Oxidative stress impairs multiple regulatory events to drive persistent cytokine-stimulated STAT3 phosphorylation

Biochimica et Biophysica Acta 1843 (2014) 483–494

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbamcr

Oxidative stress impairs multiple regulatory events to drive persistent cytokine-stimulated STAT3 phosphorylation Ivan H.W. Ng a,b, Yvonne Y.C. Yeap a, Lynette S.R. Ong a,1, David A. Jans b,⁎, Marie A. Bogoyevitch a,⁎⁎ a b

Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and Biotechnology Institute, University of Melbourne, Victoria 3010, Australia Department of Biochemistry and Molecular Biology, Monash University, Victoria 3800, Australia

a r t i c l e

i n f o

Article history: Received 20 August 2013 Received in revised form 31 October 2013 Accepted 19 November 2013 Available online 25 November 2013 Keywords: STAT3 Oxidative stress Cytokines Nuclear transport

a b s t r a c t Although cytokine-driven STAT3 phosphorylation and activation are often transient, persistent activation of STAT3 is a hallmark of a range of pathologies and underpins altered transcriptional responses. As triggers in disease frequently include combined increases in inflammatory cytokine and reactive oxygen species levels, we report here how oxidative stress impacts on cytokine-driven STAT3 signal transduction events. In the model system of murine embryonic fibroblasts (MEFs), combined treatment with the interleukin-6 family cytokine Leukemia Inhibitory Factor (LIF) and hydrogen peroxide (H2O2) drove persistent STAT3 phosphorylation whereas STAT3 phosphorylation increased only transiently in response to LIF alone and was not increased by H2O2 alone. Surprisingly, increases in transcript levels of the direct STAT3 gene target SOCS3 were delayed during the combined LIF + H2O2 treatment, leading us to probe the impact of oxidative stress on STAT3 regulatory events. Indeed, LIF + H2O2 prolonged JAK activation, delayed STAT3 nuclear localisation, and caused relocalisation of nuclear STAT3 phosphatase TC-PTP (TC45) to the cytoplasm. In exploring the nuclear import/ export pathways, we observed disruption of nuclear/cytoplasmic distributions of Ran and importin-α3 in cells exposed to H2O2 and the resultant reduced nuclear trafficking of classical importin-α/β-dependent protein cargoes. CRM1-mediated nuclear export persisted despite the oxidative stress insult, with sustained STAT3 Y705 phosphorylation enhancing STAT3 nuclear residency. Our studies thus reveal for the first time the striking impact of oxidative stress to sustain STAT3 phosphorylation and nuclear retention following disruption of multiple regulatory events, with significant implications for STAT3 function. © 2013 Elsevier B.V. All rights reserved.

1. Introduction The proteins of Janus Kinase (JAK)/Signal Transducer and Activator of Transcription (STAT) pathway, initially identified as important direct signalling components linking cytokine receptor activation Abbreviations: JAK, Janus kinase; STAT, Signal Transducer and Activator of Transcription; TC-PTP, T-cell protein tyrosine phosphatase; SOCS3, suppressor of cytokine signalling 3; H2O2, hydrogen peroxide; LIF, leukemia inhibitory factor; OSM, oncostatin M; MAPK, mitogen-activated protein kinase; MEF, murine embryonic fibroblast; FCS, foetal calf serum; CLSM, confocal laser scanning microscopy; GFP, green fluorescent protein; NES, nuclear export sequence; NLS, nuclear localisation sequence; Fn, nuclear fluorescence; Fc, cytoplasmic fluorescence; Fn/c, the nuclear to cytoplasmic ratio; SEM, standard error of the mean ⁎ Correspondence to: D.A. Jans, Nuclear Signalling Laboratory, Department of Biochemistry and Molecular Biology, Monash University, Clayton, VIC 3800, Australia. Tel.: +61 3 9901 9341; fax: +61 3 9902 9500. ⁎⁎ Correspondence to: M.A. Bogoyevitch, Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and Biotechnology Institute, University of Melbourne, 30 Flemington Road, Parkville, VIC 3010, Australia. Tel.: +61 3 8344 2289; fax: +61 3 9348 1421. E-mail addresses: [email protected] (D.A. Jans), [email protected] (M.A. Bogoyevitch). 1 Present Address: Centre for Animal Biotechnology, School of Veterinary Science, University of Melbourne, Parkville, VIC 3010, Australia. 0167-4889/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbamcr.2013.11.015

to downstream transcriptional changes, have been increasingly recognised as playing major roles in cancer and pathologies associated with enhanced inflammation [1,2]. The STAT transcription factor proteins are largely inactive and cytoplasmic under normal conditions, but upon JAK-stimulated tyrosine phosphorylation, translocate to the cell nucleus to participate directly in the expression of the STATspecific gene targets [3]. One member of the STAT family, STAT3, has attracted attention as a transcription factor mediating the gene expression changes upon stimulation by the interleukin-6 family of cytokines [4]. In this classical model of activation, STAT3 phosphorylation and nuclear localisation are usually transient, but persistent STAT3 phosphorylation and activation have frequently been linked with events of cellular injury, inflammation and cancer [3,5,6]. Although the mechanisms leading to persistent STAT3 activation remain largely unclear, several regulatory events control the magnitude and extent of STAT3 activation under control conditions. These regulatory events include the elevation of cytokine levels that promote receptor activation with downstream activation of tyrosine kinases [7], and conversely the involvement of negative regulators such as the tyrosine phosphatase T-cell protein tyrosine phosphatase (TC-PTP) [8–10] and the suppressor of cytokine signalling 3 (SOCS3) protein [11]. Notably, SOCS3 is itself a gene target of active STAT3 and its direct regulation

484

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

by STAT3 provides an immediate negative feedback mechanism attenuating signalling by the JAK/STAT3 pathway following cytokine activation [11]. Increased intracellular reactive oxygen species levels, leading to oxidative stress, have also been linked to disease pathologies, often promoting inflammatory processes and thus further elevating proinflammatory cytokine levels [12,13]. Whilst recent studies have demonstrated STAT3 activation by hydrogen peroxide (H2O2) as an oxidative stress trigger [14–16], the impact of oxidative stress on the magnitude and kinetics of cytokine-stimulated STAT3 phosphorylation and activation has not been evaluated. In this study, we focus on the consequences of oxidative stress on STAT3 activation by the interleukin-6 family cytokine Leukemia Inhibitory Factor (LIF). Our studies provide the first evidence that oxidative stress prolongs cytokine-driven STAT3 activation by deregulating multiple regulatory events. 2. Materials and methods 2.1. Cell culture, transfection and treatments Wild-type (WT) murine embryonic fibroblasts (MEFs), STAT3−/− MEFs, or HeLa cells were maintained in high glucose Dulbecco's modified Eagle's medium (DMEM) containing L-glutamine (Gibco) supplemented with 10% [v/v] foetal calf serum (FCS; DKSH Australia) and 100 U/ml penicillin/streptomycin (Gibco) then changed to serum-free media conditions 16 h before treatment with either 10 ng/ml murine recombinant LIF (Sigma), 1 mM H2O2 (Ajax Finechem), or a combination of 10 ng/ml LIF and 1 mM H2O2 (LIF + H2O2). For selected studies, 50 μM menadione was used to generate superoxide in place of oxidative stress initiated by H2O2. For protein kinase inhibitor studies, WT MEFs were pre-treated with a Mitogen-Activated Protein Kinase (MAPK) pathway inhibitor (20 μM JNK Inhibitor VIII [17]; 20 μM MEK inhibitor, UO126 [18]; 20 μM p38 inhibitor, SB203580 [19]), 0.5 μM JAK inhibitor I [20] or 10 μM Src family kinase inhibitor SU6656 [21] for 1 h prior to LIF, H2O2, or LIF + H2O2 treatment. Cells were pre-treated with the STAT3 pathway inhibitor STATTIC [22] (20 μM) 30 min prior to treatment with LIF, H2O2, or LIF + H2O2. All inhibitors were from Calbiochem. For the delivery of anti-importin-α3 or anti-importin-β1 antibodies (Abcam) into intact cells, the Chariot peptide system was used according to the manufacturer's instruction (Active Motif) [23,24]. Briefly, antibodies (1 μg in 100 μl PBS) were mixed with 3.5 μl Chariot peptide solution and 100 μl distilled deionised H2O, added to 150 μL of serumfree culture medium, and then preincubated with cells (12 well dishes) for 4 - 12 h prior to exposure to LIF. Transient transfection studies were carried out with the expression of myc epitope-tagged phosphatase TCPTP [25], or Green Fluorescent Protein (GFP)-tagged nuclear export sequence (NES) and nuclear localisation sequence (NLS) constructs: GFP–Rev–NES [26], GFP-T-antigen (GFP-T-ag(114–135)) [27], GFPpUL54(1145–1161) [28], GFP-VP3(74–121), GFP–VP3–NES mutant [29,30] or a control GFP-only construct using Lipofectamine™ LTX with Plus™ reagent and Opti-MEM according to the manufacturer's instructions (Invitrogen) prior to further treatments. In control experiments, GFP–Rev–NES cells were additionally pre-treated with 10 ng/ml Leptomycin B, a CRM1 inhibitor [31] for 60 min prior to analysis. 2.2. Cell lysate preparation and immunoblot analysis Following treatment, cell lysates were prepared using RIPA buffer (50 mM Tris–HCl, pH 7.3, 150 mM NaCl, 0.1 mM EDTA, 1% [v/v] sodium deoxycholate, 1% [v/v] Triton X-100, 0.2% [w/v] NaF and 100 μM Na3VO4) supplemented with complete protease inhibitor mix (Roche Diagnostic). Protein samples, diluted with 3 × protein sample buffer, were resolved by SDS-PAGE (10% [v/v] polyacrylamide gels) and the separated proteins were transferred onto PVDF membranes

