Hyaluronan (HA) has been shown to play crucial roles in the tumorigenicity of malignant tumors. Previous studies demonstrated that inhibition of HA suppressed the tumorigenicity of various malignant tumors including breast cancer. 4-methylumbelliferone (MU) has been reported to inhibit HA synthesis in several cell types. However, few studies have focused on the effects of HA inhibition in breast cancer cells by MU, nor the effects on bone metastasis. We hypothesized that MU would suppress the progression of bone metastasis via inhibition of HA synthesis. Here, we investigated the effects of MU on HA expression in MDA-MB-231 breast cancer cell line in addition to their tumorigenicity in vitro and in vivo. HAS2 mRNA expression was downregulated after 6 and 24 hr treatment with MU. Quantitative analysis of HA revealed that MU significantly inhibited the intracellular and cell surface HA. MU significantly inhibited cell growth and induced apoptosis as determined by cell proliferation and TUNEL assays, respectively. Phosphorylation of Akt was suppressed after 12 and 24 hr treatment with MU. MU treatment also inhibited cell motility as well as cell invasiveness. MU also inhibited cell growth and motility in murine fibroblast cell line NIH3T3. In vivo, administration of MU inhibited the expansion of osteolytic lesions on soft X-rays in mouse breast cancer xenograft models. HA accumulation in bone metastatic lesions was perturbed peripherally. These data suggest that MU might be a therapeutic candidate for bone metastasis of breast cancer via suppression of HA synthesis and accumulation.
Breast cancer is the most common cancer in women particularly in western countries.1, 2 Approximately 6% of breast cancer patients present to hospital with distant metastasis2 and about 25% despite neoadjuvant or adjuvant systemic treatment develop distant metastasis.3 In a retrospective study of patients with metastatic breast cancer, bone was the most common site of metastatic spread and 70% of these patients developed metastases in one or more bones before they died.4 Bone metastases occasionally result in hypercalcemia, pathologic fractures and spinal cord compression,5 and substantially reduce patients' quality of life.6
Hyaluronan (HA) is a high molecular weight linear glycosaminoglycan, the structure of which is composed of repeating disaccharides of D-glucuronic acid and N-acetyl-D-glucosamine. HA is a ubiquitous extracellular and cell surface associated matrix component of connective, epithelial and neural tissues.7 HA is abundant in surrounding migrating and proliferating cells during morphogenesis and wound healing.8, 9 Increased HA levels are also observed, both in stroma of malignant areas and in a part of tumor parenchyma in some malignant tumors including colorectal, ovarian and breast cancer and liposarcoma.10–14 Not only the HA accumulation level in the stroma of malignancy11, 15, 16 but also the HA level in tumor parenchyma10, 13 has been reported to associate with a poor clinical outcome. In in vitro experiments, the interactions between tumor cells and fibroblasts were reported to stimulate HA synthesis17, 18 but tumor cells themselves also have a potential to increase HA synthesis with increasing phenotypical aggressiveness.19, 20 Three eukaryotic HA Synthase (HAS) isoforms have been identified, and termed HAS1, HAS2 and HAS3.21 Several studies have demonstrated that the manipulation of HAS genes alters the tumorigenicity of malignant tumors,22–27 including breast cancer.28, 29 However, genetic manipulation cannot be easily applied clinically.
4-Methylumbelliferone (MU) has been reported to inhibit HA synthesis dose-dependently in several cell types.30–33 This inhibitory effect of MU on HA synthesis shows anti-tumor effects on cell proliferation, migration and invasion in vitro.32, 33 In breast cancer, additive effects of MU on trastsuzumab treatment were reported in trantsuzumab resistant tumor in vitro and in vivo.34 The mechanism of MU underlying the inhibition of HA synthesis has not been completely elucidated. A recent report showed that MU inhibited HA synthesis by depletion of cellular UDP-glcuronic acid and downregulation of HAS2 and/or HAS3 in tumor cells.32 Another recent report described that MU increased UDP-glucuronyl transferase 1 enzymes, which can reduce UDP- glcuronic acid, and reduced HAS-2 expression in human smooth muscle cells.31 We questioned and investigated in this study whether MU inhibits HA synthesis in breast cancer cells and if so whether this is mediated by inhibition of HAS gene expression, and whether HA inhibition by MU administration suppresses tumorigenicity of breast cancer cells and progression of bone metastasis in an in vivo model. Further, to investigate MU effects on tumor stromal cells, HA accumulation, cell growth and motility were analyzed using murine fibroblast cell line NIH3T3 cells.
