Alternative Signaling Mechanism of Leukemia Inhibitory Factor ...

2 downloads 0 Views 4MB Size Report
ABSTRACT. Leukemia inhibitory factor (LIF) is a cytokine that plays an im- portant role during mouse embryogenesis. We showed that adenovi- rus E1A ...
0013-7227/97/$03.00/0 Endocrinology Copyright © 1997 by The Endocrine Society

Vol. 138, No. 7 Printed in U.S.A.

Alternative Signaling Mechanism of Leukemia Inhibitory Factor Responsiveness in a Differentiating Embryonal Carcinoma Cell TAKASHI TAKEDA, HIROHISA KURACHI, TOSHIYA YAMAMOTO, HIROAKI HOMMA, KAZUSHIGE ADACHI, KENICHIROU MORISHIGE, AKIRA MIYAKE, AND YUJI MURATA Department of Obstetrics and Gynecology, Osaka University Medical School, Osaka 565, Japan ABSTRACT Leukemia inhibitory factor (LIF) is a cytokine that plays an important role during mouse embryogenesis. We showed that adenovirus E1A represses the interleukin-6 signal transduction pathway that uses the same JAK tyrosine kinase and STAT (signal transducer and activator of transcription) transcription factor as LIF. Here, we report that the LIF-JAK-STAT signal transduction pathway is blocked in cellular E1A-expressing undifferentiated F9 cells, and that the block is overcome by retinoic acid-induced differentiation. LIF failed to stimulate the expression of the acute phase response element (APRE)driven luciferase gene in undifferentiated F9 cells, whereas the luciferase activity was remarkably increased by LIF treatment in dif-

ferentiated F9 (dF9) cells. We analyzed the mechanism of the APRE regulation and found that the LIF-induced APRE-binding activity was regulated in a differentiation-dependent manner. The protein levels and the tyrosine phosphorylation of JAK1, JAK2, and STAT3 in F9 cells were not different from those in dF9 cells. The exogenous expression of activated c-Ha-ras partially recovered the LIF responsiveness of the APRE-luciferase gene in F9 cells, but the dominant negative ras N-17 did not repress the LIF-induced activation of APREluciferase in dF9 cells. These results suggested that an unknown coactivation process that is partially compensated by Ras is required for STAT3-APRE binding in F9 cells. (Endocrinology 138: 2689 –2696, 1997)

L

to respective receptor complexes, both of which contain the same signal transducer gp130 (12). Both cytokines use the recently characterized JAK protein tyrosine kinases (JAK1, JAK2, and Tyk2), as well as the STAT (signal transducer and activator of transcription) transcription factor signal transduction pathway and activate the same acute phase response element (APRE) in the rat a2-macroglobulin promoter by binding the acute phase response factor (APRF) (13–16). It is reported that although F9 cells have high affinity LIF receptors, as do ES cells (2), LIF has no biological effect on undifferentiated F9 cells (17). This suggests that the E1A-like activity might repress the LIF-JAK-STAT pathway in undifferentiated F9 cells. In this study, we examined whether the LIF-JAK-STAT signal transduction pathway is regulated in a differentiationdependent manner in F9 cells. We further analyzed the mechanism of this regulation.

EUKEMIA inhibitory factor (LIF) is a multifunctional cytokine that plays important roles in a wide range of biological activities, such as the synthesis of acute phase proteins in hepatocytes, osteoblast formation, hematopoiesis, and neuronal differentiation (1). In addition, LIF maintains the developmental potential of embryonic stem (ES) cells (2), and it has been reported that the transient expression of LIF is essential for implantation in mice (3). These facts suggest a critical role for this molecule in early embryogenesis. The embryonic carcinoma cell line, F9 is a useful model system with which to analyze early differentiation events that are similar to those of the early mammalian embryogenesis. Retinoic acid (RA)-treated F9 cells differentiate into parietal extraembryonic endoderm-like cells (4). The activated c-Ha-ras oncogene (Ha-rasVal-12) also induces the differentiation of F9 cells (5). Undifferentiated F9 cells contain an E1A-like activity that is lost upon RA-induced differentiation (6 – 8). The exogenous expression of the adenovirus E1A gene reverses the terminal differentiation and the expression of differentiation specific genes in F9 cells (7). A cellular E1A-like activity and an adenoviral E1A are thought to use similar mechanisms to repress the interferon-a (IFNa) signal transduction pathway (9). We showed that adenovirus E1A represses interleukin-6 (IL-6) induced gene activation in the hepatoma cell line HepG2 (10). LIF shares striking similarities with IL-6 in terms of biological activities (reviewed in Ref. 11). LIF and IL-6 bind Received January 2, 1997. Address all correspondence and requests for reprints to: Dr. Takashi Takeda, Department of Obstetrics and Gynecology, Osaka University Medical School, 2–2 Yamadaoka, Suita, Osaka 565, Japan.

Materials and Methods Cell and cell culture The embryonal carcinoma cell line F9 (18) was cultured in aMEM (Life Technologies, Grand Island, NY) containing 10% FCS (CSL, Parkville, Australia). Terminally differentiated F9 cells (dF9) were obtained by growing F9 cells in medium containing RA (0.1 mm; Sigma Chemical Co., St. Louis, MO) and (Bu)2cAMP (0.1 mm; Sigma).

