Localization of uPAR and MMP-9 in lipid rafts is critical for migration, invasion and angiogenesis in human breast cancer cells

MDA MB231, ZR 751, HMEC cells were cultured according to methods established in our laboratory (24). MβCD, nystatin and NAC were purchased from Sigma (St. Louis, MO). Anti-Src-pY416, anti-Src-pY527, and pAkt were purchased from Cell Signaling (Danvers, MA). Antibodies for the phosphorylated and total forms of ERK, FAK, PI3K, Akt and Cav were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies against MMP-9, flotillin and CD71 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The uPAR antibody was purchased from R&D Systems (Minneapolis, MN). Secondary antibodies were also purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Alexa Fluor antibodies and Vybrant lipid raft labeling kit were purchased from Invitrogen (Carlsbad, CA). Caveolae/raft isolation kit was purchased from Sigma (St. Louis, MO).

Target cells were incubated with Dulbecco's modified Eagle medium (DMEM) containing different concentrations of MβCD for 1, 4, 24 and 48 hrs. Supernatants were collected and assessed for cellular toxicity using a LDH cytotoxicity assay kit (Promega, Madison, WI) in accordance with the manufacturer's instructions.

MDA-MB-231 and ZR 751 cells were washed twice with phosphate-buffered saline (PBS) and then incubated at 37°C for 1 hr with different concentrations of MβCD in DMEM. After two washes with PBS, cells were harvested for cholesterol measurement assays as per the manufacturer's instructions (Molecular Probes Inc, Eugene, OR).

Confluent MDA-MB-231 and ZR 751 cells grown in eight-well chamber slides were left untreated or pretreated with different concentrations of MβCD for one hour at 37°C, washed and labeled for lipid raft marker flotiilin or with GM1 (lipid raft labeling kit, Molecular Probes) as per the manufacturer's instructions. The cells were washed, fixed with 2% paraformaldehyde, permeabilized in 0.5% Triton X-100, and stained with primary antibodies against uPAR, MMP-9, pFAK, pCav, pPI3K, and GM1 overnight at 4°C. The cells were then washed, incubated with appropriate secondary antibodies for 1 hr at room temperature, washed, mounted with anti-fading agent containing DAPI, and examined under a confocal fluorescence microscope. For colocalization studies of uPAR and LAMP-1, MDA-MB-231 cells grown in eight-well chamber slides were left untreated or treated with 7.5 mM MβCD, fixed with 2% paraformaldehyde at indicated time points (1, 8 and 24 hrs), and incubated with primary antibodies against uPAR and LAMP-1. The cells were then washed and further incubated with appropriate secondary Alexa Fluor antibodies, and examined under a Olympus fluoview confocal microscope.

MDA-MB-231 and ZR 751 cells (1.5 × 10⁴) were suspended in the appropriate medium, seeded onto 0.5% agar-coated plates, and cultured until spheroids formed. Intact tumor spheroids were selected and transferred to six-well plates. The spheroids were either left untreated or treated with different concentrations of MβCD and followed by incubation for 48 hrs. Then, the spheroids with migrated cells were fixed with 10% buffered formalin in PBS and stained using crystal violet staining solution. The spheroids were observed under a normal light microscope and photographed at 20× and 40× magnification using the DP controller imaging software.

Conditioned medium was collected from MDA-MB-231 and ZR 751 control cells and cells that were treated with different concentrations of MβCD. MDA MB231 and ZR 751 cells pretreated with ruPA (10 ng/μl) and rMMP-9 (2.5 ng/μl) were used as positive controls. The conditioned media was subjected to fibrin and gelatin zymography as described by Gondi et al. [23]. MMP-9 activity was quantified as arbitrary units and compared with controls.

Isolation of total RNA from treated and untreated cells was carried out using the TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions.

Total RNA was isolated from cells in all treatment conditions using the TRIzol reagent as per the standard protocol. Total RNA was treated with DNAse I (Invitrogen, Carlsbad, CA) to remove contaminating genomic DNA. PCR analysis was done using the one-step reverse transcription-PCR kit (Invitrogen, Carlsbad, CA). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control. The following primers were used:

5'-GATCCAGCCAGGGCAGAG-3' antisense

5'-TGCCTTTGGACACGCACGAC-3' antisense

The PCR conditions were as follows: 95°C for 5 min, followed by 30 cycles of 95°C for 1 min, and annealing temperature set according to the AT and GC content of the primers.

