Substrate-Induced Conformational Change and Isomerase Activity of Dienelactone Hydrolase and its Site-Specific Mutants

Article abstract of DOI:10.1002/cbic.201200232

Studies of the interactions of dienelactone hydrolase (DLH) and its mutants with both E and Z dienelactone substrates show that the enzyme exhibits two different conformational responses specific for hydrolysis of each of its substrate isomers. DLH facili

Full text of DOI:10.1002/cbic.201200232

DOI: 10.1002/cbic.201200232
Substrate-Induced Conformational Change and Isomerase
Activity of Dienelactone Hydrolase and its Site-Specific
Mutants
[
a]
Ian Walker, James E. Hennessy, David L. Ollis, and Christopher J. Easton*
Studies of the interactions of dienelactone hydrolase (DLH)
and its mutants with both E and Z dienelactone substrates
show that the enzyme exhibits two different conformational
responses specific for hydrolysis of each of its substrate iso-
mers. DLH facilitates hydrolysis of the Z dienelactone through
an unusual charge-relay system that is initiated by interaction
between the substrate carboxylate and an enzyme arginine
residue that activates an otherwise non-nucleophilic cysteine.
The E dienelactone does not display this substrate–arginine
binding interaction, but instead induces an alternate confor-
mational response that promotes hydrolysis. Furthermore, the
substitution of cysteine 123 for serine (C123S) in DLH, instead
of inactivating the enzyme as is typical for this active-site mu-
tation, changes the catalysis from substrate hydrolysis to iso-
merisation. This is due to the deacylation of the acyl–enzyme
intermediates being much slower, thereby increasing their life-
times and allowing for their interconversion through isomerisa-
tion, followed by relactonisation.
Introduction
Dienelactone hydrolase (E.C. 3.1.1.45) from Pseudomonas sp.
B13 (DLH type III) is a monomeric enzyme found on the chloro-
[
1]
catechol branch of the b-ketoadipate pathway, serving an
important role in the biodegradation of toxic aromatic com-
[
2]
pounds. It catalyses the hydrolysis of dienelactones 1 and 2
[
3]
to yield maleylacetate 3 (Scheme 1).
Scheme 2. Mechanism of DLH activation by Z dienelactone 2.
Scheme 1. Hydrolysis of E and Z dienelactones 1 and 2 catalysed by DLH,
and the substrate analogue, Z dienelactam 4.
late due to a substrate-induced activation mechanism occur-
[7]
ring within the DLH active site (Scheme 2). Accordingly, it
was proposed that in the resting state, the C123 neutral thiol
is positioned pointing towards the interior of the enzyme
away from the active site. In the active state, this thiol is acti-
vated to the thiolate and rotated around its CaÀCb bond to
point into the active site, aligned with the H202 imidazolium
and positioned to attack the acyl carbon of substrates 1 and 2.
In the absence of substrate, an ion pair between R206 and E36
Crystal-structure analysis has confirmed the presence of
a DLH catalytic triad consisting of an aspartate (D171), a histi-
[
4,5]
dine (H202) and a nucleophilic cysteine (C123).
A mecha-
nism of substrate hydrolysis was therefore proposed resem-
bling that of the cysteine proteases with a resting-state ion
pair between the C123 thiolate and the H202 imidazolium fa-
cilitating nucleophilic attack on dienelactone substrates 1 and
[
5,6]
2
.
However, further crystal-structure analyses and inhibitor
[
a] Dr. I. Walker, Dr. J. E. Hennessy, Prof. Dr. D. L. Ollis, Prof. Dr. C. J. Easton
Research School of Chemistry, Australian National University
Canberra, ACT 0200 (Australia)
binding studies using Z dienelactam substrate analogue 4 later
indicated that, in fact, C123 is present as the inactive thiol in
the resting state and is only converted to the nucleophilic thio-
E-mail: easton@rsc.anu.edu.au
ChemBioChem 2012, 13, 1645 – 1651
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1645

C. J. Easton et al.
prevents deprotonation of the C123 thiol by E36. Rotation of
the C123 side chain around its CaÀCb bond in the absence
of substrate is also hindered due to a close steric interaction
Results
DLH and its mutants that retain cysteine at the active site hy-
drolyse Z dienelactone 2 to maleylacetate 3. These interactions
follow simple Michaelis–Menten kinetics, and the data show
linear correlations on both Hanes and Lineweaver–Burk plots.
