Báo cáo khoa học: A search for synthetic peptides that inhibit soluble N-ethylmaleimide sensitive-factor attachment receptor-mediated membrane fusion pot

As an initial attempt to identify the peptide sequences that block SNARE assembly and membrane fusion, we created thirteen 17-residue synthetic peptides derived from the SNARE motifs of

N-ethylmaleimide sensitive-factor attachment

receptor-mediated membrane fusion

Chang H Jung1,*, Yoo-Soo Yang1,*, Jun-Seob Kim1, Jae-Il Shin1, Yong-Su Jin1, Jae Y Shin1, Jong

H Lee2, Koo M Chung2, Jae S Hwang3, Jung M Oh1, Yeon-Kyun Shin4and Dae-Hyuk Kweon1

1 School of Biotechnology and Bioengineering, Sungkyunkwan University, Gyeonggi-do, Korea

2 School of Bioresource Sciences, Andong National University, Gyeongsangbuk-do, Korea

3 Amorepacific R&D Center, Yongin, Korea

4 Department of Biochemistry, Biophysics, and Molecular Biology, Iowa State University, Ames, IA, USA

Soluble N-ethylmaleimide-sensitive factor attachment

protein receptor (SNARE) proteins have central roles

in neurotransmission At the synapse, membrane

fusion, which is required for neurotransmitter release,

is mediated by SNAREs Vesicle-associated membrane

protein 2 (v)-SNARE synaptobrevin (VAMP2)

associ-ates with target membrane (t)-SNAREs syntaxin 1a

and synaptosome-associated protein of 25 kDa (SNAP-25) [1–3] to form the highly stable ternary SNARE complex [4–6] Cumulative evidence has shown that the SNARE complex forms the core of the machine that generates the energy required for mem-brane fusion, while other accessory proteins are involved in docking, tethering, Ca2+-sensing and

Keywords

exocytosis; membrane fusion;

neurotransmitter release; SNARE;

SNARE-patterned peptide

Correspondence

D.-H Kweon, School of Biotechnology and

Bioengineering, Sungkyunkwan University,

Suwon, Gyeonggi-do 440-746, Korea

Fax: +82 31 290 7870

Tel: +82 31 290 7869

E-mail: dhkweon@skku.edu

*These authors contributed equally to this

work

(Received 28 December 2007, revised 9

March 2008, accepted 10 April 2008)

doi:10.1111/j.1742-4658.2008.06458.x

Soluble N-ethylmaleimide sensitive-factor attachment receptor (SNARE) proteins have crucial roles in driving exocytic membrane fusion Molecular recognition between vesicle-associated (v)-SNARE and target membrane (t)-SNARE leads to the formation of a four-helix bundle, which facilitates the merging of two apposing membranes Synthetic peptides patterned after the SNARE motifs are predicted to block SNARE complex formation by competing with the parental SNAREs, inhibiting neuronal exocytosis As

an initial attempt to identify the peptide sequences that block SNARE assembly and membrane fusion, we created thirteen 17-residue synthetic peptides derived from the SNARE motifs of v- and t-SNAREs The effects

of these peptides on SNARE-mediated membrane fusion were investigated using an in vitro lipid-mixing assay, in vivo neurotransmitter release and SNARE complex formation assays in PC12 cells Peptides derived from the N-terminal region of SNARE motifs had significant inhibitory effects on neuroexocytosis, whereas middle- and C-terminal-mimicking peptides did not exhibit much inhibitory function N-terminal mimicking peptides blocked N-terminal zippering of SNAREs, a rate-limiting step in SNARE-driven membrane fusion Therefore, the results suggest that the N-terminal regions of SNARE motifs are excellent targets for the development of drugs to block SNARE-mediated membrane fusion and neurotransmitter release

Abbreviations

DOPS, 1,2-dioleoyl-sn-glycero-3-phosphatidylserine; POPC, 1-palmitoyl-2-dioleoyl-sn-glycero-3-phosphatidylcholine; SNAP-25, synaptosome-associated protein of 25 kDa; SNARE, soluble N-ethylmaleimide sensitive factor attachment receptor; VAMP2, vesicle-synaptosome-associated membrane protein 2.

recycling A fusion pore is created as a result of

mem-brane fusion and neurotransmitters are released

through this pore [7–12]

