Molecular Cell, Vol. 5, 639–648, April, 2000, Copyright 2000 by Cell Press
Dynamics of Substrate Denaturation
and Translocation by the ClpXP
Degradation Machine
Yong-In Kim,*† Randall E. Burton,* Briana M. Burton,*
Robert T. Sauer,* and Tania A. Baker*†‡
* Department of Biology
† Howard Hughes Medical Institute
Massachusetts Institute of Technology
Cambridge, Massachusetts 02139
Summary
ClpXP is a protein machine composed of the ClpX
ATPase, a member of the Clp/Hsp100 family of remodeling enzymes, and the ClpP peptidase. Here, ClpX
and ClpXP are shown to catalyze denaturation of GFP
modified with an ssrA degradation tag. ClpX translocates this denatured protein into the proteolytic chamber of ClpP and, when proteolysis is blocked, also
catalyzes release of denatured GFP-ssrA from ClpP in
a reaction that requires ATP and additional substrate.
Kinetic experiments reveal that multiple reaction steps
require collaboration between ClpX and ClpP and that
denaturation is the rate-determining step in degradation. These insights into the mechanism of ClpXP explain how it executes efficient degradation in a manner
that is highly specific for tagged proteins, irrespective
of their intrinsic stabilities.
Introduction
The Clp/Hsp100 ATPases are a ubiquitous family of enzymes that unfold proteins, dismantle protein multimers,
and solubilize aggregates (Squires and Squires, 1992;
Levchenko et al., 1995; Rohrwild et al., 1996; Schirmer
et al., 1996; Gottesman et al., 1997; Pak and Wickner,
1997; Glover and Lindquist, 1998; Weber-Ban et al.,
1999). For example, E. coli ClpX catalyzes disassembly
of hyperstable complexes of MuA transposase tetramers bound to DNA (Levchenko et al., 1995; Kruklitis et
al., 1996). Yeast Hsp104 mediates the resolubilization
of heat-induced aggregates in the cell and modulates
the transition between the prion and nonprion forms of
the Sup35 protein (Chernoff et al., 1995; Newnam et al.,
1999). Some family members also play a role in degrading specific substrate proteins. For example, a protease
composed of the ClpX ATPase and ClpP peptidase
(Wojtkowiak et al., 1993) degrades the CtrA transcription
factor in Caulobacter crescentus, a reaction critical for
progression through the cell cycle (Jenal and Fuchs,
1998). ClpXP also plays a role in intracellular protein
quality control by degrading proteins modified by C-terminal addition of the ssrA degradation tag (Tu et al.,
1995; Keiler et al., 1996; Gottesman et al., 1998). This
peptide tag is cotranslationally added to nascent protein
fragments when ribosomes stall during translation
(Keiler et al., 1996; Roche and Sauer, 1999).
Electron microscopy and crystallographic studies
‡ To whom correspondence should be addressed (e-mail: tabaker@
mit.edu).
have defined the basic architecture of the Clp/Hsp100
ATPases and associated peptidases. The active ATPases
are hexameric rings (Parsell et al., 1994; Kessel et al.,
1995; Rohrwild et al., 1997; Grimaud et al., 1998; Bochtler et al., 2000). The ClpP peptidase consists of two
heptameric rings, which enclose a large central chamber
with two, small axial pores (Wang et al., 1997). The 14
active site serines in the ClpP multimer are sequestered
on the inner surface of this chamber. By itself, ClpP can
degrade peptides but not native proteins (Maurizi et al.,
1994), which are too large to pass though the axial pores
(Wang et al., 1997). To degrade protein substrates, ClpP
forms a complex with one or two rings of ClpX or ClpA
to generate the ClpXP or the ClpAP proteases. In these
complexes, ATP is required for stable multimerization
of the ATPases, for unfolding and chaperone activity,
and for degradation of protein substrates (Kessel et al.,
1995; Grimaud et al., 1998; Hoskins et al., 1998; Maurizi
et al., 1998; Pak et al., 1999). Biochemical analysis
of ClpAP has led to an appealing model in which the
ATPase rings bind native substrates, alter their conformation, and translocate them through the axial pores of
ClpP and into the proteolytic chamber for destruction
(Gottesman et al., 1997). Two recent studies support
this model. ClpA was shown to promote global unfolding
of a model substrate, GFP carrying an ssrA degradation
tag (Weber-Ban et al., 1999), and translocation of substrates from ClpA to ClpP was also demonstrated (Hoskins et al., 1998).
Here, we examine the dynamics of substrate processing by the ClpXP degradation machine. Despite being only about half the size of ClpA, ClpX also dramatically accelerates the unfolding of a GFP-ssrA fusion
and transports the denatured polypeptide to ClpP. In
addition, we show that ClpX catalyzes substrate release
from ClpP and demonstrate that substrate denaturation
is the rate-determining catalytic step in the overall degradation cycle. Finally, we find that the sequence of the
C-terminal degradation tag, rather than the structure or
stability of the attached protein, determines the efficiency of substrate recognition.
Results
To investigate the dynamics of substrate processing by
the ClpXP complex, green fluorescent protein bearing
a C-terminal ssrA degradation tag (GFP-ssrA) was chosen as a model substrate (Andersen et al., 1998). GFP
is useful for studies of this type because its conformation
can be monitored in real time by changes in fluorescence
and absorbance (Palm et al., 1997). The ssrA peptide
tag converts GFP into a substrate for ClpXP. Initial experiments with purified components established that
GFP-ssrA was efficiently degraded in a reaction requiring ClpX, ClpP, and ATP, both as assayed by the release
of acid-soluble peptides from 35S-labeled GFP-ssrA (Figure 1A) and by a decrease in native GFP fluorescence
at 511 nm (Figure 1B). Untagged GFP was not degraded
by ClpXP.
Molecular Cell
640
Figure 2. Denatured GFP-SsrA Copurifies with DFP-Modified ClpP
(A) [35S]GFP-ssrA was denatured by ClpXPDFP and chromatographed
on a Superdex 200 column. The elution positions of molecular weight
standards are shown.
(B) SDS-PAGE of fractions 10–19.
(C) Autoradiograph of the gel shown in (B).
Figure 1. ClpX and ClpXPDFP Catalyze Denaturation but Not Degradation of GFP-SsrA
(A) Release of acid-soluble [35S]peptides from 0.25 ␮M [35S]GFPssrA following incubation with ClpXP or ClpXPDFP.
(B) Change in fluorescence at 511 nm of 0.21 ␮M GFP-ssrA following
incubation with ClpXP or ClpXPDFP.
(A and B) The ClpX6 concentration was 0.3 ␮M, the concentration
of modified or unmodified ClpP14 was 1 ␮M, and the concentration
of GroEL14 trap was 10 ␮M. Gel filtration experiments confirmed that
denaturated GFP-ssrA was bound to the GroEL trap when the latter
molecule was present (data not shown).
(C) Rate constants for GuHCl denaturation of GFP-ssrA. The dashed
line shows a semilogarithmic fit of the data (r ⫽ 0.998; intercept ⫽
6.4 ⫻ 10⫺8 s⫺1).
