Ultrastructural Immunolocalization of Basic Fibroblast Growth Factor in Mast Cell Secretory Granules: Morphological Evidence for bFGF Release Through Degranulation

Abstract

We previously reported that mast cells (MCs) serve as a source of basic fibroblast growth factor (bFGF), a potent angiogenic and mitogenic polypeptide, suggesting that bFGF may mediate MC-related neovascularization and fibroproliferation. Unlike many other growth factors, bFGF lacks a classic peptide sequence for its secretion, and the mechanism(s) for its release remains controversial. Because MCs release a wide spectrum of bioactive products via degranulation, we hypothesized that MC degranulation may be a mechanism of bFGF release and used ultrastructural immunohistochemistry to test the hypothesis. We reasoned that if bFGF is released through degranulation, it should be localized to MC secretory granules. Human tissues with chronic inflammation and rat/mouse tissues with anaphylaxis were studied. In all tissue samples examined, positive staining (or immunogold particle localization) for bFGF in MCs was predominantly in the cytoplasmic granules. Moderate bFGF immunoreactivity was also found in the nucleus, whereas the cytosol and other subcellular organelles exhibited minimal immunogold particle localization. In contrast, no immunogold particle localization for bFGF was observed in lymphocytes or plasma cells. In rat/mouse lingual tissue undergoing anaphylaxis, immunogold particle localization for bFGF was found not only in swollen cytoplasmic granules but also in the extruded granules of MCs. Three different anti-bFGF antibodies gave similar immunogold particle localization patterns, whereas all controls were negative. These results provide morphological evidence suggesting that, despite the lack of a classic secretory peptide in its structure, bFGF is localized to the secretory granules in MCs and may be released through degranulation.

Keywords

mast cell basic fibroblast growth factor (bFGF)electron microscopy degranulation

Basic fibroblast growth factor (bFGF) is a member of a heparin binding growth factor family (fibroblast growth factor family) that includes at least eight other members sharing 20-40% homology in their gene structure (Burgess and Maciag 1989; Baird 1991). Basic FGF is a potent mitogenic factor for a broad spectrum of cell types, including endothelial cells, fibroblasts, chondrocytes, osteoblasts, and synoviocytes (Melnyk et al. 1990; Goddard et al. 1992). Basic FGF is also one of the most potent angiogenic factors known and has been shown to induce neovascularization in vitro (Montesano et al. 1986) and in vivo (Lindner and Reidy 1991). bFGF is characterized by its high affinity for heparin. Binding to heparin also stabilizes bFGF and potentiates its bioactivities (Saksela et al. 1988). Unlike other polypeptide growth factors that exert their biological effects through a paracrine mechanism, bFGF lacks a conventional secretory sequence in its structure, and its release mechanism is still a subject of controversy. Its lack of a conventional secretory peptide, together with its increased extracellular level after cell injury, has led to the hypothesis that compromised plasma membrane integrity is the mechanism for bFGF release.

Several lines of evidence indicate that the release of bFGF at both the cellular and the tissue level is tightly controlled and that loss of such control may be deleterious. First, increased free bFGF has been linked to neoplasms. For example, increased levels of bFGF were found in cerebral spinal fluid (CSF) in 62% patients with brain tumors and correlated well with the vascular density of the tumors. In contrast, bFGF was not detectable in the normal controls (Li et al. 1994). High-level expression of bFGF occurs in primary Kaposi's sarcoma (AIDS-KS), and tumor cells from Kaposi's sarcoma constitutively produce and release bioactive bFGF in the absence of cell death (Bailer et al. 1995; Samaniego et al. 1995). Second, a switch to the export of bFGF has been implicated in malignant transition in vivo. In a transgenic mouse model of fibrosarcoma, Kandel et al. (1991) found that although bFGF is expressed in normal dermis and in the early mild lesion of fibromatosis, a switch to the export of intracellular bFGF occurs concomitantly both with the neovascularization characteristic of aggressive fibrosarcoma seen in vivo and with the tumorigenicity of cell lines derived from the lesion. Third, NIH 3T3 fibroblasts transfected with native bFGF exhibit no morphological alteration or tumorigenicity despite expression of high levels of bFGF (Blam et al. 1988; Rogelj et al. 1988). However, cells transfected with the bFGF gene fused with a secretory signal sequence undergo morphological alteration and display tumorigenicity. The transformed cells cause rapidly growing tumors within 10 days in 100% of syngenic and nude mice. In nude mice, the tumors are highly vascularized (Rogelj et al. 1988,1989). Similar results are obtained with the human endometrial adenocarcinoma HEC-1-B cell line (Coltrini et al. 1995). When introduced into Xenopus embryos, bFGF mRNA for a secretory form of bFGF has been shown to be 100-fold more potent in eliciting aberrant mesoderm formation than its native counterpart (Thompson and Slack 1992). Results from these studies indicate that the mechanism(s) of bFGF secretion exerts a critical control of the bioactivities of this multipotent growth factor.

