CD44 enhances invasion of basal-like breast cancer cells by upregulating serine protease and collagen-degrading enzymatic expression and activity
- Nicola Montgomery†1,
- Ashleigh Hill†1,
- Suzanne McFarlane1,
- Jessica Neisen1,
- Anthony O'Grady2,
- Susie Conlon2,
- Karin Jirstrom3,
- Elaine W Kay2 and
- David JJ Waugh1Email author
© Waugh et al.; licensee BioMed Central Ltd. 2012
Received: 7 December 2011
Accepted: 23 May 2012
Published: 23 May 2012
Basal-like breast cancers (BL-BCa) have the worst prognosis of all subgroups of this disease. Hyaluronan (HA) and the HA receptor CD44 have a long-standing association with cell invasion and metastasis of breast cancer. The purpose of this study was to establish the relation of CD44 to BL-BCa and to characterize how HA/CD44 signaling promotes a protease-dependent invasion of breast cancer (BrCa) cells.
CD44 expression was determined with immunohistochemistry (IHC) analysis of a breast cancer tissue microarray (TMA). In vitro experiments were performed on a panel of invasive BL-BCa cell lines, by using quantitative polymerase chain reaction (PCR), immunoblotting, protease activity assays, and invasion assays to characterize the basis of HA-induced, CD44-mediated invasion.
Expression of the hyaluronan (HA) receptor CD44 associated with the basal-like subgroup in a cohort of 141 breast tumor specimens (P = 0.018). Highly invasive cells of the representative BL-BCa cell line, MDA-MB-231 (MDA-MB-231Hi) exhibited increased invasion through a basement membrane matrix (Matrigel) and collagen. In further experiments, HA-induced promotion of CD44 signaling potentiated expression of urokinase plasminogen activator (uPA) and its receptor uPAR, and underpinned an increased cell-associated activity of this serine protease in MDA-MB-231Hi and a further BL-BCa cell line, Hs578T cells. Knockdown of CD44 attenuated both basal and HA-stimulated uPA and uPAR gene expression and uPA activity. Inhibition of uPA activity by using (a) a gene-targeted RNAi or (b) a small-molecule inhibitor of uPA attenuated HA-induced invasion of MDA-MB-231Hi cells through Matrigel. HA/CD44 signaling also was shown to increase invasion of MDA-MB-231 cells through collagen and to potentiate the collagen-degrading activity of MDA-MB-231Hi cells. CD44 signaling was subsequently shown to upregulate expression of two potent collagen-degrading enzymes, the cysteine protease cathepsin K and the matrix metalloprotease MT1-MMP. RNAi- or shRNA-mediated depletion of CD44 in MDA-MB-231Hi cells decreased basal and HA-induced cathepsin K and MT1-MMP expression, reduced the collagen-degrading activity of the cell, and attenuated cell invasion through collagen. Pharmacologic inhibition of cathepsin K or RNAi-mediated depletion of MT1-MMP also attenuated MDA-MB-231Hi cell invasion through collagen.
HA-induced CD44 signaling increases a diverse spectrum of protease activity to facilitate the invasion associated with BL-BCa cells, providing new insights into the molecular basis of CD44-promoted invasion.
Breast cancer is a heterogeneous disease, currently defined as a minimum of five distinct molecular subtypes . Of these subtypes, "basal-like" breast cancer (BL-BCa) has the worst clinical outcome and is associated with an increased risk of hematogenous metastasis, predominantly to the lungs and liver . An enhanced understanding of the mechanisms and factors that underpin the local invasion and the capacity of BL-BCa cells to escape from the primary tumor, or invade secondary tumor sites, would have significant impact on improving the outcomes for this disease subtype.
Hyaluronan (HA) is a constituent of extracellular matrix, which can induce marked effects on cell behavior by binding to its predominant cell-surface receptor CD44 . Before the era and definition of the molecular subtypes, elevated levels of HA in tumor stroma were shown to correlate with poorly differentiated tumors, auxiliary lymph node status, and short overall survival in breast cancer [4, 5]. Klingbeil and colleagues  recently determined that CD44 expression associates with the BL-BCa subtype. Furthermore, we recently determined that CD44 is inversely associated with estrogen receptor (ER) expression, with strong expression localized to basal cells [McFarlane S, Conlon S, O'Grady A, Kay EW, Waugh DJJ, unpublished observations]. Consistent with an association with the most clinically aggressive tumors, in vitro studies have demonstrated the role of HA and CD44 in stimulating breast cancer cell migration and cell invasion. We have shown that tetracycline-induced expression of CD44 in the noninvasive, luminal MCF-7 breast cancer cell line is alone sufficient to induce cell invasion in response to HA in vitro . The induction of CD44 also was sufficient to promote the spontaneous metastasis of these noninvasive luminal breast cancer cells to the liver in vivo . Clinical studies have also confirmed the enrichment of CD44 expression in disseminated tumor cells resident in secondary tissue sites [9, 10].
Metastasis demands that cancer cells invade through the physical barriers provided by the extracellular matrix of the primary and secondary tumor sites and the basement membranes present within each of these tissue sites. Activation of proteolytic enzymes is thought to be essential in facilitating the degradation of the proteins that constitute these physical structures. Interestingly, two enzymes of the matrix metalloproteinase (MMP) family, MMP-9 and MMP-7, were previously shown to complex with the ectodomain of CD44 on the surface of malignant cells, suggesting that CD44 acts in a structural capacity to concentrate protease activity on the surface of actively invading cells [11, 12]. Moreover, CD44 itself is a substrate of a further MMP, which complexes with the membrane-tethered enzyme MT1-MMP and is then cleaved by the enzyme [13, 14]. Although these studies suggest that CD44 cooperates with MMPs to regulate cell invasion, the relation of HA-induced CD44 signaling to the regulation of protease expression and activity in invasive breast cancer cells is poorly defined. Moreover, it is well known that malignant breast cancer cells express other protease species in addition to those of MMPs, including the serine protease urokinase plasminogen activator (uPA), a marker of poor prognosis and associated with BL-BCa [15, 16].
