Background
Recurrences at metastatic sites represent a major cause of mortality in breast cancer patients [
1,
2]. It is reported that 20–45 % of breast cancer patients will relapse years or even decades after the resection of the primary tumor [
3]. Only a small number of the disseminated tumor cells that lodge in secondary organs will eventually grow to form a clinically evident metastasis; cancer cells can remain dormant in secondary organs for years [
4,
5]. The existence of such dormant cancer cells at metastatic sites has been described previously as quiescent solitary cells that neither proliferate nor undergo apoptosis [
1,
6]. Therefore, a comprehensive understanding of the “switch” from a dormant to a growth state is crucial to elucidate the mechanism of cancer progression and recurrence, might lead to the development of novel treatments for cancer metastasis.
The cancer-associated Epithelial-to-Mesenchymal Transition (EMT) has been strongly correlated with metastasis and shortened life expectancy of many carcinomas, has been proposed as a mechanism for enabling cancer cell invasion and dissemination [
7,
8]. Nevertheless, EMT is reversible and that a reversion back towards the epithelial phenotype may occur at the secondary metastatic site (MErT) [
9]. Current opinion and our previous studies revealed that metastatic breast cancer dormancy is likely not sustainable by the invasive, mesenchymal phenotype but rather through a partial epithelial reversion in which the cells are in a quiescent state [
9,
10]. Accumulating evidence suggests that MErT may be critical for breast cancer ectopic survival and dormancy once a distant metastasis is involved. Furthermore, studies have shown that a secondary epithelial to mesenchymal transition is thought to underlie latent metastatic outgrowth [
10‐
12]. Therefore, EMT and MErT may determine dormant or active states of the tumour, respectively, and allow for an indeterminate number of metastases formation.
It is established that distant metastases involves disseminated tumor cells adapting to the foreign environment, suggesting that the microenvironment is capable of regulating a series of switches between EMT and MErT phenotypes [
13‐
15]. The triggers for the second mesenchymal transition of these dormant cells are not known though initial studies in an ex vivo microphysiological system suggest that inflammatory signals may underlie this [
16,
17]. In breast cancer stroma, key cells of the innate inflammatory process, macrophages, can occupy more than 50 % of the breast tumour mass and influence breast cancer prognosis [
18,
19]. Macrophages are heterogeneous in population and can be classified within a spectrum of M1 or M2, polarising to each dependent on the stimuli present at time of activation. Recently, it has been shown that tumor-associated macrophages (TAMs), which are characterized by M2 macrophages, contribute to EMT and cancer metastasis from primary tumor to a distant tissue [
20,
21]. However, the roles of macrophages on modulating the balance between EMT and MErT of breast cancer cells in response to a metastatic microenvironment remain unclear.
The goal of this study was to determine the functional contributions of M1/M2 macrophages to epithelial-mesenchymal plasticity in breast cancer cells, and to elucidate the underlying effects of macrophage polarization on tumor dormancy or growth state for emergency in metastatic tumors. This work might highlight a novel function of macrophage for colonization in the distant metastatic site, and serve as a foundation to explore mechanisms to either maintain metastatic dormancy or induce emergence.
Methods
Antibodies and reagents
Phorbol 12-myristate 13-acetate (PMA), Lipopolysaccharides (LPS) and Hoechst 33258 dye were from Sigma-Aldrich (St. Louis, MO, USA). The cytokines for interferon γ (IFNγ), interleukin-4 (IL-4) and interleukin-13 (IL-13) were from R&D Systems (Minneapolis, MN, USA). The antibody for CD163 was from Abcam (Cambridge, MA, USA). The primary antibodies for E-cadherin, GAPDH were from Cell Signaling Technology (Danvers, MA, USA). Antibodies for fluorescein isothiocyanate (FITC)-conjugated CD68, FITC-conjugated CD206, FITC-conjugated CD163, FITC-conjugated HLA-DR and isotype control were from BD Biosciences (San Jose, CA, USA). FITC-conjugated E-cadherin antibody was from BioLegend (San Diego, CA, USA). Hepatocyte maintenance medium (HMM) and SingleQuots were from Lonza (Anaheim, CA, USA).
