Interferon-gamma-induced dephosphorylation of STAT3 and apoptosis are dependent on the mTOR pathway

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / y e x c r

Research Article

Interferon-gamma-induced dephosphorylation of STAT3 and apoptosis are dependent on the mTOR pathway Peng Fang a,⁎, Vivian Hwa b , Ron G. Rosenfeld a,b,c a

Department of Pediatrics, Oregon Health and Science University, Portland, OR 97239-3098, USA Lucile Packard Foundation for Children's Health, Palo Alto, CA 94304, USA c Department of Pediatrics, Stanford University, Stanford, CA 94305, USA b

ARTICLE INFORMATION

ABS T R AC T

Article Chronology:

Interferon-gamma (IFN-γ) exhibits diverse biological activities, including control of cell

Received 25 August 2005

growth and tumor suppression. Here, we report that the treatment of M12 cells, a human

Revised version received

metastatic prostate cancer cell line, with IFN-γ, resulted in marked inhibition of cell

9 November 2005

proliferation and induced apoptosis. These effects were not seen with either IFN-α or IFN-β.

Accepted 6 December 2005

M12 cells, like many other human cancer cells, contain constitutively activated signal

Available online 19 January 2006

transducer and activator of transcription 3 (STAT3). The basal levels of both Akt and ERK1/2

Keywords:

and growth inhibition of M12 cells were associated with persistent suppression of the

Interferon-gamma

constitutive tyrosine-phosphorylated STAT3 (pY-STAT3). The IFN-γ-induced

STAT3

dephosphorylation of pY-STAT3, however, was inhibited when the mTOR pathway was

STAT1

specifically blocked by rapamycin. Inhibition of PI-3K with low-dose LY294002, or MAPK with

mTOR

PD98059 also suppressed the mTOR/p70 S6k pathway, and correlated with the blockage of

Apoptosis

IFN-γ-induced dephosphorylation of pY-STAT3. Simultaneously, treatment with LY294002,

Abbreviations:

mTOR pathway, however, did not affect IFN-γ-induced activation of STAT1 pathway, and

IFN-γ, interferon-gamma

suppression of STAT1 expression by siRNA had no effect on IFN-γ-induced

IFN-α, interferon-alpha

dephosphorylation of pY-STAT3. Taken together, these results demonstrate that an intact

phosphorylation are also markedly elevated in M12 cells. Strikingly, IFN-γ-induced apoptosis

PD98059, or rapamycin abolished IFN-γ-induced apoptosis in M12 cells. The inhibition of the

IFN-β, interferon-beta

mTOR pathway is critical for IFN-γ-induced suppression of pY-STAT3 and apoptosis. Our

JAK, Janus kinase

study thus provides novel insights into the contributions of signaling pathways other than

STAT, signal transducers and

the classical JAK/STAT1 pathway in the anti-proliferative, proapoptotic actions of IFN-γ.

activators of transcription SOCS, the suppressors of cytokine signaling siRNA, small interfering RNA PI-3K, phosphatidylinositol 3-kinase MAPK, mitogen-activated protein kinases ERK, extracellular signal-regulated kinase IRF-1, interferon regulatory factor-1 mTOR, mammalian target of rapamycin IL, interleukin

⁎ Corresponding author. Fax: +1 503 494 0428. E-mail address: [email protected] (P. Fang). 0014-4827/$ – see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2005.12.011

© 2005 Elsevier Inc. All rights reserved.

1230

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

Introduction

Materials and methods

Interferon-gamma (IFN-γ), in addition to its well-known antiviral activities, also exhibits anti-proliferative, proapoptotic effects on various tumor cells [1], including prostate cancer cells [2]. The anti-proliferative, proapoptotic effects of IFN-γ have been largely attributed to the activation of the Janus kinase (JAK) and Signal Transducer and Activator of Transcription-1 (STAT1) signaling pathway [3–5]. The cascade of signal transduction is initiated upon the binding of dimeric IFN-γ to its receptor, followed by the activation of the receptorassociated Jak1 and Jak2, which in turn phosphorylate tyrosines on the receptor. The phospho-tyrosines serve as the docking sites for multiple cytosolic proteins, including STAT1. Subsequent tyrosine-phosphorylation of STAT1 by the Jaks leads to homodimerization of phosphorylated STAT1 and translocation into the nucleus, where STAT1 functions as a transcription factor. In addition to the JAK/STAT pathway, IFN-γ can also activate the MAPK pathway and the PI-3K/mTOR pathway (reviewed recently in [6]). The role of these pathways in IFNγ-induced biological effects, however, has been less clearly defined. The mTOR pathway, critically involved in cell proliferation, cell cycle progression, and cell survival, regulates ribosome biogenesis and protein synthesis in response to the availability of nutrients and stimulation from growth factors and cytokines [7]. The upstream signaling pathways affecting the activity of mTOR include the PI-3K and MAPK/ERK pathways, both of which have been shown to regulate the mTOR pathway through modulating the activity of the Tuberous Sclerosis tumor suppressor (TSC1/2) complex [7–9]. The direct targets activated by mTOR include p70 S6K, 4EBP1, and eIF4G [7]. The activation of the mTOR pathway by IFN-γ appears to be mediated through the PI-3K pathway, while STAT1 activation and gene transcription via GAS elements are mTOR-independent [10]. The majority of evidence indicates that an activated mTOR pathway positively regulates cell growth and protects cells from apoptosis [7]. However, one recent study suggests that, on the contrary, an intact PI3-K/mTOR pathway was required by IFN-α to induce apoptosis in multiple myeloma cells, although the underlying mechanism(s) remains to be established [11]. In normal cells, STAT activation is transient and tightly controlled, whereas in a large number of primary tumors and cancer-derived cell lines, including prostate cancer cells, STAT3, in particular, is deregulated and remains constitutively activated [12–14]. STAT3 has been considered a potential oncogene because of its ability to induce tumorigenic transformation and block apoptosis [15]. Direct inhibition of constitutively tyrosine-phosphorylated STAT3 can elicit growth inhibition and apoptosis in various cancer cells, including prostate cancer cells [16,17]. In this study, we demonstrate that, in M12 cells, a human metastatic prostate cancer cell line, the IFN-γ-induced apoptosis, and inhibition of cell proliferation are associated with persistent suppression of constitutively tyrosine-phosphorylated STAT3, and furthermore, an intact mTOR pathway is required for the IFN-γ-induced apoptosis and suppression of STAT3.

