Cellular Signalling 26 (2014) 1604–1615

Contents lists available at ScienceDirect

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

TGF-β1–ROS–ATM–CREB signaling axis in macrophage mediated migration of human breast cancer MCF7 cells Rajshri Singh, Bhavani S. Shankar ⁎, Krishna B. Sainis Radiation Biology & Health Sciences Division, Bio-Science Group, Bhabha Atomic Research Centre, Mumbai 400 085, India

a r t i c l e

i n f o

Article history: Received 4 February 2014 Received in revised form 28 March 2014 Accepted 30 March 2014 Available online 3 April 2014 Keywords: Macrophage CREB EMT DNA damage Oxidative stress TGF-β1

a b s t r a c t Macrophages in the tumor microenvironment play an important role in tumor cell survival. They influence the tumor cell to proliferate, invade into surrounding normal tissues and metastasize to local and distant sites. In this study, we evaluated the effect of conditioned medium from monocytes and macrophages on growth and migration of breast cancer cells. Macrophage conditioned medium (MϕCM) containing elevated levels of cytokines TNF-α, IL-1β and IL-6 had a differential effect on non-invasive (MCF7) and highly invasive (MDA-MB-231) breast cancer cell lines. MϕCM induced the secretion of TGF-β1 in MCF7 cells. This was associated with apoptosis in a fraction of cells and generation of reactive oxygen and nitrogen species (ROS and RNS) and DNA damage in the remaining cells. This, in turn, increased expression of cAMP response element binding protein (CREB) and vimentin resulting in migration of cells. These effects were inhibited by neutralization of TNF-α, IL-1β and IL-6, inhibition of ROS and RNS, DNA damage and siRNA mediated knockdown of ATM. In contrast, MDA-MB-231 cells which had higher basal levels of pCREB were not affected by MϕCM. In summary, we have found that pro-inflammatory cytokines secreted by macrophages induce TGF-β1 in tumor cells, which activate pCREB signaling, epithelial–mesenchymal-transition (EMT) responses and enhanced migration. © 2014 Elsevier Inc. All rights reserved.

1. Introduction The tumor microenvironment composed of malignant and immune cells, cytokines, chemokines and stromal components including extracellular matrix (ECM) plays an important role to facilitate cancer progression and metastasis. Interactions between tumor and immune cells and the soluble factors they secrete, influence the tumor cell survival and proliferation, integrity of the ECM, invasion, angiogenesis and metastasis. Based on the elucidation of the roles of individual players in the tumor microenvironment, signaling pathways such as those modulated by transcription factors NF-κB and STAT3 are emerging as important targets for chemotherapeutics [1,2]. The importance of macrophages, one of the prominent infiltrating immune cells, in growth and metastasis of breast cancer is well documented. Focal macrophage infiltration is an important prognostic factor in breast invasive carcinoma and reduced survival is associated with high infiltration rates [3]. Tumor cell metastasis to lungs was significantly reduced in polyoma middle T antigen [PyMT] transgenic mice susceptible to breast cancer, when they were crossed with mice deficient in macrophage colony-stimulating factor 1 (CSF-1), whereas over-expression of CSF-1 in these mutants accelerated tumor cell dissemination and was ⁎ Corresponding author at: Immunology Section, Radiation Biology & Health Sciences Division, Bio-Science Group, Bhabha Atomic Research Centre, Modular Laboratories, Trombay, Mumbai 400 085, India. Tel.: +91 22 25593706; fax: +91 22 25505326. E-mail address: [email protected] (B.S. Shankar).

http://dx.doi.org/10.1016/j.cellsig.2014.03.028 0898-6568/© 2014 Elsevier Inc. All rights reserved.

associated with high macrophage infiltration [4]. Macrophages directly enhanced the metastatic growth through their effects on tumor cell extravasation, survival and subsequent growth [5]. Crosstalk between tumor cells and M2-macrophages increased the production of various pro-inflammatory cytokines including TNF-α, resulting in infiltration of CD45+ leukocytes into tumor tissues. Tumor cell proliferation, angiogenesis, and lymph angiogenesis also followed, thereby accelerating solid tumor growth and lung metastasis [6,7]. Some in vitro studies also have addressed the mechanistic aspects of these phenomena. Co-culture of cancer cells with macrophages resulted in an increased secretion of pro-inflammatory cytokines like tumor necrosis factor (TNF-α), interleukin-1β (IL-1β) and interleukin-6 (IL-6) [7,8]. A variety of cytokines and growth factors, such as TNF-α, transforming growth factor-β (TGF-β), hepatocyte growth factor (HGF), epidermal growth factor (EGF) etc., have been implicated in tumor–stroma crosstalk. The TGF-β-pathway is one of the major pathways altered in tumors, including breast cancer [9,10]. A critical step to establish metastasis is the cancer cell migration and invasion, which accounts for a large portion of cancer related deaths. An essential and initial process leading to the tumor invasion is epithelial–mesenchymal transition (EMT) [11]. During the EMT process, cells lose their epithelial properties such as cell polarity, normal cell–cell contact and acquire mesenchymal properties like fibroblast morphology, invasion and expression of mesenchymal markers including vimentin and N-cadherin [12,13]. Mechanisms that induce EMT involve multiple extracellular triggers and intracellular signaling pathways [14–16]. These include oncogenic signaling [17], Wnt3/β

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

catenin signaling, increased reactive oxygen species (ROS) [16] as well as DNA damage [15]. cAMP-response-element-binding protein (CREB) signaling is associated with cancer development and poor clinical outcome in leukemogenesis [18], breast cancer [19], and melanoma [20]. CREB is activated by a number of growth factors, hormones and stress signals that trigger its phosphorylation at Ser-133 and its association with the co-activators, CREB binding protein (CBP) and p300. However, the exact mechanisms by which the macrophage derived soluble factors increase tumor growth and migration and whether CREB could be a potential player remain to be elucidated. In the present study, we show that macrophages induce tumor derived TGF-β1. This further results in oxidative stress as well as DNA damage mediated signaling in tumor cells. This signaling activates CREB, EMT responses and increased migration. Neutralization of proinflammatory cytokines TNF-α, IL-1β and IL-6 independently or in combination as well as inhibition of oxidative stress or DNA damage abrogates MϕCM induced phosphorylation of CREB. Abrogation of CREB phosphorylation is also associated with decreased migration of MCF7 cells. These results thus emphasize the role of pro-inflammatory cytokines secreted by macrophages resulting in TGF-β1 signaling, CREB phosphorylation and metastatic phenotype in MCF7 cells.

