Cell, Vol. 99, 703±712, December 23, 1999, Copyright 1999 by Cell Press
Transport-Dependent Proteolysis of SREBP:
Relocation of Site-1 Protease from Golgi to ER
Obviates the Need for SREBP Transport to Golgi
complex is formed by an interaction between the cyto-
solic COOH-terminal regulatory domain of the SREBPs
and the cytosolic WD40 domain of SCAP. SCAP targets
theSREBPs toS1P, a membrane-bound serineprotease
whose active site faces the lumen of the ER and post-
ER organelles (Sakai et al., 1998b). S1P cleaves the
luminal loop of the SREBPs following the tetrapeptide
sequence Arg-Xaa-Xaa-Leu, thereby separating SREBP
into two halves (Duncan et al., 1997). This event allows
a second protease, designated Site-2 protease (S2P),
to cleave the NH2-terminal transmembrane domain,
therebyliberating theNH2-terminalsegments ofSREBPs
from the membrane and allowing them to enter the nu-
cleus (Sakai et al., 1996; Rawson et al., 1997).
This complex proteolytic activationmechanismallows
lipid synthesis tobe controlled by the lipid contentofcell
membranes (Brownand Goldstein, 1997).Whencells are
depleted of sterols, the cleavage reactions are rapid,
and the NH2-terminal segments of SREBPs readily enter
the nucleus. When cells are overloaded with sterols, the
Site-1cleavage reactionis blocked.The SREBPs remain
bound to cell membranes, and the target genes are no
longeractivated. Cholesterol synthesis stops as a result
of decreased transcription of genes encoding 3-hydroxy-
3-methylglutaryl coenzyme A (HMG CoA) synthase,
HMG CoA reductase, and other enzymes of the cho-
lesterol biosynthetic pathway. Cholesterol uptake is
blocked by a downregulationof the low-density lipopro-
tein (LDL) receptor, and the synthesis of unsaturated
fatty acids is partially decreased as a result of reduced
transcription of genes encoding acetyl CoA carboxyl-
ase, fatty acid synthase, and stearoyl CoA desaturase
(Brown and Goldstein, 1997; Edwards and Ericsson,
Biochemical and genetic data show that SCAP is the
key to the regulation of Site-1 cleavage (Hua et al.,
1996a; Nohturfft et al., 1996, 1999; Rawson et al., 1999).
Mutant Chinese hamster ovary (CHO) cells that lack
SCAP fail to cleave SREBPs at Site-1, and they are
therefore auxotrophic for cholesterol and unsaturated
fatty acids (Rawson et al., 1999). Sterols block Site-1
cleavage by reducing the activity of SCAP. This inhibi-
tiondoes notoccurwhenSCAP bears substitutionmuta-
tions at either of two positions in its membrane attach-
ment domain (Tyr298Cys or Asp443Asn) (Hua et al.,
1996a; Nohturfft et al., 1996, 1998b). This portion of the
membrane attachment domainhas beendesignated the
sterol-sensing region. Mutant CHO cells that produce
these altered forms of SCAP overproduce cholesterol
and fail to slow this synthesis when cholesterol overac-
cumulates in the cell.
Evidence indicates that SCAP acts by escorting
SREBPs to a post-ER compartment where the active
S1P resides and that sterols may prevent cleavage by
blocking this transport (Nohturfft et al., 1998b, 1999).
This hypothesis emerged from studies of the N-linked
carbohydrates attached to a luminal loop in the poly-
topic membrane domain of SCAP. When cells are de-
pleted of sterols, these carbohydrates are largely resis-
tant to cleavage by endoglycosidase H (endo H),
Russell A. DeBose-Boyd, Michael S. Brown,*
Wei-Ping Li, Axel Nohturfft, J oseph L. Goldstein,*
and Peter J . Espenshade
Department of Molecular Genetics
University of Texas
Southwestern Medical Center
Dallas, Texas 75390-9046
Cholesterolhomeostasis in animalcells is achieved by
regulated cleavage of membrane-bound transcription
factors, designated SREBPs.Proteolytic release ofthe
active domains ofSREBPs frommembranes requires a
sterol-sensing protein, SCAP, which forms a complex
with SREBPs. In sterol-depleted cells, SCAP escorts
SREBPs from ER to Golgi, where SREBPs are cleaved
by Site-1 protease (S1P). Sterols block this transport
and abolish cleavage. Relocating active S1P from
Golgi to ER by treating cells with brefeldin A or by
fusing the ER retention signal KDEL to S1P obviates
the SCAP requirement and renders cleavage insensi-
tive to sterols. Transport-dependent proteolysis may
be a common mechanism to regulate the processing
of membrane proteins.
Cleavage of sterol regulatory element binding proteins
(SREBPs)by Site-1 protease (S1P)initiates a process by
whichtheactive fragments oftheSREBPs translocateto
the nucleus and activate genes controlling the synthesis
and uptake of cholesterol and unsaturated fatty acids
are synthesized as tripartite membrane-bound proteins
averaging 1150 amino acids in length. The NH2-terminal
segment of ?480 amino acids is a transcription factor
of the basic helix-loop-helix-leucine zipper (bHLH-Zip)
family. This is followed by an 80±amino acid membrane
attachment domain that comprises two transmembrane
helices separated by a short hydrophilic loop of ?30
amino acids. This is followed by a COOH-terminal regu-
latory domain of ?590 amino acids. Newly synthesized
SREBPs are bound to membranes of the nuclear enve-
lope and endoplasmic reticulum(ER)ina hairpinfashion
with the NH2-terminal and COOH-terminal segments
projecting into the cytosol and the hydrophilic loop pro-
jecting into the lumen (Brown and Goldstein, 1997).
The newly synthesized SREBPs form complexes with
a polytopic membrane protein designated SCAP (SREBP
cleavage-activating protein) (Sakai et al., 1997, 1998a).
The NH2-terminal domain of SCAP contains eight trans-
membrane helices, and the COOH-terminaldomaincon-
tains five WD40 repeats that mediate protein±protein
interactions (Nohturfft et al., 1998a). The SREBP/SCAP
*To whom correspondence should be addressed (e-mail: mbrow1@
indicating that SCAP has reached the Golgi apparatus
where it has been modified by Golgi mannosidase II
(Kornfeld and Kornfeld, 1985). When cells are over-
loaded withsterols, the N-linked sugars of SCAP remain
in an endo H±sensitive form, indicating that SCAP has
not reached the Golgi. In sterol-depleted cells, most of
the SCAP cofractionates with ER markers even though
it has been modified by Golgi enzymes (Nohturfft et al.,
1999). These findings have led to a modelinwhichSCAP
carries SREBP from the ER to the Golgi in sterol-
depleted cells, after which SCAP recycles to the ER
following Site-1 cleavage. Insterol-overloaded cells, the
SCAP/SREBP complex does notleave the ER and Site-1
cleavage cannot take place.
A corollary of the sterol-regulated cycling model is
that S1P is not active in the ER and that it becomes
activated only after transport to a post-ER compart-
ment. Partial support for this model comes from experi-
ments on the activation process of S1P. S1P, like other
mammalian subtilases, is synthesized in the ER as an
inactive precursor(Espenshade et al., 1999). It becomes
active only afterit is autocatalytically cleaved, liberating
an NH2-terminal propeptide. Most of the inactive S1P
precursor, designated S1P-A, is endo H sensitive, sug-
gesting that the protein is largely located in the ER. On
the otherhand, the majoractive formofS1P, designated
S1P-C, is endo H resistant, suggesting that it resides in
the Golgi (Espenshade et al., 1999).
If the primary function of SCAP is to transport SREBP
to the post-ER compartment that contains active S1P,
the requirement for SCAP should be abolished if we
could generate active S1P in the ER. The current experi-
ments were designed to test this hypothesis using mu-
tant SRD-13A cells thatlack SCAP (Rawsonetal., 1999).
We used two approaches to trap active S1P in the ER.
First, we treated the SRD-13A cells with brefeldin A, an
agent that blocks anterograde movement of proteins
fromER to Golgi, thus causing Golgi proteins to translo-
cate back to the ER (Lippincott-Schwartz et al., 1989).