Cell culture M12 cells are a highly metastatic derivative of a SV40-T antigen transformed, low tumorigenic human prostate cancer cell line [18]. M12 cells were cultured in defined RPMI 1640 supplemented with 5 μg/ml insulin–transferrin– sodium selenite (ITS, Sigma, Aldrich, St. Louis, MO), 0.2 μM dexamethasone, 10 ng/ml epidermal growth factor (EGF) at 37°C and 5% CO2. For experiments detecting changes in signaling pathways in response to various treatments, M12 cells were grown to 60–70% confluence on 6-well plate and starved in RPMI 1640 overnight before being incubated in RPMI 1640 supplemented with 50 ng/ml bovine serum albumin (BSA) plus human IFN-αA/D (Sigma-Aldrich, St. Louis, MO), human IFN-β (Sigma-Aldrich, St. Louis, MO), human IFN-γ (Roche, Mannheim, Germany), LY294002, PD98059 (Calbiochem, La Jolla, CA), or Rapamycin (Cell Signaling Technology, Beverly, MA) as indicated. DU145 cells or SK-HEP-1 cells were grown to 80% confluency in RPMI medium supplemented with 10% fetal bovine serum, starved and treated with IFN-γ in the presence of LY294002 or Rapamycin as described for M12 cells.

Cell proliferation assay M12 cells were plated on 96-well plates (2900 cells/well) and grown for 16 h. The cells were then synchronized in RPMI for 4 h before treatment with defined RPMI culture medium supplemented with 50 ng/ml BSA plus either IFNα, IFN-β, IFN-γ, or LY294002 at the concentrations indicated. Cell proliferation was quantitated at 48 h, 72 h, or 96 h post-treatment with the CellTiter 96® AQueous NonRadioactive Cell Proliferation Assay kit (MTS/PMS assay) (Promega, Madison, WI) according to the manufacturer's instructions.

Apoptosis assay M12 cells (2900 cells/well) or SK-HEP-1 cells (3000 cells/well) were plated, cultured, and treated on 96-well plates as described above. Apoptosis-induced DNA fragmentation was quantitated at 48 h, 72 h, or 96 h post-treatment with a Cell Death Detection ELISAPLUS assay system (Roche Diagnostics, Indianapolis, IN) according to the manufacturer's protocol. The assay quantitatively measures cytoplasmic histoneassociated DNA fragments (mono- and oligo-nucleosomes) generated in the early phase of apoptosis.

Western immunoblot Cells were washed once with ice-cold phosphate-buffered saline and lysed for 20 min at 4°C in the buffer containing 50 mM Tris–HCl (pH7.4), 1% NP-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 1 mM NaF, and a cocktail of protease inhibitors (Roche, Indianapolis, IN). Western immunoblot analysis was performed as described previously [19]. The density of the bands on X-ray film was determined by using

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

1231

IMAGEQUANT 5.1 (Molecular Dynamics). The antibodies used for Western immunoblot analyses in this study were: rabbit polyclonal IgG against phospho-Tyr701-STAT1, phosphoTyr705-STAT3, STAT3, phospho-Ser473-Akt, Akt, phosphothreonine389 p70 S6K, phospho-threonine421, serine 424 p70 S6K, or p70 S6K from Cell Signaling Technology (Beverly, MA); rabbit polyclonal anti-phospho-Ser727-STAT1 from Biosource International (Camarillo, CA); rabbit polyclonal anti-IRF-1 (C-20) from Santa Cruz Biotechnology (Santa Cruz, CA); mouse monoclonal IgG against phospho-Thr202-Tyr204ERK1/2 from Cell Signaling Technology; mouse monoclonal anti-STAT1 (C-136) from Santa Cruz Biotechnology (Santa Cruz, CA); mouse monoclonal anti-ERK2 from Upstate Biotechnology (Lake Placid, NY). Secondary antibodies (antimouse IgG and anti-rabbit IgG) were obtained from Amersham Biosciences (Piscataway, NJ). All immunoblot data shown are representative of at least four independent experiments.

Small-interfering RNA transfection M12 cells were seeded at 60% confluency on 12-well plates. After 16 h of incubation, siRNA transfection was performed. For each well, 4 μl TransIT-TKO® Transfection Reagent (Mirus, Madison, WI) was mixed with 50 μl RPMI medium and incubated at room temperature for 15 min. Either 7 μl of water, or 7 μl of 10 μM negative control siRNA (Silencer™ Negative Control #1 siRNA), or 7 μl of 10 μM STAT1 siRNA solution was diluted in 50 μl RPMI (Silencer™ Validated siRNA: STAT1, ID#42860, Ambion, Austin, TX. The STAT1 siRNA is a mixture containing two siRNAs targeting two coding regions on stat1 gene), mixed with the diluted transfection reagent, and incubated at room temperature for another 15 min. 1 ml of prewarmed P69 culture medium was then mixed with the siRNA suspension solution and dispensed onto cell monolayer on 12-well plate. After 24 h, 1 ml of fresh P69 culture medium was added into each well. Cells were incubated for another 20 h prior to synchronization in RPMI for 6 h and treated with defined RPMI culture medium supplemented with 50 ng/ml BSA plus IFN-γ as indicated.

