JOURNAL CLUB
11/17/15
ZHENGFENG YANG
SARAF Inactivates the Store Operated
Calcium Entry Machinery to Prevent
Excess Calcium Refilling
Raz Palty,1 Adi Raveh,1 Ido Kaminsky,1 Ruth Meller,1 and Eitan Reuveny1,*
1Department of Biological Chemistry, Weizmann Institute of Science, Rehovot 76100, Israel
*Correspondence: [email protected]
DOI 10.1016/j.cell.2012.01.055
SUMMARY
Store operated calcium entry (SOCE) is a principal
cellular process by which cells regulate basal
calcium, refill intracellular Ca2+ stores, and execute
a wide range of specialized activities. STIM and
Orai proteins have been identified as the essential
components enabling the reconstitution of Ca2+
release-activated Ca2+ (CRAC) channels that mediate SOCE. Here, we report the molecular identification of SARAF as a negative regulator of SOCE.
Using heterologous expression, RNAi-mediated
silencing and site directed mutagenesis combined
with electrophysiological, biochemical and imaging
techniques we show that SARAF is an endoplasmic
reticulum membrane resident protein that associates
with STIM to facilitate slow Ca2+-dependent inactivation of SOCE. SARAF plays a key role in shaping
cytosolic Ca2+ signals and determining the content
of the major intracellular Ca2+ stores, a role that is
likely to be important in protecting cells from Ca2+
overfilling.
INTRODUCTION
Store operated calcium entry (SOCE) represents a key mechanism, by which cells convey Ca2+ signals and maintain Ca2+
homeostasis (Parekh and Putney, 2005). Despite its initial
characterization more than two decades ago, when SOCE was
characterized mainly via mechanistic and biophysical means,
only recently have the molecular components essential for its
activity been identified. The molecular identification revealed
the discrete events associated with SOCE activation and regulation (Cahalan, 2009; Lewis, 2007; Putney, 1986). STIM and Orai
proteins are the essential components that enable the reconstitution of Ca2+ release-activated Ca2+ (CRAC) channels underlying SOCE activity (Feske et al., 2006; Liou et al., 2005; Peinelt
et al., 2006; Prakriya et al., 2006; Roos et al., 2005; Soboloff
et al., 2006; Vig et al., 2006; Zhang et al., 2006; Zhang et al.,
2005). Although other proteins, such as members of the TRPC
family, might also be involved in SOCE (Hardie and Minke,
1993; Kiselyov et al., 1998; Liu et al., 2007; Pani et al., 2009),
STIM, the Ca2+ sensor (Liou et al., 2005; Roos et al., 2005;
Stathopulos et al., 2008; Zhang et al., 2005), and Orai, the
channel pore forming subunit (Prakriya et al., 2006; Vig et al.,
2006; Yeromin et al., 2006), have been shown to be sufficient
for this activity. STIM is an endoplasmic reticulum (ER) single
pass membrane protein (Hogan et al., 2010), which detects
changes in ER Ca2+ levels through a conserved Ca2+ binding
domain (EF hand) (Stathopulos et al., 2009; Stathopulos et al.,
2008). ER Ca2+ depletion leads to STIM oligomerization and
relocalization to specialized regions close to the plasma
membrane (ER-PM junctions), where it directly binds to Orai,
causing the opening of the channel and initiation of Ca2+ entry
(Baba et al., 2006; Barr et al., 2008; Frischauf et al., 2009; Kawasaki et al., 2009; Liou et al., 2007; Liou et al., 2005; Muik et al.,
2008; Navarro-Borelly et al., 2008; Park et al., 2009; Wu et al.,
2006; Yuan et al., 2009; Zhang et al., 2005; Zhou et al., 2010).
Once SOCE is activated, it is subjected to various direct and
indirect regulatory processes that determine the duration and
magnitude of the Ca2+ influx, which set downstream cellular
events (Hogan et al., 2010; Parekh and Putney, 2005). Two definitive Ca2+-dependent modes of regulation of CRAC channels
were previously reported: fast inactivation that depends largely
on STIM1 and calmodulin binding to Orai1 (Derler et al., 2009;
Hoth and Penner, 1993; Lee et al., 2009; Litjens et al., 2004;
Mullins et al., 2009) and a slow Ca2+-dependent inactivation
process with an unknown molecular mechanism (Louzao et al.,
1996; Parekh, 1998; Zweifach and Lewis, 1995).
Using a functional-based high-throughput screen, we have
identified the gene product of TMEM66, named hereafter as
SARAF (for SOCE-associated regulatory factor) as a regulator
of cellular Ca2+ homeostasis. SARAF is a highly conserved
protein in vertebrates and has poor functional annotation. In
mammals, SARAF is ubiquitously expressed but has exceptionally high transcript levels in the immune and neuronal tissues
(Su et al., 2004). In this study we show that SARAF is an ER resident protein, which responds to cytosolic Ca2+ elevation after
ER Ca2+ refilling by promoting a slow inactivation process of
STIM2-dependent basal SOCE activity, as well as STIM1-mediated SOCE activity. These actions collectively impose stringent
conditions for maintaining proper intracellular Ca2+ levels in
both resting and stimulated cells.
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. 425
RESULTS
SARAF Regulates Basal ER and Cytosolic Ca2+ Levels
We performed a functional expression screen in an attempt to
isolate cDNA candidates that affect mitochondrial Ca2+ homeostasis (Figures S1A and S1B, available online). These efforts lead
to the identification of a 2 kilobase cDNA clone with a 1020 bp
open reading frame coding for SARAF (accession #JQ348891)
(also known as TMEM66, XTP-3, FOAP-7, HSPC035,
MGC8721 or FLJ22274) (Figure 1A). Transfection of HEK293
cells with SARAF cDNA alone resulted in lowered basal mitochondrial Ca2+ levels compared to control cells (Figures S1A
and S1B).
Because changes in basal mitochondrial Ca2+ levels may
reflect changes in global cellular Ca2+ homeostasis, we also
tested the effect of SARAF expression on cytosolic and ER
Ca2+ levels. These parameters were measured with a fura-2 or
FRET-based D1ER indicator (Palmer et al., 2004), respectively.
Transfection of HeLa or HEK293-T cells with SARAF cDNA, or
with an empty vector (control), showed that basal [Ca2+]ER was
significantly decreased in SARAF-overexpressing cells (Figure 1B). In order to assess the relationship between SARAF
expression levels and [Ca2+]cyto, we expressed SARAF C terminally tagged with yellow fluorescent protein (YFP) (SARAF-YFP)
or YFP (control), and measured [Ca2+]cyto as a function of protein
expression, as judged by YFP fluorescence. We found that
decreased [Ca2+]cyto highly correlated with YFP fluorescence in
cells expressing SARAF-YFP but not in cells expressing YFP
alone (Figure 1C), indicating that SARAF expression primarily
affected global Ca2+ homeostasis. In order to assess the role
of endogenous SARAF expression, a siRNA silencing approach
against SARAF was used in both HeLa and Jurkat cells. Reduced
expression of SARAF resulted in a significant elevation of basal
[Ca2+]cyto in both cell types (Figures 1D–1F). Likewise, [Ca2+]ER
levels, estimated by organelle content release using ionomycin
(Brandman et al., 2007) or with the D1ER indicator, were higher
in siRNA-treated cells, (Figure 1F inset and Figure S1D). The
specificity and efficiency of siRNA-mediated silencing of SARAF
mRNA and protein levels were determined by western blotting,
semiquantitative RT-PCR, and a fluorescent-based assay employing YFP fluorescence as reference for protein expression
(Figure 1D and Figure S2). The functionality of SARAF-YFP suggested that the protein is correctly expressed and folded and
therefore encouraged us to study its cellular localization. When
a fluorescent-tagged chimera of SARAF was coexpressed with
either that of STIM1 or SEC61, both ER resident proteins, SARAF
fluorescence strongly overlapped with both markers (Figure 1G
and Figure S1E), but did not colocalize with that of mitochondrial
markers (data not shown), indicating that the ER is the primary
site for SARAF subcellular localization.
At resting conditions, SARAF modulation of [Ca2+]i must
involve a change of Ca2+ fluxes at the level of the plasma
membrane (Rı́os, 2010). This can be achieved either through
the regulation of plasma membrane Ca2+ permeation or extrusion pathways. In order to distinguish between these two possibilities, we overexpressed SARAF in HEK293-T or HeLa cells
and tested whether SARAF-mediated reduction of [Ca2+]cyto depended on the presence of Ca2+ in the growth media. Control
426 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
and SARAF overexpressing cells grown on a low Ca2+-containing media (0.1 mM), exhibited similar cytosolic Ca2+ levels (Figure S1C), suggesting that increased SARAF expression is linked
to inhibition of Ca2+ entry from an extracellular source rather than
the activation of a Ca2+-extrusion mechanism. Recent findings
by Meyer and colleagues showed that STIM2 plays a central
role in maintaining basal [Ca2+]cyto and [Ca2+]ER by acting as
a feedback regulator that senses small changes in ER Ca2+
and translates them into differential gating of Orai1 (Brandman
et al., 2007). To test whether SARAF regulation of basal [Ca2+]cyto
is linked to the STIM2-Orai1 pathway, we used siRNA-mediated
knock down of STIM2, either alone, or together with SARAF, and
determined basal [Ca2+]cyto in HeLa cells. In agreement with
previous studies (Bird et al., 2009; Brandman et al., 2007), we
observed a significant decrease in [Ca2+]cyto after STIM2
silencing, compared to control (Figure 1H). More importantly,
silencing both STIM2 and SARAF did not result in an additive
effect on resting [Ca2+]cyto. These results suggest that SARAF,
with STIM2, may act on a similar Ca2+ permeation pathway.
SARAF Regulates STIM1-Orai1-Mediated Store
Operated Calcium Entry
STIM2 and STIM1 share moderate sequence similarity and maintain similar domain architecture (Hogan et al., 2010). To test
whether SARAF regulates STIM1-Orai1-mediated SOCE, we
heterologously expressed STIM1 and Orai1 alone or with SARAF
in HEK293 or HEK293-T cells and measured ICRAC, with the
whole-cell variation of the patch clamp technique by passively
depleting Ca2+ from the ER by using ethylene glycol-bis(2-aminoethylether)-N,N,N0 ,N0 -tetraacetic acid (EGTA) (1.2 mM) in the
recording pipette (Zweifach and Lewis, 1995). After 3 min under
whole-cell configuration in zero extracellular Ca2+, addition of
10 mM Ca2+ to the bath solution initiated La3+-sensitive strong
inwardly rectifying currents typical of ICRAC (Figures S2A and
S2B). These had similar magnitudes in both control- expressing
(HEK293 22.6 ± 2 pA/pF and HEK239-T 40.7 ± 3.5 pA/pF) and
SARAF- expressing (HEK293 27.1 ± 6pA/pF and HEK293-T
41.8 ± 3.3 pA/pF) cells. Strikingly, we found clear differences in
current inactivation. In control HEK293 cells, currents partially
inactivated to steady-state levels (35.6% ± 3.2% of peak levels),
whereas in SARAF-expressing cells, the extent of current
inactivation was greatly increased (17.3% ± 1% of peak levels,
Figure 2A). An additional period of passive store depletion,
when no external Ca2+ was present, partially restored ICRAC in
control cells (46.5% ± 4.7% of peak levels), whereas in
SARAF-expressing cells, currents remained inactivated (21% ±
3% of peak levels). These results indicate that part of the
SARAF-mediated ICRAC inhibition does not require the continuous presence of elevated Ca2+. Similar results were also
obtained from HEK293-T cells under the same experimental
conditions (Figures S3A and S3B). Intriguingly, a similar ‘‘slow
inactivation’’ process of native ICRAC currents in Jurkat
T-lymphocytes and RBL cells has been observed (Parekh,
1998; Zweifach and Lewis, 1995); however, the molecular
component responsible for this process remained unknown.
The continuous nature of the SARAF-mediated ICRAC inhibition
suggests that inhibition may be Ca2+ dependent. We therefore
performed similar experiments as described above using the
A
B
C
D
E
F
G
H
Figure 1. SARAF Regulates Basal Cytosolic and ER Ca2+ Concentrations by Inhibiting STIM2-Dependent Basal SOCE
(A) The predicted topology and domain structure of the human SARAF. Human SARAF has 339 amino acids with a validated signal peptide (S.Pep) (1–30
[Zhang and Henzel, 2004]) and a putative single membrane spanning domain (173–195, predicted by the SOSUI program). Initiation point of SARAF short isoform
is indicated (aa 173). Also marked are clusters of serine/proline (284–310) and arginine rich regions.
(B) ER Ca2+ levels were measured in control and heterologously expressing SARAF cells, together with D1ER. Averaged D1ER emission ratios from individual
cells are shown (HEK293-T [control, n = 76; SARAF, n = 86], HeLa [control, n = 71; SARAF n = 61]).
