Published online before print April 30, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
INSERM UMRS 872, Université Pierre et Marie Curie-Paris 6, Université René Descartes-Paris 5, Centre de Recherches Biomédicales des Cordeliers, Paris, France
1 Correspondence: INSERM UMRS 872, Université Pierre et Marie Curie-Paris 6, Université René Descartes-Paris 5, Centre de Recherches Biomédicales des Cordeliers, 15, Rue de l’Ecole de Médecine, 75006 Paris, France. E-mail: wing-hong.kwan{at}u255.bhdc.jussieu.fr
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: growth factor • cytokine • differentiation • activation • tolerance
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mice lacking functional M-CSF (op/op) have reduced numbers of Langerhans cells (LC), the epidermal DC subset [4
], and mice deficient in M-CSF receptor (M-CSFR) are devoid completely of LC [5
]. Moreover, M-CSFR–/– mice appear to lack dermal DC. Thus, M-CSF seems to be involved in the development and/or maintenance of LC and dermal DC, possibly at the stage of DC precursor. In human models of DC development, it has been shown that cord blood monocytes cultured in M-CSF and IL-4 produce more IL-10 in response to LPS than when cultured in GM-CSF and IL-4 [6
]. This suggests that M-CSF instructs myeloid cells to an anti-inflammatory program. However, the M-CSF/IL-4-cultured monocytes are fully differentiated DC and cannot address the role of DC precursors in the immune system. Alternately, monocytes cultured in M-CSF retain their capacity to differentiate to DC in the presence of GM-CSF and IL-4 [7
], and we have shown recently that cord blood progenitor-derived CD14+ cells kept in M-CSF retain the ability to differentiate to dermal DC and LC [8
, 9
]. These M-CSF-conditioned CD14+ cells show phenotypic homologies with dermal CD14+ cells, including expression of DC-SIGN, an important pathogen attachment factor, and are permissive to HIV and Dengue virus infection [8
]. As these cells can differentiate to dermal DC or LC-type DC, we called them preDC. Although DC release TNF-
and undergo maturation, when infected by Dengue virus, the preDC release IL-10 but no TNF- and do not undergo maturation [8
]. This suggested that preDC, recognizing infectious organisms, could down-modulate the immune response through IL-10 production.
Therefore, we investigated IL-10 production by monocytes in response to LPS when precultured in M-CSF or GM-CSF and studied how LPS affected preDC, which are M-CSF-conditioned myeloid cells, in their ability to differentiate to DC, to mature, and to activate T cells. Our results suggest that M-CSF instructs myeloid cells to an anti-inflammatory function in response to LPS, and GM-CSF conditions an inflammatory program. The preDC released early IL-10 and could modulate the immune response by down-regulating T cell proliferation and DC maturation. However, GM-CSF could revert the anti-inflammatory program installed by M-CSF. Thus, we confirm that M-CSF enhances the production of IL-10 by differentiating monocytes [6 ] and extend the previous work by detailed IL-10 secretion kinetics and by characterizing the immune properties of M-CSF-conditioned DC precursors.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(R&D Systems), and 200 U/ml GM-CSF (Schering-Plough, Kenilworth, NJ, USA). The cells were then washed and cultured in 25 ng/ml M-CSF (R&D Systems) for 6 days with refreshing M-CSF at Day 3, giving rise to a cell population containing mainly CD14+ cells, which were purified using anti-CD14-coated magnetic beads (Miltenyi) and are referred to as preDC. To obtain DC, the preDC were differentiated in the presence of 500 U/ml (45 ng/ml) GM-CSF (Schering-Plough), 5 ng/ml IL-4 (Schering-Plough), and 2 ng/ml TGF-ß (R&D Systems) for 2–3 days. Peripheral blood buffy coats from healthy donors were obtained from the Établissement Français du Sang, Hôpital Hôtel-Dieu (France). Monocytes were enriched from mononuclear cells by centrifugation through a 52% Percoll gradient (Sigma-Aldrich, Saint Quentin Fallavier, France). Floating cells were collected and negatively selected using a monocyte isolation kit (Dynal, Compiègne, France). Monocytes were used directly or after a 5-day preculture in 50 ng/ml M-CSF or 100 ng/ml GM-CSF in complete medium with cytokine refreshment at Day 3. For data presented in Table 1 , monocytes were cultured for 18 h in 25, 50, or 100 ng/ml M-CSF or in 100 ng/ml GM-CSF.