1605

washed to remove PMA and were further cultured in serum containing growth medium for another 24 h. The medium was filtered and stored in − 20 °C. MCF7 and MDA-MB-231 cells were treated with 10% of MCM and MϕCM in the entire study. 2.3. Small interfering RNA (siRNA) treatment Cells were transfected at 70% confluence with ATM siRNA (ATM KD) or scrambled siRNA (ATM scr) using X-treme GENE transfection reagent according to the manufacturer's instructions and incubated for 72 h. 2.4. Colony forming assay

2. Materials and methods

MCF7 and MDA-MB-231 cells were seeded at a density of 103 cells/well in a 6 well plate in complete medium. Twenty four hours later, cells were incubated with MCM or MϕCM for a period of 6 days. Once colonies were formed, cells were washed with phosphate buffered saline (PBS) and fixed with methanol:acetone (7:3) at −20 °C. Fixed colonies were stained with crystal violet. The stained colonies were counted using Gelcount™ cell counter (Oxford Optronix) and data were analyzed by Gelcount™ software. Each experiment was carried out in triplicates/group and was repeated three times. Cells not treated with any supernatant served as controls.

2.1. Chemicals and antibodies

2.5. Assay for apoptosis

Dulbecco's modified Eagle's medium (DMEM), RPMI 1640 and fetal bovine serum (FBS) were purchased from HiMedia (Mumbai, India). Anti-vimentin antibody (clone V-9; sc-6260), anti-iNOS antibody (N-20; sc-651), anti-rabbit IgG-HRP conjugated antibody (sc-2030), ATM siRNA (sc-29761) and scrambled control siRNA (sc-37007) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antiγ-H2AX antibody (Ser 139, clone JBW 301; 05-636), PVDF membrane and chemiluminescent HRP substrate were purchased from Millipore (Billerica, MA, USA). Anti-PARP antibody (clone 46D11; 9532), antiCREB antibody (clone 48H2; 9197) and anti-phospho CREB antibody (S133, clone 87G3; 9198) were from Cell Signaling Technology (Danvers, MA, USA). Anti-phospho ATM (ser 1981, clone 10H11.E12; 05-740) was obtained from Upstate cell signaling solutions (Lake Placid, NY, USA). Alexa Fluor 488 conjugated secondary antibodies, 4-amino-5methylamino-2′,7′-difluorofluorescein (DAF-FM), JC-1, anti-fade mountant and calcein-AM were procured from Invitrogen (Grand Island, NY, USA). Dichlorofluoresceine diacetate (DCF DA) was obtained from Fluka (Bucks, Switzerland). Antibiotics, trypsin, anti-TGF-β, pan (T9429), phorbolmyristate acetate (PMA), propidium iodide, RNase A, crystal violet, 1400 W, CGK733, N-acetyl cysteine (NAC) and Tween 20 were purchased from Sigma (St. Louis, MO, USA). Protease, phosphatase inhibitor cocktail and X-treme GENE transfection reagent were procured from Roche Applied Science (Germany). Human TNF-α, human IL-1β, human IL-6, human IFN-γ, human TGF-β1 ELISA kits, GolgiPlug (555029) and transwell inserts (353097) were obtained from BD Biosciences (Franklin Lakes, NJ, USA).

MCF7 and MDA-MB-231 cells were seeded at a density of 104 cells/well in 6 well plates in complete medium. Twenty four hours later, cells were treated with MCM or MϕCM and incubated for a period of 5 days. The cells were subsequently harvested using trypsin and fixed with 70% ethanol at − 20 °C overnight. Cells were washed with PBS and re-suspended in 50 μg/ml propidium iodide and 50 μg/ml RNase A at 37 °C for 30 min. Samples were acquired in a Partec CyFlow Space™ flow cytometer and data were analyzed using FCS Express™ software. Cells with less than G1 DNA content were enumerated as apoptotic cells. 2.6. Quantitation of cytokines The cytokines TNF-α, IL-1β, IL-6, TGF-β1 and IFN-γ in MCM or MϕCM were estimated using ELISA kits according to the manufacturer's instructions. The cytokine TGF-β1 was estimated in CM derived from MCM or MϕCM treated MCF7 cells. 2.7. Measurement of ROS and reactive nitrogen species (RNS) MCF7 and MDA-MB-231 cells were seeded at a density of 104 cells/well in a 6 well plate in complete medium. Twenty four hours later, cells were incubated with MCM or MϕCM for 24 h, 48 h and 5 days. Cells were treated with 20 μM DCF DA for detection of ROS and 10 μM DAF-FM for detection of RNS for 30 min at 37 °C. Cells were subsequently harvested using trypsin, washed with PBS and acquired in a flow cytometer and data were analyzed by FCS Express™ software.

2.2. Cell culture and generation of conditioned media 2.8. Intracellular labeling for flow cytometry The human breast adenocarcinoma cell lines MCF7 and MDA-MB-231 and monocyte cell line U937 were obtained from National Centre for Cell Sciences, Pune, India. The breast cancer cell lines were maintained in DMEM and U937 cells were maintained in RPMI 1640 supplemented with 10% FBS, 100 units/ml of penicillin and 100 μg/ml of streptomycin (complete medium) in a humidified atmosphere at 37 °C in 5% CO2. For the preparation of monocyte conditioned medium (MCM), U937 cells were grown in complete media for 24 h at 37 °C. After 24 h, culture supernatant was collected, filtered using 0.2 μm membranes and stored at −20 °C. For macrophage conditioned medium (MϕCM), U937 cells were treated for 48 h with 160 nM PMA. The adherent cells were

MCF7 and MDA-MB-231 cells (104/well) were incubated with MCM or MϕCM for 5 days. For measuring intracellular TGF-β1 expression in MCF7 cells following MϕCM treatment, the cells were incubated with GolgiPlug (a protein transport inhibitor) for 4 h prior to harvesting for flow cytometric analysis. Cells were harvested using trypsin and fixed with 1 ml 70% ethanol in −20 °C overnight. Cells were further washed with PBS and non-specific binding was blocked with blocking buffer (5% FBS in PBS) for 30 min. The cells were incubated with anti-pan-TGFβ1 antibody diluted in blocking buffer for 1 h at room temperature (RT). After three washes with PBS, the cells were incubated with Alexa Fluor