The objective of our study was to increase our understanding of CD44-promoted breast cancer cell invasion by defining the effect of CD44 signaling on protease gene expression and activity, and defining the role of these proteases in underpinning HA-induced invasion. Informed by studies on breast cancer tissue reaffirming the association of CD44 with the basal-like subtype of breast cancer, we show that CD44 signaling amplifies serine protease, MMP, and/or cysteine cathepsin gene expression and activity, all of which contribute to the invasion of BL-BCa cells through a specialized matrix. Our studies thus provide a new molecular insight to substantiate the association of CD44 with the metastasis of breast cancer.
Materials and methods
Tissue microarray construction and immunohistochemistry
Breast tumor samples were collected and data recorded as described previously . Then 4 μm sections were cut from the TMA and immunostained with anti-CD44 on an automated platform (Bond system; Vision BioSystems, Mount Waverley, Victoria, Australia). In brief, cut sections were subjected to on-board dewaxing (Dewax solution; Vision BioSystems) and antigen retrieval (Epitope Retrieval 1 solution; Vision BioSystems) for 20 minutes before application of primary antibody (1:200) and detection by using the Bond polymer Refine detection system (Vision BioSystems). All sections were counterstained with hematoxylin. Negative controls were included for all sections by the omission of primary antibody. Positive control tissue (normal thymus) was also used. Immunostained slides were scored on the proportion of positive tumor cells (range, 0 to 4) and the average intensity of staining (range, 0 to 3). These values were added to obtain a total score (range, 0 to 7). All scores were examined by two independent observers (SMcF and SC), and 10% were also scored by a third independent observer (EK). An expression correlation analysis between CD44 and the basal-like subtype in breast cancer cells and primary tumors was also performed by using Oncomine (Compendia Bioscience, Ann Arbor, MI, USA).
The MDA-MB-157 cell line was purchased from American Type Culture Collection (ATCC) (Manassas, VA, USA) and cultured as previously described . Hs578T cells were provided by Dr Paul Mullan (CCRCB, Queen's University Belfast) and cultured as described . The highly invasive clone, MDA-MB-231Hi, was provided, complete with matched parental cells by Prof. Toshiyuki Yoneda (University of Health Sciences, San Antonio, TX, USA) and were cultured in DMEM supplemented with 10% vol/vol fetal calf serum (FCS) (Invitrogen Life Technologies, Paisley, UK) . CD44-depleted MDA-MB-231 cells (termed MDA-MB-231 sh#1) were generated in our laboratory by stable transfection with a CD44 shRNA; parallel transfection with a nontargeting shRNA was used as a control to generate MDA-MB-231 NT cells (McFarlane S, et al., unpublished data) and cultured in DMEM supplemented with 10% vol/vol FCS, 0.2 μg/ml puromycin. All cell lines were grown to 70% confluence before experimentation.
Reagents and antibodies
Chemicals were supplied by Sigma Chemical Co. (St. Louis, MO, USA) unless otherwise stated. Hyaluronan of molecular mass 220 kDa and medical-grade purity was purchased from Lifecore Biomedical Inc. (Chaska, MN, USA). The cathepsin K inhibitor was purchased from Calbiochem (La Jolla, CA, USA). Pefabloc uPA inhibitor was supplied by DSM Nutritional Products Ltd Branch, Pentapharm (Basel, Switzerland). Specific RNAi SMARTpools for CD44, uPA, and MT1-MMP, the nontargeting siCONTROL RNAi oligonucleotide, and Dharmafect 2 transfection reagent were obtained from Dharmacon (Lafayette, CA, USA). The mouse anti-human CD44 monoclonal antibody (1:500), mouse anti-human uPAR mAb (1:500 dilution), and mouse anti-human serpin E1/PAI-1 mAb (1:500) were obtained from R&D Systems (Abingdon, UK). Abcam (Cambridge, UK) supplied the rabbit anti-cathepsin K pAb (1:100). The mouse anti-human urokinase mAb (1:500) and the mouse anti-human PAI-2 mAb (1:3,000) were provided by American Diagnostica (Stamford CT, USA). The mouse anti-MMP14 (MT1-MMP) mAb (1:1,000) was obtained from Chemicon (Watford, UK). Mouse anti-human GAPDH mAb (1:3,000) was supplied by AbD Serotec (Oxford, UK). Sigma (UK) supplied the mouse anti-human β-tubulin mAb (1:1,000). The sheep anti-mouse IgG/horseradish peroxidase conjugate (1:2,000) and donkey anti-rabbit IgG/horseradish peroxidase conjugate (1:2,000) secondary antibodies were purchased from Amersham (Amersham, UK). The anti-p38 MAPK antibody (1:1,000) was supplied by Cell Signaling Technologies (Beverly, MA, USA).
Quantitative PCR analysis
RNA was harvested and cDNA was synthesized from 10 to 20 μg total RNA, as previously described . Analysis was performed on a Roche LC480 Light Cycler (Roche Diagnostics GmbH, Mannheim, Germany), with product amplification determined by SYBR Green 1 fluorescence detection. Forward (Fwd) and reverse (Rev) primers are as shown: CD44 Fwd TTTGCATTGCAGTCAACAGTC; CD44 Rev GTTACACCCCAATCTTCATGTCCAC; Cathepsin K Fwd AGGCTTCTCTTGGTGTCCATA; Cathepsin K Rev CCTTTCTTTCGATAGTCGACA; MT1-MMP Fwd AATATGGCTACCTGCCTCCC; MT1-MMP Rev TTGCCATTTGAGACCCTGGAT; uPA Fwd GAGGCCCCGCTTTAAGATTA; uPA Rev TGGAGTTAAGCCTTGAGCGA; uPAR Fwd AAGATCACCAGCCTTACCGA; uPAR Rev CCTTCTTCACCTTCCTGGAT; PAI-1 Fwd CTGACAACAGGAGGAGAAAC; PAI-1 Rev GGAACAGCCTGAAGAAGTGG; PAI-2 Fwd ACCCAGAACCTCTTCCTCTC; PAI-2 Rev TGGTAAAGTTCTCTGGAGTCA; 18S Fwd CATTCGTATTGCGCCGCTA; 18S Reverse CGACGGTATCTGATCGTC. Reactions were conducted by using 1 μl of cDNA reverse transcribed from 10 to 20 μg total RNA, 0.4 μM final concentration of forward and reverse primers and 2 × SYBR Green 1 master mix (Roche Diagnostics). Standard cycling procedures were used, with annealing temperatures of 51°C used for the CD44 and 18S, 54°C for uPAR, 55°C for cathepsin K, 55.7°C for MT1-MMP, and 57°C for the uPA, PAI-1, and PAI-2 primer pairs. Specific amplicon formation with each primer pair was confirmed by melt-curve analysis. Gene expression was quantified relative to an 18S housekeeping gene.
