The synaptotagmin (Syt) family comprises evolutionarily conserved ...

J. Biochem.

Calcium-Dependent

and

Synaptotagmin

-Independent

128,637--645

Hetero-Oligomerization

in

(2000)

the

Family1

Mitsunori

Fukuda*,2

and

Katsuhiko

Mikoshiba*,•õ,•ö

* Laboratory for Developmental Neurobiology Ch , Brain Science Institute, RIKEN (The Institute of Physical and emical Research), 2-1 Hirosawa, Wako , Saitama 351-0198; •õDepartment of Molecular Neurobiology, The Institute of Medical Science, The University of Tokyo , 4-6-1 Shirokanedai, Minato-ku, Tokyo 108-8639; and •öCalciosignal Net Project, Exploratory Research for Advanced Technology (ERATO) 2 -28-8 Honkomagome , Japan Science and Technology Corporation (JST), , Bunkyo-ku, Tokyo 113-0021 Received

June

29, 2000;

accepted

Synaptotagmins

constitute

transmembrane

region

have

indicated

of function

that

totagmin cial

the

in

V, VI,

minal

and

Mikoshiba, show

K.

(1999)

moderate are

K.

(2000)

apparent

can

hetero-oligomerize that create

Key

words:

J.

Biol.

of Chem.

the

with

J. Biol.

Chem.

weak

VII

Ca2+-dependent a variety Ca2+

31421-31427].

C2

EC50

values

275,

activity,

of but and/or

presence (Syts

has about some

exocytosis,

This work was supported in part by grants from the Science and Technology Agency of Japan (to K.M.) and from the Ministry of Ed ucation, Science, Sports and Culture of Japan [to K.M., and M.F. (11780571 and 12053274)]. 2 To whom correspondence should be addressed . Tel: +81-48-4679745, Fax: +81-48-467-9744, E-mail: [email protected] Abbreviations: cyto, cytoplasm; HRP, horseradish peroxidase; lum, lumen; Mr, molecular weight; N, amino; N-Gly, N-glycosylation; PAGE, polyacrylanvde gel electrophoresis; PCR, polymerase chain reaction; Syt(s), synaptotagmin(s). c 2000 by The Japanese Biochemical Society. 637

A Syts an

no

VIII,

N-ter E.,

IV, VIII, sets

-independent

self-oligomerization,

Ca2+

Group

homo

XI did

of synaptotagmin manner.

XI)

C syn

homo-oligomer

[Fukuda,

and

and

and

Ca2+-dependent

150ƒÊM Syts

(I, II,

or

at

Kanno,

of Ca2+. IX)

unique

hetero-oligomerization

The synaptotagmin (Syt) family comprises evolutionarily conserved proteins that are thought to be involved in vesic ular trafficking including synaptic vesicle exocytosis (re viewed in Refs. 1-3). They are characterized by a short amino (N)-terminus, a single transmembrane region, and two highly conserved C2 domains (named the C2A and C2B domains) in the large carboxyl terminus. The C2 do mains of Syt I, a well characterized isoform, are essential for synaptic vesicle exocytosis and endocytosis (4-9 and reviewed in Refs. 1-3) and this region is likely to be func tional in all synaptotagmin family proteins. To date, 12 syn-1

Vol. 128, No. 4, 2000

D)

28180-28185].

a Ca2+-dependent

domain,

of the

M.,

B Syts

all the proteins based on

Group

bonds

[Fukuda,

Group

Ca2+-dependent

and -independent of Ca2+-sensors.

sensor,

of Ca2+

disulfide

hetero

the synap an artifi

examined family groups

activity.

by

one

expression

about

study

oligomerization

presence

(Group

hetero-oligomerization in

for

is known

, and systematically among synaptotagmin family into four distinct

irrespective

Syt

studies

, little

hetero-oligomers

274,

by

IV).

by

biochemical

domain

and

irrespective

this

characterized

, we showed that (Nlumen/Ccytoplasm) by introducing

-independent

homo-

properties

any

gest may

and

strong

(Syt

oligomerization

In

are and

I is important

However

protein

N-terminal

characterized

activity

Mikoshiba,

family.

that genetic

(Syt)

vesicles.

homo-oligomerization

aptotagmins ization

the

Ca2+-dependent motif

Recent

of synaptotagmin of synaptic

at

proteins

domains.

of hetero-oligomerization the synaptotagmin

X) form

cysteine

of membrane C2

I membrane

site

possible combinations (Syts I-XI). We classified (III,

two

synaptotagmin

is a type

N-glycosylation

differences

a family and

exocytosis in

family

2 , 2000

oligomerization

during

oligomerization

August

M., not

and show

isoforms Our

data

sug

of synaptotagmins

synaptotagmin.

aptotagmin isoforms have been described in rat and mouse (Ref 10 and references therein), and 7 and 5 isoforms respectively in Drosophila and Caenorhabditis elegans (11, 12). These isoforms can be divided into subclasses based on the phylogenetic relationship (10), gene structures (13) and differences in the biochemical natures of the C2 domains, such as Ca2+-dependent phospholipid or syntaxin binding to the C2A domain (14-16), and inositol 1,3,4,5-tetrak isphosphate binding to the C2B domain (17-19). Syts III, V, VI, and X form a small branch in the phyloge netic tree (10). This class of synaptotagmins form hetero oligomers through the N-terminal cysteine motif by disul fide bonding (10) and, except for Syt III, interact with Syt VII in a Ca2+-dependent manner at the cytoplasmic domain (20). Syts III, V, VI, and X are also characterized by weak or no inositol 1,3,4,5-tetrakisphosphate binding activity (19). Syts IV and XI are another class of synaptotagmins, char acterized by deficient phospholipid (phosphatidylcholine/ phosphatidylserine, 1/1 liposome) binding due to the substi tution of Ser for Asp at position 244 (or 247) of Syt IV (or XI) (14, 16). Syts I, II and IX also form a small branch in the phylogenetic tree (10). Although the functions of the other classes of synaptotagmins largely remain unknown,

638

M. Fukuda

Syt

I (or

II) is thought

probably the basis min

mutants

studies tion

(reviewed

(20, 22-26). properties:

erization the

to function

in

Syt

I (or

20),

mine

we

isoforms

21)

and

classes

(10,

20,

that

self-oligomeriza

just

multim

and

by the C2B also showed

domain (22Ca2+-depen

the

be

properties

interesting

would

In the

present

tematically

study,

we

determined

isoforms

examined

the

(Syts

homo-

properties of synaptotagmin two different epitope (T7

and

the

membrane

III-XI),

and

topo then

sys

EXPERIMENTAL

enzymes

were

CA, USA) Polyclonal

I-XI)

by using synaptotag

obtained

and and

tively.

