Cancer Research Prevention Award Metabolism
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bossard, C.
Right arrow Articles by Prats, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bossard, C.
Right arrow Articles by Prats, H.
[Cancer Research 64, 7507-7512, October 15, 2004]
© 2004 American Association for Cancer Research

Regular Articles

Antiangiogenic Properties of Fibstatin, an Extracellular FGF-2–Binding Polypeptide

Carine Bossard1, Loic Van den Berghe1, Henrik Laurell1, Caroline Castano1, Martine Cerutti2, Anne-Catherine Prats1 and Hervé Prats1

1 Institut National de la Santé et de la Recherche Médicale (INSERM) U589, C.H.U. Rangueil, Toulouse, France; and 2 INRA-CNRS UMR Saint Christol Lès Alès, France


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
By using the two-hybrid system with basic fibroblast growth factor (FGF-2) as bait, we isolated and characterized fibstatin, an endogenous Mr 29,000 human basement membrane-derived inhibitor of angiogenesis and tumor growth. Fibstatin, a fragment containing the type III domains 12–14 of fibronectin, was produced as a recombinant protein and was shown to inhibit the proliferation, migration, and differentiation of endothelial cells in vitro. Antiangiogenic activity of fibstatin was confirmed in a Matrigel angiogenesis assay in vivo, and electrotransfer of the fibstatin gene into muscle tissue resulted in reduced B16F10 tumor growth. Taken together, these results suggest that fibstatin could act as a powerful molecule for antiangiogenic therapy.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Angiogenesis, the outgrowth of new capillaries from preexisting vessels, is not only essential for a number of physiologic processes but also occurs in many pathological conditions including tumor growth. This phenomenon is tightly controlled by numerous angiogenic factors the effects of which are counterbalanced by inhibitory molecules such as endostatin and angiostatin, which are proteolytic polypeptides derived from collagen XVIII and plasminogen, respectively (1 , 2) . Among the angiogenic growth factors, fibroblast growth factor 2 (FGF-2) is one of the most efficient (3) .

It can exert its activity both as an endogenous (intracrine) and an exogenous (auto-/paracrine) factor through different pathways. Exogenous FGF-2 act via cell surface receptors linked to conventional signal transduction pathways, but also through direct association with the nuclei of target cells after internalization and nuclear translocation during the G1 phase of the cell cycle (4) . We have recently shown that this nuclear translocation is necessary for the mitogenic activity of FGF-2 and is mediated by Translokin, isolated by the two-hybrid system with FGF-2 Mr 18,000 as bait (5) .

In the same screening, we isolated two independent clones encoding fragments of fibronectin that also showed a strong interaction with FGF-2. It has been previously reported that the NH2- and COOH-terminal heparin-binding domains of human plasma fibronectin appear to be rather specific potent inhibitors of bovine aortic endothelial cell growth in culture by an unknown mechanism, whereas the plasma fibronectin itself is much less inhibitory (6) . We thus evaluated whether the shortest of the two FGF-2 interacting fragments could counteract the angiogenic properties of FGF-2. Because of its antiangiogenic activity, we named the fragment fibstatin.

Because angiogenesis represents a cascade of cellular processes that includes endothelial cell proliferation, migration, and tube formation (7) , we used multiple in vitro and in vivo assays related to angiogenesis to confirm the bioactivity of fibstatin.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Yeast Two-Hybrid System
The two-hybrid expression-cloning system was carried out as described by Van den Berghe et al. (8) . Briefly, the Mr 18,000 FGF-2 was used as bait, and the corresponding cDNA was inserted in the pAS2 vector. A human placenta library (Clontech, Palo Alto, CA) cloned into the pACT2 vector was screened in Saccharomyces cerevisiae strain Y190. ß-galactosidase activity of colonies growing on selection medium (DOBArg/-Trp/-Leu) was assessed by blue/white coloring (X-gal assay) and in yeast extracts with Galacto-Light (Tropix Inc. Bedford, MA) as described by the manufacturer. The activity was normalized for protein contents.

Plasmids Encoding Fibstatin or Endostatin
The coding sequence of fibstatin was amplified by PCR from the pACT2 plasmid containing the fibstatin cDNA isolated by the two-hybrid screening. The coding sequence of endostatin was amplified by reverse transcription-PCR from RNA extracted from NIH-3T3 cells. The two cDNAs were then ligated into pET-15b (Novagen, Madison, WI) in frame with the histidine tag for recombinant protein production in Escherichia coli or cloned in the eukaryotic expression vectors, pUHD10–3 (9) and pSCT (10) in frame with the signal peptide of vascular endothelial growth factor A (VEGF-A; ref.11 ) and the HA epitope, for use in conditioned medium preparation and electrotransfer, respectively. For the structure-functional analysis of the FGF2-fibstatin complex, given regions of fibstatin were amplified by PCR, respecting the limits of each type-III module (Fig. 1)Citation . The amplified products were cloned in frame with the transactivating domain of Gal4 in pACT2 vector. All of the sequences were confirmed by automatic sequencing with the ABI Prism Dye Terminator kit (Applied Biosystems, Foster City, CA). Cloning details can be provided on request.



