Eur. J. Biochem. 194, 825-829 (1990) 0FEBS 1990
Serotonin organelles of rabbit platelets contain synaptophysin Martin BAHLER’, Andrea M. CESURA’, Gunther FISCHER’, Herbert KUHN’, Ronald L. KLEIN’ and Mosi: Da PRADA2
’ Laboratory of Molecular and Cellular Neuroscience, The Rockefeller University, New York, USA Pharmaceutical Research Department, F. Hoffmann-La Roche Ltd, Basel, Switzerland (Received June 5/August 17, 1990) - EJB 90 0633
Synaptophysin, an integral membrane protein of synaptic vesicles in nerve terminals and a class of small translucent vesicles in neuroendocrine cells, was detected in intact rabbit platelets by immunoblotting, immunofluorescence staining and immuno-electron microscopy. In a highly purified preparation of serotonin organelles isolated from rabbit platelets, synaptophysin was enriched approximately 10 - 15-fold over platelet homogenate. About 80% of total platelet synaptophysin was present in this purified fraction. The apparent molecular mass ( x 38 kDa) and the extent of glycosylation of platelet-derived synaptophysin was more similar to the neuronal than to the neuroendocrine form of the protein. Immunofluorescence microscopy revealed that synaptophysin was compartmentalized in intact rabbit platelets and immuno-electron microscopy of subcellular fractions showed that it was localized exclusively to the membrane surface of serotonin organelles. No synaptophysin-like immunoreactivity was detected in platelets from other species such as human, guinea pig and rat. Another integral membrane protein of synaptic vesicles, p65, and a family of synaptic vesicle-associated phosphoproteins, the synapsins, were not detected in platelets of any species tested. These results provide evidence that serotonin organelles from rabbit platelets share a subset of protein components with synaptic vesicles from neurons. Synaptophysin in serotonin organelles from rabbit platelets, as suggested for small synaptic vesicles in neurons, might play a role in the formation of protein channels for the exocytotic release of serotonin.
Synaptic vesicles are secretory organelles (40 - 60 nm in diameter), which are clustered in presynaptic nerve terminals, take up classical neurotransmitter molecules, store them and release their content via exocytosis upon stimulation. Several proteins specific for these organelles, have been identified and characterized, e. g. synaptophysin, p65 and the synapsin protein family [I - 51. Synaptophysin and p65 have also been shown to be expressed in neuroendocrine cells [3, 61, in which synaptophysin was found to be localized to a newly discovered class of clear small vesicles apparently concentrated in the Golgi area [6]. Other reports, however, suggest that synaptophysin and p65 might not be exclusively localized on these structures, but that they are also present on secretory dense core organelles of larger diameter ( 2 60 nM), which store peptides [7, 81. Due to the presence of similar biochemical systems for the uptake, storage and metabolism of the neurotransmitter serotonin (5-hydroxy-tryptamine), blood platelets have often been used as a model for the study of serotoninergic neurons of the central nervous system [9, 101. Platelets contain at least two secretory organelles, namely the a-granules, with a diameter of about 300 nm, in which several peptides are stored, and the serotonin organelles (dense bodies, diameter 150 nm) which store serotonin and other monoamines together with a non-metabolic pool of 5’-phosphonucleotides (ATP, ADP, GTP, etc.) [II, 121. The monoamine content of these granules is depleted by reserpine and by benzo[u]quinolizine derivatives which have been shown to bind to the granular membrane [13] and to inhibit the granular amine Correspondence to A. M. Cesura, c/o Pharmaceutical Research Department, F. Hoffmann-La Roche Ltd, CH-4002 Basel, Switzerland
carrier [14]. In addition, various agents (e. g. thrombin) acting through specific receptors on the external cell membrane, cause the exocytotic release of the content of the platelet granules. The biochemical composition of platelet serotonin organelles is only poorly characterized. Because there are a number of similarities between neurons and platelets we investigated the possibility that serotonin organelles from blood platelets might share proteins with synaptic vesicles in neurons, namely synaptophysin, p65 and the synapsins. In the present report we show that rabbit blood platelets do in fact contain synaptophysin which appears to be localized exclusively on the serotonin organelles. However, no p65 or synapsins I and I1 was detected. MATERIALS AND METHODS Preparation of platelets
Blood from rabbits, rats, guinea pigs and human volunteers was collected using NazEDTA as anticoagulant; platelets were isolated according to standard procedures and washed twice in modified Tyrode buffer [I 51. The serotonin-storing organelles were isolated from rabbit platelets by ultracentrifugation of the platelet homogenate on a continuous density gradient of Urografinm [15]. Details about the isolation procedure and the morphological and biochemical characteristics of the serotonin organelles have been described elsewhere [12]. Antibodies and immunoblotting techniques
Mouse monoclonal and rabbit serum antibodies against rat synaptophysin were generated and characterized pre-
826
69 -
kDa
46
92-
-
69 46 30 -
30-
21 -
1 2
1
4
3
Fig. 1. S ~ ~ n i i p ~ o p / ~is~ prcsent .sin in rabbit platelets. Platelet proteins (120 pg/lane) from guinea pig (lane l), human (lane 2), rat (lane 3) and rabbit (lane 4) were separated by SDSjPAGE and then electrophoretically transferred to nitrocellulose. Blots were incubated with polyclonal antihodies followed by 1251-proteinA as described in Materials and Methods. The autoradiogram is shown. The molecular mass of standard proteins (in kDa) is shown on the left
2
3
4
5
6
7
Fig. 3. Synaptopliysin is mriched in serotonin orgnnelles purifiecl,fionr rabbit platelets. Autoradiogram of an immunoblot (see Materials and Methods) incubated with a monoclonal antibody directed against synaptophysin (SY38, Progen). Lane 1, purified serotonin organelles: 2, a-granules; 3, a-granules + heavy mitochondria; 4, light mitochondria; 5 , 6, microsomes and membrane debris; lane 7, rabbit platelet homogenate. Samples of 100 pg protein wcre loaded in each lane except in lane 1 where 30 pg protein was applied
