Signal transduction in endothelial cells by the angiogenesis inhibitor histidine-rich glycoprotein targets focal adhesions

Signal transduction in endothelial cells by the angiogenesis inhibitor histidine-rich glycoprotein targets focal adhesions

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6 a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m w w w. e l s...

446KB Sizes 2 Downloads 26 Views

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / y e x c r

Research Article

Signal transduction in endothelial cells by the angiogenesis inhibitor histidine-rich glycoprotein targets focal adhesions ○

Chunsik Lee, Johan Dixelius 1 , Asa Thulin, Harukiyo Kawamura, Lena Claesson-Welsh, Anna-Karin Olsson⁎ Department of Genetics and Pathology, Rudbeck Laboratory, Uppsala University, Dag Hammarskjölds väg 20, SE-751 85 Uppsala, Sweden

ARTICLE INFORMATION

ABS T R AC T

Article Chronology:

Histidine-rich glycoprotein (HRGP) is an abundant heparin-binding plasma protein. We have

Received 2 March 2006

shown that a fragment released from the central histidine/proline-rich (His/Pro-rich)

Revised version received

domain of HRGP blocks endothelial cell migration in vitro and vascularization and growth of

7 April 2006

murine fibrosarcoma in vivo. The minimal active HRGP domain exerting the anti-angiogenic

Accepted 19 April 2006

effect was recently narrowed down to a 35 amino acid peptide, HRGP330, derived from the

Available online 13 May 2006

His/Pro-rich domain of HRGP. By use of a signal transduction antibody array representing 400 different signal transduction molecules, we now show that HRGP and the synthetic

Keywords:

peptide HRGP330 specifically induce tyrosine phosphorylation of focal adhesion kinase and

Anti-angiogenesis

its downstream substrate paxillin in endothelial cells. HRGP/HRGP330 treatment of

Endothelial cell

endothelial cells induced disruption of actin stress fibers, a process reversed by treatment

Histidine-rich glycoprotein

of cells with the FAK inhibitor geldanamycin. In addition, VEGF-mediated endothelial cell

HRGP/HPRG/HRG

tubular morphogenesis in a three-dimensional collagen matrix was inhibited by HRGP and

Focal adhesion

HRGP330. In contrast, VEGF-induced proliferation was not affected by HRGP or HRGP330,

Tubular morphogenesis

demonstrating the central role of cell migration during tube formation. In conclusion, our data show that HRGP targets focal adhesions in endothelial cells, thereby disrupting the cytoskeletal organization and the ability of endothelial cells to assemble into vessel structures. © 2006 Elsevier Inc. All rights reserved.

Introduction Angiogenesis, de novo synthesis of blood vessels from the preexisting vasculature, is required during embryonic development and in pathophysiological conditions such as wound healing, tumor growth, and rheumatoid arthritis [1,2]. Angiogenesis is tightly controlled by a number of pro- and antiangiogenic molecules. The local change in balance between positive and negative factors serves to regulate the angiogenic activity of the microvasculature. A growing tissue such as a

tumor will induce angiogenic activity in the neighboring vasculature, e.g. due to hypoxia-induced expression of angiogenic growth factors such as vascular endothelial growth factor (VEGF) [3]. On the other hand, endogenous inhibitors of angiogenesis, such as endostatin and thrombospondin, have been shown to efficiently suppress growth of murine tumor models [4,5]. Proteins with transforming capacity such as Ras and Myc have been shown to downregulate the expression of thrombospondin, thereby facilitating tumor angiogenesis and growth [6]. Proof of the concept that endogenous angiogenesis

⁎ Corresponding author. Fax: +46 18 55 89 31. E-mail address: [email protected] (A.-K. Olsson). 1 Present address: MBB, Karolinska Institute, SE-171 77 Stockholm, Sweden. 0014-4827/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2006.04.022

2548

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

inhibitors act to suppress tumor angiogenesis has been provided in studies of gene targeted animals [7,8]. Histidine-rich glycoprotein (HRGP; alternatively, HRG/ HPRG) has been identified as an angiogenesis inhibitor in vitro and in vivo by us and others [9,10]. HRGP is a 75 kDa single chain heparin-binding plasma protein produced by the liver and stored in platelets [11]. Structurally, HRGP consists of three distinct parts. The NH2-terminal part contains two cystatin (cysteine proteinase inhibitor)-like domains, which classifies HRGP as a member of the cystatin superfamily together with, for example, kininogen and fetuin. The central part of HRGP consists of a unique histidine/proline-rich (His/ Pro-rich) domain organized in tandem repeats of a consensus GHHPH motif. The His/Pro-rich domain, which may be proteolytically released, is followed by a C-terminal stretch of 68 amino acid residues. The detailed structure of HRGP has not been resolved due to the intrinsically unstructured features of the His/Pro-rich domain [12]. HRGP is known to engage in a wide variety of interactions and bind divalent cations as well as heparan sulfate [13–16]. Its biological function has been obscure, but HRGP has been implicated in regulation of blood coagulation, immune complex formation and clearance [17,18]. In accordance, HRGP gene inactivation leads to enhanced coagulation and fibrinolytic activity [19]. We have previously shown that HRGP exerts anti-angiogenic effects established as inhibition of endothelial cell migration, reduced tumor growth rate and decreased vascular density of murine tumor tissue [10]. The anti-angiogenic properties are contained within the central His/Pro-rich domain of HRGP [9,10,20,21]. We have recently identified the minimal active region of HRGP as a 35 amino acid residue peptide, HRGP330, derived from the His/Pro-rich domain. HRGP330 retains the anti-angiogenic properties in vitro and in vivo [21]. HRGP and HRGP330 block adhesion of endothelial cells to vitronectin in a manner implicating vitronectinbinding integrins as transducers of the anti-angiogenic effect [21]. In addition, binding to heparan sulfate/heparan sulfate proteoglycans on the endothelial cell surface is required for HRGP/HRGP330 to exert its anti-angiogenic effect [22]. The mechanism of action of HRGP330 involves disruption of VEGFinduced complex formation between paxillin and integrinlinked kinase in focal adhesions, an interaction required for cell migration. Moreover, HRGP330 induces phosphorylation of α-actinin in endothelial cells and hence a less motile cytoskeleton [21]. The present study presents a pan-analysis of signal transduction pathways in endothelial cells induced by HRGP330, demonstrating narrow and specific targeting of the actin cytoskeleton.

