Halomethane biosynthesis: Structure of a SAM-dependent halide methyltransferase from arabidopsis thaliana

Article abstract of DOI:10.1002/anie.201000119

It's a gas ! The structure of the halomethane-producing halo/thiocyanate methyltransferase enzyme from plants has been determined. The halide ion and the methyl group of S-adenosyl-L-methionine (SAM) were modeled into the active site (see picture; chloride: green sphere; SAM: C green, O red, S yellow, N blue), which indicated their predisposition for reaction. (Figure Presented)

Full text of DOI:10.1002/anie.201000119

Communications
DOI: 10.1002/anie.201000119
Structural Biology
Halomethane Biosynthesis: Structure of a SAM-Dependent Halide
Methyltransferase from Arabidopsis thaliana**
Jason W. Schmidberger, Agata B. James, Robert Edwards, James H. Naismith, and
David OꢀHagan*
Of the gaseous halomethanes (MeCl, MeBr, MeI) found in
the atmosphere, chloromethane is the major contributor.
About 4 ꢀ 106 tonnes of chloromethane are estimated to be
produced globally per year, which is contributed to by
terrestrial plants (and fungi) and by photosynthetic micro-
organisms, algae, and marine plants.[1] Chloromethane is the
most significant naturally produced volatile chlorocarbon,
contributing up to 15% of stratospheric chlorine.[2] Bromo-
Scheme 1. The halomethane gases are generated in a nucleophilic
methane is significant too, but less abundant, and is estimated
to be produced at about 1.8 ꢀ 105 tonnes per year[1] and
contributing up to 55% of stratospheric bromine.[2] Iodo-
methane is generated at a higher level than bromomethane at
about 8 ꢀ 105 tonnes per year,[1] but it appears to be the least
significant with respect to atmospheric chemistry because of
photolysis, resulting in a low stability and a short half-life.[3]
Higher plants are estimated to account for 30–50% of the
global production of chloromethane,[4] with its biogenesis
receiving considerable attention owing to its role in ozone
depletion. The genes responsible for halomethane biosynthe-
sis in plants have been named the HOL (harmless to ozone
layer) genes,[5,6] because of inactivation of the associated
biosynthetic pathway when they are deleted. The halome-
thane gases almost certainly play some regulatory role too
within the producing organisms, as methyl transfer vehicles,
although their metabolic role is unclear.[7–9]
In plants, fungi, and bacteria, the enzyme product of the
HOL gene is known to be an S-adenosyl-l-methionine
(SAM)-dependant methyltransferase, which combines
halide ion and SAM in a nucleophilic substitution reaction
to generate halomethane.[10] Although this enzyme is respon-
sible for halomethane production (Scheme 1), it was recently
shown to have a preference for thiocyanate as a nucleo-
phile.[11] The promiscuity shown by the enzyme for the halides
(excluding fluoride) and thiocyanate renders its physiological
role unclear, although its particular ability to methylate
reaction between halide ion and SAM to generate S-adenosyl-l-
homocysteine (SAH). The reaction is catalyzed by the action of halide
methyltransferases.
thiocyanate implies an intracellular role in glucosinolate
metabolism.[12]
Over the last decade, there has been a substantial
development in our understanding of biohalogenation and
particularly enzymatic chlorination, and an intriguing range
of different halogenation enzymes have been identified that
À
are responsible for generating the C Cl bond in secondary
metabolites.[13–15] The earliest characterized are the haloper-
oxidases, found typically in marine plants, which oxidize
chloride, bromide, and iodide ions at the expense of hydrogen
peroxide to generate X+ for reactions with aromatics and
other unsaturated organics. These haloperoxidases fall into
two categories, and are either vanadium- or heme-iron-
dependent.[16] More recently, FAD-dependent chlorination
enzymes have been characterized, which also mediate elec-
trophilic reactions such as the chlorination of tryptophan.[17,18]
These speocies are typically found in microbes rather than
plants. Walsh has recently identified a class of iron–sulfur
halogenation enzymes responsible for the selective chlorina-
tion of unactivated carbon atoms (for example, methyl
groups).[19,20] These enzymes essentially generate chlorine
radicals, and they are particularly relevant to the aliphatic
chlorination of bacterial secondary metabolites in both
terrestrial and marine environments. Nucleophilic chlorina-
tion beyond chloromethane production is rare, although
Moore reported in 2008[21] the identification of the chlorinase,
an enzyme responsible for the production of 5-chloro-5’-
deoxyadenosine (ClDA), the first-formed intermediate in the
biosynthesis of salinosporamide-A, a metabolite of the deep-
sea soil bacterium Salinospora tropica. This enzyme is chemi-
