Some approaches to glycosylated versions of methyl β-acarviosin

Article abstract of DOI:10.1071/CH03203

In a first approach to a glycosylated version of 'methyl β-acarviosin', a putative inhibitor of cellulases, cellobiose was converted into a carbocyclic enone that could not be transformed into the required amine for a subsequent alkylation. Alternatively,

Full text of DOI:10.1071/CH03203

Full Paper
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2004, 57, 197–205
Some Approaches to Glycosylated Versions of Methyl β-Acarviosin
Jon K. Fairweather,A Matthew J. McDonough,A Robert V. Stick,A,B and
D. Matthew G. TilbrookA
A Chemistry, School of Biomedical and Chemical Science M313, University of Western Australia,
Crawley WA 6009, Australia.
B Author to whom correspondence should be addressed (e-mail: rvs@chem.uwa.edu.au).
In a first approach to a glycosylated version of ‘methyl β-acarviosin’, a putative inhibitor of cellulases, cellobiose was
converted into a carbocyclic enone that could not be transformed into the required amine for a subsequent alkylation.
Alternatively, methyl β-acarviosin itself was glycosylated at C4ꢀ, using a ‘glycosynthase’, to provide the ‘trisac-
charide’ (and some ‘tetrasaccharide’). Both of these molecules were effective inhibitors of various cellulases. In a
related approach to a regioisomer of the above ‘trisaccharide’, a selectively protected derivative of 1-epivalienamine
was alkylated with a carbohydrate triflate to give a ‘disaccharide’ that could not be glycosylated to give the desired
‘trisaccharide’. Another unsuccessful approach to this molecule is also reported.
Manuscript received: 3 August 2003.
Final version: 15 September 2003.
We have reported a synthesis of ‘methyl β-acarviosin’ 1, a
diastereoisomer of methyl acarviosin 2, one of the degrada-
tion products from the treatment of acarbose with methanol
(Diagram 1).[1,2] Much to our disappointment, the amine 1
proved to be an ineffective inhibitor of various cellulases,
unlike the quite striking inhibitory action of 2 against some
hydrolases that process α-d-glucosidic linkages.[2]
We have also shown that the inhibitory action of various
imino sugars can be markedly improved upon by glyco-
sylation. For example, isofagomine 3 is a weak inhibitor
of a xylanase from Cellulomonas fimi but the glycosylated
versions 4 and 5 exhibit a marked improvement in inhibition
(Diagram 2).[3] It thus seemed a reasonable proposal to pre-
pare the glycosylated methyl β-acarviosin derivative 6, in the
hope of developing an inhibitor of cellulases. As well, we
included in our proposal a synthesis of the regioisomer 7,
potentially an inhibitor of 1,3:1,4-β-d-glucan hydrolases.[4]
There are two general approaches possible for a synthesis
of 6, namely the glycosylation of a properly functionalized
methyl β-acarviosin or the alkylation of a suitable amine
OH
OH
HO
O
HO
HO
HO
HO
HO
O
NH
NH
OH
3
4
OH
OH
HO
O
HO
HO
O
HO
HO
O
O
NH
OH
OH
OH
5
HO
OH
HO
HO
OMe
HO
NH
O
HO
HO
O
CH3
HO
O
HO
OMe
HO
HO
HO
NH
1
O
CH3
HO
OH
OH
OH
6
HO
HO
OH
CH3
HO
HN
HO
HO
O
O
O
HO
HO
HO
OMe
HO
NH
O
HO
OH
HO
CH3
OMe
2
7
Diagram 1.
Diagram 2.
10.1071/CH03203
© CSIRO 2004
0004-9425/04/030197

198
J. K. Fairweather et al.
(Diagram 3). In light of our rather inefficient synthesis of
methyl β-acarviosin,[1] we chose to investigate the latter
approach first.
Thus, towards an amine such as 8, methyl β-cellobioside
was converted into its benzylidene acetal and thence the
iodide 9 (Scheme 1). Treatment of 9 with 1,8-diazabicyclo-
[5.4.0]undec-7-ene (DBU) then gave the alkene, and protect-
ing group interchange provided 10. The alkene 10 was then
converted into the enone 11.
We next needed to add a one-carbon unit to the ketone
group of 11. Treatment of 11 under a variety of conditions
with the Grignard reagent derived from benzyl chloromethyl
ether never yielded a single product—the diastereoisomers
12 and 13 were generally obtained in equal amounts. For 12,
there was a noticeable nuclear Overhauser effect between H5
and a H7, which was absent in 13. The lack of stereoselec-
tivity in this addition was both unexpected and frustrating,
especially in light of a successful addition to the enone 14
(Diagram 4).[5] Still and McDonald have commented upon
similar results in a related system.[6]
With small amounts of the desired diastereoisomer 12 in
hand, acetylation under forcing conditions gave the acetate
15 (Scheme 1). The subsequent treatment of 15 with sodium
azide under Pd(0) catalysis never realised the azide 16, a
transformation accomplished satisfactorily with the unglyco-
sylated cyclohexene.[5]
At this stage we decided to abandon the ‘amine alkylation’
approach and turn to the alternative, namely the glycosylation
of a methyl β-acarviosin (Diagram 3).Although conventional
glycosylation methodology could be followed to achieve this
aim, we had available some pure methyl β-acarviosin 1 (from
our earlier synthesis) and decided to utilize an enzyme-
assisted process for the glycosylation. Although glycoside
hydrolases have been used extensively in the past for such
glycosylations,[7] we were attracted to the obvious benefits
of ‘glycosynthases’.[8]
OP
OP
OP
PO
PO
OX
HO
PO
OMe
OMe
PO
NH
O
O
OP
CH3
PO
OP
O
PO
O
PO
PO
PO
O
PO
BnO
BnO
PO
NH2
O
PO
CH3
BnO
OP
XO
8
14
Diagram 3.
Diagram 4.
OAc
O
Ph
O
O
OMe
AcO
O
(a)
(b)
methyl ꢀ-cellobioside
AcO
O
OAc
I
9
OBn
O
OBn
O
O
Ph
Ph
O
O
O
BnO
OMe
(c)
BnO
O
O
BnO
O
BnO
OBn
OBn
O
10
11
OBn
OBn
BnO
O
O
Ph
Ph
O
O
O
BnO
O
O
BnO
BnO
O
HO
(d )
ꢁ
BnO
OBn
OBn
1
HO
OBn
13
12
(e)
OBn
O
Ph
OBn
BnO
O
Ph
O
O
N3
BnO
O
O
O
BnO
O
BnO
BnO
OBn
OBn
OBn
AcO
15
16
Scheme 1. (a) PhCH(OMe)2, PTSA, DMF, then Ph3P, imidazole, dioxan, then Ac2O, pyridine, DMAP; (b) DBU, THF,
then Na, MeOH, then NaH, BnBr, DMF; (c) Hg(OCOCF3)2, H2O, acetone, then MsCl, Et3N, CH2Cl2; (d) Mg, BnOCH2Cl,
HgCl2, THF; (e) Ac2O, pyridine, DMAP.

Glycosylated Versions of Methyl β-Acarviosin
199
OAc
AcO
1
1؆
AcO
AcO
OMe
O
AcO
AcO
NH
AcO
AcO
O
O
1؅
CH3
OAc
17
OR
OR
OR
OR
O
O
OMe
O
RO
RO
NH
RO
O
RO
RO
O
CH3
OR
OR
OR
18 R = Ac
20 R = H
OAc
OAc
OMe
AcO
AcO
AcO
NH
O
CH3
OAc
19
Diagram 5.
Thus, methyl β-acarviosin 1 and α-d-glucopyranosyl
fluoride in aqueous ammonium bicarbonate were treated
with the glycosynthase AbgGlu358Ser.[9] After three days
all of the fluoride had been consumed and the mixture
appearedtocontaintwonewcompounds, isolatedastheirper-
acetates. The major product (42%) was the ‘trisaccharide’17
(Diagram 5), along with the ‘tetrasaccharide’ 18 (6%) and
some acetylated methyl β-acarviosin 19 (24%). The 1H
NMR spectrum of 17 revealed both the stereoselectivity and
regioselectivity of the glycosylation—the magnitude of the
Table 1. Inhibition constants of methyl β-acarviosin 1 and the
polyols 6 and 20 against various cellulases
Cellulases
Ki [µM]
1
6
20
EGII (T. reesei)
–
–
–
–
410
200
–
60
90
–
100
–
CelC (C. thermocellum)
CBHII (T. reesei)
CBHI (T. reesei)
EGI (T. reesei)
CelD (C. thermocellum)
–
>1000
>1000
–
–
15
ꢀꢀ ꢀꢀ
coupling constant, J1 ,2 (8.0 Hz) confirmed that a new β-d-
linkage had been formed, and the chemical shift of the signal
for H4ꢀ (δ 4.20) indicated glycosylation (and not acetylation)
at O4ꢀ.
The per-acetates 17 and 18 were each subjected to treat-
ment with sodium methoxide in methanol, presumably to
formthepolyols6and 20, andthesetwopolyols, togetherwith
methyl β-acarviosin 1, were assayed as inhibitors of various
cellulases. The results are presented in Table 1 and, perhaps
expectedly, reveal the polyol 20 to be the most effective
inhibitor.
