This Article Abstract Full Text (PDF) Supplemental Material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by García, R. A. Articles by Godwin, H. A. Search for Related Content PubMed PubMed Citation Articles by García, R. A. Articles by Godwin, H. A.

High Metal Concentrations Are Required for Self-Association of Synaptotagmin II

Ricardo A. García ^* and Hilary Arnold Godwin ^* 

^* Department of Biochemistry, Molecular Biology, and Cell Biology, and Department of Chemistry, Northwestern University, Evanston, Illinois

Correspondence: Address reprint requests to Hilary Arnold Godwin, Dept. of Chemistry, Northwestern University, 2145 Sheridan Rd., Evanston IL 60208-3113. Tel.: 847-467-3543; Fax: 847-491-7713; E-mail: h-godwin{at}northwestern.edu.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

 
Several members of the synaptotagmin (syt) family of vesicle proteins have been proposed to act as Ca²⁺ sensors on synaptic vesicles. The mechanism by which calcium activates this class of proteins has been the subject of controversy, yet relatively few detailed biophysical studies have been reported on how isoforms other than syt I respond to divalent metal ions. Here, we report a series of studies on the response of syt II to a wide range of metal ions. Analytical ultracentrifugation studies demonstrate that Ca²⁺ induces protein dimerization upon exposure to 5 mM Ca²⁺. Whereas Ba²⁺, Mg²⁺, or Sr²⁺ do not potentiate self-association as strongly as Ca²⁺, Pb²⁺ triggers self-association of syt II at concentrations as low as 10 µM. Partial proteolysis studies suggest that the various divalent metals cause different changes in the conformation of the protein. The high calcium concentrations required for self-association of syt II suggest that the oligomerized state of this protein is not a critical intermediate in vesicle fusion; however, low-affinity calcium sites on syt II may play a critical role in buffering calcium at the presynaptic active zone. In addition, the high propensity of lead to oligomerize syt II offers a possible molecular explanation for how lead interferes with calcium-evoked neurotransmitter release.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

 
Calcium-triggered release of neurotransmitter from synaptic vesicles is a tightly regulated process involving a large number of proteins on the synaptic vesicle surface (Brunger, 2000; Fon and Edwards, 2001; Lin and Scheller, 2000; Misura et al., 2000; Mochida, 2000). Substantial efforts have been put forth to identify the Ca²⁺ sensor(s) that mediate activation and fusion of synaptic vesicles with the presynaptic membrane. Strong evidence points to the synaptotagmin (syt) proteins as the primary receptors of the Ca²⁺ signal on synaptic vesicles (Brose et al., 1992; Elferink et al., 1989; Fernandez-Chacon et al., 2001; Geppert et al., 1994; Perin et al., 1990; Reist et al., 1998). Despite numerous studies detailing the activities of the protein components of the fusion apparatus, the exact mechanism(s) by which Ca²⁺ regulates the fusogenic activity of syt remains the subject of intense debate (Damer and Creutz, 1996; Davletov et al., 1998; Davletov and Sudhof, 1994; García et al., 2000; Osborne et al., 1999; Shao et al., 1997). Possible mechanisms of Ca²⁺ activation include metal-mediated phospholipid/membrane binding/penetration (Bai et al., 2000; Chae et al., 1998; Chapman et al., 1998; Chapman and Jahn, 1994; Davis et al., 1999; Davletov et al., 1998, 1993; Li et al., 1995a; Perin et al., 1990; Schiavo et al., 1996; Shin et al., 2002; Sugita et al., 1996, 2002; Zhang et al., 1998), conformational changes (Davletov and Sudhof, 1994; García et al., 2000), binding to protein partners (Chapman et al., 1998, 1995; Leveque et al., 2000; Matos et al., 2000; Schiavo et al., 1996; Shao et al., 1997; Sugita and Sudhof, 2000; Verona et al., 2000; von Poser et al., 2000; Zhang et al., 1994), changes in electrostatics (Shao et al., 1997), phosphorylation (Verona et al., 2000), and/or oligomerization (Chapman et al., 1996, 1998; Damer and Creutz, 1996; Desai et al., 2000; Fukuda and Mikoshiba, 2000a,b; Littleton et al., 2001, 1999; Osborne et al., 1999; Sugita et al., 1996).

To date, 15 isoforms of syt have been identified in neuronal and nonneuronal tissues (Schiavo et al., 1998; Sudhof, 2002; von Poser and Sudhof, 2001; Fukuda 2003a,b), suggesting a more global role for syts in intracellular membrane transport. The observation that individual isoforms of syt are expressed at varied levels in different tissue types has led to the hypothesis that each syt isoform may exhibit a distinct Ca²⁺ response and/or perform unique functions (Littleton et al., 2001, 1999; Osborne et al., 1999; Schiavo et al., 1998; Sudhof, 2002; Sugita et al., 2002; Sugita and Sudhof, 2000). Of all of the syt family members, syt I is the best-characterized isoform. Upon binding of Ca²⁺, syt I binds to the SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) complex (Chapman et al., 1995; Shao et al., 1997; Sugita and Sudhof, 2000) at the nerve terminal and triggers fusion of synaptic vesicles with the presynaptic membrane. Syt II (Fukuda and Mikoshiba, 2000c), originally identified by Sudhof and co-workers, shares a high degree of sequence homology with syt I, but exhibits different expression patterns (Mochly-Rosen et al., 1992). As is true for syt I, syt II spans the synaptic vesicle membrane once via an N-terminal transmembrane domain and contains two cytosolic Ca²⁺-binding C2 domains (C2A and C2B). The C2 domains are each highly homologous with the C2 regulatory region of protein kinase C (Mochly-Rosen et al., 1992; Perin et al., 1990), the protein from which the term "C2 domain" was originally coined. C2 domains have subsequently been identified in over 130 proteins in signaling pathways (http://www.expasy.ch/cgi-bin/prosite-search-ac?PS50004), with functions ranging from phospholipid binding and Ca²⁺ signaling to ubiquitination (Nalefski and Falke, 1996; Rizo and Sudhof, 1998).

