Content uploaded by Alfred L Goldberg
Author content
All content in this area was uploaded by Alfred L Goldberg on Jul 24, 2015
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
THE JOURNAL
OF
BIOLOGICAL
CHEMISTRY
(0
1992
by
The American Society for Biochemistry and Molecular Biology, Inc.
Vol.
261,
No.
28,
Issue
of
October
5,
pp.
20429-20434,1992
Printed
in
U.
S.A.
The Heat-shock Protein
ClpB
in
Escherichia coli
Is
a
Protein-activated ATPase*
(Received for publication, January 27, 1992)
Kee Min
Woos,
Keun I1 Rims, Alfred
L.
Goldberg& Doo Bong Has, and Chin Ha Chungsll
From the $Department
of
Molecular Biology, College
of
Natural Sciences, Seoul National Uniuersity, Seoul
151-742,
Korea and
the §Department
of
Molecular and Cellular Physiology, Haruard Medical School, Boston, Massachusetts
02115
The
clpB
gene in
Escherichia coli
encodes
a
heat-
shock protein that is
a
close homolog
of
the
clpA
gene
product. The latter
is
the ATPase subunit of the mul-
timeric ATP-dependent protease Ti (Clp) in
E.
coli,
which also contains the 21-kDa proteolytic subunit
(ClPP).
The
clpB
gene product has been purified to near
homogeneity by DEAE-Sepharose and heparin-aga-
rose column chromatographies. The purified ClpB con-
sists of a major 93-kDa protein and a minor 79-kDa
polypeptide
as
analyzed by polyacrylamide gel electro-
phoresis in the presence of sodium dodecyl sulfate.
Upon gel filtration on
a
Superose-6 column, it behaves
as
a
350-kDa protein. Thus, ClpB appears to be a
tetrameric complex of the 93-kDa subunit.
The purified ClpB has ATPase activity which
is
stim-
ulated S-10-fold by casein. It
is
also activated by in-
sulin, but not by other proteins, including globin and
denatured bovine serum albumin. ClpB cleaves aden-
osine 5’-(cr,&methylene)-triphosphate as rapidly
as
ATP, but not adenosine
5’-@,y-methylene)-triphos-
phate. GTP, CTP, and UTP
are
hydrolyzed 15-25%
as
well
as
ATP. ADP strongly inhibits ATP hydrolysis
with a
Ki
of 34
PM.
ClpB has a
K,
for ATP of
1.1
mM,
and casein increases
its
V,,,
for ATP without affecting
its
K,.
A Mg2+ concentration of
3
mM
is
necessary for
half-maximal ATP hydrolysis. Mn2+ supports ATPase
activity
as
well
as
Mg2+, and Ca2+ has about 20% their
activity. Anti-ClpB antiserum does not cross-react
with ClpA nor does anti-ClpA antiserum react with
ClpB. In addition, ClpB cannot replace ClpA in sup-
porting the casein-degrading activity of ClpP. Thus,
ClpB
is
distinct from ClpA in its structural and bio-
chemical properties despite the similarities in their
sequences.
Escherichia coli contain at least two distinct endoproteases,
proteases La and
Ti,
that require
ATP
and
Mg2+
for activity.
The first ATP-requiring protease discovered was protease La,
the
lon
gene product (1,
2).
This enzyme is a heat-shock
protein (3, 4) and plays an essential role in the degradation
of most abnormal proteins (1-6) and certain short-lived reg-
ulatory proteins (6). It is composed of four identical subunits,
each
(87
kDa) of which contains a site for ATP hydrolysis
(7-
9).
One important feature of this enzyme is that protein
*This work was supported by Korea Science and Engineering
Foundation through Research Center for Cell Differentiation and
The Ministry of Education. The costs of publication of this article
were defrayed in part by the payment of page charges. This article
must therefore be hereby marked “aduertisement” in accordance with
18
U.S.C. Section 1734 solely to indicate this fact.
ll
To whom correspondence should be addressed.
substrates stimulate ATP-hydrolyzing activity of protease La,
and this process is linked stoichiometrically to proteolysis (7).
The second ATP-requiring enzyme, protease Ti (Clp), con-
sists of two different polypeptide subunits, both of which are
required for proteolysis (10-13). When isolated, component
A
(ClpA) behaves as a dimeric complex of 81-kDa subunits and
contains the ATP-hydrolyzing site. Component
P
(ClpP),
which also is a heat-shock protein (14), is a multimer of
21-
kDa subunits and contains the proteolytic site sensitive to
diisopropylfluorophosphate.
The isolated ClpA also shows
protein-activated ATPase activity, which in the reconstituted
enzyme is linked to protein breakdown
(12).
Recently, ClpA has been shown to be a member of a family
of highly conserved proteins with no previously described
function
(15,
16). Members of the family include a second
E.
coli gene called clpB and genes from other bacteria, trypano-
somes, and plants. The
DNA
sequences of this family have
two regions of particularly high homology, each of which
contains a consensus sequence for an adenine nucleotide
binding site. Squires and co-workers (17,
18)
have recently
shown that the clpB gene encodes an 84.1-kDa protein (ClpB)
as well as an additional protein of 68.5 kDa, both of which
are heat-shock proteins. The relationship between these two
products of the clpB gene is unclear
(18).
Because of the
sequence similarity of ClpB with that of ClpA, particularly of
the adenine nucleotide binding motifs (16), it has been sug-
gested that ClpB may play an important role in ATP-depend-
ent proteolytic processes in
E.
coli (17,
18).