Christofferson, Department of Medical Cell Biology, Uppsala University, Sweden) were cultured in DMEM/10% newborn calf serum (Invitrogen) with 2 ng/ml FGF-2 (Peprotech).

Preparation of HRGP and HRGP330 Recombinant His-tagged full-length HRGP was generated and purified as described previously [10]. The synthetic peptide HRGP330 (35 amino acid residues; DLHPHKHHSHEQHPHGHHQ PHAHHPHEHDTHRQHPH-COOH, derived from the His–Pro-rich domain of HRGP) was synthesized by Innovagen AB (Lund, Sweden).

Antibody array screening TIME cells were cultured on vitronectin-coated cell culture dishes, serum-starved overnight, and treated with 100 ng/ml HRGP330 for 10 min. This concentration of HRGP330 was determined as optimal in inhibition of endothelial cell chemotaxis [21]. The cells were washed twice with icecold Tris-buffered saline (TBS; 25 mM Tris–HCl, pH 7.5, 150 mM NaCl) and lysed in Triton extraction buffer containing 15 mM Tris–HCl, pH 7.5, 120 mM NaCl, 25 mM KCl, 2 mM EDTA, 0.1 mM DTT, 0.5% Triton X-100, 10 μg/ml leupeptin, 100 μM Na3VO4, and 0.5 mM phenylmethylsulfonyl fluoride (PMSF). The antibody array (HM3000 Signal Transduction Antibody Array; Hypromatrix, Worcester, MA) was incubated in TBST buffer (25 mM Tris–HCl, pH 7.5, 150 mM NaCl, 0.05% Tween-20) containing 5% BSA for 1 h to block unspecific reactivity. This was followed by incubation of the filters with cleared total cell extracts at room temperature for 2 h. The antibody array filter was washed with TBST and blotted with HRP-conjugated anti-phosphotyrosine monoclonal antibody (HM2040; Hypromatrix, Worcester, MA) for 2 h. Anti-phosphotyrosine reactivity was visualized by enhanced chemiluminescence (ECL; Amersham Biosciences). The filter was scanned and quantified by Image processing tool kit® (Reindeer Graphics, Asheville, NC) software. Immunoreactivity on the filter incubated with lysate from control-treated cells was set to 1 for each spotted antibody. Table 1A,B shows the result of the Hypromatrix array experiment for a selected group of proteins which may be tyrosine-phosphorylated and have been reported to be expressed in endothelial cells. The phosphotyrosine signal from the epidermal growth factor receptor (EGFR) antibody on the array could not be quantified due to locally high background on the filter and was therefore excluded. For a complete list of antibodies present on the filter, see www.hypromatrix.com.

Materials and methods

Immunoprecipitation and western blotting

Cell culture

Starved TIME and BCE cells were treated or not with 100 ng/ ml full-length HRGP or HRGP330 for indicated time periods before lysis. When included, geldanamycin (No. G3381; Sigma-Aldrich) was used at 3 nM and cells were preincubated for 1 h before HRGP or HRGP330 treatment. Cells were lysed in NP-40 lysis buffer (1% Nonidet P-40, 150 mM NaCl, 10% glycerol, 20 mM HEPES, 1 mM PMSF, 2.5 mM EDTA, 100 μM Na3VO4, and 1% aprotinin). For immunoprecipitation,

Telomerase-immortalized microvascular endothelial (TIME) cells [23] (a kind gift from Dr. Martin McMahon, Cancer Research Institute, UC San Francisco, USA) were cultured in endothelial cell basal medium (EBM) MV2 with supplements (PromoCell, Heidelberg, Germany). Bovine adrenal cortex capillary endothelial (BCE) cells (a kind gift from Dr. Rolf

2549

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

Table 1A – HRGP330-induced changes in tyrosine phosphorylation level of tyrosine kinases

antibody and visualized by enhanced chemiluminescence (ECL; Amersham Biosciences).

Tyrosine kinases Designation Jak3 Jak1 FAK Jak2 b Fyn FGFR2 b Met Flt-3/2 Csk erbB4 Syk BMX erbB2 b c-Fgr FGFR3 Flt-4 FGFR1 c-Kit EphA1 b EphA4 b erbB3 b FGFR4 Tyk2 VEGFR1 c-Src c-Abl b EphB1 Pyk2 Yes b VEGFR2 b Lyn

Description Janus kinase-3 Janus kinase-1 Focal adhesion kinase Janus kinase-2 Proto-oncogene tyrosine kinase Fibroblast growth factor receptor-2 Hepatocyte growth factor receptor Stem cell tyrosine kinase 1 c-src tyrosine kinase Epidermal growth factor receptor Spleen tyrosine kinase BMX non-receptor tyrosine kinase Epidermal growth factor receptor Feline sarcoma viral oncogene homolog Fibroblast growth factor receptor-3 Vascular endothelial growth factor receptor-3 Fibroblast growth factor receptor-1 Stem cell factor kinase receptor Eph receptor Eph receptor Epidermal growth factor receptor Fibroblast growth factor receptor-4 Cytoplasmic tyrosine kinase Vascular endothelial growth factor receptor-1 Rous sarcoma oncogene V-abl Abelson murine leukemia viral oncogene homolog Eph receptor Proline-rich tyrosine kinase 2 Yamaguchi sarcoma viral (v-yes) oncogene homolog Vascular endothelial growth factor receptor-2 Yamaguchi sarcoma viral (v-yes-1) oncogene homolog

Relative ratio a 1.97 1.94 1.37 1.36 1.28 1.27 1.22 1.19 1.16 1.14 1.11 1.11 1.11 1.10 1.09 1.03 0.97 0.90 0.90 0.90 0.86 0.82 0.81 0.77 0.69 0.65

