Src-family kinases play an essential role in differentiation signaling ...

Leukemia (2008) 22, 161–169 & 2008 Nature Publishing Group All rights reserved 0887-6924/08 $30.00 www.nature.com/leu

ORIGINAL ARTICLE Src-family kinases play an essential role in differentiation signaling downstream of macrophage colony-stimulating factor receptors mediating persistent phosphorylation of phospholipase C-c2 and MAP kinases ERK1 and ERK2 C Bourgin-Hierle1,2, S Gobert-Gosse1, J The´rier, M-F Grasset and G Mouchiroud Centre de Geńe´tique Molećulaire et Cellulaire, Universite´ de Lyon, UMR 5534, Universite´ Claude Bernard Lyon 1, CNRS, Villeurbanne, France

Macrophage colony-stimulating factor (M-CSF) has been found to be involved in multiple developmental processes, especially production of cells belonging to the mononuclear phagocyte system. The decision of myeloid progenitor cells to commit to differentiation depends on activation levels of the mitogenactivated protein kinases (MAPK), ERK1 and ERK2. Using the murine myeloid progenitor cell line FD-Fms, we show here that persistent activity of Src-family kinases (SFK) is necessary for FD-Fms cell differentiation to macrophages in response to MCSF. Chemical inhibition of SFK blocked FD-Fms cell differentiation while it caused strong inhibition of the late phosphorylation of phospholipase C (PLC)-c2 and MAPK. The PLC inhibitor U73122, previously shown to block M-CSF-induced differentiation, strongly decreased long-term MAPK phosphorylation. Interestingly, inhibiting SFK with SU6656 or the MAPK kinases MEK with U0126 significantly impaired development of mononuclear phagocytes in cultures of mouse bone marrow cells stimulated with M-CSF. Collectively, results support a model in which SFK are required for sustained PLC activity and MAPK activation above threshold required for commitment of myeloid progenitors to macrophage differentiation. Leukemia (2008) 22, 161–169; doi:10.1038/sj.leu.2404986; published online 1 November 2007 Keywords: Src-family kinases; MAP kinases; myelopoiesis; differentiation

Introduction A key issue in cell signaling is how ubiquitous signal transduction pathways can lead to a specific cellular response. It is now well admitted that signal strength and/or duration is crucial to a variety of biological responses. Sustained activation of the mitogen-activated protein kinases (MAPK), ERK1 and ERK2, has been found to be involved in various cellular systems, such as fibroblast proliferation,1 neuronal,2 lymphoid,3,4 megakaryocyte5,6 and myeloid7–9 differentiation. Multiple mechanisms have been found to be involved in the fine tuning of MAPK activation, including differential usage of Ras effectors, formation of specific molecular complexes around scaffolding proteins, and/or negative regulator effects.10–12 Thus, there appear to be no general rules determining the level of MAPK activation upon cytokine stimulation. Correspondence: Dr G Mouchiroud, Centre de Geńe´tique Molećulaire et Cellulaire, CNRS Universite´ Lyon 1, UMR 5534, 43 Bd du 11 novembre 1918, Ba´timent Gregor Mendel, Villeurbanne 69622, France. E-mail: [email protected] 1 These authors contributed equally to this work 2 Current address: The Burnham Institute for Medical Research, Pathology Department, 10901 N. Torrey Pines Rd., La Jolla, CA 92037, USA Received 3 January 2007; revised 3 September 2007; accepted 14 September 2007; published online 1 November 2007

Macrophage colony-stimulating factor (M-CSF) and its receptor are strongly involved all along the monocyte-macrophage lineage, providing cells with survival, proliferation, differentiation and activation signals.13,14 The physiological effects of M-CSF could be reproduced in hematopoietic cell lines through enforced macrophage colony-stimulating factor receptor (M-CSFR) expression. Of particular interest were derivatives of the myeloid FDC-P1 cell line originally established from interleukin-3 (IL-3)-stimulated long-term cultures of mouse bone marrow cells.15 FDC-P1 cells expressing M-CSFR (FD-Fms cells) differentiate to macrophage-like cells in response to M-CSF,16,17 whereas those expressing the G-CSF receptor express neutrophil-specific genes in response to G-CSF.18 These data support a model of instructive lineage decision in FDC-P1 cells. In this respect, FDC-P1 cells behave like native myeloid progenitors, whose commitment to granulocyte or macrophage lineages is controlled by specific cytokines.19 In the FD-Fms model, activation of MAPK8 or phospholipase C-g2 (PLC-g2)20 is required for macrophage differentiation. Studies with M1 and 32D murine myeloid cell lines demonstrated Src-family kinases (SFK) binding to activated M-CSFR21 and partial impairment of M-CSF-induced differentiation by the src kinase inhibitor PP2.22 In addition to activation of specific signaling pathways, another important aspect of M-CSF signaling specificity is the strength and duration of signaling. It has long been known that continuous M-CSF stimulation is required for progression throughout the G1 phase of cell cycle and induction of early response genes.23 We have previously found that macrophage differentiation of FD-Fms cells depends on sustained activation of the Ras–MAPK pathway.8 In the present study, we used the FD-Fms cell model and bone marrow cell cultures to clarify the role of SFK in M-CSF differentiation signaling. Next, we asked whether signal transduction molecules involved in FD-Fms cell differentiation, SFK and PLC-g2, participate in sustained MAPK activation, and whether they act in separate pathways or are components of a single pathway dedicated to differentiation commitment. For that purpose, we investigated the activation kinetics of these molecules, their role in persistent MAPK activation using specific signaling inhibitors, and possible functional relationships between them. The results support the model of a single signal transduction pathway connecting M-CSF receptors to sustained activation of MAPK, and thereby differentiation.

Materials and methods

Cytokines and signaling inhibitors X63-IL-3 cell conditioned medium was used as a source of IL-3.24 The source of M-CSF was conditioned medium from Sf9

Src kinases regulate M-CSF differentiation signal C Bourgin-Hierle et al

162

insect cells expressing murine M-CSF25 for FD-Fms studies or purified recombinant murine M-CSF (Peprotech, Rocky Hill, NJ, USA) for bone marrow cell cultures. PP2, SU6656, U0126, U73122 and U73343 were purchased from Calbiochem (San Diego, CA, USA).