The SNARE complex is a parallel four-helical

bun-dle [4,5] The four core helices are provided by the

three SNARE proteins; one each from syntaxin and

VAMP2, and two from SNAP-25 [4,5] Syntaxin and

VAMP2 are transmembrane proteins with single

trans-membrane helices and are anchored to the presynaptic

membrane and vesicular membrane, respectively

SNAP-25 is peripherally attached to the presynaptic

membrane Thus, formation of a parallel coiled-coil

would result in close apposition of two membranes,

and facilitate membrane fusion The core complex has

been shown to be linked tightly to the membranes so

that the force generated by SNARE complex

forma-tion can be faithfully transferred to the membranes

[13–15]

Impaired SNARE function is known to block

neuro-nal exocytosis For example, synaptic SNARE proteins

are the specific substrates of eight clostridial

neuro-toxins (one tetanus neurotoxin and seven botulinum

neurotoxins; BoNT⁄ A–G) [16,17] The neurotoxins

bind specifically to the nerve terminals and deliver the

N-terminal catalytic domain of the zinc-endopeptidase

into the cytosol, where the catalytic domain specifically

cleaves a SNARE protein at a single site within its

cytosolic portion Such specific cleavage leads to an

inhibition of neuroexocytosis and results in the

para-lytic syndromes of botulism and tetanus Low doses of

BoNT⁄ A are now widely used to alleviate the

symp-toms of various disorders, including paralytic

strabis-mus, blephoraspasm, cervical dystonia and severe

hyperhydrosis [18,19] The neurotoxin is also widely

used for cosmetic purposes Clostridial toxins have

pro-ven very versatile with therapeutic and cosmetic uses

Inhibition of neurotransmission might be achieved

not only by the specific cleavage of SNARE proteins,

but also by blocking SNARE complex formation, and

several peptide inhibitors have been developed for this

purpose [20–27] The peptides are mostly modeled on

the sequences of the SNARE motifs in SNAP-25

These peptides are thought to competitively inhibit

SNARE complex formation by interfering with

inter-actions between parental SNARE proteins However,

no systematic study has evaluated the efficacy of

SNARE-patterned sequences from all four SNARE

motifs As such, there is no design principle to guide

the development of potent peptide SNARE inhibitors

Because individual sequences modeled on the SNARE

motifs are expected to inhibit SNARE complex

forma-tion to differing degrees, careful assessment of the

effect of SNARE-patterned peptides on SNARE

complex formation might help us better understand SNARE-driven neuronal exocytosis Such efforts will also help us identify the potent peptide sequences that interfere with neurotransmission In this study, we designed and synthesized small peptides derived from SNARE motifs We assessed their inhibitory activities

on membrane fusion, neuroexocytosis of digitonin-permeabilized PC12 cells and SNARE complex forma-tion in PC12 cells We found that the peptide sequences derived from the N-terminal regions of SNARE motifs show high inhibitory activity

Results

Design of peptides modeled on the SNARE motifs

Peptide sequences modeled on SNARE motifs are most likely to compete with parental SNARE proteins for SNARE complex formation, inhibiting SNARE assembly In an initial attempt to search for the potent peptide sequences that inhibit SNARE complex forma-tion and neurotransmitter release, we synthesized thir-teen 17-mer peptides representing the various regions

in the SNARE motifs of individual SNAREs Three peptide sequences were derived from each SNARE motif, giving 12 from four SNARE motifs Three sequences from each SNARE motif represent the N-terminal, middle and C-terminal regions (Fig 1A) Sequences representing the middle regions contained the amino acid Q or R that is present in the zero-layer

of the SNARE complex (bold italics in Table 1) [5] Individual peptide sequences are shown in Table 1 One additional 17-mer peptide (designated as LIB), which was previously selected as a SNARE inhibitor from an a-helix-constrained combinatorial peptide library [24], was prepared as a positive control

Inhibition of SNARE-driven membrane fusion by synthetic peptides

In order to measure the efficacy of 13 synthetic pep-tides, a fluorescence lipid-mixing assay was performed using SNARE-reconstituted liposomes (see Experi-mental procedures) [3,28] The lipid-mixing assay is a well-established method that has frequently been used

to show various features of SNARE-driven membrane fusion [7,12,28–33]

