Tài liệu Báo cáo khoa học: C-Terminal extension of a plant cysteine protease modulates proteolytic activity through a partial inhibitory mechanism doc

C-Terminal extension of a plant cysteine protease
modulates proteolytic activity through a partial inhibitory
mechanism
Sruti Dutta, Debi Choudhury, Jiban K. Dattagupta and Sampa Biswas
Crystallography and Molecular Biology Division, Saha Institute of Nuclear Physics, Kolkata, India
Keywords
C-terminal extension; cysteine proteases;
modulation of proteolytic activity;
papain-like; thermostable
Correspondence
S. Biswas, Crystallography and Molecular
Biology Division, Saha Institute of Nuclear
Physics, 1 ⁄ AF Bidhannagar, Kolkata
700 064, India
Fax: +91 332 337 4637
Tel: +91 332 337 5345
E-mail:
(Received 14 March 2011, revised 16 May
2011, accepted 22 June 2011)
doi:10.1111/j.1742-4658.2011.08221.x
The amino acid sequence of ervatamin-C, a thermostable cysteine protease
from a tropical plant, revealed an additional 24-amino-acid extension at its
C-terminus (CT). The role of this extension peptide in zymogen activation,
catalytic activity, folding and stability of the protease is reported. For this
study, we expressed two recombinant forms of the protease in Escherichia
coli, one retaining the CT-extension and the other with it truncated. The
enzyme with the extension shows autocatalytic zymogen activation at a
higher pH of 8.0, whereas deletion of the extension results in a more active
form of the enzyme. This CT-extension was not found to be cleaved during
autocatalysis or by limited proteolysis by different external proteases.
Molecular modeling and simulation studies revealed that the CT-extension

blocks some of the substrate-binding unprimed subsites including the speci-
ficity-determining subsite (S2) of the enzyme and thereby partially occludes
accessibility of the substrates to the active site, which also corroborates the
experimental observations. The CT-extension in the model structure shows
tight packing with the catalytic domain of the enzyme, mediated by strong
hydrophobic and H-bond interactions, thus restricting accessibility of its
cleavage sites to the protease itself or to the external proteases. Kinetic
stability analyses (T
50
and t
1 ⁄ 2
) and refolding experiments show similar
thermal stability and refolding efficiency for both forms. These data suggest
that the CT-extension has an inhibitory role in the proteolytic activity of
ervatamin-C but does not have a major role either in stabilizing the enzyme
or in its folding mechanism.
Structured digital abstract
l
ErvC cleaves ErvC by protease assay (View interaction)
l
trypsin cleaves ErvC by protease assay (View interaction)
Introduction
Papain-like cysteine proteases (EC 3.4.22) from plant
sources are of industrial and biotechnological impor-
tance because these enzymes are better suited to various
industrial processes [1]. A cysteine protease is expressed
as an inactive precursor in a pre-proenzyme form
which contains a signal peptide (pre-), an inhibitory
Abbreviations
CT, C-terminal; E-64, 1-[L-N-(trans-epoxysuccinyl)leucyl]amino-4-guanidinobutane; Erv-C, ervatamin-C; pNA, p-nitroanilide; rmErv-C

+CT
,
recombinant mature ervatamin-C with C-terminal extension; rmErv-C
DCT
, recombinant mature ervatamin-C without C-terminal extension;
rproErv-C
+CT
, recombinant proervatamin-C with C-terminal extension; rproErv-C
DCT
, recombinant proervatamin-C without C-terminal
extension.
3012 FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS
pro-region and a mature catalytic domain [2–4]. Fol-
lowing synthesis, the pre-peptide is removed during
passage to the lumen of the endoplasmic reticulum [2].
The inactive proenzyme subsequently undergoes prote-
olytic processing to produce an active mature enzyme
by autocatalytic cleavage of the propeptide part at the
N-terminus [2]. It is known that the propeptide at the
N-terminus of the protease acts as an intramolecular
chaperone to mediate correct folding of the protease
[2]. The mature catalytic domain of the enzyme of this
family has a molecular mass of  21–30 kDa and
shares a common fold with papain, the archetype
enzyme of the family. These proteases are folded into
two compact interacting domains of comparable size,
delimiting a cleft which contains the active site residues
cysteine and histidine, forming a zwitterionic catalytic
dyad (Cys
)

His
+
) [5].
Sometimes a larger precursor is also synthesized
which contains a C-terminal (CT) extension ⁄ propep-
tide in addition to the abovementioned N-terminal
propeptide flanking the mature protease domain [6,7].
Unlike N-terminal propeptides, the role of CT-exten-
sions (or propeptides) is not yet well established.
Sometimes an endoplasmic reticulum retention signal
Lys-Asp-Glu-Leu (KDEL) is found in the CT-propep-
tide which regulates the delivery of protease precursor
to other cellular compartments [8]. Some other mem-
bers of the papain family from plant sources also con-
tain a larger CT-propeptide domain which shares a
similarity with animal epithelin ⁄ granulin and the func-
tion of this domain is reported to be involved in leaf
senescence [6]. In addition, a CT-propeptide without
any specific domain or motif is observed in some
papain-like cysteine proteases from plants like Nicoti-
ana tabacam (Q84YH7), Actinidia chinensis (P00785)
and Vicia sativa (Q41696). In most of the cases, such a
CT-extension contains the vacuolar sorting signal and
is cleaved inter- or intramolecularly after sorting [9].
No conserved sequence motif has been found in the
vacuolar sorting signal at CT-propeptide, rather an
amphipathic-like (hydrophobic and acidic) motif is
generally observed [9,10] at the core of such peptides.
Other than plant systems, a CT-extension found in
a lysosomal cysteine protease (Lpcys2) of Leish-

mania pifanoi, plays a role in the regulation of enzyme
activity [11]. CT-extension in mammalian and yeast
bleomycin hydrolase [12,13] is a key factor which regu-
lates their endo-peptidase or exo-aminopepetidase
activity by blocking the unprimed subsites in the
enzymes.
Ervatamin-C (Erv-C) is a papain-like cysteine prote-
ase (
EC 3.4.22) with high stability purified from the
latex of a tropical plant Ervatamia coronaria [14]. The
3D structure of Erv-C reveals an extra disulfide bond,
shorter loop regions and additional electrostatic inter-
actions in the interdomain space, which are thought to
be responsible for its high stability [15]. Sequencing
of the cDNA (from mRNA) of Erv-C from the leaf
of the plant in our laboratory [16], and comparison of
the cDNA-derived amino acid sequence with other
members of the family reveal that Erv-C is synthesized
as a precursor protein and in addition to the pre- (19
amino acids), pro- (114 amino acids) and mature (208
amino acids) parts, it contains an extension of 24
amino acids at the CT of the mature enzyme [16]
(Fig. 1A). This CT-extension was not observed, how-
ever, when the mature Erv-C was purified directly
from the latex of the plant [15].
In this article, we attempt to understand the role of
this CT-extension in zymogen activation, enzyme activ-
ity, folding and stability in vitro at the molecular level
from structural and functional points of view.
Results

