Published online before print July 20, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Biological Sciences, University of Toledo, Ohio; and
Department of Cancer Biology, Lerner Research Institute, Cleveland Clinic Foundation, Ohio
1Correspondence: Department of Biological Sciences, University of Toledo, 2801 West Bancroft Street, Toledo, OH 43606. E-mail: fdong{at}utnet.utoledo.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: cytokines • signal transduction • granulopoiesis
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
G-CSF stimulates the activation of Stat3, Stat5, and to a lesser extent, Stat1. Stat3 was identified as being critical for G-CSF-induced cell proliferation, differentiation, and survival [4 , 5 ], but a recent study using conditional gene ablation in transgenic mice indicated that Stat3 may negatively regulate granulopoiesis [6 ]. Stat5 exists in two major forms, i.e., Stat5a and Stat5b, which are encoded by two closely related genes, and has been shown to positively regulate cell proliferation and granulocytic differentiation induced by G-CSF [7 , 8 ]. Although Stat5a/b-deficient mice were not neutropenic, bone marrow cells from these mice exhibited reduced response to G-CSF stimulation in in vitro colony formation assay [9 ]. Apart from its role in normal granulopoiesis, Stat5 also appears to play a part in leukemogenesis. Expression of leukemogenic proteins such as Bcr-Abl, Tel-Jak2, and FMS-like fetal liver tyrosine kinase 3/internal tandem duplication is associated with constitutive activation of Stat5 [10 11 12 13 ]. Notably, deregulated activation of Stat5 alone can convert interleukin (IL)-3-dependent cells to growth factor independence [14 ].
Mutations in the G-CSF-R gene have been identified in a subgroup of patients with acute myelogenous leukemia (AML) evolving from severe congenital neutropenia (SCN) [15 16 17 18 ]. These mutations, which result in the carboxy terminal truncation of the G-CSF-R, are almost always present in the patients who developed AML after a history of SCN, strongly suggesting that the mutations have a role in leukemogenesis. Expression of the truncated G-CSF-R in hematopoietic cells leads to augmented cell proliferation and survival but defective granulocytic differentiation [15 , 19 ]. Notably, G-CSF-stimulated activation of Stat5 is markedly enhanced and prolonged in these cells [7 , 20 ], indicating that the carboxy terminal region of the G-CSF-R negatively regulates Stat5 activation. It has been shown that Src homology 2 (SH2)-containing protein tyrosine phosphatase-1 (Shp-1) negatively regulates the intensity, but not the duration, of G-CSF-stimulated Stat5 activation [21 ]. The negative effect of Shp-1 on Stat5 activation requires the carboxy terminus of the G-CSF-R, but the tyrosine residues located in the receptor’s carboxy terminus are dispensable for Shp-1 action [21 , 22 ]. In this paper, we show that tyrosine (Tyr) 729 of the G-CSF-R, located in the receptor’s carboxy terminus, controls the duration of G-CSF-stimulated activation of Stat5, Akt, and extracellular signal-regulated kianse 1/2 (Erk1/2). Our data also strongly suggest that suppressor of cytokine signaling (SOCS)1 and SOCS3 are involved in the negative regulation of G-CSF-R signaling, presumably through interacting with Tyr 729 of the G-CSF-R.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Reagents
Antibodies against Akt and Erk1/2 and phospho-specific antibodies against Akt, Erk1/2, and Stat5 were purchased from Cell Signaling Technology (Beverly, MA). Anti Stat5 (C-17) antibody was from Santa Cruz Biotechnology (CA). Monoclonal antibody to the G-CSF-R was obtained from BD Biosciences (San Diego, CA). Anti-FLAG (M2) antibody was from Sigma Chemical Co. (St. Louis, MO). [
-32P]Adenosine 5'-triphosphate (ATP) and an enhanced chemiluminescence (ECL) kit were purchased from PerkinElmer Life and Analytical Sciences (Boston, MA) and Pierce Biotechnology (Rockford, IL), respectively. FuGENE 6 transfection reagent was from Roche Diagnostics (Indianapolis, IN). Fluorophore-labeled goat anti-mouse immunoglobulin G (IgG) antibody (Alexa Fluor® 488) was purchased from Molecular Probes (Eugene, OR).
