STAT-3 Activation Is Required for Normal G-CSF-Dependent Proliferation and Granulocytic Differentiation

Immunity, Vol. 14, 193–204, February, 2001, Copyright 2001 by Cell Press

STAT-3 Activation Is Required for Normal G-CSF-Dependent Proliferation and Granulocytic Differentiation Morgan L. McLemore, Satkiran Grewal, Fulu Liu, Angela Archambault, Jennifer Poursine-Laurent, Jeff Haug, and Daniel C. Link* Division of Oncology Department of Internal Medicine Washington University School of Medicine St. Louis, Missouri 63110

Summary To investigate the role of signal transducer and activator of transcription (STAT) proteins in granulocyte colony-stimulating factor (G-CSF)-regulated biological responses, we generated transgenic mice with a targeted mutation of their G-CSF receptor (termed d715F) that abolishes G-CSF-dependent STAT-3 activation and attenuates STAT-5 activation. Homozygous mutant mice are severely neutropenic with an accumulation of immature myeloid precursors in their bone marrow. G-CSF-induced proliferation and granulocytic differentiation of hematopoietic progenitors is severely impaired. Expression of a constitutively active form of STAT-3 in d715F progenitors nearly completely rescued these defects. Conversely, expression of a dominant-negative form of STAT-3 in wild-type progenitors results in impaired G-CSF-induced proliferation and differentiation. These data suggest that STAT-3 activation by the G-CSFR is critical for the transduction of normal proliferative signals and contributes to differentiative signals. Introduction Granulocyte colony-stimulating factor (G-CSF) is the principal hematopoietic growth factor regulating the production of neutrophils, and it is widely used to treat neutropenia in a variety of clinical settings (Welte et al., 1996). The effects of G-CSF are solely mediated through its interaction with the granulocyte colony-stimulating factor receptor (G-CSFR), a member of the hematopoietic cytokine receptor superfamily. G-CSF has three major biologic actions. It increases the production and release of neutrophils from the marrow, mobilizes hematopoietic progenitor cells from the marrow to the blood, and modulates certain effector functions of mature neutrophils (de Haas et al., 1994; Bober et al., 1995; Welte et al., 1996; Carulli, 1997; Hoglund et al., 1997). The importance of G-CSF to the regulation of basal granulopoiesis has been confirmed by the observation of severe neutropenia in mice carrying homozygous deletions of their G-CSF or G-CSFR genes (Lieschke et al., 1994; Liu et al., 1996). The contribution of G-CSFR signals to the regulation of granulocytic differentiation is controversial. Early studies showed that treatment of certain immortalized * To whom correpondence should be addressed (e-mail: dlink@im. wustl.edu).

hematopoietic cells lines with G-CSF resulted in their granulocytic differentiation (Dong et al., 1993; Fukunaga et al., 1993; Ziegler et al., 1993). These studies led to the hypothesis that G-CSFR signals were actively directing terminal granulocytic differentiation. However, in mice lacking G-CSF or the G-CSFR, morphologically mature neutrophils are produced (albeit at reduced levels), showing that G-CSFR signals are not required for granulocytic differentiation (Lieschke et al., 1994; Liu et al., 1996). Although not required, these observations do not exclude a significant role for G-CSFR signals in the regulation of granulocytic differentiation. Indeed, recent studies suggest that G-CSFR signals may regulate expression of PU.1 and CCAAT enhancer binding protein ⑀ (C-EBP⑀), two key transcription factors controlling granulocytic differentiation (H. Nakajima et al., 1998, Blood, abstract; Lutz et al., 2000). These latter observations may provide a potential mechanism for the longstanding observation that G-CSF treatment reduces the “transit time” for myeloid precursors to differentiate to mature neutrophils (Lord et al., 1989). To investigate the specificity of cytokine receptor signals in the regulation of hematopoietic differentiation, we recently generated mice carrying a targeted mutation of their G-CSFR such that the cytoplasmic (signaling) domain of the G-CSFR is replaced with that of the erythropoietin receptor (Semerad et al., 1999). These mice are predicted to express, in a myeloid-specific fashion, a chimeric receptor (G:EpoR) that is activated by G-CSF but transduces eryhtropoietin-specific signals. We showed that morphologically mature neutrophils were still produced in these mice and that G:EpoR signals could support terminal granulocytic differentiation in vitro. However, mutant neutrophils displayed defective chemotaxis, and their trafficking from the bone marrow to blood was impaired. Collectively, these data suggested that the G-CSFR is generating unique signals that are required for certain specialized hematopoietic cell functions but are not required for granulocytic differentiation. Signal transducer and activator of transcription (STAT) proteins are latent cytoplasmic transcription factors that are activated by a broad range of hematopoietic cytokine receptors (Chakraborty and Tweardy, 1998; Ward et al., 2000). Recent data suggest that STAT proteins may play an important role in the regulation of hematopoietic cell proliferation, differentiation, and function (Smithgall et al., 2000; Ward et al., 2000). Moreover, perturbations in the regulation of their activation may contribute to the pathogenesis of certain hematopoietic disorders, including myeloproliferative and lymphoproliferative disorders (Gouilleux-Gruart et al., 1997; CatlettFalcone et al., 1999; Ward et al., 2000). STAT-3 is the principal STAT protein activated by the G-CSFR, with STAT-5 and STAT-1 also activated to a lesser degree (Tian et al., 1994; Avalos, 1996; Chakraborty et al., 1996; Ward et al., 1999). The contribution of STAT-3 to the signal transduction of the G-CSFR’s diverse biological activities is largely unknown. Expression of dominantnegative forms of STAT-3 in certain myeloid cell lines leads to sustained G-CSF proliferation without granulo-

Immunity 194

cytic differentiation, indicating that STAT-3 activation is required for G-CSF-dependent granulocytic differentiation (Shimozaki et al., 1997; de Koning et al., 2000). In contrast, in transgenic mice in which the STAT-3 gene was deleted in a myeloid-specific fashion, the production of morphologically mature neutrophils was apparently normal, although significant functional defects were detected in mature neutrophils (Takeda et al., 1999). To further characterize the contribution of STAT-3 to the regulation of granulopoiesis, we generated a targeted mutation in mice that results in the production of a mutant G-CSFR (termed d715F) in which the distal 98 amino acids are deleted and the sole remaining cytoplasmic tyrosine (Y704) is mutated to phenylalanine. Importantly, we previously showed that mice expressing an identically truncated G-CSFR but with Y704 intact (termed d715) had normal granulopoiesis and activated STAT-3 in response to G-CSF (McLemore et al., 1998). Herein, we demonstrate that activation of STAT-3 by the d715F G-CSFR is absent in primary hematopoietic cells and STAT-5 activation markedly attenuated. G-CSFinduced cellular proliferation and differentiation by the d715F G-CSFR is impaired leading to defective granulopoiesis in vivo. These defects could be partially rescued by expression of a constitutively active form of STAT3, suggesting that activation of STAT-3 is critical for G-CSFinduced cellular proliferation and possibly granulocytic differentiation. Conversely, expression of a dominantnegative form of STAT-3 in wild-type progenitors results in impaired G-CSF-induced proliferation and differentiation, providing further support for the hypothesis that STAT-3 activation plays a critical role in G-CSF-dependent proliferation and granulocytic differentiation. Results Generation of G-CSFR Mutant Mice Mice carrying two different targeted G-CSFR mutations were examined in this study (Figure 1A). The generation and characterization of mice homozygous for the d715 G-CSFR mutation has been described previously (McLemore et al., 1998). The d715 G-CSFR mutation introduces a premature stop codon that truncates the carboxy-terminal 96 amino acids of the G-CSFR and reproduces the mutation found in a patient with severe congenital neutropenia. The d715F G-CSFR allele was generated by introducing into the d715 allele an A→T change at nucleotide 2362 via homologous recombination in embryonic stem (ES) cells (Figure 1B). Thus, the d715F allele is predicted to produce a truncated G-CSFR in which the sole remaining cytoplasmic tyrosine (Y704) is mutated to phenylalanine. Mice carrying the d715F allele were generated from a single targeted ES clone. The fidelity of the d715F G-CSFR allele was confirmed by sequence analysis of PCR-amplified genomic DNA isolated from mutant mice (data not shown). The d715F G-CSFR allele was inherited in a Mendelian fashion. Homozygous d715F mice have normal growth, development, and fertility and are grossly indistinguishable from wild-type littermates. Cell surface expression of the d715F G-CSFR on hematopoietic cells from d715F mice was analyzed using