Synaptotagmins exhibit a high degree of specificity in binding to target molecules. The C2A and C2B domains are each responsible for distinct binding interactions in both Ca²⁺-dependent and Ca²⁺-independent manners (Sugita et al., 1996). For syts I and II, the C2A domain binds negatively charged phospholipids (particularly phosphatidylserine) at low Ca²⁺ concentrations (5–10 µM), and syntaxin (a member of the SNARE complex) at higher Ca²⁺ concentrations (>200 µM) (Chae et al., 1998; Chapman et al., 1998; Davletov and Sudhof, 1993; Kee and Scheller, 1996; Sugita and Sudhof, 2000; Sutton et al., 1995; Verona et al., 2000; Zhang et al., 1998). It seems likely that Ca²⁺ potentiates these binding events at least in part via an electrostatic switch mechanism (Davletov et al., 1998; Shao et al., 1997), where Ca²⁺ binding to syt shields the negatively charged residues within the binding regions of the C2 domains and facilitates interactions with the negatively charged binding sites of target molecules such as proteins and phospholipids (Chae et al., 1998; Chapman and Jahn, 1994; Shao et al., 1997; Zhang et al., 1998). Fluorescence resonance energy transfer studies with syt I revealed that Ca²⁺ binding triggers global changes in conformation via rearrangement of C2 domains that are driven by changes in electrostatic charge within each C2 domain (García et al., 2000). The C2A domain of syt I has also been shown to penetrate lipid bilayers upon exposure to Ca²⁺, suggesting a more direct role of C2A in catalyzing membrane fusion (Bai et al., 2000; Chapman et al., 1998). Both C2 domains are necessary for high-affinity binding of syt I to syntaxin and to SNAP-25 (a member of the SNARE complex), although significant binding is maintained in the isolated C2A domain (Chapman et al., 1996; Gerona et al., 2000). The C2B domain of syt I binds to phosphoinositides in a strictly Ca²⁺ dependent manner (Schiavo et al., 1998). Binding of C2B of syt I to the clathrin adaptor protein, AP-2 (Chapman et al., 1998; Zhang et al., 1994), and to SNAP-25 (Schiavo et al., 1998) occurs independent of Ca²⁺. The C2B domain has also been implicated in mediating protein homo- and heterodimerization of various syt isoforms upon exposure to Ca²⁺ (Chapman et al., 1996, 1998; Fukuda et al., 1999; Osborne et al., 1999; Sugita et al., 1996).

The propensity of syts to self-associate has led to the assertion that oligomerization of syt is a prerequisite for subsequent interactions with the SNARE complex, and is thus necessary for membrane fusion (Chapman et al., 1996, 1998; Damer and Creutz, 1996; Desai et al., 2000; Fukuda et al., 1999; Fukuda and Mikoshiba, 2000a, 2001; Littleton et al., 2001, 1999; Osborne et al., 1999; Sugita et al., 1996). However, for isoforms other than syt I, little is known about the metal dependence or metal specificity for triggering self-association, and no reports have been made about the extent toxic metal exposure can affect this process. The effect of toxic metals is of particular interest given the recent discovery that syt I is a target for Pb²⁺ (Bouton et al., 2001). Given the high level of sequence identity between syt isoforms, other members of the family are likely to be targets for Pb²⁺ as well. To address these issues, we have analyzed the association state of syt II in the presence of various divalent metal cations, including Pb²⁺. In addition, we have further characterized the Ca²⁺-activated state of syt II by assessing metal-induced conformational changes by partial proteolysis coupled with MALDI-TOF mass spectrometry.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

 
Preparation of recombinant syt II
The cDNA encoding mouse syt II (residues 1–422) was obtained from Dr. Mitsunori Fukuda (RIKEN, Brain Science Institute, Wako City, Saitama, Japan). The cytosolic domain of syt II (residues 104–422), C2A (residues 104–267), and C2B (residues 262–422) were polymerase chain reaction (PCR) amplified using PCR primers that contained restriction sites for BamHI (5') and EcoRI (3'). The PCR-derived DNA was digested by BamHI and EcoRI endonucleases to create compatible ends for insertion into the GST-expression vector pGEX-PKT (Sehgal et al., 2000) at these restriction sites. Successful insertion of the gene was verified by restriction analysis, DNA sequencing, and SDS-polyacrylamide gel electrophoresis (SDS-PAGE) of the protein products. The GST-fusion proteins were overexpressed in Escherichia coli BL21/DE3 cells and purified by affinity chromatography using glutathione-sepharose beads (Amersham Pharmacia, Amersham, UK). The GST moiety was removed by thrombin cleavage on the resin. Proteins were further purified by size-exclusion chromatography using a Superdex 75 column (Amersham Pharmacia). Fractions containing pure protein, as determined by SDS-PAGE, were dialyzed against 10 mM Bis-Tris, 100 mM KCl, 2 mM EDTA, pH 7.0, and then against 10 mM Bis-Tris, 100 mM KCl, pH 7.0. MALDI-TOF mass spectrometry was used to confirm that the proteins obtained were of the correct molecular mass: syt II (residues 104–422), predicted molecular weight = 36,079 mol wt, observed = 36,081 mol wt; C2A (residues 104–267), predicted molecular weight = 18,528 mol wt, observed = 18,532 mol wt; C2B (residues 262–422) predicted molecular weight = 18,313 mol wt, observed = 18,310 mol wt. The concentrations of standardized solutions of syt II (residues 104–422) were calibrated by amino acid analysis (Keck Biophysics Facility at Yale University, New Haven, CT) and used to calculate the extinction coefficient of the protein (43,980 M^-1 cm^-1). Protein concentrations of C2A and C2B were determined using the Bradford protein concentration assay (Bio-Rad, Hercules, CA).

