ReportHeat activates the AAA+ HslUV protease by melting
an axial autoinhibitory plugGraphical AbstractHighlightsd A trimeric structure can plug the axial channel of the AAA+
HslU ring hexamer
d This autoinhibitory plug is observed in crystal structures and
by electron microscopy
d Temperature-induced melting of the plug activates ATP
hydrolysis and proteolysis
d Low-temperature autoinhibition limits rogue degradation
after recovery from heat shockBaytshtok et al., 2021, Cell Reports 34, 108639
January 19, 2021 ª 2020 The Author(s).
https://doi.org/10.1016/j.celrep.2020.108639Authors
Vladimir Baytshtok, Xue Fei,
Tsai-Ting Shih, Robert A. Grant,
Justin C. Santos, Tania A. Baker,
Robert T. Sauer
Correspondence
bobsauer@mit.edu
In Brief
Baytshtok et al. demonstrate that the
activity of HslUV, a AAA+ heat shock
protease, is regulated by thermal melting
of its autoinhibitory axial plug, which
activates ATP hydrolysis, substrate
binding, and energy-dependent
proteolysis and ensures that robust
protein degradation by HslUV occurs only
at elevated temperatures in the cell.ll
OPEN ACCESS
llReport
Heat activates the AAA+ HslUV protease
by melting an axial autoinhibitory plug
Vladimir Baytshtok,1,2,4 Xue Fei,1,2 Tsai-Ting Shih,1,2 Robert A. Grant,1 Justin C. Santos,1,5 Tania A. Baker,1
and Robert T. Sauer1,3,6,*
1Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
2These authors contributed equally
3Senior author
4Present address: Structural Biology Program, Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, New York, NY 10065, USA
5MD/PhD Program, Emory University School of Medicine, 100 Woodruff Circle, Atlanta, GA 30322, USA
6Lead Contact
*Correspondence: bobsauer@mit.edu
https://doi.org/10.1016/j.celrep.2020.108639SUMMARYAt low temperatures, protein degradation by the AAA+ HslUV protease is very slow. New crystal structures
reveal that residues in the intermediate domain of the HslU6 unfoldase can plug its axial channel, blocking
productive substrate binding and subsequent unfolding, translocation, and degradation by the HslV12 pepti-
dase. Biochemical experiments with wild-type and mutant enzymes support a model in which heat-induced
melting of this autoinhibitory plug activates HslUV proteolysis.INTRODUCTION
By degrading incomplete, unfolded, or unneeded proteins,
intracellular AAA+ proteases serve important roles in pro-
tein-quality control, often removing proteins damaged by
heat or other destabilizing stresses (Sauer and Baker, 2011).
HslUV consists of one or two AAA+ HslU6 ring hexamers
and the double-ring HslV12 peptidase (Figure 1A; Rohrwild
et al., 1997; Sousa et al., 2000; Wang et al., 2001). In Escher-
ichia coli, expression of the HslU and HslV enzymes increases
~10-fold at high temperatures (Rohrwild et al., 1996), and
HslUV proteolysis is substantially faster at high than low tem-
peratures in vitro (Burton et al., 2005). ATP-dependent prote-
olysis requires HslU6 to bind the degrons of target proteins in
its axial channel and then unfold and translocate the substrate
into the proteolytic chamber of HslV12 (Sundar et al., 2012). To
reach the axial channel of the HslU hexamer, substrates must
transit a funnel-like structure formed by segments of its inter-
mediate (I) domain (Figures 1A and 1B; Sousa et al., 2000;
Wang et al., 2001). Both the I-domain and this funnel-like
structure are unique to the HslU subfamily of AAA+ unfoldases
(Sauer and Baker, 2011). An I-domain L199Q mutation en-
hances degradation of some protein substrates and the rate
of ATP hydrolysis (Baytshtok et al., 2016), possibly by relief
of autoinhibition. However, the molecular mechanism by
which Leu199 and surrounding residues affect proteolysis
and ATP hydrolysis is unknown, as amino acids 177–212 are
disordered in known structures (Bochtler et al., 2000; Sousa
et al., 2000; Wang et al., 2001). Here, we present evidence
for a model in which HslU hexamers are autoinhibited at low
temperatures by a trimeric I-domain plug that blocks the axial
channel but melts at high temperatures to activate proteolysis.C
This is an open access article undRESULTS AND DISCUSSION
Autoinhibited structures of HslU and HslUV
We determined structures for three new crystal forms of E. coli
HslU or HslUV by molecular replacement (Table 1).
