Cancer-associated fibroblasts induce high mobility group box 1 and contribute to resistance to doxorubicin in breast cancer cells

The human breast cancer cell line MDA-MB-231 was obtained from ATCC-LGC (#HTB-26, Middlesex, UK). Cells were cultured in DMEM (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; Gibco), 100 U/ml penicillin, 100 μg/ml streptomycin (Gibco), and anti-fungal agent. Cells were cultured in a 5% CO₂ in air incubator at 37°C and passaged by 0.25% trypsin-EDTA when they reached confluence. Cells of more than 90% viability evaluated by trypan blue dye exclusion were used in further experiments.

Primary cultures of breast cancer-associated fibroblasts (BCFs) were isolated from patients who underwent surgery at Siriraj Hospital, Mahidol University, Bangkok, and non-tumor-associated fibroblasts (NTFs) were isolated in each case from adjacent uninvolved breast tissue. The protocol for tissue collection was approved by the Siriraj Institutional Review Board (si498/2010) and informed consent was obtained from each of the six patients enrolled for this study. All breast cancers were of the hormone receptor positive luminal subtype. Briefly, sterile fresh surgical tissue was placed on ice in DMEM/F12 (Gibco) supplemented with 10X penicillin-streptomycin (1,000 U/ml penicillin and 1,000 μg/ml streptomycin) (Gibco). Tissues were washed 2 to 3 times by 1X phosphate buffered saline (PBS) to remove blood contamination. Tissues were then minced finely using a sterile surgical blade followed by enzymatic dissociation in collagenase type 1A (1,140 U/ml) (Sigma-Aldrich, St. Louis, MO, USA) diluted in DMEM/F12 supplemented with 10% FBS for 2 h at 37°C with agitation every 20 min. Next, tissues were digested in 0.05% trypsin-EDTA (Gibco) in serum-free DMEM/F12 for 10 min. The digestion solution was removed and fragments were washed with DMEM/F12 containing no FBS. The cell suspension was sequentially filtered through 100 μm and 70 μm nylon meshes (BD Biosciences, San Jose, CA, USA) and centrifuged at 2,000 xg for 5 min. The cell pellet was resuspended in complete DMEM/F12 media and cultured in a 25 cm² culture flask (Corning, NY, USA). Cells isolated from tissue samples were incubated in DMEM/F12 media containing 10% FBS for 10–14 days to allow attachment and the formation of colonies; these primary cultures were designated as passage 0. All cultures were kept in a humidified incubator with 5% CO₂ in air at 37°C. Cells were subcultured when 80% confluent, banked and used for characterization and experimental studies at passages 5–13.

To discriminate BCFs and NTFs from cancer cells, immunohistochemical staining for epithelial CK19 and mesenchymal ASMA markers were performed. The breast cancer cell line, MDA-MB-231 was used as a positive control for CK19 detection. In brief, around 4,000 cells were plated into each well of a 96-well plate and cultured for 24-h to allowed for cell adhesion. Cells were then fixed with 4% paraformaldehyde. Non-specific binding was blocked by incubating cells with 1% BSA in 1X PBS for ASMA detection or 5% FBS in 1X PBS for CK19 detection. Mouse anti-human CK19 antibody (SC-6278; 1:100 dilution, Santa Cruz Biotechnology Inc., Dallas, TX, USA) or mouse anti-human ASMA antibody (A5228, 1:200 dilution, Sigma-Aldrich) was added for 3 h at room temperature. A blocking reagent was used as the negative control in place of the primary antibody. After washing with 1X PBS, goat anti-mouse IgG-Cy3 antibody (#115-166-071, 1:2,000, Jackson ImmunoResearch Laboratories Inc, West Grove, PA, USA) was applied for 1 h at RT. The signals were detected by fluorescence microscopy.

Cultures of BCF and NTF were grown in 75-cm² flasks to reach 90-95% confluency in DMEM (containing 10% FBS) which is a suitable media for MDA-MB-231 cells. The conditioned-media (CM) were collected 24 h following addition of fresh complete medium and designated as 24-h CM. CMs were centrifuged at 2,000 g for 5 min to remove cell debris and the suspension stored at -80°C or -20°C until use.

MDA-MB-231 cells were cultured in a 6-well plate until approximately 90-100% confluent. A reference line was drawn across the bottom of the plate. A scratch wound was made in the cell monolayer with a sterile 200-μl pipette tip and the culture was then washed three times with serum-free medium to remove the detached cells. The cells were then treated with BCF-CM, NTF-CM or complete medium as a negative control. The scratch wound indicated by the reference line was imaged at the start of the treatment and 6 h later. The cell migration efficiency was determined as a percentage of wound healing calculated by the following formula using three different zones for each condition:

