Regulation of Vesicle Docking, Priming, and Fusion by Synaptotagmin. Zhao Wang. A dissertation submitted in partial fulfillment of

Transcription

1 Regulation of Vesicle Docking, Priming, and Fusion by Synaptotagmin By Zhao Wang A dissertation submitted in partial fulfillment of the requirements for the degree Doctor of Philosophy (Cellular and Molecular Biology) at the UNIVERSITY OF WISCONSIN-MADISON 2012 Date of final oral examination: 8/22/12 The dissertation is approved by the following members of the Final Oral Committee: Edwin R. Chapman, Professor, Neuroscience Meyer B. Jackson, Professor, Neuroscience William M. Bement, Professor, Zoology James C. Weisshaar, Professor, Chemistry Guy E. Groblewski, Associate Professor, Nutritional Sciences

2 i Acknowledgements I like to thank my thesis advisor Dr. Edwin Chapman. It has been a great honor and pleasure to work with him for five years. Dr. Chapman has been a role model for me to learn how to be a diligent, productive, and intelligent scientist. His passion and dedication towards science deeply inspired me. I also want to thank my thesis committee members: Dr. Meyer Jackson, Dr. James Weisshaar, Dr. William Bement, and Dr Guy Groblewski. They gave me a lot of great advices and help on my thesis projects. I also want to thank CMB program coordinator Michelle Holland, for giving me a lot of kind help during my first year in Madison. I would like to thank all the lab members of Chapman lab Dr. Min dong, Dr. Camin Dean, Dr. Jon Gaffaney, Dr. Huisheng Liu, Dr. Mark Dunning, Dr. Colin Johnson, Dr. Jun Yao, Dr. Mike Chicka, Dr. Enfu Hui, Dr. Yao Wu, Dr. Felix Yeh, Dr. Sam Kwon, Dr. Shihu Sun, Hua Bai, Lindsey Roper, Chantell Evans, Annette Figueroa-Bernier, Ewa Bomba, Yiwen Gu, Jon Rehfuss, Deepshika Ramanan, Jennifer Levenson, Monica Strathman, and Reid Alderson. I want to thank Dr. Jon Gaffaney, a brilliant Biochemist, guided me into Dr. Chapman s lab and showed me how to do all kinds of Biochemistry assays. I also want to thank Enfu Hui, Felix Yeh and Hua Bai for discussing scientific questions with me, which help me acquire deeper understanding of many scientific questions. I would like to thank Annette Figueroa-Bernier and Ewa Bomba as well for their help on the confocal imaging. I want to thank my parents, who unconditionally support me. Especially, my mom always encourages me to be a scientist.

3 ii Finally, I would like to thank my dear wife and daughter. I would not be able to accomplish my thesis work without your selfless support. My wife gives me a lovely home, where I can always find peace and relaxation. I also want to thank my one-year old daughter, who brings endless joy into my life.

4 iii TABLE OF CONTENTS Abstract...viii Chapter I Background and Significance 1.1) Introduction ) Synaptic Vesicle Exoctyosis.3 An Overview of Synaptic Vesicle Recycle 3 Vesicle Docking. 4 Vesicle Priming..7 Vesicle Fusion and fusion pore ) In Vitro Functional Study of Synpatotagmin Unique Prospective of Synaptotagmin Studies of Synaptotagmin 4 s Localization and Function ) Studies on In Vitro Fusion Assay..13 Studies of Using C2AB in Fusion Assay.13

5 iv An Advanced split SNARE assay Early Studies of Using Full Length Reconstituted Synaptotagmin..15 The role of PIP 2 in exocytosis ) Figures Chapter II Rat and Drosophila Synaptotagmin 4 Have Opposite Effect during SNARE Catalyzed Membrane Fusion 2.1) Summary ) Introduction ) Results.36 Opposite effects of rat and fly syt 4 in reconstituted membrane fusion reactions..36 Rat syt 1 and syt 4 chimeras fail to stimulate fusion..38 Functional comparison of the isolated C2-domains from rat and fly syt Reversal of the S244 mutation endows rat syt 4 with the ability to function as a Ca 2+ sensor for fusion ) Discussion ) Experimental procedures..47

6 v 2.6) Figures ) Appendix.75 Chapter III Reconstituted Synaptotagmin I Mediates Vesicle Docking, Priming, and Fusion. 3.1) Summary ) Introduction ) Results.80 Effect of PIP 2 on Ca 2+ syt regulated fusion..80 Specificity of the phosphatidylinositol bisphosphate requirement for regulated fusion. 81 Optimal syt density and the Ca 2+ sensitivity of fusion 82 Topological requirements for PS, PIP 2, and syt..83 Mutational analysis of syt during regulated fusion.84 Context-dependent mixed antagonist/agonist activity of C2AB...85 Systematic comparison of fusion reactions regulated by FL or the cytoplasmic domain of syt.87 Distinct mechanism of FL syt and C2AB mediated fusion ) Discussion.. 89

7 vi 3.5) Experimental procedures ) Figures ) Appendix Chapter IV Defining the roles of reconstituted FL syt1 in three discrete steps in the secretory pathway 4.1) Summary ) Introduction ) Preliminary Data, Experimental Procedures, Proposed experiments ) Discussion ) Figures ) Appendix Chapter V Conclusions and Future Perspectives 5.1) Conclusions ) Future Perspectives.149 References 150

8 vii Abbreviations list SV, synaptic vesicle; LDCV, large dense core vesicle; SNARE, soluble N-ethyl maleimide-sensitive factor attachment protein receptor; NMJ, neuromuscular junction; BDNF, brain derived neurotrophic growth factor; Syt 1, synapatotagmin 1; C2AB domain, cytosolic domain of syt; C2A domain, membrane proximal C2-domain of syt; C2B domain, C-terminal membrane distal C2-domain of syt; NBD-PE, 1,2-Dipalmitoyl-sn-Glycero-3-Phospho-ethanolamine-N-(7-nitro-2-1,3-benzoxadiazol- 4-yl); Rhodamine-PE, 1,2-Dipalmitoyl-sn-Glycero-3-Phosphoethanolamine-N-(Lissamine Rhodamine B Sulfonyl); PS, 1,2-Dioleoyl-sn-Glycero-3-[Phospho-L-Serine]; PC, 1-Palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphocholine; PE, 1-Palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphoethanolamine; Tr, t-snare vesicle; Vr, v-snare vesicle; PIP 2, phosphatidylinositol 4,5-bisphosphate; FL syt, full-length synapatotagmin I

9 viii Abstract Ca 2+ dependent exocytosis is responsible for neuronal communications. During this process, the synaptic vesicle undergoes several distinct steps, including docking, priming, and fusion, to release its cargo. The basic machinery that mediates membrane fusion has been proved to be a set of SNARE proteins. Synaptotagmin I is thought serve as the Ca 2+ sensor that regulates neuronal exocytosis. Besides synaptotagmin 1, there are 16 other isoforms in verterbrate have been identified. The functions of most of these proteins are unknown. The major goal of my thesis is to characterize the biological function and molecular mechanism of action of the synaptotagmin (syt) family of proteins. In the second chapter, I worked on an interesting syt isoform, syt 4. This protein has been implicated in regulating synaptic transmission, but with apparently conflicting results in different organisms. Using a direct reconstitution approach, I discovered that vertebrate (murine) syt 4 serves as an inhibitory isoform that prohibit exocytosis, whereas invertebrate (fly) syt 4 acts in a positive manner to couple Ca2+ to membrane fusion, fully resolving the controversies regarding the function of this protein in different species. In the next chapter, I focus on the Ca2+ sensor of exocytosis, syanptotagmin I. The progress in understanding of the function(s) of this protein has been slow, due to the lack of a functional in vitro assay, in which syt1 regulates membrane fusion in a manner analogous to its function in vivo. By including a critical lipid, PIP2, in the system, and allowing for a novel priming step, I successfully established an in vitro fusion assay that more accurately recapitulates syt 1 function in vivo, thus providing a powerful tool to study the mechanism of action of syt 1. Finally, to determine whether synaptotagmin enhances membrane fusion by directly regulating SNARE complex formation, I have developed a fluorescent reporter system (based on transition metal FRET) that can precisely monitor sub-domain structural changes of SNARE proteins. In a

10 ix separate study, I utilized different diameters of lipids-coated glass beads to explore the role of membrane curvature, and bending, on the kinetics of membrane fusion.

11 1 Chapter I Background and Significance

12 2 1.1) Introduction Important brain functions, such as thinking and memory, depend on neuron communication. Neurons communicate with each other by synaptic transmission. The process of synaptic vesicle release has been elucidated in numerous studies(chapman, 2008). During synaptic transmission, an electrical pulse along the axon arrives in the end of synaptic terminal, which triggers the opening of Ca 2+ channels and Ca 2+ influx. The increase of Ca 2+ initiates synaptic vesicle exocytosis (Fig. 1-1). In order for the neurons to sense the change of calcium signals during synaptic transmission, a calcium sensor on synaptic vesicle is required, and was later identified as a 65 kda protein, synaptotagmin 1 (syt 1) (Matthew et al., 1981; Perin et al., 1990). The fast synchronized phase of exocytosis, which is introduced by Ca 2+ influx is abolished when syt 1 is knocked-out from neurons (Maximov and Sudhof, 2005; Nishiki and Augustine, 2004b). Although the mechanism of how syt 1 regulates exocytosis is not clear, it has been shown that syt 1 binds to anionic lipids(brose et al., 1992) and SNARE (soluble N-ethyl maleimide-sensitive factor attachment protein receptor) proteins in a Ca 2+ dependent manner(bhalla et al., 2005a; Mahal et al., 2002). It has also been suggested that syt 1 is able to promote membrane fusion by bending the membrane, and thus lowering the energy requirement of lipid mixing step (Hui et al., 2009). Besides syt 1, which functions as the major Ca 2+ sensor, 16 other synaptotagmin isoforms have been identified (Mittelsteadt et al., 2009). The mrna of most of isoforms were identified in brain. The overall structure of synaptotagmin family members is highly conserved. It contains a luminal domain, a transmembrane domain, and double C2 domains. Despite the similarity of their structures, the localization and function of these isoforms are different. Some isoforms (syt 1, 2, 9, and 12) (Sudhof, 2002; Takamori et al., 2006a) are suggested to localize to synaptic vesicles, others (syt 3,

13 3 6) are indicated majorly localize to the plasma membrane(butz et al., 1999), while the majority localization still remains unclear. For instance, syt 7 has been suggested to localized predominately to the plasma membrane (Sugita et al., 2001) and large dense core vesicles (Wang et al., 2005). 1.2) Synaptic Vesicle Exocytosis An Overview of Synaptic Vesicle Cycle The synaptic vesicle cycle is localized in prsynaptic terminal (Ceccarelli et al., 1973; Heuser and Reese, 1973). The SV cycle includes several steps, including exocytosis, endocytosis, and vesicle reacidification and refilling (Fig. 1-2). Synaptic vesicles undergo exocytosis upon arrival of an action potential. After fusion, endocytosis is followed, in which membrane is retrieved from plasma membrane and new vesicles. Those vesicles are then acidified by proton pumps and reloaded with neurotransmitters. Vesicle recycling can go through two pathways, so called fast and slow pathway. In fast pathway, vesicles do kiss and run during membrane fusion, and are then reused for next round of fusion. During the slow pathway, vesicles are not resused right away. They have to go through endosomal intermediates before being reloaded with neurotransmitters. There are extensive studies that address synaptic vesicle exocytosis. Although the molecular mechanism is still not completely clear, the basic steps have been elucidated, including vesicle docking, priming, and fusion(chapman, 2008).

14 4 Vesicle Docking Before membrane fusion, a synaptic vesicle is targeted to the plasma membrane, in a process is referred to vesicle docking. Docking normally defined as synaptic vesicle is in a very close proximity to the plasma membrane (less than 30 nm). Although its morphology is well defined, its molecular mechanism remains unknown. Many proteins have been shown to be critical for vesicle docking. 1) Sec 1/ Munc 18 Munc 18 has been shown to be required for Ca 2+ dependent exocytosis, although its exact function is still unclear. It binds to and changes structural confirmation of Syntaxin to a ready for fusion state. Knock-out of Munc 18 in mouse chomaffin cells reduced the amount of docked vesicles ten fold, without affecting the total vesicle population or altering the kinetics of single vesicle fusion event (Voets et al., 2001b). Overexpression of Munc 18 in C. elegans enhanced vesicle docking, while deletion of Munc 18 dramatically impaired vesicle docking (Weimer et al., 2003). However, the effect Munc 18 has on vesicle docking is unclear. One possibility, is that Munc 18 changes the structure and membrane distribution of Syntaxin. Overexpression of Munc 18 mutant that does not bind to Syntaxin, reduced Munc 18 s plasma membrane association in PC 12 cells (Schütz et al., 2005). In yeast, overexpression of Syntaxin (Sso1p or Sso2p) could rescue the defect of exocytosis from deletion of Munc 18 (Sec1 p) (Aalto et al., 1993) and similar result was observed in C. elegans, wher a constitutively active form of Syntaxin fully rescued the docking defects of Munc 18 mutants (Hammarlund et al., 2007).

15 5 However, it should be noted that some studies also suggess Munc 18 does not affect Syntaxin cellular function. In C. elegans, knock-out of Munc18 did not affect Syntaxin s conformational change, localization, or expression level (Weimer et al., 2003). 2) Syntaxin The role of Syntaxin in docking is controversial. Early studies indicate docking is not altered by deletion of Syntaxin. Injection of botulinium toxin C, which reduces the level of Syntaxin, into squid giant presynaptic terminal significantly decreased exocytosis, without altering the docking of vesicle (O'Connor et al., 1997). In Drosophila, deletion of Syntaxin did not change vesicle targeting to the plasma membrane, despite the complete loss of synaptic transmission (Broadie et al., 1995). However, more recent studies hold a different view and suggest that Syntaxin is absolutely required for vesicle docking. Deletion of Syntaxin significantly affected docking of dense core vesicle in chromaffin cells, although this defect was not observed at synapse (de Wit et al., 2006). Another study also indicates that Syntaxin directly mediates docking. Overexpression of a constitutively open form of Syntaxin fully rescued the docking defects of Munc 18 mutants (unc-13) in C elegans (Hammarlund et al., 2007). The discrepancies between the recent and earlier studies may be due to the advancement in experimental techniques. The data in the recent studies was generated by using high pressure freezing electron microscopy, which has fewer chances to generate artifacts from sample fixation, compared to traditional EM in earlier studies. 3) Synaptobrevin (VAMP)

16 6 Early studies indicate that synaptobrevin is not involved in vesicle docking. Injection of tetanus, which can cleave synaptobrevin, in squad giant synapses, did not reduce the number of docked vesicles by using electron microscopy, despite the blocks of neuronal secretion (Hunt et al., 1994). Similar results were also observed when synaptobrevin was deleted in Drosophila strains. Vesicles were still targeted to the presynaptic memberane and docked normally at the release sites (Broadie et al., 1995). These results are further confirmed in mouse chromaffin cells (Borisovska et al., 2005). Double knock-out of synaptobrevin and cellubrevin did not change the number, size or spatial distribution of chromaffin granules (Gerber et al., 2008). Similar to Syntaxin, all the early studies on Synaptobrevin were based on traditional electron microscopy, which might introduce artifacts in cellular structure morphology during fixation and sample preparation, and subsequently alter the vesicle distribution. A new study using total internal reflection fluorescence microsopy (TIRFM) suggests a different role of synaptobrevin in vesicle docking (Wu, 2012). In PC 12 cells, cleavage of synaptobrevein with botulinum neurotoxins strongly reduced vesicle docking. Mutagenesis studies indicate the docking function is dependent on the membrane distal region of the SNARE motif. This result was further confirmed by using high pressure freezing electron microscopy. 4) Syt 1 The Ca 2+ sensor, syt 1, has also been suggested to be critical for vesicle docking. In Drosophila, knock-out of syt 1 significantly decreased the number of total vesicles as well as docked vesicles. Introduction of a syt mutant, which restored the number of total vesicles in nerve terminals, did not rescue the defect on vesicle docking, indicating the synaptotagmin play a critical role in synaptic vesicle docking (Reist et al., 1998). In a different study, the docking of large dense core

17 7 vesicles was also impaired when using chromaffin cells from syt 1 knock-out mice (de Wit et al., 2009). A synaptotagmin mutant, in which the positive lysine patch is mutated, displayed a slow recovery from synaptic depression, suggesting the membrane trafficking step before fusion is disrupted (Loewen et al., 2006). FM dye uptake experiments indicate endocytosis was not altered. Thus, synaptotagmin must be involved in steps between endocytosis and membrane fusion, which are the docking and priming. Vesicle Priming After a vesicle is docked to the plasma membrane, there is an energy preparation step before the final fusion step, called priming. This step is thought to involve a rearrangement of proteins and lipids, and it was shown to require ATP. It renders vesicles the competence of fusing with plasma membrane upon Ca 2+ influx. The priming step was originally identified in permeabilized chromaffin cells (Holz et al., 1989). The Ca 2+ dependent secretion was sensitive to the level of MgATP due to the fact that pre-incubation with MgATP significantly enhanced exocytosis. The priming step was further shown to be reversible and could be inhibited by protein kinase inhibitors in PC 12 cells (Hay and Martin, 1992). This process was later on found to be dependent on PIP 2 levels in the plasma membrane. In chomaffin cells, changing PIP 2 concentration by using bacterial phospholipase (PLC) or removal of ATP, alters priming activities (Eberhard et al., 1990). In PC 12 cells, PIP 2 antibody and PLC also inhibit Ca 2+ regulated secretion by reducing priming activity (Hay et al., 1995). Besides PIP 2, N-ethyl maleimide- sensitive Factor (NSF) was also suggested to be required for the priming step. The NSF, as an ATPase, was proposed to catalyze the activation of SNARE proteins in the absence of Ca 2+, and make vesicles competent for fusion.

18 8 Although priming was shown to be essential for Ca 2+ triggered exocytosis, its definition is still vague and under debate. It is arguable whether it happens before or after the docking step. It is even regarded to happen simultaneously with docking, and therefore it can not be separated from the docking step. It also should be noted that majority of studies on vesicle priming are from research that is based on large dense core vesicle. Whether it is essential for synaptic vesicle exocytosis is still uncertain. Future studies will shed more light on this open question. Vesicle Fusion and fusion pore The final step of exocytosis is membrane fusion. Membrane fusion is energy cost step. Energy is required during membrane fusion for removal of water and dehydration of polar phospholipid headgroups. SNAREs have been suggested to be the basic machinery for membrane fusion (Jackson and Chapman, 2006). The formation of SNARE complex supplies sufficient energy for the following lipid mixing step and overcome the energy barrier of membrane fusion. SNARE was first proposed to be critical for membrane fusion by Rothman group (Söllner et al., 1993). Since the NSF and SNAP protein were known to be crucial for membrane traffic, Rothman et al. used an affinity purification procedure to identify the target of these proteins and found a trimeric protein complex, which is formed by SNAREs. Because numerous SNAREs are localized to different cellular compartments, it is proposed that SNARE interactions ensure the vesicle to target specificity. Studies using neurotoxins are consistent with this hypothesis. Cleavage of each individual SNARE by neurotoxins significantly blocks neurotransmitter release (Blasi et al., 1993a; Blasi et al., 1993b; Schiavo et al., 1992). During the fusion of vesicles with plasma membrane, a transient luminal channel is formed, named fusion pore, which allows cargos inside of vesicles to diffuse out into the cell exterior (Fig.

