Mutation of glu-166 blocks the substrate-induced dimerization of SARS coronavirus main protease

Article abstract of DOI:10.1016/j.bpj.2009.12.4272

The maturation of SARS Coronavirus involves the autocleavage of polyproteins 1a and 1ab by the main protease (Mpro) and a papain-like protease; these represent attractive targets for the development of anti-SARS drugs. The functional unit of Mpro is a hom

Full text of DOI:10.1016/j.bpj.2009.12.4272

Biophysical Journal Volume 98 April 2010 1327–1336
1327
Mutation of Glu-166 Blocks the Substrate-Induced Dimerization of SARS
Coronavirus Main Protease
Shu-Chun Cheng, Gu-Gang Chang, and Chi-Yuan Chou*
Department of Life Sciences and Institute of Genome Sciences, National Yang-Ming University, Taipei, Taiwan, Republic of China
ABSTRACT The maturation of SARS coronavirus involves the autocleavage of polyproteins 1a and 1ab by the main protease
(Mpro) and a papain-like protease; these represent attractive targets for the development of anti-SARS drugs. The functional unit
of Mpro is a homodimer, and each subunit has a His-41/Cys-145 catalytic dyad. Current thinking in this area is that Mpro dimer-
ization is essential for catalysis, although the influence of the substrate binding on the dimer formation has never been explored.
Here, we delineate the contributions of the peptide substrate to Mpro dimerization. Enzyme kinetic assays indicate that the mono-
meric mutant R298A/L exhibits lower activity but in a cooperative manner. Analytical ultracentrifugation analyses indicate that in
the presence of substrates, the major species of R298A/L shows a significant size shift toward the dimeric form and the mono-
mer-dimer dissociation constant of R298A/L decreases by 12- to 17-fold, approaching that for wild-type. Furthermore, this
substrate-induced dimerization was found to be reversible after substrates were removed. Based on the crystal structures,
a key residue, Glu-166, which is responsible for recognizing the Gln-P1 of the substrate and binding to Ser-1 of another protomer,
will interact with Asn-142 and block the S1 subsite entrance in the monomer. Our studies indicate that mutation of Glu-166 in the
R298A mutant indeed blocks the substrate-induced dimerization. This demonstrates that Glu-166 plays a pivotal role in connect-
ing the substrate binding site with the dimer interface. We conclude that protein-ligand and protein-protein interactions are closely
correlated in Mpro.
INTRODUCTION
Severe acute respiratory syndrome (SARS) is an emerging
infectious disease of this century and is caused by a novel
coronavirus (CoV) termed SARS-CoV (1). During the
outbreak in 2003, this virus infected >8000 people and the
fatality rate in humans was as high as 15% (World Health
Organization). In 2004 and 2005, the discovery of two
more species of CoV that infect humans, NL-63 and
HCoV-HKU1, confirm the high mutating rate and genetic
recombination within Coronaviridae (2,3). The fact that the
virus is easily transmitted between humans makes the ree-
mergence of SARS and other human CoVs a distinct possi-
bility and has resulted in an urgent need to understand these
viruses and their coding proteins.
Mpro is a homodimer in which the two subunits are ar-
ranged perpendicular to each other. Each protomer
comprises three distinct structural domains. The first two
domains (residues 8–101 for domain I and residues 102–
184 for domain II) have an antiparallel b-barrel structure,
which forms a folding scaffold similar to other viral chymo-
trypsinlike proteases (7,12,13). Each subunit has its own
substrate binding site consisting of a His-41$$$Cys-145 cata-
lytic dyad located at the interface between domains I and II
(Fig. 1). An oxyanion hole is formed by the main-chain
amides of Gly-143, Ser-144, and Cys-145 (6). Interestingly,
Mpro contains an extra domain (III), which consists of five
a-helices (residues 201–306), and this is a specific feature
of CoV Mpro (6–9).
SARS-CoV main protease (Mpro) cleaves the virion poly-
proteins (pp1a and 1b) at 11 sites that contain the canonical
L-Q-Y-(A/S) sequence (4,5). Mpro was the first of the
SARS-CoV proteins to have its three-dimensional structure
solved by crystallography (6). Recently, structures of other
CoV Mpros such as TGEV, IBV, and HCoV-HKU1 have
also been solved (7–9). Although the overall sequence iden-
tity of these Mpros is only 40–50%, the three-dimensional
structures, especially the regions of dimer interface, catalytic
dyad, and substrate binding site, are highly conserved (10).
Therefore, the design of broad-spectrum inhibitors of Mpro
appears to be feasible for drug development (11).
The catalytic N-terminal domain and C-terminal domain
III can fold independently. The N-terminal domain is
a monomer that is able to fold correctly but is catalytically
inactive (14). The extra domain III increases the structural
stability of the catalytic domain by increasing the folding
rate. Furthermore, domain III is involved in the dimerization
of Mpro, which has important functional implications for
this enzyme (15). The side chain of Arg-4 at the N-finger
(residues 1–7) of protomer A fits into a pocket of protomer
B and forms a salt bridge with Glu-290; this constitutes
one of the major interactions between the two subunits
(16). Our previous studies have shown that N-terminal trun-
cation of the whole N-finger results in almost complete loss
of enzymatic activity (17). Critical N-terminal amino acid
residues up to Arg-4 and C-terminal residues up to Gln-
299 have been identified as involved in dimerization and
thus in generating the correct conformation of the active
Submitted October 20, 2009, and accepted for publication December 7,
2009.
*Correspondence: cychou@ym.edu.tw
Editor: Patrick Loria.
Ó 2010 by the Biophysical Society
0006-3495/10/04/1327/10 $2.00
doi: 10.1016/j.bpj.2009.12.4272

1328
Cheng et al.