(Amersham Life Science). Proteins of interest were blotted using the following primary antibodies: anti-STAT3 N-terminus (BD Bioscience), anti-phospho-Y705 STAT3 or phospho-S727 STAT3 (Cell Signaling Technology), anti-TC-PTP (R & D systems), anti-α-tubulin (Sigma) and anti-c-Myc epitope (Santa Cruz Biotechnology). After incubation with horseradish peroxidase-linked secondary antibody (Thermo Scientific), immunoreactive proteins were visualized using an enhanced chemiluminescence detection system (Thermo Scientific) and quantitation carried out using ImageJ 1.38 public domain software (National Institutes of Health, USA).

2.3. Immunofluorescence, confocal laser scanning microscopy and image analysis Samples were prepared and analysed as described previously [32]. Briefly, treated cells on coverslips were washed with cold phosphate-buffered saline (PBS) before fixation using 4% [w/v] paraformaldehyde (15 min, 37 °C) and permeabilisation in 0.2% [v/v] Triton X-100/PBS (15 min, room temperature). Non-specific binding was blocked by incubation in 1% [w/v] bovine serum albumin/PBS (30 min, room temperature), then coverslips incubated with primary antibodies [STAT3 (C-20) (Santa Cruz), Ran (BD Bioscience), importin-α3/KPNA4 or importin-β1/NTF97 (Abcam), 1:200 dilution in sterile filtered 1% [w/v] bovine serum albumin/PBS] then washed with PBS before incubation with Cy2/Cy3- (Millipore) or Alex488coupled (Invitrogen) secondary antibodies. Nuclei were visualized using DAPI (Sigma, 1:15000 in PBS). Coverslips were mounted (Biomeda Gel Mount, ProSciTech) onto glass slides and confocal laser scanning microscopy was performed using a Leica TCS SP2 imaging system with a 100 × 1.35 NA objective. Quantitation of relative fluorescence intensities in digitised confocal images was carried out by measuring fluorescence in an area of the nucleus and cytoplasm of cells in 10 different fields (ImageJ 1.38 public domain software) from 3 independent experiments (n = 3). The values for the nuclear (Fn ) and cytoplasmic (F c) fluorescence, subsequent to subtraction of background fluorescence, enabled the specific nuclear to cytoplasmic ratio (Fn/c) to be calculated [32].

2.4. RNA preparation and quantitative real-time PCR Samples were prepared and analysed as described previously [32]. Briefly, total RNA was extracted from MEFs using the Purelink RNA mini-kit (Invitrogen) and reverse-transcribed to cDNA using the RT High Capacity kit (Applied Biosystems) according to the manufacturer's protocols. Quantitative real-time TaqMan® PCR to determine the regulation of SOCS3 transcript levels was performed using 50 ng cDNA in a 20 μl reaction containing TaqMan® Gene Expression Master Mix and specific TaqMan® Gene Expression Assay (AssayID: SOCS3, Mm00545913_s1; Applied Biosystems). Amplification of cDNA was carried out in a 48-well Step One real-time PCR system (Applied Biosystems) using the following PCR conditions: 2 min at 50 °C and 10 min at 95 °C, followed by 40 cycles of 15 s at 95 °C and 1 min at 60 °C. The data was normalised to GAPDH (AssayID: Mm99999915_g1) in the respective samples and data quantitation was carried out using the 2−ΔΔCT method and expressed relative to the control sample. RNA isolation and expression analysis were performed on 3 independent occasions.

2.5. Statistical analysis Statistical analysis was carried out using Graphpad Prism 5 software (Windows version 5.00, GraphPad Software) and the unpaired student T-test was used to compare data from control cells versus treated cells at each corresponding time point.

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

not the result of irreversible damage to the cell regulatory system during oxidative stress (Fig. 1E, and quantitative data in Supp. Fig. 1C). To determine the impact of oxidative stress on STAT3 transcriptional activity, we used RT-PCR to evaluate the expression of SOCS3, a wellknown STAT3 direct target gene [11]. In the presence of LIF alone, a robust increase in SOCS3 was observed at 15 min (Fig. 2), corresponding with the peak in STAT3 pY705 and pS727 (Fig. 1). H2O2 treatment alone did not induce SOCS3 expression, but LIF + H2O2 treatment resulted in a delayed increase in SOCS3 expression. We confirmed STAT3 inhibition by STATTIC (20 μM) [22] (Supp. Fig. 2), and demonstrated the requirement for STAT3 in SOCS3 expression by showing that pre-treatment with STATTIC blocked SOCS3 induction in LIF + H2O2-treated WT MEFs, and that there was a lack of SOCS3 induction in response to LIF/H2O2/LIF + H2O2 in STAT3−/− MEFs (Fig. 2). Taken together, these results indicate a delayed STAT3-dependent regulation of SOCS3 expression following the combined cytokine and oxidative stress treatment despite persistent STAT3 phosphorylation.

3. Results 3.1. Treatment with LIF and H2O2 in combination results in sustained STAT3 Y705 and S727 phosphorylation and delayed expression of SOCS3 To test the impact of oxidative stress on STAT3 signalling, we exposed MEFs to 1 mM H2O2 or 10 ng/ml Leukemia Inhibitory Factor (LIF), as well as a combination of both stimuli (LIF + H2O2), for up to 2 h. STAT3 activation was analysed by immunoblotting for phosphorylation of STAT3 at Y705 (pY705 STAT3, required for dimerization and nuclear translocation [3]), and S727 (pS727 STAT3, linked to increased STAT3 transactivation [33]). Consistent with LIF-stimulated STAT3 activation, we observed rapid and transient increases in both pY705 and pS727 STAT3 following exposure to LIF (Fig. 1A). Exposure to H2O2 for up to 2 h stimulated a detectable increase in pS727 STAT3 without an accompanying increase in pY705 STAT3, but the combined treatment of LIF + H2O2 resulted in increased and sustained pY705 STAT3 together with delayed pS727 STAT3 (Fig. 1A). Densitometric analysis and statistical evaluation of the quantitative data from three independent experiments confirmed these observations of sustained pY705 STAT3 and delayed pS727 STAT3 following LIF + H2O2 co-exposure (Fig. 1B). To ensure this sustained STAT3 Y705 phosphorylation was not reliant on the MEF cell background, we also exposed HeLa cells to cotreatment with oncostatin M + H2O2 and demonstrated that the presence of H2O2 could sustain the STAT3 phosphorylation in response to the cytokine oncostatin M (Fig. 1C, and quantitative data in Supp. Fig. 1A). Similarly, to ensure that the responses were not specific for H2O2, we also exposed MEFs to the co-treatment of LIF + menadione, a superoxide generator, and again demonstrated the sustained STAT3 phosphorylation (Fig. 1D, and quantitative data in Supp. Fig. 1B), emphasizing that different initiators of oxidative stress can promote prolonged STAT3 activation in different cell types in the presence of IL6 family cytokines. Furthermore, replacing LIF + H2O2-containing medium with control serum-free medium could reverse the changes, implying that the observed increases in STAT3 phosphorylation are LIF

