Identification of heparin affin regulatory peptide domains with potential role on angiogenesis

Identification of heparin affin regulatory peptide domains with potential role on angiogenesis

The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966 Identification of heparin affin regulatory peptide domains with potentia...

328KB Sizes 0 Downloads 5 Views

The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

Identification of heparin affin regulatory peptide domains with potential role on angiogenesis Apostolos Polykratis a,b , Jean Delbé c , José Courty c , Evangelia Papadimitriou a,∗ , Panagiotis Katsoris b a

Laboratory of Molecular Pharmacology, Department of Pharmacy, University of Patras, Patras, GR 26504 Greece b Laboratory of Cell Biology, Department of Biology, University of Patras, Patras, Greece c Laboratoire de Recherche sur la Croissance Cellulaire, la Réparation et la Régénération Tissulaire (CRRET), CNRS FRE 2412, Université Paris XII Créteil, France Received 9 May 2003; received in revised form 12 February 2004; accepted 13 February 2004

Abstract Heparin affin regulatory peptide (HARP) is a growth factor displaying high affinity for heparin. It is present in the extracellular matrix of many tissues, interacting with heparan sulfate and dermatan/chondroitin sulfate glycosaminoglycans. We have previously shown that HARP is implicated in the control of angiogenesis and its effects are mimicked, at least in part, by synthetic peptides that correspond to its N and C termini. In the present work, we show that HARP is cleaved by plasmin, leading to the production of five peptides that correspond to distinct domains of the molecule. Heparin, heparan sulfate and dermatan sulfate, at various HARP to glycosaminoglycan ratios, partially protect HARP from plasmin degradation. The molecules with higher affinity to HARP are the more protective, heparin being the most efficient. The peptides that are produced from cleavage of HARP by plasmin, affect in vivo and in vitro angiogenesis and modulate the angiogenic activity of vascular endothelial growth factor on human umbilical vein endothelial cells. Similar results were obtained in vitro with recombinant HARP peptides, identical to the peptides generated after treatment of HARP with plasmin. These results suggest that different regions of HARP may induce or inhibit angiogenesis. © 2004 Elsevier Ltd. All rights reserved. Keywords: HARP; Pleiotrophin; Plasmin; Angiogenesis; Endothelial cells

1. Introduction Heparin affin regulatory peptide (HARP) (Courty, Dauchel, Caruelle, Perderiset, & Barritault, 1991), also known as pleiotrophin (Li et al., 1990) or heparin-binding growth-associated molecule (Rauvala, 1989), is an 18 kDa growth factor that has a high ∗ Corresponding author. Tel.: +30-2610-969336; fax: +30-2610-997665. E-mail address: [email protected] (E. Papadimitriou).

affinity for heparin. It constitutes with midkine and retinoic acid heparin-binding protein, a family of structurally related heparin-binding growth factors, with an overall 50% amino acid identity (Courty, Milhiet, Delbe, Caruelle, & Barritault, 2000). HARP has been purified from perinatal rat brain as a molecule that induces neurite outgrowth, suggesting that it is involved in the maturation of neuronal cells (Rauvala, 1989). Later studies have shown that it is also present in non-neuronal tissues, including uterus (Milner et al., 1989), heart (Hampton, Marshak,

1357-2725/$ – see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2004.02.012

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

& Burgess, 1992), cartilage (Neame et al., 1993) and bone extracts (Gieffers, Engelhardt, Brenzel, Matsuishi, & Frey, 1993), suggesting that the function of HARP is not restricted to the neurite outgrowth promoting activity. Studies on the in vitro biological properties of HARP have been performed by several laboratories and despite the existence of controversial results, a growing body of evidence indicates that HARP is involved in the control of cellular proliferation (Papadimitriou et al., 2000; Souttou, Raulais, & Vigny, 2001), migration (Maeda & Noda, 1998; Papadimitriou et al., 2001; Souttou et al., 2001) and differentiation (Papadimitriou et al., 2001; Raulo, Julkunen, Merenmies, Pihlaskari, & Rauvala, 1992; Szabat & Rauvala, 1996). Moreover, there is a strong correlation between HARP expression and tumor growth and angiogenesis (Chauhan, Li, & Deuel, 1993; Czubayko, Riegel, & Wellstein, 1994; Zhang, Zhong, Wang, & Deuel, 1997). The biological activities of HARP can be attributed to interactions with cell surface proteoglycans or binding to cell surface receptors. N-syndecan (Raulo, Chernousov, Carey, Nolo, & Rauvala, 1994), receptor-type protein tyrosine–phosphatase ␤/␨ (RPTP␤/␨) and its secreted variant phosphacan (Maeda, Nishiwaki, Shintani, Hamanaka, & Noda, 1996; Meng et al., 2000), as well as anaplastic lymphoma kinase (ALK) (Bernard-Pierrot et al., 2001; Bernard-Pierrot et al., 2002; Stoica et al., 2001), have been reported to bind HARP and to be implicated in its signaling. Moreover, HARP has a well-established affinity for heparin (Hampton et al., 1992; Rauvala, 1989), heparan sulfate (HS) and dermatan sulfate (DS) proteoglycans (Vacherot et al., 1999). Glycosaminoglycans (GAG) bind HARP, store it in the extracellular matrix of many cell types and tissues (Papadimitriou et al., 2000; Vacherot et al., 1999) and contribute to the molecule dimerization (Bernard-Pierrot et al., 1999). The identification of HARP domains that are responsible for its biological activity, is a matter of investigation of several groups (Courty et al., 2000; Deuel, Zhang, Yeh, Silos-Santiago, & Wang, 2002). Several data suggest that the different regions of HARP may exert distinct or even opposite effects. We have previously shown that the two lysine-rich terminal peptides of HARP have an angiogenic effect in vitro and in vivo (Papadimitriou et al., 2000). Furthermore, the last 25 aminoacids of the C-terminal region of HARP are

