Recombinant expression of a unique chloromuconolactone dehalogenase ClcF from Rhodococcus opacus 1CP and identification of catalytically relevant residues by mutational analysis

Article abstract of DOI:10.1016/j.abb.2012.07.007

Chloromuconolactone dehalogenase ClcF plays a unique role in 3-chlorocatechol degradation by Rhodococcus opacus 1CP by compensating the inability of its chloromuconate cycloisomerase ClcB2 to dechlorinate the chemically stable cycloisomerization product (4R,5S)-5-chloromuconolactone (5CML). High sequence similarities showed relatedness of ClcF to muconolactone isomerases (MLIs, EC 5.3.3.4) of the 3-oxoadipate pathway. Although both enzyme types share the ability to dechlorinate 5CML, comparison of kcat/Km indicated a significant extent of specialization of ClcF for dechlorination. This assumption was substantiated by an almost complete inability of ClcF to convert (4S)-muconolactone and the exclusive formation of cis-dienelactone from 5CML. Mutational analysis of ClcF by means of variants E27D, E27Q, Y50A, N52A, and A89S indicated relevance of some highly conserved residues for substrate binding and catalysis. Based on the putative isomerization mechanism of MLI, evidence was provided for a role of E27 in initial proton abstraction as well as of Y50 and N52 in substrate binding. In case of N52 substrate binding is likely to occur to the carboxylic group of 5CML as indicated by a significant change of product specificity. Expression in Escherichia coli BL21-CP(DE)-RIL followed by a three-step purification procedure with heat treatment is a convenient strategy to obtain recombinant ClcF and variants thereof.

Full text of DOI:10.1016/j.abb.2012.07.007

Archives of Biochemistry and Biophysics 526 (2012) 69–77
Contents lists available at SciVerse ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/locate/yabbi
Recombinant expression of a unique chloromuconolactone dehalogenase ClcF from
Rhodococcus opacus 1CP and identification of catalytically relevant residues
by mutational analysis
Janosch A.D. Gröning , Christian Roth , Stefan R. Kaschabek , Norbert Sträter , Michael Schlömann a
a
b
a,
⇑
b
a
Environmental Microbiology, TU Bergakademie Freiberg, Leipziger Str. 29, 09599 Freiberg, Germany
Center for Biotechnology and Biomedicine, Institute for Structural Analytics of Biopolymers, University of Leipzig, Deutscher Platz 5, 04103 Leipzig, Germany
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Chloromuconolactone dehalogenase ClcF plays a unique role in 3-chlorocatechol degradation by Rhodo-
coccus opacus 1CP by compensating the inability of its chloromuconate cycloisomerase ClcB2 to dechlo-
rinate the chemically stable cycloisomerization product (4R,5S)-5-chloromuconolactone (5CML). High
sequence similarities showed relatedness of ClcF to muconolactone isomerases (MLIs, EC 5.3.3.4) of the
Received 21 April 2012
and in revised form 16 July 2012
Available online 27 July 2012
3-oxoadipate pathway. Although both enzyme types share the ability to dechlorinate 5CML, comparison
Keywords:
of kcat/Km indicated a significant extent of specialization of ClcF for dechlorination. This assumption was
substantiated by an almost complete inability of ClcF to convert (4S)-muconolactone and the exclusive
formation of cis-dienelactone from 5CML. Mutational analysis of ClcF by means of variants E27D,
E27Q, Y50A, N52A, and A89S indicated relevance of some highly conserved residues for substrate binding
and catalysis. Based on the putative isomerization mechanism of MLI, evidence was provided for a role of
E27 in initial proton abstraction as well as of Y50 and N52 in substrate binding. In case of N52 substrate
binding is likely to occur to the carboxylic group of 5CML as indicated by a significant change of product
specificity. Expression in Escherichia coli BL21-CP(DE)-RIL followed by a three-step purification procedure
with heat treatment is a convenient strategy to obtain recombinant ClcF and variants thereof.
Ó 2012 Elsevier Inc. All rights reserved.
5
-Chloromuconolactone
Dehalogenation
Muconolactone isomerase
3
Modified ortho-cleavage pathway
Actinobacteria
-Chlorocatechol degradation
Introduction
been described in detail for representatives of the phylum Proteo-
bacteria [6,7].
Dehalogenation mechanisms probably play the most critical
role in the detoxification and mineralization of chloroorganic com-
pounds by microorganisms. Organically bound chlorine results in
an increase of persistence, toxicity, and tendency of bioaccumula-
tion and thus is responsible for the high environmental relevance
of this heterogeneous group of substances [1]. Incomplete elimina-
tion of chlorine may result in the formation of intermediates of
even increased toxicity, as it is the case for the accumulation of
mutagenic vinyl chloride during dehalorespiration of tetrachloro-
ethene [2,3] or it may lead to an inhibition of degradation path-
ways not suited for dehalogenation [4,5].
Various dechlorination mechanisms have been identified in the
aerobic degradation of chloroaromatic compounds. These either
occur at an early point of peripheral degradation and/or at a late
stage in the central catabolism of chlorocatechol intermediates.
The modified ortho-cleavage pathway is probably the most com-
mon route for the aerobic degradation of chlorocatechols and has
Investigations into the biochemistry and genetics of chloroaro-
matic degradation in the actinobacterium Rhodococcus opacus 1CP
recently pointed to an alternative type of modified ortho-cleavage
pathway which obviously evolved independently from that found
in Proteobacteria by functionally convergent evolution [8–10]. In
addition to a low phylogenetic relatedness, the chloromuconate
cycloisomerases of strain 1CP and Proteobacteria fundamentally
differ in their ability to dechlorinate. While proteobacterial chlo-
romuconate cycloisomerases lactonize and dechlorinate 2-chloro-
cis,cis-muconate (an intermediate of 3-chlorocatechol catabolism)
into trans-dienelactone ((2Z)-2-(5-oxo-2(5H)-furanylidene)acetic
acid) (Fig. 1A), the corresponding enzyme ClcB2 of strain 1CP only
catalyzes lactonization to (4R,5S)-5-chloromuconolatone1 (5CML)
as a stable intermediate without further dehalogenation (Fig. 1B)
[11]. The inability of ClcB2 to dechlorinate is shared by muconate
cycloisomerases of Proteobacteria and Actinobacteria [8,12], a fact
consistent the latter type of enzyme representing the putative ances-
1
Abbreviations used: 5CML, 5-chloromuconolactone; ELH, 3-oxoadipate enol-
⇑
Corresponding author. Fax: +49 3731 393012.