We now shifted our attention to the amine 7 and lamented
the fact that there was no readily available glycoside hydro-
lase, letaloneaglycosynthase, thatcouldinstallaβ-d-glucose
residue at the 5ꢀ-OH of methyl β-acarviosin 1.∗ Our approach
had to be purely synthetic, and we chose to glycosylate the
alcohol 21 (Diagram 6).Thisalcoholshould beavailablefrom
the amine 22 and the β-d-galacto triflate 23.
OBn
OBn
OP
OBz
O
O
BzO
NH
BnO
HO
OMe
PO
OX
PO
CH3
TfO
OP
21
OBn
OBn
Me
O
BnO
PMBO
BnO
HO
BzO
OMe
NH2
N3
OBz
OBn
OBn
24
22
23
Diagram 6.
failed to give the methyl α-d-glucoside 25 (Diagram 7)—it
probably gave just methyl α-d-glucopyranoside.
As an alternative, we decided to investigate the use of stan-
nylene chemistry for the preparation of a molecule related
to 25. Although methyl 2-O-benzyl-4,6-O-benzylidene-α-
d-glucoside 26 (a product of stannylene chemistry) could
be converted into its p-methoxybenzyl ether 27, subsequent
A logical precursor of the amine 22 was the azide 24,
and this seemed a realistic task in view of our experiences in
the synthesis of other cyclohexenyl azides from monosaccha-
rides.Thus, inpursuitof 24, weconverted‘diacetoneglucose’
into its p-methoxybenzyl ether but methanolysis of this ether
∗ Later on, such a glycosynthase did become available,[26a] but failed to couple the amine 1 and α-laminaribiosyl fluoride.[26b]

200
J. K. Fairweather et al.
treatment with LiAIH4/AlCl3 (to transform the benzylidene
ring) caused loss of the newly introduced protecting group.
The selective acid hydrolysis of 27 (to remove the benzyli-
dene group) fared no better, so we prepared the more labile
p-methoxybenzylidene acetal 28 and attempted a selective
alkylation with p-methoxybenzyl chloride in the presence of
dibutyltin oxide—the so-formed mixture of 2- and 3-O-(p-
methoxybenzyl) ethers was inseparable. Finally, we found
that iodine in refluxing methanol[10] smoothly converted the
ether 27 into the diol 29.
range of d-glucosyl donors failed to yield the desired ‘trisac-
charide’ in pure form. Those donors tried were the bromide
35 (Diagram 8; virtually no reaction in the presence ofAgOTf
and tetramethylurea[18]), the trichloroacetimidate 36 (again
virtuallynoreactioninthepresenceofborontrifluorideether-
ate; probable formation of the trimethylsilyl (TMS) ether
of 21 when TMSOTf was used as the promoter; probable
formation of an ortho ester, not amenable to rearrangement
with TMSOTf, when AgOTf was used as the promoter),
Although we gained access to 29 there was still a problem
in obtaining multi-gram amounts of material—the prob-
lem resided in the selective benzylation (dibutyltin oxide,
benzyl bromide, tetrabutylammonium iodide) of methyl 4,6-
O-benzylidene-α-d-glucoside. Boons and coworkers[11,12]
have reported an 83% yield for such a selective benzyla-
tion in benzene [(1t3o,14p]roduce the 2-O-benzyl ether 26) but
cess. However, the problem was solved when we conducted
the benzylation in heptane in the absence of tetrabutyl-
ammonium iodide (conditions that favour a dimeric structure
for the intermediate stannylene acetal[14])—the 2-O-benzyl
ether was isolated in 76% yield.
Returning to the synthesis of 24 (Scheme 2), the diol 29
was converted into the iodide 30 and, somewhat gratifyingly,
benzylation and alkene formation proceeded in a single, sub-
sequent step. The alkene 31 was then converted into the
azide 24 in a straightforward sequence of reactions.[5] The
desired amine 22 was then obtained in an oxidation–reduction
sequence.
With the amine 22 in hand, we then required the triflate
23. The iodide 32[15] (Scheme 3) was converted into the
6-deoxy sugar 33 and subsequently the alcohol 34.Treatment
of this alcohol with triflic anhydride and pyridine then gave
the crystalline and rather stable triflate 23. Both the alcohol
34 and the triflate 23 have been prepared previously by either
circuitous or poorly described routes,[16,17] and the present
procedures are a vast improvement.
OH
O
Ph
O
O
(a)
(b)
O
HO
PMBO
HO
HO
BnO
OMe
OMe
29
I
we, and others,
have been unable to match this suc-
O
O
(d)
(c)
BnO
PMBO
HO
PMBO
BnO
BnO
OMe
OMe
30
24
31
OBn
OBn
(e)
BnO
PMBO
BnO
HO
N3
NH2
BnO
BnO
22
Scheme 2. (a) Bu2SnO, C6H6, followed by BnBr, heptane, then NaH,
p-MeOC6H4CH2Cl, DMF, then I2, MeOH; (b) I2, Ph3P, imidazole,
PhMe; (c) NaH, BnBr, DMF; (d) Hg(OCOCF3)2, H2O, acetone, then
MsCl, Et3N, CH2Cl2, then Mg, BnOCH2Cl, HgCl2, THF, then Ac2O,
pyridine, DMAP, then NaN3, Pd(PPh3)4, H2O, THF; (e) DDQ, H2O,
CH2Cl2, then HS(CH2)3SH, Et3N, MeOH.
I
O
O
CH3
O
O
(a)
(b)
In the crucial alkylation step, treatment of the amine 22
with the triflate 23 gave the alcohol 21 in an acceptable yield
(36%) for this sort of procedure (Scheme 4).
O
O
OMe
OMe
OH
OBz
32
33
Wewerenowreadyfortheevenmorecrucialglycosylation
step and, to our dismay, treatment of the alcohol 21 with a
HO
TfO
CH3
CH3
O
O
(c)
BzO
BzO
OMe
OMe
OH
O
Ph
OBz
OBz
O
O
HO
PMBO
O
34
23
RO
HO
BnO
Scheme 3. (a) H2, Pd/C, Et2NH, EtOAc/petrol, then BzCl, pyridine;
(b) AcOH, H2O, then BzCl, pyridine; (c) Tf2O, pyridine, CH2Cl2.
OMe
OMe
25
26 R ꢂ H
27 R ꢂ PMB
OH
OBn
OBn
p-MeOC6H4
O
O
O
OBz
O
HO
PMBO
O
(a)
BnO
HO
BnO
HO
BzO
NH
HO
OMe
NH2
BnO
HO
OBn
OBn
Me
OMe
OMe
28
29
22
21
Diagram 7.
Scheme 4. (a) 23, DMI.

Glycosylated Versions of Methyl β-Acarviosin
201
the thioglycoside 37 (either no reaction or decomposition
with varying amounts of N-iodosuccinmide/triflic acid as
promoter), and the sulfoxide 38.
treated with the trichloroacetimidate 36 in the presence of
TMSOTf—the azide 42 was formed in good yield. This suc-
cess contrasts markedly with the attempts described in the
paragraphs above and, no doubt, belies the absence of a
(basic) nitrogen atom in the acceptor molecule.
However, oursuccesswastobeshort-lived—theattempted
reduction of the azide 42 to the amine 40, using propane-1,3-
dithiol in methanol, gave a complex mixture (deacetylation?),
and a resultant effort to change the protecting groups of
the pyranose ring (acetyl to benzyl) foundered when the
sodium hydride necessary for the benzylations seemed to
cause reduction of the azide group as well.
With the sulfoxide 38 as donor, the procedure recom-
mended by Kahne and coworkers was followed[19]—half an
equivalent of triflic anhydride was added to the sulfoxide
in dichloromethane at −60◦C, followed by the alcohol 21.
Gratifyingly, TLC analysis of the reaction mixture showed
rapid conversion of the alcohol into a new, less polar, com-
pound. Disappointingly, though, this ‘compound’ proved to
be an inseparable mixture of what appeared to be the two
anomers of the desired ‘trisaccharide’.
Although the reasons for these unsuccessful glycosyla-
tions are not clear, the answer perhaps lies in the presence
of a (basic) nitrogen atom in the alcohol 21. However,
such a nitrogen atom gave no problems for Miyamoto and
Ogawa (working with the validamycins)[20] and Rukhman
and coworkers (glycosylating a derivative of morphine).[21]
It is somewhat ironic that the very atom that seemed to be
at the root of our problem is very difficult to protect—by
way of example, the amine 39 is totally unreactive towards
trifluoroacetic anhydride.
At this stage of our work we were very keen to follow the
approach that had been first tried in pursuit of the ‘trisaccha-
ride’ 6, namely the alkylation of a regioisomer of the amine
8 (Diagram 3). Unfortunately, whereas cellobiose is a readily
available disaccharide, laminaribiose is not and so a glyco-
sylation route had to be used for the synthesis of an amine
such as 40 (Diagram 9). To this end, the alcohol 41 was
Experimental
General experimental procedures have been given previously.[5]
Methyl β-Cellobioside
Cellobiose (20 g, 58 mmol) and NaOAc (10 g) were heated in Ac2O
(180 mL) under reflux until the mixture became clear (30 min). The
mixture was cooled and then poured into ice-cold water, resulting in the
formation of a milky precipitate. This precipitate was collected and sub-
jected to a normal workup (CH2Cl2) to give a colourless solid.This solid
in HOAc (100 mL) was treated with HBr in HOAc (30% w/v, 19 mL)
overnight. The solution was diluted with ice-cold water and subjected
to a normal workup (CH2Cl2) to give an off-white solid. This solid was
added to Ag2CO3 (12 g, 44 mmol) and powdered 3 Å molecular sieves
(10 g) in MeOH (100 mL) and CH2Cl2 (100 mL). The resultant mix-
ture was stirred (3 h) in the absence of light. The mixture was filtered
through a plug of silica (EtOAc) and the filtrate was concentrated to give
a colourless solid. A small piece of Na was added to this solid in MeOH
(100 mL). Shortly afterwards, a solid began to crystallize from solution.
