FADH2-dependence of tryptophan 7-halogenase

Article abstract of DOI:10.1002/adsc.200505029

Tryptophan 7-halogenase (Trp 7-hal) catalyses the regioselective chlorination and bromination of tryptophan. For halogenating activity, Trp 7-hal requires FADH₂ produced from FAD and NADH by a flavin reductase, halide ions (chloride or bromide), molecular oxygen and tryptophan as the organic substrate. Investigations of the flavin dependence showed that purified Trp 7-hal itself does not contain flavin. Keeping the Trp 7-hal separated from the flavin reductase during the reaction revealed that Trp 7-hal can use diffusible FADH₂ produced by a flavin reductase showing that direct contact between the halogen ase and the flavin reductase is not absolutely necessary. Thus, the reaction also proceeds when chemically reduced flavin is used. For the catalytic regeneration of FADH₂, the organometallic complex (pentamethylcyclopentadienyl)rhodium-bipyridine {[Cp*Rh(bpy) (H₂O)]²⁺} can be employed as the redox catalyst with formate as the electron donor. With this chemoenzymatic system about 85% yields of the product formed by the two-component enzyme system consisting of Trp 7-hal and a flavin reductase were obtained.

Full text of DOI:10.1002/adsc.200505029

FULL PAPERS
FADH -Dependence of Tryptophan 7-Halogenase
2
a
b, c
b, d
Susanne Unversucht, Frank Hollmann, Andreas Schmid,
a,
Karl-Heinz van P e´ e *
a
Institut fꢀr Biochemie, Technische Universitꢁt Dresden, Bergstraße 66, 01062 Dresden, Germany
Fax: (þ49)-351-463-35506, e-mail: karl-heinz.vanpee@chemie.tu-dresden.de
Institut fꢀr Biotechnologie, ETH-Hçnggerberg, HPT, 8093 Zꢀrich, Switzerland
Present address: Max-Planck-Institut fꢀr Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45470 Mꢀlheim an der Ruhr,
Germany
b
c
d
Lehrstuhl fꢀr Biotechnik, Universitꢁt Dortmund, Emil-Figge-Straße 66, 44221 Dortmund, Germany
Received: January 21, 2005; Accepted: March 29, 2005
th
Dedicated to Prof. Franz Lingens on the occasion of his 80 birthday.
Abstract: Tryptophan 7-halogenase (Trp 7-hal) cataly- reduced flavin is used. For the catalytic regeneration
ses the regioselective chlorination and bromination of of FADH , the organometallic complex (pentame-
2
tryptophan. For halogenating activity, Trp 7-hal re- thylcyclopentadienyl)rhodium-bipyridine
2
þ
quires FADH produced from FAD and NADH by {[Cp*Rh(bpy)(H O)] } can be employed as the re-
2
2
a flavin reductase, halide ions (chloride or bromide), dox catalyst with formate as the electron donor.
molecular oxygen and tryptophan as the organic sub- With this chemoenzymatic system about 85% yields
strate. Investigations of the flavin dependence showed of the product formed by the two-component enzyme
that purified Trp 7-hal itself does not contain flavin. system consisting of Trp 7-hal and a flavin reductase
Keeping the Trp 7-hal separated from the flavin re- were obtained.
ductase during the reaction revealed that Trp 7-hal
can use diffusible FADH produced by a flavin reduc-
2
tase showing that direct contact between the halogen- Keywords: enzymatic halogenation; FADH regenera-
2
ase and the flavin reductase is not absolutely necessa- tion; flavoenzyme; monooxygenase; organometallic
ry. Thus, the reaction also proceeds when chemically catalyst
[
3]
Introduction
pounds. Whereas a number of three-dimensional
structures of haloperoxidases are available and their re-
Living organisms have an enormous potential for the action mechanisms have been elucidated, hardly any-
[
1]
formation of halogenated organic compounds. In con- thing is known about the reaction mechanism of the
trast to chemical synthesis of organohalogens, enzyme- new type of halogenases.
catalysed halogenation reactions occurring in biosyn-
Tryptophan 7-halogenase (Trp 7-hal) from the pyrrol-
thetic pathways proceed under mild conditions and nitrin-producing soil bacterium Pseudomonas fluores-
with very high regioselectivity and substrate specificity. cens is the only halogenase whose purification in active
For almost 40 years, it was assumed that biological halo- form has been published up to now. This enzyme cataly-
genation reactions are catalysed by haloperoxidases. ses the incorporation of the halogen atoms chlorine and
However, halogenation reactions catalysed by haloper- bromine into the 7-position of its natural substrate tryp-
oxidases lack regioselectivity and substrate specificity. tophan in the presence of oxygen and an additional en-
This is due to the formation of free hypochlorite orhypo- zyme component, a flavin reductase. This flavin reduc-
bromite, respectively, as the actual halogenating agent tase produces the reduced flavin (FADH ) required by
2
[
4]
of haloperoxidases. A few years ago, a new type of hal- the halogenase.
