MOLECULAR REGULATION OF MEMBRANE RESEALING IN 3T3 FIBROBLASTS

Transcription

1 JBC Papers in Press. Published on November 9, 2004 as Manuscript M Molecular regulation of membrane repair MOLECULAR REGULATION OF MEMBRANE RESEALING IN 3T3 FIBROBLASTS Sheldon S. Shen 1,3, Ward C. Tucker 2, Edwin R. Chapman 2, and Richard A. Steinhardt 3* From the 1 Department of Genetics, Development and Cell Biology, Iowa State University, Ames, Iowa 50011, the 2 Department of Physiology, University of Wisconsin School of Medicine, Madison, Wisconsin and the 3 Department of Molecular and Cell Biology, University of California, Berkeley, California *To whom correspondence should be addressed: Richard A. Steinhardt, Telephone: , Mail room Fax: rsteinha@socrates.berkeley.edu Running title: Molecular regulation of membrane repair Copyright 2004 by The American Society for Biochemistry and Molecular Biology, Inc.

2 Summary Membrane resealing in mammalian cells after injury depends on Ca 2+ -dependent fusion of intracellular vesicles with the plasma membrane. When cells are wounded twice, the subsequent resealing is generally faster. Physiological and biochemical studies have shown the initiation of two different repair signaling pathways, which are termed facilitated and potentiated responses. The facilitated response is dependent on the generation and recruitment of new vesicles, while the potentiated response is not. Here, we report that the two responses can be differentially defined molecularly. Using recombinant fragments of synaptobrevin-2 and synaptotagmin C2 domains we were able to dissociate the molecular requirements of vesicle exocytosis for initial membrane resealing, and the facilitated and potentiated responses. The initial resealing response was blocked by fragments of synaptobrevin-2 and the C2B domain of synaptotagmin VII. Both the facilitated and potentiated responses were also blocked by the C2B domain of synaptotagmin VII. While the initial resealing response was not blocked by the C2AB domain of synaptotagmin I or the C2A domain of synaptotagmin VII, recruitment of new vesicles for the facilitated response was inhibited. We also used Ca 2+ -binding mutant studies to show that the effects of synaptotagmins on membrane resealing are Ca 2+ dependent. The pattern of inhibition by synaptotagmin C2 fragments that we observed cannot be used to specify a vesicle compartment, such as lysosomes, in membrane repair. 2

3 Introduction Puncture of the plasma membrane invokes a rapid resealing response that depends on a source of intracellular membrane and delivery of this membrane to the plasma membrane by exocytosis to ensure cell survival 1. Repair signaling pathways are initiated by the sudden rise in cytosolic Ca 2+ due to influx through the puncture site 2. When cells are wounded twice, the second wound generally reseals more quickly. Two different signaling responses, facilitated and potentiated, have been described and depend on the site of the second wound in relation to the first. The facilitated response is characterized by a faster resealing response to a second wound at the same initial wound site and depends on the generation of new vesicles in a protein kinase C (PKC) and Golgi-dependent manner 3. Cells also have a globally increased exocytotic response to a second wound that increases the rate of repair to a second wound at a different site from the initial wound. This potentiated response does not depend on the generation of new vesicles 4. The global increase in exocytosis and repair is camp and protein kinase A (PKA) dependent, in the first few minutes, becomes dependent on protein synthesis within a few hours, and with CREB activation can last for 24 hours or more 4. The molecular mechanism mediating Ca 2+ -dependent exocytosis has been extensively investigated and appears to require assembly of soluble N- 3

4 ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) 5-7. The fusion event requires coupling of the plasma membrane-localized heterodimer target SNAREs (t-snares), syntaxin and synaptosome-associated protein of 25 kda (SNAP-25), and the vesicle-localized SNARE (v-snare), synaptobrevin 2 (syb2) 8, 9. However, fusion of purified SNAREs in an in vitro vesicle-vesicle fusion assay occurs with slow (minutes) kinetics and is Ca 2+ -independent 10. The addition of soluble synaptotagmin I (syt I) markedly enhanced both rate and extent of vesicle fusion in a Ca 2+ -dependent manner 11. Thus, synaptotagmin and SNAREs probably comprise the minimal proteins required for Ca 2+ -mediated exocytosis. Synaptotagmin (syt) is a large family of at least 15 identified isoforms that are expressed at different levels in a variety of tissues and implicated in numerous types of vesicular traffic 7, 12. All syts share the structural organization of tandem C2 domains 13. The C2 domain is a motif of approximately 130 residues, initially identified in PKC and later in more than 100 proteins 14, 15. The tandem C2 domains of syt are separated into the membrane proximal C2 domain called C2A, and the membrane distal C2 domain called C2B. Similar to its function in some PKC isoforms, the C2AB domain of syt I (C2AB1) binds Ca 2+ ions and phospholipids, and is thought to act as the Ca 2+ sensor for regulating secretion 11, 16. While C2AB domains from nearly half of the identified isoforms of syts do not bind Ca 2+ /phospholipids, nearly all do bind the t-snare 4

5 heterodimer 17 ; thus syts may trigger Ca 2+ -dependent exocytosis either as the Ca 2+ sensor or in collaboration with an additional Ca 2+ sensor. Participation of SNARE proteins in membrane repair has been suggested by the use of neurotoxins, recombinant protein fragments and inhibitory antibodies 2, 3, Membrane resealing of 3T3 fibroblast cells is inhibited by the action of clostridial neurotoxins via proteolytic cleavage of SNAREs. Cleavage of SNAP-25 by botulinum neurotoxin A 21 or syb2 by either botulinum neurotoxin B or tetanus toxin 22 inhibited 3T3 membrane resealing after micropuncture of the plasma membrane 2, 3. Recombinant syt I and VII C2 domains, and also an antibody against syt VII C2A (C2A7) domain have been tested on plasma membrane repair in NRK fibroblasts. Recombinant fragments of C2A7, as well as the antibody against this domain blocked plasma membrane repair, while recombinant C2A1 had no effect 20. Since these C2 domains also had differential effects on Ca 2+ -dependent exocytosis of secretory lysosomes 23, it has been proposed that lysosomal fusion is the sole source of membrane for membrane resealing 20, 24. An important experimental support for this proposal was that C2B7 did not block lysosome exocytosis 23 or membrane resealing 24. These results are surprising since the two C2 domains share many interactions with other molecules 16. An exclusive use of lysosomes in membrane repair would also appear to conflict with earlier evidence of the sensitivity of resealing and exocytosis to botulinum neurotoxin B and tetanus toxin and the sensitivity of the 5

