Pseudoglycosyltransferase catalyzes nonglycosidic C-N coupling in validamycin a biosynthesis

Article abstract of DOI:10.1021/ja203574u

Glycosyltransferases are ubiquitous in nature. They catalyze a glycosidic bond formation between sugar donors and sugar or nonsugar acceptors to produce oligo/polysaccharides, glycoproteins, glycolipids, glycosylated natural products, and other sugar-containing entities. However, a trehalose 6-phosphate synthase-like protein has been found to catalyze an unprecedented nonglycosidic C-N bond formation in the biosynthesis of the aminocyclitol antibiotic validamycin A. This dedicated 'pseudoglycosyltransferase catalyzes a condensation between GDP-valienol and validamine 7-phosphate to give validoxylamine A 7′-phosphate with net retention of the 'anomeric configuration of the donor cyclitol in the product. The enzyme operates in sequence with a phosphatase, which dephosphorylates validoxylamine A 7′-phosphate to validoxylamine A.

Full text of DOI:10.1021/ja203574u

ARTICLE
pubs.acs.org/JACS
Pseudoglycosyltransferase Catalyzes Nonglycosidic CꢀN Coupling in
Validamycin A Biosynthesis
Shumpei Asamizu, Jongtae Yang, Khaled H. Almabruk, and Taifo Mahmud*
Department of Pharmaceutical Sciences, Oregon State University, Corvallis, Oregon 97331-3507, United States
S Supporting Information
b
ABSTRACT: Glycosyltransferases are ubiquitous in nature.
They catalyze a glycosidic bond formation between sugar
donors and sugar or nonsugar acceptors to produce oligo/
polysaccharides, glycoproteins, glycolipids, glycosylated natural
products, and other sugar-containing entities. However, a
trehalose 6-phosphate synthase-like protein has been found to
catalyze an unprecedented nonglycosidic CꢀN bond formation
in the biosynthesis of the aminocyclitol antibiotic validamycin A.
This dedicated ‘pseudoglycosyltransferase’ catalyzes a conden-
sation between GDP-valienol and validamine 7-phosphate to give validoxylamine A 70-phosphate with net retention of the
‘anomeric’ configuration of the donor cyclitol in the product. The enzyme operates in sequence with a phosphatase, which
dephosphorylates validoxylamine A 70-phosphate to validoxylamine A.
’ INTRODUCTION
underlying this coupling reaction had been poorly understood,
stimulating speculations around how the nitrogen bridge in this
compound is formed.2,3,6 Here, we describe the first biochemical
evidence of an unprecedented nonglycosidic CꢀN coupling in
validamycin biosynthesis catalyzed by a trehalose 6-phosphate
synthase-like enzyme, VldE. This dedicated ‘pseudoglycosyl-
transferase’ recognizes GDP-valienol and validamine 7-phos-
phate as substrates and operates in sequence with a
phosphatase (VldH), which dephosphorylates the VldE product
validoxylamine A 70-phosphate to validoxylamine A, a process
that is similar to that of trehalose biosynthesis by trehalose
6-phosphate synthase and trehalose 6-phosphate phosphatase.
Glycosyltransferases (EC 2.4.x.y) are among the most abun-
dant enzymes in nature and play a central role in structural and
physiological aspects of living organisms. They catalyze the
transfer of a sugar moiety from an activated donor sugar onto
sugar or nonsugar acceptors, resulting in sugar-containing pro-
ducts, such as oligo/polysaccharides, glycoproteins, glycolipids,
glycosylated natural products, and many others. To date, there
are more than 44 000 genes which encode proteins across the 92
known families of glycosyltransferases listed in the NCBI data-
base, and this number is fast growing.1 However, only a fraction
of them have actually been functionally characterized.
Whereas all glycosyltransferases characterized to date catalyze
glycosylation reactions, some glycosyltransferase-like proteins
have recently been implicated in nonglycosidic CꢀN bond
formations in natural products biosyntheses. Examples include
ValL/VldE, trehalose 6-phosphate synthase-like proteins in-
volved in the biosynthesis of the crop protectant validamycin A
(1) (Figure 1) in several Streptomyces hygroscopicus sp.,2,3 and
AcbS/GacS, glycosyltransferase-like enzymes proposed to be
involved in the biosynthesis of the antidiabetic drug acarbose
(2) in Actinoplanes sp. SE50/110 and Streptomyces glaucescens.4,5
Validamycin A (1) is a bacterial-derived natural product that
consists of an aminocyclitol moiety, validoxylamine A, and
glucose. The validoxylamine A core structure is built of an
unsaturated cyclitol and a saturated cyclitol, which are connected
through a shared single nitrogen bridge. Previous investigations
using isotopically labeled precursors and biochemical experi-
ments demonstrated that these cyclitol units are derived from
sedoheptulose 7-phosphate via a set of cyclic intermediates,6ꢀ9
and further downstream the pathway requires a coupling reaction
between the two cyclitols. However, the biochemical mechanism
’ RESULTS
Validation of VldB as a Cyclitol Nucleotidyltransferase.
Biosynthetic gene clusters of validamycin have been identified in
a number of Streptomyces strains, for example, S. hygroscopicus
subsp. jinggangensis 5008 and S. hygroscopicus subsp. limoneus
KCCM 11405 (IFO 12704).2,3 Direct comparison of the former
(the val cluster) and the latter (the vld cluster) has shown that
both clusters contain almost identical sets of genes necessary for
the biosynthesis of validamycin A (Figure 2A),10 suggesting that
both producers may be closely related. As part of our investiga-
tion on validamycin biosynthesis in S. hygroscopicus 5008, we
recently established the role of ValB, a cyclitol nucleotidyltrans-
ferase that catalyzes the conversion of valienol 1-phosphate to
GDP-valienol.11 However, its homologue (VldB) from S. hygro-
scopicus subsp. limoneus KCCM 11405 was reported as a glucose
Received: April 18, 2011
Published: July 18, 2011
r
2011 American Chemical Society
12124
dx.doi.org/10.1021/ja203574u J. Am. Chem. Soc. 2011, 133, 12124–12135
|

Journal of the American Chemical Society
ARTICLE
Figure 1. Chemical structures of validamycin A, acarbose, and related natural products.
1-phosphate uridylyltransferase that catalyzes the formation of
UDP-glucose.12 To validate the function of VldB as a cyclitol
nucleotidyltransferase, the recombinant protein was tested using
synthetic valienol 1-phosphate11 as substrate in the presence of
all five naturally occurring NTPs (ATP, GTP, CTP, dTTP, and
UTP). The results showed that VldB, in reactions containing
1 mM Mg2+, effectively catalyzes the conversion of valienol
1-phosphate to GDP-valienol and less effectively to ADP-valie-
nol or other NDP derivatives (Figure 2C). However, interest-
ingly, in reactions containing 10 mM Mg2+, the enzyme more
effectively used ATP as nucleotidyl donor than GTP or other
NTPs (Figure 2C). Although the reason for this nucleotidyl
donor preference switch is unclear, the results confirm our
hypothesis that VldB is a cyclitol nucleotidyltransferase that
plays a critical role in supplying NDP-valienol, most likely
GDP-valienol, in validamycin biosynthesis. Therefore, the for-
mation of validoxylamine A may involve a coupling reaction between
GDP-valienol and an amino-bearing cyclitol counterpart.
Syntheses of Validamine (16) and Validamine 7-Phos-
phate (21). To explore the above-proposed coupling reaction,
we synthesized two potential aminocyclitol substrates, valida-
mine (16) and validamine 7-phosphate (21). The two com-
pounds were prepared from validoxylamine A (8) as shown in
Scheme 1. Thus, complete benzylation of validoxylamine A using
BnBr and NaH furnished perbenzylated validoxylamine A (9),
which was then treated with NBS to give a mixture of four
products, tetrabenzylvalienone (10), tetrabenzylvalidone (11),
tetrabenzylvalidamine (12), and tetrabenzylvalienamine (13).
The benzylated validamine and valienamine products were
chromatographically separated from their keto-analogues, and
the mixture was subjected to Cbz-protection to give 1-Cbz-
2,3,4,7-tetrabenzylvalienamine (14) and 1-Cbz-2,3,4,7-tetraben-
zylvalidamine (15).Thecompoundswerethenchromatographically
separated, and complete deprotection of 15 by hydrogenolysis
afforded validamine (16) in a quantitative yield.
For the synthesis of validamine 7-phosphate (21), 15 was
regioselectively debenzylated and acetylated using ZnCl2, Ac2O,
and AcOH, followed by a deacetylation step to furnish a free
alcohol product, 1-Cbz-2,3,4-tribenzylvalidamine (19). The
alcohol 19 was then treated with dibenzyl diisopropyl phosphor-
amidite in the presence of 1H-tetrazole to form the phosphite
intermediate, which was readily oxidized by m-chloroperoxyben-
zoic acid (m-CPBA) to afford a phosphate product (20). Finally,
debenzylation was achieved by hydrogenolysis with Pd/C to give
the desired product validamine 7-phosphate (21).