MATERIALS AND METHODS
Preparation of recombinant SARS-CoV Mpro
The construction of the expression plasmids for wild-type Mpro, R298A,
R298L, and R298A/Q299A mutants have been described previously
(16,18). For the site-directed mutagenesis (22) of E166A, the forward primer
was 50-tgctatatgcatcatatggcgcttccaacaggagtacac and the reverse primer was
50-gtgtactcctgttggaagcgccatatgatgcatatag. The polymerase chain reaction
products were then treated with DpnI and transformed into Escherichia
coli cells and checked by autosequencing. Next, the plasmids were trans-
formed into E. coli strain BL21 (DE3) cells. The cells were grown in
Luria-Bertani medium with 50 mg/ml Kanamycin at 37ꢀC. After 3 h incuba-
tion, the cells were induced overnight by 0.4 mM isopropyl-1-thio-b-D-
galactoside at 20ꢀC. For purification, all operations were performed at
4ꢀC. The cells were centrifuged at 6000 ꢁ g for 10 min. The supernatant
was removed and the pellets were resuspended in the lysis buffer (20 mM
Tris-Cl, 250 mM NaCl, 2 mM b-mercaptoethanol (b-ME), and 0.2% Triton
X-100, pH 8.5). The cells were then sonicated using sixty 10-s bursts at
300 W with a 10-s cooling period between each burst. The cell debris was
removed by centrifugation (10,000 ꢁ g for 25 min). Lysis buffer-equili-
brated Ni-NTA slurry (Qiagen, Hilden, Germany) (1 ml) was then added
to the soluble lysate, followed by gentle mixing for 1 h. The mixture was
then loaded into a column and washed with the washing buffer (20 mM
Tris-Cl, 250 mM NaCl, 8 mM imidazole, and 2 mM b-ME, pH 8.5). Finally,
the protease was eluted with elution buffer (20 mM Tris-Cl, 30 mM NaCl,
150 mM imidazole, and 2 mM b-ME, pH 8.5).
FIGURE 1 Active center of the SARS-CoV Mpro. The interactions
between the P1 substrate-binding subsite from chain A (cyan) and the
N-finger and domain III from chain B (magenta) of SARS-CoV Mpro
(PDB code 1UK4) are shown. The substrate analog and the side chain of
the Gln-P1 residue are yellow. The R298A monomeric structure (green)
(PDB code 2QCY) is superimposed on the wild-type chain A. The dashed
lines indicate hydrogen bonds for 1UK4 (red) and 2QCY (black). This figure
was produced using PyMOL (35).
The purified enzyme was concentrated and the buffer replaced using phos-
phate-buffered saline (20 mM sodium phosphate buffer, 150 mM NaCl, and
2 mM b-ME, pH 7.6) and an Amicon Ultra-4 centrifugal filter with a mass
cutoff at 10 kDa (Millipore, Bedford, MA). Samples from this purification
step were subjected to SDS-PAGE to check for homogeneity (>95% purity).
Typical yields of the enzyme after purification were 5–10 mg from 1 liter of
E. coli culture medium.
site (17,18). In addition, the interactions between the two
helices A0 (residues 11–15) and the S1 substrate-binding
subsite, consisting of Phe-140, His-163, Met-165, Glu-166,
and His-172, are also regarded as major components of the
dimer interface (19,20).
Enzyme kinetic assay
The enzymatic activity of Mpro was measured by a colorimetry-based
peptide cleavage assay involving the 6-mer peptide substrate TSAVLQ-
para-nitroanilide (TQ6-pNA) (purity 95–99% by high-performance liquid
chromatography; GL Biochem, Shanghai, China) (23). This substrate is
cleaved at the Gln-pNA bond to release free pNA, which turns the solution
color to yellow. The increase in absorbance at 405 nm was continuously
monitored using a Jasco V-550 UV/VIS spectrophotometer (Tokyo, Japan).
The amount of pNA released from the proteolysis can be calculated with
a standard curve generated by analytical-grade pNA, which is consistent
with the absorbance reported in the literature (A405 nm ¼ 9.8 at 1 mM) (24).
The protease activity assay was performed in 10 mM phosphate (pH 7.6)
at 30ꢀC. The substrate stock solution was 1 mM and the working concentra-
tions were 5–950 mM. In the substrate titration assay, the concentrations of
Mpro, E166A, and R298A/L mutants were 1, 5.7, and 2.6 mM, respectively.
The steady-state enzyme kinetic parameters were obtained by fitting the
initial velocity (n0) data to the Michaelis-Menten equation (Eq. 1):
Compared with functional dimeric Mpro, the crystal struc-
tures of monomeric Mpro (G11A, S139A, and R298A) have
provided direct structural evidence for the catalytic incompe-
tence of the dissociated monomer (19–21). In the monomer
mutants, the oxyanion loop (Ser-139 to Leu-141) is con-
verted into a short 310-helix and completely collapses
inward, as exemplified by the large movement of Asn-142
and Leu-141. The slipped Asn-142 interacts with the side
chain of Glu-166 and blocks entry to the S1 subsite; this
results in enzyme inactivation (Fig. 1) (20). All experimental
results have indicated that the dimerization of Mpro is essen-
tial for catalysis. Here, we provide what we believe is a novel
approach that shows the influence of substrate binding on
Mpro dimerization. Various biochemical and biophysical
techniques are used to demonstrate the importance of
substrate-induced dimerization to the catalytic mechanism
of Mpro. By mutagenesis studies, a key residue, Glu-166,
is found to play a connecting role between the substrate
binding site and the dimer interface. We believe that this
study will help deepen our understanding of the correlation
between protein-ligand and protein-protein interactions in
Mpro. Such an understanding will aid the development of
new approaches to control SARS-CoV and other CoVs.
kcat½Eꢂ½Sꢂ
n0 ¼
;
(1)
Km þ ½Sꢂ
where kcat is the catalytic constant, [E] is the enzyme concentration, [S] is the
substrate concentration, and Km is the Michaelis constant of the substrate.
The program SigmaPlot (Systat Software, Richmond, CA) was used for
the data analysis.
