Sustained neurotransmitter release at synapses during high-frequency synaptic activity involves the mobilization of synaptic vesicles
(SVs) from the tightly clustered reserve pool (RP). Synapsin I (Syn I), a brain-specific peripheral membrane protein that undergoes
Deletion or mutational inactivation of the ALPS impairs the ability of Syn I to associate with highly curved membranes and with SVs.
in neurons and fails to induce vesicle clustering in vitro. Our data suggest a crucial role for ALPS-mediated sensing of membrane
Neurotransmission requires the exo-endocytic cycling of small
(40 nm) synaptic vesicles (SVs) (Murthy and De Camilli, 2003).
While the readily releasable pool of SVs resides at or near active
zones, a distantly located, tightly clustered reserve pool (RP) of
SVs is required for sustained neurotransmitter release during
high-level activity (Rizzoli and Betz, 2005). The organization of
the RP of SVs is linked to the peripheral SV-associated phospho-
protein synapsin I (Syn I) (Cesca et al., 2010; Shupliakov et al.,
Synapsins (I–III) are evolutionary conserved neuron-specific
proteins (Cesca et al., 2010). Syn I associates with SVs, partly via
resents a major component of purified SVs (Takamori et al.,
2006). Syn I is surface-active, covers a large proportion of the SV
et al., 1993; Pieribone et al., 1995; Rosahl et al., 1995; Ryan et al.,
Syn I dissociates from SVs upon stimulation, allowing mobi-
lization of SVs from the RP during activity (Hosaka et al., 1999;
Chi et al., 2001; Menegon et al., 2006; Cesca et al., 2010). After
cessation of the stimulus, Syn I reassociates with SVs, promoting
their reclustering within the RP (Chi et al., 2001; Menegon et al.,
tional cycle remain enigmatic. Syn I-binding proteins such as
CaMKII or Src are not exclusively associated with SVs and are
present in insufficient copy numbers (Cesca et al., 2010). Syn I
also directly and saturably associates with (Benfenati et al., 1989)
and clusters (Benfenati et al., 1993) acidic liposomes. Because of
into curved membranes via an amphipathic helix rich in hydro-
philic polar amino acids (Bigay et al., 2005; Antonny, 2011) have
been identified in several proteins (Drin et al., 2007), including
putative motifs in Syn II and Syn III. Here, we show that the
major synapsin isoform, Syn I, contains an ALPS motif that, by
sensing membrane curvature (MC), facilitates Syn I association
with SVs and regulates its vesicle clustering activity. A Syn I mu-
tant lacking the ALPS displays defects in its ability to associate
with highly curved membranes, including SVs, and to undergo
activity-induced cycles of dispersion and reclustering in primary
neurons. Moreover, mutant Syn I is partially defective in facili-
Author contributions: L.K., A.P., O.S., D.S., F.B., and V.H. designed research; L.K., A.F., V.K.B., A.P., F.O., M.F.,
(Torino) and Telethon-Italy (to F.B.), EU-FP7 [HEALTH-F2-2009-242167 (SynSys) to O.S.], the Swedish Research
TheJournalofNeuroscience,December7,2011 • 31(49):18149–18154 • 18149
for ALPS-mediated sensing of MC in regulating synapsin
Peptides and antibodies. Peptide sequences were as follows: rat Syn Ia
KASFLAATGGSSTG; Kes1p (7–29, yeast): SSSWTSFLKSIASFNGDLSS
LSA; AP180 (753–70, human): LGSDLSSLASLVGNLGI.
For single liposome assays, peptides were N-terminally tagged with
CAAP. Antibodies used were Syn I, bassoon (Synaptic Systems), synap-
tophysin (Sigma-Aldrich), and GFP (Stressgen).
