Vol. 5, 259-265, March 1994 Cell Growth & Differentiation 259

Activation of Early Growth Response 1 GeneTranscription and pp90rsk during Inductionof Monocytic Differentiation1

Surender Kharbanda,2 A. Saleem, Masanori Hirano,Yutaka Emoto, Vikas Sukhatme, John Blenis, andDonald KufeDivision of Cancer Pharmacology, Dana-Farber Cancer Institute,Boston, Massachusetts 02115 IS. K., A. S., M. H., Y. E., D. K.J; Division ofNephrology, Beth Israel Hospital, Boston, Massachusetts 02215

IV. 5.1; and Department of Cellular and Molecular Physiology, Harvard

Medical School, Boston, Massachusetts 021 1 5 Ii. B.I

Abstrad

The present work has studied mechanisms responsiblefor indudion of early growth response 1 (EGR-1) geneexpression during monocytic differentiation of U-937myeloid leukemia cells. Differentiation of U-937cells with 1 2-O-tetradecanoylphorbol-1 3-acetate (IrA),an adivator of the serine/threonine protein kinase C,was associated with transcriptional adivation of EGR-1promoter-reporter construds. The EGR-1 promotercontains six CC(A/T)6GG (CArG) motifs. The two

5’-most distal CArG sequences conferred TPAinducibility. In contrast, there was little effed of TPA onEGR-1 transcription in a TPA-resistant U-937 cellvariant, designated TUR. Treatment of both U-937 andTUR cells with okadaic acid, an inhibitor of serine/threonine protein phosphatases 1 and 2A, was associatedwith indudion of monocytic differentiation and EGR-1transcription through the 5’-most CArG element.Since these findings supported the involvement ofserine/threonine protein phosphorylation in theregulation of EGR-1 expression, we studied adivation ofthe 405 ribosomal protein 56 serine/threonine kinases,pp7O� and pp9O�. Although both kinases participatein regulating cell growth, there was no detedableadivation of pp7O� during TPA- or okadaicacid-induced monocytic differentiation. Moreover,rapamycin, an inhibitor of pp7O� adivation, had noeffed on indudion of EGR-1 expression. In contrast,analysis of pp9O� adivity by phosphorylation of apeptide derived from 56 protein demonstratedstimulation of this kinase in TPA-treated U-937, and notTUR, cells. Okadaic acid treatment of both cell typeswas associated with adivation of pp9O�. Since serumresponse factor interacts with CArG elements and isdiredly phosphorylated by pp9O�, the present findingssuggest that adivation of the pp9O�’ cascade maycontribute to indudion of EGR-1 transcription andappearance of the differentiated monocytic phenotype.

lntrodudion

Previous work has demonstrated that treatment of humanmyeloid leukemia cell lines with TPA3 is associated withinduction of differentiation along the monocytic lineage(1-3). TPA activates the serine/threonine-specific PKC (4),and thus certain insights are available regarding early signalsin this process. Other studies have demonstrated that TPA-induced differentiation of U-937 leukemia cells is accom-panied by increased expression ofthe early growth response1 (EGR-l , zif/268, NGF-IA, Krox24, TlS-8) gene (5-1 0). Thefinding that this gene is activated by other agents which in-duce monocytic differentiation of both leukemic cells andnormal monocytes suggested that EGR-l may function intransducing early nuclear signals into longer term changesin gene expression associated with the monocytic pheno-type (1 1-1 3). Indeed, recent studies have suggested thatEGR-l is required for differentiation along the monocyticlineage (1 4). EGR-l expression, however, is not restricted tomyeloid cells. Neuronal differentiation of PCi 2 rat pheo-chromocytoma cells by nerve growth factor and cardiac dif-ferentiation of the pluripotent embryonal carcinomaP19518O1A1 line are associated with increases in EGR-ltranscripts (1 0). Moreover, the EGR-l gene is induced duringmitogenic stimulation of diverse cell types, including fibro-blasts, epithelial cells, and lymphocytes (1 5, 1 6). Althoughthese findings have indicated that EGR-1 is involved in bothgrowth and differentiation, few insights are available regard-ing the signaling mechanisms responsible for induction ofthis gene.

Recent attention has been directed toward defining cy-toplasmic and nuclear signaling pathways that contribute tothe regulation of immediate-early genes. This effort has inpart been focused on MAP kinases. Several serine/threonineprotein kinases have been implicated in the control of earlyresponse gene expression on the basis of their spatial dis-tribution and temporal regulation. These include the 405ribosomal protein S6 kinases, designated pp9O�� andpp7oSbK (1 7-20). Activation of both pp9O�’ and pp7#{216}SbK

involves serine/threonine phosphorylation but occursthrough distinct signaling pathways (1 8, 20-22). Althoughthe kinases responsible for activation of pp7OS6K remain un-known, the MAP kinases pp44/42maPk/erk appear to regulatepp9O�’ activity (23, 24). Activated MAP kinases have beenshown to phosphorylate transcription factors such as Jun andFos in vitro (25, 26). Moreover, regulation of nuclear MAPkinase and pp9orsk activities by growth factors is coordi-nated with induction of certain early response genes (27).These findings have suggested that MAP kinases and pp9#{216}r�

Received 9/23/93; revised 1 1/1 7/93; accepted 1 2/1 5/93.1 This investigation was supported by USPHS Grant CA42802 awarded by the

National Cancer Institute, Department of Health and Human Services.2 To whom requests for reprints should be addressed, at Division of Cancer

Pharmacology, Dana-Farber Cancer Institute, Harvard Medical School, Bos-

ton, MA 02115.

3 The abbreviations used are: TPA, 1 2-O-tetradecanoylphorbol-1 3-acetate;PKC, protein kinase C; EGR, early growth response; MAP, mitogen-activatedprotein; CArG, CC(AIF)6GG; SRF, serum responsefactor; CAT, chlorampheni-

col acetyltransferase; SRE, serum response element; PBS, phosphate-bufferedsaline; kd, kilodalton(s); kb, kilobase(s); SDS, sodium dodecyl sulfate.

