Heparanase Regulates Secretion, Composition, and Function of Tumor Cell-derived Exosomes^,

Background: Heparanase drives the progression of many tumor types.

Results: Heparanase enhances exosome secretion and alters exosome composition and function.

Conclusion: Heparanase promotes tumor progression by regulating exosome secretion and composition.

Significance: Therapeutic inhibition of heparanase may decrease exosome secretion and slow tumor progression.

Abstract

Emerging evidence indicates that exosomes play a key role in tumor-host cross-talk and that exosome secretion, composition, and functional capacity are altered as tumors progress to an aggressive phenotype. However, little is known regarding the mechanisms that regulate these changes. Heparanase is an enzyme whose expression is up-regulated as tumors become more aggressive and is associated with enhanced tumor growth, angiogenesis, and metastasis. We have discovered that in human cancer cells (myeloma, lymphoblastoid, and breast cancer), when expression of heparanase is enhanced or when tumor cells are exposed to exogenous heparanase, exosome secretion is dramatically increased. Heparanase enzyme activity is required for robust enhancement of exosome secretion because enzymatically inactive forms of heparanase, even when present in high amounts, do not dramatically increase exosome secretion. Heparanase also impacts exosome protein cargo as reflected by higher levels of syndecan-1, VEGF, and hepatocyte growth factor in exosomes secreted by heparanase-high expressing cells as compared with heparanase-low expressing cells. In functional assays, exosomes from heparanase-high cells stimulated spreading of tumor cells on fibronectin and invasion of endothelial cells through extracellular matrix better than did exosomes secreted by heparanase-low cells. These studies reveal that heparanase helps drive exosome secretion, alters exosome composition, and facilitates production of exosomes that impact both tumor and host cell behavior, thereby promoting tumor progression.

Introduction

Exosomes are membrane-bound vesicles that are released into the extracellular space when a multivesicular body fuses with the plasma membrane (1). Exosomes are defined predominantly by their size (∼30–120 nm) and by the presence of specific proteins, for example, tetraspanins (CD63, CD81), clathrin, and flotillin-1 (2). Exosomes contain cargo of microRNA, mRNA, lipids, and proteins that can be delivered to local or distant sites within the body, thereby providing a means for intercellular communication (3, 4). When exosomes come in contact with recipient cells, the cells can undergo immediate stimulation through cell signaling mediated by exosomal proteins, or cell behavior can be modified by the proteins, microRNA, or mRNA delivered by the exosome. Most cells secrete exosomes, but their secretion is up-regulated in disease states. In cancer, secretion of exosomes often increases as tumors transit toward a more aggressive phenotype. The tumor-host cross-talk mediated by exosomes has multiple effects that can influence processes such as formation of the premetastatic niche, angiogenesis, and host immune function (5–9). Additionally, exosomes can carry tumor-specific biomarkers useful for early detection and/or diagnosis and, because they may be taken up selectively by some cells within tumors, exosomes have the potential to be utilized as drug delivery vehicles (6). Although research on exosomes has intensified over the last five years, there are still many unanswered questions regarding what regulates exosome secretion, their contents, and their downstream effects on recipient cells. Several Rab protein family members have been shown to control exosome formation, and more recently, it was reported that syndecan proteoglycans, through their interaction with the syntenin-Alix complex, influence the biogenesis of exosomes (9–12).

Heparanase is an endoglycosidase that cleaves heparan sulfate and is up-regulated in many cancers. High levels of heparanase in cancer patients are associated with shorter postoperative survival time as compared with patients with low levels of heparanase (13). Although some of the tumor-promoting effects of heparanase can be attributed to its ability to remodel the extracellular matrix barrier by cleaving heparan sulfate, heparanase is also known to regulate cell signaling and gene transcription (13–16). Elevation of heparanase levels in myeloma cells, either by transfection of cells or by addition to cells of recombinant active heparanase enzyme, up-regulates expression of matrix metalloproteinase-9 (MMP-9), VEGF, HGF,² and receptor activator of nuclear factor κ-B ligand (RANKL), which together drive an aggressive tumor phenotype (17–20). Through these multiple functions that facilitate cross-talk between tumor and host cells, heparanase via both enzyme activity-dependent and enzyme activity-independent mechanisms regulates tumor growth, angiogenesis, and metastasis (13, 21).