Stable transfection
The cDNAs encoding the different tyrosine-to-phenylalanine substitution mutants of the G-CSF-R (kindly provided by Dr. Ivo P. Touw, Erasmus University, Rotterdam) were cloned in the pBabe-puro retroviral expression vector as described [23
]. 32D cells were transfected by electroporation and selected in medium containing puromycin (1 µg/ml), 24 h after transfection. Individual clones were expanded and examined for expression of transfected proteins by Western blotting. Three independent clones were pooled and used in subsequent experiments.
Transient transfection
The expression vectors encoding FLAG-tagged cytokine-inducible SH2-domain-containing protein (CIS), SOCS1, SOCS2, SOCS3, and Stat5a have been described [7
, 24
]. 293T cells were plated on 24-well plates and transfected with 0.1 µg pBabe G-CSF-R, 0.5 µg pXM-Stat5a, and 0.1 µg SOCS expression vectors using FuGENE 6. In some experiments, higher amounts of SOCS expression vectors were used to see the expression of different SOCS proteins. Twenty hours after transfection, cells were deprived of serum for 4 h and stimulated with G-CSF for the times as indicated. Ba/F3 cells were transiently transfected with 10 µg empty vector or expression constructs for SOCS1 and SOCS3 by electroporation as described previously [25
].
Western blot analysis
Cells were starved in the absence of serum for 4 h and subsequently stimulated with G-CSF (20 ng/ml) for the times indicated. Cells were washed with ice-cold phosphate-buffered saline and resuspended in lysis buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 10 mM NaF, 0.5 mM dithiothreitol, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, and 1 mM vanadate). After incubation on ice for 20 min, lysates were cleared by centrifugation at 12,000 rpm for 20 min at 4°C. Supernatants were collected, and proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis prior to transfer to Immobilon membranes. The membranes were incubated with the appropriate antibodies. Western blots were developed using ECL.
Northern blot analysis
Total RNA was extracted with RNA-BeeTM RNA isolation reagent (Tel-Test Inc., Friendswood, TX). RNA (10 µg) was run on a 1% agarose gel and transferred to a Hybond-N membrane (Amersham Biosciences, Piscataway, NJ). The membrane was prehybridized for 1 h at 42°C in NorthernMaxTM hybridiztion buffer (Ambion, Austin, TX) and hybridized overnight at 42°C after addition of the [-32P]-labeled probe. The membrane was washed and developed.
Electrophoretic mobility shift assay (EMSA)
EMSAs were performed as described previously using whole cell extracts [7
]. The ß-casein probe (5'-AGATTTCTAGGAATTCAAATC-3'), derived from the promoter of the bovine ß-casein gene, was end-labeled using polynucleotide kinase and [-32P]ATP.
Flow cytometry
Cells (106) were incubated at 4°C for 30 min sequentially with the anti-G-CSF-R antibody (1 µg/100 µl) and the fluorophore-labeled goat anti-mouse IgG antibody (2 µg/100 µl), with washing between each step. Samples were analyzed by flow cytometry using a FACScan (Becton Dickinson, San Jose, CA).
Cell proliferation assay
Cell proliferation was measured using 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS; CellTiter 96® aqueous nonradioactive cell proliferation assay, Promega, Madison, WI). After extensive washing, 2 x 104 cells were incubated in triplicate in 100 µl RPMI-1640 medium in 96-well plates in the presence of G-CSF (20 ng/ml) for 24 h. MTS (10 µl) was added to each well, and the plates were read 2 h later at 490 nm wavelength in a microplate luminometer (Molecular Devices, Sunnyvale, CA).
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (39K): [in a new window] |
Figure 1. Expression of the different G-CSF-R in 32D cells. (A) Schematic representation of the different forms of the G-CSF-R. Shown are the transmembrane and cytoplasmic domains. Boxes 1–3 denote regions that are conserved among some members of the cytokine receptor superfamily. Cytoplasmic tyrosine residues are also indicated. (B) Flow cytometric analysis of G-CSF-R expression in 32D cells. Cells were incubated sequentially with the anti G-CSF-R antibody and the fluorophore-labeled goat anti-mouse IgG antibody at 4°C. Samples were analyzed by flow cytometry. Ctr, control.