After 1 h, the mixture was cooled and the solid collected. The filtrate
was neutralized with resin [Dowex 50W-X8 (H+)], filtered, and the
filtrate concentrated to give a yellow residue. This residue and the pre-
viously collected solid were combined and recrystallized to give methyl
β-cellobioside as a microcrystalline solid (13.9 g, 67%), mp 192–194◦C
(EtOH/MeOH; lit.[22] 198◦C).
OAc
OAc
O
O
AcO
AcO
AcO
AcO
OC(NH)CCl3
AcO
OAc
Br
35
36
Methyl 2,3-Di-O-acetyl-6-deoxy-4-O-(2,3-di-O-acetyl-4,6-O-
benzylidene-β-D-glucosyl)-6-iodo-β-D-glucoside 9
OBz
OPiv
Iodine (800 mg, 3.4 mmol) was added to methyl 4-O-(4,6-O-
benzylidene-β-d-glucosyl)-β-d-glucoside (prepared from the above
methyl β-cellobioside;[23] 1.0 g, 2.2 mmol), Ph3P (890 mg, 3.4 mmol),
and imidazole (460 mg, 6.8 mmol) in dioxan (5 mL). The mixture was
then heated (65◦C, 30 min). After the mixture was cooled, H2O (2 mL)
was added. The mixture was concentrated and the residue treated with
Ac2O/pyridine (15 mL, 1 : 2) and 4-dimethylaminopyridine (DMAP;
50 mg) for 5 h. MeOH (3 mL) was added to the mixture, which was
then concentrated. Normal workup (CH2Cl2) gave a crystalline mass
that was recrystallized to give the iodide 9 as needles (1.2 g, 72%),
mp 275◦C (dec.) [EtOH/CHCl3; lit.[23] 268–270◦C (dec.)], [α]D −48.5◦
(lit.[23] −47.8◦). δH (500 MHz) 2.00, 2.03, 2.03, 2.12 (4 s, 12 H, Ac),
3.23–3.30 (m, H5,6), 3.48–3.52 (m, H5ꢀ), 3.51 (s, OMe), 3.56–3.73
O
O
BzO
BzO
PivO
PivO
SPh
S(O)Ph
OBz
OPiv
37
38
OAc
OAc
O
AcO
AcO
OMe
AcO
NH
OAc
CH3
39
(m, H4, 6, 4ꢀ, 6ꢀ), 4.33–4.37 (m, H6ꢀ), 4.44 (d, J1,2 7.9, H1), 4.71 (d, J1 ,2
ꢀ
ꢀ
Diagram 8.
ꢀ
7.8, H1ꢀ), 4.88 (dd, J2,3 9.5, H2ꢀ), 4.94 (dd, J2 ,3 9.2, H2 ), 5.17 (dd, J3,4
ꢀ
ꢀ
ꢀ
ꢀ
8.9, H3), 5.28 (dd, J3 ,4 9.3, H3 ), 5.47 (s, CHPh), 7.34–7.42 (m, Ph). δC
(125 MHz) 4.25 (C6), 20.77, 20.90, 20.92 (4 C, COMe), 57.01 (OMe),
66.32 (C5ꢀ), 68.36 (C6ꢀ), 71.55, 71.81, 72.57, 72.59 (C2, 3, 2ꢀ, 3ꢀ), 73.47
(C5), 77.86, 80.40 (C4, 4ꢀ), 100.92 (C1), 101.34 (C1ꢀ), 101.50 (CHPh),
126.05–136.52 (Ph), 169.25, 169.48, 169.61, 170.15 (4 C, C=O).
OBn
OBn
OAc
O
AcO
AcO
BnO
BnO
HO
O
X
N3
OAc
OBn
OBn
Methyl 2,3-Di-O-acetyl-6-deoxy-4-O-(2,3-di-O-acetyl-4,6-O-
benzylidene-β-D-glucosyl)-β-D-xylo-hex-5-enoside
40 X ꢂ NH2
42 X ꢂ N3
41
The iodide 9 (1.0 g, 1.4 mmol) was treated with DBU (690 µL,
4.60 mmol) in THF (10 mL) at reflux (30 min). The solution was con-
centrated to give an orange oil. This oil was dissolved in EtOAc (40 mL)
Diagram 9.

202
J. K. Fairweather et al.
and the resultant solution was washed with cold, aqueous HCl (1 M),
followed by a normal workup to give an off-white solid. Recrystalliza-
tion gave methyl 2,3-di-O-acetyl-6-deoxy-4-O-(2,3-di-O-acetyl-4,6-O-
benzylidene-β-d-glucosyl)-β-d-xylo-hex-5-enoside as needles (0.71 g,
86%), mp 175◦C (EtOH), [α]D −112◦ (Found C 56.6, H 5.9. C28H34O14
requires C 56.5, H 5.9%). δH (300 MHz) 2.04, 2.06, 2.07, 2.08 (4 s, 12
H, Ac), 3.45–3.82 (m, H4ꢀ, 5ꢀ, 6ꢀ), 3.53 (s, OMe), 4.32–4.42 (m, H1,
(900 mg, 40%) as needles, mp 124◦C (Pri O), [α]D +2.7◦ (Found C 75.0,
H 6.5. C56H56O10 requires C 75.2, H, 62.5%). δH (500 MHz) 3.16–3.24
(m, H5ꢀ), 3.24 (d, J7,7 9.1, H7), 3.45–3.52 (m, H2ꢀ, 6ꢀ), 3.60–3.67 (m,
H7, 3ꢀ, 4ꢀ), 3.94 (dd, J4,5 7.2, J5,6 9.6, H5), 4.09 (d, H6), 4.13–4.17 (m,
H4), 4.23–4.32 (m, H6ꢀ), 4.33–4.98 (10H, m, CH2Ph), 4.39 (d, J1 ,2
ꢀ
ꢀ
7.7, H1ꢀ), 5.52 (s, CHPh), 5.63 (dd, J2,3 10.1, J2,4 2.0, H2), 5.90 (dd,
J3,4 2.2, H3), 7.21–7.52 (m, Ph). δC (125 MHz) 65.81, 78.29, 78.77,
79.36, 81.46, 81.85, 81.87 (C4, 5, 6, 2ꢀ, 3ꢀ, 4ꢀ, 5ꢀ), 68.77, 72.04, 72.73,
73.20, 74.60, 74.90, 75.45 (7 C, C7, 6ꢀ, CH2Ph), 72.04 (C1), 101.08,
103.25 (C1ꢀ, CHPh), 125.98–137.88 (C2, 3, Ph).
6ꢀ), 4.62–4.68 (m, H2, 6), 4.73 (d, J1 ,2 7.8, H1ꢀ), 4.80 (dd, J4,6 ≈ J6,6
ꢀ
ꢀ
1.2, H6), 4.92 (dd, J2,3 ≈ J3,4 5.1, H3), 4.98–5.07 (m, H4, 2ꢀ), 5.30
(dd, J2 ,3 ≈ J3 ,4 9.3, H3ꢀ), 5.50 (s, CHPh), 7.32–7.48 (m, Ph). δC
(75.5 MHz) 14.54, 20.71, 20.81, 21.00 (4 C, COMe), 56.99 (OMe),
65.30, 66.40, 68.47, 71.87, 72.28, 77.41, 78.05 (C2, 3, 4, 2ꢀ, 3ꢀ, 4ꢀ, 5ꢀ),
68.37 (C6ꢀ), 96.61, 100.86, 101.48 (C1, 1ꢀ, CHPh), 126.08–144.34 (C5,
6, Ph), 169.39, 169.44, 169.94, 170.03 (4 C, C=O).
ꢀ
ꢀ
ꢀ
ꢀ
Next to elute was the alcohol 13 as needles (760 mg, 34%), mp
141◦C (Pri2O), [α]D −13◦ (Found C 75.2, H 6.6. C56H56O10 requires
C 75.2, H, 6.5%). δH (500 MHz) 3.37–3.43 (m, H5ꢀ), 3.50–3.54 (m,
H7,2ꢀ), 3.59 (d, J7,7 9.2, H7), 3.66 (dd, J5 ,6 ≈ J6 ,6 10.3, H6ꢀ), 3.72
ꢀ
ꢀ
ꢀ
ꢀ
(dd, J3 ,4 ≈ J4 ,5 9.3, H4ꢀ), 3.87 (dd, J2 ,3 9.0, H3ꢀ), 4.03 (dd, J4,5 6.6,
J5,6 10.2, H5), 4.10 (d, H6), 4.20–4.26 (m, H4, 6ꢀ), 4.24–5.00 (m, 11
H, H1ꢀ,CH2Ph), 5.56 (s, CHPh), 5.65 (dd, J2,3 10.4, J2,4 2.0, H2), 5.75
(dd, J3,4 2.5, H3), 7.23–7.50 (m, Ph). δC (125 MHz) 65.81, 78.29, 78.77,
79.36, 81.46, 81.84, 81.87 (C4, 5, 6, 2ꢀ, 3ꢀ, 4ꢀ, 5ꢀ), 68.75, 72.04, 72.73,
73.20, 74.57, 74.87, 75.45 (7 C, C7, 6ꢀ, CH2Ph), 72.73 (C1), 101.08,
103.25 (C1ꢀ, CHPh), 125.98–137.88 (C2, 3, Ph).
