Mechanism of Transcriptional Activation of the ...

From www.bloodjournal.org by guest on July 19, 2016. For personal use only.

Mechanism of Transcriptional Activation of the Immediate Early Gene Egr-1 in Response to PIXY321 By Robert C. Mignacca, Hu-Jung Julie Lee, Evelyn M. Kwon, and Kathleen M. Sakamoto Studies with the granulocyte-macrophage colony-stimulating factor (GM-CSF)/interleukin-3 (IL-3) fusion protein, PIXY321, demonstrated enhanced biological activity of this molecule in comparison with GM-CSF or IL-3 alone or in combination. Experiments were performed t o study the mechanisms resulting in PIXY321-induced egr-I expression in human myeloid leukemic cells (TF-1).Transfections of egr1promoter constructs revealed that PIXY321 stimulation resulted in fourfold induction of the -116 and -600 nucleotide (nt) constructs. We transfected a -116 nt construct containing a deletion of the cyclic AMP response element (CRE) or mutation in the serum response element (SRE) and demonstrated that both the SRE and CRE are necessary for maximal induction. However, PIXY321stimulation resulted in 2.5fold induction of a SRE-CRE-containing construct ( P < .05), suggesting that the SRE and CRE are sufficient for PIXY321

responsiveness. Electrophoretic mobility shift assays (EMSA) revealed that the CRE binding protein (CREB) was phosphorylated on serine 133 in PIXY321-stimulated but not -unstimulated extracts from cells cultured in GM-CSF. By Western analysis and EMSA, CREB was constitutively phosphorylated in TF-1 cells grown on PIXY321 before growth factor and serum starvation. However, in TF-1 cells grown on GM-CSF before starvation, CREB phosphorylation was observed 10 minutes after PIXY321stimulation. Furthermore, EMSAs with PIXY321-stimulated and -unstimulated extracts demonstrated the presence of specific proteins that recognize the SRE. Our data demonstrate that transcriptional regulation of egr-1 by PIXY321 is mediated by the CRE and SRE. 0 1996 by The American Society of Hematology.

T

dromes demonstrated that PIXY321 has the potential to improve both platelet and neutrophil recovery.’-’” PIXY321 has the ability to bind to both the GM-CSF and IL-3 receptor^.^ The GM-CSF and IL-3 receptors are members of the hematopoietin receptor superfamily, which also includes the receptors for granulocyte colony-stimulating factor; erythropoietin; IL-2, -4, -5, -6, and -7; prolactin; and growth hormone. This family of receptors is notable for lack of kinase activity in their cytoplasmic regions.” A novel 100-kD protein associates with the receptor complex in cross-linking experiments in myeloid cells stimulated with PIXY321 (Linda Park, personal communication, July 1995). As with GM-CSF and IL-3, PIXY321 stimulation most likely results in the activation of signaling intermediates in the Jak/ STAT pathway and the Ras/Raf pathway via involvement of Shc or another as yet undefined pathways. Activation of these signal transduction cascades results in the induction of primary response genes.” We previously demonstrated that GM-CSF and IL-3 rapidly and transiently induce the primary response gene, egr1 (early growth response gene- l), in the human factor-dependent cell line, TF- 1, I 2 by activating signaling pathways that lead to phosphorylation of the cyclic AMP response element binding protein (CREB). CREB forms part of a transcription factor complex that requires both the cyclic AMP response element (CRE) and the serum response element (SRE) for activation of egr-1.I2 Additionally, we reported that GMCSF and IL-3 stimulation results in the phosphorylation of CREB on serine 133.” PIXY321 has the ability to bind both the GM-CSF and IL-3 receptors, and TF-1 cells express receptors for both. Therefore, the following experiments were undertaken to determine whether the greater activity and potency of PIXY321 could be explained by induction of egr-l via unique pathways of transcription factor binding sites. Our studies indicate that, like GM-CSF and IL-3, PIXY321 stimulation of TF- I cells results in phosphorylation of CREB on serine 133 through a transcription factor complex that requires both the CRE and the SRE for induction of egr- 1. However, the SRE-CRE sequence alone is sufficient for transcriptional activation of egr- 1 by PIXY 32 1 but not by

HE HYBRID molecule PIXY321 connects granulocytemacrophage colony-stimulating factor (GM-CSF) and interleukin-3 (IL-3) via a flexible protein linker. GM-CSF produces rapid increases in neutrophil numbers in the peripheral blood.’.2 IL-3 produces a slower but more sustained increase in neutrophils and, according to some studies, also leads to increased platelet production in PIXY32 1 therefore provides the potential to increase platelet and neutrophil numbers through the administration of a single agent. In thymidine incorporation assays with the AML-193 cell line, PIXY321 had 10- to 20-fold greater specific activity than GM-CSF andlor IL-3.’ In vitro clonogenic assays for erythroid progenitor cells, multilineage progenitor cells, and granulocyte-macrophage colony-forming units, using human bone marrow cells, also revealed 10- to 20-fold greater potency of PIXY321 in stimulating these cell types in comparison to GM-CSF and/or IL-3.‘ Phase I and I1 clinical trials using PIXY32 1 in patients receiving chemotherapy after bone marrow transplant or with bone marrow failure syn-

