Heparanase modulation by Wingless/INT (Wnt)

Heparanase is an endo-beta-glucuronidase, the only enzyme in mammals capable of cleaving heparan sulfate/heparin chains from proteoglycans. The oligosaccharides generated by heparanase present extensive biological functions since such oligosaccharides interact with adhesion molecules, growth factors, angiogenic factors and cytokines, modulating cell proliferation, migration, inflammation, and carcinogenesis. However, the regulation of heparanase activity is not fully understood. It is known that heparanase is synthesized as an inactive 65 kDa isoform and that post-translation processing forms an active 50 kDa enzyme. In the present study, we are interested in investigating whether heparanase is regulated by its own substrate as observed with many other enzymes. Wild-type Chinese hamster (Cricetulus griséus) ovary cells (CHO-K1) were treated with different doses of heparin. Heparanase expression was analyzed by Real-time PCR and flow cytometry. Also, heparanase activity was measured. The heparanase activity assay was performed using a coated plate with biotinylated heparan sulfate. In the present assay, a competitive heparin inhibition scenario was set aside. Exogenous heparin trigged a cell signaling pathway that increased heparanase mRNA and protein levels. The Wnt/beta-catenin pathway, judged by TCF-driven luciferase activity, seems to be involved to enhance heparanase profile during treatment with exogenous heparin. Lithium chloride treatment, an activator of the Wnt/beta-catenin pathway, confirmed such mechanism of transduction in vivo using zebrafish embryos and in vitro using CHO-K1 cells. Taken together the results suggest that heparin modulates heparanase expression by Wnt/beta-catenin.