19 9 1-3). Upon forming, fusion pore could go through two different pathways. It could be open and closed transiently and vesicles are reloaded with neurotransmitters, and reused for the next round of membrane fusion. This process is named as kiss and run event. Another possibility is that fusion pore is widely dilated and vesicles completely merge with the plasma membrane, which is named as full fusion. A fusion pore can be studied with three methods, including capacitance measurement by patch clamping, amperometry, and fluorescent imaging. Membrane capacitance is proportional to the cell somatic membrane area (Gentet et al., 2000). Thus, the increases of membrane area caused by vesicle fusion results an increase of membrane capacitance. The changes of capacitance reflect sizes of vesicles. Kiss and run events were first identified through capacitance measurements on mast cells (Fernandez et al., 1984). The capacitance recording shows a similar pattern of upwards and downwards step distribution. This observation suggests that the reversals of fusion pore during membrane fusion. This conclusion was also confirmed by capacitance measurement on small dense core vesicles that have comparable size to synaptic vesicles in nerve terminals (Klyachko and Jackson, 2002). It was also found that during the opening of fusion pore, a transient current is generated and can be recorded by measuring the conductance change of the cell (Breckenridge and Almers, 1987). The conductance of fusion pores is comparable to that of gap junction channel, indicating the earliest formation of a fusion pores has a similar structure to an ion channel. The variations of conductance recording from different cell types indicate fusion pore has diverse structures and molecular compositions (Lindau and Almers, 1995). Amperometry can be also used to study fusion pores. Vesicle fusion is monitored electrochemically with a carbon-fiber microelectrode placed adjacent to the cell. The release of oxidized cargo from vesicles results in individual spikes to estimate the kinetics of fusion pore. It

20 10 was first utilized to estimate catecholamine secretion from chormaffin cells and observed the quantal vesicle secretion (Wightman et al., 1991). The technique was then used to identify a prespike foot (PSF) at the onset of quantal release, which represents the slow leak of catecholamine molecules out of a narrow 'fusion pore' before the pore dilates for complete exocytosis (Chow et al., 1992). The amplitude of PSF is correlated to the size of the fusion pore and the duration of PSF indicates the stability of fusion pore, which could be used for analyzing the fusion pore kinetics. In line with the capacitance recordings, the amperometry study, using PC 12 cells, also indicates that fusion sizes are in great variation (Alvarez de Toledo et al., 1993). Another method to study fusion pore is fluorescent imaging. The lipophilic fluorescent dye, FM1-43, and its variants are loaded into secretion vesicles to monitor their recycling kinetics. It was firstly used to study the kinetics of synaptic vesicle recycling at presynaptic bouton (Ryan et al., 1993). The studies, using FM1-43, also suggest synaptic vesicle fusion consist of two types of events: kiss and run and full fusion. Rat hippocampal neurons were loaded with FM1-43 and secretion was triggered by single action potentials. In only a small number of cases, complete loss of the fluorescence signal was observed, which represents full fusion events. While in most cases, only partial loss of fluorescence was trigged by action potentials, indicating vesicular retention of membrane marker, consistent with kiss and run events (Aravanis et al., 2003). The fusion pore in spontaneous release and stimulated exocytosis has been shown to be different. Studies on lactotroph, a secreting cell of the anterior pituitary, showed that stimulated hormone release is ten times faster than spontaneous secretion when using FM dye imaging (Stenovec et al., 2004). The slow hormone release is due to the smaller size of the fusion pore in spontaneous secretion (Vardjan et al., 2007) by measuring the release rates of different sizes of cargo or capacitance measurement.

21 11 Several proteins have been proposed to regulate the kinetics and size of fusion pores. Different synaptotagmin isoforms have been indicated to alter fusion pores. In PC 12 cells, overexpression of synaptotagmin 1 increased the time duration of the PSF of dense core vesicles in amperometry measurements, suggesting that synaptotagmin stabilizes the fusion pore (Wang et al., 2001). On the other hand, synaptotagmin 4 plays an opposite role in regulating fusion role. PSF was shortened by overexpression of synaptotagmin 4 in PC 12 cells. Synaptotagmin 4 can also induce a unique form of exocytosis, which is mainly involved kiss and run events (Wang et al., 2003a). This type of fusion pore opens 100 times longer than the normal PSF duration. Similar results were also observed in either Drosophila or pancreatic cells (Pawlu et al., 2004) (Tsuboi and Rutter, 2003). SNARE proteins are also indicated to be crucial for fusion pores. A mutation of the transmembrane domain of Syntaxin, to alter the size of side chains of amino acid, changed the rate of vesicle release (Han et al., 2004). 1.3) In vivo functional study of Synpatotagmin 4 (syt 4) Unique Perspectives of Syt 4 As an early immediate gene, syt 4 has drawn a lot attention. It is essential for brain function and has been shown to be required for motor neuron coordination and hippocampal-dependent learning and memory (Ferguson et al., 2000a). It was suggested to be related to psychiatric diseases, schizophrenia, and bipolar disorders (Ferguson et al., 2000b). It was originally identified in mouse brain and PC 12 cells (Hilbush and Morgan, 1994). Later on, it was found in Astrocytes and is required for glutamate secretion (Zhang et al., 2004) Studies of Syt 4 s Localization and Function

22 12 The studies about the localization and function of syt 4 have been controversial. Early works on syt 4 indicated that it was localized to synaptic vesicles (Osborne et al., 1999) (Ferguson et al., 1999) (Littleton et al., 1999a) and PC 12 cells (Hilbush and Morgan, 1994) (Wang et al., 2001) (Zhang et al., 2010) (Machado et al., 2004). It was predicted to co-regulate secretion of synaptic vesicles with synaptotagmin 1. However, recent studies exclude the possibility that syt 4 is localized to synaptic vesicles of presynaptic membrane. Syt 4 was shown to localize to large dense core vesicles on the postsynaptic membrane (Adolfsen et al., 2004; Dean et al., 2009). The functional studies of syt 4 on synaptic transmission draw different conclusions. First, several groups have reported syt 4 plays a negative role in regulating Ca 2+ triggered exocytosis. Studies in mice posterior pituitary nerve terminals indicate that syt 4 reduced the size of the fusion pore through the use of aperometry (Zhang et al., 2009). In mouse neurons, syt 4 was shown to reduce synaptic vesicles exocytosis by altering the release of postsynaptic BDNF vesicles. Studies on PC 12 cells also indicate syt 4 reduces the secretion of large dense core vesicle by using different approaches. A study, utilizing aperometry, found that overexpression of syt 4 could significantly reduced PC 12 cell s secretion by shortening the time from fusion pore opening to dilation (Wang et al., 2001). Similar results were acquired when using patch clamp to measure the capacitance change of the membrane (Wang et al., 2003a) (Zhang et al., 2010). Overexpression of syt 4 decreased the secretion of large dense core vesicles by reducing the size of vesicles and increasing the frequency and duration of kiss and run events. Another study used the human growth hormone as a reporter system and found that overexpression of syt 4 reduced depolarization induced secretion (Machado et al., 2004). Studies on Drosophila draw a completely different conclusion. The majority of studies indicate syt 4 plays a positive role. For example, syt 4 was shown to rescue exocytosis at neuromuscular

23 13 junctions in a syt 1 knockout background, despite the fact that syt 4 was later shown to not be on synaptic vesicles (Robinson et al., 2002a). In another study, syt 4 was shown to enhance presynaptic synaptic vesicle release by affecting postsynaptic retrograde signaling by activating the cyclic adenosine monophosphate (camp) dependent protein kinase pathway (Yoshihara et al., 2005) (Barber et al., 2009). A following up study indicated both C2 domains are required for syt 4 s ability to enhance exocytosis (Barber et al., 2009). From the discussion above, it seems that fly syt 4 plays a positive role in secretion, while rat syt 4 inhibits Ca 2+ triggered exocytosis. However, it should be noted that not all the studies agrees with these conclusions. For example, a study at the Drosophila neuromuscular junction indicated that synaptic neurotransmitter release was inhibited by up-regulation of syt 4 (Littleton et al., 1999a). Moreover, there are studies on vertebrate cells indicating syt 4 plays positive role. Syt 4 was suggested to be essential for Ca 2+ triggered NGF (nerve growth factor) release in PC 12 cells (Mori et al., 2008), glutamate secretion from astrocytes (Zhang et al., 2004), and exocytosis at auditory ribbon synapses (Johnson et al., 2010). 1.4) membrane fusion Studies by using In Vitro Fusion Assay To monitor fusion in a simplified system, a lipids mixing assay, named in vitro fusion assay, was initially utilized to demonstrate membrane fusion is mediated by SNARE proteins (Weber et al., 1998). This assay was used to study the fusion of vesicles to cultured fibroblasts (Struck et al., 1981). v-snare and t-snare proteins were first reconstituted into artificial liposomes, respectively. A pair of lipid fluorescent probes (NBD-PE/ Rhodamine-PE) were included in the v- SNARE imbedded liposomes during the reconstitution step. When membrane fusion occurs, the density of the lipid quencher is reduced and thus the distance between donor and receptor

24 14 fluorophores is increased. The membrane fusion then can be estimated by measuring the dequeching of the donor fluorophore signal. Studies of Using C2AB in Fusion Assay Due to the lack of ability to stimulate membrane fusion of full length syanptotagmin proteins, the cytoplamic domain of the protein (C2AB) was utilized in the in vitro fusion assay to address the fusion regulation of synaptotagmin (Tucker et al., 2004). In the presence of Ca 2+, C2AB strongly increase the rate and extent of membrane fusion, thus confirming C2AB as the Ca 2+ sensor of exocytosis in a functional in vitro assay (Fig. 1-4). Furthermore, this study also suggested synaptotagmin s function is largely dependent on its ability to bind membranes and SNARE proteins by varying the lipids composition of the vesicles, and by using a SNAP25 truncation mutant that mimics cleavage of SNAP-25 by botulinum neurotoxin A. This C2AB based fusion assay has been utilized in a few follow up studies. In one study, it was used to demonstrate the specificity of divalent cation coupling of synaptotagmin isoforms on membrane fusion regulation (Bhalla et al., 2005b). Syt 7 and syt 9 are more sensitive to Ca 2+ than syt 1 in regulating exocytosis. Furthemore, both of these syt isoforms stimulate fusion in response to Sr 2+, while syt 1 fails to activate fusion in the presence of Sr 2+. The significance of t-snare interaction of synaptotagmin was also studied by using this assay. Firstly, t-snare interaction was shown to be essential for synaptotagmin to stimulate membrane fusion (Bhalla et al., 2006a) through the use of yeast SNARE proteins. The vertebrate syt 1 failed to render robust Ca 2+ dependent fusion on yeast SNARE mediated membrane fusion. The SNARE binding was then purposed to be critical for vesicle docking by holding SNARE zippering in a partially assembly way (Chicka et al., 2008b). This system was further improved by including

25 15 phosphatidylethanolamine in the SNARE bearing vesicles, which significantly improved the rate and efficiency of membrane fusion (Gaffaney et al., 2008). An Advanced Split SNARE Assay A recent study based on the above mentioned in vitro fusion assay correlated fusion efficiency to aggregation activity, suggesting that C2AB s function is actually strictly dependent on its ability to aggregate vesicles in this assay (Hui et al., 2011). A few of other vesicles aggregators (ie. CPLA2 or Biotin/ Avidin) were also examined using the same assay and displayed robust stimulation of membrane fusion. This raised the possibility that synaptotagmin stimulates membrane fusion by merely aggregating vesicles. Syt 1 has been shown to drive the assembly of SNAP 25 onto Syntaxin to form a stable t-snare heterodimer. To study synaptotagmin s ability to assembly SNAREs, this fusion assay was further modified. Instead of using vesicles that bear t-snare heterodimer (syntaxin/ SNAP 25) in the assay, vesicles bearing syntaxin alone were used, and soluble SNAP25 were directly added in the buffer. Among all the aggregator, examined in this modified assay, only syt 1 was able to drive robust fusion in response to Ca 2+, while other aggregators (ie. CPLA2 or avidin/biotin) failed to stimulate fusion, indicating aggregation alone is not sufficient to stimulate fusion (Fig. 1-5). The SNARE assembly ability of synaptotagmin is absolutely required for its ability to stimulate membrane fusion. Early Studies of Using Reconstituted Full Length Syt Protein Several groups are trying to address syt s ability to regulate membrane fusion by using membrane embedded full length proteins, while none of these systems are physiologically relevant. In one study, full length synaptotagmin was able to stimulate fusion when it was reconstituted into v- SNARE vesicles. However, addition of Ca 2+ failed to increase this response (Mahal et al., 2002). In another study, full length syt was able to stimulate fusion in response to Ca 2+ only when PS

26 16 was omitted from v-snare vesicles (Stein et al., 2007). However, PS has been shown to be present on synaptic vesicles (Takamori et al., 2006b). Surprisingly when PS was included on v- SNARE vesicles, syt- Ca 2+ reduced membrane fusion. There are two studies showing that membrane anchored syt 1 stimulated fusion in response to Ca 2+. However, one study shows syt only stimulates fusion in a very narrow range of Ca 2+ around 10 µm (Lee et al., 2010). Stimulation was not observed when Ca 2+ raised beyond 25 µm, whereas robust Ca 2+ triggered exocytosis is detected in neuronal and neuroendocrine cells at these concentrations (Llinas et al., 1992, 1995). In another study, the minimal concentration that Ca 2+ stimulates fusion has to be above 2 mm, which is well above the physiological relevant [Ca 2+ ] during exocytosis (Kyoung et al., 2011). The Role of PIP 2 in Exocytosis PIP 2 has been indicated to be involving in many cellular functions. It is localized to plasma membrane and was shown to be crucial for Ca 2+ triggered exocytosis (Fig. 1-6). The majority of these works are based on the study of large dense core vesicle exocytosis. Eberhard et al. firstly showed that by reducing PIP 2 levels, using bacterial phospholipase C or removal of ATP, the Ca 2+ triggered secretion in chromaffin cells were largely reduced, indicating PIP 2 is required for exocytosis (Eberhard et al., 1990). PIP 2 was then suggested to be required for ATP-dependent vesicles priming step in PC 12 cell (Hay et al., 1995). The Ca 2+ triggered exocytosis in PC 12 cell was inhibited, when PIP 2 was blocked or reduced by either mouse antibody or recombinant PLC through reducing the priming activity. This priming effect of PIP 2 was further confirmed by transient expression of pleckstrin homology domain (PH domain) in chromaffin cells, which specifically binds to PIP 2. The block of exocytosis by expression of the PH domain is confirmed by measuring the secretion of growth hormone and membrane capacitance changes. The study of

27 17 PIP5K, an enzyme that is responsible for the synthesis of PIP 2, also confirmed the importance of PIP 2 in exocytosis (Aikawa and Martin, 2003). The expression of an ARF6 mutant inhibited exocytosis by reducing the activity of PIP5K. There are also studies that suggest PIP 2 is crucial for synaptic vesicle exocytosis. PIPKlγ is critical for PIP 2 systhesis. PIPKIγ knock-out mice displayed defects in synaptic transmission which involve several synaptic vesicle cycles, such as the reduction of frequency of miniature current and the ready releasable pool, the enhancement of synaptic depression, and slower recycling kinetics (Di Paolo et al., 2004). The molecular mechanism of PIP 2 regulating vesicle exocytosis is still not clear. However, several mechanisms have been indicated. First, PIP 2 was indicated to form microdoamins on the membrane(laux et al., 2000). This microdomain formation is regulated by three proteins, GAP43, MARCKS, and CAP23. These proteins have a common effector binding (ED) domain, which is indicated to be responsible for binding and regulating the distribution of PIP 2 on the plasma membrane(laux et al., 2000). In PC12 cells, the t-snare protein, syxntaxin, was shown to be clustered into this PIP 2 microdomain. Increase in the formation of PIP 2 microdomain, by overexpression of phosphatidylinositol-4-phosphate 5-kinase, enhanced Ca 2+ regulated exocytosis, indicating these microdomains are crucial for vesicle secretion (Aoyagi et al., 2005). In another study, PIP 2 microdomains were shown to be related to the size of the vesicle pool(milosevic et al., 2005). In chromaffin cells, regulation of the microdomain formation by changing PIP 2 levels in the cell membrane alters the number of vesicles ready for fusion. PIP 2 was also proposed to regulate exocytosis by interacting with proteins that are important for exocytosis, including synaptotagmin 1 (Schiavo et al., 1996) (Tucker et al., 2003) (Bai et al., 2004), calcium-activated protein for secretion (CAPS/ Munc 13) (Loyet et al., 1998) (Grishanin et al., 2004) and rabphilin (Montaville et al., 2008). Synaptotagmin 1 was suggested to interact with

28 18 PIP 2 through its C2B domain (Schiavo et al., 1996) by liposome binding assay. This was further studied in cracked PC 12 cells, where soluble syt fragment protein, C2B domain, effectively inhibited Ca 2+ regulated exocytosis (Tucker et al., 2003). Furthermore, PIP 2 is able to steer syt to penetrate the plasma membrane (Bai et al., 2004). The steering ability is dependent on a lysine patch on the side of the C2B domain of synaptotagmin. Mutations in this region severely affect the binding and steering ability of synaptotagmin to PIP 2 containing membranes (Bai et al., 2004).