Cloning, expression, purification and refolding of
rproErv-C
DCT
and rproErv-C
+CT
Both the proteins, recombinant proervatamin-C with-
out the CT-extension (rproErv-C
DCT
) and recombinant
proervatamin-C with the CT-extension (rproErv-
C
+CT
), were expressed in E. coli as inclusion bodies
with an apparent molecular mass of  41 and
 43 kDa (Fig. 1B), respectively, which is consistent
with the estimated molecular masses of their deduced
amino acid sequences. Correct refolding was checked
by gelatin gel assay. The condition and efficiency of
refolding were almost similar for both forms with
> 90% recovery of the folded form from Ni-NTA
purified protein for each (Table S1).
Activation to mature protease
The purified refolded rproErv-C
+CT
could be con-
verted into its mature active form (rmErv-C
+CT
)by
using cysteine (20 m
M) as the activator in 50 mM Tris

buffer, pH 8.00, at 60 °C for 25–30 min, whereas the
purified refolded rproErv-C
DCT
could be converted
into its mature form (rmErv-C
DCT
) by the same activa-
tor in 50 mM Na-acetate buffer, pH 4.5, at 60 °C for
45 min (Fig. 2A). Thus the zymogen activation process
occurs at different pH and time of maturation for the
two enzymes. The molecular mass of the mature
enzyme rmErv-C
+CT
is higher ( 27 kDa) than that
of rmErv-C
DCT
( 25 kDa) as observed in the
SDS ⁄ PAGE analyses (Fig. 2A). This difference in
S. Dutta et al. Role of C-terminal extension in a cysteine protease
FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS 3013
molecular mass almost fits the theoretically calculated
value for the 24-amino-acid CT-extension. We
expected, therefore, that the CT-extension continued to
remain attached with the mature enzyme even after
autocatalytic processing of rproErv-C
+CT
. Gelatin gel
assay (Fig. 2B) and western blot analyses (Fig. 2C)
also confirmed the retention of the CT-extension in the
mature rmErv-C

+CT
.
Because Erv-C isolated from the latex of the same
plant does not show the CT-extension, the possibility
19 aa 114 aa
208 aa
24 aa
Pre N-Pro Protease domain
CT - ex
365 aa
H1
I
N
Pro CT ex
114 aa
208 aa
24 aa
BamH
34 aa
Start
Stop
Xho1
N
-Pro Protease domain
Protease domain
CT - ex
380 aa
1
II
NP

114 aa
208 aa
BamH1
34 aa
Start
Stop
Xho1
SS SS
G
SSS
S
N-Pro
356 aa
III
MSTLFII
S
ILLFLAS
F
SYAMDI
S
TIEYKYDKSS
AWRTDEEVKEIYELWLAKHDKVY
SG
LVEYEKRFEIFKDNLKFIDEH
NSENHTYKMGLTPYTDLTNEEFQAIYLGTRSDTIHRLKRTINISERYAYEAGDNLPEQIDWRKKGAVTPVKNQGKCG
SCWAFSTVSTVESINQIRTGNLISLSEQQLVDCNKKNHG
CKGGAFVYAYQYIIDNGGIDTEANYPYKAVQGPCRAAK
KVVRIDGYKGVPHCNENALKKAVASQPSVVAIDASSKQF
QHYKSGIFSGPCGTKLNHGVVIVGYWKDYWIVRNSW
GRYWGEQGYIRMKRVGGCGLCGIARLPYYPTKA

AGDENSKLETPELLQWSEEAFPLA
IV
A
B
66 kDa
45 kDa
36 kDa
29 kDa
24 kDa
20 kDa
321
Fig. 1. (A) (I) Open reading frame of Erv-C precursor, pre-pro-ErvC. (II) Recombinant ervatamin-C with C-terminal extension, rproErv-C
+CT
.
(III) rproErv-C
DCT
, recombinant ervatamin-C without C-terminal extension. Red indicates vector portion and ‘aa’ stands for amino acids. (IV)
The amino acid sequence of the open reading frame. The sequence of CT-extension is in red. (B) Lanes 1 and 2, Purified proteins rproErv-
C
+CT
and rproErv-C
DCT
, respectively; lane 3, Molecular mass markers.
Role of C-terminal extension in a cysteine protease S. Dutta et al.
3014 FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS
of removal of the CT-extension by other plant prote-
ases in vivo could not be ruled out. To explore this
possibility, we performed a trans mode activation of
rproErv-C
+CT

in vitro using seven different proteases
(four cysteine proteases Erv-A, -B, -C and papain from
the plant latex; two serine proteases trypsin and chy-
motrypsin from bovine pancreas; one aspartic protease
pepsin from porcine stomach mucosa), each having
sequences specific for their cleavage in the amino acid
sequence of the CT-extension. The activated mature
protein thus generated in each case shows a band at
the same position ( 27 kDa) like that in the autoacti-
vated enzyme, as observed in SDS ⁄ PAGE analyses
(Fig. S1), indicating that these external enzymes can
not cleave the CT-extension. Even with a prolonged
digestion time (24 h), the same result was obtained for
all proteases except trypsin (Fig. S1). Trypsin digestion
for 24 h resulted in a truncated protein at  26 kDa,
slightly above the activated mature rmErv-C
DCT
( 25 kDa). This result probably indicates that trypsin
has some accessibility to its sites of specificity (Lys,
which is in position 7 of the CT-extension) (Fig. S2)
and only after a prolonged incubation time can it
result in a band at a slightly lower molecular mass
position, as observed in the SDS ⁄ PAGE analysis
(Fig. S1).
Specific activity and optimum temperature of
activity
The optimum temperature of activity, T
opt
(Fig. 3), for
both forms is 65 °C. Interestingly, it was observed that

rproErv-C
+CT
shows no activity below 45 °C and then
activity rises sharply from 60 °C onwards, reaching a
maximum at 65 °C. In the case of rproErv-C
DCT
,a
gradual increase in activity with temperature was
observed until it reached its maximum at 65 °C.
At T
opt
(65 °C), the specific activity of rmErv-C
DCT
was found to be almost double that of rmErv-C
+CT
(Table 1). At 37 °C, however, rproErv-C
DCT
shows
measurable proteolytic activity, whereas no activity
was seen for rmErv-C
+CT
.
Kinetic measurements of the recombinant
proteins
Kinetic constants of rproErv-C
+CT
and rproErv-C
DCT
were measured at room temperature against N-ben-
zoyl-Phe-Val-Arg-p-nitroanilide (pNA), a tripeptide

substrate with a valine at the P2 position which is
known to act as a substrate for Erv-C [17]. The kinetic
constants of the two recombinant enzymes (Table 1)
clearly show that rproErv-C
DCT
has almost 10 times
97 kDa
66 kDa
66 kDa
45 kDa
36 kDa
29 kDa
24 kDa
20 kDa
43 kDa
rproErv-C
+CT
rproErv-C
+CT
rproErv-C
ΔCT
rmErv-C
+CT
rmErv-C
ΔCT
rproErv-C
ΔCT
rmErv-C
ΔCT
rmErv-C