Introduction
Heparanase-1 or heparanase (HPSE), an endo-beta-glucuronidase, is the only enzyme in mammals capable of cleaving heparan sulfate/heparin chains from proteoglycans. The oligosaccharides generated by heparanase present extensive biological functions since such oligosaccharides interact with adhesion molecules, growth factors, angiogenic factors and cytokines, modulating cell proliferation, migration, inflammation, and carcinogenesis. Furthermore, heparanase is capable of digesting heparan sulfate in the extracellular matrix (ECM), thereby allowing cell invasion [1,2]. Increased expression of heparanase is closely related to poor prognosis in several pathological conditions, such as the process of tumor development, epithelial-mesenchymal transition, inflammation, and angiogenesis [3][4][5][6][7][8][9][10]. Higher levels of heparanase expression and enzymatic activity are frequently used as a marker for diagnosis and prognosis of malignant tumors [7,11].
Several strategies for inhibiting heparanase activity have been developed, including anti-heparanase antibodies, DNA aptamers displaying high affinity and specificity for the active site of the enzyme, modified heparins, and synthetic polysaccharides PI-88 (a mixture of highly sulfated, monophosphorylated mannose oligosaccharides), SST0001 (chemically modified heparin), and M402 (heparan sulfate mimetic). Some of these compounds are undergoing clinical trials with good results, PI-88 (Phase-III studies in hepatocellular carcinoma), PG-545 (Phase-I in patients with advanced solid tumors, SST0001 (Phase-I studies with multiple myeloma) and M-402 (Phase-II studies with pancreatic cancer) [11][12][13]. Weissmann et al. using specific antibodies demonstrated an inhibitory effect of heparanase activity and a significant decrease in the tumor growth of non-Hodgkin's B lymphoma in mice model [14].
Heparanase is synthesized as an inactive 65 kDa isoform, which is cleaved by proteolysis in late endosomes/lysosomes by cathepsin L, releasing an active enzyme, which consists of a heterodimer of 50 kDa and 8 kDa [4,15,16]. Both 50 kDa and 8 kDa subunits form a heterodimer that is 100fold more active than the pro-enzyme (65 kDa) [17]. Thus, the post-translation processing of heparanase regulates its enzymatic activity. Therefore, slight differences in mRNA expression may not be enough to change heparanase activity due to post-translational activation mechanisms [18].
Previous results obtained by our group [19] demonstrated that in CHO-K1, heparanase is located mostly in the ECM and cell surface. However, heparanase is present inside the cell in mutant CHO-745, which has a significantly lower amount of glycosaminoglycans, suggesting that sulfated glycosaminoglycans can modulate heparanase location. In addition, CHO-K1 cell line was treated with 4-methylumbelliferone (7-hydroxy-4-methyl coumarin or 4-MU, Sigma-Aldrich), a potent chemotherapeutic agent that completely inhibits cellular synthesis of glycosaminoglycans, due to inhibition of the enzyme glucuronosyltransferase (UGT) [20][21][22]. Surprisingly, heparanase expression significantly decreased after treatment with 4-MU. Moreover, the heparanase levels were not related to the sulfation of glycosaminoglycans [19].
It is well known that heparin can inhibit the heparanase enzyme by competing with the substrate heparan sulfate [23]. Also, it has already been demonstrated that heparanase expression can be stimulated by Early Growth Response 1 (EGR1), estrogen and p53 [24,25]. Breast cancer patients treated with tamoxifen, an estrogen receptor inhibitor, have decreased heparanase expression in the circulating lymphocytes [26]. In addition, cathepsin L activity is also regulated by glycosaminoglycans [18,[27][28][29]. Despite all our knowledge about heparanase, the regulation of enzymatic activity is not fully understood. Therefore, we decided to investigate the mechanisms of heparanase modulation by exogenous heparin.
In the present study, we are interested in investigating whether heparanase levels are regulated by its own substrate as observed with many other enzymes. Also, we evaluated if Wnt/beta-catenin might be involved in such regulation. We have used techniques that allow us to suggest the involvement of the Wnt signaling in our model. The Wnt/betacatenin pathway acts to control the transcription of genes through the binding of a complex of beta-catenin and Transcription Factor (TCF) to specific promoter elements. The activity of this final step in the Wnt/beta-catenin pathway can be measured using a luciferase reporter construct [30]. Thus, CHO-K1 cells were transiently transfected with the pTOPFLASH or pFOPFLASH (containing a 'far from optimal' TCF binding site). In addition, we also investigated the response of cells to lithium chloride (LiCl), which is a known inhibitor of Glycogen Synthase Kinase 3 (GSK-3) phosphorylation, thus stimulating the canonical Wnt signaling pathway. We believe that these combined results allowed us to evidence that heparin can modulate the expression of heparanase, suggesting that the Wnt pathway is involved in such modulation.

Treatment with heparin
Heparin (porcine intestinal mucosa, Sigma-Aldrich, USA) was added to the culture medium of CHO-K1 (100 µg/mL, 18 U/mL and 500 µg/mL, 90 U/mL) and incubated for 18 h at 37 °C. This CHO-K1 treatment was performed to evaluate the effect of exogenous heparin on heparanase expression. These dosages were used since some studies in the literature found competition with endogenous heparan sulfate using this amount range [31,32].

Enzymatic activity of heparanase using biotinylated heparan sulfate
The cellular fraction used to perform the enzymatic assay was scraped from culture plates with sodium acetate buffer, pH 5.5, and protease inhibitor cocktail (Sigma-Aldrich). Total cell protein was estimated by Pierce™ Coomassie Plus (Bradford) Assay Kit (Thermo Fisher Scientific), as described by the manufacturer. It is important to point out that the values that express heparanase activity were adjusted by the total protein of cellular extract, to eliminate bias due to some possible anti-proliferative or pro-proliferative effect. Enzymatic activity of heparanase was measured using 1 µg heparan sulfate 15% biotinylated (bovine kidney, Sigma-Aldrich) immobilized on a 96-well plate previous coated with protamine sulfate (Sigma-Aldrich), following the instructions described by Bouças and coworkers [33]. Subsequently, the cellular fraction (15 µg of total protein) was added to the plate containing immobilized biotinylated heparan sulfate and maintained for 4 h at 37 °C, followed by streptavidin-europium incubation (Perkin-Elmer, Finland), for 40 min, at room temperature. Finally, a Delfia® enhancement solution (Perkin-Elmer) was added and fluorescence was quantified following the protocol described by Melo et al. [34].