View larger version (39K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. Characterization of fibstatin, an FGF-2 interacting protein. A, domain structure of human fibronectin and relative position of the two FGF-2 binding clones (Clone A and Clone B) isolated by the two-hybrid system. B, amino acid sequence of the two FGF-2 binding clones (bold letters, A and B). Arrows, the extremities of each clone. Italics, the type III (12) domain; bold, the type III (13) domain; underlined, the type III (14) domain. (Hep, heparin-binding domain.)

 
Recombinant Protein Expression
Using E. coli BL21 (DE3).
The production and purification of recombinant proteins were performed with Ni2+-NTA beads (Qiagen, Courtaboeuf, France) as described by Holzinger et al. (12) .

Using Baculovirus.
The coding sequences for fibstatin and endostatin were subcloned into the transfer vector p119 designed for recombination in the P10 locus as described by Marchal et al. (13) . Briefly, by using DOTAP (Roche), Sf9 cells were cotransfected with p119-Fib or p119-Endo and purified viral DNA of baculovirus AcSLP10 expressing the polyhedrin gene under the control of the P10 promoter (14) . The screening and purification of the recombinant virus were carried out as reported by Summers and Smith (15) . Fibstatin and endostatin were purified from the serum- free supernatant of Sf9 cells. The supernatant (1 liter) was slowly mixed with ammonium sulfate to obtain a final concentration of 50%. The ammonium sulfate/supernatant solution was stirred 2 hours at 4°C and was centrifuged at 5,000 rpm for 30 minutes. The precipitated proteins were resuspended in 0.1 starting volumes of PBS. The solution was dialyzed against PBS and the proteins were purified by using Ni2+-NTA beads according to the manufacturer (Qiagen).

Cell Culture
NIH-3T3, HeLa HT (16) , and ABAE cells were grown in DMEM (Invitrogen, Cergy Pontoise, France) containing antibiotics, 1% glutamine, and 10% fetal calf serum or 10% calf serum for ABAE. B16F10 cells were grown in RPMI 1640 (Biowhittaker, Emerainville, France) containing antibiotics, 1% glutamine, and 10% fetal calf serum. Cells were maintained in a 37°C and 10% CO2 (for ABAE) or 5% CO2 (for NIH-3T3, HeLa, and B16F10 cells) humidified incubator.

In vitro Transfection and Conditioned Media Preparation
HeLa HT cells, seeded at 1.2 x 106/100-mm tissue culture dish, were transfected with 17 µg of plasmid pUHD-fibstatin, pUHD-endostatin, or "empty" pUHD by using Fugene 6 reagent according to the manufacturer’s recommendations (Roche Diagnostics, Mannheim, Germany). Twenty-four hours after the transfection, subconfluent cells were washed with DMEM and were incubated in DMEM supplemented with 1% fetal calf serum. Twenty-four hours later, cell-conditioned medium was collected, centrifuged to remove cell debris, and stored at –80°C until use.

Proliferation Assay
Using Conditioned Medium.
ABAE cells were seeded at 5 x 104/mL of medium. Twenty-four hours later, cells were washed with DMEM and were incubated with different dilutions of transfected-HeLa cells–conditioned medium so that the final calf serum concentration was 10%. The proliferation was stimulated with 0.5 ng/mL FGF-2. The medium was changed twice, and after 72 hours, cells were counted with a Coulter counter ZM (Beckman Coulter, Villepinte, France).

Using Recombinant Protein.
ABAE cells, seeded at 5 x 103/mL, were stimulated by 0.5 ng/mL FGF-2 with different concentrations of fibstatin or endostatin added at days 1 and 3. At day 5, cells were trypsinized and counted with a Coulter counter ZM.

Cell Migration Assay
ABAE cells were grown to confluence in complete medium. The cells were then incubated in serum-free DMEM for 48 hours before the experiment. A scratch injury was applied on the cell monolayer with a plastic pipette tip to generate a wound. The debris were removed by washing the cells with PBS, and the cells were incubated for 14 hours in serum-free DMEM supplemented with FGF-2 (3 ng/mL) in the presence or not (control) of fibstatin (1 µg/mL). Migrating cells were then counted under microscope.

In vitro Angiogenesis Assay
Twenty-four-well plates were coated with 300 µL of Growth Factor Reduced Matrigel (B&D, Bedford, MA) per well, which was allowed to polymerize for 1 hour at 37°C. ABAE cells (2 x 105 cells/mL) suspended in 500 µL of culture medium were then added to each well. FGF-2 (0.5 ng/mL) with or without (control) fibstatin (1 µg/mL) was added to the wells, and the plates were incubated overnight at 37°C. After removal of the medium, the culture was fixed, and the length of the tube network was quantified with the Q Win Leica system (Leica Microsystems, Rueil-Malmaison, France).