For the cooperativity effect, the kinetic parameters were obtained by
fitting the initial velocities to the Hill equation (Eq. 2):
h
kcat½Eꢂ½Sꢂ
n0 ¼
;
(2)
h
K0 þ ½Sꢂ
Biophysical Journal 98(7) 1327–1336

Substrate-Induced Dimerization of Mpro
1329
where K0 is a constant that relates to the dissociation constant and h is the
Hill constant.
of enzyme was performed in parallel to correct for the dilution of heat. The
data were then analyzed by integrating the heat effects normalized to the
amount of injected proteins using curve fitting based on a 1:1 binding model.
This involved the use of Digitam software (TA instruments).
The dependence of the proteolytic activity on enzyme concentration was
investigated for the wild-type Mpro and R298A/L mutants, at a substrate
concentration of 600 mM TQ6-pNA. The initial velocity of the reaction at
various concentrations of each enzyme was determined and fitted to the
nonlinear dependence equation (Eq. 3):
RESULTS AND DISCUSSION
ꢀ
ꢁ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
.
Cooperative effect of initial velocity curves
of R298A and R298L mutants
n0 ¼ kcat Kd þ 4½Eꢂ ꢃ Kd2 þ 8Kd½Eꢂ
where Kd is the monomer-dimer dissociation constant.
Analytical ultracentrifugation analysis
8;
(3)
To measure the activity of Mpro and its mutants, we used the
6-mer substrate peptide (TSAVLQ) attached to a pNA group
(23). This peptide is specifically cleaved by SARS-CoV
Mpro at the designated site (Gln-pNA) to release free
pNA, which results in an increased absorbance at 405 nm.
Besides wild-type Mpro as a dimeric target, two single
mutants, R298A and R298L, and one double mutant,
R298A/Q299A, were chosen as the monomeric targets. Ac-
cording to previous studies (19), mutation of R298 or Q299
will induce dimer dissociation and result in an ~10-fold
decrease in proteolytic activity (18). When R298 and Q299
are both mutated, the enzyme activity decreases by 100-fold.
Other studies also suggest that the monomeric Mpro by other
mutations shows very low or no proteolytic activity
(16–21,31). Indeed, in this study, we were not able to detect
any enzyme activity associated with the R298A/Q299A
double mutant. However, for the single mutants, R298A or
R298L, the initial velocity pattern at various substrate
concentrations displayed a sigmoid curve (Fig. 2, B and C).
The dimeric Mpro, on the other hand, exhibited a classical
saturation curve (Fig. 2 A). These results were then fitted
to the Michaelis-Menten or Hill equations to evaluate the
kinetic parameters. The best-fit results are shown in Table 1.
The Km (223 mM) and kcat (0.63 sꢃ1) of wild-type Mpro with
TQ6-pNA substrate are close to observations from other
laboratories (23,32). After fitting to the Hill equation, the
kcat of R298A was sixfold lower than that of the wild-type
enzyme. In contrast, R298L showed a kcat close to that of
the wild-type enzyme. The Hill constants of R298A and
R298L were 2.0 and 1.8, respectively. This nonunity number
suggested that there is a strong positive cooperativity among
the Mpro protomers. However, since the cooperativity
phenomenon is not compatible with a monomeric form, we
sought other evidence to examine the possibility of dimeric
Mpro formation during the catalytic process.
The analytical ultracentrifugation (AUC) experiments were performed on an
XL-A analytical ultracentrifuge (Beckman, Fullerton, CA) with an An-50 Ti
rotor (17). The sedimentation velocity (SV) experiments were performed in
a double-sector epon charcoal-filled centerpiece at 20ꢀC with a rotor speed
of 42,000 rpm. The sample (330 ml) and reference (370 ml) solutions with or
without different concentrations of TQ6-pNA substrate were loaded into the
centerpiece. We found that the TQ6-pNA was cleaved and free pNA was
released in the process of centrifugation (detected by absorbance at
405 nm). The absorbance spectrum of free pNA interfered with protein
absorbance at 280 nm. Therefore, absorbance at 250 nm was chosen to
detect the protein, which was monitored in a continuous mode with a time
interval of 480 s and a step size of 0.003 cm. Three different protein concen-
trations (from 1.4 to 57.2 mM) were used to estimate the dynamic monomer-
dimer content. Multiple scans at different time intervals were then fitted to a
continuous c(s) distribution model using the SEDFIT program (25,26). The
partial specific volume of Mpro, the solvent density, and the viscosity were
calculated by SEDNTERP (http://www.jphilo.mailway.com/download.htm)
(cited Oct. 20, 2009).
The sedimentation equilibrium (SE) experiments were performed in a six-
channel centerpiece. Three different samples (0.10–0.12 ml) were loaded
into the sample channels and 0.11–13 ml buffers were loaded into the refer-
ence channels. The cells were then loaded into the rotor and run at multi-
speeding (8000, 12,000, and 15,000 rpm), each for 12 h at 20ꢀC. Ten scans
of absorbance at 250 nm at time intervals of 10 min were measured for every
rotor speed to check the status of SE. In our studies, all Mpro and its mutants
were able to achieve equilibrium state after 12 h. The SV results at three
protein concentrations and the multispeed SE data were then globally
analyzed using a monomer-dimer equilibrium model by the SEDPHAT
program (27), which gives a precise measurement for Kd and the dissociation
rate constant (koff) (18,28,29).
Analytical size-exclusive chromatography
Size-exclusive chromatographic experiments were performed using a GE
¨
Healthcare AKTA purifier system (Pittsburgh, PA) with a Superose 12
(10/300) column preequilibrated with phosphate-buffered saline (pH 7.6).
Mpro and its mutants without or with 600 mM of substrate preincubation
for 20 min were injected into the column separately. The elution was carried
out at a flow rate of 0.5 ml/min and the absorbance at 280 nm was monitored
continuously.
