Muscle-Like Contractile Proteins and Tubulin in Synaptosomes (Brain/Peptide Mapping/Actin, Myosin, Tropomyosin/Membrane-Associated Protein) A

Proc. Nat. Acad. Sci. USA - Vol. 71, No. 11, pp. 4472-4476, November 1974

Muscle-Like Contractile Proteins and Tubulin in Synaptosomes (brain/peptide mapping/actin, myosin, tropomyosin/membrane-associated protein) A. L. BLITZ AND R. E. FINE Department of Physiology, Boston University School of Medicine, 80 East Concord Street, Boston, Massachusetts 02118 Communicated by Frances 0. Schmitt, August 16, 1974

ABSTRACT Material in major bands with molecular MATERIALS AND METHODS weights corresponding to those of actin, brain tropomyosin, and myosin is present in purified rat synaptosomes Isolation and Fractionation of Synaptosomes. Rat cortical dissolved in sodium dodecyl sulfate and subjected to synaptosomes were purified from washed brain mitochondrial electrophoresis on dodecyl sulfate-acrylamide gels. A band pellets by centrifugation through sucrose gradients (10) or corresponding to tubulin appears to be the major con- stituent of synaptosomes, confirming the work of Feit and Ficoll gradients (11), or by flotation on Ficoll (12). his coworkers. We have demonstrated by peptide mapping Synaptosomal membranes were obtained by lysis of synapto- that the proteins in these bands have strong chemical somes in distilled water (13) or in 1 mM sodium phosphate similarities to actin, brain tropomyosin, myosin, and containing 0.1 mM EDTA (12) and centrifugation through tubulin. We have prepared synaptic membrane, vesicle, and soluble fractions from synaptosomes. The polypeptide discontinuous sucrose gradients or through continuous sucrose composition of synaptic membranes, as determined by gradients (14). dodecyl sulfate-acrylamide gel electrophoresis, is similar Crude synaptic vesicles were obtained from the upper phase to that of synaptosomes, with tubulin, actin, and tropo- of the continuous sucrose gradients used to purify membranes. myosin being major constituents. Synaptic vesicles have Purer vesicles were obtained by lysis of a brain mitochondrial as their major polypeptide an unidentified protein with a molecular weight of 50,000; they also have many bands in pellet followed by centrifugation through a discontinuous common with synaptosomes. The soluble fraction pre- gradient (13). dominantly contains actin and tubulln. The possibility that the muscle-like contractile proteins and tubulin are Electron Microscopy. Pellets were fixed in Karnovsky's membrane-associated in various cell types is discussed, as fixative (15), post-fixed in Millonig's buffered 1% osmium is their possible role in neurotransmitter release. tetroxide (16), dehydrated in graded ethanol solutions, and Increasingly, reports have appeared of muscle-like proteins embedded in Epon-araldite. identified in or isolated from brain. Fine and Bray (1) have Electro- and Sodium Dodecyl Sulfate (NaDodSO4)-Acrylamide shown an actin-like protein in embryonic chick brain, phoresis of Synaptosomal Proteins. Pellets containing synapto- more recently, Fine et al. (2) have isolated a tropomyosin-like somes, membranes, or vesicles were dissolved in 2% NaDod- protein from the same source. The isolation from brain of a at 1000 for biochem- S04 containing 5% 2-mercaptoethanol by heating calcium-activated adenosinetriphosphatase similar 2-5 min. The samples were subjected to electrophoresis at pH ically to muscle actomyosin has also been reported (3). 7.2 on 7.5% acrylamide gels containing 0.8% N,N'-methylene- Tubulin, the microtubule subunit protein, exists in highest bisacrylamide (17). concentration in whole brain (4, 5) and has also been reported in nerve endings (6). Recovery of Proteins from NaDodSO4-Acrylamide Gels. The Since both muscle contraction and release of neurotrans- desired bands were cut from the gels, divided into small cubes, mitter from axon termini are dependent on calcium ion (7, 8), and extracted for 2 days at 370 with 1-2 ml of 50 mM sodium it is of interest to search for a role for brain contractile proteins phosphate (pH 7.2) containing 0.1% NaDodSO4 (18). The in synaptic transmission. These considerations led Berl et al. resulting blue extracts were dialyzed against methanol con- (9) to the identification of a calcium-stimulated, myosin-like taining 5% acetic acid until colorless and then further dialyzed adenosinetriphosphatase activity associated with synaptic against 0.1 M ammonium bicarbonate overnight. The dialyzed vesicles which is stimulated by synaptic membrane compo- solutions were lyophilized, and the residues were dissolved in nents. a minimum volume of distilled water and lyophilized again. We offer here more rigorous criteria for the presence and distribution of proteins similar to actin, myosin, and tropo- Proteolytic Digestion of Proteins and Peptide Mapping. The myosin in rat cortical synaptosomes and in the vesicle, mem- lyophilized proteins were digested with either chymotrypsin brane, and soluble fractions of lysed synaptosomes. We also or iA-tosylamido-2-phenylethyl chloromethyl ketone offer evidence that confirms an earlier report of tubulin in (TPCK)-trypsin by standard procedures. For peptide map- nerve endings (6), as well as evidence concerning its localiza- ping, 30-50 .g of protein, estimated with flourescamine (19), tion. were applied to 20 X 20-cm silica gel G thin-layer chromatography plates (250 Mm in thickness) and subjected to chro- Abbreviation: NaDodSO4, sodium dodecyl sulfate. matography followed by electrophoresis at pH 6.5 (2). 4472 Downloaded by guest on September 26, 2021 Proc. Nat. Acad. Sci. USA 71 (1974) Contractile Proteins in Synaptosomes 4473