Second, we transfected the SRD-13A cells with a cDNA
encoding a truncated, soluble luminal form of S1P with
an ER retention signal inserted at its COOH terminus.
Both of these treatments led to Site-1 cleavage of
SREBPs in the SCAP-deficient SRD-13A cells, and in
both cases this SCAP-independent cleavage was not
inhibited by sterols. These data provide strong experi-
mental support forthe sterol-regulated transport model
of SCAP function.
Figure 1. Brefeldin A Blocks Sterol-Mediated Suppression of
SREBP Cleavage in CHO Cells
On day 0, CHO-7 cells were set up at 5 ? 105cells per 10 cm dish
in medium A supplemented with 5% fetal calf serum. On day 2, the
cells were switched to medium A supplemented with 5% newborn
calf lipoprotein-deficient serum and compactin/mevalonate in the
absence (A) or presence (B) of sterols. After 16 hr, the cells were
switched to mediumA supplemented with5% newborncalflipopro-
tein-deficient serum and compactin/mevalonate in the absence or
presence of sterols and 1 ?g/ml brefeldin A as indicated. After 5
hr, the cells were harvested, and membrane and nuclear extract
fractions were prepared as described in the Experimental Proce-
dures. Aliquots of the membranes (50 ?g protein) and nuclear ex-
tracts (100 ?g protein) were subjected to SDS-PAGE. Immunoblot
analysis was performed with 5 ?g/ml of monoclonal IgG-7D4, which
recognizes SREBP-2. The filters were exposed to Kodak X-Omat
Blue XB-1 film for 10 s (membranes) and 1 min (nuclear extracts) at
room temperature. P, I, and N denote the precursor, intermediate,
and nuclear forms of SREBP, respectively.
incubated in the absence of sterols (lower panel, lane
1) but not in the presence of sterols (lane 2). When the
final 5 hr incubation was conducted in the presence of
brefeldinA, nuclearSREBP-2 continued to be visualized
in the sterol-depleted cells (lane 3), but there was no
longer any decrease when sterols were added (lane 4).
The amount of the membrane-bound precursor did not
change under any of these conditions (upper panel).
Brefeldin A did produce an increase in the membrane-
bound intermediate of SREBP (designated I in Figure
1), which is the immediate product of Site-1 cleavage.
This experiment shows that brefeldin A blocks the sup-
pression of Site-1 cleavage by sterols.
Panel B of Figure 1 shows that brefeldin A induces
SREBP cleavage in cells that were previously inhibited
by sterols. In this experiment, the cells were incubated
for 16 hr in the presence of sterols, and then some of
the dishes were switched to sterol-free medium in the
absence or presence of brefeldin A. In the absence of
brefeldin A, we found no nuclear SREBP-2 in the sterol-
treated cells (panel B, lane 2). In the presence of brefel-
din A, nuclear SREBP-2 appeared even when sterols
were present (lane 4). Again, brefeldin A blocked sterol
suppression of SREBP cleavage.
A trivial explanation for the findings in Figure 1 would
exist if brefeldin A blocked the movement of the exoge-
nous sterols to the ER. This possibility is rendered un-
likely by thepriorfindings ofRidgway and Lagace(1995),
who showed that brefeldin A does not interfere with
the ability of 25-hydroxycholesterol to stimulate the ER
synthesis of cholesteryl esters in CHO cells. This result
was confirmed in our laboratory (Nohturfft et al., 1999).
Figure 1 shows an experiment designed to test the ef-
fects of brefeldin A on sterol-regulated cleavage of
SREBP-2 in wild-type CHO-7 cells. In panel A, the cells
were incubated for16 hrwithout sterols to induce cleav-
age of SREBPs. The cells were then incubated for an
additional5 hrinthe absence orpresence ofa mixture of
25-hydroxycholesteroland cholesterol, whichefficiently
suppresses Site-1 cleavage.Extracts ofcellmembranes
and nuclei were then subjected to SDS-PAGE and blot-
ted with an antibody against the NH2-terminal segment
of SREBP-2. As expected, the NH2-terminal segment
was found in the nuclear extract when the cells were
Transport of SREBPs from ER to Golgi
Figure 2. Brefeldin A Induces S1P-Depen-
dent Cleavage of SREBPs in Transfected
(A) On day 0, SCAP-deficient SRD-13A cells
were set up at a density of 4 ? 105cells/60
mm dish in medium B. On day 1, cells were
transfected with 2 ?g/dish of pTK-HSV-BP2
together with 0.25 ?g/dish pTK-SCAP. The
total amount of DNA was adjusted to 4 ?g/
dish with pcDNA3 empty vector and pTK
mock vector.Onday 2,thecells werecultured
for16 hrin medium A supplemented with 5%
newborn calf lipoprotein-deficient serum and
compactin/mevalonate in the absence or
presence ofsterols.Onday 3, cells were incu-
bated with or without 1 ?g/ml brefeldin A for
5 hr, after which they were harvested and
fractionated as described in Figure 1. Ali-
quots of membranes (6 ?g protein) and nuclear extracts (10 ?g protein) were subjected to SDS-PAGE, and immunoblot analysis was carried
out with 0.1 ?g/ml of IgG-HSV-Tag for pTK-HSV-BP2 and 5 ?g/ml IgG-R139 for SCAP. Filters were exposed to film for 1 s (top and bottom
panels) and 30 s (middle panel). Molecular mass standards are expressed in kilodaltons.
(B) On day 0, S1P-deficient SRD-12B cells were set up at a density of 3 ? 105cells/60 mm dish in medium B. On day 1, cells were transfected
with 2 ?g/dish of pTK-HSV-BP2 together with 0.5 ?g/dish pCMV-Myc-S1P. The total amount of DNA was adjusted to 4 ?g/dish with pcDNA3
empty vector and pTK mock vector. On day 2, the cells were cultured for 16 hr in the absence or presence of sterols as described in (A). On
day 3, cells were treated with 1 ?g/ml brefeldin A for 5 hr prior to harvest and fractionation. Aliquots of membranes (14 ?g protein) and nuclear
extract (33 ?g protein) were subjected to SDS-PAGE, and immunoblot analysis was carried out with 0.1 ?g/ml of IgG-HSV-Tag for pTK-HSV-
BP2 and 0.5 ?g/ml anti-Myc 9E10 for Myc-S1P. Filters were exposed to film for 5 s (top panel), 15 s (middle panel), and 1 s (bottom panel).
Molecular mass standards are expressed in kilodaltons.
If the effect of brefeldin A on SREBP cleavage is
caused byrelocationofactiveS1P totheER,thenbrefel-
din A should relieve the requirement for SCAP. To test
this hypothesis, we measured the cleavage of epitope-
tagged SREBP-2 in transfected SRD-13A cells, which
lack SCAP as a result of mutations in the SCAP gene
(Figure 2A). In the absence of SCAP and brefeldin A,
these cells showed no nuclear SREBP-2 whether incu-
bated in the absence or presence of sterols (middle
panel, lanes 1 and 2). Treatment with brefeldin A caused
the appearance of nuclear SREBP-2, and this was not
prevented by sterols (lanes 3and 4).As a positivecontrol
for these experiments, we transfected the cells with a
cDNA encoding wild-type SCAP (lanes 5±8). In the ab-
sence of brefeldin A, the SCAP cDNA restored nuclear
SREBP-2, and cleavage was blocked by sterols (lanes
5 and 6). Brefeldin A allowed cleavage to persist in the
presence of sterols (lanes 7 and 8). To confirm that
the nuclear SREBP-2 resulted from cleavage by Site-1
protease, we performed a similar experiment in SRD-
12B cells, which lack S1P as a result of mutations in the
S1P gene (Figure 2B). In the absence of S1P, brefeldin
A failed to cause the appearance of nuclear SREBP-2
either in the absence or presence of sterols (middle
panel, lanes 1±4). When we cotransfected a cDNA en-
coding S1P, we restored cleavage of SREBP-2, and this
was inhibited by sterols (lanes 5 and 6). Sterol suppres-
sion was blocked by brefeldin A (lanes 7 and 8). Thus,
treatment of SCAP-deficient cells with brefeldin A res-
cues S1P-dependent cleavage of SREBPs.