Nonlinear dependence of initial velocity
on protease concentration
Isothermal titration calorimetry
The dependence of the initial velocity on protease concentra-
tion was analyzed (Fig. 2, D–F). If the monomeric Mpro has
an activity identical to that of the dimeric one, a linear pattern
should be obtained (28). However, a nonlinear positive
correlation was observed for the wild-type and R298A/L
monomeric mutants. After fitting to the nonlinear depen-
dence equation (Eq. 3), the kcat (Table 1) and Kd for the
monomer-dimer equilibrium (Table 2) were calculated.
The isothermal titration calorimetry (ITC) protocol followed was that of
Sondermann et al. (30) with some modifications. Apparent dissociation
constants and stoichiometry of the enzyme-ligand interactions were
measured by a thermal activity monitor (2277, TA instruments, New Castle,
DE). Calorimetric titrations of the peptide substrate TQ6-pNA (1 mM in
a 250-ml syringe) and Mpro (5.7 mM for wild-type and 28.6 mM for mutants
in a 4-ml ampoule) were carried out at 25ꢀC in 10 mM phosphate buffer
(pH 7.6). The peptides were titrated into the enzyme in 10-ml aliquots per
injection with a time interval of 20 min. A control experiment in the absence
Biophysical Journal 98(7) 1327–1336

1330
Cheng et al.
FIGURE 2 Initial velocity patterns of Mpro (A–C) and
dependence of the initial velocity of Mpro on enzyme
concentration (D–F). (A–C) Plots of the velocity difference
at various substrate concentrations for wild-type, R298A,
and R298L mutants, respectively. The lines represent
results fitted according to the Michaelis-Menten equation
(Eq. 1) for wild-type and the Hill equation (Eq. 2) for
mutants. The kinetic parameters are shown in Table 1.
(D–F) Difference in velocity at various enzyme concentra-
tions for wild-type, R298A, and R298L mutants, respec-
tively. The concentration of substrate was 600 mM. The
line represents the best fit to the nonlinear dependence
equation (Eq. 3).
The kcat for wild-type and the mutants did not show any
significant difference, whereas the Kd values for R298A/L
were increased from 14- to 17-fold, indicating that the
R298A/L mutants are more easily dissociated. The nonlinear
dependence detected here suggests that proteolytic activity is
actually related to the dimer content, which in turn is influ-
enced by the protein concentration. Based on this, we next
performed AUC experiments on Mpro with or without the
presence of peptide substrates to evaluate the quaternary
structural changes during the catalytic process.
Dimeric R298A and R298L mutants in the
presence of peptide substrates can be
detected by AUC
Interestingly, wild-type Mpro and the R298 mutants dis-
played different monomer-dimer equilibrium with or without
substrates. A typical SV experiment of AUC is shown in
Fig. 3 A. The cumulative spectra were analyzed using the
continuous size distribution model (26), which shows the
quaternary structure distribution and the sedimentation coef-
ficients (S) (Fig. 3, B–E). Although wild-type Mpro main-
tained a stable dimeric form, the R298A/Q299A double
mutant existed exclusively in monomeric form irrespective
of the presence of high concentrations of substrate (Fig. 3,
B and E). Here, we confirmed that the monomeric and
dimeric Mpro sedimented at 2.9 and 4.1 S, respectively.
These values are consistent with previous observations
(16,18,21).
TABLE 1 Kinetic parameters of Mpro and its mutants
k
cat (sꢃ1) by
nonlinear
fittingy
k
cat/Km
Protein
Km (mM)*
kcat (sꢃ1)*
(sꢃ1 Mꢃ1
)
Wild-type 222.6 5 19.0 0.63 5 0.02
E166A
2830 5 303
877 5 132
1.4 5 0.2
NDz
353.4 5 48.0 0.31 5 0.01
K0 (104 mM)
kcat (sꢃ1
)
h
R298A
R298L
3.9 5 2.3
9.5 5 6.2
0.10 5 0.004
0.48 5 0.06
2.0 5 0.1
1.8 5 0.1
0.7 5 0.4
2.3 5 0.9
*Data of wild-type Mpro and E166A mutant were fitted to the Michaelis-
Menten equation (Eq. 1) and the Rsqr values were 0.994 and 0.987, respec-
tively. Those of mutants were fitted to the Hill equation (Eq. 2), and the Rsqr
values were 0.998 and 0.996, respectively. All the assays were repeated
several times to ensure reproducibility.
However, a different story was found for the R298 single
mutants. In the presence of substrates, the R298A and R298L
mutants showed a significant shifting of the major species
(Fig. 3, C and D). At 600 mM TQ6-pNA, the major species
yCatalytic constant was calculated by the nonlinear dependence equation
(Eq. 3). The best-fitting Rsqr values for wild-type, R298A, and R298L
were 0.995, 0.99 and 0.993, respectively.
zNot determined.
Biophysical Journal 98(7) 1327–1336

Substrate-Induced Dimerization of Mpro
1331
TABLE 2 Dissociation of Mpro and its mutants with and without substrates
AUC analysis
Nonlinear fitting
No substrate
With 600 mM substrate
Kd (mM)y
Protein
Kd (mM)*
Kd (mM)y
k
off (sꢃ1 y
)
koff (sꢃ1 y
)
Wild-type
R298A
0.8 5 0.4
11.4 5 9.7
13.8 5 8.7
NDc
2.0 5 0.01
81.1 5 3.3
30.8 5 0.4
85.7 5 3.1
3.7 5 0.2
0.1 5 0.007
0.1 5 0.004
0.1 5 0.001
0.1 5 0.004
0.04 5 0.001
0.1 5 0.009
1.7 5 0.03
4.7 5 0.1
2.6 5 0.1
0.1 5 0.002
0.08 5 0.001
0.1 5 0.003
0.05 5 0.002
0.1 5 0.002
0.06 5 0.005
R298L
R298A/Q299A
E166A
115.0 5 4.4
0.4 5 0.01
103.7 5 8.3
ND
E166A/R298A
ND
141.1 5 12.7
*Kd was calculated by nonlinear dependence equation (Eq. 3).
