Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha

CLS-08443; No of Pages 13 Cellular Signalling xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Cellular Signalling journal homepage: www.elsevier.com/locate/cellsig

2

Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha

F

1Q1

5

Department of Molecular Biology and Genetics, Aarhus University, 8000 Aarhus C, Denmark

6

a r t i c l e

7 8 9 10 11

Article history: Received 6 November 2014 Received in revised form 14 March 2015 Accepted 29 March 2015 Available online xxxx

12 13 14 15 16 17

Keywords: Folate Receptor Folic acid gp130 Vitamin B9 STAT3

R O

i n f o

a b s t r a c t

C

T

E

D

P

The signal transducer and activator of transcription 3 (STAT3) is a well-described pro-oncogene found constitutively activated in several cancer types. Folates are B vitamins that, when taken up by cells through the Reduced Folate Carrier (RFC), are essential for normal cell growth and replication. Many cancer cells overexpress a glycophosphatidylinositol (GPI)-anchored Folate Receptor α (FRα). The function of FRα in cancer cells is still poorly described, and it has been suggested that transport of folate is not its primary function in these cells. We show here that folic acid and folinic acid can activate STAT3 through FRα in a Janus Kinase (JAK)-dependent manner, and we demonstrate that gp130 functions as a transducing receptor for this signalling. Moreover, folic acid can promote dose dependent cell proliferation in FRα-positive HeLa cells, but not in FRα-negative HEK293 cells. After folic acid treatment of HeLa cells, up-regulation of the STAT3 responsive genes Cyclin A2 and Vascular Endothelial Growth Factor (VEGF) were verified by qRT-PCR. The identification of this FRα-STAT3 signal transduction pathway activated by folic and folinic acid contributes to the understanding of the involvement of folic acid in preventing neural tube defects as well as in tumour growth. Previously, the role of folates in these diseases has been attributed to their roles as one-carbon unit donors following endocytosis into the cell. Our finding that folic acid can activate STAT3 via FRα adds complexity to the established roles of B9 vitamins in cancer and neural tube defects. © 2015 Published by Elsevier Inc.

37 35 34

R

36

1. Introduction

39

1.1. The signal transducer and activator of transcription 3 (STAT3)

40

STAT3 is a pro-oncogene constitutively active in several cancer types, for example ovarian, breast, prostate, lung, renal, colon, gastric and cervical cancers [1,2]. STAT3 contributes to tumourigenesis through stimulation of cell division, angiogenesis, and metastasis as well as

N C O

R

38

Abbreviations: CNTF(R), Ciliary neurotrophic factor (receptor); EGF, Epidermal growth factor; ERK, Extracellular signal-regulated kinase; FRα, Folate receptor alpha; GPI, glycophosphatidylinositol; IFN, Interferon; IL-6(R), Interleukin 6 (receptor); ISG, Interferon stimulated gene; JAK, Janus kinase; LIF(R), Leukaemia inhibitory factor (receptor); RFC, Reduced folate carrier; STAT, Signal transducer and activator of transcription; VEGF, Vascular Endothelial Growth Factor. ⁎ Corresponding author at: Department of Molecular Biology and Genetics, Aarhus University, C. F. Mollers Allé 3, 8000 Aarhus C, Denmark. Tel.: +45 871 55484; fax: +45 8619 6500. E-mail address: [email protected] (P.M. Martensen). 1 Present address: Department of Clinical Biochemistry, Aarhus University Hospital, Norrebrogade 44, 8000 Aarhus C, Denmark. 2 Present address: Department of Animal Science, Aarhus University, 8830 Tjele, Denmark. 3 Present address: Department of Oncology, Aarhus University Hospital, 8000 Aarhus C, Denmark.

U

43

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

E

33

41 42

O

4

Mariann F. Hansen, Eva Greibe 1, Signe Skovbjerg, Sarah Rohde, Anders C.M. Kristensen, Trine R. Jensen, Charlotte Stentoft 2, Karina H. Kjær, Camilla S. Kronborg 3, Pia M. Martensen ⁎

3Q2

inhibition of apoptosis [3]. In addition, STAT3 has a direct inhibitory effect on the transcription of p53, one of the most important defenses against cancer [4]. STAT3 is considered an important contributor to the oncogenic transformation in a number of cancer types and is therefore considered to be a putative molecular target in cancer therapy [1–6]. STAT3 is activated upon ligand binding to specific receptors, including receptors for interferon (IFN) and epidermal growth factor (EGF), as well as upon activation of gp130, the common signal transducing receptor for interleukin 6 (IL-6) type cytokines, such as IL-6, IL-11, leukaemia inhibitory factor (LIF), oncostatin M, ciliary neurotrophic factor (CNTF), and cardiotrophin [7,8]. STAT3 associates intracellularly with the receptors, and is phosphorylated on Tyr705 by one of the Janus Kinases (JAKs). Subsequently, STAT3 dimerises, migrates to the nucleus and binds to SIE/GAS elements in the promoters of target genes thus activating transcription. The proteins encoded by these STAT3 activated genes can then lead to the carcinogenic effects observed in cancer cells [2,9].

44

1.2. Folic acid, folate and folate receptors

61

45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Folate, also called vitamin B9, acts as a one-carbon unit donor in im- 62 portant metabolic pathways including the biosynthesis of nucleotides 63 and Methionine, and is thus highly essential to the organism. The 64

http://dx.doi.org/10.1016/j.cellsig.2015.03.020 0898-6568/© 2015 Published by Elsevier Inc.

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

118

HeLa (human epithelial cervical carcinoma cells) (ATCC, CCL-2), and HEK293 (human embryonic kidney cells) (ATCC, CRL-1573) cell lines were cultured in DMEM or RPMI medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. The FRα cDNA was obtained from the image clone (IRAUp969F0432D/IMAGE ID:3956659). The FRα cDNA was excised from the plasmid pOTB7 with EcoR1 and Xho1 restriction enzymes, and ligated into the same restriction sites in the pcDNA3 expression vector (Invitrogen). The FRα-pcDNA3 and the control pcDNA3 plasmids were linearized with BglII and transfected into HeLa cells. Geneticin at

90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114

119 120 121 122 123 124 125 126 127

C

88 89

E

86 87

R

84 85

R

82 83

O

80 81

C

78 79

N

76 77

U

74 75

2.2. Immunoblotting analysis

158

Cell extracts were subjected to 10% SDS-PAGE, and the proteins transferred to PVDF membranes by electroblotting. In the case of FRα analysis, the proteins were kept under non-reduced conditions. The immunoreactions on the PVDF membranes were carried out in Trisbuffered saline (TBS) with 0.1% Tween-20 and 5% non-fat dry milk as blocking, and TBS with 0.1% Tween-20 and 5% BSA as antibody dilution buffer. The following primary antibodies from Santa Cruz were used at dilutions 1:1000: STAT3 STAT1 and phosphoSTAT1 (Tyr701), and the anti-gp130 antibody was used at dilution 1:5000. From Cell Signalling two antibodies were used against phosphoSTAT3 (Tyr705) #9131 (1:1000) and #9145 (XpSTAT3) (1:5000). Also from Cell Signalling antibodies against pERK1/2 and ERK1/2 were both used at 1:5000. The Mov18/ZEL (Alexis Biomedical) was used 1:2000. The incubations with the primary antibodies were carried out overnight at 4 °C. The secondary antibodies were HRP-conjugated goat-anti-rabbit (DakoCytomation) (1:5000), goat-anti-mouse (Amersham Biosciences) (1:5000) or anti-rabbit (#7074S from Cell Signalling, 1:1000). Proteins were visualised by ECL plus Western Blotting Detection System (GE Healthcare), and quantified using Adobe Photoshop, version 7.0 (Adobe Systems; Mountain View, CA.) For quantification the scanned images were inverted before the protein bands were selected. The intensity of the bands in arbitrary units was then found by multiplying the mean intensity with the number of pixels.

