Localization of Basic Fibroblast Growth Factor in ... - Wiley Online Library

2 downloads 0 Views 1MB Size Report
Abstract: We have investigated basic fibroblast growth factor (FGF-2) localization in and release from isolated bovine adrenal chromaffin cells. In contrast to ...
Journal of Neurochemistry

Raven Press, Ltd ., New York C 1995 International Society for Neurochemistry

Localization of Basic Fibroblast Growth Factor in Bovine Adrenal Chromaffin Cells Sophie C. Bieger, *A. W. Henkel, and K. Unsicker Department of Anatomy and Cell Biology, University of Heidelberg, Heidelberg, Germany ; and *Department of Physiology, University of Colorado Health Sciences Center, Denver, Colorado, U.S.A .

also be partly artefactual . It has been shown that immunolocalization of FGF-2 is strongly dependent on antibodies and tissue fixation techniques used by each investigator (Hanneken and Baird, 1992) ; thus, an unambiguous localization of this factor is extremely difficult in situ . Possibly the biochemical characteristics of the FGF-2 protein itself (e .g ., basic isoelectric point, affinity for polyanionic molecules such as heparin) contribute to artefactual localizations . Not only is there discordance in immunohistochemical data, but there are conceptual disagreements as well . It is difficult to explain how FGF-2, which lacks a signal peptide (the prerequisite for Golgi-mediated secretion of proteins), could be released from the cells that synthesize it . If FGF-2 is indeed stored in the extracellular matrix and in basement membranes (Vlodavsky et al ., 1987 ; Flaumenhaft et al ., 1989 ; Gonzalez et al ., 1990), then some mechanism by which this protein is liberated from the cell must be postulated . To date, only one study (Westermann et al ., 1990) has indicated the possible "regulated," i .e ., Golgi vesicle-mediated, release of FGF-2 from cells . Isolated bovine adrenal medullary cells, viewed under immunoelectron microscopy with anti-FGF-2 antibodies, showed a significant accumulation of gold-labeled antibody over chromaffin granules, the catecholaminestoring secretory organelles . Purification of these granules by centrifugation through a sucrose cushion was reported to bring about an enrichment of FGF-2 activity in this cellular compartment . Because these data strongly suggested that bovine adrenal chromaffin cells are capable of storing FGF-2 in secretory granules and thus releasing the factor by exocytosis, we decided to validate this hypothesis using other investigative techniques . However, these

Abstract : We have investigated basic fibroblast growth factor (FGF-2) localization in and release from isolated bovine adrenal chromaffin cells . In contrast to previous reports, we found no evidence of fibroblast growth factor (FGF) storage in catecholamine-containing chromaffin granules . Subcellular fractionation studies did not show enrichment of FGF-2 immunoreactivity in granules, and cholinergic stimulation failed to release FGF-2 into the medium . Our results suggest that adrenal chromaffin cells resemble other FGF-2-synthesizing cell types with respect to FGF storage and secretion . Key Words : Bovine adrenal chromaffin cell-Basic fibroblast growth factor . J . Neurochem. 64, 1521-1527 (1995) .

Although basic fibroblast growth factor (FGF-2 ) was one of the first heparin-binding growth factors to be described, it remains one of the more enigmatic members of the fibroblast growth factor (FGF) family, which now encompasses at least nine different proteins (for reviews, see Baird and B6hlen, 1990 ; Coulier et al ., 1993) . Perhaps the most controversial issue concerns the localization of FGF-2 at both cell and tissue levels . In the central nervous system, for example, different investigators have described a primarily neuronal (Pettmann et al ., 1986) or primarily nonneuronal (Woodward et al ., 1992) distribution of FGF-2 immunoreactivity . In the developing rat adrenal gland, Gonzalez et al . (1990) have localized FGF-2 in the zona reticularis and zona fasciculata of the adrenal cortex, but not in the adrenal medulla, whereas Grothe and Unsicker (1990) found FGF-2 mainly in the zona glomerulosa of the adrenal cortex as well as in the adrenal medulla. At the cellular level, FGF-2 has been alternately described as predominantly intracellular (Whittemore et al., 1991) ,predominantly extracellular (Healy and Herman, 1992), and ubiquitous (Brigstock and Klagsbrun,1991) . The discrepant results may arise from the various animal species, developmental stages, or cell types investigated in each study, i .e ., FGF-2 synthesis and deposition may actually be regulated in different fashions for different cells . However, the discrepancies may