To study the subcellular localization of S1P directly,
we performed double-label immunofluorescence stud-
ies ofCHO/pS2P cells and S1P-deficientSRD-12B cells,
which were derived from the CHO/pS2P cells (Figure 3).
When stained with an antibody against S1P, the CHO/
pS2P cells exhibited intense staining of a juxtanuclear
organelle whose appearance was consistent with that
of the Golgi complex (Figure 3A). This intense staining
was absent in the SRD-12B cells (Figure 3D), indicating
that it was specific for S1P. We also observed a faint
reticular staining pattern that appeared nonspecific
since it was present in the SRD-12B cells (Figure 3D)
as well as the CHO/pS2P cells (Figure 3A). Because of
this nonspecific staining, we could not determine
whether any S1P is located in the ER. Incubation of
cells with rhodamine-conjugated wheat germ agglutinin
(WGA)stained the trans-Golginetwork inthe CHO/pS2P
cells (Figure 3B) and the SRD-12B cells (Figure 3E).
Overlay ofthe greenS1P image withthe red WGA image
revealedthatS1P andtheWGA Golgimarkercolocalized
in CHO/pS2P cells, as indicated by the merged yellow
image (Figure 3C). No colocalization was observed in
the SRD-12B cells, whichis consistent withthe absence
of S1P in these cells (Figure 3F).
We used the same technique to trace the fate of S1P
in brefeldin A±treated CHO/pS2P cells (Figure 4). In the
absence of brefeldin A, the anti-S1P antibody showed
intense Golgifluorescence and a smallamountofreticu-
larfluorescence (Figure 4A). Whenthe cells were treated
with brefeldin A, the Golgi fluorescence disappeared,
and the reticular pattern increased (Figure 4D). Staining
with WGA revealed a Golgi pattern (Figure 4B), and this
was unaffected by brefeldin A (Figure 4E). This finding
is consistent with previous data showing that brefeldin
A causes the cis- and medial-Golgi to collapse into the
ER, but it does not affect the trans-Golgi (Lippincott-
Schwartz et al., 1989). Overlay of the S1P and WGA
images confirmed the Golgi localization of S1P in the
untreated cells (Figure 4C), and this was abolished in
the brefeldin A±treated cells (Figure 4F). This finding
suggests that S1P is located in the cis- and medial-
Golgi, but not in the trans-Golgi.
We previously reported thatS1P is processed to three
forms, designated S1P-A, S1P-B, and S1P-C (Espen-
shade etal., 1999). The NH2-terminalsequences ofthese
Figure 3. Endogenous S1P Localizes to the
Golgi Apparatus in CHO Cells
Wild-type CHO/pS2P cells (A±C) orS1P-defi-
cient SRD-12B cells (D±F) were grown in me-
dium B, fixed, and stained with antibodies
to S1P (A and D) or rhodamine-conjugated
wheat germagglutinin (B and E)as described
in the Experimental Procedures. Merged im-
ages (C and F) show colocalization of S1P
and the Golgi marker, wheat germ agglutinin.
Bar, 20 ?m.
proteins were examined, and this revealed that S1P-A
is the product of cleavage by signal peptidase. S1P-A
retains the propeptide, and it is enzymatically inactive.
S1P-B and S1P-C are produced by autocatalytic re-
moval of the propeptide, and they are enzymatically
active.TheN-linked carbohydrates onS1P-A and S1P-B
are mostly endo H sensitive, indicating that the bulk
of these proteins have not reached the medial-Golgi.
S1P-C is endo H resistant, implying that it has reached
the medial-Golgi. All of these experiments were per-
formed with transfected cells that overexpressed epi-
tope-tagged S1P. To confirm that these same findings
apply to endogenous S1P, we performed immunoblot-
ting experiments with anti-S1P in nontransfected CHO
cells (Figure 5). As reported previously, we observed
S1P-A, S1P-B, and S1P- C inSRD-12B cells thatoverex-
press transfected S1P (Figure 5A, lane 2). In contrast,
the nontransfected CHO cells showed only two bands,
corresponding in size to S1P-A and S1P-C (Figure 5A,
lane 3). This indicates that S1P-C is the major form
of active S1P in nontransfected cells and that S1P-B
becomes prominent only when the protein is overex-
pressed as a resultoftransfection.WhenCHO cellmem-
branes were treated with peptide N-glycosidase F
(PNGase F), the mobility ofthe A and C forms increased,
indicating that both forms have N-linked sugars (Figure
5B, lanes 2 and 3). Endo H treatment increased the
mobility of S1P-A to approximately the same extent as
PNGase F, indicating that all of the N-linked sugars
were endo H sensitive (Figure 5B, lane 4). The small
difference in mobility between PNGase F±treated and
endo H±treated S1P-A may be due to the fact that endo
H does not remove the asparagine-linked N-acetylglu-
cosamine whereas PNGase F does. The mobility ofS1P-C
was only partially increased by endo H, and none of
it comigrated with the PNGase F product. These data
indicate that all molecules of S1P-C contain at least one
endo H±resistant carbohydrate chain, implying that this
protein has reached the medial-Golgi. Similar results
were obtained when the cells were incubated with ste-
rols (Figure 5B, lanes 7±10), which is consistent with
previous evidence that sterols do not affect the autocat-
alytic processing or subcellular localization of S1P (Es-
penshadeetal., 1999).Together,thesedata suggestthat
in nontransfected cells the active form of S1P, namely
S1P-C, resides in the Golgi.
In addition to brefeldin A treatment, another means
to relocate active S1P from the Golgi to the ER is to
Figure 4. Brefeldin A
Golgi Localization of S1P in CHO Cells
Wild-type CHO/pS2P cells were growninme-
dium B and cultured for 1 hr in the absence
(A±C) or presence (D±F) of 2 ?g/ml brefeldin
A. Cells were fixed and stained with antibod-
ies to S1P (A and D)orrhodamine-conjugated
wheat germagglutinin (B and E)as described
in the Experimental Procedures. Merged im-
ages (C and F) show localization of S1P and
the Golgi marker, wheat germagglutinin. Bar,
Transport of SREBPs from ER to Golgi
SRD-12B cells, which lack endogenous S1P, and we
isolated permanent clones of cells that grow in the ab-
sence of cholesterol as a result of expressing active
Figure 6B shows SREBP-2 processing in the parental
SRD-12B cells and the two lines of permanent transfec-
tants, designated S1P-KDEL and S1P-KDAS. When
SRD-12B cells were incubated inthe absence ofsterols,
the amount of SREBP-2 precursor was relatively low,
and no cleaved nuclearSREBP-2 was visualized (Figure
6B, lane 1). Addition of sterols caused an increase in
the amount of precursor, and there was still no nuclear
SREBP-2(lane 2).(We previously observed low amounts
of endogenous SREBP precursors in sterol-deprived
SRD-12B cells (Rawson et al., 1998), which we attribute
to degradation of the precursor by pathways that are
independent of S1P and do not lead to release of the
NH2-terminal fragment.) When the SRD-12B cells ex-
pressed the KDEL-terminated S1P, SREBP-2 was cleaved
to the nuclearform (lane 3), and cleavage was not inhib-
ited by sterols (lane 4). Expression of the KDAS-termi-
nated S1P also restored SREBP-2 processing (lane 5),
but in this case the SREBP-2 cleavage was abolished
by sterols (lane 6).
To follow the fate of the expressed S1P in the perma-
nent cell lines, we subjected aliquots of whole-cell ex-
tracts and culture medium to electrophoresis, followed
by immunoblotting with an antibody against the Myc
epitope tag (Figure 6C). As expected, we visualized no
Myc-tagged S1P in the parental SRD-12B cells (lanes 1
and 2). In the cells expressing the KDEL construct, we
visualized all three forms (S1P-A, S1P-B, and S1P-C).