OAcv)’ I

0) .C .C .C .C .C

I i-Cv) CO 04�

� Ii_28SEGR-1

TPAA. �I ‘-

28S- ,� � . �

18S-

TPAB. F- � �

� .C .C� 1� Cv)

28S-�,

OA

EGR-1

260 Induction of EGR-1 Gene Transcription

18S-

are involved in transducing growth-regulatory signals fromthe cell membrane to the nucleus. The potential role of thissignaling pathway in the induction ofcellular differentiation,however, is unclear.

The present studies demonstrate that activation of theEGR-1 gene during TPA-induced monocytic differentiationis mediated through serum response or CArG elements. Pre-vious work has demonstrated that pp901� directly phospho-rylates SRF (28). Since this phosphorylation of SRF enhancesthe affinity and rate with which it interacts with the CArGbox (28), we asked whether pp9orsk is activated duringmonocytic differentiation. Using both TPA and okadaicacid, an inhibitor of serine/threonine protein phosphatasesPP-i and PP-2A, we demonstrate that activation of nuclearpp90� is associated with induction of EGR-1 and thedifferentiated monocytic phenotype.

Results

Previous work has demonstrated that treatment of humanmyeloid leukemia cells is associated with increased EGR-1mRNA levels (6). Similar findings were obtained in thepresent studies with U-937 cells (Fig. 1A). Although the3.4-kb EGR-1 transcripts were at low to undetectable levelsin untreated U-937 cells, exposure to 32 nsa TPA resulted inincreased EGR-1 expression, which was detectable at 1 hand maximal at 3-6 h (Fig. 1 A). Exposure of U-937 cells to40 ng/ml okadaic acid was also associated with a transientinduction of EGR-1 expression, which was maximal at 3-6h (Fig. 1 A). These findings were compared with the effectsofTPA and okadaic acid on a U-937 cell variant, designatedTUR, which was selected for resistance to TPA-inducedgrowth inhibition and other characteristics, such as adher-ence, ofthe differentiated monocytic phenotype. In contrastto the effects of TPA on U-937 cells, induction of EGR-lexpression was attenuated during TPA treatment ofthe TURclone (Fig. 1 B). However, the finding that TUR cells respondto okadaic acid with increases in EGR-l mRNA levels

�1

.C .C .C .C .CD�- #{216}�CsJ �l-� ‘-04

- 1 8 S Fig. 1. Induction o EGR-1 expres-sion by okadaic acid and TPA.U-937 (A) and TUR (B) cells were

treated with 32 n� TPA or 40 nWmlokadaic acid (OA) for the indicatedtimes. Total cellular RNA (20 pg) washybridized to the 32P-labeled EGR-1probe. Hybridization to the actin

probe demonstrated equal loadingof the lanes.

indicated that certain signaling pathways associated withinduction of this gene are intact in the variant (Fig. 1 8).

In order to determine the mechanisms responsible forTPA-induced EGR-l mRNA levels, we used a construct con-taming the EGR-1 promoter region (positions -957 to + 248)ligated to the CAT reporter gene (pEGR-l P1 .2). This region

includes several potential regulatory elements with two Jun/AP-1 binding sites and six CArG boxes (Fig. 2A). Exposureof pEGR-1 P1 .2-transfected U-937 cells to TPA was associ-ated with a 4.5 ± 1 .7-fold (mean ± SD of three determi-nations) increase in CAT activity compared with transfectedcells not exposed to this agent (Fig. 28). In contrast, therewas no detectable induction ofthe minimal pE7O-CAT con-struct (Fig. 28). Similar studies were performed with pE425(positions -425 to +65; six CArG boxes without the twoAP-l sites). pE425 was somewhat more responsive to TPA

than pEGR-l P1 .2, with a 7.2 ± 2.1 -fold (mean ± SD of threedeterminations) increase in TPA-induced CAT activity com-pared with that obtained with pE425 transfected, but un-treated, cells (Fig. 28). Transfection of pE395, which con-tains the five proximal CArG boxes, also conferred TPAinducibility, whereas deletion ofthe next CArG box (pE359)resulted in little if any TPA-induced CAT activity (Fig. 28).

These findings indicated that the region encompassing thefirst two CArG elements mediates TPA-induced expressionof the EGR-l gene.

Similar transient expression studies were performed withTUR cells. However, there was no detectable inducibihity ofthe EGR-1 promoter-CAT constructs during TPA treatment ofthese cells. For example, percentage conversion with pE425was 0.16 and 0.19 in control and TPA-induced TUR cells,respectively. In contrast, these cells responded to okadaicacid with activation of EGR-1 transcription. Transfection ofthe pE425 construct demonstrated okadaic acid responsive-ness with a 4.4 ± 1 .1 -fold increase in CAT activity, whereaslittle inducibihity by okadaic acid was observed with bothpE395 and pE359 (Fig. 3). Other studies with U-937 cells

-.500 -250 0 +248I as I

A. -iooo -750

-957

pEGR-1 P1 .2 1-

B.

-II II +1liii II �

pE7O - ‘

pE425

pE395

pE359

C

pEGR-1P1.2

TPA

pE7O

C TPA

pE425

C TPA

pE395

C TPA

pE359

C TPA

#{149}#{149}

.

#{149}�sI

.�

#{149}#{149}.f #{149}.

% Conversion 0.21 0.97 0.11 0.14 0.80 5.70 O.161.20 0.13 0.17

.

0�

% Conversion 0.72 3.19 0.11 0.15 0.16 0.12

Fig. 3. Okadaic acid-induced EGR-1 transcription is conferred by the 5’-

most CArG element. TUR cells were transfected with the constructs shownin Fig. 2A and maintained in medium for 36 h. Okadaic acid (OA( was thenadded, and the cells were harvested after 9 h for analysis ofCAT activity. The

percentage conversion of chloramphenicol to the acetylated forms as deter-mine(I I)y scintittation counting is shown for control (C) and okadaic acid

(OA(-treated cells. The relative fold increase (mean ± SD) in CAT activity for

control versus OA-treated cells was calculated as: pE425 In = 3), 4.4 ± 1.1;

pE395 (n = 3(, 1 .4 ± 0.33; pE359 (n = 3), 0.76 ± 0.11.

