Proteinase K

Establishment of a heterotypic 3D culture system to evaluate the interaction of TREG lymphocytes and NK cells with breast cancer

Tanya N. Augustine a,⁎, Thérèse Dix-Peek b, Raquel Duarte b, Geoffrey P. Candy b

Abstract

Three-dimensional (3D) culture approaches to investigate breast tumour progression are yielding information more reminiscent of the in vivo microenvironment. We have established a 3D Matrigel system to determine the interactions of luminal phenotype MCF-7 cells and basal phenotype MDA-MB-231 cells with regulatory T lymphocytes and Natural Killer cells. Immune cells were isolated from peripheral blood using magnetic cell sorting and their phenotype validated using flow cytometry both before and after activation with IL-2 and phytohaemagglutinin. Following the establishment of the heterotypic culture system, tumour cells displayed morphologies and cell–cell associations distinct to that observed in 2D monolayer cultures, and associated with tissue remodelling and invasion processes. We found that the level of CCL4 secretion was influenced by breast cancer phenotype and immune stimulation. We further established that for RNA extraction, the use of proteinase K in conjunction with the Qiagen RNeasy Mini Kit and only off-column DNA digestion gave the best RNA yield, purity and integrity. We also investigated the efficacy of the culture system for immunolocalisation of the biomarkers oestrogen receptor-α and the glycoprotein mucin 1 in luminal phenotype breast cancer cells; and epidermal growth factor receptor in basal phenotype breast cancer cells, in formalin-fixed, paraffinwax embedded cultures. The expression of these markers was shown to vary under immune mediation. We thus demonstrate the feasibility of using this co-culture system for downstream applications including cytokine analysis, immunolocalisation of tumour biomarkers on serial sections and RNA extraction in accordance with MIQE guidelines.

Keywords:
Regulatory T cells
Natural Killer cells
Matrigel
Three-dimensional culture
Breast cancer

1. Introduction

The breast tumour microenvironment is an intricate network involving multiple cell types, the interactions of which play pivotal roles in cancer progression. Ex vivo experimentation has attempted to construct tumour microenvironments to better investigate cellular interactions; however, the reductionist approach of two-dimensional (2D) culture systems fails to recreate the complexity of the in vivo microenvironment (Pinto et al., 2011). While three-dimensional (3D) culture systems are more successful in this regard, studies of breast tumourigenesis have focused on metastatic processes primarily using homotypic cultures to investigate breast cancer cell invasion (Poincloux et al., 2011; Chandrasekaran et al., 2012; Yu and Machesky, 2012) or heterotypic cultures investigating breast tumour cell and fibroblast interaction (Olsen et al., 2010). While few studies have assessed the invasion capacity of lymphocytes in 3D cultures (Albertsson et al., 2007; Edsparr et al., 2011), to date, no studies have investigated the heterotypic interactions between immune cells and breast cancer cells in such a system.
Immunity is a fundamental determinant of tumour progression and response to therapy. The vast majority of studies have noted regulatory T (TREG) lymphocyte accumulation in the tumour infiltrating lymphocyte (TIL) population and peripheral blood of breast cancer patients (Bates et al., 2008; Pooi et al., 2006; Bohling and Allison, 2008; Decker et al., 2012). TREG lymphocytes, functionally described as the principal T cell subset responsible for maintenance of self-tolerance; are however, also associated with the induction of tumour tolerance and the suppression of anti-tumour responses (Sakaguchi et al., 1995; Beyer and Schultze, 2006; Mougiakakos et al., 2010). Elevated TREG lymphocyte numbers are associated with poor prognosis in oestrogen receptor (ER)+ invasive ductal carcinoma and human epidermal growth factor receptor 2 (HER2/neu)+ carcinomas (Pooi et al., 2006; Bates et al., 2008). In triple negative breast cancers, TREG lymphocytes have traditionally been linked with poor overall, and relapse-free survival (Bohling and Allison, 2008; Cimino-Mathews et al., 2013). However, recent data indicates that a TREG lymphocyte infiltrate conversely correlates with good prognosis in triple-negative cancers (West et al., 2013).
The immunosuppressive action of TREG lymphocytes against effector T lymphocytes and Natural Killer (NK) cells has been documented in both in vitro and in vivo studies (Ghiringhelli et al., 2005; Ralainirina et al., 2007; Salagianni et al., 2011); however, interleukin (IL)-activated NK cells have been shown to be resistant to these inhibitory effects (Chikileva et al., 2010). The primary mechanisms of NK cell killing, facilitated by direct contact with stressed, infected or malignant cells, include the activation of death receptors on target cells and the secretion of degradative granules including perforin, a membranedisrupting protein and granzymes, a family of proteases (Smyth et al., 2005). NK cells further produce a range of cytokines and chemokines implicated in controlling tumour progression (Wilk et al., 2008). These include: interferon-γ (IFN-γ), IL-2, IL-8, IL-15, tumour necrosis factor (TNF), transforming growth factor-β (TGF-β) and CCL3, amongst others; which allow for interaction with other cell types including malignant cells and TREG cells of the adaptive immune system (Chambers, 2010).
In patients presenting with breast cancers, NK cells are scarce in TIL populations (Georgiannos et al., 2003; Macchetti et al., 2006), despite an elevated presence in peripheral blood (Mozaffari et al., 2007). This may indicate inadequate NK cell homing mechanisms to tumour sites (Albertsson et al., 2003) and possibly the dominance of adaptive immune responses. Alterations in NK cell phenotype and functionality are dependent on tumour stage and presentation, with NK cells exhibiting poor cytotoxic capacity dominating NK populations in advanced breast cancer immune infiltrates (Dewan et al., 2009; Mamessier et al., 2011, 2013). The prognostic significance of NK cells in breast cancer has yet to be established (Roberti et al., 2012), however, gene expression studies have indicated that signatures associated with NK cells are predicative of relapse free-survival in primary breast cancer patients (Ascierto et al., 2013). The scarcity of NK cells and the dominance of TREG lymphocytes in advanced TIL populations may be linked to breast cancer immune evasion and tumour tolerance strategies, where the induction of a local inflammatory response is employed for tumour progression.
The reciprocal interaction between cancer cells, immune cells and the extracellular matrix (ECM) is mediated by a variety of growth factors, cytokines and chemokines. These secreted factors allow, either directly or indirectly, for the hallmarks of tumour progression to occur, including invasion and remodelling of the ECM, cellular proliferation, neovascularisation, immune evasion and extravasation into the vascular or lymphatic system for the establishment of distant secondary sites (de Visser and Coussens, 2005). There is an increasing body of evidence indicating that tumour cells are not only able to induce tumour tolerance and suppress or evade cytotoxic lymphocyte function, but also to induce TILs to become active participants in tumour progression (Whiteside et al., 1992; Drescher and Lynch, 2005; Prestwich et al., 2008; Emens et al., 2012; Cimino-Mathews et al., 2013; West et al., 2013). This stresses the need to investigate the interaction between TREG cells and NK cells in breast cancer in order to shed more light on tumour escape from immunological control, which may or may not be dependent on breast cancer phenotype.
In order to compare the response of luminal and basal phenotype breast cancer cells to immune infiltration, the luminal phenotype ER+ progesterone receptor (PR)+ MCF-7 cell line and the basal phenotype phenotype (ER−PR−HER2/neu−) MDA-MB-231 cell line (Lostumbo et al., 2006; Neve et al., 2006; Prat and Perou, 2010) were thus selected for this study. To recreate a geometric space more reminiscent of the tumour microenvironment, the commercially available laminin-rich extract established from Engelbreth–Holm–Swarm tumour cell-secreted basement membrane (Matrigel, BD Biosciences) (Kleinman and Martin, 2005; Nyga et al., 2011) was used as a scaffold within which to create 3D heterotypic cultures. Matrigel is regarded as a more biologically relevant scaffold for both normal mammary and cancerous mammary epithelial cell cultures compared to synthetic scaffolds (Ampuja et al., 2013), with morphology therein directly linked with phenotype (Kenny et al., 2007).
We thus present the feasibility of using a 3D heterotypic culture system to investigate the reciprocal interactions of TREG lymphocytes and NK cells with luminal and basal phenotype breast cancers. The effectiveness of downstream applications including RNA extraction for RT-PCR, immunocytochemistry for selected biomarkers and chemokine analysis is demonstrated.