159

2.3. STAT3 transcription factor assay

182

The STAT3 Transcription Factor Assay (Chemicon) was performed according to the manufactures' instructions. During the assay, the Capture Probe, a double stranded biotinylated oligonucleotide containing the STAT consensus sequences for STAT binding (5′-TTC CCG TAA-3′ (hSIE67) and 5′-TTC CGG GAA-3′ (ICAM-pIRE)) was mixed with the nuclear extract. The mixture was transferred to a streptavidin-coated micro plate immobilizing the biotinylated Capture probe. Bound

183 184

F

2.1. Cell culture

72 73

O

117

71

R O

2. Materials and methods

69 70

128 129

P

116

67 68

500 μg/ml was used to select HeLa cells with integrated plasmid, and resistant HeLa cells were collected in bulk thus obtaining the HeLa-FRα cell line overexpressing FRα, and the control HeLa-pcDNA3 cell line. Overexpression of FRα was verified by immunoblot analysis using the Mov18/Zel antibody (Alexis Biochemical). To reduce the level of active components, FBS was incubated with 2% active charcoal overnight (Sigma). Cells pre-treated for three to four hours with folate-free RPMI medium containing 10% charcoal stripped FBS and 1% P/S were treated for 5–60 min with 0–800 U/ml IL-6 (Sigma), 0–1000 U/ml IFNα or IFNβ (Schering-Plough), 0–600 μg/ml folic acid (Pteroylglutamic acid) (Sigma), or 250 μg/ml folinic acid (N5-formyl-5,6,7,8-tetrahydropteroyl-L-glutamic acid) (Sigma) in folate-free RPMI medium (Gibco). For immunoblotting analysis, cells were lysed in 62.5 mM Tris–HCl (pH 7.5), 2% SDS, 10% glycerol with 1 μl/ml Protease Inhibitor Cocktail P8340 (Sigma), 1 μl/ml phosphatase inhibitor cocktail II (Sigma), and 5 μl/ml Benzonase (Merck) overnight at 4 °C. The protein extracts were centrifuged at 13.000 rpm for 5 min and the pellet discarded. For nuclear extractions, the NE-PER Nuclear and Cytoplasmic Extraction reagent (Pierce) was used, and protein concentration was determined by Bio-Rad Protein Assay (Bio-Rad). For the JAK-inhibitor experiments, cells were pre-treated for 30 min with 100 μM AG490 (α-Cyano-(3,4-dihydroxy)-N-benzylcinnamide) (Calbiochem) prior to treatment. For the receptor blocking experiments, cells were pre-treated for 30 min with 3 μM AG1478 (Calbiochem) to inactivate the EGF receptor, and as control treated with 25 ng/ml epidermal growth factor (EGF) (Sigma). Furthermore, cells were pre-treated for 60 min with 8 μg/ml anti-gp130 monoclonal antibody clone 28126 (R&D systems) or for 60 min with 16 μg/ml anti-FRα antibodies: Mov18/ZEL from Alexis Biochemical, and BGN64, Sf99 and Sf55 from AbD SeroTec.

T

115

Reduced Folate Carrier (RFC), present on all cell surfaces, mediates uptake of folates from food such as vegetables and liver [10]. In contrast, most healthy tissues either do not express any or express only negligible levels of the Folate Receptor α (FRα) [11–13], with the exception of kidneys and a few fast growing cell types [14]. In addition, FRα is highly expressed on the surface of a number of cancer cells including ovarian, endometrial, pancreas, breast and cervical cancers [11,14]. In general, aggressive cancer types express the highest amounts of FRα [11,15,16]. While RFC has a high affinity for reduced folates and a relatively poor affinity for the non-reduced synthetic form folic acid, FRα has a high affinity for both forms, although with a slightly higher affinity for the nonreduced folic acid [14,17]. The presence of highly up-regulated FRα on a number of cancer cell types, combined with the high affinity for folic acid, has made FRα a possible target for cancer specific chemo and immune therapy, as well as for folate-targeted imaging [18–20]. FRα has been associated with internalizing folate via endocytosis [14]. The FRs are endocytosed via a cell division cycle 42 proteindependent mechanism [21], and folates are released from mildly acidic endosomes, as very elegantly shown by Yan et al. 2007 [22]. FRα is specifically involved in renal tubular reabsorption of folate by endocytosis [23], and it has been shown that the megalin and cubilin endocytosis receptors expressed in the kidneys are most likely involved in this mechanism [24,25]. In the choroid plexus, FRα has been found to mediate the transport of folates into the cerebrospinal fluid [26]. Other results indicate that transport of folate is not the primary function of FRα in ovarian cancer cells, since folate uptake in these cells is primarily mediated by RFC, in spite of highly up-regulated FRα [27]. Also, the uptake of the folate antagonist pemetrexed is not dependent of FRα in HeLa cells [18,19,28,29]. The development of folate antagonists as cancer drugs has been targeted towards RFC and another wellcharacterized folate transporter, the proton-coupled folate transporter [30,31]. The recent structural determination of FRα bound to folic acid in combination with mutational analyses revealed that in particular the amino acid residue D81 is required for folic acid binding to FRα [32]. Moreover, the combined structural studies of FRα and FRβ revealed a distinct conformational change at pH 6.5 involving the formation of an inhibitory loop covering the ligand binding site, thereby explaining the 2000-fold decrease in affinity of folic acid for FRα at this pH [33]. This gives a plausible explanation for how FRα bound folate is released in acidified endosomes. To assess whether FRα could be involved in a signal transduction pathway affecting cell proliferation, we initially investigated the effect of folic acid (non-reduced form) and folinic acid (reduced form) on the activation of the STAT3 oncogene in FRα-positive HeLa and FRαnegative HEK293 cells. Furthermore, we investigated whether folic acid could promote cell proliferation in these two cell lines, and verified the up-regulation of two STAT3 regulated genes in folic acid treated HeLa cells. Finally, we identified co-receptors and signalling molecules involved in the folic acid initiated FRα-STAT3 signalling transduction pathway.

D

65 66

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

E

2

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157

160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181

185 186 187 188 189

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

2.8. Flow cytometry

251

195

2.4. Cell proliferation assay

252

196

2.9. RT-PCR

258

205

HeLa cells and HEK293 cells were seeded at 20% confluence in 96well microplates, starved for folic acid and then treated for 48 h with different concentrations of folic acid. The number of proliferating cells was estimated by the use of Delfia® Cell Proliferation Kit (PerkinElmer) according to manufactures' instructions. The cells were labelled for two hours with 5-Bromo-2′-deoxyUridine (BrdU), followed by incubation with Eu-labelled anti-BrdU antibody, which was subsequently detected by a Fluostar Optima multiplate reader (BMG Labtech). The fluorescence measured is proportional to the DNA synthesis in the cell population of each well.

Cells were harvested by trypsination and incubated for 30 min with the anti-FRα antibody Mov18/ZEL (1:500). As a staining and washing buffer, PBS with 1% BSA was used. After incubation with the secondary FITC-conjugated Anti-mouse IgG (Sigma) (1:200) for 20 min, the cells were fixed in 2% paraformaldehyde and Flow Cytometry was performed using a FACS Calibur Analyzer (Becton Dickinson).

206

2.5. siRNA assays

207

218

HeLa cells in 12-well dishes were transiently transfected twice with 30 pmol FRα duplex oligo siRNAs (a pool of three siRNAs (target FRα sequences): 5′-GAA GAA UGC CUG CUG UUC U-3′; 5′-GCA AUG GUG GGA AGA UUG U-3′; 5′-CCA CUG UUC UGU GCA AUG A-3′ from Santa Cruz Biotechnology), 30 pmoles gp130 duplex oligo from Santa Cruz Biotechnology, or 30 pmoles GFP duplex oligo non-target siRNA (Dharmacon), with 24 hour interval using Lipofectamine™ 2000 (Invitrogen). After additional 24 h, the cells were starved for folic acid for 4 h, treated for 10 min with IL-6, folic acid, or left untreated. Statistical analysis was performed on quantified data from three biologically independent experiments using paired t-tests. Statistical analysis was performed using GraphPad Prism Software.

219

2.6. Co-immunoprecipitation assay

220 221

228

HeLa cells at 80% confluence in p10 dishes were lysed in 1 ml buffer (1% IGEPAL CA-630, 20 mM Tris-HCl2, 150 mM NaCl, 10 mM EDTA, 2 mM EGTA, 10% glycerol, pH 7.4) for 30 min on ice. Samples were sonicated for 5 s before immunoprecipitation with the Mov18/Zel antibody using Dynabeads protein G immunoprecipitation kit (Invitrogen) according to the manufacturer's recommendations. Samples were eluted in two subsequent steps, first with elution buffer alone (native condition) or heated for 5 min at 95 °C with elution buffer added SDS sample buffer without DTT (denaturing condition).