Preparation of Protein Standards. Brain myosin was isolated from calf brain by the method of Adelstein and coworkers (20) used for the preparation of myosin from fibroblasts. This method was first applied to brain by Dennis Bray and coworkers (unpublished). This preparation was resolved by NaDodSO4-acrylamide electrophoresis into two bands containing material of high molecular weight (200,000 and 170,000) and one lower molecular weight component. It displayed a potassium-EDTA ATPase activity of 0.7 ,omol/- min per mg of protein. Chick brain tropomysin was isolated by the method of Fine and coworkers (2). Rat brain myelin was prepared by the method of Norton and Poduslo (21). Rabbit muscle actin and rat brain tubulin were the generous gifts of Dr. William Lehman and Dr. Clara Asnes, respectively. Rat brain actin and tubulin were also obtained from a brain homogenate by pre- cipitation with vinblastine sulfate (1).

RESULTS Morphological Purity of Synaptosomes and Their Subfrac- tions. The electron micrographs of the synaptosome pellet (Fig. la) displays a nearly homogeneous preparation of well- preserved nerve-ending particles. There is little myelin, a frequent contaminant of synaptosome preparations, few free mitochondria or synaptic vesicles, and a few empty synaptosomes. The most prevalent contaminant is a structure that appears to derive from post-synaptic material, either dendrites or axons. The membrane fraction (Fig. lb) shows mostly empty membranes, which vary widely in diameter. Some membranes have vesicles still attached, and some free vesicles can be seen. The vesicle fraction (Fig. ic) is perhaps not so pure as the other two fractions. It appears to be contaminated with small (about 1500 A in diameter) membranes. These structures may, however, be large synaptic vesicles. They are only about 50% larger than the vesicles isolated from the Torpedo electroplax (22). Presence in Synaptosomes and Subfractions of Proteimns Similar to Myosin, Tubulin, Actin, and Tropomyosin. The photograph in Fig. 2a shows the banding patterns of NaDod- S04-acrylamide gels of NaDodSO4-soluble proteins of synaptosomes, synaptic membranes, synaptic vesicles, and the soluble proteins, for synaptosomes obtained after lysis and centrifugation to remove particulate material. A diagram of the gels in Fig. 2a is also presented (Fig. 2b) to clarify the poor contrast of the photographs*. There are major components in some or all of these fractions that migrate with actin, tubulin, and brain tropomyosin, and less intense bands that migrate with the two high-molecular-weight components of brain myosin. Two components are also present in platelet myosin (23). These gel patterns are highly reproducible from FIG. 1. Electron micrographs of synaptosomes and synapto- one preparation to the next. Also, synaptosomes prepared by some subfractions. (a) Synaptosomes prepared by flotation on three different procedures all yield nearly identical gel pat- Ficoll (12). X23,500. (b) Synaptic membrane fraction prepared terns. by lysis of synaptosomes followed by purification on discontinuous The gels of membranes prepared by two different procedures sucrose gradients (13). X 19,300. (c) Synaptic vesicle fraction are nearly identical to those of whole synaptosomes except for prepared by lysis of a rat brain mitochondrial pellet followed by the absence of a few minor bands which may represent vesicle purification on discontinuous sucrose gradients (13). X26,900. or mitrochondrial proteins. Membranes prepared by lysis of synaptosomes in phosphate buffer containing EDTA (12) contain less actin and tropomyosin. * The original photograph of the purified vesicle fraction is not Most of the material in the bands in the gels of the crude shown because of its poor contrast. vesicle preparation migrate with those of whole synaptosomes Downloaded by guest on September 26, 2021 4474 Biochemistry: Blitz and Fine Proc. Nat. Acad. Sci. USA 71 (1974)

