Granulocytic di erentiation of myeloid progenitor cells by p130 ... | Nature

2 downloads 113 Views 627KB Size Report
The retinoblastoma protein (pRB) and the related pocket proteins, p107 and p130, play crucial roles in mammalian cell cycle control. Recent studies indicate that ...
Oncogene (1999) 18, 6209 ± 6221 ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $15.00 http://www.stockton-press.co.uk/onc

Granulocytic di€erentiation of myeloid progenitor cells by p130, the retinoblastoma tumor suppressor homologue Akio Mori1, Hideaki Higashi1, Yutaka Hoshikawa1, Masahiro Imamura2, Masahiro Asaka3 and Masanori Hatakeyama*,1 1

Department of Viral Oncology, Cancer Institute, Japanese Foundation for Cancer Research, 1-37-1 Kami-Ikebukuro, Toshima-Ku, Tokyo 170-8455, Japan; 2Department of Hematology and Oncology, Hokkaido University School of Medicine, Sapporo 060-8638, Japan; 3Third Department of Internal Medicine, Hokkaido University School of Medicine, Sapporo 060-8638, Japan

The retinoblastoma protein (pRB) and the related pocket proteins, p107 and p130, play crucial roles in mammalian cell cycle control. Recent studies indicate that these pocket proteins are also involved in cellular di€erentiation processes. We demonstrate in this work that the pRB-related p130 selectively accumulates during the in vitro di€erentiation of the myeloid progenitor cell, 32Dcl3, into granulocyte in response to granulocytecolony stimulating factor (G-CSF). This G-CSF-dependent granulocytic di€erentiation is blocked by the adenovirus E1A oncoprotein, which binds to and inactivates the pRB family of pocket proteins including p130. Furthermore, enforced overexpression of p130 but not pRB inhibits the myeloid cell proliferation that is concomitantly associated with granulocytic di€erentiation morphologically characterized by nuclear segmentation. However, simple G1-cell cycle arrest induced by cytokine deprivation or ectopic overexpression of the p27 cyclin-dependent kinase inhibitor, or inhibition of E2F activities by dominant negative DP-1 is not sucient to trigger granulocytic di€erentiation. The di€erentiationpromoting activity of p130 in myeloid cells requires both the pocket domain and the spacer domain. Our results indicate that the pRB-related p130 plays a critical role in myeloid cell di€erentiation and suggest that coupling of cell cycle exit with the cellular di€erentiation program may be speci®cally achieved by p130. Keywords: p130; pRB; granulocytic di€erentiation; 32Dcl3 cell; E1A

Introduction The retinoblastoma tumor suppressor protein (pRB) and the structurally related p107 and p130 proteins are considered to play crucial roles in mitotic cell cycle control (Weinberg, 1995; Sherr, 1996; Beijersbergen and Bernards, 1996). This pocket protein-dependent cell cycle control is primarily mediated by their interaction with a class of cell cycle-regulating transcriptional factors collectively denoted E2F. The pRB family members bind and convert E2F from a transcriptional activator to a transcriptional repressor, thereby inhibiting promoters whose activation is required for cell cycle progression from G1 to S phase

*Correspondence: M Hatakeyama Received 6 April 1999; revised 16 June 1999; accepted 23 June 1999

(Weintraub et al., 1992; Nevins et al., 1997; Zhang et al., 1999). Progressive phosphorylation of the pRB family of pocket proteins by G1 cyclin ± cyclin dependent kinase (Cdk) complexes (Hatakeyama et al., 1994; Matsushime et al., 1994; Meyerson and Harlow, 1994; Resnitzky and Reed, 1995; Baldi et al., 1995; Beijersbergen et al., 1995; Mayol et al., 1995; Xiao et al., 1996; Knudsen and Wang, 1996; Zarkowska and Mittnacht, 1997; Lundberg and Weinberg, 1998) is considered to abolish their ability to form complexes with E2F. Furthermore, disruption of pocket protein complexes with E2F promotes a series of transcriptional events leading to G1 to S phase transition and subsequent cell division (Nevins et al., 1997). Among the pRB family proteins, p107 and p130 are structurally closer to each other than to pRB and share unique properties allowing interaction with cyclins and Cdks (Ewen et al., 1992; Faha et al., 1992; Lees et al., 1992; Cobrinik et al., 1993; Zhu et al., 1995; Smith and Nevins, 1995). Through this interaction, p107 and p130 are capable of inhibiting the kinase activity of cyclin ± Cdk complexes (Zhu et al., 1995; Woo et al., 1997; Castano et al., 1998). In addition, p107 and p130 selectively bind E2F-4 and E2F-5 (Beijersbergen et al., 1994; Hijmans et al., 1995; Sardet et al., 1995; Vairo et al., 1995), whereas pRB preferentially binds E2F-1, E2F-2 and E2F-3 (Ikeda et al., 1996; Moberg et al., 1996). Together with the di€erential regulation of protein levels during cell cycle (Garriga et al., 1998), each member of the pRB family appears to perform shared as well as unique cell cycle-regulatory roles within a single cell. Furthermore, whereas growth of certain cells is strongly inhibited by pRB (Tam et al., 1994; Koh et al., 1995; Lukas et al., 1995a,b; Medema et al., 1995), proliferation of hematopoietic cells is e€ectively blocked by p130 but not by pRB (Hoshikawa et al., 1998a). These observations suggest that a di€erent member of the pRB family may be employed as a key cell cycle regulator among di€erent cell types. In addition to cell cycle control, it has been suggested that pRB plays a role during cellular di€erentiation (Yee et al., 1998). Genetically engineered mice lacking pRB exhibit defects in erythropoiesis and extensive cell death in the central nervous system (Jacks et al., 1992; Lee et al., 1992; Clarke et al., 1992). In in vitro di€erentiation systems, pRB is required for adipocyte, neuron or muscle cell differentiation. Furthermore, pRB binds a transcription factor termed CCAAT enhancer binding protein (C/ EBP) and positively regulates the transcriptional

p130-mediated granulocytic differentiation A Mori et al

6210

activity that promotes adipocyte di€erentiation (Chen et al., 1996). pRB also functionally interacts with MyoD, a muscle speci®c transcription factor, and potentiates its activity for muscle cell di€erentiation (Gu et al., 1993; Zacksenhaus et al., 1996). More recently, a sequence-speci®c HMG transcription factor, HBP-1, was reported to interact with pRB and p130 but not with p107 (Tevosian et al., 1997). Since HMG1 is upregulated during muscle di€erentiation and is capable of repressing the N-myc promoter, it may act as a target of the pocket proteins in the muscle di€erentiation process. A distinct role of pRB in cell di€erentiation is further suggested by a recent observation that certain pRB mutants are unable to bind E2Fs, yet they retain the ability to promote osteoblastic and myogenic di€erentiation in vitro (Sellers et al., 1998). In contrast to pRB, little is known about the role of p130 and p107 in cellular di€erentiation. Disruption of either p107 or p130 in mice of the C57BL/6J background does not produce any detectable developmental defects. However, mice lacking both p130 and p107 in the same genetic background show defects in endochondral bone formation and die shortly after birth (Cobrinik et al., 1996). This indicates a unique role for p107 and p130 in bone development which cannot be compensated by pRB. However, the e€ects of p107 and/or p130 in mouse development appear to be more complicated, because mice nullizygous for p130 gene in an enriched Balb/cJ genetic background display embryonic lethality (LeCouter et al., 1998a) and those lacking p107 display impaired growth in the Balb/cJ background (LeCouter et al., 1998b). These observations indicate the existence of mouse strainspeci®c modi®er genes that have epistatic relationships with p130 and p107. Together with the functional redundancy among pRB family proteins, existence of such modi®ers makes the knock-out phenotype of pRB family proteins in mice dicult to interpret. In this work, we addressed the role of the pRB family of pocket proteins in the growth and di€erentiation of hematopoietic cells by ectopically expressing these proteins in 32Dcl3 myeloid cells that di€erentiate into granulocytes upon granulocyte-colony stimulating factor (G-CSF) treatment. We demonstrate that the pRB-related p130 protein, but not pRB, is capable of inhibiting proliferation and promoting granulocytic di€erentiation in the myeloid progenitor cell. Our results suggest that a particular member of pRB family protein, p130, participates in dual functions, one linked to cell cycle control and the other to the cellular di€erentiation program, thereby coordinating these processes.

