Evidence for Synaptotagmin As an Inhibitory Clamp on Synaptic Vesicle Release in Aplysia Neurons KELSEY C

Proc. Natl. Acad. Sci. USA Vol. 92, pp. 11307-11311, November 1995 Neurobiology

Evidence for synaptotagmin as an inhibitory clamp on synaptic vesicle release in Aplysia neurons KELSEY C. MARTIN*, YINGHE Hu*t, BETH A. ARMITAGE*, STEVEN A. SIEGELBAUM, ERIC R. KANDEL, AND BONG-KIUN KAANG* Howard Hughes Medical Institute and Center for Neurobiology and Behavior, College of Physicians and Surgeons of Columbia University, 722 West 168th Street, New York, NY 10032 Contributed by Eric R. Kandel, August 22 1995

ABSTRACT While previous studies have demonstrated effect of our manipulations on synaptic strength. Our results that synaptotagmin plays an essential role in evoked neuro- are consistent with synaptotagmin acting as an inhibitor of transmitter release, it has been difficult to determine whether neurotransmitter release.§ it acts to facilitate or inhibit release. To address this question, we used acute genetic manipulations to alter the expression-of synaptotagmin in Aplysia neurons. Transient overexpression MATERIALS AND METHODS ofsynaptotagmin in acutely dissected cholinergic neurons and Cloning. The coding region of synaptotagmin A from the in cultured glutaminergic neurons decreased the amplitude of marine ray Discopyge ommata was amplified by PCR (Perkin- the excitatory postsynaptic potential (EPSP) by 32% and 26%, Elmer/Cetus) and used to screen an Aplysia nervous system respectively. In contrast, treatment of cultured presynaptic cDNA library cloned in the A-ZAP vector (Stratagene) as neurons with synaptotagmin antisense oligonucleotides in- described (16). creased the amplitude of the EPSP by 50-75%. These results Antibodies. Antibodies were generated by Berkeley Anti- are consistent with a role of synaptotagmin as an inhibitor of body (Richmond, CA) against a maltose binding protein- release. synaptotagmin fusion protein (New England Biolabs) containing residues 1-67 fused to residues 93-426 of synaptotagmin. Neurotransmitter release is distinguished from other forms of The antiserum was purified using a fusion protein affinity exocytosis by its tight spatial and temporal regulation. Since column (Pierce). calcium influx is the signal for release, there must be one or Overexpression. The entire coding region ofAplysia synap- more calcium-sensing proteins present at the active zone, totagmin was amplified from the cDNA by PCR (Perkin- acting either to facilitate fusion in the presence of calcium or Elmer/Cetus) and ligated into the pNEX6 expression vector to inhibit fusion in its absence. Synaptotagmin has many of the (17) to create pNEX8-p65. Two complementary synthetic characteristics expected of such a calcium sensor. It contains oligonucleotides encoding the influenza hemagglutinin (HA) two sequences homologous tothe C2 domain of protein kinase epitope YPYDVPDYA were ligated into Sal I-digested C and has been found to bind calcium in a phospholipid- pNEX8-p65 to insert the tag between Arg-411 and Pro-412 of dependent manner in vitro (1-3). Synaptotagmin coprecipi- Aplysia synaptotagmin. Generation of pNEX&-lacZ has been tates with calcium channels (4) and with several plasma described (18). membrane proteins, including syntaxin (5) and the neurexins DNA was prepared for microinjection as described (17) and (6). Synaptotagmin is also a receptor for clathrin AP-2, injected by brief pressure pulses applied to microelectrodes suggesting that it may also be involved in the endocytosis of beveled to a resistance of 20 Mfl. In culturedAplysia neurons, synaptic vesicles (3, 7). overexpression of synaptotagmin was verified by indirect im- Several previous studies have addressed the function of munofluorescence using affinity-purified anti-synaptotagmin synaptotagmin in neurotransmitter release. Synthetic C2 do- antibodies and expression of lacZ was verified by enzymatic main peptides and antibodies