ORIGINAL ARTICLE
C-terminal phosphorylation of Hsp70 and Hsp90 regulates
alternate binding to co-chaperones CHIP and HOP to
determine cellular protein folding/degradation balances
P Muller1
, E Ruckova1
, P Halada2
, PJ Coates3
, R Hrstka1
, DP Lane4
and B Vojtesek1
Heat shock proteins (Hsp)90 and Hsp70 facilitate protein folding but can also direct proteins for ubiquitin-mediated degradation.
The mechanisms regulating these opposite activities involve Hsp binding to co-chaperones including CHIP and HOP at their
C-termini. We demonstrated that the extreme C-termini of Hsp70 and Hsp90 contain phosphorylation sites targeted by kinases
including CK1, CK2 and GSK3-b in vitro. The phosphorylation of Hsp90 and Hsp70 prevents binding to CHIP and thus enhances
binding to HOP. Highly proliferative cells contain phosphorylated chaperones in complex with HOP and phospho-mimetic and nonphosphorylable
Hsp mutant proteins show that phosphorylation is directly associated with increased proliferation rate. We also
demonstrate that primary human cancers contain high levels of phosphorylated chaperones and show increased levels of HOP
protein and mRNA. These data identify C-terminal phosphorylation of Hsp70 and Hsp90 as a switch for regulating co-chaperone
binding and indicate that cancer cells possess an elevated protein folding environment by the concerted action of co-chaperone
expression and chaperone modifications. In addition to identifying the pathway responsible for regulating chaperone-mediated
protein folding/degradation balances in normal cells, the data provide novel mechanisms to account for the aberrant chaperone
activities observed in human cancer cells and have implications for the application of anti-chaperone therapies in cancer treatment.
Oncogene (2012) 0, 000–000. doi:10.1038/onc.2012.314
Keywords: Hsp70; Hsp90; CHIP; HOP; chaperone; folding
INTRODUCTION
Chaperones are required for the folding and conformational
regulation of proteins. Two of the major cytosolic chaperones,
heat shock protein (Hsp)70 and Hsp90, are particularly responsible
for the production of signal transduction proteins and cell cycle
regulators.1–3
However, their roles extend beyond the correct
folding of client proteins to regulate degradation of unfolded
clients.3,4
These degradation and folding properties of chaperones
are essential to maintain native proteins and prevent aggregation
of denatured proteins.5
To achieve folding or degradation of client
proteins, Hsp90 and Hsp70 cooperate with co-chaperones and
Hsp90 is known to interact with at least 20 co-chaperones that are
variously involved in coordinating the interaction of Hsp90 with
Hsp70, in regulating the ATPase activity of Hsp90, and in the
recruitment of specific client proteins into the chaperone system.6
Although co-chaperones may bind to the N-terminal or middle
regions of HSPs, the largest group of co-chaperones bind to the
highly conserved sequence EEVD-COOH in the C-terminal
domains of Hsp90 and Hsp70.7–9
These co-chaperones contain a
so-called tetratricopeptide repeat (TPR) domain that interacts with
this C-terminal sequence. Within the group of TPR-containing cochaperones
are proteins that can assist with protein folding
activity, such as HOP, or that have ubiquitin ligase activity, such as
CHIP. Thus, the balance between the binding of Hsp’s to these cochaperones
is a potential mechanism that may contribute to the
regulation of whether proteins are folded and stabilized or are
targeted for degradation.10,11
Such a process must be highly
regulated to ensure correct cell functionality, but how the
dynamic interactions between chaperones and TPR-containing
co-chaperones are regulated are unclear.
Here, we provide evidence that chaperone binding to cochaperones
is a regulated process and searched for posttranslational
modifications in the C-terminus of Hsp70 and Hsp90 that
regulate binding to the CHIP or HOP TPR domains. Our results
show that phosphorylation of the C-terminal region of both
chaperones is critical for the decision between folding or
degradation of client proteins by altering the binding of CHIP
and HOP. We also show increased phosphorylation in cancer cells
together with increased levels of the co-chaperone HOP. The data
define a physiological pathway for the regulation of protein
folding of importance for normal cellular growth and for cancer.
RESULTS
Prediction of phosphorylation of C-terminus of Hsp70 and Hsp90
As it is known that the C-terminal regions of Hsp70 and Hsp90 are
critical for binding to the TPR domain of co-chaperones, we
scanned this region and identified a highly homologous region at
the extreme C-terminus of both Hsp70 and Hsp90 that contains
serine and threonine residues that are potential phosphorylation
sites for CK212
(Figures 1a and b). The evolutionarily younger
Hsp90a protein contains a second potential phosphorylation site
that is not present in vertebrate Hsp90b or in Hsp90 of nonvertebrate
species. We also used NetPhosK 1.0 and NetPhos 2.0
1
Masaryk Memorial Cancer Institute, Brno, Czech Republic; 2
Institute of Microbiology, Academy of Sciences of the Czech Republic, Prague, Czech Republic; 3
Centre for Oncology
and Molecular Medicine, University of Dundee, Scotland, UK and 4
p53 Laboratory (A*STAR), 8A Biomedical Grove, Immunos, Singapore. Correspondence: Dr B Vojtesek, Masaryk
Memorial Cancer Institute, Zluty kopec 7, 65653 Brno, Czech Republic.
E-mail: vojtesek@mou.cz
Received 13 January 2012; revised 30 May 2012; accepted 13 June 2012
Oncogene (2012), 1–10
& 2012 Macmillan Publishers Limited All rights reserved 0950-9232/12
www.nature.com/onc
Server, which uses artificial neural network methods,13
to provide
additional evidence for potential phosphorylation of these sites in
the C-terminus of Hsp70 and Hsp90. Figure 1c shows that the
probability of phosphorylation of C-terminal residues of Hsp70
and Hsp90 is between 60 and 99% and it is predicted that Hsp90a
will be phosphorylated at both sites. NetPhos 1.0 also predicted
that CK1 and GSK3-b are capable of phosphorylating the C-termini
of Hsp70 and Hsp90.
CK1, CK2 and GSK3-b phosphorylate Hsp70 and Hsp90 in vitro
To investigate whether CK2 can phosphorylate the C-terminus of
Hsp70 and/or Hsp90, we performed in vitro phosphorylation
assays using recombinant Hsp70 and Hsp90a and CK2. We used
mass spectrometry coupled with sequential proteolytic digestion
(Supplementary Data S2) to detect C-terminal phosphorylation.