Since both ClpP
and ClpB are heat-inducible proteins and under the control
of the htpR (rpoH) gene product (14,17,18), it has also been
proposed that a model ATP-dependent protease, a ClpB-ClpP
complex, may exist and play an essential role in protection
against thermal stress in
E.
coli (17,
18).
In order to elucidate
the function of ClpB, we purified the ClpB protein from an
E.
coli strain that contains the clpB gene in a multicopy
plasmid. In the present study, we show that the purified ClpB,
like ClpA, exhibits a protein-stimulated ATP-hydrolyzing
activity and yet cannot replace ClpA in supporting the ATP-
dependent degradation of proteins by ClpP.
EXPERIMENTAL PROCEDURES
Materiak-E. coli strain, MC1009/pClpB (18), was kindlyprovided
by Dr. Catherine Squires (Columbia University). The cells were grown
at 35
“C
in Luria broth until to reach an optical density
(600
nm)
of
1.0 and shifted to
45
“C. After culturing for
2
h at this temperature,
the cells were harvested and stored at -70
“C
until
use.
The clpA and
3H]Casein was prepared as described by Jentoft and Dearborn (19).
clpP
gene products were purified as described previously (12). [methyl-
Globin and denatured bovine serum albumin (BSA)’ were prepared
The abbreviations used are: BSA, bovine serum albumin; DTT,
dithiothreitol; SDS, sodium dodecyl sulfate; MES, 2-(N-morpho-
1ino)ethanesulfonic acid; MOPS,
3-(N-morpholino)propanesulfonic
acid; FPLC, fast protein liquid chromatography; Ap(CH,)pp, adeno-
sine
5’-(cu,@-methylene)-triphosphate;
App(CHa)p, adenosine
5’-(@,y-
methylene)-triphosphate.
20429
20430
ClpR
Is
a Prokin-activated ATPase
:IS
rlescril)ed previously
(20).
DEAE-Sepharose, Superose-6, and
Mono
Q
were purchased from Pharmacia; heparin-agarose from
Re-
t
hesda
Research I,ahoratories; and ["H]formaldehvde from
Ih
I'ont-
New
England Nuclear.
All
other chemicals were ohtained from Sigma.
Assny.s"ATPase was assayed
by
incuhating the reaction mixtures
(0.1
ml) containing appropriate amounts ofthe
ClpH
fractions,
5
mM
ATP,
100
mM 'I'ris-HCI
(pH
8),
10
mM
MgCI,,
1
mM dithiothreitol
(DTT), and
1
mM EIITA
(12).
ATP was titrated to pH
'i.5
with
0.1
N
NaOH prior
to
the incuhation. After the incuhation
at
37
"C
for
30
min, the reaction
was
terminated hv adding
0.2
ml
of
lob
(w/v)
sodium
dodecyl
sulfate
(SDS).
The phosphate released was then
determined
as
tlescri1)e.d
hv
Ames
(21).
To
assay
ATP
hydrolysis at
different pH values, the following huffers were used: sodium acetate
(pH
4.5);
MES
(pH
5.5);
MOPS,
(pH
6.5); Tris-HCI, (pH
75
and
X.5);
glycine HCI, (pH
9.5
and
10.5).
Hvdrolvsis of radioactive proteins
and lluorogenic peptides were determined
as
descrihed previouslv
(12.
22).
Proteins were
assnved
hv following the method of
Hradford
using
IiSA
as
a stnndnrd
(23).
I'rrpnrotion
of
('rudr
Extracts
of
E.
coli-The frozen
E.
coli
cells
(20
g)
were thawed tlnd resuspended in
50
ml
of
20
mM Tris-HCI (pH
X)
containing
X0
mM
NaCI.
5
mM
MgCI,,
1
mM DTT,
0.5
mM
EDTA,
and
20";'
(v/v)
glycerol.
The cells were then disrupted with a French
press
at
14,000
p.s.i. and centrifuged at
100,000
X
F:
for
3
h. The
supernatant
WHS
dialyzed against the same huffer and referred to
as
crude
ext ract.
Immunochrmicnl
Annlysis-The
ClpA
and
ClpR
preparations oh-
tained from the linal purification steps
(Ref. 12
and see helow) were
rlectrophoresed in
10";
(w/v) polvacrvlamide
gels
in the presence
of
SI)S
(24).
After staining the
gels
briefly with ('oomassie Hlrle
R-250,
the protein hands of X1-kDn
('IpA
and 93-kl)a
ClpH
were
cut
out and
crushed. The samples were then suhcutaneously injected into alhino
rat)l)its
for
the prnhct ion of ant ihodies.
For immunohlot analvsis. proteins were electrophoresed as
ahove
and transferred to nitrocellulose memlmnes. The membranes were
incuhated with nnti-ClpA
or
anti-ClpH antiserum and then with anti-
ralhit
IgG
conjugated with alkaline phosphatase. Protein hands were
visualized
hv
the phosphatase reaction
(25).
Amino
Acid
S(*c/ucnw
and
Composition
Andy.&-The purified
ClpR
preparation
was
rlectrophoresed in the presence of
SDS,
and
the
93-
and 'i9-kt)a protein hands were transferred
to
polvvinvlidene
dillrroride memhranes. Ihch pol?lpeptide was then suhjected to Edrnan
degradation for thr determination
of
its N-terminal amino acid
se-
quence using I'roSequencer (model 6600; MilliGen).
For
analvsis of
amino acid composition, each of the protein hands
was
excised
out,
minced, and extracted in
0.If'A
SIX.
The protein samples were
hvtlrolyzed in
6
N
HCI
at
110
*C
for
24
h
in
LWCUO.