Knockdown of CD44, uPA, and MT1-MMP expression in cancer cells
Cells were seeded to approximately 70% confluence before transfection with the relevant RNAi SMARTPool or the nontargeting RNAi oligonucleotide by using DharmaFECT-2 according to the manufacturer's instructions.
Protein samples were collected, quantified, and blotted as previously described  by using the antibodies described. Immunoreactivity was detected by using chemiluminescence (Supersignal; Pierce). Equal loading of the protein samples was assessed by reprobing the membrane with either a GAPDH or β-tubulin antibody.
uPAR flow cytometry
Samples were analyzed for cell-surface uPAR expression as previously described  by using a fluorescein-conjugated murine anti-human monoclonal uPAR antibody (American Diagnostica Inc., Stamford, CT, USA) and an IgG control antibody (R&D Systems, Abingdon, UK).
Invasion chambers were prepared by coating cell-culture inserts (12-μm pore size; Costar) with either 2 μg/cm2 of Collagen I (BD Biosciences, Erembodegem, Belgium) or 100 μg/cm2 of Matrigel (BD Biosciences) alone or supplemented with HA at a final concentration of 100 μg/cm2 in phenol red-free DMEM. Assays were conducted as previously described . In some experiments, the invasive capacity of the cells was determined by using the xCELLigence cell analyzer system (Roche Diagnostics). The top chambers of CIM invasion plates were coated with Matrigel (5%) and allowed to dry for 4 hours at 37°C. Prewarmed phenol red-free DMEM supplemented with FCS was added to the lower chambers before the upper chamber was locked into place and 30-μl serum-free medium added to each well. The invasion apparatus was allowed to equilibrate at 37°C for 1 hour, after which 1 × 105 cells in serum-free medium were seeded into the top wells and allowed to settle for 30 minutes. The plates were then loaded onto the xCELLigence analyzer, and electrical impedance was measured every 15 minutes over a 36-hour period.
uPA activity assays
uPA activity was determined by using a uPA Activity Assay Kit (Chemicon International) according to the manufacturer's instructions on either cell lysate or cell media samples. In brief, a standard curve was generated by the addition of varying amounts of the uPA-positive control (10 to 160 μl) in duplicate to a clear 96-well plate. Then 50μl cell lysate samples or 10 μl of cell media samples was added to the 96-well plate along with a either a media or RIPA buffer blank. Sufficient deionized water was added to bring the total volume of each well to 160 μl. Then 20 μl assay buffer was added to each well, followed by 20 μl of chromogenic substrate. The plate was incubated in the dark at 37°C for periods between 10 minutes and 8 hours, depending on uPA levels contained in samples. The optical density of each well was determined by using a standard microplate reader (405 nm). Duplicate OD readings for each sample were averaged, and the relevant blanks were subtracted. Samples were compared with the standard curve to obtain relative uPA activity. Results are expressed as fold change over control cells.
Plasmin activity assays
Cells were plated at a density of 2 × 104 cells per well in 200 μl of phenol red-free medium and allowed to reach approximately 70% confluence. Cell-free medium (200 μl) was added to a well as a negative control, and 3 μg human plasminogen (Hyphen BioMed, ZAC Neuville University, France) was added to each well and incubated at 37°C for 30 minutes, and 65 μl of assay buffer (100 mM Tris/0.5% TritonX-100, pH 8.8) and 15 μl S-2251 substrate (25 μg) (Chromogenix, Milan, Italy) was added to the wells of a separate 96-well plate. After a 30-minute incubation period, 20 μl of each sample was added in duplicate to wells. Human plasmin was implemented to generate a standard curve, from which plasminogen activation rates of unknown samples were determined. Plasmin of varying amounts was added to the wells of the 96-well plate containing assay buffer and S-2251 substrate. The plate was covered, and the optical density for each well at 405 nm was determined by using a standard microplate reader at regular intervals. Plasminogen is activated to form plasmin that is able to cleave the chromogenic substrate S-2251. The method for determination of activity is based on the difference in absorbance at 405 nm. Duplicate OD readings for each sample were averaged, and the mean media blank subtracted. Samples were compared with the plasmin standard curve to obtain relative plasmin units of activity. Results were then quantified relative to cell number for each sample, and expressed as relative plasmin units of activity per 1 × 106 cells.
Cathepsin K ELISA
Breast cancer cells were seeded at a density of 1 × 105 cells/500 μl in a 24-well plate and allowed to reach 70% confluence before experimentation. Cell-culture supernatants were collected and cell debris removed by centrifugation. Cell counts were taken from each experimental well to allow normalization of cathepsin K concentration to cell number (106 cells). The cathepsin K ELISA (Biomedica Gruppe, Vienna, Austria) was conducted as per the manufacturer's instructions. In brief, 50 μl of the sample, standards or cathepsin K control were added to the polyclonal sheep anti-cathepsin K-coated microtiterstrips and incubated at room temperature for 20 to 24 hours in the dark. The samples, standards, and control were aspirated and wells washed 5 times with wash buffer. The 200-μl substrate was added to each well and incubated for 30 minutes at room temperature in the dark. The 50-μl STOP solution was added to each well, and the absorbance was immediately read. The sample concentration was calculated from the standard curve, and the concentration normalized to 106 cells by using previously determined cell counts.