For the lipid-mixing assay, full-length t-SNARE complex was reconstituted into large unilamellar vesi-cles (100 nm diameter) composed of 1-palmitoyl-2-dioleoyl-sn-glycero-3-phosphatidylcholine (POPC) and 1,2-dioleoyl-sn-glycero-3-phosphatidylserine (DOPS) in

a molar ratio of 65 : 35 Full-length v-SNARE was

reconstituted into a separate population of the same

POPC⁄ DOPS vesicles containing 1.5 mol% each of

flu-orescence lipids When the t-SNARE and v-SNARE

vesicles were mixed without adding the synthetic

peptides there was an increase in the fluorescence signal (Fig 2A), indicating that lipid mixing had occurred Several synthetic peptides showed rather poor solu-bility in water, whereas others had good solusolu-bility in buffer Therefore, dimethylsulfoxide was used to dis-solve peptides with poor water solubility and peptide⁄ dimethylsulfoxide solutions were injected directly into the fusion reaction The added dimethylsulfoxide influ-enced lipid mixing somewhat (Fig 2A); therefore, the inhibition efficacies of the peptides were tested in two different batches depending on the solvent conditions (Fig 2)

Some synthetic peptides had profound effects on membrane fusion when tested at a peptide concentra-tion of 200 lm (Fig 2A,B) Peptides representing the N-terminal region of the SNARE motifs SCN (from SNAP-25C), SynN (from Syntaxin 1a) and VpN (from VAMP2) reduced membrane fusion significantly, whereas the other peptides were much less effective The most effective peptides (VpN and SynN) decreased membrane fusion by as much as 60–70% of the con-trol (Fig 2B)

The amphipathic character of the SNARE-patterned peptides may make them fusogenic [34–36] In order to rule out this possibility, the fusion activities of the pep-tides were tested in the absence of SNAREs (Fig 2B, gray bar) None of the 17-mer peptides had significant membrane fusion activity without SNAREs The cyto-plasmic domain of VAMP2 (VpS, amino acids 1–96) has been used frequently as an inhibitor for SNARE-driven membrane fusion [3] VpS suppressed mem-brane fusion very effectively at lower concentrations (20 lm) (Fig 2A,B) and served as a good positive control to compare the inhibitory effect of the syn-thetic peptides

37

kDa

25

15

10

SNAP-25

Syntaxin 1a VA

MP-2

20

A

B

Fig 1 Design of SNARE motif-patterned peptides and purification

of full-length SNARE proteins (A) Synthetic peptides patterned

after the a-helical regions of SNARE motifs Peptide names are

designated in the gray box Hexamer peptides were also

synthe-sized patterned after VpN, VpM and VpC sequences The amino

acid sequences of synthetic peptides are shown in Table 1 (B)

SDS ⁄ PAGE analysis of purified recombinant SNARE proteins used

in this study.

Table 1 Amino acid sequences of synthetic peptides Glutamine and arginine residues in the zero layer are shown in bold italics.

SNN (7–23)

MRNELEEMQRRADQLAD

SNM (45–61) RTLVMLDEQGEQLERIE

SNC (65–71) DQINKDMKEAEKNLTDL SCN (140–156)

DARENEMDENLEQVSGI

SCM (166–182) DMGNEIDTQNRQIDRIM

SCC (190–206) TRIDEANQRATKMLGSG SynN (192–208)

LSEIETRHSEIIKLENS

SynM (218–224) DMAMLVESQGEMIDRIE

SynC (245–261) ERAVSDTKKAVKYQSKA VpN (30–46)

RRLQQTQAQVDEVVDIM

VpM (48–64) VNVDKVLERDQKLSELD

VpC (76–92) QFETSAAKLKRKYWWKN LIB

SAAEAFAKLYAEAFAKG

Argireline EEMQRR VpN1

RRLQQT

VpN2 QAQVDE

VpN3 EVVDIM VpM1

VNVDKV

VpM2 LERDQK

VpM3 KLSELD VpC1

QFETSA

VpC2 KLKRKY

VpC3 KYWWKN

Fig 2 Inhibitory effects of the synthetic peptides on SNARE-driven membrane fusion Peptides were added at 200 l M (A) Percent of maxi-mum fluorescence intensity was plotted as a function of time in the presence or absence of the peptides The maximaxi-mum fluorescence intensity was obtained by adding 0.1% Triton X-100 (upper) for peptides dissolved in dimethylsulfoxide (DMSO) and (lower) for peptides dissolved in dis-tilled water (DW) VpS, soluble domain of VAMP2 (amino acids 1–96) (B) The inhibitory effect of synthetic peptides was converted to ‘% of con-trol’ (black bar) Percentage of maximum fluorescence intensity in the presence of synthetic peptides was divided by that of dimethylsulfoxide

or distilled water (DW) depending on the dissolving solvents Each bar represents mean ± SD of three independent experiments *P < 0.05 ver-sus control SNARE-independent membrane fusion was shown as ‘% of maximum fluorescence intensity’ (gray bar) (C) Western blot analysis

of SNARE complex formation was performed after the membrane fusion assay in order to confirm that the N-terminal-mimicking peptides inhib-ited SNARE complex formation Numbers below the figure are relative band intensities Asterisks denote peptides dissolved in dimethylsulfox-ide and should be compared with dimethylsulfoxdimethylsulfox-ide as the control; others were dissolved in water and should be compared with distilled water.

