Extending the biocatalytic scope of regiocomplementary flavin-dependent halogenase enzymes

Article abstract of DOI:10.1039/c5sc00913h

Flavin-dependent halogenases are potentially valuable biocatalysts for the regioselective halogenation of aromatic compounds. These enzymes, utilising benign inorganic halides, offer potential advantages over traditional non-enzymatic halogenation chemistry that often lacks regiocontrol and requires deleterious reagents. Here we extend the biocatalytic repertoire of the tryptophan halogenases, demonstrating how these enzymes can halogenate a range of alternative aryl substrates. Using structure guided mutagenesis we also show that it is possible to alter the regioselectivity as well as increase the activity of the halogenases with non-native substrates including anthranilic acid; an important intermediate in the synthesis and biosynthesis of pharmaceuticals and other valuable products. This journal is

Full text of DOI:10.1039/c5sc00913h

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
Extending the biocatalytic scope of
regiocomplementary flavin-dependent halogenase
enzymes†
Cite this: Chem. Sci., 2015, 6, 3454
Sarah A. Shepherd,ab Chinnan Karthikeyan,ab Jonathan Latham,ab
Anna-Winona Struck,ab Mark L. Thompson,ab Binuraj R. K. Menon,ab
ab
Matthew Q. Styles,ab Colin Levy,bc David Leysbc and Jason Micklefield
*
Flavin-dependent halogenases are potentially valuable biocatalysts for the regioselective halogenation of
aromatic compounds. These enzymes, utilising benign inorganic halides, offer potential advantages over
traditional non-enzymatic halogenation chemistry that often lacks regiocontrol and requires deleterious
reagents. Here we extend the biocatalytic repertoire of the tryptophan halogenases, demonstrating how
these enzymes can halogenate a range of alternative aryl substrates. Using structure guided mutagenesis
we also show that it is possible to alter the regioselectivity as well as increase the activity of the
halogenases with non-native substrates including anthranilic acid; an important intermediate in the
synthesis and biosynthesis of pharmaceuticals and other valuable products.
Received 12th March 2015
Accepted 9th April 2015
DOI: 10.1039/c5sc00913h
www.rsc.org/chemicalscience
specicity and regioselectivity, presenting similar issues of by-
product separation as those encountered with non-enzymatic
halogenations.16,17 However, in recent years evidence has
Introduction
Halogenated aromatic compounds are widely used as pharma-
ceuticals, agrochemicals, polymers and in other materials.1–3
Aryl and heteroaryl halides are also key synthetic precursors in
many of the most powerful transition metal catalysed cross-
coupling reactions which are ubiquitous in synthesis today.4–10
Despite this, current methods for producing haloaromatic
compounds oen involve toxic or deleterious reagents, catalysts
and solvents. Traditional halogenation methods can also lack
regiocontrol rendering some regioisomers inaccessible and
resulting in unwanted by-products, including polyhalogenated
compounds, that require separation and careful disposal due to
their toxicity and/or persistence in the environment.11–15 The
development of regioselective halogenase enzymes, utilising
benign inorganic halides to deliver a range of desirable hal-
oaromatic molecules, would present an attractive alternative
that may alleviate difficulties encountered with non-enzymatic
halogenation chemistry.
emerged that it is Fe2+/a-ketoglutarate (a-KG) and particularly
avin-dependent halogenases, which are most widely utilised in
nature to catalyse regioselective halogenation reactions during
the biosynthesis of halogenated natural products.18–21 Flavin-
dependent halogenases require an associated reductase to
reduce FAD to FADH2, which is then oxidised by the halogenase
and O2 to give C4a-hydroperoxyavin (FAD-OOH) (Fig. 1A). It is
then suggested that chloride attacks the distal oxygen atom of
FAD-OOH to generate hypochlorous acid (HOCl),22–24 which H-
bonds with an active site lysine residue positioning the elec-
trophile in a spatially dened orientation relative to the
aromatic substrate.22 Alternatively, the active site lysine may
react with HOCl to generate an electrophilic chloroamine
(Fig. 1B).24 Both proposed mechanisms indicate that the lysine
residue controls the regiochemistry of the electrophilic attack
and the resulting s-complex is then deprotonated, by an active
site general base, to give the halogenated product. All of the
avin-dependent halogenases investigated to date will accept
bromide, as well as chloride, to yield either brominated or
chlorinated products.25
The heme- and vanadium-dependent haloperoxidases were
considered for potential biocatalysis applications, but their
development has been hampered by a lack of substrate
Many of the Fe2+/a-KG and avin-dependent halogenases,
process substrates which are tethered to carrier proteins of
polyketide synthase or nonribosomal peptide synthetase
assembly-line enzymes26,27 limiting their scope for synthetic
applications. On the other hand, the avin-dependent trypto-
phan halogenases can regioselectively halogenate free trypto-
phan which makes them more viable candidates for further
development as biocatalysts. Moreover, the tryptophan
aSchool of Chemistry, The University of Manchester, 131 Princess Street, Manchester,
M1 7DN, UK. E-mail: jason.mickleeld@manchester.ac.uk
bManchester Institute of Biotechnology, The University of Manchester, 131 Princess
Street, Manchester, M1 7DN, UK
cFaculty of Life Sciences, The University of Manchester, 131 Princess Street,
Manchester, M1 7DN, UK
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c5sc00913h
3454 | Chem. Sci., 2015, 6, 3454–3460
This journal is © The Royal Society of Chemistry 2015

View Article Online
Edge Article
Chemical Science
Fig. 2 Substrates and products from reactions with PyrH and PrnA.