Results Expression of the pRB family of pocket proteins during granulocytic di€erentiation 32Dcl3 cell is a myeloid cell line derived from normal mouse bone marrow cells and is dependent on interleukin 3 (IL-3) for growth and survival (Greenberger et al., 1983). As previously reported, 32Dcl3 cells cultured in medium containing G-CSF in the absence of IL-3 cease proliferation and undergo

granulocytic di€erentiation, which is morphologically characterized by nuclear segmentation (Valtieri et al., 1987) (Figure 1a,b). It is generally observed that cell cycle withdrawal precedes terminal di€erentiation. Since the pRB family of pocket proteins are considered to play crucial roles in exiting cell cycle, we wondered if they are involved in the processes of hematopoietic di€erentiation at least by being negative regulators of cell cycle. To this end, we ®rst examined levels of the pRB family proteins during granulocytic di€erentiation of the 32Dcl3 cells. Immunoblot analyses revealed that hyperphosphorylated forms of pRB and p107 are abundantly present, whereas p130 was virtually undetectable in the exponentially growing 32Dcl3 cells in the presence of IL-3 (Figure 1c). When 32Dcl3 cells were switched to medium containing G-CSF instead of IL-3, proliferation ceased (Figure 1a) and p130 accumulated dramatically by day 3 (Figure 1c). Conversely, expression of p107 and pRB was strongly downregulated in cells treated with G-CSF. The observation indicates that, if any of pRB family members has a role in the granulocytic di€erentiation, it may be p130. Alternatively, decreased levels of pRB and/or p107 might play a role in inducing granulocytic differentiation. E€ect of adenovirus E1A oncoprotein on G-CSF-induced granulocytic di€erentiation Selective accumulation of p130 during granulocytic di€erentiation suggests that p130 might have an active role in myeloid di€erentiation. However, since p130 levels generally rise when cycling cells are forced to exit the cell cycle (Mayol et al., 1995, 1996), it is equally possible that the elevated p130 is simply due to cell cycle exit and does not play any substantial role in di€erentiation. To address this, the e€ect of the adenovirus E1A oncoprotein on G-CSF-dependent granulocytic di€erentiation was next examined. The E1A gene produces two forms of mRNA, 12S and 13S. The 12S and 13S mRNA products have two common regions termed CR1 and CR2, while a third region, CR3, is unique to E1A 13S product (Velcich and Zi€, 1988; Whyte et al., 1989). Transformation and immortalization functions of E1A have been localized to CR1 and CR2 (Moran et al., 1986; Lillie et al., 1987), whereas CR3 is known to act as a transcriptional activator of viral and cellular promoters (Green et al., 1983; Kao and Nevins, 1983). The E1A 12S product is capable of physically interacting with and functionally neutralizing all of the pRB family proteins but lacks the transcriptional activation functions provided by the CR3 domain (Egan et al., 1988; Jelsma et al., 1989; Whyte et al., 1989; Cobrinik et al., 1993). Accordingly, we generated stable transfectants that express the E1A 12S product by transfecting pOPTET-E1A12S into 32D-B5 cell, a 32Dcl3-derived transfectant clone that ectopically co-express the tetracycline-repressible transactivator (tTA) and the E. coli lac repressor as described (Figure 2a) (Hoshikawa et al., 1998b). The E1A expressing cells were morphologically indistinguishable from parental cells and were strictly dependent on IL-3 for their growth and survival. Parental 32Dcl3 cells treated with G-CSF in the absence of IL-3 for 5 days exhibited

p130-mediated granulocytic differentiation A Mori et al

granulocytic di€erentiation as characterized by nuclear segmentation. Like the parental cells, the E1A expressing cells ceased proliferation in the presence of G-CSF. However, they were morphologically of a myeloblastic phenotype and failed to undergo differentiation (Figure 2b). These results indicate that the pRB family proteins are actively involved in granulocytic di€erentiation and support for the notion that p130 plays an important role in myeloid di€erentiation. E€ect of ectopically expressed pRB-family proteins on the growth and di€erentiation of 32Dcl3 myeloid cells To directly address the role of the pRB family proteins, particularly p130, in myeloid di€erentiation, we next generated 32Dcl3-derived stable transfectants that ectopically overexpress p130 (clones 130-1 and 130-3) or a phosphorylation-resistant pRB,

pRBDp34HA (clones H1-12 and H3-8) in an inducible manner. As demonstrated in Figure 3a,b, overexpression of p130 substantially inhibited IL-3dependent growth of 32Dcl3 cells. Furthermore, a certain fraction of the p130 expressing cells underwent morphological di€erentiation into granulocytes in IL-3-containing medium despite the absence of GCSF (Figure 3e, Table 1). In contrast, ectopic overexpression of pRBDp34HA in 32Dcl3 cell neither inhibited proliferation nor promoted differentiation (Figure 3c ± e). The results indicate that 32Dcl3 cells are resistant to pRB but are sensitive to p130 for their IL-3-dependent growth as is the case of another IL-3-dependent hematopoietic cell line, BaF3 (Hoshikawa et al., 1998a). They further suggest that elevated levels of p130 per se are sucient to trigger morphological di€erentiation in 32Dcl3 myeloid progenitor cells.

