| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 275, Issue 35, 27013-27020, September 1, 2000
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,,From the Department of Biological Sciences, Columbia University, New York, New York 10027 and the § Department of Biological Chemistry and Howard Hughes Medical Institute, University of Michigan Medical Center, Ann Arbor, Michigan 48109
Received for publication, April 18, 2000, and in revised form, June 13, 2000
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
ATF6 is a member of the basic-leucine zipper
family of transcription factors. It contains a transmembrane domain and
is located in membranes of the endoplasmic reticulum. ATF6 has been
implicated in the endoplasmic reticulum (ER) stress response pathway
since it can activate expression of GRP78 and other
genes induced by the ER stress response. ER stress appears to activate
ATF6 by cleavage from the ER membrane and translocation to the nucleus. However, direct DNA binding by ATF6 had not been demonstrated. In this
report, we have identified a consensus DNA binding sequence for ATF6.
This site is related to but distinct from ATF1/CREB binding sites. The
site was placed in a reporter gene and was specifically activated by
ATF6 overexpression and was strongly induced by the ER stress response.
A dominant negative form of ATF6 blocked ER stress induction of both
ATF6 site and GRP78 reporter genes. We further found that
GAL4-ATF6 could be activated by ER stress. These results demonstrate
that ATF6 is a direct target of the ER stress response. A proximal
sensor of the ER stress response, human IRE1 (hIRE1), was
sufficient to activate the ATF6 reporter gene, while a dominant
negative form of hIRE1 blocked ER stress activation, suggesting that
hIRE1 is upstream of ATF6 in the ER stress signaling pathway.
The endoplasmic reticulum
(ER)1 stress response is a
mechanism by which cells protect themselves from many noxious
insults that cause protein unfolding in the ER (reviewed in
Ref. 1). Inducers of this response include inhibitors of
glycosylation (tunicamycin), dithiothreitol, agents that affect calcium
homeostasis such as calcium ionophores and thapsigargin (an inhibitor
of an ER calcium-ATPase), and agents that perturb ER function and
protein movement such as brefeldin A.
The ER stress response (also known as the unfolded protein response)
causes an increase in gene expression of a number of ER chaperones,
such as GRP78/BiP and GRP94,
and erp72, which is related to protein-disulfide isomerase
(1). These factors probably allow the cell to refold unfolded proteins
in the ER. Analysis of the promoters of these genes has revealed a
consensus sequence element, CCAATN9CCACG, that is required
for ER stress induction of the promoters (2, 3). Mutation of this ER
stress response element (ERSE) has shown that both the 5' CCAAT and 3'
CCACG boxes are required. The variable central region was not sensitive
to point mutations; however, some multiple base changes abolished activity (2, 3).
Not surprisingly, NF-Y/CBP binds to the 5' CCAAT box of the ERSE (4,
5); however, it has been less clear which factor(s) bind the element to
mediate the ER stress response. In order to identify factors binding
the ERSE, Yoshida et al. (2) utilized a yeast one-hybrid
approach and found ATF6. While they were not able to show that ATF6
bound the ERSE directly, overexpression of ATF6 activated
GRP78 reporter genes in mammalian cells in an ERSE
site-dependent manner. They suggested that ATF6 can
interact with the ERSE directly or indirectly in mammalian cells.
ATF6 is a member of the ATF/CREB basic-leucine zipper (bZIP)
DNA-binding protein family (6). It is a 90-kDa protein with 670 amino
acids (7). ATF6 contains a transmembrane domain at amino acids 378-398
with the N terminus facing the cytoplasm (8). By immunofluorescence and
cell fractionation, Haze et al. (8) found that ATF6 resides
in the ER. They were also able to identify an ER stress-induced 50-kDa
cleavage product of ATF6, which is proposed to be an active form of the
factor. Consistent with this model, the 50-kDa protein was found in
nuclear fractions. In addition, expression of the cytoplasmic domain of
ATF6 (aa 1-373) resulted in a protein that localized entirely in the
nucleus and that activated expression of the GRP78 gene (8).
We identified ATF6 in a yeast two hybrid-like screen interacting with
the serum response factor, a transcription factor that controls
growth factor induced expression of the c-fos gene (7). It
is thus possible that ATF6 activation may affect other targets besides
ERSE-containing genes.
In order to characterize the function and mechanism of activation of
ATF6, we describe here the identification of a consensus DNA binding
site for ATF6. This site is activated by overexpression of ATF6 and by
ER stress inducers. In addition, a GAL4-ATF6 fusion protein is
regulated by ER stress, demonstrating that ATF6 can respond to this signal.
Binding Site Selection and Gel Mobility Shift Assays
Bacterial expression vector pET28bATF6bZIP was constructed by
PCR amplification of the region of ATF6 encoding amino acids 287-432.
This fragment was cloned into BamHI to XhoI
sites of pET28b (Novagen) downstream of a polyhistidine tag. ATF6bZIP
protein was expressed in Escherichia coli strain BLR
(Novagen) from the pET28bATF6bzip vector and purified on
Ni2+-agarose (Invitrogen) as described (9).
Selection of ATF6bZIP binding to random oligonucleotides was done by
gel mobility shift assay essentially as described (10). Briefly, the
reaction conditions were 10 mM Tris-HCl, pH 7.5, 2 µg of
herring sperm DNA, 1 mM DTT, and 20 ng of ATF6bZIP protein
in 20 µl. This mixture was incubated with 1 ng of
32P-labeled N15 probe for 1 h at room temperature and
loaded on a 4% polyacrylamide gel in 0.25× TBE (25 mM
Trizma base, 25 mM boric acid, 1 mM EDTA). The
N15 probe was
5'-TACGGATCCCTACAGGTGCN15GCAATCCAGGAATTCGT-3', where the central 15 nucleotides were random. It was
annealed to PCR primer 2, 5'-ACGAATTCCTGGATTCG-3', and extended with
the Klenow fragment of E. coli DNA polymerase. The
double-stranded oligonucleotide was then labeled with
[ The following double-stranded oligonucleotides were synthesized for
binding experiments with ATF6bZIP and ATF1.
INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
EXPERIMENTAL PROCEDURES
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
-32P]ATP and T4 polynucleotide kinase. The position
of the ATF6bZIP-DNA complex was estimated by comparison with the low
affinity binding of the protein to a similarly labeled probe containing
the c-fos AP1 site in the presence of 0.5 µg of herring
sperm DNA. This position of the gel was excised from the gel, and the
protein-bound DNA was soaked out of the gel slice in 10 mM
Tris-HCl, pH 8.0, 1 mM EDTA. The excised DNA was
PCR-amplified using PCR primer 2 and PCR primer 1 (5'-TACGGATCCCTACAGGTGC-3'). The PCR product was
32P-labeled and used for the next round of selection. This
process was repeated for seven cycles. The PCR product from the sixth round of selection was subcloned into the BamHI and
EcoRI sites of pBluescript II SK+ and sequenced.
The nucleotide mutated in ATF6m1 is underlined. Full-length ATF1
with a polyhistidine tag at the N terminus was cloned into pET28b at
the NcoI and SalI sites. Histidine-tagged ATF1
was purified from E. coli on Ni2+-agarose
as described (9). Purified ATF6bZIP and ATF1 proteins were assayed for
binding to the 32P-labeled oligonucleotides by gel mobility
shift assay as described above. For competitions in Fig. 3A,
increasing amounts of the ATF6 oligonucleotide were added or, as a
control, 100 ng of oligonucleotide RP4
(5'-TCGAGAAGCGCCCAGGCCCGCGCGCA-3').
Plasmids
Luciferase Reporter Genes-- Oligonucleotides ATF6 and ATF6m1 containing one of the selected ATF6 sites and a single base mutation (underlined) were used in reporter genes. Their sequences are as follows.
|
|
|
53 to +45 of the human
c-fos promoter) and the firefly luciferase gene. Plasmid
p5×ATF6GL3 contains five repeats of the ATF6 oligonucleotide, while
pATF6GL3 and pATF6m1GL3 contain one copy each of their respective
sites. The p4×CRE-luciferase reporter was from Stratagene. The
p5×GAL4-E1b-luc reporter contains five GAL4 sites upstream of the
adenovirus E1b minimal promoter and the firefly luciferase gene. The
rat GRP78-luciferase reporter gene with 304 to +7 of the
rat GRP78 promoter was as described (2).
Expression Vectors--
pCGNATF6 contains full-length ATF6 (aa
1-670) with an HA epitope tag at the N terminus as described (7).
pCGNATF6-(1-373) contains amino acids 1-373 of ATF6 (generated
by PCR amplification) in the XbaI-BamHI sites of
pCGN (12). pCGNATF6-(1-373)m1 was made from pCGNATF6-(1-373) with the
QuikChange site-directed mutagenesis kit (Stratagene). Amino acids
315-317 were changed from KNR to TAA. pSV40Gal4ATF6 contains the SV40
early promoter driving expression of GAL4's DNA binding domain (aa
1-147) fused to full-length ATF6 (aa 1-670). The expression vector
for hIRE1
, pED-hIRE1, was as described (13). The dominant negative
form of hIRE1 with the deletion of the C-terminal kinase and
nuclease domains, pED-hIRE1
C, contains the coding region for amino
acids 1-492 of hIRE1. The cAMP-dependent protein kinase
expression vector, pFC-PKA, was from Stratagene.
Transfections and Luciferase Assays
HeLa cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. Cells were transfected by the calcium phosphate coprecipitation method (14). Each transfection mixture for a 60-mm diameter plate contained 0.5 µg of luciferase reporter gene, 0.1 µg of pRLSV40P as an internal control (15), the indicated amounts of expression constructs, and herring sperm DNA to give a total of 10 µg of DNA. HeLa cells were transfected for 16-20 h and then induced with 2 µg/ml tunicamycin for 12 h. The transfected cells were lysed and assayed for firefly and Renilla luciferase activity using the dual luciferase kit (Promega). The results were normalized to the Renilla luciferase activity of the internal control. Each experiment was repeated four or more times. The average and the S.E. are shown.
Immunoblotting
Cell lysates were prepared from transfected HeLa cells by
resuspending the cells on the plate in 0.2 ml of 3× SDS sample buffer (6% SDS, 180 mM Tris-HCl (pH 6.8), 30% glycerol, 0.003%
bromphenol blue, 1% 
-mercaptoethanol) and boiled for 5 min. The
lysates were analyzed by immunoblotting using a 1:1000 dilution of
anti-HA monoclonal antibody (Babco). Horseradish peroxidase-conjugated goat anti-mouse IgG (Sigma) was used as a secondary antibody at a
1:4000 dilution, and the signal was visualized using the ECL chemiluminescence kit (Amersham Pharmacia Biotech).
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
Selection of an ATF6 DNA Binding Site-- Since ATF6 is a bZIP protein of the ATF/CREB family, it was likely that it could bind directly to DNA. We expressed a fragment of ATF6 (aa 287-432) spanning the bZIP domain with a polyhistidine tag. We found that the purified protein was able to bind weakly to ATF-related sites such as the c-fos FAP1 site, which is immediately 3' to the c-fos SRE (data not shown). Binding was readily apparent using 0.5 µg of herring sperm DNA as nonspecific competitor in the gel mobility shift assay but was not observed under more stringent conditions where we used 2 µg of herring sperm DNA. The binding was nearly undetectable under conditions where similar amounts of ATF1 bound a consensus ATF1/CREB site well (data not shown).
We decided to look for higher affinity ATF6 DNA binding sites by
selecting from a pool of random oligonucleotides by gel mobility shift
assays (10). We used a random oligonucleotide with 15 central random
nucleotides and fixed ends allowing it to be amplified by PCR. We mixed
the 32P-labeled random oligonucleotides with the purified
ATF6bZIP protein in a gel mobility assay using the more stringent
conditions of 2 µg of herring sperm DNA. The region of the gel where
ATF6bZIP bound to the FAP1 site was excised, and the DNA was isolated
and amplified by PCR. This amplified product was
32P-labeled and used in another round of gel mobility shift
assay selection. This process was repeated seven times. The results from cycles 2, 6, and 7 are shown in Fig.