From the Division of Hematology/Oncology, Gwynne Hazen Cherry Memorial Laboratories, Department of Pediatrics, A2-312 MDCC, and Department of Pathology and Laboratory Medicine, UCLA School of Medicine and Jonsson Comprehensive Cancer Center, Los Angeles, CA. Submitted July 28, 1995; accepted March 20, 1996. Supported by a National Cancer Institute Clinical Investigator Award CA59463, James A. Shannon Director’s Award lR55CA/ 0068221-01, Concern I1 Foundation, and a Leukemia Society of America Special Fellowship (K.M.S.). R.C.M. is a postdoctoral trainee supported by USHHS National Institutional Research Service Award CA-09056. Address reprint requests to Kathleen M. Sakumoto, MD, Department of Pediatrics, A2-312, Division of Hematology-Oncology, UCLA School of Medicine, 10833 Le Conte Avr, Los Angeles. CA 90024- 1752. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. section 1734 solel! to indicate this fact. 0 1996 by The American Society of Hematology. 0006-4~71/96/8803-0029$3.00/0

848

Blood, Vol 88,No 3 (August I ) , 1996:pp 848-854


PIXY321 AND REGULATION OF egr-1 EXPRESSION

GM-CSF or IL-3. These studies showed that PIXY321 enhances myeloid cell proliferation through activation of different transcription factor complexes associating with the SRE and CRE. Furthermore, by Western blot analysis and electrophoretic mobility shift assays (EMSA), we demonstrated that cells cultured in PIXY321 before serum and growth factor starvation demonstrated constitutive phosphorylation of CREB, suggesting that TF-1 cells cultured in PIXY321 have different biological properties than cells cultured in GM-CSF. MATERIALS AND METHODS Cell culture. TF-I cells were obtained from T. Kitamura (DNAX, Palo Altos, CA)’4and cultured in RPMI medium containing 10% fetal calf serum, L-glutamine (2 mmol/L), penicillin (100 U/ mL) and streptomycin (100 mglmL) at a ratio of 1: I, and gentamicin (10 mmol/L) in nonadherent tissue culture plates. The cells required 500 pmol/L recombinant human GM-CSF (rhGM-CSF, a gift from Larry Souza at Amgen, Thousand Oaks, CA) or 200 pmol/L PIXY321 (generously provided by Linda Park at Immunex, Seattle, WA) for growth. Recombinant plasmid construction and oligonucleotide construct synthesis. Construction of deletion mutants of the human egr-1 promoter (containing -600, -480, -235, -180, - 116, and -56 nt) and synthesis of the constructs containing a mutated SRE (mutSRE) or a deleted CRE (delCRE) were described previously.” The SRE-CRE plasmid was constructed by synthesizing the SRE from nt - 113 to -88 (5’-AGCTTTCCTGCCATA~AGGGC~T/3’) in tandem with the region from nt -76 to -57 (5’-ATGCCATGTACGTCACGACG-3‘). which contains the downstream CRE (CRE2). The sequences of the SRE and CRE, respectively, are in italics. Hind111 and Xba I sites were created at the 5’ and 3’ ends, respectively. The oligonucleotides were annealed and ligated into the Hindlll and Xba I sites in the pTE2 vector.” The - 116 nt construct with deletion of the CRE was prepared by deleting nts -57 to -76 in the egr-l promoter, using the Amersham (Arlington Heights, IL) mutagenesis kit. For the - I 16 mutSRECAT construct, the SRE core element was mutated as previously described.” All oligonucleotide and promoter constructs have been sequenced using the Sequenase kit (US Biochemical, Cleveland, OH). Transient transfections and chloramphenicol acepltransferase (CAT) assays. TF-I cells grown in GM-CSF (500 pmoVL) were washed three times with phosphate-buffered saline (PBS) and placed in serum- and factor-free RPMI-0.5% bovine serum albumin (BSA) (Sigma Chemical CO, St Louis, MO) 24 hours before transfection. DNA (25 pg: 23 pg of the egr-1 construct and 2 pg of the CMV P-galactosidase plasmid) were electroporated at 200 V into five million cells per sample. Cells were then resuspended in RPMI0.5% BSA and stimulated with 1 nmol/L rhGM-CSF, 1 nmol/L recombinant human IL-3 (rhlL-3), 200 pmol/L PIXY32 I , or diluent control (PBS-0.02% BSA) for 4 hours. Cells were harvested for CAT or &galactosidase activity as previously described.’“’* Acetylation of ‘‘C-chloramphenicol was determined by liquid scintillation counting. Fold induction was determined by dividing the percent acetylation of the growth factor-stimulated construct by the diluent control for the same construct. Fold stimulation by the growth factors was corrected for transfection efficiency with the P-galactosidase assay (Promega, Madison, WI). Statistical analysis with the paired Student’s t-test was done using the STATWORKS program. EMSA. Probe was prepared using the region of the human egr1 pronwter from nt -76 to -57. which contains the downstream CRE (in italics) (5’-ATGCCATGTACGTCACGACG-3’).The SRI of the egr-1 promoter between nts -90 and - I10 was used as the probe for SRE EMSAS.” A complementary single-stranded oligonu-

