Biosynthetic studies on the α-glucosidase inhibitor acarbose: The chemical synthesis of isotopically labeled 2-epi-5-epi-valiolone analogs

Article abstract of DOI:10.1016/S0008-6215(03)00315-X

2-epi-5-epi-Valiolone is a cyclization product of the C₇ sugar phosphate, sedoheptulose 7-phosphate, involved in the biosynthesis of the aminocyclitol moieties of acarbose, validamycin, and pyralomicin. As part of our investigation into the pat

Full text of DOI:10.1016/S0008-6215(03)00315-X

Carbohydrate Research 338 (2003) 2075ꢀ2082
/
www.elsevier.com/locate/carres
Biosynthetic studies on the a-glucosidase inhibitor acarbose:
the chemical synthesis of isotopically labeled 2-epi-5-epi-valiolone
analogs
Kenji Arakawa, Simeon G. Bowers, Benjamin Michels, Vu Trin, Taifo Mahmud*,$
Department of Chemistry, University of Washington, Box 351700, Seattle, WA 98195-1700, USA
Received 30 April 2003; accepted 9 June 2003
Abstract
2-epi-5-epi-Valiolone is a cyclization product of the C7 sugar phosphate, sedoheptulose 7-phosphate, involved in the biosynthesis
of the aminocyclitol moieties of acarbose, validamycin, and pyralomicin. As part of our investigation into the pathway from 2-epi-5-
epi-valiolone to the valienamine moiety of acarbose, we prepared 1-epi-5-epi-(6-2H2)valiolol [(6-2H2)-6], 5-epi-(6-2H2)valiolol
[(6-2H2)-17], 1-epi-2-epi-5-epi-(6-2H2)valiolol [(6-2H2)-12] and 2-epi-5-epi-(6-2H2)valiolamine [(6-2H2)-11]. Compounds (6-2H2)-6
and (6-2H2)-17 were synthesized from 2,3,4,6-tetra-O-benzyl-D-glucopyranose in 10 and seven steps, respectively, whereas (6-2H2)-12
and (6-2H2)-11 were synthesized from 2,3,4,6-tetra-O-benzyl-D-mannopyranose in eight and 10 steps, respectively.
# 2003 Elsevier Ltd. All rights reserved.
Keywords: 2-epi-5-epi-Valiolone analogs; Acarbose biosynthesis; Valiolol; Valiolamine
1. Introduction
the biosynthesis of validamycin A (2) by Streptomyces
hygroscopicus var limoneus, the 2-epi-5-epi-valiolone is
epimerized at the C-2 stereocenter to give 5-epi-valio-
lone followed by dehydration between C-5 and C-6 to
yield valienone.7,10 However, feeding experiments with
(6-2H2)-5-epi-valiolone and (1-13C)valienone to the
acarbose (1) producer revealed that neither of these
ketocyclitols was incorporated into the valienamine
moiety of 1.6 Further studies using isotopically labeled
2-epi-valiolone, 2-epi-valienone, as well as valiolamine
and valienamine yielded similar results, leaving the
pathway from 2-epi-5-epi-valiolone to the valienamine
moiety of 1 purely conjectural.
One explanation that we previously proposed was
that the transformation of 2-epi-5-epi-valiolone (4) to
the valienamine moiety of 1 involves a substrate-
channeling mechanism in which enzyme-bound inter-
mediates are directly transferred from one enzyme active
site to the next in a multi-enzyme complex. However,
while this proposal is mechanistically attractive, there is
no hard evidence to support it. On the other hand,
several alternatives are certainly also plausible, for
example, one cannot rule out the possibility that the
nitrogen is introduced into the cyclitol moiety in an
The a-glucosidase inhibitor acarbose and related com-
pounds are members of the C7N aminocyclitol family of
natural products that are increasingly gaining recogni-
tion due to their significant biomedical and agricultural
uses.1ꢀ4 Their chemical structures most commonly
contain an unsaturated aminocyclitol moiety, the so
called valienamine moiety (Fig. 1). In our earlier work
on the biosynthesis of acarbose (1),5,6 validamycin A
(2),5,7 and pyralomicin (3)8 we demonstrated that the
valienamine moiety of these compounds is derived from
2-epi-5-epi-valiolone (4), an intramolecular aldol cycli-
zation product of sedoheptulose 7-phosphate. The
cyclization reaction has been proposed to occur via a
dehydroquinate (DHQ) synthase-like mechanism.9 In
* Corresponding author. Tel.: ꢁ
/
1-206-5439950; fax: ꢁ
/
1-
206-5438318.
E-mail address:
taifo.mahmud@oregonstate.edu (T.
Mahmud).
$
Current address: College of Pharmacy, Oregon State
University, 203 Pharmacy Building, Corvallis, OR 97331-3507,
USA. Tel.: ꢁ
/
1-541-737-9679; fax: ꢁ1-541-737-3999.
/
0008-6215/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0008-6215(03)00315-X

2076
K. Arakawa et al. / Carbohydrate Research 338 (2003) 2075ꢀ2082
/
Fig. 1. Chemical structures of acarbose (1), validamycin A (2), and pyralomicin 1a (3).
earlier step (Scheme 1). Alternatively, the coupling may
occur by an SN2 displacement of an activated hydroxyl
group on the cyclitol by the nitrogen of the amino sugar
moiety.11 The most plausible intermediate for this
mechanism was 1-epi-valienol, which is found in the
acarbose production culture at the early stage of
fermentation, qualifying it as a biosynthetic intermedi-
ate. However, when 1-epi-(1-13C)valienol was synthe-
sized and fed, no incorporation into acarbose was
observed.6 Thus, if this mode of coupling does operate
in 1 biosynthesis, the coupling must occur at an earlier
stage, that of 1-epi-5-epi-valiolol or 1-epi-2-epi-5-epi-
valiolol (Scheme 1). On such a hypothetical pathway,
the coupling product might undergo epimerization and
dehydration catalyzed by flavin-containing redox en-
zyme which transiently generates an imine in its active
site. In the present paper, we report the synthesis of
isotopically labeled 1-epi-5-epi-valiolol (6), 5-epi-valio-
lol (17), 1-epi-2-epi-5-epi-valiolol (12), and 2-epi-5-epi-
valiolamine (11). The examination of their possible
involvement in the biosynthesis of acarbose is also
discussed.