In order to verify whether the reduced membrane

fusion was caused by inhibition of SNARE complex

formation by the peptides, western blot analysis was

performed for six peptides after membrane fusion The

SDS-resistant SNARE complex was detected using

anti-(SNAP-25) IgG SCN, SynN and VpN reduced

SNARE complex formation, whereas SNN, SCM and

LIB were not effective (Fig 2C), supporting the idea

that the inhibitory activities were due to inhibition of

SNARE complex formation by the peptides

Interestingly, SCM increased membrane fusion by

 70% (Fig 2A,B), although it did not enhance

SNARE complex formation (Fig 2C) It has

previ-ously been shown that a synthetic peptide from the

C-terminal region of VAMP2 promoted

SNARE-medi-ated fusion [37] Pobbati et al [37] showed that this

stimulatory effect might be due to its role in refolding

the t-SNARE complexes into an active conformation

We wonder whether SCM has a similar effect

in SNARE-mediated membrane fusion, warranting

further investigation

Inhibition of neurotransmitter release from PC12

cells by the synthetic peptides

In order to measure the effects of synthetic peptides on

neuronal exocytosis, we prepared permeabilized PC12

cells loaded with [3H]-noradrenaline Depolarization of

these cells by high K+ concentrations (50 mm KCl)

results in the release of neurotransmitters such as

nor-adrenaline, acetylcholine and arachidonic acid [38,39]

Inhibition of this release by synthetic peptides could be

assessed by measuring [3H]-noradrenaline release after

stimulation with high concentrations of K+, in the

presence or absence of synthetic peptides In the

permeabilized PC12 cells, stimulation with high

concentrations of K+ significantly increased [3

H]-noradrenaline release, compared with the basal levels

We tested the effects of the synthetic peptides on the

high K+ concentration-stimulated release of [3

H]-nor-adrenaline Addition of SCN, SynN and VpN

effi-ciently inhibited neuroexocytosis even at a final

concentration of 10 lm, whereas other peptides

showed no significant effect (Fig 3A), consistent with

results from the lipid-mixing assay Inhibition of

[3H]-noradrenaline release by the

N-terminal-mimick-ing peptides SCN, SynN and VpN was as much as

20–30%, when compared with controls (Fig 3A)

Verapamil, an L-type calcium-channel blocker,

inhib-ited neurotransmitter release by  70% at the same

concentration (10 lm) [40,41] Thus,

N-terminal-mimicking peptides serve as good SNARE-targeting

neuroexocytosis inhibitors We were not able to test

A

B

C

Fig 3 Inhibitory effects of the synthetic peptide on [ 3 H]-noradrena-line release in high K+-stimulated PC12 cells (A) The high K+ -stimu-lated [ 3 H]-noradrenaline release over 15 min was measured in the presence or absence of synthetic peptides The amount of [ 3 H]-nor-adrenaline released measured in the presence of an inhibitory pep-tide was divided by that measured in the absence of the peppep-tide, where the amount of [ 3 H]-noradrenaline release is (cpm of high

K+-stimulated sample – cpm of basal level release) per mg of pro-tein Each bar represents mean ± SD of 5–7 observations.

*P < 0.05 versus K + -evoked control (B, C) Inhibition of SNARE complex formation by the synthetic peptides in K + -stimulated PC12 cells SNARE complex was detected by western blotting using an anti-(SNAP-25) IgG (B) and relative band densities are represented

by the bar graph (C) Cells were pretreated with synthetic peptides for 2 h prior to stimulation with K+(50 m M KCl), and then further incubated for 15 min.

higher peptide concentrations because some peptides

were dissolved in dimethylsulfoxide which killed PC12

cells at higher concentrations

In order to verify whether the peptides inhibited

SNARE assembly in PC12 cells, membrane proteins

were extracted from the cells and analyzed by western

blotting using anti-(SNAP-25) IgG (Fig 3B,C)

Consis-tent with other results, SCN, SynN and VpN inhibited

SNARE complex formation significantly (Fig 3B,C)