b MYI 4 200,000 MY2 4 170,000 a MY 1 MY 2

I TUB 4 55,000 * TUB ACT 4 45,000 ACT I I I BTM 4 29,000 ! BTM

I m a pp. 4 i S-T mol. wt. SYN MEM VES SOL STD SYN MEM VESc VESp SOL STD FIG. 2 (a). NaDodSO4-acrylamide gels of synaptosomes and subfractions. (b) Diagram of NaDodSO4 gels shown in (a). NaDodSO4- acrylamide gels containing 7.5% acrylamide, 0.8% N,N'-methylene-bisacrylamide, and 0.1% NaDodSO4 were subjected to electrophoresis at pH 7.2 (17) and stained with Coomassie blue R. SYN, synaptosomes prepared on Ficoll gradients (11). MEM, synaptic membranes obtained by lysing synaptosomes in distilled water followed by purification on continuous sucrose gradients (14). VESc, a crude synaptic vesicle preparation obtained from the gradients used to purify MEM. VESp, a purer synaptic vesicle fraction obtained by lysis of a brain mitochondrial pellet in distilled water and purified on discontinuous sucrose gradients (13) (a photograph of this gel is not shown because of lack of contrast). SOL, soluble synaptosome proteins obtained as a 10% trichloroacetic acid precipitate from the upper portion of the gradients used to obtain MEM and VESc. STD, standards: MY1 and MY2, large-molecular-weight components of calf brain myosin; TUB, rat brain microtubule subunit (tubulin); ACT, rabbit muscle actin subunit; BTM, calf brain tropomyosin. The arrows indicate bands that migrate with the four major polypeptides of rat brain myelin.

and membranes. This fraction also contains four major bands the standards to establish a strong similarity between the pairs that migrate with the four major proteins of rat brain myelin, of proteins. The standard maps also contain peptides that do including a band that migrates between tubulin and actin not appear in the maps of the synaptosomal proteins. These and corresponds to a molecular weight of about 50,000. A purer peptides may represent real differences between the pairs vesicle fraction prepared by lysis of a mitochondrial pellet in of proteins, but may result from small differences in the EDTA containing phosphate buffer contained little or no amount of protein applied to the chromatography plates.t myelin when examined in the electron microscope (Fig. lc). A small number of peptides appearing in the maps of synapto- Gels of this fraction contained reduced amounts of the three somal proteins do not appear in the maps of the standards. bands containing low molecular weight material, but the band These may be due to contaminating proteins that comigrate containing 50,000 molecular weight material was still a major on the gels or, again, may represent real differences in primary component (Fig. 2b). Neither preparation of vesicles contains sequence. Because of an insufficient amount of synaptosomal appreciable bands corresponding to actin and tropomyosin. material, the maps of the myosins are intended to serve as a In addition, the purer preparation of vesicles lacks the band preliminary comparison showing only a small number of the containing 200,000 molecular weight material corresponding expected peptides; however, several correspondences can be to the heavy chain of myosin (Fig. 2b). seen. The soluble fraction of lysed synaptosomes contains pri- and Biochemical Purity of Synaptosomal to actin and tubulin. Distribution marily material corresponding in size Proteins. Table 1 represents the results of an attempt to These proteins may originate from the synaptoplasm or they contractile and membrane quantitate the distribution of the proteins may be removed from the presynaptic during tubulin in the synaptosomes and synaptosomal subfractions by lysis. scanning the gels shown in Fig. 2 with a gel scanner and com- Peptide Maps of Synaptosomal and Standard Proteins. In puting the areas under the peaks of the recorder readout. order to demonstrate more rigorously that protein bands with The results must be considered approximate since the exact material corresponding in molecular weight to myosin, actin, l)ositioll of thie base line in the recorder trace was difficult to tropomyosin, and tubulin were indeed composed of these pro- determine, and since Coomassie blue does not stain all proteins, we removed these bands from gels of synaptosomes, teins linearly and with equal intensity. It is reported, however, prepared two-dimensional tryptic or chymotryptic peptides from these proteins, and compared them to corresponding t We have found that slight differences in the amount of a peptide maps of characterized proteins (Fig. 3). protein spotted on the plates can produce differences in the The peptide maps of bands corresponding to tubulin, number of peptides seen; even though we attempted to spot tropomyosin, and actin all contain a sufficient number of identical amounts of both unknown and standard in each case, peptides that correspond in position to peptides in the maps of small differences will certainly occur. Downloaded by guest on September 26, 2021 Proc. Nat. Acad. Sci. USA 71 (1974) Contractile Proteins in Synaptosomes 4475