In the cells expressing S1P-KDAS, we observed a strik-
ingly differentresult. The cellextracts contained equiva-
lent amounts of S1P-A and S1P-B, but there was much
less S1P-C than we observed in the cells expressing
S1P-KDEL. Instead, we found S1P-C in the culture me-
dium of the S1P-KDAS cells (lanes 5 and 6). Sterols had
no effect on the amount of S1P in either cell line. The
data of Figure 6 are consistent with the hypothesis that
S1P-KDEL is retained in the cells as a result of binding
to theKDEL receptorand retrievalto theER fromtheGolgi
and that this retrieval leads to cleavage of SREBP-2 that
cannot be blocked by sterols.
The experiments of Figure 7 were designed to deter-
mine whether S1P-KDEL obviates the requirement for
SCAP and to confirm that cleavage by S1P-KDEL re-
quires the RSVL recognition sequence in the luminal
loop of SREBP-2. For this purpose, we transfected
SCAP-deficient SRD-13A cells with a plasmid encoding
wild-type HSV-tagged SREBP-2 or a mutant version
containing an Ala in place of the Arg of the S1P recogni-
tion sequence, RSVL (Figure 7A). In the absence of any
cotransfected plasmid, the SRD-13A cells failed to gen-
erate the nuclear form of SREBP-2 (middle panel, lanes
2 and 3). Cotransfection of a plasmid encoding SCAP
restored SREBP-2 cleavage, and this was suppressed
by sterols (lanes 4 and 5). Cotransfection of S1P-KDEL
also restored cleavage, and this was no longer sup-
pressed by sterols (lanes 6 and 7). The S1P-KDAS con-
struct was unable to restore cleavage in SRD-13A cells
(lanes 8 and 9). The S1P-KDEL plasmid did notstimulate
cleavage of the R519A version of SREBP-2 (lanes 11
and 12). The bottom panel of Figure 7A shows that
Figure 5. ImmunoblotAnalysis ofS1P inTransfected and Nontrans-
fected CHO Cells
(A) Membrane fractions were prepared from eithertransfected S1P-
deficient SRD-12B cells (lanes 1 and 2) or nontransfected CHO/
pS2P cells (lane 3) as described in Experimental Procedures. SRD-
12B cells were transfected as described in Figure 2 with either 0.5
?g of pcDNA3 empty vector (lane 1) or pCMV-S1P (lane 2), and the
total amount of transfected DNA was adjusted to 3 ?g/dish by the
addition ofpcDNA3 empty vector. Transfected cells were incubated
in medium A supplemented with 5% fetal calf serum for 16 hr prior
to harvest and cell fractionation. Aliquots of membrane protein (3
?g, lanes 1 and 2; 15 ?g, lane 3) were subjected to SDS-PAGE and
immunoblotted with a 1:500 dilution of anti-S1P antiserum. Filter
was exposed to film for 10 s.
(B) CHO/pS2P and SRD-12B cells were cultured in medium A sup-
plemented with 5% newborn calf lipoprotein-deficient serum and
compactin/mevalonate in the absence (lanes 1±5) or presence of
sterols (lanes 6±10) for 17 hr. Cells were harvested, and membrane
fractions were prepared as described in the Experimental Proce-
dures. Aliquots of membrane protein (15 ?g) were treated with the
following glycosidases: lanes 1, 2, 5, 6, 7, and 10, none; lanes 3 and
8, 0.26 IU/ml peptide N-glycosidase F (PNGase F); lanes 4 and 9,
0.83 IU/ml endo H. Following enzymatic treatment, samples were
subjected to SDS-PAGE and immunoblotted with a 1:500 dilution
of anti-S1P antiserum. Filter was exposed to film for 15 s.
transfect cells with a mutant form of S1P that contains
an ER retention/retrieval signal. For this purpose, we
prepared a cDNA encoding S1P that was truncated just
ure 6A). Prior studies have shown that such a truncated
form of S1P is catalytically active (Cheng et al., 1999).
At the COOH terminus, we inserted three copies of a
Myc epitope tag followed by the sequence Lys-Asp-
Glu-Leu (KDEL). This mutant version of S1P should be
inserted into the ER lumen. After it exits from the ER, it
should bind to the KDEL receptor, which should trans-
port it back to the ER from the Golgi (Pelham, 1992). As
a control, we prepared a similar construct containing
the COOH-terminal sequence Lys-Asp-Ala-Ser (KDAS),
which is not recognized by the KDEL receptor (Munro
and Pelham, 1987). We transfected these cDNAs into
Figure 6. Stable Expression of S1P(1-997)
Myc-KDEL in S1P-Deficient SRD-12B Cells
Produces Unregulated Cleavage of SREBP-2
(A) Structure of S1P(1-997)Myc-KDEL and
S1P(1-997)Myc-KDAS. The propeptide (Pro)
and the subtilisin-like catalytic domain are
serine (Ser-414), Myc epitope tags, and the
COOH-terminal four amino acids (KDEL or
KDAS)are indicated.Amino acid numbers are
shown below the diagram.
(B) Cleavage of SREBP-2 in SRD-12B cells
stably transfected with pCMV-S1P(1-997)
Myc-KDEL. Clones ofSRD-12B cells thatsta-
bly express either S1P(1-997)Myc-KDEL or
S1P(1-997)Myc-KDAS were generated as de-
scribed in the Experimental Procedures.
SRD-12B cells (lanes 1 and 2) and SRD-12B
cells stably transfected with pCMV-S1P
(1-997)Myc-KDEL (lanes 3 and 4) or pCMV-
S1P(1-997)Myc-KDAS (lanes 5 and 6) were
cultured for16 hrin medium A supplemented
with 5% newborn calf lipoprotein-deficient
serum and compactin/mevalonate in the ab-
sence or presence of sterols, and whole-cell
extracts were prepared as described in the Experimental Procedures. Aliquots of cell extract (65 ?g protein) were subjected to SDS-PAGE
and immunoblotted with 5 ?g/ml of monoclonal IgG-7D4 anti-SREBP-2. Filters were exposed to film for 5 s (top panel) and 30 s (bottom
panel). P and N denote the precursor and cleaved nuclear forms of SREBP-2, respectively.
(C) Secretion of S1P(1-997)Myc-KDEL or S1P(1-997)Myc-KDAS in stably transfected SRD-12B cells. SRD-12B cells (lanes 1 and 2) and SRD-
12B cells stably transfected with pCMV-S1P(1-997)Myc-KDEL (lanes 3 and 4) or pCMV-S1P(1-997)Myc-KDAS (lanes 5 and 6) were cultured
for 16 hr in medium A supplemented with 5% newborn calf lipoprotein-deficient serum and compactin/mevalonate in the absence or presence
of sterols. Samples of medium and whole-cell extracts were prepared as described in the Experimental Procedures. Aliquots of cell extracts
(40 ?g protein) and medium (corresponding to an equivalent amount of cells) were subjected to SDS-PAGE and immunoblotted with 2.5 ?g/
ml anti-Myc 9E10 antibody. Filters were exposed to film for 1 s. Molecular mass standards are expressed in kilodaltons.
S1P-C was detectable in membrane vesicles derived
fromthecells expressing S1P-KDEL, butnotS1P-KDAS.
Figure 7B shows a similar transient transfection ex-
perimentinS1P-deficientSRD-12B cells.Inthe absence
ofa cotransfected plasmid, thesecells failed togenerate
the nuclearformofHSV-tagged SREBP-2(middle panel,
lanes 2 and 3). Expression of wild-type S1P restored
sterol-regulated cleavage (lanes 4 and 5). Expression of
S1P-KDEL also restored cleavage, but there was no
suppression by sterols (lanes 6 and 7). S1P-KDAS re-
stored cleavage to a lesser extent, and sterol suppres-
sion was preserved (lanes 8 and 9). Neither S1P-KDEL
norS1P-KDAS permitted cleavage of the R519A version
of SREBP-2 (lanes 11±14). Again, immunoblotting with
anti-Myc revealed abundant S1P-C in the cells express-
ing S1P-KDEL, but much lower amounts in the cells
expressing S1P-KDAS (lowerpanel of Figure 7B). These
data indicate that expression of S1P-KDEL in the ER
bypasses the SCAP requirement for SREBP cleavage.
