Transforming Growth Factor-@ Inhibits Phosphorylation of the ...

Vol. 267, No. 24, Issue of August 25, pp. 17121-17127,1992 Printed in U.S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY (0 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Transforming Growth Factor-@Inhibits Phosphorylation of the Retinoblastoma Susceptibility Gene Productin Human Monocytic Leukemia CellLine JOSK-I* (Received for publication, February 3, 1992)

Yusuke Furukawa, Shoichi Uenoyama, Masatsugu Ohta, Atsuko Tsunoda, James D. Griffin$, and Masaki Saiton From the Division of Hemopoiesis, Institute of Hematology, Jichi Medical School, Tochigi, Japan andthe $Division of Tumor Immunology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115

Proliferation of the human monocytic leukemia cell line JOSK-I is inhibited by transforming growth factor+ (TGF-8). Growth inhibition by TGF-j3 was not due to eithera toxic effector to inductionof differentiation. TGF-8 induced a cell cycle arrest at late GI phase and was not found to be inhibitory to JOSK-I cells in S phase or G2/M. This Gl cell cycle arrest was associated with an accumulation of the unphosphorylated form of the retinoblastoma susceptibility gene product(Rb)in good correlationwithinhibition of DNA synthesis. In contrast to the effects of TGF-8, two other agentswhich induceda G1 arrest of JOSK-I cells had a different effecton Rb. Aphidicolin blocked cells a t G1/S but could not reduce Rb phosphorylation as great as that seen with TGF-8. 12-0-Tetradecanoylphorbol-13-acetate, an inducer of differentiation, did reduce Rb phosphorylation, but not until 72 h, when differentiation had already occurred. The identities of the Rb kinases are unknown, but recent evidence suggests that the cdc2 gene product could participate in Rb phosphorylation. Although cdc2 mRNA and total protein levels were not affected, TGF-8 inhibited the rate of translation and kinase activity of cdc2 in JOSKI cells. These results suggestthat growth inhibitionof hematopoietic cells by TGF-8 is linked to suppression of Rb phosphorylation to retain Rb in an unphosphorylated, growth-inhibitory state. The suppression of Rb phosphorylation is suggested to be mediated through inhibition of cdc2 kinase activityby TGF-8.

/3 may have important physiological roles in theregulation of

embryogenesis, tissue repair, or cell growth and differentiation, however, little is known about its mechanism of action (2, 3). The wide diversity of action suggests that there may be multiple mechanisms which can be varied with the cell type. Inhibition of growth-related genes such as c-myc, KC, a n d J Eappears toplay an important role in certain endothelial cells and keratinocytes (4, 5 ) but not in fibroblasts(6, 7). TGF-P does not, ingeneral, inhibit cell growth by inactivating growth factor receptors, although this mechanism was observed in some systems including interleukin 1 receptor on T-lymphocytes (8). Therefore, it is likely of value to investigate the particular mechanism of action of TGF-P ina variety of different hematopoieticcell lines. The retinoblastoma susceptibility gene (RB) is the prototype for a class of tumor suppressorgenes (9-11). Mutational inactivation of RB hasbeen implicated in thedevelopment of a wide variety of tumors including retinoblastoma, osteosarcoma (12), soft-tissue sarcomas (13), small cell lung cancer (14), breast cancer (15), bladder cancer (16),hematologic and malignancies (17). This suggests that theloss of RB function can contribute to the loss of growthregulation in several different cell lineages. Viral transforming proteins such as SV40 large T antigen form specific complexes with the RB gene product (18, 19). Thiscomplex formation contributes to the transforming action of DNA tumor viruses, possibly by inactivating the growth-suppressing activityof RB (20). The mechanism by which RB regulates cell proliferation is not yet understood, but recent evidence suggests that phosphorylation of Rb protein may be linked to control of its function TGF-P’ is a polypeptide factor which has complex positive (21-24). Rb phosphorylation iscell cycle associated, i.e. Rb is or negative effects on a wide variety of cell types including unphosphorylated in Go/G, phase of the cell cycle and specifhematopoietic cells (1).A series of studies indicate that TGF- ically phosphorylated at the GI/S boundaryby a kinase that is inactive during Go/G1 phase (25). These observationshave *This workwas supported in part by a Grant-in-Aid from the led to a model of RB function inwhich unphosphorylated Rb Ministry of Education, Science and Culture of Japan and by a grant cell cycle block can be from the Yamanouchi Foundation for Research on Metabolic Disor- inhibits G1/S transition, and this ders. The costs of publication of this article were defrayed in part by released by inactivating Rb by phosphorylation or a complex the payment of page charges. This article must therefore be hereby formation with viral transforming proteins. marked “advertisement” in accordance with 18 U.S.C. Section 1734 A recent study showed that growth inhibition by TGF-P solely to indicate this fact. was linked to suppressionof Rb phosphorylation to retain Rb ll To whom correspondence should be addressed: Div. of Hemopoiesis, Institute of Hematology, Jichi Medical School, Minamika- protein in an unphosphorylated, growth-inhibitory state in wachi-machi, Kawachi-gun, Tochigi 329-04, Japan. Tel.: 81-285-44- MvlLu lung epithelial cells (26). We have previously shown 2111; Fax: 81-285-44-1317. that phosphorylation of Rb plays an important role in cell ’ The abbreviations are: TGF-@, transforminggrowth factor-@; Rb, cycle control and terminal differentiationof human hematoretinoblastoma susceptibility gene product; RB, retinoblastoma suspoietic cells (27). Thus, itis interesting to investigate whether ceptibility gene; TPA, 12-0-tetradecanoylphorbol-13-acetate; TdR, the suppression of Rb phosphorylation also plays an importhymidine; TK, thymidine kinase; TBS, tris-buffered saline; Hepes, inhibition of hematopoietic cells by 4-(2-hydroxyethyl)-l-piperazine-ethanesulfonic acid; EGTA, [ethyl- tant role ingrowth enebis(oxyethylenenitrilo)]tetraacetic acid; MOPS, 4-morpholinepro- TGF-P. panesulfonic acid kb, kilobase(s). JOSK-I isa human monocytic leukemiacell line established