29 19 1.5) Figures Fig. 1-1

30 20 Fig. 1-1 views of freeze-substituted nerve terminals. High magnification images of nerve terminals are taken in frog neuromuscular junctions at rest (upper panel) and stimulated (lower panel) states. Synaptic vesicle fusion causes many pockets on the plasma membrane of the stimulated nerve terminal. This figure is adapted from (Heuser et al. 1981)

31 Fig

32 22 Fig. 1-2 synaptic vesicle recycles SV vesicles are acidified by protein pumps, which results the uptake of neurotransmitters. The loaded vesicles are then targeted to the active zones, and undergo priming steps that render them the competence of membrane fusion. The final fusion steps are triggered by the influx of Ca 2+ ion that is initiated by an action potential. The SV vesicles then are retrieved by endocytosis and reloaded with neurotransmitters for the next round of fusion. This figure is adapted from (Chapman ER. 2008)

33 Fig

34 24 Fig. 1-3 two fusion pore models. Left panel shows the proteinaceous fusion pores, which assemble like a gap junction channel. Fusion pore opening and dilating involves the structural rearrangement of SNARE protein. The lipids fusion pores model is shown in the right panels. The states of fusion pore are indicated as: (i) before forming; (ii) fusion pore forming; (iii) fusion pore opening; (iv) fusion pore dilating. This figure is adapted from (Jackson et al. 2008)

35 Fig

36 26 Fig. 1-4 Reconstitution of Ca 2+ -dependent membrane fusion by using C2AB. C2AB stimulates SNARE-mediated fusion between v-snare vesicles and t-snare vesicles in the presence of 1mM Ca 2+ but not in the presence of 0.2 mm EGTA. The NBD fluorescence was normalized by using the maximum fluorescence obtained after addition of the detergent. This figure is adapted from (Tucker et al. 2004)

37 Fig

38 28 Fig. 1-5 models of split SNARE assay (a) Syx vesicles, v-snare vesicle and soluble SNAP25 were incubated with different membrane aggregators. C2AB efficiently stimulated fusion in presence of Ca2+, while other aggregators failed. (b) Schematics of split SNARE assay. C2AB stimulates fusion by aggregating vesicles and driving the assembling of t-snare heterodimers. This figure is adapted from (Hui et al. 2011)

39 Fig

40 30 Fig. 1-6 Localization of Phosphatidylinositol 4,5 P2 in chromaffin cell PH-GFP was expressed in chromaffin cells. It specifically binds to PIP2 and is used to estimate the cellular localization of PIP 2. Wildtype PH-GFP proteins were specifically localized to plasma membrane (left panel). A mutant of PH-GFP, which does not bind to PIP2, was transfected into chromaffin cells (right panel). This mutant protein failed to localize to plasma membrane. This figure is adapted from (Holz et al, 2000)

41 31 Chapter II Rat and Drosophila Synaptotagmin 4 Have Opposite Effect during SNARE Catalyzed Membrane Fusion

42 32 2.1) Summary Synaptotagmins (syt) are a large family of proteins that regulate membrane traffic in neurons and other cell types. One isoform that has received considerable attention is syt 4, with apparently contradictory reports concerning the function of this isoform in fruit flies and mice. Syt 4 was reported to function as a negative regulator of neurotrophin secretion in mouse neurons and as a positive regulator of secretion of a yet-to-be identified growth factor from muscle cells in flies. Here, we have directly compared the biochemical and functional properties of rat and fly syt 4. We report that rat syt 4 inhibited SNARE (soluble N-ethyl maleimide-sensitive factor attachment protein receptor) catalyzed membrane fusion in both the absence and presence of Ca 2+. In marked contrast, fly syt 4 stimulated SNARE-mediated membrane fusion in response to Ca 2+. Analysis of chimeric molecules, isolated C2-domains, and point mutants, revealed that the C2B domain of the fly protein senses Ca 2+ and is sufficient to stimulate fusion. Rat syt 4 was able to stimulate fusion, in response to Ca 2+, when the conserved D-to-S Ca 2+ -ligand substitution in its C2A domain was reversed. In summary, rat syt 4 serves as an inhibitory isoform while fly syt 4 is a bona fide Ca 2+ sensor capable of coupling Ca 2+ to membrane fusion.

43 33 2.2) Introduction Neurons harbor two distinct major classes of secretory vesicles: small synaptic vesicles (SV) that mediate rapid synaptic transmission, and large dense core vesicles (LDCV) that secrete neuropeptide growth factors as well as other hormones (Chicka and Chapman, 2009). Significant progress has been made in the identification and function of positive and negative regulators of SV exocytosis (Lin and Scheller, 2000) but, at present, less is known concerning the regulation of LDCV exocytosis. Early studies indicated that a member of the synaptotagmin family of proteins, syt 4 (Takahashi et al., 2002), was a component of SVs (Wiedemann et al., 1998), while other reports ruled-out the presence of syt 4 on SVs (Diao et al., 2010; van den Bogaart et al., 2010). More recent studies, carried out using fruit flies and mouse neurons, suggest that syt 4 might play a direct role in the regulation of secretion from LDCVs in both organisms (Basu et al., 2005; Shen et al., 2007; Wang et al., 2011). More specifically, syt 4 was shown to be expressed in muscle cells at the Drosophila neuromuscular junction (NMJ) (Shen et al., 2007). Loss of post-synaptic syt 4 resulted in defects in pre-synaptic nerve terminal growth and plasticity and it was proposed that syt 4 functions as a positive regulator of an unidentified retrograde messenger (Shen et al., 2007). A different view arose from similar studies carried out using cultured mouse neurons (Basu et al., 2005). In this case, syt 4 was shown to co-localize with brain-derived neurotrophic factor (BDNF) in LDCVs that are targeted to both pre- and post-synaptic compartments. Knock-out and over-expression experiments revealed that syt 4 serves as a negative regulator of BDNF secretion at both loci. syt 4 is rapidly upregulated in response to activity (Karatekin et al., 2010; Wu, 2012; Zimmerberg et al., 2006) and it was proposed that up-regulation of syt 4 serves to down regulate

44 34 synaptic function via inhibition of BDNF (Basu et al., 2005). Indeed, syt 4 serves to limit the magnitude of LTP via its ability to inhibit the release of BDNF (Basu et al., 2005). Collectively, it appears that the positive role played by syt 4 at the fly NMJ, and the negative role of syt 4 in the release of BDNF from mouse neurons, are contradictory findings. The focus of the current study is to use direct approaches to resolve this apparent controversy. Before delving further it is useful to review a few key findings regarding the syt family. All syt isoforms have similar overall structures: a short luminal domain, a single transmembrane domain, and a cytoplasmic domain comprised of tandem C2-domains called C2A and C2B (Stein et al., 2009). In syt 1, each C2-domain binds 2-3 Ca 2+ ions via 4-5 acidic residues in two flexible loops (Shao et al., 1998; Sutton et al., 1995; Ubach et al., 1998); upon binding Ca 2+ these loops partially insert into membranes that harbor acidic phospholipids (Bai et al., 2004; Chapman and Davis, 1998). The insertion step drives localized bending of the target membrane to facilitate fusion (Hui et al., 2009; Sahoo et al., 2000). Syt has also been shown to mediate vesicle aggregation, and this could juxtapose SNAREs to facilitate fusion (Domanska et al., 2009; Popoli and Mengano, 1988). In addition, syt 1 binds to target membrane SNARE proteins in a Ca 2+ promoted manner (Chapman et al., 1995; Schiavo et al., 1997); it is thought that the Ca 2+ - independent component of this interaction serves to clamp or inhibit the fusion complex (Chicka et al., 2008a). Then, in response to Ca 2+, syt 1 can influence the folding and assembly of SNAREs to facilitate fusion (Bhalla et al., 2006a). In contrast to syt 1, SYT4 harbors a D-to-S substitution of one of the Ca 2+ ligands in its C2A domain (Takahashi et al., 2002). Interestingly, fly syt 4 harbors the same D-to-S mutation within its C2A domain, but, in contrast to the rat protein, fly syt 4 appears to retain some degree of Ca 2+, and Ca 2+ -dependent effector, binding activity (Dai et al., 2004; Littleton et al., 1999b). This observation is consistent with the notion that fly syt 4 might serve to facilitate SNARE-mediated

45 35 membrane fusion reactions while rat syt 4 inhibits these reactions. Here, we directly test this idea by assaying both proteins in a reconstituted membrane fusion system. Our results confirmed that rat syt 4 is a negative regulator of SNARE catalyzed membrane fusion, and also directly revealed - for the first time - that fly 4 serves as a positive regulator of fusion. Hence, despite their ~50% sequence identity and the conserved substitution of a Ca 2+ ligand in their C2A domains, mouse and fly syt 4 are not functional orthologs of one-another. Further experiments provided insights into the molecular basis for these functional differences.

46 36 2.3) Results Opposite effects of rat and fly syt 4 in reconstituted membrane fusion reactions - Rat and fly syt 4 exhibit ~50% sequence identity within their cytoplasmic domains and, as their names indicate, are thought to be encoded by orthologous genes. Compared to syt 1, they both bear a D- to-s substitution of one of the acidic Ca 2+ ligands within the C2A domain. However, recent studies suggest that rat and fly syt 4 have distinct physiological functions (Barber et al., 2009; Basu et al., 2005; Shen et al., 2007; Wang et al., 2011). These differences in function may be due to distinct biochemical properties of these two proteins. Here, we address this question by carrying out direct comparisons of rat and fly syt 4 in a well characterized reconstituted membrane fusion assay (Chicka et al., 2008a; Gaffaney et al., 2008; Tucker et al., 2004). SNAREs, from either rat or fly, were reconstituted into PS /PC /PE (15/ 55/ 30%) liposomes. v- SNARE vesicles bear a FRET donor-acceptor pair (NBD [7-nitro-2-1,3-benzoxadiazol-4-yl] /Rhodamine) attached to the head-group of PE. Fusion of v-snare-bearing vesicles with unlabeled t-snare vesicles serves to dilute the donor-acceptor pair, resulting in dequenching of the NBD signal, which is monitored using a plate reader (Fig. 2-1A). Rat and fly syt 4 were pre-incubated with SNARE-bearing vesicles for 20 min. Ca 2+ was then added to yield a final free concentration of 1 mm and fusion was monitored for another 120 min. Rat and fly syt 4 exhibited marked functional differences in this fusion assay (Fig. 2-1B, C). Rat syt 4 inhibited fusion in the absence and presence of Ca 2+, whereas fly syt 4 promoted robust fusion in response to Ca 2+ ; fusion was not observed using protein free v- or t-snare vesicles (data not shown). As a positive control we included rat and fly syt 1 in parallel fusion assays. As expected, rat syt 1 gave rise to efficient Ca 2+ -promoted fusion activity. Interestingly, similar results were obtained using the fly ortholog of syt 1. Since syts operate, in part, by engaging

47 37 SNARE proteins (Bai et al., 2004; Bhalla et al., 2006a; Chicka et al., 2008a; Gaffaney et al., 2008; Zhang et al., 2002), these findings further suggest that the determinants that underlie syt-snare interactions are conserved across species. These results also help to validate the use of vertebrate SNAREs to study invertebrate syts. However, in order to ensure these results are not specific for rat SNAREs, we repeated these experiments with fly SNARE-bearing vesicles and similar results were obtained. Rat syt 4 inhibited fly SNARE mediated fusion whereas fly syt 4 stimulated fly SNARE mediated fusion (Fig. S2-1); both proteins were able to clamp fusion in EGTA. As alluded to above, syts have been proposed to regulate fusion by interacting with PS and t- SNAREs (Jackson, 2009). We therefore compared the PS- and t-snare binding activities of fly and rat syt 1 and 4 to determine whether these effector interactions were indeed correlated with Ca 2+ -regulated fusion activity. PS binding activity was monitored using a co-sedimentation assay in which syt was incubated with increasing amounts of PS bearing liposomes. Liposomes were then pelleted via centrifugation. To avoid the loss of material in the pellet due to washing steps, we monitored the depletion of syt from the supernatant, and used this information to calculate the amount of bound material. Using this approach we found that rat syt 4 failed to bind to PS in either the absence or presence of Ca 2+. In contrast, fly syt 4, as well as rat and fly syt 1, bound to PS in a Ca 2+ -dependent manner (Fig. 2-2A). These data are consistent with previous reports indicating that fly, but not rat syt 4, binds to PS in response to Ca 2+ (Chapman and Davis, 1998; Dai et al., 2004). Next, we examined the t-snare binding activities of rat and fly syt 4 using a co-flotation assay in which t-snares are reconstituted into liposomes that lack PS. Soluble syt fragments were incubated with the liposomes and centrifuged through an Accudenz step gradient. Syt that was bound to vesicles, via interactions with t-snares, co-floated with the proteoliposomes to the top of the 0-30% Accudenz interface; vesicles were collected from this interface and

48 38 subjected to SDS-PAGE and stained with Coomassie blue. Rat syt 4 strongly bound to t-snare heterodimers in the absence Ca 2+ ; binding was not influenced by the addition of Ca 2+. In contrast, fly syt 4, as well as rat and fly syt 1, bound to t-snares weakly in the absence of Ca 2+ ; addition of Ca 2+ resulted in a significant increase in t-snare binding activity (Fig. 2-2B). Together with the PS-binding data detailed above, we propose that rat syt 4 is unable to stimulate fusion because it fails to interact with either PS or t-snare heterodimers in response to Ca 2+. Rat syt 1 and syt 4 chimeras fail to stimulate fusion - Both rat and fly syt 4 bear the same aspartate to serine substitution in one of the Ca 2+ ligands in C2A (Fig. 2-4A), yet one protein is an active Ca 2+ -sensor and the other is not. To gain greater insight into this critical difference, we made use of chimeric proteins that harbored tethered C2-domains from different syt isoforms. First, we note that in the case of rat syt 1, the isolated C2B domain alone is able to stimulate fusion in the reconstituted fusion assay (Gaffaney et al., 2008), and thus differences in the C2B domain between rat and fly syt 4 could potentially contribute to the observed differences in our fusion assays. However, it is also possible that an inactive C2A domain, in rat syt 4, might be able to down-regulate the activity of its adjacent C2B domain (e.g. as shown in Fig. S2-2, the D230S mutation in the C2A-domain of the cytoplasmic domain of rat syt 1 significantly reduced the ability of adjacent C2B-domain to stimulate fusion). To begin to discern between these possibilities, we built two chimeras between rat syt 1 and rat syt 4: the C2A domain from syt 1 was fused with the C2B domain from rat syt 4 (designated rat1a 4B). This chimera was used to address the question of whether the C2A domain of syt 1 can activate the C2B domain of syt 4 (Bai et al., 2002). The second chimera, in which the C2A domain from syt 4 was fused with the C2B domain from syt 1 (designated rat4a 1B), was used to determine if an active C2B domain can still regulate fusion when tethered to a dead C2A domain (Fig. 2-3A).

49 39 We tested these two chimeras in the fusion assay and found that neither was able to stimulate fusion to any significant degree, although there is a minor increase in fusion when using relatively high concentrations (1-2 µm) of rat1a 4B (Fig. 2-3B). These experiments suggest that both C2- domains from rat syt 4 are defective in terms of stimulating membrane fusion in response to Ca 2+. Functional comparison of the isolated C2-domains from rat and fly syt 4 - As shown above, rat and fly syt 4 have distinct effects in membrane fusion reactions. To further elucidate the molecular basis for this difference, we purified the isolated C2-domains from each protein and analyzed them in the fusion assay. Interestingly, the isolated C2B from fly syt 4 was able to stimulate fusion in response to Ca 2+. In contrast, rat syt 4 C2B, as well as rat and fly syt 4 C2A, failed to stimulate fusion (Fig. 2-4B). These results confirm that rat syt 4 fails to stimulate membrane fusion due to a lack of activity in both of its C2 domains. We then tested the PS binding activities of the isolated syt 4 C2-domains using the supernatant-depletion liposome co-sedimentation assay (Fig. 2-4C, Fig. S2-3A). Fly syt 4 C2B bound to PS in response to Ca 2+ ; in contrast, rat syt 4 C2B, as well as rat and fly syt 4 C2A, failed to bind to PS in either the presence or absence of Ca 2+. We also tested the t-snare binding activities of each C2 domain using the reconstituted t- SNARE liposome co-flotation assay. Analogous to the PS binding activity of each C2-domain above, fly syt 4 C2B bound to t-snare heterodimers in a Ca 2+ -promoted manner. In contrast, rat syt 4 C2B exhibited robust Ca 2+ independent t-snare binding activity. The rat and fly syt 4 C2A domains bound to t-snare heterodimers weakly in the absence and presence of Ca 2+ (Fig. 2-4D, Fig. S2-3B). In light of these findings - using isolated C2-domains - we built and tested another chimeric protein: the C2A domain of rat syt 4 was fused to the C2B domain of fly syt 4 (designated rat4a fly4b) (S2-4A). We found that this chimera was not able to stimulate fusion when tested at

50 40 concentrations up to 1 µm (Fig. S2-4B). These findings indicate that rat syt 4 C2A is able to shut off a functional adjacent C2B domain. These data are consistent with our findings in Fig. 2-3B demonstrating that rat syt 4 C2A is also able to inhibit the function of an adjacent syt 1 C2B domain. For completeness, we tested an additional chimera: the C2A domain of fly syt 4 was fused to the C2B domain of rat syt 4 (designated fly4a rat4b). This chimera also failed to stimulate fusion, in agreement with the observation that each of the isolated C2-domains used to build the construct were without activity in the reconstituted fusion assay (Fig. S2-4A and B). Reversal of the S244 mutation endows rat syt 4 with the ability to function as a Ca 2+ sensor for fusion - As detailed above, syt 4 - in both fly and rat - harbors a D-to-S substitution of one of the Ca 2+ ligands within its C2A domain. Previous studies have examined the functional consequences of this mutation in a neuroendocrine cell line. First, amperometric recordings of LDCV exocytosis in PC12 cells over-expressing wild type rat syt 4 (which harbors S244) revealed an increase in stand-alone foot (SAF) signals as compared to cells over-expressing syt 1. SAF represent events in which fusion pores open and close again without dilating. The S244D mutation in rat syt 4 reduced the duration of SAF to that in cells overexpressing syt 1, but the frequency of SAF remained as high as that in cells overexpressing wild type syt 4 (Wang et al., 2003b). These data suggest that the S244D mutation converts, at least partially, the biological activity of syt 4 so that it is more similar to syt 1. We therefore tested whether the S244D mutation might be able to rescue the ability of rat syt 4 to stimulate membrane fusion in vitro. As shown in Fig. 2-5A, we found that the S244D mutant form of rat syt 4 was in fact able to stimulate membrane fusion. This gain-of-function suggests that rat syt 4 evolved to inhibit membrane fusion. For comparison we generated and tested a fly syt 4 mutant that carries an analogous mutation (fly syt 4 S284D). The mutant fly syt 4 stimulated fusion in response to Ca 2+ to levels similar to wild-type fly syt 4 and no apparent gain-of-function was observed (Fig. 2-5B).

51 41 To further understand how the S244D mutation activates rat syt 4, we examined its PS and t-snare binding activity, again using the co-sedimentation and co-flotation assays detailed above. These experiments revealed that the point mutation gave rise to Ca 2+ -regulated binding of the intact cytoplasmic domain of rat syt 4 to both of these effectors (Fig. 2-6A and B, Fig. S2-5A and B). The gain-of-function brought about by the S-to-D mutation could be restricted to the C2A domain of rat syt 4. Alternatively, and as noted above, this mutation could act by altering the ability of C2A to influence the adjacent C2B domain (Fig. S2-2). To address this question, we analyzed the isolated C2A domain, which harbored this mutation, in the fusion assay and found that it stimulated fusion in response to Ca 2+ (Fig. 2-5C). Hence, the S244D mutation endows the C2A domain of rat syt 4 with the ability to function as a Ca 2+ -sensor capable of regulating SNARE catalyzed fusion reactions. Next, we examined the PS and t-snare binding activity of the isolated C2A domain that bore the S244D mutation using co-sedimentation and co-immunoprecipitation assays. Isolated S244D syt 4 C2A bound to PS-bearing liposomes in response to Ca 2+ (Fig. 2-6A, Fig. S2-5A), consistent with a previous report (von Poser et al., 1997). Since the t-snare binding activity of the rat syt 4 C2A S244D mutant was too weak to be quantified using co-flotation assays (where low affinity interactions can result in significant levels of dissociation during centrifugation), we used a co-immunoprecipitation (IP) approach to monitor binding. Isolated S244D rat syt 4 C2A was incubated with full-length t-snare heterodimers in detergent and immunoprecipitated using an anti-syntaxin antibody. Bound material was subjected to SDS-PAGE and the gels stained with Coomassie blue. Rat syt 4 S244D C2A bound to t-snare heterodimers weakly in the absence of Ca 2+ ; binding was significantly increased by addition of Ca 2+. Syt 1 served as a control and exhibited robust Ca 2+ -dependent t-snare binding activity (Fig. 2-6C, Fig. S2-5C).

