Background
Stress has been described as a promoter of tumor growth and angiogenesis in different
in vivo models [
1]. Thus, it has been considered that during chronic stress and depression, the persistent activation of the hypothalamic-pituitary-adrenal (HPA) axis is probably responsible of an impaired immune response, contributing to the development and progression of several types of cancer [
2].
Corticotropin Releasing Factor (CRF) was the first peptide isolated from the now named CRF-related peptides family that also includes urocortin 1, urocortin 2 and urocortin 3. This family of peptides exerts its biological actions through the activation of two receptors: CRF receptor 1 (CRF
1) and CRF receptor 2 (CRF
2). CRF exert its effect primarily via CRF receptor 1 and at a lesser extent via CRF
2 [
3], exhibiting a 10 fold higher affinity for the former.
CRF has been described to be present not only in the central nervous system, its primary site of expression, but also in peripheral tissues and organs [
3]. Indeed, multiple studies have shown that CRF mediates endocrine responses to stress, not only by activating the HPA axis but also via direct actions in the periphery [
4‐
6]. In this regard, the CRF-based paracrine activity has been postulated to participate in the modulation of stress effects on the gastrointestinal system [
5]. Moreover, CRF-related peptides exert direct actions on cardiomyocytes mediating the adaptive response of the cardiovascular system to stressful conditions such as ischemia and reperfusion [
7,
8].
In the tumor microenvironment, CRF is released by endothelial and immune cells and by the local neuronal innervation [
9‐
11]. Moreover, peptides of the CRF family and their receptors have been also found expressed by several cancer cells [
12], such as human renal cell carcinoma [
13], tumorous adrenocortical cells [
14], human endometrial, prostate, ovarian and breast cancer cells [
14‐
19], human pheochromocytoma cells and melanomas [
20‐
22] and the murine melanoma cell line B16F10 [
23]. However, the effects exerted by CRF in cancer cells range from promotion of cancer cell proliferation and migration to inhibition of proliferation and induction of angiogenesis. Thus, CRF has been described to inhibit cell proliferation via CRF1 in the endometrial adenocarcinoma cell line Ishikawa [
24] and in the human HaCaT keratinocytes [
25]. In contrast, in the Y79 retinoblastoma cell line CRF suppresses apoptosis via downregulation of pro-caspase 3 cleavage and activation [
26] and in the B16F10 murine melanoma cell line it enhances cell migration through the ERK1/2 pathway [
23]. Moreover, in the human breast cancer MCF7 cells, an estrogen-dependent tumor cell line, CRF inhibits cell proliferation but promotes motility and invasiveness via the activation of CRF
1 [
17,
18]. In addition, CRF induces local immunosuppression by promoting apoptosis of cytotoxic T-cell via the prduction of Fas ligand (FasL) in ovarian cancer cells [
19].
The aim of the present study was to test the role of peripheral CRF as a mediator of stress response on breast cancer cell growth using both
in vivo and
in vitro studies on the 4T1 breast cancer cell line. In the first part of this work we evaluated the direct effects of CRF on this cell line in culture. In the second part, we used a mouse model of orthotropic injection of breast cancer cells in the mammary fat pad of Balb/c mice. In this model we studied the effect of stress on tumor growth and we evaluated the impact of inhibition of peripheral CRF. For this purpose we administered antalarmin intraperitoneally, which does not affect stress-induced Hypothalamus-pituitary-adrenal (HPA) axis responses [
27]. In this way, we determined the effect of peripheral CRF inhibition on tumor growth in the presence or absence of stress exposure.
Our results showed that CRF increased proliferation, migration and actin polymerization in 4T1 cells. Moreover, it modified the expression of several molecules involved in tumor growth and metastasis. Two of them, SMAD2 and β-Catenin, transcription factors connected with the TGFβ and the Wnt signaling pathways respectively [
28,
29], were increased following CRF treatment. Finally,
in vivo studies demonstrated that peripheral CRF induced angiogenesis and tumor growth
in vivo.