351
46-
synaptophysin arnine granular carrier
301
2
3
4
5
6
Fig. 2. The estent of glyc,a.sjlation of rabbit platelet synaptophysin is more sirnilur to /lie neuronul than to the neuroendocrine f o r m qf’ the protein. Particulate membranes (30 pg) from rabbit platelet serotonin organellcs (lanes 1 2 ) , rat brain (lanes 3,4) and PC12 cells (lanes 5 , 6 ; kindly donated by Miss F. Ferracin, Dept. Forschung, Kantonsspital, Basel) were incubated in 0.1 M sodium phosphate pH 8.6 (containing 20 m M Na2EDTA. 0.6Y0 Triton X-100, 0.1% SDS, 2% 2mercaptoethanol and 0.5 mM phenylmethylsulfonylfluoride) with (even lanes) or without (odd lanes) 1.0 unit N-glycosidase I; (Boehringer Mannhcim, FRG) for 18 h a t 37 C. Samples were then clectrophoresed and blotted on nitrocellulose as above. Synaptophysin was detected using polyclonal antibodies. The molecular mass (in kDa) of two standard proteins is shown on the left
viously [l] and were kindly donated by Dr R. Jahn (Max Planck Institute, Munich. FRG). A commercially obtained monoclonal antibody against synaptophysin (SY38, Progen, Heidelberg, FRG) was also used. The monoclonal antibody against p65 was a gift of D r L. Reichardt (Howard Hughes Medical Institute, San Francisco, CA, USA). Antisera against synapsin I and I 1 were raised in rabbits using highly purified bovine synapsin 1 and 11 as antigens. Sodium dodecyl sulfate/ polyacrylamide gel electrophoresis (SDS/PAGE) was run according to Liiminli [I 61 and proteins were blotted on nitrocellulose filters according t o Towbin et al. (171. Blots were incu-
-
1 2 3 4 5 6 Fig. 4. Enriclimcnt ef’synaptoplzj,sin and ainine granular transporter in rabbit platelet .suhcelliihr ,f).actions. The content in platelet homogenates was taken as control value. In rabbit platelet homogenates, the amount of synaptophysin (determined by dotimmunoassay) and of the amine granular transporter (B,,, of [3H]dihydrotetrabenazine binding) were 30 ng/mg protein and 7.2 pmol/mg protein, respectively. The various fractions are numbered as in Fig. 3. The figure shows the results from a typical cxpcriment
bated with polyclonal antibodies and then, after extensive washing, reincubated with Iz5I-protein A as described [6]. When monoclonal antibodies were used, a secondary lz5Ilabeled sheep anti-mouse antibody was used. The nitrocellulose filters were then exposed to X-ray films for autoradiography. Quantification of synaptophysin was performed using a dot-immunobinding assay [18] with rat brain synaptophysin as a standard. By this method as little as 0.5 ng of the protein can be detected.
827
Fig. 5. Indirect immunofluorescencestaining of’ rabbit platelets for synaptophysin ( A ) is compared with background staining ( C j obtuined by the fluorescein-isothiocyanute-conjugated second antibody alone. In (B) and (D) the respective differential interference contrast micrographs are shown. Note in (A) the locally higher fluorescence intensities in subcellular compartments. Bar = 10 pm
Light and electron microscopy imrnunocytochernistry
For immunofluorescence experiments, washed rabbit platelets resuspended in modified Tyrode solution were plated on collagen-coated coverslips. After drying the coverslips at room temperature for 15 min, the platelets were fixed with 4% paraformaldehyde for 5 min and then incubated for 1 min with methanol at -20°C. Platelets were then stained for synaptophysin with an indirect fluorescence method using a monoclonal mouse anti-synaptophysin antibody (1 : 5 , SY38, Progen) and a fluorescein-isothiocyanate-conjugated rabbit anti-mouse antibody (1 :40, Dakopatts, via InstrumentenGesellschaft, Zurich, Switzerland). The different monoclonal antibodies against synaptophysin gave identical staining patterns. For immuno-electron microscopy with the immuno-gold technique [19], a suspension ( 5 pl) of purified serotonin organelles or other subcellular fractions from rabbit platelets in Tyrode buffer was applied to nickel grids, precoated with collodium and carbon. After 30 s, the liquid was removed and the grids, with the adherent platelet subcellular organelles, were immediately transferred and incubated for 15 min on a drop of fixing solution consisting of 2% paraformaldehyde, 0.1 YOglutaraldehyde in phosphate-buffered saline (20 mM KH2P04in 0.9% NaCI, pH 7.4, NaC1/Pi). After rinsing with buffer containing 0.1 % goat serum, the grids were incubated (2 h at 22°C) with a monoclonal antibody against synaptophysin (SY38, Progen). The same preparation omitting the anti-synaptophysin antibody was used as a control. The grids were then rinsed in NaCl/Pi and incubated (1 h at 22°C) with goat anti-mouse IgG gold (10-nm gold particle, Janssen
Pharmaceutica, Beerse, Belgium) at a dilution of 1 : 50. After washing with NaC1/Pi, air-dried grids were finally investigated in the transmission mode with a Philips CM12/STEM electron microscope. [3H]Dihydrotetrabenazine binding
The amount of amine granular transporter in rabbit platelet subcellular fractions was estimated by determining the maximal binding (Bmax) of the selective ligand [3H]dihydrotetrabenazine (CEA, Gyf-sur-Yvette, France; 15 Ci/mmol) [20]. B,,, values were calculated from saturation experiments using nine different radioactivity concentrations. Protein determination
Protein content was determined by a modification [21] of the method of Lowry et al. [22] and by the method of Bradford ~31. RESULTS AND DISCUSSION We tested, by immunoblotting with antibodies directed against synaptophysin, p65 and the synapsins, whether platelets from a variety of species contain detectable levels of these synaptic vesicle proteins. Interestingly, appreciable amounts of synaptophysin were found to be present in rabbit platelets, but not in platelets from other animal species, i.e. human, guinea pig and rat (Fig. 1). The identity of the rabbit platelet protein recognized by the antibodies as synaptophysin
828
Fig. 6. Electron microscopic localization Of'synuptoph~isinon rabbit platelet serotonin organelles ( A and B ) and in other subcellulur organelles ( C , 1-granules; D ,mitochondria; E , microsomes) . In A the isolated serotonin organelles were incubated with a monoclonal antibody against synaptophysin (SY38. Progen) followed by an antibody against mouse IgG labeled with colloidal gold (10-nm gold particles). In B the incubation with the primary antibody was omitted. Gold labeling is only visible in A but not in B. Bar = 0.5 pm
