Jacques Pouysségur, Edurne Berra, Emmanuel Gothié, Julie Milanini, Darren Richard, and Gilles Pagès

Angiogenesis and permeability of blood vessels are regulated by vascular endothelial growth factor (VEGF) via its two known receptors VEGFR-1 and VEGFR-2. Recently, additional VEGF genes have been cloned and new insight has been obtained of the molecular mechanisms regulating the function of endothelial cells of lymphatic vessels. VEGF-C and VEGF-D have been shown to stimulate lymphangiogenesis and their receptor VEGFR-3 has been linked to human hereditary lymphoedema, although there is evidence that also other genes are involved. Such molecules may allow the regulation of angiogenesis and lymphangiogenesis as well as tissue edema involved in many diseases.

The VEGFR-3 receptor tyrosine kinase is related to the VEGF receptors, but does not bind VEGF and its expression becomes restricted mainly to lymphatic endothelia during development. We have found that homozygous VEGFR-3 targeted mice die around day 10 of embryonic development due to failure of cardiovascular development and that heterozygous missense mutations of VEGFR-3 inactivating the tyrosine kinase activity are associated with human hereditary lymphedema and two mouse models for lymphedema reproduce essential features of this disease. We have also purified and cloned the VEGFR-3 ligand, VEGF-C, which is made as a precursor protein having an extended N-terminus and a C-terminal half containing extra cysteine-rich motifs characteristic of a protein component of silk. Transgenic mice expressing VEGF-C developed a hyperplastic lymphatic vessel network and show evidence of lymphangiogenesis. However, proteolytically processed VEGF-C was also capable of stimulating VEGFR-2 and was weakly angiogenic. VEGF-C induced vascular permeability, but its point mutant, which retained lymphangiogenic properties and activated only VEGFR-3 did not. VEGF-D is closely related to VEGF-C, similarly processed and binds to the same receptors. Thus, VEGF-C and VEGF-D appear to be both angiogenic and lymphangiogenic growth factors. When overexpressed as a transgene in the RIP-Tag model of pancreatic ß-cell tumors, VEGF-C induced the growth of peritumoral lymphatic vessels and was associated with lymphatic metastasis. VEGF-C overexpression also led to lymphangiogenesis and intralymphatic tumor growth in an orthotopic model of human breast carcinoma in SCID mice. Furthermore, soluble VEGFR-3 blocked these changes. However, VEGFR-3 is also induced in blood vessels of various types of human cancer. Ongoing experiments address the role of the VEGFR-3 signaling pathway in embryonic and tumor angiogenesis and the mechanisms of lymphatic metastasis.

Marco Presta

Unit of General Pathology and Immunology, Department of Biomedical Sciences and Biotechnology, University of Brescia, 25123 Brescia, Italy

AUTOCRINE AND PARACRINE ROLES OF FGF2 IN ANGIOGENESIS

Tumor cells of different origin, macrophages, and T lymphocytes express fibroblast growth factor-2 (FGF2) in vitro and in vivo. FGF2 lacks a classic signal peptide for secretion. However, cell damage may cause the release of FGF2 from producing cells. Also, an alternative mechanism of exocytosis of FGF2, independent of the endoplasmic reticulum/Golgi pathway, has been proposed. Accordingly, FGF2 has been found associated with the extracellular matrix (ECM) of cell cultures and located in the basement membranes of blood vessels. On this basis, FGF2 is thought to exert its effects on endothelial cells via a paracrine mode consequent to its release by other cells and/or mobilization from ECM.

Besides experimental evidence for a paracrine mode of action for FGF2, some observations raise the hypothesis that FGF2 may also play an autocrine role in endothelial cells. Indeed, endothelial cells produce FGF2 that modulates cell proliferation and migration, as well as the production of proteinases and their receptors. In vivo, FGF2 expression occurs in the endothelium adjacent to neoplastic cells in several human tumor types. These neoplasms include neuroblastoma, astrocytoma, glioblastoma, meningioma, pheochromocytoma, melanoma, carcinomas of the stomach and colon, and adenocarcinomas of the larynx, endometrium, and cervix. Thus, FGF2 expression is a common feature of vascular endothelium during tumor angiogenesis.

These observations strongly support the hypothesis that neovascularization may be triggered by molecule(s) released by tumor cells and/or infiltrating inflammatory cells that induce FGF2 upregulation in the quiescent endothelium. In keeping with this hypothesis is the observation that tumor cells of different origin release molecule(s) able to interact with endothelium and to upregulate the expression of FGF2 that, in turn, stimulates the fibrinolytic potential of the endothelial cell in an autocrine manner. In addition, FGF2 itself, thrombin, nitric oxide, and interleukin-2 stimulate FGF2 production in endothelial cells.

FGF2 OVEREXPRESSION IN ENDOTHELIAL CELLS

To investigate the biological consequences of endothelial cell activation by endogenous FGF2, immortalized Balb/c mouse aortic endothelial cells (MAE cells) were transfected with a retroviral expression vector harboring a human FGF2 cDNA. FGF2 transfectants express all FGF2 isoforms and are characterized by a transformed morphology and an increased saturation density. FGF2 transfectants show invasive and morphogenetic behavior in three-dimensional gel that is prevented by anti-FGF2 antibody, revealing the autocrine modality of the process. The biological consequences of this autocrine activation were investigated in vivo. FGF2-transfected MAE cells induce the growth of highly vascularized tumors. In agreement with these observations, FGF2-transfected MAE cells induce an angiogenic response when implanted in the avascular rabbit cornea. Also, they cause an increase in vascular density and formation of hemangiomas in the chorioallantoic membrane (CAM) when injected into the allantoic sac of the chick embryo. Thus, the data demonstrate that pZipbFGF2-MAE cells induce highly vascularized spindle-cell hemangioendotheliomas in immunodeficient mice that are sustained by recruitment of host elements, including endothelial cells.

Cidofovir has been approved for the treatment of cytomegalovirus (CMV) retinitis in AIDS patients and possesses potent inhibitory activity against various papillomavirus (HPV)-induced tumors in animal models and patients. In addition, cidofovir inhibits the development of murine polyomavirus (PyV)-induced hemangiomas in rats by an as yet uncharacterized antiviral-independent mechanism. We investigated the effect of cidofovir on virus-independent vascular tumors originated by FGF2-T-MAE cells, a subclone of pZipbFGF2-MAE cells. In vitro, cidofovir was cytostatic for FGF2-T-MAE cells. Cidofovir did not affect FGF2-T-MAE cell sprouting in 3D fibrin gel and morphogenesis on Matrigel at non-cytotoxic concentrations. In vivo, cidofovir completely suppressed hemangioma formation on CAM induced by intra-allantoic injection of FGF2-T-MAE cells, without affecting the normal CAM vessels. Intratumoral or systemic administration of cidofovir caused a significant inhibition of the growth of subcutaneous, intraperitoneal, or intracerebral FGF2-T-MAE-xenografts in nude and SCID mice. Drug-induced apoptosis was observed in FGF2-T-MAE tumors as soon as 2 days after the beginning of treatment. Thus, cidofovir appears to inhibit the growth of endothelial-derived tumors via induction of apoptosis without exerting a direct anti-angiogenic activity. Cidofovir may be explored for the treatment of tumors that are not associated with an oncogenic virus.

FGF2 OVEREXPRESSION IN TUMOR CELLS

Various tumor cell lines express FGF2 in vitro. In situ hybridization and immunolocalization experiments have shown the presence of FGF2 mRNA and/or protein in neoplastic cells, endothelial cells, and infiltrating cells within human tumors of different origin. Antisense-FGF2 and FGF receptor-1 cDNAs inhibit neovascularization and growth of human melanomas in nude mice. Also, a significant correlation between the presence of FGF2 in cancer cells and advanced tumor stage has been reported. Recent observations have shown that a secreted FGF-binding protein can serve as an angiogenic switch for different tumor cell lines, including squamous cell carcinoma and colon cancer cells. Interestingly, targeting of FGF-binding protein with specific ribozymes reduces significantly the growth and vascularization of xenografted tumors in mice despite the high levels of VEGF produced by these cells. These data suggest that modulation of FGF2 expression, release, and mobilization may allow a fine-tuning of the angiogenesis process even in the presence of significant levels of VEGF. This hypothesis is supported by the capacity of the two factors to act synergistically in stimulating angiogenesis in vitro and in vivo.

