J. Mol. Biol. (1996) 262, 71–76
COMMUNICATION
Molecular Symmetry of the ClpP Component of the
ATP-dependent Clp Protease, an Escherichia coli
Homolog of 20 S Proteasome
Dong Hae Shin1, Cheol Soon Lee2, Chin Ha Chung2 and
Se Won Suh1*
1
Department of Chemistry
Center for Molecular
Catalysis and 2Department of
Molecular Biology, Seoul
National University, Seoul
151-742, Korea
The ClpP component Clp protease from Escherichia coli has been
crystallized and examined by X-ray crystallography and self-rotation
function calculations. The crystal belongs to the monoclinic space
group P21 with unit cell dimensions of a = 196.9 Å, b = 104.3 Å, c = 162.4 Å
and b = 98.3°. The X-ray diffraction pattern extends at least to 2.5 Å Bragg
spacing when exposed to CuKa X-rays. Self-rotation function analyses
indicate that the ClpP oligomer has 72-point group symmetry. This
symmetry suggests that the ClpP oligomer is a tetradecamer, (ClpP)14 ,
consisting of two heptamers, (ClpP)7 stacked on top of each other in a
head-to-head fashion. The measurement of crystal density indicates that
two independent copies of the ClpP oligomers are present in the
asymmetric unit, giving a crystal volume per protein mass (VM ) of
2.73 Å3/Da and a solvent content of 54.9% (v/v). Self-rotation function
calculations are consistent with the presence of two ClpP tetradecamers in
the asymmetric unit. The Patterson function suggests that a translation of
x = 0.5 and y = 0.5 relates a pair of ClpP oligomers in one asymmetric unit
to another pair in the other asymmetric unit. And the two independent
tetradecamers in one asymmetric unit are related by a relative rotation of
about 18° around the 7-fold axis.
7 1996 Academic Press Limited
*Corresponding author
Keywords: ClpP; TiP; protease; crystal; X-ray crystallography
Since the early 1980s, at least nine distinct
endoproteases have been isolated from Escherichia
coli. Seven of these (Do, Re, Mi, Fa, So, La, and Ti)
are serine proteases and two others (Ci and Pi) are
metalloproteases (Chung, 1993). The proteolytic
activities of two of these proteases, La and Ti, are
ATP-dependent (Menon & Goldberg, 1987; Hwang
et al., 1987). Protease Ti, also called Clp (for
caseinolytic protease), is a high molecular mass
(0700,000 Da), ATP-dependent protease found in
the cytoplasm of E. coli. The Clp protease consists
of two components: the protease component ClpP,
and the ATPase component, ClpA (Hwang et al.,
1988).
The DNA sequence of the clpP gene in E. coli
predicts a polypeptide chain of 207 amino acid
residues, including a 14 residue leader peptide,
Abbreviations used: Mes, 2-[N-morpholino]ethanesulfonic acid; MPD, 2-methyl-2,4-pentanediol.
0022–2836/96/370071–06 $18.00/0
which is rapidly cleaved in vivo to yield the 193
residue protein of molecular mass 21,567 Da
(Maurizi et al., 1990a). Ser111 and His136 of ClpP
were found to be essential for the protease activity
by site-directed mutagenesis (Maurizi et al., 1990b).
They represent two elements of the catalytic triad
found in most serine proteases. The amino acid
sequence around these active-site residues suggests
that ClpP represents a unique class of serine
protease. ClpP-like proteins are also coded by
chloroplast genomes (Maurizi et al., 1990b). The
ClpP has a native molecular mass of about 240,000
under high salt conditions (above 0.1 M KCl),
whereas the native mass under lower salt
conditions is approximately doubled. This was
originally interpreted to suggest the formation of a
dodecamer under high salt conditions (Maurizi
et al., 1990a; Maurizi, 1991). A recent study by
electron microscopy, however, showed that ClpP
subunits are arranged in two heptagonal rings
directly superimposed on each other (Kessel et al.,
7 1996 Academic Press Limited
72
1995). The ClpP by itself is capable of degrading
small peptides (for example, succinyl-Leu-Tyr-amidomethylcoumarin; Woo et al., 1989) and small
proteins such as insulin (Maurizi et al., 1990a).