Plasmids The 43APRE-luciferase gene containing four repeats of APRE (IL-6 response element of the rat a2-macroglobulin promoter), was constructed by subcloning two repeats of the oligonucleotides (59-TCGACATCCTTCTGGGAATTCTGATCCTTCTGGGAATTCTGGGTAC-39) (19) in front of the minimal junB promoter-luciferase gene as previously

2689

2690

DIFFERENTIATION-DEPENDENT REGULATION OF JAK-STAT PATHWAY

Endo • 1997 Vol 138 • No 7

described (10). The 43APRE-luciferase gene containing four repeats of mutated APRE (43mAPRE) was constructed by subcloning two repeats of the oligonucleotides (59-TCGACATCCTTCTCTAGATTCTGATCCTTCTCTAGATTCTGGGTAC-39) (13) in front of the minimal junB promoter-Luciferase gene. The activated c-Ha-ras (Ha-rasVal-12) expression vector was pSVEJ6.6 (20). The dominant negative ras N-17 mutant (Asn-17 ras) expression vector was pZIPRasN17 (21). The mouse c-jun expression vector was pRSV-c-jun (22).

nuclear extracts from 2 3 107 cells were immunoprecipitated with the appropriate antibodies. Immune complexes were separated by SDSPAGE (7.5%) and transferred to an Immobilon-P nylon membrane (Nihon Millipore, Tokyo, Japan), which was immunoblotted with the relevant primary antibody as previously described (10). Proteins were detected by enhanced chemiluminescence (ECL, Amersham International, Aylesbury, UK). Rehybridization was performed as previously described (26).

DNA transfection and luciferase assays

Results LIF-mediated induction of the APRE-luciferase gene is impaired in F9 cells

DNA transfection and luciferase assays were performed as previously described (10, 23). In each experiment, cells were seeded at 5 3 105 cells/6-cm dish, and 24 h later, 2.9 mg of the luciferase reporter plasmid, 1.0 mg pEFlacZ (23) (internal control for transfection efficiency), and 0.7–2.1 mg expression vector were used as required. The total amount of transfected DNA was adjusted to 6.1 mg with pTZ19R (Pharmacia, Milwaukee, WI). Cells were extracted, and luciferase activities were assayed in a MicroLumat luminometer (EG&G Berthold, Postfach, Germany) as previously described (10). b-Galactosidase activity was determined to normalize the transfection efficiency. All experiments were performed in triplicate and repeated at least three times with essentially similar results. Results are expressed as relative luciferase activity. Data are shown as averages of three independent experiments, with sds indicated by bars.

We tested the LIF-induced activation of the APRE-luciferase gene in undifferentiated and RA-differentiated F9 (dF9) cells. The 43APRE-luciferase or 43mAPRE-luciferase plasmid was transiently transfected into F9 and dF9 cells, and luciferase activity was assayed after 5 h of LIF stimulation (Fig. 1A). Transfection efficiency was normalized by b-galactosidase activity as described in Materials and Methods. Both cells showed almost the same transfection efficiency.

Cytokines and antibodies The cytokines used in this study were human recombinant LIF (Pepro Tech, Rocky Hill, NJ) and human recombinant IL-6 (a gift from Dr. T. Hirano, Osaka University Medical School, Osaka, Japan). Anti-ISGF3 (STAT1-a/b) monoclonal antibody (mAb; Transduction Laboratories, Lexington, KY) and the antiserum against STAT3 (a gift from Dr. K. Nakajima, Osaka University Medical School) have been previously described (10, 24). The antibodies used for Western blotting and immunoprecipitation were anti-JAK1 polyclonal antibody (HR-785, Santa Cruz Biotechnology, Santa Cruz, CA), anti-JAK2 polyclonal antibody (Upstate Biotechnology, Lake Placid, NY), antiphosphotyrosine mAb (4G10, Upstate Biotechnology), and anti-STAT3 polyclonal antibody (C-20, Santa Cruz Biotechnology).

Electrophoretic mobility shift assays (EMSAs) Nuclear and cytoplasmic extracts were prepared from F9 and dF9 cells according to the method of Sadowski et al. (25). The double stranded oligonucleotides used as probes or competitor in the EMSAs were 59GCGCCTTCTGGGAATTCCTA-39 and 59-GCGCTAGGAATTCCCAGAAG-39 (10). Binding reactions and electrophoresis proceeded as previously described (23). In the competition analysis, cell extracts were incubated with a 10-fold molar excess of unlabeled oligonucleotides for 5 min before adding the labeled oligonucleotides. In the supershift assays, anti-STAT1 mAb, anti-STAT3 antiserum, or a control rabbit serum was added to the binding reaction mixture in the indicated amount.

LIF binding assay LIF was iodinated, and binding was carried out as previously described (2). Five million F9 or dF9 cells were incubated at 37 C for 40 min in 100 ml RPMI (Life Technologies) containing 10% FCS, 20 mm HEPES (pH 7.4), and [125I]LIF (2 3 103 to 8 3 105 cpm; SA, ;40,000 cpm/ng) with or without excess amounts (500-fold that of [125I]LIF) of unlabeled LIF. After incubation, cell-associated and free radioactivity were separated by centrifugation, and the radioactivity was counted by a g-counter.

