Published online before print September 12, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Allergy and Clinical Immunology, King’s College Hospital NHS Foundation Trust, and
King’s College London, Breast Cancer Biology Group, Division of Cancer Studies, Guy’s Hospital Campus, London, United Kingdom
1Correspondence at current address: Department of Cancer Immunology and AIDS, Dana-Farber Cancer Institute, Jimmy Fund Building, Room 707, Harvard Medical School, 44 Binney Street, Boston, MA 02115, USA. E-mail: agnes_lo{at}dfci.harvard.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: macrophage • monocyte • apoptosis
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CSF-1 plays a central role in orchestrating the survival and differentiation of cells of the monocytic lineage. These activities are mediated by a single high-affinity receptor tyrosine kinase (CSF-1R), encoded by the c-fms proto-oncogene [8
]. The CSF-1R is widely expressed throughout the mononuclear phagocyte system, including monocytes, macrophages, osteoclasts, DCs, and Langerhans cells (LCs) [9
10
11
]. Upon binding of ligand, the CSF-1R undergoes tyrosine autophosphorylation, followed by activation of signaling intermediates, which include PI-3K, the MAPK/ERK pathway, and phospholipase C-
(PLC-) [12
13
14
15
]. CSF-1 is found at low but detectable levels in serum from normal donors. However, pronounced overproduction of this cytokine is a feature of pregnancy [16
17
18
] and many pathological states, including malignancy, trauma, and chronic inflammatory disease [19
20
21
22
23
]. In these settings, there is considerable evidence that CSF-1 may contribute to depressed APC function, as a result of skewed differentiation of monocytic precursors toward macrophage-like cells rather than DCs [21
, 24
, 25
].
The inhibitory effect of CSF-1 excess upon DC differentiation was demonstrated first using tumor cell-derived conditioned medium [24 ]. Using CD34+-derived [24 ] or monocyte-derived DC [25 ], it has been shown that CSF-1 and IL-6 promote the commitment of these cells toward a macrophage-like phenotype, characterized by high phagocytic activity but poor APC function. Overproduction of CSF-1 has been reported in association with several solid tumors, most notably epithelial ovarian, renal cell, and breast carcinomas [19 ]. Consequently, the suppressive effect of CSF-1 upon DCs may contribute widely to immune evasion by tumor cells [20 , 23 , 24 ].
Recently, there has been increased understanding of the pathways that regulate monocytic differentiation. At the transcriptional level, regulation is mediated by the opposing influences of PU.1 (promotes DC fate) and MafB (promotes macrophage fate) [26 ]. Caspase activity has also been implicated in this process. Conditional deletion of the caspase 8 gene in the myelomonocytic lineage results in a marked impairment of macrophage differentiation and a much more subtle reduction in the DC compartment [27 ]. Furthermore, differentiation of human monocytes to macrophages in vitro is accompanied by and dependent on the activation of several caspase family members [28 ]. Recent evidence indicates that caspase 8 lies at the apex of this pathway [29 , 30 ]. By contrast, caspase activity is regulated in precisely the opposite manner during DC differentiation. Immature DCs exhibit low levels of caspase activity and further reduction upon maturation. Furthermore, pharmacological inhibition of caspase activity can promote DC maturation [31 ].
In the present study, we have investigated the ability and mechanisms by which CSF-1 inhibits the differentiation of myeloid DCs, focusing on the role of proximal signaling intermediates and caspase activation. We present evidence that CSF-1 activates a PI-3K-dependent pathway, which leads to delayed caspase activation, thereby directing differentiation away from DCs and toward a unique macrophage-like phenotype.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and Fas:Fc fusion protein (BD PharMingen, Oxford, UK); CD40 ligand (CD40L) and IFN- (Peprotech, Rocky Hills, NJ, USA); CSF-1 (Cetus, Emeryville, CA, USA); neutralizing anti-human TNF- antibody (MAB210; R&D Systems, Abingdon, UK). Pharmacological inhibitors were all purchased from Calbiochem (Nottingham, UK), dissolved in DMSO, and used at the indicated concentrations: Wortmannin (1 µM unless otherwise stated), LY294002 (0.3–10 µM), PD98059 (50 nM), U73122 (5 nM), Z-Ile-Glu-Thr-Asp (IETD)-fluoromethylketone (FMK; 20 µg/ml), and Z-Leu-Glu-His-Asp (LEHD)-FMK (3–20 µg/ml). All other chemicals were from Sigma-Aldrich (Poole, UK).