Cell cultures FDC-P1 cells constitutively expressing mouse M-CSFR (FD-Fms cells)17 were maintained in Iscove’s modified defined medium (IMDM) (Invitrogen, Cergy-Pontoise, France) supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich, Saint-Quentin Fallavier, France) and 5% X63-IL-3 cell conditioned medium. For differentiation, FD-Fms cells were washed twice in IMDM, seeded at 2 104 cells per ml in IMDM containing 10% FBS and 2500 U ml1 of M-CSF, and subcultured when they reached a density of 5 105 cells per ml. One unit of M-CSF was defined as the amount able to stimulate formation of one colony from 50 000 bone marrow cells plated in standard agar culture. For primary cell cultures, bone marrow cells from femurs of 6- to 8-week-old C57BL/6 mice were collected in IMDM-10% FBS and subjected to centrifugation onto Lympholyte-M (Cedarlane, Burlington, Ontario, Canada) according to manufacturer’s instructions. Resulting bone marrow mononuclear cells (BMMC) were resuspended at 5 105 per ml in IMDM supplemented with 10% FBS, 12.5 ng ml1 M-CSF and 50 mM bmercaptoethanol. After 6-day cultivation, both adherent and non-adherent cells were harvested and subjected to cell counting, cytospin preparations, and flow cytometry analyses as specified below. Signaling inhibitors were added at the onset of cultures unless otherwise specified.

Morphological and flow cytometric analyses For morphological studies, cells were cytocentrifuged onto glass slides, air-dried, and stained with May–Gru¨nwald-Giemsa reagent (MGG, Sigma-Aldrich). Flow cytometry determination of Mac-2 expression on FD-Fms cells was performed by indirect immunofluorescence as previously described.26 Mac2 antibody and FITC-conjugated F(ab0 )2 fragment goat anti-rat immunoglobulin G were from Cedarlane and Caltag Laboratories (Burlingame, CA, USA), respectively. Cultivated BMMC were labeled by direct immunofluorescence according to instructions from antibody manufacturers. Antibodies were as follows: FITCconjugated anti-Ly-6C (clone ER-MP20, Acris Antibodies, Hiddenhausen, Germany), phycoerythrin (PE)-Cy7-conjugated anti-CD31 (eBiosciences, San Diego, CA, USA), PE-Cy5conjugated anti-CD11b, and Alexa488-conjugated anti-F4/80 (Caltag). Fc receptors were blocked with anti-CD16/32 monoclonal antibody (Caltag). All samples were analyzed after gating of viable cells, as determined by propidium iodide exclusion, using FACSCalibur (FD-Fms cells) or FACSCanto (cultured BMMC) apparatus (Becton Dickinson, Le Pont de Claix, France).

Signaling studies FD-Fms cells were washed twice in phosphate-buffered saline to remove IL-3 and incubated for 3 h at 37 1C in IMDM containing 0.1% bovine serum albumin (Sigma) for phosphotyrosine immunoblotting or 0.5% fetal calf serum (Sigma) for all other signaling studies. Starved cells were then stimulated with 2500 U ml1 of M-CSF at 37 1C for times as indicated in the text. Signaling inhibitors were added to cells 30 min before M-CSF. Protocols for immunoprecipitation and western blotting have been previously described.8 Monoclonal antibody to phosphotyrosine (clone 4G10) was from Upstate (Charlottesville, VA, USA). Monoclonal antibody against Leukemia

phosphorylated ERK1/2 and polyclonal antibodies against M-CSFR, ERK1, and PLC-g2 were from Santa Cruz Biotechno logy (Santa Cruz, CA, USA). Horseradish peroxidase-conjugated secondary antibodies were from Bio-Rad Laboratories (Hercules, CA, USA). SFK kinase activity was assessed using a commercially available, radioactive, assay kit based on phosphorylation of a specific Src-substrate peptide derived from p34cdc2 (Upstate). Briefly, Src-family proteins were precipitated from cell lysates (300 mg of total protein) using anti-c-Src polyclonal antibody (SRC2, Santa Cruz Biotechnology). Immunoprecipitates were washed twice with lysis buffer, twice in 50 mM Hepes, pH 7.4, 10 mM MnCl2, and incubated with substrate peptide and [g32P] ATP for 20 min at 30 1C. SFK kinase activity was determined by subtracting counts per min (cpm) in samples without peptide from cpm in samples with peptide.

Data analyses and statistics In western blot experiments, ERK1/2 phosphorylation levels were determined by quantitation of phospho-ERK and ERK signals using Molecular Analyst software and normalization of pERK signals to corresponding ERK signals for each experimental point. Flow cytometry data were analyzed with CellQuest or Diva softwares. Unless otherwise specified, results were expressed as the mean±standard error of mean (s.e.m.) for at least three independent experiments. Statistical significance was determined by Student’s t-test (paired-data analysis). P-values o0.05 were considered as statistically significant.

Results

Effects of the SFK inhibitor PP2 on SFK and M-CSFR kinase activities We first verified that SFK activity could be inhibited without affecting M-CSFR activity. In these preliminary experiments, SFK inhibition was achieved using PP2, a potent inhibitor of SFK activity.27,28 For that, FD-Fms cells were stimulated with M-CSF in time-course studies in the presence or not of PP2 (10 mM) then lysed for SFK immunoprecipitation. Kinase activity was assessed in each immunoprecipitate using a radioactive, SFK-specific, assay. In two independent experiments, M-CSF stimulated biphasic SFK activity, with peaks at 5 min and 2 h of M-CSF stimulation (Figure 1a, open bars). As expected, PP2 completely blocked SFK activity, whatever the duration of M-CSF stimulation (Figure 1a, hatched bars). Same dose of PP2 had no inhibitory effect on M-CSFR phosphorylation as probed by phosphotyrosine immunodetection in M-CSFR immunoprecipitates (Figure 1b). On the contrary, PP2 treatment apparently increased receptor phosphorylation and subsequent kinase activity, especially after 15 min of M-CSF stimulation. Consistent with this observation, it was previously shown that mutation of M-CSFR on the SFK binding site resulted in delayed receptor internalization and degradation.21 With this restriction in mind, we concluded that SFK inhibitors provide a valuable approach to study SFK-mediated signaling events as they are able to inhibit the kinase activity of SFK, not that of M-CSFR, in response to M-CSF.

