5-epi-Deoxyrhamnojirimycin is a potent inhibitor of an α-L- rhamnosidase: 5-epi-Deoxymannojirimycin is not a potent inhibitor of an α- D-mannosidase

Article abstract of DOI:10.1016/S0957-4166(98)00317-6

Whereas deoxyrhamnojirimycin (LRJ) 1 shows no significant inhibition of naringinase (an α-L-rhamnosidase), its C-5 epimer 2 is a potent and specific inhibitor of the enzyme and demonstrates the value of unambiguous chemical synthesis of such materials in the evaluation of their biological properties. In contrast, moderately weak inhibition towards an α-D-mannosidase is shown by both deoxymannojirimycin (DMJ) 5 and its C-5 epimer 6. Mimics of L- rhamnose which are recognised by enzymes that synthesise or process L- rhamnose may inhibit either the biosynthesis of the sugar or its incorporation into mycobacterial cell walls, providing new strategies for the treatment of diseases such as tuberculosis and leprosy. Molecular modelling studies provide a rationale for the surprisingly potent activity of the C-5 epimer 2 compared with LRJ 1 and support a general hypothesis that potent piperidine glycosidase inhibitors mimic the ⁴H₃ conformation of the relevant glycopyranosyl cation intermediate.

Full text of DOI:10.1016/S0957-4166(98)00317-6

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 9 (1998) 2947–2960
5-epi-Deoxyrhamnojirimycin is a potent inhibitor of an
α-L-rhamnosidase: 5-epi-deoxymannojirimycin is not a potent
inhibitor of an α-D-mannosidase
Benjamin G. Davis,a Andrew Hull,a Colin Smith,b Robert J. Nash,c Alison A. Watson,c
David A. Winkler,d Rhodri C. Griffiths c and George W. J. Fleet a,∗
aDyson Perrins Laboratory, Oxford University, South Parks Road, Oxford OX1 3QY, UK
bGlaxoWellcome Medicines Research Centre, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, UK
cInstitute of Grassland and Environmental Research, Plas Gogerddan, Aberystwyth SY23 3EB, UK
dCSIRO Division of Molecular Science, Private Bag 10, Clayton South MDC, Clayton 3169, Australia
Received 24 July 1998; accepted 5 August 1998
Abstract
Whereas deoxyrhamnojirimycin (LRJ) 1 shows no significant inhibition of naringinase (an α-L-rhamnosidase),
its C-5 epimer 2 is a potent and specific inhibitor of the enzyme and demonstrates the value of unambiguous
chemical synthesis of such materials in the evaluation of their biological properties. In contrast, moderately weak
inhibition towards an α-D-mannosidase is shown by both deoxymannojirimycin (DMJ) 5 and its C-5 epimer 6.
Mimics of L-rhamnose which are recognised by enzymes that synthesise or process L-rhamnose may inhibit either
the biosynthesis of the sugar or its incorporation into mycobacterial cell walls, providing new strategies for the
treatment of diseases such as tuberculosis and leprosy. Molecular modelling studies provide a rationale for the
surprisingly potent activity of the C-5 epimer 2 compared with LRJ 1 and support a general hypothesis that potent
piperidine glycosidase inhibitors mimic the 4H3 conformation of the relevant glycopyranosyl cation intermediate.
© 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction
There exists a very wide range of naturally occurring1 and synthetic2 polyhydroxylated piperidines and
pyrrolidines that mimic individual sugar moieties. Piperidine analogues of carbohydrates in which the
ring oxygen of a sugar is replaced by nitrogen and the anomeric hydroxyl group is replaced by hydrogen
are usually effective inhibitors of the corresponding glycosidases,3 but a change of configuration of one
∗
Corresponding author.
0957-4166/98/$ - see front matter © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00317-6

2948
B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
of the ring carbon atoms usually causes a dramatic loss of glycosidase inhibition. Thus deoxynojirimycin
(DNJ) is a very good inhibitor of many α-glucosidases whereas 5-epi-DNJ is only a very weak inhibitor.4
Accordingly, L-deoxyrhamnojirimycin (LRJ) 1, the corresponding analogue of L-rhamnose, might be
predicted to inhibit naringinase [an α-L-rhamnosidase (EC 3.2.1.40)]. However, a paper reporting the
synthesis of LRJ 1 from D-gulonolactone stated that the compound had no significant inhibition of
naringinase.5 In contrast, it was claimed that LRJ prepared by an alternative procedure was a potent
inhibitor of naringinase with a Ki of 34 µM towards naringinase.6 The apparent discrepancy in these
results was investigated by Wong. He repeated both syntheses and found that, whilst the materials had
essentially identical physical data, the sample prepared from D-gulonolactone showed no significant
inhibition of naringinase (Ki 490 µM) whereas that prepared by a sequence involving an aldolase-
catalysed step7 was a good inhibitor of the enzyme (Ki 62 µM).8 Wong suggested that the difference
between the two materials might be due to traces of an impurity which was a strong inhibitor of
naringinase. On the basis of the possible ambiguities in the aldolase-mediated syntheses, he proposed
that the inhibition of the enzyme was due to the presence of a small, undetectable9 amount of 5-epi-LRJ
2,10 the C-5 epimer of 1.
Mimics of L-rhamnose, as well as — or in preference to — their ability to inhibit L-rhamnosidase
activity, may be recognised by enzymes which are involved in the biosynthesis of rhamnose or associated
with its incorporation into cell walls of mycobacteria. Such materials may provide novel chemothera-
peutic strategies for the treatment of diseases such as tuberculosis or leprosy.10a,11 L-Rhamnose is the
enantiomer of 6-deoxy-D-mannose; on the basis of known potent inhibitors of α-D-mannosidases, some
powerful inhibitors of L-rhamnosidase have been described.11d Deoxymannojirimycin (DMJ) 5, first
isolated from Lonchocarpus sericeus,12 is a potent inhibitor of mannosidase I of glycoprotein processing
but otherwise is a rather weak inhibitor of most other α-mannosidases;13 indeed, DMJ is generally a
more potent inhibitor of α-L-fucosidases than of mannosidases.14 By analogy with the relative qualities
of 1 and 2 as inhibitors of naringinase, it was possible that 5-epi-DMJ 615 may be a stronger inhibitor
in general of mannosidases than is DMJ 5. This paper reports the unambiguous synthesis of 5-epi-LRJ
2 and confirms Wong’s suggestion that 2 is a very good inhibitor of naringinase. The corresponding
lactam 3 and tetrazole 4 were also prepared as potential L-rhamnose mimics; other lactams and tetrazoles
have been shown to be effective glycosidase inhibitors.10a,11a,d,16 The preparation of epi-DMJ 6 from
mannose is also described; there is little difference between the ability of DMJ 5 and 5-epi-DMJ 6 to