To investigate the impact of the modulation of FGF2 expression on the neovascularization at different stages of tumor growth, we generated stable transfectants (Tet-FGF2) from the human endometrial adenocarcinoma HEC-1-B cell line in which FGF2 expression is under the control of the tetracycline-responsive promoter (Tet-off system). After transfection, independent clones were obtained in which FGF2 mRNA and protein were upregulated compared to parental cells. Also, the conditioned medium of Tet-FGF2 transfectants caused proliferation, urokinase-type plasminogen activator upregulation, migration, and sprouting of cultured endothelial cells. A 3 day-treatment of Tet-FGF2 cell cultures with tetracycline abolished FGF2 overexpression and the biological activity of the conditioned medium without affecting their proliferative capacity. Tet-FGF2 cells formed tumors when injected s.c. in nude mice. Administration of 2.0 mg/ml tetracycline in the drinking water prior to cell transplantation and continued throughout the whole experiment inhibited FGF2 expression in Tet-FGF2 tumor lesions. This was paralleled by a significant decrease in the rate of tumor growth and vascularization to values similar to those observed in lesions generated by parental HEC-1-B cells. Tetracycline administration 20 days after tumor cell implant, although equally effective in reducing FGF2 expression and inhibiting tumor vascularity, only minimally impaired the growth of established Tet-FGF2 tumors.

The results indicate that FGF2 expression deeply affects the initial tumor growth and neovascularization of HEC-1-B human endometrial adenocarcinoma in nude mice. On the contrary, the growth of established tumors appears to be independent of the inhibition of FGF2 expression and decreased vascular density. The possibility that a significant reduction of angiogenesis may not affect the progression of large tumors points to the use of anti-angiogenic therapy in early tumor stage.

CONCLUDING REMARKS

FGF2 exerts angiogenic activity in vivo and induce a pro-angiogenic phenotype in cultured endothelial cells. In vivo, FGF2 exerts paracrine effects on endothelial cells when released by tumor and/or inflammatory cells. FGF2 may also play an autocrine role in endothelial cells in vitro and in vivo. FGF2 may therefore represent a target for anti-angiogenic therapies. In order to assess the angiostatic potential of different classes of compounds, novel experimental models can be developed based on the autocrine and/or the paracrine capacity of FGF2.

ACKNOWLEDGEMENTS

The work from our laboratory was supported by grants from Associazione Italiana per la Ricerca sul Cancro, Istituto Superiore di Sanità (AIDS Project), C.N.R. (Target Project Biotechnology), and M.U.R.S.T (Cofin 1999).

Tie2/Angiopoietin pathway in embryonic vascular development

Tie2 is a novel member of a large family of receptor tyrosine kinases that includes the PDGF, FGF and VEGF receptors. Tie2 is highly conserved across species and is expressed predominantly in embryonic endothelial cells in all vertebrate species examined including the zebrafish.^1-4 Disruption of Tie2 function in transgenic mice leads to embryonic lethality due to defects in the embryonic microvasculature characterized by incomplete vascular remodeling and disruption of endothelial-mesenchymal interactions. There are currently at least three ligands for Tie2, Angiopoietin 1-3 (Ang1-3). As expected, Ang1 binding results in Tie2 activation and phosphorylation. Convsersely, Ang2 competes with Ang1 for Tie2 binding but does not stimulate Tie2 phosphorylation suggesting that Ang2 may be a naturally occurring inhibitor of Tie2/Ang1 activity. Consistent with this notion, disrupting the function of Ang1 or overexpressing Ang2 produced similar phenotypes and further established the importance of Tie2 signaling in the formation of the embryonic vasculature.

Tie2/Angiopoietin pathway in tumor angiogenesis

Based on the role of the Tie2/Ang pathway in the development of the embryonic vasculature, we hypothesized that Tie2 might also play a role during physiologic and pathologic angiogenesis in adult tissues. Supporting this hypothesis, Tie2 was expressed in the endothelium of rodent and human tumors suggesting a role in tumor angiogenesis.^5-7 To further explore the role of Tie2 in tumor angiogenesis, a recombinant, soluble form of the extracellular domain of Tie2 (ExTek) was made as an inhibitor of Tie2 action.⁵ ExTek potently inhibited Ang1 binding to immobilized Tie2 in a Biacore assay and blocked Ang1 mediated Tie2 autophosphorylation in cultured endothelial cells. Direct administration of ExTek to a rat mammary carcinoma in a cutaneous tumor window inhibited both tumor angiogenesis and tumor growth. Interestingly, a soluble VEGF receptor yielded similar inhibition suggesting that, as in the embryonic vasculature, both the VEGF and Tie2/Angiopoietin pathways are required for tumor angiogenesis, a finding that was recently confirmed in a human melanoma xenograph model.^8,9 Demonstrating the potential clinical application of Tie2 inhibition, systemic administration of ExTek using a recombinant adenovirus (AdExTek) inhibited growth of two different murine tumors, a melanoma (B16F10.9) and a mammary carcinoma (4T1). ¹⁰ Taken together, these findings demonstrate a role for the Tie2 pathway in tumor angiogenesis and suggest a role for Tie2 during angiogenesis in other pathologic and non-pathologic adult tissues.

Cellular mechanisms of Tie2/Angiopoietin action

Two recent studies suggested that the Tie2/Angiopoietin pathway played an important role in the morphogenetic changes that occur during endothelial sprout formation in vitro.^11,12 Consistent with these studies, in the corneal micropocket assay, ExTek almost completely blocked sprouting of neovessels stimulated by tumor cell conditioned media, an effect that was similar to blocking the VEGF pathway using a soluble VEGF receptor.^5,8 Moreover, in the rat aortic ring assay, infection of aortic rings with AdExTek almost completely inhibited formation of endothelial sprouts in spite of previous studies demonstrating the dependence of this model on the VEGF pathway. These results suggested that, in adult tissues, the Tie2 pathway, like the VEGF pathway, was required during the earliest stages of vascular sprout development. In cultured endothelial cells, Ang1*-mediated activation of Tie2 failed to stimulate mitogenesis or chemotaxis, two endothelial responses that are required for vascular sprouting and elongation. However, Tie2 stimulation inhibited endothelial apoptosis under conditions of serum deprivation, suggesting that one role of Tie2 was to enhance endothelial survival in nascent, unperfused vascular sprouts (C.K. and K.P. unpublished data and see also reference 12).

Molecular mechanisms of Tie2/Angiopoietin action

Using the Tie2 kinase domain as a bait to screen a human heart cDNA library, the p85 subunit of PI-3 kinase was identified as a potential signaling molecule in the Tie2/Angiopoietin pathway.¹³ Mutational analysis of Tie2 demonstrated that tyrosine 1112 was the major site of association of PI3 kinase with Tie2 and that this association was required for the activation of PI-3 kinase upon Tie2 activation. Importantly, tryosine 1112 was also required for Tie2-mediated activation of AKT, a serine/threonine kinase known to be activated downstream of PI-3 kinase and to be involved in PI-3 kinase-mediated cell survival. Wortmannin, a selective inhibitor of PI-3 kinase, inhibited both Tie2-mediated activation of AKT and Tie2-mediated endothelial cell survival. Together with the studies presented in the preceding paragraph, these findings suggest that the Tie2/PI-3 kinase/AKT pathway is a crucial pathway for endothelial cell survival during vascular sprout formation in the adult vasculature.

Shp2 is another signaling molecule that associates with activated, autophosphorylated Tie2 kinase.¹⁴ Interestingly, Shp2 can have either a negative or positive role in various signaling pathways. Mutating the site of association of Tie2 with Shp2 (Y1112F) results in ligand-mediated receptor hyperphosphorylation compared with the wild type Tie2. The mutated receptor also appears to be a more potent activator of downsteam signaling pathways (PI-3 kinase/Akt and map kianse) than the wild type receptor. Moreover, recombinant wild-type Tie2 kinase is a much better substrate for recombinant Shp2 than the mutated kinase. These findings are consistent with a predominantly negative regulatory role for Shp2 in the Tie2 pathway. Whether other signaling molecules also contribute to the function of Tie2 during angiogenesis remains to be determined.^14,15

Tie2/Angiopoietin pathway in vascular maintenance and physiologic angiogenesis

As in the tumor vasculature, Tie2 expression was localized to the endothelium during hormone-stimulated angiogenesis in the in the rat ovary and uterus.¹⁶ In addition, Tie2 was localized to the endothelium of neovessels in rat excision skin wounds. Immunoprecipitated Tie2 from the rat wound tissues was tyrosine-phosphorylated indicating that Tie2 was playing an active role in angiogenesis throughout the process of wound healing. Unexpectedly, Tie2 was also expressed and phosphorylated in the vasculature of all quiescent adult vascular tissues suggesting a role for the Tie2/Angiopoietin pathway maintenance of adult blood vessels (but see also reference 17).

Summary

The studies described here indicate that the Tie2/Angiopoietin pathway is required for the earliest stages of vascular sprout formation in adult tissues. One function of the Tie2 pathway is to enhance the survival of the endothelial cells comprising the early vascular sprout at least in part by activation of the PI-3 kinase/AKT pathway. Tie2 is expressed in the endothelium of tumor vessels and blocking the Tie2 pathway blocks the growth of a number of rodent tumors suggesting that targeting the Tie2 pathway may offer an effective antiangiogenic treatment for cancer. However, the expression and activation of Tie2 in the quiescent vasculature suggests that Tie2 may play a role in the maintenance of the adult vasculature and that therapeutic blockade of the Tie2 pathway should be done with caution.