The kinetic parameters for the interactions were determined
using nonlinear-regression analysis, and the values are shown
in Table 1. These proteins also hydrolyse E dienelactone 1 to
maleylacetate 3, but in some of these cases a pronounced de-
viation from Michaelis–Menten kinetics is apparent at low sub-
strate concentrations. This is particularly marked with DLH and
its R206A mutant, for which the corresponding data are illus-
trated in Figure 1. Nonlinear regression was used to derive the
kinetic parameters for the hydrolysis of lactone 1, which are
shown in Table 2. These values were not significantly affected
by either including or excluding the data obtained at low sub-
strate concentrations.
(
<1.5 ꢁ) between the side chain methylene of C123 and the
backbone amide of L124. However, upon binding of the sub-
strate, the ion pair between R206 and E36 is weakened, due to
a new ion pair formed between R206 and the substrate car-
boxylate. This allows E36 to shift towards the C123 thiol and
abstract its proton. Free rotation of this newly generated thio-
late to align with H202 in the active conformation is also pro-
moted by the L124 amide being displaced away from the
C123 methylene by the substrate-ring carbonyl oxygen.
This mechanism was proposed based on crystal-structure
[
7]
analysis of both DLH and its C123S mutant. DLH C123S was
used as a model for inhibitor binding studies to overcome
problems of C123 oxidation during X-ray crystallographic anal-
[
5]
ysis of DLH. Chemical evidence for this mechanism was ob-
served in the unusual resistance of the DLH C123 side chain to
chemical modification, corroborating its location in a poorly
accessible region towards the interior of the enzyme, in the
[
7]
absence of substrate. Additional theoretical analysis has also
supported the occurrence of substrate-induced DLH activa-
Table 1. Kinetic parameters for interactions of DLH and a range of its
active site mutants with E and Z dienelactones 1 and 2.
[
a]
[
8]
tion.
Z dienelactone 2
E dienelactone 1
À1
À1
We now report a mutagenic analysis of the DLH active site
that provides a clearer picture of this unusual catalytic mecha-
nism. This has involved a detailed kinetic investigation of the
catalytic activities of DLH and DLH C123S and their mutants
with both E and Z dienelactones 1 and 2. Whilst the compara-
tive kinetic analyses of DLH and its site-specific mutants cor-
roborate the proposed substrate-induced enzyme-activation
hypothesis for hydrolysis of Z dienelactone 2, we find that this
mechanism does not apply with E dienelactone 1, with R206
having no binding interaction with this substrate. It is instead
apparent that, in order to hydrolyse E lactone 1, DLH under-
goes a substrate-induced conformational change different to
that observed during hydrolysis of Z isomer 2. Given
the important role of R206 in the substrate-induced
activation of DLH upon binding Z lactone 2, it is sur-
prising that DLH is at all active towards E isomer 1, in
the absence of a binding interaction with R206. Even
more intriguing is the ability of DLH to exhibit differ-
ent conformational responses specific for hydrolysis
of each of its substrate isomers 1 and 2.
kcat [min
]
Km [mm]
kcat [min
]
Km [mm]
DLH
1120Æ15
50Æ1
11Æ1
18Æ1
870Æ20
44Æ1
180Æ10
260Æ10
240Æ20
1550Æ90
170Æ10
210Æ10
240Æ10
1730Æ80
390Æ10
6870Æ70
6190Æ280
DLH E36D
DLH R206A
DLH R81A
DLH R206K
DLH R81K
DLH Y85F
DLH W88A
DLH S203A
DLH S203H
DLH S203D
37Æ1
350Æ30
160Æ5
25Æ1
6.0Æ0.2
1070Æ10
96Æ1
650Æ15
41Æ1
940Æ20
440Æ10
270Æ5
510Æ10
26Æ1
1250Æ10
670Æ10
690Æ20
820Æ10
450Æ10
16Æ1
19Æ1
340Æ20
1.9Æ0.1
160Æ10
12Æ0.5
5070Æ400
4.0Æ0.1
[
a] In HEPES buffer (20 mm, pH 7.0) containing 1 mm EDTA at 258C.