ogenating enzymes was detected. In contrast to haloper- The flavin reductases present in the halogenase pro-
oxidases, halogenation reactions catalysed by these hal- ducer can be substituted by other flavin reductases,
ogenases proceed with very high regioselectivity and like the flavin reductase from the ribonucleotide reduc-
[
2]
[4]
substrate specificity.
tase complex (Fre) from E. coli suggesting that a spe-
During the last few years, it became evident that these cific interaction between the reductase and the halogen-
halogenases are the major type of halogenating enzymes ase is not necessary. This is analogous to the situation
involved in the biosynthesis of halogenated com- found in a group of monooxygenases which are also a
Adv. Synth. Catal. 2005, 347, 1163–1167
DOI: 10.1002/adsc.200505029
ꢂ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1163

FULL PAPERS
Susanne Unversucht et al.
[
4–7]
two-component system.
In the case of these en- the two enzymes and by the use of chemically synthes-
zymes, it is believed that they use freely diffusible flavin. ised FADH clearly prove that the halogenase can use
2
Interestingly, the genes of these monooxygenases and freely diffusible FADH for the catalysis of the chlorina-
2
the flavin reductases are clustered, which does not tion of tryptophan. When FAD was substituted by FMN,
[
4,5]
seem to be the case with the halogenases.
no reaction occurred, showing that the halogenase is
The amino acid sequences of these monooxygenases specific for FADH2.
show hardly any similarities to those of the halogenases.
Even an absolutely conserved motif for a nucleotide-
binding site (GxGxxG) near the amino terminus of the Halogenation using an FADH Regeneration System
2
halogenases is only presentin afew of these monooxyge-
nases. This nucleotide-binding site could be a binding The results described above suggested that it might be
site for FADH or an additionally required NADH.
possible to establish a continuous reaction system on
2
For the elucidation of the reaction mechanism, the in- the basis of a non-enzymatic FADH regeneration. For
2
terplay of the two protein components with respect to this purpose an enzyme-independent in situ regene-
the FADH transfer and the NADH requirement was in- ration system consisting of the organometallic (penta-
2
vestigated in detail.
methylcyclopentadienyl)rhodium-bipyridine complex
2
þ
{
[Cp*Rh(bpy)(H O)] } as the catalyst and formate as
2
[
11]
the electron donor was used. In this system, by which
the flavin reductase as well as NADH could be substitut-
Results and Discussions
ed, FADH is provided for the halogenase by the reac-
2
tion of the organometallic catalyst (Scheme 1).
Direct Interaction of the Flavin Reductase and the
Halogenase Component
In comparison with the two-component enzyme sys-
tem flavin reductase/tryptophan halogenase, the
amount of product obtained in the chemoenzymatic as-
To investigate whether a direct contact between the fla- say was about 85% when catalyst concentrations of 0.2–
vin reductase and the halogenase is necessary, the two 0.4 mM were used. The addition of flavin reductase to
protein components were separated by a dialysis mem- the chemoenzymatic system had no influence on the
brane. Even under these conditions, chlorination of product formation, also suggesting that the role of the
tryptophan occurred, although only with about 60% flavin reductase used (Fre) was just that of a FADH2
(
130 mU compared to 220 mU of halogenating activity) provider. The product formation decreased with catalyst
of the efficiency seen under normal conditions. From concentrations below 0.2 mM from 53 mM 7-chloro-
corresponding results obtained for the 4-hydroxyphenyl- tryptophan at 0.2 mM catalyst concentration to 38 mM
[
9]
acetate monooxygenase, Galan et al. have concluded 7-chlorotryptophan at 50 mM catalyst concentration, in-
that direct contact between the reductase and the mono- dicating that FADH became rate-limiting. The reduc-
2
oxygenase is not necessary and that this monooxygenase tion in product formation in the chemoenzymatic system
is able to bind freely diffusible FADH . The result ob- compared to the enzymatic two-component system is
2
tained with Trp 7-hal separated from the flavin reduc- partially due to the fact that the use of formate in the
tase suggests a similar interpretation for the halogenase chemoenzymatic system leads to a decrease in Trp 7-
system.
