Annexins
245
different acidic lipids. In addition, the four
homologous repeats may have different
speciFcities, since the repeats are only 40
to 50% identical in sequence.
The crystal structure of annexin V has
also been interpreted in terms of the ion
channel–forming properties of the pro-
tein. The molecule has a hydrophilic pore
perpendicular to the face of the membrane
(±ig. 2). Because the molecule has no hy-
drophobic external surfaces, it is assumed
that it cannot actually enter the bilayer in
this conformation. However, it has a cal-
culated external electrical Feld similar to
the Feld strength necessary to punch a
hole through a membrane by electropo-
ration. Therefore, it has been suggested
tha
tthemo
le
cu
les
i
t
sonthesu
r
fa
ceo
f
the membrane and leads to a disordered
state in the lipids immediately below it.
In this disordered state, the membrane
conducts ions that must also pass through
the hydrophilic pore in the molecule. The
molecule thus provides the ion selectiv-
ity of the overall transmembrane channel.
A similar mechanism of membrane per-
meabilization has been suggested for the
action of cholera toxin.
An alternative mechanism for the ability
of annexins to promote ion permeation has
been proposed that involves a signiFcant
reorganization of the molecule at low pH to
form a transmembrane channel. However,
a crystallographic model of this proposed
structure has not yet been determined. It
is also not known if the channel activity
that annexins exhibit
in vitro
corresponds
to a speciFc physiological channel in cells.
Since the calcium/lipid binding sites
all appear to be on one side of the an-
nexin molecule, it is not obvious how
the annexins aggregate membranes. It
seems likely that a self-association of
annexin molecules attached to different
membranes is required for membrane
aggregation (±ig. 3). The concave faces
of the molecules that face the cytoplasm
might interlock during this self-association
event. In the case of synexin acting
in vitro
on chromafFn granules (secretory vesicles
of the adrenal medulla), binding to mem-
branes occurs at low levels of calcium (less
than 10
µ
mol), but membrane aggrega-
tion and fusion depend on higher levels of
calcium (greater than 100
µ
mol). The cal-
cium dependence of granule aggregation
correlates exactly with the calcium depen-
dence of synexin self-association in the ab-
sence of membranes, suggesting a mecha-
nism for membrane aggregation whereby
membrane-bound synexin molecules un-
d
e
r
g
os
e
l
f
-
a
s
s
o
c
i
a
t
i
o
nt
ob
r
i
n
gt
h
etw
o
membranes together. If this mechanism
is correct, then the ‘‘bottleneck’’ in the sys-
tem is the high level of calcium needed
to promote annexin self-association. The
annexin II (calpactin) tetramer promotes
the aggregation and fatty acid–dependent
fusion of chromafFn granules at the lowest
level of calcium of any annexin (approxi-
mately 1
µ
mol). This may occur because
the tetramer represents a permanently
self-associated annexin (±ig. 1). The two
heavy chains might be associated through
the light-chain dimer in such a way that
each heavy chain can bind a different
membrane and thus pull them together.
Therefore, the processes of overall mem-
brane aggregation and fusion are catalyzed
by the low levels of calcium needed simply
to promote membrane binding.
3
Regulation
The N-terminal domains of the annexins
are the major sites of sequence divergence
in the family (±ig. 1). They may be the
site of interaction with other proteins, as
previous page 245 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online next page 247 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online Home Toggle text on/off