E-mail address: stefan.kaschabek@ioez.tu-freiberg.de (S.R. Kaschabek).
lactone hydrolase; MLI, muconolactone isomerase, IPTG, isopropyl-b-D-thiogalacto-
pyranoside; SDS, sodium dodecyl sulfate.
0
003-9861/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.abb.2012.07.007

7
0
J.A.D. Gröning et al. / Archives of Biochemistry and Biophysics 526 (2012) 69–77
Fig. 1. (A) Function of proteobacterial chloromuconate cycloisomerases CMCI in 3-chlorocatechol degradation, (B) the ClcF-comprising variant of this pathway in R. opacus
CP, and (C) the physiological role of muconolactone isomerases in catechol degradation. The reactions above represent only an excerpt of the corresponding pathways. CMCI,
1
chloromuconate cycloisomerase; CMLD, 5-chloromuconolactone dehalogenase; MCI, muconate cycloisomerase; MLI, muconolactone isomerase. Abbreviations for the gene
product in strain 1CP are given in parenthesis.
tor. In strain 1CP dechlorination of 5CML is achieved by a hitherto
unique type of 5-chloromuconolactone dehalogenase, encoded by
gene clcF, to yield cis-dienelactone ((2E)-2-(5-oxo-2(5H)-furanylid-
ene)acetic acid) [10].
Sequence alignment of ClcF against proteins of the non-redun-
dant NCBI database indicated a high relatedness of the dehalogen-
ase to the class of muconolactone delta-isomerases. Latter
enzymes are found in the regular ortho-cleavage- or 3-oxoadipate
pathway of catechol degradation and catalyze the conversion of
purpose of biochemical and X-ray crystallographic analysis. Mu-
tants to be constructed were chosen based on the multiple se-
quence alignment with (methyl)muconolactone isomerases, on a
low-resolution X-ray structure for muconolactone isomerase from
P. putida [20] as well as on preliminary structural information for
ClcF [21]. Kinetic studies of the wild-type protein and of these vari-
ants were to be performed in order to support crystallographic
analysis and to interpret the catalytic function of certain amino
acids. In addition, kinetic data were compared to those available
for (methyl)muconolactone isomerases.
(
4S)-muconolactone into 3-oxoadipate enol-lactone (Fig. 1C). Like
ClcF, all (methyl)muconolactone isomerases that have been inves-
tigated in that respect are able to dechlorinate (4R,5S)-5-chlorom-
uconolactone to a mixture of cis- and trans-dienelactone [13,14].
However, this dechlorinating capability has only been shown
in vitro for this class of enzymes and due to the restricted substrate
range of catechol dioxygenases and muconate cycloisomerases
does not play a physiological role. The same is true for the
in vitro demonstration of muconolactone isomerase from Cupriavi-
dus necator JMP134 to dechlorinate (4R,5S)- and (4R,5R)-5-chloro-
Materials and methods
Strains and plasmids and cultivation conditions
R. opacus 1CP was originally isolated from an enrichment cul-
ture with 2,4-dichlorophenol as the sole carbon source [22]. For
DNA isolation, strain 1CP was grown at 30 °C in a mineral medium
(
pH 7.2) [23] with a doubled phosphate buffer concentration sup-
3
-methylmuconolactone, potential intermediates from 5-chloro-3-
À1
À1
plied with 20 g l glycine and 2 g l glucose. P. putida PRS2000
methylcatechol degradation [13,15,16]. Both reactions may be
attributed as promiscuous activity towards these substrates. Thus,
ClcF is likely to be an example of an evolutionary specialization of
muconolactone isomerases for the promiscuous dechlorination
activity. This hypothesis is strengthened by the observations that
ClcF has almost completely lost the ability to convert (4S)-muco-
nolatone to 3-oxoadipate enol-lactone [10] and, as already men-
tioned, that ClcF forms solely cis-dienelactone from 5CML-
dechlorination whereas MLI-catalyzed dechlorinations result in
mixtures of the cis- and trans-isomer of dienelactone.
was grown at 30 °C in mineral medium (pH 7.2) [23] using 5 mM
benzoate as the sole carbon source. Escherichia coli DH5
BRL) and E. coli BL21-CP(DE3)-RIL (Agilent Technologies, Strata-
gene) were grown aerobically with constant shaking at 37 °C in
a (GIBCO
À1
À1
Luria-Bertani (LB) medium [24] containing 100
l
l
g ml ampicillin
g ml chloram-
À1
or a mixture of 100
lg ml ampicillin and 35
phenicol, respectively.
Plasmids used in this study are summarized in Table S1 (Sup-
plementary material).
Although a low-resolution structure of muconolactone isomer-
ase of Pseudomonas putida is available for almost 30 years [17], the
detailed enzymatic mechanism has not yet been elucidated. Based
on reports of Stanier and Ornston [18] and Chari et al. [19], Pieper
and co-workers [15] assumed that proton abstraction from C4 and
proton addition to C2 would bring about the isomerization of muc-
onolactone. They postulated that with 5-chloro-3-methylmucono-
lactone (and consequently also 5-chloromuconolactone) the
proton abstraction from C4 must result in chloride elimination
from C5.
Cloning of the clcF gene
Genomic DNA from R. opacus 1CP was obtained as described
previously [10] and was used as the template in PCR amplification.
The reaction mix for the PCR contained 0.5 lg template DNA, 5% (v/
v) DMSO, 60 pmol of primer clcF_fw (GCTTCGA CATATGTTGTACC-
TAGTTC) and 60 pmol clcF_rev (ATGGATCCTCAGTCTTTGCCGAC)
(Eurofins MWG-Biotech), 40 nmol dNTPs, 150 nmol Mg2+, and
2.5 U Taq polymerase (Thermo Fisher Scientific, Fermentas) in a to-
In order to elucidate catalytically relevant residues, we report
on the preparation and purification of mg amounts of recombinant
ClcF as well as single amino acid exchange mutants of ClcF for the
tal volume of 100 ll. Both oligonucleotides contained recognition
sites for the restriction endonucleases NdeI and BamHI, respec-
tively (underlined sequence). The PCR product was initially ligated

J.A.D. Gröning et al. / Archives of Biochemistry and Biophysics 526 (2012) 69–77
71
into the T-tailed EcoRV-site of pBluescript II KS(+), yielding
pBSclcF1a, and transformed into E. coli DH5 . The gene was then
Healthcare) pre-equilibrated with high-salt buffer (25 mM Tris–
HCl, pH 7.5, containing 1.6 M (NH SO ). ClcF and its variants were
a
4
)
2
4
excised by digestion with NdeI and BamHI (Thermo Fisher Scien-
tific, Fermentas) and ligated into the expression vector pET11a*.