In the present work, we demonstrate that heparanase causes a marked increase in the secretion of exosomes. Exosome protein composition was also altered by heparanase as demonstrated by increased levels of syndecan-1, VEGF, and HGF, three proteins associated with an aggressive tumor phenotype. Moreover, the addition of exosomes secreted by cells expressing high levels of heparanase altered the behavior of both tumor and host cells. These studies provide the first evidence that heparanase can regulate exosome secretion, composition, and function and reveal a novel mechanism whereby heparanase governs tumor-host cross-talk, thereby enhancing aggressive tumor behavior.

EXPERIMENTAL PROCEDURES

Cells and Cell Cultures

CAG (22), ARH-77 (ATCC), and MDA-MB-231 (ATCC) cells were cultured in RPMI 1640 growth medium (Cellgro) supplemented with 10% fetal bovine serum, 1% antibiotic/antimycotic, and l-glutamine (Mediatech, Herndon, VA). As described previously, CAG and ARH-77 cells were transfected with empty vector or vector containing the cDNA for human heparanase to prepare heparanase-low (HPSE-low) and heparanase-high (HPSE-high) cells, respectively (23). CAG cells were also transfected with a cDNA for heparanase that was mutated at either amino acid 225 or amino acid 343 (designated M225 and M343; glutamic acid to alanine substitution) (23). In some experiments, recombinant heparanase (rHPSE) (24) was added to cells 48 h prior to seeding in serum-free medium and again at the time of seeding in serum-free medium. Human umbilical vein endothelial cells (Cambrix Bioscience) were cultured as described previously (18).

Exosome Isolation and Characterization

Cells were washed twice with PBS and grown in serum-free medium for 24–42 h. In some experiments, bacterial heparinase III (Hep III; 12 μg/ml) was introduced at the beginning of the incubation period. When comparing exosome secretion levels of HPSE-high and HPSE-low cells, plates were seeded with equal numbers of cells, and it was confirmed that cell numbers were still equivalent at the time of harvesting the conditioned medium. Exosomes secreted into the medium was isolated by differential ultracentrifugation (2). Briefly, media were centrifuged at 300 × g for 10 min to clear cells and large debris. The supernatant was then centrifuged at 2000 × g for 20 min and then at 10,000 × g for 30 min to remove residual membranous debris. The remaining supernatant was then subjected to ultracentrifugation at 100,000 × g for 70–120 min to pellet the exosomes. The pellets were resuspended in PBS and repelleted at 100,000 × g for 70–120 min to remove contaminating proteins and resuspended in PBS for further analysis. In some experiments, resuspended exosome pellets were layered on top of a 40% iodixanol cushion (Sigma) and centrifuged at 100,000 × g for 120 min, and the remaining exosome fraction excluded by the cushion was analyzed. The amount of protein present in exosome pellets was determined using a BCA protein assay kit (Pierce), and the number and size of particles was assessed by NanoSight particle tracking (NanoSight Ltd.). Particles of size 30–120 nm were designated as exosomes. As described previously (25), for electron microscopy, 3 μl of exosomes suspended in PBS were placed on a glow-discharged Formvar carbon-coated grid and negatively stained with 2% uranyl acetate solution. For cryo-electron microscopy, 3 μl of exosomes were placed on C-flat holey film, blotted, and frozen in liquid ethane. Images were taken using FEI Tecnai F20 electron microscope operated at 200 kv, and images were captured on a 4k × 4k CCD camera. For Western blots of exosome proteins, samples were loaded onto a 10% or a 4–20% gradient SDS-polyacrylamide gel (Bio-Rad), transferred to a positively charged nylon membrane (Nytran SPC, Schleicher & Schuell), and probed with antibody as described (26). Antibodies used were against: heparanase (affinity-purified polyclonal antibody 1453 (27)), flotillin-1 (Abcam), clathrin heavy chain (Abcam), and CD63 (Abcam). Western blots of exosome protein probed with antibody to calnexin (Cell Signaling) were negative, indicating that preparations were free of endoplasmic reticulum contamination (e.g. microsomes).³

ELISA

ELISAs were utilized to quantify syndecan-1 (Cell Sciences), VEGF (BIOSOURCE), and HGF (R&D Systems) following the manufacturer's instructions. For each molecule tested, an equivalent amount of exosome protein isolated from medium conditioned by HPSE-high or HPSE-low cells was utilized.