However, rapid degradation of higher molecular
mass proteins is dependent on the hydrolysis
of ATP by ClpP (Hwang et al., 1987; KatayamaFujimura et al., 1987). ClpP and ClpA associate in
the presence of ATP to form an active proteolytic
complex, ClpAP, which is composed of a tetradecameric ClpP and a hexameric ClpA (Kessel et al.,
1995; Maurizi, 1991). Most large peptides and
proteins are degraded at multiple sites of the ClpAP
complex without release of high molecular mass
intermediates. Processive degradation of protein
substrates has been shown to be a function of the
multiple array of proteolytic active sites within the
ClpP tetradecamer (Thompson et al., 1994). ClpAP
is also an essential component for the degradation
of N-end rule protein substrates in E. coli (Tobias
et al., 1991).
The ATP-dependent proteolytic activity and the
structural organization of the ClpAP complex
suggest that it is functionally related to the
eukaryotic proteasome (Rechsteiner et al., 1993).
Recently, the crystal structure of 20 S proteasome
from the archaeon Thermoplasma acidophilum has
been determined at 3.4 Å resolution. In this
structure, 14 copies of the two different subunits a
and b are arranged in a7 b7 b7 a7 stoichiometry with
72-point group symmetry (Löwe et al., 1995). The
eukaryotic 20 S proteasome, which is composed of
more than ten distinct polypeptides with molecular
masses in the range of 20,000 to 35,000 Da, displays
a similar subunit arrangement upon electron
microscopic examination (Rivett, 1993; Yoshimura
et al., 1993). Although a eukaryotic 20 S proteasome
has been crystallized (Hwang et al., 1994), its X-ray
structure has not been reported. The tetradecameric
subunit arrangement of ClpP is similar to that of the
inner (b-type) subunits of the eukaryotic and
archaeal 20 S proteasomes (Kessel et al., 1995; Löwe
et al., 1995). Moreover, in both the ClpAP complex
and the eukaryotic 26 S proteasome, an oligomeric
ATPase, ClpA or 19 S particle, is attached to one or
both layers of the protease catalytic core, ClpP or
20 S proteasome (Kessel et al., 1995; Peters et al.,
1993). It is interesting to note that ClpA and subunit
4 of the 26 S proteasome show weak sequence
similarity (Dubiel et al., 1992).
ClpA, alone and as a component of the ClpAP
complex, functions like the ATP-dependent chaperones DnaK and DnaJ (Wickner et al., 1994). A heat
shock protein, ClpX, which is an alternative
counterpart of ClpP, also performs chaperone
functions independent of ClpP (Wawrzynow et al.,
1995). When bound to ClpP, both ClpA and ClpX
play a role as protein specificity factors by
presenting different polypeptide substrates in a
form competent for proteolysis by ClpP (Wickner
et al., 1994).
In order to provide a structural basis for
understanding the proteolytic activity of the ClpP
Molecular Symmetry of ClpP
component of Clp protease, the determination of its
three-dimensional structure is necessary. Here, we
report the molecular symmetry of the ClpP
component of E. coli Clp protease as determined by
X-ray crystallography.
Plate-shaped crystals of the ClpP component of
Clp protease from E. coli have been grown in three
weeks to dimensions of 0.8 mm × 0.5 mm × 0.2 mm.
When these crystals were exposed to CuKa X-rays,
their diffraction spots were observable to at least
2.5 Å Bragg spacing. A set of X-ray data has been
collected to approximately 4 Å from a native
crystal. A primitive monoclinic unit cell with
dimensions of a = 196.9 Å, b = 104.3 Å, c = 162.4 Å
and b = 98.3° was derived by the autoindexing and
parameter refinement procedure of the MADNES
software (Messerschmidt & Pflugrath, 1987). The
final merged data set consists of 95,798 measurements of 44,430 unique reflections with an Rmerge
(on intensity) of 8.8% (rejecting 6.9% outliers). The
merged data set is 77.6% complete to 4.0 Å (the
completeness between 4.6 and 4.0 Å is 64.6%). The
space group was determined to be P21 by
inspecting the intensity distribution of the X-ray
data. The crystal density measurement gave the
average crystal density of 1.18(20.01) g/cm3.
