S-adenosyl-L-methionine:Hydroxide adenosyltransferase: A SAM enzyme

Article abstract of DOI:10.1002/anie.200800794

(Chemical Equation Presented) Not so DUF: A DUF62 enzyme from the archaeon Pyrococcus horikoshii OT3 converts S-adenosyl-L-methionine (SAM) into adenosine through the nucleophilic attack of a hydroxide ion derived from water (see picture of the active site). The highly conserved nature of Asp68, Arg75, and His127 throughout the DUF62 protein superfamily suggests the wide-spread distribution of this novel catalytic activity in microorganisms. DUF = domain of unknown function.

Full text of DOI:10.1002/anie.200800794

Angewandte
Chemie
DOI: 10.1002/anie.200800794
Enzyme Catalysis
S-Adenosyl-l-methionine:Hydroxide Adenosyltransferase: A SAM
Enzyme**
Hai Deng, Catherine H. Botting, John T. G. Hamilton, Rupert J. M. Russell, and
David OꢀHagan*
Dedicated to the memoryof Jonathan B. Spencer
S-Adenosyl-l-methionine (SAM, 1) has a variety of roles in
enzymology. It is most commonly involved in methyl-transfer
reactions,[1] in which it transfers its methyl group to O, N, S,
and C atoms of various substrates. SAM also donates a
methylene group in the cyclopropanation of unsaturated fatty
acids.[2] There is a SAM decarboxylase,[3] which initiates
polyamine biosynthesis, and, perhaps most exotically, SAM is
the source of 5’-deoxyadenosyl radicals in at least three iron–
sulfur enzymes.[4] SAM acts as a precursor in the biosynthesis
of several metabolites, including ethylene in plants,[5] biotin,
and epoxyqueuosine.[6] Fluorination[7] and chlorination[8]
enzymes have been reported that utilize SAM as a substrate;
in each case, the halide ion mediates a nucleophilic attack at
the 5’-position of SAM (1) to generate the corresponding
5’-halo-5’-deoxyadenosine (5’XDA) product 2 or 3 and
l-methionine (l-Met; Scheme 1). These reactions, with
SAM-dependent methyl transferases,[1] are among the few
enzymatic SN2-type reactions known.[9,10]
Examination of the fluorinase and chlorinase amino acid
sequences with data derived from genome sequencing reveals
an identity (< 36%) to the DUF62 superfamily (Figure 1;
DUF = domain of unknown function).[11] No function has
been assigned to the DUF62 proteins. The distribution of
these genes is restricted; in general they are found only in
extremophile and pathogen-related microorganisms. The
protein products of four different duf62 genes from four
different extremophiles have been the subject of overexpres-
sion and X-ray crystal-structure evaluation in structural-
proteomics screening programs.[12]
The four DUF62 protein structures[12] are nearly identical
with each other and almost superimposable on the previously
reported fluorinase[10] and chlorinase[8] structures. One of
these enzymes is from P. horikoshii OT3, a microorganism
with an optimum growth temperature of 988C, which was
isolated at a depth of 1395 m from an Okinawa trough vent in
the Japanese Pacific Ocean.[13] Examination of the structure
reveals an adenosine molecule coordinated at an intersubunit
position corresponding to the fluorinase active site (Figure 2).
The fluorinase and chlorinase also cocrystallize with adeno-
sine, and the adenosine molecule in the re-refined P. hori-
koshii structure is almost superimposable on that in the
structures of the fluorinase and the chlorinase. For example,
an aspartate carboxylate (Asp7) anchors the 2’- and 3’-OH
groups of the adenosine ribose moiety. With this background,
we recloned the DUF62 gene of P. horikoshii OT3 into E. coli
and overexpressed and purified the protein.[14] The protein
was unable to mediate a fluorination or chlorination reaction
with SAM after incubation at high concentrations (> 10 mm)
of halide ions.[15] However, a novel activity was apparent in
that the enzyme could catalyze the conversion of SAM into
adenosine (4) by the attack of a hydroxide ion (from water) at
C5’ of SAM (1), in analogy with nucleophilic halide reactions
(Scheme 2).