6 facilitated response to brefeldin-a 3. These sensitivities are also unexpected if enlargosomes are exclusively used in membrane repair 25. Here, we have extended the use of recombinant protein domains in tests of membrane resealing of 3T3 fibroblasts with the goal of better understanding which intracellular compartments are used in repair and to find which functional properties of the C2 domains are essential for successful resealing. We focused mainly on tests of competitive inhibition using recombinant C2 domains of syts I and VII, mutant variants of these domains that are unable to bind Ca 2+, and the cytoplasmic domain of syb2 (cd-syb2). Our results support the view that the pool of membrane vesicles for membrane repair is a heterogeneous population rather than a specific sub-type, that the Ca 2+ binding properties of syt C2 domains are essential for inhibition of membrane resealing, and that different vesicle pools have different sensitivities to exogenous syt C2 domains. Experimental Procedures Cell Preparation and Reagents Swiss 3T3 fibroblasts were cultured in DMEM (Invitrogen, Carlsbad, CA) containing 8% fetal bovine serum (FBS; Atlanta Biologicals, Norcross, GA) and 50 µg/ ml normocin (Invitrogen). Cells for experiments were plated on glass coverslip-inserted plastic dishes (35 x 10 mm), and were grown for 1 to 3 days 6

7 before use. Fura-2 acetoxymethyl ester (fura-2 AM) was purchased from Molecular Probes (Eugene, OR). Preparation of Recombinant Fragments cdna encoding rat synaptotagmin I (syt I) 26 and mouse synaptotagmin VII (syt VII) 27 were kindly provided by T.C. Sudhof (University of Texas Southwestern Medical Institute, Dallas, TX) and M. Fukuda (Institute of Physical and Chemical Research, Saitama, Japan), respectively. The C2B ( ) and C2A-C2B (96-421) domains of syt I and the C2A domain (residues ) of syt VII were prepared as previously described The C2B domain (residues ) of syt VII was subcloned into pgex 2T-1 (Amersham Pharmacia Biotech, Piscataway, NJ) as previously describe 32. Ca 2+ -binding mutants of syt I C2AB (D230,232,363,365N), syt VII C2A (D225,227N) and syt VII C2B (D357,359N) domains, and a control mutant of syt VII C2B (K320,321N) domain were generated by PCR. All recombinant syt constructs were confirmed by DNA sequencing and expressed as glutathione S-transferase fusion proteins. Proteins were purified by affinity chromatography with glutathione Sepharose 4B (Amersham Pharmacia Biotech) and washed with nucleases and high salt buffer to remove bacterial contaminants as described in Wu et al. 33. Beads were washed twice more in Buffer A (12.4 mm Hepes-NaOH (ph 7.4), 138 mm NaCl, and 2.4 mm KCl) before generating soluble syt fragments by thrombin cleavage 32. The cytoplasmic domain (residues 1-94) of synaptobrevin 2 9 was expressed as His6 fusion proteins and purified by Ni-NTA agarose (Qiagen, 7

8 Valencia, CA). The proteins were step-eluted with 400 mm imidazole and dialyzed against Buffer A. All proteins were stored on ice and used within 7-10 days. Assay of Membrane Resealing Membrane resealing was monitored by measuring the fluorescence of the calcium sensitive dye fura-2 as described previously 2. Fura-2 was introduced into the cells by AM-ester loading of 1 µm Fura-2-AM at 25 C for 1hr and washed with Ringer s solution containing 1.8 mm Ca 2+. During experiments the cells were maintained in 1.8 mm Ca 2+ Ringer s solution. Ca 2+ -free Ringer s solution contained 138 mm NaCl, 2.7 mm KCl, 1.06 mm MgCl 2, 5.6 mm D- glucose and 12.4 mm HEPES (ph 7.25). A stock solution of 100 mm CaCl 2 was used to adjust the concentration of Ca 2+ to 1.8 mm in all experiments. Prior to the wounding experiments the purified recombinant fragments, which ranged in concentration from µm in Buffer A, were diluted into the Rodent Ringer s to a final concentration of 10, 20 or 30 µm, and Ca 2+ was adjusted to 1.8 mm. 10 µm recombinant syt C2 domains maximally inhibited exocytosis in PC12 cells (Tucker and Chapman, unpublished observations). Similar concentrations have previously been tested on secretory lysosome exocytosis 23 and membrane resealing after glass bead wounding 20 in NRK fibroblasts. Fura-2-loaded cells were wounded with a 0.5 micron-tipped diameter solid glass needle pulled with a PE-2 vertical puller (Narishige, Tokyo) and manipulated using an Eppendorf 5242 microinjector and 5170 manipulator (Brinkmann Instruments, Westbury, NY) 8

9 mounted on a Zeiss IM-35 inverted microscope. The time setting for wounding was 0.3 seconds. Resealing was monitored by photometric measurements of fura-2 fluorescence at nm. Wounding was marked by a persistent decrease in the Ca 2+ insensitive 357 nm excited fluorescent intensity (as an indicator of dye loss) together with an increase in the ratio of fluorescent intensity excited by 357/385 nm light (as an indicator of increased intracellular Ca 2+ activity). Membrane resealing was marked by a cessation of dye loss. The duration between wounding and cessation of dye loss is the resealing time, and the resealing rate is defined as the inverse of the resealing time in seconds. Thus a faster resealing time is reflected as a greater resealing rate. For cells that failed to reseal after the 2 nd wounding, the rate was defined as zero. In these experiments the recombinant proteins were added to the bath prior to wounding. Since diffusion of these fragments into the cell after micropuncture occurs in tens of milliseconds and membrane resealing requires tens of seconds, the addition of fragments in the bath prior to wounding should not significantly alter their effects on the kinetics of membrane resealing events. Exocytosis in cracked PC12 cells was inhibited regardless of the addition of recombinant fragments prior, during or after the addition of Ca 2+ to stimulate fusion 31. All experiments were performed at C. 9