1. There was no clear band in cycles
1-4. A faint band was apparent at cycle 5. A strong band was observed
at cycle 6 and was not further increased at cycle 7, suggesting that
further rounds would not select for higher affinity sites (Fig. 1 and
data not shown).
|
We subcloned the PCR products from the sixth round of selection and
sequenced 20 individual clones. These results are shown in Fig.
2. Two sequences were found twice with
all the others being unique. We tested all of the 20 products in a gel
mobility shift assay for binding to ATF6bZIP protein and found they all bound similarly, although it was difficult to compare them
quantitatively, since the amount of each product and the efficiency of
32P labeling varied (data not shown). Notably, all the
subclones contained the sequence TGACGT. The sequence TGACGT is also
the core of ATF1/CREB sites (known as CREs), which have the consensus TGACGTCA (16). We aligned the sequences according to the TGACGT core to
determine a consensus binding site (Fig. 2). This suggests a consensus
of (G)(G)TGACGTG(G/A), where the nucleotides in parentheses are less
strongly maintained. We found that 18 of 20 had a G following the core
sequence, suggesting that this nucleotide is particularly important.
The flanking sequence that was fixed in the design of the random
oligonucleotides may also contribute to binding by ATF6, since in 17 of
20 cases the core TGACGTG sequence was found juxtaposed to the
3'-flanking sequence. We have not included bases from the flanking
sequence in the consensus; however, the final base of the consensus was
G in two of four cases, and G is also the first base in the 3'-flanking
sequence, suggesting a consensus of TGACGTGG. The only two sequences
without GG at the end (sequence 2 in Fig. 2) seem to be exceptions to
the rule and contain a perfect inverted repeat sequence TGACGTCA, which is identical to the ATF1/CREB consensus.
|
Specific DNA Binding by ATF6--
To demonstrate that binding by
ATF6 was specific, we synthesized a double-stranded oligonucleotide
corresponding to sequence 1 in Fig. 2 and will refer to this as the
ATF6 site. This oligonucleotide was 32P-labeled, and strong
binding of ATF6 was observed in a gel mobility shift assay (Fig.
3A). This binding was competed
away by excess unlabeled ATF6 site oligonucleotide but not by 100 ng of
a nonspecific oligonucleotide (Fig. 3A).
|
Since the ATF6 consensus site is similar to CRE sites, we compared
binding of ATF6 to these sites. We used a CRE oligonucleotide corresponding to the 60 CRE from the c-fos promoter, which
contains the sequence TGACGTTT (17). ATF6 bound this site significantly more weakly than the ATF6 site (Fig. 3B, lanes
1 and 3). We also mutated the ATF6 site from
TGACGTGG to TGACGTTG (referred to as the ATF6m1
site). ATF6 bound the ATF6m1 site significantly more weakly than the
ATF6 site (Fig. 3B, lanes 1 and
2), suggesting that the G flanking the TGACGT core is
critical to ATF6-specific binding. In contrast, we found that purified
ATF1 bound all three sites, with stronger binding to the CRE and ATF6m1
sites (Fig. 3B, lanes 4-6).
It is also apparent that ATF1 binding to the CRE site is similar to ATF6 binding to the ATF6 site (Fig. 3B, compare lanes 6 and 1). Since high affinity binding of ATF1 to CRE sites is well characterized (16), this suggests that ATF6 binds its site with high affinity.
We next compared binding of ATF6 and ATF1 to the ATF6 site to further show that that ATF6 binds the site we have identified with high affinity compared with ATF1. A titration of similar amounts of these factors showed that ATF6 binds its site significantly better than ATF1 (Fig. 3C).
Activation of the ATF6 Site by ER Stress and ATF6 in Vivo--
To
check whether the ATF6 consensus DNA binding site can function in
vivo, we constructed a luciferase reporter gene with five ATF6
sites in front of the c-fos minimal promoter. This reporter was transfected into HeLa cells along with a full-length (aa 1-670) ATF6 expression vector. ATF6 strongly activated the 5× ATF6 site reporter (Fig. 4A). There was
no effect of ATF6 on the internal control of the SV40 promoter driving
a Renilla luciferase gene (data not shown). In order to
determine whether the ATF6 sites were targets for the ER stress
response, we treated the transfected cells with or without tunicamycin.
Tunicamycin strongly activated the 5× ATF6-luciferase gene 30-fold
(Fig. 4A). Tunicamycin also further increased ATF6
activation of the reporter, but only by about 2-fold. A reporter gene
lacking the ATF6 sites was not induced by tunicamycin (data not shown).
The internal control of SV40-Renilla luciferase was
consistently reduced by tunicamycin treatment by about 50%, perhaps
due to toxicity, and we have normalized the results to the
Renilla luciferase activity to account for possible cell
loss and variations in transfection efficiencies. We also found that
other ER stress inducers, including thapsigargin and brefeldin A,
induced expression of the 5× ATF6 reporter gene (data not shown).
These results suggest that ATF6 is a general target of agents that
induce the ER stress response and that the ATF6 site is a direct target
of ATF6 in cells.
|
Haze et al. (8) found that the cytoplasmic domain of ATF6 localized to the nucleus and activated GRP78 gene expression. We tested whether a similar ATF6 construct (aa 1-373) lacking the transmembrane and ER luminal domains could activate the 5× ATF6 site reporter. We used increasing amounts of ATF6-(1-373) and full-length ATF6 to compare their ability to activate the reporter. ATF6-(1-373) strongly activated the 5× ATF6 site reporter with as little as 10 ng of plasmid, activating as strongly as 3 µg of full-length ATF6 (Fig. 4B). To correlate this with ATF6 protein expression, we assayed transfected cell lysates by immunoblotting with antisera to the HA epitope tag at the N terminus of the ATF6 constructs. We found that transfection of 30 ng of ATF6-(1-373) gave a similar level of protein expression as 1 µg of full-length ATF6 (Fig. 4C). However, 30 ng of ATF6-(1-373) strongly activated reporter gene expression, while 1 µg of full-length ATF6 had little effect (Fig. 4B). This demonstrates that ATF6-(1-373) is an activated form of ATF6.
We also tested by immunofluorescence whether the ATF6 constructs localized to the ER or the nucleus. Similar to the results of Haze et al. (8), full-length ATF6 was localized outside the nucleus, consistent with ER localization, while ATF6-(1-373) was entirely localized to the nucleus (data not shown).