17121

17122

Inhibition of Rb Phosphorylation by TGF-p in Leukemic Cells

each of aprotinin, phenylmethylsulfonyl fluoride, and leupeptin (Sigma). The lysates were then centrifuged for 15min, and the supernatants were applied onto 7.5% SDS-polyacrylamide gels. After electrophoresis, the proteins were transferred to nitrocellulose membranes (Schleicher & Schuell) in buffer containing 25 mM Tris-HC1, 192 mM glycine, and 20% methanol. After blocking in TBS with 4% bovine serum albumin (fraction V, Sigma), the membrane was incubated for 12 h with 5pg/ml anti-Rb monoclonal antibody PMG3-245 (Pharmingen, San Diego, CA) in TBS containing 0.05% Tween 20 (Bio-Rad) and 4% bovine serum albumin. Then, the membrane was probed with alkaline phosphatase-conjugated rabbit anti-mouse IgG (Promega, Madison, WI) and developed with nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate. The expression of the gene products of c-myc and cdc2 was assayed by immunoblotting using monoclonal antibodies against c-myc (Cambridge Research Biochemicals, Wilmington, DE) and cdc2 (MBL, Nagoya, Japan), respectively (32). Metabolic Labeling-Rate of cdc2 protein synthesis was measured MATERIALS AND METHODS by [35S]methionine pulse-labeling followed by an immunoprecipitaCells and Factors-The JOSK-I cell line was established in our tion. TGF-@-treatedcells were incubated for a final 2 hof the culture laboratory and maintained as previously described (28). The human with 200 pCi of [35S]methionine (Du Pont-New England Nuclear) in methionine-free medium. Samples were adjusted for counts per min, myeloblastic leukemia cell line KG-1 and thepromyelocytic leukemia ~ ~ immunoprecipitated ' as previously described (34). cell line HL-60 were also used as othertypes of leukemia cells. Highly and ~ 3 4 "was purified human TGF-@1was purchased from Collaborative Research Histone H1 Kinase Assay-Histone H1 kinase activity was measInc. (Two Oak Park, Bedford, MA). TPA (12-0-tetradecanoylphor- ured as described (35). Briefly, 10 pg of protein lysate was incubated bol-13-acetate) and aphidicolin were obtained from Sigma and Wako in 40 p1 of kinase buffer containing 20 mM Hepes, pH 7.5, 15 mM Pure Chemicals (Ohsaka, Japan), respectively. EGTA, 20 mM MgCl,, 1 mM dithiothreitol, 500 nM A-kinase inhibiCell Cultures-JOSK-I cells were seeded at aninitial concentration tory peptide (Sigma), 20 Fg of histone H1 (type 111-S, Sigma), and 15 of 1 X lo6 cells/ml inRPMI 1640 medium (Flow Laboratories, FCi of [-y-3ZP]ATPfor 20 min at 30 "C. The reaction was terminated McLean, VA) supplemented with 5% heat-inactivated (56 'C for 30 by adding of 20 pl of 3 X loading buffer and boiling for 5 min. Samples min) fetal calf serum (Commonwealth SerumLaboratories, Mel- were analyzed by 12% SDS-PAGE and autoradiography. bourne, Australia) and grown in the absence or presence of varying concentrations of factors as indicated. Cell proliferation was deterRESULTS mined using [3H]thymidine (TdR) incorporation which was pulsed (0.2 pCi/well) 6 h before the end of incubation (TdR obtained from Effect of TGF-/3 on JOSK-I Cell Proliferation"TGF-/3 inDu Pont-New England Nuclear). hibited the growth of JOSK-I cells in a dose-dependent manCell Synchronization-JOSK-I cells were synchronized at the GI/ S boundary of the cell cycleby aphidicolin (30). The cells in the ner (Fig. 1).JOSK-I cells were more sensitive to TGF-/3than logarithmic growth phase were incubated for 12 h in the presence of other myeloid leukemia cell lines such as HL-60 and KG-1 on 2.5 pg/ml aphidicolin. They were washed twice with fresh medium, the basis of the inhibition of [3H]TdR incorporation (IC&; 1 resuspended at 5 X lo5 cells/ml, and then reincubated at 37 "C in ng/ml for JOSK-I, 50 ng/ml for KG-1, and more than 100 Falcon 3072 plates. TGF-@,and aphidicolin were added at thebeginng/ml for HL-60). The viability of JOSK-I cells was not ning of reculture at a concentration of 10ng/ml and 1.0 Fg/ml, respectively. Synchrony was monitored by serial cell counts, measurement of [3H]TdR incorporation, and cell cycle analysis. For [3H] TdR incorporation, the cells were pulsed with 0.2 pCi of [3H]TdR per well for 30 min, and radioactivity was determined by liquid scintillation counting. Cell cycle analysis was performed by staining DNA with propidium iodide (Sigma) in preparation for flow cytometry (21). Thymidine Kinase Assay-Thymidine kinase (TK) activity was determined with the cell extracts according to themethod of Johnson et al. (31). The specific activity of T K was expressed as picomoles of phosphorylated [3H]TdR (TMP) formed/min/mg protein. Transferrin Receptor Expression-The presence of cell surface expression of transferrin receptors reactive to monoclonal anti-transferrin receptor antibody, OKT9 (Ortho Diagnostics, Westwood, MA) was investigated by the indirect immunofluorescence method using Ortho Spectrum 111. Expression of HLA-DR antigen was determined simultaneously using OKIal antibody. Northern Blotting-Total cellular RNA was isolated by cesium chloride centrifugation after lysing cells in guanidium isothiocyanate solution. 10-pg RNA samples were electrophoresed in a 1.0% agarose gel containing 6% formaldehyde, 20 mM MOPS, 5 mM sodium acetate, and 1 mM EDTA and blotted onto a synthetic nylon membrane in 10 X SSC (1 X SSC = 150 mM NaC1, 15 mM sodium citrate). c-myc mRNA expression was detected using a 1.8-kb ClaI/EcoRI fragment of c-myc exon 3 (Oncor Inc., Gaithersburg, MD) which was labeled with [3'P]dCTP by the oligonucleotide random priming method. The cdc2 mRNA was detected with a 0.9-kb KpnI/PuuII fragment of human cdc2 cDNA as previously described (32). To monitor the equal loading of RNA, the membranes were rehybridized with glyceraldehyde-3-phosphate dehydrogenase probe (Oncogene Science, ManhasFIG. 1. Effect of TGF-BI on leukemic cell proliferation. Triset, NY), since the expression of glyceraldehyde-3-phosphate dehydrogenase mRNA is reportedto be stable after TGF-B treatment (33). tiated thymidine incorporation of JOSK-I (O), KG-1 (O),and HL-60 (A) cells was determined after 72 h of incubation in the presence of Western Blotting-Cells were washed with ice-cold TBS (25 mM Tris-HC1, pH 8, 150 mM NaCl) and lysed for 30 min at 4 "C with varying amounts of TGF-P1 and expressed in terms of the percentage EBC buffer (50 mM Tris-HC1, pH 8, 120 mM NaCl, 0.5% Nonidet P- of the control value. [3H]TdR uptake in the absence of TGF-B, was 40, 100 mM NaF, 200 p~ sodium orthovanadate) containing 10 pg used as a control for each cell line.