The MS/MS analysis of recombinant Hsp90a phosphorylated by
CK2 in vitro confirmed that both serines and threonines in close
proximity to the EEVD sequence are phosphorylated. Using mass
spectrometry we could not detect the C-terminal peptide of
Hsp70. Therefore, we developed monoclonal antibodies against
phosphorylated peptides specific for either Hsp70 or Hsp90. These
monoclonal antibodies were rigorously tested for specificity to
recognize C-terminal phosphorylation of Hsp70 or of Hsp90a and
were shown to be highly selective for the phosphorylated forms
(see Supplementary Data S1). The purified monoclonal antibodies
were also used to demonstrate in vitro phosphorylation of
C-terminal Hsp70 and Hsp90a by CK1, CK2, GSK3b, but not by
p38 mitogen-activated protein kinase (MAPK) or cyclin-dependent
kinase (CDK)1/cyclinB (Figure 1d).
Phosphorylation of Hsp70 and Hsp90 alters their binding to
co-chaperones
To test the influence of C-terminal phosphorylation of Hsp70 and
Hsp90 on their ability to bind HOP and CHIP, biotin-labeled Hsp70
and Hsp90 peptides and phosphopeptides were adsorbed to
streptavidin-coated plates and recombinant glutathione S-transferase
(GST)-tagged HOP and CHIP proteins were added. Figure 2a
shows that the non-phosphorylated C-terminal peptides of both
Hsp70 and Hsp90 interact more strongly with CHIP than the
phosphorylated forms of these peptides. In contrast, HOP binds
more strongly to the phosphorylated peptide of Hsp90a although
the interaction between Hsp70 and HOP was not significantly
influenced by phosphorylation (Figure 2b). As an independent
method to assess the strength of binding of CHIP and HOP to the
C-terminal phosphorylated forms of Hsp70 and Hsp90, we
used fluorescence polarization assay to confirm the above results
Figure 1. Phosphorylation of Hsp70 and Hsp90 C-terminus. (a) Alignment of eukaryotic Hsp90 sequences. Potentially phosphorylated serines
or threonines at seventh and eighth position from C-terminus of Hsp90 are lined with red. (b) Alignment of eukaryotic Hsp90 sequences.
Eukaryotic Hsp70, except green plants, contains serine or threonine at sixth position from its C-terminus. (c) Prediction of phosphorylation of
C-terminus of Hsp70 and Hsp90. The graph shows the probability of phosphorylation of C-terminal residues of Hsp70 and Hsp90 based on
neural network prediction. The index 0–1 of listed kinases reflects the probability that the kinase phosphorylates the serine or threonine.
(d) In vitro phosphorylation of recombinant Hsp70 and Hsp90a protein. The phosphorylation of Hsp70 and Hsp90a was detected using
phospho-specific monoclonal antibodies GGS2.1 and GDD8.2, respectively.
Balance between folding and degradation
P Muller et al
2
Oncogene (2012), 1 – 10 & 2012 Macmillan Publishers Limited
and to quantify the dissociation constant between Hsp’s
C-terminal peptides and HOP and CHIP proteins (Supplementary
Data S3). The data from fluorescence polarization are consistent
with data obtained by enzyme-linked immunosorbent assay.
The calculated dissociation constants indicate that phosphorylated
Hsp70 has an B20-fold lower affinity for CHIP than nonphospho-Hsp70,
whereas the affinity of HOP for Hsp70 is not
increased by phosphorylation of Hsp70. For Hsp90, the phosphorylated
C-terminal peptide shows both a higher affinity for
HOP and a lower affinity for CHIP (Figure 2c). To test whether
these observations are also seen in vivo, A375 cell lysates were
incubated with magnetic beads coated with peptides corresponding
to the C-terminus of Hsp70 and Hsp90 and the amounts of
bound CHIP and HOP measured by western blotting. As with the
in vitro data, CHIP preferentially binds to the non-phosphorylated
C-terminal peptides of both Hsp70 and Hsp90, whereas HOP is
preferentially bound to the phosphorylated C-terminal peptide of
Hsp90 and phosphorylated and non-phosphorylated Hsp70
peptides bind equivalent levels of HOP (Figure 2d). These results
indicate that the abilities of CHIP and HOP to bind Hsp70 and
Hsp90 are influenced by C-terminal phosphorylation of the
chaperones.
Structural comparison of the HOP and CHIP TPR domains
To address the question of how the interactions between CHIP
and the C-termini of Hsp70 and Hsp90 are disrupted by C-terminal
phosphorylation, we compared the crystal structures of the HOP
TPR (PDB: 1ELW and 1ELR) and CHIP TPR (PDB: 2C2L) domains.14,15
Despite limited amino-acid sequence similarity, the alignment of
CHIP TPR with HOP TPR1 and TPR2A shows high structural
similarity, especially in the N-terminal parts of the domain that are
responsible for binding the most C-terminal part of EEVD-COOH
peptides.16
Figure 3a demonstrates the conserved basic residues
that are responsible for anchorage of the C-terminal carboxyl
groups of the EEVD domain. The first five helices of the TPR
domains contain conservative basic residues responsible for
interaction with C-termini of the EEVD domains. However, the
structure of the TPR domain of CHIP differs from HOP in the
C-terminal part. The most prominent difference is loop 6, which
connects helix 6 and 7 of the TPR domain of CHIP and also
interacts with serine and threonine of the Hsp90 C-terminus. This
loop is longer than the corresponding loop of HOP TPR1 or TPR2A
and protrudes into the groove of the TPR domain in CHIP
(Figure 3b). Most importantly, the asparagine 131 of loop 6 in the
CHIP TPR is able to form hydrogen bonds with the C-terminal
serine and threonine of Hsp90a. Phosphorylation of the C-terminal
serine and threonine of Hsp90 or Hsp70 will prevent the formation
of hydrogen bonds with loop 6 of the CHIP TPR domain and
disrupt protein binding.
With regard to the regulation of binding to HOP, phosphorylated
groups in Hsp90 can contact one or more of the basic
groups in the groove of the elongated HOP TPR. Among the basic
groups protruding into the space in close proximity to the
phosphorylated serine and threonine of Hsp90 are lysines 237, 238
and 239 of the human HOP TPR2A domain. Thus, phosphorylation
of Hsp90 is predicted to enhance binding to these lysines in the
HOP TPR2A domain (Figure 3b). Despite an overall similarity of
structure in the TPR domains of HOP and CHIP, these three lysines
in loop 1 of HOP TPR2A are not present in the corresponding parts
of CHIP TPR and HOP TPR1, so that phosphorylation of Hsp90 or
Hsp70 is predicted not to increase the interaction with CHIP or
with the TPR1 domain of HOP, to which Hsp70 binds. Thus,
phosphorylation is predicted to enhance Hsp90 binding to HOP
but not to directly influence binding of Hsp70, in keeping with the
findings above.
Figure 2. The impact of C-terminal phosphorylation of Hsp70 and Hsp90 on binding of CHIP and HOP proteins. Enzyme-linked
immunosorbent assay analysis of complex formation between C-terminal peptides of Hsp70 and Hsp90 with CHIP and HOP proteins.