Immunofluorescence TIME cells were plated on vitronectin-coated cell culture slides, serum-starved overnight and treated for 10 min with 100 ng/ml HRGP or HRGP330. When included, geldanamycin was used at 3 nM and cells were pre-incubated for 1 h before HRGP or HRGP330 treatment. The slides were washed with chilled TBS and fixed in Zn-fix solution (0.05% Ca acetate, 0.5% ZnCl2, 0.5% Zn acetate, and 0.2% Tween-20 in TBS). PhosphoFAK was detected using an anti-phosphotyrosine FAK antibody (44–652; Biosource) directed against tyrosine 576 in the kinase domain and an Alexa-568-conjugated secondary antibody (A-11011; Molecular Probes). F-actin was detected by using Texas Red-X phalloidin (T7471; Molecular Probes) diluted 1:200 in TBS/1% fetal calf serum. Nuclei were stained with Hoechst 33342 (1 μg/ml; Molecular Probes). The samples were examined and photographed using a Nikon Eclipse microscope and a Nikon DXM1200 camera.

Quantification of cells with disrupted stress fibers The number of cells with disrupted actin stress fiber formation was calculated from 1000 randomly selected cells from each treatment and presented as the average number of cells in three independent experiments, ±SD. Statistical significance was determined by Student's t test and differences were considered significant at P < 0.05.

Tube formation assay 0.64 0.57 0.47 0.41

Collagen type I (Vitrogen; Cohesion Technologies Inc.) was mixed with 0.1 M NaOH and 10× Ham's F12 medium (8:1:1), and the following components were added to the indicated final concentration: 20 mM HEPES, 0.1% w/v bicarbonate, 2 mM

0.28

a

TIME cells treated with 100 ng/ml HRGP330 for 10 min were lysed and tyrosine phosphorylation events examined using the Hypromatrix antibody array. Ratios represent the phosphotyrosine signal in HRGP330-treated cells in relation to phosphotyrosine signals in control cells. b These spots were quantified individually due to varying background conditions on the filter, and their relative ratio may be less reliable.

cell lysates were incubated with anti-FAK antibody (SC-557; Santa Cruz Biotechnology) or anti-phosphotyrosine antibody (p-Tyr-102; Cell Signaling) for 2 h on ice and precipitated with protein–A Sepharose. Immunoprecipitated samples and total cell lysates were separated by SDS-PAGE in 10% or 8% gels and transferred to a nitrocellulose membrane. The membranes were incubated with anti-phosphotyrosine antibody (pY99; Santa Cruz Biotechnology), anti-Grb2 antibody (SC-255; Santa Cruz Biotechnology) or anti-FAK antibody (SC-557; Santa Cruz Biotechnology). Immunoreactivity was detected using horseradish-peroxidase (HRP)-conjugated secondary

Table 1B – HRGP330-induced changes in tyrosine phosphorylation level of cell adhesion/migration molecules Cell adhesion/migration Designation Grb2 Paxillin Integrin β1 FAK Grb7 Fyn Ezrin Csk Nck p-c-Raf-1 p130Cas Crk Integrin β3 c-Src

Description

Relative ratio

Growth factor receptor bound protein-2 Cell adhesion/motility Integrin beta subunit Focal adhesion kinase Growth factor receptor bound protein-7 Proto-oncogene tyrosine kinase Microvillar cytoplasmic peripheral membrane protein c-src tyrosine kinase SH2/SH3 adaptor protein NCK-alpha Raf proto-oncogene Crk-associated substrate v-crk sarcoma virus CT10 oncogene homolog Integrin beta subunit Rous sarcoma oncogene

1.94 1.60 1.56 1.37 1.32 1.28 1.26 1.16 1.15 1.08 1.05 0.91 0.84 0.69

2550

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

Glutamax-I (Invitrogen). The collagen mix was added into 6well plates and incubated at 37°C overnight. Serum-starved TIME cells were seeded at a density of 650,000 cells/well on top of the collagen layer and allowed to adhere at 37°C for 2 h. A second layer of collagen was added on top of the cells, which was allowed to solidify for 1 h. After solidification, top medium containing 2% FCS and VEGF-A (50 ng/ml), with or without HRGP (100 ng/ml) or HRGP330 (100 ng/ml), was added. Tube formation was photographed 24 h later in a 20× objective. Quantification of total tube length was performed on three randomly selected fields from each treatment using the Image processing tool kit® (Reindeer Graphics, Asheville, NC) software. The software “skeleton” function was used that estimates the total length of connected structures and therefore excludes single cells. Data are presented as a mean length ± SEM, n = 3 experiments. Statistical significance was determined by one-way analysis of variance (ANOVA), and differences were considered significant at P < 0.05.

Fig. 1 – Signal transduction antibody array analysis of HRGP330-treated endothelial cells. Serum-starved TIME cells were untreated (left) or treated (right) with HRGP330 (100 ng/ml) for 10 min. The antibody array filter was incubated with total cell lysates before detection of tyrosine phosphorylated proteins with an HRP-conjugated anti-phosphotyrosine monoclonal antibody. Immunoreactive sites were visualized by ECL. Areas of the filter arrays containing FAK, paxillin, and Grb2 immunoreactive sites are shown.

Proliferation assay TIME cells were plated on gelatin- or vitronectin-coated cell culture dishes and serum-starved for 18 h in 1% FCS. Triplicate samples were stimulated for 22 h with 4 ng/ml VEGF-A, in the presence or absence of 100 ng/ml HRGP330. 1 μCi/ml [methyl3 H] thymidine (Amersham Biotech; TRK686) was included during the last 6 h of stimulation. After the incubation, cells were put on ice and washed twice with ice-cold PBS. DNA was precipitated with 10% trichloroacetic acid (TCA) for 20 min on ice followed by two washes with 99% ethanol. Precipitated DNA was solubilized with 0.2 M NaOH at room temperature for 10 min and the amount of incorporated radioactivity was determined by scintillation counting.