cally similar and almost structurally identical to the fluori-
nase, an enzyme from Streptomyces cattleya that utilizes both
fluoride and to a lesser extent chloride ion in a similar
nucleophilic attack to C5’ carbon of SAM.[22–24] X-ray
structural data is in place for the nucleophilic fluorinase[23]
and chlorinase[21] enzymes, as well as representatives of
almost all of the other known halogenases.[15] Therefore, the
[*] Dr. J. W. Schmidberger, Prof. Dr. J. H. Naismith, Prof. Dr. D. O’Hagan
University of St Andrews
Department of Chemistry, Centre for Biomolecular Sciences
North Haugh, St Andrews, Fife, KY16 9ST (UK)
Fax: (+44)1334 463-808
E-mail: do1@st-andrews.ac.uk
A. B. James, Prof. Dr. R. Edwards
University of Durham, School of Biological & Medical Sciences
South Road, Durham, DH1 3LE (UK)
[**] We thank the BBSRC for a grant (BB/F007426/1) and J.W.S. thanks
the Wellcome Trust for the Value in People (VIP) Scheme for
financial support. SAM=S-adenosyl-l-methionine.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201000119.
3646
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 3646 –3648

Angewandte
Chemie
plant halo/thiocyanate methyltransferases remain the last
major group of halogenase enzymes that have not been
structurally characterized. Given the importance of gaseous
halomethane production to atmospheric chemistry,[1–3] and
the developing recognition of the role of these enzymes in
thiocyanate metabolism in plants,[11] we present herein the
first structural study on this enzyme group.
Plant methyltransferases that are active towards halide/
thiocyanate nucleophiles have been isolated and studied from
Batis maritima,[5] Brassica oleracea,[25] and Arabidopsis thali-
ana.[5,6] Collectively, they have been termed halide methyl-
transferases (HMT), or halide/thiocyanate methyltransfer-
ases (HTMT), on the basis of their activity with halides alone,
or additionally with thiol substrates, such as bisulfide or
thiocyanate.[11] A phylogenetic analysis using the A. thaliana
structural gene (AtHOL1) suggests a wide distribution of
these enzymes amongst the plant kingdom, with two further
family members encoded within the A. thaliana genome
(AtHOL2 and AtHOL3).[12]
Figure 1. Stylized ribbon diagram of the structure of the Aribidopsis
thaliana halomethyl transferase (AtHTMT1), solved crystallographically
to a resolution of 1.8 ꢀ and colored in a spectrum from N-terminus
(blue) to C-terminus (red). S-Adenosyl-l-homocysteine (SAH) is bound
to the putative active site of the enzyme.
Herein the coding sequence of AtHOL1 (gene
At2g43910; accession AY044314), now termed AtHTMT,
was isolated by PCR[26] and the gene cloned and protein over-
expressed in Rosetta II (DE3) cells with a C-terminal 6-His
tag for ease of purification. For assay work (to remove the
influence of the His tag), the AtHTMT1 gene was cloned into
a pEHISTEV vector[27] to give an N-terminal poly His tag
with a TEV protease cleavage site. The two mutants (V23C
and Y172F) were also constructed and cloned into the same
vector. Purification of the over-expressed proteins then
included a TEV digestion. The purified enzyme was concen-
trated (10 mgmLÀ1), and was co-crystallized with S-adenosyl-
l-homocysteine (SAH). The structure was solved to a
resolution of 1.8 ꢁ using the related (20% amino acid
sequence identity) mouse thiopurine methyltransferase
(PDBId: 2GB4) structure as a molecular replacement
model. The low identity necessitated considerable model
manipulation. The SAH ligand locates the active site of the
enzyme, as illustrated in Figure 1, and the active site is shown
more closely in Figure 2. The methyl group was modeled at
the sulfonium stereogenic center to represent SAM (Fig-
ure 2b). It is well-established that SAM synthase only
generates the (S)-SAM configuration at sulfur, and all
enzymes appear to utilize this isomer of SAM.[28] The
trajectory of the modeled methyl group projects into the
active site. Furthermore, the Trp47 residue would prevent the
diastereomeric (R)-SAM from binding, as there is an obvious
clash with the methyl group. The modeled methyl group
shown in Figure 2b occupies a rather open active site, which is
consistent with the promiscuity of the enzyme, particularly for
large nucleophiles.
genesis (Tyr172Phe) of the tyrosine residue to phenylalanine
led to a functional enzyme, but with a reduced efficiency with
chloride ions (Vmax drops from 2.43 to 0.92 nmolminÀ1 mgÀ1
protein), but otherwise a similar efficiency for bromide and
thiocyanate (Table 1). This analysis suggests that the effi-
ciency of the larger nucleophiles, bromide and thiocyanate, is
not particularly compromised, and perhaps ordered hydrogen
bonding, from tyrosine 172, through a bridging water
molecule is important for orientating the smaller chloride
nucleophile.