Discussion
The impact of stress on the development of cancer has been widely proposed [
1,
2,
36]. The stress response involves the activation of cascades in both the central and the peripheral nervous systems. CRF is the main hypothalamic stress-induced neuropeptide and its peripheral effect has also been reported in several systems [
3,
5,
8,
9,
31,
37‐
39]. Thus, the objective of this work was to analyze the role of peripheral CRF as a mediator of stress effects on cancer cells in a murine model of breast cancer. To this aim, we first analyzed the expression of CRF receptors in 4T1 cells in order to assess any direct effect of CRF on this system. In the present study we found that 4T1 cells expressed high levels of CRF
1 receptor and low levels of CRF
2b receptor. The expression of CRF receptors have been described in other cancer cell lines. In fact, previous studies from our group had shown that MCF7 breast cancer cells also express CRF
1 receptor and low levels of CRF
2. However, in the present work we found that in 4T1 cells CRF induced cell proliferation, whereas in MCF7, and others cell lines such as the adenocarcinoma cell line Ishikawa and the human HaCaT keratinocytes [
18,
24,
25], proliferation was suppressed by CRF. In contrast, CRF induced proliferation of the At20 corticotrophic adenocarcinoma cell line and primary canine corticotrophic adenoma cells [
40,
41]. This discrepancy is in accordance with previous works describing that the phenotypic effects of CRF on cell proliferation were dependent on both cell type and nutrition conditions [
42]. Therefore using non malignant cells it has been shown that CRF stimulated dermal fibroblasts proliferation while it inhibited cell proliferation in keratinocytes [
42].
Since different reports support positive or negative actions of CRF on cancer cell growth and metastasis we assessed the effect of CRF on the expression pattern of genes involved in cancer cell metastasis. For this purpose we used gene-specific oligo microarrays. Our results demonstrated that CRF treatment increased expression of Smad2 and β-catenin, and suppressed the expression of the angiogenesis inhibitor Bai1, the metastasis suppressor Brms1 and the cell cycle regulator Cdkn2a/p16. In addition, CRF also enhanced the expression of molecules involved in cell cycle, proliferation and apoptosis, such as Ha-ras1, Myb, Pten, Rb1 and RhoC. Our studies focused on the impact of CRF on SMAD2 and β catenin, being molecules involved in two central signaling pathways regulating breast cancer growth and metastasis, these of TGFβ and Wnt respectively [
28,
29].
We therefore confirmed the effects of CRF on SMAD2 and β-catenin expression at protein levels. SMAD2 and β-catenin are two major transcription factors involved in metastasis. SMAD2, together with SMAD3, is associated with the TGF-β receptor. When TGF-β binds to its receptor, SMAD2 and SMAD3 are phosphorylated and form a complex with SMAD4 that translocates to the nucleus. In the nucleus, an activated SMAD complex is formed which regulates gene expression and ultimately cell growth [
28]. Regarding β-catenin, apart from being a cell-cell adhesion protein, is also an important signal transduction molecule in the Wnt signaling pathway [
43]. Induction of Wnt signaling, mostly by affecting β-catenin, has been described as a hallmark of colon, breast, prostate and ovarian cancer [
29]. Interestingly, recent evidence described a link between the TGF-β and the Wnt-signaling pathways, since receptor-activated SMAD2 synergistically enhances the Wnt/β-catenin pathway in epithelial cancer cells [
44]. Thus, the potential effect of CRF on SMAD2 and β-catenin, and subsequently TGF-β and Wnt signaling, may confer a novel mechanism for crosstalk between cancer cells and stress neuropeptides.
Moreover, it has been reported that TGF-β promotes cell motility and invasiveness in epithelial cancer cells. In addition, β-catenin is also involved in cytoskeletal changes characterized by actin polymerization, cell adhesion and motility [
30]. Therefore, we analyzed the effect of CRF on actin polymerization in 4T1 cells. Our results showed higher levels of polymerized actin as well as an increase of actin stress fibers. This suggests that CRF could promote changes in cytoskeletal structures that allow cells to migrate and metastasize [
45].
The results of the present and our earlier study [
18] suggest distinct effects of CRF on breast cancer cells. Several reports have indicated either tumor promoting or tumor-inhibitory effects of neuropeptides. Oxytocin has been shown to suppress proliferation while ghrelin promotes proliferation in breast cancer cell lines [
46,
47]. Moreover, the effect of grelin on the phenotype depends on the expression of Estrogen Receptor [
47,
48]. In our case both MCF7 and 4T1 are ER+ cell lines suggesting that the discrepancy of the effects does not depend on ER but on other genetic differences.