2. Materials and methodology

2.1. Isolation of TREG lymphocyte and NK cell populations

Approximately 30 ml blood from seemingly healthy female volunteers (n = 13) between ages 18–35 (exclusion criteria included pregnancy, autoimmune diseases, immunodeficiency, cancer and a previous history of cancer) was collected via venipuncture in ethylenediaminetetraacetic acid (EDTA)-coated Vacutainers (BD Biosciences, Woodmead, South Africa). Following 1:1 dilution with phosphate buffered saline (PBS), peripheral blood mononuclear cells (PBMCs) were obtained via density gradient centrifugation using Ficoll-Hypaque (1.077 g/cm3) (GE Healthcare Biosciences AB, Sweden, 17-1440-03). PBMCs were washed in PBS and resuspended in 80 μl PBS (pH 7.2) supplemented with 0.5% bovine serum albumin (BSA) (Sigma-Aldrich, A9418) and 2 mM EDTA, in preparation for magnetic cell sorting. All reagents used in magnetic cell sorting were maintained at a temperature of 8 °C as per the manufacturer’s instruction.
The CD4+ Multisort Microbead Kit (Miltenyi Biotec, 130-055-101) was employed for positive selection of the CD4+ lymphocyte compartment followed by removal of residual magnetically labelled cells and enrichment of the CD4+ population using MACS Multisort Release agent (Miltenyi Biotec, 130-055-101) and MACS Multisort Stop Reagent (Miltenyi Biotec, 130-055-101). CD4+ lymphocytes were thereafter labelled with CD25 Microbeads (Miltenyi Biotec, 130-055-101) followed by positive selection and magnetic isolation of the prototypic TREG lymphocyte population. From the unlabelled PBMC fraction, NK cells were labelled with APC-NKp46 (Miltenyi Biotec, 130-092-609) and magnetically isolated using anti-APC microbeads (Miltenyi Biotec, 130-090-855). The median yield of live cells and viability were assessed using the trypan blue exclusion assay and the Bio-Rad Automated Cell Counter TC-20 (Bio-Rad, Parkwood, South Africa).
For activation of lymphocyte populations, TREG cells and NK cells were resuspended at a minimum of 1 × 105 cells/100 μl in RPMI medium 1640 (Lonza, Bloemfontein, South Africa) supplemented with 0.1% penicillin/ streptomycin (P/S), 10% foetal bovine serum (FBS), 1 μg/ml phytohaemagglutinin (PHA) (Sigma-Aldrich, L1668) and 0.8 ng/ml IL-2 (Miltenyi Biotec, 130-093-901) for 18 h. Subsequently, cell suspensions were harvested, centrifuged at 400 ×g for 5 min and resuspended in RPMI 1640 supplemented with 0.1% P/S, 10% FBS and 0.8 ng/ml IL-2 for a further 30 h after which the establishment of co-cultures was undertaken (Domaica et al., 2009).

2.1.1. Flow cytometry analysis of lymphocyte subpopulations

Flow cytometry (LSRFortessa, BD Biosciences, South Africa) was used to analyse the samples of isolated TREG and NK cell populations to ascertain the efficacy of the magnetic cell sorting technique and the activation procedure. Cells were resuspended at a maximum of 1 × 106 cells/100 μl PBS. Magnetically isolated TREG cells were incubated with 10 μl CD4-APC (BD Biosciences, Woodmead, South Africa, 555349) and 10 μl CD25-PE (BD Biosciences, 555432). Magnetically isolated NK cells were incubated with 10 μl CD56-PE-Cy7 (BD Biosciences, 557747) and 10 μl NKp46-APC (BD Biosciences, 558051). Both cell populations were incubated with their respective antibodies for 15 min at room temperature in the dark, washed in PBS and fixed in 1% paraformaldehyde for 10 min at room temperature. Subsequently, the cell suspensions were washed in PBS, centrifuged at 400 ×g and the resulting cell pellet resuspended in PBS and stored at 4 °C in the dark until flow cytometry analysis.
To establish the appropriate compensation to be applied prior to data collection, single-stained compensation beads (BD Biosciences, 552843) were used to assess fluorescence spill-over across channels. Each fluorescent-conjugated antibody at the concentrations used as above (10 μl antibody/100 μl PBS) was incubated with compensation beads in the dark for 15 min at room temperature. Negative compensation beads were left unstained, and thus used as a negative control. Suspensions were washed in 2.5 ml PBS and centrifuged at 400 ×g for 10 min. The supernatant was aspirated and the compensation beads resuspended in 500 μl PBS for subsequent flow cytometry analysis. FlowJo software (Free Trial) was used to analyse data.

2.2. Establishment of heterotypic 3D co-culture models

GFRM (BD Biosciences, BD354230), stored at −20 °C, was defrosted at 8 °C overnight prior to the establishment of co-cultures. All solutions and pipette tips were placed on ice and plating of cells within the GFRM was conducted on a cooling block. Luminal-phenotype culture models (LPCMs) and basal-phenotype culture models (BPCMs) were established by co-culturing the luminal phenotype, hormone-dependent MCF-7 cell line (Sigma Aldrich, St. Louis, USA, 86012803, passage number P11) and the basal phenotype, hormone independent, MDA-MB-231 cell line (Sigma Aldrich, St. Louis, USA, 92020424, passage number P46) respectively, with lymphocyte subgroups (Table 1). Specifically, lymphocyte subgroups (TREG and/ or NK cells from the same donor) were resuspended with breast cancer cells (MCF-7 or MDA-MB-231) at a ratio of 1:1:2 in a volume of 66.5 μl RPMI 1640 culture media supplemented with 10% FBS and 0.1% P/S, to yield experimental and control culture groups (Table 1).
Cell suspensions for each culture group were mixed with 200 μl GFRM and plated at ~65 μl/well (to yield a minimum 1 × 105 cell concentration) in 2× 96-well plates, in duplicate for each downstream application: ICC, RNA extraction and cytokine analysis. The cultures were allowed to solidify at room temperature for 5 min, followed by incubation at 37 °C for 30 min. Samples were thereafter incubated at 37 °C, 5% CO2 for 96 h in 100 μl RPMI 1640 culture media supplemented with 10% FBS and 0.1% P/S. Culture media levels were checked daily with a further 100 μl added over the 96 h period.
Images were captured with an Olympus iX51 Inverted Microscope using cellSens Software V10. All images were obtained at 20× or 40× magnification. Adobe Photoshop Elements 10 was used to generate the figures.

2.3. Analysis of CCL4 expression in supernatant

Culture media from each well was aspirated and centrifuged at 400 ×g for 5 min to pellet cellular debris. The supernatant was collected, snap frozen in liquid N2 and stored at −80 °C until use. The Bio-Plex Pro cytokine assay (Bio-Rad, Parkwood, South Africa, M50000175D) was used to detect alterations in the expression of the chemokine CCL4 in 50 μl of undiluted sample, as per the manufacturer’s instructions. Appropriate standards and two media controls i.e. culture media and GFRM-conditioned culture media were used. Data were acquired using the Bio-Plex 200 system and Bio-Plex Manager software. Assay sensitivity was noted as per the manufacturer’s notation.

