Synthesis of (±)-isofagomine and its stereoisomers from arecoline

Article abstract of DOI:10.1039/a909472e

(±)-Isofagomine (1) and all its stereoisomers were prepared in a short synthesis from arecoline. The double bond of arecoline was isomerised to be no longer conjugated to the carboxy, and the ester reduced to (hydroxymethyl)tetrahydropyridine 5 which afte

Full text of DOI:10.1039/a909472e

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis of ( )-isofagomine and its stereoisomers from arecoline
Steen Uldall Hansen and Mikael Bols*
Department of Chemistry, Aarhus University, Langelandsgade 140, DK-8000 Aarhus,
Denmark. E-mail: mb@kemi.aau.dk; Tel: ϩ45 89 42 39 64; Fax: ϩ45 86 19 61 99
1
Received (in Lund, Sweden) 26th November 1999, Accepted 23rd January 2000
Published on the Web 24th February 2000
( )-Isofagomine (1) and all its stereoisomers were prepared in a short synthesis from arecoline. The double bond of
arecoline was isomerised to be no longer conjugated to the carboxy, and the ester reduced to (hydroxymethyl)tetrahydro-
pyridine 5 which after dihydroxylation of the alkene, separation and N-demethylation gave 1 and its isomers.
Introduction
It is becoming increasingly evident that sugar derivatives, in
which the anomeric carbon has been replaced with nitrogen, are
transition state analogues of enzymatic glycoside cleavage, as
seen by the fact that these compounds are strong inhibitors of
glycosidases and related enzymes.1 Thus isofagomine (1) is the
strongest known inhibitor of β-glucosidase,2 possibly because it
resembles the cation 2. Isofagomine and its stereoisomers are
therefore important building blocks for rationally designed
inhibitors. The synthesis of isofagomine analogues is however
relatively long, and an alternative shorter synthesis is of inter-
est. Arecoline (3), a natural product that is isolated from palm
trees and is commercially available, resembles these compounds
and could therefore potentially be a useful starting material for
the synthesis of isofagomine and its analogues. In the present
paper we report a very short synthesis of all the isofagomine
stereoisomers, in racemic form, starting from arecoline.
Scheme 1 Synthesis of tetrahydropyridine 6 from arecoline.
Results and discussion
It is known that arecoline (3) can be demethylated in high yields
with various reagents including α-chloroethyl chloroformate3
and 2,2,2-trichloroethyl chloroformate (TrocCl).4 It is also
known that the double bond in arecoline can be isomerised.4,5 It
therefore seemed straightforward to perform these two steps
and reduce the ester, thereby obtaining a potential precursor
to isofagomine and its galacto analogue. Arecoline could be
isomerised by following the known procedure4 to give a 20:1
ratio of 4 and arecoline (3) in a quantitative yield (Scheme 1).
The isomerized product 4 could not be separated chrom-
atographically from the 5% content of arecoline (3) and was
therefore used as such in the next step.
Demethylation of 4 was tried with 2,2,2-trichloroethyl
chloroformate (TrocCl) in toluene at 80 ЊC, but gave a poor
yield (20%) of the corresponding N-2,2,2-trichloroethylcarb-
amate apparently due to the occurence of side reactions. In
general, the success of demethylation of the piperidine deriv-
atives was found to be very structure-dependent as has been
seen by others.6 Thus arecoline (3) itself could be demethylated
in good yield with TrocCl;4 however that reaction was not use-
ful in the present case, because the double-bond isomerisation
failed on the N-trichloroethylcarbamate, as the Troc group is
base-labile.
Instead, reduction of 4 was carried out using lithium
aluminium hydride giving the alcohol 5 in 89% yield (Scheme
1). Demethylation of 5 with 2.1 equivalents of TrocCl in
toluene at 80 ЊC was then successful, giving a 60% yield of
the N,O-bis(Troc)-protected derivative 6. The carbonate could
selectively be saponified with K2CO3 in water–MeOH, giving
the mono N-Troc derivative 7 in 92% yield. The primary alcohol
functionality in 7 was protected with a tert-butyldiphenyl-
silyl (TBDPS) group by treatment with tert-butyldiphenylsilyl
chloride (TBDPSCl) and imidazole to give the silyl ether 8 in
97% yield.