849

cleotide probe was synthesized with an Applied Biosystems (Foster City, CA) synthesizer and then end-labeled using c-”P-dATP and T4 polynucleotide kinase. Labeled probe was purified from unlabeled oligonucleotide using Nuctrap Push Columns (Stratagene). Five million TF-I cells cultured in GM-CSF ( I nmol/L) or PIXY321 (200 pmol/L) were serum- and factor-starved for 24 hours and then stimulated with diluent control, 200 pmol/L PIXY321, or 50 nglmL tetradecanoyl phorbol acetate (TPA). Cells were lysed by sonication methods as described.” EMSA were performed by incubation of 1 ng of probe with 2 pL of cell lysate in the presence of I pg of poly(d1-dC):(dl-dC) and 5 pL of gel shift buffer (50 mmollL HEPES [pH 7.8],5 mmol/L spermidine, 15 mmol/L MgCI2, 36% glycerol, 3 mg/mL BSA, 0.3% NP40 and 15 mmol/L dithiothreitol) in 20 pL total volume (brought to full volume with water) for 30 minutes on ice. Excess (50- or 500-fold) of specific or nonspecific competitor was also included in the incubation mix. In supershift EMSA, rabbit antirat CREB (2 pL of a 1 : l O dilution) and antirat phosphorylated CREB (2 pg) antibodies (Upstate Biotechnology Inc. Lake Placid, NY) were added to the reaction. Nonspecific polyclonal rabbit IgG (Sigma, St Louis, MO) was used as the control antiserum. The antibody was incubated with gel shift reaction mix on ice for 30 minutes. After incubation, the samples were loaded onto a 4% polyacrylamide gel and run at a constant 110 V in 0.4X Tris-borateEDTA. Gels were dried and placed on film at -70°C. Western analysis. TF-I cells (4 X IO5) grown in rhGM-CSF (1 nmol/L), PIXY321 (200 pmol/L), or both rhGM-CSF (1 nmol/L) and rhlL-3 ( I nmol/L) were washed and serum- and factor-starved for 24 hours before stimulation with 200 pmol/L PIXY321 for 2, 5, 10, or 15 minutes or with TPA (50 ng/mL) for 10 minutes. Cells were lysed with boiling loading buffer (5% sodium dodecyl sulfate [SDS], 20% glycerol, 140 mmol/L Tris-HCI [pH 6.61, bromphenol blue and 10% P-mercaptoethanol) and boiled for 5 minutes according to methods described by Ginty et al.” Total cell lysate was loaded onto a 10% SDS-polyacrylamide gel and transferred onto nitrocellulose (Amersham) for 1 hour at 600 mA. Rabbit antirat CREB antibody at a 1:7,500 dilution or antirat phosphorylated CREB antibody (Upstate Biotechnology Inc) at a concentration of 0.14 pglmL were used as primary antibodies, and biotinylated, donkey antirabbit IgG was used as a secondary antibody (Amersham). The streptavidin-horseradish peroxidase conjugate (Amersham ECL Western kit) was used as the tertiary detection method. Film was exposed at room temperature for between 5 seconds and 3 minutes. RESULTS

Stimulation of the egr-1 promoter constructs by PIXY321. We transiently transfected the egr- 1 promoter constructs (Fig

-235

CREI

SRE CRU

TATA

cRE1

SRE C R U

TATA

RE CRU

TATA

e-[-1-

+-H-t-*+ 416

+-[-*+

*

-+ TATA

Fig 1. Egr-1 promoter constructs. The -600 nt Hinfl fragment of the human egr-1 promoter was digested with restriction endonucleases and subcloned into the pCATvector, resulting in the -480, -235, -180, -116, and -56 nt fragments shown. Previously known transcription factor binding sites are labeled Egr-1 binding site (EBS), SP1, SRE, and the CRE.


MIGNACCA ET AL

850

1) into TF-1 cells and stimulated them with PIXY321. Transfections of the egr- 1 constructs revealed a fourfold induction (corrected) of the -116 nt construct in comparison with pCAT control in cells stimulated with PIXY321 (P < .05). This was comparable with the -600 nt construct, which demonstrated a 4.2-fold induction with PIXY321 stimulation. The constructs containing -480, -235, and -180 nt of the egr-1 promoter demonstrated less than a twofold induction. The -56 nt construct was not stimulated by PIXY321 (Fig 2). The mean basal activity (represented by the average uncorrected percent acetylation of unstimulated cells) was 4% for -600 nt and 0.96% for the - 116 nt construct (Fig 2). which was not statistically significant (P = .268). In the case of the -56 CAT construct, the basal activity was 0.2% compared with 0.96% for the - 116 nt or 4% for the -600 nt CAT constructs, and these differences were not found to be statistically significant (P = .231 and .266, respectively). Thus, our data revealed that the degree of stimulation of the - 116 nt construct was most similar to that seen with the -600 nt construct; however, the percent acetylation (basal and stimulated activity) was lower in the - 1 16 nt construct. In previous experiments, we determined that the relative CAT activities of the egr-1 promoter constructs in cells treated with GM-CSF and IL-3 were identical to those seen in cells treated with GM-CSF alone (data not shown). TF-1 cells were cultured in GM-CSF before growth factor and serum starvation to maintain consistency throughout the transfections. These data represent the average of 3 to 13 independent experiments. We previously demonstrated that the CRE in the - 116 nt region was critical for GM-CSF and IL-3 responsiveness.

30

I

0

DILUENT (PBS+BSA.02%) PIXY321 (2W)PM)

I

z

P

4F +

20

+ U 0

U

s

10

CONSTRUCTS Fig 2. Transient transfections of egr-1 constructs in TF-1 cells. The egr-1 promoter constructs I23 pg) and the CMV P-galactosidase plasmid (2 pg) were transiently transfected into serum- and factorstarved TF-1 cells and stimulated with PIXY321 (200 pmol/L). Percentages of acetylation of constructs were calculated in cells stimulated with diluent or PIXY321. The asterisks indicatestatistical significance ( P < .05). P values were calculated by Student's paired t-test. These results represent the average of 3 to 13 independent experiments.