The dried
hvdrolysates
were
tlerivatized with phenylisothincvanate and
sepa-
rated on
a
I'iroTag column (Waters) according
to
the operator's
manual.
RESULTS
Purification
of
ClpR-The crude extract
(2.0
g)
obtained
from
20
g
of heat-shocked
E.
coli
MC1009/pClpB was loaded
on
a
DEAE-Sepharose column
(2.5
x
20
cm) equilibrated with
20
mM Tris-HCI (pH 7.8) buffer containing
80
mM NaCI,
5
mM MgCI?,
1
mM DTT,
0.5
mM EDTA, and
20%
glycerol.
After washing extensively the column with this buffer, and
proteins were eluted with a linear gradient of
80-200
mM
NaCI. Most of the
93
kDa-protein (ClpR), which was identi-
fied
by
SDS-polyacrylamide gel electrophoresis (Fig.
lA),
was
eluted at the salt concentration of
160-180
mM.
Since ClpR
is
homologous to ClpA
(Ifi),
which
is
a protein-
activated ATPase
(12),
we tested whether the fractions con-
taining ClpR may also exhibit an ATPase activity that can
be
stimulated
by
proteins, such as casein. As shown in Fig.
lR,
a
distinct peak of casein-stimulated ATPase was evident.
in the fractions where the peak of ClpR protein appeared.
Much
less
of this ATPase act.ivity
or
of the ClpR protein was
found in these DEAE-fractions, when the extracts were pre-
pared from these cells hut, without, heat treatment (data not
shown). Thus,
it
appears likely that ClpR
is
responsible for
the casein-act,ivated ATPase activity. This result also indi-
cates that the ATPase activit.y
is
not due
to
any ClpA that
5:
w
c
0
E
a
1.0
0.8
0.8
0.4
0.2
0.0
-
66
-
45
-
29
I
1.0
IB
t
I
I
,
.\
-.
0.2
00
0
30
80
00
120
150
FRACTION
NUMBER
FIG.
1.
Isolation
of
ClpR
by
DEAE-Sepharone chromatog-
raphy.
The
crude
extract
of
b,'.
coli
\lClOO9/p('lpl3
was
suhirctrrl
to
Dl<AIMepharose chromatography, and Ill-ml tractinns
wrrc
col-
lected at
a
llow rate of
100
ml/h.
A,
aliqunts of
the
fractions were
analyzed
hv
polvarrylamide
pel
electrophoresis in
IO';
slnh
gels
containing
SDS.
The
gel
was
then stninrd with Coornnwie
I3lr1r
It-
250.
The
nrrotr'
indicates the hand mrrespondinc to thr 9:t-kDn
c'lpl3.
H,
the fractions were
also
assayed
for their ahility to hydrnlyze
A'l'l'
hv
incuhating with
:i
m%r
A'TI'
in the presence
(01
and ahrenrP
(Q)
of'
10
pg
of
casein. Elution protile
of
proteins
1
. . .
.)
was
drtermined
hy follnwing the method
ol'
Hratlfnrtl
(231.
'The slnshrd
/in(,
indicntrs
the salt gradient.
might have heen contaminated in the ClpR fractions, since
ClpA
is
not a heat-shock protein
(13,
15).
The fractions
(108-126
in Fig.
1R)
containing ClpR were
pooled, dialyzed against
20
mM 'I'ris-HCI (pH
8)
containing
5
mM MgCL,
100
mM NaCI,
1
mM DTT,
0.5
mM EDTA, and
20%
glycerol, and loaded on a heparin-agarose column
(1.5
x
13
cm) equilibrated with this buffer. After washing the col-
umn, proteins were eluted with a linear gradient of
100-400
mM NaC1. Fig.
2
again shows that the elution profile of
ClpH
protein overlaps closely with that of the casein-activated
ATPase. Thus,
ClpR
is
responsihle for the ATPase activity.
However, a minor protein hand with 79 kDa
WAS
evidently
present in all fractions containing ClpR (Fig.
2R).
Moreover,
the 79-kDa protein was always co-eluted with the 93-kDa
protein upon further purification of ClpR using Superose-6
and Mono
Q
column (data not shown). Thus, the 79-kDa
protein must share many physical properties with ClpR
or
he
part of the same multimeric complex. The fractions with
highest activity from the heparin-agarose column were pooled
and kept frozen at -70
"C.
Summary of the purification of
ClpR
is
shown in Tahle
I.
Physicoch~rnical
PropPrtim
of
CIpR-To determine the size
of ClpR under nondenaturing condition. the purified protein
was suhjected to analvsis
by
FPLC on a Superose-6 column.
ClpR was eluted
as
a
symmet,ric peak in the fractions corre-
sponding to an apparent molecular weight
of
350,000
(data
not shown). This result suggests that
ClpR
under nondena-
ClpR
Is
a
Protein-activated ATPase
20431
0.4
0.3
-
0
0
v
2
2
0.2
W
c
0
a
a
0.1
0.0
A
.\
,*
*..
0
,
0
0
0
20
40
eo
FRACTION
NUMBER
21
27
33
39
45
st
1.2
-
0
0.9
;
v
m
0.6
5
m
0
m
a
*
0.3
a
X
c
4
0.0
kDa
-
116
-
97
66
45
29
-
-
0.4
h
v
a
G
-
0.2
g
-
0.0
FIG.
2.
Elution profile of ClpR from heparin-agarose
col-
umn.
Proteins ohtained from DEAR-Sepharose column were suh-
jected
to
heparin-agarose chromatography. Fractions
of
3
ml
were
collected at a flow rate of
20
ml/h.