Expected crystal density for one, two, three or four
molecules of 302,000 Da in the asymmetric unit is
calculated to be 1.10, 1.17, 1.25 or 1.32 g/cm3,
respectively. Therefore, two ClpP tetradecamers are
present in the asymmetric unit (or four tetradecamers in the unit cell), with the corresponding
crystal volume per protein mass (VM ) of 2.73 Å3/Da
and a solvent content of 54.9% (v/v). These are
within the observed ranges (Matthews, 1968).
Self-rotation function calculations are consistent
with the presence of two ClpP tetradecamers in the
asymmetric unit.
A strong peak at f = 4.5°, c = 90.0° (or equivalently at f = 184.5°, c = 90.0°) in the k = 51.5°
section of the self-rotation function clearly indicates
the presence of a 7-fold rotation symmetry in the
ClpP oligomer (Figure 1(a)). A single strong peak
for the 7-fold rotation symmetry suggests that the
7-fold symmetry axes of all four ClpP oligomers in
the unit cell are parallel with each other. And
f = 4.5° means that the non-crystallographic 7-fold
symmetry axes lie roughly parallel with the
crystallographic a axis, with an inclination of about
4.5°. Moreover, when the height around the peak
maximum at f = 4.5°, c = 90° (or equivalently at
f = 184.5°, c = 90.0°) is plotted as a function of k
(Figure 1(b)), three sets of peaks, with a separation
in k of 51.4° (=360°/7) can be recognized. The first
series of peaks occur at k = 51.5°, 102.5°, 154.5°, the
second series at k = 18.5°, 69.0°, 121.0°, 171.0°, and
the third series at k = 33.5°, 83.5°, 135.5°. The latter
two series reflect the intermolecular relationship
between a pair of ClpP oligomers. The second series
of peaks indicate that, between a certain pair of
ClpP oligomers in the unit cell, there is a relative
rotation of approximately 18° around the 7-fold
axis. The third series of peaks correspond to a
73
Molecular Symmetry of ClpP
(a)
(b)
Figure 1. Results of self-rotation
function calculations using the programs GLRF (Tong & Rossmann,
(c)
1990) and X-PLOR (Brünger, 1992).
Different resolution intervals in the
range from 50 to 4.0 Å and integration radii from 30 to 110 Å were
tried, with some limitations depending on the coupling between
radius of integration and the maximum usable resolution. The highest
signal-to-noise ratio was obtained
with data between 50 and 5 Å
resolution and with an integration
radius of 50 Å. (a) The k = 51.5°
section of the self-rotation function.
X-ray data between 50 and 5 Å were
used with a Patterson cut-off radius
of 50 Å. Contouring starts at the
6.0s level with an interval of 0.25s.
The directions of rotation axes are
plotted in spherical polar co-ordinates where the tilt of the rotation
axes away from the crystallographic
b axis, c, is plotted latitudinally and
the rotation within the ac* plane, f,
longitudinally. The a axis corresponds to f = 0°. The k rotation is applied around an axis defined by f and c angles.
(b) A plot of the peak height at f = 4.5°, c = 90.0° (or equivalently at f = 184.5°, c = 90.0°) as a function of k. The
resolution of the data and the integration radius for the calculations are the same as in (a). (c) The k = 180.0° section
of the self-rotation function. The data and the integration radius used are the same as in (a). Contouring starts at the
3.5s level with an interval of 0.25 s. ClpP, overproduced in E. coli, was purified as described (Hwang et al., 1988).