Material and Methods
The human breast cancer cell line, MDA-MB-231 and the murine fibroblast cell line, NIH3T3 were obtained from the American Type Culture Collection (Manassas, VA). The cells were grown as monolayers in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% Fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin. The cultures were maintained in a humidified atmosphere with 5% CO2 at 37°C.
MU stock solution for vitro experiments was dissolved in Dimethylsulfoxide (DMSO), and the final concentration of DMSO in medium was adjusted to 1.0%. All incubations were performed in the presence of 10% FBS.
Real time RT-PCR analysis
Expression levels of HAS1, HAS2, HAS3 and CD44 were determined in MDA-MB-231 human breast cancer cell line after treatment with or without 0.1 mM and 1.0 mM MU for 6 and 24 hr. Total cellular RNA was isolated using RNeasy Mini Kit (Qiagen), following the manufacturer's instructions. Following the conventional reverse transcriptase-polymerase chain reaction (RT-PCR), cDNA was subjected to real time RT-PCR for semi-quantification of HAS1, HAS2, HAS3 and CD44 mRNAs using a LightCycler (Roche Diagnostics, Mannheim, Germany). The relative levels of HAS1, HAS2, HAS3 and CD44 mRNA in a sample were expressed in relative quantification normalized against GAPDH mRNA. The HAS1, HAS2, HAS3, CD44 and GAPDH primer pairs were as follows: HAS1 sense; 5′-CAGACCCACTGCGATGAGAC-3′, HAS1 antisense; 5′-CCACCAGGTGCGCTGAAA-3′ (predicted PCR product of 218bp), HAS2 sense; 5′-TCAGAGCACTGGGACGAAG-3′, HAS2 antisense; 5′-CCCAACACCTCCAACCAT-3′ (predicted PCR product of 125bp), HAS3 sense; 5′-CAGCAACTTCCATGAGGC-3′, HAS3 antisense; 5′-CACAGTGTCAGAGTCGCA-3′ (predicted PCR product of 202bp), CD44 sense; 5′-TTGCAGTCAACAGTCGAA-3′, CD44 antisense; 5′-TTCTGACGACTCCTTGTTC-3′ (predicted PCR product of 155bp) and GAPDH sense; 5′-TGAACGGGAAGCTCACTGG-3′, antisense; 5′-TCCACCACCCTGTTGCTGTA-3′ (predicted PCR product of 307bp). The mRNA level at each time was shown as a percent of the control with culture including 1% DMSO.
Quantifications of HA
The subconfluent MDA-MB-231 cells were incubated with or without 1.0 mM MU for 6, 12 and 24 hr. The isolation of HA was based on the methods reported by Tammi et al.35 Briefly, the conditioned medium was collected and designated as “medium.” To remove the cell-surface associated HA, the cells were incubated for 10 min at 37°C with trypsin-EDTA and washed with PBS. The trypsin solution and combined washes were designated as “pericellular.” After cell counts, the cells were placed in Protease K solution (0.15 M Tris-HCl, pH 7.5, 0.15 M NaCl, 10 mM CaCl2 and 5 mM deferoxamine mesylate containing 20 units of protease K) and incubated for 2 hr at 55°C and the solution was designed as “intracellular.” All samples were heated at 100°C for 15 min to inactivate protease activity and centrifuged at 15,000g for 30 min at 4°C, and the supernatants were analyzed. HA concentrations were measured using a sandwich enzyme-linked immunosorbent assay. The method was described previously.36 Briefly, each wall in a Maxisorp microtiter plate (Nunc, Roskilde, Denmark) was coated with 50 μl of HABP (Seikagaku, Tokyo, Japan; 0.4 μg/ml in 0.1 M NaHCO3, pH9.2), incubated overnight at 4°C, and followed by blocking with 200 μl each of 2% BSA in phosphate-buffered saline with 0.1% (v/v) Tween 20 (PBS-T) at room temperature for 2 hr. After washing with PBS-T, samples (50 μl each) and standard HA solutions were applied to the wells, followed by incubated by at 37°C for 1 hr. After washing, 50 μl of biotinylated-HABP (Seikagaku, Tokyo, Japan; 0.3 μg/ml, diluted with 1% BSA/PBS-T) was added to each well, followed by incubation at 37°C for 1 hr. Then, peroxidase-conjugated streptavidin (Jackson ImmunoResearch Laboratories; diluted 1:2,000 with 1% BSA/PBS-T) was added (50 μl each), followed by incubation at 37°C for 1 hr. Finally, color development was achieved by incubation with 50 μl of 3,3′,5,5′-tetramethylbenzidine substrate (KPL, Gaithesburg, CA) at 37°C for 10 min and then stopped by adding 50 μl of 1 M HCl. The absorbance at 450 nm was measured on a VERSA max microplate reader (Molecular Devices, Sunnyvale, CA).