Bibliography:

1.    Hanahan, D.: Signaling vascular morphogenesis and maintenance. Science 277:48-50, 1997.

2.    Korpelainen, E.I. and Alitalo, K..: Signaling angiogenesis and lymphangiogenesis. Current Opinion in Cell Biol. 10:159-164, 1998.

3.    Gale, N.W. and Yancopoulos, G.D.: Growth factors acting via endothelial cell-specific receptor tyrosine kinases: VEGFs, Angiopoietins, and ephrins in vascular development. Genes and Devel. 13:1055-1066, 1999.

4.    Lyons, M.S., Bell, B., Stainier, D., Peters, K.G.: Isolation of the zebrafish homologues for the tie-1 and tie-2 endothelium specific receptor tyrosine kinases. Dev. Dynamics, 212:133-140, 1998.

5.    Lin, C., Shan, S., Dewhirst, M., Rao, P., Peters, K.G.: Inhibition of Tumor Growth by Targeting Tumor Endothelium Using a Soluble Form of the Tie-2/Tek Receptor Tyrosine Kinase. J. Clin. Invest. 100:2072-2078, 1997.

6.    Peters, K.G., Coogan, A., Berry, D., Marks, J., Iglehart, J.D., Kontos, C.D., Trogan, E., Rao., P.: Expression of the Endothelial Receptor Tyrosine Kinase Tie-2/Tek in Breast Tumors Suggests a Role in Tumor Angiogenesis. Br. J. Cancer, 77:51-56, 1998.

7.    Stratmann, A., Risau, W., Plate, K.H.: Cell type-specific expression of angiopoietin-1 and angiopoietin-2 suggests a role in glioblastoma angiogenesis. Am. J. Pathol., 153:1459-1466, 1998.

8.    Lin, C., Sankar, S., Shan, S., Dewhirst, M., Quinn, T., Peters, K.G.: Inhibition of Tumor Growth by Targeting Tumor Endothelium Using a Soluble VEGF Receptor. Cell Growth & Differentiation, 9:49-58, 1998.

9.    Siemeister, G., Schirner, M., Weindel, K., Reusch, P., Menrad, A., Marme, D., Martiny-Baron, G.: Two independent mechanisms essential for tumor angiogenesis: Inhibition of human melanoma xenograft growth by interfering with either the vascular endothelial growth factor pathway or the Tie-2 pathway. Cancer Res., 59:3185-3191, 1999.

10. Lin, C., Buxton, J.A., Acheson, A., Redziejewski, C., Maisonpierre, P.C., Yancopoulos, G.D., Channon, K.M., Hale, L. P., Dewhirst, M.W., George, S.E., Peters, K.G.: Anti-angiogenic gene therapy targeting the endothelium-specific receptor tyrosine kinase Tie2. Proc. Natl. Acad. Sci., USA, , 95:8829-8834, 1998.

11. Kobilzek, T.I., Weiss, C., Yancopoulos, G.D., Deutsch, U., Risau, W.: Angiopoietin-1 induces sprouting angiogenesis in vitro. Current Biol. 8:529-532, 1998.

12. Papapetropoulos A., Garcia-Cardena, G., Dengler, T.J., Maisonpierre, P.C., Yancopoulos, G.D., Sessa, W.C.: Direct actions of angiopoietin-1 on human endothelium: Evidence for network stabilization, cell survival, and interaction with other angiogenic growth factors. Lab. Invest. 79:213-223, 1999.

13. Kontos, C.D., Stauffer, T.P., Yang, W-P., Huang, L., Blanar, M.A., Meyer, T., Peters, K.G.: Activation of p85 with a Non-concensus Binding Motif on Tie2 is Required for Activation of Phosphatidylinositol 3-Kinase and Akt. Molecular & Cellular Biology, 18:4131-4140, 1998.

14. Huang, L., Turck, C.W., Rao, P., Peters, K.G.: GRB2 and SH-PTP2: Potentially important endothelial signaling molecules downstream of the TEK receptor tyrosine kinase. Oncogene. 11:2097-2103, 1995.

15. Jones, N., Dumont, D.J. The Tek/Tie2 receptor signals through a novel Dok-related docking protein, Dok-R. Oncogene. 17:1097-1108, 1998.

16. Wong, A.L., Haroon, Z.A., Werner, S., Dewhirst, M.W., Greenberg, C.S., Peters, K.G.: Tie2 expression and phosphorylation in angiogenic and quiescent adult tissues. Circulation Res. 87:567-574, 1997.

17. Koblizek, TI., Runting, A.S., Stacker, S.A., Wilks, A.F., Risau, W., Deutsch, U. Tie2 receptor expression and phosphorlyation in cultured cells and mouse tissues. Eur. J. Biochem. 244:774-779, 1997.

Sprouting of new capillaries from preexisting vessels, or angiogenesis, occurs in several physiological or pathological conditions such as tumor progression, diabetic retinopathy or rheumatoïd arthritis. This local hypervascularization is thought to result from release by the tissues of growth factors interacting with their receptors on endothelial cells which in turn migrate, proliferate and differentiate into new capillaries. Both fibroblast growth factor 2 (FGF2) and vascular endothelial growth factor exert a potent pharmacological angiogenic activity as determined by the usual corneal pocket assay but the actual role of FGF2 in pathological angiogenesis remains questionable.

The concept that inhibitors of angiogenesis might be usefull as antitumor therapeutic agents has been launched almost thirty years ago by J. Folkman. Several products can inhibit tumor angiogenesis and entered clinical trials recently. The widespread expectation is that these antiangiogenic compounds have the potential to eradicate tumors by killing genetically stable target cells without inducing drug tolerance. However most of them might not be as specific for tumor endothelial cells as expected from preclinical data. Therefore another field of investigation focused on the identification of agents targeting only angiogenic vessels but not established vessels. Such a goal requires the identification of ports of entry (receptors, antigens) which either are not expressed or are not activable by their cognate ligand in the normal adult vasculature.

Theorical issues

There is no rationale to discover new antiangiogenic agents. The demonstration that a given growth factor exerts a pharmacological angiogenic activity does not preclude that its inhibition might be beneficial to reduce pathological angiogenesis. For instance basic peptides encoded by exon 6 of VEGF 189 aa can chase FGF2 from its storage sites in the corneal stroma and elicit a strong dose and sequence dependent angiogenic response although they do not bind to VEGF receptors. The straightforward demonstration that heparin binding growth factors such as VEGF or FGF2 act by activating their receptors and not by releasing other growth factors from the extracellular matrix raises several methodologic points. Systemic delivery of growth factors is inefficient because they are sequestered onto heparan sulfates in the extracellular matrix of the vascular wall and therefore do not reach their target receptors.

The large number of heparin binding growth factor receptors makes the deciphering of their functions a difficult task. The measurement of the remaining functions exerted by mutant growth factors deleted of their binding domain to only one receptor provides an interesting clue but the problem of the bioavailability is not solved. the Preclinical and clinical data can demonstrate unambiguously that a growth factor or a growth factor receptor is up-regulated in pathological vessels and thus would represent a potential therapeutic target. This strategy has led to the discovery of many anti-angiogenic agents. However long lasting treatments might affect normal endothelial cells. For instance the inhibition of the kinase activity of VEGFR2 induces pulmonary emphysema. Thus the elucidation of the functions mediated by a single factor requires the preparation of circulating agonists in order to inhibit only those which are not functional in the normal vasculature.

Mechanisms of action of VTA targeting known receptors.

We had to construct circulating agonists mimicking the distinct domains of VEGF and FGF2 interactions with their receptors and therefore we relied on the anti-idiotypic strategy. We raised several anti-idiotypic antibodies by priming lymphocytes in the lymph nodes with neutralizing IgG against VEGF or FGF2. The anti-idiotypic IgG against VEGF (AId-V) or FGF2 (AId-F) were further purified by affinity chromatography for the anti VEGF or anti-FGF2 neutralizing IgG previously conjugated to CN-Br sepharose. Their specificity for VEGF-R2 or FGF-R1 was ascertained by radioreceptorassay using heparan sulfates deficient CHO cells transfected with the corresponding cDNA sequences. AId-V and AId-F2 induced a similar mitogenic effect on microvascular endothelial cells which was inhibited only by their corresponding immunogen, thus demonstrating the anti-idiotypic property of AId-F2 and AId-V.