10 Data Analysis All data are indicated as mean ± standard error of the mean (sem) and the number of observations. Statistics were calculated by Student s t-test for paired data and Mann-Whitney (nonparametric) test for unpaired data using InStat 2.00 (GraphPad Software, San Diego, CA). Results Swiss 3T3 fibroblasts rapidly reseal after puncture of the plasma membrane. To analyze membrane resealing, fura-2 loaded cells were wounded as previously described 2. Wounding (arrows in Fig. 1) by a 0.3 sec jab of a microneedle is indicated by a sharp rise in the Ca 2+ -sensitive fura-2 ratio (lower trace) and loss of fura-2 dye, which results in a decrease in the fluorescent intensity of fura-2 (upper trace) excited at a Ca 2+ -insensitive wavelength (357 nm). When the cell resealed, the decrease in fluorescence intensity stopped (bars in Fig. 1). In control 1.8 mm Ca 2+ Ringer s the average resealing time for an initial wound was ±.88 (n=51) secs or an average resealing rate of ±.004 sec -1 (Table 1). Swiss 3T3 fibroblasts are able to withstand multiple punctures and usually reseal even faster after a second wounding. The faster resealing to a second wounding at the same site is called a facilitated response 3. In the example shown in Figure 1A the resealing time is the 27 s between the initial wound (first arrow) and the cessation of fura-2 dye loss (first bar). The 10

11 second wounding (arrow) required only 16 s to reseal (bar). Each 0.3 s poke stimulated a rise in intracellular Ca 2+, which is seen as a rise in ratio of the 520 emission to excitation at 357 and 385 nm. When cells were wounded at the same site, the average initial resealing rate of 0.08 ±.006 (n=29) was significantly (P <.0001, Student s t-test) increased to a resealing rate of ±.009 for the second wound. When plotted as the rate of membrane resealing for the second wound on the Y axis to the rate of membrane resealing for the first wound on the X axis, most of the data fall above the diagonal (Fig. 1B, ). A second wounding at a different site also shows a faster resealing rate, but uses a different pathway and is called a potentiated response 4. When cells were wounded at different sites, the average initial resealing rate of ±.007 (n=22) was observed to significantly (P <.003) increase to ±.013 for the second wound. When plotted as the rate of the 2 nd membrane resealing to the rate of the 1 st membrane resealing, most of the data fall above the diagonal (Fig. 1B, ), reflecting the faster resealing response to the second wound. The average ratios of the 2 nd with respect to the 1 st resealing rates of control cells for facilitation and potentiation were 1.81 ±.16 (n=29) and 1.9 ±.14 (n=22), respectively. In nucleated animal cells wound resealing requires Ca 2+ -triggered exocytosis 1. Synaptotagmins are a family of membrane proteins proposed to regulate Ca 2+ -triggered membrane traffic to the plasma membrane 34, 35. Recombinant fragments of synaptotagmin domains would be expected to inhibit 11

12 membrane repair depending on their ability to disrupt endogenous synaptotagmin-effector interactions required for vesicle fusion with the plasma membrane. Two well-studied synaptotagmins, syt I 36 and syt VII 37, are expressed in NIH 3T3 fibroblasts, and we focused on the Ca 2+ binding domains of these two isoforms. The addition of 30 µm recombinant C2B domain of syt I (C2B1) did not inhibit the resealing rate to the first wound or the facilitation response to a second wound at the same site (Fig. 2, ). The 1 st resealing rate was ±.006 (n=16), which is not significantly (P = 0.23, Mann-Whitney, unpaired nonparametric test) changed from that of untreated control cell (0.077 ±.004, n=51). The resealing rate of C2B1-treated cells to a 2 nd wounding at the same site was observed to increase to ±.015 or a 2 nd /1 st resealing ratio of 2.06 ±.185. Thus similar to a previous report for C2A1 20, the single recombinant C2B1 domain did not inhibit membrane resealing. Inhibition of exocytosis in PC12 cells has been shown to require fragments containing the tandem C2 domains of C2A-C2B of syt I 31, which is required for efficiently binding t-snares, syntaxin and SNAP-25 30, 32, 38. For 3T3 cells in the presence of 30 µm C2AB1, the resealing rate to the first wound was 0.08 ±.005 (n=54), which was not significantly (P = 0.497) changed from that of untreated control cells (Table1). But the facilitated resealing response to a 2 nd wounding at the same site was blocked in the C2AB1-treated cells (Fig. 2, ). The 2 nd resealing rate was ±.012 (n=29), which was not significantly (P <.35) increased from the resealing rate of the initial wound (0.081 ±.008). The 12

13 potentiated response to a second wounding at a different site was also blocked by the addition of 30 µm C2AB1 to the medium (data not shown). The 1 st resealing rate of ±.006 (n=25) was decreased to ±.006 resealing rate for a 2 nd wounding at a different site on the cell. With C2AB1 competitive inhibition the vesicle fusions that increase the resealing rate to a second wound appear to be inhibited while the initial response remains intact. The role of synaptotagmins in membrane fusion depends on their function as Ca 2+ -binding modules mediating Ca 2+ -dependent interactions with other fusion proteins 16. Disruption of the Ca 2+ binding properties of syt I impairs synaptic exocytosis The effect of C2AB1 on the facilitated membrane resealing response was dependent on its Ca 2+ binding properties. The syt I Ca 2+ ligand mutant, C2AB (D230,232,363,365N) 30, did not inhibit the facilitated membrane resealing response. The 2 nd resealing rates of 3T3 cells treated with 30 µm syt I C2AB (D230,232,363,365N) was significantly (P <.0002) increased from the 1 st resealing rates (Fig. 2, ), and were ±.018 and ±.022 (n=28), respectively. These resealing rates were similar to those of control cells. Synaptotagmin VII C2 domains have been previously used to conclude that lysosomes are exclusively used in membrane repair 24. We repeated and extended tests of syt VII C2 domains and reached different conclusions. Recombinant C2AB7 fragment had no measurable effect on membrane resealing to an initial wounding, but did block the facilitated response. The initial resealing 13