1955

considered important for HARP binding to ALK and the exertion of its angiogenic activity (Bernard-Pierrot et al., 2002). The two termini of the molecule are also necessary for the successful transformation of cells (Zhang, Zhong, & Deuel, 1999). A chemically synthesized C-domain of HARP (last 43 aminoacids) enhanced plasminogen activator activity and decreased plasminogen activator inhibitor levels in bovine aortic endothelial cells, suggesting a stimulatory effect on angiogenesis (Kojima et al., 1995). On the other hand, amino acid residues 41–45 and 64–68, which are “internal repeats” (GAECK), together function to suppress the transforming potential of HARP and are important regulatory domains (Zhang et al., 1999). HARP also contains two central ␤-sheet domains with thrombospondin type 1 (TSR-1) homology motifs, implicated in its binding to heparin (Kilpelainen et al., 2000; Rauvala et al., 2000). Finally, it is also suggested that HARP over-expression may be implicated in cellular quiescence, rather than an oncogenic phenotype (Corbley, 1997; Szabat & Rauvala, 1996). Extracellular proteolytic enzymes, such as plasmin, play a key role in both induction or inhibition of tumor growth and angiogenesis, giving rise to active proteolytic fragments of larger molecules (Pepper, 2001). Considering that HARP is implicated in angiogenesis and that different regions of HARP could have biological activities, as well as the presence of proteolytic forms of HARP in conditioned media of cells, including endothelial cells (Papadimitriou et al., 2000 and unpublished data), we questioned whether plasmin could generate HARP peptides that regulate angiogenesis. We identified five plasmin-generated HARP peptides with potential activity on angiogenesis in vivo and in vitro. Recombinant HARP peptides with sequences identical to those generated by plasmin had identical activities. Our results indicate that different regions of HARP differentially affect angiogenic processes of endothelial cells.

2. Materials and methods 2.1. Production and purification of human recombinant HARP Expression of human recombinant HARP was induced in Esherichia coli BL21 pLys cells transformed

1956

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

with the human HARP-pETHH8 plasmid (kindly provided by P. Bohlen), as previously described (Papadimitriou et al., 2000). Expression of the recombinant protein was induced for 2 h at 37 ◦ C, by the addition of 2 mM IPTG. Purification of the molecule was achieved by a two step procedure, including heparin–Sepharose affinity and Mono-S chromatography. The protein content of each collected fraction was quantified with a BCA assay (Pierce, Rockford, IL, USA), using bovine serum albumin (BSA) as standard protein. 2.2. Production and purification of recombinant HARP peptides H9–59, H60–110 and H9–110 Glutathione S-transferase (GST) fusion peptides were constructed by polymerase chain reaction (PCR), using the human HARP cDNA as a template. The oligonucleotides were synthesized by incorporating a BamHI restriction site at the 5 end and a stop-codon followed by an EcoRI restriction site at the 3 end. After digestion by EcoRI and BamHI, the PCR products were further purified and subcloned into the pGEX 2T vector (Amersham Pharmacia Biotech, Orsay, France) linearized using the same enzymes. The GST-HARP fusion peptides were isolated by glutathione-agarose affinity, and the GST tag was cleaved by thrombin treatment. Purification of the resulting peptides was achieved by Mono-S chromatography, and the purity (>95%) and integrity of the peptides were routinely checked by SDS–PAGE (data not shown). 2.3. Treatment of HARP with plasmin Ten micrograms (650 pmoles) of human recombinant HARP were incubated for various time periods at 37 ◦ C with 8 pmoles of human plasmin. Aprotinin (16.7 ␮g/ml) was also added when required. In order to test the possible protective effect of GAG, HARP was pre-incubated at room temperature for 30 min with each one of the GAG, including heparin, HS, DS, chondroitin sulfate C (CS-C) and keratan sulfate (KS), in the range of concentrations 1.7–830 ␮g/ml, before the addition of plasmin and in a final volume of 60 ␮l phosphate buffered saline (PBS), pH 7.4. Samples were then incubated with plasmin for 2 h at 37 ◦ C and the reaction was terminated by addi-

tion of 15 ␮l of 5× Laemmli buffer supplemented with ␤-mercaptoethanol. The reaction products were analyzed on 17.5% SDS–PAGE and transferred to Immobilon P membranes. Blocking was performed by incubating the polyvinylidene difluoride membranes with Superblocker (Pierce), overnight at 4 ◦ C under continuous agitation. Western blot analysis was performed by incubating the membranes with specific polyclonal affinity purified anti-HARP antibodies (Papadimitriou et al., 2000) at a final concentration of 1 ␮g/ml in Tris-buffered saline (TBS), pH 7.4, containing 0.05% Tween-20 (TBS-T), for 1 h at room temperature under continuous agitation and then with horseradish peroxidase conjugated rabbit anti-goat IgG (Sigma, Athens, Greece) at a dilution of 1:7500 in TBS-T, for 1 h at room temperature under continuous agitation. Detection of immunoreactive bands was performed by ECL (AmershamPharmaciaBiotech, Buckinghamshire, UK), according to the manufacturers’ instructions. In order to purify the peptides generated after treatment of HARP with plasmin, 1.75 mg HARP were incubated with 1.4 nmoles plasmin in a final volume of 10.5 ml for 20 min at 37 ◦ C. The reaction was ended by the addition of aprotinin (16.7 ␮g/ml). The ionic capacity of the reaction mixture was fixed to 0.5 M NaCl and HARP peptides were purified using Mono-S chromatography and elution with a NaCl linear gradient of 0.5–2 M. Each fraction was analyzed using SDS–PAGE and its protein content was quantified with a BCA assay (Pierce), using BSA as standard protein. Determination of the N-terminal sequence of each proteolytic peptide was performed using the Protein Microsequencing Service at the Pasteur Institute (Paris, France). 2.4. Cell culture Human umbilical vein endothelial cells (HUVEC) were isolated from human umbilical cords, cultured as previously described (Papadimitriou et al., 2000) and used at passages 1–3. The cells were grown as monolayers in medium M199 supplemented with 15% fetal calf serum (FCS), 150 ␮g/ml endothelial cell growth supplement, 5 U/ml heparin sodium, 100 U/ml penicillin-streptomycin and 50 ␮g/ml gentamycin. Cultures were maintained at 37 ◦ C, 5% CO2 and 100% humidity.