The C123S mutation changes the catalytic triad of
DLH to an aspartate (D171), a histidine (H202) and
a serine (S123), which is also observed with some
[
9,10]
serine proteases.
Our preliminary studies of DLH
C123S showed quite surprising activity. Whereas DLH
catalyses only the hydrolysis of dienelactones 1 and
Figure 1. Hanes plots demonstrating the deviation from Michaelis–Menten kinetics for
hydrolysis of E dienelactone 1 by A) DLH and B) DLH R206A, in HEPES buffer (20 mm,
pH 7.0) containing 1 mm EDTA at 258C.
2
, DLH C123S promotes a completely different class
of reaction and instead catalyses the isomerisation of
[
11]
these same substrates without hydrolysis.
A de-
tailed analysis of the interactions of lactones 1 and 2
with site-specific mutants of both DLH and DLH C123S now
allows us to also provide an explanation of this unusual behav-
iour.
Interactions of DLH C123S and of the double mutants DLH
C123S/R206A and DLH C123S/R81A with lactones 1 and 2
were also studied. In initial experiments, it was observed that
the UV absorbance at 280 nm of a solution of Z lactone 2 and
1
646
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2012, 13, 1645 – 1651

Dienelactone Hydrolase
DLH C123S increased with time, which is clearly inconsistent
with hydrolysis of substrate 2 to maleylacetate 3. Closer analy-
sis using HPLC showed that instead Z lactone 2 was being con-
verted to E isomer 1. Consequently the ratios of lactones
substantially deviate from linearity when correlated using
a Hanes plot (Figure 3), but the relevant kinetic parameters
1
and 2 and acetate 3 in solutions, beginning with either of
the substrates (1 or 2) and either one of the C123S, C123S/
R206A and C123S/R81A mutants of DLH, were determined
using HPLC, as illustrated in Figure 2. For comparison, the cor-
1
responding data for DLH are also shown. H NMR spectroscopy
and thin layer chromatography were used to confirm the inter-
conversion between lactones 1 and 2, as well as their hydroly-
sis to maleylacetate 3.
The interactions of Z lactone 2 with DLH C123S, C123S/
R206A and C123S/R81A obey Michaelis–Menten kinetics, and
the constants characterising this behaviour were determined
using nonlinear-regression analysis and are shown in Table 2.
The data for the interaction of E lactone 1 with DLH C123S
Figure 3. Hanes plot demonstrating the deviation from Michaelis–Menten ki-
netics for isomerisation of E dienelactone 1 by DLH C123S, in HEPES buffer
(
20 mm, pH 7.0) containing 1 mm EDTA at 258C.
[
a]
Table 2. Kinetic parameters for the interactions of DLH C123S, DLH C123S/R206A and DLH C123S/R81A with E and Z dienelactones 1 and 2.
[
b]
k
cat
k’cat
(hydrolysis)
[min
K
m
Ratio
v
v’
Ratio
v/v’
[
c]
[c]
(isomerisation)
[mm]
k
cat/k’cat
(isomerisation)
[ꢂ10 min
(hydrolysis)
[ꢂ10 min
À1
À1
2
À1
2
À1
[min
]
]
]
]
Z dienelactone 2
DLH C123S
DLH C123S/R206A
DLH C123S/R81A
E dienelactone 1
DLH C123S
11Æ0.5
1.4Æ0.1
4.6Æ0.3
<0.01
0.4+0.1
0.35+0.03
2.7Æ0.1
36Æ4
>1100
3.5
660Æ1
<0.1
>6600
5.0Æ0.1
12Æ0.5
1.3Æ0.05
0.8Æ0.05
4
15
8.6Æ1.5
13
38Æ1.0
<0.01
4.4Æ0.2
>3800
620Æ80
9.6Æ0.1
41Æ2
<0.1
>6200
16
DLH C123S/R206A
DLH C123S/R81A
0.60Æ0.01
3.0Æ0.1
14
[
a] In HEPES buffer (20 mm, pH 7.0) containing 1 mm EDTA at 258C. [b] Isomerisation and hydrolysis occur with the same K , implying that both processes
m
h
involve a common binding event. For lactone 1, K
m
was determined as K0 by nonlinear regression to best fit the data to the Hill equation. [c] Determined
:5
by using an initial substrate concentration of 6.8 mm.