Figure 1 Expression of pRB family pocket proteins during granulocyte di€erentiation of 32Dcl3 cells. (a) E€ect of G-CSF on 32Dcl3 cell growth. 32Dcl3 cells (26105 /ml) were cultured in medium containing IL-3 (20% WEHI) (*) or G-CSF (*). Because doubling time of the 32Dcl3 cells in the IL-3 medium was approximately 15 ± 18 h, cell concentrations were adjusted to 26105 /ml at 24 h intervals. Viable cell number was counted every 24 h by Trypan-blue dye exclusion methods. Cell viabilities after 5 days culture were approximately 100% for IL-3 medium and 90% for G-CSF medium, respectively. (b) Granulocytic di€erentiation of 32Dcl3 cells in response to G-CSF. 32Dcl3 cells cultured in IL-3 medium (left) or treated with 2.5 ng/ml recombinant G-CSF in the absence of IL-3 for 5 days (right) were ®xed and subjected to May-Grunwald-Giemsa staining. Dead cells were removed by density centrifugation. (c) Detection of pocket proteins in 32Dcl3 cells. Cell lysates (50 mg each) were prepared from cells cultured in IL-3 medium or G-CSF medium for 3 days, resolved in 7.5% SDS-polyacrylamide gel (SDS ± PAGE), and immunoblotted with antipRB, anti-p107 and anti-p130, respectively. Positions for each pocket protein were indicated

6211

p130-mediated granulocytic differentiation A Mori et al

6212

p130 was capable of inhibiting 32Dcl3 cell growth whereas pRB failed to do so. This raises a possibility that p130 promotes cell di€erentiation by simply inhibiting the cell cycle. If this is the case, then inhibition of cell growth by mechanisms other than p130 overexpression should also induce granulocytic di€erentiation. To address this, we arrested 32Dcl3 cells in the G1 phase by cytokine starvation. Whereas a high-dose of cytokine supports cell proliferation, a low-dose of cytokine acts as a survival factor (or anti-apoptotic factor) rather than a growth factor in many cytokine-dependent cells. Indeed, 32Dcl3 cells cultured in medium containing extremely low levels of IL-3 (15 pg/ml IL-3; denote here as `cytokinestarved condition') displayed prolonged survival without proliferation for at least 5 days, arresting cells in the G0/G1 phase of the cell cycle (Figure 4a,b). Consistent with the cell cycle analysis by FACS, DNA synthesis of these cells as examined by measuring [3H]thymidine incorporation into DNA was strongly reduced in this cytokine-starved condition when compared with that of the cycling cells (Figure 4c). As demonstrated in Figure 4d, parental 32Dcl3 or 32D-B5 cells that were growth-arrested for 5 days under the cytokine-starved condition were morphologically of a myeloblastic phenotype without showing any nuclear segmentation. Hence, a simple G0/G1 cell cycle arrest is not sucient for triggering granulocytic di€erentiation. In striking contrast, cells

ectopically overexpressing p130 under the cytokinestarved condition provoked a dramatic increase in the proportion of di€erentiated granulocytes that were morphologically characterized by segmented nuclei (Figure 4d,e, Table 1). From these observations, we concluded that p130 performs additional functions which contribute to morphological di€erentiation other than simply inhibiting the cell cycle in 32Dcl3 cells. However, in contrast to G-CSF stimulation, ectopic expression of p130 did not induce myeloid di€erentiation markers such as myeloperoxidase and CD14 in 32Dcl3 cells (data not shown). Cdk inhibition is not sucient for triggering granulocytic di€erentiation p130 is known to bind cyclin E-Cdk2 or cyclin ACdk2 and, as a result, is capable of inhibiting Cdk2 activity at least in vitro (Zhu et al., 1995; Woo et al., 1997; Castano et al., 1998). This Cdk inhibitory activity requires N-terminal and spacer regions of p130 which are not conserved in pRB. To address if such a Cdk inhibitory activity is sucient in triggering granulocytic di€erentiation, we generated 32Dcl3derived stable transfectants, 27 ± 61 and 27 ± 111, that conditionally overexpress p27 Cdk inhibitor upon IPTG treatment in the absence of tetracycline (Figure 5a). Induction of p27 in 32Dcl3 cells gave rise to strong cell growth inhibition despite the

Figure 2 E€ect of adenovirus E1A on G-CSF-dependent granulocytic di€erentiation of 32Dcl3 cells. (a) Ectopic expression of adenovirus E1A protein. The parental 32Dcl3 cell or 32D-B5 cell, a 32Dcl3-derived clone that stably co-express the lac repressor and the tetracycline-repressible transactivator, were cultured in IL-3 medium (20% WEHI) in the presence of tetracycline. The two E1A transfectants, E1A-4 and E1A-5, were cultured in IL-3 medium in the presence of IPTG for the protein induction. After 24 h culture, cells were lysed and the lysates were subjected to 10% SDS ± PAGE, followed by anti-E1A immunoblotting. (b) Morphological examination of E1A expressing cells treated with G-CSF. The parental 32Dcl3 cells and the E1A expressing transfectant clones, E1A-4 and E1A-5, were cultured in medium supplemented with 2.5 ng/ml G-CSF in the absence of IL-3. On day 5, cells were harvested and subjected to May-Grunwald-Giemsa staining. Dead cells were removed by density centrifugation

p130-mediated granulocytic differentiation A Mori et al

continued presence of IL-3 (Figure 5b). However, these cells did not exhibit any morphological signs of granulocytic di€erentiation after 5 days of culture in IL-3-containing medium (Figure 5c). Since p130dependent granulocytic di€erentiation was strongly promoted under the cytokine-starved condition, we also inducibly expressed p27 in cells cultured with cytokine-starved medium. Despite strong G0/G1 cell cycle arrest (Figure 6a ± c), the p27 expressing cells did not undergo di€erentiation (Figure 6d). We note here that cells growth-arrested by cytokine starvation hardly induced p130 (Figure 6c). In contrast, cell arrested by p27 expressed signi®cant levels of p130 on day 5 after the onset of p27 induction. In these cells, p130 may be secondarily induced in response to p27mediated cell cycle arrest, which is quicker and stronger than that mediated by cytokine deprivation. Yet, the accumulated p130 may be insucient to accomplish morphological changes within 5 days of culture in our experimental condition. These p27 overexpressing cells rapidly lost viability after 5 days in culture.

6213

Role of E2F in granulocytic di€erentiation The best studied target of pRB family proteins is the E2F family of transcription factors. We have shown previously that p130 inhibits cytokine-dependent proliferation of BaF3 pro-B cell by interacting with E2F-4 (Hoshikawa et al., 1998). As is the case of BaF3, E2F-4 was predominantly expressed throughout the cell cycle of 32Dcl3 and the activities of other E2Fs except E2F-3 were hardly detectable in an E2F gel shift-supershift assay (Figure 7). Almost exclusive expression of E2F-4 is consistent with the observation that the growth of 32Dcl3 cells is eciently inhibited by p130 but not by pRB (Figure 3). To address more directly if the inhibition of E2F transcriptional activities by p130 plays a role in cell growth arrest and/or cell di€erentiation, we generated 32Dcl3-derived stable transfectants that conditionally overexpress a dominant negative form of DP-1. The dominant negative DP-1 (dnDP-1) lacks amino acid residues between 103 and 126 and binds E2Fs, thereby forming a defective E2F/DP-1 heterodimer that is