Purified ClpP was concentrated with an Amicon YM10 Diaflo membrane and the concentrated protein solution was
dialysed against 10 mM Tris-HCl (pH 7.9), 5 mM MgCl2 . The protein concentration was adjusted to 7 mg/ml using
the above dialysis buffer (assuming 0.36 A1 cm,280 nm = 1 mg/ml). This was used for the crystallization experiment by the
hanging drop vapour diffusion method at 23(21)°C. The coverglass with a hanging drop was placed over the 1 ml
reservoir solution in each well of the tissue culture plate and an air-tight seal was made with grease. Initial
crystallization conditions were established by the incomplete factorial method (Carter & Carter, 1979). The best crystals
were obtained under the following conditions. The reservoir solution contained 200 mM Mes-KOH (pH 6.0), 50% (w/v)
MPD. The hanging drop was prepared by mixing 2.5 ml each of the above protein solution and a solution containing
3% (w/v) MPD, 12.5 mM benzene sulphonic acid, 12.5 mM benzoic acid, 12.5 mM sodium tartrate, 12.5 mM trisodium
citrate, 200 mM Mes-KOH at a final pH of 6.2. In order to determine the number of molecules per asymmetric unit,
the measurement of crystal density was achieved by the procedure of Low & Richards (1952). A density gradient
formed by placing p-xylene over bromobenzene was calibrated with droplets of standard potassium bromide solutions.
X-ray diffraction data were collected on the FAST diffractometer system (Enraf-Nonius) using graphite-monochromatized CuKa X-rays from a rotating anode generator (Rigaku RU-200), running at 40 kV and 60 mA, with a 0.3 mm
focus cup and a 0.6 mm collimator. The MADNES software (Messerschmidt & Pflugrath, 1987) was used for data
collection. The reflection intensities were obtained by the profile fitting procedure (Kabsch, 1988) and the data were
scaled by the Fourier scaling program (Weissman, 1982).
74
Molecular Symmetry of ClpP
Figure 2. A stereo plot of a possible packing arrangement of four ClpP oligomers in the unit cell. For clarity, ClpP
oligomers drawn as a stack of heptagonal rings are scaled down. Dotted lines are drawn within the oligomers to
represent one of the 2-fold symmetry axes. One of the 2-fold axes of ClpP oligomers, molecules 1 and 4 (M1 and M4),
is exactly parallel with the crystallographic b axis. That of the others, molecules 2 and 3 (M2 and M3), is rotated around
the molecular 7-fold axes by either +18° or 33° (equivalent to −18°).
rotation of approximately 33°, or equivalently −18°
(333° − 51.4°), around the 7-fold axis of one ClpP
oligomer relative to the other in the unit cell. The
relative rotation of 218° between a pair of the ClpP
oligomers is also suggested by the k = 180° section
of the self-rotation function, as discussed below.
The first series of peaks account for the intramolecular 7-fold symmetry of the four ClpP oligomers in
the unit cell, as well as the intermolecular
relationship between a pair of oligomers with no
relative rotation around the 7-fold axis. Therefore,
the height of the peaks in the first series is greater
than those of the other two series.