Recently Altenbach and Blanda have reported the syn-
thesis of the N-phenylcarbamate analogue of 7, using
methyl N-benzyl-4-oxopiperidine-3-carboxylate as the starting
material.7 They were also able to enzymatically resolve this
analogue of 7 using enantioselective esterification with a
lipase. Our strategy provides an alternative high yielding route
DOI: 10.1039/a909472e
J. Chem. Soc., Perkin Trans. 1, 2000, 911–915
This journal is © The Royal Society of Chemistry 2000
911

View Article Online
Table 1 Results of epoxidation or dihydroxylation of 6–8. 6: R =
COOCH2Cl3, 7: R = H, 8: R = TBDPS
The Os-catalysed dihydroxylation, which would give the syn-
diol, was also attempted (Table 1). Dihydroxylation of 6 gave
a 1:1 mixture of syn:anti products that, as in the case of the
epoxides 9, was inseparable.
R-group
Reagent
Yield (%)
anti:syn ratio
Since good selectivity could not be obtained, a 1:1 mixture
of the epoxides 9 was hydrolysed with a 2.3% HClO4 solution at
100 ЊC. This gave an inseparable mixture of triols 10 and 11 in
68% yield. These were subjected to HCl hydrolysis to give a
mixture of racemic isofagomine (1) and its gulo isomer 12.
These could be separated chromatographically to give 1 and 12
in 14 and 44% yield, respectively. Only the -enantiomer of
isofagomine2 and the -enantiomer of 1210 are known.
The lack of selectivity in the addition reactions to the alkenes
6–8 was puzzling, since the literature contains a number of
examples where alkenes of related structure are dihydroxylated
with high anti selectivity.11 Thus 3-methylcyclohexene gives only
the anti diastereomer when reacted with OsO4.11b Since the main
difference between substrates 6–8 and cyclohexene is the geom-
etry at the amide nitrogen, it was suggested that dihydroxyl-
ation of 5 might give a higher selectivity. Indeed, treatment of 5
with OsO4 and NMO gave, after peracetylation, a 5:1 mixture
of the anti and syn triacetates 13 and 14. The diastereomers
could be separated to give pure 13 and 14 in 44 and 9%
yield, respectively, thus a combined yield of 53% (Scheme 3).
H
CF3(CH3)CO2
3-ClC6H4CO3H
CF3(CH3)CO2
(CH3)3C(CH3)CO2
Fructosedioxiranea
3-ClC6H4CO3H
CF3(CH3)CO2
OsO4/NMO
91
72
90
42
54
82
1:1
2:1
3:2
5:2
3:1
1:1
CCl3CH2OCO
CCl3CH2OCO
CCl3CH2OCO
CCl3CH2OCO
(CH3)3CPh2Si
(CH3)3CPh2Si
CCl3CH2OCO
Quantitative 2:1
86b
1:1
a Dioxirane generated from Oxone and 1,2:4,5-di-O-isopropylidene-
-arabino-hexo-2,6-pyran-2,3-diulose according to ref. 9. b Yield of
corresponding syn-diols.
Scheme 2 Epoxidation of alkenes 6–8. 6: R = COOCH2Cl3, 7: R = H,
8: R = TBDPS. R1 or R2 = Me, CF3, t-Bu or diisopropylidenefructose.
Scheme 3 Dihydroxylation of 5.
to similar 3-hydroxymethyl-1,2,3,6-tetrahydropyridine deriv-
atives, which also should be good substrates for enzymatic
resolution.
Deacetylation of 13 was performed with NaOMe in MeOH to
provide 15 in 90% yield. Similar deacetylation of 14 gave 16 in
93% yield.
Methods for dihydroxylation of the piperidine alkenes were
investigated next (Scheme 2, Table 1). If epoxidation could be
carried out from the less hindered side, the resulting trans
oriented epoxide would be expected to give the gluco-type trans,
trans-diol upon subsequent hydrolysis. Epoxidation of 6 was
carried out with either trifluoromethylmethyldioxirane8 or
m-chloroperbenzoic acid (MCPBA), however the diastereo-
selectivity was 1.5 or less. The anti selectivity could be improved
to 3 by using dioxiranes prepared from the more sterically
hindered ketones tert-butyl methyl ketone and 1,2:4,5-di-O-
isopropylidene--arabino-hexo-2,6-pyran-2,3-diulose,9 but the
yields of these reactions were considerably lower. When the
most reactive of these reagents, trifluoromethylmethyldi-
oxirane, was used on 7, a lower selectivity was obtained relative
to epoxidation of 6 (anti:syn 1:1) with the same reagent. In
contrast, similar epoxidation of 8 gave a somewhat higher anti
selectivity (anti:syn 2:1, Table 1). These results fit well with the
relative sterical hindrance of the allylic substituent in the sub-
strates. The diastereomers of 9 could not be separated chrom-
atographically. The diastereomeric ratios were determined by
NMR.