Results IFN-γ inhibited cell proliferation and induced apoptosis in M12 cells M12 cells were treated with either type I interferon (IFN-α or IFN-β) or type II interferon, IFN-γ, at a concentration of 20 U/ml, 100 U/ml, 500 U/ml, or 2500 U/ml. Proliferation of M12 cells was dose-dependently inhibited by IFN-γ, with 100 U/ml inducing greater than 70% growth inhibition, relative to untreated cells, 96 h post-treatment (Fig. 1A). Higher doses of IFN-γ suppressed growth by more than 85%. In contrast, IFN-α or IFN-β, at a high concentration of 2500 U/ml, only modestly suppressed cell growth by no more than 30% (Fig. 1A). Correlated with the inhibition of cell growth, IFN-γ at 100 U/ml also induced significant apoptosis (3.8 ± 0.7-folds over control) 96 h post-treatment, whereas no increase in apoptosis was observed

Fig. 1 – The anti-proliferative, proapoptotic effect of IFN-γ on M12 cells. (A) M12 cells in 96-well plates were treated with either 5 ng/ml bovine serum albumin (BSA), IFN-α, IFN-β, or IFN-γ at the concentrations indicated. Cell proliferation was measured 96 h post-treatment as described in Materials and methods. The results are expressed as the percentages of the proliferation measured in the control (treated with BSA), which was given an arbitrary value of 100%. The data presented here are the mean ± SE from seven independent experiments performed in triplicates. (B) M12 cells in 96-well plates were treated with either BSA (5 ng/ml), IFN-α (1000 U/ml), IFN-β (1000 U/ml), or IFN-γ (100 U/ml). Apoptosis of M12 cells was measured 96 h post-treatment as described in Materials and methods. The results are expressed relative to control (treated with BSA), which was assigned a value of 1. The data presented here are the mean ± SE from three independent experiments performed in triplicates.

with IFN-α or IFN-β at 1000 U/ml, relative to untreated cells (Fig. 1B).

IFN-γ suppressed constitutively tyrosine-phosphorylated STAT3 The STAT3 pathway is constitutively activated in M12 cells (Fig. 2A), consistent with the observations made in many primary human tumors and in derived cancer cells [12], including prostate cancer cells [17]. Surprisingly, treatment with IFN-γ, but not with IFN-α or IFN-β, suppressed tyrosine-

1232

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

Fig. 2 – IFN-γ suppressed constitutively activated STAT3 in M12 cells. (A) M12 cells were starved in RPMI 1640 for 16 h before treatment with either BSA (5 ng/ml), IFN-α (1000 U/ml), IFN-β (1000 U/ml), or IFN-γ (100 U/ml). Cell lysates were collected at each time point, and were subject to Western immunoblot analysis with specific antibodies against proteins indicated at the left to each blot. (B) DU145 cells were grown to 80% confluency in 6-well plates, starved in RPMI 1640 for 16 h, and treated with either BSA (50 ng/ml) or IFN-γ (5000 U/ml). The cell lysates were collected at each time point and were subjected to Western immunoblot analysis with specific antibodies against proteins indicated at the left to each blot.

phosphorylation of STAT3 (pY-STAT3) in M12 cells (Fig. 2A). The suppression appeared to be a late event, since significant suppression was observed 24 h post-treatment (Fig. 2A and data not shown). Treatment with IFN-γ, however, did not suppress serine phosphorylation of STAT3 in M12 cells, or alter total STAT3 levels (Fig. 2A). The observation of IFN-γ-induced suppression of pYSTAT3 was not unique to M12 cells, as a similar observation was made in another prostate cancer line, DU145 (Fig. 2B). Compared with M12 cells, however, a higher concentration of IFN-γ (5000 U/ml) was required to see 40% dephosphorylation of pY-STAT3 48 h post-treatment (Fig. 2B).

Inhibition of STAT1 activation did not impair dephosphorylation of STAT3 by IFN-γ The JAK/STAT1 pathway is the major signaling pathway activated by IFN in many cell types [3,4]. In M12 cells, IFN-γ (100 U/ml) elicited a robust tyrosine phosphorylation of STAT1, as does IFN-α and IFN-β (Fig. 3A). Compared with IFN-α or IFN-β, however, IFN-γ also significantly elevated the level of serine phosphorylation of STAT1 (Fig. 3A). Similarly, in DU145 cells, IFN-γ elicited a robust tyrosinephosphorylation of STAT1; but unlike in M12 cells, activation of STAT1 was sustained for at least 48 h (Fig. 3B). Although IFN-γ-induced pY-STAT1 in DU145 cells was sustained compared to that in M12 cells, suppression of constitutively activated pY-STAT3 in DU145 cells was modest (Fig. 2, compare panels A and B), suggesting that STAT1 activation may not play a direct role in IFN-γinduced dephosphorylation of STAT3. To further evaluate whether IFN-γ-induced STAT1 activation might be responsible for the observed dephos-

phorylation of pY-STAT3 in M12 cells, STAT1 expression was blocked with siRNA. Total STAT1 expression in M12 cells was significantly reduced (90%) by STAT1 siRNA (Figs. 3C and D). A nonspecific siRNA (negative control), in contrast, did not alter the level of STAT1 protein (Fig. 3C). Corresponding to the reduced total STAT1 expression, IFNγ-induced pY-STAT1 was reduced by 80% (Figs. 3C and D). This specific and significant inhibition of STAT1 activation, however, did not interfere with the dephosphorylation of pY-STAT3 induced by IFN-γ (Fig. 3D). These results suggest that a fully activated STAT1 pathway is not required for IFN-γ to induce dephosphorylation of pY-STAT3 in M12 cells and thus imply the involvement of other signaling pathways in the dephosphorylation of STAT3.

