activated M2-like monocyte/macrophage subsets in humans. (B) Microarray analysis data were confirmed by protein quantification (ELISA) for two selected chemokines (CCL 18 and CCL 23). (C and D) IL-4– (C) and IFN-g–dependent (D) gene expression analysis (RT-PCR) in human monocytes at different time points. Bars represent the relative n-fold regulation (mean 6 SEM) of three independent experiments. *p , 0.05, **p , 0.01, ***p , 0.001 compared with untreated monocytes. (E) Effects of GM-CSF activation on already primed monocytes. Monocytes were primed for 24 h toward M1 or M2 and subsequently treated for 24 h 6 GM-CSF. Expression (mean 6 SEM) of IL-1b, TNF-a, IL-10, and CD206 is shown gated on CD14+ cells of three independent experiments. *p , 0.05, **p , 0.01, ***p , 0.001.

8


Table II. M1-/M2-like differentiation and polarization of GMaMs Gene Symbol

M1/M2

Regulation

n-Fold

p

Chemokine (C-X-C motif) ligand 10 Chemokine (C-X-C motif) ligand 11 Chemokine (C-X-C motif) ligand 13 IFN, g TNF a TNF ligand superfamily NO synthase 2, inducible B7-1 IL 1 b IL 6 IL 8 IL 12 IL 18 IL 23 IFN regulatory factor 5 PG-endoperoxidase synthase 2 Colony stimulating factor 2 Colony stimulating factor 3 Chemokine (C-X-C motif) ligand Chemokine ligand 1 Chemokine ligand 13 Chemokine ligand 23 TGF b Arginase Mannose receptor C type 1 (MRC1) CD163 molecule DC-SIGN IL 1 receptor, type 2 IL 10 IFN regulatory factor 4

CXCL10 CXCL11 CXCL13 IFNg TNFa TRAIL iNOS CD80 IL1b IL6 IL8 IL12 IL18 IL23 IRF5 COX2 GM-CSF G-CSF CXCL5 CCL1 CCL13 CCL23 TGFb ARG1 CD206 CD163 CD209 CD121b IL10 IRF4

M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M1 M2 M2 M2 M2 M2 M2 M2 M2 M2 M2 M2 M2

↓ ↓ ↓↓ ↔ ↑ ↓ ↔ ↑ ↑ ↑ ↔ ↔ ↔ ↔ ↔ ↔ ↓ ↓ ↑ ↑↑ ↑↑↑ ↑↑ ↔ ↑ ↑↑ ↔ ↑ ↑ ↔ ↔

22.36 22.75 25.27 21.20 4.50 22.03 1.94 4.22 2.81 2.42 1.15 1.43 1.59 21.05 21.07 21.03 21.36 23.81 3.87 6.82 21.06 6.87 1.12 3.90 6.45 21.90 2.65 2.70 21.25 1.01

0.005 0.004 0.000 0.700 0.000 0.001 0.230 0.005 0.000 0.005 1.000 0.700 0.100 1.000 1.000 1.000 0.000 0.000 0.015 0.000 0.000 0.000 0.694 0.005 0.000 0.113 0.009 0.001 0.206 1.000

similarities between GMaMs and alternatively activated (M2-like) macrophages, which have been originally described as “woundhealing macrophages,” we analyzed the influence of GMaMs on epithelial healing. In epithelial cell (Caco-2) monolayers, preactivation of monocytes with GM-CSF significantly accelerated wound closure compared with unstimulated monocytes (Fig. 3F). GMaMs attract T cells and induce Tregs. Another important function of monocytes is the cross talk to adaptive immunity. Monocytes and macrophages serve as APCs, but they also shape lymphocyte activation by a whole battery of different costimulatory molecules and cytokines. As shown in Fig. 2, gene expression (Fig. 2A) and protein production (Fig. 2B) of chemokines CCL18 and CCL23 were strongly increased in GMaMs. Because CCL18 and CCL23 have been shown to attract naive and resting T cells (43, 44), we analyzed the capacity of GMaMs to attract T cells by using a modified Boyden chamber. T cells were allowed to migrate toward the culture supernatants of GMaMs or untreated monocytes. We observed a significantly increased T cell migration toward cell supernatants of GMaMs (Fig. 3G). In addition, we have addressed effects of GM-CSF on Treg differentiation. To this end, we stimulated human monocytes with GMCSF and cocultured monocytes and autologous T cells for 7 d, and analyzed resulting T cells for Foxp3 expression to evaluate Treg differentiation. CD25 and Foxp3 expression in T cells were already increased upon interaction with monocytes that had been stimulated with GM-CSF for only 24 h. The induction of Foxp3 expression in T cells cocultured with GMaMs was further increased when monocytes were stimulated for 48 h with GM-CSF (Fig. 3H, 3I). In vivo effects of GMaMs in experimental T cell–induced colitis GMaMs alleviate CD4+ T cell–induced colitis. Having confirmed a specific activation pattern of human monocytes in response to GM-CSF, with the augmentation of host immune defense functions,

tissue repair capacities, and a positive effect on T cell recruitment, we sought to address the functionality of these cells in vivo in the context of CD. Because systematic analyses in the human system are not feasible, we went on analyzing the effects of GMaMs in the murine system. Analysis by qRT-PCR showed that murine GMaMs had a similar gene regulation profile when compared with the GM-CSF–dependent gene expression in human monocytes (Fig. 4A). We chose the CD4+ T cell–dependent experimental colitis model as an acceptable surrogate of human CD. In this model, adoptive transfer of syngeneic CD4+CD252 T cells into Rag12/2 mice (which lack mature T cells) induces severe colitis (45). The onset of colitis is monitored clinically by weight loss. Untreated monocytes (control) or ex vivo GMaMs were injected i.v. after mice had lost weight on consecutive days (∼5–6 wk after eliciting colitis by injection of CD4+CD252 T cells). Migration of injected GMaMs to the inflamed gut was confirmed by in vivo imaging. In agreement with the in vitro data on cell migration, we observed an increased infiltration of GMaMs into MLNs compared with control monocytes in two independent experiments (Fig. 5). Mice that had received GMaMs showed no weight loss at all over a period of 12 d after monocyte transfer (Fig. 4B). Normally, the inflamed colon becomes shorter and presents with reduced length, and thus shortening of the colon is a measure of inflammation. Also, in this study, mice that received GMaMs did not show relevant shortening of the colon, whereas all other groups were not protected from colitis (Fig. 4B, 4C). Control mice that had received untreated monocytes or no monocytes showed progressive weight loss and signs of intestinal inflammation, and had to be euthanized on day 12 (Fig. 4B, 4C). Animals that had received untreated monocytes had significantly severe histopathologic alterations of the colon, most evident in the distal part (Fig. 4D, 4E). In addition, we performed high-resolution colonoscopy and graded inflammation (MEICS-Score) (34). The colon of mice receiving no treatment presented with a vulnerable and bleeding mucosa, rarefication of


