C O M M U N I C A T I O N S
oxygen functionality turned out to occupy almost the identical
position as that of the amide bond of 3 on the hydrophobic scaffold.
Thus, such a hydrogen-bonding interaction appears to be significant
in the substrate recognition by VinC. This notion was supported
by the rather poor reactivity of 10, probably due to the lack of
hydrogen bonding by the substitution at C-17 with a hydrophobic
alkyl side chain. These results appear to imply that a variety of
glycosyl acceptors can be designed to fulfill these structural
demands. This turned out to be the case since a synthesized racemic
mimic (21) and its further simplified alcohol (22)13 were in fact
reactive in the VinC reaction to form their corresponding vice-
nisaminides 23 ([M + H]+ ) 517, 95% conversion by LC-MS
analysis) and 24 ([M + H]+ ) 359), respectively, although
conversion of 14 was not sufficiently high enough (less than 5%
conversion).
Figure 3. Superimposed view of vicenilactam 3 with (A) mimic 21, (B)
brefeldin A (5), (C) R-zearalenol (6), (D) â-estradiol (8), (E) pregnenolone
(9), and (F) 3-O-acetyl-â-estradiol (11).
In summary, the present studies have successfully demonstrated
for the first time significant potential of an enzymatic approach to
unnatural glycosides having diverse aglycon structures apart from
those found in nature. Particularly, VinC was shown to be useful
for the preparation of hydrophobic glycosides with different aglycon
scaffolds.
glycosyl acceptor. Transglycosylation to cholesterol 10 turned out
to be extremely uneffective, and no glycoside formation was
observed at all in the reactions with 12 and 13.
All the vicenisaminide products (14-20) were isolated by large-
scale enzyme reactions and characterized chemically on the basis
of spectroscopic analysis. The glycoside bonds were all in â-form
as determined by 1H NMR spin-spin coupling constants of
anomeric positions (14, δH 4.63, J ) 9.6 and 2.0 Hz; 15; δH 4.76,
J ) 9.6 and 2.0 Hz; 16, δH 4.87, J ) 9.6 and 2.0 Hz; 17, δH 4.82,
J ) 9.6 and 2.0 Hz; 18, δH 5.42, J ) 9.2 and 2.0 Hz; 19, δH 4.84,
J ) 9.6 and 2.4 Hz; 20, δH 4.97, J ) 9.2 and 1.6 Hz; vicenistatin,
δH 5.29, J ) 9.5, 3.0 Hz). The site of glycosylation of each
vicenisaminide was determined as follows, based on the observation
Acknowledgment. This work was supported in part by Grants-
in-Aid for Scientific Research from JSPS and a 21st Century COE
program from MEXT.
Supporting Information Available: Experimental procedures and
characterization data for new compounds (PDF). This material is
References
(1) Weymouth-Wilson, A. C. Nat. Prod. Rep. 1997, 14, 99-110.
(2) (a) Me´ndez, C.; Salas, J. A. Trends Biotechnol. 2001, 19, 449-456 and
references therein. (b) Yamase, H.; Zhao, L.; Liu, H.-w. J. Am. Chem.
Soc. 2000, 122, 12397-12398.
1
of chemical shift changes in H NMR (15, C-7, ∆δH-7 0.07 ppm,
and ∆δH-4 0 ppm; 16, C-6′, ∆δH-6′ 0.13 ppm; 17, C-6′, ∆δH-6′
0.05 ppm). The site of glycosylation in 18 was also determined to
be at the C-3 position because of apparent chemical shift changes
among the A-ring protons: ∆δH-1 0.12 ppm, ∆δH-2 0.22 ppm,
∆δH-4 0.21 ppm, and ∆δH-17 0.01 ppm. Glycosylation to the C-17
position of â-estradiol was also successfully achieved using 3-O-
acetyl-â-estradiol (11) as a glycosyl acceptor.