To determine whether the CRE is required for PIXY321 induction of the - 116 nt region of the egr-1 promoter, transient transfections using the - 116 nt construct containing a delCRE were performed (for CRE sequence deleted in the egr-1 promoter, see Materials and Methods). We did not observe stimulation of the - 1 16delCRE CAT construct with PIXY321 stimulation (Fig 3A). This decrease was statistically significant in three separate experiments (P < .05). However, - 116delCRECAT had greater activity in PIXY321-treated cells than GM-CSF- or IL-3-treated cells, suggesting transcription factors interact with sites other than the CRE in the - 116 nt region, thereby producing greater transcriptional activation. Mutation of the SRE core element in the context of the surrounding -116 nucleotides (- 116mutSRECAT) produced a statistically significant decrease in fold stimulation compared with wild-type - 116CAT (P< .05;Fig 3A). The sequence of the SRE that was mutated was previously described.'* Experiments with GM-CSF or IL-3 stimulation demonstrated 0.7- to 1.3-fold induction of - 116mutSRECATconstruct compared with the 2.5-fold with PIXY321 stimulation. The fold induction of mutant constructs stimulated with GM-CSF + IL-3 were identical to those with GM-CSF alone (data not shown). These results represent the average of 11 independent experiments. The basal activity between the wild-type - 116 nt construct and the mutated SRE construct was not significantly different (P = .230). To determine whether the SRE-CRE sequence alone was sufficient for transcriptional activity, we transiently transfected a construct containing the SRE-CRE sequence upstream of tkCAT into TF-1 cells. The SRE-CRE sequence contains nts -57 to -1 10 of the egr-1 promoter (see Materials and Methods). The SRE-CRE construct produced a 2.4fold induction of CAT activity after PIXY321 stimulation in comparison with pTE2 control (P< .05;Fig 3B). We were unable to demonstrate significant induction by PIXY321 of constructs containing either two copies of the SRE or the CRE alone (data not shown). PIXY321 stimulation results in CREB phosphorylation. We previously showed that GM-CSF and IL-3 stimulation of TF-1 cells resulted in phosphorylation of CREB on serine 133.13 To determine whether CREB binds to the CRE in PIXY321-stimulated TF-1 cell extracts, we performed EMSA using cell lysates from unstimulated and PIXY321stimulated TF-1 cells. The CRE probe contains sequences between nts -57 and -76 of the egr-1 promoter (see Materials and Methods). The TF-1 cells used for these experiments were cultured in GM-CSF (500 pmol/L) before 24-hour factor and serum starvation. The EMSA demonstrated that the CRE-CREB complex can be supershifted by the anti-CREB antibody. Using an antibody that specifically recognizes CREB phosphorylated on serine 133, we observed a supershift with PIXY321-stimulated cell extracts but not with unstimulated cell extracts (Fig 4).19 In contrast, in cells cultured in PIXY321 before starvation, low levels of phosphorylated CREB were seen with diluent stimulation (Fig 5, lane 5, complex 11), suggesting constitutive phosphorylation of CREB. TPA is known to activate CREB through a protein kinase C-dependent pathway. Thus, TF-1 cells were treated


851


1 2 -3- -4 --

A

5 6 7 8 9 10 111213 14

-- 1m

116CAT (PIXY321)

5

z 0 Io

0 IlWelCRE (PIXY321) ll6delCRE (GMCSF)

-1

Ul IlWelCRE (11-3) 0 1lGmmSRE (PIXY321) E ll6mutSRE (GM-CSF)

4

3

P

z

3

-PIXY 321

9

P

2

1

0

CONSTRUCTS

B

m

+ PIXY321

t TPA

Fig 4. EMSA with the CRE probe and PIXY321-stimulated TF-1 cell extracts. TF-1 cells (5 x 10') cultured in GM-CSF before growth factor and serum starvation were stimulated with diluent control (0.02% BSA in PBS), PIXY321 (200 pmol/L), or TPA (50 ng/mL) for 10 minutes. Cells were lysed by sonication. Probe (1 ngl (CRE from nt -76 t o -57 of the egr-1 promoter) was incubated with 2 p L of cell lysate, 1 p g poly(didC):(didC), and 5 p L of gel shift buffer. Lanes 2 and 7 contain 500-fold excess of cold CRE; lanes 4,9, and 13 contain anti-CREB antibody 12 p L of a 1 : l O dilution); lanes 5, 10, and 14 contain antiphosphorylated CREB antibody (2 pg); lanes 3, 8, and 12, extracts were incubated with equal volumes of rabbit IgG; and lane 11contains control with an unrelated egr-1 antibody (4 pL). Complex I represents specific complex, complex II with antiphosphorylated CREE antibody, and complex 111 with anti-CREE antibody.