+CT
29 kDa
20 kDa
M
1
123
2
45 30 20 10 0 C25 15 5
M4530 20 10 0 C25 15 5
A
B
C
Fig. 2. (A) Time course of activation to the mature form. (I) rproErv-C
+CT
. (II) rproErv-C
DCT
as discussed in Materials and methods. Time
intervals of 0–45 min are indicated for the respective lanes. Untreated proteins (control) are labelled as ‘C’, ‘M’ denotes the molecular mass
marker. (B) Gelatin gel assay of activated rproErv-C
DCT
(lane 1) and rproErv-C
+CT
(lane 2). (C) Western blot analysis. Lane 1, Erv-C purified
from the plant latex; lanes 2 and 3, purified and refolded rproErv-C
+CT
and rproErv-C
DCT
. [Correction added on 26 July 2011 after original
online publication: in the figure, labelling for part C was changed from ‘rproErv-C
+CT

, rproErv-C
DCT
, rproErv-C
+CT
and rproErv-C
DCT
to rproErv-C
+CT
,
rproErv-C
DCT
, rmErv-C
+CT
and rmErv-C
DCT
’].
S. Dutta et al. Role of C-terminal extension in a cysteine protease
FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS 3015
higher activity than rproErv-C
+CT
. One can probably
conclude that the CT-extension has some inhibitory
effect on the activity of the enzyme against a small
peptide.
Thermal stability
Temperatures of maximum proteolytic activity (T
max
)
for rproErv-C
+CT

and rproErv-C
DCT
are 50 and 45 °C
(Table 2 and Fig. 4) and they retain > 90% activity
up to 65 and 60 °C, respectively. These data indicate a
good thermotolerance for these enzymes. The pattern
of retention and fall in activity beyond T
max
was more
or less the same for both enzymes. The T
50
values for
rproErv-C
+CT
and rproErv-C
DCT
were 76 and 72 °C
(Fig. 4), respectively. The half lives (t
1 ⁄ 2
)at65°C were
 400 min (Fig. 5) for both enzymes.
Molecular modeling studies
To gain insight into the stability and dynamic proper-
ties of the structure, solvent MD simulation was
110
100
rproErv-C
+CT
rproErv-C
ΔCT

90
80
70
60
Residual activity (%)
Temperature (°C)
50
40
30
20
10
0
30 35 40 45 50 55 60 65 70 75 80 85 90 10095
Fig. 3. Determination of optimum temperature of activity (T
opt
)of
rproErv-C
+CT
and rproErv-C
DCT
. Purified proenzymes (10–20 lg)
were converted to their respective mature forms and the percent-
age residual enzyme activities were determined with respect to the
maximum activity using an azocasein assay at different tempera-
tures, as described in Materials and methods. Each data point is an
average of three independent experiments having similar values
(Table S3).
Table 1. Kinetic constants using the substrate N-benzoyl-Phe-Val-Arg-pNA. Specific activity using azocaesin and IC
50
value for the inhibitor

E-64. ND, not determined.
k
cat
(s
)1
)
a
K
m
(lM)
a
k
cat
⁄ K
m
(s
)1
ÆmM
)1
)
Specific activity
at 37 °C
b
(UÆmg
)1
)
Specific activity
at 65 °C
b
(UÆmg

)1
)
IC
50
against
E-64 (n
M)
c
rproErv-C
+CT
0.0170 ± 0.004 88.33 ± 56.77 0.193 No activity 15.87 ± 1.36 482.5 ± 108.0
rproErv-C
DCT
0.2295 ± 0.057 127.3 ± 48.46 1.803 13.21 ± 2.32 35.27 ± 1.78 349.1 ± 62.0
Latex Erv-C [17] 9.312 1063 8.76 75.0 ND 225.0
a
Given standard errors were calculated based on nonlinear fitting of the Michaelis–Menten saturation curve using the software Graphpad
PRISM ( />b
Each value of specific activity of rproErv-C+CT and rproErv-CDCT is a mean of three independent
experiments ± SD.
c
Given standard deviations were calculated from linear regression plot of residual activity and inhibitor concentration.
Table 2. Kinetic stabilities. ND, not determined.
T
max
(°C) T
50
(°C)
t
1 ⁄ 2

at
65 °C (min) T
opt
(°C)
rproErv-C
+CT
(activity at 65 °C)
50 76  400 65
rproErv-C
DC
(activity at 65 °C)
45 72  400 65–70
Native mature Erv-C
(latex) [14,32]
70 76 ND 50
100
rproErv-C
+CT
rproErv-C
ΔCT
80
60
Residual activity (%)
Temperature (°C)
40
20
0
40 45 50 55 60 65 70 75 80 85 90
Fig. 4. Effect of temperature on activity of rproErv-C
+CT

and rpr-
oErv-C
DCT
. Each purified proenzyme (10–20 lg) was treated for
10 min at different temperatures followed by activation of the pro-
proteins to their respective mature forms. The percentage residual
enzyme activities (at each temperature) were determined with
respect to the maximum activity using an azocasein assay at 65 °C
as described in Materials and methods. Each data point is an aver-
age of three independent experiments having similar values
(Table S3).
Role of C-terminal extension in a cysteine protease S. Dutta et al.
3016 FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS
performed. The total energy of the whole system and
root mean square deviation (RMSD) from the starting
structure are essential to determine the sustainability
and convergence of MD simulation. Figure 6A shows
the RMSD of backbone atoms of the CT-extension in
association with a mature domain as well as in an iso-
lated form. The graph shows that the RMSD reached
below  0.4 A
˚
when the extension is attached to the
mature domain and is > 1 A
˚
when the extension is on
its own. The fluctuation in the radius of gyration was
also analyzed (Fig. 6B) as a measurement of the over-
all stability of the CT-extension for both forms. These
analyses show that the extension achieves a relatively

more stable conformation when it is attached to the
mature domain.
The modeled structure of the CT-extension with the
mature catalytic domain (rmErv-C
+CT
) shows that the
extension peptide blocks some of the unprimed sub-
sites of the enzyme (Fig. 7). The interface area of the
CT-extension and the mature catalytic domain is
1037 A
˚
2
, which is  56% of the total surface area of
the CT-extension. The Leu side chain at the position
23 of the CT-extension occupies specificity pocket S2
of the catalytic domain and is stabilized mainly by
hydrophobic interactions with S2 subsite residues A67,
F68, A131, L155 and L201 (Fig. 8) [17]. Residue
Phe21 of the CT-extension is buried inside a hydro-
phobic pocket of mature domain formed by V69, L201
and Y203 (Fig. 8B). Other residues of the CT-exten-
sion are stabilized by electrostatic and hydrophobic
interactions with the mature catalytic domain of the
enzyme. Superposition of the crystal structure of the
complex of mature Erv-C (without CT-extension) with
the inhibitor 1-[L-N-(trans-epoxysuccinyl)leucyl]amino-
4-guanidinobutane (E-64; Protein Data Bank ID
2PRE) and the modeled structure of Erv-C with
CT-extension (rmErv-C
+CT

) reveals that the Leu of
100
rproErv-C
+CT
rproErv-C
ΔCT
90
80
70
60
Residual activity (%)
Time in min
50
40
30
20
10
0
0
10
20
30
40
50
60
120
180
240
300
360