Circular dichroism spectroscopy. Peptides (3 mM) were dialyzed into
circular dichroism (CD) buffer (10 mM Tris, pH 7.5, 150 mM NaF) and
measured on a Jasco J-810 spectrometer at 20°C. Spectra represent aver-
ages of 10 scans recorded from 194 to 240 nm. Control spectra were
Plasmids, generation of stable cell lines, protein purification. eGFP-Syn
Ia was purified from stably transfected doxycycline-inducible HEK293
cells. Pellets were resuspended in 20 mM HEPES, pH 7.4, 100 mM NaCl,
50 mM KCl, 5 mM MgCl2, 5 mM DTT, 0.3% CHAPS plus protease inhib-
mM NaCl, 2 mM DTT. Cleared supernatant was applied to GFP-TrapA
columns, washed, eluted (0.2 M glycine, pH 2.5), and neutralized. Purity
was assessed by measuring (A260/A280) and silver-stained SDS-PAGE.
Single liposome curvature assays. Single liposome binding assays were
done as in Bhatia et al. (2009) with liposomes of a SV-like composition
[69% DOPC (1,2-dioleoyl-sn-Glycero-3-Phosphocholine), 30% DOPS
(1,2-dioleoyl-sn-glycero-3-phospho-L-serine), 0.5% DOPE-Biotin (1,2-
Dioleoyl-sn-Glycero-3-Phosphoethanolamine-N-(Cap Biotinyl)), 0.5%
DID (1,10-dioctadecyl-3,3,30,30-tetramethyl-indodicarbocyanine per-
chlorate); Invitrogen]. For control experiments using eGFP-PH, lipo-
somes contained 5% PI(4,5)P2at the expense of DOPS. Samples were
analyzed as described previously (Bhatia et al., 2009).
Analysis of Syn I-induced liposome clustering. SV-like small liposomes
(Ø 60 nm) [phosphatidylcholine (PC):phosphatidylethanolamine (PE):
phosphatidylserine (PS):phosphatidylinositol (PI):cholesterol ? 27:30:
10:3:30; 1% rhodamine-PE] were prepared by dehydration/rehydration
in HEPES-buffered saline (HBS) (20 mM HEPES, pH 7.4, 200 mM NaCl)
followed by sonication in HBS for 10 min under argon flow while cool-
peptide (3 min, 37°C) before addition of 75 nM eGFP-Syn I (WT or ?).
Samples incubated for 10 min at 37°C were imaged by fluorescence mi-
croscopy and quantitatively analyzed using Volocity (Improvision).
Mean particle sizes were determined from ?10 fields of view (n ? 2
independent experiments). Data were analyzed statistically using a one-
way ANOVA with Bonferoni’s multiple comparison as a post-test.
al., 1986) or sucrose-loaded liposomes (PC:PE:PS:PI:cholesterol ? 27:
30:10:3:30) (Benfenati et al., 1989) were incubated for 1 h at 4°C with
wild-type (WT) or mutant eGFP-Syn I (40–200 nM) in 220 mM glycine,
30 mM NaCl, 5 mM Tris-HCl, 4 mM HEPES, pH 7.4, 0.22 mM NaN3, and
100 ?g/ml BSA. SVs or liposomes were reisolated by ultracentrifugation
(400,000 ? g, 30 min) through a 10% (w/w) sucrose cushion. Samples
were immunoblotted and recovery of14C-PC was measured.
Optical recordings in primary neurons. Hippocampal neurons from
lipofectamine 2000, fixed (4% PFA 4 d posttransfection), and immuno-
(3–5 d posttransfection) as described previously (Menegon et al., 2006).
Fluorescence changes of eGFP-Syn Ia at synaptic boutons were normal-
ized to the starting fluorescence (F0). Time constants and plateaus for
dispersion and reassociation were determined by fitting the respective
curves with a single exponential function.