229

2.7. Immunostaining

230

Cells were grown on cover slips and fixed with 4% paraformaldehyde (Bie and Berntsen) for 10 min at room temperature. The coverslips were blocked in 0.1% Tween-20 in phosphate-buffered saline (PBS) with 3% BSA. For experiments shown in Fig. S1 the cells were incubated for one hour with the anti-FRα antibody Mov18/ZEL (1:200). Cy3 conjugated anti-mouse IgG (Sigma) (1:250) was used as secondary antibody together with nuclear DAPI (Invitrogen) (1:1000) staining. Cells were visualised with a Leica DMR fluorescent microscope and photographed with a Leica DC200 CCD camera. The pictures were acquired and processed by Adobe Photoshop. For the co-localization experiments shown in Fig. 7, the anti-gp130 antibody (C20, Santa Cruz) was used (1:100) together with Mov18/Zel (1:500). As secondary antibodies FITCconjugated anti-mouse IgG (Sigma) and Cy3-conjugated anti-rabbit IgG (Sigma) were used (both 1:250). The coverslips were mounted with ProLong Gold with DAPI (Molecular Probes, Invitrogen) for visualization of the nuclei. Regular wide field epifluorescence microscopy was performed on a Zeiss Axiovert 200 M Inverted Fluorescence Microscope equipped with a 63x oil-immersion lens (NA 1.4) and an HBO lamp.

214 215 216 217

222 223 224 225 226 227

231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247

255 256 257

2.10. Quantitative RT-PCR

269

P

R O

O

259 260

D

212 213

253 254

mRNA was purified from 1 × 104 cells using Dynabeads® mRNA DIRECT™ Micro Kit (Dynal Biotech) and cDNA was prepared using MMLV reverse transcriptase (Epicentre). The PCR primers used were FRα forward, 5′-ATG GCT CAG CGG ATG ACA AC-3′, FRα reverse, 5′TCA GCT GAG CAG CCA CAG CA-3′, GAPDH forward, 5′-GGT CGG AGT CAA CGG ATT T-3′ and GAPDH reverse, 5′-CCA GCA TCG CCC CAC TTG G-3′. The PCRs were carried out in a Peltier Thermal Cycler from Merck Eurolab with 40 cycles of denaturation at 95 °C for 10 min for the first cycle and then 30 s for the following cycles, annealing at 62 °C for 30 s, and elongation at 72 °C for 30 s.

261 262 263 264 265 266 267 268

HeLa cells were grown to 80% confluence in 100 mm culture dishes before treatment with 250 μg/ml folic acid for 24 h in folate-free RPMImedia, supplemented with antibiotics and active charcoal treated FBS. Cells were then washed in PBS and harvested by scraping, and 3 × 106 cells were pelleted by centrifugation for 5 min at 1700 rpm. RNA was purified using the RNeasy Mini Kit (Qiagen) according to the manufacturer's instructions. Total RNA purified from three independent experiments was pooled to reduce variations between the experiments. 1500 ng RNA of these pooled RNA samples were used for cDNA synthesis using M-MLV reverse transcriptase (Invitrogen) with Oligo dT primers according to the manufacturer's recommendations. mRNA expression was measured using Platinium SYBR Green q-PCR SuperMixudg (Invitrogen) with 0.2% ROX Reference Dye (Invitrogen) according to the manufacturer's recommendations. The primer sequences were: Cyclin A2: Forward primer: 5′-AGC TGG CCT GAA TCA TTA-3′, Reverse primer: 5′-TGG CCA CAA CTT CTG TAT-3′. VEGF: Forward primer: 5′GGA GTA CCC TGA TGA GAT-3′ and Reverse primer: 5′-GGT CTG CAT TCA CAT TTG TT-3′ and GAPDH: Forward primer: 5′-GGT CGG AGT CAA CGG ATT T-3′ and Reverse: 5′-CCA GCA TCG CCC CAC TTG-3′. Cyclin A2 and VEGF expression were normalized to GAPDH expression.

270 271

3. Results

290

E

210 211

T

208 209

C

203 204

E

201 202

R

199 200

R

197 198

N C O

192 193

F

Images were collected with a Photometric CoolSnapTM HQ camera 248 from Roper Scientific controlled by MetaMorph acquisition software. 249 The images were analysed with ImageJ [34] and Adobe Photoshop. 250

194

STAT3 transcription factor subunits were detected with specific STAT3 antibodies, and a HRP secondary antibody was used for detection in a spectrophotometric plate reader. Results from a biotinylated nonspecific double stranded oligonucleotide used as a negative control was subtracted.

U

190 191

3

272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289

3.1. STAT3 activation by folic acid and folinic acid in HeLa and HEK293 cells 291 We have previously observed that folic acid was able to induce the transcription of ISG12A (interferon stimulated gene of 12 kDa) in HeLa cells after 24 h of folic acid treatment [35]. We hypothesized that folic acid binding to the glycophosphatidylinositol (GPI)-anchored FRα could be involved in this up-regulation through STAT3 activation. First, the presence of FRα on the HeLa cell surface was found by flow cytometry and immunofluorescence microscopy (Figs. 1A and S1A, and verified in Fig. 7B). We then searched for FRα negative cell lines and found that HEK293 cells did not express FRα on the cell surface (Figs. 1A and S1A). These two cell lines were stimulated with folic acid and folinic acid for 15 min and the phosphorylation of STAT3 on Tyr705 investigated (Fig. 1B). In the FRα positive HeLa cells, both folic acid and folinic acid treatment induced phosphorylation of STAT3,

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

292 293 294 295 296 297 298 299 300 301 302 303 304

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

U

N

C

O

R

R

E

C

T

E

D

P

R O

O

F

4

Fig. 1. STAT3 activation by folic acid and folinic acid in HeLa and HEK293 cells. A. Cells were incubated with (black lines and filled area) or without (grey lines and clear area) the anti-FRα antibody (Mov18/ZEL). The cells were subjected to flow cytometry showing the expression of FRα on the surface of HeLa and not on HEK293 cells. The experiment was performed three times. B. Immunoblot analysis of protein extracts from HeLa and HEK293 cells treated for 15 min with 800 U/ml IL-6, 250 μg/ml folic acid or folinic acid or left untreated (control) using an antibody recognizing STAT3 phosphorylated on tyrosine Y705 (pSTAT3(Y705)). Immunoblot analysis with the anti-STAT3 antibody was used as a loading control. Protein bands from biologically independent experiments were quantified, and pSTAT3 levels relative to STAT3 levels were calculated. Untreated controls were set to 1, and fold induction (FI) indicated below the blots. SD: standard deviation of the repeated measurements, HeLa (n = 2) and HEK293 (n = 3). C. Binding of active pSTAT3-dimers to the STAT3 DNA consensus sequences in nuclear extracts from HeLa cells treated with IL-6, folic acid, folinic acid or left untreated (control) as in B. Data are represented as means ± SEM of OD450 values from triplets. Statistical analysis was performed using unpaired t-tests. *: P b 0.03 (treament vs. control), and **: P b 0.006 (treatment vs. control). D. Immunoblot analysis with the anti-pSTAT3(Y705) antibody of protein extracts from HeLa cells transiently transfected with non-target EGFP-siRNA (−) or FRα-siRNA (+) as indicated, followed by treatment with IL-6, folic acid and folinic acid, or left untreated (control) as in B. Immunoblot analysis with the anti-STAT3 antibody was used as a loading control. Protein bands from the presented blot were quantified, and pSTAT3 levels relative to STAT3 levels were calculated. Untreated control was set to 1, and fold induction (FI) indicated below the blots. E. Protein bands from biologically independent experiments shown in Fig. 7B were quantified, and pSTAT3 levels relative to STAT3 levels are depicted including siRNA treatments with non-target control (con) EGFP-siRNA and FRα-siRNA (FRα). The relative pSTAT3/ STAT3 level in untreated cells transfected with the control EGFP siRNA was set to 1. Error bars represent SEM from n = 3. Statistical analysis was performed using paired t-tests. *: P b 0.05; ns = not significantly different.