Analytical ultracentrifugation
Sedimentation equilibrium experiments were conducted using a Beckman XLA-70 analytical ultracentrifuge (Fullerton, CA). Syt II (residues 104–422) was centrifuged at protein concentrations of 2.5 µM and 7.5 µM and at rotor speeds of 19,000 and 22,000 rpm in a four-hole Beckman An60 Ti rotor cooled to 4°C. Samples were prepared in buffer (10 mM Bis-Tris, 100 mM KCl, pH 7.0) with either no divalent metal added, or in the presence of various concentrations of CaCl₂, BaCl₂, MgCl₂, and SrCl₂ (110 µM, 550 µM, 1.1 mM, 5 mM, 10 mM). Sedimentation equilibrium experiments conducted in the presence of Pb²⁺ were prepared in buffer (10 mM Bis-Tris, pH 7.0) with either no metal added, or in the presence of various concentrations of Pb(NO₃)₂ (10 µM, 50 µM, 110 µM, 550 µM, 1.1 mM). Parallel sedimentation equilibrium experiments were also performed with C2A (residues 104–267) and C2B (residues 262–422) at concentrations of 8.5 µM and 26 µM for C2A and 5 µM and 15 µM for C2B, and at rotor speeds of 27,000 and 32,000 rpm. Samples of C2A and C2B were prepared in buffer (10 mM Bis-Tris, 100 mM KCl, pH 7.0) with either no divalent metal added, or in the presence of various concentrations of CaCl₂ (110 µM, 550 µM, 1.1 mM, 5 mM, 10 mM). Attainment of equilibrium was assessed by analysis of difference plots of successive absorbance scans. Data taken at all speeds and all protein concentrations were fit simultaneously using a global fitting procedure to single-species or associative models. Data were analyzed using WinNonlin2 1.03 (http://spin3.mcb.uconn.edu/) and Sedntrp 1.01 (ftp://alpha.bbri.org/rasmb/spin/ms_dos/sedntrp-philo/) software. For the associative model, the theoretical molecular weight of monomeric syt II was input as a known value and the association constant was varied to obtain the best fit to the data. The association constants obtained in this manner reflect the net equilibrium for both the affinity of syt II for Ca²⁺ and the affinity of Ca²⁺-syt II for itself:

In the absence of independent measures of the affinity of syt II for Ca²⁺ (K_a(Ca-syt)), it is not possible to extract the value for the self-association of syt from the affinity constant obtained from the fit, except at the highest Ca²⁺ concentrations (where all syt molecules are fully saturated with Ca²⁺). Thus, to assess more quantitatively how Ca²⁺ affects the self-association of syt II, the fraction of dimer was calculated from the apparent molecular weight obtained from the single ideal species fit (MW_observed) at each Ca²⁺ concentration:

where MW_monomer and MW_dimer are the calculated molecular weights for syt II (residues 104–422) monomer (36,079 mol wt) and dimer (72,148 mol wt), respectively.

Atomic force microscopy of syt II
Protein samples were analyzed in air with a Multimode atomic force microscope (AFM) (Digital Instruments, Santa Barbara, CA) operated in tapping mode using Digital Instruments n+ silicon nitride tapping mode probes with a force constant of C = 40–100 N/m and a resonance frequency of n_o = 300–400 kHz (García et al., 1996; Schneider et al., 1998). Samples of syt II (50–500 nM) were prepared in 20 µl of buffer (10 mM Bis-Tris, 100 mM KCl, pH7.0) with either no divalent metal added, or in the presence of 5 mM CaCl₂. Protein samples were incubated for 5 min at 20°C. Samples were deposited onto freshly cleaved ruby mica (New York Mica, New York, NY), washed with 15–20 drops of deionized water, blotted with filter paper, and blown dry under a stream of N₂ gas. All samples were imaged at a scan rate of 2.2–3.2 lines/s (512 x 512 pixels per image). The force exerted on the sample was minimized by retracting the tip as far from the surface as possible without a loss of resolution in the images. All images were acquired without online filtering.

AFM image analysis
Protein heights and diameters were measured from the unflattened images using the section analysis function of the Nanoscope III offline data analysis software (version 4.32r3; Digital Instruments). Proteins were measured at half-maximal height to compensate for the lateral broadening effects of AFM raster scanning (Schneider et al., 1998). Measurements of molecular volume of the scanned proteins were calculated from the height and diameter measurements as described previously (Schneider et al., 1998). The theoretical molecular volume of syt II was also calculated according to the method of Edstrom et al. (1990). Captured images were flattened to remove the background slope.