Generation of human monocyte-derived DC and macrophages
Human PBMCs were obtained from anonymized blood packs purchased from the UK National Blood Transfusion Service. Following Ficoll separation, T cells were depleted using sheep red cell rosetting or anti-CD3-coated paramagnetic beads (Invitrogen, Paisley, UK). In some experiments, B cells were depleted using anti-CD19-coated paramagnetic beads (Invitrogen). To induce DC or macrophage differentiation, purified monocytes were cultured in RPMI 1640 with L-glutamine (Gibco-BRL, Paisley, UK), sodium pyruvate (110 mg/l, Sigma-Aldrich), gentamicin (0.1 mg/ml, Sigma-Aldrich), 2-ME (0.05 mM, Gibco-BRL), and 10% FBS (Gibco-BRL). To induce DC differentiation, GM-CSF (500 U/ml) + IL-4 (250 U/ml) were added on Day 0. In some cases, CSF-1 (5000 U/ml, unless otherwise stated) was also added on Day 0 (DC-M cultures). Where stated, TNF- (100 U/ml) was added on Day 5 to induce maturation. To induce macrophage differentiation, cells were cultured in CSF-1 (5000 U/ml, M-Mac) or GM-CSF (500 U/ml, GM-Mac) alone. All cultures were fed by replacing half of medium and cytokines at Days 3 and 5.
Flow cytometry analysis
Immunophenotyping of cultures was performed using the following antibodies: CD95-FITC, HLA-DR-PE, CD14-allophycocyanin, CD1a-FITC, CD80-PE, CD86-allophycocyanin, CD40-FITC, CD11c-PE, CD3-allophycocyanin, and isotype controls (BD Biosciences, Oxford, UK). For detection of intracellular Bcl-2, cells were fixed in 4% paraformaldehyde, permeabilized in 0.5% saponin + 1% BSA, and stained with FITC-conjugated anti-Bcl-2 (BD PharMingen) in permeabilization buffer. Fas ligand (FasL) expression was detected using FITC-conjugated antibody (Calbiochem). Prior to staining, cells were incubated with 10 µM matrix metalloproteinase inhibitor KB8301 (BD PharMingen) for 8 h. In all cases, 10,000 events were acquired using a FACSCaliber flow cytometer and analyzed by CellQuest software (BD Biosciences). The CSF-1R was detected using the 3-4A4 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA), followed by FITC-conjugated anti-rat IgG (Serotec, Kidlington, UK). In this case, flow cytometry was performed using a Coulter EPICS XL cytometer with Expo32 ADC software. Data are expressed as percentage positive events or mean fluorescence intensity (MFI) of staining.
MLR
DC and DC-M were harvested on Day 7, irradiated at 3000 cGy, and added to 1 x 105 allogeneic T cells at the indicated ratios. After 6 days, cultures were pulsed with 1 µCi tritiated thymidine (Amersham, Arlington Heights, IL, USA) for 18 h. Cells were harvested onto glass fiber filters using a Micromate 96 cell harvester (Canberra Packard, Pangbourne, Berkshire, UK) and counted using a matrix 96 direct β-counter (Canberra Packard).
Dextran endocytosis assay
Day 7 DC and DC-M were incubated with 1 mg/ml FITC-dextran (Sigma-Aldrich) [32
] for 1 h at 37°C. After washing, MFI in the fluorescence 1 (FL-1) channel was measured by flow cytometry. Data were corrected by subtraction of background uptake (uptake of FITC-dextran by the same cells placed on ice for 1 h).
May-Grünwald Giemsa staining of DC
Cytospins were fixed with methanol and stained with May-Grünwald solution for 2 min, followed by Giemsa for 4 min. After washing, cytospins were air-dried, dipped in xylene, and mounted.