Analysis of Exosome Functions

Tumor cell spreading on fibronectin-coated wells was performed as described (28). Cells were stained with rhodamine-phalloidin to assess their phenotype. The effect of isolated exosomes on the invasion of human umbilical vein endothelial cells was assessed using Biocoat Matrigel invasion chambers (BD Biosciences) as described (18).

RESULTS

Heparanase Enhances Exosome Secretion

To begin exploring the relationship between heparanase and exosomes, we isolated exosomes from medium conditioned by the CAG human myeloma cell line expressing heparanase at either high levels (HPSE-high) or low levels (HPSE-low). The level of heparanase expressed in the HPSE-high cells is similar to that found in some myeloma patient tumors, thereby lending physiological relevance to their use (29, 30). We discovered that HPSE-high cells secreted ∼6-fold higher levels of total protein in exosomes per million cells than did the HPSE-low cells (Fig. 1A). This was due to an increase in the number of exosomes in medium from HPSE-high cells as confirmed by counting the particles of the size 30–120 nm using a NanoSight nanoparticle analysis system (123,798 particles/million HPSE-high cells; 20,063 particles/million HPSE-low cells). In addition to NanoSight analysis, the fidelity of the exosome preparations was confirmed by electron microscopy and Western blotting. Electron microscopy of negatively stained exosomes revealed a “cup shape” typical of exosomes, and cryo-electron microscopy demonstrated that the particles isolated from CAG cells were within the size range (30–120 nm) characteristic of exosomes (Fig. 1B). Next, exosomes were isolated from equivalent volumes of conditioned medium from HPSE-high and HPSE-low cells and layered on the top of a 40% iodixanol cushion and centrifuged. Western blots of material excluded by the iodixanol layer confirm that higher levels of the exosome markers flotillin-1, clathrin heavy chain, and CD63 are present in medium conditioned by HPSE-high cells as compared with HPSE-low cells (Fig. 1B, right panel). Because the amount of material loaded in each lane of the gel is from an equal amount of conditioned medium, the result demonstrates the presence of substantially more exosomes in the medium from HPSE-high cells than in the medium of HPSE-low cells, further supporting the conclusion that heparanase expression by cells enhances exosome secretion. In addition, pro-heparanase was readily detected in the exosomes secreted by HPSE-high cells.

FIGURE 1.

Heparanase enhances the amount of exosomes secreted by tumor cells. A, CAG human myeloma cells were transfected with the cDNA for human heparanase (HPSE-high cells) or control vector (HPSE-low cells), and the amount of exosome protein accumulated over 42 h in the cell medium was quantified by BCA protein assay. In addition, exosome protein levels were measured following the addition of rHPSE to HPSE-low cells. *, p < 0.05 versus level of exosome protein in HPSE-low cells. B, characterization of exosomes. Micrographs from electron microscopy of negative stained particles (left panel, upper micrograph) or cryo-electron microscopy (lower panel) were of the size (30–120 nm) and shape consistent with their identity as exosomes (bar = 100 nm). Cryo-electron microscopy also demonstrated that the exosomes isolated from the HPSE-high and HPSE-low cells were similar in size (see Footnote 3), Right panel, Western blots of proteins (flotillin-1, clathrin heavy chain, and CD63) present in exosome preparations excluded by an iodixanol cushion. Each lane is loaded with material purified from an equivalent volume of conditioned medium. Thus, the much higher levels of the exosome markers seen in medium conditioned by CAG HPSE-high cells as compared with medium conditioned by HPSE-low cells indicate that HPSE-high cells are secreting substantially more exosomes. Pro-heparanase (65 kDa) was also detected in exosomes from HPSE-high cells. C, left panel, ARH-77 human lymphoblastoid cells also exhibited enhanced exosome secretion following transfection with the cDNA for heparanase. Right panel, exosome secretion by the human breast carcinoma cell line MDA-MB-231 was enhanced in response to the addition of 125 ng/ml recombinant heparanase. *, p < 0.05. D, left panel, decrease in exosome secretion following treatment of CAG HPSE-high with Hep III. *, p < 0.05. Right panel, a robust enhancement of exosome secretion is dependent on the active form of the heparanase enzyme. CAG cells transfected with a cDNA coding for mutated, enzymatically inactive forms of heparanase (M225 and M343) secrete low levels of exosomes as compared with cells transfected with the cDNA coding for the active form of heparanase (HPSE-high). *, p < 0.01 as compared with M225, M343 and HPSE-low cells. Error bars in panels A, C, and D indicate ± S.D.