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

2.5. Migration assay Migration assays were performed as previously described (Papadimitriou et al., 2001), in 24-well microchemotaxis chambers (Costar, Avon, France), using uncoated polycarbonate membranes with 8 ␮m pores. Briefly, HUVEC were harvested and resuspended at a concentration of 105 cells/0.1 ml in M199 containing 0.25% BSA. The bottom chamber was filled with 0.6 ml of M199 containing 0.25% BSA and the tested peptides. The upper chamber was loaded with 105 cells and incubated for 4 h at 37 ◦ C. After completion of the incubation, the filters were fixed with saline-buffered formalin and stained with 0.33% toluidine blue solution. The cells that migrated through the filter were quantified by counting the entire area of each filter, using a grid and an Optech microscope at a 20× magnification. 2.6. MatrigelTM tube formation assay The matrigel tube formation assay was performed as previously described (Papadimitriou et al., 2001). Briefly, MatrigelTM was used to coat the wells of 96-well tissue culture plates (0.04 ml per well) and was left to polymerize for 1 h at 37 ◦ C. After polymerization, 15,000 cells suspended in 0.15 ml of M199 supplemented with 5% FCS were added to each well. The peptides were added to the corresponding wells simultaneously with the cells. After 6 h of incubation at 37 ◦ C, the medium was removed, the cells were fixed and stained and the length of the tube network was measured in the total area of the wells, as previously described (Papadimitriou et al., 2001). 2.7. Cell adhesion assay One hundred microliters per well of each peptide, diluted in dH2 O at 4 ◦ C, were added in triplicate into the wells of 96-well tissue culture plates. The plates were sealed and incubated overnight at 4 ◦ C. The coating solution was removed, the wells were washed twice with sterile water and blocking was achieved by the addition of 100 ␮l per well of 1% BSA diluted in dH2 O for 4 h at 4 ◦ C. Blocking solution was removed and the wells were washed twice. One hundred microliters of the cell suspension containing 2×105 cells/ml in serum-free M199 were added in each well and the

1957

plate was incubated for 45 min at 37 ◦ C. The non adherent cells were removed, 100 ␮l of freshly prepared saline-buffered formalin were added slowly in each well and cells were fixed for 10 min at room temperature. Staining was performed by adding 100 ␮l per well of freshly filtered solution of 0.1% crystal violet in dH2 O for 25 min at room temperature. The wells were washed twice with distilled water and allowed to dry for 5–10 min at room temperature. Fifty microliters of 0.5% Triton X-100 (diluted in dH2 O) were added into each well and crystal violet was solubilized overnight, under continuous agitation at room temperature. Plates were read in an ELISA microplate reader (Biorad) at 600 nm. 2.8. Chorioallantoic membrane (CAM) assay The in vivo chicken embryo CAM angiogenesis model was used, as previously described (Papadimitriou et al., 2001). Leghorn fertilized eggs (Pindos, Greece) were incubated for 4 days at 37 ◦ C, when a window was opened on the egg shell, exposing the CAM. The window was covered with tape and the eggs were returned to the incubator. The tested agents were diluted in 20 ␮l H2 O and applied on an area of 1 cm2 (restricted by a plastic ring) of the CAM at day 9 of embryo development. In order to evaluate the effect of each substance on angiogenesis, 48 h after treatment and subsequent incubation at 37 ◦ C, CAMs were fixed in situ, excised from the eggs, placed on slides and left to air-dry. Pictures were taken through a stereoscope equipped with a digital camera and the total length of the vessels was measured using image analysis software, as previously described (Papadimitriou et al., 2001). Assays for each test sample were carried out thrice and each experiment included 8–10 eggs per data point. 2.9. Statistical analysis The significance of variability between the results from each group and the corresponding control was determined by unpaired t-test. Each experiment included triplicate wells for each condition tested, unless otherwise indicated, and all results are expressed as mean ± S.E.M. from at least four independent experiments.

1958

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

3. Results 3.1. Characterization of HARP fragments generated after in vitro cleavage by plasmin Plasmin has been shown to control the availability and activation of growth factors and to play an important role during angiogenesis (Pepper, 2001). In the present study, we questioned if plasmin could control the biological activity of HARP during angiogenesis. In vitro incubation of human recombinant HARP with plasmin, led to the generation of five major peptides that can be easily separated by SDS–PAGE. These peptides, named P5, P7, P10, P13 and P16 had estimated molecular weights of 5.5, 7, 10, 13 and 16 kDa, respectively (Fig. 1). When 10 ␮g (about 650 pmoles) of HARP were incubated with 8 pmoles of plasmin, all peptides were present after 5 min of incubation. During this incubation period, the intact molecule of HARP was also present. Complete cleavage of HARP occurred after 20 min, when only the five peptides could be detected. If HARP was incubated with plasmin for a period of 120 min, only the two smaller peptides (P5 and P7) could be detected by SDS–PAGE (Fig. 1). After plasmin cleavage, all five peptides were detectable in Western blots, using an affinity purified anti-HARP polyclonal antibody (Fig. 3, lane 9). As shown in Fig. 2, only two different N-terminal sequences were identified from the five peptides shown. The first one is VKKSD, leading to a cleavage site between K9 and V10 in the N-terminal ran-

dom coiled region of HARP. The second N-terminal sequence is KQFGA, leading to a cleavage site identified as K59–K60, located in the flexible linker between the two ␤-sheet domains of HARP. 3.2. Heparin and GAG protect HARP from plasmin degradation HARP has a well-established affinity for heparin, HS and DS through interaction with its two TSR-1 homology sequence domains (Kilpelainen et al., 2000; Rauvala et al., 2000). In order to further characterize plasmin-generated HARP peptides and to study a possible protective effect of GAG on the degradation of HARP by plasmin, we studied the in vitro effect of GAG on the plasmin-mediated proteolysis of HARP. HARP was incubated with heparin, HS, KS, DS or CS-C, prior to the addition of plasmin for 120 min. Interestingly, as shown in Fig. 3, heparin, DS and HS partially protected HARP from cleavage by plasmin, yielding peptide P13. No protective effect was observed in the presence of CS-C or KS. Dose–response studies indicated that partial inhibition of HARP degradation occurred at concentrations starting from 16.7 ␮g/ml for heparin and 167 ␮g/ml for HS. DS protected HARP to a smaller degree than the other two GAG, at a concentration of 833 ␮g/ml. These results suggest that the GAG protect HARP at different GAG to HARP ratios. The most protective was heparin, followed consecutively by HS and DS.