Gene Name


9

vascular pattern, presence of fibrin, and ulcerations. Mice treated with GMaMs depicted a transparent colonic mucosa with a regular vascular pattern resembling healthy animals (Fig. 4F). Taken together, treatment of an established CD4+ T cell–induced colitis with GMaMs alleviates inflammation of the colon, resulting in significantly improved clinical parameters and histology, suggesting that GMaMs potentially exert regulatory effects on T cells in vivo. GMaMs regulate T cell responses in vivo and in vitro. After termination of colitis experiments, LPMCs and MLNs were harvested. Migration of injected GMaMs to gut tissue and MLNs was confirmed by in vivo imaging (Fig. 5). Single-cell suspensions were restimulated with anti-CD3/anti-CD28 to explore how the capacity and strength of T cell cytokine production has changed in vivo

during colitis and treatment with GMaMs. After restimulation for 24 h, we tested supernatants for production of IFN-g (Th1 cell response) and IL-4, IL-10, and IL-13 (Th2 cell response). As shown in Fig. 6, GMaM transfer led to significantly reduced IFN-g production in T cells from LPMCs (Fig. 6A) and MLNs (Fig. 6B). The Th2 cytokines IL-4, IL-10, and IL-13, however, were slightly increased in supernatants from LPMCs of animals treated with GMaMs (Fig. 6A), and IL-4 and IL-13 were also increased in supernatants from MLNs of animals treated with GMaMs (Fig. 6B). In summary, treatment of mice suffering from colitis with GMaM results in a shift in cytokine production of T cells in vivo. To confirm these in vivo data, we performed coculture experiments with anti-CD3e/anti-CD28–stimulated T cells in vitro. Also, in this study, GMaMs skewed the T cell response and led


FIGURE 3. Functional properties of GMaMs and interaction of human GMaMs with T cells. (A) Human monocytes were activated with GM-CSF or left untreated (control) for 24/48 h in Teflon bags and subsequently allowed to adhere to multiwell plates for 4 h. Adherent cells were stained with crystal violet, and staining was quantified measuring the OD at 560 nm. (B) Monocytes were activated with GM-CSF or left untreated for 24 h in Teflon bags and placed into the upper chamber of a Transwell filter. The lower chamber contained monocyte medium with the addition of LTB4, MCP-1/CCL2, IL-8/CXCL8, or no attractants. After 4 h cells that had migrated into the lower compartment were counted, and numbers are presented as the percentage of cells, which migrated in the absence of any chemotactic stimulus. The p values refer to the migration of untreated (control) cells in the absence of any chemotactic stimuli (w/o). (C) Oxidative burst of GMaMs or untreated monocytes was initiated by the addition of PMA. Isoluminol chemiluminescence was measured in PMA-treated cells and control cells after induction of oxidative burst. (D) Cytokine secretion was measured in supernatants of GMaMs (24 h) after exposure to LPS (4 h) and compared with untreated monocytes exposed to LPS. **p , 0.01, ***p , 0.001 compared with LPS-treated control. (E) The gene expression of GMaMs (24 h) after exposure to LPS (4 h) was assessed by qRT-PCR. (F) Shown is the extent of wound closure in scratch assays of Caco-2 monolayers at 24 and 48 h in the absence of monocytes (control; n = 30), the presence of untreated monocytes (monocytes; n = 45), or the presence of monocytes preactivated for 48 h with GM-CSF (GMaM; n = 45). (G) T cell migration was analyzed. Monocytes were activated with GM-CSF or left untreated for 24 h in Teflon bags. Fifty thousand T cells were added to top chamber, and cell culture supernatants of untreated monocytes or GMaMs, as well as CCL18, CCL23, and medium (control), were added to the bottom chamber. T cells that had migrated into the lower compartment within 4 h were counted. *p , 0.05, **p , 0.01, ***p , 0.001 compared with untreated monocytes. (H and I) Human autologous T cells were cocultured with untreated monocytes (control) or GMaMs (24 and 48 h) at a ratio of 10:1 for 7 d. Cells were stained for CD4, CD25, and intracellular for Foxp3 expression and analyzed by flow cytometry. (H) Representative dot plots are shown for 48 h. (I) Cells were gated on CD4+ cells and analyzed for CD25 and Foxp3 expression. Data shown are the means (6 SEM) of three independent experiments. (A–I) *p , 0.05, **p , 0.01, ***p , 0.001.

10


to significant upregulation of Th2 cytokines IL-4, IL-13, and IL-10, whereas the Th1 cytokine IFN-g was downregulated (Fig. 6C). These data demonstrate that GMaM cross talk with T cells results in a phenotype shift that attenuates classical Th1 responses, which may contribute to an immunomodulatory effect. We did not observe an increase of Tregs, represented by Foxp3 expression in CD4+ T cells, in MLNs or LPMCs of mice with experimental T cell transfer colitis after the transfer of GMaMs (Fig. 6D, 6E). However, we were able to demonstrate that murine GMaMs induce Tregs in vitro (Fig. 6F). Peripheral blood monocytes from patients with CD behave like GMaMs We next studied phenotypic and functional features of untreated versus GM-CSF–activated peripheral blood monocytes of 18 patients with quiescent CD by analyses of cell adherence, migration, chemotaxis, phagocytosis, oxidative burst, and cytokine expression and secretion. Collectively, our data suggest that the effects of GM-CSF activation of peripheral monocytes of patients with CD (Fig. 7) are similar to the observed effects in GMaMs from healthy donors (Figs. 1–3). This includes the GM-CSF–induced increase in adherence, migration, chemotaxis, and oxidative burst, as well as the priming of monocytes to secondary microbial stimuli (Fig. 7A–E). In addition, changes in GM-CSF–dependent mRNA expression of selected key inflammatory cytokines were in agreement with our transcriptomic data obtained from GMaMs of healthy individuals (Fig. 7F). Importantly, there was no evidence that GM-CSF activation had different effects on monocytes when compared between individual patients.