a quantitative flow cytometric method to detect specific binding of biotinylated G-CSF. Specific G-CSF binding was detected on d715F neutrophils (Figure 1C) but not lymphocytes (data not shown), indicating myeloid-specific expression. Relative to wild-type neutrophils, a modest increase in G-CSF binding was observed with d715F and, as previously described, with d715 neutrophils. Since the observed increase in G-CSF binding could represent an increase in receptor number or receptor affinity for ligand, ligand binding studies on wildtype or d715F bone marrow cells were performed using radiolabeled G-CSF. The calculated dissociation constants (Kd) for wild-type and d715F cells were similar (137 and 155 pM, respectively) and are consistent with previously published data for wild-type murine G-CSFR (Nicola and Metcalf, 1985; Watanabe et al., 1991). Collectively, these data suggest that the number of G-CSFR receptors is modestly increased on d715F neutrophils but their binding affinity for G-CSF is normal. STAT-3 and STAT-5 Activation by the d715F G-CSFR Are Markedly Impaired Previous studies have shown that stimulation of the d715 G-CSFR results in near normal STAT-3 activation and sustained activation of STAT-5 (Hermans et al., 1998, 1999). To assess the ability of the d715F G-CSFR to activate STAT proteins, bone marrow cells were harvested from mice of each genotype, stimulated with G-CSF (or control cytokines), and the activation of STAT proteins assessed using an electrophoretic mobility shift assay (Figure 2A). As expected, STAT-3 activation was readily observed 10 min after G-CSF stimulation of wildtype and d715 bone marrow cells. In contrast, no STAT-3 activation was detected in d715F bone marrow cells anytime from 5–120 min after G-CSF stimulation (Figure 2A; data not shown). Surprisingly, STAT-5 activation by the d715F G-CSFR also was markedly impaired. Although a low level of STAT-5 activation could be detected after prolonged exposure to G-CSF (at least 30 min), no STAT-5 activation was detected 10 min after stimulation with G-CSF (Figure 2B; data not shown). STAT-1 activation by the d715F G-CSFR, though not readily apparent from Figure 2A, was consistently observed at a level similar to wild-type after extended autoradiography (data not shown). The failure of the d715F G-CSFR to activate STAT-3 or STAT-5 was not secondary to a loss of myeloid cells or STAT protein expression, since IL-6 and IL-3 induced a similar degree of STAT-3 and STAT-5 activation in all samples. The d715F Mice Have an Isolated Defect in Granulopoiesis Homozygous d715F mice have normal white blood cell, red blood cell, and platelet counts (Table 1). Moreover, the number and morphology of circulating lymphocytes, monocytes, and eosinophils are normal (data not shown). However, homozygous d715F mice are neutropenic, with levels of circulating neutrophils ⵑ15% that of wild-type mice and similar to that observed in G-CSFRdeficient mice (Liu et al., 1996). In the bone marrow, the number and morphology of the erythroid, megakaryocytic, and lymphoid lineages were within normal ranges. In contrast to the severity of the neutropenia, only a

STAT-3 Activation by G-CSF 195

Figure 1. Generation of Targeted Transgenic Mice and Characterization of Mutant G-CSFR Expression (A) G-CSFR mutations. Schematic representations of the wild-type and mutant G-CSFRs. The extracellular (ED), transmembrane (TM), and cytoplasmic (CD) domains of the G-CSFR are shown. The positions of cytoplasmic tyrosines (Y) and conserved box 1 and box 2 motifs are indicated. (B) Targeting strategy. The genomic organization of the 3⬘ region of the murine G-CSFR gene is shown in the upper panel. Coding exons are shown as black boxes and the 3⬘ untranslated region as a hatched box. The targeting vector is shown in the middle panel. The asterisk indicates the C→G and A→T mutations at nucleotides 2403 and 2362, respectively. The recombined allele (Mt) is shown in the lower panel. Note, the neomycin phosphotransferase gene driven by the phosphoglycerate kinase I promoter (neo) is present in the 3⬘ untranslated region. (C) Flow cytometric analysis of biotinylated G-CSF binding to peripheral blood leukocytes. Peripheral blood leukocytes isolated from wild-type, d715, and d715F mice were incubated with biotinylated G-CSF in the absence (unshaded histogram) or presence (shaded histogram) of a 100 M excess of nonlabeled G-CSF. Specific binding of biotinylated G-CSF is represented by the difference between the two curves. Results shown are gated on the neutrophil population.

modest reduction in neutrophils was observed in the bone marrow of d715F mice. Interestingly, a relative accumulation of myeloid precursors was observed in the bone marrow of d715F mice. In contrast, G-CSFRdeficient mice demonstrate a uniform reduction in myeloid precursors, suggesting that G-CSF-induced granulocytic differentiation may be impaired in d715F mice (Liu et al., 1996). To assess the in vivo response to G-CSF, mice were treated with G-CSF (10 ␮g/kg daily) for 7 days and the neutrophil response in the blood assessed. As reported previously, d715 mice have an exaggerated increment in circulating neutrophils in response to G-CSF (Figure 3) (McLemore et al., 1998). In contrast, the increment in circulating neutrophils in d715F mice was markedly less than that observed with d715 mice and actually less than that observed with wild-type mice. Even after considering the 2- to 3-fold difference in the degree of leukocytosis induced by G-CSF in C57BL/6 versus 129SvJ mice (Roberts et al., 1997), the difference between 715 and d715F mice appears significant (all mice in this study are outbred on a C57BL/6 ⫻ 129SvJ background). Interestingly, in contrast to wild-type and d715 mice, no evidence of STAT-3 activation was observed in bone

marrow cells isolated from 715F mice after 7 days of treatment with G-CSF (Figure 2C), suggesting that STAT-3-independent mechanisms were responsible for the neutrophil response in d715F mice.

G-CSF-Induced Cellular Proliferation Is Deficient in d715F Progenitor Cells To further characterize the ability of the d715F G-CSFR to mediate G-CSF biological responses, the G-CSFdependent proliferation and differentiation of hematopoietic progenitors isolated from these mice was examined. The number of IL-3-responsive progenitor cells in the bone marrow of d715F mice was similar to that observed with wild-type and d715 cells, indicating that the production of myeloid progenitors was normal in these mice (Figure 4). In contrast, G-CSF-induced colony formation was markedly impaired. In fact, no G-CSFdependent colonies (based on standard criteria, defined as containing at least 50 cells) were observed in cultures of d715F bone marrow cells (Metcalf, 1984). Instead of the typical G-CSF-induced colony that averages approximately 1000 cells per colony, d715F cells produced only small clusters (15–20 cells) in response to G-CSF.

Immunity 196

cells were transduced with STAT-3C or vector alone (MSCV) and their ability to form colonies in response to G-CSF assessed. Transduction efficiency of hematopoietic progenitor cells, based on EGFP expression in the lineage-negative cell compartment, was 30%–40% for STAT-3C and 50%–60% for MSCV (data not shown). In wild-type or d715 cells, expression of STAT-3C did not result in cytokine-independent colony formation nor did it consistently effect IL-3 or G-CSF responsiveness (Figure 5). Likewise, in d715F cells, expression of STAT-3C did not result in cytokine-independent colony formation nor significantly effect IL-3-dependent colony formation. However, expression of STAT-3C did partially restore their ability to form colonies in response to G-CSF. In fact, after accounting for transduction efficiency (30%–40% for STAT-3C), it appears that the G-CSF responsiveness of STAT-3C expressing d715F progenitor cells is similar to that of wild-type progenitor cells. Consistent with this observation, the G-CSF-dependent colonies produced by STAT-3C-transduced d715F progenitor cells were of nearly normal size; the number of cells per colony (mean of three independent experiments) for d715F cells transduced with MSCV or STAT-3C was ⬍100 and 462, respectively (compared with 1126 and 827 cells per colony for wild-type cells transduced with MSCV or STAT-3C, respectively). Collectively, these data show that the expression of STAT-3C in d715F progenitor cells partially rescues G-CSF-induced cellular proliferation.