Blockade of the PI-3K/Akt pathway or the MAPK/ERK pathway inhibited dephosphorylation of pY-STAT3 induced by IFN-γ In addition to constitutively activated STAT3, the basal levels of Akt and MAPK/ERK phosphorylation are also elevated in M12 cells (Fig. 4A). Treatment with IFN-α, IFN-β, or IFN-γ did not further activate Akt or ERK1/2 (Fig. 4A). However, inhibition of either the PI-3K pathway with LY294002 or the MAPK/ERK pathway with PD98059, affected IFN-γ-induced dephosphorylation of pY-STAT3 (Fig. 4). LY294002 at 5 μM or 20 μM completely blocked phosphorylation of Akt, with no effect on ERK1/2 phosphorylation (Fig. 4B). Conversely, PD98059 at 75 μM completely inhibited the MAPK/ERK pathway, but did not affect Akt phosphorylation (Fig. 4B). Inhibition of either pathway significantly blocked IFN-γinduced suppression of pY-STAT3, while having no effect on serine phosphorylation status of STAT3 24 h post-treatment (Fig. 4C).

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

1233

Fig. 3 – The effect of attenuating STAT1 synthesis with siRNA on dephosphorylation of pY-STAT3 induced by IFN-γ. (A) M12 cells were treated with interferons and the cell lysates were collected and analyzed as described in Fig. 2A. (B) DU145 cells were treated with IFN-γ (5000 U/ml) and the cell lysates were collected and analyzed as described in Fig. 2B. (C–D) M12 cells were seeded at 60% confluency on 12-well plates. After 16 h of incubation, siRNA transfection was performed with TransIT-TKO® Transfection Reagent (Mirus, Madison, WI) and/or negative control siRNA, or STAT1 siRNA (Ambion, Austin, TX) as described in Materials and methods. After 24 h, 1 ml of fresh P69 culture medium was added into each well. After another 20 h, the cells were synchronized in RPMI for 6 h before treatment with defined RPMI culture medium supplemented with BSA (50 ng/ml) plus IFN-γ (100 U/ml) as indicated, and the cell lysates were collected at 1 h (C) or 24 h (D) post-treatment with IFN-γ, and were subjected to Western immunoblot analysis with specific antibodies against proteins indicated at the left to each blot.

The involvement of the PI-3K/Akt pathway in IFN-γinduced de-phosphorylation of STAT3 was also observed in DU145 cells. Blocking the PI-3K/Akt pathway with 5 μM LY294002 suppressed IFN-γ-induced dephosphorylation of

pY-STAT3 (Fig. 4D), consistent with that observed in M12 cells. Inhibition of the MAPK/ERK pathway with PD98059, however, did not suppress IFN-γ-induced dephosphorylation of pY-STAT3 (data not shown).

1234

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

Dephosphorylation of pY-STAT3 required an intact mTOR pathway The mTOR pathway lies downstream of the PI-3K/Akt pathway [7,8], and LY294002 is a known dual inhibitor for both PI-3K and mTOR. The MAPK/ERK pathway can also regulate mTOR through modulating the activity of the Tuberous Sclerosis tumor suppressor (TSC1/2) complex [9]. The activated mTOR may regulate several downstream targets, including p70 S6K, a kinase regulating the synthesis of components in the translation machinery [20]. In M12 cells, treatment with IFN-γ for 5 to 60 min appears not to further activate p70 S6K (Fig. 5A). LY294002, at 5 or 20 μM, as expected, completely inhibited phosphorylation of p70 S6K at threonine389, and significantly impaired the phosphorylation of p70 S6K at threonine421 and serine 424 (Fig. 5B, lanes 2 and 3), even in the presence of IFN-γ (Fig. 5B, lanes 6 and 7). Similarly, PD98059 (75 μM) with or without IFN-γ, also significantly inhibited the phosphorylation of p70 S6K at threonine 389 and 421, and serine 424 (Fig. 5B, lanes 4 and 8). To investigate the role of the mTOR pathway in IFN-γinduced dephosphorylation of pY-STAT3, M12 cells were treated with rapamycin, the specific inhibitor to mTOR, in the absence or the presence of IFN-γ. Treatment with rapamycin at 50 nM completely inhibited phosphorylation of p70 S6K at threonine389, and significantly impaired the phosphorylation of p70 S6K at threonine421 and serine 424 (Fig. 5C), but did not interfere with the phosphorylation of Akt or ERK1/2 (Fig. 5C). Importantly, direct inhibition of mTOR with rapamycin significantly inhibited IFN-γ-induced dephosphorylation of pY-STAT3, but did not alter the level of serine phosphorylation of STAT3 or the level of total STAT3 protein in M12 cells (Fig. 5D) or in DU145 cells (Fig. 5E). Taken together, the observations shown in Figs. 4 and 5 suggest that the mTOR pathway is required to mediate, at least in part, the dephosphorylation of pY-STAT3 induced by IFN-γ.

Inhibition of the mTOR pathway did not interfere with the activation of STAT1 pathway by IFN-γ

Fig. 4 – Blockade of the PI3K pathway or the MAPK/ERK pathway inhibited dephosphorylation of pY-STAT3 induced by IFN-γ. (A) M12 cells were treated with interferons and the cell lysates were collected and analyzed as described in Fig. 2. (B–C) M12 cells were starved in RPMI 1640 for 16 h and pretreated for 1 h with either 0.1% (v/v) DMSO, 5 μM or 20 μM LY294002, or 75 μM PD98059 as indicated, prior to treatment with either BSA (5 ng/ml) or IFN-γ (100 U/ml). The cell lysates were collected at each time point and were subjected to Western immunoblot analysis with specific antibodies against proteins indicated at the left to each blot. (D) DU145 cells were grown to 80% confluency in 6-well plates, starved in RPMI 1640 for 16 h, and pretreated for 1 h with either 0.1% (v/v) DMSO or 5 μM LY294002 prior to treatment with either BSA (50 ng/ml) or IFN-γ (5000 U/ml). The cell lysates were collected at each time point and were subjected to Western immunoblot analysis with specific antibodies against proteins indicated at the left to each blot.