14 rate of 3T3 cells in 30 µm C2AB7 was ±.004 (n=32), which is similar to the untreated control cell resealing rate (Table 1). But the rate of membrane resealing in the presence of C2AB7 after a 2 nd wounding at the same site was significantly (P<.03) decreased, such that the 2 nd resealing rate fell to 0.06 ±.009. Furthermore 7 of 32 cells failed to reseal after the second wounding. We then examined separately the effects of recombinant fragments of the two Ca 2+ binding domains. We observed that the addition of 30 µm C2A7 to the medium did not inhibit the rate of membrane resealing to an initial wound (0.077 ±.004, n=59), which was similar (P = 0.78) to the rate of resealing in untreated control cells (Table 1). This observation was in contrast to the partial inhibition of membrane resealing after glass bead-wounding in NRK fibroblasts 20. However, membrane resealing to a second wound at the same site was significantly impaired in 3T3 fibroblasts. Half of the cells were unable to reseal to a second wound at the same site (23 of 46 cells) and the resealing rate of cells that did reseal was significantly (P<.0001) decreased to ±.007 (n=23) from the initial membrane resealing rate of ±.004 (n = 46). In the example shown in Figure 3A, the membrane resealed 17 s after cell wounding, but a significantly longer time of 51 s was necessary for membrane resealing after a 2 nd wounding. The reduction of the rate of membrane resealing is seen as data points below the diagonal (Fig. 3B, ) and a 2 nd /1 st resealing ratio of ±.09 (n = 46). The presence of C2A7 did not block membrane resealing to a 2 nd wound at a different site (Fig. 3B, ). 13 of 13 cells resealed after a 2 nd wound, but potentiation of 14

15 the response was not observed. The rate of membrane sealing to the 1 st and 2 nd wounds were 0.09 ±.012 and ±.014 (n = 13), respectively. Plot of the 2 nd and 1 st resealing rates at different sites of C2A7-treated cells were scattered on both sides of the diagonal (Fig. 3B, ) and a 2 nd /1 st resealing ratio of 1.28 ±.23 (n = 13). The effect of C2A7 was dependent on its Ca 2+ -binding function since a putative non Ca 2+ -binding mutant, syt VII C2A (D225, 227N), did not block the facilitated response (Fig. 3B, ) and the resealing rate (0.134 ±.016, n=28) was significantly (P<.001) increased in response to a 2 nd wounding at the same site when compared with the initial resealing rate of ±.007 (n=28). Important Ca 2+ -dependent effector functions of synaptotagmins are associated with the C2B-domain 16, 44. We tested the effect of adding 30 µm recombinant C2B7 to the bath prior to cell wounding. Many cells were observed not to reseal following an initial wounding in the presence of C2B7. Of the cells that did reseal, the fragment significantly (P=0.014) reduced the rate of membrane repair to an initial wounding from that of untreated control cells, ±.005 (n=45) and ±.004 (n=51), respectively (Table 1). Our observation of inhibition of membrane resealing by C2B7 was in direct contrast to an absence of effect on membrane resealing in glass bead-wounded NRK fibroblasts 20. Furthermore membrane resealing after a 2 nd wound either at the same ( ) or different ( ) site was significantly impaired (Fig. 4A). The rate of membrane resealing to a 2 nd wound at the same site was reduced to ±.008 (n=15) from a resealing rate to the 1 st wound of ±.008 (n=26) with 11 of 26 cells 15

16 not resealing after the 2 nd wounding. Cells treated with C2B7 also failed to reseal following a 2 nd wounding at a different site in 11 of 19 recordings, and the rate of membrane resealing was reduced to ±.01 (n=8) from a resealing rate to the 1 st wound of ±.007 (n=19). When the 2 nd resealing rate was plotted against the 1 st resealing rate (Fig. 4A) most of the data points were near or below the diagonal for wounding at the same and different sites. Thus the ratio of 2 nd /1 st rate of membrane resealing were 0.47 ±.13 (n=26) for same sites and 0.53 ±.21 (n=19) for different sites, which were significantly reduced from the 2 nd /1 st ratio for control facilitation and potentiation. The effect of C2B7 on membrane resealing was Ca 2+ -dependent. A putative non Ca 2+ -binding mutant, syt VII C2B (D357, 359N), did not have a deleterious effect on membrane repair after wounding and a facilitated response was observed (Fig. 4B, ). The membrane resealing rate for a 2 nd wounding at the same site increased from ±.013 (n=16) for a 1 st wounding to ±.016 or a 2 nd /1 st resealing ratio of 1.80 ±.32, which is not significantly (P=.126) different from that of untreated control cells (1.75 ±.16, n=20). As a control we used the mutant fragment, syt VII C2B (K320, 321A), which still binds Ca 2+, and observed inhibition of membrane repair and the facilitated response (Fig. 4B, ). The initial resealing rate of syt VII C2B (K320, 321A)-treated cells was reduced to ±.009 (n=24) and 8 of 24 cells failed to reseal after a 2 nd wounding at the same site. Of the cells that resealed, the resealing rate was further decreased to 16