Three lines ofevidencesupportthese conclusions.First,
previous studies ofSCAP glycosylationhave shownthat
SCAP moves from the ER to the Golgi when cells are
grown in the absence, but not the presence of sterols
(Nohturfft et al., 1998b, 1999). Second, S1P-C, the pre-
dominant form of active S1P, is found in or near the
Golgi apparatus, as shown by immunocytochemistry
(Figures 3 and 4) and studies of glycosylation patterns
(Figure 5). Third, retrograde transport of active S1P to
the ER by two independent methods elicits unregulated
SREBP processing and obviates the requirement for
SCAP in SRD-13A cells (see Figure 8B).
The currentfindings withbrefeldinA provide anexpla-
nation for the earlier findings of Ridgway and Lagace
(Ridgway and Lagace, 1995), who observed that brefel-
din A blocks sterol-mediated suppression of transcrip-
tion in CHO cells. By redistributing active S1P from the
Golgito the ER, brefeldinA eliminates the need to trans-
port SREBPs to the Golgi, allowing cleavage of SREBP
in the presence as well as absence of sterols. This con-
clusion is supported by the demonstration that a similar
effect can be achieved by attaching a KDEL signal to
S1P, thereby trapping it in the ER.
Thecurrentstudies also provideinformationaboutthe
Site-2 protease (S2P), which cleaves the NH2-terminal
intermediate formofSREBPs that is produced by cleav-
age at Site-1. If Site-1 cleavage normally occurs in a
Golgi compartment, then S2P must also be located in
this compartment or a more distal compartment in the
secretory pathway. When Site-1 cleavage takes place
in the ER, as in brefeldin A±treated wild-type CHO cells,
it is followed by Site-2 cleavage, thereby generating the
Thecurrentresults supportthehypothesis thatcleavage
of SREBPs at Site-1 requires vesicular transport of
SREBPs from the ER to the Golgi and that this transport
is blocked by sterols. The key to regulated transport is
SCAP, which forms a tight complex with SREBPs. In
sterol-depleted cells, SCAP escorts SREBP to a post-
ER compartment that contains active S1P (see model
in Figure 8A). In the presence of sterols, the SCAP/
SREBP complex remains in the ER, and SREBP never
reaches the compartment that contains active S1P.
Transport of SREBPs from ER to Golgi
Figure 7. Transfected S1P(1-997)Myc-KDEL Rescues Cleavage of SREBP-2 in Cells Deficient in SCAP (A) and S1P (B)
(A) Immunoblot analysis of SREBP-2 cleavage in transfected SCAP-deficient SRD-13A cells. SRD-13A cells were set up on day 0 in medium
B at 4 ? 105cells/60 mm dish and transfected on day 1 as described in the Experimental Procedures. The indicated plasmid encoding either
wild-type (lanes 2±9) or mutant R519A (lanes 11±14) HSV-tagged SREBP-2 (1.75 ?g/dish) was cotransfected into SRD-13A cells with empty
vector (lanes 1±3, and 10), pTK-SCAP (0.05 ?g/dish; lanes 4 and 5), pCMV-S1P(1-997)Myc-KDEL (0.25 ?g/dish; lanes 6, 7, 11, and 12), or
pCMV-S1P(1-997)Myc-KDAS (0.25 ?g/dish; lanes 8, 9, 13, and 14) as indicated. The total amount of transfected DNA was adjusted to 3.75
?g/dish by the addition of pTK empty vectorand pcDNA3 empty vector. Transfected cells were incubated for18 hrin medium A supplemented
with 5% newborn calf lipoprotein-deficient serum and compactin/mevalonate in the absence or presence of sterols, and the cells were
harvested and fractionated as described in the Experimental Procedures. Aliquots of membranes (3 ?g protein) and nuclear extract (15 ?g
protein) were subjected to SDS-PAGE and immunoblotted with 0.5 ?g/ml IgG-HSV-Tag antibody (top and middle panels) or 2.5 ?g/ml anti-
Myc 9E10 antibody (bottom panel). Filters were exposed to film for 1 s. P and N denote the precursor and cleaved nuclear forms of SREBP-2,
respectively. Molecular mass standards are expressed in kilodaltons.
(B) Immunoblot analysis of SREBP-2 cleavage in transfected S1P-deficient SRD-12B cells. SRD-12B cells were set up on day 0 in medium B
at 4 ? 105cells/60 mm dish and transfected on day 1 as described in the Experimental Procedures. The indicated plasmid encoding either
wild-type (lanes 2±9) or mutant R519A (lanes 11±14) HSV-tagged SREBP-2 (1.75 ?g/dish) was cotransfected into SRD-12B cells with empty
vector (lanes 1±3, and 10), pCMV-S1P-Myc (0.25 ?g/dish; lanes 4 and 5), pCMV-S1P(1-997)Myc-KDEL (0.25 ?g/dish; lanes 6, 7, 11, and 12),
or pCMV-S1P(1-997)Myc-KDAS (0.25 ?g/dish; lanes 8, 9, 13, and 14) as indicated. The total amount of transfected DNA was adjusted to 3.75
?g/dish by the addition of pTK empty vector and pcDNA3 empty vector. Cells were cultured and harvested as described in (A). Aliquots of
membranes (4 ?g protein) and nuclear extracts (11 ?g protein) were subjected to SDS-PAGE and immunoblotted with 0.5 ?g/ml IgG-HSV-
Tag antibody (top and middle panels) or 2.5 ?g/ml anti-Myc 9E10 antibody (bottom panel). Filters were exposed to film for 1 s. P and N denote
the precursor and cleaved nuclear forms of SREBP-2, respectively. Molecular mass standards are expressed in kilodaltons.
nuclear form of SREBP (Figure 1). This implies either
that some active S2P normally resides in the ER or that
S2P is translocated there along with S1P following bre-
feldin A treatment. We also generated the nuclear form
of SREBP after cleavage by the KDEL-terminated S1P
(Figures 6 and 7). In this case, Site-2 cleavage might
have been catalyzed by the fraction of active S2P that
resides in the ER, or the cleaved intermediate form of
SREBP might have left the ER and reached the Golgi
compartment where S2P resides.
Figure 8. A Model Illustrating the Requirement of SCAP for Cleavage of SREBPs
(A) Wild-type cells. In the absence of sterols, SCAP transports SREBPs to the Golgi where S1P cleaves SREBPs. Transport of SCAP is blocked
in the presence of sterols.
(B) SRD-13A cells (SCAP-deficient cells). In the absence of SCAP, SREBPs cannot gain access to active S1P, thus preventing cleavage (left
panel). Treatment of cells with brefeldin A translocates S1P from the Golgi to the ER, restoring cleavage of SREBPs (middle panel). Retrieval
of active, soluble S1P from the Golgi to the ER by the KDEL receptor also restores cleavage of SREBPs (right panel).
If active S1P and S2P are both located in the same
compartment, this compartment is in or near the Golgi,
as indicated by immunofluorescence studies (Figure 3).
This compartment has the properties of the cis- or me-
dial-Golgistacks inthatitcollapses into theER following
treatment with brefeldin A (Figure 4). For this reason,
we refer to this compartment as the Golgi. However,
this compartmentmay notbe composed ofclassic Golgi
stacks. It might be part of the transitional network of
tubules and vesicles that exists between the ER and
partment awaits the development of antibodies capable
of visualizing S2P in nontransfected cells that express
physiologic amounts of this protein. Cell fractionation
studies and ultrastructural analysis by electron micros-
copy are also required.
Certain parallels exist between the role of SCAP in
SREBP processing and the role played by presenilin in
the processing of the amyloid precursor protein (APP),
which is the only other animal cell protein that is clearly
shown to be cleaved within a transmembrane region
(Selkoe, 1996). Presenilin, like SCAP, is a polytopic
membrane protein that is required for proteolytic pro-
cessing of APP (Haass and De Strooper, 1999). Some
(Weidemann et al., 1997; Xia et al., 1997), but not all
(Thinakaran et al., 1998) studies show that presenilin
forms a complex withfull-lengthAPP priorto proteolytic
cleavage. In addition, presenilin is necessary in order
for APP to be transported to the cellular compartment
where the ?-secretase resides (Naruse et al., 1998).