(C) Cytosolic Ca2+ levels and YFP expression were measured in HEK293-T cells transfected with SARAF-YFP or YFP (control). Expression units that cover the
entire YFP fluorescence range were created by grouping 35–50 cells distributed around similar YFP fluorescent values (SARAF-YFP, n = 228; control-YFP,
n = 217). The averaged basal Ca2+ levels are plotted as a function of expression groups.
(D) Immunoblot analysis of SARAF, STIM1 and Orai1 expressed in HeLa cells before or after treatment with siSARAF or siGFP (control).
(E) Cytosolic Ca2+ levels of HeLa cells transfected with siSARAF (n = 118) or siGFP (n = 185), and of Jurkat cells (siSARAF, n = 236) (siGFP, n = 247).
(F) Averaged ER Ca2+ measured in individual cells transfected with siSARAF or siControl (nontarget, see Extended Experimental Procedures). ER Ca2+ was
estimated by monitoring the peak of ionomycin-induced Ca2+ released from the ER in the presence of extracellular EGTA (3 mM, inset).
(G) Fluorescent images of a HEK293-T cell expressing SARAF-GFP (green) and STIM1-mCherry (red); the scale bar represents10 mm. (H) Cytosolic Ca2+ levels of
HeLa cells transfected siSARAF (n = 250), si STIM2 (n = 417), siSTIM2+siSARAF (n = 481), or siControl (n = 289). p value: * < 0.05, ** < 0.01, *** < 0.001.
faster Ca2+ chelator, 1,2-bis(o-aminophenoxy)ethane-N,N,N’,N’tetra-acetic acid (BAPTA) (10 mM), in the pipette solution. Under
these experimental conditions, ICRAC currents were of similar
magnitude for both control and SARAF-expressing cells and
did not undergo inactivation (Figure 2B). These results suggest
that the onset of SARAF-mediated ICRAC inactivation depends
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. 427
Normalized I
-20
-40
-60
-80
control
SARAF
-100
0 100 200 300 400 500 600 700
Time (s)
0
10mM
0 10mM
Ca2+
20
0
-20
-40
-60
-80
-100
-120
0
control
SARAF
80 160 240 320 400 480 560 640
Time (s)
D
2mM Ca
0 Ca
2mM Ca
0.6 ATP
Tg
+Crb
0.4
0.2
control
SARAF
non-transfected
0
-80
Time (s)
2+
2+
0 Ca
0.4
0.3
0.2
CD3
0.6
0.4
0.2
siRNA SERAF
siRNA control
0
0
200 400 600 800 1000
BHQ, t0
BHQ, t30
*
1.5
0 Ca
2+
2+
2mM Ca 0 Ca
0.1
80
160 240 320 400
Time (s)
2+
0 Ca
2+
20mM Ca
2+
0
1
BHQ
0.5
0
0
Time (s)
control
SARAF
non-transfected
H
2+
pA/pF
2mM Ca
Ratio (340/380)
Ratio (340/380)
2+
2mM Ca
2+
+ serotonin 0 Ca
2+
0.5
0
G
1 0 Ca
0.8
0 Ca
Time (s)
F
control
SARAF
0 100 200 300 400 500 600 700
200 400 600 800 1000 1400 1600 1800 2000 2200 2400
0
-20
-100
2+
0.8
Ca2+
-60
Ratio (340/380)
0 Ca
2+
10mM
0
0
-40
-0.5
-1
-1.5
siSeraf + SERAF
siSERAF
siControl
-2
0
200 400 600 800 10001200
Time (s)
siRNA SARAF
siRNA control
80
160
240
Time (s)
Tg, t30
BHQ
J
Tg
Orai1/
STIM1
** *
*
* *
*
*
428 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
Orai1/
STIM1/
SARAF
control
Nuclear translocation of
NFAT(% of total)
Ratio (340/380)
2+
10mM
0
E
2+
I
(%)
Ca2+
EGTA+Tg
Normalized I
10mM
0
(%)
(%)
CRAC
Normalized I
10mM
0
0
C
Bapta
CRAC
B
EGTA
CRAC
A
control
Orai1/STIM1
Orai1/STIM1/SARAF
100
80
60
40
**
20
0
BHQ
Tg
320
on the presence of free cytosolic Ca2+ ions. To address more
specifically whether ER luminal Ca2+ content affects SARAFmediated ICRAC inhibition, we repeated the experimental paradigm as in Figure 2A with the addition of 2 mM Thapsigargin
(Tg), an irreversible SERCA (sarcoplasmic/endoplasmic reticulum Ca2+-ATPase), in the patch pipette solution. In the absence
of ER Ca2+ refilling, both control and SARAF-expressing cells
exhibited reduced ICRAC inactivation (Figure 2C) compared to
conditions that allow ER Ca2+ refilling (Figure 2A), indicating
that store refilling partially contributes to the inactivation
process. In the absence or presence of Tg, the extent of current
inactivation in control cells was 35.6% ± 3.2% (from Figure 2A)
and 80.6% ± 4.4% of peak levels, respectively. Similar changes
were seen in cells expressing SARAF, with 17.3% ± 1% (from
Figure 2A) and 40% ± 1.6% of peak levels, respectively. These
data suggest that there is a substantial component of ICRAC inactivation that is dependent on luminal Ca2+, which in turn may be
partially modulated by SARAF. In order to further confirm that
SARAF is indeed involved in regulating Ca2+ entry upon ER
Ca2+ depletion, we monitored SOCE-induced [Ca2+]i changes
under conditions that either allowed or prevented ER refilling.
To equilibrate both [Ca2+]cyto and [Ca2+]ER in control and
SARAF-expressing cells, we incubated cells in a Ca2+-free extracellular solution (1 mM EGTA) (Brandman et al., 2007) for 45 min
prior to the initiation of experiment. Transient ER Ca2+ depletion
was induced by bath application of 100 mM ATP and 100 mM
carbachol (Carb) (Figure 2D). In cells coexpressing Orai1,
STIM1, and SARAF, the rise in [Ca2+]cyto after the addition of
extracellular Ca2+ was about 2-fold lower than in STIM1- and
Orai1-coexpressing cells (Figure 2D). In contrast, under conditions where SOCE activity was induced by Tg, the apparent
ability of SARAF to inhibit SOCE activity was greatly diminished.
It should be noted that a small but persistent difference in
[Ca2+]cyto rise under these conditions was maintained between
SARAF-expressing and control cells (an average of 13.9% ±
2.2% in the relative Rmax value, from seven independent experiments, p < 0.05, t test analysis). The apparent nonlinear correlation between the effects of SARAF on steady-state ICRAC
currents and on [Ca2+]cyto may arise because of the nature of
the recording method, i.e., direct Ca2+ ion flow versus cumulative
determination [Ca2+]cyto with electrophysiology and fluorescence Ca2+ imaging, respectively. Nevertheless, to exclude the
possibility that SARAF somehow interfered nonspecifically with
cellular Ca2+ signaling, we monitored the serotonin-dependent
Ca2+ influx through the 5HT3 receptor (5HT3R), in HEK293-T
cells. SARAF expression did not affect 5HT3R-mediated Ca2+
influx and thus a nonspecific effect of SARAF on Ca2+ permeation is ruled out (Figure 2E). To substantiate that indeed SARAF
is playing an important role in regulating SOCE activity under
native conditions, we measured SOCE activity in Jurkat T cells
where endogenous SARAF levels were reduced by siRNA. In
these cells, we performed both transient activation of SOCE
(using anti-CD3) and permanent activation (using Tg) by using
a similar experimental paradigm to that described in Figure 2D
(Figure 2F and Figure S3E). SOCE activation by CD3-induced
store depletion in siSARAF-treated Jurkat cells resulted in prolonged elevation of [Ca2+]cyto, whereas control cells exhibited
similar initial Ca2+ elevation, which later decreased to levels
that were about 2-fold lower than in SARAF-silenced cells.
SOCE induction by Tg showed no significant differences
between control and siSARAF-treated cells, which is similar to
the experiments with HEK293 cells (Figure S3E). Similar results
were also obtained from HeLa cells in which SARAF expression
was either elevated via SARAF cDNA expression or reduced by
siSARAF (Figures S3C and S3D), thus emphasizing the ubiquitous nature of SARAF action. To confirm that endogenous
SARAF was indeed the molecular component responsible for
SOCE inactivation, we tested the ability of the murine form of
SARAF (mSARAF), whose mRNA is not targeted by the human
siSARAF used in the above experiments (Figure S2F), to rescue
the SARAF phenotype. Jurkat cells were cotransfected with siSARAF with or without mSARAF cDNA, and SOCE was activated
by the reversible SERCA inhibitor BHQ (2,5-Di-t-butyl-1,4-benzohydroquinone). mSARAF rescued SOCE inactivation, indicating that SARAF is the molecular component responsible for
SOCE inactivation (Figure 2G and Figures S3F–S3H). Similarly,
recording of ICRAC from Jurkat cells treated with siSARAF also
showed a reduction in current inactivation; in control cells,
currents remaining after 4 min in the presence of high external
Ca2+ were 15.2% ± 5% (n = 15), whereas in siSARAF-treated
cells, currents were 82% ± 4% (n = 9, p < 0.05) (Figure 2H).
Current densities in control and siSARAF-treated Jurkat cells,
however, were similar, with values of 1.35 ± 0.10 pA/pF and
1.17 ± 0.06 pA/pF, respectively.
Figure 2. SARAF Mediates ER Ca2+ Refilling-Dependent Regulation of SOCE
(A–C) Normalized whole cell current levels at 100 mV of cells expressing YFP-STIM1, mCherry-Orai1 with (black) or without SARAF (red). (A) Pipette solutions
contained (A) 1.2 mM EGTA, SARAF (n = 6), control (n = 9) (B) 10 mM Bapta, SARAF (n = 6), control (n = 8) (B), or (C) 1.2 mM EGTA plus 0.2 mM Tg, SARAF (n = 11),
control (n = 9).
(D) Mean intracellular Ca2+ traces recorded from individual cells (SARAF, n = 36; control, n = 46) after activation of SOCE by ATP + Charbachol (100 mM each) or Tg
(0.2 mM).
(E) Analysis of 5-HT3 ionophoric receptor activity (1 mM serotonin) expressed with (n = 11, black), without SARAF (n = 14, red) or nontransfected (n = 8, green) in
HEK293-T; response was triggered by 1 mM of serotonin in the presence of 2 mM Ca2+.
(F) Average CD3-induced SOCE activity in Jurkat T cells transfected with siSARAF (n = 45, black) or nontarget siControl (n = 52, red).
(G) Averaged BHQ-induced SOCE response in Jurkat T cells transfected with siSARAF (black, n = 41), siSARAF with mouse-siSARAF (blue, n = 15) and nontarget
siControl (red, n = 44).
(H) Averaged whole cell ICRAC currents at 100 mV recorded from Jurkat T cells treated with siSARAF (black, n = 9) or siControl (red, n = 15). Pipette solutions
containing 1.2 mM EGTA.
(I) Representative fluorescent images of GFP-NFAT localization in HEK293-T cells expressing GFP-NFAT (control), with STIM1/Orai1 or with STIM1/Orai1/
SARAF. Arrows indicate cells with nuclear localization and asterisks indicate cytosolic localization of NFAT.
(J) Quantitation of GFP-NFAT nuclear localization in transfected cell (as in I) Tg- (n = 2) and BHQ-induced SOCE(n = 6). (n = number of independent experiments,
60–100 cells analyzed in each experiment). p value: * < 0.05, ** < 0.01.
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. 429
A
SARAF
Epi
Tg
C
SARAF
TIRF
Tg
Tg
SARAF + Orai1
SARAF
Tg
SARAF + STIM1
B
Tg
SARAF
Overlay
Tg
Tg
Overlay
Tg
SARAF + Orai1, siRNA STIM1/2
D
Orai1
STIM1
Tg
Orai1
Tg
Overlay
Tg
Figure 3. SARAF Translocates to ER-PM Regions in a STIM1-Dependent Manner after Store Depletion
(A) Fluorescent images of an HEK293 cell expressing SARAF-GFP, taken under epi illumination or TIRF configurations. Images of the same cell taken after the
application of Tg (0.2 mM) (lower).
(B) TIRF images of a cell expressing SARAF-GFP and STIM1-mCherry before (upper) and after incubation with Tg (lower).
(C) Images of a cell expressing SARAF-GFP and Orai-mCherry, taken under epi illumination, before (upper) and after the application of Tg (lower).
(D) Images of a cell expressing SARAF-GFP, Orai-mCherry treated with siSTIM1 and siSTIM2, taken under epi illumination, before (upper) and after the application
of Tg (lower).