Figure 2. Activation of STAT Proteins by the d715F G-CSFR (A and B) Bone marrow cells isolated from wild-type (Wt) and homozygous d715 and d715F G-CSFR mutant mice were stimulated with G-CSF (100 ng/ml), IL-6 (20 ng/ml), IL-3 (10 ng/ml), or no cytokine for 10 min. Nuclear extracts were incubated with radiolabeled m67 probe (A and C) or ␤-cas probe (B) and complexes resolved on a nondenaturing polyacrylamide gel. The identity of specific complexes was confirmed by supershifting with specific antibodies (data not shown). (C) The indicated mice were treated with G-CSF (10 ␮g/kg/day) for 7 days. Bone marrow cells were isolated 4 hr after the last injection and analyzed for STAT-3 activation, as described above.

Expression of a Constitutively Active Form of STAT-3 in d715F Bone Marrow Cells Restores G-CSF-Induced Cellular Proliferation To determine if the loss of G-CSF-induced cellular proliferation in d715F progenitor cells was due to loss of STAT-3 activation, a constitutively active form of STAT-3 was introduced into bone marrow cells via retroviral transduction. This mutant STAT-3 protein (termed STAT3C) contains two cysteine substitutions in its SH2 domain, which results in spontaneous dimerization (Bromberg et al., 1999). STAT-3C was introduced into a murine stem cell virus (MSCV) retroviral vector containing an internal ribosomal entry sequence (IRES) motif linked to the enhanced green fluorescent protein (EGFP) to track transduced cells. Wild-type, d715, or d715F bone marrow

Activation of STAT-3 May Contribute to G-CSFInduced Granulocytic Differentiation As noted above, there is an accumulation of myeloid precursors in the bone marrow of d715F mice, suggesting that granulocytic differentiation may be impaired. However, the neutrophils present in the blood of d715F mice following G-CSF treatment are morphologically normal and express normal levels of MMP-9 (gelatinase B), a protease expressed at the metamyelocyte/band stage of differentiation and widely considered a marker of terminal granulocytic differentiation (data not shown) (Borregaard and Cowland, 1997). Since many cytokines contribute to granulopoiesis in vivo, we examined G-CSF-dependent granulocytic differentiation of d715F progenitor cells. Morphologic analysis of cells contained in colonies generated after 7 days in G-CSF-stimulated methylcellulose cultures revealed that, while wild-type or d715 colonies contained 40%–50% mature neutrophils, d715F “colonies” contained less than 5% neutrophils (data not shown). To further examine the contribution of STAT-3 to G-CSFdependent granulocytic differentiation, a system was developed to examine in detail the differentiation of d715F progenitor cells transduced with STAT-3C or MSCV alone. Retrovirally transduced (EGFP⫹), lineagenegative bone marrow cells were sorted by flow cytometry and cultured for 7 days in the presence of G-CSF (100 ng/ml) and stem cell factor (SCF) (100 ng/ml); the lineage-negative cell population is enriched for progenitors and, importantly, contains no myeloid precursors or mature myeloid cells. Stem cell factor was added to the culture to support the growth of early progenitors,

STAT-3 Activation by G-CSF 197

Table 1. Blood and Bone Marrow Analysis

Blood Number of mice analyzed WBC (⫻ 10⫺9/L) RBC (⫻ 10⫺12/L) Platelets (⫻ 10⫺9/L) Neutrophils (%) Neutrophils (⫻ 10⫺9/L) Bone Marrow Total nucleated cells (⫻ 10⫺6) (per femur) Band and segmented neutrophil (%) Metamyelocyte neutrophil (%) Myelocyte neutrophil (%) Promyelocytes (%) Myeloblasts (%) Eosinophil lineage (%) Lymphoid lineage (%) Erythroid lineage (%) Myeloid/erythroid ratio

Wild-Type

d715

d715F

15 9.74 ⫾ 1.06 8.92 ⫾ 0.26 1,025 ⫾ 66 10.9 ⫾ 4.6 1.07 ⫾ 0.27

10 10.00 ⫾ 1.04 9.01 ⫾ 0.19 980 ⫾ 72 8.8 ⫾ 2.6 0.88 ⫾ 0.24

9 9.56 ⫾ 2.60 10.20 ⫾ 0.56 1,098 ⫾ 194 2.0 ⫾ 1.2* 0.16 ⫾ 0.04*

28.1 ⫾ 6.0 28.2 ⫾ 5.0 13.6 ⫾ 0.9 6.6 ⫾ 1.1 2.4 ⫾ 0.4 0.1 ⫾ 0.0 2.4 ⫾ 0.4 18.1 ⫾ 3.9 28.5 ⫾ 2.8 2.9 ⫾ 0.3

32.9 ⫾ 5.8 33.7 ⫾ 2.8 14.5 ⫾ 0.9 7.6 ⫾ 1.0 2.5 ⫾ 0.4 0.1 ⫾ 0.1 3.3 ⫾ 0.6 13.6 ⫾ 0.9 25.1 ⫾ 1.2 3.5 ⫾ 0.4

29.3 ⫾ 6.0 19.3 ⫾ 8.9* 8.8 ⫾ 1.6* 12.8 ⫾ 4.4* 3.5 ⫾ 1.9 0.2 ⫾ 0.0 3.1 ⫾ 0.6 23.8 ⫾ 3.8 26.8 ⫾ 6.7 2.2 ⫾ 0.9

Manual 300-count leukocyte differentials were performed on blood smears from 5- to 6-week-old mice. Bone marrow analysis was based on 500-count manual leukocyte differentials performed on nucleated cells recovered from the femurs of six to eight age-matched mice of each genotype. Data represent the mean ⫾ SD. p ⬍ 0.05 compared to wild-type and d715.

since preliminary studies showed minimal cell growth with G-CSF or stem cell factor alone (data not shown). These culture conditions support a robust and nearly synchronized wave of granulopoiesis that allows for the detailed study of G-CSF-dependent granulocytic differentiation. In cultures of MSCV- or STAT-3C-transduced wild-type cells, ⵑ50% of the cells present after 7 days

were mature neutrophils (Figure 6). In contrast, in cultures of MSCV-transduced d715F cells, the majority of cells were myeloid precursors with only 5% mature neutrophils. Expression of STAT-3C in d715F progenitors led to an increase in both total cell number and percentage of mature neutrophils. Importantly, no significant growth of STAT-3C-transduced bone marrow cells was observed in cultures stimulated with SCF alone. Collec-

Figure 4. Hematopoietic Progenitor Assays Figure 3. In Vivo Response to G-CSF Five age- and sex-matched mice of each genotype were treated with human G-CSF (10 ␮g/kg/day) for 7 days. The absolute neutrophil count in the blood was determined before and 4 hr after the last injection of G-CSF. Data represent the mean ⫾ SD. *P ⬍ 0.05 versus G-CSF-treated d715 mice.

Bone marrow mononuclear cells were plated in methylcellulosecontaining media supplemented with 100 ng/ml of G-CSF or 10 ng/ ml of IL-3. Hematopoietic colonies containing 50 or more cells were scored on day 7. The data were normalized to the number of colonies present in cultures of wild-type cells. Data represent the mean ⫾ SD. ⫹P ⬍ 0.05 versus wild-type and d715F; *only small clusters containing 15–20 cells were observed. Data represent the mean ⫾ SD.