Discussion Despite the fact that the concept of CD as a chronic granulomatous Th1-driven disease shifts toward a theory of CD as an immunodeficiency of macrophages, while we only begin to understand the regulatory or suppressive functions of our immune system, our general approach to chronic inflammatory diseases including CD is still mainly based upon the paradigm of immunosuppression as the primary therapeutic intervention. In line with that, phagocytes are primarily seen as driving forces of inflammation that need to be inhibited. This traditional view of immune interventions, however, is in sharp contrast with our currently changing view of immunity. It is now accepted that cells of the monocyte–macrophage lineage are characterized by considerable diversity and plasticity that may encompass, as an example, classical M1-macrophage differentiation (when stimulated by IFN-g) or alternative M2 differentiation (when stimulated by IL-4/IL-13) as outer margins of a broad phenotypical plasticity (2). Serving another example, the population resulting from GM-CSF–stimulated human monocytes has been referred to as M1like macrophages with a proinflammatory cytokine profile (41, 46). As an attempt to introduce a novel concept based on stimulating rather than suppressing immunity, GM-CSF has been used both in animal IBD models and in human patients with CD (47). Intraperitoneal administration of GM-CSF alleviated acute DSSinduced colitis in mice, resulting in decreased proinflammatory cytokine release, improved clinical and histologic parameters, as well as more rapid ulcer healing, and facilitated epithelial regeneration (15, 16). Importantly, transfer of splenic GM-CSF– induced CD11b+ myeloid cells into DSS-exposed mice improved


FIGURE 4. Treatment with GMaMs protects from experimental colitis. (A) GM-CSF–dependent gene regulation in murine monocytes derived from bone marrow of C57BL/6 mice and human peripheral blood monocytes compared with unstimulated monocytes. Shown is the mean relative n-fold regulation (6 SEM) of three independent experiments. *p , 0.05 compared with GM-CSF–treated human monocytes. (B) Rag12/2 mice were injected i.v. with CD4+CD252 T cells. After 40 d, when weight loss of the animals was severe, we injected: 1) GMaM i.v., or 2) untreated monocytes i.v., or 3) GM-CSF i.p. (daily for 7 consecutive days), or, as a control, 4) PBS i.v. or 5) PBS i.p. Body weight of mice was subsequently monitored daily for 12 d. (C) On day 12, colons were removed for histology. The graph shows the mean colon lengths of each experimental group. (D) Representative macroscopic and microscopic (H&E staining) images of mice with colitis injected with control monocytes or GMaMs. Original magnification 3100. (E) Intestinal inflammation scores of the proximal and distal colon of mice with colitis injected with control monocytes or GMaM (0, no inflammation; 1, mild inflammation; 2, moderate inflammation; 3, severe inflammation; 4, extreme inflammation). (F) MEICS score and representative pictures of high-resolution colonoscopy showing the colon of a mouse receiving no treatment with a vulnerable and bleeding mucosa, rarefication of vascular pattern, fibrin, and ulcerations. Mice treated with GMaMs depicted a transparent mucosa with a regular vascular pattern resembling healthy animals. Graphs in (B)–(E) show mean values (6 SEM) of 10 control mice and 14 mice injected with GMaMs from 3 independent experiments. *p , 0.05, **p , 0.01 compared with untreated monocytes; #p , 0.05 compared with T cells only.


11

colitis, and GM-CSF–expanded CD11b+ splenocytes were shown to promote in vitro wound repair (16). Furthermore, it has been shown that: 1) neutralization of GM-CSF increases intestinal permeability and bacterial translocation in mice; and 2) increased levels of GM-CSF autoantibodies are associated with an increase in bowel permeability, disease relapse, stricturing ileal disease, and surgery in patients with CD (48–50). As a therapy that could help in overcoming insufficient macrophage functions, GM-CSF had strikingly beneficial effects in subgroups of CD patients (17). Seemingly, these beneficial effects were rather unexpected in light of the previously reported proinflammatory polarization of macrophages upon stimulation with GM-CSF (41, 46). We speculated that GM-CSF exerts its beneficial effects in intestinal inflammation in vivo by specific activation of monocytes that combines innate immune activation, thus facilitating antiinfectious defense, with a simultaneous regulatory function serving to limit adaptive immunity and excessive inflammation. We thus set off in an unbiased systems biology approach to comprehensively study the many facets of monocyte activation in vitro, ranging from gene expression to innate immune functions and the interplay with adaptive immunity. All these aspects have previously been studied on monocytes but separately and independently from each other (41, 46, 51–54). Our findings suggest that the early imprinting of monocytes after activation with GM-CSF is of crucial importance, because monocytes play an important role during the recruitment phase of the innate immune response and have the potential to regulate adaptive immune mechanisms. GMCSF has been shown to have a pleiotropic role in inflammation and autoimmunity (23). Data from other groups suggest that cell culture conditions, concentration, time point, and duration chosen for GM-CSF stimulation of human monocytes may determine transcriptional outcomes relating to M1/M2 polarity (53–55). In this respect, it has been described that different biologic responses

induced by GM-CSF depend on its concentration (53), and that the time point chosen for the CSF treatment of human monocytes can markedly determine the relative expression of cytokine genes (41). Collectively, it is conceivable that the described population of GMaMs in this study represents an intermediate cell type, combining cell-surface expression characteristics and functional features of different M2 macrophage subsets including CD206 and CD209 cell-surface expression (M2a), Th2 responses/activation (M2a/b), killing, and type II inflammation (M2a) and immunoregulation (M2b). In contrast with proinflammatory and antimicrobial responses of classically activated monocytes, M2-like phenotypes are broadly anti-inflammatory and play important roles in wound healing (54). GMaMs combine: 1) features of augmented host defense functions; 2) the ability to facilitate epithelial healing; and 3) the regulatory potential on adaptive immunity. Specifically, we found a GMaM-dependent accelerated wound closure in Caco-2 monolayers using an in vitro scratch closure assay and in addition an upregulation of genes involved in cell proliferation (e.g., FGF13, CDKN1C, TGF-b). Furthermore, we found that the expression of a number of genes that are associated with M2 polarization is increased in human monocytes after activation with GM-CSF. In particular, we found a significant regulation of chemokines and chemokine receptors in human monocytes. In this study, we found a GM-CSF–dependent downregulation of the M1 chemokines CXCL9, CXCL10, CXCL11, CXCL13, and CXCL16 in monocytes. At the same time, GM-CSF significantly induced expression of chemokines CXCL3, CCL1, CCL13, CCL17, CCL18, CCL23, and CCL24 in monocytes, which are characteristic for M2 macrophages (39, 54, 56). In addition, we showed that a number of other M2a/b macrophage markers were significantly upregulated in human monocytes after GM-CSF activation (e.g., CD206, CD209, CD121b, ARG1). As mentioned earlier, we also found an enhanced production of proinflammatory cytokines (IL-1b, TNF-a,


FIGURE 5. GMaMs rapidly infiltrate the intestine. (A) Mesenteric lymph node single-cell suspensions of congenic CD45.2 mice, suffering from colitis (and treated as indicated), were analyzed for infiltration of injected donor monocytes (CD45.1+) using CD45.1 Ab by flow cytometry. (B) Results of two independent experiments are shown as percent infiltrated donor CD45.1+ cells. *p , 0.05 compared with untreated monocytes. (C) GMaMs were labeled with 1,1-dioctadecyl-3,3,3,3-tetramethylindotricarbocyanine iodide, and 2 3 106 cells were injected according to the standard treatment regimen. Monocyte infiltration in the intestine was visualized after 24 h by a planar small-animal fluorescence-mediated tomography system, and representative pictures are depicted.