Our new structures were similar to previous ones with the
notable exception that a trimeric plug, formed by residues
183–206 of the I-domain, completely blocked the axial channel
of the HslU ring (Figures 1C–1E). Because protein substrates
must access the axial channel of HslU to be engaged, unfolded,
and translocated into the peptidase chamber of HslV, this plug
would inhibit substrate degradation. The plug consisted of amino
acids from three of the six HslU subunits (Figures 1C–1E), each
comprising an extended region (residues 183–188), a sharp
turn (residues 189–191), and an a helix (residues 192–206). We
could not reliably model connections between plug elements
and other regions of the I-domain in specific subunits. Impor-
tantly, however, a crystallographic 3-fold axis related plug ele-
ments in the 6PXI structure, indicating that I-domain elements
from three alternating subunits of the HslU hexamer must form
the plug.
The three a helices that comprise the plug center packed
together via hydrophobic contacts involving the side chains of
Met192, Met195, Leu199, and Phe203. Hence, replacing the
nonpolar Leu199 with a polar Gln199 in the L199QHslU variant
should destabilize helix-helix packing and plug stability, thereby
reducing autoinhibition and providing a structural rationale for
the ability of L199QHslUV to degrade some protein substrates
more rapidly than wild-type HslUV (WTHslUV; Baytshtok et al.,
2016). In prior biochemical experiments, the subtilisin and
chymotrypsin endoproteases were found to cleave sites within
the plug region of L199QHslU faster than in WTHslU (Baytshtokell Reports 34, 108639, January 19, 2021 ª 2020 The Author(s). 1
er the CC BY license (http://creativecommons.org/licenses/by/4.0/).
ll
OPEN ACCESS Report
A B C
D E
F
Figure 1. HslUV with open and closed axial channels
(A) Surface-representation side view of the HslUV protease (PDB: 5JI3).
(B) Surface-representation top view of open-channel HslUV structure (PDB: 5JI3).
(C) Surface-representation top view of plugged-channel HslUV structure (PDB: 6PXI).
(D) Top view of axial plug (PDB: 6PXI) in cartoon representation. The side chains of Arg101, Ile186, and Leu199 are shown as spheres. The axial channel of HslU is
shown in surface representation.
(E) Cutaway side view of axial plug (PDB: 6PXI) in cartoon representation with Arg101, Ile186, and Leu199 shown as spheres.