from the peripheral blood of a patient with acute myelomonocytic leukemia in our laboratory (28). The cells possess immature monocytic features and were considered as a useful model of monocyte-macrophage lineage hematopoietic cells (29). We have found that the growth of JOSK-I cells was inhibited by the addition of TGF-/3 in a liquid culture system. Although JOSK-I cell line is not as sensitive as lung epithelial cells and keratinocytes, this line is still useful to investigate the mechanism of TGF-@action onhematopoietic cells which may be different from those of other cell types. Using JOSKI cells, we have investigated the mechanism of TGF-/3 action on hematopoietic cells with special reference to its effect on the cell cycle regulatory elements such as Rb protein, c-myc, and cdc2.

Inhibition of Phosphorylation Rb

by TGF-p in Leukemic Cells

17123

Effect of TGF-@on Phase-specific Functions-To identify where TGF-@ acts in thecell cycle, we investigated changes in phase-specific functions in the presence of TGF-@. The induction of thymidine kinase, a marker of cells entering the S phase (31), was inhibited by TGF-P1 (TK activity was 32.5 f 1.4 pmol of TMP/min/mg protein in the absence of TGF@, whereas 13.7 f 0.2 in the presence of TGF-@ at day 3). We then examinedtransferrin-receptor expression which normally occurs in thelate G1 phase and continues during active cell proliferation (36). Transferrin receptor was expressed in 97.8% of JOSK-I cells at the logarithmic growth phase. The expression was reduced by TGF-PI to16.7%,whereas expression of HLA-DR antigen was not affected. This suggests that TGF-@ arrests the cells in late GI phase prior to transferrinreceptor expression. To confirm that the block is actually occurring at late G1,we examined the kinetics of entry to DNA synthesis following the release from TGF-@ growth arrest. As shown in Fig. 2B, JOSK-I cells showed a significant increase of [3H]TdR uptake after2 h of the release, indicating that theblock by TGF-@occurred at late GIphase of the cell cycle. Effect of TGF-@ on c-mycmRNA Expression-Previous studies indicate that TGF-@effect on c-myc expression may be a central event in the growth-inhibitory response in kera0 None tinocytes (37, 38), whereas induction of c-myc was not inhibTGF-E ited in fibroblasts(6, 7). Thus, therole of c-myc in the growth inhibition of hematopoietic cells was investigated using JOSK-I cells. When JOSK-I cells were growth arrested by differentiation induced by 4 nM TPA, c-myc showed a transient decrease in expression at 6 h (Fig. 3 A ) , suggesting that c-myc plays some role in the commitment of differentiation as previously shown in mouse erythroleukemia (MEL) cells (39, 40). When cells were arrested in early S phase by aphidicolin, c-myc expression decreased after 36 h of incubation when [3H]TdR uptake was almost completely inhibited. In contrast with these results, TGF-@did not reduce the c-myc expression within 36 h of culture when [3H]TdRincorporation was significantly inhibited (Fig. 3A). A more detailed time36 I 24 12 0 course experiment confirmed that c-myc expression was not Time after release of aphidicolin block I h ) affected within 48 h but decreased after a prolonged exposure to TGF-@(Fig. 4). To confirm the lack of effect of TGF-@on c-myc in the earlier time course, the amountof c-myc protein (B) was examined by immunoblotting. As shown in Fig. 5B, the level of c-myc protein, p62c"''yc,was not altered by TGF-@ within 72 h of culture. These results indicate that c-myc does k not play a pivotal role in the early step of TGF-@-induced i growth arrest of JOSK-I. G Effect of TGF-/3 on Phosphorylation ofRb Protein-Since phosphorylation of Rb is a critical step for G1/S transition, we investigated the effect of TGF-@on the expression and phosphorylation of Rb by immunoblotting. The specificity and sensitivity of this assay have been previously demonstrated and the phosphorylation state of Rb can be distinguished by the mobility on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (19-21). In the exponentially growing phase, JOSK-I cells showed mainly phosphorylated Rb protein which consists of multiple bands at the molecular 8 12 16 mass range of 112-114 kDa (pp112-114). As early as 18 h Tim0 eftor releame 01 TGF-p block (h) FIG. 2. A , effect of TGF-B1 on the synchronized JOSK-I cells. after addition of TGF-@,, a decrease in the level of phosAfter 12-h treatment with 2.5 pg/ml of aphidicolin, growth of the phorylated Rb started and unphosphorylated Rb species apresuspended cells was arrested after one synchronous division (0). peared at the molecular mass of 110 kDa (p110) (Fig. 5A). TGF-PI ( 0 ) and aphidicolin (A) were added at the beginning of Unphosphorylated Rb became apparent after 72 h of culture reculture after washing the cells twice. [3H]TdR uptake was deterwhen DNA synthesis decreased to 33% that of the control. mined at given time points. B, kinetics of entry to DNA synthesis following release from TGF-8 growth arrest. JOSK-I cells were cul- TPA also induced dephosphorylation of Rb (Fig. 6B). Howtivated with 10 ng/ml of TGF-0 for 3 days, then washed, and resus- ever, accumulation of unphosphorylated Rb was not evident pended in the medium with (+) or without (-) TGF-0. until 72 h, suggesting that dephosphorylation of Rb was a