420
Fig. 5. Time course of thermal inactivation of rproErv-C
+CT
and
rproErv-C
DCT
at 65 °C. An aliquot of the purified proenzymes
(10–20 lg) was treated at 65 °C for 0 min to 8 h followed by activa-
tion of the pro-proteins to their respective mature forms. At each
indicated time, the residual enzyme activity was determined using
an azocasein assay at 65 °C, as described in Materials and meth-
ods, and values are expressed as percentage of the initial activity
of the respective enzymes. The experiment was carried out in
duplicate for each data point (Table S3).
Fig. 6. 2 ns molecular dynamics trajectory of CT-extension part in
association with the mature Erv-C domain (red) and in an isolated
form (black). (A) Backbone RMSD and (B) radius of gyration.
AB
Fig. 7. Surface presentation of mature Erv-C. (A) Mature Erv-C
without the CT-extension (Protein Data Bank ID
2PNS), the catalytic
cleft is marked in red. (B) Mature Erv-C with modeled CT-exten-
sion, the CT-extension is displayed in magenta.
S. Dutta et al. Role of C-terminal extension in a cysteine protease
FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS 3017
E-64 at P2 position of the inhibitor lies close to Leu23
of the CT-extension of the rmErv-C
+CT
molecule with
a minimum distance of 1.9 A

˚
between two atoms of
the two leucines (Fig. 8A). Because the Leu23 residue
of the CT-extension of the rmErv-C
+CT
molecule is
buried inside the S2 pocket, it restricts access of Leu at
the P2 position of E-64 and results in a higher IC
50
value (482.5 nM) of E-64 inhibition compared with
rmErv-C
DCT
(349.1 nM) (Table 1) in which the exten-
sion is truncated. Modeling studies also reveal that
although this CT-extension blocks some of the
unprimed subsites beyond S2, access to the catalytic
centre (Cys-His dyad) is not totally blocked.
Discussion
It is known that the autocatalytic processing of
papain-like cysteine proteases from pro- to mature
form generally occurs at acidic pH [18]. The 3D struc-
tures of papain-like cysteine proteases in the pro-form
[19–23] reveal that N-terminal propeptide part adopts
a specific globular structure which is conserved among
the family despite a relatively low homology in their
amino acid sequences. Previous reports suggest that an
acidic pH induces a conformational change in the
N-terminal propeptide domain, resulting in a molten
globule state and thereby the activation process is trig-
gered [24]. The molten globule state of the N-terminal

propeptide domain results in a reduction in the associ-
ation affinity of the propeptide towards the protease
domain and cleavage of the propeptide occurs leading
to a mature active enzyme. To date, there has been
practically no detailed report available in the literature
on the role of the CT-extension (or CT-propeptide) in
the maturation process for proteases in this family.
There is one study on kiwifruit cysteine protease,
actinidin, which shows that its C-terminal propeptide
is required for correct processing [7]. In our studies,
we have observed that for the precursor rproErv-
C
+CT
, in vitro autoactivation essentially removes the
N-terminal propeptide part leading to a mature active
protease (rmErv-C
+CT
) with a molecular mass of
 27 kDa (Fig. 2) with the CT-extension remaining
attached. Moreover, this autocatalytic processing does
not occur at acidic pH, instead autoactivation is found
to occur at a basic pH of 8.0. By contrast, in rproErv-
C
DCT
, in vitro activation to mature enzyme (rmErv-
C
DCT
) occurs at an acidic pH of 4.5, like in other
members of the family [17]. The 24-amino-acid
CT-extension contains six negatively charged residues

(one aspartate and five glutamates) (Fig. 1A) and our
previous molecular modeling studies [16] indicated that
this negatively charged region of the C-terminus could
A
B
Fig. 8. (A) The structure of the catalytic cleft and unprimed sub-
sites region of mature Erv-C with modeled CT-extension. The
mature catalytic domain is represented as an electrostatic potential
surface and the CT-extension as a stick model with C atoms in
green. The inhibitor E-64 (taken from the structure of Erv-C and
E-64 complex; Protein Data Bank ID
2PRE) is also docked for com-
parison and is represented as a stick model with C atoms in brown.
The catalytic cysteine residue (C25) of Erv-C is represented as a
ball and stick model (light blue) and the S2 subsite residues are rep-
resented as a stick model (light pink). The minimum distance
(1.9 A
˚
) between P2 (Leu23) of the CT-extension and P2(Leu) of
E-64 are indicated by the dotted line. (B) The last four residues,
F21P22L23A24, of the CT-extension domain (C atoms are shown in
magenta) and the neighboring residues of Erv-C mature domain
within 4.2 A
˚
(C atoms are shown in deep bottle green). Interdomain
distances within 4.2 A
˚
are shown in green.
Role of C-terminal extension in a cysteine protease S. Dutta et al.
3018 FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS

be positioned structurally in such a way that it could
interact with a positively charged zone of the N-termi-
nal propeptide in the zymogen. Therefore, one may
postulate that at acidic pH, this electrostatic inter-
action between the N- and C-terminal propeptides ⁄
extensions may be strong enough to restrict the confor-
mational change in the N-terminal propeptide,
required for the activation process. But at basic pH,
this interaction may become weak, leading to a free
and flexible C-terminus, and the presence of the six
acidic residues in this region may locally mimic an
acidic environment to trigger the activation process.
This may be the explanation for autocatalytic activa-
tion at a higher pH for rproErv-C
+CT
which is not
observed when the protein is expressed without the
extension part in rproErv-C
DCT
.
Because the C-terminal extension of 24 residues
remains intact in the mature form of the rmErv-C
+CT
molecule (Fig. 2) after in vitro processing, autocatalytic
trimming does not occur in this case. It should be
mentioned, however, that in some vacuolar papain-like
cysteine proteases the CT-extension (propeptide), car-
rying a vacuolar sorting signal, is cleaved by intermo-
lecular proteolysis by a different vacuolar protease
in vivo [9,10]. So, in this study, we performed an

in vitro transactivation experiment using three latex
enzymes Erv-A, -B and -C from the same plant, and
papain from papaya latex and two other serine (tryp-
sin and chymotrypsin) and one aspartic (pepsin) prote-
ases. The results showed that this extension is also not
cleaved by any of these proteases in their optimum
condition of activity. However, because no experiment
has been carried out in vivo, the possibility of intermo-
lecular cleavage of the CT-extension by a different
vacuolar protease under specific in vivo conditions can
not be ruled out.
The results of enzyme kinetic studies show that the
proteolytic activities of the two recombinant forms
(rproErv-C
+CT
and rproErv-C
DCT
) are not of the same
order (Table 1). In optimum temperature determina-
tion using azocasein as a substrate, rmErv-C
+CT
does
not show any proteolytic activity when activity is
assayed below 45 °C, but activity increases suddenly
beyond 60 °C and sufficient activity is retained in the
temperature range 65–75 °C (Fig. 3) with the highest
being at 65 °C. This activity profile is different for
rmErv-C
DCT
where activity increases systematically

with temperature, although the highest activity in this
case is also at 65 °C (Fig. 3). However, the specific
activity of rmErv-C
DCT
is almost twice as high as that
of rmErv-C
+CT
at their optimum temperature of activ-
ity (65 °C) (Table 1). When the activity is measured
with a small peptide like N-benzoyl-Phe-Val-Arg-pNA,
both forms, rmErv-C
+CT
and rmErv-C
DCT
, show
activity at room temperature although the former has
a 10 times lower K
cat
⁄ K
m
value (Table 1). Thus, for a
small peptide, enzymatic activity is observed at room
temperature for rmErv-C
+CT
(around 25 °C), which is
totally absent when this form is assayed with azocaesin
(Table 1). These data suggest that the CT-extension
interferes with a protein substrate at a lower tempera-
ture and the enzyme active site is not accessible to the
protein substrate. But at a higher temperature (beyond

