Chen M, Gutierrez GJ, Ronai ZA.. Ubiquitin-recognition protein Ufd1 couples the endoplasmic reticulum (ER) stress response to cell cycle control. Proc Natl Acad Sci USA 108: 9119-9124

ArticleinProceedings of the National Academy of Sciences 108(22):9119-24 · May 2011with43 Reads
Impact Factor: 9.67 · DOI: 10.1073/pnas.1100028108 · Source: PubMed
Abstract

The ubiquitin-recognition protein Ufd1 facilitates clearance of misfolded proteins through the endoplasmic reticulum (ER)-associated degradation (ERAD) pathway. Here we report that prolonged ER stress represses Ufd1 expression to trigger cell cycle delay, which contributes to ERAD. Remarkably, down-regulation of Ufd1 enhances ubiquitination and destabilization of Skp2 mediated by the anaphase-promoting complex or cyclosome bound to Cdh1 (APC/C(Cdh1)), resulting in accumulation of the cyclin-dependent kinase inhibitor p27 and a concomitant cell cycle delay during the G1 phase that enables more efficient clearance of misfolded proteins. Mechanistically, nuclear Ufd1 recruits the deubiquitinating enzyme USP13 to counteract APC/C(Cdh1)-mediated ubiquitination of Skp2. Our data identify a coordinated cell cycle response to prolonged ER stress through regulation of the Cdh1-Skp2-p27 axis by Ufd1 and USP13.

Full-text

Available from: Gustavo J Gutierrez, Jul 24, 2014
Ubiquitin-recognition protein Ufd1 couples the
endoplasmic reticulum (ER) stress response to
cell cycle control
Meifan Chen, Gustavo J. Gutierrez, and Zeev A. Ronai
1
Signal Transduction Program, Sanford-Burnham Medical Research Institute, La Jolla, CA 92037
Edited* by Aaron Ciechanover, Technion-Israel Institute of Technology, Bat Galim, Haifa, Israel, and approved April 15, 2011 (received for review
January 2, 2011)
The ubiquitin-recognition protein Ufd1 facilitates clearance of mis-
folded proteins through the endoplasmic reticulum (ER)-associated
degradation (ERAD) pathway. Here we report that prolonged ER
stress represses Ufd1 expression to trigger cell cycle delay, which
contributes to ERAD. Remarkably, down-regulation of Ufd1 enhan-
ces ubiquitination and destabilization of Skp2 mediated by the
anaphase-promoting complex or cyclosome bound to Cdh1 (APC/
C
Cdh1
), resulting in accumulation of the cyclin-dependent kinase in-
hibitor p27 and a concomitant cell cycle delay during the G1 phase
that enables more efcient clearance of misfolded proteins. Mech-
anistically, nuclear Ufd1 recruits the deubiquitinating enzyme USP13
to counteract APC/C
Cdh1
-mediated ubiquitination of Skp2. Our data
identify a coordinated cell cycle response to prolonged ER stress
through regulation of the Cdh1-Skp2-p27 axis by Ufd1 and USP13.
R
egulated protein degradation by the ubiquitin-proteasome
system (UPS) plays a central role in diverse cellular processes.
An effort to dissect this proteolytic system in Saccharomyces
cerevisiae uncovered Ufd1 in a genetic screen for mutations in the
ubiquitin fusion degradation (UFD) pathway responsible for the
degradation of a synthetic substrate N-terminally fused to one
ubiquitin moiety (1). Functional and structural evidence suggests
that Ufd1 acts as an ubiquitin-recognition protein with putative
monoubiquitin and polyubiquitin binding sites (2).
Ufd1 is best characterized as an adaptor protein that, together
with Npl4, confers AAA-ATPase Cdc48/p97/VCPspecic activity
in endoplasmic reticulum (ER)-associated degradation (ERAD)
(3), a process that functions constitutively to export misfolded
proteins from the ER to the cytosol for UPS-dependent degra-
dation. ERAD is part of the unfolded protein response (UPR),
which is critical for restoring homeostasis in the ER when its
function is perturbed, such as by the accumulation of misfolded
proteins. Analyses of ERAD performed primarily in yeast suggest
a role for the Ufd1-Npl4-p97 complex in recognizing and extract-
ing polyubiquitinated misfolded proteins from the ER of mis-
folded proteins. Intriguingly, in mammalian cells, Ufd1 also di-
rectly enhances the activity of gp78, an ubiquitin ligase involved in
ERAD, independently of p97 and Npl4 (4). Studies in yeast have
reported impaired degradation of the ER proteins HMG-CoA
reductase, H-2Kb (MHC class I heavy chain), and CPY* (an ab-
errant form of carboxypeptidase Y) in the Ufd1 mutant, sug-
gesting an essential role for Ufd1 in promoting ERAD (3, 5).
However, RNAi-mediated depletion of Ufd1 in mammalian cells
has yielded opposite results. While some studies reported im-
paired ERAD in Ufd1-depleted cells (6, 7), others showed ac-
celerated degradation of the classical ERAD substrates, such as
cholera toxin and T-cell receptors (8, 9). Although these seemingly
contradictory observations might stem from the use of distinct
cellular systems, they point to greater complexity in the regulation
and function of Ufd1 that impinge on the ER stress response.
Furthermore, In Xenopus laevis egg extracts, Ufd1-Npl4-p97
promote chromatin decondensation (10) and regulate spindle
disassembly (11).
In the present study, we examined the function of Ufd1 in
maintaining steady-state levels of Skp2 (the F-box adaptor of the
E3 ubiquitin ligase SCF
Skp2
) in mammalian cells by regulating its
ubiquitination. We report that Ufd1 acts as a scaffold for Skp2
and the deubiquitinating enzyme (DUB) USP13 to antagonize
anaphase-promoting complex or cyclosome bound to Cdh1
(APC/C
Cdh1
)-mediated ubiquitination of Skp2. Our ndings also
show that prolonged ER stress down-regulates Ufd1 levels,
triggering Skp2 destabilization and accumulation of p27, which
contribute to delayed progression through G1. Our data also
demonstrate facilitated degradation of misfolded proteins in G1-
arrested cells, suggesting that the link between Ufd1 and cell
cycle control serves to optimize ERAD.
Results
Prolonged Treatment with Tunicamycin Down-Regulates Ufd1 Expression.
To investigate the role of Ufd1 in the UPR of mammalian cells,
we rst examined the expression of endogenous Ufd1 protein in
HeLa cells treated with tunicamycin (TM), a glycosylation in-
hibitor. To our surprise, we found decreased Ufd1 expression at
20 h after treatment with 0.5 μg/mL of TM, concomitant with
the accumulation of GRP78/BiP, a known marker of the UPR
(Fig. 1A). Down-regulation of Ufd1 in mammalian cells exposed
to prolonged TM treatment may accommodate a novel function
of Ufd1 in the UPR, distinct from its direct role in the retro-
translocation of misfolded proteins.
Ufd1 Regulates the Cdh1-Skp2-p27 Axis. To understand the signi-
cance of ER stress-dependent down-regulation of Ufd1, we rst
examined the effect of Ufd1 down-regulation under nonstressed
conditions by depleting Ufd1 from HeLa cells. Knockdown of
Ufd1 (75% reduction) was achieved in a stable Ufd1-decient
cell line (Ufd1-KD) and was validated at both the mRNA and
protein levels (Fig. S1 A and B; drug-selected pool 2).
Given the link between Ufd1 and the cell cycle, we asked
whether Ufd1 knockdown would affect cell cycle progression. By
examining the levels of major cell cycle markers in Ufd1-KD
cells synchronized in G1/S by double-thymidine block and re-
lease, we observed an overall down-regulation of Skp2 accom-
panied by an up-regulation of p27, a major substrate of SCF
Skp2
(12) (Fig. 1B). Accordingly, we found that although the order of
activation of cyclin-dependent kinase (CDK) 1 and CDK2 was
preserved in Ufd1-KD cells, their activation was delayed (Fig.
1C). Expression of exogenous Ufd1 in Ufd1-KD cells restored
Author contributions: M.C., G.J.G., and Z.A.R. designed research; M.C. and G.J.G. per-
formed research; M.C., G.J.G., and Z.A.R. analyzed data; and M.C., G.J.G., and Z.A.R. wrote
the paper.
The authors declare no conict of interest.
*This Direct Submission article had a prearranged editor.
1
To whom correspondence should be addressed. E-mail: ronai@sanfordburnham.org.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1100028108/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1100028108 PNAS
|
May 31, 2011
|
vol. 108
|
no. 22
|
91199124
CELL BIOLOGY
Page 1
Skp2 and p27 levels to those seen in control cells (Fig. 1D). The
effect of Ufd1 on Skp2 was conrmed in an independently
established stable Ufd1-KD clone, as well as in cells with tran-
sient knockdown of Ufd1 by siRNA (Fig. S2A).
Down-regulation of Skp2 in Ufd1-KD cells can occur at the
level of transcription, translation, or protein stability. Whereas
qRT-PCR analysis revealed a twofold increase in Skp2 transcrip-
tion in Ufd1-KD cells compared with controls (Fig. 2A), kinetics
performed with cycloheximide (CHX) demonstrated reduced
protein stability of endogenous Skp2 in Ufd1-KD cells (Fig. S2B).
Using X. laevis egg extracts supplemented with exogenous Ufd1
in the presence of recombinant Cdh1, the adaptor of the APC/C
ubiquitin ligase responsible for Skp2 degradation in the cell cycle
(1317), we conrmed that Ufd1 is a negative regulator of Skp2
degradation (Fig. 2B). Accordingly, the stability of endogenous
Skp2 on exit from mitosis, when APC/C
Cdh1
is active in the cell
cycle, was increased in cells that overexpressed Ufd1 (Fig. 2C).
Given that ubiquitination of Skp2 is required for its protea-
somal degradation, its down-regulation in Ufd1-KD cells might
result from enhanced activation of APC/C
Cdh1
. To test this, we
compared the levels of Skp2 ubiquitination in control and Ufd1-
KD cells. We found enhanced ubiquitination of Skp2 in Ufd1-
KD cells (Fig. 2D), which could be rescued by reexpression of the
full-length Ufd1 (Fig. 2E) or depletion of Cdh1 from Ufd1-KD
cells (Fig. S2C). These ndings suggest that Ufd1 stabilizes Skp2
by antagonizing the ubiquitination of Skp2 by APC/C
Cdh1
.
Ufd1 Acts as a Scaffold for Skp2USP13 Interaction. We explored
whether Ufd1 protects Skp2 from ubiquitin-dependent degrada-
tion through recruitment of a DUB that counteracts Skp2 ubiq-
uitination. Consistent with data from a recent proteomic survey of
interaction partners of human DUBs (18), we observed robust in
vivo binding between Ufd1 and USP13 (Fig. 3A, lane 2). Using
a series of Ufd1 truncation mutants, we found that amino acids
261280 of Ufd1 were required for the in vivo binding of Ufd1 to
USP13 (Fig. 3A, lanes 28), which was further conrmed by the
inability of the Ufd1 mutant containing the internal deletion
Δ261280 to interact with USP13 in vivo (Fig. S3A). Notably, the
N241-Ufd1 mutant containing the p97 binding site (19) interacts
with p97 but not with USP13, indicating that Ufd1 has distinct
binding sites for both proteins (Fig. 3A, lane 5). We then sought
to determine whether interaction with USP13 enables Ufd1 to
protect Skp2 from ubiquitination by testing the ability of the
mutant Ufd1-Δ261280 to rescue the low protein expression and
enhanced ubiquitination of Skp2 in Ufd1-KD cells. Ufd1-Δ261
280 was unable to rescue the reduced levels of endogenous Skp2
(Fig. 3B) or the enhanced ubiquitination of Skp2 in Ufd1-KD
cells (Fig. 3C), indicating that Ufd1USP13 interaction is im-
portant in controlling Skp2 stability.