Received June 17, 1994 ; revised manuscript received September 5, 1994; accepted September 22, 1994 . Address correspondence and reprint requests to Dr. K . Unsicker at Department of Anatomy and Cell Biology, University of Heidelberg, Im Neuenheimer Feld 307, 69120 Heidelberg, Germany . Abbreviations used: BSA, bovine serum albumin ; FGF, fibroblast growth factor ; FGF-2, basic fibroblast growth factor ; PBS, phosphate-buffered saline ; TBS, Tris-buffered saline . 1521

S. C. BIEGER ET AL.

1522

stimulation experiments as well as subcellular fractionation studies using isolated bovine chromaffin cells failed to support the hypothesis that FGF-2 is colocalized with catecholamines and is released by the "regulated" pathway of secretion. MATERIALS AND METHODS Cell culture Chromaffin cells were isolated from bovine adrenal medullae by collagenase digestion and Percoll gradient centrifugation essentially as described by Unsicker et al . (1980) and were seeded into plastic culture flasks (Nunc) at densities of 1 .5-2 X 10 5 cells/CMZ . The cultures routinely had a purity of at least 90-95% . Culture medium consisted of Dulbecco's modified Eagle's medium with NI supplements (Bottenstein et al ., 1980) and 100 Ng/ml bovine serum albumin (BSA) . For stimulation experiments, the culture medium was replaced after 18-30 h with prewarmed stimulation buffer (Hanks' balanced salt solution containing 3 mM CaCl z ) . After a 30-min acclimation period at 37°C, the cells were given fresh stimulation buffer containing nicotine (50 IM) or carbachol (100 pM) or no additives and incubated for an additional 20 min at 37°C . The stimulation buffer was then collected, an aliquot was removed for determination of catecholamines by HPLC (as described by Müller and Unsicker, 1981), and the remainder was prepared for protein analysis (i .e ., precipitation of proteins with trichloroacetic acid and solubilization in sample buffer for electrophoresis) . Immunocytochemistry Chromaffin cells were seeded onto polyornithine-coated coverslips . After at least 18 h of culture, cells were fixed with 4% paraformaldehyde (in some cases containing 0.5% Triton X-100), blocked with 1 % BSA in phosphate-buffered saline (PBS), incubated 4-10 h with primary antibody (diluted 1 :50- 1 :100), and then incubated for 1 h with fluorescein-isothiocyanate-conjugated secondary antibody . Controls included omission of primary antibody or preabsorption of the primary antibody with excess antigen. The polyclonal FGF-2 antibody was prepared from an antiserum directed against the 24 N-terminal amino acids of FGF-2 (generously supplied by P. Bohlen) . A crude immunoglobulin fraction was obtained from this antiserum by ammonium sulfate precipitation (Harlow and Lane, 1988) . Alternatively, commercial monoclonal anti-FGF-2 antibody (UBI ; Frankfurt, Germany) was used . The anti-chromogranin A antibody (rabbit polyclonal) was purchased from DAKO (Hamburg, Germany) and the anti-dopamine ß-hydroxylase was a gift from R. Fischer-Colbrie (Innsbruck, Austria) . Northern blot RNA was isolated from purified chromaffin cells according to the method of Chomczynski and Sacchi (1987) . RNA samples (25 leg) were fractionated on a 1% agarose, 2.2 M formaldehyde gel, transferred to nylon membranes (Hybond-N; Amersham), and hybridized with "P-labeled FGF-2 probe using the protocol of Church and Gilbert (1984) . We used a cDNA probe (pRObFGF-477) containing the complete rat FGF-2 cDNA (Shimasaki et al ., 1988), which was kindly provided by A. Baird (La Jolla, CA, U.S .A .) . Exposures of at least 3 days were routinely required to obtain detectable signals. To confirm that equivalent amounts of RNA had been loaded, the membrane was .1.