BLAST searches for presynaptic proteins that contain potential
ALPSs (Drin et al., 2007) identified Syn Ia/b, a protein known to
2001), i.e., high curvature membranes (Fig. 1A). The putative
ALPS within mammalian Syn Ia corresponds to amino acids
69–96 belonging to its domain B and is rich in polar hydrophilic
and hydrophobic residues (Fig. 1A,B). Sequences with similar
features are found within rat or human Syn I, Syn II, and Syn III
and within Syn I from lamprey and Xenopus, suggesting that the
ALPSs adopt a random conformation in solution, but fold
18150 • J.Neurosci.,December7,2011 • 31(49):18149–18154Krabbenetal.•CurvatureSensingbySynapsin
(Drin et al., 2007). To test whether the putative ALPS within Syn
I is able to sense MC, we incubated a peptide corresponding to
amino acids 69–96 of rat Syn I (WTALPS) with liposomes of dif-
ferent diameters and analyzed its secondary structure by circular
tide displayed a random conformation (Fig. 1C, black line). The
?-helical (negative ellipticity ?202 nm and a strong increase in
positive ellipticity at 195 nm) and random coil conformation
(Fig. 1C, red line). Simulation of the CD spectrum predicted
?30% ?-helical content, consistent with a weak association of
ALPS with membranes (Antonny, 2011). The CD spectrum
therefore likely represents a weighted average of vesicle-bound
and free soluble Syn I-derived peptide. No induction of helix
formation was observed with 400 nm liposomes (Fig. 1C, blue
line), indicating that the Syn I-derived peptide indeed is able to
To corroborate this, we synthesized a peptide in which S78
and Q86 were exchanged for helix-breaking proline residues
tide (scr). 2PALPSand scr appeared unstructured in CD with or
without high-curvature liposomes (Fig. 1D,E). Both peptides
thus lack the ability to undergo high-
The above data indicate that the Syn
I-derived ALPS peptide is able to sense
MC. To test whether full-length Syn Ia
can sense MC, we purified recombinant
eGFP-Syn Ia from stably transfected
HEK293 cells (Fig. 2A). eGFP-Syn Ia was
soluble and able to bind and cluster lipo-
sense MC was then analyzed using quan-
titative fluorescence imaging of single li-
posomes with various diameters and MC
(Bhatia et al., 2009). The pleckstrin ho-
mology domain (PH) of phospholipase
C?, a phosphatidylinositol 4,5-bisphos-
phate binding domain unable to sense
curvature (Bhatia et al., 2009), served as
the negative control. Quantitative confocal
imaging was used to record the fluores-
cence intensity of eGFP-tagged proteins
colocalized with DID-labeled liposomes
of various sizes. Plots of the relative
some size revealed a remarkable prefer-
ence of WT eGFP-Syn I for liposomes
ability to sense MC (Fig. 2B), whereas
eGFP-PH bound equally well to lipo-
somes of all sizes (Fig. 2C). Two mutant
variants of Syn I in which the putative
curvature-sensing ALPS was either deleted
(?ALPS) or mutationally inactivated (2P-
ALPS) showed a greatly reduced ability to
associate with high-curvature liposomal membranes (Fig.
2D,E) or with small SV-like membrane vesicles mimicking
size and phospholipid composition of native SVs (Fig. 2H).
The fluorescently labeled Syn I-derived ALPS peptide fully
retained the ability to sense MC, similar to holo-Syn I (Fig. 2F).
The 2PALPSpeptide or a peptide carrying substitutions in the
MC, resulting in their inability to stably associate with liposomal
membranes (Fig. 2G).