from

other

from Sigma

Biosystems

from

chemicals

were

cel A10

(Millipore

System

Construction

of

Bedford,

the

T7-tag

underlined

in bold

letters]

was

MA,

template

(10) by

following

oligonucleotides:

tions

were

naturation extension

and

and

for

at 94•Ž

for 30s, for 30s.

gel by

PCR

sequencing

Other

plasmids

and

the

X, and full

were

length

excised

Madison, NotI

WI, site

pEF-T7(or

prepared DNA was

using

a Hitachi

in.

day

before

XI were

similarly

from

the by

of pEF-BOS

or a QIAGEN

Maxi

prep

then

substituted IX

pGEM-T NotI

DNA

Easy

(10).

30)

I-XI

verified

-Syts

previously (10, using Wizard-mini

by

this

method. tag

I-XI-cyto

dish) as

F

buffer

to

(5

was

described cells were

and

10%

and

cells

SDS-

immuno

of

pEF-T7-

x

105 cells,

carried

out

previously harvested,

by (10). and

with buffer containing 1% Triton 50mM HEPES-KOH, pH 7.2, 0.1 fluoride, 10ƒÊM leupeptin, and

pepstatin

by anti-T7 tag antibody-conjugated and immunoblotting analyses were

A at

4•Ž

previously

Membrane Synaptotagmin brane

for

1h.

Immunoprecipitation

of

agarose, SDSalso performed

(10).

Topology family

region

Syts

with

I and

of Mouse Synaptotagmins proteins have a single

a short

containing

N-terminus

tandem

II have

a single

N-terminus,

and

this

C2

the

N-terminal

domain

domain

of Syt

domains.

N-glycosylation been

has

carboxyl

In vertebrates, shown

to

F digestion (27, 29) In addition, antibodies I were

at

be

N-

or siteagainst into

hip

after membrane depolar of antibodies against the

squid

giant

presynapse

proteins, orientation.

showing Nlumen (or extracellular domain)/Ccytoplasm (Nlum/Ccyto) However, the Syt isoforms other than Syts I

II do

minus therefore

we

Gly-Ser)

the into

is no same

site

region

III-XI)

membrane -XI proteins

in such

cells

N-terminal

each

by

(Fukuda tagging

topology T7-Syts were transiently

immunoprecipitated

the

other To

N-ter

(10),

and

isoforms

address

this

site

(Asn-

N-glycosylation of

PC12

at

synaptotagmin

iso

(Fig . 1A). We first confirmed (T7 , T7-N-Gly or FLAG) to the N the subcellular localization of

of a short tag did not affect

isoforms so that

that

orientation.

artificial

N-terminus

ata),

evidence

Nlum/Cyto

the

several

the synap I membrane

N-glycosylation

direct an

results, type

transmembrane

(T7-N-Gly-Syts

III-

and

an

of the

introduced

were

the

contain

there

show

issue, form

not

upstream

on these to comprise

visualized

vesicles family

and

(4). Based is thought

incorporated

site

synaptic totagmin

that addition -terminus

20). The plasmid preps (Promega)

I XItransmem

a large

site

of Syt

I into

and

potential

glycosylated by N-glycosidase directed mutagenesis (31).

into

(Promega;

subcloned

sample

(PAGE)

COS-7

then

boiling

was added to for 1h at 37•Ž.

T7-Syts PAGE,

also III-

a T7-N-Gly

(pEF-T7-N-Gly-Syts and

for The

sequencer.

vector and

with

Syts by

with

digestion

(29,

kits.

was

constructed inserts

FLAG)Syts

as described prepared by

and

T7-N-Gly-tagged

synaptotagmin USA)

digested

by

of N-glycosidase

SDS

transfection/10cm method transfection,

of 20 CA-

A, and

subjected

electrophoresis elsewhere (10).

C2A

and from

were

unit

Germany) incubated

into

described

100ƒÊl IGEPAL

pepstatin

10ƒÊM

as described

as

in 0.5%

Procedures-Co-transfection

Reac

Pharma

UK),

one

were

pEF-FLAG-Syts

dish)

antibody-conju

denaturation

by adding

Proteins

of de

purified

(Amersham

SQ-5500

the

37•Ž,

COS-7

phenylmethylsulfonyl

10ƒÊM

pocampal neurons by endocytosis ization (32-34), and microinjection

for 30s,

products,

IX plasmid

encoding

the

5'-CGGA

consisting

of pGEM-T-T7-Syt

pGEM-T-T7-N-Gly-Syt

DNA

using

and

at 60•Ž

Column above)

insert

each

annealing The

(italics

site

IX as a

(PCR)

(sense)

cycles,

to

in

Novagen)

After

Mannheim, mixtures were

polyacrylamide gel blotting as described Syts

and

stopped

3

tag

0.1mM

microtubes.

were for

anti-T7

resuspended 0.1% SDS,

EDTA,

cooling

expressed

transfeetion/10cm

30ƒÊl,

leupeptin

two and

Reactions

the

(N-Gly)

reaction

Buckinghamshire, BamHI

III-

(antisense).

25

MicroSpin

ApaI-BamHI

reagent

pGEM-T-T7-Syt

chain

out

Biotech;

resulting

XI).

3min,

terminus

N-glycosylation

T7 primer

carried

of

[MASMTGGQQMG

from

polymerase

at 72•Ž

agarose

Apal

the

10ƒÊM into

II, T7-

RESULTS

water prepared and Milli-Q Bio

Synaptotagmins

TCCGTTGCGACCCATTTGCTGTCC-3'

The

fluoride,

20mM

I,

All

USA).