Partial proteolysis of syt II
Proteolysis experiments were conducted with 30 µM syt II and 4 µg of trypsin (Sigma, St. Louis, MO) in buffer (10 mM Bis-Tris, 100 mM KCl, pH 7.0) with either no divalent metal added, or in the presence of various concentrations of CaCl₂ (110 µM, 550 µM, 1.1 mM, 5 mM, 10 mM). Experiments designed to assess metal-specific conformational changes by trypsin proteolysis were conducted with either no divalent metal, or in the presence of 5 mM CaCl₂, SrCl₂, BaCl₂, MgCl₂, or 1.1 mM Pb(NO₃)₂. Samples were prepared in 20-µl volumes, gently vortexed, and incubated at 37°C for 1 h. Reactions were stopped by the addition of SDS-sample buffer (Laemmli, 1970) followed by boiling at 95°C and were analyzed by SDS-PAGE. Control experiments with 30 µM bovine serum albumin (Sigma) were conducted under identical conditions to those described for syt II with either no divalent metal, or in the presence 5 mM CaCl₂, SrCl₂, BaCl₂, or MgCl₂; or 1.1 mM Pb(NO₃)₂.

MALDI-TOF mass spectrometry of syt II
Samples of intact syt II (residues 104–422) and trypsin-digested syt II (residues 104–422) were analyzed by MALDI-TOF mass spectrometry using a Voyager DE Pro spectrometer from PE Biosystems (Foster City, CA) operated in linear mode. The MALDI matrix was prepared by dissolving 10 mg of sinapinic acid (Sigma) in 1 ml of 50% acetonitrile in water. Samples of intact syt II were prepared by adding 1 µl of protein (80 µM) to 9 µl of matrix solution. Trypsin digests of syt II in the presence of 5 mM CaCl₂, SrCl₂, BaCl₂, MgCl₂, or 1.1 mM Pb(NO₃)₂ were prepared as described in the previous section and stopped after 1 hr by boiling the samples for 5 min at 95°C. The samples were briefly centrifuged and a 4-µl aliquot of each trypsin digest was added to 11 µl of matrix solution. All samples were briefly vortexed, spotted on a 100-well stainless steel sample plate (1–2 µl of analyte per well), and allowed to air dry. MALDI-TOF mass spectra of all samples were collected in negative ion mode with delayed extraction (300 ns), and at an acceleration voltage of 25 kV. Internal and/or external mass calibrations were performed using bovine carbonic anhydrase II (MW = 28,980 mol wt) or horse heart cytochrome c (MW = 12,361 mol wt) (Sigma). Fragment masses were compared to predicted cleavage products of syt II (residues 104–422) produced by trypsin cleavage using the MS Digest program on the ProteinProspector internet site (version 3.4.1; http://prospector.ucsf.edu/).

Modeling of syt II
The coordinates of the crystal structure of syt III (Sutton et al., 1999) were downloaded from the Protein Data Bank (http://www.rcsb.org/pdb/cgi/explore.cgi?pid=25992988744610&pdbId=1DQV) and used as a scaffold for threading of the sequence comprising the cytosolic region of syt II (residues 104–422) using the Swiss-PdbViewer program (http://www.expasy.ch/spdbv/mainpage.html). The Swiss-PdbViewer was also used to align the cytosolic domains of syt II (residues 104–422) and syt III (residues 295–566) before threading the syt II amino acid sequence to the syt III scaffold.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

 
To determine the propensity of Ca²⁺ to mediate self-association of syt II, sedimentation equilibrium experiments were conducted on the full cytosolic domain of syt II (residues 104–422) in the absence of divalent metal, and in the presence of increasing concentrations of Ca²⁺. These data were analyzed to determine the apparent molecular weight of syt II using a single ideal species model in the program WinNonlin2 (see Materials and Methods). The results of analyses for syt II are provided in tabular form in the supplemental materials (see Supplementary Material). In the absence of Ca²⁺, the apparent molecular weight of syt II (37,300 mol wt) is consistent with the calculated molecular mass of the polypeptide (36,079 Da), indicating that the apo form of the protein is monomeric. A comparison of the equilibrium scans of syt II taken without divalent metal in the buffer, and in the presence of 550 µM and 5 mM Ca²⁺ reveals a dramatic shift in the absorbance profiles (Fig. 1): at higher Ca²⁺ concentrations, the absorbance increases more dramatically as a function of increasing cell radius. Fits to these data using a single ideal species model reveal that increasing the Ca²⁺ concentration results in a corresponding increase in molecular weight (see supplemental materials). The fraction of dimer observed as a function of Ca²⁺ concentration is plotted in Fig. 2. From the plot, the EC₅₀ for dimerization occurs at 1 mM Ca². Complete formation of a protein dimer (calculated MW = 72,158; observed MW = 75,100 mol wt) is observed at 5mM Ca²⁺. At the concentration at which syt II is saturated with Ca²⁺ (5 mM CaCl₂), the associative model yields a dissociation constant for the dimer of 0.5 µM.

View larger version (23K): [in this window] [in a new window]   FIGURE 1  Sedimentation equilibrium analysis of syt II. Samples of syt II (2.5 and 7.5 µM) were centrifuged at two speeds (19,000 and 22,000 rpm) as described in Materials and Methods. For clarity, only the absorbance scans taken at 22,000 rpm are shown. Absorbance scans of syt II taken at equilibrium in the absence of divalent metal (blue trace) and in the presence of 0.55 mM (green trace) or 5 mM Ca2+ (red trace) are shown in the bottom panel and the corresponding residuals are displayed above. The transition from monomer (blue trace) to dimer (red trace) correlates with the increase in Ca2+ concentration.

View larger version (16K): [in this window] [in a new window]   FIGURE 2  Comparative plot of percent syt II oligomerization state as a function of divalent metal concentration. (A) The percent dimer present at different metal concentrations was calculated from the molecular weights obtained from a single ideal species fit to the sedimentation equilibrium data. Essentially 100% dimer is present at a Ca2+ concentration (•) of 5 mM; Sr2+ () promotes dimer formation almost as well as Ca2+. By contrast, Mg2+ () and Ba2+ () are much weaker promoters of dimer formation, and even 10 mM Mg2+ or Ba2+ does not yield 100% protein dimer. (B) Pb2+ () is a more potent mediator of self-association than Ca2+ (•); complete dimer formation is seen at 1 mM Pb2+.

