CRACM1, CRACM2, and CRACM3 are store-operated Ca2+ channels with distinct functional properties

Abstract

STIM1 in the endoplasmic reticulum and CRACM1 in the plasma membrane are essential molecular components for controlling the store-operated CRAC current. CRACM1 proteins multimerize and bind STIM1, and the combined overexpression of STIM1 and CRACM1 reconstitutes amplified CRAC currents. Mutations in CRACM1 determine the selectivity of CRAC currents, demonstrating that CRACM1 forms the CRAC channel's ion-selective pore, but the CRACM1 homologs CRACM2 and CRACM3 are less well characterized. Here, we show that both CRACM2 and CRACM3, when overexpressed in HEK293 cells stably expressing STIM1, potentiate I(CRAC) to current amplitudes 15-20 times larger than native I(CRAC). A nonconducting mutation of CRACM1 (E106Q) acts as a dominant negative for all three CRACM homologs, suggesting that they can form heteromultimeric channel complexes. All three CRACM homologs exhibit distinct properties in terms of selectivity for Ca(2+) and Na(+), differential pharmacological effects in response to 2-APB, and strikingly different feedback regulation by intracellular Ca(2+). Each of the CRAC channel proteins' specific functional features and the potential heteromerization provide for flexibility in shaping Ca(2+) signals, and their characteristic biophysical and pharmacological properties will aid in identifying CRAC-channel species in native cells that express them.

Full text

Current Biology 17, 794–800, May 1, 2007 ª2007 Elsevier Ltd All rights reserved DOI 10.1016/j.cub.2007.03.065
Report
CRACM1, CRACM2, and CRACM3
Are Store-Operated Ca
2+
Channels
with Distinct Functional Properties
Annette Lis,
1,2,3
Christine Peinelt,
1,2,3
Andreas Beck,
1,2
Suhel Parvez,
1,2
Mahealani Monteilh-Zoller,
1,2
Andrea Fleig,
1,2
and Reinhold Penner
1,2,
*
1
Center for Biomedical Research
The Queen’s Medical Center
Honolulu, Hawaii 96813
2
John A. Burns School of Medicine
University of Hawaii
Honolulu, Hawaii 96813
Summary
STIM1 in the endoplasmic reticulum and CRACM1 in
the plasma membrane are essential molecular compo-
nents for controlling the store-operated CRAC current
[1–4]. CRACM1 proteins multimerize and bind STIM1
[5, 6], and the combined overexpression of STIM1
and CRACM1 reconstitutes amplified CRAC currents
[7–10]. Mutations in CRACM1 determine the selectivity
of CRAC currents, demonstrating that CRACM1 forms
the CRAC channel’s ion-selective pore [11, 5, 6], but
the CRACM1 homologs CRACM2 and CRACM3 are
less well characterized [7, 12]. Here, we show that
both CRACM2 and CRACM3, when overexpressed in
HEK293 cells stably expressing STIM1, potentiate
I
CRAC
to current amplitudes 15–20 times larger than
native I
CRAC
. A nonconducting mutation of CRACM1
(E106Q) acts as a dominant negative for all three
CRACM homologs, suggesting that they can form het-
eromultimeric channel complexes. All three CRACM
homologs exhibit distinct properties in terms of selec-
tivity for Ca
2+
and Na
+
, differential pharmacological
effects in response to 2-APB, and strikingly different
feedback regulation by intracellular Ca
2+
. Each of the
CRAC channel proteins’ specific functional features
and the potential heteromerization provide for flexibil-
ity in shaping Ca
2+
signals, and their characteristic
biophysical and pharmacological properties will aid
in identifying CRAC-channel species in native cells
that express them.
Results and Discussion
In many cell types, store-operated Ca
2+
entry represents
the primary mechanism underlying long-lasting eleva-
tions in intracellular Ca
2+
, elevations that follow IP
3
-
mediated release of Ca
2+
from intracellular stores [13–
16]. Previous investigations have identified CRACM1
(or Orai1) as the calcium-release-activated calcium
(CRAC) channels in the plasma membrane [3, 4]. There
are three mammalian homologous CRAC channel
proteins, CRACM1, CRACM2, and CRACM3, and all ho-
mologs are widely expressed at the mRNA level [12].
CRACM Homologs Represent
Store-Operated Channels
To assess the functional properties of CRACM proteins,
we overexpressed all three CRACM species in HEK293
cells that stably overexpress STIM1 and measured
CRAC currents in response to store depletion. Upon
store depletion with IP
3
, all CRACM homologs produced
large membrane currents (Figure 1A) with inwardly recti-
fying current-voltage (I/V) relationships characteristic of
I
CRAC
(Figure 1B). Although our data substantiate that
CRACM2 represents a store-operated channel, they
are at variance with the reported inability of CRACM3
to increase store-operated Ca
2+
entry or CRAC currents
[7]. We should note, however, that our own attempts
with the commercial CRACM3 vector used by Mercer
et al. also failed to produce enhanced CRAC currents
when expressed in STIM1-expressing cells (data not
shown). However, after subcloning the CRACM3 se-
quence into another vector (see the Experimental Proce-
dures in the Supplemental Data online), we consistently
observed significant CRACM3 currents.
The average current amplitudes of CRACM2 and
CRACM3 at 280 mV were approximately 3-fold smaller
than the corresponding amplitude of CRACM1, but
they are still 15- to 20-fold larger than native CRAC cur-
rents in wild-type or STIM1-expressing HEK293 cells.