1606

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

488 secondary antibody. Cells labeled only with secondary antibody served as a negative control. Samples were acquired in a flow cytometer and were analyzed by FCS Express™ software. 2.9. Immunofluorescence MCF7 and MDA-MB-231 cells grown on glass coverslips (104/well) were incubated with MCM or MϕCM for 5 days. For antibody neutralization experiments, MϕCM was treated with 1 μg of anti-TNF-α, IL-1β, IL-6 alone or in combination for 3 h at 37 °C. The treated CM was filtered through a 0.22 μM filter prior to treatment of MCF7 cells. After treatment for 5 days, the cells were permeabilized with methanol: acetone (7:3) for 20 min at − 20 °C. After 30 min of treatment with blocking buffer, cells were incubated with primary antibodies (pCREB, CREB, pATM, γ-H2AX, vimentin) for 1 h at RT. The cells were then stained with Alexa Fluor 488 conjugated secondary antibody. Cells labeled only with the secondary antibody served as a negative control. The coverslips were mounted on glass slides with anti-fade mountant having DAPI. Images were acquired using a Nikon Eclipse Ti™ inverted microscope equipped with a Nikon digital camera using NIS elements™ software. The intensity of the fluorescence labeling of cells was quantitated using Image J software in a minimum of 50 cells. 2.10. Wound healing and migration assays The mobility of the breast cancer cells in the presence of MCM and MϕCM was assessed by a wound healing assay (2 × 104 cells/well). MCF7 and MDA-MB-231 cells were grown in growth medium containing MCM or MϕCM for 5 days. After formation of a confluent monolayer, wounds were created using a sterile pipette tip and the cells were incubated with serum free media. The cells were stained with calcein AM and images of the wound were acquired using a Nikon Eclipse Ti™ microscope equipped with a Nikon digital camera. The ability of cells to migrate into the wound area was assessed by comparing micrographs at 0 and 24 h along the wounded area. The number of cells in the wounded area was calculated using CellProfiler image analysis software. Migration of MCF7 and MDA-MB-231 cells was studied using a transwell chamber with 8 μm pores. MCF7 cells were treated with scrambled control or ATM siRNA for 72 h prior to the experiment. The cells (105) were seeded in the top chamber of each transwell and allowed to migrate for 72 h. The bottom chamber contained MCM or MϕCM alone or along with NAC (5 mM), 1400 W (10 μM) or CGK733 (2 μM). The cells on the upper side of the membrane were removed with a cotton swab. The membranes were fixed in methanol:acetone (7:3) for 20 min at − 20 °C and stained with crystal violet for 5 min. The number of cells that migrated to the lower surface of the membrane was counted in a minimum of ten fields by light microscopy using CellProfiler image analysis software. 2.11. Statistical analysis All results are expressed as mean ± S.E.M. Statistical difference between means was assessed using Student's t test, with a P value less than 0.05 considered significant. 3. Results 3.1. Differential modulation of growth pattern of breast cancer cells MCF7 and MDA-MB-231 by MϕCM To assess the effect of macrophages on growth of breast cancer cells, we treated MCF7 and MDA-MB-231 cells with MCM or MϕCM. Treatment with MϕCM resulted in a decreased colony forming ability of both breast cancer cell lines (Fig. 1a and e). There was greater than a 2 fold decrease in the number of colonies (P b 0.01; Fig. 1c and g). When these cells were observed under a microscope, we noticed that in MCF7 cells, though

there was a decrease in the number of colonies following MϕCM treatment, the smaller colonies had merged with each other forming a network (Fig. 1a inset). These observations were validated by the change in scatter properties of MCF7 cells treated with MϕCM. There were two distinct populations, viable cells with higher FSC and dead cells with lower FSC and SSC (Fig. 1b). Apoptotic cells are characterized by lower FSC and SSC [21]. However, in MDA-MB-231, though the number of colonies was less, the cells had increased in size and had not merged with the neighboring colonies like MCF7 cells (Fig. 1e inset). The increased size was substantiated by the population with increased FSC and SSC observed in MDA-MB-231 cells treated with MϕCM (Fig. 1f). Increased forward scatter indicates changes in cell size and have been associated with differentiation [22] as well as senescence [23]. Next we addressed the role of apoptosis in this effect of MCM and MϕCM on breast cancer cells. There was a marginal increase in cell death in MϕCM treated MCF7 cells 24 h after treatment which gradually increased up to two fold by day 5 (P b 0.01; Fig. 1d). In contrast, there was no increase in cells undergoing apoptosis in MDA-MB-231 cells even 5 days after MϕCM treatment (Fig. 1h). We further examined mitochondrial membrane potential (ΔΨm) in these breast cancer cells treated with MCM and MϕCM since depolarization of the mitochondrial membrane is established to be part of the mitochondria-mediated apoptotic pathway [19]. ΔΨm was determined by flow cytometry using the fluorescent dye JC-1 and fluorescence was analyzed as the ratio of FL2 (585 nm)/FL1 (525 nm). Mitochondrial membrane potential has a critical role in the intrinsic pathway of apoptosis with an initiating cascade of apoptosis linked to Δψm dissipation [24]. In MCF7 cells, the FL2/FL1 fluorescence ratio of JC-1 increased following treatment with MϕCM (P b 0.01; Fig. S1). This intrinsic ratio (basal level between the two cell lines) was higher in MDA-MB-231 cells as compared to MCF7 cells and was not significantly altered following treatment with MϕCM (Fig. S1). Further, this was not accompanied by an increase in the expression of pro-apoptotic Bax family members or a decrease of antiapoptotic Bcl-2 family members (Fig. S2). 3.2. Pro-inflammatory cytokines in MϕCM induced TGF-β1 in MCF7 cells We detected the presence of some cytokines in MCM and MϕCM. IFN-γ and TGF-β1 were not present in both the conditioned media. In MCM, TNF-α, IL-1β and IL-6 were undetectable and in MϕCM, TNF-α (300–700 pg/ml), IL-1 β (700–900 pg/ml) and IL-6 (400–600 pg/ml) were present (Fig. 2a). We also estimated the levels of TGF-β1 in MCF7 cells treated with MCM and MϕCM. This was undetectable in the control cells as well as in MCM treated cells. However there was a significant increase (P b 0.01) in MCF7 cells treated with MϕCM (Fig. 2b). We confirmed this by detection of intracellular accumulation of TGF-β1. There was no significant change in the intracellular levels of TGF-β1 between untreated and MϕCM treated MCF7 cells. When the secretion of proteins was blocked using a GolgiPlug, there was no difference in untreated (32%) and MCM treated cells (33%). However, with this treatment, there was an increase in the percentage of cells expressing intracellular TGF-β1 (50%) in MϕCM treated cells (Fig. 2c). 3.3. Increase in TGF-β1 was associated with an increase in ROS, iNOS and RNS and DNA damage in MCF7 cells Cytokines like TNF-α, IL-1β, and IFN-γ as well as TGF-β1 have been known to increase production of ROS [16,25]. We, therefore, estimated the generation of ROS and RNS in cancer cells following treatment with MϕCM by using specific dyes like DCFDA and DAF. The cell permeable fluorogenic probe 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) diffuses into cells and is deacetylated by cellular esterases to nonfluorescent 2′,7′-dichlorodihydrofluorescein (DCFH), which is rapidly oxidized to highly fluorescent 2′,7′-dichlorodihydrofluorescein (DCF) by ROS. The fluorescence intensity is proportional to the ROS levels