Total RNA was extracted from MDA-MB-231 breast cancer cells using the PerfectPure RNA Cultured Cell Kit (5 Prime; Gaithersburg, MD, USA) as per the manufacturer’s instructions. The OD260/280 and OD260/230 were measured to ensure the quality of extracted RNA. Complementary DNA was synthesized from 1 μg of total RNA using the SuperScript™ III First-Strand Synthesis System for RT-PCR (M-MLV) (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Expression levels of HMGB1 were determined by SYBR Green-based real time PCR (Roche Applied Sciences, Indianapolis, IN, USA) in a Light Cycler® 480 II machine (Roche Applied Sciences). Optimal primers were designed using the nucleotide database in PubMed and Primer 3 software. Sequences of primers were: HMGB1 (NM_002128.4): forward primer: 5'-CACTGGGCGACTCTGTGCCTCG-3', reverse primer: 5'-CGGGCCTTGTCCGCTTTT-GCCA-3'. β-actin (ACTB) served as an internal control to adjust the amount of starting cDNA. The expression of each gene in breast cancer cells was calculated by the 2^-ΔCp equation. In this case, ΔC _p= C_p(HMGB1) - C_p(ACTB). The expression of HMGB1 in breast cancer cells treated with fibroblast CM or doxorubicin (Dox) (Pfizer, Perth Pty Ltd, Bentley WA, Australia) compared with that in untreated control cells was calculated by the 2^-ΔΔCp equation. In this case, ΔC _p= C_p(HMGB1) - C_p(ACTB) and ΔΔCp= ΔC _p(treated cells) - ΔC _p(control cells).

MDA-MB-231 breast cancer cells were treated with BCF-CM or NTF-CM for 48 hr. Cell suspensions were centrifuged at 2,000 ×g for 5 min in a refrigerated centrifuge. The cell pellets were collected and rinsed in cold 1X PBS twice before lysis in 1X sample buffer containing 50 mM Tris–HCl pH 6.8, 2% (w/v) SDS, 10% (v/v) glycerol, 5% (v/v) β-mercaptoethanol and 0.05% (w/v) bromophenol blue. Cell lysates were boiled for 10 min and centrifuged at 10,000 rpm for 1 min to remove undissolved proteins and cell debris. Cell extracts were then separated by 10% SDS-PAGE and transferred onto PVDF membranes (GE Healthcare, Buckinghamshire, UK). Membranes were blocked in 5% skimmed milk containing TBST (TBS containing 0.1% Tween 20) for 1 h at room temperature. Membranes were then washed 3 times with 1X TBST and incubated with 1:1,000 rabbit anti-human HMGB1 (ab18256, Abcam, Cambridge, CB4 OFL, UK) at 4°C overnight. After washing, membranes were incubated with 1:2,000 goat anti-rabbit IgG-HRP (Abcam) at room temperature for 1 h. The blots were visualized by enhanced chemiluminescence (Thermo Scientific, Rockford, IL, USA). Using 1:10,000 anti-β-actin antibody (sc47778, Santa Cruz Biotechnology Inc.) or 1:10,000 anti-β-actin antibody (ab8226, Abcam) with the suitable HRP conjugated secondary antibody, β-actin protein levels were used as an internal control to confirm equal protein loading. β-actin normalized HMGB1 levels in breast cancer cells without any CM treatment were used as controls.

The same procedure was used for the measurement of extracellular HMGB1 except for the process of sample collection. To measure the amount of HMGB1 released from cells, the media from cells stimulated with either BCF-CM or NTF-CM were collected and concentrated using a 10 kDa cut-off Vivaspin concentrator (Sartorius Stedim Biotech, Goettingen, Germany) at 3,500 ×g. Protein concentration was determined by a Coomassie Plus (Thermo Scientific) assay kit. The same amount of total protein was loaded into SDS-PAGE gels and the procedure for detection of HMGB1 above.

For LC3B autophagy protein detection, cells were treated with 100 ng/ml human rHMGB1 (#1690-HMB-050, R&D Systems, Minneapolis, MN, USA) with or without anti-HMGB1 neutralizing antibody (H00003146-M08, Novus, Littleton, CO, USA) before exposure to Dox for 24 h. Cells were harvested and total proteins were separated and blotted on to the membrane as described above. The rabbit anti-human LC3B (#2775, Cell Signaling Technology, Danvers, MA, USA) was then incubated at 1:2,000 dilution with the membranes at 4°C overnight followed by goat anti-rabbit IgG-HRP (#7044, Cell Signaling Technology; 1:1,200 dilution) at room temperature for 1 h. The blots were visualized as described previously.

MDA-MB-231 breast cancer cells were cultured in 6-well plates (3 × 10⁵/well) for 48 h, then the growth medium was removed and the cells washed thoroughly with PBS. CM from either BCF or NTF was added to the cells which were incubated for a further 48 h at 37°C in a 5% CO₂ in air incubator. Negative controls were cultured in parallel in DMEM plus 10% FBS. Cells were then harvested to determine the level of intracellular HMGB1 expression by real time PCR and Western blot analysis. At the same time, the culture media were also collected to investigate the level of extracellular HMGB1. Similarly, 1 and 5 μM doxorubicin Dox (Pfizer) was used to treat MDA-MB-231 breast cancer cells for 48 h. Dox-treated cells were harvested to determine the level of intracellular HMGB1 by real time PCR and the culture medium was also collected to investigate the level of extracellular HMGB1. In addition, cancer cells with or without pre-treatment with fibroblast CMs were then exposed to Dox for 48 h to induce cell death and the release of HMGB1. Cell viability was checked by trypan blue exclusion and the levels of HMGB1 were determined by Western blot analysis as above.