Light microscopy
Light microscopy was performed using an Eclipse TE300 inverted microscope (Nikon, Tokyo, Japan) equipped with a 40x/0.75 objective lens and 10x/0.30 objective lens. Images were captured using a Nikon N70 camera and processed using Preview 2.1.0 software (Apple, Cupertino, CA, USA).
Measurement of DNA fragmentation in apoptosis
To detect DNA fragmentation resulting from apoptosis, an apolipoprotein-BrdU kit (BD PharMingen) was used as described by the manufacturer.
Cytokine secretion and ELISA
Days 5 and 7 DCs (3x105 cells) were stimulated with 1 µg/ml soluble CD40L in the presence or absence of 1000 U/ml IFN- for 24 h. For the detection of IL-12p70 secretion, Day 5 DCs (3x105 cells) were stimulated with 1 µg/ml LPS (Sigma-Aldrich) for 24 h. Harvested supernatants were analyzed by sandwich ELISA for TNF-, IL-6, IL-10, IL-12p40, and IL-12p70 using paired capture and biotinylated antibody sets (BD PharMingen).
Measurement of caspase activity by flow cytometry
Activity of caspases 3, 8, and 9 was detected using CaspaTag kits [33
] purchased from Intergen (Purchase, NY, USA), as recommended by the manufacturer. Caspase-related fluorescence was measured in the FL-1 channel. Propidium iodide+ events were excluded from the analysis.
Detection of mitochondrial membrane potential (
m) by flow cytometry
m was assayed using a "DePsipher" kit (TA700), purchased from R&D Systems, according to the manufacturer’s instructions. Fluorescence was measured in the FL-1 channel.
Statistical analysis
Statistical significance was determined using the two-tailed Student’s t-test. Values were considered statistically significant if P < 0.05.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
was added to promote maturation of cultures. After 7 days, cultures established without CSF-1 ("DC") consisted of nonadherent cells with smooth, uniformly stained cytoplasm and abundant dendritic processes (Fig. 1A
). Cells expressed CD1a, CD80, CD86, CD40, and high levels of HLA-DR, consistent with DC differentiation (Fig. 1B
and 1C)
. By contrast, when CSF-1 was also present ("DC-M"), cultures exhibited increased adherence, vacuolation, reduced cytoplasmic staining (Fig. 1A)
, and markedly altered cell surface immunophenotype (Fig. 1B
and 1C
, and Table 1
). In most cases, DC-M cultures expressed little CD14 by Day 7 (Table 1)
. However, CD14 expression increased thereafter and was generally detectable on DC-M by Day 9. The effects of CSF-1 upon DC differentiation were irreversible; removal of this cytokine from Day 7 cultures did not permit reacquisition of a DC-like phenotype (data not shown). Furthermore, following treatment with TNF-, up-regulation of the activation marker CD83 was greater in Day 7 DC (24%) than DC-M cultures (6%).
 View larger version (69K): [in a new window] |
Figure 1. Inhibitory effect of CSF-1 on differentiation of monocyte-derived DC. (A) Human monocytes were cultured with GM-CSF, IL-4, and TNF- (Day 5) in the absence (DC) or in the presence (DC-M) of CSF-1. On Day 7, stained cytospins were prepared (left panels; original magnification, x400; May-Grünwald Giemsa). Panels on the right show typical DC and DC-M cultures by inverted light microscopy (original magnification, x100). (B) Representative surface immunophenotype of DC and DC-M, assayed on Day 7. Black histogram, DC; gray histogram, DC-M; filled histogram, isotype control. Percentage-positive events for each marker are shown. (C) MFI expression of the designated marker on DC-M cultures has been expressed as a percentage of the corresponding value from DC cultures. Data shown are the mean ± SD of six independent experiments. For each marker, statistical significance of the difference between DC and DC-M is indicated at the P < 0.005 (*), P < 0.001 (**), and P < 0.0005 (***) levels. (D) Following activation with TNF- (Day 5), irradiated Day 7 DC and DC-M were established in a MLR with 1 x 105 allogeneic T cells for 6 days and then pulsed overnight with 3H-thymidine. Data presented (cpm) are the mean ± SD obtained from triplicate cultures. Similar findings were obtained in three independent experiments. (E) Day 7 DC, DC-M, and macrophages (generated using CSF-1 alone; M-Mac) were generated in the absence of exogenous TNF-, irradiated, and established in a MLR with 1 x 105 allogeneic T cells as described in D. Statistically different results between DC and DC-M are indicated in panels D and E where P < 0.05 (*), and P < 0.01 (**).