A,
aliquots
of
them were assayed
l'or proteins
(.
..
.) and for ATP hydrolysis in the presence
(0)
and
nhsence
of
(0)
of
10
pg
of
casein.
H,
they were also suhjected to
~mlyarrylamide gel electrophoresis in the presence
of
SDS. The
slnshrd
/inc
indicates the salt gradient.
TABLE
I
Summnry
of
thp
purification of
ClpR
Assays were performed
as
in Fig.
1
in the presence and ahsence of
0.1
mg/ml
of
casein. The specific activity
or
total
units
of
ClpR
could
not
he estimated prior
to
the DEAF,-Sepharose step due
to
the
presence
of
manv other ATP-cleaving activities.
Total activity Specific activity
-Casein +Casein -Casein +Casein
Steps Protein Yield"
mp
unit?
unitslmp
Oh
Crude extract
1,960
DEAR-Sepharose
143 386 2,874 2.7 20.1
100
Heparin-agarose
71
227 1,974 3.2 27.8
69
'I
Determined in the presence
of
casein only.
One unit
is
defined
as
1
pmol
of
ATP
hydrolyzed per h.
turing conditions may behave as a tetrameric complex of the
93-kDa subunit.
To
clarify further the structural organization of ClpB, the
purified protein was electrophoresed in duplicates in 7% poly-
acrylamide gels under nondenaturing condition (26). After
the electrophoresis at
4
"C at
pH
8.9, one of the gels was
stained wit,h Coomassie Blue R-250. The other gel was cut
into 2-mm pieces, and the proteins in these pieces were
extracted overnight in 100 mM Tris-HCI (pH
8)
buffer con-
taining 10 mM MgC12,
1
mM DTT, and
0.5
mM EDTA.
Aliquots of the extracts were then assayed for
ATP
hydrolysis.
As shown in Fig.
3A,
a single band with a relative mobility of
0.38 was evident in the Coomassie Blue-stained gel and, at
the same position, a peak of ATPase activity was
also
found.
Because the purified ClpB preparation consisted of a major
93-kDa protein and low amounts (5-10%) of 79-kDa polypep-
tide, we analyzed the protein composition of the band by
0.0
0.2
0.4
0.0
0.0
1.0
RELATIVE
YOBILITY
FIG.
3.
Analysis
of
ClpR on a nondenaturing
polyacryl-
amide slab
gel.
The
ClpH
preparation
(30
pc~ ohtalned from the
heparin-agarose column was elertrophoresed
on
T,'
slat) gels in the
ahsence of
SDS.
After the electrophoresis, the
gels
were stained with
Coomassie
Hlue
R-2550
or cut into
2-mm
pieces and extracted
as
descrihed in the
text.
Aliquots
of
the extracts were then
assayed
fnr
ATP hydrolysis
(A).
The protein hand seen in the stained
pcI
was
cut out and electrophoresed again on
a
IO?
slah
gel
rontnininx
SI)S
(H).
The resulting gel was stained
as
above
(lnnr,
0).
The
pllrifird
ClpR
was
also
run
as
a
marker
(Innv
h).
The arrow indiratc whrare
the
93-
and
79-kDa
protein hands migrate. The intensity
of
the upper
hands was ahout
14-fold
greater than that of the lower tmnd.;
$IS
analyzed by densitometric scanning.
cutting out from the nondenaturing gel and running again on
a 10% polyacrylamide slab gel containing SDS. Fig.
3fj
shows
that the single band from the nondenaturing gel consists
of
both the 93-kDa ClpR and 79-kDa polypeptide. These results
suggest that the ClpR may form homo-tetramers of the
93-
kDa proteins and/or hetero-oligomeric complexes toget her
with the 79-kDa polypeptide.
To determine the N-terminal amino acid sequences of the
93- and 79-kDa pol.ypeptides, they were electrophoresed
in
the presence of SDS, transferred to polwinvlidene difluoride
membrane, and subjected to Edman degradation. The
13
N-
terminal amino acid sequence of the 9:bkDa protein was
MRLDRLTNKFQLA, and this was identical to the N-ter-
minal amino acid sequence deduced from the nucleotide se-
quence of the clpR gene
(16,
18,
27,
28). On the other hnnd,
t.he
8
N-terminal amino acid sequence of the 79-kDn polypep-
tide was MNDQGAED. In accord with the earlier finding
with
R:,
protein
(28),
t,his sequence nearly matches with
VNDQGAED, which
is
the deduced internal amino acid se-
quence of ClpR (from
149
to
1.56).
Furthermore, the amino
acid composition of each polypeptide, determined experimen-
tally after acid hydrolysis, was in
good
agreement
with
that
predicted from the reading frame (Table
11).
These results
indicate that the 79-kDa pol-ypeptide
is
in the same reading
frame as the 93-kDa ClpR, as was suggested by Squires and
co-workers
(18).
Immunochemical
Ana/.vsis
of
CIpR-Recently,
it
has been
reported that ClpA and ClpR have more than
60"6
similar
amino acids over the full length of the proteins
(16).
There-
fore, we examined the cross-reactivity of the proteins with
the antibodies raised against ClpA and ClpR. The proteins
were electrophoresed in the presence
of
SDS,
transferred to
nitrocellulose membranes, and subjected
to
immunohlot
analysis. As shown in Fig.
4,
ClpA did not interact with the
anti-ClpR antiserum (Ianr
r)
nor did ClpR react with the
anti-ClpA antiserum (lane
d).