Streptavidin-coated plates were incubated with biotinylated peptides corresponding to phosphorylated and non-phosphorylated C-terminal
region of Hsp70 and Hsp90. Purified recombinant protein GST-CHIP (a) and GST-HOP (b) were tested for interaction with these peptides.
(c) Equilibrium dissociation constants between proteins HOP and CHIP and phosphorylated or non-phosphorylated fluorescent peptides were
measured by fluorescence polarization binding assay. (d) Peptide pull-down assay of CHIP and HOP protein from cell lysate. Biotinylated
phospho- and non-phosphorylated C-terminal peptides of Hsp70 and Hsp90 were incubated with A375 cell lysates. The figure shows the level
of HOP and CHIP in cell lysate and their interaction with C-terminal peptides of Hsp70 and Hsp90 origin. HOP and CHIP proteins were detected
by monoclonal antibodies HOP1.1 and CHIP3.1.
Balance between folding and degradation
P Muller et al
3
& 2012 Macmillan Publishers Limited Oncogene (2012), 1 – 10
Hsp C-terminal phosphorylations are higher in proliferating cells
These data demonstrate that phosphorylation of the C-termini of
Hsp70 and Hsp90 alters their binding to co-chaperones with
opposite functions in protein folding and stability. Proliferating
cells are known to require high levels of protein synthesis to allow
increased cell growth, reflected by an increase in ribosomal mass
in proliferating cells compared with non-proliferating cells.17
To
assess the physiological relevance of our observations we
therefore initially tested the association of Hsp C-terminal
phosphorylation with proliferative state in A375 melanoma cells
cultured in medium containing 10% fetal bovine serum or without
serum. We performed dephosphorylation of each sample to
demonstrate the specificity of the phospho-Hsp antibodies.
Figure 4a shows that the level of C-terminal phosphorylation of
both Hsp70 and Hsp90 is markedly increased in proliferating cells
compared with serum-starved cells. The specificity of phosphospecific
antibodies was also confirmed by the absence of signal
following alkaline phosphatase treatment. These data indicate
that C-terminal phosphorylation of Hsp70 and Hsp90 is altered
according to cellular environment. Co-immunoprecipitation of
Hsp70 or Hsp90 demonstrated that Hsp90 preferentially binds
HOP in proliferating cells, in contrast to serum-starved cells in
which both Hsp’s preferentially complex with CHIP (Figures 4b and
c). We also used proximity ligation assay as another independent
method to confirm that Hsp90 and Hsp70 in proliferating cells
preferentially bind HOP in contrast to CHIP (Figure 4d and
Supplementary Data S4). These data indicate that Hsp binding to
these two co-chaperones is a regulated process in cells.
C-terminal phosphorylation increases proliferation rate
To further investigate the physiological relevance of Hsp
C-terminal phosphorylation in regulating cellular activities and
the relationship with cell proliferation, we determined whether
phosphorylation directly influences cell proliferation. We designed
mutations of Hsp70, Hsp90a and Hsp90b to mimic the phosphorylated
and non-phosphorylated status of their C-termini.
Hsp70, Hsp90 and their phospho-mimetic variants were produced
as N-terminal streptavidin-binding protein-tagged proteins in
cells. Figure 5a shows that both co-chaperones CHIP and HOP
specifically bind to Hsp70 and Hsp90 in contrast to GFPQ1 , which
served as a negative control. The ratios between HOP and
CHIP indicate their affinity for wild-type, non-phospho- or
phospho-mimetic forms of Hsp70, Hsp90a and Hsp90b. Phosphorylated
Hsp90 has increased affinity to HOP compared with
non-phosphorylated forms of Hsp90 proteins, whereas CHIP has
higher affinity for non-phosphorylated than for the negatively
charged phosphorylated C-terminus of both Hsp70 and Hsp90. To
investigate the effects of C-terminal Hsp90 phosphorylation on
proliferation, we transfected cells with non-tagged Hsp70, Hsp90a
and Hsp90b and their phospho-mimetic mutants. Expression of
the non-phosphorylable Hsp70 and Hsp90 mutants significantly
prolonged the doubling time compared with cells transfected
with wild-type protein (Po0.001 for each), whereas expression of
phospho-mimetic mutants of Hsp90 reduced the doubling time
(Po0.01 for Hsp90a and Hsp90b phospho mimetics; Figure 5b).
Doubling times and statistical testing were computed from
exponential growth curves described in Supplementary Data S5.
The regulation of C-terminal phosphorylation of Hsp70 and Hsp90
in vivo involves numerous kinases and phosphatase activity
We had identified that CK1, CK2 and GSK3b are able to directly
phosphorylate the C-terminus of Hsp70 and Hsp90 in vitro. To
investigate whether other kinases are responsible for C-terminal
phosphorylation in vivo, we screened a panel of kinase inhibitors
applied to proliferating cells and measured C-terminal phosphorylation
of Hsp90 and Hsp70 by western blotting with phosphospecific
monoclonal antibodies (Figure 6a). Inhibition of phosphatidylinositol
3-kinase, mitogen-activated protein kinase)/extracellular
signal-regulated kinase or AKT but not tyrosine kinase or
CDKs reduced C-terminal Hsp90 phosphorylation, while Hsp70
phosphorylation was reduced in the presence of phosphatidylinositol
3-kinase inhibitors only. Treatment with a specific inhibitor
of p38MAPK, SB202190, was a very effective inhibitor of Hsp90a
C-terminal phosphorylation and of Hsp70 to a lesser extent
(Figure 6b). These data identify a complexity of the regulation of
Hsp70 and Hsp90 C-terminal phosphorylation that is modified by
a variety of kinases, and a prominent role for p38MAPK in
selectively regulating the C-terminal phosphorylation of Hsp90,
despite the inability of p38MAPK to directly phosphorylate these
residues (Figure 1d).
To investigate whether the in vivo phosphorylation status of
Hsp70 and Hsp90 is also subject to regulation by protein
phosphatases we treated cells with serine/threonine phosphatase
inhibitors Calyculin A (inhibits PP1 and PP2A) and Okadaic acid
Figure 3. Structural alignment of the CHIP TPR, HOP TPR1 and HOP TPR2A domains. (a) The alignment of TPR domains shows the conservative
residues responsible for binding of EEVD of Hsp70 or Hsp90. The loop 6 of protein CHIP TPR domain is highlighted to show the hydrogen
bonds between asparagine and C-terminal serine and threonine of Hsp90a. (b) Alignment of PDB structures 2c2L (complex between CHIP TPR
and Hsp90a peptide DTSRMEEVD), 1ELR (complex between HOP TPR2A and Hsp90 peptide AceMEEVD) and 1ELW (complex between HOP
TPR1 and Hsp70 peptide GPTIEEVD). The lysines K238-239 of HOP TPR2A are highlighted.