As an independent third measure for the possible action of
SARAF on SOCE activity, we assessed the ability of SARAF to
interfere with the Ca2+-dependent translocation of the nuclear
factor of activated T cells (NFAT-GFP) to the cell nucleus (Tomida
et al., 2003). In unstimulated cells, coexpressing STIM1 and Orai1
with or without SARAF, NFAT-GFP was predominantly found in the
cytoplasm. After ER Ca2+ depletion by BHQ and the reintroduction
of external Ca2+, in cells expressing STIM1 and Orai1, 73% ± 2.8%
of the cells exhibited nuclear NFAT-GFP localization, compared to
only 44% ± 3.1% of the cells expressing STIM1, ORAI1, and
SARAF. In contrast, when ER refilling was abolished by Tg, nearly
all cells exhibited nuclear localization of NFAT and the expression
of SARAF no longer made a significant difference (Figures 2I and
2J). The experiments described above indicate that SARAF-mediated SOCE inactivation primarily depends on cytosolic Ca2+ rise
and is partially affected by the ER Ca2+ refilling state.
SARAF Translocates to ER-PM Regions in a
STIM1-Dependent Manner after Store Depletion
One of the most prominent manifestations of STIM1-Orai1induced SOCE is the puncta formation of a multimeric STIM1Orai1 protein aggregate at specialized ER-PM junctions (Liou
et al., 2005). If SARAF directly inhibits SOCE, it should be found
at these active sites as well. Indeed, in cells coexpressing
430 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
SARAF-GFP together with STIM1-mCherry and Orai1, but not
in cells expressing SARAF-GFP alone, upon depletion of ER
stores with Tg, a significant translocation of SARAF-GFP and
STIM1-mCherry to the plasma membrane was apparent (Figure 3A–3C and Movie S1). SARAF-GFP translocation into
ER-PM regions did not depend on the [Ca2+]cyto because chelation of free cytosolic Ca2+ did not prevent it (Figure S4D). These
results indicate that ER depletion alone is not sufficient for
SARAF translocation and that either Orai1 or STIM1 were essential for this process. To test whether endogenous STIM played
a role in the Tg-induced SARAF translocation, we coexpressed
SARAF-GFP and Orai1-mCherry in siSTIM1 and siSTIM2-treated
cells (Figures S4A–S4C) and monitored their cellular localization.
Silencing STIM1 and STIM2 protein levels inhibited the Tg-induced SARAF-GFP translocation in the presence of Orai1mCherry (Figure 3D). These results indicate that after store
depletion, SARAF ability to translocate into ER-PM regions
depends primarily on the presence of STIM proteins.
SARAF Luminal Domain Regulates Its Activity
and the Cytosolic Domain Is Responsible
for SOCE Modulatory Function
Full-length human SARAF cDNA encodes a 339 amino acid (aa)
protein that contains a cleavable signal peptide (aa 1–30)
600
800 1000 1100 1200 1300 1400 1500 1600
0
C' SARAF
-20
-40
SARAF
-60
-80
control
-120
0
50 100 150 200 250 300
Time (s)
+STIM1
+STIM1
Overlay
+STIM1
E
F
control
SARAF
E148A
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0
0.5
0 Ca
2+
STIM1+Tg
0.45
0.4
0.35
BHQ
ΔR
0.3
0.25
200 400 600 800 1000 1200
Time (s)
STIM1+Tg
**
2+
2mM Ca
α-GFP
α-STIM1
- + - +
- +
G
ΔR
SARAF C’
Ratio (340/380)
Orai1
IB
α-HA
Orai1
100
75
Tg: - +
-100
Time (s)
D
kDa
50
25
*
120
100
80
**
60
40
20
0
wt
400
1OmM Ca
E148A
200
O Ca
20
control
0
0
control
C' SARAF
N' SARAF
Input
IP GFP
C’ SARAF + + - - + + - -YFP
2+
Nuclear translocation of
NFAT (% of control)
Tg
0.4
0.2
2+
wt
ATP
+Crb
C
EGTA+Tg
E148A
0.6
2mM Ca
2+
CRAC
0.8
0 Ca
2mM Ca
2+
Normalized I
Ratio (340/380)
0 Ca
2+
control
2+
(%)
B
A
STIM1+Tg
Figure 4. The ER Luminal Domain of SARAF Regulates Its Activity and the Cytosolic Domain Mediates Its SOCE Inhibitory Activity
(A) Mean intracellular Ca2+ traces of individual HEK293-T cells expressing STIM1+Orai1 alone (n = 46) or with N0 (SARAF(1-191)) (n = 38) or C0 (SARAF(196-339))
(n = 29) termini of SARAF constructs. SOCE was induced by ATP+Charbachol (100 mM each) or Tg (0.2 mM).
(B) Mean normalized whole-cell ICRAC currents (at 100 mV, 1.2 mM EGTA, and 2 mM Tg in the pipette) of cells expressing STIM1 and Orai1 with (black, n = 8) or
without C0 -SARAF-GFP (n = 14). Dashed trace is taken from Figure 2C for comparison.
(C) Western blot analysis of input or immune-precipitated material (IP with anti-GFP) prepared from cells expressing Orai1and STIM1 before and after SOCE
induction by Tg, supplemented with purified YFP-SARAF(196-339) or GIRK1(365-501) (negative control) proteins.
(D) Fluorescent images of HEK293-T cells coexpressing YFP-C0 -SARAF and mCherry-Orai1 alone or with STIM1. (Lower) TIRF images of a cell expressing
mCherry-Orai1/STIM1/YFP-C0 SARAF after SOCE induction.
(E) Mean intracellular Ca2+ traces of individual HEK293-T cells expressing STIM1+Orai1 alone (red), with SARAF (black), or with SARAF(E148A) mutant (blue).
SOCE was induced with BHQ (2 mM) as indicated.
(F) Quantitation of steady-state SOCE activity (DR) as indicated in (E), data obtained from three independent experiments (15–45 cells each).
(G) Normalized percentage of GFP-NFAT nuclear localization induced by BHQ in cells expressing STIM1/Orai1 alone, with SARAF, or SARAF(E148A). Data
obtained from three independent experiments (60–100 cells each). p values: * < 0.05, ** < 0.01, *** < 0.001.
(Zhang and Henzel, 2004) and a single putative transmembrane
domain (aa 173–195). SARAF topology predicts a cytoplasm
facing domain (aa 196–339) and an ER-luminal facing domain
(aa 31–172) (Figure 1A). We cloned an additional splice isoform
of the human SARAF gene, SARAF-S, from HEK293 cells (accession number JQ348892). This splice isoform lacks the second
exon of the full-length SARAF and results in a shorter open
reading frame (ORF), corresponding to aa 173–339 (the ORF is
identical to AK315956). The short isoform exhibits similar cellular
distribution and translocates to ER-PM regions after SOCE activation, similar to the full-length protein (Figure S5A). To elucidate
the potential differential roles played by either the luminal or the
cytosolic domains of SARAF, we created YFP fusions with either
N-terminal (aa 1–191) or C-terminal (aa 196–339) domains and
tested their ability to regulate SOCE. We repeated the same
experimental paradigm as in Figure 2D and measured SOCE
under both store refilling and nonrefilling conditions. Remarkably, truncation of SARAF C-terminal cytosolic domain eliminated SOCE modulation. In contrast, truncation of SARAF
luminal N-terminal domain resulted in strong inhibition of SOCE
that exhibited little dependence on store refilling (Figure 4A).
To further address this issue we measured ICRAC currents in
the absence of store refilling in cells expressing C0 -SARAF using
the same experimental conditions as in Figure 2C. Strikingly,
in the absence of store refilling, in cells expressing C0 -SARAF
isoform currents inactivated to 20.8% ± 2.8% of peak levels,
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. 431
whereas in control cells the extent of current inactivation was
more than 3-fold lower (64.6% ± 5.2%, Figure 4B). Peak current
densities were similar in both control-expressing (31 ± 1.4 pA/
pF, n = 14) and C0 -SARAF- expressing (29 ± 2.3 pA/pF, n = 8)
cells, indicating that SOCE inactivation by the C0 -SARAF
remained dependent on cytosolic Ca2+. To test whether the
C0 -SARAF interacts directly with STIM1 or Orai1, we performed
immuno-pull down analysis by using purified YFP-C0 -SARAF
(aa 196–339) and extracts from HEK293-T cells coexpressing
STIM1 and Orai1-HA. We found that YFP-C0 -SARAF mainly
precipitated STIM1; traces of Orai1 immunopercipitate were
also detected, which may be attributed to its native interaction
with STIM1 (Figure 4C). In contrast, similar immuno-pull down
analysis using N0 -SARAF-YFP were inconclusive and indicated
a very weak, if any, interaction with STIM1 (Figure S5F). From
the above functional assays, it was expected that C0 -SARAF
would colocalize at ER-PM junctions upon store depletion in
a STIM1-dependnet manner. Indeed, cells expressing both
YFP-C0 -SARAF and mCherry-Orai1 exhibited dispersed cytosolic staining that did not overlap with that of mCherry-Orai1
(Figure 4D). In contrast, when STIM1 was also coexpressed,
YFP-C0 -SARAF could be found in reticular structures that moved
close to plasma membrane regions upon store depletion to form
the classical puncta as described in Figure 3B (see also Figure 4D). The reticular like structures of YFP-C0 -SARAF colocalized with STIM1-mCherry, and formation of these structures
did not require Orai1 (Figures S5C–S5E). Taken together, the
above results suggests that the SARAF C-terminal domain
recognizes a domain on STIM1 that enables it to control SOCE
activity, whereas the luminal N-terminal domain is responsible
for the regulation of SARAF inhibitory activity. To further test
this hypothesis, we mutated a conserved glutamate (E148) in
the N-terminal region of SARAF and tested its effect on SARAF
activity. SARAF(E148A) dominantly inactivated SARAF phenotypic activity both on SOCE and on NFAT translocation (Figures
4E–4G), similar to the phenotype seen after its silencing with
siRNA. Moreover, the SARAF(E148A)-YFP mutant exhibited
normal translocation into Orai1 puncta at ER-plasma membrane
junctions upon store depletion (Figure S5B).
SARAF Associates with STIM1 at Rest
To further test the mode of interaction between SARAF and
STIM1, we coexpressed SARAF-GFP and STIM1-mCherry and
used a FRET (Förster resonance energy transfer)-based
approach to test for intermolecular interactions between the
two proteins. Remarkably, resting cells without SOCE stimulation, exhibited a significant FRET signal between SARAF-GFP
and STIM1-mCherry. The specificity of this signal was validated
in several ways. First, to check for random collisions that may
occur between the two constructs, we replaced the donor
SARAF-GFP with a cytosolic expressed GFP. Under these
conditions, no FRET could be detected between GFP and
STIM1-mCherry (Figures 5A and 5B). Next, to test whether true
donor and acceptor interactions were responsible for the FRET
signal within the oligomer, we coexpressed SARAF-GFP with
STIM1-mCherry and added untagged SARAF to compete with
SARAF-GFP on STIM1-mCherry binding. Adding untagged
SARAF resulted in a decrease in the FRET signal between
432 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
SARAF-GFP and STIM1-mCherry (Figure 5C). Likewise, the
reverse competition with untagged STIM1, underlining the
specific interaction between the two molecules. Finally, acceptor
photobleaching was used to verify that indeed, destruction of the
acceptor dequench donor fluorescence (Figures S6A and S6B)
(Riven et al., 2003). Furthermore, independent measurements
of FRET were also obtained using the time-domain approach,
by measuring fluorescent donor life-time changes, fluorescence
lifetime imaging microscopy (FLIM) (Raveh et al., 2010) (Figures
S6C–S6F). These measurements verified our initial observation
and confirmed that STIM1 and SARAF interact at rest. We then
measured FRET changes during SOCE activation after the application of a combination of ATP, Carb, and Tg to cells that
express SARAF-GFP and STIM1-mCherry in the presence or
absence of Orai1. Remarkably, about a 20% FRET change in
the mCherry/GFP ratio was observed only in cells that contained
both STIM1 and Orai1 (Figure 5D). The increase in FRET between
STIM1 and SARAF after store depletion was reversible when
the stores were allowed to refill (Figures S6E and S6F). These
results indicate that SARAF and STIM1 associate under resting
conditions and that SOCE activation promotes molecular rearrangements between STIM1 and SARAF C termini in an Orai1dependent manner.
SARAF Inhibits Spontaneous Activation of STIM1
In order to study the functional consequences of STIM1 and
SARAF interaction under resting conditions, we silenced endogenous SARAF expression and analyzed STIM1 and Orai1 activity
and distribution. In control cells under resting conditions, we
observed a reticular staining pattern for YFP-STIM1 that was
also evident, in part, near plasma membrane regions with a
uniform Orai1-mCherry plasma membrane staining. After ER
depletion, this morphology changed into intense colocalized
punctate fluorescence signals from both YFP-STIM1 and
Orai1-mCherry at plasma membrane regions (Figure 6A). When
cells where endogenous SARAF expression was reduced were
at rest, YFP-STIM1 fluorescence near the plasma membrane
was significantly increased with both STIM1-YFP and Orai1mCherry forming fluorescent puncta (Figure 6A). After the
depletion of Ca2+ from the ER by Tg, a smaller additional recruitment of YFP-STIM1 into puncta was apparent but with an overall
effect that was similar to control cells under similar conditions.