Characterization of the Glycosyltransferase-like Protein
VldE. As validoxylamine A is structurally similar to trehalose, the
trehalose 6-phosphate synthase-like protein VldE was suspected
to catalyze this coupling reaction. Trehalose 6-phosphate
synthases (EC 2.4.1.15) catalyze the synthesis of trehalose
6-phosphate from UDP-glucose and glucose 6-phosphate with
net retention of the anomeric configuration of the donor sugar in
the product (Scheme S1). The notion that its homologue is
involved in validoxylamine formation is intriguing, but raises
fundamental questions as to how a glycosyltransferase-like en-
zyme catalyzes a condensation of two nonsugar molecules. To
gain insight into the catalytic function of VldE and its involve-
ment in the coupling reaction, we cloned and heterologously
expressed the gene in E. coli and characterized its biochemical
activity. As the potential donor substrate GDP-valienol is rela-
tively unstable, the compound was prepared in situ using VldB,
valienol 1-phosphate, and GTP. In addition to these compo-
nents, recombinant VldE and the synthetic amino-bearing cycli-
tol, either validamine or validamine 7-phosphate, were included
in the reaction mixture (Scheme 2). The results showed that
only the incubation containing GDP-valienol and validamine
12125
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
Figure 2. Biosynthetic gene clusters for validamycin A and characterization of VldB. (A) Genetic organizations of the validamycin biosynthetic gene
clusters from S. hygroscopicus 5008 and S. hygroscopicus subsp. limoneus; (B) SDSꢀPAGE of recombinant VldB, VldE, VldH, OtsA (Escherichia coli
trehalose 6-phosphate synthase), and OtsB (E. coli trehalose 6-phosphate phosphatase) purified by Ni-NTA affinity chromatography; (C) relative
specificity for nucleotidyl donors of VldB in the presence of 1 mM and 10 mM Mg2+.
7-phosphate gave a product (m/z 416 [M + H]+), which is
consistent with the molecular mass of the expected product
validoxylamine A 70-phosphate (Figure 3G, and Figures S2 and
S3). MS/MS analysis of the product gave fragment ions con-
sistent with those observed in authentic validoxylamine A, except
that most of the fragment ions of validoxylamine A 70-phosphate
are 80 amu higher than those of validoxylamine A, indicating the
presence of a phosphate moiety in the former compound
(Figure 3L). HR-ESTOFMS analysis of the product revealed a
molecular formula of C14H26NO11P, consistent with that of
validoxylamine A 70-phosphate. Reactions containing validamine
did not give any corresponding product, suggesting that VldE
does not recognize an unphosphorylated substrate (Figure S3).
Synthesis of Validoxylamine A 70-Phosphate (26) and
Confirmation of the VldE Product. To further confirm the
product of VldE, we synthesized validoxylamine A 70-phosphate
(26) and compared the VldE product with the synthetic com-
pound. The synthesis was carried out using validoxylamine A (8),
which was prepared from acid hydrolysis of validamycin A (1), as
starting material. Validoxylamine A 70-phosphate (26) was
previously synthesized by Davies and co-workers in 10 steps
starting from 8.13 However, using a modified procedure, we were
able to obtain 26 from 8 in 5 steps (Scheme 3). Thus, 8 was
treated with benzaldehyde dimethyl acetal in the presence of
p-TsOH to selectively protect the 40- and 70-hydroxy groups.
The product was then perbenzylated with benzyl bromide and
NaH to give 28. Compound 28 was treated with DIBAL-H for
selective acetal opening to give 29. The free alcohol functionality
was then phosphorylated as described for the synthesis of
validamine 7-phosphate (21) and the product was debenzylated
using BBr3 at ꢀ40 °C to give validoxylamine A 70-phosphate
(26). This compound was used as reference for the assay of VldE
catalysis. HPLC, MS, and MS/MS analyses of the VldE product
gave chromatographic and spectral data identical to those of the
synthetic compound, confirming its identity as validoxylamine A
70-phosphate (26).
Kinetic Properties of VldE. The kinetic properties of VldE
were determined using a pyruvate kinase/lactate dehydrogenase
(PK/LDH) coupled colorimetric assay (Scheme 2).14 Reactions
were initiated by addition of a saturating amount of either GDP-
valienol (25) or validamine 7-phosphate (21), and the oxidation
of NADH to NAD+, which is stoichiometrically equivalent to the
production of GDP, was measured by the decrease in A340
.
Compound 25 was obtained from valienol 1-phosphate (24) and
GTP catalyzed by VldB and the concentration was determined
using a coupled colorimetric assay containing inorganic pyropho-
sphatase and purine nucleoside phosphorylase (PNP) [EnzChek
Pyrophosphate Assay Kit15 (Molecular Probes)].11 For the
kinetic studies, a protein concentration of 1ꢀ3 μM, a pH of 7.5,
and a reaction time of 1ꢀ3 min, which showed linear decrease of
12126
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
Scheme 1. Synthesis of Validamine (16) and Validamine 7-Phosphate (21)
absorbance, were chosen. The apparent kinetic values, obtained
from Hanes-Woolf plots, were Km 308.0 ( 34.5 μM andkcat 7.0 (
0.4 minꢀ1 for 21 and Km 59.7 ( 11.9 μM and kcat 7.0 ( 0.7
minꢀ1 for 25 (Figure 4).
not give any hybrid products (Figure S8), indicating that OtsA
does not recognize any activated cyclitols as substrate.
VldE Is Not a Validoxylamine A Glycosyltransferase. Inter-
estingly, VldE has previously been reported as a glycosyltransfer-
ase that catalyzes the glucosylation of validoxylamine A to give
validamycin A.16 However, a wealth of evidence is available to
conclude that another glycosyltransferase, ValG (VldK), is
responsible for this conversion.2,17 Nevertheless, we explored
the reported glycosyltransferase activity of VldE by incubating
the enzyme with validoxylamine A and UDP-glucose and analyz-
ing the product by ESIMS. No validamycin A could be identified
in this reaction (Figure S9). The results confirmed the distinct
role of VldE as a dedicated validoxylamine A 70-phosphate
synthase and not a validoxylamine A glycosyltransferase.
The Role of the Phosphatase VldH. The production of
validoxylamine A 70-phosphate by VldE suggests that a depho-
sphorylation reaction is necessary downstream in the pathway to
provide a phosphate-free substrate for the glycosyltransferase ValG
(VldK).2,17 The only candidate phosphatase gene within the
cluster is vldH, which encodes a protein that has high identity with
proteins from the haloacid dehalogenase-like superfamily, which
includes L-2-haloacid dehalogenase, epoxide hydrolase, phospho-
serine phosphatase, phosphomannomutase, phosphoglycolate
phosphatase, P-type ATPase, and many others.18 Interestingly,
the G + C content of the vldH gene is only 60%, relatively low
compared to that of other genes in the cluster (68ꢀ72%).
VldE Is a Dedicated Pseudoglycosyltransferase. Whereas
VldE is genetically similar to trehalose 6-phosphate synthases, it
does not have any trehalose 6-phosphate synthase activity.
Incubation of the enzyme with UDP- or GDP-glucose and
glucose 6-phosphate did not give any product (Figure S5). The
same result was observed when only one of the cyclitol deriva-
tives was replaced by its glucose analogue (Figure S6). These
results suggest that VldE is a dedicated pseudoglycosyltransferase
that only recognizes GDP-valienol and validamine 7-phosphate
as substrates, and not a trehalose 6-phosphate synthase with
relaxed substrate specificity.
The E. coli Trehalose 6-Phosphate Synthase OtsA Does
Not Recognize GDP-Valienol and Validamine 7-Phosphate
as Substrates. To investigate if a trehalose 6-phosphate synthase
can alternatively catalyze the formation of validoxylamine A 70-
phosphate, we cloned and expressed in E. coli the trehalose
6-phosphate synthase gene otsA and characterized its catalytic
function. While the recombinant OtsA can catalyze the forma-
tion of trehalose 6-phosphate, the enzyme was not able to
catalyze the formation of validoxylamine A 70-phosphate
(Figure S7). Replacements of GDP-valienol with UDP-/GDP-
glucose or validamine 7-phosphate with glucose 6-phosphate did
12127
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
Scheme 2. Biosynthetic Pathway to Validamycin A and Characterization of VldB, VldE, and VldH
To investigate the function of VldH as a phosphatase, the
recombinant protein was produced (Figure 2B) and co-incu-
bated with VldB and VldE in the presence of valienol 1-phos-
phate, GTP, and validamine 7-phosphate (Scheme 2). The
results showed that addition of VldH to the mixture resulted in
a phosphate-free product, validoxylamine A (m/z 336 [M + H]+)
(Figure 3D,I). In addition, MS/MS analysis of the product gave
fragment ions identical to those of authentic validoxylamine A
(Figure 3N,O), providing compelling evidence for the critical
roles of VldB, VldE, and VldH in the formation of validoxylamine
A. However, in contrast to VldE, VldH appears to have broader
substrate permissiveness, as it could also convert trehalose
6-phosphate to trehalose (Figure S10). On the other hand, the
trehalose 6-phosphate phosphatase OtsB, which was recombi-
nantly produced in E. coli (Figure 2B), does not dephosphorylate
validoxylamine A 70-phosphate (Figure S4), suggesting that the
role of VldH in validamycin A biosynthesis cannot be assumed
by OtsB.
