Transcriptional regulation of c-Fes in myeloid leukemia cells

ELSEVIER

Biochi~ic~a et BiophysicaActa Biochimica et Biophysica Acta 1306 (1996) 179-186

Transcriptional regulation of c-Fes in myeloid leukemia cells Yufang He 1, Flavia Borellini 2, Walter H. Koch 3, Kai-Xing Huang 4, Robert I. Glazer * Department of Pharmacology amt Lombardi Cancer Center, Georgetown Unil'ersiO' Medical Center, 3900 Reservoir Road, NW, Washington, DC 20007, USA Received 31 July 1995; revised 26 December 1995; accepted 27 December 1995

Abstract

The c-Fes proto-oncogene encodes a myeloid-specific protein-tyrosine kinase that is expressed preferentially in differentiated myeloid cells, but not in early myeloblast progenitor cells. To examine the basis for the phenotypic expression of c-Fes, the transcription initiation sites of the human c-Fes gene were mapped in myeloid leukemia cells and regulatory elements in the genomic c-Fes sequence were characterized. Two major transcription initiation sites were found in the myeloid leukemia cell line THP-I which delineated exon 1 to be 72-83 bp. When the activity of the CAT reporter gene under the control of the c-Fes promoter region, untranslated exon 1 and intron 1 was measured in TF-1, K562 and MCF-7 cells, only TF-1 cells exhibited chloramphenicol acetyltransferase activity. In contrast, all cell lines supported reporter gene activity when intron 1 was deleted. Deletion analyses revealed a negative regulatory region in intron l, which was localized by Southwestern analysis and DNA footprinting to a 14 bp region. This negative regulatory region suppressed reporter CAT activity in K562 and TF-I cells when inserted downstream to the SV40 early promoter. These results suggest that the tissue-specific expression of c-Fes may result, in part, from the negative regulation of transcription in myeloid and nonmyeloid cells. Keywords: c-Fes; Promoter region; Transcription initiation; Negative regulatory element; TF-I cell; K562 cell; Myeloid leukemia cell

1. Introduction c-Fes is the cellular homolog of the transforming oncogene found in the avian (v-fps) and feline (v-fes) sarcoma retroviruses [1-4]. c-Fes expression is restricted to normal [5,6] and primary leukemia [7-9] cells of granulocytic and monocytic lineage as well as in myeloid leukemia cell lines of a more mature pheaotype [10,11]. Previous studies by our laboratory [12-17] have shown that c-Fes m R N A expression and protein-tyrosine kinase activity are limited to cells which undergo myeloid differentiation and that

Abbreviations: CAT, chloramphenicol acetyltransferase. * Corresponding author. Fax: 1 (202) 6878324; e-mail: glazerr @gunet.georgetown.edu. i Present address: National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Building 10, Room 8N252, Bethesda, MD 20892, USA. 2 Present address: Microbiological Associates, 9900 Blackwell Rd., Rockville, MD 20850, USA. 3 Present address: Food and [)rug Administration, Washington, DC, USA. 4 Present address: National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD 20892, USA. 0167-4781/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S 0 1 6 7 - 4 7 8 1 ( 9 6 ) 0 0 0 0 5 - K

stable expression of c-Fes in K562 myeloblast progenitor cells induces the functional characteristics of mature granulocytic cells [17,18]. Little is known about the transcriptional regulation of c-Fes and the mechanisms that govern its tissue- and phenotype-specific pattern of expression. Transgenic animals carrying multiple copies of the 13.2 kb genomic c-Fes fragment expressed c-Fes only in the bone marrow which paralleled the expression pattern of the endogenous gene [19]. These results suggest that the c-Fes sequence is a self-contained genetic element with all the promoter and enhancer regions that are required for conferring myeloid tissue specificity. Despite these data, there is no direct evidence to suggest that the c-Fes promoter region controls its tissue-specific expression. Previous studies showed that promoter sequences immediately upstream to exon 1 were active in driving reporter gene expression in nonmyeloid cells [20], and that monkey kidney fibroblasts (Cos-1 cells) transfected with the genomic c-Fes sequence did not express c-Fes m R N A [21]. In this report, the transcriptional regulatory sequences that control the expression of c-Fes were examined in myeloid and nonmyeloid cells of different c-Fes phenotypes. Our results identify promoter sequences upstream to

180

Y.He et al. / Biochimica et BiophysicaActa 1306 (1996) 179-186

exon 1 that are preferentially active in c-Fes-expressing cells and a negative regulatory region within intron 1 that is active only in myeloid cells.

