Gurgiset al. MK2-HuR pathway in glioblastoma inflammation

Supplementary Figures Legends

Supplementary Figure 1. (A) U87 and U87-EGFRvIII (5 x 104) cells were treated with IL-1β (0-100 ng/mL) for 24 h and IL-8 secretion was determined by ELISA (for details see Methodology below). Values were normalised to cell viability (MTT) and expressed as fold-increase (mean ± SEM) compared to untreated U87 cells. (B-C) U87 and U87-EGFRvIII cells were incubated with fluorescein-labelled p38 MAPK inhibitor SB203580 (10μM) for 30 and 60 min. Cells were lysed (B) and fluorescence intensity of 100 μg whole cell lysate was measured (for details see Methodology below). Alternatively, (C) cells were trypsinised, resuspended and fluorescence intensity was analysed by flow cytometry. (D) U87 and U87-EGFRvIII cells were starved for 2 h, pre-treated with SB203580 (SB, 10μM) for 60 min and incubated with IL-1β (10 ng/mL) for 24 h. IL-6 in supernatants (extracellular IL-6) and cell lysates (intracellular IL-6) was determined by ELISA and normalised to total protein concentration (BCA assay). Data are expressed relative to IL-6 levels in IL-1β treated cells and represent the mean ± SEM from 3 independent experiments performed in duplicates. (E) U87 and U87-EGFRvIII (3 x 105) cells were starved for 2 h, pre-treated with 10 μM SB203580 (SB) and treated with IL-1β (10 ng/mL) for 24 h. IL-8 mRNA was determined by RT-PCR and normalised to 18s mRNA as described(1-3). Data are expressed as fold-increase compared to untreated cells and represent the mean ± SEM from at least 3 independent experiments performed in duplicates. (**P < 0.01, ***P < 0.001, 1-way ANOVA followed by Newman-Keuls post-test using Prism 5 GraphPad Software).

Supplementary Figure 2. (A-C) U87, U87-EGFRvIII and U87-EGFRwt (4 x 105) cells were starved for 2 h and treated with IL-1β (10 ng/mL) for 0-120 min.Whole cell lysates were prepared and protein concentration was determined (for details see Methodology below). 25 μg cellular protein were subjected to Western blot analysis as described(4) and analysed forp38 MAPK phosphorylation (p-p38), total p38 MAPK (p38), phosphorylated MK2 (p-MK2), total MK2 (MK2), phosphorylated Hsp27 (p-Hsp27) and total Hsp27 (Hsp27) as indicated. Representative blots of three independent experiments are shown.

Supplementary Figure 3. (A) U87 and U87-EGFRvIII (4 x 105) cells were starved for 2 h and treated with IL-1β (10 ng/mL) for 2 h. Whole cell lysates from U87 (lanes 1-2) and U87-EGFRvIII (lanes 3-4) were analysed for protein concentration. 25 μg cellular protein were subjected to Western blot analysis as described (4) and analysed forCREB and ATF-1 phosphorylation (p-CREB, p-ATF-1), respectively, and total CREB (CREB) as indicated. Representative blots of three independent experiments are shown. (B) U87 and U87-EGFRvIII (3x105) were transfected with 1 μg IL-6-luc651 and 1 μg pSV-β-galactosidase for 24 h. Cells were then starved for 2 h, pre-treated with 10 μM SB203580 for 30 min, followed by IL-1β (10 ng/mL) stimulation for 24 h. Cells were harvested, and luciferase and β-galactosidase activities determined as described (1).