60 °C), this interference is partly removed and the
enzyme can show proteolytic activity to the protein
substrate. This behavior of the enzyme differs for a
smaller peptide substrate, where the enzyme can work
on this peptide even at lower temperatures, although
to a lesser extent. Molecular modeling studies show
that the C-terminal tail blocks the unprimed subsites
beyond S2 of the enzyme thus inhibiting the endopep-
tidase activity at temperatures below 55 °C for azocae-
sin. Perhaps, at a higher temperature, a reduction of
the association affinity for this tail towards the
unprimed subsites occurs leaving the subsites partially
free for substrate binding. This is in conformity with a
report [25] that the CT-extension sometimes has an
inhibitory property towards some enzymes. Blocking
of unprimed subsites is also found in other papain-like
proteases by different mechanisms either by a mini-
chain like cathepsin-H [26] or an exclusion domain in
cathepsin-C [27] or by C-terminal extension in bleomy-
cin hydrolase [12,13]. In all these cases, each protease
loses its endo-peptidase activity and functions as an
aminopeptidase because their unprimed subsites are
blocked as found here. A superposition of the struc-
tures of cathepsin-H and bleomycin hydrolase on the
modeled structure of rmErv-C
+CT
revealed that
although the blocking strategy of these peptides is sim-
ilar, in the first two proteases the blocking peptides
extend closer to the active site than in rmErv-C

+CT
(Fig. 9).
We also note from this study that the catalytic activ-
ity and maturation kinetics (both pH of activation and
time of activation) of the two forms vary (Fig. 2A).
Alternative splicing in plant systems has been reported
[28] to affect the stability and translatability at the
RNA level and produce truncated or extended proteins
with altered (increased, decreased or loss of) activity,
cellular localization, regulation and ⁄ or stability. Here,
we used the clone of Erv-C which was constructed
from the cDNA (mRNA) of the leaf of the plant [16]
and the deletion mutant of that has been generated by
eliminating the last 24 amino acids at the C-terminus.
The mature Erv-C isolated from the latex of the plant
also does not have the CT-extension [15]. Because both
S. Dutta et al. Role of C-terminal extension in a cysteine protease
FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS 3019
enzymes, latex Erv-C and recombinant mature rmErv-
C
DCT
without a CT-extension, appear to be the same,
having a similar molecular mass (Fig. 2C) and similar
enzymatic properties [29], we cannot ignore the possi-
bility that Erv-C may exist in two isoforms (splice vari-
ants) in different plant tissues, and that the activation
requirements of the two isoforms are different depend-
ing on the presence or absence of a 24-amino-acid neg-
atively charged CT-extension.
We have noted that both recombinant protease

forms can fold efficiently in vitro. So the CT-extension
does not appear to have any noticeable effect on proper
folding of the protein in vitro. However, our observa-
tions also indicate that this extension has some effects
in the maturation process and activity of the protease.
It suggests two possibilities for the presence of the CT-
extension in the enzyme: either it carries the signal for
vacuolar sorting and after that is cleaved by another
enzyme in vivo, or an mRNA (cDNA) exists in the leaf
of the plant that essentially encodes an isoform of Erv-
C which is present in the latex of the plant.
Materials and methods
Cloning, expression, purification and refolding of
rproErv-C
+CT
The entire open reading frame of the Erv-C precursor had
been cloned previously in our laboratory in pTZ57R ⁄ T
vector [16]. A fragment of the original cDNA encoding the
prodomain, the mature domain and CT-extension of Erv-C
was PCR-amplified from this clone using primers (Forward:
5¢-CCC
GGATCCATGGACATATCTACC-3¢ and Reverse:
5¢-GGT
CTCGAGTTAAGCAAGTGGAAAAGCT-3¢)desig-
ned to delete the pre-peptide (the signal peptide) and to
include the restriction sites for BamHI and Xho1 (under-
lined) to facilitate cloning into pET-28a(+) expression vec-
tor (Novagen, Madison, WI, USA). The amplified product
was then subcloned into respective restriction sites of pET-
28a(+) expression vector. The resultant plasmid was trans-

formed in E. coli strain DH10B. Insertion of the correct
gene ⁄ transcript was confirmed by DNA sequencing, restric-
tion digestion and colony PCR with gene ⁄ vector-specific
primers. This subcloned transcript was named rproErv-
C
+CT
and it was expressed using E. coli strain BL21(DE3).
Hexa-His-tagged recombinant protein expression was car-
ried out as described earlier [29] except that the cells were
grown for 5 h after induction with 0.5 m
M isopropyl b-D-
thiogalactoside instead of overnight.
The overexpressed recombinant rproErv-C
+CT
was puri-
fied by Ni-NTA affinity chromatography (Qiagen, Hilden,
Germany) under a denaturing condition and refolding of
the eluted purified protein was done as for rproErv-C
DCT
by dialysis method [29]. The refolded protein was concen-
trated by Amicon Ultra-4 (10 kDa cut-off) for further stud-
ies. The generation of rproErv-C
DC
clone and its expression
and purification was done as described previously [29].
Conversion of rproErv-C
+CT
to its mature form
Autocatalytic processing
In vitro conversion of the purified and refolded rproErv-

C
+CT
into its mature and active form was performed by
optimizing several parameters like proper activator (reduc-
ing agent), concentration of the activator, pH, temperature
and time. At a designated time interval, aliquots of the
sample were collected from the reaction mixture and mixed
with an equal volume of 2 · SDS ⁄ PAGE gel loading sam-
ple buffer containing 2–3 m
M irreversible inhibitor E-64
and analyzed by SDS ⁄ PAGE gel to optimize the time
required for complete maturation. Proenzyme processing of
rproErv-C
+CT
was compared with rproErv-C
DCT
by
SDS ⁄ PAGE analysis. Purity of the mature forms of rpro-
Erv-C
+CT
and rproErv-C
DCT
was established by western
blot analysis using rabbit antiserum raised against pure and
mature Erv-C from the plant latex as primary antibody
(Bangalore Genei, Bangalore, India) using the protocol
described earlier [29]. Wild-type Erv-C from plant latex was
also used for comparison.
Effect of external proteases on C-terminal processing
To determine whether the CT-extension of rproErv-C