against synaptotagmin inhibit staining for ,B-galactosidase as described (17). Immunofluo- release when microinjected into individual cells (8, 9). Cae- rescence was visualized on a Bio-Rad MRC1000 confocal norhabditis elegans, Drosophila, and mice bearing mutations in microscope mounted on a Zeiss Axiovert and images were synaptotagmin all had impaired synaptic function (10-13). taken using fixed settings. In the buccal ganglia, pNEXS-lacZ Biochemical studies of the interaction of synaptotagmin with was coinjected with pNEX&p65 and enzymatic staining by other synaptic membrane proteins suggest that it functions f3-galactosidase was used as an indication of synaptotagmin prior to the final fusion step (14, 15). The results of these overexpression. studies have led to three alternative views of the function of Antisense Oligonucleotides. Oligonucleotides were synthe- synaptotagmin: it docks vesicles at the plasma membrane, it sized by Oligos Etc. (Wilsonville, OR). The following se- promotes vesicle fusion, or it inhibits vesicle fusion. quences were used: antisense, 5'-GAG AAG CGG TCA AAG We have designed experiments using acute genetic manip- TC-3'; control (reverse order), 5'-CTG AAA CTG GCG ulations to increase or decrease synaptotagmin expression in AAG AG-3' (base pairs 718-734 after the initiating ATG). Aplysia californica neurons. First, we cloned the Aplysia ho- Oligos containing three thiol groups at the 3' and 5' ends molog of synaptotagmin and then overexpressed it in cholinergic neurons of acutely dissected Aplysia ganglia and in Abbreviations: EPSP, excitatory postsynaptic potential; HA, hemag- glutaminergic sensory neurons cocultured with a single motor glutinin. neuron. Second, we used antisense oligonucleotides to reduce *K.C.M., Y.H., and B.A.A. made equal contributions to this work. the expression of endogenous synaptotagmin in Aplysia sen- tPresent address: Miles, Inc., Institute for Metabolic Disorders, 400 sory-motor neuron cocultures. In both cases, we measured the Morgan Lane, Building 24, West Haven, CT 06516. *To whom reprint requests should be sent at the present address: Laboratory of Molecular Neurobiology, Institute for Molecular Bi- The publication costs of this article were defrayed in part by page charge ology and Genetics, Seoul National University, Seoul 151-742, Korea. payment. This article must therefore be hereby marked "advertisement" in §The sequence reported in this paper has been deposited in the accordance with 18 U.S.C. §1734 solely to indicate this fact. GenBank data base (accession no. U03125). 11307 Downloaded by guest on September 29, 2021 11308 Neurobiology: Martin et al. Proc. Natl. Acad. Sci. USA 92 (1995)

(end-capped) were found to be effective at a dose of 30 ,uM MDSLLARVKRAADADALNPAQEGVTGGPDAAGLPDVSTSSPGGGGAGDKL 50 without being toxic. For bath application of antisense oligo- KEKMDEYKDKLINEIENLPIWAIVLIIAGSLLFLVCCVYCVCRRSCRKRK 100 nucleotides, a lOX solution of oligonucleotides suspended in L15 medium (Sigma) was added to the culture medium KKEGKKGLKGAVDLKSVQLLGNSYKEKPDLDELPVNMEDNEDAESTKSEV 150 immediately prior to plating of the sensory neuron and was KLGKLQFSLDYDFQKGELSVNVIQAADLPGMDMSGTSDPYVKVYLLPDKK 200 replenished daily. In other experiments, oligonucleotides were KKYETKVHRKTLNPVFNESFTFKVPYAEVGSKILTFAVYDFDRFSKHDQI 250 injected into sensory cells 3-12 hr after plating by pressure GQVQVPLNSIDLGRVVEDWKDLQSPDTESEKENKLGDICFSLRYVPTAGK 300 application to microelectrodes containing 300 ,uM oligonucle- LTVVILEAKNLKKMDVGGLSDPYVKIALLQGTKRLKKKKTTIKKNTLNPY 350 otides in 0.5 M potassium acetate/10 mM Tris, pH 7.2. The final intracellular concentration of oligonucleotides was esti- FNESFGFEVPFEQIQKVTLIITVVDYDRIGTSEPIGRCVLGCNSSGTELR 400 mated to be 10-30 ,uM. HWSDMLANPRRPIAQWHTLQEVPEKN 426 Immunocytochemistry and Immunoblotting. CulturedAply- sia neurons were prepared for immunohistochemistry as de- FIG. 1. Amino acid sequence of Aplysia synaptotagmin. scribed (19). Antibody specificity was determined by immu- were identical. Only one gene was detected onAplysia genomic noblotting as described (19). In competition experiments, Southern blots probed at low and high stringency with the affinity-purified anti-synaptotagmin antibodies were preincu- Aplysia synaptotagmin cDNA, and only one transcript was bated with a 10-fold molar excess of maltose binding protein- detected on Northern blots (data not shown). These findings synaptotagmin fusion protein. Immunoblots of pedal ganglia suggest that only one synaptotagmin isoform exists in Aplysia. neurons microinjected with pNEX8-p65 HA were incubated However, in rat and mouse there is <40% sequence identity with mouse monoclonal anti-HA antibody 12CA5 (Berkeley among the multiple synaptotagmin isoforms (3). It remains Antibody, Richmond, CA) diluted 1:2500 followed by horse- possible, therefore, that other synaptotagmin isoforms exist in radish peroxidase-conjugated goat anti-mouse antibody Aplysia that were not detected. (Pierce) diluted 1:5000. Immunoblots were stripped according Polyclonal antibodies were generated in rabbits against a to the manufacturer's instructions (Amersham). synaptotagmin fusion protein. On Western blots of Aplysia Aplysia Culture, Dissection, and Electrophysiology. Sensory central nervous system, the affinity-purified antisera recog- cells (one or two) isolated from the pleural ganglia of adult nized a set of at least four proteins between 55 and 75 kDa as (80-100 g) animals were plated with L7 motor cells from the well as two smaller breakdown products, both of which in- abdominal ganglia of juvenile (1.5-4 g) animals as described creased in intensity in the absence of protease inhibitors (Fig. (20, 21). In overexpression studies, electrophysiological re- 2A). To test the specificity of this antibody, we performed cordings were as described (21, 22). In some antisense exper- immunoblots ofAplysia neurons overexpressing a HA epitope- iments, the sensory neuron was also impaled with a micro- tagged Aplysia synaptotagmin construct and compared stain- electrode to record the action potential triggered by a brief ing with our antibody to that of a monoclonal anti-HA injection of depolarizing current (5-10 msec, 2-6 nA). antibody. As shown in Fig. 2B, lane 1, the anti-HA antibody The buccal ganglia were dissected from 70- to 100-g animals at kDa. The same immu- and desheathed, and either cell B4 or B5 was randomly chosen recognized a very sharp band -65 to be injected with plasmid. Synaptic potentials were recorded noblot reprobed with our anti-synaptotagmin antibody pro- as described in Kaang et al. (17) and Gardner and Kandel (23). duced a broad set of bands around 55-75 kDa as well as two Because of an upward drift in the excitatory postsynaptic smaller breakdown products (lane 2). This strongly suggests potential (EPSP) amplitude with time, the effect of overex- that our antibody either recognizes several Aplysia synapto- pression was normalized by the change in the uninjected tagmin isoforms or cross-reacts with proteins sharing common presynaptic cell: % change = [(D2 x C1)/(D1 x C2) - 1] x epitopes, such as rabphilin-3A, which also contains C2 do- 100, where Dl = the initial EPSP amplitude of the injected mains (26). cell, D2 = the final EPSP amplitude of the injected cell, Cl = Immunohistochemical assays with anti-Aplysia synaptotag- the initial EPSP amplitude of the uninjected cell, and C2 = the min antibody showed intense staining in the neuropil in final EPSP amplitude of the uninjected cell. cryostat sections of dissected ganglia (data not shown) and The statistical significance of antisense treatment in senso- punctate staining in the neurites of cultured Aplysia neurons ry-motor cocultures and of overexpression in the buccal gan- (see Fig. SB). The staining pattern was indistinguishable from glia was tested using an unpaired Student's t test. The statistical that observed using an