hal activity. Formate concentrations above 200 mM
Although the separation of the two enzymes showed even result in total inhibition of Trp 7-hal. Additionally,
that direct contact of the two protein components is
not absolutely required for the halogenation reaction
to proceed, it did not show whether additional NADH
was required by the halogenase. This question could
be unequivocally answered by the use of chemically re-
duced FAD, since in this experiment NADH was not
present at all. When chemically reduced FAD (80 mM)
was added to the halogenase assay under anaerobic con-
ditions at room temperature and then oxygen was al-
lowed to slowly diffuse into the solution, chlorination
of tryptophan (8.6 mM) occurred. Since FADH is con-
2
sumed through the very rapid, spontaneous reaction
with molecular oxygen to 4a-flavin hydroperoxide
[
10]
which then decomposes into FAD and H O2 the reac-
2
tion stops after a short period of time whereby only
about 10% of the reduced flavin were used for the halo- Scheme 1. The chemoenzymatic system of Trp 7-hal and the
2
þ
genation reaction. The results obtained by separating cofactor regeneration complex [Cp*Rh(bpy)(H O)]
.
2
1164
ꢂ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1163–1167

FADH -Dependence of Tryptophan 7-Halogenase
FULL PAPERS
2
the substrates for Trp 7-hal, tryptophan and chloride,
also have a negative influence on the rate of cofactor re-
generation by the organometallic complex.
Although these results leave no doubt that Trp 7-hal
can use freely diffusible FADH and that direct interac-
2
tion between the flavin reductase and Trp 7-hal is not ab-
solutely necessary, it is difficult to believe that in vivo
such a “waste” of in situ formed flavin reduction equiv-
alents should occur. The very low turnover number of
ꢀ
1
Trp 7-hal of 6.4 h in the enzymatic two-component
system with both enzymes in the same solution indicates
that this in vitro system is far from optimal. In this sys-
tem, incubation for 6 h would result in the formation
Figure 1. Time-dependent chemoenzymatic formation of 7-
of 1 mM FADH which could lead to the formation of
2
chlorotryptophan (^) and theoretical amount of FADH re-
2
1
mM 7-chlorotryptophan, however, only 200 mM 7-
2
þ
generated by [Cp*Rh(bpy)(H O)] (~). General conditions:
2
chlorotryptophan are formed indicating that 80% of
1
1
0 mM phosphate buffer (pH 7.2), 0.25 mM l-tryptophan,
the formed FADH are lost through autooxidation; in
ꢀ
2
0 mM FAD, 25 mM Cl , 40 mM formate, 308C; (^): 85 mM
the chemoenzymatic system even 90% are lost (Fig-
ure 1). This corresponds to a loss of energy that would
be extremely uneconomical for living cells. Thus, a direct
2þ
Trp 7-hal, 0.2 mM [Cp*Rh(bpy)(H O)]
;
(~): 0.2 mM
2
2
þ
[
Cp*Rh(bpy)(H O)] .
2
transfer from the FADH -producing enzyme, the flavin
2
reductase, to the FADH -consuming enzyme, the halo- NADH. The FADH must be protected from the unpro-
2
2
genase or a monooxygenase, for example, through the ductive spontaneous reaction with molecular oxygen by
formation of an enzyme complex, seems much more ef- binding to flavin-free Trp 7-hal. This hypothesis is sup-
fective. Such a protein-protein interaction was thought ported by the fact that all halogenases purified in active
to occur in the case of the FADH -dependent 4-hydroxy- form so far, do not contain flavin (unpublished results).
2
phenylacetate monooxygenase (HpaB) from E. coli W Through the reaction of the halogenase-bound FADH2
(
(
ATCC 11105) and the corresponding flavin reductase with oxygen, enzyme-bound 4a-flavin hydroperoxide
HpaC). However, the search for the existence of a pro- would be formed which could attack the organic sub-
tein-protein complex using size-exclusion chromatogra- strate. This would lead to activation for the attack of
phy under anaerobic conditions has not provided con- chloride as a nucleophile. The resulting halohydrin
[
6]
clusive proof for the existence of such a complex.
could give rise to the end product 7-chlorotryptophan
[
13]
Also the search for a potential enzyme complex between via dehydration. Hubbard and Walsh suggested an al-
the styrene monooxygenase (StyA) and the flavin re- ternative for the actual halogenation step in which chlor-
ductase (StyB) from Pseudomonas sp. VLB120 via ul- ide would react with the 4a-flavin hydroperoxide and
tracentrifugation gave no indications for such a protein thus form a flavin-bound, activated electrophilic halide
[
8]
interaction in FADH transfer. Only in the luciferase/ species (Scheme 2). Both hypotheses are based on the
2
flavin reductase system from Vibrio harveyi, could an initial formation of enzyme-bound 4a-flavin hydroper-
FMNH -tunnel system in the donor-acceptor enzyme oxide. Hopefully, further investigations, based on the re-
2
[
12]
complex be shown to exist in vitro and in vivo.
cently elucidated three-dimensional structure of Trp 7-
[
14]
Another possibility would be that a third, yet un- hal will help to elucidate the reaction mechanism of
known, component functions as an FADH -shuttle and this new type of enzymes.