|
View this table: [in a new window] |
Table 1. Presence of IL-10 (pg/ml) 4 h after Stimulation by Indicated Doses of LPS
|
(R&D Systems), 10 ng/ml TNF- (R&D Systems), 5 ng/ml IL-1ß (R&D Systems), 10 ng/ml IL-6 (R&D Systems), or 1 x 10–6 M PGE2 (Sigma-Aldrich). For CD40 ligand (CD40L) simulation, cells were cocultured with mitomycin C-treated CD40L+ mouse fibroblasts at a ratio of five preDC:one fibroblast. IL-10R-blocking experiments were performed using the anti-IL-10R1 antibody (MAB274; R&D Systems). For this, two million cells were incubated with 15 µg/ml (2x) of blocking antibody for 30 min at 37°C, and then medium alone or supplemented with LPS was added to obtain a final concentration of 7.5 µg/ml neutralizing antibody. Recombinant human (rh)IL-10 was purchased from R&D Systems and used at 10 ng/ml. The mycobacterium Mycobacterium bovis bacillus Calmette-Guerin (BCG) was kindly provided by Nathalie Winter (Institut Pasteur, Paris, France). preDC or DC were incubated for 4 h in complete medium lacking antibiotics at a multiplicity of infection (MOI) of one cell to five bacteria. The cells were then washed extensively in PBS and cultured for 20 h in complete medium containing antibiotics.
Cytokine measurements
All cell culture (106 cells/ml) supernatants were collected and stored at –20°C. IL-10 measurements were done by the sandwich ELISA technique using antihuman IL-10-specific antibody (Diaclone, Besançon, France). The IL-6 ELISA set was from Immunotools (Freiburg, Germany). The data are reported for 1 x 106 cells/ml.
Phenotypic analysis
Expression of specific markers was determined by flow cytometry on a FACSCaliburTM (Becton Dickinson, San Jose, CA, USA) and the following antibodies from BD PharMingen (San Diego, CA, USA): CD14-FITC [macrophage P9 (M
P9)], CD1a-PE (HI149), CD83-FITC (HB15e), and CD86-FITC (FUN-1); from Beckman-Coulter (Roissy, France): HLA-DR-FITC (Immu-357) and CD80-FITC (MAB104).
MLR
Naïve T cells were purified from cord blood CD34+ progenitor-depleted PBMC by positive cell sorting using anti-CD4-coated magnetic beads (Dynal). After selection, beads were detached using the provided detach-bead enzyme mix. The T cell population purity was routinely >95%. The MLR medium was complete medium supplemented with 5% human AB serum (AbCys, Paris, France) and 1 mM sodium pyruvate (Invitrogen). For the MLR, stimulator cells, washed free of cytokines or LPS, were distributed in triplicate in graded doses in 96-well, round-bottom plates (Falcon, Becton Dickinson). To facilitate comparison of preDC and DC with respect to T cell stimulation, same donor CD34+ cord blood progenitors were used to generate preDC and DC at different time intervals. Cord blood CD4+ T cells (2x104) were added to each well in a final volume of 200 µl. T cell proliferation was evaluated after 5 days of culture following an overnight incubation with 1 µCi [3H]methylthymidine. Results are expressed as mean cpm ± SD of triplicate cultures.