Fig. 1 The mechanism of the tryptophan 7-halogenase PrnA. (A) The
flavin reductase from Escherichia coli (Fre) and glucose dehydroge-
nase (GDH2) from Bacillus megaterium were used to recycle FAD and
NAD+. (B) PrnA active site with the tryptophan substrate bound.
Residues which stack above and below the substrate indole are
removed for clarity. The hypohalous acid generated (step i), reacts with
the amino group of K79 (step ii) resulting in a chloroamine electrophile
which attacks the indole C7 (step iii). E346 acts as a general base to
deprotonate the s-complex.
family of Trp halogenases have not been extensively evaluated.
Accordingly, the regiocomplementary tryptophan 5- and 7-hal-
ogenases PyrH39 and PrnA,22 were overproduced in E. coli and
screened against a number of alternative non-indolic substrates
including kynurenine (2), anthranilamide (3), anthranilic acid
(4) and other anilines (Fig. 2). With the natural substrate tryp-
tophan (1), PyrH gave 5-chlorotryptophan (1a) and PrnA gave 7-
chlorotryptophan (1b) exclusively as reported previously.22,28
With PrnA, the carboxylic acid and the amino group of the Trp
substrate are engaged in a series of hydrogen bonds to tyrosine
(Y443 and Y444), glutamate (E450) and phenylalanine (F454)
residues; the indole moiety is stacked between aromatic amino
acids (H101, F103 and W455), with an hydrogen bond between
the indole NH and a backbone carbonyl group (E346). These
interactions serve to orientate the C7 of the indole towards the
key catalytic lysine (K79) and glutamate (E346) residues which
are suggested to govern the observed regioselectivity (Fig. 1).22
As we move away from the natural substrate to non-indolic
molecules the regiocomplementarity is maintained for kynur-
enine (2), with PyrH affording 5-chlorokynurenine (2a), whereas
PrnA halogenation gives 3-chlorokynurenine (2b) (Table 1).
Positioning of kynurenine into the active site of PrnA (Fig. S1†),
reveals, as anticipated, that 2 is likely to engage in the same
hydrogen bonding and p-stacking interactions as the natural
substrate (Trp), and an additional contact between the keto
group of 2 and an active site asparagine (N459) may also be
possible. With the smaller substrate anthranilamide (3), both
PyrH and PrnA favour halogenation para to amino group; PyrH
exclusively produced 5-chloroanthranilamide (3a), while PrnA
formed 5-chloroanthranilamide (3a, 86%) with 3-chloro-
halogenases characterized to date demonstrate exquisite regio-
complementarity and include: PyrH a tryptophan 5-halogen-
ase;28 SttH, Thal and KtzR tryptophan 6-halogenases;29–31 and
PrnA, KtzQ and RebH tryptophan 7-halogenases.31–33 In prin-
ciple these enzymes would provide a useful starting point from
which to develop and evolve more generic regioselective halo-
genase enzymes. Despite this potential, progress in the devel-
opment of avin-dependent halogenases for biocatalysis, has
been hampered by limited substrate scope, poor catalytic
activity22 and enzyme instability.34,35 In the case of the RebH
halogenase, notable recent reports have demonstrated how
directed evolution,34 or cross-linked enzyme aggregates,35 can
be used to improve the productivity of halogenase enzymes for
synthetic applications. In this paper we describe how Trp hal-
ogenases can be engineered, extending their biocatalytic
repertoire, to accept a wider range of aryl substrates with
improved activity and altered regioselectivities.