+CT
can be cleaved by external proteases in trans mode, four
Fig. 9. Superposition of the catalytic cleft and the unprimed subsite
region of cysteine proteases blocked by different peptides like the
mini-chain in cathepsin H (Protein Data Bank ID
8PCH; sky blue),
CT-extension in bleomycin hydrolase (Protein Data Bank ID
1CB5;
green) and the modeled CT-extension in Erv-C (magenta). The E-64
molecule (brown) in complex with Erv-C (Protein Data Bank ID
2PRE) is also used in the superposition for comparison. The respec-
tive catalytic domains of the proteases are shown in Ca presenta-
tion in corresponding colors. The Ca atoms of the catalytic dyad
residues Cys and His are shown in yellow and blue, respectively.
Role of C-terminal extension in a cysteine protease S. Dutta et al.
3020 FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS
cysteine proteases [Erv-A, -B and –C (isolated from the
latex of Ervatamia coronaria in our laboratory) and papain
from Carica papaya (Merck, Kenilworth, NJ, USA)], two
serine proteases [trypsin and chymotrypsin from bovine
pancreas (Sigma-Aldrich, St. Louis, MO, USA)], one aspar-
tic protease [pepsin from porcine stomach mucosa (SRL,
Mumbai, India)] were used. Purified rproErv-C
+CT
(3 mg)
was digested using the abovementioned proteases (90 lg) at
the optimum pH and temperature of activity of each prote-
ase. In brief, digestion by Erv-A, Erv-B, Erv-C were carried
out at pH 8.0 and 37 °C for 2 h and blocked by E-64;
digestion by papain was carried out at pH 6.5 and 60 °C

for 0.5 h with 20 m
M cysteine, 2 mM EDTA and blocked
by E-64; trypsinization was carried out at pH 8.0 and
37 °C for 2 h and blocked by complete mini EDTA free
cocktail inhibitor (Roche, Mannheim, Germany); digestion
by chymotrypsin was carried out at pH 8.0 at 50 °C for 1 h
with 2 m
M CaCl
2
and blocked by complete mini EDTA free
cocktail inhibitor; pepsinization was carried out at pH 4.0
and 37 °C for 2 h and blocked by E-64 and pepstatin. Puri-
fied rproErv-C
+CT
was also digested using the abovemen-
tioned enzymes separately at 20 °C for 24 h under the same
conditions. The proteolytic digestion in each experiment
was checked by SDS ⁄ PAGE analysis.
Measurement of proteolytic activity
Substrate gel zymography using 0.1% gelatin as a substrate
was used to demonstrate the protease activity of recombi-
nant refolded proteases (rproErv-C
DCT
and rproErv-C
+CT
)
using a protocol described previously [29].
The specific activity of the recombinant proteases rpro-
Erv-C
DCT

and rproErv-C
+CT
was determined using sub-
strate azocasein. The specific activity of the native mature
Erv-C isolated from the plant latex was also determined
using the same protocol for comparison (Table 1). For this
assay, a reaction mixture containing 0.5 mL of 0.2% azoca-
sein in Tris ⁄ HCl buffer (pH 8.0), 0.5 mL of recombinant
proenzyme (10–20 lg) activated in Tris ⁄ HCl buffer
(pH 8.0) and incubated for 30 min. The reaction was then
terminated with 5% trichloroacetic acid. The mixture was
centrifuged at 9300 g for 5 min to remove precipitate and
the absorbance of the supernatant was measured at 366 nm
to determine the amount of released azopeptides using the
specific absorption coefficient (A
1%
366
= 40) for azocasein
solution [30]. One enzyme unit was defined as the amount
of soluble protease required to release 1 lg of soluble azo-
peptidesÆmin
)1
. The specific activity was the number of
units of activity per milligram of protein. For specific activ-
ity measurements the concentration of pro-protease has
been used in the calculations because once the proteases are
autocatalytically activated as described above, it is difficult
to isolate the active mature enzyme from degraded propep-
tide parts.
Determination of optimum temperature for

enzymatic activity of rproErv-C
DCT
and
rproErv-C
+CT
T
opt
or ‘temperature optima’ is the temperature at which
the enzyme shows maximum activity. Proteolytic activity of
both the recombinant proteins (rproErv-C
DCT
and rproErv-
C
+CT
) was measured in the range 30–90 at 5 °C intervals
to determine the optimum temperature of activity (T
opt
) for
both recombinant proteins. The pro-proteases were first
converted to mature enzymes and the proteolytic activity
was then measured using azocasein as a substrate as
described above at specific temperatures.
Kinetic measurements using a chromogenic
peptide
The mature form of rproErv-C
+CT
, activated from 1 lM
proenzyme rproErv-C
+CT
, was used for this assay. An

Erv-C-specific chromogenic substrate N-benzoyl-Phe-Val-
Arg-pNA (Sigma-Aldrich) [17] was used in this study for
kinetic measurements of the recombinant proteins. Liber-
ated pNA was monitored for 15 min at 410 nm on a
UV ⁄ Vis spectrophotometer (Nicolet Evolution 100;
Thermo Electron Corporation, Rockville, MD, USA).
Conditions for the measurement of the kinetic parameters
of rproErv-C
+CT
are as described earlier [29]. K
m
and
V
max
values of rproErv-C
+CT
were calculated by nonlinear
fitting of the Michaelis–Menten saturation curve using the
software Graphpad
PRISM ( />prism). The k
cat
value was calculated by using the equa-
tion k
cat
=V
max
⁄ [E]
T
where [E]
T

is the total concentration
of the active enzyme, the values of which were mea-
sured by active-site titration with the irreversible inhibitor,
E-64 using the above mentioned pNA containing sub-
strate, as described below. An extinction coefficient
of 8800
M
)1
Æcm
)1
at 410 nm for pNA was used for the
calculations.
Measurement of IC
50
value of E-64
Aliquots of 1 lM proenzyme (rproErv-C
+CT
and rproErv-
C
DCT
) were converted into their respective mature forms.
The irreversible inhibitor, E-64 was added in increasing
concentrations to the aliquots until the residual activity
reached 0. The residual activity (DA
410 nm
Æmin
)1
) was deter-
mined with respect to the activity of the enzyme (carried
out without any inhibitor) as described in the previous sec-

tion, against the peptide substrate (N-benzoyl-Phe-Val-Arg-
pNA) for both the proteases. This residual activity of the
enzyme was plotted against the inhibitor concentration.
The inhibitor concentration required for half-maximal inhi-
bition (IC
50
) of E-64 for these enzymes were determined
from these plots.
S. Dutta et al. Role of C-terminal extension in a cysteine protease
FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS 3021
Thermal stability of rproErv-C
DCT
and
rproErv-C
+CT
The effect of temperature on the kinetic stability of the
purified enzymes was investigated by incubating the pro-
teins at different temperatures. Purified proenzyme was
incubated at different temperatures in the range of
40–90 °Cat5°C intervals for 10 min each and then imme-
diately chilled on ice. The activity of the proteases was
measured at 65 °C (optimum temperature of activity for
the proteases) against azocasein as substrate, as described
previously. T
max
was expressed as the temperature of incu-
bation at which the enzyme showed maximum activity and
T
50
was expressed as the temperature of incubation at

which the enzyme showed 50% of maximum activity. The
half-life (t
1 ⁄ 2
) of an enzyme is the time it takes for the
residual activity to reduce to half of the original activity
after the enzyme is incubated at a particular temperature,
generally at T
max
. The t
1 ⁄ 2
values of two enzymes were esti-
mated by incubating the proteins at 65 °C(T
max
for both
the enzymes) for periods between 0 min and 8 h and then
the residual enzymatic activity was determined by azocasein
assay at 65 °C.
Molecular modeling of the mature Erv-C with
CT-extension (rmErv-C
+CT
)
The X-ray structure of mature domain of Erv-C, without
the CT-extension, was solved previously in our laboratory
(Protein Data Bank ID
2PNS). We have now performed a
BLASTP ( protein sequence
search of the amino acid sequence of CT-extension part
using RCSB Protein Data Bank ( />and found no significant sequence homology with any
structure in the database. Therefore, homology modeling to
generate the structure of the CT-extension domain did not