The differences in amplitudes may be due to different
expression levels but could also reflect differences in
single-channel conductance or open probability. The
activation kinetics of the CRACM homologs were dis-
tinctly different, with half-maximal activation times
(6SEM) of CRACM1 at 35 67 s (n = 12), CRACM2 at
21 63 s (n = 8), and CRACM3 at 63 67 s (n = 9). These
were unrelated to current magnitude because they were
preserved when analyzing currents with matched ampli-
tudes (Figure 2) or at various [Ca
2+
]
i
(see Figure 3). We
also examined whether CRACM currents were activated
when preventing store refilling with 20 mM BAPTA in the
pipette. Indeed, all three CRACM species produced
CRAC-like currents with a characteristic delay that pre-
sumably reflects the time needed to deplete stores
through leak pathways (Figure S1A). Under these condi-
tions, CRAC currents developed with a similar time
course, thus indicating that store depletion is likely to
be the rate-limiting step for CRAC activation. The I/V re-
lationships (Figure S1B) confirm that these currents also
have the typical shape of CRAC currents. These results
demonstrate that all three CRACM homologs can gener-
ate amplified store-operated CRAC currents and that
they possess characteristic kinetics of activation, thus
possibly indicating differences in the binding or inter-
action with STIM1.
CRACM Homologs Form Heteromeric Channels
Given that all three homologs produced store-operated
channels and CRACM1 has been shown to form
multimeric channel complexes, we used a nonconduct-
ing CRACM1 (E106Q) pore mutation that confers a
*Correspondence: rpenner@hawaii.edu
3
These authors contributed equally to this work.

dominant-negative phenotype on native CRAC channels
[5] to assess whether CRACM1 can assemble into
heteromeric channel complexes with CRACM2 and
CRACM3. Figures 1A and 1C illustrate that the co-over-
expression of CRACM1-E106Q in equal amounts with
the three wild-type homologs essentially abolished
Figure 1. All CRACM Homologs Produce
Store-Operated Currents
(A) Average CRAC current densitiesat 280 mV
induced by IP
3
(20 mM) in stable STIM1-
expressing HEK293 cells transiently overex-
pressing CRACM1 (black, n = 12), CRACM2
(blue, n = 7), and CRACM3 (red, n = 9). Open
symbols represent cells that were cotrans-
fected with the WT constructs of the three ho-
mologs plus the dominant negative E106Q
mutant of CRACM1 (CRACM1-E106Q +
CRACM1, n = 6; + CRACM2, n = 6; +
CRACM3, n = 7). [Ca
2+
]
i
was clamped to
near zero with 20 mM BAPTA. Error bars indi-
cate SEM.
(B) Average current-voltage (I/V) relation-
ships of CRAC currents extracted from repre-
sentative HEK293 cells shown in (A) obtained
at 120 s. Data represent leak-subtracted
current densities (pA/pF) evoked by 50 ms
voltage ramps from 2150 to +150 mV cor-
responding to CRACM1 (black, n = 11),
CRACM2 (blue, n = 6), and CRACM3 (red,
n = 9).
(C) Average CRAC current densities at 280 mV in cells expressing STIM1 alone (n = 13) or additional ly with CRACM1-E106Q + CRACM1/CRACM
2/CRACM 3; data points correspond to currents analyzed from (A) at 120 s. Error bars indicate SEM.
(D) Coimmunoprecipitation of CRACM1 with CRACM2 and CRACM3. Wild-type HEK293 cells were cotransfected with CRACM1-Myc in com-
bination with HA-CRACM1, HA-CRACM2, or HA-CRACM3. Lanes 1 and 2 show nontransfected HEK293 cells. Lanes 3 and 4 show that
CRACM1-Myc can co-IP HA-CRACM3, HA-CRACM2 (lanes 5 and 6), and HA-CRACM1 (lanes 7 and 8). The resulting immune complexes
were immunoblotted with HA antibody, thus revealing bands with molecular weights of w33, w28, and w31 kDa for CRACM1, CRACM2, and
CRACM3, respectively.
Figure 2. CRACM Homologs Have Distinct Fast and Slow Ca
2+
-Dependent-Inactivation Properties
(A) Average CRAC current densities at 280 mV induced by IP
3
(20 mM) with 10 mM EGTA in stable STIM1-expressing HEK293 cells transiently
overexpressing CRACM1 (n = 3; total n = 8 and three cells with the smallest current densities were averaged to approximate the lower current
densities of CRACM2 and CRACM3). CRAC currents were monitored continuo usly by voltage ramps spanning 2100 mV to +100 mV over 50 ms
delivered at a rate of 0.5 Hz. After CRAC currents were fully activated (120 s), rectangular voltage pulses of 2 s duration were delivered to various
negative voltages (see [D]–[F]) interspaced by 10 ramps. Error bars indicate SEM.
(B) Experimental protocol as described in (A), but for CRACM2-expressing cells. Note the minor fast inactivation and virtual absence of slow
inactivation. Error bars indicate SEM.
(C) Experimental protocol as described in (A), but for CRACM3-expressing cells. Note the significant fast inactivation and virtual absence of slow
inactivation. Error bars indicate SEM.
(D) Average CRAC currents evoked by step pulses (2 s duration) to 220 mV (green), 240 mV (blue), 260 mV (red), and 280 mV (black) in cells
expressing CRACM1 (n = 3, same cells as in [A]). At the beginning of each pulse, 2.5 ms were blanked out so that residual capacitative artifacts
could be eliminated.
(E) Average CRAC currents evoked by step pulses from 220 mV to 280 mV in cells expressin g CRACM2 (n = 4, same cells as in [B]).
(F) Average CRAC currents evoked by step pulses from 220 mV to 280 mV in cells expressing CRACM3 (n = 5, same cells as in [C]).
Functional Properties of CRACM1, CRACM2, and CRACM3
795

CRAC currents, suggesting that the CRACM1 pore
mutant indeed confers a dominant-negative effect.
Coimmunoprecipitation experiments confirmed that
CRACM1 can form stable heteromeric complexes with
both of its homologs (Figure 1D).
