Halogenation of Unactivated Carbon Centers in Natural Product
Biosynthesis: Trichlorination of Leucine during Barbamide Biosynthesis
Danica P. Galonic ´, Fre ´de ´ric H. Vaillancourt,†and Christopher T. Walsh*
Department of Biological Chemistry and Molecular Pharmacology, HarVard Medical School,
Boston, Massachusetts 02115
Received January 9, 2006; E-mail: email@example.com
Halogenation is a common structural feature of more than 4500
natural products of both marine and terrestrial origin.1This
biosynthetic modification has a profound influence on the biological
activity of the resulting compounds.2Recently, our laboratory
described a novel class of halogenating enzymes capable of carrying
out halogenations at aliphatic carbon centers of peptidyl carrier
protein-linked amino acid residues.3These enzymes, represented
by SyrB2 and CmaB, are nonheme FeIIhalogenases that require
oxygen, R-ketoglutarate (R-KG), and chloride for their activity. In
analogy to related FeII-R-KG-dependent hydroxylases,4it is pos-
tulated that the initial generation of a high-valent oxoiron species
(FeIVdO) is followed by hydrogen atom abstraction from an
aliphatic carbon center in the substrate and subsequent chlorina-
tion of the substrate radical.5Herein, we report that the trichlori-
nation of the C5 methyl group of L-leucine attached to a peptidyl
carrier protein BarA in the biosynthesis of barbamide is carried
out by tandem action of two nonheme FeIIhalogenases, BarB1 and
Barbamide (3, Scheme 1) is a potent molluscicidal agent isolated
from the marine cyanobacterium Lyngbya majuscula.6An intriguing
structural feature of this natural product is the presence of a
trichloromethyl moiety. Feeding experiments with isotopically
labeled precursors have established that the trichloromethyl group
is biosynthetically derived from the pro-R methyl group of
L-leucine.7Furthermore, high incorporation of labeled exogenous
[2-13C]-5,5,5-trichloroleucine implies a direct role for trichloroleu-
cine as an intermediate in the biosynthesis of barbamide.8Analysis
of the barbamide gene cluster revealed the presence of only two
candidate halogenase genes, barB1 and barB2, whose putative
products are predicted to be highly similar to syringomycin
halogenase SyrB2,3,9making triple halogenation of the unactivated
methyl group of leucine an interesting biosynthetic problem.
To study the trichlorination event, synthetic genes (DNA 2.0,
Menlo Park, CA) encoding for the peptidyl carrier protein BarA
(T domain), the two putative halogenases BarB1 and BarB2, and
the adenylation (A) domain BarD optimized for E. coli overex-
pression were obtained. All four proteins were overproduced in E.
coli and purified as N-His-tagged fusions. The 12.9 kDa BarA was
obtained in its apo form in a yield of 15 mg/L of culture, while
overexpression of the 60.3 kDa BarD yielded 4 mg/L. The apo
BarA (100 µM) was subsequently post-translationally modified with
the phosphopantetheine arm through in situ incubation with
coenzyme A (200 µM) and Sfp (2.5 µM),10providing the desired
holo protein. To minimize any adventitious deactivation of oxygen-
labile putative nonheme FeIIhalogenases, crude cell lysates
containing His-tagged BarB1 and BarB2 were kept under nitrogen
and purified in an anaerobic glovebox as apo proteins. Proteolytic
removal of BarB1 and BarB2 His tags was followed by further
purification by gel filtration chromatography and reconstitution with
FeIIin the presence of R-KG and chloride to provide the desired
holo BarB1 (yield 0.4 mg/L, iron content 115-130%) and holo
BarB2 (yield 1.2 mg/L, iron content 70%).
In trans loading of holo BarA with L-[14C]Leu (2 mM) was
accomplished by incubation with BarD (20 µM) and ATP (2 mM).
L-[14C]Leu-S-BarA (1, Scheme 1) was incubated with reconstituted
BarB1 (40 µM) and BarB2 (40 µM) either separately or in the
presence of both enzymes, together with R-KG (2 mM) and chloride
(∼15 mM), and incubated for 1 h at room temperature. The
modified amino acids were released by treatment of T domain-
tethered products with the type II thioesterase TycF11and further
converted to isoindole derivatives as previously described.3Products
were identified by radio-HPLC coelution with isoindole-derivatized
authentic standards of (2S,4S)-5-chloroleucine, (2S,4S)-5,5-dichlo-
roleucine, and (2S,4S)-5,5,5-trichloroleucine (Figure 1Aa).8,12
When L-[14C]Leu-S-BarA was presented to BarB1 in the presence
of O2, Cl-, and R-KG, only trace amounts of trichloroleucine and
no mono- or dichloroleucine were observed (Figure 1Ab), suggest-
ing that leucine-loaded BarA is a poor substrate for this halogenase.