2. Materials and methods 2.1. Cell culture

Human chronic myelogenous leukemia cell line K562, human monocytic leukemia cell line THP-1 and human breast carcinoma cell line MCF-7 were obtained from the American Type Culture Collection, Rockville, MD. Human erythroleukemia cell line TF-1 [22] was generously provided by Dr. Toshio Kitamura, DNAX, Inc. THP-1 and TF-1 cells express c-Fes mRNA, whereas K562 and MCF-7 do not [12]. K562, THP-1 and TF-1 cells were maintained at a density of 10s-106 cells/ml in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 40 mM Hepes (pH 7.4), 2 mM glutamine and 50 /~g/ml gentamicin. TF-1 cells were also supplemented with 5 n g / m l granulocyte/macrophage colony-stimulating factor. MCF-7 cells were grown in Iscove's DMEM supplemented with 10% heat-inactivated fetal bovine serum, 40 mM Hepes (pH 7.4), 2 mM glutamine and 50 /zg/ml gentamicin.

mega). Five deletion constructs (see Fig. 2) were prepared by digestion with HindIII and ApaI ( [ - 2 1 3 / + 574]CAT), HindIII and XmaIII ( [ - 131 / + 574]CAT), AccI and PstI ([-425/+483]CAT), AccI and C~'nI ( [ - 4 2 5 / + 245]CAT), and AccI and KpnI ( [ - 4 2 5 / + 9 1 ] C A T ) . Three deletion constructs were derived from plasmid [ - 4 2 5 / + 91]CAT by digestion with HindIII and ApaI ( [ - 2 1 3 / + 9 1 ] C A T ) , HindIII and XmaIII ( [ - 1 3 1 / + 574]CAT), and HindIII and SstII ( [ - 5 / + 91]CAT). Plasmids [ + 4 3 3 / + 4 8 3 ] O B C A T 5 and [ + 4 8 3 / + 433]OBCAT6 were constructed by first amplifying sequence + 4 3 3 / + 483 from [ - 4 2 5 / + 574]CAT by PCR and subcloning the fragment into the AccI site of the pCAT-Promoter. The plasmid was subsequently digested with BamHI and SalI to remove sequence + 4 3 3 / + 483 and subcloned into the polylinker site of pOBCAT5 or pOBCAT6 (kindly provided by Dr. Carl Baker, National Cancer Institute, NIH [25]), which contained the polylinker in opposite orientations (see Fig. 5). The polylinker in these vectors is contained within a splice acceptor and splice donor site downstream to the SV40 early promoter. Plasmids [ + 4 4 1 / + 454]OBCAT5 and [ + 4 5 4 / + 441]OBCAT6 were constructed by PCR amplification of sequence + 4 3 3 / + 483 with primers containing BamHI and SalI restriction sites. 2.4. Transcription initiation mapping

2.2. Preparation of nuclear extracts

Nuclear extracts were prepared by the procedure described previously [23] except that 1 mM phenylmethylsulfonyl fluoride, 50 p,g/ml aprotinin, 200 p,g/ml leupeptin, 400 /xg/ml soybean trypsin inhibitor and 10 /zM pepstatin were added to all extraction solutions. 2.3. Plasmid constructs

The genomic c-Fes 5' sequence, - 4 2 5 / + 5 7 4 was amplified by PCR from plasmid pECE/fes [18] which contained the 13.2 kb EcoRI genomic c-Fes sequence [4,24]. The forward primer, 5'-TAAAGCTTTCTCTTCTGCGGTGG-3' and the reverse primer, 5'-TAGTCGACAGAAAAGGGGCCATC-3' contained HindIII and AccI restriction sites, respectively, to facilitate subcloning. Amplification was carried out for 30 cycles in the presence of 10 mM Tris-HC1 (pH 7.5), 50 mM KCI, 1.5 mM MgCI 2, 0.1% gelatin, 200 /zM dATP, dTTP and dCTP, 150 /zM dcvGTP, 50 p.M dGTP, 50-100 pmol of each primer and 2.5 units of native Taq polymerase (Perkin Elmer) per 100 p.l reaction. The sequence of the amplified fragment was confirmed by sequencing with Sequenase (USB) and was found to agree with the published sequence [24]. The PCR-amplified c-Fes sequence ( - 4 2 5 / + 574) was subcloned into the HindIII and AccI polylinker sites in the promoterless reporter gene plasmid, pCAT-Basic (Pro-