[
a]
[b]
Figure 2. Concentrations of E and Z dienelactones 1 (
), 2 (*) and maleylacetate 3 (ꢂ) in mixtures obtained by treatment of A) lactone 2 with DLH
&
(
(
12 nm), B) lactone 1 with DLH (20 nm), C) lactone 2 with DLH C123S (10.7 mm), D) lactone 1 with DLH C123S (10.7 mm), E) lactone 2 with DLH C123S/R206A
80 mm), F) lactone 1 with DLH C123S/R206A (80 mm), G) lactone 2 with DLH C123S/R81A (59.7 mm), H) lactone 1 with DLH C123S/R81A (59.7 mm). [a] Deter-
mined by HPLC. [b] In HEPES buffer (20 mm, pH 7.0) containing 1 mm EDTA at 258C.
ChemBioChem 2012, 13, 1645 – 1651
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1647

C. J. Easton et al.
[
12]
were obtained using the Hill equation (Table 2). DLH C123S/
R206A and C123S/R81A bind lactone 1 only weakly, and it was
therefore impractical to prepare solutions containing sufficient
concentrations of lactone 1 to be able to determine the cata-
lytic rate constant (k ) and Michaelis–Menten constant (K )
tant contribution of E36 and R206 to the hydrolysis of lactone
1. In any case, it is clear that events facilitating hydrolysis of E
dienelactone 1 by DLH are different from those for Z isomer 2.
The roles of R81 and R206 of DLH to bind Z dienelactone 2
and of R81 to bind E dienelactone 1, appear to be maintained
by lysine incorporated in DLH R81K and DLH R206K. The Km
values for the three proteins interacting with lactone 2 are
quite similar, as are those for DLH and DLH R81K with lactone
1. However, the lysine mutant DLH R206K does not retain the
catalytic activity, as demonstrated by its much lower kcat values
than those for DLH for processing both substrates. Previous
crystallographic analysis of DLH and DLH C123S with the
bound inhibitor, Z dienelactam 4, indicated that Y85 is posi-
tioned with its hydroxyl group in the active site, 3.74 ꢁ from
cat
m
values for these interactions. Instead, the specific activities of
these enzymes in isomerising and hydrolysing lactone 1 were
determined using an initial substrate concentration of 6.8 mm.
For comparison, the corresponding data were obtained for the
interactions of DLH C123S with lactone 1, and DLH C123S,
C123S/R206A and C123S/R81A with lactone 2 (Table 2).
Discussion
[13]
The kcat and Km values in Table 2 are consistent with the mech-
anism proposed for the hydrolysis of Z dienelactone 2 by DLH
Scheme 2). Both R81 and R206 are shown to be important for
the exocyclic methine of the inhibitor. In this position, Y85
might therefore stabilise enolate formation during ring open-
ing of lactones 1 and 2 by proton donation. However, substi-
tuting this tyrosine for phenylalanine and thereby removing
the hydroxyl group had little effect on either the kcat or Km
values of DLH Y85F compared to those of DLH with either sub-
strate. Therefore, should this tyrosine be involved in the re-
action as an acid catalyst, it is not during a rate-determining
step. A substrate-binding role for W88 is also apparent, as a de-
crease in binding affinities is observed when that tryptophan is
replaced with alanine in DLH W88A. W88 is the only residue
found in close proximity to the substrate ring hydrogens of Z
dienelactam 4 bound in the DLH active site. It therefore
appears that W88 is able to stabilise the Michaelis complex
through the formation of p–p interactions with substrates
1 and 2. Previously it has been proposed that the S203 hydrox-
yl group is positioned to bind to the carboxylate of Z dienelac-
(
substrate binding, since their replacement with alanine in DLH
R81A and R206A increases the K value more than tenfold. The
m
role of R206 and E36, but not R81, in catalysing reaction of the
bound substrate is illustrated by the close similarity in the kcat
values for DLH and the R81A mutant, and the much lower
values for DLH R206A and E36D. This is consistent with R206
and E36 being involved in the activation of C123 for nucleo-
philic attack on lactone 2. By contrast, the kinetic parameters
summarised in Table 1 are not in accord with the same mecha-
nism operating for the hydrolysis of dienelactone (E)-1 by DLH.