Animals Studies
Six-to-10-week-old C57Bl/6 mice were housed in stainless-steel cages in groups of five, were kept in a temperature-controlled facility on a 12-hour light-dark cycle, and were fed normal laboratory mouse chow diet. All of the procedures were performed in accordance with the recommendations of the European Accreditation of Laboratory Animal Care.

In vivo Angiogenesis Assay
C57Bl/6 female mice (8 to 10 weeks old) received a subcutaneous injection (0.3 mL) of Growth Factor Reduced Phenol-red-free Matrigel (BD Biosciences, Bedford, MA) supplemented with FGF-2 (100 ng/mL) alone (control) or in the presence of fibstatin or endostatin (10 µg/mL). Matrigel plugs were removed 7 days later, were weighed, and were dissolved in 1.5 mL of Matrisperse cell release solution (BD Biosciences) during 2 hours at 4°C. After dispersion from the gel, microvascular endothelial cells were counted under a microscope with a Malassez cell, and the cell number was normalized or not by the weight of the Matrigel. The phenotype of the cells was verified by the staining with anti-CD31 (PharMingen, San Diego, CA).

Electrotransfer of Plasmid Encoding Fibstatin or Endostatin into Quadriceps of Mice Bearing B16F10 Melanoma
Fifteen mice were used in each control or experimental group. The mice were anesthetized with ketamin (150 mg/kg) and subjected to a dorsal subcutaneous injection of 1 x 106 B16F10 cells in 100 µL PBS– (i.e., standard PBS without Ca2+ and Mg2+). At the same time, 50 µg of pSCT DNA plasmid, encoding or not fibstatin (pSCT Fib) or endostatin (pSCT Endo), were mixed with 5 µg of DNA plasmid encoding the ß-galactosidase and were injected in 50 µL of PBS– into the quadriceps. Only endotoxin-free plasmids were used (Endofree Plasmid Maxi kit, Qiagen). The quadriceps were then entirely covered with ultrasound gel and placed between 10-mm-diameter caliper-type electrodes set 4 mm apart, and were electroporated with 8 pulses (80 V, 20 milliseconds) from an ElectroSquare Porator ECM830 (BTX, a division of Genetronics, San Diego, CA). Electroporation was repeated 5 days later. Tumors were excised and weighed 10 days after the first injection. P < 0.05 was considered as statistically significant with a one-factor ANOVA. Endostatin concentration was assayed by ELISA (Accucyte Murine Endostatin Immunoassay kit Cytimmune Science, College Park, MD). For determination of ß-galactosidase activity, the electroporated quadriceps were excised and placed in 350 µL of ß-galactosidase lysis solution (Tropix) and homogenized. The homogenate was centrifuged (13,000 rpm, 10 minutes) and 20 µL of supernatant were used to measure the ß-galactosidase activity with the commercial kit Galacto Light (Tropix).

CD31 Immunostaining
Intratumoral microvessel density was analyzed on frozen sections (5 µm) of B16 melanoma with a rat antimouse CD31 monoclonal antibody (PharMingen, San Diego, CA). After fixation in 4% paraformaldehyde for 1 hour at 4°C, the sections were washed three times with PBS and were permeabilized with PBS-0.25% Triton X100–3% bovine serum albumin for 1 hour. After blocking with 5% rabbit serum/PBS-Tween 0.1% for 1 hour at room temperature, the sections were incubated overnight with the antimouse CD31 antibody, diluted 1:50 in antibody diluent solution (DAKO, Carpinteria, CA). The sections were washed three times with PBS and incubated with biotinylated antirat IgG (Vector Labs, Burlingame, CA) diluted 1:200 in antibody diluent solution during 2 hours at room temperature. After three washings with PBS, a Streptavidin Alexa Fluor 488 conjugate (Molecular Probes, Eugene, OR), diluted 1:200 in antibody diluant solution, was used for detection of bound antibodies. The Visiolab 2000 software (Biocom, Paris, France) was used to quantify the anti-CD31–labeled surface.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Isolation of a Novel FGF-2 Interacting Protein.
In a two-hybrid screening, we isolated two independent clones that interacted strongly with FGF-2 (Fig. 1A)Citation . These two clones named clone A and clone B, contain the heparin-binding domain 2 (Hep2) of fibronectin and mainly differ in the length of their NH2 extremities (Fig. 1B)Citation . After a more substantial characterization of the properties of the shortest clone B (see below), which overlaps with the three type-III modules (FNIII 12–14) of fibronectin, we named the corresponding peptide fibstatin.