Figure 3 E€ect of the pRB family pocket proteins on the growth and di€erentiation of 32Dcl3 cells. (a) Ectopic expression of p130. The parental cells (32Dcl3 and 32D-B5) and the two p130 transfectants, p130-1 and p130-3, were cultured in IL-3 medium (20% WEHI). The p130-transfectants were treated with IPTG for 24 h for the protein induction. The cell lysates were then prepared and subjected to 7.5% SDS ± PAGE, followed by anti-p130 immunoblotting. Position of p130 was indicated. (b) E€ect of the ectopically expressed p130 on the growth of 32Dcl3 cells. The parental cells (32Dcl3 and 32D-B5) and the p130 transfectants (130-1 and 130-3) were seeded at the concentration of 16105 /ml and were cultured in IL-3 medium in the presence of IPTG. At 24 h intervals, viable cell number was counted by Trypan-blue dye exclusion methods. Cell culture was diluted tenfold every 2 ± 3 days so as not to exceed the cell density of 16106 /ml. (c) Ectopic expression of a phosphorylation-resistant pRB, pRBDp34HA. The parental cells (32Dcl3 and 32D-B5) and the pRBDp34HA transfectants, H1-12 and H3-8, were cultured in IL-3 medium. The pRBDp34HA transfectants were treated with IPTG for 24 h for the protein induction. The cell lysates were then prepared and subjected to 7.5% SDS ± PAGE, followed by anti-pRB immunoblotting. Positions of the hypophosphorylated pRB (pRB) that corresponds to pRBDp34HA and the hyperphosphorylated form of the endogenous pRB (ppRB) were indicated. (d) E€ect of pRBDp34HA on the growth of 32Dcl3 cell. The parental cells (32Dcl3 and 32D-B5) and the two pRBDp34HA transfectants (H1-12 and H3-8) were seeded at the concentration of 16105 /ml and were cultured in IL-3 medium (20% WEHI) in the presence of IPTG. At 24 h intervals, cell number was counted by the Trypan-blue dye exclusion methods. (e) Morphological examination of 32Dcl3 cells that ectopically express p130 or pRBDp34HA. Cells used for (b) and (d) were collected after 5 days culture in IL-3 medium (20% WEHI), ®xed and subjected to May-Grunwald-Giemsa staining. Dead cells were removed by density centrifugation

p130-mediated granulocytic differentiation A Mori et al

6214

Figure 4 Promotion of p130-mediated granulocytic di€erentiation in 32Dcl3 cells growth-arrested at G0/G1. (a) Prolonged cell survival in medium supplemented with low concentrations of IL-3. Parental cells (32Dcl3 and 32D-B5) and the p130 transfectants (130-1 and 130-3) were cultured in medium containing 15 pg/ml IL-3. The p130 transfectants were treated with IPTG for 24 h for the protein induction. Viable cell number was counted after testing for cell viability by Trypan-blue dye exclusion every 24 h. Under this culture condition, cell viability on day 5 was approximately 80%. Notably, most cells die within 24 h in IL-3-free medium. (b) G0/G1 cell cycle arrest by the low-dose IL-3 treatment. Cells cultured in medium containing 15 pg/ml IL-3 were collected on day 3 and their cell cycle distribution was examined by ¯ow cytometric analysis. Exponentially growing 32Dcl3 cells in IL-3 medium (20% WEHI) were used as a control. (c) Inhibition of DNA synthesis by the low-dose IL-3 treatment. Cells (16104) used for the cell cycle analysis in (b) were pulsed with 1 mCi 3H-thymidine for 4 h and the 3H-thymidine incorporation into DNA was measured. Exponentially growing 32Dcl3 cells in IL-3 medium were used as a control. (d) Morphological examination of p130 expressing 32Dcl3 cells under growth-inhibited state. The p130 transfectants were cultured in medium supplemented with 15 pg/ml IL-3 in the presence of IPTG for 5 days, collected, ®xed and subjected to May-Grunwald-Giemsa staining. The parental 32Dcl3 and 32D-B5 cells were also cultured in medium supplemented with 15 pg/ml IL-3 for 5 days. Dead cells were removed by density centrifugation. (e) Detection of ectopic p130 in the di€erentiated cells. Lysates were prepared from the cells used in (d), resolved by 7.5% SDS ± PAGE and subjected to anti-p130 immunoblotting. Position of p130 is indicated

Table 1 Morphological stages of 32Dcl3 myeloid cells expressing p130 Treatment/Stage (%) Myeloblast Promyelocyte Myelocyte Metamyelocyte Band cell Polymorphonuclear leukocyte a

G-CSF

32Dcl3 IL-3

Ld. IL-3

1.5 16.2 21.7 30.0 24.0 6.6

98.5 1.5 0.0 0.0 0.0 0.0

93.1 3.9 1.5 1.5 0.0 0.0

a

Cell clone 32DB5 IL-3 Ld. IL-3a

IL-3

97.9 2.1 0.0 0.0 0.0 0.0

19.2 20.1 21.3 24.1 12.3 3.0

94.6 3.3 1.8 0.3 0.0 0.0

130-1 Ld. IL-3a 0.0 1.8 6.0 20.4 48.1 23.7

IL-3 16.5 17.4 16.2 35.2 12.6 2.1

130-3 Ld. IL-3a 0.0 1.8 8.1 30.9 43.9 15.3

Cells were cultured for 5 days in medium containing 15pg/ml recombinant IL-3 Ld; Low-dose

incapable of binding E2F-speci®c DNA sites (Wu et al., 1996). Like p130, induced expression of the dnDP-1 e€ectively inhibited IL-3-dependent proliferation of 32Dcl3 cells (Figure 8a,b). This reinforces the importance of the E2F transcriptional activity in cell cycle progression. However, the dnDP-1 overexpression did not provoke any signs of morphological di€erentiation of the myeloid progenitor cells (Figure 8c). Hence, inhibition of the E2F transcriptional activity suppresses

the myeloid cell growth but does not suce granulocytic di€erentiation. Delineation of p130 regions involved in granulocytic di€erentiation It has been demonstrated that the p130-mediated inhibition of the cell cycle requires the C-terminal domain, which binds E2F-4 and E2F-5, but not the

p130-mediated granulocytic differentiation A Mori et al

spacer domain that is involved in the regulation of cyclin ± Cdk2 activity (Hoshikawa et al., 1998a). To investigate the contribution of these p130 domains in granulocytic di€erentiation, we ectopically overexpressed two p130 mutants; one which lacks the cyclin A/E ± Cdk2 binding spacer domain, p130(D620-697) and the other mutant, p130(D846-1119), which is incapable of binding E2F but retains the cyclin ± Cdk binding activity (Lacy and Whyte, 1997; Hoshikawa et al., 1998a) (Figure 9a,b). Despite comparable levels of mutant p130 protein expression with that of wild type p130, stable transfectants expressing each of the p130 mutants exhibited little tendency to di€erentiate, as judged by the absence of segmented nuclei following 5 days culture in cytokinestarved condition (Figure 9b). This result indicates that both of the p130 domains are indispensable for the p130dependent granulocytic di€erentiation. Discussion We demonstrate in this work that granulocytic di€erentiation of myeloid progenitor cells is under the control of the pRB family pocket protein and that a particular member of the family, p130, plays a dual role in regulating myeloid cell cycle and differentiation. We delineate two distinct p130 domains that are indispensable for promoting granulocytic differentiation.