T7-N-Gly-tag

amplified

USA).

products

T7-N-Gly-Tagged

encoding

WI,

made up in deionized Purification System

RNGS;

VIII,

(Sigma),

volume

anti-T7

(Madison,

commercial

were Water

City,

(San Fran USA), respec

(HRP)-conjugated

Novagen

grade. Solutions with an Elix10

(Foster

(Tokyo), respectively (M2) against FLAG

before

Beads were pH 7.2,

proteins were solubilized X-100, 250mM NaCl, mM phenylmethylsulfonyl

andand polymerase restriction

Zymed Laboratories (St. Louis, MO,

peroxidase

was

XI-cDNA

PE

Toyobo Biochemicals monoclonal antibodies

Horseradish antibody

polymerase

III-XI

day

by

(wet

Miscellaneous

PROCEDURES

DNA

peptide were obtained cisco, CA, USA) and

the

630


family (Syts and FLAG)-tagged

the

the DEAE-dextran Three days after

Materials-AmpliTaq

cia

agarose

previously (10). mM HEPES-KOH,

(Roche Diagnostics, one tube, and the of

T7-Synaptotagmins III-XI-Recombinant

immunoprecipitated

for

variety

of

T7 N-Gly-Syts

other

a

Digestion

105 cells,

deter-

mins.

an

II, and

boiling

of synaptotagmin

tag

(5 x

to

Ca2+-sensors. logy

I,

cells

divided

of differ

create

Syts

with


synaptotagmins

F

T7-N-Gly-Synaptotagmins

gated of

a Ca2+-de-

hetero-oligomerize

and

were

downstream

27, 28)

it would

N-Gyyeosidase

on

biochemical

Ca2+-independent

I or II can

whether of

in vitro two

state,

release synaptotag


thought Syt

and

mediated isoforms

-independent

whether

ent

oligomerized

by a region

region

pendent oligomerization 26). Since several other (10,

an

11) shows

mediated

transmembrane

and

Ref

A SDS-resistant

probably

dent

in

as a Ca2+-sensor for neurotransmitter of the genetic analysis of Drosophila

and K. Mikoshiba

, M., unpublished d is unlikely to affect

I , II, and expressed

anti-T7

tag

T7-N-Gly-Syts III in COS -7 cells antibody-conju-

J. Biochem.

Ca2+-Dependent and -Independent Oligomerization of Synaptotagmins

Fig. 1. Determination of the membrane topology of synap totagmins I-XI. (A) Schematic representation of T7-N-Gly-tagged Syt IX.Top indicates the nucleotide and amino acid sequences around the T7-N-Gly-tag. T7-tag and N-glycosylation sites (N-Gly) are shown by open box and double lines, respectively. Non-coding regions and the major restriction enzyme sites are shown by small letters and underlining, respectively. Nucleotide and amino acid numbers are given on both sides. Bottom is a schematic representation of T7-N-Gly-Syt IX: T7-tag (black box), N-glycosylation (Y), transmembrane region (TM; open box), and two C2 domains (C2A and C2B; hatched boxes). (B) N-

gated showed

agarose. Although several isoforms heterogeneous bands which probably

translational

modifications,

(M,)

of

expressed

that

of T7-Syts

N-Gly-Syts

the

III-XI

III-XI

(data

were

were

then

treated

cleaves

the N-linked

sugar,

without

all

As

isoforms

cosidase

and

shown

decreased

when These

forms

via the

C2

co-expression

assay

T7-

and

the

association

these

immunoprecipitation vivo when

This

assay

system

T7-

and

in

the

FLAG-Syts

Vol. 128, No. 4, 2000

by

apparent

treated

Mr of

with

N-gly show

that

studies

because

on iso

all

isoforms

Hetero-Oligomeriza examine

the

Ca2+-de

hetero-oligomerization of synap used a T7- and FLAG-tagged Syts described

previously

co-expressed two

in the

interactions

with

synaptotagmin

meaningful

were of

isoform compared

indicate

-Independent

as

FLAG-Syts

which

synaptotagmins

Isoforms-To

and -independent isoforms, we

the

F,

topology.

and

of Synaptotagmin

pendent totagmin

Ca2+.

are

membrane

Ca2+-Dependent tion

were

between

domains

same

this,

1B, the

results


show

T7-

To confirm

was

of the

than that

the Mr of each

they all

weight

larger

N-glycosidase

Fig.

VII) post

suggesting

treatment in

that

Nlum/Ccyto orientation. the

with F

F, indicating

was

not shown),

N-glycosidase

immunoblotting.

molecular

III-XI

N-glycosylated.

beads or

apparent

T7-N-Gly-Syts

(e.g., Syt reflect

in

proteins

presence

has cells

were

was merits.

to

be

separately

Briefly, cells

and

evaluated

or absence

three seem

(10). COS-7

by

of 500ƒÊM

First, detected, expressed

only

in

because and

639

glycosidase F treatment of T7-Syts I, II and T7-N-Gly-Syts III-XI. Anti-T7 tag antibody immunoprecipitants were treated with or with out N-glycosidase F and analyzed by immunoblotting with HRP-con jugated anti-T7 tag antibody (1:1,000 dilution). Note that the apparent M, of all synaptotagmin isoforms was decreased by N-glycosi dase F treatment. A proportion of N-linked sugars of T7-N-Gly-Syts VIII and IX were further converted to a complex form because they showed broad bands and were resistant to endoglycosidase H treat ment (data not shown). The positions of M, markers (x10-3) are shown on the right.

mixed after solubilization, T7- and FLAG-Syts interactions were in many cases not detectable (10, 20). Second, since native synaptotagmins undergo posttranslational modifica tions such as palmitoylation (35, 36) or disulfide bond for mation (10), it would be better to use a mammalian rather than bacterial expression system. Third, significant levels of endogenous Syts I-XI are not expressed in COS-7 cells (Fukuda, M., unpublished data), so that the involvement of endogenously expressed Syts in oligomerization need not be considered. The only problem with this assay is that the expression levels of the two isoforms may be very different. The amounts of each immunoprecipitant from each isoform were different even when the same amount of plasmid was transfected (Fig. 2). For instance, Syts I, II, VIII, and IX were expressed at relatively high levels in COS-7 cells, Syts VII and XI were moderately expressed, but the expression levels of Syts III-VI, and X were at least 5-10 times lower than that of Syt I (Fig. 2A). To overcome this, we initially analyzed 1/10 the volume of immunoprecipitants by 10% SDS-PAGE and normalized them based on the immunore activity of anti-T7 tag antibodies in immunoblotting. Then, equivalent amounts of anti-T7 tag antibody immunoprecip itants were subjected to 10% SDS-PAGE and analyzed by immunoblotting using anti-FLAG rabbit antibody (Fig. 3, upper four panels and summarized in Table I). To confirm the amounts of anti-T7 tag antibody immunoprecipitants, the blots were stripped and reprobed with HRP-conjugated

640

M. Fukuda

anti-T7 tag antibody (Fig. 3, bottom panel). The results are summarized in Table I, and representa tive immunoblots are shown in Fig. 3. Based on the differ ences in oligomerization activity, we classified the synap totagmin family into four distinct groups. Group A Syts

and K Mikoshiba

(III, V, VI, and X) formed stable hetero-ohgomers with each other irrespective of the presence of Ca2+ (10). In addition, except for Syt III, group A synaptotagmins strongly associ ated with Syt VII (group D) in a Ca2+-dependent manner as described previously (Fig. 3, third and fourth panels) (20).