VpN inhibited SNARE complex formation with a

sim-ilar efficiency to verapamil N-Terminal-mimicking

peptides inhibited SNARE complex formation by

directly interacting with native SNARE proteins,

whereas verapamil did the same by blocking calcium

entry By contrast, other peptides did not have

notice-able inhibitory effects on SNARE complex formation

Therefore, the results show that the reduction in

neu-roexocytosis by SCN, VpN and SynN was a direct

consequence of the inhibition of SNARE complex

formation

Smaller peptides might be beneficial in terms of

syn-thesis costs, as well as membrane penetration

There-fore, we synthesized nine additional 6-residue peptides

derived from v-SNARE VAMP2 (Fig 1A and

Table 1) We also made the well-known Argireline

[26,42] We did not observe any measurable inhibitory

effects with the 6-mer peptides when tested using the

lipid-mixing assay (data not shown) and noradrenaline

release assay (Fig 3)

Time-dependent effects of inhibitory peptides on

membrane fusion and neurotransmitter release

N-Terminal zippering may be the rate-limiting step in

SNARE complex formation The N-terminal-mimicking

peptides may have inhibited this N-terminal zippering,

resulting in the most significant inhibition of membrane

fusion and neurotransmitter release Conversely,

N-ter-minal-mimicking peptides might not inhibit SNARE

complex formation and membrane fusion when partial

N-terminal zippering has occurred In order to test this,

five selected synthetic peptides were added to the

membrane fusion reaction mixture before or after

prein-cubation of t- and v-SNARE vesicles at 4C overnight

[3] SNN and SCC were selected as negative controls because they did not inhibit membrane fusion and neurotransmitter release (Figs 2 and 3)

Preincubation of t- and v-SNARE vesicle mixture at

4C overnight accelerated the membrane fusion reac-tion by  40% when the reaction was induced by increasing the temperature to 37C, consistent with previous results (Fig 4A; upper left) [3] When SCN, SynN and VpN were added before preincubation, membrane fusion was consistently inhibited However, the inhibitory effect of those peptides was much lower when they were added after preincubation SNN and SCC did not inhibit membrane fusion, regardless of preincubation Thus, N-terminal-mimicking peptides seem to inhibit membrane fusion by hindering N-ter-minal zippering When membrane fusion was complete, the amount of SNARE complex was measured by wes-tern blotting (Fig 4B) SNN and SCC did not change the amount of SNARE complex formed, regardless of preincubation However, there was a dramatic change

in the amount of SNARE complex depending on the addition time of SCN, SynN and VpN When SCN, SynN and VpN were added after preincubation they did not reduce SNARE complex formation much By contrast, these peptides inhibited SNARE complex for-mation very efficiently when added before preincuba-tion Therefore, N-terminal-mimicking peptides inhibited SNARE complex formation by hindering N-terminal zippering, which is consistent with the membrane fusion result

These experiments show that N-terminal-mimicking peptides do not inhibit membrane fusion if the N-ter-minals of t- and v-SNARE proteins are already zipped together, suggesting that the target of the peptides is newly forming SNARE complex In order to confirm this in PC12 cells, already-loaded neurotransmitters were depleted by pretreating with high concentrations

of K+before the addition of inhibitory peptides After pretreatment with high concentrations of K+, inhibi-tory peptides were added and cells were cultured for

24 h to regenerate synaptic vesicles By doing so, we are able to measure noradrenaline release from the newly docking vesicles only In Fig 4C, it is clearly shown that the noradrenaline release is dramatically

Fig 4 Time-dependent effects of inhibitory peptides on membrane fusion and neurotransmitter release (A) Synthetic peptides were added

to the membrane fusion reaction mixture before or after preincubation of t- and v-SNARE vesicles at 4 C overnight (B) After completion of the membrane fusion reaction in (A), western blot analysis was performed to measure the amount of SNARE complex formation +, peptide added after preincubation; ), peptide added before preincubation (C) Neurotransmitter release from PC12 cells was measured after depleting already-docked synaptic vesicles Already-docked synaptic vesicles were depleted by pretreating with a high K + concentration before the addition of synthetic peptides After addition of the peptides, PC12 cells were cultured for an additional 24 h for regeneration of the SNARE complex and neurotransmitter release was measured again (D) PC12 cells of (C) were subjected to western blot analysis to measure the amount of SNARE complex formation.