was established by using two different polyclonal antisera and two monoclonal antibodies directed against synaptophysin. Furthermore, the molecular mass of the protein recognized by the antibodies in rabbit platelets ( z38 kDa), as determined by SDS/PAGE. is very similar if not identical to the molecular mass of authentic synaptophysin from rat brain. Notably, synaptophysin from neuroendocrine cells exhibits a slightly lower migration rate than the corresponding protein from brain; this has been shown to be due to a difference in the glycosylation state of the protein [6]. After treatment of rabbit platelet serotonin organelles, rat brain synaptosomes and PC12 cell membranes with N-glycosidase F, a molecular mass of about 33 - 34 kDa for the deglycosylated protein was observed in all three tissues (Fig. 2 ) . An identical molecular mass value for deglycosylated synaptophysin has been previously reported [6, 241. Therefore, based on electrophoretic mobility, the extent of glycosylation of rabbit platelet synaptophysin seems to be more similar to that of the neuronal than of the neuroendocrine form of synaptophysin. The amount of synaptophysin in rabbit platelets, as estimated by dot-imniunoassay, was determined to be about 20 30 ng/mg total platelet protein. Compared to a value of 3.2 pg/ mg total protein found in rat forebrain homogenate [6], the amount of synaptophysin in rabbit platelets is about 1% of the amount present in rat forebrain.
After subcellular fractionation of rabbit platelets. synaptophysin was enriched about 10 - 15-foid in the fraction corresponding to pure serotonin organelles (Figs 3 and 4). The same amount of synaptophysin was also found after hypotonic lysis and repeated washing of these organelles, suggesting that the protein is intrinsic to the granular membrane. Some immunoreactivity was also detected in other subcellular fractions, but it was estimated that about 80% of the total synaptophysin along the gradient was present in the serotonin organelle fraction. As shown in Fig. 4, the profile of synaptophysin enrichment was similar to that obtained for the granular amine transporter, measured by [3H]dihydrotetrabenazine binding [20, 251. However, in the latter case a higher purification factor in respect to rabbit platelet homogenate was found. The reasons for this discrepancy is not understood. It might indicate that synaptophysin is also localized on other platelet organelles. In fact, conflicting results have been obtained concerning the presence of synaptophysin in peptide-storing large dense core granules of neurons and of neuroendocrine cells [2, 7, 8, 261, organelles which might be related to the platelet a-granules. Indirect immunofluorescence staining of intact rabbit platelets revealed that synaptophysin is localized in subcellular compartments (Fig. 5 ) . Immuno-electron microscopic localization of synaptophysin using a secondary antibody labeled
829 with colloidal gold, clearly showed the presence of gold particles on the surface of intact serotonin organelles purified from rabbit platelet (Fig. 6A), whereas no staining was seen when the primary antibody was omitted (Fig. 6B). Immunoelectron micrographs of the other subcellular fractions of rabbit platelets showed no evidence for the presence of synaptophysin on @-granules,mitochondria or microsomes. The sparse labeling sometimes observed was due to the presence of contaminating serotonin organelles. A similar immuno-gold staining pattern was observed when ultrathin sections of the various subcellular structures embedded in LR White (The London Resins Co. Ltd, Hampshire, UK) medium were analyzed (data not shown). These results would rule out the possibility that synaptophysin is also present on other platelet organelles. The finding that, using the SY38 monoclonal antibody, the immunogold label was exclusively present on the outer surface of the granules is in accordance with previous reports that assign the synaptophysin epitope recognized by the SY38 antibody (i.e. the near-carboxy-terminal region) to the cytoplasmic surface of the vesicles [27]. It is surprising that synaptophysin was only detected in rabbit platelets but not in platelets from other animal species tested (Fig. 1). This does not appear to be attributable to a lack of crossreactivity of the antibodies with synaptophysin of these species, since the antibodies did recognize synaptophysin in brain homogenates from these species [6] (and data not shown). However, since rabbit platelets show the highest number of serotonin granules/cell, accompanied by much higher concentration of monoamines and 5'-phosphonucleotides [28], our failure to detect the presence of synaptophysin in platelets of other species may be attributed to the very low amounts possibly present. Another interesting finding was that the integral membrane protein of synaptic vesicles, p65, which is normally coexpressed with synaptophysin, could not be detected by immunoblotting in platelets of any species tested, not even in platelets or purified serotonin organelles from rabbits (data not shown). In conclusion, our findings provide strong evidence that synaptophysin is a constituent of the membrane of the serotonin organelles of rabbit blood platelets. It has been suggested that synaptophysin might be involved in the exocytotic release process by forming specific hexameric protein channels [29, 301. The presence of synaptophysin in platelets suggests that serotonin-storing organelles of these cells and small synaptic vesicles of neurons might release neurotransmitter molecules by a related mechanism. In addition, since synaptophysin has so far been detected only in neurons and in neuroendocrine cells, platelets may either be of the same ontogenetic origin as neurons or else be a part of the diffuse neuroendocrine tissue. In this context, it is worth remembering that a neuron-specific enolase [31] has also been found in blood platelets [32]. The authors wish to thank Mrs Martine Buhler, Mr J.C. Brun and Mr R. Krauer for skilful technical assistance. M. B. and R. L.K. thank Paul Greengard for support.