T cell activation
Peripheral adult T cells from buffy coats were isolated with anti-CD4 (Dynal) magnetic beads. Anti-CD3 antibody (10 µg/ml) in PBS was added to a flat-bottom, 96-well plate (Costar, Corning, NY, USA), incubated for 3 h at room temperature, and then seeded with 100 µl containing 105 T cells in complete medium, supplemented with 2 µg/ml anti-CD28 (Serotech, Oxford, UK) and 10% human AB serum. Finally, 100 µl supernatants from resting or LPS-stimulated preDC or medium, with or without IL-10 (20 ng/ml) or IL-2 (400 IU/ml), were added. After 5 days, T cells proliferation was assessed as described above. To avoid contact of LPS with T cells, which may have an effect on regulatory T cells [10
], the preDC were pulsed for 3 h with LPS, washed extensively, and then kept in culture for 24 h. Still, in control experiments, we found that the addition of LPS to CD3/CD28-activated T cells had no effect on T cell proliferation (data not shown).
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
/IL-6 [11
]. In view of this dichotomous cytokine response, with important implications with regard to monocyte differentiation into suppressor-type macrophages [12
] or DC, we asked if monocytes can be instructed to trigger an anti-inflammatory or an inflammatory program exclusively. Monocytes were cultured for 5 days in M-CSF or in GM-CSF and then stimulated by LPS. Production of IL-10 and IL-6 was measured with time (Fig. 1A
). M-CSF-conditioned monocytes produced IL-10 rapidly but little IL-6. In contrast, when precultured in GM-CSF, the monocytes released IL-6 rapidly but little IL-10. TNF- secretion was detected in both conditions after 3 h of stimulation without a clear difference between M-CSF or GM-CSF pretreatment (data not shown). Culturing monocytes in M-CSF or GM-CSF alone did not lead to IL-10 or IL-6 release. To further analyze the potential of M-CSF to instruct monocytes to secrete early IL-10, we cultured monocytes only for 18 h in the presence of different concentrations of M-CSF or GM-CSF and then measured IL-10 4 h after LPS stimulation. As shown in Table 1
, exposing monocytes to M-CSF for 18 h was sufficient to induce IL-10 secretion, whereas no IL-10 was released by control monocytes. However, concomitant stimulation of LPS + M-CSF did not lead to IL-10 secretion (data not shown), suggesting that M-CSF triggers a cell differentiation program and does not simply provide a synergistic signal to LPS. The data show that M-CSF prepares monocytes to an IL-10 anti-inflammatory program, and GM-CSF instructs for an IL-6 inflammatory response.
 View larger version (16K): [in a new window] |
Figure 1. M-CSF instructs monocytes to rapid IL-10 production. (A) Adult monocytes were cultured for 5 days with M-CSF or GM-CSF and then stimulated with 50 ng/ml LPS, and release of IL-10 and IL-6 was measured. The data are expressed as OD of the ELISA assay and are representative of three donors. (B) preDC or monocytes were stimulated with 2.5 ng/ml or 50 ng/ml LPS, and IL-10 production was measured after 24 h (mean±SD of IL-10 for 106 cells from five different donors). (C) preDC and monocytes are stimulated with 2.5 ng/ml LPS, and production of IL-10 was measured at different time-points for 24 h. The data are the mean ± SD of duplicates and representative of three donors.
|
Proinflammatory signals trigger IL-10 release
We next tested stimuli other than LPS for their ability to trigger IL-10 production. preDC were cultured for 24 h in GM-CSF, TNF-, IL-6, IL-1ß, IFN-, as well as PGE2, LPS, or CD40L-transfected fibroblasts or a combination thereof. The combination of TNF-, IL-6, IL-1ß, and PGE2 was reported to trigger DC maturation efficiently [13
]. As shown in Figure 2A
, GM-CSF did not induce IL-10, but all of the other cytokines triggered IL-10 production. We then assessed these stimuli for their ability to convey to preDC the capacity to stimulate naive T cells in a MLR (Fig. 2B)
. In almost a reciprocal manner to IL-10 release, none of the stimuli except GM-CSF instructed preDC to prime naive T cell proliferation. Therefore, preDC responded to a variety of proinflammatory cytokines by IL-10 production and remained unable to prime T cells. Reflecting the lower IL-10 release, it is likely that TNF- is a weak stimulus for preDC. However, GM-CSF could instruct the preDC to activate T cells.