It has been well documented that cell injury results in bFGF release. However, injury also results in release of many other intracellular components and growth factors that are normally secreted, such as transforming growth factorb (TGF-β) and plateletderived growth factor (PDGF). Although the exact role(s) of bFGF remains unknown, its wide tissue distribution, broad spectrum of target cells, and mutation lethality in transgenic animals strongly suggest its important roles in normal homeostasis. It is reasonable to believe that release mechanisms other than cell injury play the major role in controlling the export of bFGF in normal homeostasis. A study by Mignatti et al. (1991) shows that bFGF can be released from single uninjured cells and mediates functions in an autocrine manner. Aortic smooth muscle cells release bFGFlike mitogen under shear stress (Sterpetti et al. 1994). In bFGF-transfected COS-1 cells, an energy-dependent, non-ER/Golgi pathway is reported to direct the export of bFGF (Florkiewicz et al. 1995). We and others demonstrated that mast cells express bFGF and serve as a major source of tissue bFGF in several disease conditions (Qu et al. 1995; Reed et al. 1995). Mast cells have been linked for almost two decades to neovascularization and fibroproliferative diseases. However, the exact contribution of mast cells to these disease processes remains largely unknown. Previous studies by Reed et al. and by us strongly suggest that bFGF mediates angiogenesis and fibroproliferation associated with mast cells (Qu et al. 1995; Reed et al. 1995). Because mast cells release several bioactive substances by degranulation, we hypothesize that bFGF can be released by mast cells through degranulation. In the present study we tested this hypothesis by immunolocalization of bFGF in secretory granules in both resting and activated mast cells. We found that bFGF immunoreactivity was localized in the cytoplasmic granules of mast cells in human tissues from lung, synovium, nasal mucosa, and skin, and in lingual tissues from rats and mouse. Immunogold particle localization for bFGF was also present in the extruded secretory granules of mast cells in rat/mouse tongue undergoing passive anaphylactic reaction. The results provide morphological evidence for bFGF release through a noninjury mechanism.

Materials and Methods

Tissue Processing

Fresh synovium (n = 4), lung (n = 3), and nasal mucosal membrane (n = 2) samples were obtained during surgery from patients with rheumatoid arthritis, pulmonary fibrosis, and nasal polyposis, respectively. All samples were divided into two parts and processed for light and electron microscopy separately. For light microscopy (LM), the samples were fixed in neutralized buffered formalin for 24-30 hr at 4C, dehydrated, and embedded in paraffin.

For electron microscopy (EM), tissue samples were sliced into 1-3-mm³ pieces and fixed with 4% paraformaldehyde-0.5% glutaraldehyde in 0.1 M Sörensen's phosphate buffer for 4 hr at 4C. From each case, some sample pieces were also postfixed in 1% OsO₄ for 1 hr at room temperature (RT). After extensive washes in the phosphate buffer, the samples underwent ethanol dehydration to propylene oxide and then were embedded in Araldite 502 (Electron Microscopic Sciences; Ft Washington, PA).

In Vivo Mast Cell Degranulation

To induce degranulation in vivo, direct lingual anaphylactic assay was performed on female Brown Norway rats weighing 250-300 g (n = 4). Rats were anesthetized with isoflurane/O₂, delivered with an anesthetic vaporizer, and about 15 μg goat anti-rat IgE antibody in 100 μl PBS was injected into the tongue along the midline. The control rats (n = 3) were injected with an equal amount of nonimmune goat IgG. Anesthetized rats were sacrificed 1 or 2 hr after the injection by iv perfusion with 0.1 M Na-citrate containing 15% glycerol at 37C, pH 7.4, followed by 4% paraformaldehyde-0.2% glutaraldehyde in 0.1 M PBS at 4C. The rats were kept at 4C overnight. The following day, lingual tissue was dissected out and processed for both LM and EM as described above.