B

D

C

reduced by N-terminal-mimicking peptides after

pre-treatment with high K+ concentrations SCN, SynN

and VpN reduced noradrenaline release to the level of

verapamil, which may represent the minimum level of

stimulated neuroexocytosis SNN and SCC did not

inhibit noradrenaline release regardless of pretreatment

with high K+ concentrations Verapamil, a

calcium-channel blocker, did not have an additional inhibitory

effect on noradrenaline release Also, we were able to

confirm that SNARE complex formation was

dramati-cally reduced by N-terminal peptides, whereas other

peptides were almost neutral with regard to SNARE

complex formation in PC12 cells (Fig 4D) In

conclu-sion, N-terminal peptides inhibit newly forming

SNARE complex in PC12 cells and block the

regenera-tion of neuroexocytosis in synaptic vesicles We

specu-late that the neurotransmitter release and SNARE

complex band shown in Fig 3 were the sum of effects

derived from partially zipped SNARE complex and

newly forming complex By eliminating the effect

derived from already-zipped complex, the effect of

N-terminal peptides on SNARE complex formation

and neuroexocytosis could be more distinctive

Discussion

Depolarization of PC12 cells induces the influx of

Ca2+, leading to the fusion of synaptic vesicles which

are docked at the active zone of neurotransmitter

release [43] When the PC12 cells were depolarized with a short-pulse high concentration of K+, the syn-thetic peptides derived from the N-terminal regions of the SNARE motifs SCN, SynN and VpN inhibited [3H]-noradrenaline release in detergent-permeabilized PC12 cells (Fig 3) These results correlate strongly with the reduction in SNARE complex formation in the PC12 cells (Fig 3B,C), suggesting that the peptides inhibit [3H]-noradrenaline release by competing with parental SNAREs for complex formation

Recently, it has been shown that N-terminal nucle-ation of the SNARE complex can promote rapid mem-brane fusion [37] Conversely, one could predict that the disruption of N-terminal nucleation would affect the rate of membrane fusion significantly (route A in Fig 5) A variety of experiments including the lipid-mixing assay, the neurotransmitter release assay in PC12 cells and western blot analysis consistently con-firmed that SNARE-mediated fusion is profoundly affected by N-terminal patterned peptides Therefore,

it appears that the N-terminal regions of SNARE motifs are the potential targets for developing potent blockers of SNARE-mediated membrane fusion The relative inefficiency of other peptides in inhi-biting fusion can be rationalized in the context of the zipper model for SNARE complex formation [31,44– 47] The zipper model predicts that SNARE assembly starts in the N-terminal region and proceeds towards the C-terminal region One may envisage that a

B

C

D

A

X

X

Fig 5 Schematic presentation showing the effect of SNARE motif-patterned peptides

on the inhibition of membrane fusion (A) N-Terminal docking, which is believed to

be the rate-limiting step, is inhibited in the presence of N-terminal-mimicking peptide (e.g VpN in the figure), resulting in pro-foundly reduced SNARE-driven membrane fusion (B) In the absence of inhibitory pep-tide, SNARE complex drives membrane fusion (C) Middle- or C-terminal-mimicking peptides (e.g VpC in the figure) did not inhi-bit membrane fusion indicating that the pep-tide might be easily removed from t-SNARE complex [37] (D) In the presence of SynN, t-SNARE complex cannot accept VAMP2 for binding as in (A) Yellow, membranes; red, SNARE motif of Syntaxin or Syntaxin-pat-terned peptides; green, two SNARE motifs

of SNAP-25; blue, SNARE motif of VAMP2

or VAMP2-patterned peptides.

peptide bound to the middle or C-terminal region of

t-SNARE complex (route C in Fig 5) is peeled off

eas-ily as zippering progresses [37], resulting in no or only

moderate effects on membrane fusion (Figs 2–4)

A current model for SNARE assembly predicts that

VAMP2 in synaptic vesicles associates with the binary

complex of t-SNARE Syntaxin 1a and SNAP-25 in

presynaptic membranes Thus, it is likely that

VAMP2-patterned peptides bind to the t-SNARE

com-plex and compete with VAMP2 for binding Because

Syntaxin 1a and SNAP-25 can form a complex with

2 : 1 stoichiometry, it is likely that syntaxin-patterned

peptides also have their binding sites within

the t-SNARE complex When SynN is bound to the

1 : 1 t-SNARE complex, however, the structure of

the t-SNARE complex might change to that of the

2 : 1 complex, which is known to be a dead-end

prod-uct (route D in Fig 5) [48] Such a conformational

change would definitely inhibit the association with

VAMP2, and thus membrane fusion

By contrast, N- and C-terminal-mimicking peptides

from SNAP-25 (SNN and SCN, respectively) had

rather intricate effects on neuroexocytosis (Figs 2–4)