SRECRE

I

response to PIXY321 stimulation, we performed a Western blot analysis and timecourse with TF-I cells. Cells cultured in PIXY32 1 resulted in constitutive phosphorylation of CREB (Fig 6A). In contrast, in TF-1 cells cultured in GMCSF 24 hours before factor and serum starvation, phosphorylation of CREB on serine 133 was observed after PIXY321

z 0 c 0

3 0

I

9

P

1 2 3 4 5 6 7 8 910

1112Q

CONSTRUCTS Fig 3. (A) Fold induction by PIXY321 of the vector pCAT, -116 nt egr-1 promoter construct (wild-type), the deleted CRE construct (delCRE), and the mutated SRE construct b"SRE) in response t o PIXY321 (200 pmol/L), GM-CSF (1nmol/L), or 11-3 (1nmol/L) stimulation. (BI The SRE-CRE sequence and the pTE2 vector are also shown. The egr-1 constructs (23 p g ) and the CMV 0-galactosidase plasmid (2 p g l were transiently transfected into serum- and factor-starved TF-1 cells and then stimulated with PIXY321 or diluent control (0.02% ESA in PBS). The fold stimulation of each construct was determined by dividing the percent acetylation of the PIXY321-stimulated sample by the percent acetylation of the diluent control sample. The fold induction was then corrected using the 0-galactosidase activity as the internal control. P values were calculated by paired Student's ttest. These data represent an average of 3 t o 11 separate experiments. The asterisks represent statistical significance ( P < .05).

with TPA for 5 minutes, and EMSAs were performed with these nuclear extracts. We used the potent activation of CREB in response to TPA as a positive control (Figs 4 and 5).

To determine the kinetics of CREB phosphorylation in

U PIX321 + PIX321

-

t TPA

Fig 5. EMSA with the CRE probe and PIXY32l-stimulated TF-1 cell extracts. TF-1 cells (5 x 10') cultured in PIXY321 before growth factor and serum starvation were stimulated with diluent control (0.02% BSA in PBS), PIXY321 (200 pmol/L), or TPA 150 ng/mL) for 10 minutes. Cells were lysed by sonication. Probe (1 ngl (CRE from nt -76 t o -57 of the egr-1 promoter) was incubated with 2 pL of cell lysate, 1 p g poly(didC):(didC), and 5 p L of gel shift buffer. Lanes 2 and 7 contain 500-fold excess of cold CRE; lanes 4,9, and 13 contain anti-CREB antibody (2 p L of a 1:lO dilution); lanes 5, 10, and 14 contain antiphosphorylated CREE antibody (2 pg); lanes 3, 8, and 12, extracts were incubated with equal volumes of rabbit lgG; and lane 11contains control with an unrelated egr-1 antibody (4 pL). Complex I represents specific complex, complex II with antiphosphorylated CREB antibody, and complex 111 with anti-CREE antibody. EMSAs with cells cultured in PIXY321 before factor or serum starvation demonstrate low levels of phosphorylated CREE with diluent stimulation (lane 5, complex 11).


852

A Min

:

0

2

5 IO 15 TPA 0

2

==:U anti-CREE

:

0

2

5 10 15 TPA 0

-=rz;*rcL

f i -46kDa

I

anti- pCREE

B Min

5 ( 0 15 TPA

2

-

Lanti-CREB

5 10 15 TPA #

-46kDa

U anti-CREE

anti-pCREB

Fig 6. Western blot analysis and timecourse with PIXY321-stimulated TF-1 cells. (A) TF-1 cells (4 x ’10’) cultured in PIXY321 before growth factor and serum starvation were stimulated with PIXY321 (200 pmol/L) for 2,5,10, or 15 minutes or with TPA (50 ng/mL). Cells were lysed with boiling SDS for 5 minutes. Total cell lysate was loaded onto a 10% SDS-polyacrylamide gel and transferred onto nitrocellulose. The blot was probed with anti-CREB antibody (1:7,500 dilution) or antiphosphorylated CREB antibody (0.14 pglmL). (B) TF1 cells (4 x 10’) cultured in GM-CSF before growth factor and serum starvation were stimulated with PIXY321 I200 pmollL). Factor- and serum-starved TF-1 cells were stimulated with PIXY321 (200 pmoll L) for 2, 5, 10, or 15 minutes or with TPA (50 ng/mL). Cells were lysed with boiling SDS. Total cell lysate was loaded onto a 10% SDSpolyacrylamide gel and transferred onto nitrocellulose. The blot was probed with anti-CREB antibody (1:7,500 dilution) or antiphosphorylated CREB antibody 10.14 pglmL).

stimulation for IO minutes (Fig 6B). These data confirmed our results with EMSA, where cells grown in PIXY321 had evidence of phosphorylated CREB after 24 hours of serum and growth factor starvation (Fig 5, lane 5, complex 11). For Westem blots, extracts from TF-I cells treated with TPA were used as a positive control. A doublet was observed that represents differential phosphorylation on CREB, resulting in a different migration pattern on the gel (Fig 6, A and B). The bands representing phosphorylated CREB are fainter with cells cultured in PIXY321 compared with GM-CSF before starvation (Fig 6, A and B). This finding is confirmed in the EMSA experiments, where generally the bands representing a supershifted complex with phosphorylated CREB are weaker (Fig 5) in PIXY321-cultured cells than GMCSF-cultured cells (Fig 4)before starvation. This may reflect decreased stability of phosphorylated CREB in cells grown in PIXY321. Furthermore, in TF-I cells cultured in both GM-CSF and IL-3 before serum and growth factor starvation, we did not observe evidence of phosphorylated CREB with diluent (Fig 7). Interestingly, a doublet was not seen in this experiment, suggesting that differential phosphorylation of CREB may vary, depending on specific conditions used (this finding was not consistently observed between experiments). These results suggested that cells cultured with PIXY321 before growth factor and serum starvation appeared to be biologically different from cells cultured in GM-CSF, IL-3, or both GM-CSF and IL-3. We also observed that in cells cultured in GM-CSF before starvation, phosphorylated CREB decreased from 10 to 15 minutes after PIXY321 stimulation (Fig 6B). This may be the re-