To test the functional consequences of this effect on SOCE,
we measured NFAT translocation and [Ca2+]cyto levels under
similar conditions. Basal [Ca2+]cyto was significantly elevated in
siSARAF-treated cells, a difference that could be reversed by
either rescuing SARAF expression by mSARAF or by omitting
external Ca2+ from the incubation medium (Figure 6C). In agreement with the increased basal SOCE, enhanced nuclear localization of NFAT was also observed in siSARAF-treated cells under
either basal conditions or after reversal depletion of the stores
(Figure 6D).
SARAF Enhances STIM1 Deoligomerization upon Store
Refilling
To gain further insight into the mechanism by which SARAF
modulates SOCE, we raised the question whether STIM1STIM1 oligomeric interactions that underlie SOCE activation
C
D
19500
18000
F
2+
2+
2 mM Ca
ATP/Crb/Tg
1.28
1.2
1.12
1.04
0.96
0.88
- Orai1
+ Orai1
0.8
0.72
0
400
800
Time (s)
1200
3600
/F
Cherry
Ratio (F
0.3
0.25
0.2
**
**
0.15
0.1
ol
1
AF
TIM
+S
ntr
co
AR
GFP
4200
+S
Cherry
4800
0 Ca
)
STIM1-Cherry
+ SARAF-GFP
GFP
B
GFP
STIM1-Cherry (F)
16500
Cherry
STIM1-SARAF
/FGFP (R/R0)
SARAF-GFP+
STIM1-Cherry
Cherry
STIM1-Cherry
+ GFP
SERAF-GFP (F)
A
Figure 5. Resting and Dynamic Interaction of SARAF with STIM1
(A and B) Fluorescent images of HEK293-T cells expressing STIM1-mCherry together with GFP (A) or with SARAF-GFP (B). Images are of the same field in
both GFP and mCherry channels acquired simultaneously. Fluorescent signals in the mCherry channel result specifically from FRET between SARAF-GFP and
STIM1-mCherry (B).
(C) Quantitation of mCherry and GFP fluorescence levels and mCherry/GFP ratios measured from individual cells expressing STIM1-mCherry and SARAF-GFP
(n = 232), alone or with untagged STIM1 (n = 120) or SARAF (n = 86).
(D) Time course of mean mCherry/GFP ratios measured from individual cells expressing STIM1-mCherry and SARAF-GFP alone (red, n = 18) or with Orai1
(black, n = 44) after the application of ATP, Carb (100 mM each) and Tg (0.2 mM). p value: ** < 0.01.
and inactivation, are affected by the presence of SARAF. For this
purpose, we expressed STIM1-GFP, STIM1-mCherry, and Orai1
with or without SARAF and monitored the dynamics of FRET
signals between STIM1-GFP/mCherry during ER Ca2+ depletion
and refilling. In agreement with previous studies (Liou et al.,
2007), we observed an increase in FRET after ER Ca2+ depletion
and a subsequent decrease after ER Ca2+ refilling. In both
control and SARAF overexpressing cells, the rate of FRET
increase during ER Ca2+ depletion and decrease after refilling
were similar. With or without SARAF after ER refilling, the time
constants of FRET decrease were 11.9*103 ± 0.6*103%/s
and 12.9*103 ± 0.7*103%/s, respectively. In contrast, the
extent to which the FRET signal was decreased was different,
possibly reflecting a reduced amount of STIM1 oligomers in
the presence of SARAF (Figure 7A). Because the extent of ER
Ca2+ refilling mirrors the oligomeric state of STIM1 (Shen et al.,
2011), these results are in apparent contradiction with reduced
ICRAC monitored in SARAF-expressing cells. To test this theoretical discrepancy, we determined the levels of ER Ca2+ levels
under the same experimental conditions as in Figure 7A. Results
from this analysis indicated that in agreement with reduced ICRAC
and prior to the onset of the STIM1 deoligomerization process,
[Ca2+]ER was stabilized at lower levels in cells overexpressing
SARAF compared to the levels in control cells (Figure 7B). In
conclusion, these results indicate that SARAF reduces active
STIM1 oligomers upon Ca2+ entry and store refilling even at
lower [Ca2+]ER (Figures 7C and 7D).
DISCUSSION
By screening for genes that shape mitochondrial Ca2+ levels, we
serendipitously identified SARAF as a regulator of store operated
calcium influx. Increasing or decreasing SARAF expression
levels resulted in opposite effects on global levels of intracellular
Ca2+ at resting and activated states, indicating an important role
for this gene product in the regulation of cellular Ca2+ homeostasis. SARAF shares similar topology and overall structurefunction relationships with STIM proteins and consists of a single
pass ER membrane protein with a cytosolic facing segment
responsible for activity, whereas its luminal facing segment
engages in regulatory function. Moreover, like STIM1, SARAF
also contains a serine-proline rich domain followed by a cluster
of basic residues at its C-terminal tail that may aid its interaction
with the plasma membrane phospholipids (Huang et al., 2006;
Korzeniowski et al., 2009; Park et al., 2009) (Figure 1A). Whether
this similarity bears functional significance is currently unclear;
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. 433
siRNA SARAF
STIM1
Orai1
Overlay
Tg
Tg
Tg
Tg
C
siRNA SARAF
siRNA control
0.7
0.6
0.5
*
0.4
0.3
0.2
0.1
0
Basal
Tg
Tg
D
2+
2+
2mM Ca
0 Ca
1.2 si control
1 si SARAF
si SARAF
0.8 +m-SARAF
Ratio (340/380)
PM STIM1
(F STIM1/ F Orai1)
B
siRNA Control
STIM1
Orai1
Overlay
0.6
Tg
0.4
0.2
Tg
0
200 400 600 800 1000
Time (s)
2+
2mM Ca
siRNA SARAF
siRNA control
Nuclear translocation of
NFAT (% of total)
A
100
*
80
60
40
*
20
Basal
BHQ
Figure 6. SARAF inhibits basal activation of SOCE
(A) Representative TIRF images of cells treated with siSARAF or siControl, before and after treatment with Tg (bottom).
(B) Quantitation cell surface puncta formed by STIM1-YFP and mCherry-Orai1 in cells treated with siSARAF (n = 24) or with siControl (n = 25), before and after the
application of Tg (0.2 mM).
(C) Average of intracellular Ca2+ traces of HeLa cells transfected with siSARAF (n = 20), siSARAF+cDNA of mSARAF (n = 38) and siControl (n = 30).
(D) Summary of nuclear localization of GFP-NFAT in siControl or siSARAF-treated cells, before and after BHQ application. Data obtained from three independent
experiments, 60–100 cells each. p value: * < 0.05.
however, our results strongly suggest that STIM is the major
target for SARAF regulation. This conclusion is supported by
several lines of evidence, as both proteins colocalize to the
same organelle where they reside within a short distance from
each other (5–10 nm) and undergo dynamic physical interaction
in response to Ca2+ store depletion. The results presented in this
work, however, do not rule out additional functional interaction
between SARAF and Orai1, and further studies are needed to
address this point.
SARAF is expressed in the ER membrane and regulates cellular
Ca2+ influx mediated by STIM1 activation of Orai1. SARAF regulation over SOCE, however, is largely dependent on Ca2+ entry.
The robust ability and the nature by which SARAF regulates ICRAC
currents is in favor of the idea that SARAF may be part of the
molecular entity responsible for one form of slow Ca2+-dependent
inactivation process that was previously identified in cultured
cells and native tissues (Louzao et al., 1996; Parekh, 1998;
Zweifach and Lewis, 1995). Previous studies that analyzed slow
Ca2+-dependent inactivation of ICRAC yielded somewhat different
findings. In rat basophilic leukemia (RBL) cells inactivation onset
depended on cytosolic Ca2+, but no appreciable involvement of
store refilling was observed. In Jurkat cells, however, a strong
434 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
dependence on cytosloic Ca2+ for CRAC channel inactivation
was also reported, but about 50% of inactivation was traced to
the refilling of stores. Similar to the findings in both studies
mentioned above, we find that the onset of CRAC channel inactivation mediated by SARAF critically depends on cytoslic Ca2+.
Involvement of store refilling, however, is only apparent with the
full-length SARAF but not with its short splice isoform. Moreover,
we find that the depletion of SARAF expression in Jurkat T cells
also diminishes ICRAC inactivation (Figure 2H). It is therefore
conceivable that differential cell-specific alternative splicing of
SARAF may explain some of the discrepancies observed in
previous studies and thus may also underlie a mechanism for
regulating cell Ca2+ homeostasis at the posttranscriptional level.
The precise molecular mechanism that underlies the ability of
SARAF to sense cytosolic Ca2+ elevations and induce ICRAC inactivation is unclear at this stage. Recently, via a proteomics
approach, a modulator for SOCE activity, an EF-hand containing
a cytosolic protein termed CRACR2A, was identified (Srikanth
et al., 2010). CRACR2A binds both Orai1 and STIM1 to aid the
stabilization of the active CRAC complexes, mainly under conditions of low abundance of STIM1 and Orai1. Upon the elevation
of cytosolic Ca2+, CRACR2A is believed to dissociate from the
A
Normalized FRET
2mM Ca
2+
Inactivation
Orai1
PM
1.1
1.08
1.06
ERM
1.04
1.02
1
0.98
0.96
0
Lumen
- SARAF
+ SARAF
80 160 240 320 400 480 560 640
Time (s)
Activation
D
B
2+
Ratio (YFP/CFP)
Activation
C
2+
0 Ca
1.12
ATP/Crb/BHQ
0 Ca
2.15
ATP/Crb/BHQ
Inactivation
Orai1
PM
2+
2mM Ca
2.1
ERM
2.05
2
Lumen
1.95
1.9
1.85
0
- SARAF
+ SARAF
80 160 240 320 400 480 560 640
Time (s)
STIM
SARAF
Ca
2+
Figure 7. SARAF Enhances STIM1 Disaggregation after Store Refilling
(A) Time course of mean GFP/mCherry FRET signal measured from individual cells expressing Orai1, STIM1-mCherry and STIM1-GFP alone (n = 41) or with
SARAF (n = 21). SOCE was induced by ATP, Carb (100 mM each) and BHQ (2 mM).
(B) Dynamics of ER Ca2+ levels measured using D1ER fluorescence ratio of individual cells expressing Orai1, STIM1, D1ER alone (n = 211), or with SARAF
(n = 189), SOCE induction like in (A) and store refilling by the application of 2 mM Ca2+.
(C and D) Cartoons depicting the action of SARAF on SOCE activity. (C) Upon Ca2+ depletion from the ER (depicted as a yellow gradient in the ER lumen), the EF
hand of STIM molecules (residing in the ER membrane, ERM) lose their Ca2+ bound ions (yellow balls). The dissociation of Ca2+ ions promotes the dimerization/
aggregation of STIM molecules and to induce a conformational rearrangement of the C terminal of STIM enabling their interaction with Orai channels at the
plasma membrane (PM). STIM-Orai interaction promotes channel activation and entry of Ca2+ ions into the cell. After refilling of the ER with the entering Ca2+ ions
by SERCA (not shown), Ca2+ levels in the ER lumen rise to reverse the above process. In the absence of SARAF (green) the reverse process is less efficient
to cause elevation in intracellular Ca2+ concentration above normal levels. (D), SARAF association with STIM proteins affect the above process by facilitating
the disaggregation of STIM molecules to efficiently turn off Orai when the ER lumen is filled with the appropriate Ca2+ levels, and thus preventing the overload of
the cell with excessive Ca2+ ions. Events read left to right.
complex to promote the destabilization of the SOCE complex.
Like SARAF it also responds to local elevation of Ca2+ near the
plasma membrane. It remains to be seen if CRACR2A has a corresponding effect on ICRAC. Whether the two proteins are part of
the same protein complex that regulates CRAC channels is an
appealing possibility that awaits future studies.
The ER luminal facing segment of SARAF regulates the inhibitory action mediated by the cytosolic facing segment, which has
the capability to interact with the C-terminal cytosolic domain of
STIM1. The apparent action of SARAF in facilitating ER Ca2+ refilling-dependent inactivation is, however, likely a consequence of
its ability to indirectly respond to changes in [Ca2+]ER and
promote an efficient disaggregation of STIM molecules (Figure 7).