REFERENCES 1. Jahn, R., Schiebler, W., Ouimet, C. & Greengard, P. (1985) Proc. Nut1 Acad. Sci. USA 82,4137-4141.
2. Wiedenmann, B. & Franke, W. W. (1985) Cell41, 1017-1028. 3. Matthew, W. D., Tsavaler, L. & Reichardt, F. L. (1981) J . Cell. Biol. 91, 257 - 269. 4. De Camilli, P., Harris, S. M., Huttner, W. B. & Greengard, P. (1983) J . Cell. Biol. 96, 1355-1373. 5. Huttner, W. B., Schiebler, W., Greengard, P. & De Camilli, P. (1983) J . Cell. Bid. 96, 1374-1388. 6. Navone, F., Jahn, R. D., Gioia, G., Stukenbrok, H., Greengard, P. & De Camilli, P. (1986) J. Cell. Biol. 102, 2511 -2527. 7. Lowe, A. W., Madeddu, L. & Kelly, R. B. (1988) J. Cell. Biol. 106, 51 - 59. 8. Fournier, S., Novas, M. L. & Trifaro, J.-M. (1989) J . Neurochem. 53,51- 59. 9. Pletscher, A. (1978) in Essay in neurochemistry and neuropharmacology (Youdim, M. B. H., Lovenberg, W., Sharman, D. F. & Lagnado, J. R., eds) vol. 3, pp. 49-99, John Wiley and Sons, New York. 10. Da Prada, M., Lorez, H. P. & Richards, J. G. (1982) in The secretory granule (Poisner, A. M. & Trifaro, J.-M., eds) pp. 279- 316, Elsevier Biomedical Press, Amsterdam. 11. Da Prada, M., Cesura, A. M., Launay, J.-M. & Richards, J. G. (1988) Experientia 44, 115-126. 12. Da Prada, M., Richards, J. G. & Lorez, H. P. (1978) in Platelets: a multidisciplinary approach (de Gaetano, G. & Garattini, S., eds) pp. 331 -353, Raven Press, New York. 13. Da Prada, M. & Pletscher, A. (1969) Eur. J . Pharmacol. 7, 4548. 14. Henry, J. P., Gamier, B., Roisin, M. P., Isambert, M. F. & Sherman, D. (1987) Ann. N. Y. Acad. Sci. 493, 194-206. 15. Da Prada, M. & Pletscher, A. (1968) Br. J . Pharmacol. 34, 591 597. 16. Limmli, U. K. (1970) Nature 227, 680-685. 17. Towbin, H., Staehelin, H. & Gordon, J. (1979) Proc. Nut1 Acad. Sci. USA 76, 4350-4354. 18. Jahn, R., Schiebler, W. & Greengard, P. (1984) Proc. Natl Acad. Sci. USA 81, 1684-1687. 19. Faulk, W. & Taylor, G. (1971) Immunocytochemistry 8, 10811083. 20. Scherman, D., Jaudon, P. & Henry, J. P. (1983) Proc. Nut/ Acad. Sci. USA 60, 584-588. 21. Markwell, M. A. K., Haas, S. M., Bieber, L. L. & Tolbert, N. E. (1978) Anal. Biochem. 87,206-210. 22. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J . (1951) J . Bid. Chem. 193,265-275. 23. Bradford, M. M. (1976) Anal. Biochem. 72,248-254. 24. Rehm, H., Wiedenmann, B. & Betz, H. (1986) EMBO J . 5,535 541. 25. Cesura, A. M., Bertocci, B. & Da Prada, M. (1990) Eur. J . Pharmacol. 186,95 - 104. 26. Navone, F., Di Gioia, G., Matteoli, M. & De Camilli, P. (1988) Proc. Alfred Benzon Symp. 25,433 -450. 27. Leube, R. E., Kaiser, P., Seiter, A., Zembelmann, R., Franke, W. E., Rehm, H., Knaus, P., Prior, P., Betz, H., Beyreuther, K. & Wiedenmann, B. (1987) EMBU J . 6, 3261 -3268. 28. Da Prada, M., Picotti, G. B., Kettler, R. & Launay, J.-M. (1980) in Platelets: cellular response mechanism and their biological significance (Rotham, A,, Meyer, F. A., Gitler, C. & Silberberg, A , , eds) pp. 277-288, John Wiley and Sons, New York. 29. Thomas, L., Hartung, K., Langosch, D., Rehm, H., Bamberg, E., Franke, W. W. & Betz, H. (1988) Science 242, 1050-1053. 30. Breckenridge, L. J. & Almers, W. (1987) Nuture328, 814-817. 31. Marangos, P. J., Schmechel, D., Parma, A,, Clark, R. L. & Goodwin, F. K. (1979) J . Neurochem. 33, 319-329. 32. Campbell, J. C., Marangos, P. J., Murphy, D. L. & Pearse, A. G. E. (1981) Adv. Biosci. 31,203-211.