Scheme 1. Nucleophilic substitution of SAM with halide ions.[7,8]
[*] Dr. H. Deng, Dr. C. H. Botting, Dr. R. J. M. Russell,
Prof. Dr. D. O’Hagan
Centre for Biomolecular Sciences
University of St Andrews
St Andrews KY16 9ST (UK)
Fax: (+44)1334-463-808
E-mail: do1@st-and.ac.uk
Dr. J. T. G. Hamilton
The enzyme was assayed[16] at 378C and found to have
optimal activity at pH 8.5, and to be inactive below pH 5 and
above pH 10 (pH 8.5, Km(SAM) = 39.2 mm, kcat = 0.14 sÀ1).
Consistent with a hyperthermophilic protein, the enzyme
retains full activity after being heated at 808C for 30 min and
then cooled to 378C. An assay of the P. horikoshii SAM
hydroxide adenosyltransferase in a buffer enriched with
H218O resulted in isotopically enriched [5’-18O]-adenosine
([M+2]+: 48%), as determined by GC–MS analysis after
pertrimethylsilylation at the three hydroxy groups and the
Food Chemistry Branch
Agriculture, Food and Environmental Science Division
Agri-Food Biosciences Institute
Newforge Lane, Belfast BT9 5PX (UK)
[**] This research is supported by the Biotechnological and Biological
Science Research Council. We thank Dr. K. Fujita from NBRC (NITE-
DOB, Japan) for providing the clone containing ORF PH0463. We
thank the BBSRC for a research grant and the Wellcome Trust for
mass-spectrometry support. SAM=S-adenosyl-l-methionine.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2008, 47, 5357 –5361
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5357

Communications
Figure 1. Amino acid sequence alignment and identity (%) between the fluorinase from S. cattleya and DUF62 proteins from various organisms.
I) Fluorinase from Streptomyces cattleya;[7] II) chlorinase from Salinispora tropica;[8] III) Acidobacteria bacterium; IV) Salinibacter ruber; V) Prostheco-
chloris aestuarii; VI) Thermotoga maritima; VII) Listeria monocytogenes EGD-e; VIII) Chlorobium ferrooxidans; IX) Staphylothermus marinus F1;
X) Pyrococcus horikoshii OT3; XI) Escherichia coli O157:H7; XII) Staphylococcus aureus subsp. aureus. Star: conserved amino acids in DUF62 proteins
(Asp68, Arg75, and His127).
Figure 2. Adenosine bound to SAM hydroxide adenosyltransferase
from P. horikoshii.
Scheme 3. Two putative mechanisms for the P. horikoshii SAM hydrox-
ide adenosyltransferase reaction with SAM.
in the catalytic cycle. This ester is then hydrolyzed to release
an alcohol and recover the carboxylate group.[18] Putative
mechanism B in Scheme 3 presents an analogous possibility
for SAM hydroxide adenosyltransferase. To distinguish
mechanisms A and B and probe the role of Asp68 in this
reaction, we examined isotopic labeling of the protein.
Assays were carried out with H218O (64 atom%). Mech-
anism A would label the product adenosine with 18O at C5’,
but not the enzyme. Mechanism B would load the Asp
carboxylate group with 18O, and multiple turnover would
result in both enzyme and product labeling (at C5’). To probe
isotope incorporation into the protein, subsequent trypsin
digestion of the DUF62 protein in heavy-isotope-free buffer
was followed by analysis of a fragment containing the Asp68
residue by mass spectrometry.[17] MS/MS analysis did not
provide any evidence for heavy-isotope incorporation into the
peptide fragment containing Asp68. In a control experiment,
the enzyme was digested with trypsin in H218O (95 atom%)
buffer (in-gel digestion),[19] and the fragment containing the
Scheme 2. The SAM hydroxide adenosyltransferase reaction.