12


and IL-6) in GMaMs, a feature also seen in M2b macrophages upon exposure to immune complexes and LPS (56). In addition to these strengthened innate immune functions, GMaMs simultaneously indeed have a regulatory potential on adaptive immunity. Overall, our data indicate that GMaMs represent a distinctive cell population with characteristics of both M1 and M2 cells. GM-CSF also stimulated functions that are typically assigned to classically activated monocytes. It has previously been published that GM-CSF increases the adherence of purified peripheral blood monocytes to plastic surfaces and to monolayers of HUVECs (55), and that GM-CSF can prime monocytes for increased transendothelial migration (57). Furthermore, reported results on GMCSF effects on other functions such as oxidative metabolism, cytotoxicity, phagocytosis, and the in vitro response to other stimuli are conflicting to some extent (23, 52, 58, 59). We confirm in this study that short-term treatment (24 h) of human monocytes with GM-CSF promotes: 1) cell adherence and migration, 2) production of ROS, and 3) response to a second microbial stimulus (LPS). Thus, GM-CSF enhances selective effector functions of monocytes, an effect of GM-CSF previously described for tissue-derived macrophages (60). In addition, our data indicate that GM-CSF–activated peripheral blood monocytes from patients with CD behave the same way as GMaMs (Fig. 7) and that GMCSF may regulate the homing molecules CCR2 and CCR6, which are involved in regulating several aspects of mucosal immunity, including the ability to mediate the recruitment of innate immune cells to the sites of epithelial inflammation (61, 62).

Monocytes may also significantly regulate the immune context by interaction with other cells. In particular, chemokines and chemokine receptors have a key role in intestinal epithelial barrier repair and maintenance (63, 64). We found a significant regulation of chemokines and chemokine receptors with GM-CSF–dependent downregulation of the chemokines CXCL9, CXCL10, and CXCL11 in monocytes. These factors are known to be increased in IBD and to attract Th1 and NK cells (65). At the same time, GM-CSF significantly induced a short-termed expression of the chemokines CCL13, CCL17, CCL18, CCL23, and CCL24 in monocytes, which are known to attract naive T cells, Th2 cells, and/or Tregs (43, 44, 63, 65). Because of the observed upregulation of costimulatory molecule CD80 and the chemotactic factors for naive and quiescent T cells CCL18 and CCL23 (43, 44), we next analyzed the interaction of GMaMs with T cells. Indeed, migration of naive, autologous T cells toward GMaMs was accelerated. Our data suggest that particularly CCL18 and CCL23 might be responsible for the increased T cell migration. However, our transcriptomic data suggest that other GM-CSF–induced chemokines (e.g., CCL13, CCL17, CCL24) might also be responsible for the increased T cell migration because they are known to attract T cells. Studies to address this question more specifically are beyond the scope of this study. It has been shown that treatment of human monocytes with GM-CSF generates a subtype of cells that regulate CD4+ T cell proliferation partially via production of IL-10 (52), and that GM-CSF may sustain Treg homeostasis and enhance their suppressive functions (47), indi-


FIGURE 6. GMaMs alter T cell cytokine responses in vivo and in vitro. LPMCs and MLNs were isolated from the colon of mice that were treated as indicated. Expression of cytokines after 24-h stimulation with anti-CD3e/anti-CD28 Abs is shown for LPMCs (A) and after 96 h for MLNs (B). Graphs show mean values (6 SEM) from 3 independent experiments and n = 6–8 per group. (C) Naive pan T cells were cocultured for 4 d with GMaM, control monocytes, or left alone. T cells were stimulated with anti-CD3e/anti-CD28 Abs. Ratio of respective monocytes to T cells was 1:10, and cytokines were measured in supernatants. Data shown are the means (6 SEM) of nine independent experiments. *p , 0.05, **p , 0.01, ***p , 0.001 compared with untreated monocytes. Cell populations from the lamina propria (D) and MLNs (E) were stained for CD4 and Foxp3 expression and analyzed by flow cytometry. Bars refer to mean 6 SEM of three independent experiments. (F) Murine T cells were cocultured with GMaMs or control monocytes at a ratio of 10:1 (T cells/monocytes). Cells were stained for CD4 and Foxp3 expression and analyzed by flow cytometry. Bars refer to mean 6 SEM of three independent experiments. **p , 0.01 compared with untreated monocytes (control).


13

cating the regulatory potential of GMaMs toward CD4+ T cells and Tregs. When we cocultured GMaMs with syngeneic CD4+ T cells, we observed that GMaMs shape T cell response toward a Th2 phenotype and induce Tregs. After having analyzed the programming of monocytes with GM-CSF in vitro, we next aimed at analyzing the therapeutic effects of this cell population in a model of CD. We found that T and B cell–deficient (Rag12/2) mice in which Crohn-like colitis was induced with the CD4+CD252 T cell transfer model (66, 67) were protected from disease progression when they received GMaMs (but not untreated monocytes). Interestingly, Rag12/2 mice that did not receive GMaMs but i.p. GM-CSF injections after the transfer of T cells were not completely protected from colitis but showed reduced disease severity. This is in agreement with earlier work demonstrating positive effects of GM-CSF administration (i.p.) in DSS-induced colitis in BALB/c, and more importantly, in Rag12/2 mice (15, 16). We postulate that the alleviating effects of GM-CSF in experimental colitis are due to direct modulation of monocyte/macrophage functions including accelerated epithelial healing. Our data demonstrate that the protective effect of monocytes depends on their GM-CSF prestimulation in a T cell– dependent model of colitis. The therapeutic mechanism of action of GMaM thus involves the regulation of T cell responses, which

includes a down-toning of classical Th1 responses. In this regard, our results reconfirm that monocytes harbor important functions regarding polarization and expansion of lymphocytes and may also contribute to shaping T cell responses (22). The in vivo effects of GMaMs showed increased levels of Th2 cytokines in LPMCs (IL-4, IL-10, IL-13) and MLNs (IL-4, IL-13), but this trend was not statistically significant. However, together with the observed significantly reduced production of IFN-g in T cells from LMPCs and MLNs, we concluded that treatment of mice suffering from colitis with GMaMs results in a shift toward Th2 cytokine production of T cells in vivo. Our in vitro experiments confirmed that GMaMs skew the T cell response and lead to significant upregulation of Th2 cytokines (IL-4, IL-10, IL-13), whereas the Th1 cytokine IFN-g was significantly downregulated. This might be explained by our observation that GMaMs display characteristics of both M2-like/IL-4–induced macrophages and M1-like/ IFN-g–induced macrophages (as discussed earlier). Interestingly, it has recently been shown that the therapeutic transfer of glucocorticoid-stimulated monocytes (GCsMs) in the T cell transfer colitis model also resulted in a strongly downregulated release of IFN-g by T cells from LPMCs and MLNs. The production of IL-4 and IL-13 was not influenced in single-cell suspensions from LPMCs and MLNs after treatment of mice