Serotonin organelles of rabbit platelets contain synaptophysin Martin BAHLER’, Andrea M. CESURA’, Gunther FISCHER’, Herbert KUHN’, Ronald L. KLEIN’ and Mosi: Da PRADA2
’ Laboratory of Molecular and Cellular Neuroscience, The Rockefeller University, New York, USA Pharmaceutical Research Department, F. Hoffmann-La Roche Ltd, Basel, Switzerland (Received June 5/August 17, 1990) - EJB 90 0633
Synaptophysin, an integral membrane protein of synaptic vesicles in nerve terminals and a class of small translucent vesicles in neuroendocrine cells, was detected in intact rabbit platelets by immunoblotting, immunofluorescence staining and immuno-electron microscopy. In a highly purified preparation of serotonin organelles isolated from rabbit platelets, synaptophysin was enriched approximately 10 - 15-fold over platelet homogenate. About 80% of total platelet synaptophysin was present in this purified fraction. The apparent molecular mass ( x 38 kDa) and the extent of glycosylation of platelet-derived synaptophysin was more similar to the neuronal than to the neuroendocrine form of the protein. Immunofluorescence microscopy revealed that synaptophysin was compartmentalized in intact rabbit platelets and immuno-electron microscopy of subcellular fractions showed that it was localized exclusively to the membrane surface of serotonin organelles. No synaptophysin-like immunoreactivity was detected in platelets from other species such as human, guinea pig and rat. Another integral membrane protein of synaptic vesicles, p65, and a family of synaptic vesicle-associated phosphoproteins, the synapsins, were not detected in platelets of any species tested. These results provide evidence that serotonin organelles from rabbit platelets share a subset of protein components with synaptic vesicles from neurons. Synaptophysin in serotonin organelles from rabbit platelets, as suggested for small synaptic vesicles in neurons, might play a role in the formation of protein channels for the exocytotic release of serotonin.
Synaptic vesicles are secretory organelles (40 - 60 nm in diameter), which are clustered in presynaptic nerve terminals, take up classical neurotransmitter molecules, store them and release their content via exocytosis upon stimulation. Several proteins specific for these organelles, have been identified and characterized, e. g. synaptophysin, p65 and the synapsin protein family [I - 51. Synaptophysin and p65 have also been shown to be expressed in neuroendocrine cells [3, 61, in which synaptophysin was found to be localized to a newly discovered class of clear small vesicles apparently concentrated in the Golgi area [6]. Other reports, however, suggest that synaptophysin and p65 might not be exclusively localized on these structures, but that they are also present on secretory dense core organelles of larger diameter ( 2 60 nM), which store peptides [7, 81. Due to the presence of similar biochemical systems for the uptake, storage and metabolism of the neurotransmitter serotonin (5-hydroxy-tryptamine), blood platelets have often been used as a model for the study of serotoninergic neurons of the central nervous system [9, 101. Platelets contain at least two secretory organelles, namely the a-granules, with a diameter of about 300 nm, in which several peptides are stored, and the serotonin organelles (dense bodies, diameter 150 nm) which store serotonin and other monoamines together with a non-metabolic pool of 5’-phosphonucleotides (ATP, ADP, GTP, etc.) [II, 121. The monoamine content of these granules is depleted by reserpine and by benzo[u]quinolizine derivatives which have been shown to bind to the granular membrane [13] and to inhibit the granular amine Correspondence to A. M. Cesura, c/o Pharmaceutical Research Department, F. Hoffmann-La Roche Ltd, CH-4002 Basel, Switzerland
carrier [14]. In addition, various agents (e. g. thrombin) acting through specific receptors on the external cell membrane, cause the exocytotic release of the content of the platelet granules. The biochemical composition of platelet serotonin organelles is only poorly characterized. Because there are a number of similarities between neurons and platelets we investigated the possibility that serotonin organelles from blood platelets might share proteins with synaptic vesicles in neurons, namely synaptophysin, p65 and the synapsins. In the present report we show that rabbit blood platelets do in fact contain synaptophysin which appears to be localized exclusively on the serotonin organelles. However, no p65 or synapsins I and I1 was detected. MATERIALS AND METHODS Preparation of platelets
Blood from rabbits, rats, guinea pigs and human volunteers was collected using NazEDTA as anticoagulant; platelets were isolated according to standard procedures and washed twice in modified Tyrode buffer [I 51. The serotonin-storing organelles were isolated from rabbit platelets by ultracentrifugation of the platelet homogenate on a continuous density gradient of Urografinm [15]. Details about the isolation procedure and the morphological and biochemical characteristics of the serotonin organelles have been described elsewhere [12]. Antibodies and immunoblotting techniques
Mouse monoclonal and rabbit serum antibodies against rat synaptophysin were generated and characterized pre-
826
69 -
kDa
46
92-
-
69 46 30 -
30-
21 -
1 2
1
4
3
Fig. 1. S ~ ~ n i i p ~ o p / ~is~ prcsent .sin in rabbit platelets. Platelet proteins (120 pg/lane) from guinea pig (lane l), human (lane 2), rat (lane 3) and rabbit (lane 4) were separated by SDSjPAGE and then electrophoretically transferred to nitrocellulose. Blots were incubated with polyclonal antihodies followed by 1251-proteinA as described in Materials and Methods. The autoradiogram is shown. The molecular mass of standard proteins (in kDa) is shown on the left
2
3
4
5
6
7
Fig. 3. Synaptopliysin is mriched in serotonin orgnnelles purifiecl,fionr rabbit platelets. Autoradiogram of an immunoblot (see Materials and Methods) incubated with a monoclonal antibody directed against synaptophysin (SY38, Progen). Lane 1, purified serotonin organelles: 2, a-granules; 3, a-granules + heavy mitochondria; 4, light mitochondria; 5 , 6, microsomes and membrane debris; lane 7, rabbit platelet homogenate. Samples of 100 pg protein wcre loaded in each lane except in lane 1 where 30 pg protein was applied