* 0 0 * 0 .. 0@ 0 0 * @ * Q * 0 0: a * S *00

0 * * * .0 0 *0 * *D 0 0

8 D

* 0.0 * * * 00 0 * * 0 a 0 0 0 * 0 0 0 0 0 * *. 0 * 0 * * 0 *0 0.0. G~~~~~~ 0 E H FIG. 3. Two-dimensional peptide maps of synaptosomal proteins and standards. The proteins were recovered from NaDodSO4- acrylamide gels and digested with trypsin or chymotrypsin. Chromatography was performed on silica gel G thin-layer chromatography plates with n-propanol-concentrated ammonium hydroxide (7:3, v/v) followed by electrophoresis at pH 6.5, with 10% pyridine, 3% acetic acid as buffer. The plates were developed by spraying with 0.03% flourescamine in acetone. The fluorescent spots were viewed under UV light and marked for later viewing with 0.04% bromphenol blue. The positions of the bromphenol blue spots are displayed in the figure. (A) Tryptic peptides of tubulin precipitated from a whole brain supernatant with vinblastine sulfate. (B) Tryptic peptides of band having the same mobility as tubulin, recovered from acrylamide gels of synaptosomal protein. (C) Tryptic peptides of vinblastine-precipitated brain actin. (D) Tryptic peptides of synaptosomal band with the same mobility as actin, recovered from acrylamide gels. (E) Chymo- tryptic peptides of isolated chick brain tropomyosin. (F) Chymotryptic peptides of synaptosomal band with the same mobility as brain tropomyosin, recovered from acrylamide gels. (G) Chymotryptic peptides of calf brain myosin. (H) Chymotryptic peptides of band with the same mobility as myosin, recovered from acrylamide gels. Filled circles, peptides that appear in corresponding positions in the maps of both the standard proteins and the synaptosomal proteins; half-filled circles, peptides that appear only in the maps of the standards; open circles, peptides that appear only in the maps of the synaptosomal proteins.