Immunity 198

Figure 5. Expression of STAT-3C in d715F Hematopoietic Progenitor Cells Restores Their G-CSF Responsiveness Bone marrow cells from the indicated genotype were retrovirally transduced with MSCV or STAT-3C and their growth in response to no cytokine, G-CSF (10 ng/ml), or IL-3 (10 ng/ml) measured in standard methylcellulose cultures. 2000 (IL-3) or 5000 (no cytokine and G-CSF) cells per dish were plated and the total number of colonies scored on day 7. Transduction efficiency of hematopoietic progenitors, as estimated by EGFP expression in lineage-negative cells, was 30%–40% for STAT-3C and 50%–60% for MSCV (data not shown). *No colonies were observed in the absence of cytokine. The data are representative of three independent experiments.

tively, these data suggest that STAT-3 activation contributes to G-CSF-dependent granulocytic differentiation. Expression of a Dominant-Negative Form of STAT-3 in Wild-Type Bone Marrow Cells Impairs Their G-CSF-Dependent Proliferation and Differentiation To further examine the contribution of STAT-3 to G-CSFdependent granulocytic differentiation and proliferation, a well-characterized dominant-negative form of STAT-3 (STAT-3D) was introduced into wild-type bone marrow cells via retroviral transduction, as described above (Nakajima et al., 1996). To characterize STAT-3D expression, EGFP-positive lineage-negative cells were sorted by flow cytometry and expanded for 5 to 6 days in media supplemented with IL-3 and SCF ligand. STAT-3 and

STAT-5 activation by G-CSF in these cells were assayed using gel shift assays to monitor DNA binding activity (Figures 7B and 7C). In cells transduced with STAT-3D, G-CSF-dependent activation of STAT-3 was significantly attenuated, although residual activation was detected after stimulation with 10 ng/ml of G-CSF. G-CSF-induced activation of STAT-5 was modest but comparable in wild-type cells transduced with MSCV or STAT-3D. Furthermore, no consistent effect of STAT-3D expression was observed in IL-3-induced activation of STAT-5, suggesting that STAT-3D had no significant effect on STAT-5 activation. We next examined the effect of STAT-3D expression on IL-3 and G-CSF-induced colony formation (Figure 7A). Expression of STAT-3D led to a modest, but nonsignificant, reduction in the number of IL-3-dependent col-

Figure 6. G-CSF-Induced Granulocytic Differentiation Bone marrow cells were transduced with MSCV or STAT-3C. Transduced (EGFP-positive) lineage-negative cells were sorted by flow cytometry and cultured in the presence of G-CSF (100 ng/ml) and stem cell factor (100 ng/ml) for 7 days. (A) 200-count manual leukocyte differentials of cells present on day 7 of culture. (B) Photomicrograph of Wright-Giemsa stained cells. Original magnification: 100⫻ (upper panel) and 250⫻ (lower panel). Arrows indicate mature neutrophils. The data are representative of three independent experiments.

STAT-3 Activation by G-CSF 199

Figure 7. Expression of STAT-3D in WildType Hematopoietic Progenitor Cells Inhibits G-CSF-Dependent Colony Formation (A) Bone marrow cells were retrovirally transduced with MSCV or STAT-3D, and their growth in response to IL-3 and G-CSF was measured in standard methylcellulose cultures. Transduction efficiency of hematopoietic progenitors, as estimated by EGFP expression in lineage-negative cells, was 60%–80% for STAT-3D and 80%–90% for MSCV (data not shown). Due to significant interexperimental variability, the colony data for each experiment were normalized to the number of colonies observed in cultures of MSCV-transduced cells stimulated with IL-3. Data represent the mean ⫾ SD of five independent experiments. *P ⬍ 0.05 versus MSCV. (B and C) Transduced (EGFP-positive) lineage-negative cells were sorted by flow cytometry, cultured in the presence of kit ligand (100 ng/ml) and IL-3 (20 ng/ml) for 5 to 6 days, and then analyzed for STAT-3 and STAT-5 activation. In (B), the cells were stimulated with no cytokine (0) or G-CSF (at the indicated concentration in ng/ml) for 10 min; in (C), the cells were stimulated with no cytokine, G-CSF, or IL-3 (10 ng/ml) for 10 min. Equal amounts of nuclear extracts (except for IL-3-stimulated cells in which 5-fold less extract was used) were incubated with radiolabeled m67 probe (B) or ␤-cas probe (C) and complexes resolved on a nondenaturing polyacrylamide gel.

onies. In contrast, STAT-3D expression resulted in a dose-dependent reduction in both the number and size of colonies formed in response to G-CSF (Figure 7A; data not shown). Cytological examination of G-CSF-dependent colonies revealed a partial block in differentiation in the STAT-3D-transduced colonies; in one representative experiment, MSCV-transduced colonies contained 64% neutrophils and bands, 34% metamyelocytes and myelocytes, 0% promyelocytes and blasts, and 2% macrophages, while STAT-3D-transduced colonies contained 5% neutrophils and bands, 76% metamyelocytes and myelocytes, 17% promyelocytes and blasts, and 2% macrophages. Collectively, these data show that expression of dominant-negative STAT-3 in murine progenitors inhibits their G-CSF-dependent proliferation and differentiation. Discussion The G-CSFR activates STAT-3, STAT-5, and to a lesser degree STAT-1 (Avalos, 1996; Avalos et al., 1997). Previous studies of the signal transduction pathways utilized by the G-CSFR to activate STAT proteins have been

performed predominantly with immortalized cell lines. These studies suggest at least two mechanisms of STAT-3 activation by the G-CSFR (Ward et al., 1999). At low concentrations of G-CSF (1 ng/ml), STAT-3 activation is dependent upon receptor tyrosines, presumably through a direct docking mechanism with Y704 and Y744 of the G-CSFR. In contrast, at saturating concentrations of G-CSF (⬎10 ng/ml), STAT-3 activation also occurs in a tyrosine-independent fashion (Chakraborty et al., 1999; Ward et al., 1999). In the present study, we show that mutation of Y704 in the context of the d715 G-CSFR truncation mutant abrogates G-CSF-dependent STAT-3 activation in primary myeloid cells even at saturating (100 ng/ml) concentrations of G-CSF. Thus, these data provide additional evidence that Y704 of the G-CSFR is an important mediator of STAT-3 activation in vivo. Recent studies have suggested that the membraneproximal region of the G-CSFR, containing the conserved box 1 and box 2 motifs but lacking tyrosines, is sufficient to activate STAT-5 and STAT-1 (de Koning et al., 1996; Avalos et al., 1997; Dong et al., 1998; Ward et al., 1999). Consistent with these studies, STAT-1 activation by the d715F G-CSFR is normal, confirming that its