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
Methyl 2,3-Di-O-benzyl-6-deoxy-4-O-(2,3-di-O-benzyl-
4,6-O-benzylidene-β-D-glucosyl)-β-D-xylo-hex-5-enoside 10
A small piece of Na was added to methyl 2,3-di-O-acetyl-6-deoxy-4-
O-(2,3-di-O-acetyl-4,6-O-benzylidene-β-d-glucosyl)-β-d-xylo-hex-5-
enoside (590 mg, 1.0 mmol) in MeOH (10 mL) at 0◦C. The solution
was left to stand (room temp., 1 h) and was then concentrated,
co-evaporating with toluene (2 × 10 mL). NaH (220 mg, 60% dispersion
in oil, 5.6 mmol) was added to a solution of the residue in dimethyl
formamide (DMF; 10 mL). BnBr (520 g µL, 4.4 mmol) was then added
dropwise and the resultant mixture was stirred (30 min). EtOH (1 mL)
was then added. Concentration of the mixture and normal workup
(CH2Cl2) gave a yellow oil that was purified by rapid silica filtra-
tion (RSF) (toluene to EtOAc/toluene, 1 : 3) to give the alkene 10 as
a microcrystalline powder (760 mg, 97%), mp 93–95◦C (Pr2i O), [α]D
−47◦ (Found C 73.2, H 6.6. C48H50O10 requires C 73.3, H 6.4%). δH
(300 MHz) δ 3.36–3.83 (m, H2, 3, 2ꢀ, 3ꢀ, 4ꢀ, 5ꢀ, 6ꢀ), 3.48 (s, OMe), 4.32
(3R,4S,5R,6S)-3-Acetoxy-5,6-dibenzyloxy-3-benzyloxymethyl-4-
(2,3-di-O-benzyl-4,6-O-benzylidene-β-D-glucosyloxy)cyclohexene 15
The alcohol 12 (540 mg, 0.62 mmol) was treated with Ac2O (1 mL) and
pyridine (10 mL) in the presence of DMAP (120 mg, 1.0 mmol) at 60◦C
for4 h. MeOH(1 mL)wasaddedtothenowdarksolution. Concentration
of the mixture followed by a normal workup (EtOAc) gave a brown oil.
This oil was purified by flash chromatography (EtOAc/petrol, 1 : 4) to
give the acetate 15 as a colourless oil (440 mg, 79%), [α]D −2.2◦. δH
(600 MHz) 1.98 (s, Ac), 3.30–3.35 (m, H5ꢀ), 3.43 (dd, J1 ,2 7.6, J2 ,3
ꢀ
ꢀ
ꢀ
ꢀ
8.5, H2ꢀ), 3.64 (dd, J5 ,6 ≈ J6 ,6 10.4, H6ꢀ), 3.68 (dd, J3 ,4 ≈ J4 ,5 9.3,
H4ꢀ), 3.78 (dd, H3ꢀ), 3.82, 3.87 (2 H, ABq, J 10.0, H7), 4.15 (dd, J4,5
9.6, J5,6 7.4, H5), 4.26–4.32 (m, H6, 6ꢀ), 4.28–4.96 (10 H, m, CH2Ph),
4.80 (d, H1ꢀ), 4.93 (d, H4), 5.53 (s, CHPh), 5.82 (dd, J1,2 10.4, J1,6 2.4,
H1), 5.89 (dd, J2,6 2.1, H2), 7.25–7.50 (m, Ph). δC (150 MHz) 22.08
(COCH3), 66.20, 77.19, 79.11, 81.28, 81.43, 81.66, 82.26 (C4, 5, 6,
2ꢀ, 3ꢀ, 4ꢀ, 5ꢀ), 68.83, 71.75, 72.24, 73.43, 74.68, 75.03, 75.07 (7 C, C7,
6ꢀ, CH2Ph), 85.06 (C3), 101.06 (CHPh), 102.44 (C1ꢀ), 125.96–138.31
(C1, 2, Ph), 169.70 (C=O). HRMS (FAB) m/z 919.4050 [C57H59O16
(dd, J5 ,6 6.0, J6 ,6 11.6, H6ꢀ), 4.43 (d, J1,2 6.9, H1), 4.60–4.96 (m, 12
H, H4, 6, 1ꢀ, CH2Ph), 5.58 (s, CHPh), 7.16–7.52 (m, Ph). δC (75.5 MHz)
56.90 (OMe), 66.00, 75.67, 80.81, 81.04, 81.20, 81.80, 82.05 (C2, 3,
4, 2ꢀ, 3ꢀ, 4ꢀ, 5ꢀ), 68.85, 73.77, 73.92, 75.21, 75.71 (5 C, C6ꢀ, CH2Ph),
95.98, 100.80, 101.16 (C1, 1ꢀ, CHPh), 126.01–152.30 (C5, 6, Ph).
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀꢀ
ꢀ
ꢀ
ꢀ
ꢀ
(4S,5R,6S)-6-(2,3-Di-O-benzyl-4,6-O-benzylidene-β-D-
glucosyloxy)-4,5-(dibenzyloxy)cyclohex-2-enone 11
•
(M + H)+ requires 919.4057].
The alkene 10 (460 mg, 0.60 mmol) was treated with Hg(OCOCF3)2
(50 mg) in acetone/H2O (9 mL, 2 : 1) overnight. Volatile solvents were
removed and the resultant aqueous mixture was subjected to a normal
workup(EtOAc), followedbyRSF(EtOAc/petrol, 1 : 2), togiveacolour-
less oil. MsCl (95 µL, 1.2 mmol) was added to this oil in Et3N (330 µL,
2.4 mmol) and CH2Cl2 (10 mL) at −10◦C. The solution was then left
to stand (room temp., 1 h). Normal workup (CH2Cl2) followed by flash
chromatography (EtOAc/petrol, 1 : 9) gave the enone 11 as a colour-
less oil (590 mg, 80%), [α]D +5.7◦ (Found C 74.6, H 6.1. C47H46O9
requires C 74.8, H 6.1%). δH (500 MHz) 3.38–3.44 (m, H5ꢀ), 3.57 (dd,
Glycosylation of Methyl 4,6-Dideoxy-4-[1ꢀR,4ꢀR,5ꢀS,6ꢀS)-
4ꢀ,5ꢀ,6ꢀ-trihydroxy-3ꢀ-(hydroxymethyl)cyclohex-2ꢀ-enyl]amino-
β-D-glucoside (Methyl β-Acarviosin) 1
α-d-Glucopyranosyl fluoride (64 mg, 0.32 mmol) and methyl
β-acarviosin 1 (80 mg, 0.23 mmol) in NH4HCO3 (3 mL of 0.15 M)
were treated withAbgGlu358Ser (1 mg) at room temperature for 3 days.
The solution was lyophilized and the resultant residue was treated with
Ac2O (1 mL) and pyridine (3 mL). Concentration of the mixture fol-
lowed by flash chromatography (EtOAc/petrol, 4 : 6 to 6 : 4) gave, first,
some acetylated methyl β-acarviosin[1] (32 mg, 24%), followed by the
‘trisaccharide’ 17 as a colourless oil (88 mg, 42%), [α]D −82◦ (Found
C 50.2, H 6.0. C38H54NO24 requires C 50.2, H, 6.0%). δH (500 MHz)
1.33 (d, J5,6 6.2, 3 H, H6), 1.98, 2.00, 2.03, 2.03, 2.04, 2.05, 2.08, 2.10
(8 s, 27 H, Ac), 2.47 (dd, J3,4 ≈ J4,5 9.3, H4), 3.18–3.28 (m, H5, 1ꢀ),
J1 ,2 7.3, J2 ,3 8.2, H2ꢀ), 3.68 (dd, J5 ,6 ≈ J6 ,6 10.3, H6ꢀ), 3.73 (dd,
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
J3 ,4 ≈ J4 ,5 9.4, H4ꢀ), 3.83 (dd, H3ꢀ), 3.94 (dd, J4,5 8.1, J5,6 10.6,
ꢀ
H5), 4.29 (dd, J5 ,6 5.0, H6ꢀ), 4.42 (ddd, J2,4 ≈ J3,4 2.2, H4), 4.54 (d,
H6), 4.73–5.13 (9H, m, H1ꢀ, CH2Ph), 5.50 (s, CHPh), 6.06 (dd, J2,3
10.4, H3), 6.85 (dd, H2), 7.25–7.50 (m, Ph). δC (125 MHz) 65.72, 78.47,
80.08, 80.72, 81.30, 82.30, 83.75 (C4, 5, 6, 2ꢀ, 3ꢀ, 4ꢀ, 5ꢀ), 69.92, 73.81,
74.77, 75.13, 75.39 (5 C, C6ꢀ, CH2Ph), 101.02, 102.48 (C1ꢀ, CHPh),
125.98–148.24 (C2, 3, Ph), 196.01 (C1).