351
46-
synaptophysin arnine granular carrier
301
2
3
4
5
6
Fig. 2. The estent of glyc,a.sjlation of rabbit platelet synaptophysin is more sirnilur to /lie neuronul than to the neuroendocrine f o r m qf’ the protein. Particulate membranes (30 pg) from rabbit platelet serotonin organellcs (lanes 1 2 ) , rat brain (lanes 3,4) and PC12 cells (lanes 5 , 6 ; kindly donated by Miss F. Ferracin, Dept. Forschung, Kantonsspital, Basel) were incubated in 0.1 M sodium phosphate pH 8.6 (containing 20 m M Na2EDTA. 0.6Y0 Triton X-100, 0.1% SDS, 2% 2mercaptoethanol and 0.5 mM phenylmethylsulfonylfluoride) with (even lanes) or without (odd lanes) 1.0 unit N-glycosidase I; (Boehringer Mannhcim, FRG) for 18 h a t 37 C. Samples were then clectrophoresed and blotted on nitrocellulose as above. Synaptophysin was detected using polyclonal antibodies. The molecular mass (in kDa) of two standard proteins is shown on the left
viously [l] and were kindly donated by Dr R. Jahn (Max Planck Institute, Munich. FRG). A commercially obtained monoclonal antibody against synaptophysin (SY38, Progen, Heidelberg, FRG) was also used. The monoclonal antibody against p65 was a gift of D r L. Reichardt (Howard Hughes Medical Institute, San Francisco, CA, USA). Antisera against synapsin I and I 1 were raised in rabbits using highly purified bovine synapsin 1 and 11 as antigens. Sodium dodecyl sulfate/ polyacrylamide gel electrophoresis (SDS/PAGE) was run according to Liiminli [I 61 and proteins were blotted on nitrocellulose filters according t o Towbin et al. (171. Blots were incu-
-
1 2 3 4 5 6 Fig. 4. Enriclimcnt ef’synaptoplzj,sin and ainine granular transporter in rabbit platelet .suhcelliihr ,f).actions. The content in platelet homogenates was taken as control value. In rabbit platelet homogenates, the amount of synaptophysin (determined by dotimmunoassay) and of the amine granular transporter (B,,, of [3H]dihydrotetrabenazine binding) were 30 ng/mg protein and 7.2 pmol/mg protein, respectively. The various fractions are numbered as in Fig. 3. The figure shows the results from a typical cxpcriment
bated with polyclonal antibodies and then, after extensive washing, reincubated with Iz5I-protein A as described [6]. When monoclonal antibodies were used, a secondary lz5Ilabeled sheep anti-mouse antibody was used. The nitrocellulose filters were then exposed to X-ray films for autoradiography. Quantification of synaptophysin was performed using a dot-immunobinding assay [18] with rat brain synaptophysin as a standard. By this method as little as 0.5 ng of the protein can be detected.
827
Fig. 5. Indirect immunofluorescencestaining of’ rabbit platelets for synaptophysin ( A ) is compared with background staining ( C j obtuined by the fluorescein-isothiocyanute-conjugated second antibody alone. In (B) and (D) the respective differential interference contrast micrographs are shown. Note in (A) the locally higher fluorescence intensities in subcellular compartments. Bar = 10 pm
Light and electron microscopy imrnunocytochernistry
For immunofluorescence experiments, washed rabbit platelets resuspended in modified Tyrode solution were plated on collagen-coated coverslips. After drying the coverslips at room temperature for 15 min, the platelets were fixed with 4% paraformaldehyde for 5 min and then incubated for 1 min with methanol at -20°C. Platelets were then stained for synaptophysin with an indirect fluorescence method using a monoclonal mouse anti-synaptophysin antibody (1 : 5 , SY38, Progen) and a fluorescein-isothiocyanate-conjugated rabbit anti-mouse antibody (1 :40, Dakopatts, via InstrumentenGesellschaft, Zurich, Switzerland). The different monoclonal antibodies against synaptophysin gave identical staining patterns. For immuno-electron microscopy with the immuno-gold technique [19], a suspension ( 5 pl) of purified serotonin organelles or other subcellular fractions from rabbit platelets in Tyrode buffer was applied to nickel grids, precoated with collodium and carbon. After 30 s, the liquid was removed and the grids, with the adherent platelet subcellular organelles, were immediately transferred and incubated for 15 min on a drop of fixing solution consisting of 2% paraformaldehyde, 0.1 YOglutaraldehyde in phosphate-buffered saline (20 mM KH2P04in 0.9% NaCI, pH 7.4, NaC1/Pi). After rinsing with buffer containing 0.1 % goat serum, the grids were incubated (2 h at 22°C) with a monoclonal antibody against synaptophysin (SY38, Progen). The same preparation omitting the anti-synaptophysin antibody was used as a control. The grids were then rinsed in NaCl/Pi and incubated (1 h at 22°C) with goat anti-mouse IgG gold (10-nm gold particle, Janssen
Pharmaceutica, Beerse, Belgium) at a dilution of 1 : 50. After washing with NaC1/Pi, air-dried grids were finally investigated in the transmission mode with a Philips CM12/STEM electron microscope. [3H]Dihydrotetrabenazine binding
The amount of amine granular transporter in rabbit platelet subcellular fractions was estimated by determining the maximal binding (Bmax) of the selective ligand [3H]dihydrotetrabenazine (CEA, Gyf-sur-Yvette, France; 15 Ci/mmol) [20]. B,,, values were calculated from saturation experiments using nine different radioactivity concentrations. Protein determination
Protein content was determined by a modification [21] of the method of Lowry et al. [22] and by the method of Bradford ~31. RESULTS AND DISCUSSION We tested, by immunoblotting with antibodies directed against synaptophysin, p65 and the synapsins, whether platelets from a variety of species contain detectable levels of these synaptic vesicle proteins. Interestingly, appreciable amounts of synaptophysin were found to be present in rabbit platelets, but not in platelets from other animal species, i.e. human, guinea pig and rat (Fig. 1). The identity of the rabbit platelet protein recognized by the antibodies as synaptophysin
828
Fig. 6. Electron microscopic localization Of'synuptoph~isinon rabbit platelet serotonin organelles ( A and B ) and in other subcellulur organelles ( C , 1-granules; D ,mitochondria; E , microsomes) . In A the isolated serotonin organelles were incubated with a monoclonal antibody against synaptophysin (SY38. Progen) followed by an antibody against mouse IgG labeled with colloidal gold (10-nm gold particles). In B the incubation with the primary antibody was omitted. Gold labeling is only visible in A but not in B. Bar = 0.5 pm