|
 View larger version (20K): [in a new window] |
Figure 2. Activation of Stat5, Akt, and Erk1/2 by G-CSF in 32D cells expressing the different forms of the G-CSF-R. (A) 32D cells expressing the WT, D715, or mA receptor were starved in serum-free medium for 4 h prior to G-CSF stimulation for 15 min. Phosphorylation (p) of Stat5, Akt, and Erk1/2 was determined by Western blotting using the phospho-specific antibodies. The membranes were reprobed with antibodies to Stat5, Akt, and Erk1/2. (B) Survival of 32D cells upon G-CSF withdrawal. Cells, as indicated, were incubated with G-CSF for 8 h and then cultured in quadruplicate in medium containing no growth factors. Cell viability was determined by exclusion of trypan blue staining at the indicated times. The data are presented as mean ± SD of quadruplicate determinations. *, Significant differences as compared with the survival of 32D/WT cells (Student’s t-test; P<0.05). Comparable results were obtained in four independent experiments.
|
To determine which carboxy terminal tyrosine residue(s) of the G-CSF-R are involved in controlling the duration of G-CSF-activated signaling pathways, we compared Stat5 activation by G-CSF in 32D cells expressing the G-CSF-R mutants that contained single or double tyrosine-to-phenylalanine substitutions. Stat5 tyrosine phosphorylation and subsequent activation of its DNA binding activity in response to G-CSF were significantly prolonged when Tyr 729 was mutated to phenylalanine (Fig. 3A and 3B ). In contrast, substitution of Tyr 744 and/or Y764 had no significant effect on Stat5 activation. Mutation of Tyr 729 also significantly prolonged the activation of Akt and Erk1/2 by G-CSF. It appeared that prolongation of Erk1/2 activation by mA and Y729/764F mutants was less significant than that by the Y729F and Y729/744F mutant (Figs. 2A and 3A) , presumably, as Tyr 764 plays an important role in the activation of Erk1/2 by G-CSF [26 , 27 ]. Upon G-CSF withdrawal, 32D cells expressing the Y729F mutant (32D/Y729F) died at a rate almost identical to that of 32D/mA cells (Fig. 2B) . In addition to 32D cells, Stat5 activation was more sustained in Ba/F3 cells expressing the Y729F mutant than in cells expressing the WT G-CSF-R (Fig. 3C) . Together, these data established that Tyr 729 is critical for regulating the duration of G-CSF-activated signaling pathways.
 View larger version (48K): [in a new window] |
Figure 3. Activation of Stat5, Akt, and Erk1/2 by the different G-CSF-R forms. 32D and Ba/F3 cells expressing the different forms of the G-CSF-R were stimulated with G-CSF after starvation in serum-free medium for 4 h. (A) Activation of Stat5, Akt, and Erk1/2 in 32D cells was examined by immunoblotting with the phospho-specific antibodies as indicated. The membranes were subsequently probed with antibodies to Stat5, Akt, and Erk1/2. (B) Stat5 DNA binding activity in 32D cells was assessed by EMSA using the ß-casein probe. (C) Activation of Stat5 in Ba/F3 cells was determined by immunoblotting with the phospho-specific Stat5 antibody. The membrane was reprobed for Stat5.
|
 View larger version (54K): [in a new window] |
Figure 4. Effects of actinomycin D and cycloheximide on G-CSF-induced activation of Stat5, Akt, and Erk1/2. After starvation in serum-free medium for 2 h, 32D/WT cells were treated with actinomycin D (5 µg/ml) or cycloheximide (30 µg/ml) for an additional 2 h prior to stimulation with G-CSF. Whole cell extracts were prepared and examined for Stat5 tyrosine phosphorylation and DNA binding activity using the ß-casein probe and for Akt and Erk1/2 phosphorylation. The membranes were reprobed with the antibodies to Stat5, Akt, and Erk1/2.
|
 View larger version (29K): [in a new window] |
Figure 5. Effects of the different SOCS proteins on G-CSF-induced Stat5 activation. 293T cells were transiently transfected with cDNAs encoding the WT G-CSF-R and Stat5a, together with cDNAs encoding the different FLAG-tagged SOCS proteins or the empty vector. Twenty hours after transfection, cells were starved for 4 h and stimulated with G-CSF for the indicated times. Whole cell extracts were examined by Western blotting (WB) for Stat5 tyrosine phosphorylation (p-Stat5; top panel) and for expression of Stat5 (middle panel) and the different SOCS proteins using the anti-FLAG antibody (bottom panel).
|
45 min for SOCS3 and 60 min for SOCS1, suggesting that Tyr 729 of the G-CSF-R is required for prompt induction of SOCS3 and SOCS1 expression.