Fig. 1. Proteolytic cleavage of HARP by plasmin. Human recombinant HARP (10 ␮g) was incubated with 8 pmoles of human plasmin at 37 ◦ C for various time periods. HARP degradation led to the generation of five peptides named P16, P13, P10, P7 and P5. HARP was incubated with plasmin for 0 (lane 1), 5 (lane 2), 10 (lane 3), 20 (lane 4), 30 (lane 5), 60 (lane 6) and 120 (lane 7) min.

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

1959

Fig. 2. Schematic representation of HARP protein and plasmin-generated peptides. Arrows indicate the sites of cleavage by plasmin and numbers indicate the corresponding aminoacids. Hatched boxes correspond to the TSR-1 domains of HARP. Question mark corresponds to the unknown site of cleavage at the C-terminus of HARP. Also shown are the generated N-terminal sequences of the corresponding peptides.

Fig. 3. Heparin and GAG partially protect HARP from degradation by plasmin. HARP was incubated with plasmin for 120 min in the presence (lanes 4–8) or absence (lanes 1–3) of 833 ␮g/ml of different GAG. (1) HARP alone, (2) HARP incubated with plasmin for 120 min, (3) HARP incubated with plasmin for 120 min in the presence of aprotinin, (4) HARP incubated with plasmin for 120 min in the presence of KS, (5) HARP incubated with plasmin for 120 min in the presence of CS-C, (6) HARP incubated with plasmin for 120 min in the presence of DS, (7) HARP incubated with plasmin for 120 min in the presence of HS, (8) HARP incubated with plasmin for 120 min in the presence of heparin, (9) HARP incubated with plasmin for 20 min.

3.3. Biological activity of plasmin-generated HARP peptides in vitro Endothelial cells are the main cellular component of the vessel wall and have important roles in the process of angiogenesis. In order to investigate if and how plasmin-generated HARP peptides influence angiogenic processes, we studied their effect on the migration, adhesion and differentiation of HUVEC, using in vitro assays as previously described (Papadimitriou et al., 2001). As shown in Fig. 4A, the peptides generated after cleavage of HARP with plasmin differentially modulated the migration of HUVEC. Peptides P5,

P7 and P10 significantly induced HUVEC migration (35.8 ± 6.9%, 33.1 ± 4.0% and 32.9 ± 14.2% increase, respectively, compared to unstimulated cells). In contrast, peptides P13 and P16 inhibited in a statistically significant way the migration of HUVEC (18.8 ± 2.2 and 22.6 ± 9.4% decrease, respectively, compared to unstimulated cells). In the same set of experiments, full length HARP molecule had no statistically significant effect on the migration of HUVEC (−3.7±11.5% compared to unstimulated cells), in line with previously published data (Papadimitriou et al., 2001). We next evaluated the effect of the peptides on the ability of HUVEC to form tubes when grown on

1960

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

Fig. 4. Effect of plasmin-generated (A–C) or recombinant (D–F) HARP peptides on the migration (A and D), differentiation on MatrigelTM (B and E) and adhesion (C and F) of HUVEC. For the migration assays, peptides (100 ng/ml) were added at the lower chamber in serum free M199 supplemented with 0.25% BSA. Cells were added at the upper chamber and incubated for 4 h, as described in Section 2. For the matrigel assays, cells were added to each coated well of 96-well plates, in the presence or not of HARP peptides (100 ng/ml). After 6 h, cells were fixed and the tube network was quantified as described in Section 2. For the adhesion assays, culture wells of 96-well tissue culture plates were coated with various peptides before addition of cells. Adherent cells were stained and measured as described in Section 2. In all cases, results are expressed as the mean ± S.E.M. (%) change of control (unstimulated cells), of at least four independent experiments performed in triplicates (*P < 0.05, **P < 0.01, ***P < 0.001).

MatrigelTM . As shown in Fig. 4B, peptides P5, P7, P10, and P16 induced in a statistically significant manner the formation of tubes by HUVEC (16.1 ± 8.3%, 21.8 ± 3.4, 12.9 ± 6.1 and 12.3 ± 1.7% increase, re-

spectively, compared to unstimulated cells). Peptide P13 and HARP had no significant effect (11.5 ± 5.6 and 10.2 ± 6.6% decrease, respectively, compared to unstimulated cells).

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

We also tested the effect of HARP peptides on the adhesion of HUVEC. Each peptide was tested at the concentrations of 10 and 100 ng per well. As shown in Fig. 4C, all plasmin-generated peptides significantly induced HUVEC adhesion and the induction was not statistically different among the peptides. Similarly, intact HARP significantly induced HUVEC adhesion at levels comparable to its peptides (Fig. 4F). To confirm these results, we tested the effect on the migration, differentiation and adhesion of HUVEC of the recombinant peptides H9–110, H9–59 and H60–110, which correspond to the sequences of the peptides P13, P5 and P7, respectively. As shown in Fig. 4D, peptides H9–59 and H60–110 induced (35.0±5.7 and 21.5±10.7% increase, respectively, compared to unstimulated cells), while the peptide H9–110 inhibited HUVEC migration (21.2±6.8% decrease, compared to unstimulated cells). On MatrigelTM , H9–59 and H60–110 caused a small but statistically significant increase in the total length of the tubes (10.5 ± 2.4 and 10.0 ± 3.3% increase, respectively, compared to unstimulated cells), while H9–110 had no significant effect (3.6 ± 2.3% compared to unstimulated cells) (Fig. 4E). Finally, all recombinant peptides caused an increase of HUVEC adhesion at levels comparable to those of the corresponding plasmin-generated peptides (Fig. 4F). Taken together, the HUVEC adhesion results (Fig. 4C and F) suggest that this biological activity of HARP could be mediated by cellular proteoglycan receptors, since each proteolytic fragment or recombinant peptide contain at least one ␤-sheet domain displaying affinity for heparin and GAG. 3.4. Effect of HARP peptides on the in vitro chemotactic and differentiation activities of VEGF165 on HUVEC We have recently shown that HARP directly interacts with VEGF165 via the TSR-1 sequence motif present on both ␤-sheets (Heroult et al., 2004). As the HARP peptides generated from plasmin degradation contain at least one of these domains, we questioned whether P16, P13, P10, P7 and P5 affect VEGF165 -induced HUVEC migration. As shown in Fig. 5A, VEGF165 significantly stimulated (47.1±7.5% increase compared to unstimulated cells) HUVEC migration at a concentration of 10 ng/ml.