Cell Growth & Differentiation 261

Fit,’. 2. TPA-induced EGR-1 tran-

scription is regulated by CArG ele-nients. A, u-q37 cells were trans-fected with the indicated constructs

and maintained in medium for 36 h.TPA was then added, and the cellswere harvested after 9 h for analysis

ofCAT activity. Hatchedbars, AP-1sites; solid bars, CArG boxes. B,

conversion of chloramphenicol tothe acetylated tornis is shown for

control (C( and TPA-treated cellstransfected with the indicated con-structs. Percentage conversion wasdetermi ned by scintillation count-ing. The relative fold increase

(mean � SD) in CAT activity for

control versus TPA-treated cellswas calculated as: pEGR-1 P1 .2

(n = 3(, 4.5 ± 1 .7; pE7O (n = 2),1.24 � 0.16; pE425 In = 3), 7.2 ±

2.1; pE395 (n - 3(, 5.7 ± 2.1;pE359 (n = 3(, 1 .31 ± 0.22.

pE 425 pE 395 pE 359

C OA C OA C OA

.. .#{149} ..#{149}- . o

have demonstrated that the first CArG element linked to theheterologous herpes simplex virus thymidine kinase mini-mal promoter and CATgene is able to confer okadaic acidresponsiveness, whereas there was no detectable inducibil-ity with a similar heterologous promoter construct contain-ing a mutated CC(A/V)6GG sequence (1 3). Similar findingswere obtained in okadaic acid-treated TUR cells (data notshown). Taken together, these findings indicated that the first

CArG element mediates okadaic acid-induced expression ofthe EGR-l gene in the wild-type U-937 and variant TURcells.

One of the early responses to mitogen-induced cellulargrowth isthe phosphorylation ofserine/threonine residues incertain translational components such as the ribosomal 56protein (29). Although activation of pp70S6K results in thephosphorylation of 56 (30), it is not known whether othersubstrates exist for this kinase. Since previous studies havedemonstrated that TPA activates pp70S6K in quiescent 3T3cells (31), we used immunoblot analysis to study pp7056K inlysates from U-937 cells. The results demonstrate little if any

change in pp70S6�< levels or electrophoretic mobility follow-ing treatment with TPA (Fig. 4A). Similar findings were ob-tamed during exposure to okadaic acid (data not shown).There was also no detectable change in pp7OS6K duringtreatment of TUR cells with TPA or okadaic acid (data notshown). Recent studies have shown that the macrolide ra-pamycin specifically blocks signaling by pp70S6K (3 1-33).Consequently, we asked whether rapamycin inhibits eventsassociated with induction of U-937 cell differentiation. Thisagent had no detectable effect on TPA-induced growth arrestor adherence of U-937 cells (data not shown). Moreover,rapamycin had little effect on TPA- or okadaic acid-inducedEGR-l expression (Fig. 48). These findings and the absenceof pp7oS6K activation indicated that pp70S6K signaling is notessential for induction of the EGR-l gene and other char-acteristics of monocytic differentiation.

Previous studies have demonstrated that pp70S6K andpp9O#{176}� are differentially regulated by a variety of mitogens

A.

0.4

#{149}03

Fraction Number

A.�-. TPAC’) jia,..

I 00

- �

69-�f’ pp7OS6K

.(

< Q.

B. � 20) 0. < Q.1< <Q.<

�o � i- �

28S-

�, .4 �4 .4 EGR-118S-

Fin. 4. Activation of pp7OSt�K is not associated with TPA- or okadaic acid-induced monocytic differentiation. A, U-937 cells were treated with TPA for

the indicated times. Soluble proteins (75 pg( were subjected to electrophoresisin 7.5/, SDS-polyacrylamide gels and transferred to nitrocellulose paper. The

filters were incubated with an anti�pp70S�K antiserum, and reactivity was

visualized by chemiluminescence. Position of the 69 kd molecular weightmarker is shown. B, U-937 cells were treated with TPA or okadaic acid (OA(for 6 h. Cells were also pretreated with 80 nsa rapamycin (RAP) for 3 h andthen exposed to TPA or OA. Total cellular RNA (20 pg( was hybridized to the

#{176}P-labeled EGR-1 probe. Hybridization to the actin probe demonstrated

equal loading of the lanes.

(1 8, 20, 31 , 34, 35). Consequently, we asked whether TPA-induced differentiation of U-937 cells is associated with ac-tivation of pp9orsk. These studies were performed by pun-fying pp90rsk and then assaying for phosphorylation of 56peptide. Extracts of U-937 cells treated with TPA were par-tially purified by Q Sepharose Fast Flow chromatography.Increases in 56 peptide phosphorylation were detectable at30 mm of TPA treatment (data not shown). This activity wasfurther purified by absorption to Mono Q anion exchangebeads and then elution with a linear NaCI gradient. TheTPA-induced pp90rsk activity was detectable in a peak thateluted with 0.1 3 M NaCI (Fig. 5A). In contrast, there was littleif any 56 peptide phosphorylation when using protein pu-rified from uninduced U-937 cells (Fig. 5A). Immunoblotanalysis of the Mono Q eluate with the anti-pp90�� anti-serum revealed a major species at approximately 90 kilo-daltons in the active fractions (Fig. 58). Similarfindings wereobtained in okadaic acid-treated U-937 cells. Treatmentwith okadaic acid was associated with an increase in pp9orsk

activity compared to that in control cells (Fig. 6). In contrast,assays of ppgorsk activation in TUR cells revealed a differentpattern. Treatment of TUR cells with TPA had no detectableeffect on pp9orsk phosphotransferase activity, whereas phos-phorylation of 56 peptide was stimulated after adding oka-daic acid (Fig. 7). These findings and the results obtainedwith U-937 cells indicated that pp9orsk activation is asso-ciated with induction of EGR-l expression and monocyticdifferentiation.