The isomers 13 and 14 were also demethylated. This was
done by treatment of each compound with 1-chloroethyl
chloroformate in dichloroethane, giving a 1-chloroethyl carb-
amate that was deprotected with MeOH and NaOMe to obtain
either allo-isofagomine 17 or galacto-isofagomine 18 both in
81% yield.
The compounds 12 (one enantiomer of 12 is known10) and
15–17 are new and were therefore tested for inhibition of
glycosidases (Table 2). Particularly noteworthy was the observ-
ation that ( )-16 is a relatively poor β-galactosidase inhibitor
compared with galacto-isofagomine (18), which has been
reported to be extremely potent.12 The fact that 16 is racemic
and 18 sterically pure does not affect this observation, since
even if one antipode of 16 is completely inactive the other will
only be twice as potent (which is still much less than the
potency of 18). It was also noteworthy that ( )-17 was a
rather potent inhibitor of galactosidase, and that its N-methyl
analogue ( )-15 is a poor inhibitor. It suggests that N-methyl-
ation decreases glycosidase inhibition.
In conclusion we have reported a short synthesis of three of
the four possible diastereomers of isofagomine starting from
912
J. Chem. Soc., Perkin Trans. 1, 2000, 911–915

View Article Online
Table 2 Glycosidase inhibition by 12, 15, 16 and 17. Figures show Ki values in µM
Ki/µM
Compound
( )-12
( )-15
( )-16
( )-17
β-Glucosidase
α-Glucosidase
β-Galactosidase
α-Galactosidase
22
>1000
405a
>1000
>1000
530
235
>1000
77
76
>1000
9
>1000
>1000
74
37
a Non-competitive (or mixed) inhibition.
arecoline, and also the synthesis of the N-methyl analogue of
the fourth diastereomer. Arecoline has been found to be a use-
ful source for the synthesis of these compounds in a racemic
form. The method may also be extended to the synthesis of the
enantiomerically pure compounds, since the enzymatic reso-
lution method reported by Altenbach and Blanda7 probably
can be applied to substrates 5 and 7.
room temperature. The mixture was filtered, the filtrate washed
with HCl (1 M, 50 mL), NaOH (1 M, 50 mL), water (50 mL),
dried and evaporated to leave a crude product. Flash chromato-
graphy in CH2Cl2–pentane (1:1) gave 6. Yield: 6.55 g (60%). 1H
NMR (CDCl3): δ 5.79 (m, 2H, H4, H5), 4.73 (m, 4H, CH2-
CCl3’s), 4.2–3.5 (m, 6H, H3Ј, H2, H6), 2.69 (m, 1H, H3). 13C
NMR (CDCl ): δ 154.4, 154.2, 154.0 (rotamer C᎐O’s), 127.7,
᎐
3
126.9, 125.0, 124.6 (rot. C4, rot. C5), 96.0, 94.8 (CH2CCl3’s),
78.2 (CH2CCl3), 75.6, 75.4 (rot. CH2CCl3), 69.5, 69.0 (rot.
C3Ј), 44.3, 43.9, 42.9, 42.8 (rot. C2, rot. C6), 35.3, 35.1 (rot.
C3). HRMS (ES): m/z 483.8815 (M ϩ Naϩ), calcd for C12H13-
NO5Cl6 ϩ Naϩ: 483.8823.
Experimental
General
Solvents were distilled under anhydrous conditions. Thus THF
was distilled from sodium–benzophenone and used directly. All
reagents were used as purchased without further purification.