B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
2949
inhibit mannosidases, but the latter is a significantly weaker inhibitor of α-L-fucosidases; some of this
work has appeared as a preliminary publication.10a
Molecular modelling studies are reported which provide a good rationale for the much stronger
inhibition of naringinase by the 5-epi compound 2 than by 1 itself. Further modelling studies may give
rise to predictions of novel structures that should be good rhamnosidase inhibitors and thus provide
guides for structures that interact with enzymes that handle L-rhamnose, and might give clues to new
approaches to the study of mycobacterial cell wall biosynthesis.
2. Synthesis
The synthesis of 5-epi-LRJ 2, the lactam 3, and the tetrazole 4 requires introduction of nitrogen with
inversion of configuration at C-5 of L-rhamnose (Scheme 1). Thus, L-rhamnose was first converted, as
previously described,17 to the lactone acetonide 817a in which only the side chain C-5 hydroxyl group
is unprotected. Esterification of 8 with triflic anhydride in the presence of excess pyridine gave the
corresponding triflate which, with sodium azide in DMF, afforded the fully protected azide 9 in 67%
yield. The azide 9 is the key divergent intermediate in the synthesis of all three 5-epi-L-rhamno targets 2,
3 and 4.
Scheme 1. (i) Tf2O/pyridine/CH2Cl2/−40°C; (ii) NaN3/DMF; (iii) H2/10%Pd–C/MeOH; (iv) CF3COOH:H2O (3:2);
(v) BH3·Me2S/THF then conc. HCl/EtOH; (vi) NH3/MeOH; (vii) (CF3CO)2O/pyridine/−30°C; (viii) ∆/toluene; (ix)
CF3COOH:H2O (1:1)
Hydrogenation of a solution of 9 in methanol in the presence of palladium on carbon caused reduction
to the corresponding amine which underwent intramolecular acylation to give the lactam 10 (74% yield).
Subsequent deprotection with aqueous trifluoroacetic acid gave the easily crystallised unprotected 1,5-
lactam 3 in 89% yield. The protected lactam 10 was reduced with borane:dimethylsulphide complex
in THF. Treatment of the crude product mixture with methanolic hydrogen chloride caused both the
decomposition of the borane adduct formed as well as removal of the acetonide protecting group to
produce, after purification by ion-exchange chromatography, 5-epi-LRJ 2 in 71% yield.
The bicyclic tetrazole 4 was synthesised in a manner essentially analogous to methods described
previously,10a,11a,d,16i and required the synthesis of an open chain 5-azidonitrile derivative 12 as a
substrate for a key thermally induced intramolecular [1,3]-dipolar cycloaddition. Reaction of azido-

2950
B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
lactone 9 with a methanolic solution of ammonia gave the ring opened primary amide 11 in 99% yield.
Dehydration of 11 was achieved by treatment with excess trifluoroacetic anhydride in pyridine, followed
by methanolic work-up, and gave the azidonitrile 12 (84% yield). Cyclisation of nitrile 12 to give the
protected tetrazole 13 (71% yield) was induced by heating in anhydrous toluene; the formation of a
small amount of azidolactone 9 (16% yield) was also observed. Hydroxynitriles are particularly prone to
hydrolysis; intramolecular closure of the hydroxyl group onto the electrophilic nitrile group leads to the
formation of a cyclic imidate ester which is readily converted to the corresponding lactone.18 Treatment
of protected tetrazole 13 with aqueous trifluoroacetic acid afforded target 5-epi-tetrazole 4 in 79% yield.
For the synthesis of 5-epi-DMJ 6 and the corresponding δ-lactam 7,19 diacetone mannose 14 was
converted to the azidolactone 15 as previously described (Scheme 2).20 Hydrogenation of the azide 15
afforded the protected lactam 16 via the corresponding 4-amino lactone in 80% yield. Deprotection of 16
to L-gulono-1,5-lactam 7 was achieved using aqueous trifluoroacetic acid in 90% yield. Reduction of the
protected lactam 16 using borane as its dimethylsulphide complex in THF followed by treatment with
aqueous acid in ethanol afforded 5-epi-DMJ 6 in 92% yield.
Scheme 2. (i) H2/10% Pd–C/EtOH; (ii) CF3COOH:H2O (2:1); (iii) BH3·Me2S/THF then conc. HCl/EtOH
2.1. Effect on glycosidases
The inhibitory effects of LRJ 1, 5-epi-LRJ 2, DMJ 5, and 5-epi-DMJ 6, the corresponding lactam
3, and the tetrazole derivative of 3 towards a panel of glycosidases have been studied. Only 5-epi-
LRJ 2 showed significant inhibition against an α-L-rhamnosidase from Penicillium decumbens with a
Ki of 1 µM (naringinase from Sigma). None of 1, 3, 4, 5 and 6 were inhibitory at 770 µM using p-
nitrophenyl α-L-rhamnopyranoside as substrate. 2 was weakly inhibitory to almond β-glucosidase at
970 µM (60% inhibition). 5-epi-DMJ 6 was a poor inhibitor of jack bean α-mannosidase (52%) and
bovine epididymis α-L-fucosidase (40%) at 770 µM; DMJ 5 has a Ki of 5 µM against the bovine
fucosidase used here and is reported as an inhibitor of mannosidases (Ki 110 µM against jack bean
mannosidase)14 but there was no inhibition by 775 µM concentrations of the compounds 1–6 of the
following enzymes: yeast α-glucosidase, green coffee bean α-galactosidase, E. coli β-galactosidase
and jack bean β-N-acetylglucosaminidase. The substrates were all 5 mM p-nitrophenylglycosides and
enzyme concentrations were 1.4 µg/ml in the assay mixture.14 All inhibition of enzymes shown by these
compounds is competitive.
The low level of inhibition of naringinase shown by LRJ 1 confirmed the results of previous assays of
chemical syntheses. Wong’s suggestion that the inhibition by the product of the aldolase-catalysed routes
was due to 5-epi-LRJ’s 2 potency is correct and shows that the inhibitory activity shown by enzymatically

