Catalytic analysis of the validamycin glycosyltransferase (ValG) and enzymatic production of 4″-epi-validamycin A

Article abstract of DOI:10.1021/np800185k

ValG is a glycosyltransferase (GT) that is responsible for the glucosylation of validoxylamine A to validamycin A. To explore the potential utilization of ValG as a tool for the production of validamycin analogues, a number of nucleotidyldiphosphate-sugars were evaluated as alternative substrates for VaIG. The results indicated that in addition to its natural substrate, UDP-glucose, ValG also efficiently utilized UDP-galactose as sugar donor and resulted in the production of an unnatural compound, 4″-epi-validamycin A. The new compound demonstrated a moderate growth inhibitory activity against the plant fungal pathogen Rhizoctonia solani (=Pellicularia sasakii). A comparative analysis of ValG with its homologous proteins revealed that ValG contains an unusual DTG motif, in place of the DXD motif proposed for metal ion binding and/or NDP-sugar binding and commonly found in other glycosyltransferases. Site-directed mutagenesis of the DTG motif of ValG to DCD altered its preferences for metal ion binding, but did not seem to affect its substrate specificity.

Full text of DOI:10.1021/np800185k

J. Nat. Prod. 2008, 71, 1233–1236
1233
Catalytic Analysis of the Validamycin Glycosyltransferase (ValG) and Enzymatic Production of
4′′-epi-Validamycin A
Hui Xu,†,‡ Kazuyuki Minagawa,† Linquan Bai,‡ Zixin Deng,‡ and Taifo Mahmud*,†
Department of Pharmaceutical Sciences, Oregon State UniVersity, CorVallis, Oregon 97331-3507, and Laboratory of Microbial Metabolism
and School of Life Science & Biotechnology, Shanghai Jiaotong UniVersity, Shanghai 200030, People’s Republic of China
ReceiVed March 24, 2008
ValG is a glycosyltransferase (GT) that is responsible for the glucosylation of validoxylamine A to validamycin A. To
explore the potential utilization of ValG as a tool for the production of validamycin analogues, a number of
nucleotidyldiphosphate-sugars were evaluated as alternative substrates for ValG. The results indicated that in addition
to its natural substrate, UDP-glucose, ValG also efficiently utilized UDP-galactose as sugar donor and resulted in the
production of an unnatural compound, 4′′-epi-validamycin A. The new compound demonstrated a moderate growth
inhibitory activity against the plant fungal pathogen Rhizoctonia solani ()Pellicularia sasakii). A comparative analysis
of ValG with its homologous proteins revealed that ValG contains an unusual DTG motif, in place of the DXD motif
proposed for metal ion binding and/or NDP-sugar binding and commonly found in other glycosyltransferases. Site-
directed mutagenesis of the DTG motif of ValG to DCD altered its preferences for metal ion binding, but did not seem
to affect its substrate specificity.
Combinatorial biocatalysis is an emerging technology in the field
of drug discovery and development. Some of the approaches include
the use of enzymatic, chemoenzymatic, and microbial transforma-
tions to generate libraries of new chemical entities from lead
compounds.1–3 Since many bioactive natural products contain sugar
moieties, which play a critical role in their pharmacological
activities,4,5 glycosylation processes of natural products have
become a prime subject of investigation.6
Glycosyltransferases (GTs) catalyze the transfer of a sugar moiety
from a nucleotidyldiphosphate (NDP)-sugar to an acceptor, which
could be a growing oligosaccharide, a lipid, a protein, or a small
molecule. On the basis of tertiary structure analysis, GTs have been
divided into two superfamilies, known as GT-A and GT-B.7
Members of the GT-A superfamily contain two dissimilar domains,
one involved in the recognition of the NDP-sugar and the other in
the recognition of the acceptor molecule. Most of the GTs in the
Leloir pathway that reside in the Golgi apparatus and the endo-
plasmic reticulum belong to this family. The GT-B superfamily is
remarkably diverse and contains members that are extremely
promiscuous to their NDP-sugar donors.8,9 The GT-B family
consists mostly of prokaryotic enzymes that glycosylate secondary
metabolites to produce active natural products, as well as some
glycosyltranferases from the primary pathways.
Figure 1. Biosynthesis of validamycin A.