In vivo mast cell degranulation in mice was induced using a passive cutaneous anaphylactic assay previously described by Wershil et al. (1991). Briefly, 20 μl serum-free Dulbecco's modified Eagle's medium (DMEM) containing 100 ng of mouse monoclonal IgE anti-dinitrophenyl (DNP) (Sigma; St Louis, MO) was injected into the ear pinna of 3-month old Balb-c/CBYJ male mice (Jackson Labs; Bar Harbor, ME) anesthetized with isoflurane/O₂. After 30 hr, the mice were challenged iv with 100 μg of DNP conjugated to human serum albumin (DNP-HSA) (Sigma) in 200 μl of serum-free DMEM containing 0.5% Evan's blue, pH 7.4. Extravasation of Evan's blue dye was used as an indicator of the anaphylaxis reaction. Control animals were given either the IgE or DNP-HSA only. Anesthetized mice were sacrificed at 1, 2, or 3 hr after the challenge by iv perfusion and the ear pinna was processed as described above.

Histochemical and Immunohistochemical Staining

Paraffin-embedded tissues were cut into 2-3-μm sections and laid on poly-l-lysine-coated slides. The sections were deparaffinized in xylene and rehydrated through graded alcohol. To visualize tissue mast cells, the sections were stained with 0.1% toluidine blue (Sigma) in 0.1 N HCl for 1 hr at RT. Adjacent sequential sections were stained for bFGF as described previously (Qu et al. 1995). In brief, the sections were treated for 30 min at RT with hyaluronidase (Sigma Cat. 6254; Type V from sheep testis) at 2 mg/ml in 0.1 M acetate buffer, pH 5.2, to unmask antigens. After washing with PBS containing 0.05% Tween-20, the sections were treated with blocking solution preabsorbed with heparin gel (Bio-Rad Laboratories; Hercules, CA) for 20 min at RT to block nonspecific binding sites in the tissue. The blocking solution consists of 0.1 M PBS, 2% crystal bovine serum albumin, 2% normal goat serum, 0.1% Tween-20, and 0.05% NaN₃. Then sections were incubated with the primary antibodies overnight at 4C. The antibodies used include two rabbit anti-human bFGF antibodies (Biomedical Technologies; Stoughton, MA, and Sigma) at 1 μg/ml, mouse monoclonal antibody to human tryptase at 0.1 μg/ml (Dako; Carpinteria, CA), and a mouse monoclonal anti-human bFGF (148.6.1.1.1) in ascites fluid (a gift from Dr. C. Hart; Zymo-Genetics, Seattle, WA) at 1:75,000 dilution. A three-step avidin-biotin complex (ABC) (Vector Labs; Burlingame, CA) method or the APAAP method (Dako) was carried out to detect the antigen-antibody complex according to the vendors' instructions. The antibody-antigen complexes were visualized by incubation for 30 min in Fast Red substrate (Bio-Genex; San Ramon, CA) or Vector Blue (Vector) containing 2 mM levamisole according to the suppliers' instructions. The sections were counterstained with Gill III hematoxylin and mounted in Crystal Mount (Biømeda; Foster City, CA).

Postembedding Immunogold Labeling for Electron Microscopy

Ultrathin sections were cut at 50-70 nm, collected on 200- mesh nickel grids coated with Coat-Quick “G” (Electron Microscopic Sciences), and immunolabeled with three different antibodies specific for bFGF, using a two-step procedure. All immunogold steps were carried out at RT in 70-μl reagent drops on moisture-resistant parafilm (American National Can; Greenwich, CT). The grids were first treated with hyaluronidase at 2 mg/ml in 0.1 M acetate buffer, pH 5.2, for 15 min to unmask antigens (Qu et al. 1995). After two 5-min washes in Tris-buffered saline (TBS), pH 7.5, containing 0.1% Triton X-100, nonspecific binding sites were blocked by incubation for 20 min in blocking solution as described above. Then the grids were incubated in a primary anti-bFGF antibody (see Table 1) in the blocking solution for 2 hr, followed by two 5-min washes and one 30-min wash with TBS and 7.5- and 5-min washes in TBS buffer, pH 8.2. The grids were incubated for 1 hr with a goat antirabbit IgG antibody conjugated with 5- or 10-nm gold particles (Goldmark Biologicals; Philipsburg, NY) in TBS buffer, pH 8.2. After several washes in TBS and dH₂O, the grids were air-dried and then counterstained with aqueous uranyl acetate. Controls included substitution of nonimmune IgG for primary antibodies, preabsorption of the primary antibodies with recombinant bFGF, or omission of the primary antibody.