affinity-purified antibody against Aply- significance of overexpression in sensory-motor cocultures was sia synaptobrevin, another synaptic vesicle protein (data not tested using a one-way analysis of variance followed by a shown). This suggested to us that our anti-synaptotagmin Newman-Keuls range of means test. All results are as given A Aplysia CNS B Pleural ganglia means ± standard error. overexpressing p65HA 1 1 2 RESULTS kDa 2 Cloning of the Aplysia Synaptotagmin Homolog and Iden- 69-W - tification of the Endogenous Gene Product. We isolated ..... Aplysia synaptotagmin cDNA clones using low stringency 46- hybridization to a marine ray (D. ommata) synaptotagmin A-specific probe (24). TheAplysia synaptotagmin open reading frame encodes a polypeptide of 426 residues with a calculated molecular mass of 47,455 Da (Fig. 1). Aplysia synaptotagmin FIG. 2. Immunoblots ofAplysia synaptotagmin. (A)Aplysia central shares the putative domain structure of all other synaptotag- nervous system (CNS) probed with affinity-purified polyclonal anti- min isoforms: two cytoplasmic repeats of the C2 domain, a Aplysia synaptotagmin antibody (lane 1) or with polyclonal anti-Aplysia single transmembrane spanning region, and an intravesicular synaptotagmin antibody depleted of specific antibodies by preincuba- tion with maltose binding protein-Aplysia synaptotagmin fusion pro- amino terminus (3, 25).Aplysia synaptotagmin shares sequence tein (lane 2). (B) Pedal ganglia injected with plasmids encoding HA identity of 55.6% with synaptotagmin A from D. ommata and epitope-taggedAplysia synaptotagmin (pNEXS-p65 HA), probed with 56.1% with rat synaptotagmin 1. monoclonal anti-HA antibody 12CA5 (lane 1) and then stripped and From 300,000 clones screened with the synaptotagmin A reprobed with polyclonal anti-Aplysia synaptotagmin antibody (lane clone from D. ommata, the five positively hybridizing clones 2). Downloaded by guest on September 29, 2021 Neurobiology: Martin et al. Proc. Natl. Acad. Sci. USA 92 (1995) 11309 antibody most likely recognized multiple synaptotagmin iso- fluorescence using affinity-purified anti-synaptotagmin anti- forms or closely related synaptic proteins. body or for enzymatic staining for ,3-galactosidase. As shown Overexpression of Synaptotagmin cDNA in Presynaptic in Fig. 4 A and C, synaptotagmin and lacZ were expressed at Neurons Decreases Release. To analyze the function of syn- very high concentrations and were present in the synapse at the aptotagmin in vesicle release, we first transiently overex- time of fixation. Overexpression of full-length synaptotagmin pressed the cDNA by injecting pNEX8-p65 into identified in sensory cells depressed the initial EPSP by 26% [P < 0.01 cholinergic presynaptic neurons in the buccal ganglion. As a vs. injected but not overexpressing cells (impaled); P < 0.05 vs. control for injection and overexpression, pNEX6-lacZ was lacZ-expressing cells; Fig. 3D]. Overexpression of synaptotag- injected into some presynaptic cells. Baseline recordings were min had no effect on the degree of homosynaptic depression made 1-2 hr after plasmid injection, when the protein product (data not shown) nor did overexpression qualitatively affect of the injected plasmid was weakly expressed, if at all, and was the morphology of the presynaptic cell. present only in the perinuclear region. Recordings were re- Antisense Oligonucleotides Enhance Neurotransmitter Re- peated 8-12 hr later to allow time for further expression and lease. Through the application of antisense oligonucleotides, transport of the protein out to the synapse. The synaptic we attempted to reduce the amount of endogenous synapto- strength of control-injected and uninjected cells gradually tagmin in cultured cells. When Aplysia neurons are cultured, increased over the 8-12 hr between recordings, possibly the cell body and initial axon segment are pulled out of the triggered by the acute injury of dissection (27). Normalizing for