To test the specificity of the observed interaction between fibstatin and FGF-2, we coexpressed fibstatin in yeast with a panel of representative FGF-family members (FGF-1, FGF-3, FGF-6, FGF-12). Among the FGFs tested, only FGF-2 was able to interact with fibstatin (Fig. 2A)Citation . To ascertain that the fibstatin/FGF-2 interaction was direct and independent of a third partner, we expressed the two proteins in E. coli, purified them, and tested their interaction in an in vitro ELISA assay, as described by Bossard et al. (5) . Fibstatin and FGF-2 interact strongly in a direct and dose-dependent manner (Fig. 2B)Citation . We next determined which of the three type-III modules (FNIII 12–14) in fibstatin, contained the FGF-2 interacting site. Different portions of fibstatin were subcloned in pACT2, and their capacity to interact with FGF-2 was tested in the two-hybrid system. We found that both FNIII 13 and 14 are necessary to maintain a similar level of binding as full-length fibstatin (Fig. 2C)Citation . When tested separately, the two modules were not able to interact with FGF-2, except for a residual interaction (10%) for FNIII 14. These results suggest that two structural motifs in fibstatin, the type-III modules 13 and 14, cooperate in the binding to FGF-2.



View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 2. Specific and direct interaction between fibstatin and FGF-2. A, specificity of the interaction between fibstatin and FGF-2. Several FGFs and fibstatin were fused to DNA binding (DB) or activating (AD) domains of Gal4 and subjected to a yeast two-hybrid assay to evaluate the degree of interaction. B, direct interaction between FGF-2 and fibstatin by ELISA. FGF-2 and bovine serum albumin were coated on a 96-well plate and were incubated with different concentration of His-tagged recombinant fibstatin. The association was revealed by using a horseradish-peroxidase–conjugated antibody against the (His)6 tag and was quantified at 490 nm in an ELISA microplate reader. Binding was calculated by the subtraction of nonspecific binding obtained with bovine serum albumin. Data are the mean ± SD of a representative triplicate sample. C, analysis of structural elements of fibstatin involved in the interaction with FGF-2. The capacity of modules type III 12 to 14, alone or together, to interact with FGF-2 was assessed in the yeast two-hybrid system. Horizontal bars, the ß-galactosidase activity ± SD in yeast extracts from three independent transformations.

 
Fibstatin Inhibits Endothelial Cells Proliferation.
Because FGF-2 is a strong inducer of endothelial cell proliferation, fibstatin was tested for its activity on FGF-2–induced proliferation of ABAE cells and was compared with the inhibitory activity of endostatin. For this purpose, we prepared conditioned medium from transiently transfected HeLa HT. The presence of fibstatin or endostatin, which have an apparent molecular weight of 29,000 and 20,000, respectively, was confirmed by SDS-PAGE (data not shown). The conditioned medium was then applied at different dilutions to FGF-2–stimulated ABAE cells, in a 72-hour proliferation assay. Fibstatin and endostatin, expressed in the conditioned media, inhibited, in a dose-dependent manner, the ABAE proliferation induced by FGF-2 (Fig. 3A)Citation . A similar inhibitory effect on ABAE cells was observed when recombinant fibstatin produced in E. coli was used in the above assay (Fig. 3B)Citation . Interestingly, the inhibitory effect of fibstatin was greater, compared with that of endostatin, at every dose tested. On the contrary, no significant inhibition of proliferation was observed when cells of nonendothelial origin, such as fibroblasts NIH-3T3 or B16F10, were tested (Fig. 3C)Citation .



View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 3. Fibstatin inhibits endothelial cell proliferation. A, inhibition of bovine aortic endothelial cell proliferation by conditioned medium from HeLa cells expressing or not (Control) fibstatin or endostatin. Diluted conditioned medium [one-plus-one volume (1:2) or three-plus-one volumes (3:4)], supplemented with serum and 0.5 ng/mL FGF-2, was applied to bovine aortic endothelial cells in a 72-hour proliferation assay. Data are expressed as percentage of the proliferation of FGF-2–stimulated control cells ± SD. B, inhibition of aortic endothelial cell proliferation by recombinant fibstatin or endostatin. Different concentrations of recombinant fibstatin or endostatin were applied to ABAE cells stimulated with 0.5 ng/mL of FGF-2 at days 1 and 3. At day 5, cells were trypsinized and were counted with a Coulter counter ZM. Data are expressed as percentage of the proliferation of FGF-2–stimulated control cells ± SD. C, effect of recombinant fibstatin on non-endothelial cell proliferation. One µg/mL of recombinant fibstatin was applied to B16F10 or NIH-3T3 cells, stimulated with 1 ng/mL of FGF-2. No significant inhibition of proliferation was observed at day 5.