Balance between folding and degradation
P Muller et al
4
Oncogene (2012), 1 – 10 & 2012 Macmillan Publishers Limited
(selective for PP2A). Figure 6c shows that treatment with Calyculin
A in contrast to Okadaic acid increased the steady state levels of
C-terminal phosphorylation of both Hsp70 and Hsp90, indicating a
selective role for PP1 in regulating Hsp C-terminal phosphorylation
in vivo.
Inhibiting phosphorylation synergizes with Hsp90 inhibition to
reduce client protein stability
Our data indicate that C-terminal phosphorylation prevents the
interaction of Hsp’s with the ubiquitin ligase CHIP and should
therefore prolong the half-life of Hsp90 clients by inhibiting their
degradation. In particular, C-terminal phosphorylation should
restrict degradation after inhibition of Hsp90 by 17AAG. To
investigate this possibility, we initially analyzed the stability of
Hsp90 clients in T47D breast cancer cells that express estrogen
receptor, CDK4 and mutant p53, all well-studied Hsp90 client
proteins. Figure 7a shows that inhibiting Hsp90 with 17-AAG
reduces the levels of these three client proteins, confirming the
involvement of Hsp90 in regulating their production and stability.
Inhibition of p38 MAPK by SB202190 reduces Hsp90 phosphorylation
and reduces the levels of client proteins. Most notably,
combined treatment causes a further reduction in protein levels,
indicating synergistic effects of inhibiting Hsp90 phosphorylation
and inhibiting Hsp90 activity.
However, these experiments only show an association between
phosphorylation and client protein levels and do not show a direct
role for C-terminal phosphorylation. Therefore, we performed
experiments with point mutants of Hsp70 and Hsp90a to directly
investigate the effect of C-terminal phosphorylation on client
protein stability. For these experiments, we used the p53 mutant
R175H as a model client protein that relies on Hsp90 for stability.18
The p53-null cell line H1299 was co-transfected with p53 mutant
R175H, Hsp70, Hsp90a and their phospho-mimetic or nonphosphorylable
mutants. Figure 7b shows that the level of p53
mutant is decreased when coexpressed with non-phosphorylable
alanine mutants of Hsp70 and Hsp90a but is stabilized in the
presence of phospho-mimetic mutants.
Primary cancers show altered C-terminal phosphorylated
chaperones and increased levels of HOP
The previous experiments had shown that C-terminal phosphorylation
of Hsp70 and Hsp90 induces preferential binding of HOP
and directly influences growth of tumor cells. To detect
differences in phosphorylation of Hsp’s in primary tumors, we
used surgically removed human breast tissues from which we took
a sample of healthy and cancerous tissue. Western blotting shows
that both cancer and normal cells express comparable levels of
Hsp70 and Hsp90a. However, the C-terminal phosphorylation of
Hsp’s is significantly increased in cancer cells compared with
normal breast tissue (Figure 8a). Previous work had suggested that
co-chaperone balances may be disrupted in cancer cells,19
prompting us to investigate whether the levels of CHIP or HOP
are different between normal and tumor tissues. Unlike previous
data,19
we did not find decreased levels of CHIP protein in cancer
compared with normal tissues but levels of HOP were
considerably increased in cancer cells. We also interrogated the
expression profiling of cancer tissues contained with the
Oncomine database20
for altered expression of CHIP and HOP
Figure 4. Phosphorylation of Hsp70 and Hsp90 are regulated by proliferation. (a) The lysates of proliferating cells or cells grown without serum
were treated by alkaline phosphatase. The C-terminal phosphorylations of Hsp70 and Hsp90a were detected by phospho-specific monoclonal
antibodies GGS2.1 and GDD8.2. (b) Differential binding of chaperones to co-chaperones in proliferating and serum-starved cells. Coimmunoprecipitation
of complexes between Hsp’s and CHIP and HOP using rabbit polyclonal antibodies recognizing Hsp70 or Hsp90. (c) The
intensity of CHIP and HOP was quantified by densitometry. (d) The individual complexes of Hsp70-CHIP, HSP70-HOP, Hsp90-HOP and Hsp90CHIP
were analyzed using proximity ligation assay in situ and rolling circle amplification. The graph shows the mean number of signals per cell.
Balance between folding and degradation
P Muller et al
5
& 2012 Macmillan Publishers Limited Oncogene (2012), 1 – 10
mRNAs. These data confirmed that HOP (STIP1) mRNA is
overexpressed in many human cancers compared with
corresponding normal tissues (melanoma, bladder, gastric, lung,
breast and pancreas), whereas the levels of CHIP (STUB1) mRNA
are not commonly altered (Supplementary Material S6). We also
showed that the levels of HOP mRNA are higher in proliferating
cells compared with serum-starved cells (Po0.05) but there was
no difference in CHIP mRNA levels (Figure 8b).
DISCUSSION
Chaperones act to promote the folding and activation of
polypeptide substrates, but are also involved in targeting proteins
for degradation.3,4
Given their crucial importance for cellular
function, chaperone activities must be tightly regulated according
to cellular requirements for proliferation and survival. Previous
data have shown that Hsp90 is phosphorylated in the N-terminal
and linker regions by protein kinases including CK2, Akt, Wee1 and
DNA-dependent protein kinase.21–23
CK2 also phosphorylates the
co-chaperone Cdc37 to enhance the interaction with Hsp90, leading
to increased activity of numerous kinases that are specifically
chaperoned by Cdc37 and have pro-proliferative actions.21
However, numerous co-chaperones with different activities bind
to the C-termini of Hsp70 and Hsp90 through the TPR domains of
the co-chaperones. Two such C-terminal binding co-chaperones,
CHIP and HOP, represent a paradigm for the overall protein
folding or protein degradation activities of Hsp70/Hsp90.24
HOP is
important in coordinating protein folding and binds Hsp70 and
Hsp90 together only when Hsp90 is in an open, client-free
state.25,26
In contrast, the co-chaperone CHIP targets client
proteins for degradation. Thus, we hypothesized that
interactions between the Hsp’s and TPR-domain co-chaperones
represent an important regulator of chaperone activities.