Results and discussion
Although PrnA36 and RebH37,38 have been shown to halogenate
indoles as well as Trp (1), the substrate specicities of the larger
This journal is © The Royal Society of Chemistry 2015
Chem. Sci., 2015, 6, 3454–3460 | 3455

View Article Online
Chemical Science
Edge Article
Table 1 Conversions and regioselectivity of PyrH and PrnA with 1–4
Table 2 Kinetics for PrnA wild type and mutants with tryptophan (1),
kynurenine (2), anthranilamide (3), and anthranilic acid (4) as substrates
Substrate
Enzyme
Conversiona
Productb (o/p-ratio)
kcat/Km  10À3
Sub
Enzyme
Km (mM)
kcat (minÀ1
)
(minÀ1 mMÀ1
)
1
2
3
4
PyrH
PrnA
PyrH
PrnA
PyrH
PrnA
PyrH
PrnA
100%
59%
67%
76%
46%
19%
<1%
1%
1a (p-100%)
1b (o-100%)
2a (p-100%)
2b (o-100%)
3a (p-100%)
3a (p-86%), 3b (o-14%)
4a (p-100%)
1
2
3
4
4
4
4
4
WT
WT
WT
WT
0.7 Æ 0.1
19.0 Æ 2.2
3267 Æ 491
3161 Æ 986
384 Æ 93
1.1 Æ 0.05
3.7 Æ 0.1
1700 Æ 200
200 Æ 10
0.7 Æ 0.1
0.16 Æ 0.05
2.4 Æ 0.6
1.0 Æ 0.4
8.9 Æ 0.7
3.4 Æ 0.6
2.1 Æ 0.1
0.51 Æ 0.07
0.93 Æ 0.05
3.66 Æ 0.67
1.82 Æ 0.04
1.27 Æ 0.07
E450K
F454K
E450K, F454K
E450K, F454R
4a (p-16%), 4b (o-84%)
3628 Æ 1108
205 Æ 16
a
Conversion aer 30 minutes with halogenase enzyme (10 mM) and
378 Æ 59
b
substrate (0.5 mM). Ratios of halogenation ortho or para to the aryl
NH/NH2 group was determined by HPLC.
is clear that although tryptophan binds most tightly to the
enzyme, kynurenine exhibits the highest turnover number. This
may be due to increased exibility of the kynurenine aryl group
within the active site which could reduce binding affinity, on
the one hand, but might also enable the aryl group to adopt a
geometry that would lead to a more stable transition state upon
attack of the chloroamine electrophile. Smaller substrates
anthranilamide (3) and anthranilic acid (4) exhibit turnover
numbers that are similar to the native substrate, but as expected
have much higher Km values than tryptophan (1) or kynurenine
(2), probably due to reduced potential for hydrogen bonding
within the active site of PrnA. Overall from kcat/Km values it is
clear that anthranilic acid 4, is the least efficient substrate with
PrnA.
Anthranilic acid is a good target for biohalogenation, given
that halogenated anthranilic acids are widely used scaffolds in
the development of pharmaceuticals, agrochemicals and other
materials of industrial importance (for some recent examples
see Fig. S3 and references in ESI†). Anthranilic acid is also a key
precursor in the biosynthesis of many bioactive natural prod-
ucts.42–44 Consequently, haloanthranilates, produced via
fermentation from microbial strains possessing an appropriate
engineered halogenase, could be useful precursors for incor-
poration into biosynthetic pathways leading to new halogenated
secondary metabolites.45,46 In light of this, we sought to engi-
neer PrnA to improve its activity and regioselectivity with
anthranilic acid through structure-guided mutagenesis. It
seems unlikely that anthranilic acid could contact PrnA active
site residues Y443, Y444, E450 and F454, as tryptophan does,
while remaining proximal to the catalytic side chains of K79 and
E346 (Fig. 1). Consequently, a series of PrnA mutants were
produced with selected active site residues mutated to lysine or
arginine. It was reasoned that the additional basic amino acid
residues might make contact with the carboxylate group of
anthranilic acid to enable tighter substrate binding and effect
the orientation of anthranilic acid relative to the catalytic active
site residues, thus altering the regioselectivity of halogenation.