These findings have direct parallels with SCAP. A corol-
lary of this hypothesis is that the absence of presenilin
prevents APP from reaching the compartment where
the ?-secretase resides (Naruse et al., 1998). This is
analogous to the fate of SREBP in SCAP-deficient cells.
AlthoughSCAP and presenilinsharecertainfunctions,
the proteins have no sequence resemblance. Moreover,
they facilitate different reactions. SCAP facilitates the
first cleavage of SREBP, which occurs in the luminal
domain of SREBP. Presenilin facilitates the second
cleavage of APP, namely the intramembrane cleavage
thatfollows the luminalcleavage. Indirect evidence sug-
gests that presenilin may be an aspartyl protease that
cleaves APP at the ? site (Wolfe et al., 1999). Like Site-2
in SREBP, the ? cleavage site of APP is located within a
transmembrane domain. However, there is no evidence
that SCAP is a protease, and indeed we have isolated
a protein with the characteristics of a hydrophobic zinc
metalloprotease that appears to carry out Site-2 cleav-
age of SREBPs (Rawson et al., 1997).
In a more general sense, the results with SCAP and
ular transport in carrying membrane proteins to their
sites of processing. The postulated transport roles of
SCAP and presenilin raise the possibility that polytopic
membrane proteins may function as escort molecules
that carry specific proteins to their sites of processing.
Schekman and colleagues have postulated the exis-
tence ofthese molecularsorters inyeast(Springeret al.,
1999). It will be important to determine whether escort
proteins like SCAP or presenilin target proteins to dedi-
cated vesicles, or whether they simply select individual
proteins for incorporation into a single set of common
carrier vesicles that are targeted to a particular organ-
elle. This situation would be analogous to the function
of cell surface clathrin-coated pits, which are common
carriers whose cargo is selected by sorting molecules.
We obtained monoclonalanti-HSV-Tag (IgG1)fromNovagen, mono-
clonal anti-Myc (clone 9E10) from Roche Molecular Biochemicals,
horseradish peroxidase-conjugated donkey anti-mouse and anti-
rabbit IgG (affinity-purified) from J ackson Immunoresearch Labora-
tories, horseradishperoxidase-conjugateddonkey anti-rabbitwhole
antibody from Amersham Pharmacia Biotech, glycosidases from
New England Biolabs, and brefeldin A from Calbiochem. The follow-
ing recombinant expression plasmids were previously described:
pTK-HSV-BP2 (WT and R519A), encoding wild-type and mutant
HSV-tagged humanSREBP-2(Hua etal., 1996b);pTK-SCAP, encod-
ing hamster SCAP (Nohturfft et al., 1998b); and pCMV-Myc-S1P,
pCMV-S1P-Myc, and pCMV-S1P, encoding hamster S1P (Espen-
shade et al., 1999). Other reagents were obtained from sources as
described previously (Wang et al., 1994). Newborn and fetal calf
lipoprotein-deficient sera (d ? 1.215 mg/ml) were prepared as de-
scribed (Goldstein et al., 1983).
Cells were maintained in monolayer culture at 37?C in 8%-9% CO2.
CHO-7 cells are a clone of CHO-K1 cells selected for growth in
lipoprotein-deficient serum (Metherall et al., 1989). CHO/pS2P cells
(Rawson et al., 1998) are a clone of CHO-7 cells stably transfected
with pCMV-HSV-S2P, a plasmid that encodes human Site-2 prote-
ase undercontrolofthe cytomegalovirus (CMV)promoterenhancer.
SRD-12B cells (deficient inSite-1 protease)and SRD-13A cells (defi-
cientinSCAP)are previously described cholesteroland unsaturated
fatty acid auxotrophs derived from ?-irradiated CHO/pS2P cells
(Rawson et al., 1998, 1999). In experiments using SRD-12B or SRD-
13A cells, the parentalline, CHO/pS2P, was used as a control. Stock
cultures of CHO-7 cells were maintained in medium A (1:1 mixture
of Ham's F12 medium and Dulbecco's modified Eagle medium con-
taining 100 U/ml penicillin and 100 ?g/ml streptomycin sulfate) sup-
plemented with 5% (v/v) newborn or fetal calf lipoprotein-deficient
serum. Stock cultures of CHO/pS2P cells were maintained in me-
diumA supplemented with5% fetalcalflipoprotein-deficientserum,
2 ?M compactin, and 500 ?g/ml G418. Stock cultures of SRD-12B
and SRD-13A cells were maintained inmediumB (mediumA supple-
mented with 5% (v/v) fetal calf serum, 5 ?g/ml cholesterol, 1 mM
sodium mevalonate, and 20 ?M sodium oleate).
Construction of pCMV-S1P(1-997)Myc-KDEL
The expressionvectorpCMV-S1P(1-997)Myc-KDXX encodes amino
acids 1-997 of hamster S1P followed by three tandem copies of the
c-Myc epitope tag (GGRSEQKLISEEDLNGEQKLISEEDLNGEQKLI
SEEDLNSSGR)andtheaminoacids KDEL orKDAS underthecontrol
of the CMV promoter enhancer. To generate these plasmids, we
used the QuikChange site-directed mutagenesis kit (Stratagene)
to mutagenize pCMV-S1P (Sakai et al., 1998b) using the following
GACGAGCTGTGATAAGTGGGCCAGACCATCCC-3? for the KDEL
plasmid and 5?-GCCGCTACAACCAAGAGGGCGGCCGCAAGGAC
GCCTCCTGATAAGTGGGCCAGACCATCCC-3? for the KDAS plas-
mid. Following mutagenesis, a 120 bp fragmentflanked by NotIsites
encoding three tandem copies of the c-Myc epitope (Sakai et al.,
1998b)was inserted into the NotIsite ofthe newly created plasmids.
Inall plasmid constructions, mutations were confirmed by sequenc-
ing the relevant regions.
Stable Transfection of SRD-12B Cells
On day 0, cholesterol auxotrophic SRD-12B cells were plated at a
density of 4 ? 105cells/60 mm dish in medium B. On day 1, cells
were transfected with 0.5 ?g of either pCMV-S1P(1-997)Myc-KDEL
or pCMV-S1P(1-997)Myc-KDAS plus 2.5 ?g of pcDNA3 (Invitrogen)
Transport of SREBPs from ER to Golgi
per dish using the lipofectamine method as described previously
(Sakai et al., 1998b). On day 2, the medium was switched to medium
A supplemented with 5% fetal calf lipoprotein-deficient serum (con-
taining no added cholesterol). The medium was changed as needed
for 12 days until individual colonies were visible. Stable expression
of either S1P(1-997)Myc-KDEL or S1P(1-997)Myc-KDAS permitted
growth of SRD-12B cells in the absence of exogenous cholesterol.
Single-cell clones that stably expressed S1P(1-997)Myc-KDEL and
S1P(1-997)Myc-KDASwere isolated by
screened for S1P expression by immunoblotting with anti-Myc
(clone 9E10)monoclonalantibody. Two celllines expressing equiva-
lent levels of S1P(1-997)Myc-KDEL and S1P(1-997)Myc-KDAS were
selected for further studies and designated S1P-KDEL and S1P-
KDAS cells, respectively.
For experiments using SRD-12B stable cell lines, cells were set
up on day 0 at a density of 7 ? 105cells/10 cm dish in medium B.
On day 1, the medium was switched to medium A supplemented
and 50 ?M sodiummevalonate inthe absence orpresence ofsterols
(1 ?g/ml 25-hydroxycholesterol plus 10 ?g/ml cholesterol added in
a final concentrationof0.2% ethanol). After15 hr, N-acetyl-leucinal-
leucinal norleucinal (ALLN) was added to a final concentration of
25 ?g/mlfor1 hrpriorto harvest. To prepare whole-cellextracts, cell
monolayers werewashed threetimes incold Dulbecco's phosphate-
buffered saline containing 1 mM sodium EDTA and 1 mM sodium
EGTA. To lyse the cells, 0.25 ml of SDS lysis buffer containing 1?
protease inhibitor cocktail (1 mM dithiothreitol, 1 mM PMSF, 0.5
mM Pefabloc, 10 ?g/ml leupeptin, 5 ?g/ml pepstatin A, 25 ?g/ml
ALLN, and 10 ?g/ml aprotinin) was added to each dish. Cell lysate
was passed five times through a 22.5-gauge needle and five times
though a 25-gauge needle. Whole-cell lysates were mixed with 5?