The limited residual curvature-sensing ability of ?ALPS- and
2P-ALPS mutant Syn I (Fig. 2D,E) may be explained by other
potential curvature-sensitive elements not encoded within the
identified ALPS. In line with this, multiple determinants within
domains B, C, and E cooperate in targeting Syn I to presynaptic
A central function of Syn I is to organize the RP of SVs via its
liposome binding and clustering is saturable (Benfenati et al.,
1993) and depends on Syn I oligomerization via C- and
Syn I (D); 2P-ALPS-eGFP-Syn I (E); Atto-labeled Syn I-derived ALPS peptide (WTALPS; F). Plots display protein/peptide density
liposomes (Ø 50–70 nm; 20 ?g of phospholipid/sample). Bound Syn I (left) is expressed in percentage of total Syn I added
Membrane curvature sensing and liposome binding by Syn I involves an ALPS. A, Silver-stained gel illustrating
Krabbenetal.•CurvatureSensingbySynapsinJ.Neurosci.,December7,2011 • 31(49):18149–18154 • 18151
E-domain-mediated interactions (Hosaka and Su ¨dhof, 1999;
vature sensing in synapsin function, we monitored the ability of
WT and ALPS mutant Syn I to cluster liposomes mimicking the
size and lipid composition of native SVs in vitro. Purified WT or
mutant eGFP-Syn Ia (?ALPS) was incubated with fluorescently
labeled SV-like liposomes and analyzed by fluorescence micros-
larger assemblies. Mutant eGFP-Syn I (?ALPS) lacking its ALPS
gesting that ALPS-mediated curvature sensing modulates the
further support of this, addition of ALPS peptides derived from
mammalian (WTALPS) or lamprey (WTlALPS) Syn I potently in-
hibited Syn Ia-induced liposome clustering, whereas a mutant
version of the ALPS peptide unable to sense curvature (2PALPS;
icle clustering was also inhibited by a known ALPS peptide de-
taken from AP180. Thus, the ability to sense MC is critical for
peptide competition of Syn I-induced liposome clustering.
Syn I reversibly binds to small SVs and undergoes activity-
dependent association/dissociation cycles (Cesca et al., 2010). It
is thus conceivable that MC sensing participates in the ability of
synapsin to reassociate with SVs following its axonal dispersion.
ing SV binding, highly purified rat SVs depleted of endogenous
synapsin (Schiebler et al., 1986) were incubated with WT eGFP-
Syn I or mutants, in which the ALPS was either deleted (?ALPS)
or inactivated (2P-ALPS). While WT-Syn I avidly bound to SVs
in a saturable and concentration-dependent manner, 2P-ALPS-
Syn I displayed a strongly reduced SV binding ability. The
?ALPS-mutant virtually failed to associate with SV membranes
(Fig. 4A,B). Hence, the ability of Syn I to associate with native
SVs strongly depends on the presence of an intact ALPS.
Syn I association with SVs is critically involved in its activity-
induced axonal dispersion and presynaptic reclustering, a func-
tion linked to the organization of the SV RP (Chi et al., 2001;
Menegon et al., 2006). To assess the role of Syn I-mediated cur-
and mutant Syn I were observed (Fig. 4D, left). eGFP-Syn Ia
aptic boutons positive for the active zone marker bassoon (Fig.
4C). A similar, although significantly less intense, presynaptic
staining was seen for ?ALPS-eGFP-Syn I (Fig. 4D, right). Thus,
targeting of Syn I, although it improves targeting efficacy.
Synaptic activity induces the axonal dispersion of Syn I, re-
Syn I-mediated vesicle clustering is modulated by ALPS-mediated membrane curvature sensing. A, Representative images of liposome (Lipos) clusters induced by WT or mutant
18152 • J.Neurosci.,December7,2011 • 31(49):18149–18154Krabbenetal.•CurvatureSensingbySynapsin
D, Reduced presynaptic expression of eGFP-Syn Ia (?ALPS). Left, Expression level of WT and ?ALPS eGFP-Syn Ia quantified by total GFP fluorescence (mean ? SEM; n ? 2 independent
Krabbenetal.•CurvatureSensingbySynapsinJ.Neurosci.,December7,2011 • 31(49):18149–18154 • 18153
boutons due to its rebinding to small SVs (Chi et al., 2001; Cesca
et al., 2010). We monitored the distribution of eGFP-Syn I by
optically recording fluorescence changes elicited by neuronal ac-
tivity within presynaptic boutons. Chemical stimulation with
KCl caused a loss of fluorescence intensity of presynaptic eGFP-
Syn I with a half-time (?) of ?15 s, similar to Chi et al. (2001)
(Fig. 4E,F), reflecting its axonal dispersion. Following cessation
of the stimulus, eGFP-Syn I fluorescence recovered with a ? of
18–20 s to reach ?80% of its initial value (Fig. 4E,G). A much
less pronounced activity-induced dispersion was observed for
however, indistinguishable between WT and mutant synapsin (?