Results

Localization of bFGF in Rat and Mouse Tissue After Cutaneous Anaphylaxis

On intralingual injection of anti-Ig E antibody or iv challenge of antigen-sensitized rats or mice, the site of antigen injection became blue within 30 min. The local coloration by Evan's blue remained unaffected by iv perfusion with glycerol/citrate buffer and fixative solution, suggesting extravasation of Evan's blue dye into the extravascular compartment. No such change was observed in nonsensitized animals or sensitized animals challenged with Evan's blue only (data not shown). When the Evan's blue dye-containing tissues were examined microscopically, extrusion of mast cell granules was readily observed (Figures 1A and 1B). Strong immunohistochemical staining for bFGF was found in both mast cells and their extruded granules (Figure 1A), which were identified by toluidine blue stain on an adjacent section (Figure 1B). In contrast, bFGF immunoreactivity was not found to be associated with neutrophils, lymphocytes, or plasma cells. Similar findings were observed in the lingual tissue after local injection of anti-IgE antibody. Increased extracellular staining for bFGF was also found in areas near the degranulating mast cells. In the lingual tissues from control rats, positive staining for bFGF was mainly associated with the cytoplasmic granules confined within the cells that were also labeled by toluidine blue dye (Figures 1C and 1D), indicating that they are mast cells. No apparent granule extrusion or increased extracellular staining for bFGF were present in the control animals. In all tissues from rats and mice examined, no tissue necrosis was seen.

Table 1

Antibodies used for the study

		Working concentrations
Anti-bFGF antibody	Source	LM	EM
Rabbit IgG	Biomedical Technologies, Stoughton, MA	1 μg/ml
Rabbit IgG	Santa Cruz Biotechnologies, Santa Cruz, CA	1 μg/ml	3 μg/ml
Rabbit IgG	Sigma Chemical, St Louis, MO	0.2 mg/ml	3.5 μg/ml
Mouse IgG	Sigma	0.5 mg/ml	ND
Mouse ascites	ZymoGenics, Seattle, WA	1:75,000	1:500

To identify the subcellular compartment associated with bFGF-like immunoreactivity, tissue samples were examined by EM. Tissue mast cells were identified by metachromasia after toluidine blue staining of semithin sections and by characteristic ultrastructure, such as typical plicated membrane folds, conspicuous membrane- bound cytoplasmic granules, and nonlobed nucleus (Figure 2). Immunogold particle localization of bFGF was found predominantly in the cytoplasmic granules of MCs (Figure 2). However, no particular distribution pattern of the bFGF immunoreactivity within the cytoplasmic granules was observed. Similarly, moderate bFGF immunoreactivity was also found in the nucleus, whereas the cytosol and other subcellular organelles were essentially devoid of immunogold particles for bFGF in MCs. Only background level of immunogold particle localization was found in cytoplasm and mitochondria (Figure 3A). Moderate to little immunogold particle localization was present in the cytoplasm and nucleus of fibroblasts and endothelial cells (data not shown). In contrast, no immunogold particle localization for bFGF was observed in lymphocytes or plasma cells.

Figure 1

Localization of mast cell-associated bFGF by immunohistochemistry. Paired sequential sections at 2-μm thickness of lingual tissue from rat with anaphylaxis (A,B) and from control rat (C,D) were stained with anti-bFGF antibody (A,C). Positive staining for bFGF was present in MCs (A,C) that were identified by toluidine blue dye in adjacent sections (B,D). Arrows indicate the same cells on two sequential sections. Dissemination of mast cell granules was evident in the anaphylactic (A,B) but not in the control animal (C,D). Note the association of bFGF-like immunoreactivity with the extruded mast cell granules (A). (A) Brightfield, (B-D) Nomarski DCI. Bars = 20 μm.

Figure 2

Immunogold particle localization of bFGF in MC of mouse lingual tissue stimulated with anti-IgE antibody for 2 hr in vivo: Characteristic MC membrane folds (F), granules (G), and round nucleus (N). In addition, this MC shows degranulation features such as channel formation (C), swollen granules (L), and granule extrusion (E). Immunogold particle localization for bFGF is present in both the cytoplasmic granules (G) and extruded granules (E). Bar = 1.2 μm. (Inset) bFGF immunogold particle localization (arrowheads) in one of these extruded granules at higher magnification (box). Bar = 290 nm.

After degranulation, mast cells in the anaphylactic tissue exhibited morphological changes characteristic of degranulation, such as degranulation channel formation, swollen granules, shed surface folds, and granule extrusion (Figures 2 and 3A). The extruded granules were membrane-free and exhibited reduced matrix density. Extensive granule extrusion was observed (Figure 3A). Immunogold particle localization of bFGF was present in cytoplasmic secretory granules of the degranulating mast cells. Notably, bFGF immunogold particle localization was also present in the extruded granules (Figures 2 and 3A).