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350

F

O

To further clarify the pathway for the activation of STAT3, we examined the effect of the JAK-inhibitor AG490 on the activation of STAT3 by folic acid and folinic acid (Fig. 4A). As STAT3-activation was reduced in the presence of AG490, we concluded that the activation of STAT3 by folic acid and folinic acid is dependent on JAK. JAK is one of the mediators of STAT3 activation by IL-6 [1], as well as by EGF [37]. The high activation of STAT3 by IL-6 in this assay may have caused saturation of the signal and inaccurate quantifications, and this may explain the no apparent effect of AG490 treatment on reduction of IL-6 induced STAT3 activation. The STAT3 activation by folinic acid was low in this experiment, and it is reduced by the addition of AG490, although not significantly. Upon ligand binding, type 1 IFN receptors can activate STAT3 as well as STAT1 through activation of JAKs [38], thus we investigated the effect of folic acid and folinic acid on STAT1 activation (Fig. 4B). Treatment with folic acid or folinic acid was unable to stimulate STAT1 Tyrosine phosphorylation (Y701) in HeLa cells. IFNα stimulated cells were used as a positive control for STAT1-activation, and the lack of pSTAT1 in the non-treated control cells suggested that STAT1 is not constitutively active in HeLa cells, which confirmed earlier studies [39].

R O

324 325

354

P

322 323

STAT3 activation leads to multiple cellular responses. Fig. 3 shows that folic acid could promote cell proliferation in HeLa cells in a dosedependent manner, whereas proliferation of the FRα-negative HEK293 cells was independent of folic acid concentrations.

D

320 321

353

E

318 319

3.2. Folic acid stimulates cell proliferation of HeLa but not HEK293 cells

Fig. 2. Up-regulation of STAT3 regulated genes. qRT-PCR analysis of Cyclin A2 and VEGF expression in folic acid treated (250 μg/ml) and untreated HeLa cells. RNA isolated from HeLa cells from three independent experiments was pooled before cDNA synthesis, and qRT-PCR was performed in triplicates. Expression levels were normalized to GAPDH. Error bars represent SD. Both genes were significantly up-regulated by folic acid treatment (*: P b 0.01, unpaired t-test).

355 356 357

3.3. Folic acid and folinic acid activate STAT3 through a JAK-dependent and 358 a STAT1 independent mechanism 359

T

316 317

C

314 315

E

312 313

R

311

R

309 310

inducible genes Cyclin A2 (CycA2) and Vascular Endothelial Growth 351 Factor (VEGF) in HeLa cells analysed by quantitative RT-PCR (Fig. 2). 352

360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379

3.4. Time course and dose responses of folic acid induced STAT3 activation 380 In order to investigate the signal transduction pathway more thoroughly, we performed time course and dose response experiments for folic acid induced STAT3 activation in HeLa cells. First, we determined the optimal folate starvation time to be between two and four hours, as stimulation with folic acid after this starvation period had the

N C O

307 308

whereas no STAT3 activation was detected in FRα negative HEK293 cells. Cell extracts, non-treated or treated with IL-6, were used as negative and positive controls, respectively. The observed pSTAT3 background levels in control cells throughout our studies correlated with the presence of constitutively active STAT3 previously observed in HeLa cells [36]. In addition, the observed IL-6 mediated STAT3 activation in HEK293 cells shows that the STAT3 activation mechanism in these cells is functional. In these initial experiments very high doses of IL-6 were used, hence the pSTAT3 levels in the IL-6 treated cells were very high leading to inaccurate quantifications (e.g. high SD for IL-6 induced STAT3 activation in HEK293 cells (Fig. 1B)). We then investigated whether STAT3, when activated by folic acid and folinic acid in HeLa cells, could bind to STAT3 consensus DNA sequences. Fig. 1C shows that increased levels of pSTAT3 indeed bound the SIE/GAS DNA enhancer elements after folic acid or folinic acid treatment. Even though the effect of the folates on STAT3 activation is lower than that of IL-6, the activity is significantly higher than that of the non-treated controls. Moreover, the relative levels of pSTAT3 binding observed in Fig. 1C, correlated with the corresponding levels detected in the pSTAT3 immunoblotting analyses in Fig. 1B. To further substantiate the involvement of FRα in the activation of STAT3 by folic acid and folinic acid, HeLa cells were transfected with siRNA against FRα prior to treatment with folic acid and folinic acid. The knock-down of FRα was verified by immunofluorescence staining of the cells and RT-PCR analysis (Fig. S1B and C) as well as immunoblotting analysis (Fig. 7C). Knock-down of FRα was able to diminish folic acid and folinic acid mediated STAT3 activation, thereby showing a direct involvement of FRα in this signalling pathway (Fig. 1D). Quantification and statistical analyses of the experiments shown in Fig. 7 are depicted in Fig. 1E for repeated FRα knock-down experiments. It can be concluded from this experiment that knock-down of FRα decreases the folic acid induced activation of STAT3 significantly. The very high background level of pSTAT3 in the non-treated control cells (seen in Fig. 7) may be caused by the presence of dsRNA from the siRNA treatment, or general stress from the two successive transfections, which might have activated cellular defence mechanisms, leading to the activation of STAT3. To confirm the role of FRα in folic acid and folinic acid mediated STAT3 activation, we transiently expressed FRα in HEK293 cells to investigate if this would enable STAT3 activation in these cells following folic acid and folinic acid treatment. Unfortunately, no cell surface FRα expression was detected in HEK293 cells despite the presence of FRα protein in cell extracts, suggesting incomplete processing and cellular targeting of FRα in these cells (data not shown). In agreement with folic acid induced STAT3 activation in HeLa cells, we observed that folic acid increased transcription of the STAT3

U

305 306

5

Fig. 3. Folic acid stimulates cell proliferation of HeLa but not HEK293 cells. Cell proliferation of HeLa and HEK293 cells treated for 48 h with or without the indicated amounts of folic acid. The number of cells were measured by 5-Bromo-2′-deoxyuridine (BrdU) incorporation for two hours and then harvested. The number of folic acid treated cells is depicted as relative to the number of untreated cells. The experiment was done in 6 (HEK293) or 24 replicas (HeLa) for each concentration. Error bars represent SD. The folic acid correlation was significant * (P = 0.01) for HeLa cells and not significant (ns) for HEK293 cells. Statistical analysis of individual points was performed using Two-way Anova. ****: P b 0.0001.

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

381 382 383 384 385

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

E

D

P

R O

O

F

6

386 387

R

E

C

T

Fig. 4. Folic acid and folinic acid activate STAT3 through a JAK-dependent and a STAT1 independent mechanism. A. Immunoblot analysis of protein extracts from HeLa cells treated with (+) or without (−) 100 μM JAK-inhibitor AG490 prior to treatment with 800 U/ml IL-6, 250 μg/ml folic acid, 250 μg/ml folinic acid or left untreated (control) using the anti-pSTAT3(Y705) antibody. Immunoblot analysis with the anti-STAT3 antibody was used as a loading control. B. Immunoblot analysis of protein extracts from HeLa cells treated with 1000 U/ml IFNα, 250 μg/ml folic acid, 250 μg/ml folinic acid, or left untreated (control) using an antibody recognizing STAT1 phosphorylated on Tyrosine 701 (pSTAT1(Y701)). Immunoblot analysis with the anti-STAT1 antibody was used as a loading control. Protein bands from biologically independent experiments were quantified, and pSTAT3 or pSTAT1 levels relative to STAT3/ or STAT1 levels were calculated and depicted below the blots. Untreated controls were set to 1. Error bars represent SEM from n = 3 in A and n = 2 in B. Statistical analyses in A were performed using paired t-tests. *: P b 0.04, ns = not significantly different.