was established by using two different polyclonal antisera and two monoclonal antibodies directed against synaptophysin. Furthermore, the molecular mass of the protein recognized by the antibodies in rabbit platelets ( z38 kDa), as determined by SDS/PAGE. is very similar if not identical to the molecular mass of authentic synaptophysin from rat brain. Notably, synaptophysin from neuroendocrine cells exhibits a slightly lower migration rate than the corresponding protein from brain; this has been shown to be due to a difference in the glycosylation state of the protein [6]. After treatment of rabbit platelet serotonin organelles, rat brain synaptosomes and PC12 cell membranes with N-glycosidase F, a molecular mass of about 33 - 34 kDa for the deglycosylated protein was observed in all three tissues (Fig. 2 ) . An identical molecular mass value for deglycosylated synaptophysin has been previously reported [6, 241. Therefore, based on electrophoretic mobility, the extent of glycosylation of rabbit platelet synaptophysin seems to be more similar to that of the neuronal than of the neuroendocrine form of synaptophysin. The amount of synaptophysin in rabbit platelets, as estimated by dot-imniunoassay, was determined to be about 20 30 ng/mg total platelet protein. Compared to a value of 3.2 pg/ mg total protein found in rat forebrain homogenate [6], the amount of synaptophysin in rabbit platelets is about 1% of the amount present in rat forebrain.
After subcellular fractionation of rabbit platelets. synaptophysin was enriched about 10 - 15-foid in the fraction corresponding to pure serotonin organelles (Figs 3 and 4). The same amount of synaptophysin was also found after hypotonic lysis and repeated washing of these organelles, suggesting that the protein is intrinsic to the granular membrane. Some immunoreactivity was also detected in other subcellular fractions, but it was estimated that about 80% of the total synaptophysin along the gradient was present in the serotonin organelle fraction. As shown in Fig. 4, the profile of synaptophysin enrichment was similar to that obtained for the granular amine transporter, measured by [3H]dihydrotetrabenazine binding [20, 251. However, in the latter case a higher purification factor in respect to rabbit platelet homogenate was found. The reasons for this discrepancy is not understood. It might indicate that synaptophysin is also localized on other platelet organelles. In fact, conflicting results have been obtained concerning the presence of synaptophysin in peptide-storing large dense core granules of neurons and of neuroendocrine cells [2, 7, 8, 261, organelles which might be related to the platelet a-granules. Indirect immunofluorescence staining of intact rabbit platelets revealed that synaptophysin is localized in subcellular compartments (Fig. 5 ) . Immuno-electron microscopic localization of synaptophysin using a secondary antibody labeled
829 with colloidal gold, clearly showed the presence of gold particles on the surface of intact serotonin organelles purified from rabbit platelet (Fig. 6A), whereas no staining was seen when the primary antibody was omitted (Fig. 6B). Immunoelectron micrographs of the other subcellular fractions of rabbit platelets showed no evidence for the presence of synaptophysin on @-granules,mitochondria or microsomes. The sparse labeling sometimes observed was due to the presence of contaminating serotonin organelles. A similar immuno-gold staining pattern was observed when ultrathin sections of the various subcellular structures embedded in LR White (The London Resins Co. Ltd, Hampshire, UK) medium were analyzed (data not shown). These results would rule out the possibility that synaptophysin is also present on other platelet organelles. The finding that, using the SY38 monoclonal antibody, the immunogold label was exclusively present on the outer surface of the granules is in accordance with previous reports that assign the synaptophysin epitope recognized by the SY38 antibody (i.e. the near-carboxy-terminal region) to the cytoplasmic surface of the vesicles [27]. It is surprising that synaptophysin was only detected in rabbit platelets but not in platelets from other animal species tested (Fig. 1). This does not appear to be attributable to a lack of crossreactivity of the antibodies with synaptophysin of these species, since the antibodies did recognize synaptophysin in brain homogenates from these species [6] (and data not shown). However, since rabbit platelets show the highest number of serotonin granules/cell, accompanied by much higher concentration of monoamines and 5'-phosphonucleotides [28], our failure to detect the presence of synaptophysin in platelets of other species may be attributed to the very low amounts possibly present. Another interesting finding was that the integral membrane protein of synaptic vesicles, p65, which is normally coexpressed with synaptophysin, could not be detected by immunoblotting in platelets of any species tested, not even in platelets or purified serotonin organelles from rabbits (data not shown). In conclusion, our findings provide strong evidence that synaptophysin is a constituent of the membrane of the serotonin organelles of rabbit blood platelets. It has been suggested that synaptophysin might be involved in the exocytotic release process by forming specific hexameric protein channels [29, 301. The presence of synaptophysin in platelets suggests that serotonin-storing organelles of these cells and small synaptic vesicles of neurons might release neurotransmitter molecules by a related mechanism. In addition, since synaptophysin has so far been detected only in neurons and in neuroendocrine cells, platelets may either be of the same ontogenetic origin as neurons or else be a part of the diffuse neuroendocrine tissue. In this context, it is worth remembering that a neuron-specific enolase [31] has also been found in blood platelets [32]. The authors wish to thank Mrs Martine Buhler, Mr J.C. Brun and Mr R. Krauer for skilful technical assistance. M. B. and R. L.K. thank Paul Greengard for support.