Immunoprecipitation, SDS-PAGE, and Western blotting Cells (2 3 107) were harvested and lysed for 30 min in 1 ml lysis buffer [50 mm Tris (pH 7.5), 150 mm NaCl, 1 mm EDTA, 0.5% Nonidet P-40, 1 mm sodium orthovanadate, 1 mm NaF, 0.75 mm phenylmethylsulfonylfluoride, 15% glycerol, and 10 mg/ml each of aprotinin, pepstatin, and leupeptin] as previously described (15). Proteins in the lysates or the

FIG. 1. LIF-mediated induction of the APRE-luciferase gene in undifferentiated and differentiated F9 cells. A, Undifferentiated (F9) and differentiated F9 (dF9) cells were transfected with the 43APREluciferase gene (APRE) or the 43mAPRE-luciferase gene (mAPRE). The transfected cells were not stimulated (2) or were stimulated (1) with LIF (10 ng/ml) for 5 h, and luciferase activity was determined. B, Dose-dependent activation of the APRE-luciferase gene by LIF in dF9 cells. The dF9 cells were transfected with the 43APRE-luciferase gene, then stimulated with LIF or IL-6 (at the concentrations indicated) for 5 h. Thereafter, luciferase activities were determined.

DIFFERENTIATION-DEPENDENT REGULATION OF JAK-STAT PATHWAY

The data are presented by calculating the luciferase activity of unstimulated F9 cells as 1. The basal APRE activity was slightly (;2.5-fold) increased by the differentiation process. Although F9 cells have high affinity LIF receptors (2, 17), the APRE-luciferase gene was not activated by LIF in F9 cells. In dF9 cells, however, LIF largely (;18-fold) activated APRE activity. As LIF did not activate the mAPRE-luciferase (Fig. 1A) and minimal junB promoter-luciferase (data not shown) genes in dF9 cells, LIF-induced activation of the APRE-luciferase gene is through activation of the APRE site. We examined the dose dependence of the LIF-induced APRE activation (Fig. 1B). LIF activated the APRE in a dose-dependent manner in dF9 cells. LIF-induced activation reached the maximal level at a concentration of 10 ng/ml. In F9 cells, APRE-luciferase was not activated even at a concentration of 50 ng/ml (data not shown). IL-6 also uses the gp130-JAK-STAT (STAT3) signal transduction pathway and activates the APRE (14, 15). We, therefore, examined whether this factor activates the APRE-luciferase gene in F9 and dF9 cells. IL-6 did not activate the APRE in F9 cells as LIF (data not shown). IL-6 activated the APRE to a much lesser extent than LIF in dF9 cells even at a concentration of 100 ng/ml (Fig. 1B). IL-6-induced APRE activation reached the maximal level at 100 ng/ml (data not shown). These findings suggest that the region upstream of the JAK kinase is impaired in IL-6 signal transduction pathways in dF9 cells. The a-subunit function of the IL-6 receptor may be impaired in dF9 cells. APRF activation is regulated in a differentiation-dependent manner in F9 cells

To examine which part of the LIF-JAK-STAT signal transduction pathway is disturbed in F9 cells, we investigated whether APRF is activated in LIF-stimulated F9 cells using EMSAs (Fig. 2, A and B). LIF-induced DNAbinding activity of dF9 cells was detected in the nucleus (Fig. 2A, lanes 3 and 4) and in the cytoplasm (Fig. 2B, lanes 3 and 4). This binding complex was competed out by an excess of the unlabeled APRE oligonucleotides (Fig. 2, A and B, lane 5). In F9 cells, LIF-induced DNA-binding activity was not detected in either the nucleus (Fig. 2A, lanes 1 and 2) or the cytoplasm (Fig. 2B, lanes 1 and 2). These results suggest the absence of an APRE activation process in F9 cells. The LIF-induced APRE binding complex in dF9 cells was supershifted by anti-STAT3, whereas anti-STAT1 and control serum had no effect (Fig. 2C). These results indicated that STAT3 is dominant in LIF-induced DNAbinding activity in dF9 cells. Tyrosine phosphorylation and the protein levels of JAK1, JAK2, and STAT3 are the same in undifferentiated F9 and dF9 cells

We analyzed which part of the LIF-JAK-STAT signal transduction pathway was disturbed in undifferentiated F9 cells. To examine the possible absence of LIF receptor in F9 cells, we conducted [125I]LIF binding studies (Fig. 3). Dissociation constants (Kd) were 100 pm in both F9 and dF9 cells. The maximal numbers of binding sites were 380 and 330/cell in

2691

F9 and dF9 cells, respectively. Specific [125I]LIF binding was observed at similar levels in both F9 and dF9 cells. These results suggested the disturbance of LIF-JAK-STAT pathway in undifferentiated F9 cells was not due to the absence of LIF receptors. Next we examined whether LIF induces the tyrosine phosphorylation of JAK1 and JAK2 in F9 and dF9 cells (Fig. 4, A and B). LIF induced the tyrosine phosphorylation of JAK1 and JAK2 in F9 cells as effectively as in dF9 cells (Fig. 4, A and B, upper panel), and the protein levels of these kinases in F9 and dF9 cells were the same (Fig. 4, A and B, lower panel). These results suggested that the LIF-induced JAK activation process is not disturbed and that functional LIF receptors are present in F9 cells. We then examined whether STAT3 is tyrosine phosphorylated in LIF-stimulated F9 cells, and the STAT3 protein levels in undifferentiated and differentiated F9 cells were compared (Fig. 5A). LIF induced the phosphorylation of STAT3 to a similar extent in F9 and dF9 cells, and the protein level in F9 cells was not different from that in dF9 cells. Besides STAT3 protein, we detected three tyrosine-phosphorylated proteins (190, 145, and 130 kDa). It has been reported that the LIF receptor b-subunit (LIFR-b) and gp130 coprecipitate with STAT3 under nonreducing conditions and that they correspond to the 190- and 145-kDa proteins, respectively (15). The similarity of the molecular mass suggests that the 130-kDa protein is a JAK kinase. These three proteins were also tyrosine phosphorylated by LIF stimulation in both F9 and dF9 cells to a similar extent. These findings indicate that in F9 cells, LIF induces the tyrosine phosphorylation of gp130, LIFR-b, JAKs, and STAT3. Under unstimulated conditions, a small amount of tyrosine-phosphorylated proteins was found in F9, but not dF9, cells (Fig. 5A). They were slightly activated in the unstimulated condition by an unknown mechanism. In the JAK-STAT signal transduction pathway, STATs are tyrosine phosphorylated in the cytoplasm and translocate to the nucleus (27). As it is possible that the translocation of the phosphorylated STAT3 protein from cytoplasm to nucleus is impaired in F9 cells, we analyzed whether phosphorylated STAT3 is present in the nucleus of undifferentiated cells. As shown in Fig. 5B, STAT3 phosphorylated by LIF stimulation was present in the nuclear protein fraction from undifferentiated F9 cells. These findings together with the results of the EMSA (Fig. 2) suggested that tyrosine-phosphorylated STAT3 cannot bind to the APRE site in undifferentiated F9 cells. The presence of a binding inhibitor or the absence of a coactivation mechanism in undifferentiated F9 cells is thus indicated. Negative transcriptional regulator was not found in F9 cells