Considering the results obtained
in vitro, and our earlier studies in the human breast cancer cell line MCF7 [
18], we used an
in vivo model of breast cancer in which we exposed mice to chronic stress (see Materials and methods). Antalarmin was administered intraperitoneally and did not affect chronic stress-induced corticosterone levels ([
27] and present report) but was able to inhibit its action on tumor cells. Indeed, earlier studies showed that intraperitoneal administration of antalarmin inhibited the proinflammatory role of CRF in toxin A-induced intestinal secretion and inflammation [
49] or in the adjuvant induced arthritis model with Lewis rats [
50]. Furthermore, inhibition of peripheral CRF with i.p. administration of antalarmin resulted in an increased survival after LPS-induced endotoxic shock, without affecting the production of corticosterone [
34]. Accordingly, our results showed that administration of antalarmin intraperitoneally did not affect the elevation of corticosterone following stress exposure. Once confirmed that in our system the HPA axis was not affected, we analyzed the effects of peripheral CRF inhibition on tumor growth. We observed that i.p. administration of antalarmin in stressed animals resulted in significant reduction of tumor burden, which suggests that peripheral CRF promoted the growth or tumor cells also
in vivo. Moreover, we quantitatively evaluated the extent of neoangiogenesis in the 4T1 tumors, as an essential process for the tumor growth and metastasis. Histological analysis did not reveal any other changes in the tumors, such as apoptotic/necrotic lesions. Our experiments showed that treatment of mice exposed to stress with antalarmin resulted in reduced angiogenesis compared to stressed mice injected with vehicle. This suggests that peripheral CRF significantly contributes to neoangiogenesis observed after stress. Moreover, our results illustrated that this effect of peripheral CRF is exerted via CRF receptor 1, since it was inhibited by the selective CRF
1 antagonist antalarmin. Interestingly, previous reports have shown a suppressive effect of Urocortin2 on tumor vascularization via CRF receptor 2 [
51,
52] and depletion of CRF1 in mice suppresses intestinal angiogenesis while ablation of CRF
2 augments it, supporting a role for CRF
1 signals in angiogenesis [
35]. Also, peripheral CRF has been shown to enhance local angiogenesis and vascular permeability in skin via a CRF receptor-dependent mechanism [
10,
53]. This indicates that different CRF receptors may have different effects on neoangiogenesis. Expression of Cox-2 and VEGF have been associated with neoangiogenesis. In the case of 4T1 cells CRF induced Cox-2 but not VEGF expression suggesting that it utilizes a Cox2-dependent, VEGF-independent mechanism to promote angiogenesis.
Experimental Procedures
Cell Culture
The mouse mammary tumor cell line 4T1 was cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 1% penicillin/streptomycin (all purchased from GIBCO, UK) at 37°C in a 5% CO2 humidified atmosphere. For cell stimulations, 4T1 cells were plated one day before stimulation at 500.000 or 250.000 cells per well in 6-well or 24-well plates, respectively. Subsequently, medium was refreshed and supplemented with (or without for controls) synthetic rat/human CRF (Tocris, Bioscience, UK) at a concentration of 10-8M.
Reverse transcriptase PCR
Total cellular RNA was isolated using Trizol reagent (Invitrogen, UK). cDNA was prepared by reverse transcriptase PCR (Superscript RT, Invitrogen) and amplified by PCR using the following primer pairs: CRFR1: fwd 5'- GCC GCC TAC AAT TAC TTC CA-3', rev 5'- CGG AGT TTG GTC ATG AGG AT - 3' and CRFR2: fwd 5'- GGA GCC CTA GTG GAG AGA CC -3', rev 5'- AGG TGG TGA TGA GGT TCC AG -3', VEGF: 5'- GTACCTCCACCATGCCAAGT-3', 5'- ACTCCAGGGCTTCATCGTTA -3', Cox-2: 5'- GCTTGCATTGATGGTGGCTG-3', 5'- CCAGATGCTATCTTTGGGGAGAC-3'. For each PCR reaction, 1 μl of cDNA was used together with primers indicated above, at 45 cycles and an annealing temperature of 60°C. 10 μl of amplified products were separated on a 1.5% agarose gel and visualized by ethidium bromide staining, or subjected to real-time PCR using SYBRgeen method as previously reported [
54].