The resulting plasmid, pETRoclcF, was transformed into E. coli
BL21-CP(DE)-RIL and used for expression purposes.
obtained as almost homogeneous preparations (purity P97%
according to SDS–PAGE) by means of a linear decreasing salt gradi-
ent (from 1.6 to 0 M ammonium sulfate). The enzyme eluted at
ꢀ1 M (NH
4 2 4
) SO and was precipitated by addition of solid ammo-
nium sulfate to give a final concentration of 75% saturation. The
suspension was incubated on ice for about 120 min and centri-
fuged (50,000Âg, 15 min, 4 °C). Precipitated protein was stored
either for short term at +4 °C or for longer times at À80 °C.
Site-directed mutagenesis to generate ClcF mutants
Mutants of ClcF were created according to the procedure de-
scribed by Weiner et al. [25]. Briefly, primers carrying the desired
mutation (Table S2, Supplementary material) were synthesized
Determination of ClcF activity and protein concentration
(
biomers.net GmbH, Ulm, Germany) and the mutagenesis PCR with
the wild-type plasmid as template was set up as described.
Remaining parental plasmid DNA was removed by a digestion with
DpnI and 1 ll of the mutagenic PCR was directly transformed into
ClcF activity was determined by measuring the formation of cis-
dienelactone from (4R,5S)-5-chloromuconolactone at 280 nm in a
Varian Cary 50 UV–Vis spectrophotometer using a specific extinc-
tion coefficient of 13.9 mM cm (this study). The same value
was used for mutants E27D and N52A which yielded a mixture
of cis- and trans-dienelactone. The standard assay consisted of
À1
À1
ultracompetent E. coli DH5
a
. Cells were plated onto LB-agar con-
À1
taining 100
lg ml
ampicillin. Several clones were picked and
DNA was isolated (MinElute Plasmid Kit, Quiagen) and sequenced
Eurofins MWG-Biotech) in order to verify the incorporated muta-
(
54 mM Na/K-phosphate buffer (pH 7.2), 100 lM (4R,5S)-5-chlo-
tion and the integrity of the gene sequence.
romuconolactone [10], and was started by addition of the ClcF-
containing sample.
Expression of ClcF and its variants in E. coli BL21-CP-RIL
Kinetic parameters Km and vmax were determined by means of
the initial rate method assuming Michaelis–Menten kinetics. Reac-
tion rates were fitted by non-linear least-square regression using
the DynaFit 3.28 software [26]. The turnover number kcat was cal-
culated with the molecular weight of 11,193.7 g mol for one ClcF
subunit. Initial reaction rates were determined in triplicates using
the standard assay and substrate concentrations between 40 and
Due to the occurrence of codons AGG, AUA, and CUA which are
rarely used by wild-type E. coli, strain BL21-CP(DE3)-RIL harbour-
ing corresponding tRNA genes was chosen as an expression host
for ClcF.
À1
Freshly grown transformants of E. coli BL21-CP(DE3)-RIL har-
bouring the corresponding expression plasmid (e.g. pETRoclcF)
500 lM (4R,5S)-5-chloromuconolactone.
Protein concentrations were determined with the Bradford Pro-
tein Assay (BioRad) using bovine serum albumin as standard.
À1
were used to inoculate 10-ml LB-medium with 100
lg ml ampi-
À1
cillin and 50
l
g ml chloramphenicol. After overnight cultivation
at 30 °C under constant shaking (130 rpm), 5 ml of that pre-culture
were used to inoculate 500 ml of the same medium to be incubated
under similar growth conditions. At an optical density OD(600 nm)
Product analysis of 5-chloromuconolactone turnover
of 0.6–0.7, expression was induced by addition of 100
pyl-b- -thiogalactopyranoside (IPTG). After incubation for further
at 30 °C, the culture was harvested by centrifugation
l
M isopro-
ClcF variants were analyzed by reversed phase HPLC for their
ability to dehalogenate 5-chloromuconolactone into cis- and
trans-dienelactone. Twenty microliter samples from standard
reaction assays (see below) were checked for complete transfor-
mation by UV/VIS-spectrophotometry and applied onto an Euro-
spher-100 C18 column (4 mm internal diameter, 125 mm
length, Knauer, Berlin, Germany) as the stationary phase. An
D
4
h
(
5000Âg, 15 min, 4 °C), washed with 54 mM Na/K-phosphate buf-
fer (pH 7.2), and biomass was stored at À20 °C for further use.
Preparation of cell extracts
aqueous solution of 30% (v/v) methanol and 0.2% (w/v) H
3
PO
4
À1
À1
To freshly thawed biomass of expression clones, 0.5 U ml
served as the mobile phase at a flow rate of 0.7 ml min . Absorp-
tion of the eluate was recorded at multiple wavelengths. Signals
were assigned by comparison to external standards for cis-diene-
lactone (net retention volume 2.55 ml), trans-dienelactone
(0.98 ml), (4R,5S)-5-chloromuconolactone (0.77 ml), and 2-chlo-
romuconolactone (1.40 ml).
Dnase I was added and disruption was achieved by two passages
through a French pressure cell (1500 psi). The mixture was centri-
fuged (100,000Âg, 45 min, 4 °C) and the clear supernatant served
as cell-free crude extract for further protein purification. In cases,
extracts were not immediately subjected to purification, the crude
extract was stored at À20 °C.
Determination of muconolactone-isomerizing activity
Purification and storage of recombinant wild-type ClcF and its variants
The ability of ClcF and its mutants to convert (4S)-muconolac-
tone into 3-oxoadipate enol-lactone was assayed by a HPLC-based
modification of the method of Ornston (1966) [27]. A typical reac-
For an initial heat treatment cell-free crude extract of the corre-
sponding expression clone was incubated at 65 °C for 10 min. After
rapid cooling to 4 °C on ice, precipitated protein was separated by
centrifugation and removed (10,000Âg, 15 min, 4 °C) resulting in a
tion mixture contained in 1 ml: 25
lmol Tris–HCl (pH 7.5),
0.2 mol (4S)-muconolactone, and approximately 0.01 U of par-
l
3
-fold increased specific activity of ClcF. The supernatant was then
tially purified 3-oxoadipate-enol-lactone hydrolase (ELH) from P.