(4.01 ppm, d, J 2.6 Hz) and H-2 (3.48 ppm, dd, J 2.6
and 7.8 Hz), as well as the strong nOe correlation
between H-1 and H-2 in (6-2H2)-16. Large coupling
constants between H-2 and H-3 (3.69 ppm, dd, J 7.8, 8.3
Hz) and between H-3 and H-4 (3.50 ppm, d, J 8.3 Hz)
suggest that (6-2H2)-16 adopts a chair conformation.
The stereoselectivity of the borohydride reduction of
ketone (6-2H2)-15 is most likely due to steric hindrance,
especially by the C-2 benzyloxyl group, that allows the
nucleophilic attack to proceed only from the b face. A
small quantity of b-alcohol was also isolated, but all
attempts to increase the proportion of the b-alcohol
using different reducing agents and reaction conditions
were unsuccessful. It was therefore necessary to invert
the stereochemistry at the 1-position by acetate displa-
cement of a C-1 tosylate. Thus, a-alcohol (6-2H2)-16 was
reacted with p-toluenesulfonyl chloride in pyridine to
give tosylate (6-2H2)-18 in 82% yield. Subsequent
reaction with KOAc in DMF afforded acetate (6-2H2)-
19 along with elimination product (6-2H2)-20. The
mixture of (6-2H2)-19 and (6-2H2)-20 was reacted with
NaOMeꢀ
MeOH to give alcohol (6-2H2)-21. A large
coupling constant (J 8.8 Hz) between H-1 and H-2 of
/
alcohol (6-2H2)-21 confirms the b-OH configuration.
Finally, debenzylation of (6-2H2)-21 with PdÃ
/
C under
2. Results and discussion
an atmosphere of hydrogen gave 1-epi-5-epi-(6-2H2)va-
liolol [(6-2H2)-6] in 96% yield. Likewise, a-alcohol
(6-2H2)-16 was debenzylated under identical conditions
to give 5-epi-(6-2H2)valiolol [(6-2H2)-17].
2.1. Synthesis of 1-epi-5-epi-(6-2H2)valiolol [(6-2H2)-6]
and 5-epi-(6-2H2)valiolol [(6-2H2)-17]
1-epi-5-epi-(6-2H2)Valiolol [(6-2H2)-6] and 5-epi-
(6-2H2)valiolol [(6-2H2)-17] were prepared from 2,3,4,7-
2.2. Synthesis of 1-epi-2-epi-5-epi-(6-2H2)valiolol [(6-
2H2)-12] and 2-epi-5-epi-(6-2H2)valiolamine [(6-2H2)-11]
tetra-O-benzyl-5-epi-(6-2H2)valiolone
[(6-2H2)-15]
(Scheme 2). The latter compound was generated from
2,3,4,6-tetra-O-benzyl-D-glucopyranose (14) in five
steps.10 Reduction of the ketone (6-2H2)-15 with sodium
borohydride gave predominantly the undesired cyclitol
(6-2H2)-16 in 86% yield. The stereochemistry at C-1 was
determined based on the coupling constant between H-1
2,3,4,7-Tetra-O-benzyl-2-epi-5-epi-(6-2H2)valiolone [(6-
2H2)-23], prepared from commercially available 2,3,4,6-
tetra-O-benzyl-D-mannopyranose,6 was used for the
synthesis of 1-epi-2-epi-5-epi-(6-2H2)valiolol [(6-2H2)-
12] and 2-epi-5-epi-(6-2H2)valiolamine [(6-2H2)-11]

K. Arakawa et al. / Carbohydrate Research 338 (2003) 2075ꢀ
/2082
2077
Scheme 1. Originally considered possible pathways from 2-epi-5-epi-valiolone (4) to dTDP-acarbiose (13), the precursor of
acarbose (1).
(Scheme 3). Compound (6-2H2)-23 was treated with
methanol to give predo-
comparison of its NMR data with those of racemic
compound 12 and its 1-epimer, synthesized from p-
benzoquinone12 and provided by Dr Oliver Block.
The synthesis of 2-epi-5-epi-(6-2H2)valiolamine
[(6-2H2)-11] also proceeded from alcohol (6-2H2)-24.
The latter compound was tosylated to afford (6-2H2)-25
in 90% yield. Displacement of the C-1 tosylate with
azide was surprisingly difficult, given that the stereo-
configurations of the C-1 tosyloxyl and C-2 benzyloxyl
moieties would usually allow an SN2 substitution to
proceed from the a face. After many attempts, (6-2H2)-
sodium borohydride in THFꢀ
/
minantly the desired alcohol (6-2H2)-24 in 86% yield.
This stereoselectivity may be due to the strong steric
effect of the benzyloxyl group at C-2 that shields the b
face and directs the hydride attack at C-1 to proceed
exclusively from the a face.
Alcohol (6-2H2)-24 was subsequently subjected to
catalytic hydrogenation to afford (6-2H2)-12 in quanti-
tative yield. The stereochemistry at C-1 in (6-2H2)-12
was concluded to be as shown in Scheme 3 based on a

2078
K. Arakawa et al. / Carbohydrate Research 338 (2003) 2075ꢀ2082
/
Scheme 2. The synthesis of 1-epi-5-epi-(6-2H2)valiolol [(6-2H2)-6] and 5-epi-(6-2H2)valiolol [(6-2H2)-17]. Reagents: (a) NaBH4/
THFꢀEtOH, 86%; (b) H2, 10% PdꢀC/95% EtOH, 96%; (c) pTsCl, DMAP/pyridine, 82%; (d) KOAc/DMF; (e) NaOMe/methanol,
34% in 2 steps.