Neurochem ., Vol . 64, No. 4, 1995

stained with 0.02% methylene blue in 0.3 M sodium acetate, pH 5.5, after hybridization and stripping. Western blot Suspensions of freshly isolated chromaffin cells were counted in a hemocytometer prior to lysis in electrophoresis sample buffer. Samples of bovine adrenal medulla were homogenized in electrophoresis sample buffer and protein concentrations were determined by the method of Sportsman and Elder (1984) . Cell and tissue samples were then separated by electrophoresis and transferred to nitrocellulose (Hybond-ECL; Amersham) . The nitrocellulose membranes were blocked with 3% skim milk powder/0 .1 % BSA in Trisbuffered saline (TBS) buffer (150 in M NaCI, 10 mM Tris, pH 7.3), incubated overnight with primary antibody (diluted 1 :500-1 :1,000 in TBS/I% BSA) followed by peroxidaseconjugated secondary antibody, and finally developed using the Amersham enhanced chemiluminescence (ECL) detection system . For reprobing, the membranes were stripped by incubation in 2% sodium dodecyl sulfate, 100 mM 2mercaptoethanol, and 62 .5 mM Tris, pH 6 .7, for 30 min at 37°C and washed in TBS (3 X 15 min) prior to blocking and incubation with antibodies as described above. Subcellular fractionation After 18 h of culture, chromaffin cells were washed twice with ice-cold PBS and once with cold homogenization buffer (10 mM HEPES, pH 7.4, 1 mM EDTA, 250 mM sucrose) prior to being disrupted by means of a ball bearing homogenizer (described by Balch and Rothman, 1985) . A postnuclear supernatant was generated by centrifuging the homogenate at 1,000 g (10 min, 4°C) ; this supernatant was layered on top of a 0.4-2 .4 M continuous sucrose gradient (in 10 mM HEPES, pH 7.4, and 1 mM EDTA) . After centrifugation at 100,000g (4°C, 6 h), 13 1-ml fractions were collected from the gradient, precipitated with trichloroacetic acid, solubilized in electrophoresis sample buffer, and analyzed by western blot, as above. Immunoelectron microscopy Bovine chromaffin cells cultivated on polyornithinecoated glass coverslips for 24-36 h were fixed with 2% paraformaldehyde, 0.5% glutaraldehyde, 0.05% saponin in PBS for 45 min at 4°C, washed with PBS, and blocked with 0.1% NaBH 4 (10 min) followed by 5% BSA, l Ic normal goat serum, and 0.1% gelatin in PBS (15 min) . After being washed again with PBS, the coverslips were incubated sequentially with anti-FGF-2 antibody (UBI ; dilution 1 :1,000 ; overnight at 4°C) and with ultrasmall gold-conjugated secondary antibody (Bio-Trend ; dilution 1 :50; 3 h at 4°C) . The coverslips were then postfixed with 2% glutaraldehyde in PBS, silver enhanced according to Danscher (1981), postfixed with Os04 , and block stained with uranyl acetate prior to dehydration and embedding in Epon . Ultrathin sections were cut using an LKB ultrotome III for electron microscope examination . RESULTS Expression of FGF-2 in chromaffin cells RNA isolated from chromaffin cells after various times in culture and hybridized to radiolabeled FGF2 probe yielded characteristic bands of 7 and 3 .7 kb

CHROMAFFIN CELL FIBROBLAST GROWTH FACTOR

1523

FIG. 3. Analysis of FGF-2 immunoreactivity in chromaffin culture supernatants after stimulation with 100 pM carbachol (Garb), 50 N.M nicotine (Nic), or no secretagogue (Ctrl) . After a 20-min stimulation period, the medium was precipitated with trichloroacetic acid and analyzed by western blot as described .

FIG. 1 . A: Northern blot analysis of FGF-2 in chromaffin cells. Samples of total RNA (25 N,g) isolated from chromaffin cells after the culture times indicated were probed with radiolabeled FGF2 cDNA . B: After hybridization, the membrane was stained with methylene blue to confirm that equal quantities of RNA had been loaded .