Results Global signal transduction analysis of HRGP330-treated endothelial cells To map the spectrum of targets of HRGP and HRGP330 in angiogenesis inhibition, signal transduction antibody arrays containing 400 different antibodies were probed with lysates derived from vehicle and HRGP330-treated human microcapillary endothelial (TIME) cells. Subsequent incubation with anti-phosphotyrosine antibody allowed identification of tyrosine phosphorylation events induced by HRGP330. Table 1A shows the effect of HRGP330 on tyrosine phosphorylation of protein tyrosine kinases, whereas Table 1B displays the tyrosine phosphorylation of signaling molecules functionally categorized as regulating cell adhesion/migration. The tyrosine phosphorylation level of the kinases in Table 1A showed moderate changes in response to HRGP330; the most marked increase was seen for Janus kinase 1, 2, and 3 (Jak1/2/3) and focal adhesion kinase (FAK). Among the cell adhesion/migration signaling molecules, increased tyrosine phosphorylation of the potential FAK substrates paxillin and Grb2 was detected. Tyrosine phosphorylation of the integrin β1 subunit was also increased. Fig. 1 shows selected areas of the arrays including spots representing reactivity for FAK, paxillin, and Grb2.

HRGP330 induces tyrosine phosphorylation of FAK and Grb2 Increased tyrosine phosphorylation detected by an antibody on the array could represent direct phosphorylation of the protein bound to the antibody or of an associated protein. To further investigate these possibilities and to validate the array data, we examined the phosphorylation status of Jaks, FAK, and integrin β1 in endothelial cells after HRGP330 stimulation. We have previously demonstrated a direct phosphorylation of paxillin after HRGP stimulation of endothelial cells [10], in agreement with the array data presented here. Tyrosine phosphorylation of FAK after HRGP330 stimulation was analyzed both in TIME cells and primary endothelial cells (bovine capillary endothelial; BCE) by immunoprecipitation of FAK and blotting with an anti-phosphotyrosine antibody (Fig. 2A). HRGP330-induced FAK phosphorylation was confirmed in primary BCE cells (Fig. 2A, left panel). The phosphorylation was sustained for at least 2 h in TIME cells after treatment with HRGP330, as demonstrated by Fig. 2A, right panel. We were not able to detect HRGP330-induced tyrosine phosphorylation of integrin β1 or Jak1 in endothelial cells (data not shown), indicating that the increased tyrosine phosphorylation signal detected on the antibody array was derived from molecules associated with these proteins. Immunoprecipitation with an anti-phosphotyrosine antibody and subsequent blotting for Grb2 showed increased amounts of Grb2 in the precipitates after HRGP or HRGP330 treatment of TIME cells for 10 min. This result confirms the array data and shows that Grb2 either binds to a protein which is phosphorylated after HRGP/ HRGP330 treatment, or is directly phosphorylated (Fig. 2B).

HRGP and HRGP330 increase the number of phospho-FAK containing focal adhesions FAK is a central mediator of integrin and growth factor signals and regulates the turnover of focal adhesions in cells attached to extracellular matrix [24]. We examined the subcellular distribution of tyrosine phosphorylated FAK in endothelial (TIME) cells after HRGP and HRGP330 treatment for 10 min. As

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

2551

HRGP330-induced formation of focal adhesions is connected to an increase in FAK tyrosine phosphorylation.

Geldanamycin inhibits HRGP/HRGP330-induced phosphorylation of FAK and restores the integrity of disrupted actin stress fibers

Fig. 2 – Tyrosine phosphorylation of FAK is induced by treatment with HRGP330. (A) BCE (left panel) or TIME (right panel) cells were treated with 100 ng/ml HRGP330 for 30 min or 2 h, respectively. FAK was immunoprecipitated (IP:FAK) from total cell lysates followed by immunoblotting with anti-phosphotyrosine antibody (IB:pY). Immunoblotting for FAK (IB:FAK) in the precipitated samples and for actin (IB:actin) in total cell lysates was performed to verify equal amount of protein in the samples. (B) TIME cells were treated with 100 ng/ml HRGP (H) or HRGP330 (P) for 10 min. Total cell lysates were immunoprecipitated with anti-phosphotyrosine antibody (IP:pY) and immunoblotted for Grb2 (IB:Grb2). Immunoblotting for actin (IB:actin) shows equal amount of protein in total cell lysates from the different treatments.

shown in Fig. 3, both treatments induced an increase in the number of phospho-FAK containing focal adhesions compared to control cells. Using paxillin as a marker, we have previously demonstrated an up-regulation of the number of focal adhesions after treatment of endothelial cells with HRGP [10] or HRGP330 [22]. The present data demonstrate that HRGP/

The actin cytoskeleton is anchored to focal adhesions, which function as cellular contact points where stress fibers communicate with the extracellular matrix to allow changes in response to integrin- and growth-factor-mediated signals [25]. Based on the effect of HRGP/HRGP330 on focal adhesion formation, we analyzed whether the actin cytoskeleton in endothelial cells was also affected. In TIME cells treated for 10 min with HRGP or HRGP330, the integrity of the cytoskeleton was disturbed as indicated by disruption of actin stress fibers, compared to vehicle-treated cells where the actin stress fibers were aligned in parallel bundles (Fig. 4A, left panels). To demonstrate a role for FAK in the cytoskeletal changes induced by HRGP/HRGP330, we used the FAK inhibitor geldanamycin, a benzoquinone ansamycin antibiotic, which has been shown to stimulate proteolysis of FAK. Endothelial cells were pre-incubated with geldanamycin for 1 h and then treated with HRGP or HRGP330 for 10 min. As shown in Fig. 4B (upper panel), geldanamycin efficiently attenuated the pool of tyrosine phosphorylated FAK in HRGP/ HRGP330-treated cells. During these conditions of geldanamycin treatment, total levels of FAK protein did not decrease (Fig. 4B, middle panel). However, incubation of the cells with the inhibitor for longer time periods also reduced FAK expression (data not shown). Co-treatment with HRGP/ HRGP330 and geldanamycin normalized the appearance of the actin stress fibers and restored their integrity (Fig. 4A, right panels). Fig. 4C shows a quantification of the number of cells with disrupted actin stress fibers under the conditions used in Figs. 4A and B. These data indicate that HRGP/ HRGP330-induced tyrosine phosphorylation and activation of FAK are connected to disruption of the actin cytoskeleton in endothelial cells.