B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
2951
synthesised samples of LRJ 1 is consistent with the presence of 2 as an impurity. However the potent (Ki
1.0 µM) action of 2 towards naringinase is in stark contrast to the complete lack of inhibition shown by
the corresponding 5-epi-lactam 3 and 5-epi-tetrazole 4.
5-epi-DMJ 6 inhibits the action of jack bean α-mannosidase but is weaker than DMJ itself and 6
is ineffective as an inhibitor of the bovine α-L-fucosidase compared with DMJ 5. Therefore, 5-epi-
LRJ 2 is a more potent and specific inhibitor than 5-epi-DMJ 6 as evidenced by its strong activity
against naringinase and its weak inhibition of the almond β-glucosidase. The unambiguous synthesis
of 6 explains the inhibition of α-mannosidase noted by Legler and Jülich13c which was thought to be due
to contamination by DMJ 5.
2.2. Molecular orbital calculations
Molecular modelling techniques were used to rationalise the surprising biological activity of 5-epi-
LRJ 2 in comparison to LRJ 1. It has been suggested that inhibitors of glycosidases are effective because
they have structures which resemble those of the respective glycopyranosyl cations that are the puta-
tive intermediates of enzyme-catalysed hydrolysis.21,22 A recent molecular graphics structure–activity
study23,24 of inhibitors of α-D-mannosidases (EC 3.2.1.24) has suggested that they do indeed mimic the
mannopyranosyl cation reaction intermediate. DFT molecular orbital methods were used to determine
the geometries of the two possible half-chair forms (4H3 and 3H4) of the rhamnopyranosyl cation.
2.3. Molecular modelling
The simulated annealing calculations with subsequent minimisation located the lowest energy con-
1
formations of LRJ 1 and its C-5 epimer 2. For LRJ 1, the lowest energy structure was the C4 chair
conformation in which the substituents are all equatorial except 2-OH. The lowest energy conformer of
the C-5 epimer 2 was the 4C1 (opposite) chair conformation in which the 2-OH and 5-Me substituents are
equatorial and the remainder axial. This conformation is also supported by a 3J1,2 proton NMR coupling
constant value of 10.3 Hz for 2 which is consistent with H-1 and H-2 in a diaxial arrangement. As
in previous studies,23,24 each ring conformation was used as a potential model to rationalise inhibitor
structure–activity relationships. The lowest energy conformation of the biologically active C-5 epimer
4
of LRJ 2 was found to be an excellent match with the H3 half-chair structure of the rhamnopyranosyl
cation. The 3H4 rhamnopyranosyl cation geometry was a very poor fit to the lowest energy C-5 epimer
geometry. The lowest energy conformation of the biologically inactive LRJ 1 was a very poor match with
the 4H3 rhamnopyranosyl structure. Interestingly, a 4H3 half-chair conformation for the glucosyl cation
was also proposed by Sinnott25 as being consistent with studies of the mechanism of β-glycosidases; it
was also found that the 4H3 conformation provided the most useful SAR model for mannosidases. Fig. 1
shows the superimposition of the lowest energy conformations of LRJ 1 and its C-5 epimer 2 on the 4H3
structure of the rhamnopyranosyl cation.
For the C-5 epimer of LRJ 2 the ring hydroxyl oxygen atoms and ring heteroatoms superimpose well,
in spite of the conformation of the piperidine ring not being identical to the half-chair conformation of
the glucosyl cation pyran ring. The ring carbon atoms also superimpose quite well.

2952
B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
Fig. 1. Superimposition of lowest energy conformers of LRJ 1 (left) and C-5-epi-LRJ 2 (right) on the 4H3 conformation of the
rhamnopyranosyl cation. Hydrogens have been omitted for clarity
The dominant factors which modulate rhamnosidase inhibition appear to be a ring conformation which
matches the 4H3 ring conformation of the rhamnopyranosyl cation, and hydroxyl substituents which are
topographical analogues of those in the rhamnopyranosyl cation. The 5-methyl group of LRJ 1 and the
epimer 2 appear to modulate inhibition by controlling the conformation of the piperidine ring via steric
interactions with the hydroxyl groups. In the case of the C-5 epimer 2, the piperidine ring adopts a
conformation which causes the hydroxyl groups to closely match the orientation in the rhamnopyranosyl
cation. In LRJ 1, the ring conformation is such that the hydroxyls adopt orientations clearly different
from those of the cation. Presumably this interferes with the recognition processes in the active site, as
previous work has shown that the stereochemistry and positions of the hydroxyl groups of substrates and
inhibitors of glycosidase inhibitors are crucial for activity. The apparent tolerance of the active site at
certain positions in inhibitors (e.g. the 5-methyl group in LRJ 1 and its epimer 2) has been observed in
other glycosidases. For example, the topographical analogue of the 4-OH in mannosidase inhibitors has
only a small influence on inhibitory activity,23 a finding in agreement with those of kinetic studies of
β-glucosidase inhibitors by Dale et al.26
The model based on the geometry of the rhamnopyranosyl cation provides a useful rationale to
explain the surprising biological activity of LRJ and its epimer, and extends the paradigm for glycosidase
4
inhibition, providing a third example where the H3 form of the glycopyranosyl cation is the favoured
form for templating of inhibitors. We are currently extending this model for other rhamnosidase inhibitors
and are assessing its value in designing other inhibitors of this enzyme.
Simulated annealing experiments were also carried out on DMJ 5 and its C-5 epimer 6. In both cases
4
the lowest energy conformations were C1, in which the C-3 and C-4 hydroxyl groups are equatorial.
Such conformations do not superimpose well (Fig. 2) on the mannopyranosyl cation model published
previously.23,24 For DMJ 5 this conformation is further stabilised by the formation of an intramolecular
hydrogen bond in low polarity environments, as has been shown for several glycosidase inhibitors by
NMR experiments.27 Superimposition of the heteroatoms of DMJ 5 onto the topographically equivalent
heteroatoms in the mannopyranosyl cation model gave an RMS error of 0.9 Å. This relatively poor
alignment is consistent with the low activity of compounds 5 and 6 against jack bean α-mannosidase.