MTT viability assay
To determine the effect of CRF on cell proliferation, MTT (3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide) cell viability assays were performed. 4T1 cells were plated in flat-bottomed 96-well plates at a 5000 cells/well concentration and allowed to adhere overnight. The following day the medium was changed by fresh DMEM supplemented or not with 10-8 M CRF. To determine cell growth, after 24 and 48 hours, 50 μg MTT (Sigma-Aldrich, USA) was added to each well and the plates were incubated an additional 4 h at 37°C and 5% CO2. After 4 h the supernatant was removed and the formed crystals were dissolved in 100 μl 0,04N HCL in isopropanol. The plates were analyzed at 570 nm with a microplate reader (Bio-Rad, UK). All assays were performed in quadruplicate and the mean values for each data point was calculated from the combined data.
Gene expression arrays
Total RNA from 4T1 cells was isolated using Trizol reagent (Invitrogen), following the manufacturer's recommendations. Using the TrueLabeling-AMP™ 2.0 kit, (Superarray Bioscience Corp., Frederick, Md.), the RNA was reversely transcribed to obtain cDNA and converted into biotin-labeled cRNA using biotin-16-UTP (Roche, Mannheim, Germany) by in vitro transcription. cRNA probes were then purified with the ArrayGrade cRNA cleanup kit (Superarray) and hybridized to the pretreated Oligo GEArray® Mouse Tumor metastasis microarray (Superarray). Following washing steps, array spots binding cRNA were detected using alkaline phosphatase-conjugated streptavidin and CDP-Star as chemiluminescent substrate. Signal was detected by exposure to high-performance chemiluminescence films (Amersham Biosciences, UK). The image data were transformed into numerical data using GEArray Expression Analysis Suite software (SuperArray Bioscience). To normalize the data, background signal was subtracted and the intensity of all genes was referred to GAPDH as an endogenous control. Data filtering criteria were as follows: at least one of the spot intensities to be compared had to be more than twice the background intensity, and the spot intensity ratios had to be ± 40% in all set of samples analyzed to consider up or down-regulation. Finally, mean expression of each gene was transformed into color intensity using the program TIGRMultiExperiment Viewer V4.5.1.
Western blot analysis
Western blot analysis for the detection of SMAD2 (sc-6200, Santa Cruz, US) and β-catenin (sc-7199, Santa Cruz) were performed. After treatment, cells were harvested and lysed in buffer, containing 1.5 mM Tris (Bio-Rad Labs), 150 mM NaCL, 0.1% SDS, 1% NP-40, 0.02% Sodium Azide, pH 8, with proteinase inhibitors 4% complete and 1% PMSF (phenylmethylsulfonyl fluoride) (all from Sigma) as previously described [
55]. Cell lysates were sonicated for 4 seconds and solid cellular debris were removed by centrifugation at 12.000 rpm for 10 min. Lystates were stored at -80°C until use. 20 ug of lysate was loaded in a 12% SDS-polyacrylamide gel, transferred to nitrocellulose membranes and processed according to standard Western blotting procedures. To normalize for protein content the blots were stripped in buffer containing 62.5 mM Tris-HCl, pH 6.7, 2% SDS, 100 mM β-mercaptoethanol and stained with anti-tubulin antibody (T4026, Sigma). The concentration of each target protein was normalized
vs tubulin. NIH image software (ImageJ) was used to quantify the intensity of each band.
Immunofluorescence
4T1 cells were cultured at a concentration of 30.000 cells per well in 8-well chamber slides. After 24 h fresh medium supplemented with CRF at a concentration of 2 × 10-8M was added. After 2 or 4 h cells were fixed with 3.7% formaldehyde in PBS for 10 min, permeabilized with acetone for 4 min, washed with PBS and blocked with 1.5% FCS in PBS for 15 min. The chamber slides were subsequently incubated with rhodamine-phalloidin (Sigma) at a 1:100 dilution in 1.5% FCS in PBS, for 30 min at dark. Cells probed with rhodamine-phalloidin were washed with PBS and immediately mounted and stored at -20°C until observation with confocal laser scanning microscopy.
Wound healing assay
Cells were cultured in 60 mm plates until they fromed a monolayer. A small area was then disrupted and a group of cells was destroyed or displaced by scratching a line through the layer with a tip [
18]. The culture medium was replaced with serum free medium and cells received the stimulus (CRF and/or Antalarmin). The open gap was then inspected microscopically (Leica, Germany) over time as the cells moved in and filled the damaged area. Images were captured at the beginning and at regular time points during cell migration and the cell migration was quantified by measuring the distance with the program Image J
http://rsbweb.nih.gov/ij/ between two certain points on either side of the gap. For proper statistical evaluation, at least three measurements at different points were performed at each image.