Mutation of the active site tyrosines (Y443 or Y444) to either
lysine or arginine had little effect on the relative activity or
regioselectivity with 4. However, the mutation E450K not only
increased relative activity ca. 8 fold, but also enhanced ortho
selectivity producing 93% 3-chloroanthranilate (4b) compared
anthranilamide (3b, 14%) as a minor product. Presumably
many of the key substrate interactions in the active site of PrnA
and PyrH are lost, leading to more exibility in the binding
position of anthranilamide. For example with PrnA p-stacking
interactions and contact of the aniline amino group with the
backbone carbonyl of E346 may be maintained to allow for C3
(ortho) halogenation (Fig. S2A†). However, the NH2 of the amide
may also participate in hydrogen bonding to E346, which could
enable an alternative binding mode, that would position C5 of
anthranilamide closest to the key lysine residue (K79)
(Fig. S2B†). Anthranilic acid (4) proved to be a poorer substrate
than anthranilamide for both PyrH and PrnA with conversions
of 1% or less for both enzymes (Table 1). Whilst PyrH gave
exclusively 5-chloroanthranilate, PrnA gave predominately 3-
chloroanthranilate (4b, 84%) with 5-chloroanthranilate (4a,
16%) the minor product which is notably opposite to the
regioselectivity observed for PrnA with anthranilamide.
Interestingly, non-enzymatic halogenation of anthranilates
using traditional chemistry favours C5 over C3, and can oen
result in mixtures including C3 and C5 disubstituted prod-
ucts.40,41 We do not observe any dihalogenation of 3 or 4 with the
halogenases. Moreover the fact PrnA predominately haloge-
nates anthranilic acid at C3, whilst the C5 position para to the
activating amino group is intrinsically the most reactive posi-
tion,40,41 suggests that substrate binding still governs regiose-
lectivity. Presumably while anthranilate (4) may have greater
exibility in the PrnA active site, compared with the native
substrate (Trp), it is preferentially bound in an orientation that
would position C3 closer to the catalytic amino and carboxyl
groups of K79 and E346 respectively. Several other anilines also
proved to be halogenase substrates; N-phenylanthranilic acid
(5) resulted in a single chlorinated product 5a with both halo-
genases; 2-amino-4-methylbenzamide (6) producing exclusively
the 5-chloro derivative (6a) with PyrH, whereas PrnA gave a
mixture of both the 3- and 5-halogenated products (6a and 6b);
PrnA was also able to halogenate 2-amino-N-ethylbenzamide
(7), yielding a mixture of 5- and 3- halogenated derivatives (7a)
and (7b). Although only the chlorination reactions are discussed
here, similar regioselectivities were observed for bromination
reactions with various alternative non-indolic substrates.
Catalytic parameters were determined for chlorination reac-
tions catalysed by PrnA with selected substrates 1–4 (Table 2). It
3456 | Chem. Sci., 2015, 6, 3454–3460
This journal is © The Royal Society of Chemistry 2015

View Article Online
Edge Article
Chemical Science
Fig. 3 The % conversion of anthranilic 4, by PrnA mutants, and ratios
of products 5- and 3-chloroanthranilic acid (4a & 4b) after 1 hour with
halogenase enzyme (10 mM) and substrate (0.5 mM).
Fig. 4 X-ray crystal structure of PrnA F454K mutant (PDB 4Z44) with
anthranilic acid (4) positioned in the active site showing possible H-
bonding interactions with the substrate.
with 84% observed with the wild-type PrnA (Fig. 3 & Table S1†).
On the other hand, mutation of F454 to lysine and arginine
shied the regioselectivity towards para, with increasing 5-
chloroanthranilate (4a) produced (38% and 46% respectively),
while also increasing relative activity by 4-fold in the case of
F454K compared to the wild-type PrnA. Following further
mutagenesis PrnA double mutant (E450K/F454K) was found to
display a 16-fold increase in activity compared to the wild-type,
producing predominantly the para chlorinated product (4a,
54%). Additionally PrnA (E450K/F454R), also exhibited
increased activity, with a notable further shi towards para
chlorination yielding 62% 5-chloroanthranilate (4a).
The kinetic parameters were determined for anthranilic acid
(4) with the most interesting mutants (Table 2). The double
mutant (E450K/F454K) has the lowest Km of the mutants tested
and highest kcat/Km (55-fold higher than the wild type) which is
comparable with % conversion observed in Fig. 3. PrnA E450K
also shows a lower Km of 384 mM and a kcat/Km improvement
compared to the wild type enzyme, which may be due the
mutation (E450K) alleviating an unfavourable interaction
between carboxylate groups of E450 and anthranilic acid. The
F454K mutant on the other hand shows no improvement in Km,
but exhibits a kcat 7-fold higher than the parent enzyme. These
results show that structure-guided mutagenesis can signi-
cantly increase PrnA activity as well as altering regioselectivity
from predominantly ortho-halogenation with E450K to para-
halogenation with PrnA (E450K/F454R).
acid bound were unsuccessful most likely due to low binding
affinity of the substrate (Table 2). However, if anthranilic acid is
positioned in the active site (Fig. 4) it can be seen that K454
could make a salt bridge to the carboxylate group, whilst E346
contacting the amino group of anthranilate such that electro-
philic attack of the postulated K79 chloroamine would occur
preferentially at C3 (also see Fig. S4†). In the E450K structure
(PDB 4Z43) the loop region, from residues 435–445, is
completely disordered and electron density from several active
site residues is not dened. This may suggest that the E450K
mutation has destabilized the structure preventing formation of
the helical region between residues 435 and 445 observed in the
wild type structure.