ganglion, leaving the synaptic terminals behind. Antisense this upward drift (see Materials and Methods), we found a 32% oligonucleotides applied immediately might therefore block decrease in the postsynaptic response in cells that were injected synthesis of the synaptic proteins required for the new synapse with the synaptotagmin plasmid compared to those injected formed in vitro. with control plasmid or buffer alone (P < 0.05; Fig. 3C). We used a 17-bp antisense oligonucleotide complementary To rule out the possibility that the decreased EPSP resulted to a sequence that is conserved among the Aplysia homolog from an altered response to the acute injury of dissection, we and isoforms 1 and 2 of rat synaptotagmin. We also performed overexpressed synaptotagmin in cultured glutaminergic sen- a set of experiments using antisense oligonucleotides comple- sory neurons, in which the synaptic strength of control synapses mentary to a sequence spanning the initiation codon. Oligo- does not change significantly over a 24-hr period. EPSPs were nucleotides were either added to the culture medium or recorded from the motor neuron in response to a series of four injected into the presynaptic cell. action potentials evoked in the sensory neuron once every 20 We examined the effects of the oligonucleotides on synaptic sec. This provided a measure of baseline synaptic strength as transmission by recording from the synapse on day 4 or day 5. well as of the degree of homosynaptic depression (28, 29). The amplitude of the postsynaptic potential was 1.5-1.75 times Sensory cells were then microinjected with pNEX8-p65 or greater in the antisense treated synapses than in synapses pNEX6-lacZ. Fifteen to 20 hr later, the stimulus series was treated with control oligonucleotides (P < 0.01; Fig. 5) repeated and the cells were processed for indirect immuno- whether the antisense was applied to the culture medium or A Buccal Ganglia B Sensory-motor cultures 1st Recording 2nd Recording 1st Recording 2nd Recording FIG. 3. Transient overexpression of Aplysia synaptotagmin in dissected buccal ganglia. (A) ApSyt /f Impaled Control Examples of the EPSPs recorded in B7 after stimulating one of two presynaptic cells, B4 or B5, before (recording 1) and after (recording 2) expression of plasmids encoding Aplysia synap- ApSyt totagmin (ApSyt) in one of the presynaptic cells; Control control recording is from the uninjected presynaptic cell. (B) Examples of EPSPs recorded in J10 mv motor neuron L7 after stimulation of the sensory 100 msec lac Z neuron before (1st recording) or 15-20 hr after (2nd recording) injection of plasmids encoding C Buccal Ganglia synaptotagmin (ApSyt) or lacZ into the sensory cell. Impaled cells were injected with plasmid but 30 - D Sensory-motor cultures did not express the encoded protein. (C and D) Effects of transient overexpression of synapto- 15 - tagmin (ApSyt) or lacZ on synaptic transmission 20 in buccal ganglia (C) and sensory neurons (D). The height of each bar corresponds to the mean T + SEM in the EPSP 10 l a) percentage change ampli- cm cm 5- n=9 n=1 1 tude between the 1st and 2nd recordings (in C, n=11 .Ccu cu lac Z ApSyt the EPSP has been normalized as described in the C.) ApSyt A of variance followed 'IO o _ -t///v1 text). one-way analysis by Impaled a of Impalkled lac Z ~0 n=21 I comparison the mean (Newman-Keuls mul- n=18 n=7 tiple range test) indicates that overexpression of cn synaptotagmin in B4 or B5 (C) significantly de- . -10 -15 creased EPSP amplitude relative to impalement E C') (P < 0.01) and expression of lacZ (P < 0.01). In :01 contrast, of lacZ did not a. -20 'a expression significantly Cn alter EPSP amplitude relative to impalement. In w sensory-motor cultures (D), overexpression of -30 - -25 - synaptotagmin significantly decreased EPSP amplitude relative to impalement (P < 0.01) and to expression of lacZ (P < 0.05). In contrast, ex-