Figure 3

Immunogold particle localization for bFGF in MCs. (A) Mast cell with extensive degranulation in lingual tissue from anti-DNP-sensitized rat challenged with DNP-HAS. This MC exhibits marked shedding of membrane folds (F), channel formation (C), swollen granule (L), normal granules (G), and granule extrusion (E), showing massive degranulation compared to that seen in Figure 2. A mitochondrion (M) with only one gold particle is present. Note the heavy positive labeling in the granules in contrast to minimal background immunogold particle localization in surrounding cytoplasm and extracellular matrix. Bar = 500 nm. (B) Immunogold localization of bFGF in human rheumatoid synovial MC. Strong labeling (gold particles) is present in the cytoplasmic granules (G). Lighter immunogold particle localization is also seen in the nucleus (N), whereas the immunogold particle localization in the cytosol is minimal. Note the characteristic plicated membrane folds (F) and cytoplasmic granules (G). Bar = 250 nm. (C) Immunolabeling of MC with anti-bFGF antibody preabsorbed with recombinant human bFGF. Human rheumatoid synovial mast cell labeled with bFGF exhibited greatly reduced immunogold particle localization in the cytoplasmic granules of the MCs compared with Figure 2. Note the characteristic plicated membrane folds (F) and cytoplasmic granules (G). Bar = 125 nm.

Localization of bFGF in Human Tissues

In all samples examined, immunogold particle localization for bFGF was present in mast cells. Specific bFGF staining in the human tissues at the LM level was reported in detail our previous article (Qu et al. 1995). To determine subcellular localization of bFGF-like molecules, tissue samples were examined by immunoelectron microscopy. Tissue mast cells were identified by their metachromasia and morphological features as described above, and by positive labeling of the cytoplasmic granules by anti-tryptase antibody (data not shown). When examined by EM after immunogold labeling with anti-bFGF antibodies, immunogold particle localization in mast cells was found in all tissue specimens, and all mast cells examined demonstrated immunoreactivity to the anti-bFGF antibody. The localization was predominantly present in the cytoplasmic secretory granules and nucleus (Figure 3B). Immunogold particle localization of bFGF did not appear to assume any particular distribution pattern within the cytoplasmic granules. No apparent immunogold particle localization of bFGF was found in the cytoplasm or cell plasma membrane (Figure 3B).

Controls to Confirm the Specificity of bFGF Staining

To confirm the specificity of bFGF staining present in mast cells, cornu ammon 1 and cornu ammon 2 regions from primate and rat hippocampus known to contain bFGF were used as a positive control tissue. Previous studies by several independent research groups showed that both mRNA for bFGF and immunoreactivity, as well as biological activity, were present in neurons in these regions (Emoto et al. 1989; Eckenstein et al. 1994). Strong immunolabeling by antibFGF antibody was observed in the nucleus of cells in these regions (data not shown), consistent with previous findings reported by others (Emoto et al. 1989; Eckenstein et al. 1994). All anti-bFGF antibodies tested gave similar immunogold particle localization patterns at tissue and subcellular levels in both human, rat, and mouse tissues. Preabsorption of anti-bFGF antibody (Sigma F-3393) with recombinant human bFGF greatly reduced the positive labeling intensity in both cytoplasmic granules and nucleus of MCs (Figure 3C). Omission of primary antibody or substitution of a bFGF-specific antibody with nonimmune IgG or an irrelevant antibody from the same species as the specific antibody completely abolished the immunogold particle localization (data not shown). Specific immunogold particle localization for bFGF was not affected by preabsorption of any protein-containing solution with heparin beads, suggesting that the immunogold particle localization was not due to exogenous bFGF from the solution/reagents used.

Discussion

In this article we present morphological evidence suggesting that mast cell-derived bFGF can be released through degranulation, a classical exocytotic process characteristic of noninjured, properly stimulated mast cells. Because bFGF is often found extracellularly and may be readily available to mast cells, one can argue that mast cells could take up extracellular bFGF and store it in their granules. Two findings appear to refute such an explanation. First, unlike epithelial cells of the tongue and airways that express both bFGF and FGF receptor-1, −2, and −3 (Hughes and Hall 1993), mast cells exhibit no apparent immunostaining for these receptors (our unpublished observation). Second, exogenous bFGF uptake by cells in culture via either FGF receptors or heparin sulfate proteoglycans is directed into multivesicular bodies in endothelial cells (Gleizes et al. 1995), which are morphologically distinct from secretory granules and were not seen in mast cells examined. Although crossreaction of the anti-bFGF antibodies to other members of the FGF family with secretory peptides cannot be completely ruled out, the identical immunogold particle localization patterns given by different anti-bFGF antibodies and the abolishment of the selective localization by preabsorbing the antibody with excess bFGF indicate that this is unlikely. Previous studies by Reed at al. and by us have also shown that both tissue MCs and MC lines contain mRNA for bFGF (Reed et al. 1995; Qu et al., 1995,1998). Finally, because of high bFGF-heparin affinity, exogenous bFGF from BSA and normal sera used for blocking nonspecific binding may bind to mast cell granules and give rise to false but highly specific immunogold particle localization. However, the bFGF level in the blocking solution used was undetectable using a commercial ELISA kit with detection sensitivity at 5 pg/ml (unpublished data). Moreover, the blocking solution was absorbed with heparin acrylamide beads before use. Therefore, possible contamination of tissue sections by exogenous bFGF is unlikely.