Haze et al. (8) detected a 50-kDa cleavage product of ATF6 that was induced by tunicamycin treatment. This was observed using antisera to endogenous ATF6. We have not been able to detect this 50-kDa fragment using our anti-ATF6 serum (7), at least partially due to the high background observed with this serum (data not shown). Upon overexpression of ATF6, Haze et al. (8) did not observe tunicamycin-induced ATF6 cleavage but did detect the constitutive presence of the 50-kDa cleavage product using anti-HA epitope sera and the same ATF6 construct as we have used. We have not clearly detected the 50-kDa form of ATF6 using the anti-HA sera (Fig. 4C, lanes 9 and 10). Using 3 µg of ATF6 plasmid, which is the amount that activates reporter gene expression, we do detect a generally higher background (Fig. 4C, lane 10) such that there may be basal cleavage at multiple sites leading to translocation of the ATF6 cytoplasmic domain to the nucleus. While a band is seen near the 50-kDa marker in lane 10, upon longer exposure this band is also present in all of the other lanes. It is possible that a fragment of ATF6 comigrates with the background band at about 55 kDa or that it is simply below the limits of our detection. This is possible, since low levels (3 ng) of ATF6-(1-373) partially activated the reporter gene (Fig. 4B) but were nearly undetectable in the immunoblot (Fig. 4C, lane 3). Further work will be required to resolve this issue (see "Discussion").
Specificity of Activation of the ATF6 Site--
Since the ATF6
site we have identified is similar to the consensus CRE site, we
compared the ability of a 4× CRE luciferase reporter gene (with the
multimerized sequence TGACGTCA) to be induced by tunicamycin and
activated by ATF6. In contrast to the 5× ATF6 reporter, the CRE
reporter was not induced by tunicamycin or by overexpression of
ATF6-(1-373) (Fig. 5A). As a
control, the 4× CRE reporter was strongly activated by the catalytic
subunit of cAMP-dependent protein kinase.
|
The consensus ATF6 site contrasts with CRE sites in that it contains a G 3' to the TGACGT core CRE sequence. We therefore tested the importance of this nucleotide by comparing reporter genes with one copy of the ATF6 site (TGACGTG) or the ATF6m1 mutant site (TGACGTT). In vitro ATF6 bound the ATF6m1 site significantly more weakly than the ATF6 site, while ATF1 bound the ATF6m1 site somewhat better (Fig. 3B). The 1× ATF6 site reporter gene was induced by tunicamycin and ATF6-(1-373), although the activation was weaker than the 5× ATF6 reporter (Fig. 5B). In contrast, the ATF6m1 reporter was neither induced by tunicamycin nor activated by ATF6-(1-373). These results indicate that the activation of the ATF6 consensus site by tunicamycin and ATF6 is specific to the ATF6 consensus site and correlates with binding of ATF6 to the sites in vitro.
ATF6 Is Required for Activation of the ATF6 Consensus DNA Binding
Site--
Since the ATF6 site reporter genes were activated by ER
stress inducers, we sought to test whether endogenous ATF6 is required for this activation using a potentially dominant negative form of ATF6.
We made point mutations in the basic region of ATF6 in the context of
ATF6-(1-373). These mutations (KNR to TAA at aa 315-317) would be
predicted to disrupt DNA binding activity of the cytoplasmic domain. It
would act as a dominant negative if it dimerized with endogenous ATF6
and prevented its binding to ATF6 DNA binding sites. This construct,
ATF6-(1-373)m1, did not activate the 5× ATF6 site reporter, as
expected, and completely inhibited tunicamycin induction of the
reporter (Fig. 6A). There was
no effect of ATF6-(1-373)m1 on the SV40 promoter-Renilla
luciferase internal control. The inhibition of tunicamycin induction of
the ATF6 site by the ATF6 dominant negative mutants strongly suggests that endogenous ATF6 is mediating tunicamycin induction through this
site.
|
To test whether ATF6 is required for ER stress induction of the GRP78 gene, which contains three ERSE elements, we transfected ATF6-(1-373)m1 with a rat GRP78 promoter reporter gene. This reporter was induced 3.5-fold by tunicamycin, and this induction was reduced to 1.6-fold by ATF6-(1-373)m1 (Fig. 6B).
Activation of GAL4-ATF6 by Tunicamycin--
Since levels of
endogenous ATF6 appear to be sufficient to mediate tunicamycin
induction of the ATF6 site, we were not able to show directly by
transfection of ATF6 that it is required for induction of the site. In
order to confirm that ATF6 can be activated by the ER stress response,
we fused it to the GAL4 DNA binding domain and tested whether a 5×
GAL4 site-luciferase reporter gene could then be induced by
tunicamycin. While the GAL4 reporter was not induced by tunicamycin,
cotransfection of increasing amounts of GAL4-ATF6 resulted in
tunicamycin induction of the reporter (Fig.
7). Only the highest amounts of GAL4-ATF6
activated the reporter without tunicamycin induction. This is similar
to the activation we see with full-length ATF6 (Fig. 4B) and
may represent basal cleavage of the protein, releasing the cytoplasmic
domain from the membrane. The tunicamycin induction of GAL4-ATF6
clearly shows that ATF6 is a target of the ER stress response.
|
IRE1 Is Required for Tunicamycin Induction of ATF6--
Ire1p is
an ER transmembrane protein that is required for the ER stress response
in the budding yeast, Saccharomyces cerevisiae (18, 19).
Ire1p contains a cytoplasmic protein kinase domain along with a domain
similar to endoribonuclease RNase L (20). Yeast Ire1p can be activated
by ER stress signals, which leads to activation of its ribonuclease
activity. This activity causes the cleavage of an mRNA for the
transcription factor Hac1p at two sites removing an intron. The cleaved
mRNA is subsequently religated by a tRNA ligase. This results in
enhanced translation of Hac1p, which activates ER stress-responsive
genes in yeast (20-23). Two mammalian homologues of yeast Ire1p have
been cloned that have similar domains and ER membrane localization (13, 24). Overexpression of mammalian IRE1 (hIRE1 and mIRE1) activated GRP78 gene expression, and dominant negative IRE1 was able
to inhibit tunicamycin induction of GRP78 expression (13,
24). These results suggest that IRE1 is also a critical mediator of the
ER stress response in mammalian cells.