/
/
25 was eventually treated with a large excess of sodium
azide in the presence of crown ether13 and the reaction
was carried out under vigorous reaction conditions
(90 8C for 52 h) to furnish azidocyclitol (6-2H2)-26 in
32% yield. Further extension of the reaction time or
increasing the temperature leads to extensive decom-
position. Similar conditions applied to the triflate
derivative of alcohol (6-2H2)-24 did not give the desired
product. Finally, azide (6-2H2)-26 was subjected to
catalytic hydrogenation with 20% Pd(OH)2 on charcoal
to afford (6-2H2)-11 as an amorphous solid in quanti-
tative yield.
2.3. Feeding experiments in resting cells with labeled
compounds
The cyclitols (6-2H2)-6, (6-2H2)-12, and (6-2H2)-17, as
well as the aminocyclitol (6-2H2)-11 (15 mg each) were
fed to resting cell cultures of Actinoplanes sp. SN223/29
(60 mL each) along with (U-13C3)glycerol and 2-epi-5-
Scheme 3. The synthesis of 1-epi-2-epi-5-epi-(6-2H2)valiolol [(6-2H2)-12] and 2-epi-5-epi-(6-2H2)valiolamine [(6-2H2)-11]. Reagents:
(a) NaBH4/THFꢀEtOH, 86%; (b) H2, 10% PdꢀC/95% EtOH, quantitative; (c) pTsCl, DMAP/pyridine, 90%; (d) NaN3, 18-crown-6/
DMF, 32% (38% recovered); (e) H2, 20% Pd(OH)2ꢀC/95% EtOH, quantitative.
/
/
/

K. Arakawa et al. / Carbohydrate Research 338 (2003) 2075ꢀ
/2082
2079
epi-(6-2H2)valiolone as positive controls. After 24 h
incubation, the resulting acarbose was isolated and
purified as described in the previous paper.6 Analysis
of the acarbose by ESI-MS indicated that except the
positive controls, none of the cyclitols was significantly
incorporated into acarbose. The non-incorporation of 1-
epi-2-epi-5-epi-(6-2H2)valiolol [(6-2H2)-12] rules out the
possibility of an immediate reduction of 2-epi-5-epi-
valiolone (4) prior to a SN2 coupling between the
alcohol (or its activated form) with dTDP-4-amino-
from their unphosphorylated counterparts. A similar
instance, where none of the plausible intermediates,
except 2-epi-5-epi-valiolone and to some extent 5-epi-
valiolone, was incorporated has also been observed in
the biosynthesis of pyralomicin 1a.8
3. Experimental
3.1. General
4,6-dideoxy-D
-glucose11 (Scheme 1). Likewise, the non-
incorporation of 1-epi-5-epi-(6-2H2)valiolol [(6-2H2)-6]
and its C-1 isomer, 5-epi-(6-2H2)valiolol [(6-2H2)-17],
excludes the possibility of 4 being epimerized at C-2 to
give 5-epi-valiolone (5) and subsequently reduced to 6
prior to the coupling reaction. Evidence suggesting that
5 is not directly involved in the biosynthesis of acarbose
has also been obtained in previous feeding experiments
with 5-epi-(6-2H2)valiolone to the acarbose producer,
which yielded a negative incorporation result (T.
Mahmud, J. Xu, Y. U. Choi, unpublished data).
After none of the possible cyclitols used in this study
nor the ketocyclitols described in our previous report6
(except 2-epi-5-epi-valiolone) was incorporated into
acarbose, only a very short list of compounds could
still be considered as plausible intermediates. Although
a reduction of 2-epi-5-epi-valiolone (4) in theory will
give either 1-epi-2-epi-5-epi-valiolol (12) or its 1-epimer,
2-epi-5-epi-valiolol, the involvement of the latter com-
pound in acarbose biosynthesis is less likely since a
coupling between 2-epi-5-epi-valiolol or its activated
form with dTDP-4-amino-4,6-dideoxy-D-glucose will
lead to the opposite C-1 stereochemistry in the product
as in acarbose. Although this inverted stereochemistry
could be corrected by a further imine formation, which
will also facilitate the 2-epimerization and 5,6-dehydra-
tion to give dTDP-acarbiose, such inefficiency is unu-
sual in a biological system. In addition, it is evident that
hydride reduction of the protected 2-epi-5-epi-valiolone
from the pro-S face to give protected 2-epi-5-epi-valiolol
is rather difficult, presumably due to steric hindrance,
which most likely will also be the case if the reduction of
2-epi-5-epi-valiolone has to occur in the biological
system.
The H, H, and 13C NMR spectra were recorded on
Bruker AF-300 or AM500 NMR spectrometers with
MacNMR 5.5 PCI as the instrument controller and data
processor. Optical rotations were measured on a JASCO
DIP-370 digital polarimeter. Low-resolution electro-
1
2
spray mass spectra were recorded on a Brukerꢀ
/
Esquire
Liquid ChromatographꢀIon Trap Mass Spectrometer.
/
A PerSeptive Biosystems Mariner Electrospray Time of
Flight Mass Spectrometer (ESI-TOF) was used for the
high resolution mass spectrometry. A ISF-4-V shaker,
Adolf Kuhner AG was used for the fermentations. All
synthetic reactions were carried out under an atmo-
sphere of dry Ar at room temperature (rt) in oven-dried
glassware unless otherwise noted. Reactions were mon-
itored by TLC (silica gel 60 F254, Merck) with detection
by UV light or by alkaline permanganate spray or
Ce(SO4)2/H2SO4 solution. Column chromatography was
performed on 230ꢀ400 mesh silica gel (Aldrich), Sepha-
/
dex LH-20 (Sigma) or Sephadex CM-25 (Pharmacia).