MDA-MB-231 and ZR 751 cells were either left untreated or pretreated with different concentrations of MβCD for 1 hr at 37°C. Total cell lysates were prepared and resolved on SDS-polyacrylamide gels, transferred onto nitrocellulose paper, and immunoblotted with appropriate primary antibodies.

Untreated MDA-MB-231 and ZR 751 cells and cells pretreated with different concentrations of MβCD for 1 hr at 37°C were allowed to incubate with serum-free medium for 24 and 48 hrs. Conditioned medium collected from the treated and untreated cells were subjected to uPAR immunoassay as per the manufacturer's instructions (R&D Systems, Minnepolis, MN).

MDA-MB-231 cells (2 × 10⁴) and ZR 751 cells were seeded in 100-mm plates and either left untreated or treated with different concentrations of MβCD for 1 hr at 37°C. After treatment, the medium was removed; washed and serum-free medium was added. Conditioned medium was collected after 24 hrs. Human microvascular endothelial cells (HMEC) were cultured in the conditioned medium for 24 hrs. After the incubation period, the medium was removed, and the cells were stained with Hema-3 stain (Fisher Diagnostics, Fisher Scientific Company, Middletown, VA) and examined under a microscope. Image Pro software (Media Cybernetics, Silver Spring, MD) was used for quantification of angiogenesis. The degree of angiogenesis was measured by the following method: number of branch points and the total number of branches per point, with the product indicating the degree of angiogenesis.

MDA-MB 231 cells (1 × 10⁶) were left untreated or pretreated with different concentrations of MßCD for 1 h at 37°C, washed and allowed to invade through matrigel - coated transwell inserts (8 μM pores) for 24 hrs. Cells which invaded through the Matrigel coated inserts were stained, counted and photographed under an Olympus light microscope at 20× magnification. Quantitative estimation of invasion was done as per the method previously established in our laboratory [24].

MDA-MB-231 and ZR 751 cells (1-2 × 10⁷/25-cm² culture flasks) were serum starved for 14 to 16 hrs, and cells were left untreated, pretreated with 7.5 mM MβCD or 10 mM MβCD for 1 hr at 37°C to allow for cholesterol depletion (60%). Lipid raft fractions were isolated using the Raft Isolation Kit (Sigma, Inc., St.Louis, MO) Rafts were also isolated using the procedure of Raghu et al. [3]. The whole procedure was performed at 0-4°C. DRM and soluble fractions were collected and subjected to electrophoresis and Western blotting. Fractions 3, 4, and 5 were pooled and marked as raft fractions and fractions 6, 7, and 8 were labeled as non-raft fractions. Raft fractions and non-raft fractions were characterized based on the specificity to caveolin and transferrin in MDA-MB-231 cells and flotillin and transferrin in ZR 751 cells.

Cholesterol depletion was done by treating cells with different concentrations of MβCD for 1 hr at 37°C followed by washing and incubation with serum-free DMEM for the indicated time periods. For cholesterol supplementation, the cells were treated initially with 10 mM MβCD, and then the cells were incubated for different time periods with or without the cholesterol MβCD complexes. At the end of each indicated time period, the cells were washed and incubated with serum-free medium, and the lysates were collected for western blot analysis.

As previously reported [6, 7], uPAR, a GPI-anchored receptor protein, is a lipid raft-associated protein. In Figure 1A, GM1 staining, which is an LR specific marker, was observed predominantly in the cell periphery of unstimulated MDA-MB-231 cells (Figure 1A, panel a). Also, uPAR staining was seen more near the plasma membrane (Figure 1A, panel b). A very clear margin of colocalization of uPAR the with LR marker GM1 was observed near the plasma membrane of MDA-MB-231 cells (Figure 1A, panel c). Colocalization studies were done with GM1 at 1 h and 24 h time points after 1 h treatment with MβCD. In cells pretreated with 7.5 mM MβCD, we found disrupted GM1 spots in the cells (Figure 1A, panel d). In addition, we observed inhibition of uPAR staining at the plasma membrane (Figure 1A, panel e). Moreover, we did not observe colocalization of GM1 with uPAR in MβCD-treated cells (Figure 1A, panel f). Morphological analysis revealed the association of uPAR with another lipid raft marker, flotillin, in ZR 751 cells (Figure 1B, panels a to c). In cells pretreated with MβCD, we observed significant inhibition of the colocalization of flotillin with uPAR in ZR 751 cells (Figure 1B, panels d to f).