yParameters were derived from a global fit of the SV and SE data to a monomer-dimer self-association model by SEDPHAT. The SE experiments for the assay
were obtained at protein concentrations of 5.7–28.6 mM, whereas SV experiments were at 1.4–57.2 mM.
of R298A/L shifted from 2.9 S to 3.8–3.9 S. The distribution
now became a broad peak between the monomer and dimer,
but closer to the dimer. In a fast dissociated and associated
system, a broad size distribution will appear between the
monomer and dimer when the protein concentration is close
to the Kd (27). Thus, SARS-CoV Mpro acts as a rapid mono-
mer-dimer self-association equilibrium system. Furthermore,
we also measured the size distribution of R298A at various
substrate concentrations. The results clearly indicated that
the distance of major species shifting was related to the
substrate dosage (Fig. 4 A). More substrates resulted in
more of the dimer form and the major species became closer
and closer to the dimer. The trend was a linear positive corre-
lation (Fig. 4 B). Therefore, we conclude that the presence of
substrate can induce a quaternary structural change in the
R298A/L mutants. The system described in this article
prevents any further monomer-dimer equilibrium analysis
of the broad size distribution among the monomer and dimer
species. We therefore performed the next set of experiments
to further delineate the monomer-dimer equilibrium.
We next analyzed the size distribution of the R298A mutant
at various substrate concentrations in D2O (Fig. S2 A). The
results clearly indicate that the increase in dimeric form
content is closely correlated with the substrate concentration
(Fig. S2 B).
In summary, the measurements obtained by the SV exper-
iments proved that the R298A/L mutants are able to associate
into a dimeric form during the catalytic process. This
substrate-induced dimerization is dose-dependent.
Dramatic decrease in dissociation constant
in the presence of peptide substrate
To quantitatively characterize the monomer-dimer equilib-
rium of Mpro and its mutants, SV experiments were carried
out at the three different enzyme concentrations with or
without 600 mM TQ6-pNA. In addition, SE experiments
were executed. The three SV and one SE datasets were then
globally fitted to the monomer-dimer self-association model
by SEDPHAT to estimate the Kd and koff. The best-fit results
are shown in Table 2. The koff values for wild-type Mpro and
its three mutants were all close to 0.1 sꢃ1, which indicates that
the monomer-dimer exchange is instantaneous (27). The Kd
values of wild-type Mpro without and with substrates were
2.0 and 1.7 mM, respectively, when those for the R298A/
Q299A double mutant were 85 and 115 mM. Not surprisingly,
the Kd of the R298A/L mutants was dramatically decreased,
by 12- to 17-fold, in the presence of substrate, approaching
the values for the wild-type (Table 2). The Kd rescued by
substrates suggested that the R298A/L mutants display
a protein association comparable to that of the wild-type in
the presence of substrates. On the other hand, by nonlinear
dependence fitting, the Kd of wild-type in the presence of
substrates is 2.5-fold lower than in the absence of substrates
(Table 2). The influence of substrates on the dimerization of
wild-type is minor because of the stronger association
between its protomers. In contrast, the R298A/Q299A double
mutant lost two important ion pairs (Ser-123 (A)$$$Arg-298
(B) and Ser-139 (A)$$$Gln-299 (B)) in the dimer interface
(18) and results in complete dissociation, even in the presence
of substrates.
Monomer and dimer distribution of R298A/L
mutants can be clearly delineated in deuterium
water
To clearly separate the monomeric and dimeric signals, the
speed of the protein self-association needs to be slowed
down. To do this, we performed the same SV experiments
in deuterium water (D2O), which has a higher density
(Fig. S1 A in the Supporting Material). All the buffer solu-
tions and substrate stocks were prepared in D2O to give
the same ionic strengths. In the D2O environment, the mono-
meric and dimeric Mpro sedimented at 1.9 and 2.6 S, respec-
tively (Fig. S1, B and E). For the R298A mutant, there was a
single monomeric form in the absence of substrate and a
single dimeric form in the presence of substrate (Fig. S1 C).
In addition, under the same conditions, both monomeric and
dimeric forms of R298L could be observed (Fig. S1 D). This
demonstrates that the high density of D2O is able to slow
down the exchange rate within the monomer-dimer equilib-
rium and clearly separate the signals from the two species.
Biophysical Journal 98(7) 1327–1336

1332
Cheng et al.
FIGURE 4 Effects of substrate concentration on the quaternary structure
of the R298A mutant. (A) Continuous (c(s)) distribution of the R298A
mutant at TQ6-pNA concentrations of 0 (solid circles), 20 (open circles),
120 (solid triangles), 250 (open triangles), 400 (solid squares), and
600 mM (open squares). (B) Sedimentation coefficient shift of R298A major
species at different TQ6-pNA concentrations.
Dimerization induced by peptide substrate
is reversible
To further evaluate whether the substrate-induced dimeriza-
tion is a dynamic equilibrium or a permanent, irreversible
process, we performed size exclusion chromatography on
Mpro and its mutants. The protease was preincubated with
or without 600 mM TQ6-pNA for 20 min and then injected
into the Superose 12 column (Fig. 5). The running buffer
was substrate-free. The results showed that after removing
the substrate, wild-type Mpro still maintained a dimeric
form, but the three mutants were all monomeric. This
suggests that substrate-induced dimerization is a reversible
process. When the substrate leaves, the R298A/L dimers
dissociate back to being a monomer. A similar result was
obtained when the substrates were removed by the HiTrap
desalting column (GE Healthcare, Pittsburgh, PA) and then
analyzed by AUC (Fig. S3).