32Dcl3 is a non-tumorigenic myeloblastoid cell line established from mouse bone marrow and is totally dependent on interleukin-3 (IL-3) for its proliferation and survival (Greenberger et al., 1983). When the 32Dcl3 cells are treated with G-CSF in the absence of IL-3, they exit from the cell cycle and terminally di€erentiate to granulocytes. As are the cases of muscle or neuronal di€erentiation (Gu et al., 1993; Mymryk et al., 1992; Kalman et al., 1993; Webster et al., 1988; Boulukos and Zi€, 1993; Tiainen et al., 1996), this GCSF-dependent granulocytic di€erentiation is blocked by ectopic expression of the E1A oncoprotein, which binds and functionally inactivates the pRB family of pocket proteins (Whyte et al., 1988; Giordano et al., 1991; Ewen et al., 1991; Cobrinik et al., 1993). The observation indicates a requirement for the biologically active pRB family proteins in the process of granulocytic di€erentiation. In particular, selective accumulation of p130 that is concomitantly associated with down-regulation of pRB and p107 strongly suggests that p130 is a physiologically relevant pocket protein involved in G-CSF-mediated granulocytic di€erentiation. Ectopic overexpression of p130 in exponentially growing 32Dcl3 cells gave rise to cell growth inhibition that was followed by morphological di€erentiation characterized by nuclear segmentation. We have previously shown that the growth of another hematopoietic cell line BaF3 is also sensitive

Figure 5 E€ect of p27 Cdk inhibitor expressed in cycling 32Dcl3 cells. (a) Detection of ectopic p27. p27 transfectant clones, 27-61 and 27-111, were cultured in IL-3 medium (20% WEHI) in the presence of IPTG for 24 h. Cell lysates were then prepared and subjected to 10% SDS ± PAGE, followed by anti-p27 immunoblotting. Position of p27 is indicated. (b) Growth inhibition of 32Dcl3 cells by the ectopic expression of p27. Cells were seeded at 56104 /ml in IL-3 medium (20% WEHI) in the presence of IPTG. Cell number was determined at 24 h intervals by Trypan-blue dye exclusion methods. Cell culture was diluted tenfold every 2 ± 3 days so as not to exceed the cell density of 16106 /ml. (c) Morphological examination of p27 expressing cells. The p27 transfectants, 27-61 and 27-111, were cultured in IL-3 medium in the presence of IPTG for the protein induction. On day 5, cells were collected, ®xed and subjected to May-Grunwald-Giemsa staining. The parental 32Dcl3 cell cultured in IL-3 medium was shown as a negative control of di€erentiation. Dead cells were removed by density centrifugation

6215

p130-mediated granulocytic differentiation A Mori et al

6216

Figure 6 E€ect of p27 Cdk inhibitor expressed in growth-inhibited 32Dcl3 cells. p27 does not promote granulocytic di€erentiation in the growth-arrested 32Dcl3 cells. (a) Strong G0/G1 cell cycle arrest by p27 in IL-3-starved 32Dcl3 cells. p27 was inducibly expressed in the p27 transfectants, 27-61 and 27-111, in medium containing 15 pg/ml IL-3 in the presence of IPTG. After 3 days culture, cells were collected and were subjected to ¯ow cytometric analysis. At this point, more than 80% of cell were viable as determined by Trypan-blue dye exclusion methods. (b) Inhibition of DNA synthesis by p27. The same cells used for (a) was subjected to 3H-incorporation assay. Exponentially growing 32Dcl3 cells in IL-3 medium (20% WEHI) were employed as a control. (c) Anti-p27 and anti-p130 immunoblot analyses. Cells were cultured in cytokine-starved medium (15 pg/ml IL-3) in the presence of IPTG. After 5 days culture, cells were harvested and lysates were prepared. The lysates were then subjected to SDS ± PAGE, followed by anti-p27 (upper) or anti-p130 (lower) immunoblotting. Positions of p27 and p130 are indicated. (d) Morphological examination of cells ectopically expressing p27 under growth-inhibited state. The p27 transfectants were cultured in cytokine-starved medium (15 pg/ml IL-3) in the presence of IPTG. After 5 days culture, cells were harvested, ®xed and subjected to May-GrunwaldGiemsa staining. The parental 32Dcl3 cell cultured in medium containing 15 pg/ml IL-3 was shown as a negative control of di€erentiation. Dead cells were removed by density centrifugation

Figure 7 E2F-DNA binding activity in 32Dcl3 cells. Whole cell extracts prepared from exponentially growing 32Dcl3 cells were subjected to an E2F gel assay. Wild-type (wt) or mutant (mt) E2F gel shift oligonucleotides were used as competitors. The positions of free E2F-4 and E2Fs complexed with p107 in cycling 32Dcl3 cells are indicated. Anti-p130 (SC-317) used in the assay crossreacts with p107

to p130 but is resistant to pRB (Hoshikawa et al., 1998a). Again, in the present work, we demonstrate that IL-3-dependent growth of the 32Dcl3 myeloid cell is resistant to pRB even in its phosphorylationresistant form. Our results thus provide additional evidence for the di€erential sensitivity of cells to pRB family members and indicates the existence of a cell cycle control mechanism that does not depend on pRB. Based on the observation with BaF3 cells, p130-mediated inhibition of 32Dcl3 cells is likely to involve interaction of p130 with E2F. This is particularly true for E2F-4, because both cells predominantly express E2F-4 whereas expression of pRB-controlled E2Fs such as E2F-1 and E2F-3 is extremely low (Figure 7). In 32Dcl3 cells treated with G-CSF, accumulated p130 may interact with E2F, converting it from a transcriptional activator to a transcriptional repressor (Weintraub et al., 1992; Nevins et al., 1997; Zhang et al., 1999). Such transcriptional regulation is considered to underlie

p130-mediated granulocytic differentiation A Mori et al

6217

Figure 8 E€ect of dominant negative DP-1 (dnDP-1) on granulocytic di€erentiation of 32Dcl3 cells. (a) Ectopic expression of dnDP-1. Two dnDP-1 transfectant clones, dD1-36 and dD1-42, were cultured in IL-3 medium (20% WEHI) in the presence of IPTG for 24 h. Cell lysates were then prepared and subjected to 10% SDS ± PAGE, followed by anti-DP-1 immunoblotting. Position of dnDP-1 is indicated. (b) Growth inhibition by ectopic dnDP-1. The dnDP-1 transfectants (dD1-36 and dD1-42) were seeded at the concentration of 16105 /ml and were cultured in IL-3 medium in the presence of IPTG. At 24 h intervals, viable cell number was counted by Trypan-blue dye exclusion methods. (c) Morphological examination of dnDP-1 expressing cells. The dnDP1 transfectants were cultured in cytokine-starved (15 pg/ml IL-3) medium in the presence of IPTG for the protein induction. On day 5, cells were collected, ®xed and subjected to May-Grunwald-Giemsa staining. Dead cells were removed by density centrifugation

the p130-mediated cell growth suppression of hematopoietic cells. Concomitant association of granulocytic differentiation with growth inhibition by p130 raises a possibility that growth inhibitory activity is a prerequisite for triggering granulocytic di€erentiation by the pRB family pocket protein. Indeed, the importance of cell cycle arrest in promoting di€erentiation is supported by the observation that di€erentiation-inducing activity of p130 is strongly potentiated when cells are arrested in G0/G1 phase by cytokine starvation. However, a simple cell cycle arrest at G0/G1 phase by cytokine starvation or overexpression of the p27 Cdk inhibitor is incapable of triggering granulocytic di€erentiation. Accordingly, p130 plays an additional role in cell di€erentiation other than simply acting as a cell cycle inhibitor. In contrast to p130, overexpression of pRB in cycling 32Dcl3 cells neither inhibited cell proliferation nor induced morphological di€erentiation. Together with the observation that pRB is downregulated during G-CSF-induced di€erentiation, this argues against the possibility that pRB plays a central role in granulocytic di€erentiation. However, we also found that ectopically expressed pRB was capable of inducing morphological di€erentiation, although less