Conclusions
In summary, we have expanded the range of substrates accepted
by tryptophan halogenases, PrnA and PyrH, to include alterna-
tive aryl substrates. We show that structure guided mutagenesis
of PrnA not only resulted in an increase of activity, but also
provides an example of how regioselectivity can be improved
and regiocomplementary enzymes can be created from a single
parent enzyme. The concept of generating enantiocomple-
mentary enzymes was introduced in recent years, and its
application is now more widespread with increasingly signi-
cant results.47,48 The ability to manipulate enzymatic regiocon-
trol, in this case the orientation of aryl halogenation, could offer
considerable benet for the preparation of halogenated thera-
peutic and agrochemical scaffolds and other materials1–3 (also
see Fig. S3†). Moreover, integration of engineered halogenases
into synthetic pathways including powerful chemocatalytic
cross-coupling chemistries,4–10 or biosynthetic pathways in
In order to rationalise the effects of the selected mutations
on the activity of PrnA, X-ray crystal structures for F454K and
˚
E450K were determined at 2.4 and 2.3 A respectively. The F454K
structure (PDB 4Z44) is similar to the wild type structure, except
that an extended loop is formed instead of an a-helix from
residues 435–445. Electron density for all active site residues is
present and it is evident that K454 is orientated towards the
substrate binding site and the catalytic residues K79 and E346
(Fig. 4). Attempts to obtain a structure of F454K with anthranilic
This journal is © The Royal Society of Chemistry 2015
Chem. Sci., 2015, 6, 3454–3460 | 3457

View Article Online
Chemical Science
Edge Article
vivo,42–46 could potentially be useful for production of other overexpression using 0.2 mM IPTG. Cultures were incubated
valuable synthetic or natural products.
overnight at 18 ꢀC before harvesting cells and purication of
recombinant His tagged GDH2.
Experimental
Cloning, expression, and mutagenesis of halogenases (PrnA
Protein purication
Cell pellets derived from the E. coli protein expression, were
resuspended in 25 mL lysis buffer (50 mM phosphate, 500 mM
and PyrH)
NaCl, and 10 mM imidazole, pH 7.4). Lysozyme (1 mg mLÀ1
)
A synthetic gene for the halogenase from Streptomyces rugo-
sporus (PyrH) was codon optimised using GeneArt® and
obtained from Invitrogen (UK). The pyrH genes were sub-cloned
into the pET28a vector containing an N-terminal His-tag using
the restriction enzymes NdeI and XhoI. The PrnA encoding gene
was amplied by PCR from the genomic DNA of Pseudomonas
uorescens BL915, with the primers prnA F and prnA R (Table
S1†) using the Phusion High-Fidelity PCR master mix with GC
buffer (New England Biosciences) according to the manu-
facturer's protocol. The PCR product was digested with NdeI
and NotI and ligated into the pET28a expression vector. E. coli
Arctic Express cells were subsequently transformed with the
resulting PyrH- and PrnA-containing vectors for overexpression
of the recombinant halogenases. LB medium containing
kanamycin (50 mg mLÀ1) was inoculated with the transformants
was added to the cell resuspension which was then incubated at
30 ꢀC for 1 h. Cells were disrupted by sonication and the lysate
was claried by centrifugation (4 ꢀC, 40 min, 10 000Âg). The
soluble cell extract was loaded onto a HisTrap™ FF crude
column and puried by FPLC. The column was washed with 20
mM phosphate buffer (pH 7.4) containing 20 and 60 mM
imidazole and 500 mM NaCl. Puried PyrH and PrnA were
eluted in phosphate buffer containing 500 mM imidazole, Fre
and GDH2 at 250 mM imidazole. Protein samples were sub-
jected to buffer exchange with phosphate buffer containing 10%
glycerol using a Vivaspin 20ꢀcentricon (10 000 MWCO) before
subsequently storing at À20 C.