Columns were packed with silica gel 60 (230–400 mesh) as the
stationary phase. TLC-plates (Merck, 60F254) were visualised
either by spraying with ninhydrin (2% in butanol) or CEMOL
(1% ceric sulfate and 1.5% ammonium molybdate in 10%
( )-3-Hydroxymethyl-1-(2,2,2-trichloroethoxycarbonyl)-1,2,3,6-
tetrahydropyridine (7)
To 1.93 g (4.16 mmol) of 6 was added 10 mL MeOH and a
solution of K2CO3 (5 g) in 10 mL water. After stirring for 1
hour the mixture was extracted with ether (2 × 20 mL). The
combined organic phases were washed with HCl (1 M, 10 mL),
NaOH (1 M, 2 × 10 mL), water (10 mL), dried and evaporated
aqueous H2SO4) and heating until coloured spots appeared. 13
C
1
and H NMR spectra were recorded on a Varian Gemini 200
instrument. Mass spectra were obtained on a Micromass LCT-
QTOF instrument. Products were concentrated on a rotary
evaporator below 40 ЊC.
1
to leave 7 as an oil. Yield: 1.11 g (92%). H NMR (CDCl3):
δ 5.75 (m, 2H, H4, H5), 4.73 (m, 2H, CH2CCl3), 4.1–3.4 (m,
6H, H3’s, H2’s, H6’s), 2.47 (m, 1H, H3). 13C NMR (CDCl3):
δ 155.0, 154.1 (rot. C᎐O), 126.8, 126.5, 126.3, 125.7 (rot. C4,
᎐
( )-3-Hydroxymethyl-1-methyl-1,2,3,6-tetrahydropyridine (5)
rot. C5), 96.1 (CCl3), 75.6, 75.5 (rot. CH2CCl3), 63.8, 63.5 (rot.
C3Ј), 44.5, 44.3, 43.1, 42.9 (rot. C2, rot. C6), 38.5, 38.4 (rot.
C3). HRMS (ES): m/z 309.9776 (M ϩ Naϩ), calcd for C9H12-
NO3Cl3 ϩ Naϩ: 309.9781.
A solution of 9.50 g (61.2 mmol) of 4 (containing 5% of 3) in
25 mL of THF was added during 60 min to 2.32 g (61.2 mmol)
of LiAlH4 in 35 mL of THF at 0 ЊC. After stirring for another
90 min at room temperature, the mixture was quenched by the
careful addition of water (50 mL). Extraction with EtOAc
(8 × 50 mL), drying and evaporation left 5 as a yellow oil.
Yield: 6.90 g (89%). This substance contained however a 5%
impurity of 4. This could be removed by purification with flash
chromatography in EtOAc–MeOH 5:1. 1H NMR (CDCl3):
δ 5.79 (m, 1H, H5, J5,4 10.3 Hz), 5.64 (m, 1H, H4), 4.46 (br s, 1H,
3Ј-OH), 3.76 (dd, 1H, H3Јa, J3Јa,3Јb 10.3, J3Јa,3 3.7 Hz), 3.58
(ddd, 1H, H3Јb, J3,3Јb 3.7 Hz, J3Јb,3ЈOH 1.5 Hz), 3.06 (m, 1H,
H6eq, J6eq,6ax 16.1, J6eq,5 4.0, J6eq,4 1.1 Hz), 2.65 (m, 2H, H2ax,
H6ax), 2.49 (dd, 1H, H2eq, J2eq,2ax 11.0, J2eq,3 4.0 Hz), 2.27
(m, 1H, H3), 2.26 (s, 3H, 1-Me). 13C NMR (CDCl3): δ 127.5,
126.5 (C4, C5), 66.7 (C3Ј), 57.1, 54.8 (C2, C6), 46.0 (C1Ј),
37.7 (C3). HRMS (ES) m/z: 150.0889 (M ϩ Naϩ), calcd for
C7H13NO ϩ Naϩ: 150.0895.