ꢀ
ꢀ
3.46 (s, OMe), 3.67–3.71 (m, H5ꢀꢀ), 4.05 (dd, J5 ,6 2.3, J6 ,6 12.4,
ꢀꢀ ꢀꢀ
ꢀꢀ ꢀꢀ
H6ꢀꢀ), 4.20 (br d, J4 ,5 5.7, H4ꢀ), 4.28 (d, J1,2 7.7, H1), 4.33 (dd, J5 ,6
ꢀ
ꢀ
ꢀꢀ ꢀꢀ
6.6, H6ꢀꢀ), 4.49, 4.57 (2 H, ABq, J 13.0, H7ꢀ), 4.68 (d, J1 ,2ꢀꢀ 8.0, H1ꢀꢀ),
ꢀꢀ ꢀꢀ
(1R,4S,5R,6S)-4,5-Dibenzyloxy-1-benzyloxymethyl-6-(2,3-di-O-
benzyl-4,6-O-benzylidene-β-D-glucosyloxy)cyclohex-2-en-1-ol 12
and (1S,4S,5R,6S)-4,5-Dibenzyloxy-1-benzyloxymethyl-6-(2,3-di-O-
benzyl-4,6-O-benzylidene-β-D-glucosyloxy)cyclohex-2-en-1-ol 13
4.81–4.95 (m, H2, 3, 6ꢀ, 2ꢀꢀ), 5.05 (dd, J3 ,4 ≈ J4 ,5 9.4, H4 ), 5.15 (dd,
ꢀꢀ ꢀꢀ
ꢀꢀ
ꢀꢀꢀ ꢀꢀ
ꢀ
ꢀꢀ ꢀꢀ
ꢀ
ꢀ
J2 ,3 9.4, H3 ), 5.36 (dd, J5 ,6 8.5, H5 ), 5.84 (br s, H2 ). δC (125 MHz)
17.78(C6), 20.54, 20.58, 20.66, 20.69, 20.75, 20.81, 20.88(9C, COMe),
56.75 (OMe), 56.88 (C1ꢀ), 61.64 (C6ꢀꢀ), 62.49 (C4), 63.76 (C7ꢀ), 67.91
(C4ꢀꢀ), 70.86 (C5ꢀ), 71.51, 71.78, 71.92, 72.09 (C2, 6ꢀ, 2ꢀꢀ, 5ꢀꢀ), 73.03
(C3ꢀꢀ), 73.89 (C5), 74.82 (C3), 76.84 (C4ꢀ), 101.16 (C1), 101.43 (C1ꢀꢀ),
128.60 (C2ꢀ), 131.16 (C3ꢀ), 169.30, 169.36, 169.54, 169.68, 170.02,
170.06, 170.26, 170.62, 171.37 (9 C, C=O).
Magnesium (370 mg, 15 mmol), benzyl chloromethyl ether (1.9 mL,
14 mmol)andHgCl2 (50 mg)inTHF(15 mL)werestirred(90 min, 0◦C).
The enone 11 (2.0 g, 2.6 mmol) was added and the mixture was stirred
(1 h). The mixture was poured into saturated NaHCO3 solution and the
suspension was stirred (10 min). Normal workup (CH2Cl2) followed by
flash chromatography (EtOAc/petrol, 15 : 85) gave, first, the alcohol 12
Next to elute was the ‘tetrasaccharide’18 as a colourless oil (16 mg,
6%), [α]D −28◦. δH (600 MHz) 1.33 (d, J5,6 6.1, 3 H, H6), 1.97, 1.98,

Glycosylated Versions of Methyl β-Acarviosin
203
2.00, 2.00, 2.07, 2.11, 2.14 (7 s, 21 H, Ac), 2.03–2.04 (m, 15 H, Ac),
2.51 (br s, H4), 3.07–3.14 (m, H5), 3.34 (br s, H1ꢀ), 3.47 (s, OMe),
and then cooled and diluted with saturated NaHCO3 solution (100 mL).
I2 was added until the solution remained dark (5 min). Na2S2O3 solu-
tion (1 M) was then added to decolourize the mixture. Normal workup
(toluene) followed by RSF (EtOAc/toluene, 1 : 4) gave the iodide 30
as a colourless oil (25.7 g, 87%), [α]D +15◦ (Found C 51.5, H 5.4.
C22H27IO6 requires C 51.4, H 5.3%). δH (300 MHz) 3.11–3.22 (m, H2,
5), 3.32 (dd, J5,6 2.1, J6,6 9.4, H6), 3.34 (s, OMe), 3.38–3.46 (m, H4,
6), 3.65 (dd, J2,3 ≈ J3,4 9.0, H3), 3.72 (s, ArOMe), 4.48–4.88 (4 H, m,
CH2Ar), 4.54 (d, J1,2 3.5, H1), 6.80–7.36 (9 H, m, Ar). δH (75.5 MHz)
6.98 (C6), 55.27, 55.53 (2 C, OMe), 69.78, 73.60, 79.83, 80.30 (C2, 3,
4, 5), 73.14, 74.96 (2 C, CH2Ar), 98.13 (C1), 114.09–159.45 (Ar).
3.60–3.75 (m, H4ꢀꢀ, 5ꢀꢀ, 5ꢀꢀꢀ), 4.04 (dd, J5 ,6 2.2, J6 ,6 12.5, H6ꢀꢀ),
ꢀꢀ ꢀꢀ
ꢀꢀ ꢀꢀ
4.08 (dd, J5
5.5, J6
ꢀꢀ ꢀꢀ
12.1, H6ꢀꢀꢀ), 4.19 (br s, H4ꢀ), 4.31 (br s,
ꢀꢀꢀ,6ꢀꢀꢀ
ꢀꢀꢀ,6ꢀꢀꢀ
H7ꢀ), 4.37 (dd, J5 ,6 4.3, H6ꢀꢀ), 4.39 (br d, J7 ,7 10.8, H7ꢀ), 4.46 (d,
ꢀ
ꢀ
J1,2 7.9, H1), 4.58 (br d, H6ꢀꢀꢀ), 4.64 (br d, J1 ,2 7.9, H1ꢀꢀ), 4.83–4.94
ꢀꢀ ꢀꢀ
(m, H2, 3, 6ꢀ, 2ꢀꢀ, 1ꢀꢀꢀ, 2ꢀꢀꢀ), 5.06 (dd, J3
≈ J4
9.8, H4ꢀꢀꢀ), 5.13
ꢀꢀꢀ,5ꢀꢀꢀ
ꢀꢀꢀ,4ꢀꢀꢀ
(m, H3ꢀꢀ,3ꢀꢀꢀ), 5.33 (dd, J4 ,5 5.7, J5 ,6 8.4, H5ꢀ), 5.83 (br s, H2ꢀ). δC
(150 MHz) 17.77 (C6), 20.47, 20.53, 20.59, 20.66, 20.70, 20.77, 20.81,
20.91 (12 C, COMe), 56.77 (OMe), 57.00 (C1ꢀ), 61.46, 62.15, 63.74
(C7ꢀ, 6ꢀꢀ, 6ꢀꢀꢀ), 62.57 (C4), 67.69, 71.06, 71.52, 71.69, 71.87, 71.96,
72.10, 72.70, 72.81, 72.90, 73.92, 74.80, 76.34, 76.94 (C2, 3, 5, 4ꢀ, 5ꢀ,
6ꢀ, 2ꢀꢀ, 3ꢀꢀ, 4ꢀꢀ, 5ꢀꢀ, 2ꢀꢀꢀ, 3ꢀꢀꢀ, 4ꢀꢀꢀ, 5ꢀꢀꢀ), 100.88, 101.18, 101.34 (C1, 1ꢀꢀ, 1ꢀꢀꢀ),
128.61 (C2ꢀ), 131.34 (C3ꢀ), 169.14, 169.31, 169.48, 169.57, 169.69,
169.80, 169.95, 170.05, 170.21, 170.35, 170.52, 171.39 (12 C, C=O).
ꢀ
ꢀ
ꢀ
ꢀ
Methyl 2,4-Di-O-benzyl-6-deoxy-3-O-(4-methoxybenzyl)-
α-D-xylo-hex-5-enoside 31
•
m/z (FAB) 1164.3886 [(M + H)+ requires 1164.3904].
The iodide 30 (25.7 g, 46 mmol) was treated with NaH (8.0 g, 60% dis-
persioninoil, 200 mmol)andBnBr(7.0 mL, 65 mmol)inDMF(250 mL,
18 h). MeOH (10 mL) was added cautiously and the solution was con-
centrated. Normal workup (CH2Cl2) gave a dark oil that was purified
by flash chromatography (EtOAc/petrol, 3 : 17) to give the alkene 31 as
a colourless oil (16.3 g, 82%), [α]D −38◦. δH (300 MHz) 3.42 (s, OMe),
3.59 (dd, J1,2 3.4, J2,3 9.0, H2), 3.80 (s, ArOMe), 3.90 (ddd, J3,4 9.0,
J4,6 ≈ J4,6 1.8, H4), 4.23 (m, H3), 4.61 (d, H1), 4.65–4.87 (8 H, m, H6,
CH2Ar), 6.83–7.38 (14 H,Ar). δC (75.5 MHz) 55.24, 55.43 (2 C, OMe),
73.59, 75.46, 76.57 (3 C, CH2Ar), 79.24, 79.51, 80.92 (C2, 3, 4), 96.77
(C6), 99.04 (C1), 113.76–159.18 (C5, Ar).