SFK inhibitors prevent macrophage differentiation of FD-Fms cells When maintained in IL-3, FD-Fms cells show a blast-like morphology after staining with MGG (Figure 2a, IL-3). Accordingly, cells displayed homogenous cell size (forward scatter or


163

Figure 1 PP2 inhibits SFK activity, not M-CSFR phosphorylation. (a) Kinase activity was determined in SFK immunoprecipitates from lysates of FD-Fms cells stimulated with 2500 U ml1 of M-CSF with or without 10 mM PP2, for times as indicated. Data are representative of two independent experiments. (b) Growth factor-deprived FD-Fms cells were treated or not for 30 min with 10 mM PP2 then 2500 U ml1 M-CSF were added for times as indicated. Cell lysates were subjected to M-CSFR immunoprecipitation, followed by electrophoresis on a 7.5% SDS–PAGE, and then probed with anti-phosphotyrosine (aPY) or anti-M-CSFR (aM-CSFR) antibody.

FSC) and granularity (side scatter or SSC) parameters in flow cytometry analyses (Figure 2b, IL-3). After 3 days in the presence of M-CSF, FD-Fms cells displayed the typical changes associated with macrophage differentiation, exhibiting heterogeneous size, enlarged cytoplasm containing vacuoles and granules, and increased cytoplasm:nucleus ratio (Figure 2a, M-CSF). Consistent with previous report,17 cells showed meaningful increase in both FSC and SSC parameters (Figure 2b, 48.4% in M-CSF vs o0.1% in IL-3), correlating with the morphological changes associated to differentiation. Finally, macrophage differentiation of FD-Fms cells was confirmed by induction by M-CSF of Mona protein expression, a molecular marker of monocytes and their precursors,29 and detection of the Mac-2 antigen on the surface of FD-Fms cells cultivated for 3 days in M-CSF,17 as shown in Figure 2c and Figure 3, respectively. FD-Fms cells cultivated for 3 days in the presence of M-CSF and PP2 (10 mM) looked like IL3 cultivated cells after MGG staining (Figure 2a, M-CSF þ PP2), suggesting lack of differentiation. Moreover, PP2 treatment prevented increase in FSC and SSC parameters (Figure 2b, 4.2% in M-CSF þ PP2 vs 48.4% in M-CSF), induction of Mona protein expression (Figure 2c), and surface expression of Mac-2 antigen (Figure 3). In order to confirm the effects of SFK inhibition on FD-Fms cell response to M-CSF, we used SU6656, another selective and potent inhibitors of SFK activity, which was described to have a higher specificity than PP2 for SFK.30,31 We found that SU6656 (1 mM) prevented the M-CSF-induced increase in FSC/SSC parameters (Figure 2b, 5.2% in MCSF þ SU6656 vs 48.4% in M-CSF) and induction of Mona protein expression (Figure 2c). Therefore, blocking SFK activity with PP2 or SU6656 totally abrogated M-CSF differentiation signal in FD-Fms cells, which represents a more dramatic

Figure 2 PP2 and SU6656 inhibit macrophage differentiation of FDFms cells in response to M-CSF. (a) Effects of PP2 on FD-Fms cell morphology. Cells were cultivated for 3 days with interleukin-3 (IL-3) or with 2500 U ml1 M-CSF in the presence or not of 10 mM PP2. (b) Flow cytometry analyses. Cells were cultivated for 3 days with IL-3 or with 2500 U ml1 M-CSF in the presence or not of SFK inhibitors (10 mM PP2 or 1 mM SU6656). Undifferentiated cells were gated as indicated (low FSC/SSC) and the percentage of differentiated cells (that is, outside the gate) is shown in the upper right part of each dot-plot. (c) PP2 prevents induction of Mona expression in response to M-CSF. Cells were treated as above. For each time point, cell lysates were probed by immunoblotting with anti-Mona or anti-ERK1/2 antibody. ERK1/2 were used here as internal loading control.

response than that previously reported in M1-Fms cells.22 These unequivocal data also support the idea that SFK inhibitors would specifically target the signaling pathways responsible for M-CSFinduced differentiation.

Macrophage differentiation requires late and sustained SFK activity We previously showed that commitment of FD-Fms cells to macrophage differentiation involves late and persistent ERK phosphorylation.8,26 Thus, addition of the MEK inhibitor U0126 could be delayed by 24 h after the beginning of M-CSF stimulation while being still able to block M-CSF differentiation signal.8 This prompted us to perform delayed PP2 addition experiments on M-CSF stimulation of FD-Fms cells. As described above, adding PP2 at the onset of M-CSF stimulation completely prevented Mac-2 expression and the increase in FSC/SSC parameters (Figure 3, t ¼ 0). PP2 could still block FDFms cell differentiation when added to cultures 8 h after M-CSF (Figure 3, t ¼ 8 h). Delaying PP2 addition by 24 h allowed differentiation of a minority of cells, which was restored when PP2 was introduced into cultures at 48 h of M-CSF stimulation Leukemia


164

Figure 3 Effects of delayed PP2 addition on the macrophage differentiation of FD-Fms cells. Cells were cultivated for 3 days in the presence of interleukin-3 (IL-3) or 2500 U ml1 M-CSF. PP2 (10 mM) was added at various times after onset of M-CSF-stimulated cultures, as indicated. In histograms, dotted lines represent control cells labeled with secondary antibody alone. Bold lines show cells labeled using anti-Mac-2 antibody. The percentage of Mac-2 þ cells was calculated by subtracting overlayed histograms using the CellQuest software. The percentage of differentiated cells in FSC/SSC diagrams was calculated as in Figure 2. Data are representative of three independent experiments.