ClpX and ClpXP Catalyze Denaturation of GFP-SsrA
Reactions containing ClpX, ATP, and GFP-ssrA were
also performed in the absence of ClpP or with a derivative in which the active site serine was inactivated
by modification with diisopropylfluorophosphate (DFP)
(Hwang et al., 1988; Maurizi et al., 1990). We refer to
this inactive peptidase as ClpPDFP and to its complex
with ClpX as ClpXPDFP. As expected, neither ClpX alone
nor the ClpXPDFP complex degraded [35S]GFP-ssrA (Figure 1A; data not shown). However, ClpXPDFP did catalyze
an ATP-dependent reduction of 50%–60% in GFP-ssrA
fluorescence (Figure 1B), as well as a decrease in native
absorbance at 500 nm, and the appearance of nonnative
absorbance peaks at 395 and 402 nm (data not shown).
These spectral changes are similar to those observed
for denatured GFP and are diagnostic of a major disruption of the native structure around the GFP chromophore, which is normally buried in the protein core (Palm
et al., 1997). These observations suggest that ClpXPDFP
denatures GFP-ssrA. No change in GFP-ssrA fluorescence was observed upon incubation with ClpPDFP in
the absence of ClpX.
ClpX alone also promoted a small decrease in GFPssrA fluorescence, which was magnified to about half
the efficiency of the ClpXPDFP-catalyzed reaction by
addition of a GroEL variant that binds irreversibly to
nonnative proteins (Fenton et al., 1994) (Figure 1B).
GroEL, by itself, did not change the fluorescence of
GFP-ssrA. These results provide evidence that ClpX,
like ClpA (Weber-Ban et al., 1999), denatures GFP-ssrA.
The half-life of spontaneous GFP-ssrA denaturation,
estimated from extrapolation of GuHCl unfolding data,
was approximately 20 years (Figure 1C). Hence, ClpXPDFP
actively promotes denaturation of GFP-ssrA, increasing
the uncatalyzed rate by almost seven orders of magnitude.
The unmodified ClpXP complex is undoubtedly also able
to catalyze denaturation of protein substrates.
Substrate Proteins Are Trapped by Inactive ClpP
To determine the fate of GFP-ssrA following denaturation, the products of reactions containing [35S]GFPssrA, ClpXPDFP, and ATP were separated by gel filtration,
and column fractions were analyzed by scintillation
counting and SDS-PAGE. Under the chromatographic
conditions used, ClpPDFP and ClpX eluted at distinct positions. When the initial denaturation reaction contained
ClpXPDFP and ATP, about 70% of the [35S]GFP-ssrA coeluted with ClpPDFP, and ⵑ25% eluted at a position expected for unbound GFP (Figure 2). When ClpX, ClpPDFP,
or ATP was absent from the initial reaction, more than
Dynamics of the ClpXP Degradation Cycle
641
95% of the [35S]GFP-ssrA eluted as unbound protein
(data not shown). Addition of chymotrypsin to the purified ClpPDFP·[35S]GFP-ssrA complex did not significantly
degrade GFP-ssrA, suggesting that this substrate was
sequestered within the ClpP active site chamber (data
not shown). These results support a mechanism in which
ClpX mediates ATP-dependent denaturation and translocation of GFP-ssrA into the central chamber of the
ClpPDFP 14-mer, where it can become stably associated
or “trapped.” Stoichiometry experiments performed
with high concentrations of the substrate and enzyme
indicated that ⵑ0.8 moles of GFP-ssrA were trapped
per mole of ClpXPDFP complex.
ClpX Promotes Release of Substrate Trapped
in ClpP
Denaturation reactions containing GFP-ssrA, ATP, and
ClpXPDFP showed an initial loss of fluorescence but then
reached a plateau, even when enzyme was present in
significant excess over substrate (Figure 1B; data not
shown). This observation suggested that denaturation
might reach an equilibrium in which trapped GFP-ssrA
was released and refolded at a rate equal to the rate of
denaturation of new GFP-ssrA molecules.
To investigate whether GFP-ssrA “cycles” though
ClpXPDFP, denaturation and trapping of GFP-ssrA were
allowed to proceed for 30 min and the resulting complexes were challenged with a large excess of another
substrate, the N-terminal domain of ␭ cI repressor containing an ssrA tag (␭cI-N-ssrA; Gottesman et al., 1998).
Within minutes of adding ␭cI-N-ssrA, GFP-ssrA fluorescence began to increase, and more than half of the
fluorescence initially lost as a result of ClpXPDFP-mediated denaturation was regained after 30 min (Figure
3A). This restoration of native GFP fluorescence was
accompanied by accumulation of free, monomeric GFPssrA, showing that substrate was physically released
from ClpXPDFP (Figure 3B). In a control reaction, a poorly
recognized variant of ␭cI-N-ssrA in which the C-terminal
residues were DD instead of AA was ineffective at increasing GFP-ssrA fluorescence (Figure 3A). Renaturation of GFP-ssrA following dilution from 6 M GuHCl
(half-life ⬇ 3 min) was faster than GFP-ssrA release and
refolding from ClpXPDFP. These results suggest that the
rate of reappearance of fluorescent GFP-ssrA is largely
determined by its rate of release, assuming that the
chemically denatured protein is similar to that denatured
by ClpXPDFP.
These experiments show that denatured GFP-ssrA
trapped in ClpXPDFP can be released into solution to
refold. In experiments similar to those of Figure 3A, very
little restoration of GFP-ssrA fluorescence was observed after addition of ␭cI-N-ssrA when ATP was depleted by addition of hexokinase and its substrate glucose (data not shown). This observation indicates that
release requires the continual presence of ATP and thus
the active participation of ClpX.
Does second substrate facilitate reappearance of native GFP-ssrA by directly stimulating the release of kinetically trapped GFP-ssrA? Or does this substrate simply
compete for binding to ClpXPDFP, thereby reducing the
rate of further GFP-ssrA uptake and the amount of denatured GFP trapped at steady state? These two models
have a simple distinguishing feature. By the former
Figure 3. Trapped GFP-SsrA Is Released from ClpXPDFP by Addition
of a Second Substrate
(A) Time course of GFP-ssrA denaturation and second substrate–
mediated renaturation. Denaturation reactions (100 ␮l) containing
0.2 ␮M GFP-ssrA, 0.3 ␮M ClpX6, and 1.3 ␮M ClpPDFP14 were allowed
to proceed for ⵑ30 min at 30⬚C, and 20 ␮l of a second substrate (␭cIN-ssrA or ␭cI-N-ssrA-DD) was added to a concentration of 83 ␮M.
(B) Gel filtration (Superdex 200) of a sample 60 min after addition
of ␭cI-N-ssrA, showing that trapped GFP-ssrA is largely released
from ClpXPDFP. Reactions as in (A), except [35S]GFP-ssrA was used.
model, the second substrate should be actively required
for GFP-ssrA release. By the latter model, GFP-ssrA
should be released from ClpXPDFP even in the absence
of the second substrate and, therefore, be detectable
under conditions where uptake and denaturation of new
GFP-ssrA is slowed. In this second model, the second
substrate simply competes for uptake or “space” within
ClpXPDFP, thereby causing a net increase in the amount
of native GFP-ssrA present, without effecting its rate of
release per se.