Methyl 2-O-Benzyl-4,6-O-benzylidene-α-D-glucoside 26
Methyl 4,6-O-benzylidene-α-d-glucoside[24] (2.5 g, 8.7 mmol) and
Bu2SnO (2.4 g, 9.6 mmol) in C6H6 (100 mL) were heated under reflux
(12 h) with the azeotropic removal of water. The now clear solution
was concentrated. Heptane (50 mL) and BnBr (2.1 mL, 17 mmol) were
added and the solution was heated under reflux (3 days). Concentra-
tion of the solution followed by flash chromatography (EtOAc/toluene,
1 : 9 to 3 : 17) gave, first, the alcohol 26 as a colourless solid (2.48 g,
76%), mp 129–130.5◦C (EtOAc/petrol; lit.[25] 129.5◦C), [α]D +35◦
(lit.[25] +35◦), followed by methyl 3-O-benzyl-4,6-O-benzylidene-α-
d-glucoside, also as a colourless solid (230 mg, 7%), mp 182–183◦C
(EtOAc/petrol; lit.[25] 185◦C), [α]D +71◦ (lit.[25] +84◦).
(4S,5R,6S)-4,6-Dibenzyloxy-5-[(4-methoxybenzyl)oxy]-
cyclohex-2-enone
The alkene 31 (16.1 g) was treated with Hg(OCOCF3)2 (500 mg) in
acetone/water (300 mL, 2 : 1) for 12 h. The mixture was concentrated
somewhat and the resultant aqueous solution was extracted with CH2Cl2
(3 × 70 mL). The combined extracts were dried, filtered, and concen-
trated. RSF (EtOAc/petrol, 1 : 2) of the residue gave a colourless oil.
MsCl (11 mL) was added dropwise to this oil and Et3N (35 mL) in
CH2Cl2 (200 mL) at −10◦C. The solution was then stirred (1 h, room
temp.). Normal workup (CH2Cl2) gave an oil that was purified by flash
chromatography (EtOAc/petrol, 1 : 9 to 1 : 4) to give (4S,5R,6S)-4,6-
dibenzyloxy-5-[(4-methoxybenzyl)oxy]cyclohex-2-enone as a colour-
less oil (12.0 g, 80%), [α]D +62◦ (Found C 76.0, H 6.1. C28H28O5
requires C 75.7, H 6.4%). δH (300 MHz) 3.78 (s, OMe), 3.91 (dd, J4,5
7.7, J5,6 10.7, H5), 3.99 (d, H6), 4.30 (ddd, J2,4 2.0, J3,4 2.4, H4), 4.68–
5.08 (6 H, m, CH2Ar), 5.99 (dd, J2,3 10.4, H3), 6.77 (dd, H2), 6.79–7.46
(14 H, m, Ar). δC (75.5 MHz) 55.22 (OMe), 73.57, 74.48, 75.36 (3 C,
CH2Ar), 78.96, 83.88, 84.31 (C4, 5, 6), 113.74–159.25 (C2, 3, Ar),
197.38 (C1).
Methyl 2-O-Benzyl-4,6-O-benzylidene-3-O-(4-methoxybenzyl)-
α-D-glucoside 27
The alcohol 26 (24.1 g, 60 mmol) was treated with NaH (3.2 g, 60%
dispersion in oil, 84 mmol) and 4-methoxybenzyl chloride (10.0 mL,
72 mmol) in DMF (250 mL) for 2 h. MeOH (2 mL) was added and the
mixture was concentrated to give an orange oil. This oil was subjected
to a normal workup (CH2Cl2) to give a crystalline mass, which was
used without further purification in the next step. A small portion was
recrystallized to give the ether 27 as needles, mp 93◦C (Pr2i O), [α]D
−33◦ (Found C 70.5, H 6.6. C29H32O7 requires C 70.5, H 6.6%). δH
(300 MHz) 3.42 (s, OMe), 3.55 (dd, J1,2 3.7, J2,3 9.6, H2), 3.58 (dd,
J3,4 ≈ J4,5 9.6, H4), 3.72 (dd, J5,6 ≈ J6,6 9.6, H6), 3.80 (s, ArOMe),
3.79–3.85 (m, H5), 4.03 (dd, H3), 4.27 (dd, J5,6 4.7, H6), 4.60 (d, H1),
4.68–4.89 (4 H, m, CH2Ar), 5.56 (s, CHPh), 6.88–7.51 (m, 14 H, Ar).
δC (75.5 MHz) 55.25, 55.33 (2 C, OMe), 62.34, 78.28, 79.14, 82.12
(C2, 3, 4, 5), 69.06, 73.79, 75.04 (3 C, C6, CH2Ar), 99.26 (C1), 101.24
(CHPh), 113.72–159.18 (Ar).
(1R,4S,5R,6S)-4,6-Dibenzyloxy-1-benzyloxymethyl-5-
[(4-methoxybenzyl)oxy]cyclohex-2-en-1-ol
Methyl 2-O-Benzyl-3-O-(4-methoxybenzyl)-α-D-glucopyranoside 29
The crude ether 27 was treated with I2 (2 g) in MeOH (200 mL) at reflux
(5 h). Powdered Na2S2O3 was added. The mixture was filtered and the
filtrate was concentrated to give a colourless oil. This oil was purified
by flash chromatography (EtOAc/petrol, 1 : 1) to give the diol 29 as a
colourless oil (22.9 g, 88% over two steps), [α]D +12◦ (Found C 65.3, H
7.0. C22H28O7 requires C 65.2, H 6.9%). δH (300 MHz) 3.36 (s, OMe),
3.36–3.62 (m, H2, 4, 5), 3.68–3.80 (3 H, m, H3, 6), 3.79 (s, ArOMe),
4.59 (d, J1,2 3.5 Hz, H1), 4.61–4.96 (4 H, m, CH2Ar), 6.88–7.36 (m, 9
H, Ar). δC (75.5 MHz) 55.14, 55.20 (2 C, OMe), 70.27, 70.63, 79.71,
80.86 (C2, 3, 4, 5), 62.34, 73.09, 74.96 (3 C, C6, CH2Ar), 98.13 (C1),
113.98–159.31 (Ar).
Magnesium (4.5 g, 180 mmol), benzyl chloromethyl ether (11.0 mL)
and HgCl2 (200 mg) were stirred in THF (100 mL) at 0◦C for
90 min. The above enone (11.8 g, 30 mmol) in tetrahydrofuran (50 mL)
was added and stirring was continued at 0◦C for 90 min. Saturated
NaHCO3 solution (100 mL) was added and the mixture left to
stand (10 min). The mixture was then filtered (Celite) and the fil-
trate was extracted with CH2Cl2. The combined organic extracts
were dried and filtered. Concentration of the filtrate gave an oil
that was purified by flash chromatography (EtOAc/petrol, 3 : 17)
to give (1R, 4S, 5R, 6S)-4,6-dibenzyloxy-1-benzyloxymethyl-5-[(4-
methoxybenzyl)oxy]cyclohex-2-en-1-ol as a colourless oil (9.5 g, 64%),
[α]D −58◦ (Found C 76.2, H 7.0. C36H38O6 requires C 76.3, H 6.8%).
δH (300 MHz) 3.63 (d, J7,7 9.2, H7), 3.65–3.90 (m, H4, 5, 7), 3.81
(s, OMe), 4.18 (d, J5,6 6.8, H6), 4.58–4.88 (8 H, m, CH2Ar), 5.73 (s,
H2, 3), 6.81–7.42 (m, Ar). δC (75.5 MHz) 55.25 (OMe), 72.08, 73.08,
73.95, 74.77, 75.59 (5 C, C7, CH2Ar), 75.70 (C1), 79.63, 81.27, 83.91
(C4, 5, 6), 113.72–159.15 (C2, 3, Ar).
Methyl 2-O-Benzyl-6-deoxy-6-iodo-3-O-(4-methoxybenzyl)-
α-D-glucopyranoside 30
Iodine (18.7 g, 74 mmol) was slowly added to imidazole (10.9 g,
160 mmol), Ph3P (20.7 g, 78 mmol), and the diol 29 (22.7 g, 52.5 mmol)
in toluene (200 mL). The mixture was vigorously stirred (1 h, 70◦C),

204
J. K. Fairweather et al.
(3R,4S,5R,6S)-3-Acetoxy-4,6-dibenzyloxy-3-benzyloxymethyl-
5-[(4-methoxybenzyl)oxy]cyclohexene
(40 mL, 1 : 1) under an atmosphere of H2. The mixture was filtered
(Celite) and the filtrate concentrated. The residue in CH2Cl2 (20 mL)
was treated with BzCl (900 µL, 7.7 mmol) and Et3N (2.0 mL, 14 mmol)
overnight. MeOH (1 mL) was added and the solution was stirred for a
further 5 min. Normal workup (CH2Cl2) gave a colourless solid that
was taken up in Et2O/petrol and filtered through a plug of silica. Con-
centration of the filtrate followed by recrystallization of the residue gave
the benzoate 33 as cubes (1.6 g, 71%), mp 102–103◦C (Et2O/petrol),
[α]D +59◦. δH (300 MHz) 1.35, 1.63 (2 s, CMe2), 1.46 (3 H, d, J5,6
6.6, H6), 3.46 (s, OMe), 3.95 (dq, J4,5 2.1, H5), 4.08 (dd, J3,4 5.4,
H4), 4.32 (dd, J2,3 7.4, H3), 4.59 (d, J1,2 8.2, H1), 5.23 (dd, H2), 7.31–
7.80 (m, Ph). δC (75.5 MHz) 16.55 (C6), 26.38, 27.78 (CMe2), 56.50
(OMe), 69.02, 73.60, 76.52, 77.27 (C2, 3, 4, 5), 101.38 (C1), 110.28
(CMe2),128.24–132.96 (Ph), 165.45 (C=O).