(Figure 3, t ¼ 24 h and t ¼ 48 h). This indicates that commitment of FD-Fms cells to macrophage differentiation requires SFK activity after 8 h of M-CSF stimulation, which corresponds to the persistent phase of M-CSFR signaling.8

SFK activity regulate persistent phosphorylation of PLCg2 Phosphorylation of PLC-g2 has been associated with M-CSF differentiation signaling, and impaired PLC-g2 recruitment to activated M-CSFR, as well as treatment of cells with the PLC inhibitor U73122, correlated with deficient maturation of FDFms cells.20 PLC-g2 is rapidly and transiently phosphorylated and redistributed within the cytoplasm in M-CSF-stimulated myeloid cells, which is strongly dependent on SFK activity.8,31 However, no information was available on late or persistent PLC-g2 activation in response to M-CSF. Here, we looked at long-term phosphorylation of PLC-g2 in response to M-CSF, then asked whether late PLC-g2 phosphorylation would be altered in FD-Fms cells impaired for SFK activation. Starved FD-Fms cells were stimulated with 2500 U ml1 of M-CSF for various lengths of time, and PLC-g2 immunoprecipitates were analyzed by western blotting using anti-phosphotyrosine or anti-PLC-g2 antibodies (Figure 4, control). The phosphotyrosine blots clearly confirmed transient PLC-g2 phosphorylation within 10 min of M-CSF stimulation and also revealed later induction of PLC-g2 phosphorylation, from 2 h of M-CSF stimulation. When similar experiments were performed in the presence of PP2, the first peak of PLC-g2 phosphorylation was slightly decreased, whereas late PLC-g2 phosphorylation was strongly inhibited (Figure 4, PP2). Therefore, sustained PLC-g2 phosphorylation in response to M-CSF depends on SFK activity.

SFK and PLC are modulators of late MAPK phosphorylation M-CSF induces biphasic MAPK phosphorylation in FD-Fms cells: the first phase is rapid, culminates at about 2 min and terminates within 30 min of M-CSF stimulation, whereas the second one begins after 1 h of M-CSF stimulation and persists for at least 24 h.8,32 The second phase of MAPK activation is essential to M-CSF differentiation signaling, and previous experiments also suggested that maintaining MAPK phosphorylation above a certain threshold is essential to commit FDFms cells to macrophage differentiation.8 Based on these Leukemia

Figure 4 Time-course studies of PLC-g2 phosphorylation in the presence or absence of PP2. Growth factor deprived FD-Fms cells were stimulated with M-CSF (2500 U ml1) for indicated times. PP2 (10 mM) or DMSO (0.1%, control) was added to cells 30 min before M-CSF. PLC-g2 immunoprecipitates were probed by western blotting using either anti-phosphotyrosine (aPY) or anti-PLC-g2 (aPLC-g2). The experiment shown here is representative of three independent experiments.

observations, we asked whether inhibition of SFK and PLC would inhibit persistent MAPK phosphorylation in parallel to macrophage differentiation in M-CSF-stimulated FD-Fms cells. For that purpose, starved FD-Fms cells were pretreated with SFK or PLC inhibitors for 30 min then M-CSF was added for various lengths of time. MAPK phosphorylation was determined by western blotting with phospho-specific anti-ERK antibodies at time points representative of the first (2 min) and second (8 h) phases. As shown in Figure 5a, PP2 (10 mM) and SU6656 (1 mM) were very effective at inhibiting MAPK phosphorylation at 8 h of M-CSF stimulation (upper panel), resulting in statistically significant decreases in phospho-ERK signals as compared to dimethyl sulfoxide control (on average, 60% reduction for PP2 and 74% for SU6656; lower panel). SFK inhibitors also affected early MAPK phosphorylation, but effects were at the limit of statistical significance (22% reduction for PP2 and 39% for SU6656) (Figure 5a, lower panel). In contrast, ERK phosphorylation was strongly and significantly reduced at both time points by U73122, as compared to cells treated with its inactive homolog U73343 (69% reduction at 2 min and 64% at 8 h) (Figure 5b). Thus, in agreement with their role in macrophage differentiation of FD-Fms cells, activities of SFK and PLC are required for maximal activation of the MAPK pathway in


165

Figure 5 Quantitative control of MAPK phosphorylation in FD-Fms cells stimulated by M-CSF in the presence or not of (a) SFK inhibitors or (b) PLC inhibitor. Growth factor deprived FD-Fms cells were pretreated for 30 min with 10 mM PP2, 1 mM SU6656, or the corresponding amount of DMSO (0.1%, control) (a), or with 0.5 mM of U73122 or of its inactive analog, U73343 (b), before addition of M-CSF for times as indicated. Cell lysates were subjected to immunoblotting using either anti-ERK (a-ERK) or anti-phosphoERK (a-pERK) antibodies. Experiments shown in upper panels are representative of three independent experiments. Diagram in the lower panel shows quantitation of data from these three experiments (*Po0.05; **Po0.01, comparing inhibitor treated cells with corresponding control cells). For that, pERK signals were normalized to ERK signals. Note in (a) that strong phosphorylation of ERK2 (lower band) resulted in poor detection with polyclonal a-ERK antibody. Therefore, only ERK1 signals were used here to normalize pERK signals.

response to M-CSF. However, SFK appeared to be mainly implicated in persistent MAPK phosphorylation, in contrast to PLC that caused a significant reduction of MAPK phosphorylation at early and late times of M-CSF stimulation.