52 42 Among all the syt isoforms, syt 11 has the greatest degree of homology to rat syt 4, and harbors the D-to-S substitution of the same Ca 2+ ligand as syt 4 (Craxton, 2004). Therefore, in the final series of experiments, we compared rat syt 4 and syt 11 in the reconstituted fusion assay. Analogous to syt 4, wild type syt 11 failed stimulate fusion, while reversal of the Ca 2+ ligand mutation in the C2A domain of syt 11, back to an aspartic acid residue, endowed the protein with the ability to stimulate membrane fusion in response to Ca 2+ (Fig. 2-7).

53 43 2.4) Discussion Syt 4 is an interesting member of the syt family because its expression is induced by seizures and activity (Wu, 2012). Indeed, syt 4 KO mice exhibit defects in memory and learning tasks and syt 4 has emerged as a critical regulator of synaptic plasticity in mice (Basu et al., 2005; Ferguson et al., 2000a; Ferguson et al., 2004a; Ferguson et al., 2004b). syt 4 has been studied in detail in fruit flies, where it also affects aspects of synaptic plasticity and neuronal growth (Barber et al., 2009; Shen et al., 2007). However, studies based on mice and flies have resulted in apparently contradictory findings. Namely, studies of mammalian syt 4 indicate that it plays an inhibitory role in secretion. For instance, over-expression of syt 4 reduced the frequency of LDCV fusion events in PC12 cells, shortened the time from fusion pore opening to dilation (Wang et al., 2001), increased the frequency and duration of kiss-and-run events (Wang et al., 2003b), modulated LDCV exocytosis in peptidergic nerve terminals of the neurohypophysis (Wang et al., 2011), and inhibited pre- and postsynaptic BDNF release in neurons (Basu et al., 2005). In contrast, studies using fruit flies as a model system suggest that fly syt 4 is a positive regulator of secretion. Syt 4 in muscle cells at the fly NMJ was required for normal pre-synaptic growth and for aspects of synaptic plasticity. It was therefore proposed that fly syt 4 serves to promote the release of a retrograde factor from muscle cells to influence pre-synaptic structure and function (Shen et al., 2007). The goal of the current study was to investigate the seemingly disparate functions reported for fly and rat syt 4. Using a reconstituted system, we directly compared the effects of fly and rat syt 4 on SNARE-mediated membrane fusion reactions. First, we confirmed that rat syt 4 failed to stimulate fusion in response to Ca 2+ (Bhalla et al., 2008). Strikingly, and in marked contrast to the rat protein, fly syt 4 promoted robust Ca 2+ stimulated fusion in vitro (Fig. 2-1B). These results

54 44 help to resolve the disparate findings regarding the function of syt 4; the fly and rat proteins are not functional orthologs of one-another. We note that cytoplasmic domain of fly syt 4 is homologous to rat syt 11 (49% identity) (Craxton, 2004). However, since rat syt 11 is an inhibitory isoform (Bhalla et al., 2008), as in the case of rat syt 4, it too does not appear to serve as a functional ortholog of fly syt 4. Previous studies have established that rat syt 1 plays a dual role in regulating membrane fusion reactions in vitro and at synapses (Chicka et al., 2008a; DiAntonio and Schwarz, 1994; Gaffaney et al., 2008; Geppert et al., 1994; Jackson, 2009; Littleton et al., 1994; Mackler et al., 2002). While rat syt 1 promotes fusion in the presence of Ca 2+, this isoform also serves to clamp or inhibit fusion in the absence of Ca 2+. This clamping activity appears to be mediated by the Ca 2+ -independent component of syt t-snare interactions (Chicka et al., 2008a). While rat syt 4 does not stimulate fusion in response to Ca 2+, it strongly clamps fusion in the absence of Ca 2+ (Fig. 2-1B, right panel) (Bhalla et al., 2008). Consistent with syt 4 s strong clamping function, its Ca 2+ - independent t-snare binding activity is also very robust (Fig. 2-2B) (Wang et al., 2003b). While one of the acidic Ca 2+ ligands within the C2A domain of syt 4 has been replaced over the course of evolution with a serine residue, all of the acidic Ca 2+ ligands are conserved within the C2B domain. Thus, the apparently obvious explanation for the functional difference between rat syt 1 and syt 4 is the S-to-D mutation within C2A. However, our in vitro fusion data - using chimeras between syt 1 and 4 - indicate that both C2-domains of rat syt 4 exhibit functional deficiencies (Fig. 2-3). This notion was confirmed by analysis of the isolated C2-domains. In contrast to syt 1, whose isolated C2B domain is able to regulate fusion (Gaffaney et al., 2008), neither isolated C2A nor C2B from rat syt 4 were able to stimulate fusion. When we examined the isolated C2-domains of fly syt 4 in the fusion assay, both fly and rat C2A failed to stimulate fusion (owing to the S-to-D mutation), but, interestingly, the isolated C2B

55 45 of fly syt 4 was able to stimulate fusion, while rat C2B did not. The fact that the C2B domains of rat and fly syt 4 have such divergent properties in reconstituted fusion reactions was highly unexpected; these proteins exhibit 56% sequence identity and their isolated C2B domains have perfectly conserved acidic Ca 2+ ligands. Our biochemical analysis revealed that the C2B domain of rat syt 4 does not possess Ca 2+ dependent PS or t-snare binding activity. In contrast, the C2B domain from fly syt 4 exhibits Ca 2+ -promoted interactions with both of these effectors (Fig. 2-4C and D). Moreover, a gain-of-function mutation, in which a Ca 2+ ligand is restored by replacing the serine at position 244 with an aspartate residue in rat syt 4, endowed the protein with the ability to stimulate membrane fusion in response to Ca 2+. Further analysis showed that this functional rescue was due to the gain of Ca 2+ -regulated PS and t-snare binding activity exhibited by the C2A domain. We point out that expression levels of rat syt 4 are controlled by neuronal activity; increases in activity result in upregulation, thus allowing syt 4 to compete with positive regulators of LDCV secretion to down-regulate exocytosis (Wang et al., 2001; Wang et al., 2011). In neurons, this down-regulation results in a reduction in BDNF secretion to dampen neuronal activity and to place an upper limit on long term potentiation (Basu et al., 2005). Rat syt 4 therefore appears to have evolved as a homeostatic regulator of synaptic activity. In contrast, fly syt 4 appears to serve as a positive regulator of secretion. Whether flies express inhibitory isoforms of syt remains an open issue. In this light it is notable that seventeen isoforms of syt have been identified in vertebrates, but only seven isoforms are expressed in flies. Future studies, in which each isoform of fly syt is screened in fusion reactions using a variety of fly SNARE proteins, should reveal inhibitory isoforms of syt (Bhalla et al., 2008). In summary, we have directly addressed the abilities of rat and fly syt 4 to regulate SNAREcatalyzed membrane fusion reactions. The major finding here is that fly syt 4 was able to respond

56 46 to Ca 2+ to drive rapid and robust membrane fusion. This sharply contrasts the inability of rat syt 4 to couple Ca 2+ to fusion. However, both rat and fly syt 4 were able to clamp fusion to some extent in the absence of Ca 2+, and our biochemical data indicate this is due to relatively efficient binding of these proteins to t-snares to shut them off. Finally, while most attention has been focused on the conserved D-to-S substitution of a Ca 2+ ligand in the C2A domain of fly and rat syt 4, the data reported here indicate that functional differences between these proteins extend to their C2B domains.

57 47 2.5) Experimental Procedures DNA Constructs - A plasmid for the expression of mouse synaptobrevin 2 in E. coli was provided by J.E. Rothman (Yale University, New Haven, CT) (Weber et al., 1998); full-length t- SNARE heterodimers were generated as described previously by subcloning cdna encoding full length rat SNAP-25B and rat syntaxin 1A into the prsfduet-1 vector (Novagen) (Chicka et al., 2008a). cdna encoding fly syt 1, fly syt 4, fly n-syb, fly syntaxin 1A and fly SNAP-25 were provided by J.T. Littleton (Massachusetts Institute of Technology, Boston, MA). For expression in bacteria, the cytoplasmic domains of fly syt 1 ( ) and fly syt 4 ( ) were subcloned into pgex-4t vectors. Point mutations were generated by Quick Change mutagenesis (Stratagnene). Rat chimeras were subcloned into pgex-4t vectors as follows: rat1a 4B: residues of syt 1 and residues of syt 4; rat4a 1B: residues of syt 4 and residues of syt 1. Full length fly n-syb was subcloned into the pet-28a vector using the EcoRI and BamHI sites. To generate fly t-snare heterodimers, cdna encoding full-length fly SNAP- 25 was subcloned into prsfduet-1 (Novagen) using the EcoRI and NotI sites; full-length fly syntaxin was subcloned into a downstream site via BgI II and KpnI sites. The rat and fly syt 4 chimeras were subcloned into pgex-4t vectors as follows: rat4a fly4b: residues of rat syt 4 and residues of fly syt 4; fly4a rat4b: residues of fly syt 4 and residues of rat syt 4. Cytoplasmic domain of syt 11 was subcloned into pgex-4t vectors. Protein Expression and Purification Recombinant proteins were generated as described previously (Gaffaney et al., 2008). Liposome preparation - Proteoliposomes were prepared as described (Tucker et al., 2004). Briefly, lipids were dried under a stream of nitrogen and resuspended in elution buffer (25 mm HEPES, 400 mm KCl, 10% glycerol, 1 mm dithiothreitol, 1% n-octylglucoside) containing

58 48 SNARE proteins. Mixtures were diluted with dialysis buffer (25 mm HEPES, 100 mm KCl, 10% glycerol, 1 mm dithiothreitol) and centrifuged for five hours at 41,000 rpm in an Accudenz gradient. Liposomes were collected (1.2 ml) from the 0 and 30% Accudenz interface. For preparation of protein free liposomes, lipids (15% PS, 30% PE, 55% PC) were dried under a stream of nitrogen and resuspended in dialysis buffer. The mixtures were then extruded through polycarbonate membranes (100 nm, GE healthcare) to form unilamellar liposomes. In vitro fusion assays - Fusion assays were performed as described (Gaffaney et al., 2008) but using 10-fold lower amounts of vesicles and proteins. Under these conditions, less syt is needed to drive fusion at rates comparable to our earlier studies. Briefly, 75 µl fusion reactions were prepared, including 4.5 µl of t-snare vesicles or protein-free vesicles, 0.5 µl of NBD- Rhodamine-labeled v-snare vesicles and 1 µm syt. The mixtures were pre-incubated at 37 C for 20 minutes in the presence of 0.2 mm EGTA, followed by injection of Ca 2+ (1 mm); fusion was monitored for an additional hour. At the end of each run, 20 µl of the detergent n-dodecyl β- d-maltoside was added to each reaction to yield the maximum fluorescence signals at infinite dilution of the FRET donor-acceptor pair. NBD-dequenching was monitored using a BioTek Synergy HT plate reader. Co-sedimentation assays Syt proteins (4 µm) were incubated with increasing concentrations of liposomes (15% PS (phosphatidylserine), 55% PC (phosphatidylcholine), 30% PE (phosphatidylethanolamine)) for 15 min at room temperature in a final reaction volume of 100 µl. The mixtures were then centrifuged at 70,000 rpm for 1 hour. The supernatant of each sample was collected, mixed with 25 µl 3X SDS loading buffer, and boiled for 5 min. Samples were subjected to SDS-PAGE and stained with Coomassie blue. Co-flotation assays - 45 µl of PS-free t-snare vesicles were mixed with the indicated syt proteins (10 µm final concentration) in the presence of either 1 mm Ca 2+ or 0.2 mm EGTA in a

59 49 total reaction volume of 100 µl at room temperature for 30 minutes. Samples were mixed with 100 µl of 80 % Accudenz and transferred to centrifuge tubes (Beckman Instruments. Inc). 35%, 30% and 0% Accudenz were sequentially added to form step gradients. Samples were centrifuged at 55,000 rpm for 105 min. 40 µl of each sample was collected at the interface between the 30% and 0% Accudenz layers and analyzed by SDS-PAGE and staining with Coomassie blue. Immunoprecipitation - Syt proteins (2 µm) were mixed with t-snare heterodimers (2 µm) in TBS buffer (20 mm Tris, 150 mm NaCl, ph 7.4) plus 0.5% Triton X-100 in a total volume of 150 µl for 1 hour at 4 ºC in the presence of 2 mm EGTA or 1 mm Ca 2+. T-SNARE heterodimers were immunoprecipitated with 2 µl of an anti-syntaxin antibody (HPC-1) for two hours followed by addition of 40 μl of a 50% slurry of Protein G Sepharose Fast-flow beads (Pharmacia). The mixtures were incubated for one hour and beads collected by centrifugation; pellets were washed with TBS buffer four times, boiled in sample buffer, subjected to SDS-PAGE, and stained with Coomassie blue.

60 50 2.6) Figures Figure 2-1

61 51 Figure 2-1. Effect of rat and fly syt 4 on reconstituted SNARE-mediated membrane fusion reactions. A, a schematic diagram of the in vitro fusion assay. B, left, the cytoplasmic domain of rat or fly syt 4 (1µM) was added to SNARE-bearing liposome fusion reactions; samples were incubated for 20 min prior to the addition of Ca 2+ ; after injection of Ca 2+ (1mM [final], indicated by arrows), fusion was monitored for another 120 min at 37 C. As controls, rat and fly syt 1 were assayed under identical conditions. NBD dequenching signals were normalized to the maximum fluorescence signal, obtained by adding detergent, and plotted as a function of time. Right, experiments were also carried out in the continued presence of 0.2 mm EGTA. C, the final extent of fusion, regulated by syt 1 or syt 4, was plotted against protein concentration in the presence of 1 mm Ca 2+ (left panel) or 0.2 mm EGTA (right panel) (n = 3).

62 Figure

63 53 Figure 2-2. PS and t-snare binding activities of rat and fly syt 4. A, left, representative gel of the co-sedimentation assay used to monitor syt-membrane interactions, in the presence of 0.2 mm EGTA or 1 mm Ca 2+. Syt was incubated with liposomes (15% PS/ 30% PE/ 55% PC) for 20 min; bound and free proteins were then separated via sedimentation as described in Methods. The supernatant fraction, which is depleted of syt protein upon binding liposomes, was subjected to SDS-PAGE and stained with Coomassie blue. Rat syt 4 failed to bind PS-harboring liposomes; in contrast, fly syt 4, rat syt 1, and fly syt 1 all bound to PS-harboring vesicles in a Ca 2+ dependent manner. Right, the amount of bound protein was calculated and plotted against [liposome] (n = 3). B, left, the t-snare binding activities of syt 4 and syt 1 were monitored using a co-flotation assay as described in Methods. Syt proteins were incubated with PS-free t-snare vesicles (30% PE /70% PC) in 0.2 mm EGTA and then floated through a density gradient with or without added Ca 2+ (1 mm). Syt that was associated with vesicles (via interactions with t-snare) was collected from the top layer of the gradient, subjected to SDS-PAGE, and stained with Coomassie blue. A representative gel is shown. Rat syt 4 exhibited strong t-snare binding activity in EGTA; This binding was unaffected by Ca 2+. In contrast, rat syt 1, fly syt 1 and fly syt 4 all exhibited weak t- SNARE binding activity in the absence of Ca 2+ and much stronger binding in the presence of Ca 2+. Neither rat syt 1 nor rat syt 4 bound to protein-free liposomes that lacked PS. Right, the molar ratio of syt to syntaxin in each sample was plotted. Tr: t-snare vesicles

64 Figure

65 55 Figure 2-3. Rat syt 1 and syt 4 chimeras have modest effects on reconstituted membrane fusion reactions. A, a schematic diagram summarizing the structural organization of rat syt 1 and syt 4 chimeras. B, the extent of fusion regulated by two rat chimeric proteins: rat1a 4B and rat4a 1B in the presence of 1 mm Ca 2+, was plotted against [syt]. Neither chimera stimulated significant levels of fusion (n = 3).

66 Figure

67 57 FIigure 2-4. The C2B domain of fly syt 4 directly promotes fusion in response to Ca 2+. A, alignment of Ca 2+ ligands in rat and fly syts. The conserved D-to-S Ca 2+ -ligand substitution is indicated by an open rectangular box. B, fusion reactions were carried out as described in Fig. 2-1B, using isolated fly and rat syt 4 C2A or C2B domains. Normalized fluorescence signals were plotted against [C2-domain]. Fly syt 4 C2B triggered fusion in response to Ca 2+ ; rat syt 4 C2A, C2B and fly syt 4 C2A all failed to stimulate fusion. C, the ability of isolated C2-domains from rat and fly syt 4 to bind PS-bearing liposomes was monitored using a co-sedimentation assay. The amount of bound protein was determined and plotted against [liposome]. Fly syt 4 C2B bound to PS-bearing liposomes in response to Ca 2+ ; the other three C2-domains failed to exhibit significant levels of binding. D, the t-snare binding activities of the isolated C2-domains from rat and fly syt 4 were monitored using a co-flotation assay as described in Fig. 2-2B. The molar ratio of syt to syntaxin in each sample was plotted. Both C2-domains of rat syt 4, as well as fly syt 4 C2A, bound to t-snares in a Ca 2+ independent manner. In contrast, fly syt 4 C2B bound to t-snare in a Ca 2+ dependent manner (n = 3). For representative gels of these experiments, please see Fig. S2-3A and B.

68 Figure

69 59 Figure 2-5. The S244D mutation in rat syt 4 results in robust Ca 2+ -sensor function during reconstituted membrane fusion. A, Fusion assays were carried out as described in Fig. 2-1B, but with the indicated syt constructs. In contrast to wild type rat syt 4, the S244D mutant stimulated membrane fusion in the presence of Ca 2+. B, An analogous mutation was also analyzed in fly syt 4 (S284D); this mutation was largely without effect. C, The S244D mutation endowed the isolated C2A domain of syt 4 with the ability to stimulate fusion in response to Ca 2+.