Validamycin A (2), a commercially used agricultural antifungal
antibiotic, is a microbial-derived pseudotrisaccharide, which con-
tains a core aminocyclitol moiety, validoxylamine A (1), and
glucose. The core aminocyclitol moiety is derived from two units
of 2-epi-5-epi-valiolone, each of which is a cyclization product of
the C7-sugar phosphate sedoheptulose 7-phosphate. The glucose
attachment is critical for the intake of the drug by the fungal cells
through the common oligosaccharide transport system. Inside the
cells, the compound is hydrolyzed to 1 and serves as a competitive
inhibitor of trehalase.10 In fungi, trehalose is commonly used as a
storage carbohydrate, which can be hydrolyzed by trehalase to two
molecules of glucose for energy supply and other physiological
purposes. Therefore, inhibition of the trehalase activity is detrimental
to fungal growth.
As part of our ongoing study on the biosynthesis of validamycin,
we have identified the complete biosynthetic gene cluster in
Streptomyces hygroscopicus subsp. jinggangensis 5008.11 Among
the proteins believed to be directly involved in the biosynthesis,
ValG has been characterized both in ViVo and in Vitro as a
glycosyltransferase that catalyzes the conversion of 1 to 2 using
UDP-glucose as the sugar donor (Figure 1).11 Interestingly, ValG
belongs to the GT-A family of glycosyltransferases, even though
it functions in secondary metabolism. ValG is also unique in that
it has an unusual DTG motif instead of the DXD motif common in
closely related proteins. While generally there is no significant
identity between different GT families, the acidic DXD motif is
highly conserved in almost all GTs12 and is predicted to participate
in the coordination of a divalent metal ion (most commonly Mn2+
)
and/or in the binding of the NDP-sugar. Mutagenesis of a
mannosyltransferase has shown that altering either of these aspar-
tates completely eliminates the enzymatic activity without causing
the protein to misfold or denature.13 In this study, we investigated
* Corresponding author. Phone: 1-543-737-9679. Fax: 1-541-737-3999.
E-mail: Taifo.Mahmud@oregonstate.edu.
† Oregon State University.
‡ Shanghai Jiaotong University.
10.1021/np800185k CCC: $40.75
2008 American Chemical Society and American Society of Pharmacognosy
Published on Web 06/19/2008

1234 Journal of Natural Products, 2008, Vol. 71, No. 7
Xu et al.
Figure 2. Bioconversion of 1 to 3 using the glycosyltransferase ValG.
the utilization of ValG as a tool for generating analogues of
validamycin by testing a number of commercially available NDP-
sugars as substrates. The involvement of the DTG sequence of ValG
in its catalytic activity or substrate specificity was explored by
replacing the DTG sequence with a DCD motif.
Results and Discussion
Enzymatic Assays of ValG with Different NDP-Sugars.
To explore the possibility of employing ValG as a tool for
generating analogues of validamycin, a number of commercially
available NDP-sugars were tested as sugar donors. The recombinant
histidine-tagged protein was prepared heterologously in E. coli
BL21Gold(DE3)pLysS and purified on a BD TALON affinity
column. The enzyme was incubated with validoxylamine A as a
sugar acceptor and either UDP-glucose, UDP-galactose, UDP-N-
acetylglucosamine, UDP-glucuronic acid, or GDP-mannose as a
sugar donor. The reactions were carried out in the presence of Mg2+
for 1-6 h and monitored by TLC. UDP-glucose is the natural sugar
donor of ValG, which converts validoxylamine A (1) to validamycin
A (2), and was used in this study as a positive control to validate
the activity of the enzyme. The results revealed that, in addition to
UDP-glucose, ValG also utilizes UDP-galactose as substrate to
produce a new validamycin analogue, 4′′-epi-validamycin A (3)
(Figure 2), whereas incubations with UDP-N-acetylglucosamine,
UDP-glucuronic acid, or GDP-mannose did not give any products
at a detectable level.
Interestingly, in contrast to the conversion of 1 to 2, a complete
conversion of 1 to 3 can be achieved only when a purified ValG
was used. When a cell-free extract of E. coli harboring ValG was
used for the enzyme reaction, only part of 1 was converted to 3
after 3 h, and most of the product was hydrolyzed back to 1 after
6 h (as monitored by TLC and MS). We hypothesized that this
hydrolysis was due to an endogenous ꢀ-galactosidase activity of
E. coli origin. Therefore, both 2 and 3 were incubated individually
with either cell-free extract of E. coli harboring the ValG gene or
cell-free extract of E. coli harboring the pRSET-B vector. The
reactions were carried out with or without an addition of uridyl-
diphosphate (UDP). Validoxylamine A (1) was not produced during
any of the incubations with 2, whereas a significant amount of 1
was produced in incubations with 3. This activity was not dependent
on the presence of additional UDP (data not shown).