applied onto a Q-Sepharose column (1.6 cm  4.0 cm, GE Health-
care) pre-equilibrated with 25 mM Tris–HCl, pH 7.5. The same buf-
fer was used for the elution of unbound proteins which were
shown to comprise also ClcF in concordance with its relatively high
pI of 6.91. The wash eluate was then supplemented at 4 °C with so-
lid ammonium sulfate to a final concentration of 1.6 M. After cen-
trifugation (10,000Âg, 15 min, 4 °C) the clear supernatant was
loaded onto a Phenyl Sepharose column (2.6 cm  2.8 cm, GE
putida PRS2000 as an auxiliary enzyme. In order to determine
the blank reaction rate, four 10-ll-samples were taken in the
course of 90 min, quenched by addition of 5% (v/v) phosphoric
À1
acid and subjected to RP-HPLC (15% (v/v) MeOH, 2 g l
Eurospher-100 C18). Afterwards, 10
buffer (pH 7.2), containing 5–10 mU (activity determined with
200 M 5CML as the substrate) wild-type- or mutated ClcF were
added to the assay mixture and RP-HPLC analysis was continued
3 4
H PO ,
l
l of 54 mM Na/K-phosphate
l

7
2
J.A.D. Gröning et al. / Archives of Biochemistry and Biophysics 526 (2012) 69–77
for 90 min. The decrease of substrate (net retention volume
.3 ml) was detected at 210 nm. A very slight decrease of (4S)-
cis-Dienelactone and trans-dienelactone were available by chemi-
cal synthesis [34].
1
muconolactone, even in the absence of ClcF, indicated traces of
muconolactone isomerase in the auxiliary enzyme preparation,
however, this contamination was considered as a blank reaction
rate.
Results and discussion
Construction of a ClcF expression clone and preparation of
recombinant 5-chloromuconolactone dehalogenase
The ELH was partially purified from a crude extract of benzoate-
grown P. putida PRS2000 as described previously [10] using anion-
exchange chromatography (Q-Sepharose, GE Healthcare) and
hydrophobic interaction chromatography (Phenyl Sepharose, GE
Healthcare). The final preparation of enriched 3-oxoadipate enol-
lactone hydrolase was stored at À20 °C in 25 mM Tris–HCl (pH
In order to obtain sufficient quantities of ClcF to investigate its
kinetic parameters and for future structure determination, the cor-
responding ORF was cloned into the cloning vector pBluescript II
KS(+). Amplification by two clcF-specific primers clcF_fw and
clcF_rev yielded an amplicon of expected size (303 bp) whose iden-
tity was confirmed by DNA sequence analysis. The gene clcF was
subsequently ligated into the expression vector pET11a* and desig-
nated as pETRoclcF. Expression of the wild-type clcF gene in E. coli
7
.5) in the presence of 50% (v/v) glycerol.
Purity control by SDS–PAGE
In order to determine the purity of wild-type ClcF and ClcF vari-
ants, enzyme preparations were subjected to discontinuous so-
dium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis.
Slab gels with 15% acrylamide were prepared by using a Tris-/Tri-
cine buffer system [28] and stained with Coomassie Brilliant Blue
R-250.
BL21-CP(DE3)-RIL at 30 °C in the presence of 100 lM IPTG led to
almost exclusive formation of the protein in soluble active form.
Thus, the chosen expression strain with t-RNAs for codons not usu-
ally occurring in E. coli was obviously suitable for expression of R.
opacus clcF.
Partial purification of muconolactone isomerase from P. putida
PRS2000
Relatedness of ClcF to (methyl)muconolactone isomerases and
identification of conserved residues for site-directed mutagenesis
Muconolactone isomerase (MLI) from P. putida was partially
purified in order to measure and compare relative activities of
this type of enzyme towards (4S)-muconolactone and 5CML.
MLI was obtained during the purification process of enol-lactone
hydrolase described before. During anion-exchange chromatogra-
phy on Q-Sepharose, MLI eluted together with ELH at approxi-
mately 230 mM NaCl. Active fractions were pooled and after
ammonium sulfate precipitation (75% saturation) proteins were
subjected to hydrophobic interaction chromatography on Phenyl
In order to identify further representatives of ClcF-like proteins,
a BLASTP search was made against available protein sequences of
the non-redundant database of the NCBI. ClcF was found to match
the superfamily of muconolactone delta-isomerase (EC 5.3.3.4)
[27] since a large number of (putative) representatives of this en-
zyme family showed high similarities (46–61% identity) over a
query coverage of 95–100% (Fig. 2A). Analyses of the relatedness
between ClcF and some representatives of these matches showed
that the dehalogenase branches deep within the dendrogram of
regular (methyl)muconolactone isomerases (Fig. 2B). That all
hypothetical muconolactone isomerases used in the dendrogram
indeed belong to the regular EC-5.3.3.4-type of isomerase was as-
sured by considering only those representatives genes of which
were part of a complete catechol-catabolic operon. None of the ob-
tained protein hits with similarity to ClcF was found to be encoded
in close neighborhood to genes of chlorocatechol degradation. This
finding and the fact that the similarity of the best hit is restricted to
61% identical positions makes a physiological function in dehalo-
genation at least unlikely. Thus, ClcF of R. opacus 1CP is suggested
to be the only available representative of a 5CML dehalogenase, so
far.
Homology of ClcF and muconolactone isomerase is substanti-
ated by comparison of the tertiary and quaternary protein struc-
tures of both enzymes. MLI of P. putida has been structurally
resolved at 7 Å [17] and later on at 3.3 Å [20] and was found to ex-
hibit a decameric structure with 5-fold and 2-fold symmetries.
Each subunit is composed of a body with two alpha-helices, an
anti-parallel twisted beta-sheet of four strands, and an extended
arm. The helices and the sheet form a two-layered structure with
an enclosed hydrophobic core and a partially formed putative ac-
tive-site pocket. The latter one is completed by means of the C-ter-
minal arm of another subunit.
Sepharose with a decreasing gradient of 1.2–0 M (NH
4
)
2
SO
5 mM Tris–HCl (pH 7.5). MLI eluted at approximately 670–
80 mM (NH SO followed by ELH (420–430 mM). Relative sub-
4
in
2
6
4
)
2
4
strate specificities of partially purified MLI were tested with the
standard assay for ClcF and the assay of the muconolactone-isom-
erizing activity, respectively, using 5CML and (4S)-muconolactone
as substrates.
Sequence alignment and computational modelling
Conserved amino acids in (methyl)muconolactone isomerases
and ClcF were determined by multiple sequence alignment using
ClustalW2 [29] (at http://www.ebi.ac.uk/Tools/msa/clustalw2/; de-
fault values). In order to visualize phylogenetic relatedness, a den-
drogram was calculated by means of programs Protdist and Fitch of
the PHYLIP package (version 3.69). Bootstrap values of which only
those >50% are shown were derived from Seqboot (PHYLIP) calcula-
tions with 100 replicates.