Cell culture and differentiation
The breast cancer cell lines MCF7, MDA-MB-231 and the human THP-1 cell line were purchased from ATCC (Manassas, VA, USA). MDA-MB-231 and MCF7 cell lines were transfected with red fluorescent protein (RFP) as previously described [
9]. To maintain selection for RFP positive breast cancer cells, MCF-7 cells were cultured with 900 μg/ml G418, and MDA-MB-231 were cultured with 5 μg/ml puromycin in RPMI-1640 (Life Technologies, Carslbad, CA) supplemented with 10 % FBS until used in the experiments. THP-1 cells were cultured in RPMI 1640 supplemented with 10 % FBS, 100 mg/ml penicillin, and 100 mg/ml streptomycin (Gibco, Life Technologies, Grand Island, NY, USA).
THP-1 cells were differentiated and polarized according to established protocols [20, 27], with minor modifications. Briefly, THP-1 cells were plated in RPMI media and treated with 320 nM PMA for 72 h, cells that adopted an adherent macrophage-like phenotype were selected as M0 macrophage. To generate M1-polarized macrophages, THP-1 cells were treated with 320 nM PMA for 6 h and then cultured with PMA plus 100 ng/ml LPS and 20 ng/ml IFNγ for the remaining 66 h of a total incubation time of 72 h. To generate M2-polarized macrophages, THP-1 cells were treated with 320 nM PMA for 6 h, and then cultured with PMA plus 20 ng/ml IL-4 and 20 ng/ml IL-13 for another 66 h of a total incubation time of 72 h.
Peripheral blood mononuclear cells (PBMCs) were obtained using a previously described method [
22,
23]. Briefly, we mixed the human peripheral blood sample with an equal volume of room temperature PBS, pelleted the leukocyte/RBC fraction by centrifuging the cells 15 min at 200 ×
g, at room temperature. The supernatant was removed and the leukocyte/RBC cell suspension brought to a final volume of 40 ml with 1× PBS. The 10 ml leukocyte/RBC/PBS mixture was layered over 3 ml Ficoll-Hypaque (10771 SIGMA) solution and centrifuged 20 to 30 min 900 g, 18° to 20 °C, with no brake. The mononuclear lymphocyte cell were washed by adding HBSS (∼3 times the volume of the mononuclear cell layer) and centrifuging 10 min at 450 to 600 ×
g, 18° to 20 °C. Mononuclear cells were resuspended in RPMI-1640 medium and seeded on poly-L-Lysine coated coverslips. Media was changed after 4 h to remove un-attached cells. For polarization we added 50 ng/ml GM-CSF for M1 or 50 ng/ml M-CSF for M2, and replenished after 48 h for total 96 h. Media was then supplemented with LPS (100 ng/ml) and IFNγ (20 ng/ml) for M1 or IL-4 (20 ng/ml) and IL-13 (20 ng/ml) for M2 for 72 h. We seeded 200,000 RFP-labeled MCF-7 or MDA-MB-231 onto the macrophages monolayer in HMM medium after thoroughly wash to remove cytokines. Cells were fixed after 5 days.
Macrophage and breast cancer cell co-culture
One million THP-1 cells, or primary human monocyte-derived macrophages, were plated in 6-well plates in 2-mL RPMI1640 with 10 % FBS plus PMA and M1-trophic (IFN-γ and LPS) or M2-trophic (IL-4 and IL-13) cytokines treatment for 72 h to induce differentiation into M1/M2 macrophages. M0 macrophages, cells treated with only PMA were taken as controls. After a thorough wash to remove PMA and cytokines, 2 × 105 per well MCF7 or MDA-MB-231 cells were seeded onto the M1/M2, M0 macrophages monolayer and co-cultured in HMM medium for additional 5 days. Breast cancer cell lines were alone planted in HMM medium as the control group.
In addition, the M0, M1 and M2- conditioned media (CM) was collected everyday for breast cancer cells treatment. The culture supernatants from macrophages were collected 24 h later, by centrifuging at 500 × g for 5 min, as the macrophage CM. After 24 h, the normal growth medium RPMI1640 was discarded, MCF-7 or MDA-231 cells were treatment M0, M1 or M2 CM was added totally for 5 days, and changed M0, M1 or M2 CM everyday.