CRACM Homologs Differ in Ca
2+
-Dependent
Inactivation
Native CRAC currents are regulated by [Ca
2+
]
i
and sub-
ject to both fast and slow Ca
2+
-dependent inactivation
[14, 17–20]. Fast inactivation, occurring in the millisec-
ond range, is believed to result from Ca
2+
binding to
the channel itself [17, 18, 20], whereas slow inactivation
over tens of seconds may result from store refilling or
regulatory mechanisms through cellular-feedback
mechanisms on the channel [21, 22, 19].Figure 2 illus-
trates IP
3
-induced CRAC currents with intracellular so-
lutions that contained 10 mM EGTA, which is slower in
chelating Ca
2+
than BAPTA and therefore less efficient
in suppressing fast Ca
2+
-dependent inactivation [17,
20]. If sufficient Ca
2+
accumulates intracellularly, it may
overpower the buffering capacity and then reveal slow
Ca
2+
-dependent processes as well. CRAC currents
were monitored continuously by voltage ramps span-
ning 2100 mV to +100 mV over 50 ms delivered at a
rate of 0.5 Hz. After CRAC currents were fully activated,
we delivered rectangular voltage pulses of 2 s duration
and increasing hyperpolarizations so as to increase
Ca
2+
entry. Figures 2A–2C illustrate that each hyperpola-
rizing pulse caused a fast drop in CRACM1-current am-
plitude that slowly, but not completely, recovered before
the next pulse was delivered. The fast drop in current is
due to fast inactivation, and the recovery is likely to be
the net result of two opposing effects, recovery of chan-
nels from fast inactivation and slow inactivation pro-
ceeding over tens of seconds (see also Figure 3). In
the case of CRACM1, the slow inactivation resulting
from the five hyperpolarizing pulses resulted in w50%
reduction in CRAC current over a period of w100 s.
The same experimental protocol performed in cells ex-
pressing CRACM2 or CRACM3 revealed only fast inacti-
vation of currents with no significant slow inactivation
(Figures 2B and 2C). CRACM2 appeared fairly resistant
to Ca
2+
-induced inactivation in general, with only a small
component of fast inactivation, whereas CRACM3 dis-
played a much greater degree of fast inactivation. In
both cases, recovery from fast inactivation was essen-
tially complete within 20 s.
Figures 2D–2F illustrate averages of high-resolution
CRAC currents produced by the hyperpolarizing pulses
in (A)–(C), revealing the degree of fast Ca
2+
-dependent
Figure 3. CRACM1, but Not CRACM2 or CRACM3, Is Inhibited by Increased [Ca
2+
]
i
(A) Average CRAC-current densities at 280 mV induced by IP
3
(20 mM) in stable STIM1-expressing HEK293 cells transiently overexpressing
CRACM1 and perfused with increasing [Ca
2+
]
i
(n = 5–12). Error bars indicate SEM.
(B) Experimental protocol as described in (A), but for CRACM2-expressing cells (n = 5–8). Error bars indicate SEM.
(C) Experimental protocol as described in (A), but for CRACM3-expressing cells (n = 9–15). Error bars indicate SEM.
(D) Average current densities of CRACM1 (black), CRACM2 (blue), and CRACM3 (red) at 280 mV extracted at 120 s (150 s for CRACM3) from the
cells shown in (A)–(C) and plotted versus [Ca
2+
]
i
. Error bars indicate SEM.
(E) Half-maximal activation time of CRACM1 (black, n = 5–12), CRACM2 (blue, n = 5–8), and CRACM3 (red, n = 9–15) plotted versus [Ca
2+
]
i
. Data
were derived from the cells shown in (A)–(C). All cells had similar series resistances in the range of 4–6 MU. Error bars indicate SEM.
(F) Average changes in [Ca
2+
]
i
induced by store depletion in stable STIM1-expressing HEK293 cells transfected with empty vector (green, n = 14),
or transiently overexpressing CRACM1 (black, n = 23), CRACM2 (blue, n = 39), or CRACM3 (red, n = 27). The arrows indicate application of thap-
sigargin (2 mM) in Ca
2+
-free solution to induce store depletion and readmission of 2 mM Ca
2+
to probe Ca
2+
entry. The inset represents rates of
[Ca
2+
]
i
obtained by differentiating the trace segment enclosed by the rectangle.
Current Biology
796

inactivation of the three homologs. CRACM3 currents
exhibit a striking Ca
2+
-dependent inactivation that
at 280 mV is characterized by a predominant exponen-
tial decay by w80% with a time constant of t=17ms
and a very small slow component of t
2
= 130 ms. We
confirmed that this dramatic inactivation of CRACM3 is
in fact entirely due to Ca
2+
in experiments in which we
delivered a hyperpolarizing voltage pulse to 280 mV in
the presence of 10 mM Ca
2+
and after switching to
DVF solution. This revealed a rapidly inactivating current
while Ca
2+
was present, and a sustained, noninacti-
vating current when divalent cations were absent
(Figure S2). CRACM2 exhibits moderately quick Ca
2+
-
dependent inactivation, decaying with two time con-
stants of t
1
= 80 ms and t
2
= 900 ms that both contribute
in roughly equal amounts to total fast inactivation of
w50%. CRACM1 exhibits complex behavior that may
reflect three Ca
2+
-dependent feedback effects and
therefore cannot be readily assessed quantitatively in
terms of time constants. Presumably, this channel
quickly inactivates in a similar manner as CRACM2 with
two fast inactivation time courses [17, 18, 20]; however,
it appears that the second phase of fast inactivation is
partially masked by a slower wave of reactivation. This
reactivation is most pronounced at the more negative
voltage pulses and appears to be both Ca
2+
and voltage
dependent, as it was significantly attenuated, but not
abolished, when exposing cells to DVF solutions
(Figure S2B). In the absence of Ca
2+
, both CRACM1
and CRACM3 currents still increase slightly, probably
because of voltage-dependent facilitation. The slow in-
activation of CRACM1 currents is not obvious in the re-
cordings shown in Figure 2D because it occurs over tens
of seconds (see Figure 2A). However, slow inactivation
is reflected by the lower initial current amplitudes in-
duced by the most negative pulses.