70 Figure

71 61 Figure 2-6. PS and t-snare binding activities of rat and fly syt 4 S-to-D mutants. A, Liposome binding assays were carried out by co-sedimentation and analyzed as described in Fig. 2-2A. Rat syt 4 failed to bind; rat syt 4 S244D, wild type fly syt 4, and fly syt 4 S284D all bound to PS-bearing liposomes in response to Ca 2+. B, The molar ratio of syt to syntaxin, as determined from co-flotation assays as described in Fig. 2-2B, was plotted (n = 3). Wild type and S244D mutant rat syt 4 bound to t-snares in a Ca 2+ independent manner; in contrast, wild type fly syt 4 and the S284D mutant exhibited Ca 2+ -promoted t-snare binding activity. C, Coimmunoprecipitation of syt with t-snare heterodimers composed of full-length SNAP-25B and syntaxin 1A. The isolated C2A domain of rat syt 4 S244D bound to t-snares in a Ca 2+ - dependent manner. Rat syt 1 was analyzed in parallel and served as a control. For representative gels of these experiments, please see Fig. S2-5A-C.

72 Figure

73 63 Figure 2-7. Analysis of wild type and mutant syt 11 in reconstituted membrane fusion assays. Wild type rat syt 11 failed to stimulate membrane fusion; the S246D mutation, which restores a Ca 2+ -ligand within the C2A domain, endowed the protein with the ability to stimulate fusion in response to Ca 2+.

74 64 SUPPLEMENTAL FIGURE Figure S2-1

75 65 FIGURE S2-1. Effect of rat and fly syt 4 on membrane fusion reactions mediated by reconstituted fly SNARE proteins. A, experiments were carried as described in Fig. 2-1B but using fly, rather than rat, v- and t-snare-bearing vesicles. NBD dequenching signals were normalized to the maximum fluorescence signals, obtained by addition of detergent, and plotted against time. B, experiments were carried out in the continued presence of 0.2 mm EGTA.

76 Figure S2-2 66

77 67

78 68 FIGURE S2-2. A D230S mutation in the C2A domain of syt 1 inhibited the ability of the adjacent C2B domain to stimulate fusion in response to Ca 2+. A, representative traces of Ca 2+ - triggered fusion, regulated by wild type syt 1 C2AB, C2AB that harbors the D230S mutation, or the isolated C2B domain of the protein, are shown. The concentrations of each protein are indicated on the right side of each panel. B, the final extent of fusion, regulated by syt 1 C2AB (D230S), C2AB, or C2B, was plotted against protein concentration.

79 Figure S2-3 69

80 70 FIGURE S2-3. PS and t-snare binding activities of the isolated C2-domains from rat and fly syt 4. A, representative gel of the co-sedimentation assay used to monitor syt-ps interactions. Experiments were carried out as described in Fig. 2-2A. Fly syt 4 C2B bound to PS bearing liposomes in the presence of 1 mm Ca 2+. In contrast, the other three C2 domains failed to bind PS-bearing liposomes. B, the t-snare binding activities of isolated rat and fly syt 4 C2 domains were monitored using a co-flotation assay as described in Fig. 2-2B. Fly syt 4 C2B exhibited weak t-snare binding activity in the absence of Ca 2+ and much stronger binding in the presence of Ca 2+. Rat syt 4 C2A, C2B and fly syt 4 C2A exhibited varying degrees of Ca 2+ independent t- SNARE binding activity that was not affected by addition of Ca 2+.

81 Figure S2-4 71

82 72 FIGURE S2-4. Rat and fly syt 4 chimeras failed to stimulate membrane fusion. A, a schematic diagram summarizing the structural organization of the rat syt 4 and fly syt 4 chimeras. B, the extent of membrane fusion regulated by the rat/fly syt 4 chimeras - fly4a rat4b and rat4a fly4b - in the presence of 1 mm Ca 2+, was plotted as a function of [chimera]. Neither chimera was able to stimulate fusion to a significant degree (n = 3).

83 Figure S2-5 73

84 74 FIGURE S2-5. PS and t-snare binding activities of rat and fly syt 4 proteins that harbor the S-to-D mutation. A, syt-ps interactions were monitored using the co-sedimentation assay as described in Fig. 2-2A; a representative gel - showing depletion of syt protein from the supernatant due to sedimentation with liposomes - is shown. The D-to-S mutant forms of rat and fly syt 4, as well as the mutant form of the isolated C2A domain of rat syt 4, bound to PS bearing liposomes in a Ca 2+ dependent manner. B, the t-snare binding activities of rat and fly syt 4, which harbor the S-to-D mutation, were monitored using a co-flotation assay as described in Fig. 2-2B; a representative gel is shown. C, the t-snare binding activity of the rat syt 4 C2A S-to-D mutant was monitored using a co-immunoprecipitation assay as described in Fig. 2-6C; a representative gel is shown.

85 75 2.7) Appendix 1) We thank C. Dean, J.D. Gaffaney and E. Hui for helpful suggestions. This study was supported by a grant from the NIH (MH 61876) to E.R.C. E.R.C. is an Investigator of the Howard Hughes Medical Institute. 2) Work in this chapter has been published in Journal of Biological Chemistry. RAT AND DROSOPHILA SYNAPTOTAGMIN 4 HAVE OPPOSITE EFFECTS DURING SNARE CATALYZED MEMBRANE FUSION Wang Z, Chapman ER. J Biol Chem Oct 1;285(40): Epub 2010 Aug 5 3) My contribution to all the data and figures in this chapter

86 76 Chapter III Reconstituted Synaptotagmin I Mediates Vesicle Docking, Priming, and Fusion.

87 77 3.1) Summary The synaptic vesicle protein synaptotagmin I (syt) promotes exocytosis via its ability to penetrate membranes in response to binding Ca 2+ and through direct interactions with SNARE proteins. However, studies using full-length, membrane-embedded syt in reconstituted fusion assays have yielded conflicting results, including a lack of effect, or even inhibition of fusion, by Ca 2+. Here, we show that reconstituted, full-length syt promoted rapid docking of vesicles (< 1 min), followed by a priming step (3-9 min) that was required for subsequent Ca 2+ -triggered fusion between v- and t-snare liposomes. Moreover, fusion occurred only when phosphatidylinositol 4,5-bisphosphate (PIP 2 ) was included in the target membrane. This system also recapitulates some of the effects of syt mutations that alter synaptic transmission in neurons. Finally, we demonstrate that the cytoplasmic domain of syt exhibited mixed agonist/antagonist activity during regulated membrane fusion in vitro and in cells. Together, these findings reveal further convergence of reconstituted and cell-based systems.

88 78 3.2) Introduction Elucidation of the molecular mechanisms that underlie Ca 2+ -triggered membrane fusion, and neurotransmitter release at synapses, can be directly addressed through in vitro fusion assays using reconstituted SNARE proteins. SNAREs form the core of a conserved membrane fusion complex in neurons, with vesicle SNAREs (v-snare, synaptobrevin (syb)) binding to target membrane SNAREs (t-snares, syntaxin and SNAP-25), thereby pulling the membranes together to catalyze fusion (Weber et al., 1998). This system has been used to study accessory proteins that regulate fusion, including the Ca 2+ sensor for exocytosis, synaptotagmin I (syt). Syt is anchored to synaptic vesicles (SV) via a single membrane spanning domain. To simplify the study of syt, most studies make use of the cytoplasmic domain of protein (which harbors both Ca 2+ sensing motifs, C2A and C2B, and is therefore designated C2AB) (Chicka et al., 2008b; Gaffaney et al., 2008; Schaub et al., 2006; Stein et al., 2007; Tucker et al., 2004; Xue et al., 2008). Recent studies have attempted to address the impact of full-length membrane-embedded syt on fusion in vitro. In one study, Ca 2+ was without effect (Mahal et al., 2002), while in another study Ca 2+ syt inhibited fusion. In this latter study, Ca 2+ syt was able to stimulate fusion only when phosphatidylserine (PS) was removed from the v-snare vesicle membrane (Stein et al., 2007); the physiological relevance of this finding is unclear as PS is present on both the SV and target membrane in vivo (Takamori et al., 2006b). A third study reported Ca 2+ triggered fusion using reconstituted full-length syt (FL syt), but in this case fusion was triggered by only a narrow range of [Ca 2+ ], centered around 10 µm (Lee et al., 2010). At [Ca 2+ ] > 25 µm, stimulation of fusion was not observed even though higher concentrations of Ca 2+ are achieved at release sites (Llinas et al., 1992, 1995) and robust neurotransmitter release occurs at tens to hundreds of µm

89 79 [Ca 2+ ] (Bollmann et al., 2000; Heidelberger et al., 1994; Heinemann et al., 1994; Thomas et al., 1993; Voets, 2000). Finally, in the most recent study, Ca 2+ -triggered fusion occurred, but only at Ca 2+ concentrations > 2 mm (Kyoung et al., 2011), a value far above the physiological range. To date, reconstituted membrane fusion systems incorporating FL syt, which mimic the native state, have yet to be described. Here, we define a FL syt-regulated membrane fusion assay that more accurately recapitulates a number of fundamental aspects syt-regulated exocytosis at synapses.

90 80 3.3) Results Effect of PIP 2 on Ca 2+ syt regulated fusion In some of the earlier studies of FL syt, a critical lipid, phosphatidylinositol 4,5-bisphosphate (PIP 2 ), was not included in the reconstituted vesicles (Mahal et al., 2002; Stein et al., 2007). PIP 2 plays an essential role in the Ca 2+ triggered exocytosis of large dense core vesicles (LDCV) in neuroendocrine cells (Eberhard et al., 1990; Hay et al., 1995), and might also play a key role in SV exocytosis (Zheng et al., 2004), although this latter issue remains to be fully explored. In neurons and neuroendocrine cells, PIP 2 is concentrated on the inner leaflet of the plasma membrane and is absent from secretory vesicles (Holz et al., 2000; Micheva et al., 2001). Ca 2+ - independent interactions with PIP 2 have been shown to steer the membrane penetration activity of syt toward the PIP 2 -harboring membrane (i.e. the plasma membrane), rather than the vesicle membrane, in response to Ca 2+ (Bai et al., 2004). Since syt stimulates fusion by selectively acting on the target membrane (Chicka et al., 2008b), and since interactions with the vesicle membrane are favored kinetically (Bai et al., 2000), we hypothesized that PIP 2 -mediated steering of syt would be essential for productive fusion (Bai et al., 2004). To test this, we titrated [PIP 2 ] in t- SNARE vesicles (Tr), and reconstituted syt in v-snare vesicles (Vr-syt) (Fig. 3-1 A-D). Fusion between Vr-syt and Tr vesicles was monitored by loss of FRET between a lipidic donor-acceptor pair as shown in Fig. 3-1 A; briefly, vesicles were mixed and monitored for 20 min, Ca 2+ was injected, and fusion monitored an additional 60 min (Fig. 3-1 B). When PIP 2 was <1% (molar ratio relative to total lipid), Ca 2+ -stimulated fusion was not observed; at >1%, Ca 2+ -triggered membrane fusion became apparent and both extent and rate of the fusion were further enhanced by increasing the PIP 2 on Tr vesicles up to 5%, the highest concentration tested (Fig. 3-1 B-D; we note that PIP 2 has been estimated to reach 6% of the total lipid within rafts in cells (James et al.,

91 ). Ca 2+ -triggered fusion was not observed when syt was not present on v-snare vesicles (Vr) (Fig. 3-1 E). These data indicate that PIP 2 is a critical effector for the action of FL syt during regulated fusion. The Ca 2+ -free pre-incubation step was critical; omission of this step resulted in fusion reactions that were not triggered by Ca 2+ to any appreciable degree (Fig. 3-1 F). The half maximal rate and extent of fusion accord with pre-incubations of 9 and 3 min, respectively (Fig. 3-1 G and H), so a 20 minute incubation time was selected for all subsequent reactions. To determine whether the membrane fusion signal observed in these experiments reflected hemi- or full-fusion, we used dithionite to selectively quench fluorophores on the outer leaflet, thereby revealing the fusion signal from only the inner leaflet. The fluorescence signal was reduced by ~60% at all stages of the fusion reaction (Fig. S3-1), demonstrating equal lipid mixing in both the inner and outer leaflets. These data indicate that full fusion occurs during all phases of these reconstituted vesicle fusion reactions. Specificity of the phosphatidylinositol bisphosphate requirement for regulated fusion To further probe the role of PIP 2 during fusion, we used the drug, neomycin, which is an antibiotic that specifically binds to PIP 2 (Griffin et al., 1980), and has been shown to inhibit synaptic transmission (Zheng et al., 2004). We titrated neomycin into FL syt regulated fusion reactions and found that 30 µm drug abolished Ca 2+ -triggered fusion (Fig. 3-2 A-B; IC 50 = 6.9 µm), a concentration that was without effect on C2AB-regulated fusion reactions. Complete inhibition of Ca 2+ C2AB regulated fusion required mm concentrations of neomycin (Fig. 3-2 C-D; IC 50 = 357 µm), and is likely to be a non-specific effect. Cell membranes contain several phosphatidylinositol bisphosphates, but only Ptdlins(4,5)P 2 is required for exocytosis (Eberhard et al., 1990; Hay et al., 1995; Tucker et al., 2003). To

92 82 determine whether syt binds to PIP 2 specifically to regulate fusion, three different phosphatidylinositol bisphosphates were reconstituted into Tr (Fig. 3-2 E). Only Ptdlins(4,5)P 2 was able to stimulate fast and robust Ca 2+ dependent fusion; Ptdlins(3,5)P 2 or Ptdlins(3,4)P 2 were significantly less effective (extent and rate of fusion reduced >35% compared to PIP 2 ) (Fig. 3-2 F and G). Optimal syt density and the Ca 2+ sensitivity of fusion To determine the number of copies of syt per vesicle required for optimal fusion activity, we titrated the amount of reconstituted syt molecules incorporated into v-snare vesicles (Vr) (Fig. 3-3 A-B); 3% PIP 2 was included in all t-snare vesicles (Tr). Vr vesicles lacking syt exhibited only small responses to Ca 2+ ; however, inclusion of syt in these vesicles, even at relatively low copy numbers (six per vesicle), resulted in fusion that was strongly stimulated, in terms of both rate and extent, by Ca 2+. Increasing the amount of syt further enhanced the Ca 2+ -dependent response until saturation was reached at thirty copies of syt per vesicle; the maximal response occurred between twelve and thirty copies. Interestingly, this range coincides with finding that SVs harbor fifteen copies of syt (Takamori et al., 2006b). The Ca 2+ sensitivity of syt-promoted fusion was determined using Vr vesicles that harbored thirty copies of syt and Tr vesicles with 3% PIP 2. Corrections for these measurements are detailed in Fig. S3-2 A-C and the dose-response curve is shown in Fig. 3-3 C. The [Ca 2+ ] 1/2 was 250 µm and the Hill coefficient was 1.5, indicating some degree of cooperativity. This Ca 2+ -sensitivity is similar to the half-maximal [Ca 2+ ] for exocytosis from goldfish retinal bipolar neurons (194 µm (Heidelberger et al., 1994)) but is less than estimates from autaptic cultures of hippocampal neurons (Burgalossi et al., 2010) or the calyx of Held (Bollmann et al., 2000; Schneggenburger and Neher, 2000). It should be noted, however, that the Ca 2+ -sensitivity is directly related to the

93 83 concentration of anionic phospholipids in the in vitro system (Tucker et al., 2004), and lower [Ca 2+ ] 1/2 values can be obtained using a higher mole fraction of PS (Tucker et al., 2004). In stark contrast to a previous report in which a lack of response was observed at high [Ca 2+ ] (Lee et al., 2010), we found robust fusion activity at all Ca 2+ concentrations tested, a result that more accurately reflects the in vivo behavior of the regulated fusion machinery (Bollmann et al., 2000; Heidelberger et al., 1994; Heinemann et al., 1994; Voets, 2000). We also confirmed that fusion was mediated by trans-snare pairing, as the cytoplasmic domains of the t-snare hetero-dimer (cd t-snares; Fig. 3-3 D and F) and synaptobrevin (cd syb; Fig. 3-3 E and F) completely blocked lipid mixing. Topological requirements for PS, PIP 2, and syt PS is the major acidic phospholipid in neurons, and is crucial for syt to penetrate and bend membranes to promote fusion (Bhalla et al., 2005b; Hui et al., 2009). Interestingly, when PS was absent from Tr vesicles, regulated fusion was abolished (Fig. 3-4 A). In contrast, fusion was largely unaffected by omission of PS from Vr vesicles (Fig. 3-4 A). These results are consistent with a model in which the C2-domains of FL syt act, in trans, on the target membrane to stimulate fusion. Moreover, these findings contrast the mechanism of fusion regulated by the cytoplasmic domain of syt, designated C2AB, which requires the presence of PS on both Vr and Tr membranes (Bhalla et al., 2005b). Analogous to PS, PIP 2 must also be present on the target membrane in order for Ca 2+ syt to stimulate fusion (Fig. 3-4 B). This finding further indicates that syt executes its function by acting on the t-snare membrane, as predicted from earlier biochemical studies (Chicka et al., 2008b; Hui et al., 2009).

94 84 We also addressed the topological requirements for syt and found that fusion was stimulated by Ca 2+ only when syt was present on Vr vesicles (Fig. 3-4 C). The finding that syt must be localized to the vesicle membrane is consistent the targeting of this protein to vesicles in vivo, and with the notion that syt acts as a docking factor via interactions with t-snares and PIP 2 (Bai et al., 2004, Liu, 2009 #94; de Wit et al., 2009; Reist et al., 1998). Mutational analysis of syt during regulated fusion To further address the mechanism by which reconstituted syt regulates fusion, we examined mutant forms of the protein. A positively charged patch on the side of C2B plays a critical role in PIP 2 -mediated steering of syt (Mackler and Reist, 2001). Steering activity in vitro (Bai et al., 2004), and SV exocytosis in vivo (Loewen et al., 2006; Takamori et al., 2006b) were both impaired by mutations - K326,327A - that neutralize these positive charges (Fig. 3-5 A). More specifically, these mutations resulted in a 40% reduction in neurotransmitter release at the Drosophila neuromuscular junction (Loewen et al., 2006) and a 50% reduction in autaptic cultures of hippocampal neurons (Takamori et al., 2006b). These findings are in reasonable agreement with the ~70% reduction observed in our simplified, reduced in vitro fusion assay (Fig. 3-5 B). A number of additional syt mutations have been characterized. One such mutant, designated AD1, which has been studied in detail in Drosophila (DiAntonio and Schwarz, 1994; Yoshihara and Littleton, 2002) (Broadie et al., 1994), lacks the C2B domain, resulting in a strong loss of function phenotype (Fig. 3-5 A). Consistent with the fly physiology, the AD1 mutant was unable to promote membrane fusion in response to Ca 2+ (Fig. 3-5 C), thus confirming that the C2B domain of syt is indispensable for regulated fusion (Broadie et al., 1994; Gaffaney et al., 2008; Yoshihara and Littleton, 2002).