|
|
View this table: [in a new window] |
Table 1. Immunophenotype of DC-M, Macrophages, and DC
|
To examine the functional effects of CSF-1 upon DC differentiation, three experiments were performed. First, when tested in a one-way MLR, DC-M exhibited lower allostimulatory activity than DCs. These differences were apparent whether DC/DC-M had undergone maturation with TNF- (Fig. 1D)
or not (Fig. 1E)
. Allostimulatory activity of DC-M cultures was comparable with that of CSF-1-generated macrophages (Fig. 1E)
. Second, DC-M had a markedly altered ability to produce cytokines compared with DCs, with reduced secretion of IL-10, TNF-, IL-12 p40, and IL-12 p70 but increased production of IL-6, even in the unstimulated state (Table 2
). Third, DC-M exhibited increased endocytic activity compared with DCs (Fig. 2E
). Taken together, these data confirm that CSF-1 profoundly alters the differentiation of monocyte-derived DC, skewing cells toward a distinctive, macrophage-like phenotype.
|
View this table: [in a new window] |
Table 2. Cytokine Production by DC and DC-M Cultures
|
 View larger version (43K): [in a new window] |
Figure 2. CSF-1 alters differentiation of monocyte-derived DC—effect of signaling inhibitors. (A) A representative experiment in which DC and DC-M cultures were supplemented with wortmannin (PI-3K inhibitor), PD98059 (MEK inhibitor), or U73122 (PLC- inhibitor) on the day of culture initiation. MFI of CD1a expression was determined on Day 7 by flow cytometry. (B) DC and DC-M were cultured in the presence of wortmannin (or DMSO as control), harvested on Day 7, and analyzed for CD1a expression. To normalize MFI data, levels found in DMSO-treated DC cultures were set at 100% (mean±SD of data obtained from five independent donors). (C) DC and DC-M were cultured ± wortmannin (w) for 7 days, irradiated, and then established in a MLR at a 1:100 ratio with 1 x 105 allogeneic T cells for 6 days. Data presented are the mean ± SD obtained from triplicate cultures. DCs cultured in wortmannin alone were not used, owing to extensive apoptosis. (D) DC-M cultures were supplemented with LY294002 and harvested after 9 days. Expression of CD14 and CD1a was determined by flow cytometry. (E) DC and DC-M were cultured in the presence or absence of wortmannin for 7 days. Endocytic uptake of FITC-dextran was determined by flow cytometry. In all cases, statistical significance is indicated at the P < 0.05 (*), P < 0.01 (**), and P < 0.0005 (***) levels.
|
(U73122). Experiments were focused on CD1a expression, which provided the most consistent marker of the inhibitory effect of CSF-1 upon DC differentiation (Fig. 1C)
. Neither inhibition of MEK or PLC- markedly influenced the effect of CSF-1 on DC differentiation, as measured by CD1a expression (Fig. 2A)
. It is striking, however, that wortmannin reversed the sensitivity of DC cultures completely to CSF-1 in a dose-dependent manner (Fig. 2A
and 2B
, and Table 3
). Wortmannin also reduced the acquisition of CD14 by DC-M cultures and adherence (data not shown) and abrogated the stimulatory effect of CSF-1 on endocytic activity (Fig. 2E)
. This action of wortmannin did not result from reduced CSF-1R expression on DC or DC-M cultures (expressed as MFI or percentage-positive cells; Supplemental Fig. 1). Furthermore, when DC-M were cultured with wortmannin, allostimulatory activity was comparable or even greater than that of DCs (Fig. 2C)
. |
View this table: [in a new window] |
Table 3. Dose-Dependent Effect of Wortmannin on CD1a Expression
|
CSF-1 promotes activation of several caspases and mitochondrial depolarization in DC-M cultures
Caspase activity was examined in human DC and DC-M cultures. It was found that when DCs differentiate in the presence of CSF-1 excess, marked activation of caspases 3, 8, and 9 was detectable by Day 7 (Fig. 3A
). Caspase activation increased progressively over the course of the culture period (Fig. 3B)
but was not accompanied by apoptotic cell death, as measured by TUNEL analysis (Fig. 3C)
or electron microscopy (data not shown). In agreement with previous reports [28
], when DCs differentiate in the absence of exogenous CSF-1, only low levels of caspase activity were detected (Fig. 3A
and 3B)
.