EnzChek collagen degradation assay
Collagenase activity was measured by using the EnzChek Gelatinase/Collagenase assay kit (Molecular Probes, Eugene, OR, USA) according to the manufacturer's instruction. In brief, cells were grown in P90s in serum-containing media and allowed to reach 80% confluence before concentration of the cell culture media by using Centriplus centrifugal filter units (Millipore, Billerica, MA, USA). The concentrated protein samples were quantified by using the BCA assay, and 1 mg protein was loaded in duplicate onto a 96-well black fluorescence plate. Then 20 μl of DQ Collagen type I fluorescein conjugate (Molecular Probes) was loaded into each well along with 80 μl 1 × Running Buffer. The collagenase Clostridium (0.025 U/ml) was run as a positive control and running buffer acted as a negative control. As an additional control, cell-culture media (in the absence of cells) was assessed to ensure that the collagenase activity was solely restricted to enzymes secreted by the breast cancer cells. The fluorescence intensity was read at 37°C in a fluorescence microplate reader (Cytofluor 4000; Applied Biosystems, Warrington, UK) over a 6- to 8-hour period with fluorescent readings taken every 30 minutes. Collagenase activity over the time period was reported as fluorescence on subtraction of the cell-culture media fluorescence.
Statistical analysis of experimental data
Differences between data points (invasion and enzyme activity assays, ELISA experiments) were assessed for statistical significance by using the two-tailed Student t test comparisons (GraphPad Prism 5 software).
CD44 expression correlates with triple-negative, basal-like breast cancers
Association of CD44 with molecular characteristics of breast tumors
Association of CD44 expression with tumor characteristics
0.019 t test
0.025 t test
0.021 t test
0.021 t test
Expression of CD44 is expressed as a percentage of patients in individual clinical classifications (first column).
Characterization of increased CD44 expression and importance to invasion in basal-like breast cancer cells
Given the strong correlation of CD44 expression to the basal-like subgroup of breast cancers, a series of in vitro experiments was performed by using cell lines representative of this subtype, primarily by using the invasive MDA-MB-231 cell line and a highly metastatic derivative MDA-MB-231 cell line (MDA-MB-231Hi). After qPCR and immunoblotting analysis, we confirmed that MDA-MB-231Hi cells were heavily enriched for CD44 at both mRNA (Figure 1C) and protein level (Figure 1D) when compared with their parental MDA-MB-231 counterparts, and were more invasive through Matrigel (Figure 1E). The importance of CD44 and its elevated expression in underpinning the more-efficient invasion of MDA-MB-231Hi cells was confirmed by using an RNAi strategy targeting this cell-surface receptor; knockdown of CD44 (validated in Additional file 1, Figure S1A) significantly reduced invasion of the MDA-MB-231Hi cells through Matrigel (Figure 1F).
Characterization of elevated uPA activity in highly invasive basal-like breast cancer cells
HA and CD44 regulate uPA expression and activity in basal-like breast cancer cells
The role of CD44 in regulating uPA and its associated proteins was next investigated. A short-hairpin strategy was used to downregulate CD44 mRNA and protein expression in the MDA-MB-231Hi cell line (termed MDA-MB-231 sh#1 cells). The effects of CD44 knockdown were determined by comparison with MDA-MB-231 NT cells, transfected with a nontargeting short hairpin and which retained CD44 expression (see Additional file 3, Figures S3A and S3B). CD44 knockdown retarded cell invasion through Matrigel (see Additional file 3, Figure S3C). Relative to a nontargeting short hairpin (NT), loss of CD44 also coincided with a decrease in the transcript levels for uPA, uPAR, PAI-1, and PAI-2 (all P < 0.001) (see Additional file 3, Figure S3D) and a decreased expression for each of these proteins (see Additional file 3, Figure S3E).
To associate CD44 further in mediating HA-induced increases in uPA gene expression and activity, we performed another series of experiments using the MDA-MB-231 sh#1 cells. In these experiments, the effect of CD44 knockdown on the expression and activity of uPA was determined in the absence and presence of an exogenous HA stimulus. Loss of CD44 not only coincided with a decreased transcript level of uPA relative to that in NT-transfected cells, as observed before, but also attenuated the HA-promoted increases in uPA gene expression (Figure 3C). When further experiments were conducted to measure uPA activity, MDA-MB-231 sh#1 cells had reduced cell-surface uPA activity under unstimulated conditions, whereas the HA-induced increase in uPA activity was severely truncated in the absence of CD44 (Figure 3D).
Consistent with the observed decreases in uPA and uPAR mRNA and protein expression, and the reduction in uPA activity, the loss of CD44 was also shown to correlate with a decrease in the plasmin activity associated with the MDA-MB-231Hi cells (Figure 3E). The functional relevance of this uPA-generated plasmin activity to HA-promoted cell invasion was determined with a further invasion assay in which plasmin activity was inhibited by using aprotinin. HA failed to stimulate invasion in the presence of 1 μM aprotinin (Figure 3F).
uPA activity underpins HA-promoted invasion of basal-like breast cancer cells
A pharmacologic approach used a small-molecule uPA inhibitor to attenuate uPA activity. Administration of this inhibitor to MDA-MB-231Hi cells reduced both cell surface-associated uPA activity and uPA activity in the supernatant in a concentration-dependent manner; at a concentration of 30 μM, this inhibitor reduced supernatant activity to <5% of vehicle-treated control and reduced cell surface-associated activity to 22 ± 6.1% of vehicle-treated control (Figure 4D). In cell-invasion assays, the addition of the uPA inhibitor (30 μM) abrogated the HA-stimulated invasion of the MDA-MB-231Hi cells (Figure 4E).
CD44 signaling also increases expression of collagenolytic proteases of the cysteine cathepsin and matrix-metalloproteinase families in basal-like breast cancer cells
uPA has no intrinsic ability to degrade collagen. Therefore, we undertook experiments to identify additional differentially expressed proteolytic species that can directly promote collagen degradation and whose altered expression/activity may explain the marked difference in cell invasion through collagen. Two of the most potent collagen-degrading proteases are the membrane-tethered matrix metalloproteinase MT1-MMP  and the cysteine protease, cathepsin K . qPCR analysis demonstrated an increased mRNA transcript level for cathepsin K (Figure 5D) and MT1-MMP in MDA-MB-231Hi cells over parental cells (Figure 5E). Immunoblotting also confirmed an elevated expression of MT1-MMP and cathepsin K protein expression in the MDA-MB-231Hi cells (Figure 5F), whereas an ELISA revealed an increased secretion of cathepsin K in MDA-MB-231Hi cells relative to parental cells (Figure 5G).