B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
2953
Fig. 2. Superimposition of lowest energy conformer of deoxymannojirimycin 5 on the 4H3 conformation of the mannopyranosyl
cation model. Hydrogens have been omitted for clarity
3. Experimental
Melting points were recorded on a Kofler hot block and are uncorrected. Proton nuclear magnetic
resonance (δH) spectra were recorded on a Varian Gemini 200 (200 MHz), Bruker AC 200 (200 MHz)
or Bruker AM 500 (500 MHz) spectrometer. Carbon-13 nuclear magnetic resonance (δC) spectra were
recorded on a Varian Gemini 200 (50.3 MHz), Bruker AC 200 (50.3 MHz) or Bruker AM 500 (125 MHz)
spectrometer and multiplicities were assigned using DEPT sequence. All chemical shifts are quoted on
the δ-scale using residual solvent as an internal standard; for carbon-13 nuclear magnetic resonance
spectra run in D2O, 1,4-dioxan (δC 67.3 ppm) or methanol (δC 49.6 ppm) were used. The following
abbreviations were used to explain multiplicities: s, singlet; d, doublet; t, triplet; q, quartet; qu, quintet;
m, multiplet; br, broad; p, pseudo. Infra-red spectra were recorded on a Perkin–Elmer 1750 FT-IR or
Perkin–Elmer Paragon 1000 spectrophotometer. Mass spectra (m/z) were recorded on a VG Micromass
20-250, ZAB1F, VG Platform, or VG Autospec spectrometer using desorption chemical ionisation (NH3,
DCI), chemical ionisation (NH3, CI), electrospray (ES), or atmospheric pressure chemical ionisation
(APCI) as stated. Optical rotations were measured on a Perkin–Elmer 241 polarimeter with a path
length of 1 dm; concentrations are given in g/100 ml. Hydrogenations were executed at atmospheric
pressure under an atmosphere of hydrogen gas maintained by an inflated balloon. The removal of water,
aqueous acetic acid or aqueous trifluoroacetic acid as solvents was aided by co-evaporation with toluene.
Microanalyses were performed by the microanalysis service of the Dyson Perrins Laboratory. Thin
layer chromatography (TLC) was carried out on aluminium or plastic sheets coated with 60F254 silica.
Plates were developed using a spray of 0.2% w/v cerium(IV) sulphate and 5% ammonium molybdate
in 2 M sulphuric acid. Flash chromatography was carried out using Sorbsil C60 40/60 silica. Aqueous
orthophosphate solution buffering to pH ∼7 (pH 7 buffer) was prepared through the dissolution of 85 g
KH2PO4 and 14.5 g NaOH in 950 ml distilled water. Solvents and commercially available reagents were
dried and purified before use according to standard procedures; hexane was distilled at 68°C before use
to remove involatile fractions. All solvents were removed in vacuo.
3.1. 5-Azido-5,6-deoxy-2,3-O-isopropylidene-D-gulono-1,4-lactone 9
Triflic anhydride (0.54 ml, 1.7 equiv.) was added dropwise to a solution of pyridine (0.58 ml, 3.6
equiv.) and 2,3-O-isopropylidene-L-rhamnono-1,4-lactone 817 (400 mg, 1.98 mmol) in freshly distilled
dichloromethane (20 ml) under nitrogen at −40°C. After 2 h, TLC (ethyl acetate:hexane, 1:1) showed the
presence of a small amount of starting material (Rf 0.2) and a major product (Rf 0.7). Further pyridine
(0.12 ml, 0.75 equiv.) and triflic anhydride (0.1 ml, 0.3 equiv.) were added to the reaction solution.

2954
B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
After 4 h, TLC showed no starting material and a major product. The reaction solution was diluted with
dichloromethane (50 ml) and then washed with dilute hydrochloric acid (30 ml, 2 M), distilled water
(30 ml), buffer (30 ml) and then brine (30 ml), dried (magnesium sulphate), filtered and the solvent
removed. The residue was dissolved in N,N-dimethylformamide (10 ml) under nitrogen, and sodium
azide (390 mg, 3.0 equiv.) was added to the resulting solution. After 26 h, TLC (ethyl acetate:hexane,
1:1) showed the consumption of triflate and the formation of a major product (Rf 0.55). The solvent
was removed and the residue dissolved in ethyl acetate (60 ml). The resulting solution was washed with
brine (3×20 ml) and the aqueous layers were further extracted with ethyl acetate (60 ml). The organic
fractions were combined, dried (magnesium sulphate), filtered and the solvent removed. The residue was
purified by flash chromatography (ethyl acetate:hexane, 1:3) to give the azidolactone 9 (303 mg, 67%) as
a white solid, m.p. 78–81°C (ether/hexane). [α]D −118.8 (c, 1.33 in CHCl3). νmax (KBr) 3300 cm−1
21
(OH), 2125 cm−1 (N3), 1779 cm−1 (C_O). m/z (APCI+): 232 (M+H+−N2+MeOH, 8), 200 (M+H+−N2,
100%). δH (CDCl3, 500 MHz) 1.35 (d, 3H, H-6, J5,6 6.7 Hz), 1.41, 1.50 (s×2, 3H×2, C(CH3)2), 3.91
(dq, 1H, H-5, J5,6 6.7 Hz, J4,5 9.0 Hz), 4.26 (dd, 1H, H-4, J3,4 3.4 Hz, J4,5 9.2 Hz), 4.75 (dd, 1H, H-3,
J2,3 5.1 Hz, J3,4 3.6 Hz), 4.85 (d, 1H, H-2, J2,3 5.2 Hz). δC (CDCl3, 50.3 MHz) 15.3 (q, C-6), 25.9, 26.7
(q×2, C(CH3)2), 57.6, 75.7, 76.2, 82.0 (d×4, C-2, C-3, C-4, C-5), 114.7 (s, C(CH3)2), 173.3 (s, C-1).
Anal. found: C, 47.79; H, 5.69; N, 18.47%; C9H13N3O4 requires C, 47.57; H, 5.77; N, 18.49%.
3.2. 6-Deoxy-2,3-O-isopropylidene-D-gulono-1,5-lactam 10
Palladium on carbon (10%, 30 mg) was added to a solution of azidolactone 9 (190 mg, 0.837 mmol)
in anhydrous methanol (6 ml). The solution was thoroughly degassed and stirred under hydrogen. After
7 h, TLC showed the consumption of starting material [Rf 0.55 in ethyl acetate:hexane (1:1); Rf 0.9
in methanol:ethyl acetate (1:9)] and the formation of a major product [Rf 0.2 in methanol:ethyl acetate
(1:9)]. The reaction solution was filtered through a pad of Celite and the solvent removed. The residue
was purified by flash chromatography (ethanol:ethyl acetate, 1:9) to give the lactam 10 (125 mg, 74%) as
21
a white crystalline solid, m.p. 151–152°C (ethyl acetate). [α]D +39.0 (c, 0.52 in CHCl3). νmax (KBr)
3300 cm−1 (OH), 1667 cm−1 (C_O). m/z (APCI+): 202 (M+H+, 100), 144 (22%). δH (CDCl3, 500 MHz)
1.35 (d, 3H, H-6, J5,6 6.8 Hz), 1.39, 1.48 (s×2, 3H×2, C(CH3)2), 2.63 (br s, 1H, OH), 3.81 (qt, 1H, H-5,
J 1.6 Hz, J5,6 6.9 Hz), 3.85 (br s, 1H, H-4), 4.48 (dd, 1H, H-3, J2,3 6.7 Hz, J3,4 3.3 Hz), 4.53 (dd, 1H,
H-2, J 0.7 Hz, J2,3 6.7 Hz), 5.86 (br s, 1H, NH). δC (CDCl3, 50.3 MHz) 16.2 (q, C-6), 26.4, 26.4 (q×2,
C(CH3)2), 46.9, 69.7, 72.8, 76.4 (d×4, C-2, C-3, C-4, C-5), 110.1 (s, C(CH3)2), 170.0 (s, C-1). Anal.
found: C, 53.62; H, 7.67; N, 6.69%; C9H15NO4 requires C, 53.72; H, 7.51; N, 6.96%.
3.3. 6-Deoxy-D-gulono-1,5-lactam 3
The protected lactam 10 (55 mg, 0.274 mmol) was dissolved in aqueous trifluoroacetic acid (60% v/v,
2 ml). After 5 h, TLC (methanol:ethyl acetate, 1:9) showed the consumption of starting material (Rf 0.6)
and the formation of a major product (Rf 0.1). The solvent was removed and the residue crystallised from
ethanol/ethyl acetate to give 6-deoxy-D-gulono-1,5-lactam 3 (39 mg, 89%) as a white crystalline solid,
25
m.p. 192–193°C (ethanol/ethyl acetate). [α]D +66.9 (c, 0.86 in H2O). νmax (KBr) 3400 cm−1 (OH),
1670 cm−1 (C_O). m/z (APCI+): 162 (M+H+, 100), 126 (87%). δH (D2O, 500 MHz) 1.12 (d, 3H, H-6,
J5,6 6.9 Hz), 3.80 (qd, 1H, H-5, J4,5 3.0 Hz, J5,6 6.9 Hz), 3.90 (dd, 1H, H-4, J3,4 4.8 Hz, J4,5 3.0 Hz),
4.17 (dd, 1H, H-3, J2,3 3.8 Hz, J3,4 4.6 Hz), 4.25 (d, 1H, H-2, J2,3 3.6 Hz). δC (D2O, 50.3 MHz) 15.0 (q,
C-6), 47.9, 66.0, 69.4, 70.1 (d×4, C-2, C-3, C-4, C-5), 173.9 (s, C-1). Anal. found: C, 44.73; H, 6.80; N,
8.49%; C6H11NO4 requires C, 44.72; H, 6.88; N, 8.69%.