SDS loading bufferand boiled for5 minpriorto SDS-PAGE. Aliquots
of the culture medium were collected, precipitated with acetone,
and resuspended in SDS lysis buffer as previously described (Es-
penshade et al., 1999).
the SuperSignal CL-HRP substrate system (Pierce) according to the
manufacturer's instructions. Gels were calibrated with prestained
molecularmass markers (Bio-Rad). Filters were exposed to X-Omat
Blue XB-1 film (Kodak) at room temperature forthe indicated times.
Glycosidase Sensitivity of S1P
SRD-12B and CHO/pS2P cells were cultured in medium A supple-
mented with 5% newborn calf lipoprotein-deficient serum, 50 ?M
compactin, and 50 ?M mevalonate in the absence or presence of
sterols (1 ?g/ml 25-hydroxycholesterol plus 10 ?g/ml cholesterol
added in a final concentration of 0.2% ethanol). After 16 hr, ALLN
was added to a final concentration of 25 ?g/ml, and the cells were
harvested 1 hr later. Membrane and nuclear extract fractions were
prepared as described above. Aliquots of solubilized membrane
proteins (15 ?g) were treated with either peptide N-glycosidase F
or endoglycosidase H (endo H) as described previously (Sakai et
Indirect Immunofluorescence Microscopy
CHO/pS2P and SRD-12B cells were set up on glass coverslips in
medium B. Cells were washed in buffer C containing 0.1 M sodium
phosphate at pH 7.4, 0.15 M NaCl, 4 mM KCl, 2 mM MgCl2, and
0.02% (w/v) sodium azide and then fixed and permeabilized by
incubating for 10 min in solution containing 60% (v/v) methanol,
10% (v/v) glacial acetic acid, and 30% (v/v) 1,1,1-trichloroethane.
After a brief rinse with buffer C, cells were incubated for 30 min at
room temperature in blocking buffer (0.1 M Tris±HCl at pH 9.0, 1%
[w/v] bovine serum albumin, 0.1 M NaCl, 0.02% [w/v] NaN3). Cells
were incubated at4?C overnightwithaffinity-purified anti-S1P rabbit
polyclonalantibody (20 ?g/mlinblocking buffer).Primary antibodies
were localized by incubating cells for 2 hr in 20 ?g/ml affinity-puri-
fied, goatanti-rabbitIgG conjugated toAlexa 488(MolecularProbes,
Inc.)in blocking buffer. ForGolgi compartment staining, rhodamine-
labeled wheatgermagglutinin(VectorLaboratories, Inc.)was added
during the second antibody incubation at a concentration of 5 ?g/
ml. Cells were washed three times after each antibody or lectin
incubation and analyzed with a Leica TCS SP confocal microscope.
S1P-specific antibodies were isolated from rabbit serum by affinity
chromatography using a cross-linked agarose resin conjugated to
the two immunizing synthetic peptides (Espenshade et al., 1999;
Harlow and Lane, 1999). Peptides were conjugated to SulfoLink
Coupling Gel (Pierce) according to the manufacturer's instructions.
For the brefeldin A experiment, cells were treated with brefeldin A
at a final concentration of 2 ?g/ml in 0.2% methanol for 1 hr prior
to fixation of cells.
Immunoblot Analysis of SREBP Processing
On day 0, SRD-12B and SRD-13A cells were set up at the indicated
density in medium B in 60 mm dishes. On day 1, cells were trans-
fected with4 ?g ofDNA/dishusing 12 ?lofFugene 6 reagent(Roche
Molecular Biochemicals) in a final volume of 0.2 ml. Transfection
was performed as previously described (Rawson et al., 1999). Cells
were incubated at 37?C for 16±24 hr in medium A supplemented
with 5% fetal calf serum. On day 2, the medium was removed, and
the cells were cultured inmediumA supplemented with5% newborn
calf lipoprotein-deficient serum, 50 ?M compactin, and 50 ?M so-
dium mevalonate in the absence or presence of sterols (1 ?g/ml
25-hydroxycholesterol plus 10 ?g/ml cholesterol added in a final
concentrationof0.2% ethanol).Afterincubationat37?C for16hr, the
cells received ALLN at a finalconcentration of 25 ?g/ml. Following a
1 hr incubation, the cells were harvested and processed as pre-
viously described (Sakai et al., 1996) with minor modifications. Har-
vested cells were resuspended in 0.4 ml of bufferA (10 mM HEPES-
KOH at pH 7.4, 10 mM KCl, 1.5 mM MgCl2, 1 mM sodium EDTA, 1
mM sodium EGTA, 250 mM sucrose), passed through a 22.5-gauge
needle 30 times, and centrifuged at 1000 g for 7 min at 4?C. The
1000 g pellet was resuspended in 0.1 ml of bufferB (20 mM HEPES-
KOH at pH 7.6, 2.5% (v/v) glycerol, 0.42 M NaCl, 1.5 mM MgCl2, 1
mM sodium EDTA, 1 mM sodium EGTA), rotated at 4?C for 1 hr, and
centrifuged at 105g for 30 min at 4?C in a Beckman TLA 100.2 rotor.
Fractionation buffers contained 1? protease inhibitor cocktail. The
supernatant from this spin was designated the nuclear extract. The
supernatant from the original 1000 g spin was used to prepare the
membrane fraction by centrifugation at 2 ? 104g for 15 min at 4?C.
The resulting membrane pellets were resuspended in 0.1 ml SDS-
lysis buffer (Espenshade et al., 1999).
Protein concentration in nuclear extract and membrane fractions
was measured using the BCA Kit (Pierce), and the samples were
mixed with5? SDS loading buffer(Bollag and Edelstein, 1991). After
boiling for 5 min, the proteins were subjected to SDS-PAGE and
transferred to Hybond C-Extra nitrocellulose filters (Amersham). The
filters were incubated with the antibodies described in the figure
legends. Bound antibodies were visualized with peroxidase-conju-
gated affinity-purified donkey anti-mouse or anti-rabbit IgG using
We thank Tammy Dinh for excellent technical assistance and Lisa
Beatty and Anna Fuller for invaluable help with tissue culture. This
work was supported by grants from the National Institutes of Health
(NIH) (HL-20948) and the Perot Family Foundation. R. A. D.-B. is the
recipient of a postdoctoral fellowship for the J ane Coffin Childs
Memorial Fund. P. J . E. is the recipient of an NIH Research Science
Fellowship Grant (HL-09993).
Received November 9, 1999; revised December 3, 1999.
Bollag, D.M., and Edelstein, S.J . (1991). Protein Methods (New York:
Wiley-Liss, Inc.), pp. 1±230.
Brown, M.S., and Goldstein, J .L. (1997). The SREBP pathway: regu-
lation of cholesterol metabolism by proteolysis of a membrane-
bound transcription factor. Cell 89, 331±340.
Brown, M.S., and Goldstein, J .L. (1999). A proteolytic pathway that
controls the cholesterol content of membranes, cells, and blood.
Proc. Natl. Acad. Sci. USA 96, 11041±11048.
Cheng, D., Espenshade, P.J ., Slaughter, C.A., J aen, J .C., Brown,
M.S., and Goldstein, J .L. (1999). Secreted Site-1 protease cleaves
peptides corresponding to luminalloop ofsterolregulatory element-
binding proteins. J . Biol. Chem. 274, 22805±22812.
Cell Download full-text
Duncan, E.A., Brown, M.S., Goldstein, J .L., and Sakai, J . (1997).
Cleavage site for sterol-regulated protease localized to a Leu-Ser
bond in lumenal loop of sterol regulatory element binding protein-2.
J . Biol. Chem. 272, 12778±12785.
Edwards, P.A., and Ericsson, J . (1999). Sterols and isoprenoids:
signaling molecules derived from the cholesterol biosynthetic path-
way. Annu. Rev. Biochem. 68, 157±185.