(F) WebLogo representation of conservation of plug residues.f indicates positions of hydrophobic residues (Val184, Ile186, Met187, Met192, Met195, Leu199, Met202,
and Phe203) that engage in packing within the plug; asterisk (*) indicates the positions of negatively charged residues (Glu193 and Glu194) that appear to form
favorable electrostatic interactions with Arg101 and Lys293 in the axial channel of the HslU hexamer.et al., 2016), supporting a model in which destabilization of the
mutant plug results in partial melting and increased susceptibility
to proteolytic cleavage. Additional hydrophobic packing con-
tacts within the plug included Val184 and Ile186 from the extended
region, which occupy peripheral grooves between plug helices,
andMet202 from the helix. Notably, residues that form hydropho-
bic interactions within the plug are highly conserved in HslU or-
thologs (Figure 1F).2 Cell Reports 34, 108639, January 19, 2021A Pro189-Pro190-Gly191 tripeptide formed a sharp turn between
the extended region and a helix of the plug. The negatively
charged side chains of Glu193 and Glu194 following this turn
were also close to the positively charged side chains of Arg101
(94% conserved) and Lys293 (92% conserved) in the HslU axial
channel, suggesting that favorable electrostatic interactions
also stabilize the plugged conformation. Conservation of the
plug residues that form the turn and electrostatic interactions
ll
Report OPEN ACCESS
Table 1. Crystallographic statistics
PDB entry 6PXI 6PXL 6PXK
Protein HslUV HslU HslU
Resolution (Å) 3.45 3.74 3.65
Space group P321 C2 P2
Unit cell
A (Å) 168.60 414.59 200.03
B (Å) 168.60 92.33 91.20
C (Å) 162.89 200.85 201.80
a 90
 90
 90

b 90
 99.43
 99.43

g 120
 90
 90

No. of unique 29,212 68,314 15,427
reflections
Redundancy 2.2 3.8 5.0
Completeness (%) 81.7 (59.3) 95.3 (77.1) 98.4 (92.5)
CC-1/2 0.985 (0.593) 0.981 (0.669) 0.998 (0.609)
Rsym 0.114 0.121 0.105
Rpim 0.081 0.068 not calculated
Rwork/Rfree 0.257/0.285 0.220/0.265 0.223/0.271
MolProbity 1.04 (100) 1.31 (100) 1.30 (100)
score (percentile)
Clash score 2.54 (100) 5.71 (100) 5.9 (100)
(percentile)
Ramachandran 0 0 0
outliers (%)
Ramachandran 98.58 98.15 99.26
favored (%)
Bad bonds/angles 0/0 0/0 0/0
Cb deviations 0 0 0
Completeness and CC-1/2 values in parentheses are for the highest res-
olution shell.within the HslU channel (Figure 1F) provide additional support for
a model in which the autoinhibited conformation is functionally
important.
Electron microscopy of closed- and open-channel HslU
structures
To investigate channel accessibility in WTHslU and L199QHslU in a
non-crystalline environment, we used negative-stain electron
microscopy. Both enzyme samples were equilibrated at room
temperature prior to application to grids and staining. Class-
average images of WTHslU, viewed approximately down the
axis of the AAA+ ring, showed a shallow axial depression or
blocked channel (Figure 2A, top row), similar to projections of
plugged crystallographic HslU hexamers filtered to 15 Å (Fig-
ure 2A, bottom row). Most class averages for L199QHslU hexam-
ers (Figure 2B, top row), by contrast, resembled open-channel
crystal structures filtered to 15 Å (Figure 2B, bottom row). These
results provide further evidence that the axial channel of WTHslU
at room temperature is plugged in most enzymes and that the
L199Q mutation shifts the equilibrium to favor the open-channel
conformation.Plug stability affects temperature-dependent activity
We propose that high temperature destabilizes the plug and fa-
vors the active open-channel conformation of HslUV, whereas
low temperature stabilizes the inactive plugged-channel confor-
mation (Figure 3A). As a first test of this model, we measured the
rates of ATP hydrolysis by WTHslUV and L199QHslUV at 25
C,
35
C, 45
C, and 55
C without or with the I37AArc-cp6GFP-st11-
ssrA protein substrate (Figures 3B and 3C). These data fit well
tomelting curves calculated assuming that the high-temperature
conformations of WTHslUV and L199QHslUV had the same
ATPase activity, whereas the low-temperature conformations
were inactive in ATP hydrolysis. Without protein substrate, the
fitted temperatures of half-maximal activity (TM) were 42

C for
L199QHslUV and 52
C for WTHslUV (Figure 3B). Protein substrate
reduced the TM values by 10

C–12
C for both enzymes (Fig-
ure 3C), indicating that binding of the I37AArc-cp6GFP-st11-ssrA
substrate preferentially stabilizes the active enzyme conforma-
tion. The lower TM values for
L199QHslUV than for WTHslUV,
both with and without protein substrate, strongly support our
model that plug melting results in temperature-dependent relief
of autoinhibition.