In addition, the use of the aminocyclitol unit as pseudosugar
acceptor by VldE to form N-pseudoglycosyl linkage is a signifi-
cant departure from the OtsA catalyzed reaction. In comparison
to the O-glycosyltransferases such as OtsA, glycogen synthase,
cellulose synthase, and others, N-glycosyltransferase enzymes are
less abundant, but play significant roles, in biological systems.
Among the known N-glycosyltransferases are the eukaryotic
oligosaccharyltransferase (OST), which catalyzes asparagine-
linked glycosylation of proteins, and UGT72B1, a bifunctional
N- and O-glycosyltransferase that is involved in xenobiotic
metabolism in plants.19,20 Other N-glycosyltransferases that are
involved in natural products biosynthesis include the indolocar-
bazole N-glycosyltransferases RebG and StaG and the ansami-
tocin N-glycosyltransferase Asm25.21ꢀ23 However, most, if not
all, of these enzymes catalyze the formation of glycosidic bonds
with inversion of configuration at the anomeric center of the
sugar donors.
Whereas inverting glycosyltransferases generally operate
through a direct displacement, SN2-like mechanism, retaining
glycosyltransferases are believed to function via either a two-step
double-displacement mechanism or an internal return (SNi)-like
mechanism, which is also known as a DN*ANss ion pair
mechanism.24 Structural studies of OtsA suggest that its activity
is likely to occur via the latter mechanism.13 As VldE shares a high
number of conserved active site residues with OtsA, they may
have similar catalytic mechanisms. However, the question re-
mains how a glycosyltransferase-like enzyme catalyzes a condensa-
tion of two nonglycosyl units with retention of stereochemistry.
Whereas the detailed mechanism underlying this unique cap-
ability is yet to be determined, we believe that the olefinic moiety
of GDP-valienol plays a critical role in facilitating the coupling
’ DISCUSSION
The biochemical analysis of VldB, VldE, and VldH and the
establishment of their roles in validoxylamine A biosynthesis
have opened up a new dimension in our understanding of
pseudoglycoside biosynthesis. Particularly, the involvement of
VldE and VldH, which resemble OtsA and OtsB, in validoxyla-
mine A formation is evolutionarily and mechanistically intri-
guing. VldE might have evolved from an ancestral trehalose
6-phosphate synthase to the extent that it no longer recognizes
activated sugars as substrates, but still retains some of the
catalytic properties, including the net retention of the anomeric
configuration of the product.
12128
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
Figure 3. HPLC and ESI-MS analyses of VldB, VldE, and VldH reaction products. (A) HPLC traces of reaction mixture containing VldB, valienol
1-phosphate, and GTP. GDP-valienol (cyan block arrow) eluted in two tautomeric forms; (B) HPLC trace of reaction mixture containing VldB, VldE,
valienol 1-phosphate, GTP, and validamine 7-phosphate; (C) HPLC trace of authentic (synthetically prepared) validoxylamine A 70-phosphate;
(D) HPLC trace of reaction mixture containing VldB, VldE, VldH, valienol 1-phosphate, GTP, and validamine 7-phosphate; (E) HPLC trace of
authentic validoxylamine A; (F) (ꢀ)ESIMS of reaction mixture containing VldB, valienol 1-phosphate, and GTP; (G) (+)ESIMS of reaction mixture
containing VldB, VldE, valienol 1-phosphate, GTP, and validamine 7-phosphate; (H) (+)ESIMS of authentic validoxylamine A 70-phosphate;
(I) (+)ESIMS of reaction mixture containing VldB, VldE, VldH, valienol 1-phosphate, GTP, and validamine 7-phosphate; (J) (+)ESIMS of authentic
validoxylamine A; (K) ESI-MS/MS of GDP-valienol (m/z 600, cyan block arrow) from panel F; (L) ESI-MS/MS of validoxylamine A 70-phosphate
(m/z 416, red block arrow) from panel G; (M) ESI-MS/MS of authentic validoxylamine A 70-phosphate; (N) ESI-MS/MS of validoxylamine A (m/z
336, green block arrow) from panel I; (O) ESI-MS/MS of authentic validoxylamine A. Yellow circle, GTP; open circle, GDP.
reaction. This is somewhat analogous to the involvement of the
allylic moiety in farnesyl diphosphate (FPP) synthase-catalyzed
reactions (Figure 5).25,26 In both VldE and FPP synthase, and
also in glycosyltransferases, a nucleophilic substitution reaction
takes place at a carbon center with a diphosphate or nucleotidyl
diphosphate acting as leaving group. Therefore, it may be
proposed that at the enzyme active site the π-electrons of the
unsaturated cyclitol resonate to form stabilized short-lived ion
12129
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
Scheme 3. Synthesis of Validoxylamine A 70-Phosphate (26)
Figure 4. Kinetic studies of VldE. (A) MichaelisꢀMenten curve for validamine 7-phosphate; (B) MichaelisꢀMenten curve for GDP-valienol.
pair intermediates similar to the oxocarbenium-ion like transition
states and the diphosphate group acts as a base to deprotonate
the pyralomicins (7) (Figure 1).27 Therefore, the results have not
only laid a foundation for understanding and discovering new
classes of pseudoglycosyltransferases, but may also facilitate the
development of new tools for generating carbohydrate mimetics
and redesigning glycoconjugates and glycosylated natural
products.
+
the ammonium (ꢀNH3 ) group of validamine 7-phosphate. The
primary amine then attacks the ‘pseudoanomeric’ carbon, leading
to a product with retained configuration.
Adoption of the SNi-like mechanism in VldE catalysis, how-
ever, may have a consequence, as the ion pair intermediate could
collapse to give back starting material, which may affect the rate
constant of the enzyme.24 In fact, we observed a rather low
turnover rate of VldE as evidenced from the MS data and its
’ EXPERIMENTAL SECTION
General. All chemical reactions were performed under an argon or
nitrogen atmosphere employing oven-dried glassware. Analytical thin-
layer chromatography (TLC) was performed using silica plates (60 Å)
with a fluorescent indicator (254 nm), which were visualized with a UV
lamp and ceric ammonium molybdate (CAM) solution. Chromato-
graphic purification of products was performed on silica gel (60 Å,
72ꢀ230 mesh). Optical rotations were measured on a Jasco P1010
polarimeter (100 mm cells were used) at the sodium D line. Proton
NMR spectra were recorded on Bruker 300 or 400 MHz spectrometers.
Proton chemical shifts are reported in parts per million (ppm) (δ)
relative to the residual solvent signals as the internal standard (CDCl3:
kinetic values (kcat/Km 0.023 minꢀ1 μMꢀ1 for validamine
3
7-phosphate and kcat/Km 0.12 minꢀ1 μMꢀ1 for GDP-valienol).
3
However, this observation might be circumstantial; therefore, a
more detailed study is necessary to determine the catalytic
mechanism of this unusual enzyme. Nevertheless, the results of
the present study suggest that certain glycosyltransferase-like
enzymes are capable of catalyzing coupling reactions between
nonglycosidic allylic cation-forming donors and amino-bearing
acceptors. Some, if not all, of them are similar to members of
certain groups of glycosyltransferases, for example, VldE to
trehalose 6-phosphate synthases (GT20 family) and AcbS/GacS
to glycogen synthases (GT5 family),4,5 suggesting that this
poorly recognized class of enzymes may have evolved from or
shared ancestral traits with known glycosyltransferases. Similar
glycosyltransferase-like enzymes may be involved in the bio-
synthesis of other natural products, for example, the adiposins
(3), salbostatin (4), the trestatins (5), the amylostatins (6), and
1
δH 7.26; D2O: δH 4.79). Multiplicities in the H NMR spectra are
described as follows: s = singlet, bs = broad singlet, d = doublet, bd =
broad doublet, t = triplet, bt = broad triplet, q = quartet, m = multiplet;
coupling constants are reported in hertz (Hz). Carbon NMR spectra
were recorded on a Bruker 300 (75 MHz) spectrometer with complete
proton decoupling. Carbon chemical shifts are reported in parts per
million (ppm) (δ) relative to the residual solvent signal as the internal
standard (CDCl3: δC 77.16), or with sodium 2,2-dimethylsilapentane-5-
12130
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
Figure 5. Proposed catalytic mechanism for VldE and its similarity to that of farnesyl diphosphate (FPP) synthase and trehalose 6-phosphate synthase
(OtsA). (A) Catalytic mechanism of FPP synthases; (B) proposed SNi-like mechanism for OtsA; (C) proposed catalytic mechanism of VldE adopting
those of FPP synthase and OtsA.
sulfonate (DSS) (δ 0.0) as an external standard. Phosphorus NMR
spectra were recorded on a Bruker 400 (122 MHz) spectrometer with
complete proton decoupling. Phosphorus chemical shifts are reported in
parts per million (ppm) (δ) relative to an 85% H3PO4 (δ 0.0) external
standard. Low-resolution electrospray ionization (ESI) mass spectra
were recorded on a ThermoFinnigan liquid chromatograph-ion trap
mass spectrometer, and high-resolution electrospray mass spectra were
recorded on a Waters/Micromass LCT spectrometer. Size exclusion
chromatography was done on Sephadex LH-20 (Pharmacia).