SNN did not inhibit SNARE complex formation at

all SNN might not bind to any of the SNARE

com-plexes involved in the SNARE assembly pathway

However, SCN was as effective as SynN and VpN

SCN may bind to the partially assembled t-SNARE

complex in which the C-terminal SNARE motifs of

SNAP-25 are not yet engaged [49], competing with the

binding of the C-terminal motif of SNAP-25

The ineffectiveness of several well-known peptides

that are known to regulate neuronal exocytosis is

inter-esting For example, a peptide mimicking the

C-termi-nal domain of SNAP-25 (corresponding to SCC in this

study) reportedly blocked Ca2+-dependent exocytosis

in chromaffin cells with an IC50 of 20 lm [20]

How-ever, we did not observe such an inhibitory effect of

SCC on SNARE assembly and neuroexocytosis in

PC12 cells (Figs 2–4) In addition, LIB, which was

selected from an a-helical peptide library of 137 180

sequences, was also not effective in the lipid-mixing

assay or the noradrenaline release and SNARE

com-plex formation assays in PC12 cells (Figs 2 and 3)

Previously, LIB was screened by measuring the

SDS-resistant complex formation with the purified SNARE

proteins We speculate that the discrepancy between

our results and those found previously on the efficacies

of SCC and LIB arises from differences in the

measurement scale of the assays used In a screening

process, determining the selection criteria depends on

the level of the background and experimental

condi-tions As such, a potential inhibitory peptide screened

from an experiment may not necessarily be reproduced

in other experiments using different assays or differently prepared cells Therefore, a fair comparison of the effectiveness of different SNARE-patterned peptides can only be made under identical experimental condi-tions In this study, the 13 representative peptides were tested under the same experimental conditions and their relative inhibition efficacies were ranked Our results showed that SCN, SynN and VpN are much more effective than any other SNARE-patterned peptides

It is also noteworthy that we were able to reproduce part of the previous results with LIB When LIB was tested for its inhibitory effect on SDS-resistant com-plex formation of purified cytoplasmic domains of SNARE proteins in vitro it reduced SNARE complex formation by 80% at 200 lm (data not shown) This

is comparable with previous results in which LIB inhibited SNARE complex formation by  95% at 0.5 mgÆmL)1 ( 250 lm) [24] However, LIB did not show any significant effects on membrane fusion and neurotransmitter release in our study (Figs 2 and 3) Interestingly, the inhibition of SNARE complex for-mation by SCN, SynN and VpN was less than that by LIB when purified SNAREs were used (data not shown), although these peptides were much more potent in inhibiting membrane fusion and release in PC12 cells These results suggest that the peptides screened by SDS-resistant complex formation using purified SNARE proteins might not correlate well with the inhibition of neuroexocytosis There is evidence that the membrane has crucial roles in SNARE com-plex formation and membrane fusion [14,50] Mem-branes may also restrict the geometry in which SNAREs assemble Thus, inhibition of SDS-resistant complex formation of purified soluble proteins does not necessarily mirror the inhibition of membrane-anchored SNARE complex formation

Because neuronal exocytosis is triggered on a milli-second time scale, synaptic vesicles are believed to be present at the active zone in a ready-to-fire state How-ever, we do not know in which state SNAREs are arrested before formation of the fusion pore: at the free protein stage, when N-terminal tips are already zipped or when the full complex is formed but the fusion pore is not yet made In Fig 4, we show that SCN, SynN and VpN inhibited N-terminal zippering and that the already-zipped complex was not inhibited

by those peptides (Fig 4A,B) Together with the fact that N-terminal peptides target newly forming SNARE complex (Fig 4C,D), we might rule out the possibility that SNARE proteins are arrested before fusion at the state of free proteins In conclusion, the inhibitory effects of the N-terminal-mimicking peptides on

neuro-transmitter release are very promising in the context of

their applications in pharmaceuticals and cosmetics

Experimental procedures

Materials

POPC, DOPS,

1,2-dioleoyl-sn-glycero-3-phosphoserine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) and

1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissamine rhodamine B

sulfonyl) (rhodamine-PE) were obtained from Avanti Polar

Lipids Inc (Alabaster, AL, USA) RPMI-1640, penicillin–

streptomycin, horse serum and fetal bovine serum were

purchased from GIBCO-BRL (Grand Island, NY, USA)