Since GRP78 is regulated by ATF6 (2), we sought to determine
whether IRE1 is necessary or sufficient for ATF6 activation as well.
Overexpression of hIRE1 resulted in strong activation of the 5×
ATF6 site reporter, which could be increased another 2.5-fold by
tunicamycin treatment (Fig.
8A). We tested a dominant negative hIRE1 construct, hIRE1C, which has a deletion of the C-terminal kinase and RNase L-like domains, for an effect on ATF6 activation. We found that hIRE1C strongly inhibited tunicamycin induction of the 5× ATF6 site reporter gene (Fig. 8A). We
tested point mutants, K599A and K907A, that abolish the kinase or
nuclease activities,
respectively,2 for their
activation of the 5× ATF6 reporter gene. Neither of these were able to
activate the reporter, suggesting that both activities are required for
IRE1 function (Fig. 8B). Both of these point mutants only
weakly inhibited tunicamycin induction; however, they were both
expressed weakly (as assayed by immunoblotting) compared with hIRE1C
(data not shown). The K599A mutant was previously shown to block
tunicamycin induction of the GRP78 gene in COS cells (13).
This may be due to the higher expression of transfected genes in COS
cells compared with HeLa cells. In addition, the hIRE1C construct
was found to be a more effective inhibitor of GRP78
induction in COS cells.3
These results show that hIRE1 overexpression is sufficient to induce ATF6 and suggest that IRE1 is required for ATF6 activation. This
suggests that IRE1 is upstream of ATF6 in the ER stress response pathway. Further, both the kinase and nuclease activities of IRE1 appear to be required for ATF6 activation.
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
We have found that ATF6 binds directly to a consensus DNA binding site TGACGTG(G) in vitro. This site could also be activated by overexpression of ATF6 in vivo. When placed upstream of a reporter gene, the consensus ATF6 site was activated by agents that induce the ER stress response. Endogenous ATF6 appears to mediate this response, since dominant negative forms of ATF6 blocked the induction. A fusion of ATF6 to GAL4's DNA binding domain showed that it could mediate ER stress induction of a GAL4 reporter gene. These results firmly show that ATF6 is a target of the ER stress response.
Mori's group (2) identified ATF6 in a yeast one hybrid screen as interacting with the consensus ERSE CCAATN9CCACG. We found that dominant negative ATF6 could also block ER stress induction of the GRP78 promoter that contains three ERSE elements. In addition, ATF6-(1-373) activated expression of the GRP78 reporter, but not of one with mutations in the ERSEs (data not shown). These results further suggest that ATF6 is involved in ER stress induction of the GRP78 gene. NF-Y (CBP) can bind to the CCAAT box of the ERSE site (4, 5), and it has been proposed that ATF6 interacts with NF-Y and binds to the CCACG part of the site (8). Interestingly, this site is the complement of the 3'-half of the ATF6 consensus site we have identified, TGACGTGG. As a member of the bZIP family, ATF6 probably binds to the ATF6 site as a dimer. In fact, using different sized variants of ATF6, we have found that ATF6 can homodimerize, while it did not heterodimerize with either ATF1 or ATF2 in vitro.4 In the ERSE, it is likely that ATF6 only binds to the DNA through a single subunit with the binding stabilized by its interaction with NF-Y or another factor. This would explain why ATF6 does not bind to this site alone (2).5 A similar mechanism is used by Elk-1 binding with serum response factor to the serum response element of the c-fos gene (25).
We have used the isolated bZIP domain of ATF6 to identify its DNA binding site. It is possible that other sequences in the cytoplasmic domain (aa 1-373) will affect the DNA binding specificity. We have not been able to express the full cytoplasmic domain in a soluble form in E. coli in order to test this directly. However, the DNA binding site was activated strongly and specifically by ATF6-(1-373), suggesting that the full domain retains the specificity of the bZIP domain. Regions outside the bZIP domain, however, may affect its association with other factors such as NF-Y and thus affect its binding to other sites. We have also been unable to detect endogenous ATF6 binding activity in nuclear extracts of uninduced or tunicamycin induced HeLa cells (data not shown). Using the ATF6 binding site in a gel mobility shift assay with nuclear extracts, we detect binding of ATF1 and CREB, but not ATF6. This probably reflects the abundance of ATF1 and CREB and the low levels of the ATF6 cleavage product. Nevertheless, the cell can distinguish the ATF6 site from ATF1/CREB binding sites, since we found that only the former site was activated by ATF6-(1-373) and tunicamycin (Fig. 5A).
We propose that some ER stress-inducible genes are activated by direct
ATF6 binding sites rather than the consensus ERSE sites. However, our
analysis of the promoter regions of ER stress-inducible genes that have
been identified has not revealed an ATF6 consensus site that is clearly
required for ER stress induction. A match of TGACGTG was found in the
mouse and rat GRP78 promoters at 344 and 378 of the
transcriptional initiation site, respectively, but was not conserved in
the human GRP78 promoter. This region of the rat
GRP78 promoter was not required for tunicamycin induction of
a reporter gene, since it contains three ERSE-like elements from 163
to 98. However, deletion of 457 to 154 in the rat GRP78 promoter strongly reduced the level of
tunicamycin-induced expression from 40 to 7 relative units although the
-fold tunicamycin induction was relatively unaltered, changing from 8- to 7-fold (3). Thus, further work will be required to show whether the putative ATF6 site at 378 contributes to ER stress induction of the
rat GRP78 reporter gene. An ATF6 site may also be present outside the sequenced promoter regions or in other ER stress-inducible genes where the promoters have not been characterized. We feel it
likely that ER stress induction of the ATF6 site occurs in natural
promoters, since the induction was very specific with the model
reporter gene; a single point mutation in the site abolished induction,
and the similar CRE site was not induced at all by tunicamycin.