It appears therefore that VinC can accommodate various
molecules as glycosyl acceptors, including structurally related
compounds, such as 4, but structurally different molecules, such
as 5-9 and 11, as well, while 10 in a lesser extent. In other words,
even structurally unrelated alcohols to the natural aglycon can be
transglycosylated by certain glycosyltransferase as exemplified by
VinC, and diverse glycosides of structural and biological interest
may well be developed through the present approach.
To figure out the determinative elements of aglycon recognition
by VinC, in addition to apparent hydrophobicity, three-dimensional
structures of the accepted aglycons were estimated and compared
with naturally derived vicenilactam 3. The structures of all accepted
aglycons, except for 5, were deduced from the molecular mechanics
calculation,10 and the structure of 5 was adopted from its crystal
structure (Figure 3).12 Interesting structural features have been borne
out. First of all, gross molecular size may be important since the
sizes of all accepted aglycons appeared to be almost the same as
3. Second, the spatial arrangement of a few polar groups may also
be significant. When the glycosyl accepting hydroxy groups of these
molecules were superimposed in their 3D structures, the additional
(3) Thorson, J. S.; Barton, W. A.; Hoffmeister, D.; Albermann, C.; Niklov,
D. B. Chembiochem 2004, 5, 16-25.
(4) Losey, H. C.; Jiang, J.; Biggins, J. B.; Oberthu¨r, M.; Ye, X.-Y.; Dong, S.
D.; Kahne, D.; Thorson, J. S.; Walsh, C. T. Chem. Biol. 2002, 9, 1305-
1314.
(5) Feel Mayers, C. L.; Oberthu¨r, M.; Anderson, J. W.; Kahne, D.; Walsh,
C. T. Biochemistry 2003, 42, 4179-4189.
(6) Tang, L.; McDaniel, R. Chem. Biol. 2001, 8, 547-555.
(7) Shindo, K.; Kamishohara, M.; Okagawa, A.; Matsuoka, M.; Kawai, H. J.
Antibiot. 1993, 46, 1076-1081.
(8) Ogasawara, Y.; Katayama, K.; Minami, A.; Otsuka, M.; Eguchi, T.;
Kakinuma, K. Chem. Biol. 2004, 11, 79-86.
(9) In a typical assay, a reaction mixture containing dTDP-vicenisamine (0.9
mM), a glycosyl acceptor in DMSO (0.2 mM for final concentration),
and VinC (0.6 mM) was incubated on a shaker at 28 °C for 1.5 hr. The
reaction was terminated by the addition of ethyl acetate. The organic layer
was separated, concentrated, and redissolved in 30 µL of methanol. The
sample was analyzed with a LCQ mass spectrometer (Finnigan) coupled
to a NANOSPACE HPLC (SHISEIDO) equipped with a RP-18 GP
column (KANTO). HPLC conditions were as follows. After injection of
a sample, the column was first washed with 10% MeOH and 0.1% TFA
in water for 10 min, and was then eluted with 90% MeOH and 0.1%
TFA in water, at a flow rate 50 µL/min. Elution was monitored with a
NANOSPACE SI-1 UV detector (SHISEIDO). The percent conversion
ratio was calculated by using eq 1, where AP represents the integration of
the product peak, and AT represents the integration of the unreacted aglycon
peak; % conversion ) [AP/(AP + AT)] × 100 (eq 1).
(10) Conformational search and energy minimizations were performed using
the MM2 force field in Macromodel version 6.5.11
(11) Macromodel, version 6.5; Department of Chemistry, Columbia University,
New York, 10027.
(12) Weber, H. P.; Hauser, D.; Sigg, H. P. HelV. Chim. Acta 1971, 54, 293-
294.
(13) A part of this study was presented in the annual meeting of The Chemical
Society of Japan: Uchida, R.; Nakayama, T.; Matsushima, Y.; Eguchi,
T.; Kakinuma, K. The Chemical Society of Japan, 81st. Spring Meeting,
Tokyo; Abstract, 4A2-08, 2002.
JA042848J
9
J. AM. CHEM. SOC. VOL. 127, NO. 17, 2005 6149