FIGURE 7. Features of GM-CSF–activated peripheral blood monocytes of patients with quiescent CD (n = 18). (A) Adhesion of untreated (w/o) versus GM-CSF–activated patient monocytes (24 h) to fibronectin-coated plastic surface. Adhering cells were stained with 0.5% crystal violet, and staining was quantified measuring the OD at 560 nm. (B) Migration and chemotaxis studies of untreated (w/o) versus GMaMs (24 h) using a modified Boyden chamber and LTB4 (100 nM) as an additional chemoattractant. After 4 h, cells that had migrated into the lower compartment were counted, and numbers are presented as the percentage of untreated cells, which migrated in the absence of any chemotactic stimuli. (C) Phagocytosis of fluorescein-labeled E. coli by untreated (w/o) versus GMaMs (24 h). E. coli phagocytosis was analyzed by flow cytometry, and phagocytic internalization was confirmed by fluorescence microscopy. (D) Production of ROS by untreated (w/o) versus GMaMs (24 h) with and without further LPS stimulation for 2 h in the presence of rhodamine for the final 15 min. Oxidative burst was analyzed by flow cytometry. (E) Cytokine secretion of untreated (w/o) versus GMaMs (24 h) with and without further LPS stimulation for 2 h. (F) Gene expression (relative n-fold regulation) in untreated (w/o) versus GMaMs (24 h). Bars refer to mean 6 SEM. *p , 0.05, **p , 0.01, ***p , 0.001 compared with untreated monocytes; ###p , 0.001 compared with controls without LTB4 or LPS.

14

Acknowledgments We thank Melanie Saers, Susanne Schleifenbaum, Andrea Dick, Claudia Sole´, Eva Nattkemper, Andrea Stadtba¨umer, and Sonja Dufentester for excellent technical work. We also thank Dr. Tilmann Spieker (Department of Pathology, University Hospital M€unster, Germany, and St. FranziskusHospital, Institute of Pathology, M€unster, Germany), who helped with histopathology. We also thank the American Gastroenterological Association Research Foundation for the Moti L and Kamla Rustgi Awards, which supported the presentation of the research at the 50th and 51st Digestive Disease Week annual meetings.

Disclosures The authors have no financial conflicts of interest.

References 1. Shi, C., and E. G. Pamer. 2011. Monocyte recruitment during infection and inflammation. Nat. Rev. Immunol. 11: 762–774. 2. Sica, A., and A. Mantovani. 2012. Macrophage plasticity and polarization: in vivo veritas. J. Clin. Invest. 122: 787–795. 3. Lawrence, T., and G. Natoli. 2011. Transcriptional regulation of macrophage polarization: enabling diversity with identity. Nat. Rev. Immunol. 11: 750–761. 4. Marks, D. J. 2011. Defective innate immunity in inflammatory bowel disease: a Crohn’s disease exclusivity? Curr. Opin. Gastroenterol. 27: 328–334. 5. Casanova, J. L., and L. Abel. 2009. Revisiting Crohn’s disease as a primary immunodeficiency of macrophages. J. Exp. Med. 206: 1839–1843. 6. Rahman, F. Z., D. J. Marks, B. H. Hayee, A. M. Smith, S. L. Bloom, and A. W. Segal. 2008. Phagocyte dysfunction and inflammatory bowel disease. Inflamm. Bowel Dis. 14: 1443–1452. 7. Segal, A. W., and G. Loewi. 1976. Neutrophil dysfunction in Crohn’s disease. Lancet 2: 219–221. 8. Marks, D. J., M. W. Harbord, R. MacAllister, F. Z. Rahman, J. Young, B. AlLazikani, W. Lees, M. Novelli, S. Bloom, and A. W. Segal. 2006. Defective acute inflammation in Crohn’s disease: a clinical investigation. Lancet 367: 668– 678. 9. Smith, A. M., F. Z. Rahman, B. Hayee, S. J. Graham, D. J. Marks, G. W. Sewell, C. D. Palmer, J. Wilde, B. M. Foxwell, I. S. Gloger, et al. 2009. Disordered macrophage cytokine secretion underlies impaired acute inflammation and bacterial clearance in Crohn’s disease. J. Exp. Med. 206: 1883–1897. 10. Maloy, K. J., and F. Powrie. 2011. Intestinal homeostasis and its breakdown in inflammatory bowel disease. Nature 474: 298–306. 11. Ja¨ger, S., E. F. Stange, and J. Wehkamp. 2013. Inflammatory bowel disease: an impaired barrier disease. Langenbecks Arch. Surg. 398: 1–12. 12. Norling, L. V., and M. Perretti. 2013. Control of myeloid cell trafficking in resolution. J. Innate Immun. 5: 367–376. 13. Ortega-Go´mez, A., M. Perretti, and O. Soehnlein. 2013. Resolution of inflammation: an integrated view. EMBO Mol. Med. 5: 661–674. 14. Mantovani, A., S. K. Biswas, M. R. Galdiero, A. Sica, and M. Locati. 2013. Macrophage plasticity and polarization in tissue repair and remodelling. J. Pathol. 229: 176–185. 15. Sainathan, S. K., E. M. Hanna, Q. Gong, K. S. Bishnupuri, Q. Luo, M. Colonna, F. V. White, E. Croze, C. Houchen, S. Anant, and B. K. Dieckgraefe. 2008. Granulocyte macrophage colony-stimulating factor ameliorates DSS-induced experimental colitis. Inflamm. Bowel Dis. 14: 88–99. 16. Bernasconi, E., L. Favre, M. H. Maillard, D. Bachmann, C. Pythoud, H. Bouzourene, E. Croze, S. Velichko, J. Parkinson, P. Michetti, and D. Velin. 2010. Granulocyte-macrophage colony-stimulating factor elicits bone marrowderived cells that promote efficient colonic mucosal healing. Inflamm. Bowel Dis. 16: 428–441. 17. Roth, L., J. K. MacDonald, J. W. McDonald, and N. Chande. 2012. Sargramostim (GM-CSF) for induction of remission in Crohn’s disease: a cochrane inflammatory bowel disease and functional bowel disorders systematic review of randomized trials. Inflamm. Bowel Dis. 18: 1333–1339. 18. Smythies, L. E., A. Maheshwari, R. Clements, D. Eckhoff, L. Novak, H. L. Vu, L. M. Mosteller-Barnum, M. Sellers, and P. D. Smith. 2006. Mucosal IL-8 and TGF-beta recruit blood monocytes: evidence for cross-talk between the lamina propria stroma and myeloid cells. J. Leukoc. Biol. 80: 492–499. 19. Geissmann, F., M. G. Manz, S. Jung, M. H. Sieweke, M. Merad, and K. Ley. 2010. Development of monocytes, macrophages, and dendritic cells. Science 327: 656–661. 20. Zhou, L., H. Braat, K. N. Faber, G. Dijkstra, and M. P. Peppelenbosch. 2009. Monocytes and their pathophysiological role in Crohn’s disease. Cell. Mol. Life Sci. 66: 192–202. 21. Auffray, C., M. H. Sieweke, and F. Geissmann. 2009. Blood monocytes: development, heterogeneity, and relationship with dendritic cells. Annu. Rev. Immunol. 27: 669–692. 22. Geissmann, F., C. Auffray, R. Palframan, C. Wirrig, A. Ciocca, L. Campisi, E. Narni-Mancinelli, and G. Lauvau. 2008. Blood monocytes: distinct subsets, how they relate to dendritic cells, and their possible roles in the regulation of T-cell responses. Immunol. Cell Biol. 86: 398–408. 23. Hamilton, J. A. 2008. Colony-stimulating factors in inflammation and autoimmunity. Nat. Rev. Immunol. 8: 533–544. 24. Ehrchen, J., L. Steinm€uller, K. Barczyk, K. Tenbrock, W. Nacken, M. Eisenacher, U. Nordhues, C. Sorg, C. Sunderko¨tter, and J. Roth. 2007. Glucocorticoids induce differentiation of a specifically activated, antiinflammatory subtype of human monocytes. Blood 109: 1265–1274. 25. Schwarzmaier, D., D. Foell, T. Weinhage, G. Varga, and J. Da¨britz. 2013. Peripheral monocyte functions and activation in patients with quiescent Crohn’s disease. PLoS ONE 8: e62761. 26. Viemann, D., M. Goebeler, S. Schmid, K. Klimmek, C. Sorg, S. Ludwig, and J. Roth. 2004. Transcriptional profiling of IKK2/NF-kappa B- and p38 MAP kinase-dependent gene expression in TNF-alpha-stimulated primary human endothelial cells. Blood 103: 3365–3373. 27. Viemann, D., A. Strey, A. Janning, K. Jurk, K. Klimmek, T. Vogl, K. Hirono, F. Ichida, D. Foell, B. Kehrel, et al. 2005. Myeloid-related proteins 8 and 14 induce a specific inflammatory response in human microvascular endothelial cells. Blood 105: 2955–2962. 28. Quackenbush, J. 2002. Microarray data normalization and transformation. Nat. Genet. 32(Suppl.): 496–501. 29. Da¨britz, J., F. Friedrichs, T. Weinhage, J. Hampe, T. Kucharzik, A. L€ugering, U. Broeckel, S. Schreiber, T. Spieker, M. Stoll, and D. Foell. 2011. The functional -374T/A polymorphism of the receptor for advanced glycation end