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

1607

Fig. 1. MϕCM decreases colony forming ability of breast cancer cells. MCF7 and MDA-MB-231 cells were grown for 6 days in the presence of MCM or MϕCM. (a, e) Photomicrographs of the plate in duplicates. Inset: 10× photomicrograph of the respective colonies showing the morphology of cells. (c, g) The number of colonies was quantified in Oxford Optronix GelCount™ using Gelcount™ software. Data from one representative experiment are shown. The experiment was repeated three times. **P b 0.01. (b, f) After 5 days of treatment, twenty thousand cells were harvested and analyzed in a Partec CyFlow Space™ flow cytometer and density plots of forward and side scatter analyzed using FCS Express™ software. (d) Percent apoptosis in MCF7 cells. On day 5 following treatment, cells stained with propidium iodide were acquired in a CyFlow Space™ flow cytometer and analyzed using Cyflogic software. Cells with less than G1 DNA content were enumerated as apoptotic cells. (h) Percent apoptosis in MDA-MB-231 cells.

1608

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

Fig. 2. Pro-inflammatory cytokines in MϕCM induce secretion of TGF-β1 in MCF7 cells. (a) MCM and MϕCM were assayed for the presence of pro-inflammatory cytokines (TNF-α, IL-1β, IL-6, TGF-β1 and IFN-γ) by ELISA. Cytokines were estimated in MϕCM independently collected in three different experiments. (b) Expression of TGF-β1 in the CM of MCF7 cells treated with MCM or MϕCM for 5 days was carried out by ELISA. TGF-β1 was estimated in supernatants of three independent experiments. Data from one representative experiment are shown. (c) MCF7 cells were treated with MCM or MϕCM for 5 days and then treated with GolgiPlug, a protein transport inhibitor for 4 h. The treated cells were fixed with 70% ethanol at −20 °C overnight. The samples were blocked with 5% FBS in PBS and then labeled with anti-TGF-β1 antibody. Cells were acquired in Partec CyFlow Space™ flow cytometer, and data were analyzed by FCS Express™ software. This experiment was repeated twice in duplicates.

within the cell cytosol [26]. DAF-FM diacetate is cell-permeant and passively diffuses across cellular membranes. Once inside cells, it is deacetylated by intracellular esterases to become nonfluorescent DAF-FM which react with NO to form a fluorescent benzotriazole. The fluorescence intensity is proportional to the RNS levels within the cell cytosol. When MCF7 cells were exposed to MϕCM, there was a gradual increase in the generation of ROS and RNS as well as expression

of iNOS. MϕCM treatment of MCF7 cells resulted in a 70% surge in ROS generation (P b 0.01; Fig. 3a and b). These cells also had increased expression of iNOS (S3). The iNOS positive MCF7 cells increased from 25% to 65% upon MϕCM treatment with the MFI increasing from 155 to 369 (Fig. S3). This was associated with a 3 fold increase in RNS generation as measured by DAF fluorescence (P b 0.01; Fig. 3d and e). However, in parallel experiments conducted

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

1609

Fig. 3. Estimation of ROS, RNS and DNA damage in MCF7 cells. Breast cancer cells MCF7 and MDA-MB-231 were incubated with MCM or MϕCM for 5 days. Cells were labeled with 20 μM DCFDA or 10 μM DAF and increase in fluorescence intensity measured using flow cytometry. (a, b) DCF fluorescence in MCF7. (d, e) DAF fluorescence in MCF7. (c) DCF fluorescence in MDA-MB-231. (f) DAF fluorescence in MDA-MB-231. (g) MCF7 cells were labeled with pATM or (h) γ H2AX antibodies and visualized in a fluorescence microscope. Images were acquired using a Nikon Eclipse Ti inverted microscope equipped with a Nikon digital camera using the NIS elements™ software. Data from one representative experiment are shown. The experiment was repeated three times.

in MDA-MB-231 cells, there was no change in generation of ROS or RNS (Fig. 3c and f). Persistent increase in the level of ROS and RNS in the cells would lead to oxidative stress. Exposure of tumor cells to oxidative stress causes double-stranded DNA breaks (DSBs), triggering DNA damage response through activation of the ataxia telangiectasia mutated (ATM) kinase, which induces cell-cycle arrest and also promotes DNA repair to maintain chromosome stability. When activated, ATM autophosphorylates itself at Ser1981 and phosphorylates the histone variant H2AX at Ser139 around DSBs, which also recruits other DNA repair proteins to

sites of DNA damage. The phosphorylated histone H2AX (known as γH2AX) and the recruited DNA repair proteins form discrete nuclear foci at DSBs, providing surrogate markers to characterize the dynamic process of DNA repair [27]. To further characterize the effects of MϕCM induced oxidative stress, we assessed the DNA damage in MCF7 cells by monitoring the expression of pATM, γ-H2AX as well as PARP. In the untreated and MCM treated MCF7 cells, there was no labeling for pATM or γ-H2AX. However, following MϕCM treatment, there was an increase in the expression of pATM (Fig. 3g) as well as γ-H2AX (Fig. 3h). A majority of the cells showed DNA damage and the foci were very intense