95 85 We also mutated Ca 2+ ligands in either the C2A or C2B domain of FL syt (Fig. 3-5 A). In earlier in vitro fusion assays using C2AB, Ca 2+ -ligand mutations in the C2A domain resulted in a more severe loss of activity than analogous mutations in the C2B domain (Bhalla et al., 2005b; Stein et al., 2007). However, in neurons, mutation of Ca 2+ ligands in C2A are tolerated, but Ca 2+ - ligand mutations in C2B completely disrupt the ability of syt to drive synchronous SV exocytosis (Mackler et al., 2002; Nishiki and Augustine, 2004a). We found that the disparity between cellbased and in vitro fusion assays was partially resolved via the use of full-length reconstituted syt; mutations in the C2B domain disrupted most of the ability of membrane embedded syt to stimulate fusion in response to Ca 2+, whereas mutations in C2A were less deleterious (35% reduction in the extent of fusion) (Fig. 3-5 D). Despite this convergence regarding the C2Bdomain, the reconstituted fusion assay still fails to recapitulate the lack of effect, or gain-offunction, reported for Ca 2+ ligand mutations in the C2A domain (Robinson et al., 2002b; Stevens and Sullivan, 2003). The transmembrane domain (TMD) of syt is not necessary for syt to regulate neuronal exocytosis (Hui et al., 2009). For example, when C2AB was targeted to synaptic vesicles by fusing it with the synaptic vesicle protein synaptophysin, synchronous neurotransmitter was fully restored in syt I knock out neurons (Hui et al., 2009). To extend this observation to the reconstituted system, we linked C2AB to Vr vesicles via conjugation with maleimidephosphatidylethanolamine. Interestingly, conjugated C2AB stimulated fusion in response to Ca 2+ in a manner analogous to FL syt (Fig. S3-3 A-C). These data further validate the observation that the TMD domain of syt is dispensable for fusion. Context-dependent mixed antagonist/agonist activity of C2AB

96 86 C2AB stimulates fusion in vitro (Tucker et al., 2004), but has been shown to inhibit fusion in PC12 (Desai et al., 2000) and chromaffin cells (Rickman et al., 2004). While these findings might appear to be contradictory, we note that C2AB was only able to inhibit fusion in cells to a limited degree, suggesting that this protein fragment might have mixed agonist/antagonist activity. In addition, it is possible that C2AB is less efficacious in terms of regulating fusion than the fulllength protein. In this case, and in the presence of intact syt, addition of C2AB would be predicted to diminish fusion to some degree. The new system reported here, based on the reconstitution of active FL syt, makes it possible to test these ideas. We titrated C2AB into fusion assays that contained FL syt; at relatively low concentrations, C2AB slightly, but reproducibly, inhibited Ca 2+ promoted membrane (Fig. 3-6 A-C). In contrast, in the absence of FL syt, C2AB only stimulated fusion, and this effect required relatively high concentrations of the protein (>3µM)(Fig. S3-4 A and B). Interestingly, the rate of fusion was reduced by increasing [C2AB] in the presence of FL syt. These results agree with our general finding that in response to Ca 2+, FL syt-regulated fusion occurs with faster kinetics than C2ABregulated fusion reactions (e.g. see Fig. 3-7, further below). To extend these experiments to a native system, we expressed C2AB in cultured chromaffin cells (Fig. S3-4 C). In wild type (WT) cells, over-expression of C2AB reduced the rate of exocytosis from the readily releasable pool of vesicles (RRP) (Fig. 3-6 D-F). Interestingly, overexpression of C2AB in syt I KO chromaffin cells did not inhibit fusion, but rather restored the size the RRP and partially restored the fast rate of release (Fig. 3-6 D-F). Over-expression of C2AB, in either wt and syt I KO cells, had no significant effect on the size or release rate of the slowly releasable pool of vesicles (SRP) or the sustained phase of release (Fig. S3-4 D-G). Together, these results are consistent with a previous report demonstrating that endogenous syt functions to regulate release from the RRP (Voets et al., 2001a). Thus, in reconstituted systems,

97 87 and in cells, C2AB can partially inhibit fusion in the presence of FL syt, but acts only to stimulate fusion in the absence of the full-length protein. These findings are interpreted further below. Systematic comparison of fusion reactions regulated by FL or the cytoplasmic domain of syt In the course of analyzing the impact of FL syt on fusion, we incorporated three important modifications in the fusion assay: incorporation of PIP 2 in Tr vesicles, addition of PE to both Tr and Vr vesicles, and the inclusion of a Ca 2+ -free pre-incubation step. To clarify potential differences between the full-length protein versus C2AB, each of these conditions were systematically explored using both proteins (Fig. 3-7 A-C, Fig. S3-5 A-C). PIP 2 and the preincubation step were both essential for FL syt to stimulate fusion in response to Ca 2+, but were not required for C2AB to regulate fusion. These findings suggest that the FL protein might act via a somewhat distinct mechanism than C2AB, as detailed in the next section. Inclusion of PE enhanced fusion reactions regulated by both FL syt and C2AB. Finally, in response to Ca 2+, FL syt-regulated fusion was less efficient than C2AB-regulated fusion, but occurred with faster kinetics. Distinct mechanism of FL syt and C2AB mediated fusion C2AB promotes fusion, in part, by aggregating vesicles in response to Ca 2+ and thereby enhancing v- and t-snare pairing (Hui et al., 2011). Our finding that FL syt stimulates fusion in response to Ca 2+ only after v- and t-snare vesicles have been pre-incubated together in EGTA, in conjunction with a report indicating that membrane embedded syt mediates secretory vesicle docking in cells (via interactions with t-snares) (de Wit et al., 2009), prompted experiments to determine whether FL syt promotes docking in our in vitro assay. To test this, we measured

98 88 vesicle aggregation in fusion reactions that contained FL syt or C2AB and found that the full length membrane-embedded protein drove rapid aggregation (<1 min) in EGTA; addition of Ca 2+ did not result in further vesicle aggregation but did stimulate fusion, presumably by acting on predocked vesicle complexes (Fig. 3-8 A-B). When FL syt was omitted from the system, we did not observe appreciable aggregation (Fig. 3-8 A). In sharp contrast to experiments using FL syt, aggregation was not observed in C2AB-regulated reactions until Ca 2+ was added (Fig. 3-8 B). To confirm that vesicle aggregation involved docking between Tr vesicles and Vr vesicles, a docking assay was utilized (Fig. 3-8 C). Vr vesicles, harboring either syt, syb or both proteins, were mixed with Tr vesicles that were immobilized on beads using avidin and biotin. Vesicles that harbored either syt or syb were pulled-down to some extent by Tr vesicles, but much more robust docking was observed when both proteins were present on the Vr vesicles. Inclusion of PIP 2 further enhanced docking mediated by both vesicular proteins (P < 0.05) (Fig. 3-8 D). These findings agree with studies reporting that native syt I mediates LDCV docking in chromaffin cells (de Wit et al., 2009) and SV docking in neurons (Reist et al., 1998), and with recent findings that syb is critical for docking of LDCVs in PC12 cells (Y.W. and E.R.C. unpublished data). Interestingly, PIP 2 failed to promote docking when Vr vesicles harbored only FL syt. This might due to the relatively weak interaction between syt and PIP 2 under Ca 2+ -free conditions, such that putative docking interactions were disrupted during the washing steps. Our comparisons of FL syt and C2AB are summarized in Fig. 3-8 E, illustrating that they act via somewhat distinct mechanisms. Finally, we note that aggregation occurs more rapidly (complete in < 1 min) than the priming step characterized above (t 1/2 3-9 min), suggesting the existence of a post docking step that has yet to be defined in molecular terms, but might involve the assembly of trans SNARE complexes.

99 89 3.4) Discussion In the current study we draw six major conclusions: first, PIP 2 is absolutely required for membrane-embedded FL syt to regulate SNARE-mediated membrane fusion; moreover, v- and t- SNARE vesicles must be pre-incubated together to prime fusion prior to the Ca 2+ trigger. Second, PIP 2 and PS are required only in the target membrane, consistent with models in which syt acts on the plasma membrane. Third, the FL syt system described here recapitulates three steps in the secretory pathway that occur in vivo: docking, priming, and subsequent Ca 2+ triggered fusion. In contrast, C2AB promotes all three steps only in response to Ca 2+ (Hui et al., 2011). Fourth, the effects of a number of syt mutations, in the FL syt fusion system described here, more closely mirror the effects of these mutations on synaptic transmission in vivo, as compared to previous work focused on C2AB. For example, Ca 2+ ligand mutations in C2B completely disrupt Ca 2+ - triggered fusion in our FL syt-regulated fusion assay and in vivo, but have little effect on C2AB regulated fusion reactions. Fifth, the number of syt molecules needed per vesicle to drive efficient fusion closely mirrors the syt density on SVs in vivo (~fifteen copies/vesicle); in contrast, higher concentrations of C2AB are needed to drive fusion (e.g. 1 µm C2AB versus 10 nm FL syt). Finally, we also addressed the ability of C2AB to both inhibit as well as stimulate fusion in vitro and in cells; in the presence of membrane embedded or native syt, this protein fragment exhibits mixed agonist/antagonist activity and in the absence of FL syt this fragment acts only to stimulate fusion. One of the most striking findings in the current study was the absolute requirement for a preincubation step prior to the Ca 2+ trigger; if syt/v- and PIP 2 /t-snare vesicles are not allowed to interact prior to the Ca 2+ signal, regulated fusion was not observed. Hence, there appears to be a novel priming step that involves both syt and PIP 2. Interestingly, the t 1/2 for priming is longer (3-9

100 90 min) than the time requirements for docking (< 1 min), indicating a second slow step that has yet to be elucidated in molecular terms. We speculate that this might involve the relatively slow partial assembly of trans SNARE pairs, and this idea will be tested in future studies using fluorescence probes to monitor SNARE structure. Further analysis of this step will be of interest as it is known that in cells, vesicles must undergo priming reactions, after docking, in order to become fusion competent (Zenisek et al., 2000). Priming in vivo involves several additional factors that are not included in our fusion system (e.g. Munc13 etc. (Brose et al., 2000; Martin, 2002)), so the priming step reported here does not reflect all the priming reactions that have been identified in cells. We note that when PIP 2 was included in both Tr and Vr vesicles, Ca 2+ triggered fusion was not compromised, as compared to the condition where PIP 2 was only present on Tr vesicles (Fig. 3-4 B). If weak Ca 2+ -independent interactions with PIP 2 serve to steer the C2-domains of syt toward the target membrane (i.e. the plasma membrane) (Bai et al., 2004), inclusion of PIP 2 in the Vr membrane might have been expected to inhibit fusion, but this did not occur. This is probably because PIP 2 and t-snares, on the target membrane, act in a synergistic manner to steer the Ca 2+ -triggered membrane penetration activity of syt to the target membrane, even when PIP 2 is present on both membranes. Indeed, recent studies indicate that PIP 2 interacts with t-snares (Murray and Tamm, 2009), and the relevant target for syt might correspond to a complex composed of these components (Tucker et al., 2003) that binds the C2-domains of FL syt avidly enough to mediate efficient steering. Another surprising finding was that FL syt, when reconstituted into only Tr vesicles, failed to promote fusion in response to Ca 2+ (Fig. 3-4 C). This result contrasts the ability of C2AB, when fused to a plasma membrane targeting motif, to rescue the syt I knockout (KO) phenotype in neurons (Hui et al., 2009). As noted above, some degree of targeting to SVs cannot be ruled-out

101 91 in the rescue experiments, but an alternative interpretation is that the N-terminal region of syt, that contains the sole TMD of the protein, prevents syt from stimulating fusion when reconstituted into the target membrane. This possibility is consistent with the finding that C2AB can partially rescue exocytosis in syt I KO chromaffin cells; clearly this soluble protein fragment is active, in reconstituted systems and in living cells. So, it is plausible that C2AB- when in the plasma membrane- might be rendered inactive by inclusion of the N-terminal domain. These findings raise the issue of whether the fraction of syt isoforms that are localized to the plasma membrane at steady state (presumably after fusion), have any function in membrane fusion reactions in vivo. A weakness of earlier in vitro fusion studies concerned the disparity between the effects of syt mutations on fusion in vitro versus the effects of these mutations on exocytosis from cells as determined using genetic and electrophysiological approaches. Namely, in neurons, Ca 2+ ligand mutations in C2A are tolerated, or even lead to a slight gain-of-function (Robinson et al., 2002b; Stevens and Sullivan, 2003), while analogous mutations in the C2B domain completely disrupt function (Mackler et al., 2002; Nishiki and Augustine, 2004a). In contrast, similar mutations in C2AB analyzed in reconstituted fusion reactions led to markedly different results; Ca 2+ -ligand mutations in C2A resulted in greater losses in activity than did mutations in C2B (Bhalla et al., 2005b; Stein et al., 2007). Here, we show that in the FL syt/pip 2 fusion assay, mutations in the C2B domain completely disrupt function, and thus mimic observations based on intact synapses. Although mutations in C2A do not yet recapitulate the synaptic physiology phenotype, they are clearly less deleterious than mutations in C2B. However, it should also be noted that expression of FL syt that harbors a mutation in a Ca 2+ ligand in the C2A domain does result in reductions in secretion in PC12 cells (Wang et al., 2006).

102 92 A key concern regarding earlier in vitro studies based on the cytoplasmic domain of syt was the fact that over-expression of C2AB in wild type PC12 cells (Desai et al., 2000; Tucker et al., 2003), or chromaffin cells (Rickman et al., 2004), inhibits exocytosis to some extent, suggesting that C2AB might not provide a valid means to study the positive role played by syt during fusion. Here we resolved this controversy by documenting the mixed agonist/antagonist activity of C2AB in both in vitro and cell-based experiments. Namely, we utilized two systems in which FL syt was functional: the reconstituted fusion assay described here, and wild type chromaffin cells. We also had variants of each system that lacked FL syt (i.e. omission of FL syt in Vr vesicles, and use of syt I KO chromaffin cells). We found that in the absence of FL syt, C2AB stimulated fusion in both systems and did not exhibit any inhibitory activity. In contrast, when FL syt was present, low concentrations of C2AB partially inhibited fusion in both reconstituted fusion reactions and in chromaffin cells. The implication from this latter experiment is that, in some ways, FL syt works better than C2AB, and that C2AB can interfere, to some degree, with the action of the intact protein. Indeed, FL syt and C2AB appear to regulate fusion via somewhat distinct mechanisms (Fig. 3-8). Together, these results indicate that C2AB has mixed agonist/antagonist activity in the presence of FL syt, but acts only as an agonist in the absence of the full length protein. We note that another tandem C2-domain protein Doc2 (Groffen et al., 2010; Orita et al., 1996; Yao et al., 2011), thought to regulate SV and LDCV exocytosis in a manner analogous to syt I, lacks a membrane anchor and is soluble. Also, a number of syt isoforms have potential splice variants lacking a transmembrane domain. Hence, it will be interesting to determine whether these soluble proteins regulate fusion in a manner analogous to the C2AB domain of syt. Unlike our previous work on C2AB, FL syt appeared to be unable to clamp fusion in the reconstituted system. In fact, in the absence of Ca 2+, membrane fusion was enhanced by

103 93 increasing the syt copy number on Vr vesicles. This lack of clamping activity is probably due to the strong spontaneous fusion of SUV vesicles during the Ca 2+ -free docking step. Indeed, we have recently shown that aggregation of v- and t-snare SUVs is sufficient to stimulate fusion to some extent, and this Ca 2+ -independent component of fusion would obscure the potential clamping activity of FL syt (Hui et al., 2011; Loewen et al., 2006; Stein et al., 2007). Future studies using GUV target membranes, or using lower the temperatures, might reduce the Ca 2+ - independent fusion rate, making it possible to probe for clamping activity by comparing the fusion of v-snare vesicles that do and do not harbor FL syt. It should be noted that vesicles prepared using different batches of lipids exhibited different degrees of Ca 2+ independent fusion, but all of the data in each individual panel are generated from the same stock of lipids. In a previous study (Chicka et al., 2008b), under Ca 2+ -free conditions, C2AB was proposed to clamp SNARE assembly at a step after vesicle docking. Our vesicle aggregation data (Fig. 3-8 B) indicate that in EGTA, vesicles were largely non-aggregated/undocked (Fig. 3-8 B). Thus, a more plausible explanation for C2AB-mediated clamping activity might be that this protein fragment down regulates fusion upstream of the docking/aggregation step. In summary, we have reconstituted active FL syt and found that this protein is required for docking, priming and fusion in an in vitro system. The next avenue of study will be to determine how each of these steps is related to changes in the structure of SNARE proteins and the assembly of SNARE complexes.