The tertiary structure modelling of ClcF was conducted applying
the muconolactone isomerase from P. putida (PDB_ID:1MLI) as
template [20]. The protein fold recognition server PHYRE2 [30]
was used in an automated procedure and images were generated
with Pymol [31]. Structural manipulations were made using the
crystallographic object-oriented tool kit Coot [32].
A basically similar structure was found for ClcF [21] and calcu-
lation of the alpha-C-backbone of the dehalogenase allowed to
identify those amino acids which are closely located to the putative
active site and thus eventually are involved in substrate binding
and catalysis (Fig. 3). Residues Glu27, Tyr50, Asn52, and His87
are of particular interest, since they are highly conserved in MLIs
and are shared by ClcF. Another interesting residue is Ser89. The
observation that this residue is highly conserved in MLIs but
Chemicals
(
4R,5S)-5-Chloromuconolactone, 2-chloromuconolactone, and
(
4S)-muconolactone were enzymatically prepared from 2-chloro-
cis,cis-muconate and cis,cis-muconate [33], respectively, using
purified muconate cycloisomerase from P. putida PRS2000 [12].

J.A.D. Gröning et al. / Archives of Biochemistry and Biophysics 526 (2012) 69–77
73
Fig. 2. (A) Identification of conserved amino acids in (putative) (methyl)muconolactone isomerases and the novel 5-chloromuconolactone dehalogenase ClcF and (B)
dendrogram showing the relatedness between selected representatives of these proteins. Highly conserved amino acids glutamic acid (pos. 27), tyrosine (pos. 50), asparagine
(
pos. 52), histidine (pos. 87), and serine (pos. 89), which were subjected to modification in the present study are marked. Protein accession numbers are as follows:
Acinetobacter lwoffii K24, CatC, AAC46227; A. lwoffii K24, CatC2, AAC31768; Acinetobacter sp. ADP1, AAC46431; Arthrobacter sp. BA-5-17, BAD11153; Burkholderia gladioli
BSR3, AEA64672; Burkholderia sp. NK8, BAB21459; C. necator 335T, AAG42025; C. necator JMP134, CatC, AAZ61055; C. necator JMP134, MmlJ, CAA67959; C. necator N-1,
AEI81695; Chromohalobacter sp. HS2, ABV82779; Methylobacterium nodulans ORS 2060, ACL57349; Mycobacterium smegmatis MC2 155, ABK73752; Novosphingobium
aromaticivorans DSM 12444, ABP64688; Polaromonas naphthalenivorans CJ2, ABM37427; Pseudomonas fluorescens WH6, EEY97020; P. putida PRS2000, AAA25767;
Pseudomonas reinekei MT1, ABI93946; Pseudomonas sp. CA10, BAB32457; Pseudonocardia dioxanivorans CB1190, MLI Psed_3095, AEA25292; P. dioxanivorans CB1190, MLI
Psed_6063, AEA28172; Psychrobacter cryohalolentis K5, ABE75050; Ralstonia solanacearum CFBP2957, CBJ43151; R. solanacearum PSI07, CBJ51252; Rhodococcus erythropolis
CCM2595, CAX36484; Rhodococcus jostii RHA1, ABG94176; Rhodococcus opacus 1CP: ClcF, CAD28145; CatC, CAA67935; Rhodococcus sp. AN-22, BAE46389; Saccharomonospora
azurea SZMC 14600, EHK84170.
exchanged by alanine in ClcF might be of relevance for the almost
complete inability of ClcF to convert muconolactone or for the
high product specificity of ClcF to yield pure cis-dienelactone
from 5CML-turnover. Glu27, His87, and Ser89 have already been
proposed by Katti and coworkers [20] to be of catalytic relevance
in muconolactone isomerase.
A simple acid-base-catalyzed 1,3-proton shift was identified
as the mechanism for muconolactone isomerization, in which

7
4
J.A.D. Gröning et al. / Archives of Biochemistry and Biophysics 526 (2012) 69–77
Purification of recombinant ClcF and its variants
ClcF and variants thereof were expressed recombinantly in
E. coli and purified by a procedure similar to the one used for the
enrichment of the wild-type enzyme from R. opacus 1CP [11].
The combination of a heat treatment (65 °C), an anion-exchange-
and a hydrophobic interaction chromatography resulted in a prep-
aration that contained 37% of the initial ClcF activity. SDS–PAGE of
the preparation yielded a single band of approx. 11 kDa (theoreti-
cal MW 11,193.72 Da) and indicated homogeneity sufficient for
crystallisation approaches (P97%, data not shown).
The purification of all ClcF variants was achieved by basically
the same procedure. While mutants E27Q, E27D, Y50A, N52A,
and A89S were obtained as soluble proteins, expression of variant
H87A yielded solely inclusion bodies.
Kinetics of 5CML-conversion by (methyl)muconolactone isomerases
from C. necator JMP 134 and by ClcF and its variants
Comparison of kinetic parameters of ClcF for 5CML-conversion
to available data of (methyl)muconolactone isomerases should
indicate an evolutionary specialization of the unique 5-chloromuc-
onolactone dehalogenase for the process of dechlorination. In fact,
ClcF was shown to exhibit the highest specificity
Fig. 3. Homology model of ClcF showing the location of mutated residues around
the putative active-site pocket. One subunit was expanded to asymmetry using a
fragment of a second neighboured monomer to model the position of the C-
terminal residues H87 and A89 in the active site.
6
À1 À1
(
1.4 Â 10 M
s ) towards (4R,5S)-5-chloromuconolactone when
deprotonation at C-4 is followed by a shift of the double bond and
a final protonation of C-2 [15,19] (Fig. 4A). Despite the fact that a
similar mechanism has been reported for ketosteroid isomerase
5
À1 À1
compared to MLI (5.8 Â 10 M
s
)
as well as of MMLI
4
À1 À1
(
6.1 Â 10 M
s ) from C. necator JMP134 (Table 1). The extre-
mely high maximum turnover number of 5CML by MLI
(
[
35], sequence alignment and structural comparison of this
À1
À1
5950 ± 770 s ) which exceeds that of ClcF (529 ± 26 s ) by a fac-
type of enzyme with ClcF and MLIs do not indicate relatedness
tor of 11.2 is compensated by its lower substrate affinity as it is ex-
(
(
data not shown). Early attempts to construe the peptide
Tyr-Gly-Asn-Ile-Ser) of MLI of P. putida to be homologous to an
pressed by
a
27.4-fold higher Michaelis constant of MLI
lM).
(
10,300 ± 1800
active-site tetrapeptide (Tyr-Ala-Asn-Ser) of ketosteroid isomer-
ase [20,36] could not be confirmed in the light of advanced
structural studies [37].