2.4. Immunolocalisation of selected biomarkers

Cultures were rinsed in PBS prior to incubation in 30 μl FBS at 37 °C for 1 min. Cultures were then incubated with 20 U/ml thrombin (SANBS, Pretoria, South Africa) for approximately 30 s at 37 °C, followed by incubation at room temperature for a further 5 min to ensure clot formation. Cultures were fixed in 200 μl 4% paraformaldehyde in PBS for 20 min. The fixative was thereafter drained and eosin applied directly to the cultures to allow for visualisation of the samples (adapted from Graham et al., 2009). Samples were removed from the wells and placed into a 0.45 μm filter paper envelope for automatic tissue processing. The samples were embedded in paraffin wax and stored at 4 °C. Given the 3D nature of the cultures, serial sections were obtained at a thickness of 3 μm and collected on glass slides. Slides were dried at 60 °C on a hot-plate and stored at 4 °C until use.
Sections were dewaxed in xylene followed by rehydration in a decreasing series of alcohol. Sections underwent antigen retrieval in freshly prepared 0.1 M Tris, 5% urea buffer (pH 9.5) at 95 °C for 10 min in an oven, followed by 3× 5 min washes in distilled water. Thereafter, sections were incubated in 1% BSA in PBS–Tween for 30 min at room temperature. Subsequently, sections were incubated with relevant first and second sequence primary antibodies (where applicable) in PBS–Tween at 4 °C overnight. For double-labelling of the biomarkers MUC-1 and ERα in the LPCM, a monoclonal mouse anti-MUC1 antibody (Abcam, Pretoria, South Africa, AB28081) and a polyclonal rabbit anti-ER-α antibody (Abcam, AB31315) at a concentration of 1 ng/ml were used. EGFR in the BPCM was localised using a polyclonal rabbit anti-EGFR antibody (Abcam, AB2430) at a concentration of 4 ng/ml. Following overnight incubation with primary antibodies, sections were washed in PBS (3× 5 min). Sections were incubated for 2 h at room temperature with either Heterotypic 3D culture models based on the MCF-7 (luminal phenotype, hormonedependent) and MDA-MB-231 (basal phenotype, hormone-independent) breast cancer cell lines were developed. The experimental culture groups (EXPs) consisted of TREG lymphocytes and NK cells, co-cultured with either MCF-7 cells or MDA-MB-231 cells. The control groups included NK cells cultured with either MCF-7 or MDA-MB-231 cells (NK-BC); TREG lymphocytes cultured with either MCF-7 or MDA-MB-231 cells (TREG-BC) and MCF-7 or MDA-MB-231 cells cultured alone (BC).
Alexa Fluor 488 anti-rabbit (Life Technologies, Johannesburg, South Africa, A11008) and/or Alexa Fluor 594 anti-mouse (Life Technologies, A11005) secondary antibodies diluted to a concentration of 1:1000 in 1% BSA in PBS. Subsequently sections were washed in PBS (3× 5 min). Nuclei were thereafter counterstained with DAPI diluted to 1:50,000 in PBS for 5 min followed by two washes in PBS (5 min/wash). Sections were mounted in Fluoromount (Sigma-Aldrich, F4680) and stored at 4 °C till viewing.
Controls were conducted to ascertain antibody specificity, and identify background and non-specific fluorescence. These included: negative controls, in which primary antibodies were omitted and replaced with buffer; and isotype controls, in which each primary antibody was replaced with the same concentration of its matched isotype control (mouse IgG3 (Abcam AB18392) for MUC1, and rabbit IgG (Abcam, AB27478) for ER-α and EGFR) followed by incubation with the secondary antibody/antibodies. Images were captured with an Olympus iX51 Inverted Fluorescent Microscope using cellSens Software. Exposure time was set at 200 ms for all images. All images were obtained at 20× magnification. Adobe Photoshop Elements was used to formulate the figures.

2.5. RNA extraction

Various methods of RNA extraction were attempted to determine that which would result in the best RNA yield, purity and integrity. These methods included the use of the Qiagen RNeasy Mini Kit (Qiagen, Hilden, Germany, 74104) with or without the addition of proteinase K (Qiagen, 19131), TRIzol (Ambion, Texas, USA, 15596), and variation in off-column or on-column DNA digestion. Eluted RNA samples were all analysed individually. A NanoDrop Spectrophotometer (Thermo Fisher Scientific, Wilmington, USA) was used to determine the concentration and quality of eluted RNA and the Agilent 2100 Bioanalyzer and the RNA 600 Pico LabChip Kit (Agilent technologies, Waldbronn, Germany) used to further ascertain RNA purity and integrity.

2.6. Statistical analyses

Between-subject variability regarding the yield of lymphocyte populations was analysed using the Kruskal–Wallis ANOVA by Ranks method with STATISTICA v12 (Statsoft, Inc., Tulsa, OK, USA). The Spearman rank order correlation test was used to determine whether correlates exist between lymphocyte populations. For CCL4 cytokine analysis, the data were first standardised. Since normality tests indicated the nonparametric nature of the data, the Kruskal–Wallis ANOVA by Ranks method with Bonferroni adjustment was employed to assess whether culture groups within the LPCM and BPCM were significantly different from each other with regard to cytokine expression. Multiple comparison post-hoc tests (2-tailed with Bonferroni adjustment) were used thereafter to determine which culture groups exhibited the differences found by the Kruskal–Wallis test. Specifically, these statistical tests considered all culture groups including the blank media control as comparator groups for analysis. For quantitative analysis of cell masses in the LPCM, cellSens software was used to manually define cellular area. Since the data did not follow a normal distribution the Kruskal–Wallis by Ranks method as indicated above was also used to assess whether culture groups in the LPCM presented with significantly different cell masses. For all analyses, p b 0.05 was regarded as statistically significant.

3. Results & discussion

3.1. Flow cytometry analysis of sorted and activated populations

The critical factor of any study investigating TREG lymphocytes remains phenotypic characterisation. In this study, the prototypic TREG lymphocyte population (CD4+CD25+) was obtained using magnetic bead-based technology, the most commonly employed method for lymphocyte isolation for in vitro assays (Mandapathil et al., 2009). One of the limiting factors of this isolation method is the inability to ascertain levels of marker expression. While flow cytometry isolation overcomes this problem and allows for increased sample purity, it too is still plagued by arbitrary designations as to what levels constitute high and low expressions of designated markers (Whiteside, 2012). Flow cytometry analysis confirmed the CD4+CD25+ profile of magnetically isolated TREG lymphocyte populations, and the CD56+/NKp46+ profile of NK cells. Cells exhibited slightly altered scatter properties possibly activation with IL-2 and PHA are noted. due to the effects of fixation thus only those cells which exhibited the typical side scatter (SSC) properties, associated with internal complexity and forward scatter (FSC) properties, associated with cell size, of lymphocytes were analysed further (Figs. 1.1A, C and 1.2A, C).
In this study flow cytometry analysis affirmed the selected profiles of magnetically isolated TREG lymphocytes. It was noted that CD4 expression was clearly delineated into CD4dim and CD4bright subsets (77.6% of the total lymphocyte population) with CD25 co-expression. Thus two distinct subsets, designated CD4dimCD25+ and CD4bright + CD25 (Fig. 1.1B) were noted, with the latter dominating the total lymphocyte population. Following activation with IL-2 and PHA, TREG lymphocytes maintained these populations (Fig. 1.1D). While period.
upregulation of CD25, the α chain of the IL-2 receptor, is commonly associated with IL-2 signalling (Cheng et al., 2011), CD25 expression has been shown to be reduced following permeabilisation of the cell membrane (Lee et al., 2011). While the fixative used in this study does not incorporate a detergent, fixation in paraformaldehyde causes some permeabilisation of the cell membrane, and may thus affect CD25 expression. Analysis of the CD4 population both postmagnetic cell sorting and following activation revealed a negligible contaminating population of CD4− cells. Thus, studies investigating the TREG subset must take cognizance that without a definitive phenotype, isolated populations may contain contaminating effector cells which in turn may contribute to the plethora of conflicting evidence regarding the role of TREG lymphocytes in breast cancer (Yu and Fu, 2006; Bohling and Allison, 2008; Cimino-Mathews et al., 2013; West et al., 2013).
Flow cytometry analysis of magnetically isolated NK cells demonstrated four distinct subsets based on the total lymphocyte population. These subsets were designated CD56brightNKp46bright; CD56bright+ NKp46bright; CD56dimNKp46dim and CD56brightNKp46dim (Fig. 1.2B). A small contaminating population of CD56− cells which nevertheless exhibited NKp46dim and NKp46bright expression was identified. Following in vitro activation with PHA and IL-2, only two distinct populations, the CD56dimNKp46bright and CD56dimNKp46dim subsets were noted (Fig. 1.2D). While CD56bright NK cells are considered to have poor cytolytic capacity, high proliferative capacity and are proposed as the dominant NK subset in cytokine secretion, the downregulation of CD56 is proposed to be indicative of the acquisition of specialised function (Bottino et al., 2000; Miller, 2001; Piersma et al., 2008; Moretta, 2010). Moreover, other studies have indicated that the CD56dim subset is the major producer of proinflammatory cytokines and chemokines upon target cell recognition, of particular interest to this study (Fauriat et al., 2010; Moretta, 2010).