1961

Fig. 5. Modulation of the VEGF165 -induced migration of HUVEC by HARP peptides. (A) Peptides generated after cleavage with plasmin. (B) Recombinant HARP peptides (V—VEGF165 , H—HARP). Results are expressed as the mean ± S.E.M. (%) change of control, of at least three independent experiments performed in triplicates (**P < 0.01 and ***P < 0.001).

When VEGF165 was pre-incubated for 30 min at 4 ◦ C with 100 ng/ml of each peptide, a decreased stimulatory effect of VEGF165 was observed. Peptides P5, P7, and P10 partially (17.6 ± 3.5, 26.3 ± 3.7 and 31.6 ± 5.5% increase, respectively, compared to unstimulated cells), while peptides P13 or P16 completely reversed VEGF165 -induced migration (5.2 ± 3.8 and −37.9 ± 9.7%, respectively, compared to unstimulated cells). Similar results were obtained with the recombinant HARP peptides. Pre-incubation of VEGF165 with peptides H9–59 and H60–110 resulted in a partial decrease of the VEGF165 -induced migration of HUVEC (34.7 ± 19.1 and 25.3 ± 12.0%

1962

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

increase, respectively, compared to unstimulated cells), while the peptide H9–110 completely abolished the effect of VEGF165 (20.8 ± 17.2% decrease, compared to unstimulated cells) (Fig. 5B). These data suggest that the presence of both ␤-sheet domains of HARP is necessary to completely reverse the activity of VEGF165 on HUVEC migration. In line with this, full length HARP completely abolished VEGF-induced HUVEC migration (48.7 ± 14.8% decrease compared to unstimulated cells) (Fig. 5A). VEGF165 also induced differentiation of HUVEC grown on MatrigelTM . When 50 ng/ml of VEGF165 were added in the cell culture medium, a statistically significant increase of 17.2 ± 2.9% in the total length of the formed tubes was observed, compared to unstimulated cells. When VEGF165 was pre-incubated

Fig. 7. Effect of plasmin-generated HARP peptides on physiological in vivo angiogenesis in the chicken embryo CAM. One microgram of the tested agents in the same final volume of 20 ␮l were applied on an area of 1 cm2 restricted by a plastic ring, at CAMs of day 9, as described in Section 2. After 48 h of incubation at 37 ◦ C, the CAMs were fixed, excised from the eggs, photographed and the total length of the vessel network was measured using image analysis software. Results are expressed as mean ± S.E.M. (%) change of the number of vessels in treated compared to untreated tissue (control). Asterisks denote a statistically significant difference (unpaired t-test) from the control (***P < 0.001).

with 100 ng/ml of each peptide or HARP for 30 min at 4 ◦ C, the length of the tubes decreased at the levels of control (2.7 ± 5.1, 7.6 ± 5.2, 1.3 ± 3.6, −1.8 ± 3.3, −4.4 ± 3.6 and −4.6 ± 1.7%, compared to unstimulated cells, for P5, P7, P10, P13, P16 and full-length HARP, respectively) (Fig. 6A). Similarly, the recombinant HARP peptides also reversed VEGF165 -induced tube formation (7.4 ± 2.2, 0.4 ± 3.3 and −4.2 ± 1.6% compared to unstimulated cells, for peptides H9–59, H60–110 and H9–110, respectively) (Fig. 6B). 3.5. Angiogenic activity of plasmin-generated HARP peptides in vivo

Fig. 6. Modulation of the VEGF165 -induced differentiation of HUVEC grown on MatrigelTM by HARP peptides. (A) Peptides generated after cleavage with plasmin. (B) Recombinant HARP peptides (V—VEGF165 , H—HARP). Results are expressed as the mean ± S.E.M. (%) change of control, of at least three independent experiments performed in triplicates (**P < 0.01 and ***P < 0.001).

In order to investigate if and how plasmin-generated HARP peptides influence angiogenesis in vivo, we studied their effect on the number of vessels using the in vivo model of the chicken embryo CAM, as previously described (Papadimitriou et al., 2001). Surprisingly, as can be seen in Fig. 7, all plasmin-generated HARP peptides induced angiogenesis with similar efficiency as the full length HARP in this system (Papadimitriou et al., 2001).

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

4. Discussion Although several reports indicate a positive correlation between HARP and in vivo or in vitro angiogenesis (Czubayko, Schulte, Berchem, & Wellstein, 1996; Papadimitriou et al., 2000, 2001), some data suggest that HARP may also have an inhibitory effect in this process (Corbley, 1997). Controversial results have been attributed to differences in the experimental procedures of the production and purification of HARP or the system used. Alternatively, they could be explained by the existence of various forms of HARP that could be obtained after proteolytic degradation of the whole molecule and could have distinct or/and opposite activities. The detection of HARP in the extracellular matrix or serum makes this molecule a possible substrate for proteolytic enzymes, such as plasmin, that are present in the cellular microenvironment. This hypothesis is reinforced by the presence of proteolytic forms of HARP in conditioned media from human and bovine endothelial cells (Papadimitriou et al., 2000). Previous studies have shown that HARP is a substrate for trypsin or chymotrypsin, whose degradative effects are not inhibited by heparin (Delbe, Vacherot, Laaroubi, Barritault, & Courty, 1995). Plasmin is a serine proteinase that catalyzes the hydrolysis of peptide bonds on the C-terminal side of lysine and arginine residues. The presence of many lysine residues in the protein sequence of HARP (Hampton et al., 1992) makes the molecule a potent substrate of plasmin. The first cleavage site, as identified after N-terminal sequencing of the generated peptides, is at the K9–V10 peptide bond within the cluster of lysine residues located at the N-terminal region of the molecule. HARP degradation at this region results to the removal of a significant portion of the N-terminal region of the growth factor, which has been considered important for its biological activity (Papadimitriou et al., 2000, 2001). The second cleavage site has been identified at the K59–K60 peptide bond, located at the flexible linker between the two globular domains of the HARP molecule (Kilpelainen et al., 2000). Our results suggest that the plasmin activity at this region of HARP could be critical for the biological activity of the resulting peptides. Indeed, HARP peptides containing only one ␤-sheet domain seem to positively regulate in vitro angiogenesis processes, while HARP peptides containing both ␤-sheet