Whilst the binding role of R81 observed for lactone 2 is also
shown for lactone 1, R206 does not retain substrate-binding
properties. This is apparent, since the Km values of DLH and
DLH R206A are similar, while that of DLH R81A is substantially
higher. R206 does however retain its catalytic role for reaction
with lactone 1 once bound, which is shown by the kcat value
for DLH R206A being less than 1% of that of DLH. A large
reduction in kcat for DLH E36D confirms this glutamate is also
important for reaction of lactone 1.
These results indicate that the mechanism illustrated in
Scheme 2 is not adequate to explain the hydrolysis of E lac-
tone 1. For reaction of Z isomer 2, its binding to R206 is cou-
pled to the role of R206 and E36 in deprotonating C123, and
the realignment of the resultant thiolate with H202 and D171
of the catalytic triad for hydrolysis of the bound substrate.
However, this is not the case for E isomer 1, where binding
seems not to involve R206. Nevertheless, for hydrolysis of this
substrate, there must be a rearrangement of the active-site res-
idues from their orientations observed in the crystal structure
of free DLH to those required for catalysis. The deviations from
Michaelis–Menten kinetics observed for interactions of E lac-
tone 1 with DLH and its mutants (Figure 1) indicate that this
occurs through the enzyme undergoing a conformational tran-
sition on substrate binding. These deviations only occur at low
substrate concentrations, where the enzyme reverts to its inac-
tive form between binding events. It seems likely that this sub-
strate-induced conformational transition disrupts the ion pair
between R206 and E36, thus allowing C123 activation without
formation of an R206–substrate carboxylate ion pair, since the
kcat values of DLH, DLH E36D and DLH R206A show an impor-
[7]
tone 2. Under these circumstances, substituting the serine for
alanine in DLH S203A would have been expected to lead to
a decrease in catalytic efficiency. However, in terms of k /K ,
cat
m
the mutation has produced an enzyme that is four times more
efficient in processing lactone 2. It appears that rather than
providing a substrate-binding role, S203 instead weakens sub-
strate binding through interaction with the side chain of R206.
Mutation of S203 to alanine therefore releases the R206 guani-
dinium to enhance binding of lactone 2, as seen in the tenfold
reduction in Km for DLH S203A with this substrate (Table 1).
This binding effect is not observed during hydrolysis of E dien-
elactone 1, since this substrate does not interact with R206.
The only reaction of lactones 1 and 2 observed with DLH
and its mutants that retain C123 is hydrolysis to maleylacetate
3. There is no evidence of interconversion between substrates
1 and 2 in the presence of these enzymes (Figure 2A and B for
DLH). In complete contrast, DLH C123S catalyses the intercon-
version of lactones 1 and 2 to the equilibrium position, where
they are present in a 53:47 ratio, and negligible hydrolysis of
either substrate is discernible, even long after equilibration of
substrates 1 and 2 is complete (Figure 2C and D). On this
basis, the kcat values for hydrolysis of lactones 1 and 2 by DLH
C123S are at least 1000 times less than those for isomerisation.
DLH C123S/R206A and C123S/R81A also catalyse the intercon-
version of dienelactones 1 and 2 (Figure 2E–H). Again this is
the dominant reaction in each case, but hydrolysis to maleyla-
1
648
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2012, 13, 1645 – 1651

Dienelactone Hydrolase
cetate 3 is also observed as a competing process. The ratios of
the kcat values for isomerisation and hydrolysis are 3.5 and 13
for the interaction of lactone 2 with DLH C123S/R206A and
C123S/R81A, respectively. The corresponding ratios of the iso-
merisation and hydrolysis rates for reactions of a 6.8 mm solu-
tion of E lactone 1 with DLH C123S/R206A and C123S/R81A
are 16 and 14, respectively. Compared to DLH C123S, DLH
change of reaction pathway from hydrolysis to isomerisation,
observed with DLH and DLH C123S, is unprecedented.
It is conceivable that DLH C123S, DLH C123S/R206A and
DLH C123S/R81A catalyse the isomerisation by a reversible Mi-
chael addition of the S123 nucleophile onto the exocylic acry-
late moiety of lactones 1 and 2. However, this seems unlikely,
because a similar process involving the thiolate of DLH and its
mutants that retain C123 would also be expected. Further,
analysis of the crystal structure of Z dienelactam 4-bound DLH
C123S indicates that the distance between the serine and the
acrylate is too great for nucleophilic attack (4.2 ꢁ). Alternative-
ly, it seems more reasonable that, whereas the reactions cata-
lysed by DLH and its mutants that retain C123 are understood
to proceed via thioesters 5a–7a, DLH C123S, DLH C123S/
R206A and DLH C123S/R81A are instead acylated by lactones
1 and 2 to give the corresponding esters 5b–7b (Scheme 3).