The regulatory role of the SARAF ER luminal facing domain is
underlined by the enhanced ability of the short isoform of SARAF
to inhibit ICRAC independently of store refilling. The fact that
a mutation in this region has a dominant negative effect on
SOCE activity also suggests that SARAF may be functional as
a dimer or in a higher-order complex. The ER luminal facing
segment of SARAF, however, does not show clear similarity
with any known Ca2+ binding motifs. In addition, ICRAC measurements in the absence or presence of store refilling yield similar
2-fold differences in both control and SARAF-expressing cells
(Figures 2A and S2C), and translocation of SARAF to ER-PM
regions depends on that of STIM1 but not on store depletion
per se (Figure 3). It is therefore less likely that SARAF directly
binds Ca2+ ions and thus raises the possibility that SARAF may
respond to changes in [Ca2+]ER indirectly through interaction
with STIM or other ER Ca2+-sensitive proteins or is regulated
by other physiological mechanisms (Enyedi et al., 2010; Hawkins
et al., 2010). The functional interaction of SARAF with STIM1 may
point toward the former possibility, were coupling between
STIM1 Ca2+ occupancy and the SARAF inhibitory mode, found
to occur. Such a scenario is in agreement with spontaneous activation of STIM1 proteins without ER Ca2+ depletion, observed
after knockdown of SARAF expression or with a more efficient
disaggregation and stabilization of inactive STIM1 proteins at
lower ER Ca2+ levels, after an increase in SARAF expression.
The direct mode by which SARAF affects SOCE inactivation after
ER Ca2+ refilling is currently unresolved and awaits future studies.
Although no direct evidence provided to date shows that
mutations in SARAF are linked to any disease state in humans,
studies employing various differential expression or transcriptome-profiling strategies have recently identified SARAF as a
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. 435
biomarker linked to prostate cancer (Romanuik et al., 2009a),
Alzheimer’s disease (Twine et al., 2011), and dilated cardiomyopathy (DCM) (Camargo and Azuaje, 2008). Notably, most of
these disease states are accompanied by abnormal intracellular
Ca2+ handling. It is also interesting to note that both STIM1 and
SARAF transcripts have been shown to be positively modulated
by androgens, implying common regulatory mechanisms of
these genes expression levels (Berry et al., 2011; Romanuik
et al., 2009b).
In summary we have identified the gene product of TMEM66,
SARAF, as a regulator of store operated calcium entry. SARAF
promotes slow ICRAC inactivation after Ca2+ entry and refilling
of the endoplasmic reticulum. This physiological role places
SARAF as a dominant player in a protective mechanism designed to prevent rapid overfilling of cells with Ca2+ ions.
ACCESSION NUMBERS
The GeneBank accession number for SARAF is JQ348891, and for SARAF
short isoform it is JQ348892.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Extended Experimental Procedures,
six figures, and one movie and can be found with this article online at
doi:10.1016/j.cell.2012.01.055.
ACKNOWLEDGMENTS
The authors would like to thank Elisha Shalgi for technical help and
Dr. Dan Minor for his helpful comments and critical reading of the manuscript.
The work was supported in part by the Clore postdoctoral fellowship (RP), the
Josef Cohn Center for Biomembrane Research, The Israeli Science Foundation (ISF grant 207/09) and the Minerva Foundation (Munich) all to ER.
EXPERIMENTAL PROCEDURES
Fluorescent Measurements of Intracellular Ions
Typically, 6–8 hr after plasmid transfection and 17–24 hr before starting the
experiments, cells were plated onto 24 mm cover glass coated with L-polylysine. On the day of the experiments, the cover glass was mounted on an
imaging chamber and washed with Ringer solution. Fura-2 loading of cells
was performed as previously described (Palty et al., 2010). Cytosolic Ca2+
levels were recorded from fura-2 loaded cells, excited at wavelengths of
340/20 and 380/20 nm and imaged with 510/80 nm filters.
Mitochondrial Ca2+ levels were monitored in cells expressing the mitochondrial-targeted ratiometric-pericam (RP-mt) protein via excitation wavelengths
of 436/20 nm and 480/20 nm and emission collected with a 540/80 nm filter.
Endoplasmic reticulum Ca2+ levels were recorded in cells expressing D1ER
via a dual view device with a CFP/YFP filter set.
All experiments were conducted with Ringer’s solutions containing 130 mM
NaCl, 20 mM HEPES,15 mM Glucose, 5 mM KCl, and 0.8 mM MgCl2 with the
pH adjusted to 7.4. Ringer’s solution was supplemented with 2 mM CaCl2 or
1 mM EGTA (Ca2+ free) as indicated. For all single cell imaging experiments,
traces of averaged responses, recorded from 10 to 50 cells in each experiment, were plotted with KaleidaGraph. All experiments were repeated 2 to
8 times. Statistical significance was determined with t test analysis; p < 0.05
was considered significant. All data are shown as average ± SEM.
NFAT Translocation Assay
Monitoring of nuclear translocation of NFAT-GFP was performed 30–48 hr
posttransfection in HEK293-T or HeLa cells. At the start of each experiment,
cells were washed twice with Ca2+-free Ringer’s solution, and 2 mM BHQ or
0.2 mM Tg were added for 5 min. Cells were then washed with Ca2+-free
Ringer’s solution for additional 5 min and incubated for 30 min in a 2 mM
Ca2+-containing Ringer’s solution. Localization of NFAT-GFP in cells was
examined by fluorescent microscopy before and after SOCE induction.
Electrophysiological Recordings
Recorded HEK293-T or HEK293 cells were transfected 24–36 hr prior to
electrophysiology experiments. Each transfection contained STIM1 and
Orai1 with or without SARAF (1 mg each). Cells were transferred to the
measurement chamber more than 6 hr before measurements for HEK293 cells
and 10–20 min for Jurkat cells. Membrane currents were recorded under
voltage-clamp conditions with whole-cell variation of the patch-clamp configuration by using Axopatch 200B (Axon Instruments). Patch pipettes were
fabricated from borosilicate glass capillaries (2–5 MU). Signals were analog
filtered with a 1 kHz low-pass Bessel filter. Data acquisition and analysis
were done with pCLAMP 9 software (Axon Instruments). All data were leakcorrected with the current elicited in Ca2+-free Ringer’s solution at the beginning of each experiment. More detailed information and solutions used are
in Extended Experimental Procedures.
436 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
Received: July 15, 2011
Revised: November 24, 2011
Accepted: January 26, 2012
Published online: March 29, 2012
REFERENCES
Baba, Y., Hayashi, K., Fujii, Y., Mizushima, A., Watarai, H., Wakamori, M.,
Numaga, T., Mori, Y., Iino, M., Hikida, M., and Kurosaki, T. (2006). Coupling
of STIM1 to store-operated Ca2+ entry through its constitutive and inducible
movement in the endoplasmic reticulum. Proc. Natl. Acad. Sci. USA 103,
16704–16709.
Barr, V.A., Bernot, K.M., Srikanth, S., Gwack, Y., Balagopalan, L., Regan, C.K.,
Helman, D.J., Sommers, C.L., Oh-Hora, M., Rao, A., and Samelson, L.E.
(2008). Dynamic movement of the calcium sensor STIM1 and the calcium
channel Orai1 in activated T-cells: puncta and distal caps. Mol. Biol. Cell 19,
2802–2817.
Berry, P.A., Birnie, R., Droop, A.P., Maitland, N.J., and Collins, A.T. (2011). The
calcium sensor STIM1 is regulated by androgens in prostate stromal cells.
Prostate 71, 1646–1655.
Bird, G.S., Hwang, S.Y., Smyth, J.T., Fukushima, M., Boyles, R.R., and Putney,
J.W., Jr. (2009). STIM1 is a calcium sensor specialized for digital signaling.
Curr. Biol. 19, 1724–1729.
Brandman, O., Liou, J., Park, W.S., and Meyer, T. (2007). STIM2 is a feedback
regulator that stabilizes basal cytosolic and endoplasmic reticulum Ca2+
levels. Cell 131, 1327–1339.
Cahalan, M.D. (2009). STIMulating store-operated Ca(2+) entry. Nat. Cell Biol.
11, 669–677.
Camargo, A., and Azuaje, F. (2008). Identification of dilated cardiomyopathy
signature genes through gene expression and network data integration. Genomics 92, 404–413.
Derler, I., Fahrner, M., Muik, M., Lackner, B., Schindl, R., Groschner, K., and
Romanin, C. (2009). A Ca2(+)release-activated Ca2(+) (CRAC) modulatory
domain (CMD) within STIM1 mediates fast Ca2(+)-dependent inactivation of
ORAI1 channels. J. Biol. Chem. 284, 24933–24938.
Enyedi, B., Várnai, P., and Geiszt, M. (2010). Redox state of the endoplasmic
reticulum is controlled by Ero1L-alpha and intraluminal calcium. Antioxid.
Redox Signal. 13, 721–729.
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 abrogating CRAC channel function. Nature
441, 179–185.
Frischauf, I., Muik, M., Derler, I., Bergsmann, J., Fahrner, M., Schindl, R.,
Groschner, K., and Romanin, C. (2009). Molecular determinants of the
coupling between STIM1 and Orai channels: differential activation of Orai1-3
channels by a STIM1 coiled-coil mutant. J. Biol. Chem. 284, 21696–21706.
Hardie, R.C., and Minke, B. (1993). Novel Ca2+ channels underlying transduction in Drosophila photoreceptors: implications for phosphoinositide-mediated Ca2+ mobilization. Trends Neurosci. 16, 371–376.
Hawkins, B.J., Irrinki, K.M., Mallilankaraman, K., Lien, Y.C., Wang, Y., Bhanumathy, C.D., Subbiah, R., Ritchie, M.F., Soboloff, J., Baba, Y., et al. (2010).
S-glutathionylation activates STIM1 and alters mitochondrial homeostasis.
J. Cell Biol. 190, 391–405.
Hogan, P.G., Lewis, R.S., and Rao, A. (2010). Molecular basis of calcium
signaling in lymphocytes: STIM and ORAI. Annu. Rev. Immunol. 28, 491–533.
Hoth, M., and Penner, R. (1993). Calcium release-activated calcium current in
rat mast cells. J. Physiol. 465, 359–386.
Huang, G.N., Zeng, W., Kim, J.Y., Yuan, J.P., Han, L., Muallem, S., and Worley,
P.F. (2006). STIM1 carboxyl-terminus activates native SOC, I(crac) and TRPC1
channels. Nat. Cell Biol. 8, 1003–1010.
Kawasaki, T., Lange, I., and Feske, S. (2009). A minimal regulatory domain
in the C terminus of STIM1 binds to and activates ORAI1 CRAC channels.
Biochem. Biophys. Res. Commun. 385, 49–54.
Kiselyov, K., Xu, X., Mozhayeva, G., Kuo, T., Pessah, I., Mignery, G., Zhu, X.,
Birnbaumer, L., and Muallem, S. (1998). Functional interaction between
InsP3 receptors and store-operated Htrp3 channels. Nature 396, 478–482.
Korzeniowski, M.K., Popovic, M.A., Szentpetery, Z., Varnai, P., Stojilkovic,
S.S., and Balla, T. (2009). Dependence of STIM1/Orai1-mediated calcium
entry on plasma membrane phosphoinositides. J. Biol. Chem. 284, 21027–
21035.
Lee, K.P., Yuan, J.P., Zeng, W., So, I., Worley, P.F., and Muallem, S. (2009).
Molecular determinants of fast Ca2+-dependent inactivation and gating of
the Orai channels. Proc. Natl. Acad. Sci. USA 106, 14687–14692.
Lewis, R.S. (2007). The molecular choreography of a store-operated calcium
channel. Nature 446, 284–287.
Liou, J., Kim, M.L., Heo, W.D., Jones, J.T., Myers, J.W., Ferrell, J.E., Jr., and
Meyer, T. (2005). STIM is a Ca2+ sensor essential for Ca2+-store-depletiontriggered Ca2+ influx. Curr. Biol. 15, 1235–1241.
Liou, J., Fivaz, M., Inoue, T., and Meyer, T. (2007). Live-cell imaging reveals
sequential oligomerization and local plasma membrane targeting of stromal
interaction molecule 1 after Ca2+ store depletion. Proc. Natl. Acad. Sci.
USA 104, 9301–9306.
Palty, R., Silverman, W.F., Hershfinkel, M., Caporale, T., Sensi, S.L., Parnis, J.,
Nolte, C., Fishman, D., Shoshan-Barmatz, V., Herrmann, S., et al. (2010). NCLX
is an essential component of mitochondrial Na+/Ca2+ exchange. Proc. Natl.
Acad. Sci. USA 107, 436–441.
Pani, B., Ong, H.L., Brazer, S.C., Liu, X., Rauser, K., Singh, B.B., and Ambudkar, I.S. (2009). Activation of TRPC1 by STIM1 in ER-PM microdomains
involves release of the channel from its scaffold caveolin-1. Proc. Natl.
Acad. Sci. USA 106, 20087–20092.
Parekh, A.B. (1998). Slow feedback inhibition of calcium release-activated
calcium current by calcium entry. J. Biol. Chem. 273, 14925–14932.
Parekh, A.B., and Putney, J.W., Jr. (2005). Store-operated calcium channels.
Physiol. Rev. 85, 757–810.