1963

domains seem to have an inhibitory effect. A similar difference was observed when the proteolytic or recombinant peptides were tested on the in vitro chemotactic activity of VEGF165 . These last results could be explained by the requirement of both ␤-sheet domains of HARP in the binding of VEGF165 (Heroult et al., 2004). The third cleavage results to the removal of an unknown portion from the C-terminal region of HARP. The C terminus of HARP is implicated in its angiogenic activity (Bernard-Pierrot et al., 2001; Papadimitriou et al., 2000, 2001) and is considered necessary for the binding of HARP to its receptor ALK (Bernard-Pierrot et al., 2002). Thus, proteolysis of HARP by plasmin could be a critical regulatory step during angiogenesis, which results in the production of peptides with distinct biological activities. The molecular weights of the peptides generated after cleavage of HARP with plasmin were estimated by SDS–PAGE as 16, 13, 10, 7 and 5.5 kDa. The existence of HARP peptides with molecular weights 16 and 13 kDa have been previously detected in vivo (Bohlen et al., 1991; Hampton et al., 1992). In those reports, a partial removal of the C-terminal region has been suggested as the mechanism of generation of those smaller forms of HARP. A 15 kDa form of HARP has also been detected in cases when protease inhibitors were omitted from the purification procedure (Hampton et al., 1992). We are also continuously obtaining from cell culture media or tissue extracts, bands that are specifically recognized by anti-HARP antibodies in Western blots, whose molecular weights are smaller than that of the intact growth factor (unpublished data). To what extent these forms of HARP could be attributed to the activity of plasmin, is further being investigated but other proteolytic enzymes like metalloproteinase associated with the remodeling of the extracellular compartment could also be implicated. Indeed, we have recently shown that HARP is a substrate of metalloproteinase 2 (unpublished data). The protection conferred from the different GAG corresponds to their affinity for HARP (Vacherot et al., 1999). However, in all cases, protection was not total and always led to the appearance of the band that corresponds to the peptide P13. This suggests that the GAG protect the molecule from cleavage at the K59–K60 peptide bond, which is located between the two ␤-sheet domains of HARP, but do not protect from cleavage at the other two sites, which are located at

1964

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

the two termini of the molecule. This is probably due to the fact that the two termini of the molecule do not bind heparin (Kilpelainen et al., 2000). The protection of HARP by HS or DS associated with the extracellular matrix and/or cellular membrane proteoglycans, could be of major importance for the biological activities of HARP during angiogenesis, since depending on the presence or absence of these GAG in the cellular microenvironment, HARP could positively or negatively regulate this process. Indeed, all peptides containing either the N-central (P5 and H9–59) or the C-central (P7, P10 and H60–110) domain of HARP stimulate HUVEC migration. On the contrary, peptides containing both central domains (P13, P16 and H9–H110) have an inhibitory effect. Similar results were observed when the peptides were tested for their effect on HUVEC differentiation (matrigel assay), with the exception of peptide P16, which is the only peptide tested that contains the entire C-terminal domain of HARP and has a significant stimulatory effect in this assay. The C-terminal region of HARP has been shown to be responsible for its angiogenic and transforming activities (Bernard-Pierrot et al., 2001; Bernard-Pierrot et al., 2002) and its presence on peptide P16 may overcome in this system the inhibitory effect of the two central domains of HARP. We have previously shown that HARP directly interacts with VEGF165 and inhibits its activity on HUVEC. For this interaction to be measurable, both central domains of HARP are required, as mentioned above. In line with this, proteolytic and recombinant peptides containing one of the central domains of HARP (P5, P7, P10, H9–59 and H60–110) only partially, while peptides that contain both central domains of HARP (P13, P16 and H9–110) totally inhibited the activity of VEGF on HUVEC. The partial inhibition caused by the peptides corresponding to only one of the central domains of HARP might be attributed to dislocation of VEGF from the cell surface due to common binding sites, or to a possible interference with VEGF signaling. In contrast to the in vitro results, in the in vivo chicken CAM model of angiogenesis, all plasmin-generated HARP peptides induced angiogenesis in a statistically significant manner, in agreement with the effect of full length HARP in this system (Papadimitriou et al., 2001). The mechanisms involved in the angiogenic action of HARP in the CAM are not