seem to be possible. However, we did find sequence homol-
ogy for the three stretches: the first residues, residues 11–
19, and the last four residues (21–24) of the CT-extension,
separately with some corresponding stretches of three dif-
ferent proteins having Protein Data Bank IDs
2X7X, 1PV8
and
3OHM respectively (Table S2). We therefore decided
to build the structures of the three stretches of the CT-
extension from the corresponding parts of the abovemen-
tioned Protein Data Bank structures after replacing mis-
matched residues appropriately. As a next step, we fitted
the three structural stretches individually at the mature
domain of Erv-C (Protein Data Bank ID
2PNS), as
described below, and finally the stretches were threaded by
the connecting amino acid residues (8–10 and 20) to gener-
ate the entire structural model of the CT-extension in asso-
ciation with mature Erv-C. The last four residues (24–21)
generated from the template (Protein Data Bank ID
3OHM, 444–447) were initially docked manually into the
unprimed subsites of Erv-C with Leu23 in the S2 pocket,
considering the specificity of Erv-C towards a Leu residue
[17]. The mode of interactions of mini-chain of human lyso-
somal cysteine protease cathepsin H (Protein Data Bank
ID
8PCH) and CT-extensions of human and yeast cysteine
proteases bleomycin hydrolases (Protein Data Bank IDs
2CB5 and 1GCB) with the unprimed subsites of their cog-
nate enzymes were also considered during this manual

docking. The docked tetrapeptide fragment of the CT-
extension was used as an anchor position and the next
stretch of 19–11 residues, which showed a helical nature
from both the template conformation (Protein Data Bank
ID
1PV8, residues 320–329 of chain B) and the secondary
structure prediction using the program
DSC (Discrimination
of Protein Secondary Structure Class) [31] implemented in
DISCOVERY STUDIO 2.5 (Accelrys Inc., San Diego, CA,
USA), were positioned and fitted, taking into account
charge compatibility and hydrophobic interactions with
respect to the structure of mature Erv-C. The template
(Protein Data Bank ID
2X7X, residues 44–50 of chain A)
of the first seven residues of the CT-extension is a connect-
ing polypeptide between a b sheet and an a helix. In the
structure of the mature domain of Erv-C (Protein Data
Bank ID
2PNS) the C-terminus ends with a b sheet and
therefore we connected this seven-residue fragment of the
CT-extension to this b sheet of Erv-C here also, and ori-
ented it towards the previously fitted helical stretch (19–11).
Finally, these three fragments were joined by the connecting
sequences (residues 8–10 and 20) and the rotamer confor-
mation of the side chain of each residue was optimized.
The resulting complex of Erv-C together with the CT-exten-
sion was then globally optimized using ‘Smart Minimizer’
protocol in
DISCOVERY STUDIO 2.5 (Accelrys Inc.) keeping

the Ca position of the mature domain of Erv-C fixed. The
model was then solvated with a 30 A
˚
water sphere. A few
cycles of minimization were performed without any restric-
tion to optimize the model and the position of the water
molecules. In the last step, the entire assembly was simu-
lated at 300 K for 2.6 ns using ‘Standard Dynamic Cas-
cade’ in
DISCOVERY STUDIO (40 ps heating from 50 to
300 K, 20 ps equilibration at 300 K and 2 ns production
run at 300 K), keeping all Ca atoms of the mature part of
the enzyme fixed. Simulations were carried out at constant
volume and temperature (NVT ensemble) through the
Leapfrog Verlet integrator. A time-step of 2 fs for integra-
tion was used and a bond constraint was applied for all
covalent bonds with H atoms through the SHAKE algo-
rithm implemented in the program. Coordinates were saved
at 10-ps intervals for a 2-ns product run for analysis. All
simulations were carried out with CHARMm force field
implemented in
DISCOVERY STUDIO 2.5 with a non-bonded
cut-off of 14 A
˚
. The last 100-ps average structure from the
simulation was further optimized for 100 steps Steepest
Descent without any restraint and used as the model for
structural interpretations in this study. In the model thus
generated, the length of the helix of CT-extension was
Role of C-terminal extension in a cysteine protease S. Dutta et al.

3022 FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS
reduced (residues 12–16) from the starting structure (resi-
dues 11–19). At the end of the simulation, the CT-extension
was isolated by deleting the mature domain and the exten-
sion was allowed to relax in a same simulation protocol as
described above for comparison. The simulation trajectory
was analyzed by ‘Analysis’ protocol of
DISCOVERY STUDIO
2.5.
The geometrical and structural consistencies of the model
structure of Erv-C together with CT-extension were evalu-
ated by different approaches. The stereochemical quality of
the model was evaluated using
PROCHECK [32] and VERIFY3D
[33]. This structure was further validated using web-based
software
QMEAN [34], PROQ [35] and SOLX [36]. All these val-
idations (Table S2) signify that the model is of good quality
and is reliable.
Acknowledgement
The work was partially supported by CSIR (Grant
No. 21 ⁄ 0653 ⁄ 06 ⁄ EMR-II), Govt. of India.
References
1 Haki GD & Rakshit SK (2003) Developments in indus-
trially important thermostable enzymes: a review.
Bioresour Technol 89, 17–34.
2 Wiederanders B, Kaulmann G & Schilling K (2003)
Functions of propeptide parts in cysteine proteases.
Curr Protein Pept Sci 4, 309–326.
3 Cygler M & Mort JS (1997) Proregion structure of

members of the papain superfamily. Mode of inhibition
of enzymatic activity. Biochimie 79, 645–652.
4 Turk B, Turk D & Turk V (2000) Lysosomal cysteine
proteases: more than scavengers. Biochim Biophys
Acta 1477, 98–111.
5 Kamphuis IG, Drenth J & Baker EN (1985) Thiol
proteases. Comparative studies based on the high-reso-
lution structures of papain and actinidin, and on amino
acid sequence information for cathepsins B and H, and
stem bromelain. J Mol Biol 182, 317–329.
6 Yamada K, Matsushima R, Nishimura M & Hara-
Nishimura I (2001) A slow maturation of a cysteine
protease with a granulin domain in the vacuoles of
senescing Arabidopsis leaves. Plant Physiol 127,
1626–1634.
7 Paul W, Amiss J, Try R, Praekelt U, Scott R & Smith
H (1995) Correct processing of the kiwifruit protease
actinidin in transgenic tobacco requires the presence
of the C-terminal propeptide. Plant Physiol 108, 261–
268.
8 Grudkowska M & Zagdanska B (2004) Multifunctional
role of plant cysteine proteinases. Acta Biochim Pol 51,
609–624.
9 Nakamura K & Matsuoka K (1993) Protein targeting
to the vacuole in plant cells. Plant Physiol 101, 1–5.
10 Okamoto T, Yuki A, Mitsuhashi N & Mimamikawa T
(1999) Asparaginyl endopeptidase (VmPE-1) and auto-
catalytic processing synergistically activate the vacuolar
cysteine proteinase (SH-EP). Eur J Biochem 264,
223–232.