SFK and MAPK regulate monocytic development of mouse bone marrow cells in vitro Next, we asked whether SFK and MAPK might act as regulators of M-CSF signaling in the more physiological context of in vitro development of the mononuclear phagocyte system. For that purpose, signaling inhibitors were tested in liquid cultures of mouse BMMC established in the presence of M-CSF, which were expected to homogeneously develop into cells of the monocyte/macrophage lineage within 5–7 days.33 In our conditions, populations obtained at day 6 of culture were almost entirely composed of CD11b þ cells (96.1±0.5%) and were predominantly expressing F4/80 and Ly-6C used here, respectively, as macrophage and monocyte-specific markers34 (Supplementary Figure 1). Interestingly, small cells were F4/80/low and Ly-6Chigh, whereas larger cells were F4/80high and Ly-6C/low, indicating they represented mid-stage macrophage precursors and mature macrophages, respectively.35 Preliminary experiments were conducted to determine optimal conditions for testing signaling inhibitors. For optimal effects as described below, SU6656 and U0126 were used at 1 and 10 mM, respectively, but with a re-addition of U0126 at day 3 of cultures (data not shown). Note that U73122 proved to be very toxic, hampering further investigations on the role of PLC in this culture system. After 6-day cultivation, cultures established in the presence of SU6656 or U0126 showed significantly decreased cell

production (Figure 6a). Differentiation in these cultures was first assessed using MGG staining of cytospin preparations. Consistent with immunophenotype (Supplementary Figure 1), cell morphology after 6 days of culture in the presence of M-CSF was relatively homogeneous and 485% of cells possessed a monocyte/macrophage morphology (Figure 6b, upper panel; Figure 6c). In contrast, cultures treated with U0126 or SU6656 showed decreased proportion of macrophages and increase in granulocytic cells and monocytes (Figures 6b and c). Effects of U0126 and SU6656 on macrophage differentiation were quantitated by flow cytometric analyses of Ly-6C and F4/80. Figure 7a clearly shows the significantly diminished expression of both markers in the presence of U0126 or SU6656, as compared to control cultures established in the presence of M-CSF alone. Next, we reasoned that impairment of macrophage differentiation by signaling inhibitors would result in the emergence of monocyte/macrophage precursors. Accordingly, cells were analyzed for expression of Ly-6C and CD31, a marker of immature hematopoietic cells. BMMC cultivated for 6 days in the presence of M-CSF yielded few CD31high Ly-6C cells that represent earliest cells of the macrophage differentiation pathway.35 The proportion of CD31high Ly-6C cells was strongly and significantly increased by both inhibitors (Figure 7b). Interestingly, these cells were only found among small cells, which is consistent with their immature state, whereas large, mature cells, showed increased CD31 expression, suggesting maturation delay (Supplementary Figure 2). In conclusion, introducing SFK or MAPK inhibitors into M-CSF-stimulated cultures of mouse BMMC had two major effects that severely impaired macrophage production: total cell production was strongly reduced, and the proportion of immature or maturing macrophagic cells was increased. Therefore, our studies strongly Leukemia


166

Figure 6 SU6656 and U0126 affect cell number and morphology in cultures of mouse BMMC stimulated for 6 days with M-CSF. (a) Viable cell counts, as determined by trypan blue exclusion (***Po0.001, comparing inhibitor treated cells with corresponding control cells in six independent experiments). (b) Cytospin preparations were stained with May–Gru¨nwald-Giemsa reagent. (c) Major cell types were quantitated by counting at least 250 cells in more than five independent fields.

suggest that SFK and MAPK are implicated both in proliferation and differentiation signals of M-CSF.

Discussion In the FD-Fms cell differentiation model, long-lasting stimulation of MAPK8 and SFK (this study) is required by M-CSF to induce commitment to macrophage differentiation. Delayed addition of MEK or SFK inhibitors helped to define a critical time window for commitment between 8 and 24 h of M-CSF stimulation. Moreover, chemical inhibition of MEK, SFK and PLC resulted in differentiation blockade in FD-Fms cells, while inducing a 50–70% decrease in long-term MAPK phosphorylation (8, 17, this study). This indicates that FD-Fms cell commitment to macrophage differentiation is triggered when MAPK phosphorylation reaches a certain threshold. Latter conclusion is consistent with the observation that constitutive Ras activation provokes macrophage differentiation of native human myeloid progenitor cells, whereas establishment of moderate Ras activity by combining mutant Ras expression and farnesyltransferase inhibitor results in proliferation without differentiation.7 Therefore, tight control of the kinetics and strength of MAPK activation is required for the control of the balance between proliferation and differentiation in myeloid cells. The importance of MAPK activity in the myelomonocytic lineage came mainly from studies with immortalized cell lines or primary cells expressing mutated Ras proteins.7–9 This role of MAPK has been documented in physiologically more relevant experimental conditions. For instance, mice deficient for the Leukemia

T-cell protein tyrosine phosphatase shows hyperphosphorylated M-CSFR, subsequent increased recruitment of Gab2/Shp2 complexes and enhanced activation of MAPK in response to M-CSF, which in turn is likely the cause of a strong bias of myelopoiesis toward macrophage lineage in these mice.36 A requirement for MAPK activation was found in the granulocytic differentiation of bone marrow-derived human CD34 þ cells as induced by G-CSF stimulation.9 Thus, in normal myeloid progenitor cells, MAPK activity is necessary to differentiationinducing properties of G-CSF and M-CSF, but increased MAPK activation apparently favors macrophage differentiation.7,9,35 Here, we observed that the MEK inhibitor U0126 not only reduced macrophage differentiation after cultivating mouse bone marrow cells for 6 days in the presence of M-CSF, but also favored granulocytic cells. This suggests that reduced MCSFR signaling may favor commitment of myeloid progenitor cells to the granulocyte lineage at the expense of the macrophage lineage. In support to this hypothesis, we recently observed that differentiation response to M-CSF of conditionally immortalized hematopoietic progenitors could be shifted from macrophage to granulocyte lineage by addition of U0126.37 Therefore, current data support the conclusion that myeloid progenitor cell fate depends exclusively on the activation levels of the MAPK pathway. A paradigm for M-CSF stimulation of the MAPK pathway is based on the binding of Grb2/Sos complexes to the phosphorylated Y697 of the activated M-CSF receptor.38 However, it has been shown that M-CSFR lacking the kinase insert, and thereby Y697, is still able to activate MAPK in 32D-Fms cells through SFK-dependent mechanism.39 Studies with murine myeloid cells