In contrast, when halogenation of the L-[14C]Leu-S-BarA was
examined in the presence of BarB2, (2S,4S)-5,5-dichloroleucine was
formed as the major product, accompanied by a small amount of
(2S,4S)-5,5,5-trichloroleucine (Figure 1Ac). Interestingly, no (2S,4S)-
5-monochloroleucine accumulated in this experiment. When both
BarB1 and BarB2 were incubated with the leucine-loaded T domain,
both (2S,4S)-5,5-di- and (2S,4S)-5,5,5-trichloroleucine were detected
(Figure 1Ad). An increase in the intensity of the trichloroleucine
peak was noted when compared to incubation with BarB2 alone.
A further increase in the production of trichloroleucine was observed
when BarB1 was added to the L-[14C]Leu-S-BarA that had been
preincubated with BarB2 (Figure 1Ae). In the absence of its
preferred substrate, it is believed that BarB1 undergoes inactiva-
tion via either FeIIoxidation13aor radical-induced self-hydrox-
ylation,13b,ctwo common uncoupling pathways. We postulate that
these modes of deactivation are responsible for the decrease in the
amount of trichloroleucine product in the assay where both
halogenases are present from the very beginning of incubation
(Figure 1Ad) when compared to pregeneration of dichloroleucine-
BarA prior to the addition of BarB1 (Figure 1Ae). Collectively,
these data suggest that BarB2 is an efficient dihalogenating enzyme
working twice in trans on the L-Leu-S-BarA to yield (2S,4S)-5,5-
†Present address: Boehringer Ingelheim (Canada) Ltd., Research and
Development, Dept. of Biological Sciences, Laval, Qc, H7S 2G5, Canada.
Published on Web 03/08/2006
3900 9 J. AM. CHEM. SOC. 2006, 128, 3900-3901
10.1021/ja060151n CCC: $33.50 © 2006 American Chemical Society
dichloro-Leu-S-BarA. The inefficiency of BarB2 in catalyzing
trichlorination is overcome by the presence of the second haloge-
nase, BarB1, which converts dichloroleucine-loaded BarA to the
To further investigate the substrate specificity of each of the
halogenating enzymes, holo BarA was loaded with authentic
(2S,4S)-5-mono- and (2S,4S)-5,5-dichloroleucine in the presence
of BarD. The resulting aminoacyl-S-BarA derivatives were sepa-
rately incubated with each halogenase in the presence of O2, R-KG,
and Cl-and products analyzed as described previously (Figure 1B).
While BarB1-catalyzed chlorination of monochloroleucyl-S-BarA
provided (2S,4S)-5,5,5-trichloroleucine (Figure 1Bb and Supporting
Information), BarB2-catalyzed halogenation of this substrate led
to the formation of dichloroleucine as the major product (Figure
1Bc). When (2S,4S)-5,5-dichloro-Leu-S-BarA was subjected to
enzymatic halogenation with BarB1, formation of (2S,4S)-5,5,5-
trichlorinated product was observed (Figure 1Bd). In contrast,
dichloroleucine-loaded BarA is not a good substrate for BarB2
(Figure 1Be). Together, these data suggest that BarB2 not only
dichlorinates BarA-tethered L-leucine 1 but also converts monochloro-
Leu-S-BarA 4 to dichloroleucine derivative 5 (Scheme 2A).
Inefficient trichlorination by this enzyme is circumvented by the
presence of the second halogenase BarB1, able to convert both
mono- and dichloro derivatives 4 and 5 to (2S,4S)-5,5,5-trichloro-
Leu-S-BarA 2 (Scheme 2B).
In summary, in vitro reconstitution of the formation of the
trichloroleucine moiety has been achieved. It was demonstrated that
the triple chlorination of the unactivated methyl group of BarA-
tethered L-leucine substrate is mediated by the tandem action of
two nonheme FeIIhalogenases, BarB1 and BarB2, establishing
complementary roles of these two enzymes in the generation of
(2S,4S)-5,5,5-trichloroleucine in barbamide biosynthesis and setting
up the system for subsequent mechanistic analysis.
Acknowledgment. We thank Ms. Ellen Yeh for providing TycF
and helpful discussions. This work was supported in part by NIH
grants to C.T.W. D.P.G. is supported by the Damon Runyon Cancer
Research Foundation Fellowship (DRG-1893-05). F.H.V. is sup-
ported by a Merck-sponsored Fellowship of Helen Hay Whitney
Foundation and NSERC Postdoctoral Fellowship.
Supporting Information Available: Experimental details. This
material is available free of charge via the Internet at http://pubs.acs.org.
(1) Gribble, G. W. J. Chem. Educ. 2004, 81, 1441-1449.