( )-3-(tert-Butyldiphenylsiloxy)methyl-1-N-(2,2,2-trichloro-
ethoxycarbonyl)-1,2,3,6-tetrahydropyridine (8)
To 0.560 g (1.94 mmol) of 7 in 2 mL DMF was added 0.246 g
(3.88 mmol) of imidazole and 510 µL (1.96 mmol) of tert-
butyldiphenylsilyl chloride. The solution was stirred for 16 h
under nitrogen. Quenching with water (10 mL) was followed
by extraction with pentane (2 × 10 mL), drying (MgSO4)
1
and evaporation, leaving 8 as an oil. Yield: 0.995 g (97%). H
NMR (CDCl3): δ 7.20 (m, 10H, phenyl), 5.58 (m, 2H, H4, H5),
4.62 (m, 2H, CH2CCl3), 4.2–3.2 (m, 6H, H3’s, H2’s, H6’s),
2.42 (m, 1H, H3), 0.98 (s, 9H, tert-butyl). 13C NMR (CDCl3):
δ 153.3, 153.2 (rot. C᎐O), 135.0, 132.3, 129.2, 127.1 (phenyl),
᎐
126.2, 125.7, 124.5, 124.0 (rot. C4, rot. C5), 95.1 (CCl3), 74.5
(CH2CCl3), 64.6, 64.2 (rot. C3Ј), 43.4, 43.0, 42.3 (rot. C2,
rot. C6), 37.6, 37.2 (rot. C3), 26.1 (tert-butyl). HRMS (ES):
m/z 548.0964 (M ϩ Naϩ), calcd for C25H30NO3SiCl3 ϩ Naϩ:
548.0958.
( )-3-(2,2,2-Trichloroethoxycarbonyloxymethyl)-1-(2,2,2-tri-
chloroethoxycarbonyl)-1,2,3,6-tetrahydropyridine (6)
To 3.02 g (23.7 mmol) of 5 in 30 mL toluene was added 4.13 mL
(23.7 mmol) of diisopropylethylamine. Over a period of 15 min
a solution of 10.54 g (49.8 mmol) 2,2,2-trichloroethyl chloro-
formate in 10 mL toluene was added. The mixture was heated
to 80 ЊC for 20 hours. A solid precipitate formed on cooling to
Epoxidation or dihydroxylation of 6–8. General procedure
a) Epoxidations using MCPBA. To 1 mmol of the alkene in
4 mL dichloromethane was added 1.2–1.5 mmol of MCPBA,
J. Chem. Soc., Perkin Trans. 1, 2000, 911–915
913

View Article Online
and the solution was stirred for 20 h. The organic phase was
washed with NaHSO3 (10%), NaHCO3 (sat.), water, dried
(MgSO4) and evaporated. The inseparable diastereomeric
mixture of epoxides was purified by flash chromatography if
necessary.
b) Epoxidations using dioxiranes. To 1 mmol of the alkene
in a mixture of water–CH3CN (2:3, 10 mL) was added 1,1,1-
trifluoroacetone (10 mmol) and NaHCO3 (8 mmol). To this
mixture at 0 ЊC was added Oxone (6 mmol) in small portions
during 10 min. The solution was stirred at room temperature
for 1–20 h. Quenching with water (20 mL), extraction with
dichloromethane (2 × 30 mL), drying and evaporation gave the
inseparable diastereomeric mixture of epoxides, which could be
purified by flash chromatography if necessary.
calcd for C13H21NO6 ϩ Hϩ: 288.1447. Yield of 14: 70 mg (9%).
1H NMR (CDCl3): δ 5.41 (m, 1H, H4), 4.98 (ddd, 1H, H3, J3,4
3.3, J3,2eq 4.3, J3,2ax 10.1 Hz), 4.04 (dd, 1H, H5Јa, J5Јa,5Јb 10.3,
J5Јa,5 7.7 Hz), 3.90 (dd, 1H, H5Јb, J5Јb,5 6.7 Hz), 2.1–2.7 (m, 5H,
H2’s, H5, H6’s), 2.36 (s, 3H, Me), 2.09, 2.03, 2.00 (3s, 9H, Ac’s).
13C NMR (CDCl ): δ 170.2, 169.4, 169.3 (C᎐O’s), 68.6, 65.6
᎐
3
(C3, C4), 61.3 (C5Ј), 52.4, 51.3 (C2, C6), 44.9 (Me), 37.0 (C5),
20.1 (Ac’s). HRMS (ES): m/z 310.1288 (M ϩ Naϩ), calcd for
C13H21NO6Naϩ: 310.1267.