Immunity 200

activation predominantly occurs in a receptor tyrosineindependent fashion. In contrast, we show that STAT-5 activation by the d715F G-CSFR in primary myeloid cells is markedly attenuated, suggesting that in vivo, Y704 of the G-CSFR is an important mediator of STAT-5 activation. STAT proteins have been implicated in the regulation of cellular proliferation and survival. In fact, constitutive activation of STAT-5 and STAT-3 has been observed in certain human malignancies, suggesting that their inappropriate activation may contribute to neoplastic transformation (Gouilleux-Gruart et al., 1997; Bowman et al., 2000; Song et al., 2000; Ward et al., 2000). With respect to the G-CSFR, both STAT-3 and STAT-5 have been implicated in the transduction of proliferative signals. In mice lacking STAT5a and STAT5b, even though circulating numbers of neutrophils were normal, the number of colonies produced in response to G-CSF was reduced 2-fold (Teglund et al., 1998). Although the targeted disruption of the STAT-3 gene leads to embryonic lethality, mice with a conditional knockout of STAT-3 resulting in its myeloid-specific deletion are viable (Takeda et al., 1997, 1999). In these mice, granulopoiesis appeared normal, as evidenced by normal neutrophil numbers and morphology in the blood. In this elegant study, the STAT-3 gene was conditionally deleted using the Cre-lox recombination system in which the Crerecombinase was expressed under the control of the murine lysozyme M gene regulatory region. Since lysozyme M is expressed at the promyeloctye stage of granulocytic differentiation (Faust et al., 2000), and since there is a delay between gene excision and the loss of gene product, it is likely that STAT-3 function was lost very late during granulocytic differentiation. Thus, it is difficult to draw any conclusions about the role of STAT-3 in granulopoiesis from this transgenic mouse model. However, it should be noted that this study did show that the suppressive effects of IL-10 on mutant neutrophils were completely abolished, demonstrating that STAT-3 is required for normal neutrophil function. In the present study, we show that mice homozygous for the d715F G-CSFR have impaired basal granulopoiesis. They are neutropenic with levels of circulating neutrophils similar to that observed in G-CSFR-deficient mouse (Liu et al., 1996). Moreover, G-CSF-induced colony formation in vitro is markedly impaired. Together, these data suggest that the transduction of cellular proliferative signals by the d715F G-CSFR is impaired. In contrast to the markedly impaired G-CSF-induced cellular proliferation observed in vitro, the in vivo neutrophil response to G-CSF in d715F mice was not significantly different from that observed in wild-type mice, although it was dramatically less than that seen in the d715 mice. This apparent discrepancy between in vitro and in vivo G-CSF responsiveness suggests that G-CSFR-independent signals may be contributing to the stimulation of granulopoiesis in G-CSF-treated d715F mice. Interleukin-6 and granulocyte-macrophage colony stimulating factor are potential candidates to provide these signals, since both cytokines have been shown to contribute to granulopoieisis in mice lacking the G-CSFR or G-CSF, respectively (Liu et al., 1997; Seymour et al., 1997). Interestingly, these putative signals appear to be STAT-3 independent, since no evidence of STAT-3 activation

was detected in bone marrow cells isolated from d715F mice after in vivo treatment with G-CSF (Figure 2C). To assess the contribution of STAT-3 to the defect in G-CSF-induced cellular proliferation, a constitutively active form of STAT-3 (STAT-3C) was expressed in d715F myeloid cells. Expression of STAT-3C restored nearly normal G-CSF responsiveness of d715F progenitor cells, suggesting that the defect in G-CSF-induced cellular proliferation was due to a loss of STAT-3 activation. Alternatively, it is possible that the observed rescue of G-CSF responsiveness by STAT-3C may represent activation of nonphysiological signaling pathways. This possibility, although not excluded, is unlikely for the following reasons. First, although STAT-3C has been demonstrated to transform fibroblast cell lines, expression of STAT-3C in primary myeloid cells did not result in cytokine-independent growth (Bromberg et al., 1999). Second, STAT-3C expression in wild-type or d715 bone marrow cells had no effect on their growth or differentiation in response to G-CSF or IL-3. Finally, expression of a dominant-negative form of STAT-3 in wild-type progenitor cells resulted in impaired G-CSF responsiveness, providing independent evidence that STAT-3 activation is required for the transduction of normal proliferative signals by the G-CSFR. In contrast to this conclusion, two previous studies showed that expression of dominant-negative forms of STAT-3 (including STAT-3D) in myeloid cell lines resulted in enhanced G-CSF-dependent proliferation and a block in granulocytic differentiation (Shimozaki et al., 1997; de Koning et al., 2000). Although a definitive explanation for the apparent discrepancy in these results is not known, it is possible that the mechanisms regulating cellular proliferation and differentiation may differ significantly between these cell lines and primary myeloid cells. The residual neutrophils present in d715F mice are phenotypically normal, as evidenced by their normal morphology and expression of MMP-9 (data not shown). However, there is a significant accumulation of myeloid precursors in the bone marrow of these mice. Together, these observations suggested a delay, but not absolute block, in granulocytic differentiation in d715F mice. Since multiple cytokines contribute to granulopoiesis in vivo, we developed a system to study the G-CSFdependent granulocytic differentiation of murine hematopoietic progenitor cells in vitro. In cultures of d715F progenitor cells, a relative accumulation of myeloid precursors and loss of mature neutrophils, indicative of a partial block in granulocytic differentiation, was observed. Moreover, expression of STAT-3C in d715F progenitor cells partially rescued their G-CSF-dependent granulocytic differentiation. These results are consistent with at least two possibilities. The transduction of differentiation signals by the d715F G-CSFR may be impaired. Alternatively, the transduction of survival signals by the d715F G-CSFR may be impaired, leading to a selective loss of mature neutrophils. The latter results seem less likely for several reasons. First, no increase in the number of morphologically apoptotic neutrophils was observed in cultures of d715F cells. Second, the normal blood neutrophil count in mice lacking STAT-3 in their neutrophils suggests that STAT-3 expression is not required to maintain normal neutrophil survival in vivo

STAT-3 Activation by G-CSF 201

(Takeda et al., 1999). Collectively, these data suggest that, although not required, STAT-3 activation by G-CSF may contribute significantly to the regulation of granulocytic differentiation. As outlined in the introduction, the contribution of G-CSFR signals to the regulation of granulocytic differentiation is controversial. The presence of morphologically mature neutrophils in G-CSFR-deficient mice demonstrates that G-CSFR signals are not required for granulocytic differentiation. Indeed, the accumulation of myeloid precursors in the bone marrow of d715F mice is not present in G-CSFR-deficient mice (Liu et al., 1996). These observations suggest that it is not simply the loss of differentiative signals but rather an imbalance between proliferative and differentiative signals by the d715F G-CSFR that leads to the partial block in granulocytic differentiation in d715F mice. It is likely that the d715F G-CSFR mutation may alter other signaling pathways besides STAT-3 that contribute to G-CSF-dependent proliferation and differentiation. Indeed, STAT-5 activation by the d715F G-CSFR is both diminished and delayed. Although the findings in the present study suggest that STAT-3 is the principal STAT protein mediating G-CSF-induced proliferation and differentiation, these data do not exclude a minor role for STAT-5. It is possible that the failure of STAT3C expression to completely restore normal G-CSF responsiveness in d715F progenitors is due to impaired STAT-5 activation. In summary, to examine the contribution of STAT proteins to the transduction of biological signals by the G-CSFR, we generated transgenic mice containing a targeted mutation of their G-CSFR in which G-CSFdependent activation of STAT-3 is absent and STAT-5 activation markedly impaired. We provide evidence that STAT-3 activation is critical for the transduction of proliferative signals by the G-CSFR and may also contribute to the regulation of granulocytic differentiation. The STAT-3-responsive targets that mediate these biological responses are currently under investigation. Experimental Procedures Generation of d715F Mice The d715 G-CSFR mutant mice were generated in our laboratory, as previously described (McLemore et al., 1998). To generate the d715F targeting vector, an A→T substitution at nucleotide 2362 of the murine G-CSFR gene was introduced by site-directed mutagenesis into the d715 targeting vector using the following primers: forward, 5⬘-GGTTCAGGCCTTTGTGCTCCAAGGAGATC-3⬘; and reverse, 5⬘-TGGAGCACAAAGGCCTGAACCAGGGCT-3⬘. Thus, the d715F targeting vector is designed to introduce two separate mutations into the wild-type G-CSFR gene: a C→G mutation at nucleotide 2403 (also present in the d715 allele) that introduces a premature stop codon leading to the truncation of the distal 96 amino acids of the G-CSFR; and the A→T substitution at nucleotide 2362 (present only in the d715F allele) that mutates the sole remaining cytoplasmic tyrosine (Y704) to phenylalanine. RW4 embryonic stem (ES) cells (a gift from T.J. Ley, Washington University, St. Louis, MO) were transfected with the Nsi-1-linearized d715F targeting vector, and G418-resistant clones were isolated essentially as described (Hug et al., 1996). Two independent clones that had undergone homologous recombination were identified by Southern analysis, as described previously (McLemore et al., 1998). C57BL/6 blastocysts were microinjected with ES cells from each of these clones and implanted into pseudopregnant Swiss Webster foster females, as described previously (Hug et al., 1996). Only one