Initially, we predicted that evolutionarily conserved phosphorylation
sites in the C-termini of Hsp70 and Hsp90 might influence
differential binding activities and showed that the interactions of
Hsp70 and Hsp90 with CHIP and HOP are altered by different
growth conditions (Figure 9). We have shown that a variety of
kinases including CK1, CK1 and GSK3-b are able to directly
phosphorylate the C-termini of Hsp70 and Hsp90 in vitro and that
p38MAPK also regulates the levels of C-terminal phosphorylated
Hsps in vivo but does not directly phosphorylate these residues
in vitro. Together with the observations that phosphatases are
involved in regulating phosphorylation levels in vivo, these
findings indicate a complex regulation of C-terminal Hsp70 and
Hsp90 phosphorylation that involves both direct and indirect
regulatory factors, as would be expected for a fundamental
physiological process involved in determining the overall protein
production/degradation balance in cells. In view of this complexity,
it will be a challenging but important task to identify the roles
of these and potentially other kinases and phosphatases in Hsp
C-terminal phosphorylation in different physiological and pathological
states in vivo. However, to confirm the physiological
relevance of Hsp C-terminal phosphorylation, we demonstrated
that proliferating cells contain higher levels of C-terminally
phosphorylated Hsp70 and Hsp90. Moreover, the introduction of phospho-mimetic
or non-phosphorylatable Hsp70 and Hsp90 proteins
demonstrated that C-terminal phosphorylation directly affects
proliferation. As these forms are preferentially bound to HOP
rather than CHIP, the data indicate that dividing cells exhibit a profolding
environment, in keeping with the requirement for
increased protein production in proliferating cells.17
Finally, we
demonstrated that Hsp70 and Hsp90 exist predominantly as
C-terminal phosphorylated forms in primary tumor tissues
compared with non-neoplastic tissue, suggesting that tumors
exhibit a dominant folding environment. The importance of the
Figure 5. Point mutations of C-terminal serine and threonine of Hsp70, Hsp90a and Hsp90b influence co-chaperone binding and cell
proliferation. (a) Pull-down assay of complexes between Hsp90 and Hsp70 phospho-mimetic mutants and co-chaperones HOP and CHIP.
Streptavidin-binding protein (SBP)-tagged GFP, Hsp70 wild-type, non-phosphorylable Hsp70 mutant T636A (Hsp70A), phospho-mimetic
Hsp70 mutant T636D (Hsp70D), Hsp90a wild-type, non-phosphorylable Hsp90a mutant T725A, S726A (Hsp90a AA), phospho-mimetic Hsp90a
mutant T725D, S726D (Hsp90a DD), Hsp90b wild-type, non-phosphorylable Hsp90b mutant S718A (Hsp90bA) and phospho-mimetic Hsp90b
mutant S718D (Hsp90bD) were transfected into HEK-293 cells to perform pull-down assays on streptavidin sepharose. The interaction of these
proteins with HOP and CHIP was assessed by monoclonal antibodies against CHIP and HOP using western blotting. (b) The effect of point
mutations of Hsp70 and Hsp90 was assessed in transfected HEK293 cells using the xCELLigence SystemQ17 monitoring cell proliferation in real
time. The graph shows the doubling time extrapolated from growth curve. The P-value shows the statistical significance of difference between
point mutants and appropriate wild-type.
Balance between folding and degradation
P Muller et al
6
Oncogene (2012), 1 – 10 & 2012 Macmillan Publishers Limited
altered interactions between Hsp70, Hsp90, CHIP and HOP in
cancers is supported by evidence for additional mechanisms that
disrupt this pathway. Thus, HOP mRNA and protein are higher in
primary human cancers compared with matched normal tissues,
reinforcing the pro-folding environment. In addition, CHIP has
been suggested to function as a tumor suppressor and can inhibit
metastatic spread.19
Our data indicated that the levels of CHIP
protein were lower in proliferating cells than in non-proliferating
cells but mRNA levels were similar, suggesting posttranslational
regulation. However, we could not confirm from our own data or
from publicly available gene expression profiling databases that
CHIP protein or mRNA levels are commonly lower in human
cancers. Taken together, the data indicate that the upregulation of
Hsp90 chaperone activity that is observed in cancer cells is related
both to increased levels of HOP and the increased proportions of
C-terminally phosphorylated Hsp70 and Hsp90, together with a
reduced amount of Hsp90-bound CHIP. In some tumors, there
may also be a reduction in CHIP levels to allow for a more folding
environment.19
In conclusion, it is known that Hsp90a is selectively activated in
human cancer and that cancer cells become addicted to increased
chaperoning activity, such that cancers can be effectively and
selectively treated with agents that inhibit Hsp90 or Hsp70.27–29
We have identified that a key mechanism for the dynamic
regulation of chaperone activity is C-terminal phosphorylation,
which regulates binding to co-chaperones that either fold or
degrade client proteins. These co-chaperones are themselves
regulated, such that replicating tumor cells possess a dominant
pro-folding environment and non-proliferating cells exhibit a
dominant protein degradation phenotype (summarized in
Table 1). In addition to influencing co-chaperone binding, it is
possible that C-terminal phosphorylation of Hsps might influence
other properties such as selective binding of clients and/or ATPase
activity, either directly or indirectly. Finally, the data provide a
framework both for understanding normal protein homeostasis
and the disruptions that occur to chaperone activity in cancer, and
allow for the identification of novel chemotherapies that can
restore the normal balance of protein folding/degradation in
cancer.
MATERIALS AND METHODS
Cell culture and treatment
Human cancer cell lines A375, H1299, T47D and HEK-293 were cultured in
Dulbecco’s modified Eagle’s medium, 10% fetal bovine serum and 300 mg/l
L-glutamine and penicillin/streptomycin. Cells were grown to 80%
confluence before experimental treatments. In specified experiments
serum was removed for 72 h to block the cell cycle at G0 phase.
Antibodies
The antibodies used in the study are described in Supplementary Data S1.
Plasmid constructs
CHIP and HOP cDNAs were cloned into pDEST-15 that is designed to
express the proteins containing an N-terminal fusion with GST in
Escherichia coli. Hsp70 and Hsp90a cDNAs were cloned into pDEST17,
which expresses the proteins with N-terminal 6 Â His tag in E. coli. All
Figure 6. Inhibition of phosphatases and p38MAPK influence the phosphorylation of Hsp70 and Hsp90. (a) Detection of C-terminal
phosphorylation in A375 cells treated with different protein kinase inhibitors for 8 h. (b) Detection of phosphorylation of Hsp70 and Hsp90a in
A375 cells treated with different concentrations of SB202190. (c) Detection of C-terminal phosphorylation of Hsp70 and Hsp90a in A375 cells
treated with different concentration of phosphatase inhibitors Calyculin A and Okadaic acid for 1 h.
Balance between folding and degradation
P Muller et al
7
& 2012 Macmillan Publishers Limited Oncogene (2012), 1 – 10
vectors designed for expression in E. coli contained TEV protease cleavage
site ENLYFQG between N-terminal tag and first methionine.
The vectors designed for pull-down assay contained CMV promoter and
streptavidin-binding protein tag MDEKTTGWRGGHVVEGLAGELEQLRARLE
HHPQGQREPS30
fused to N-terminus of EmGFP, Hsp’s and their point
mutants.