(3,4-cis-4,5-trans)-5-Hydroxymethyl-1-methyl-3,4-dihydroxy-
piperidine (15)
To 70 mg (0.25 mmol) of 13 in 2 mL dry MeOH was added a
small piece of sodium. After stirring for 1 h the solution was
evaporated. Flash chromatography in EtOH–NH4OH (25%,
1
(3,4-trans-4,5-trans)-5-Hydroxymethyl-1-(2,2,2-trichloroethoxy-
carbonyl)-3,4-dihydroxypiperidine (10) and (3,4-trans-4,5-cis)-5-
hydroxymethyl-1-(2,2,2-trichloroethoxycarbonyl)-3,4-dihydroxy-
piperidine (11)
9:1) gave 15. Yield: 36 mg (90%). H NMR (D2O): δ 3.65 (dd,
1H, H3, J3,4 4.0, J3,2 5.9 Hz), 3.49 (dd, 1H, H4, J4,5 11.4 Hz),
3.33 (dd, 1H, H5Јa, J5Јa,5Јb 11.0, J5Јa,5 6.2 Hz), 3.29 (dd, 1H,
H5Јb, J5Јb,5 3.3 Hz), 2.58, 2.1–1.7 (m, 5H, H2’s, H6’s, H3),
1.97 (s, 3H, Me). 13C NMR (D2O): δ 68.3, 66.4 (C3, C4), 60.1
(C5Ј), 56.9, 54.0 (C2, C6), 43.4 (C1Ј), 37.9 (C5). HRMS
(ES): m/z 162.1143 (M ϩ Hϩ), calcd for C7H15NO3 ϩ Hϩ:
162.1130.
To 1.062 g (3.69 mmol) of 9 (R = H, syn:anti ratio 1:1) in 30
mL water was added 1 mL HClO4 (70%). The solution was
heated to reflux for 90 min and then brought to a neutral pH by
addition of K2CO3. Salts were removed by filtration, and the
filtrate extracted with EtOAc (5 × 25 mL). Drying and evapor-
ation left a 1:1 ratio of 10 and 11 as a mixture of inseparable
diastereomers. Yield: 0.810 g (68%).
(3,4-cis-4,5-cis)-5-Hydroxymethyl-1-methyl-3,4-dihydroxypiper-
idine (16)
To 62 mg (0.22 mmol) of 14 in 2 mL dry MeOH was added a
small piece of sodium. After stirring for 1 h the solution was
evaporated. Flash chromatography in EtOH–NH4OH (25%)
(3,4-trans-4,5-trans)-5-Hydroxymethyl-3,4-dihydroxypiperidine
(1)2a and (3,4-trans-4,5-cis)-5-hydroxymethyl-3,4-dihydroxy-
piperidine (12)10
1
(9:1) gave 16. Yield: 33 mg (93%). H NMR (D2O): δ 3.74 (t,
1H, H4, J4,3 and J4,5 2.2), 3.49 (ddd, 1H, H3, J3,2ax 11.4 Hz, J3,2eq
5.1 Hz), 3.41 (dd, 1H, H5Јa, J5Јa,5Јb 11.0, J5Ј,5 7.0 Hz), 3.29 (dd,
1H, H5Јb, J5,5Јb 6.7 Hz), 2.49 (dd, 1H, H2eq, J2eq,2ax 11.0 Hz),
2.38 (dd, 1H, H6eq, J6eq,6ax 11.0, J6eq,5 3.3 Hz), 2.04 (s, 3H, Me),
2.00 (t, 1H, H2ax), 1.78 (t, 1H, H6ax, J6ax,5 11.1 Hz), 1.68 (m,
1H, H5). 13C NMR (D2O): δ 67.1, 65.4 (C3, C4), 59.6 (C5Ј),
52.6, 49.0 (C2, C6), 43.1 (Me), 39.8 (C5). HRMS (ES): m/z
162.1119 (M ϩ Hϩ), calcd for C7H15NO3 ϩ Hϩ: 162.1130.