Treatment of M12 cells with rapamycin did not affect the IFN-γ-induced activation of the STAT1 pathway, as both tyrosine and serine phosphorylation of STAT1 were not altered upon inhibition of the mTOR pathway (Fig. 6A), nor was the synthesis of the STAT1 target gene IRF-1 [21] (Fig. 6B), or the induction of total STAT1 expression changed (Fig. 6C). Conversely, blocking the STAT1 pathway did not impair the activation of the mTOR/p70 S6K pathway (Fig. 6D). Similar observations were also made with DU145 cells (data not shown). These data suggest that the mTOR pathway contributes to IFN-γ-induced dephosphorylation of pY-STAT3, independent of STAT1 activation.

Inhibition of the mTOR pathway abolished the apoptosis induced by IFN-γ Accumulated evidence has suggested that constitutively activated STAT3 could protect cancer cells against apoptosis [12,22]. The effect of inhibiting the mTOR pathway on IFN-γinduced dephosphorylation of STAT3, therefore, suggested

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

that a fully activated mTOR pathway may also be required for IFN-γ to induce apoptosis in M12 cells (Fig. 1B). The effect of LY294002, PD98059, or rapamycin on cell death in the absence or the presence of IFN-γ was examined. LY294002, at a concentration of 20 μM, induced a moderate level of apoptosis in M12 cells at 72 h post-treatment (1.9-fold relative to control) (Fig. 7A). The sensitivity of M12 cells to 20 μM of LY294002 is consistent with those reported in the literature. The lower concentration (5 μM) of LY294002, as well as PD98059 (75 μM) or rapamycin (50 nM), in contrast, minimally affected cell proliferation (data not shown) and did not induce apoptosis in M12 cells (Fig. 7A). LY294002, PD98059, or rapamycin, however, totally abolished IFN-γ-induced apoptosis at 72 h post-treatment (Fig. 7A). These observations thus support the hypothesis that IFN-γ-induced apoptosis in M12 cells is dependent on a fully activated mTOR signaling pathway, and correlates with the persistent suppression of constitutively activated STAT3. The requirement of the mTOR pathway for IFN-γ-induced apoptosis was not specific to M12 cells, as the similar results were also observed in another cell line, SK-HEP-1 heptocarcinoma cells, which do not contain a high basal level of pYSTAT3 (data not shown). Treatment of SK-HEP-1 cells with IFN-γ (4000 U/ml), induced significant apoptosis (2.7 ± 0.1-folds over control) 48 h post-treatment (Fig. 7B). Treatment with rapamycin (50 nM) did not induce apoptosis relative to untreated cells, but totally suppressed the apoptosis induced by IFN-γ in SK-HEP-1 cells (Fig. 7B), consistent with that observed in M12 cells (Fig. 7A). These results further suggest that an intact mTOR pathway is generally required for IFN-γ to induce apoptosis in certain tumor cells.

Discussion In this study, we have demonstrated that the potent inhibitory effect on cell growth and the induction of apoptosis exhibited by IFN-γ are associated with the persistent suppression of constitutive pY-STAT3 in M12 cells, a derived human metastatic prostate cancer cell line. IFN-γ primarily activates STAT1 [4], although STAT3 and 5 can also be activated [19,23]. In M12 cells, IFN-γ could induce phosphorylation of STAT1 (Fig. 3A) and STAT5 (data not shown). The immediate effect of IFN-γ on STAT3 phosphorylation, however, was not obvious, as STAT3 is constitutively tyrosine- and serinephosphorylated in M12 cells (Fig. 2A and data not shown). Surprisingly, persistent dephosphorylation of pY-STAT3, but not pS-STAT3, was observed 24 h–48 h post-IFN-γ treatment (Fig. 2A), and preceded subsequent growth inhibition and apoptosis (Fig. 1 and data not shown). The importance of this correlation was supported by the demonstration that: (i) IFN-α or IFN-β, which did not affect cell proliferation nor induce apoptosis, did not elicit dephosphorylation of pY-STAT3 and (ii) restoration of pY-STAT3 by blocking the mTOR pathway abrogated IFN-γ-induced apoptosis. Altogether, our results are consistent with the prosurvival, anti-apoptotic role of phospho-STAT3 well established in human cancer. In normal cells, STAT3 is predominantly activated by the cytokines of IL-6 family. In prostate cancer cells, IL-6 has been implicated as an autocrine/paracrine factor for prostate cancer cells [24], although the role of IL-6 on the growth of prostate