REFERENCES 1. Jahn, R., Schiebler, W., Ouimet, C. & Greengard, P. (1985) Proc. Nut1 Acad. Sci. USA 82,4137-4141.
2. Wiedenmann, B. & Franke, W. W. (1985) Cell41, 1017-1028. 3. Matthew, W. D., Tsavaler, L. & Reichardt, F. L. (1981) J . Cell. Biol. 91, 257 - 269. 4. De Camilli, P., Harris, S. M., Huttner, W. B. & Greengard, P. (1983) J . Cell. Biol. 96, 1355-1373. 5. Huttner, W. B., Schiebler, W., Greengard, P. & De Camilli, P. (1983) J . Cell. Bid. 96, 1374-1388. 6. Navone, F., Jahn, R. D., Gioia, G., Stukenbrok, H., Greengard, P. & De Camilli, P. (1986) J. Cell. Biol. 102, 2511 -2527. 7. Lowe, A. W., Madeddu, L. & Kelly, R. B. (1988) J. Cell. Biol. 106, 51 - 59. 8. Fournier, S., Novas, M. L. & Trifaro, J.-M. (1989) J . Neurochem. 53,51- 59. 9. Pletscher, A. (1978) in Essay in neurochemistry and neuropharmacology (Youdim, M. B. H., Lovenberg, W., Sharman, D. F. & Lagnado, J. R., eds) vol. 3, pp. 49-99, John Wiley and Sons, New York. 10. Da Prada, M., Lorez, H. P. & Richards, J. G. (1982) in The secretory granule (Poisner, A. M. & Trifaro, J.-M., eds) pp. 279- 316, Elsevier Biomedical Press, Amsterdam. 11. Da Prada, M., Cesura, A. M., Launay, J.-M. & Richards, J. G. (1988) Experientia 44, 115-126. 12. Da Prada, M., Richards, J. G. & Lorez, H. P. (1978) in Platelets: a multidisciplinary approach (de Gaetano, G. & Garattini, S., eds) pp. 331 -353, Raven Press, New York. 13. Da Prada, M. & Pletscher, A. (1969) Eur. J . Pharmacol. 7, 4548. 14. Henry, J. P., Gamier, B., Roisin, M. P., Isambert, M. F. & Sherman, D. (1987) Ann. N. Y. Acad. Sci. 493, 194-206. 15. Da Prada, M. & Pletscher, A. (1968) Br. J . Pharmacol. 34, 591 597. 16. Limmli, U. K. (1970) Nature 227, 680-685. 17. Towbin, H., Staehelin, H. & Gordon, J. (1979) Proc. Nut1 Acad. Sci. USA 76, 4350-4354. 18. Jahn, R., Schiebler, W. & Greengard, P. (1984) Proc. Natl Acad. Sci. USA 81, 1684-1687. 19. Faulk, W. & Taylor, G. (1971) Immunocytochemistry 8, 10811083. 20. Scherman, D., Jaudon, P. & Henry, J. P. (1983) Proc. Nut/ Acad. Sci. USA 60, 584-588. 21. Markwell, M. A. K., Haas, S. M., Bieber, L. L. & Tolbert, N. E. (1978) Anal. Biochem. 87,206-210. 22. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J . (1951) J . Bid. Chem. 193,265-275. 23. Bradford, M. M. (1976) Anal. Biochem. 72,248-254. 24. Rehm, H., Wiedenmann, B. & Betz, H. (1986) EMBO J . 5,535 541. 25. Cesura, A. M., Bertocci, B. & Da Prada, M. (1990) Eur. J . Pharmacol. 186,95 - 104. 26. Navone, F., Di Gioia, G., Matteoli, M. & De Camilli, P. (1988) Proc. Alfred Benzon Symp. 25,433 -450. 27. Leube, R. E., Kaiser, P., Seiter, A., Zembelmann, R., Franke, W. E., Rehm, H., Knaus, P., Prior, P., Betz, H., Beyreuther, K. & Wiedenmann, B. (1987) EMBU J . 6, 3261 -3268. 28. Da Prada, M., Picotti, G. B., Kettler, R. & Launay, J.-M. (1980) in Platelets: cellular response mechanism and their biological significance (Rotham, A,, Meyer, F. A., Gitler, C. & Silberberg, A , , eds) pp. 277-288, John Wiley and Sons, New York. 29. Thomas, L., Hartung, K., Langosch, D., Rehm, H., Bamberg, E., Franke, W. W. & Betz, H. (1988) Science 242, 1050-1053. 30. Breckenridge, L. J. & Almers, W. (1987) Nuture328, 814-817. 31. Marangos, P. J., Schmechel, D., Parma, A,, Clark, R. L. & Goodwin, F. K. (1979) J . Neurochem. 33, 319-329. 32. Campbell, J. C., Marangos, P. J., Murphy, D. L. & Pearse, A. G. E. (1981) Adv. Biosci. 31,203-211.