17 0.039 ±.01 (n=16). Thus the C2B7 inhibitory effect on membrane repair was dependent on Ca 2+ binding. Other Ca 2+ -dependent or independent effects of syt VII C2 domains are suggested by comparing syt VII C2B (K320,321A) (Fig. 4B, ) with C2B7 (Fig. 4A, ) or syt I C2AB Ca mutant (Fig. 2, ) with either syt VII C2A Ca mutant (Fig. 3B, ) or syt VII C2B Ca mutant (Fig. 4B, ). The inhibition of facilitation by syt VII C2B (K320,321A) appears to be less than that by C2B7, although both fragments would have intact Ca 2+ binding. Similarly the syt VII C2 Ca mutants appear to still slightly inhibit facilitation when compared with syt I C2AB Ca mutant despite the loss of Ca 2+ binding by the fragments. However, Mann- Whitney statistical analysis of these data sets did not show significance in their differences. Membrane vesicle fusion in neurons is dependent on the formation of the SNARE complex 8, which can be blocked by the cytoplasmic domain of synaptobrevin 2 (cd-syb2) 9, 11. Syb2 is ubiquitously expressed 45 and can be specifically cleaved by botulinum neurotoxin B or tetanus toxin, both of which inhibit resealing in a site specific manner 2, 3, thus we investigated the effect of cdsyb2 on membrane resealing. Cd-syb2 at 10 µm ( ) and 20 µm ( ) were tested with no statistical significant differences between the two concentrations, thus the data was combined (Fig. 5A). Cd-syb2 significantly (P=.0001) inhibited membrane resealing after an initial wound (Table 1). The membrane resealing 17

18 rates in cd-syb2 was ±.005 (n=46), which was significantly reduced from that of untreated control cells. The membrane resealing rate following a second wound at the same site was also decreased when compared to that of untreatedcontrol cells but facilitation was often still observed (25 of 39 recordings are above the diagonal in Fig. 5A). As seen in the example in Fig. 5B after an initial wounding (arrow) the membrane resealed after 54 sec (bar), but a second wounding (arrow) resealed in only 14 sec (bar). The membrane resealing rate was significantly (P<.02) increased following the 2 nd wounding at the same site in cd-syb2 treated 3T3 cells to ±.008 (n=44) or a 2 nd /1 st resealing ratio of ±.309. Discussion Recombinant C2 domains of syts have been used as a dominant negative tool for examining the possible role of syts in a variety of Ca 2+ -mediated exocytotic events 20, 29, 31, The rationale for these studies is that the fragments block the interactions of endogenous syts with their effectors. In this study we have used the recombinant domains to define Ca 2+ -dependent events during membrane resealing in NIH 3T3 fibroblasts after micropunctures. Our results suggest: (1) Inhibition by syt C2 domains depends on their Ca 2+ binding properties. (2) Inhibition by syt C2 domains cannot be used to determine the role of any specific vesicle compartment for membrane resealing. (3) The pool of 18

19 vesicles for an initial membrane resealing response differs from vesicles recruited for facilitation or potentiation of a 2 nd resealing response. Syts have been proposed to act as the Ca 2+ sensor for a variety of Ca 2+ - dependent exocytotic events including synaptic transmission 46, 50, PC12 secretion 29, 31, 49, 51, and endocrine secretion 47, 48. Our results suggest syts may regulate Ca 2+ dependent membrane resealing after micropuncture woundings. The inhibition of membrane resealing by C2AB1 (Fig. 2), C2A7 (Fig. 3B) and C2B7 (Fig. 4B) were all dependent on the Ca 2+ binding property of the recombinant fragment. In all cases the non-ca 2+ binding mutants of inhibitory C2 domains no longer blocked membrane resealing. The amino acid sequences of the C2AB domains of syt I and VII show only a 44% identity and 65% similarity, but both tandem C2 domains conserve the acidic residues 17, 52 required in coordinating Ca Neutralization of these conserved acidic residues disrupts the putative Ca 2+ binding activity of the C2 domains and inhibition of membrane resealing, which suggests the syt fragments act either directly or indirectly as the Ca 2+ sensors for membrane resealing. Inhibition by syt VII C2 fragments may not implicate particular syt isoforms in normal function. Recombinant fragments of syt VII C2 domains are potent inhibitors of PC12 secretion 31, 53, yet PC12 cells express syt VII at very low levels 31. A recent gene silencing approach suggested syt IX, not syt VII, is indispensable for PC12 exocytosis 51. Other studies suggest that syt VII C2 domains are non-selective due to binding with a variety of other molecules 31, 49, 52 29, 54, 55, including oligomerization with other subclasses of syts 19

20 and competition with native syts for binding to effectors, such as SNARE proteins and phosphatidylinositol 4,5-bisphosphate 31. These other binding properties of syt VII C2 domains may also be important in membrane repair. The source of intracellular compartments for membrane repair is uncertain. The lysosome compartment has been suggested as the vesicular source of membrane for resealing 24 due to a correlation of C2A7 fragment inhibition of lysosome exocytosis 23 and membrane resealing of glass beadwounded NRK fibroblasts 20, and failure of both C2B7 and C2A1 fragments to inhibit either lysosome exocytosis 23 or membrane resealing 20. We have found that C2A7 fragment does not inhibit membrane repair after the initial wound (Table 1) and the C2B7 fragment is an even more potent inhibitor than C2A7 fragment of both membrane resealing after the 1 st wounding (Table 1) and the facilitated or potentiated response after the 2 nd wounding. We do not know why C2B7 inhibited membrane resealing in 3T3 fibroblasts, but not in NRK fibroblasts 20. Syt VII has been shown to partially co-localize to lysosomes in both 3T3 37 and NRK 23 fibroblasts. The C2B domain engages a wider array of effector molecules than the C2A domain 16, 44. At least in 3T3 fibroblasts, C2B7 cannot be used as supporting evidence that lysosomes are the exclusive compartment for resealing. This conclusion is corroborated by the recent report that syt VII knockout mice exhibit enhanced lysosome fusion 56, but show a somewhat reduced membrane repair response 57. In addition, vacuolin-1, which can completely block Ca 2+ -dependent exocytosis of lysosomes, did not affect the rate 20