Mast cells have been implicated in a wide spectrum of diseases characterized by fibroproliferation and neovascularization (for review see Rother et al. 1990; Galli et al. 1993). Notable examples include pulmonary fibrosis (Schulman 1993), rheumatoid arthritis (Gruber and Kaplan 1993), scleroderma, and GVHD (Claman 1990). Active mast cell degranulation has been noted at least in scleroderma (Claman 1990; Irani et al. 1992). Studies have directly linked mast cell degranulation to fibrotic and angiogenic changes in vivo. For example, Gordon and Galli (1994) have demonstrated that mast cell degranulation elicited by passive cutaneous anaphylaxis resulted in increased collagen gene expression by mouse fibroblasts. Such an increase occurs before apparent influx of neutrophils and other inflammatory cells, indicating that factors derived from the resident cells are responsible. Although the PCA animal model was used to demonstrate the release of bFGF through mast cell degranulation in the present study, extensive studies by Dvorak et al. have shown that a slow degranulation process, piecemeal degranulation, is also present and is clinically relevant to diseases characterized by fibroproliferation and neovascularization (for review see Dvorak 1992).

In addition to cell injury, bFGF release has been described in two other experimental systems in which noninjury release is also contemplated (Mignatti et al. 1991; Sterpetti et al. 1994). Because bFGF is localized in the cytoplasmic secretory granules of mast cells that also contain other secretory proteins such as tryptase, the intracellular trafficking and release mechanism of bFGF in mast cells may differ from those described in two other cell types. Therefore, it appears that different types of cells may use different transport-export systems for bFGF release. Unlike the studies described above, our findings suggest that bFGF is sorted into the classical endoplasmic reticulum/Golgi route. Results from a study by Rubartelli et al. (1990) have indicated that interleukin-1 β, a multifunctional cytokine that also lacks a classical secretory peptide, is also targeted into the classical protein secretory pathway. It is still unknown from the present study how bFGF is targeted into the classical secretory pathway. We believe that there are several plausible mechanisms. The first potentially explanatory finding emerges from a study of another prominent member of the heparin binding growth factor family, acidic FGF, which also lacks a classical secretory sequence. Preliminary results from the study by LaVallee et al. (1996) indicate that intracellular aFGF forms a complex with a protein with high homology to synaptotagmin, a synaptic vesicle protein involved in exocytosis by regulating neurotransmitter uptake, vesicle targeting, and fusion with presynaptic membranes (for review see Elferink and Scheller 1995). Interestingly, the Type II synaptotagmin mRNA was also found in cornu ammon 1-3 areas of the hippocampus of the brain where bFGF is localized (Emoto et al. 1989; Marqueze et al. 1995; Williams et al. 1996). Synaptotagmin is also found in the anterior pituitary of rat, hamster, and guinea pig, a location in which bFGF is present (Redecker et al. 1995). “Secretory lysosomes” may offer a second plausible explanation. Compelling evidence has suggested that cells of the hematopoietic lineage may use special sorting and secretory mechanisms in which the lysosome functions as a regulated secretory granule (for review see Griffiths 1996). In these cells, the secretory lysosomes may transport the proteins that lack a classical secretory sequence as a sorting signal directly into the ER/Golgi route by fusing with different compartments of the pathway. Finally, mast cells are rich in heparin and chondroitin sulfate, which are localized in the secretory granules. It is possible that bFGF employs these proteoglycans as transporting vehicles owing to their high bFGF binding affinities. Although results from the present study cannot provide the answer to how bFGF is sorted into the classical secretory pathway, they strongly suggest that, despite the lack of a classic secretory peptide in its structure, bFGF is localized to the secretory granules in MCs and is released together with the granules via degranulation.

Footnotes

Acknowledgements

Supported by NIH grant EY 10572, by a grant from the Gerlinger Fund, and by an unrestricted grant from Research to Prevent Blindness, New York. JTR is a senior scholar supported by Research to Prevent Blindness, New York. SRP is a recipient of an RPB Special Scholar Award, New York.

We wish to acknowledge the helpful critical review and important insights freely given this manuscript by Dr Ann Dvorak.