1610

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

in some of the cells indicating complex DNA damage. This was confirmed by western blot as well as flow cytometry (Figs. S4 and S5). MFI of γ-H2AX labeled MCF7 cells increased from 62.50 to 83.44 after treatment with MϕCM (Fig. S5) as compared to that of cells treated with MCM. When the expression of PARP was assessed by western blot, we could observe the full-length PARP in all the treatment groups, whereas the cleaved PARP was seen only in MϕCM treated MCF7 cells (Fig. S6). 3.4. MϕCM induced increased CREB phosphorylation in MCF7 cells Deregulation of the response to ROS or DNA damage could result in activation of survival pathways leading to EMT [15,16]. One of the important survival pathways that get activated is CREB mediated signaling. Hence we assessed the effect of MϕCM on the phosphorylation as well as expression of CREB. There was a difference in the basal level of expression of pCREB (Fig. 4a–d) as well as total CREB (Fig. 4e–h) between the two cell lines. No labeling was seen in control MCF7 cells (Fig. 4a and c) whereas control MDA-MB-231 cells were positive for pCREB (Fig. 4b and d). However, upon treatment with MϕCM, MCF7 cells showed a dramatic increase in the phosphorylation of CREB and also a corresponding increase in the total protein (Fig. 4a, c, e and g). This was in contrast with MDA-MB-231 cells where the higher levels present in the untreated cells did not further change after treatment with MϕCM (Fig. 4b, d, f and h). 3.5. MϕCM induced expression of vimentin and EMT responses in MCF7 cells CREB binding proteins (CBP) mediate interactions between β-catenin and transforming growth factor-β signaling pathways and participate in regulation of EMT [28]. The expression of TGF-β1 downstream protein vimentin, a well known marker of EMT [17] was increased in MCF7 cells after treatment with MϕCM (Fig. 5a) whereas in MDA-MB-231 cells, it was not affected though the basal expression was very high (Fig. 5b). EMT responses lead to changes in migratory properties of cells. We evaluated these changes using two well established in vitro model systems, the wound healing assay as well as migration of cells through a transwell insert. MϕCM treatment resulted in increased closure of the wound area (Fig. 5c and d) as well as migration through the transwell in MCF7 cells (Fig. 5e and f). In MCF7 cells, there was a 4 fold augmentation in the cells moving into the wound area (P b 0.01; Fig. 5d) and a 6 fold increase in the cells migrating through the transwell following MϕCM treatment (P b 0.01; Fig. 5f). However, in highly invasive MDA-MB-231 cells, the treatment did not further enhance cell migration (Fig. 5g and h). 3.6. Neutralization of pro-inflammatory cytokines or inhibition of ROS, RNS or DNA damage could abrogate pCREB expression and migration To delineate the signaling pathways leading to activation of pCREB, we assessed the effect of neutralization of pro-inflammatory cytokines as well as the ability of various inhibitors to block this expression. Neutralization of the cytokines by pre-treatment of MϕCM with the respective antibodies alone or in combination also resulted in abrogation of the increase in pCREB expression (Fig. 6a; P b 0.01). Inhibition of change in the colony morphology was also observed. The scattering of cells and the inability to form colonies in the presence of MϕCM was inhibited and compact colonies were observed when any one of the proinflammatory cytokines was neutralized (Fig. S7). Neutralization of all three cytokines did not lead to any additive effect indicating the redundant function of these cytokines. N-acetyl cysteine (NAC) was used to

1611

scavenge ROS. 1400 W, a specific inhibitor of iNOS as well as CGK733, an inhibitor of ATM/ATR were also used. Pre-treatment with inhibitors of ROS, RNS or DNA damage substantially decreased pCREB expression induced by MϕCM treatment (Fig. 6b). The intensity of pCREB expression in cells was quantitated using ImageJ software. There was a 70% decrease in the intensity of pCREB labeling on pre-treatment with 1400 W (P b 0.01), a 3 fold decrease with NAC (P b 0.01) and a N2 fold reduction with CGK733 (P b 0.01) (Fig. 6b). A 30% decrease in pCREB expression was observed in ATM KD cells as compared to ATM scr cells in the presence of MϕCM (P b 0.01) (Fig. 6c). There was a 40% decrease in pATM labeling in ATM KD cells exposed to 5 Gy ionizing radiation as compared to the ATM scr cells (Fig. S8). Inhibition of iNOS with other inhibitors like NG-nitro-L-arginine (L-NNA), and L-NG-monomethyl arginine citrate (L-NMMA) also resulted in diminished phosphorylation of CREB (data not shown). Inhibition of ROS, RNS or DNA damage also inhibited the MϕCM induced migration in MCF7 cells (Fig. 6d). There was a significant decrease in migration of MCF7 cells in ATM KD following MϕCM treatment as compared to the scrambled siRNA treated control cells (P b 0.01, Fig. 6e). The schematic representation of the interaction between macrophages and tumor cells leading to EMT responses is given in Fig. 6f. 4. Discussion Macrophages and tumor cells mutually influence each other's behavior in majority of cancers, with the tumor cell attracting macrophages and sustaining their survival and they, in turn, producing a myriad of factors to promote or regulate tumor growth and angiogenesis. The main finding of this study is that the pro-inflammatory cytokines secreted by macrophages induce secretion of TGF-β1 in MCF7 cells. This results in apoptosis in a fraction of cells. In the remaining cells, there is increase in oxidative stress and DNA damage which trigger CREB mediated survival signaling inducing EMT responses. The data presented herein not only provide evidence that macrophage mediated release of soluble factors result in EMT responses in tumor cells but also point out to a differential effect on tumors depending on their invasive nature. The less invasive tumors may turn more aggressive whereas the already invasive tumors may not be significantly affected by the presence of soluble factors from macrophages. This is in agreement with an earlier report on the invasion of these cell lines in vivo after somatic hybridization with macrophages [29]. In the present study, we have observed changes in growth characteristics in both cell lines following MϕCM treatment. Increased apoptosis and merging of the colonies were observed in MCF7 cells accompanied by EMT responses. In MDA-MB-231 cells, there was an increase in size of the cells resembling a senescence phenotype. Pro-inflammatory cytokines like TNF-α, IL-1β and IL-6 were detected in the MϕCM. We demonstrate that these macrophage derived cytokines, in turn, induce the secretion of TGF-β1 from MCF7 breast cancer cells. Though about 30% of MCF7 cells were positive for intracellular TGF-β1, there was no detectable secretion in the supernatant. With MϕCM treatment of MCF7 cells, there was increased secretion of TGF-β1 as well as intracellular accumulation in the presence of GolgiPlug. The pro-inflammatory cytokine TNF-α, induces secretion of various cytokines depending upon the cell type involved: TGF-β1 in lung fibroblasts [30,31], IL-6 in breast cancer cells [32], IL-8 in endothelial cells [33], and IL-1 and IL-6 in cardiac fibroblasts [34]. TNF-α induced secretion of TGF-β2 in MCF7 breast cancer cells has been reported due to changes in post-translational control mechanisms [35]. In a panel of breast cancer cell lines with increasing invasive ability, the levels of TGF-β1, TGF-β2, TGF-βRI, and TGF-βRII mRNA increased in accordance with their