The functional importance of MT1-MMP and cathepsin K in underpinning invasion of these cells through a collagen-I matrix was then investigated by using molecular or pharmacologic approaches. RNAi was used to suppress MT1-MMP expression in the MDA-MB-231Hi cells (see Additional file 1, Figure S1B), whereas a commercially available pharmacologic inhibitor was used to antagonize cathepsin K activity in these cells. Inhibition of MT1-MMP (Figure 7E) or cathepsin K (Figure 7F) reduced the invasive capacity of these MDA-MB-231Hi cells through collagen.
HA induces transcriptional regulation of proteases through a p38 mitogen-activated protein kinase pathway
Molecular stratification of gene-expression data has defined a minimum of five distinct subtypes of breast cancer . Our analysis of breast tumor material determined that CD44 was inversely correlated with expression of the ER and PR, with no correlation to Her2, and that enrichment of CD44 was associated with BL-BCa, the subtype associated with the poorest clinical outcome. Our tissue-based characterization is consistent with other recent studies reporting the localization of CD44 in BL-BCa . The association of CD44 with the most aggressive subtype of disease is consistent with the well-established cellular functions of CD44 in promoting cell invasion and cell adhesion in vitro, and the capacity for CD44 to increase spontaneous metastasis of breast cancer cells in vivo [7,8; McFarlane S, Coulter J, Waugh DJJ, unpublished observations]. Bourguignon and colleagues [30–32] provided a detailed understanding of CD44-promoted activation of Rho-GTPase family signaling, underpinning the reorganization of the cell cytoskeleton to facilitate the active migration of cells. The Stamenkovic and Seiki laboratories also reported that CD44 can complex with secreted and membrane-tethered matrix metalloproteinases, localizing their proteolytic activity to the invasive edge of tumor cells [11–14]. In this study, we now add a further dimension to our understanding of CD44-promoted invasion of breast cancer. In a series of experiments, conducted primarily in MDA-MB-231 breast cancer cells, but supported by additional observations in other well-established models of ER-negative and BL-BCa, we show that HA and CD44 can upregulate expression of key proteases that underpin the invasion of BL-BCa cells through experimental matrices. Importantly, our studies demonstrate that CD44 affects the expression and activity of members of the serine protease, cysteine cathepsin, and matrix metalloproteinase family. Collectively, these enzymes may contribute a spatial and temporal function throughout the metastatic cascade, providing invasive breast cancer cells with the essential complement of proteolytic activity to traverse successfully the physical barriers provided by the tissue matrix and basement membranes in both the primary and secondary tissue sites.
The serine protease uPA and its inhibitor PAI-1 are bona fide markers of poor prognosis that predict distant metastasis in breast cancer [15, 16], whereas expression of uPA is upregulated in tumor-initiating breast cancer cells . Highly invasive MDA-MB-231Hi cells were shown to have elevated expression of uPA, uPAR, and PAI-1 at transcript and protein levels, and had elevated cell surface-associated and extracellular uPA activity. We also detected increased plasmin activity in the MDA-MB-231Hi cells, consistent with a more-efficient uPA-mediated cleavage of its substrate plasminogen. The addition of HA further increased the mRNA transcript levels for uPA, uPAR, and PAI-1 and increased the levels of the corresponding proteins in a time-dependent manner in these BL-BCa cells. Coincident with the increased expression of uPA and uPAR, HA increased the cell surface-associated activity of uPA in these cells.
Conversely, the loss of CD44 in MDA-MB-231Hi cells coincided with a decreased expression of uPA, uPAR, PAI-1, and PAI-2, and attenuated the capacity of HA to increase expression of uPA. Consequently, CD44-depleted cells were associated with a decreased level of plasmin activity.
The importance of these serine proteases to BL-BCa cell invasion was investigated in vitro. Consistent with the HA-induced promotion of initially uPA and subsequently plasmin activity, the serine protease inhibitor aprotinin abrogated the HA-induced invasion of the MDA-MB-231Hi cells through Matrigel. Similarly, RNAi-mediated knockdown of uPA attenuated the invasion of these BL-BCa cells, whereas an inhibition of MDA-MB-231Hi cell invasion was also observed by using a small-molecule inhibitor of uPA. In addition to this role for uPA in promoting localized invasion of BL-BCa cells, increased uPA activity was detected in highly disseminating tumor cells and implicated in regulating the intravasation of these disseminating cells [23–25]. Accordingly, CD44-positive BL-BCa cells may be more efficient in completing intravasation, consistent with the role of CD44 in promoting the spontaneous metastasis of breast cancer in vivo  and the increased detection of CD44-enriched cells to disseminate to distant organs [9, 10]. Targeting of the enzymatic activity of uPA by using small-molecule inhibitors may therefore be an appropriate and effective adjuvant therapeutic strategy to use in BL-BCa, or those breast cancers showing strong expression of CD44, with the intent to reduce distant metastasis and prolong survival in this poor-prognosis group of patients. Importantly, the small-molecule inhibitor used in this study was capable of antagonizing both cell-surface and extracellular activities of uPA, suggesting that its inhibitory potential is not restricted by specific localization of the enzyme.
Collagen is a major constituent of the extracellular matrix of primary and secondary tumor sites. MDA-MB-231Hi cells showed increased capacity to degrade and invade through collagen, a response also attenuated by CD44 knockdown. As uPA has no capacity to degrade collagen directly, a further analysis characterized additional differential protease expression consistent with this increased collagen-degrading activity in MDA-MB-231Hi cells. Immunoblotting confirmed increased expression of two potent collagen-targeting proteases, MT1-MMP and cathepsin K, in the more invasive cells. Furthermore, HA increased (a) the transcript levels of each of the MT1-MMP and cathepsin K genes, (b) increased expression of the corresponding proteins, and (c) promoted the extracellular secretion of cathepsin K. As before, depletion of CD44 reduced the expression of MT1-MMP and the secretion of cathepsin K from the MDA-MB-231Hi cells, consistent with an overall reduction in the capacity of these cells to degrade collagen in the absence of CD44. Furthermore, loss of CD44 attenuated the HA-induced increases in protease expression, underlying the importance of this receptor in mediating the response. Loss of either MT1-MMP or cathepsin K enzymatic activity was shown to reduce the HA-promoted invasion of MDA-MB-231Hi cells through collagen.