3.2. Effects of magnetic cell sorting and subsequent activation on TREG lymphocyte and NK cell yield

No statistically significant differences were found with regard to the yields (total and live) of the PBMCs (live mean 2.09 × 107) and lymphocyte subgroups following magnetic cell sorting and activation procedures (Fig. 2.1), highlighting the efficacy of the technique. Subsequent to magnetic isolation, the median yield of live TREG lymphocytes and NK cells was 1.55 × 106 (88.5% viable, range 3.22 × 105–7.33E × 106) and 1.1 × 106 (83.2% viable, range 1.55E × 105–2.51 × 106) respectively. Following the first 18 h period of activation the average percentage of live NK cells decreased to 68% (median yield 1.75 × 105, range 1.39 × 105–1.86 × 106) and stabilised at 62% over the next 30 h activation period (median yield 5.37E + 05, range 1.50 × 105–1.86 × 106) (Fig. 2.2). The average percentage live cell count for TREG lymphocytes post-magnetic cell sorting was 88% (median yield 3.34 × 105, range 6.11 × 104–1.43 × 106) and stabilised at 78% (median yield 8.94 × 105, range 9.44 × 104–3.41 × 106) during both activation periods. For subsequent co-culture procedures, TREG lymphocytes and NK cells were plated only if the sample percentage viability exceeded 70% and 60% respectively.
The yield of TREG lymphocytes obtained correlated with that of CD4+ lymphocytes post-magnetic cell sorting. However, no correlation was found between the aforementioned subsets after the first activation procedure, with the TREG subset instead correlating with the PBMCs. Interestingly, following the second activation procedure the yield of the TREG subset correlated with that of the initial post-magnetic cell sorting as well as that of the first activation procedure. This data may have further value for in vitro assays in raising the possibility of predicting the yield of TREG cells post-activation. No correlations were found regarding the NK cell subset.

3.3. Morphological analysis of heterotypic 3D cultures & CCL4 expression

Molecular profiling of the MCF-7 and MDA-MB-231 cell lines has confirmed their representation of infiltrating ductal carcinoma, specifically with reference to their respective luminal and basal phenotypes (Neve et al., 2006; Kao et al., 2009). MCF-7 and MDA-MB-231 cells were initially propagated as 2D monolayers in plastic culture dishes to establish stock populations. Under these conditions, MCF-7 cells (Fig. 3.1A) exhibited a characteristic polygonal morphology with vesicular nuclei and prominent nucleoli. Cells evidenced blunt cytoplasmic connections and numerous pseudopodia, forming a cobble stone-like appearance typical of their luminal phenotype. MDA-MB-231 cells in monolayer (Fig. 3.1B) were primarily spindle-shaped, as per their basal phenotype, and associated with each other in an irregular, lattice-like manner.
In our 3D culture system, by recreating the spatial dimensions associated with the in vivo environment, tumour cells are allowed to assume a morphology and molecular phenotype, more representative of the in vivo tumour environment than those cultured in 2D systems (Kenny et al., 2007; Han et al., 2010; Krause et al., 2010; Pinto et al., 2011). One of the limitations of this study and others is the visualisation of live unstained 3D culture systems using 2D microscopy (Naber et al., 2011). Despite these limitations, qualitative assessment alone reveals distinct morphological differences compared to 2D culture systems, and the effects of immune mediation on breast cancer cells. In the 3D culture system, both cell lines required between 24 h and 48 h to attach and spread within the GFRM, with MDA-MB-231 cells establishing themselves at a marginally slower rate. In the LPCM (Fig. 3.2.1), MCF7 cells assumed a globular polygonal morphology, with large nuclei and numerous nucleoli, and exhibited connecting cytoplasmic projections by 48 h. In the BPCM (Fig. 3.3), MDA-MB-231 cells established themselves at a slower rate, assuming a stellate morphology with cytoplasmic extensions. In the LPCM BC control group, in which MCF-7 cells were cultured without immune mediation, cells formed putative tubular and acinar glandular structures (Fig. 3.2.1), indicating a cavitation process despite the lack of definitive lumina (do Amaral et al., 2011); a phenomenon noted in other studies (Krause et al., 2010). The formation of glandular structures, specifically with lumen formation, is indicative of a well-differentiated tumour (Krause et al., 2010) traditionally associated with better prognosis. In the BPCM BC control group, the more aggressive MDA-MB-231 cells exhibited an organised lattice-like network of cytoplasmic projections (Fig. 3.3). While this morphological association was distinctly dissimilar to another 3D study in which MDAMB-231 cells formed stellate clusters (Ampuja et al., 2013); our results do reflect invasion tunnels formed by MDA-MB-231 cells linking into chains, allowing leading cells to migrate within the tunnel through the GFRM (Yu and Machesky, 2012), a process dependent on upregulation of chemokines and matrix metalloproteinases (MMPs).
MMPs are noted to facilitate the migration of IL-2 activated-NK cells through GFRM (Edsparr et al., 2010). In this study, the possibility of MMP induction was highlighted by the reduction in viscosity of GFRM in culture groups under NK cell mediation (EXP and NK-BC culture groups), and to a lesser extent under TREG lymphocyte mediation, in both the LPCM and BPCM systems. These culture groups exhibited a less gel-like consistency than the other culture groups, which caused difficulty in harvesting samples from microwells. Using light microscopy, it was evident that GFRM underwent remodelling exhibiting a more filamentous appearance in those culture groups containing lymphocytes as opposed to the relatively homogenous appearance in BC control groups (data not shown). In the LPCM, NK cell presence was further associated with the disruption of MCF-7 cell masses (Fig. 3.2.1A, B) regardless of TREG lymphocyte presence. Notably under TREG lymphocyte mediation alone (Fig. 3.2.1C), cell clusters were more reminiscent of the BC control group (Fig. 3.2.1D). We investigated this phenomenon quantitatively using cellSens software to define cell mass area in the LPCM. A polygon tool was used to manually define the area parameters (Fig. 3.2.2A); however, no area could be calculated for the EXP culture group given the extreme reduction of cell masses. Notably, quantitative assessment showed that NK cell presence significantly disrupted the presentation of MCF-7 cell masses compared to TREG-mediation alone and the BC control group (Fig. 3.2.2B). Furthermore, no significant difference in cell mass area was noted between the TREG-BC control group and the BC control group (Fig. 3.2.2B). The quantitative assessment echoes that of the qualitative assessment; however, that cells also lie beneath the plane of view must be noted as a caveat.
In the BPCM group, qualitative assessment indicated that immune mediation resulted in effects similar to that seen in the LPCM, albeit regarding network formation. NK cells disrupted MDA-MB-231 cell networks regardless of TREG lymphocyte presence (Fig. 3.3A, B). MDAMB-231 cells under TREG lymphocyte-mediation alone (Fig. 3.3C) showed distinct stellate networks, reminiscent of the BC control groups (Fig. 3.3D). Cellular area could not be accurately assessed quantitatively in the BPCM group given that the unstained networks did not lend themselves to precise measurement.
The analysis of CCL4 secretion corroborated our findings regarding the remodelling of the GFRM. CCL4 is implicated in tissue remodelling via MMP induction and is associated with pro-tumourigenic processes (Vicari and Caux, 2002; Ben-Baruch, 2003; Mantovani et al., 2010). No significant difference was found between the two media controls, i.e. GFRM-conditioned culture media and culture media alone, thus for subsequent analyses data were combined. CCL4 analysis of the BC control groups indicates that MDA-MB-231 cells produce a higher mean concentration of CCL4, as opposed to MCF-7 cells (Fig. 4). This pattern was expressed in the TREG-BC control groups, indicating that TREG cells may maintain the intrinsic capacity of breast cancer cells to produce chemokines, as indicated by other studies (Robinson and Coussens, 2005), for the induction of a proinflammatory microenvironment.
Within the LPCM, the EXP culture group in which MCF-7 cells were under the influence of both TREG lymphocytes and NK cells presented significantly higher levels (p b 0.05) of CCL4 than the blank media control and control culture groups TREG-BC and BC (Fig. 4). While CCL4 detection in the NK-BC control was significantly higher (p b 0.05) than that of the blank control and culture group BC, it was not significantly different to the EXP group. This may indicate that the interaction between the lymphocyte subgroups induces increased CCL4 production. Conversely, within the BPCM CCL4 expression was lower in the EXP culture group as compared to the NK-BC control culture group, albeit not significantly so. While IL-2 activated NK cells are noted for their CCL4 production (Sanchez et al., 2010), overall the variation in CCL4 secretion was more apparent in culture groups within the LPCM, indicating the possible influence of cell phenotype on the secretion of this chemokine. We propose that breast cancer cells subvert both immune cell subsets to induce tumour tolerance and potentially to facilitate tumour progression via remodelling of the extracellular matrix and the induction of a proinflammatory microenvironment.