Transcription initiation sites were mapped by RNase protection analysis. A riboprobe was constructed by subcloning an E c o R I / K p n I fragment ( - 4 2 5 / + 91) of the genomic c-Fes sequence into pGEM4Z (Promega) to obtain plasmid pFes4ZEK [9]. pFes4ZEK was linearized at the 5' EcoRI site and in vitro transcription was carried out as described previously [9]. RNA was isolated from THP-1 and MCF-7 cells with guanidinium thiocyanate-CsC1 [9]. RNA protection analysis (RPAII kit, Ambion) was carried out with 20 p.g of total RNA. RNA was dissolved in 20/zl of RNA hybridization buffer containing 2 X 10 6 dpm of riboprobe, and after hybridization for 4 h at 42°C, samples were digested by the addition of 200 /xl of a solution containing 5 units RNase A and 20 units RNase T~ for 30 rain at 37°C. The protected double-stranded RNA fragments were resolved on a denaturing 8% polyacrylamide/urea gel and visualized by autoradiography of the dried gel. 2.5. Transient expression assay

TF-1 cells at a concentration of 5 X 10 7 cells per 0.5 ml of RPMI 1640 medium containing 10 n g / m l granulocyte/macrophage colony-stimulating factor were transfected by electroporation with a Bio-Rad Gene Pulser set at 960 /zF and 300 V. Cells were then incubated for 15 min on ice, diluted to 30 ml with RPMI 1640 medium containing 10% heat-inactivated fetal bovine serum and 10

Y. He et al. / Biochimica et Biophysica Acta 1306 (1996) 179-186

n g / m l granulocyte/macrol?hage colony-stimulating factor and incubated at 37°C for :!4 h. K562 cells at a concentration of 107 cells per 0.25 ml RPMI 1640 medium were transfected by electroporation at 960 /zF and 250 V. Cells were then incubated at room temperature for 10 min, diluted to 25 ml with RPMI 1640 medium containing 10% fetal bovine serum and incubated at 37°C for 24 h. MCF-7 cells were transfected by lipofection (Lipofectin, BRL) as described previously [26]. All test reporter gene plasmids were cotransfected with 5 /xg of pSVlacZ (Promega) to normalize for transfection efficiency. Cells were lysed by sonication in 250 mM Tris-HC1 (pH 7.8) and CAT activity in whole cell extracts was determined by a modifcation of the method of Nordeen et al. [27]. CAT activity was measured in a reaction mixture containing: 0.015 units of acetylCoA synthetase (Sigma), 0.5 mM sodium acetate, 0.15 mM CoA, 3.75 mM ATP, 0.2 tzCi [3H]chloramphenicol (10 Ci/mmol, Amersham) and 0.05 mM chloramphenicol. The reaction was preincubated for 5 min at 37°C to generate acetylCoA, and then extracts from transfected and mock-transfected cells were added and incubated at 37°C for 20 h. Reaction kinetics were linear with both protein concentration and time over 20 h. [3H]Acetylchloramphenicol was extracted with xylene, back-extracted with water to remove trace amounts of [3H]chloramphenicol, and quantitated by scintillation counting. Transfection efficiencies were normalized for /3-galactosidase activity th~Ltwas assayed spectrophotometrically with o-nitrophenyl-fl-o-galactopyranoside as substrate.

181

2.6. DNase footprinting and Southwestern blotting DNase footprinting was carried out by digesting [ + 4 3 3 / + 4 8 3 ] C A T in the polylinker with XbaI, endlabeling with T4 polynucleotide kinase and digesting with HindlII. Varying amounts of nuclear extract from K562 cells were incubated with 1 ng of probe and digested with DNase (Stratagene) as described previously [23]. Digested samples were separated in a 6% polyacrylamide-urea gel and detected by autoradiography. The sequence of the protected region was determined by sequencing utilizing an end-labeled primer complementary to the HindlII site in the polylinker sequence of [ + 4 3 3 / + 483]CAT. Southwestem blotting was carried out with doublestranded probes containing sequence + 2 4 6 / + 407 or a tandem repeat of sequence + 4 3 3 / + 483 that were endlabeled with T4 polynucleotide kinase [23].