Since both anti-idiotypic antibodies stimulated endothelial cell proliferation we determined whether systemic activation of endothelial cell proliferation could modulate tumor angiogenesis. MCF-7 cells grafted in ovariectomized nude mice were challenged with PI-IgG, AId-F2, AId-V or VEGF. We found that VEGF could not increase tumor growth whereas AId-V or AId-F2 treatments resulted in a similar increase of tumor volume (doubling time ca 6 days as compared to PI-Ig treated mice). Tumor sections immunostained with anti-CD31 antibodies demonstrated an increase of vascularization upon AId-V, but not AId-F2 treatment. Image analysis of low magnification photomicrographs demonstrated striking differences in the distribution of endothelial cells in tumor sections. In contrast no difference in apoptotic cell counts was observed. Upon estrogen addition xenografts of MCF-7 led to tumor formation which was increased through distinct mechanisms by AId-V (angiogenesis) and AId-F2 (stroma reaction).

The most striking result was that none of these angiogenic factors receptors could be activated unless a converting factor had previously switched the phenotype of the endothelial cells.

Organ specificity of VTA

Although the VEGF promoter does not contain consensus steroid- responsive sequences, estrogen can up-regulate VEGF expression at a level comparable to that obtained upon FGF2 or AId-F2 in MCF-7. Estrogen exposure of endothelial cells which do not express VEGF induced a dose-dependent increase of VEGF expression and a mitogenic effect which was abolished by neutralizing anti VEGF antibody. In contrast anti-FGF2 neutralizing antibody was inefficient on E2-stimulated endothelial cell proliferation. We looked if an increase of VEGF bioavailability might be sufficient to allow the tumor take. When ovariectomized mice did not receive an estrogen pellet, they did not develop tumors whether they had been challenged with anti-idiotypic antibodies or not. Similarly VEGF¹⁶⁵-transfected MCF-7 cells did not develop tumors in the absence of estrogen whereas when estrogen was present their doubling time was even lower (5 days) than those growing in AId-V or AId-F2 challenged mice. Thus, although in vitro experiments have clearly shown that VEGF or estrogen stimulates the expression of VEGFR2, neither constitutive expression of VEGF in MCF-7 cells nor systemic activation of VEGFR2 could circumvent estrogen ablation. This indicates that VEGFR2 and FGFR1 are not functional in skin vessels from which is derived the tumor vascularization, Taken together these data suggest that the permissive effect of estrogen on tumor angiogenesis resulted from the release by MCF-7 cells of soluble factors which would induce a phenotypic switch of dermal endothelial cells into tumor endothelial cells.

We previously showed that AId-V failed to activate VEGF receptors in normal vessels, and did not induce vasodilatation leading to hypotension (Malavaud, 1997) or vascular permeability. However AId-V induced corneal angiogenesis when inserted in the cornea (Ortéga, 1996). More surprisingly the surgical traumatism made during the insertion of the pellet in the corneal stroma is sufficient to induce the angiogenic switch of endothelial cells in the limbal vessels (Ortéga, 1997). As expected VEGF and FGF2 induced a similar corneal angiogenesis in normal or castrated animals.

Thus the converting factors in inflammatory and estrogen-dependent tumor take are different.

VTA targeted on VEGFR2

Knowing that VEGFR2, despite it is expressed in the normal vasculature, does not constitute a target for agonistic antibodies, we conjugated Aid-V to a toxin.

In our laboratory we selected from retina-derived capillaries clones of non angiogenic cells which do not differentiate in 3D cultures in the presence of VEGF (Figure 3) or angiogenic cells which do differentiate. Although both strains differentiate upon FGF2 addition, VEGF induces an increase of bcl-2 expression and prevents TNF-a dependent apoptosis only in angiogenic cells. The immunoconjugate kills only angiogenic endothelial cells but not the non angiogenic ones.

When injected in nude mice bearing PC3 (human prostate cancer cells) tumors it inhibited by more than 80% their progression. Conversely a similar regimen of immunotoxin did not affect the normal vasculature nor the immunity of wild-type mice.

It has been taken for granted until recently that VTA should bind only to antigens or receptors specifically expressed in angiogenic vessels. The proof of concept has been mainly demonstrated by the work of P. Thorpe who inserted foreign genes in tumor endothelial cells and therefore induced tumor infarction by bispecific antibodies recognizing tissue factor (Huang, 1997). Promising candidate molecules are represented by endoglin, endosialin, fibronectin isoform. The strategy of phage display (Pasqualini, 1999) or gene soustraction () has also provided evidences that the number of gene selectively expressed in tumor endothelial cells must have been underscored because the endothelial compartment represents a very low amount of the tumor and thus the specific genes may have escape to tumor libraries screening. However the concept of specific expression should be revisited by functional data to achieve a safe Trojan horse strategy.

We, among others, developed a method that allows the identification of peptides with homing capability to different vascular beds after in vivo administration of a phage display random peptide library. The selection is based on the use of peptides that are expressed on the surface of the bacteriophage. Extensive previous work has established that peptide libraries can be used to probe tissue-specific and angiogenesis-related vascular homing. Taken together, this work has uncovered a novel vascular address system. We have isolated peptides that can home to normal blood vessels or sites of angiogenesis through the circulation via this vascular address system. Every normal or diseased organ appears to display a unique signature on its blood vessels that selected peptides can use as a target. We have also developed complementary methods of assessing the distribution of peptides guided to the vasculature, their tissue-specificity, and their target cells.

Angiogenic vasculature as a therapeutic target. It is well established that angiogenesis, the recruitment of new blood vessels, is an important rate-limiting step in solid tumor growth. New antitumor therapies based on the premise that inhibiting angiogenesis suppresses tumor growth are currently being tested in clinical trials. Angiogenesis is a multi-stage process that involves the release and activation of angiogenic factors, endothelial cell migration and proliferation, and differentiation into newly formed capillaries (6-11). The neovasculature of diseases with an angiogenic component differentially expresses many cell surface receptors in the endothelium (3,12,13). Identification of novel molecules characteristic of angiogenic vasculature by techniques such as in vivo phage display will improve our understanding of the plasticity of the activated endothelial phenotype and suggest new therapeutic strategies (14-20). Thus far, identification and isolation of such molecules has been slow, in large part because endothelium-derived cells undergo marked phenotypical changes when grown in culture (21). Angiogenic vasculature is an appealing target for cancer therapy since it is composed of non-malignant cells; such endothelium-derived cells are genetically stable and presumably less prone to acquire drug resistance. Targeting angiogenic blood vessels has other advantages such as an improved accessibility to the drug, and an intrinsic amplification mechanism (6,7,19). The idea of therapy directed at angiogenic vasculature has been proposed for more than a decade but its impact has not yet been fully realized. However, an intense effort is currently underway for the development of antiangiogenic strategies because of the fact that solid tumors cannot progress without new blood vessels, (6-10,15,19).

In vivo selection of peptides from phage display libraries enables homing probes to be isolated through a functional screening: the ability to home selectively to blood vessels of an organ, tumor, or defined site. The strategy has allowed the identification of peptides that selectively target the vasculature of normal tissues (4,5,22-26) and that of tumors. These results indicate that the vascular endothelium is modified by the tissue microenvironment in ways that allow differential targeting with circulating peptide ligands (23,26-29).

The system can also permit the identification endothelial cell surface receptors expressed in vivo. This can be accomplished by the discovery of the receptors corresponding to selected ligands. We have assembled a panel of peptide motifs that target the blood vessels of tumor xenografts. These motifs include the sequences ACDCRGDCFCG (termed RGD-4C), NGR, and GSL. The RGD-4C peptide binds selectively to av integrins and it has been shown to home to angiogenic vasculature (13). We identified an aminopeptidase (CD13) as a vascular receptor for the NGR motif (30). CD13 is strongly expressed in angiogenic blood vessels. Peptides that home to tumor vasculature have been used as carriers to deliver cytotoxic drugs (26), proapoptotic peptides (29), cytokines (33), and gene therapy (27,31). Generally, the coupling of homing peptides can yield targeted agents that are more effective and less toxic than the parental agents. Taken together, these results indicate that it is possible to isolate ligand-receptor pairs for targeted therapy by in vivo phage display.

Each of these peptides binds to different receptors that are selectively expressed on the vasculature of the target tissues. It is interesting to note that many of these tumor vascular markers are proteases. This finding is not surprising, given the invasive features of malignant tumors; in fact, some of the selective vascular markers are vascular proteases that not only serve as receptors for circulating ligands but also modulate angiogenesis. Also, it is intriguing that some of the markers also serve as viral receptors: alpha v integrins are receptors for adenoviruses, CD13 are receptors for coronaviruses, and MMP-2/MMP9 are receptors for echoviruses (4). It is tempting to speculate that bacteriophage--prokaryotic viruses--may use the same cellular receptors of eukaryotic viruses. In fact, the structure of the phage capsid protein provides good evidence bacteriophage share ancestry with animal viruses. More than an evolutionary biology footnote, these findings do suggest that the receptors isolated will have internalization capability, a key feature if one wishes to utilize peptide motifs as therapy carriers targeted to certain cell populations.