 
Fibstatin Inhibits Endothelial Cell Migration and Tubulogenesis.
To test the ability of recombinant fibstatin to inhibit endothelial cell migration induced by FGF-2, we used an in vitro wounding assay. The addition of 1 µg/mL fibstatin greatly reduced the number of migrating ABAE cells on FGF-2 stimulation (Fig. 4A)Citation . To determine the in vitro antiangiogenic effects of fibstatin, we tested its ability to interfere with endothelial cell tube formation in Matrigel, a solid gel of basement membrane proteins derived from mouse sarcoma. When FGF-2–stimulated ABAE cells were cultured on Matrigel, they rapidly aligned and formed hollow tube-like structures. In the presence of fibstatin produced in E. coli, the tube formation induced by FGF-2 was inhibited, and the cells appeared clustered, as in the absence of FGF-2 (Fig. 4B)Citation .



View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 4. Fibstatin inhibits endothelial cell migration and tubulogenesis A, Fibstatin effect on endothelial cell migration, tested in an in vitro wounding assay. Serum-starved cells were stimulated with 3 ng/mL FGF-2 in the presence or not (Control) of 1 µg/mL fibstatin. After 14 hours, quantification of migrating cells into the wound area was determined under an inverted microscope. The data shown are the mean of a representative duplicate experiment. B, tubulogenesis assay. ABAE cells were plated on Matrigel-coated wells in the presence or not of 0.5 ng/mL FGF-2 with or without 1 µg/mL fibstatin during 12 hours. Tube formation was quantified with a Leica Q Win system. The data shown are the means of a representative duplicate experiment. (AU, arbitrary unit.)

 
In vivo Antiangiogenic Activities of Fibstatin.
We then tested the ability of recombinant fibstatin to inhibit in vivo angiogenesis. For this purpose, Matrigel plugs containing FGF-2 with or without baculovirus-expressed fibstatin or endostatin were injected subcutaneously in C57Bl/6 mice. One week after implant, the endothelial cells that had colonized the plugs were counted. Under these conditions, more than 90% of the cells in the Matrigel were positive for the endothelial marker CD 31 (data not shown). The number of cells within the plugs containing both FGF-2 and fibstatin was 50%, compared with the control containing only FGF-2. Interestingly, fibstatin was more potent than endostatin in inhibiting the angiogenic activity of FGF-2 (Fig. 5A)Citation . One could argue that this result was due to an effect of fibstatin on matrix-degrading protease, provoking Matrigel degradation and the apparent decrease in cell number per unit Matrigel weight. We then considered the cell numbers without normalizing them. As shown in Fig. 4ACitation , even if the inhibition is lower in this condition (40%), it is still substantial, consistent with an antiangiogenic activity of fibstatin.



View larger version (34K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 5. Fibstatin inhibits angiogenesis and tumor growth in vivo. A, angiogenesis assay in Matrigel. Quantification of endothelial cells from Matrigel plugs injected subcutaneously in C57Bl6 mice, with FGF-2 alone (Control), or in the presence of fibstatin (Fib) or endostatin (Endo). FGF-2–induced angiogenesis was significantly inhibited by fibstatin compared with endostatin. Data are expressed as percentage ± SE, of the cell number obtained with FGF-2 alone, normalized (Norm; left, +) or not (right, –) with respect to the weight of the Matrigel plug. Each experimental group contained five mice, and the experiment was repeated twice with similar results. *, P < 0.05 was considered statistically significant by using a one-factor ANOVA. B, inhibition of B16F10 melanoma growth after gene electrotransfer of fibstatin or endostatin. C57Bl6 mice were inoculated subcutaneously with a suspension of B16F10 cells. Just after inoculation, therapy by electrotransfer was initiated. Each experimental group contained 15 animals. Data represent the average tumor mass (± SE) evaluated 10 days after the B16F10 inoculation. The difference between the Fib and the control group were statistically significant. *, P < 0.05 was considered statistically significant by using a one-factor ANOVA. C, reporter gene (LacZ) activity after electrotransfer into quadriceps. Ten days after the first electroporation, quadriceps were excised and ß-galactosidase activity was determined in tissue homogenates with suitable commercial kit Galacto Light (Tropix). Bars, the average ß-galactosidase (ß-Gal) activity ± SD per 1 mg of muscle from 15 mice. D, Endostatin concentrations in serum after electroporation of plasmid DNA containing or not (Control) endostatin gene, into quadriceps. Serum concentration of endostatin was measured 10 days after the first injection with a commercial kit (Accucyte Murine endostatin Immunoassay kit, Cytimmune Science). Each group contained eight animals.

 
Electrotransfer of Fibstatin-Encoding cDNA into Muscles Inhibits B16F10 Tumor Growth.
We then evaluated the in vivo effect of fibstatin on tumor growth after DNA electrotransfer. Fibstatin- or endostatin-encoding plasmid DNA was mixed with a reporter plasmid encoding LacZ and was injected and electroporated into quadriceps of C57Bl/6 mice, previously inoculated with B16F10 cells. The electroporation was repeated 5 days later. In the group of mice treated with the fibstatin-coding gene, there was a 50% inhibition of tumor growth compared with control mice that received injections of empty vector. This inhibition was observed at day 10 (Fig. 5B)Citation and also at day 14 (data not shown). Tumor mass in the endostatin-treated group was reduced, but the difference did not reach statistical significance (Fig. 5B)Citation . To assess the efficiency of the gene transfer, the ß-galactosidase activity was measured in muscle lysate. The results, shown in Fig. 5CCitation , reveal a significant ß-galactosidase activity in all of the electroporated mice.