1235

cancer cells has been controversial [25,26] and may be dependent on the type of prostate cancer [17]. M12 cells also secret IL-6 into its culture medium (data not shown). Treatment with IFN-γ, however, neither reduced the amount of secreted IL-6 (data not shown), nor suppressed constitutively activated PI-3K/Akt pathway and MAPK/ERK pathway (Fig. 5B and data not shown), implicating that treatment with IFN-γ did not generally suppress the signaling pathways likely activated by IL-6, but specifically de-phosphorylated pY-STAT3. One common mechanism utilized by cytokines to modulate Jak/STAT signaling is through induction of negative regulators, such as SOCS or tyrosine phosphatases [27,28]. In M12 cells, our preliminary data suggest that the dephosphorylation of pY-STAT3 by IFN-γ appears not to involve SOCS, such as SOCS-1 and SOCS-3, as SOCS-3 mRNA expression was only transiently up-regulated by IFN-γ, and SOCS1 mRNA was not detected (data not shown). In addition, IFN-γ-induced dephosphorylation of pY-STAT3 appears not to involve the tyrosine phosphatases sensitive to orthovanadate as orthovanadate treatment did not prevent IFN-γ-induced dephosphorylation of pY-STAT3, although activation of pY-STAT1 was prolonged (data not shown). Furthermore, the phosphorylation status of Jak2 in M12 cells appeared not to be significantly altered by IFN-γ-treatment (data not shown), and together with the observation of continued activation of the PI-3K/Akt and ERK1/2 pathways (data not shown and see discussion above), suggesting that inactivation of Jak2 was unlikely to play a direct role in the dephosphorylation of pYSTAT3. The possibility that STAT3 dephosphorylation is a result of STAT3-specific degradation was also ruled out by the observation that the level of total STAT3 remained constant with or without IFN-γ treatment. Further investigation, however, is required to define the exact role for these Jak/ STAT negative regulators in IFN-γ-elicited pY-STAT3 suppression in M12 cells. Although the STAT1 pathway has been regarded as the major pathway activated in many cell types upon IFN-γ treatment [3,4], it has also been shown that IFN-γ could regulate gene expression independent of STAT1 activation [29]. Our results indicate that an 80% suppression of STAT1 activation did not prohibit IFN-γ-induced dephosphorylation of pY-STAT3 (Fig. 3), suggesting that gene expression required for STAT3 dephosphorylation may not be directly regulated by STAT1. The possibility that the residual IFN-γactivated STAT1 after siRNA suppression may still be involved in IFN-γ-induced dephosphorylation of pY-STAT3 was not supported by our data: (1) robust activation of STAT1 by IFN-α or IFN-β did not result in dephosphorylation of pY-STAT3 (Fig. 3A); and (2) suppression of the mTOR pathway, which restored pY-STAT3, did not interfere with IFN-γ-activated STAT1 pathway (Fig. 6). These results thus indicate that IFN-γ-induced-dephosphorylation of pY-STAT3 does not involve the STAT1 pathway. The mTOR pathway is one of several signaling pathways that has been shown to be activated by IFN-γ [10]. In M12 cells, activation of mTOR by IFN-γ is not obvious, as indicated by the phosphorylation status of p70 S6K, the direct target of mTOR (Fig. 5A). This could be due to the fact that mTOR is already highly activated in M12 cells, most probably because of the elevated levels of active PI-3K or MAPK/ERK (Fig. 4A), both of

1236

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

which have been shown to positively regulate mTOR activity [7–9]. The role of the mTOR pathway in IFN-γ-induced biological effects is poorly defined. Several reports implicate the involvement of mTOR in regulation of serine phosphorylation of STAT1 [30] or STAT3 [31], and inhibition of mTOR with rapamycin

enhanced IFN-γ-induced apoptosis [32], consistent with the established, prosurvival, anti-apoptotic roles of mTOR [7]. Our data, on the contrary, indicate that suppression of the mTOR pathway in M12 cells or DU145 cells did not alter the serine phosphorylation of STAT1 or STAT3 (Figs. 4 and 6), but inhibited the dephosphorylation of pY-STAT3 induced by IFN-γ (Fig. 5).

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

Fig. 6 – Blockade of mTOR pathway did not affect STAT1 activation. (A–C) M12 cells were treated with DMSO, 50 nM rapamycin in the absence or the presence of IFN-γ (100 U/ml), and the cell lysates were collected and analyzed as described in Fig. 5C. (D). M12 cells were transfected with siRNA and treated with or without IFN-γ (100 U/ml) and the cell lysates were collected and analyzed as described in Fig. 3C. This inhibitory effect, furthermore, was not through impairment of IFN-γ-induced STAT1 activation (Fig. 6). In addition, suppression of STAT1 synthesis by siRNA did not inhibit dephosphorylation of pY-STAT3 (Fig. 3D). These data suggest that it is the mTOR pathway (and possibly other signaling pathways), and not the STAT1 pathway, that plays a pivotal role in dephosphorylation of pY-STAT3 induced by IFN-γ. Previous studies have established the essential role of STAT1 activation in IFN-γ-induced anti-proliferative, proapoptotic effects [5]. Our preliminary data also indicated that

1237

Fig. 7 – The mTOR pathway is involved in IFN-γ-induced apoptosis. (A) M12 cells on 96-well plates were pretreated for 1h with 0.1% (v/v) DMSO, 5 μM or 20 μM LY294002, 75 μM PD98059, or 50 nM rapamycin and then treated with either BSA (5 ng/ml) or IFN-γ (100 U/ml) as indicated. Apoptosis in M12 cells at 72 h post-treatment was measured and the results are expressed as in Fig. 1B. The data presented here are the mean ± SE from three independent experiments performed in triplicates. (B) SK-HEP-1 cells on 96-well plates were pretreated for 1 h with 0.1% (v/v) DMSO or 50 nM rapamycin and then treated with either BSA (5 ng/ml) or IFN-γ (4000 U/ml) as indicated. Apoptosis in SK-HEP-1 cells 48 h post-treatment was measured and the results are expressed as in Fig. 1B. The data presented here are the mean ± SE from three independent experiments performed in triplicates.

suppressing STAT1 expression with STAT1 siRNA inhibited IFN-γ-induced apoptosis in M12 cells (data not shown), suggesting that the apoptosis induced by IFN-γ still required STAT1 activation in M12 cells. However, activation of STAT1 pathway alone may not be sufficient for IFN-γ to induce

Fig. 5 – IFN-γ-induced dephosphorylation of STAT3 requires the mTOR pathway. (A) M12 cells were treated with IFN-γ (100 U/ml) and the cell lysates were collected and analyzed as described in Fig. 2A. (B) M12 cells were treated with DMSO, LY294002, or PD98059 in the absence or the presence of IFN-γ (100 U/ml) and the cell lysates were collected and analyzed as described in Fig. 4B. (C–D) M12 cells were starved in RPMI 1640 for 16 h and pretreated for 1 h with either 0.1% (v/v) DMSO or 50 nM rapamycin prior to treatment with either BSA (5 ng/ml) or IFN-γ (100 U/ml). The cell lysates were collected at each time point and were subjected to Western immunoblot analysis with specific antibodies against proteins indicated at the left to each blot. (E) DU145 cells were treated with DMSO, 100 nM rapamycin in the absence or the presence of IFN-γ (5000 U/ml), and the cell lysates were collected and analyzed as described in panel C.