Exploring the heterogeneity of the blood vessels may further our understanding of tumor endothelium specificity and may define the role of endothelial cell markers play in angiogenesis and metastasis.

References.

(1)       Scott, J.K. and Smith, G.P., Science 249 (1990) 386.

(2)       Smith, G.P. and Scott, J.K., Methods in Enzymology 217 (1993) 228.

(3)       Arap, W., Pasqualini, R. and Ruoslahti, E., Current Opinion in Oncology 10 (1998) 560.

(4)       Koivunen, E., Arap, W., Rajotte, D., Lahdenranta, J. and Pasqualini, R., Journal of Nuclear Medicine 40 (1999) 883.

(5)       Pasqualini, R., Quarterly Journal of Nuclear Medicine 43 (1999) 159.

(6)       Folkman, J., Nature Medicine 1 (1995) 27.

(7)       Folkman, J., Nature Biotechnology 15 (1997) 510.

(8)       Rak, J.W., St Croix, B.D. and Kerbel, R.S., Anti-Cancer Drugs 6 (1995) 3.

(9)       Zetter, B.R., Annual Review of Medicine 49 (1998) 407.

(10)     Carmeliet, P. and Jain, R.K., Nature 407 (2000) 249.

(11)     Iruela-Arispe, M.L. and Dvorak, H.F., Thrombosis & Haemostasis 78 (1997) 672.

(12)     Mustonen, T. and Alitalo, K., Journal of Cell Biology 129 (1995) 895.

(13)     Brooks, P.C., Clark, R.A. and Cheresh, D.A., Science 264 (1994) 569.

(14)     Fidler, I.J. and Ellis, L.M., Nature Medicine 6 (2000) 500.

(15)     Jain, R.K., Microcirculation 4 (1997) 1.

(16)     McLean, J.W., Fox, E.A., Baluk, P., Bolton, P.B., Haskell, A., Pearlman, R., Thurston, G., Umemoto, E.Y. and McDonald, D.M., American Journal of Physiology 273 (1997) H387.

(17)     Thurston, G., McLean, J.W., Rizen, M., Baluk, P., Haskell, A., Murphy, T.J., Hanahan, D. and McDonald, D.M., Journal of Clinical Investigation 101 (1998) 1401.

(18)     O'Reilly, M.S., Boehm, T., Shing, Y., Fukai, N., Vasios, G., Lane, W.S., Flynn, E., Birkhead, J.R., Olsen, B.R. and Folkman, J., Cell 88 (1997) 277.

(19)     Folkman, J., in V.T. DeVita, Jr., S. Hellman, S.A. Rosenberg (Eds.), Cancer: Principles and Practice. Lippincott-Raven, Philadelphia-New York, 1997, p. 3075.

(20)     Folkman, J., Molecular Medicine 1 (1995) 120.

(21)     Watson, C.A., Camera-Benson, L., Palmer-Crocker, R. and Pober, J.S., Science 268 (1995) 447.

(22)     Pasqualini, R. and Ruoslahti, E., Nature 380 (1996) 364.

(23)     Pasqualini, R., Koivunen, E. and Ruoslahti, E., Nature Biotechnology 15 (1997) 542.

(24)     Rajotte, D., Arap, W., Hagedorn, M., Koivunen, E., Pasqualini, R. and Ruoslahti, E., Journal of Clinical Investigation 102 (1998) 430.

(25)     Rajotte, D. and Ruoslahti, E., Journal of Biological Chemistry 274 (1999) 11593.

(26)     Arap, W., Pasqualini, R. and Ruoslahti, E., Science 279 (1998) 377.

(27)     Trepel, M., Grifman, M., Weitzman, M.D. and Pasqualini, R., Human Gene Therapy 11 (2000) 1971.

(28)     Koivunen, E., Arap, W., Valtanen, H., Rainisalo, A., Medina, O.P., Heikkila, P., Kantor, C., Gahmberg, C.G., Salo, T., Konttinen, Y.T., Sorsa, T., Ruoslahti, E. and Pasqualini, R., Nature Biotechnology 17 (1999) 768.

(29)     Ellerby, H.M., Arap, W., Ellerby, L.M., Kain, R., Andrusiak, R., Rio, G.D., Krajewski, S., Lombardo, C.R., Rao, R., Ruoslahti, E., Bredesen, D.E. and Pasqualini, R., Nature Medicine 5 (1999) 1032.

(30)     Pasqualini, R., Koivunen, E., Kain, R., Lahdenranta, J., Sakamoto, M., Stryhn, A., Ashmun, R.A., Shapiro, L.H., Arap, W. and Ruoslahti, E., Cancer Research 60 (2000) 722.

(31)     Wickham, T.J., Haskard, D., Segal, D. and Kovesdi, I., Cancer Immunology, Immunotherapy 45 (1997) 149.

(32)     Burg, M.A., Pasqualini, R., Arap, W., Ruoslahti, E. and Stallcup, W.B., Cancer Research 59 (1999) 2869.

(33)     Curnis, F., Sacchi, A., Borgna, L., Magni, F., Gasparri, A., Corti, A., Nature Biotechnology 18 (2000) 1185.

 Human methionine aminopeptidase -- a novel target associated with inhibition of angiogenesis

In the last decade, the recognition of angiogenesis as a promising new anticancer target has proceeded simultaneously with the discovery and characterization of a novel anti-angiogenic anticancer drug class (fumagillin) and its target enzyme (methionine aminopeptidase 2, MetAp2). The parallel progress made in initial studies of fumagillin/angiogenesis and MetAp2 is reviewed below, followed by results of more recent studies conducted after the establishment of their enzyme-substrate relationship. Finally, some current ideas concerning cellular mechanisms and consequences of MetAp2 inhibition by fumagillin and related compounds are discussed.

MetAp2. Eukaryotic methionine aminopeptidases were cloned and characterized in the mid-1990s. Their reaction mechanisms have been extensively described and reviewed. These enzymes catalyse the co-translational removal of initiator methionines at N-termini of a subset of eukaryotic proteins. To be a substrate, a nascent protein must, at a minimum, contain a small, uncharged amino acid next after methionine at the N-terminus. This substrate specificity is highly conserved; the identity of actual cellular substrates remains under investigation.

In eukaryotes, two types (1 and 2) of methionine aminopeptidase have been defined, based upon sequence alignments. They are more similar to their prokaryotic counterparts than they are to each other, although their catalytic domains are homologous. In SDS-polyacrylamide gels, purified MetAp1 migrates as a polypeptide of molecular mass somewhat smaller than that of MetAp2 (ca. 40 kilodaltons vs ca. 60 kilodaltons). Purified MetAp2 is dependent on cobalt, as well as on zinc, manganese, and nickel; which of these metals are critical physiologically has not been established unambiguously.

In yeast, neither type 1 nor type 2 methionine aminopeptidase is essential, but the double null construct is lethal, suggesting that the two classes have overlapping roles. In the absence of evidence from knockout animals or highly selective inhibitors, it is not known whether similar functional redundancy exists in higher eukaryotes, including mammalian cells.

            Fumagillin and its analogues. Fumagillin is a natural product identified originally from a fungal (Aspergillus fumigatus) contamination of an endothelial cell culture, producing local rounding up of the endothelial cells. Fumagillin and certain structurally similar molecules, notably the semisynthetic analogue TNP-470, have become important tool compounds in the study of angiogenesis. They inhibit the growth of cultured endothelial cells at nanomolar or greater potency, but are inactive against most other cell lines at concentrations less than several micromolar. Selectivity for endothelial lines is not absolute, however, as recently a few non-endothelial cell lines have been found to be inhibited by nanomolar concentrations of fumagillin. Thus, although the cellular mechanism of fumagillin is not completely understood, it is characterised by inhibition of DNA synthesis (late G1 block) and by expression or activation of various cyclins and cyclin-dependent kinases. Cell strain-selective transcriptional effects are considered to be the ultimate step in the cell biology of fumagillin-like compounds. TNP-470 inhibits angiogenesis in vivo, and has demonstrated antitumour activity in several animal models. It is currently in Phase II/III clinical trial as a single agent as well as an agent in combination with conventional cytotoxics such as paclitaxel.

            Fumagillin and MetAp2. In the late 1990s, the cellular target of fumagillin was established to be MetAp2. Mammalian cell extracts subjected to conventional affinity chromatography yielded MetAp2 as the fumagillin binding protein. A second part of the study reported that among various methionine aminopeptidase-engineered yeast strains evaluated for sensitivity to fumagillin, only a deletion strain lacking the type 1 enzyme (DmetAp1) was inhibited. This finding is consistent with the inability of fumagillin to inhibit MetAp1 and with lethality of the double mutant (redundancy of function of type 1 and type 2 enzymes). Taken together, these biochemical and genetic data strongly indicate that fumagillin inhibits MetAp2 in yeast, leading to growth inhibition and ultimately to the death of the organism. Coincidentally, fumagillin was shown to be an irreversible inhibitor of MetAp2, but not MetAp1. Crystal structure studies of enzyme and enzyme-inhibitor complexes have provided elegant substantiation of this finding.