To assess the slow Ca
2+
-dependent inactivation of
CRAC currents quantitatively, we perfused cells with
20 mM BAPTA and appropriate amounts of CaCl
2
so
that free [Ca
2+
]
i
was clamped to defined levels between
0 and 1 mM. Figure 3A shows that [Ca
2+
]
i
dose-depen-
dently inhibited CRACM1 currents but had little or no
significant effect on CRACM2 or CRACM3 (Figures 3B
and 3C). The absence of significant slow inactivation
seen with CRACM2 or CRACM3 is likely to be of some
importance in the physiological context because inter-
mediate [Ca
2+
]
i
levels occurring physiologically (300–
500 nM) would tend to maintain activity of CRACM2
and CRACM3 channels, whereas CRACM1 currents
would be significantly reduced. Only at 1 mM [Ca
2+
]
i
were the CRACM2 and CRACM3 currents suppressed
almost as strongly as those carried by CRACM1. It re-
mains to be determined whether the inhibitory effect
seen at this high concentration reflects direct channel
inhibition, is due to decreased coupling of STIM1 and
CRACM proteins, or is caused by decreased IP
3
efficacy
and refilling of stores.
We also examined the effect of [Ca
2+
]
i
on the kinetics
of CRAC-current activation by determining the time to
half-maximal activation (t
1/2
). We found this parameter
to be predominantly independent of [Ca
2+
]
i
for CRACM2
and CRACM1, which both had similarly fast activation
kinetics (Figure 3E). At low [Ca
2+
]
i
levels, CRACM3 cur-
rents activated significantly slower than those of the
other homologs, but they accelerated at intermediate
[Ca
2+
]
i
of 150–300 nM (Figure 3E).
Slow Ca
2+
-dependent inactivation would be expected
to at least partially affect the amount of Ca
2+
entry ob-
served in intact cells, where [Ca
2+
]
i
increases because
of CRAC-channel activity. We assessed and compared
this by monitoring fura-2 signals in cells overexpressing
the various CRACM proteins, and we subjected them to
a standard protocol where store-depletion was induced
by thapsigargin in the absence of extracellular Ca
2+
; this
was followed by readmission of 2 mM Ca
2+
for probing
store-operated Ca
2+
entry (Figure 3F). In empty-vector-
transfected cells, Ca
2+
readmission caused a moderate
increase in [Ca
2+
]
i
by store-operated entry through
endogenous CRAC channels. Cells overexpressing
CRACM homologs produced significantly greater
[Ca
2+
]
i
changes that are even more impressive when an-
alyzing the rate of Ca
2+
entry by differentiation of the
fura-2 signals (see inset in Figure 3F). Although CRACM1
is capable of generating 3-fold larger currents compared
to CRACM2 or CRACM3 when [Ca
2+
]
i
is buffered to near
zero (see Figure 1A), all three homologs achieve similar
absolute levels in [Ca
2+
]
i
and initial rates of Ca
2+
entry
when assessed by fura-2 in intact cells. Although
[Ca
2+
]
i
signals in intact cells are complex and subject
to numerous feedback mechanisms, slow Ca
2+
-depen-
dent inactivation may account at least partially for the
relatively lesser increase in [Ca
2+
]
i
observed with
CRACM1. Thus, the [Ca
2+
]
i
signals obtained in intact
cells, where global [Ca
2+
]
i
increases into the range of
300–500 nM, are comparable to the amplitudes of
CRAC currents observed when clamping global [Ca
2+
]
i
to defined levels of that range (see Figure 3D).
CRACM Homologs Differ in Selectivity
Previous work on CRACM1 has identified critical resi-
dues in three regions that affect selectivity of the
channel. Glutamate residue 106 in transmembrane
(TM) segment 1 [11, 5, 6] and glutamate residue 190 in
TM 3 [11, 5] are thought to form a ring of negatively
charged amino acids lining the pore of the channel.
Both of these residues are conserved identically in all
three CRACM homologs and are therefore unlikely to
account for differential selectivity. However, we have
previously identified a third region, located in the loop
between TM 1 and TM 2, that affects selectivity of
CRACM1 [5]. This region has three key aspartate resi-
dues (D110/D112/D114) that we have proposed to
form a second ring of negative charges that coordinate
a second Ca
2+
ion to the CRACM1 pore, and those res-
idues differ in the three homologs (CRACM2: E110/
Q112/Q114; CRACM3: E110/D112/E114). We therefore
analyzed and compared the selectivity profiles of all
three proteins with respect to Ca
2+
,Ba
2+
, and Na
+
permeation (Figure 4). In the presence of 10 mM extra-
cellular Ca
2+
, all three homologs generated large inward
currents at 280 mV (Figure 4A) and exhibited similar in-
wardly rectifying I/V relationships (Figures 4B and 4C).
When removing extracellular Ca
2+
, inward currents
were suppressed to the same degree in the three chan-
nel species (Figures 4A–4C), demonstrating that they
share similarly high Ca
2+
selectivity and discriminate
against Na
+
ions as long as Mg
2+
ions (2 mM) are
present.
Functional Properties of CRACM1, CRACM2, and CRACM3
797

We next tested whether the CRACM homologs might
exhibit different selectivities for Ba
2+
ions. Figure 4D
illustrates that equimolar substitution greatly reduces
inward currents in CRACM1, suggesting that this protein
can discriminate Ca
2+
ions against Ba
2+
. Remarkably, in
cells overexpressing CRACM2 or CRACM3, there re-
mains significantly more inward current when Ba
2+
is
used as charge carrier, and this finding at first glance
would indicate higher Ba
2+
permeation. However,
because Na
+
ions remain present in the extracellular so-
lution, there is also the possibility that Na
+
might contrib-
ute to inward current when Ba
2+
is present. Indeed,
when performing the same experiments as in Figure 4D,
but additionally replacing Na
+
with TEA, the inward cur-
rents through all three homologs were essentially abol-
ished (Figure 4E); this indicated that Na
+
ions or a mixture
of Na
+
and Ba
2+
may be carrying the current seen in
Figure 4D. Native CRAC currents in Jurkat T cells and
RBL cells have been considered to carry Ba
2+
ions [23,
18]; however, this was determined in solutions in which
both Na
+
and Ba
2+
were present. We re-examined Ba
2+
permeation in Jurkat T cells by replacing 10 mM Ca
2+
equimolarly with Ba
2+
in the presence and absence of
Na
+
and find that significant inward currents through
native CRAC channels are only recorded when both
ions are present and are absent when Ba
2+
is used as
the sole charge carrier (see Figure S3).