398

3.5. Co-receptors for folic and folinic acid induced STAT3 activation

399

Since FRα is attached extracellularly to the cell by a GPI-anchor, we investigated the mechanism behind FRα mediated STAT3-activation upon stimulation by folic acid and folinic acid. Based on the knowledge of traditional receptor-mediated STAT3 activation mechanisms, it seemed most likely that FRα upon extracellular binding of folic acid and folinic acid should activate a transmembrane co-receptor on the cell surface. Therefore, we wanted to explore which co-receptors could be responsible for this signal transduction pathway. The common signal transducing receptor for IL-6 type cytokines called gp130, the EGF receptor, and the type 1 IFN receptor are known STAT3 activating transmembrane receptors. As folic acid and folinic acid lacked the ability to activate STAT1 (Fig. 5B), and the time course of STAT3 activation after IFNβ treatment was

394 395

400 401 402 403 404 405 406 407 408 409 410

O

C

392 393

N

390 391

U

388 389

R

396 397

strongest relative effect (Fig. S2). The time course of folic acid induced STAT3 activation (Fig. 5A) showed a pattern similar to IL-6 induced STAT3 activation, i.e. maximum at 10 min and a decline to control levels after 45 to 60 min. This pattern was different from the IFNβ induction, which showed an increasing STAT3 activation up to 20 min of treatment, and this level remained high throughout the 60 min of treatment. Thus the folic acid induced STAT3 activation resembles IL-6 induction more than IFNβ induction. The dose response of folic acid induced STAT3 activation showed a significant dose dependent increase in STAT3 activation (Fig. 5B). As expected, there were clear dose responses of IL-6 and IFNβ induced STAT3 activation in HeLa cells.

different from the IL-6 and folic acid induced STAT3 activation, the type 1 IFN receptor was excluded from further analysis. To investigate the role of the EGF receptor, we pre-treated HeLa cells with the EGF receptor inhibitor AG1478 prior to addition of folic acid and folinic acid. The results show that the EGF receptor was not involved, since addition of AG1478 did not influence STAT3 activation induced by either folic acid or folinic acid (Fig. 6A). In contrast, STAT3 activation by EGF was diminished by pretreating HeLa cells with AG1478 (Fig. 6B). The AG1478 inhibited the EGF induced autophosphorylation of the 170 kDa EGF receptor, which was recognized by this anti-pSTAT3 antibody (as indicated in the data sheet for the phospho-STAT3 (Tyr705) antibody #9131 from Cell Signalling used in this experiment). We thus concluded that the EGF receptor was not the co-receptor for folic acid induced FRα-STAT3 signalling. Pretreatment of HeLa cells with the anti-FRα Mov18 antibody surprisingly increased the STAT3 activation even without addition of folic acid (Fig. 6C). Addition of either folic acid or folinic acid together with the Mov18 antibody seemed to increase the STAT3 activation, although not significantly. In order to analyse this phenomenon further, HeLa cells were pretreated with various antibodies against FRα (Fig. 6D). Increased STAT3 activation was seen with most of these antibodies without addition of folic acid. One anti-FRα antibody, BGN64 did not exhibit this STAT3 activation property, showing that STAT3 activation is not a general capacity of all anti-FRα antibodies, and thus indicating that antibodies binding to specific epitopes on FRα may be required to facilitate STAT3 activation.

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

Since the level of folic acid used in our experiments was higher than physiological serum concentrations, we wanted to investigate whether the level of FRα on the cell surface of HeLa cells was a limiting factor. Thus, we constructed a HeLa cell line overexpressing FRα (HeLa-FRα) and analysed whether this cell line could activate STAT3 at lower folic

458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492

498 499 500

C

456 457

E

454 455

R

452 453

R

450 451

N C O

448 449

U

446 447

4. Discussion

509

4.1. gp130 as a co-receptor for folic acid induced FRα-STAT3 signalling

510

We have depicted a new signal transduction pathway initiated by binding of folic acid or folinic acid to the FRα and eliciting a cellular response through the co-receptor gp130 followed by activation of JAK and STAT3. Knocking down the FRα reduced the folic acid induced STAT3 activation. The fact that both STAT3 and ERK1/2 can be activated by folic acid supports the role of gp130 as a co-receptor because this receptor is capable of activating both pathways. Together with the coimmunoprecipitation of FRα and gp130, our results allowed us to establish that the common gp130 receptor acts as a transducing co-receptor for the folic acid induced FRα-STAT3 signalling pathway. Blocking the gp130 receptor with the monoclonal 28126 antibody inhibited the STAT3 activation by folinic acid, whereas the activation by folic acid seemed to be increased. This difference may be explained by sterical hindrance due to the lower affinity of FRα for the larger reduced form folinic acid, compared with that of the small non-reduced form folic acid [14]. Binding of anti-gp130 antibody to gp130 may potentially inhibit the formation of the FRα-folinic acid complex, due to the larger size and lower affinity of folinic acid, thus reducing STAT3 activation, whereas the smaller size and higher affinity of folic acid enables it to associate with FRα despite the binding of antibody to the gp130 receptor chain. This hypothesis can be supported by the recent structural models of FRα and FRβ with and without folic acid [32,33]. The amino acid residues H135 and R103 in FRα are important for the ligand binding. However, when the 5-methyl-tetrahydrofolate is modelled in the structure, little extra room for the methyl group is observed in the binding pocket compared with the model made with folic acid [32]. The same residues, numbered H157 and R125 in FRα and H151 and R119 in FRβ were identified in the folic acid binding pocket by Wibowo et al. (2013), and in particular H157 was suggested to undergo a rotamer shift to interact with folic acid at neutral pH [33]. In our experiments we used the 5-formyl-tetrahydrofolate (folinic acid), which has a side chain in the same position as 5-methyltetrahydrofolate. The binding of the anti-gp130 antibody might distort the binding pocket of FRα in a way that allows folic acid binding and blocking of folinic acid binding. However, further in-depth investigations are required to fully explain our results. The observation that STAT3, but not STAT1, was activated by folic acid and folinic acid, not only correlates with the two STATs having opposing actions [41], but also shows that STAT3 activation by folic acid and folinic acid is not mediated by the IFN induced STAT1-dependent JAK-STAT pathway. Blocking the EGF receptor did not inhibit folic acid induced STAT3 activation, demonstrating that this receptor was not involved in the signalling transduction pathway.

511 512

4.2. STAT3 activation by anti-FRα antibodies

554

Some anti-FRα antibodies were able to induce STAT3 activation independent of folic acid treatment; however, the presence of folic acid increased the level of STAT3 activation. Other studies have shown that specific anti-gp130 monoclonal antibodies are able to induce STAT3 activation [42]. In this particular study, they were able to demonstrate that specific anti-IL6-receptor monoclonal antibodies were able to induce gp130 dimerisation on the plasma membrane of XG-2 cells; however,

555 556

F

496 497

444 445

O

3.7. Folic acid induced STAT3 activation in FRα overexpressing HeLa cells

443

R O

495

441 442

P

493 494

The involvement of the gp130 transmembrane receptor in the signalling was then investigated. Initially, pre-treatment of HeLa cells with blocking anti-gp130 monoclonal antibody reduced STAT3 activation by folinic acid, and surprisingly increased the STAT3 activation by folic acid in the presence of the same antibody although not significantly (Fig. 7A). A significant reduction in STAT3 activation was observed for IL-6 and anti-gp130 antibody treated HeLa cells, compared with IL-6 treated without anti-gp130 antibody. This complies with IL-6 mediating the activation of STAT3 through the gp130 receptor [7,8] and validates the effectiveness of blocking this pathway with anti-gp130 antibody. To further substantiate the role of gp130 as a co-receptor for folic acid induced STAT3 activation, siRNA knock-down of gp130 was carried out (Fig. 7B). A clear knock-down of FRα was obtained, whereas a residual amount of gp130 was seen (Fig. 7C). The transfection procedure increased the background level of activated STAT3 substantially in control cells, and knock-down of gp130 reduced the activated STAT3 levels considerably in control cells (Fig. 7B). After quantification of repeated experiments it was observed that folic acid significantly increased STAT3 activation in control and FRα knock down cells, whereas this effect was diminished when gp130 was knocked down alone or in combination with FRα. The IL-6 induced STAT3 activation was not found to be significantly different from the untreated controls in the siRNA treated cells. As a result, it was not possible to significantly evaluate the effect of knockdown of gp130 and FRα on IL-6 induced STAT3 activation. However, a tendency towards a decrease in IL-6 induced STAT3 activation was seen when gp130 was knocked down alone or in combination with FRα. Of note is that siRNA treatment against FRα resulted in a significant increase in STAT3 activation following IL-6 treatment compared with the untreated cells. In conclusion, folic acid induced STAT3 activation is significantly reduced when gp130 is knocked down. To finally establish the role of gp130 as a co-receptor for folic acid signalling, immunoprecipitation studies of FRα and gp130 were performed (Fig. 8A). Here, we found that gp130 indeed coimmunoprecipitated with FRα. Two successive elutions of FRα from the immunoprecipitate were performed, the second being more harsh causing the elution of the FRα antibody. Although a proportion of FRα could be eluted in the first elution, the remaining FRα required the release of FRα antibody in order to be released (top panel). When investigating for the presence of gp130 in the immunoprecipitate (lower panel), gp130 could be detected in the second but not in the first elution, suggesting that gp130 may stabilize FRα binding to the FRα antibody. A fraction of the present gp130 protein was found not to coimmunoprecipitate with FRα. Since gp130 can engage with multiple other receptors e.g. IL-6R, CNTFR and LIFR [40], it is likely that only a percentage of total cellular gp130 will associate with FRα. This is indeed observed with the co-immunostaining of HeLa cells (Fig. 8B). Some colocalization of FRα and gp130 can be seen, but also areas in the cell and cell membrane without co-localization can be detected. Activation of the gp130 receptor can result in ERK1/2 activation in addition to STAT3 activation. To test if ERK1/2 was activated in HeLa cells after folic acid treatment, a time course experiment was performed (Fig. 8C). The activation of ERK1/2 is evident from 10 to 20 min after addition of folic acid, thus a similar activation pattern was elicited as for folic acid induced STAT3 activation. Taken together, our results indicate that gp130 is a co-receptor for folic acid induced STAT3 activation via FRα.