21 of membrane resealing 58. Thus lysosomes do not appear to be essential for the process of membrane resealing, and inhibition by syt VII C2 fragments does not now appear to be consistent with a major role for lysosomes in membrane repair. From studies of Ca 2+ -triggered exocytosis in chromaffin cells, intracellular vesicles have been proposed to be subdivided into separate pools of rapidly releasable, slowly releasable and unprimed vesicles 59. A proposed distinguishing difference between the pools is the formation of the SNARE complex between v-snare and t-snare proteins, and interacting partners such as synaptotagmins 59. The role of syts in the fusion process is unclear but our studies suggest that vesicle pools used during membrane resealing may also be similarly subdivided. The differential effects of C2A7 and C2B7 suggest different effectors for syts in membrane fusion of releasable and unprimed vesicles. C2A7 had no effect on the initial rate of membrane resealing (Table 1), but significantly prevented both facilitation and potentiation (Fig. 3B). An important difference in C2A7 inhibition of facilitation and potentiation was that 65% of the second woundings at the same site either failed to reseal or had a 2 nd resealing rate less than 0.01 sec -1, while all the second woundings at a different site resealed. Since facilitation, but not potentiation, requires recruitment of new vesicles 3, 4, these results suggest that C2A7 did not block membrane resealing by releasable vesicles, but the recruitment of newly synthesized or unprimed vesicles. Facilitation also depends on protein kinase C activity 3, which enhances exocytosis from chromaffin cells by increasing the size of the readily releasable 21

22 pool 60. In contrast to C2A7, C2B7 significantly inhibited the initial rate of membrane resealing (Table 1). The rate of membrane resealing for a 2 nd wounding at either the same or different site was further reduced. More than 60% of the second wounding at same or different sites had a 2 nd resealing rate less than 0.01 sec -1. A possible explanation is that C2B7 was able to disrupt the fusion of vesicles in releasable pools, with the plasma membrane, as well as, the recruitment of unprimed vesicles. These observations suggest that the C2B fragment engages a wider array of effectors than the C2A fragment, which is similar to a more crucial role for C2B in synaptic transmission 44. A difference in vesicle pools seen with syt VII C2 domain inhibitions of first and second wounds can also be detected by syt I C2 domains. We found that CAB1 had no effect on the rate of resealing to an initial wound but significantly inhibited both the facilitated and potentiated responses to a second wound. The different sensitivies of first and second wounds again suggest that more than one vesicle pool is used in membrane repair. The dependence of the initial wound resealing upon vesicles of the releasable pool is further supported by our observations with the recombinant cd-syb2, which would be expected to disrupt the coupling between v-snare and t-snare 9. The addition of cd-syb2 caused a dramatic delay in membrane repair, such that the rate of membrane resealing after the initial wound was significantly decreased from that of control cells (Table 1). Interestingly membrane resealing after a 2 nd wounding at the same site was also significantly delayed when 22

23 compared to that of control, but nonetheless facilitation was observed. The ratio of 2 nd /1 st resealing rates at the same site for cd-syb2-treated cells was similar to that of control cells, ±.309 (n=43) and ±.164 (n=29), respectively. This implies that cd-syb2 interferes with the trans-snare complex formation and vesicle-membrane fusion required for membrane repair, but does not prevent the recruitment of new vesicles required for the facilitated response. Wounding the plasma membrane evokes a series of membrane trafficking events at a distinct time and place, ranging from the immediate Ca 2+ -dependent exocytotic fusion of docked vesicles, to the generation, recruitment and transport of new ones. Our studies of resealing with recombinant syt and syb2 fragments reveal a high degree of similarity with the fusion mechanisms of neurosecretion, and suggest that membrane repair will be an important adjunct to understanding the molecular events in vesicle trafficking. The use of different vesicle pools for initial, facilitated and potentiated membrane resealing responses is suggested by the differential effects of the recombinant syt and syb2 fragments on these physiological processes. Membrane resealing can be useful for further characterization of the molecular properties of the different vesicle pools. Acknowledgements This study was supported in part by grants from the National Institute of Health (EY 13436) to R.A.S. and (NIGMS GM and NIMH MH 61876) to 23

24 E.R.C. E.R.C. is a Pew Scholar in the Biomedical Sciences. W.C.T. is supported by a Postdoctoral Individual NRSA from the NIH. References 1. McNeil, P.L., and Steinhardt, R.A. (2003) Annu Rev Cell Dev Biol. 19, Steinhardt, R.A., Bi, G.-Q., and Alderton J.M. (1994) Science 263, Togo, T., Alderton J.M., Bi G.-Q., and Steinhardt R.A. (1999) J Cell Sci. 112, Togo T., Alderton J.M., and Steinhardt R.A. (2003) Mol Biol Cell. 14, Jahn, R., and Südhof, T.C. (1999) Annu. Rev. Biochem. 68, Augustine, G.J. (2001) Curr. Opin. Neurobio. 11, Chapman, E.R. (2002) Nature Rev. 3, Sollner, T., Whiteheart, S.W., Brunner, M., Erdjument-Bromage, H., Geromanos, S., Tempst, P., and Rothman, J.E. (1993) Nature 362, Weber, T., Zemelman, B.V., McNew, J.A., Westermann, B., Gmachl, M., Parlati, F., Sollner, T., and Rothman, J.E. (1998) Cell 92, Parlati, F., Weber, T., McNew, J.A., Westermann, B., Sollner, T.H., and Rothman, J.E. (1999) Proc Natl. Acad. Sci. U.S.A. 96, Tucker, W.C., Weber, T., and Chapman, E.C. (2004) Science 304, Marqueze, B., Berton, F., and Seagar, M. (2000) Biochimie 82,