H2O2

LIF + H2O2

STAT3

α -tubulin

120 min

∗∗

3

∗∗

∗∗

1

∗

0

HeLa OSM

H2O2

OSM + H2O2

D

∗∗ ∗

1

∗∗∗

∗∗

∗∗

∗

∗ ∗∗

LIF

MEN

LIF + H2O2

LIF + MEN

Control

H2O2

LIF

E

LIF

H2O2

LIF + H2O2

LIF + H2O2

SF Media:

STAT3 (pY705)

STAT3 (pY705)

STAT3 (pS727)

STAT3 (pS727)

STAT3 (pS727)

STAT3

STAT3

STAT3

α -tubulin

α -tubulin

α -tubulin

∗∗∗

2

0

0 15 60 120 15 60 120 15 60 120 (min)

0 15 60 120 15 60 120 15 60 120 (min)

3

120 min

60 min

15 min

4

∗∗∗

2

Control

C

60 min

15 min

4 Relative Density (A.U) (pY705 STAT3/STAT3)

STAT3 (pS727)

We next evaluated whether upstream regulators of STAT3 phosphorylation could be responsible for the altered STAT3 phosphorylation and transcriptional activation following exposure to LIF + H2O2. There were no changes in the levels of JAK1 or 2, or in the kinetics of phosphorylation of JAK1 (pY1022) or JAK2 (pY1007) in the combined presence of LIF + H2O2 detected by immunoblotting (Supp. Fig. 3). The pan JAK inhibitor I prevented pY705 phosphorylation in response to LIF or LIF + H2O2 treatment (Fig. 3A, and quantitative data in Supp. Fig. 4A), consistent with the central role for the JAKs as the key mediators of STAT3 Y705 phosphorylation status. Since oxidative stress can also upregulate Src family kinase activity [34], we tested whether the Src family kinase inhibitor SU6656 could alter STAT3 phosphorylation. Markedly reduced levels of STAT3 Y705 phosphorylation upon LIF + H2O2 treatment were observed in the presence of SU6656

B

0 15 60 120 15 60 120 15 60 120 (min) STAT3 (pY705)

3.2. JAKs and Src family kinases, but not MAPKs, contribute to STAT3 phosphorylation during combined LIF + H2O2 exposure

Relative Density (A.U) (pS727 STAT3/STAT3)

A

485

15 60 0 15 60 120 (min) 0 30 60 0 30 60

STAT3 (pY705)

Fig. 1. Oxidative stress drives reversible persistent cytokine-stimulated STAT3 phosphorylation. (A & B) WT MEFs were treated with 10 ng/ml LIF, 1 mM H2O2 or LIF + H2O2 for the times indicated. (A) Lysates were subjected to immunoblot analysis for activated STAT3 (pY705 or pS727), total STAT3, or α-tubulin as a loading control. (B) Quantitative analysis of the immunoblots for STAT3, STAT3 pY705 and STAT3 pS727 (n = 3) was performed using ImageJ software. Results for STAT3 pY705 and STAT3 pS727 are expressed relative to those for total STAT3. Data represent the mean ± SEM. Asterisks denote statistically significant differences between cells treated compared to untreated cells (*, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001). (C) HeLa cells were treated with 10 ng/ml oncostatin M (OSM), 1 mM H2O2 or OSM + H2O2 for the times indicated, or (D) MEFs were treated with 10 ng/ml LIF, 50 μM menadione (MEN) or LIF + MEN. Lysates were subjected to immunoblot analysis as per (B). (E) MEFs were treated with LIF + H2O2 (0–120 min as indicated) prior to exposure to serum-free (SF) media (0–60 min as indicated). Lysates were subjected to immunoblot analysis as per (B).

486

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494 0 min

Relative Quantity SOCS3/GAPDH mRNA (A.U)

20

15 min

60 min

120 min

15

10

5

∗∗ ∗∗ ∗∗

∗∗

0 LIF

H2O2

LIF + H2O2

LIF

H2O2

LIF + H2O2

20 μ M STATTIC

Control

LIF

H2O2

LIF + H2O2

STAT3 -/- MEF

Fig. 2. Oxidative stress slows cytokine-stimulated STAT3-mediated changes in SOCS3 expression. WT MEFs, WT MEFs pre-treated (30 min) with 20 μM STATTIC, or STAT3−/− MEFs were treated with agents as per the legend to Fig. 1. SOCS3 and GAPDH mRNA levels were measured by quantitative real-time PCR and values normalised to the levels detected in the cells under control conditions (n = 3). Data represent the mean ± SEM. Asterisks denote statistically significant differences as per the legend to Fig. 1B.

(Fig. 3B, and quantitative data in Supp. Fig. 4B), implicating the involvement of Src family kinases in upstream regulatory events leading to STAT3 Y705 phosphorylation induced by LIF in the absence or presence of H2O2. As LIF + H2O2 treatment also increased pS727 STAT3 levels (Fig. 1A), we evaluated the mitogen-activated protein kinases (MAPKs) that have been implicated as direct regulators of STAT3 pS727 [33]. We examined the effects of inhibitors of the different MAPK pathways (JNK Inhibitor VIII directed at the JNKs, SB203580 directed at p38 MAPK, and UO126 directed at MEK1/2 within the ERK pathway). None of the inhibitors altered the LIF-, H2O2- or LIF + H2O2-stimulated increases in pY705 or pS727 STAT3, although

A

Control LIF

H2O2

0.5 µM JAK Inhibitor I LIF + H2O2

LIF

H2O2

LIF + H2O2

0 15 60 15 60 15 60 0 15 60 15 60 15 60 (min) STAT3 (pY705)

STAT3

α -tubulin

B

Control LIF

H2O2

10 µM SU6656 LIF + H2O2

0 15 60 15 60 15 60

LIF

H2O2

LIF + H2O2

0 15 60 15 60 15 60 (min)

STAT3 (pY705) STAT3

α -tubulin

Fig. 3. JAK and Src family tyrosine kinases contribute to persistent STAT3 Y705 phosphorylation driven by cytokine and oxidative stress stimulation. MEFs were pre-treated (1 h) with DMSO as a control, (A) 0.5 μM JAK Inhibitor I, or (B) 10 μM SU6656 before exposure to 10 ng/ml LIF, 1 mM H2O2 or LIF + H2O2 for the times indicated. Lysates were subjected to immunoblot analysis as per the legend to Fig. 1.

clearly being active in inhibiting phosphorylation of their respective specific pathway components (Fig. 4A, B and C, and quantitative data in Supp. Fig. 5A, B and C, respectively). Thus, proline-directed kinases other than these MAPKs are likely responsible for mediating the increases in pS727 STAT3 observed in response to LIF, H2O2 or LIF + H2O2.

3.3. H2O2 induces delayed LIF-dependent STAT3 nuclear translocation and mislocalisation of STAT3 phosphatase TC-PTP (TC45) to the cytoplasm As STAT3 phosphorylation is required for cytokine-stimulated STAT3 nuclear localisation [3], we evaluated STAT3 localisation, under conditions of combined cytokine stimulation and oxidative stress, using both total STAT3 and STAT3 pY705 antibodies. In contrast to H2O2 alone that had no effect, LIF induced a rapid increase in nuclear and phosphorylated STAT3 in the first 15 min of exposure followed by a decrease in nuclear levels (Fig. 5). Combined LIF + H2O2 treatment appeared to delay nuclear STAT3 localisation but resulted in prolonged nuclear STAT3 levels over the 2 h treatment that correlated with the increasing nuclear pY705 staining over the 2 h treatment (Fig. 5A and C). This was confirmed by quantitative analysis (Fig. 5B and D). Clearly, LIFinduced nuclear translocation kinetics are altered in the presence of H2O2. In addition to upregulated tyrosine kinase activity, sustained STAT3 phosphorylation can also result from reduced STAT3 phosphatase activity [35,36], with oxidative stress known to impact on tyrosine phosphatases in particular [37]. Since the nuclear phosphatase T-cell protein tyrosine phosphatase (TC-PTP) can dephosphorylate nuclear pY705STAT3 [9,35,38,39], we examined TC-PTP under conditions of LIF, H2O2 or LIF + H2O2 exposure. Immunoblot analysis revealed a significant (up to 60%) decrease in the total levels of detectable endogenous TCPTP following exposure of MEFs to H2O2 or LIF + H2O2 (Fig. 6A, and quantitative data in Fig. 6B); loss could be reduced by the inclusion of the proteasomal inhibitor MG132 (Fig. 6C, and quantitative data in Fig. 6D), thus implying proteasomal degradation of TC-PTP under the conditions of oxidative stress. The nuclear/cytoplasmic distribution of TC-PTP under conditions of oxidative stress, a factor of key importance in providing access to its nuclear or cytoplasmic substrates [25], was also examined by transiently transfecting MEFs to express myc epitope tagged-TC-PTP and performing immunostaining in the absence or presence of cytokine and H2O2. Our analysis confirmed the exclusive nuclear localisation of myc-TC-PTP in either the absence or presence of LIF alone, but revealed significant (p ≤ 0.001) relocalisation of myc-TC-PTP from the nucleus