(Fig . 1 ), indicating that chromaffin cells contain FGF2 mRNA. The weak signal suggests that FGF message is present at very low levels in freshly isolated chromaffin cells . Because mRNA levels declined during a 7-day culture period, all further experiments were conducted with cells remaining no more than 1-2 days in culture . To investigate the expression of FGF-2 protein, samples of whole adrenal medulla or freshly isolated chromaffin cells were lysed directly in electrophoresis sample buffer and analyzed by western blot. The antiFGF-2-antibody yielded immunoreactive bands corresponding to molecular masses of - 18, 21, and 23 kDa (Fig. 2) . Immunoreactivity could be detected in as few as 4 X 10 4 chromaffin cells . By comparison with the staining intensity of purified (recombinant) FGF-2, it could be estimated that 17 leg of adrenal medulla protein or -v5 X 10 4 chromaffin cells contain approximately 100-200 pg of FGF. The limit of detection in the western blot was routinely about 20-40 pg of FGF-2 . Effect of secretagogues on release of FGF-2 from chromaffin cells When chromaffin cells were stimulated for 20 min with either nicotine or the cholinergic agonist carba-

FIG . 2. Western blot analysis of FGF-2 in adrenal medulla and in isolated chromaffin cells . The indicated amounts of protein or of FGF-2 standard were separated by sodium dodecyl sulfatepolyacrylamide gel electrophoresis, blotted onto nitrocellulose, and probed with anti-FGF-2 antibody as described .

chol to induce exocytosis of catecholamines, no FGF2 immunoreactivity could be recovered in the supernatants . As shown in Fig . 2, the western blot was capable of detecting 18-, 21-, and 23-kDa FGF-2 contained in fewer than 10 5 cells ; however, the combined supernatants from 9 X 10' cells yielded no detectable FGF-2 immunoreactivity (Fig . 3) . Similarly, stimulation with 12-O-tetradecanoylphorbol 13-acetate (250 nM) or KCl (50 mM) failed to induce detectable release of FGF-2 by chromaffin cells . The possibility that FGF-2 is secreted upon stimulation but remains associated with chromaffin cells after release was investigated by immunocytochemistry . Chromaffin cells, when fixed with 4% paraformaldehyde (in the presence or the absence of 0.5% Triton X-100) and stained with anti-FGF-2 antibody followed by fluorescenn isothiocyanate-labeled secondary antibody, showed a diffuse and relatively weak fluorescence (Fig . 4C) . Nuclei were also frequently stained . Preabsorption of the anti-FGF-2 antibody with I leglml recombinant FGF-2 abolished the fluorescence almost completely, and in the absence of primary antibody no fluorescent signal was obtained . Pretreatment of cells with carbachol for 20 min prior to fixation and processing for immunocytochemistry did not significantly enhance the immunofluorescence (Fig . 4D) . Similar results were obtained with commercial anti-FGF-2 antibody (UBI) . An antibody directed against dopamine ß-hydroxylase, on the other hand, yielded an intense and characteristic "granular" signal, with no significant nuclear fluorescence (Fig. 4A) . The signal was dramatically increased in carbachol-stimulated cells (Fig . 4B) . Localization of FGF-2 immunoreactivity by subcellular fractionation of chromaffin cells Results of the subcellular fractionation experiments are shown in Fig . 5 . As expected, chromogranin A, a marker for chromaffin granules, was enriched in dense fractions of the gradient (1 .7-1 .9 M sucrose ; fractions 7-10), demonstrating that granules had largely remained intact during the homogenization procedure . The same blots, when stripped of antibody and reprobed with anti-FGF-2 antibody, did not show enrichment of FGF immunoreactivity in fractions 7-10, conJ. Neuiochem. . Vol. 64 . No. 4, 1995

152 4

S. C. BIECER ET AL.

FIG . 4 . Dopamine-0-hydroxylase immunofluorescence in unstimulated chromaffin cells (A) and in cells stimulated with 100 pM carbachol (B) . FGF-2 immunofluorescence in unstimulated chromaffin cells (C) and in carbachol-stimulated cells (D) . Photographic exposure times were 4 s for A and B, and 60 s for C and D .

tradicting the hypothesis of colocalization with chromogranin A . Instead, FGF-2 was found primarily in lighter fractions of the gradient, with 18-kDa FGF distributed throughout fractions 1-8 and higher molecular mass (21 and 23 kDa) forms of FGF, peaking in fractions 4-5 . Significant FGF immunoreactivity (18, 21, and 23 kDa) was also detected in the 1,000-g pellet, which was also strongly positive for anti-histone antibody (not shown) . This suggests that FGF-2 may also be associated with nuclei . Localization of FGF-2 by immunoelectron microscopy Electron microscopic inspection of chromaffin cells treated with anti-FGF-2 antibody and gold-conjugated

FIG . 5. Distribution of FGF-2 in bovine chromaffin cells . A postnuclear supernatant was prepared and separated by centrifugation (100,000 g for 6 h) on a continuous 0 .4-2 .4 M sucrose gradient . One-milliliter fractions were then collected and analyzed by western blot using, sequentially, anti-FGF-2 and anti-chromogranin A antibody .