Chemical Synthesis of Validamine (16) and Validamine
7-Phosphate (21). Benzyl ((1S,2S,3S,4R,5R)-2,3,4-Tris(benzyloxy)-
5-((benzyloxy)methyl)cyclohexyl)-carbamate (15). To a solution of
(1S,4R,5S,6S)-4,5,6-tris(benzyloxy)-3-((benzyloxy)methyl)-N-((1S,2S,
3S,4R,5R)-2,3,4-tris(benzyloxy)-5-((benzyloxy)methyl)-cyclohexyl)-
cyclohex-2-enamine (9)28 (1.77 g, 1.676 mmol) in CH3CN/H2O (4:1;
40 mL) was added NBS (447 mg, 2.51 mmol), and the reaction mixture
was stirred for 15 h at room temperature (rt). The reaction mixture was
diluted with water (150 mL) and EtOAc (150 mL), and the aqueous
layer extracted with EtOAc (150 mL). The combined organic layer was
washed with brine (150 mL), dried over MgSO4, filtered and concen-
trated under vacuum to give crude product. Column chromatography
(silica gel, Hex/EtOAc = 3:1 and then EtOAc/MeOH = 4:1) of the
product yielded a mixture of amine products (626 mg).
products (626 mg) and NaHCO3 (491 mg, 5.85 mmol) in MeOH
(16 mL). The reaction mixture was stirred for 15 h at rt and diluted with
EtOAc (150 mL) and water (150 mL). The organic layer was washed
with brine (150 mL), dried over MgSO4, filtered and concentrated under
vacuum. Column chromatography (silica gel, Hex/EtOAc = 5:1 and
then toluene/EtOAc = 10:1) yielded 15 (8.7 mg, 17% from 9); [R]20
=
D
+47.1° (c 0.72, CHCl3); 1H NMR (300 MHz, CDCl3): δ 7.25ꢀ7.42 (m,
25H), 5.17 (d, J = 1.5, 2H), 5.12 (m, 1H), 4.98 (d, J = 10.5 Hz, 1H), 4.92
(d, J = 10.8 Hz, 1H), 4.82 (d, J = 10.8 Hz, 1H), 4.75 (d, J = 11.4 Hz, 1H),
4.60 (d, J = 11.7 Hz, 1H), 4.59 (d, J = 10.8 Hz, 1H), 4.48 (s, 2H),
3.55ꢀ3.66 (m, 3H), 3.48 (dd, J = 9.0, 2.4 Hz, 1H), 2.30 (bd, J = 14.7 Hz,
1H), 1.95 (m, 1H), 1.62 (m, 1H); 13C NMR (75 MHz, CDCl3): δ 156.5,
138.94, 138.85, 138.6, 138.0, 136.6, 128.6, 128.5, 128.4, 128.19, 128.15,
128.08, 127.97, 127.96, 127.8, 127.6, 129.6, 83.8, 80.9, 80.3, 75.8, 75.3,
73.2, 71.7, 69.8, 66.8, 48.1, 37.5, 28.7. HRMS (ESI) m/z 672.3338 (calcd
for C43H46NO6 [M + H]+: 672.3325).
Validamine (6). A mixture of 15 (15 mg, 22.33 μmol) and 10% Pd/C
(15.0 mg) in MeOH (1.8 mL) and EtOAc (1.2 mL) containing 12 M
HCl (30 μL) was hydrogenated under atmospheric pressure for 20 h.
The reaction mixture was filtered through Celite and a cellulose syringe
filter, washed with MeOH, and concentrated. The crude product was
further purified with silica gel column chromatography (CHCl3/
MeOH/25% NH4OH = 2:3:1) to give the pure product (3.7 mg, 95%).
To compare its physicochemical properties with those reported in the
literature,29,30 the product (1.0 mg, 5.6 μmol) was acetylated using Ac2O
Benzyl chloroformate (439 mg, 2.57 mmol; 50% solution in toluene)
in CHCl3 (8 mL) was slowly added to a solution of the above amine
12131
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
(0.12 mL) in the presence of DMAP (1.4 mg, 11.3 μmol) in pyridine
(0.8 mL) and the product was purified by silica gel column chromatog-
raphy (CHCl3/MeOH = 15:1) to give pentaacetylvalidamine (17) (1.1
mg, 52%). The physicochemical data for the product are in agreement
with those reported in the literature.29,30 [R]19D= +59.1° (c 0.19,
CHCl3) (lit.30 [R]20D= +60.2° (c 0.6, CHCl3)); 1H NMR (300 MHz,
CDCl3): δ 5.55 (m, 1H), 5.21 (dd, J = 9.9, 10.2 Hz, 1H), 4.94ꢀ5.02 (m,
2H), 4.53 (m, 1H), 4.13 (dd, J = 4.5, 11.4 Hz, 1H), 3.90 (dd, J = 3.6, 11.1
Hz, 1H), 2.14ꢀ2.24 (m, 2H), 2.06 (s, 3H), 2.05 (s, 3H), 2.03 (s, 3H),
2.02 (s, 3H) (ꢁ2), 1.60ꢀ1.71 (m, 1H); 13C NMR (75 MHz, CDCl3): δ
170.9, 170.53, 170.47, 170.0, 169.7, 71.7, 71.5, 63.3, 47.0, 35.3, 23.7,
20.88, 20.82, 20.76 (ꢁ3).
(s, 2H), 5.02 (d, J = 9.0 Hz, 5H), 4.90 (d, J = 10.8 Hz, 1H), 4.82 (d, J =
10.8 Hz, 1H), 4.73 (d, J = 10.8 Hz, 1H), 4.67 (d, J = 11.7 Hz, 1H),
4.51ꢀ4.56 (m, 2H), 4.25ꢀ4.31 (m, 1H), 4.13 (bs, 1H), 3.95ꢀ3.99 (m,
1H), 3.45ꢀ3.58 (m, 2H), 3.29 (bt, J = 8.4 Hz, 1H), 2.15 (d, J = 14.4 Hz,
1H), 1.86 (m, 1H), 1.27 (bt, J = 13.8, 1H); 13C NMR (75 MHz, CDCl3):
δ 156.5, 138.8, 138.5, 137.9, 136.5, 136.03 (d, Jc-p = 1.8 Hz), 135.95 (d,
Jc-p = 2.0 Hz), 128.75, 128.71, 128.59, 128.53, 128.49, 128.3, 128.22,
128.12, 128.08, 128.02, 127.9, 127.8, 127.7, 83.5, 80.5, 79.3, 75.8, 75.4,
71.7, 69.53 (d, Jc-p = 1.9 Hz), 69.46 (d, Jc-p = 1.9 Hz), 67.5 (d, Jc-p = 5.85
Hz), 66.9, 47.9, 37.4 (d, Jc-p = 8.76 Hz), 27.8. HRMS (ESI) m/z 842.3435
(calcd for C50H52NO9 [M + H]+: 842.3458).