To test these models, reactions containing ClpXPDFP
and GFP-ssrA were allowed to reach equilibrium and
were then diluted 10-fold into a reaction mixture containing ATP and ClpX with or without ␭cI-N-ssrA. In the
absence of second substrate, only a small amount of
trapped GFP-ssrA was released and refolded to give
fluorescent protein after dilution (Figure 4A). This dilution
step decreased the rate of uptake of new GFP-ssrA by
a factor of 10, based on an independently determined
substrate concentration versus reaction rate curve (data
not shown, but see Figure 5; the dependence of rate on
[substrate] was the same for degradation and denaturation). Therefore, under these conditions, GFP-ssrA
should have been efficiently released if the trapped substrate was able to escape in a reaction independent of
the substrate binding or uptake process. Because it did
Molecular Cell
642
Figure 4. ATP and Second Substrate Directly
Stimulate Release of Trapped GFP-SsrA
(A) Denaturation of GFP-ssrA by ClpXPDFP was
carried out for 30 min (see Figure 3A legend),
and reactions were diluted 10-fold into PD
buffer (which contains ATP) with or without
0.3 ␮M ClpX and/or 25 ␮M ␭cI-N-ssrA. The
fluorescence scale is 10-fold smaller for the
data from 30–90 min to account for the dilution. In reconstruction experiments, the excess free ClpX present in this experiment had
a negligible effect on the amount of native
GFP-ssrA detected.
(B) Release of trapped GFP-ssrA from purified ClpPDFP complexes. Reactions contained
5 mM ATP or ATP␥S, 2 ␮g of purified
ClpPDFP·GFP-ssrA, 0.3 ␮M ClpX, with or without 25 ␮M ␭ cI-N-ssrA.
(C) Proposed pathway of substrate trapping
by and release from ClpXPDFP. GFP-ssrA is
denatured by ClpX and translocated to ClpP,
where it resides stably enough to copurify
by gel filtration. In the presence of ATP and
additional substrate, ClpX actively promotes
release of the denatured GFP-ssrA from ClpP
into solution, where it can refold.
not, these results suggest that the second substrate
is required actively, supporting the facilitation model.
Indeed, as expected by this model, when the initial reaction was diluted into buffer containing high concentrations of ␭cI-N-ssrA, more than 95% of the bound GFPssrA was released and refolded (Figure 4A).
In a second test, the ClpPDFP·GFP-ssrA complex was
purified from free components by gel filtration and was
then reconstituted with ClpX and ATP, either in the presence or absence of ␭cI-N-ssrA (Figure 4B). Again, the
results clearly reveal stimulation of release by ␭cI-NssrA. Because the initial concentration of free GFP-ssrA
was negligible in this experiment, the rate of GFP-ssrA
binding and uptake would also be insignificant. Some
GFP-ssrA was released in the absence of ␭cI-N-ssrA,
but at a much slower rate. Taken together, these experiments demonstrate the existence of a significant kinetic
barrier to release of GFP-ssrA trapped in ClpXPDFP; this
barrier is lowered by the presence of ATP and substrate
interacting with ClpX.
The experiments presented above establish that substrates cycle through the ClpXPDFP complex. Each reaction cycle must minimally consist of: binding of substrate
to ClpX to form a productive enzyme·substrate complex,
substrate denaturation, translocation of denatured substrate from ClpX to ClpPDFP, release of denatured substrate, and refolding. The experiments described here
establish that efficient GFP-ssrA release from ClpPDFP
requires ClpX, ATP, and a second substrate molecule
(Figure 4C). These results, in turn, indicate that catalytic
events within the ClpX portion of this modular protease
influence the way in which the ClpP component interacts
with denatured substrates within its proteolytic chamber.
Substrate Denaturation Is Rate Limiting
for Degradation
Substrate degradation by unmodified ClpXP must also
be a multistep reaction. In this case, the minimal model
includes five steps: (1) productive binding of substrate
Dynamics of the ClpXP Degradation Cycle
643
Figure 5. Substrate Denaturation and Degradation Occur at the Same Rate
(A) Rates of degradation of [35S]GFP-ssrA by
ClpXP were assayed at different substrate
concentrations by the release of acid-soluble
peptides (closed triangle), and rates of denaturation/degradation of unlabeled GFP-ssrA
were assayed by changes in fluorescence
(closed circle). The solid line is a fit of the
combined data to a Michaelis-Menten model
(r ⫽ 0.99; Km ⫽ 1.95 ⫾ 0.26 ␮M; Vmax/
ClpX6-total ⫽ 0.94 ⫾ 0.04 min⫺1). (Inset) Time
courses of the fluorescence and peptide release assays at ⵑ6 ␮M GFP-ssrA. Fluorescence change is shown by the small symbols
and peptide release by the larger filled symbols. Predicted values for fluorescence and
degradation, calculated from the kinetic
model with denaturation as the slow step, are
shown by the gray and black lines, respectively.
(B) Kinetic scheme for GFP-ssrA degradation.
ATP is always present, but the ATPase cycle
is not represented; for simplicity, the equilibrium for ClpX·ClpP complex formation is also
not shown. Degradation is initiated by binding
of the ssrA peptide tag of the substrate to
the ClpX component of the ClpXP complex.
In the slowest step of the overall reaction,
ClpX catalyzes denaturation of bound substrate. Denatured protein is translocated to
the ClpP chamber in a reaction that may be
facilitated by ClpX-mediated opening of the
axial pore of ClpP. The denatured and translocated protein is degraded to peptide fragments by the ClpP active sites. Following
degradation, peptides are released from the
complex. Although not shown, based on its
stimulation of protein release from ClpPDFP,
ClpX is likely to also facilitate this release
reaction. The assay used to measure degradation will detect both free peptides and
those still bound to ClpP; therefore, we have
not measured release, independent from
degradation.
to ClpXP, (2) substrate denaturation, (3) translocation, (4)
substrate cleavage, and (5) release of peptide fragments.
To address which of these steps might limit the overall
rate of protein degradation, we determined the kinetics
of GFP-ssrA degradation by ClpXP both by the disappearance of GFP fluorescence and by the appearance
of acid soluble 35S-labeled GFP peptides at a series of
substrate concentrations (Figure 5A). At each substrate
concentration tested, denaturation and degradation occurred with very similar kinetics (Figure 5A). The combined data from both assays were fit to a kinetic model
consisting of the five reaction steps listed above (Figure
5B). A satisfactory fit of the experimental data over the
entire range of substrate concentrations could only be
obtained when the equilibrium constant for substrate
binding was about 2 ␮M and the rate constant for enzyme-mediated GFP-ssrA denaturation was roughly 0.9
min⫺1 and at least 10-fold slower than the rate constants
for translocation or degradation (Figure 5B; see Experimental Procedures). These findings suggest strongly
that, once GFP-ssrA is engaged by the ClpXP complex,
substrate denaturation is the slowest step in the overall
reaction and therefore determines the rate of degradation.
Productive Interaction of Substrate with ClpXP
Depends on the Degradation Signal and Is
Independent of Protein Structure
It has been established that addition of an ssrA tag or
the last ten residues of MuA transposase to the C terminus of unrelated proteins converts these proteins into
substrates for ClpXP (Levchenko et al., 1997a; Gottesman et al., 1998), but it is not known whether the nature
of the “host” protein influences recognition. The ability
of a “second” substrate to promote release and renaturation of GFP-ssrA from ClpXPDFP·GFP-ssrA complexes
provided a convenient assay to quantitate how different
substrate proteins interact with ClpXPDFP. Moreover, because ClpPDFP is proteolytically inactive, this assay
avoids complications that might be introduced by degradation of either substrate.