3.2. Materials
Actinoplanes sp. SN223/29 was obtained from Bayer
AG, Wuppertal, Germany. All chemicals were pur-
chased from Aldrich or Sigma and used without further
purification unless otherwise noted. Ingredients for
fermentations, soybean meal (fat free), maltzin powder,
and yeast extract, were obtained from Bayer AG,
Germany and NZ-Amine A from ICN Biochemicals.
3.3. Culture conditions for feeding experiments in resting
cells
The non-incorporation of all plausible cyclitols that
have been tested suggests the involvement of a rather
unusual mechanism in the biosynthesis of acarbose.
Recently, the group of Piepersberg reported the identi-
fication of a kinase activity in the acarbose producer,
which modifies 2-epi-5-epi-valiolone to its 7-phosphate
derivative.14 The product is then epimerized at the C-2
position to give 5-epi-valiolone 7-phosphate.15 These
results suggest that, the intermediates involved in the
biosynthesis of the valienamine moiety of acarbose are
phosphorylated cyclitols which, except for 2-epi-5-epi-
valiolone 7-phosphate, cannot be generated directly
For the resting cell experiments, production cultures in
complex medium6 were incubated for 28ꢀ
30 h and
/
centrifuged (4000g for 5 min) to remove the super-
natant. The remaining cells were washed twice with cold
100 mM potassium phosphate buffer (pH 6.9) and then
suspended in 50 mM potassium phosphate (pH 6.9)
containing maltose (1.5%) and bactopeptone (0.15%) to
make 75 mL of cell suspension from 50 mL of
production culture. Each 60ꢀ/75 mL of resting cells
suspension in a 500-mL flask was incubated for 22ꢀ
/24 h
on a rotary shaker at 220 rpm and 28 8C with labeled
substrates added at 0 h (1/3) and the other 2/3 at 8 h.

2080
K. Arakawa et al. / Carbohydrate Research 338 (2003) 2075ꢀ2082
/
Isolation and purification procedures of acarbose were
same as described in our previous paper.6
MHz, CDCl3): dC 21.8, 73.0, 73.3, 73.7, 74.7, 75.7, 76.4,
77.9, 78.0, 78.1, 78.3, 127.7ꢀ129.9, 134.1, 137.8, 138.1,
138.3, 138.6, 144.7. Electrospray-MS: m/z 733.4 [Mꢁ
Na]ꢁ. Electrospray-HRMS: m/z 733.2787 [MꢁNa]ꢁ
Calcd for C42H42D2NaO8S: 733.2778.
/
/
3.4. Synthesis of labeled substrates
/
3.4.1. 2,3,4,7-Tetra-O-benzyl-5-epi-(6-2H2)valiolol
[(6-2H2)-16]. 2,3,4,7-Tetra-O-benzyl-5-epi-(6-2H2)valio-
lone [(6-2H2)-15] (926 mg, 1.67 mmol) was dissolved in
3.4.3. 2,3,4,7-Tetra-O-benzyl-1-epi-5-epi-(6-2H2)valiolol
[(6-2H2)-21]. 2,3,4,7-Tetra-O-benzyl-1-O-(p-tolylsulfo-
nyl)-5-epi-(6-2H2)valiolol [(6-2H2)-18] (497 mg, 699
mmol) was dissolved in DMF (50 mL) and KOAc (207
mg, 2.11 mmol), was added. The reaction mixture was
stirred at 60 8C for 48 h after which TLC [eluent 2:1 (v/
THFꢀ
was added. The mixture was stirred for 30 min after
which TLC [eluent 2:1 (v/v) hexaneꢀEtOAc] indicated
/
EtOH (50 mL) and NaBH4 (250 mg, 6.64 mmol)
/
complete conversion to a lower running product. The
solution was diluted with EtOAc (100 mL) and washed
v) hexaneꢀ
the starting material. The solution was diluted with
water (20 mL) and extracted with EtOAc (3ꢂ50 mL),
the organic layer was washed with water (2ꢂ25 mL),
/EtOAc] indicated complete consumption of
with 1 M HCl (2ꢂ
/
50 mL), satd NaHCO3 (3ꢂ50 mL),
/
and brine (50 mL), dried (MgSO4), filtered and the
filtrate concentrated under reduced pressure. The result-
ing syrup was purified by column chromatography
/
/
and brine (25 mL), dried (MgSO4), filtered and the
filtrate concentrated under reduced pressure. The result-
ing residue was purified by column chromatography
[eluent 5:1ꢀ
of the appropriate fractions afforded (6-2H2)-16 (797
mg, 1.43 mmol, 86%) as a clear syrup; [a]D23
1.8 (c 1,
CHCl3); 1H NMR (300 MHz, CDCl3): dH 3.48 (dd, 1 H,
1,2 2.6, J 2,3 7.8 Hz, H-2), 3.50 (d, 1 H, J 3,4 8.3 Hz, H-
4), 3.69 (dd, 1 H, J 7.8, J 8.3 Hz, H-3), 3.71 (s, 2
/
4:1 (v/v) hexaneꢀ
/
EtOAc] and concentration
ꢃ
/
[eluent 4:1 (v/v) hexaneꢀ
the appropriate fractions afforded (6-2H2)-19; TLC Rf
EtOAc], contaminated
/EtOAc] and concentration of
J
0.61 [eluent 2:1 (v/v) hexaneꢀ
/
with elimination product (6-2H)-20; Rf 0.55. This
mixture was dissolved in MeOH (20 mL) and NaOMe
(5 mg, 92.5 mmol) was added. The reaction mixture was
stirred at rt for 18 h after which CH2Cl2 (100 mL) was
added. The resulting solution was washed with 1 M HCl
2,3
3,4
H, H-7a, H-7b), 4.01 (d, 1 H, J
2.6 Hz, H-1), 4.57ꢀ
/
1,2
4.81 (m, 8 H, CH2Phꢂ
4); 13C NMR (75 MHz, CDCl3): dC 36.2, 65.7, 72.9,
73.4, 73.7, 75.4, 75.5, 75.6, 79.2, 81.8, 84.3, 127.7ꢀ128.7,
138.7, 138.4, 138.5, 138.9. Electrospray-MS: m/z 557.3
/
4), 7.19ꢀ
/
7.48 (m, 20 H, C6H5ꢂ
/
/
(50 mL), satd NaHCO3 (2ꢂ50 mL), brine (50 mL),
/
[Mꢁ
/
H]ꢁ. Electrospray-HRMS: m/z 579.2686 [Mꢁ
/
dried (MgSO4), filtered and the filtrate was concentrated
under reduced pressure. The resulting residue was
Na]ꢁ Calcd for C35H36D2NaO6: 579.2689.