A solution of 8 mL aqueous HCl (3 M) and 0.142 g (0.44
mmol) of the mixture of 10 and 11 (ratio, 1:2) was heated to
reflux for 24 h. After evaporation, ( )-isofagomine (1) was sep-
arated from the gulo isomer (12) by flash chromatography in
EtOH–NH4OH (1%) (72:28). The yield of pure 1 was 12 mg
(14%), while 39 mg of pure 12 (44%) were isolated together with
11 mg of a mixture of 1 and 12. Total yield: 62 mg (70%). NMR
data of 1 were identical to literature values.2a Data for the
hydrochloride of 12:10 1H NMR (D2O): δ 3.77 (m, 2H, H3, H4),
3.46 (dd, 1H, H5Јa, J5Јa,5Јb 11.4, J5Јa,5 6.6 Hz), 3.32 (dd, 1H,
H5Јb, J5,5Јb 7.0 Hz), 3.03 (m, 3H, H2’s, H6eq), 2.75 (t, 1H,
H6ax, J6ax,6eq 12.5, J6ax,5 12.5 Hz), 2.14 (m, 1H, H5). 13C NMR
(D2O): δ 64.0, 63.9 (C3, C4), 59.0 (C5Ј), 43.0, 39.4 (C2, C6),
33.7 (C5). HRMS (ES): m/z 170.0800 (M ϩ Naϩ), calcd for
C6H13NO3Naϩ: 170.0793.
(3,4-cis-4,5-trans)-5-Hydroxymethyl-3,4-dihydroxypiperidine
(17)
To 78 mg (0.27 mmol) of 13 in 2 mL of dry 1,2-dichloroethane
was added 50 µL (0.4 mmol) of 1-chloroethyl chloroformate
and 50 µL of ethyldiisopropylamine. The solution was heated
to reflux for 3 h. After evaporation the residue was dissolved in
2 mL of dry MeOH and heated to reflux for 1 h. Heating was
stopped and a small piece of sodium was added. After stirring
for 1 h the solvent was evaporated, and the residue purified by
flash chromatography in EtOH–NH4OH (25%) 9:1. Yield: 32
mg (81%). 1H NMR (D2O): δ 3.60 (dd, 1H, H3, J3,4 2.9, J3,2 4.8
Hz), 3.40 (m, 3H, H4, H5’s), 2.77 (m, 2H, H2eq, H6eq), 2.40
(dd, 1H, H2ax, J2ax,2eq 14.3 Hz), 2.10 (dd, 1H, H6ax, J6ax,6eq
13.2, J6ax,5 11.0 Hz), 1.66 (m, 1H, H5). 13C NMR (D2O): δ 68.7,
66.3 (C3, C4), 59.8 (C5), 47.5, 44.4 (C2, C6), 38.4 (C5). HRMS
(ES): m/z 170.0786 (M ϩ Naϩ), calcd for C6H13NO3Naϩ:
170.0793.
(3,4-cis-4,5-trans)-5-Acetoxymethyl-1-methyl-3,4-diacetoxy-
piperidine (13) and (3,4-cis-4,5-cis)-5-acetoxymethyl-1-methyl-
3,4-diacetoxypiperidine (14)
To 0.334 g (2.63 mmol) of 5 in acetone–water (1:1, 4 mL) was
added 0.886 g (6.56 mmol) 4-methylmorpholine N-oxide
monohydrate. To this solution was added by syringe 1.67 mL
(2.5 mol%) of a 1% solution of OsO4 in tert-butyl alcohol and it
was stirred for 5 days. NaHSO3 was added in excess and the
mixture was stirred for 30 min. The mixture was purified by ion
exchange giving a crude product of inseparable diastereomers.
To allow separation, the mixture was acetylated by dissolving in
pyridine (3 mL) and acetic anhydride (3 mL), and was stirred
for 45 min. After evaporation the two diastereomers 13 and 14
could be separated by flash chromatography in EtOAc. Yield of
(3,4-cis-4,5-cis)-5-Hydroxymethyl-3,4-dihydroxypiperidine (18)
To 12 mg (0.042 mmol) of 14 in 1 mL of dry 1,2-dichloroethane
was added 20 µL (0.16 mmol) of 1-chloroethyl chloroformate
and 20 µL of ethyldiisopropylamine. The solution was heated
to reflux for 3 h. After evaporation the residue was dissolved in
1 mL of dry MeOH and heated to reflux for 1 h. The heating
was stopped and a small piece of sodium was added. After
stirring for 1 h the solvent was evaporated, and the residue
purified by flash chromatography in EtOH–NH4OH (25%) 9:1.