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

A

Control LIF H2O2

B

20 µM JNK Inhibitor VIII LIF + H2O2

LIF H2O2

LIF + H2O2

Control LIF H2O2

0 15 60 15 60 1560 0 15 60 15 60 15 60 (min)

C

20 µM SB 203580 LIF + H2O2

LIF H2O2

487

LIF + H2O2

Control

0 15 60 15 60 1560 0 15 60 15 60 15 60 (min)

STAT3 (pY705)

STAT3 (pY705)

STAT3 (pS727)

STAT3 (pS727)

STAT3 ERK1/2 (pT202/pY204)

JNK1/2

ERK1/2

MAPKAP2 p38 (pT180/pY182)

α -tubulin

p38

LIF + H2O2

STAT3 (pS727)

MAPKAP2 (pT334)

c-Jun

H2O2

0 15 6015 60 15 60 0 15 60 15 60 15 60 (min)

MAPKAP2 (pT222)

JNK1/2 (pT183/pY185)

LIF

STAT3 (pY705)

STAT3

STAT3

20 µM UO126 LIF + H2O2

LIF H2O2

α -tubulin

α -tubulin

Fig. 4. Inhibition of MAPK pathways does not abrogate persistent STAT3 phosphorylation in the presence of cytokine and oxidative stress. MEFs were pre-treated with DMSO, (A) 20 μM JNK Inhibitor VIII, (B) 20 μM SB203580 (p38 inhibitor) or (C) 20 μM UO126 (MEK1/2 inhibitor) for 1 h before treatment with 10 ng/ml LIF, 1 mM H2O2 or LIF + H2O2 (15 or 60 min, as indicated). Cell lysates were prepared and subjected to immunoblot analysis for activated STAT3 (pY705 and pS727), total STAT3, or α-tubulin as a loading control. (A) Immunoblot analysis for phosphorylated JNK1/2 (pT183/pY185), total JNK1/2, and total c-Jun, confirmed JNK Inhibitor VIII actions. (B) Immunoblot analysis for phosphorylated MAPKAP2 (pT222 and pT334), total MAPKAP2, phosphorylated p38 (pT180/pY182), and total p38, confirmed SB203580 actions. (C) Immunoblot analysis for phosphorylated ERK1/2 (pT202/pY204) and total ERK1/2, confirmed UO126 actions.

to the cytosol at 60 min after treatment with H2O2 in either the absence or presence of LIF (Fig. 6E, and quantitative data in Fig. 6F). 3.4. Oxidative stress disrupts nuclear transport protein distributions and functions With our results highlighting altered cytoplasmic/nuclear distribution of both STAT3 and TC-PTP, we extended our analyses to 0 min

15 min

C

120 min

60 min

STAT3

0 min

15 min

120 min

60 min

pSTAT3

B

60 min

15 min

LIF

LIF

H2O2

H2O2

LIF + H2O2

LIF + H2O2

D

120 min

6

120 min

60 min

15 min

∗∗∗

∗∗∗ n= 74

∗∗∗

4

n= 77

STAT3 Fn/c 2

n= 63 n = 72

Control

0

LIF

n= 68

∗∗∗ ∗∗∗ n = ∗∗∗ ∗∗∗ n= n= 61

73

H2O2

n= 65

pSTAT3 Y705 (mean nuclear fluorescence/µm2)

200 n = 65

150

n= 72

n= 65

∗∗∗

100

n= 75

50

69

n= 58

LIF + H2O2

∗∗∗

n= 68

n= 70

n= 63

∗∗∗ n= 66

n= 57

n= 69

0 Control

A

explore more broadly the possibility that oxidative stress disrupts importin-dependent nuclear transport. Importins-β1 and -α3 and Ran have previously been implicated in STAT3 nuclear import during cytokine-stimulation [40–43], and we confirmed the reliance of STAT3 nuclear entry on importin-β1 and importin-α3 by using the Chariot peptide system to deliver importin-β1 and importin-α3 antibodies to cells. This showed the reduced nuclear accumulation of STAT3 in response to LIF stimulation (Supp. Fig. 6). We next

LIF

H2O2

LIF + H2O2

Fig. 5. Oxidative stress delays cytokine-stimulated STAT3 nuclear accumulation. MEFs were treated with agents as per the legend to Fig. 1. Confocal laser scanning microscopy analysis using cells immunostained with either (A) anti-STAT3-C20 or (C) anti-pSTAT3 Y705 primary and Cy2-coupled secondary antibodies. Bars represent 10 μm. (B) Quantitative analysis of STAT3 localisation using the ImageJ software to determine the nuclear to cytoplasmic ratio (Fn/c), representing measurements on at least 8 cells from 10 different microscopic fields for each time point for the nuclear (Fn) and cytoplasmic (Fc) fluorescence above background. (D) Quantitative analysis of pSTAT3 Y705 nuclear staining using the ImageJ software to determine the mean nuclear fluorescence intensity per μm2. Data represent the mean ± SEM. Asterisks denote statistically significant differences as per the legend to Fig. 1B.

488

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

LIF

A

H2O2

LIF + H2O2

B Relative Density (A.U) (TC-PTP/tubulin)

TC-PTP STAT3 (pY705)

STAT3

1.0

∗

∗ ∗∗

∗∗ ∗∗

0.5

α-tubulin

∗∗

0 LIF

Control

Control LIF

H2O2

20 µM MG132 LIF + H2O2

LIF

H2O2

D

LIF + H2O2

0 15 60 15 60 15 60 0 15 60 15 60 15 60 (min) TC-PTP

STAT3

Relative Density (A.U) (TC-PTP/tubulin)

C

120 min

60 min

15 min

1.5

0 15 60 120 15 60 120 15 60 120 (min)

H2O2

15 min

1.5

LIF + H2O2

60 min

1.0

∗∗

∗∗

∗∗

0.5

∗∗

α-tubulin LIF

H2O2 LIF + H2O2

LIF

Control

Control

0

20 μM MG132

Control

E

0 m in

15 min

60 m in

H2O2 LIF + H2O2

F

MYC-TC-PTP

15 min

LIF

H2O2

Myc-TC-PTP Fn/c

10

n= 51

8

n= 46

n= 41

60 min n= 43 n= 63

6 4

∗∗∗

2

n= 41

∗∗∗ n= 56

LIF + H2O2

Control

0 LIF

H2O2

LIF + H2O2

Fig. 6. Oxidative stress induces TC-PTP degradation and mislocalisation to the cytoplasm. MEFs were treated with agents as per the legend to Fig. 1. (A & B) Cell lysates were prepared and subjected to immunoblot analysis for TC-PTP, activated STAT3 (pY705), total STAT3, and α-tubulin as a loading control. (C & D) MEFs were pre-treated with either DMSO or 20 μM MG132 (proteasomal inhibitor) for 6 h before treatment with 10 ng/ml LIF, 1 mM H2O2 or LIF + H2O2 (as indicated). Cell lysates were prepared and subjected to immunoblot analysis for TC-PTP, STAT3, or α-tubulin as a loading control. (B & D) Quantitative analysis of the immunoblots for the levels of TC-PTP (n = 3) was performed using ImageJ software as per the legend to Fig. 1B, where data represent the mean ± SEM. Asterisks denote statistically significant differences as per the legend to Fig. 1B. (E & F) MEFs were transiently transfected to express Myc-tagged TC-PTP prior to treatment with 10 ng/ml LIF, 1 mM H2O2 or LIF + H2O2 for 15 or 60 min as indicated. (E) Confocal laser scanning microscopy analysis on cells immunostained using anti-myc primary and Cy2-coupled secondary antibodies. Bar represents 10 μm. (F) Image analysis of Myc-TC-PTP localisation using ImageJ software as described in the legend to Fig. 5B; data represent the mean ± SEM. Asterisks denote statistically significant differences between treated and control cells as per the legend to Fig. 1B.

examined the subcellular localisations of importins-β1 and -α3 and Ran in response to treatment with LIF, H2O2 or LIF + H2O2. Whilst importin-β1 localisation remained predominantly cytoplasmic/ nuclear envelope-localised throughout 60 min of each treatment (Fig. 7A, and quantitative data in Fig. 7B), both importin-α3 and Ran showed altered localisation even after 15 min of H2O 2 treatment, with importin-α3 accumulating in the nucleus (Fig. 7C, and quantitative data in Fig. 7D) and Ran mislocalising to the cytoplasm in the absence or presence of LIF (Fig. 7E, and quantitative data in Fig. 7F). Indeed, the aberrant localisation of both importin-α3 and

Ran proteins could be observed within 5 min of H2 O2 treatment (Ivan Ng, unpublished data). Quantitative image analysis to determine F n/c values confirmed these results, with importin-α3 and Ran showing significantly (p b 0.001) altered F n/c values in the presence of H2O2 regardless of LIF stimulation (Fig. 7D and F). Comparable nuclear and cytoplasmic mislocalisation of importin-α3 and Ran respectively was observed following exposure to 50 μM menadione to induce oxidative stress (Supp. Fig. 7). Thus, oxidative stress has clear effects on localisation of key components of the cellular nuclear transport machinery.