To further assess the selectivity of CRACM channels,
we tested for possible differences in Na
+
permeation in
divalent-free solutions and 10 mM EDTA. Under these
conditions, CRAC channels become permeable to Na
+
[17], thus typically generating a 2-fold increase in inward
current in HEK293 cells overexpressing CRACM1 (Fig-
ure 4F). The fact that the same experimental protocol
produces slightly larger CRACM2 currents, whereas
CRACM3 generates a significantly larger monovalent
current again suggests that CRACM homologs exhibit
slightly different selectivities for Na
+
ions. Although mu-
tational analysis is required to identify the contributions
of the amino acid residues responsible for these differ-
ences, it seems likely that the 110/112/114 residues
may be involved because those have been determined
Figure 4. CRACM Homologs Have Distinct
Ion Selectivity and Pharmacology
(A) Average normalized CRAC currents
at 280 mV induced by IP
3
(20 mM) in stable
STIM1-expressing HEK293 cells transiently
overexpressing CRACM1 (black, n = 12,
data taken from [5]) CRACM2 (blue, n = 8),
or CRACM3 (red, n = 10). Currents of individ-
ual cells were normalized to the current be-
fore solution change at 120 s (I/I
120s
). [Ca
2+
]
i
was clamped to near zero with 20 mM
BAPTA. The bar indicates application of
nominally Ca
2+
-free external solution. Error
bars indicate SEM.
(B) Average I/V relationships of CRACM2 cur-
rents extracted from representative cells
shown in (A) obtained at 120 s and 180 s
(n = 7). Data represent leak-subtracted cur-
rent densities (pA/pF) evoked by 50 ms volt-
age ramps from 2150 to +150 mV.
(C) Average I/V relationships of CRACM3
currents extracted from representative cells
shown in (A) at 120 s and 180 s into the exper-
iment (n = 9).
(D) Average normalized CRAC currents
(I/I
120s
)at280 mV induced by IP
3
(20 mM) in
stable STIM1-expressing HEK293 cells tran-
siently overexpressing CRACM1 (black, n =
5), CRACM2 (blue, n = 7), or CRACM3 (red,
n = 10). The bar indicates application of an ex-
ternal solution containing 10 mM Ba
2+
in the
presence of Na
+
. Error bars indicate SEM.
(E) Average normalized CRAC currents
(I/I
120s
)at280 mV induced by IP
3
(20 mM) in
stable STIM1-expressing HEK293 cells tran-
siently overexpressing CRACM1 (black, n = 9;
data taken from [5]), CRACM2 (blue, n = 6),
or CRACM3 (red, n = 6). The bar indicates
application of an external solution containing
10 mM Ba
2+
with external Na
+
being replaced
by TEA
+
. Error bars indicate SEM.
(F) Average normalized CRAC currents (I/I
120s
)at280 mV induced by IP
3
(20 mM) in stable STIM1-expressing HEK293 cells transiently over-
expressing CRACM1 (black, n = 3, data taken from [5]), CRACM2 (blue, n = 5), or CRACM3 (red, n = 10). The bar indicates application of
divalent-free external solution. Error bars indicate SEM.
(G) Average normalized CRAC currents (I/I
120s
)at280 mV induced by IP
3
(20 mM) in stable STIM1-expressing HEK293 cells transiently over-
expressing CRACM1 (black, n = 8), CRACM2 (blue, n = 4), or CRACM3 (n = 9). The bar indicates application of external solution containing
50 mM 2-APB. Error bars indicate SEM.
Current Biology
798

to contribute to monovalent permeation [5] and they are
different in the three homologs.
CRACM Homologs Differ in Pharmacology
Finally, we tested for pharmacological differences be-
tween the CRACM homologs. 2-APB has been found
to potentiate CRAC currents at low concentrations
(%5mM) and inhibit them at high concentrations
(R10 mM) [24–26]. We previously demonstrated that
CRACM1 is indeed completely inhibited by 50 mM
2-APB [8] (see Figure 4G). However, CRACM2 appears
to be significantly less sensitive because the same con-
centration reduced the current only by approximately
50%. The most striking effect, however, was observed
with CRACM3, which was not inhibited at all and instead
greatly potentiated by 50 mM 2-APB. Although the mech-
anism of action of 2-APB remains unknown and it cannot
be considered a specific pharmacological tool for CRAC
channels, the compound clearly has differential effects
on the three homologs. If these effects also apply to
native CRACM homomeric channels, it may currently
represent the best pharmacological tool to identify en-
dogenous CRAC-channel species expressed in various
cell types.
In summary, our data present a comprehensive char-
acterization of the three CRACM channels and reveal
distinct biophysical properties such as activation kinet-
ics, selectivity, Ca
2+
-dependent inactivation, and phar-
macology (Table 1). Finally, we demonstrate that the
three homologs can form heteromeric channel com-
plexes that may endow cells to express tailor-made
CRAC channels for specific Ca
2+
signaling needs. The
specific properties of CRACM channels described here
may serve as a reference for future studies aimed at
classifying the CRAC-channel composition of native
cell types as well as guidance for site-directed-
mutagenesis studies designed to localize the sites
responsible for the differences in functional and phar-
macological properties of the CRACM channels.
Supplemental Data
Experimental Procedures and three figures are available at http://
www.current-biology.com/cgi/content/full/17/9/794/DC1/.
Acknowledgments
We thank M. Bellinger for help with cell culture. This work was sup-
ported in part by National Institutes of Health grants R01-AI050200
(R.P.). C.P. was supported by a fellowship from the Deutsche
Forschungsgemeinschaft (PE-1478/1-1).
Received: March 7, 2007
Revised: March 27, 2007
Accepted: March 28, 2007
Published online: April 19, 2007
References
1. Roos, J., DiGregorio, P.J., Yeromin, A.V., Ohlsen, K., Lioudyno ,
M., Zhang, S., Safrina, O., Kozak, J.A., Wagner, S.L., Cahalan,
M.D., et al. (2005). STIM1, an essential and conserved compo-
nent of store-operated Ca
2+
channel function. J. Cell Biol. 169,
435–445.