Cell growth assay
MDA-MB-231 cells and NIH 3T3 cells were seeded in 96-well plates at 1 × 104/well in medium supplemented with 10% FBS and allowed to adhere for 6 hr. The subconfluent cells were exposed to 10% FBS medium with and without DMSO containing 0–1.0 mM MU. After treatment for 24, 48 and 72 hr, cell proliferation was measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5,diphenyl tetrazolium bromide (MTT) colorimetric assay using Cell proliferation Kit I (Roche Diagnostics). Color intensity was determined on a microculture plate reader (TOSOH, Tokyo, Japan) at 550 nm. We also investigated whether exogenously added high-molecular-weight hyaluronan (HMWHA, 600–1,200 kDa, Seikagaku, Tokyo, Japan) neutralize the effect of MU. MDA-MB-231 cells were incubated with 0–1.0 mM MU with or without 200 μg/ml of HMWHA for 24, 48 and 72 hr.
Cell cycle analysis
The effect of MU on the MDA-MB-231 cell cycle was evaluated by flow cytometry. MDA-MB-231 cells were treated with or without 1.0 mM MU for 24 hr. The cells were washed with phosphate-buffered saline (PBS), trypsinized and stained with propidium iodide using a CycleTEST PLUS DNA reagent kit (Becton Dickinson Immunocytometry Systems). The DNA content of the stained cells was immediately analyzed using a FACSCalibur (Becton Dickinson). The percentages of cells in G0/G1 phase, S-phase and G2/M phase were determined using ModiFit LT software (Verity Software House, Topsham, ME).
TUNEL (TdT: terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling) staining was used to evaluate the apoptotic effect of MU on MDA-MB-231 cells. The subconfluent cells were incubated with or without 0.1 and 1.0 mM MU for 72 hr, and subjected to TUNEL staining using an In Situ Cell Death Detection Kit, POD (Roche Diagnostics). Cells with brown-stained nuclei in ten different fields were counted under a light microscope at 400× magnification, and the percentage of positive-staining cells was calculated.
Western blot analysis
For protein extraction, MDA-MB-231 cells were incubated with or without 1.0 mM MU for 30 min, 3, 6, 12 and 24 hr and the cells were lysed on ice in RIPA buffer (Santa Cruz Biotechnology). After centrifugation at 10,000g for 10 min, the supernatant was subjected to Western blot analysis with rabbit anti human p-Akt (Ser473) polyclonal antibody (Cell Signaling Technology, Beverly, MA; 1:1,000 dilution), anti human Akt polyclonal antibody and anti human β-actin monoclonal antibody (Cell signaling Technology; 1:1,000 dilution, respectively), was finally visualized with enhanced chemiluminescence reagents using Super Signal West Pico Trial Kit (Thermo) and immediately photographed with a charge coupled device (CCD) digital scan camera (Cool Saver, Rise & ATTO Corporation, Tokyo, Japan). Densitometric analysis of immunoreactive areas was performed, and the results are expressed as the index (p-Akt/Akt).