Biotransformations, and characterisation of activity and
regioselectivity
ꢀ
and incubated at 37 C overnight. The cells were then diluted
1 : 100 in fresh LB medium and incubated shaking at 30 ꢀC
until an optical density (OD600nm) of 0.6. The cells were subse-
quently incubated at 4 ꢀC for 30 minutes for cold-shock, and
protein expression was induced with addition of IPTG (0.1 mM),
before growing overnight at 15 ꢀC, followed by harvesting of
cells (4 ꢀC, 20 min, 4000Âg). Construction of halogenase
mutants was achieved using the QuickChange® site-directed
mutagenesis kit (Stratagene). The previously described
constructs were used as templates and mutations were intro-
duced via the manufacturer's protocol using the primers shown
in Table S1.†
The following conditions were used for assays to determine
regioselectivity and %conversions (Table 1 and Fig. 3). Puried
halogenase enzyme (10 mM) was incubated at 30 ꢀC with
shaking for 1 h with Fre (5 mM), FAD (1 mM), NADH (2.5 mM),
MgCl2 (10 mM) and substrate (0.5 mM) in a total volume of 100
mL in 10 mM potassium phosphate buffer, pH 7.4. Reactions
were stopped by incubating at 95 ꢀC for 5 min and precipitated
protein was removed by centrifugation before analysis via HPLC
on an Agilent Technologies 1260 system using an Agilent Zorbax
Eclipse Plus C18 4.6  100 mm  3.5 mm column. For trypto-
phan, absorbance was measured at 280 nm. Kynurenine,
anthranilamide and anthranilic acid absorbance were
measured at 254 nm, with a 5 min gradient 5–75% H2O/aceto-
nitrile + 0.1% formic acid.
Cloning and expression of the avin reductase (Fre) and
glucose dehydrogenase (GDH2)
The gene coding for the avin reductase (Fre) from E. coli49 was
To obtain kinetic parameters for selected reactions (Table 2)
amplied from E. coli BL21 genomic DNA using the oligonu- the concentration of the assay components were varied
cleotides fre F and fre R (Table S1†), then digested using the according to conditions required to provide the best t for the
restriction enzymes KpnI and XhoI, before ligating into the Michaelis–Menten curve. However in each case the total assay
pET45b expression vector containing an N-terminal His-tag. E. volume was 150 mL and the Fre concentration was always in
coli BL21 (DE3) cells transformed with the recombinant excess in order to ensure the production of reduced aꢀvin was
Fre plasmid were initially grown overnight at 37 ꢀC in LB not a rate limiting factor. Assays were performed at 30 C with
medium containing ampicillin (50 mg mLÀ1) before diluting shaking at 800 rpm. Plates and assay components were pre-
1 : 100 in fresh LB medium the following day. Cultures incubated at 30 ꢀC. Assays were started by the addition of
were subsequently incubated at 37 ꢀC with shaking until an substrate using a multi-channel pipette. Substrate was added at
OD600 of 0.6. Recombinant protein overexpression was induced 15 second intervals and the reaction terminated with the
with IPTG (1 mM). Cultures were incubated for a further 5 h at addition of formic acid. All assays were performed in triplicate.
ꢀ
ꢀ
30 C before harvesting cells (4 C, 20 min, 4000Âg) and puri-
Larger scale assays were carried out to obtain chlorinated
cation of recombinant N-terminal His tagged Fre. A pET products for characterisation using halogenase (25 mM), Fre (2.5
21b vector containing a gene encoding GDH2 from Bacillus mM), GDH2 (12.5 mM), substrate (3 mM), MgCl2 (100 mM), FAD
megaterium50 was kindly provided by Prof. Nigel Scrutton (10 mM), NADH (10 mM), glucose (200 mM) in 10 mM potassium
(University of Manchester). E. coli BL21 (DE3) cells transformed phosphate buffer, pH 7.4. Assays were run at 30 ꢀC with shaking.
with the recombinant GDH2 plasmid were cultivated using Reactions were stopped by incubating at 95 C for 5 min and
the same method as described previously with induction of precipitated protein was removed by centrifugation (4 ꢀC,
ꢀ
3458 | Chem. Sci., 2015, 6, 3454–3460
This journal is © The Royal Society of Chemistry 2015

View Article Online
Edge Article
Chemical Science
10 min, 12 000Âg) before analysis by HPLC. For reactions with 21 S. Cadel-Six, C. Dauga, A. M. Castets, R. Rippka, C. Bouchier,
low yield, multiple 10 mL reactions were performed and
combined prior to purication by semi-prep HPLC. For trypto-
N. Tandeau de Marsac and M. Welker, Mol. Biol. Evol., 2008,
25, 2031–2041.