purified by column chromatography [2:1 (v/v) hexaneꢀ
/
3.4.2. 2,3,4,7-Tetra-O-benzyl-1-O-(p-tolylsulfonyl)-5-
EtOAc] and concentration of the appropriate fractions
afforded (6-2H2)-21 (132 mg, 237 mmol, 34% over two
epi-(6-2H2)valiolol [(6-2H2)-18]. 2,3,4,7-Tetra-O-benzyl-
5-epi-(6-2H2)valiolol [(6-2H2)-16] (497 mg, 893 mmol)
was dissolved in Py (80 mL) and p-toluenesulfonyl
chloride (0.85 g, 4.47 mmol) and DMAP (54 mg, 447
mmol) were added. The reaction mixture was stirred for
15 h at rt followed by the addition of more p-
toluenesulfonyl chloride (0.85 g, 4.47 mmol) and
DMAP (54 mg, 447 mmol). The mixture was stirred at
steps) as a white solid; TLC Rf 0.21 [2:1 (v/v) hexaneꢀ
/
1
EtOAc); [a]2D3
ꢃ39.88 (c 0.6, CHCl3); H NMR (300
/
MHz, CDCl3): dH 2.16 (d, 1 H, J 1,OH 2.0 Hz, OH), 2.56
(s, 1 H, OH), 3.32 (dd, 1 H, J 1,2 8.8, J 2,3 9.3 Hz, H-2),
3.38 (d, 1 H, J
9.3 Hz, H-7b), 3.42 (dd, 1 H, J
3,4
7a,7b
9.8, J 2,3 9.3 Hz, H-3), 3.47 (dd, 1 H, J 1,2 8.8, J 1,OH 2.0
Hz, H-1), 3.58 (d, 1 H, J 9.3 Hz, H-7a), 3.65 (d, 1
7a,7b
70 8C for 5 h after which TLC [eluent 2:1 (v/v) hexaneꢀ
/
H, J
9.8 Hz, H-4), 4.44, 4.55, 4.60, 4.65, 4.70, 4.78,
3,4
EtOAc] indicated that the reaction was complete. The
reaction mixture was diluted with CH2Cl2 (100 mL) and
water (20 mL) was added. The organic layer was washed
4.81, 4.88 (8 H, 8ꢂ
20 H, ArH); 13C NMR (75 MHz, CDCl3): dC 68.3, 71.5,
73.8, 74.5, 75.5, 75.6, 75.7, 75.9, 82.4, 86.4, 127.7ꢀ128.9,
138.5, 137.9, 138.6, 138.9. Electrospray-MS: m/z 579.3
[Mꢁ
Na]ꢁ. Electrospray-HRMS: m/z 579.2690 [Mꢁ
Na]ꢁ Calcd for C35H36D2NaO6: 579.2689.
/
1/2 AB, CH2Phꢂ
/
4), 7.53ꢀ
/
7.10 (m,
/
with 1 M HCl (2ꢂ
/
50 mL), NaHCO3 (2ꢂ50 mL), and
/
brine (50 mL), dried (MgSO4), filtered and the filtrate
concentrated under reduced pressure. The resulting
residue was purified by column chromatography [eluent
/
/
5:1 (v/v) hexaneꢀ
appropriate fractions afforded (6-2H2)-18 (523 mg, 736
mmol, 82%) as a clear syrup; [a]D23
8.58 (c 0.6, CHCl3);
1H NMR (300 MHz, CDCl3): dH 2.35 (s, 3 H, CH3),
3.41ꢀ3.48 (m, 2 H), 3.52 (dd, 1 H, J 1,2 3.1, J 2,3 7.3 Hz,
H-2), 3.59ꢀ3.72 (m, 2 H), 4.36ꢀ4.59 (m, 8 H, CH2Phꢂ
4), 4.84 (d, 1 H, J 3.1 Hz, H-1), 7.12ꢀ7.36 (m, 22 H,
ArH), 7.76 (d, 2 H, J 8.1 Hz, ArH); 13C NMR (75
/
EtOAc] and concentration of the
3.4.4. 1-epi-5-epi-(6-2H2)Valiolol [(6-2H2)-6]. 2,3,4,7-
Tetra-O-benzyl-1-epi-5-epi-(6-2H2)-valiolol [(6-2H2)-21]
ꢁ
/
(107 mg, 192 mmol) was dissolved in 95% EtOHꢀ
(10 mL) and PdÃC (107 mg) was added. The suspension
was stirred under an atmosphere of H2 at rt for 28 h
after which TLC [eluent 4:1:1 (v/v/v) 2-propanolꢀ
waterꢀAcOH] indicated complete conversion to a
slower running product. The product was filtered
/H2O
/
/
/
/
/
/
/
/
1,2

K. Arakawa et al. / Carbohydrate Research 338 (2003) 2075ꢀ
/2082
2081
through Celite and concentrated under reduced pressure
to give [(6-2H2)-6] (36 mg, 184 mmol, 96%) as a white
spray-HRMS: m/z 557.2874 [Mꢁ
C35H37D2O6: 557.2870.