104 94 Experimental Procedure DNA Constructs. cdna encoding rat syt I was provided by T. C. Südhof (Stanford University, Menlo Park, CA). The D374 mutation was corrected by substituting this residue with glycine. A plasmid for the expression of recombinant mouse synaptobrevin 2 was provided by J.E. Rothman (Yale University, New Haven, CT) (Weber et al., 1998); full-length t-snare heterodimers were generated as described previously by subcloning cdna encoding full-length rat SNAP-25B and rat syntaxin 1A into the prsfduet-1 vector (Novagen) (Chicka et al., 2008b). Point mutations were generated by Quick Change mutagenesis (Stratagene). Protein expression and purification. Recombinant proteins were purified as described previously (Gaffaney et al., 2008). Briefly, Escherichia coli were grown at 37 C to an A 600 of 0.8, and protein expression was induced with 0.4 mm isopropyl 1-thio-β-d-galactopyranoside. After four hours, bacteria were collected by centrifugation, lysed via sonication, and then extracted with Triton X-100 for three hours at 4 C. Insoluble material was removed by centrifugation (17K rpm, 25 minutes), and the supernatant was applied to a Ni 2+ column using an AktaFPLC (GE- Amersham Biosciences). Bound protein was washed extensively with resuspension buffer (25 mm HEPES-KOH, 400 mm KCl, 50 mm imidazole, 10% glycerol, 5 mm 2-mercaptoethanol) containing 1% Triton X-100 followed by a wash buffer (25 mm HEPES-KOH, 400 mm KCl, 50 mm imidazole, 10% glycerol, 5 mm 2-mercaptoethanol, 1% n-octylglucoside); his-tagged proteins were eluted in the wash buffer with 500 mm imidazole. Vesicle preparation. Lipids were purchased from Avanti Polar Lipids (Alabaster, AL). Proteoliposomes were prepared as described (Tucker et al., 2004). Briefly, lipids (Vr vesicles: 15% PS, 30% PE, 55% PC; Tr vesicles: 25% PS, 30% PE, 42% PC, 3%PIP 2 ) were dried under a stream of nitrogen and resuspended in elution buffer (25 mm HEPES, 400 mm KCl, 10%

105 95 glycerol, 1 mm dithiothreitol, 1% n-octylglucoside) plus the indicated proteins. Mixtures were diluted in dialysis buffer (25 mm HEPES, 100 mm KCl, 10% glycerol, 1 mm dithiothreitol) and centrifuged for five hours at 41,000 rpm in an Accudenz gradient. Vesicles were collected (1.2 ml) from the 0 and 30% Accudenz interface. In vitro fusion assay. Fusion reactions (100 µl total volume) were composed of 8 µl of t-snare vesicles (8 nm final concentration), 1 µl of NBD-rhodamine-labeled v-snare vesicles (1 nm final concentration) that bear syt, and buffer (25 mm HEPES, 100 mm KCl, 1 mm dithiothreitol). Mixtures were pre-incubated at 37 C for 20 minutes in the presence of 0.2 mm EGTA, followed by injection of Ca 2+ (1 mm); fusion was monitored for an additional hour. At the end of each run, 20 µl of the detergent n-dodecyl β-d-maltoside was added to each reaction to yield the maximum fluorescence signals at infinite dilution of the FRET donor-acceptor pair. NBD-dequenching was monitored using a BioTek Synergy HT plate reader. Statistical significance was evaluated by using the two-tailed unpaired Student s t-test: ***p < All data shown are represented as mean ± SEM. Conjugation of C2AB to Vr vesicles. The lone endogenous cysteine in C2AB (C277) was mutated to an alanine, and glycine 96 at the N-terminus was mutated to cysteine, using the mutagenesis method described above. For labeling, lipid mixtures that contained maleimide-pe (Avanti Polar Lipids) were incubated with the mutated form of C2AB(G96C, C277A) for 30 min. DTT was then added to block the remaining maleimide functional groups. Then, syb was incubated with the mixtures for 20 min, and samples were diluted with fusion assay buffer and dialyzed against this buffer overnight. The dialyzed vesicles were purified on an Accudenz (Accurate Chemical & Scientific Corporation, NY) gradient. Preparation of mouse chromaffin cells. Adrenal glands were removed from newborn WT and syt I KO mice and digested with 1 ml Dispase II (Roche) for 20 min at 37 C to obtain isolated

106 96 chromaffin cells. Cells were incubated in DMEM supplemented with penicillin/streptomycin (40,000 U/L and 40 mg/l; Invitrogen, San Diego, CA) and 10% FBS at 37 C with 5% CO 2. Tails from syt KO pups were kept for genotyping. All procedures involving animals were performed in accordance with the guidelines of the National Institutes of Health, as approved by the Animal Care and Use Committee of the University of Wisconsin-Madison. Lentivirus constructs and infection of chromaffin cells. GFP was fused to the N-terminus of the cytoplasmic domain of syt I, C2AB (residues ) and subcloned into the bicistronic lentiviral vector, plox Syn-DsRed-Syn-GFP lenti-viral vector (provided by Dr. Francisco Gomez-Scholl (Serville, Spain)). Because cultured chromaffin cells are viable for relatively short periods of time (5 days) and protein expression using the synapsin promoter in plox is slow, we replaced the synapsin promoter, via XbaI and EcoRI restriction sites, with a CMV promoter which results in faster protein expression (2 days). The GFP tag was used to identify infected chromaffin cells for recordings. Lentiviral particles were generated by transfecting HEK293T cells with the modified lentiviral construct plus two other packaging vectors encoding VSV-G and Δ8.9. The supernatant was collected after h, purified by filtration through a 0.45 µm filter and centrifuged at 70,000g for 2 h to concentrate the virus. Viral particles were resuspended in phosphate-buffered saline and the titer determined. For overexpression of GFP-C2AB, cells were plated on poly-lysine coated cover slips, and infected with virus for 3 days. Electrophysiological recordings were carried out between DIV3-4. Ca 2+ -uncaging and [Ca 2+ ] i measurement. Homogenous global elevation of [Ca 2+ ] i was achieved by photolysis of the caged-ca 2+ compound, nitrophenyl-ethylene glycol-bis (β-aminoethyl ether)- N,N,N,N -tetraacetic acid (NP-EGTA, Molecular Probes, Carlsbad, CA), with a UV light source

107 97 as previously described (Xu et al., 1997). In brief, step-like elevations of [Ca 2+ ] i were elicited via a UV flash generated from a Rapp flash lamp (Rapp Optoelektronik, Hamburg, Germany). The flash was followed by excitation, via a Till Polychrome V monochromator (Till Photonics GmnH, Grafelfing, Germany), that alternated between 350 and 380 nm, allowing for ratiometric determination of the Ca 2+ concentration according to the equation (Grynkiewicz et al., 1985): [Ca 2+ ] i = K eff (R R min )/(R max R), where K eff, R min, and R max are constants obtained from in vivo calibration. R was calculated as F 350 /F 380 after background subtraction. Fluorescence signals were monitored using a photodiode detector (Till Photonics GmnH, Grafelfing, Germany). The NP-EGTA pipette solution contained (in mm): 110 CsCl, 5 NP-EGTA, 2 NaCl, 4 CaCl 2, 2 MgATP, 0.3 GTP, 0.2 Fura-6F, and 35 HEPES, adjusted to ph 7.2 using CsOH or HCl (osmolarity, 300 mosm). The free Ca 2+ concentration in the pipette solution was determined to be ~200 nm. Membrane capacitance (C m ) measurements. The C m of chromaffin cells was monitored in real time using an EPC10 Double amplifier (HEKA Electronics, Lambrecht, Germany) using a conventional whole-cell patch clamp configuration. A sine + dc protocol was applied using the Lockin extension of the Pulse program (HEKA Electronics). Chromaffin cells were voltage clamped at a holding potential of 70 mv and a sine wave voltage command (20 mv, 977 Hz) was applied. Currents were filtered at 2.9 khz and sampled at 15.6 khz. The bath solution contained (in mm): 140 NaCl, 2.5 KCl, 1.3 CaCl 2, 1 MgCl 2, 10 HEPES, and 10 Glucose (adjusted to ph 7.4 with NaOH, 308 mosm). Vesicle aggregation assay. To monitor vesicle aggregation, the optical density of samples was measured at 405 nm using an Eppendorf Biophotometer Plus (Eppendorf, NY). 5 µl Tr vesicles were incubated with either 5 µl Vr-syt vesicles, or Vr vesicles plus C2AB (1 µm), in the presence

108 98 of 0.2 mm EGTA for 20 min. Then, Ca 2+ (1 mm) was injected and aggregation was monitored for 40 min. EGTA was added at the end of each run to assay for reversibility. Data analysis for membrane capacitance. Data analysis was carried out using IGOR Pro 5.05 software (WaveMetrics, Lake Oswego, OR). Statistical significance was evaluated using the Kruskal-Wallis test for multiple comparisons of groups with non-normal distributions. Off-line analysis of [Ca 2+ ] i data was performed by measuring the fluorescence intensity from individual chromaffin cells; data were analyzed using IGOR Pro 5.05 software. [Ca 2+ ] i was calculated from the equation derived by Grynkiewicz et al. (Grynkiewicz et al., 1985) as detailed above. For Cm responses in flash photolysis experiments, the size and release rate of three distinct release components: the readily releasable pool (RRP), the slowly releasable pool (SRP), and the sustained release of vesicles, were determined as described in previous reports (Sorensen et al., 2003; Voets, 2000). In brief, a triple exponential function was used to fit the Cm responses Where A 0 is the capacitance of the cell before flash, and t 0 is the time of flash. The amplitudes (A i ) and time constants ( i) of the two faster exponentials define the size and release kinetics of the fast and the slow exocytotic burst, respectively. The third exponential represents the sustained component. Bead pull-down docking assay. Avidin beads (Thermo scientific, IL) were first blocked with protein free liposome (PS:PC:PE/15%:55%:30%) at 4 ºC overnight. Beads then washed with fusion assay buffer three times, and 40 µl of bead slurry was incubated with 10 µl of Tr vesicles - that bear biotin-pe - at room temperature for 15 min. The samples were then washed with with the same buffer and incubated with Vr vesicles, that did and did not harbor reconstituted

109 99 proteins (i.e. FL syt and syb, as indicated in Fig. 3-8C) labeled with Rh-PE (1.5%) at 4 ºC for 30 min. After washing three times with fusion buffer, 50 µl of 2.5% n-dodecyl beta-d-maltoside (Sigma, St. Louis) was added to solubilize the Rh-PE. Beads were removed by centrifugation (4600 rpm in a microfuge for 1 min), and the Rh signal in the supernatant was measured using a plate reader. Immunostaining. Chromaffin cells were co-stained with a mouse monoclonal antibody directed against the C2AB domain of syt I (41.1, provided by Reinhard Jahn) and, to visualize LDCVs, a rabbit polyclonal antibody directed against chromogranin B (CgB). The cells were fixed in PBS with 4% paraformaldehyde, permeabilized and blocked with 0.1% Triton X-100 plus 10% goat serum, stained with primary antibodies for 2 hours, washed with PBS three times, and then incubated with either Cy3-tagged anti-mouse or Alexa 647-tagged anti-rabbit (Jackson ImmunoResearch Laboratories) secondary antibodies for 1 hour at room temperature. Coverslips were then mounted in Fluoromount (Southern Biotechnology Associates) and images were acquired on an Olympus FluoView 1000 upright confocal microscope with a 60x 1.10 numerical aperture water-immersion lens.

110 ) Figures Figure 3-1

111 101 Figure 3-1. Reconstitution of active FL syt. (A) Schematic diagram of the in vitro FL sytregulated fusion assay. (B) Titration of PIP 2 in t-snare vesicles (Tr). Vesicles bearing FL syt and syb (Vr-syt) were preincubated with t-snare vesicles (Tr), that harbored the indicated amount of PIP 2, for 20 min in the presence of 0.2 mm EGTA before addition of 1 mm Ca 2+. Fusion was monitored for another 60 min after Ca 2+ injection. (C, D) Increasing [PIP 2 ] resulted in increases in the extent (C) and initial rate (D) of regulated fusion. (E) In reactions lacking FL syt, membrane fusion was only slightly promoted by Ca 2+, and this small effect occurred when the PIP 2 levels were > 1 %. Experiments for panels B-E were repeated twice and similar results were obtained. (F) Optimization of the pre-incubation time. Fusion was monitored as in (B). The duration of the pre-incubation step was varied as indicated. (G, H) the extent (G) and the initial rate (H) of regulated fusion were plotted versus the preincubation time (n = 4).

112 Figure

113 103 Figure 3-2. Phosphatidylinositol bisphosphate specificity for syt regulated fusion (A) Fusion assays were carried out as described in Fig. 3-1B, but in the presence of the indicated [neomycin]; (B) quantification of the data from panel (A). (C,D) The same experiments were carried out as in panels (A) and (B), but using fusion assays in which FL syt was replaced by 3 µm C2AB. (E) Tr vesicles containing three different phosphatidylinositol bisphosphates were mixed with Vr-syt vesicles, and fusion was monitored as in Fig. 3-1 B. (F, G) The extent (F) and initial rate (G) of regulated fusion were calculated and plotted. All experiments were carried out in > 3 independent trials.

114 Figure

115 105 Figure 3-3. Physiological densities of FL syt regulate fusion via a SNARE-dependent mechanism. (A) Titration of reconstituted FL syt in v-snare vesicles (Vr) alters the rate and extent of Ca 2+ -promoted membrane fusion. The experiments were carried out as described as Fig. 3-1 B; the average copy number of syt per vesicle is indicated. (B) Samples from (A) were subjected to SDS-PAGE and stained with Coomassie blue. (C) Ca 2+ dose-response from fusion assays containing thirty copies of syt in Vr vesicles and 3% PIP 2 in Tr vesicles. The corrected data were fitted with a sigmoidal curve; the [Ca 2+ ] 1/2 was 250 µm, and the Hill coefficient was 1.5. Experiments were repeated twice and similar results were obtained. (D-F) The cytoplasmic domains (cd) of SNAREs block syt-promoted fusion. (D, E) Increasing concentrations of cd t- SNAREs (D) or cd syb (E) were preincubated with Tr and Vr-syt. Membrane fusion was monitored using the same protocol as shown in Fig. 3-1 B. (F) The final extent of fusion at 80 min was plotted as a function of [cd SNARE] (n=3).

116 Figure

117 107 Figure 3-4. Topological requirements for PS, PIP 2, and syt during reconstituted membrane fusion. (A, B) PS (A) or PIP 2 (3%) (B) was reconstituted into either Vr vesicles with thirty copies of syt, or Tr vesicles. In panel (A), all Tr vesicles contained 3% PIP 2 ; in panel (B), all Tr vesicles contained 25% PS and all Vr contained 15% PS. Fusion assays were carried out as described in Fig 1 B. Both PS and PIP 2 are required in the target, but not the vesicle, membrane. (C) Thirty copies of syt were reconstituted into either Vr or Tr vesicles; syt must be present on Vr in order to stimulate fusion. Representative examples from > 3 independent trials are shown.

118 Figure

119 109 Figure 3-5. Syt loss-of-function mutations that impair synaptic transmission also disrupt syt function in vitro. (A) Schematic diagram of syt mutants; each mutant has been previously characterized in intact synapses. (B-D) Fusion assays, using Vr vesicles harboring each mutant shown in panel (A), were carried out as described in Fig. 3-1 B. Representative examples from > 3 independent trials are shown.

120 Figure

121 111 Figure 3-6. C2AB exhibits context-dependent mixed antagonist/agonist activity. (A) C2AB was titrated into fusion assays that contained FL syt in Vr vesicles. C2AB inhibited FL syt and Ca 2+ promoted fusion at relatively low concentrations (< 1 µm), but promoted fusion at relatively high concentrations (> 1 µm). (B, C) the extent (B) and initial rate (C) of fusion (only the Ca 2+ dependent component was analyzed) were plotted versus [C2AB] (n=3). (D) Average membrane capacitance (C m ) traces of WT and syt I KO chromaffin cells, that did and did not over-express GFP-C2AB, in response to a step-like elevation of [Ca 2+ ] i generated by flash photolysis of caged Ca 2+ (arrow indicates the flash). (E, F) The amplitudes (the size of the readily releasable pool (RRP)) and rates of the fast release component, were plotted for each condition (n > 10). C2AB inhibited release in the presence of endogenous syt, but partially rescued fast release in the absence of native syt.

122 Figure

123 113 Figure 3-7. PIP 2 and a preincubation step are required for FL syt-regulated fusion, but are not need for C2AB-regulated fusion activity. Fusion assays were carried out as described in Fig. 3-1 B, with modifications. FL syt was used in all experiments shown in the left panels; C2AB was used in all experiments shown in the right panels. Assays were carried out with or without PIP 2 (A), with or without 30% PE (B), and with or without a 20 min pre-incubation step in EGTA prior to the Ca 2+ trigger (C). Other than these specific omissions, all other components remained constant. Injection of Ca 2+ is indicated by arrows (grey arrow: Ca 2+ injected at 0 min; black arrow: Ca 2+ injected at 20 min).

124 Figure

125 115 Figure 3-8. FL syt and C2AB promote fusion via different mechanisms. (A and B) FL syt, but not C2AB, mediates rapid Ca 2+ -independent docking/aggregation of SNARE bearing vesicles. Vesicle aggregation (OD 405 ) was monitored as a function of time; injection of Ca 2+ or EGTA is indicated by arrows. (C) Illustration of the bead pull-down docking assay. (D) Direct measure of v- and t-snare vesicle docking. t-snare vesicles were tethered to beads and used to pull down Rh-labeled v-snare vesicles that harbored either FL syt, syb, or both proteins together. Protein free Rh-labeled vesicles were used as a control; signals obtained under this condition were used correct for background binding, so the plotted data report only specific binding (n = 3). (E) Model summarizing the mechanism of action of FL syt versus C2AB.

126 116 Supplemental Figures Figure S3-1.

127 117 Figure S3-1. Syt promotes fusion of both the inner and outer leaflets of SNARE-bearing liposomes. To determine whether Ca 2+ syt promoted full fusion or hemi-fusion, dithionite quenching experiments were carried out as previously described (Bhalla et al., 2006b). Dithionite (10 mm) was incubated with Vr-syt to destroy fluorophores on the outer leaflet, thereby revealing the fusion signal for the inner leaflet. Fusion of untreated and dithionite treated vesicles was monitored using the protocol described in Fig. 3-1 B. Approximately 60% of the fusion signal, at all stages of the fusion reaction, was quenched by dithionite, demonstrating that Ca 2+ syt promotes full fusion (i.e. fusion of both the outer and inner leaflets).

128 Figure S

129 119 Figure S3-2. Ca 2+ sensitivity of syt-promoted fusion. (A) Vr-syt vesicles were incubated with Tr vesicles for 20 min followed by injection of Ca 2+ ; the final [Ca 2+ ] in each reaction is indicated. Membrane fusion was then monitored for 60 min as described in Fig. 3-1 B. (B) As a control, fusion reactions were carried out in parallel using Vr vesicles lacking syt. (C) The final extent of fusion extent from (A) and (B) was plotted against the [Ca 2+ ]. Circles, fusion between Vr-syt and Tr; squares, fusion between Vr and Tr; triangles, difference in the extent of fusion using Vr-syt versus Vr vesicles (these data are re-plotted in Fig. 3-3 C).

130 Figure S

131 121 Figure S3-3. C2AB, covalently linked to phosphoplipids, mimics FL syt during regulated membrane fusion. (A) Vr vesicles bearing C2AB that was conjugated to the headgroup of PE and incorporated into vesicles, were mixed with Tr vesicles containing PIP 2. Fusion was monitored as described in Fig. 3-1B. (B) Samples from (A) were subjected to SDS-PAGE and stained with Coomassie blue. (C) The extent of fusion was plotted versus protein copy number. Conjugated C2AB simulated fusion in a manner comparable to FL syt.

132 Figure S

133 123 Figure S3-4. (A) C2AB was incubated with Vr vesicles, that lacked syt, and Tr vesicles that harbored 3% PIP 2. C2AB alone stimulated fusion; so, in contrast to Fig. 3-6 A-B, in the absence of full-length syt, inhibition was not observed. (B) The extent of fusion from (A) was plotted versus [C2AB] (n=3). (C) Syt KO primary chromaffin cells, that over-expressed GFP-C2AB, were analyzed by immunocytochemistry. Antibodies against GFP and syt were used to confirm the expression of exogenous GFP-C2AB; large dense core vesicles were stained using as antibody against chromagranin B (CgB). (D-G) The sizes and rates of the slow release component (SRP) and sustained release component are plotted for each of the indicated conditions (n>10); data regarding the readily releasable pool (RRP) are shown in Fig. 3-6 D-F.

134 Figure S

135 125 Figure S3-5. Systematic comparison of FL syt-regulated fusion versus C2AB-regulated fusion. (A-C) The data (only the Ca 2+ dependent component was analyzed) from Fig. 3-7 was quantified and plotted (n=3).

136 ) Appendix 1) We thank the Chapman lab, E. Smith, J. Weisshaar, and J. Audhya for discussions and critical comments regarding this manuscript, and M. Dong for the modified plox vector used in this study. This work was supported by a grant from the NIH (MH to E.R.C.). E.R.C. is an Investigator of the Howard Hughes Medical Institute. 2) This work has been published at Journal of cell biology. 3) My contributions to this work are: Fig 3-1 to Fig 3-7; Fig 3-8 A, B, E; Fig S3-1 to S3-3; Fig S3-4 A, B; Fig S3-5.