FIGURE 3 SV patterns of Mpro. (A) Typical trace of absorbance at
250 nm of the enzyme during the SV experiment. Symbols represent exper-
imental data and lines the results fitted to the Lamm equation using the
SEDFIT program (25,26). (B–E) Continuous (c(s)) distribution of wild-type,
R298A, R298L, and R298A/Q299A mutants, respectively. The residual
bitmaps are shown in the insets. Protein concentration was 1.0 mg/ml.
The distributions in 10 mM phosphate buffer are shown by solid lines and
those in 10 mM phosphate containing 600 mM TQ6-pNA substrate by
dashed lines. The left vertical dotted line indicates the monomer, and the
right the dimer position.
Biophysical Journal 98(7) 1327–1336

Substrate-Induced Dimerization of Mpro
1333
type. Although this might be an artifact, because the enzy-
matic hydrolysis during the titration might produce addi-
tional heat that will contribute to the overall DH, the DH,
N, and Kd values from our ITC observations still suggest
that the wild-type Mpro and its mutants, especially the
monomeric R298A/Q299A mutant, have comparable sub-
strate-binding affinities.
Mechanism for substrate-induced dimerization
of SARS-CoV Mpro
On the basis of the solved monomeric structures (PDB codes
2QCY, 2PWX, and 3F9E), the loop of Ser-139 to Leu-141 is
able to form a 310-helical structure, which results in an Asn-
˚
142 shift by 5.8 A (20). This shift helps the side-chain amide
of Asn-142 to accept hydrogen bonding from the side chain
of Glu-166 (Fig. 1). In the dimeric Mpro (PDB code 1UK3),
the side chain of Glu-166 interacts with the main-chain
amide of Ser-1 from protomer B (6). In the substrate-bound
form (PDB code 1UK4), in protomer A, Gln-P1 enters the S1
binding pocket and its side-chain amide and carbonyl donate
a hydrogen bond to the side chains of Glu-166 and His-163,
respectively. Comparing these structures, the Glu-166
residue should play a distinct role. In the dimer, Glu-166
binds to the Gln-P1 of the substrate and the N-finger of pro-
tomer B. In the monomer, it shifts to interact with Asn-142,
which is able to stabilize the 310-helix of Ser-139 to Leu-141.
To obtain direct evidence for the role of Glu-166, we
mutated and purified E166A and E166A/R298A mutants
and then performed the kinetic and quaternary structure anal-
yses. The E166A single mutant had an ~3-fold loss in
enzyme activity based on kcat/Km (Table 1), whereas the
enzyme activity of the E166A/R298A double mutant was
not detectable. Furthermore, the experimental data from
ITC for the E166A point mutant and the E166A/R298A
double mutant suggest that these two mutants have compa-
rable substrate binding affinities (Fig. S4). This indicates
that without Glu-166, the substrate can still bind to the active
site. On the other hand, the SV experiments suggest that
E166A is a dimer with or without substrates (Fig. 7 A).
Furthermore, the Kd of E166A was 3.7 mM and decreased
to 0.4 mM in the presence of substrates, which was very close
to the Kd for wild-type (Table 2). It is reasonable in that Glu-
166 (A)$$$Ser-1 (B) is not the main force stabilizing the
dimer interface. However, when Glu-166 and Arg-298
were both mutated, the dimerization was not detected, and
the Kd did not decrease in the presence of substrates
(Fig. 7 B and Table 2). The results suggest that mutation
of Glu-166 in R298A mutant will block the substrate-
induced dimerization. This demonstrates that Glu-166 plays
a pivotal role in connecting the substrate binding site with the
dimer interface. The highly conserved Glu-166 in most
CoVs, including SARS CoV, MHV, TGEV, HCoV, and
IBV (8), implies that involvement of this residue in the pro-
tomeric communication may be universal in all virion Mpro.
FIGURE 5 Size-exclusion chromatography of Mpro. The enzyme in
phosphate-buffered saline buffer was applied to a preequilibrated Superose
12 10/300 GL column. The Mpro and its mutants, without or with preincu-
bation in 600 mM substrate, are wild-type (solid circles), wild-type with
substrate (open circles), R298A (solid triangles), R298A with substrate
(open triangles), R298A/Q299A (solid squares), and R298A/Q299A with
substrate (open squares).
ITC assay showed that both dimeric and
monomeric Mpro can bind to the peptide
substrates
Based on the structure of the SARS-CoV Mpro with its
substrate analog (PDB code 1UK4), the substrate can bind
by different modes to the active and inactive protomer (6).
With the active protomer, the side chain of Gln-P1 accepts
hydrogen bonding from the side chain of His-163 and Glu-
166, resulting in specificity of Gln-P1 for the S1 subsite
(Fig. 1). On the other hand, with the inactive protomer, the
S1 pocket is not opened and thus does not allow the Gln-
P1 to enter. Instead, Leu-P2 and Ser-P4 bind to the appro-
priate specific pocket. These differences imply that the
substrate binding site is flexible and can undergo a significant
conformational change.
To further delineate the binding of substrate to the Mpro
and its mutants, we employed ITC to measure the dissocia-
tion constant (Kd for the substrate-enzyme complex), the
binding enthalpy change (DH), and the binding stoichiom-
etry (N) of the substrate-enzyme interaction. In the ITC
experiments, 1 mM TQ6-pNA was injected into the protein
solution and curve fitting based on a 1:1 binding model
was carried out (Fig. 6). Compared with wild-type Mpro
(Fig. 6 A), all three mutants (Fig. 6, B–D) exhibited a similar
N and Kd. The N values are close to 1, which means that one
substrate may bind one enzyme molecule. The Kd of
substrate to wild-type Mpro was 34 mM, whereas those of
substrate to R298A, R298L, and R298A/Q299A mutants
were 81, 114, and 30 mM, respectively. However, the DH
values for substrate binding to Mpro mutants were decreased
by three- to sixfold compared with substrate binding to wild-
Biophysical Journal 98(7) 1327–1336

1334
Cheng et al.