 View larger version (5K): [in a new window] |
Figure 2. Proinflammatory and microbial signals but not GM-CSF trigger IL-10 release. (A) preDC were cultured for 24 h in the presence of the indicated cytokines, LPS, or in the presence of CD40L+ fibroblasts. IL-10, in the cell supernatant, was measured by ELISA and expressed as the mean ± SD for 106 cells of at least three different donors. (B) preDC were stimulated by the indicated cytokines or LPS for 24 h and then washed. They were added to allogenic cord blood CD4+ T cells at a ratio of 1:1. After 5 days, T cell proliferation was measured by incorporation of [3H]methylthymidine for 24 h. The data, representative of two experiments, are expressed as mean cpm ± SD of triplicate wells.
|
|
View this table: [in a new window] |
Table 2. Conversion of preDC into CD1a+CD14– DC
|
 View larger version (17K): [in a new window] |
Figure 3. Differentiation to DC abolishes IL-10 production and allows maturation in response to LPS and BCG. (A, left panel) preDC and DC derived from the same preDC in GM-CSF/IL-4/TGF-ß were stimulated for 24 h with 1 µg/ml LPS or left untreated, and secretion of IL-10 was measured. The data represent the mean ± SD for 106 cells from five different donors. Right panel, preDC or the derived DC were incubated with live BCG for 3 h at MOI = 5 and then were washed and cultured in complete medium containing antibiotics for 24 h. The data are representative of two donors. (B) preDC or the derived DC were stimulated with 1 µg/ml LPS or infected by live BCG at MOI = 5. After washing, the cells were then added in graded doses to allogenic cord blood CD4+ T cells. After 5 days, T cell proliferation was measured by incorporation of [3H]methylthymidine. The data are representative of at least two donors. NT, not treated.
|
 View larger version (23K): [in a new window] |
Figure 4. Blocking IL-10 enables LPS-stimulated preDC to convert into professional APC. (A, left panel) preDC were incubated with an anti-IL-10R antibody (-IL-10R) or its isotype control before exposing the cells to 1 µg/ml LPS. After 24 h, expression of cell surface markers was assessed by FACS together with control, unstimulated (no LPS) preDC and as positive control, DC and LPS-matured DC derived from preDC. Live cells were electronically gated; white histograms represent isotype controls, and specific labeling is shown in gray. (Right panel) Expression of CD1a versus CD14 was measured by FACS for unstimulated preDC—preDC treated with LPS in the presence of control antibody of anti-IL-10R antibody. The percentage of CD1a+CD14+ cells and CD1a–CD14+ cells is shown. Also, the mean fluorescence intensity (MFI) of CD14 expression of CD1a–CD14+ cells is indicated. (B) Cells treated as in A were washed and added in graded doses to allogeneic cord blood CD4+ T cells. After 5 days, T cell proliferation was measured by incorporation of [3H]methylthymidine and expressed as mean cpm ± SD of triplicate wells. The data are representative of three experiments.
|
 View larger version (15K): [in a new window] |
Figure 5. DC derived from LPS-stimulated preDC are immature. (A) DC were derived from preDC under the following conditions: 3-day culture in GM-CSF/IL-4/TGF-ß in the absence or in the immediate presence of 100 ng/ml LPS or when LPS addition was delayed by 48 h. Expression of CD1a versus CD14 as well as that of other cell surface markers was analyzed by FACS without electronically gating for live cells. (B) The level of naïve T cell priming was compared between DC derived from preDC in the presence of GM-CSF/IL-4/TGF-ß (DC) and LPS added during DC differentiation (LPS-DC) or added after 48 h (DC>LPS). Also, LPS-DC were re-exposed to 100 ng/ml LPS for 24 h (LPS-DC>LPS). Shown is the T cell proliferation at an APC:T cell ratio of 1:1. Mean cpm ± SD from triplicate wells; each condition is representative of at least two experiments.