 View larger version (29K): [in a new window] |
Figure 3. Exogenous CSF-1 activates caspases 3, 8, and 9 during differentiation of human monocyte-derived DC (A), which were cultured for 7 days in the presence or absence of CSF-1. The activity of caspases 3, 8, and 9 was determined by flow cytometry in DCs (filled histograms) and DC-M (open histograms), using the fluorochrome-labeled inhibitors of caspase-propidium iodide technique. (B) Time course of activation of caspases 3, 8, and 9 in DC and DC-M cultures. (C) Detection of DNA fragmentation in DC and DC-M cultures on Day 4 using the TUNEL assay. Control cells were provided with the kit. Similar results were obtained when cultures were tested after 3, 5, or 7 days.
|
m [28
]. Figure 4A
shows that DC-M cultures also exhibit a significant disruption of m when compared with DCs (increase in MFI of 174.0±24.9% when analyzed on Day 7; n=3 experiments). Mitochondrial membrane depolarization was detectable as early as Day 4 in DC-M cultures (data not shown). It is notable that this effect was observed, despite the ability of CSF-1 to up-regulate Bcl-2 (Fig. 4B)
.
 View larger version (13K): [in a new window] |
Figure 4. Mitochondrial depolarization in DC-M cultures. (A) A representative experiment in which m was measured in Day 7 DC and DC-M cultures by flow cytometry. Valinomycin was used as positive control (right panel). (B) Bcl-2 was detected by flow cytometry in DC (dotted line) and DC-M (solid line) cells following permeabilization. Isotype control is shown as the filled histogram. Similar findings were obtained in two experiments.
|
 View larger version (56K): [in a new window] |
Figure 5. Effect of caspase inhibition on DC differentiation. A representative experiment in which DC and DC-M cultures were established in the presence of wortmannin, Z-IETD-FMK (caspase 8 inhibitor), Z-LEHD-FMK (caspase 9 inhibitor), or DMSO as solvent control. Cultures were also supplemented with Fas-Fc fusion protein at the indicated concentrations. After 7 days, cells were harvested and analyzed for expression of CD1a and CD14 by flow cytometry.
|
 View larger version (31K): [in a new window] |
Figure 6. Caspase 9 inhibition achieves a dose-dependent effect in DC-M cultures. (A) DC and DC-M were established in Z-LEHD-FMK (caspase 9 inhibitor) at the indicated doses or DMSO as solvent control. Wortmannin is shown as positive control. After 7 days, cells were harvested and analyzed for expression of CD1a and CD14 by flow cytometry. (B) DC and DC-M cultures were established in the presence or absence of Z-LEHD-FMK (20 µg/ml). Cells were harvested on Day 7 and analyzed for CD1a expression. To normalize MFI, levels found in DMSO-treated DC cultures were set at 100%. Data shown are the mean ± SD from eight independent experiments. ***, P < 0.0005.
|
 View larger version (28K): [in a new window] |
Figure 7. Wortmannin inhibits caspase activation in DC-M cultures. DC and DC-M cultures were generated in the presence or absence of wortmannin and harvested on Day 7. Caspases 8 and 9 activity was measured by flow cytometry. Similar findings were observed in two (caspase 8) and three (caspase 9) independent experiments.
|
|
View this table: [in a new window] |
Table 4. Expression of Fas and FasL by DC and DC-M Cultures
|
in CSF-1-triggered caspase activation was added to cultures on Day 5 to induce maturation. As TNF- promotes caspase activation, we wondered if this might be required for the actions of CSF-1 upon DC differentiation. To investigate this, we established DC and DC-M cultures without addition of exogenous TNF- and analyzed CD1a expression on Day 7 (Supplemental Table 1). These experiments confirmed that CSF-1 inhibits DC differentiation in the absence of TNF-induced maturation and that wortmannin could counteract this.