2. Zhang, S.L., Yu, Y., Roos, J., Kozak, J.A., Deerinck, T.J., Ellis-
man, M.H., Stauderman, K.A., and Cahalan, M.D. (2005).
STIM1 is a Ca
2+
sensor that activates CRAC channels and mi-
grates from the Ca
2+
store to the plasma membrane. Nature
437, 902–905.
3. Feske, S., Gwack, Y., Prakriya, M., Srikanth, S., Puppel, S.H.,
Tanasa, B., Hogan, P.G., Lewis, R.S., Daly, M., and Rao, A.
(2006). A mutation in Orai1 causes immune deficiency by abro-
gating CRAC channel function. Nature 441, 179–185.
4. Vig, M., Peinelt, C., Beck, A., Koomoa, D.L., Rabah, D., Koblan-
Huberson, M., Kraft, S., Turner, H., Fleig, A., Penner, R., et al.
(2006). CRACM1 is a plasma membrane protein essential for
store-operated Ca
2+
entry. Science 312, 1220–1223.
5. Vig, M., Beck, A., Billingsley, J.M., Lis, A., Parvez, S., Peinelt, C.,
Koomoa, D.L., Soboloff, J., Gill, D.L., Fleig, A., et al. (2006).
CRACM1 multimers form the ion-selective pore of the CRAC
channel. Curr. Biol. 16, 2073–2079.
6. Yeromin, A.V., Zhang, S.L., Jiang, W., Yu, Y., Safrina, O., and
Cahalan, M.D. (2006). Molecular identification of the CRAC
channel by altered ion selectivity in a mutant of Orai. Nature
443, 226–229.
7. Mercer, J.C., Dehaven, W.I., Smyth, J.T., Wedel, B., Boyles,
R.R., Bird, G.S., and Putney, J.W., Jr. (2006). Large store-oper-
ated calcium selective currents due to co-expression of Orai1
or Orai2 with the intracellular calcium sensor, Stim1. J. Biol.
Chem. 281, 24979–24990.
8. Peinelt, C., Vig, M., Koomoa, D.L., Beck, A., Nadler, M.J.,
Koblan-Huberson, M., Lis, A., Fleig, A., Penner, R., and Kinet,
J.P. (2006). Amplification of CRAC current by STIM1 and
CRACM1 (Orai1). Nat. Cell Biol. 8, 771–773.
9. Soboloff, J., Spassova, M.A., Tang, X.D., Hewavitharana, T., Xu,
W., and Gill, D.L. (2006). Orai1 and STIM reconstitute store-oper-
ated calcium channel function. J. Biol. Chem. 281, 20661–20665.
10. Zhang, S.L., Yeromin, A.V., Zhang, X.H., Yu, Y., Safrina, O.,
Penna, A., Roos, J., Stauderman, K.A., and Cahalan, M.D.
(2006). Genome-wide RNAi screen of Ca
2+
influx identifies genes
that regulate Ca
2+
release-activated Ca
2+
channel activity. Proc.
Natl. Acad. Sci. USA 103, 9357–9362.
11. Prakriya, M., Feske, S., Gwack, Y., Srikanth, S., Rao, A., and
Hogan, P.G. (2006). Orai1 is an essential pore subunit of the
CRAC channel. Nature 443, 230–233.
12. Gwack, Y., Srikanth, S., Feske, S., Cruz-Guilloty, F., Oh-Hora, M.,
Neems, D.S., Hogan, P.G., and Rao, A. (2007). Biochemical and
functional characterization of Orai family proteins. J. Biol.
Chem., in press. Published online February 9, 2007. 10.1074/
jbc.M609630200.
Table 1. Properties of the Mammalian CRACM Proteins
CRACM1 CRACM2 CRACM3
Store-operated Yes Yes Yes
Activation time (t
1/2
)3567s 2163s 6367s
Ca
2+
-dependent inactivation (fast) Moderate Moderate Strong
Ca
2+
-dependent inactivation (slow) Strong None None
Ca
2+
-dependent reactivation Yes No No
Selectivity Ca
2+
>> Na
+
,Ba
2+
Ca
2+
>> Na
+
,Ba
2+
Ca
2+
>> Na
+
,Ba
2+
Monovalent permeation in DVF solutions Moderate Moderate Strong
2-APB effect at 50 mM Block Reduction Potentiation
Functional Properties of CRACM1, CRACM2, and CRACM3
799

13. Penner, R., Matthews, G., and Neher, E. (1988). Regulation of
calcium influx by second messengers in rat mast cells. Nature
334, 499–504.
14. Hoth, M., and Penner, R. (1992). Depletion of intracellular cal-
cium stores activates a calcium current in mast cells. Nature
355, 353–356.
15. Parekh, A.B., and Penner, R. (1997). Store depletion and calcium
influx. Physiol. Rev. 77, 901–930.
16. Parekh, A.B., and Putney, J.W., Jr. (2005). Store-ope rated
calcium channels. Physiol. Rev. 85, 757–810.
17. Hoth, M., and Penner, R. (1993). Calcium release-activated
calcium current in rat mast cells. J. Physiol. 465, 359–386.
18. Zweifach, A., and Lewis, R.S. (1995). Rapid inactivati on of deple-
tion-activated calcium current (ICRAC) due to local calcium
feedback. J. Gen. Physiol. 105, 209–226.
19. Parekh, A.B. (1998). Slow feedback inhibition of calcium release-
activated calcium current by calcium entry. J. Biol. Chem. 273,
14925–14932.
20. Fierro, L., and Parekh, A.B. (1999). Fast calcium-dependent in-
activation of calcium release-activated calcium current (CRAC)
in RBL-1 cells. J. Membr. Biol. 168, 9–17.
21. Parekh, A.B., and Penner, R. (1995). Depletio n-activated calcium
current is inhibited by protein kinase in RBL-2H3 cells. Proc.