.1. Neuro, hem., Vol. 64, No. 4, 1995

secondary antibody revealed significant accumulation of gold label over nuclei and in extracellular spaces at cell-to-cell contacts (Fig . 6) . No gold label was found to be associated with electron-dense catecholaminestoring granules . However, gold grains were occasionally found in a class of large membrane-bound organelles of lysosomal appearance (having intermediate electron density and often containing membranous inclusions) . DISCUSSION The series of experiments described here cast doubt on the previous finding (Westermann et al ., 1990) that

CHROMAFFIN CELL FIBROBLAST GROWTH FACTOR

1525

FIG. 6. Immuno-gold labeling of chromaffin cells using commercial anti-FGF-2 antibody . Cells were immunolabeled prior to embedding and processing for electron microscopy as described in Materials and Methods. A: Overview of gold particle localization in cells. Note labeling of nucleus (n), of membrane (arrowhead), and of vesicular compartment (arrows) (x25,300) . B: Higher magnification of goldlabeled vesicles (x31,500) .

FGF-2 is localized in the secretory granules of chromaffin cells . Using several techniques different from those used by Westermann et al ., we could confirm the occurrence of FGF-2 immunoreactivity in chromaffin cells, but we could not find evidence that this immunoreactivity is concentrated in chromaffin granules . The conflicting data from the two studies are somewhat difficult to reconcile; however, given the sensitivity of FGF immunolocalization to tissue fixatives and to antibody type, the variations in immunocytoehemical protocols and the different antibodies used (e.g ., monoclonal versus polyclonal) in each study may be significant . Another methodological difference is in the isolation of chromaffin granules . In the "classical" method employed by Westermann et al . (1990) [based on the method of Winkler and Smith (1975)], whole adrenal medullae are homogenized in a blender, a crude vesicle pellet is obtained by centrifugation, and granules are purified by centrifugation of the vesicle preparation through a 1 .7 M sucrose cushion . We found that it is preferable to disrupt chromaffin cells by means of a ball bearing homogenizer, as Potter-Elvehjem or UltraTurrax devices caused significant rupture of granules (hence potential relocalization of FGF-2) . Separation of the postnuclear supernatant on a sucrose gradient

extending from 0.4 to 2.4 M (rather than on a sucrose cushion) enabled us to cleanly resolve granules from other organelles (endoplasmic reticulum, mitochondria, residual nuclei) and thus to identify more clearly the cellular compartments containing FGF-2 immunoreacti vi ty. Our results from the subcellular fractionation indicate that FGF is localized in several compartments . The amino terminally extended forms of FGF-2 (21 and 23 kDa) seem to have a more restricted distribution than the 18-kDa protein, supporting the possibility that the different forms may have different intracellular functions (Brigstock and Klagsbrun, 1991) . Both 21and 23-kDa FGF-2 are found only in fractions 3-5 (-0.7- 1 .1 M sucrose) and in the 1,000 g pellet, which contains nuclei as well as unbroken cells . The 18-kDa FGF-2 has a broader intracellular distribution : some of the protein is apparently cytosolic, as immunoreactivity was found in the top fraction of the sucrose gradient . A major portion of the total 18-kDa FGF immunoreactivity is found throughout the upper region of the gradient (fractions 2-8 ; -0.4-1 .3 M sucrose), where small vesicles, Golgi apparatus, and endoplasmic reticulum are also located . Finally, a significant portion is found in the nuclear pellet . Despite the slight overlap between the 18-kDa FGF and the granule J. Ncurochern ., Vol. 04, No . 4, 1995