137 127 Chapter IV Defining the roles of reconstituted FL syt 1 in three discrete steps in the secretory pathway

138 ) Summary We have recently reconstituted active FL syt 1 into SNARE mediated membrane fusion reactions, and our data indicate that, in vitro, this protein plays crucial roles in three steps involved in secretion: vesicle docking, priming, and Ca 2+ -triggered fusion(wang et al., 2011). A major goal is to determine whether each of these steps is associated with changes in the interactions between SNARE proteins and changes in their structures (i.e. the assembly of SNARE motifs into helices) via transition metal FRET, cerulean/citrine fluorescent protein FRET, and pyrene excimer fluorescence; our preliminary data establish feasibility of the FRET experiments. In the long term, a number of additional regulatory factors (e.g. complexin, Munc13, nsec1) will be added back to this system, and their impact on docking/priming/fusion, and SNARE complex assembly, will also be assessed. A novel aspect of this study will be to utilize bilayer-coated glass beads of different diameters to discern the relationship between starting membrane curvature and the rate of membrane fusion. For these experiments the docking and fusion of small unilamellar (SUV) v- SNARE vesicles with SNARE/bilayer-coated beads, that range from 150 nm to >7 µm (roughly the size of a typical cell), will be utilized; in effect this system can be configured to represent inside-out cells. This bead-approach will also enable single vesicle studies of docking and fusion using fluorescence microscopy, making it possible to resolve docking and fusion, optically(liu et al., 2005).

139 ) Introduction The molecular mechanism that underlies regulated SV membrane fusion remains unknown. Elucidation of this mechanism can be directly addressed through in vitro fusion assays using reconstituted SNARE proteins. SNAREs form the core of a conserved membrane fusion complex, with vesicle SNAREs (the v-snare, synaptobrevin (syb)) binding to target membrane SNAREs (the t-snares, syntaxin (syx) and SNAP-25), thereby pulling the membranes together to catalyze fusion(weber et al., 1998). This system has been used to study accessory proteins that regulate fusion, including the Ca 2+ sensor for rapid exocytosis, syt1. Syt1 is anchored to SVs via a single membrane-spanning domain. To simplify the study of syt1, most laboratories make use of the cytoplasmic domain of protein (which harbors both Ca 2+ sensing motifs, C2A and C2B, and is therefore designated C2AB)(Chicka et al., 2008b; Gaffaney et al., 2008; Schaub et al., 2006; Stein et al., 2007; Tucker et al., 2004; Xue et al., 2008). Recent studies have attempted to address the impact of membrane-embedded FL syt1 on fusion in vitro, but with conflicting results. We have addressed this issue by recently reporting a FL syt1-regulated membrane fusion system that more accurately recapitulates a number of fundamental aspects syt1-regulated exocytosis at synapses(wang et al., 2011); this system will be used to directly address mechanistic questions regarding the fusion pathway. To determine whether synaptotagmin enhances membrane fusion by facilitating SNARE complex formation, we will use a fluorescent reporter system that can precisely monitor sub-domain structural changes of SNARE proteins by using transitional metal ion FRET. Furthermore, we will investigate whether the membrane bending ability of synaptotagmin could contribute to membrane fusion, by utilizing different diameters of lipids-coated glass beads.

140 ) Preliminary Data, Experimental Procedures, Proposed experiments Rate of fusion, with and without FL syt1, as a function of initial membrane curvature. We have recently shown that syt1 can lower the energy barrier for fusion by bending the target membrane(hui et al., 2009), as predicted from theoretical studies(monck and Fernandez, 1994; Zimmerberg et al., 2006). However, little empirical data are available regarding the relationship between membrane curvature and membrane fusion; at present, there is no method to generate SNARE-bearing vesicles with graded curvatures, we can only build small (~50 nm) or giant unilamellar (several um) vesicles (SUVs and GUVs, respectively), but little in-between. This is because protein-bearing vesicles cannot be prepared by extrusion; moreover, post insertion of SNAREs into extruded vesicles of different diameters results in a shift of vesicle size into the SUV range. Here, we will use a new innovation to generate target membranes with known uniform curvature. Namely, we will use glass beads of precise dimensions ranging from 0.15 to 7.27 µm (obtained from Bangs Laboratories, Fisher, IN), and coat them with t-snare bilayers. Since this method is new, we provide a detailed protocol here: for each prep, combine 6 µl of a 10% bead slurry with 52 µl of buffer (20 mm HEPES, ph 7.5, 150 mm NaCl); bath sonicate for up to 5 minutes to reduce clumping. Add 12 µl of 4 mm SUVs, mix gently, sonicate for 30 seconds, rotate for 15 minutes to maintain dispersion of beads, add 750 µl MilliQ water, then spin at 200 rpm for 1 min at rt in micro-centrifuge. Remove supernatant and add 750 µl buffer or MilliQ water, resuspend, repeat 5x, then resuspend in a final volume of 200 µl of buffer (should correspond to ~ 5 x 10 8 beads/ml) and bath sonicate for up to 5 minutes before use. Small

141 131 perturbations in the bilayer - due to syt1 mediated bending - would occur on the nm-scale and should not be impaired by the underlying bead; this will be confirmed via EM as described(hui et al., 2009). For the t-snare bead vesicles, the lipid composition will be: 25% PS, 26% PE, 45% PC, and 4% PEG-PE (note: inclusion of PEG-PE, which is PEG2000 covalently bound to the head-group of PE, has been shown to render fusion assays, in which bilayers are formed on glass surfaces, dependent on SNAP-25, and not just on syx and syb, as detailed in ref. (Karatekin et al., 2010)). For v-snare SUVs, the lipid composition will be: 15% PS, 27% PE, 55% PC, 1.5% NBD-PE, 1.5% Rh-PE. We have used this protocol with success to prepare glass beads that are coated with lipid bilayers plus SNARE proteins, and preliminary FRAP experiments indicate that the lipids and SNAREs are mobile (Fig. 4-1A) and that these bead-supported bilayers readily fuse with incoming v-snare SUVs in a manner that requires SNAP-25 and that is stimulated by Ca 2+ and syt1 (Fig. 4-1B,C; the low level of fusion observed with syx alone will be addressed by finetuning the PEG-PE content (ref. (Karatekin et al., 2010) and data not shown). We predict that in fusion assays using SNAREs alone, there will be a relatively steep relationship between starting curvature and the rate of fusion, but in the presence of reconstituted FL syt1, the curvature dependence will be reduced due to the ability of syt1 to bend membranes to lower the energy barrier for fusion. Our preliminary studies have begun to reveal the curvature-dependence for the rate of fusion using SNAREs alone (Fig. 4-1D,E). We note that the FL syt1-regulated fusion system described here does not yet recapitulate the ability of C2AB to clamp fusion prior to the Ca 2+ signal(chicka et al., 2008b). The apparent lack of clamping activity might result from weak Ca 2+ -independent trans interactions between FL syt1 on v-snare vesicles with PIP 2 and t-snares on the surface of t-snare vesicles. These weak

142 132 interactions bring v- and t-snare vesicles together (Fig. 4-2) to favor the assembly of SNARE complexes, thus driving some degree of fusion prior to the Ca 2+ trigger(hui et al., 2011). We predict that the use of glass-bead GUVs as the target membrane will raise the energy requirements for fusion (as Ca 2+ -dependent bending of the target membrane is required for efficient fusion to occur using GUVs(Hui et al., 2009)), potentially slowing - or preventing - fusion from occurring prior to the Ca 2+ trigger. Assaying for structural changes in SNARE proteins during docking, priming, and fusion. We hypothesize that the docking and priming steps that occur in FL syt1-regulated fusion reactions involve structural rearrangements among SNARE proteins. This idea will be tested using fluorescent probes to determine whether docking involves interactions between v- and t- SNAREs, potentially via zippering of the N-terminal regions of their SNARE motifs, and to determine whether the slow Ca 2+ -independent priming step involves further assembly of the SNARE motifs. We further hypothesize that in response to Ca 2+, FL syt1 drives complete zippering of the SNARE complex to accelerate fusion. Regardless of whether our hypotheses are correct, the experiments proposed in this section will shed new light on the relationship between SNARE complex assembly and the docking, priming and fusion steps in the secretory pathway, as direct measurements of these structural transitions have not been carried out. In one approach we will use GFP variants to measure SNARE complex assembly; we have fused cerulean (the donor; excited at 433 nm) to the N-terminus of syb 2 and citrine (the acceptor; emission peak = 529 nm) to the N-terminus of SNAP-25 (note: R. Zucker at Berkeley has shown that these constructs are fully active in neurons and has used them to measure dynamic

143 133 changes in FRET during secretion in live cells; it will be of interest to compare our in vitro data with his cell-based findings). We have also generated a parallel set of labeled SNAREs, but with the fluorescent proteins at the C-termini. The goal of these experiments is to test the hypothesis that docking involves interactions between the N-termini of syb and SNAP-25, but that interactions between the C-termini of these proteins do not occur until during or after fusion, when SNAREs form cis-complexes. To establish feasibility of these experiments, we incubated v- and t-snare vesicles, with N-terminal tags on syb and SNAP-25, together at 4 C, conditions that allow SNARE mediated docking, but not fusion (Fig. 4-3A)(Weber et al., 1998; Wu, 2012). As shown in Fig. 4-3B, N-terminal FRET was readily detected in these samples, indicating that during the low temperature incubation step the N-termini of v- and t-snares interact with oneanother. The next goal will be to time-resolve changes in FRET using both N-terminal and C- terminal FRET reporters, again to test a model in which N-terminal FRET occur before the Ca 2+ trigger, while C-terminal FRET occurs upon addition of Ca 2+. We have also confirmed that the fluorescent tags (citrine and cerulean) used in Fig. 4-3A,B do not disrupt function and can be readily used to study membrane fusion in vitro (Fig. 4-3C). We note that in order to measure fusion using the lipidic probes (i.e. FRET from NBD-PE to Rh-PE), it was necessary to monitor changes in the sensitized emission of the acceptor, Rh-PE, rather than changes in the donor, NBD-PE (which is what is typically done in this assay system); this is due to spectral overlap with the lipidic and protein donor/acceptor pairs. Therefore, identical samples will be run at the same time on two parallel plate readers (or one after another in the same plate reader), rather than monitoring the lipidic and protein probes in the same wells at the same time. In parallel we will use pyrene excimer fluorescence(panda and Bhattacharyya, 1992; Sahoo et al., 2000; Santra et al., 2006) to monitor, with better spatial resolution, assembly of syb into SNARE complexes; we will focus on syb as syt1 does not form direct contacts with this

144 134 SNARE (so labeling should not interfere with interactions), and because the cytoplasmic domain of syb has a random structure until it assembles into SNARE complexes(fasshauer et al., 1997). Excimer fluorescence occurs when two pyrenes come into close proximity (e.g. when syb assembles into an helix), and simply involves labeling two cys residues with the same probe. We have begun to place pairs of cys residues, for labeling with pyrene, in syb as follows: the first pyrene will be placed at position A37 and the second pyrene will be placed at each of the following positions: D44, D51, D65, A72, T79. Thus, the pyrene pairs will be 11, 21, 42, 53 and 63 Å away from each other; this is optimal as the R o for the excimer is ~ 30 Å. Pyrene is excited at 345 nm, and the monomer emits at 375 and 395 nm, while the excimer emits at 460 nm, so the ratio of monomer to excimer can be readily determined to monitor structural changes in syb. For all of these experiments, as well as those below, we will be careful to label only the outside surface of syb using the crystal structure of the SNARE complex as a guide(sutton et al., 1998). We will also employ a third approach, transition metal FRET(Taraska et al., 2009), which has an extremely short R o (12 Å using Ni 2+, which does not bind to or affect syt1 (data not shown)) and is therefore particularly useful for fine-mapping the assembly of syb into a helix. Because of this short R o, these experiments will be iterative, beginning with placement of the FRET acceptor, mbbr (monobromobimane), at position A37 (excite at 395 nm, monitor emission at 490 nm), and the Ni 2+ binding site, composed of two histidine residues, will be placed at each of the following sites via mutagenesis: Q38/E41, D40/D44, D44/V48. FRET is monitored as a function of the distance between the metal and the mbbr, and the initial distances will be 4.4, 7.5 and 13.5 Å. In this first approximation, the maximum distance between the mbbr at position 37 and the metal binding site at position 44/48 would be 13.5 Å in a syb α-helix, so once this mapping is done we will move the mbbr down to position 48 with the metal binding sites placed further along the helix in the same fashion as stated above. The experimental scheme is illustrated in Fig. 4-4.In the

145 135 figure, a cys is placed at position 37, and his residues are placed at positions 38/41; once syb forms an -helix the two his residues come into register and allow Ni 2+ to bind, thus resulting in FRET with the covalently bound FRET acceptor, mbbr, which in turn results in quenching of the mbbr.feasibility is established using the cytoplasmic domain (cd) of SNARE proteins in Fig. 4-4B,C; FRET does not occur unless cd-syb is folded into SNARE complexes by addition of the cytoplasmic domain of the syntaxin/snap-25 heterodimer, denoted cd-syx/sp25. We then fixed the [Ni 2+ ] at 5 µm and added-back Ca 2+ plus the cytoplasmic domain of syt1; these experiments clearly revealed that even in the absence of membranes, Ca 2+ syt1 can significantly accelerate the assembly of syb (residues 37-41) into an -helix upon assembly into SNARE complexes. We predict that the effect of Ca 2+ and syt1 on the folding of SNAREs will be greatly enhanced by acidic phospholipids(bhalla et al., 2006b). Again, the goal of all three approaches is to test the hypothesis that docking involves only partial assembly of syb into four-helix bundles, priming involves further assembly, and fusion involves the assembly of the C-terminal portion of the syb SNARE motif into a helix(stein et al., 2009). If our earlier study of docking in PC12 cells is correct(wu, 2012), we predict that there will be a break point in the structure of syb during docking, and these fluorescence mapping studies should reveal that break-point. Also, experiments using all three approaches outlined above will be conducted in stages: the first stage will be to determine whether we can assemble SNARE complexes in vitro using the cytoplasmic domains of all proteins. This will reveal whether our reporters work (i.e. Fig. 4-4). The second stage will be to use reconstituted syb, but to add-back cytoplasmic domains of t-snares and syt1 to determine whether Ca 2+ and C2AB can drive assembly of ternary SNARE complexes, as shown in our earlier papers using only biochemical methods(bhalla et al., 2006b). Finally, the last stage will be to use reconstituted t- SNARE vesicles and v-snare vesicles that harbor FL syt1.

146 ) Discussion For the fluorescence experiments, one concern stems from the fact that for a given vesicle, only a small number of SNAREs are needed for fusion (estimates in the literature range from 1-8 SNAREs per fusion event(domanska et al., 2009; van den Bogaart et al., 2010)), and these act in a background of a considerable number of labeled bystander SNAREs. So, once fusion occurs, bystander SNAREs can come together to form SNARE complexes, potentially drowning out the signal of interest. We are aware of this problem, and have begun to address it by empirically determining the minimal number of SNAREs per vesicle in fusion reactions; our preliminary data (not shown) suggest this corresponds to ~12 SNAREs per vesicle, thus helping to mitigate this concern (i.e. by using only 12 copies, we can decrease the background signals from bystanders). Second, we reiterate that the relevant signal will occur before the assembly of bystander SNAREs into cis SNARE complexes; by time resolving our signals - via stopped-flow rapid mixing (which can be used to monitor fusion (data not shown)) - we should be able to dissociate these events. Finally, we can scrub out bystander SNAREs by allowing docking to occur (but not fusion) at low temperature and using the light chain of botulinum toxin D to remove free syb, as described previously(weber et al., 1998).

147 ) Figures Fig. 4-1

148 138 Fig Using glass bead GUVs to study the curvature dependence of SNARE-mediated membrane fusion reactions. A) GUVs, formed over glass bead templates using fluorophorelabeled lipids or SNAREs (as indicated in the figure), were bleached with a laser (boxed area); recovery of the fluorescence signal in the bleached area was monitored over time. B) schematic diagram of the bead-guv/suv fusion assay. C) GUV vesicles (Tr), bearing either t-snare heterodimer (syntaxin and SNAP-25 are abbreviated as Syx r /SP25) or syntaxin alone (Syx r ), were preincubated with v-snare SUVs (Vr) plus 3 µm of the cytoplasmic domain of syt1 for 20 min in 0.2 mm EGTA; fusion was then triggered by addition of 1 mm Ca 2+ and monitored for another 120 min. D) Different sizes of t-snare glass-bead GUV vesicles were incubated with v-snare SUVs, in the absence of syt1 or Ca 2+. The sizes of the bead-guvs are indicated in the panel; fusion was monitored for 140 min. E) The rate of fusion was plotted versus the diameter of the t- SNARE glass bead GUVs.

149 Fig

150 140 Fig Reconstituted FL syt1 mediates rapid vesicle docking prior to the Ca 2+ signal. A) Tr vesicles were incubated with either Vr vesicles and the cytoplasmic domain of syt1 (C2AB), or Vr-syt vesicles, in the presence of EGTA for 20 min before addition of Ca 2+. B) The optical density (OD) of each sample was measured at 400 nm to monitor vesicle aggregation/docking. The reaction was continually monitored for another 40 min. Then, EGTA was added and the OD 400 was measured for another 20 min. The results are also summarized in the schematic diagram shown in panel (A).

151 Fig

152 142 Fig Docking and membrane fusion mediated by fluorescent SNARE fusion proteins. A) Schematic diagram of docking reactions for vesicles bearing fluorescent SNARE fusion proteins; only in the case where docking involves close interactions between v- and t-snares will FRET occur. B) Incubation of v- and t-snare vesicles together, under conditions that prevent fusion, results in FRET between N-terminal fluorescent proteins on v- and t-snare proteins. 20 µl of cerulean-syb vesicles were mixed with 20 µl of citrine-snap25/syx vesicles. Samples were incubated at 4 ºC overnight, which has been shown to effectively block fusion, but allows for putative SNARE-mediated docking reactions(wang et al., 2011; Weber et al., 1998). To probe for N-terminal interactions between v- and t-snares under these conditions, the samples were excited at 433 nm (cerulean) and the fluorescent emission spectra were collected from 448 nm to 550 nm, and normalized to the cerulean emission peak at 475 nm (the citrine emission peak is ~527 nm). As shown in the figure, an increase in the sensitized emission of the citrine acceptor occurred (indicated by the arrow), due to contact between v- and t-snares during docking. C) Citrine and cerulean-tagged SNAREs are active in the in vitro fusion assay. Briefly, 0.5 µl of vesicles bearing cerulean-syb were incubated with 4.5 µl of vesicles bearing citrine-snap25/syx, in the presence of 3 µm cytoplasmic domain of syt1 and 0.2 mm EGTA, for 20 min before addition of 1 mm Ca 2+ ; fusion was monitored for another 60 min after Ca 2+ injection. Detergent was injected at the end of each experiment to yield the FRET signals at infinite fusion. Due to spectral overlap with the fluorescent fusion proteins, changes in the sensitized emission of Rh-PE were monitored (excitation, 460 nm; emission, 620 nm), as opposed to the usual method of monitoring changes in the de-quenching of the NBD-PE donor, due to loss of FRET, during fusion.

153 Fig