2
thus could guarantee the safe and protected transport
of the reduction equivalents. However, so far there are
no data available to support this assumption, either on Conclusions
a genetic or on an enzymatic basis.
Tryptophan 7-halogenase is the first member of the new
typeofhalogenatingenzymes, theflavin-dependenthalo-
Hypothetical Reaction Mechanism of the Two-
Component Halogenase System
genases. The flavin-free enzyme requires FADH for the
regioselective chlorination or bromination of trypto-
2
phan. It could be shown that FADH can be used by
2
Based on the data obtained for the interaction of the fla- this enzyme in freely diffusible form; direct contact be-
vin reductase and Trp 7-hal and the resulting FADH de- tween the flavin reductase providing FADH and the ac-
2
2
pendence a monooxygenase-like mechanism can be tual halogenase is not absolutely necessary. The flavin
postulated for Trp 7-hal (Scheme 2). In a first reaction, reductase and NADH can be substituted by a chemoen-
preceding the actual halogenation reaction, FADH is zymatic FADH regeneration system consisting of the
2
2
2
þ
produced by a flavin reductase with oxidation of organometallic catalyst [Cp*Rh(bpy)(H O)] and for-
2
Adv. Synth. Catal. 2005, 347, 1163–1167
asc.wiley-vch.de
ꢂ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1165

FULL PAPERS
Susanne Unversucht et al.
Scheme 2. Proposed hypothetical reaction mechanisms of FADH -dependent Trp 7-hal with (A) nucleophilic attack of chlor-
2
[
3]
[13]
ide or (B) via formation of an electropilic, flavin-bound chlorine.
mate as the electron donor for FAD reduction. These re- (standard Trp 7-hal assay). The reaction was stopped and ana-
[4]
lysed by HPLC as described by Keller et al.
sults suggest similarities in the reaction mechanism with
a group of monooxygenases which can also use freely
diffusible FADH and they are an important step to-
2
wards the elucidation of the reaction mechanism of the
new type of halogenases.
Reaction with Separated Flavin Reductase and Trp 7-
hal
Trp 7-hal and the flavin reductase (Fre) were separated with
the help of a Serva dialysis tubing 20/32 membrane with an ex-
clusion limit of 12–14 kDa. The dialysis tubing was filled with
Experimental Section
0
.55 mU Fre in a total volume of 100 mL in 10 mM potassium
phosphate buffer (pH 7.2). The dialysis tubing was then placed
in a solution of 110 mL Trp 7-hal-containing protein solution
Purification of Trp 7-hal and Enzyme Assay
ꢀ
1
(0.3 mg mL ), 10 mM FAD, 0.25 mM l-tryptophan, 12.5 mM
[4]
Trp 7-hal was purified as described previously. Chlorinating
activity was determined by incubating 55 mL of Trp 7-hal-con-
taining protein solution with 0.55 mU flavin reductase (Fre)
from E. coli, 10 mM FAD, 0.25 mM l-tryptophan, 12.5 mM
MgCl , and 2.4 mM NADH in a total volume of 700 mL in
2
10 mM potassium phosphate buffer (pH 7.2) and incubated
at 308C for 6 h. As a reference, an assay was used in which
the dialysis tubing had been filled with 100 mL 10 mM potassi-
um phosphate buffer (pH 7.2) and Fre had been added to the
Trp 7-hal-containing solution. Detection and quantification
MgCl , and 2.4 mM NADH in a total volume of 200 mL in
2
10 mM potassium phosphate buffer (pH 7.2) at 308C for 6 h
1166
ꢂ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1163–1167

FADH -Dependence of Tryptophan 7-Halogenase
FULL PAPERS
2
of the formed 7-chlorotryptophan was done by HPLC analysis References
[4]
as described previously.
[
1] G. W. Gribble, in: The Handbook of Environmental
Chemistry, Vol. 3, Natural Production of Halogen Com-
pounds, (Ed.: G. W. Gribble), Springer Verlag, Berlin,
Reactions without Flavin Reductase
2
003, pp. 1–15.