152 6

S. C. BIEGER ET AL .

marker chromogranin A, which is enriched in fractions 7-10, the distribution patterns of the two proteins are clearly distinct from each other . Thus, the subcellular fractionation experiments indicate that FGF-2 is unlikely to be costored with catecholamines. Although our results contradict the hypothesis of FGF-2 localization in chromaffin granules, they cannot precisely identify the cellular compartment(s) in which FGF-2 is found . Consistent with several other reports (Brigstock and Klagsbrun, 1991 ; Florkiewicz et al., 1991 ; Woodward et al ., 1992), we have found FGF-2, particularly the higher molecular mass forms (21 and 23 kDa), in the nucleus, suggesting that this growth factor has autocrine functions in chromaffin cells . The identity of the "light vesicles" containing FGF immunoreactivity remains uncertain . Endoplasmic reticulum (density of 0.5- 1 .6 M in sucrose) and Golgi membranes (density of 0.8-1 .2 M in sucrose) would be expected to lie in the same region of the gradient in which FGF-2 is found, but it is unlikely that FGF-2 (lacking a signal peptide) would be located in these cellular compartments . Nevertheless, it is interesting that the localization of 21 and 23 kDa FGF2 (fractions 3-5) overlaps with that of the 6-amyloid precursor protein (fractions 4-5 ; not shown), which is secreted constitutively by chromaffin cells (Bieger et al., 1993) . Plasma membrane sheets and membrane vesicles (density of 0.8-1 .4 M in sucrose) would also be expected to lie in the same region of the gradient, and we suggest that it is these membranes, or possibly heparin sulfate proteoglycans associated with these membranes, to which FGF-2 is bound. Although we cannot exclude the possibility that such an association of FGF with membranes could take place "artificially" during the homogenization procedure, we can exclude a specific association of FGF-2 with the catecholamine-storing organelles of chromaffin cells . Immunoelectron microscopy observations (obtained using a commercial monoclonal antibody) support the biochemical data. Electron-dense chromaffin granules are unlabeled by anti-FGF-2 antibody ; only nuclei and plasma membranes (on the extracellular side) show significant accumulation of gold label. The nature of the large, relatively electron-transparent organelles which also show gold labeling is unclear . If these vesicles correspond to the "light vesicles" found to contain FGF immunoreactivity in the subcellular fractionation, their buoyant density makes them unlikely to be lysosomes, but endosomes are a possibility . Further studies should clarify whether chromaffin cells internalize and transport FGF by endosomes or other vesicles . The stimulation experiments clearly exclude cosecretion of FGF-2 with catecholamines as a major mechanism of FGF release . Less than 20 pg of FGF2 is released from 9 X 10' chromaffin cells upon stimulation with secretagogue, i .e., fewer than eight molecules per cell are released into the medium by exocytosis. In light of the strong FGF-2 signal obtained J. Neurohem . . Vol. 64, No . 4, 1995

in the western blot with 10' cells (corresponding to roughly 200 pg of FGF), this would imply that less than 0.01% of' the theoretical FGF-2 content is released . Thus, the main site of FGF-2 storage is unlikely to be the chromaffin granule. Further evidence for a "nongranular" storage site for FGF-2 is the following : (a) Other cell types such as bovine adrenocortical cells and bovine aortic endothelial cells, both of which have been shown to synthesize FGF-2 (Schweigerer et al ., 1987; Vlodavsky et al ., 1987), do not contain granules analogous to the chromaffin catecholamine storage organelles . The constitutive, rather than the regulated, route of secretion thus appears to be the "default" pathway for FGF-2 . We have subjected cells isolated from the bovine adrenal cortex to the same subcellular fractionation procedure used for chromaffin cells and found a similar distribution of 18-kDa FGF2 immunoreactivity in the sucrose density gradient (data not shown) . (b) During the collagenase digestion of bovine adrenal medullae, large quantities of FGF-2 are released into the digestion supernatant. We have estimated that 90% of the 18-kDa FGF-2 immunoreactivity found in bovine adrenal medullae is recovered in the noncellular fraction (on an FGF per gram of protein basis) . In part, the FGF-2 found in the digestion supernatant could be released from ruptured chromaffin cells, but significant quantities may also be stored in the extracellular matrix of chromaffin cells and liberated from the cells by enzymatic (collagenase) treatment . This hypothesis agrees with results obtained by several investigators using vascular endothelial cells, who have shown conclusively that FGF-2 is colocalized with heparan sulfates in basement membranes and in the extracellular matrix synthesized by these cells (Baird and Ling, 1987; Flaumenhaft et al ., 1989; Vlodavsky et al ., 1991a,b ; Healy and Herman, 1992) . In summary, the present findings bring FGF-2 synthesis and "release" by bovine adrenal chromaffin cells in line with data obtained using other cell types such as bovine adrenocortical cells and bovine aortic endothelial cells, by suggesting that FGF-2 is not released by regulated secretion, but is sequestered extracellularly by some unknown mechanism . Whether cosecretion of FGF-2 with heparan sulfate proteoglycans (Baird and Ling, 1987 ; Gordon et al., 1989), externalization of FGF-2 via some membrane transporter (Muesch et al., 1990; Healy and Herman, 1992), release by cell lysis (Clarke et al., 1993), or perhaps some novel secretory path is involved must be determined by a more detailed molecular analysis. REFERENCES Baird A . and Bohlen P . ( 1990) Fibroblast growth factors . Handbook Exp. Pharmacol. 95, 369-418 . Baird A . and Ling N . (1987) Fibroblast growth factors are present in the extracellular matrix produced by endothelial cells in vitro : implications for a role of heparinase-like enzymes in the neovas-