Some human cancer cell lines express a negative transcriptional regulator, termed transcriptional knockout, that causes a lack of IFN-stimulated gene factor-3; thus, these cells show defects in IFN-induced gene expression (28). We examined whether F9 cells contain such a direct inhibitor that prevents APRF binding to APRE. The addition of the nuclear extracts prepared from untreated F9 cells had no effect on the LIF-induced binding complex in dF9 cells (Fig. 6, lanes 1 and

2692

DIFFERENTIATION-DEPENDENT REGULATION OF JAK-STAT PATHWAY

Endo • 1997 Vol 138 • No 7

FIG. 2. Differentiation-dependent activation of APRF in F9 cells. Undifferentiated F9 (lanes 1 and 2) and dF9 (lanes 3–5) cells were not stimulated (lanes 1 and 3) or were stimulated (lanes 2, 4, and 5) with LIF (10 ng/ml) for 15 min, then nuclear and cytoplasmic extracts were prepared. For each reaction, 12 mg nuclear (A) or 20 mg cytoplasmic (B) extracts were incubated with a 32P-labeled APRE probe. Competition analysis (lane 5) proceeded as described in Materials and Methods. The reactions were followed by electrophoresis in a 4.5% polyacrylamide gel. The arrows in A and B indicate the position of APRF in dF9 cells. C, Analysis of APRE-binding proteins in dF9 cells. Nuclear extract (12 mg protein) prepared from dF9 (lanes 1–5) cells stimulated with LIF (10 ng/ml) for 15 min was incubated with an APRE probe. Competition analysis (lane 2) proceeded as described in Materials and Methods. Where indicated, control rabbit serum (cont Ab, lane 3), anti-STAT3 antiserum (aSTAT3, lane 4), or anti-STAT1 mAb (aSTAT1, lane 5) was added at final dilutions of 1:50, 1:50, and 1: 5, respectively. Arrow 1 indicates the APRF in dF9 cells, as described in A and B. Arrow 2 indicates the position of the supershifted complex of APRF by anti-STAT3 antiserum.

2). Moreover, addition of the nuclear extracts from LIFtreated F9 cells did not affect the LIF-induced DNA-binding activity in dF9 cells (data not shown). These results suggested that a negative regulator, such as transcriptional knockout, is not present in F9 cells. In IFN-a signaling, E1A-mediated repression is attributed to a defect in the DNA-binding subunit (IFN-stimulated gene factor-3 g component) that is restored by the complementation of this component (29). F9 cells might have such a defective DNA-binding subunit. Therefore, we tested whether the nuclear extract from the untreated dF9 cells restores the LIF-induced DNA-binding activity in F9 cells (Fig. 6, lanes 3 and 4). Addition of the nuclear extract prepared from the dF9 cells did not restore the LIF-induced DNA-binding activity in F9 cells, suggesting that F9 cells do not have a defective DNA-binding subunit.

Ha-rasVal-12 restored the LIF-induced APRE activation in F9 cells, but the Ras itself was not involved in LIF-JAKSTAT pathway in dF9 cells

The above data suggest a possible defect in the coactivation pathway for APRE activation in undifferentiated cells. As the stable expression of Ha-rasVal-12 induces the differentiation of F9 cells to endoderm-like cells as does RA (5), it was interesting to examine whether Ha-rasVal-12 could induce the LIF responsiveness of the APRE-luciferase gene in F9 cells. The 43APRE-luciferase or 43mAPRE-luciferase plasmid was transiently transfected with a Ha-rasVal-12 expression vector (0.7 mg) into F9 cells, and the luciferase activities were assayed after 5 h of LIF stimulation. As shown in Fig. 7A, the LIF-induced APRE activity was partially restored by Ras expression; although the LIF-induced APRE activation

DIFFERENTIATION-DEPENDENT REGULATION OF JAK-STAT PATHWAY

2693

FIG. 3 Scatchard analysis of [125I]LIFspecific binding in F9 and dF9 cells. Cells were incubated with increasing amounts of [125I]LIF as described in Materials and Methods. The ratio of the amount of [125I]LIF specifically bound vs. free [125I]LIF was plotted. Scatchard plots for F9 cells (A) and dF9 cells (B) are shown.