Hydrolysis of thioesters 5a–7a then proceeds through forma-
tion of the tetrahedral intermediate 8a, which collapses with
preferential loss of the best leaving group, the enzyme thiol or
thiolate, to give maleylacetate 3. By contrast, the tetrahedral
intermediate 8b for hydrolysis of the esters 5b–7b preferen-
tially expels water or hydroxide, thus reforming the esters 5b–
7b and making them more resistant to hydrolysis, as observed
with the acylated serine mutants of UDP-glucose dehydrogen-
C123S/R206A exhibits a 13-fold increase in K and an eightfold
m
decrease in kcat for isomerisation of lactone 2, whilst DLH
C123S/R81A exhibits a threefold increase in Km and a twofold
decrease in kcat for isomerisation of this same substrate.
Though more modest, these effects of the DLH C123S muta-
tion on the Km values are in line with those observed with
identical mutations of DLH (Table 1), in that they indicate the
roles of R206 and R81 in facilitating binding of lactone 2 are
shared by both DLH and DLH C123S. As observed during the
reaction of lactone 1 catalysed by DLH, the deviations from Mi-
chaelis–Menten kinetics (Figure 3) indicate that DLH C123S-cat-
alysed isomerisation of E dienelactone 1 is dependent on a sub-
strate-induced conformational change occurring in the enzyme
active site, and this most likely corresponds to a rearrangement
of the active site residues from their orientations in the most
stable, catalytically inactive state to those required for cataly-
sis.
[17]
Any isomerisation of lactones 1 and 2 catalysed by DLH and
its mutants that retain C123 would have been detectable, par-
ticularly in the reactions of Z lactone 2. With most of these en-
zymes, E lactone 1 is hydrolysed much less efficiently (k /K )
ase.
This increases the lifetime of esters 5b–7b, allowing
more time for their interconversion, leading to isomerisation,
before recyclisation.
cat
m
than Z isomer 2. Consequently, if lactone 1 formed through
simultaneously occurring isomerisation and hydrolysis of Z
isomer 2, it would accumulate as a proportion of lactones
1
and 2 remaining, at least up to the equilibrium value of
53%. None was observed, even when mixtures were analysed
after more than 90% of lactone 2 had hydrolysed, and the
limits of detection of lactone 2 as a fraction of total residual
substrate concentration were then less than 1%. This indicates
that, should DLH and its mutants that retain C123 catalyse sub-
strate isomerisation, the rate of this process is negligible in
comparison to that of substrate hydrolysis. This is completely
different to DLH C123S, with this enzyme instead catalysing
substrate isomerisation with no evidence of substrate hydroly-
sis.
There have been other reports of the conversion of an en-
[
14–19]
zyme’s active-site cysteine to serine.
showed greatly reduced catalytic activity or were inactive. With
The resultant proteins
[
17]
UDP-glucose dehydrogenase, it was found that replacing the
active-site C260 with serine gave a modified protein, that
formed an ester instead of a thioester with the substrate, but
the ester was not labile and, therefore, the mutant protein
showed no substrate turnover. Formation of an unreactive co-
valent adduct was also indicated as a result of replacing the
Scheme 3. Proposed mechanisms for the reactions of lactones 1 and 2 cata-
lysed by DLH and its mutants.
[18]
active-site C55 with serine in nitrile oxidoreducatase QueF. In
these cases, the wild-type enzymes and their serine mutants
both form covalent enzyme–substrate intermediates, with
adduct breakdown being slowed or stopped with the mutants.