Fig. 4. MϕCM induces expression of pCREB and CREB in MCF7 cells. MCF7 and MDA-MB-231 cells were incubated with MCM or MϕCM for 5 days. The cells were labeled with the respective antibodies and visualized in a fluorescence microscope. Images were acquired using a Nikon Eclipse Ti inverted microscope equipped with a Nikon digital camera using the NIS elements™ software. The intensity of pCREB expression was quantified using Image J software in a minimum of 50 cells. Expression of (a, c) pCREB in MCF7, (b, d) pCREB in MDA-MB-231, (e, g) CREB in MCF7, and (f, h) CREB in MDA-MB-231. Data from one representative experiment are shown. The experiment was repeated three times.

1612

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

metastasis potential with MCF7 having the lowest levels [36]. TGF-β1 production in MCF7 cells was associated with an increase in oxidative stress, DNA damage and CREB mediated survival signaling. Though there was a two fold increase in apoptosis of MCF7 cells, the remaining surviving cells exhibited an increase in oxidative stress and DNA damage. As a multifunctional factor, TGF-β1 is involved in the regulation of many biological processes and induced concomitant apoptosis and EMT responses in hepatocytes [37] and this was closely related to the cell cycle stage. TGF-β1 induced apoptosis in cells synchronized at G2/M phase and EMT responses in unsynchronized cells and cells at G1/S phase of the cell cycle [38]. It has been referred to as a “double edged sword” because of its dual function as a tumor suppressor and a tumor promoter (reviewed in [39]). These data further confirm that it can induce apoptosis as well as EMT responses at the same time in a population of breast cancer cells, though we have not looked at changes in specific cell cycle stages. ROS and RNS play important roles in regulation of cell survival. In general, moderate levels of ROS/RNS may function as signals to promote cell proliferation and survival, whereas a severe increase of ROS/RNS can induce cell death. ROS include radical species such as superoxide (O− 2 ) and hydroxyl radical (HO•), along with non-radical species such as hydrogen peroxide (H2O2). In non-phagocytic cells like tumor cells, increased ROS generation can occur in mitochondria as a by-product of the respiratory chain or via the NADPH oxidase pathway [40,41]. RNS include nitric oxide (NO•) and peroxynitrite (ONOO−) and are generated through specific nitric oxide synthase isoenzymes (reviewed in [42]). Increased ROS/RNS production following MϕCM treatment was observed only in MCF7 cells and not in MDA-MB-231 cells. Expression of iNOS and secretion of NO antagonized TGF-β1 induced apoptosis and EMT in hepatocytes [43]. In contrast, however, we found that TNF-α, IL-1β and IL-6 induced TGF-β1 expression were associated with an increase in ROS as well as RNS production. In response to DNA damage or extracellular signals, expression of several transcription factors like CREB, NF-κB, c-fos and c-jun allows the cells to overcome a stressful or deleterious environment [44–47]. ATM is regarded as the major regulator of the cellular response to DNA double strand breaks (DSBs). Furthermore, ATM can also be activated directly by oxidative stress independent of DSBs by a mechanism distinct from MRN/DSB-dependent activation [48]. The observed ATM activation following MϕCM treatment could be directly activated by oxidative stress in a MRN independent pathway or as a result of MRN dependent DNA damage response. CREB belongs to the bZIP superfamily of transcription factors, which include CREB and the closely related factors CREM (cAMP response element modulator) and ATF-1 (activating transcription factor 1) [49]. Canonical activation of CREB occurs in response to cAMP, which induces PKA-dependent Ser-133 phosphorylation [50]. The phosphorylation of CREB on Ser-133 promotes recruitment of additional proteins or co-activators like CBP [51] and p300 [52]. In addition to its regulation by metabolic and growth signals, CREB is also a target of the DNA damage response [44,53,54]. We observed higher basal levels of DNA damage markers like pATM, γ-H2AX and transcription factor pCREB, CREB in MDA-MB-231 cells. When the pro-inflammatory cytokines TNF-α, IL-1β and IL-6 in MϕCM were neutralized using specific antibodies, there was a decrease in downstream pCREB expression as well as abrogation of the change in

1613

colony morphology induced by MϕCM. Pre-treatment with agents like NAC, that inhibit ROS generation or specific inhibitors of iNOS or ATM significantly reduce MϕCM induced pCREB expression in MCF7 cells. Similar results were also observed in ATM KD cells. It can be argued based on these results that DNA damage mediated ATM signaling indeed resulted in phosphorylation of CREB on ser-133. The functional implications of the ATM–CREB pathway have not been fully elucidated. The impact of DNA damage mediated CREB signaling has not been understood so far and the possibility of DNA damage mediated ser-133 phosphorylation could be an interesting possibility based on our results. Increased CREB stability as a result of phosphorylation of CREB on ser-133 has been reported [55]. We also observed increased total CREB expression in MCF7 cells following MϕCM treatment whereas the basal levels were high in MDA-MB-231 cells. These results implicate the stabilization of total CREB due to phosphorylation on ser-133. These data confirm an earlier report on differences between these two cell lines in terms of total CREB expression [56]. The argument, ‘oxidative stress induced DNA damage survival signaling was associated with increased EMT responses’, was strengthened by the use of inhibitors of this pathway. MϕCM treatment induced increased migration was significantly decreased following treatment with NAC or iNOS inhibitors or ATM/ATR inhibitor. The decreased migration was also seen in ATM KD cells. ATM mediated phosphorylation of CREB closely coincided with increased migration of cells, thus implicating CREB as a direct target for EMT responses in human breast cancer cells. Our results provide new evidence supporting the complex cytokine signaling network involved between the cancer cells and the macrophages in the tumor microenvironment. The pro-inflammatory cytokines secreted by the macrophages induce secretion of TGF-β1, which being a multi-functional cytokine induces apoptosis in MCF7 cells. The remaining fraction of cells responds to various signaling such as TGF-β1, oxidative stress and DNA damage. These changes ultimately culminate in the activation of EMT responses mediated by activation of CREB. Thus, our results have identified a distinct survival pathway that establishes a direct link between the cytokine signaling network in the tumor microenvironment, oxidative stress, DNA damage, pCREB and the selective control of human breast cancer cell migration.