-40- pression of lacZ did not significantly alter EPSP -35 amplitude relative to impalement. Downloaded by guest on September 29, 2021 11310 Neurobiology: Martin et al. Proc. Natl. Acad. Sci. USA 92 (1995) DISCUSSION Does Synaptotagmin Facilitate or Inhibit Release? Our studies confirm previous work showing that synaptotagmin plays an important role in synaptic transmission. From this earlier work, however, it was difficult to determine whether synaptotagmin acts to facilitate or inhibit release. In three genetically modified animals, C. elegans, Drosophila, and mouse, loss of synaptotagmin largely abolished evoked neurotransmitter release (10-13). However, recordings from the Drosophila neuromuscular junction of third instar larvae of synaptotagmin hypomorphs revealed that while evoked release was diminished, spontaneous release was increased (10, 11, 30). These findings are consistent with synaptotagmin acting to inhibit vesicle fusion. In its complete absence, vesicles would constantly undergo spontaneous fusion, thus depleting the pool of vesicles available for synchronous release in response to the calcium signal (31). That synaptotagmin may inhibit release is further supported by the findings of Sollner et al. (15), showing that synaptotagmin forms a complex with synaptobrevin, syntaxin, and SNAP-25 and is displaced from this complex by a-SNAP prior to formation of a 20S complex containing the fusion factor NSF. In mice bearing mutations of synaptotagmin 1, however, spontaneous release was unaffected, the rapid phase of evoked release was eliminated, and there were no obvious alterations in active zone morphology (13). In contrast to the findings in Drosophila, the findings in mice are more consistent with synaptotagmin acting to promote vesicle fusion. The discrep- ancy between the two mutant animals may reflect differences between vertebrates and invertebrates. More likely, however, it may reflect differences in the physiological function of the particular isoform eliminated in each mutant animal. Synaptotagmin Acts as an Inhibitor of Neurotransmitter Release in Aplysia. We have attempted to increase and decrease synaptotagmin expression and to assess the effects of such changes on synaptic function. In the Aplysia culture FIG. 4. Detection of endogenous and overexpressed protein in system, changes in EPSP amplitude are likely to reflect cultured neurons. (A and B) Sensory neurons were injected with changes in transmitter release because there is neither the plasmids (pNEXS-65) encoding synaptotagmin. After the final record- complication of multiple synaptic inputs nor that of nervous ing, cells were fixed and stained with affinity-purified anti-Aplysia system development. In addition, because the cells are elec- synaptotagmin antibody. Both sensory neurons in A (arrows) overex- trically silent in the absence of extracellular stimulation, we can press synaptotagmin, whereas the sensory neuron in B (arrow) does interpret changes in the size of the initial EPSP as changes in not. Solid arrowheads indicate motor cell body and hollow arrowheads exocytosis as opposed to endocytosis of synaptic vesicles indicate motor neuron initial axon segment. The staining in B repre- following release. sents endogenous Aplysia synaptotagmin. (C) Sensory neurons were injected with plasmids encoding lacZ (pNEX&.lacZ). After the final When we overexpressed synaptotagmin, immunostaining recording, cells were fixed and stained enzymatically with ,3-galacto- indicated a large increase in protein concentration throughout sidase. Both sensory neurons (arrows) express lacZ. (A and B, bar = the cell. This resulted in a 26% decrease in EPSP amplitude in 12.5 ,um; C, bar = 50 ,um.) cultured cells and a 32% decrease in EPSP amplitude in dissected ganglia. There are several explanations for why such injected into the presynaptic cell. By contrast, there was no a large increase in protein expression should result in a significant difference in EPSP amplitude between cells in- relatively