with injection of GCsMs, which is also in contrast with the in vitro system, where cytokine production of IFN-g, IL-4, and IL-13 by T cells was significantly regulated in cocultures with GCsMs (68). The same study also showed that in the T cell transfer colitis model, CD4+Foxp3+ Tregs accumulate locally in the colon after treatment with GCsMs and that repetitive stimulation of naive splenic T cells with GCsMs induces Tregs in vitro. However, Tregs also did not expand in draining MLNs and LPMCs of animals treated with GCsMs during T cell transfer colitis (68). Collectively, GMaMs induce the differentiation of CD4+Foxp3+ Tregs in vitro, but we found no evidence that the GMaM-mediated protection from colitis is facilitated via expansion of mucosal Tregs in vivo. Limitations of our study include the lack of in vivo data exploring specifically GMaMs in human CD, which may become feasible in the future. Although we address the tissue context by analyzing LPMCs and MLN, the further differentiation of ex vivo– activated monocytes after transfer in vivo, especially after infiltrating the inflamed tissue, is certainly a complex issue. The circulating monocyte population is not homogenous but consists of both inflammatory and regulatory populations that counterbalance each other. We speculate that GM-CSF exerts its beneficial effects in intestinal inflammation in vivo by specific activation of monocytes that combines innate immune activation, facilitating anti-infectious defense and a simultaneous regulatory function serving to limit adaptive immunity and excessive inflammation rather short term. The pleiotropic GM-CSF functions on monocyte activation and their consequences for innate and adaptive immunity range from activation of M2-like monocytes, chemotactic migration, and antimicrobial response to mucosal healing, but also encompasses regulation of adaptive immunity by attraction of, for example, T cells with the possibility to differentiate Th cells and subsequently limit inflammation induced by adaptive immunity. However, it remains a question for further studies whether and how monocytes differentiate long term at sites of inflammation and within damaged tissue. It is also important to investigate the immunomodulatory properties of monocytes in other animal models of experimental colitis, for example, in DSS-induced colitis. Recent work by Kurmaeva et al. (69) has demonstrated that immunosuppressive monocytes (CD11b+Ly6GnegLy6Chigh cells) accumulate in the spleen and inflamed intestine during experimental colitis not only in T cell–induced colitis but also in an acute model induced by administration of DSS and a TNFDARE model of chronic ileitis. In conclusion, while taking the shifting paradigms of CD pathogenesis and immune regulation into account, our findings support the exploration of stimulating rather than suppressive therapies with the potential to more specifically reprogram monocytes toward immunomodulatory functions to alleviate chronic inflammatory bowel disease.



30.

31.

32.

33.

34.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54. 55.

56.

57.

58. 59.

60.

61. 62. 63.

64.

65.

66.

67.

68.