CHROMAFFIN CELL FIBROBLAST GROWTH FACTOR cular response . Biochem. Biophys. Res. Commun. 142, 428435 . Balch W . E. and Rothman J . (1985) Characterization of protein transport between successive compartments of the Golgi apparatus : asymmetric properties of donor and acceptor activities in a cell-free system . Arch . Biochem . Biophys. 240, 413-425 . Bieger S ., Klafki H .-W ., and Unsicker K . (1993) Synthesis and release of the ß-amyloid precursor protein by bovine chromaffin cells . Neurosci . Lett. 162, 173-175 . Bottenstein J . E ., Skaper S . D ., Varon S . S ., and Sato G. ( 1980) Selective survival of neurons from chick embryo sensory ganglionic dissociates utilizing serum-free supplemented medium . Exp . Cell Res . 125, 183-190 . Brigstock D . R . and Klagsbrun M . (1991) Subcellular distribution of multiple molecular weight forms of bFGF in hepatoma cells . Ann. NY Acad. Sci . 638, 427-429 . Chomczynski P . and Sacchi N . (1987) Single-step method of RNA extraction by acid guanidinium thiocyanate-phenol-chloroform extraction . Anal. Biochem. 162, 156-159 . Church G . M . and Gilbert W . ( 1984) Genomic sequencing . Proc. Nall. Acad. Sci . USA 81, 1991-1995 . Clarke M . S . F., Khakee R ., and McNeill P. L. (1993) Loss of cytoplasmic basic fibroblast growth factor from physiologically wounded myofibers of normal and dystrophic muscle . J. Cell Sci . 106, 121-133 . Coulier F ., de Lapeyrière O ., and Birnbaum D . ( 1993) Complexit6 de la famille des facteurs de croissance FGF: la preuve par 9 . Medecine et Sciences 9, 1113-1115 . Danscher G . (1981) Localization of gold in biological tissue: a photochemical method for light and electron microscopy . Hi .stochemistry 71, 81-88 . Flaumenhaft R ., Moscatelli D ., Saksela O ., and Rifkin D . B . (1989) Role of extracellular matrix in the action of basic fibroblast growth factor : matrix as a source of growth factor for long term stimulation of plasminogen activator production and DNA synthesis . J. Cell. Physiol. 140, 75-81 . Florkiewicz R . Z ., Shibata F ., Barankiewicz T ., Baird A ., Gonzalez A . M ., Florkiewicz E ., and Shah N. (1991 ) Basic fibroblast growth factor gene expression . Ann. NY Actin Sci. 638, 109126 . Gonzalez A .-M ., Buscaglia M ., Ong M ., and Baird A . (1990) Distribution of basic fibroblast growth factor in the 18-day rat fetus : localization in the basement membranes of diverse tissues . J . Cell Biol. 110, 753-765 . Gordon P . B ., Choi H . U ., Conn G ., Ahmed A ., Ehrman B ., Rosenberg L ., and Hatcher V . B . (1989) Extracellular matrix heparin sulfate proteoglycans modulate the mitogenic capacity of acidic fibroblast growth factor . J. Cell . Physiol. 140, 584-592 . Grothe C . and Unsicker K . (1990) Immunocytochemical localization of basic fibroblast growth factor in the developing rat adrenal gland . Histochemistry 94, 141-147 . Hanneken A . and Baird A . ( 1992) Immunolocalization of basic fibroblast growth factor : dependence on antibody type and tissue fixation . Exp. Eye Res. 54, I011-1014. Harlow E . and Lane D . (1988) Antibodies: A Laboratory Manual . Cold Spring Harbor Laboratory, New York.