The phospho-mimetic and non-phosphorylable point mutants of Hsp70,
Hsp90a and Hsp90b were constructed by changing their C-terminal serines
or threonines to alanine (Hsp70A, Hsp90aAA and Hsp90bA) or to aspartic
acid (Hsp70D, Hsp90aDD and Hsp90bD) using the QuikChange SiteDirected
Mutagenesis Kit (Agilent Q2). We used primer pair 50
-ctgggtcaggcccc
gacattgaggaggtag-30
and 50
-ctacctcctcaatgtcggggcctgacccag-30
to make
phospho-mimetic mutant of Hsp70 (T636D), primer pair 50
-ctgggtcaggcccc
gccattgaggaggtag-30
and 50
-ctacctcctcaatggcggggcctgacccag-30
to make
non-phosphorylable mutant of Hsp70 (T636A), primer pair 50
-gaaggagatga
cgacgatgatcgcatggaagaagta-30
and 50
-tacttcttccatgcgatcatcgtcgtcatctcct
tc-30
to make phospho-mimetic mutant of Hsp90a (T725D, S726D), primer
pair 50
-gaaggagatgacgacgcagcacgcatggaagaagta-30
and 50
-tacttcttc-
catgcgtgctgcgtcgtcatctccttc-30
to make non-phosphorylable mutant of
Hsp90a (T725A, S726A), primer pair 50
-gcgatgaggatgcggaccgcatggaagaag-
30
and 50
-cttcttccatgcggtccgcatcctcatcgc-30
to make phospho-mimetic
mutant of Hsp90b (S718D) and primer pair 50
-gcgatgaggatgcggctcgcatg-
gaagaag-30
and 50
-cttcttccatgcgagccgcatcctcatcgc-30
to make non-phosphorylable
mutant of Hsp90b (S718A).
Enzyme-linked immunosorbent assay
Biotinylated peptides biotin-GGSGSPTIEEVD, biotin-GGSGSP(pT)IEEVD, biotin-GDDTSRMEEVD
and biotin-GDDD(pT)(pS)RMEEVD correspond to the
non-phosphorylated and phosphorylated C-termini of human Hsp70
Figure 7. The effect of C-terminal phosphorylation on stability of
Hsp90 clients. (a) The effect of inhibition of Hsp90 by 17AAG and
inhibition of phosphorylation by SB202190 on the degradation of
CDK4, mutant p53 and estrogen receptor in T47D cells. (b) The
effect of point mutations mimicking phosphorylated or nonphosphorylated
status of Hsp70 or Hsp90 C-terminus in p53-null
H1299 cells. Mutant p53 was co-transfected with Hsp70 and Hsp90a
and their point mutants. The levels of 53 were detected by DO-1
antibody. The phospho-mimetic mutants of Hsp70 and Hsp90 were
detected by phospho-specific antibodies GGS2.1 and GDD8.2,
respectively.
Figure 8. The chaperone status in cancer versus normal cells. (a) The expression levels of Hsp90a, Hsp70, their C-terminal phosphorylations
and the expression levels of co-chaperone HOP and CHIP in primary human breast tumors (T) and paired normal tissues (N). Actin was
detected as a loading control, proliferating cell nuclear antigen (PCNA) was detected as a marker of proliferation.
Figure 9. The balance between folding and degradation. The
C-terminal phosphorylation of Hsp70 and hsp90 in proliferating
cancer cells enhance the assembly with co-chaperone HOP. The nonphosphorylated
chaperones preferentially bind CHIP mediating
degradation of client proteins.
Balance between folding and degradation
P Muller et al
8
Oncogene (2012), 1 – 10 & 2012 Macmillan Publishers Limited
(NP_005336) and non-phosphorylated and phosphorylated C-termini of
human Hsp90a (NP_005339), respectively. Streptavidin-coated 96-well
plates were incubated for 1 h with biotinylated peptides (50 ml per well at a
final concentration 2 mM) dissolved in buffer-containing phosphatebuffered
saline, 0.1% NP-40 and 3% immunoglobulin-free AlbuminQ3 (Sigma
A3803). The plates were washed three times using 200 ml washing buffer
(phosphate-buffered saline þ 0.1% Tween 20). The peptide-coated plates
were incubated for 2 h with purified GST-tagged HOP and CHIP proteins.
After three washes in washing buffer the protein binding was detected by
horseradish peroxidase-conjugated anti-GST antibody diluted 1 mg/ml in
buffer-containing phosphate-buffered saline, 0.1% Tween 20 and 3%
Albumin. The signal was detected using the TMB Substrate Kit (Thermo
PierceQ4 ) following the manufacturer instructions.
Fluorescence polarization binding assay
The equilibrium bindings between a fixed concentration (30 nM) of a
fluorescent ligand and increasing concentrations of either HOP or CHIP in
buffer containing 150 mM NaCl, 50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic
acid (HEPES) pH 7.2, 1 mM dithiothreitol and 0.05%
Tween-20 were monitored by both total fluorescence intensity and
fluorescence polarization on the plate reader Filtermax F5 (Molecular
DevicesQ5 ). The ligands corresponding to C-terminal peptide of Hsp70
(GGSGSPTIEEVD), phospho-Hsp70 (GGSGSP(pT)IEEVD), Hsp90a
(GDDTSRMEEVD) or phospho-Hsp90a (GDDD(pT)(pS)RMEEVD) were
labeled by fluorescein on its N-terminus. The equilibrium dissociation
constant (Kd) was then calculated by fitting the sigmoidal dose-dependent
FP increases as a function of protein concentrations using Graphpad
Prism.31
Pull-down assay
Pull-down assay of streptavidin-binding protein-tagged GFP, Hsp70,
Hsp90a and Hsp90b was performed in H1299 and HEK-293 cells
transfected with plasmids expressing the corresponding proteins or their
point mutants. Cells were lysed in buffer containing 0.5% Tween 20,
150 mM NaCl, 50 mM HEPES (pH 7.2) and protease and phosphatase
inhibitors. Cell lysate (200 ml) containing 200–800 mg of total protein was
incubated with 15 ml of high capacity streptavidin agarose resin (Thermo
ScientificQ6 ) or with 50 ml of streptavidin-conjugated magnetic beads
(InvitrogenQ7 ). After three washes in lysis buffer, the beads were resuspended
in 60 ml of NuPAGE LDS Sample Buffer (Invitrogen). A total of 8 ml of
prepared samples were used for electrophoresis.
Co-immunoprecipitation
Cell pellets containing 107
were lysed in 600 ml lysis buffer containing 0.5%
3-((3-cholamidopropyl)dimethylammonio)-1-propane sulfonate, 0.1% NP-
40, 100 mM KCl, 50 mM Na-HEPES pH 7.3, 1 mM MgCl2, phosphatase and
protease inhibitor cocktails and 25 U Benzonase (Sigma-AldrichQ8 ). The cell
lysate was sonicated and centrifuged at 14 000 g for 15 min. The
supernatant was separated and the total protein concentration was
adjusted to 5 mg/ml in each sample. Protein G Sepharose (GE HealthcareQ9 )
was pre-incubated with excess rabbit polyclonal antibody and washed in
washing buffer containing 0.1% NP-40, 100 mM KCl and 50 mM Na-HEPES
pH 7.3. Protein lysate (180 ml) was incubated with 20 ml protein G sepharose
for 2 h at 4 1C. The immunoprecipitates were washed four times in washing
buffer (1 ml), eluted in 60 ml of NuPAGE LDS Sample Buffer (Invitrogen) and
10 ml used for electrophoresis.