To address whether features of the substrate protein
other than the C-terminal tag influence recognition by
Molecular Cell
644
Figure 6. Rate of Release of Trapped GFP-SsrA Depends on the
Second Substrate Concentration and the Degradation Signal
Denaturation reactions (100 ␮l) containing 0.3 ␮M ClpX6, 0.8 ␮M
ClpPDFP14, and 0.2 ␮M GFP-ssrA were performed for 30 min at 30⬚C;
different concentrations of ␭cI-N-ssrA (closed triangle), Arc-ssrA
(closed circle), or Arc-MuA10 (closed square) were added (in 20 ␮l);
and initial rates of GFP-ssrA release/refolding (see Figure 3A and
4A) were determined by linear regression of fluorescence data. The
rates shown were obtained by dividing the initial rates by the concentration of ClpX6PDFP14·GFP-ssrA (13.6 pmol/120 ␮l), as estimated
from the fluorescence value at the time of addition of the second
substrate. The solid lines are fits to the equation: rate ⫽ Vmax·(1 ⫹
K1/2/[second substrate])⫺1.
ClpXP, release assays were performed using three second substrates: ␭cI-N-ssrA, Arc-ssrA (P22 Arc repressor
with an ssrA tag), and Arc-MuA10 (Arc repressor with
the ten C-terminal residues of MuA transposase). In each
case, the ability of different concentrations of these proteins to promote release of trapped GFP-ssrA was measured (Figure 6). Arc-ssrA and ␭cI-N-ssrA showed similar behavior in these assays, with protein concentrations
supporting half-maximal stimulation of 1.1 ⫾ 0.1 ␮M
and 1.3 ⫾ 0.4 ␮M, respectively. For Arc-MuA10, the
concentration required to achieve half-maximal stimulation of the release rate was at least 10-fold higher (14 ⫾
3 ␮M) than for the ssrA-tagged proteins (Figure 6). ArcssrA and Arc-MuA10 are identical, except for the sequences of their C-terminal peptide tags, yet 13-fold
higher concentration of Arc-MuA10 than Arc-ssrA was
required to achieve the same rate of release. Moreover,
Arc-ssrA and ␭cI-N-ssrA are unrelated except for their
ssrA tags and yet interact with the ClpXP complex in a
manner almost indistinguishable by these measurements.
Based on these data, we conclude that the C-terminal
peptide tag sequences of these substrates are the principal factor determining the strength of productive
ClpXP interactions.
To elucidate aspects of the ssrA tag that contribute
to its interaction with ClpX, three variants of the ssrApeptide sequence were tested for their ability to inhibit
ClpXPDFP-mediated denaturation of GFP-ssrA. ClpXPDFP
was used to avoid complications associated with peptide degradation. The peptides assayed were: NH2CAANDENYALAA-COOH (this sequence is that of the
Figure 7. Inhibition of GFP-SsrA Denaturation by SsrA Peptides
Kinetics of ClpXPDFP denaturation of GFP-ssrA in the presence of
no peptide (⫺), wild-type ssrA peptide (ssrA), carboxamide ssrA
peptide (CONH2), and the AA→DD ssrA peptide derivative (see text).
Reaction mixtures with 0.3 ␮M ClpX6 and 0.8 ␮M ClpPDFP14 were
preincubated for 2 min at 30⬚C, and a mixture of GFP-ssrA (1.3 ␮M)
and peptide (280 ␮M) was added. Inset shows a Lineweaver-Burk
plot for the inhibition of GFP-ssrA denaturation by the wild-type
ssrA peptide and ␭cI-N-ssrA, both at 13 ␮M.
wild-type ssrA tag with an additional N-terminal cysteine); NH2-CAANDENYALAA-CONH2 (identical to the first
peptide but with a carboxamide group instead of an
␣-carboxyl group at the C terminus), and NH2-CAANDENYALDD-COOH (identical to the first peptide but with
DD instead of AA as the final two residues). The wildtype ssrA peptide clearly inhibited denaturation of GFPssrA (Figure 7); the extent of inhibition by this peptide
was very similar to that observed by the same concentrations of ␭cI-N-ssrA (see inset). The carboxamide peptide inhibited somewhat less well, and the DD-peptide
was an extremely poor inhibitor (Figure 7). Experiments
performed using a range of peptide concentrations
showed that the carboxamide peptide was about 10fold less effective than the wild-type peptide and the
DD-peptide was at least 50-fold less effective. Hence,
these data confirm that the sequence at the C-terminal
end of the ssrA tag is important for interactions with
ClpXP (Levchenko et al., 1997b; Gottesman et al., 1998;
Smith et al., 1999) and also demonstrate that a free
␣-carboxyl group is an important molecular determinant
of this interaction.
Discussion
ClpX Is a Protein Unfoldase
Enzyme catalyzed protein unfolding is an important aspect of many cellular processes and appears to be the
unifying biochemical activity of the Clp/Hsp100 family
of ATPases (Squires and Squires, 1992; Schirmer et al.,
1996; Gottesman et al., 1997). The first direct demonstration of catalyzed denaturation for this family was recently provided by Weber-Ban et al. (1999), who showed
Dynamics of the ClpXP Degradation Cycle
645
that E. coli ClpA—a class I family member with two
ATPase domains—catalyzed the global unfolding of
GFP-ssrA. ClpX is a class II family member that contains
a single ATPase domain and lacks several additional
domains present in ClpA (Schirmer et al., 1996). Nevertheless, ClpX and ClpXP also catalyze ATP-dependent
denaturation of GFP-ssrA. ClpXPDFP accelerates denaturation of GFP-ssrA, an extremely stable protein, by a
factor of approximately 107. This rate enhancement is
comparable to that observed in 7 M GuHCl, a strong
chemical denaturant, and corresponds to a decrease of
9.8 kcal/mol in the activation free-energy barrier between the native state and the transition state for denaturation. Thus, we conclude that ClpX and ClpXP have
the capacity to catalyze ATP-dependent denaturation of
ssrA-tagged protein substrates and suggest that other
native substrate proteins are similarly altered by the
catalytic action of these proteins.
Substrate Denaturation Is the Slow Step
during GFP Degradation by ClpXP
The rate of ClpXP-mediated denaturation of GFP-ssrA,
as measured by loss of fluorescence, and the rate of
ClpXP-mediated degradation of GFP-ssrA, as measured
by appearance of labeled peptide fragments, were
found to be indistinguishable within experimental error.
These observations suggest that once GFP-ssrA is productively bound by the ClpXP complex, denaturation is
the slow, rate-determining step in the overall degradation reaction (Figure 5). Cleavage of small molecule substrates by ClpP is known to be very fast (⬎104 min⫺1)
compared to the rates of ClpXP degradation of protein
substrates (Grimaud et al., 1998), consistent with models
in which the rate-limiting step for protein degradation
is prior to peptide bond hydrolysis. Our experiments
further argue that translocation from ClpX to ClpP is fast
compared to denaturation. Although a detailed comparison has not been reported, the denaturation and degradation reactions of GFP-ssrA by ClpAP at one substrate
concentration also appear to occur with roughly similar
kinetics (Weber-Ban et al., 1999), suggesting that denaturation is also the slow step in this reaction.