In addition, the sizes of ClpA
(81
kDa) and ClpR
(93
kDa) were clearly different. These
results indicate that ClpA and ClpR are at least immunolog-
ically distinct proteins, despite the similarities in their amino
20432
ClpR
Is
a Protein-activated ATPase
Ala
Arg
Asp/Asn
112Cys
Clu/Gln
GlV
His
Ile
Leu
I,vs
Met
I'
he
I'ro
Thr
Ser
mol
G
9.2
10.1
1.1
6.9
9.6
9.5
0.3
0.5
1.53
15.4
I
..1
8.3
1
.<I
1.8
6.4
6.1
1
1.6
11.2
5.0
4.7
2.4 2.4
2.2
9.4
3.2
3.9
4.5
3.9
3.7
3.5
"
I.
mol
I
8.7
8.1
8.9
0.2
15.9
7.0
2.1
6.7
11.0
*5.6
2.5
2.1
3.3
4.0
3.3
TABLE
I1
AminfJ
ncid
COmpfJ.SitifJf1 /Jf
('/pH
93-kI)n
protein
79-kDa
protein
Ikdurrd"
I)eterminedh
Deduced
Determined
Amino acid
9.7
7.2
8.9
0.3
15.9
7.5
1.9
10.8
6.5
5.4
2.7
3.3
4.0
3.6
Trp
3.3
0.2
ND'
0.2
ND
Tyr
1
.<I
2.4
2.3
Val
6.7
5.9
7.0
6.5
2.5
''
Deduced
from
nucleotidc sequcwce.
"
Ikterminetl
hv
amino acid composition analysis.
'
Not determined.
Ma
b
cd
ef
kDa
116-
97
-
66
-
-
-
-
45-
-
29-
FIG.
4.
Immunochemical analysis of
ClpA
and
ClpR.
The
purified
CIpA
and
(Ypl3
(4
pg
each) were electrophoresed in
10%
slah
gels containing
SDS.
The proteins in the
gels
were
stained with
(homassie
Hlue
I{-2.50
(loncs
a
and
h)
or
suhjected to immunohlot
nnalysis
(r-f).
Lnnrs
a,
c,
and
c,
ClpA;
lnnrs
h.
d.
and
f,
ClpR.
Lone
M
shows the size markers: +galactosidase
(M,
116,000).
phosphoryl-
ase
b
(M,
97,000),
I3SA
(M,
(ifi,OOO),
ovalhumin
(M,
4.5,000),
and
rarhonic anhydrase
(M,
29.0001
from top to hottom.
acid sequences. On the other hand, the antibody raised against
the 93-kDa
ClpR
also
interacted with the 79-kDa polypeptide.
These findings are consistent with the 79-kDa protein being
a
fragment of the 93-kDa ClpR,
as
suggested by the DNA
sequence
(18).
Protein-activated
ATPase
Associated
with ClpR-To char-
acterize further the casein-stimulated ATPase activity of
ClpR,
the purified protein was incubated for various periods
in the presence and ahsence of casein. As shown in Fig.
5.4,
casein stimulated t.he ATPase activity by about 5-fold. Fur-
thermore, the magnitude of this stimulation rose with increas-
ing concentrations of casein, until
a
maximal effect was
reached (Fig.
Fill).
Noteworthy
is
the finding that the increase
in ATP hydrolysis by ClpR
is
not linear with time hut
is
rather curvilinear. However,
it
is
at least certain this effect of
incubation period
is
not due to
a
time-dependent inactivation
of
ClpR,
because
ClpR
is
completely stable for
at
least several
hours under the incubation conditions tested in the presence
and absence of casein (data not shown). We then tested
whether other pol-ypeptides are also capable of stimulating
the ATP hydrolysis. While RSA, denatured RSA, globin, and
y-globulin showed little
or
no effect, insulin also stimulated
0
20
40
EO
EO
0.1
0.2
0.3
0.4
TIME
(mln)
CASEIN
(mg/ml)
FIG.
.5.
Effect of casein on
the
ATPase activity of
ClpR.
I\.
ATP
hydrolvsis was
assay~d
t)y incubating
3
pg
of
('lpli
and
>
m\f
ATP
in the presence
(0,
and ahsence
IC))
of
10
pg
of
casein.
H.
the
ATPase
artivitv was
alw
measured with inrreasing ronrent rat ions
of
casein. Incuhations wt're
at
:17
"C
for
:X0
min.
'rAH1.E
111
Kffrcls
of
tarious prolrins on
7'1'
hydrfJ/y~is
hy
ATPase
activity
of
ClpR
was measured hy incuhating
:I
pg
of
the
purified protein at
37
"C
for
RO
min in the prescnce
and
~Ihwnce
of
10
pg
of
the listed proteins. The acti\'ity
seen
in the
al~sencr
of
nnv
added pol-ypept ide
WRS
expressed
as
100";.
Polvpepcide
added
l',
relensed
None
Casein
Insulin
y-Glohulin
Lysozyme
HSA
Denatured
RSA
Glohin
20
-
!-
16
ct
e
1
O
10
v
Ei
>
0
-1.0
-0.5
0.0
0.5
1.0
1/s
(mu)"
FIG.
6.
Effect
of
ADP
on the casein-activated ATPase activ-
ity of
ClpR.
Reaction mixtures contained
:X
pg
of'
('lpli,
Io
pg
of
casein, and increasing amounts
of
ATI'.
Assays
were performed
nt
87
"C
for
30
min in the presence
(0)
and
ahsence
10,
of
0.5
m\f
A1)I'.