Nevertheless, there is a similarity in mechanism, whereas the
Consistent with this rationale, the rates of hydrolysis of lac-
tones 1 and 2 by DLH C123S, DLH C123S/R206A and DLH
C123S/R81A are at least five orders of magnitude lower than
those with DLH. The totality of the data shown in Tables 1 and
2 also strongly supports this proposed mechanism. The Km
values demonstrate that R206 and R81 are important in bind-
ChemBioChem 2012, 13, 1645 – 1651
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1649

C. J. Easton et al.
ing lactones 1 and 2, presumably through interactions with
the substrate carboxylate groups. It is likely that analogous
interactions constrain the conformation of esters 5b–7b to
maintain the keto–enol moiety near the ester-group carbon,
thereby facilitating the recyclisation with DLH C123S. With the
double mutants, DLH C123S/R206A and DLH C123S/R81A,
there are less of these interactions, resulting in a less constrain-
ed conformation. As a consequence, cyclisation is less favoured
and the isomerisation rates are lower with the R206A and
R81A mutants, while the ester-group carbon is more exposed
to attack by water, and the hydrolysis rates are higher. This ex-
plains why isomerisation of lactones 1 and 2 is the only pro-
cess observed with DLH C123S, while hydrolysis and isomerisa-
tion are competing processes catalysed by DLH C123S/R206A
and C123S/R81A.
Table 3. Oligonucleotides used for site-directed mutagenesis of DLH and
DLH C123S. The replaced codon is underlined.
Mutation
Oligonucleotide
E36D
R206A
CATGAACGCGTTCACACCAAATATTTGTTGAGCGATCAC
CAACGCCGCGGCACTCGCCACATAGCCCGAACTGCTCGTCGC
GGCGAACGAG
R81A
R206K
R81K
CAGGATGAGGCGCAGGCAGAGCAAGCCTAC
GAGGCCGGACACTCGTTCGCCAAGACGAGCAGTTC
GTCGAAGGCCTGCCAGAGCTTGTAGGCTTGCTCTTTCTGCGCTC
AGAGAGCAAGCCTTCAAGCTCTGGCAG
Y85F
W88A
S203A
S203H
S203D
GCCTACAAGCTCGCGCAGGCCTTCGAC
GAGGCCGGACACGCGTTCGCCAGGACG
GAGGCCGGACACCATTTCGCCAGGACG
GAGGCCGGACACGATTTCGCCAGGACG
characterised using procedures similar to those employed with
[4]
DLH.
Conclusions
The interactions of the proteins with substrates 1 and 2 were ex-
amined in HEPES buffer (20 mm, pH 7.0) containing EDTA (1 mm)
and BSA (25 mgmL ) at 258C. The products were characterised
In summary, our studies of the interactions of lactones 1 and 2
with DLH and a range of its site-specific mutants have resulted
in the identification of several unusual features. While our data
are consistent with the mechanism proposed for catalysis of
the hydrolysis of Z lactone 2 by DLH, they show that hydrolysis
of E isomer, 1, requires a different active-site conformational
change in the enzyme to align the catalytic residues in their
active orientation. A conformational change is also apparent in
the kinetics of the interaction of DLH C123S with lactone 1. In
addition, our kinetic analyses with mutants of DLH and DLH
C123S now allow us to explain why the cysteine-to-serine sub-
stitution in DLH C123S changes the catalysis displayed by the
proteins from hydrolysis of lactones 1 and 2 to interconversion
of these same species.
À1
using HPLC and a diode array detector, thin layer chromatography
1
and H NMR spectroscopy, by comparison with authentic samples.
The kinetics of the reactions were studied by monitoring changes
in absorbance at 280 nm by UV spectroscopy, taking into account
the formation of different products as determined by HPLC. Lac-
tones 1 and 2 have lmax values of 276 and 277 nm (e=17200 and
À1
À1
1
7550m cm respectively), while maleylacetate 3 shows negligi-
[1,22]
ble absorbance in this region.
Acknowledgements
We gratefully acknowledge financial support received from the
Australian Research Council, including the ARC Centre of Excel-
lence for Free Radical Chemistry and Biotechnology.
Experimental Section
Keywords: enzyme catalysis
kinetics · mutagenesis
· hydrolases · isomerases ·
[
20]
The synthesis of lactones 1 and 2 has been reported previously.
Each was obtained in analytically pure form. An authentic sample
of maleylacetate 3 was prepared by treatment of lactone 1 with
aqueous base. DLH and DLH C123S were obtained using reported
[
1] E. Schmidt, G. Remberg, H. J. Knackmuss, Biochem. J. 1980, 192, 331–
37.