In order to gain evidence for an involvement of the above men-
tioned amino acid residues in substrate binding and catalytic
mechanism, ClcF mutants E27Q, E27D, Y50A, N52A, H87A, and
A89S were created. Except mutant H87A all variants could be ex-
pressed as soluble proteins by means of a pET11a*/E. coli BL21-
CP(DE3)-RIL expression system.
The kinetic parameters of all ClcF variants for 5CML-conversion
were found to be significantly changed. Compared to the five mu-
tants E27D, E27Q, Y50A, N52A, and A89S, wild-type ClcF shows the
À1
highest kcat value of 529 ± 26 s . However, the mutant A89S
turned out to be slightly more specific for (4R,5S)-5-chloromucono-
lactone conversion compared to the wild-type enzyme. The slight
increase of the specificity constant of the A89S variant by a factor
of 1.2 results mainly from a significantly reduced Km value,
which is compensated in part by the reduced turnover number.
Fig. 4. (A) Proposed mechanism of muconolactone isomerase catalysis [20,38] and (B) hypothetical adaptation of the initial proton abstraction in the dehalogenation of 5-
chloromuconolactone. The 1,3-proton transfer in muconolactone involves an initial proton abstraction at C-4 by a basic amino acid residue, followed by protonation of the
carbanion intermediate at C-2 through the action of an acidic amino acid residue [38]. During 5CML-conversion the negative charge of carbanion transition state is removed
by chloride elimination. Pointed lines indicate negatively charged resonance structures.

J.A.D. Gröning et al. / Archives of Biochemistry and Biophysics 526 (2012) 69–77
75
Table 1
Comparison of kinetic parameters for the conversion of (4R,5S)-5-chloromuconolactone by wild-type ClcF, some variants, and (methyl)muconolactone isomerases from C. necator
JMP134.
À1
kcat [s ]b
À1
À1 À1
Enzyme
Km [l
M]
vmax [U mg
]
kcat/Km [M
s ]
6
ClcF wild-type
ClcF (E27D)
ClcF (E27Q)
ClcF (Y50A)
ClcF (N52A)
ClcF (H87A)
ClcF (A89S)
376 ± 32
303 ± 10
307 ± 12
1556 ± 127
1250 ± 112
N. d. (inclusion bodies)
215 ± 25
10,300 ± 1800
5200 ± 800
2838 ± 141
11.5 ± 0.2
0.45 ± 0.01
137 ± 9
529 ± 26
1.4 Â 10
3
2.10 ± 0.04
0.080 ± 0.002
25.6 ± 1.8
5.0 ± 0.4
6.9 Â 10
261
4
1.6 Â 10
3
26.8 ± 2.0
4.0 Â 10
6
1934 ± 109
35,300 ± 4600
1850 ± 200
361 ± 20
5950 ± 770
317 ± 35
1.7 Â 10
a
5
MLI JMP134
5.8 Â 10
MMLI JMP134a
6.1 Â 10
4
N. d. not determined.
a
[
14,15].
b
kcat-values were calculated on the basis of a subunit size of 11.2 kDa (ClcF and variants), 10.1 kDa (muconolactone isomerase, MLI), and 10.3 kDa (methylmuconolactone
isomerase, MMLI).
Considering that serine is highly conserved among all available
reported for muconolactone isomerase and methylmuconolactone
isomerase from C. necator JMP134 [13,15].
(
methyl)muconolactone isomerases the fact that alanine instead
of serine at this position does positively affect the performance
of ClcF to convert 5CML raises the question about evolutionary
reasons for this exchange. Possible explanations may be related
to the inability of ClcF to convert (4S)-muconolactone or to the
product specificity of dechlorination (cis- and trans-isomer of
dienelactone).
A drastic effect on activity was observed for the substitution of
the highly conserved E27 against aspartic acid or glutamine lead-
ing to a decrease of the turnover number by a factor of around
In order to identify single residues which are responsible for the
increase of substrate specificity of ClcF, mutants were character-
ized for their ability to convert (4S)-muconolactone. Unlike the
dehydrochlorination of 5-chloromuconolactone to cis-dienelac-
tone, which is an irreversible reaction, the isomerization of (4S)-
muconolactone to 3-oxoadipate enol-lactone has a pronounced
equilibrium character making it necessary to determine kinetics
in the presence of an excess of 3-oxoadipate enol-lactone hydro-
lase [27]. Specific activities of wild-type ClcF and variants towards
2
50 (E27D) or 6600 (E27Q), respectively. Obviously the acidic res-
(4S)-muconolactone were determined in the presence of 200
this substrate. Compared to the wild-type enzyme (1.9 U mg
l
M of
À1
idue at position 27 plays a crucial role in the turnover of (4R,5S)-5-
chloromuconolactone. The substrate affinities were not signifi-
cantly affected by both mutations, a fact which suggests that E27
may not primarily be involved in substrate binding. With a pKa va-
lue of 4.2 the carboxyl group should be almost completely ionized
under physiological conditions (pH 7.0) and thus act primarily as a
base. For the mechanism of muconolactone isomerization a simple
acid-base catalysis via a proton transfer from C-4 to C-2 has been
suggested (Fig. 4A) [19,20] and it is very likely that an initial step
of proton abstraction is also involved in the dechlorination of
)
only mutant A89S showed an increased specific activity of
À1
3.0 U mg for the non-chlorinated parent compound whereas all
other mutants were shown to fall below these values for two to
six orders of magnitude (Table 2). A quotient of specific activity to-
wards ML and 5CML was calculated and normalized for wild-type
ClcF in order to better visualize changes in substrate preference.
Mutants E27D and N52A showed a relative increase of activity
for ML, however, due to the strong decrease of overall activity is
practically irrelevant when been compared to the corresponding
activity of the wild-type. Only for mutant A89S, of which the spec-
5
CML (Fig. 4B). Despite of the presence of a b-carboxyl function
6
À1 À1
with similar acidity (pKa = 3.9) in the aspartic acid residue of mu-
tant E27D, the catalytic activity is drastically reduced when com-
pared to the enzyme with glutamic acid. This presumably is a
result of an insufficient length of the aspartate side chain impeding
contact to the substrate (4R,5S)-5-chloromuconolactone.
The exchanges of Y50 and N52 for alanine were accompanied
with decreases of activity by factors 21 and 106, respectively. In
addition to the strong effects on enzyme activity, the effects of
both exchanges on Km were significant with increases of around
three (N52A) to four (Y50A). Thus, N52 and Y50 may be involved
in substrate fixation via hydrogen bonding interactions.
ificity constant of 1.7 Â 10 M
s
exceeds that of wild-type ClcF
by a factor of 1.2 (Table 1), an 1.6-fold increase of muconolactone-
isomerizing activity is accompanied with an absolute enhance-
ment of specific activity under the conditions of measurement.