As a second test of our temperature-dependent activation
model, we used SDS-PAGE to measure HslUV degradation of
an Arc-st11-ssrA substrate at 25
C or 55
C (Figure 3D). As ex-
pected from findings of a previous study (Burton et al., 2005),
degradation by the WT enzyme was very slow at 25
C and
much faster at 55
C. We also assayed degradation of the
same substrate by R101AHslUV, constructed to remove favor-
able electrostatic interactions between Arg101 in the axial chan-
nel and glutamates in the autoinhibitory plug, and by
I186NHslUV, which replaces a hydrophobic residue that normally
stabilizes plug packing with a polar residue (Figures 1D and
1E). At 25
C, both the R101AHslUV and I186NHslUV enzymes
degraded Arc-st11-ssrA substantially faster than did WTHslUV
(Figure 3D), as expected by our model. At 55
C, all three en-
zymes degraded Arc-st11-ssrA more rapidly than at the lower
temperature, with WTHslUV degradation being slower than
degradation by the mutant enzymes. In each case, faster sub-
strate degradation at the higher temperature probably results
from a combination of reduced autoinhibition and reduced sta-
bility of the native portion of the Arc-st11-ssrA substrate.
Importantly, these results support a model in which the
R101A and I186N mutations destabilize the autoinhibited
conformation at low temperatures and thereby increase prote-
ase activity relative to the WT enzyme.
As a third test of our model, we determined steady-state
degradation rates of different concentrations of the
I37AArc-cp6GFP-st11-ssrA substrate at a temperature (37
C) at
which both WTHslUV and I186NHslUV have substantial activity
and fit the data to the Michaelis-Menten equation to determine
KM and Vmax values (Figure 3E). At low substrate concentrations,
degradation by I186NHslUV was ~6-fold faster than by WTHslUV,
largely as a consequence of a tighter KM for the mutant (Fig-
ure 3E). By contrast, degradation of high concentrations of this
substrate by I186NHslUV was only ~1.2-fold faster than by
WTHslUV. These results support a model in which plug destabili-
zation, as a consequence of reduced hydrophobic packing,
makes the axial channel accessible in a larger fraction of theCell Reports 34, 108639, January 19, 2021 3
ll
OPEN ACCESS Report
A Figure 2. Effects of L199Q mutation on HslU
channel size
(A) Class-average negative-stain electron micro-
scopy (EM) images of WTHslU (top row) compared
with projections of a plugged hexamer (PDB: 6PXK)
filtered to 15-Å resolution (bottom row). Samples
were prepared at room temperature.
(B) Class-average images of L199QHslU (top row)
compared with filtered images of an open-channel
hexamer (PDB: 5JI3) or the plugged hexamer (PDB:
6PXK).
B
population of I186N mutant enzymes at low substrate concen-
trations, thereby resulting in tighter substrate binding and faster
degradation. Very high substrate concentrations, in turn, appear
to shift the equilibrium toward the open-channel conformation of
WTHslUV, thereby reducing the activity difference between the
mutant and WT enzymes.Conclusions and biological inferences
We have identified a new conformation of the HslUV protease in
which part of the I-domain forms a plug that blocks the axial
channel of the AAA+ HslU ring and prevents productive sub-
strate engagement and thus degradation. This plugged-channel
inactive conformation is in dynamic equilibrium with an open-
channel active conformation, with higher temperature and/or
higher protein-substrate concentrations favoring the active
conformation.