The resultingdatn were then used
In
estimation
of
kinetic parameters
hv
a
douhle-reciprocal plot.
the ATPase activity, although the extent of the stimulation
was less than that shown by casein (Table
111).
Kinetic Parameters-To define more precisely the stimu-
latory effect of casein, the rates of
ATP
hydrolysis
by
ClpR
were determined in the presence of increasing concentrations
of ATP. Using double-reciprocal plots of the data, the
K,
for
ATP was estimated to be
1.1
mM (Fig.
6).
This value did not
change significantly when determined in the absence of
CAS-
ein. However, the protein increased the
V,,,",
of
ClpR
by
5-
10-fold (data not shown). Addition of
ClpP
to
ClptZ
alters
its
protein-act.ivated ATPase activitv
(12).
In contrast,
ClpJ'
did
ClpB
Is
a
Protein-activated
ATPase
20433
not change the kinetic parameters of ClpB.
ADP has been reported to inhibit the casein-activated
ATPase activity of protease La
(7).
ADP also effectively
inhibited the ATPase activity of ClpA (data not shown). To
examine if ADP is also capable of inhibiting the casein-
activated ATPase activity of ClpB, the rates of ATP hydrol-
ysis were determined as above but in the presence of ADP.
Fig. 6 shows that ADP shows relatively little effect on
Vmax
but markedly increases the K,,, of ATP. Half-maximal inhi-
bition of ClpB by ADP (i.e.
Ki)
occurred at 34
WM,
indicating
that the affinity of ADP for ClpB is at least 30-fold greater
than that of ATP. These results suggest that ADP molecules,
generated from ATP under typical assay conditions, can
negatively modulate the ATPase activity of ClpB. This feed-
back, inhibitory effect of ADP may be responsible for the lack
of linearity in the time-dependent hydrolysis of
ATP
by ClpB
(see Fig. 5A).
Effects of Different Nucleotides and
Other
Agents-The
effects of various nucleotides were compared to characterize
further the ATPase (Table
IV).
Both ATP and adenosine 5’-
(a,@-methylene)-triphosphate (Ap(CHz)pp) were good sub-
strates for the ATPase, whereas adenosine 5’-(@,y-methyl-
ene)-triphosphate (App(CHz)p), ADP, and AMP were not
hydrolyzed. GTP, UTP, and CTP were hydrolyzed at 15-25%
rate of the ATP cleavage. Nearly identical data in relative
activity were obtained when hydrolysis of these nucleotides
was measured in the absence of casein (data not shown).
We then examined the effects of various divalent cations
on the ATPase activity of ClpB. ATP hydrolysis reached a
maximal rate at 10 mM concentration of M$+. At 10 mM,
Mn2+ supported the ATPase activity as well as
Me,
and
Ca2+ showed about 25% of their activity.
No
ATPase activity
was observed with other cations, including Fez+, CoZf, Zn2+,
and Cu2+. Hydrolysis of ATP was also examined in the pres-
ence and absence of casein at different pH values. ClpB was
maximally active at pH
8
and was nearly inactive at pH below
6
and above
10.
Can ClpB Support Proteolysis by ClpP?-To examine
whether ClpB can replace ClpA in supporting the proteolytic
activity of ClpP as has been proposed (17, 18), protein break-
down was assayed using [3H]casein as a substrate. As shown
in Table
V,
either the mixture of ClpB and ClpP or ClpB
alone did not have any casein-degrading activity, unlike the
complex of ClpA and ClpP. A number of other proteins, such
as
insulin, globin, denatured BSA, and lysozyme, were neither
hydrolyzed by the mixture of ClpB and ClpP to acid-soluble
materials or to the fragments that can be resolved by poly-
acrylamide gels containing SDS (data not shown). In addition,
ClpB showed little or no effect on the hydrolysis of succinyl-
Leu-Tyr-amidomethylcoumarin
by ClpP in the presence or
TABLE
IV
Hydrolysis of different nucleotides by
ClpB
Assays were performed as described in the legend to Fig.
5,
but in
the presence
of
10
pg
of casein and various nucleotides. The activity
seen with ATP was expressed as
100%.
All nucleotides were titrated
to pH
7.5
prior to the incubation with ClpB.
Nucleotides
(5
mM)
P,
released Relative activity
nmol
%
ATP
24.9 100
ADP
0.0
AMP
0
0.0
0
AP(CH~PP
28.2 113
APP(CH~)P
GTP
0.0
0
5.7 23
CTP
3.7 15
UTP
4.0 16
TABLE
V
Hydrolysis of casein and
succinyl-leu-Tyr-amidomethylcoumarin
Hydrolysis of [3H]casein and the peptide were determined by
incubating
0.2
pg
of
CIpA,
0.05
pg
of
ClpP, and/or
3
pg
of
ClpB at
37
‘C for
1
h in the presence
of
5
mM ATP. Nearly identical data
were obtained for the peptide hydrolysis when the assays were done
without ATP.
Additions Casein hydrolyzed Peptide cleaved
ClpA
ClpB
ClpP
ClpA
+
ClpB
ClpA
+
ClpP
7%
pmol
0 0
0
0
0
75
0 0
23
78
CGB
+
CGP
0
73
ClpA
+
ClpB
+
ClpP
21
77
absence of ATP, ClpA, or both. Thus, it appears unlikely that
ClpB interacts with ClpP under these conditions to form an
active proteolytic complex.