amino group.[17] This result suggests that the general mech-
anism A in Scheme 3 operates. However, an examination of
the active-site residues of the protein from its X-ray crystal
structure (Figure 2) reveals an aspartate carboxylate residue,
Asp68, in relatively close contact to the ribose ring of
adenosine (bound product). Asp and Glu carboxylate groups
are known to be involved as nucleophiles in haloacid
dehalogenase reactions to generate a substrate-bound ester
5358
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 5357 –5361

Angewandte
Chemie
Asp68 residue was analyzed by MS/MS mass spectrometry. In the fluoride-ion-desolvation process of the fluorinase, to force
this case, heavy-isotope incorporation was very clear: [M+2]+ two water molecules into the active-site cavity.[15] The inner
and [M+4]+ ions were observed, which is consistent with the water molecule is hydrogen-bonded to the Asp68 carboxylate
trypsin-mediated incorporation of 18O into the newly released group. We propose that the outer water molecule is made
terminal carboxylate residue. The MS method is clearly nucleophilic by coordination to both a backbone amide
capable of detecting 18O incorporation into the protein; carbonyl group and the inner water molecule (present in the
however, in the SAM hydroxide adenosyltransferase assays, product crystal structure). On SAM binding, the outer water
no incorporation of 18O was observed. These experiments molecule is also forced close to the electrophilic carbon atom
suggest that the enzyme operates by general mechanism A of SAM. Also, the basicity of the Asp68 carboxylate group
(Scheme 3).
will increase upon the weakening of the Asp68/Arg75 ion
Examination of the active-site residues of the enzyme in pair. Such a weakening will be promoted by hydrogen
the X-ray crystal structure reveals an intriguing juxtaposition bonding of the guanidinium residue to the imidazole residue
of substituents. Asp68 forms an ion-pair dimer with the of His127. The subtlety of this hydrogen-bonding arrange-
guanidinium residue of Arg75. The guanidinium residue is ment in activating a water molecule for nucleophilic attack in
H bonded to the imidazole unit of His127. These three amino the enzyme remains to be evaluated; however, the weakening
acid residues are absolutely conserved in all of the putative of very similar ion-pair interactions in this manner has been
hydroxide adenosyltransferases ( ꢀ 200) identified in the focus of model reactions in supramolecular chemistry.[21,22]
BLAST[20] searches of the DUF62 enzymes (see the Support- Scheme 4 illustrates the putative interplay of the active-site
ing Information), ten of which are shown in Figure 1. residues for catalytic turnover.
Interestingly, there are no equivalent residues in the fluo-
The enzyme was assayed with the adenosine analogues 2,
rinase or chlorinase enzymes, and the arrangement of the 3, and 5 in place of SAM to explore its ability to catalyze the
active site for nucleophilic attack in this enzyme bears no displacement of halides as leaving groups. Intriguingly,
obvious resemblance to those of the halide transferases. It is 5’-ClDA (3)[23] served as a substrate; however, adenosine
anticipated that substantial active-site desolvation occurs as a was generated with much lower efficiency than from SAM (1;
consequence of efficient SAM binding, in a similar manner to Km(3) = 410 mm,
k
cat = 0.31 minÀ1). Thus, chlorine could
Scheme 4. Topographical images of the active-site residues of SAM hydroxide adenosyltransferase: A putative reaction cycle.
Angew. Chem. Int. Ed. 2008, 47, 5357 –5361 ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5359

Communications
[3] M. S. Cohn, C. W. Tabor, H. Tabor, J. Biol. Chem. 1977, 252,
8212 – 8216.
[4] M. Fontecave, M. Atta, E. Mulliez, Trends Biochem. Res. 2004,
29, 243 – 249.
[5] A. J. Hamilton, M. Bouzayen, D. Grierson, Proc. Natl. Acad. Sci.
USA 1991, 88, 7434 – 7437.
[6] S. G. Van Lanen, S. D. Kinzie, S. Matthieu, T. Link, J. Culp, D.
Iwata-Reuyl, J. Biol. Chem. 2003, 278, 10491 – 10499.