MDA-MB-231 cells were seeded into 96-well plates at 6,000 cells per well and cultured overnight in DMEM + 0.1% FBS. Cells in triplicate wells were then treated with 100 ng/ml rHMGB1 (R&D Systems) in the presence or absence of 10 mg/ml HMGB1-neutralizing antibody (Novus) and then 5 μM Dox or vehicle. Cell viability was measured 24 h later by trypan blue exclusion.

MDA-MB-231 cells were plated at a density of 6,000 cells/well in 96-well plates. These cells were treated for 24 h with CM collected from MDA-MB-231 cells exposed to 5 μM Dox (for 48 h), designated as ‘dead cancer-CM’, with or without anti-HMGB1-neutralizing antibody (10 mg/ml) or isotype-matched control IgG (X0943, Dako, Agilent Technologies, Glostrup, Denmark). Cell viability after exposure to 5 μM Dox was analyzed by erythrosine B dye exclusion and compared with control untreated cells. Three independent experiments were performed.

The values are expressed as mean ± SD. Statistical significance was determined by Student’s t-test. A p-value of equal to or less than 0.05 was defined as statistically significant.

BCFs and NTFs were characterized by their expression of the mesenchymal marker, ASMA, and absence of the epithelial marker, CK19. Immunocytochemical staining revealed that all cancer-associated and ‘normal’ breast fibroblasts from six different patients were negative for CK19 compared with the positive control MDA-MB-231 breast cancer cells (Figure 1) and were positive for ASMA. Thus we confirmed that both BCFs and NTFs were mesenchymally-derived cells with no epithelial cell contamination.To ensure that the BCFs were activated and capable of promoting malignant potential, the effects of CM on MDA-MB-231 breast cancer cell migration were tested. The results indicated that the BCF-CM promoted cancer cell migration to a significantly greater degree than NTF-CM (Figure 2). Indeed, NTF-CM had a minimal effect compared with untreated control cells.

BCF-CM significantly induced intracellular HMGB1 protein expression in MDA-MB-231 breast cancer cells as detected by Western blot analysis at all time points tested (Figure 3). The effect was time-dependent and since the greatest differential induction (BCF-CM vs NTF-CM) was observed at 48 h, this time period was selected for further studies. As a further quality control, the CMs of BCF and NTF isolated from the same patient were used to treat MDA-MB-231 cells and HMGB1 gene expression was analyzed by real time PCR. The results showed that BCF-CM induced HMGB1 mRNA to a significantly greater degree than NTF-CM (Figure 4A). Western blot analysis confirmed that the protein levels of HMGB1 induced by BCF-CM were statistically significantly higher than those induced by patient-matched NTF-CM (Figure 4B). In addition, HMGB1 protein levels in MDA-MB-231 cells treated with BCF-CMs from different patients were consistently significantly higher than those treated with NTF-CMs.

Doxorubicin is commonly used in breast cancer treatment and our results using real time PCR showed that this drug could induce intracellular HMGB1 expression in MDA-MB-231 cells in a concentration-dependent manner (Figure 5A). The maximal level of HMGB1 was induced with 5 μM which was statistically significantly different from untreated controls. Moreover, cancer cells killed by Dox exposure released HMGB1 into the culture media and the level was again increased in a concentration-dependent manner (Figure 5B).BCF-CM-pretreated cancer cell cultures showed less cell death in response to Dox than cells pre-treated with NTF-CM (Figure 5C). In a second study, we found that BCF-CM treated cells also released more HMGB1 than those pre-treated with NTF-CM when treated with equitoxic concentrations of Dox (80% cell death) (Figure 5D). No HMGB1 was detected in the culture media when cell viability was greater than 95% but in contrast, Dox-induced release of HMGB1 was related to the degree of cell death (data not shown).

MDA-MB-231 cells exposed to Dox together with rHMGB1 showed statistically significantly higher viability than those treated with Dox alone (Figure 6A). This effect was reversed by the addition of an HMGB1 neutralizing antibody. The fact that LC3B-I converts to LC3B-II and levels of LC3B-II increase over time suggests that autophagy occurs in MDA-MB-231 cells treated with rHMGB1 (Figure 6B).

Dead cancer-CM, (shown to increase the level of extracellular HMGB1; Figure 5B) also increased the survival of MDA-MB-231 cells during subsequent Dox treatment (Figure 6C) and this effect was reversed by an HMGB1 neutralizing antibody (**p-value = 0.006). The blocking antibody also showed a small but significant reduction in cell viability in the absence of CM (p-value = 0.017).