Motility and matrigel invasion assays
Chemotactic motility of MDA-MB-231 cells and NIH3T3 cells were investigated using 12-well cell culture chambers containing inserts with 12-μm pores (Millipore, Billerica, MA). Invasion of MDA-MB-231 cells was assayed in the same chambers that contained the inserts with 12-μm pore membrane coated with matrigel. The cells were added to the upper chamber at a density of 5 × 105 cells/insert in the presence or absence of 0.1, and 1.0 mM MU with or without 200 μg/ml HMWHA, and chemotaxis buffer containing 10 μg/ml of fibronectin was placed in the lower chamber. After 24 hr of incubation, cells on the upper surface were wiped off with a cotton swab. Cells that had invaded the lower surface were fixed with 70% ethanol and stained with hematoxylin, and ten different fields were counted under a light microscope at 200× magnification.
Effects of MU in vivo
To establish a murine bone metastasis model, using a 29-gauze sterile needle with syringe (Becton, Dickinson and Company, Franklin Lakes, NJ), a single-cell suspension (1 × 105 cells in 10 μl PBS) of MDA-MB-231 cells was carefully injected into the bone marrow cavity of tibial plateau in anesthetized 4-week-old nude mice (BALB/C nu/nu mice). After tumor cell inoculation, formation of osteolytic lesions was evaluated with Industrial X-ray Film IX FR (FUJIFILM, Tokyo, Japan) by soft X-rays (SOFTEX, Yokohama, Japan) every 7 days. After confirmation of the development of osteolytic lesions, the mice were randomly divided into two groups, a control group and a group treated with MU, each containing 10 mice. MU (10 mg per mouse) with 100 μl of 0.4% Carboxymethyl cellulose sodium (CMC) solution (control; 0.4% CMC solution only) was administered daily to mice intraperitoneally. After 14 days of consecutive administration, the mice were subjected to soft X-ray evaluation and sacrificed for histological examination. The soft X-ray evaluation was performed by lateral views of tibia, and the osteolytic areas on radiographs were obtained with a scanner (ES-8000; EPSON, Tokyo, Japan) and measured using image analysis software (Scion Image, Scion Corporation, Frederick, MD). Ratios of osteolytic area to whole tibia before and after treatments were calculated and compared. The body weight of mice was measured once a week to evaluate the side effects of MU. Further, to evaluate the effects of consecutive administration of MU on normal tissues synthesizing HA, elbow joints in mice treated or untreated with MU for 14 days were obtained and subjected to staining for Alcian blue. All animal care and experimentation were performed according to the study guidelines established by the Division of Experimental Animals in Nagoya University for animal care, handling and termination.
HA staining for cells and tissues
HA accumulation in MDA-MB-231 cells, NIH3T3 cells, and local tumor in vivo model was visualized using biotinylated Hyaluronic Acid Binding Protein (b-HABP; Seikagaku, Tokyo, Japan). The cells were seeded into chamber slides (BD Biosciences) and treated with or without 0.1 and 1.0 mM MU for 24 hr. Moreover, cells were incubated with 1.0 mM MU with or without 200 μg/ml HMWHA for 24 hr to determine the exogenous HA effects on HA accumulation. Cultured MDA-MB-231 cells were fixed with 4% of paraformaldehyde buffered with PBS at room temperature for 2 hr. The tissue sections of cartilage were pretreated with 1 U/ml Chondroitinase ABC (pH 8.0) for 2 hr at 37°C. The cells and tissue sections of in vivo model were subjected to the incubation with 2.0 μg/ml b-HABP probe for 2 hr at room temperature. Bounded b-HABP was detected by the addition of streptavidin-peroxidase reagents (Nichirei, Tokyo, Japan) and diaminobenzidine-containing substrate solution (Nichirei). The skin tissues and cartilages of mice treated or untreated with MU were also stained using b-HABP for the evaluation of the systemic effects of MU.
CD44 staining for cells
CD44 expression on MDA-MB-231 cells was visualized using biotin-conjugated anti-CD44 monoclonal antibody, IM7 (StemCell Technologies, Seattle, WA). The cells were seeded into chamber slides (BD Biosciences) and treated with or without 1.0 mM MU for 24 hr, and subjected to immunocytochemistry using 5.0 μg/ml of anti-CD44 antibody.