elucidated and it remains unknown whether its angiogenic effect is due to interactions with cell surface proteoglycans (Vacherot et al., 1999) or binding to another cell surface receptor, such as N-syndecan (Raulo et al., 1994), receptor protein tyrosine phosphatase ␤/␨ (RPTP␤/␨) (Deuel et al., 2002; Maeda et al., 1996; Meng et al., 2000) and anaplastic lymphoma kinase (ALK) (Stoica et al., 2001). It is possible that HARP has an indirect effect, releasing growth factors sequestered in the ECM, such as FGF-2 or VEGF, by inducing stimulation of proteinases (Kojima et al., 1995). Alternatively, since HARP induces proliferation of human peripheral blood mononuclear cells (Achour, Laaroubi, Caruelle, Barritault, & Courty, 2001), it might be possible that at least part of the angiogenic effect of HARP and HARP peptides on the chicken embryo CAM is due to activation of CAM blood cells and not due to a direct effect on CAM endothelial cells. Finally, it may be possible that other proteinases present in the chicken embryo CAM, further cleave plasmin-generated HARP peptides, leading to a molecular form of HARP that is angiogenic. Proteinases are known to contribute both to induction and down-regulation of angiogenesis. Almost half of the known inhibitors of angiogenesis are proteolytic fragments from larger angiogenic molecules, such as endostatin and angiostatin, the 16 kDa N-terminal fragment of prolactin and the PEX metalloproteinase 2 fragment (Cao, 2001). Thus, it appears that the generation of endogenous inhibitors in vivo from large precursor proteins with distinct functions is a recurrent theme in the inhibition of angiogenesis. In accordance with these examples, HARP could act as pro- or anti-angiogenic factor, depending on the system used and the cell microenvironment (types of proteoglycans and their proportional ratio, receptors, proteinases etc), which could each time regulate the predominant form of HARP. In conclusion, we present for the first time that HARP is a substrate for plasmin and that the peptides produced by such a cleavage may act either as inducers or inhibitors of angiogenesis. Furthermore, molecules of the extracellular matrix affect the production of these peptides and contribute to the effect of HARP in a more complicated manner than previously suggested (Bernard-Pierrot et al., 2001; Vacherot et al., 1999). Further studies are in progress in order to clarify the physiological significance of these data.

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

Acknowledgements This work was in part supported by grants from the Research Committee of the University of Patras (Karatheodoris) and Empeirikio Foundation, Greece and from Association pour la Recherche sur le Cancer (Jose Courty No. 4248). We are grateful to the Private Maternity Clinic of Patras for providing us with umbilical cords. A. Polykratis is a recipient of a fellowship from the State Scholarships Foundation. References Achour, A., Laaroubi, D., Caruelle, D., Barritault, D., & Courty, J. (2001). The angiogenic factor heparin affin regulatory peptide (HARP) induces proliferation of human peripheral blood mononuclear cells. Cell and Molecular Biology (Noisy-le-grand), 47 (2001) Online Pub: OL73–7. Bernard-Pierrot, I., Heroult, M., Lemaitre, G., Barritault, D., Courty, J., & Milhiet, P. E. (1999). Glycosaminoglycans promote HARP/PTN dimerization. Biochemical and Biophysical Research Communications, 266, 437–442. Bernard-Pierrot, I., Delbe, J., Caruelle, D., Barritault, D., Courty, J., & Milhiet, P. E. (2001). The lysine-rich C-terminal tail of heparin affin regulatory peptide is required for mitogenic and tumor formation activities. Journal of Biological Chemistry, 276, 12228–12234. Bernard-Pierrot, I., Delbe, J., Rouet, D. V., Vigny, M., Kerros, M. E., & Caruelle, D. et al., (2002). Dominant negative effectors of heparin affin regulatory peptide (HARP) angiogenic and transforming activities. Journal of Biological Chemistry, 277, 32071–32077. Bohlen, P., Muller, T., Gautschi-Sova, P., Albrecht, U., Rasool, C. G., & Decker, M. et al., (1991). Isolation from bovine brain and structural characterization of HBNF, a heparin-binding neurotrophic factor. Growth Factors, 4, 97–107. Cao, Y. (2001). Endogenous angiogenesis inhibitors and their therapeutic implications. International Journal of Biochemistry and Cell Biology, 33, 357–369. Chauhan, A. K., Li, Y. S., & Deuel, T. F. (1993). Pleiotrophin transforms NIH 3T3 cells and induces tumors in nude mice. Proceedings of the National Academy of Sciences of the United States of America, 90, 679–682. Corbley, M. J. (1997). Transformation by ras suppresses expression of the neurotrophic growth factor pleiotrophin. Journal of Biological Chemistry, 272, 24696–24702. Courty, J., Dauchel, M. C., Caruelle, D., Perderiset, M., & Barritault, D. (1991). Mitogenic properties of a new endothelial cell growth factor related to pleiotrophin. Biochemical and Biophysical Research Communications, 180, 145–151. Courty, J., Milhiet, M. E., Delbe, J., Caruelle, D., & Barritault, D. (2000). Heparin-affin regulatory peptide, HARP. In Comprehensive vascular biology and pathology—an encyclopedic reference (pp. 145–152). Berlin: Springer-Verlag.