MT1-MMP promotes the cleavage of the ectodomain of CD44 [13, 14]. Our data suggest the existence of a positive-feedback mechanism by which the engagement of extracellular HA fragments with CD44 induces the transcription of the protease responsible for promoting the shedding of CD44 from the cell surface. Overall, CD44 cleavage would attenuate cell attachment to HA in the matrix and facilitate cell invasion. Moreover, CD44, MT1-MMP, and cortactin have been shown to localize to invadopodia , the membrane protrusions that localize enzymes required for ECM degradation at the leading edge of invasive cells. We suggest that CD44 may play an important hierarchic role in facilitating invadopodia organization; for example, we showed that CD44/HA-signaling increases the transcription and expression of both MT1-MMP and EMS1/cortactin in invasive BL-BCa cells . Consistent with the past observations of Bourguignon and colleagues , we also confirmed that HA-induced CD44 signaling induces a posttranslational activation of cortactin signaling in these invasive breast cancer cells [30, McFarlane et al., unpublished data]. The capacity for CD44 to increase MT1-MMP expression and promote its localization within invadopodia would be consistent with the capacity of CD44 to induce breast cancer cell invasion. MT1-MMP has multiple substrates within the extracellular matrix, including collagen I, II, and III; fibronectin; and each of laminin 1 and 5; in addition to having an enzyme convertase function that leads to an activation of the gelatinase MMP-2 and the collagenase MMP-13 [26, 35]. Consequently, MT1-MMP activity is associated with localized invasion but also contributes to the degradation of basement membranes, a process essential to successful intravasation and extravasation of tumor cells. The capacity for CD44 to increase an enzyme that can activate a downstream cascade of further MMP activity would also be consistent with the role of this receptor in enabling cells to complete metastasis to different tissue sites.
Cathepsin K is a member of the papain/cysteine protease superfamily . Cathepsin K expression has previously been reported in primary human breast tumors and their metastases, including breast cancer cells within bone metastatic lesions . Analysis of bone metastases has reported CD44 expression on resident breast cancer cells , whereas recent data from our laboratory confirm that loss of CD44 on systemically administered MDA-MB-231 cells reduces secondary tumor formation in athymic nude mice, including a reduction in osteolytic metastases [McFarlane et al., unpublished data]. Furthermore, the inhibition of hyaluronan synthesis in MDA-MB-231 cells has been reported to reduce subsequent bone metastasis in vivo . Thus CD44-promoted increases in cathepsin K expression and secretion may assist the initial invasion and colonization of BrCa cells within collagen-enriched secondary sites, including the bone.
In summary, we define the importance of CD44 in underpinning the invasion of BL-BCa cells, and illustrate the importance of HA and CD44 in upregulating a diverse spectrum of protease expression and activity that can function in a spatial or temporal manner to enable invasive cells to remodel their localized environment or perform specialist functions, including the successful completion of intravasation and extravasation. The capacity of CD44 to induce protease activity is consistent with the aggressive clinical characteristics of BL-BCa and the increased propensity to invade locally through primary breast tissue and to colonize collagen-enriched organs, including the liver, lungs, brain, and skin. Thus the detection of high CD44 expression in tumor biopsy tissue may be a suitable biomarker to identify patients who may benefit from the provision of MT1-MMP- or uPA-targeting therapeutics to reduce the risk of intravasation and distant metastasis. Moreover, as CD44-enriched "stem cell-like" breast cancer cells have been shown to disseminate early in the course of disease [9, 10], inhibitors of these key protease activities may be a relevant adjuvant approach to counteract the successful invasion and colonization of secondary tissues by these invasive breast cancer cells.
basal-like breast cancer
plasminogen activator inhibitor-1/2
small interfering RNA
urokinase plasminogen activator
urokinase plasminogen activator receptor.
The authors gratefully acknowledge the kind gift of the MDA-MB-231HI and parental cell lines from Professor Toshiyuki Yoneda (University of Health Sciences, San Antonio, TX, USA) and Hs578T cells from Dr Paul Mullan (QUB). This work was supported by a research grant from Breast Cancer Campaign (DJJW), the Association for International Cancer Research (DJJW), and was supported by a Scholarship Award from the Department of Employment and Learning, Northern Ireland, to NM.
- Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, Hastie T, Eisen MB, van de Rjin M, Jeffrey SS, Thorsen T, Quist H, Matese JC, Brown PO, Botstein D, Lonning PE, Borresen-Dale A: Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci USA. 2001, 98: 10869-10874. 10.1073/pnas.191367098.View ArticleGoogle Scholar
- Minn AJ, Gupta GP, Siegel PM, Bos PD, Shu W, Giri DD, Viale A, Olshen AB, Gerald WL, Massagué R: Genes that mediate breast cancer metastasis to lung. Nature. 2005, 436: 518-524. 10.1038/nature03799.View ArticleGoogle Scholar
- Gotte M, Yip GW: Heparanase, hyaluronan and CD44 in cancers: a breast carcinoma perspective. Cancer Res. 2006, 66: 10233-10237. 10.1158/0008-5472.CAN-06-1464.View ArticleGoogle Scholar
- Auvinen P, Tammi R, Parkinnen J, Tammi M, Agren U, Johansson R, Hirvikoski P, Eskelinen M, Kosma VM: Hylauronan in peritumoral stroma and malignant cells associates with breast cancer spreading and predicts survival. Am J Pathol. 2000, 156: 529-536. 10.1016/S0002-9440(10)64757-8.View ArticleGoogle Scholar
- Karihtala P, Soini Y, Auvinen P, Tammi R, Tammi M, Kosma VM: Hyaluronan and breast cancer. J Histochem Cytochem. 2007, 55: 1191-1198. 10.1369/jhc.7A7270.2007.View ArticleGoogle Scholar
- Klingbeil P, Natrajan R, Everitt G, Vatcheva R, Marchio C, Palacios J, Buerger H, Reis-Filho JS, Isacke CM: CD44 is overexpressed in basal-like breast cancers but is not a driver of 11p13 amplification. Breast Cancer Res Treat. 2010, 120: 95-109. 10.1007/s10549-009-0380-7.View ArticleGoogle Scholar
- Hill A, McFarlane S, Mulligan K, Gillespie H, Draffin JE, Trimble A, Outhit A, Johnston PG, Harkin DP, McCormick D, Waugh DJJ: Cortactin underpins CD44-promoted invasion and adhesion of breast cancer cells to bone marrow endothelial cells. Oncogene. 2006, 25: 6079-6091. 10.1038/sj.onc.1209628.View ArticleGoogle Scholar
- Ouhtit A, Abd Elmageed ZY, Abdraboh ME, Lioe TF, Raj MH: vivo evidence for the role of CD44s in promoting breast cancer metastasis to the liver. Am J Pathol. 2007, 171: 2033-2039. 10.2353/ajpath.2007.070535.View ArticleGoogle Scholar
- Balic M, Lin H, Young L, Hawes D, Giuliano A, McNamara G, Datar RH, Cote RJ: Most early disseminated cancer cells detected in bone marrow of breast cancer patients have a putative breast cancer stem cell phenotype. Clin Cancer Res. 2006, 12: 5615-5621. 10.1158/1078-0432.CCR-06-0169.View ArticleGoogle Scholar
- Abraham BK, Fritz P, McClellan M, Hauptvogel P, Athelogou M, Brauch H: Prevalence of CD44+/CD24-/low cells in breast cancer may not be associated with clinical outcome but may favor distant metastasis. Clin Cancer Res. 2005, 11: 1154-1159.Google Scholar
- Yu Q, Stamenkovic I: Localization of matrix metalloproteinase-9 to the cell surface provides a mechanism for CD44-promoted tumor invasion. Genes Dev. 1999, 13: 35-48. 10.1101/gad.13.1.35.View ArticleGoogle Scholar
- Yu WH, Woessner JF, McNeish JD, Stamenkovic I: CD44 anchors the assembly of matrilysin/MMP-7 with heparin-binding epidermal growth factor precursor and ErbB4 and regulates female reproductive organ remodeling. Genes Dev. 2002, 16: 307-323. 10.1101/gad.925702.View ArticleGoogle Scholar
- Kajita M, Itoh Y, Chiba T, Mori H, Okada A, Kinoh H, Seiki M: Membrane-type 1 metalloproteinase cleaves CD44 and promotes cell migration. J Cell Biol. 2001, 153: 893-904. 10.1083/jcb.153.5.893.View ArticleGoogle Scholar
- Ueda J, Kajita M, Suenaga N, Fujii K, Seiki M: Sequence specific silencing of MT1-MMP expression suppresses tumor cell migration and invasion: importance of MT1-MMP as a therapeutic target for invasive tumors. Oncogene. 2003, 22: 8716-8722. 10.1038/sj.onc.1206962.View ArticleGoogle Scholar
- Duffy MJ, O'Grady P, Devaney D, O'Siorain L, Fennelly JJ, Lijnen HJ: Urokinase-plasminogen activator, a marker for aggressive breast carcinomas: preliminary report. Cancer. 1988, 62: 531-533. 10.1002/1097-0142(19880801)62:3<531::AID-CNCR2820620315>3.0.CO;2-B.View ArticleGoogle Scholar
- Shimizu M, Cohen B, Golvasser P, Berman H, Virtanen C, Reedijk M: Plasminogen activator uPA is a direct transcriptional target of the JAG1-Notch receptor signaling pathway in breast cancer. Cancer Res. 2011, 71: 277-286. 10.1158/0008-5472.CAN-10-2523.View ArticleGoogle Scholar
- Molerstrom E, Kovacs A, Lovgren K, Nemes S, Delle U, Danielsson A, Parris T, Brennan DJ, Jirstrom K, Karlsson P, Helou K: Upregulation of cell cycle arrest protein BTG2 correlates with increased overall survival in breast cancer as detected by immunohistochemistry using tissue microarray. BMC Cancer. 2010, 10: 296-306. 10.1186/1471-2407-10-296.View ArticleGoogle Scholar
- Neve RM, Chin K, Fridlyand J, Yeh J, Baehner FL, Fevr T, Clark L, Bayani N, Coppe JP, Tong F, Speed T, Spellman PT, DeVries S, Lapuk A, Wang NJ, Kuo WL, Stilwell JL, Pinkel D, Albertson DG, Waldman FM, McCormick F, Dickson RB, Johnson MD, Lippman M, Ethier S, Gazdar A, Gray JW: A collection of breast cancer cell lines for the study of distinct cancer subtypes. Cancer Cell. 2006, 10: 515-527. 10.1016/j.ccr.2006.10.008.View ArticleGoogle Scholar
- Yoneda T, Williams PJ, Hiraga T, Niewolna M, Nishimura R: A bone seeking clone exhibits different biological properties from the parental MDA-MB-231 human breast cancer cells and a brain seeking clone in vivo and in vitro. J Bone Miner Res. 2001, 16: 1486-1495. 10.1359/jbmr.2001.16.8.1486.View ArticleGoogle Scholar