Park, C.Y., Hoover, P.J., Mullins, F.M., Bachhawat, P., Covington, E.D.,
Raunser, S., Walz, T., Garcia, K.C., Dolmetsch, R.E., and Lewis, R.S. (2009).
STIM1 clusters and activates CRAC channels via direct binding of a cytosolic
domain to Orai1. Cell 136, 876–890.
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.
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.
Putney, J.W., Jr. (1986). A model for receptor-regulated calcium entry. Cell
Calcium 7, 1–12.
Raveh, A., Cooper, A., Guy-David, L., and Reuveny, E. (2010). Nonenzymatic
rapid control of GIRK channel function by a G protein-coupled receptor kinase.
Cell 143, 750–760.
Rı́os, E. (2010). The cell boundary theorem: a simple law of the control of cytosolic calcium concentration. J. Physiol. Sci. 60, 81–84.
Riven, I., Kalmanzon, E., Segev, L., and Reuveny, E. (2003). Conformational rearrangements associated with the gating of the G protein-coupled potassium
channel revealed by FRET microscopy. Neuron 38, 225–235.
Romanuik, T.L., Ueda, T., Le, N., Haile, S., Yong, T.M., Thomson, T., Vessella,
R.L., and Sadar, M.D. (2009a). Novel biomarkers for prostate cancer including
noncoding transcripts. Am. J. Pathol. 175, 2264–2276.
Romanuik, T.L., Wang, G., Holt, R.A., Jones, S.J., Marra, M.A., and Sadar,
M.D. (2009b). Identification of novel androgen-responsive genes by
sequencing of LongSAGE libraries. BMC Genomics 10, 476.
Litjens, T., Harland, M.L., Roberts, M.L., Barritt, G.J., and Rychkov, G.Y.
(2004). Fast Ca(2+)-dependent inactivation of the store-operated Ca2+ current
(ISOC) in liver cells: a role for calmodulin. J. Physiol. 558, 85–97.
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 component of store-operated Ca2+ channel function.
J. Cell Biol. 169, 435–445.
Liu, X., Cheng, K.T., Bandyopadhyay, B.C., Pani, B., Dietrich, A., Paria, B.C.,
Swaim, W.D., Beech, D., Yildrim, E., Singh, B.B., et al. (2007). Attenuation of
store-operated Ca2+ current impairs salivary gland fluid secretion in
TRPC1(-/-) mice. Proc. Natl. Acad. Sci. USA 104, 17542–17547.
Shen, W.W., Frieden, M., and Demaurex, N. (2011). Local cytosolic Ca2+
elevations are required for stromal interaction molecule 1 (STIM1) de-oligomerization and termination of store-operated Ca2+ entry. J. Biol. Chem.
286, 36448–36459.
Louzao, M.C., Ribeiro, C.M., Bird, G.S., and Putney, J.W., Jr. (1996). Cell typespecific modes of feedback regulation of capacitative calcium entry. J. Biol.
Chem. 271, 14807–14813.
Soboloff, J., Spassova, M.A., Tang, X.D., Hewavitharana, T., Xu, W., and Gill,
D.L. (2006). Orai1 and STIM reconstitute store-operated calcium channel function. J. Biol. Chem. 281, 20661–20665.
Muik, M., Frischauf, I., Derler, I., Fahrner, M., Bergsmann, J., Eder, P., Schindl,
R., Hesch, C., Polzinger, B., Fritsch, R., et al. (2008). Dynamic coupling of the
putative coiled-coil domain of ORAI1 with STIM1 mediates ORAI1 channel
activation. J. Biol. Chem. 283, 8014–8022.
Srikanth, S., Jung, H.J., Kim, K.D., Souda, P., Whitelegge, J., and Gwack, Y.
(2010). A novel EF-hand protein, CRACR2A, is a cytosolic Ca2+ sensor that
stabilizes CRAC channels in T cells. Nat. Cell Biol. 12, 436–446.
Mullins, F.M., Park, C.Y., Dolmetsch, R.E., and Lewis, R.S. (2009). STIM1 and
calmodulin interact with Orai1 to induce Ca2+-dependent inactivation of
CRAC channels. Proc. Natl. Acad. Sci. USA 106, 15495–15500.
Navarro-Borelly, L., Somasundaram, A., Yamashita, M., Ren, D., Miller, R.J.,
and Prakriya, M. (2008). STIM1-Orai1 interactions and Orai1 conformational
changes revealed by live-cell FRET microscopy. J. Physiol. 586, 5383–5401.
Palmer, A.E., Jin, C., Reed, J.C., and Tsien, R.Y. (2004). Bcl-2-mediated alterations in endoplasmic reticulum Ca2+ analyzed with an improved genetically
encoded fluorescent sensor. Proc. Natl. Acad. Sci. USA 101, 17404–17409.
Stathopulos, P.B., Zheng, L., Li, G.Y., Plevin, M.J., and Ikura, M. (2008).
Structural and mechanistic insights into STIM1-mediated initiation of storeoperated calcium entry. Cell 135, 110–122.
Stathopulos, P.B., Zheng, L., and Ikura, M. (2009). Stromal interaction molecule (STIM) 1 and STIM2 calcium sensing regions exhibit distinct unfolding
and oligomerization kinetics. J. Biol. Chem. 284, 728–732.
Su, A.I., Wiltshire, T., Batalov, S., Lapp, H., Ching, K.A., Block, D., Zhang, J.,
Soden, R., Hayakawa, M., Kreiman, G., et al. (2004). A gene atlas of the mouse
and human protein-encoding transcriptomes. Proc. Natl. Acad. Sci. USA 101,
6062–6067.
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. 437
Tomida, T., Hirose, K., Takizawa, A., Shibasaki, F., and Iino, M. (2003). NFAT
functions as a working memory of Ca2+ signals in decoding Ca2+ oscillation.
EMBO J. 22, 3825–3832.
Twine, N.A., Janitz, K., Wilkins, M.R., and Janitz, M. (2011). Whole transcriptome sequencing reveals gene expression and splicing differences in brain
regions affected by Alzheimer’s disease. PLoS ONE 6, e16266.
Vig, M., Peinelt, C., Beck, A., Koomoa, D.L., Rabah, D., Koblan-Huberson, M.,
Kraft, S., Turner, H., Fleig, A., Penner, R., and Kinet, J.P. (2006). CRACM1 is
a plasma membrane protein essential for store-operated Ca2+ entry. Science
312, 1220–1223.
Wu, M.M., Buchanan, J., Luik, R.M., and Lewis, R.S. (2006). Ca2+ store depletion causes STIM1 to accumulate in ER regions closely associated with the
plasma membrane. J. Cell Biol. 174, 803–813.
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.
Yuan, J.P., Zeng, W., Dorwart, M.R., Choi, Y.J., Worley, P.F., and Muallem, S.
(2009). SOAR and the polybasic STIM1 domains gate and regulate Orai
channels. Nat. Cell Biol. 11, 337–343.
438 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
Zhang, Z., and Henzel, W.J. (2004). Signal peptide prediction based on
analysis of experimentally verified cleavage sites. Protein Sci. 13, 2819–2824.
Zhang, S.L., Yu, Y., Roos, J., Kozak, J.A., Deerinck, T.J., Ellisman, M.H.,
Stauderman, K.A., and Cahalan, M.D. (2005). STIM1 is a Ca2+ sensor that
activates CRAC channels and migrates from the Ca2+ store to the plasma
membrane. Nature 437, 902–905.
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.
Zhou, Y., Meraner, P., Kwon, H.T., Machnes, D., Oh-hora, M., Zimmer, J.,
Huang, Y., Stura, A., Rao, A., and Hogan, P.G. (2010). STIM1 gates the
store-operated calcium channel ORAI1 in vitro. Nat. Struct. Mol. Biol. 17,
112–116.
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.
Supplemental Information
EXTENDED EXPERIMENTAL PROCEDURES
Creating HEK293-RPmt Cells
Recombinant lentiviruses were produced by transient transfection in HEK293-T as previously described (Tiscornia et al., 2006). Viral
particles were used to infect HEK293 cells and single colonies were selected for 10 days after a preclearing step using flow-cytometry
cell sorting to collect (see below) cells expressing RP-mt and to dispose of uninfected cells.
Identification and Isolation of SARAF cDNA
We performed a functional expression screen in attempt to isolate cDNA candidates that regulate mitochondrial calcium homeostasis. For this purpose, a cDNA library derived from adult human heart (Invitrogen) was transiently expressed in human embryonic
kidney-derived cells stably expressing the mitochondrial calcium indicator ratiomeric pericam (Nagai et al., 2001) (HEK293-RPmt),
and cells that exhibited either low or high basal mitochondrial calcium levels were pooled using fluorescence activated cell sorting.
cDNAs from each sorted sub-population of cells were extracted and iteratively re-pooled. In this way an individual 1.8 kilobase
cDNA clone with a 1020 bp open reading frame coding for SARAF was identified, in the cDNA subpopulation enriched by a phenotypically low basal Ca2+ levels.
Flow Cytometry Cell Sorting and Analysis of Mitochondrial Ca2+
To evaluate mitochondrial Ca2+ transfected HEK293-RPmt cells were collected 48 hr after transfection and analyzed by flow
cytometry using a FACSAriaII instrument. Criteria used for sorting cDNA library transfected cells was determined as the values
encompassing the 15% of cells exhibiting either lowest or highest basal Ca2+ mitochondrial levels in the first sorting procedure.
Data were processed with FlowJo software and presented as the ratio between fluorescence emission intensities measured with
480nm and 430nm laser excitation.
Cell Culture and Transfection Procedures
HEK293-T, HEK293 and HeLa cells were cultured in DMEM and Jurkat cells were cultured in RPMI as previously described (Palty
et al., 2010). For cells cultured under external Low Ca2+ conditions Minimum Essential Medium Eagle or DMEM supplemented
with EGTA (2 mM) were used. Plasmid transfection of HEK293-T, HEK293 or HeLa cells was performed using Ca2PO4 as described
previously (Palty et al., 2010). siRNA transfection (100 nM in total) of HeLa, HEK293-T or Jurkat cells were performed using Dharmafect 1 or 4 siRNA transfection reagent or Dharmafect Duo for siRNA and plasmid transfection according to the manufacturer’s
protocol (Dharmacon). Fluorescent ion measurements and cell harvesting were conducted 24-48 hr after plasmid transfection or
96 hr after siRNA transfection. Twenty-four hrs after Jurkat cells transfection 10 mM LaCl3 were added to the media.
Plasmid and Small Interfering RNAs
The full-length human SARAF cDNA was isolated from an adult human heart derived cDNA expression library (Invitrogen). The fulllength mouse SARAF corresponds to clone no. 1810045K07 from the RIKEN Fantom library. The 1091 bp open reading frame of this
clone was inserted into pCDNA3.1 using EcoRI/HindIII.
The SARAF short splice isofrm was cloned from HEK293 cells. For this purpose total RNA was extracted from HEK293 cells using
RNeasy Mini kit (QIAGEN), followed by one-step reverse transcriptase reaction (QIAGEN) using 500 ng of total RNA and the following
primers:
(forward primer) AAAAAACCCGGGATGGCCGCAGCCTGCGGGCC
(reverse primer) AAAAAACCGGTGGTCGTCTCCTGGTACCACCAT
Two transcripts of 845 and 1024 bp were amplified. Both transcripts were ligated into pGEM-T vector and sequenced. The short
SARAF transcript was excised with SpeI and AgeI from the pGEM-T vector and inserted into EGFP-N1 using NheI/AgeI.
SARAF-EGFP/EYFP were created by inserting the full-length coding region of SARAF cDNA into EGFPN1 or EYFP-N1 vector using
HindIII/AgeI. SARAF-mKok was created by inserting the mKok cDNA in frame into a BsrGI site in SARAF cDNA. The SARAF(196-339)
was inserted into EYFP-C1 using BglII/HindIII. For expression in bacteria the YFP-SARAF(196-339) construct was inserted into
pET-TEV(6XHis) vector using BamHI/HindIII. SARAF(1-191)-YFP was created by inserting SARAF(1-191) into EYFP-N1 vector using
HindIII/AgeI. Single point mutations in SARAF were performed by a single or double PCR-based method and mutated SARAF cDNAs
were inserted into EYFP-N1 or into the same vector backbone that lacks the coding region for the fluorescent yellow protein. STIM1
D76A-YFP plasmid was obtained from Prof. Murali Prakriya lab at Northwestern University (Chicago, USA). STIM1-myc, Orai1-HA,
STIM1-YFP, STIM2-YFP and Orai1-mCherry were obtained from Prof. Shmuel Muallem lab at the National Institutes of Health
(NIH, USA). STIM1-mCherry and STIM1-GFP were created by inserting the STIM1 cDNA into mCherry-N1 using EcoRI/BamHI sties.