Co-culture systems were established by using transwell inserts (0.4 mm pore, polycarbonate membrane; Costar, Cambridge, MA, USA) and transferred to six-well culture plates. M0, M1 or M2 macrophages (1 × 106 cells) were loaded in the upper inserts, and MCF7 or MDA-MB-231 cells (2 × 105 cells) were put into the lower compartment of the culture well and serum-starved for at least 12 h before co-culture. Macrophages co-cultures with MCF-7 or MDA-MB-231 cells were performed in serum free HMM for 5 days. Total cellular protein was extracted from MCF7 or MDA-MB-231 cells exposed to different experimental conditions.
Flow cytometry
Phenotypic changes in macrophages were verified by flow cytometry using the BD LSRFortessa flow cytometer (BD Biosciences, Franklin Lakes, NJ). After blocking human FcRs, cells were washed and resuspended in PBS supplemented with 1 % heat-inactivated FBS and 0.01 % NaN3. For CD68 staining, cells were fixed and permeabilized with a BD Cytofix/Cytoperm Fixation/Permeabilization Solution Kit (BD Biosciences, Franklin Lakes, NJ). Cells were then incubated with the FITC-CD68 antibody. For surface markers, cells were incubated with FITC-coupled antibodies to CD206, CD163, HLA-DR, as well as isotype matched control antibodies, for 30 min at 4 °C in the dark. Flow cytometry analyses of co-cultures were accomplished by non-enzymatic dissociation of cells from the culture plates. The cells were fixed in 2 % Paraformaldehyde for 30 min, permeabilized with 1 % Triton for 3 min, and incubated with a FITC-conjugated E-cadherin antibody for 30 min. Following the final washing step, labeled cells were analyzed by flow cytometry using BD FACSDiva Software (BD Biosciences, Franklin Lakes, NJ). The mixed macrophage-breast cancer cell suspension was gated as to identify breast cancer cells using RFP fluorescence versus side scatter. Gates were set on the RFP+ population, and this gated population was analyzed for E-Cadherin expression. The RFP- population was analyzed for macrophage marker expression.
Imaging
Phase contrast images of whole cell morphology were captured by an Olympus inverted scope and digitally analyzed using Spot Advanced software (Diagnostics Instruments, Macomb, MI). Immunofluorescent staining of co-cultures was conducted to evaluate E-cadherin expression. Cell slides were incubated with the primary antibody for E-cadherin (1: 500) at 4 °C overnight and then with Alexa Fluo® 488-conjugated secondary antibody at room temperature for 1 h. Counterstains used were 0.1 % Hoechst 33258 dye for nuclei. Immunofluorescent images were captured on an Olympus Fluoview 1000 scope (Olympus, Center Valley, PA) and captured using Fluoview Viewer.
Images from co-cultures in HMM versus MCF-7 or MDA-MB-231 in RPMI growth medium were imported into ImageJ Version 1.44i (U. S. National Institutes of Health, Bethesda, Maryland). Two fields per experimental condition were analyzed. Cell perimeters and midpoint widths were manually traced and measured in pixel units using the ImageJ functions. The ratio of width versus perimeter was computed for each cell and the mean values, standard deviations, and Student’s t test (2 tailed) were calculated using Prism 5 (GraphPad Software, San Diego, CA, USA).
Breast cancer cells viability was assessed as the quantity of RFP+ cells to the total cell number of co-cultures. Manual calculations of the RFP+ cells within 10 random fields with approximately 500–1000 cells were performed.
Xenografts
Metastatic samples in mice: Prostate cancer (PCa) DU145 cells were obtained from ATCC, cultured in RPMI-1640 media supplemented with 10 % fetal bovine serum (FBS) in the presence of 100 u/mL penicillin and 0.1 mg/mL streptomycin. Seven-week-old male NOD/SCID gamma mice were purchased from the Jackson Laboratories (Bar Harbor, ME). For the in vivo hepatic metastasis model cells were injected into the spleen and both liver and spleen examined. After anesthetizing the mice, a transverse incision in the left flank was made to expose the spleen; a half a million viable PCa DU145 cells were injected into the spleen using a 27-gauge needle. Five weeks after inoculation, the tumor xenograft tissues of NOD/SCID gamma mice were fixed in 10 % neutral buffered formalin for immunohistochemical staining. The Mouse on Mouse Kit (Vector Labs, Berlingame, CA), was used for positive labeling by comparing serial sections incubated with the primary antibodies for E-cadherin or CD163, and the biotinylated secondary antibody alone. Labeling was visualized with the Vectastain Elite kit (Vector Labs). The AAALAC-accredited Institutional Animal Care and Use Committees of the Veteran’s Administration Pittsburgh Health System approved all animal studies and procedures.