If Ufd1 facilitates interaction between USP13 and Skp2, then
it should be able to bind Skp2. Indeed, such an interaction was
observed in vivo (Fig. 3D, lane 2). In our effort to map the do-
main of Ufd1 required for the binding of Ufd1 to Skp2, we again
identied amino acids 261280, which also are required for the
binding of Ufd1 to USP13 (Fig. 3D, lanes 28). A schematic
summarizing the ability of the Ufd1 variants to bind USP13,
Skp2, and p97 is presented in Fig. S3B. In addition, Ufd1
interacts with free Skp2 proteins (i.e., those not in complex with
Skp1-Cul1), as demonstrated by a lack of endogenous Cul1 in
complex with Ufd1-Skp2 in vivo (Fig. S3C, Left) and the binding
of Ufd1 with equal afnity to both Skp2 and the Skp2ΔF-box
mutant, which cannot be incorporated into SCF complexes (12)
(Fig. S3C, Right). In support of our hypothesis that Ufd1-de-
pendent regulation of Skp2 is mediated by USP13, we also ob-
served in vivo interaction between USP13 and Skp2 (Fig. 3E)
and among these three proteins (Fig. 3 F and G), suggesting the
formation of a functional ternary complex in cells. Indeed, we
conrmed the existence of a complex containing Ufd1, Skp2, and
USP13 in vivo by sequential immunoprecipitations (Fig. 3G ).
Next, to assess the role of Ufd1 as a mediator of Skp2USP13
interaction, we compared the binding of USP13 to Skp2 in
control and Ufd1-KD cells. A decreased amount of Skp2-bound
USP13 was found in Ufd1-KD cells (Fig. 3H,
Middle). Moreover,
although exogenous Ufd1 can restore the interaction between
Skp2 and USP13 in Ufd1-KD cells, the ability of the mutant
Ufd1-Δ261280 to do so was attenuated (Fig. 3H, Right). Taken
together, our data suggest that Ufd1 acts as a scaffolding protein
that enables a functional interaction between USP13 and Skp2.
USP13 Deubiquitinates Skp2. Although USP13 shares sequence
and structural homology with the ubiquitin protease USP5, it has
been reported only as an ISG15-reactive protease (20). To di-
C
B
A
D
TM (0.5μg/ml; 20h) - +
Cdh1
G1/S release 0 2 4 6 9 12 15 0 2 4 6 9 12 15 h
control Ufd1-KD
cyclin B1
p27
β-actin
Ufd1
AU 1.00 1.05 1.06 1.19 1.40 1.33 1.21 0.53 0.60 0.47 0.46 0.49 0.54 0.46
Skp2
Ufd1
β-actin
GRP78
Skp2
HeLa
CHL1
β−actin
FLAG
p27
FLAG-Ufd1 + +
empty vector + +
Ufd1-KD + +
control-shRNA + +
control Ufd1-KD
G1/S release 0 2 4 6 7 8 9 10 11 12 13 14 0 2 4 6 7 8 9 10 11 12 13 14 h
cyclin B1
CDK2
control Ufd1-KD
Skp2
β−actin
G1/S release 0 2 4 6 7 8 9 10 11 12 13 14 0 2 4 6 7 8 9 10 11 12 13 14 h
CDK2 kinase activity
CDK1 kinase activity
β−actin
cyclin E2
AU 5.5 7.1 5.6 7.6 7.6 7.7 5.9 6.0 6.6 6.3 4.8 4.9 1.7 2.6 2.2 2.1 2.4 1.8 2.9 1.9 2.5 1.6 1.8 2.0
Fig. 1. Prolonged ER stress down-regulates Ufd1, and
depletion of Ufd1 by RNAi affects cell cycle progression.
(A) Protein levels of endogenous Ufd1 were compared
in HeLa cells treated with DMSO ()or0.5μg/mL of TM
(+) for 20 h by immunoblot analysis. GRP78 was used as
a marker of UPR, and β-actin served as a loading con-
trol. (B) Immunoblot analysis of cell cycle markers in
control and Ufd1-KD cells released from G1/S arrest
after double-thymidine block. Control refers to cells
infected with an empty shRNA plasmid. Levels of Ufd1
were quantied and are shown in arbitrary units (AU).
(C)(Upper) Comparison of CDK1 and CDK2 kinase ac-
tivity in control and Ufd1-KD cells released from G1/S
arrest. CDK2 kinase reactions from control and Ufd1-KD
cells were run in separate gels but were processed
identically otherwise. Same for CDK1 kinase reactions.
(Lower) Protein levels of Skp2 in control and Ufd1-KD
cells. (D) Analysis of cell cycle markers in control and
Ufd1-KD cells expressing FLAG-Ufd1. The melanoma
cell line CHL1 expressing high levels of p27 was used
solely as a positive control for p27 immunodetection.
9120
|
www.pnas.org/cgi/doi/10.1073/pnas.1100028108 Chen et al.
Page 2
rectly assess whether USP13 has DUB activity, we performed
deubiquitination assays using K48-linked di-ubiquitins as sub-
strates. Immunopuried USP13 (Fig. S4A) elicited a steady
cleavage of K48-linked di-ubiquitins (Fig. S4B).
We next evaluated the effect of USP13 on Skp2. We found that
USP13 overexpression increased the level of endogenous Skp2
protein (Fig. 4A). In agreement, USP13 knockdown reduced the
level of endogenous Skp2 and consequently increased p27 expres-
sion (Fig. S4C and Fig. 4B), supporting a positive role of USP13
in the regulation of Skp2 protein levels. To directly test whether
USP13 exhibits DUB activity toward Skp2, we incubated FLAG-
USP13 with ubiquitinated-Skp2, both immunopuried from cells.
The amount of ubiquitin conjugates on Skp2 was reduced in the
presence of USP13 (Fig. 4C). Taken together, our data show that
USP13 has DUB activity toward both K48-linked di-ubiquitin
and ubiquitinated Skp2.
ER Stress Regulates the Ufd1-Skp2-p27 Axis and G1 Cell Cycle
Progression.
The observation that Ufd1 knockdown induced
Skp2 down-regulation and consequently p27 up-regulation under
nonstressed conditions (Fig. 1 B and D) led us to examine
whether ER stress-dependent Ufd1 down-regulation (Fig. 1A)
would result in similar biochemical changes modulating cell cycle
progression. Indeed, we observed that TM triggered Ufd1 down-
regulation in a dose-dependent manner that was correlated with
Skp2 clearance and p27 accumulation after 20 h of treatment
(Fig. 5A). ER stress-dependent regulation of Ufd1-Skp2-p27 was
observed in other cell lines as well, including HeLa-S3, HEK-
A
C
D
B
0 30 60 90 120 min
S-Skp2
Interphase + Cdh1
+ 100ng GST
+ 100ng GST-Ufd1
+ 25ng GST-Ufd1
35
0.0
0.5
1.0
1.5
2.0
2.5
Ufd1-KDmock
Relative quantity of Skp2 mRNA
FLAG-Ufd1
HSP90
Skp2
Noco release,
CHX (25 g/ml) 0 1.5 3.5 5.5 0 1.5 3.5 5.5 h
empty vector FLAG-Ufd1
μ
60
80
100
120
FLAG-Ufd1
empty vector
Relative abundance (%)
normalized to HSP90
Quantication of endogenous Skp2 protein amount
noco release, CHX (h)
0 1.5 3.5 5.5
E
IB: myc
IB: myc
myc-Skp2 + + + +
HA-Ub + +
empty vector + +
control
Ufd1-KD
IP myc
lysate
(Ub) -Skp2
n
(Ub) -Skp2
n
Skp2
IB: Ub
34
55
43
95
72
130
170
34
55
43
95
72
130
170
IB: Ub
IP myc
FLAG-Ufd1
empty vector + +
myc-Skp2
HA-Ub
+
+ + +
+ + +
control
Ufd1-KD
IB: myc
IB: FLAGlysate
(Ub) -Skp2
n
55
72
95
130
170
43
34
26
MW (kDa)
MW (kDa)
Fig. 2. Ufd1 interferes with the ubiquitination of
Skp2 in vivo. (A)Quantication of Skp2 mRNA levels
in control and Ufd1-KD cells by SYBR-green qRT-PCR.
(B) Skp2 degradation assays in interphase Xenopus
egg extracts supplemented with recombinant Cdh1,
in the presence of the indicated amount of bacteri-
ally puried recombinant GST-Ufd1. Samples were
obtained at the indicated times after
35
S-Skp2 was
added to the extract. (C)(Upper) Half-life analysis of
endogenous Skp2 protein with CHX treatment in
empty vector- or FLAG-Ufd1transfected HeLa cells
released from nocodazole (noco) arrest. Total lysates
were immunoblotted for Skp2 or FLAG. Quantica-
tion is shown below the immunoblots. (D)Invivo
ubiquitination of Skp2. myc-Skp2 and HA-Ub were
expressed in control and Ufd1-KD cells. Immuno-
precipitates obtained with c-myc antibodies were
immunoblotted for either ubiquitin (Ub) or c-myc to
detect polyubiquitinated Skp2, denoted by (Ub)
n
-
Skp2. Control refers to cells infected with empty
shRNA plasmid. (E) In vivo ubiquitination of Skp2.
myc-Skp2 and HA-Ub were coexpressed with either
empty vector or FLAG-Ufd1. Immunoprecipitates ob-
tained with c-myc antibodies were immunoblotted
for Ub. (Bottom) Expression of FLAG-Ufd1 in total cell
extracts.
Chen et al. PNAS
|
May 31, 2011
|
vol. 108
|
no. 22
|
9121
CELL BIOLOGY
Page 3
293T, and HFF-1 (Fig. S5A), although the degree of response
differed among different cell lines subjected to the same TM
treatment. Thapsigargin, an inhibitor of ER Ca
2+
ATPase that
also induces ER stress, elicited similar responses (Fig. S5B).
Remarkably, the extent of the G1 delay was correlated with the
dose- and time-dependent regulation of Ufd1, Skp2, and p27
levels after TM treatment (Fig. 5B and Fig. S5C).
ER stress-dependent Ufd1 down-regulation is a response to
prolonged UPR activation (Figs. 1A and 5 A and B). Notably,
changes in the expression of Ufd1 and Skp2 occur mainly in the
nuclear fraction (Fig. S5D), suggesting that the pool of Ufd1
contributing to cell cycle control is distinct from that involved in
the retrotranslocation of misfolded proteins in the cytosol.
