|Home | About | Journals | Submit | Contact Us | Français|
Heparan sulfate (HS) mediates the activity of various growth factors including TGF‐β. Heparanase is an endo‐glucuronidase that specifically cleaves and modifies HS structure. In this study, we examined the effect of heparanase expression on TGF‐β1‐dependent signaling activities. We found that overexpression of heparanase in human tumor cells (i.e., Fadu pharyngeal carcinoma, MCF7 breast carcinoma) attenuated TGF‐β1‐stimulated Smad phosphorylation and led to a slower cell proliferation. TGF‐β1‐stimulated Akt and Erk phosphorylation was also affected in the heparanase overexpression cells. This effect involved the enzymatic activity of heparanase, as overexpression of mutant inactive heparanase did not affect TGF‐β1 signaling activity. Analysis of HS isolated from Fadu cells revealed an increase in sulfation of the HS that had a rapid turnover in cells overexpressing heparanase. It appears that the structural alterations of HS affect the ability of TGF‐β1 to signal via its receptors and elicit a growth response. Given that heparanase expression promotes tumor growth in most cancers, this finding highlights a crosstalk between heparanase, HS, and TGF‐β1 function in tumorigenesis.
The complex and heterogeneous heparan sulfate proteoglycans (HSPG) are ubiquitous macromolecules expressed on the cell membrane and in the extracellular matrix, having essential roles in development and homeostasis 1, 2, 3. One of the important functions of HSPG is to mediate growth factor‐stimulated cell signaling, through interaction of the heparan sulfate (HS) side chains with growth factors and their receptors 4. The biological activity of several growth factors (e.g., FGF, PDGF, VEGF) involves dual cell surface receptor systems consisting of a tyrosine kinase‐type receptor along with a HSPG coreceptor 5. The interaction of HS with growth factors is dependent on its molecular structure that is generated through a strictly regulated biosynthesis process and postsynthesis modifications. Accumulated evidence shows that HS has enormous structural diversity, expressed in a tissue/cell‐specific manner, enabling the interaction of HS with a wide spectrum of protein ligands 6.
Heparanase is an endo‐glucuronidase that modifies HS structure through cleavage of the long HS polysaccharide chains to shorter fragments. This unique mammalian enzyme is expressed at essentially nondetectable amounts in normal tissues, but is elevated in a number of pathological conditions such as cancer and inflammation 7, 8, indicating that the enzyme has important functions in pathophysiology. Our earlier studies revealed that overexpression of heparanase in mice not only led to production of fragmented HS chains but also altered HS structure 9, 10. Increased sulfation of HS in cells overexpressing heparanase promotes FGF2 binding to its receptor and formation of a ternary complex 9.
Involvement of HS in TGF‐β‐induced signaling has been reported 11, 12, 13, 14, 15; however, information regarding the HS molecular structure in TGF‐β‐stimulated cellular activity is lacking. Heparanase was found to regulate TGF‐β expression and activity in renal fibrosis, proposing a role of heparanase in the axis of HS structure and TGF‐β activity 16. In the present study, we found reduced phosphorylation of Smad, Akt, and Erk in response to TGF‐β1 stimulation of cells overexpressing heparanase. This effect is apparently not a direct function of the overexpressed heparanase protein, but is mediated by modification of the HS structure expressed in the cells. The data provide evidence displaying that increased sulfation degree in HS is not favored by TGF‐β1, highlighting, for the first time, a crosstalk between heparanase, HS, and TGF‐β1 signaling in cancer cells.
Antibodies against P‐Smad2 (cat#3101) and total Smad 2/3 (cat#3102), p‐Akt (cat#9271S) and Akt (cat# 9272), P‐Erk (cat#9101) and total Erk (cat# 9107) were purchased from Cell Signaling Technology® (Danvers, MA, USA); β‐actin antibody (Sc‐69879) was from Santa‐Cruz Biotechnology (Dallas, TX, USA); Recombinant human TGF‐β1(cat#100‐21) was from PeproTech (Rocky Hill, NJ, USA). Anti heparanase antibody (1453) has been described 17. The cell lines used are Fadu (human pharyngeal carcinoma), MCF7 (human breast carcinoma), and CHO (Chinese hamster ovary), described previously 18, 19, 20. The cells were either stably (Fadu and MCF7) or transiently (CHO‐K1) overexpressing human heparanase. CHO‐K1 cells stably overexpressing double mutant enzymatically inactive heparanase were described earlier 20, 21. Cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% FBS and antibiotics. Cells were passed in culture no more than 2 months after being thawed from authentic stocks.