1965

Czubayko, F., Riegel, A. T., & Wellstein, A. (1994). Ribozyme-targeting elucidates a direct role of pleiotrophin in tumor growth. Journal of Biological Chemistry, 269, 21358– 21363. Czubayko, F., Schulte, A. M., Berchem, G. J., & Wellstein, A. (1996). Melanoma angiogenesis and metastasis modulated by ribozyme targeting of the secreted growth factor pleiotrophin. Proceedings of the National Academy of Sciences of the United States of America, 93, 14753–14758. Delbe, J., Vacherot, F., Laaroubi, K., Barritault, D., & Courty, J. (1995). Effect of heparin on bovine epithelial lens cell proliferation induced by heparin affin regulatory peptide. Journal of Cellular Physiology, 164, 47–54. Deuel, T. F., Zhang, N., Yeh, H-J., Silos-Santiago, I., & Wang, Z. Y. (2002). Pleiotrophin: A cytokine with diverse functions and a novel signaling pathway. Archives of Biochemistry and Biophysics, 397(2), 162–171. Gieffers, C., Engelhardt, W., Brenzel, G., Matsuishi, T., & Frey, J. (1993). Receptor binding of osteoblast-specific factor 1 (OSF-1/HB-GAM) to human osteosarcoma cells promotes cell attachment. European Journal of Cell Biology, 62, 352–361. Hampton, B. S., Marshak, D. R., & Burgess, W. H. (1992). Structural and functional characterization of full-length heparin-binding growth associated molecule. Molecular Biological Cell, 3, 85–93. Heroult, M., Bernard-Pierrot, I., Delbe, J., Hamma-Kourbali, Y., Katsoris, P., & Barritault, D. et al., (2004). Heparin affin regulatory peptide binds to vascular endothelial growth factor (VEGF) and inhibits VEGF-induced angiogenesis. Oncogene, 23, 1745–1753. Kilpelainen, I., Kaksonen, M., Avikainen, H., Fath, M., Linhardt, R. J., & Raulo, E. et al., (2000). Heparin-binding growth-associated molecule contains two heparin-binding beta-sheet domains that are homologous to the thrombospondin type I repeat. Journal of Biological Chemistry, 275, 13564–13570. Kojima, S., Inui, T., Kimura, T., Sakakibara, S., Muramatsu, H., & Amanuma, H. et al., (1995). Synthetic peptides derived from midkine enhance plasminogen activator activity in bovine aortic endothelial cells. Biochemical and Biophysical Research Communications, 206, 468–473. Li, Y. S., Milner, P. G., Chauhan, A. K., Watson, M. A., Hoffman, R. M., & Kodner, C. M. et al., (1990). Cloning and expression of a developmentally regulated protein that induces mitogenic and neurite outgrowth activity. Science, 250, 1690–1694. Maeda, N., & Noda, M. (1998). Involvement of receptor-like protein tyrosine phosphatase zeta/RPTPbeta and its ligand pleiotrophin/heparin-binding growth-associated molecule (HBGAM) in neuronal migration. Journal of Cell Biology, 142, 203–216. Maeda, N., Nishiwaki, T., Shintani, T., Hamanaka, H., & Noda M, M. (1996). 6B4 proteoglycan/phosphacan, an extracellular variant of receptor-like protein-tyrosine phosphatase zeta/RPTPbeta, binds pleiotrophin/heparin-binding growth-associated molecule (HB-GAM). Journal of Biological Chemistry, 271, 21446–21452. Meng, K., Rodriguez-Pena, A., Dimitrov, T., Chen, W., Yamin, M., & Noda, M. et al., (2000). Pleiotrophin signals

1966

A. Polykratis et al. / The International Journal of Biochemistry & Cell Biology 36 (2004) 1954–1966

increased tyrosine phosphorylation of beta beta-catenin through inactivation of the intrinsic catalytic activity of the receptor-type protein tyrosine phosphatase beta/zeta. Proceedings of the National Academy of Sciences of the United States of America, 97, 2603–2608. Milner, P. G., Li, Y. S., Hoffman, R. M., Kodner, C. M., Siegel, N. R., & Deuel, T. F. (1989). A novel 17Kd heparin-binding growth factor (BGF-8) in bovine uterus: Purification and N-terminal amino acid sequence. Biochemical and Biophysical Research Communications, 165, 1096–1103. Neame, P. J., Young, C. N., Brock, C. W., Treep, J. T., Ganey, T. M., & Sasse, J. et al., (1993). Pleiotrophin is an abundant protein in dissociative extracts of bovine fetal epiphyseal cartilage and nasal cartilage from newborns. Journal of Orthopedic Research, 11, 479–491. Papadimitriou, E., Heroult, M., Courty, J., Polykratis, A., Stergiou, C., & Katsoris, P. (2000). Endothelial cell proliferation induced by HARP: Implication of N or C terminal peptides. Biochemical and Biophysical Research Communications, 274, 242–248. Papadimitriou, E., Polykratis, A., Courty, J., Koolwijk, P., Heroult, M., & Katsoris, P. (2001). HARP induces angiogenesis in vivo and in vitro: Implication of N or C terminal peptides. Biochemical and Biophysical Research Communications, 282, 306–313. Pepper, M. S. (2001). Extracellular proteolysis and angiogenesis. Thrombosis and Haemostasis, 86, 346–355. Raulo, E., Julkunen, I., Merenmies, J., Pihlaskari, R., & Rauvala, H. (1992). Secretion and biological activities of heparin-binding growth-associated molecule. Neurite outgrowth-promoting and mitogenic actions of the recombinant and tissue-derived protein. Journal of Biological Chemistry, 267, 11408–11416. Raulo, E., Chernousov, M. A., Carey, D. J., Nolo, R., & Rauvala, H. (1994). Isolation of a neuronal cell surface receptor of heparin binding growth-associated molecule (HB-GAM). Identification

as N-syndecan (syndecan-3). Journal of Biological Chemistry, 269, 12999–13004. Rauvala, H., Huttunen, H. J., Fages, C., Kaksonen, M., Kinnunen, T., & Imai, S. et al., (2000). Heparin-binding proteins HB-GAM (pleiotrophin) and amphoterin in the regulation of cell motility. Matrix Biology, 19, 377–387. Rauvala, H. (1989). An 18-kd heparin-binding protein of developing brain that is distinct from fibroblast growth factors. EMBO Journal, 8, 2933–2941. Souttou, B., Raulais, D., & Vigny, M. (2001). Pleiotrophin induces angiogenesis: Involvement of the phosphoinositide-3 kinase but not the nitric oxide synthase pathways. Journal of Cellular Physiology, 187, 59–64. Stoica, G. E., Kuo, A., Aigner, A., Sunitha, I., Souttou, B., & Malerczyk, C. et al., (2001). Identification of anaplastic lymphoma kinase as a receptor for the growth factor pleiotrophin. Journal of Biological Chemistry, 276, 16772–16779. Szabat, E., & Rauvala, H. (1996). Role of HB-GAM (heparin-binding growth-associated molecule) in proliferation arrest in cells of the developing rat limb and its expression in the differentiating neuromuscular system. Developmental Biology, 178, 77–89. Vacherot, F., Delbe, J., Heroult, M., Barritault, D., Fernig, D. G., & Courty, J. (1999). Glycosaminoglycans differentially bind HARP and modulate its biological activity. Journal of Biological Chemistry, 274, 7741–7747. Zhang, N., Zhong, R., Wang, Z. Y., & Deuel, T. F. (1997). Human breast cancer growth inhibited in vivo by a dominant negative pleiotrophin mutant. Journal of Biological Chemistry, 272, 16733–16736. Zhang, N., Zhong, R., & Deuel, T. F. (1999). Domain structure of pleiotrophin required for transformation. Journal of Biological Chemistry, 274, 12959–12962.