Substrate Release Involves Collaboration
between the ClpX and ClpP Modules
The use of DFP-inactivated ClpP allowed us to study
the properties of translocated GFP-ssrA. For example,
[35S]GFP-ssrA copurified with ClpPDFP in a form resistant
to digestion by chymotrypsin, suggesting that it resides
within the active site chamber of the ClpP 14-mer. This
central chamber has a radius of ⵑ25 Å (Wang et al.,
1997), adequate for a single molecule of denatured GFPssrA (calculated radius of gyration 24 Å), and stoichiometry experiments indicate that one molecule of GFPssrA is trapped within the ClpXP complex. Wang et al.
(1997) suggested that the hydrophobic surface of the
internal chamber of ClpP could help keep substrates in
an unfolded and easily degradable state, consistent with
our observation that trapped GFP-ssrA remains largely
nonfluorescent.
In addition to its role in catalyzing substrate denaturation and translocation, the ClpX ATPase also catalyzed
active release of trapped GFP-ssrA from ClpPDFP. The
latter role for ClpX in modulating ClpP interactions with
trapped molecules was unanticipated but can be rationalized in both mechanistic and functional terms. For
example, Wang et al. (1997) proposed that the Clp
ATPase cycle opens the narrow axial pores of the ClpP
cylinder to allow denatured substrates to enter the degradation chamber. We suggest that the same mechanism could also permit egress of undegradable proteins
or cleaved peptide products. It has been assumed that
peptide fragments exit the ClpP proteolytic chamber by
passive diffusion (Thompson et al., 1994; Wang et al.,
1997), but this may only be true for very small peptides.
In the absence of a Clp ATPase, for example, only peptides of five or fewer residues appear to enter ClpP
efficiently—as judged by their ability to be degraded
(Thompson and Maurizi, 1994; Thompson et al., 1994)—
and the same limitation could exist for peptide exit. In
this case, a mechanism to facilitate peptide release
would be necessary. Peptide fragments between seven
and ten residues are common products of degradation
and substantially larger peptide products have also
been detected (Thompson et al., 1994). Moreover, if
ClpP binding to denatured proteins is required to prevent refolding and position the substrate for multiple
rounds of processive hydrolysis (Wang et al., 1997), then
a role for the Clp ATPases in promoting product release
may also be important for maintaining efficient turnover
during multiple rounds of catalysis.
An interesting aspect of the finding that ClpX promotes release of substrate trapped in ClpP is the stimulation of this reaction by additional substrate. There are
two models that might explain this stimulation. First,
simply by binding to ClpX, additional substrate may enhance the rate of ATP-hydrolysis, which in turn could
promote axial pore opening and efficient substrate release. Second, uptake and translocation of the additional substrate might be required to help push the
trapped substrate out of the ClpP chamber. Experiments
designed to test these models are in progress. By either
model, however, this feature of the system would allow
the degradation machine to attempt to degrade trapped
substrates as completely as possible and only release
undegraded protein, or larger peptide products, when
the need for degradation of fresh substrates arises.
Peptide Tail·ClpX Interactions Determine the Strength
and Specificity of Substrate Binding
Using the GFP-ssrA-based denaturation and release
assays described here, we were able to compare how
different ssrA-tagged substrates interact with ClpXP.
We found that the concentrations required to support
half-maximal reaction rates for GFP-ssrA, Arc-ssrA, and
␭cI-N-ssrA were all between 1 and 2 ␮M, despite the
fact that these proteins differ dramatically in sequence,
structure, and stability. For example, Arc repressor and
the N-terminal domain of ␭ repressor are predominantly
␣-helical proteins that unfold with half-lives of less than
1 min (Milla and Sauer, 1994; Huang and Oas, 1995),
whereas GFP is mainly a ␤ sheet protein that unfolds
with a half-life estimated to be roughly 20 years. We
conclude that the ssrA tag is the principal factor contributing to productive association with ClpXP and that additional sequence or structural information from the protein that is tagged contributes relatively little information
that influences ClpXP recognition.
Molecular Cell
646
The finding that the 11-residue ssrA tag is the principal
feature within tagged proteins recognized by ClpXP
makes biological sense, as this would ensure that addition of the ssrA tag to a variety of unrelated proteins
during translation would be sufficient to mark these proteins for proteolytic destruction. Although protein fragments generated by ssrA-mediated tagging during
stalled translation would be expected to display a wide
range of stabilities, this appears to play little or no role
in the recognition of these proteins by ClpXP. Clearly,
this aspect of the ClpXP system is quite different from
many chaperone systems, in which hydrophobic residues exposed in denatured substrates are the principal
factor recognized by the enzyme (Fenton and Horwich,
1997; Rüdiger et al., 1997). Although our results indicate
that substrate stability does not influence ClpXP recognition, it is possible that ssrA-tagged substrates with
different stabilities are denatured and therefore degraded at different rates. This model could explain, for
example, why the rates of degradation of ␭cI-N and Arc
variants in E. coli show a strong correlation with protein
stability (Parsell et al., 1990; Milla et al., 1993).
Inhibition of ClpXPDFP-mediated denaturation of GFPssrA by variants of the ssrA peptide further delineates
properties of this sequence important for recognition.
For example, these experiments show that mutation of
the last two residues of the peptide from AA→DD decreases binding to below detectable levels. The AA→DD
mutation also dramatically slows the rate of ␭cI-N-ssrA
degradation in vivo and in vitro (Gottesman et al., 1998),
reduces affinity for the substrate binding domain of ClpX
(Levchenko et al., 1997b), and drastically reduces the
ability of ␭cI-N-ssrA to catalyze release of trapped GFPssrA from ClpXPDFP. All of these properties are readily
explained if the DD mutation causes a large decrease
in productive binding to ClpX. The peptide inhibition
experiments also reveal that a free ␣-carboxylate at the
peptide terminus contributes to ClpX·substrate interactions, although replacing the charged -COO- group with
the neutral -CONH2 group reduced apparent affinity by
only 10-fold. We note that some ClpXP substrates appear to be recognized via peptide signals that are not
located at the protein C terminus (Schweder et al., 1996;
Gonciarz-Swiatek et al., 1999), which is consistent with
our finding that a free ␣-carboxylate is not essential for
ClpX recognition.
What Is the Mechanism of Protein-Catalyzed
Protein Denaturation?
How do the Clp/Hsp100 ATPases promote the denaturation of substrate proteins? Models in which these enzymes simply bind to and trap spontaneously denatured
molecules are ruled out by the remarkably large enhancements in the rate of substrate denaturation catalyzed by ClpX and ClpA. Furthermore, it seems unlikely
that the Clp/Hsp100 enzymes could provide a chemical
environment that would function to denature substrate
proteins in a fashion analogous to chaotrophic agents
such as urea or GuHCl, which reduce the strength of
hydrophobic forces. The most likely models would
therefore be those in which the Clp/Hsp100 ATPases
catalyze denaturation by exerting mechanical stress on
native protein substrates. Such mechanisms require a
way to secure the native substrate to the enzyme and
then to generate and apply force to this molecule via
conformational rearrangements in the Clp enzyme,
which are presumably driven by ATP binding or hydrolysis. It is, of course, the structural and mechanistic details
of such models that will need to be elucidated to provide
a molecular understanding of this example of enzymecatalyzed protein denaturation.