Cells were seeded into six‐well plates at a density of 3–6 × 105 cells per well in 2 mL of DMEM supplemented with 10% FBS. After 24 h, the medium was replaced by starvation medium (DMEM without FBS) for 24 h. Then the cells were changed to fresh starvation medium containing TGF‐β1. Following stimulation for 30–60 min, medium was removed and cells were washed twice with PBS before lysis in 100 μL of RIPA buffer (50 mm Tris pH 7.5, 150 mm NaCl, 1%Triton X‐100, 1% Na‐deoxycholate, 1 mm EDTA and 0.1% SDS, Protease inhibitor, 1 mm NaF, 1 mm Na3VO4). The lysate was kept on ice for 30 min followed by ultrasonication for 3 min and centrifugation for 10 min at 16 000 g. The supernatants were collected and protein concentration was determined (bicinchoninic acid assay). Samples of 20 μg of total protein were separated by electrophoresis on SDS/PAGE (10%) and electroblotted onto a Nitrocellulose membrane. The membrane was probed with antibodies and the signals were developed using Super Signal West Duration Substrate (Thermo Scientific, Waltham, MA, USA) and Bio‐Rad CCD camera. The results were analyzed by image lab™ Software (Bio‐Rad, Hercules, CA, USA). The intensity of each band was normalized with that of β‐actin or total Smad, total Akt or total Erk. The average of relative intensity is a mean of two to three blots from independent cell experiments. The intensity of Mock samples without TGF‐β1 stimulation is defined as 1.
To reduce the high endogenous signaling of Erk in Fadu cells, a TGF‐β1 inhibitor (GW6604) was included in the starvation medium (final concentration of 3 μm).
Fadu cells (Mock and Hpa) were cultured to 95% confluence and Na35SO4 (specific activity 1500 Ci·mmol−1, Perkin Elmer, Waltham, MA, USA) was added to the culture (100 μCi·mL−1) for 24 h before harvesting. HSPG was purified from both medium and cell fractions essentially as described 22. Briefly, the cells were lysed in buffer containing 4M Urea, 1% Triton X‐100, 50 mm Tris‐HCI, pH 7.4, 0.25 m NaCl and centrifuged. The cell lysates and medium were applied to DEAE‐Sephacel columns (GE Healthcare Biosciences, Uppsala, Sweden) pre‐equilibrated with 50 mm Tris‐HCl, 0.25 m NaCl, pH 7.4. The columns were washed with 50 mm NaAc, 0.25 m NaCl, pH 4.5 and proteoglycans were eluted with 50 mm NaAc containing 2 m NaCl, pH 4.5. The eluted material was desalted on a PD‐10 column (GE Healthcare Biosciences), followed by lyophilization to dryness. The samples were treated with Chondroitinase ABC (0.1 U per sample, Seikagaku, Tokyo, Japan) to degrade chondroitin sulfate (CS). The purified HSPG was incubated in 0.5 m NaOH on ice, to obtain free HS chains. The samples (10 000 cpm) were analyzed on a Superose 12 column (GE Healthcare Biosciences) to examine chain length or on a DEAE‐Sepharcel column (1‐mL) to assess charge density. The effluent fractions from the chromatographic separation were mixed with scintillation cocktail and counted in a scintillation counter.
Fadu cells were cultured in the flasks (T‐75) to 90% confluency and then changed to starvation medium for 24 h in serum‐free medium. Then, the cells were collected and seeded in 96‐well plates (10 000 cell per well) in the starvation medium in the presence of TGF‐β1 at the concentrations indicated for 24 h. Cell proliferation was determined applying the MTS assay (Promega) by measuring absorbance at 490 nm using TECAN Plate reader. The experiments were repeated two times and the results are expressed as the mean ± SE of two independent experiments (total of 10 wells).
Previous studies elucidated the involvement of heparanase in signal transduction and response to growth‐promoting factors 19, 23, 24, 25, 26. To find out the effect of heparanase expression on TGF‐β‐induced signaling, we examined phosphorylation of Smad in Fadu cells that stably overexpress human heparanase (Fig. (Fig.1A,1A, lower panel) in comparison to mock‐transfected Fadu cells. Smad2 phosphorylation was examined by western blot analysis using an antibody that recognizes the C‐terminal diphosphorylated serine motif of Smad2 (Fig. (Fig.1A,1A, upper panel). The results show that stimulation with TGF‐β1 led to phosphorylation of Smad2 in Fadu cells essentially in a dose‐dependent manner up to a concentration of 5 ng·mL−1 TGF‐β1 (Fig. (Fig.1B).1B). Notably, cells overexpressing heparanase (Hpa) displayed a lower degree of Smad2 phosphorylation (Fig. (Fig.1A,B)1A,B) as compared to mock‐transfected cells. Neither did TGF‐β1 stimulation affect expression of heparanase.