3.4. Immunolocalisation of tumour biomarkers in the LPCM and BPCM

In this study we present a qualitative interpretation of the immunostaining in order to demonstrate that (double) immunolocalisation of tumour biomarkers is possible using this 3D model. Since our system allows for the obtaining of serial sections we plan to use quantitative procedures, specifically CellProfiler (Carpenter et al., 2006) to assess differences in biomarker expression, as we have conducted in previous studies in our laboratory (Gil et al., 2013). Of particular importance is evidence that biomarker expression alters in a 3D environment, in a manner not recapitulated by 2D studies. In this study ER-α expression in the LPCM was of intermediate intensity (Fig. 5), appearing reduced, possibly as a function of the 3D culture scenario, as compared with other studies of MCF-7 cells in monolayer (Gil et al., 2013; Sharan et al., 2013). Furthermore, we identified not only nuclear localisation in all culture groups as expected, but found that immune-mediation induces a shift in ER localisation to the cytoplasmic compartment (Fig. 5). The importance of this finding is linked to the differing function of ER-α dependent on its cellular localisation. Clinically, downregulation of ERα expression is associated with increased recurrence and metastasis (Dhasarathy et al., 2007); however, diagnostic procedures are essentially more concerned with nuclear ER-α presentation. Cytoplasmic expression is associated with an increase in invasive profile which in 2004). Areas of high intensity nuclear and perinuclear EGFR expressions were noted primarily in the EXP, NK-BC and TREG-BC culture groups (Fig. 5). Nuclear translocation of the EGFR is associated with invasive processes, including MMP induction, proliferation, and immune evasion (Lechner et al., 2005; Nyga et al., 2011; Brand et al., 2013). Distinct regions of extracellular EGFR expression, indicating secreted EGFR, were noted primarily in the EXP and NK-BC culture groups. These regions either appeared diffuse or as clusters. However, these regions could not be accurately distinguished from the cytoplasm since cell boundaries could not be resolved. Furthermore, overlay of fluorescence data with light microscopy photomicrographs proved inadequate in distinguishing cell boundaries within the filamentous GFRM. Since the antibody employed recognises the 170 kDA EGFR isoform via western blot, it is expected that the staining in the extracellular matrix is not due to the soluble form of EGFR but rather may correspond to that sequestered in exosomes, implicated in immune suppression and cellular communication necessary for tumour progression, trapped within the GFRM (Adamczyk et al., 2011; Zhang and Grizzle, 2011).

3.5. RNA extraction

In order to identify alterations in gene expression associated with metastasis and cell death, RNA extraction from 3D cultures was optimised. The vast majority of studies using Matrigel for invasion assays (Pollard et al., 2003; Selander et al., 2004; Jin et al., 2010; Chandrasekaran et al., 2012; Hoover et al., 2012) or to establish xenografts, typically conducted further assessments using immunohistochemistry (Elstner et al., 1998; Pollard et al., 2003; Engelmann et al., 2008; Tate et al., 2012). There is a distinct paucity of studies extracting RNA for further analysis from Matrigel 3D cultures. With the number of available RNA extraction kits on the market, it is understandable that different research groups use different kits. However, few researchers report in detail, any variation to their extraction methodology or the methods of RNA quality control conducted (Herrera et al., 2006; Lee et al., 2007; Graham et al., 2009; Ampuja et al., 2013).
The advent of the MIQE (Minimum Information for Publication of Quantitative Real-Time PCR Experiments) guidelines in 2009, aims to ensure the repeatability of experimentation (Bustin et al., 2009). Details essential to reporting nucleic acid extraction methods include modification to extraction procedures or kits and additional DNase or RNase treatments for example. It is further suggested that while absorbance ratios as determined using spectrophotometry (NanoDrop) provide an indication of RNA purity, the integrity of the RNA to be investigated using microfluidic analysis (Agilent Technologies Bioanalyzer).
We confirm the overexpression of EGFR in the MDA-MB-231 cell line as indicated by numerous other studies (Foley et al., 2010; Gumuskaya et al., 2010; Ye et al., 2013). In the BPCM, EGFR expression was found predominantly in the cytoplasmic compartment in all culture groups, but was of a higher intensity in the TREG-BC and NK-BC culture groups (Fig. 5). EGFR localisation to the cytoplasm has been noted in breast tumours, with a correlation between such staining and increased invasiveness established in pancreatic adenocarcinomas turn involves the upregulation of MMPs and chemokines (Kumar purity with a value close to 1 indicative of large scale RNA degradation.
A number of techniques were attempted to optimise the yield, purity and integrity of the extracted RNA (Table 2). Some studies have advocated the removal of cell colonies from Matrigel with EDTA (Kenny et al., 2007; Lee et al., 2007; Ampuja et al., 2013) prior to extraction. However, it was of concern that EDTA treatment would alter gene expression beyond the dictates of the co-culture system. As such, this method was not attempted. This study found that the TRIzol technique, used in studies involving Matrigel-coated flasks or 3D Matrigel cultures (Herrera et al., 2006; Lee et al., 2007), whether used in conjunction with the Qiagen RNeasy Mini Kit or not, did not yield RNA with suitable integrity for further analysis (Table 2, Fig. 6). The efficacy of sample lysis in RLT buffer followed by extraction using the Qiagen RNeasy Mini Kit was determined; however, unlike other studies (Han et al., 2010), it was necessary to include proteinase K for digestion of residual GFRM comparable to the protocol for RNA extraction from fibrous tissue (Table 2, Fig. 6). Furthermore, it was noted that lysis in RLT buffer at 55 °C followed by both off and on-column DNA digestion, resulted in varying yields of RNA, but no RNA integrity number (Table 2 and Fig. 6). Similar to other studies (Kenny et al., 2007; Han et al., 2010), it was determined that only one DNA digestion (off-column in this study) was necessary to remove residual DNA but also to ensure no further damage to the integrity of the RNA. It was thus established that the use of the Qiagen RNeasy Mini Kit protocol with lysis of samples in RLT buffer at room temperature followed by digestion of residual GFRM with proteinase K and off-column DNA digestion alone presented a fair RNA yield, given the low cell numbers. The value of the primary measure of purity (260/280 ratio), as determined using the NanoDrop, was 1.98, indicative of good purity, with Agilent analysis providing an RNA integrity number of 9.2 (Table 2, Fig. 6).

4. Conclusion

Three-dimensional culture systems investigating breast cancer are primarily directed at homotypic models, or heterotypic models involving fibroblast interaction. We have developed the first 3D coculture system for the investigation of heterotypic interactions between TREG lymphocytes, NK cells and either hormone-dependent or independent breast cancer cells. We have shown that immunemediation affects morphological presentation of tumour masses. Since tumour cell morphology in 3D is linked to a molecular phenotype and physiologic behaviour that recapitulates the in vivo tumour microenvironment more faithfully, we further demonstrate that the culture system is suitable for downstream applications including gene expression analysis, the generation of serial sections for subsequent immunolocalisation of tumour biomarkers and cytokine analysis.