In cells, direct thermal activation of HslUV by melting of the
autoinhibitory plug at heat shock temperatures (Figure 3A) would
increase protease activity almost immediately. Temperature-
induced expression increases of HslU and HslV would then
further amplify HslUV proteolytic capacity. Following a return
to lower temperatures, levels of the HslUV enzyme would be ex-
pected to remain high for several generations before reduced
expression and cell division allowed a return to the low-temper-4 Cell Reports 34, 108639, January 19, 2021ature steady state. During this transition from high to low temper-
atures, increased plug-mediated autoinhibition could contribute
to cell fitness by minimizing rogue HslUV proteolysis and/or
reducing excessive ATP hydrolysis.STAR+METHODS
Detailed methods are provided in the online version of this paper
and include the following:
d KEY RESOURCES TABLE
d RESOURCE AVAILABILITYB Lead Contact
B Materials Availability
B Data and Code Availability
d EXPERIMENTAL MODEL AND SUBJECT DETAILS
d METHOD DETAILS
d QUANTIFICATION AND STATISTICAL ANALYSISACKNOWLEDGMENTS
This work was supported by NIH grant AI-15706. NECAT beamline studies at
APS were supported by NIGMS (P41 GM103403) and DOE (DE-AC02-
06CH11357) grants.
ll
Report OPEN ACCESS
A B C
D
E
Figure 3. Activities of WTHslUV and mutants with destabilized plugs
(A) Model for activation of HslU by thermal melting of the autoinhibitory plug.
(B) Temperature dependence of ATP hydrolysis by WTHslUV or L199QHslUV (0.3 mM HslU6, 0.9 mM HslV12). Values are averages (±SD) of three independent
experiments. The dotted lines are fits to melting curves.
(C) Same as in (B) but in the presence of the I37AArc-cp6GFP-st11-ssrA substrate (50 mM).
(D) SDS-PAGEwas used to assay the kinetics of degradation of the Arc-st11-ssrA substrate (5 mM) at 25
Cor 55
CbyWTHslUV, R101AHslUV, or I186NHslUV (0.3 mM
HslU6, 0.9 mM HslV12). In each subpanel, a representative gel is shown at the top, and the mean (±SD) of substrate remaining in three independent replicates is
shown in the graph below.
(E) Michaelis-Menten plots and steady-state kinetic parameters for I37AArc-cp6GFP-st11-ssrA degradation at 37
C by WTHslUV or I186NHslUV (0.3 mM HslU6,
0.9 mM HslV12). Values plotted are averages (±SD) of three independent replicates.AUTHOR CONTRIBUTIONS
All authors contributed to experimental design. R.T.S. and T.A.B. supervised
research. V.B., X.F., T.-T.S., T.A.B., and R.T.S. wrote the manuscript. V.B.,
X.F., T.-T.S., R.A.G., J.C.S., and R.T.S. performed experiments.
DECLARATION OF INTERESTS
We, the authors and our immediate family members, have no competing finan-
cial interests to declare.
Received: October 17, 2019
Revised: November 24, 2020
Accepted: December 21, 2020
Published: January 19, 2021
REFERENCES
Adams, P.D., Afonine, P.V., Bunkóczi, G., Chen, V.B., Davis, I.W., Echols, N.,
Headd, J.J., Hung, L.W., Kapral, G.J., Grosse-Kunstleve, R.W., et al. (2010).PHENIX: a comprehensive Python-based system for macromolecular struc-
ture solution. Acta Crystallogr. D Biol. Crystallogr. 66, 213–221.
Baytshtok, V., Fei, X., Grant, R.A., Baker, T.A., and Sauer, R.T. (2016). A struc-
turally dynamic region of the HslU intermediate domain controls protein degra-
dation and ATP hydrolysis. Structure 24, 1766–1777.
Bochtler, M., Hartmann, C., Song, H.K., Bourenkov, G.P., Bartunik, H.D., and
Huber, R. (2000). The structures of HsIU and the ATP-dependent protease
HsIU-HsIV. Nature 403, 800–805.
Burton, R.E., Baker, T.A., and Sauer, R.T. (2005). Nucleotide-dependent sub-
strate recognition by the AAA+ HslUV protease. Nat. Struct. Mol. Biol. 12,
245–251.
Chen, V.B., Arendall, W.B., 3rd, Headd, J.J., Keedy, D.A., Immormino, R.M.,
Kapral, G.J., Murray, L.W., Richardson, J.S., and Richardson, D.C. (2010).
MolProbity: all-atom structure validation for macromolecular crystallography.