Figure 1(c) is a plot of the k = 180° section of the
self-rotation function, where 2-fold rotation symmetry will appear. Three sets of seven peaks, with
a uniform peak separation in c of 25.7° (=360°/14)
in each set, can be identified at f = 94.5°. They are
at (c = 0.0°, 25.5°, 51.5°, 77.5°, 103.0°, 128.5°, 154.5°),
(c = 9.0°, 34.0°, 60.5°, 86.0°, 112.0°, 138.5°, 163.5°)
and (c = 17.0°, 41.5°, 68.0°, 94.0°, 119.5°, 146.0°,
171.0°). This result indicates that the ClpP oligomer
has a set of seven 2-fold rotation axes, all
perpendicular to the non-crystallographic 7-fold
axis, which was indicated by a strong peak at
f = 4.5°, c = 90° (or equivalently at f = 184.5°,
c = 90.0°) in Figure 1(a). The above three sets of
2-fold peaks confirm our interpretation of the
k = 51.5° section of the self-rotation function, which
indicated the relative rotations among the four
ClpP oligomers in the unit cell, and suggest the
orientations of the molecular 2-fold rotation axes of
the ClpP oligomers as follows. The first set of seven
peaks indicates that one of the 2-fold axes of the
ClpP oligomers in the unit cell is virtually exactly
parallel with the crystallographic b axis. The
crystallographic symmetry of space group P21
dictates a multiple of a pair of the oligomers with
such 2-fold axes in the unit cell. Therefore, among
the four oligomers in the unit cell, only two
oligomers can possibly have one of their molecular
2-fold axes exactly parallel with the crystallographic b axis. The intermolecular 2-fold relationships between these ClpP oligomers and each of the
other two oligomers in the unit cell, which are
rotated around the molecular 7-fold axis by either
+18° or 33° (equivalent to −18°) with respect to the
first two ClpP oligomers, explain each of the latter
two sets. The peaks in the first set are stronger than
those in the other two sets, because they include
contributions from the intermolecular 2-fold relationship as well as the intramolecular relationship.
In summary, our self-rotation function analyses
indicate that the ClpP oligomer has 72-point group
symmetry. This symmetry suggests that the ClpP
oligomer is a tetradecamer, (ClpP)14 , consisting of
two heptamers, (ClpP)7 , stacked on top of each
other in a head-to-head fashion. A recent electron
microscopic study has suggested the same molecular symmetry (Kessel et al., 1995). And the
relationship between a pair of tetradecamers in the
unit cell is a relative rotation of either 0° or 218°
around the non-crystallographic 7-fold axis. The
7-fold symmetry axes of all four ClpP oligomers in
the unit cell are parallel with each other, lying
roughly parallel with the crystallographic a axis,
with an inclination of about 4.5°.
The native Patterson maps were calculated at low
resolution using 50 to 10 Å, to 50 to 5 Å data. As
expected from the fact that a local 2-fold is virtually
parallel with the crystallographic b axis, a strong
peak appears at (u, v, w) = (0.5, 0.5, 0.0), with its
height being 61% of the origin peak. It indicates
pseudo C2 space group symmetry. Indeed, our
X-ray data to approximately 10 Å resolution show
the systematic absences consistent with pseudo C2
symmetry. This is, however, broken at higher
resolution and the systematic absence of the present
data is consistent with the space group P21 .
75
Molecular Symmetry of ClpP
One possible packing of four ClpP oligomers in
the unit cell shown in Figure 2 illustrates what has
been stated above. The two independent ClpP
oligomers in the asymmetric unit (molecules 1 and
2 or molecules 3 and 4 in Figure 2) are related to
each other by a 2-fold symmetry, which is
approximately similar to that in the space group C2.
The two oligomers in one asymmetric are related to
the other pair in the other asymmetric unit by a
translation of x = 0.5 and y = 0.5, in addition to the
relative rotation of 218° around the molecular
7-fold axis.
When a subunit of ClpP is modelled as a sphere,
its radius is estimated to be about 37 Å. Assuming
a spherical subunit and an eclipsed configuration of
the two layers, the ClpP oligomer is estimated to
have a diameter of 0110 Å and a height of 075 Å.
This is in broad agreement with the molecular
dimensions estimated by electron microscopy, a
diameter of 0110 Å and a height of 0100 Å (Kessel
et al., 1995). A height less than 0100 Å is apparent
in the averaged image of the ClpAP complex
produced by electron microscopy (Figure 2b in
Kessel et al., 1995). The authors mention a source of
a possible error in the estimated height, i.e. a
staining artifact. We note that the length of the b
axis (0104 Å) is similar to the diameter of the ClpP
oligomer.
It is interesting that the structural organization of
the ClpP oligomer resembles the subunit assembly
of 20 S proteasome from the archaeon T. acidophilum
(Löwe et al., 1995). This may reflect a possible
common functional mechanism between the two
protease complexes in cellular protein degradation.