Fig. 3 – HRGP and HRGP330 induce an increased number of phospho-FAK containing focal adhesions. TIME cells treated with HRGP (100 ng/ml) or HRGP330 (100 ng/ml) as indicated were fixed after 10 min and stained with anti-phospho-FAK antibody (red). Nuclei were stained with Hoechst 33342 (blue). Arrows indicate focal adhesions. Scale bar, 20 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

2552

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

HRGP and HRGP330 attenuate VEGF-A-induced tube formation in endothelial cells Endothelial cells may form tubular structures when provided with appropriate components of extracellular matrix proteins and growth factors in a 3-D context. The tube formation is a multi-step biological process involving cell–cell and cell– matrix adhesion, cell migration, cell fusion, and cell differen-

tiation [26]. We have previously shown that tubular morphogenesis is accompanied by growth arrest of the endothelial cells [27]. We asked whether HRGP/HRGP330 would affect the capacity of endothelial cells to form tubular structures. Endothelial cells were cultured between two layers of collagen for 24 h in the presence of VEGF-A, with or without HRGP or HRGP330. As shown in Fig. 5A, endothelial cells treated with VEGF-A fused to continuous vessel-like tubular structures. By contrast, co-treatment of VEGF-A with HRGP or HRGP330 efficiently inhibited tubular morphogenesis. Quantification of the total tube length under these different conditions was performed using a computer software as described in the Materials and methods section, and the results are displayed in Fig. 5B.

HRGP330 does not affect VEGF-induced proliferation of endothelial cells We have previously reported that HRGP does not affect proliferation or apoptosis of endothelial cells cultured on gelatin [10]. Since the inhibitory effect of HRGP on endothelial cell adhesion is most pronounced when using vitronectin as a matrix [10], we tested whether HRGP would inhibit VEGF-induced proliferation, in a matrixdependent manner. Starved TIME cells were stimulated for 22 h with 4 ng/ml VEGF-A, in the presence or absence of 100 ng/ml HRGP330. During the last 6 h of stimulation, 1 μCi/ml [methyl-3H] thymidine was included. As shown in Fig. 6, VEGF stimulated an approximately two-fold increase in [3 H] thymidine incorporation in cells cultured on vitronectin, while cells on gelatin displayed a more modest increase (1.3-fold). The presence of HRGP330 did not affect VEGF-induced S-phase entry of TIME cells, regardless of the matrix, demonstrating that inhibition of proliferation is not likely the mechanism of action of HRGP330. This result

Fig. 4 – HRGP and HRGP330 induce disruption of actin stress fibers, which is reverted by the FAK-inhibitor geldanamycin. (A) TIME cells on vitronectin-coated culture slides were pre-incubated with geldanamycin (3 nM) for 1 h before treatment with 100 ng/ml HRGP or HRGP330 for 10 min. Actin stress fibers were visualized by Texas Red-X phalloidin staining, and nuclei were stained with Hoechst 33342 (blue). The arrows indicate disrupted stress fiber organization. Scale bar, 25 μm. (B) Total cell lysates from TIME cells treated as in A were immunoprecipitated with anti-FAK antibody (IP:FAK) and blotted for anti-phosphotyrosine (IB:pY). C: control, H: HRGP, and P: HRGP330. Immunoblotting for FAK (IB:FAK) and for actin (IB:actin) in total cell lysates was performed to verify equal amount of protein in the samples. (C) Quantification of A. The number of cells with disrupted stress fibers was calculated from 1000 randomly selected cells for each treatment. Each bar represents the average number of cells in three independent experiments, ±SD. Differences were considered significant at P < 0.05 (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

2553

Fig. 5 – HRGP and HRGP330 interfere with tube formation of endothelial cells. (A) TIME cells were cultured in between two layers of collagen matrix and treated with VEGF-A only (50 ng/ml) or co-treated with VEGF-A (50 ng/ml) and HRGP (100 ng/ml) or VEGF-A (50 ng/ml) and HRGP330 (100 ng/ml) for 24 h, as indicated. Arrows show fused endothelial cells forming tubes. Scale bar, 100 μm. (B) Quantification of the total tube length for each treatment was performed using an image analysis software as described in the Materials and methods section. Data are presented as the mean tube length ± SEM from three independent experiments. Differences were considered significant at P < 0.05.

highlights the dependence on proper cell migration during endothelial tube formation.

Discussion HRGP and its minimal active domain HRGP330 interfere with cell adhesion in a manner implicating integrin αvβ3 as

important in its mechanism of action [21]. Moreover, HRGP and HRGP330 bind with high affinity to heparan sulfate, which is also of consequence for the ability of HRGP/ HRGP330 to exert its anti-angiogenic effect on endothelial cells [22]. Thus, these and ongoing studies of the cell surface structure mediating the anti-angiogenic effect of this inhibitor strongly suggest the involvement of more than one molecular component. However, the present study shows

2554

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

Fig. 6 – HRGP330 does not affect VEGF-induced proliferation of endothelial cells. Starved TIME cells cultured on either gelatin or vitronectin were treated with 4 ng/ml VEGF-A in the absence or presence of 100 ng/ml HRGP330 for 22 h. 1 μCi/ml [3H] thymidine was included during the last 6 h of stimulation. The amount of incorporated radioactivity was determined by scintillation counting, and data are presented as relative counts per minute (cpm) ± SD, where the value 1 represents the number of cpm in control samples. C = control, P = HRGP330, and V = VEGF-A.