To ensure that the increase in exosome secretion in cells transfected with the cDNA for heparanase was not due to an artifact of transfection or the cell selection process, we used a second approach in which rHPSE was added to HPSE-low cells. Previous studies have demonstrated that the addition of rHPSE can impact cell behavior and closely mimic the effect of cells expressing heparanase (17, 31, 32). The addition of rHPSE stimulated exosome secretion in a concentration-dependent manner (Fig. 1A). To determine whether the effect of heparanase on exosome secretion also occurred in cancer types other than myeloma, we examined ARH-77, a human lymphoblastoid cell line. Similar to what was seen with myeloma cells, heparanase enhanced exosome secretion (Fig. 1C, left panel). In addition, recombinant heparanase elevated exosome secretion by MDA-MB-231 human mammary carcinoma cells (Fig. 1C, right panel), further indicating that heparanase exerts its effects on exosomes in a variety of cancers. To extend the in vitro findings, we also analyzed levels of exosomal protein in serum pooled from five normal and five heparanase-transgenic animals (33) and found levels ∼60% higher in the mice overexpressing heparanase (90 μg/ml versus 150 μg/ml of exosomal protein/ml serum from normal versus heparanase transgenic mice, respectively).

It was recently demonstrated that exosome biogenesis in MCF-7 breast cancer cells is controlled by syndecan and also dependent on the presence of heparan sulfate for the formation of a syndecan-syntenin-Alix complex (11). This complex supports intraluminal budding of endosomal membranes, a critical step in exosome formation. To determine whether the heparanase-mediated increase in exosome secretion may also involve this complex, we assessed exosome secretion by CAG HPSE-high cells following treatment with Hep III. (It should be noted that Hep III is a bacterial enzyme that degrades heparan sulfate predominantly into disaccharides, whereas in contrast, human Hep releases fragments of heparan sulfate 5–7 kDa in size, leaving an intact proteoglycan still containing some heparan sulfate.) Similar to what was seen with the MCF-7 cells, degradation of heparan sulfate by Hep III resulted in ∼50% reduction in exosome secretion (Fig. 1D, left panel).

Heparanase is known to have both enzyme-dependent and enzyme-independent effects on tumor cells. To determine whether the robust up-regulation of exosome secretion mediated by heparanase required enzyme activity, conditioned medium was harvested from CAG HPSE-high and HPSE-low cells (as shown in Fig. 1A) and from cells expressing high levels of enzymatically inactive heparanase (mutated heparanase designated M225 and M343). M343 failed to stimulate exosome secretion, whereas M225 had a mild stimulatory effect. However, active heparanase stimulated robust exosome secretion that was significantly greater than both mutated forms of the proteoglycan (Fig. 1D, right panel).

Heparanase Regulates Exosome Protein Cargo

The proteomic content of tumor cell-derived exosomes has been shown to dictate, at least in part, their functional role in cancer progression (34). To determine whether altered heparanase expression impacted the protein composition of exosomes, we harvested exosomes secreted by CAG HPSE-low and HPSE-high cells and analyzed the level of three proteins known to be important in myeloma progression: syndecan-1, VEGF, and HGF (18, 35–37). ELISA results revealed that all three of these molecules were more abundant in exosomes secreted by HPSE-high cells as compared with exosomes from HPSE-low cells (Fig. 2A).

FIGURE 2.