Fig. 2. Expression of T7tagged Syts I-XI (A) and -Syts I-XI-cyto (B). T7tagged Syts were expressed in COS-7 cells as described previ ously (10). Cells were homoge nized in 1% SDS using a 27 gauge syringe. The solubilized proteins were boiled for 3min, subjected to 10% SDS-PAGE, and transferred to a polyvinyl idene difluoride membrane (Millipore). Expressed T7-Syts were detected by HRP-conju gated anti-T7 tag antibody (1/ 1,000 dilution). Since the expo sure time of (B) is less than half of that in (A), comparable amounts of Syt cytoplasmic domains were expressed in COS-7 cells. Notably, Syts IV and VI cytoplasmic domains were more highly expressed than full length proteins. Similar protein expression levels were observed in the case of FLAG-Syts markers (x 10-3) are shown on the left. Lanes 1-11, T7-Syts I-XI (or I-XI-cyto) and lane 12, vector

Fig.

3.

tion

properties

Ca2+-dependent

I-XI

and

pEF-FLAG-Syt

A), or -Syt combinant sence of followed were

and

VII

(Group

proteins 500ƒÊM by

detected

-independent

of synaptotagmins I (Group D) were

by

B), -Syt

co-transfected

were immunoprecipitated Ca2+ by anti-T7 tag

immunoblotting. exposure

hetero-oligomeriza

1, TV, V, and

VII.

IV (group into

C), COS-7

in the antibody-conjugated

Co-immunoprecipitated to X-ray

film

(X-OMATAR,

pEF-T7-Syts -Syt

V (Group

cells, presence

and

re

or ab agarose

FLAG-Syts Kodak)

for

2.5

I-XI (data (pEF-BOS)

not shown).The positions transfected control.

of Mr

min using an Enhanced chemiluminescence kit (Amersham Pharma cia Biotech). Bottom panel indicates the immunoprecipitated T7-Syts I-XI of the third panel from the top visualized by HRP-conjugated anti-T7 tag antibody Total (right most lane) includes 1/40 volume of reaction mixtures used for immunoprecipitation . Arrowheads indi cate the positions of the FLAG-Syt I , IV, V, or VII. Bars shown on the ri ght are Mr markers, corresponding to 97.4, 66.2, and 45.0 kDa from the top, respectively.

J. Biochem.

Ca2+-Dependent and -Independent Oligomerization of Synaptotagmins T7-Syt V was shown to interact well with FLAG-Syt II irre spective of the presence of Ca2+ (Table I) . In contrast, T7-S yt II-FLAG-Syt V interactions were difficult to detect unless a large amount of T7-Syt II immunoprecipitant was loaded on SDS-PAGE. This was probably due to the differ ence in expression levels of the two molecules; the amount of Syt II expressed was at least 10 times greater than that

TABLE

I. Summary

of Ca2+-dependent

Ca2+-dependent independent

(500ƒÊM) experiments

from

extracts

ray

the

cell

films

by

(Version "++++"

eye,

1.0) and

and (see

-independent immunoblots

co-transfected

but

and

in some

gomerization.

++,

weak

(2mM of Fig.

with

cases,

were

like

Syt

presence of Ca2+. +, very weak association, ing conditions, like Syt IV homo-oligomerization.

TABLE

II. Summary

Ca2+-dependent three

transfected in

some

Image)

with cases, as

dependent detected on

were

did

like

Syt

II

prolonged

exposure

-,

undetectable

under

show

Vol. 128, No. 4, 2000

apparent

to X-ray our

binding


-,

was

(Bio

very

conditions (see

weak weak like Table

and

were

quantified

association,

like

Ca2+-dependent association, Syt

when

IV homo-oligomerization.

I). a, some

fractions

mainly by

III,

and

Syt

Basic

V, VI,

"+++" like

and

from

Quantifier

software between

like

-,

undetectable

were

determined

Basic

Syt amount

ND,

of oligomerization

not

IX

from from X-ray

software

at cell films

in the our

least

bind-

because

Ca2+-independent.

two

by

co

eye,

but

1.0)

(Bio Ca2+-

which was the

or

extracts

moderate


are

I homo-oli

(Version ++,

of immunoprecipitant determined

X-

I-XI.

judged

from

Quantifier

homo-oligomerization. like

Syt

under

of synaptotagmins

mainly

a large

determined

X). +++•},

"++",

immunoprecipitants

VII

at least two or three immunoprecipitants

IX homo-oligomerization

film.

T7-Syts-cyto

using

by Syt

association, only

"-",

to X-ray

and

activities

Image)

(Syts

domains

by

were

quantified

between and

exposure

determined

Oligomerization

+,

film.

prolonged

and A Syts

++•}, "--+"

of FLAG-Syts-cyto level

Ca2+-dependent

homo-oligomerization.

after

not

control.

Image)

VIII).

judged from by using the

activities

group

between

I -XI.

T7-Syts were was determined

the cytoplasmic

association

2000i/VGA

strong

after

between

background

Print

(Bio like

II and +-,

only

and level

2000i/VGA

(Syts

synaptotagmins

Oligomerization

association,

B Syts

detected

EGTA)

a vector Gel

+++,

control.

Print

strong

group

was

(2mM

by (20).

Gel

between

of FLAG-Syts background

a vector

homo-oligomerization.