References

Acconcia, F., Kumar, R., 2006. Signaling regulation of genomic and nongenomic functions of oestrogen receptors. Cancer Lett. 238, 1.
Adamczyk, K.A., Klein-Scory, S., Moradian, M., Warnken, U., Schmiegel, W., Schnölzer, M., Schwarte-waldhoff, I., 2011. Characterization of soluble and exosomal forms of the EGFR released from pancreatic cancer cells. Life Sci. 89 (9–10), 304.
Albertsson, P.A., Basse, P.H., Hokland, M., Goldfarb, R.H., Nagelkerke, J.F., Nannmark, U., Kuppen, P.J.K., 2003. NK cells and the tumour microenvironment: implications for NK-cell function and anti-tumour activity. Trends Immunol. 24 (11), 603.
Albertsson, P.A., Basse, P.H., Edsparr, K., Kim, M.H., Golfarb, R.H., Kitson, R.P., Lennernäs, B., Nannmark, U., Johansson, B.R., 2007. Differential locomotion of long- and short-term IL-2-activated murine natural killer cells in a model matrix environment. Scand. J. Immunol. 66, 402.
Ampuja, M., Jokimäki, R., Juuti-Uusitalo, K., Rodriguez-Martinez, A., Alarmo, E.L., Kallioniemi, A., 2013. BMP4 inhibits the proliferation of breast cancer cells and induces an MMP-dependent migratory phenotype in MDA-MB-231 cells in 3D environment. BMC Cancer 13, 429.
Ascierto, M.L., Idowu, M.O., Zhao, Y., Khalak, H., Payne, K.K., Wang, X., Dumur, C.I., Bedognetti, D., Tomei, S., Ascierto, P.A., Shanker, A., Bear, H.D., Wang, E., Marincola, F.M., De Maria, A., Manjili, M.H., 2013. Molecular signatures mostly associated with NK cells are predictive of relapse free survival in breast cancer patients. J. Transl. Med. 11 (1), 145.
Bates, G.J., Fox, S.B., Han, C., Leek, R.D., Garcia, J.F., Harris, A.L., Banham, A.H., 2008. Quantification of regulatory T cells enables the identification of high-risk breast cancer patients and those at risk of late relapse. J. Clin. Oncol. 24 (34), 5373.
Ben-Baruch, A., 2003. Inflammatory cells, cytokines and chemokines in breast cancer progression: reciprocal tumor – microenvironment interactions. Breast Cancer Res. 5, 31. Beyer, M., Schultze, J.L., 2006. Regulatory T cells in cancer. Blood 108, 804.
Björnström, L., Sjöberg, M., 2004. Estrogen receptor-dependent activation of AP-1 via non-genomic signalling. Nucl. Recept. 2, 3.
Bohling, S.D., Allison, K.H., 2008. Immunosuppressive regulatory T cells are associated with aggressive breast cancer phenotypes: a potential therapeutic target. Mod. Pathol. 21 (12), 1527.
Bottino, C., Biassoni, R., Millo, R., Moretta, L., Moretta, A., 2000. The human natural cytotoxicity receptors (NCR) that induce HLA class I-independent NK cell triggering. Hum. Immunol. 61, 1.
Brand, T.M., Iida, M., Luthar, N., Starr, M.M., Huppert, E.J., Wheeler, D.L., 2013. Nuclear EGFR as a molecular target in cancer. Radiother. Oncol. 108 (3), 370.
Bustin, S.A., Benes, V., Garson, J.A., Hellemans, J., Huggett, J., Kubista, M., Mueller, R., Nolan, T., Pfaffl, M.W., Shipley, G.L., Vandesompele, J., Wittwer, C.T., 2009. The MIQE guidelines: Minimum Information for publication of Quantitative real-time PCR Experiments. Clin. Chem. 55 (4), 611.
Carpenter, A.E., Jones, T.R., Lamprecht, M.R., Clarke, C., Kang, I.H., Friman, O., Guertin, D.A., Chang, J.H., Lindquist, R.A., Moffat, J., Golland, P., Sabatini, D.M., 2006. Cell Profiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biol. 7, R100. http://dx.doi.org/10.1186/gb-2006-7-10-r100.
Chambers, B.J., 2010. NK cell–T cell interactions. In: Lotze, M.T., Thomson, A.W. (Eds.), Natural Killer Cells, Basic Science and Clinical Application. Academic Press Elsevier Ltd., London, p. 297.
Chandrasekaran, S., Geng, Y., DeLouise, L., King, M.R., 2012. Effect of homotypic and heterotypic interaction in 3D on the E-selectin mediated adhesive properties of breast cancer cell lines. Biomaterials 33 (35), 9037.
Cheng, G., Aixin, Y., Malek, T.R., 2011. T-cell tolerance and the multi-functional role of IL2R signalling in T-regulatory cells. Immunol. Rev. 241, 63.
Chikileva, I.O., Shubina, I.Z., Baronzio, G., Kiselevsky, M.V., 2010. Is it necessary to deplete the lymphokine activated killers’ populations of CD4+CD25+ lymphocytes? Regulatory Foxp3-positive T cells within lymphokine activated killers. Biomed. Pharmacother. 64, 379.
Cimino-Mathews, A., Ye, X., Meeker, A., Argani, P., Emens, L., 2013. Metastatic triplenegative breast cancers at first relapse have fewer tumour-infiltrating lymphocytes than their matched primary breast tumours: a pilot study. Hum. Pathol. 44 (10), 2055.
de Visser, K., Coussens, L., 2005. The interplay between innate and adaptive immunity regulates cancer development. Cancer Immunol. Immunother. 25, 1143.
Decker, T., Fischer, G., Bücke, W., Bücke, P., Stotz, F., Grüneberger, A., Gropp-Meier, M., Wiedemann, G., Pfeiffer, C., Peschel, C., Götze, K., 2012. Increased number of regulatory T cells (T-regs) in the peripheral blood of patients with Her-2/neu-positive early breast cancer. J. Cancer Res. Clin. Oncol. 138 (11), 1945.
Dewan, Z., Takada, M., Terunuma, H., Deng, X., Ahmed, S., Yamamoto, N., Toi, M., 2009. Natural killer activity of peripheral-blood mononuclear cells in breast cancer patients. Biomed. Pharmacother. 63 (9), 703.
Dhasarathy, A., Kajita, M., Wade, P.A., 2007. The transcription factor Snail mediates epithelial to mesenchymal transitions by repression of oestrogen receptor alpha. Mol. Endocrinol. 21 (12), 2907.
do Amaral, J.B., Rezende-Teixeira, P., Freitas, V.M., Machado-Santelli, G.M., 2011. MCF-7 cells as a three-dimensional model for the study of human breast cancer. Tissue Eng. C 17 (11), 1097.
Domaica, C.I., Fuertes, M.B., Rossi, L.E., Rabinovich, G.A., Zwirner, N.W., 2009. Tumour experienced T cells promote NK cell activity through trogocytosis of NKG2D and NKp46 ligands. EMBO Rep. 10, 908.
Drescher, K.M., Lynch, H.T., 2005. Tumour infiltrating Proteinase K lymphocytes (TILs): lessons learned in 30 years of study. Clin. Appl. Immunol. Rev. 5, 149.
Edsparr, K., Speetjens, F.M., Mulder-Stapel, A., Goldfarb, R.H., Basse, P.H., Lennernäs, B., Kuppen, P.J.K., Albertsson, P., 2010. Effects of IL-2 on MMP expression in freshly isolated human NK cells and the IL-2-independent NK cell line YT1. J. Immunother. 33 (5), 475.
Edsparr, K., Basse, P.H., Goldfarb, R.H., Albertsson, P., 2011. Matrix metalloproteinases in cytotoxic lymphocytes impact on tumour infiltration and immunomodulation. Cancer Microenviron. 4, 351.
Elstner, E., Muller, C., Koshizuka, K., Wimmilianson, E.A., Park, D., Asou, H., Shintaku, P., Said, J.W., Heber, D., Koeffler, H.P., 1998. Ligands for peroxisome proliferatoractivated receptor y and retinoic acid receptor inhibit growth and induce apoptosis of human breast cancer cells in vitro and in BNX mice. Proc. Natl. Acad. Sci. U. S. A. 95 (July), 8806.
Emens, L.A., Silverstein, S.C., Khleif, S., Marincola, F.M., Galon, J., 2012. Toward integrative cancer immunotherapy: targeting the tumour microenvironment. J. Transl. Med. 10 (1), 70. http://dx.doi.org/10.1186/1479-5876-10-70.