Consistent with this idea are observations that both USP13 (20)
and Skp2 (Fig. S5D) are localized mainly in the nucleus.
Cdh1-Skp2-p27 Axis Contributes to ER-Induced G1 Delay. Under
nonstressed conditions, Ufd1 knockdown destabilized the Skp2
protein. To examine whether Skp2 also is regulated at the level of
protein stability under ER stress conditions that down-regulate
Ufd1, we rst treated cells with TM and the proteasome inhibitor
MG-132. The addition of MG-132 prevented TM-induced down-
regulation of Skp2, indicating that ER stress accelerates protea-
somal degradation of Skp2 (Fig. 5C). Interestingly, treatment with
a low dose of MG-132 (2 μM) for 12 h stabilized Skp2 and
abolished p27 accumulation after TM treatment, suggesting that
ER stress-dependent p27 accumulation is a consequence of Skp2
destabilization, and that the coregulation of Skp2 and p27 in re-
sponse to TM occurs at the level of protein stability (Fig. 5C, Left,
lane 3 and Fig. S6A;Fig.5C, Right, lanes 3 and 7).
We observed that although MG-132 attenuated TM-induced
Skp2 down-regulation, the effect was nevertheless incomplete.
Indeed, analysis of Skp2 and p27 transcript levels after TM
treatment revealed a 30% decrease in Skp2 mRNA and a 1.6-fold
increase in p27 mRNA (Fig. S6B). These data suggest ER stress-
dependent transcriptional regulation of Skp2 and p27. In contrast,
Ufd1 mRNA levels were comparable before and after TM treat-
ment (Fig. S6B). Further, TM-dependent Ufd1 down-regulation
was not restored by MG-132 (Fig. 5C, Left), suggesting translation
or proteasome-independent degradation.
Next, to determine whether APC/C
Cdh1
targets Skp2 for
ubiquitin-mediated degradati on in response to ER stress, we
A
p97
FLAG
USP13
FLAG
USP13
IP: FLAG
lysate
USP13 + + + + + + + +
FLAG-Ufd1 +
FLAG-Ufd1-N215 +
FLAG-Ufd1-N220 +
FLAG-Ufd1-N241 +
FLAG-Ufd1-N260 +
FLAG-Ufd1-N280 +
FLAG-Ufd1-N300 +
1 2 3 4 5 6 7 8
empty vector +
B
FLAG-Ufd1 +
empty vector + +
FLAG- 261-280 +
Skp2
Ufd1
HSP90
cont Ufd1-KD
C
FLAG-Ufd1
FLAG- 261-280 +
empty vector + +
myc-Skp2
HA-Ub
+
+ + + +
+ + + +
IP myc
IB: Ub
IB: Skp2
IB: Ufd1
cont Ufd1-KD
lysate
AU 1.0 2.4 1.5 2
170
MW (kDa)
130
72
95
E
USP13
Skp2
USP13 + + + + + +
myc-Skp2 + + + +
lysate IP: myc
empty vector + +
endog USP13
endog USP13
endog Ufd1
FLAG-Skp2
IgG
FLAG-Skp2 + +
empty vector + +
lysate IP: FLAG
*
**
F
D
Skp2
Skp2
FLAG
FLAG
Skp2 + + + + + + + +
FLAG-Ufd1 +
FLAG-Ufd1-N215 +
FLAG-Ufd1-N220 +
FLAG-Ufd1-N241 +
FLAG-Ufd1-N260 +
FLAG-Ufd1-N280 +
FLAG-Ufd1-N300 +
empty vector +
IP: FLAG
lysate
1 2 3 4 5 6 7 8
G
step 1
IP FLAG-Ufd1
from cell lysate
step 2
FLAG peptide
elution
step 3
IP myc-Skp2
from elution
lysate
USP13 + + + + + + + +
empty vector + + + +
myc-Skp2 + + + +
FLAG-Ufd1 + + + +
10%
FLAG-IP
elution
10%
myc-IP
USP13
Skp2
Ufd1
IgG
H
Ufd1
β-actin
Ufd1-KD +
control-shRNA +
USP13
Skp2
USP13 + + + + + +
Ufd1-KD + +
control-shRNA + + + +
myc-Skp2 + + + + + +
* ** * **
lysate IP: myc
USP13 + + + + + + + +
Ufd1-KD + + + + + +
FLAG-Ufd1 ev ev Ufd1 261 ev ev Ufd1 261
-280 -280
control-shRNA + +
myc-Skp2 + + + + + + + +
lysate IP: myc
USP13
Skp2
FLAG
Fig. 3. Ufd1 interferes with the ubiquitination of Skp2 by
recruiting the deubiquitinating enzyme USP13. (A)Mappingof
the Ufd1 binding site to USP13 in vivo. USP13 was expressed
with empty vector (lane 1) or with the indicated FLAG-Ufd1 C-
terminal truncation mutants in HEK-293T cells (lanes 28). Ufd1-
N215 denotes an Ufd1 truncation mutant containing the N-
terminal amino acids 1215, and so on. FLAG-Ufd1 immuno-
precipitates were immunoblotted for USP13, p97, and FLAG. (B)
Western blots of endogenous Skp2 in control (cont) and Ufd1-
KD cells transfected with empty vector, FLAG-Ufd1, or mutant
FLAG-Ufd1Δ261280. (C) In vivo ubiquitination of Skp2 in con-
trol and Ufd1-KD cells overexpressing myc-Skp2 and HA-Ub
with empty vector, FLAG-Ufd1, or FLAG-Ufd1Δ261280. c-myc
immunoprecipitates were immunoblotted for Ub. Levels of
ubiquitination were quantied and are shown as AUs. (D)
Mapping of the Ufd1 binding site to Skp2 in vivo. myc-Skp2 was
expressed in HEK-293T cells with empty vector (lane 1) or the
indicated FLAG-Ufd1 C-terminal truncation mutants (lanes 28).
FLAG immunoprecipitates were immunoblotted for Skp2 and
FLAG. (E) Coimmunoprecipitation of USP13 and Skp2 in vivo.
USP13 was expressed with empty vector (lane 1) or together
with myc-Skp2 (lanes 2 and 3). c-myc immunoprecipitates were
immunoblotted for USP13 or Skp2. (F) Semiendogenous coim-
munoprecipitation. HeLa cells were transfected with FLAG-
Skp2, and FLAG immunoprecipitates were immunoblotted for
endogenous USP13 and Ufd1. *Short exposure; **long expo-
sure. (G) Evidence of ternary complex formation among USP13,
Ufd1, and Skp2 by sequential immunoprecipitations. HeLa cells
were transfected with FLAG-Ufd1, myc-Skp2, and USP13. FLAG
immunoprecipitates were eluted with FLAG peptide. Eluates
were then used for c-myc immunoprecipitation. Then 10% of
FLAG-IP, 10% of eluate, and all of myc-IP were immunoblotted
for USP13, Skp2, and Ufd1. (H)(Left) Immunoblots of endoge-
nous Ufd1 in control and Ufd1-KD cells. (Middle) USP13 and
myc-Skp2 were coexpressed in control cells (lanes 1 and 2) or
Ufd1-KD cells (lane 3). Control cells were transfected with either
half (*) or an equal amount (**) of the DNA used for trans-
fection in Ufd1-KD cells. c-myc immunoprecipitates were
immunoblotted for USP13. (Right) Control and Ufd1-KD cells
were transfected with USP13 and myc-Skp2 together with
empty vector (ev), FLAG-Ufd1, or the mutant FLAG-Ufd1Δ261
280. c-myc immunoprecipitates were immunoblotted for USP13
and Skp2.
9122
|
www.pnas.org/cgi/doi/10.1073/pnas.1100028108 Chen et al.
Page 4
knocked-down Cdh1 by RNAi, followed by treatment with TM.
Notably, Cdh1-RNAi alone attenuated TM-induced protea-
somal degradation of Skp2 and diminished TM-induced up-
regulation of p27 (Fig. S6A,lane10),demonstratingthatAPC/
C
Cdh1
targets Skp2 for degradation during ER stress, which in
turn leads to p27 accumulation. Cdh1 knockdown i ncompletely
prevented TM-induced Skp2 down-regulation (Fig. 5C and Fig.
S6A) because TM treatment also transcriptionally repressed
Skp2 (Fig. S6B). Compared with the addition of MG-132
alone, the combination of Cdh1-RNAi and MG-132 did not
further stabilize S kp2 under ER stress, indicating that APC/
C
Cdh1
is the primary E3 ligase responsible for targeting Skp2
for degradation under such condi tions (Fig. 5C , Right, lanes
1115). Taken together, our data show that E R stress directly
regulates the Cdh1-Skp2-p27 axis through Ufd1 to delay cell
cycle progression. Finally, the ability of Ufd1 overexpression or
p27 knockdown to overcome TM-induced G1 delay, although
only partial, fu rther supports a role of these protei ns in regu-
lating the UPR-induced cell cycle response (Fig. 5 D and E and
Fig. S6C).
Because it was previously reported that cyclin D1 down-reg-
ulation also can mediate TM-dependent G1 delay in NIH 3T3
cells (21, 22), we sought to analyze the contribution of cyclin D1
down-regulation and p27 up-regulation to ER stress-induced cell
cycle arrest. Interestingly in HeLa cells, overexpression of cyclin
D1 could not overcome TM-induced G1 delay (Fig. S7). Of note,
Skp2 expression was still reduced regardless of cyclin D1 over-
expression, indicating that Skp2 and cyclin D1 down-regulation
after TM treatment are independent. This points to the existence
of complementary mechanisms in ER stress-dependent cell cycle
arrest, which may be cell typespecic. Our data suggest that in
the absence of cyclin D1driven arrest, Ufd1-Skp2-p27 contrib-
utes to cell cycle delay.
G1 Cell Cycle Phase Facilitates Clearance of ERAD Substrates. Given
the link between ER stress and cell cycle, we next asked
whether ER stress-ind uced G1 delay affects ERAD. Toward
this end, we examined the efciency of ERAD in different
phases of the cell cycle by comparing the rate of degradation of
two classic ERAD substratescystic brosis tran smembrane
conductance regulator (CFTR) and ER luminal protein null
Hon g-Kong α1-antitryp sin (NHK)in G1/S- and G2/M-arres-
ted cells. Notably, accelerated degradation of both proteins was
AB C
USP13
Skp2
β-actin
empty vector
HeLa
USP13
p27
USP13
mock 1 2 3 mix
USP13 shRNA
β-actin
Skp2
(Ub)n-Skp2
IP’d myc-Skp2-ub + +
FLAG-USP13 +
control elution +
eluted FLAG-USP13
IB: Ub
IB: Skp2
IP: myc
170
130
72
95
170
130
72
95
MW (kDa)
Fig. 4. USP13 controls Skp2 levels via deubiquitination. (A) Immunoblots of
endogenous Skp2 and exogenous USP13 in HeLa cells transfected with
empty vector or untagged USP13. (B) Knockdown of USP13 in HeLa cells
infected with lentivirus packaged with either empty vector (mock) or one of
three individual USP13-specic shRNA vectors, as indicated. Immunoblots of
endogenous USP13, Skp2, and p27 are shown. (C) In vitro deubiquitination
of Skp2 by USP13 as described in Materials and Methods. Immunopuried
FLAG-USP13 and immunoprecipitated myc-Skp2 were obtained from HeLa
cells. Reactions were immunoblotted for Ub and Skp2.