|
 View larger version (14K): [in a new window] |
Figure 6. LPS-stimulated preDC reduce DC maturation and T cell proliferation. (A) DC were cocultured with preDC (ratio of 1:1) and 50 ng/ml LPS. As control, DC were left untreated or treated with LPS in the absence of preDC. After 24 h, the cells were stained for CD1a and CD83. The percentage of CD1a+ cells expressing CD83 is shown. The results are the mean ± SD from three experiments. Statistic significance was assessed with the Student’s paired t-test. (B) Adult CD4+ T cell was activated by anti-CD3/CD28 antibodies for 5 days in the presence of medium only, supernatant (SN) from preDC, supernatant from LPS-treated preDC, or medium containing IL-2 or IL-10. T cell proliferation was measured by incorporation of [3H]methylthymidine. The data are expressed as mean cpm ± SD from triplicate wells, representative of at least three experiments. Statistic significance was assessed with the Student’s paired t-test.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We observed that conversion into DC by GM-CSF/IL-4/TGF-ß abolished IL-10 production almost completely, and the cells acquired potent T cell priming activity in response to LPS or BCG mycobacteria. These data show that according to the type of differentiation factor (M-CSF vs. GM-CSF), myeloid cells acquire an anti-inflammatory or an inflammatory activity, respectively. Yet, the M-CSF-cultured anti-inflammatory preDC remained fully capable of converting into DC, even during LPS-triggered IL-10 release. Thus, preDC are like monocytes in terms of capacity to differentiate into DC but are like anti-inflammatory macrophages in terms of cytokine expression. However, it is probable that a prolonged exposure to M-CSF will result in progressive incapacity to differentiate into DC [7 ]. Although simultaneous exposure of LPS and GM-CSF/IL-4/TGF-ß still allowed DC formation, the DC displayed an immature phenotype and were unable to prime T cells. In addition, their re-exposure to LPS could not induce T cell priming. This phenomenon of endotoxin tolerance has also been observed with LPS-stimulated monocytes and may be related to reduced association of TLR4 with MyD88 and decreased IL-1R-associated kinase 1 activity [16 ]. Thus, autocrine IL-10 prevented the formation of mature DC and hence, T cell priming. In addition, we provided evidence that the supernatant of LPS-preDC antagonized maturation of bystander DC. This may be relevant in conditions of a mixed myeloid population such as in the skin, where CD14+ cells colocalize with CD1a+ dermal and Langerhans DC. Furthermore, the supernatant of LPS-stimulated preDC prevented polyclonal CD4+ T cell proliferation, so that even under conditions of T cell activation, preDC can down-modulate the formation of a T cell-mediated immune response.
Collectively, the data suggest that myeloid DC precursors, when exposed to M-CSF, respond to LPS and probably to other danger signals by installing a tolerogenic environment, which inhibits formation of professional APC in an autocrine and a paracrine manner and prevents bystander T cell activation. We have not detected CCR7 expression on LPS-stimulated preDC, indicating that tolerance would be restricted to tissue and not carried to draining lymph nodes, unless mediated by tolerized, immature or semimature, migratory DC or via soluble IL-10. Our work is in line with a report showing that M-CSF/IL-4 but not GM-CSF/IL-4 induces cord blood monocytes to differentiate into cells that secrete IL-10 but not IL-12 upon LPS stimulation [6 ]. However, while the CD14+ cells used here can convert readily into CD1a+ DC in GM-CSF/IL-4, the cells used by Li et al. [6 ] lack CD14 and CD1a and can no longer differentiate into CD1a+ DC when subsequently exposed to GM-CSF. Also, adult monocytes cultured in M-CSF for 3 days produced more IL-10 than when cultured in M-CSF/IL-4 (data not shown). Xu et al. [17 ] showed that M-CSF but not GM-CSF induces the differentiation of monocytes into IL-10-producing macrophages with intrinsic anti-inflammatory properties and are specialized in clearing apoptotic cells. This indicates that M-CSF may have a function in tissue homeostasis and peripheral tolerance to self-antigens.