[7] D. OꢀHagan, C. Schaffrath, S. L. Cobb, J. T. G. Hamilton, C. D.
Murphy, Nature 2002, 416, 279.
[8] A. Eustµquio, F. Pojer, J. P. Noel, B. S. Moore, Nat. Chem. Biol.
2008, 4, 69 – 74.
[9] H. M. Senn, D. OꢀHagan, W. Thiel, J. Am. Chem. Soc. 2005, 127,
13643 – 13655; C. D. Cadicamo, J. Courtieu, H. Deng, A.
Meddour, D. OꢀHagan, ChemBioChem 2004, 5, 685 – 690.
[10] C. Dong, F. L. Huang, H. Deng, C. Schaffrath, J. B. Spencer, D.
OꢀHagan, J. H. Naismith, Nature 2004, 427, 561 – 565.
[11] Wellcome Trust Sanger Institute, Cambridge, UK; accession
number: PF01887.
[12] There is no literature on these structures; however, the DUF62
X-ray crystal-structure coordinates have been deposited rou-
tinely in the Protein Data Bank. The four structures have the
following PDB codes: Methanococcus Jannaschii DSM 2661,
PDB 2F4N; Pyrococcus horikoshii OT3, PDB 1WU8; Thermus
Thermophilus Hb8, PDB 2CW5; Thermotoga maritima, PDB
2ZBU.
[13] J. M. Gonzalez, Y. Masuchi, F. T. Robb, J. W. Ammerman, D. L.
Maeder, M. Yanagibayashi, J. Tamaoka, C. Kato, Extremophiles
1998, 2, 123 – 130.
[14] The clone containing open reading frame (ORF) PH0463 from
the marine archaeon P. horikoshii OT3 was purchased from
NBRC (NITE-DOB, Japan; NBRC No.: NBRC G01-000-971,
clone ID: S2_2219, EMBL accession No.: BAA29549). The ORF
PH0463 was subcloned into pET28a(+) vector (Novagen) by
using the primers PH0463-F (GCAGGAGGAATTCATATGA-
replace the methionyl moiety as a leaving group. No new
products were detected by HPLC or NMR spectroscopy after
the incubation of the enzyme with SAM and FÀ, ClÀ, or NH4
+
ions at high mm concentrations, although in all cases
adenosine (4) was generated in a manner consistent with
normal enzymatic turnover, with water providing the nucle-
ophile. S-Adenosyl-l-homocysteine (SAH, 6), the neutral,
demethylated form of SAM, is often a potent inhibitor of
SAM-dependent enzyme reactions[24] and also emerged as an
efficient inhibitor of this enzyme (Ki(6) = 3.6 mm).
In conclusion, we have identified a SAM enzyme that
generates adenosine and l-methionine in a reaction with
water. SAM is more commonly converted into adenosine,
after demethylation to SAH (6), by the action of SAH
lyase.[25] This latter enzyme does not catalyze a direct
nucleophilic reaction but utilizes NAD+ in an oxidation–
elimination–reduction process. Thus, there is no mechanistic
analogy between SAM hydroxide adenosyltransferase and
SAH lyase. The metabolic role of SAM hydroxide adenosyl-
transferase is not clear. The enzyme takes a high-energy
compound (SAM) and generates lower-energy metabolites
(adenosine and l-methionine), which are available from
other metabolic sources. A similar observation was made with
NAD glycohydrolase (NADase), the role of which is also not
clear, although recent studies are revealing the sophisticated
regulation of its activity in bacteria.[26] Intriguingly, SAM
hydroxide adenosyltransferase generates a proton with each
turnover, and the enzyme becomes deactivated at pH 5. The
enzyme is most active at pH 8.5, which perhaps suggests a role
in regulating intracellular pH values in these extremophiles.
To date, the duf62 gene has been found in up to 200
microorganisms, most of which are extremophiles. Despite its
very close (superimposable) structural similarity to the
fluorinase and chlorinase enzymes, the DUF62 enzyme
activates the nucleophile in a very different manner.[9] SAM
hydroxide adenosyltransferase appears to have a novel mode
of action in this respect.