B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
2955
3.4. 5-epi-LRJ (1,5,6-trideoxy-1,5-imino-D-gulitol) 2
A solution of borane:dimethylsulphide complex in THF (0.4 ml, 2 M) was added dropwise to a solution
of the protected lactam 10 (106 mg, 0.527 mmol) in freshly distilled THF (2 ml) under nitrogen. After
25 h, anhydrous methanol was added dropwise until no further effervescence was observed. The reaction
solvent was removed and the residue repeatedly dissolved in methanol (3×5 ml) and evaporated under
reduced pressure. The residue was then dissolved in ethanol (1 ml), a small amount of concentrated
hydrochloric acid added (3 drops) and the resulting solution cooled to −15°C. After 147 h, the solvent
was removed, the residue purified by ion-exchange chromatography [Amberlite IR-120(plus), eluent
aqueous ammonia solution 1 M] and freeze-dried to give 5-epi-LRJ 2 (55 mg, 71%) as an amorphous
solid. [α]D24 +6.3 (c, 0.96 in H2O) {lit.,10b [α]D −3 (c, 1 in H2O)}. νmax (KBr) 3400 cm−1 (OH, NH).
20
m/z (APCI+): 148 (M+H+, 100), 130 (M+H+−H2O, 5%). δH (D2O, 500 MHz) 0.94 (d, 3H, H-6, J5,6
6.9 Hz), 2.59 (dd, 1H, H-1, J1,1 12.8 Hz, J1,2 10.3 Hz), 2.71 (dd, 1H, H-10, J1,1 12.9 Hz, J1 ,2 4.7 Hz),
0
0
0
2.91 (m, 1H, H-5), 3.61 (dd, 1H, H-4, J 2.0 Hz, J 4.4 Hz), 3.77 (ddd, 1H, H-2, J2,3 3.4 Hz, J1,2 10.0 Hz,
J
1 ,2 4.4 Hz), 3.81 (m, 1H, H-3), consistent with literature.10b δC (D2O, 50.3 MHz) 15.7 (q, C-6), 44.8 (t,
0
C-1), 49.3 (d, C-5), 66.6, 71.3, 72.9 (d×3, C-2, C-3, C-4), consistent with literature.10b HRMS m/z (CI,
NH3): found 148.096890 (M+H+); C6H14NO3 requires 148.097368.
3.5. 5-Azido-5,6-dideoxy-2,3-O-isopropylidene-D-gulononamide 11
Ammonia was bubbled through a solution of 5-azido-5,6-dideoxy-2,3-O-isopropylidene-D-gulono-
1,4-lactone 9 (160 mg, 0.705 mmol) in a freshly prepared saturated solution of ammonia in anhydrous
methanol (7 ml). After 1 h, TLC (ethyl acetate) showed the complete conversion of starting material (Rf
0.8) to a single product (Rf 0.4). The solvent was removed to give the open chain amide 11 (170 mg, 99%)
as a hygroscopic white solid, m.p. 78–81°C (ethyl acetate). [α]D23 +13.1 (c, 1.13 in acetone). νmax (KBr)
3400 cm−1 (OH, NH), 2111 cm−1 (N3), 1684 cm−1 (amide I), 1586 cm−1 (amide II). m/z (APCI+): 245
(M+H+, 46), 217 (M+H+−N2, 100%). δH (CD3CN, 500 MHz) 1.23 (d, 3H, H-6, J5,6 6.7 Hz), 1.37, 1.56
(s×2, 3H×2, C(CH3)2), 3.60 (pqu, 1H, H-5, J 6.4 Hz), 3.68 (dd, 1H, H-4, J3,4 3.1 Hz, J4,5 5.9 Hz), 4.46
(dd, 1H, H-3, J2,3 7.9 Hz, J3,4 3.1 Hz), 4.50 (d, 1H, H-2, J2,3 7.9 Hz), 6.06, 6.63 (br s×2, 1H×2, NH2).
δC (CD3CN, 125 MHz) 16.3 (q, C-6), 24.7, 26.6 (q×2, C(CH3)2), 60.9, 73.3, 76.6, 77.9 (d×4, C-2, C-3,
C-4, C-5), 110.5 (s, C(CH3)2), 173.3 (s, C-1). Anal. found: C, 44.25; H, 6.56; N, 22.51%; C9H16O4N4
requires C, 44.26; H, 6.60; N, 22.94%.
3.6. 5-Azido-5,6-dideoxy-2,3-O-isopropylidene-D-gulononitrile 12
Trifluoroacetic anhydride (0.43 ml, 5 equiv.) was added dropwise to a solution of 5-azido-5,6-dideoxy-
2,3-O-isopropylidene-D-gulononamide 11 (150 mg, 0.615 mmol) in dry pyridine (5 ml) under nitrogen
at −35°C. After 2 h, TLC (ethyl acetate) showed the consumption of starting material (Rf 0.4) and the
formation of a major product (Rf 0.9). Methanol (1 ml) was added, the reaction solution warmed to room
temperature and the solvent removed. The residue was dissolved in ethyl acetate (15 ml), washed with
brine (10 ml), dried (magnesium sulphate), filtered and the solvent removed. The residue was purified
by flash chromatography (ethyl acetate:hexane, 2:3) to give the azidonitrile 12 (117 mg, 84%) as an
24
amorphous solid. [α]D −82.1 (c, 0.78 in CH3CN). νmax (KBr) 3450 cm−1 (OH), 2111 cm−1 (N3). m/z
(APCI+): 227 (M+H+, 100), 201 (8%). δH (CD3CN, 500 MHz) 1.36 (d, 3H, H-6, J5,6 6.7 Hz), 1.38, 1.54
(s×2, 3H×2, C(CH3)2), 3.50 (qd, 1H, H-5, J4,5 3.4 Hz, J5,6 6.4 Hz), 3.58 (d, 1H, OH, J4,OH 6.5 Hz), 3.79
(ptd, 1H, H-4, J4,5 3.4 Hz, J 6.6 Hz), 4.32 (dd, 1H, H-3, J2,3 5.5 Hz, J3,4 6.7 Hz), 4.95 (d, 1H, H-2, J2,3