/
H]ꢁ Calcd for
solid; TLC Rf 0.63 [eluent 4:1:1 (v/v/v) 2-propanolꢀ
/
waterꢀ
/
AcOH]; [a]2D3
(300 MHz, CD3OD): dH 3.28 (dd, 1 H, J
ꢃ
/
7.38 (c 0.3, H2O); 1H NMR
3.5. 1-epi-2-epi-5-epi-(6-2H2)Valiolol [(6-2H2)-12]
9.3, J
3.45 (m, 1 H, H-3), 3.52 (d, 1 H, J
2,3
1,2
8.6 Hz, H-2), 3.40ꢀ
/
A mixture of benzyl ether (6-2H2)-24 (200 mg, 0.359
3,4
9.9 Hz, H-4), 3.57 (d, 1 H, J
11.7 Hz, H-7b), 3.59
mmol) and wet 10% PdꢀC (180 mg) in 95% aq EtOH (20
/
7a,7b
(d, 1 H, J 1,2 8.6 Hz, H-1), 3.85 (d, 1 H, J 7a,7b 11.7 Hz,
H-7a); 13C NMR (75 MHz, CD3OD): dC 39.2 (br, m),
64.6, 69.8, 74.5, 75.3, 79.8, 80.5. Electrospray-MS: m/z
mL) was stirred under a hydrogen atmosphere at rt for
14 h. The mixture was passed through a pad of Celite
with 1:1 MeOHꢀwater, and the filtrate and washings
/
218.9 [Mꢁ
/
Na]ꢁ. Electrospray-HRMS: m/z 219.0817
were concentrated to dryness. The residue was dissolved
in water and was passed through a membrane filter. The
filtrate was concentrated to dryness to afford (6-2H2)-12
[Mꢁ
/
Na]ꢁ Calcd for C7H12D2NaO6: 219.0812.
3.4.5. 5-epi-(6-2H2)Valiolol [(6-2H2)-17]. 2,3,4,7-Tetra-
O-benzyl-5-epi-(6-2H2)-valiolol [(6-2H2)-16] (101 mg,
(69.0 mg, quant.) as a white solid; [a]D25
ꢃ
/
6.038 (c 0.318,
H2O); 1H NMR (300 MHz, D2O): d 3.56 (d, 1 H, J
7a,7b
181 mmol) was dissolved in 95% EtOHꢀ
and PdÃC (101 mg) was added. The suspension was
stirred under an atmosphere of H2 at rt for 24 h after
which TLC [4:1:1 (v/v/v) 2-propanolꢀwaterꢀAcOH]
/
H2O (10 mL)
12 Hz, H-7a), 3.57 (dd, 1 H, J 2,3 3.0, J 3,4 10.4 Hz, H-3),
3.75 (d, 1 H, J 7a,7b 12 Hz, H-7b), 3.78 (d, 1 H, J 3,4 10.4
Hz, H-4), 3.79 (d, 1 H, J 1,2 3.0 Hz, H-1), 4.03 (t, 1 H, J
/
/
/
3.0 Hz, H-2); 13C NMR (75 MHz, D2Oꢀ
/
1,2/2,3
indicated complete conversion to a lower running
product. The product was filtered through Celite and
concentrated under reduced pressure to give (6-2H2)-17
(34 mg, 173 mmol, 96%) as a white solid; TLC Rf 0.60
CD3ODꢄ
74.6, 76.7. Electrospray-MS: 219 [Mꢁ
spray-HRMS: m/z 219.0814 [Mꢁ
Na]ꢁ Calcd for
C7H12D2NaO6: 219.0812.
/
2:1): dC 34.3 (br, m), 63.3, 67.0, 72.0, 73.6,
/
Na]ꢁ. Electro-
/
[eluent 4:1:1 (v/v/v) 2-propanolꢀ
/
waterꢀ
/
AcOH]; [a]2D3
ꢁ
/
1
18.08 (c 0.6, H2O); H NMR (300 MHz, CD3OD): dH
3.52 (d, 1 H, J 3,4 7.3 Hz, H-4), 3.60 (dd, 1 H, J 1,2 3.6, J
2,3 6.7 Hz, H-2), 3.63 (d, 1 H, J 7a,7b 10.9 Hz, H-7b), 3.74
(d, 1 H, J 7a,7b 10.9 Hz, H-7a), 3.83 (dd, 1 H, J 2,3 6.7, J
3.5.1. 2,3,4,7-Tetra-O-benzyl-1-O-(p-tolylsulfonyl)-1-
epi-2-epi-5-epi-(6-2H2)valiolol [(6-2H2)-25]. To a solu-
tion of alcohol (6-2H2)-24 (373 mg, 0.671 mmol) and
DMAP (165 mg, 1.35 mmol) in Py (13 mL) was added
p-toluenesulfonyl chloride (1.41 g, 7.40 mmol) at rt, and
the mixture was stirred at rt for 10 h. Water (10 mL) was
added to the mixture, and the product was extracted
with EtOAc. The combined organic phase was washed
with 2 M HCl, satd NaHCO3, and brine, dried
(Na2SO4), filtered and concentrated to dryness. The
residue was chromatographed over silica gel with 4:1
7.3 Hz, H-3), 4.02 (d, 1 H, J
3.6 Hz, H-1); 13C
3,4
1,2
NMR (75 MHz, CD3OD): dC 36.5 (br, m), 66.8, 68.3,
73.2, 75.4, 76.4, 78.5. Electrospray-MS: m/z 219.0 [Mꢁ
Na]ꢁ. Electrospray-HRMS: m/z 219.0816 [MꢁNa]ꢁ
/
/
Calcd for C7H12D2NaO6: 219.0812.