affected, and differentiation was not induced by TGF-@ (data not shown). Effect of TGF-@on S Phase Synchronized Cells-We examined the effect of TGF-@on the synchronized population of JOSK-I cells. The cells were arrested at theG1/S boundary by treatmentwith aphidicolin. The peak [3H]TdR uptakewas at 8 h after the beginning of reculture, and the length of the S phase was considered to be 6-8 h (Fig. 2A). TGF-B1 did not affect the length of the S phase and [3H]TdR uptakewithin 24 h of the reculture. This indicates that TGF-@ did not affect the growth of cells after they had already entered theS phase. In contrast,aphidicolin which blocks cell proliferation at the early S phase via inhibition of DNA polymerase (Y completely abrogated DNA synthesis in synchronized JOSK-I cells. At 48 h, when substantial numbers of the cells entered thesecond S phase after the passage through the first S and G1 phases, ["HITdR uptake was significantly inhibited by TGF-P1. This suggests that TGF-@affects the cells in G1 phase of the cell cycle. TGF-@did not seem to inhibit the mitotic process since the cell number at 18-24 h of culture with TGF-@was the same as thatof the control (data not shown).

I

1

* I

17124

Inhibition of Phosphorylation Rb Aphdiiolin Control TPA

FIG. 3. Effect of various growth inhibitors on c-mvc and cdc2 mRNA expression. JOSK-I cells were cultured in

absence or presence of TGF-pl at

6 ng/ml, TPAat 4 nM and aphidicolin at 1 pg/ml. Total cellular RNA was iso-

and

latedpoints, a t given time expression- of c-myc ( A ) and cdc2 ( B ) mRNA was evaluated by Northern blotloading of ting. To monitortheequal RNA, the membraneswere rehybridized with glyceraldehyde-3-phosphate dehydrogenase probe (C).

(B)

&,pm .

by TGF-p in Leukemic Cells TGF-f3

rp@( ib.44

GAPDH

c

(1.2kb)

0 6 18 36 0 6 36 18 Time in culture (h)

2

4

0

6 12 24 48 72 (A)

YYWUUOr*rY..

0 6

36 18

Time in culture (h)

Time in culture (h) 1

cdc2 (1.8kb) -(1.6kb)

(c)

0 6 18 36

0

-

6

18

36

72

r pp 112-1 14

-

GAPDH (1.2kb)

- p34 cdc2

FIG.4. Effect of TGF-8 on c-myc mRNA expression. c-myc mRNA expression was examined by Northern blotting as described in Fig. 3 legend.

consequence of cell cycle arrest due to differentiation rather than a primary event induced by TPA. Dephosphorylation was not as great as thatseen with TGF-8 when JOSK-I cells were treated with aphidicolin (Fig. 6C). Accumulation of unphosphorylated Rb was minimal after 72 h of culture, although DNA synthesis was almost completely inhibited by aphidicolin. The same result was obtained with hydroxyurea treatment (datanot shown). These reagents did not primarily inhibit the phosphorylation of Rb because they blocked cell cycle at early S phase after thecompletion of Rb phosphorylation. Effect of TGF-/3 on cdc2 Kinase Actiuity-Dephosphorylation of Rb by TGF-0 might be due to either inhibition of kinase activity or induction of phosphatase activity. It is very difficult to assess the specific phosphatase activity against Rb since the phosphatase involved in Rb dephosphorylation has not yet been determined. Thus, we evaluated the effect of TGF-/3 on kinase activity. Although the identities of the Rb kinases are unknown, recent studies including our own suggest that the cdc2gene product could participate in Rb phosphorylation (32, 41, 42). The modulation of cdc2 mRNA expression was evaluated by Northern blotting. As shown in Fig. 3B, expression of cdc2 mRNA was not inhibited by TGFp. Additionally, both TPA and aphidicolin could not downregulate the cdc2 mRNA level despite the significant growth inhibition. This suggests that thelevel of cdc2 mRNA is not affected by the change in cell proliferation of JOSK-I. Next, we examined the level of cdc2 gene product, p34cdc2, byim-

