Fgfs In Tumor Angiogenesis

Formation of new blood vessels from pre-existing blood vessels (angiogenesis) is a multistep process. Stages in this process include the activation of quiescent endothelial cells in a pre-existing vessel, degradation of the basement membrane, migration of endo-thelial cells into the interstitial space and sprouting, endothelial cell proliferation at the migrating tip, lumen formation, generation of new basement membrane with the recruitment of pericytes, formation of anastomoses, and, finally, blood flow (75).

FGF-1 and FGF-2 are mitogenic for endothelial cells and stimulate endothelial cell migration (1,2,75). They induce endothelial cell production of collagenase and plasminogen activator that are capable of degrading basement membranes (76). FGF-1 and FGF-2 also cause endothelial cells to migrate and form capillary-like tubes in three dimensional culture systems in vitro (77). Both FGF-1 and FGF-2 exhibit potent angiogenic activity in all in vivo assays tested. In fact, they are usually employed in these bioassays as positive references. Endothelial cells derived from large or small vessels are not only a target for FGF-2, but they also synthesize high amounts of FGF-2 in vitro (78,79). Most of the FGF-2 produced by the cells is found associated with the extracellular matrix and subendothelial basement membrane (53). Spontaneous migration of endothelial cells was inhibited by neutralizing antibodies to FGF-2, suggesting an autocrine role of FGF synthesized and released by the endothelial cells themselves (76). A similar autocrine role of endothelial cell-produced FGF-2 has been shown to promote tube formation in collagen gels (77).

Angiogenic sprouting from rat aorta specimen, cultured in collagen, was stimulated by addition of exogenous FGF-2, and was inhibited by neutralizing antibodies to FGF-2 (80). Apparently, FGF-2 was released by experimental wounding, and the released FGF-2 mediated autocrine stimulation of angiogenesis after injury. FGF-2, packaged into polymer-based pellets and implanted under the kidney capsule, in the mouse cornea, the periadventitial space of rat carotid artery, or in the rat mesenteric window, caused a dose-dependent stimulation of angiogenesis (81). The results summarized above clearly demonstrate that FGF-1 or FGF-2 are able to directly induce angiogenesis in vitro and in vivo. However, it is not known how FGFs interact with the target endothelial cells in vivo, or which of their pleiotropic activities is actually required for the induction of angiogenesis in vivo. For instance, it has not been demonstrated that FGF-1 or FGF-2 are directly involved in physiological or pathological angiogenesis, such as ovulation, implantation, wound healing, or vascular disease. Most of the accumulated data has been based predominantly on expression studies that attempted to correlate expression of FGFs with angiogenic processes. Although the involvement of members of the FGF family in embryonic development is undisputed, a functional role for FGFs in developmental angiogen-esis has not been unequivocally demonstrated. Conclusions are based chiefly on the spatial and temporal correlation of FGF expression and neovascularization.

Since FGF-2 and, to a lesser extent, FGF-1 are expressed by many tumors in vivo, and by tumor cell lines in vitro, it has been assumed that they were responsible for the induction of angiogenesis during tumor progression. This assumption, however, is oversimplified, since, in some cases, other angiogenic factors, such as VEGF, have been shown to play a lead role. In some instances, expression of FGF-1 or FGF-2 in vivo correlates with the degree of vascularity of the tumors. However, at the same time, FGFs have other target cell specificities. In most cases, the targets are the tumor cells themselves, or cells within remodeling stroma, leading to an autocrine or paracrine stimulation of tumor-cell proliferation. Therefore, it has been difficult to dissect the pleiotropic activities of FGFs in functional terms, and to assess their actual role in tumor angiogen-esis in vivo.

In order to induce tumor angiogenesis, FGFs produced by the tumor cells need to be released to bind FGF receptors on the surface of endothelial cells. However, both FGF-1 and FGF-2 lack a signal sequence for secretion, and, although several mechanisms for their release from cells have been proposed, their export pathway has not been elucidated (see Section 5.3.). It is possible that FGFs are produced by cells other than tumor cells. For example, an increase in FGF-2 synthesis in the tumor vasculature has been observed, indicating that endothelial cells themselves could be a source of FGF-2 (81). Tumors also recruit macrophages, and activate them to secrete FGF-2 (82). Furthermore, mast cells may also be recruited by tumors. They are loaded with heparin, which they might release in the tumor, and amplify the effects of FGF-1 or FGF-2 (83).