Next, we considered whether endogenous (DC-derived) TNF- could play a role in caspase activation. Even in the absence of stimulation, DC and DC-M cultures produce readily detectable levels of TNF- (Table 2)
. To block this, cultures were supplemented with a neutralizing anti-TNF- antibody. When analyzed on Day 7, DC-M expressed lower levels of CD1a than DCs, as is typically seen. Addition of anti-TNF- did not reverse the effect of CSF-1 and indeed, resulted in lowered CD1a expression in DC and DC-M cultures (Supplemental Table 2).
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The finding that CSF-1 can impair the differentiation of monocyte-derived DC contrasts with two earlier studies [34 , 35 ]. Differences in relative levels of CSF-1 and IL-4/IL-13 may underlie these discrepant findings. By contrast, Menetrier-Caux and colleagues [24 , 25 ] have shown that CSF-1 + IL-6 can inhibit the differentiation of DCs. In man, the importance of this phenomenon is illustrated following trauma, where high levels of CSF-1 "prime" monocytes so that subsequent ex vivo DC differentiation is impaired [21 ]. The data presented here extend these findings, as we show that CSF-1 can influence DC differentiation profoundly, even in the presence of the antagonistic cytokine IL-4.
In considering the in vivo counterpart of the DC-M cell, one possible candidate is the LC. Recently, Ginhoux and colleagues [11 ] have shown that monocytes are direct LC precursors. Furthermore, using CSF-1R knockout mice, they have provided evidence that CSF-1R plays a pivotal role in LC development. In keeping with this proposal, DC-M exhibit high endocytic capacity and are allostimulatory, albeit to a lesser extent than DCs [36 ]. Furthermore, serum CSF-1 levels are increased substantially in LC histiocytosis [37 ]. Although we did not observe other typical features of LC, such as Birbeck granules (data not shown), expression of such features may require an appropriate environment such as epidermis or other mucosa. Further investigation of the possibility that LCs represent an intermediate cell type between macrophages and DCs would appear warranted.
To explore the molecular mechanisms by which CSF-1 can alter DC differentiation, studies were performed using signaling inhibitors. These experiments implicated PI-3K as a key effector of CSF-1 on monocyte differentiation, complementing the known ability of this signaling molecule to inhibit DC maturation [38 , 39 ]. Indeed, in the presence of PI-3K inhibitors, CSF-1 frequently up-regulated CD1a expression. In some experiments, we found that MEK inhibition could also weakly antagonize some effects of CSF-1 upon DC differentiation (data not shown). This finding is consistent with data indicating that PI-3K lies proximal to and regulates the activity of the MAPK pathway in CSF-1-stimulated monocytes [15 ].
Xie and colleagues [40 ] have also examined the effect of PI-3K inhibition on monocyte-derived DCs. They similarly observed that PI-3K inhibition resulted in appreciable death of DCs as a result of apoptosis. By contrast, they found variable down-regulation of DC markers in wortmannin-treated cultures, including CD1a. This discrepancy may reflect the fact that in both of our studies, the analysis of DCs was hampered by toxicity of inhibitors, leading to reduced cell yield at Day 7. By contrast, we observed that DC-M cultures were much more resistant to toxicity induced by PI-3K and caspase inhibitors. This allowed us to obtain robust and reproducible data when such inhibitors were added to DC-M cultures.
DC-M cultures exhibited a progressive increase in the activity of caspases 3, 8, and 9, in association with depolarization of m. These data are reminiscent of the role played by caspases in macrophage differentiation [28
29
30
] and raise the possibility that caspases may also mediate the effect of CSF-1 on DC differentiation. In agreement with this, inhibition of caspase 8 or 9 reversed the suppressive effect of CSF-1 on DCs. Furthermore, CSF-1-dependent caspase activity was abrogated by PI-3K inhibition.