The amino acid sequence homology is generally high
between HMT/HTMT proteins, except for the first 30 or so
residues of the N-terminus (blue helices in Figure 1). This
region creates a cap over the active site, forming key contacts
to the nucleophile during the reaction. When the residues
lining the active site of AtHTMT1 are compared (through a
sequence alignment) to those of the HTMT from Batis
maritima (BmHTMT),[8] which features a much greater
activity for ClÀ, Val-23 (of AtHTMT1) is the only active
site residue that is not conserved. In BmHTMT, the equiv-
alent residue is a cysteine. Accordingly this valine residue of
AtHTMT1 was mutated to cysteine. Despite removing a
hydrophobic active site residue in close contact with the
nucleophile, the resultant V23C mutant remained functional,
displaying a slightly improved activity for all of the nucleo-
philes explored, including chloride ions (Table 1). However, a
lowered stability of this mutant was noticeable by decom-
position on SDS-PAGE.
Three crystallographically identifiable water molecules
occupy the cavity in the SAH–enzyme co-crystal. One water
molecule (W198) will be displaced by the methyl group of
SAM and the central water molecule (W68) occupies the
predicted location of the halide nucleophile. The remaining
water molecule (W35) is hydrogen bonded to the side chain of
Tyr172. Therefore, a model emerges in which this water
moelcule is bridging, which helps to orient the nucleophile
(bromide or chloride ion; Figure 2b). Site-directed muta-
In conclusion, the structure of a plant halomethane-
producing enzyme is presented and a model for substrate/
nucelophile binding and reaction at the active site rational-
ized. The Arabidopsis thaliana enzyme presents the reactive
sulfonium methyl group into a large cavity that accounts for
its promiscuous nature with respect to a variety of nucleo-
philes. The enzyme promotes reaction most efficiently with
Angew. Chem. Int. Ed. 2010, 49, 3646 –3648
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3647

Communications
Figure 2. a) The Aribidopsis thaliana halomethyl transferase active site showing the amino acid side groups forming the putative nucleophile-
binding site (shown). Three water molecules (W35, W68, and W198) occupy a channel from the bulk solvent to a proposed nucleophilic binding
site occupied by W198. b) The missing methyl group of (S)-SAM is modeled into the structure with space-filling chloride (green sphere, 1.75 ꢀ
radius) and bromide ions (larger pale red sphere, eclipsed by green sphere, 1.85 ꢀ) to simulate preorganization for reaction.
[7] X. H. Ni, L. P. Hager, Proc. Natl. Acad. Sci. USA 1998, 95,
Table 1: Enzyme kinetic data for AtHTMT variants for the three
nucleophiles NCSÀ, BrÀ, and ClÀ.
12866 – 12871.
[8] X. H. Ni, L. P. Hager, Proc. Natl. Acad. Sci. USA 1999, 96, 3611 –
3615.
AtHTMT variant
Nucleophile
Km [mm]
Vmax [nmolminÀ1
mgproteinÀ1
]
[9] R. C. Rhew, B. R. Miller, R. F. Weiss, Nature 2000, 403, 292 – 295.
[10] T. S. Bayer, D. M. Widmaier, K. Temme, E. A. Mirsky, D. V.
Santi, C. A. Voigt, J. Am. Chem. Soc. 2009, 131, 6508 – 6515.
[11] N. Itoh, H. Toda, M. Matsuda, T. Negishi, T. Taniguchi, N.
Ohsawa, BMC Plant Biol. 2009, 9, 116.
[12] Y. Nagatoshi, T. Nakamura, J. Biol. Chem. 2009, 284, 19301 –
19309.
[13] C. D. Murphy, Nat. Prod. Rep. 2006, 23, 147 – 152.
[14] C. S. Neumann, D. C. Fujimori, C. T. Walsh, Chem. Biol. 2008,
15, 99 – 109.