Interestingly, this serine residue was shown to be highly conserved
among all muconolactone isomerases and only in ClcF is ex-
changed by alanine.
Effect of amino acid exchanges on product formation from 5CML
As mentioned before, isomerization of (4S)-muconolactone
comprises a deprotonation at C-4 and, in order to compensate
the resulting negative charge of the carbanion, a subsequent pro-
tonation of C-2 (Fig. 4A). Only the first of both steps should be nec-
essary for the conversion of 5CML, since here it is the release of
chloride which removes the negative charge of the transition state
(Fig. 4B). For 5CML-conversion the second step is supposed to be a
spontaneous process which in principal does not need the presence
of or an interaction with a certain amino acid residue, a hypothesis
which is strengthened by the observation that slightly alkaline
conditions (pH 9) are sufficient to cause an immediate dehydro-
halogenation of 5CML in the absence of an enzyme [12].
Influence of amino acid exchanges on (4S)-muconolactone conversion
Initial attempts to characterize the substrate specificity of ClcF
pointed to a complete inability to transform (4S)-muconolactone
[
10]. The availability of larger amounts of the recombinant enzyme
now allowed the detection of
a very low specific activity
À1
(
1.9 U mg ) which corresponds to 0.2% of the activity determined
À1
with (4R,5S)-5-chloromuconolactone (988 U mg ) at identical
substrate concentration (200 M) (Table 2). In contrast, mucono-
l
lactone isomerase of P. putida PRS2000 converts (4S)-muconolac-
tone compared to 5CML with a relative activity of 170% under
similar conditions. A similar capability to transform both (4R,5S)-
However, comparison of chemical and enzymatic 5CML-conver-
sion indicates a distinctive influence of protein structure on prod-
uct configuration. While alkaline treatment of 5CML yields
5
-chloromuconolactone and (4S)-muconolactone has also been

7
6
J.A.D. Gröning et al. / Archives of Biochemistry and Biophysics 526 (2012) 69–77
Table 2
Conversion of (4S)-muconolactone by wild-type ClcF, some variants, and muconolactone isomerase from P. putida PRS2000, and comparison of product specificity from (4R,5S)-5-
chloromuconolactone turnover.
a
a
Enzyme
Sp. activity for ML
Sp. activity for 5CML
Normalized ratio (sp. act. for 5CML)/(sp.
act. for ML)b
Product(s) from 5CML-conversion
cis-DL [%]/trans-DL [%]
À1
À1
[U mg
]
[U mg
]
ClcF wild-type
ClcF (E27D)
ClcF (E27Q)
ClcF (Y50A)
ClcF (N52A)
1.9
975
1
100/0
100/0
92/8
100/0
75/25
100/0
65/35
0.023
3.7 Â 10
6.9 Â 10
0.025
3.0
2.81
0.177
15.7
3.76
926
4.1
0.01
0.22
3.3
1.6
À6
À3
ClcF (A89S)
MLI P. putida PRS2000
c
N. d.
N. d.
172
N. d. not determined.
a
Substrate concentration = 200 lM.
The ratio (sp. act. for 5CML)/(sp. act. for ML) of a mutant was calculated and referred to that of ClcF wild-type (normalized as 1).
Determined from partially purified enzyme on the basis of volume activities.
b
c
predominantly the trans isomer of dienelactone ((2Z)-2-(5-oxo-
purification method for ClcF and single-position mutants. All
variants investigated in this study showed significant to drastic
effects either on substrate binding (Y50A, N52A), on the reaction
rate of catalysis (E27D, E27Q) or on both processes (A89S). In
addition, N52 was found to strongly affect product specificity of
dehydrochlorination probably by the hindrance of rotatability of
the exocyclic C–C single bond of the carbanion transition state
prior to elimination. This fixation could be achieved by hydrogen
bonds between arginine and the carboxyl group of 5CML. None
of the investigated variants was able to fully regain the evolution-
arily lost ability of ClcF to convert (4S)-muconolactone and the
relative activity towards this compound could only be increased
by a few percent for variants E27D and N52A. In combination with
crystallographic data of ClcF which are currently in progress the
obtained data will help to interpret the exact role of residues in
the catalytic process.
2
(5H)-furanylidene)acetic acid) (ꢀ87% trans- and ꢀ13% cis-diene-
lactone [this study]; >95% trans-dienelactone [13]) via anti-elimi-
nation, enzymatic conversions favour formation of the cis-isomer
(
=(2E)-2-(5-oxo-2(5H)-furanylidene)acetic acid) via syn-elimina-
tion. MLI of P. putida PRS2000 was shown in this study to convert
CML into 65% cis- and 35% trans-dienelactone (Table 2), a prefer-
5
ence similar to the ones reported for MLI and MMLI from C. necator
JMP134 yielding mixtures of cis- and trans-dienelactone of 3:1 and
2
.5:1, respectively [14,15]. A remarkable difference in product
specificity was found for ClcF of R. opacus 1CP, which yields almost
pure cis-dienelactone. Since the product configuration does reflect
the localization of the carboxylic group in the moment of chloride
elimination, product specificity is likely to be influenced by the ex-
tent of rotatability of the exocyclic C4–C5 single bond.
During the conversion of muconolactone by MLI, a free rotat-
ability of this single bond does not have an effect on product for-
mation, since this structural moiety is not affected during the
course of reaction. Hence, a fixation of the carboxylic acid residue
should be necessary for the purpose of substrate binding rather
than for the reaction mechanism itself.
Acknowledgment
We would like to thank Robert Zeidler for his support in prepar-
ing mutants.
The remarkably high product specificity of ClcF makes sense
from a physiological point of view, since the activity of the follow-
ing 3-chlorocatechol-catabolic enzyme dienelactone hydrolase
Appendix A. Supplementary data
(
ClcD2) is restricted to cis-dienelactone [10]. Recruitment of a reg-
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.abb.2012.07.007.
ular MLI for the purpose of dechlorination might lead to the forma-
tion of trans-dienelactone as a dead-end metabolite in strain 1CP.
Thus, a high product specificity of ClcF could be easily interpreted
as an evolutionary adaptation of MLI in order to meet the narrow
substrate requirements of the following catabolic enzyme.
Available variants were tested for an effect of the mutated sin-
gle amino acid residue on product specificity.