that only a very narrow spectrum of intracellular signaling pathways is activated in HRGP330-treated endothelial cells, resulting in inhibition of actin reorganization. A signal transduction antibody array presenting 400 distinct antibodies to signal transduction molecules showed a marked specificity in HRGP330 signaling to a group of proteins involved in cell adhesion/migration, namely FAK and its substrates Grb2 and paxillin, as well as integrin β1. In addition to FAK, Jak1/2/3 displayed an increased phosphotyrosine signal among the tyrosine kinases. We have verified tyrosine phosphorylation of FAK and paxillin after HRGP/HRGP330 treatment of endothelial cells, but we have not been able to detect tyrosine phosphorylation of integrin β1 or Jak1. HRGP/HRGP330 also induced increased amounts of Grb2 protein in anti-phosphotyrosine immunoprecipitates from treated cells, indicating either direct phosphorylation of Grb2 or that it associates with a protein that is phosphorylated after HRGP/HRGP330 treatment, possibly FAK. Increased immunoreactivity for phosphorylated tyrosine on the antibody array could either reflect tyrosine phosphorylation of the protein directly bound to the antibody on the array or of an associated protein since the incubations were not performed under reducing conditions. We are currently investigating the phosphorylation status of Jak1/3 associated proteins after HRGP/HRGP330 stimulation of endothelial cells. FAK, Grb2, and paxillin may directly or indirectly associate with the integrin β1 subunit in focal contacts and could potentially be responsible for the phosphotyrosine signal detected at the site of the integrin β1 antibody. Whereas integrin β1 function is known to be regulated by threonine phosphorylation [28–30], there is no consensus concerning its ability to serve as a substrate also for tyrosine kinases. Further studies are needed to establish whether HRGP330-induced signal transduction in endothelial cells involves integrin β1.

FAK is a central regulator of cell motility and integrates signals from growth factor receptors and integrins. FAK is activated and autophosphorylated on tyrosine 397 in response to matrix-ligation of most integrins [31]. Phosphorylation of Y397 creates a docking site for a number of Src homology 2 (SH2) containing proteins such as Src, PLCγ, Shc, and the p85 subunit of phosphoinositide 3 kinase [24]. After binding to FAKY397, Src mediates phosphorylation of Y576/577 in the kinase domain of FAK, which is required for maximal kinase activity. Tyrosine 861 and 925 in the C-terminal part of FAK are also phosphorylated by Src, which creates binding sites for several additional proteins, among them p130Cas, Grb2, and paxillin. Phosphorylated FAK localizes to multiprotein complexes, focal adhesions, which connect the extracellular matrix to the actin cytoskeleton of cells. FAK has for long been considered a stimulator of cell migration [32,33], but data are accumulating showing that FAK may have a dual role in regulating motility of cells. Several inhibitors of angiogenesis have been reported to induce tyrosine phosphorylation of FAK, such as angiostatin [34], endostatin [35], thrombospondin [36,37] and endorepellin [38]. Like HRGP330, these molecules are also potent inhibitors of endothelial cell migration. HRGP and HRGP330 induce an increased number of focal adhesions [10,22], and we now show that phosphorylated FAK localize to these sites after HRGP330 treatment. FAK activity is however not essential for the formation of focal adhesions, as evidenced by the properties of FAK−/− fibroblasts [32] and endothelial cells [39]. Instead, FAK regulates the turnover and maturation of focal contacts [40,41]. It is important that the level of FAK activity is finely tuned for the cell to respond correctly to a migratory stimulus. Cells lacking the tyrosine phosphatase SHP2 (SH2-domain-containing protein tyrosine phosphatase 2) have hyperactive FAK and a high level of focal contact turnover [41]. In FAK−/− cells, focal contact turnover and maturation are inhibited [40]. However, SHP2−/− and FAK−/− cells both display an accumulation of immature focal contacts and migratory defects, demonstrating that too much or too little FAK activity has negative effects on cell motility. By which mechanism could FAK activation inhibit cell motility under certain circumstances and stimulate it in other situations? The answer may lie in the many phosphotyrosinedependent interactions that activated FAK may engage in, which differently regulate Rac1 activity. It has been suggested that FAK stimulates Rac1 activity and hence motility via interaction with p130Cas [42] while formation of FAK/paxillin complexes inhibits migration by lowering the levels of active Rac1 [43]. Indeed, paxillin overexpressing cells display reduced haptotactic migration towards collagen I [44]. In agreement with the above data, we could not detect any increase in tyrosine phosphorylation of p130Cas after HRGP330 treatment of endothelial cells (Table 1B). Both HRGP and HRGP330 treatment disturbed the actin stress fiber formation in endothelial cells. These effects were abrogated by the FAK inhibitor geldanamycin, indicating a role for FAK in HRGP/HRGP330-induced disruption of actin stress fibers. In addition to FAK, geldanamycin has also been reported to inhibit epidermal growth factor receptors, nuclear hormone receptors, and Hsp90 function. However, in the present study, geldanamycin restores the integrity of actin

E XP E RI ME N TA L CE L L RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

stress fibers disrupted by HRGP/HRGP330 treatment, demonstrating a “gain-of-function” phenotype and not a general toxicity. Geldanamycin is known to stimulate proteolysis of FAK, but the length of incubation with the inhibitor we employed only reduced phosphorylation of FAK and not total expression levels. We have previously reported that HRGP inhibits chemotaxis of endothelial cells, without affecting the rate of proliferation or apoptosis in vitro [10]. We now confirm and extend these data by showing that VEGF-induced proliferation of endothelial cells, cultured either on gelatin or vitronectin, is not affected by the presence of HRGP330. The present data showing inhibition of tube formation by HRGP330 implicate migration of endothelial cells as a crucial process for the formation of tube-like structures in this assay and further confirm the anti-angiogenic properties of this 35 amino acid residue peptide. In conclusion, our data implicate FAK as a mediator of HRGP330-induced inhibitory effects on endothelial cells. Thus, HRGP/HRGP330 exhibit a striking specificity in targeting focal adhesions, apparently leading to subversion of endothelial cell movement and eventually disturbed formation of new blood vessels.

[8]

[9]

[10]

[11]

[12] [13]

[14]

Acknowledgments [15]

Funding was provided by grants from the Swedish Research Council, Magnus Bergvalls Foundation and Erik, Karin and Gösta Selanders Foundation to AKO and from the Swedish Cancer Society (3820-B04-09XAC) and Innoventus Project AB to LCW. The study was in part supported by the sixth EU Framework Programme (Integrated Project ‘Angiotargeting; contract no 504743’) in the area of ‘Life sciences, genomics and biotechnology for health’.