The above alcohol (9.3 g, 16.4 mmol) was treated with Ac2O (5.0 ml,
48 mmol) and DMAP (2.0 g, 24 mmol) in pyridine (40 mL) at 55◦C
for 3 h. MeOH (3 mL) was added and then the dark solution was con-
centrated. Normal workup (EtOAc) followed by flash chromatography
(EtOAc/petrol, 1 : 9) gave (3R,4S,5R,6S)-3-acetoxy-4,6-dibenzyloxy-
3-benzyloxymethyl-5-[(4-methoxybenzyl)oxy]cyclohexene as a colour-
less oil (8.2 g, 82%), [α]D +63◦ (Found C 74.9, H 6.6. C38H40O7
requires C 75.0, H 6.6%). δH (300 MHz) 1.86 (s, Ac), 3.80 (s, OMe),
3.88 (m, 2 H, H7), 4.07 (dd, J4,5 10.3, J5,6 7.8, H5), 4.32 (ddd, J1,6
2.0, J2,6 1.7, H6), 4.50 (d, H4), 4.56–4.89 (8 H, m, CH2Ar), 5.83 (dd,
J1,2 10.5, H1), 5.90 (dd, H2), 6.83–7.42 (m, Ar). δC (75.5 MHz) 22.02
(COCH3), 55.24 (OMe), 71.95, 72.24, 73.62, 74.97, 75.55 (5 C, C7,
CH2Ar), 79.93, 82.46, 85.07 (C4, 5, 6), 80.95 (C3), 113.68–159.10
(C1, 2, Ar), 170.01 (C=O).
Methyl 2,3-Di-O-benzoyl-6-deoxy-β-D-galactopyranoside 34
The benzoate 33 (580 mg, 1.8 mmol) was treated with aqueous
CH3COOH (20 mL, 80%) at 100◦C for 20 min. The solution was
concentrated, followed by repeated co-evaporation with toluene. The
residue in CH2Cl2 (10 mL) was treated with BzCl (230 µL, 2.0 mmol)
and pyridine (500 µL) at −50◦C. The solution was allowed to warm
to room temperature overnight, after which time MeOH (1 mL) was
added. Normal workup (CH2Cl2) followed by flash chromatography
(EtOAc/petrol, 1 : 3 to 1 : 1) gave the alcohol 34 as needles (600 mg,
86%), mp 131◦C (Pri O/petrol), [α]D +121◦ (Found C 65.4, H 5.7.
C21H22O7 requires C 265.3, H 5.7%). δH (300 MHz) 1.42 (d, J5,6 6.5, 3
H, H6), 3.53 (s, OMe), 3.89 (dq, J4,5 0.9, H5), 4.08–4.11 (m, H4), 4.59
(d, J1,2 8.0, H1), 5.30 (dd, J2,3 10.4, J3,4 3.2, H3), 5.69 (dd, H2), 7.31–
7.80 (m, 10H, Ph). δC (75.5 MHz) 16.14 (C6), 56.82 (OMe), 69.48,
70.16, 70.55, 74.85 (C2, 3, 4, 5), 102.11 (C1), 128.28–133.36 (Ph),
165.44, 165.96 (2 C, C=O).
(3R,4S,5S,6R)-3-Azido-4,6-dibenzyloxy-1-benzyloxymethyl-5-
[(4- methoxybenzyl)oxy]cyclohexene 24
Tetrakis(triphenylphosphino)palladium(0) (800 mg, 0.61 mmol) was
added to the above acetate (8.0 g, 13.2 mmol) and NaN3 (5.2 g, 80 mmol)
in deoxygenated THF/water (240 mL, 2 : 1). The yellow solution was
heated at reflux (2 h) under Ar. Volatile solvents were removed and the
resultant solution was extracted with CH2Cl2. The combined extracts
were dried and filtered. The filtrate was concentrated to give an orange
oil that was purified by flash chromatography (EtOAc/petrol, 1 : 9) to
give the azide 24 as needles (4.7 g, 59%), mp 50–51◦C (Et2O/petrol),
[α]D −58◦. δH (300 MHz) 3.62 (dd, J3,4 8.5, J4,5 10.5, H4), 3.80 (s,
OMe), 3.83 (dd, J5,6 7.8, H5), 3.70 (d, J7,7 12.2, H7), 4.06–4.22 (m,
H3,7), 4.24 (d, H6), 4.42–4.89 (8 H, m, CH2Ar), 5.80 (br s, H2), 6.82–
7.41 (m, Ar). δC (75.5 MHz) 55.26 (OMe), 63.52 (C3), 69.62, 72.39,
74.91, 75.14, 75.39 (5 C, C7, CH2Ar), 79.76, 82.61, 84.05 (C4, 5, 6),
113.85–138.46 (C1, 2, Ar).
Methyl 2,3-Di-O-benzoyl-6-deoxy-4-O-trifluoromethanesulfonyl-
β-D-galactoside 23
(1S,2R,5R,6S)-5-Azido-2,6-dibenzyloxy-3-(benzyloxymethyl)-
cyclohex-3-en-1-ol 41
The alcohol 34 (600 mg, 1.55 mmol) was treated with Tf2O (330 µL,
2.0 mmol) and pyridine (1 mL) in CH2Cl2 (10 mL) at 0◦C for 20 min.
The addition of saturated NaHCO3 solution, followed by normal workup
(CH2Cl2) and flash chromatography (EtOAc/petrol, 1 : 4), gave the
triflate 23 as a microcrystalline powder (700 mg, 87%), mp 128◦C
(Et2O/petrol), [α]D +64◦. δH (300 MHz) 1.48 (d, J5,6 6.6, 3 H, H6),
3.55 (s, OMe), 4.09 (br q, H5), 4.68 (d, J1,2 7.8, H1), 5.24 (br d, J3,4 3.0,
H4), 5.57 (dd, J2,3 10.8, H3), 5.68 (dd, H2), 7.32–8.03 (10 H, m, Ph).
δC (75.5 MHz) 16.40 (C6), 56.98 (OMe), 68.49, 68.86, 71.14, 84.90
(C2,3,4,5), 101.97 (C1), 118.32 (q, JC,F 318.9, CF3), 128.28–133.36
(Ph), 165.01, 165.78 (2 C, C=O).
The azide 24 (4.5 g, 7.6 mmol) was treated with 2,3-dichloro-5,6-
dicyano-1,4-benzoquinone (DDQ; 2.6 g, 11.4 mmol) in CH2Cl2/water
(100 mL, 95 : 5) for 2 h. Normal workup (CH2Cl2) followed by flash
chromatography (CH2Cl2/petrol, 2 : 3 to EtOAc/petrol, 1 : 4) gave
(1S,2R,5R,6S)-5-azido-2,6-dibenzyloxy-3-(benzyloxymethyl)cyclohex-
3-en-1-ol 41 as a colourless oil (2.4 g, 74%), [α]D −49◦ (Found C 71.4,
H 6.0. C28H29N3O4 requires C 71.3, H 6.2%). δH (300 MHz) 3.48 (dd,
J1,6 10.4, J5,6 8.5, H6), 3.90 (dd, J1,2 7.8, H1), 3.98 (br d, J7,7 12.8,
H7), 4.08–4.22 (m, H2, 5, 7), 4.47–5.02 (6 H, m, CH2Ph), 5.65–5.69
(br s, H4), 7.22–7.45 (15 H, Ph). δC (75.5 MHz) 63.05 (C5), 69.54,
72.53, 74.43, 75.20 (4 C, C7, CH2Ph), 76.13, 79.34, 82.11 (C1, 2, 6),
127.73–138.38 (C3, 4, Ph).
Methyl 2,3-Di-O-benzoyl-4-[(1ꢀR,4ꢀR,5ꢀS,6ꢀS)-4ꢀ,6ꢀ-dibenzyloxy-
3ꢀ-benzyloxymethyl-5-hydroxycyclohex-2ꢀ-enyl]amino-4,6-
dideoxy-β-D-glucoside 21
(1S,2R,5R,6S)-5-Amino-2,6-dibenzyloxy-3-(benzyloxymethyl)-
cyclohex-3-en-1-ol 22
The azide 41 (1.25 g, 2.65 mmol) was treated with Et3N (3.7 mL,
26 mmol) and propane-1,3-dithiol (2.9 mL, 26 mmol) in MeOH (20 mL)
at reflux for 5 h.The solution was concentrated to give an orange residue
that was purified by flash chromatography (EtOAc/toluene/Et3N,
20 : 79 : 1 to EtOAc/toluene/EtOH/Et3N, 30 : 64 : 5 : 1) to give the
amine 22 as a colourless oil (1.15 g, 97%), [α]D −71◦. δH (300 MHz)
3.23 (dd, J1,6 9.8, J5,6 7.2, H6), 3.43–3.50 (m, H5), 3.92 (br d, J7,7 11.7,
H7), 3.95 (dd, J1,2 9.5, H1), 4.20–4.28 (m, H2, 7), 4.48–4.93 (6 H, m,
CH2Ph), 5.62 (br s, H4), 7.21–7.42 (15 H, m, Ph). δC (75.5 MHz) 52.99
(C5), 70.22, 70.55, 73.99, 74.98 (4 C, C7, CH2Ph), 76.20, 80.25, 85.23
The amine 22 (1.92 g, 4.32 mmol) and the triflate 23 (900 mg,
1.74 mmol) were stirred in 1,3-dimethylimidazolidin-2-one (DMI;
3 mL) at room temperature for 4 days.The deep orange solution was con-
centrated (70◦C, 1 mmHg) and the residue partitioned between CH2Cl2
and saturated NaHCO3 solution. The organic layer was separated and
the aqueous layer was extracted with more CH2Cl2. The combined
organic extracts were dried and concentrated. Flash chromatography
(EtOAc/petrol, 3 : 17 to 1 : 3) of the residue gave, first, the alcohol 21 as
a pale yellow oil (520 mg, 36%), [α]D −12◦. δH (300 MHz) 1.51 (d, J5,6
6.5, 3 H, H6), 2.69 (br t, J3,4 ≈ J4,5 9.3, H4), 3.17–3.28 (m, H1ꢀ, 6ꢀ, 7ꢀ),
3.42–3.48 (m, H5), 3.47 (s, OMe), 3.80–3.84 (2 H, m, H5ꢀ, CH2Ph),
3.98–4.03 (m, H1, 4ꢀ, 7ꢀ), 4.58–5.10 (5 H, m, CH2Ph), 5.34–5.46 (m,
H2, 3), 5.57 (br s, H2ꢀ), 6.95–8.02 (m, Ph). δC (75.5 MHz) 18.12 (C6),
56.80 (OMe), 60.11, 62.96, 72.37, 74.65, 75.15, 75.67, 78.87, 82.10
(C2, 3, 4, 5, 1ꢀ, 4ꢀ, 5ꢀ, 6ꢀ), 70.53, 71.91, 73.92, 75.91 (4 C, C7ꢀ, CH2Ph),
101.50 (C1), 127.39–138.66 (C2ꢀ, 3ꢀ, Ph), 165.32, 165.21 (2 C, C=O).