While mutants E27D, Y50A, and A89S behaved like wild-type
ClcF and did not produce measurable amounts of trans-dienelac-
tone, variants E27Q and N52A led to the formation of 8% and
References
[1] W. Reineke, C. Mandt, S.R. Kaschabek, D.H. Pieper, Chlorinated Hydrocarbon
Metabolism, eLS, John Wiley
& Sons Ltd., Chichester, 2011. <http://
www.els.net> http://dx.doi.org/10.1002/9780470015902.a0000472.pub3.
2] T.D. DiStefano, J.M. Gossett, S.H. Zinder, Appl. Environ. Microbiol. 57 (1991)
2287–2292.
3] P.J.M. Middeldorp, M.L.G.C. Luiyten, B.A. van de Pas, M.H.A. van Eekert, S.W.M.
Kengen, G. Schraa, A.J.M. Stams, Biorem. J. 3 (1999) 151–169.
4] I. Bartels, H.J. Knackmuss, W. Reineke, Appl. Environ. Microbiol. 47 (1984) 500–
505.
[
[
[
[
[
2
5%, respectively, of the trans isomer. The behavior of the latter
variant can be interpreted, if it is assumed that asparagine is in-
volved in substrate binding of the carboxyl group by hydrogen
bonds. However, since N52 is conserved in all MLIs other structural
features than this residue are responsible for the high specificity of
ClcF.
5] C.A. Pettigrew, B.E. Haigler, J.C. Spain, Appl. Environ. Microbiol. 57 (1991) 157–
162.
6] W. Reineke, H.J. Knackmuss, Annu. Rev. Microbiol. 42 (1988) 263–287.
[7] M. Schlömann, Biodegradation 5 (1994) 301–321.
[
[
8] I.P. Solyanikova, O.V. Maltseva, M.D. Vollmer, L.A. Golovleva, M. Schlömann, J.
Bacteriol. 177 (1995) 2821–2826.
9] D. Eulberg, E.M. Kourbatova, L.A. Golovleva, M. Schlömann, J. Bacteriol. 180
(1998) 1082–1094.
Conclusions
[
[
[
[
10] O.V. Moiseeva, I.P. Solyanikova, S.R. Kaschabek, J. Gröning, M. Thiel, L.A.
Golovleva, M. Schlömann, J. Bacteriol. 184 (2002) 5282–5292.
11] O.V. Moiseeva, O.V. Belova, I.P. Solyanikova, M. Schlömann, L.A. Golovleva,
Biokhimiya 66 (2001) 678–687.
The ClcF protein of R. opacus 1CP could successfully be ex-
pressed and purified from E. coli BL21-CP-RIL(pETRoclcF) in mg-
amounts and to homogeneity, sufficient for further structural stud-
ies. A three-step procedure with heat treatment, Q-Sepharose- and
Phenyl Sepharose chromatography turned out to be a reliable
12] M.D. Vollmer, P. Fischer, H.J. Knackmuss, M. Schlömann, J. Bacteriol. 176
(
1994) 4366–4375.
13] M. Prucha, V. Wray, D.H. Pieper, Eur. J. Biochem. 237 (1996) 357–366.
[14] M. Prucha, A. Peterseim, D.H. Pieper, Arch. Microbiol. 168 (1997) 33–38.

J.A.D. Gröning et al. / Archives of Biochemistry and Biophysics 526 (2012) 69–77
77
[
15] M. Prucha, A. Peterseim, K.N. Timmis, D.H. Pieper, Eur. J. Biochem. 237 (1996)
50–356.
[26] P. Kuzmic, Anal. Biochem. 237 (1996) 260–273.
[27] L.N. Ornston, J. Biol. Chem. 241 (1966) 3795–3799.
3
[
[
[
[
16] K. Pollmann, V. Wray, D.H. Pieper, J. Bacteriol. 187 (2005) 2332–2340.
17] B.A. Katz, D. Ollis, H.W. Wyckoff, J. Mol. Biol. 184 (1985) 311–317.
18] R.Y. Stanier, L.N. Ornston, Adv. Microb. Physiol. 9 (1973) 89–151.
19] R.V.J. Chari, C.P. Whitman, J.W. Kozarich, K.-L. Ngai, L.N. Ornston, J. Am. Chem.
Soc. 109 (1987) 5520–5521.
20] S.K. Katti, B.A. Katz, H.W. Wyckoff, J. Mol. Biol. 205 (1989) 557–571.
21] C. Roth, S.R. Kaschabek, J.A.D. Gröning, T. Handrek, M. Schlömann, N. Sträter,
Sect. F: Struct. Biol. Cryst. Commun. F68 (2012) 591–595.
[28] H. Schägger, G. von Jagow, Anal. Biochem. 166 (1987) 368–379.
[29] M. Goujon, H. McWilliam, W. Li, F. Valentin, S. Squizzato, J. Paern, R. Lopez,
Nucleic Acids Res. 38 (2010) W695–W699.
[30] L.A. Kelley, M.J.E. Sternberg, Nat. Protocols 4 (2009) 363–371.
[31] W.L. DeLano, The PyMOL Molecular Graphics System, DeLano Scientific, Palo
Alto, CA, USA, 2002.
[32] P. Emsley, B. Lohkamp, W.G. Scott, K. Cowtand, Acta Crystallogr. D. Biol.
Crystallogr. 66 (2010) 486–501.
[
[
[
[
[
[
22] S.N. Gorlatov, O.V. Maltseva, V.I. Shevchenko, L.A. Golovleva, Mikrobiologiya
[33] S.R. Kaschabek, W. Reineke, J. Org. Chem. 59 (1994) 4001–4003.
[34] R.A. Massy-Westropp, M.F. Price, Aust. J. Chem. 33 (1980) 333–341.
[35] S.K. Malhotra, H.J. Ringold, J. Am. Chem. Soc. 87 (1965) 3228–3236.
[36] T.M. Penning, P. Talalay, J. Biol. Chem. 256 (1981) 6851–6858.
[37] N.-C. Ha, G. Choi, K.Y. Choi, B.-H. Oh, Curr. Opin. Struct. Biol. 11 (2001) 674–
678.
5
8 (1989) 802–806.
23] E. Dorn, M. Hellwig, W. Reineke, H.J. Knackmuss, Arch. Microbiol. 99 (1974)
1–70.
6
24] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: A Laboratory Manual,
Cold Spring Harbor Laboratory Press, Plainview, New York, 1989.
25] M.P. Weiner, G.L. Costa, W. Schoettlin, J. Cline, E. Mathur, J.C. Bauer, Gene 151
[38] K.-L. Ngai, Enzymes of the beta-ketoadipate pathway in Pseudomonas putida,
PhD-thesis (1981) University of Pennsylvania.
(
1994) 119–123.