Supplementary Figure 4. (A) Whole cell lysates fromU87 and U87-EGFRvIII were analysed for protein concentration and 20 – 40μg cellular protein were subjected to Western blot analysis as described (4). Immunoblots were analysed for total p38 MAPK (p38), phosphorylated and total MK2 (p-MK2 and MK2). Protein expression was determined and normalised against β-tubulin. Representative titration blots and quantification of relative p38 MAPK (p38) and MK2 protein expression from three independent experiments are shown.(B) U87 cells were transfected with HuR siRNA or scrambled siRNA as a negative control for 72 h(for details see Methodology below). Whole cell lysates were subjected to Western blot analysis under non-reducing conditions to detect HuR monomer (37 kD) and dimer (75 kD). Rabbit polyclonal anti-HuR (#20694, Santa Cruz) and mouse monoclonal anti-HuR (#5261, Santa Cruz) antibodies were used as indicated. β-tubulin served as a loading control.

Supplementary Figure 5. (A-B) U87 and U87-EGFRvIII (1 x 107) cells were activated with IL-1β (10 ng/mL) for 0-30 min. Nuclearand cytoplasmic fractions were prepared and analysed by Western blotting for phosphorylated and total MK2(p-MK2, MK2), phosphorylated and total Cdk1 (p-Cdk1, Cdk1),phosphorylated and total HuR (pHuR(S202), HuR), lamin A/C and β-tubulin (for details see Methodology below).Relative p-MK2 and HuR levels were quantified and normalised to lamin A/C (nuclear fractions) and β-tubulin (cytosolic fractions). Relative p-Cdk1 and p-HuR(S202) levels were quantified and normalised to total Cdk1 and total HuR, respectively. Representative blots and data (mean ± SEM) from two independent experiments are shown. (C)U87 and U87-EGFRvIII (1x107) were starved for 2 h, pre-treated with SB203580 (10 μM) for 60 min and activated with IL-1β (10 ng/mL) for 8 h. Protein concentration of nuclear and cytosolic fractions was determined and 25 μg of nuclear protein fraction was subjected to Western blot analysis. Nuclear distribution of HuR, lamin A/C and β-tubulin is shown (representative for three independent experiments).

Supplementary Table 1.

U87 and U87-EGFRvIII (5 x 104) cells were starved for 2 h, pre-treated with compounds (10 μM) inhibiting p38 MAPK (SB203580), MEK1/2 (PD98059) or JNK (SP600125) for 60 min and treated with IL-1β or TNFα (10 ng/mL) for 24 h. Secretion of IL-6 and IL-8 was determined by ELISA, respectively, and normalised to cell viability as above. See Methodology below for further details.

Methodology

Reagents and antibodies. Recombinant IL-1β and TNFα were obtained from R&D Systems. All chemicals including inhibitors for p38 MAPK (SB203580), JNK (SP600125), MEK1/2 (PD098059) were obtained from Sigma Aldrich. p38 MAPK inhibitor MW-01-SRM-2-069A (069A) was developed in our laboratory and has been described in detail(5). MK2 inhibitors sc-48 (#204048), sc-38 (#203138), mouse monoclonal antibody against EGFR (#365829), rabbit polyclonal antibody against HuR (#20694), mouse monoclonal antibody against HuR (#5261) and rabbit polyclonal antibody against CRM1 (#5595) were from Santa Cruz. Rabbit monoclonal antibodies against phosphorylated p38 MAPK (#9215), β-tubulin (#2128), phosphorylated CREB (#9197), total CREB (#9198); rabbit polyclonal antibodies against total p38 MAPK (#9212), MK2 (#3042), phosphorylated Hsp27 (#2401), lamin A/C (#2032), total Cdk1 (#9112), phosphorylated Cdk1 (Tyr15) (#9111); mouse monoclonal antibody against total Hsp27 (#2402); the secondary anti-mouse IgG, HRP-linked monoclonal antibody (#7076) and the secondary anti-rabbit IgG, HRP-linked monoclonal antibody (#7074) were obtained from Cell Signaling Technology.pHuR(S202) antibody was kindly provided by Myriam Gorospe (National Institutes of Health, Baltimore, Maryland, USA).