In many naturally growing, as well as in experimentally induced, tumors, the extent of vascularization correlates with the expression of angiogenic FGF family members. For example, neural transplants that have been retrovirally transfected with the FGF-4 gene exhibit abundant capillary proliferation, and induce the formation of capillary angio-

sarcomas, suggesting that FGF-4 may have a direct role in angiogenesis and endothelial cell transformation (84). Moreover, in glioblastoma and meningioma, the levels of expression of FGF-2 and FGFR-1 correlate not only with tumor cell proliferation, but also with vascularity of tumors (15). FGF-1 and FGF-2 have been found to be highly expressed in KS lesions of AIDS patients. Indeed, FGF-2 injected into nude mouse elicited the formation of KS-like lesions regarding the extent of vascularization, the morphology of proliferating cells, and general pathology. In these experiments, FGF-2 synergized with the Tat gene product of human immunodeficiency virus and caused KS-like lesions. Fibronectin could replace Tat, suggesting that FGF-2 induced the lesions, and that Tat enhanced its activity by mimicking the effect of ECM molecules (85).

Quantification of angiogenic proteins in the blood and urine of cancer patients may measure progression of disease and guide therapy. Based on detection of endothelial cell stimulators in the urine of cancer patients (86), an immunoassay for FGF-2 was developed that revealed elevated levels of the angiogenic protein in the serum of patients with renal cell carcinoma (87). Elevated levels of FGF-2, but not FGF-1, were found in the serum and urine of bladder and hepatocellular carcinoma patients, with the highest levels in patients with active metastatic disease (87,88). High levels of FGF-2 were found in the serum of approx 10% of a wide spectrum of cancer patients (89), and in the urine of more than 37% of cancer patients (90). Biologically active FGF-2 was abnormally elevated in the CSF of children with brain tumors, but not in children with hydrocephalus or malignant disease outside of the CNS (91). The FGF-2 level in CSF correlated with microvessel density in histologic sections, which itself provided a prognostic indicator of risk of mortality (91). Also, FGF-2 levels in the urine of children with Wilms' tumor correlated with stage of disease and tumor grade (92). In some tumors, tissue levels of angiogenic proteins have correlated with severity of disease or outcome. Immunohistochemical levels of FGF-2 in renal carcinoma correlated with the risk of death (93). FGF-2 mRNA was detected in 80% of patients with renal cell carcinoma. FGF was suggested to induce microvessel tube formation in these patient tumors (94). In infants with hemangiomas, urinary FGF-2 levels were abnormally elevated, and returned toward normal with involution of the lesions. In life-threatening hemangiomas treated with IFN-a-2a, quantification of urine FGF-2 has been a useful way of determining an effective dose, and of differentiating between hemangioma and vascular malformation (95).

In an experimental setting, Balb/c 3T3 cells that expressed a mutant, but active, form of FGF-2 were transplanted into athymic mice, resulting in elevated levels of the mutant form of FGF-2 in the urine of tumor bearing mice (96). These experiments suggest that the source of circulating FGF in the serum and urine originated from the tumor itself. However, in cancer patients, in addition to the export of FGF-2 from tumor cells, FGF-2 may also be mobilized from extracellular matrix by tumor-derived heparinases or colla-genases. FGF-2 and other angiogenic peptides could also be released from host cells, such as macrophages, recruited into the tumor (82). Another puzzle is why FGF-2 remains elevated in the serum of tumor patients when it is normally cleared within approx 30 min after intravenous injection (97). The normal clearance mechanisms for FGF-2 (98) may be saturated or disturbed in cancer patients, but these systems remain to be studied. Circulating FGF-2 may be bound to soluble receptors. It is not known whether abnormally elevated levels of circulating FGF-2 maintained over prolonged periods of time may potentiate growth of dormant metastases. The presence of FGF-2 in body fluids of cancer patients and experimental animals correlated significantly with the extent of vascularization and metastasis, suggesting that expression and release of FGFs might be involved in tumor angiogenesis. Based on these results, the detection of FGF in the urine of cancer patients might be a useful diagnostic and prognostic tool. The results also indicate that FGF-2 is released by producer cells, despite the lack of a signal peptide.

0 0

Post a comment