Discussion

The present studies have used U-937 myeloid leukemia cellsand a TPA-resistant U-937 variant to study signaling eventsassociated with induction of monocytic differentiation. The

C

0

- _�I>a’.-�� 2

� 0.2� I-I

� -�o_� 0�1%�

(0(I)

B.

kD � � 6 7 8 9 10111 1 6- . �..._ �,

94---.

69-

Fig. 5. Effects of TPA on pp9Oask activity in U-937 cells. U-937 cells weretreated with TPA for 30 mm. A, the cells were lysed, and the soluble fractionwas applied to a Q Sepharose column. Protein eluting at 0.2 si NaCI was

adjusted to 0.02 M NaCI and then applied to a Mono Q column. Fractions wereassayed for phosphorylation of 56 peptide after elution with a 0.0-0.5 xi NaCIlinear gradient. EL control; #{149},TPA-treated. B, fractions from the Mono Qcolumn were subjected to electrophoresis in 7.5% SDS-polyacrylamide gels

and transferred to nitrocellulose filters. The filters were incubated with ananti-pp90’�’ antiserum, and reactivity was visualized by a peroxidase detec-

tion system.

C

.2 1.4

�T�E�:i.o2.Eo.Eca-..O.6 -

� 2a.a.� 0.2Co

C,) 4

Fraction Number

Fig. 6. Effects ofokadaic acid on pp9O’�’ activity in U-937 cells. U-937 cellswere treated with okadaic acid for 3 h. Soluble proteins (Li,control; 0, OA-

treated) were separated by Q Sepharose and then Mono Q column chroma-tography. The fractions were assayed for phosphorylation of S6 peptide.

TUR clone was isolated from a single U-937 cell selected forresistance to the growth-inhibitory effects of 5 flM TPA. Thisvariant fails to exhibit growth arrest, adherence, or increasesin a-naphthylacetate esterase staining, all characteristics ofthe differentiated monocytic phenotype, when exposed toTPA at concentrations as high as 500 nsi (36). Other work hasdemonstrated that monocytic differentiation of U-937 cellsis also associated with increased expression of the EGR-1gene (6). Similar findings were obtained with additional in-ducers of U-937 cell differentiation, including okadaic acid,and during activation of resting human peripheral blood

5 6 7 8 9 10 11

262 Induction of EGR-1 Gene Transcription

C0

� zI- m0’� 0

.C.- -

o.E I-’

��E I

0.a.Co(1)

Fig. 7. Activation of pp9O� in TUR cells. TUR cells were treated with TPAor okadaic acid for 30 mm and 3 h, respectively. Soluble proteins (0, control;

., TPA-treated; 0, OA-treated) were separated by Q Sepharose and thenMono Q column chromatography. The fractions were assayed for phospho-

rylation of 56 peptide.

Fraction Number

Cell Growth & Differentiation 263

monocytes (6, 1 3). More importantly, recent work has dem-onstrated that EGR-1 expression is required for induction ofmonocytic differentiation (1 4). In the present studies, TPAinduced little EGR-1 expression in TUR cells, whereas oka-daic acid treatment was associated with induction of thisgene. Okadaic acid is an inhibitor of serine/threonine pro-tein phosphatases (37-39) and in earlier studies was foundto induce monocytic differentiation of U-937 cells (40).Treatment of TUR cells with okadaic acid was also associ-

ated with growth arrest and the appearance of a differenti-ated monocytic phenotype (36). These findings indicatedthat signaling events associated with induction ofthe EGR-1

gene and monocytic differentiation are regulated at least inpart by serine/threonine protein phosphorylation.

Several serine/threonine protein kinases have been iden-tified that participate in the regulation of cell growth and

differentiation. TPA activates the serine/threonine-specificPKC (4), and thus this kinase functions in initial signals thatcontribute to monocytic differentiation. However, studieswith the TPA-resistant TUR clone suggest that other events,

potentially downstream of PKC, can be activated by okadaicacid and thereby induce a monocytic phenotype. Indeed,characterization of the TUR variant has demonstrated de-creased levels of PKC isoforms (36). Similar decreases in PKCexpression have been reported for TPA-resistant HL-60myeloid leukemia cells (41 , 42). The present studies havetherefore focused on two other senine/threonine protein ki-nases, pp70S6K and pp9O��, that phosphorylate the 40S ri-bosomal 56 protein. Previous work has shown thatmitogen-induced growth is associated with phosphoryla-tion of 56 protein (29). TPA-induced mitogenesis of quies-cent cells is associated with activation of pp7OSbK (31).Moreover, selective inhibition of pp7OSbK with rapamycinblocks interleukin 2-induced entry of T-cells into S phase(32). Of interest, TPA treatment of U-937 cells had no de-tectable effect on pp7056K. Furthermore, rapamycin, an in-hibitor of pp7056K, failed to block TPA-induced EGR-1 ex-pression and appearance of the monocytic phenotype. Thefinding that okadaic acid treatment of both U-937 andTUR cells has no effect on this kinase also supported thelack of involvement of pp7OS6K in monocytic differentia-tion. Thus, pp70S6�( would appear to be activated duringmitogenesis, but not differentiation of certain cell types.