Cell culture. Primary human astrocytes were purified from 14 to 19 week old aborted foetuses, collected after therapeutic termination after obtaining informed consent (approved by the Human Ethics Committee of the University of Sydney (HREC 2013/131) and University of New South Wales (HREC08284) as described (6). U87 and U87-EGFRvIII glioblastoma cells were provided by their laboratory of origin (7) and tested for EGFR mutation and stable expression by Western blotting (Figure 3C) and immunostaining. The cumulative culture length of all cell lines was less than 3 months. GBM cell lines (U87, U87-EGFRvIII) and primary human astrocytes were cultured in DMEM and RPMI-1640 (Invitrogen), respectively, supplemented with 10 % (v/v) foetal bovine serum (FBS, Invitrogen) and antibiotic-antimycotic solution (Gibco) at 37 C and 5 % CO2.

Cytokine expression. Cytokine production was determined by ELISA using BD OptEIA Human IL-6 and IL-8 kits (BD Biosciences), following manufacturer’s instructions.

Western blotting.Cells were harvested in lysis buffer (20 mM Tris-HCl, pH 7.5, 2 mM EDTA, 100 mM NaCl, 5 mM MgCl2, 1% (v/v) Triton X-100, 5 mM NaF, 10% (v/v) glycerol, 0.5% (v/v) 2-mercaptoethanol, 0.1 mM Na3VO4 and protease inhibitors (Sigma Aldrich P8340). After centrifugation at 10.000 x g the protein concentration of the cleared cell lysate was determined using a Bicinchoninic Acid (BCA) Protein Assay Kit (Pierce), following manufacturer’s instructions. Cell lysates were separated by 10 – 12.5 % SDS-PAGE and transferred to Immobilion-P (Millipore) as described (4). Proteins were detected using their specific primary antibodies, respectively, followed by goat anti-rabbit horseradish peroxidase–conjugated secondary antibody or horse anti-mouse horseradish peroxidase–conjugated secondary antibody, and detection using the SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) and the ChemiDoc MP System (Bio-Rad).

HuRdetection by Western blotting. In order to detect HuR monomer and dimer in nuclear, cytosolic fractions or whole cell lysates, subcellular fractions or whole cell lysates were prepared in 4x Bolt Lithium Dodecyl Sulphate (LDS) sample buffer (Life Technologies) without a reducing agent. Subcellular fractions or whole cell lysates were separated by 10 – 12.5 % SDS-PAGE and transferred to Immobilion-P (Millipore) as described (4). Proteins were detected using freshly prepared rabbit polyclonal antibody against HuR (#20694, Santa Cruz, overnight at 4 °C at a 1:500 dilution), followed by goat anti-rabbit horseradish peroxidase–conjugated secondary antibody, and detection using the SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) and the ChemiDoc MP System (Bio-Rad). In some experiments (Figure 3 and Suppl Figure 5), HuR immunoblots were run under the same conditions, but DTT (50 mM) was added to the LDS buffer. The identity of HuR monomer and dimer was verified in lysates from HuR-depleted cells (see HuR knockdown studies below).

Subcellular fractionation.Subcellular fractionation of U87, U87-EGFRvIII and primary astrocytes (1 x 107)was performed using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific) following manufacturer’s instructions. The extracts were stored at -80 ˚C prior to Western blot analysis. The purity of nuclear and cytoplasmic fractions was verified using anti-lamin A/C and β-tubulin antibodies.

HuR knockdown studies. 1 x 106 U87 and U87-EGFRvIII cells were transfected with 75 pmol HuR siRNA targeting human HuR from position 325 to 345 (Invitrogen #4390824, sense 5’-ACUUAUUCGGGAUAAAGUATT-3’, antisense 5’-UACUUUAUCCCGAAUAAGUTT-3’) and Lipofectamine RNAiMAX (Invitrogen #13778) following manufacturer’s protocol. After 72 h, cells were stimulated with IL-1β (10 ng/mL) for further 24 h. Supernatants were collected for IL-6 quantification and cell lysates were subjected to Western blot analysis. Scrambled siRNA (Invitrogen #4390843) served as a negative control.