69.

colony-stimulating factor autoantibodies and increased intestinal permeability in Crohn disease. J. Pediatr. Gastroenterol. Nutr. 52: 542–548. Da¨britz, J., E. Bonkowski, C. Chalk, B. C. Trapnell, J. Langhorst, L. A. Denson, and D. Foell. 2013. Granulocyte macrophage colony-stimulating factor autoantibodies and disease relapse in inflammatory bowel disease. Am. J. Gastroenterol. 108: 1901–1910. Sierra-Filardi, E., A. Puig-Kro¨ger, F. J. Blanco, C. Nieto, R. Bragado, M. I. Palomero, C. Bernabe´u, M. A. Vega, and A. L. Corbı´. 2011. Activin A skews macrophage polarization by promoting a proinflammatory phenotype and inhibiting the acquisition of anti-inflammatory macrophage markers. Blood 117: 5092–5101. Ganguly, D., K. Paul, J. Bagchi, S. Rakshit, L. Mandal, G. Bandyopadhyay, and S. Bandyopadhyay. 2007. Granulocyte-macrophage colony-stimulating factor drives monocytes to CD14low CD83+ DCSIGN- interleukin-10-producing myeloid cells with differential effects on T-cell subsets. Immunology 121: 499–507. Guthridge, M. A., J. A. Powell, E. F. Barry, F. C. Stomski, B. J. McClure, H. Ramshaw, F. A. Felquer, M. Dottore, D. T. Thomas, B. To, et al. 2006. Growth factor pleiotropy is controlled by a receptor Tyr/Ser motif that acts as a binary switch. EMBO J. 25: 479–489. Murray, P. J., and T. A. Wynn. 2011. Protective and pathogenic functions of macrophage subsets. Nat. Rev. Immunol. 11: 723–737. Gamble, J. R., M. J. Elliott, E. Jaipargas, A. F. Lopez, and M. A. Vadas. 1989. Regulation of human monocyte adherence by granulocyte-macrophage colonystimulating factor. Proc. Natl. Acad. Sci. USA 86: 7169–7173. Mantovani, A., A. Sica, S. Sozzani, P. Allavena, A. Vecchi, and M. Locati. 2004. The chemokine system in diverse forms of macrophage activation and polarization. Trends Immunol. 25: 677–686. Shang, X. Z., and A. C. Issekutz. 1999. Enhancement of monocyte transendothelial migration by granulocyte-macrophage colony-stimulating factor: requirement for chemoattractant and CD11a/CD18 mechanisms. Eur. J. Immunol. 29: 3571–3582. Armitage, J. O. 1998. Emerging applications of recombinant human granulocytemacrophage colony-stimulating factor. Blood 92: 4491–4508. Robin, G., S. Markovich, A. Athamna, and Y. Keisari. 1991. Human recombinant granulocyte-macrophage colony-stimulating factor augments viability and cytotoxic activities of human monocyte-derived macrophages in long-term cultures. Lymphokine Cytokine Res. 10: 257–263. Coleman, D. L., J. A. Chodakewitz, A. H. Bartiss, and J. W. Mellors. 1988. Granulocyte-macrophage colony-stimulating factor enhances selective effector functions of tissue-derived macrophages. Blood 72: 573–578. Ito, T., W. F. Carson, IV, K. A. Cavassani, J. M. Connett, and S. L. Kunkel. 2011. CCR6 as a mediator of immunity in the lung and gut. Exp. Cell Res. 317: 613–619. Thomas, S., and D. C. Baumgart. 2012. Targeting leukocyte migration and adhesion in Crohn’s disease and ulcerative colitis. Inflammopharmacology 20: 1–18. Zimmerman, N. P., R. A. Vongsa, M. K. Wendt, and M. B. Dwinell. 2008. Chemokines and chemokine receptors in mucosal homeostasis at the intestinal epithelial barrier in inflammatory bowel disease. Inflamm. Bowel Dis. 14: 1000– 1011. Tokuyama, H., S. Ueha, M. Kurachi, K. Matsushima, F. Moriyasu, R. S. Blumberg, and K. Kakimi. 2005. The simultaneous blockade of chemokine receptors CCR2, CCR5 and CXCR3 by a non-peptide chemokine receptor antagonist protects mice from dextran sodium sulfate-mediated colitis. Int. Immunol. 17: 1023–1034. Koelink, P. J., S. A. Overbeek, S. Braber, P. de Kruijf, G. Folkerts, M. J. Smit, and A. D. Kraneveld. 2012. Targeting chemokine receptors in chronic inflammatory diseases: an extensive review. Pharmacol. Ther. 133: 1–18. Powrie, F., J. Carlino, M. W. Leach, S. Mauze, and R. L. Coffman. 1996. A critical role for transforming growth factor-beta but not interleukin 4 in the suppression of T helper type 1-mediated colitis by CD45RB(low) CD4+ T cells. J. Exp. Med. 183: 2669–2674. Mudter, J., S. Wirtz, P. R. Galle, and M. F. Neurath. 2002-2003. A new model of chronic colitis in SCID mice induced by adoptive transfer of CD62L+ CD4+ T cells: insights into the regulatory role of interleukin-6 on apoptosis. Pathobiology 70: 170–176. Varga, G., J. Ehrchen, A. Brockhausen, T. Weinhage, N. Nippe, M. Belz, A. Tsianakas, M. Ross, D. Bettenworth, T. Spieker, et al. 2014. Immune suppression via glucocorticoid-stimulated monocytes: a novel mechanism to cope with inflammation. J. Immunol. 193: 1090–1099. Kurmaeva, E., D. Bhattacharya, W. Goodman, S. Omenetti, A. Merendino, S. Berney, T. Pizarro, and D. V. Ostanin. 2014. Immunosuppressive monocytes: possible homeostatic mechanism to restrain chronic intestinal inflammation. J. Leukoc. Biol. 96: 377–389.


35.