3. Results 3.1. Transcription initiation sites of the human c-Fes gene RNase protection analysis was carried out to determine the transcription initiation sites in the c-Fes genomic sequence (Fig. 1A). Total cellular RNA was hybridized to an antisense riboprobe that encompassed the EcoRI-digested sequence upstream to intron 1. RNA from THP-1 cells, which express high levels of c-Fes [12], indicated multiple transcription initiation sites which existed in two clusters

A B +iiiii!+ii!iii~i~i!i+ii;/

..73 '-72

-425 -365 -305 -245 -185 -125 -65 -5 +56 +116 +176 +236 +296 +356 +416 +476 +536

aattccgtgaggtggggagggctgggaccagggttccctctttctcttctgcggtggccc tggcctggtgctaggactgcgcgcctcccctcagtacccgcggacaccctgggcttccct gggcccagcatctgcctggggcctcgcctgggctccccctcctgacccccaccttg=gcc ccttcccggtgttcccggggcgctgccgggccctggggcctgcggggcgcgggcggctct tggctgggccattctttcccggccccctcctcccttccgtttccgtggccgtgcggcc~g ctagaggctgcggcccagcgcggagcaggggggctggcaggcgtcgggacggtcgggccg gtcccgcccgccccttcccctccacaggcccgccccggggcctgggccaactgaaaccgc gggag~AGGAAGCGC~AATCAGGAACTGGCCGGGGTCCGCACCCGGGCCTGAGTCGGTC CGAGGCCGTCCCAGGAGCAGCTGCCCGTGCGGgtacctctagcccCggggcctggaggag cggtgggagctgggggcgcggcaggcaggggcagagcaggcgttccgagggccagagacc cacccaggtcggggtaggggccgcggaagggcggggatggccgcaggggcagggctcagg ctgtgggcgcctgaggcttcagctggggcaggcttggcctgtcgaggacctgggcagggt gtccctgcaaggggtggtgggtggaagggcctggggagggaggctccaggttggctcctg cttcccgaacgtgcggaggagaccctgacgctaaggaagcaatgagggccagtccccagg ccaggctgctgctgggtacccatggctgcgtgtgagcgaggcaggaccccacctcctccc cgtctgcagtccatcctgaccctacagtccccagtctcctcgtcccatgcctccgtctcc agctgctgccttgcctccagggatggccccttttctgtccccagAACAGCACT~=~GG..

Fig. 1. Identification of the transcriptional initiation sites of the c-Fes gene by RNase protection analysis. A: Total RNA (20 /zg) from THP-1 and MCF-7 cells was hybridized to an antisense [nP]riboprobe containing the c-Fes genomic sequence - 4 2 5 / + 91, and digested with RNase A and T I. The digestion products were resolved on an 8% polyacrylamide-urea sequencing gel and the protected fragments are indicated by arrows. The numbers 72, 73 and 83 indicate the number of nucleotides upstream to the known 3' boundary of exon 1 where transcription begins. B: c-Fes genomic sequence. Exon 1 and the beginning of exon 2 are indicated by upper case letters. The transcription initiation sites in exon 1 and the translation start codon in exon 2 are underlined.

Y. He et al. / Biochimica et Biophysica Acta 1306 (1996) 179-186

182

72-83 bp upstream to the known 3' boundary of exon 1 [24], whereas no hybridization signal was detected with RNA from MCF-7 cells (Fig. 1). Fig. 1B presents a genomic map of c-Fes with the first transcription initiation site as + 1. This sequence indicates that the promoter region is 425 bp and that the largest exon 1 is 83 bp. The translation start site is + 589 in exon 2.

[ - 4 2 5 / + 245]CAT and [ - 4 2 5 / + 483]CAT showed the most striking differences in reporter gene activity. By deletion of intron sequence + 246 to + 483, CAT activity was increased 13- to 14-fold in both TF-1 and K562 cells. In summary, these data suggested the presence of a strong promoter within sequence - 131 / + 91 which exhibits a c-Fes myeloid phenotype specificity, and a negative regulatory region within sequence + 2 4 6 / + 483 of intron 1.