Acta Crystallogr. D Biol. Crystallogr. 66, 12–21.
Emsley, P., Lohkamp, B., Scott, W.G., and Cowtan, K. (2010). Features and
development of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501.Cell Reports 34, 108639, January 19, 2021 5
ll
OPEN ACCESS ReportGoddard, T.D., Huang, C.C., Meng, E.C., Pettersen, E.F., Couch, G.S., Morris,
J.H., and Ferrin, T.E. (2018). UCSF ChimeraX: Meeting modern challenges in
visualization and analysis. Protein Sci. 27, 14–25.
Kabsch, W. (2010). XDS. Acta Crystallogr. D Biol. Crystallogr. 66, 125–132.
McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C.,
and Read, R.J. (2007). Phaser crystallographic software. J. Appl. Cryst. 40,
658–674.
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.
Mindell, J.A., andGrigorieff, N. (2003). Accurate determination of local defocus
and specimen tilt in electron microscopy. J. Struct. Biol. 142, 334–347.
Nørby, J.G. (1988). Coupled assay of Na+,K+-ATPase activity. Methods Enzy-
mol. 156, 116–119.
Otwinowski, Z., and Minor, W. (1997). Processing of x-ray diffraction data
collected in oscillation mode. Methods Enzymol. 276, 307–325.
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 to the eukaryotic proteasome.
Proc. Natl. Acad. Sci. USA 93, 5808–5813.
Rohrwild, M., Pfeifer, G., Santarius, U., Mu€ller, S.A., Huang, H.C., Engel, A.,
Baumeister, W., and Goldberg, A.L. (1997). The ATP-dependent HslVU prote-6 Cell Reports 34, 108639, January 19, 2021ase from Escherichia coli is a four-ring structure resembling the proteasome.
Nat. Struct. Biol. 4, 133–139.
Sauer, R.T., and Baker, T.A. (2011). AAA+ proteases: ATP-fueled machines of
protein destruction. Annu. Rev. Biochem. 80, 587–612.
Scheres, S.H. (2012). RELION: implementation of a Bayesian approach to
cryo-EM structure determination. J. Struct. Biol. 180, 519–530.
Sousa, M.C., Trame, C.B., Tsuruta, H., Wilbanks, S.M., Reddy, V.S., and
McKay, D.B. (2000). Crystal and solution structures of an HslUV protease-
chaperone complex. Cell 103, 633–643.
Sundar, S., Baker, T.A., and Sauer, R.T. (2012). The I domain of the AAA+
HslUV protease coordinates substrate binding, ATP hydrolysis, and protein
degradation. Protein Sci. 21, 188–198.
Tang, G., Peng, L., Baldwin, P.R., Mann, D.S., Jiang, W., Rees, I., and Ludtke,
S.J. (2007). EMAN2: an extensible image processing suite for electron micro-
scopy. J. Struct. Biol. 157, 38–46.
Wang, J., Song, J.J., Seong, I.S., Franklin, M.C., Kamtekar, S., Eom, S.H., and
Chung, C.H. (2001). Nucleotide-dependent conformational changes in a pro-
tease-associated ATPase HsIU. Structure 9, 1107–1116.
Yakamavich, J.A., Baker, T.A., and Sauer, R.T. (2008). Asymmetric nucleotide
transactions of the HslUV protease. J. Mol. Biol. 380, 946–957.