It is tempting to make the extrapolation that in vivo and clinical activity of TNP-470 is due to inhibition of MetAp2, and that later generation inhibitors of this enzyme will also be clinically active antitumour agents by virtue of their anti-angiogenic activity. While it is clear that MetAp2 is a target of fumagillin in higher eukaryotes, a compelling genetic component of the argument is lacking. Biochemical and biological data have established two important correlations, however, in the case of mammalian cells. First, limited structure-activity data based on TNP-470 show a correlation between MetAp2 inhibition and antiproliferative activity. Second, TNP-470 inhibits endothelial cell proliferation and fumagillin-MetAp2 binding with the same dose dependence. Moreover, antisense oligonucleotides to human MetAp2 block endothelial cell proliferation. All of this evidence suggests that MetAp2 is a critical target for the anti-angiogenic, and thus antitumour activity of fumagillin and related compounds.

            It remains, however, to establish how inhibition of MetAp2 by fumagillin results in the observed biological and pharmacological activity in mammalian cells. No correlation exists between cell strain sensitivity to fumagillin and related compounds and cellular content of MetAp1 and MetAp2, as both fumagillin sensitive and fumagillin insensitive strains contain similar amounts of type 1 and type 2 enzyme. Two attempts to explore the relationship between MetAp2 inhibition and biological effect have recently been described. Proteomics methodology has been employed to determine critical substrates of MetAp2 in cells by subtractive analysis (cells +/- fumagillin) of proteins subjected to two-dimensional gel electrophoresis. Several proteins were identified which appeared to be altered as anticipated under conditions of MetAp2 inhibition, but none had any obvious relevance to the DNA/cell cycle effects of fumagillin. A second approach, based on an early observation that TNP-470 inhibits activation of cyclin dependent kinases, led to the finding that this fumagillin analogue induces p53 activation and expression of p21CIP/WAF, a natural inhibitor of cyclin-dependent kinases. These inductions are reportedly specific for endothelial cells, providing a possible explanation for selective anti-angiogenic activity of fumagillin and related compounds. A clear relationship between this effect and MetAp2 inhibition is not evident from these data, however, and it is necessary to postulate that an event downstream of MetAp2 action is cell-type dependent and favors p53 activation and cell cycle blockade in endothelial cells. Additional studies of this sort should lead to a clearer understanding of the operation of fumagillin-like compounds on their target MetAp2 in a cellular setting.

Conclusion. Based on current knowledge and understanding, it appears very likely that in vivo and clinical activity of fumagillin analogues such as TNP-470 arises from inhibition of MetAp2 in cells, and that later stage inhibitors of this enzyme will also be clinically active antitumour agents by virtue of their anti-angiogenic activity. Current efforts are aimed at identifying such inhibitors and exploring the cellular consequences of MetAp2 inhibition to find the molecular basis of its link with angiogenesis.

Bradshaw, RA, Brickey, WW & Walker, KW (1998). N-terminal processing: the methionine aminopeptidase and Na acetyl transferase families. TIBS 23: 263 - 267.

Ingber, D, Fujita, T, Kishimoto, S, Sudo, K, Kanamaru, T, Brem, H & Folkman, J (1990). Synthetic analogues of fumagillin that inhibit angiogenesis and suppress tumour growth. Nature 348: 555 - 557.

Liu,S, Widom, J, Kemp, CW, Crews, CM & Clardy, J (1998). Structure of human methionine aminopeptidase-2 complexed with fumagillin. Science 282: 1324 - 1327.

Sin, N, Mheng, L, Wang, MQW, Wen, JJ, Bornmann, WG & Crews, CM (1997). The anti-angiogenic agent fumagillin covalently binds and inhibits the methionine aminopeptidase, MetAp2. PNAS(US)94: 6099 - 6103.

Yeh, J-R, Mohan, R & Crews, CM (2000). The antiangiogenic agent TNP-470 requires p53 and p21CIP/WAF for endothelial cell growth arrest. PNAS(US)97: 12782 - 12787.

            More recently, bone marrow (BM) transplantation (T) experiments have demonstrated the incorporation of BM-derived EPCs into foci of physiological and pathological neovascularization . Wild type mice were lethally irradiated and transplanted with BM harvested from transgenic mice in which constitutive LacZ expression is regulated by an EC-specific promoter, Flk-1 or Tie-2. The tissues in growing tumor, healing wound, ischemic skeletal and cardiac muscles and cornea micropocket surgery, have shown localization of Flk-1 or Tie-2 expressing endothelial lineage cells derived from BM in blood vessels and stroma around vasculatures. The similar incorporation was observed in physiological neovascularization in uterus endometrial formation following induced ovulation as well as estrogen administration.

Previous investigators have shown that wound trauma causes mobilization of hematopoietic cells, including pluripotent stem or progenitor cells in spleen, BM, and peripheral blood. Consistent with EPC/HSC common ancestry, recent data from our laboratory has shown that mobilization of BM-derived EPCs constitutes a natural response to tissue ischemia. The former murine BMT model presented the direct evidence of enhanced BM-derived EPC incorporation into foci of corneal neovascularization following the development of hindlimb ischemia. Light microscopic examination of corneas excised 6 days after micropocket injury and concurrent surgery to establish hindlimb ischemia demonstrated a statistically significant increase in cells expressing b-galactosidase in the corneas of mice with versus those without an ischemic limb. This finding indicates that circulating EPCs are mobilized endogenously in response to tissue ischemia following which they may be incorporated into neovascular foci to promote tissue repair.

These experimental findings call into question certain fundamental concepts regarding blood vessel growth and development in adult organisms. Postnatal neovascularization has been previously considered synonymous with proliferation and migration of pre-existing, fully differentiated ECs resident within parent vessels, i.e. angiogenesis. The finding that circulating EPCs may home to sites of neovascularization and differentiate into ECs in situ is consistent with vasculogenesis, a critical paradigm for establishment of the primordial vascular network in the embryo. While the proportional contributions of angiogenesis and vasculogenesis to postnatal neovascularization remain to be clarified, our findings together with the recent reports from other investigators suggest that growth and development of new blood vessels in the adult is not restricted to angiogenesis but encompasses both embryonic mechanisms. As a corollary, augmented or retarded neovascularization - whether endogenous or iatrogenic - likely includes enhancement or impairment of vasculogenesis.

Moreover, the observation that circulating EPCs home to foci of neovascularization suggests potential utility as autologous vectors for gene therapy. For treatment of regional ischemia, neovascularization could be amplified by transfection of EPCs to achieve highly localized constitutive expression of angiogenic cytokines and/or provisional matrix proteins. For anti-neoplastic therapies, EPCs could be transfected with or coupled to anti-tumor drugs or angiogenesis inhibitors.

            Future studies will clarify the mechanisms and circumstances that may be responsible for modulating the contribution of vasculogenesis to postnatal neovascularization. Specifically in this regard, it is intriguing to consider the possibility that certain angiogenic growth factors which are acknowledged to promote both angiogenesis and vasculogenesis in the embryo, but have been assumed to promote neovascularization exclusively by angiogenesis in the adult, may in fact promote migration, proliferation, and mobilization of EPCs from BM. The possibility that modulation of vasculogenesis can be used therapeutically to augment as well as inhibit neovascularization deserves further investigation.

            The recently discovered Angiopoietins join the vascular endothelial growth factor (VEGF) family as the only known growth factor families largely specific for vascular endothelial cells. A single member of the very large Ephrin family of growth factors, EphrinB2, also appears to have selective actions on blood vessels.

            Emerging data indicates that VEGF and the Angiopoietins work in complementary and coordinated fashion during normal vascular development and remodeling. While VEGF is critical for the initiation of vessel formation, Angiopoietin-1 seems to serve an important later role during vessel maturation and stabilization, by optimizing formation of the vessel wall. While the hypervascularity formed in the presence of excess VEGF is leaky and fragile (and associated with tissue edema and hemmorhage), vessels made in the presence of excess Angiopoietin-1 are actually resistant to vascular leak induced by VEGF or inflammatory mediators. Correcting an imbalance towards excess VEGF seen in many pathologic states, by either blocking the excess VEGF or administering additional Angiopoietin-1, could decrease plasma leakage and the resulting edema and thus have important clinical benefit in numerous disease settings, including diabetic retinopathy, tumor-associated ascites, brain edema associated with tumors or ischemic stroke, as well as in arthritis and other inflammatory conditions.