In untreated cells, we observed increased activity of uPA in MDA-MB-231 and ZR 751 cells as assessed by fibrin zymography (Figure 1C, lane 2 top panel (MDA MB231), bottom panel, lane 2(ZR 751). In cells treated with MβCD, inhibition of uPA activity was observed at 24 h time point (Figure 1C top panel lane 3 & 4 (MDA MB 231) and bottom panel lane 3 & 4 (ZR 751). Conditioned medium from cells treated with ruPA (10 ng/μl) was taken as a positive control (Figure 1C lane 1, top and bottom panel). Lysates from two time points namely at 1 hr and 24 hrs showed significant decreases in uPAR protein levels in LR-disrupted cells when compared with untreated controls. The decrease in expression levels was also dose-dependent; 10 mM MβCD treatment resulted in 40% decrease in expression levels while 7.5 mM MβCD treatment resulted in 30% decrease in expression levels in MDA-MB-231 cells (Figure 1D, lanes 2 and 3). Equal loading of total lysates was confirmed with GAPDH. In ER-positive breast cancer cells, we observed a similar pattern of uPAR expression in unstimulated cells. In cells treated with different concentrations of MβCD, we observed a 30-35% reduction in uPAR protein levels (Figure 1E, lanes 2 and 3). uPAR expression in ZR 751 cells was observed for up to 24 hrs (Figure 1E, lanes 2 and 3). These results suggest that lipid raft disruption by plasma membrane cholesterol depletion plays an active role in uPA and uPAR expression levels in both MDA-MB-231 and ZR 751 cells.

Previous studies have shown that MMP-9 is associated with lipid rafts (14). To ascertain if such an association exists in breast carcinoma cells, we used untreated and MβCD-treated breast carcinoma cells and checked for colocalization of MMP-9 with GM1 in MDA-MB-231 cells and with flotillin in ZR 751 cells. As expected, we observed lipid raft staining at the cell periphery in MDA-MB-231 cells (Figure 2A, panel a), and MMP-9 staining was more concentrated at the leading edge of the cells (Figure 2A, panel b). In the merged image, we observed colocalization of MMP-9 with the lipid raft marker GM1 at the leading edge and at the margin of the cell, suggesting that MMP-9 associates with lipid rafts at the leading edge and in the periphery of the cells. When the cells were pretreated with 7.5 mM MβCD for 1 hr at 37°C, we did not observe colocalization of MMP-9 and GM1 (Figure 2A, panels d to f). In ZR 751 cells, we observed colocalization of MMP-9 with the lipid raft marker flotillin in untreated cells (Figure 2B, panels a to c) while in cells treated with 7.5 mM MβCD, we observed significant inhibition of the colocalization of MMP-9 with flotillin (Figure 2B, panels d to f). These results suggest that MMP-9 colocalizes with lipid rafts, and lipid raft disruption by cholesterol depletion inhibits this association.