Reverse transcriptase-polymerase chain reaction (RT-PCR). Total RNA was extracted using the RNeasy Micro Kit (Qiagen) according to manufracturer’ protocol.The integrity of total RNA was verified on agarose gels and 1 μg RNA was reverse transcribed using the Fermentas RevertAidTM First strand cDNA Synthesis Kit following manufacturer’ protocol. RT-PCR to amplify cDNA fragments encoding IL-6 and IL-8 were performed and quantified as described (2, 3).

IL-6 mRNA stability assay. U87 and U87-EGFRvIII (4 x105) were starved for 2 h, pre-treated with MK2 inhibitor sc-48 (10 μM) for 60 min and activated with IL-1β (10 ng/mL) for 4 h. Cells were then washed and incubated with actinomycin D (5 μg/mL) to inhibit further transcription. Cells were harvested after 0, 0.5, 1, 2 and 4 hours of incubation with actinomycin D. IL-6 mRNA was quantified by real-time PCR as described (2). Results are presented as % IL-6 mRNA remaining compared with mRNA expression after 4 h of IL-1β treatment. Decay constants (k) were calculated by nonlinear regression (one phase exponential decay) of the % IL-6 mRNA remaining versus time after actinomycin D (3) using Prism 5 (GraphPad Software, San Diego, CA).

Luciferase activity assays. 3 x 105 U87 and U87-EGFRvIII cells were transfected with 1 μg pGL3 luciferase reporter containing a 651-bp fragment of the human IL-6 gene promoter (pIL-6-luc 651) together with 1μg pSV-β-Galactosidase (Promega) and 4 μL Lipofectamine® 2000 (Invitrogen) following manufacturer’s protocol. After 48 h, cells were lysed in 150 μL reporter lysis buffer (Promega, #E4030). Luciferase activity (Promega, #E4030) was measured at 490 nm and normalized to β-galactosidase activity (Promega, #E2000) as described (1).

SB203580 internalisation.U87 and U87-EGFRvIII (2 x 105) cells were starved for 2 h and fluorescein-labeled SB203580 (10 μM) was added for 30 and 60 min, respectively. Cells were washed twice with PBS and lysed in 100μL lysis buffer (20 mM Tris-HCl, pH 7.5, 2 mM EDTA, 100 mM NaCl, 5 mM MgCl2, 1% (v/v) Triton X-100, 5 mM NaF, 10% (v/v) glycerol, 0.5% (v/v) 2-mercaptoethanol, 0.1 mM Na3VO4 and protease inhibitors). Cells lysates were cleared by centrifugation for 5 min at 10,000 x g and fluorescence intensity of duplicate samples was measured at 382 nm using the Novostar plate reader (BMG Labtech). Fluorescence relative units (FRU) were normalised to total cellular protein.Alternatively, SB203580 internalisation was determined by flow cytometry. 4 x 105 cells were incubated with 10 μM fluorescein-labelled SB203580 for 30 and 60 min as above. Cells were trypsinised and resuspended in PBS with 0.1% (w/v) BSA. Identical cell numbers (1x105) were analysed with the FACSCalibur flow cytometer (Becton Dickinson) using CellQuestPro Software. Data analysis was performed using FlowJo Software.