Although pp9orsk also phosphorylates S6 protein, this ki-nase, in contrast to pp7056K, exhibits phosphotransferase

activity toward a variety of other substrates. For example,activated pp9O� and its homologues phosphorylate MAP2(31 ), c-fos (27), histone H3 (27), lamin C (43), and glycogensynthase (44). Previous studies have demonstrated that thepp9Or� signaling cascade transduces growth-regulatory sig-nals to the nucleus, whereas less is known about the role ofthis kinase in the induction of differentiation. The presentresults demonstrate that pp90�’ is also activated duringmonocytic differentiation. TPA treatment of U-937 cells, butnot the TUR variant, was associated with activation ofpp9O�� as determined by increased phosphorylation of 56peptide. Activation ofpp9O�’ was also found during okadaicacid-induced differentiation of both U-937 and TUR cells.Although the mechanisms responsible for this activation ofpp9#{216}rsk in these TPA- and okadaic acid-treated cells havenot been defined in the present studies, previous work hasshown that pp9orsk is activated by MAP kinase (23, 24).Moreover, MAP kinase and pp9O�� exhibit coordinate tem-poral regulation and spatial distribution following additionof growth factors to serum-deprived cells (27). Other studieshave demonstrated that MAP kinase is activated in TPA-induced U-937 cells(26), and we haveobtained similar find-ings in okadaic acid-treated U-937 and TUR cells (data not

shown). Stimulation of MAP kinase activity was detectableat low levels in TPA-treated TUR cells and was less pro-nounced than that obtained with okadaic acid or during in-duction of U-937 cells (data not shown). Thus, activation ofMAP kinase may be responsible for stimulation of ppg#{248}rsk

activity during both mitogenesis and induction of monocyticdifferentiation.

Certain insights are available regarding the relationshipbetween activation ofthe MAP kinase/pp90�’� signaling cas-cade and induction of early response gene expression. MAPkinase phosphorylates c-Jun and thereby may contribute toJun/AP-1 -mediated activation of c-jun gene transcription(26). pp9O�az phosphorylates a COOH-terminal serine in

c-Fos which is necessary for transcriptional transrepressionof the c-fos promoter (27, 45). This COOH-terminal regionof c-Fos is also required for down-regulation of EGR-1 genetranscription after mitogenic stimulation (46). The target forthis repression is the SRE or CArG box (46). pp9Orsk phos-phorylates SRF in vitro at the same site (serine 1 03) that isphosphorylated in vivo after growth factor stimulation (28).Since SRF also interacts with the SRE (47, 48), and phos-phorylation at serine 1 03 enhances the affinity of SRF bind-ing to the SRE (28), pp9orsk may coordinate the transcriptionof genes regulated by this element. Other studies have dem-onstrated that the SRE interacts with both SRF and p62TcF,

and that phosphorylation of p62TcF by MAP kinase stimu-lates formation of this ternary complex (49). The presentwork demonstrates that TPA-induced EGR-l gene transcrip-tion in U-937 cells is conferred by the region of the EGR-1promoter containing the first and second CArG elements. Incontrast, there was no detectable activation of EGR-lpromoter-reporter constructs in TPA-treated TUR cells.Moreover, as previously demonstrated in okadaic acid-treated U-937 cells (1 3), the present work shows thatthe firstCArG element confers okadaic acid inducibility of EGR-ltranscription in TUR cells.

The findings thus suggest that activation of the MAPkinase/pp90�� cascade may contribute to induction ofEGR-l transcription through phosphorylation of serum re-sponse factor and p62TcF. Direct support for this mechanismwould be provided by increased binding of these factors tothe proximal CArG sites in the EGR-l promoter. However,

264 Induction of EGR-1 Gene Transcription

we have been unable thus far to demonstrate a significantand reproducible shift in electrophoretic mobility patternsusing the proximal CArG box and nuclear proteins from in-duced U-937 cells. Finally, although MAP kinase/pp90rskactivation and EGR-l induction have been associated withmitogenesis, the present findings support the involvement ofthese signaling pathways in the differentiation of myeloidleukemia cells.

Materials and Methods

Cell Culture. U-937 cells (50) and a TPA-unresponsiveU-937 variant, designated TUR (36), were grown in RPMI1 640 medium containing 1 0% heat-inactivated fetal bovineserum supplemented with 1 00 units/mI penicillin, 1 00 �ig/mlstreptomycin, and 2 m�a i-glutamine. Cells were exposed to32 nsi TPA (Sigma Chemical Co., St. Louis, MO), 40 ng/mlokadaic acid (Moana Bioproducts, Honolulu, HI), and 80 nMrapamycin (Drug Synthesis and Chemistry Branch, NationalCancer Institute, Bethesda, MD).

RNA Isolation and Northern Blot Analysis. Total cellularRNA was isolated by the guanidine isothiocyanate-cesiumchloride method (51 ). The RNA (20 pig) was analyzed byelectrophoresis in 1% agarose-2.2 M formaldehyde gels andtransferred to nitrocellulose filters. Hybridizations were per-formed as previously described (6) with the following 32P-labeled DNA probes: (a) the 0.7-kb non-zinc finger insert ofa murine Egr-1 complementary DNA (1 0); and (b) the 2.0-kbPstl insert of a chicken g3-actin gene purified from the pAlplasmid (52).

Reporter Assays. The pEGR-1P1.2, pE425, pE395,pE359, and pE7O constructs were prepared as described(46). The vectors were transfected into cells using the DEAE-dextran technique (53). Briefly, 2 x 1 0� cells were harvestedand washed with PBS. The cells were resuspended in 1 mlof 20 m� Tris-HCI (pH 7.4) buffer containing 0.5 mg/mIDEAE-dextran, 8 pg plasmid, 1 40 mvi NaCI, 5 m�i KCI, 375

hiM Na2HPO47H2O, 1 mM MgCI2, and 675 �M CaCl2. Thecells were incubated at 37#{176}Cfor 45 mm, washed with mediacontaining 1 0% fetal bovine serum, resuspended in com-plete media, and then incubated at 37#{176}C.Thirty-six h aftertransfection, the cells were distributed equally into two ali-quots; one ahiquot served as a control, and the other wastreated with TPA or okadaic acid. The cells were harvestedafter 9 h and lysed by three cycles of freezing and thawingin 0.25 M Tris-HCI (pH 7.8) and 1 msa phenylmethylsulfonylfluoride. Equal amounts of the cell extracts were incubatedwith 0.025 pCi [14C]chloramphenicol, 0.1 5 si Tris-HCI (pH7.8), and 0.4 mM acetyl-CoA for 1 h at 37#{176}C.The enzymeassay was terminated by addition of ethyl acetate. The or-ganic layer containing the acetylated [i 4Clchloramphenicolwas separated by thin-layer chromatography using chloro-form-methanol (95%/5%; v/v). Following autoradiography,both acetylated and unacetylated forms of [‘4Cjchloram-phenicol were cut from the plates, and the conversion ofchloramphenicol to the acetylated form was calculated by