Morphological studies have shown that MMP-9, a zinc endopeptidase, associates with lipid rafts at the periphery and at the leading edge of the cells. To ascertain if lipid raft disruption by cholesterol depletion inhibits MMP-9 activity, we determined MMP-9 activity using gelatin zymography. Consistent with our previous results [24], we observed increased MMP-9 activity in conditioned medium at the 24 hr time point in untreated cells (Figure 2C, lane 1 top panel (MDA MB231) and Figure 2C lane 2 bottom panel (ZR 751). However, in the MβCD-treated cells, there was 25-35% decrease in MMP-9 activity (Figure 2C, lanes 2 and 3 (MDA MB 231 and Figure 2C lane 3, 4 (ZR 751). As a positive control the cells were treated with rMMP-9 (2.5 ng/μl) (Figure 2C lane 4 top panel (MDA MB231) and Figure 2C lane 1 bottom panel (ZR 751). To further substantiate our results, we wanted to determine if lipid raft disruption also affected the MMP-9 protein expression levels in MDA-MB-231 and in ZR 751 cells. MMP-9 levels in the total cell lysates were measured using western blotting. As shown previously, we observed a higher level of MMP-9 expression in untreated cells (Figures 2D-E, lane 1) as compared with MDA-MB-231 and ZR 751 cells pretreated with different concentrations of MβCD (Figure 2D, lanes 2 and 3). The MMP-9 expression levels in MDA-MB-231 cells showed a decrease in MMP-9 protein levels up to 24 hrs in MβCD-treated cells. Equal loading of lysates is demonstrated by GAPDH expression (Figure 2D, top and bottom panel). The MMP-9 expression levels in ZR 751 cells showed a decrease in MMP-9 protein levels up to 1 hr in MβCD-treated cells. At the 24 hr time point, we did not observe significant changes in MMP-9 levels in untreated and lipid raft-disrupted ZR 751 cells (Figure 2E, lanes 2 and 3). These results show that in both ER-positive and ER-negative cell lines, lipid rafts are critical for uPAR and MMP-9 expression and activity. The differences in MMP-9 protein expression levels in MDA-MB-231 and ZR 751 cells demonstrate the cell type variation in MMP-9 levels mediated by lipid raft disruption.

To determine whether treatment of MDA-MB-231 and ZR 751 cells with various concentrations of MβCD and nystatin efficiently extracted cholesterol and to check whether the cholesterol levels remained depleted even after 24 hrs, we assayed cellular cholesterol levels. As shown in Figures 3A and 3B, after treatment with non-toxic doses of the drugs, cellular cholesterol levels decreased in a dose-dependent manner at both the 1 hr and 24 hr time points in both of the cell lines tested. With 50 and 100 μg concentrations of nystatin, we observed a 20-30% reduction in ZR 751 cells at 1 hr and 24 hrs; in MDA-MB-231 cells, we observed a 25-40% reduction in cellular cholesterol levels (Figures 3A-B). With 5 mM, 7.5 mM and 10 mM MβCD, we found 20-30% reduction in cellular cholesterol levels in ZR 751 cells and 25-55% reduction in MDA-MB-231 cells (Figure 3A). In contrast, we did not observe a significant change in cellular cholesterol levels when cells were treated with 20 mM NAC (n-acetyl cysteine (a known antioxidant) for 1 hr at 37°C (Figures 3A-B). To further confirm that concentration of MβCD used is nontoxic to the cells, we did the cellular cytotoxicity assay using the Lactate dehydrogenase assay kit. Results showed that both MDA MB231 and ZR 751 cells tolerated 5 mM, 7.5 mM and 10 mM concentrations of MβCD upto 48 h time points and the concentrations used were nontoxic to the cells. (Figure 3C). These results demonstrate that both MβCD and nystatin were efficient in depleting or chelating cholesterol from the lipid rafts of MDA-MB-231 and ZR 751 cells under our testing conditions.

As shown previously in Figure 1C and Figure 2C, when lipid rafts were disrupted by depletion of plasma membrane cholesterol levels in MDA-MB-231 and ZR 751 cells, we observed decreased uPA and MMP-9 activity in the conditioned medium at the 24 hr time point. In addition, we found significant decreases in uPAR and MMP-9 protein levels in the LR-disrupted lysates as compared with untreated lysates. Next, we determined whether uPAR and MMP-9 are associated with LR microdomains by biochemical isolation of lipid raft fractions from both cell types. We carried out biochemical isolation of lipid rafts using the detergent method [25]. Fractions 3, 4 and 5, which were positive for the lipid raft marker caveolin (Figure 3A, left upper panel) and negative for transferrin (Figure 4A, left bottom panel), were pooled and labeled as lipid raft fractions. Fractions 6, 7 and 8, which were positive for transferrin, a non-raft marker, (Figure 4A, right bottom panel) and negative for caveolin (Figure 4A, right upper panel), were pooled and labeled as non-raft fractions. uPAR and MMP-9 in the pooled raft fractions (3 to 5) and non-raft fractions (6 to 8) were examined by immunoprecipitation assays. Higher levels of uPAR and MMP-9 were observed in raft fractions of both MDA-MB-231 and ZR 751 cells (Figures 4A-B) than in the non-raft fractions. MβCD treatment significantly decreased the association of uPAR and MMP-9 with the raft fractions, resulting in little or no activity similar to non-raft fractions. These results suggest that both uPAR and MMP-9 associate with lipid rafts in MDA-MB-231 and ZR 751 cells.