The circulating endostatin level was determined by the use of a commercially available ELISA kit. Endostatin level in the blood of mice electrotransferred with a plasmid that encoded endostatin was markedly higher than in the blood of control mice (Fig. 5D)Citation . To detect the level of circulating fibstatin, we developed a rabbit polyclonal antibody that specifically recognized the protein but was insufficient to detect low concentrations of the protein (data not shown). Even though we were unable to quantify the amount of circulating fibstatin, the efficiency of the electrotransfer was ascertained by the quantification of the ß-galactosidase activity in the transfected muscles (Fig. 5C)Citation . Moreover, the transgene expression from both plasmids was similar when transfected in mammalian cells (data not shown).

Fibstatin Inhibits Tumor Angiogenesis.
To elucidate whether the effect of fibstatin gene delivery on melanoma growth was due to an inhibition of angiogenesis, tumor capillaries were examined by anti-CD31 immunostaining. As shown in Fig. 6Citation , representative tumor sections from fibstatin- and endostatin-treated mice showed a significant decrease in CD31 staining, compared with the tumor sections from control mice. Quantification of microvessels at x200 indicated that the CD31-positive surface in tumors from control mice was 2.2 ± 0.58% (mean ± SD), compared with 0.74 ± 0.23% and 0.94 ± 0.39% in fibstatin-and endostatin-treated tumors, respectively (Fig. 6B)Citation . Taken together, these data demonstrated that fibstatin gene delivery inhibits the angiogenic response in vivo in a B16-F10 melanoma immunocompetent mouse model.



View larger version (72K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 6. Effect of fibstatin gene delivery on tumor angiogenesis A, representative histologic tumor sections were stained with an antimouse CD31. The immunostaining shows the rarefication of tumor blood vessels in fibstatin (Fib)- and endostatin (Endo)-treated mice as compared with control mice. B, quantitative analysis of CD31 immunostaining. The number of CD31-stained blood vessels was obtained from 10 different microscopic fields (x200) from at least 5 different tumor sections. The data are represented as mean ± SD. *, P < 0.05 was considered statistically significant by using a one-factor ANOVA.

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In this article, we report the identification of fibstatin, a human fibronectin fragment that interacts specifically with FGF-2 and corresponds to the major heparin-binding domain (Hep2) of fibronectin. Interestingly, fibstatin is highly similar to the Mr 29,000 fragment that is generated when fibronectin is digested by Dispase (17) . This indicates that such a fragment could be produced in physio(patho)logic conditions. Here, we demonstrate that fibstatin functions as an antiangiogenic molecule that inhibits in vitro FGF-2 dependent proliferation, migration, and tube formation of endothelial cells. This may indicate that, similar to FGF-2, fibstatin interferes with more than one step in the angiogenic process. In vivo, fibstatin inhibits Matrigel neovascularization and the growth of B16F10 melanoma. Tumor angiogenesis in melanoma was examined with CD31 immunostaining. Our results show a rarefaction of small blood vessels in both Fibstatin- and endostatin-treated tumor tissue sections, as compared with the control, indicating a role of fibstatin in suppressing tumor angiogenesis. Because fibstatin did not show any in vitro activity on the proliferation of B16F10 cells, its inhibitory effects on tumor growth is likely due to its antiangiogenic properties. To evaluate the antiangiogenic activities of fibstatin, we made a comparison with endostatin, one of the most described antiangiogenic factor. In our study, fibstatin is more efficient than endostatin when used as recombinant protein. However, the activity of endostatin is sensitive on its interaction with zinc (18) . The results obtained with recombinant protein show that the effect of fibstatin is always efficient, independently of its mode of preparation. The process of protein purification may denature the endostatin protein, and yield rates may be low. Gene delivery should, in theory, abrogate this problem (19) . Moreover, because angiostatic therapy would require a prolonged maintenance of therapeutic levels in vivo, the continuous delivery of a recombinant protein would be expensive and cumbersome. In vivo transfer of the corresponding genes constitutes an attractive solution to this problem.

For this reason, we evaluated the respective effect of endostatin and fibstatin on tumor growth after gene delivery. In this experiment, the antitumorigenic effect of fibstatin is still more powerful than the effect of endostatin.