All of the quantitative experiments in vitro were performed more than three times and analysis of variance, followed by Bonferroni-Dunn post hoc test, was used to assess differences between means. Statistical comparison between two groups (n = 10/group) in vivo was made using an unpaired Student's t test.
Effect of MU on HA accumulation in MDA-MB-231 cells
Abundant accumulation of cell associated HA was shown in MDA-MB-231 cells after incubation with control medium with or without DMSO (Figs. 1a and 1b). Stainable HA was decreased after 24 hr treatment with MU, and the decrease of HA accumulation was significant in 1.0 mM MU (Fig. 1c). The MU treatment also caused a loss of lamellipodia, and the cells were circular. Decrease of HA accumulation was not recovered by exogenously added 200 μg/ml of HMWHA (Fig. 1d). HA quantifications revealed that there was more intracellular HA than pericellular HA and HA in medium in MDA-MB-231 cells (Figs. 1e–1g). MU suppressed the mean amount of HA in medium and intracellular and pericellular HA at every time point, except at 24 hr of pericellular HA. Intracellular and pericellular HA were significantly decreased after 6 and 12 hr treatment with 1.0 mM MU, respectively (p < 0.05 respectively, compared with control of 1% DMSO).
HAS and CD44 mRNA levels after MU treatment in MDA-MB-231 cells
MDA-MB-231 cells are known to express HAS2 and HAS3,20, 29 which was consistent with the results that no expression of HAS1 was detected in MDA-MB-231 cells in the current study. Treatment with 0.1 and 1.0 mM MU significantly downregulated mRNA expression of HAS2 after 6 and 24 hr of treatment (Fig. 2a). After 6 hr, the 0.1 and 1.0 mM MU exhibited the significant effect of 18 and 84% inhibition of HAS2 mRNA, respectively (both p < 0.0001, compared with the 1% DMSO control), and a 17 and 52% reduction following treatment for 24 hr, respectively (p < 0.05 and p < 0.0001, respectively). There was no difference in HAS3 mRNA expression between the treatment with 0.1 mM and 1.0 mM MU and control at 24 hr (Fig. 2a). Expression level of CD44 mRNA in cells treated with 1.0 mM MU was significantly lower by 30% as compared with the control at 24 hr (p < 0.0001; Fig. 2b). CD44 protein expression was evaluated by immunocytochemistry. There was no evident difference of positivity between cells with and without MU treatment (Figs. 2d and 2c, respectively).
Effects of MU on cell growth in MDA-MB-231 cells
MU inhibited MDA-MB-231 cell proliferation dose-dependently (Fig. 2e). The 0.8 and 1.0 mM MU exhibited the greatest effect of 36 and 43% inhibition following 48 hr of treatment, respectively (p < 0.001 and p < 0.0001, respectively, compared with control with DMSO), and a 38 and 29% reduction following treatment of 72 hr, respectively (p < 0.0001 and p < 0.001, respectively). Less inhibition was observed following treatment of 48 and 72 hr with 0.4 mM MU (both p < 0.05). Exogenously added HMWHA did not recover the inhibition of cell growth by MU (Fig. 2h). On the basis of the results of cell growth, the following experiments were analyzed with culture containing 0.1 or 1.0 mM of MU as compared with the control culture.
Effects of MU on cell migration and invasiveness in MDA-MB-231 cells
The treatment of 0.1 mM MU exhibited a statistically significant reduction (18%) in migration activity at the 24-hr time point compared with the 1% DMSO control in MDA-MB-231 cells (p < 0.01; Fig. 2f). The treatment with 1.0 mM MU also showed a 19% reduction in migration activity at the 24-hr time point compared with control (p < 0.001). Exogenously added HMWHA did not recover the suppressed migration by MU (Fig. 2i). In the invasion assay, the 0.1 mM and 1.0 mM MU-treated cells exhibited a significantly lower capacity (51% and 39% compared with control, respectively) to pass through the matrigel-coated filters as compared with control cells in MDA-MB-231 cells (p < 0.01 and p < 0.001, respectively; Fig. 2g).