measurement of radioactivity in a a-scintillation counter.Subcellular Fradionation. Cells were washed twice with

ice-cold PBS and lysed in 1 ml of hypotonic lysis buffer (1mM ethyleneglycol bis(�-aminoethyl ether)-N,N,N’,N’-tetraacetic acid, 1 m� EDTA, 1 0 msi �3-gIycerophosphate, 1mM sodium orthovanadate, 2 ms� MgCI2, 1 0 msi KCI, 1 mt�idithiothreitol, 40 pg/mI phenylmethylsulfonyl fluoride, and10 pg/mI of both leupeptin and aprotinin). The cell suspen-sion was incubated on ice for 30 mm to allow swelling. AfterDounce homogenization of the cells (20-25 strokes, tight

pestle A), the lysate was loaded onto 1 ml of 1 M sucrose inlysis buffer and centrifuged at 1 ,600 X g for 1 5 mm. Thesupernatant above the sucrose gradient was again centri-fuged at 1 50,000 X gfor 30 mm. The resulting soluble frac-tion was adjusted to 0.5% Nonidet P-40, 0.1 % deoxy-chohate, and 0.1 % Brij 35 and used as the cytosolic fraction.

Immunoblot Analysis. Phosphorylation of pp7OS6K andpp9O�k were performed as described (1 8). Proteins wereseparated in 7.5% SDS-polyacrylamide gels and transferredto nitrocellulose paper. The residual binding sites were firstblocked by incubating thefilter in 5% dry milk in PBST (PBS-0.05% Tween 20) for 1 h at room temperature. The filterswere then incubated with either a rabbit anti�pp70S6K or arabbit anti-pp90�� antiserum for 1 h with shaking. Afterwashing three times with PBST, the blots were incubatedwith anti-rabbit lgG peroxidase conjugate (Sigma). Theantigen-antibody complexes were visualized by chemilu-minescence (ECL detection system; Amersham, ArlingtonHeights, IL).

Partial Purification of pp9O� Activity. The cytosolicfraction was applied to a Q Sepharose Fast Flow column(1 .0 ml; Pharmacia, Piscataway, NJ). After washing withbuffer A [20 mM Tris-HCI (pH 7.4), 0.5 mr�i ethyleneglycolbis(�3-aminoethyl ether)-N,N,N’,N’-tetraacetic acid, 0.5 mt�aEDTA, and 1 0 mM f3-mercaptoethanohl, proteins were elutedwith 3 ml of 0.2 M NaCI-buffer A, adjusted to 0.02 M NaCIwith buffer A, and applied to a Mono Q column (HR 5/5;Pharmacia). The column was washed with buffer A. Proteinwas eluted with a linear 0.0-0.5 M NaCI-buffer A gradient.Fractions were assayed for pp9O�� activity by incubation inthe presence of 50 �M S6 peptide (RRRLSSLRA), 20 �M

[‘y-32P]ATP, 10 m� MgCI2, and 0.4 m�i dithiothreitol. Afterincubation for 1 5 mm at 30#{176}C,25 p1 ofthe reaction mixturewere loaded onto phosphocellulose discs (GIBCO-BRL,Grand Island, NY). The discs were washed three times with1 % phosphoric acid and then with water before scintillationcounting.

References

1 . Huberman, E., and Callahan, M. P. Induction ofterminal differentiation ofhuman promyelocytic leukemia cells by tumor-promoting agents. Proc. NatI.Acad. Sci. USA, 76: 1293-1297, 1979.

2. Lotem, H., and Sachs, L. Regulation of normal differentiation in mouse and

human myeloid leukemic cells by phorbol esters and the mechanism of tumorproduction. Proc. NatI. Acad. Sci. USA, 76: 51 58-51 62, 1979.

3. Rovera, G., O’Brien, T. A., and Diamond, L. Induction of differentiation

in human promyelocytic leukemia cells by tumor promoters. Science (Wash-ington DC), 204: 868-870, 1979.

4. Nishizuka, Y. Intracellular signaling by hydrolysis of phospholipids and

activation of protein kinase C. Science (Washington DC), 258: 607-614,1992.

5. Christy, B. A., Lau, L. R., and Nathans, D. A gene activated in mouse 3T3

cells by serum growth factors encodes a protein with “zinc finger” sequences.Proc. NatI. Acad. Sci. USA, 85: 7857-7861, 1988.

6. Kharbanda, S., Nakamura, T., Stone, R., Hass, R., Bernstein, S., Sukhatme,V., and Kufe, D. Expression ofthe EGR-1 and EGR-2 zinc finger genes during

induction of monocytic differentiation. J. Clin. Invest., 88: 571-577, 1991.

7. Lemaire, P., Revelant, 0., Bravo, R., and Charney, P. Two mouse genes

encoding potential transcription factors with identical DNA-binding domainsare activated by growth factors in cultured cells. Proc. NatI. Acad. Sci. USA,

85: 4691-4695, 1988.

8. Lim, R. W., Varnum, B. C., and Herschman, H. R. Cloning of tetra-decanoylphorbol ester induced “primary response” sequences and theirexpression in density-arrested Swiss 3T3 cells and a TPA nonproliferativevariant. Oncogene, 1: 263-270, 1987.

9. Milbrandt, J. A nerve growth factor-induced gene encodes a possible tran-

scriptional regulatory factor. Science (Washington DC), 238: 797-799, 1987.