The Lenti virus plasmid construct of the mitochondrial pericam was obtained from Dr. Daniel Bano from the Deutsches Zentrum für
Neurodegenerative Erkrankungen (Bonn, Germany). SEC61-GFP was obtained from Prof. Gerardo Lederkremer’s lab at Tel-Aviv
University (Tel-Aviv, Israel). NFAT-GFP was purchased from Addgene (originally provided by Dr. A. Rao).
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. S1
siRNA Sequence
All siRNAs were purchased from Sigma-Aldrich (MISSION Predesigned siRNA) and sequences for small interfering RNAs used in
this work are:
STIM1:
1. (SASI_Hs01_00107804) GAGATTGTGTCTCCCTTGT
2. (SASI_Hs01_00107805) CTCACTTCATCATGACTGA
3. (SASI_Hs01_00107803) GGGATTTGACCCATTCCGA
STIM2:
1. (SASI_Hs02_00354130) CTTCATTTATGATCTCCCA
2. (SASI_Hs01_00146914) CTTCAGTGGTTGATAGAGT
3. (SASI_Hs02_00354131) GATCTGTGGCTTTCAGATA
SARAF:
1. (SASI_Hs01_00070453) CTCTTACCCTCCACTATGA
2. (SASI_Hs02_00348914) CCTTTGTAGTCTATAAGCT
3. CGGACTTAGATATTGCATACA
Control siRNA used were either universal non target (SIC-001) or Cy3 tagged siRNA duplex targeting GFP: [cy3]
GCAAGCTGACCCTGAAGTTCAT.
Semiquantitative RT-PCR
Total RNA was isolated from Jurkat cells transfected with control or SARAF siRNAs using RNeasy Mini kit (QIAGEN). The cDNAs were
synthesized with 0.5 mg of total RNA using reverse transcriptase and random oligos primers (High Capacity cDNA Reverse Transcription Kit, Applied Biosystems). Transcripts of Actin and SARAF were amplified from equal amounts of cDNA templates using 15, 25,
and 32 cycles. Mouse SARAF transcripts were amplified using 40 PCR cycles. The amplification primers used were:
d
d
d
d
d
d
(Human-Actin -Forward primer) GGCTACAGCTTCACCACCAC
(Human-Actin -Reverse primer) TACTCCTGCTTGCTGATCCTC
(Human-SARAF - Forward primer) TGGAACGACCCTGACAGAATG
(Human-SARAF - Reverse primer) GAGGAGTACTGCCCGTCAA
(Mouse-SARAF - Forward primer) ATGGTGGGTTCCTGTGGTCGC
(Mouse -SARAF - Reverse primer) CTAGGAATCTGAGGAGTACAGCTT
Immunoblot and Immunoprecipitation
HEK293-T, HeLa or Jurkat cells were washed once with ice-cold PBS solution and placed into RIPA buffer on ice for 15 min. Samples
were then homogenized by sonication and centrifugated for 10 min to dispose cell debris and nuclei. Total protein was quantified by
the Bradford dye-binding procedure (Bio-Rad). Proteins were resolved by SDS-PAGE and transferred onto nitrocellulose
membranes. Immunoblot analysis was carried out as previously described (Palty et al., 2010), using antibodies targeted against
STIM1 (Alomone Labs, 1:1000), Orai1 (Alomone Labs, 1:500), HA tag (Pierce, 1:500), SARAF (Abcam, 1:1000), GFP (MBL, 1:500)
and Na/K ATPase (Gift from Prof. Steve Karlish, Weizmann Inst., 1:3000). The SARAF(196-339)-YFP-6Xhis and GIRK1(365-501)6Xhis recombinant proteins were expressed in BL21 cells, after sonication and clearing, the supernatant was incubated with
Ni-NTA agarose resin and eluted in buffer containing 250 mM imidazole. Immunoprecipitation was performed using agarose
conjugated Anti-GFP beads (MBL) according to manufacturer protocol.
Calibration of Fura-2 Response
The Fura-2 ratio was calibrated by a similar method described in Brandman et al. (Brandman et al., 2007) using:
2+ = Kd Q ðR Rmin Þ=ðRmax RÞ
Ca
Where: Kd = 0.114 mM; Rmin = 0.05; Rmax = 4.89 and Q = 9.55.
Microscopy
Epifluorescent and total internal reflection fluorescence microscopy (TIRFM) were performed using a fluorescence imaging workstation (Intelligent Imaging Innovation) consisting of an inverted Axio Observer Z1 (Zeiss) microscope combined with a TIRFM module
and equipped with a cooled CCD cascade QuantEM 512 SC camera (Photometrics), z-step motor, 6 channel filter wheel and
S2 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
a 300W fast excitation Xenon lamp, all controlled by SlideBook 4.2 software. The TIRF laser modules consisted of a 3-line AOTF solid
state laser launch system (473nm 40 mW, 561nm 50mW and 661nm 45mW). Fluorescence was collected by using various objectives
(20X for fura-2, 40X for fura-2, D1ER and mCherry/GFP FRET measurements and 100X N.A. 1.45 for TIRF and high-resolution
imaging) and filter sets consisting of 510/80 nm for fura-2 emission, 540/80 nm for RP-mt emission and EGFP/DsRed (51019, chroma)
for GFP and mCherry emissions.
For FRET measurements, fluorescent signals were collected through the objective and split to CFP (465/30) and YFP (535/30)
channels or to GFP (525/50) and mCherry (650/70) channels through a Dual-View device. Images were processed using SlideBook
software and exported as Microsoft Excel files. To obtain high-resolution images of cells, serial 2-D images were recorded at 150-nm
intervals and Z-stack of images was then deconvoluted using a nearest neighbors method.
Fluorescence Lifetime Measurements
For fluorescence lifetime measurements (FLIM), 470 nm ps diode laser (FWHM < 90 ps) was used, driven by a 40 MHz pulse
controller, PDL 800-B. Single photons were collected using PMA-165P photon counter and processed using TimeHarp
200 PC-board. Data were acquired under TIRF and analyzed using SymPhoTime software (PicoQuant, Germany) as described
previously (Raveh et al., 2010).
Solutions Composition for Electrophysiological Recordings
Ringer’s solution containing 2 mM Ca2+ was replaced to Ca-free Ringer’s solution immediately after break-in to the whole-cell
configuration, to allow passive store depletion for 3 min (for HEK293) or 5 min (for Jurkats). Voltage protocols, 20-ms step
to 100 mV followed by a 50-ms ramp from 100 to +100 mV, were delivered every 2 s from the holding potential of 0 mV. Analysis
was performed on steady-state current signals at 100 mV. Current densities were calculated by normalizing currents to cell
capacitance. EGTA internal solution contained 150 mM Cs aspartate, 8 mM MgCl2, 1.2 mM EGTA, 10 mM HEPES and 2 mM Mg-ATP
(pH 7.2 with CsOH). BAPTA internal solution contained 150 mM Cs aspartate, 8 mM MgCl2, 12 mM BAPTA, 10 mM HEPES and 2 mM
Mg-ATP (pH 7.2 with CsOH). External solution contained 145 mM NaCl, 10 mM CaCl2, 2 mM MgCl2, 2.8 mM KCl, 10 mM HEPES
and 10 mM Glucose (pH 7.4 with NaOH). 10 mM CaCl2 (20 mM for Jurkat cells recordings) or 10 mM MgCl2 (20 mM for Jurkat cells
recordings) were added to the external solution for high-Ca2+ or Ca-free solution, respectively.
SUPPLEMENTAL REFERENCES
Brandman, O., Liou, J., Park, W.S., and Meyer, T. (2007). STIM2 is a feedback regulator that stabilizes basal cytosolic and endoplasmic reticulum Ca2+ levels.
Cell 131, 1327–1339.
Nagai, T., Sawano, A., Park, E.S., and Miyawaki, A. (2001). Circularly permuted green fluorescent proteins engineered to sense Ca2+. Proc. Natl. Acad. Sci. USA
98, 3197–3202.
Palty, R., Silverman, W.F., Hershfinkel, M., Caporale, T., Sensi, S.L., Parnis, J., Nolte, C., Fishman, D., Shoshan-Barmatz, V., Herrmann, S., et al. (2010). NCLX is
an essential component of mitochondrial Na+/Ca2+ exchange. Proc. Natl. Acad. Sci. USA 107, 436–441.
Raveh, A., Cooper, A., Guy-David, L., and Reuveny, E. (2010). Nonenzymatic rapid control of GIRK channel function by a G protein-coupled receptor kinase. Cell
143, 750–760.
Tiscornia, G., Singer, O., and Verma, I.M. (2006). Production and purification of lentiviral vectors. Nat. Protoc. 1, 241–245.
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. S3
B
2+
Mito [Ca ] (488/430)
D
2.5
Ratio (YFP/CFP)
mito
1.3
control
SERAF
Ca
2+
control
SARAF
i
70
60
1.2
50
1.1
**
1
*
*
40
30
0.9
20
0.8
HEK293
2
Hela
0.1
2
0.1
HEK293-T
Ca
2+
e
E
Ca2+ ER
2
C
2+
Ratio (480/430)
Counts
Sort High
Pre-sort
Sort low
Ca
[nM]
A
*
SEC61
SARAF
Overlay
1.5
1
0.5
0
siRNA siRNA
SARAF control
Figure S1. SARAF Is an ER Resident Protein Identified after a Screening of Cells for Changes in Intracellular Ca2+, ‘‘Related to Figure 1’’
(A) Flow cytometry analysis of mitochondrial calcium levels in HEK293-RPmt cells transfected with complete cDNA expression library (dashed gray) or with the
subpopulation of cDNAs after two rounds of sorting procedure (see methods) for low (green solid line) or high (orange solid line) mitochondrial Ca2+ levels.
(B) Enhanced expression of SARAF decreases basal mitochondrial Ca2+ levels in HEK293-RPmt cells. Mitochondrial Ca2+ levels measured in HEK293-RPmt cells
transfected with SARAF cDNA (n = 11) or empty vector (control, n = 12). **p < 0.01.
(C) Comparison of basal Ca2+ levels in HeLa and HEK293-T cells transfected with SARAF cDNA or empty vector (control) and cultured under normal (1.5-2 mM) or
low (0.1 mM) Ca2+ containing growth media. HeLa (control n = 51, SARAF n = 54) and HEK293-T (control n = 154, SARAF n = 136). *p < 0.05.
(D) ER Ca2+ levels were measured in control HeLa cells (nontarget siRNA) and in HeLa cells depleted from SARAF expression (using SARAF siRNAs) both
heterologously expressing the ER-targeted cameleon calcium indicator (D1ER). Averaged D1ER emission ratios from individual cells are shown (control n = 42,
SARAF n = 36).
(E) Fluorescent images of a HEK293-T cell expressing SARAF-mKok (Red) and SEC61-GFP (Green).
S4 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
B
YFP-C’ SARAF SARAF-GFP SARAF-GFP SARAF-GFP SARAF-GFP
+siRNA #2
+siRNA #3
+siRNA #2
+siRNA #1 +siRNA control
3770
120
Averaged
GFP Fluorescence
A HEK293-T
100
GFP
0
80
60
40
20
0
+ + + +
SARAF
-GFP: si-C si- si- sio 1 2 3
DIC
ntr
ol
α−SEARF
-C
ns
AR
AF
F Jurkat
+ + + - + + + -
RT
700
500
M 15 25
32
32 15
siRNA control
25
α-Orai1
siRNA: control 1
SA SA pC
RA RA DN
F F- A3
G
FP
’S
bp
1000
700
α-STIM1
37
25
RT
+ - + -
RT
Actin
25cyc
Actin
hSARAF
25cyc
hSARAF
NA
siR
si
+m RNA
-S hnt
AR SA
ro
AF RA
l
co
F
G
siRNA control
2+
Ratio (340/380)
1.4
0 Ca
siRNA 1
2+
2mM Ca
2+
0 Ca
siRNA 2
2+
2mM Ca
2+
0 Ca
0.8
BHQ
BHQ
2+
2mM Ca
2+
H
2+
0 Ca
2mM Ca
BHQ
BHQ
ΔR
0.6
0.4
0.2
00 200 400 600 80010001200
Time (s)
0 200 400 600 80010001200
Time (s)
si
+m RNA
co -SA h-S
RA AR
nt
ro
F AF
l
siRNA 3
1.2
1
mSARAF
NA
siR
siRNA SARAF
+ + -
40cyc
32 32 :Cycle
3
2
0 200 400 600 80010001200
Time (s)
0 200 400 600 80010001200
Time (s)
0.8
0.75
0.7
0.65
0.6
0.55
0.5
0.45
0.4
**
*
siRNA 3
o
fec l
ted
n.s.