501 502

D

440

acid concentrations, compared with a control cell line (HeLa-pcDNA3). As seen in Fig. 9, the lowest concentration required for an increase in STAT3 activation is lower (50 μg/ml) for the HeLa-FRα cell line compared with HeLa-pcDNA3 control cells (200 μg/ml). Both cell lines exhibit a significant dose response to folic acid. However, the responses of the two cell lines are significantly different, demonstrating that overexpression of FRα significantly affects the folic acid induced STAT3 activation in HeLa cells.

E

3.6. The co-receptor gp130 in the FRα-STAT3 signalling transduction pathway

T

438 439

7

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

503 504 505 506 507 508

513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553

557 558 559 560 561

8

C

O

R

R

E

C

T

E

D

P

R O

O

F

crosslinking the gp130 receptor, then this may explain why FRα, gp130 565 and anti-FRα antibodies eluted together in the immunoprecipitation 566 experiment. 567

N

564

the STAT3 activation by these antibodies was not investigated. The mechanism of anti-FRα antibody induced STAT3 activation needs further investigation. Moreover, if anti-FRα antibodies can activate STAT3 by

U

562 563

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

9

R

E

C

T

E

D

P

R O

O

F

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

N C O

R

Fig. 6. Co-receptors for folic and folinic acid induced STAT3 activation. Immunoblot analysis with the anti-pSTAT3(Y705) antibody of protein extracts from HeLa cells treated with (+) or without (−) A. and B. 3 μM EGF receptor inhibitor AG1478, C. 16 μg/ml of the anti-FRα antibody Mov18/ZEL, or D. 16 μg/ml of the indicated anti-FRα antibodies prior to treatment with 25 ng/ml epidermal growth factor (EGF), 250 U/ml interferon β (IFN), 600 U/ml IL-6, 250 μg/ml folic acid, 250 μg/ml folinic acid, or left untreated (control). Immunoblot analysis with the anti-STAT3 antibody was used as a loading control. The arrows indicate the position of the EGF induced autophosphorylated EGF receptor of 170 kDa recognized by the pSTAT3(Y705) antibody (#9131) used in this experiment. Protein bands from the presented blots (A, B and D) and from biologically independent experiments (C) were quantified, and pSTAT3 levels relative to STAT3 levels were calculated. Untreated controls were set to 1. Fold induction (FI) indicated below the blots (A, B and D) and the relative STAT3 activation depicted below the blots (C). Error bars represent SEM for n = 3. Statistical analysis was performed using paired t-tests. *: P = 0.005, ns = not significantly different.

4.3. Other signals initiated by folates

569

Two studies have addressed other signalling pathways initiated by folates in neural stem cells. Firstly, the Notch signalling system was altered in neural stem cells during folate stimulated proliferation [43]. Neither FRs nor STAT3 was investigated, and both these factors might be involved. Secondly, folate was found to stimulate ERK1/2 phosphorylation and cell proliferation in neural stem cells [44]. However, this study did not investigate if FRα or STAT3 was involved in the signalling

570 571 572 573 574 575

U

568

pathway, therefore it cannot be excluded that both ERK1/2 and STAT3 576 could be activated by folate in these cells. 577 4.4. Folic acid levels

578

The concentrations of folic acid used in these experiments were higher than physiological serum levels. When using the HeLa-FRα cell line with higher amounts of FRα on the cell surface, we were however able to lower the amounts of folic acid required for STAT3 activation,

579

Fig. 5. Time course and dose responses of folic acid induced STAT3 activation. A. Immunoblot analysis of protein extracts from HeLa cells treated with 500 μg/ml folic acid, 600 U/ml IL-6 or 1000 U/ml IFNβ for the indicated time using the anti-pSTAT3(Y705) antibody. Immunoblot analysis with the anti-STAT3 antibody was used as a loading control. Protein bands from three biologically independent experiments were quantified, and pSTAT3 levels relative to STAT3 levels are depicted to the right. The relative pSTAT3/STAT3 level at time 0 for each treatment was set to 1. B. Immunoblot analysis of protein extracts from HeLa cells treated with the indicated amounts of folic acid (μg/ml), IL-6 (U/ml) or IFNβ (U/ml) for 10 min using

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

580 581 582

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

C

O

R

R

E

C

T

E

D

P

R O

O

F

10

U

N

Fig. 7. The gp130 as a co-receptor for folic acid induced STAT3 activation. A. Immunoblot analysis with the anti-pSTAT3(Y705) antibody of protein extracts from HeLa cells treated with (+) or without (−) 8 μg/ml neutralizing anti-gp130 antibody, prior to treatment with 600 U/ml IL-6, 250 μg/ml folic acid, 250 μg/ml folinic acid, or left untreated (control). Immunoblot analysis with the anti-STAT3 antibody was used as a loading control. Protein bands from biologically independent experiments were quantified, and pSTAT3 levels relative to STAT3 levels were calculated. Untreated controls were set to 1. The relative STAT3 activation is depicted below the blot. Error bars represent SEM for n = 3. Statistical analysis was performed using paired ttests. *: P = 0.002, ns = not significantly different. B. Immunoblot analysis of protein extracts from HeLa cells transfected with siRNAs targeting gp130, FRα, or both gp130 and FRα in half amounts (gp130/FRα) using the anti-pSTAT3 antibody. Non-targeting control (Con) was EGFP siRNA. Cells were treated with 600 μg/ml folic acid, 200 U/ml, IL-6 or left untreated. Protein bands from biologically independent experiments were quantified, and pSTAT3 levels relative to STAT3 levels are depicted below the blots. The relative pSTAT3/STAT3 level in untreated cells transfected with the non-target control EGFP siRNA was set to 1. Error bars represent SEM from n = 3. Statistical analysis was performed using paired t-tests. *: P b 0.05, ns = not significantly different. C. Immunoblot analyses with antibodies against FRα (upper panel) or gp130 (lower panel) of protein extracts from HeLa cells transfected with siRNAs targeting gp130 or FRα, or a control siRNA (eGFP), as indicated. Equal amounts of protein extracts were applied to the SDS-PAGE.