2956
B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
5.5 Hz). δC (CD3CN, 125 MHz) 15.4 (q, C-6), 25.7, 27.1 (q×2, C(CH3)2), 59.0, 66.4, 73.8, 78.8 (d×4,
C-2, C-3, C-4, C-5), 112.7 (s, C(CH3)2), 117.9 (s, C-1). Anal. found: C, 48.05; H, 6.03; N, 28.01%;
C9H14N4O3 requires C, 47.78; H, 6.24; N, 24.76%.
3.7. (5R,6S,7R,8S)-5,6,7,8-Tetrahydro-7,8-O-isopropylidene-5-methyl-tetrazolo[1,5-a]pyridine-6,7,8-
triol 13
The amide 12 (110 mg, 0.486 mmol) was dissolved in deuteriated toluene (1.65 ml) and the resulting
solution heated to 102°C under nitrogen. After 33 h, proton NMR and TLC (ethyl acetate:hexane,
1:1) showed no starting material (Rf 0.45) and the formation of a major product (Rf 0.2) and a minor
product (Rf 0.7). The solvent was removed and the residue purified by flash chromatography (ethyl
acetate:hexane, 1:1) to give the lactone 9 (18 mg, 16%) as a white solid and the protected tetrazole 13 (78
24
mg, 71%) as a white foam. [α]D −23.4 (c, 0.84 in CHCl3). νmax (KBr) 3400 cm−1 (OH). m/z (APCI+):
227 (M+H+, 100%). δH (CDCl3, 500 MHz) 1.24, 1.47 (3H×2, s×2, C(CH3)2), 1.87 (d, 1H, CH3, J5,Me
6.8 Hz), 2.65 (br s, 1H, OH), 4.42 (m, 1H, H-6), 4.67 (qd, 1H, H-5, J5,Me 6.8 Hz, J5,6 2.1 Hz), 4.72 (dd,
1H, H-7, J6,7 3.8 Hz, J7,8 6.0 Hz), 5.59 (d, 1H, H-8, J7,8 6.0 Hz). δC (CDCl3, 125 MHz) 13.9 (q, CH3),
25.1, 26.9 (q×2, C(CH3)2), 52.7, 66.6, 70.4, 75.0 (d×4, C-5, C-6, C-7, C-8), 111.6 (s, C(CH3)2), 150.2
(s, C-8a). Anal. found: C, 47.76; H, 6.16; N, 24.95%; C9H14N4O3 requires C, 47.78; H, 6.24; N, 24.76%.
3.8. (5R,6S,7R,8S)-5,6,7,8-Tetrahydro-5-methyl-tetrazolo[1,5-a]pyridine-6,7,8-triol 4
The protected tetrazole 13 (68 mg, 0.301 mmol) was dissolved in aqueous trifluoroacetic acid (50%
v/v, 3 ml). After 5 h, TLC (ethyl acetate) showed no starting material (Rf 0.6) and the formation of
a major product (Rf 0.1). The solvent was removed and the residue purified by flash chromatography
23
(methanol:ethyl acetate, 1:19) to give the tetrazole 4 (44 mg, 79%) as a colourless oil. [α]D +24.3 (c,
0.6 in acetone). νmax (film) 3400 cm−1 (OH). m/z (APCI−): 221 (M+Cl−, 13), 185 (M−H+, 100), 149
(18), 125 (34), 113 (57%). δH (CD3OD, 500 MHz) 1.71 (d, 3H, CH3, J5,Me 6.8 Hz), 4.19 (dd, 1H, H-6,
J5,6 2.8 Hz, J6,7 5.1 Hz), 4.28 (dd, 1H, H-7, J6,7 5.0 Hz, J7,8 3.7 Hz), 4.68 (qd, 1H, H-5, J5,Me 6.8 Hz,
J5,6 2.7 Hz), 5.11 (d, 1H, H-8, J7,8 3.6 Hz). δC (CD3OD, 125 MHz) 14.7 (q, CH3), 54.7, 63.9, 71.2,
72.0 (d×4, C-5, C-6, C-7, C-8), 155.2 (s, C-8a). HRMS m/z (CI, NH3): found 187.083165 (M+H+);
C6H11N4O3 requires 187.083115.
3.9. 6-O-tert-Butyldimethylsilyl-2,3-O-Isopropylidene-L-gulono-1,5-lactam 16
The lactone 1520 (307 mg, 0.86 mmol) was dissolved in ethanol (8 ml) and palladium on charcoal
(25 mg) was added. The solution was degassed and stirred under hydrogen. After 23 h, TLC (ethyl
acetate:hexane, 1:1) showed one major product (Rf 0.2) and no starting material (Rf 0.8). The reaction
mixture was passed through a Celite plug and the solvents removed. Purification by flash chromatography
22
gave the gulono-1,5-lactam 16 (227 mg, 80%) as a colourless oil. [α]D −10.6 (c, 1.16 in CHCl3). νmax
(CHCl3) 3300 cm−1 (OH), 1673 cm−1 (C_O). m/z (APCI+): 332 (M+H+, 100%). δH (CDCl3, 500 MHz)
0.12, 0.12 (s×2, 3H×2, Si(CH3)2), 0.90 (s, 9H, SiC(CH3)3), 1.39, 1.47 (s×2, 3H×2, C(CH3)2), 3.66–3.68
(m, 2H, OH, H-4/H-5), 3.91 (dd, 1H, H-6, J5,6 3.4 Hz, J6,6 10.9 Hz), 3.98 (dd, 1H, H-60, J5,6 4.5 Hz,
0
0
0
J6,6 10.9 Hz), 4.13–4.15 (m, 1H, H-4/H-5), 4.43 (dd, 1H, H-3, J2,3 6.7 Hz, J3,4 3.8 Hz), 4.54 (d, 1H,
H-2, J2,3 6.7 Hz), 6.42 (br s, 1H, NH). δC (CDCl3, 50.3 MHz) −5.64, −5.56 (q×2, Si(CH3)2), 18.1 (s,
SiC(CH3)2), 24.5, 26.7 (q×2, C(CH3)2), 25.8 (q, SiC(CH3)2), 52.2, 68.5, 73.4, 76.1 (d×4, C-2, C-3,