- Histone H1 100 92 76 53 33 DNA synthesis (% Control) FIG. 5. Effect of TGF-f3 on various cell cycle control elements. JOSK-I cells were cultured in the presence of TGF-01 a t 5 ng/ml for 72 h. The effect on cell proliferation was monitored by [3H] TdR uptake and shown as percentage of control. The cell lysates were isolated a t given time points, and theexpression of Rb protein ( A ) ,~62'-"'~( B ) ,~34'~''(C), and histone H1(D) kinase activity were evaluated simultaneously. The extent of phosphorylation of Rb was determined by its mobility on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Phosphorylated Rb migrated at the molecular mass of112-114 kDa (pp112-114), and unphosphorylated Rb migrated a t 110 kDa ( ~ 1 1 0 ) .

munoblotting. Total amountof ~34'~'' in whole celllysate was not reduced by TGF-8 (data not shown). However, the rate of cdc2 translation as determined by [35S]methioninepulse labelingwas inhibitedby TGF-/3 (Fig.5C).The kinase activity of the cdc2 geneproduct was inhibited by TGF-P when histone H1 was used as a specific substrate in vitro, probably due to the reduction of the amountof newly synthesized cdc2 protein which is now known to be modifiedand activated for a kinase activity (43). As shown in Fig. 5 0 , phosphorylation of histone H1 in vitro started to decrease after 18-h treatment with TGF-8 concomitant withthe appearance of unphosphorylated

Inhibition of Phosphorylation Rb by 6

18

36

17125

GI but an event in the TGF-/3 response which leads to cell cycle arrest in late GI.

Time in culture (h) 0

TGF-p in Leukemic Cells

72

(A) Control

DISCUSSION

.

Jpp112-114 TGF-P is now considered to be an important - 110

regulatory factor which modulatesthe growth anddifferentiation of many cell types including normal, neoplastic, and immunocompetent cells. Recently, many observationsof the actionof TGF-@ havebeen noted, and the therapeutic potential of (B)TPA TGF-P has been considered. However, little is known about 'Y.U. the mechanism of action, and previous studies indicate that ?E@"p110 there may be many mechanisms which may be different in DNA synthesis each cell type (2,3). (% Control) 100 78 21 27 65 Here, we report the actionof TGF-P onhematopoietic cells using monocytic leukemia cell line JOSK-I cells. Cell growth of JOSK-I cells was markedly inhibited by the addition of (C)Aphidicolin TGF-P1. We tested other human leukemic cell lines KG-1 and pp 112-1 l4 HL-60 simultaneously. The HL-60cell line hasbeen reported -p 110 to lack TGF-P receptor (44) and was resistant to the effects DNA synthesis of TGF-@,. KG-1 wasslightlyless sensitive than JOSK-I. Thus, JOSK-I isconsidered to be a good model to investigate (% Control) 100 69 22 54 8 the mechanism of action of TGF-P on hematopoietic cells. FIG. 6. Effect of TPA and aphidicolin on phosphorylation of morphological observations,TGF-P1 was of Rb protein. JOSK-I cells were cultured in the absence ( A ) or Onthebasis cell viabilities nor inducecellular presence of TPA at 4 nM ( R ) and aphidicolin a t 1 pg/ml (C). The foundtoneitheraffect expression of Rhprotein was determined by immunoblotting as differentiation of JOSK-I cells. Previous studies suggest that described in Fig. 5 legend. TGF-0 might inhibit cell proliferation by affecting the cell cycle. For example, TGF-@blocked the GI to S transition in early or middle GI phase in mitogen-stimulated B-lymphoTime after release of aphidicolin block cytes (45) or fibroblasts (6). Therefore, we first investigated 48 36 24648 36 246 (h) the effects of TGF-/3 on synchronized cells and found that it did not inhibit the growth of cells when they had already entered the S phase. JOSK-I cells started DNA synthesis "pllORb soon after release from the growth arrest by TGF-8. These results lead to the conclusion that TGF-/3 arrests the cells in late GI phaseof the cell cycle. 15 56 79 44 42 19 55 6 9 GoIG 1 Recently, TGF-@ has been reported to directly modulate 18 46 5 1 7 9 4 0 29 82 38 S some cell cycle control elements to inhibit cell proliferation. For example, growth inhibition of skin keratinocytes is me6 3 3 102 75 2 G2lM diated through suppression of c-myc mRNA by TGF-@ (37, Control TGF-P 38). But c-myc mRNA wasdown-regulatedonly after the growth of JOSK-I was fully inhibited, and the level of c-myc FIG. 7. Kinetics of Rb dephosphorylationby TGF-B. Rb protein was analyzed for its phosphorylation status after the release of protein was almost stable in the experimental condition deaphidicolin block. Aphidicolin block was performed as described in scribed here. TPA induced a transient inhibition of c-myc Fig. 2. Cell cycle analysis was performed simultaneously by staining mRNA expression, indicating thevalidity of the role of c-myc DNA with propidium iodide. in the commitmentof differentiation in JOSK-Icells. Thus, inhibition of c-myc is unlikely to play a central role in TGFRb. This suggests that inhibitionof cdc2 kinase activity is, a t p action in this cell line especially in the early time course. least partially, involved in dephosphorylation of Rb by TGF- The lack of effect on c-myc by TGF-/3 is also reported in p, but theinvolvement of another kinase and/or phosphatases fibroblasts (6, 7). This again suggests that the mechanism of action of TGF-P iscomplex and varies with thecell type. cannot be ruled out. Another cell cycle control elementwhich can be modulated Kinetics of Rb Dephosphorylation by TGF-p-Rb protein was analyzed for its phosphorylation status after the release by TGF-fi is the RB gene product. Phosphorylation of Rb of aphidicolin block. As shown in Fig. 7, dephosphorylation protein has been reported to be a critical step of GI/S transiof human cells examined so far (21-25). Since of Rb protein was observed either in the presence or absence tion in all types of TGF-@ with similar levels after 24 h from therelease when the unphosphorylated form of Rb is active as a growth supsubstantial numbersof the cells were at the GI phase. Rb was pressor to keep cells in a resting state, dephosphorylation of mainly observed as a heavily phosphorylated form after 36 h Rb may be a target of TGF-P action. In JOSK-I cells, TGF-/3 when the cells re-entered S phase in the absence of TGF-@. induced a reduction in the amountof phosphorylated Rb and TGF-P prevented the phosphorylationof Rb protein after36 arrested cells in late GI as recently reported in MvlLu cells h from the release of aphidicolin block, whereas it could not (26). The reduction of phosphorylated Rb occurred as early induce dephosphorylation of Rb at 6 h when the majority of as 18 h which was in good correlation with inhibition of DNA the cells were at the S phase. This means that TGF-0 could synthesis. Dephosphorylation of Rb preceded down-regulanot act on Rb proteinat various cell cycle times and that the tion of c-myc. This may suggest the primary role of dephossignal might be transduced only a t late GI via inhibition of phorylation of Rb in growth inhibition by TGF-/3 but cannot fully rule out the possibility that inhibitionof Rb phosphorylR b kinase. This alsosuggests that inhibition of Rb phosphorylation by TGF-/3 is not a consequence of arrest in early ation is a consequence of growth arrest. The following two y