Validamine 7-Phosphate (21). A mixture of 20 (10 mg, 11.9 μmol)
and 10% Pd/C (5.0 mg) in MeOH (0.6 mL) and EtOAc (0.4 mL)
containing 12 M HCl (10 μL) was hydrogenated under atmospheric
pressure for 20 h. The reaction mixture was filtered through Celite and a
cellulose syringe filter, washed with MeOH and concentrated. The crude
product was further purified with Sephadex LH-20 to give the pure
((1R,2R,3S,4S,5S)-2,3,4-Tris(benzyloxy)-5-(((benzyloxy)carbonyl)-
amino)-cyclohexyl)methyl Acetate (18). To a solution of 15 (135 mg,
0.20 mmol) in Ac2O/AcOH (2:1; 2 mL) was added ZnCl2 (328 mg,
2.4 mmol), and the reaction mixture was stirred for 5 h at rt. The reaction
mixture was diluted with water (100 mL) and EtOAc (100 mL), and the
organic layer washed with sat. aq. Na2CO3 (3 ꢁ 50 mL). The organic
layer was dried over MgSO4, filtered and concentrated under reduced
pressure to give a crude product. Column chromatography (silica gel,
1
product 21 (2.1 mg, 70%); [R]23D= +46.7° (c 0.13, H2O); H NMR
(300 MHz, D2O): δ 4.0ꢀ4.02 (m, 2H), 3.79ꢀ3.86 (m, 2H), 3.57 (t, J =
9.3 Hz, 1H), 3.46 (t, J = 9.0 Hz, 1H), 1.82ꢀ1.90 (m, 2H); 13C NMR (75
MHz, D2O; DSS): δ 76.3, 74.2, 72.7, 67.5 (d, Jc-p = 4.5 Hz), 54.0, 39.8 (d,
Jc-p = 7.5 Hz), 28.6; 31P NMR (122 MHz, D2O; H3PO4): δ 1.70. HRMS
(ESI) m/z 256.0576 (calcd for C7H15NO7P [M ꢀ H]ꢀ: 256.0586).
Chemical Synthesis of Validoxylamine A 70-Phosphate
(26). (4aR,6S,7S,8S,8aR)-7,8-Bis(benzyloxy)-2-phenyl-N-((1S,4R,5S,6S)-
4,5,6-tris(benzyloxy)-3-((benzyloxy)methyl)cyclohex-2-en-1-yl)hex-
ahydro-4H-benzo[d][1,3]dioxin-6-amine (28). To a solution of vali-
doxylamine A (8) (320 mg, 0.95 mmol) in 5 mL dry DMF, benzaldehyde
dimethyl acetal (189 mg, 1.24 mmol), and p-TsOH (181 mg, 0.95 mmol)
were added and the mixture was stirred at 60 °C for 3 h. The reaction was
quenched with diethylamine, and the pH was adjusted to 8. To the same
pot, NaH (275 mg, 11.5 mmol, 12 equiv) was added at 0 °C and stirred
for 30 min. The mixture was brought to rt and benzyl bromide (1.4 mL,
11.5 mmol, 12 equiv) was added and the mixture was stirred for 3 h. The
reaction was quenched with sat. aq. NH4Cl and extracted with EtOAc
(3 ꢁ 5 mL). The EtOAc fractions were combined and washed with brine
(3 ꢁ 5 mL), dried over anhydrous MgSO4, filtered, and concentrated
under reduced pressure. The residue was subjected to column chroma-
tography (silica gel, hexane/EtOAc = 7:1) to afford 28 (420 mg, 2 steps,
Hex/EtOAc = 3:1) of the product yielded 18 (82 mg, 66%); [R]20
=
D
+29.3° (c 1.19, CHCl3); 1H NMR (300 MHz, CDCl3): δ 7.27ꢀ7.38 (m,
20H), 5.14 (s, 2H), 5.10 (m, 1H), 4.95 (d, J = 10.8 Hz, 1H), 4.90 (d, J =
10.8 Hz, 1H), 4.78 (d, J = 10.5 Hz, 1H), 4.71 (d, J = 11.4 Hz, 1H), 4.56
(d, J = 11.4 Hz, 2H), 4.23ꢀ4.28 (m, 2H), 4.13 (dd, J = 11.1, 2.7 Hz, 1H),
3.57ꢀ3.68 (m, 2H), 3.38 (dd, J = 8.1, 10.5 Hz, 1H), 2.26 (bd, J = 14.1
Hz, 1H), 2.03 (m, 1H), 2.00 (s, 3H), 1.37 (bt, J = 14.7 Hz, 1H); 13C
NMR (75 MHz, CDCl3): δ 170.9, 156.5, 138.71, 138.33, 137.83, 136.5,
128.7, 128.54, 128.49, 128.18, 128.15, 128.09, 128.0, 127.95, 127.87,
127.7, 83.6, 80.8, 79.9, 75.8, 75.2, 71.7, 64.4, 66.9, 47.8, 36.2, 28.4, 20.9.
HRMS (ESI) m/z 624.2982 (calcd for C38H42NO7 [M + H]+:
624.2961).
Benzyl ((1S,2S,3S,4R,5R)-2,3,4-Tris(benzyloxy)-5-(hydroxymethyl)-
cyclohexyl)carbamate (19). To a solution of 18 (82 mg, 0.13 mmol) in
MeOH (6.0 mL) was added 30% NaOMe in MeOH (2.4 mL) via
syringe at 0 °C, and the reaction mixture was stirred for 4 h at rt. The
reaction was quenched by addition of water (3.0 mL) and sat. aq. NH4Cl
(5.0 mL), and diluted with water (50 mL) and EtOAc (50 mL). The
organic layer was washed with brine (50 mL), dried over MgSO4, filtered
and concentrated under reduced pressure to give a crude product.
Column chromatography of the product (silica gel, Hex/EtOAc = 1:1)
yielded 19 (68 mg, 90%); [R]20D= +59.3° (c 1.44, CHCl3); 1H NMR
(300 MHz, CDCl3): δ 7.31ꢀ7.37 (m, 20H), 5.08ꢀ5.18 (m, 3H), 4.96
(d, J = 10.8 Hz, 1H), 4.94 (d, J = 11.1 Hz, 1H), 4.78 (d, J = 10.8 Hz, 1H),
4.68 (dd, J = 8.7, 11.1 Hz, 2H), 4.55 (d, J = 11.4 Hz, 1H), 4.23 (bs, 1H),
3.55ꢀ3.72 (m, 4H), 3.41 (bt, J = 8.7 Hz, 1H), 2.17 (bd, J = 14.4 Hz, 1H),
1.85 (m, 1H), 1.31 (td, J = 13.7, 2.7 Hz, 1H); 13C NMR (75 MHz,
CDCl3): δ 156.5, 138.8, 138.4, 137.9, 136.5, 128.7, 128.55, 128.50,
128.3, 128.2, 128.11, 128.05, 128.02, 127.9, 127.7, 83.8, 81.6, 81.0, 75.8,
75.1, 71.7, 66.9, 64.3, 47.9, 38.6, 28.2. HRMS (ESI) m/z 582.2853 (calcd
for C36H40NO6 [M + H]+: 582.2856).
1
46%). H NMR (300 MHz, CDCl3): 7.13ꢀ7.54 (m), 5.92 (brs, 1H),
5.56 (s, 1H, HCPh), 4.89 (d, J = 11 Hz, 1H), 4.34ꢀ4.80 (m, 16H), 4.21
(d, J = 12 Hz, 1H), 3.81ꢀ4.14 (m, 6H), 3.62 (m, 1H), 3.51 (bt, J = 9 Hz),
3.31 (m, 3H), 2.54 (m, 1H), 1.88 (m, 1H), 1.63 (bd, J = 14 Hz, 1H), 1.54
(S, 1H), 0.823 (t, J = 14 Hz); 13C NMR (300 MHz, CDCl3); 139,1,
138.8, 138.4, 135.25, 127.3 - 128.1, 126.9, 126.2, 101.1, 84.6, 82.8, 82.3,
81.16, 81.16, 80.2, 76.6, 75.4, 74.6, 74.6, 73.3, 72.7, 72.3, 71.4, 71.3, 52.9,
51.4, 31.8, 28.7, 26.7. Electrospray MS: m/z 964.40 [M + H]+. HRMS
(ESI) m/z 964.4743 (calcd for C63H66NO8 [M + H]+: 964.4788).