FLAG-Syts-cyto and

captured

previously

association,

SDS-PAGE.

proteins

The

by

oligomerization

independent

pEF-FLAG-Syts-cyto bands

described

only

and

experiments.

like XI

which

of Ca2+-dependent

(500ƒÊM)

independent

association FLAG-Syts and

(20). ++++,

association

association,

EGTA) 3). The

captured

previously

of Syt V (Fig. 2A), and accordingly T7-Syt II homo-oligo mers were predominantly immunoprecipitated rather than T7-Syt II-FLAG-Syt V hetero-oligomers. Group B was com posed of Syts I, II, VIII, and XI, which form homo-oligomers irrespective of the presence of Ca2+. Syts I and II associated with each other irrespective of the presence of Ca2+ (Fig. 3, upper panel), consistent with the previous report that

oligomerization

pEF-FLAG-Syts

bands

(Bio Image) as described "+++" +++ , moderate

-independent

641

was loaded

full-length

642

M. Fukuda

Fig. 4. Summary -independent

and K Mikoshiba

of Ca2+-dependent oligomerization

and proper.

ties of synaptotagmins in COS-7 cells, Synaptotagmin isoforms are classified into four distinct groups according to their abil ity to self-oligomerize (Group A > D ? B > C; see text and Tables I and II). Isofonns that show strong or weak Ca2+-dependent homo-oligomerization mediated by cytoplasmic domains are indicated by red or green letters, respectively. Syt VII is further shaded because it shows robust Ca2+-depen dent oligomerization even when the fulllength proteins are used. Isoforms that do not essentially show hetero-oligomerization are indicated by blue letters (Syts VIII and XI). An isoform that does not essentially show homo-oligomerization is indicated by black letters (Syt IV). Lines indicate the het ero-oligomerization of two molecules and their thickness indicates the relative strength of the oligomerization, although the weak hetero-oligomerizations (e.g. "++" or "+" in Table I) are omitted in this figure. Hetero-oligomerization that was activated by Ca2+ is colored in red. Broken lines indi cate that the interaction of two isoforms was detected in only one way (e.g., T7-Syt VFLAG-Syt II interaction).

native

Syts

I and

other

(26).

these

associations

Compared ties

oligomerize, mer,

we

used

ing

activity

showed

of

weak

M Ca2+, acted

Syt

IX

not

Syts

Ca2+. VII

final

plasmic domains combinant Syt

were

Syt

the full of T7-(or

summarized Syts

of the

presence

equally

well

and

X in

mic

domain

ious

Syts

I protein

COS-7

evident

at

weakly

cytoplasmic

500ƒÊ

a

from

cells

II.

Syt

II

but

not

V were

with

Syt

VII

a Ca2+-dependent showed (I, II, V, VI,

the

weakly

manner. broadest

IX, and

The

com-

greater domains

results

are

V, though

the

full-

The

Syts Syt

different

VII

and

III, VI,

IX,

cytoplas bound

affinities.

var

synaptotagmin

the

show

four

distinct

properties. associated with motif

by

the

with differing

previous

disulfide

XI were

I and

classed

as at

I itself

whereas

Syt (e.g.,

because

Syts

ally

equivalent

based

fact

that

tion ties

in brain (37), and of the C2 domains

I and

Syts

I only Syts on

the

were

the

20).

cytoplasmic very

Al-

N-terminus

weak

domain Ca-de-

conditions. In addition, we with a number of different

associated

with

, VII, and IX). I and II are thought high

B and (10,

at

II

II show

anal

1,3,4,5-tet

as group

showed

in our binding II can associate

in This

phylogenetic

to ƒÀ-mercaptoethanol as well

other

homo-oligomerization,

II hetero-oligomerize Syt

of isoforms

each

Ca2+these

affinities.

of inositol

Ca2+-independent

insensitive

Syts

with

with

and

by

Since

cysteine

weakness or absence binding ability (19).

I, II, VIII,

pendent activity found that Syt

17, 38).

into

hetero-oligomerize

is consistent

Syts

family

manner

(Ca2+ -dependent),

ber

isoforms

properties by using two dif FLAG). As illustrated in Fig. 4,

N-terminal

(Ca2+-independent)

surprising

irrespective

to hetero-oligomerize with

the

can

was

though

isoforms,

hetero-oligomerize

Syt

specificity

X) with

As or

that

(Nlum/Ccyto) at the N-

and accordingly their associations were At the cytoplasmic domain, however,

characterized which

synaptotagmin

their homo-oligomerization V, VI, and X) strongly

Ca2+-dependent

Syts cyto

(22-26).

associated

V seemed and

the

interaction of rewith detergent

2B).

could

strongly

Syt

via

ysis (10) and rakisphosphate

Cal-dependent

brain

(Fig.

of

I-XI-In the

by the bacteria

in

IX,

of Ca2+.

other

that

I membrane topology N-glycosylation site

synaptotagmin

on (III,

synaptotagmins

interdomain

demonstrated

and

the

based A Syts

bonding (10), independent.

of synaptotagmin

and

II and

classified

groups Group each

association

have

show type an artificial

domain,

its

Properties

examined

Table

I, VII,

length

IX also

the

we IV

or 250mM

length protein, comparable FLAG)-Syts I-XI cytoplasmic

in in

Syts

Syt at

Syt

we

complex hetero-oligomerization ferent epitopetags (T7 and

self-oligomeriz

Synaptotagmins

we

as demonstrated I or II expressed

expressed

with

was

properties

native to

which

study,

grouping

from


pared amounts

weak,

iso

manner.

of experiments,

solubilized

X-100

Hetero-Oligomerization

Domains

set

that

the

quite

250ƒÊM

Ca2+-Dependent Cytoplasmic

was

other

this

terminal

self-

as a mono

possibility

1% Triton

I, II, V, and

in a Ca2+-dependent

with

In

family proteins by introducing

expression

present

Although

Ca2+-dependency,

but

with

to

the

more affini

XI could

interact

out

solubilization.

in

DISCUSSION

Ca2+.

not detected

and

mainly

and

by

various

were

VIII

each

VII,

interacted

with

differences

essentially

is sensitive for

II

Syts

rule

Syt

activated

interactions

IV was

cannot

oligomerization

slightly

the

with

with

isoforms to

not

C Syt

associate

FLAG-Syt

In contrast,

did

group

though

NaCl

I,

due

2A). but

The

to be

these

probably Fig.

brain

associated

T7-tagged

I). However,

(see

forms.