3
[4]
methods. Other mutants of DLH were prepared from either DLH
or DLH C123S, through either PCR-based mutagenesis involving
amplification of the reverse complement oligonucleotide strand
[2] L. N. Ornston, W. K. Yeh in Biodegradation and Detoxification of Environ-
mental Pollutants (Ed.: A. M. Chakrabarty), CRC Press, Boca Raton, 1982,
p. 105.
[
3] K. L. Ngai, M. Schlçmann, H. J. Knackmuss, L. N. Ornston, J. Bacteriol.
987, 169, 699–703.
(
5’–3’) incorporating the desired mutated codon (DLH E36D, R81K
1
and R206A), or by using the Quick-Change site-directed mutagene-
sis kit (Stratagene). The oligonucleotides consisting of the 5’–3’
primer with the replaced codon underlined are shown in Table 3,
and these were used with the exact complement in the Quick-
Change protocol. The oligonucleotides and primers were obtained
from Bresatec Pty. Ltd. (Adelaide–Thebarton, Australia) or Gene-
[
[
[
4] D. Pathak, K. L. Ngai, D. Ollis, J. Mol. Biol. 1988, 204, 435–445.
5] D. Pathak, D. Ollis, J. Mol. Biol. 1990, 214, 497–525.
6] D. Pathak, G. Ashley, D. Ollis, Proteins: Struct. Funct. Bioinf. 1991, 9, 267–
279.
[
7] E. Cheah, C. Austin, G. W. Ashley, D. Ollis, Protein Eng. 1993, 6, 575–583.
[8] A. J. Beveridge, D. L. Ollis, Protein Eng. 1995, 8, 135–142.
[9] C. S. Wright, R. A. Alden, J. Kraut, Nature 1969, 221, 235–242.
[10] D. M. Blow, J. J. Birktoft, B. S. Hartley, Nature 1969, 221, 337–340.
[21]
Works (Hindmarsh, Australia). The plasmid pND704
and the
E. coli strain AN1459 for these experiments were provided by Prof.
Nick Dixon of the School of Chemistry, University of Wollongong.
The gene encoding DLH (clcD) was cloned into the pND704 plas-
mid, comprising a heat-inducible overexpression system, and trans-
formed into the E. coli strain AN1459. To confirm the constructs,
the full coding region of each mutant plasmid was sequenced
using the Big Dye Terminator reaction at the Biomolecular Re-
source Facility of the John Curtin School of Medical Research, Aus-
tralian National University. The mutant proteins were purified and
[
[
[
11] I. Walker, C. J. Easton, D. L. Ollis, Chem. Commun. 2000, 671–672.
12] A. V. Hill, J. Physiol. 1910, 40(Suppl.), iv–vii.
13] E. Cheah, G. W. Ashley, J. Gary, D. Ollis, Proteins: Struct. Funct. Bioinf.
1
993, 16, 64–78.
[
[
14] W. E. Karsten, R. E. Viola, Biochim. Biophys. Acta Protein Struct. Mol. Enzy-
mol. 1992, 1121, 234–238.
15] J. Farres, T. T. Y. Wang, S. J. Cunningham, H. Weiner, Biochemistry 1995,
34, 2592–2598.
[16] Z.-Y. Zhang, L. Wu, Biochemistry 1997, 36, 1362–1369.
1
650
www.chembiochem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2012, 13, 1645 – 1651

Dienelactone Hydrolase
[
[
[
[
17] X. Ge, R. E. Campbell, I. van de Rijn, M. E. Tanner, J. Am. Chem. Soc.
998, 120, 6613–6614.
18] B. W. K. Lee, S. G. Van Lanen, D. Iwata-Reuyl, Biochemistry 2007, 46,
2844–12854.
19] M. Ghanevati, J. G. Jaworski, Biochim. Biophys. Acta Mol. Cell Biol. Lipids
001, 1530, 77–85.
20] R. A. Massy-Westropp, M. F. Price, Aust. J. Chem. 1980, 33, 333–341.
[21] C. A. Love, P. E. Lilley, N. E. Dixon, Gene 1996, 176, 49–53.
[22] M. Schlçmann, E. Schmidt, H. J. Knackmuss, J. Bacteriol. 1990, 172,
5112–5118.
1
1
2
Received: April 4, 2012
Published online on July 3, 2012
ChemBioChem 2012, 13, 1645 – 1651
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1651