D
G
B
C
A
E
Skp2
p27
Cdh1
α−tubulin
mock empty vector Cdh1 RNAi
TM (2.5μg/ml) + + + + + + + + + + + +
DMSO + + +
MG-132
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Skp2
p27
β−actin
TM (2.5μg/ml)
DMSO
MG-132
+ + +
+
Ufd1
1 2 3 4
F
prolonged ER stress
Skp2 ubiquitination
p27 protein stabilization
G1 delay
cyclin D1 downregulation
and other pathways
USP13
nuclear Ufd1
Skp2
transcription
p27
transcription
0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24 h
0.5μg/ml TM 2.5μg/ml TM
DMSO
GRP78
Skp2
Ufd1
α−tubulin
p27
NHK
HSP90
G1/S
CHX (25 g/ml) 0 1 2 3 0 1 2 3 h
7h release from G1/S
μ
G1/S arrest
G2/M arrest
GFP (CFTR)
β−actin
CHX (25 g/ml) 0 0.5 1 2 0 0.5 1 2 h
NHK
CHX (25 g/ml) 0 1 2 3 0 1 2 3 h
G1/S arrest
G2/M arrest
HSP90
μ
μ
Ufd1
Skp2
p27
GRP78
β−actin
0.05 0.1 0.25 0.5
1 2 3 4 5
TMg/ml)
DMSO
H
Ufd1
tubulin
TM (1μg/ml) + +
DMSO + +
empty vector FLAG-Ufd1
0
1
2
3
4
5
6
7
8
FLAG-Ufd1control
% increase in G1
p27
β−actin
TM (2.5μg/ml) 0 8 h
control
p27
β−actin
TM
(2.5μg/ml)
control
p27-KD 1
p27-KD 2
% increase in G1
0
5
10
15
20
2.5
μ
g/ml, 8h
2.5
μ
g/ml, 4h
p27-kD, #2p27-KD, #1control
Fig. 5. ER stress regulates the Ufd1-Cdh1-Skp2-p27 axis to delay progression
through G1. (A) HeLa cells were treated with DMSO or increasing concen-
trations of TM (0.05, 0.1, 0.25, and 0.5 μg/mL) for 20 h. Lysates were
immunoblotted for Ufd1, Skp2, p27, and GRP78 (a marker of UPR). (B) Cells
were treated with DMSO or with 0.5 μg/mL or 2.5 μg/mL of TM over a 24-h
course. At the indicated time points, samples were collected for immunoblot
and FACS analyses. Fig. S5C shows the corresponding quantication of G1
delay based on FACS analysis. (C)(Left) HeLa cells were treated with DMSO
or with 2.5 μg/mL of TM alone or together with MG-132 (2 or 5 μM) for 12 h,
and then collected for immunoblot analysis for the indicated proteins.
(Right) HeLa cells transfected with control or Cdh1-specic shRNA for 24 h
were treated with DMSO or with 2.5 μg/mL of TM alone or together with
MG-132 (5, 10, or 15 μM) for 12 h. Total cell extracts were immunoblotted for
the indicated proteins. (D) HeLa cells transfected with empty vector or FLAG-
Ufd1 were treated with DMSO or TM (1 μg/mL for 8 h). (Upper) Western blot
of Ufd1. (Bottom) Increase in G1 at 8 h after the addition of TM compared
with DMSO treatment. Fig. S6C shows FACS histograms. (E) HeLa cells were
transfected with either control or two different p27-specic shRNAs, as in-
dicated, for 24 h. (Upper) Western blot analysis of p27 in control and p27-KD
cells treated with TM (2.5 μg/mL for 8 h). (Bottom) Percent increase in G1 at 4
h and 8 h after TM addition relative to 0 h. The percentage of cells in G1 at
0 h is set as the baseline, with a 0% increase in G1. (F) Half-life analyses of
GFP-CFTR and NHK proteins with CHX treatment in double thymidine-
arrested (G1/S) or nocodazole-arrested (G2/M) HeLa cells. Total lysates were
immunoblotted for GFP or α1-antitrypsin/NHK. Fig. S8A veries cell cycle
synchronization by FACS, and Fig. S8B quanties protein half-lives. (G) Half-
life of NHK protein in double thymidine-arrested G1/S HeLa cells or cells
released from double-thymine arrest for 7 h (approximately corresponding
to G2/M). Total lysates were immunoblotted for α1-antitrypsin/NHK. Fig S8C
presents verication of cell cycle synchronization by FACS, and
Fig. S8D
shows quantication of NHK half-life. (H) In the proposed model, down-
regulation of Ufd1 after prolonged ER stress reduces recruitment of USP13
to Skp2, there by resulting in ubiquitin-dependent degradation of Skp2 by
APC/C
Cdh1
. Consequently, levels of p27 increase, contributing to G1 arrest that
supports degradation of misfolded proteins.
Chen et al. PNAS
|
May 31, 2011
|
vol. 108
|
no. 22
|
9123
CELL BIOLOGY
Page 5
observed in G1/S-arrested cells compared with the G2/M-syn-
chronized cells (Fig. 5F and Fig. S8 A and B). The impaired
clearance of ERAD substrates seen in G2/M-arrested cells did not
result from nocodazole-induced toxicity, as the degradation of
NHK was still slower in the G2/M population obtained after a 7-h
release from G1 arrest (Fig. 5G and Fig. S8 C and D). These data
suggest that ER stress-induced G1 delay serves to facilitate deg-
radation of misfolded proteins.
This hypothesis prompted us to examine the cell cycle status of
broblasts bearing two alleles of the mutant CFTR gene (CFTR-
ΔF508) isolated from a cystic brosisaffected individual. If G1
phase of the cell cycle is more conducive to ERAD, then the
constitutive expression of endogenous mutant CFTR targeted to
ERAD in these cells may render them more prone to G1 delay.
We found that these cells not only had a slightly higher basal G1
population compared with normal human broblasts (Fig. S9A),
but also had a greater increase in G1 when treated with TM (Fig.
S9B). These observations further support the physiological rel-
evance of ER stress-induced cell cycle arrest.
Discussion
The present study provides insight into cell cycle control in re-
sponse to ER stress. We have demonstrated that ER stress-de-
pendent control of Ufd1 modulates Skp2 protein expression and
consequently p27, to delay progression through G1 and facilitate
the clearance of misfolded proteins. Mechanistically, Ufd1 pro-
motes the interaction between USP13 and Skp2 to maintain
steady-state Skp2 levels. Fig. 5H presents a proposed model of
this process.
In support of our ndings, TM treatment has been reported
to trigger G1 cell cycle arrest by inhibiting cyclin D1 translation
through the activation of PKR-like endoplasmic reticulum kinase
(PERK) (22). Our ndings reveal a cyclin D1-independent mech-
anism for attaining G1 delay after protein misfolding through
regulation of the Cdh1-Skp2-p27 axis at the level of protein sta-
bility. We have begun to examine the efciency of ERAD in the
context of the cell cycle, a currently underexplored area. Our data
raise the intriguing possibility that cell cycle arrest is not simply
a passive response that provides cells time to restore homeostasis,
but creates conditions conducive to the degradation of misfolded
proteins. Considering that CFTR and NHK are characterized
ERAD substrates subjected to degradation by the ER resident
E3 ligases gp78, RNF5/RMA1, and Hrd1 (23, 24), the cell cycle
dependent degradation of ERAD substrates we observed may
be the result of cell cycleregulated expression or activity of these
ligases and other ERAD components.
We also have presented an initial characterization of USP13 as
a di-ubiquitin able to process K48-linked ubiquitins and antag-
onize APC/C
Cdh1
-mediated ubiquitination of Skp2 in an Ufd1-
dependent manner. Our mapping of Ufd1-USP13 interaction
sites shows that Ufd1 binds p97 and USP13 through distinct
domains. Our observation that Ufd1 binds Skp2 proteins not in
complex with Skp1-Cul1 suggests that Ufd1 positively regulates
the abundance of free Skp2 proteins to control the formation of
SCF
Skp2
complexes.
Based on our ndings, we propose a model in which the functions
of Ufd1 are regulated temporally and spatially under conditions
activating the UPR. In the immediate phase, Ufd1 in cooperation
with p97 and Npl4 in the cytosol contribute to the retrotranslocation
of misfolded proteins from the ER, whereas in a delayed response,
down-regulation of nuclear Ufd1 mediates G1 cell cycle delay. Both
responses serve to clear misfolded proteins. Such temporal and
spatial modulation of Ufd1 expression shown here could not have
been detected in previous yeast studies using loss-of-function
mutants that yielded all-or-none phenotypes (3, 5).
The identication of Ufd1-dependent recruitment of USP13
to control Skp2 stability and G1 delay provides mechanistic
insights into cell cycle regulation under ER stress while also
pointing to the implications of a cell cycledependent control
of ERAD.
Materials and Methods
Protocols for the immunoblotting, immunoprecipitations, cell cycle syn-
chronization, siRNA/shRNA knockdown, qRT-PCR, assays for measuring ki-
nase, degradation, ubiquitination and deubiquitination, cell fractionation
and the details of cell lines and expression plasmids used in this study can be
found in SI Materials and Methods.
ACKNOWLEDGMENTS. We thank R. Agami, O. Coux, P. Hydbring, R.
Kopito, H . Meyer, M. Pagano, D. Ron, and D. Wolf for the essen tial
reagents used in this study. We also thank Y. Altman for help with FACS
analysis, and members of the Ronai laboratory for stimulating discus-
sions. T his work was supported by Nati onal Cancer Institu te Grants
CA097105 and CA78419 (to Z.A.R. ). M.C. has been part of the Molecular
Pathology PhD Program at the University of California San Diego and was
supported in part by Molecular Patholog y C ancer Training Grant
5T32CA077109. G.J.G. was supported in part by a Sass Foundation
postdoctoral fell owship.
1. Johnson ES, Ma PC, Ota IM, Varshavsky A (1995) A proteolytic pathway that
recognizes ubiquitin as a degradation signal. J Biol Chem 270:1744217456.
2. Park S, Isaacson R, Kim HT, Silver PA, Wagner G (2005) Ufd1 exhibits the AAA-ATPase
fold with two distinct ubiquitin interaction sites. Structure 13:9951005.
3. Ye Y, Meyer HH, Rapoport TA (2001) The AAA ATPase Cdc48/p97 and its partners
transport proteins from the ER into the cytosol. Nature 414:652656.
4. Cao J, et al. (2007) Ufd1 is a cofactor of gp78 and plays a key role in cholesterol
metabolism by regulating the stability of HMG-CoA reductase. Cell Metab 6:115128.
5. Bays NW, Wilhovsky SK, Goradia A, Hodgkiss-Harlow K, Hampton RY (2001) HRD4/
NPL4 is required for the proteasomal processing of ubiquitinated ER proteins. Mol
Biol Cell 12:41144128.
6. Ballar P, Shen Y, Yang H, Fang S (2006) The role of a novel p97/valosin-containing
protein-interacting motif of gp78 in endoplasmic reticulumassociated degradation.