Transfections and luciferase assay
4T1 cells were transfected with a pGL3 plasmid containing a WNT-reporter sequence harbouring 3 TCF binding sites linked to the luciferase by lipofectamine™ 2000 (Invitrogen), according to the manufacturer's instructions. In parallel experiments, the same plasmid was used that carries a point mutation in the TCF binding site. In brief, one day before transfection, cells were plated at 1,5 × 105 cells/well in 500 μl medium in 24-well plates. Plasmid DNA and lipofectamine both diluted in Opti-MEM I reduced serum medium without serum (Invitrogen), were mixed at a 1:2 ratio (1 μg DNA: 2 μl lipofectamine/well) and incubated for 20 min at room temperature. After 20 min incubation, 100 μl of plasmid/liposome complex was added to each well, and cells were incubated for 24 hours at 37°C and 5% CO2. Medium was refreshed after 4 hours with normal culture medium containing 10-8 M CRF, cells were lysed after 18 hours. Luciferase assay was performed with the Dual-Luciferase® Reporter Assay System (Promega), according to the manufacturer's manual. Each transfection was performed in triplicate to allow statistical evaluation and control for possible variations in transfection efficiency.
RIA
Corticosterone was measured by RIA in serum collected at the indicated time points. Five animals per group were used. Sera were frozen at -70°C and analyzed as recommended by the manufacturer (ICN, USA).
Animals
Six to eight weeks old Balb/c female mice were purchased from the Hellenic Pasteur Institute (Athens, Greece). All procedures described below were approved by the Animal Care Committee of the University of Crete School of Medicine, Heraklio, Crete, Greece and from the Veterinary Department of the Heraklion Prefecture, Heraklio, Crete, Greece.
In vivo model of breast tumor and restraint stress
One million 4T1 cells were implanted in the mammary fat pad of Balb/c mice and three different groups were created. One group was injected intraperitoneally with 20 mg/kg antalarmin everyday. A second group was exposed to restraint stress for 3 hours for 4 consecutive days following a 5 day interval. The third group was exposed to the same type of restraint stress and in addition received 20 mg/kg antalarmin daily, dissolved in cremaphor (Sigma). Control group was subjected to the same surgical procedure, without the implantation of tumor cells.
We used antalarmin for inhibition of CRF receptors since CRF functions primarily via CRF1, the target of antalarmin. Mice that did not receive antalarmin received an injection of vehicle at the same time points. At different time points (either before or immediately after the restraint period) samples were collected by the retroorbital route to measure corticosterone in the plasma. The experiment was terminated 6 weeks later. At the end of the experiment mammary glands were visualized on the animal to determine the extent of neoangiogenesis and samples were collected from the different groups and histological analysis was performed.
Angiogenesis determination
To quantitatively measure angiogenesis an in-house developed method was used. Briefly, this method utilized the contrast of autofluorescence between the mammary tissue and the blood vessels. Blood vessels greatly absorb visible light, while mammary gland and mammary tumors are strongly autofluorescent. User-friendly software developed in-house was used in order to quantify the area of vascularization vs total area.
Histological analysis
Mammary glands samples were collected as specified above and fixed in formalin. Sections were stained with Haematoxylin-Eosin using standard techniques. Presence of tumors was determined by the same pathologist blinded to the treatment conditions. Percentage of tumor bearing mice was calculated per each group.
Statistical analysis
Comparison between groups was made using the Student's t-test and ANOVA test, and p < 0.05 was considered significant.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AAr performed the in vivo studies, the gene expression array analysis, SMAD and β-catenin expression studies, the angiogenesis studies and drafted the manuscript. MV participated in the in vivo experiments, performed corticosterone measurements, participated in data evaluation, co-ordination of the study and manuscript preparation. BM performed SMAD experession analysis, cell proliferation studies and confocal studies. ED participated in confocal microscopy analyses and co-ordination of the study. AAn participated in the in vivo studies data evaluation and co-ordination of the study. OR participated in the cell proliferation studies and receptor expression analysis. ENS participated in the histological analysis and evaluation. RG participated in the array analysis and drafting the manuscript. JR designed the software and participated in the angiogenesis analysis. ANM participated in the design of the study and drafting of the manuscript. CT conceived of the study, participated in its design and drafted the manuscript. All authors read and approved the final manuscript.