Y T

p

-

- Jpp112-114

4-

Jpp112-114

17126

Inhibition of Rb Phosphorylation by TGF-p in Leukemic Cells

evidences seem to support the notion that Rb dephosphorylation by TGF-P is a primary eventrather thana consequence. First, dephosphorylation of Rb was not induced after 6 h of the release from aphidicolin block where the cells were mainly a t the S phase. This means that TGF-/3 could not act on Rb protein at various cell cycle times and that the signal itself might be transduced only at late GI via inhibition of Rb kinase, suggesting that inhibition of Rb phosphorylation is not a consequence of arrest at early GI. Second, Laiho et al. (26) showed that the SV40 T antigen, a viral transforming protein which selectively binds to unphosphorylated Rb and perturbs growth-inhibitory function,abrogates the growth inhibitory response to TGF-P without affecting dephosphorylation of Rb. Their study clearly demonstrates that the effect of TGF-P on Rb phosphorylation is a primary event rather than a secondary response to growth arrest. TPA also induced dephosphorylation of Rb, but thisoccurred relatively late (after48 h of culture). Similar resultshave been observed with other differentiationinducers such as dimethyl sulfoxide and retinoic acid (23, 24). This suggests that dephosphorylation by TPA is a consequence of differentiation rather thana primary target of TPA action. Since phosphorylation of Rb plays an important role in cell cycle control of human hematopoietic cells (27), this can be a reasonable target for TGF-/3 which is known as an important mediator of negative regulation of hemopoiesis (1,44). Dephosphorylation of Rb by TGF-8 might be mediated either by inhibition of kinases or activation of phosphatases or both. The identities of both kinase and phosphatase are virtually unknown, but recent studies suggest that the cdc2 gene product can act asa Rb kinase (32, 41). cdc2was originally identified in fission yeast as an essential cell cycle controlelement for both Gl/S and G2/M transition (46). Subsequently, cdc2 has been found to be conserved in a wide variety of species (47, 48). The product of the cdc2 gene, designated p3PdcZ,is a serine-threonine protein kinase which controls entry of eukaryotic cells into cell cycle (49). p34'd'2 is now known to be a catalytic subunitof the mitosis-regulating protein kinase complex known as maturation promoting factor or growth-associated histone H1 kinase (50, 51). The function of cdc2 in the promotion of mitosis is firmly established and a number of substrates for ~ 3 4 at~ G2/M ~ ~ transi' tion have been reported (52). In contrast, its function at GI/ S transitionin higher eukaryotes is not well understood. Recently, we reported that blocking ~34'~''up-regulation by antisense oligonucleotides resulted in inhibition of S-phase entry and reduction of the amount of phosphorylated Rb in human T lymphocytes (32). Lin et al. (41) showed that cdc2 kinase efficiently phosphorylates unphosphorylated Rb prepared from insect cells infected with an RB-recombinant baculovirus. These results suggest a role for cdc2 in G1/S transition and Rb phosphorylation. With these backgrounds, we tested the hypothesis that dephosphorylation of Rb by TGF-/3 was mediated through inactivation of cdc2 kinase. ' not afBoth cdc2 mRNA and total amount of ~ 3 4 ' ~ "were fected by TGF-P as well asTPAor aphidicolin. This is consistent with the previous report that the amount of cdc2 mRNA and ~34'~'' remained constant throughout the cell cycle in growingcell lines such as HeLa (53). Thus, we examined the change of cdc2 kinase activity using histone H1 as a substrate and found that histone H1 kinase activity was inhibited by TGF-P in parallelwith the dephosphorylation of Rb. The inhibition of histone H1 kinase activity is, at least in part, due to the reduction of the newly synthesized cdc2 protein, because the rate of cdc2 translation was inhibited by TGF-P after 18 h of the culture. Recent evidence suggests