A BLAST search of the Eukaryotic Promotor Database at the National Center for Biotechnology Information revealed four mammalian promoters that match the ATF6 consensus site TGACGTG. Since the promoter data base is relatively small and limited to 600 bases of the promoter, this search is fairly limited. In addition, the statistical significance of a seven-base match is not very high. Nevertheless, the best match that contains the TGACGTGG sequence was the promoter of the human ADP-ribosylation factor 1 gene. ADP-ribosylation factor 1 is a small GTPase localized to the Golgi that affects vesicular transport (26) such that it is a reasonable candidate for a gene induced to compensate for ER perturbations. Other promoters matching the TGACGTG consensus were the mouse c-mos, human apolipoprotein CII, and human HMG-14 genes. Further work will be required to test these possible ATF6 targets.
The model for ER stress regulation of ATF6 proposed by Haze et al. (8) is that ATF6 is localized on the ER membrane and that upon ER stress it is cleaved off the membrane such that the N-terminal cytoplasmic domain can move to the nucleus and activate gene expression. Although we have not been able to confirm the cleavage of ATF6, our results are generally consistent with this model. We find by immunofluorescence of epitope-tagged ATF6 that it localizes to the ER or cytoplasm (data not shown). When the cytoplasmic domain of ATF6 (aa 1-373) was expressed separately, it localized completely to the nucleus. In addition, this fragment of ATF6 was a potent activator of transcription. Our inability to detect the 50-kDa cleavage product of ATF6 may be due to limits of sensitivity and the small amounts of this fragment that are generated. We used transfection of epitope-tagged ATF6 to look for cleavage, while Haze et al. used antisera to endogenous ATF6. It is possible that transfection of ATF6 results in high levels per cell and that only very low levels of 50-kDa ATF6 are generated in each cell. Nevertheless, Haze et al. also used the same epitope-tagged ATF6 construct in the same cells (HeLa) and detected basal levels of the ATF6 50-kDa cleavage product that we have not detected (Fig. 4C). We cannot account for this difference. Unfortunately, our antisera to ATF6 (7) had a high background in immunoblots such that we were not able to observe a clear 50-kDa product. Further experiments and improved reagents will be required for us to critically judge the presence and importance of the 50-kDa ATF6 cleavage product. Since we and Mori's group (8) have found that ATF6 is an activator of gene expression, the model of cleavage of ATF6 and movement to the nucleus is much more plausible than the alternative, which is that ATF6 acts on transcription as a transmembrane protein on the inner surface of the nuclear membrane or that the full-length protein dissociates from the membrane and moves to the nucleus.
A precedent for activation of a transcription factor by cleavage from the ER membrane is SRE-BP (reviewed in Ref. 27). SRE-BP is an ER transmembrane protein that is activated in response to low cholesterol. This results in cleavage of SRE-BP at two sequential sites. First it is cleaved in the luminal domain by site 1 protease followed by cleavage by site 2 protease in the transmembrane domain near the cytoplasmic face. This results in release of the DNA binding domain of SRE-BP, which moves to the nucleus. The site 1 protease site in SRE-BP was found to have the sequence RSVLS with cleavage between the leucine and serine (28). The arginine and leucine residues were absolutely required, while the other residues could be exchanged freely such that the requirement for the cleavage site is RXXL (28). Interestingly, a similar sequence, RRHLL, is present in the luminal domain of ATF6. We are currently testing whether this site and site 1 protease are required for ATF6 activation. The cleavage of SRE-BP is further regulated by its association with SREBP cleavage-activating factor, which senses cholesterol levels and controls its accessibility to the proteases (27). The precedent of SRE-BP activation raises the interesting questions of whether ATF6 activation is controlled by complexing factors and whether cleavage need occur in the luminal domain as well as near the junction of the cytoplasmic domain with the ER membrane.
The most proximal sensor of the ER stress response that has been identified is Ire1p. In yeast, this protein is required for transmission of the ER stress response (18, 19). It is an ER transmembrane protein with two catalytic domains in the C-terminal cytoplasmic domain: a protein kinase domain and a domain with endoribonuclease activity that is similar to RNase L (20). Ire1p has the remarkable ability to regulate the levels of Hac1p, a transcription factor, by inducing nonconventional splicing of its message, resulting in a transcript that is more efficiently translated (20-23).
Two mammalian homologues of Ire1p have been cloned, hIRE1 and
mIRE1, which have similar ER localization and domain structure as
yeast Ire1p (13, 24). Mammalian IRE1 appears to be required for the ER
stress response, since it can induce GRP78/BiP and CHOP
mRNA levels, two target genes of the ER stress response. In
addition, dominant negative forms of IRE1 defective in the kinase and
endonuclease domains blocked ER stress induction of target genes (13,
24).3 We similarly found that hIRE1 overexpression
activated the ATF6 reporter and that point mutations in the kinase or
endonuclease domains abolished this activation. An IRE1 construct with
a deletion of the kinase and nuclease domains in fact acted as a
dominant negative to block ER stress activation of the ATF6 reporter
gene. These results suggest that IRE1 is upstream of ATF6 in the ER stress signaling pathway. How direct IRE1 activation of ATF6 is remains
to be determined. Since we have had difficulty detecting the ATF6
cleavage product, we have not been able to test whether IRE1 induces
ATF6 cleavage. Our preliminary results have also not detected
tunicamycin-induced phosphorylation of ATF6 such that direct
phosphorylation of ATF6 by IRE1 may not be involved.
One mechanism for activation of ATF6 by IRE1 could be through activation and phosphorylation by the stress-activated MAP kinases JNK and p38. It was recently found that tunicamycin and IRE1 can activate JNK kinase activity (29). It has also been shown that p38 can phosphorylate ATF6 in vitro and activate its transcription activity in myocytes (30). However, we found that JNK and p38 activators (activated MEKK and MKK6) did not activate the ATF6 reporter gene (data not shown). These results suggest that JNK and p38 activation are not sufficient for ER stress activation of ATF6. In addition, dominant negative TRAF2 blocked IRE1 activation of JNK (29) but had no effect on IRE1 activation of the ATF6 reporter by tunicamycin, suggesting that JNK activation is not required (data not shown).
Mouse IRE1 overexpression also induced apoptosis (24). Since ER
stress inducers can cause apoptosis as well as induction of ER
chaperones that protect cells from apoptosis, it is intriguing to
conclude that IRE1 can mediate both of these signals. It was recently
found that caspase-12 is localized on the ER membrane and is partially
required for induction of apoptosis by ER stress (31). ER stress
induced the cleavage of caspase-12 to a fragment that was no longer
associated with the ER (31). Since IRE1 can induce apoptosis, it will
be interesting to determine whether it or ATF6 is required for
caspase-12 activation.