DISCUSSION
One unusual and essential activity of several heat-shock
proteins in
E.
coli is that they are protein- or peptide-activated
ATPases. With protease La and the ClpA subunit of protease
Ti, protein substrates stimulate the ATPase activity of the
enzymes, and this reaction is tightly linked to protein break-
down (7, 12). The major heat-induced protein, DnaK, in
E.
coli (29) and its eukaryotic homolog, Hsp70 (30), as well as
GroEL in
E.
coli
(31), function in protein refolding and
assembly. They also are ATPases that are activated by un-
folded protein or peptide substrates.
The present studies demonstrate that ClpB, a member of
well conserved ClpA family (16), also
is
a protein-activated
ATPase.
In
its structural organization and detailed mecha-
nism, however, ClpB appears to differ from ClpA (10-13) in
a number of respects.
1)
While ClpB appears to be a 350-kDa
tetramer containing primarily 93-kDa subunits, ClpA (when
isolated) behaves as a 140-kDa dimer of 81-kDa subunits.
2)
An anti-ClpB antibody does not cross-react with ClpA and
neither does anti-ClpA antiserum react with ClpB. 3) The
stimulatory effect of casein on ATP hydrolysis is much greater
for ClpB (5-10-fold) than for ClpA (less than 2-fold in the
absence of ClpP and 2-3-fold in its presence). 4) The
K,,,
of
ATP for ClpB (1.1 mM) is much higher than that for ClpA
(0.21 mM). 5) Relative to these
K,,,
values for ATP, the affinity
of ADP for ClpB
(Kc
=
34
WM)
is significantly higher than
that for ClpA (102
p~)
(data not shown). 6) ClpA is maximally
active over a wide pH range from 7 to 10, whereas ClpB has
a sharp pH optimum of
8.
7) While ClpB hydrolyzes
Ap(CHz)pp as rapidly as ATP, ClpA cleaves the analog at
10% as fast as ATP.
(8)
Mn2+ supports ATP hydrolysis as
well as
M$+
for ClpB, but it does about 20% as well as Me
for ClpA.
ClpB has initially been identified as a 84.1-kDa protein
using two-dimensional gel electrophoresis (18), despite the
fact that the size of ClpB predicted from the nucleotide
sequence
is
95.4 kDa (16,18,27). The size of ClpB determined
in the present study by one-dimensional SDS-gel electropho-
resis is 93 kDa, which is much similar to the predicted size.
In addition, the amino acid composition
of
the 93-kDa ClpB
is in close agreement with that calculated from the nucleotide
sequence of the clpB gene (16, 18, 27). Thus, it appears that
the discrepancy in the size of ClpB
is
due to a possible
abnormal behavior of the protein in two-dimensional gels and
that both the 84.1- and 93-kDa proteins represent the full
ClpB sequence.
20434
ClpB
Is
a Protein-activated ATPase
Of interest are the findings that the ClpB preparation
consisting of both 93- and 79-kDa proteins runs as a single
band on a nondenaturing polyacrylamide gel and that the
same preparation is eluted as a single symmetric peak with
an apparent molecular weight of about 350,000 upon analysis
by
FPLC on a Superose-6 column. Therefore, it appears likely
that the 79-kDa protein interacts with the 93-kDa ClpB to
form hetero-oligomeric complexes. However, the portion
of
such hetero-oligomer among the oligomeric complexes of ClpB
should be minor, since the 93-kDa ClpB appears to be 10-20-
fold excess of the 79-kDa protein in the purified ClpB prep-
aration (see Figs.
2
and
4).
It is unknown whether the 79-kDa protein in such a com-
plex
or
by itself has any catalytic activity
(e.g.
as an ATPase).
If indeed the 79-kDa protein is an abbreviated part of the 93-
kDa ClpB
(18),
the 79-kDa protein lacking the 148 N-terminal
amino acids of ClpB should still retain the consensus se-
quences for adenine nucleotide binding sites within two re-
gions of high homology of the ClpA family (16). Thus, the 79-
kDa protein may be a functional protein.
The physiological role(s) of the 93-kDa ClpB and the 79-
kDa protein are also presently unclear, except
clpB
mutant
strains have reduced growth rate and survival at high temper-
atures (17,
18).
ClpB has widely been assumed to be involved
in intracellular proteolysis, because of its sequence similarity
to the ClpA subunit of protease Ti (16-18). In addition,
because both ClpP and ClpB are HtpR-dependent heat-shock
proteins
(14,
17),
it
has been suggested that an ATP-depend-
ent protease containing the ClpB and ClpP polypeptides may
be formed during heat-shock and play an important role
against thermal stress in
E.
coli
(17). However, we failed to
obtain any cleavage of proteins upon incubation of ClpP with
ClpB (unlike the results with ClpP and ClpA). Therefore, it
remains possible that
E.
coli
may contain another ClpP-like
proteolytic component (homologous to ClpP) that associates
with ClpB to form a new type of ATP-dependent protease.
ATP hydrolysis by protease La and Ti are stimulated only
by
those proteins that are degraded by these enzymes, and
protein and ATP hydrolysis seem to occur as a coupled process
(7,
12).
In fact with protease La, these processes are linked
stoichiometrically (9). Of interest is the finding that a limited
number of proteins are capable of activating ATP hydrolysis
by
ClpB. Thus, the proteins that stimulate the ATPase activ-
ity may serve as the substrates for an unknown counterpart
protease component of ClpB. It is also noteworthy that the
proteins that stimulate the ATPase activity of ClpB differ
from those that activate ClpA. For example, denatured BSA,
which is a good substrate for protease Ti, is able to stimulate
ATP hydrolysis by ClpA, but not by ClpB, whether
or
not
ClpP is present. Thus, it is attractive to believe that an
additional protease component, which specifically interacts
with ClpB, may exist and play an important role in intracel-
lular protein breakdown in
E.
coli,
particularly under thermal
stress. Alternatively, ClpB may function as a molecular chap-
erone, like DnaK
or
GroEL, and may catalyze the
ATP-
dependent refolding or reassembly of unfolded proteins in
heat-shocked cells.