of the ES clones yielded germline transmission of the d715F allele. All mice were housed in a specific-pathogen free environment and examined daily by veterinary staff for signs of illness. Flow Cytometry G-CSF Binding To assess surface G-CSFR expression, peripheral blood mononuclear cells were incubated at 4⬚C for 1 hr with biotinylated G-CSF (generated as described [Liu et al., 1996], 5 ng per 106 cells) in the presence or absence of a 100-fold molar excess of nonlabeled G-CSF, followed by incubation with PE-conjugated streptavidin (Gibco BRL, Rockville MD). The neutrophil population was gated based on forward- and side-scatter characteristics. Transduction Efficiency Nonadherent cells were harvested from the retroviral-producing monolayer by gentle tituration and incubated with a cocktail of phycoerythrin (PE)-conjugated rat anti-mouse lineage markers, including Gr-1 (myeloid cells), B220 (B lymphocytes), and CD3 (T lymphocytes). After incubation at 4⬚C for 30 min cells were washed; they were then analyzed using a FACScan flow cytometer and CellQuest version 1.2.2 software (Becton-Dickinson, Mansfield MA). Transduction efficiency was defined as the percentage of lineage-negative cells that expressed EGFP. All antibodies were purchased from PharMingen (San Diego, CA). Cell Sorting Nonadherent cells harvested from the retroviral-producing monolayer were incubated with the same cocktail of lineage markers, and lineage-negative EGFP-positive cells were sorted using a Coulter Elite ESP cytometer (Coulter, Hialeah, FL). G-CSF Binding Experiments Binding experiments were performed as previously described with minor modifications (Semerad et al., 1999). In brief, 5.0 ⫻ 106 bone marrow cells were incubated in triplicate with eight different concentrations of 125I-[Tyr1, Tyr3]rhG-CSF for 3 hr at 15⬚C in the presence or absence of at least a 50-fold excess of unlabeled rhG-CSF. Cell bound versus free 125I-[Tyr1, Tyr3] rhG-CSF was measured using a LKB-Wallac CliniGamma 1272–004 ␥ counter (Wallac Oy, Finland). Dissociation constants (Kd) were calculated from two independent experiments. 125I-[Tyr1, Tyr3] rhG-CSF was kindly provided by Pharmacia (Peapack, NJ) Electrophoretic Mobility Shift Assay Bone marrow cells (1 ⫻ 107) were resuspended in Opti-MEM 1 (Gibco BRL, Grand Island, NY) media with 1% fetal calf serum and stimulated at 37⬚C with buffer alone or with the indicated cytokines for the indicated period of time. Nuclear extracts were prepared as previously described (Hermans et al., 1999). One microgram of nuclear extract was incubated for 20 min at room temperature with 25,000 CPM of P32-labeled double-stranded oligonucleotide probe and 1 ␮g of poly (dI:dC) in 20 ␮l of binding buffer (10 mM Tris [pH 7.5], 100 mM KCL, 5 nM MgCL, 10% Glycerol, and 1 mM DTT). Complexes were resolved on a 5% nondenaturing polyacrylamide gel. The gel was subsequently dried and analyzed by autoradiography. The following double-stranded oligonucleotides were used: m67 (5⬘-CATTTCCCGTAAATCAT-3⬘), a high-affinity mutant of the sis-inducible element of the human c-fos gene; and ␤-casein (5⬘-GATTTCTAGGAATTCAATCC-3⬘), the STAT-5 binding site of the bovine ␤-casein promoter. For supershift analysis, nuclear extracts were preincubated for 1 hr at 4⬚C with specific antibodies to STAT-3 (#ST3-5G7, ZYMED, San Francisco, CA) or STAT-5 (#sc-835, Santa Cruz Biotechnology, Santa Cruz, CA) prior to the addition of radiolabeled oligonucleotide. For STAT-3D- and MSCV-transduced cells, lineage-negative EGFP-positive cells were sorted by flow cytometry and expanded in Myelocult 5300 (Stem Cell Technologies) supplemented with IL-3 (10 ng/ml) and SCF (100 ng/ml). After 5 to 6 days in culture, cells were washed extensively and then incubated for 4 hr in Opti-MEM with 1% FCS prior to cytokine stimulation. Peripheral Blood and Bone Marrow Analysis Blood was obtained by retro-orbital venous plexus sampling in polypropylene tubes containing EDTA. Complete blood counts were determined using a Hemavet automated cell counter (CDC Technol-

Immunity 202

ogies, Inc., Oxford, CT). Bone marrow was harvested by flushing both femoral bones with ␣-minimum essential medium (␣-MEM) containing 10% fetal bovine serum. Manual leukocyte differentials were performed on Wright-stained blood smears or cytospin preparations of bone marrow mononuclear cells.

Hematopoietic Progenitor Cell Assays A total of 5–10 ⫻ 104 bone marrow mononuclear cells were plated in 2.5 ml of methylcellulose media (MethoCult 3230; Stem Cell Technologies, Vancouver, BC, Canada) supplemented with human G-CSF (100 ng/ml, Amgen, Thousand Oaks, CA) or with murine IL-3 (10 ng/ ml, R&D Systems, Minneapolis, MN) and placed in a humidified chamber with 5% CO2. Colonies containing at least 50 cells were scored on day 7 of culture. For retrovirally transduced bone marrow cells, a total of 4–10 ⫻ 103 mononuclear cells were plated in 2.5 ml of methylcellulose media.

G-CSF Administration to Mice Recombinant human G-CSF diluted in PBS with 1% low endotoxin bovine serum albumin (Sigma, St. Louis, MO) was administered by daily subcutaneous injection at a dose of 10 ␮g/kg per day for 7 days. Peripheral blood was obtained before the first G-CSF dose and 4 hr after the final injection on day 7.

MMP-9 Immunocytochemistry MMP-9 immunohistochemistry was performed on cytospin preparations of peripheral blood mononuclear cells isolated from G-CSFtreated mice, as described previously (Semerad et al., 1999). The murine MMP-9 antiserum was provided by J. Michael Shipley and Robert M. Senior (Washington University, St. Louis, MO).

In Vitro Suspension Culture System 10,000 EGFP-positive lineage-negative cells were suspended in Myelocult 5300 containing 100 ng/ml SCF and 100 ng/ ml of G-CSF and incubated for 7 days at 37⬚C in a 5% CO2 humidified chamber. Statistical Analysis Data are presented as mean ⫾ standard deviation (SD). Statistical significance was assessed by Student’s t test. Acknowledgments The authors thank Tim Graubert, Jessica Pollock, Tim Ley, and Christine Pham for their thoughtful discussions and advice. This work was supported by the Edward Mallinckrodt, Jr., Foundation (DCL), by a grant from the National Institutes of Health NHLBI (R01 HL60772-01A1 [D. C. L.]), and by a training grant from the National Institutes of Health NHLBI (T32 HL 07088-23 [M. L. M.]). Received September 18, 2000; revised January 18, 2001. References Avalos, B.R. (1996). Molecular analysis of the granulocyte colonystimulating factor receptor. Blood 88, 761–777. Avalos, B.R., Parker, J.M., Ware, D.A., Hunter, M.G., Sibert, K.A., and Druker, B.J. (1997). Dissociation of the Jak kinase pathway from G-CSF receptor signaling in neutrophils. Exp. Hematol. 25, 160–168. Bober, L.A., Grace, M.J., Pugliese-Sivo, C., Rojas-Triana, A., Waters, T., Sullivan, L.M., and Narula, S.K. (1995). The effect of GM-CSF and G-CSF on human neutrophil function. Immunopharmacology 29, 111–119. Borregaard, N., and Cowland, J.B. (1997). Granules of the human neutrophilic polymorphonuclear leukocyte. Blood 89, 3503–3521.