1238

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

apoptosis, as our data indicated that, even while STAT1 can be fully activated, IFN-γ-induced apoptosis was abolished when the mTOR pathway was blocked (Fig. 7). It is apparent, therefore, that an intact mTOR pathway is also required for IFN-γ to induce apoptosis. More importantly, the mTOR pathway not only mediated apoptosis induced by IFN-γ, but has been shown to also mediate the apoptosis induced by type I interferon, IFN-α [11], thus suggesting that the mTOR pathway, which normally supports cell survival, may participate in cell death induced by interferons. Several mechanisms may account for how the mTOR pathway mediates IFN-γ-elicited suppression of activated STAT3 and apoptosis. The mTOR pathway is well characterized for regulating protein synthesis [7]. Our data suggested that an impaired mTOR pathway only affected the synthesis of a subset of proteins, with late induction of IRF-I and STAT1 protein synthesis unaffected by rapamycin treatment (Fig. 6). Furthermore, blocking the mTOR pathway with rapamycin (50 nM) in the absence of IFN-γ did not induce apoptosis (Fig. 7), while blocking protein synthesis with cycloheximide, the potent general protein synthesis inhibitor, induced apoptosis and dephosphorylation of STAT3 even in the absence of IFN-γ (data not shown). Hence, inhibition of the mTOR pathway may significantly affect synthesis of those proteins transcriptionally regulated by IFN-γ that are required for dephosphorylation of STAT3 and apoptosis. It is also possible that blocking the mTOR pathway affects both transcription and translation of proteins involved in IFN-γ-induced STAT3 dephosphorylation and apoptosis. The involvement of mTOR in transcriptional regulation has been demonstrated in one study where the mTOR pathway is involved in the transcriptional regulation of human inducible nitric oxide synthase by IFN-γ and lipopolysaccharide (LPS) [30]. In summary, our studies demonstrate, for the first time, that a fully active mTOR pathway is required for IFN-γ to suppress constitutively activated pY-STAT3 and to induce apoptosis, suggesting that the signaling pathways traditionally regarded as survival pathways may positively contribute to the anti-proliferative, proapoptotic effects elicited by IFN-γ. The further understanding of the underlying mechanism may greatly assist in the development of new strategies for clinical applications of IFN-γ.

[3]

[4]

[5]

[6] [7] [8]

[9]

[10]

[11]

[12] [13] [14]

[15]

[16]

Acknowledgments [17]

This work was supported by National Institutes of Health Grant CA58110 (to RGR), and by Department of Defense Grant W81XWH-04-1-0107 (to PF). [18] REFERENCES

[1] H. Ikeda, L.J. Old, R.D. Schreiber, The roles of IFN gamma in protection against tumor development and cancer immunoediting, Cytokine Growth Factor Rev. 13 (2002) 95–109. [2] R.J. van Moorselaar, P. van Stratum, G. Borm, F.M. Debruyne, J.A. Schalken, Differential antiproliferative activities of alphaand gamma-interferon and tumor necrosis factor alone or in

[19]

[20]