-

MIGNACCA ET AL

-46koO

-J

\L

-

-46kDa

onti- pCREB

Fig 7. Western blot analysis demonstrating phosphorylation of CREB after both GM-CSF and IL-3 stimulation. TF-1 cells (4 x 10’) were cultured in GM-CSF and IL-3 before growth factor and serum starvation for 24 hours. Cells were treated with diluent (PBS 0.2% BSA) or both GM-CSF (1 nmol/L) and IL-3 (1 nmollL) for 10 minutes and lysed with boiling SDS.Total cell lysate was loaded onto a 10% SDS-polyacrylamide gel and transferred onto nitrocellulose. The blot was probed with anti-CREB antibody (AI or antiphosphorylated CREB antibody (B). No phosphorylated CREB was observed in the diluent lane (B).

+

sult of a more unstable form of phosphorylated CREB or increased phosphatase activity in GM-CSF- compared with PIXY321-cultured cells. Specifc proteins recognize the SRE in unsrimulated or PIXY32I-stimulated TF-I cells. Because the SRE is necessary for maximal induction of - 1 16CAT, we performed EMSAS to determine whether specific proteins recognized the SRE in TF-I extracts (Fig 8) The SRE probe used in the gel

lane:

1 2 3 4 5

6 7 8 910

4-

U -

-PIXY321

t PIXY321

Fig 8. EMSA with the SRE probe and unstimulated or PIXY-stimulated TF-1 cell extracts. TF-1 cells (5 x 10’) cultured in GM-CSF before growth factor and serum starvation were stimulated with diluent (0.02% BSA in PBS) or PIXY321 (200 pmol/L) for 10 minutes. Cells were lysed by sonication. One nanogram of the SRE probe (SRE from nts -90 t o -110 of the egr-1 promoter) was incubated with 8 p L of cell lysate, 2 p g poly(didcl:(didc), and 2 pL of gel shift buffer. Lanes 2 and 7 contain 50-fold excess of cold SRE, lanes 3 and 8 contain 500fold excess cold SRE, lane 1 contains no competitor, lanes 4 and 9 contain 50-fold nonspecific competitor, and lanes 5 and 10 contain 500-fold nonspecific competitor. The arrows represent specific bands.



shift experiments included nts between -90 and -110 of the egr-1 promoter (see Materials and Methods). In both unstimulated and PIXY321-stimulated nuclear extracts, two specific complexes were observed. These results are consistent with previous data obtained using GM-CSF- and IL3-stimulated TF- 1 cell extracts (K. Sakamoto, unpublished results, September 1994). The patterns of specific complexes in the unstimulated and stimulated cell extracts were identical, suggesting that protein-protein interactions and posttranslational modification are critical in regulating transcriptional activity. In addition, there were no differences in SRE complexes observed between TF-1 cells cultured in both GM-CSF and IL-3 or PIXY321 before growth factor or serum starvation (data not shown). DISCUSSION

PIXY321 was developed to potentiate the known biological activities of GM-CSF and IL-3. Previous studies demonstrated that PIXY32 1 has greater activity than GM-CSF and IL-3 alone or in c ~ m b i n a t i o n .We ~ , ~performed transient transfections, EMSAs, and Western analyses in TF- 1 cells to determine whether the effect of PIXY321 could be explained by different molecular pathways than those of GMCSF or IL-3. We examined the regulatory regions of the primary response gene, egr- 1, to identify the transcription factor binding sites responsive to PIXY321 stimulation. Transient transfections with the various egr- 1 promoter constructs stimulated with PIXY321 revealed a different pattern than with GM-CSF, IL-3, or the two growth factors combined. Maximal induction was observed with either the -600 nt or the - 116 nt constructs, whereas the -480 and -235 nt constructs had decreased activity. This pattern is different from experiments we performed previously with GM-CSF or IL-3. '* GM-CSF stimulation produced sixfold induction of the -600 nt construct and 2.5- to 3.5-fold induction with the -480, -235, - 180, and - 116 nt constructs. IL-3 stimulation produced similar induction (3.2- to 3.7-fold) of all five constructs. These results suggest that GM-CSF, IL-3, and PIXY321 may use different elements of the egr1 promoter to activate transcription of the egr- 1 gene. Several SRE-like sequences are contained within the -480 to -235 nt region of the egr- 1 promoter. Our transfection experiments demonstrated that these SREs may not play a significant role during PIXY321 signal transduction, as in the case with GMCSF or IL-3. The significant increase in fold stimulation between the - 180 and - 116 nt constructs might be explained by deletion of the CREl (Fig 1). This CREl may be competing for the CREB protein, so that there is relatively poor interaction of CRE2 with CREB and, hence, the transcriptional machinery. Our results indicate both similarities and differences in egr-1 induction between PIXY321, GM-CSF, and IL-3. As we reported previously, both the CRE and the SRE are necessary for maximal induction of the egr-1 promoter after stimulation with GM-CSF, IL-3, and PIXY321. However, we found that the - 116 nt construct containing the mutant SRE has greater activity in response to PIXY321 than to GMCSF or IL-3. This suggests that different factors interact with CREB or that CREB may be modified differently in response