eciently than p130, once 32Dcl3 cells were growtharrested at G0/G1 phase by cytokine starvation (data not shown). Thus, p130 and pRB share certain activities in promoting cellular di€erentiation and the failure of ectopic pRB to trigger granulocytic di€erentiation appears to be primarily due to its inability to inhibit the growth of 32Dcl3 cells. It was previously reported that 32Dcl3 cells ectopically expressing anti-apoptotic oncoprotein Bcl2 have a prolonged survival in cultures without IL-3 and undergo granulocytic di€erentiation which is characterized by nuclear segmentation (Rodel and Link, 1996). As is the case of p130, this Bcl2-mediated morphological di€erentiation is not associated with the induction of di€erentiation markers such as myeloperoxidase and cathepsin G. Thus the Bcl2dependent di€erentiation may be explained by an accumulation of p130 in G0/G1-arrested 32Dcl3 cells surviving long-term culture without cytokine because of the presence of Bcl2. We also observed in this work that p130 accumulated in cells growth-arrested by p27 overexpression. Although these p27 expressing cells did not show any morphological changes during 5 days of culture and rapidly lost viability thereafter, such cells may eventually undergo di€erentiation by

p130-mediated granulocytic differentiation A Mori et al

6218

Figure 9 E€ect of p130 mutants expressed in 32Dcl3 cells. (a) Expression of the p130(D846-1119) mutant that lacks the C-terminal pocket domain of p130. The two transfectant clones, dE34 and dE41, were cultured in IL-3 medium (20% WEHI) in the presence of IPTG for 24 h. Cell lysates were then prepared and were subjected to 7.5% SDS ± PAGE that was followed by anti-p130 immunoblotting. Position for the p130(D846-1119) mutant is indicated. (b) Expression of the p130(D620-697) mutant that lacks the spacer domain of p130. The two transfectant clones, dS15 and dS142, were cultured in IL-3 medium (20% WEHI) in the presence of IPTG for 24 h. Cell lysates were then prepared and were subjected to 7.5% SDS ± PAGE that was followed by anti-p130 immunoblotting. Position for the p130(D620-697) mutant is indicated. (c) Morphological examination of 32Dcl3 cells expressing the p130 mutants. Cells were cultured in cytokine-starved medium containing 15 pg/ml IL-3 in the presence of IPTG for 5 days and were subjected to May-Grunwald-Giemsa staining. The parental 32Dcl3 cell and the p130-expressing 130-1 cell cultured in medium containing 15 pg/ml IL-3 for 5 days were shown as a negative and a positive controls for granulocytic di€erentiation, respectively. Dead cells were removed by density centrifugation

Table 2

Morphological stages of 32Dcl3 myeloid cells expressing p27

Treatment/Stage (%) Myeloblast Promyelocyte Myelocyte Metamyelocyte Band cell Polymorphonuclear leukocyte

Cell clone 27-61 27-111 a low-dose IL-3 low-dose IL-3a 91.0 4.2 4.2 0.6 0.0 0.0

90.1 5.7 4.2 0.0 0.0 0.0

a

p27 was inducibly expressed by treating cells with IPTG. The p27 expressing cells were cultured for 5 days in medium containing lowdose (15pg/ml) IL-3

virtue of the elevated p130 if they could have prolonged survival in culture. In contrast, ectopic overexpression of cyclin D2 or D3 in 32Dcl3 cells shortens their G1 phase and prevents their ability to di€erentiate into granulocytes in response to G-CSF (Kato and Sherr, 1993). In such cells, p130 may be inactivated through phosphorylation by deregulated cyclin D-associated kinase.

Although a molecular mechanism that turns on the granulocytic di€erentiation program in the myeloid progenitor cells remains obscure, it is most likely to involve the pocket function of p130 because E1A e€ectively blocks the di€erentiation process. While p130-E2F interaction appears to play a role in cell growth inhibition, our study does not necessarily support the idea that the p130-E2F complex also directly contributes to granulocytic di€erentiation from the following reasons; First, a p130 mutant that still interacts with E2Fs is incapable of promoting di€erentiation. Second, inhibition of E2F activities by dominant negative DP-1 does not suce di€erentiation under the growth-inhibited state. Accordingly, p130 appears to regulate the activities of molecules distinct from E2Fs during granulocytic di€erentiation. In cells treated with G-CSF or ectopically expressing p130, elevated p130 which is in relative excess to E2F may start to interact with such di€erentiation regulators and turn on programs for granulocytic di€erentiation. Consistent with this idea, pRB is demonstrated to interact with cell-type speci®c transcription factors in muscle cells or adipocytes and the interaction is suggested to promote cell type-speci®c di€erentiation (Gu et al.,

p130-mediated granulocytic differentiation A Mori et al

1993; Chen et al., 1996; Zacksenhaus et al., 1996; Tevosian et al., 1997). We found in this work that the two independent domains of p130 (Lacy and Whyte, 1997; Hoshikawa et al., 1998a), the spacer domain and the pocket domain, are indispensable for promoting granulocytic di€erentiation in 32Dcl3 cells. Given the facts that pRB is capable of inducing morphologic differentiation in the growth-arrested 32Dcl3 cells and that a common structural feature between p130 and pRB is the pocket domain, it is reasonable to speculate that the pocket domains are primarily responsible for induction of di€erentiation in growth-inhibited state. In addition, the p130 but not pRB pocket domain appears to play a role in cell growth inhibition by virtue of its interaction with E2F-4. The growth inhibitory activity of p130 may be further potentiated by its CDK inhibitory activity via the p130speci®c spacer region. Importantly, this spacer region is also required for promoting granulocytic differentiation in growth-inhibited state. Thus, the two independent domains appear to play active roles in both cell growth control and cell di€erentiation by p130. Genetically engineered mice lacking p130 do not manifest any hematological abnormality (Cobrinik et al., 1996). The discrepancy between our results and those from the p130 nullizygous mice may be explained by the existence of p107 and/or pRB that can at least in part function redundantly with p130. In fact, pRB is capable of compensating for p130 function in myeloid di€erentiation once the cell cycle is arrested by another mechanism(s). Alternatively, as recently suggested, additional pRB family pocket proteins that share redundant functions with p130 might also exist (Hurford et al., 1997). Such molecules could be upregulated in p130 nullizygous mice and complement p130 function. Existence for the functional modi®ers of the pocket proteins is also suggested by the recent observation that p130 nullizygous mice with di€erent strain background display embryonic lethality (LeCouter et al., 1998a). Such modi®ers might alter biological activities of the pocket proteins in a mouse-strain speci®c manner. Our works presented here provide evidence for a critical role of a particular member of the pRB family pocket protein, p130, in hematopoietic cell differentiation. Our observations indicate that p130 possesses two critical activities in myeloid cells, one linked to cell cycle progression and the other to cellular differentiation. In response to cell fate decisions, the bifunctional nature of p130 indicates that pocket proteins may act as critical coordinators of cell growth and differentiation. Materials and methods Cell culture Interleukin-3 (IL-3)-dependent 32Dcl3 cell was maintained in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS) and 20% WEHI-3B conditioned media (20% WEHI) as a source of IL-3. 32D-B5 cell is a 32Dcl3-derived transfectant clone that stably co-expresses the tetracyclinerepressible transactivator (tTA) and the E. coli lac repressor as described (Hoshikawa et al., 1998b).