preDC share cell surface markers with dermal CD11b+CD1a– cells [8 , 9 ], and human dermal CD11b+CD1a– mono/macrophages produce high levels of IL-10 rapidly after UV irradiation [18 ]. It is therefore plausible that the dermal CD11b+ mono/macrophages are DC precursors and that IL-10 influences their conversion into immunostimulatory, mature dermal DC. In support of this notion are the observations that intradermal injection of live bacteria in mice results in little DC mobilization to draining lymph nodes [19 , 20 ], and that subcutaneous LPS triggers a higher immune response in mice when the IL-10R is blocked [21 ]. Recently, Bogunovic et al. [22 ] presented evidence that dermal DC, like Langerhans DC, arise from a tissue-residing, cycling precursor under steady-state conditions. This supports our use of the cycling CD34+ progenitor to generate dermal-like CD14+ cells and suggests that dermal mono/macrophages may be exposed to M-CSF throughout their development.
Yet, the fact that an adaptive immune response can be raised in response to immunization with adjuvants containing microbial products shows that a local immune tolerance can be broken. We have found that a means to convert LPS-stimulated preDC directly into professional APC is by blocking IL-10 signaling. Also, CD40L can override the IL-10 block and likewise induce formation of CD14+ APC with potent T cell priming activity [23
]. The Th2 cytokine IL-4 does not revert tolerance, although M-CSF/IL-4 monocytes have DC features [6
], but the Th1 cytokine IFN- can convert LPS-stimulated monocytes into CD14–CD83+ mature DC through down-regulation of IL-10 production [24
]. Alternatively, it is possible that the inflammation increases GM-CSF levels so that attracted monocytes will be conditioned to release proinflammatory cytokines and would differentiate into DC. In this context, it is relevant to point out that CD16+ monocytes (Gr-1+ in mouse) are preferentially recruited to inflamed sites [25
]. These cells not only appear to release more proinflammatory cytokines than their CD16– counterpart [26
] but also differentiate more rapidly to DC [27
].
In conclusion, we propose that under steady-state conditions, M-CSF may condition tissue-residing monocytes or DC precursors developing in tissue to an anti-inflammatory program. This would curtail an efficient, T cell-mediated immune response but avert pathogen dissemination and immunopathology.
Received April 12, 2006; revised March 13, 2007; accepted March 14, 2007.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
- and IL-4-independent development of Langerhans cell-like dendritic cells from M-CSF-conditioned precursors J. Invest. Dermatol. 126,114-120[CrossRef][Medline] and endogenously produced IL-10 Eur. J. Immunol. 31,3773-3781[CrossRef][Medline]RIII(+)) subset of human monocytes preferentially becomes migratory dendritic cells in a model tissue setting J. Exp. Med. 196,517-527This article has been cited by other articles:
 |
|
 |
|
  B. B. Ganesh, D. M. Cheatem, J. R. Sheng, C. Vasu, and B. S. Prabhakar GM-CSF-induced CD11c+CD8a--dendritic cells facilitate Foxp3+ and IL-10+ regulatory T cell expansion resulting in suppression of autoimmune thyroiditis Int. Immunol., March 1, 2009; 21(3): 269 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. P. Sarangi, B. Kim, E. Kurt-Jones, and B. T. Rouse Innate Recognition Network Driving Herpes Simplex Virus-Induced Corneal Immunopathology: Role of the Toll Pathway in Early Inflammatory Events in Stromal Keratitis J. Virol., October 15, 2007; 81(20): 11128 - 11138. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