Staining of protein complexes by proximity ligation assay in situ
Cells grown on poly-lysine-coated glass slides were fixed in AcetoneMethanol
(1:1) at À 20 1C for 10 min. Slides were rehydrated in phosphatebuffered
saline and incubated for 1 h in blocking solution (Duolink II-Olink
Bioscience Q10). The slides were incubated with the mixture of primary
antibodies containing 1 mg/ml of mouse monoclonal antibody recognizing
CHIP or HOP and rabbit polyclonal sera (1:10 000) recognizing Hsp70 or
Hsp90. Protein complexes were detected and visualized by proximity
ligation assay and rolling circle amplification using the Duolink II kit (Olink
Bioscience). The fluorescent signals were quantified using the Nikon Q11
Eclipse Ti fluorescence microscope and NIS Q12-Elements image analysis
software.
Protein purification
His–Hsp70 and His–Hsp90a fusion proteins were produced in BL21CodonPlus
(DE3)-RIPL (Agilent) and purified using the Ni–NTA bead Q13
(Qiagen). Cells were lysed in buffer containing 300 mM NaCl, 50 mM Tris pH
8.0, 20 mM Imidazole, 0.3 mM tris(2-carboxyethyl)phosphine, 1 mM phenylmethylsulfonyl
fluoride and lysozyme (1 mg/ml). The lysates were
sonicated and centrifuged (14 000 g) before separation on the Ni-NTA
columns. The His-tagged proteins were eluted in Imidazole gradient
(20–500 mM). GST-CHIP and GST-HOP proteins were produced in BL21CodonPlus
(DE3)-RIPL (Agilent) and purified using glutathione sepharose
(GE Healthcare). The cells were lysed in buffer containing 150 mM NaCl,
50 mM HEPES pH 7.5, 0.3 mM tris(2-carboxyethyl)phosphine and 1 mM
phenylmethylsulfonyl fluoride and lysozyme (1 mg/ml). The lysate was
incubated with glutathione sepharose after sonication and centrifugation
(14 000 g). The GST-fusion proteins were eluted in buffer containing
150 mM NaCl, 50 mM HEPES pH 7.5, 0.3 mM tris(2-carboxyethyl)phosphine
and 15 mM glutathione.
In vitro phosphorylation
The phosphorylation reaction mediated by CK1, CK2, GSK3B, CDK1/cyclinB
(New England Biolabs Q14) and activated p38 (Biaffin GmbH
Q15
) contained 1 mg of
recombinant Hsp90a or Hsp70, reaction buffer supplied by the manufacturer,
500 U of enzyme and 300 mM ATP in total volume 25 ml. The
reaction incubated at 30 1C was stopped after 30 min by addition of the
NuPAGE LDS Sample Buffer (Invitrogen).
Protein dephosphorylation
Cell pellet containing 5 Â 106
cells was lysed in 200 ml lysis buffer
containing 0.04% sodium dodecyl sulfate, 0.5% sodium deoxycholate,
1% Triton X-100, 150 mM NaCl, 2 mM MgCl2, 50 mM HEPES pH 7.3, protease
inhibitor cocktail (Sigma Aldrich) and 1 mM N-ethylmaleimide for irreversible
inhibition of natural phosphatases. N-ethylmaleimide was neutralized
after 0.5 h by addition of 2 mM dithiothreitol. Nucleic acids were digested
by 250 U of Benzonase Nuclease (Sigma Aldrich) for 0.5 h. The soluble
fraction of lysate was separated by centrifugation (14 000 g). We used Zeba
Spin Desalting Columns, 7K MWCO (Thermo Pierce) to remove lowmolecular-weight
compounds and exchange the lysis buffer to calf
intestinal phosphate Q16reaction buffer (100 mM NaCl, 50 mM Tris–HCl, 10 mM
MgCl2, 1 mM dithiothreitol and pH 7.9). The samples containing proteins
dissolved in reaction buffer were treated by 50 U of alkaline phosphatase
calf intestinal phosphate (New England Biolabs) for 1 h at 30 1C. The
reaction was stopped by addition of the NuPAGE LDS Sample Buffer
(Invitrogen).
Table 1. Differences in Hsp70 and Hsp90 chaperones in normal and cancer cells
Proliferating cancer cells Differentiated post-mitotic cells
Higher expression levels of Hsp70, Hsp90 and HOP Lower expression of Hsp70, Hsp90 and HOP, higher expression of CHIP
C-terminal phosphorylation of Hsp70 and Hsp90 Dephosphorylation of Hsp70 and Hsp90 C-terminus
Hsp70 and Hsp90 preferentially bind HOP Hsp70 and Hsp90 preferentially bind CHIP
Formation of multichaperone complexes The connection of Hsp70 and Hsp90 is reduced
Stabilization and folding of oncogenic client proteins Degradation of unfolded proteins
High folding capacity is essential to buffer genetic instability Hsp-mediated degradation maintains protein homeostasis
Malignant transformation, proliferation and metastasis Prevention of protein aggregation
Balance between folding and degradation
P Muller et al
9
& 2012 Macmillan Publishers Limited Oncogene (2012), 1 – 10
Kinase inhibitor screen
The kinase screening was performed with phosphatidylinositol 3-kinase
inhibitors Wortmanin (10 mM) and LY294002 (10 mM), mitogen-activated
protein kinase)/extracellular signal-regulated kinase inhibitors PD98059
(100 mM) and UO126 (50 mM), AKT inhibitors Triciribin (10 mM) and Tetrandrin
(10 mM), tyrosine kinase inhibitor Genistein (50 mM), CDK inhibitor Roscovitine
(50 mM), CDK9 and CK2 inhibitor DRB (50 mM) and p38 MAPK inhibitor
SB202190 (50 mM).
Human tumor tissues
Breast tissues after cancer surgery were divided into pieces that consist of
tumor or healthy tissue. These tissue samples were lysed by homogenization
(TissueLyser, Qiagen) in radioimmuno precipitation assay buffer
containing 50 mM Tris pH 7.4, 150 mM NaCl, 0.1% sodium dodecyl sulfate,
0.5% sodium deoxycholate, 1% Triton X 100 and protease and
phosphatase inhibitor cocktails (Sigma-Aldrich; nos. P8340, P0044 and
P5726). The soluble fraction was separated by centrifugation at 14 000 g
and the total concentration of soluble protein was adjusted to 2 mg/ml in
all samples.
CONFLICT OF INTEREST
Borivoj Vojtesek and Petr Muller work as consultants for Moravian Biotechnologies
that has developed some antibodies used in this study.