Interestingly, the chaperonin GroEL and its
cochaperonin GroES, which participate in promoting protein folding, have a common architecture
with a ring or a dome structure of 7-fold symmetry
(Braig et al., 1994; Hunt et al., 1996). However, there
is a fundamental difference between the GroEL and
20 S proteasome from T. acidophilum in the size of
entrance to the active sites (Weissman et al., 1995).
The 20 S proteasome has a smaller opening at its
entrance and thus most proteins are protected from
entering the active-site cavity. It will be interesting
to compare the structures of E. coli ClpP and the
bacterial 20 S proteasome to uncover structural
differences as well as similarities in more detail. In
order to determine the three-dimensional structure
of the ClpP component, the multiple isomorphous
replacement method of X-ray crystallography will
be attempted.
Acknowledgements
We thank the Inter-University Center for Natural
Science Research Facilities for providing the X-ray
equipment. This work was supported by the Basic
Science Research Institute grant from Korea Ministry of
Education (S.W.S.), center for Molecular Catalysis
(S.W.S.) and Research Center for Cell Differentiation
(C.H.C.).
References
Braig, K., Otwinowski, Z., Hegde, R., Boisvert, D. C.,
Joachimiak, A., Horwich, A. L. & Sigler, P. B. (1994).
The crystal structure of the bacterial chaperonin
GroEL at 2.8 Å. Nature, 371, 578–586.
Brünger, A. T. (1992). X-PLOR Version 3.1. A System for
X-ray Crystallography and NMR. Yale University
Press, New Haven and London.
Carter, C. W., Jr & Carter, C. W. (1979). Protein
crystallization using incomplete factorial experiments. J. Biol. Chem. 254, 12219–12223.
Chung, C. H. (1993). Proteases in Escherichia coli. Science,
262, 372–374.
Dubiel, W., Ferrell, K., Pratt, G. & Rechsteiner, M. (1992).
Subunit 4 of the 26 S protease is a member of a novel
eukaryotic ATPase family. J. Biol. Chem. 267,
22699–22702.
Hunt, J. F., Weaver, A. J., Landry, S. J., Gierasch, L. &
Deisenhofer, J. (1996). The crystal structure of the
GroES co-chaperonin at 2.8 Å resolution. Nature, 379,
37–45.
Hwang, B. J., Park, W. J., Chung, C. H. & Goldberg, A. L.
(1987). Escherichia coli contains a soluble ATP-dependent protease (Ti) distinct from protease La. Proc.
Natl Acad. Sci. USA, 84, 5550–5554.
Hwang, B. J., Woo, K. M., Goldberg, A. L. & 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.
Hwang, K. Y., Kim, K. K., Chung, C. H., Tanaka, K. &
Suh, S. W. (1994). Crystallization of ‘20 S’ proteasome from rat liver. Mol. Cells, 4, 273–275.
Kabsch, W. (1988). Evaluation of single-crystal X-ray
diffraction data from a position-sensitive detector.
J. Appl. Crystallog. 21, 916–924.
Katayama-Fujimura, Y., Gottesman, S. & Maurizi, M. R.
(1987). A multiple-component, ATP-dependent
protease from Escherichia coli. J. Biol. Chem. 262,
4477–4485.
Kessel, M., Maurizi, M. R., Kim, B., Kocsis, E., Trus, B. L.,
Singh, S. K. & 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.
Low, B. W. & Richards, F. M. (1952). The use of the
gradient tube for the determination of crystal
densities. J. Am. Chem. Soc. 74, 1660–1666.
Löwe, J., Stock, D., Jap, B., Zwickl, P., Baumeister, W. &
Huber, R. (1995). Crystal structure of the 20 S
proteasome from the archaeon T. acidophilum at 3.4 Å
resolution. Science, 268, 533–539.
Matthews, B. W. (1968). Solvent content of protein
crystals. J. Mol. Biol. 33, 491–497.