((1R,2R,3S,4S,5S)-2,3,4-Tris(benzyloxy)-5-(((1S,4R,5S,6S)-4,5,6-tris-
(benzyloxy)-3-((benzyloxy)methyl)cyclohex-2-en-1-yl)amino)cyclohexyl)-
methanol (29). To a solution of 28 (200 mg, 0.21 mmol) in CH2Cl2
(4 mL) DIBAL-H (1.0 M in hexane, 311 μL, 0.311 mmol) was added at
0 °C. After stirring at rt for 3 h, more DIBAL-H (150 μL) was added. The
reaction mixture was stirred for 3.5 h and quenched by adding few drops
of saturated solution of potassium sodium tartrate at 0 °C. The mixture
was brought to rt with continuous addition of potassium sodium tartrate
(5 mL). The mixture was diluted with EtOAc and vigorously stirred for
2.5 h. The aqueous phase was extracted three times with EtOAc. The
organic fraction was washed with brine, dried over anhydrous MgSO4,
filtered, and concentrated under reduced pressure. The residue was
subjected to column chromatography (silica gel, hexane/EtOAc = 3:1)
Benzyl ((1S,2S,3S,4R,5R)-2,3,4-Tris(benzyloxy)-5-(((bis(benzyloxy)-
phosphoryl)oxy)methyl)cyclohexyl)carbamate (20). To a solution of
19 (42 mg, 72.7 μmol) and 1H-tetrazole (12.6 mg, 0.18 mmol) in
CH2Cl2 (2 mL) was added dibenzyl N,N-diisopropylphosphoramidite
(50 mg, 0.144 mmol) in CH2Cl2 (1 mL) at rt under Ar, and the reaction
mixture was stirred for 1 h at rt. The reaction mixture was cooled to 0 °C
and m-CPBA (35.6 mg (70%), 0.144 mmol) was added. After 10 min
stirring at 0 °C, the reaction mixture was diluted with EtOAc (25 mL)
and washed with NaHCO3 (2 ꢁ 25 mL) and brine (3 ꢁ 25 mL). The
organic layer was dried over Na2SO4, filtered and concentrated under
reduced pressure. Column chromatography of the product (silica gel,
Hex/EtOAc = 3:2) yielded 20 (41 mg, 68%); [R]19D= +27.1° (c 0.72,
1
to afford 29 (150 mg, 74.9%) as a colorless oil. H NMR (300 MHz,
CDCl3): 7.25 (m), 5.87 (brs, 1H), 4.90 (bt, J = 11 Hz, 2H), 4.58ꢀ4.87
1
CHCl3); H NMR (300 MHz, CDCl3): δ 7.19ꢀ7.37 (m, 30H), 5.10
(m, 13H), 4.22 (d, J = 12 Hz, 1H), 3.84ꢀ4.09 (m, 4H), 3.60 (m, 1H),
12132
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
3.51 (m, 1H), 3.34 (m, 3H), 2.27 (m, 1H) 1.74 (brd, J = 14 Hz), 1.52 (s),
1.03 (t, J = 12 Hz); 13C NMR (300 MHz, CDCl3): 139.0, 138.9, 138.7,
135.6, 128.0ꢀ128.9, 127.9, 127.8, 102.2, 100.6, 83.9, 83.8, 83.5, 78.2,
78.1, 75.9, 75.2, 74.1, 73.8, 72.9, 72.3, 65.3, 51.4, 51.2, 37.7, 28.1, 21.2.
Electrospray MS: m/z 966.47 [M+H]+. HRMS (ESI) m/z 966.4984
(calcd for C63H68NO8 [M + H]+: 966.4945).
OtsA and OtsB Expression Plasmid Construction. The otsA
and otsB genes from E. coli MG1655 were amplified by PCR from
plasmid pTrc99A-otsAB (a gift from Dr. C-J. Kim) using the following
primer pairs: 50-AGG CTC GAG GCA TAT GAG TCG TTT AGT
CGT-30 and 50-GAC GAA TTC TCA CGC AAG CTT TGG AAA
GGT-30 for otsA, 50-GAC CTC GAG ACA TAT GAC AGA ACC GTT
AAC-30 and 50-GCA GAA TTC TCA GAT ACT ACG ACT AAA CGA-
30 for otsB (XhoI and EcoRI sites are underlined). The otsA and otsB PCR
fragments were digested with XhoI and EcoRI and ligated into pRSET-B
(Invitrogen) to give pRSET-B-otsA and pRSET-B-otsB. DNA se-
quences of otsA and otsB were confirmed by CGRB.
Dibenzyl (((1R,2R,3S,4S,5S)-2,3,4-Tris(benzyloxy)-5-(((1S,4R,5S,6S)-
4,5,6-tris(benzyloxy)-3-((benzyloxy)methyl)cyclohex-2-en-1-yl)amino)-
cyclohexyl)methyl) Phosphate (30). A solution of29 (70 mg, 0.0725 mmol)
in CH2Cl2 (2 mL) was added to a mixture of dibenzyl N,N-diisopro-
pylphosphoramidite (37.5 mg, 0.108 mmol) and 1Hꢀtetrazole (15.2 mg,
0.21 mmol) in CH2Cl2 (2 mL), which had been stirred for 40 min, at rt and
the reaction mixture was further stirred for 3 h. The reaction mixture was
cooled to 0 °C and m-CPBA (12.5 mg, 0.0725 mmol) was added. After the
cooling bath was removed, the reaction mixture was stirred for an additional
30 min at rt. The solvent was evapored in vacuo, and the residue was
redissolved in EtOAc (5 mL). The EtOAc solution was washed with 10%
aqueous Na2SO3 (5 mL), saturated aqueous NaHCO3 (5 mL) and brine
(5 mL), dried over anhydrous MgSO4, filtered and concentrated in a rotary
evaporator. The product was purified using column chromatography (silica
gel, tolueneꢀEtOAc = 3: 1) to afford 30 (65mg, 73%). 1HNMR(300MHz,
CDCl3): 7.25 (m), 5.92 (br s, 1H), 4.96 (d, J = 9 Hz, 4H), 4.85 (t, J = 10 Hz,
2H), 4.38 - 4.75 (m, 14H), 4.22 (d, J = 12 Hz, 2H), 4.02 (br d, J = 4 Hz, 1H),
3.88 (m, 3H), 3.71 (m, 1H), 3.55 (m, 1H), 3.38 (dd, J = 9 Hz, J = 4 Hz, 1H),
3.26 (m, 3H), 2.28 (m, 1H), 1.67 (m, 1H), 1.12 (t, J = 15 Hz); 13C NMR
(300 MHz, CDCl3) 138.6ꢀ139.4, 126.4ꢀ128.9, 83.6, 83.2, 80.3, 77.9, 77.6,
77.4, 76.8, 75.7, 75.3, 74.2, 73.6, 72.6, 72.5, 71.1, 69.5, 69.4, 68.3, 51.5, 51.3,
36.7, 36.6, 27.9, 20.7. Electrospray MS: m/z 1226.53 [M+H]+. HRMS (ESI)
m/z 1226.5573 (calcd for C77H81NO11P [M + H]+: 1226.5542).
VldB, VldE, and VldH Production and Purification. pET20-
vldB, pRSETB-vldE and pRSETB-vldH were used to transform E. coli
BL21(DE3) pLysS. Transformants were grown overnight at 37 °C on
LB agar plate containing 100 μg/mL ampicillin and 25 μg/mL chlor-
amphenicol. A single colony was inoculated into 3 mL of LB medium
and cultured at 37 °C for 6 h and then 1 mL of seed culture was
transferred into 100 mL of LB medium in a 500 mL flask and grown at
28 °C until OD600 reached 0.6. Then, the temperature was reduced to
18 °C, and after 2 h of adaptation, 0.1 mM IPTG was added to induce the
C-terminal hexahistidine-tagged VldB and the N-terminal hexahistidine-
tagged VldE and VldH proteins. After further growth for 14 h, the cells
were harvested by centrifugation (5000 rpm, 10 min, 4 °C) and stored at
ꢀ80 °C until used. Cell pellets from 50 mL of culture were washed with
1 mL of binding (B) buffer (40 mM HEPES (pH 8.0), 300 mM NaCl,
10% glycerol, and 10 mM imidazole) and centrifuged (6000 rpm, 3 min,
4 °C). Then, 1 mL of B buffer was added and the mixtures were
sonicated (8 W, 15 s, 4 times). After centrifugation (14 500 rpm, 20 min,
4 °C), 0.8 mL of the supernatants was each mixed with 0.2 mL of B
buffer-equilibrated Ni-NTA resin (QIAGEN) and incubated for 1 h at
4 °C. After incubation, the mixtures were centrifuged (4000 rpm, 3 min,
4 °C) and the supernatants were discarded. One milliliter of washing
(W) buffer (40 mM HEPES pH 8.0, 300 mM NaCl, 10% glycerol, and
20 mM imidazole in the case of VldE and VldH or 100 mM imidazole in
the case of VldB) was added and the mixture centrifuged (4000 rpm, 3
min, 4 °C). This washing step was repeated three times. Subsequently,
0.5 mL of elution (E) buffer (40 mM HEPES pH 8.0, containing
300 mM NaCl, 10% glycerol, and 500 mM imidazole) was added to elute
the desired proteins. This elution step was repeated twice. Eluted VldB,
VldE, and VldH were dialyzed against 1 L of dialysis buffer (10 mM Tris-
HCl pH 7.5, 0.1 mM DTT, 1 mM MgCl2) 3 times for 3 h each. Purified
proteins were analyzed by SDSꢀPAGE and concentrated by ultrafiltra-
tion using Amicon YM-10 (Millipore). Protein concentration was
determined using the Bradford assay (BIO-RAD) with BSA as standard.