Wnt activity
Luciferase reporter assay was used to evaluate Wnt activation. The cells were transiently transfected using Fugene with the TCF firefly luciferase reporters (pTOPFLASH, TCF Optimal motif) and pFOPFLASH (containing a 'Far from Optimal' TCF binding site), as described by van de Wetering et al. [30]. pTOPFLASH and pFOPFLASH were gifts from Prof. Maikel Peppelenbosch (Erasmus University Medical Center, Rotterdam, The Netherlands). After specific treatments, the cells were lysed and luciferase activity was measured using the Steadylite Plus reporter gene assay (Perkin-Elmer, The Netherlands), according to the manufacturer's instructions. The results were calculated by the ratio of TOP/FOP to rule out some bias due to alteration in cell death or proliferation after treatments. The result was expressed relative to the control group.

LiCl treatment
To evaluate the effect of Wnt signaling on heparanase expression in CHO-K1 lithium chloride (molecular biology grade 99% was purchased from Sigma-Aldrich, catalog number L9650) (activator of Wnt signaling through inhibition of GSK-3) was added to the culture medium of CHO-K1 cells (1 mM) and incubated for 6 h at 37 °C, followed by flow cytometry analysis. Also, zebrafish with 3 days postfertilization were treated with LiCl (2 mM) in water containing 1% DMSO for 6 h, 28 °C. Fifty embryos per group were used to RNA extraction. Even though it is recognized that LiCl can inhibit other targets, its major effect has long been demonstrated to be GSK-3 inhibition in diverse organisms and in vitro [35][36][37].

Statistics
Statistical analysis was performed using the SPSS® 13 program (SPSS® Inc; Illinois, USA). The variables were tested using the Kolmogorov-Smirnov test. The Mann-Whitney test was used to determine the relation 1 3 between non-parametric variables. A value of P < 0.05 was considered statistically significant. To non-parametric was used Kruskal-Wallis test and to parametric was used ANOVA.

Effect of exogenous heparin on the expression of heparanase
Heparin upregulated heparanase mRNA expression. Figure 1a shows an increase of 250-fold on heparanase mRNA expression with 500 µg/mL of heparin compared to 100 µg/ mL. Moreover, heparin compared to 100 µg/mL treatment produced an increase of 4 times over the control. Figure 1b shows that exogenous heparin also upregulates heparanase activity in the CHO-K1 cell line. The enzymatic activity of heparanase was enhanced 2 times with 500 µg/ mL of heparin over 100 µg/mL of heparin, while an increase of 1.5 was observed between 100 µg/mL of heparin and the control (Fig. 1b). It is important to point out that heparin is a competitive inhibitor of heparanase. However, after heparin treatment, the culture medium was removed, and the cells harvested to determine the enzymatic activity using biotinylated heparan sulfate. Therefore, a competitive heparin inhibition scenario was set aside in the present assay.
These results suggest that the effect of heparin on mRNA expression and heparanase protein levels is concentration dependent.