Natl. Acad. Sci. USA 92, 7907–7911.
22. Zweifach, A., and Lewis, R.S. (1995). Slow calcium-dependent
inactivation of depletion-activated calcium current. Store-
dependent and -independent mechanisms. J. Biol. Chem. 270,
14445–14451.
23. Hoth, M. (1995). Calcium and barium permeation through
calcium release-activated calcium (CRAC) channels. Pflugers
Arch. 430, 315–322.
24. Braun, F.J., Broad, L.M., Armstrong, D.L., and Putney, J.W., Jr.
(2001). Stable activation of single Ca
2+
release-activated Ca
2+
channels in divalent cation-free solutions. J. Biol. Chem. 276,
1063–1070.
25. Prakriya, M., and Lewis, R.S. (2001). Potentiation and inhibition
of Ca
2+
release-activated Ca
2+
channels by 2-aminoethyldi-
phenyl borate (2-APB) occurs independently of IP
3
receptors.
J. Physiol. 536, 3–19.
26. Hermosura, M.C., Monteilh-Zoller, M.K., Scharenberg, A.M.,
Penner, R., and Fleig, A. (2002). Dissociation of the store-
operated calcium current I
CRAC
and the Mg-nucleotide-
regulated metal ion current MagNuM. J. Physiol. 539, 445–458.
Current Biology
800

Supplemental Data S1
CRACM1, CRACM2, and CRACM3
Are Store-Operated Ca
2+
Channels
with Distinct Functional Properties
Annette Lis, Christine Peinelt, Andreas Beck,
Suhel Parvez, Mahealani Monteilh-Zoller,
Andrea Fleig, and Reinhold Penner
Supplemental Experimental Procedures
Subcloning and Overexpression
Full-length human CRACM1 and CRACM1-E106Q were subcloned
as described [S1]. Full-length human CRACM2 (accession no.
NM_032831) and CRACM3 (accession no. NM_152288) were ampli-
fied from cDNAs (purchased from OriGene) with Pfu Ultra High Fidel-
ity polymerase (Stratagene) and subcloned into a pCAGGS-
IRES-GFP vector [S2]. We introduced the ribosome-binding site
ACC GCC ACC and a HA-tag in frame immediately 50to the start co-
don of CRACM2 and CRACM3 cDNAs, which were subsequently
cloned into pCAGGS-IRES-GFP for transient dicistronic expression
of CRACM2 and CRACM3 together with the green fluorescent pro-
tein (GFP). For electrophysiological analysis, CRACM proteins were
overexpressed in HEK293 cells stably expressing STIM1 [S3] with
lipofectamine 2000 (Invitrogen), and the GFP expressing cells were
selected by fluorescence. Experiments were performed 24–48 hr
after transfection.
Immunoprecipitation
HEK293 cells were transiently cotransfected with CRACM1-Myc [S4]
and HA-CRACM1, HA-CRACM2, and HA-CRACM3 (described
above). Forty-eight hours after transfection, cells were harvested
in PBS and lysed in 1 ml lysis buffer with the following: 75 mM
NaCl, 40 mM NaF, 20 mM Iodocetamide, 50 mM HEPES, 1%
IGEPAL, 1 mM phenylmethylsulfonyl fluoride (PMSF), 0.25 mM
sodium orthovanadate, and protease inhibitor cocktail (Sigma). The
cell lysates were precipitated with anti-HA rat monoclonal antibody
(2.5 mg, Roche) or anti-c-Myc mouse monoclonal antibody (2.5 mg,
Calbiochem) for 2 hr at 4C. Samples were resolved by SDS-PAGE
and analyzed with anti-HA rat monoclonal antibody at a dilution
1:1000. Anti-Rat IgG (whole molecule) peroxidase conjugate (Sigma)
were used as secondary antibody in accordance with the manufac-
turer’s instructions. Proteins were detected by development with
the ECL Plus Western Blotting Detection System (Amersham).
Electrophysiology
Patch-clamp experiments were performed in the tight-seal whole-
cell configuration at 21C–25C. High-resolution current recordings
were acquired with the EPC-9 (HEKA). Voltage ramps of 50 ms du-
ration spanning a range of 2150 to +150 mV were delivered from
a holding potential of 0 mV at a rate of 0.5 Hz over a period of
100–300 s. All voltages were corrected for a liquid junction potential
of 10 mV. Currents were filtered at 2.9 kHz and digitized at 100 ms
intervals. Capacitive currents were determined and corrected be-
fore each voltage ramp. Extracting the current amplitude at 280
mV from individual ramp current records assessed the low-resolu-
tion temporal development of currents. Where applicable, statistical
errors of averaged data are given as means 6SEM with n determi-
nations. Standard external solutions were as follows: 120 mM NaCl,
2 mM MgCl
2
, 10 mM CaCl
2
, 10 mM TEA-Cl, 10 mM HEPES, 10 mM
glucose, pH 7.2 with NaOH, 300 mOsm. In some experiments, we
applied Na
+
-free solutions, where NaCl was replaced equimolarly
by tetraethylammonium-chloride (TEA-Cl). For Ca
2+
-free external
solutions CaCl
2
was omitted, but Mg
2+
was retained. The divalent-
free external solution (DVF) was based on the standard external so-
lution but in the absence of CaCl
2
and MgCl
2
and was additionally
supplemented with 10 mM EDTA. Divalent replacement solutions
were based on the standard external solution but with 10 mM
CaCl
2
replaced by 10 mM BaCl
2
. In some experiments, 2-aminoe-
thyldiphenyl borate (2-APB) was added to the standard external
solution at a final concentration of 50 mM. Standard internal solu-
tions were as follows: 120 mM Cs-glutamate, 20 mM Cs$BAPTA,
3 mM MgCl
2
, 10 mM HEPES, 0.02 mM IP
3
, pH 7.2 with CsOH,
300 mOsm. In the experiments of Figure 2, 10 mM EGTA was used,
and in Figure 3, [Ca
2+
]
i
was buffered to defined levels with 20 mM
Cs$BAPTA, and appropriate concentrations of CaCl
2
as calculated
with WebMaxC (http://www.stanford.edu/wcpatton/webmaxcS.
htm). For passive-depletion experiments, IP
3
was omitted from
the internal solution. All chemicals were purchased from Sigma-
Aldrich.