[15] L. C. Blasiak, C. L. Drennan, Acc. Chem. Res. 2009, 42, 147 – 155.
[16] J. Littlechild, Curr. Opin. Chem. Biol. 1999, 3, 28 – 34.
[17] C. Dong, S. Flecks, S. Unversucht, C. Haupt, K. H. van Pee, J. H.
Naismith, Science 2005, 309, 2216 – 2219.
[18] E. Yeh, L. C. Blasiak, A. Koglin, C. L. Drennan, C. T. Walsh,
Biochemistry 2007, 46, 1284 – 1292.
[19] F. H. Vaillancourt, J. Yin, C. T. Walsh, Proc. Natl. Acad. Sci. USA
2005, 102, 10111 – 10116.
[20] D. P. Galonic, E. W. Barr, C. T. Walsh, J. M. Bollinger, C. Krebs,
Nat. Chem. Biol. 2007, 3, 113 – 116.
Native
V23C
Y172F
Native
V23C
Y172F
Native
V23C
NCSÀ
NCSÀ
NCSÀ
BrÀ
0.099Æ0.020
0.103Æ0.009
0.141Æ0.009
24.87Æ2.785
21.06Æ1.883
30.16Æ2.942
145.2Æ26.56
122.0Æ25.31
137.2Æ20.60
43.6Æ2.52
43.4Æ0.93
46.0Æ1.11
11.4Æ0.31
12.4Æ0.25
11.4Æ0.33
2.43Æ0.12
2.78Æ0.16
0.91Æ0.04
BrÀ
BrÀ
ClÀ
ClÀ
Y172F
ClÀ
thiocyanate and then the halides, with an efficiency order of
NCSÀ > IÀ > BrÀ > ClÀ (but not FÀ). The dual role of this
enzyme class is intriguing and there remains much to learn
regarding the relationship between halomethane biosynthesis
and thiocyanate/methyl thiocyanate metabolism within a
single organism.
[21] A. Eustꢂquio, F. Pojer, J. P. Noel, B. S. Moore, Nat. Chem. Biol.
2008, 4, 69 – 74.
[22] D. OꢃHagan, C. Schaffrath, S. L. Cobb, J. T. G. Murphy, C. D.
Murphy, Nature 2002, 416, 279.
Received: January 8, 2010
Revised: February 22, 2010
Published online: April 7, 2010
[23] C. Dong, F. L. Huang, H. Deng, C. Schaffrath, J. B. Spencer, D.
OꢃHagan, J. H. Naismith, Nature 2004, 427, 561 – 565.
[24] H. Deng, S. L. Cobb, A. R. McEwan, R. P. McGlinchey, J. H.
Naismith, D. D. OꢃHagan, D. A. Robinson, J. B. Spencer, Angew.
Chem. 2006, 118, 773 – 776; Angew. Chem. Int. Ed. 2006, 45, 759 –
762.
[25] J. M. Attieh, A. D. Hanson. H. S. Saini, J. Biol. Chem. 1995, 270,
9250 – 9257.
[26] D. P. Dixon, B. G. Davis, R. Edwards, J. Biol. Chem. 2002, 277,
30859 – 30869.
Keywords: atmospheric organohalogens · biohalogenation ·
chloromethane · methyltransferases · structural biology
.
[1] D. B. Harper, Nat. Prod. Rep. 2000, 17, 337 – 348.
[2] M. L. Cox, P. J. Fraser, G. A. Sturrock, S. T. Siems, L. W. Porter,
Atmos. Environ. 2004, 38, 3839 – 3852.
[3] M. L. Cox, G. A. Sturrock, P. J. Fraser, S. T. Siems, P. B.
Krummell, J. Atmos. Chem. 2005, 50, 59 – 77.
[4] T. Saito, Y. Yokouchi, Y. Yosugi, M. Tani, E. Philip, T. Okuda,
Geophys. Res. Lett. 2008, 35, L19812.
[5] Y. Nagatoshi, T. Nakamura, Plant Biotechnol. 2007, 24, 503 – 506.
[6] R. C. Rhew, L. Ostergaard, E. S. Saltzman, M. F. Yanofsky, Curr.
Biol. 2003, 13, 1809 – 1813.
[27] H. Liu, J. H. Naismith, Protein Expression Purif. 2009, 63, 102 –
111.
[28] J. Komoto, T. Yamada, Y. Takata, G. D. Markham, F. Takusa-
gawa, Biochemistry 2004, 43, 1821 – 1831.
3648 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 3646 –3648