References

Bailer

Lazo

Ng-Bautista

Hout

Ness

Hegtvedt

Blakeslee

Stephens

Brierley

Mallery

(1995) Comparison of constitutive cytokine release in high and low histologic grade AIDS-related Kaposi's sarcoma cell strains and in sera from HIV + /KS + and HIV + /KS-patients. J Interferon Cytokine Res 15:473–483.

Baird

(1991) The fibroblast growth factor family. Cancer Cells 3:239–243.

Blam

Mitchell

Tischer

Rubin

Silva

Silver

Fiddes

Abraham

Aaronson

(1988) Addition of growth hormone secretion signal to basic fibroblast growth factor results in cell transformation and secretion of aberrant forms of the protein. Oncogene 3:129–136.

Burgess

Maciag

(1989) The heparin-binding (fibroblast) growth factor family of proteins. Annu Rev Biochem 58:575–606.

Claman

(1990) Mast cells and fibrosis. The relevance to scleroderma. Rheumatic Dis Clin North Am 16:141–151.

Coltrini

Gualandris

Nelli

Parolini

Molinari-Tosatti

Quarto

Ziche

Giavazzi

Presta

(1995) Growth advantage and vascularization induced by basic fibroblast growth factor overexpression in endometrial HEC-1-B cells: an export-dependent mechanism of action. Cancer Res 55:4729–4738.

Dvorak

(1992) Basophils and mast cells: piecemeal degranulation in situ and ex vivo: a possible mechanism for cytokineinduced function in disease. Immunol Ser 57:169–271.

Eckenstein

Kuzis

Nishi

Woodward

Meshul

Sherman

Ciment

(1994) Cellular distribution, subcellular localization and possible functions of basic and acidic fibroblast growth factors. Biochem Pharmacol 47:103–110.

Elferink

Scheller

(1995) Synaptic vesicle proteins and regulated exocytosis. Prog Brain Res 105:79–85.

10.

Emoto

Gonzalez

Walicke

Wada

Simmons

Shimasaki

Baird

(1989) Basic fibroblast growth factor (FGF) in the central nervous system: identification of specific loci of basic FGF expression in the rat brain. Growth Factors 2:21–29.

11.

Florkiewicz

Majack

Buechler

Florkiewicz

(1995) Quantitative export of FGF-2 occurs through an alternative, energy-dependent, non-ER/Golgi pathway. J Cell Physiol 162:388–399.

12.

Galli

Gordon

Wershil

(1993) Mast cell cytokines in allergy and inflammation. Agents Actions Suppl 43:209–220.

13.

Gleizes

Noaillac-Depeyre

Amalric

Gas

(1995) Basic fibroblast growth factor (FGF-2) internalization through the heparan sulfate proteoglycans-mediated pathway: an ultrastructural approach. Eur J Cell Biol 66:47–59.

14.

Goddard

Grossman

Newton

Clark

Bomalaski

(1992) Regulation of synovial cell growth: basic fibroblast growth factor synergizes with interleukin 1 beta stimulating phospholipase A2 enzyme activity, phospholipase A2 activating protein production and release of prostaglandin E2 by rheumatoid arthritis synovial cells in culture. Cytokine 4:377–384.

15.

Gordon

Galli

(1994) Promotion of mouse fibroblast collagen gene expression by mast cells stimulated via the Fc epsilon RI. Role for mast cell-derived transforming growth factor beta and tumor necrosis factor alpha. J Exp Med 180:2027–2037.

16.

Griffiths

(1996) Secretory lysosomes-a special mechanism of regulated secretion in haemopoietic cells. Trends Cell Biol 6:329–332.

17.

Gruber

Kaplan

(1993) Mast cells and rheumatic diseases. In McCarty

, ed. Arthritis and Allied Conditions. Philadelphia, Lea & Febiger, 417–436.

18.

Hughes

Hall

(1993) Immunolocalization of fibroblast growth factor receptor 1 and its ligands in human tissues. Lab Invest 69:173–182.

19.

Irani

Gruber

Kaufman

Kahaleh

Schwartz

(1992) Mast cell changes in scleroderma. Presence of MCT cells in the skin and evidence of mast cell activation. Arthritis Rheum 35:933–939.

20.

Kandel

Bossy-Wetzel

Radvanyi

Klagsbrun

Folkman

Hanahan

(1991) Neovascularization is associated with a switch to the export of bFGF in the multistep development of fibrosarcoma. Cell 66:1095–1104.

21.

LaVallee

Tarantini

Gamble

Burgess

Maciag

(1996) Synaptotagmin-1 and FGF-1 homodimer are present as a high molecular weight complex in neural tissue. FASEB J 10: A1377.