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

A

0 m in

15 min

489

60 m in

IMP-β 1

H2O2

B

0.3

Importin-α 1 Fn/c

LIF

n= 53

15 min

n= 60 n= 55

n= 66

n= 58 n= 45

n= 55

60 min

0.2

0.1

Control

0 LIF + H2O2

C

0 m in

15 min

LIF

H2O2

LIF + H2O2

60 m in

IMP-α3 LIF

∗∗∗

D

15 min

∗∗∗ Importin- α 3 Fn/c

3

H2O2

0 m in

15 min

n= 66

n= 45

n= 67

60 min

2

1

n= 53

0 Control

LIF + H2O2

E

n= 45

n= 58

n= 63

LIF

H2O2

LIF + H2O2

60 m in

Ran LIF

∗∗∗

F 3

Ran Fn/c

n= 53

H2O2

15 min

∗∗∗ n= 51

60 min

n= 56

2

1

n= 43

n= 56

n= 48

n= 51

Control

0 LIF + H2O2

LIF

H2O2

LIF + H2O2

Fig. 7. Oxidative stress disrupts importin-α3 and Ran but not importin-β1 subcellular localisation. MEFs were treated with agents as per the legend to Fig. 1. Confocal laser scanning microscopy analysis was performed on immunostained cells using specific primary antibodies for (A & B) importin-β1 (IMP-β1), (C & D) importin-α3 (IMP-α3) or (E & F) Ran, followed by the appropriate fluorescently labelled secondary antibodies. Bars represent 10 μm. (B, D & F) Quantitative results for importin-β1, importin-α3 and Ran localisation using ImageJ software to determine the Fn/c as per the legend to Fig. 5B; data represent the mean ± SEM. Asterisks denote statistically significant differences between treated and control cells as per the legend to Fig. 1B.

To investigate further the extent to which the effect of H2O2 on LIFdependent STAT3 nuclear translocation might relate to the mislocalisation of key components of STAT3's nuclear transport pathway, we tested the effect of pre-treatment with H2O2 on STAT3 trafficking.

Strikingly, even though H2O2 pre-treatment for 15 or 60 min caused redistribution of Ran to the cytoplasm and importin-α3 to the nucleus to similar extents, LIF-induced translocation of STAT3 from the cytoplasm to the nucleus was not abrogated (Fig. 8A, and quantitative data in

490

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

A

H2O2: + LIF:

No pre-treatment 0 min

60 min pre-treatment

15 min pre-treatment 0 min

60 min

15 min

15 min

0 min

60 min

15 min

STAT3

Ran

IMP-α3

B

+ LIF 0 min

∗∗∗ 6

n= 54

∗∗∗

∗∗∗

5

∗∗∗

Fn/c 3

n= 47

2 1

n= 50

n= 45 n= 42

n= 51

0 H2O2 Pre-treatment: 0 min

n= 54 n = 46

n= n = 42 n = 56 67

n= 56

15 min

∗∗∗

∗∗∗

n= n = 46 48

4

n= 45

n= n= n= 43 50 n = 56 n = n = 48 53 47

60 min

0 min

STAT3

C

+ LIF 60 min

+ LIF 15 min

∗∗

15 min

60 min

LIF:

0 min

15 min

60 min

0 min

0 min

0 min

n= 66

n= n= n= 48 51 53

0 min

15 min

60 min

Importin-α3

Ran

+ H2O2:

n= 48 n = 57

15 min pre-treatment 15 min

30 min

45 min

STAT3

Ran

IMP-α 3

∗∗∗

D

6

n= 53 n= 48

5

∗∗∗

n= 51 n = 42

∗∗∗

4

n= 50

Fn/c

3 n= 56

2 1 0 LIF pre-treatment: + H2O2:

n= 48

n= 42 n = 53

n= 51

n= 43 n= 42

n= n= n= 44 47 53

0 15 60 15 0 0 0 15 30 45

0 15 60 30 0 0 0 15 30 45

STAT3

Ran

n= n= n= 48 41 42

0 15 60 45 (min) 0 0 0 15 30 45 Importin-α3

60 min

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

A

H2O2 Pre-treatment : + LIF:

0 0

15

15 60

0

60

15 60

0

15

(min)

B 4 Relative Density (A.U) (pY705 STAT3/STAT3)

STAT3 ERK1/2 (pT202/pY204) ERK1/2

∗∗∗ ∗∗∗ ∗

3

2

1

0 H2O2 Pre-treatment:

α -tubulin

-

15 60 -

15

(min)

STAT3 (pY705) STAT3 ERK1/2 (pT202/pY204) ERK1/2

α -tubulin

0 min

15 min

60 min

D

- 15 30 45

4 Relative Density (A.U) (pY705 STAT3/STAT3)

LIF : 0 + H2O2 :

+ LIF 60 min

+ LIF 15 min

60

STAT3 (pY705)

C

+ LIF 0 min

491

∗∗∗

3

2

1

0 LIF pre-treatment: + H2O2:

0 0

15 60 15 0 0 15 30 45

Fig. 9. Phosphorylated pY705 STAT3 levels are sustained in different cytokine and oxidative stress pre-treatment protocols. (A) MEFs were pre-treated with 1 mM H2O2 (0–60 min as indicated) prior to exposure to 10 ng/ml LIF for 15 or 60 min in the continued presence of 1 mM H2O2. (C) MEFs were either treated with 10 ng/ml LIF (0–60 min as indicated) or pre-treated with 10 ng/ml LIF (15 min) prior to exposure to 1 mM H2O2 for 15 or 60 min in the continued presence of 10 ng/ml LIF. (A & C) Cell lysates were prepared and subjected to immunoblot analysis for activated STAT3 (pY705), total STAT3, activated ERK1/2 (pT202/pY204), total ERK1/2 or α-tubulin as a loading control. (B & D) Quantitative analysis of the immunoblots for STAT3 and STAT3 pY705 (n = 3) was performed using ImageJ software, where data represent the mean ± SEM. Asterisks denote statistically significant differences as per the legend to Fig. 1B.

Fig. 8B). Further, H2O2 pre-treatment for 15 min resulted in higher STAT3 nuclear accumulation 60 min after LIF addition compared to cells not pre-treated with H2O2. Experiments were also performed where H2O2 was added 15 min post-LIF treatment, conditions which again mislocalised Ran and importin-α3 to the cytoplasm and nucleus respectively (Fig. 8C, and quantitative data in Fig. 8D); in this case, H2O2 treatment post-LIF addition prolonged the high levels of STAT3 in the nucleus (Fig. 8C and D; Fn/c ~ 4–5) up to 60 min. Evaluation of pY705 STAT3 levels under these different conditions confirmed sustained phosphorylation of STAT3 (Fig. 9) as the likely basis of the prolonged nuclear retention of STAT3. To confirm that oxidative stress-induced alterations in the nuclear transport machinery impacted on cellular nuclear transport generally, we examined the effect of H2O2 on the subcellular localisation of the NLS-containing GFP-tagged nuclear import cargoes GFP-conjugated SV40 large T-antigen (GFP-T-ag) and human cytomegalovirus polymerase UL54 (GFP-pUL54), representative of importin-α/β1-dependent transport [27,28], and the importin-β1-specific nuclear import cargo GFP-VP3 (chicken anaemia virus viral protein 3) [29,30]; since VP3 also contains a CRM1 (exportin 1)-recognised nuclear export signal (NES), we used a NES-inactivated mutant derivative, whilst a VP3 construct containing an intact NES, in addition to the NLS/NES-containing