 View larger version (24K): [in a new window] |
Figure 6. Northern blot analysis of SOCS1 and SOCS3 transcripts. (A) 32D cells expressing the WT, D715, and Y729F forms of the G-CSF-R were stimulated with G-CSF for the indicated times following starvation in serum-free medium for 4 h. Total RNA extracted from the cells was electrophoresed in a 1% agarose gel, blotted to membrane, and probed with [-32P]-labeled probes for SOCS1 (top panel) and SOCS3 (middle panel). Sample loadings were determined by ethidium bromide staining of 18S rRNA (bottom panel). (B and C) The intensities of the bands for SOCS1 and SOCS3 transcripts were quantitated using a phosphoimager.
|
 View larger version (37K): [in a new window] |
Figure 7. Requirement of Tyr 729 of the G-CSF-R for SOCS1- and SOCS3-mediated inhibition of Stat5 activation. 293T cells were transiently transfected with Stat5a expression construct and the expression constructs for the WT and the different tyrosine-to-phenylalanine substitution forms of the G-CSF-R together with constructs for SOCS1, SOCS3, or empty vector. G-CSF-induced tyrosine phosphorylation of Stat5 was examined by Western blotting using the phospho-specific anti-Stat5 antibody. Expression of SOCS1 and SOCS3 was determined using the anti-FLAG antibody.
|
50% (data not shown). Cell proliferation in response to G-CSF was examined by MTS uptake assay. Expression of SOCS1 or SOCS3 inhibited G-CSF-dependent proliferation of BaF/WT cells (Fig. 8
). It is notable that the inhibitory effect of the two SOCS proteins was significantly weaker in BaF/Y729F cells than in BaF/WT cells, indicating that effective inhibition of G-CSF-dependent cell proliferation by SOCS1 and SOCS3 requires Tyr 729 of the G-CSF-R.
 View larger version (12K): [in a new window] |
Figure 8. Effects of SOCS1 and SOCS3 on G-CSF-dependent proliferation. BaF/WT and BaF/Y729F cells were transiently transfected with the empty vector (Ctr) or expression constructs for SOCS1 and SOCS3. Six hours after transfection, cells were washed and incubated in triplicate in serum-free medium for 24 h in the presence of G-CSF (20 ng/ml). Cells were then pulsed with MTS for 2 h, and MTS uptake was determined. Data are presented as percentage (mean±SD) of MTS uptake by cells transfected with the empty vector. Comparable results were obtained in three independent experiments. The differences in the rate of inhibition by SOCS1 and SOCS3 are statistically significant between BaF/WT and BaF/Y729F cells (Student’s t-test; P<0.01).
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our results are consistent with recent studies showing that SOCS3 binds to Tyr 729 of the G-CSF-R and inhibits Stat-dependent luciferase activity stimulated by G-CSF [35 36 37 ]. However, these studies did not examine Stat5 tyrosine phosphorylation and DNA binding activity in hematopoietic cells, and therefore, it remains unknown as to which aspects of Stat5 activation are affected by SOCS3. The data presented here strongly suggest that SOCS3 and SOCS1 control the duration, but not the magnitude, of Stat5 activation in hematopoietic cells. It appears that Tyr 729 controls the functions of SOCS3 and SOCS1 by at least two distinct mechanisms. Removal of Tyr 729 from the G-CSF-R through carboxy terminal truncation or amino acid substitution not only delays the induction of SOCS3 and SOCS1 transcripts by G-CSF but also diminishes the negative effect of SOCS3 and SOCS1 on Stat5 activation. Thus, Tyr 729 of the G-CSF-R regulates the rates at which SOCS3 and SOCS1 are up-regulated by G-CSF and the efficacies with which the two SOCS proteins inhibit G-CSF-R signaling. Conceivably, mutation of Tyr 729 causes prolonged G-CSF-R signaling by disrupting both negative regulatory mechanisms.
In addition to deregulated Stat5 activation, carboxy terminal truncation of the G-CSF-R results in sustained activation of Akt by G-CSF [38 , 39 ]. Our results indicate that Tyr 729 of the G-CSF-R is critically involved in controlling the duration of G-CSF-stimulated activation of Akt and Erk1/2. Activation of Akt and Erk1/2 by G-CSF was also prolonged upon inhibition of protein or RNA synthesis, suggesting that SOCS1 and SOCS3 may also exert a negative effect on the activation of Akt and Erk1/2. Indeed, expression of SOCS1 or SOCS3 in 293T cells attenuated G-CSF-induced Akt phosphorylation (data not shown). It is likely that SOCS1 and SOCS3 may target a common upstream component of the Stat5, Akt, and Erk1/2 pathways, although one cannot exclude the possibility that the two SOCS proteins may directly inhibit the activation of individual components within the three pathways. As the Jak and Src family kinases are differentially required for G-CSF-stimulated activation of Stat5 and Akt, respectively [38 ], one potential, common target of SOCS1 and SOCS3 is the G-CSF-R. In support of this, we have observed that tyrosine phosphorylation of the G-CSF-R induced by G-CSF was completely blocked upon expression of SOCS1 or SOCS3 in 293T cells (data not shown).