J Biol Chem 281:3535935368.
7. Lin PH, Chiang MT, Chau LY (2008) Ubiquitin-proteasome system mediates heme
oxygenase-1 degradation through endoplasmic reticulumassociated degradation
pathway. Biochim Biophys Acta 1783:18261834.
8. McConnell E, Lass A, Wójcik C (2007) Ufd1-Npl4 is a negative regulator of cholera
toxin retrotranslocation. Biochem Biophys Res Commun 355:10871090.
9. Nowis D, McConnell E, Wójcik C (2006) Destabilization of the VCP-Ufd1-Npl4 complex
is associated with decreased levels of ERAD substrates. Exp Cell Res 312:2921 2932.
10. Ramadan K, et al. (2007) Cdc48/p97 promotes reformation of the nucleus by
extracting the kinase Aurora B from chromatin. Nature 450:12581262.
11. Cao K, Nakajima R, Meyer HH, Zheng Y (2003) The AAA-ATPase Cdc48/p97 regulates
spindle disassembly at the end of mitosis. Cell 115:355367.
12. Carrano AC, Eytan E, Hershko A, Pagano M (1999) SKP2 is required for ubiquitin-
mediated degradation of the CDK inhibitor p27. Nat Cell Biol 1:193199.
13. Pines J (2006) Mitosis: A matter of getting rid of the right protein at the right time.
Trends Cell Biol 16:5563.
14. Peters JM (2006) The anaphase-promoting complex/cyclosome: A machine designed
to destroy. Nat Rev Mol Cell Biol 7:644656.
15. Gutierrez GJ, Ronai Z (2006) Ubiquitin and SUMO systems in the regulation of mitotic
checkpoints. Trends Biochem Sci 31:324332.
16. Wei W, et al. (2004) Degradation of the SCF component Skp2 in cell-cycle phase G1 by
the anaphase-promoting complex. Nature 428:194198.
17. Bashir T, Dorrello NV, Amador V, Guardavaccaro D, Pagano M (2004) Control of the SCF
(Skp2-Cks1) ubiquitin ligase by the APC/C(Cdh1) ubiquitin ligase. Nature 428:190193.
18. Sowa ME, Bennett EJ, Gygi SP, Harper JW (2009) Dening the human deubiquitinating
enzyme interaction landscape. Cell 138:389403.
19. Bruderer RM, Brasseur C, Meyer HH (2004) The AAA ATPase p97/VCP interacts with its
alternative co-factors, Ufd1-Npl4 and p47, through a common bipartite binding
mechanism. J Biol Chem 279:4960949616.
20. Catic A, et al. (2007) Screen for ISG15-crossreactive deubiquitinases. PLoS ONE 2:e679.
21. Brewer JW, Hendershot LM, Sherr CJ, Diehl JA (1999) Mammalian unfolded protein
response inhibits cyclin D1 translation and cell-cycle progression. Proc Natl Acad Sci
USA 96:85058510.
22. Brewer JW, Diehl JA (2000) PERK mediates cell-cycle exit during the mammalian
unfolded protein response. Proc Natl Acad Sci USA 97:1262512630.
23. Christianson JC, Shaler TA, Tyler RE, Kopito RR (2008) OS-9 and GRP94 deliver mutant
α1-antitrypsin to the Hrd1-SEL1L ubiquitin ligase complex for ERAD. Nat Cell Biol 10:
272282.
24. Morito D, et al. (2008) Gp78 cooperates with RMA1 in endoplasmic reticulum
associated degradation of CFTRDeltaF508. Mol Biol Cell 19:13281336.
9124
|
www.pnas.org/cgi/doi/10.1073/pnas.1100028108 Chen et al.
Page 6
Supporting Information
Chen et al. 10.1073/pnas.1100028108
SI Materials and Methods
Cell Culture and Transfection. HeLa, HeLa-S3, HEK-293T, and
melanoma CHL1 cells were grown in DMEM with 10% bovine
serumat37°Cwith5%CO
2
. HFF-1, BJT, and cystic brosis -
broblasts (GM01348; Coriell Cell Repositories) were grown in
DMEM with 10% FBS. HeLa cells stably expressing Ufd1-specic
or USP13-specic shRNAs (MISSION pLKO.1-Puro
3
shRNA
bacterial glycerol stock; Sigma-Aldrich) were prepared by lenti-
viral infection (with lentiviral particles packaged with the shRNA
vectors generated in HEK-293T cells) and selected with 1 μg/mL
of puromycin. Drug-selected pools were isolated, and the ef-
ciency of knockdown was validated by qRT-PCR and immunoblot
analysis for expression of Ufd1 or USP13. HeLa cells infected with
empty pLKO.1-Puro
3
vector served as a control. Analysis of in-
dividual drug-resistant Ufd1-KD pools conrmed the major bio-
chemical phenotypes shown in Fig. 1B (Fig. S2).
For transient transfection of plasmids in HeLa cells, 80%-
conuent cells grown in 10-cm dishes were transfected with 5 μg
of each of the indicated plasmids, using Lipofectamine 2000 (In-
vitrogen) following the manufacturersprotocol.Cellswerehar-
vested within 2436 h after transfection. For transient transfection
of siRNA, 60%-conuent cells grown in 60-mm plates were
transfected with 100 nM siRNA with RNAiMAX (Invitrogen)
according to the manufacturersspecications. NonspecicsiRNA
served as a control. For transient transfection in HEK-293T cells,
50%-conuent cells grown in 10-cm dishes were transfected with
10 μgofDNAcalcium phosphate complexes, generated by bub-
bling the DNA, CaCl
2
, and Hepes-buffered saline mixture. Cells
were analyzed 24 h after transfection.
Expression Plasmids. Ufd1 was subcloned from pET26-Ufd1-His (a
generous gift from Dr. Hemmo Meyer, Institute of Biochemistry,
Zurich, Switzerland) into pGEX4T3. Ufd1 (a 940-bp insert in-
cluding restriction sites for BamH1 on the 5 end and EcoRV on
the 3 end) was subcloned into the BamH1-EcoRV sites of either
the pEF-FLAG vector or the pEF-HA vector for mammalian
expression. C-terminal truncation variants of pEF-FLAG-Ufd1
(N215, N220, N241, N260, N280, and N300) were generated by
the insertion of two stop codons after the indicated amino acid
from the N terminus using site-directed mutagenesis. pEF-FLAG-
ΔUfd1-261280 was generated by site-directed mutagenesis using
the Stratagene QuikChange Site-Directed Mutagenesis Kit. The
integrity of all constructs was conrmed by sequencing. pCDNA3-
FLAG-Skp2 and pCDNA3-FLAG-Skp2ΔF-box were a generous
gift from Dr. Michele Pagano (New York University, New York,
NY). pCDNA-myc-Skp2 was a generous gift from Dr. Dieter
Wolf (Sanford-Burnham Medical Research Institute, La Jolla,
CA). The pEF-HA-ubiquitin vector has been described previously
(1). pCMV6-XL-USP13 was obtained from OriGene (SC117705)
and was cloned into the pEF-FLAG vector using the restriction
enzymes SmaI (5) and NotI (3). pSUPER-Cdh1-shRNA was a
generous gift from Dr. Reuven Agami (Netherlands Cancer In-
stitute, Amsterdam, The Netherlands). pLKO.1-USP13-shRNAs
were purchased from Open Biosystems (RHS4533). pRS-p27-
shRNA and control pRS-shRNAs were purchased from OriGene
(TR314022). GFP-CFTR and pCDNA3-NHK constructs were
generous gifts from Dr. Ron Kopito (Stanford University, Stan-
ford, CA). pPL8-cyclin D1 (from mouse) was a generous gift from
Dr. Per Hydbring (Dana-Farber Cancer Institute, Boston, MA).
siRNA/shRNA Target Sequences. CDKN1B/p27-shRNAs: CGGC-
TAACTCTGAGGACACGCATTTGGTG, AAGGAAGCGA-
CCTGCAACCGACGATTCTT; Ufd1-shRNAs: CTGGCAAT-
AGACTGGATGGAA, TGGGCTACAAAGAACCCGAAA,
CAGCCGACTTAACATTACCTA, CCCAATCAAGCCTGG-
AGATAT; Ufd1-3UTR-siRNAs: CAGCAGGAGCAATTCT-
GCATCCCTA, CCTATGAAGGTGACTAAATTGTCTA;
USP13-shRNAs: CGCCTGATGAACCAATTGATA, CGATT-
TAAATAGCGACGATTA, CCGGTGAAATCTGAACTCAT-
T; Cdh1-shRNA: TGAGAAGTCTCCCAGTCAG (1).
Antibodies. The following commercially available antibodies were
used: Ufd1L (19, 1:2,000 dilution; BD Transduction Laborato-
ries), Skp2 (H-435, 1:500; Santa Cruz Biotechnology), p27 (57,
1:500; BD Transduction Laboratories), Cdh1 (DCS-266, 1:1,000;
Abcam), cyclin B1 (GNS1, 1:1,000; Santa Cruz Biotechnology),
cyclin E2 (1142-1, 1:1,000; Epitomics), Cdk2 (M2, 1:500; Santa
Cruz Biotechnology ), GRP78 (N-20, 1:1,000; Santa Cruz Bio-
technology), FLAG (M2, 1:5,000; Sigma-Aldrich), GFP (mouse,
1:1,000; Zymed), α1-antitrypsin/NHK (mouse, 1:1,000; U.S. Bi-
ological), ubiquitin (FL-76, 1:1,000; Santa Cruz Biotechnology),
c-myc (9E10, 1:2,000; Santa Cruz Biotechnology), USP13
(HPA004827, 1:1,000; Sigma-Aldrich and 2141C, 1:500; Ab-
gent), HSP90 (F-8, 1:1,000; Santa Cruz Biotechnology), Lamin
A/C (636, 1:1,000; Santa Cruz Biotechnology), cyclin D1 (2922,
1:1,000; Cell Signaling), Npl4 (H-300, 1:1,000; Santa Cruz Bio-
technology), p97 (ab19444, 1:1,000; Abcam), and α-tubulin (TU-
02, 1:2,000; Santa Cruz Biotechnology).
Immunoblotting and Immunoprecipitation. For Western blot anal-
ysis of total cell lysates, cells were detached from plates with
0.05% trypsin-EDTA, washed with cold PBS, and lysed on ice for
20 min after resuspension in IP buffer [50 mM Tris-HCl (pH 7.5),
150 mM NaCl, 0.5% Nonidet P-40, 5 mM EDTA, 5 mM EGTA,
20 mM NaF, 100 μM sodium-orthovanadate, 2 mM β-glycer-
ophosphate, 1 mM DTT, 1 mM PMSF, 4 mg/mL of aprotinin,
100 μM leupeptin, and 2 mg/mL of pepstatin A). After 20 min of
centrifugation (at 16,000 × g and 4 °C), supernatants were saved,
and the protein concentration was performed using Coomassie
Plus protein assay reagent (Thermo Scientic).