B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
2957
C-4, C-5), 63.5 (t, C-6), 110.2 (s, C(CH3)2), 170.1 (s, C-1). Anal. found C, 54.31; H, 9.10; N, 4.21%;
C15H29NO5Si requires C, 54.35; H, 8.82; N, 4.21%.
3.10. L-Gulono-1,5-lactam 7
The protected lactam 16 (73 mg, 0.22 mmol) was dissolved in trifluoroacetic acid (4 ml) and water (2
ml) and stirred for 4 h, after which time TLC (ethyl acetate) showed one product (Rf 0.1) and no starting
material (Rf 0.4). The solvents were removed and the residue coevaporated with toluene (2×5 ml). The
residue was dissolved in the minimum amount of hot ethanol and a few drops of ethyl acetate were added.
After 3 days, the resulting crystals were washed with ethyl acetate giving L-gulono-1,5-lactam 7 (35 mg,
90%), m.p. 171–173.5°C {lit.,19 could not be crystallised}. [α]D −20.6 (c, 0.65 in H2O) {lit.,19 [α]D
22
−27.0 (c, 0.5 in H2O)}. δH (D2O, 500 MHz) 3.61–3.67 (m, 1H, H-5), 3.72–3.78 (m, 2H, H-6, H-60),
4.16 (dd, 1H, H-4, J4,5 2.8 Hz, J3,4 4.8 Hz), 4.19 (dd, 1H, H-3, J2,3 3.5 Hz, J3,4 4.8 Hz), 4.35 (d, 1H, H-2,
J2,3 3.5 Hz), consistent with literature.19
3.11. 5-epi-DMJ (1,5-dideoxy-1,5-imino-L-gulitol) 6
The protected lactam 16 (127 mg, 0.38 mmol) was dissolved in dry tetrahydrofuran (2.5 ml) and the
solution stirred under nitrogen. Borane:dimethyl sulphide (2 M in THF, 0.75 ml) was added. After 16
h, methanol was added to the reaction mixture until effervescence ceased. The solvents were removed
and further methanol (4×5 ml) distilled off the residue, which was then dissolved in ethanol (3 ml). A
few drops of concentrated hydrochloric acid were added and the mixture stirred for 3 h. The solvent was
removed and the residue dissolved in water (1 ml) and purified by ion-exchange chromatography to give
5-epi-DMJ 6 (57 mg, 92%) as a colourless oil. [α]D21 +8.8 (c, 0.25 in H2O) {lit.,15c,d [α]D +9 (c, 0.58 in
H2O), lit.,15b [α]578 −21 (c, H2O)}. δH (D2O, 500 MHz) 2.86–2.91 (m, 1H, H-1), 3.06 (dd, 1H, H-10,
20
0
0
0
J1,1 12.3 Hz, J1 ,2 4.9 Hz), 3.20–3.23 (m, 1H, H-5), 3.67 (dd, 1H, H-6, J5,6 8.0 Hz, J6,6 11.8 Hz), 3.73
(dd, 1H, H-60, J5,6 5.6 Hz, J6,6 11.8 Hz), 3.95–3.98 (m, 2H, H-3, H-4), 4.06 (ddd, 1H, H-2, J1,2 11.2
0
0
Hz, J1 ,2 4.9 Hz, J2,3 2.9 Hz), consistent with literature.15c,d
0
3.12. Enzyme assays14
α-Glucosidase (brewer’s yeast), α-fucosidase (bovine epididimus), β-glucosidase (almond emulsin),
α-galactosidase (green coffee bean), β-galactosidase (E. coli), α-mannosidase (jack bean), β-N-
acetylglucosaminidase (bovine and jack bean) and naringinase (Penicillium decumbens) (20 µl of 10
µg/ml commercially available solution) were purchased from Sigma and assayed using the appropriate
p-nitrophenylglycopyranoside (5 mM in 100 µl) as a substrate, at the optimum pH of the enzyme at 30°C
in the absence and presence of each of the compounds to be tested and quenched after a period of 10 min
by the addition of glycine solution (pH 10.4).
3.13. Molecular orbital calculations
Molecular orbital calculations were performed on a Cray Y-MP4 computer using the DGausss DFT
molecular orbital package in the UniChem suite of programs.28 Full geometry optimisation was carried
out on each half-chair form of the rhamnopyranosyl cation. The calculations used the DZVP basis
functions with BLYP non-local functionals. The starting geometries for the two possible half-chair
forms of the rhamnopyranosyl cation were derived from the corresponding mannopyranosyl cation