25 13. Perin, M.S., Brose, N., Jahn, R., and Südhof, T.C. (1991) J. Biol. Chem. 266, Nalfeski, E.A., and Falke, J.J. (1996) Protein Sci. 5, Rizo, J. and Südhof, T.C. (1998) J. Biol. Chem. 273, Tucker, W.C. and Chapman, E. R. (2002) Biochem. J. 366, Rickman, C., Craxton, M., Osborne, S., and Davletov, B. (2004) Biochem. J. 378, Bi, G.-Q., Alderton, J.M., and Steinhardt, R.A. (1995) J. Cell Biol. 131, Miyake, K. and McNeil, P. L. (1995) J. Cell Biol. 131, Reddy, A., Caler, E.V., and Andrews, N.W. (2001) Cell 106, Blasi, J., Chapman, E.R., Yamasaki, S., Binz, T., Niemann, H., and Jahn, R. (1993) EMBO J. 12, Schiavo, G., Benfanati, F., Poulain, B., Rosetto, O., Polverino de Laaureto, P., DasGupta, B.R., and Montecucco, C. (1992) Nature 359, Martinez, I., Chakrabarti, S., Hellevik, T., Morehead, J., Fowler, K., Andrews, N.W. (2000) J. Cell Biol. 148, Andrews, N.W. (2000) Trends Cell Biol. 10, Borgonovo B., Cocucci E., Racchetti G., Podini P., Bachi A., and Meldolesi J. (2002) Nature Cell Biol. 4, Perin, M.S., Fried, V.A., Mignery, G.A., Jahn, R., and Südhof, T.C. (1990) Nature 345,

26 27. Fukuda, M., Kanno, E., and Mikoshiba, K. (1999) J. Biol. Chem. 274, Littleton, J.T., Sereno, T.L., Rubin, G.M., Ganetzky, B., and Chapman, E.R. (1999) Nature 400, Desai, R.C., Vyas, B., Earles, C.A., Littleton, J.T., Kowalchyck, J.A., Martin, T.F.J., and Chapman, E.R. (2000) J. Cell Biol. 150, Earles, C.A., Bai, J., Wang, P., and Chapman, E.R. (2001) J. Cell Biol. 154, Tucker, W. C., Edwardson, J.M., Bai, J., Kim, H.-J., Martin, T.F.J., and Chapman, E.R. (2003) J. Cell Biol. 162, Chapman, E. R., An, S., Edwardson, J.M., and Jahn, R. (1996) J. Biol. Chem. 271, Wu, Y., He, Y., Bai, J., Ji, S.-R., Tucker, W.C., Chapman, E.R., and Sui, S.- F. (2003) Proc. Natl. Acad. Sci. U.S.A. 100, Südhof, T.C. (2002) J. Biol. Chem. 277, Koh, T.W., and Bellen, H.J. (2003) Trends Neurosci. 26, Tarantini, F., La Vallee, T., Jackson, A., Gamble, S., Carreira, C.M., Garfinkel, S., Burgess, W.H., and Maciag, T. (1998) J. Biol. Chem. 273, Caler, E.V., Chakrabarti, S., Fowler, K.T., Rao, S., and Andrews, N.W. (2001) J. Exp. Med. 193, Bai, J., Wang, C.T., Richards, D.A., Jackson, M.B., and Chapman, E.R. (2004) Neuron 41,

27 39. Fernandez-Chacon, R., Konigstorfer, A., Gerber, S.H., Garcia, J., Matos, M.F., Stevens, C.F., Brose, N., Rizo, J., Rosenmund, C., and Südhof, T. C. (2001) Nature 410, Littleton, J.T., Bai, J., Vyas, B., Desai, R., Baltus, A.E., Garment, M.B., Carlson, S.D., Ganetzky, B., and Chapman, E.R. (2001) J. Neurosci. 21, Mackler, J.M., Drummond, J.A., Loewen, C.A., Robinson, I.M., and Reist, N.E. (2002) Nature 418, Robinson, I.M., Ranjan, R., and Schwarz, T.L. (2002) Nature 418, Stevens, C.F., and Sullivan, J.M. (2003) Neuron 39, Bai, J., and Chapman, E.R. (2004) Trends Biochem. Sci. 29, Rossetto, O., Gorza, L., Schiavo, G., Schiavo, N., Scheller, R.H., and Montecucco, C. (1996) J. Cell Biol. 132, Bommert, K., Charlton, M.P., DeBello, W.M., Chin, G.J., Betz, H., and Augustine, G.J. (1993) Nature 363, Baram, D., Mekori, Y.A., and Eisenberg, R.S. (2001) Immunol. Rev. 179, Gut, A., Kiraaly, C.E., Fukuda, M., Mikoshiba, K., Wollheim, C.B., and Lang, J. (2001) J. Cell Sci. 114, Sugita, S., Han, W., Butz, S., Liu, X., Fernandez-Chacon, R., Lao, Y., and Südhof, T.C. (2001) Neuron 30, Geppert, M., Goda, Y., Hammer, R.E., Li, C., Rosahl, T.W., Stevens, C., and Südhof, T.C. (1994) Cell 79,

28 51. Fukuda, M. (2004) Biochem. J. 380, Li, C., Ullrich, B., Zhang, J.Z., Anderson, R.G.W., Brose, N., and Südhof, T.C. (1995) Nature 375, Sugita, S., Shin, O.-H., Han, W., Lao, Y., and Südhof, T.C. (2002) EMBO J. 21, Fukuda, M.,and Mikoshiba, K. (2000) J. Biol. Chem. 275, Fukuda, M.,and Mikoshiba, K. (2001) J. Biol. Chem. 276, Jaiswal, J.K., Chakrabarti, S., Andrews, N.A., and Simon, S.M. (2004) PloS 2, e Chakrabarti, C., Kobayashi, K.S., Flavell, R.A., Marks, C.B., Miyake, K., Liston, D.R., Fowler, K.T., Gorelick, F.S., and Andrews, N.W. (2003) J. Cell Biol. 162, Cerny, J., Feng, Y., Yu, A., Miyake, K., Borgonovo, B., Kluperman, J., Meldolesi, J., McNeil, P.L., and Kirchausen, T. (2004) EMBO reports. 5, Rettig, J., and Neher, E. (2002) Science 298, Gillis, K.D., Mobner, R., and Neher, E. (1996) Neuron 16,