combinations against two prostate cancer xenografts transplanted in nude mice, Prostate 8 (1991) 331–344. J.E. Darnell Jr., I.M. Kerr, G.R. Stark, Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins, Science 264 (1994) 1415–1421. G.R. Stark, I.M. Kerr, B.R. Williams, R.H. Silverman, R.D. Schreiber, How cells respond to interferons, Annu. Rev. Biochem. 67 (1998) 227–264. J.F. Bromberg, C.M. Horvath, Z. Wen, R.D. Schreiber, J.E. Darnell Jr., Transcriptionally active Stat1 is required for the antiproliferative effects of both interferon alpha and interferon gamma, Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 7673–7678. L.C. Platanias, Mechanisms of type-I- and type-II-interferonmediated signalling, Nat. Rev., Immunol. 5 (2005) 375–386. N. Hay, N. Sonenberg, Upstream and downstream of mTOR, Genes Dev. 18 (2004) 1926–1945. B.D. Manning, L.C. Cantley, United at last: the tuberous sclerosis complex gene products connect the phosphoinositide 3-kinase/Akt pathway to mammalian target of rapamycin (mTOR) signalling, Biochem. Soc. Trans. 31 (2003) 573–578. L. Ma, Z. Chen, H. Erdjument-Bromage, P. Tempst, P.P. Pandolfi, Phosphorylation and functional inactivation of TSC2 by ERK implications for tuberous sclerosis and cancer pathogenesis, Cell 121 (2005) 179–193. F. Lekmine, A. Sassano, S. Uddin, J. Smith, B. Majchrzak, S.M. Brachmann, N. Hay, E.N. Fish, L.C. Platanias, Interferongamma engages the p70 S6 kinase to regulate phosphorylation of the 40S S6 ribosomal protein, Exp. Cell Res. 295 (2004) 173–182. L. Thyrell, L. Hjortsberg, V. Arulampalam, T. Panaretakis, S. Uhles, M. Dagnell, B. Zhivotovsky, I. Leibiger, D. Grander, K. Pokrovskaja, Interferon alpha-induced apoptosis in tumor cells is mediated through the phosphoinositide 3-kinase/ mammalian target of rapamycin signaling pathway, J. Biol. Chem. 279 (2004) 24152–24162. T. Bowman, R. Garcia, J. Turkson, R. Jove, STATs in oncogenesis, Oncogene 9 (2000) 2474–2488. J. Bromberg, Stat proteins and oncogenesis, J. Clin. Invest. 109 (2002) 1139–1142. V. Calo, M. Migliavacca, V. Bazan, M. Macaluso, M. Buscemi, N. Gebbia, A. Russo, STAT proteins: from normal control of cellular events to tumorigenesis, J. Cell. Physiol. 197 (2003) 157–168. J.F. Bromberg, M.H. Wrzeszczynska, G. Devgan, Y. Zhao, R.G. Pestell, C. Albanese, J.E. Darnell Jr., Stat3 as an oncogene, Cell 98 (1999) 295–303. Z. Ni, W. Lou, E.S. Leman, A.C. Gao, Inhibition of constitutively activated Stat3 signaling pathway suppresses growth of prostate cancer cells, Cancer Res. 60 (2000) 1225–1228. L.B. Mora, R. Buettner, J. Seigne, J. Diaz, N. Ahmad, R. Garcia, T. Bowman, R. Falcone, R. Fairclough, A. Cantor, C. Muro-Cacho, S. Livingston, J. Karras, J. Pow-Sang, R. Jove, Constitutive activation of Stat3 in human prostate tumors and cell lines: direct inhibition of Stat3 signaling induces apoptosis of prostate cancer cells, Cancer Res. 62 (2002) 6659–6666. V.L. Bae, C.K. Jackson-Cook, S.J. Maygarden, S.R. Plymate, J. Chen, J.L. Ware, Metastatic sublines of an SV40 large T antigen immortalized human prostate epithelial cell line, Prostate 34 (1998) 275–282. V. Hwa, B. Little, E.M. Kofoed, R.G. Rosenfeld, Transcriptional regulation of insulin-like growth factor-I by interferongamma requires STAT-5b, J. Biol. Chem. 279 (2004) 2728–2736. J. Chung, C.J. Kuo, G.R. Crabtree, J. Blenis, Rapamycin-FKBP specifically blocks growth-dependent activation of and signaling by the 70 kd S6 protein kinases, Cell 69 (1992) 1227–1236.

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 1 2 2 9 –12 3 9

[21] X. Li, S. Leung, S. Qureshi, J.E. Darnell Jr., G.R. Stark, Formation of STAT1-STAT2 heterodimers and their role in the activation of IRF-1 gene transcription by interferon-alpha, J. Biol. Chem. 271 (1996) 5790–5794. [22] R. Catlett-Falcone, T.H. Landowski, M.M. Oshiro, J. Turkson, A. Levitzki, R. Savino, G. Ciliberto, L. Moscinski, J.L. FernandezLuna, G. Nunez, W.S. Dalton, R. Jove, Constitutive activation of Stat3 signaling confers resistance to apoptosis in human U266 myeloma cells, Immunity 10 (1999) 105–115. [23] E.M. Kofoed, V. Hwa, B. Little, K.A. Woods, C.K. Buckway, J. Tsubaki, K.L. Pratt, L. Bezrodnik, H. Jasper, A. Tepper, J.J. Heinrich, R.G. Rosenfeld, Growth hormone insensitivity associated with a STAT5b mutation, N. Engl. J. Med. 349 (2003) 1139–1147. [24] M. Okamoto, C. Lee, R. Oyasu, Interleukin-6 as a paracrine and autocrine growth factor in human prostatic carcinoma cells in vitro, Cancer Res. 57 (1997) 141–146. [25] W. Lou, Z. Ni, K. Dyer, D.J. Tweardy, A.C. Gao, Interleukin-6 induces prostate cancer cell growth accompanied by activation of stat3 signaling pathway, Prostate 42 (2000) 239–242. [26] M.T. Spiotto, T.D. Chung, STAT3 mediates IL-6-induced growth inhibition in the human prostate cancer cell line LNCaP, Prostate 42 (2000) 88–98.

1239

[27] H. Yasukawa, A. Sasaki, A. Yoshimura, Negative regulation of cytokine signaling pathways, Annu. Rev. Immunol. 18 (2000) 143–164. [28] S. Wormald, D.J. Hilton, Inhibitors of cytokine signal transduction, J. Biol. Chem. 279 (2004) 821–824. [29] C.V. Ramana, M.P. Gil, R.D. Schreiber, G.R. Stark, Stat1-dependent and -independent pathways in IFN-gamma-dependent signaling, Trends Immunol. 23 (2002) 96–101. [30] A.S. Kristof, J. Marks-Konczalik, E. Billings, J. Moss, Stimulation of signal transducer and activator of transcription-1 (STAT1)-dependent gene transcription by lipopolysaccharide and interferon-gamma is regulated by mammalian target of rapamycin, J. Biol. Chem. 278 (2003) 33637–33644. [31] K. Yokogami, S. Wakisaka, J. Avruch, S.A. Reeves, Serine phosphorylation and maximal activation of STAT3 during CNTF signaling is mediated by the rapamycin target mTOR, Curr. Biol. 10 (2000) 47–50. [32] N. El-Hashemite, H. Zhang, V. Walker, K.M. Hoffmeister, D.J. Kwiatkowski, Perturbed IFN-gamma-Jak-signal transducers and activators of transcription signaling in tuberous sclerosis mouse models: synergistic effects of rapamycin-IFN-gamma treatment, Cancer Res. 64 (2004) 3436–3443.