FIG. 4. Tyrosine phosphorylation of JAK family protein kinases by LIF. F9 and dF9 cells were incubated for 10 min without (2) or with (1) LIF (50 ng/ml). Cell lysates were immunoprecipitated with anti-JAK1 (A) or anti-JAK2 (B). Immune complexes were separated by SDS-PAGE and analyzed by blotting with antiphosphotyrosine (upper panel). The blots were then stripped and reprobed sequentially with anti-JAK1 (A, lower panel) or anti-JAK2 (B, lower panel).

was less in the Ras-expressed F9 cells than in dF9 cells (;18fold, see Fig. 1), approximately 4-fold induction was obtained by Ha-rasVal-12 transfection in F9 cells. As LIF failed to induce mAPRE activity by Ras expression (Fig. 7A), Ras-induced restoration of the APRE-luciferase gene may be through activation of the APRE site. When we transfected higher doses of Ras expression vector (up to 2.2 mg), the LIF responsiveness of the APRE-luciferase in F9 cells did not change (data not shown). The dominant negative Ras (N-17) (22) did not restore the LIF responsiveness of APRE (Fig. 7A). These results show that the Ras-dependent pathway might be involved in the coactivation process of the APRE activation in F9 cells. Next we examined the effect of the N-17 on the LIF-induced activation of APRE in dF9 cells (Fig. 7B). The 43APRE-luciferase was transiently transfected with an N-17 expression vector (2.1 mg) into dF9 cells, and the luciferase activities were assayed after 5 h of LIF stimulation. As shown in Fig. 7B, the LIF-induced APRE activity was not repressed by N-17. The effect of N-17 in dF9 cells was confirmed by the fact that N-17 completely repressed the c-jun-induced activation of the activator protein-1 site-luciferase (30) (data not shown). These results suggested that Ras itself was not involved in the LIF-JAK-STAT signal transduction pathway. An unknown coactivation process that could be partially

compensated by Ras expression may be involved in this process. Discussion

Here, we showed that 1) the LIF-JAK-STAT signal transduction pathway is blocked in undifferentiated F9 cells; 2) LIF-induced APRF activation is regulated in a RA-induced, differentiation-dependent manner; 3) tyrosine-phosphorylated STAT3 is present in the nuclei of undifferentiated F9 cells, but it does not bind to the APRE site; 4) the exogenous expression of Ha-rasVal-12 partially rescues the responsiveness to LIF in undifferentiated F9 cells; but 5) Ras itself is not involved in LIF-JAK-STAT signal transduction pathway in dF9 cells. F9 cells have high affinity LIF receptors, like ES cells, and both cells bind similar levels of LIF (2). Although LIF is necessary for maintenance of the developmental potential of ES cells (2, 31), LIF has no biological effect on F9 cells (17). Both F9 and dF9 cells have the same number of LIF receptors and level of affinity (17). Moreover, under our conditions, LIF activated JAK1 and JAK2 in F9 cells to the same extent as in dF9 cells, suggesting that the LIF receptor is also functional in F9 cells. Therefore, the LIF-JAK-STAT signal transduction

2694

DIFFERENTIATION-DEPENDENT REGULATION OF JAK-STAT PATHWAY

FIG. 5. Tyrosine phosphorylation of STAT3 by LIF. F9 and dF9 cells were treated for 10 min without (2) or with (1) LIF (50 ng/ml). Cell lysates (A) or nuclear extracts (B) were immunoprecipitated with anti-STAT3. Immune complexes were separated by SDS-PAGE and blotted with antiphosphotyrosine (A, upper panel, and B). The blots were then stripped and reprobed with anti-STAT3 (A, lower panel).

pathway in F9 cells is likely to be blocked downstream of the LIF receptor. Our data showed that the critical step of this repression was the disturbance of tyrosine-phosphorylated STAT3 binding to the APRE site and suggested the presence of an unknown coactivation process that is regulated in a differentiation-dependent manner. It has been reported that the H7 (serine/threonine kinase inhibitor)-sensitive pathway is involved in the IL-6, LIF signal transduction pathway (23, 32, 33) and that serine phosphorylation at the mitogen-activated protein kinase (MAPK) site of STAT3 is required for STAT3-DNA complex formation (34). Therefore, an H7-sensitive coactivation process may be required for STAT3-DNA complex formation, which is disturbed in undifferentiated F9 cells. As Ras activates the Raf/ MAPK signal transduction pathway (35–37), we studied the exogenous expression of Ha-rasVal-12 in F9 cells. We found that Ras restored LIF-induced APRE activation in undifferentiated F9 cells, suggesting that STAT3 phosphorylation at the MAPK site may be required to bind to APRE. We also found that the LIF-JAK-STAT pathway was Ras independent

Endo • 1997 Vol 138 • No 7

FIG. 6. Absence of a negative transcriptional regulator in F9 cells. Nuclear extract from untreated F9 cells (F9 2, 8 mg) was mixed with nuclear extract (8 mg) from untreated (dF9 2) or 10 ng/ml LIF-treated (dF9 1) dF9 cells (lanes 1 and 2). Nuclear extract from untreated dF9 cells (dF9 2, 8 mg) was mixed with nuclear extract (8 mg) from untreated (F9 2) or 10 ng/ml LIF-treated (F9 1) F9 cells (lanes 3 and 4). These nuclear extract mixtures were incubated with a 32P-labeled APRE probe. The arrow indicates the same complexes as those described in Fig. 3, A and B.