            Recent gene knockout studies suggest that Angiopoietin-2 also plays a key role within the vessel wall, in regulating vessel de-stabilization and vessel regressions, and in regulating development of lymphatics. Interestingly, recent work with EphrinB2 also suggest a key role in vessel wall formation, particularly for arterial vessels.

            Re-examination of tumor angiogenesis, in the context of considering the roles of these various factors, has led to a new view of how tumors interact with the vasculature. In contrast to prevailing dogma suggesting that tumors arise as avascular masses that require new angiogenesis for their initial vasculaturization and further growth, we suggest that tumors can instead grow by coopting existing vessels. Development of a potential therapeutic we term our VEGF Trap, which is perhaps the most potent VEGF antagonist described and can completely block tumor angiogenesis and endothelial proliferation, has allowed us to conclusively demonstrate that tumors can indeed grow invasively by vessel cooption in the absence of angiogenesis or endothelial proliferation; tumor growth is however limited by using the VEGF Trap to block new angiogenesis. In addition to regulating vessel formation and survival, the balance between VEGF and Angiopoietins seems to regulate the function and quality of tumor vessels - e.g., tumor vessels are often subject to excess VEGF and are thus leaky and fragile (and associated with tissue edema and hemmorhage), due to relative lack of the vessel stabilization, maturation and anti-permeability functions provided by Angiopoietin-1. In addition, Angiopoietin-2 and EphrinB2 are dramatically induced in tumor vessels, providing perhaps the best early markers of coopted or angiogenic tumor vessels, and thus targets of anti-angiogenesis approaches.

            The discovery of multiple new angiogenesis regulatory factors, together with novel characterization approaches made possible by the use of knockout and transgenic technologies, is clearly leading to a new understanding of the molecular basis of blood vessel development that seems likely to have important therapeutic implications.

References:

1.         Suri, C., Jones, P.F., Patan, S., Batunkova, S., Maisonpierre, P.C., Davis, S., Sato, T.N. & Yancopoulos, G.D. (1996) Requisite Role of Angiopoietin-1, a Ligand for the TIE2 Receptor, During Embryonic Angiogenesis. Cell, 87:1171-1180.

2.         Maisonpierre, P.C., Jones, P.F., Wiegand, S.J., Radziejewski, C., Compton, D., Aldrich, T.H., Papadopoulos, N., Daly, T.J., Sato, T.N., Davis, S. & Yancopoulos, G.D. (1997) Angiopoietin-2: A Natural Antagonist for Tie2 That Disrupts in vivo Angiogenesis. Science, 277:55-60.

3.         Holash, J., Maisonpierre, P.C., Compton, D., Boland, P., Alexander, C.R., Zagzag, D., Yancopoulos, G.D., & Weigand, S.J. (1999) Vessel Cooption, Regression, and Growth Tumors Mediated by Angiopoietins and VEGF. Science, 284:1994-1998.

4.         Thurston, G., Rudge, J., Ioffe, E., Zhou, H., Ross, L., Croll, S.D., Glazer, N., Holash, J., McDonald, D.M. & Yancopoulos, G.D. (2000) Angiopoietin-1 Protects the Adult Vasculature Against Plasma Leakage. Nature Medicine, 6:460-463.

5.  Yancopoulos, G.D., Davis, S., Gale, N.W., Rudge, J.S., Wiegand, S.J., Holash, J., (2000) Vascular-Specific Growth Factors and Blood Vessel Formation. Nature. 407:242-248

Background. Recombinant fibroblast growth factor-2 (rFGF-2) is a potent endothelial mitogen that improves perfusion in animal models of myocardial and hindlimb ischemia. This trial evaluates the effect of intraarterial rFGF-2 on a defined group of patients with intermittent claudication (IC).

Hypothesis: Intraarterial rFGF-2 increases exercise capacity in patients with moderate to severe IC due to infra-inguinal peripheral artery disease (PAD).

Methods: TRAFFIC is a Phase II, multicenter, randomized, double-blind, placebo-controlled, regimen-finding study of rFGF-2 or placebo. Eligible patients have stable moderate-severe IC for at least 4 months with reproducible baseline impairment in treadmill performance on paired examinations, depressed ankle-brachial index at rest, and angiographically-confirmed infrainguinal obstructive atherosclerosis without significant aortoiliac obstruction. Patients are ineligible if treadmill performance is limited by symptoms other than claudication, or if they have a history of prior malignancy, proteinuria, or other anticipated risks for investigational therapeutic angiogenesis. One third receive intraarterial rFGF-2 (30 mg/kg) on days 1 and 30, one third receive rFGF-2 on day 1 and placebo on day 30, and one-third receive placebo on both days. The primary endpoint is change in peak walking time (PWT) at day 90 on a Gardner ETT. Secondary endpoints include safety, change in claudication onset time and PWT, quality of life assessed using validated instruments, hemodynamic changes assessed using blood pressure indices and hyperemic plethysmography, and pharmacokinetics during a 180 day follow-up interval.

Discussion: TRAFFIC is the largest trial of therapeutic angiogenesis undertaken to date in the treatment of IC due to peripheral artery disease. Results of preliminary analysis after 90 and 180 days will be presented.

Tumour vasculature represents an appealing target for the development of new cancer treatments. Most of the research effort has been focused on antiangiogenic therapy. Many agents are currently undergoing clinical evaluation and even more are at the preclinical development phase. These agents target one or more of the key processes involved in angiogenesis e.g. endothelial cell proliferation, migration, and basement membrane degradation. Such agents can prevent the growth of new blood vessels but will have little or no effect on the vasculature already preexisting in the tumour at the time treatment commences. As a result, recently there has been an increased interest in developing agents, which irreversibly damage the already formed neovasculature in tumours. One of the most effective group of agents identified to date are certain tubulin depolmerising agents the most studied compound being Combretastatin A4 phosphate (CA4P) which is in Phase I trials.

CA4P has been demonstrated to induce rapid, selective and extensive blood flow reductions in a large number of experimental tumour systems. These effects have been observed in spontaneous as well as transplanted tumour models. More recently ongoing clinical studies have demonstrated that tumour blood flow reductions are also seen in humans following administration of CA4P. Although the mechanisms responsible for these effects have not been completely defined it has been shown that CA4P induces rapid shape changes in proliferating endothelial cells in culture. The fact that these changes occur over a similar time course to that seen with blood flow effects in vivo and that quiescent endothelial cells are much more resistant to such changes provides further evidence that this process is a key component in its action. These selective effects on cell shape reflect the critical role the tubulin cytoskeleton plays in maintenance of the elongated shape of endothelial cells when they are newly formed. Whilst a change in cell shape is not critical event for many cellular functions for an endothelial cell in a vessel such a change can result in rapid alterations in vessel function.

The ability of CA4P to induce blood flow changes in tumours at doses below that required to see antiproliferative/cytotoxic effects in normal tissues is in contrast to other tubulin depolymerising agents such as colchicine and the vinca alkaloids. In vitro the effect of CA4P on endothelial cell shape is reversed 4 hours after drug exposure unlike colchicine and vinblastine where changes persist for at least 24hours following drug exposure, reflecting the rapid reversibility of the tubulin binding properties of CA4P. In addition CA4P has a more rapid distribution and terminal half life in the plasma. These two factors reduced drug exposure to critical tissues and reversibility of action will minimize the antiproliferative efects of CA4P. However, since the vascular changes once initiated in vivo may lead to irreversible vascular dysfunction in the individual vessels due to additional effects like coagulation the vascular effects of CA4P can be fully retained.

Although CA4P can induce significant and extensive vascular shutdown in tumours it at best can induce stable disease when administered in a daily schedule to tumor bearing mice. This lack of overt response reflects the fact that even in tumours where complete vascular shutdown is induced a viable rim of tumour cells feeding off the normal vasculature surrounding the tumour remains. These cells and the stimulus from the localized hypoxia induced as a result of the ischemia result in continued growth and revascularisation. The most effective strategy to eliminate these remaining cells is to combine CA4P with conventional radiation and cytotoxic treatments. Several studies including those from our own laboratories have shown significant benefit using this approach. Interestingly whilst little or no response to CA4P has been observed in experimental systems when used as a single agent, clinical responses have been reported .

CA4P has been shown to be an effective neovascular damaging agent in a large number of experimental tumour systems however it is clear that there is a heterogeneous response with some tumours being more sensitive than others. One protective factor, which appears to be a determinant of response to CA4P and other vascular damaging agents, is nitric oxide (NO). NO plays a key role in several functions including preventing neutrophil adhesion to the vessel wall, increasing vessel patency and increasing cell survival during ischemic insult. Several studies in experimental tumors have estabished that NO synthase inhibitors can significantly enhance the activity of CA4P . Another factor, which may be a determinant of vascular response, is vessel maturity as determined by the number of vessels staining for smooth muscle actin. Identification of these modulators of response should help identify new strategies to enhance the activity of tumor vascular damaging agents.