(2) (a) Gribble, G. W. Acc. Chem. Res. 1998, 31, 141-152. (b) van Pe ´e,
K.-H.; Unversucht, S. Chemosphere 2003, 52, 299-312. (c) Renner, M.
K.; Jensen, P. R.; Fenical, W. J. Org. Chem. 1998, 63, 8346-8354.
(3) (a) Vaillancourt, F. H.; Yin, J.; Walsh, C. T. Proc. Natl. Acad. Sci. U.S.A.
2005, 102, 10111-10116. (b) Vaillancourt, F. H.; Yeh, E.; Vosburg, D.
A.; O’Connor, S. E.; Walsh, C. T. Nature 2005, 436, 1191-1194.
(4) (a) Pavel, E. G.; Zhou, J.; Busby, R. W.; Gunsior, M.; Townsend, C. A.;
Solomon, E. I. J. Am. Chem. Soc. 1998, 120, 743-753. (b) Hegg, E. L.;
Whiting, A. K.; Saari, R. E.; McCracken, J.; Hausinger, R. P.; Que, L.,
Jr. Biochemistry 1999, 38, 16714-16726. (c) Schofield, C. J.; Zhang, Z.
Curr. Opin. Struct. Biol. 1999, 9, 722-731. (d) Price, J. C.; Barr, E. W.;
Tirupati, B.; Bollinger, J. M., Jr.; Krebs, C. Biochemistry 2003, 42, 7497-
7508. (e) Hausinger, R. P. Crit. ReV. Biochem. Mol. Biol. 2004, 39, 21-
(5) Vaillancourt, F. H.; Yeh, E.; Vosburg, D. A.; Garneau-Tsodikova, S.;
Walsh, C. T. Chem. ReV. In press.
(6) Orjala, J.; Gerwick, W. H. J. Nat. Prod. 1996, 59, 427-430.
(7) Sitachitta, N.; Rossi, J.; Roberts, M. A.; Gerwick, W. H.; Fletcher, M.
D.; Willis, C. L. J. Am. Chem. Soc. 1998, 120, 7131-7132.
(8) Sitachitta, N.; Ma ´rquez, B. L.; Williamson, R. T.; Rossi, J.; Roberts, M.
A.; Gerwick, W. H.; Nguyen, V.-A.; Willis, C. L. Tetrahedron 2000, 56,
(9) Chang, Z.; Flatt, P.; Gerwick, W. H.; Nguyen, V.-A.; Willis, C. L.;
Sherman, D. H. Gene 2002, 296, 235-247.
(10) Quadri, L. E.; Weinreb, P. H.; Lei, M.; Nakano, M. M.; Zuber, P.; Walsh,
C. T. Biochemistry 1998, 37, 1585-1595.
(11) Yeh, E.; Kohli, R. M.; Bruner, S. D.; Walsh, C. T. ChemBioChem 2004,
(12) (a) Gerwick, W. H.; Leslie, P.; Long, G. C.; Marquez, B. L.; Willis, C.
L. Tetrahedron Lett. 2003, 44, 285-288. (b) Arda ´, A.; Jime ´nez, C.;
Rodrı ´guez, J. Tetrahedron Lett. 2004, 45, 3241-3243.
(13) (a) Vaillancourt, F. H.; Labbe, G.; Drouin, N. M.; Fortin, P. D.; Eltis, L.
D. J. Biol. Chem. 2002, 277, 2019-2027. (b) Liu, A.; Ho, R. Y. N.; Que,
L., Jr.; Ryle, M. J.; Phinney, B. S.; Hausinger, R. P. J. Am. Chem. Soc.
2001, 123, 5126-5127. (c) Liu, P.; Mehn, M. P.; Yan, F.; Zhao, Z.; Que,
L., Jr.; Liu, H.-w. J. Am. Chem. Soc. 2004, 126, 10306-10312.
Figure 1. Analysis of the reactions catalyzed by BarB1 and BarB2. (A)
UV 338 nm traces of standards (a);14C radio-HPLC traces of products
obtained in L-[14C]Leu-S-BarA incubation with BarB1 (b), BarB2 (c), BarB1
and BarB2 (d), BarB2, followed by BarB1 (e). The 0.4 min shift between
retention times of trace a (UV) and b-e (radioactivity) is caused by the
positioning of the two detectors. (B) UV 338 nm traces of standards (a),
products of (2S,4S)-5-chloro-Leu-S-BarA incubation with BarB1 (b) and
BarB2 (c), and products of (2S,4S)-5,5-dichloro-Leu-S-BarA incubation with
BarB1 (d) and BarB2 (e).
C O M M U N I C A T I O N S
J. AM. CHEM. SOC. 9 VOL. 128, NO. 12, 2006 3901