Experimental Procedures
Strains and Plasmids
E. coli strain JB401 (a gift of A. J. Anderson, The Technical University
of Denmark) expresses GFP-ssrA with mutations (S65G and S72A)
that affect its spectral properties and may improve folding in the
cell (Cormack et al., 1996). A plasmid expressing E. coli ClpP with
a C-terminal His6 tag (pYK133) was constructed by PCR amplification of the clpP gene, cleavage with SphI and BglII, and cloning into
the SphI to BglII backbone fragment of pQE70 (Qiagen). The clpP
gene in pYK133 was sequenced to ensure the absence of mutations.
Buffers
PD buffer contains 25 mM HEPES-KOH (pH 7.6), 5 mM KCl, 5 mM
MgCl2, 0.032% NP-40, 10% glycerol, 5 mM ATP, and an ATP-regenerating system consisting of 16 mM creatine phosphate and 0.32
mg/ml creatine kinase. Clp buffer contains 50 mM Tris-HCl (pH 7.5),
200 mM KCl, 25 mM MgCl2, 1 mM DTT, 0.1 mM EDTA, and 10%
glycerol. S buffer contains 50 mM Na-phosphate (pH 8.0), 1 M NaCl,
5 mM imidazole, and 10% glycerol. W20 buffer contains 50 mM Naphosphate (pH 8.0), 1 M NaCl, 10% glycerol, and 20 mM imidazole;
W500 buffer is the same but contains 500 mM imidazole. Q50 buffer
contains 50 mM Tris (pH 8.0), 5 mM DTT, 10 mM MgCl2, 10% glycerol,
and 50 mM KCl; Q1000 buffer is the same but contains 1 M KCl.
Proteins and Peptides
E. coli ClpX protein with an N-terminal His6 tag was expressed and
purified as described (Levchenko et al., 1997a). E. coli ClpP protein
with a C-terminal His6 tag was expressed in E. coli strain DH5␣/
pYK133/pRep4. Cultures were grown at 30⬚C to an OD600 of 0.5 in
LB with antibiotics, and IPTG was added to 0.5 mM. Cells were
harvested by centrifugation 2–3 hr later and resuspended in 3 ml of
S buffer per gram of cells. Following sonication, the lysate was
centrifuged for 20 min at 17,000 ⫻ g, and the supernatant (about
30 ml) was added to 1.5 ml nickel-NTA resin (Qiagen), preequilibrated with S buffer. After mixing for 1 hr at 4⬚C, the resin was
packed into a column and washed with 200 ml S buffer and 100 ml
W20 buffer. Protein was eluted with W500 buffer, and fractions
containing ClpP were pooled and desalted into Q50 buffer. This
material was loaded onto a MonoQ HR5/5 FPLC column (Pharmacia)
and eluted with a 20 ml gradient from Q50 to Q1000 buffer. Fractions
containing ClpP were pooled, dialyzed against Clp buffer at 4⬚C,
and stored at ⫺80⬚C. Concentrations of ClpP14 (⑀280 ⫽ 125160
M⫺1cm⫺1) and ClpX6 (⑀280 ⫽ 98340 M⫺1cm⫺1) were determined by UV
absorbance or dye binding (Bradford, 1976).
GFP-ssrA containing the S65G and S72A mutations (⑀280 ⫽ 23380
M⫺1cm⫺1) was purified from E. coli JB401 cells by organic extraction
(Yakhnin et al., 1998) and chromatography on phenyl sepharose
(Pharmacia). 35S labeling was performed as previously described
(Baker et al., 1993). Fluorescence of GFP-ssrA (excitation maximum,
467 nm; emission maximum, 511 nm) was measured at 30⬚C using
a SPEX FluoroMax2 instrument. Slit widths were adjusted to ensure
that the maximum fluorescence did not exceed 4 ⫻ 106 cps.
A plasmid (a gift from A. Horwich, Yale) expressing the GroEL
D87K variant (⑀280 ⫽ 12200 M⫺1cm⫺1) was transformed into E. coli
strain DH5␣. Cells in LB ⫹ 100 ␮g/ml ampicillin were grown at 37⬚C
to OD650 of 0.9, harvested by centrifugation, resuspended, and lysed
by French press. The lysate was applied to a 50 ml FFQ column
(Pharmacia), equilibrated in 50 mM Tris HCl (pH 7.4), and protein
was eluted with a gradient to 0.5 M NaCl in the same buffer. GroEL
was precipitated with 60% (NH4)2SO4, resuspended and dialyzed
against 50 mM Tris (pH 7.4), 50 mM KCl, and mixed for 24 hr at 4⬚C
with AffiGel Blue (Bio-Rad) using 1 ml of resin per 20 mg protein.
Dynamics of the ClpXP Degradation Cycle
647
GroEL in the unbound fraction was concentrated with Centriprep
30. ␭cI-N-ssrA (⑀280 ⫽ 8940 M⫺1cm⫺1), Arc-ssrA (⑀280 ⫽ 8250
M-1cm-1), and Arc-MuA10 (⑀280 ⫽ 6970 M⫺1cm⫺1) were expressed
and purified as described (Milla et al., 1993; Gottesman et al., 1998).
The peptides (⑀280 ⫽ 1215 M⫺1cm⫺1) described in Results were synthesized by the MIT Biopolymers Facility.
Active Site Modification of ClpP
Purified ClpP (2 mg/ml) was chemically modified in Clp buffer containing 10 mM DFP (Sigma Chemical Co.); the volume of DFP added
was less than 1% of the total reaction volume. Modification was
allowed to proceed for 2–3 hr at 30⬚C and then for an additional 2–3
hr at 4⬚C. The reaction solution was extensively dialyzed against
Clp buffer at 4⬚C, and the DFP-modified ClpP protein was stored
at ⫺80⬚C.
Degradation and Denaturation Assays
Unless noted, all degradation and denaturation assays were performed in PD buffer at 30⬚C. Degradation assays monitored by release of acid-soluble peptides contained 0.3 ␮M ClpX6, 0.8 ␮M
ClpP14 and different concentrations of [35S]GFP-ssrA in a volume of
20 ␮l. The reaction mix without [35S]GFP-ssrA was preincubated for
2 min at 30⬚C, [35S]GFP-ssrA was added, and reactions were stopped
by addition of 10 ␮l 50% trichloroacetic acid and 10 ␮l BSA (10 mg/
ml). After centrifugation for 10 min at 15,000 rpm, the radioactivity
in the supernatant was measured by scintillation counting. Degradation assays monitored by fluorescence were performed in the same
way, except the reaction volume was 100 ␮l, unlabeled GFP-ssrA
was used, and there was a lag (ⵑ20 s) between addition of GFP-ssrA
and monitoring fluorescence. To avoid substantial concentrations of
free ClpX, reactions were generally performed with ClpP in excess
over ClpX. Apparent turnover numbers were normalized to [ClpX6].
Denaturation and degradation experiments were all performed using
GFP-ssrA concentrations between 0.05 and 20 ␮M. Subsaturating
concentrations of GFP-ssrA (e.g., 0.2 ␮M) were commonly used
when substrate release and refolding were assayed, to minimize
the presence of high levels of native, free GFP-ssrA.