To verify whether this effect of heparanase is specific to Fadu cells, we examined additional cell lines. MCF‐7 cells showed essentially no response to TGF‐β1 at lower concentrations, irrespective of heparanase overexpression (Fig. (Fig.2).2). A higher dose of TGF‐β1 (5 ng·mL−1) stimulated Smad2 phosphorylation in both Mock and Hpa cells. Again, the Hpa cells displayed a lower degree of Smad2 phosphorylation in comparison to Mock cells. In CHO cells, the same trend was observed in heparanase high (Hpa) cells, however, the reduction in phosphorylated Smad2 was not as strong as in the human cell models (Fig. S1).
To find out whether the TGF‐β1‐induced Smad phosphorylation has impact on cellular activities, we examined cell proliferation by the MTS assay. In agreement with the TGF‐β1‐stimulated phosphorylation of Smad2 (Fig. (Fig.1B),1B), prolonged incubation (24 h) with TGF‐β1 stimulated the proliferation of Mock cells, but had no effect on Hpa cells (Fig. (Fig.33).
To verify if heparanase expression has a global effect on TGF‐β1‐induced signaling activity, we checked Akt and Erk signaling in the cells and found the same effect of heparanase on the TGF‐β1‐induced phosphorylation of Akt and Erk (Fig. (Fig.4).4). It should be noted that phosphorylation of Akt and Erk in Fadu cells is sensitive to the concentration of TGF‐β1, where 5 ng·mL−1 of the cytokine showed less activity than the concentrations of 0.5 and 1 ng·mL−1.
To investigate whether the reduced TGF‐β1‐dependent phosphorylation in heparanase overexpressing cells is associated with the fine structure of HS, Fadu cells were cultured in the presence of 35S, and metabolically labeled HSPG/HS was purified. Analysis on a gel filtration chromatographic column (Superose 12) showed a marginally reduced overall molecular size of HSPG (Fig. (Fig.5A)5A) isolated from the conditioned medium of heparanase overexpressing cells (Hpa), accompanied by elevation in the amount of smaller fragments. In the cell‐derived fractions, although the overall size of HSPG in Mock vs. Hpa cells was unchanged (Fig. (Fig.5B),5B), the Hpa cells also displayed an accumulation of smaller fragments. This indicates an increased fragmentation of HS by overexpressed heparanase. Analysis of the free HS chains released from HSPG by alkali treatment revealed, again, accumulation of smaller fragments in both the medium (Fig. (Fig.5C)5C) and cell (Fig. (Fig.5D)5D) fractions. Quantification of the total 35S (cpm) in HS and CS fractions from the peaks showed that the ratio of HS/CS is 4.4 in the conditioned medium of Mock vs. 14.6 in the conditioned medium of Hpa‐tg Fadu cells, indicating a more than threefold increase in shedding of HS. At the same time, the HS/CS ratio was 5.4 in Mock vs. 2.7 in the Hpa high cells, indicating an increased turnover of HS in the heparanase overexpression cells (Table 1). A similar pattern of the ratio between HSPG/CSPG confirms the effect of heparanase on the intact proteoglycans.
The free HS chains were further subjected to analysis on a DEAE‐Sepharcel column to evaluate the overall charge density. The degraded CS disaccharides were eluted at low salt concentrations. HS from heparanase overexpressing cells and their culture medium showed a significant retardation in the anion exchange column in comparison to the samples isolated from Mock cells, indicating a higher overall sulfation of HS from the heparanase‐overexpressing cells (Fig. (Fig.66).
Given that the HS structure is altered, the effect of heparanase expression on TGF‐β1‐dependent phosphorylation is presumable due to enzymatic modification of HS in the Hpa cells. To verify this, we examined Smad2 phosphorylation in CHO cell stably overexpressing mutant (Glu225, Glu343) heparanase that lacks catalytic activity 19, 21. Western blot analysis revealed that overexpression of mutant heparanase had essentially no effect on TGF‐β1‐simulated Smad2 C‐terminal phosphorylation (Fig. (Fig.7),7), confirming the notion that HS molecular structure plays pivotal roles in TGF‐β1 signaling activity.
Cell surface HS has important functional roles in formation of cytokine gradients 2 and in mediating the activities of different protein ligands including growth factors 6. The role of HS in growth factor signaling has been most studied in relation to the FGF family. In comparison, information regarding HS function in TGF‐β activity is limited. Nonetheless, several recent studies have shown that HS is involved in TGF‐β functions 11, 12, 13, 14, 15, 27. Likewise, heparanase has been shown to play a key role in renal fibrosis by regulating TGF‐β expression and activity 16. It also promotes bronchiolitis and lung fibrosis by enhancing the release and activation of ECM‐stored TGF‐β through the cleavage of HS 28. Since TGF‐β is a pluripotent cytokine that can promote or suppress cancer progression and metastasis 29, it is important to elucidate the involvement of heparanase and HS in TGF‐β‐induced signaling and cellular effects.