3.4.6. 2,3,4,7-Tetra-O-benzyl-1-epi-2-epi-5-epi-
(6-2H2)valiolol [(6-2H2)-24]. To a solution of 2,3,4,7-
tetra-O-benzyl-2-epi-5-epi-(6-2H2)valiolone [(6-2H2)-23]
hexanesꢀ
90%) as a colorless syrup; [a]D26
1H NMR (300 MHz, CDCl3): d 2.37 (s, 3 H, CH3Ã
C6H4), 3.42 (d, 1 H, J 9.3 Hz, H-7a), 3.49 (d, 1
/
EtOAc to give tosylate (6-2H2)-25 (429 mg,
ꢃ
/
12.838 (c 0.6, CHCl3);
(696 mg, 1.25 mmol) in THFꢀ
was added NaBH4 (200 mg, 5.28 mmol) in THFꢀ
/
MeOH (7 mL, 1:1 v/v)
MeOH
/
/
7a,7b
(3 mL, 1:1 v/v) at 0 8C, and the mixture was stirred at rt
for 2.5 h. The mixture was concentrated in vacuo, and
the resulting residue was diluted with ether and water.
The organic phase was washed with 2 M HCl, satd
NaHCO3, and brine, dried (Na2SO4), filtered and
concentrated to dryness. The residue was chromato-
H, J 7a,7b 9.3 Hz, H-7b), 3.67 (dd, 1 H, J 2,3 3.0, J 3,4 8.8
Hz, H-3), 3.96 (d, 1 H, J 3,4 8.8 Hz, H-4), 3.98 (br, 1 H),
4.48ꢀ
/
4.81 (m, 10 H), 7.10ꢀ
7.70 (ArH); 13C NMR (75
/
MHz, CDCl3): dC 21.4, 34.3 (br, m), 71.6, 72.8, 72.9,
73.0, 75.1, 76.1, 76.5, 78.6, 127.3, 127.4, 127.4, 127.5,
127.5, 127.6, 128.0, 128.1, 128.1, 128.3, 129.5, 133.8,
137.6, 138.0, 138.3, 138.5, 144.4. Electrospray-MS: m/z
graphed over silica gel with hexanesꢀ
to give alcohol (6-2H2)-24 (600 mg, 86%) as a colorless
syrup; [a]2D4 11.38 (c 0.5, CHCl3); 1H NMR (300 MHz,
CDCl3): d 2.80 (1 H, br, 1-OH), 3.39 (d, 1 H, J 7a,7b 9.3
Hz, H-7a), 3.56 (d, 1 H, J 9.3 Hz, H-7b), 3.75 (t, 1
/
EtOAc (4:1ꢀ/2:1)
711 [Mꢁ
/
H]ꢁ. Electrospray-HRMS: m/z 733.2780 [Mꢁ
/
ꢃ
/
Na]ꢁ Calcd for C42H42D2NaO8S: 733.2778.
3.5.2. 2,3,4,7-Tetra-O-benzyl-1-azido-1-deoxy-2-epi-5-
7a,7b
H, J 1,2/2,3 3.1 Hz, H-2), 3.80 (dd, 1 H, J 3,4 5.7, J 2,3 3.1
Hz, H-3), 4.01 (d, 1 H, J 3,4 5.7 Hz, H-4), 4.05 (br, 1 H,
epi-(6-2H2)valiolol [(6-2H2)-26]. A mixture of tosylate
(6-2H2)-25 (43.5 mg, 0.0612 mmol), NaN3 (250 mg, 3.85
mmol), and 18-crown-6 (15 mg, 0.0567 mmol) in DMF
(1 mL) was stirred at 90 8C. Additional NaN3 (250 mg,
3.85 mmol) and 18-crown-6 (15 mg, 0.0757 mmol) were
added to the reaction mixture every 16 h period. After
H-1), 4.55ꢀ
NMR (75 MHz, CDCl3): d 34.9, 67.2, 71.7, 72.9, 73.4,
73.5, 73.5, 74.0, 76.3, 77.2, 78.5, 127.6ꢀ128.2, 137.9ꢀ
138.2. Electrospray-MS: m/z 557 [Mꢁ
H]ꢁ. Electro-
/
4.73 (CH2Phꢂ
/
4), 7.10ꢀ
/
7.45 (ArH); 13C
/
/
/

2082
K. Arakawa et al. / Carbohydrate Research 338 (2003) 2075ꢀ2082
/
52 h stirring, water (3 mL) was added to the mixture,
and the product was extracted with EtOAc. The
combined organic phase was washed with brine, dried
(Na2SO4), filtered and concentrated to dryness. The
residue was chromatographed over silica gel with 4:1
her colleagues at Bayer AG for cultures and materials,
and Dr Ross F. Lawrence, University of Washington
Mass Spectrometry Center, for his help with the ESI MS
analysis, as well as Dr Oliver Block for providing us
1
reference H NMR data. This work was supported by
Bayer AG, Wuppertal, Germany and by NIH grant
AI20264.