The high affinity for heparin on the part of both fibstatin and FGF-2 raises the question as to whether heparin acts as a mediator in the interaction between fibstatin and FGF-2. Our results indicate that this interaction could be independent of the heparin binding of fibstatin because the interaction occurs in vitro between recombinant proteins without any addition of heparin or heparan sulfates proteoglycans (HSPG). Fibstatin could interact directly with FGF-2 and, thus, would inhibit, like PF-4 or thrombospondin, the interaction between FGF-2 and its receptors, or the dimerization or internalization of the growth factor (20 , 21) . Furthermore, we have mapped the fibronectin modules that are required for the binding with FGF-2. The module 13 plays a major role in the interaction with heparin (22 , 23) . However, this module alone is unable to interact with FGF-2. Indeed, both module 13 and module 14 are required for an efficient binding of FGF-2. Nevertheless, most of the endogenous antiangiogenic proteins identified thus far, like endostatin, angiostatin, PF-4, and kallistatin, bind to heparin (1 , 2 , 6 , 24 , 25) . This property is collectively interpreted as involved in the inhibition of FGF/VEGF activity by competing for their interaction with HSPG. Indeed, these two angiogenic factors present a strong heparin-binding activity.

It thus appears that at least two modes of action could be invoked for fibstatin: the binding to HSPG and the direct protein–protein interaction with FGF-2. In addition, we could also speculate that fibstatin acts by binding to integrins. Indeed fibstatin corresponds to the 12–14 type-III domain, which binds to {alpha}IIbß3 (26) and {alpha}4ß1 (27) . Moreover, it has been recently shown that two other antiangiogenic factors, tumstatin and endostatin, exhibit their activities through {alpha}vß3 and {alpha}5ß1 integrins (28) .

The characterization of novel antiangiogenic proteins and the elucidation of their mode of action is of great importance. Indeed, it would allow the consideration of associative strategies by using the electrotransfer of polycistronic vectors expressing several antiangiogenic peptides at the same time, with different properties to obtain an optimal angiogenesis inhibition.


   ACKNOWLEDGMENTS
 
We thank Alexia Schambourg, Catherine Zanibellato, Marcel Mariller, and Annick Ozil for technical assistance.


   FOOTNOTES
 
Grant support: Supported by grants from Institut National de la Santé et de la Recherche Médicale, University Paul Sabatier Toulouse, la Ligue Contre le Cancer, Conseil Régional Midi-Pyrénées, European Commission FP5 (QOL-2000-3.1.2, consortium CONTEXTH contract QLRT-2000-00721) and the French Ministry of Research (decision No 01H0387). C. Bossard was supported by the Association pour la Recherche contre le Cancer.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Hervé Prats, INSERM U589, C.H.U. Rangueil, Bat L3, 31403 Toulouse Cedex 04, France. Phone: 33-561-322-144; Fax: 33-561-322-141; E-mail: pratsh{at}toulouse.inserm.fr