ACKNOWLEDGEMENTS
This work was supported by grants from GACR P301/11/1678, GACR P206/12/G151,
IGA MZ CR NS/9812-4, by the European Regional Development Fund and the State
Budget of the Czech Republic RECAMO CZ.1.05/2.1.00/03.0101 and research program
of the Ministry of Health of the Czech Republic FUNDIN MZ0MOU2005.
REFERENCES
1 Buchner J. Hsp90 & Co—a holding for folding. Trends Biochem Sci 1999; 24: 136–141.
2 Mayer MP, Bukau B. Molecular chaperones: the busy life of Hsp90. Curr Biol 1999;
9: R322–R325.
3 Taipale M, Jarosz DF, Lindquist S. HSP90 at the hub of protein homeostasis:
emerging mechanistic insights. Nat Rev Mol Cell Biol 2010; 11: 515–528.
4 Young JC, Agashe VR, Siegers K, Hartl FU. Pathways of chaperone-mediated
protein folding in the cytosol. Nat Rev Mol Cell Biol 2004; 5: 781–791.
5 Hohfeld J, Cyr DM, Patterson C. From the cradle to the grave: molecular chaperones
that may choose between folding and degradation. EMBO Rep 2001; 2:
885–890.
6 Krukenberg KA, Street TO, Lavery LA, Agard DA. Conformational dynamics of the
molecular chaperone Hsp90. Q Rev Biophys 2011; 44: 229–255.
7 Liu FH, Wu SJ, Hu SM, Hsiao CD, Wang C. Specific interaction of the 70-kDa heat
shock cognate protein with the tetratricopeptide repeats. J Biol Chem 1999; 274:
34425–34432.
8 Russell LC, Whitt SR, Chen MS, Chinkers M. Identification of conserved residues
required for the binding of a tetratricopeptide repeat domain to heat shock
protein 90. J Biol Chem 1999; 274: 20060–20063.
9 Ramsey AJ, Russell LC, Whitt SR, Chinkers M. Overlapping sites of tetratricopeptide
repeat protein binding and chaperone activity in heat shock protein 90. J Biol
Chem 2000; 275: 17857–17862.
10 Peterson LB, Blagg BS. To fold or not to fold: modulation and consequences of
Hsp90 inhibition. Future Med Chem 2009; 1: 267–283.
11 Hutt D, Balch WE. Cell biology. The proteome in balance. Science 2010; 329:
766–767.
12 Meggio F, Pinna LA. One-thousand-and-one substrates of protein kinase CK2?
FASEB J 2003; 17: 349–368.
13 Blom N, Gammeltoft S, Brunak S. Sequence and structure-based prediction of
eukaryotic protein phosphorylation sites. J Mol Biol 1999; 294: 1351–1362.
14 Scheufler C, Brinker A, Bourenkov G, Pegoraro S, Moroder L, Bartunik H et al.
Structure of TPR domain-peptide complexes: critical elements in the assembly of
the Hsp70-Hsp90 multichaperone machine. Cell 2000; 101: 199–210.
15 Zhang M, Windheim M, Roe SM, Peggie M, Cohen P, Prodromou C et al. Chaperoned
ubiquitylation—crystal structures of the CHIP U box E3 ubiquitin ligase
and a CHIP-Ubc13-Uev1a complex. Mol Cell 2005; 20: 525–538.
16 Brinker A, Scheufler C, Von Der Mulbe F, Fleckenstein B, Herrmann C, Jung G et al.
Ligand discrimination by TPR domains. Relevance and selectivity of
EEVD-recognition in Hsp70  Hop  Hsp90 complexes. J Biol Chem 2002; 277:
19265–19275.
17 Grummt I. Life on a planet of its own: regulation of RNA polymerase I transcription
in the nucleolus. Genes Dev 2003; 17: 1691–1702.
18 Muller P, Hrstka R, Coomber D, Lane DP, Vojtesek B. Chaperone-dependent stabilization
and degradation of p53 mutants. Oncogene 2008; 27: 3371–3383.
19 Kajiro M, Hirota R, Nakajima Y, Kawanowa K, So-ma K, Ito I et al. The ubiquitin
ligase CHIP acts as an upstream regulator of oncogenic pathways. Nat Cell Biol
2009; 11: 312–319.
20 Rhodes DR, Yu J, Shanker K, Deshpande N, Varambally R, Ghosh D et al.
ONCOMINE: a cancer microarray database and integrated data-mining platform.
Neoplasia 2004; 6: 1–6.
21 Miyata Y. Protein kinase CK2 in health and disease: CK2: the kinase controlling the
Hsp90 chaperone machinery. Cell Mol Life Sci 2009; 66: 1840–1849.
22 Lees-Miller SP, Anderson CW. The human double-stranded DNA-activated protein
kinase phosphorylates the 90-kDa heat-shock protein, hsp90 alpha at two NH2terminal
threonine residues. J Biol Chem 1989; 264: 17275–17280.
23 Mollapour M, Tsutsumi S, Donnelly AC, Beebe K, Tokita MJ, Lee MJ et al. Swe1Wee1-dependent
tyrosine phosphorylation of Hsp90 regulates distinct facets of
chaperone function. Mol Cell 2010; 37: 333–343.
24 Kundrat L, Regan L. Balance between folding and degradation for Hsp90dependent
client proteins: a key role for CHIP. Biochemistry 2010; 49: 7428–7438.
25 Li J, Soroka J, Buchner J. The Hsp90 chaperone machinery: conformational
dynamics and regulation by co-chaperones. Biochim Biophys Acta 2011; 1823:
624–635.
26 Mayer MP. Gymnastics of molecular chaperones. Mol Cell 2010; 39: 321–331.
27 Workman P. Altered states: selectively drugging the Hsp90 cancer chaperone.
Trends Mol Med 2004; 10: 47–51.
28 Powers MV, Jones K, Barillari C, Westwood I, van Montfort RL, Workman P. Targeting
HSP70: the second potentially druggable heat shock protein and molecular
chaperone? Cell Cycle 2010; 9: 1542–1550.
29 Pacey S, Wilson RH, Walton M, Eatock MM, Hardcastle A, Zetterlund A et al. A
phase I study of the heat shock protein 90 inhibitor alvespimycin (17-DMAG)
given intravenously to patients with advanced solid tumors. Clin Cancer Res 2011;
17: 1561–1570.
30 Keefe AD, Wilson DS, Seelig B, Szostak JW. One-step purification of recombinant
proteins using a nanomolar-affinity streptavidin-binding peptide, the SBP-Tag.
Protein Expr Purif 2001; 23: 440–446.
31 Zhang R, Mayhood T, Lipari P, Wang Y, Durkin J, Syto R et al. Fluorescence
polarization assay and inhibitor design for MDM2/p53 interaction. Anal Biochem
2004; 331: 138–146.
Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)
Balance between folding and degradation
P Muller et al
10
Oncogene (2012), 1 – 10 & 2012 Macmillan Publishers Limited