1
Figure 3. Partial H NMR spectra of validamycin A (2) and 4′′-
epi-validamycin A (3).
ValG.11 Complete structure elucidation and assignments of the
1
protons and carbons were based on direct comparisons of the H
and 13C NMR spectroscopic data with those for 2, as well as 2D
1
NMR, including H-1H COSY, HSQC, and HMBC.
Fungal Growth Inhibitory Assay of 4′′-epi-Validamycin
A. It has been reported that 1 has a stronger in Vitro inhibitory
activity against trehalase than 2. However, 2 is more active in ViVo,
as it is efficiently taken up by the fungal cells and then hydrolyzed
by an intracellular ꢀ-glucosidase to give the active pharmacophore
1.15 Given that 3 is readily hydrolyzed by the E. coli enzymes to
give 1, it was anticipated that 3 might be comparably active to or
more active than 2 if a similar ꢀ-galactosidase activity is present
in the fungal cells. To compare their fungal growth inhibitory
activity, 2 and 3 were tested against Pellicularia sasakii using an
agar plug assay.16 Agar plugs with diameter of 5 mm and containing
P. sasakii mycelia were placed in the center of agar plates that
contain 1 mg of each compound, and the growth of the pathogen
was monitored after 2 days. Active compounds were determined
on the basis of their ability to inhibit the expansion of the fungal
mycelia on the Petri dishes. Consistent with the reported results,
validamycin A (2) demonstrated a potent fungal growth inhibitory
activity (Figure 4A), which validated the sensitivity of the assay.
While 3 demonstrated a moderate growth inhibitory activity against
P. sasakii, it is unexpectedly less active than 2 (Figure 4A and B).
On the other hand, the negative control (Figure 4C) showed
extensive fungal mycelial expansion. The lower activity of 3 may
be due to a number of reasons, including less efficient uptake by
the fungal cells and/or less effective intracellular hydrolysis by the
fungal enzyme.
Chemical Structure Determination of 4′′-epi-Validamycin
A. For a complete characterization of 4′′-epi-validamycin A (3), a
large-scale enzyme reaction of purified ValG with validoxylamine
A (1) and UDP-galactose was carried out and the product was
purified by ion-exchange chromatography. The high-resolution mass
spectrum of 3 revealed a pseudomolecular ion m/z 498 [M + H]+,
1
which is identical to that of validamycin A (2). However, the H
and 13C NMR spectra of 3 are slightly different from those of 2.14
In particular, the C-4′′ proton of 3 appeared as a small doublet (J
) 3 Hz), as opposed to the large doublet (J ) 12 Hz) of 2 (Figure
3), which confirmed the identity of a galactosyl moiety in 3. On
the other hand, the coupling constant of the anomeric proton (H-
1′′) of 3 is 7.5 Hz, which is consistent with that of 2, indicating
a ꢀ-glycosidyl linkage of galactose. Furthermore, a correlation
between H-1′′ (δH 4.42 ppm) and C-4′ (δC 83.8 ppm) in the HMBC
analysis provides evidence that the sugar moiety attaches at the
C-4′ position of 3, which is consistent with the regioselectivity of
Site-Directed Mutagenesis of ValG. BLAST analysis of ValG
showed that it is a member of the glycosyltransferase family 2 (GT-
2), classified by the CAZy system. This family is the largest and
evolutionarily the most ancient of the inverting enzyme families
with over 8000 members to date, and it utilizes a vast array of
potential sugar donors and acceptors. The SpsA from Bacillus
subtilis is one of the few well-studied glycosyltransferases with
crystal structures in this most widespread family.17 The N-terminal
domains of this family often display a much higher similarity since

Enzymatic Production of 4′′-epi-Validamycin A
Journal of Natural Products, 2008, Vol. 71, No. 7 1235
Figure 4. Fungal growth inhibitory assays of validamycin A (A), 4′′-epi-validamycin A (B), and the negative control, water (C).