3.2. Analysis of the genomic c-Fes 5' sequence To characterize the cis regulatory sequences in the c-Fes sequence, several deletion constructs were linked to the promoterless CAT reporter gene and transiently expressed in TF-I, K562 and MCF-7 cells (Fig. 2). The entire c-Fes sequence - 4 2 5 / + 574 exhibited promoter activity in TF-I cells but not in K562 or MCF-7 cells, a result consistent with the c-Fes-expressing myeloid phenotype. Constructs containing deletions to - 213 and - 131 were also preferentially active in TF-1 cells, and these constructs which also lacked intron 1 exhibited increased CAT activity with a similar cell line specificity. It was also noted that the promoter activity of sequence - 4 2 5 / + 91 in TF-I cells was comparable to the activity of the SV40 early promoter (data not shown), and was the only construct exhibiting activity in MCF-7 cells. Plasmids

3.3. Characterization of the negative regulatory region in intron 1 Nuclear extracts from K562 and TF-1 cells were next examined for DNA-binding activity to sequence + 2 4 6 / + 483. To eliminate interference from an AP-2 site at +410 to +417, DNA-binding activity was measured with sequences + 2 4 6 / + 407 and + 4 3 3 / + 483 which flanked the AP-2 binding site. Southwestern blotting with sequence + 2 4 6 / + 407 did not detect DNA-binding activity in any of the cell lines, whereas sequence + 4 3 3 / + 483 detected proteins with molecular weights of 105, 68, 50 and 40 kDa in K562 and MCF-7 cells and 105 and 40 kDa in TF-1 cells (Fig. 3). DNase footprinting was then carried out with nuclear extracts from K562 cells and sequence + 4 3 3 / + 483 (Fig.

u n i t s C A T / u n i t 8 G a l x 10 "a ~ o Exon 1 m

. . . .

[-425/+ 5741CAT

~o

O 0

0

!

. . . .

i

,=

1-2131 + 5741CAT

BB-

,,,,

[-131 / + 574]CAT

I

=,=

1-4251+

4831CAT

,.,,

[425/+

2451CAT

[-425/+ 91 ]CAT

,=

[-213/+911CAT

j

[-131/+ 91]CAT

..Jm

0 0

. . . .

!

0 0 . . . .

!

B~

m

~ o

0 0

TF.1 K562 MCF-7

m

[-5/+ 911CAT

Fig. 2. Reporter gene activity in TF- 1, K562 and MCF-7 cells. Cells were transfected with the CAT reporter gene constructs and CAT activity was assayed as described in Section 2. CAT plasmids were cotransfected with pSVlacZ, and CAT activity was normalized to /3-galactosidase activity to correct for transfection efficiency. Each value is the mean + standard error of four separate experiments or the mean of two experiments.

183

Y. He et al. / Biochimica et Biophysica Acta 1306 (1996) 179-186

Sequence +247/+407

Sequence +433/+483

4). These results revealed two protected regions from + 4 4 1 to + 4 5 4 ( C T G C G T G T G A G C G A ) and from + 4 5 9 to + 4 7 7 ( G G A C C C C A C C T C C T C C C C G ) . The cis regulatory activity of sequences + 4 3 3 / + 483 and + 4 4 1 / + 454 were next assessed b y reporter gene activity in a p l a s m i d c o n t a i n i n g the heterologous SV40 early promoter (Fig. 5). These sequences were inserted into p l a s m i d p O B C A T 5 or p O B C A T 6 [25] in the sense or antisense orientation, respectively, d o w n s t r e a m to the SV40 early promoter and within a polylinker site flanked by a splice donor and acceptor site. Sequences + 4 4 1 / + 4 5 4 and + 4 3 3 / + 483 inhibited C A T activity by 50% in K562 cells, but did not affect C A T activity in the antisense orientation, suggesting that these negative regulatory sequences function in an orientation-dependent manner.

ex

ii!iiiiiiiiiiiiiii!iiiiiiiiill ~i

!,i:iii i iiiiiiiiiiiiiiii,:ii

i~i~~ii~ii!i~!ii!i~ii~i~~i!~i!~i!i ~ ~.... ~~ii!~

~;~ ~i~i~ / : ~

~

~i~iiiiiii!!!iiiiiiii~iiii!!i!ii!iiiii!ill ~~

68 kDa~ !:!

iii!i

43 kDa-

4. Discussion The remarkable feature of the h u m a n c - F e s gene is that its tissue- and phenotype-specific expression is contained within a 1 kb sequence upstream to exon 2, the first translated exon. The present study is the first to identify regulatory regions in the c - F e s promoter and intron 1

Fig. 3. Southwestern analysis 3f intron 1. Probes corresponding to + 246/+ 407 or + 433/+ 483 were used for analysis of nuclear extracts from K562, MCF-7 and TF-1 cells. Each lane contained either 100 p.g (K562 and MCF-7 cells) or 50 tag (TF-1 cells) of protein.