A recent study showed that CIS was also up-regulated by G-CSF and bound to phosphopeptides corresponding to Tyr 729 and Tyr 744 of the G-CSFR [40 ]. However, we consistently observed that expression of CIS enhanced Stat5 activation by G-CSF in 293T cells, consistent with the report by van de Geijn et al. [37 ]. It remains to be determined, with respect to the relative contribution of SOCS1 and SOCS3 to the attenuation of G-CSF-R signaling and the precise mechanism by which the two SOCS proteins down-regulate the activation of G-CSF-stimulated pathways. In addition to Tyr 729, SOCS3 has been shown to bind to Tyr 704 of the G-CSF-R with a reduced affinity [35 ], which may explain the partial inhibitory effect of SOCS3 on Stat5 activation mediated by the Y729F mutant. In contrast to SOCS3, it is generally believed that SOCS1 inhibits cytokine receptor signaling via interacting directly with the Jak family kinases but not with the cytokine receptors [34 , 41 ]. In fact, it has been shown that SOCS1 fails to bind to any of the cytoplasmic tyrosine residues including Tyr 729 of the G-CSF-R [35 ]. We also have been unable to detect any interaction between SOCS1 and the G-CSF-R (unpublished data). However, our data clearly show that Tyr 729 of the G-CSF-R is required for the maximal inhibitory effect of SOCS1 on G-CSF-stimulated Stat5 activation. Whether SOCS1 may interact indirectly with Tyr 729 of the G-CSF-R remains to be examined.
Regardless of the mechanisms whereby SOCS3 and SOCS1 exert their negative effect on G-CSF-R signaling, the results presented here together with the previously report [21 ] demonstrate that the intensity and duration of G-CSF-stimulated Stat5 activation are regulated by at least two distinct regulatory mechanisms, i.e., the one mediated by Shp-1, which reduces the magnitude of Stat5 activation, and the one mediated by SOCS3 and/or SOCS1, which controls the duration of Stat5 activation (Fig. 9 ). To our knowledge, such a regulatory feature has not been reported for other cytokine receptors. Additionally, our data reveal an important molecular mechanism that explains, at least in part, why truncation of the carboxy terminal region of the G-CSF-R leads to prolonged receptor signaling and may thus contribute to leukemogenesis.
 View larger version (22K): [in a new window] |
Figure 9. A model of negative regulation of G-CSF-stimulated Stat5 activation. The G-CSF-R form dimers or oligomers upon ligand binding. Shp-1 is constitutively expressed in myeloid cells and controls the magnitude of Stat5 activation via interacting with the carboxy terminus of the G-CSF-R. Carboxy terminal tyrosine residues of the receptor appear dispensable for Shp-1 action [21
, 22
]. SOCS3 and SOCS1 are up-regulated rapidly by signals transduced from the G-CSF-R and control the duration of Stat5 activation. Tyr 729 of the G-CSF-R is required for the maximal inhibitory activities of SOCS3 and SOCS1.
|
Received January 19, 2005; revised June 12, 2005; accepted June 16, 2005.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. M. Cheng, J. C. B. Li, S. S. Lin, D. C. W. Lee, L. Liu, Z. Chen, and A. S. Y. Lau HIV-1 transactivator protein induction of suppressor of cytokine signaling-2 contributes to dysregulation of IFN{gamma} signaling Blood, May 21, 2009; 113(21): 5192 - 5201. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Jiang, N. Hein, K. Eckert, J. Luscher-Firzlaff, and B. Luscher Regulation of the MAD1 promoter by G-CSF Nucleic Acids Res., March 1, 2008; 36(5): 1517 - 1531. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Zhuang, Y. Qiu, S. C. Kogan, and F. Dong Increased CCAAT Enhancer-binding Protein {epsilon} (C/EBP{epsilon}) Expression and Premature Apoptosis in Myeloid Cells Expressing Gfi-1 N382S Mutant Associated with Severe Congenital Neutropenia J. Biol. Chem., April 21, 2006; 281(16): 10745 - 10751. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