Funding This work was supported by funding from Bhabha Atomic Research Centre, Government of India.

Conflict of interest statement None declared.

Acknowledgements We thank Mr. Prayag Amin for assistance in flow cytometry and Mr. N. S. Sidnalkar for his technical assistance. We also thank Dr. Savita Kulkarni and Mr. P.K. Gupta, RMC for help with cytokine ELISA.

Fig. 5. MϕCM treatment induces expression of vimentin and EMT responses in MCF7 cells. (a) MCF7 and (b) MDA-MB-231 cells were treated with MCM or MϕCM for 5 days. The cells were labeled with vimentin antibody and visualized in a fluorescence microscope. Images were acquired using a Nikon Eclipse Ti inverted microscope equipped with a Nikon digital camera using the NIS elements™ software. (c) MCF7 cells were grown in the presence of MCM or MϕCM for 5 days. After the formation of a confluent monolayer, wounds were created using a sterile pipette tip and the cells were incubated with serum free media for 24 h. The cells were stained with calcein AM and images of the wound were acquired using a Nikon Eclipse Ti microscope equipped with a Nikon digital camera. The ability of cells to migrate into the wound area was assessed by comparing micrographs at 0 and 24 h along the wounded area. Images of the wound following various treatments from a representative experiment are shown. (d) The number of cells in the wounded area calculated using CellProfiler image analysis software. (e) Migration of MCF7 and (g) migration of MDA-MB-231 were assessed by plating cells in the top chamber of transwell inserts that were allowed to migrate for 72 h. The bottom chamber contained MCM or MϕCM. The cells on the upper side of the membrane were removed with a cotton swab. The membranes were fixed in methanol:acetone (7:3) for 20 min and stained with crystal violet for 5 min. Images from a representative experiment are shown. (f) The number of MCF7 cells and (h) the number of MDA-MB-231 cells that migrated to the lower surface of the membrane were counted in ten randomly selected fields by light microscopy using CellProfiler software.

1614

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

R. Singh et al. / Cellular Signalling 26 (2014) 1604–1615

Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.cellsig.2014.03.028. References [1] S.I. Grivennikov, M. Karin, Cytokine Growth Factor Rev. 21 (1) (2010) 11–19. [2] W.W. Lin, M. Karin, J. Clin. Invest. 117 (5) (2007) 1175–1183. [3] R.D. Leek, C.E. Lewis, R. Whitehouse, M. Greenall, J. Clarke, A.L. Harris, Cancer Res. 56 (20) (1996) 4625–4629. [4] E.Y. Lin, A.V. Nguyen, R.G. Russell, J.W. Pollard, J. Exp. Med. 193 (6) (2001) 727–740. [5] B. Qian, Y. Deng, J.H. Im, et al., PLoS ONE 4 (8) (2009) e6562. [6] H.J. Cho, J.I. Jung, D.Y. Lim, G.T. Kwon, S. Her, J.H. Park, Breast Cancer Res. 14 (3) (2012) R81. [7] T. Hagemann, S.C. Robinson, M. Schulz, L. Trümper, F.R. Balkwill, C. Binder, Carcinogenesis 25 (8) (2004) 1543–1549. [8] Z. Hou, D.J. Falcone, K. Subbaramaiah, A.J. Dannenberg, Carcinogenesis 32 (5) (2011) 695–702. [9] L.M. Wakefield, A.B. Roberts, Curr. Opin. Genet. Dev. 12 (1) (2002) 22–29. [10] A.B. Roberts, L.M. Wakefield, Proc. Natl. Acad. Sci. U. S. A. 100 (15) (2003) 8621–8623. [11] J.P. Thiery, J.P. Sleeman, Nat. Rev. Mol. Cell Biol. 7 (2) (2006) 131–142. [12] A.K. Perl, P. Wilgenbus, U. Dahl, H. Semb, G. Christofori, Nature 392 (6672) (1998) 190–193. [13] F. van Roy, G. Berx, Cell. Mol. Life Sci. 65 (23) (2008) 3756–3788. [14] K. Polyak, R.A. Weinberg, Nat. Rev. Cancer 9 (4) (2009) 265–273. [15] N. Chiba, V. Comaills, B. Shiotani, et al., Proc. Natl. Acad. Sci. U. S. A. 109 (8) (2012) 2760–2765. [16] D.Y. Rhyu, Y. Yang, H. Ha, et al., J. Am. Soc. Nephrol. 16 (3) (2005) 667–675. [17] K. Vuoriluoto, H. Haugen, S. Kiviluoto, et al., Oncogene 30 (12) (2011) 1436–1448. [18] D.B. Shankar, K.M. Sakamoto, Leuk. Lymphoma 45 (2) (2004) 265–270. [19] C.F. Fan, X.Y. Mao, E.H. Wang, Mol. Med. Rep. 5 (2) (2012) 357–362. [20] V.O. Melnikova, A.S. Dobroff, M. Zigler, et al., PLoS ONE 5 (8) (2010) e12452. [21] F.J. Chrest, M.A. Buchholz, Y.H. Kim, T.K. Kwon, A.A. Nordin, Cytometry 14 (8) (1993) 883–890. [22] M. Daigneault, J.A. Preston, H.M. Marriott, M.K. Whyte, D.H. Dockrell, PLoS ONE 5 (1) (2010) e8668. [23] W. Wagner, P. Horn, M. Castoldi, et al., PLoS ONE 3 (5) (2008) e2213. [24] J.D. Ly, D.R. Grubb, A. Lawen, Apoptosis 8 (2) (2003) 115–128. [25] D. Yang, S.G. Elner, Z.M. Bian, G.O. Till, H.R. Petty, V.M. Elner, Exp. Eye Res. 85 (4) (2007) 462–472. [26] E. Eruslanov, S. Kusmartsev, Methods Mol. Biol. 594 (2010) 57–72. [27] S. Burma, B.P. Chen, M. Murphy, A. Kurimasa, D.J. Chen, J. Biol. Chem. 276 (45) (2001) 42462–42467.