K562 7.5

15

+478

+459 +454

+441

acccatggctgcgtgtgagcgaggcaggaccccacctcctccccgt ct tgggtaccqacqcacact cqctccgt cctaqqataaaaqaqqqqcaga +441

+454

+459

+478

Fig. 4. DNase footprinting of intron 1. The DNase I protection pattern is shown for the non-coding strand of sequence + 433/+ 483 incubated with 0, 7.5 and 15 /zg of nuclear extract fro~a K562 cells. The protected regions are indicated by boxes with the corresponding sequence. Sequence + 4 3 3 / + 483 is indicated below the figure and the protected regions are underlined.

184

Y. He et al. / Biochimica et Biophysica Acta 1306 (1996) 179-186

which may account, in part, for its phenotypic expression in myeloid cells. RNase protection analysis revealed two major transcription initiation regions in THP-1 mRNA which revealed exon 1 to be 72 to 83 bp. Exon 1 was originally estimated to be 200 bp [24], but a later study using primer extension analysis found exon 1 to be only 14 bp in HL-60 cell mRNA [20]. The presence of multiple initiation sites was not unexpected since the c-Fes genomic sequence does not

A

contain either a TATA box or an initiator (Inr) sequence [28,29]. The disparity between previous findings and our results may be attributed to the cell lines used as well as differences in methodology. The RNase protection technique used in the present study is not subject to interference by hairpin loop structures in contrast to the primer extension method used previously [20]. c-Fes expression appears to be regulated by phenotype-specific factors that are present in more differ-

,AF

pOBCAT5

+433//~ +441/

+483

\~ +454

P~

.A=

pOBCAT6

+ 48 3 / ~ 433 + 454/ ~ + 441

B 125 100 . ~

. ~

o

< i-<

75

//

50

G)

25

rr

Ctl

+441/+454 +433/+483 +4541+441 +483/+433

Fig. 5. Negative regulation of the heterologous SV40 early promoter by sequence + 4 4 1 / + 454. A: SV40CAT constructs. The sense and antisense orientations of sequences + 4 3 3 / + 483 and + 4 4 1 / + 454 were cloned into pOBCAT5 and pOBCAT6, respectively, downstream to the SV40 early promoter. The SV40 early promoter and polyadenylation sites are identified by PE and A E, respectively, and splice donor and acceptor sites are identified by ( O ) and ( v ) , respectively. B: Reporter gene activity in K562 cells. Activity was normalized for /3-galactosidase activity after cotransfection with pSVlacZ. CAT activity is normalized to the activity of pOBCAT5 or pOBCAT6 which were equal to 100%. Each value is the mean + range of two experiments.

185

Y. He et al. / Biochimica et Biophysica Acta 1306 (1996) 179-186 E

SB

SSBSPexl

Ell

$

A

ex2

j\ CTGCGTGTGAGCGA Fes LD8 7 u IL- 3 consensus

TGCGTGTGAGCGAG AGCATGACAGCATC AGCATGGCAGCAGG GC TG AGC

Fig. 6. Organization of the genoraic c-Fes 5' sequence. The c-Fes negative regulator 3' sequence + 4 4 1 / + 454 is compared with the negative regulatory elements in the LD78ot and IL-3 promoters [38]. The following consensus elements and their location are: E, ETF ( - 154, - 155, - 5 3 ) ; S, Spl ( - 195, -57, -30, -52, +204); P, PU.1 (-3); Eg, EGR-1 ( + 125); A, AP2 (+409). ex, exon.