´
phan and N-phenylanthranilic acid, absorbance was measured 22 C. Dong, S. Flecks, S. Unversucht, C. Haupt, K.-H. van Pee
at 280 nm. All other substrate and products absorbance were and J. H. Naismith, Science, 2005, 309, 2216–2219.
measured at 254 nm, with a 5 min gradient 5–75% H2O/aceto- 23 E. Yeh, L. Cole, E. Barr and J. Bollinger, Biochemistry, 2006,
nitrile + 0.1% formic acid. Products were subsequently charac- 45, 7904–7912.
terised using 1D and 2D NMR, LRMS, HRMS and UV and are in 24 E. Yeh, L. C. Blasiak, A. Koglin, C. L. Drennan and
good agreement with the literature data (see ESI†).
C. T. Walsh, Biochemistry, 2007, 46, 1284–1292.
25 A rare example of a avin-dependent halogenase with
specicity for bromide and iodide, but not chloride, was
recently reported: V. Agarwal, A. A. El Gamal,
K. Yamanaka, D. Poth, R. D. Kersten, M. Schorn,
E. E. Allen and B. S. Moore, Nat. Chem. Biol., 2014, 10,
640–647.
Acknowledgements
We acknowledge BBSRC (grants BB/K00199X/1 & BB/I020764/1),
GlaxoSmithKline and CoEBio3 for nancial support. We
acknowledge Diamond Light Source for time on beamlines i02,
i03, i04 & i04-1 under proposal MX8997.
26 D. Khare, B. Wang, L. Gu, J. Razelun, D. H. Sherman,
˚
W. H. Gerwick, K. Hakansson and J. L. Smith, Proc. Natl.
Acad. Sci. U. S. A., 2010, 107, 14099–14104.
27 P. C. Dorrestein, E. Yeh, S. Garneau-Tsodikova, N. L. Kelleher
and C. T. Walsh, Proc. Natl. Acad. Sci. U. S. A., 2005, 102,
13843–13848.
Notes and references
1 Y. Lu, Y. Liu, Z. Xu, H. Li, H. Liu and W. Zhu, Expert Opin.
Drug Discovery, 2014, 7, 375–383.
´
¨
28 S. Zehner, A. Kotzsch, B. Bister, R. D. Sussmuth, C. Mendez,
2 P. Jeschke, Pest Manage. Sci., 2009, 66, 10–27.
3 O. Bolton, K. Lee, H.-J. Kim, K. Y. Lin and Ji. Kim, Nat. Chem.,
2011, 3, 207–212.
´
J. A. Salas and K.-H. van Pee, Chem. Biol., 2005, 12, 445–452.
29 J. Zeng and J. Zhan, Biotechnol. Lett., 2011, 33, 1607–1613.
4 K. C. Nicolaou, P. G. Bulger and D. Sarlah, Angew. Chem., Int. 30 C. Seibold, H. Schnerr, J. Rumpf, A. Kunzendorf,
Ed. Engl., 2005, 44, 4442–4489.
C. Hatscher, T. Wage, A. J. Ernyei, C. Dong, J. H. Naismith
´
5 F.-S. Han, Chem. Soc. Rev., 2013, 42, 5270–5298.
6 J. F. Hartwig, Nature, 2008, 455, 314–322.
7 C. Valente, S. Calimsiz, K. H. Hoi, D. Mallik, M. Sayah and
M. G. Organ, Angew. Chem., Int. Ed., 2012, 51, 3314–3332.
8 S. Z. Tasker, E. A. Standley and T. F. Jamison, Nature, 2014,
509, 299–309.
and K.-H. Van Pee, Biocatal. Biotransform., 2006, 24, 401–408.
31 J. R. Heemstra and C. T. Walsh, J. Am. Chem. Soc., 2008, 130,
14024–14025.
32 S. Kirner, P. E. Hammer, D. S. Hill, A. Altmann, I. Fischer,
´
L. J. Weislo, M. Lanahan, K. Van Pee and J. M. Ligon, J.
Bacteriol., 1998, 180, 1939–1943.
´
˜
9 Z. Zuo, D. T. Ahneman, L. Chu, J. A. Terrett, A. G. Doyle and 33 C. Sanchez, L. Zhu, A. F. Brana, A. P. Salas, J. Rohr,
´
D. W. C. MacMillan, Science, 2014, 345, 437–440.
10 E. J. Cho, T. D. Senecal, T. Kinzel, Y. Zhang, D. A. Watson and
S. L. Buchwald, Science, 2010, 328, 1679–1681.
11 F. Alonso, I. P. Beletskaya and M. Yus, Chem. Rev., 2002, 102,
4009–4091.