Figure 5. (A) SDS-PAGE of ValG and ValG mutant: M, protein
marker; I, cell-free extract of E. coli harboring ValG; II, purified
ValG; III, cell-free extract of E. coli harboring mutated ValG; IV,
purified ValG mutant. (B) TLC analysis (nPrOH/AcOH/H2O, 4:1:
1) of ValG reactions: a, validoxylamine A; b, glycosylation product
of validoxylamine A with UDP-galactose; c, glycosylation product
of validoxylamine A with UDP-glucose.
they bind a common nucleotidyl moiety, and the C-terminal
domains show enormous diversity probably because of the variety
of sugar acceptors. Alignment of ValG showed that the N-terminal
domain (amino acid residues 12-110) has high homology with the
sugar donor-binding domain of SpsA, and the C-terminal domain
shows very low similarity to other proteins.
Structurally, ValG belongs to the GT-A superfamily. A general
Figure 6. HPLC analyses of the reaction products of ValG (A)
and mutated ValG (B) with different divalent metal ions.
feature of all enzymes of the GT-A fold family is the presence of
A comparative study between the native and mutated ValG
a common DXD motif, which interacts with a divalent metal ion
revealed that the native DTG-bearing ValG efficiently used a
and/or an NDP-sugar. In the DXD motif, the third aspartate residue
number of divalent metal ions, e.g., Mn2+, Mg2+, Co2+, and Ca2+
,
binds the divalent metal ion, typically Mn2+, and the middle amino
acid residue binds the hydroxyl groups on the ribose moiety of the
nucleotide moiety. Alignments of the amino acid sequence of ValG
with its homologous proteins revealed that instead of having the
common DXD motif at the position 99-101, ValG contains a DTG
sequence. Using Hydrophobic Cluster Analysis (HCA),18 which is
commonly used for predicting structural similarities in protein
sequences with low levels of identity, we were able to predict that
the DTG sequence in ValG is indeed located at the DXD motif
position in other glycosyltransferases. To investigate the role of
this DTG sequence, as opposed to the common DXD motif, the
D99/T100/G101 motif was replaced with DCD residues by site-directed
mutagenesis. An overlapping PCR method was used for this
experiment, and the mutated gene was confirmed by DNA sequenc-
ing. The original DNA sequence ACTGGT encoding threonine (T)
and glycine (G) was replaced with TGCGAC for cysteine (C) and
aspartic acid (D).
Effects of DTG f DCD Mutation on ValG Catalytic
Activity. Recombinant wild-type and mutant forms of ValG were
produced in E. coli BL21 Gold(DE3)pLysS, purified by nickel
columns, and tested for their catalytic activity using the suite of
different sugar donors described above. While the enzyme main-
tained its catalytic activity with UDP-glucose and UDP-galactose
(Figure 5), no other NDP-sugars were recognized as alternative
substrates. On the other hand, compared to the native ValG, the
mutated enzyme appeared to have different preferences toward
divalent metal ions.
as cofactor, whereas the DCD-bearing ValG could effectively use
only Mn2+, and to some extent Co2+, as cofactor (Figure 6). The
result is consistent with the nature of other DXD-containing
glycosyltransferases, which mostly require Mn2+ for their catalytic
activities. The DTG sequence of ValG may provide greater cofactor
flexibility for the enzyme; however, it is not clear as to what extent
this would benefit the host bacteria in relation to validamycin A
biosynthesis.
Experimental Section
General Experimental Procedures. [R]D was recorded on a P-1010
polarimeter (JASCO Corporation). 1H and 13C NMR were recorded on
a Bruker DPX300 NMR spectrometer. Low-resolution electrospray
ionization (ESI) mass spectra were recorded on a ThermoFinnigan liquid
chromatograph-ion trap mass spectrometer. High-resolution electrospray
mass spectra were recorded on Waters/Micromass LCT spectrometer.
Reactions were monitored by TLC (silica gel 60 F254, Merck) with
detection by use of Ce(SO4)2 solution. Column chromatography was
performed on Dowex 50Wx8-200 (Fluka). All chemicals were pur-
chased from Aldrich or Sigma and were used without further purification
unless otherwise noted. Reversed-phase preparative HPLC was carried
out on a C18 column (Atlantis C18 5 mm column, Waters) with CH3CN/
KH2PO4 buffer (pH 6.92, 1 mM) (3:97) as mobile phase at 0.6 mL/
min. Cloning and overexpression of the ValG gene and purification of
the recombinant His6-tagged ValG were carried out as described
previously.11
Enzyme Assays. The enzyme assays were carried out at 30 °C for
3-12 h in a 100 µL volume of 25 mM Tris-HCl (pH 7.5), 10 mM

1236 Journal of Natural Products, 2008, Vol. 71, No. 7
Xu et al.