TAACCTTAACCACCGACTTCGG)
and
PH0463-R
(CGCCGCTCGAGTCATAGCAACCTCACCCTTAGCTC)
with restriction sites NdeI and XhoI, respectively. The resultant
plasmid pET-PH0463 was transferred into competent E. coli
BL21 C43(DE3) cells and grown in Luria broth containing
kanamycin (100 mgmLÀ1) at 378C until an absorbance of 0.6 at
l = 600 nm was reached. Overexpression was induced by adding
isopropylthiogalactoside (IPTG) to 1 mm and continuing the
incubation at 168C for 24 h. Cells were collected and lysed by
sonication, and the apoenzyme was purified to homogeneity as
described previously.[15]
[15] X. F. Zhu, D. A. Robinson, A. R. McEwan, D. OꢀHagan, J. H.
Naismith, J. Am. Chem. Soc. 2007, 129, 14597 – 14604.
[16] The HPLC assay was adapted from that used for the fluori-
nase.[15] Briefly, the apoenzyme (0.02 mgmLÀ1) was incubated at
378C with SAM and Tris-HCl buffer (pH 8.5; Tris = 2-amino-2-
hydroxymethylpropane-1,3-diol). The sample was quenched by
adding cold ethanol (500 mL) and then freeze-dried. It was then
added to water (100 mL) and subjected to centrifugation to
remove precipitated protein. The supernatant was analyzed by
HPLC to monitor adenosine production (Varian 9012 UV/Vis
detector at 260 nm).
Received: February 18, 2008
Published online: June 13, 2008
[17] See the Supporting Information for GC–MS, LC–ESI–MSMS,
and MALDI–MSMS experimental details.
Keywords: archaea · enzyme catalysis · fluorinase ·
nucleophilic substitution · SAM enzymes
[18] V. Nardi-Dei, T. Kurihara, C. Park, M. Miyagi, S. Tsunasawa, K.
Soda, N. Esaki, J. Biol. Chem. 1999, 274, 20977 – 20981.
[19] I. I. Stewart, T. Thomson, D. Figeys, Rapid. Commun Mass.
Spectrom. 2001, 15, 2456 – 2465; S. Shevchenko, M. Wilm, O.
Vorm, M. Mann, Anal. Chem. 1996, 68, 850 – 858.
.
[1] W. A. M. Loenen, Biochem. Soc. Trans. 2006, 34, 330 – 333.
[2] D. W. Gronan, J. E. Cronan, Jr., Microbiol. Mol. Biol. Rev. 1997,
61, 429 – 441.
[20] BLAST= Basic Local Alignment Search Tool; NIH, US
National Center for Biotechnology Information (NCBI),
Bethesda, MD 20894.
5360
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 5357 –5361

Angewandte
Chemie
[21] C. Schmuck, Coord. Chem. Rev. 2006, 250, 3053 – 3067.
[22] C. Schmuck, W. Wienland, J. Am. Chem. Soc. 2003, 125, 452 –
459.
[23] H. Deng, S. L. Cobb, A. R. McEwan, R. P. McGlinchey, J. H.
Naismith, D. OꢀHagan, D. A. Robinson, J. B. Spencer, Angew.
Chem. 2006, 118, 773 – 776; Angew. Chem. Int. Ed. 2006, 45, 759 –
762.
[24] A. K. Ghosh, Y. Wang, J. Chem. Soc. Perkin Trans. 1 1999, 24,
3597 – 3601.
[25] J. L. Palmer, R. H. Abeles, J. Biol. Chem. 1979, 254, 1217 – 1226.
[26] H. Kimoto, Y. Fujii, S. Hirano, Y. Yokota, A. Takto, J. Biol.
Chem. 2006, 281, 9181 – 9189.
Angew. Chem. Int. Ed. 2008, 47, 5357 –5361
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5361