To identify whether a particular domain of syt II is responsible for self-association, sedimentation equilibrium experiments were also performed with isolated C2A and C2B domains and analyzed as described above using a single ideal species model. The results of these fits for C2A and C2B are given in Table S2 of the supplemental materials. Sedimentation equilibrium results (19.1–21.6 kDa) for C2A (residues 104–267) indicate that the domain is predominantly monomeric, even when exposed to increasing Ca²⁺ concentrations (see Fig. 3). Parallel experiments conducted with the isolated C2B domain (residues 262–422) indicate that C2B also exists as a monomer (17.5–19.7 kDa), irrespective of Ca²⁺ concentration (0–10 mM). These results show that neither C2A nor C2B independently has a propensity for self-association and suggest that both domains and/or an intact linker between domains are required for Ca²⁺-triggered association of C2A-C2B cytosolic region of syt II.

View larger version (15K): [in this window] [in a new window]   FIGURE 3  Association state of individual C2 domains from syt II as a function of calcium concentration. The observed molecular mass for isolated C2A (•) and C2B () are close to the calculated masses for the monomers (18.5 kDa and 18.3 kDa, respectively) and remain constant as a function of added calcium, suggesting that the isolated domains are not oligomerization competent.

View larger version (40K): [in this window] [in a new window]   FIGURE 4  AFM images of syt II in air. Individual protein molecules of syt II (A) without Ca2+ and (B) with 5 mM Ca2+ are shown. Addition of Ca2+ to the protein samples increases the size dimensions of the molecules (see Table 1). Proteins for analysis were visualized at 800–3000-nm scan sizes. For comparisons, representative images were magnified to 500 nm x 500 nm.

View larger version (28K): [in this window] [in a new window]   FIGURE 5  Calcium-dependent protease protection of syt II. Proteolysis experiments were carried out with 30 µM syt II and 2 µg of trypsin in the absence (lane 0) and presence of increasing millimolar concentrations of Ca2+ (lanes 0.11, 0.55, 1.1, 5, 10). Control lane (-T) shows syt II (denoted by the arrow) without treatment with trypsin. Protein products were analyzed by SDS-PAGE and observed by Coomassie blue staining. Molecular weight markers are shown on the left.

View larger version (22K): [in this window] [in a new window]   FIGURE 6  Metal-specific protease protection of syt II. Proteolysis experiments were carried out with 30 uM syt II and 2 µg of trypsin in the absence (-) and presence (+) of 5 mM Ca2+, Sr2+, Ba2+, Mg2+,or 1.1 mM Pb2+. Intact syt II is denoted by the arrow. Trypsin cleavage in the presence of Pb2+ gives rise to a distinct proteolytic fragment below the intact syt II band (denoted by the asterisk). Protein products were analyzed by SDS-PAGE and observed by Coomassie blue staining.

To further characterize the nature of the conformational change(s) induced by addition of divalent metal ions to syt II, molecular weights of the protein fragments derived from the trypsin digests were determined using MALDI-TOF mass spectrometry and the cleavage sites of the fragments were mapped using the program MS Digest (Table 2). The MALDI-TOF mass spectrum of syt II in the absence of both trypsin and divalent metal reveals a peak at the molecular weight of the monomeric protein (Fig. 7, top panel; Table 2). Trypsin digestion in the absence of metal or in the presence of Mg²⁺ results in two protease-resistant fragments (F1 = 17,730 ± 11 mol wt; F2 = 18,742 ± 8 mol wt). Consistent with the results seen by SDS-PAGE, no peak corresponding to intact monomer is observed in either of these cases. The masses of F1 and F2 correspond to major segments of the C2A and C2B domains, respectively (Fig. 7, second and third panels; Table 2). By contrast, when Ca²⁺ (or Ba²⁺ or Sr²⁺, data not shown) is present during the digestion, intact syt II (Fig. 7, fourth panel; Table 2) is observed in addition to the two protease-resistant fragments observed for the apo protein. In the presence of Pb²⁺, only one of the protease-resistant fragments is present (F1), along with intact syt II (Fig. 5, bottom panel; Table 2). It is not clear why the additional proteolytic fragment observed by SDS-PAGE for the trypsin digest of the Pb²⁺-exposed sample (Fig. 6; band highlighted by asterisk) is undetectable by mass spectrometry; presumably this fragment does not ionize or desorb efficiently in the mass spectrometer. Comparison of the relative intensities of the tryptic fragments (F1, F2, and intact syt II) in the Ca²⁺- and Pb²⁺-exposed samples reveal that the amount of the specific proteolytic fragments produced is dependent on the specific divalent metal present in the proteolysis reaction (Fig. 7).

View larger version (18K): [in this window] [in a new window]   FIGURE 7  Identification of major proteolytic fragments of syt II. MALDI-TOF mass spectrum of apo-syt II in the absence of trypsin (-T,-M) reveals a peak at the molecular weight of the intact monomer (36,079 mol wt), and a less intense peak corresponding to its doubly protonated form (+2). Trypsin-digested samples in the absence of divalent metal (+T,-M), and in the presence of magnesium (+T,+Mg2+), calcium (+T,+Ca2+), and lead (+T, +Pb2+) reveal the major protease-resistant polypeptide fragments (F1, F2) and the remaining intact protein (monomer'). The molecular weights of the various fragments are given in Table 2.