seemed

other

reverse,

level

from

II also

FLAG-Syt

with

(Table

in

I and

to

broadly

II proteins

Syts

sequence

almost

a limited

num

This finding is to be function-

similarity

complimentary

(10),

the

distribu

the fact that the biochemical proper tested so far are quite similar (15-

I and

II are

thought

to

be Ca2+-sensors

2 Biochem,

Ca2+-Dependent and -Independent Oligomerization of S

ynaptotagmins

643

for neurotransmitter release (1-3) , our results raise the modulate the function of other neuronal (I, II, V, X) and possibility that their hetero-oligomerization with other iso non-neuronal (VI, IX) isoforms without tethering at the Nforms having distinct biochemical activities may the variety of Ca2+-sensors produce terminal domain. While we were preparing this manu . In contrast, Syts VIII and XI script, Martinez et al. reported that Syt VII is involved in are thought to be present in the homo-oligomer or mono Ca2+-dependent exocytosis of lysosomes in fibroblasts (43). mer state. Therefore, it will be interesting to see whether Syt I-Syt Is Ca2+-dependent oligomerization of Syt I C2B domain VII hetero-oligomers are involved in neurotransmitter indeed essential for synaptic vesicle exocytosis in vivo? Al release in the future. though the importance of the Syt I C2B domain in synaptic Among the synaptotagmin isoforms, Syts I and II are vesicle exocytosis and endocytosis has been well demon known to be localized at synaptic vesicles as well as secre strated by antibody-loading experiments (5-7) and analysis tory vesicles of some endocrine cells (44). Recently, several of Drosophila syt mutants (21), there is no direct evidence synaptotagmin isoforms other than Syts I and II in brain that Ca2+-dependent oligomerization of the Syt I C2B do have been reported to have different subcellular localiza main itself is essential for synaptic vesicle exocytosis in tion: Syt III is mainly localized at the synaptic plasma vivo. Recently, however, we found that a Syt II mutant car membrane (45); Syt VI‡™TM, which lacks the transmem rying a Tyr-312 to Asn substitution in the C2B domain domain, is associated with various membrane frac , which c brane orresponds to the Drosophila AD3 mutation , had tions (45,46); and Syt IV is localized at the Golgi and distal completely impaired self-oligomerization activity but could part of neurites in nerve growth factor-differentiated PC12 weakly hetero-oligomerize with wild-type proteins in a cells and cultured hippocampal neurons (39, 40). These Ca2+-dependent manner (Fukuda , M., unpublished data). observations are consistent with our results that Syt I did This result is consistent with the fact that this syt allele not essentially interact with Syts III, IV, and VI even in the could complement another mutant phenotype , although it i presence of Ca2+. At this stage, we cannot completely rule s homozygous lethal (21) . Under our binding conditions , h out the possibility that certain adaptor proteins [e .g, syn owever, the Syt I cytoplasmic domain showed only very taxin IA, which binds Syt I in a Ca2+-dependent manner weak Ca2+-dependent self-oligomerization activity How (15)] bridges different Syt isoforms. However, this possibil ever, given that Syt I proteins are enriched on synaptic ves ity is unlikely, because we could only detect Syts and IgG icles at the preterminal (i.e., local concentrations of this used for immunoprecipitation, when immunoprecipitants protein are relatively high), it is possible ttt Ca2+-depen were visualized by Coomassie Brilliant Blue R-250 staining dddddddnt oligomerization of the two C2B domains that are pre (data not shown). Furthermore, the T7and FLAG-Syts co assembled at the N-terminal domain may occur in intact - transfection assay needs to be performed in other types of preterminals (20). cells, such as PC12 cells, in the future, because of the possi Group C synaptotagmins were characterized by weak bility that subcelluar localization of Syt isoforms may differ (Syt IX) or no homo-oligomerization activity (Syt IV) under between cell types. our binding conditions. Because of the weak oligomeriza In summary, we investigated the Ca2+-dependent and tion activity, we could not detect their SDS-insensitive -independent homoand hetero-oligomerization of synap homo-oligomerization on SDS-PAGE (10). Although the totagmin family proteins and showed that some isoforms Ca2+-independent homo-oligomerization activity of Syt IX can oligomerize via the N-terminal domain (Ca2+-indepen was very weak, Syt IX weakly associated with Syts I, II, V, dent) and/or the cytoplasmic domain (Ca2+-dependent). Our and VII in a Ca2+-dependent manner. In contrast, Syt IV is results and heterologous expression of synaptotagmin fam ily in brain (37, 47, 48) suggest that hetero-oligomerization probably present mainly as a monomer irrespective of the presence of Ca2+. When Syt IV is expressed in COS-7 cells, of a possible neuronal Ca2+-sensor, Syt I (or II), with other it is predominantly localized at the Golgi-like perinuclear isoforms that have different biochemical properties, may compartment, whereas other Syt isoforms are localized at create a variety of Ca2+-sensors. tubular or vesicular structures and/or plasma membranes (39,40 and Fukuda, M., unpublished data). Thus, this dif We thank Dr. Shigekazu Nagata for the expression vector, Dr. ferent compartmentalization of Syt IV to other isoforms in Masao Sakaguchi for helpful advice, Eiko Kanno for technical as sistance, Hiroyuki Kabayama, Keiji Ibata, and Fumiaki Hamazato intact cells may restrict the hetero-oligomerization of Syt for valuable discussions, and Akihiro Mizutani for critical com IV with other informs (Fig. 3, 2nd panel). In contrast to our ments on the manuscript. findings, however, Chapman et a). reported that Syts I and IV hetero-oligomerize in a Ca2+-dependent manner by REFERENCES using recombinant proteins from bacteria (25). This dis crepancy may be due to differences in binding conditions, 1. Sudhof, T.C. and Rizo, J. (1996) Synaptotagmin: C2-domain pro such as different recombinant proteins, amount of proteins teins that regulate membrane traffic. Neuron 17, 379-388 or incubation time. Alternatively, in vitro oligomenzation of 2. Schiavo, G., Osborne, S.L., and Sgouros, J.G. (1998) Synap recombinant synaptotagmin cytoplasmic domains from bac totagmins: More isoforms than functions? Biochen. Biophys. teria may be non-selective as in the case of the SNARE Res. Commun. 248,1-8 3. Fukuda, M. and Mikoshiba, K. (1997) The function of inositol [SNAP (soluble NSF attachment protein) receptors] com high polyphosphate binding proteins, BioEssays 19,593-603 plex (41, 42). 4. Mikoshiba, K., Fukuda, M., Moreira, J.E., Lewis, F.M.T., Sugi Syt VII, the only member of group D, showed unique mori, M., Niinobe, M., and Llinas, R. (1995) Role of the C2A Ca2+-dependent oligomerization properties. The Syt VII domain of synaptotagrnin in transmitter release as determined cytoplasmic domain oligomerizes in a Ca2+-dependent man by specific antibody injection into the squid giant synapse pre ner without being tethered at the N-terminal domain (20). terminal. Proc. Natl. Acad. Sci. USA 92,10703-10707 5. Fukuda, M., Moreira, J.E., Lewis, F.M.T., Sugimori, M., NiiSince Syt VII is ubiquitously expressed (15), Syt VII can Vol. 128, No. 4, 2000

644

M. Fukuda

nobe,

M.,

domain

Mikoshiba,

K.,

and

of synaptotagmin

determined synapse

Llinds,

in

by specific preterminal.