10. Sukhatme, V. P., Cao, X., Chang, L. C., Tsai-Morris, C-H., Stamenkovich,

D., Ferreira, P. C. P., Cohen, D. R., Edwards, S. A., Shows, T. B., Curran, T.,

Cell Growth & Differentiation 265

LeBeau, M. M., and Adamson, E. D. A zinc finger-encoding gene coregulatedwith c-fos during growth and differentiation, and after cellular depolarization.

Cell, 53: 37-43, 1988.

1 1 . Bernstein, S. H., Kharbanda, S. M., Sherman, M. 1., Sukhatme, V. P., andKufe, D. W. Posttranscriptional regulation of the zinc finger-encoding EGR-1

gene by granulocyte-macrophage colony-stimulating factor in human U-937monocytic leukemia cells: involvement of a pertussis toxin-sensitive G pro-tein. Cell Growth & Differ., 2: 273-278, 1 991.

1 2. )ingwen, L., Lacy, I. Sukhatme, V. P., and Coleman, D. L. Granulocyte-macrophage colony-stimulating factor induces transcriptional activation of

Egr-1 in murine peritoneal macrophages. J. Biol. Chem., 266: 5229-5933,1991.

1 3. Kharbanda, S., Rubin, E., Datta, R., Hass, R., Sukhatme, V., and Kufe, D.

Transcriptional regulation of the early growth response 1 gene in humanmyeloid leukemia cells by okadaic acid. Cell Growth & Differ., 4: 1 7-23,

1993.

14. Nguyen, H. Q., Hoffman-Liebermann, B., and Liebermann, D. A. The

zinc finger transcription factor Egr-1 is essential for and restricts differentiation

along the macrophage lineage. Cell, 72: 197-209, 1993.

1 5. Seyfert, V. L., Sukhatme, V. P., and Munroe, I. G. Differential expression

of a zinc finger-encoding gene in response to positive versus negative sig-naling through receptor immunoglobulin in murine B lymphocytes. Mol. Cell.

Biol., 9: 2983-2088, 1989.

1 6. Sukhatme, V. P., Kartha, S., Toback, F. G., Taub, R., Hoover, R. G., and

Tsai-Morris, C-H. A novel early growth response gene rapidly induced byfibroblast, epithelial cell and lymphocyte mitogens. Oncogene Res., 1: 343-355, 1987.

1 7. Banerjee, P., Ahmad, M. F., Grove, J. R., Kozlosky, C., Price, D. J., andAvruch, J. Molecular structure of a major insulin/mitogen-activated 70-kDa

S6 protein kinase. Proc. NatI. Acad. Sci. USA, 87: 8550-8554, 1990.

1 8. Chen, R-H., and Blenis, J. Identification of Xenopus 56 protein kinasehomologs (pp9O�’) in somatic cells: phosphorylation and activation during

initiation of cell proliferation. Mol. Cell. Biol., 10: 3204-3215, 1990.

19. Kozma, S. C., Ferrari, S., Bassand, P., Siegmann, M., Tony, N., andThomas, G. Cloning of the mitogen-activated S6 kinase from rat liver reveals

an enzyme of the second messenger subfamily. Proc. NatI. Acad. Sci. USA,

87: 7365-7369, 1990.

20. Sweet, L. J., Alcorta, D. A., Jones, S. W., Erikson, E., and Erikson, R. L.Identification of mitogen-responsive ribosomal protein 56 kinase pp90�, ahomolog of Xenopus 56 kinase II, in chicken embryo fibroblasts. Mol. Cell.

Biol., 10:2413-2417, 1990.

21. BalIou, L. M., Luther, H., andThomas,G. MAP2 kinaseand 70k S6 kinase

lie on distinct signalling pathways. Nature (Lond.), 349: 348-350, 1990.

22. Chen, R-H., Chung, I., and Blenis, J. Regulation of pp90rsk phosphoryl-ation and 56 phosphotransferase activity in Swiss 3T3 cells by growth factor-,phorbol ester- and cyclic AMP-mediated signal transduction. Mol. Cell. Biol.,11:1861-1867, 1991.

23. Chung. J., Pelech, S., and Blenis, J. Mitogen-activated Swiss mouse 3T3

RSK kinase I and II are related to pp44mPk from sea staroocytes and participate

in the regulation of pp90ask activity. Proc. NatI. Acad. Sci. USA, 88: 4981-4985, 1991.

24. Sturgill, T. W., Ray, L. B., Erikson, E., and Mailer, J. L. Insulin-stimulatedMAP-2 kinase phosphorylates and activates ribosomal protein 56 kinase I.

Nature ILond.), 334: 715-718, 1988.

25. Alvarez, E., Northwood, I. C., Gonzalez, F. A., Latour, D. A., Seth, A.,Abate, C., Curran, T., and Davis, R. I. Pro-Leu-Serflhr-Pro is a consensusprimary sequence for substrate protein phosphorylation. J. Biol. Chem., 266:15277-15285, 1991.

26. Pulverer, B. I., Kyriakis, I. M., Avruch, J., Nikolakaki, E., and Woodgett,J- R. Phosphorylation ofc-Jun mediated by MAP kinases. Nature (Lond.), 353:

670-674, 1991.

27. Chen, R-H., Sarnecki, C., and Blenis, J. Nuclear localization and regu-lation oferk- and rsk-encoded protein kinases. Mol. Cell. Biol., 12:915-927,

1992.

28. Rivera, V. M., Miranti, C. K., Misra, R. P., Ginty, D. D., Chen, R-H., Blenis,J., and Greenberg, M. E. A growth factor-induced kinase phosphorylates theserum response factor at a site that regulates its DNA-binding activity. Mol.

Cell. Biol., 13:6260-6273, 1993.

29. Hershey, J. W. B. Protein phosphorylation controls translational rates. J.Biol.Chem., 264:20823-20826, 1989.

30. Ferrari, S., Bannwarth, W., Morley, S. J., Tony, N. F., and Thomas, G.

Activation of p7OS6K is associated with phosphorylation offour clustered sitesdisplaying Ser/l’hr-Pro motifs. Proc. NatI. Acad. Sci. USA, 89: 7282-7286,

1992.