severely reduced ability to recluster within presynaptic boutons
geting and impaired activity-induced dispersion of ?ALPS-Syn I
were also seen in neurons from Syn I/II/III?/?/?mice (Fig. 4H),
indicating that these phenotypes do not reflect differential asso-
ciation of expressed WT versus mutant Syn I with endogenous
synapsins. However, recovery of ?ALPS-Syn I was only slightly
and nonsignificantly impaired compared to WT-Syn I, suggest-
ing that the absence of competition by endogenous ALPS-
containing synapsins might facilitate reassociation with SVs.
Thus, the ability of Syn I to sense MC via its ALPS critically
activity-induced dispersion and reclustering. These observations
bind to SVs (Fig. 4A,B).
We have identified and functionally characterized an evolutionary
conserved ALPS within the SV-associated protein Syn I. Our data
indicate that ALPS-mediated curvature sensing is required for the
cles and native SVs. Furthermore, we demonstrate a crucial func-
tional role for ALPS-mediated recognition of MC in regulating the
ability of Syn I to induce vesicle clusters in vitro and to reversibly
associate with SVs in live neurons. Our data provide a mechanistic
explanation for the selective association of Syn I with small SVs
for MC sensing by Syn I in presynaptic membrane organization.
Induction and sensing of MC by exo-endocytic proteins is a key
cling of SV membranes (McMahon et al., 2010). The results re-
Syn with SV membranes, and thus, to the Syn-dependent orga-
et al., 1996; Cesca et al., 2010). Since Syn I undergoes cycles of
ation/dissociation (Cesca et al., 2010), it is tempting to speculate
that ALPS-mediated curvature sensing by Syn I might be modu-
lated by phosphorylation-induced conformational changes.
Interestingly, ?-synuclein, a presynaptic SV-associated pro-
its overexpression inhibits neurotransmitter release due to de-
fects in SV reclustering (Nemani et al., 2010), similar to loss of
Syn I (Rosahl et al., 1995; Ryan et al., 1996). Together with our
findings reported here, these results are consistent with the hy-
pothesis that SV clustering is regulated by curvature sensors on
the SV membrane, potentially regulating SV pool sizes, synaptic
transmission, and short-term plasticity.
Antonny B (2011) Mechanisms of membrane curvature sensing. Annu Rev
Benfenati F, Greengard P, Brunner J, Ba ¨hler M (1989) Electrostatic and hy-
drophobic interactions of synapsin I with phospholipid vesicles. J Cell
Benfenati F, Valtorta F, Rossi MC, Onofri F, Sihra T, Greengard P (1993)
Interactions of synapsin I with phospholipids: possible role in synaptic
vesicle clustering and in the maintenance of bilayer structures. J Cell Biol
Bhatia VK, Madsen KL, Bolinger PY, Kunding A, Hedegård P, Gether U,
Stamou D (2009) Amphipathic motifs in BAR domains are essential for
membrane curvature sensing. EMBO J 28:3303–3314.
BigayJ,CasellaJF,DrinG,MesminB,AntonnyB (2005) ArfGAP1responds
to membrane curvature through the folding of a lipid packing sensor
motif. EMBO J 24:2244–2253.
Cesca F, Baldelli P, Valtorta F, Benfenati F (2010) The synapsins: key actors
of synapse function and plasticity. Prog Neurobiol 91:313–348.
Chi P, Greengard P, Ryan TA (2001) Synapsin dispersion and reclustering
during synaptic activity. Nat Neurosci 4:1187–1193.