REFERENCES

[1] P. Carmeliet, R.K. Jain, Angiogenesis in cancer and other diseases, Nature 407 (2000) 249–257. [2] M. Papetti, I.M. Herman, Mechanisms of normal and tumor-derived angiogenesis, Am. J. Physiol.: Cell Physiol. 282 (2002) C947–C970. [3] N. Ferrara, Vascular endothelial growth factor: basic science and clinical progress, Endocr. Rev. 25 (2004) 581–611. [4] M.S. O'Reilly, T. Boehm, Y. Shing, N. Fukai, G. Vasios, W.S. Lane, E. Flynn, J.R. Birkhead, B.R. Olsen, J. Folkman, Endostatin: an endogenous inhibitor of angiogenesis and tumor growth, Cell 88 (1997) 277–285. [5] T. Hawighorst, H. Oura, M. Streit, L. Janes, L. Nguyen, L.F. Brown, G. Oliver, D.G. Jackson, M. Detmar, Thrombospondin-1 selectively inhibits early-stage carcinogenesis and angiogenesis but not tumor lymphangiogenesis and lymphatic metastasis in transgenic mice, Oncogene 21 (2002) 7945–7956. [6] R.S. Watnick, Y.N. Cheng, A. Rangarajan, T.A. Ince, R.A. Weinberg, Ras modulates Myc activity to repress thrombospondin-1 expression and increase tumor angiogenesis, Cancer Cell 3 (2003) 219–231. [7] T. Hawighorst, P. Velasco, M. Streit, Y.K. Hong, T.R. Kyriakides, L.F. Brown, P. Bornstein, M. Detmar, Thrombospondin-2 plays a protective role in multistep

[16]

[17] [18]

[19]

[20]

[21]

[22]

[23]

[24]

2555

carcinogenesis: a novel host anti-tumor defense mechanism, EMBO J. 20 (2001) 2631–2640. M. Sund, Y. Hamano, H. Sugimoto, A. Sudhakar, M. Soubasakos, U. Yerramalla, L.E. Benjamin, J. Lawler, M. Kieran, A. Shah, R. Kalluri, Function of endogenous inhibitors of angiogenesis as endothelium-specific tumor suppressors, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 2934–2939. J.C. Juarez, X. Guan, N.V. Shipulina, M.L. Plunkett, G.C. Parry, D.E. Shaw, J.C. Zhang, S.A. Rabbani, K.R. McCrae, A.P. Mazar, W.T. Morgan, F. Donate, Histidine–proline-rich glycoprotein has potent antiangiogenic activity mediated through the histidine–proline-rich domain, Cancer Res. 62 (2002) 5344–5350. A.K. Olsson, H. Larsson, J. Dixelius, I. Johansson, C. Lee, C. Oellig, I. Björk, L. Claesson-Welsh, A fragment of histidine-rich glycoprotein is a potent inhibitor of tumor vascularization, Cancer Res. 64 (2004) 599–605. A.L. Jones, M.D. Hulett, C.R. Parish, Histidine-rich glycoprotein: a novel adaptor protein in plasma that modulates the immune, vascular and coagulation systems, Immunol. Cell Biol. 83 (2005) 106–118. P. Tompa, Intrinsically unstructured proteins, Trends Biochem. Sci. 27 (2002) 527–533. W.T. Morgan, The histidine-rich glycoprotein of serum has a domain rich in histidine, proline, and glycine that binds heme and metals, Biochemistry 24 (1985) 1496–1501. B.A. Kluszynski, C. Kim, W.P. Faulk, Zinc as a cofactor for heparin neutralization by histidine-rich glycoprotein, J. Biol. Chem. 272 (1997) 13541–13547. H.R. Lijnen, D. Collen, Interaction of heparin with histidine-rich glycoprotein, Ann. N. Y. Acad. Sci. 556 (1989) 181–185. A.L. Jones, M.D. Hulett, C.R. Parish, Histidine-rich glycoprotein binds to cell-surface heparan sulfate via its N-terminal domain following Zn2+ chelation, J. Biol. Chem. 279 (2004) 30114–30122. L.L. Leung, Interaction of histidine-rich glycoprotein with fibrinogen and fibrin, J. Clin. Invest. 77 (1986) 1305–1311. N.N. Gorgani, C.R. Parish, S.B. Easterbrook Smith, J.G. Altin, Histidine-rich glycoprotein binds to human IgG and C1q and inhibits the formation of insoluble immune complexes, Biochemistry 36 (1997) 6653–6662. N. Tsuchida-Straeten, S. Ensslen, C. Schafer, M. Woltje, B. Denecke, M. Moser, S. Graber, S. Wakabayashi, T. Koide, W. Jahnen-Dechent, Enhanced blood coagulation and fibrinolysis in mice lacking histidine-rich glycoprotein (HRG), J. Thromb. Haemost. 3 (2005) 865–872. F. Donate, J.C. Juarez, X. Guan, N.V. Shipulina, M.L. Plunkett, Z. Tel-Tsur, D.E. Shaw, W.T. Morgan, A.P. Mazar, Peptides derived from the histidine–proline domain of the histidine–proline-rich glycoprotein bind to tropomyosin and have antiangiogenic and antitumor activities, Cancer Res. 64 (2004) 5812–5817. J. Dixelius, A.K. Olsson, Å. Thulin, C. Lee, I. Johansson, L. Claesson-Welsh, Minimal active domain and mechanism of action of the angiogenesis inhibitor histidine-rich glycoprotein, Cancer Res. 66 (2006) 2089–2097. M. Vanwildemeersch, A.K. Olsson, E. Gottfridsson, L. Claesson-Welsh, U. Lindahl, D. Spillmann, Angiogenesis inhibition by histidine-rich glycoprotein involves Zn2+-dependent binding of heparan sulfate to the His/Pro-rich domain, J. Biol. Chem. 281 (2006) 10298–10304. E. Venetsanakos, A. Mirza, C. Fanton, S.R. Romanov, T. Tlsty, M. McMahon, Induction of tubulogenesis in telomerase-immortalized human microvascular endothelial cells by glioblastoma cells, Exp. Cell Res. 273 (2002) 21–33. S.K. Mitra, D.A. Hanson, D.D. Schlaepfer, Focal adhesion kinase: in command and control of cell motility, Nat. Rev., Mol. Cell Biol. 6 (2005) 56–68.