Construction of plasmids pOPTET-BSD is an inducible cDNA expression vector having the TcIP promoter and the blasticidin-resistance gene as a drug-selection marker (Hoshikawa et al., 1998a,b). pOPTET-puro was constructed from pOPTET-BSD by replacing the blasticidin-resistance gene with the puromycinresistance gene. cDNAs were subcloned into the pOPTET vectors as described and the resulting plasmids were transfected into 32D-B5 cells by electroporation. Transfected cells were selected 48 h after transfection by adding 20 mg/ml blasticidin S (Nakarai) or 1.5 mg/ml puromycin to the culture medium containing 1 mg/ml tetracycline (Tc). Two to three weeks after transfection, drug-resistant cells were single-cell cloned by limiting dilution methods. Protein induction and immunoblot analysis Stable transfectants were cultured in medium supplemented with 1 mg/ml Tc to suppress cDNA expression. For protein induction, cells were washed twice with phosphate-bu€ered saline (PBS) and cultured in medium containing 5 mM isopropyl-thiogalactopyranoside (IPTG) without Tc for 24 h before cell lysates were prepared. Cells were lysed in ELB bu€er (250 mM NaCl, 5 mM EDTA, 50 mM HEPES/pH 7.0, 0.5% NP40, 1 mM PMSF, 10 mg/ml leupeptin, 10 mg/ml aprotinin) and the cell lysates were loaded on a 7.5, 10 or 15% SDS-polyacrylamide gel (SDS ± PAGE) and blotted onto PVDF membrane ®lter as described (Mizuguchi and Hatakeyama, 1998). Proteins were visualized using ECL detection system (NEN). Anti-p130 (SC-317) and anti-p27 (SC-527) were purchased from Santa Cruz Biotechnology. Anti-pRB (14001A) and anti-E1A (14161A) were purchased from PharMingen. Cell growth assay After washing twice with PBS, cells were cultured in medium containing 2.5 ng/ml recombinant murine G-CSF or 15 pg/ml recombinant murine IL-3 or 20% WEHI as an IL-3 source in the presence of Tc or IPTG. For cell growth curves, cells were harvested at 24 h intervals, and viable cell number was counted by the Trypan-blue dye exclusion method. Morphologic characterization Cells were collected and cytospins of 16106 cells were made using the Auto Smear CF 120 cytocentrifuge (Sakura). Necrotic cells were removed by centrifugation through a cushion of Histopaque 1077 (Sigma) at 400 g for 30 min. Cells were ®xed and subjected to May-Grunwald-Giemsa staining by using established method. Flow cytometry Cells were washed in PBS, ®xed by adding cold ethanol to a ®nal concentration of 70%, and stored for at least 30 min at 48C. The ®xed cells were then centrifuged, washed twice with PBS, and resuspended in 0.5 ml of PBS containing 20 mg/ml propidium iodide and 10 mg/ml RNase A. Samples were incubated for 30 min at 48C prior to ¯ow cytometric analysis with a Becton Dickinson FACS Calibur. 3

H-thymidine incorporation assay

Cells were cultured in medium containing 20% WEHI or 15 pg/ml recombinant IL-3 in the presence of Tc or IPTG. After 68 h culture, cells were re-seeded at 16104 (100 ml of the same medium) in each well of a 96-well ¯at bottom plate and pulse-labeled with 1 mCi of 3H-thymidine for 4 h as described (Hatakeyama et al., 1989).

6219

p130-mediated granulocytic differentiation A Mori et al

6220

Gel shift assays E2F-DNA gel shift assay was performed as described (Ikeda et al., 1996; Hoshikawa et al., 1998a). Antibodies used for supershift assays were purchased from Santa Cruz Biotechnology except anti-pRB (14051A, PharMingen); anti-E2F-1 (SC-193x), anti-E2F-2 (SC-633x), anti-E2F-3 (SC-878x), antiE2F-4 (SC-1082x), anti-E2F-5 (SC-1083x), anti-p107 (SC318), and anti-p130 (SC-317).

Acknowledgments We thank Dr L Yamasaki for reading of the manuscript. We also thank Dr K Ando for providing 32Dcl3 cells; Dr P Whyte for mutant p130 cDNA; and Dr T Nakamura for technical assistance. This work was supported by Research Grants from the Princess Takamatsu Cancer Research Fund and from Nippon Boehringer Ingelheim Co., Ltd.

References Baldi A, De Luca A, Claudio PP, Baldi F, Giordano GG, Tommasino M, Paggi MG and Giordano A. (1995). J. Cell Biochem., 59, 402 ± 408. Beijersbergen RL, Kerkhoven RM, Zhu L, Carlee L, Voorhoeve PM and Bernards R. (1994). Genes Dev., 8, 2680 ± 2690. Beijersbergen RL, Carlee L, Kerkhoven RM and Bernards R. (1995). Genes Dev., 9, 1340 ± 1353. Beijersbergen RL and Bernards R. (1996). Biochim. Biophys. Acta., 1287, 103 ± 120. Boulukos KE and Zi€ EB. (1993). Oncogene, 8, 237 ± 248. Castano E, Kleyner Y and Dynlacht BD. (1998). Mol. Cell. Biol., 18, 5380 ± 5391. Chen PL, Riley DJ, Chen Y and Lee WH. (1996). Genes Dev., 10, 2794 ± 2804. Clarke AR, Maandag ER, van Room M, van der Lugt NM, van der Valk M, Hooper ML, Berns A and te Riele H. (1992). Nature, 359, 328 ± 330. Cobrinik D, Whyte P, Peeper DS, Jacks T and Weinberg RA. (1993). Genes Dev., 7, 2392 ± 2404. Cobrinik D, Lee MH, Hannon G, Mulligan G, Bronson RT, Dyson N, Harlow E, Beach D, Weinberg RA and Jacks T. (1996). Genes Dev., 10, 1633 ± 1644. Egan C, Jelsma TN, Howe JA, Bayley ST, Ferguson B and Branton PE. (1988). Mol. Cell. Biol., 8, 3955 ± 3959. Ewen ME, Xing YG, Lawrence JB and Livingston DM. (1991). Cell, 66, 1155 ± 1164. Ewen ME, Faha B, Harlow E and Livingston DM. (1992). Science, 255, 85 ± 87. Faha B, Ewen ME, Tsai LH, Livingston DM and Harlow E. (1992). Science, 255, 87 ± 90. Garriga J, Limon A, Mayol X, Rane SG, Albrecht JH, Reddy EP, Andres V and Grana X. (1998). Biochem. J., 333, 645 ± 654. Giordano A, McCall C, Whyte P and Franza Jr BR. (1991). Oncogene, 6, 481 ± 485. Green MR, Treisman R and Maniatis T. (1983). Cell, 35, 137 ± 148. Greenberger JS, Sakakeeny MA, Humphries RK, Eaves CJ and Eckner RJ. (1983). Proc. Natl. Acad. Sci. USA, 80, 2931 ± 2935. Gu W, Schneider JW, Condorelli G, Kaushal S, Mahdavi V and Nadal-Ginard B. (1993). Cell, 72, 309 ± 324. Hatakeyama M, Brill JA, Fink GR and Weinberg RA. (1994). Genes Dev., 8, 1759 ± 1771. Hatakeyama M, Mori H, Doi T and Taniguchi T. (1989). Cell, 59, 837 ± 845. Hijmans EM, Voorhoeve PM, Beijersbergen RL, van't Veer LJ and Bernards R. (1995). Mol. Cell. Biol., 15, 3082 ± 3089. Hoshikawa Y, Mori A, Amimoto K, Iwabe K and Hatakeyama M. (1998a). Proc. Natl. Acad. Sci. USA, 95, 8574 ± 8579. Hoshikawa Y, Amimoto K, Mizuguchi R and Hatakeyama M. (1998b). Anal. Biochem., 261, 211 ± 218. Hurford Jr RK, Cobrinik D, Lee MH and Dyson N. (1997). Genes Dev., 11, 1447 ± 1463. Ikeda MA, Jakoi L and Nevins JR. (1996). Proc. Natl. Acad. Sci. USA, 93, 3215 ± 3220.