Drin G, Casella JF, Gautier R, Boehmer T, Schwartz TU, Antonny B (2007)
A general amphipathic alpha-helical motif for sensing membrane curva-
ture. Nat Struct Mol Biol 14:138–146.
Gitler D, Xu Y, Kao HT, Lin D, Lim S, Feng J, Greengard P, Augustine GJ
(2004) Molecular determinants of synapsin targeting to presynaptic ter-
minals. J Neurosci 24:3711–3720.
HosakaM,Su ¨dhofTC (1999) Homo-andheterodimerizationofsynapsins.
J Biol Chem 274:16747–16753.
McMahonHT,KozlovMM,MartensS (2010) Membranecurvatureinsyn-
aptic vesicle fusion and beyond. Cell 140:601–605.
Menegon A, Bonanomi D, Albertinazzi C, Lotti F, Ferrari G, Kao HT, Benfe-
natiF,BaldelliP,ValtortaF (2006) ProteinkinaseA-mediatedsynapsin
I phosphorylation is a central modulator of Ca2?-dependent synaptic
activity. J Neurosci 26:11670–11681.
Monaldi I, Vassalli M, Bachi A, Giovedì S, Millo E, Valtorta F, Raiteri R,
Benfenati F, Fassio A (2010) The highly conserved synapsin domain E
mediates synapsin dimerization and vesicle clustering. Biochem J
Murthy VN, De Camilli P (2003) Cell biology of the presynaptic terminal.
Annu Rev Neurosci 26:701–728.
Nemani VM, Lu W, Berge V, Nakamura K, Onoa B, Lee MK, Chaudhry FA,
Nicoll RA, Edwards RH (2010) Increased expression of alpha-synuclein
reduces neurotransmitter release by inhibiting synaptic vesicle recluster-
ing after endocytosis. Neuron 65:66–79.
Pieribone VA, Shupliakov O, Brodin L, Hilfiker-Rothenfluh S, Czernik AJ,
Greengard P (1995) Distinct pools of synaptic vesicles in neurotrans-
mitter release. Nature 375:493–497.
Pranke IM, Morello V, Bigay J, Gibson K, Verbavatz JM, Antonny B, Jackson
CL (2011) Alpha-synuclein and ALPS motifs are membrane curvature
sensors whose contrasting chemistry mediates selective vesicle binding.
J Cell Biol 194:89–103.
Rizzoli SO, Betz WJ (2005) Synaptic vesicle pools. Nat Rev Neurosci
Rosahl TW, Spillane D, Missler M, Herz J, Selig DK, Wolff JR, Hammer RE,
MalenkaRC,Su ¨dhofTC (1995) EssentialfunctionsofsynapsinsIandII
in synaptic vesicle regulation. Nature 375:488–493.
Ryan TA, Li L, Chin LS, Greengard P, Smith SJ (1996) Synaptic vesicle re-
cycling in synapsin I knock-out mice. J Cell Biol 134:1219–1227.
Schiebler W, Jahn R, Doucet JP, Rothlein J, Greengard P (1986) Character-
ization of synapsin I binding to small synaptic vesicles. J Biol Chem
Shupliakov O, Haucke V, Pechstein A (2011) How synapsin I may cluster
synaptic vesicles. Semin Cell Dev Biol 22:393–399.
Schenck S, Bru ¨gger B, Ringler P, Mu ¨ller SA, Rammner B, Gra ¨ter F, Hub
JS, De Groot BL, Mieskes G, Moriyama Y, Klingauf J, Grubmu ¨ller H,
HeuserJ,etal. (2006) Molecularanatomyofatraffickingorganelle.Cell
JC, Stamou DG, Steven AC, Langen R (2010) Membrane curvature in-
duction and tubulation are common features of synucleins and apolipo-
proteins. J Biol Chem 285:32486–32493.
18154 • J.Neurosci.,December7,2011 • 31(49):18149–18154Krabbenetal.•CurvatureSensingbySynapsin