Denaturation assays containing ClpX, ClpPDFP, and PD buffer were
preincubated for 3–5 min to allow ClpXPDFP complex formation, GFPssrA was added, and after a mixing lag (ⵑ20 s), fluorescence was
monitored. Denaturation by ClpX alone was preformed as described
above except ClpP was omitted and some reactions contained the
GroEL14 D87K trap (10 ␮M), which was added with ClpX during
preincubation. In some experiments, [35S]ClpPDFP·GFP-ssrA complexes were purified following the denaturation reaction by gel filtration chromatography on a Superdex 200 PC 3.2/30 column run in
Clp buffer at room temperature at a flow rate of 40 ␮l/min.
however, assuming that the initial binding step was effectively irreversible (k2 ⬎⬎ k⫺1) with a value of roughly 1.2 ⫻ 104 M⫺1s⫺1 for the
second-order rate constant k1. Fits allowing K1 and one rate constant
to vary while the other constants were fixed at 100 min⫺1 were
acceptable only when k2 was the slow step with a value of roughly
0.9 min⫺1. When k2 was fixed at this value, other rate constants
could be individually reduced to 10 min⫺1.
Acknowledgments
We wish to dedicate this work to the memory of Paul Sigler. We
thank A. J. Anderson, Steve Bell, Ilana Goldhaber-Gordon, Art Horwich, Jong Myoung Kim, Richard Klemm, Igor Levchenko, Meredith
Seidel, Frank Solomon, Kate Smith, and lab members for strains,
materials, and advice. Supported by the Howard Hughes Medical
Institute (T. A. B.) and NIH grant AI-16892 (R. T. S.). R. E. B. is an
NIH postdoctoral fellow, and Y.-I. K. and T. A. B. are employees of
the Howard Hughes Medical Institute.
Received January 21, 2000; revised March 28, 2000.
References
Andersen, J.B., Sternberg, C., Poulsen, L.K., Bjorn, S.P., Givskov,
M., and Molin, S. (1998). New unstable variants of green fluorescent
protein for studies of transient gene expression in bacteria. Appl.
Environ. Microbiol. 64, 2240–2246.
Baker, T.A., Mizuuchi, M., Savilahti, H., and Mizuuchi, K. (1993).
Division of labor among monomers within the Mu transposase tetramer. Cell 74, 723–733.
Barshop, B.A., Wrenn, R.F., and Frieden, C. (1983). Analysis of numerical methods for computer simulation of kinetic processes: development of KINSIM—a flexible, portable system. Anal. Biochem.
130, 134–145.
Bochtler, M., Hartmann, C., Song, H.K., Bourenkov, G.P., Bartunik,
H.D., and Huber, R. (2000). The structures of HslU and the ATPdependent protease HslU-HslV. Nature 403, 800–805.
Bradford, M.M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254.
Chernoff, Y.O., Lindquist, S.L., Ono, B.I., Inge-Vechtomov, S.G., and
Liebman, S.W. (1995). Role of the chaperone protein Hsp104 in
propagation of the yeast prion-like factor [psi⫹]. Science 268,
880–884.
Cormack, B.P., Valdivia, R.H., and Falkow, S. (1996). FACS optimized
mutants of the green fluorescent protein (GFP). Gene 173, 33–38.
GuHCl Denaturation of GFP-SsrA
Native GFP-ssrA (final concentration, 1 ␮M) was added to buffer
containing 25 mM HEPES-KOH (pH 7.6), 5 mM KCl, 5 mM MgCl2,
0.032% NP-40, 10% glycerol, and 4.5–8 M GuHCl, and denaturation
at 30⬚C was assayed by fluorescence at 511 nm. Kinetic trajectories
were fit to a single exponential to determine rate constants for
denaturation (r ⬎ 0.995).
Fenton, W.A., and Horwich, A.L. (1997). GroEL-mediated protein
folding. Protein Sci. 6, 743–760.
Kinetic Fitting
Kinetic simulations and fitting were performed using the KIMSIM
and FITSIM programs (Barshop et al., 1983; Zimmerle and Frieden,
1989). Combined data sets from degradation and denaturation
assays like those shown in Figure 5 were fit to the model shown
below (where S is native substrate, and D is denatured):
Gonciarz-Swiatek, M., Wawrzynow, A., Um, S.-J., Learn, B.A.,
McMacken, R., Kelley, W.L., Georgopoulos, C., Sliekers, O., and
Zylicz, M. (1999). Recognition, targeting, and hydrolysis of the ␭ O
replication protein by the ClpP/ClpX protease. J. Biol. Chem. 274,
13999–14005.
K1
k2
k3
k4
k5
XP ⫹ S ⇔ S·XP → D·XP → XP·D → XP·frag → XP ⫹ frag.
The assumptions used in these calculations were that the loss of
GFP fluorescence depends on the concentrations of four nonnative
species ([D·XP] ⫹ [XP·D] ⫹ [XP·frag] ⫹ [frag]), that denatured GFPssrA within ClpX is not fluorescent, and that the appearance of
cleavage products depends either on the concentration of one species ([frag]) or two species ([XP·frag] ⫹ [frag]). For simplicity, the
initial binding of GFP-ssrA to the enzyme was assumed to reach a
rapid preequilibrium (k⫺1 ⬎⬎ k2). The data could be fit equally well,
Fenton, W.A., Kashl, Y., Furtak, K., and Horwich, A.L. (1994). Residues in chaperonin GroEL required for polypeptide binding and
release. Nature 371, 614–619.
Glover, J.R., and Lindquist, S. (1998). Hsp104, Hsp70, and Hsp40:
a novel chaperone system that rescues previously aggregated proteins. Cell 94, 73–82.
Gottesman, S., Maurizi, M.R., and Wickner, S. (1997). Regulatory
subunits of energy-dependent proteases. Cell 91, 435–438.
Gottesman, S., Roche, E., Zhou, Y., and Sauer, R.T. (1998). The
ClpXP and ClpAP proteases degrade proteins with carboxy-terminal
peptide tails added by the ssrA-tagging system. Genes Dev. 12,
1338–1347.
Grimaud, R., Kessel, M., Beuron, F., Steven, A.C., and Maurizi, M.R.
(1998). Enzymatic and structural similarities between the Escherichia
coli ATP-dependent proteases, ClpXP and ClpAP. J. Biol. Chem.
273, 12476–12481.
Hoskins, J.R., Pak, M., Maurizi, M.R., and Wickner, S. (1998). The
Molecular Cell
648
role of the ClpA chaperone in proteolysis by ClpAP. Proc. Natl. Acad.
Sci. USA 95, 12135–12140.
to the eukaryotic proteasome. Proc. Natl. Acad. Sci. USA 93, 5808–
5813.
Huang, G.S., and Oas, T.G. (1995). Submillisecond folding of monomeric lambda repressor. Proc. Natl. Acad. Sci. USA 92, 6878–6882.
Rohrwild, M., Pfeifer, G., Santarius, U., Muller, S.A., Huang, H.C.,
Engel, A., Baumeister, W., and Goldberg, A.L. (1997). The ATPdependent HslVU protease from Escherichia coli is a four-ring structure resembling the proteasome. Nat. Struct. Biol. 4, 133–139.
Hwang, B.J., Woo, K.M., Goldberg, A.L., and Chung, C.H. (1988).
Protease Ti, a new ATP-dependent protease in Escherichia coli,
contains protein-activated ATPase and proteolytic functions in distinct subunits. J. Biol. Chem. 263, 8727–8734.