hexanesꢀ
EtOAc to give azide (6-2H2)-26 (11.3 mg, 32%)
/
as a colorless syrup and unreacted tosylate (6-2H2)-25
1
(16.7 mg, 38%); [a]D23
(300 MHz, CDCl3): d 3.35 (d, 1 H, J 7a,7b 9.9 Hz, H-7a),
ꢃ
/
15.68 (c 0.7, CHCl3); H NMR
3.56 (dd, 1 H, J
(dd, 1 H, J
2.1, J
9.9 Hz, H-7b), 3.67
7a,7b
7b,5-OH
9.9 Hz, H-3), 3.78 (d, 1 H, J
3,4 1,2
3.1, J
2,3
References
3.6 Hz, H-1), 3.84 (dd, 1 H, J 2,3 3.1, J 1,2 3.6 Hz, H-2),
4.08 (d, 1 H, J 3,4 9.9 Hz, H-4), 4.29ꢀ4.74 (CH2Phꢂ4),
4.39 (d, 1 H, J 2.1 Hz, 5-OH), 7.10ꢀ7.45 (ArH);
13C NMR (75 MHz, CDCl3): dC 57.0, 72.5, 73.2, 73.7,
74.0, 74.1, 74.9, 75.3, 75.8, 79.7, 127.6ꢀ128.5, 136.9,
137.8, 137.8, 138.1. Electrospray-MS: m/z 582 [Mꢁ
H]ꢁ
and 554 [MꢁHꢃ
N2]ꢁ. Electrospray-HRMS: m/z
582.2932 [Mꢁ
H]ꢁ Calcd for C35H36D2N3O5: 582.2935.
/
/
1. Mahmud, T. Nat. Prod. Rep. 2003, 20, 137ꢀ166.
2. Truscheit, E.; Frommer, W.; Junge, B.; Muller, L.;
¨
Schmidt, D.; Wingender, W. Angew. Chem., Int. Ed.
/
/
5-OH,7b
/
Engl. 1981, 20, 744ꢀ
/761.
/
3. Wakae, O. J. Pesticide Sci. 1981, 6, 455ꢀ
/456.
/
/
4. Xie, X.; Rivier, A.-S.; Zakrzewicz, A.; Bernimoulin, M.;
Zeng, X.-L.; Wessel, H. P.; Schapira, M.; Spertini, O. J.
/
Biol. Chem. 2000, 275, 34818ꢀ
5. Mahmud, T.; Lee, S.; Floss, H. G. Chem. Rev. 2001, 1,
300ꢀ301.
6. Mahmud, T.; Tornus, I.; Egelkrout, E.; Wolf, E.; Uy, C.;
Floss, H. G.; Lee, S. J. Am. Chem. Soc. 1999, 121, 6973ꢀ
/34825.
3.5.3. 2-epi-5-epi-(6-2H2)Valiolamine [(6-2H2)-11].
A
mixture of azide (6-2H2)-26 (110 mg, 0.189 mmol) and
wet 20% Pd(OH)2 on charcoal (315 mg) in 95% aq EtOH
(14 mL) was stirred under hydrogen atmosphere at rt for
36 h. The mixture was passed through a pad of Celite
/
/
6983.
7. Dong, H.; Mahmud, T.; Tornus, I.; Lee, S.; Floss, H. G. J.
with 1:1 MeOHꢀwater , and the filtrate and washings
/
Am. Chem. Soc. 2001, 123, 2733ꢀ2742.
/
were concentrated to dryness. The residue was dissolved
in water and passed through a membrane filter. The
filtrate was concentrated to dryness to afford (6-2H2)-11
8. Naganawa, H.; Hashizume, H.; Kubota, Y.; Sawa, R.;
Takahashi, Y.; Arakawa, K.; Bowers, S. G.; Mahmud, T.
J. Antibiot. 2002, 55, 578ꢀ
9. Stratmann, A.; Mahmud, T.; Lee, S.; Distler, J.; Floss, H.
G.; Piepersberg, W. J. Biol. Chem. 1999, 274, 10889ꢀ
/584.
(37.0 mg, quant.) as an amorphous solid; [a]D27
ꢁ
/
45.98 (c
0.7, H2O); 1H NMR (300 MHz, D2O): d 3.30 (d, 1 H, J
12 Hz, H-7a), 3.46 (d, 1 H, J 9.4 Hz, H-4), 3.49
/
10896.
10. Mahmud, T.; Xu, J.; Choi, Y. U. J. Org. Chem. 2001, 66,
7a,7b
3,4
(d, 1 H, J 7a,7b 12 Hz, H-7b), 3.75ꢀ
H-3), 3.91 (brs, 1 H, H-1); 13C NMR (75 MHz, D2Oꢀ
CD3ODꢄ8:1): dC 32.0 (br, m), 48.4, 66.8, 69.6, 70.3,
74.1, 75.6. Electrospray-MS: m/z 196 [Mꢁ
H]ꢁ. Elec-
trospray-HRMS: m/z 196.1152 [Mꢁ
H]ꢁ Calcd for
/
3.79 (m, 2 H, H-2 and
5066ꢀ
11. Bowers, S. G.; Mahmud, T.; Floss, H. G. Carbohydr. Res.
2002, 337, 297ꢀ304.
12. Block, O., Ph.D. thesis, Bergische Universitat ꢀ
/
5073.
/
/
/
/
/
GH
¨
/
Wuppertal, Germany, 2000.
13. Shing, T. K. M.; Wan, L. H. J. Org. Chem. 1996, 61,
8468ꢀ8479.
C7H14D2NO5: 196.1152.
/
14. Zhang, C.-S.; Stratmann, A.; Block, O.; Bruckner, R.;
¨
Podeschwa, M.; Altenbach, H.-J.; Wehmeier, U. F.;
Acknowledgements
Piepersberg, W. J. Biol. Chem. 2002, 277, 22853ꢀ
15. Zhang, C.-S.; Podeschwa, M.; Altenbach, H.-J.; Piepers-
berg, W.; Wehmeier, U. F. FEBS Lett. 2003, 540, 47ꢀ52.
/
22862.
The authors are indebted to Professor Heinz G. Floss
for his guidance and support, Dr Anneliese Crueger and
/