853

to PIXY321. Furthermore, we show here that the SRE-CRE sequence within the - 116 nt region is sufficient to activate the egr-1 promoter in response to PIXY321. The study of nerve growth factor activation of another early response gene, c-fos, also demonstrated that both the SRE and the CRE were required for induction.20 There have not been previous reports, however, of the SRE-CRE sequence being sufficient for primary response gene activation by hematopoietic growth factors. PIXY321 stimulation may result in a more stable interaction of CREB with the transcription initiation factor complex in PIXY321-grown cells. This complex may activate transcription of egr-l through the SRE-CRE sequence in PIXY32 1 -stimulated cells. In contrast, transcriptional activation of egr-1 by GM-CSF and IL-3 may require sequences surrounding the SRE-CRE within the - 116 nt region to stabilize the transcription initiation factor complex. This would explain why the SRE-CRE sequence alone is not sufficient for initiation of transcription of egr- 1 by GM-CSF and IL-3. Alternately, different posttranslational modification (ie, phosphorylation) of SRE-binding proteins or CREB might result from PIXY321 stimulation compared with GMCSF or IL-3. The EMSA results with the SRE probe (Fig 6) are consistent with this hypothesis. An alternative hypothesis is that the specific complex of proteins that associate with the SRE-CRE sites may differ between these three cytokines. Others have demonstrated that multiple proteins are capable of binding to SRE-like sequences.2' In addition, proteins binding to the CRE may heterodimerize with other nuclear proteins such as ATFl and ATF2.22-24Our results (Fig 4) suggest that multiple proteins bind the CRE as a complex. We are presently characterizing the proteins within these complexes in GM-CSF- or PIXY321-treated cells. Finally, the unique properties of PIXY321 may also result from the ability to form a different receptor complex involving a novel 100-kD protein (Linda Park, personal communication, July 1995), which in turn results in the activation of altemate signal transduction pathways and, hence, other immediate early genes. The EMSA and Western blots demonstrated that although CREB constitutively binds the CRE, phosphorylated CREB was observed after 10 minutes of PIXY321 stimulation in cells previously cultured in GM-CSF. These results are identical to those observed with GM-CSF, IL-3, or both GMCSF and IL-3 stimulation.'' In contrast, EMSA and Western blots with extracts from TF-I cells grown in PIXY321 before factor starvation demonstrated low levels of phosphorylated CREB in unstimulated extracts. The difference in the timecourse of CREB phosphorylation suggests that PIXY321 activates signaling pathways that lead to constitutive activation of CREB. PIXY321 may prime cells by altering cell cycle kinetics or by activating specific kinases or phosphatases. Studies to define the mechanism of constitutive phosphorylation of CREB in PIXY321-cultured cells are currently in progress. The data presented here suggest that subtle differences in protein complexes within the nucleus may contribute to unique biological effects of PIXY321. Studies to define the molecular action of PIXY321 will provide further under-


MIGNACCA ET AL

854

standing of signaling pathways activated by growth factors that control myeloid cell proliferation. ACKNOWLEDGMENT We thank Gayle Baldwin and Judith Gasson for their helpful suggestions, Anne 0’Shea-Greenfield for technical assistance, and Wendy Aft for preparation of the manuscript. We also thank Linda Park, Immunex, for providing the PIXY321 and for critical reading of the manuscript. REFERENCES I . Gerhartz HH, Engelhard M, Meusers P, Brittinger G, Wilmanns W, Schlimok G, Mueller P, Huhn D, Musch R, Siegert W, Gerhartz D, Hartlapp JH, Thiel E, Huber C, Peschl C, Spann W, Emmerich B, Schadek C, Westerhausen M, Pees H-W, Radtke H, Engert A, Terhardt E, Schick H, Binder T, Fuchs R, Hasford J, Brandmaier R, Stem AC, Jones TC, Ehrlich HJ, Stein H, Parwaresch M, Tiemann M, Lennert K: Randomized, double-blind, placebocontrolled, phase 111 study of recombinant human granulocyte-macrophage colony-stimulating factor as adjunct to induction treatment of high-grade malignant non-Hodgkin’s lymphomas. Blood 82:2329, 1993 2. Rowe JM, Andersen J, Mazza JJ, Paietta E, Bennett JM, Hayes A, Oette D, Wiemik PH: Phase I11 randomized placebo-controlled study of granulocyte-macrophage colony-stimulating factor (GMCSF) in adult patients (55-70 years) with acute myelogenous leukemia (AML). A study of the Eastern Cooperative Oncology Group (ECOG). Blood 82:329a, 1993 (suppl 1) 3. Ganser A, Lindemann A, Seipelt G, Ottman OG, Herrmann F, Eder M, Frisch J, Schulz G, Mertelsmann R, Hoelzer D: Effects of recombinant human interleukin-3 in patients with normal hematopoiesis and in patients with bone marrow failure. Blood 76:666, 1990 4. Ganser A, Lindemann A, Seipelt G, Ottman OG, Herrmann F, Eder M, Frisch J, Schulz G , Mertelsmann R, Hoelzer D: Clinical effects of recombinant human interleukin-3. Am J Clin Oncol 14:S51, 1991 5. Williams DE, Park LS: Hematopoietic effects of a granulocytemacrophage colony-stimulating factodinterleukin-3 fusion protein. Cancer 67:2705, 1991 6. Curtis BM, Williams DE, Broxmeyer HE, Dunn J, Farrah T, Jeffery E, Clevenger W, DeRoos P, Martin U, Friend D, Craig V, Gayle R, Price V, Cosman D, March C, Park LS: Enhanced hematopoietic activity of a human granulocytehacrophage colonystimulating factor-interleukin 3 fusion protein. Proc Natl Acad Sci USA 88:5809, 1991 7. Vadhan-Raj S, Papadopoulos NE, Burgess MA, Linke KA, Patel SR, Hays C, Arcenas A, Plager C, Kudelka AP, Hittelman WN, Broxmeyer HE, Williams DE, Garrison L, Benjamin RS: Effects of PIXY321, a granulocyte-macrophage colony stimulating factor interleukin-3 fusion protein, on chemotherapy-induced multilineage myelosuppression in patients with sarcoma. J Clin Oncol 12:715, 1994 8. Runowicz CD, Mandeli J, Speyer J, Wadler S, Cohen C, Hochster H, Bruckner H, Goldberg G, Smith H, Garrison L, Wallach R, Tomaino C, Demakos E, Sorich J, Holland JF: Phase VI1 study of PIXY321 in combination with cyclophosphamide (CTX) and car-