The HS endo‐glucuronidase (heparanase) is overexpressed in most tumor tissues, correlating with tumor vascular density and metastatic potential, and with patient survival 30. Our earlier study has found that overexpression of heparanase in mice led to production of highly sulfated HS that displayed a higher potency in assembling FGF2‐FGFR ternary structure 9. To investigate whether heparanase expression has a similar effect on TGF‐β1 activity, we examined TGF‐β1‐stimulated signaling pathways in human cancer cells that stably overexpress heparanase. Unexpectedly, we found that heparanase overexpression attenuated TGF‐β1‐induced Smad2, Akt, and Erk phosphorylation in heparanase‐overexpressing tumor cells, contradictory to its effect on FGF activity 9, 24. However, our results are in line with the finding that transgenic overexpression of an endo‐sulfatase (Sulf1) promoted activation of the TGF‐β1/Smad pathway 13. This mutual feature can be plausibly ascribed to the sulfation status of HS in the cells. Sulf1 specifically removes 6‐O‐sulfate in a given HS chain. Consequently, overexpression of sulf1 will lead to reduced HS sulfation and thereby higher TGF‐β1 activity 13. In contrast, overexpression of heparanase leads to production of highly sulfated HS (Fig. (Fig.5)5) 9, and hence attenuated TGF‐β1 activity.
Unlike many growth factors TGF‐β is a bifunctional regulator 31 that has both stimulatory and inhibitory activity in the same cells 32. This effect of TGF‐β is believed to be dependent on the molecular context of the cells. This feature of TGF‐β is evidenced in our experiments. Stimulation at lower concentration of TGF‐β1 increased phosphorylation of Akt and Erk in Fadu cells, but at higher concentration suppressed phosphorylation. Similarly, the cell proliferation exhibited a bell‐shape in response to the stimulation of TGF‐β1 (Figs.(Figs.3,3, ,4).4). Our results, together with the findings with Sulf1‐overexpressing cells 13, suggest that the molecular structure of HS on the cell surface plays a role in the pluripotent activity of TGF‐β1. Notably, HS biosynthesis and postmodification are spatially and temporally regulated by diverse factors 3, offering an explanation for the apparently contradictory activity of TGF‐β1 on the same cell type cultured under different conditions 32.
Two distinct types of FGF‐HS‐FGFR complexes were discerned; one structure showing two FGF–FGFR pairs that interact in a symmetrical mode with two oligosaccharides (2:2:2 complex) 33. The other model involves a single HS that interacts with two FGFs but only one of the two FGFR molecules (2 : 2 : 1 complex) 34. In comparison, the molecular feature of TGF‐β/HS/TGF‐β receptor interaction is largely unveiled. A model has been proposed, showing HS modulation of TGF‐β1 receptor complex formation 11. The absence of HS promotes formation of receptor Complex 1 that is constituted mostly of type II rather than type I TGF‐β1 receptor, leading to signal transduction. Thus, the attenuated TGF‐β1 activity found in the heparanase‐overexpressing cells may involve two independent or associated mechanisms; a highly charged HS on the cell surface traps TGF‐β1 preventing its interaction with receptors; or/and the increased sulfation of HS interferes with the formation of receptor Complex 1, leading to the degradation pathway 11. An additional parameter is that heparanase overexpression increases the turnover of HS, which may abrogate the stability of HS interaction with TGF‐β1, and potentially with its receptors.
In conclusion, our finding that TGF‐β1 activity is associated with the molecular structure of HS has conveyed a novel notion for the functions of TGF‐β family members. Given that heparanase expression is elevated in the majority of tumor tissues, and that HS isolated from tumor cells and tissues exhibit a higher sulfation content 9, it is of importance to establish the functions of heparanase, as well as HS, on TGF‐β1 activity and the related effects on tumor growth.
TB and JPF designed and performed the experiments, wrote the manuscript. UB performed experiment IV, AM and JPL designed the study, analyzed the data, and wrote the manuscript.
Fig. S1. CHO‐K1 cells were seeded at a density of 3 × 105 cells per well of six‐well plate in 3 mL of F12K supplemented with 10% FBS and cultured for 24 h.
The study is supported by grants from the Swedish Research Council (2015‐02595; K2013‐66X‐14936‐10‐5), Swedish Cancer Foundation (150815; 150834) awarded to JPL, and AM, and by grants from the Israel Science Foundation (grant 601/14) and the Israel Cancer Research Fund (ICRF) awarded to IV. TB is a PhD student supported by Erasmus Mundus, Experts for Asia program. IV is a Research Professor of the ICRF.