For coimmunoprecipitation in vivo, cells expressing the pro-
teins of interest for 24 h were washed with cold PBS and lysed with
IP buffer. The target protein was immunoprecipitated from total
cell lysates containing 2 mg of protein with 1 μg of target-specic
antibody overnight at 4 °C in a rotating wheel, followed by in-
cubation with 30 μL of a 50% slurry of protein G-agarose for 2 h
at 4 °C with rotation. Immunoprecipitates were washed four times
with IP buffer, resuspended in loading buffer, and analyzed by
SDS/PAGE and Western blot.
Sequential Immunoprecipitations. HeLa cells were transfected with
FLAG-Ufd1, myc-Skp2, and USP13, or with USP13 alone (as
a negative control). FLAG-Ufd1 was immunoprecipitated from
8 mg of total lysate by incubation with α-FLAGM2 beads for 2 h
at 4 °C, washed three times in IP buffer and twice in TBS, and
then eluted with 500 μg/mL of 3X FLAG peptide. The eluate
was subsequently used for IP of myc-Skp2 for 8 h at 4 °C. Beads
were washed three times in IP buffer, run on SDS/PAGE, and
immunoblotted for USP13, Skp2, and Ufd1.
Cell Cycle Synchronization and FACS. For G1/S synchronization of
HeLa cells by double-thymidine block, cells were incubated with
2mMthymidine for18h,releasedintofreshmediafor12h,andthen
incubated with 2 mM thymidine again for 16 h. Cells were then
washed twice with PBS and given fresh media to release them from
Chen et al. www.pnas.org/cgi/content/short/1100028108 1of10
Page 7
the arrest. For G2/M synchronization by thymidine-nocodazole
block, HeLa cells were incubated with 40 ng/mL of nocodazole for
16 h after release from the rst thymidine block as detailed earlier.
For cell cycle analysis by FACS, cells were washed with cold
PBS, xed with 70% ethanol in PBS for 1 h at 20 °C, and washed
again with cold PBS. Cells were then incubated with PBS with 50
μg/mL of propidium iodide and 100 μg/mL of RNase A for 1 h at
room temperature. DNA content was analyzed by three-color
FACS, and quantication of G1, S, and G2/M populations was
performed using FlowJo software.
SYBR-Green qRT-PCR. Total RNA was extracted from cells grown in
10-cm plates (RTN70-1KT; Sigma-Aldrich). Up to 2 μg of total RNA
was used per 20 μL of cDNA reverse transcription (Applied Bio-
systems High-Capacity cDNA Reverse-Transcription Kit). A total of
5 μL of 20% vol/vol diluted cDNA was used per 25 μLofqRT-PCR
with SYBR GreenER qPCR SuperMix Universal (Invitrogen) and
target gene-specic primers. Forty cycles of qPCR and analysis were
performed using Stratagene Mx3000p. Human cyclophilin or
GAPDH was used as a control to normalize transcript levels of
experimental samples. The specicity of all primer sets was validated
by dissociation curve analysis. qPCR primer sequences were as fol-
lows: Ufd1: forward 1, GAGGGAAGATAATTATGCCAC; re-
verse 1, CTTCCAAGAGTAAGTTCTGC; forward 2, GAGG-
GAAGAGCCGAC-TTAAC; reverse 2, CTCTGAGGTTGGA-
ATTTGGAG; forward 3, GACATGAACGTGGACTTTG; re-
verse 3, GGAATTCCTCTTTTAATATC; forward 4, GGGCCT-
AATGACAGGTCAG; reverse 4, CTTGAAGGTTGACGC-
TCTCC; Skp2: forward, GCTATGCACAGGAAGCACCT; re-
verse, CCCATGAAACACCTGGAAAG; p27: forward, CGG-
CTAACTCTGAGGACACG; reverse, CTTAATTCGAGCTG-
TTTACG; USP13: forward, GCATGGGACAGAGAATGGGC;
reverse, CCTAACTTAGTCATCTGTGTG; cyclophilin: forward,
GACCCAACACAAATGGTTC; reverse, AGTCAGCAATGGT-
GATCTTC; GAPDH: forward, GAAGGTGAAGGTCGGAGTC;
reverse, ATGGGATTTCCATTGATGAC.
Cdk Kinase Assays. Cells were synchronized in G1 and released into
fresh media. At the indicated time points, cells were harvested and
lysed with modied IP buffer [0.1% Nonidet P-40, 25 mM Tris-HCl
(pH 7.5), 50 mM NaCl, 5 mM EGTA, 60 mM β-glycerophosphate,
20 mM NaF, 100 μm sodium-orthovanadate, 1 mM DTT, 1 mM
PMSF, 100 μM leupeptin, and 4 mg/mL of aprotinin]. Approxi-
mately 5 mg of total cell lysate was used to immunoprecipitate
CDK2 with 0.5 μg of CDK2 antibody and CDK1 with 0.5 μgof
cyclin B1 antibody. Immunoprecipitates were washed three times
with histone
1
H buffer [15 mM MgCl
2
,20mMEGTA,80mM
β-glycerophosphate (pH 7.4), 1 mM DTT, 1 mM PMSF, 4 mg/mL
aprotinin, and 100 μM leupeptin). Kinase reactions were per-
formed by incubating each pellet with 35 μLofhistone
1
H buffer
supplemented with 0.1 μgofhistone
1
H, 0.2 μLof
32
P-γATP, and
50 μM cold ATP. Reactions were incubated for 30 min at 30 °C in
a shaker at 800 rpm and then stopped with sample buffer and
heat-denatured. Samples were run in SDS/PAGE gel, dried, and
analyzed with a PhosphorImager system.
APC/C
Cdh1
-Dependent Degradation Assay. Interphase Xenopus laevis
egg extracts were prepared as described previously (2) and sup-
plemented with 6xHis-Cdh1 protein puried from pelleted bacu-
lovirus-infected insect cells (a generous gift from Olivier Coux,
Centre de Recherche de Biochimie Macromoleculaire-Centre
National de la Recherche Scientique, Montpellier, France). Skp2
protein was in vitro translated with [
35
S]-methionine and added to
the extract for degradation assays as described previously (2) in
the presence of bacterially puried GST-Ufd1 as indicated. A
sample of this mixture was recovered at the indicated times, and
35
S-Skp2 levels were measured with the PhosphorImager.
In Vivo Ubiquitination Assays. After HeLa cells were transfected
with myc-Skp2 and HA-Ub plasmids for 24 h, cells were treated
with 5 μM MG-132 for 5 h, then harvested and resuspended in
50 μL of cold TBS buffer (10 mM Tris-HCl, pH 8), followed by
60 μL of 2% SDS in TBS. The resulting suspension was heat-
denatured for 10 min and then cooled on ice for 5 min. Then 900
μL of 1% Triton X-100 in TBS supplemented with protease in-
hibitors was added to the mixture and sonicated. This mixture was
spun down, and 2 mg of supernatant was incubated with 1 μgofc-
myc antibody rotating overnight at 4 °C, followed by an additional
1 h of incubation with 30 μL of a 50% slurry of G protein agarose.
Precipitates were washed once with 1 mL of 1% Triton X-100 and
0.1% SDS in TBS, twice with 0.5 M LiCl and 1% Triton X-100 in
TBS, and once with 1% Triton X-100 in PBS.
Deubiquitination Assays. Internally quenched uorescence-diubi-
quitin (IQF-DiUb, K48-2) of K48 linkage was obtained from
LifeSensors. FLAG-USP13 was expressed and immunoprecipi-
tated from HeLa cells using α-FLAG M2 afnity gel (A2220;
Sigma-Aldrich), washed three times in IP buffer and twice in
TBS [25 mM Tris-HCl (pH 7.6), 150 mM NaCl, 1 mM DTT],
and then eluted with 500 μg/mL of 3X FLAG peptide (F4799;
Sigma-Aldrich). Immunopuried FLAG-USP13 was quantied
against BSA by Coomassie blue staining (Fig. S4A). Between 20
and 30 nM USP13 was used per reaction with 200 nM K48-linked
DiUb in assay buffer [50 mM Tris-HCl (pH 8.0), 10 mM DTT]
in a 96-well plate (Greiner BioONE 655076). Kinetic reads (60
min) were performed at 25 °C using a FlexStation 3 spectrometer
(Molecular Devices) with 540/580nm excitation/emission and
mirror cutoff of 570 nm. Results were analyzed using SoftMax
Pro (Molecular Devices).
For deubiquitination of Skp2 by USP13, myc-Skp2 and HA-
ubiquitin were coexpressed in HeLa cells. Cells were treated with
20 μM MG-132 for 3 h before lysis and IP using c-myc antibodies
conjugated to agarose beads. Immunoprecipitates were washed
three times in IP buffer and twice in TBS. FLAG-USP13 was
immunopuried separately (see above). Then 2030 nM USP13
was incubated with Skp2 for 1 h at 25 °C. Reactions were stop-
ped by the addition of sample buffer, run on SDS/PAGE, and
immunoblotted with Ub and Skp2 antibodies.
CHX and Degradation Assays. Asynchronous HeLa cells were
treated with 25 μg/mL of CHX and collected at the indicated times
after the addition of CHX for immunoblot analysis of proteins of
interest. To determine the half-life of NHK and GFP-CFTR in
synchronized cell cultures, cells were rst arrested in G1/S by
double-thymidine block or in G2/M by thymidine-nocodazole
block and then transfected with either NHK or GFP-CFTR
during the release after the rst thymidine treatment, followed by
the addition of 25 μg/mL of CHX to G1/S- or G2/M-arrested cells.
Finally, cells were harvested at the indicated time points after the
addition of CHX for immunoblot analysis.
Cytosolic and Nuclear Fractionation. Fractionation was performed
using reagents from the Active Motif 40010 Nuclear Extract Kit. In
brief, cells were detached from plates with 0.05% trypsin-EDTA
and washed with cold PBS. Cells were lysed with a hypotonic buffer.
Centrifugation yielded the cytosolic fraction as the supernatant and
the nucleus fraction as the pellet. The nuclear pellet was further
lysed with Complete Lysis Buffer (included in the Active Motif
40010 Nuclear Extract Kit; Roche Applied Science), following the
manufacturers instructions, to yield the nuclear extract.
1. Brummelkamp TR, et al. (2002) A system for stable expression of short interfering RNAs
in mammalian cells. Science 296:550.
2. Castro A, et al. (2006) Ubiquitin-mediated protein degradation in Xenopus egg
extracts. Methods Mol Biol 322:223234.