583 584 585 586 587 588 589

suggesting that the level of FRα is a limiting factor on HeLa cells. In addition, we have observed that omitting serum from the medium during the folate starvation, as well as in the following folic acid treatments of HeLa cells, resulted in no STAT3 activation. This indicates that an unidentified stabilising folic acid binding protein in the medium is most likely needed in order to convey the signal, which is dependent on folic acid in a dose dependent manner. This factor could be bovine

serum albumin, since folate in serum is associated with serum albumin [10]. Another consideration could be circulating soluble FRα receptors, which have been found to be up-regulated in serum from ovarian cancer patients [45]. When present, these soluble receptors might bind folic acid in the serum and may thus inhibit or convey the STAT3 activating signal through gp130. Although soluble folic acid-FRα transsignalling is speculative at this point, similar transsignalling

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

590 591 592 593 594 595 596

11

U

N C O

R

R

E

C

T

E

D

P

R O

O

F

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

Fig. 8. The co-receptor gp130 in the FRα-STAT3 signalling transduction pathway. A. Co-immunoprecipitation of FRα and gp130 in HeLa cells. Protein extracts from HeLa cells were subjected to immunoprecipitation (IP) using the anti-FRα antibody Mov18/ZEL, and the proteins analysed by immunoblot analysis using antibodies against the indicated proteins (WB). Input 10%: The amount of protein extract loaded on the gel represents 10% of the protein extract used for IP. FT: flow-through. Proteins were eluted in two successive elutions (1 and 2) with the second being more harsh causing the elution of anti-FRα antibody. The light and heavy chain antibody fragments are indicated (*). Representative blots from three biologically independent experiments are presented. B. Immunofluorescent staining af HeLa cells using antibodies against FRα (Mov18/ZEL, green) and gp130 (red), and the nuclei were stained with DAPI (blue). The merged picture shows an overlay of the three stainings. Scale bars indicate 10 μm. C. Time course of ERK1/2 activation in HeLa cells. Immunoblot analysis of protein extracts from HeLa cells treated with 500 μg/ml folic acid for the indicated amount of time using antibodies against phosphorylated ERK1/2 (pERK1/2). Immunoblot analysis with the anti-ERK1/2 antibody was used as a loading control. Protein bands from biologically independent experiments were quantified, and pERK1/2 levels relative to ERK1/2 levels were calculated and depicted below the blot. Untreated controls were set to 1. Error bars represent SEM from n = 2.

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

600 601

608 609

New evidence behind the mechanism of tissue folate transport in the brain has recently been published, revealing a new role for FRα in transcytosis [26]. The first step in this process involves binding of serum folate to FRα expressed on choroid plexus epithelial cells, followed by transcytosis across the epithelial cell. Here, exosomes with folate bound FRα are released and specifically delivered to astrocytes within the brain parenchyma. The authors hypothesised that this complex process might be regulated by cell signalling, and although speculative at this point, it is possible that folic acid and folinic acid induced STAT3 activation may be involved in the regulation of this process.

610

4.6. Folate and neural tube defects

611 612

Inadequate serum levels of folate as well as autoantibodies against FRs have been observed in women with pregnancies complicated by neural tube defects [48,49]. Mice deficient in the Folbp1 gene (the mouse homologue of FRα) display multiple developmental abnormalities including neural and craniofacial defects, and several crucial differentiation markers are mis-expressed [50,51]. Likewise, lithium and

E T C E R R

615 616

Our finding that FRα binding of folic acid can initiate an intracellular signalling pathway leading to the activation of STAT3 and ERK1/2, upregulation of the STAT3 regulated genes VEGF and Cyclin A2, as well as increased folic acid induced proliferation of the FRα-positive HeLa cells, correlates with various studies suggesting that FRα plays a role in tumour etiology and progression [14]. In spite of highly upregulated FRα on cancer cells, this receptor only seems to contribute to a minority of the cellular uptake of folic acid at least in ovarian cancer

O

613 614

4.7. FRα is expressed in cancer cells

C

606 607

N

604 605

U

602 603

F

4.5. The FR and transcytosis

O

599

homocysteine induced cardiac abnormalities in mice could be rescued by folic acid treatment [52,53]. These studies support the importance of FRs in embryogenesis at the time of closure of the neural tube, giving rise to the central nervous system. It should be noted that STAT3 is activated at this particular time in development in the mouse [54], suggesting that the folic acid induced FRα -STAT3 signalling pathway may be crucial in the initial development of the central nervous system. Moreover, a study investigating the combined effect of folate and vitamin B12 deficiency in rat brain tissues found that lack of these B vitamins down-regulated activated STAT3 [13]. In neuronal progenitor cell lines, folic acid deficiency alone had a similar effect, suggesting that the decreased STAT3 activation seen was mediated by the lack of folic acid.

R O

pathways have been described for soluble IL-6 receptors able to activate gp130 after IL-6 binding [46,47].

D

597 598

P

12

Fig. 9. Folic acid induced STAT3 activation in FRα overexpressing HeLa cells. A. Immunoblot analysis of HeLa-FRα and HeLa-pcDNA3 protein extracts using the pSTAT3 antibody. Cells were treated with the indicated amounts of folic acid for 10 min. Immunoblot analysis with the anti-STAT3 antibody served as a loading control. Protein bands from biologically independent experiments were quantified, and pSTAT3 levels relative to STAT3 levels were calculated and depicted below the blots. Untreated controls were set to 1. Error bars represent SEM from n = 3. Folic acid dose correlations were significant for both cell lines ** (P b 0.007). Statistical analysis was performed using paired t-tests, and the two cell lines responded significantly different to folic acid treatment * (P = 0.04). B. Immunoblot analysis of protein extracts from the indicated HeLa cells using the Mov18/ZEL antibody against FRα. α-Tubulin was used as loading control.

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638

M.F. Hansen et al. / Cellular Signalling xxx (2015) xxx–xxx

643

652

In conclusion, we have depicted a new signal transduction pathway initiated by folic acid or folinic acid binding to FRα thereby eliciting a cellular response through the co-receptor gp130 followed by activation of JAK and STAT3. This activation of the STAT3 oncogene by folic acid and folinic acid adds to the increased complexity of the established roles of the B9 vitamins in one-carbon metabolism. This signal transduction pathway may influence various cellular processes including cell proliferation, development and cancer through the FRα. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.cellsig.2015.03.020.

653

Competing interests

650 651

654

The authors declare no competing interests. Author contributions

656 657

670

MFH carried out folic acid dose–response and time course experiments, gp130 knockdown and co-immunoprecipitation experiments, was involved in designing and interpreting Q-RT PCR results and drafted the manuscript. EG carried out the initial folic acid and folinic acid induced STAT3 activation studies in HEK293 and HeLa cells, performed the STAT3 transcription factor assay, analysed putative co-receptors for folic acid signalling, purified RNA for the qRT-PCR, and drafted the manuscript. ACMK carried out the proliferation assay. SS, SR, and TRJ constructed FRα-HeLa cell lines and carried out the analyses. CSN carried out the IL-6 and IFN time course and dose–response experiments. KNK carried out the qRT-PCR analyses. CSK carried out the initial experiments and conceived of the study. PMM conceived of the study, participated in its design and coordination, carried out some of the experiments, and drafted the manuscript. All authors read and approved the final manuscript.

671

Acknowledgements

672

679 680

We thank Inger Bjorndal for excellent laboratory assistance, Dr. Alexander Schmitz for expert assistance in flow cytometry analysis, Associate Professor Just Justesen for helpful discussions and Associate Professor Ernst-Martin Füchtbauer for helpful comments on this manuscript. Support from the Danish Natural Science Research Council, the Danish Cancer Society, the A.P. Moller Foundation for the Advancement of Medical Science, the Arne Hansen Foundation, Else and Mogens Wedell-Weddelsborgs Foundation and Aarhus University was highly appreciated.

681

References

673 674 675 676 677 678

682 683 684 685 686 687 688

C

668 669

E

666 667

R

664 665

R

662 663

[1] [2] [3] [4]

N C O

660 661

U

658 659

T

655

P

648 649

D

646 647

E

644 645

F

5. Conclusions

O

642

D.R. Hodge, E.M. Hurt, W.L. Farrar, Eur. J. Cancer 41 (2005) 2502–2512. H. Yu, D. Pardoll, R. Jove, Nat. Rev. Cancer 9 (2009) 798–809. N. Jing, D.J. Tweardy, Anti-Cancer Drugs 116 (2005) 601–607. G. Niu, K.L. Wright, Y. Ma, G.M. Wright, M. Huang, R. Irby, J. Briggs, J. Karras, W.D. Cress, D. Pardoll, R. Jove, J. Chen, H. Yu, Mol. Cell. Biol. 25 (2005) 7432–7440. [5] W.M. Burke, X. Jin, H.-J. Lin, M. Hunag, R. Liu, R.K. Reynolds, J. Lin, Oncogene 20 (2001) 7925–7934.