Western blot
Cells were lysed in SDS buffer (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 5 mM EDTA, 5 mM EGTA and 1 % SDS) containing complete protease inhibitor cocktail (Roche). Protein concentration was determined by the Bradford protein assay and gels loaded with equal amounts of protein per lane. A 8 % SDS–PAGE gel resolved cell lysates and were subsequently transferred to a PVDF membrane. Membranes were blocked with 5 % serum albumin for 1 h and incubated overnight with primary antibody E-cadherin or GAPDH. The membranes were washed and incubated with peroxidase-conjugated secondary antibodies for 1 h, and then detected by enhanced chemiluminescence.
Statistical analysis
Data analysis involved use of GraphPad software (GraphPad Prism version 5.00 for Windows, GraphPad Software, San Diego, CA, USA). Results are expressed as mean ± s.e.m. Differences were analyzed by t-test or ANOVA, and results were considered significant at P <0.05.
Discussion
In this study, we established in vitro models based on co-culture of breast cancer cells with macrophages in a liver microenvironment medium so as to investigate possible roles in regulating breast cancer metastasis dormancy and outgrowth. Using THP-1 cells (a human monocytic cell line) and a protocol we devised for their differentiation and polarization into M1- and M2-like populations, we found that dormancy and emergence from dormancy may hinge upon the M1 or M2 type of the macrophage polarization in the liver metastatic microenvironment, thus making the metastatic niche a potential therapeutic target to combat breast cancer metastasis.
The formation of micrometastases and the processes by which micrometastases progress to macrometastases are the main rate-limiting steps in clinically evident metastases [
26]. Our foundational model posits that EMT and its accompanying reduction in E-cadherin expression enables carcinoma cells to disseminate from the primary tumor, while MErT with upregulation of E-cadherin expression allows disseminated carcinoma cells to integrate and survive within ectopic metastatic microenvironments [
9,
12,
25,
27]. Outgrowth of these silent micrometastases is accompanied by a second EMT with subsequent formation of secondary macroscopic metastases [
11,
28]. In addition to their role in constructing connections, the formation of cell heterotypic E-cadherin adhesions in the metastatic target organ, MErT may result in dormancy and enable the tumor cells to survive at a lower metabolic load at the micrometastasis stage [
29,
30]. Thus the control of the epithelial-mesenchymal plasticity in breast cancer cells in secondary organs remains key to this mortal development. Therefore, we focused on the mechanism by which breast cancer cells regulate EMT/MErT balance in liver microenvironment.
Diverse interactions between the host microenvironment and cancer cells determine the course of tumor progression and metastasis. Macrophages are a major component of the breast cancer microenvironment. The human THP-1 cell line is a widely used monocyte/macrophage model, and the biologic behavior of THP-1 cells resembles monocyte-derived macrophages with respect to adherence, phagocytic capacity, and surface marker and cytokine expression [
31,
32]. Therefore, we used the THP-1 monocyte cell line as a model for macrophage differentiation and polarization, and one with which to assess the effects of macrophages on epithelial-mesenchymal plasticity in breast cancer cells. The phenotypic state of the tumor cells is key to the fate of the micrometastases and also the chemosensitivity thereof [
9,
25,
33].
Experimental and clinical data support the notion that the survivors persisting in metastatic sites can exist in at least three alternative states: (i) as solitary viable cancer cells in a quiescent, nonproliferative state (dormancy); (ii) as micrometastases, which remain as small, at least quasi-quiescent lesions; or (iii) as actively growing macrometastatic lesions [
34]. Thus, the critical question as to whether the micrometastatic cells are in a state of quiescence or outgrowth is the key to dealing with tumor dormancy, and should be at the forefront of new approaches to tumor management. We previously reported that perturbations of the parenchymal hepatocytes and NPCs in the liver metastatic microenvironment may differentially contribute to metastatic dormancy, stability, or emergence [
12].