FIGURE 6 Isothermal calorimetric titration for the
substrate TQ6-pNA binding to Mpro and its mutants.
(A–D) Binding of TQ6-pNA to wild-type, R298A,
R298L, and R298A/Q299A mutants, respectively. Wild-
type protein concentration was 5.7 mM and protein concen-
tration of the mutants was 28.6 mM. The TQ6-pNA (1 mM)
was titrated into 2.7 ml of protein solution using 25–30
injections at a rate of 10 ml/injection. Values for the N,
Kd, and DH were determined by ligand binding analysis
with the Digitam program (TA instruments). The solid
circles show the observed values and the lines represent
fitted results.
To summarize the above observations, a schematic
induced dimerization. On the other hand, Glu-166 is critical
in the following subunit association step, which pushes the
equilibrium to the catalytic-competent dimer formation.
Then catalysis and release of products, followed by dissoci-
ation and regeneration of the inactive protomer, complete the
catalytic cycle. This mechanism is also compatible with
many current and previous studies (18,21,33,34).
diagram showing the proposed catalytic process of SARS-
CoV Mpro is presented in Fig. 8. The wild-type is a dimer,
as shown in our AUC studies, but we are not very sure about
the quaternary structure of Mpro under the highly diluted
assay conditions (<1.4 mM). We assume that there is a rapid
equilibrium between the monomer and dimer in solution.
Mutation at Arg-298 will push the equilibrium to monomer.
There is also a rapid equilibrium between the active and inac-
tive protomers. Substrate binding to the protomer (either
a monomer or a dimer) induces a conformational change
of the enzyme that shifts the equilibrium from monomers
to the catalytic-competent dimers. The protein-concentra-
tion-dependent enzyme activity experiment indicated that
only the dimer is enzymatically active. Arg-298 and Gln-
299 are involved in this rate-limiting step for Mpro catalysis.
When both residues are mutated, there will be no substrate-
CONCLUSION
In this study, we have delineated the contributions of
substrate binding to Mpro dimerization. Enzyme kinetic
assays indicated a nonlinear upward dependence between
the enzyme activity and enzyme concentration for wild-
type and R298A/L mutants. The quaternary structure of
R298A/L was also changed at different enzyme concentra-
tions. The AUC analysis proved that in the presence of
Biophysical Journal 98(7) 1327–1336

Substrate-Induced Dimerization of Mpro
1335
tion. The connection between the substrate binding site and
the dimer interface by Glu-166 may be universal in all CoV
Mpro. Substrate binding induces and stabilizes enzyme
dimerization, which then activates the enzyme molecule.
SUPPORTING MATERIAL
Four figuresareavailableat http://www.biophysj.org/biophysj/supplemental/
S0006-3495(09)06085-8.
This study was supported by grants from the National Science Council,
Taiwan (98-2320-B-010-026-MY3), and the National Health Research Insti-
tute, Taiwan (NHRI-EX99-9947SI), to C.Y.C. We also thank National
Yang-Ming University for its financial support (Aim for Top University
Plan from the Ministry of Education).
REFERENCES
FIGURE 7 Sedimentation velocity experimental values for Glu-166
mutants E166A (A) and E166A/R298A (B) in D2O. The protein concentra-
tion was 0.33 mg/ml. The distributions in D2O are represented by solid lines
and those in D2O with 600 mM TQ6-pNA substrate by dashed lines. The left
vertical dotted line indicates the monomer position and the right, the dimer
position. (Insets) Residual bitmaps for mutants without (upper) and with
(lower) substrate.
1. Peiris, J. S., Y. Guan, and K. Y. Yuen. 2004. Severe acute respiratory
syndrome. Nat. Med. 10(12, Suppl):S88–S97.
2. van der Hoek, L., K. Pyrc, ., B. Berkhout. 2004. Identification of
a new human coronavirus. Nat. Med. 10:368–373.
3. Woo, P. C., S. K. Lau, ., K. Y. Yuen. 2005. Characterization and
complete genome sequence of a novel coronavirus, coronavirus
HKU1, from patients with pneumonia. J. Virol. 79:884–895.
4. Fan, K., P. Wei, ., L. Lai. 2004. Biosynthesis, purification, and
substrate specificity of severe acute respiratory syndrome coronavirus
3C-like proteinase. J. Biol. Chem. 279:1637–1642.
peptide substrates, the major species of the R298A/L mutants
showed significant size shifting toward the dimeric form,
which became clear in the D2O environment. The substrate-
induced dimerization of R298A/L was reversible upon
removal of the substrates. Mutation of Glu-166 in R298A
mutant significantly blocks the substrate-induced dimeriza-
5. Hegyi, A., and J. Ziebuhr. 2002. Conservation of substrate specificities
among coronavirus main proteases. J. Gen. Virol. 83:595–599.
6. Yang, H., M. Yang, ., Z. Rao. 2003. The crystal structures of severe
acute respiratory syndrome virus main protease and its complex with
an inhibitor. Proc. Natl. Acad. Sci. USA. 100:13190–13195.
7. Anand, K., G. J. Palm, ., R. Hilgenfeld. 2002. Structure of coronavirus
main proteinase reveals combination of a chymotrypsin fold with an
extra a-helical domain. EMBO J. 21:3213–3224.
8. Xue, X., H. Yu, ., Z. Rao. 2008. Structures of two coronavirus main
proteases: implications for substrate binding and antiviral drug design.
J. Virol. 82:2515–2527.
9. Zhao, Q., S. Li, ., Z. Rao. 2008. Structure of the main protease from
a global infectious human coronavirus, HCoV-HKU1. J. Virol.
82:8647–8655.
10. Yang, H., W. Xie, ., Z. Rao. 2005. Design of wide-spectrum inhibitors
targeting coronavirus main proteases. PLoS Biol. 3:e324.
11. Bacha, U., J. Barrila, ., E. Freire. 2008. Development of broad-
spectrum halomethyl ketone inhibitors against coronavirus main
protease 3CL(pro). Chem. Biol. Drug Des. 72:34–49.