small change in release. (i) Synaptotagmin may act jected with control oligonucleotides and untreated cells. We with one or more cofactors that are present at limited con- also examined the effects of synaptotagmin antisense oligo- centrations. (ii) Newly expressed synaptotagmin may not be nucleotides on two forms of short-term synaptic plasticity, efficiently incorporated into vesicles residing at the active depression and serotonin-mediated facilitation zone. (iii) The small size of the vesicles may permit only a homosynaptic limited number of proteins to be present on their surface and of the depressed synapse (dishabituation) (28, 29), neither of this maximum may normally be reached. If any of these which was altered (data not shown). explanations apply, one would predict that decreasing the Using our anti-Aplysia synaptotagmin antibody, we were not concentration of synaptotagmin would produce a much greater able to detect any consistent effect of antisense treatment on effect. protein levels. The apparent cross-reactivity of our antibody Indeed, when we used antisense oligonucleotides to block with other synaptotagmin isoforms or antigenically related synaptotagmin synthesis during synapse formation, we saw a proteins may have obscured any change in the protein con- 50-75% increase in EPSP amplitude. This occurred despite centration of the targeted isoform. However, the effect of the the fact that we were unable to determine whether antisense antisense oligonucleotides was specific: control oligonucleo- treatment reduced synaptotagmin concentration since our tides did not affect initial EPSP strength, and antisense antibody appeared to cross-react with other synaptotagmin oligonucleotides increased initial EPSP strength without af- isoforms or proteins with common epitopes. It is possible that fecting homosynaptic depression or heterosynaptic dishabitu- even a small decrease in synaptotagmin expression caused a ation. large change in release. This could occur if synaptotagmin Downloaded by guest on September 29, 2021 Neurobiology: Martin et al. Proc. Natl. Acad. Sci. USA 92 (1995) 11311

A Control B Antisense C 2.25 El Control 2.00 E2 Antisense 1.75 10 3mV| 1.50 Postsyn 1.25 a. 1.00

a. 0.75 0.50 0.25 Presynh 15 mvI 15 mVI CO n-46 n-49 n-24 n-35 100 msec 100 msec IN BATH INJECTED

FIG. 5. Treatment of cultured sensory-motor neurons with synaptotagmin antisense oligonucleotides increases transmitter release. (A and B) Examples of the EPSP and action potential in sensory-motor neuron cultures receiving control oligonucleotides (A) or antisense oligonucleotides (B) in the culture medium. (C) Averaged data. The initial EPSP amplitude of antisense treated cells was averaged and normalized to the average initial EPSP amplitude of cells treated with control oligonucleotides. Oligonucleotides were added to the culture medium (in bath) or microinjected into the sensory cell (injected). Comparison of the means (unpaired Student's t test) indicates that antisense delivered in the culture medium or by microinjection significantly increased transmitter release (P < 0.01).

acted to inhibit release and if there were normally more than 8. Bommert, K., Charlton, M. P., DeBello, W. M., Chin, G. J., Betz, one copy per vesicle. Only when calcium bound and inactivated H. & Augustine, G. J. (1993) Nature (London) 363, 163-165. all copies would fusion proceed. Then, if antisense treatment 9. Elferink, L. A., Peterson, M. R. & Scheller, R. H. (1993) Cell 72, decreased the copy number, the probability of all copies being 153-159. inactivated would increase nonlinearly, leading to an increase 10. DiAntonio, A., Parfitt, K. D. & Schwarz, T. L. (1993) Cell 73, in the total 1281-1290. number of vesicles released. 