R.

(1995)

vesicular

antibody Proc. Natl.

Role

release

injection Acad.

of the

and

C2B

recycling

into the Sci. USA

squid 92,

as

giant 10708-

10712 6.

Mochida,

S.,

Mikoshiba,

Fukuda,

M.,

(1997)

Roles

K.

neurotransmitter binding 937-943 7.

at

mammalian

Ohara-Imaizumi, Ikeda,

K. (1997)

exocytosis

C2

aptotagmin

at

low

M.,

M.,

Misonou,

and

C2B

calcium.

and

domains

and

of chro

Wollheim,

synaptotagmin

is

pancreatic ƒÀ-cells:

micromolar

H.,

K.,

in adrenal

Sci. USA 94,287-291 H., Mikoshiba, K.,

from

in

77,

Mikoshiba,

of C2A

of

and

Neuroscience

of exocytosis

domain

of insulin

domains

Niinobe,

roles

regulation

first

H.,

C2

high-polyphosphate

synapses. M.,

Natl. Acad. M., Zhang,

The

Kobayashi,

inositol

Kawasaki,

Distinct

in the

(1997)

for

T.,

cells. Proc J., Fukuda,

C.B.

M.,

synaptotagmin

and

Fukuda,

Murakami,

synaptotagmin maffin Lang,

of

cholinergic M.,

K.,

Kumakura,

8.

Niinobe,

secretion

required

Action

EMBO

J.

of 16,

syn

5837-

5846 9.

Thomas, the C2A

D.M. and Elferink, L.A. domain of synaptotagmin

regulated 10.

secretion.

Fukuda,

M.,

terminal

cysteine

formation

Kanno,

of

(1998) Functional 1: Implications

J. Neurosci. E.,

and

18,

motif

Mikoshiba,

K.

is essential

synaptotagmins

for

III,

analysis for calcium

of

3511-3520 (1999)

homo-

V, VI,

Conserved

and

and

N-

heterodimer

X. J

Biol.

Chem.

274,31421-31427 11.

Littleton,

J.T.,

Secant,

T.L.,

Chapman,

E.R.

(1999)

Synaptic

in 12.

the

Lloyd,

ratio

of synaptotagmin

T.E.,

Verstreken,

O., and

Bellen,

vesicle 13.

cycle

Fukuda,

of

and

von

15.

528-532

C.,

Ichtchenko,

The

evolutionary

synaptotagmins

with

Ca2+ binding. Li, C., Ullrich, Sudhof,

16.

270,

Poser,

(1997)

Mikoshiba,

and

Fukuda,

M.,

(1995)

(2000)

K.,

26,

Shao,

Genomic

Rizo,

to

structures

inactivate:

C2A

and

J.,

and

domain

T.C.

subclass

N.,

375,

594-599

Phospholipid

phospholipid IV. J

and

activities

Nature

of synaptotagmin

of

abolishes

Brose,

K. (1996)

Ca2+ -dependent

Res.

Sudhof, A that

-independent

Mikoshiba,

of

of

organiza Biophys.

substitution

synaptotagmins. T., and

synaptic

45-50

Biochem.

X.,

acid

Ca2+ -dependent

non-neural

Lichtarge,

for

of exon-intron

pressure

and

changes

757-760 A.,

search

family.

amino

dependence

to the

400,

Phillippi,

Comparison

an

Kojima,

composition ing

K.

by

J. Biol. Chem. 272, 14314-14319 B., Zhang, J.Z., Anderson, R.G.W,

T.C.

of neural

IV. Nature

Neuron

gene

B.,

modulated

E.J.,

in Drosophila.

synaptotagmin

Common. 14.

Ganetzky,

A genome-wide

II protein:

the

function I and

(2000)

proteins

M.

tion

G.M.,

P., Ostrin,

H.J.

synaptotagmin

Rubin,

Biol.

bind

Chem.

271,

8430-8434 17.

Fukuda, K.

18.

M.,

(1995)

Kojima,

Mutational

binding

domain.

Mehrotra,

J

B.,

Chaudhary, (1997)

Ibata,

K.,

of

Chem.

J.T.,

C2B

and

acid

20.

Fukuda, tion

that

22.

and

activities

J.D., and

Mikoshiba,

K.

of synaptotagmin

and

neuronal

X. J

Distinct

family:

properties

Chem. Inositol and amino

1,3,4,5-tetrakisphos

V,

(2000)

Biol.

(1998) of

G.D.

of conserved

inositol III,

A.A.,

polyphos

J.

activities

abolish

Profit,

Prestwich, inositol

K.

Identification

synaptotagmins

oligomerization

Chem.

Biol.

Chem.

self-oligomeriza

Unique

calcium-depen

of synaptotagmin

VII.

J.

Riot.

275,28180-28185

Littleton, and

to

M.

dent

21.

binding

substitutions

K.,

of the

Mikoshiba,

synaptotagmins:

phate binding 273,12267-12273

Olszewski,

of synaptotagmins.

M.,

Mikoshiba,

polyphosphate

26523-26527

J.,

labeling

1,3,4,5-tetrakisphosphate nonneuronal

and

of synaptotagmin high

Mikoshiba,

domains

M.,

domains

inositol

270,

Chen, M.,

photoaflinity

Fukuda,

Niinobe,

of C2

Fukuda,

Selective

J.,

diversity analysis Biol.

Elliott,

A.,

phate binding 272,4237-4244 19.

T., Arug,

Functional

family:

J.T.

and

modulates

manner.

Trends

Sugita,

S.,

dependent

Bellen,

H.J.

(1995)

synaptic-vesicle Neurosci.