12. Hegyi, A., A. Friebe, ., J. Ziebuhr. 2002. Mutational analysis of the
active centre of coronavirus 3C-like proteases. J. Gen. Virol. 83:
581–593.
13. Ziebuhr, J., S. Bayer, ., A. E. Gorbalenya. 2003. The 3C-like
proteinase of an invertebrate nidovirus links coronavirus and potyvirus
homologs. J. Virol. 77:1415–1426.
14. Chang, H. P., C. Y. Chou, and G. G. Chang. 2007. Reversible unfolding
of the severe acute respiratory syndrome coronavirus main protease in
guanidinium chloride. Biophys. J. 92:1374–1383.
15. Zhong, N., S. Zhang, ., B. Xia. 2009. C-terminal domain of SARS-
CoV main protease can form a 3D domain-swapped dimer. Protein
Sci. 18:839–844.
FIGURE 8 Schematic model for the catalytic process of SARS-CoV
Mpro. Substrate binding triggers the monomer-dimer equilibrium to favor
the active protomer (square). The active protomer now is prompt to asso-
ciate, forming the catalytic-competent dimer. After the catalytic cycle, the
associated dimer will dissociate again to form the inactive protomer (circle).
16. Chou, C. Y., H. C. Chang, ., G. G. Chang. 2004. Quaternary structure
of the severe acute respiratory syndrome (SARS) coronavirus main
protease. Biochemistry. 43:14958–14970.
Biophysical Journal 98(7) 1327–1336

1336
Cheng et al.
17. Hsu, W. C., H. C. Chang, ., G. G. Chang. 2005. Critical assessment of
important regions in the subunit association and catalytic action of the
severe acute respiratory syndrome coronavirus main protease. J. Biol.
Chem. 280:22741–22748.
26. Schuck, P. 2000. Size-distribution analysis of macromolecules by sedi-
mentation velocity ultracentrifugation and Lamm equation modeling.
Biophys. J. 78:1606–1619.
27. Schuck, P. 2003. On the analysis of protein self-association by sedimen-
tation velocity analytical ultracentrifugation. Anal. Biochem. 320:
104–124.
18. Lin, P. Y., C. Y. Chou, ., G. G. Chang. 2008. Correlation between
dissociation and catalysis of SARS-CoV main protease. Arch. Biochem.
Biophys. 472:34–42.
28. Shen, Y., C. Y. Chou, ., L. Tong. 2006. Is dimerization required for
the catalytic activity of bacterial biotin carboxylase? Mol. Cell.
22:807–818.
19. Chen, S., T. Hu, ., X. Shen. 2008. Mutation of Gly-11 on the dimer
interface results in the complete crystallographic dimer dissociation of
severe acute respiratory syndrome coronavirus 3C-like protease: crystal
structure with molecular dynamics simulations. J. Biol. Chem. 283:
554–564.
29. Bai, Y., T. C. Auperin, ., L. Tong. 2007. Crystal structure of murine
CstF-77: dimeric association and implications for polyadenylation of
mRNA precursors. Mol. Cell. 25:863–875.
20. Hu, T., Y. Zhang, ., X. Shen. 2009. Two adjacent mutations on the
dimer interface of SARS coronavirus 3C-like protease cause different
conformational changes in crystal structure. Virology. 388:324–334.
30. Sondermann, H., B. Nagar, ., J. Kuriyan. 2005. Computational dock-
ing and solution x-ray scattering predict a membrane-interacting role for
the histone domain of the Ras activator son of sevenless. Proc. Natl.
Acad. Sci. USA. 102:16632–16637.
21. Shi, J., J. Sivaraman, and J. Song. 2008. Mechanism for controlling
the dimer-monomer switch and coupling dimerization to catalysis of
the severe acute respiratory syndrome coronavirus 3C-like protease.
J. Virol. 82:4620–4629.
31. Barrila, J., U. Bacha, and E. Freire. 2006. Long-range cooperative inter-
actions modulate dimerization in SARS 3CLpro. Biochemistry.
45:14908–14916.
22. Chou, C. Y., Y. L. Lin, ., M. S. Shiao. 2005. Structural variation in
human apolipoprotein E3 and E4: secondary structure, tertiary structure,
and size distribution. Biophys. J. 88:455–466.
32. Solowiej, J., J. A. Thomson, ., B. W. Murray. 2008. Steady-state and
pre-steady-state kinetic evaluation of severe acute respiratory syndrome
coronavirus (SARS-CoV) 3CLpro cysteine protease: development of an
ion-pair model for catalysis. Biochemistry. 47:2617–2630.
23. Huang, C., P. Wei, ., L. Lai. 2004. 3C-like proteinase from SARS
coronavirus catalyzes substrate hydrolysis by a general base mecha-
nism. Biochemistry. 43:4568–4574.
33. Chen, H., P. Wei, ., L. Lai. 2006. Only one protomer is active in the
dimer of SARS 3C-like proteinase. J. Biol. Chem. 281:13894–13898.
24. Stra¨ter, N., L. Sun, ., W. N. Lipscomb. 1999. A bicarbonate ion as
a general base in the mechanism of peptide hydrolysis by dizinc leucine
aminopeptidase. Proc. Natl. Acad. Sci. USA. 96:11151–11155.
34. Hsu, M. F., C. J. Kuo, ., P. H. Liang. 2005. Mechanism of the
maturation process of SARS-CoV 3CL protease. J. Biol. Chem. 280:
31257–31266.
25. Brown, P. H., and P. Schuck. 2006. Macromolecular size-and-shape
distributions by sedimentation velocity analytical ultracentrifugation.
Biophys. J. 90:4651–4661.
35. DeLano, W. L. 2002. The PyMOL Manual. DeLano Scientific, San
Carlos, CA.
Biophysical Journal 98(7) 1327–1336