View larger version (29K): [in this window] [in a new window]   FIGURE 8  Modeled structure of syt II. Ribbon diagram of the modeled structure of syt II (residues 104–422) depicting the significant trypsin cleavage sites (yellow) contained within the C2A (blue) and C2B (red) domains, and also within the linker region separating C2A and C2B (green). Significant trypsin cleavage also occurs at Lys-115 (not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

It is interesting to compare these results for the cytosolic domain of syt II (residues 104–422) to those obtained previously from our laboratory (García et al., 2000) and by others (Chapman et al., 1996, 1998; Damer and Creutz, 1996; Desai et al., 2000; Fukuda and Mikoshiba, 2000b; Littleton et al., 2001, 1999; Osborne et al., 1999; Sugita et al., 1996) for syt I and other syt isoforms. In our hands, the cytosolic domain from syt I (residues 96–421) remains monomeric even when it is expressed, purified, and exposed to divalent cation, and analyzed (by analytical ultracentrifugation) in a manner identical to the conditions used herein to study syt II (García et al., 2000). Comparison of these data would suggest that there is something inherently different about oligomerization propensity between syt I and syt II, despite the high sequence identity (82%) of the cytosolic domains of the two isoforms. However, several other groups have reported that syt I—and other syt family members—oligomerize either constitutively or in the presence of Ca²⁺ (Chapman et al., 1996, 1998; Damer and Creutz, 1996; Desai et al., 2000; Fukuda and Mikoshiba, 2000b; Littleton et al., 2001, 1999; Osborne et al., 1999; Sugita et al., 1996). Furthermore, whereas some of these authors have found the oligomerization activity of syt to reside in the C2B domain (Chapman et al., 1996, 1998; Desai et al., 2000; Littleton et al., 2001), we find that C2B from syt II is monomeric even in the presence of high concentrations of Ca²⁺.

How can these conflicting data be reconciled? Several possibilities must be considered:

Specific regions of syt (e.g., the transmembrane domain, Fukuda and Mikoshiba, 2000b; C2B, Chapman et al., 1996) are required for oligomerization; the different results observed in different laboratories arise from the use of proteins of different lengths.

Specific mutations (e.g., G374 versus D374) or polymorphisms present in syt isoforms used by different members of the community determine whether syt is able to oligomerize (Davis et al., 1999).

"Contaminants" (Fernandez et al., 2001), membrane components, or posttranslational modifications (or the lack thereof; Fukuda and Mikoshiba, 2000b) are responsible for the propensity of syt to oligomerize; differences in purification protocols may account for the differences in oligomerization observed in different laboratories.

The ability of syts to oligomerize may be transient and hence lost with aging.

The data reported herein provide the following insights into these issues:

The transmembrane domain is not required for oligomerization. In addition to the data reported herein demonstrating that the cytosolic domain for syt II oligomerizes in response to Ca²⁺. Fukuda and co-workers have also observed oligomerization of cytosolic domains (or portions thereof) for various syt isoforms (Fukuda and Mikoshiba, 2000b).

Although C2B may be necessary for oligomerization, it is not sufficient. No oligomerization was observed for isolated C2B from syt II (residues 262–422).

The presence of a glycine (instead of glutamate) at position 374 (syt I numbering) is not sufficient to confer the ability to oligomerize to syt. Both syt I (residues 96–421) (García et al., 2000) and syt II (residues 104–422) studied in our laboratory have glycine at this position, yet only syt II was observed to oligomerize.

Contaminants are not solely responsible for the Ca²⁺-triggered oligomerization of syt II: all samples used herein were subjected to purification by gel filtration after isolation using affinity chromatography. No contaminants were observed by ultraviolet-visible (any samples with significant absorbance at 260 nm were rejected and not subjected to further analysis) or by mass spectrometry (syt II (residues 104–422), predicted molecular weight = 36,079 mol wt, observed = 36,081 mol wt). Clearly, parallel studies in the presence of membrane components are warranted, but these studies reveal that oligomerization occurs even in their absence.

The different behaviors that we observe for syt I and syt II, when isolated and studied under identical conditions, point to the clear need for detailed structural studies on these isoforms and mutagenesis studies that explore the role of specific residues in promoting oligomerization. Furthermore, studies on oligomerization of syt I and II under conditions that allow membrane association would also be of interest.

The studies reported herein also provide important insights into the metal dependence and specificity for self-association of syt II. Barium and magnesium are much weaker than Ca²⁺ at triggering dimer formation of syt II (Fig. 2 A). These data are consistent with oligomerization studies performed by Chapman and co-workers with GST-immobilized cytosolic syt I that showed complex formation with native syt I in the presence of Ca²⁺, but not with Ba²⁺or Mg²⁺ (Chapman et al., 1996). By contrast, syt II self-associates in response to Sr²⁺ in a manner similar to that of Ca²⁺ (Fig. 2 A; Table S1D in the supplemental materials), although slightly higher concentrations of Sr²⁺ (10 mM) are required to yield complete dimer formation. This result is somewhat surprising, given that Sr²⁺ does not trigger binding of syt II to syntaxin (Li et al., 1995b) and does not promote oligomerization of syt I (Chapman et al., 1996). Several possible explanations for these discrepancies must be considered: oligomerization of syt may not be a determinant (or the sole determinant) of activity; activation of syt II may not necessarily result in syntaxin binding; the different conditions used for different studies in different groups may account for the differences in observed behaviors; and/or syt II exhibits significantly different cation selectivities than syt I.