As previously shown in Figure 1C-E and in Figure 2C-E, lipid raft disruption downregulated uPAR, MMP-9 and uPA activity and protein expression in both MDA-MB-231 and ZR 751 cells. As a consequence, we assessed whether lipid raft disruption affects gene expression in breast carcinoma cells. MDA-MB-231 and ZR 751 cells were left untreated or pretreated with different concentrations of MβCD for one hour at 37°C, and total RNA was extracted at 4, 8 and 24 hrs. uPAR mRNA and MMP-9 mRNA levels were measured by semi-quantitative RT-PCR. As shown in Figures 4C-D, we observed significant inhibition of uPAR mRNA in both MDA-MB-231 and ZR 751 cells at early time points up to 8 hrs after treatment. mRNA analysis at the 24 hr time point showed no changes in uPAR mRNA or MMP-9 mRNA in LR-disrupted cells (data not shown). These results further strengthened our previous findings, and the downregulation of uPAR and MMP-9 protein levels upon lipid raft disruption could be due to downregulation of uPAR and MMP-9 gene expression at earlier time points. Taken together, these results suggest that lipid rafts play an active role in the modulation of uPAR and MMP-9 activities in both MDA-MB-231 and ZR 751 cells. The observed decrease in uPAR and MMP-9 activity upon lipid raft disruption in both cell lines could be very critical to the cellular proliferation of breast carcinoma.

Previous studies have shown that Src is associated with lipid raft domains in different cell types [3]. To determine if lipid raft disruption affected Src phosphorylation, lysates prepared from MDA-MB-231 cells before and after treatment with 7.5 and 10 mM MβCD were probed with antibodies against phospho-Src pY416. As expected, Src phosphorylation was high in untreated cells (Figure 5A, upper panel, lane 1). In cells pretreated with MβCD, there was significant downregulation of phospho-Src in both MDA-MB-231 and ZR 751 cells. Previous studies from our lab have shown that uPAR activation in breast carcinoma cells activates Src, FAK, PI3-K and Akt signaling pathways [24]. As most of these signaling molecules are associated with lipid rafts [3], we wanted to check if lipid raft disruption regulates these signaling molecules in both MDA-MB-231 and ZR 751 cells. Western blotting analysis of the lysates revealed higher levels of pFAK, pPI3-K, pAkt, pCav and pERK in untreated MDA-MB-231 and ZR 751 cells (Figures 5A-B). We observed significant reductions in the phosphorylated levels of FAK, Akt, PI3-K, caveolin and ERK in LR-disrupted cells when compared with untreated cells (Figures 5A-B, lanes 2 and 3). Equal loading of total lysates was confirmed by GAPDH. We did not find significant differences in the total levels of Src, FAK, caveolin, ERK and Akt. These results suggest that lipid rafts play a significant role in Src, FAK, PI3-K, Akt and caveolin phosphorylation in MDA-MB-231 and ZR 751 cells.

To further confirm our results, immunofluorescent analysis of untreated and LR-disrupted cells revealed association of FAK, pCav, and PI3-K with the LR marker GM1 in untreated MDA-MB-231 cells (Figure 5C, panels c, i and o) and the absence of this colocalization in MβCD-treated cells (Figure 5C, panels f, l and r). Similar results were observed with pSrc and pERK (data not shown). Overall, these results suggest that lipid rafts play a critical role in Src, FAK, caveolin, ERK and Akt phosphorylation in MDA-MB-231 and ZR 751 cells.