that muscle actin and muscle myosin stain linearly and with except vesicles. Tropomyosin is present in significant amounts nearly equal intensity (24). Also, we have observed that mus- in synaptosomes and membranes. Myosin represents a small cle tropomyosin stains linearly when between 0 and 15 Ag of fraction of the total protein in all fractions. The 50,000 mo- protein were applied to a series of gels. Tubulin represents the lecular weight component is the major protein in the vesicle major component of synaptosomes, membranes, and soluble fraction, but is present in negligible amounts in synaptosomes fractions. Actin represents a major component of all fractions and is completely absent from the membrane and soluble fractions. TABLEU 1. Distribution of myosin, tubulin, actin, and To determine whether or not the subfractions were con- tropomyosin in synaptosomes and subfractions taminated with soluble brain proteins, radioiodinated actin, tropomyosin, and tubulin (18) were added to the brain Percent of total protein homogenate before isolation of the synaptosomes. Since the Tropo- synaptosome proteins were not found to be labeled, we con- Fraction Myosin Tubulin X* Actin myosin clude that they are intrinsic to the synaptosomes. Synaptosome 3.5 27 2 12 7 DISCUSSION Membrane 2.2 23 13 8 In this report biochemical evidence is presented that the Vesicle 4.3 19t 38 5 3 termini of rat cortical neurons contain relatively large amounts Soluble 3.9 17 17 of proteins very similar to actin and to brain tropomyosin in molecular weight and in primary sequence. Also, microtubule The NaDodSO4-acrylamide gels of proteins from synapto- protein is present as the polypeptide component found in somes and several synaptosomal subfractions were scanned with highest concentration, confirming the work of Feit et al. (6). a Gilford densitometer, and the areas under the peaks were A smaller amount of a myosin-like protein is also found. computed. The percentage of the total area under all peaks that occupied the peak corresponding to the indicated protein is These proteins are associated with the synaptic membrane shown. and, in the case of actin and tubulin, may also be present as * Unidentified protein of molecular weight about 50,000. soluble components of the synaptoplasm. Recently, morpho- t Migrates slightly less rapidly than tubulin standard. logical evidence has been presented for the presence of actin -, Not detected. filaments in presynaptic endings (25). Downloaded by guest on September 26, 2021 4476 Biochemistry: Blitz and Fine Proc. Nat. Acad. Sci. USA 71 (1974)