Engelmann, K., Shen, H., Finn, O.J., 2008. MCF7 side population cells with characteristics of cancer stem/progenitor cells express the tumour antigen MUC1. Cancer Res. 68, 2419.
Fauriat, C., Long, E.O., Ljunggren, H.G., Bryceson, Y.T., 2010. Regulation of human NK-cell cytokine and chemokine production by target cell recognition. Blood 115 (11), 2167. Foley, J., Nickerson, N.K., Nam, S., Tan, K., Gilmore, J.L., Nephew, K.P., Riese II, D.J., 2010. EGFR signalling in breast cancer: bad to the bone. Semin. Cell Dev. Biol. 21 (9), 951.
Georgiannos, S.N., Renaut, A., Goode, A.W., Sheaff, M., 2003. The immunophenotype and activation status of the lymphocytic infiltrate in human breast cancers, the role of the major histocompatibility complex in cell-mediated immune mechanisms, and their association with prognostic indicators. Surgery 134, 827.
Ghiringhelli, F., Ménard, C., Terme, M., Flament, C., Taieb, J., Chaput, N., Puig, P.E., Novault, S., Escudier, B., Vivier, E., Lecesne, A., Robert, C., Blay, J., Bernard, J., Caillat-Zucman, S., Freitas, A., Tursz, T., Wagner-Ballon, O., Capron, C., Vainchencker, W., Martin, F., Zitvogel, L., 2005. CD4 CD25 regulatory T cells inhibit natural killer cell functions in a transforming growth factor-β-dependent manner. J. Exp. Med. 202 (8), 1075.
Gil, J.F., Augustine, T.N., Hosie, M.J., 2013. Anastrozole and RU486: effects on estrogen receptor α and Mucin 1 expression and correlation in the MCF-7 breast cancer cell line. Acta Histochem. 115, 851.
Graham, J.D., Mote, P.A., Salagame, U., Balleine, R.L., Huschtscha, L.I., Clarke, C.L., 2009. Hormone-responsive model of primary human breast epithelium. J. Mammary Gland Biol. Neoplasia 14, 367.
Gumuskaya, B., Alper, M., Hucumenoglu, S., Altundag, K., Uner, A., Guler, G., 2010. EGFR expression and gene copy number in triple-negative breast carcinoma. Cancer Genet. Cytogenet. 203 (2), 222.
Han, J., Chang, H., Giricz, O., Lee, G.Y., Baehner, F.L., Gray, J.W., Bissell, M.J., Kenny, P.A., Parvin, B., 2010. Molecular predictors of 3D morphogenesis by breast cancer cell lines in 3D culture. PLoS Comput. Biol. 6 (2), e1000684. http://dx.doi.org/10.1371/ journal.pcbi.1000684.
Herrera, M.B., Bruno, S., Buttiglieri, S., Tetta, C., Gatti, S., Deregibus, M.C., Bussolati, B., Camussi, G., 2006. Isolation and characterization of a stem cell population from adult human liver. Stem Cells 24, 2840.
Hoover, R.G., Gullickson, G., Kornbluth, J., 2012. Natural killer lytic-associated molecule plays a role in controlling tumour dissemination and metastasis. Front. Immunol. 3, 393. http://dx.doi.org/10.3389/fimmu.2012.0039.
Jin, Q., Yuan, L.X., Boulbes, D., Baek, J.M., Wang, Y.N., Gomez-cabello, D., Hawke, D.H., Yeung, S.C., Lee, M.H., Hortobagyi, G.N., Hung, M.C., Esteva, F.J., 2010. Fatty acid synthase phosphorylation: a novel therapeutic target in HER2-overexpressing breast cancer cells. Breast Cancer Res. 12 (6), R96. http://dx.doi.org/10.1186/bcr2777.
Kao, J., Salari, K., Bocanegra, M., Choi, Y.L., Girard, L., Gandhi, J., Kwei, K.A., HernandezBoussard, T., Wang, P., Gazdar, A.F., Minna, J.D., Pollack, J.R., 2009. Molecular profiling of breast cancer cell lines defines relevant tumour models and provides a resource for cancer gene discovery. PLoS ONE 4 (7), e6146. http://dx.doi.org/10.1371/journal. pone.0006146.
Kenny, P.A., Lee, G.Y., Myers, C.A., Neve, R.M., Semeiks, J.R., Spellman, P.T., Lorenz, K., Lee, E.H., 2007. The morphologies of breast cancer cell lines in three-dimensional assays correlate with their profiles of gene expression. Mol. Oncol. 1 (1), 84.
Khodarev, N.N., Pitroda, S.P., Beckett, M.A., Macdermed, D.M., Huang, L., Kufe, D.W., Weichselbaum, R.R., 2009. MUC1-induced transcriptional programs associated with tumorigenesis predict outcome in breast and lung cancer. Cancer Res. 69 (7), 2833.
Kleinman, H.K., Martin, G.R., 2005. Matrigel: basement membrane matrix with biological activity. Semin. Cancer Biol. 15, 378.
Krause, S., Maffini, M.V., Soto, A.M., Sonnenschein, C., 2010. The microenvironment determines the breast cancer cells’ phenotype: organization of MCF7 cells in 3D cultures. BMC Cancer 10, 263.
Kumar, R., Wang, R., Mazumdar, A., Talukder, A.H., Mandal, M., Yang, Z., Bagheri-Yarmand, R., Sahin, A., Hortobagyi, G.N., Adam, L., Barnes, C.J., Vadlamudi, R.K., 2002. A naturally occurring MTA1 variant sequesters oestrogen receptor-a in the cytoplasm. Nature 418 (August), 256.
Lechner, M., Lirk, P., Rieder, J., 2005. Inducible nitric oxide synthase (iNOS) in tumour biology: the two sides of the same coin. Semin. Cancer Biol. 15, 277.
Lee, G.Y., Kenny, P.A., Lee, E.H., Bissell, M.J., 2007. 3D culture models of normal and malignant breast epithelial cells. Nature Methods 4 (4), 359.
Lee, S.K., Kim, J.Y., Jang, B.W., Hur, S.E., Na, B.J., Lee, M., Fukui, A., Gilman-Sachs, A., KwakKim, J., 2011. Foxp3(high) and Foxp3(low) Treg cells differentially correlate with T helper 1 and natural killer cells in peripheral blood. Hum. Immunol. 72 (8), 621.
Lostumbo, A., Mehta, D., Setty, S., Nunez, R., 2006. Flow cytometry: a new approach for the molecular profiling of breast cancer. Exp. Mol. Pathol. 80, 46.
Macchetti, A.H., Marana, H.R.C., Silva, J.S., De Andrade, J.M., Ribeiro-Silva, A., Bighetti, S., 2006. Tumour-infiltrating CD4+ T lymphocytes in early breast cancer reflect lymph node involvement. Clinics 61 (3), 203.
Mamessier, E., Sylvain, A., Thibult, M., Houvenaeghel, G., Jacquemier, J., Castellano, R., Gonçalves, A., André, P., Romagné, F., Thibault, G., Viens, P., Birnbaum, D., Bertucci, F., Moretta, A., Olive, D., 2011. Human breast cancer cells enhance self tolerance by promoting evasion from NK cell antitumor immunity. J. Clin. Invest.
Mamessier, E., Pradel, L.C., Thibult, M.-L., Drevet, C., Zouine, A., Jacquemier, J., Houvenaeghel, G., Bertucci, F., Birnbaum, D., Olive, D., 2013. Peripheral blood NK cells from breast cancer patients are tumour-induced composite subsets. J. Immunol. 190 (5), 2424.
Mandapathil, M., Lang, S., Gorelik, E., Whiteside, T.L., 2009. Isolation of functional human regulatory T cells (Treg) from the peripheral blood based on the CD39 expression. J. Immunol. Methods 346 (1–2), 55.
Mantovani, A., Savino, B., Locati, M., Zammataro, L., Allavena, P., Bonecchi, R., 2010. The chemokine system in cancer biology and therapy. Cytokine Growth Factor Rev. 21 (1), 27.
Miller, J.S., 2001. The biology of natural killer cells in cancer, infection, and pregnancy.Exp. Hematol. 29, 1157.
Moretta, L., 2010. Dissecting CD56dim NK cells. Blood 116, 3689.
Mougiakakos, D., Choudhury, A., Lladser, A., Kiessling, R., Johansson, C.C., 2010. Regulatory T cells in cancer. Adv. Cancer Res. 107, 57.