entiated myeloid cells. Reporter gene expression revealed that promoter sequence - - 1 3 1 / + 91 was active preferentially in TF-1 cells. This sequence contains three CCCTTCC motifs for ETF [30], a transcription factor which is conserved in the promoter of several myeloid genes [31] including C D l l b [32], and three elements for Spl, a transcription factor which is activated in K562 cells overexpressing the genomic c-Fes sequence [23] (Fig. 6). The promoter also contains a motif at the juncture of the promoter and exon 1 for transcription factor PU.1, a member of the ets gene family [33]. The PU. 1 motif exists in a similar context in the C D l l b promoter [32] and is found in the noncoding region of exon 2 in the macrophage colony-stimulating factor gene [34]. Both PU.1 and ETF are important for myeloid phenotype specificity and developmental regulation of the integrin 1 la gene [35]. ETF also binds with comparable affinity to GC boxes recognized by Spl [36] and stimulates transcription from promoters which lack a TATA box [37]. Since the TATA-Iess c-Fes promoter contains three ETF and Spl elements in close proximity to the PU.1 motif, factors which bind to these motifs may play a role in regulating c-Fes expression in myeloid cells. Previous studies have shown that the c-Fes promoter sequence - 4 2 5 / + 87 could activate reporter gene expression in non-myeloid cells [20], and that the entire 13.2 kb genomic c-Fes sequence 'was not activated in Cos-1 cells [21]. Our data are consistent with these results in that sequence - 4 2 5 / + 574 was inactive in K562 and MCF-7 cells, but that the intronless sequence - 4 2 5 / + 91 was a strong promoter in all cell lines examined. These data suggest that tissue-specificity may be controlled by the functional interaction of :factors binding to promoter and intron regions, and that sequences in intron 1 probably exert a strong negative regulatory influence on c-Fes expression. Our data further indicate that an orientation-dependent negative effect is exerted by factors which bind to sequence + 4 4 1 / + 454. It is of interest that this sequence shares homology with negative regulatory elements present in the promoter region of the LD87ct and 1L-3 cytokine

genes (Fig. 6). The nuclear inhibitory protein, NIP (also called ICK-1) which binds to these negative regulatory elements is present in nuclear extracts from K562 and Jurkat cells [38]. It will be of interest to determine if NIP also plays a role in regulating the myeloid expression of c-Fes.

Acknowledgements This work was supported by Grant IR01CA54231 from the NCI, NIH to R.I.G.

References [1] Groffen, J., Heisterkamp, N., Grosveld, F., Van de Ven, WJ.M. and Stephenson, J.R. (1982) Science, 216, 1136-1138. [2] Groffen, J., Heisterkamp, N., Shibuya, M., Hanafusa, H. and Stephenson, J.R. (1983) Virology, 125, 480-486. [3] Hampe, A., Laprevotte, I., Galibert, F., Fedele, L.A. and Sherr, C.S. (1982) Mol. Cell. Biol., 12, 981-990. [4] Trus, M.D., Sodroski, J.G. and Haseltine, W.A. (1982) J. Biol. Chem. 257, 2730-2733. [5] Emilia, G., Donelli, A., Ferrari, S., Torelli, U., Selleri, L., Zucchini, P., Moretti, L., Venturelli, D., Ceccherelli, G. and Torelli, G. (1986) Br. J. Haematol. 6, 287-292. [6] Ferrari, S., Torelli, U., Selleri, L., DoneUi, A., Venturelli, D., Moretti, L. and Torelli, G. (1985) Br. J. Haematol. 59, 21-25. [7] Mavilio, F., Sposi, N.M., Petrini, M., Bottero, L., Marinucci, M., De Rossi, G., Amadori, S., Mandelli, F. and Peschle, C. (1986) Proc. Natl. Acad. Sci. USA 83, 4394-4398. [8] Slamon, D.J., de Kernion, J.B., Verma, I.M. and Cline, M.J. (1984) Science 224, 256-262. [9] Smithgall, T.E., Johnston, J.B., Bustin, M.and Glazer, R.I. (1991) J. Natl. Cancer Inst. 83, 42-46. [10] Feldman, R.A., Gabrilove, J.L., Tam, J.P., Moore, M.A.S. and Hanafusa, H. (1985) Proc. Natl. Acad. Sci. USA 82, 2379-2383. [11] MacDonald, L., Levy, J. and Pawson. T, (1985) Mol. Cell. Biol. 5, 2543-2551. [12] Glazer, R.I., Smithgall, T.E., Yu, G. and Borellini, F. (1991) Adv. Regul. Cell Growth 2, 41-60. [13] Smithgall, T.E., Yu, G. and Glazer, R.I. (1988) J. Biol. Chem. 263, 15050-15055.