Generation of STAT-3C and STAT-3D Retrovirus and Transduction of Bone Marrow Cells Generation of STAT-3C The STAT-3C molecule contains cysteine substitutions for residues A661 and N663 of the murine STAT-3 protein and reproduces the mutation described by Bromberg and colleagues (Bromberg et al., 1999). STAT-3C cDNA was generated by site-directed mutagenesis using the following primers: (forward, 5⬘-TGGATTGTACCTGCA TCCTGG-3⬘); and reverse, 5⬘-CCAGGATGCAGGTACAATCCA-3⬘). The STAT-3D cDNA was a kind gift of Drs. Koischi Nakajima and Toshio Hirano, Osaka University Medical School, Osaka, Japan. The STAT-3C and STAT-3D cDNAs were cloned into the MSCV retroviral vector containing an IRES motif linked to the EGFP gene (Persons et al., 1998). Production of High-Titer Ecotropic Retrovirus The MSCV, STAT-3C MSCV, and STAT-3D MSCV plasmids were transfected into the Phoenix amphotropic packing-cell line (ATCC, Rockville MD) by calcium phosphate precipitation, as described previously (Wigler et al., 1979; Persons et al., 1997). The supernatant from these cultures was harvested 48 and 72 hr after transfection and used to infect the GP⫹E86 ecotropic packaging cell line (Ref, Genetix Pharmaceuticals, Cambridge, MA) in the presence of 4 ␮g/ ml of polybrene. After expansion of the GP⫹E86 cells for 4 days, EGFP-positive cells were sorted by flow cytometry and further expanded at 37⬚C in HXM media (DMEM supplemented with 10% FCS, 15 ␮g/ml hypoxanthine, 250 ng/ml xanthine, and 25 ␮g/ml of mycophenolic acid). Viral production was tested on NIH 3T3 cells; titers of 0.5–1.0 ⫻ 106 infectious particles per ml were obtained for each construct (data not shown). Transduction of Bone Marrow Cells Bone marrow cells were harvested from wild-type or d715F mice 48 hr after treatment with 150 mg/kg of 5-fluoruracil (5-FU). Cells were stimulated for 48 hr with IL-3 (10 ng/ml), thrombopoietin (10 ng/ml), flt-3 ligand (50 ng/ml), and kit ligand (100 ng/ml) in DMEM supplemented with 15% FCS (all recombinant murine cytokines were purchased from R&D Systems). Bone marrow cells were subsequently cocultured for 48 hr with irradiated (1500 cGy) retroviralproducing cells using the above culture media supplemented with 4 ␮g/ml of polybrene.

Bowman, T., Garcia, R., Turkson, J., and Jove, R. (2000). STATs in oncogenesis. Oncogene 19, 2474–2488. Bromberg, J.F., Wrzeszczynska, M.H., Devgan, G., Zhao, Y., Pestell, R.G., Albanese, C., and Darnell, J.E., Jr. (1999). Stat3 as an oncogene. Cell 98, 295–303. Carulli, G. (1997). Effects of recombinant human granulocyte colonystimulating factor administration on neutrophil phenotype and functions. Haematologica 82, 606–616. Catlett-Falcone, R., Landowski, T.H., Oshiro, M.M., Turkson, J., Levitzki, A., Savino, R., Ciliberto, G., Moscinski, L., Fernandez-Luna, J.L., Nunez, G., et al. (1999). Constitutive activation of Stat3 signaling confers resistance to apoptosis in human U266 myeloma cells. Immunity 10, 105–115. Chakraborty, A., and Tweardy, D.J. (1998). Stat3 and G-CSF-induced myeloid differentiation. Leuk. Lymphoma 30, 433–442. Chakraborty, A., White, S.M., Schaefer, T.S., Ball, E.D., Dyer, K.F., and Tweardy, D.J. (1996). Granulocyte colony-stimulating factor activation of Stat3 alpha and Stat3 beta in immature normal and leukemic human myeloid cells. Blood 88, 2442–2449. Chakraborty, A., Dyer, K.F., Cascio, M., Mietzner, T.A., and Tweardy, D.J. (1999). Identification of a novel Stat3 recruitment and activation motif within the granulocyte colony-stimulating factor receptor. Blood 93, 15–24. de Haas, M., Kerst, J.M., van der Schoot, C.E., Calafat, J., Hack, C.E., Nuijens, J.H., Roos, D., van Oers, R.H., and von dem Borne, A.E. (1994). Granulocyte colony-stimulating factor administration to healthy volunteers: analysis of the immediate activating effects on circulating neutrophils. Blood 84, 3885–3894. de Koning, J.P., Dong, F., Smith, L., Schelen, A.M., Barge, R.M., van der Plas, D.C., Hoefsloot, L.H., Lowenberg, B., and Touw, I.P. (1996). The membrane-distal cytoplasmic region of human granulocyte colony-stimulating factor receptor is required for STAT3 but not STAT1 homodimer formation. Blood 87, 1335–1342. de Koning, J., Soede-Bobok, A., Ward, A., Schelen, A., Antonissen, C., van Leeuwen, D., Lowenberg, B., and Touw, I. (2000). STAT3mediated differentiation and survival of myeloid cells in response to

STAT-3 Activation by G-CSF 203

granulocyte colony-stimulating factor: role for the cyclin-dependent kinase inhibitor p27(Kip 1). Oncogene 19, 3290–3298.

Metcalf, D. (1984). The Hematopoitic Colony-Stimulating Factors (The Netherlands: Elsevier).

Dong, F., van Buitenen, C., Pouwels, K., Hoefsloot, L.H., Lowenberg, B., and Touw, I.P. (1993). Distinct cytoplasmic regions of the human granulocyte colony-stimulating factor receptor involved in induction of proliferation and maturation. Mol. Cell. Biol. 13, 7774–7781.

Nakajima, K., Yamanaka, Y., Nakae, K., Kojima, H., Ichiba, M., Kiuchi, N., Kitaoka, T., Fukada, T., Hibi, M., and Hirano, T. (1996). A central role for Stat3 in IL-6-induced regulation of growth and differentiation in M1 leukemia cells. EMBO J. 15, 3651–3658.

Dong, F., Liu, X., de Koning, J.P., Touw, I.P., Henninghausen, L., Larner, A., and Grimley, P.M. (1998). Stimulation of Stat5 by granulocyte colony-stimulating factor (G-CSF) is modulated by two distinct cytoplasmic regions of the G-CSF receptor. J. Immunol. 161, 6503– 6509.

Nicola, N.A., and Metcalf, D. (1985). Binding of 125I-labeled granulocyte colony-stimulating factor to normal murine hemopoietic cells. J. Cell. Physiol. 124, 313–321.

Faust, N., Varas, F., Kelly, L., Heck, S., and Graf, T. (2000). Insertion of enhanced green flourescent protein into the lysozyme gene creates mice with green flourescent granulocytes and macrophages. Blood 96, 719–726.

Persons, D.A., Allay, J.A., Allay, E.R., Smeyne, R.J., Ashmun, R.A., Sorrentino, B.P., and Nienhuis, A.W. (1997). Retroviral-mediated transfer of the green fluorescent protein gene into murine hematopoietic cells facilitates scoring and selection of transduced progenitors in vitro and identification of genetically modified cells in vivo. Blood 90, 1777–1786.

Fukunaga, R., Ishizaka-Ikeda, E., and Nagata, S. (1993). Growth and differentiation signals mediated by different regions in the cytoplasmic domain of granulocyte colony-stimulating factor receptor. Cell 74, 1079–1087.

Persons, D.A., Allay, J.A., Riberdy, J.M., Wersto, R.P., Donahue, R.E., Sorrentino, B.P., and Nienhuis, A.W. (1998). Use of the green fluorescent protein as a marker to identify and track genetically modified hematopoietic cells. Nat. Med. 4, 1201–1205.