C. Mendez and J. A. Salas, Proc. Natl. Acad. Sci. U. S. A.,
2005, 102, 461–466.
34 C. B. Poor, M. C. Andorfer and J. C. Lewis, ChemBioChem,
2014, 15, 1286–1289.
35 M. Frese and N. Sewald, Angew. Chem., Int. Ed., 2015, 54,
298–301.
12 S. A. Rowlands, A. K. Hall, P. G. McCormick, R. Street,
¨
´
R. J. Hart, G. F. Ebell and P. Donecker, Nature, 1994, 367, 36 M. Holzer, W. Burd, H.-U. Reibig and K.-H. van Pee, Adv.
223. Synth. Catal., 2001, 343, 591–595.
13 B. C. Kelly, M. G. Ikonomou, J. D. Blair, A. E. Morin and 37 J. T. Payne, M. C. Andorfer and J. C. Lewis, Angew. Chem., Int.
F. A. P. C. Gobas, Science, 2007, 317, 236–239.
Ed. Engl., 2013, 52, 5271–5274.
14 J. M. Suita, A. Horowitz, D. R. Shelton and J. M. Tiedje, 38 M. Frese, P. H. Guzowska, H. Voß and N. Sewald,
Science, 1982, 218, 1115–1117. ChemCatChem, 2014, 6, 1270–1276.
15 B. Xu, H.-Z. Zhu, Y.-L. Lin, K.-Y. Shen, W.-H. Chu, C.-Y. Hu, 39 X. Zhu, W. De Laurentis, K. Leang, J. Herrmann, K. Ihlefeld,
´
K.-N. Tian, S. A. Mwakagenda and X.-Y. Bi, Water, Air, Soil
Pollut., 2012, 223, 4429–4436.
K.-H. van Pee and J. H. Naismith, J. Mol. Biol., 2009, 391, 74–
85.
´
16 K.-H. van Pee, C. Dong, S. Flecks, J. Naismith, E. P. Patallo 40 B. Ojo and B. K. Chowdhury, Synth. Commun., 2012, 42,
and T. Wage, Adv. Appl. Microbiol., 2006, 59, 127–157.
1002–1009.
17 P. Bernhardt, T. Okino, J. M. Winter, A. Miyanaga and 41 D. Roche, K. Prasad, O. Repic and T. J. Blacklock,
B. S. Moore, J. Am. Chem. Soc., 2011, 4268–4270.
Tetrahedron Lett., 2000, 41, 2083–2085.
18 A. Butler and M. Sandy, Nature, 2009, 460, 848–854.
42 B. D. Ames and C. T. Walsh, Biochemistry, 2010, 49, 3351–3365.
19 C. S. Neumann, D. G. Fujimori and C. T. Walsh, Chem. Biol., 43 S. W. Haynes, X. Gao, Y. Tang and C. T. Walsh, J. Am. Chem.
2008, 15, 99–109. Soc., 2012, 134, 17444–17447.
20 P. Gao and Y. Huang, Appl. Environ. Microbiol., 2009, 75, 44 S. W. Haynes, X. Gao, Y. Tang and C. T. Walsh, ACS Chem.
4813–4820.
Biol., 2013, 8, 741–748.
This journal is © The Royal Society of Chemistry 2015
Chem. Sci., 2015, 6, 3454–3460 | 3459

View Article Online
Chemical Science
Edge Article
45 W. S. Glenn, E. Nims and S. E. O'Connor, J. Am. Chem. Soc., 48 Q. Wu, P. Soni and M. T. Reetz, J. Am. Chem. Soc., 2013, 135,
2011, 133, 19346–19349. 1872–1881.
46 A. Deb Roy, S. Gruschow, N. Cairns and R. J. M. Goss, J. Am. 49 G. Spyrou, E. Haggard-Ljungquist, M. Krook, H. Jornvall,
˚
¨
¨
Chem. Soc., 2010, 132, 12243–12245.
E. Nilsson and P. Reichard, J. Bacteriol., 1991, 173, 3673–
47 P. F. Mugford, U. G. Wagner, Y. Jiang, K. Faber and
3679.
R. J. Kazlauskas, Angew. Chem., Int. Ed. Engl., 2008, 47, 50 T. Nagao, T. Mitamura and X. Wang, J. Bacteriol., 1992, 174,
8782–8793.
5013–5020.
3460 | Chem. Sci., 2015, 6, 3454–3460
This journal is © The Royal Society of Chemistry 2015