Jenal, U., and Fuchs, T. (1998). An essential protease involved in
bacterial cell-cycle control. EMBO J. 17, 5658–5669.
Keiler, K.C., Waller, P.R.H., and Sauer, R.T. (1996). Role of a peptide
tagging system in degradation of proteins synthesized from damaged messenger RNA. Science 271, 990–993.
Kessel, M., Maurizi, M.R., Kim, B., Kocsis, E., Trus, B.L., Singh,
S.K., and Steven, A.C. (1995). Homology in structural organization
between E. coli ClpAP protease and the eukaryotic 26 S proteasome.
J. Mol. Biol. 250, 587–594.
Kruklitis, R., Welty, D.J., and Nakai, H. (1996). ClpX protein of Escherichia coli activates bacteriophage Mu transposase in the strand
transfer complex for initiation of Mu DNA synthesis. EMBO J. 15,
935–944.
Levchenko, I., Luo, L., and Baker, T.A. (1995). Disassembly of the
Mu transposase tetramer by the ClpX chaperone. Genes Dev. 9,
2399–2408.
Levchenko, I., Yamauchi, M., and Baker, T.A. (1997a). ClpX and MuB
interact with overlapping regions of Mu transposase: implications
for control of the transposition pathway. Genes Dev. 11, 1561–1572.
Levchenko, I., Smith, C.K., Walsh, N.P., Sauer, R.T., and Baker, T.A.
(1997b). PDZ-like domains mediate binding specificity in the Clp/
Hsp100 family of chaperones and protease regulatory subunits. Cell
91, 939–947.
Maurizi, M.R., Clark, W.P., Kim, S.-H., and Gottesman, S. (1990).
ClpP represents a unique family of serine proteases. J. Biol. Chem.
265, 12546–12552.
Maurizi, M.R., Thompson, M.W., Singh, S.K., and Kim, S.-H. (1994).
Endopeptidase Clp: ATP-dependent Clp protease from Escherichia
coli. Meth. Enzymol. 244, 315–331.
Maurizi, M.R., Singh, S.K., Thompson, M.W., Kessel, M., and Ginsburg, A. (1998). Molecular properties of ClpAP protease of Escherichia coli: ATP-dependent association of ClpA and ClpP. Biochemistry 37, 7778–7786.
Milla, M.E., and Sauer, R.T. (1994). P22 Arc repressor: folding kinetics of a single-domain dimeric protein. Biochemistry 33, 1125–1133.
Milla, M.E., Brown, B.M., and Sauer, R.T. (1993). P22 Arc repressor:
enhanced expression of unstable mutants by addition of polar C-terminal sequences. Protein Sci. 2, 2198–2205.
Newnam, G.P., Wegrzyn, R.D., Lindquist, S.L., and Chernoff, Y.O.
(1999). Antagonistic interactions between yeast chaperones Hsp104
and Hsp70 in prion curing. Mol. Cell. Biol. 19, 1325–1333.
Pak, M., and Wickner, S. (1997). Mechanism of protein remodeling
by ClpA chaperone. Proc. Natl. Acad. Sci. USA 94, 4901–4906.
Pak, M., Hoskins, J.R., Singh, S.K., Maurizi, M.R., and Wickner, S.
(1999). Concurrent chaperone and protease activities of ClpAP and
the requirement for the N-terminal ClpA ATP binding site for chaperone activity. J. Biol. Chem. 274, 19316–19322.
Palm, G.J., Zdanov, A., Gaitanaris, G.A., Stauber, R., Pavlakis, G.N.,
and Wlodawer, A. (1997). The structural basis for spectral variations
in green fluorescent protein. Nat. Struct. Biol. 4, 361–365.
Parsell, D.A., Silber, K.R., and Sauer, R.T. (1990). Carboxy-terminal
determinants of intracellular protein degradation. Genes Dev. 4,
277–286.
Parsell, D.A., Kowal, A.S., and Lindquist, S. (1994). Saccharomyces
cerevisiae Hsp104 protein. Purification and characterization of ATPinduced structural changes. J. Biol. Chem. 269, 4480–4487.
Roche, E.D., and Sauer, R.T. (1999). SsrA-mediated peptide tagging
caused by rare codons and tRNA scarcity. EMBO J. 18, 4579–4589.
Rohrwild, M., Coux, O., Huang, H.-C., Moerschell, R.P., Yoo, S.J.,
Seol, J.H., Chung, C.H., and Goldberg, A.L. (1996). HslV-HslU: a
novel ATP-dependent protease complex in Escherichia coli related
Rüdiger, S., Buchberger, A., and Bukau, B. (1997). Interaction of
Hsp70 chaperones with substrates. Nat. Struct. Biol. 4, 342–349.
Schirmer, E.C., Glover, J.R., Singer, M.A., and Lindquist, S. (1996).
HSP100/Clp proteins: a common mechanism explains diverse functions. Trends Biochem. Sci. 21, 289–296.
Schweder, T., Lee, K.-H., Lomovskaya, O., and Matin, A. (1996).
Regulation of Escherichia coli starvation sigma factor (ss). J. Bacteriol. 178, 470–476.
Smith, C.K., Baker, T.A., and Sauer, R.T. (1999). Lon and Clp family
proteases and chaperones share homologous substrate-recognition domains. Proc. Natl. Acad. Sci. USA 96, 6678–6682.
Squires, C., and Squires, C.L. (1992). The Clp proteins: proteolysis
regulators or molecular chaperones? J. Bacteriol. 174, 1081–1085.
Thompson, M.W., and Maurizi, M.R. (1994). Activity and specificity
of Escherichia coli ClpAP protease in cleaving model peptide substrates. J. Biol. Chem. 269, 18201–18208.
Thompson, M.W., Singh, S.K., and Maurizi, M.R. (1994). Processive
degradation of proteins by the ATP-dependent Clp protease from
Escherichia coli. Requirement for the multiple array of active sites
in ClpP but not ATP hydrolysis. J. Biol. Chem. 269, 18209–18215.
Tu, G.F., Reid, G.E., Moritz, R.L., and Simpson, R.J. (1995). C-terminal extension of truncated recombinant proteins in Escherichia coli
with a 10Sa RNA decapeptide. J. Biol. Chem. 270, 9322–9326.
Wang, J., Hartling, J.A., and Flanagan, J.M. (1997). The structure
of ClpP at 2.3 Å resolution suggests a model for ATP-dependent
proteolysis. Cell 91, 447–456.
Weber-Ban, E.U., Reid, B.G., Miranker, A.D., and Horwich, A.L.
(1999). Global unfolding of a substrate protein by the Hsp100 chaperone ClpA. Nature 401, 90–93.
Wojtkowiak, D., Georgopoulos, C., and Zylicz, M. (1993). Isolation
and characterization of ClpX, a new ATP-dependent specificity component of the Clp protease of Escherichia coli. J. Biol. Chem. 268,
22609–22617.
Yakhnin, A.V., Vinokurov, L.M., Surin, A.K., and Alakhov, Y.B. (1998).
Green fluorescent protein purification by organic extraction. Protein
Expr. Purif. 14, 382–386.
Zimmerle, C.T., and Frieden, C. (1989). Analysis of progress curves
by simulations generated by numerical integration. Biochem. J. 258,
381–387.