Mechanisms involved in heparanase modulation by heparin
The Wnt/beta-catenin signaling controls the transcription of genes through the binding of a complex of beta-catenin and TCF to specific promoter elements. The activity of this final step in the Wnt/beta-catenin cascade can be investigated using a luciferase reporter construct. Thus, CHO-K1 cells were transiently transfected with the pTOPFLASH or pFOPFLASH reporter [30].
Heparin increased beta-catenin/TCF signaling as shown in Fig. 2. There was a 1.5-fold increase in luciferase activity after incubation with different concentrations of heparin compared to control (Fig. 2). The stimulation of Wnt activity by heparin was similar to LiCl treatment. An increase in Wnt activity corroborates with a decreased phosphorylation of beta-catenin as shown in Fig. 3.
To confirm that the Wnt/beta-catenin pathway is involved in the activation of heparanase expression, CHO-K1 cells were treated with LiCl, as shown in Fig. 4.
In addition, zebrafish embryos were treated with LiCl, the heparanase expression increased comparing to control group (Fig. 5). Fig. 1 Expression of heparanase. a Heparanase mRNA expression in CHO-K1 was investigated by quantitative Real Time-PCR. The result was expressed by 2 −ΔΔCt compared to the control. b After treatment, the culture medium was removed, and the cellular extract was harvested. The samples obtained from cellular fraction were used to heparanase activity assay with biotinylated heparan sulfate as described in Methods. The result was expressed by the percentage of hepara-nase activity comparing with the control. Control, CHO-K1 treated with 1% DMSO; Heparin 100 µg/mL, CHO-K1 was treated with 100 µg/mL of heparin; Heparin 500 µg/mL, CHO-K1 was treated with 500 µg/mL of heparin. Bars means the average and lines represent standard deviations, *p < 0.05, Kruskal-Wallis test. Heparin stimulated mRNA expression and the protein levels of heparanase Even though it is recognized that LiCl can inhibit other targets, its major effect has long been demonstrated to be GSK-3 inhibition in diverse organisms and in vitro [36]. In addition, LiCl can modulate GSK-3/phosphorylation on Ser9/21 by a complex mechanism. In this respect, studies have shown that inhibition of GSK-3 by lithium leads to increased N-terminal phosphorylation of GSK-3 and GSKbeta, demonstrating autoregulation of GSK-3N-terminal serine phosphorylation. Thus, GSK-3 autoregulation could involve inhibition of a kinase or activation of a phosphatase. Fig. 2 Wnt activity reporter assay. Wnt/beta-catenin pathway acts to control the transcription of genes through the binding of a complex of beta-catenin and TCF to specific promoter elements. The activity of this final step in the Wnt/beta-catenin pathway can be measured using a luciferase reporter construct. Thus, CHO-K1 cells were transiently transfected with the pTOPFLASH (TCF Optimal Motif) or pFOPFLASH (containing a 'far from optimal' TCF binding site). Control, CHO-K1 without treatment; DMSO, CHO-K1 treated with 1% DMSO; Hep 100, CHO-K1 was treated with 100 µg/mL of hepa-rin; Hep 500; CHO-K1 was treated with 500 µg/mL of heparin; LiCl, CHO-K1 was treated with lithium chloride (1 mM). The result was expressed as relative quantification of TOP/FOP ratio compared with the control group. TOP/FOP Flash ratio remove some bias due to proliferation or cell death. The assay was performed in triplicate. Bars means the average and lines represent standard deviations, *p < 0.05, Kruskal-Wallis test. Heparin increased Wnt signaling at a similar level of LiCl, a well-known activator of Wnt signaling Fig. 3 Phospho beta-catenin analysis. a Flow cytometry analysis was performed using phospho-beta-catenin antibody. Control (yellow), CHO-K1 without treatment; Heparin 100 µg/mL (blue), CHO-K1 was treated with 100 µg/mL of heparin; Heparin 500 µg/mL (red), CHO-K1 was treated with 500 µg/mL of heparin. Blank (arrow); CHO-K1 cells without primary antibody. b The numbers represent median values obtained by flow cytometry analysis (FlowJo Software). Phospho-beta-catenin increased after treatment, therefore the Wnt pathway was stimulated by heparin. (Color figure online) In the response to lithium chloride, this autoregulation implies two levels of inhibition: rapid response, direct inhibition of GSK-3 by lithium, followed by inactivation of the protein phosphatase (PP1). As a consequence of PP1 inactivation there is an increase in inhibitory phosphorylation of GSK-3 [37]. Moreover, GSK-3 function in the Wnt pathway appears to be insulated from the effects of inhibitory N-terminal kinases such as Akt. It has been suggested that this inhibitory phosphorylation of GSK-3 would primarily affect Wnt-independent GSK3-regulated pathways, indicating an unexpected level of signaling pathway selectivity to lithium response.
There was a shift in the curves of heparanase fluorescence as shown in Flow Cytometer analysis, median 213 (control), 273 (treatment with heparin), 385 (treatment with LiCl), suggesting an increase in heparanase protein expression compared to CHO-K1 cells without any treatment (Control), as shown in Fig. 4.
In addition, after treatment with LiCl, zebrafish embryos presented an enhance of heparanase expression, confirming that the same mechanism occurs in vivo, as demonstrated in Fig. 5.
The mechanisms that modulate heparanase activity and heparanase expression are still unclear. It is important to emphasize that heparin is a known inhibitor of heparanase activity since it competes with heparan sulfate for the active site of the enzyme [23]. Nevertheless, in the present study the treatment of CHO-K1 cells with heparin induced an increase in mRNA expression and protein level.