Yield: 5 mg (81%). NMR data for 18 were identical to literature
values.12
1
13: 331 mg (44%). H NMR (CDCl3): δ 5.31 (dd, 1H, H3, J3,2
5.1, J3,4 3.3 Hz), 4.77 (dd, 1H, H4, J4,5 11.0 Hz), 4.06 (d, 1H,
H5Јa, J5Јa,5 3.7 Hz), 4.04 (d, 1H, H5Јb, J5Јb,5 5.2 Hz), 3.0–2.1 (m,
5H, H2’s, H6’s, H3), 2.31 (s, 3H, Me), 2.14, 2.06, 2.03 (s, 3H,
Ac’s). 13C NMR (CDCl ): δ 170.2, 169.9, 169.6 (C᎐O), 69.4,
᎐
3
66.7 (C3, C4), 61.9 (C5Ј), 56.4, 55.9 (C2, C6), 45.0 (Me), 34.9
(C5), 20.4, 20.1 (Ac’s). HRMS (ES): m/z 288.1442 (M ϩ Hϩ),
914
J. Chem. Soc., Perkin Trans. 1, 2000, 911–915

View Article Online
Enzymatic assays
References
The substrates 4-nitrophenyl α- and β--glucopyranoside, 4-
nitrophenyl α--galactopyranoside and 2-nitrophenyl β--
galactopyranoside, and the enzymes α-glucosidase (yeast,
Sigma G 5003), α-galactosidase (Green Coffee Bean), β-galacto-
sidase (Aspergillus Oryzae) and β-glucosidase (sweet almonds,
Sigma G 4511) were purchased from Sigma Chemical Co. All
assays were carried out in a sodium phosphate buffer (0.05 M,
pH 6.8) at 25 ЊC. Formation of the product, 4-nitrophenol, was
measured continually at 400 nm using a Milton Roy Genesys 5
spectrometer. In all kinetic runs less than 1% of the initial sub-
strate was consumed assuring the constancy of the substrate
concentration.
1 M. Bols, Acc. Chem. Res., 1998, 31, 1.
2 (a) T. M. Jespersen, M. Bols, M. R. Sierks and T. Skrydstrup,
Tetrahedron, 1994, 50, 13449; (b) W. Dong, T. Jespersen, M. Bols,
T. Skrydstrup and M. R. Sierks, Biochemistry, 1996, 35, 2788.
3 R. A. Olofson, J. T. Martz, J.-P. Senet, M. Piteau and T. Malfroot,
J. Org. Chem., 1984, 49, 2081.
4 R. D. Allan and J. Fong, Aust. J. Chem., 1983, 36, 601.
5 C. Müller-Uri, E. A. Singer and W. Fleischhacker, J. Med. Chem.,
1986, 29, 125.
6 G. Zhao and B. Ganem, J. Org. Chem., 1997, 62, 2298.
7 H.-J. Altenbach and G. Blanda, Tetrahedron: Asymmetry, 1998, 9,
1519.
8 D. Yang, M.-K. Wong and Y.-C. Yip, J. Org. Chem., 1995, 60,
3887.
Ki determinations were performed as follows. Two thermo-
statted solutions of 1) 1 mL of 0.1 M buffer, 800 µL substrate in
varied concentration ([S] = 0.25 KM to 4 KM) and 100 µL water
and 2) 100 µL enzyme were mixed, and the reaction was
immediately monitored. From 5 experiments with varied sub-
strate concentrations, initial reaction rates were calculated from
the slope of the 1st order plot of product absorption vs reaction
time. KM and Vmax were calculated from a Hanes plot. From 5
experiments KMЈ was calculated from the Hanes plot and from
KM and KMЈ, Ki was calculated.
9 Y. Tu, Z.-X. Wang and Y. Shi, J. Am. Chem. Soc., 1996, 118,
9806.
10 U. Katzmaier and C. Schneider, Tetrahedron Lett., 1998, 39, 817.
11 (a) B. B. Snider and K. Yang, J. Org. Chem., 1990, 55, 4392;
(b) J. K. Cha, W. J. Christ and Y. Kishi, Tetrahedron Lett., 1983, 24,
3943.
12 Y. Ichikawa, Y. Igarashi, M. Ichikawa and Y. Suhura, J. Am. Chem.
Soc., 1998, 120, 5854.
Paper a909472e
J. Chem. Soc., Perkin Trans. 1, 2000, 911–915
915