Jacks T, Fazeli A, Schmitt EM, Bronson RT, Goodell MA and Weinberg RA. (1992). Nature, 359, 295 ± 300. Jelsma TN, Howe JA, Mymryk JS, Evelegh CM, Cunni€ NF and Bayley ST. (1989). Virology, 171, 120 ± 130. Kalman D, Whittaker K, Bishop JM and O'Lague PH. (1993). Mol. Biol. Cell., 4, 353 ± 361. Kao HT and Nevins JR. (1983). Mol. Cell. Biol., 3, 2058 ± 2065. Kato JY and Sherr CJ. (1993). Proc. Natl. Acad. Sci. USA, 90, 11513 ± 11517. Knudsen ES and Wang JY. (1996). J. Biol. Chem., 271, 8313 ± 8320. Koh J, Enders GH, Dynlacht BD and Harlow E. (1995). Nature, 375, 506 ± 510. Lacy S and Whyte P. (1997). Oncogene, 14, 2395 ± 2406. LeCouter JE, Kablar B, Whyte PF, Ying C and Rudnicki MA. (1998a). Development, 125, 4669 ± 4679. LeCouter JE, Kablar B, Hardy WR, Ying C, Megeney LA, May LL and Rudnicki MA. (1998b). Mol. Cell. Biol., 18, 7455 ± 7465. Lee EY, Chang CY, Hu N, Wang YC, Lai CC, Herrup K, Lee WH and Bradley A. (1992). Nature, 359, 288 ± 294. Lees E, Faha B, Dulic V, Reed SI and Harlow E. (1992). Genes Dev., 6, 1874 ± 1885. Lillie JW, Loewenstein PM, Green MR and Green M. (1987). Cell, 50, 1091 ± 1100. Lukas J, Parry D, Aagaard L, Mann DJ, Bartkova J, Strauss M, Peters G and Bartek J. (1995a). Nature, 375, 503 ± 506. Lukas J, Bartkova J, Rohde M, Strauss M and Bartek J. (1995b). Mol. Cell. Biol., 15, 2600 ± 2611. Lundberg AS and Weinberg RA. (1998). Mol. Cell. Biol., 18, 753 ± 761. Matsushime H, Quelle DE, Shurtle€ SA, Shibuya M, Sherr CJ and Kato JY. (1994). Mol. Cell. Biol., 14, 2066 ± 2076. Mayol X, Garriga J and Grana X. (1995). Oncogene, 11, 801 ± 808. Mayol X, Garriga J and Grana X. (1996). Oncogene, 13, 237 ± 246. Medema RH, Herrera RE, Lam F and Weinberg RA. (1995). Proc. Natl. Acad. Sci. USA, 92, 6289 ± 6293. Meyerson M and Harlow E. (1994). Mol. Cell. Biol., 14, 2077 ± 2086. Mizuguchi R and Hatakeyama M. (1998). J. Biol. Chem., 273, 32297 ± 32303. Moberg K, Starz MA and Lees JA. (1996). Mol. Cell. Biol., 16, 1436 ± 1449. Moran E, Zerler B, Harrison TM and Mathews MB. (1986). Mol. Cell. Biol., 6, 3470 ± 3480. Mymryk JS, Lee RW and Bayley ST. (1992). Mol. Biol. Cell., 3, 1107 ± 1115. Nevins JR, Leone G, DeGregori J and Jakoi L. (1997). J. Cell Physiol., 173, 233 ± 236. Resnitzky D and Reed SI. (1995). Mol. Cell. Biol., 15, 3463 ± 3469. Rodel JE and Link DC. (1996). Blood, 87, 858 ± 864. Sardet C, Vidal M, Cobrinik D, Geng Y, Onufryk C, Chen A and Weinberg RA. (1995). Proc. Natl. Acad. Sci. USA, 92, 2403 ± 2407.

p130-mediated granulocytic differentiation A Mori et al

Sellers WR, Novitch BG, Miyake S, Heith A, Otterson GA, Kaye FJ, Lassar AB and Kaelin Jr WG. (1998). Genes Dev., 12, 95 ± 106. Sherr CJ. (1996). Science, 274, 1672 ± 1677. Smith EJ and Nevins JR. (1995). Mol. Cell. Biol., 15, 338 ± 344. Tam SW, Theodoras AM, Shay JW, Draetta GF and Pagano M. (1994). Oncogene, 9, 2663 ± 2674. Tevosian SG, Shih H, Mendelson KG, Sheppard KA, Paulson KE and Yee AS. (1997). Genes. Dev., 11, 383 ± 396. Tiainen M, Spitkovsky D, Jansen-Durr P, Sacchi A and Crescenzi M. (1996). Mol. Cell. Biol., 16, 5302 ± 5312. Vairo G, Livingston DM and Ginsberg D. (1995). Genes Dev., 9, 869 ± 881. Valtieri M, Tweardy DJ, Caracciolo D, Johnson K, Mavilio F, Altmann S, Santoli D and Rovera G. (1987). J. Immunol., 138, 3829 ± 3835. Velcich A and Zi€ E. (1988). Mol. Cell. Biol., 8, 2177 ± 2183. Webster KA, Muscat GE and Kedes L. (1988). Nature, 332, 553 ± 557. Weinberg RA. (1995). Cell, 81, 323 ± 330. Weintraub SJ, Prater CA and Dean DC. (1992). Nature, 358, 259 ± 261.

Woo MS, Sanchez I and Dynlacht BD. (1997). Mol. Cell. Biol., 17, 3566 ± 3579. Whyte P, Buchkovich KJ, Horowitz JM, Friend SH, Raybuck M, Weinberg RA and Harlow E. (1988). Nature, 334, 124 ± 129. Whyte P, Williamson NM and Harlow E. (1989). Cell, 56, 67 ± 75. Wu C-L, Classon M, Dyson N and Harlow E. (1996). Mol. Cell. Biol., 16, 3698 ± 3706. Xiao ZX, Ginsberg D, Ewen M, Livingston DM. (1996). Proc. Natl. Acad. Sci. USA, 93, 4633 ± 4637. Yee AS, Shih HH and Tevosian SG. (1998). Front Biosci., 3, D532 ± D547. Zacksenhaus E, Jiang Z, Chung D, Marth JD, Phillips RA and Gallie BL. (1996). Genes Dev., 10, 3051 ± 3064. Zarkowska T and Mittnacht S. (1997). J. Biol. Chem., 272, 12738 ± 12746. Zhang HS, Postigo AA and Dean DC. (1999). Cell, 97, 53 ± 61. Zhu L, Harlow E and Dynlacht BD. (1995). Genes Dev., 9, 1740 ± 1752.

6221