11 Villa MM, Vargas-Inchausteguic DA, Chaves SP, Tem-
ponea AJ, Dutraa JMF, Soarese MJ, Ueda-Nakamuraf
T, Mendonc SCF, Rossi-Bergmannd B, Soongc L et al.
(2008) The C-terminal extension of Leishmania pifanoi
amastigote-specific cysteine proteinase Lpcys2: a puta-
tive function in macrophage infection. Mol Biochem
Parasitol 162, 52–59.
12 Zheng W, Johnston SA & Joshua-Tor L (1998) The
unusual active site of Gal6 ⁄ bleomycin hydrolase can act
as a carboxypeptidase, amino peptidase and peptode
ligase. Cell 93, 103–109.
13 Koldamova RP, Lefterov IM, Gadjeva VG & Lazo JS
(1998) Essential binding and functional domains of
human bleomycin hydrolase. Biochemistry 37, 2282–
2290.
14 Sundd M, Kundu S, Pal G & Jagannadham MV (1998)
Purification and characterization of a highly stable cys-
teine protease from the latex of Ervatamia coronaria.
Biosci Biotechnol Biochem 62, 1947–1955.
15 GuhaThakurta P, Biswas S, Chakrabarti C, Sundd M,
Jagannadham MV & Dattagupta JK (2004) Structural
basis of the unusual stability and substrate specificity of
ervatamin-C, a plant cysteine protease from Ervata-
mia coronaria. Biochemistry 43 , 1532–1540.
16 Ghosh R, Dattagupta JK & Biswas S (2007) A thermo-
stable cysteine protease precursor from a tropical plant
contains an unusual C-terminal propeptide: cDNA
cloning, sequence comparison and molecular modeling
studies. Biochem Biophys Res Commun 362
, 965–970.

17 Ghosh R, Chakraborty S, Chakrabarti C, Dattagupta
JK & Biswas S (2008) Structural insight into the
substrate specificity and activity of Ervatamins: the
papainlike cysteine proteases from a tropical plant
Ervatamia coronaria. FEBS J 275, 421–434.
18 Bromme D, Nallaseth FS & Turk B (2004) Production
and activation of recombinant papain-like cysteine pro-
teases. Methods 32, 199–206.
19 Groves MR, Taylor MAJ, Scott M, Nicola J, Richard
C, Pickersgill W & Jenkins JA (1996) The prosequence
of procaricain forms an a-helical domain that prevents
access to the substrate-binding cleft. Structure 4 ,
1193–1203.
20 LaLonde JM, Zhao B, Janson CA, D’Alessio KJ,
McQueney MS, Orsini MJ, Debouck CM & Smith WW
(1999) The crystal structure of human procathepsin K.
Biochemistry 38, 862–869.
21 Coulombe R, Grochulski P, Sivaraman J, Menardl R,
Mort JS & Cygler M (1996) Structure of human procat-
hepsin L reveals the molecular basis of inhibition by the
prosegment. EMBO J 15, 5492–5503.
S. Dutta et al. Role of C-terminal extension in a cysteine protease
FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS 3023
22 Kaulmann G, Palm GJ, Schilling K, Hilgenfeld R &
Wiederanders B (2006) The crystal structure of a
Cys25!Ala mutant of human procathepsin S elucidates
enzyme–prosequence interactions. Protein Sci 15,
2619–2629.
23 Cygler M, Sivaraman J, Grochulski P, Coulombe R,
Storer AC & Mort JS (1996) Structure of rat procathep-

sin B: model for inhibition of cysteine protease activity
by the proregion. Structure 4, 405–416.
24 Luis H, Gonza
´
lez G, Rojo-Domı
´
nguez A, Cabrera-
Gonza
´
lez NE, Pe
´
rez-Montfort R & Padilla-Zu´ niga AJ
(2006) Loosely packed papain prosegment displays
inhibitory activity. Arch Biochem Biophys 446, 151–160.
25 Fairhead M & Thony-Meyer L (2010) Role of the
C-terminal extension in a bacterial tyrosinase. FEBS J
277, 2083–2095.
26 Vasiljeva O, Dolinar M, Turk V & Turk B (2003)
Recombinant human cathepsin H lacking the
mini chain is an endopeptidase. Biochemistry 42,
13522–13528.
27 Turk D, Janjic V, Stern I, Podobnik M, Lamba D,
Dahl SW, Lanritzen C, Pederson J, Turk V & Turk B
(2001) Structure of human dipeptidyl peptidase I
(cathepsin-c): exclusion domain added to an endopepti-
dase framework creates the machine for activation of
granular serine proteases. EMBO J 20, 6570–6582.
28 Palusa SG, Ali GS & Reddy ASN (2007) Alternative
splicing of pre-mRNAs of Arabidopsis serine ⁄ arginine-
rich proteins: regulation by hormones and stresses.

Plant J 49, 1091–1107.
29 Dutta S, Ghosh R, Dattagupta JK & Biswas S (2010)
Heterologous expression of a thermostable plant
cysteine protease in Escherichia coli both in soluble and
insoluble forms. Process Biochem 45, 1307–1312.
30 Chen CY, Luo SC, Kuo CF, Lin YS, Wu JJ, Lin MT,
Liu CC, Jeng WY & Chuang WJ (2003) Maturarion
processing and characterization of streptopain. J Biol
Chem 278, 17336–17343.
31 King RD & Sternberg MJE (1996) Identification and
application of the concepts important for accurate and
reliable protein secondary structure prediction. Protein
Sci 5, 2298–2310.
32 Laskowski RA, MacArthur MW, Moss DS & Thornton
JM (1993) PROCHECK: a program to check the
stereochemical quality of protein structures. J Appl
Crystallogr 26, 283–291.
33 Luthy R, Bowie JU & Eisenberg D (1992) Assessment
of protein models with three-dimensional profiles.
Nature 356, 83–85.
34 Benkert P, Tosatto SCE & Schomburg D (2008)
QMEAN: a comprehensive scoring function for model
quality assessment. Proteins 71, 261–277.
35 Wallner B & Elofsson A (2003) Can correct protein
models be identified? Protein Sci 12, 1073–1086.
36 Holm L & Sander C (1992) Evaluation of protein models
by atomic solvation preference. J Mol Biol 225, 93–105.
Supporting information
The following supplementary material is available:
Fig. S1. SDS ⁄ PAGE showing C-terminal processing of

rproErv-C
+CT
by external proteases.
Fig. S2. The mature Erv-C with CT-extension.
Table S1. Yield of rproErv-C
+CT
and rproErv-C
DCT
from 1 L culture at different purification steps from
inclusion bodies.
Table S2.
CLUSTALW alignment and structural valida-
tions of the model structure of CT-extension.
Table S3. Raw data sets related to thermal stability
experiments.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
Role of C-terminal extension in a cysteine protease S. Dutta et al.
3024 FEBS Journal 278 (2011) 3012–3024 ª 2011 The Authors Journal compilation ª 2011 FEBS