167

Figure 7 Effects of SU6656 and U0126 on macrophage differentiation in cultures of mouse bone marrow cells stimulated for 6 days with M-CSF. (a) Flow cytometry analysis of F4/80 and Ly-6C expression. The percentage of positive cells was determined from F4/80 or Ly-6C vs FSC dot-plots for each corresponding sample in order to minimize autofluorescence signal of macrophages and the fluorescence of signaling inhibitors. Gating was set to get o1.0% positive cells in unlabeled controls. (b) Flow cytometry analysis of CD31/Ly-6C expression. Dot blots are representative of three independent experiments. The percentage of CD31 þ Ly-6C cells was calculated using cell numbers in the Q4 quadrant of corresponding dot-plots. *Po0.05; **Po0.01, ***Po0.001, comparing inhibitor treated cells with corresponding control cells in three independent experiments.

have linked SFK to MAPK activation through Shc phosphorylation and association with Grb2/Sos complexes, and subsequent activation of the Ras–MAPK pathway downstream of M-CSFR.39,40 However, we have previously demonstrated that specific disruption of Grb2/Sos complexes by a cell-permeant peptide did not modify MAPK phosphorylation levels, especially at late times of M-CSF stimulation.8 Altogether, these studies revealed the existence of alternative routes for stimulating the MAPK pathway downstream of M-CSFR. In this context, the present study suggests an important role for a SFK–PLC axis in putative Grb2/Sos-independent MAPK activation. First, PLC inhibition with U73122 caused a 50% decrease of MAPK activity in M-CSF-stimulated FD-Fms cells, similar to that obtained with SFK inhibitors (Figure 5). Second, SFK inhibition strongly reduced late tyrosine phosphorylation of PLC-g2 (Figure 4), which is assumed to correlate with decreased catalytic activity.41 PLCs hydrolyze phosphatidylinositols to generate diacylglycerol and inositol 1,4,5-trisphosphate, both of which have been reported to be involved in MAPK activation through Grb2/Sos-independent mechanisms.42,43 Interestingly, SFK and PLC have been recently implicated in the M-CSFinduced proliferation of mouse bone marrow-derived macrophages.44 Thus, one might speculate that M-CSF stimulates a SFK–PLC–ERK axis and thereby controls the extent of macrophage production by regulating differentiation and proliferation along the macrophage pathway. The question of how MAPK activation may impinge on differentiation programs in myeloid progenitor cells remains largely unanswered. In the model of fibroblast proliferation, it was shown that sustained MAPK activation makes the expression of immediate early genes (IEG) durable through stabilization of IEG products.45 Two IEG products play important roles in macrophage development and are regulated

through persistent MAPK activation: ets-246,47 and egr-1.45,48 Myeloid transcription factors are also phosphorylated by MAPK, such as AML-149 and members of the C-EBP family.50,51 Thus, it is possible that the graded influence of MAPK on myeloid progenitor cell fate may be relayed through differential regulation of major myeloid transcription factors. Abnormal activation of the MAPK pathway is found in many cases of acute myeloid leukemia, especially in those exhibiting activating mutations in Ras or Flt-3.52 Constitutive activation of Flt-3 in AML results in the inhibition of C/EBPa function by MAPKdependent phosphorylation, which may in turn cause differentiation block in leukemic cells.53 Furthermore, MAPK have been recently shown to cooperate with vitamin D derivatives or retinoic acid to promote differentiation of AML blasts.54,55 Thus, MAPK activity levels prove to have a strong impact on AML blast behavior, which may lead to development of new therapeutic protocols. In this context, it is important to determine how MAPK activity is fine-tuned in myeloid progenitor cells and what may be the consequence of altering this regulation.

Acknowledgements This work was supported by grants from the Ligue Nationale Contre le Cancer (Equipe labelliseé 2000 and 2004), Association pour la Recherche contre le Cancer (grant no 4000) and the Centre National de la Recherche Scientifique. CB-H and JT were supported by fellowships from French Ministe´re de la Recherche.

References 1 Kahan C, Seuwen K, Meloche S, Pouyssegur J. Coordinate, biphasic activation of p44 mitogen-activated protein kinase and Leukemia