that only the newly synthesized cdc2 protein can be modified and activated for a kinase activity (43). However, the modulation at the posttranslational level cannot be ruled out. Although histone H1 kinase activity is not representative of all the cdc2 activity, it could be a part of activities which are important for the cell cycle control. Thus, our results support the recent report that TGF-P growth inhibition involves the regulation of cdc2 activity at the GI/S transition (54), but cannot rule out the possibility of involvement of another kinase and/or phosphatase. In summary, growth inhibition of hematopoietic cells by TGF-P seems to be linked to the suppression of Rb phosphorylation to retain Rb in an unphosphorylated, growthinhibitory state. Our present data are consistent with the previous reports using MvlLu lung epithelial cells, suggesting the generality of this phenomenon as a mechanism of TGF-/3 action, although they did not give direct evidences. Further investigation should be required to confirm this important hypothesis directly. Acknowledgment-We thank Drs. T. Sakai, E. Azuma, K. Sakoe, M. Nakamura, and S. Kitagawa of the Division of Hemopoiesis, Institute of Hematology, Jichi Medical School, for valuable advice during this work. We are also grateful to J. Yamanoi andN. Usui for technical assistance and toF. Saotome for secretarial assistance. REFERENCES 1. Ohta, M., Greenberger, J. S., Anklesaria, P., Bassols, A., and Massague, J. (1987) Nature 329.539-541 2. ~ & i i & i 5. , (1987) Ceii491437-438 3. Moses, H. L., Yang, E. Y., and Pietenpol, J. A. (1990) Cell 6 3 , 245-247 4. Takehara, K., LeRoy, E. C., and Grotendorst, G. R. (1987) Cell 4 9 , 415A33

5. Coffev. R. J.. Bascom. C. C.. Sioes. N. J.. Graves-Deal. R.. Weissman. B. E.,-and Moses, H. L. ( 1 9 6 ) Qol. 'Cell. &ol. 8, 3088-3093' 6. Chambard, J. C., and Pouyssegur, J. (1988) J. Cell. Physiol. 135,101-107 7. Sorrentino, V., and Bandyopadhyay, S. (1988) Oncogene 4,569-574 8. Dubois, C. M., Ruscetti, F. W., Palaszynski, E. W., Falk, L. A., Oppenheim, J. J.. and Keller. J. R. (1990) J. EXD.Med. 172. 737-744 9. Cavenee, W.K., D'&a, T:P., Philli s', R A., Benedict, W. F., Godbout, R., Gallie, B. L., Murphree, A.L., Jtrong, L. C., and White, R. L. (1983) Nature 3 0 5 , 779-784 10. Friend, S. H., Bernards, R., Rogelj, S., Weinberg, R. A,, Rapaport, J. M., Albert. D. M.. and Drvia. T. P. (1986) Nature 323.643-646 11. Lee, W.-H., Shew, J.-Y.;*Hong, F.'D., Sery, T. W., Donoso, L. A,, Young, L.-J., Bookstein, R., and Lee, E.-H. (1987) Science 235,1394-1399 12. Toguchida, J., Ishizaki, K., Sasaki, M., Nakamura, Y., Ikenaga, M., Kato, M., Sugimoto, M., Kotoura, Y., and Yamamuro, T. (1989) Nature 3 3 8 , 156-158 13. Cance, W. G., Brennan, M. F., Dudas, M. E., Huang, C.-M., and CordonCardo, C. (1990) N . Engl. J. Med. 323,1457-1462 14. Harbour, J. W., Lai, S.-L., Whang-Peng, J., Gazdar, A. F., Minna, J. D., and Kaye, F. J. (1988) Science 241,353-356 15. T'Ane. A,. Varlev. J. M.. Chakrabortv. S.. Muruhree, A. L., and Fung, Y. K. y. (1988) Siience 2 4 2 , 263-266. 16. Horowitz, J. M., Yandell, D.W., Park, S.-H., Canning, S., Whyte, P., Buchkovich, K., Harlow, E., Weinberg, R. A., and Dryja, T. P. (1989) Science 243,937-940 17. Furukawa. Y.,DeCaDrio. J. A., Belvin. M., and Griffin, J. D. (1991) Oncogene 6,1343-i346 18. Whyte, P., Buchkovich, K. J., Horowitz, J. M., Friend, S. H., Raybuck, M., Weinberg, R. A., and Harlow, E. (1988) Nature 3 3 4 , 124-129 19. DeCaprio, J. A,, Ludlow, J. W., Figge, J., Shew, J.-Y., Huang, C.-M., Lee, W.-H., Marsilio, E., Paucha, E., and Livingston, D. M. (1988) Cell 5 4 , 275-283 20. Ludlow. J. W.. DeCaDrio. J. A,. Huane. C.-M... Lee,. W.-H.. Paucha. E., and Livingston,b. M. 71989) Ceil56, 57-65 21. DeCaprio, J. A,,Ludlow, J. W., Lynch, D., Furukawa, Y., Griffin, J., Piwnica-Worms. H.. Huang, C.-M., and Livingston, D.M. (1989) Cell 5 8 , 1085-1095 22. Buchkovich, K., Duffy, L. A,, and Harlow, E. (1989) Cell 58,1097-1105 23. Chen, P.-L., Scully, P., Shew, J.-Y., Wang, J. Y. J., and Lee, W.-H. (1989) Cell 5 8 , 1193-1198 24. Mihara, K., Cao, X.-R., Yen, A., Chandler, S., Driscoll, B., Murphree, A. L., T'Ang, A., and Fung, Y.-K. (1989) Science 246,1300-1303 25. Ludlow, J. W., Shon, J., Pipas, J. M., Livingston, D. M., and DeCaprio, J. A. (1990) Cell 6 0 , 387-396 26. Laiho, M., DeCaprio, J. A., Ludlow,J. W., Livingston, D. M.,and Massague, J. (1990) Cell 6 2 , 175-185 27. Furukawa, Y., DeCaprio, J. A,, Freedman, A,, Kanakura, Y., Nakamura, M., Ernst, T. J., Livingston, D. M., and Griffin, J. D. (1990) Proc. Natl. Acud. Sci. U. S. A. 87,2770-2774 28. Ohta, M., Furukawa, Y., Ide, C., Akiyama, N., Utakoji, T., Miura, Y., and Saito, M. (1986) Cancer Res. 46,3067-3074 29. Furukawa, Y., Ohta, M., Kasahara, T., Miura, Y., and Saito, M. (1987) Cancer Res. 4 7 , 2589-2593 30. Ayusawa, D., Shimizu, K., Koyama, H., Takeishi, K., and Seno, T. (1983) J. Biol. Chern. 2 5 8 , 12448-12454 '