A last interesting aspect of the ER stress signaling pathway is the
role of presenilin-1 (PS1). PS1 is a multipass ER membrane protein that
is mutated in some cases of familial Alzheimer's disease (32, 33).
Cells with familial Alzheimer's disease-linked variants of PS1 have
altered processing of the amyloid precursor protein that is processed
to the amyloid- peptide (34, 35). Amyloid- peptide has been
associated with the cerebral neuritic plaques found in Alzheimer's
patients (36). PS1 is also required for cleavage of Notch, a
transmembrane protein required for developmental decisions
(37-39). Cleavage of Notch at the plasma membrane results in a
fragment that can translocate to the nucleus and activate gene
expression (40-43).
Recently, PS1 mutants were found to have a reduced ER stress response
and to bind to hIRE1 in cultured cells (44). It was also found that
hIRE1 is cleaved in response to ER stress such that the cytoplasmic
domain can move to the nucleus (45). This movement of hIRE1 induced by
ER stress was defective in PS1
/ cells, although the
induction of GRP78 mRNA was only moderately reduced
(45). These results suggest a connection of PS1 and IRE1 processing to
the ER stress response and the possibility that defects in the ER
stress pathway may contribute to Alzheimer's disease. It will be
interesting to determine whether ATF6 activation is affected by
presenilin-1 or other Alzheimer's-associated mutations.
| |
ACKNOWLEDGEMENTS |
|---|
We thank Lian Duan for aid in purifying the ATF6bZIP protein and Dr. Kazutoshi Mori for the GRP78 reporter gene.
| |
FOOTNOTES |
|---|
* This work was supported by NCI, National Institutes of Health Grant CA50329 (to R. P.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
These two authors contributed equally to this work.
¶ To whom correspondence should be addressed: Dept. of Biological Sciences, Columbia University, Fairchild 813B, MC 2420, 1212 Amsterdam Ave., New York, NY 10027. Tel.: 212-854-8281; Fax: 212-854-7655; E-mail: mrp6@columbia.edu.
Published, JBC Papers in Press, June 15, 2000, DOI 10.1074/jbc.M003322200
2 W. Tirasophon and R. J. Kaufman, submitted for publication.
3 W. Tirasophon and R. J. Kaufman, unpublished results.
4 C. Zhu and R. P., unpublished results.
5 Y. Wang, J. Shen, and R. Prywes, unpublished results.
| |
ABBREVIATIONS |
|---|
The abbreviations used are:
ER, endoplasmic
reticulum;
ERSE, ER stress response element;
bZIP, basic-leucine
zipper;
aa, amino acids;
PCR, polymerase chain reaction;
CRE, cAMP-response element;
CREB, cAMP-response element-binding protein;
HA, hemagglutinin;
hIRE1, human IRE1;
mIRE1, murine IRE1;
JNK, c-Jun
N-terminal kinase;
PS1, presenilin-1;
TM, tunicamycin.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
| 1. | Kaufman, R. J. (1999) Genes Dev. 13, 1211-1233 |
| 2. | Yoshida, H., Haze, K., Yanagi, H., Yura, T., and Mori, K. (1998) J. Biol. Chem. 273, 33741-33749 |
| 3. | Roy, B., and Lee, A. S. (1999) Nucleic Acids Res. 27, 1437-1443 |
| 4. | Roy, B., and Lee, A. S. (1995) Mol. Cell. Biol. 15, 2263-2274 |
| 5. | Roy, B., Li, W. W., and Lee, A. S. (1996) J. Biol. Chem. 271, 28995-29002 |
| 6. | Hai, T. W., Liu, F., Coukos, W. J., and Green, M. R. (1989) Genes Dev. 3, 2083-2090 |
| 7. | Zhu, C., Johansen, F. E., and Prywes, R. (1997) Mol. Cell. Biol. 17, 4957-4966 |
| 8. | Haze, K., Yoshida, H., Yanagi, H., Yura, T., and Mori, K. (1999) Mol. Biol. Cell 10, 3787-3799 |
| 9. | Wang, B. Q., Kostrub, C. F., Finkelstein, A., and Burton, Z. F. (1993) Protein Expression Purif. 4, 207-214 |
| 10. | Mavrothalassitis, G., Beal, G., and Papas, T. S. (1990) DNA Cell Biol. 9, 783-788 |
| 11. | Clarke, N., Arenzana, N., Hai, T., Minden, A., and Prywes, R. (1998) Mol. Cell. Biol. 18, 1065-1073 |
| 12. | Tanaka, M., and Herr, W. (1990) Cell 60, 375-386 |
| 13. | Tirasophon, W., Welihinda, A. A., and Kaufman, R. J. (1998) Genes Dev. 12, 1812-1824 |
| 14. | Wang, Y., Falasca, M., Schlessinger, J., Malstrom, S., Tsichlis, P., Settleman, J., Hu, W., Lim, B., and Prywes, R. (1998) Cell Growth Differ. 9, 513-522 |
| 15. | Chen, X., and Prywes, R. (1999) Mol. Cell. Biol. 19, 4695-4702 |
| 16. | Montminy, M. (1997) Annu. Rev. Biochem. 66, 807-822 |
| 17. | Fisch, T. M., Prywes, R., and Roeder, R. G. (1987) Mol. Cell. Biol. 7, 3490-3502 |
| 18. | Cox, J. S., Shamu, C. E., and Walter, P. (1993) Cell 73, 1197-1206 |
| 19. | Mori, K., Ma, W., Gething, M. J., and Sambrook, J. (1993) Cell 74, 743-756 |
| 20. | Sidrauski, C., and Walter, P. (1997) Cell 90, 1031-1039 |
| 21. | Kawahara, T., Yanagi, H., Yura, T., and Mori, K. (1997) Mol. Biol. Cell 8, 1845-1862 |