The ability of Pb²⁺ to trigger oligomerization of syt II is of particular interest, given recent studies that suggest syts may be a target for Pb²⁺ (Bouton et al., 2001). The propensity of Pb²⁺ to trigger dimerization of syt II at a concentration approximately one order of magnitude less than that observed with Ca²⁺ (550 µM Pb²⁺ vs. 5 mM Ca²⁺) indicates that Pb²⁺ is a more potent mediator of the self-association process than Ca²⁺. Exposure of syt II to Pb²⁺ triggered protein self-association at Pb²⁺ concentrations as low as 10 µM and yielded a multimeric species at 1.1 mM Pb²⁺. By contrast, previous studies with syt I demonstrate that Pb²⁺ is a potent substitute for Ca²⁺ in potentiating interactions between syt I and phospholipid liposomes but that Pb²⁺ interferes with the ability of syt I to bind to syntaxin (Bouton et al., 2001). Taken together, these results suggest that syts, as a class of proteins, may be important protein targets by which Pb²⁺ mediates its neurotoxicity and that the differences between how Pb²⁺ and Ca²⁺ interact with different syt isoforms may explain why Pb²⁺ inhibits the Ca²⁺-evoked step of neurotransmitter release and increases spontaneous release (Manalis et al., 1984).

Critically, the studies reported herein demonstrate that divalent metals also induce a structural change in the protein that does not necessarily occur concomitant with dimerization. In the absence of metal, significant cleavage occurs within the linker region at Lys-273, which separates the C2A and C2B domains, and within C2A (Lys-223) and C2B (Arg-389). Protection from digestion occurs upon binding of Ca²⁺ (5 mM), Ba²⁺ (5 mM), Sr²⁺ (5 mM), and Pb²⁺ (1 mM), due to a change in the structure or dynamics of syt II that renders the cleavage sites inaccessible to the protease. However, even though Mg²⁺ promotes dimer formation as well as Ba²⁺, Mg²⁺ does not give rise to the same level of protection from protease cleavage as does Ba²⁺. Instead, the cleavage pattern obtained with Mg²⁺ more closely resembles that obtained in the absence of divalent metal, suggesting that dimer formation and protease protection are not intimately coupled. The abilities of different divalent metal ions to protect syt II from proteolysis are consistent with previous studies reported for syt I (i.e., Ca²⁺, Sr²⁺, Ba²⁺, and Pb²⁺ protect from cleavage and Mg²⁺ does not; Bouton et al., 2001; Davletov and Sudhof, 1994). In addition, these results support the assertion that the recently reported structure of syt III with Mg²⁺ bound is likely to be in the "unliganded conformation" (Sutton et al., 1999).

	CONCLUSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

 
The high calcium concentrations required for self-association of syt II suggest the oligomerized state of this protein is not a critical intermediate in vesicle fusion at the synaptic cleft. However, the low-affinity calcium sites on syt II may play a critical role in buffering calcium at the presynaptic active zone (Felmy et al., 2003; Trommershèauser et al., 2003). In addition, the high propensity of lead to oligomerize syt II offers a possible molecular explanation for how lead interferes with calcium-evoked neurotransmitter release. Future studies that address the structural origins of the differences in oligomerization propensities between syt I and syt II are warranted.

	SUPPLEMENTAL MATERIALS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

 
An online supplement to this article can be found by visiting BJ Online at http://www.biophysj.org. The supplementary material contains: tables of results of sedimentation equilibrium studies on syt II in the presence of varying concentrations of Ca²⁺, Ba²⁺, Mg²⁺, Sr²⁺, and Pb²⁺; and a table of results of sedimentation equilibrium studies on C2A and C2B domains from syt II in the presence of varying concentrations of Ca²⁺.

	ACKNOWLEDGEMENTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

 
We are grateful to Dr. Mitsunori Fukuda for generously providing the cDNA for syt II, to Lydia Finney for her kind help with the mass spectrometer and related software, to Mathew Miller for his help with modeling of syt II, and to Jihong Bai, Edwin Chapman, and Jose Rizo for their helpful discussions. We are also thankful to Angie Ormonde for providing some materials for the AFM experiments and for advice on AFM. Analytical ultracentrifugation data were acquired in the Keck Biophysics Facility at Northwestern University. AFM experiments were conducted in a shared facility of the Catalysis Center at Northwestern University and MALDI-TOF experiments were performed in the Analytical Services Laboratory of the Department of Chemistry at Northwestern University.

This work was supported by a grant from the National Institutes of Health (1 R01 GM58183-01A1). Hilary Arnold Godwin is a recipient of a Camille and Henry Dreyfus New Faculty Award, a Burroughs-Wellcome Fund New Investigator Award in the Toxicological Sciences, a National Science Foundation CAREER Award, a Sloan Research Fellowship, and a Camille Dreyfus Teacher-Scholar Award. Support for Ricardo García was provided by a grant from the National Institutes of Health, Cellular and Molecular Basis of Disease Training Program (T32 GM08061) (http:x.biochem.nwu.edu/Keck/keckmain.html).

	FOOTNOTES

 
Abbreviations used: syt, synaptotagmin; C2, a calcium-binding domain homologous to the second regulatory domain from protein kinase C; C2A and C2B, the first and second calcium-binding domains in synaptotagmin II, respectively; GST, glutathione-S-transferase; SNARE, soluble N-ethylmaleimide-sensitive factor attachment protein receptor; syn, syntaxin.

