Abstract. Immunohistochemical methods were used to study the distribution of basic FGF in the 18-d rat fetus. The results reveal a pattern of widespread yet specific staining that is consistent with the wide distribution of basic FGE Immunoreactive basic FGF is associated with mesenchymal structures, mesodermand neuroectoderm-derived cells, and their extracellular matrices. As an example, skeletal and smooth muscle cells are strongly positive. The basement membrane underlying the epithelia always contain basic FGE In some tissues (i.e., cartilage and bone) the intensity of immunostaining is dependent on the stage of cell differentiation.Although the staining of tissues is primarily associated with the extracellular matrix, there is significant intracellular staining in various cell types. This is particularly evident in the endocrine cells of the adrenal cortex, testis, and ovary. The histochemical findings reported here support the notion that basic FGF has the characteristics required to mediate many of the effects of the mesenchyme on cell growth and differentiation. The significance of these findings in understanding the role of basic FGF in regulating cell proliferation and differentiation is discussed.

The presence of an angiogenic protein basic fibroblast growth factor (FGF) was established in juvenile nasopharyngeal angiofibroma (JNF). Extracts of these tumors have the capacity to stimulate endothelial cell proliferation. This activity is indistinguishable from basic FGF. The biological activity contained in the extracts binds to heparin-Sepharose columns and is eluted with a characteristic 2 mol sodium chloride. The exact fraction of the biological activity corresponds to the location where an immunoreactive basic FGF can be detected by radioimmunoassay. These same fractions contain an 18,000-d molecule which is identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting with an antibody to basic FGF. Indeed, immunohistochemical studies localize the growth factor to the endothelium of JNF. Although these findings do not establish that basic FGF mediates the development of this angiofibroma, they do support the possibility that the pathogenesis of JNF is associated with the presence of angiogenic factors like basic FGF. If this is the case, a comprehensive study of the etiology of JNF may lead to a better understanding of how locally produced growth factors mediate proliferative disease and how its modification might lead to better treatment on a biological basis.

A major goal of gene therapy is to improve target specificity by delivering vectors through alternative cellular receptors. We previously reported that adenoviral vector delivery through basic fibroblast growth factor (FGF2) receptors enhances both cellular transduction and in vivo efficacy. We now present studies addressing the cellular pathways and mechanisms underlying these events. Cellular receptors for adenoviruses are not required for transduction by FGF2-retargeted vectors. Moreover, alpha(V) integrins can antagonize FGF2 retargeting, in contrast to their obligatory role in non-retargeted vector delivery. By contrast, high-affinity FGF receptors, which are overexpressed on potential tumor targets, are required for FGF2-retargeted transduction. Low-affinity heparan sulfate proteoglycan interactions, however, are not a prerequisite, in marked contrast to their obligatory role in FGF2 mitogenic signaling. By comparing receptor expression and ligand binding with transgene expression, we also demonstrate that FGF2 retargeting enhances transduction by mechanisms other than increasing the number of targeted cells. Rather, the use of alternative targeting ligands supports the conclusion that specific receptor interactions and intracellular events serve to enhance transgene expression. Together, these studies highlight the unique delivery and transduction pathways used by FGF2-retargeted adenoviruses, and help define the basis for their enhanced in vivo efficacy.

Targeted vectors provide a number of advantages for systemic and local gene delivery strategies. Several groups have investigated the utility of using various ligands to alter the tropism of adenovirus (Ad) vectors. We have previously demonstrated that fibroblast growth factor (FGF) ligands can specifically target DNA transfection and Ad transduction through high-affinity FGF receptors (FGFRs). FGFRs are overexpressed in abnormally proliferating tissues, such as malignancies. The present studies explore the effects of retargeting with FGF2 on the tissue localization pattern and the systemic toxicity of Ad in mice. Results of semiquantitative PCR analyses indicate that the distribution of FGF2-Ad vector genome sequences after intravenous administration in mice is altered. Markedly lower amounts (10- to 20-fold) of FGF2-Ad localize to the liver when compared with native Ad. This decrease in liver deposition translates into a significant reduction in subsequent toxicity as measured by serum transaminases and histopathology in mice injected with FGF2-AdHSV-thymidine kinase with and without ganciclovir administration. In an intraperitoneal model of ovarian cancer, FGF2-Ad generates increased transgene expression in tumor tissue when compared with Ad. Taken together, these results indicate that the retargeting of Ad with FGF2 results in a more efficient vector system for systemic and regional gene therapy applications, with concomitant lower levels of systemic toxicity.

The neurotrophic activity of astrocytes and fibroblasts and its regulation by various cytokines were investigated. Astrocyte conditioned medium (ACM) enhanced the survival of neurons and the proliferation of astrocytes in embryonic cortical cultures grown in serum-free defined medium. However, these results were not affected by acidic fibroblast growth factor, interleukin-1 beta (IL-1 beta), tumor necrosis factor-alpha (TNF alpha), and transforming growth factor-beta 1. In contrast, ACM induced choline acetyltransferase expression in septal cholinergic neurons via nerve growth factor (NGF)-dependent and -independent mechanisms. However, neither acidic nor basic fibroblast growth factor is involved in this biological activity in ACM. The cytokines listed above mainly stimulate NGF-mediated cholinergic neurotrophic activity in ACM. A combination of IL-1 beta and TNF alpha significantly enhanced choline acetyltransferase activity in septal neurons co-cultured with astrocytes, and this effect was found to be mediated by NGF produced by activated astrocytes. Effects of astrocytes on GABAergic neurons were also examined. ACM was found to increase glutamate decarboxylase activity in neuronal cultures from septum in the presence of Ara-C. However, the cytokines did not enhance this activity in ACM. Moreover, a combination of IL-1 beta and TNF alpha had no effect on glutamate decarboxylase activity in septal neurons co-cultured with astrocytes. In a final set of experiments, cholinergic neurotrophic activity in skin-derived fibroblast conditioned medium (FCM) was examined. FCM was found to possess biological activity similar to that of ACM on septal neurons grown in serum-free defined medium with Ara-C. The cytokines also enhanced NGF-mediated cholinergic neurotrophic activity in FCM. Astrocytes and fibroblasts were found to possess NGF-type and non-NGF-type cholinergic neurotrophic activity, and various cytokines were found to regulate the NGF-type cholinergic neurotrophic activity in both types of cells. NGF produced by astrocytes and fibroblasts that are activated by cytokines is likely to be important for development and regeneration of NGF-sensitive neurons in the central and peripheral nervous systems.