At first sight, the finding that CSF-1 promotes caspase activation in a PI-3K-dependent manner appears counterintuitive. It has been shown that CSF-1 inhibits monocyte apoptosis through activation of PI-3K and Akt, associated with suppressed activity of caspases 3 and 9 [13 , 41 ]. In both studies, attenuation of caspase activity was demonstrable at 18 h and was antagonized by PI-3K inhibitors. These findings are consistent with the well-established roles of these molecules in the inhibition (PI-3K) or promotion (caspases) of apoptosis. However, they are not inconsistent with our data showing that CSF-1 promotes caspase activation at a later time-point. In agreement with this, caspase activation became detectable after 3 days during CSF-1-stimulated macrophage differentiation and was not accompanied by apoptosis [29 ].
A further dichotomy presented by our data is the observation that caspase inhibitors affect survival of DCs (low caspase activity) to a greater extent than DC-M (high caspase activity). This may be explained by the ability of caspase inhibitors to cause nonapoptotic cell death [28 ]. A key finding of our study is that in DC-M cultures, increased caspase activity has been uncoupled from apoptotic cell death.
The mechanism by which caspases are activated in response to CSF-1 remains incompletely understood. Although our inhibitor data implicate PI-3K activation in this process, the delayed kinetics of caspase activation suggest that further steps are also involved. One possibility we had considered was the role of reactive oxygen species (ROS), which is produced by CSF-1-stimulated monocytes in a PI-3K-dependent manner [15 ]. However, we could not consistently block the inhibitory effects of CSF-1 upon DC differentiation with antioxidants (data not shown).
We have also explored the role of death receptors in CSF-1-triggered caspase activation. However, inhibition of FasL or TNF- had no effect on the ability of CSF-1 to skew DC differentiation. It should be acknowledged that this experimental approach might not inhibit intracrine engagement of FasL or TNF- with its receptor. Furthermore, Fas aggregation may occur in the absence of ligand, for example, in response to ROS [42
].
Our findings indicate that caspase activation plays a key role in the differentiation of monocytes to macrophages (favored) or DCs (inhibited). Experiments performed with inhibitors implicate PI-3K in the regulation of this balance. Taken with previous reports, these data raise the possibility that caspase activity regulates the continuum of cell phenotypes among mature DCs (minimal caspase activity), immature DCs (low caspase activity), and highly phagocytic macrophage-like cells (highest caspase activity). Together with the considerable heterogeneity and plasticity seen, such a model is consistent with the proposal that the mononuclear phagocyte system consists of a spectrum of cellular subtypes, lacking clear, distinguishing boundaries [9 ].
Received March 9, 2007; revised August 16, 2007; accepted August 16, 2007.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
J. Exp. Med. 179,1109-11182 by the M-CSF receptor is necessary for differentiation signaling EMBO J. 16,5880-5893[CrossRef][Medline]
B in monocytes undergoing macrophagic differentiation Blood 109,1442-1450 induces the generation of Langerin/(CD207)+ immature Langerhans-type dendritic cells from both CD14-CD1a and CD14+CD1a– precursors derived from CD34+ cord blood cells Eur. J. Immunol. 33,2053-2063[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  M. Hiasa, M. Abe, A. Nakano, A. Oda, H. Amou, S. Kido, K. Takeuchi, K. Kagawa, K. Yata, T. Hashimoto, et al. GM-CSF and IL-4 induce dendritic cell differentiation and disrupt osteoclastogenesis through M-CSF receptor shedding by up-regulation of TNF-{alpha} converting enzyme (TACE) Blood, November 12, 2009; 114(20): 4517 - 4526. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Hume Macrophages as APC and the Dendritic Cell Myth J. Immunol., November 1, 2008; 181(9): 5829 - 5835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. G. Netea, E. C. Lewis, T. Azam, L. A. B. Joosten, J. Jaekal, S.-Y. Bae, C. A. Dinarello, and S.-H. Kim Interleukin-32 induces the differentiation of monocytes into macrophage-like cells PNAS, March 4, 2008; 105(9): 3515 - 3520. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