MgCl2, 20 mM NH4Cl, 15 mM UDP-sugar, 10 mM validoxylamine
A, and 50 µL of purified ValG solution (1 mg/mL) or 50 µL of the
cell-free extract of E. coli BL21(DE3)pLysS containing ValG. The
reaction progress was monitored by TLC analysis. For a scale-up
experiment, 5.71 mg of validoxylamine A was used. To terminate the
reaction, EtOH was added to precipitate the protein. The supernatant
was subjected to ion-exchange (Dowex 50Wx8-200 [H+ form]) column
chromatography. The column was washed with H2O and then eluted
with 0.5 M aqueous NH4OH. The elution of the product was monitored
by TLC. Fractions containing the desired product [Rf value 0.15 (nPrOH/
AcOH/H2O, 4:1:1)] were pooled and lyophilized. The product was
subsequently subjected to preparative HPLC to give 3.6 mg of pure
4′′-epi-validamycin A (3). 3: white powder, [R]23D +49.2 (c 0.07, H2O);
1H NMR (D2O, 300 MHz) δ 6.01 (1H, d, J ) 5 Hz, H-6), 4.42 (1H,
d, J ) 7.5 Hz, H-1′′), 4.21 (1H, d, J ) 14 Hz, H-7b), 4.08 (1H, d, J
) 14 Hz, H-7a), 4.05 (1H, brd, J ) 7.0 Hz, H-4), 3.89 (1H, d, J ) 3
Hz, H-4′′), 3.76-3.68 (6H, overlap, H-2′, H-7′a,b, H-5′′, H-6′′a,b), 3.60
(1H, brd, J ) 3 Hz, H-3′′), 3.58-3.56 (3H, overlap, H-2, H-3, H-2′′),
3.52 (1H, d, J ) 2.5 Hz, H-3′), 3.48 (1H, brd, J ) 2.5 Hz, H-4′), 3.34
(1H, m, H-1), 3.25 (1H, m, H-1′), 2.07 (1H, m, H-5′), 1.91 (1H, brd,
J ) 15 Hz, H-6′b), 1.29 (1H, brd, J ) 15 Hz, H-6′a); 13C NMR (D2O,
75 MHz) δ 139.1 (C-5), 123.1 (C-6), 103.3 (C-1′′), 83.8 (C-4′), 75.3
(C-5′′), 73.6 (C-3), 73.1 (C-2′), 72.7 (C-3′′), 72.6 (C-2), 71.4 (C-4),
71.2 (C-3′), 69.4 (C-2′′), 68.5 (C-4′′), 61.7 (C-7′), 61.5 (C-7), 60.9 (C-
6′′), 53.7 (C-1′), 52.3 (C-1), 37.4 (C-5′), 26.8 (C-6′); HRESIMS m/z
498.2169 [M + H]+ (calcd for C20H36NO13, 498.2187).
Eppendorf tube and dried. The product was then dissolved in 100 µL
of H2O for HPLC analysis.
Bioassay. A 1% agar solution (14 mL) was mixed with the
compounds (dissolved in 1 mL of H2O) and plated into Petri dishes.
PDA agar plugs with mycelia of P. sasakii were placed in the center
of the plates as the indicator strain for bioassay of the compounds.
The plates were incubated at 30 °C for 2 days.
Acknowledgment. The authors thank P. M. Flatt for the critical
reading of the manuscript. This work was supported by grants from
the National Institutes of Health (R01 AI061528), the Herman Frasch
Foundation, and, in part, the Mass Spectrometry Facilities and Services
Core of the Environmental Health Sciences Center, Oregon State
University, grant number P30 ES00210, National Institute of Environ-
mental Health Sciences, National Institutes of Health. H.X. was in part
supported by the Exchange Program from China Scholarship Council.
Supporting Information Available: Table 1 and Figures S1-S8.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References and Notes
(1) Michels, P. C.; Khmelnitsky, Y. L.; Dordick, J. S.; Clark, D. S. Trends
Biotechnol. 1998, 16, 210–215.
(2) Altreuter, D. H.; Clark, D. S. Curr. Opin. Biotechnol. 1999, 10, 130–
136.