in dF9 cells. Therefore, Ras itself seemed not to be involved in the coactivation pathway in dF9 cells. A Ras-independent and MAPK-dependent pathway was reported in 3T3-L1 adipocytes in the insulin signal pathway (38), and such alternate routes of signal transduction may be involved in the coactivation pathway in dF9 cells. We cannot rule out the possibility that the Ras effect is indirect, and the restoration is the result of differentiation itself, but in the previous study, the Ras-induced differentiation of F9 cells was observed in the stable transfectants and took at least 4–6 days for the morphological differentiation (5). In our study the expression of Ras was transient, the time from transfection through harvest was 40 h, and morphological differentiation was not observed. As Ras could not fully activate LIF-induced APRE in F9 cells, another coactivation process(es) might be considerable. A coactivator, cAMP response element binding protein

DIFFERENTIATION-DEPENDENT REGULATION OF JAK-STAT PATHWAY

FIG. 7. Presence of an unknown coactivation pathway in the LIFJAK-STAT pathway. A, Ha-Ras expression recovered the LIF responsiveness in F9 cells. F9 cells were transfected with the 43APREluciferase (APRE) or 43mAPRE-luciferase (mAPRE) gene and the expression vector, pSVEJ6.6 (Ha-Ras), or its dominant negative mutant pZIPRasN17 (N17). B, Dominant negative Ras (N17) did not repress the LIF-induced activation of APRE-luciferase in dF9 cells. The dF9 cells were transfected with 43APRE-luciferase and pZIPRasN17 (N17). The transfected cells were not stimulated (2) or were stimulated (1) with LIF (10 ng/ml) for 5 h. Luciferase activity was then determined.

(CREB)-binding protein (CBP) has been found in the protein kinase A-cAMP response element binding protein signal transduction pathway (39). CBP has a striking homology with p300, and adenoviral E1A binds both proteins (40, 41). Recently, it was reported that RA induces serine and threonine phosphorylation of p300 during the differentiation of F9 cells (42), and that p300/CBP cooperate with STAT2 in the IFN-a pathway (43) and with STAT1 in the IFN-g (44) signal transduction pathway. Because the LIF-JAK-STAT3 pathway closely resembles the IFN-g-JAK-STAT1 pathway, it might be interesting to examine whether a coactivator such as CBP or p300 is involved in the LIF-JAK-STAT3 pathway. It might be possible that in undifferentiated F9 cells, the p300-like coactivator is inactive; thus, the tyrosine-phosphorylated STAT3 cannot bind to the APRE site. EMSA analysis in this study showed that in F9 cells the LIF-induced APRE-binding activity was not present, and complementation of the extracts from dF9 cells could not

2695

restore this defect. These results were consistent with those of our previous study (10); in a rat fibroblast cell line stably expressing adenoviral E1A (45), the IL-6-induced STAT3 binding to APRE was greatly reduced compared with that in the parental 3Y1 cells (46). A reduction of the STAT3 protein level in E1A3Y1 cells is one of causative factors for the reduction in APRE binding. In this study, however, we observed a distinct difference; the STAT3 protein level in F9 cells with cellular E1A was not different from that in dF9 cells without E1A. These inconsistencies might suggest a difference between the cellular E1A activity present in F9 cells and the adenoviral E1A. However, in the E1A3Y1 cells, exogenous expression of STAT3 protein did not restore the IL-6induced APRE activation (our unpublished data), suggesting that the reduction in the STAT protein levels may only partly cause the repressive effect of the adenoviral E1A. The adenoviral E1A may also block the coactivation process of STAT-APRE binding as does the cellular E1A of undifferentiated F9 cells. LIF-deficient mice generated by gene targeting have shown that LIF is essential for implantation (3). LIF is expressed in the uterine endometrial glands on the fourth day of pregnancy in mice (47). The true target of the maternal LIF is not yet known, but it may directly affect the blastocyst to be implanted. The RA-induced differentiation process of F9 cells is thought to correspond to the events that occur around the day of implantation (4, 48). This differentiation process is thought to correspond to the event in the inner cell mass of the 4.5-day postcoitum embryo (4, 48). It is possible that the maternal LIF stimulates the JAK-STAT signal transduction pathway in the preimplantation embryo and that this signal is critical for implantation. An adenovirus E1A-like activity is present in preimplantation stage mouse embryos (49). The LIF-JAK-STAT pathway may be blocked in these cells. The differentiation-dependent activation of the LIF-JAK-STAT pathway may occur in the early embryo, and its regulation may play an important role in implantation. It would be of interest to test whether the regulation found in F9 cells also exists in the early embryo. Further examination of this mechanism will provide new insight into the regulation of early embryogenesis and implantation. Acknowledgments We are grateful to Dr. T. Hirano for the gift of the IL-6, and to Dr. K. Nakajima for providing the anti-STAT3 antiserum, ras, and c-jun expression vectors. We thank Ms. I. Iida and Ms. K. Ogami for excellent secretarial assistance.

References 1. Kurzrock R, Estrov Z, Wetzler M, Gutterman JU, Talpaz M 1991 LIF: not just a leukemia inhibitory factor. Endocr Rev 12:208 –217 2. Williams RL, Hilton DJ, Pease S, Willson TA, Stewart CL, Gearing DP, Wagner EF, Metcalf D, Nicola NA, Gough NM 1988 Myeloid leukaemia inhibitory factor maintains the developmental potential of embryonic stem cells. Nature 336:684 – 687 3. Stewart CL, Kaspart P, Brunet LJ, Bhatt H, Gadi I, Ko¨ntgen F, Abbondanzo SJ 1992 Blastocyst implantation depends on maternal expression of leukaemia inhibitory factor. Nature 359:76 –79 4. Strickland S, Smith KK, Marotti KR 1980 Hormonal induction of differentiation in teratocarcinoma stem cells: generation of parietal endoderm by retinoic acid and dibutyryl cAMP. Cell 21:347–355 5. Yamaguchi-Iwai Y, Satake M, Murakami Y, Sakai M, Muramatsu M, Ito Y

2696

6. 7. 8. 9. 10. 11. 12. 13.