Figure S1. Store-Operated Currents Induced by BAPTA
(A) Average CRAC current densities after store depletion with 20 mM BAPTA and omitting IP
3
in cells expressing CRACM1 (black, n = 12),
CRACM2 (blue, n = 7), and CRACM3 (red, n = 7). Currents were analyzed as shown in (A).
(B) Average I/V traces extracted from representative HEK293 cells shown in (A) at 300 s into the experiment. Traces correspond to CRACM1
(black, n = 9), CRACM2 (blue, n = 7), and CRACM3 (red, n = 7).

Fluorescence Measurements
For Ca
2+
measurements, fura-2 AM (Molecular Probes)-loaded cells
(1 mM/60 min/37C) were kept in extracellular saline containing the
following: 140 mM NaCl, 2.8 mM KCl, 2 mM MgCl
2
, 10 mM glucose,
and 10 mM HEPES$NaOH (pH 7.2). Store depletion was induced by
addition of 2 mM thapsigargin to the bath, and for assessing store-
operated Ca
2+
entry, 2 mM Ca
2+
was added. Experiments were
performed with a Zeiss Axiovert 100 fluorescence microscope
equipped with a dual excitation fluorometric imaging system
(TILL-Photonics), with a 403Plan NeoFluar objective. Data acqui-
sition and computation was controlled by TILLvisION software.
Dye-loaded cells were excited by wavelengths of 340 and 380 nm,
produced by a monochromator (Polychrome IV). The fluorescence
emission of several single cell bodies was simultaneously recorded
with a video camera (TILL-Photonics Imago) with an optical 440 nm
long-pass filter. The signals were sampled at 0.5 Hz and computed
into relative ratio units of the fluorescence intensity at the different
wavelengths (340/380 nm). Results are given as the approximate
[Ca
2+
]
i
, calculated from the 340/380 nm fluorescence values, with
an in vivo Ca
2+
calibration performed in patch-clamp experiments
with defined Ca
2+
concentrations combined with fura-2 in the patch
pipette.
Supplemental References
S1. Peinelt, C., Vig, M., Koomoa, D.L., Beck, A., Nadler, M.J.,
Koblan-Huberson, M., Lis, A., Fleig, A., Penner, R., and Kinet,
Figure S2. Ca
2+
-Dependent Inactivation of CRACM3 Currents
(A) Average CRAC current densities at 280 mV induced by IP
3
(20 mM) with 10 mM EGTA in stable STIM1-expressing HEK293 cells transiently
overexpressing CRACM1 (n = 3). CRAC currents were monitored continuously by voltage ramps spanning 2100 mV to +100 mV over 50 ms
delivered at a rate of 0.5 Hz. After CRAC currents were fully activated (120 s), a rectangular voltage pulses of 2 s duration was delivered
to 280 mV (see [B]). Then the cell was exposed to divalent-free (DVF) extracellular solution and another voltage pulse was applied.
(B) Average CRAC currents evoked by step pulses (2 s duration) to 280 mV in the presence of 10 mM Ca
2+
(black) and in DVF solution (red, n = 3,
same cells as shown in [A]). Note the loss of initial fast inactivation and subsequent reactivation in DVF solution. The remaining slow increase in
inward currents is probably voltage-dependent facilitation.
(C) Same experimental conditions and protocol as in (A), but in stable STIM1-expressing HEK293 cells transiently overexpressing CRACM3
(n = 3).
(D) Average CRAC currents evoked by step pulses (2 s duration) to 280 mV in the presence of 10 mM Ca
2+
(black) and in DVF solution (red, n = 3,
same cells as shown in [C]). Note the loss of inactivation in DVF solution, revealing the same slow facilitation as CRACM1 that is presumably
voltage dependent.
S2

J.P. (2006). Amplification of CRAC current by STIM1 and
CRACM1 (Orai1). Nat. Cell Biol. 8, 771–773.
S2. Warnat, J., Philipp, S., Zimmer, S., Flockerzi, V., and Cavalie, A.
(1999). Phenotype of a recombinant store-operated channel:
Highly selective permeation of Ca2+. J. Physiol. 518, 631–638.
S3. Soboloff, J., Spassova, M.A., Hewavitharana, T., He, L.P., Xu,
W., Johnstone, L.S., Dziadek, M.A., and Gill, D.L. (2006).
STIM2 is an inhibitor of STIM1-mediated store-operated Ca2+
entry. Curr. Biol. 16, 1465–1470.
S4. Vig, M., Peinelt, C., Beck, A., Koomoa, D.L., Rabah, D., Koblan-
Huberson, M., Kraft, S., Turner, H., Fleig, A., Penner, R., et al.
(2006). CRACM1 is a plasma membrane protein essential for
store-operated Ca2+ entry. Science 312, 1220–1223.
Figure S3. Ba
2+
Conductivity in Jurkat T Cells
(A) Average normalized CRAC currents (I/I
120s
)at280 mV induced by IP
3
(20 mM) in Jurkat T cells. The bar indicates application of an external
solution containing 10 mM Ba
2+
in the presence of Na
+
(red, n = 8) and when Na
+
was replaced by TEA
+
(black, n = 5)
(B) Average I/V relationships of CRAC currents extracted from cells shown in (A), obtained at 120 s (black) and 180 s (red) during Ba
2+
application
in the presence of Na
+
(n = 8). Data represent leak-subtracted current densities (pA/pF) evoked by 50 ms voltage ramps from 2150 to +150 mV
(voltage range shown is from 2100 to +80 mV).
(C) Average I/V relationships of CRAC currents from cells shown in (A) obtained at 120 s (black), 126 s (blue), and 180 s (red) during Ba
2+
appli-
cation when Na
+
was replaced by TEA
+
(n = 5).
S3