Site-Directed Mutagenesis of WalG. The mutagenesis in ValG was
introduced by an overlapping PCR method. For the first step, two
fragments of ValG were amplified separately by PCR with Platinum
Pfx DNA polymerase (Invitrogen) using the cosmid clone 17F2 as
template. A 323 bp DNA encoding the N-terminal fragment of ValG
(fragment 1) was amplified using primers ValGPM-1F2 (5′-AAAG-
GATCCACATATGCCCGGTGCGCATCCC-3′; the BamHI and NdeI
sites are in italics) and ValGPM-1R2 (5′-TACTGCGGCCCCGCCAG-
GACGTCGCAGTCGAGGAAGGCCAGGAGTG-3′; the introduced
mutation sites are underlined). Similarly, a 992 bp DNA encoding the
C-terminal fragment of ValG (fragment 2) was amplified using primers
(3) Woodley, J. M. Trends Biotechnol. 2008, Apr 22 [Epub ahead of print]
doi:10.1016/j.tibtech.2008.03.004.
(4) Weymouth-Wilson, A. C. Nat. Prod. Rep. 1997, 14, 99–110.
(5) Dembitsky, V. M. Chem. BiodiVersity 2004, 1, 673–781.
(6) Luzhetskyy, A.; Mendez, C.; Salas, J. A.; Bechthold, A. Curr. Top.
Med. Chem. 2008, 8, 680–709.
(7) Bourne, Y.; Henrissat, B. Curr. Opin. Struct. Biol. 2001, 11, 593–
600.
(8) Jiang, J.; Biggins, J. B.; Thorson, J. S. Angew. Chem., Int. Ed. 2001,
40, 1502–1505.
(9) Barton, W. A.; Biggins, J. B.; Jiang, J.; Thorson, J. S.; Nikolov, D. B.
Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 13397–13402.
(10) Qian, X.; Li, Z.; Liu, Z.; Song, G.; Li, Z. Carbohydr. Res. 2001, 336,
79–82.
(11) 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, 387–397.
(12) Breton, C.; Imberty, A. Curr. Opin. Struct. Biol. 1999, 9, 563–571.
(13) Wiggins, C. A.; Munro, S. Proc. Natl. Acad. Sci. U.S.A. 1998, 95,
7945–7950.
(14) Jin, W. Z.; Rinehart, K. L., Jr.; Toyokuni, T. J. Antibiot. 1987, 40,
329–339.
(15) Asano, N.; Yamaguchi, T.; Kameda, Y.; Matsui, K. J. Antibiot. 1987,
40, 526–532.
ValGPM-2F2
(5′-CACTCCTGGCCTTCCTCGACTGCGACGTC-
CTGGCGGGGCCGCAGTA-3′; the mutation sites are underlined) and
ValGPM-2R2 (5′-AAAGAATTCAGTCACCGCGAAGAGAC-3′; the
EcoRI site is in italics). For the second step, the whole modified ValG
was amplified using fragments 1 and 2 as templates with primers
ValGPM-1F2 and ValGPM-2R2. Gel-purified PCR products were
digested with BamHI and EcoRI and subsequently ligated into BamHI/
EcoRI-digested pRSET-B. The constructs were transferred into E. coli
DH10B and plated on LB agar plates containing 100 µg/mL ampicillin.
The expected mutation was confirmed by sequencing.
(16) Minagawa, K.; Zhang, Y.; Ito, T.; Bai, L.; Deng, Z.; Mahmud, T.
ChemBioChem 2007, 8, 632–641.
(17) Tarbouriech, N.; Charnock, S. J.; Davies, G. J. J. Mol. Biol. 2001,
314, 655–661.
(18) Gaboriaud, C.; Bissery, V.; Benchetrit, T.; Mornon, J. P. FEBS Lett.
1987, 224, 149–155.
Enzyme Assays with Different Metal Ions. Enzyme assays of
mutated ValG with different metal ions were carried out at 30 °C for
3 h in a 100 µL volume of 25 mM Tris-HCl (pH 7.5), 1 mM divalent
metal ion, 20 mM NH4Cl, 15 mM UDP-sugars, 10 mM validoxylamine
A, and 50 µL of protein solution (0.1 mg/mL). Subsequently, 100 µL
of MeOH was added into the reaction, and the mixture was centrifuged
for 5 min at 12 000 rpm. The supernatant was transferred into a clean
NP800185K