Acknowledgments-We are grateful to
Dr.
Catherine Squires
(Co-
lumbia University) for providing the strain carrying the
clpB
clone
and Dr. Young Mok Park (Korea Basic Science Center) for the
analysis
of
amino acid composition and N-terminal sequence of ClpB.
REFERENCES
1.
Chung, C. H., and Goldberg, A. L.
(1981)
Proc.
Natl.
Acad. Sci.
U.
S.
A.
78,
A471-A47&
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
22.
21.
23.
24.
25.
26.
27.
28.
29.
Charette, M.
F.,
Henderson,
G.
W., and Markovitz, A.
(1981)
Proc.
Natl.
Phillips,
T.
A,, VanBogelen, R. A., and Neidhardt,
F.
C.
(1984)
J.
Bacteriol.
Goff,
S.
A,, Casson, L. P., and Goldberg, A. L.
(1984)
Proc.
Natl.
Acad. Sei.
Goff,
S.
A., and Goldberg, A. L.
(1985)
Cell
41,587-595
Gottesman,
S.
(1989)
Annu.
Reu.
Genet.
23,163-198
Waxman,
L.,
and Goldberg,
A.
L.
(1982)
Proc. Natl. Acad. Sci.
U.
S.
A.
79,
Chin, D.
T.,
Goff,
S.
A., Webster,
T.,
Smith,
T.,
and Goldberg, A. L.
(1988)
Menon,
A.
S.,
and Goldberg, A. L.
(1987)
J.
Biol. Chem.
262,14921-14928
Katayama-Fujimura, Y., Gottesman,
S.,
and Maurizi,
M.
R.
(1987)
J.
Biol.
Hwang, B. J., Park, W.
J.,
Chung, C. H., and Goldberg, A. L.
(1987)
Proc.
Hwang, B.
J.,
Woo, K. M., Goldberg, A. L., and Chung, C. H.
(1988)
J.
Katayama, Y., Gottesman,
S.,
Pumphrey,
J.,
Rudikoff,
S.,
Clark, W. P.,
Kroh, H. E., and Simon, L. D.
(1990)
J.
Bacteriol.
172,6026-6034
Gottesman,
S.,
Clark, W. P., and Maurizi, M. R.
(1990)
J.
Biol. Chem.
265,
Gottesman
S
S
uires, C., Pichersky,
E.,
Carrin on, M., Hobbs,
M.,
7886-7893
Mattick,'J.
k,
%ah mple,
B.,
Kuramitsu, H., Sgroza,
T.,
Foster,
T.,
Clark, W.
C.,
Ross,
5..
Squires, C. L., and Maurizi, M. R.
(1990)
Proc.
Kitagawa, M., Wada, C., Yoshioka,
S.,
and Yura,
T.
(1991)
J.
Bacteriol.
Natl.
Acad. Sa.
U.
S.
A.
87,3513-3517
Squires, C. L., Pedersen,
S.,
Ross, B. M., and Squires, C.
(1991)
J.
Bacteriol.
173,4247-4253
Jentoft N., and Dearborn, D.
G.
(1979)
J.
Biol.
Chem.
246,4359-4365
173,4257-4262
Swamy: K. H.
S.,
Chung, C. H, and Goldberg, A. L.
(1983)
Arch. Biochem.
Woo,
K.
M., Chung, W.
J.,
Ha, D.
B.,
Goldberg, A. L., and Chung, C. H.
Ames, B.
(1966)
Methods Enzymol.
8,
115-118
Bradford M. M.
(1976)
Anal. Biochem.
72,248-254
Laemmli'
U.
K.
(1970)
Nature
227,680-685
Brake,
hi.
S.,
Johnston, K. H., Russell-Jones,
G.
J., and Gotschlich, E. C.
Davis, B.
J.
(1964)
Ann.
N.
Y.
Acud. Sci.
121,404-427
Shen, W. F., Squires, C., and Squires, C. L.
(1982)
Nucleic Acids
Res.
10,
Pontis,
E.,
Sun,
X.,
Jornvall,
H.,
Krook, M., and Reichard,
P.
(1991)
Flycn, G. C., Chappell,
T.
G., and Rothman,
J.
E.
(1989)
Science
248,850-
_""
."-"
Acad. Sci.
U.
S.
A.
78,4728-4732
159,283-287
U.
S.
A.
81,6647-6651
4883-4887
J.
Biol. Chem.
263, 11718-11728
Chem.
262,4477-4485
Natl. A~ad. Set.
U.
S.
A.
84,5550-5554
Biol.
Chem.
263,8727-8734
and Maurizi, M.
R.
(1988)
J.
Biol.
Chem.
263, 15226-15236
Biophys.
224,543-554
(1989)
J.
Biol. Chem.
264,208&2091
(1984)
Anal. Biochem.
136, 175-179
3303-3313
Biochem. Btophys.
Res.
Commun.
180,1222-1226
30.
Liberek,
K.,
Marszalek,
J.,
Ang, D., Georgopoulos, C., and Zylicz, M.
(1991)
31.
Martin,
J.,
Lan er,
T.,
Boteva, R., Schramel, A., Horwich, A. L., and Hartl,
834
Proc.
Natl.
Acad. Sct.
U.
S.
A.
88, 2874-2878
F.-U.
(1991)
kture
352,36-42