Gouilleux-Gruart, V., Debierre-Grockiego, F., Gouilleux, F., Capiod, J.C., Claisse, J.F., Delobel, J., and Prin, L. (1997). Activated Stat related transcription factors in acute leukemia. Leuk. Lymphoma 28, 83–88.

Roberts, A.W., Foote, S., Alexander, W.S., Scott, C., Robb, L., and Metcalf, D. (1997). Genetic influences determining progenitor cell mobilization and leukocytosis induced by granulocyte colony-stimulating factor. Blood 89, 2736–2744.

Hermans, M.H., Ward, A.C., Antonissen, C., Karis, A., Lowenberg, B., and Touw, I.P. (1998). Perturbed granulopoiesis in mice with a targeted mutation in the granulocyte colony-stimulating factor receptor gene associated with severe chronic neutropenia. Blood 92, 32–39.

Semerad, C.L., Poursine-Laurent, J., Liu, F. and Link, D.C. (1999). A role for G-CSF receptor signaling in the regulation of hematopoietic cell function but not lineage commitment or differentiation. Immunity 11, 153–161.

Hermans, M.H., Antonissen, C., Ward, A.C., Mayen, A.E., Ploemacher, R.E., and Touw, I.P. (1999). Sustained receptor activation and hyperproliferation in response to granulocyte colony-stimulating factor (G-CSF) in mice with a severe congenital neutropenia/ acute myeloid leukemia-derived mutation in the G- CSF receptor gene. J. Exp. Med. 189, 683–692.

Seymour, J.F., Lieschke, G.J., Grail, D., Quilici, C., Hodgson, G., and Dunn, A.R. (1997). Mice lacking both granulocyte colony-stimulating factor (CSF) and granulocyte-macrophage CSF have impaired reproductive capacity, perturbed neonatal granulopoiesis, lung disease, amyloidosis, and reduced long-term survival. Blood 90, 3037– 3049.

Hoglund, M., Hakansson, L., and Venge, P. (1997). Effects of in vivo administration of G-CSF on neutrophil functions in healthy volunteers. Eur. J. Haematol. 58, 195–202.

Shimozaki, K., Nakajima, K., Hirano, T., and Nagata, S. (1997). Involvement of STAT3 in the granulocyte colony-stimulating factorinduced differentiation of myeloid cells. J. Biol. Chem. 272, 25184– 25189.

Hug, B.A., Wesselschmidt, R.L., Fiering, S., Bender, M.A., Epner, E., Groudine, M., and Ley, T.J. (1996). Analysis of mice containing a targeted deletion of ␤-globin locus control region 5⬘ hypersensitive site 3. Mol. Cell. Biol. 16, 2906–2912.

Smithgall, T.E., Briggs, S.D., Schreiner, S., Lerner, E.C., Cheng, H., and Wilson, M.B. (2000). Control of myeloid differentiation and survival by Stats. Oncogene 19, 2612–2618.

Lieschke, G.J., Grail, D., Hodgson, G., Metcalf, D., Stanley, E., Cheers, C., Fowler, K.J., Basu, S., Zhan, Y.F., and Dunn, A.R. (1994). Mice lacking granulocyte colony-stimulating factor have chronic neutropenia, granulocyte and macrophage progenitor cell deficiency, and impaired neutrophil mobilization. Blood 84, 1737–1746. Liu, F., Wu, H.Y., Wesselschmidt, R., Kornaga, T., and Link, D.C. (1996). Impaired production and increased apoptosis of neutrophils in granulocyte colony-stimulating factor receptor-deficient mice. Immunity 5, 491–501. Liu, F., Poursine-Laurent, J., Wu, H.Y., and Link, D.C. (1997). Interleukin-6 and the granulocyte colony-stimulating factor receptor are major independent regulators of granulopoiesis in vivo but are not required for lineage commitment or terminal differentiation. Blood 90, 2583–2590. Lord, B.I., Bronchud, M.H., Owens, S., Chang, J., Howell, A., Souza, L., and Dexter, T.M. (1989). The kinetics of human granulopoiesis following treatment with granulocyte colony-stimulating factor in vivo. Proc. Natl. Acad. Sci. USA 86, 9499–9503.

Song, J.I., Grandis, J.R., Bowman, T., Garcia, R., Turkson, J., and Jove, R. (2000). STAT signaling in head and neck cancer. STATs in oncogenesis. Oncogene 19, 2489–2495. Takeda, K., Noguchi, K., Shi, W., Tanaka, T., Matsumoto, M., Yoshida, N., Kishimoto, T., and Akira, S. (1997). Targeted disruption of the mouse Stat3 gene leads to early embryonic lethality. Proc. Natl. Acad. Sci. USA 94, 3801–3804. Takeda, K., Clausen, B.E., Kaisho, T., Tsujimura, T., Terada, N., Forster, I., and Akira, S. (1999). Enhanced Th1 activity and development of chronic enterocolitis in mice devoid of Stat3 in macrophages and neutrophils. Immunity 10, 39–49. Teglund, S., McKay, C., Schuetz, E., van Deursen, J.M., Stravopodis, D., Wang, D., Brown, M., Bodner, S., Grosveld, G., and Ihle, J.N. (1998). Stat5a and Stat5b proteins have essential and nonessential, or redundant, roles in cytokine responses. Cell 93, 841–850. Tian, S.-S., Lamb, P., Seidel, H.M., Stein, R.B., and Rosen, J. (1994). Rapid activation of the STAT3 transcription factor by granulocyte colony-stimulating factor. Blood 84, 1760–1764.

Lutz, P.G., Moog-Lutz, C., Coumau-Gatbois, E., Kobari, L., Di Gioia, Y., and Cayre, Y.E. (2000). Myeloblastin is a granulocyte colonystimulating factor-responsive gene conferring factor-independent growth to hematopoietic cells. Proc. Natl. Acad. Sci. USA 97, 1601– 1606.

Ward, A.C., Hermans, M.H., Smith, L., van Aesch, Y.M., Schelen, A.M., Antonissen, C., and Touw, I.P. (1999). Tyrosine-dependent and -independent mechanisms of STAT3 activation by the human granulocyte colony-stimulating factor (G-CSF) receptor are differentially utilized depending on G-CSF concentration. Blood 93, 113–124.

McLemore, M.L., Poursine-Laurent, J., and Link, D.C. (1998). Increased granulocyte colony-stimulating factor responsiveness but normal resting granulopoiesis in mice carrying a targeted granulocyte colony-stimulating factor receptor mutation derived from a patient with severe congenital neutropenia. J. Clin. Invest. 102, 483–492.

Ward, A.C., Touw, I., Yoshimura, A., Chakraborty, A., and Tweardy, D.J. (2000). The Jak-Stat pathway in normal and perturbed hematopoiesis. Stat3 and G-CSF-induced myeloid differentiation. Blood 95, 19–29. Watanabe, M., Fukamachi, H., Uzumaki, H., Kabaya, K., Tsumura, H., Ishikawa, M., Matsuki, S., and Kusaka, M. (1991). Mutant protein

Immunity 204

of recombinant human granulocyte colony-stimulating factor for receptor binding assay. Anal. Biochem. 195, 38–44. Welte, K., Gabrilove, J., Bronchud, M.H., Platzer, E., and Morstyn, G. (1996). Filgrastim (r-metHuG-CSF): the first 10 years. Blood 88, 1907–1929. Wigler, M., Pellicer, A., Silverstein, S., Axel, R., Urlaub, G., and Chasin, L. (1979). DNA-mediated transfer of the adenine phosphoribosyltransferase locus into mammalian cells. Proc. Natl. Acad. Sci. USA 76, 1373–1376. Ziegler, S.F., Bird, T.A., Morella, K.K., Mosley, B., Gearing, D.P., and Baumann, H. (1993). Distinct regions of the human granulocytecolony-stimulating factor receptor cytoplasmic domain are required for proliferation and gene induction. Mol. Cell. Biol. 13, 2384–2390.