Previously in Figures 1 & 2 we have shown that LR disruption in MDA-MB-231 and ZR 751 cells inhibited uPAR and MMP-9 activity. Consequently, we evaluated whether LR disruption affects migration and invasion of breast carcinoma cells. Figure 6A shows spheroids, which were left untreated, had a high number of cells that migrated from the spheroids into the surrounding area. However, LR-disrupted cells failed to migrate, which resulted in no migration (Figure 6A, panels a to f). MDA-MB-231 cells were left untreated or pretreated with different concentrations of MβCD and allowed to invade through Matrigel-coated filters. Figure 6B shows the staining of LR-disrupted MDA-MB-231 cells was significantly less than that of the untreated cells. Quantitative analysis of spheroid migration in MDA MB231 and ZR 751 cells before and after treatment with MβCD showed 50-70% reduction in migration (Figure 6C). Quantitative analysis of the cells showed that only 30-40% of the LR-disrupted MDA MB231 cells invaded the Matrigel as compared to the untreated cells (Figure 6D). Taken together, these findings provide strong evidence that by modulating the lipid rafts in breast carcinoma cells, the migration and invasion of MDA-MB-231 cells and ZR 751 cells are significantly inhibited as compared to cells in which the integrity of the lipid rafts were maintained. These results indicate that lipid rafts are critical for the migratory and invasive properties of breast carcinoma cells.

As studies have shown that cholesterol depletion plays a critical role in VEGF signaling [26], we determined whether cholesterol depletion by MβCD plays a role in the angiogeneic potential of breast carcinoma cells. Conditioned medium prepared from the supernatants of MDA-MB-231 and ZR 751 cells that were left untreated or pretreated with different concentrations of MβCD were cultured with human microvascular endothelial cells. We evaluated the tumor conditioned medium-induced vessel formation using an in vitro angiogenesis assay. The results showed that conditioned medium prepared from cells that were treated with MβCD significantly reduced the capacity of endothelial cells to form capillary-like structures (Figure 6E, panels b, c, e and f). In contrast, conditioned medium from untreated cells resulted in well-formed capillary-like structures by the 48 hr time point (Figure 6E, panels a and d). The quantification of branch points and number of branches confirmed a significant reduction in the MβCD-treated samples when compared with the untreated cells (Figure 6F). These results suggest that lipid raft disruption downregulates the angiogenic potential of breast carcinoma cells.

After having demonstrated the effect of lipid raft disruption on uPAR protein expression in the cell lysates of breast carcinoma cells, we assessed the levels of soluble uPAR in the conditioned medium of LR-disrupted cells at 48 hrs. We used a commercially available uPAR immunoassay detection kit (R&D Systems) and performed the assay as per the manufacturer's instructions. As shown in Figure 6G, we observed a 7-to 8-fold increase in soluble uPAR levels in MDA-MB-231 cells pretreated with the LR-disrupting agent MβCD and a 5-to 6-fold increase in soluble uPAR levels in ZR 751 cells pretreated with MβCD. Overall, these results suggest that lipid raft disruption promotes the removal of uPAR from the plasma membrane, and the released soluble uPAR is secreted for at least 48 hrs in breast carcinoma cells.

uPAR is known to cycle back to the plasma membrane after it has been endocytosed with LRP, PAI and uPA [27]. To study the fate of uPAR after lipid raft disruption, we carried out trafficking studies of uPAR at different time points using immunofluorescent staining for LAMP-1, a lysosomal marker, and uPAR at 1, 8 and 24 hr time points. In untreated cells, we observed very little colocalization of uPAR with LAMP-1 in MDA-MB-231 cells, suggesting that the majority of uPAR is not targeted to the lysosomal pathway in LR-intact cells (Figure 7A, panels a to i). However, in LR-disrupted cells, we observed significant colocalization of uPAR with LAMP 1 as early as one hour after treatment (Figure 7B, panels a to c) and further increased by 8 hrs (Figure 7B, panels d to f) and 24 hrs (Figure 7B, panels g to i). The results suggest that lipid raft disruption causes the targeting of uPAR to the lysosome, which may contribute significantly to the loss of the uPAR function in breast cancer cells.

To further examine a possible role for cholesterol in the downregulation of uPAR in MDA-MB-231 and ZR 751 cells, MDA-MB-231 and ZR 751 cells were left untreated, treated with MβCD alone or cholesterol addition after cholesterol depletion. Lysates were prepared and subjected to immunoblotting analysis. MβCD-induced downregulation of uPAR levels in MDA-MB-231 and ZR 751 cells were reversed by the addition of cholesterol as assessed by immunoblotting (Figures 8A-B). These results indicate that cholesterol repletion reverses the MβCD-induced uPAR activity in breast carcinoma cells.