Hata, properties

Y,

Synaptotagmin

fusion

18,

in

a

controls

Ca2+-dependent

177-183

and

Sudhof,

of the

first

TC. and

(1996)

second

Distinct

C2-domains

Ca2+-

J. Boil. Che of syn-

and K. Mikoshiba

aptotagmin I. J. Biol. Chem. 271,1262-1265 23. Chapman, E.R., An, S., Edwardson, J.M., and Jahn, R. (1996) A novel function for the second C2 domain of synaptotagmn: Ca2+-triggered dimerization. J. Biol. Chem. 271, 5844-5849 24. Damer, C.K. and Creutz, C.E. (1996) Calcium-dependent self. association of synaptotagmin I. J. Neurochem. 67, 1661-1668 25. Chapman, E.R., Desai, R.C., Davis, A.F., and Tornehl, C.K. (1998) Delineation of the oligomerization, AP-2 binding, and synprint binding region of the C2B domain of synaptotagmin. J Biol. Chem. 273, 32966-32972 26. Osborne, S.L., Herreros, J., Bastiaens, P.I.H., and Schiavo, G. (1999) Calcium-dependent oligomerization of synaptotagmins I and II: Synaptotagmins I and II are localized on the same syn aptic vesicle and heterodimerize in the presence of calcium. J. Biol. Chem. 274,59-66 27. Perin, M.S., Brose, N., Jahn, R., and Sudhot T.C. (1991) Do main structure of synaptotagmin. J. Biol. Chem. 266, 623-629 28. Niinobe, M., Yamaguchi, Y., Fukuda, M., and Mikoshiba, K. (1994) Synaptotagmin is an inositol polyphosphate binding pro tein: Isolation and characterization as an Ins 1,3,4,5-P4 binding protein. Biochem. Biophys Res. Commun. 205, 1036-1042 29. Fukuda, M., Aruga, J., Niinobe, M., Aimoto, S., and Mikoshiba, K. (1994) Inositol-1,3,4,5-tetrakisphosphate binding to C2B do main of IP4BP/synaptotagmin II. J. Biol. Chem. 269, 2920629211 30. Mizushima, S. and Nagata, S. (1990) pEF-BOS, a powerful mammalian expression vector. Nucleic Acids Res. 18, 5332 31. Kida, Y., Sakaguchi, M., Fukuda, M., Mikoshiba, K., and Mihara, K. (2000) Membrane topogenesis of a type I signal anchor protein, mouse synaptotagmin II, on the endoplasmic reticulum. J. Cell Biol. 150, 719-730 32. Krazewski, K., Mundigl, O., Daniell, L., Verderio, C., Matteoli, M., and De Camilli, P. (1995) Synaptic vesicle dynamics in liv ing cultured hippocampal neurons visualized with CY3-conju gated antibodies directed against the lumenal domain of synap totagnun. J. Neurosci. 15, 4328-4342 33. Malgaroli, A., Ting, A.E., Woodland, B., Bergamaschi, A., Villa, A., Tsien, R.W., and Scheller, R.H. (1995) Presynaptic compo nent of long-term potentiation visualized at individual hippo campal synapses. Science 268, 1624-1628 34. Kraszewski, K., Daniell, L., Mundigl, O., and De Camilli, P. (1996) Mobility of synaptic vesicles in nerve endings monitored by recovery from photobleaching of synaptic vesicle-associated fluorescence. J. Neurosci. 16, 5905-5913 35. Veit, M., Sollner, T.H., and Rothman, J.E. (1996) Multiple pal mitoylation of synaptotagmin and the t-SNARE SNAP-25. FEBS Lett. 385,119-123 36. Chapman, E.R., Blasi, J., An, S., Brose, N., Johnston, P.A., Sud hof, T.C., and Jahn, R. (1996) Fatty acylation of synaptotagmin in PC12 cells and synaptosomes. Biochem. Biophys. Res. Com. mun. 225,326-332 37. Ullrich, B., Li, C., Zhang, J.Z., McMahon, H., Anderson, R.G.W, Geppert, M., and Sudhof, T.C. (1994) Functional properties of multiple synaptotagmins in brain. Neuron 13,1281-1291 38. Fukuda, M., Kojima, T., and Mikoshiba, K. (1997) Regulation by bivalent cations of phospholipid binding to the C2A domain of synaptotagmin III. Biochem. J. 323, 421-425 39. Ibata, K., Fukuda, M, Hamada, T., Kabayama, H., and Miko shiba, K. (1999) Synaptotagmin IV is present at the Golgi and distal parts of neurites . J. Neurochem. 74, 518-526 40. Berton, F., Cornet, V., Iborra, C., Garrido, J., Dargent, B., Fukuda, M., Seagar , M., and Marqueze, B. (2000) Synaptotagmin I and IV define distinct populations of neuronal transport vesicles. Eur. J. Neurosci. 12,1294-1302 41. Yang, B., Gonzalez , L.Jr., Prekeris, R., Steegmaier, M., Advani, R.J., and Scheller , R.H. (1999) SNARE interactions are not selective: Implications for membrane fusion specificity. J Biol. Chem. 274, 5649-5653 42. Fasshauer, D., Antonin, W., Margittai, M., Pabst, S., and Jahn, R. (1999) Mixed and non-cognate SNARE complexes: Charac terization of assembly and biophysical properties . m. 274, 15440-15446

J. Biochem.

Ca2+-Dependent

and -Independent

Oligomerization

of Synaptotagmins

43. Martinez, I., Chakrabarti, S., Hellevik, T, Morehead, J., Fowler, K., and Andrews, N.W. (2000) Synaptotagmin VII regulates Ca2+-dependent exocytosis of lysosomes in fibroblasts. J. Cell Biol. 148,1141-1149 44. Matthew, W.D., Tsavaler, L., and Reichardt, L.F. (1981) Identifi cation of a synaptic vesicle-specific membrane protein with a wide distribution in neuronal and neurosecretory tissue J. Cell Biol. 91, 257-269 45. Butz, S., Fernandes-Chacon, R., Schmitz, F., Jahn, R., and Sud hof, T.C. (1999) The subcellular localizations of atypical synap totagmins III and apses and synaptic

Vol. 128, No. 4, 2000

VI: Synaptotagmin plasma membranes

III is enriched in syn but not in synaptic ves-

645

icles. J. Biol. Chem. 274,18290-18296 46. Fukuda, M. and Mikoshiba, K. (1999) A novel alternatively spliced variant of synaptotagmin VI lacking a transmembrane domain: Implications for distinct functions of the two isoforms. J. Biol. Chem. 274,31428-31434 47. Berton, F., Iborra, C., Boudier, J.-A., Seagar, M.J., and Mar queze, B. (1997) Developmental regulation of synaptotagmin I, II, III, and IV mRNAs in the rat CNS. J. Neurosci. 17,12061216 48. Ullrich, B. and Sudhof TC. (1995) Differential distributions of novel synaptotagmins: Comparison to synapsins. Neurophar macolggy 34,1371-1377