Effects of MU on cell cycle in MDA-MB-231 cells
MDA-MB-231 cells treated with 1.0 mM MU for 24 hr showed a statistically significant decrease in cell number in the G2/M phases compared with control cells (p < 0.001; Fig. 3a), and the numbers of cells in the S-phase in cultures treated with 1.0 mM MU were significantly lower as compared to the control cells (p < 0.001).
Effects of MU on apoptotic activity in MDA-MB-231 cells
The percentage (mean ± SD) of apoptotic MDA-MB-231 cells exposed to 0.1 and 1.0 mM MU was significantly increased as compared with the control containing DMSO (p < 0.001, respectively; Fig. 3b).
Effects of MU on Akt phosphorylation in MDA-MB-231 cells
There was no difference in Akt expression between the treatment with and without MU at any time, but intracellular p-Akt was significantly downregulated after 12 and 24 hr treatment with MU (p < 0.001, respectively; Figs. 3c and 3d), which was consistent with the results of increased apoptotic activity at 24 hr time point with MU treatment. There was no difference in P-Akt expression after treatment for 30 min, 3 or 6 hr.
Effects of MU on HA accumulation, cell growth and migration in NIH3T3 cells
Abundant accumulation of cell-associated HA was observed in NIH3T3 mouse fibroblast cells after incubation with control of 1% DMSO (Fig. 4a), and stainable HA was decreased after 24 hr treatment with 0.1 mM MU (Fig. 4b). MU (0.2–1.0 mM) significantly inhibited the growth of NIH3T3 cells after 24 hr treatment (p < 0.001, respectively; Fig. 4c), but less inhibition was observed following treatment of 48 and 72 hr with MU (data not shown). NIH3T3 cells with treatment of 0.1 mM or 1.0 mM MU exhibited statistically significant reductions (17% or 61%, respectively) in migration activity at the 24-hr time point compared with cells with 1% DMSO control (p < 0.01 and p < 0.001, respectively; Fig. 4d)
Effect of MU on mouse xenograft model of tibia
Osteolytic areas were confirmed at an average 5.5 (4–7) and 5.9 (4–8) weeks after tumor inoculation in the control and MU-treated groups, respectively (p = 0.553; mice were equally divided in each group). The mean ratios of osteolytic area per whole tibia before treatment were 0.14 (0.07–0.22) in the control and 0.12 (0.05–0.20) in MU-treated groups (p = 0.482). No marked body weight loss was observed in the mice either treated or untreated with MU (Fig. 5e). The effect of MU was studied on the spread of osteolytic lesions by soft X-rays (Figs. 5a and 5b, control mouse, Figs. 5c and 5d, MU-treated mouse). Fourteen days administration of MU resulted in a significant 25% reduction in the progression of osteolytic lesions compared with control mice (p < 0.01; Fig. 5f).
Effect of MU on HA accumulation in tumorous tissues and extracellular matrix in articular cartilage
In control mice, HA strongly accumulated in tumor stroma, particularly at the border with host bone (Fig. 6a), and the accumulation of HA was pronounced in tumor stroma, with some tumor cells also stained at the plasma membrane and cytoplasm by HABP staining (Fig. 6b). In contrast, the accumulation of HA in tumor adjacent to bone decreased after daily administration of MU for 14 days (Fig. 6c). There was no difference either in cartilage of elbow between control and MU-treated mice by Alcian blue (Figs. 6d and 6e) and b-HABP staining (Figs. 6f and 6g). The HA accumulation in dermis slightly decreased in mice treated with MU compared with control (Figs. 6f and 6g, inset). However, no difference between control and MU-treated mice was noted under gross and histological observations of skin.