29 Figure Legends Figure 1. Facilitated and potentiated membrane resealing reponses in untreated control NIH 3T3 cells. A. A typical cell was micropunctured by a 0.3 sec jab of a microneedle at the 1 st arrow. The membrane wound caused a loss in fura-2 dye detected as a drop in intensity at the Ca 2+ insensitive λ=357 nm (top trace) and a rise in intracellular Ca 2+ activity detected as an increase in the emission ratio of fura-2 dye excited at 357 and 385 nm (bottom trace). Resealing of the membrane causes a cessation of dye loss (bar) and the rate of membrane resealing is the inverse of the duration of dye loss. In this example the duration of dye loss after the initial wound was 27 s or an initial rate of resealing of s -1. The cell was wounded a second time (2 nd arrow) at the same site, which required only 16 s to reseal. The faster resealing to a second wounding at the same or different site is called a facilitated or potentiated response, respectively. B. Plot of the rate of membrane reseal for the 2 nd wound on the Y-axis to the rate of membrane reseal for the 1 st wound on the X-axis for the same ( ) or different ( ) sites. Figure 2. Ca 2+ -binding dependent inhibition of the facilitated response by Syt I C2AB. Cells were micropunctured in the presence of 30µM C2AB1 ( ), 29

30 non-ca 2+ -binding C2AB1 mutant ( ) or C2B1 ( ). The presence of C2AB1 blocked facilitation, but C2B1 and Ca 2+ -binding mutant of C2AB1 had no effect. Figure 3. Ca 2+ -binding dependent inhibition of the facilitated and potentiated responses by Syt VII C2A. A. The resealing record of a 3T3 cell in 30 µm C2A7. Membrane resealing after the initial wound was similar to untreated control cells and in this case the membrane resealed 17 s after the initial wounding. Half of the cells wounded a second time at the same site failed to reseal and the rate was defined as zero. In most cases that did reseal, the resealing time was significantly longer and in this case the membrane resealed 51 s after the 2 nd wounding. B. Plot of the 2 nd and 1 st resealing rates at the same sites ( ) for cells treated with C2A7. All cells wounded a 2 nd time at a different site ( ) did reseal, but 5 of 13 cells had a slower rate of membrane resealing than that of the initial wound. Cells treated with the Ca 2+ -binding mutant of C2A7 ( ) showed a facilitated response to a 2 nd wounding. Figure 4. Ca 2+ -binding dependent inhibition of initial resealing, facilitated and potentiated responses by Syt VII C2B. A. The rate of membrane resealing to 1 st and 2 nd woundings of 3T3 cells in 30 µm C2B7 at same ( ) and different ( ) sites were significantly inhibited. Many cells failed to reseal after an initial micropuncture. Of those that did reseal the rate of membrane resealing was significantly increased. More than half of these cells failed to reseal after a 2 nd wounding at the same or different sites. B. The inhibitory effect of the 30

31 recombinant C2B7 peptide was dependent on Ca 2+ -binding. Cells treated with the deficient Ca 2+ -binding mutant fragment ( ) resealed normally after the initial wounding and the response to a 2 nd wounding at the same site was facilitated. A control mutant fragment ( ) still inhibited membrane repair after the initial micropuncture and 8 of 24 cells that resealed after the initial wounding failed to reseal after a 2 nd wounding at the same site. Figure 5. The rate of membrane resealing, but not the facilitated response, was inhibited by cd-syb2. A. The rate of membrane resealing to 1 st and 2 nd woundings of 3T3 cells in 10 ( ) or 20 ( ) µm cytoplasmic domain of synaptobrevin 2 (cd-syb2) at the same sites was significantly reduced. However, a facilitated response was still observed in these cells. B. An example of membrane resealing of a 3T3 cell in cd-syb2. After the initial wounding (1 st arrow) the membrane resealed (1 st bar) after 54 s. A 2 nd wounding (2 nd arrow) resealed (2 nd bar) in 14 s. 31

32 Table 1. Effect of Vesicle Fusion Fragments on Initial Resealing Rates in NIH 3T3 Cells. Cells were micropunctured in Ringer s containing different recombinant fragments and the rate of membrane resealing for an initial wounding was determined. Treatment Initial Resealing Rate 1 Mann-Whitney Paired Test 2 Control (untreated) ±.004 (51) 3 30 µm C2AB ±.005 (54) P = µm C2AB ±.004 (32) P = µm C2A ±.004 (59) P = µm C2B ±.005 (45) P = /20 µm cd-syb ±.005 (46) P = Initial membrane resealing recordings pooled from same and different sites recordings 2 Comparison with initial resealing rate of untreated control cells 3 Mean ± sem (number of recordings) 32

33 A A Intensity (357 nm) /385 Ratio Time (seconds) B 0.3 Second Resealing Rate Different Sites Same Sites Initial Resealing Rate Figure 1 33

34 0.3 Second Resealing Rate C2AB1 C2AB1 (Ca mutant) C2B Initial Resealing Rate Figure 2 34

35 A Intensity (357 nm) /385 Ratio Time (seconds) B 0.3 Second Resealing Rate C2A7: Same Sites C2A7: Different Sites Ca Mutant: Same Sites Initial Resealing Rate Figure 3 35

36 A 0.3 Second Resealing Rate C2B7: Same Sites C2B7: Different Sites Initial Resealing Rate B 0.3 Second Resealing Rate C2B7 (D357, 359N) C2B7 (K320, 321A) Initial Resealing Rate Figure 4 36

37 A 0.3 Second Resealing Rate B Initial Resealing Rate 10 µm cd-syb2 20 µm cd-syb Intensity (357 nm) /385 Ratio Time (seconds) 0 Figure 5 37