~

'

Inhibition of Rb Phosphorylation by TGF-P in Leukemic Cells 31. Johnson, L. F., Rao, L. G., and Muench, A. J. (1982)Exp. Cell. Res. 138, 79-85 32. Furukawa, Y., Piwnica-Worms, H., Ernst, T. J., Kanakura, Y., and Griffin, J. D. (1990)Science 260,805-808 33. Lund, L. R.,Romer, J., Rome, E., Ellis, V., Blasi, F., and Dano, K. (1991) EMBO J. 10,3399-3407 34. McGowan, C. H., Russell, P., and Reed, S. I. (1990)Mol.Cell.Biol. 10, 3847-3851 35. Ishida, Y., Furukawa, Y., DeCaprio, J. A., Saito, M., and Griffin, J. D. (1992)J . Cell. Physiol. 150,484-492 36. Trowbrldge, I. S., and Omary, M.B. (1981)Proc. Natl. Acad. Sci. U, S. A .

-. 37. Pieten 01, J

78.3039-3n43 -""

"

A , Holt, J. T., Stein, R. W., and Moses, H.L. (1990)Proc. Nat!A& Sci. U. S.A. 87,3758-3762 38. Pietenpol, J. A,, Stein, R.W., Moran, E.,Yaciuk P. Schle el R. L ons R. M..Pittelkow, M. R., Munger, K., Howley, k.M., a n t M o d s , L: 777-785 39. Lachman, H. M., and Skoultchi, A. I. (1984)Nature 310,592-594 40. Kume, T. U.,Takada, S., and Obinata, M. (1988)J. Mol. Biol. 202,779-

A.

""

VQC I

41. Lin, B. T.-Y., Gruenwald, S., Morla, A. O., Lee, W.-H., and Wang, J. Y. J. (1991)EMBO J. 10,857-864

17127

42. Taya, Y., Yasuda, H., Kamijo, M., Nakaya, K., Nakamura, Y., Ohba, Y., and Nishimura, S. (1989)Biochern. Biophys. Res. Cornrnun. 164, 580586 43. Welch, P. J., and Wang, J. Y. J. (1992)Proc. Natl. Acad. Sci. U. S. A. 89, 3093-3097 44. Sine, G. K., Keller, J. R.. Ellingsworth. L. R.. and Ruscetti. F. W. (1988) Bhod 72,1504-1511 45. Smeland, E. B., Blomhoff, H. K., Holte, H., Ruud, E., Beiske, K., Funderud, S.. Godal. T.. and Ohlsson. R. (1987)EXD.Cell. Res. 171.213-222 46. Nurse, P., andBissett, Y. (1981)kat& 292,558-560 47. Lee, M. G., and Nurse, P. (1987)Nature 327,31-35 48. Dunphy, W. G., Brizuela, L., Beach, D., and Newport, J. (1988)Cell 54, '

'

."

A7R-ARl ""

49. Simanis, V., and Nurse, P. (1986)Cell 45,261-268 50. Gautier, J., Norbury, C., Lohka, M., Nurse, P., and Maller, J. (1988)Cell 54,433-439 51. Arion, D., Meijer, L., Brizuela, L., and Beach, D. (1988)Cell 6 6 , 371-378 52. Moreno, S., and Nurse, P. (1990)Cell 61,549-551 53. Draetta, G., and Beach, D. (1988)Cell 64, 17-26 54. Howe, P. H., Draetta, G., and Leof, E. B. (1991)Mol. Cell. Biol. 11, 11851194