Coupled VldB/VldE Assay with ATP and GTP. To determine
whether ATP or GTP is the natural nucleotidyl donor in the coupled
VldB/VldE reaction, VldB (1 μM) and VldE (3 μM) were incubated
with 4 mM valienol 1-phosphate, 4 mM validamine 7-phosphate, 6 mM
of either ATP or GTP, and 2 mM MgCl2 in 60 mM Tris-HCl, pH 7.5.
Reaction mixtures were incubated at 30 °C for 3 h. Then, 2 vol of MeOH
were added and mixed. The mixture was centrifuged (14 500 rpm, 5 min,
4 °C) and the supernatant was subjected to MS analysis. The results
revealed that GTP is used more effectively in the VldB/VldE coupled
assay than ATP.
Validoxylamine A 70-Phosphate (26). BBr3 (417 μL of 1.0 M
solution in CH2Cl2, 0.417 mmol) was added to a solution of 30 (28.4
mg, 0.023 mmol) in CH2Cl2, and the reaction mixture was stirred for
60 min at ꢀ40 °C. The mixture was then placed in an ice bath and several
drops of saturated aqueous NaHCO3 solution were slowly added until
no more effervescence was observed. The mixture was then partitioned
between H2O (1 mL) and CH2Cl2 (1 mL). The aqueous fraction was
collected and dried in a SpeedVac rotator and the residue was dissolved
in H2O (500 μL). The solution (500 μL) was subjected to Biogel P-2
Gel fine column chromatography (10 cm ꢁ0.5 cm) and eluted with
H2O. Fractions containing the product were pooled, dried, and purified
1
on Sephadex LH-20 (75 cm ꢁ1 cm) to afford 26 (4.5 mg, 46%). H
NMR (300 MHz, D2O): 5.98 (br s, 1H), 4.33 (s, 2H), 4.22 (br d, J = 4.5
Hz, 2H), 4.13 (m, 1H), 4.04 (br s, 2H), 3.90 (m, 3H), 3.69 (t, J = 8 Hz,
1H), 3.51 (t, J = 9 Hz, 1H), 2.31 (brd, J = 14 Hz), 2.03 (m, 1H), 1.86 (t,
J = 14 Hz); 13C NMR (300 MHz, D2O): 114.9, 73.7, 71.9, 71.5, 70.7,
70.4, 66.5, 64.7, 61.2, 58.1, 56.9, 37.4, 24.5, 20.9. Electrospray MS: m/z
416.20 [M + H]+. HRMS (ESI) m/z 416.1320 (calcd for C14H27NO11P
[M + H]+: 416.1322).
VldB, VldE, and VldH Expression Plasmid Construction.
Genomic DNA of S. hygroscopicus subsp. limoneus was prepared by a
general procedure.31 The vldB, vldE, and vldH genes were amplified by
PCR from the genomic DNA using the following primer pairs: 50-TGG
GGC TCG AGG CAT ATG GAC GGA GTG CGT-30 and 50-TTC
GAA CTC GAG CAG CGC CAC-30 for vldB, 50-AAG ATC TCG AGA
CAT ATG ACC GGA TCT GAG-30 and 50-TCA GAA TTC TCA GAG
GTC TGC-30 for vldE, 50-GGT GAC TCG AGA CAT ATG TAC AAG
GTT GCA-30 and 50-TGA GAA TTC TCA GGA AGG ACC AAT ATG
CGG-30 for vldH (XhoI, NdeI, and EcoRI sites are underlined). The vldB
PCR fragment was digested with NdeI and XhoI and ligated into pET20b
(Novagen) to give pET20b-vldB. The vldE and vldH PCR fragments
were digested with XhoI and EcoRI and ligated into pRSET-B (Invitrogen)
to give pRSET-B-vldE and pRSET-B-vldH. The DNA sequences of vldB,
vldE, and vldH were confirmed by the Center for Genome Research and
Biocomputing (CGRB) at Oregon State University.
Combined VldB, VldE, and VldH Assay. Typical reaction
condition for activity assays were 60 mM Tris-HCl, pH 7.5, 4 mM
valienol 1-phosphate; 6 mM NTPs; 2 mM MgCl2; 1 μM VldB; 4 mM
validamine 7-phosphate; 3 μM VldE; and 3 μM VldH. Reaction mixtures
were incubated at 30 °C for 3 h. For enzyme inactivation, VldB, VldE,
and VldH were heated at 100 °C for 5 min.
HPLC Analysis of VldB, VldE, and VldH Products. For HPLC
analysis, the reaction was quenched by adding 2 vol of MeOH and
centrifuged, and the supernatant was analyzed by HPLC (Shimadzu
SPD-20A system) with a Gemini C18 column (4.6 ꢁ 150 mm, 5 μm,
12133
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
Phenomenex) using 0.5% MeOH in H2O containing 10 mM
NH4HCO3 as eluent at a flow rate of 1 mL/min. Elution profiles were
monitored at 210 nm.
High-Resolution Mass Spectral Analysis of VldB/VldE and
VldB/VldE/VldH Products. HR-ESTOFMS of the VldB/VldE pro-
duct: m/z 416.1315 (calcd for C14H27NO11P [M + H]+: 416.1322).
HR-ESTOFMS of the VldB/VldE/VldH product: m/z 336.1657 (calcd
for C14H26NO8 [M + H]+: 336.1658).
centrifuged (14 500 rpm, 5 min, 4 °C) and the supernatant was subjected
to MS analysis. The results revealed that VldH was able to catalyze
dephosphorylation of trehalose 6-phosphate to give trehalose.
Assay of VldE for Validoxylamine A Glycosyltransferase
Activity. To determine if VldE can catalyze the glycosylation of
validoxylamine A to validamycin A, the enzyme (50 μM) was incubated
with 1 mM validoxylamine A, 1 mM UDP-glucose, 10 mM MgCl2, and
10 mM NH4Cl in 50 mM Tris-HCl (pH 8.0). After 3 h incubation at
30 °C, 2 vol of MeOH were added and mixed. The mixture was
centrifuged (14 500 rpm, 5 min, 4 °C) and the supernatant was subjected
to MS analysis. The results show that VldE is not a validoxylamine A
glycosyltransferase.
Preparation of GDP-Valienol for Kinetic Studies. A reaction
mixture (100 μL) containing 60 mM Tris-HCl, pH 7.5; 2 mM MgCl2;
5 mM valienol 1-phosphate; 7.5 mM GTP; and 3 μM VldB was
incubated at 30 °C for 2 h and VldB was removed by ultrafiltration
(YM-10, Amicon). GDP-valienol formed in the reaction mixture was
measured by quantification of the generated inorganic diphosphate
generated using the EnzChek pyrophosphate assay kit (E-6645, In-
vitrogen). Reaction mixtures lacking VldB were used as negative control
(blank).
Km and kcat Determination for VldE. Km and kcat values for VldE
were determined by using the phosphopyruvate kinase (PK) and lactate
dehydrogenase (LDH) coupled assay (Sigma-Aldrich). The reaction
mixture (100 μL) contained 50 mM Tris-HCl, pH 7.5; 5 mM MgCl2;
1.0ꢀ3.0 μM VldE; 1.5 mM phosphoenolpyruvate; 0.5 mM NADH; 37
U of PK; and 46 U of LDH, in addition to 0, 10, 20, 30, 40, 50, 60, 90,
120, 150, 300, 600 μM GDP-valienol and 1 mM validamine 7-phosphate
for the determination of the value for GDP-valienol, and 0, 50, 100, 150,
200, 300, 500, 1000 μM validamine 7-phosphate and 300 μM GDP-
valienol for the determination of the value for validamine 7-phosphate.
Oxidation of NADH to NAD+ was monitored in 96-well plates using a
spectrophotometric microplate reader at 340 nm. Reaction mixtures
lacking VldE were used as negative control (blank). The data were
collected in duplicate (for GDP-valienol) and triplicate (for validamine
7-phosphate). Hanes-Woolf plot was used for Km and kcat value
determination.
’ ASSOCIATED CONTENT
S
Supporting Information. Comparison of catalytic activ-
b
ities of trehalose 6-phosphate synthase and VldE (Scheme S1),
complete HPLC analysis of VldB, VldE, and VldH (Figure S1),
MS analysis of VldB, VldE, and VldH (Figures S2ꢀS10), amino
acid alignment of trehalose 6-phosphate synthase and VldE
(Figure S11), 1H-, 13C-, and 31P NMR data for synthetic
compounds generated in this study (Figures S12ꢀS30). This
material is available free of charge via the Internet at http://pubs.
acs.org.