1615

[28] B. Zhou, Y. Liu, M. Kahn, et al., J. Biol. Chem. 287 (10) (2012) 7026–7038. [29] J. Ding, W. Jin, C. Chen, Z. Shao, J. Wu, PLoS ONE 7 (7) (2012) e41942. [30] D.E. Sullivan, M. Ferris, D. Pociask, A.R. Brody, Am. J. Respir. Cell Mol. Biol. 32 (4) (2005) 342–349. [31] D.E. Sullivan, M. Ferris, H. Nguyen, E. Abboud, A.R. Brody, J. Cell. Mol. Med. 13 (8B) (2009) 1866–1876. [32] C. Suarez-Cuervo, K.W. Harris, L. Kallman, H.K. Väänänen, K.S. Selander, Breast Cancer Res. Treat. 80 (1) (2003) 71–78. [33] R.Z. Zhao, X. Chen, Q. Yao, C. Chen, Biochem. Biophys. Res. Commun. 327 (4) (2005) 985–992. [34] N.A. Turner, R.S. Mughal, P. Warburton, D.J. O'Regan, S.G. Ball, K.E. Porter, Cardiovasc. Res. 76 (1) (2007) 81–90. [35] D.N. Danforth, M.K. Sgagias, Clin. Cancer Res. 2 (5) (1996) 827–835. [36] L.R. Gomes, L.F. Terra, R.A. Wailemann, L. Labriola, M.C. Sogayar, BMC Cancer 12 (2012) 26. [37] Y. Yang, X. Pan, W. Lei, et al., Cancer Res. 66 (17) (2006) 8617–8624. [38] Y. Yang, X. Pan, W. Lei, J. Wang, J. Song, Oncogene 25 (55) (2006) 7235–7244. [39] R.J. Akhurst, R. Derynck, Trends Cell Biol. 11 (11) (2001) S44–S51. [40] H.P. Indo, M. Davidson, H.C. Yen, et al., Mitochondrion 7 (1–2) (2007) 106–118. [41] M. Edderkaoui, P. Hong, E.C. Vaquero, et al., Am. J. Physiol. Gastrointest. Liver Physiol. 289 (6) (2005) G1137–G1147. [42] H. Wiseman, B. Halliwell, Biochem. J. 313 (Pt 1) (1996) 17–29. [43] X. Pan, X. Wang, W. Lei, L. Min, Y. Yang, J. Song, Hepatology 50 (5) (2009) 1577–1587. [44] Y. Shi, S.L. Venkataraman, G.E. Dodson, A.M. Mabb, S. LeBlanc, R.S. Tibbetts, Proc. Natl. Acad. Sci. U. S. A. 101 (16) (2004) 5898–5903. [45] S. Janssens, J. Tschopp, Cell Death Differ. 13 (5) (2006) 773–784. [46] B. Kaina, S. Haas, H. Kappes, Cancer Res. 57 (13) (1997) 2721–2731. [47] O. Potapova, S. Basu, D. Mercola, N.J. Holbrook, J. Biol. Chem. 276 (30) (2001) 28546–28553. [48] Z. Guo, S. Kozlov, M.F. Lavin, M.D. Person, T.T. Paull, Science 330 (6003) (2010) 517–521. [49] D. De Cesare, G.M. Fimia, P. Sassone-Corsi, Trends Biochem. Sci. 24 (7) (1999) 281–285. [50] M. Montminy, Annu. Rev. Biochem. 66 (1997) 807–822. [51] J.C. Chrivia, R.P. Kwok, N. Lamb, M. Hagiwara, M.R. Montminy, R.H. Goodman, Nature 365 (6449) (1993) 855–859. [52] R. Eckner, M.E. Ewen, D. Newsome, et al., Genes Dev. 8 (8) (1994) 869–884. [53] G.E. Dodson, R.S. Tibbetts, J. Biol. Chem. 281 (3) (2006) 1692–1697. [54] N.P. Shanware, A.T. Trinh, L.M. Williams, R.S. Tibbetts, J. Biol. Chem. 282 (9) (2007) 6283–6291. [55] L. Sheng, Y. Zhou, Z. Chen, et al., Nat. Med. 18 (6) (2012) 943–949. [56] J. Son, J.H. Lee, H.N. Kim, H. Ha, Z.H. Lee, Biochem. Biophys. Res. Commun. 398 (2) (2010) 309–314.

Fig. 6. Neutralization of pro-inflammatory cytokines or inhibition of ROS, RNS or DNA damage abrogates MϕCM induced pCREB expression and migration. (a) Pro-inflammatory cytokines TNF-α, IL-1β or IL-6 in MϕCM was neutralized by pre-treatment with the respective antibodies (1 μg/ml) alone or in combination for 3 h. MCF7 cells were incubated with pre-treated MϕCM for 5 days. The neutralizing antibody was supplemented on day 3. The cells were labeled with pCREB antibody and visualized in a fluorescence microscope. The intensity of pCREB expression was quantified using Image J software in a minimum of 50 cells. (b) Quantitation of pCREB expression in MCF7 cells pretreated with NAC (5 mM), 1400W (10 μM) and CGK733 (2 μM) followed by treatment with MCM or MϕCM for 5 days. (c) Quantitation of pCREB expression in ATM KD cells. (d) MϕCM stimulates wound healing and migration of MCF7 cells that can be abrogated by inhibitors of ROS, RNS and DNA damage. Migration of MCF7 was assessed by plating cells (105) in the top chamber of transwell inserts that were allowed to migrate for 72 h. The bottom chamber contained MCM or MϕCM supplemented with NAC (5 mM), 1400W (10 μM) or CGK733 (2 μM). The cells on the upper side of the membrane were removed with a cotton swab. The membranes were fixed in methanol:acetone (7:3) for 20 min and stained with crystal violet for 5 min. The number of cells that migrated to the lower surface of the membrane was counted in ten randomly selected fields by light microscopy using CellProfiler software. (e) MϕCM induced migration in ATM scr and ATM KD cells. Data from one representative experiment are shown. The experiment was repeated three times. **P b 0.01. (f) Schematic representation of the interaction between macrophages and tumor cells leading to EMT responses.