168

2 3 4 5

6

7

8

9 10

11 12 13 14

15 16

17

18 19

20

Leukemia

S6 kinase by growth factors in hamster fibroblasts. Evidence for thrombin-induced signals different from phosphoinositide turnover and adenylylcyclase inhibition. J Biol Chem 1992; 267: 13369–13375. Qui MS, Green SH. PC12 cell neuronal differentiation is associated with prolonged p21ras activity and consequent prolonged ERK activity. Neuron 1992; 9: 705–717. Sharp LL, Schwarz DA, Bott CM, Marshall CJ, Hedrick SM. The influence of the MAPK pathway on T cell lineage commitment. Immunity 1997; 7: 609–618. Brummer T, Shaw PE, Reth M, Misawa Y. Inducible gene deletion reveals different roles for B-Raf and Raf-1 in B-cell antigen receptor signaling. EMBO J 2002; 21: 5611–5622. Matsumura I, Nakajima K, Wakao H, Hattori S, Hashimoto K, Sugahara H et al. Involvement of prolonged ras activation in thrombopoietin-induced megakaryocytic differentiation in a human factor-dependent hematopoietic cell line. Mol Cell Biol 1998; 18: 4282–4290. Garcia J, de Gunzburg J, Eychene A, Gisselbrecht S, Porteu F. Thrombopoietin-mediated sustained activation of extracellular signal-regulated kinase in UT7-Mpl cells requires both Ras-Raf-1and Rap1-B-Raf-dependent pathways. Mol Cell Biol 2001; 21: 2659–2670. Dorrell C, Takenada K, Minden MD, Hawley RG, Dick JE. Hematopoietic cell fate and the initiation of leukemic properties in primitive primary human cells are influenced by Ras activity and farnesyltransferase inhibition. Mol Cell Biol 2004; 24: 6993–7002. Gobert-Gosse S, Bourgin C, Liu WQ, Garbay C, Mouchiroud G. M-CSF stimulated differentiation requires persistent MEK activity and MAPK phosphorylation independent of Grb2-Sos association and phosphatidylinositol 3-kinase activity. Cell Signal 2005; 17: 1352–1362. Miranda MB, Xu H, Torchia JA, Johnson DE. Cytokine-induced myeloid differentiation is dependent on activation of the MEK/ERK pathway. Leuk Res 2005; 29: 1293–1306. Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berman K et al. Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions. Endocr Rev 2001; 22: 153–183. Torii S, Nakayama K, Yamamoto T, Nishida E. Regulatory mechanisms and function of ERK MAP kinases. J Biochem (Tokyo) 2004; 136: 557–561. Dard N, Peter M. Scaffold proteins in MAP kinase signaling: more than simple passive activating platforms. Bioessays 2006; 28: 146–156. Pixley FJ. Stanley ER CSF-1 regulation of the wandering macrophage: complexity in action. Trends Cell Biol 2004; 14: 628–638. Sasmono RT, Oceandy D, Pollard JW, Tong W, Pavli P, Wainwright BJ et al. A macrophage colony-stimulating factor receptor-green fluorescent protein transgene is expressed throughout the mononuclear phagocyte system of the mouse. Blood 2003; 101: 1155–1163. Dexter TM, Garland J, Scott D, Scolnick E, Metcalf D. Growth of factor-dependent hemopoietic precursor cell lines. J Exp Med 1980; 152: 1036–1047. Rohrschneider LR, Metcalf D. Induction of macrophage colonystimulating factor-dependent growth and differentiation after introduction of the murine c-fms gene into FDC-P1 cells. Mol Cell Biol 1989; 9: 5081–5092. Bourette RP, Myles GM, Carlberg K, Chen AR, Rohrschneider LR. Uncoupling of the proliferation and differentiation signals mediated by the murine macrophage colony-stimulating factor receptor expressed in myeloid FDC-P1 cells. Cell Growth Differ 1995; 6: 631–645. Fukunaga R, Ishizaka-Ikeda E, Pan CX, Seto Y, Nagata S. Functional domains of the granulocyte colony-stimulating factor receptor. EMBO J 1991; 10: 2855–2865. Nicholls SE, Winter S, Mottram R, Miyan JA, Whetton AD. Flt3 ligand can promote survival and macrophage development without proliferation in myeloid progenitor cells. Exp Hematol 1999; 27: 663–672. Bourette RP, Myles GM, Choi JL, Rohrschneider LR. Sequential activation of phoshatidylinositol 3-kinase and phospholipase Cgamma 2 by the M-CSF receptor is necessary for differentiation signaling. EMBO J 1997; 16: 5880–5893.

21 Rohde CM, Schrum J, Lee AW. A juxtamembrane tyrosine in the colony stimulating factor-1 receptor regulates ligand-induced Src association, receptor kinase function, and down-regulation. J Biol Chem 2004; 279: 43448–43461. 22 Marks DC, Csar XF, Wilson NJ, Novak U, Ward AC, Kanagasundarum V et al. Expression of a Y559F mutant CSF-1 receptor in M1 myeloid cells: a role for Src kinases in CSF-1 receptor-mediated differentiation. Mol Cell Biol Res Commun 1999; 1: 144–152. 23 Roussel MF. Regulation of cell cycle entry and G1 progression by CSF-1. Mol Reprod Dev 1997; 46: 11–18. 24 Karasuyama H, Melchers F. Establishment of mouse cell lines which constitutively secrete large quantities of interleukin 2, 3, 4, or 5 using modified cDNA expression vectors. Eur J Immunol 1988; 18: 97–104. 25 Wang ZE, Myles GM, Brandt CS, Lioubin MN, Rohrschneider LR. Identification of the ligand-binding regions in the macrophage colony-stimulating factor receptor extracellular domain. Mol Cell Biol 1993; 13: 5348–5359. 26 Bourgin C, Bourette RP, Arnaud S, Liu Y, Rohrschneider LR, Mouchiroud G. Induced expression and association of the Mona/ Gads adapter and Gab3 scaffolding protein during monocyte/ macrophage differentiation. Mol Cell Biol 2002; 22: 3744–3756. 27 Hanke JH, Gardner JP, Dow RL, Changelian PS, Brissette WH, Weringer EJ et al. Discovery of a novel, potent, and Src familyselective tyrosine kinase inhibitor. Study of Lck- and FynTdependent T cell activation. J Biol Chem 1996; 271: 695–701. 28 Bain J, McLauchlan H, Elliott M, Cohen P. The specificities of protein kinase inhibitors an update. Biochem J 2003; 371: 199–204. 29 Bourette RP, Arnaud S, Myles GM, Blanchet JP, Rohrschneider LR, Mouchiroud G. Mona, a novel hematopoietic-specific adapter interacting with the macrophage colony-stimulating factor receptor, is implicated in monocyte/macrophage development. EMBO J 1998; 17: 7273–7281. 30 Blake RA, Broome MA, Liu X, Wu J, Gishizky M, Sun L et al. SU6656, a selective src family kinase inhibitor, used to probe growth factor signaling. Mol Cell Biol 2000; 20: 9018–9027. 31 Dearden-Badet MT, Mouchiroud G. Re-distribution of phospholipase C gamma 2 in macrophage precursors is mediated by the actin cytoskeleton under the control of the Src kinases. Cell Signal 2005; 17: 1560–1571. 32 Bourgin C, Bourette RP, Mouchiroud G, Arnaud S. Expression of Mona (monocytic adapter) in myeloid progenitor cells results in increased and prolonged MAP kinase activation upon macrophage colony-stimulating factor stimulation. FEBS Lett 2000; 480: 113–117. 33 Tushinski RJ, Oliver IT, Guilbert LJ, Tynan PW, Warner JR, Stanley ER. Survival of mononuclear phagocytes depends on a lineagespecific growth factor that the differentiated cells selectively destroy. Cell 1982; 28: 71–81. 34 Morioka Y, Naito M, Sato T, Takahashi K. Immunophenotypic and ultrastructural heterogeneity of macrophage differentiation in bone marrow and fetal hematopoiesis of mouse in vitro and in vivo. J Leukoc Biol 1994; 55: 642–651. 35 de Bruijn MF, Slieker WA, van der Loo JC, Voerman JS, van Ewijk W, Leenen PJ. Distinct mouse bone marrow macrophage precursors identified by differential expression of ER-MP12 and ER-MP20 antigens. Eur J Immunol 1994; 24: 2279–2284. 36 Simoncic PD, Bourdeau A, Lee-Loy A, Rohrschneider LR, Tremblay ML, Stanley ER et al. T-cell protein tyrosine phosphatase (Tcptp) is a negative regulator of colony-stimulating factor 1 signaling and macrophage differentiation. Mol Cell Biol 2006; 26: 4149–4160. 37 Bourette RP, Grasset MF, Mouchiroud G. E2a/Pbx1 oncogene inhibits terminal differentiation but not myeloid potential of pro-T cells. Oncogene 2007; 26: 234–247. 38 van der Geer P, Hunter T. Mutation of Tyr697, a GRB2-binding site, and Tyr721, a PI 3-kinase binding site, abrogates signal transduction by the murine CSF-1 receptor expressed in Rat-2 fibroblasts. EMBO J 1993; 12: 5161–5172. 39 Lee AW, States DJ. Both src-dependent and -independent mechanisms mediate phosphatidylinositol 3-kinase regulation of colony-stimulating factor 1-activated mitogen-activated protein kinases in myeloid progenitors. Mol Cell Biol 2000; 20: 6779–6798.