Triton X-100, 2-mercaptoethanol and all other chemicals

were purchased from Sigma-Aldrich Co (St Louis, MO,

USA) All peptides used (Fig 1A) were synthesized by

Peptron, Inc (Dajeon, Korea), and were ‡ 95% pure as

judged by MS Synthetic peptides were dissolved in distilled

water or dimethylsulfoxide depending on their solubility

Argireline [26] and LIB [23] were also synthesized as

controls

Expression and purification of recombinant

protein preparation

Protein expression and purification procedures for neuronal

SNARE proteins have been described previously [14] In

brief, the pGEX-2T vector encoding a thrombin-cleavable

N-terminal glutathione S-transferase tag was used for the

expression of the following constructs: full-length

syn-taxin 1a (amino acids 1–288), full-length VAMP2 (amino

acids 1)116) and SNAP-25 (amino acids 1)206) All

recombinant proteins were expressed in an Escherichia coli

CodonPlusRIL (DE3) (Novagen, Darmstadt, Germany)

All N-terminal glutathione S-transferase-tagged fusion

pro-teins were purified by affinity chromatography using

gluta-thione–agarose beads The protein was cleaved by thrombin

in cleavage buffer (50 mm Tris⁄ HCl, 150 mm NaCl,

pH 8.0) We added 1% n-octyl-d-glucopyranoside for

syn-taxin 1a and VAMP2 Purified proteins were examined by

12.5% SDS⁄ PAGE, and the purity was at least 90% for all

proteins (Fig 1B)

Reconstitution of SNARE proteins into

membranes

Large unilamellar vesicles with a diameter of 100 nm were

prepared as described previously [28] In brief, a mixture of

POPC and DOPS (molar ratio of 65 : 35) in chloroform

was dried in a vacuum and resuspended in buffer (50 mm

Tris⁄ HCl, 150 mm NaCl, pH 8.0) for a total lipid

concen-tration of 50 mm Protein-free large unilamellar vesicles

( 100 nm dia.) were prepared by extrusion through

poly-carbonate filters (Avanti Polar Lipids Inc., Alabaster, AL, USA) Syntaxin 1a and SNAP-25 were mixed at room tem-perature for 60 min to allow the formation of t-SNARE complex The t-SNARE complex was then mixed with the prepared large unilamellar vesicles at a 50 : 1 lipid⁄ protein molar ratio For the v-SNARE vesicles, 10 mm fluorescent liposomes, containing POPC, DOPS, 1,2-dioleoyl-sn-glycero-3-phosphoserine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) and rhodamine-PE at a molar ratio of 62 : 35 : 1.5 : 1.5, were mixed with VAMP2 at a 50 : 1 lipid⁄ protein ratio The liposome⁄ protein mixture was diluted twice, which brought the concentration of n-octyl-d-glucopyranoside below the critical micelle concentration After dialyzing against dialysis buffer (25 mm Hepes, 100 mm KCl, 5%

w⁄ v glycerin, pH 7.4) at 4 C overnight to remove detergent, the sample was treated with Bio-Beads SM2 (Bio-Rad, Hercules, CA, USA) to eliminate any remaining detergent The solution was then centrifuged at 10 000 g for 30 min to remove protein and lipid aggregates The final t-SNARE liposome solution contained  2.5 mm lipids and 1.9 mgÆmL)1 of protein, and the v-SNARE lipo-some solution contained  1 mm lipids and 0.25 mgÆmL)1

of protein The reconstitution efficiency was determined using SDS⁄ PAGE The amount of protein in the liposomes was estimated by comparing the band in the gel with that

of a known concentration of the same protein

SNARE-driven membrane fusion assay

The total lipid-mixing fluorescence assay has been described elsewhere [29] To measure total lipid mixing, v-SNARE liposomes were mixed with t-SNARE liposomes at a ratio

of 1 : 9 The final solution contained 1 mm lipids, in a total volume of 100 lL, in the presence of synthetic peptide Flu-orescence was measured at excitation and emission wave-lengths of 465 and 530 nm, respectively Changes in fluorescence were recorded with a Spectra Max M2 (Molec-ular Devices Inc., Palo Alto, CA, USA) fluorescence spec-trophotometer The maximum fluorescence intensity was obtained by adding Triton X-100 All lipid mixing experi-ments were carried out at 37C

PC12 cell culture

PC12 cells were purchased from the Korean Cell Line Bank (Seoul, Korea) The cells were plated onto poly-(d-lysine)-coated culture dishes and were kept in RPMI-1640 contain-ing 100 lgÆmL)1streptomycin, 100 UÆmL)1penicillin, 2 mm

l-glutamine, 10% heat-inactivated horse serum and 5% fetal bovine serum at 37C in a 5% CO2incubator Cell cultures were split once a week, and the medium was refreshed three times a week PC12 cells were treated with NGF (7S,

50 ngÆmL)1; Invitrogen, Carlsbad, CA, USA) for 5 days prior to [3H]-noradrenaline uptake and release