1-527

Healy A . M . and Herman I. M . ( 1992) Density-dependent accumulation of basic fibroblast growth factor in the subendothelial matrix . Ear. J . Cell Biol . 59, 56-67 . Muesch A., Hartmann E ., Rohde K ., Rubartelli A ., Sitia R ., and Rapoport T . A . (1990) A novel secretory pathway for proteins? Trends Biochem . Sci. 15, 86-88 . Midler T . H . and Unsicker K . (1981 ) High-performance liquid chromatography with electrochemical detection as a highly efficient tool for studying catecholaminergic systems . 1 . Quantification of noradrenaline, adrenaline and dopamine in cultured adrenal medullary cells . J. Neurosci . Methods 4, 36-52 . Pettmann B ., Labourdette G ., Weibel M ., and Sensenbrenner M . (1986) The brain fibroblast growth factor (FGF) is localized in neurons . Neurosci. Lett. 68, 175-180 . Schweigerer L., Neufeld G ., Friedman J ., Abraham J . A ., Fiddes J . C ., and Gospodarowicz D . (1987) Basic fibroblast growth factor : production and growth stimulation in cultured adrenal cortex cells . Endocrinology 120, 796-800. Shimasaki S ., Emoto N ., Koba A ., Mercado M ., Shibata F ., Cooksey K ., Baird A ., and Ling N . (1988) Complementary cDNA cloning and sequencing of rat ovarian basic fibroblast growth factor and tissue distribution study of its mRNA . Biochem . Biophys. Res . Commun. 157, 256-263 . Sportsman J . R . and Elder J. H . ( 1984) A microanalytical protein assay using laser densitometry . Anal. Biochem. 139, 298-300. Unsicker K ., Griesser G . H ., Lindmar R ., L6ffelholz U ., and Wolf U . (1980) Establishment, characterization and fibre outgrowth of isolated bovine adrenal medullary cells in long-term cultures . Neuroscience 5, 1445-1460 . Vlodavsky I ., Folkman J ., Sullivan R ., Fridman R ., Ishai-Michaeli R ., Sasse J ., and Klagsbrun M . (1987) Endothelial cell-derived basic fibroblast growth factor : synthesis and deposition into subendothelial extracellular matrix . Proc. Natl . Acad. Sci. USA 84, 2292-2296 . Vlodavsky I ., Bashkin P ., Ishai-Michaeli R ., Chajek-Shaul T ., BarShavit R ., Haimovitz-Friedman A ., Klagsbrun M ., and Fuks Z . (199 la) Sequestration and release of basic fibroblast growth factor. Ann. NY Acad. Sci. 638, 207-220 . Vlodavsky i., Bar-Shavit R ., Ishai-Michaeli R., Bashkin P ., and Fuks Z . (1991 b) Extracellular sequestration and release of fibroblast growth factor : a regulatory mechanism? Trends Biochem . Sci . 16, 268-271 . Westermann R ., Johannsen M ., Unsicker K ., and Grothe C . ( 1990) Basic fibroblast growth factor (bFGF) immunoreactivity is present in chromaffin granules . J . Neurochern. 55, 285-292 . Whittemore S . R., White L . A ., and Sanon H . R . (1991) Acidic and basic fibroblast growth factor levels in spinal cord cultures are not regulated by alterations in heparin sulfate proteoglycan expression. Int . J. Dev . Neurosci. 9, 521-535 . Winkler H . and Smith A . D . (1975) The chromaffin granule and the storage of catecholamines, in Handbook of Physiology, Sec. 7: Endocrinology, Vol. Vl (Greep R . O . and Astwood E . B ., eds), pp . 321-339 . Williams and Wilkins, Baltimore . Woodward W . R., Nishi R ., Meshul C . K ., Williams T . E ., Coulombe M ., and Eckenstein F . P . ( 1992) Nuclear and cytoplasmic localization of basic fibroblast growth factor in astrocytes and CA2 hippocampal neurons . .l . Neurosci. 12, 142-152 .

J . Nrnrrochem., Vol. 64, No. 4. 1995