Chen et al. www.pnas.org/cgi/content/short/1100028108 2of10
Page 8
A
)%(levelANRmUfd1evitaleR
control
1234567
drug-selected pools
0
30
60
90
120
150
***
B
drug-selected pools
0
20
40
60
80
100
120
control
control
TM
1234567
* * *
*
)%(levelUfd1 protein evitaleR
control
control+TM
1234 567 1234 567
β-actin
Ufd1
GRP78/BiP
drug-selected pool drug-selected pool
control
control+TM
Ufd1 primer pair 1
Ufd1 primer pair 2
Ufd1 primer pair 3
Ufd1 primer pair 4
Fig. S1. Validation of Ufd1 knockdown in HeL a cells by qRT-PCR and immunoblot analysis. (A) By SYBR-green qRT-PCR, four different Ufd1-specic primer
pairs were used to quantify Ufd1 transcripts in drug-selected pools of HeLa cells infected with lentivirus packaged with empty shRNA vector (control) or Ufd1-
shRNA. The amount of Ufd1 transcript was normalized against GAPDH and is presented as a percentage of the Ufd1 transcript in control cells (set as 100%).(B)
(Upper) Protein levels of Ufd1 in each drug-selected pool infected with Ufd1-shRNA were measured by immunoblot analysis. (Lower) Quantication of Ufd1
levels by immunoblot analysis using Odyssey software (LiCOR), presented as percentage of Ufd1 levels in the control (set as 100%). Asterisks denote drug-
resistant pools with a statistically signicant difference in the amount of Ufd1 transcript compared with the control. Pool 2, with a signicant down-regulation
of Ufd1, was used throughout the study and is referred to as Ufd1-KD cells. TM corresponds to treatment with TM (2.5 μg/mL for 8 h).
C
95
72
55
43
34
130
170
MW (kDa)
IP FLAG
lysate
IB: Ub
IB: FLAG
Ufd1
α-tubulin
Cdh1
(Ub) -Skp2
n
FLAG-Skp2 + + +
HA-Ub + + +
empty vector + +
Cdh1-RNAi +
control
Ufd1-KD
A
Ufd1
Skp2
p27
β-actin
siRNA
control
Ufd1, #1
Ufd1, #2
Ufd1, #1+2
HeLa-S3
control
Ufd1-KD, #1
Ufd1-KD, #2
shRNA
HeLa
Ufd1
Skp2
β-actin
B
60
80
100
120
Ufd1-KD
control
5h4h2h1h0h
Relative Skp2 protein levels
cycloheximide addition (h)
Fig. S2. Ufd1 affects Skp2 protein levels and ubiquitination. (A) Down-regulation of Skp2 in cells with stable Ufd1 knockdown by shRNA or with transient
Ufd1 knockdown by siRNA. (Left) Immunoblots of endogenous Ufd1 and Skp2 from two independently derived puromycin-resistant pools of stable Ufd1-KD
cells. (Right) Immunoblots of endogenous Ufd1, Skp2, and p27 from HeLa-S3 cells transfected with control siRNA or two individual Ufd1-specic siRNAs. (B)
Quantication of endogenous Skp2 protein levels, normalized to β-actin, in control and Ufd1-KD cells collected at the indicated times after the addition of 25
μg/mL of CHX. Relative values are presented as percentages of Skp2/β-actin at time 0 h of CHX addition (set as 100%). (C) APC/C
Cdh1
mediates ubiquitination of
Skp2 in Ufd1-KD cells. For in vivo ubiquitination of Skp2, FLAG-Skp2 and HA-Ub were expressed in control, Ufd1-KD, or Ufd1 and Cdh1 double-knockdown
cells. Immunoprecipitates obtained with FLAG antibodies were immunoblotted for Ub to detect polyubiquitinated Skp2, denoted as (Ub)n-Skp2. Control refers
to cells infected with empty shRNA plasmid. Lysates were immunoblotted for Cdh1 and Ufd1 to conrm knockdown.
Chen et al. www.pnas.org/cgi/content/short/1100028108 3of10
Page 9
endog Skp2
FLAG
FLAG-USP13 + +
empty vector + +
lysate
IP: FLAG
IgG
endog Skp2
FLAG
IgG
lysate IP: FLAG
FLAG-Ufd1 + +
empty vector + +
A
B
D
C
Binding ability of Ufd1 variants to USP13, Skp2, and p97
Ufd1(1-307)
Ufd1-N215
Ufd1-N220
Ufd1-N241
Ufd1-N260
Ufd1-N280
Ufd1-N300
Ufd1- 261-280
USP13 Skp2 p97
+ + +
- - -
- - -
- - +
- - +
+ + +
+ + +
- nd +
NC
USP13
FLAG
p97
+ + + + + + + + + +
+ +
+ +
+ +
+ +
+ +
lysate IP: FLAG
USP13
FLAG-Ufd1
FLAG- 261-280
FLAG-N260
FLAG-N280
empty vector
*
endog Cul1
myc-Skp2
FLAG-Ufd1
IgG
myc-Skp2 + + + + + + + +
+ + + +FLAG-Ufd1
10%
FLAG-IP elution
10%
myc-IP
lysate
step 1
IP FLAG-Ufd1
from cell lysate
step 2
FLAG peptide
elution
step 3
IP myc-Skp2
from elution
Skp2
IgG
Ufd1
FLAG-Skp2
HA-Ufd1
FLAG-Skp2 F-box
+ +
+
+ + +
IP
IgG FLAG FLAG
Δ
Fig. S3. Ufd1 with internal deletion of amino acids 261280 does not bind USP13 in vivo. (A) HEK-293T cells were transfected with USP13 together with empty
vector, FLAG-Ufd1, the C-terminal truncation Ufd1 mutants FLAG-N260 and FLAG-N280 (with amino acids 1260 and 1280, respectively), and the internal
deletion Ufd1 mutant of FLAG-Δ261280 that lacks amino acids 261280. FLAG immunoprecipitates were immunoblotted for USP13, p97, and FLAG. (B)
Schematic depicting the binding capacity of the various Ufd1 C-terminal deletion mutants to USP13, Skp2, and p97 based on Fig. 3 A and D and Fig. S3A. nd,
not determined. (C) Ufd1 targets free Skp2 proteins. (Left) Sequential immunoprecipitationof FLAG-Ufd1 and myc-Skp2 from HEK-293T cells, followed by
immunoblotting of endogenous Cul1. An asterisk marks the position of Cul1. (Right) HA-Ufd1 was expressed with either FLAG-Skp2 or FLAG-Skp2ΔF-box in
HEK-293T cells. FLAG immunoprecipitates were immunoblotted for Ufd1. (D) Exogenousendogenous coimmunoprecipitations. HeLa cells were transfected
with empty vector, FLAG-Ufd1, or FLAG-USP13. FLAG immunoprecipitates were immunoblotted for endogenous Skp2.
Chen et al. www.pnas.org/cgi/content/short/1100028108 4of10
Page 10
B
C
USP13 transcript levels in USP13-KD cells
Relative Quantity
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
shRNA mixshRNA 3shRNA 2shRNA 1mock
A
1 2 3 4 5 6 7 8
1: 60μg lysate, before IP
2: 60μg lysate, after IP
3: all FLAG-beads, after elution
4: 5% of total elution
(30% elution used per reaction)
5: BSA, 0.01μg
6: BSA, 0.1μg
7: BSA, 1μg
8: BSA, 3μg
Quantification of immunopurified FLAG-USP13 by signal intensity comparison using Coomassie stain
An estimated 20-30nM of FLAG-USP13 were used
in each deubiquitination reaction
USP13
-10
0
10
20
30
40
50
60
0 102030405060
buffer only
FLAG-USP13
FLAG IP
time (min)
RFU
K48-linked di-ubiquitin cleavage assay
Fig. S4. USP13 exhibits DUB activity toward K48-linked di-ubiquitins. (A) FLAG-USP13, immunopuried from HeLa cells and used in in vitro deubiquitination
assays, was quantied against BSA standards by Coomassie blue staining. Approximat ely 2030 nM FLAG-USP13 was used in each deubiquiti nation reaction
(see Materials and Methods for details). ( B) Deubiquitination of K4 8-linked di-ubiqutin substrates with internally quenched uorescence by USP13. Between 20
and 30 nM FLAG-USP13 immunopuried from HeLa cells was used per reaction with 200 nM of substrate. Buffer only and FLAG immunoprecipitates from
empty vector-transfected cells served as controls. DUB activity was monitored over 60 min by release of uorescence, measured in relative uorescence units
(RFU). (C) USP13 was knocked-down by three individual shRNAs in HeLa cells (see Fig. 4B for protein quantication of the same samples). USP13 transcripts
were quantied by SYBR-green qRT-PCR. The amount of USP13 mRNA was normalized to cyclophilin and set as 100% in control cells.
Chen et al. www.pnas.org/cgi/content/short/1100028108 5of10
Page 11
B
Ufd1
Skp2
p97
Npl4
β-actin
0h 16h .5 1 2.5 5 .5 1 2.5 5
DMSO TM (μg/ml) TG (μM)
A
HEK-293T
Counts
0 200 400 600 800 1000
FL2-A
0
20
40
60
80
100
TM
(1μg/ml)
DMSO
HeLa-S3
100
Counts
0 200 400 600 800 1000
FL2-A
0
20
40
60
80
TM
(1μg/ml)
DMSO
HFF-1
Counts
0 200 400 600 800 1000
FL2-A
0
20
40
60
80
100
TM
(1μg/ml)
DMSO
HeLa
Counts
0 200 400 600 800 1000
FL2-A
0
20
40
60
80
100
TM
(1μg/ml)
DMSO
Ufd1
Skp2
p27
β-actin
HeLa-S3
TMg/ml)
DMSO .5 1 2
HeLa
TMg/ml)
DMSO .5 1 2
HFF-1
TMg/ml)
DMSO .5 1 2
HEK-293T
TMg/ml)
DMSO .5 1 2
0 200 400 600 800 1000
0
20
40
60
80
100
Counts
0 200 400 600 800 1000
0
20
40
60
80
100
Counts
DMSO
1μg/ml TM
DMSO
1μM TG
DC
Ufd1
Skp2
HSP90
Lamin A/C
cyto
DMSO 0 24 48 0 24 48 h
nuc
Ufd1
Skp2
HSP90
Lamin A/C
TM (0.5 g/ml) 0 24 48 0 24 48 h
nuccyto
μ
relative increase in G1(%)
-15
-10
-5
0
5
10
15
20
25
30
35
40
0h 4h 8h 12h 16h 20h 24h
DMSO
HeLa, 0.5 μg/ml TM
HeLa, 2.5 μg/ml TM
Fig. S5. ER stress-dependent regulation of Ufd1-Skp2 and G1 cell cycle progression. (A) Regulation of Ufd1-Skp2 by TM in different cell types. HeLa-S3, HeLa,
HFF-1, and HEK-293T cells were treated with either DMSO (lane 1 of each panel) or TM (0.5, 1, or 2 μg/mL) for 20 h. (Upper) Immunoblots of endogenous Ufd1,
Skp2, and p27. (Lower ) FACS analysis showing TM-induced G1 delay. In each pair of superimposed histograms, the cell cycle prole of DMSO-treated cells is in
red, and that of 1 μg/mL TM-treated cells is in blue. TM-induced G1 delay (blue histograms) is better visualized as decreases in the S and G2/M populations. (B)
Thapsigargin (TG) also regulates Ufd1-Skp2 and cell cycle. HeLa cells were treated with DMSO, increasing doses of TM (0.5, 1, 2.5, and 5 μg/mL), or TG (0.5, 1,
2.5, and 5 μM) for 16 h. Total cell lysates were immunoblotted for the indicated proteins. (Right) FACS histograms comparing G1 delay in cells treated with
DMSO, 1 μg/mL TM, or 1 μM TG for 16 h. (C) G1 cell cycle delay in response to TM is dose- and time-dependent. Control cells were treated with DMSO or 0.5 μg/
mL or 2.5 μg/mL of TM over a 24-h course. Samples were collected every 4 h for immunoblot analysis (Fig. 5B) and cell cycle analysis by FACS. The graph
quanties the percent increase in G1 under each condition. (D) Nuclear Ufd1 is targeted by prolonged ER stress. HeLa cells were treated with DMSO or 0.5 μg/
mL of TM and collected at 0 h, 24 h, and 48 h for cytoplasmic (cyto) and nuclear (nuc) fractionation. Fractions were immunoblotted for Ufd1 and Skp2. HSP90
served as a cytoplasmic marker, and Lamin A/C served as a nuclear marker.