Immunohistochemistry. Formalin-fixed, paraffin-embedded tissues from 205 brain tumour patients with grade II - IV astrocytomas who underwent surgery (Unit of Neurosurgery, Tampere University Hospital, Tampere, Finland) during 1983 – 2001 were used (approved by Ethical Review Board of Tampere University Hospital, Tampere, Finland, Dnro R07042; National Authority for Medicolegal Affairs of Finland, Dnro 1502/04/046/07). Whole tissue sections of histologically normal brain tissue obtained from the University of Sydney/Neuropathology Tumour and Tissue Bank (approved by University of Sydney HREC 2013/131) served as control (n = 6). Following dewaxing and rehydration, tissue section (thickness, 5 μm; area 0.28 mm2) peroxidase activity was blocked by incubation in 0.5% H2O2 in methanol. Tissue was then treated for 30 min in citrate buffer (pH 6.0) at 121C for antigen retrieval. Activated MK2 (p-MK2) was detected using rabbit polyclonal antibody recognising pT334-MK2 (#3007, Cell Signaling Technology). The EnvisionTM+Dual Link (Dako) was applied for 20 min, followed by Dako liquid DAB+chromogen according to manufacturer’s recommendations. The tissue sections were counterstained with hematoxylin. p-MK2 immunoreactivity in TMA samples was evaluated by two pathologists (JH and MB) using a consultation microscope (Zeiss Axioscope). Immunoreactivity was scored negative (<5% positive cells) or positive (>5% positive cells). The positive samples were assigned score 1 (weak staining), score 2 (moderate staining) or score 3 (strong staining) based on the highest intensity observed in the sample(8). Examples of these intensities are in the Figure 4E.

Statistical Analysis.

ELISAdata analysis was performed using Prism 5 (GraphPad Software, San Diego, CA) and represents the mean ± SEM from at least three independent experiments. Statistical analysis of ELISA data was performed using one-way ANOVA, followed by the Newman-Keuls multiple comparison test (GraphPad Software). P values <0.05 were considered as statistically significant. Immunoblots were analysed using Image J. IC50 values were calculated by non-linear regression analysis from at least three independent experiments (each treatment data point performed in duplicate) with Excel software.

References

1.Quante T, Ng YC, Ramsay EE, Henness S, Allen JC, Parmentier J, et al. Corticosteroids reduce IL-6 in ASM cells via up-regulation of MKP-1. Am J Res Cell Mol Biol 2008; 39: 208-217.

2.Henness S, Johnson CK, Ge Q, Armour CL, Hughes JM, Ammit AJ. IL-17A augments TNF-α–induced IL-6 expression in airway smooth muscle by enhancing mRNA stability. J Allergy Clin Immunol2004;114:958-964.

3.Henness S, van Thoor E, Ge Q, Armour CL, Hughes JM, Ammit AJ. IL-17A acts via p38 MAPK to increase stability of TNF-α-induced IL-8 mRNA in human ASM. Am J Physiol Lung Cell Mol Physiol 2006;290:L1283-L1290.

4.Yeung Y, Bryce N, Adams S, Braidy N, Konayagi M, McDonald K, et al. p38 MAPK inhibitors attenuate pro-inflammatory cytokine production and invasiveness of human U251 glioblastoma cells. J Neurooncol 2012;109:35-44.

5.Munoz L, Ralay Ranaivo H, Roy SM, Hu W, Craft JM, McNamara LK, et al. A novel p38alpha MAPK inhibitor suppresses brain proinflammatory cytokine up-regulation and attenuates synaptic dysfunction and behavioral deficits in an Alzheimer's disease mouse model. J Neuroinflammation 2007;4:21.

6.Guillemin GJ, Smythe G, Takikawa O, Brew BJ. Expression of indoleamine 2,3-dioxygenase and production of quinolinic acid by human microglia, astrocytes, and neurons. Glia 2005;49:15-23.

7.Huang P, Mukasa A, Bonavia R, Flynn R, Brewer Z, Cavenee W, et al. Quantitative analysis of EGFRvIII cellular signaling networks reveals a combinatorial therapeutic strategy for glioblastoma. Proc Natl Acad Sci USA 2007;104:12867-12872.

8.Puustinen P, Junttila MR, Vanhatupa S, Sablina AA, Hector ME, Teittinen K, et al. PME-1 protects extracellular signal-regulated kinase pathway activity from protein phosphatase 2A–mediated inactivation in human malignant glioma. Cancer Res 2009;69:2870-2877.

1