products may modulate Crohn’s disease. Am. J. Physiol. Gastrointest. Liver Physiol. 300: G823–G832. Vogl, T., S. Ludwig, M. Goebeler, A. Strey, I. S. Thorey, R. Reichelt, D. Foell, V. Gerke, M. P. Manitz, W. Nacken, et al. 2004. MRP8 and MRP14 control microtubule reorganization during transendothelial migration of phagocytes. Blood 104: 4260–4268. Ehrchen, J., L. Helming, G. Varga, B. Pasche, K. Loser, M. Gunzer, C. Sunderko¨tter, C. Sorg, J. Roth, and A. Lengeling. 2007. Vitamin D receptor signaling contributes to susceptibility to infection with Leishmania major. FASEB J. 21: 3208–3218. Varga, G., J. Ehrchen, A. Tsianakas, K. Tenbrock, A. Rattenholl, S. Seeliger, M. Mack, J. Roth, and C. Sunderkoetter. 2008. Glucocorticoids induce an activated, anti-inflammatory monocyte subset in mice that resembles myeloidderived suppressor cells. J. Leukoc. Biol. 84: 644–650. Varga, G., S. Balkow, M. K. Wild, A. Stadtbaeumer, M. Krummen, T. Rothoeft, T. Higuchi, S. Beissert, K. Wethmar, K. Scharffetter-Kochanek, et al. 2007. Active MAC-1 (CD11b/CD18) on DCs inhibits full T-cell activation. Blood 109: 661–669. Becker, C., M. C. Fantini, S. Wirtz, A. Nikolaev, R. Kiesslich, H. A. Lehr, P. R. Galle, and M. F. Neurath. 2005. In vivo imaging of colitis and colon cancer development in mice using high resolution chromoendoscopy. Gut 54: 950–954. Eisenbla¨tter, M., J. Ehrchen, G. Varga, C. Sunderko¨tter, W. Heindel, J. Roth, C. Bremer, and A. Wall. 2009. In vivo optical imaging of cellular inflammatory response in granuloma formation using fluorescence-labeled macrophages. J. Nucl. Med. 50: 1676–1682. L€ ugering, A., T. Kucharzik, D. Soler, D. Picarella, J. T. Hudson, III, and I. R. Williams. 2003. Lymphoid precursors in intestinal cryptopatches express CCR6 and undergo dysregulated development in the absence of CCR6. J. Immunol. 171: 2208–2215. Ahlmann, M., G. Varga, K. Sturm, R. Lippe, K. Benedyk, D. Viemann, T. Scholzen, J. Ehrchen, F. U. M€uller, M. Seidl, et al. 2009. The cyclic AMP response element modulator alpha suppresses CD86 expression and APC function. J. Immunol. 182: 4167–4174. Martinez, F. O., S. Gordon, M. Locati, and A. Mantovani. 2006. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J. Immunol. 177: 7303–7311. Mellins, E. D., C. Macaubas, and A. A. Grom. 2011. Pathogenesis of systemic juvenile idiopathic arthritis: some answers, more questions. Nat Rev Rheumatol 7: 416–426. Fleetwood, A. J., H. Dinh, A. D. Cook, P. J. Hertzog, and J. A. Hamilton. 2009. GM-CSF- and M-CSF-dependent macrophage phenotypes display differential dependence on type I interferon signaling. J. Leukoc. Biol. 86: 411–421. Lacey, D. C., A. Achuthan, A. J. Fleetwood, H. Dinh, J. Roiniotis, G. M. Scholz, M. W. Chang, S. K. Beckman, A. D. Cook, and J. A. Hamilton. 2012. Defining GM-CSF- and macrophage-CSF-dependent macrophage responses by in vitro models. J. Immunol. 188: 5752–5765. Adib-Conquy, M., and J. M. Cavaillon. 2002. Gamma interferon and granulocyte/monocyte colony-stimulating factor prevent endotoxin tolerance in human monocytes by promoting interleukin-1 receptor-associated kinase expression and its association to MyD88 and not by modulating TLR4 expression. J. Biol. Chem. 277: 27927–27934. Adema, G. J., F. Hartgers, R. Verstraten, E. de Vries, G. Marland, S. Menon, J. Foster, Y. Xu, P. Nooyen, T. McClanahan, et al. 1997. A dendritic-cell-derived C-C chemokine that preferentially attracts naive T cells. Nature 387: 713–717. Forssmann, U., M. B. Delgado, M. Uguccioni, P. Loetscher, G. Garotta, and M. Baggiolini. 1997. CKbeta8, a novel CC chemokine that predominantly acts on monocytes. FEBS Lett. 408: 211–216. Ostanin, D. V., J. Bao, I. Koboziev, L. Gray, S. A. Robinson-Jackson, M. Kosloski-Davidson, V. H. Price, and M. B. Grisham. 2009. T cell transfer model of chronic colitis: concepts, considerations, and tricks of the trade. Am. J. Physiol. Gastrointest. Liver Physiol. 296: G135–G146. Krausgruber, T., K. Blazek, T. Smallie, S. Alzabin, H. Lockstone, N. Sahgal, T. Hussell, M. Feldmann, and I. A. Udalova. 2011. IRF5 promotes inflammatory macrophage polarization and TH1-TH17 responses. Nat. Immunol. 12: 231–238. Da¨britz, J. 2014. Granulocyte macrophage colony-stimulating factor and the intestinal innate immune cell homeostasis in Crohn’s disease. Am. J. Physiol. Gastrointest. Liver Physiol. 306: G455–G465. Han, X., K. Uchida, I. Jurickova, D. Koch, T. Willson, C. Samson, E. Bonkowski, A. Trauernicht, M. O. Kim, G. Tomer, et al. 2009. Granulocytemacrophage colony-stimulating factor autoantibodies in murine ileitis and progressive ileal Crohn’s disease. Gastroenterology 136: 1261–1271, e1–3. Nylund, C. M., S. D’Mello, M. O. Kim, E. Bonkowski, J. Da¨britz, D. Foell, J. Meddings, B. C. Trapnell, and L. A. Denson. 2011. Granulocyte macrophage-

15

Immunosuppressive monocytes: possible homeostatic mechanism to restrain chronic intestinal inflammation.

Vitamin A Impairs the Reprogramming of Tregs into IL-17-Producing Cells during Intestinal Inflammation.

Immunology: Regulatory ILCs help control intestinal inflammation.

Circadian rhythm reprogramming during lung inflammation.

Netrin-1 guides inflammatory cell migration to control mucosal immune responses during intestinal inflammation.

CD73+ Murine Monocytes Modulate Intestinal Inflammation via Induction of Regulatory T Cells.

Nerve remodelling during intestinal inflammation.

lpr mice.

GMCSF-armed vaccinia virus induces an antitumor immune response.

Interleukin-33 from Monocytes Recruited to the Lung Contributes to House Dust Mite-Induced Airway Inflammation in a Mouse Model.

Pleiotropic functions of TNF-α in the regulation of the intestinal epithelial response to inflammation.

Prolactin mediates psychological stress-induced dysfunction of regulatory T cells to facilitate intestinal inflammation.

Abdominal obesity contributes to neurocognitive impairment in HIV-infected patients with increased inflammation and immune activation.

Novel Foxp3(-) IL-10(-) Regulatory T-cells Induced by B-Cells Alleviate Intestinal Inflammation in Vivo.

Apoptotic cells ameliorate chronic intestinal inflammation by enhancing regulatory B-cell function.

Reprogramming of the human intestinal epigenome by surgical tissue transposition.

Cancer seeding contributes to intestinal anastomotic dehiscence.

Regulatory immune cells in regulation of intestinal inflammatory response to microbiota.

Natural compound methyl protodioscin protects against intestinal inflammation through modulation of intestinal immune responses.

The functions of human monocytes in normal subjects and in disorders associated with immune deficiency.

Distinct Commensals Induce Interleukin-1β via NLRP3 Inflammasome in Inflammatory Monocytes to Promote Intestinal Inflammation in Response to Injury.

Inhibition of immune functions by antiviral drugs.

VEGF Production by Ly6C+high Monocytes Contributes to Ventilator-Induced Lung Injury.

Immune surveillance of the lung by migrating tissue monocytes.