Several previous studies have reported that HA plays crucial roles in the proliferation, progression and invasion of breast cancer cells. Expression levels of HA and HAS2 are increased in highly invasive breast cancer cell lines such as MDA-MB-231 cells.20 Among three HAS enzymes, HAS2 has been shown to be critical to the progression of tumorigenicity in MDA-MB-231 cells whereas antisense inhibition of this enzyme attenuated cellular proliferation and invasiveness in vitro and inhibited the formation of primary and secondary tumors in vivo.28 In another invasive breast cancer cell line, silencing of HAS2 also suppressed anchorage-dependent and anchorage-independent cell growth, cell migration and invasion.29 MU treatment has been reported to inhibit HA synthesis by depletion of cellular UDP-glucuronic acid and downregulation of HAS2 and/or HAS3 and inhibit cell proliferation, migration and invasiveness.32 In the study, a dose-dependent reduction in the mRNA levels of HAS2 or HAS3, or both, was detected in four different cancer cell lines, and alteration of mRNA levels by MU depended on the kind of the tumor. However, no previous reports have investigated the roles of HA in the clinically devastating bone metastatic lesions of breast cancer. The current study demonstrated for the first time the significant roles of HA that MU treatment suppressed tumorigenicity including metastatic lesion of bone possibly via suppression HA deposition.
In our study, the suppression of HA by MU showed anti-tumor effects with inhibition of phosphorylation of Akt. HA-mediated CD44 activation was shown to promote phosphorylation of Akt and tumor progression such as cell growth and invasion in MDA-MB-231 cells.37 In murine lymphoma cell lines, 2 hr treatment of HA oligosaccharides, which can compete with the binding of high molecular HA to CD44 and displace HA was reported to decrease the level of phosphorylation of Akt.38 A recent report showed a strong cross-talk between focal adhesion kinase (FAK) and HA/RHAMM (receptor for HA-mediated motility), which might be important for tumor cells to regulate adhesion, migration and proliferation. In the study, knockdown of CD44 and RHAMM inhibited proliferation and migration with decreased phosphorylation of Akt, whereas only knockdown of RHAMM caused FAK cleavage.39 In our study, MU suppressed phosphorylation of Akt not in an early phase but at 12 and 24 hr after treatment whereas MU decreased HAS2 mRNA expression after 6 hr' treatment of MU, and the amount of HA was decreased by MU at every time point, except at 24 hr in the case of pericellular HA. These findings suggest that although the suppression of phosphorylation of Akt might not be a direct effect of MU, decreased HA synthesis and subsequent starvation of HA in cells might be associated with the suppression of phosphorylation of Akt after MU treatment. Together, suppression of Akt phosphorylation with MU in the current study might be due to the perturbation of HA/CD44 mediated signaling through Akt pathways, resulting in inhibition of cell growth, migration and invasion.
Several studies have shown that HA-rich cell-associated matrix has significant roles in the tumorigenicity of malignant cells, and suppression of this HA-rich cell-associated matrix leads to inhibition of malignant cell behavior.25, 40, 41 Preliminary experiments in this study revealed that MDA-MB-231 cells in vitro did not have HA rich cell-associated matrix after 24–72 hr incubation by particle exclusion assay (data not shown), suggesting that the significance of HA in the malignant cell behavior of MDA-MB-231 cells may differ from that in HA-rich cell associated matrix. In the present study, MDA-MB-231 cells have abundant intracellular HA in vitro. Increased intracellular HA was previously reported in proliferating Swiss 3T3 fibroblast cells after serum stimulation, and stimulated cells showed both enhanced uptake of exogenous fluorescein-labeled HA and intense endogenous HA staining in the cytoplasm.42 HA is internalized by cellular hyaladherines, including CD44 and RHAMM.43, 44 Intracellular HA has been reported to associate with mitosis.25, 42 These reports agree with our results that decreased accumulation of intracellular HA by MU suppressed cell proliferation. There was a report that exogenously added HA neutralized the effects of MU in prostate cancer cells,33 which is inconsistent with our results that exogenously added HA cannot neutralize the effects of MU. Results in the current study do not necessarily exclude the possibility that growth-inhibitory effects of MU are mediated by reduced HA synthesis. It may be a possible explanation that newly synthesized HA bound to the HA synthase complex rather than non-bound HA in extracellular matrix facilitates growth-related signals.