14. 15.

16.

17. 18. 19. 20. 21. 22. 23.

24.

25. 26. 27.

DIFFERENTIATION-DEPENDENT REGULATION OF JAK-STAT PATHWAY

1990 Differentiation of F9 embryonal carcinoma cells induced by the c-jun and activated c-Ha-ras oncogenes. Proc Natl Acad Sci USA 87:8670 – 8674 Weigel RJ, Nevins JR 1990 Adenovirus infection of differentiated F9 cells results in a global shut-off of differentiation-induced gene expression. Nucleic Acids Res 18:6107– 6112 Weigel RJ, Devoto SH, Nevins JR 1990 Adenovirus 12S E1A gene represses differentiation of F9 teratocarcinoma cells. Proc Natl Acad Sci USA 87:9878 –9882 LaThangue NB, Thimmappaya B, Rigby PWJ 1990 The embryonal carcinoma stem cell E1a-like activity involves a differentiation-regulated transcription factor. Nucleic Acids Res 18:2929 –2938 Kalvakolanu DVR, Sen GC 1993 Differentiation-dependent activation of interferon-stimulated gene factors and transcription factor NF-kB in mouse embryonal carcinoma cells. Proc Natl Acad Sci USA 90:3167–3171 Takeda T, Nakajima K, Kojima H, Hirano T 1994 E1A repression of IL-6Induced gene activation by blocking the assembly of IL-6 response element binding complexes. J Immunol 153:4573– 4582 Hirano T, Matsuda T, Nakajima K 1994 Signal transduction through gp130 that is shared among the receptors for the interleukin 6 related cytokine subfamily. Stem Cells 12:262–277 Hibi M, Murakami M, Saito M, Hirano T, Taga T, Kishimoto T 1990 Molecular cloning and expression of an IL-6 signal transducer, gp130. Cell 63:1149 –1157 Wegenka UM, Buschmann J, Lu¨tticken C, Heinrich PC, Horn F 1993 Acutephase response factor, a nuclear factor binding to acute-phase response elements, is rapidly activated by interleukin-6 at the posttranslational level. Mol Cell Biol 13:276 –288 Zhong Z, Wen Z, Darnell Jr JE 1994 Stat3: a STAT family member activated by tyrosine phosphorylation in response to epidermal growth factor and interleukin-6. Science 264:95–98 Lu¨tticken C, Wegenka UM, Yuan J, Buschmann J, Schindler C, Ziemiecki A, Harpur AG, Wilks AF, Yasukawa K, Taga T, Kisimoto T, Barbieri G, Pellegrini S, Sendtner M, Heinrich PC, Horn F 1994 Association of transcription factor APRF and protein kinase Jak1 with the interleukin-6 signal transducer gp130. Science 263:89 –92 Stahl N, Boulton TG, Farruggella T, Ip NY, Davis S, Witthuhn BA, Quelle FW, Silvennoinen O, Barbieri G, Pellegrini S, Ihle JN, Yancopoulos GD 1994 Association and activation of Jak-Tyk kinases by CNTF-LIF-OSM-IL-6 b receptor components. Science 263:92–95 Brown GS, Brown MA, Hilton D, Gough NM, Sleigh MJ 1992 Inhibition of differentiation in a murine F9 embryonal carcinoma cell subline by leukemia inhibitory factor (LIF). Growth Factors 7:41–52 Bernstine EG, Hooper ML, Grandchamp S, Ephrussi B 1973 Alkaline phosphatase activity in mouse teratoma. Proc Natl Acad Sci USA 70:3899 –3903 Hocke GM, Barry D, Fey GH 1992 Synergistic action of interleukin-6 and glucocorticoids is mediated by the interleukin-6 response element of the rat a2 macroglobulin gene. Mol Cell Biol 12:2282–2294 Tabin CJ, Bradley SM, Bargmann CI, Weinberg RA, Papageorge AG, Scolnick EM, Dhar R, Lowy DR, Chang EH 1982 Mechanism of activation of a human oncogene. Nature 300:143–149 Feig LA, Cooper GM 1988 Inhibition of NIH 3T3 cell proliferation by mutant ras protein with preferential affinity for GDP. Mol Cell Biol 8:3235–3243 Hirai SI, Ryseck RP, Mechta F, Bravo R, Yaniv M 1989 Characterization of junD: a new member of the jun proto-oncogene family. EMBO J 8:1433–1439 Nakajima K, Kusafuka T, Takeda T, Fujitani Y, Nakae K, Hirano T 1993 Identification of a novel interleukin-6 response element containing an Ets binding site and a CRE-like site in the junB promoter. Mol Cell Biol 13:3027–3041 Fujitani Y, Nakajima K, Kojima H, Nakae K, Takeda T, Hirano T 1994 Transcriptional activation of the IL-6 response element in the junB promoter is mediated by multiple STAT family proteins. Biochem. Biophys Res Commun 202:1181–1187 Sadowski HB, Shuai K, Darnell Jr JE, Gilman MZ 1993 A common nuclear signal transduction pathway activated by growth factor. Science 261:1739 –1744 Ruff-Jamison S, Chen K, Cohen S 1993 Induction by EGF and interferon-g of tyrosine-phosphorylated DNA-binding proteins in mouse liver nuclei. Science 261:1733–1736 Darnell Jr JE, Kerr IM, Stark GR 1994 Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins. Sciense 264:1415–1421