Received 1/29/04. Revised 6/18/04. Accepted 8/18/04.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. O’Reilly MS, Boehm T, Shing Y, et al Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell 1997;88:277-85.[CrossRef][Medline]
  2. O’Reilly MS, Holmgren L, Shing Y, et al Angiostatin: a novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell 1994;79:315-28.[CrossRef][Medline]
  3. Bikfalvi A, Klein S, Pintucci G, Rifkin DB Biological roles of fibroblast growth factor-2. Endocr Rev 1997;18:26-45.[Abstract/Free Full Text]
  4. Baldin V, Roman AM, Bosc-Bierne I, Amalric F, Bouche G Translocation of bFGF to the nucleus is G1 phase cell cycle specific in bovine aortic endothelial cells. EMBO J 1990;9:1511-7.[Medline]
  5. Bossard C, Laurell H, Van Den Berghe L, Meunier S, Zanibellato C, Prats H Translokin is an intracellular mediator of FGF-2 trafficking. Nat Cell Biol 2003;5:433-9.[CrossRef][Medline]
  6. Homandberg GA, Williams JE, Grant D, Schumacher B, Eisenstein R Heparin-binding fragments of fibronectin are potent inhibitors of endothelial cell growth. Am J Pathol 1985;120:327-32.[Abstract]
  7. Folkman J, Shing Y Angiogenesis. J Biol Chem 1992;267:10931-4.[Free Full Text]
  8. Van den Berghe L, Laurell H, Huez I, Zanibellato C, Prats H, Bugler B FIF [fibroblast growth factor-2 (FGF-2)-interacting-factor], a nuclear putatively antiapoptotic factor, interacts specifically with FGF-2. Mol Endocrinol 2000;14:1709-24.[Abstract/Free Full Text]
  9. Gossen M, Bujard H Tight control of gene expression in mammalian cells by tetracycline- responsive promoters. Proc Natl Acad Sci USA 1992;89:5547-51.[Abstract/Free Full Text]
  10. Prats AC, Vagner S, Prats H, Amalric F cis-acting elements involved in the alternative translation initiation process of human basic fibroblast growth factor mRNA. Mol Cell Biol 1992;12:4796-805.[Abstract/Free Full Text]
  11. Huez I, Bornes S, Bresson D, Creancier L, Prats H New vascular endothelial growth factor isoform generated by internal ribosome entry site- driven CUG translation initiation. Mol Endocrinol 2001;15:2197-210.[Abstract/Free Full Text]
  12. Holzinger A, Phillips KS, Weaver TE Single-step purification/solubilization of recombinant proteins: application to surfactant protein B. Biotechniques 1996;20:804-6, 808.[Medline]
  13. Marchal I, Cerutti M, Mir AM, et al Expression of a membrane-bound form of Trypanosoma cruzi trans-sialidase in baculovirus-infected insect cells: a potential tool for sialylation of glycoproteins produced in the baculovirus-insect cells system. Glycobiology 2001;11:593-603.[Abstract/Free Full Text]
  14. Chaabihi H, Ogliastro MH, Martin M, Giraud C, Devauchelle G, Cerutti M Competition between baculovirus polyhedrin and p10 gene expression during infection of insect cells. J Virol 1993;67:2664-71.[Abstract/Free Full Text]
  15. Summers M, Smith G . A manual of methods for baculovirus vectors and insect cell culture procedures 1987 Texas Agricultural Experiment Station College Station, TX
  16. Cayrol C, Flemington EK Identification of cellular target genes of the Epstein-Barr virus transactivator Zta: activation of transforming growth factor beta igh3 (TGF-beta igh3) and TGF-beta 1. J Virol 1995;69:4206-12.[Abstract]
  17. Fujita H, Mohri H, Kanamori H, Iwamatsu A, Okubo T Binding site in human plasma fibronectin to HL-60 cells localizes in the C-terminal heparin-binding region independently of RGD and CS1. Exp Cell Res 1995;217:484-9.[CrossRef][Medline]
  18. Boehm T, O’Reilly MS, Keough K, Shiloach J, Shapiro R, Folkman J Zinc-binding of endostatin is essential for its antiangiogenic activity. Biochem Biophys Res Commun 1998;252:190-4.[CrossRef][Medline]
  19. Wen XY, Bai Y, Stewart AK Adenovirus-mediated human endostatin gene delivery demonstrates strain-specific antitumor activity and acute dose-dependent toxicity in mice. Hum Gene Ther 2001;12:347-58.[CrossRef][Medline]
  20. Perollet C, Han ZC, Savona C, Caen JP, Bikfalvi A Platelet factor 4 modulates fibroblast growth factor 2 (FGF-2) activity and inhibits FGF-2 dimerization. Blood 1998;91:3289-99.[Abstract/Free Full Text]
  21. Taraboletti G, Belotti D, Borsotti P, et al The 140-kilodalton antiangiogenic fragment of thrombospondin-1 binds to basic fibroblast growth factor. Cell Growth Differ 1997;8:471-9.[Abstract]
  22. Sachchidanand, Lequin O, Staunton D, et al Mapping the heparin-binding site on the 13–14F3 fragment of fibronectin. J Biol Chem 2002;277:50629-35.[Abstract/Free Full Text]
  23. Sharma A, Askari JA, Humphries MJ, Jones EY, Stuart DI Crystal structure of a heparin- and integrin-binding segment of human fibronectin. EMBO J 1999;18:1468-79.[CrossRef][Medline]
  24. Levine SP, Wohl H Human platelet factor 4: Purification and characterization by affinity chromatography. Purification of human platelet factor 4. J Biol Chem 1976;251:324-8.[Abstract/Free Full Text]
  25. Miao RQ, Agata J, Chao L, Chao J Kallistatin is a new inhibitor of angiogenesis and tumor growth. Blood 2002;100:3245-52.[Abstract/Free Full Text]
  26. Mohri H, Tanabe J, Katoh K, Okubo T Identification of a novel binding site to the integrin alphaIIbbeta3 located in the C-terminal heparin- binding domain of human plasma fibronectin. J Biol Chem 1996;271:15724-8.[Abstract/Free Full Text]
  27. Mould AP, Humphries MJ Identification of a novel recognition sequence for the integrin alpha 4 beta 1 in the COOH-terminal heparin- binding domain of fibronectin. EMBO J 1991;10:4089-95.[Medline]
  28. Sudhakar A, Sugimoto H, Yang C, Lively J, Zeisberg M, Kalluri R Human tumstatin and human endostatin exhibit distinct antiangiogenic activities mediated by alpha v beta 3 and alpha 5 beta 1 integrins. Proc Natl Acad Sci USA 2003;100:4766-71.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 EndocrinologyHome page
O. P. Milling Smith, S. Battersby, K. J. Sales, H. O. D. Critchley, and H. N. Jabbour
Prostacyclin Receptor Up-Regulates the Expression of Angiogenic Genes in Human Endometrium via Cross Talk with Epidermal Growth Factor Receptor and the Extracellular Signaling Receptor Kinase 1/2 Pathway
Endocrinology, April 1, 2006; 147(4): 1697 - 1705.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bossard, C.
Right arrow Articles by Prats, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bossard, C.
Right arrow Articles by Prats, H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online