Submitted on March 27, 2003; accepted for publication September 22, 2003.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION SUPPLEMENTAL MATERIALS ACKNOWLEDGEMENTS REFERENCES

 
Bai, J., C. A. Earles, J. L. Lewis, and E. R. Chapman. 2000. Membrane-embedded synaptotagmin penetrates cis or trans target membranes and clusters via a novel mechanism. J. Biol. Chem. 275:25427–25435.[Abstract/Free Full Text]

Bollmann, J. H., B. Sakmann, and J. G. Borst. 2000. Calcium sensitivity of glutamate release in a calyx-type terminal. Science. 289:953–957.[Abstract/Free Full Text]

Bouton, C. M., L. P. Frelin, C. E. Forde, H. A. Godwin, and J. Pevsner. 2001. Synaptotagmin is a molecular target for lead. J. Neurochem. 76:1724–1735.[Medline]

Brose, N., A. G. Petrenko, T. C. Sudhof, and R. Jahn. 1992. Synaptotagmin: a calcium sensor on the synaptic vesicle surface. Science. 256:1021–1025.[Abstract/Free Full Text]

Brunger, A. T. 2000. Structural insights into the molecular mechanism of Ca²⁺-dependent exocytosis. Curr. Opin. Neurobiol. 10:293–302.[Medline]

Chae, Y. K., F. Abildgaard, E. R. Chapman, and J. L. Markley. 1998. Lipid binding ridge on loops 2 and 3 of the C2A domain of synaptotagmin I as revealed by NMR spectroscopy. J. Biol. Chem. 273:25659–25663.[Abstract/Free Full Text]

Chapman, E. R., S. An, J. M. Edwardson, and R. Jahn. 1996. A novel function for the second C2 domain of synaptotagmin. J. Biol. Chem. 271:5844–5849.[Abstract/Free Full Text]

Chapman, E. R., R. C. Desai, A. F. Davis, and C. K. Tornehl. 1998. Delineation of the oligomerization, AP-2 binding, and synprint binding region of the C2B domain of synaptotagmin. J. Biol. Chem. 273:32966–32972.[Abstract/Free Full Text]

Chapman, E. R., P. I. Hanson, S. An, and R. Jahn. 1995. Ca²⁺ regulates the interaction between synaptotagmin and syntaxin 1. J. Biol. Chem. 270:23667–23671.[Abstract/Free Full Text]

Chapman, E. R., and R. Jahn. 1994. Calcium-dependent interaction of the cytoplasmic region of synaptotagmin with membranes. J. Biol. Chem. 269:5735–5741.[Abstract/Free Full Text]

Damer, S. K., and C. E. Creutz. 1996. Calcium-dependent self-association of synaptotagmin I. J. Neurochem. 67:1661–1668.[Medline]

Davis, A. F., J. Bai, D. Fasshauer, M. J. Wolowick, J. L. Lewis, and E. R. Chapman. 1999. Kinetics of synaptotagmin responses to Ca2+ and assembly with the core SNARE complex onto membranes. Neuron. 24:363–376. (Published erratum appears in Neuron 1999 24:1049).[Medline]

Davletov, B., O. Perisic, and R. L. Williams. 1998. Calcium-dependent membrane penetration is a hallmark of the C2 domain of cytosolic phospholipase A2 whereas the C2A domain of synaptotagmin binds membranes electrostatically. J. Biol. Chem. 273:19093–19096.[Abstract/Free Full Text]

Davletov, B., J. M. Sontag, Y. Hata, A. G. Petrenko, E. M. Fykse, R. Jahn, and T. C. Sudhof. 1993. Phosphorylation of synaptotagmin I by casein kinase II. J. Biol. Chem. 268:6816–6822.[Abstract/Free Full Text]

Davletov, B. A., and T. C. Sudhof. 1993. A single C2 domain from synaptotagmin I is sufficient for high affinity Ca²⁺/phospholipid binding. J. Biol. Chem. 268:26386–26390.[Abstract/Free Full Text]

Davletov, B. A., and T. C. Sudhof. 1994. Ca²⁺-dependent conformational change in synaptotagmin I. J. Biol. Chem. 269:28547–28550.[Abstract/Free Full Text]

Desai, R. C., B. Vyas, C. A. Earles, J. T. Littleton, J. A. Kowalchyck, T. F. J. Martin, and E. R. Chapman. 2000. The C2B domain of synaptotagmin is a Ca²⁺-sensing module essential for exocytosis. J. Cell Biol. 150:1125–1135.[Abstract/Free Full Text]

Edstrom, R. D., M. H. Meinke, X. Yang, R. Yang, V. Elings, and D. F. Evans. 1990. Direct visualization of phosphorylase-phosphorylase kinase complexes by scanning tunneling and atomic force microscopy. Biophys. J. 58:1437–1448.[Abstract/Free Full Text]

Elferink, L. A., W. S. Trimble, and R. H. Scheller. 1989. Two vesicle-associated membrane protein genes are differentially expressed in the rat central nervous system. J. Biol. Chem. 264:11061–11064.[Abstract/Free Full Text]

Felmy, F., E. Neher, and R. Schneggenburger. 2003. Probing the intracellular calcium sensitivity of transmitter release during synaptic facilitation. Neuron. 37:801–811.[Medline]

Fernandez, I., D. Arac, J. Ubach, S. H. Gerber, O. Shin, Y. Gao, R. G. Anderson, T. C. Sudhof, and J. Rizo. 2001. Three-dimensional structure of the synaptotagmin 1 C2B-domain: synaptotagmin 1 as a phospholipid binding machine. Neuron. 32:1057–1069.[Medline]

Fernandez-Chacon, R., A. Konigstorfer, S. H. Gerber, J. García, M. F. Matos, C. F. Stevens, N. Brose, J. Rizo, C. Rosenmund, and T. C. Sudhof. 2001. Synaptotagmin I functions as a calcium regulator of release probability. Nature. 410:41–49.[Medline]

Fon, E. A., and R. H. Edwards. 2001. Molecular mechanisms of neurotransmitter release. Muscle Nerve. 24:581–601.