Maurizi, M. R. (1991). ATP-promoted interaction between
Clp A and Clp P in activation of Clp protease from
Escherichia coli. Proteinases, 19, 719–723.
Maurizi, M. R., Clark, W. P., Katayama, Y., Rudikoff, S.,
Pumphrey, J., Bowers, B. & Gottesman, S. (1990a).
Sequence and structure of Clp P, the proteolytic
component of the ATP-dependent Clp protease of
Escherichia coli. J. Biol. Chem. 265, 12536–12545.
Maurizi, M. R., Clark, W. P., Kim, S.-H. & Gottesman, S.
(1990b). Clp P represents a unique family of serine
proteases. J. Biol. Chem. 265, 12546–12552.
Menon, A. S. & Goldberg. A. L. (1987). Binding of
nucleotides to the ATP-dependent protease La from
Escherichia coli. J. Biol. Chem. 262, 14921–14928.
76
Molecular Symmetry of ClpP
Messerschmidt, A. & Pflugrath, J. W. (1987). Crystal
orientation and X-ray pattern prediction routines for
area-detector diffractometer systems in macromolecular crystallography. J. Appl. Crystallog. 20,
306–315.
Peters, J.-M., Cejka, Z., Harris, J. R., Kleinschmidt, J. A.
& Baumeister, W. (1993). Structural features of the
26 S proteasome complex. J. Mol. Biol. 234, 932–937.
Rechsteiner, M., Hoffman, L. & Dubiel, W. (1993). The
multicatalytic and 26 S proteases. J. Biol. Chem. 268,
6065–6068.
Rivett, A. J. (1993). Proteasomes: multicatalytic proteinase
complexes. Biochem. J. 291, 1–10.
Tobias, J. W., Shrader, T. E., Rocap, G. & Varshavsky, A.
(1991). The N-end rule in bacteria. Science, 254,
1374–1377.
Thompson, M. W., Singh, S. K. & Maurizi, M. R. (1994).
Processive degradation of proteins by the ATPdependent 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.
Tong, L. & Rossmann, M. G. (1990). The locked rotation
function. Acta Crystallog. sect. A, 46, 783–792.
Wawrzynow, A., Wojtkowiak, D., Marszalek, J., Banecki,
B., Jonsen, M., Graves, B., Georgopoulos, C. & Zylicz,
M. (1995). The ClpX heat-shock protein of Escherichia
coli, the ATP-dependent substrate specificity component of the ClpP-ClpX protease, is a novel
molecular chaperone. EMBO J. 14, 1867–1877.
Weissman, J. S., Sigler, P. B. & Horwich, A. L. (1995).
From the cradle to the grave: ring complexes in the
life of a protein. Science, 268, 523–524.
Weissman, L. (1982). Strategies for extracting isomorphous and anomalous signals. In Computational
Crystallography (Sayre, D., ed.), pp. 56–63, Oxford
University Press (Clarendon), Oxford.
Wickner, S., Gottesman, S., Skowyra, D., Hoskins, J.,
McKenney, K. & Maurizi, M. R. (1994). A molecular
chaperone, ClpA, functions like DnaK and DnaJ.
Proc. Natl Acad. Sci. USA, 91, 12218–12222.
Woo, K. M., Chung, W. J., Ha, D. B., Goldberg, A. L. &
Chung, C. H. (1989). Protease Ti from Escherichia coli
requires ATP hydrolysis for protein breakdown but
not for hydrolysis of small peptides. J. Biol. Chem.
264, 2088–2091.
Yoshimura, T., Kameyama, K., Takagi, T., Ikai, A.,
Tokunaga, F., Koide, T., Tanahashi, N., Tamura, T.,
Cejka, Z., Baumeister, W., Tanaka, K. & Ichihara, A.
(1993). Molecular characterization of the ‘‘26 S’’
proteasome complex from rat liver. J. Struct. Biol.
111, 200–211.
Edited by I. A. Wilson
(Received 6 February 1996; received in revised form 20 June 1996; accepted 1 July 1996)