HIV-1 Rev protein [26], were also analysed. All of the nuclear import cargoes showed significantly reduced (p ≤ 0.001) nuclear accumulation, in contrast to GFP alone which showed no effect of H2O2 treatment, implying no impact of oxidative stress on passive diffusion through the nuclear pore (Fig. 10A, and quantitative data in Fig. 10B). We confirmed that CRM1-mediated nuclear export of GFP-Rev-NES was sensitive to the CRM1 inhibitor leptomycin B (Supp. Fig. 8). GFP-Rev-NES was unaffected in the presence of H2O2 (Fig. 10), implying that nuclear export was still functional and this was further confirmed through the evaluation of the export of endogenous RanBP1 that continued in the presence of H2O2 (Supp. Fig. 9). Thus, the main impact on nucleocytoplasmic transport of Ran mislocalisation due to H2O2 treatment appears to be on importin-dependent nuclear import rather than nuclear export. 4. Discussion Cytokine-stimulated STAT3 phosphorylation has been typically shown to be both rapid and transient under normal, non-stressed conditions. Detailed studies both in vitro and in vivo have revealed a coordinated negative regulatory network contributing to the kinetics of this activation, highlighting the importance of proteins such as TC-PTP [9] and SOCS3 [11]. A persistent activation of JAK/STAT3 signalling has

Fig. 8. Oxidative stress impacts on LIF-dependent STAT3 nuclear accumulation either pre- or post-cytokine addition. (A & B) MEFs were pre-treated with 1 mM H2O2 for the indicated times prior to exposure to 10 ng/ml LIF for 15 or 60 min in the continued presence of H2O2. (A) Confocal laser scanning microscopy analysis was performed on cells immunostained using specific antibodies for STAT3, Ran and importin-α3 (IMP-α3) as per the legends to Figs. 5A and 7. (B) Quantitative results for STAT3, Ran and importin-α3 localisation using ImageJ software to determine the Fn/c as per the legend to Fig. 5B; data represent the mean ± SEM. Asterisks denote statistically significant differences for cells analysed under control conditions (**, p ≤ 0.01; ***, p ≤ 0.001). (C & D) MEFs were pre-treated with 10 ng/ml LIF for 15 min prior to exposure to 1 mM H2O2 (15–45 min as indicated) in the continued presence of LIF. (C) Confocal laser scanning microscopic analysis was performed on immunostained cells as per (A & B) above. (D) Quantitative results for STAT3, Ran and importin-α3 localisation using ImageJ software to determine the Fn/c as per the legend to Fig. 5B; data represent the mean ± SEM. (A & C) Bars represent 10 μm.

492

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

A

GFP

GFP-Rev-NES

GFP-UL54 (1145-1161)

GFP-T-ag (114-135)

GFP-VP3 (74-121)

GFP-VP3 NES mut

Control

1 mM H2O2 (60 min)

B

∗∗∗

15

n= 51

∗∗∗

-H 2O2 + H2O2

n= 49

10

∗∗∗

GFP (Fn/c)

∗∗∗ n= 50

n= 60

5 n= 55 n= 53

n= 47

n= 43

n= 46 n= 48

n= 49

n= 52

0 GFP

GFP Rev-NES

GFP T-ag

GFP UL54

GFP VP3

GFP VP3 NES mut

Fig. 10. Oxidative stress inhibits importin-α/β-dependent nuclear protein import. MEFs were transiently transfected to express GFP-tagged cargo proteins as indicated prior to treatment with 1 mM H2O2 for 60 min. (A) Live cell confocal laser scanning microscopic imaging. Bar represents 10 μm. (B) Quantitative results for the various GFP-tagged proteins using ImageJ software to determine the Fn/c as per the legend to Fig. 5B; data represent the mean ± SEM. Asterisks denote statistically significant differences for treated compared to untreated cells.

been implicated in the disease progression, and so a deeper understanding of the molecular drivers of this aberrant regulation is urgently required. Oxidative stress induced by H2O2, superoxide or ultraviolet radiation can activate STAT3 in the absence of cytokine stimulation [15,16,44,45]. Our studies here have addressed for the first time the impact of oxidative stress on cytokine-driven STAT3 signalling events.

As summarized in Fig. 11A, the key cytokine-stimulated events regulating STAT3 have been well described. However, as summarized in Fig. 11B, we have shown the simultaneous exposure to H2O2 and the cytokine LIF drives a striking and persistent phosphorylation of STAT3, but that the activation of SOCS3 expression as a direct target gene of STAT3 is slowed and the nuclear import machinery exemplified by

Fig. 11. Schematic diagram of the impact of oxidative stress on cytokine-driven STAT3 activation. (A) Under normal (non-stressed) conditions, the cytokine LIF activates the JAK-STAT3 signal transduction pathway in which negative regulators such as TC-PTP and SOCS3 limit the duration of signalling events. (B) During oxidative stress, multiple regulatory events of the STAT3 pathway are altered: TC-PTP is mislocalised and its levels are downregulated, Ran and importin-α3 distributions are altered. The result of these changes is seen in the sustained phosphorylation and nuclear localisation of STAT3.

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494

cytosolic/nuclear distributions of importin-α3 and Ran is disrupted. In combination, these changes slow the nuclear accumulation of STAT3 but also allow for its persistent nuclear retention. Studies on the impact of stress on nuclear transport events have previously defined changes in importin localisation and the Ran gradient [46–49]. Importantly, these disruptions noted under the conditions of ultraviolet radiation, oxidative stress, and heat shock are predicted to have a profound influence on the nuclear transport of the wide range of nuclear-destined cargoes dependent on these importin proteins. Previous studies have shown that STAT3 nuclear import relies on importin-α3 that interacts with STAT3's NLS ( 150 RKRQDLEQKMK 162 within the coiled-coil domain [42]), importin-β1, and a Ran gradient [41]. Here, we demonstrate the striking redistribution of importin-α3 and Ran, as well as a relocalisation of the phosphatase TC-PTP and a delay in the nuclear entry of STAT3 during oxidative stress. Paradoxically, a cytokinedriven nuclear entry of STAT3 remains possible despite the importin-α3 and Ran disruptions and can be observed even when the oxidative stress insult precedes cytokine stimulation. With such disruptions to the classical nuclear import pathway, the possibility for a non-Ran dependent nuclear transport mechanism for critical proteins such as the STAT3 transcription factor requires careful consideration. Activated STAT3 can accumulate in the nucleus through non-canonical nuclear import pathways [50–53]. Indeed, an import pathway requiring receptor-mediated endocytosis trafficking has been proposed as critical for signalling by cytokines that are weak STAT3 activators. Thus, the formation of signalling endosomes protects and prolongs the activation of signalling proteins, but this is not required for the potent activation of STAT3 by interleukin-6 family cytokines [51–53]. Alternative stress-evoked mechanisms have been proposed as critical for the nuclear transport of Hsp70 during heat stress [54,55]; whether these may apply to other types of stress, including oxidative stress, remains to be determined. This will be particularly important in understanding how critical mediators of stress responses such as signalling proteins and transcription factors can continue to enter the nucleus of the stressed cell. Indeed, a number of protein kinases are reported to show enhanced nuclear retention in response to stress [56], suggesting their nuclear entry will also require mechanisms that are not reliant on the classical Ran-dependent importin pathways. The sustained nuclear retention of cytokine-activated STAT3 we observed during oxidative stress also challenges our understanding of the STAT3 export pathway. Importantly, the dephosphorylation of STAT3 Y705 is required for its nuclear export by CRM1 [57]. The expression and localisation of TC-PTP are critical regulators of nuclear STAT3 Y705 phosphorylation [9,35,38,39]. The mislocalisation of TC-PTP has been reported previously for cells subjected to hyperosmotic stress [25]. In that study, as the hyperosmotic stress-induced nuclear exit of TC-PTP was not inhibited by leptomycin B, the nuclear export of TCPTP appeared to be independent of the exportin CRM1 [25]. Similarly, we observed TC-PTP mislocalisation during oxidative stress by tracking the localisation of myc-tagged TC-PTP. Although further studies will require more extensive evaluation of factors impacting on the localisation of the endogenous TC-PTP under conditions of oxidative stress (i.e., rather than following exogenously expressed myc-tagged TC-PTP protein, anti-TC-PTP antibodies suitable for use in immunostaining protocols will be needed), our observed loss of the endogenous TCPTP protein during oxidative stress, adds to the key stress-stimulated events impeding STAT3 dephosphorylation in the nucleus (see Fig. 11B). Our analysis of the nuclear export of the CRM1-dependent cargoes GFP-Rev-NES or RanBP1 clearly demonstrated that impairment of the Ran gradient was not sufficient to affect their predominantly cytoplasmic localisations. Thus, the prolonged STAT3 phosphorylation would appear to be the key to the enhanced nuclear localisation of STAT3 under the conditions of combined cytokine and oxidative stress stimulation.