11. Littleton, J. T., Stern, M., Schulze, K., Perin, M. & Bellen, H. J. Synaptotagmin may inhibit synaptic vesicle release in a (1993) Cell 74, 1125-1134. number of ways, all of which are consistent with our data. (i) 12. Nonet, M. L., Grundahl, K., Meyer, B. J. & Rand, J. B. (1993) It may prevent docking of synaptic vesicles at the plasma Cell 73, 1291-1305. membrdne. (ii) It may act as a physical barrier to fusion of 13. Geppert, M., Goda, Y., Hammer, R. E., Li, C., Rosahl, T. W., docked, fusion-competent vesicles (31). (iii) Synaptotagmin Stevens, C. G. & Sudhof, T. C. (1994) Cell 79, 717-729. may inhibit release by preventing the formation of a fusogenic 14. Sollner, T., Whiteheart, S. W., Brunner, M., Erdjument, B. H., complex. Thus, Sollner et al. (15) have proposed that docked Geromanos, S., Tempst, P. & Rothman, J. E. (1993a) Nature synaptic vesicles contain a complex of synaptotagmin, synap- (London) 362, 318-324. tobrevin, syntaxin, and SNAP-25 and that synaptotagmin must 15. Sollner, T., Bennett, M. K., Whiteheart, S. W., Scheller, R. H. & be replaced by a-SNAP prior to fusion of the vesicle with the Rothman, J. E. (1993b) Cell 75, 409-418. plasma membrane. It may be possible to test this particular 16. Yamasaki, S., Hu, Y., Binz, T., Kalkuhl, A., Kurazono, H., model in our system by overexpressing a-SNAP and synapto- Tamura, T., Jahn, R., Kandel, E. R. & Niemann, H. (1994) Proc. Natl. Acad. Sci. USA 91, 4688-4692. tagmin in the same presynaptic cell. 17. Kaang, B.-K., Pfaffinger, P. J., Grant, S. G. N., Kandel, E. R. & Furokawa, Y. (1992) Proc. Natl. Acad. Sci. USA 89, 1133-1137. We thank Richard Scheller for the D. ommata synaptotagmin A 18. Kaang, B.-K., Kandel, E. R. & Grant, S. G. N. (1993b) Neuron 10, clone; Oligos Etc. for technical support; Huan Yao, Rene Baston, and Mingming Ning for excellent technical assistance; Mark Barad and 427-435. Paul Skehel for critical reading of the manuscript; Charles Lam and 19. Skehel, P. A., Martin, K. C., Kandel, E. R. & Bartsch, D. (1995) Eve for of the and Harriet Ayers and Andy Science, in press. Vagg preparation figures; 20. Rayport, S. G. & Schacher, S. (1986) J. Neurosci. 6, 759-763. Krawetz for preparation of the manuscript. This research was funded 21. Montarolo, J. G., Goelet, P., Castellucci, V. F., Morgan, J., by the Howard Hughes Medical Institute and a National Institute of Mental Health Center Grant, P50 MH50733-02. Kandel, E. R. & Schacer, S. (1986) Science 234, 1249-1254. 22. Dale, N. & Kandel, E. R. (1990) J. Physiol. (London) 421, 1. Perin, M. S., Fried, V. A., Mignery, G. A., Jahn, R. & Sudhof, 203-222. T. C. (1990) Nature (London) 345, 260-263. 23. Gardner, D. & Kandel, E. R. (1971) Science 173, 550-553. 2. Brose, N., Petrenko, A. G., Sudhof, T. C. & Jahn, R. (1992) 24. Wendland, B., Miller, K. G., Schilling, J. & Scheller, R. H. (1991) Science 256, 1021-1025. Neuron 6, 993-1007. 3. Li, C., Ullrich, B., Zhang, J. Z., Anderson, R. G. W., Brose, N. & 25. Perin, M. S., Brose, N., Jahn, R. & Sudhof, T. C. (1991) J. Biol. Sudhof, T. C. (1995) Nature (London) 375, 594-599. Chem. 266, 623-629. 4. Leveque, C., Hoshino, T., David, P., Shoji, K. Y., Leys, K., Omori, 26. Shirataki, M., Kaibuchi, K., Sakoda, T., Kishida, S., Yamaguchi, A., Lang, R., El Far, O., Sato, K., Martin-Moutot, N., Newsom- T., Wada, K., Miyazaki, M. & Takai, Y. (1993) Mol. Cell. Biol. 3, Davis, J., Takahashi, M. & Seagar, M. J. (1992) Proc. Natl. Acad. 2061-2068. Sci. USA 89, 3625-3629. 27. Walters, E. T., Alizadeh, H. & Castro, G. A. (1991) Science 253, 5. Bennett, M. K., Calakos, N. & Scheller, R. H. (1992) Science 257, 797-799. 255-259. 28. Gingrich, K. J. & Byrne, J. H. (1985) J. Neurophys. 53, 652-669. 6. Petrenko, A. G., Perin, M. S., Davletov, B. A., Ushkaryov, Y. A., 29. Hochner, B., Klein, M., Schacher, S. & Kandel, E. R. (1986) Proc. Geppert, M. & Sudhof, T. C. (1991) Nature (London) 353, 65-68. Natl. Acad. Sci. USA 83, 8794-8798. 7. Zhang, J. Z., Davletov, B. A., Sudhof, T. C. & Anderson, R. G. 30. DiAntonio, A. & Schwarz, T. L. (1994) Neuron 12, 909-920. (1994) Cell 78, 751-760. 31. Popov, S. V. & Poo, M. M. (1993) Cell 73, 1247-1249. Downloaded by guest on September 29, 2021