Chen et al. www.pnas.org/cgi/content/short/1100028108 6of10
Page 12
Fig. S6. APC/C
Cdh1
mediates ER stress-dependent Skp2 degradation and subsequent p27 stabilization. (A) HeLa cells transfected with empty shRNA vector or
Cdh1-specic shRNA (for 24 h) were treated with DMSO or 2.5 μg/mL of TM alone or together with MG-132 (2 or 5 μM) for 12 h. Total cell extracts were
immunoblotted for the indicated proteins. (B) TM regulates the transcription of Skp2 and p27. Transcript levels of Ufd1, Skp2, and p27 in HeLa cells treated
with either DMSO or 0.5 μg/mL of TM for 20 h were quantied by SYBR-green qRT-PCR. The amount of each transcript was normalized to cyclophilin. (C) Ufd1
overexpression partially overcomes TM-induced G1 cell cycle delay. HeLa cells overexpressing empty vector or FLAG-Ufd1 were treated with DMSO or 1 μg/mL
of TM for 8 h. Cell cycle proles of DMSO- and TM-treated cells under each transfection condition were compared by FACS analysis. Fig. 5D quanties the
difference in the percentage of cells in G1 between DMSO- and TM-treated cells.
Chen et al. www.pnas.org/cgi/content/short/1100028108 7of10
Page 13
Fig. S7. Cyclin D1 overexpression does not overcome TM-induced G1 cell cycle delay in HeLa cells. HeLa cells transfected with empty vector or cyclin D1 were
treated with DMSO or increasin g concentrations of TM (0.5, 1, 2, and 3 μg/mL) for 17 h. Cells were collected for immunoblot analysis of the indicated proteins
and cell cycle analysis by FACS.
Chen et al. www.pnas.org/cgi/content/short/1100028108 8of10
Page 14
AB
C
D
G1 G1G2-M G2-M
G1/S arrest 7h release from G1/S arrest
CHX 0h
CHX 1h
CHX 2h
CHX 3h
20
40
60
80
100
120
G2/M
G1/S
3h2h1h0h
Quantication of NHK protein amount
relative abundance (%)
normalized to HSP90
CHX (h)
40
60
80
100
G2-M
G1-S
2h1h0.5h0h
40
60
80
100
G2-M
G1-S
3h2h1h0h
NHK, relative abundance (%)
normalized to HSP90
CFTR, relative abundance (%)
normalized to
β−actin
CHX (h)
CHX (h)
cells
G1 G2-M
- S
Fig. S8. Enhanced clearance of NHK and CFTR in G1. (A) Verication of cell cycle synchronization by FACS for Fig. 5F.(B) Protein levels of GFP-CFTR and NHK
detected by immunoblot a nalysis (Fig. 5F) were quantied with Odyssey software (LiCOR), normalized to loading controls, and presented as a percentage of
the amount of corresponding proteins at time 0 h of CHX treatment (set as 100%). (C) Verication of cell cycle synchronization by FACS for Fig. 5G.(D) Protein
levels of NHK detected by immunoblot analysis (Fig. 5G) were quantied using Odyssey software (LiCOR), normalized to HSP90, and presented as a percentage
of the amount of NHK protein at time 0 h of CHX treatment (set as 100%).
Chen et al. www.pnas.org/cgi/content/short/1100028108 9of10
Page 15
AB
0
5
10
15
20
25
30
CF broblasts
normal human broblasts
0.5
µg/ml TM, 20h
% increase in G1 population
(normalized to DMSO-treated samples)
0
200
400
600
800
# Cells
0
200
400
600
# Cells
normal human broblasts
G1: 70%
S: 20%
G2-M: 10%
CF broblasts ( F508)
G1: 76%
S: 15%
G2-M: 9%
G1 S G2/M
Fig. S9. Increased G1 and hypersensitivity to TM in cells from a cystic brosisaffected individual. (A) FACS analysis of steady-state cell cycle proles of normal
human broblasts (BJT) and broblasts from a cystic brosisaffected individual homozygou s for CFTR-ΔF508 mutation (CF broblasts). (B) Graph quantifying
the percent increase in G1 in norm al human and CF broblasts treated with DMSO or 0.5 μg/mL of TM for 20 h. The values shown were normalized to the
percentage of cells in G1 in DMSO-treated samples.
Chen et al. www.pnas.org/cgi/content/short/1100028108 10 of 10
Page 16
    • "This p97-Ufd1-Npl4 complex acts in endoplasmatic reticulum (ER) associated degradation (ERAD), extracting misfolded proteins marked with ubiquitin from ER to be degraded in cytosol by the proteasome (Ye et al. 2001). Thus, it was suggested that Ufd1 is directly associated with stress response in ERAD (Chen et al. 2011). UFD1L SNP, rs5992403, has been associated with schizophrenia (De Luca et al. 2001 ), with earlyonset of the disorder (Ota et al. 2010) and with a defi cit in the set-shifting task (which represents a defi cit in fundamental dimensions of cognition in schizophrenia) (Ota et al. 2013). "
    [Show abstract] [Hide abstract] ABSTRACT: This study aimed to investigate peripheral blood gene expression in ultra-high-risk subjects (UHR) compared to first-episode psychosis individuals (FEP) and healthy controls (HC). We enrolled 22 UHR, 66 FEP and 67 HC and investigated the expression of 12 genes using Taqman assays. We used the Univariate General Linear Model, as well as Bonferroni correction for multiple comparisons. We found that UFD1L (ubiquitin fusion degradation 1 like (yeast)) gene was upregulated in UHR group compared to HC and FEP (P = 3.44 × 10(-6) ; P = 9.41 × 10(-6)). MBP (myelin basic protein) was downregulated in UHR compared to FEP (P = 6.07 × 10(-6)). DISC1 (disrupted in schizophrenia 1) was also upregulated in UHR compared to FEP but lost statistical significance when corrected for age. These genes are directly related to neurodevelopmental processes and have been associated to schizophrenia. Recent findings described that DISC1 overexpression can disrupt MBP expression, thus, we think that these alterations in UHR individuals could be associated with a common process. UFD1L showed a different pattern of expression only for UHR group, suggesting that they can be under an acute endoplasmatic reticulum stress, demanding elevated levels of Ufd1. Further studies can improve knowledge on disease progression and putative targets to preventive strategies.
    Full-text · Article · Jun 2015 · The World Journal of Biological Psychiatry
    • "USP7 or HAUSP (herpesvirus-associated ubiquitin-specific protease) deubiquitinates SCF-í µí»½- TrCP mediated K48-linked Ub chains on claspin, the upstream regulator of Chk1 [97]. USP13 counteracts S phase kinase-associated protein 2 (Skp2) ubiquitination via the anaphase promoting complex/cyclosome (APC/C Cdh1 ), delaying cell cycle by accumulation of p27 [98] . USP19 deubiquitinates Kip1 ubiquitination-promoting complex protein 1 (KPC1) regulating p27 Kip1 [99] and some KPC1 independent cell cycle regulation also exists [100]. "
    [Show abstract] [Hide abstract] ABSTRACT: The process of cell death has important physiological implications. At the organism level it is mostly involved in maintenance of tissue homeostasis. At the cellular level, the strategies of cell death may be categorized as either suicide or sabotage. The mere fact that many of these processes are programmed and that these are often deregulated in pathological conditions is seed to thought. The various players that are involved in these pathways are highly regulated. One of the modes of regulation is via post-translational modifications such as ubiquitination and deubiquitination. In this review, we have first dealt with the different modes and pathways involved in cell death and then we have focused on the regulation of several proteins in these signaling cascades by the different deubiquitinating enzymes, in the perspective of cancer. The study of deubiquitinases is currently in a rather nascent stage with limited knowledge both in vitro and in vivo , but the emerging roles of the deubiquitinases in various processes and their specificity have implicated them as potential targets from the therapeutic point of view. This review throws light on another aspect of cancer therapeutics by targeting the deubiquitinating enzymes.
    Full-text · Article · Jul 2014 · BioMed Research International
    • "Posttranslational N-glycosylation plays a crucial role in the ERQC system to strictly preserve proteostasis [31]. The inhibition of N-glycosylation activates ER-specific stress responses in neurons , which include the ER-associated degradation (ERAD) mechanism responsible for differential and extremely efficient degradation of nonglycosylated protein by the proteasome after ubiquitination323334. Additionally, we do not observe the co-localization between Sho and PrP Sc both in the brains Author's personal copy of scrapie-infected animals and in the scrapie-infected cell line, which may indicate that it is not possible that downregulation of endogenous Sho is due to the direct interaction with exogenous prion agent. "
    [Show abstract] [Hide abstract] ABSTRACT: Shadoo (Sho) is an N-glycosylated glycophosphatidylinositol-anchored protein that is expressed in the brain and exhibits neuroprotective properties. Recently, research has shown that a reduction of Sho levels may reflect the presence of PrP(Sc) in the brain. However, the possible mechanism by which prion infection triggers down-regulation of Sho remains unclear. In the present study, Western blot and immunohistochemical assays revealed that Sho, especially glycosylated Sho, declined markedly in the brains of five scrapie agent-infected hamsters and mice at the terminal stages. Analyses of the down-regulation of Sho levels with the emergence of PrP(Sc) C2 proteolytic fragments did not identify close association in all tested scrapie-infected models. To further investigate the mechanism of depletion of Sho in prion disease, a Sho-expressing plasmid with HA tag was introduced into a scrapie-infected cell line, SMB-S15, and its normal cell line, SMB-PS. Western blot assay revealed dramatically decreased Sho in SMB-S15 cells, especially its glycosylated form. Proteasome inhibitor MG132 reversed the decrease of nonglycosylated Sho, but had little effect on glycosylated Sho. N-acetylglucosamine transferase inhibitor tunicamycin efficiently reduced the glycosylations of Sho and PrP(C) in SMB-PS cells, while two other endoplasmic reticulum stress inducers showed clear inhibition of diglycosylated PrP(C), but did not change the expression level and profile of Sho. Furthermore, immunoprecipitation of HA-Sho illustrated ubiquitination of Sho in SMB-S15 cells, but not in SMB-PS cells. We propose that the depletions of Sho in scrapie-infected cell lines due to inhibition of glycosylation mediate protein destabilization and subsequently proteasome degradation after modification by ubiquitination.
    Full-text · Article · Jan 2014 · Molecular Neurobiology
    Jin Zhang Jin Zhang Yan Guo Yan Guo Wu-Ling Xie Wu-Ling Xie +7 more authors... Yin Xu Yin Xu
Show more

Similar publications

Discover more