Further elution gave the amine 22 (1.30 g).
•
(C1, 2, 6), 127.61–138.67 (C3, 4, Ph). m/z (FAB) 446.2346 [(M + H)+
requires 446.2331].
Methyl 2-O-Benzoyl-6-deoxy-3,4-O-isopropylidene-
β-D-galactoside 33
The iodide 32[15] (2.4 g, 7.0 mmol), Et2NH (1.5 mL, 14 mmol), and Pd-
on-carbon (300 mg of 10%) were stirred overnight in EtOAc/petrol

Glycosylated Versions of Methyl β-Acarviosin
205
(3R,4S,5S,6R)-3-Azido-4,6-dibenzyloxy-1-benzyloxymethyl-
5-[(tetra-O-acetyl-β-D-glucopyranosyl)oxy]cyclohexene 42
[6] W. C. Still, J. H. McDonald, Tetrahedron Lett. 1980, 21, 1031.
doi:10.1016/S0040-4039(00)78831-9
[7] D. H. G. Crout, G. Vic, Curr. Opin. Chem. Biol. 1998, 2, 98.
doi:10.1016/S1367-5931(98)80041-0
[8] S. J. Williams, S. G. Withers, Aust. J. Chem. 2002, 55, 3.
doi:10.1071/CH02005
[9] C. Mayer, D. L. Zechel, S. P. Reid, R. A. J. Warren, S. G. Withers,
FEBS Lett. 2000, 466, 40. doi:10.1016/S0014-5793(99)01751-2
[10] W. A. Szarek, A. Zamojski, K. N. Tiwari, E. R. Ison, Tetrahedron
Lett. 1986, 27, 3827. doi:10.1016/S0040-4039(00)83890-3
[11] G.-J. Boons, G. H. Castle, J. A. Clase, P. Grice, S. V. Ley,
C. Pinel, Synlett 1993, 913. doi:10.1055/S-1993-22650
[12] G.-J. Boons, G. H. Castle, J. A. Clase, P. Grice, S. V. Ley,
C. Pinel, Synlett 1994, 764.
The alcohol 41 (180 mg, 0.39 mmol), the trichloroacetimidate 36
(380 mg, 0.78 mmol), and 4 Å molecular sieves (500 mg) were stirred in
CH2Cl2 (5 mL) for 2 h. TMSOTf (2 drops) was added and the mix-
ture was stirred (1 h). Solid NaHCO3 (500 mg) was added and the
mixture filtered through a plug of silica (EtOAc). Concentration of
the filtrate, followed by flash chromatography (EtOAc/petrol, 3 : 7) of
the residue, gave the d-glucoside 42 as a colourless oil (230 mg, 74%),
[α]D −94◦. δH (600 MHz) 1.93, 2.01, 2.03, 2.08 (12 H, 4 s,Ac), 3.57 (dd,
J3,4 8.6, J4,5 10.0, H4), 3.56–3.62 (m, H5ꢀ), 3.91 (br d, J7,7 12.4, H7),
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
4.02 (dd, J5 ,6 2.3, J6 ,6 12.4, H6 ), 4.07–4.11 (m, H3,7), 4.19 (dd, J5 ,6
4.5, H6ꢀ), 4.21 (br d, J5,6 7.4, H6), 4.23 (dd, H5), 4.43–4.97 (m, 6 H,
CH2Ph), 5.08–5.13 (m, H2ꢀ,4ꢀ), 5.17 (t, J2 ,3 ≈ J3 ,4 9.4, H3ꢀ), 5.21 (d,
ꢀ
ꢀ
ꢀ
ꢀ
[13] D. J. Jenkins, B. V. L. Potter, Carbohydr. Res. 1994, 265, 145.
doi:10.1016/0008-6215(94)00221-5
J1 ,2 8.1, H1ꢀ), 5.62 (br s, H2), 7.18–7.47 (15 H, m, Ph). δC (150 MHz)
20.57, 20.59, 20.60, 20.90 (4 C, COCH3), 61.83, 69.56, 72.64, 74.52,
75.50 (5 C, C7, 6ꢀ, CH2Ph), 63.52, 68.27, 71.87, 71.98, 72.95, 78.00,
79.81, 82.74 (C3, 4, 5, 6, 2ꢀ, 3ꢀ, 4ꢀ, 5ꢀ), 99.80 (C1ꢀ), 122.39–138.38 (C1,
2, Ph), 169.13, 169.40, 170.21, 170.61 (4 C, C=O). HRMS m/z (FAB)
ꢀ
ꢀ
[14] H. Qin, T. B. Grindley, J. Carbohydr. Chem. 1996, 15, 95.
[15] M. Trumtel, P. Tavecchia, A. Veyrières, P. Sinaÿ, Carbohydr.
Res. 1990, 202, 257. doi:10.1016/0008-6215(90)84084-8
[16] D. L. J. Clive, Y. Tao, Y. Bo, Y.-Z. Hu, N. Selvakumar,
S. Sun, S. Daigneault, Y.-J. Wu, Chem. Commun. 2000, 1341.
doi:10.1039/B001204L
[17] J. Golik, H. Wong, B. Krishnan, D. M. Vyas, T. W. Doyle,
Tetrahedron Lett. 1991, 32, 1851. doi:10.1016/S0040-
4039(00)85979-1
[18] J. Banoub, D. R. Bundle, Can. J. Chem. 1979, 57, 2091.
[19] D. Kahne, S. Walker, Y. Cheng, D. Van Engen, J. Am. Chem.
Soc. 1989, 111, 6881.
[20] Y. Miyamoto, S. Ogawa, J. Chem. Soc., Perkin Trans. 1 1991,
2121.
[21] I. Rukhman, L.Yudovich, G. Nisnevich, A. L. Gutman, Synthesis
2000, 1241. doi:10.1055/S-2000-6413
[22] F. H. Newth, S. D. Nicholas, F. Smith, L. F. Wiggins, J. Chem.
Soc. 1949, 2550.
[23] K. Takeo, T. Fukatsu, T. Yasato, Carbohydr. Res. 1982, 107, 71.
doi:10.1016/S0008-6215(00)80776-4
[24] V. Ferro, M. Mocerino, R. V. Stick, D. M. G. Tilbrook, Aust. J.
Chem. 1988, 41, 813.
•
802.3198 [(M + H)+ requires 802.3187].
Acknowledgments
We thank Professor Steve Withers (University of British
Columbia, Canada) for providing the mutant enzyme
(glycosynthase), Professor Marc Claeyssens (University of
Gent, Belgium) for assistance in the collection of enzyme
inhibition data, and the Australian Research Council for
financial support.
References
[1] J. C. McAuliffe, R. V. Stick, D. M. G. Tilbrook, A. G. Watts,
Aust. J. Chem. 1998, 51, 91.
[2] B. Junge, F.-R. Heiker, J. Kurz, L. Müller, D. D. Schmidt,
C. Wünsche, Carbohydr. Res. 1984, 128, 235. doi:10.1016/0008-
6215(84)85333-1
[3] J. M. Macdonald, R. V. Stick, D. M. G. Tilbrook, S. G. Withers,
Aust. J. Chem. 2002, 55, 747. doi:10.1071/CH02165
[4] D. M. L. Lai, P. B. Høj, G. B. Fincher, Plant Mol. Biol. 1993,
22, 847.
[25] J. M. Küster, I. Dyong, Liebigs Ann. Chem. 1975, 2179.
[26] (a) M. Hrmova, T. Imai, S. J. Rutten, J. K. Fairweather, L. Pelosi,
V. Bulone, H. Driguez, G. B. Fincher, J. Biol. Chem. 2002, 277,
30102. doi:10.1074/JBC.M203971200
[5] J. C. McAuliffe, R. V. Stick, Aust. J. Chem. 1997, 50, 193.
doi:10.1071/C96151
(b) G. B. Fincher, M. Hrmova, J. M. Macdonald, R. V. Stick,
unpublished results.