Discussion
Piva and co-workers showed that the depletion of glycosaminoglycan biosynthesis using 4-MU promoted a decrease of heparanase levels, indicating that glycosaminoglycans modulate heparanase since 4-MU inhibits glycosaminoglycans synthesis [19]. Also, the Wnt pathway and beta-catenin play a crucial role in several cellular processes, such as survival, migration, proliferation, and differentiation. It is conceivable that heparan sulfate proteoglycans act as co-receptors for a variety of ligands that regulate cell signaling. It is well known that heparan sulfate has high-affinity binding to Wnt [38]. Furthermore, exogenous heparanase has been shown to modulate cellular responses to Wnt3a [39]. In another study, Famá et al. observed that higher levels of heparan sulfate increased heparanase expression and heparanase activity in placenta tissue from patients with preeclampsia [40], supporting the hypothesis that higher levels of heparan sulfate could perhaps play a role in the positive regulation of heparanase activity.
It has been previously reported that exogenous heparanase significantly alters Wnt3a [39]. However, the modulation of heparanase by Wnt/beta-catenin signaling has not been studied so far. Thus, we set out to investigate whether the Wnt/beta-catenin pathway was associated with heparanase expression.
Another study showed heparin significantly increased Wnt/beta-catenin activity, concomitantly with a reduction in beta-catenin phosphorylation, corroborating the data obtained in the present study [41]. Also, Alexander et al. demonstrated that syndecan-1 and the soluble ectodomain of syndecan-1 were capable of increasing Wnt signaling [42]. In addition, Malinauskas et al. demonstrated that the enhancement of Wnt signaling by the heparin seems to be related to the inhibition of Wnt inhibitory factor 1 (WIF-1) [43].
Moreover, in the presence of Wnt activator (LiCl), there was an enhancement of heparanase protein levels, confirming the modulation of heparanase by this pathway.
Taken together, these results appear to show that heparan sulfate/heparin may possibly regulate heparanase transcription modulated by Wnt signaling pathway.
Although the role of Wnt/beta-catenin signaling in modulating heparanase expression needs further investigation, our study does provide insights into the mechanisms by which heparan sulfate/heparin can regulate heparanase levels.
There are several published articles that show that there is a relationship between heparanase and the modulation of Wnt. However, in such assays, recombinant heparanase has been shown to negatively modulate Wnt, possibly by degradation of the heparan sulfate chains of proteoglycan syndecan-1, which corroborates with our findings since heparin, which is a heparan sulfate-like molecule, stimulates the Wnt pathway [39].
Furthermore, it is well-known that heparan sulfate can modulate tumor development through the Wnt signaling and as mentioned previously, heparin is a heparan sulfate-like [38,45].
The present results demonstrated that heparin stimulates heparanase expression by modulating Wnt activity. Since heparanase is an enzyme directly related to carcinogenesis, our findings add important information to the current body of knowledge on cancer biology.

Conclusion
The results suggest that treatment with exogenous heparin enhanced mRNA expression and protein levels of heparanase. Furthermore, the Wnt/beta-catenin signaling pathway may be involved in the molecular mechanisms of heparanase modulation in vitro and in vivo.