Exosomes from heparanase-high cells have altered composition and regulate the behavior of tumor and host cells. A, ELISA quantification of levels of syndecan-1 (SDC1), VEGF, and HGF present in exosomes isolated from conditioned medium of CAG cells expressing high (H) or low (L) levels of heparanase. Results from each ELISA assay are mean values from three different exosome preparations ± S.D. p < 0.01 for each of the proteins quantified. B, 100 μg of exosomes isolated from HPSE-high or HPSE-low cell conditioned medium were added to HPSE-low cells growing on fibronectin-coated wells. Following overnight incubation, the cells were stained with phalloidin and photographed, and the percentage of spread cells was determined. Bar = 200 μm. Results shown are representative panels from two different experiments using two different exosome preparations. The apparent reduction in cell numbers in photos of wells in which exosomes were added (left and middle panel) is due to the fact that exosomes also cause some cell aggregation. These aggregates remain suspended during the assay, therefore resulting in fewer cells attaching to the dish. C, 8 μg of exosomes purified from medium conditioned by HPSE-high or HPSE-low cells were added to endothelial cells, and the number of cells that invaded through Matrigel-coated chambers overnight was determined. Data are mean ± S.D. of three independent experiments. p < 0.01 for HPSE-high versus HPSE-low.

Exosomes Alter Behavior of Recipient Tumor and Host Cells

To determine whether the exosomes secreted by HPSE-low and HPSE-high cells had different functional capacities, we employed two functional assays. Although CAG cells grow predominantly in suspension when in culture, we found that when plated on fibronectin-coated wells, the HPSE-high CAG cells spread extensively, whereas in contrast, the HPSE-low cells failed to spread.⁴ To determine whether exosomes isolated from medium conditioned by HPSE-high cells could transfer the spreading phenotype to HPSE-low cells, we placed HPSE-low cells in wells coated with fibronectin and added, in equal amounts, purified exosomes secreted by HPSE-high cells or HPSE-low cells. Exosomes from both cell types enhanced spreading of cells, but the exosomes from HPSE-high cells caused more cells to spread (55% spread cells) than did exosomes from HPSE-low cells (15% spread cells) (Fig. 2B).

To determine whether tumor-derived exosomes could impact the behavior of nontumor cells, we utilized an endothelial cell invasion assay. The addition of exosomes secreted by HPSE-high cells enhanced endothelial cell invasion by 70% as compared with exosomes secreted by HPSE-low cells (Fig. 2C). Thus, exosomes secreted by HPSE-high cells enhance tumor cell spreading and endothelial cell invasion, two cell behaviors key to tumor progression and metastasis.

DISCUSSION

Currently, very little is known regarding the regulation of exosome production and secretion by cells. In the present study, we discovered that the endoglycosidase enzyme heparanase dramatically up-regulates exosome secretion. This was demonstrated by up-regulation of exosome secretion in both a human myeloma cell line and a human lymphoblastoid cell line following their transfection with the cDNA for human heparanase. Exosome secretion was also enhanced following the addition of recombinant heparanase to the myeloma cells, indicating that the effect of heparanase on exosome secretion was not an artifact of the transfection or due to long term overexpression of the enzyme by the tumor cell. In addition, recombinant heparanase stimulated exosome secretion in breast carcinoma cells, thus extending our findings beyond lymphoid cells and indicating that heparanase may impact exosome biology in many tumor types. The fact that exogenous heparanase can increase exosome secretion by a tumor cell is important because it has been demonstrated that cells have receptors for heparanase that facilitate its internalization, and once taken up by the cell, the enzyme can have biological impact (31, 32). Thus, heparanase released by a tumor cell or by host cells (e.g. macrophages) could diffuse within the microenvironment and impact neighboring tumor cells and enhance, among other effects, their secretion of exosomes.