31. Chung, J., Kuo, C. J., Crabtree, G. R., and Blenis, J. Rapamycin-FKBP

specifically blocks growth-dependent activation ofand signaling by the 70 kd56 kinase. Cell, 69: 1-20, 1992.

32. Kuo, C. J., Chung, J., Florentino, D. F., Flanagan, W. M., Blenis, J., andCrabtree, G. R. Rapamycin selectively inhibits interleukin-2 activation of p7056 kinase. Nature (Lond.), 358: 70-73, 1992.

33. Price, D. J., Grove, J. R., Calvo, V., Avruch, J., and Bierer, B. E.Rapamycin-induced inhibition ofthe 70-kilodalton S6 protein kinase. Science

(Washington DC), 257: 973-977, 1992.

34. Blenis, I., Chung, I., Erikson, E., Alcorta, D. A., and Erikson, R. L. Distinctmechanisms for the activation of the RSK kinases/MAP2 kinase/pp90� and

pp7O-S6 kinase signaling systems are indicated by inhibition of protein syn-thesis. Cell Growth & Differ., 2: 279-285, 1991.

35. Chung, I., Chen, R-H., and Blenis, J. Coordinate regulation ofpp90’� and

a distinct protein-serine/threonine kinase activity that phosphorylates recom-binant pp90� in vitro. Mol. Cell. Biol., 1 1 : 1 868-1 874, 1991.

36. Hass, R., Hirano, M., Kharbanda, S., Rubin, E., Meinhardt, G., and Kufe,D. Resistance to phorbol ester-induced differentiation of a U-937 myeloid

leukemia cell variant with a signaling defect upstream to Raf-1 kinase. Cell

Growth & Differ., 4:657-663, 1993.

37. Bialojan, C., and Takai, A. Inhibitory effects of a marine sponge toxin,okadaic acid, on protein phosphatases: specificity and kinetics. Biochem. J.,256: 283-290, 1988.

38. Cohen, P., Holmes, C. F. B., and Tsukitani, Y. Okadaic acid: a new probefor the study of cellular regulation. Trends Biochem. Sci., 15: 98-1 02, 1990.

39. Haystead, T. A. J., Sim, A. T. R., Carling, D., Honnor, R. C., Tsukitani, Y.,Cohen, P., and Hardie, D. G. Effects of the tumor promoter okadaic acid on

intracellular protein phosphorylation and metabolism. Nature (Lond.), 337:

78-81, 1989.

40. Kharbanda, S., Dafta, R., Rubin, E., Nakamura, T., Hass, R., and Kufe, D.Regulation of c-jun expression during induction of monocytic differentiation

by okadaic acid. Cell Growth & Differ., 3: 391-399, 1992.

41 . Datta, R., Hallahan, D., Kharbanda, S., Rubin, E., Sherman, M., Huber-man, E., Weichselbaum, R., and Kufe, D. Involvement of reactive oxygenintermediates in the induction of c-jun gene transcription by ionizing radia-

tion. Biochemistry, 31: 8300-8306, 1992.

42. Nishikawa, M., Komada, F., Uemura, Y., Hidaka, H., and Hirakawa, S.Decreased expression of type II protein kinase C in HL-60 variant cells re-

sistant to induction of cell differentiation by phorbol ester. Cancer Res., 50:621-626, 1990.

43. Ward, G. E., and Kirshner, M. W. Identification of cell cycle-regulated

phosphorylation sites on nuclear lamin C. Cell, 61: 561-577, 1990.

44. Erickson, E., and MaIler, J. L. Substrate specificity of ribosomal protein S6kinase II from Xenopus eggs. Second Messenger Phosphoprot., 2: 135-143,

1988.

45. Ofir, R., Dwarki, V. J., Rashid, D., and Verma, I. M. Phosphorylation of

the C terminus of Fos protein is required for transcriptional transrepression of

the c-fos promoter. Nature (Lond.), 348: 80-82, 1990.

46. Gius, D., Cao, X., Rauscher, F. J., Ill, Cohen, D. R., Curran, T., and

Sukhatme, V. Transcriptional activation and repression by fos are indepen-dent functions: the C terminus represses immediate-early gene expression via

CArG elements. Mol. Cell. Biol., 10: 4243-4255, 1990.

47. Prywes, R., Dutta, A., Cromlish, J., and Roeder, R. Phosphorylation of

serum response factor, a factor that binds to the serum response element ofthe c-fos enhancer. Proc. NatI. Acad. Sci. USA, 85: 7206-7210, 1988.

48. Treisman, R. Identification of a protein-binding site that mediates tran-

scriptional response of the c-fos gene to serum factors. Cell, 46: 567-574,1986.

49. GilIe, H., Sharrocks, A. D., and Shaw, P. E. Phosphorylation of Iran-

scription factor p62T� by MAP kinase stimulates ternary complex formationat c-fos promoter. Nature (Lond.), 358: 414-417, 1992.

50. Sundstrom, C., and Nilsson, K. Establishment and characterization of ahuman histiocytic lymphoma cell line (U-937). nt. J. Cancer, 17: 565-577,

1976.

51 . Chirgwin, J. M., Przybyla, A. E., MacDonald, R. 1., and Rutter, W. J.Isolation of biologically active ribonucleic acid from sources enriched in ri-

bonuclease. Biochemistry, 18: 5294-5299, 1979.

52. Cleveland, D. W., Lopata, M. A., MacDonald, R. J., Rutter, W. J., andKirschner, M. W. Number and evolutionary conservation of a- and f3-tubulin

and cytoplasmic f3- and ‘y-actin genes using specific cloned cDNA probes.Cell, 20:95-105, 1980.

53. Grosschedl, R., and Baltimore, D. Cell-type specificity of immunoglobingene expression is regulated by at least three DNA sequence elements. Cell,41:885-897, 1985.