169 40 Csar XF, Wilson NJ, McMahon KA, Marks DC, Beecroft TL, Ward AC et al. Proteomic analysis of macrophage differentiation. p46/52(Shc) Tyrosine phosphorylation is required for CSF-1-mediated macrophage differentiation. J Biol Chem 2001; 276: 26211–26217. 41 Carpenter G, Hernandez-Sotomayor SM, Nishibe S, Todderud G, Mumby M, Wahl M. Growth factor phosphorylation of PLCgamma 1. Ciba Found Symp 1992; 164: 223–233. 42 Cullen PJ, Lockyer PJ. Integration of calcium and Ras signaling. Nat Rev Mol Cell Biol 2002; 3: 339–348. 43 Springett GM, Kawasaki H, Spriggs DR. Non-kinase secondmessenger signaling: new pathways with new promise. Bioessays 2004; 26: 730–738. 44 Takeshita S, Faccio R, Chappel J, Zheng L, Feng X, Weber JD et al. c-Fms tyrosine 559 is a major mediator of M-CSF-induced proliferation of primary macrophages. J Biol Chem 2007; 282: 18980–18990. 45 Murphy LO, Smith S, Chen RH, Fingar DC, Blenis J. Molecular interpretation of ERK signal duration by immediate early gene products. Nat Cell Biol 2002; 4: 556–564. 46 Fowles LF, Martin ML, Nelsen L, Stacey KJ, Redd D, Clark YM et al. Persistent activation of mitogen-activated protein kinases p42 and p44 and ets-2 phosphorylation in response to colony-stimulating factor 1/c-fms signaling. Mol Cell Biol 1998; 18: 5148–5156. 47 Klappacher GW, Lunyak VV, Sykes DB, Sawka-Verhelle D, Sage J, Brard G et al. An induced Ets repressor complex regulates growth arrest during terminal macrophage differentiation. Cell 2002; 109: 169–180. 48 Krishnaraju K, Hoffman B, Liebermann DA. Early growth response gene 1 stimulates development of hematopoietic progenitor cells

49

50 51

52 53

54

55

along the macrophage lineage at the expense of the granulocyte and erythroid lineages. Blood 2001; 97: 1298–1305. Tanaka T, Kurokawa M, Ueki K, Tanaka K, Imai Y, Mitani K et al. The extracellular signal-regulated kinase pathway phosphorylates AML1, an acute myeloid leukemia gene product, and potentially regulates its transactivation ability. Mol Cell Biol 1996; 16: 3967–3979. Ross SE, Radomska HS, Wu B, Zhang P, Winnay JN, Bajnok L et al. Phosphorylation of C/EBPalpha inhibits granulopoiesis. Mol Cell Biol 2004; 24: 675–686. Marcinkowska E, Garay E, Gocek E, Chrobak A, Wang X, Studzinski GP. Regulation of C/EBPbeta isoforms by MAPK pathways in HL60 cells induced to differentiate by 1,25dihydroxyvitamin D3. Exp Cell Res 2006; 312: 2054–2065. Frohling S, Scholl C, Gilliland DG, Levine RL. Genetics of myeloid malignancies: pathogenetic and clinical implications. J Clin Oncol 2005; 23: 6285–6295. Radomska HS, Basseres DS, Zheng R, Zhang P, Dayaram T, Yamamoto Y et al. Block of C/EBP alpha function by phosphorylation in acute myeloid leukemia with FLT3 activating mutations. J Exp Med 2006; 203: 371–381. Glasow A, Prodromou N, Xu K, von Lindern M, Zelent A. Retinoids and myelomonocytic growth factors cooperatively activate RARA and induce human myeloid leukemia cell differentiation via MAP kinase pathways. Blood 2005; 105: 341–349. Studzinski GP, Wang X, Ji Y, Wang Q, Zhang Y, Kutner A et al. The rationale for deltanoids in therapy for myeloid leukemia: role of KSR-MAPK-C/EBP pathway. J Steroid Biochem Mol Biol 2005; 97: 47–55.

Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)

Leukemia