Isolated synaptic vesicles also appear to contain a myosin- nents of the presynaptic membrane, and/or soluble proteins like protein and microtubule protein, but it is not possible to from the nerve ending. determine from this work whether these components merely adhere to the outer surface of the vesicles during the isolation We thank L. Taylor and L. Mengel for expert technical as- sistance. This investigation was supported by National Institutes procedure or whether they are native to the vesicle structure. of Health Grant NS10582. Many of the polypeptide components of the vesicles are also found in whole synaptosomes and in synaptic membranes, but 1. Fine, JR. E. & Bray, D. (1971) Nature New Biol. 234, 115- there are important differences in some of the major com- 118. 2. Fine, R. E., Blitz, A. L., Hitchcock, S. L. & Raminer, B. ponents. The vesicles contain little or no actin, but contain (1973) Nature New Biol. 245, 182-185. instead an unidentified protein that migrates between tubulin 3. Puszkin, W., Berl, S., Puszkin, S. & Clarke, D. (1968) and actin on NaDodSO4-acrylamide gels. This protein mi- Science 161, 170-171. grates with one of the major polypeptides of rat brain myelin, 4. Borisy, G. G. & Taylor, E. W. (1967) J. Cell Biol. 34, 525-- micrographs 533. but electron and gels of the vesicle preparation a. Dutton, G. R. & Barondes, S. H. (1971) Science 166, 1637- contain no indication of a myelin contamination. This protein 1638. also corresponds closely in molecular weight to the subunit 6. Feit, H., Dutton, G., Barondes, S. & Shelanski, M. (1971) protein of neurofilaments (26). It has been suggested (27) that J. Cell Biol. 51, 138-147. synaptic vesicles have bound neurofilaments 7. Ebashi, S. & Endo, M. (1968) Progr. Biophys. Mol. Biol. subunits. Neuro- 18, 125-183. filaments have also been seen in presynaptic nerve termini 8. Katz, B. & Miledi, R. (1965) Proc. Roy. Soc. Ser. B 161, (25). N 496-503. Although proposals have been made for the involvement of 9. Berl, S., Puszkin, W. & Nicklas, J. (1973) Science 179, contractile proteins in neurotransmitter release (9), it is clear 441-446. 10. Gray, E. C. & Whittaker, V. P. (1962) J. Anat. 96, 79-88. that the muscle-like proteins are not unique to nerve endings, 11. Kurokawa, M., Sakamoto, T. & Kato, M. (1965) Biochim. but are found in a large number of mammalian and non- Biophys. Acta 94, 307-309. mammalian cells (28). The fact that they are membrane- 12. Gurd, J. W., Jones, L. Ra., Mahler, H. R. & Moore, W. J. associated proteins is also not a unique feature of nerve end- (1974) J. Neurochecm. 22, 281-290. ings. Actin, for example, is known to be associated with 13. Lapetina, E. G., Soto, E. F. & DeRobertis, E. (1967) Bio- the chim. Biophys. Acta 135, 33-43. plasma membranes of acanthamoeba (29) and of fibroblasts 14. Whittaker, V. P. & Sheridan, M. N. (1965) J. Neurochem. (30). Acrylamide gels of the membrane proteins of whole 12, 363-372. axons (31), L cells and BHK cells (32), and brain microsomes 15. Karnovsky, M. J. (1965) J. Cell. Biol. 27, 137A-138A. (12) contain major bands that migrate similarly to tubulin, ac- 16. Millonig, S. (1961) J. Appl. Phys. 32, 1637A. 17. Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244, 4406- tin, and tropomyosin. If a calcium-mediated contractile 4412. process is involved in neurotransmitter release, this may repre- 18. Bray, 1). & Brownlee, S. M. (1973) Anal. Biochem. 55, sent a specialized function for these proteins. 213-221. The recent findings that cytochalasin thought to inter- 19. Bohlen, P., Stein, S., D)airman, W. & Udenfriend, S. (1973) B, Arch. Biochim. Biophys. 155, 213-220. fere with nonmuscle actomyosin-mediated processes, prevents 20. Adelstein, R., Conti, M., Johnson, G., Pastan, I. & Pollard, the release of neurotransmitter from sympathetic neurons (33) T. (1972) Proc. Nat. Acad. Sci. USA 69, 3693-3697. and from synaptosomes (34) provides indirect evidence for 21. Norton, W. T. & Poduslo, S. E. (1973) J. Neurochem. 21, the involvement of the muscle-like proteins in this process. 749-757. There is also indirect 22. Whittaker, V. P., Essman, W. E. & )owe, G. H. C. (1972) evidence for the involvement of Biochem. J. 128, 833-843. tubulin in neurotransmitter release. Colchicine, which irrevers- 23. Adelstein, IR. S., Pollard, T. D. & Kuehl, W. M. (1971) ibly blocks microtubule assembly, also interferes with this Proc. Nat. Acad. Sci. USA 68, 2703-2707. process (33, 35). The additional facts that calcium ion pro- 24. Morimoto, K. & Harrington, W. F. (1974) J. AMol. Biol. 83, motes both the release of neurotransmitter (8) and reversibly 83-98. 25. Metuzals, J. & Mushynski, W. E. (1974) J. Cell Biol. 61, blocks the polymerization of microtubules (36), and the fact 701-722. that low temperature causes both the reversible depolymeriza- 26. Davison, P. & Winslow, B. (1974) J. Neurobiol. 5, 119-133. tion of microtubules (36) and the release of neurotransmitter 27. Kadota, K. & Kadota, T. (1973) J. Cell Biol. 58, 135-151. from synaptosomes (37), make it tempting to postulate a role 28. Pollard, T. D. & Weihung, JR. R. (1974) CRC Crit. Rev. for microtubules neurotransmitter Bioch. 2,1-65. in release. There is, how- 29. Pollard, T. D. & Korn, F. (1973) J. Biol. Chem. 248, 448- ever, insufficient morphological evidence to do so. The existence 455. of intact microtubules [with perhaps a single exception (6)] has 30. Perdue, J. F. (1973) J. Cell Biol. 58, 265-283. not been reported in electron micrographs of either synapto- 31. Grefrath, S. P. & Reynolds, J. A. (1973) J. Biol. Chem. 248, somes or of intact axon termini. The tubulin of synaptosomes 6091-6094. 32. Greenberg, C. S. & Glick, M. C. (1972) Biochemistry 11, exists either as free or as membrane-associated subunits, or 3680-3685. else it exists in some other mode of organization that does 33. Thao, N. B., Wooten, G. F., Axelrod, J. & Kopin, I. J. not survive fixation procedures. (1972) Proc. Nat. Acad. Sci. USA 69, 520-522. The elucidation of the role of these proteins, if any, in neuro- 34. Nicklaus, W. J. & Berl, S. (1974) Nature 247, 471-473. 35. Nicklaus, W. J., Puszkin, S. & Berl, S. (1973) J. Neurochem. transmitter release may well await the development of in vitro 20, 109-121. fractionated systems consisting of transmitter containing 36. Weisenberg, R. C. (1972) Science 177, 1104-1105. synaptic vesicles together with well-characterized compo- 37. Raiteri, M. & Levi, G. (1973) Nature 243, 180-181. Downloaded by guest on September 26, 2021