’ AUTHOR INFORMATION
Corresponding Author
Taifo.Mahmud@oregonstate.edu
’ ACKNOWLEDGMENT
This paper is dedicated to Professor (Emeritus) Isao Kitagawa
on the occasion of his 80th birthday. The authors thank Chang-
Joon Kim for providing plasmid pTrc99A-otsAB, Heinz G. Floss
and Philip J. Proteau for their comments and assistance in the
preparation of this manuscript. The project described was
supported by grant R01AI061528 from the National Institute
of Allergy and Infectious Diseases and by the Herman-Frasch
Foundation. This publication was made possible, in part, by the
Mass Spectrometry Facilities and Service Core of the Environ-
mental Health Sciences Center, Oregon State University, grant
number P30 ES00210, National Institute of Environmental
Health Sciences, NIH.
Assay of VldE for Trehalose 6-Phosphate Synthase Activ-
ity. To determine if VldE has trehalose 6-phosphate synthase activity,
the enzyme (6 μM) was incubated with 1 mM donor substrate (either
GDP-valienol, UDP-glucose, or GDP-glucose), 1 mM acceptor sub-
strate (either validamine 7-phosphate or glucose 6-phosphate), 2 mM
MgCl2, 16 μM VldH or OtsB in 50 mM Tris-HCl (pH 7.5). After a 3-h
incubation at 30 °C, 2 vol of MeOH were added and mixed. The mixture
was centrifuged (14 500 rpm, 5 min, 4 °C) and the supernatant was
subjected to MS analysis. The results revealed that VldE does not
recognize UDP-glucose, GDP-glucose, or glucose 6-phosphate as
substrates.
OtsA and OtsB Production and Purification. The productions
and purifications of N-terminal hexahistidine-tagged OtsA and OtsB
were carried out by the same procedures as used for VldE and VldH.
Assay of OtsA for Validoxylamine A 70-Phosphate Synthase
Activity. To determine if OtsA has validoxylamine A 70-phosphate
synthase activity, the enzyme (6 μM) was incubated with 1 mM donor
substrate (either GDP-valienol or UDP-glucose), 1 mM acceptor
substrate (either validamine 7-phosphate or glucose 6-phosphate),
5 mM MgCl2, 16 μM OtsB or VldH in 50 mM Tris-HCl (pH 7.5).
After 3 h incubation at 30 °C, 2 vol of MeOH were added and mixed.
The mixture was centrifuged (14 500 rpm, 5 min, 4 °C) and the
supernatant was subjected to MS analysis. The results revealed that
OtsA does not recognize GDP-valienol and validamine 7-phosphate as
substrates.
’ REFERENCES
(1) Hansen, S. F.; Bettler, E.; Rinnan, A.; Engelsen, S. B.; Breton, C.
Mol. BioSyst. 2010, 6 (10), 1773–1781.
(2) Bai, L.; Li, L.; Xu, H.; Minagawa, K.; Yu, Y.; Zhang, Y.; Zhou, X.;
Floss, H. G.; Mahmud, T.; Deng, Z. Chem. Biol. 2006, 13 (4), 387–397.
(3) Singh, D.; Seo, M. J.; Kwon, H. J.; Rajkarnikar, A.; Kim, K. R.;
Kim, S. O.; Suh, J. W. Gene 2006, 376 (1), 13–23.
(4) Wehmeier, U. F.; Piepersberg, W. Appl. Microbiol. Biotechnol.
2004, 63 (6), 613–625.
(5) Rockser, Y.; Wehmeier, U. F. J. Biotechnol. 2009, 140 (1ꢀ2),
114–123.
(6) Dong, H.; Mahmud, T.; Tornus, I.; Lee, S.; Floss, H. G. J. Am.
Chem. Soc. 2001, 123 (12), 2733–2742.
(7) Yu, Y.; Bai, L.; Minagawa, K.; Jian, X.; Li, L.; Li, J.; Chen, S.; Cao,
E.; Mahmud, T.; Floss, H. G.; Zhou, X.; Deng, Z. Appl. Environ.
Microbiol. 2005, 71 (9), 5066–5076.
Characterization of OtsB and VldH as Trehalose 6-Phos-
phate Phosphatases. To determine if VldH has trehalose 6-phos-
phate phosphatase activity, the enzyme (8 μM) was incubated with
5 mM trehalose 6-phosphate, and 10 mM MgCl2 in 50 mM Tris-HCl
(pH 7.5). OtsB was used as positive control. After 1.5 h incubation at
30 °C, 2 vol of MeOH were added and mixed. The mixture was
(8) Minagawa, K.; Zhang, Y.; Ito, T.; Bai, L.; Deng, Z.; Mahmud, T.
ChemBioChem 2007, 8 (6), 632–641.
12134
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135

Journal of the American Chemical Society
ARTICLE
(9) Xu, H.; Zhang, Y.; Yang, J.; Mahmud, T.; Bai, L.; Deng, Z. Chem.
Biol. 2009, 16 (5), 567–576.
(10) Flatt, P. M.; Mahmud, T. Nat. Prod. Rep. 2007, 24 (2), 358–392.
(11) Yang, J.; Xu, H.; Zhang, Y.; Bai, L.; Deng, Z.; Mahmud, T. Org.
Biomol. Chem. 2011, 9 (2), 438–449.
(12) Seo, M. J.; Im, E. M.; Singh, D.; Rajkarnikar, A.; Kwon, H. J.;
Hyun, C. G.; Suh, J. W.; Pyun, Y. R.; Kim, S. O. J. Microbiol. Biotechnol.
2006, 16 (8), 1311–1315.
(13) Errey, J. C.; Lee, S. S.; Gibson, R. P.; Martinez Fleites, C.; Barry,
C. S.; Jung, P. M.; O’Sullivan, A. C.; Davis, B. G.; Davies, G. J. Angew.
Chem., Int. Ed. 2010, 49 (7), 1234–1237.
(14) Tamura, J. K.; Gellert, M. J. Biol. Chem. 1990, 265 (34),
21342–21349.
(15) Upson, R. H.; Haugland, R. P.; Malekzadeh, M. N.; Haugland,
R. P. Anal. Biochem. 1996, 243 (1), 41–45.
(16) Singh, D.; Kwon, H. J.; Rajkarnikar, A.; Suh, J. W. Gene 2007,
395 (1ꢀ2), 151–159.
(17) Xu, H.; Minagawa, K.; Bai, L.; Deng, Z.; Mahmud, T. J. Nat.
Prod. 2008, 71 (7), 1233–1236.
(18) Allen, K. N.; Dunaway-Mariano, D. Trends Biochem. Sci. 2004,
29 (9), 495–503.
(19) Kelleher, D. J.; Gilmore, R. Glycobiology 2006, 16 (4),
47R–62R.
(20) Brazier-Hicks, M.; Offen, W. A.; Gershater, M. C.; Revett, T. J.;
Lim, E. K.; Bowles, D. J.; Davies, G. J.; Edwards, R. Proc. Natl. Acad. Sci.
U.S.A 2007, 104 (51), 20238–20243.
(21) Zhang, C.; Albermann, C.; Fu, X.; Peters, N. R.; Chisholm, J. D.;
Zhang, G.; Gilbert, E. J.; Wang, P. G.; Van Vranken, D. L.; Thorson, J. S.
ChemBioChem 2006, 7 (5), 795–804.
(22) Salas, A. P.; Zhu, L.; Sanchez, C.; Brana, A. F.; Rohr, J.; Mendez,
C.; Salas, J. A. Mol. Microbiol. 2005, 58 (1), 17–27.
(23) Zhao, P.; Bai, L.; Ma, J.; Zeng, Y.; Li, L.; Zhang, Y.; Lu, C.; Dai,
H.; Wu, Z.; Li, Y.; Wu, X.; Chen, G.; Hao, X.; Shen, Y.; Deng, Z.; Floss,
H. G. Chem. Biol. 2008, 15 (8), 863–874.
(24) Lairson, L. L.; Henrissat, B.; Davies, G. J.; Withers, S. G. Annu.
Rev. Biochem. 2008, 77, 521–555.
(25) Kellogg, B. A.; Poulter, C. D. Curr. Opin. Chem. Biol. 1997, 1
(4), 570–578.
(26) Thulasiram, H. V.; Poulter, C. D. J. Am. Chem. Soc. 2006, 128
(49), 15819–15823.
(27) Mahmud, T. Nat. Prod. Rep. 2003, 20 (1), 137–166.
(28) Ogawa, S.; Nakajima, A.; Miyamoto, Y. J. Chem. Soc., Perkin
Trans. 1 1991, 12, 3287–3290.
(29) Shing, T. K. M.; Tai, V. W.-F. J. Org. Chem. 1995,
60, 5332–5334.
(30) Ogawa, S.; Shibata, Y.; Chida, N.; Suami, T. Bull. Chem. Soc. Jpn.
1983, 56, 494–498.
(31) Sambrook, J.; Russell, D. W., Molecular Cloning. A Laboratory
Manual. 3rd ed.; Cold Spring Harbor Laboratory Press: New York, 2001.
12135
dx.doi.org/10.1021/ja203574u |J. Am. Chem. Soc. 2011, 133, 12124–12135