186

Y. He et al. / Biochimica et Biophysica Acta 1306 (1996) 179-186

[14] Smithgall, T.E., Goswami, B.B., Nagashfar, Z., Ahmad, S. and Glazer, R.I. (1992) Biochemistry 31, 4828-4833. [15] Yu, G. and Glazer, R.I. (1987) J. Biol. Chem. 262, 17543-17548. [16] Yu, G., Grant, S. and Glazer, R.I. (1988) Mol. Pharmacol. 33, 384-388. [17] Fang, F., Ahmad, S., Lei, J., Klecker, R.W., Trepel, J.B., Smithgall, T.E. and Glazer, R.I. (1993) Biochemistry 32, 6995-7001. [18] Yu, G., Smithgall, T.E. and Glazer, R.I. (1989) J. Biol. Chem. 264, 10276-10281. [19] Greer, P., Maltby, V., Rossant, J., Bernstein, A. and Pawson, T. (1990) Mol. Cell. Biol. 10, 2521-2517. [20] Alcalay, M., Antolini, F., Van de Ven, W.J., Lanfrancone, L., Grignani, F. and Pelicci, P.G. (1990) Oncogene 5, 267-275. [21] Greer, P.A., Meckling-Hansen, K. and Pawson, T. (1988) Mol. Cell. Biol. 8, 578-587. [22] Kitamura, T., Tange, T., Terasawa, T., Chiba, S., Kuwaki, T., Miyagawa, K., Piao, Y.-F., Miyazono, K., Urabe, A. and Takaku, F. (1989) J. Cell. Physiol. 140, 323-334. [23] Borellini, F., He, Y., Aquino, A., Yu, G., Josephs, S.F. and Glazer, R.I. (1991) J. Biol. Chem. 266, 15850-15854. [24] Roebroek, A.J.M., Achalken, J.A., Bussemakers, M.J.G., van Heerikhuizen, H., Onnekink, C., Debruyne, F.M.J., Bloemers, H.P.J. and Van de Ven, W.J.M. (1986) Mol. Biol. Rep. 11, 117-125. [25] Baker, C.C. (1990) in Recombinant Systems in Protein Expression (Alitalo, K.K., Huhtala, M.L., Knowles, J. and Vaheri, A., eds.) pp. 75-86, Elsevier, New York.

[26] Ahmad, S., Trepel, J.B., Ohno, S., Suzuki, K., Tsuruo, T. and Glazer, R.I. (1992) Mol, Pharmacol. 42, 1004-1009. [27] Nordeen, S.K., Green, P.P.III. and Fowlkes, D.M. (1987) DNA 6, 173-178. [28] Ayer, D.E. and Dynan, W.S. (1988) Mol. Cell. Biol. 8, 2021-2033. [29] Smale, S.T. and Baltimore, D. (1989) Cell 57, 103-113. [30] Kageyama, R., Merlino, G.T. and Pastan, I. (1989) J. Biol. Chem. 264, 15508-15514. [31] Shapiro, L.H., Ashmun, R.A., Roberts, W.M. and Look, A.T. (1991) J. Biol. Chem. 266, 11999-12007. [32] Pahl, H.L., Scheibe, R.J,, Zhang, D.-E., Chen, H.-M., Galson, D.L., Maki, R.A. and Tenen, D.G. (1993) J. Biol. Chem. 268, 5014-5020. [33] Klemsz, M.J., McKercher, S.R., Celada, A., Van Beveren, C. and Maki, R.A. (1990)Cell 61, 113-124. [34] Zhang, D.-E., Hetherington, C.J., Chen, H.-M. and Tenen, D.G. (1994) Mol. Cell. Biol. 14, 373-381. [35] Shelley, C.S., Farkhzad, O.C. and Arnaout, M.A. (1993) Proc. Natl. Acad. Sci. USA 90, 5364-5368. [36] Kageyama, R., Merlino, G.T. and Pastan, I. (1988) Proc. Natl. Acad. Sci. USA 85, 5016-5020. [37] Chen, H.-M., Pahl, H.L., Scheibe, R,J., Zhang, D.-E. and Tenen, D.G. (1993) J. Biol. Chem. 268, 8230-8239. [38] Nomiyama, H., Hieshima, K., Hirokawa, K., Hattori, T., Takatsuki, K. and Miura, R. (1993) Mol. Cell. Biol. 13, 2787-2801.