We also found that for robust enhancement of exosome secretion, heparanase enzyme activity is required. This raises the possibility that exosome production/secretion is regulated by specific structural features of heparan sulfate that are exposed upon their cleavage by heparanase. Consistent with this idea is the previous finding that enzymatic clipping of heparan sulfate can expose cryptic sites along the oligosaccharide, thereby influencing its action as either a promoter or an inhibitor of tumor growth and metastasis (38). Another possibility is that heparanase regulates exosome levels by altering the size, amount, or location of heparan sulfate and/or syndecan-1 in a cell. We previously demonstrated that syndecan-1 isolated from the CAG HPSE-high cells used in the present work is smaller in size than the syndecan-1 isolated from the HPSE-low cells, indicative of its heparan sulfate being trimmed by heparanase (23). This smaller form of syndecan-1 is also more rapidly shed from the cell surface, indicating that heparanase also influences syndecan-1 location (23, 26). Moreover, when recombinant heparanase was added to glioma cells, it induced an accumulation of syndecan-1 within endosomes (31). Regulation of syndecan-1 localization by heparanase could stimulate exosome biogenesis because a key step in the formation of some intraluminal vesicles and exosomes is the assembly of a complex consisting of the syndecan-1 cytoplasmic domain, syntenin, and Alix (11). Because syndecan-1 is an integral membrane protein, it is also possible that heparanase-mediated remodeling of heparan sulfate size or structure alters membrane trafficking of proteins in a manner that enhances endosome production and exosome secretion. Within this context, it is important to note that heparan sulfate can influence exosome biogenesis. When heparan sulfate expression was knocked down or when heparan sulfate was extensively degraded with bacterial Hep III, exosome secretion by MCF-7 breast cancer cells (11) and by our CAG-HPSE high cells (Fig. 1D) was significantly reduced. Together these data indicate that syndecan-1 and heparan sulfate play important roles in regulating exosome secretion and that their modulation by the action of heparanase could impact this function.

In addition to the impact of heparanase on increasing the level of exosomes secreted by myeloma cells, we also discovered that the exosomes contained the heparanase enzyme as part of their cargo (Fig. 1B). Interestingly, only the precursor, latent (pro) form of the enzyme was readily detected in exosomes. This was surprising because high levels of the active heparanase enzyme are present in detergent extracts of the HPSE-high cells (30). In fact there is significantly more active form than latent form present in these cells. Thus, there appears to be preferential packing of the latent form of the enzyme within the exosomes. It may be advantageous to transport heparanase in its latent form within exosomes, later to be activated once the enzyme is delivered to a recipient cell. Of note, pro-heparanase has also been detected in exosomes isolated from a pleural effusion of a lung cancer patient,⁵ further emphasizing the clinical significance of our findings in the myeloma cells.

We previously demonstrated that heparanase plays a dynamic role in regulating the growth and progression of myeloma tumors by regulating cross-talk between the tumor and host cells (21). These interactions lead to enhanced angiogenesis, metastasis, and osteolysis, thereby implicating heparanase as a master regulator of the aggressive tumor phenotype in myeloma (21). Until now, it was assumed that heparanase action occurred via secretion of the soluble enzyme into the microenvironment or by heparanase regulation of gene expression intracellularly. However, the current finding that heparanase is present in exosomes raises the possibility that the enzyme is delivered to distal locations via exosomes. Because of the known role of heparanase in promoting angiogenesis and metastasis, delivery of the enzyme via exosomes may play a role in establishing niches to which tumor cells eventually home and grow.

The functional assays that we performed examining tumor cell spreading and endothelial invasion indicate that heparanase can enhance secretion of exosomes that interact with both tumor and host cells and drive them toward behaviors associated with an aggressive tumor phenotype. Because equal amounts of exosomes from HPSE-low and HPSE-high cells were utilized in these functional assays, the difference in their impact on cell behavior is likely due to differences in their composition. We found that the levels of syndecan-1, VEGF, and HGF are all higher in exosomes from HPSE-high cells than in exosomes from HPSE-low cells. To better understand the impact of heparanase on exosome composition, it will now be important to more fully characterize the proteome of these exosomes as well as their miRNA and mRNA composition. It has been demonstrated that a specific miRNA can regulate heparanase expression in breast cancer cells (39). Similarly, it will be important to determine whether heparanase can in turn modulate the miRNA cargo within exosomes. Also, it is important to keep in mind that the functional impact of exosomes from HPSE-high cells might be much greater than what we detected in our in vitro assays. This is because in our assays, equal amounts of exosome protein from HPSE-low and HPSE-high cells were added. However, because heparanase also dramatically enhances exosome secretion, the tumor microenvironment will be bathed in much higher levels of exosomes when tumors are expressing high levels of heparanase. This, coupled with the heparanase-driven changes in exosome composition, likely contributes to the well documented association between elevated heparanase expression and the increased morbidity and mortality of many cancers (40).