In examining NPC subsets, tissue macrophages, including Kupffer cells, stand out. Macrophages represent one of the main stromal cell populations, but are diverse as they take on various functional phenotypes, referred to as M1 and M2 phenotypes. The role of macrophages in tumor progression has been shown to be double-edged. Pro-inflammatory, or classically activated M1 macrophages, exert resistance against tumors and elicit tissue disruptive reactions, while anti-inflammatory, or alternatively activated M2 macrophages, appear to have important roles in promoting tumor growth and metastasis [
35‐
39]. But these earlier works all looked at the primary cancer situation. The actions in the metastatic site may be different if not diametric. Our study demonstrates that the epithelial, non-metastatic human immortalized breast cancer cell line MCF-7 is exacerbated by M2 macrophage co-culture, whereas the highly invasive and mesenchymal MDA-MB-231 cell outgrowth is attenuated by M1 macrophage co-culture. This provided evidence that the M2 macrophages provided a favorable outgrowth environment, but M1 macrophages served as surrogate dormant for the breast cancer cells in this culture system.
The metastatic microenvironment is composed of a complex milieu of external cues arising from the tumor, stroma, and parenchymal cells. It has been reported that stromal cells regulate proliferation and motility of cancer cells through both soluble factors and direct cell-cell interactions [
40]. In the present study, it was suggested that cell-cell contact was required for macrophage phenotypic subtypes to regulate epithelial-mesenchymal plasticity in breast cancer cells. Recent reports have shown that direct cell-cell contact between M2 macrophages and cancer cells induced migration and invasion of cancer cells while M1 macrophages reduced cancer cell invasion [
41]. However, the numerous cell surface and extracellular matrix (ECM) molecules potentially involved in cell-cell interaction are not identified yet, nor have the molecular mechanisms by which these changes actively induce or repress gene expression in malignant cells. In fact, two-dimensional (2D) culture did not recapitulate many of the complex properties of the three-dimensional (3D) in vivo microenvironment. One main issue is the stiffness of the supporting substratum. Pathological, and in tissue culture supra-physiological stiffness is well known to lead to a mesenchymal transition of even non-neoplastic breast epithelial cells [
42,
43]. Our model system for the hepatic niche moves beyond organotypic modeling into microphysiologic systems, providing the mechanical stresses and oxygenation that is so important to the liver [
17,
44]. As a part of our research we are continuing to explore the behaviors of cancer cells in the metastatic microenvironment to lie dormant or aggressively grow out through connection of this microphysiologic system with immune system activation.
There are several possible outcomes after cancer cells extravasate into a metastatic target tissue: apoptosis, dormancy, or sustained proliferation [
45]. Epithelial-mesenchymal plasticity shifts may influence the discontinuous behavior of cancers, in which some cancers remain dormant for years after therapy, or to relapse and wreak havoc [
46]. The formation of cell heterotypic E-cadherin adhesions in the metastatic target organ may result in dormancy at the micrometastasis stage. After this quiescence, a secondary insult, independent of carcinoma per se, to the local environment may induces renewed carcinoma cell proliferation and escape from E-cadherin-mediated contact inhibition during the metastatic seeding of disseminated carcinomas [
29]. In this study, we report that macrophage phenotypic subtypes diametrically regulate epithelial-mesenchymal plasticity of breast cancer cells. Hence, epithelial-mesenchymal plasticity may be appropriate as a metastasis outgrowth prevention strategies in early stage carcinomas. Inducing the MErT program in dormant micrometastases by M1 macrophages would be a novel approach to maintain the long latency of a dormant stage, and prevent occurrence of late disease recurrence. Furthermore, since an increasing number of studies suggest a role of EMT in promoting chemoresistance [
47], combining chemotherapies with M2 macrophages deletion holds promise to overcome chemoresistance in dormant tumor cells, thus providing a unique therapeutic approach to eradicate dormant tumor cells. In the near future, improving our understanding of the molecular regulation of the dynamic EMT/MErT programs during tumor metastasis will help to uncover new signaling pathways that can be therapeutically manipulated to either eliminate dormant tumor cells or to indefinitely maintain them in this dormant state, thus preventing a progressive metastatic disease.
Abbreviations
CM, conditioned media; EMT, epithelial to mesenchymal transition; HMM, hepatocyte maintenance medium; IFN, interferon; IL, interleukin; LPS, lipopolysaccharides; M1, macrophage type 1; M2, macrophage type 2; MErT, mesenchymal to epithelial reverting transition; PMA, phorbol 12-myristate 13-acetate; TAMs, tumor-associated macrophages
Acknowledgments
This work was supported by grants from a VA Merit BLRD Award and NIH NCATS and NCI (UH3TR000496). The authors thank members of Wells laboratory for discussions and suggestions.