| 22. | Chapman, R. E., and Walter, P. (1997) Curr. Biol. 7, 850-859 |
| 23. | Cox, J. S., and Walter, P. (1996) Cell 87, 391-404 |
| 24. | Wang, X. Z., Harding, H. P., Zhang, Y., Jolicoeur, E. M., Kuroda, M., and Ron, D. (1998) EMBO J. 17, 5708-5717 |
| 25. | Treisman, R. (1994) Curr. Opin. Genet. Dev. 4, 96-101 |
| 26. | Jackson, C. L., and Casanova, J. E. (2000) Trends Cell Biol. 10, 60-67 |
| 27. | Brown, M. S., and Goldstein, J. L. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 11041-11048 |
| 28. | Duncan, E. A., Brown, M. S., Goldstein, J. L., and Sakai, J. (1997) J. Biol. Chem. 272, 12778-12785 |
| 29. | Urano, F., Wang, X., Bertolotti, A., Zhang, Y., Chung, P., Harding, H. P., and Ron, D. (2000) Science 287, 664-666 |
| 30. | Thuerauf, D. J., Arnold, N. D., Zechner, D., Hanford, D. S., DeMartin, K. M., McDonough, P. M., Prywes, R., and Glembotski, C. C. (1998) J. Biol. Chem. 273, 20636-20643 |
| 31. | Nakagawa, T., Zhu, H., Morishima, N., Li, E., Xu, J., Yankner, B. A., and Yuan, J. (2000) Nature 403, 98-103 |
| 32. | Rogaev, E. I., Sherrington, R., Rogaeva, E. A., Levesque, G., Ikeda, M., Liang, Y., Chi, H., Lin, C., Holman, K., Tsuda, T., Mar, L., Sorbl, S., Nacmias, B., Piacentini, S., Amaducci, L., Chumakov, I., Cohen, D., Lannfelt, L., Fraser, P. E., Rommens, J. M., and St. George-Hyslop, P. H. (1995) Nature 376 (6543), 775-778 |
| 33. | Sherrington, R., Rogaev, E. I., Liang, Y., Rogaeva, E. A., Levesque, G., Ikeda, M., Chi, H., Lin, C., Li, G., Holman, K., Tsuda, T., Mar, L., Foncin, J.-F., Bruni, A. C., Montesi, M. P., Sorbi, S., Rainero, I., Pinessi, L., Nee, L., Chumakov, I., Pollen, D., Brookes, A., Sanseau, P., Polinsky, R. J., Wasco, W., Da Silva, H. A. R., Haines, J. L., Pericak-Vance, M. A., Tanzi, R. E., Roses, A. D., Fraser, P. E., Rommens, J. M., and St. George-Hyslop, P. H. (1995) Nature 375, 754-760 |
| 34. | Borchelt, D. R., Thinakaran, G., Eckman, C. B., Lee, M. K., Davenport, F., Ratovitsky, T., Prada, C. M., Kim, G., Seekins, S., Yager, D., Slunt, H. H., Wang, R., Seeger, M., Levey, A. I., Gandy, S. E., Copeland, N. G., Jenkins, N. A., Price, D. L., Younkin, S. G., and Sisodia, S. S. (1996) Neuron 17, 1005-1013 |
| 35. | Duff, K., Eckman, C., Zehr, C., Yu, X., Prada, C. M., Perez-tur, J., Hutton, M., Buee, L., Harigaya, Y., Yager, D., Morgan, D., Gordon, M. N., Holcomb, L., Refolo, L., Zenk, B., Hardy, J., and Younkin, S. (1996) Nature 383, 710-713 |
| 36. | Haass, C., and De Strooper, B. (1999) Science 286, 916-919 |
| 37. | De Strooper, B., Annaert, W., Cupers, P., Saftig, P., Craessaerts, K., Mumm, J. S., Schroeter, E. H., Schrijvers, V., Wolfe, M. S., Ray, W. J., Goate, A., and Kopan, R. (1999) Nature 398, 518-522 |
| 38. | Struhl, G., and Greenwald, I. (1999) Nature 398, 522-525 |
| 39. | Ye, Y., Lukinova, N., and Fortini, M. E. (1999) Nature 398, 525-529 |
| 40. | Struhl, G., and Adachi, A. (1998) Cell 93, 649-660 |
| 41. | Schroeter, E. H., Kisslinger, J. A., and Kopan, R. (1998) Nature 393, 382-386 |
| 42. | Lecourtois, M., and Schweisguth, F. (1998) Curr. Biol. 8, 771-774 |
| 43. | Kidd, S., Lieber, T., and Young, M. W. (1998) Genes Dev. 12, 3728-3740 |
| 44. | Katayama, T., Imaizumi, K., Sato, N., Miyoshi, K., Kudo, T., Hitomi, J., Morihara, T., Yoneda, T., Gomi, F., Mori, Y., Nakano, Y., Takeda, J., Tsuda, T., Itoyama, Y., Murayama, O., Takashima, A., St. George-Hyslop, P., Takeda, M., and Tohyama, M. (1999) Nat. Cell Biol. 1, 479-485 |
| 45. | Niwa, M., Sidrauski, C., Kaufman, R. J., and Walter, P. (1999) Cell 99, 691-702 |
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H.-Y. Seo, Y. D. Kim, K.-M. Lee, A.-K. Min, M.-K. Kim, H.-S. Kim, K.-C. Won, J.-Y. Park, K.-U. Lee, H.-S. Choi, et al. Endoplasmic Reticulum Stress-Induced Activation of Activating Transcription Factor 6 Decreases Insulin Gene Expression via Up-Regulation of Orphan Nuclear Receptor Small Heterodimer Partner Endocrinology, August 1, 2008; 149(8): 3832 - 3841. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Cunha, P. Hekerman, L. Ladriere, A. Bazarra-Castro, F. Ortis, M. C. Wakeham, F. Moore, J. Rasschaert, A. K. Cardozo, E. Bellomo, et al. Initiation and execution of lipotoxic ER stress in pancreatic {beta}-cells J. Cell Sci., July 15, 2008; 121(14): 2308 - 2318. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. E. Audas, Y. Li, G. Liang, and R. Lu A Novel Protein, Luman/CREB3 Reruitment Factor, Inhibits Luman Activation of the Unfolded Protein Response Mol. Cell. Biol., June 15, 2008; 28(12): 3952 - 3966. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