2556

E XP E RI ME N TA L CE LL RE S E A RCH 3 1 2 ( 2 00 6 ) 2 5 4 7 –25 5 6

[25] C. Brakebusch, R. Fassler, The integrin—Actin connection, an eternal love affair, EMBO J. 22 (2003) 2324–2333. [26] C.J. Drake, J.E. Hungerford, C.D. Little, Morphogenesis of the first blood vessels, Ann. N. Y. Acad. Sci. 857 (1998) 155–179. [27] T. Matsumoto, I. Turesson, M. Book, P. Gerwins, L. Claesson-Welsh, p38 MAP kinase negatively regulates endothelial cell survival, proliferation, and differentiation in FGF-2-stimulated angiogenesis, J. Cell Biol. 156 (2002) 149–160. [28] K. Wennerberg, R. Fassler, B. Warmegard, S. Johansson, Mutational analysis of the potential phosphorylation sites in the cytoplasmic domain of integrin beta1A. Requirement for threonines 788–789 in receptor activation, J. Cell Sci. 111 (1998) 1117–1126. [29] K. Suzuki, K. Takahashi, Reduced cell adhesion during mitosis by threonine phosphorylation of beta1 integrin, J. Cell. Physiol. 197 (2003) 297–305. [30] S.M. Kim, M.S. Kwon, C.S. Park, K.R. Choi, J.S. Chun, J. Ahn, W.K. Song, Modulation of Thr phosphorylation of integrin beta1 during muscle differentiation, J. Biol. Chem. 279 (2004) 7082–7090. [31] S.K. Hanks, L. Ryzhova, N.Y. Shin, J. Brabek, Focal adhesion kinase signaling activities and their implications in the control of cell survival and motility, Front. Biosci. 8 (2003) 982–996. [32] D. Ilic, Y. Furuta, S. Kanazawa, N. Takeda, K. Sobue, N. Nakatsuji, S. Nomura, J. Fujimoto, M. Okada, T. Yamamoto, Reduced cell motility and enhanced focal adhesion contact formation in cells from FAK-deficient mice, Nature 377 (1995) 539–544. [33] C.R. Hauck, D.A. Hsia, D.D. Schlaepfer, The focal adhesion kinase—A regulator of cell migration and invasion, IUBMB Life 53 (2002) 115–119. [34] L. Claesson-Welsh, M. Welsh, N. Ito, B. Anand-Apte, S. Soker, B. Zetter, M. O'Reilly, J. Folkman, Angiostatin induces endothelial cell apoptosis and activation of focal adhesion kinase independently of the integrin-binding motif RGD, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 5579–5583. [35] J. Dixelius, M. Cross, T. Matsumoto, T. Sasaki, R. Timpl, L. Claesson-Welsh, Endostatin regulates endothelial cell adhesion and cytoskeletal organization, Cancer Res. 62 (2002) 1944–1947.

[36] J.S. Lymn, S.J. Rao, G.F. Clunn, K.L. Gallagher, C. O'Neil, N.T. Thompson, A.D. Hughes, Phosphatidylinositol 3-kinase and focal adhesion kinase are early signals in the growth factor-like responses to thrombospondin-1 seen in human vascular smooth muscle, Arterioscler. Thromb. Vasc. Biol. 19 (1999) 2133–2140. [37] V. Gahtan, X.J. Wang, M. Ikeda, A.I. Willis, G.P. Tuszynski, B.E. Sumpio, Thrombospondin-1 induces activation of focal adhesion kinase in vascular smooth muscle cells, J. Vasc. Surg. 29 (1999) 1031–1036. [38] G. Bix, J. Fu, E.M. Gonzalez, L. Macro, A. Barker, S. Campbell, M.M. Zutter, S.A. Santoro, J.K. Kim, M. Hook, C.C. Reed, R.V. Iozzo, Endorepellin causes endothelial cell disassembly of actin cytoskeleton and focal adhesions through alpha2beta1 integrin, J. Cell Biol. 166 (2004) 97–109. [39] T.L. Shen, A.Y. Park, A. Alcaraz, X. Peng, I. Jang, P. Koni, R.A. Flavell, H. Gu, J.L. Guan, Conditional knockout of focal adhesion kinase in endothelial cells reveals its role in angiogenesis and vascular development in late embryogenesis, J. Cell Biol. 169 (2005) 941–952. [40] X.D. Ren, W.B. Kiosses, D.J. Sieg, C.A. Otey, D.D. Schlaepfer, M.A. Schwartz, Focal adhesion kinase suppresses Rho activity to promote focal adhesion turnover, J. Cell Sci. 113 (2000) 3673–3678. [41] G. von Wichert, B. Haimovich, G.S. Feng, M.P. Sheetz, Force-dependent integrin–cytoskeleton linkage formation requires downregulation of focal complex dynamics by Shp2, EMBO J. 22 (2003) 5023–5035. [42] D.A. Hsia, S.K. Mitra, C.R. Hauck, D.N. Streblow, J.A. Nelson, D. Ilic, S. Huang, E. Li, G.R. Nemerow, J. Leng, K.S. Spencer, D.A. Cheresh, D.D. Schlaepfer, Differential regulation of cell motility and invasion by FAK, J. Cell Biol. 160 (2003) 753–767. [43] H. Yano, Y. Mazaki, K. Kurokawa, S.K. Hanks, M. Matsuda, H. Sabe, Roles played by a subset of integrin signaling molecules in cadherin-based cell–cell adhesion, J. Cell Biol. 166 (2004) 283–295. [44] H. Yano, H. Uchida, T. Iwasaki, M. Mukai, H. Akedo, K. Nakamura, S. Hashimoto, H. Sabe, Paxillin alpha and Crk-associated substrate exert opposing effects on cell migration and contact inhibition of growth through tyrosine phosphorylation, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 9076–9081.