22.

Folkerth

Watanabe

Rupnick

Barnes

Scott

Black

Sallan

Folkman

(1994) Microvessel count and cerebrospinal fluid basic fibroblast growth factor in children with brain tumors. Lancet 344:82–86.

23.

Lindner

Reidy

(1991) Proliferation of smooth muscle cells after vascular injury is inhibited by an antibody against basic fibroblast growth factor. Proc Natl Acad Sci USA 88:3739–3743.

24.

Marqueze

Boudier

Mizuta

Inagaki

Seino

Seagar

(1995) Cellular localization of synaptotagmin I, II, and III mRNAs in the central nervous system and pituitary and adrenal glands of the rat. J Neurosci 15:4906–4917.

25.

Melnyk

Shipley

Sternfeld

Sherman

Rosenbaum

(1990) Synoviocytes synthesize, bind, and respond to basic fibroblast growth factor. Arthritis Rheum 33:493–500.

26.

Mignatti

Morimoto

Rifkin

(1991) Basic fibroblast growth factor released by single, isolated cells stimulates their migration in an autocrine manner. Proc Natl Acad Sci USA 88:11007–11011.

27.

Montesano

Vassalli

Baird

Guillemin

Orci

(1986) Basic fibroblast growth factor induces angiogenesis in vitro. Proc Natl Acad Sci USA 83:7297–7301.

28.

Huang

Ahmadi

Stenberg

Liebler

A-C

Planck

Rosenbaum

(1998) Expression of basic fibroblast growth factor (bFGF) by murine mast cells-regulation by transforming growth factor beta, tumor necrosis factor alpha and stem cell factor. Int Arch Allergy Immunol 115:47–54.

29.

Liebler

Powers

Galey

Ahmadi

Huang

Ansel

Butterfield

Planck

Rosenbaum

(1995) Mast cells are a major source of basic fibroblast growth factor in chronic inflammation and cutaneous hemangioma. Am J Pathol 147:564–573.

30.

Redecker

Cetin

Grube

(1995) Differential distribution of synaptotagmin I and rab3 in the anterior pituitary of four mammalian species. Neuroendocrinology 62:101–110.

31.

Reed

Albino

McNutt

(1995) Human cutaneous mast cells express basic fibroblast growth factor. Lab Invest 72:215–222.

32.

Rogelj

Weinberg

Fanning

Klagsbrun

(1988) Basic fibroblast growth factor fused to a signal peptide transforms cells. Nature 331:173–175.

33.

Rogelj

Weinberg

Fanning

Klagsbrun

(1989) Characterization of tumors produced by signal peptide-basic fibroblast growth factor-transformed cells. J Cell Biochem 39:13–23.

34.

Rother

Nowak

Kerdel

(1990) The mast cell in health and disease. J Am Acad Dermatol 23:615–624.

35.

Rubartelli

Cozzolino

Talio

Sitia

(1990) A novel secretory pathway for interleukin-1 beta, a protein lacking a signal sequence. EMBO J 9:1503–1510.

36.

Saksela

Moscatelli

Sommer

Rifkin

(1988) Endothelial cell-derived heparan sulfate binds basic fibroblast growth factor and protects it from proteolytic degradation. J Cell Biol 107:743–751.

37.

Samaniego

Markham

Gallo

Ensoli

(1995) Inflammatory cytokines induce AIDS-Kaposi's sarcoma-derived spindle cells to produce and release basic fibroblast growth factor and enhance Kaposi's sarcoma-like lesion formation in nude mice. J Immunol 154:3582–3592.

38.

Schulman

(1993) The role of mast cells in inflammatory responses in the lung. Crit Rev Immunol 131:35–70.

39.

Sterpetti

Cucina

Fragale

Lepidi

Cavallaro

Santoro

DAL

(1994) Shear stress influences the release of platelet derived growth factor and basic fibroblast growth factor by arterial smooth muscle cells. Eur J Vasc Surg 8:138–142.

40.

Thompson

Slack

(1992) Over-expression of fibroblast growth factors in Xenopus embryos. Mechan Dev 38:175–182.

41.

Wershil

Wang

Gordon

Galli

(1991) Recruitment of neutrophils during IgE-dependent cutaneous late phase reactions in the mouse is mast cell-dependent. Partial inhibition of the reaction with antiserum against tumor necrosis factor-alpha. J Clin Invest 87:446–453.

42.

Williams

Meshul

Cherry

Tiffany

Eckenstein

Woodward

(1996) Characterization and distribution of basic fibroblast growth factor-containing cells in the rat hippocampus. J Comp Neurol 370:147–158.