Mozaffari, F., Lindemalm, C., Choudhury, A., Granstam-Björneklett, H., Helander, I., Lekander, M., Mikaelsson, E., Nilsso, B., Ojutkangas, M.-L., Osterborg, A., Bergkvist, L., Mellstedt, H., 2007. NK-cell and T-cell functions in patients with breast cancer: effects of surgery and adjuvant chemo- and radiotherapy. Br. J. Cancer 97, 105.
Naber, H.P., Wiercinska, E., ten Dijke, P., van Laar, T., 2011. Spheroid assay to measure TGF-β-induced invasion. J. Vis. Exp. 57, e3337. http://dx.doi.org/10.3791/3337.
Nath, S., Mukherjee, P., 2014. MUC1: a multifaceted oncoprotein with a key role in cancer progression. Trends Mol. Med. 20 (6), 332.
Neve, R.M., Chin, K., Fridlyand, J., Yeh, J., Baehner, F.L., Fevr, T., Clark, L., Bayani, N., Coppe, J., Tong, F., Speed, T., Spellman, P.T., Devries, S., Lapuk, A., Wang, N.J., Kuo, W., Stilwell, J.L., Pinkel, D., Albertson, D.G., Waldman, F.M., Mccormick, F., Dickson, R.B., Johnson, M.D., Lippman, M., Ethier, S., Gazdar, A., Gray, J.W., 2006. A collection of breast cancer cell lines for the study of functionally distinct cancer subtypes. Cancer Cell 10 (December), 515.
Nyga, A., Cheema, U., Loizidou, M., 2011. 3D tumour models: novel in vitro approaches to cancer studies. J. Cell Commun. Signal. 5, 239.
Olsen, C.J., Moreira, J., Lukanidin, E.M., Ambartsumian, N.S., 2010. Human mammary fibroblasts stimulate invasion of breast cancer cells in a three-dimensional culture and increase stroma development in mouse xenografts. BMC Cancer 10, 444.
Piersma, S.J., Welters, M.J.P., Van Der Burg, S.H., 2008. Tumour-specific regulatory T cells in cancer patients. Hum. Immunol. 69, 241.
Pinto, M., Jacobsen, B.M., Horwitz, K.B., 2011. An immunohistochemical method to study breast cancer cell subpopulations and their growth regulation by hormones in threedimensional cultures. Cell 2 (June), 1.
Poincloux, R., Collin, O., Lizárraga, F., Romao, M., Debray, M., Piel, M., Chavrier, P., 2011. Contractility of the cell rear drives invasion of breast tumour cells in 3D Matrigel. Proc. Natl. Acad. Sci. U. S. A. 108 (5), 1943.
Pollard, J.W., Iyengar, P., Combs, T.P., Shah, S.J., Albanese, C., Flanagan, L., Tenniswood, M.P., Guha, C., Lisanti, M.P., Pestell, R.G., Scherer, P.E., 2003. Adipocyte-secreted factors synergistically promote mammary tumorigenesis through induction of antiapoptotic transcriptional programs and proto-oncogene stabilization. Oncogene 22, 6408.
Pooi, P., Mohammad, R., Ibrahim, N., Ithnin, H., Abdullah, M., Davis, W.C., Seow, H.F., 2006. Phenotyping of lymphocytes expressing regulatory and effector markers in infiltrating ductal carcinoma of the breast. Immunol. Lett. 102, 229.
Prat, A., Perou, C.M., 2010. Deconstructing the molecular portraits of breast cancer. Mol. Oncol. 5 (1), 1.
Prestwich, R.J., Errington, F., Hatfieldy, P., Merrick, A.E., Ilett, E.J., Selby, P.J., Melcher, A.A., 2008. The immune system — is it relevant to cancer development, progression and treatment? Clin. Oncol. 20, 101.
Ralainirina, N., Poli, A., Michel, T., Poos, L., Emmanuel, A., Hentges, F., Zimmer, J., 2007. Control of NK cell functions by CD4+CD25+ regulatory T cells. J. Leukoc. Biol. 81, 144.
Roberti, M.P., Mordoh, J., Levy, E.M., 2012. Biological role of NK cells and immunotherapeutic approaches in breast cancer. Front. Immunol. 3, 375.
Robinson, S.C., Coussens, L.M., 2005. Soluble mediators of inflammation during tumour development. Adv. Cancer Res. 93, 159.
Sakaguchi, S., Sakaguchi, N., Asano, M., Itoh, M., Toda, M., 1995. Immunologic selftolerance maintained by activated T cells expressing IL-2 receptor α-chains (CD25). Breakdown of a Single Mechanism of Self-tolerance Causes.
Salagianni, M., Lekka, E., Moustaki, A., Iliopoulou, E.G., Baxevanis, C.N., Papamichail, M., Perez, S. a, 2011. NK cell adoptive transfer combined with Ontak-mediated regulatory T cell elimination induces effective adaptive antitumor immune responses. J. Immunol. 186 (6), 3327.
Sanchez, C., Le Treut, T., Boehrer, A., Knoblauch, B., Imbert, J., Olive, D., Costello, R., 2010. Natural killer cells and malignant haemopathies: a model for the interaction of cancer with innate immunity. Cancer Immunol. Immunother. 60 (1), 1.
Selander, K.S., Li, L., Watson, L., Merrell, M., Dahmen, H., Heinrich, P.C., Harris, K.W., 2004. Inhibition of gp130 signalling in breast cancer blocks constitutive activation of Stat3 and inhibits in vivo malignancy. Cancer Res. 64, 6924.
Sharan, S., Nikhil, K., Roy, P., 2013. Effects of low dose treatment of tributyltin on the regulation of oestrogen receptor functions in MCF-7 cells. Toxicol. Appl. Pharmacol. 269, 176.
Smyth, M.J., Cretney, E., Kelly, J.M., Westwood, J.A., Street, S.E.A., Yagita, H., Takeda, K., van Dommelen, Serani L.H., Degli-Esposti, M.A., Hayakawa, Y., 2005. Activation of NK cell cytotoxicity. Mol. Immunol. 42, 501.
Tate, C.R., Rhodes, L.V., Segar, H.C., Driver, J.L., Pounder, F.N., Burow, M.E., Collins-Burow, B.M., 2012. Targeting triple-negative breast cancer cells with the histone deacetylase inhibitor panobinostat. Breast Cancer Res. 14 (3), R79. http://dx.doi.org/10.1186/ bcr3192.
Ueda, S., Ogata, S., Tsuda, H., Kawarabayashi, N., Kimura, M., Sugiura, Y., Tamai, S., Matsubara, O., Hatsuse, K., Mochizuki, H., 2004. The correlation between cytoplasmic overexpression of epidermal growth factor receptor and tumour aggressiveness. Pancreas 29 (1), 1.
Vicari, A.P., Caux, C., 2002. Chemokines in cancer. Cytokine Growth Factor Rev. 13, 143.
Wei, X., Xu, H., Kufe, D., 2006. MUC1 oncoprotein stabilizes and activates oestrogen receptor α. Mol. Cell 21, 295.
West, N.R., Kost, S.E., Martin, S.D., Milne, K., Deleeuw, R.J., Nelson, B.H., Watson, P.H., 2013. Tumour-infiltrating FOXP3(+) lymphocytes are associated with cytotoxic immune responses and good clinical outcome in oestrogen receptor-negative breast cancer. Br. J. Cancer 108 (1), 155.
Whiteside, T.L., 2012. What are regulatory T cells (Treg) regulating in cancer and why? Semin. Cancer Biol. 22, 327.
Whiteside, T.L., Jost, L.M., Herberman, B., 1992. Tumour-infiltrating lymphocytes: potential and limitations to their use for cancer therapy. Crit. Rev. Oncol. Hematol. 12, 25.
Wilk, E., Kalippke, K., Buyny, S., Schmidt, R.E., Jacobs, R., 2008. New aspects of NK cell subset identification and inference of NK cells’ regulatory capacity by assessing functional and genomic profiles. Immunobiology 213, 271.
Ye, Q., Kantonen, S., Gomez-Cambronero, J., 2013. Serum deprivation confers the MDAMB-231 breast cancer line with an EGFR/JAK3/PLD2 system that maximizes cancer cell invasion. J. Mol. Biol. 425, 755.
Yu, P., Fu, Y., 2006. Tumour-infiltrating T lymphocytes: friends or foes? Lab. Investig. 86, 231.
Yu, X., Machesky, L.M., 2012. Cells assemble invadopodia-like structures and invade into Matrigel in a matrix metalloprotease dependent manner in the circular invasion assay. PLoS ONE 7 (2), e30605. http://dx.doi.org/10.1371/journal.pone.0030605.
Zhang, H., Grizzle, W.E., 2011. Exosomes and cancer: a newly described pathway of immune suppression. Clin. Cancer Res. 17, 959.