boplatin (CP) in the treatment of patients (PTS) with ovarian cancer (OC). Proc Am SOCClin Oncol 12:260, 1993 (abstr) 9. Vose JM, Anderson J, Bierman PJ, Garrison L, Appelbaum F, Anderson JR, Bishop M, Larsen K, Armitage JO: Initial trial of PIXY321 (GM-CSFAL-3 fusion protein) following high-dose chemotherapy and autologous bone marrow transplantation (ABMT) for lymphoid malignancy. Proc Am SOCClin Oncol 12:366, 1993 (abstr) 10. Vadhan-Raj S, Jeha S, Broxmeyer HE, Hittelman WN. Thompson P, Arcenas A, Hoots WK, Estey E, Garrison L: Stimulation of hematopoiesis by PIXY32 I (GM-CSFAL-3 fusion protein) in patients with bone marrow failure. Blood 82:366a, 1993 (suppl I) I 1. Sakamoto KM, Mignacca RC, Gasson JC: Signal transduction by granulocyte-macrophage colony-stimulating factor and interleukin-3 receptors. Recept Channels 2: 175, 1994 12. Sakamoto KM, Fraser JK, Lee H-JJ, Lehman E, Gasson JC: Granulocyte-macrophage colony-stimulating factor and interleukin3 signaling pathways converge on the CREB-binding site in the human egr-1 promoter. Mol Cell Biol 14:5975, 1994 13. Lee H-JJ, Mignacca RC, Sakamoto KM: Transcriptional activation of Egr- 1 by granulocyte-macrophage colony-stimulating factor but not interleukin-3 requires phosphorylation of cAMP response element-binding protein (CREB) on serine 133. J Biol Chem 270:15979, 1995 14. Kitamura T, Tange T, Terasawa T, Chiba S, Kuwaki T, Miyagawa K, Piao Y-F, Miyazono K, Urabe A, Takaku F: Establishment and characterization of a unique human cell line that proliferates dependently on GM-CSF, IL-3, or erythropoietin. J Cell Physiol 140:323, 1989 15. Sakamoto KM, Nimer SD, Rosenblatt JD, Gasson JC: HTLVI and HTLV-I1 Tax frans-activate the human EGR-I promoter through different cis-acting sequences. Oncogene 7:2 125, 1992 16. Peterlin M, Luciw P, Barr P, Walker M: Elevated levels of mRNA can account for the trans-activation of human immunodeficiency virus. Proc Natl Acad Sci USA 83:9734, 1986 17. Gorman CM, Moffat LF, Howard BH: Recombinant genomes which express chloramphenicol acetyltransferase in mammalian cells. Mol Cell Biol 2:1044, 1982 18. Rosenthal N: Identification of regulatory elements of cloned genes with functional assays. Methods Enzymol 152:704, 1987 19. Ginty DD, Komhauser JM, Thompson MA, Bading H, Mayo KE, Takahashi JS, Greenberg ME: Regulation of CREB phosphorylation in the suprachiasmatic nucleus by light and a circadian clock. Science 260:238, 1993 20. Ginty DD, Bonni A, Greenberg ME: Nerve growth factor activates a Ras-dependent protein kinase that stimulates c-fos transcription via phosphorylation of CREB. Cell 77:713, 1994 21. Treisman R: The serum response element. Trends Biochem Sci 17:423, 1992 22. Karin M, Smeal T: Control of transcription factors by signal transduction pathways: The beginning of the end. Trends Biochem Sci 17:418, 1992 23. Lee KAW: Transcriptional regulation by CAMP. Curr Opin Cell Biol 3:953, 1991 24. Montminy MR, Gonzalez GA, Yamamoto KK: Characteristics of the cAMP response unit. Recent h o g Hormone Res 46:219, 1990


1996 88: 848-854

Mechanism of transcriptional activation of the immediate early gene Egr- 1 in response to PIXY321 RC Mignacca, HJ Lee, EM Kwon and KM Sakamoto

Updated information and services can be found at: http://www.bloodjournal.org/content/88/3/848.full.html Articles on similar topics can be found in the following Blood collections Information about reproducing this article in parts or in its entirety may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://www.bloodjournal.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.