In addition to cancer there are several other disease pathologies where abnormal neovascularisation is an integral part of disease progression these include ocular disorders ( macular degeneration and retinopathy), psoriasis and arthritis . Vascular Targeting approaches potentially offer a way of reversing the disease pathology by rendering non functional the newly formed vasculature. The results obtained with CA4P in an experimental model of ocular neovascularisation indicate it can not only stop but reverse the process of neovascularisation . Further studies evaluating the potential of CA4P in pathologies outside of oncology are now ongoing.

In summary, vascular targeting is a separate to antiangiogenic therapies since the primary aim is not to stop new vessel formation but to seek out and destroy newly formed and immature vessels. CA4P is the lead compound in the emerging field of tubulin binding vascular damaging agents (TBVDAs) . It has been established that the drug causes vascular dysfunction in a range of experimental tumours and also in tumours in humans. The mechanism of action centers around the critical role of tubulin in maintainance of cell shape in recently formed endothelial cells.

Fibroblast growth factors are a large family of pleiotropic regulators that comprises 23 members to date. The most extensively studied members are fibroblast growth factor-1 and 2. These factors are of average 18-30 kDa in size and have a receptor binding domain, a heparin binding domain, and potential phosphorylation sites. Some family members like FGF-4 have a classical signal sequence, while others like FGF-1 or FGF-2 do not. Besides the 18 kDa form, other forms are generated by alternate splicing or translation . The alternate translated forms found for instance in FGF-2 or FGF-3 are initiated at CUG codons and are called high molecular weight FGF (HMW FGF). These forms have an N-terminal extension with Gly-Arg-rich repeats and modified arginine residues (methyl arginines). This sequence is responsible for the nuclear accumulation of HMW FGF. Several studies on FGF-2 have attempted to define the respective roles of 18 kDa FGF and HMW FGF forms. While anyone agrees on the role of 18 kDa FGF in triggering migration and proliferation, conflicting results have been published on the role of the nuclear FGF. Some groups have claimed that HMW FGFs induce cell transformation (Arese et al, 1999 ; Bikfalvi et al, 1995 ; Galey et al., 1999), while others reported that they can trigger cell growth arrest (Dono et al., 1998). Interestingly, it has been shown that intracellular FGFs act in a receptor-independent fashion (Bikfalvi et al., 1995) and can bind transcription factors (Shen et al., 1998, H. Prats personal communication).

It has been known for long that in addition to many other mitogenic properties, exogenous FGF-2 is a good inducer of capillary formation both when added to cultured endothelial cells and in a number of in vivo situations where it induces neovascularization. However, the contribution of endogenous FGF molecules to developmental angiogenesis in vivo is still a matter of controversy.:It is true that FGF/FGF receptor expression studies could not clearly demonstrate a specific regulation during active angiogenesis whereas in the last decade, a growing family of so-called endothelial cell-specific regulatory pathways (VEGF/VEGFR, Angiopoietins/Tie and Ephrins) has emerged. All corresponding knock-outs display a clear vascularization deficiency that highlights the angiogenic step at which the molecules are involved. In the case for FGF/FGFR, gene disruptions have not been informative so far.

We undertook a series of systematic studies to evaluate the role of FGF during developmental and tumor angiogenesis. In addition, we investigated the inhibition of FGF activity by the endogenous angiogenesis inhibitor platelet factor-4 (PF-4).

I.    Role of FGF signaling during eye vascularization

The fact that fibroblast growth factor plays a crucial role in the development and maintenance of the embryonic vascular network has been recently demonstrated in vivo (Lee et al. 2000).

FGF is also implicated in vessel formation during eye organogenesis : last year, we reported that tyrp1-FGFRDN mice overexpressing a mutant FGFR1 receptor in the retinal epithelium suffer from defects of the eye vasculature (Rousseau et al. 2000). We now show that embryonic choroidal angiogenesis and retinal vasculogenesis are inhibited in these mice. The overall data allow to conclude that sprouting is significantly reduced in the choroid of transgenic embryos. Retinas on the other hand stay completely avascular in the first week after birth. Endothelial lineage labeling shows a defect in spindle cell differentiation and assembly. All in all, this model probably recapitulates major functions of FGF signaling at the onset of vasculogenesis and during angiogenesis. In future experiments we intend to elucidate the molecular mechanisms by which FGF triggers eye vascularization and will ask if these findings can be generalized to other developing organs.

II.     Role of FGF signaling in tumor growth and angiogenesis

We have then tested this model to ask whether FGF is required for tumor angiogenesis. Tyrp1-FGFRDN homozygous mice were crossed to tyrp1-SV40T hemizygotes that develop spontaneous embryonic tumors of the RPE (Penna et al. 1998). Proliferation foci and masses of neoplastic tissues at the ora serrata were identical in both tyrp1-SV40T and bigenic mice. This indicates that initial proliferation steps are not altered by FGF inhibition in SV40T transformed RPE cells. Two weeks after birth, tyrp1-SV40T tumors have switched to the angiogenic phenotype, migrate along the optic nerve to the brain and fill the entire eye posterior chamber. Mice are moribond at two months of age and are autopsied with tumors in the brain and metastasis in the inguinal lymph nodes and the spleen. We found that bigenic tumors stay small suggesting that they are inhibited at the onset of the angiogenic switch. Bigenic mice display a reduced index of secondary tumors and a doubled life-time expectancy.

Quantification of the tumor vasculature soon after switching show a decreased number and a heterogenous repartition of the vessels. We also established RPE cell lines from both groups and are currently characterizing their proliferation and angiogenic signaling properties. Recent data suggest that inhibition of tumor growth in bigenic mice is due to blockade of FGF-mediated angiogenic activity.

To examine the role of FGF signaling in a different tumor model, we chose to disrupt FGF signaling in cultured glioma tumor cells and then study the incidence of this manipulation on growth in vitro and in vivo and on angiogenesis (Auguste et al., 2001). We expressed dominant-negative FGF receptor R1 (FGFR1-DN) or R2 (FGFR2-DN) in glioma C6 cells by using stable or tetracycline regulated expression systems. Anchorage-independent growth was inhibited in FGFR1-DN or FGFR2-DN expressing cells. Tumor development after xenografting FGFR1-DN or FGR2-DN cells in immunodeficient animals was strongly inhibited. Transplantation of dominant negative FGFR expressing cells into rat brain yields to much smaller tumors than transplantation of control cells. Quantification of microvessels demonstrated a significant decrease in vessel number and density in tumors derived from dominant negative FGFR expressing cells. These results indicate that FGF signaling is involved in glioma tumor development by both angiogenesis-dependent and independent mechanisms.

Inhibition of FGF activity by platelet factor-4 or PF-4 derived peptides. Comparison with VEGF

Platelet factor-4 inhibits endothelial cell proliferation and angiogenesis in vitro and in vivo. The mechanism of action of these effects is poorly understood. We undertook a systematic study by examining in detail the interaction of PF-4 or PF-4-derived peptides with FGF-2 in comparison to VEGF (Perollet et al,1998; Jouan et al, 1999). We have shown that PF-4 inhibited binding of ¹²⁵I-FGF-2 or ¹²⁵I-VEGF to high affinity receptors, delayed ¹²⁵I-FGF-2s internalization and directly associated with FGF-2 or VEGF bound to surfaces or in solution. Furthermore, PF-4 inhibited FGF-2s dimerization. In addition, we demonstrate that a peptide between amino acid 47-70 containing the heparin-binding lysine-rich site inhibits FGF-2 or VEGF function. We have now characterized a small peptide domain (PF-4^47-70) derived from the C-terminus of PF-4 that conserves anti-angiogenic effects of the parent protein (Hagedorn et al., 2001). PF-4^47-70 inhibited internalization of ¹²⁵I-FGF-2 by endothelial cells in a time-dependent manner. The peptide reduced FGF-2-stimulated cell migration to control levels in wounded monolayers of bovine capillary endothelial cells.

PF-4^47-70 also reduced FGF-2 induced phosphorylation of MAP kinases ERK-1 and ERK-2, which are essential for migration and survival of endothelial cells. In a serum-free ex vivo angiogenesis assay, the peptide blocked microvessel outgrowth by 89%. A single amino acid substitution within PF-4^47-70 abolished all inhibitory activities. To simulate a real anti-angiogenic treatment situation, we administered PF-4^47-70 systemically to mice implanted subcutaneously with FGF-2 containing gelatin sponges with the result of sparse, scattered and immature vessel growth. The small peptide fragment derived from the angio-inhibitory CXC-chemokine PF-4 might be used as a starting point to develop anti-angiogenic designer drugs for angiogenesis-dependent pathologies like cancer, diabetic retinopathy and rheumatoid arthritis.

References