493

The significance of our findings is reinforced by the intense interest directed towards STAT3 as part of altered signalling events in a range of pathologies, and a concerted effort to define signalling nodes for therapeutic targeting [58,59]. Some studies have suggested that oxidative stress per se can induce STAT3 Y705 phosphorylation as a downstream consequence of growth factor activation [34,60]. Our results reveal a requirement for JAKs and Src family kinases in the activation of STAT3 driven by oxidative stress and cytokine stimulation and suggest that targeting these kinases will contribute to alleviating detrimental consequences of STAT3 activation noted in disease. Taken together, the results presented in our study reveal that the disruption of several regulatory mechanisms leads to a delayed nuclear accumulation and transcriptional response by STAT3. Acknowledgements We thank V. Poli (University of Turin, Italy) for the wild-type and STAT3−/− murine embryonic fibroblasts. This work was supported by an Australian Research Council (ARC) Discovery Grant (DP130100804). IHWN is a recipient of a Monash University: Monash Graduate Scholarship and Faculty of Medicine International Postgraduate Research Scholarship and DAJ is an NHMRC Senior Principal Research Fellow (#APP1002486). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.bbamcr.2013.11.015. References [1] 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. [2] C. Mertens, J.E. Darnell Jr., SnapShot: JAK-STAT signaling, Cell 131 (2007) 612. [3] D.E. Levy, J.E. Darnell Jr., Stats: transcriptional control and biological impact, Nat. Rev. Mol. Cell Biol. 3 (2002) 651–662. [4] P.C. Heinrich, I. Behrmann, S. Haan, H.M. Hermanns, G. Muller-Newen, F. Schaper, Principles of interleukin (IL)-6-type cytokine signalling and its regulation, Biochem. J. 374 (2003) 1–20. [5] V. Poli, STAT3 function in vivo, in: P.B. Sehgal, D.E. Levy, T. Hirano (Eds.), Signal Transducers and Activators of Transcription (STATs), Kluwer Academic Publishers, The Netherlands, 2003, pp. 492–512. [6] H. Yu, R. Jove, The STATs of cancer-new molecular targets come of age, Nat. Rev. Cancer 4 (2004) 97–105. [7] R. Garcia, T.L. Bowman, G. Niu, H. Yu, S. Minton, C.A. Muro-Cacho, C.E. Cox, R. Falcone, R. Fairclough, S. Parsons, A. Laudano, A. Gazit, A. Levitzki, A. Kraker, R. Jove, Constitutive activation of Stat3 by the Src and JAK tyrosine kinases participates in growth regulation of human breast carcinoma cells, Oncogene 20 (2001) 2499–2513. [8] T. Mustelin, T. Vang, N. Bottini, Protein tyrosine phosphatases and the immune response, Nat. Rev. Immunol. 5 (2005) 43–57. [9] D. Xu, C.K. Qu, Protein tyrosine phosphatases in the JAK/STAT pathway, Front. Biosci. 13 (2008) 4925–4932. [10] U. Vinkemeier, Getting the message across, STAT! Design principles of a molecular signaling circuit, J. Cell Biol. 167 (2004) 197–201. [11] J.J. Babon, N.A. Nicola, The biology and mechanism of action of suppressor of cytokine signaling 3, Growth Factors 30 (2012) 207–219. [12] N. Khaper, S. Bryan, S. Dhingra, R. Singal, A. Bajaj, C.M. Pathak, P.K. Singal, Targeting the vicious inflammation-oxidative stress cycle for the management of heart failure, Antioxid. Redox Signal. 13 (2010) 1033–1049. [13] L.A. Sena, N.S. Chandel, Physiological roles of mitochondrial reactive oxygen species, Mol. Cell 48 (2012) 158–167. [14] C. Zgheib, M. Kurdi, F.A. Zouein, B.W. Gunter, B.A. Stanley, J. Zgheib, D.G. Romero, S.B. King, N. Paolocci, G.W. Booz, Acyloxy nitroso compounds inhibit LIF signaling in endothelial cells and cardiac myocytes: evidence that STAT3 signaling is redox-sensitive, PLoS One 7 (2012) e43313. [15] S.K. Park, M.K. Dahmer, M.W. Quasney, MAPK and JAK-STAT signaling pathways are involved in the oxidative stress-induced decrease in expression of surfactant protein genes, Cell. Physiol. Biochem. 30 (2012) 334–346. [16] G. Millonig, I. Ganzleben, T. Peccerella, G. Casanovas, L. Brodziak-Jarosz, K. Breitkopf-Heinlein, T.P. Dick, H.K. Seitz, M.U. Muckenthaler, S. Mueller, Sustained submicromolar H2O2 levels induce hepcidin via signal transducer and activator of transcription 3 (STAT3), J. Biol. Chem. 287 (2012) 37472–37482. [17] B.G. Szczepankiewicz, C. Kosogof, L.T. Nelson, G. Liu, B. Liu, H. Zhao, M.D. Serby, Z. Xin, M. Liu, R.J. Gum, D.L. Haasch, S. Wang, J.E. Clampit, E.F. Johnson, T.H. Lubben, M.A. Stashko, E.T. Olejniczak, C. Sun, S.A. Dorwin, K. Haskins, C. Abad-Zapatero,

494

[18]

[19]

[20]

[21]

[22] [23] [24]

[25]

[26]

[27]

[28]

[29]

[30]

[31] [32]

[33] [34] [35]

[36] [37] [38]