[6] P.A. Johnston, J.R. Grandis, Mol. Interv. 11 (2011) 18–26. [7] C. Schindler, C. Plumlee, Semin. Cell Dev. Biol. 19 (2008) 311–318. [8] P.C. Heinrich, I. Behrmann, G. Müller-Newen, F. Schaper, L. Graeve, Biochem. J. 334 (1998) 297–314. [9] A. Jarnicki, T. Putoczki, M. Ernst, Cell Div. 5 (2010) 14. [10] M. Lucock, Mol. Genet. Metab. 71 (2000) 121–138. [11] N. Parker, M.J. Turk, E. Westrick, J.D. Lewis, P.S. Low, C.P. Leamon, Anal. Biochem. 338 (2005) 284–293. [12] S.J. Duthie, J. Inherit. Metab. Dis. 34 (2010) 101–109. [13] R. Kerek, A. Geoffroy, A. Bison, N. Martin, N. Akchiche, G. Pourie, D. Helle, J.L. Gueant, C. Bossenmeyer-Pourie, J.L. Daval, Cell Death Dis. 4 (2013) (e755). [14] L.E. Kelemen, Int. J. Cancer 119 (2006) 243–250. [15] K.R. Kalli, A.L. Oberg, G.L. Keeney, T.J.H. Christianson, P.S. Low, K.L. Knutson, L.C. Hartmann, Gynecol. Oncol. 108 (2008) 619–626. [16] L.C. Hartmann, G.L. Keeney, W.L. Lingle, T.J.H. Christianson, B. Varghese, D. Hillman, A.L. Oberg, P.S. Low, Int. J. Cancer 121 (2007) 938–942. [17] L.H. Matherly, I.D. Goldman, Vitam. Horm. 66 (2003) 403–456. [18] C. Leamon, A.L. Jackman, Vitam. Horm. 79 (2008) 203–233. [19] P.S. Low, S.A. Kularatne, Curr. Opin. Chem. Biol. 13 (2009) 256–262. [20] X. Zhao, H. Li, R.J. Lee, Expert Opin. Drug Deliv. 5 (2008) 309–319. [21] S. Sabharanjak, P. Sharma, R.G. Parton, S. Mayor, Dev. Cell 12 (2002) 411–423. [22] J.H. Yang, H. Chen, I.R. Vlahov, J.-X. Cheng, P.S. Low, J. Pharmacol. Exp. Ther. 32 (2007) 462–468. [23] H. Birn, O. Spiegelstein, E.I. Christensen, R.H. Finnell, J. Am. Soc. Nephrol. 16 (2005) 608–615. [24] E.I. Christensen, H. Birn, Nat. Rev. Mol. Cell Biol. 3 (2002) 258–268. [25] H. Birn, X. Zhai, J. Holm, S.I. Hansen, C. Jacobsen, E.I. Christensen, S.K. Moestrup, FEBS J. 272 (2005) 4423–4430. [26] M. Grapp, A. Wrede, M. Schweizer, S. Huwel, H.J. Galla, N. Snaidero, M. Simons, J. Buckers, P.S. Low, H. Urlaub, J. Gartner, R. Steinfeld, Nat. Commun. 4 (2013) 2123. [27] G. Corona, F. Giannini, M. Fabris, G. Toffoli, M. Boiocchi, Int. J. Cancer 75 (1998) 125–133. [28] S. Chattopadhyay, Y. Wang, R. Zhao, I.D. Goldman, Clin. Cancer Res. 10 (2004) 7986–7993. [29] M.D.A. Salazar, M. Ratnam, Cancer Metastasis Rev. 26 (2007) 141–152. [30] S. Chattopadhyay, R.G. Moran, I.D. Goldman, Mol. Cancer Ther. 6 (2007) 404–417. [31] J. Zain, O. O'Connor, Expert. Opin. Pharmacother. 11 (2010) 1705–1714. [32] C. Chen, J. Ke, X.E. Zhou, W. Yi, J.S. Brinzelle, J. Li, E.L. Young, H.E. Xu, K. Melcher, Nature 500 (2013) 486–489. [33] A.S. Wibowo, M. Singh, K.M. Reeder, J.J. Carter, A.R. Kovach, W. Meng, M. Ratnam, F. Zhang, C.E. Dann 3rd, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 15180–15188. [34] C.A. Schneider, R.C.W., K.W. Eliceiri, Nat. Methods 9 (2012) 671–675. [35] C.S. Kronborg, U.B. Knudsen, P.M. Martensen, Cytokine 42 (2008) 105–112. [36] C.-L. Chen, F.-C. Hsieh, J.C. Lieblein, J. Brown, C. Chan, J.A. Wallace, G. Cheng, B.M. Hall, J. Lin, Br. J. Cancer 96 (2007) 591–599. [37] V. Sriuranpong, J.I. Park, P. Amornphimoltham, V. Patel, B.D. Nelkin, J.S. Gutkind, Cancer Res. 63 (2003) 2948–2956. [38] Y. Tanabe, T. Nishobori, L. Su, R.M. Ardiuni, D.P. Baker, M. David, J. Immunol. 174 (2005) 609–613. [39] C.D. Krause, W. He, S. Kotenko, S. Pestka, Cell Res. 16 (2006) 113–123. [40] S. Xu, N. Neamati, Expert Opin. Ther. Targets 17 (2013) 1303–1328. [41] G. Regis, S. Pensa, D. Boselli, F. Novelli, V. Poli, Semin. Cell Dev. Biol. 19 (2008) 35–359. [42] P. Autissier, J.D. Vos, J. Liautard, N. Tupitsyn, C. Jacquet, N. Chavdia, B. Klein, J. Brochier, J. Gaillard, Int. Immunol. 10 (1998) 1881–1889. [43] X. Zhang, H. Liu, G. Cong, Z. Tian, D. Ren, J.X. Wilson, G. Huang, J. Nutr. Sci. Vitaminol. 54 (2008) 353–356. [44] X.-M. Zhang, G.-W. Huang, Z.-H. Tian, D.-L. Ren, J.X. Wilson, Nutr. Neurosci. 12 (2009) 226–232. [45] E. Basal, G.Z. Eghbali-Fatourechi, K.R. Kalli, L.C. Hartmann, K.M. Goodman, E.L. Goode, B.A. Kamen, P.S. Low, K.L. Knutson, PLoS One 4 (2009) (e6292). [46] H. Knüpfer, R. Preiss, Immunol. Cell Biol. 86 (2008) 87–91. [47] S. Rose-John, J. Scheller, G. Elson, S.A. Jones, J. Leukoc. Biol. 80 (2006) 227–236. [48] H.J. Blom, Birth Defects Res., Part A 85 (2009) 295–302. [49] S.P. Rothenberg, M.P.d. Costa, J.M. Sequeira, J. Cracco, J.L. TRoberts, J. Weedon, E.V. Qudros, N. Engl. J. Med. 350 (2004) 134–142. [50] J.A. Piedrahita, B. Oetama, G.D. Bennett, J.v. Waes, B.A. Kamen, J. Richardson, S.W. Lacey, R.G. Anderson, R.H. Finnell, Nat. Genet. 23 (1999) 228–232. [51] L.S. Tang, R.H. Finnell, Birth Defects Res., Part A 67 (2003) 209–218. [52] S.-P. Han, Y. Pan, Y.-Z. Peng, X.-Q. Gu, R.-H. Chen, X.-R. Guo, Int. J. Mol. Med. 23 (2009) 321–330. [53] M. Han, M.C. Serrano, R. Lastra-Vicente, P. Brinez, G. Acharya, J.C. Huhta, R. Chen, K.K. Linask, Dis. Model. Mech. 2 (2009) 1–12. [54] Y. Yan, W. Bian, Z. Xie, X. Cao, I.L. Roux, F. Guillemot, N. Jing, Dev. Dyn. 231 (2004) 248–257.

R O

641

cells [27]. Our results show for the first time that folic acid and folinic acid may contribute to tumour growth by activating the STAT3 oncogene through the cancer specific FRα.

639 640

13

Please cite this article as: M.F. Hansen, et al., Folic acid mediates activation of the pro-oncogene STAT3 via the Folate Receptor alpha, Cell. Signal. (2015), http://dx.doi.org/10.1016/j.cellsig.2015.03.020

689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 Q3 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762