Endo • 1997 Vol 138 • No 7

28. Petricoin E, David M, Fang H, Grimley P, Larner AC, Pol SV 1994 Human cancer cell lines express a negative transcriptional regulator of the interferon regulatory factor family of DNA binding proteins. Mol Cell Biol 14:1477–1486 29. Kalvakolanu DVR, Bandyopadhyay SK, Harter ML, Sen GC 1991 Inhibition of interferon-inducible gene expression by adenovirus E1A proteins: block in transcriptional complex formation. Proc Natl Acad Sci USA 88:7459 –7463 30. Nakae K, Nakajima K, Inazawa J, Kitaoka T, Hirano T 1995 ERM, a PEA3 subfamily of Ets transcription factors, can cooperate with c-Jun. J Biol Chem 270:23795–23800 31. Hooper M, Hardy K, Handyside A, Hunter S, Monk M 1987 HPRT-deficient (Lesch-Nyhan) mouse embryos derived from germline colonization by cultured cells. Nature 326:292–295 32. Lord KA, Abdollahi A, Thomas SM, DeMarco M, Brugge JS, HoffmanLiebermann B, Liebermann DA 1991 Leukemia inhibitory factor and interleukin-6 trigger the same immediate early response, including tyrosine phosphorylation, upon induction of myeloid leukemia differentiation. Mol Cell Biol 11:4371– 4379 33. Nakajima K, Matsuda T, Fujitani Y, Kojima H, Yamanaka Y, Nakae K, Takeda T, Hirano T 1995 Signal transduction through IL-6 receptor: involvement of multiple protein kinases, STAT factors and a novel H7-sensitive pathway. Ann NY Acad Sci 762:55–70 34. Zhang X, Blnis J, Li HC, Schindler C, Chen-Kiang S 1995 Requirement of serine phosphorylation for formation of STAT-promoter complexes. Science 267:1990 –1994 35. Wood KW, Sarnecki C, Roberts TM, Blenis J 1992 ras mediates growth factor receptor modulation of three signal-transducing protein kinases: MAP kinase, Raf-1, and RSK. Cell 68:1041–1050 36. Thomas SM, DeMarco M, D’Arcangelo G, Halegoua S, Brugge JS 1992 Ras is essential for nerve growth factor- and phorbol ester-induced tyrosine phosphorylation of MAP kinases. Cell 68:1031–1040 37. Howe LR, Leevers SJ, Go´mez N, Nakielny S, Cohen P, Marshall CJ 1992 Activation of the MAP kinase pathway by the protein kinase raf. Cell 71:335–342 38. Carel K, Kummer JL, Schubert C, Leitner W, Heidenreich KA, Draznin B 1996 Insulin stimulates mitogen-activated protein kinase by a Ras-independent pathway in 3T3–L1 adipocytes. J Biol Chem 271:30625–30630 39. Chrivia JC, Kwok RPS, Lamb N, Hagiwara M, Montminy MR, Goodman RH 1993 Phoshorylated CREB binds specifically to the nuclear protein CBP. Nature 365:855– 859 40. Arany Z, Newsome D, Oldread E, Livingston DM, Eckner R 1995 A family of transcriptional adaptor proteins targeted by the E1A oncoprotein. Nature 374:81– 84 41. Lundblad JR, Kwok RPS, Laurance ME, Harter ML, Goodman RH 1995 Adenoviral E1A-associated protein p300 as a functional homologue of the transcriptional co-activator CBP. Nature 374:85– 88 42. Kitabayashi I, Eckner R, Arany Z, Chiu R, Gachelin G, Livingston DM, Yokoyama K 1995 Transcriptional regulation of the c-jun gene by retinoic acid and E1A during differentiation of F9 cells. EMBO J 14:3496 –3509 43. Bhattacharya S, Eckner R, Grossman S, Oldread E, Arany Z, D’Andrea A, Livingston DM 1996 Cooperation of Stat2 and p300/CBP in signalling induced by interferon-a. Nature 383:344 –347 44. Horvai AE, Xu L, Korzus E, Brard G, Kalafus D, Mullen TM, Rose DW, Rosenfeld MG, Glass CK 1997 Nuclear integration of JAK/STAT and Ras/ AP-1 signaling by CBP and p300. Proc Natl Acad Sci USA 94:1074 –1079 45. Shimura H, Ohtsu M, Matsuzaki A, Mitsudomi T, Onodera K, Kimura G 1988 Selective cytotoxicity of phospholipids and diacyl glycerols to rat 3Y1 fibroblasts transformed by adenovirus type 12 or its E1A gene. Cancer Res 48:578 –583 46. Kimura G, Itagaki A, Summers J 1975 Rat cell line 3Y1 and its virogenic polyoma- and SV40-transformed derivatives. Int J Cancer 15:694 –706 47. Bhatt H, Brunet LJ, Stewart CL 1991 Uterine expression of leukemia inhibitory factor coincides with the onset of blastocyst implantation. Proc Natl Acad Sci USA 88:11408 –11412 48. Enders AC, Given RL, Schlafke S 1978 Differentiation and migration of endoderm in the rat and mouse at implantation. Anat Rec 190:65–77 49. Suemori H, Hashimoto S, Nakatsuji N 1988 Presence of the adenovirus E1A-like activity in preimplantation stage mouse embryos. Mol Cell Biol 8:3553–3555