I.H.W. Ng et al. / Biochimica et Biophysica Acta 1843 (2014) 483–494 E.H. Fry, C.W. Hutchins, H.L. Sham, C.M. Rondinone, J.M. Trevillyan, Aminopyridine-based c-Jun N-terminal kinase inhibitors with cellular activity and minimal cross-kinase activity, J. Med. Chem. 49 (2006) 3563–3580. M.F. Favata, K.Y. Horiuchi, E.J. Manos, A.J. Daulerio, D.A. Stradley, W.S. Feeser, D.E. Van Dyk, W.J. Pitts, R.A. Earl, F. Hobbs, R.A. Copeland, R.L. Magolda, P.A. Scherle, J.M. Trzaskos, Identification of a novel inhibitor of mitogen-activated protein kinase kinase, J. Biol. Chem. 273 (1998) 18623–18632. A. Cuenda, J. Rouse, Y.N. Doza, R. Meier, P. Cohen, T.F. Gallagher, P.R. Young, J.C. Lee, SB 203580 is a specific inhibitor of a MAP kinase homologue which is stimulated by cellular stresses and interleukin-1, FEBS Lett. 364 (1995) 229–233. J.E. Thompson, R.M. Cubbon, R.T. Cummings, L.S. Wicker, R. Frankshun, B.R. Cunningham, P.M. Cameron, P.T. Meinke, N. Liverton, Y. Weng, J.A. DeMartino, Photochemical preparation of a pyridone containing tetracycle: a Jak protein kinase inhibitor, Bioorg. Med. Chem. Lett. 12 (2002) 1219–1223. R.A. Blake, M.A. Broome, X. Liu, J. Wu, M. Gishizky, L. Sun, S.A. Courtneidge, SU6656, a selective src family kinase inhibitor, used to probe growth factor signaling, Mol. Cell. Biol. 20 (2000) 9018–9027. J. Schust, B. Sperl, A. Hollis, T.U. Mayer, T. Berg, Stattic: a small-molecule inhibitor of STAT3 activation and dimerization, Chem. Biol. 13 (2006) 1235–1242. M.C. Morris, S. Deshayes, F. Heitz, G. Divita, Cell-penetrating peptides: from molecular mechanisms to therapeutics, Biol. Cell 100 (2008) 201–217. M.C. Morris, J. Depollier, J. Mery, F. Heitz, G. Divita, A peptide carrier for the delivery of biologically active proteins into mammalian cells, Nat. Biotechnol. 19 (2001) 1173–1176. M.H. Lam, B.J. Michell, M.T. Fodero-Tavoletti, B.E. Kemp, N.K. Tonks, T. Tiganis, Cellular stress regulates the nucleocytoplasmic distribution of the protein-tyrosine phosphatase TCPTP, J. Biol. Chem. 276 (2001) 37700–37707. M.H. Lam, B. Henderson, M.T. Gillespie, D.A. Jans, Dynamics of leptomycin B-sensitive nucleocytoplasmic flux of parathyroid hormone-related protein, Traffic 2 (2001) 812–819. S. Hubner, C.Y. Xiao, D.A. Jans, The protein kinase CK2 site (Ser111/112) enhances recognition of the simian virus 40 large T-antigen nuclear localization sequence by importin, J. Biol. Chem. 272 (1997) 17191–17195. G. Alvisi, A. Ripalti, A. Ngankeu, M. Giannandrea, S.G. Caraffi, M.M. Dias, D.A. Jans, Human cytomegalovirus DNA polymerase catalytic subunit pUL54 possesses independently acting nuclear localization and pp UL44 binding motifs, Traffic 7 (2006) 1322–1332. H.V. Kuusisto, K.M. Wagstaff, G. Alvisi, D.A. Jans, The C-terminus of apoptin represents a unique tumor cell-enhanced nuclear targeting module, Int. J. Cancer 123 (2008) 2965–2969. I.K. Poon, C. Oro, M.M. Dias, J. Zhang, D.A. Jans, Apoptin nuclear accumulation is modulated by a CRM1-recognized nuclear export signal that is active in normal but not in tumor cells, Cancer Res. 65 (2005) 7059–7064. Y. Yashiroda, M. Yoshida, Nucleo-cytoplasmic transport of proteins as a target for therapeutic drugs, Curr. Med. Chem. 10 (2003) 741–748. I.H. Ng, D.C. Ng, D.A. Jans, M.A. Bogoyevitch, Selective STAT3-alpha or -beta expression reveals spliceform-specific phosphorylation kinetics, nuclear retention and distinct gene expression outcomes, Biochem. J. 447 (2012) 125–136. T. Decker, P. Kovarik, Serine phosphorylation of STATs, Oncogene 19 (2000) 2628–2637. G. Sun, D.J. Kemble, To C or not to C: direct and indirect redox regulation of Src protein tyrosine kinase, Cell Cycle 8 (2009) 2353–2355. B.J. Shields, F. Wiede, E.N. Gurzov, K. Wee, C. Hauser, H.J. Zhu, T.J. Molloy, S.A. O'Toole, R.J. Daly, R.L. Sutherland, C.A. Mitchell, C.A. McLean, T. Tiganis, TCPTP regulates SFK and STAT3 signaling and is lost in triple-negative breast cancers, Mol. Cell. Biol. 33 (2013) 557–570. T. Yi, D. Lindner, The role and target potential of protein tyrosine phosphatases in cancer, Curr. Oncol. Rep. 10 (2008) 114–121. C. Blanchetot, L.G. Tertoolen, J. den Hertog, Regulation of receptor protein-tyrosine phosphatase alpha by oxidative stress, EMBO J. 21 (2002) 493–503. R. Wakahara, H. Kunimoto, K. Tanino, H. Kojima, A. Inoue, H. Shintaku, K. Nakajima, Phospho-Ser727 of STAT3 regulates STAT3 activity by enhancing dephosphorylation of phospho-Tyr705 largely through TC45, Genes Cells 17 (2012) 132–145.

[39] T. Yamamoto, Y. Sekine, K. Kashima, A. Kubota, N. Sato, N. Aoki, T. Matsuda, The nuclear isoform of protein-tyrosine phosphatase TC-PTP regulates interleukin6-mediated signaling pathway through STAT3 dephosphorylation, Biochem. Biophys. Res. Commun. 297 (2002) 811–817. [40] S. Huang, I.S. Chang, W. Lin, W. Ye, R.Z. Luo, Z. Lu, Y. Lu, K. Zhang, W.S. Liao, T. Tao, R.C. Bast Jr., X. Chen, Y. Yu, ARHI (DIRAS3), an imprinted tumour suppressor gene, binds to importins and blocks nuclear import of cargo proteins, Biosci. Rep. 30 (2010) 159–168. [41] V. Cimica, H.C. Chen, J.K. Iyer, N.C. Reich, Dynamics of the STAT3 transcription factor: nuclear import dependent on Ran and importin-beta1, PLoS One 6 (2011) e20188. [42] L. Liu, K.M. McBride, N.C. Reich, STAT3 nuclear import is independent of tyrosine phosphorylation and mediated by importin-alpha3, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 8150–8155. [43] R. Ushijima, N. Sakaguchi, A. Kano, A. Maruyama, Y. Miyamoto, T. Sekimoto, Y. Yoneda, K. Ogino, T. Tachibana, Extracellular signal-dependent nuclear import of STAT3 is mediated by various importin alphas, Biochem. Biophys. Res. Commun. 330 (2005) 880–886. [44] M. Carballo, M. Conde, R. El Bekay, J. Martin-Nieto, M.J. Camacho, J. Monteseirin, J. Conde, F.J. Bedoya, F. Sobrino, Oxidative stress triggers STAT3 tyrosine phosphorylation and nuclear translocation in human lymphocytes, J. Biol. Chem. 274 (1999) 17580–17586. [45] T. Bito, N. Sumita, T. Masaki, T. Shirakawa, M. Ueda, R. Yoshiki, Y. Tokura, C. Nishigori, Ultraviolet light induces Stat3 activation in human keratinocytes and fibroblasts through reactive oxygen species and DNA damage, Exp. Dermatol. 19 (2010) 654–660. [46] M. Kodiha, A. Chu, N. Matusiewicz, U. Stochaj, Multiple mechanisms promote the inhibition of classical nuclear import upon exposure to severe oxidative stress, Cell Death Differ. 11 (2004) 862–874. [47] Y. Miyamoto, T. Saiwaki, J. Yamashita, Y. Yasuda, I. Kotera, S. Shibata, M. Shigeta, Y. Hiraoka, T. Haraguchi, Y. Yoneda, Cellular stresses induce the nuclear accumulation of importin alpha and cause a conventional nuclear import block, J. Cell Biol. 165 (2004) 617–623. [48] N. Crampton, M. Kodiha, S. Shrivastava, R. Umar, U. Stochaj, Oxidative stress inhibits nuclear protein export by multiple mechanisms that target FG nucleoporins and Crm1, Mol. Biol. Cell 20 (2009) 5106–5116. [49] Y. Yasuda, Y. Miyamoto, T. Saiwaki, Y. Yoneda, Mechanism of the stress-induced collapse of the Ran distribution, Exp. Cell Res. 312 (2006) 512–520. [50] J.L. Lee, M.J. Wang, J.Y. Chen, Acetylation and activation of STAT3 mediated by nuclear translocation of CD44, J. Cell Biol. 185 (2009) 949–957. [51] S. Kermorgant, P.J. Parker, Receptor trafficking controls weak signal delivery: a strategy used by c-Met for STAT3 nuclear accumulation, J. Cell Biol. 182 (2008) 855–863. [52] P.B. Sehgal, Paradigm shifts in the cell biology of STAT signaling, Semin. Cell Dev. Biol. 19 (2008) 329–340. [53] A.H. Bild, J. Turkson, R. Jove, Cytoplasmic transport of Stat3 by receptor-mediated endocytosis, EMBO J. 21 (2002) 3255–3263. [54] N. Imamoto, S. Kose, Heat-shock stress activates a novel nuclear import pathway mediated by Hikeshi, Nucleus 3 (2012) 422–428. [55] S. Kose, M. Furuta, N. Imamoto, Hikeshi, a nuclear import carrier for Hsp70s, protects cells from heat shock-induced nuclear damage, Cell 149 (2012) 578–589. [56] M. Kodiha, U. Stochaj, Nuclear transport: a switch for the oxidative stress-signaling circuit? J. Signal Transduct. 2012 (2012) 208650. [57] Y. Huang, J. Qiu, S. Dong, M.S. Redell, V. Poli, M.A. Mancini, D.J. Tweardy, Stat3 isoforms, alpha and beta, demonstrate distinct intracellular dynamics with prolonged nuclear retention of Stat3beta mapping to its unique C-terminal end, J. Biol. Chem. 282 (2007) 34958–34967. [58] E. Bournazou, J. Bromberg, Targeting the tumor microenvironment: JAK-STAT3 signaling, JAK-STAT 2 (2013) e23828. [59] P. Sansone, J. Bromberg, Targeting the interleukin-6/Jak/stat pathway in human malignancies, J. Clin. Oncol. 30 (2012) 1005–1014. [60] S. Filosto, E.M. Khan, E. Tognon, C. Becker, M. Ashfaq, T. Ravid, T. Goldkorn, EGF receptor exposed to oxidative stress acquires abnormal phosphorylation and aberrant activated conformation that impairs canonical dimerization, PLoS One 6 (2011) e23240.