ll
Report OPEN ACCESSSTAR+METHODSKEY RESOURCES TABLEREAGENT or RESOURCE SOURCE IDENTIFIER
Bacterial Strains
X90 (lDE3) slyD::kan hslUV::tet Baytshtok et al., 2016 N/A
Proteins and Enzymes
HslU (His6 tagged) Bochtler et al., 2000 N/A
L199QHslU (His6 tagged) Baytshtok et al., 2016 N/A
R101AHslU (His6 tagged) This study N/A
I186NHslU (His6 tagged) This study N/A
E257QHslU (His6 tagged) Yakamavich et al., 2008 N/A
HslV (His6 tagged) Bochtler et al., 2000 N/A
I37AArc-cp6GFP-st11-ssrA Baytshtok et al., 2016 N/A
Arc-st11-ssrA Milla et al., 1993 N/A
pyruvate kinase Sigma-Aldrich N/A
lactate dehydrogenase Sigma-Aldrich N/A
Deposited Data
E257QHsU-HslV-ADP, plugged This study PDB: 6PXI
WTHslU- ADP, plugged This study PDB: 6PXL
WTHslU- ADP, plugged This study PDB: 6PXK
Software and Algorithms
Coot Emsley et al., 2010 N/A
ChimeraX Goddard et al., 2018 N/A
Ctffind3 Mindell and Grigorieff, 2003 N/A
EMAN2 Tang et al., 2007 N/A
HKL2000 Otwinowski and Minor, 1997 N/A
Phenix Adams et al., 2010 N/A
PyMOL Schrödinger, LLC N/A
Relion Scheres, 2012 N/A
XDS Kabsch, 2010 N/ARESOURCE AVAILABILITY
Lead Contact
Requests for information and resources should be directed to and will be fulfilled by the Lead Contact, Robert T. Sauer (bobsauer@
mit.edu).
Materials Availability
This study generated newHis6-taggedmutants of E. coliHslU containing the R101A and I186Nmutations. Plasmids expressing these
mutant enzymes are available from the lead contact upon request without restriction.
Data and Code Availability
Coordinates, structure factors and electron-density maps for the 6PXI, 6PXL, and 6PXK crystal structures are available from the
RCSB Protein Data Bank (https://www.rcsb.org/). The biochemical data and electron micrographs supporting the current study
are available from the Lead Contact upon request.
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Proteins were expressed in E. coli strain X90 (lDE3) slyD::kan hslUV::tet). No additional biological strains were used in this work.Cell Reports 34, 108639, January 19, 2021 e1
ll
OPEN ACCESS ReportMETHOD DETAILS
Genes encoding H -tagged variants of E. coliHslU, E257QHslU, L199QHslU, I186NHslU, R101AHslU, HslV, I37AArc-cp66 GFP-st11-ssrA, and
Arc-st11-ssrA were expressed from pET12b or pET21b vectors (Novagen) in E. coli strain X90 (lDE3) slyD::kan hslUV::tet and pro-
teins were purified as described (Baytshtok et al., 2016).
Crystals of HslU were grown at 4
C by the hanging-drop method after mixing 0.5 mL of selenomethionine-labeled H6-tagged HslU
(15 mg/mL) in 15 mM Tris (pH 7.5), 100 mM NaCl, 20 mM MgCl2, 0.2 mM EDTA, and 5 mM ATPgS with an equal volume of well so-
lution containing 100 mM Tris (pH 7.5), 18% PEG-3350, and either 0.1 M (PDB code 6PXK) or 0.2 M (PDB code 6PXL) ammonium
sulfate. For cryo-protection, PEG-3350 was increased to 20% and 10% MPD was added to the base-well solution. HslUV crystals
were grown in the same way except the protein solution contained E257QHslU (10 mg/mL) plus HslV (9.6 mg/mL) and the well solution
contained 0.1 M Bis-Tris (initial pH 5.5), 1.85 M ammonium sulfate, 5% glycerol, and 5 mM ATPgS. For cryo-protection, glycerol was
increased to 25%. Diffraction data for the HslUV crystals were collected on our home source (Rigaku MicroMax-007HF with a Saturn
944 detector) and for the HslU crystals at the Advanced Photon Source (APS) beamline 24-ID-E and processed using HKL2000 (Ot-
winowski and Minor, 1997) or XDS (Kabsch, 2010). Structures were solved by molecular replacement with PHASER (McCoy et al.,
2007) using search models consisting of HslU hexamers alone or HslUV (PDB code 5JI3; Baytshtok et al., 2016). Anomalous differ-
ence maps – showing the selenomethionine side chains of residues 187, 192, 195, and 202 – were used to assist model building into
plug density. Phenix was used to refine structures (Adams et al., 2010), Coot was used for model building (Emsley et al., 2010), and
MolProbity was used to assess model geometry (Chen et al., 2010).