Espenshade, P.J ., Cheng, D., Goldstein, J .L., and Brown, M.S.
(1999).Autocatalytic processing ofSite-1proteaseremoves propep-
tide and permits cleavage of sterol regulatory element-binding pro-
teins. J . Biol. Chem. 274, 22795±22804.
Goldstein, J .L., Basu, S.K., and Brown, M.S. (1983). Receptor-medi-
ated endocytosis of LDL in cultured cells. Methods Enzymol. 98,
Haass, C., and De Strooper, B. (1999). The presenilins inAlzheimer's
diseaseÐ proteolysis holds the key. Science 286, 916±919.
Harlow, E., and Lane, D. (1999). Using Antibodies: A Laboratory
Manual (New York: Cold Spring Harbor), pp. 1±495.
Hua, X.,Nohturfft, A., Goldstein,J .L., and Brown, M.S.(1996a).Sterol
resistance in CHO cells traced to point mutation in SREBP cleavage
activating protein (SCAP). Cell 87, 415±426.
Hua, X., Sakai, J ., Brown, M.S., and Goldstein, J .L. (1996b). Regu-
lated cleavage of sterol regulatory element binding proteins
(SREBPs) requires sequences on both sides of the endoplasmic
reticulum membrane. J . Biol. Chem. 271, 10379±10384.
Kornfeld, R., and Kornfeld, S. (1985). Assembly of asparagine-linked
oligosaccharides. Annu. Rev. Biochem. 54, 631±664.
Lippincott-Schwartz, J ., Yuan, L.C., Bonifacino, J .S., and Klausner,
R.D. (1989). Rapid redistribution of Golgi proteins into the ER in
cells treated with brefeldin A: evidence for membrane cycling from
Golgi to ER. Cell 56, 801±813.
Metherall, J .E., Goldstein, J .L., Luskey, K.L., and Brown, M.S. (1989).
Loss of transcriptional repression of three sterol-regulated genes
in mutant hamster cells. J . Biol. Chem. 264, 15634±15641.
Munro, S., and Pelham, H.R.B. (1987). A C-terminal signal prevents
secretion of luminal ER proteins. Cell 48, 899±907.
Naruse, S., Thinakaran, G., Luo, J .-J ., Kusiak, J .W., Tomita, T., Iwat-
subo, T., Qian, X., Ginty, D.D., Price, D.L., Borchelt, D.R., et al.
(1998). Effects of PS1 deficiency on membrane protein trafficking
in neurons. Neuron 21, 1213±1221.
Nohturfft, A., Hua, X., Brown, M.S., and Goldstein, J .L. (1996). Recur-
rent G-to-A substitution in a single codon of SREBP cleavage-acti-
vating protein causes sterol resistance in three mutant CHO cell
lines. Proc. Natl. Acad. Sci. USA 93, 13709±13714.
Nohturfft, A., Brown, M.S., and Goldstein, J .L. (1998a). Topology of
SREBP cleavage-activating protein, a polytopic membrane protein
with a sterol-sensing domain. J . Biol. Chem. 273, 17243±17250.
Nohturfft, A., Brown, M.S., and Goldstein, J .L. (1998b). Sterols regu-
late processing of carbohydrate chains of wild-type SREBP cleav-
age-activating protein (SCAP), but not sterol-resistant mutants
Y298C or D443N. Proc. Natl. Acad. Sci. USA 95, 12848±12853.
Nohturfft, A., DeBose-Boyd, R.A., Scheek, S., Goldstein, J .L., and
Brown, M.S. (1999). Sterols regulate cycling of SREBP cleavage-
activating protein(SCAP)betweenendoplasmic reticulumandGolgi.
Proc. Natl. Acad. Sci. USA 96, 11235±11240.
Pelham, H.R.B. (1992). The secretion of proteins by cells. Proc. R.
Soc. Med. 250, 1±10.
Rawson, R.B., Zelenski, N.G., Nijhawan, D., Ye, J ., Sakai, J ., Hasan,
M.T., Chang, T.-Y., Brown, M.S., and Goldstein, J .L. (1997). Comple-
mentation cloning of S2P, a gene encoding a putative metallopro-
tease required for intramembrane cleavage of SREBPs. Mol. Cell 1,
Rawson, R.B., Cheng, D., Brown, M.S., and Goldstein, J .L. (1998).
Isolation of cholesterol-requiring mutant CHO cells with defects in
cleavage of sterol regulatory element binding proteins at Site-1. J .
Biol. Chem. 273, 28261±28269.
Rawson, R.B., DeBose-Boyd, R., Goldstein, J .L., and Brown, M.S.
(1999). Failure to cleave sterol regulatory element-binding proteins
(SREBPs) causes cholesterol auxotrophy in Chinese hamster ovary
cells with genetic absence of SREBP cleavage-activating protein.
J . Biol. Chem. 274, 28549±28556.
Ridgway, N.D., and Lagace, T.A. (1995). BrefeldinA renders Chinese
hamsterovary cells insensitive to transcriptionalsuppressionby 25-
hydroxycholesterol. J . Biol. Chem. 270, 8023±8031.
Sakai, J ., Duncan, E.A., Rawson, R.B., Hua, X., Brown, M.S., and
Goldstein, J .L. (1996). Sterol-regulated release of SREBP-2 from
cell membranes requires two sequential cleavages, one within a
transmembrane segment. Cell 85, 1037±1046.
Sakai, J ., Nohturfft, A., Cheng, D., Ho, Y.K., Brown, M.S., and
Goldstein, J .L. (1997). Identification of complexes between the
COOH-terminal domains of sterol regulatory element binding pro-
teins (SREBPs) and SREBP cleavage-activating protein (SCAP). J .
Biol. Chem. 272, 20213±20221.
Sakai, J ., Nohturfft, A., Goldstein, J .L., and Brown, M.S. (1998a).
Cleavage of sterol regulatory element binding proteins (SREBPs) at
site-1 requires interaction with SREBP cleavage-activating protein.
Evidence frominvivo competitionstudies. J . Biol. Chem. 273, 5785±
Sakai, J ., Rawson, R.B., Espenshade, P.J ., Cheng, D., Seegmiller,
A.C., Goldstein, J .L., and Brown, M.S. (1998b). Molecular identifica-
tion of the sterol-regulated luminal protease that cleaves SREBPs
and controls lipid composition of animal cells. Mol. Cell 2, 505±514.
Selkoe, D.J . (1996). Amyloid ?-protein and the genetics of Alzhei-
mer's disease. J . Biol. Chem. 271, 18295±19298.
Springer, S., Spang, A., and Schekman,R.(1999).A primeronvesicle
budding. Cell 97, 145±148.
Thinakaran, G., Regard, J .B., Bouton, C.M.L., Harris, C.L., Price,
D.L., Borchelt, D.R., and Sisodia, S.S. (1998). Stable association of
presenilin derivatives and absence of presenilin interactions with
APP. Neurobiol. Dis. 4, 438±453.
Wang, X., Sato, R., Brown, M.S., Hua, X., and Goldstein, J .L. (1994).
SREBP-1, a membrane-bound transcription factor released by ste-
rol-regulated proteolysis. Cell 77, 53±62.
Weidemann, A., Paliga, K., Durrwang, U., Czech, C., Evin, G., Mas-
ters, C.L., and Beyreuther, K. (1997). Formation of stable complexes
between two Alzheimer's disease gene products: presenilin-2 and
?-amyloid precursor protein. Nat. Med. 3, 328±332.
Wolfe, M.S., Xia, W., Ostaszewski, B.L., Diehl, T.S., Kimberly, W.T.,
and Selkoe, D.J . (1999). Two transmembrane aspartates in preseni-
lin-1 required for presenilin endoproteolysis and ?-secretase activ-
ity. Nature 398, 513±517.
Xia, W., Zhang, J ., Perez, R., Koo, E.H., and Selkoe, D.J . (1997).
Interaction between amyloid precursor protein and presenilins in
mammalian cells: implications for the pathogenesis of Alzheimer
disease. Proc. Natl. Acad. Sci. USA 94, 8208±8213.