Chemical reduction of FAD was performed by adding a freshly
prepared titanium(III) citrate solution (70 mM) to an FAD sol-
ution (100 mM) that had been flushed with argon. 160 mL of
this solution were added to a Trp 7-hal (0.3 mg mL ) assay sol-
ution (without Fre, FAD, and NADH) that had also been flush-
ed with argon. Oxygen was allowed to diffuse into the assay sol-
ution by incubating the open reaction tubes at 308C. After 3 h
the reaction was stopped and analysed by HPLC.
[
2] K.-H. van P e´ e, S. Unversucht, Chemosphere 2003, 52,
299–312.
[
16]
ꢀ
1
[3] K.-H. van P e´ e, S. Zehner, in: The Handbook of Environ-
mental Chemistry, Vol. 3, Natural Production of Halogen
Compounds, (Ed.: G. W. Gribble), Springer Verlag, Ber-
lin, 2003, pp. 171–199.
4] S. Keller, T. Wage, K. Hohaus, M. Hçlzer, E. Eichhorn,
K.-H. van P e´ e, Angew. Chem. Int. Ed. 2000, 39, 2300–
2302.
[
In preparation for the investigations of the cofactor regener-
ation with the organometallic complex [Cp*Rh(bpy)(H O)]
2
þ
2
and for the determination of the influence of formate on the
Trp 7-hal activity, formate was added to the Trp 7-hal standard
assay in concentrations of 5–500 mM.
[5] T. M. Louie, C. M. Webster, L. Xun, J. Bacteriol. 2002,
184, 3492–3500.
[6] T. M. Louie, X. Sunney, L. Xun, Biochemistry 2003, 42,
7509–7517.
[7] U. Kirchner, A. H. Westphal, R. Mꢀller, W. J. H. van -
Berkel, J. Biol. Chem. 2003, 278, 47545–47553.
2
þ
The organometallic complex [Cp*Rh(bpy)(H O)] was
2
[15]
prepared as described by Kçlle and Grꢁtzel. The influence
of the components of the Trp 7-hal assay on the activity of
the cofactor regeneration system was analysed using the al-
ready described concentrations of the individual components.
For the determination of the activity of the catalyst, the amount
of H O formed during the reaction was measured using a col-
[
8] K. Otto, K. Hofstetter, M. Rçthlisberger, B. Witholt, A.
Schmid, J. Bacteriol. 2004, 186, 5292–5302.
[
9] B. Gal a´ n, E. Dꢃaz, M. A. Prieto, J. L. Garcꢃa, J. Bacteriol.
2
2
2
000, 182, 627–636.
[17]
2þ
orimetric assay. The Trp 7-hal/[Cp*Rh(bpy)(H O)] assay
2
[
[
10] V. Massey, J. Biol. Chem. 1994, 269, 22459–22462.
11] F. Hollmann, P.-Ch. Li, B. Witholt, A. Schmid, J. Am.
Chem. Soc. 2003, 125, 8209–8217.
consisted of 0.2 mM organometallic complex, 10 mM FAD,
ꢀ
1
5
0
5 mL Trp 7-hal-containing protein solution (0.3 mg mL ),
.25 mM l-tryptophan, 12.5 mM MgCl , and 40 mM formate
2
[
[
[
[
[
[
12] J. C. Low, Sh.-Ch. Tu, Photochem. Photobiol. 2003, 77,
in a total volume of 200 mL in 10 mM potassium phosphate buf-
fer (pH 7.2). As references, Trp 7-hal standard assays contain-
ing the corresponding formate concentration (40 mM) were
used.
4
46–452.
13] B. K. Hubbard, C. T. Walsh, Angew. Chem. Int. Ed. 2003,
2, 730–765.
4
14] C. Dong, A. Kotzsch, M. Dorward, K.-H. van P e´ e, J. H.
Naismith, Acta Cryst. D Biol. Cryst. 2004, 60, 1438–1440.
15] A. J. B. Zehnder, K. Wuhrmann, Science 1976, 194,
Acknowledgements
1
165–1166.
16] U. Kçlle, M. Grꢁtzel, Angew. Chem. Int. Ed. Engl. 1987,
6, 567–570.
This work was supported by the Deutsche Forschungsgemein-
schaft (Pe 348/17–1 and the Graduiertenkolleg “Struktur-Ei-
genschafts-Beziehungen bei Heterocyclen”), by the Sꢀchsische
Staatsministerium fꢁr Umwelt und Landwirtschaft, and the
Fonds der Chemischen Industrie.
2
17] Y. Saito, M. Mifune, S. Nakashima, J. Odo, Y. Tanaka,
Talenta 1987, 34, 667–669.
Adv. Synth. Catal. 2005, 347, 1163–1167
asc.wiley-vch.de
ꢂ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1167