Negative-stain EM experiments were performed as described (Baytshtok et al., 2016), using WT WTHslU or L199QHslU (1.5 mM) and
ATPgS (5 mM) in 20 mM HEPES (pH 7.5), 5 mMMgCl2, 500 mM NaCl, 10% glycerol (v/v), and 0.032% Igepal CA-630. Particles with
tilted six-fold symmetry axis were removed by two rounds of 2D classification. For WTHslU, 10927 particles from 23micrographswere
used to generate representative 2D class averages; for L199QHslU, 5567 particles from 20 micrographs were used. The e2pdb2mrc
and e2project3d utilities in EMAN2 (Tang et al., 2007) were used to generate 2D projections from PDB files and Relion 3.0.8 (Scheres,
2012) was used for 2D classification.
Degradation and ATPase assays were performed in 25 mM HEPES (pH 7.5), 5 mM KCl, 20 mMMgCl2, 10% glycerol, and 0.032%
Igepal CA-630 as described (Baytshtok et al., 2016). ATPase rates at different temperatures were measured using a Spectramax M5
plate reader (Molecular Devices) by using an NADH-coupled assay (Nørby, 1988) with or without 50 mM I37AArc-cp6GFP-st11-ssrA.
Using the Solver tool of Microsoft Excel, the temperature dependencies of ATPase rates were globally fitted to the equation max/
(1+exp(DH/R$(1/TM-1/T))), where max is the maximum ATPase rate, T is the temperature in Kelvin, TM is the temperature at 50% ac-
tivity, DH is the enthalpy at TM, and R is the universal gas constant. Degradation of
I37AArc-cp6GFP-st11-ssrA by WTHslUV or
I186NHslUV was monitored by loss of fluorescence (excitation 467 nm; emission 511 nm). For degradation of Arc-st11-ssrA (5 mM)
by wild-type or mutant enzymes (0.3 mMHslU6, 0.9 mMHslV12), proteins were preincubated for 1 min at 25

C or 55
C before adding
an ATP regeneration system (2.5 mMATP, 7.5 mM phosphoenolpyruvate, 1 mMNADH, 18.8 U/ml pyruvate kinase, 21.5 U/ml lactate
dehydrogenase) to initiate degradation. Aliquots were removed at different times, quenched by addition of Tricine sample buffer (Bio-
Rad) with b-mercaptoethanol, and flash frozen in liquid nitrogen. Samples were then boiled for 5 min and electrophoresed on 15%
Tris-Tricine SDS-PAGE gels. After electrophoresis, gels were stained for 10 min using a Coomassie-blue solution, destained over-
night in water, scanned using a Typhoon FLA 9500 Imager (GE Healthcare), and the amount of substrate remaining at each time point
was quantified using ImageQuant TL (GE Healthcare).
QUANTIFICATION AND STATISTICAL ANALYSIS
Crystallographic unit-cell parameters, data completeness and redundancy, CC1/2 values, and merging R-factors (Table 1) were ob-
tained using HKL2000 (Otwinowski and Minor, 1997), which was used to index, integrate, and scale the diffraction data. Model-
refinement statistics (Rwork,Rfree; Table 1) were obtained from the phenix.refine program in Phenix (Adams et al., 2010) and are shown
in Table 1. Statistics for the quality of the final crystallographic models (Table 1) were determined usingMolProbity (Chen et al., 2010).
Biochemical experiments were performed using a minimum of three independent replicates. Biochemical data were plotted as av-
erages ± one standard deviation (SD), as described in the Figure 3 legend, using Prism (GraphPad). The errors for KM and Vmax (Fig-
ure 3E) were calculated by non-linear least-squares fitting to the Michealis-Menten equation using Prism (GraphPad).e2 Cell Reports 34, 108639, January 19, 2021