α-SARAF
siRNA 2
100
α-Na/K
ATPase No si-C si-1 si-2 si-3 si-2
n t on
YF +
ra
tr
P
F
*
control
α-GFP
kDa
50
37
SA
RA
E Hela
kDa
250
150
75
50
siRNA 1
- + + + + -
SERAF-GFP:
D
SOCE (Δ R)
C
-
si2
YF +
PC’
Figure S2. Knockdown Efficiency and Specificity of siRNA Targeting SARAF Assessed in HEK293-T, HeLa and Jurkat Cells, ‘‘Related to
Figure 1’’
(A) Representative light and GFP fluorescence images of HEK293-T cells cotransfected with siRNAs against SARAF or control together with SARAF-GFP or
YFP-C0 (196-339)SARAF as indicated. SARAF siRNA#2 targets a region that is missing from the YFP-C0 (196-339)SARAF sequence and therefore used as
a negative control for the specificity of SARAF siRNA-mediated silencing.
(B) Quantitation of SARAF-GFP or YFP-C0 (196-339)SARAF fluorescence in siRNA transfected cells (n = 8 regions).
(C) Immunoblot analysis of postnuclear lysates prepared from HEK293-T cells described in (A) using antibodies agianst GFP or Na/K-ATPase.
(D) Immunoblot analysis of postnuclear lysates prepared from HEK293-T cells transfected with SARAF, SARAF-GFP or empty vector using antibodies against
SARAF. Note that specific staining for SARAF is correlated with enhanced expression (solid arrow) while nonspecific (n.s) staining (dashed arrow) is seen in all
protein preparations.
(E) Immunoblot analysis of SARAF, STIM1 and Orai1 expression in postnuclear lysate prepared from HeLa cells transfected with siRNAs against SARAF or control.
(F) Semi Quantitative RT-PCR of SARAF mRNA expression. Jurkat T cells were transfected with siRNAs targeting either SARAF, control or with SARAF siRNA
together with cDNA coding for the mouse isoform of SARAF as indicated. Actin or SARAF cDNA transcripts were amplified using 15, 25 and 32 PCR cycles.
Mouse SARAF transcripts were amplified using higher PCR cycles (40). RT indicates when reverse transcriptase was added during cDNA synthesis.
(G and H) Intracellular Ca2+ traces recorded from individual Jurakat T cells (G) after activation of SOC by BHQ (2 mM) and re-addition of 2 mM Ca2+ to the
incubation medium (n = 30 in each). (H), Quantitation of SOCE activity measured as the averaged increase in [Ca2+]i (DR, as indicated in (G), n = 5 independent
experiments). Note that although distinct regions of the transcript had been targeted by three different siRNAs, silencing of SARAF correlates with changes in
SOC activity, indicating an on-target mechanism of action of these siRNAs.
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. S5
A
B
Orai1/STIM1
Ca2+
Ca2+
0 10mM
10mM
0
0
-20
-40
-60
-80
-90
La
2
-60
3
1
TG
0.6
0.4
1000
2000
Time (s)
G
siRNA control
Ratio (340/380)
0 Ca
2+
2+
2mM Ca 0 Ca
1.4
1.2
1
BHQ
0.8
0.6
0.4
0.2
0
0
200 400 600 800 1000 1200
-40
2
3
10
10
60 mV
-30
pA/pF
-60
1
-60
-90
-90
E
control
SARAF
0
Ratio (340/380)
Ca2+
2mM
0.8
0.2
0
-90
Time (s)
1.4
2+
2mM
Histamine
0
2mM
Ca2+
TG
1.2
1
0.8
0.6
0.4
0.2
0
1000
2+
0 Ca
2+
H
2+
2mM Ca 0 Ca
1.4
1.2
1
BHQ
0.8
0.6
0.4
0.2
0
0
1.2 0 Ca
2+
2+
0 Ca
2mM Ca
2+
Tg
0.6
0.4
0.2
siRNA SARAF
siRNA control
0
0
200 400 600 800 1000
1200 1400 1600
Time (s)
siRNA SARAF
+ mSARAF
2+
0 Ca
200 400 600 800 1000 1200
Time (s)
2mM Ca
ATP
+Crb
1
0.8
2000
Time (s)
siRNA SARAF
Ratio (340/380)
Ratio (340/380)
1.2 Histamine
1
60 mV
pA/pF
1
D
0
Orai1/STIM1/SARAF
10
-30
3
-1000 100 200 300 400 500 600
10
3+
2
-40
Time (s)
siRNA control
siRNA SARAF
0
2mM
-40
-20
-80
-100
0 100 200 300 400 500 600
C
F
0
Ratio (340/380)
0 10mM
Ratio (340/380)
STIM1/Orai1
0
10mM
pA/pF
Normalized I
CRAC
(%)
STIM1/Orai1/SARAF
Time (s)
2+
2+
2mM Ca 0 Ca
2+
1.4
1.2
1
BHQ
0.8
0.6
0.4
0.2
0
0
200 400 600 800 1000 1200
Time (s)
Figure S3. Functional Effects of SARAF on Both SOCE and Icrac, ‘‘Related to Figure 2’’
(A) Mean of normalized (Left, control n = 25, SARAF-GFP n = 23) and a representative (Right) time course of CRAC currents recorded from HEK293-T cells under
the same experimental paradigm as in Figure 2A are shown.
(B) Current/Voltage relationships (I/V) measured at the indicated time points. Left, I/V traces recorded from STIM1- and Orai1-expressing cell. I/V trace recorded
after addition of 10 mM La3+ at the end of experiment is shown as well (blue). Right I/V traces recorded from STIM1-, Orai1- and SARAF-GFP-expressing cell.
(C and D) Mean of intracellular Ca2+ traces recorded from individual HeLa cells either transfected with siRNA against SARAF or control (C) or cotransfected with
STIM1 and Orai1 with or without SARAF (D). SOC was activated by applying Histamine (100 mM) or Tg (0.2 mM) after repetitive addition and removal of 2 mM Ca2+
as indicated. siRNA SARAF (n = 74), siRNA control (n = 89), STIM1/Orai1/SARAF (n = 26) and STIM1/Orai1 (n = 39).
(E) Mean intracellular Ca2+ traces recorded from individual Jurkat cells transfected with siRNA against SARAF (n = 45) or nontarget control (n = 52) and after
activation of SOC by passive store depletion or Tg (0.2 mM) and repetitive addition and removal of 2 mM Ca2+ as indicated.
(F–H) Intracellular Ca2+ traces recorded from individual Jurakat T cells treated with (F) control siRNA, (G) SARAF siRNA and (H) SARAF siRNA together with murine
SARAF cDNA after activation of SOC by BHQ (2 mM) and re-addition of 2 mM Ca2+ to the incubation medium (n = 15 in each plot).
S6 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
siSTIM1
B
0
siSTIM2
YFP-STIM2
siControl
C
65000
3000 p=0.007
2500
2000
p=0.005
1500
1000
siRNA
control
D SARAF+STIM1+Orai1
STIM1
Tg+BAPTA
YFP-STIM1
YFP-STIM2
SARAF
Tg+BAPTA
siRNA
STIM1/2
Ratio (340/380)
siControl
Averaged
YFP Fluorescence
YFP-STIM1
A
0 Ca
0.4
2+
2mM Ca
2+
0.35
0.3
Tg
0.25
0.2
0.15
0.1
0
siRNA stim1/2
control
100 200 300 400 500 600
Time (s)
Overlay
Tg+BAPTA
Figure S4. Analysis of STIM1 and SARAF Cellular Distribution and Efficiency of siRNA-Mediated Knockdown of STIM Proteins Expression
and Function in HEK293-T Cells, ‘‘Related to Figure 3’’
(A) Representative images of YFP fluorescence in HEK293-T cells cotransfected with siRNAs against STIM1 or control together with STIM1-YFP cDNA or with
siRNAs against STIM2 or control together with STIM2-YFP cDNA.
(B) Quantitation of STIM1-YFP or STIM2-YFP fluorescence in siRNA transfected cells (n = 9 regions).
(C) Mean of intracellular Ca2+ traces recorded from individual HEK293-T cells transfected with siRNA against STIM1 and STIM2 or with control siRNA after
activation of SOCE by Tg (0.2 mM) and re-addition of 2 mM Ca2+ to the incubation medium as indicated (siRNA control n = 30, siRNA STIM1/2 n = 29).
(D) Fluorescent images of cells expressing Orai1, SARAF-GFP and STIM1-mCherry, taken under TIRF microscopy configurations. Images of the same cell taken
before and after intracellular loading with BAPTA-AM (30 min) and application of Tg for additional 15 min (0.2 mM).
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. S7
A SARAF(173-339)+STIM1+Orai1
Orai1
BHQ
SARAF(173-339)
BHQ
BHQ
C SARAF(196-339)+STIM1+Orai1
STIM1
Tg
BHQ
STIM1
Tg
Overlay
SARAF(196-339)
Tg
Tg
F
Overlay
SARAF(196-339)
IP GFP
α-GFP
Tg
BHQ
D SARAF(196-339)+STIM1
Tg
Tg
E SARAF(196-339)+STIM1
STIM1
BHQ
Overlay
E148A SARAF
Orai1
Overlay
SARAF(196-339)
Tg
B SARAF(E148A) +STIM1+Orai1
Overlay
Tg
α-STIM1
Input
50
37
100
75
.1+
A3
DN
F
pC
RA g
SA F+T
N’
RA
SA 1+
.
N’
A3
DN
pC
F
RA Tg
+
SA
N’ RAF
SA
N’
Figure S5. Analysis of the Domains Required for SARAF Interaction and Clustering with STIM1, ‘‘Related to Figure 4’’
(A and B) TIRF images of cell surface from cells expressing mCherry-Orai1, STIM1 and the (A) short isoform SARAF-GFP or (B) mCherry-Orai1, STIM1 and
SARAF(E148A)-YFP before and after SOC induction by BHQ (2 mM).
(C–E) Fluorescent images of cells expressing YFP-SARAF(196-339) and STIM1-mCherry with (C) or without Orai1 (D and E), taken under epifluorescent (C and D)
or TIRF microscopy (E) configurations. Images of the same cell taken after the application of Tg (0.2 mM). (F) Western blot analysis of input or immunoprecipitated
material (IP) prepared from cells expressing Orai1, STIM1,and SARAF (1-191)-YFP before and after SOC induction by Thapsigergin. Anti-GFP were used to
immunoprecipitate SARAF (1-191)-YFP and antibodies against GFP and STIM1 were used for protein detection.
S8 Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc.
A
B
STIM1-Cherry
SARAF-GFP
1.2
Normalized F
Normalized F
1
0.8
0.6
0.4
0.2
0
0
50
100
150
Time (s)
0.6
0.4
0.2
0(s)
IRF
1000
100
102 104 106 108 110 112 114 116 118
Relative fraction (%)
4
10
E
ATP/Crb/Tg
70
65
60
Tau donor
55
50
45
Tau FRET
40
35
30
0
50
100
150
200
Time (s)
Time (ns)
F
0 Ca
2+
2mM Ca
2+
ATP/Crb/BHQ
170
160
150
140
130
100 200 300 400 500 600 700
Time (s)
Relative fraction (%)
180
Counts
200(s)
D
SARAF-GFP + STIM1
SARAF-GFP + STIM1-mCherry
Counts
0.8
0
200
C
1
0 Ca
70
60
2+
2mM Ca
2+
ATP/Crb/BHQ
Tau donor
50
40
Tau FRET
30
0
100 200 300 400 500 600 700
Time (s)
Figure S6. Interaction between SARAF and STIM1 Assessed by FRET Analysis, ‘‘Related to Figure 5’’
(A) Time course of mCherry photo-destruction under constant illumination with a 561 nm laser.
(B) Average of normalized GFP and mCherry emissions recorded from six individual cells expressing Orai1, SARAF-GFP, and STIM1-mCherry before (t0) and after
(t200) mCherry photo-destruction.
(C) Histogram of SARAF-GFP expressed with an untagged STIM1 (green trace) displays a fluorescence lifetime with a nearly monoexponential decay (2.59 ns).
Interaction between SARAF-GFP and STIM1-mCherry (red trace) results with double-exponential lifetime histogram, the first component of 2.59 ns (63%) and the
second component of 1.2 ns (37%) wich correspond to 74% FRET efficiency.
(D) Black trace depicts the FRET-free SARAF-GFP (t of 2.59 ns fraction) and blue trace depicts the faster lifetime component (t of 1.2 ns) that results from FRET
between SARAF-GFP and STIM1-Cherry.
(E) Time course of SARAF-GFP emission from a representative cell coexpressing Orai1 and STIM1-mCherry after store depletion and refilling. Store depletion and
activation of SOCE is initiated by applying ATP, Charb (100 mM each) and BHQ. Store refilling and SOCE deactivation occurs after agonist washout and addition of
2 mM Ca2+ to the external milieu as indicated. Changes in SARAF-GFP emission are in agreement with GFP quenching and dequenching by mCherry on increase
and decrease of FRET, respectively.
(F) The same FLIM-FRET analysis as in (D) after reversible store depletion.
Cell 149, 425–438, April 13, 2012 ª2012 Elsevier Inc. S9