Morphological change of MDA-MB-231 cells, possible via HA suppression by MU, may be another mechanism underlying the anti-tumor effects. MU inhibited accumulation of cell associated HA, and the cells were circular with a loss of lamellipodia in vitro after MU treatment. Morphologic changes of cells after treatment with MU were also described in keratinocytes, which became flat and circular with a loss of lamellipodia.45 Similar morphological alterations characterize keratinocytes in which HA synthesis is inhibited by transfection of HAS2 antisense inhibition.46 HA-CD44 interactions in tumor cells are closely linked to cytoskeletal functions that involve membrane-associated cytoskeletal proteins such as ankarin and the small GTP-binding proteins such as RhoA, Rac1 and Cdc42.47 The activation of these proteins produces specific structural changes in actin assembly, cytoskeleton reorganization, transcriptional activation, tumor cell growth, survival, migration and invasion. RHAMM and CD44 act together in a HA-dependent, autocrine mechanism to coordinate sustained signaling through ERK1,2 leading to high basal motility in MDA-MB-231 and MCF7 invasive breast cancer cell lines.48 The depletion of cell associated HA by MU might affect the morphological changes and decrease motility and invasiveness in MDA-MB-231 cells.
In our study, MU decreased cell associated HA, cell growth and motility of NIH3T3 fibroblasts in vitro, and daily administration of MU decreased the accumulation of HA in both tumor and stromal cells, particularly in tumor stroma adjacent to bone in vivo. Recent reports have also reported that MU inhibits the stromal HA accumulation by melanoma cell conditioned medium, resulting in a growth inhibition of tumor cells and fibroblasts.49 Has2 knockout experiments show that HA released by stromal fibroblasts participate in tumor associated macrophage trafficking, facilitating tumor neovascularization.50 These findings suggest that inhibition of hyaluronan synthesis in tumor cells and stromal fibroblasts is a logical therapeutic tool resulting in the suppression of metastatic lesion of bone. While systemic administration of MU did not affect the structure of normal skin, articular cartilage and body weight in this study. These results suggested that systemic MU treatment is safe having minimal influence on HA-rich normal tissues. The dose of MU used for the mice in this study is equivalent to 40 g per day for humans weighing 80 kg. The metabolism of MU differs between man and mouse, and MU has been used as a choleretic agent in clinical trials for hepatitis B and C at a 2.2 g per day dose (ClinicalTrials.gov identifier, NCT00225537). The effects of MU on HA accumulation is significant in 6 (intracellular) or 12 hr (pericellular; Figs. 1f and 1g). MU is used internally twice a day for human, and continuous administration of MU has a potent enough to be appropriate for therapeutic interventions.
Although we demonstrated the inhibitory effects of MU on one of the stromal cells, NIH3T3 fibroblasts, further evaluation is necessary for other stromal cells such as osteoclasts and inflammatory cells. Systemic treatment of MU suppressed intraosseous tumor growth in vivo. Inhibitory effects of MU on bone metastatic lesions might be mediated not only by the anti-tumor response of MDA-MB-231 cells and fibroblasts but also other host stromal cells. The microenvironment of bone metastatic lesions after treatment with MU should be analyzed further. The second limitation is the lack of analysis of depletion of cellular UDP-glucuronic acid by MU treatment in MDA-MB-231 cells. Such depletion is known to be one cause of suppression of HA synthesis by MU,32 and it is very likely that not only the restraint of HAS2 mRNA expression but also depletion of cellular UDP-glucuronic acid plays an important role in the MU induced suppression of HA synthesis on MDA-MB-231 cells.
In summary, we demonstrated that MU suppresses HA synthesis and accumulation in MDA-MB-231 cells, which might be mediated partly via suppression of HAS2 mRNA expression. This leads to the suppression of cell motility and invasion in addition to cell proliferation and apoptosis with phosphorylation of Akt. We also indicated that MU suppressed cell growth and motility of NIH3T3 fibroblast cell in vitro and intraosseous tumor growth in mouse xenograft models, where MU inhibited accumulation of HA in tumor tissues at the invading border to bone in vivo. These findings show that MU suppresses intraosseous tumor growth, suggesting its potential as a therapeutic candidate for established bone metastasis of breast cancer. Considering that MU is already a commercially available agent, it could be readily applied for patients with bone metastasis, and may improve their quality of life and/or survival after the appearance of metastasis.
The authors thank Drs. Yoshitaka Suzuki, Kozo Hosono and Daizo Kato for technical assistance with the experiments.