- Draffin JE, McFarlane S, Hill A, Johnston PG, Waugh DJJ: CD44 potentiates the adherence of metastatic prostate and breast cancer cells to bone marrow endothelial cells. Cancer Res. 2004, 64: 5702-5711. 10.1158/0008-5472.CAN-04-0389.View ArticleGoogle Scholar
- Maxwell PJ, Gallagher R, Seaton A, Wilson C, Scullin P, Pettigrew J, Stratford IJ, Williams KJ, Johnston PG, Waugh DJJ: HIF-1 and NF-B-mediated upregulation of CXCR1 and CXCR2 expression promotes cell survival in hypoxic prostate cancer cells. Oncogene. 2007, 26: 7333-7345. 10.1038/sj.onc.1210536.View ArticleGoogle Scholar
- Farmer P, Bonnefoi H, Becette V, Tubiana-Hulin M, Fumoleau P, Larsimont D, Macgrogan G, Bergh J, Cameron D, Goldstein D, Duss S, Nicoulaz AL, Brisken C, Fiche M, Delorenzi M, Iggo R: Identification of molecular apocrine breast tumors by microarray analysis. Oncogene. 2005, 7: 4660-4671.View ArticleGoogle Scholar
- Madsen MA, Deryugina EI, Niessen S, Cravatt BF, Quigley JP: Activity-based protein profiling implicates urokinase activation as a key step in human fibrosarcoma intravasation. J Biol Chem. 2006, 281: 15997-16005. 10.1074/jbc.M601223200.View ArticleGoogle Scholar
- Conn EM, Botkjaer KA, Kupriyanova TA, Andreasen PA, Deryugina EI, Quigley JP: Comparative analysis of metastasis variants derived from human prostate carcinoma cells: roles in intravasation of VEGF-mediated angiogenesis and uPA-mediated invasion. Am J Pathol. 2009, 175: 1638-1642. 10.2353/ajpath.2009.090384.View ArticleGoogle Scholar
- Conn EM, Madsen MA, Cravatt BF, Ruf W, Deryugina EI, Quigley JP: Cell surface proteomics identifies molecules functionally linked to tumor cell intravasation. J Biol Chem. 2008, 283: 26518-26528. 10.1074/jbc.M803337200.View ArticleGoogle Scholar
- Sabeh F, Ota I, Holmbeck K, Birkedal-Hansen H, Soloway P, Balbin M, Lopez-Otin C, Shapiro S, Inada M, Krane S, Allen E, Chung D, Weiss SJ: Tumor cell traffic through the extracellular matrix is controlled by the membrane-anchored collagenase MT1-MMP. J Cell Biol. 2004, 167: 769-781. 10.1083/jcb.200408028.View ArticleGoogle Scholar
- Ishikawa T, Kamiyama M, Tani-Ishii N, Suzuki H, Ichikawa Y, Hamaguchi Y, Momiyama N, Shimada H: Inhibition of osteoclast differentiation and bone resorption by cathepsin K antisense oligonucleotides. Mol Carcinogen. 2001, 32: 84-91. 10.1002/mc.1067.View ArticleGoogle Scholar
- Ohno S, Im HJ, Knudson CB, Knudson W: Hyaluronan oligosaccharides induce matrix metalloproteinase 13 via transcriptional activation of NFkappaB and p38 MAPK in articular chondrocytes. J Biol Chem. 2006, 281: 17952-17960. 10.1074/jbc.M602750200.View ArticleGoogle Scholar
- Matsumoto M, Kogawa M, Wada S, Takayanagi H, Tsujimoto M, Katayma S, Hisatake K, Nogi Y: Essential role of p38 mitogen activated protein kinase in cathepsin K gene expression during osteoclastogenesis through association of NFATc1 and PU.1. J Biol Chem. 2004, 279: 45969-45979. 10.1074/jbc.M408795200.View ArticleGoogle Scholar
- Bourguignon LY, Gunja-Smith Z, Iida N, Zhu HB, Young LJ, Muller WJ, Cardiff RD: CD44v(3,8-10) is involved in cytoskeletal-mediated tumor cell migration and matrix metalloproteinase association in metastatic breast cancer cells. J Cell Physiol. 1998, 176: 206-215. 10.1002/(SICI)1097-4652(199807)176:1<206::AID-JCP22>3.0.CO;2-3.View ArticleGoogle Scholar
- Bourguignon LY, Zhu H, Shao L, Chen YW: CD44 interaction with c-Src kinase promotes cortactin-mediated cytoskeleton function and hyaluronic acid-dependent ovarian tumor cell migration. J Biol Chem. 2001, 276: 7327-7333. 10.1074/jbc.M006498200.View ArticleGoogle Scholar
- Bourguignon LY, Zhu H, Shao L, Chen YW: CD44 interaction with Tiam1 promotes Rac1 signalling and hyaluronic acid-mediated breast tumor cell migration. J Biol Chem. 2000, 275: 1829-1838. 10.1074/jbc.275.3.1829.View ArticleGoogle Scholar
- Sheridan C, Kishimoto H, Fuchs RK, Mehrota S, Bhat-Nakshatri P, Turner CH, Goulet R, Badve S, Nakshatri H: CD44+/CD24- breast cancer cells exhibit enhanced invasive properties: an early step necessary for metastasis. Breast Cancer Res. 2006, 8: R59-10.1186/bcr1610.View ArticleGoogle Scholar
- Artym VV, Zhang Y, Seillier-Moiseiwitsch F, Yamada KM, Mueller SC: Dynamic interactions of cortactin and membrane type-1 matrix metalloproteinase at invadopodia: defining the stages of invadopodia formation and function. Cancer Res. 2006, 66: 3034-3043. 10.1158/0008-5472.CAN-05-2177.View ArticleGoogle Scholar
- Hotary K, Li XY, Allen E, Stevens SL, Weiss SJ: A cancer cell metalloproteinase triad regulates the basement membrane transmigration program. Genes Dev. 2006, 20: 2673-2686. 10.1101/gad.1451806.View ArticleGoogle Scholar
- Littlewood-Evans AJ, Bilbe G, Bowler WB, Farley D, Wlodarksi B, Kokubo T, Inaoka T, Sloane J, Evans DB, Gallagher JA: The osteoclast-associated protease cathepsin K is expressed in human breast carcinoma. Cancer Res. 1997, 57: 5386-5390.Google Scholar
- Udabage L, Brownlee GR, Waltham M, Blick T, Walker EC, Heldin P, Nilsonn SK, Thompson EW, Brown TJ: Antisense mediated suppression of hyaluronan synthase 2 inhibits the tumorigenesis and progression of breast cancer. Cancer Res. 2005, 65: 6139-6150. 10.1158/0008-5472.CAN-04-1622.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.