2958
B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
geometries obtained from a molecular orbital study.23 The calculations used hydroxyl torsion angles
which maximised the number of intramolecular hydrogen bonds, as found by Csonka et al. using DFT
calculations.29
3.14. Molecular modelling
The molecular modelling was carried out on a Silicon Graphics Indigo 2xl workstation (R4400)
using Tripos Sybyl30 software (version 6.3). Crystal structures were available for L-rhamnose.31 The
geometries of the other inhibitors were obtained by modification of these crystal structures, and by the use
of the molecular modification and building capabilities of Sybyl. The derived geometries were optimised
by means of a molecular mechanics calculation using the Tripos force field.32 Simulated annealing
calculations were used to sample the conformational space accessible to the piperidine rings of LRJ and
its epimer, and to derive the lowest energy conformations. The calculations were run for 10 annealing
cycles with an initial equilibration period of 1000 fs at a temperature of 1000 K. The time increment in
the dynamics calculations was 0.5 fs and the coupling time for temperature regulations was 2.0 fs. The
annealing phase used an exponential annealing function cooling from 1000 K to 200 K over 2000 fs. Final
geometries were optimised using the Tripos force field prior to superimposition. The superimpositions
were accomplished by a least squares fit after choosing pairs of topographically equivalent atoms on the
two molecules being studied. The choice of topographically equivalent atoms for the superimpositions
was usually clear, with hydroxyl groups of inhibitors and cation being superimposed, as were the
pyranose ring oxygen atom and the heterocyclic nitrogen atoms of the inhibitors.
Acknowledgements
This work was supported by a BBSRC CASE studentship (BGD), an EPSRC post doctoral fellowship
(AAW), and assistance of the Department of Industry, Science and Technology in providing a Bilateral
Science and Technology Fellowship to Oxford (DAW).
References
1. (a) Michael, J. P., Nat. Prod. Rep., 1994, 11, 639; (b) Michael, J. P., Nat. Prod. Rep., 1994, 14, 619; (c) O’Hagan, D., Nat.
Prod. Rep., 1997, 14, 637; (d) Asano, N., Nishida, M., Kizu, H., Matsui, K., Watson, A. A., Nash, R. J., J. Nat. Prod.,
1997, 60, 98; (e) Kato, A., Asano, N., Kizu, H., Matsui, K., Watson, A. A., Nash, R. J., J. Nat. Prod., 1997, 60, 312; (f)
Asano, N., Kato, A., Miyauchi, M., Kizu, H., Tomimori, T., Matsui, K., Nash, R. J., Molyneux, R. J., Eur. J. Biochem.,
1997, 248, 296; (g) Kato, A., Asano, N., Kizu, H., Matsui, K., Suzuki, S., Arisawa, M., Phytochem., 1997, 45, 425; (h)
Watson, A. A., Nash, R. J., Wormald, M. R., Harvey, D. J., Dealler, S., Lees, E., Asano, N., Kizu, H., Kato, A., Griffiths,
R. C., Cairns, A. J., Fleet G. W. J., Phytochem., 1997, 46, 255; (i) Asano, N., Kato, A., Matsui, K., Watson, A. A., Nash, R.
J., Molyneux, R. J., Hackett, L., Topping, J., Winchester, B., Glycobiol., 1997, 7, 1085; (j) Nojima, H., Kimura, I., Chen,
F. J., Sugihara, Y., Haruno, M., Kato, A., Asano, N., J. Nat. Prod., 1998, 61, 397; (k) Asano, N., Kato A., Miyauchi, M.,
Kizu, H., Kameda, Y., Watson, A. A., Nash, R. J., Fleet, G. W. J., J. Nat. Prod., 1998, 61, 625.
2. (a) Jespersen, T. M., Dong, W. L., Sierks, M. R., Skrydstrup, T., Lundt, I., Bols, M., Angew. Chem. Int. Ed. Engl., 1994,
33, 1778; (b) Bols, M., Acc. Chem. Res., 1998, 31, 1; (c) Ichikawa, Y., Igarashi, Y., Ichikawa, M., Suhara, Y., J. Am. Chem.
Soc., 1998, 120, 3007.
3. (a) Fellows, L., Chem. Br., 1987 (Sept), 842; (b) Legler, G., Adv. Carbohydr. Chem. Biochem., 1990, 48, 319; (c) Sinnott,
M. L., Chem. Rev., 1990, 90, 1171; (d) Winchester, B., Fleet, G. W. J., Glycobiology, 1992, 2, 199; (e) Nishimura, Y., in
Studies in Natural Product Chemistry, ed. Atta-ur-Rahman, Elsevier: Amsterdam, 1992, 10, 495; (f) Look, G. C., Fotsch,
C. H., Wong, C.-H., Acc. Chem. Res., 1993, 26, 182; (g) Asano, N., Oseki, K., Kizu, H., Matsui, K., J. Med. Chem., 1994,

B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
2959
37, 3701; (h) Platt, F. M., Butters, T. D., Trends Glycosci. Glycotechnol., 1995, 7, 495; (i) Hudlicky, T., Entwistle, D. A.,
Pitzer, K. K., Thorpe, A. J., Chem. Rev., 1996, 96, 1195; (j) Ganem, B., Acc. Chem. Res., 1996, 29, 340; (k) Lundt, I.,
Topics Curr. Chem., 1997, 187, 117; (l) de Raadt, A., Ekhart, C. W., Ebner, M., Stutz, A. E., Topics Curr. Chem., 1997,
187, 157; (m) McGarvey, G. J., Wong, C.-H. Lieb Ann. Rec., 1997, 1059; (n) Sears, P., Wong, C.-H., J. Chem. Soc., Chem.
Commun., 1998, 1161.
4. Chrystal, E. J. T., Fowler, P. A., Gravestock, M. B., Haines, A. H., Taylor, R. J. K., Carbohydr. Res., 1993, 246, 377.
5. Fairbanks, A. J., Carpenter, N. C., Fleet, G. W. J., Ramsden, N. G., Cenci di Bello, I., Winchester, B. G., Al-Daher, S. S.,
Nagahashi, G., Tetrahedron, 1992, 48, 3365.
6. Zhou, P., Salleh, H. M., Chan, P. C., Lajoie, G., Honek, J. F., Nambiar, P., Ward, O. P., Carbohydr. Res., 1993, 239, 155.
7. (a) Kajimoto, T., Chen, L., Liu, K. K.-C., Wong, C.-H., J. Am. Chem. Soc., 1991, 113, 6678; (b) Liu, K. K.-C., Kajimoto,
T., Chen, L., Zhong, Z., Ichikawa, Y., Wong, C.-H., J. Org. Chem., 1991, 56, 6280.
8. Provencher, L., Steensma, D. H., Wong, C.-H., Bioorg Med. Chem., 1994, 2, 1179.
9. Although schemes in Ref. 7a indicate the formation of a 1:6 mixture of 5-epi-LRJ 2:LRJ 1, the text and later Refs. 7b, 8
indicate that only the formation of LRJ 1 was detectable.
10. (a) Shilvock, J. P., Wheatley, J. R., Davis, B., Nash, R. J., Griffiths, R. C., Jones, M. G., Müller, M., Crook, S., Watkin, D.
J., Smith, C., Besra, G. S., Brennan, P. J., Fleet, G. W. J., Tetrahedron Lett., 1996, 37, 8569; (b) Defoin, A., Sarazin, H.,
Streith, J., Helv. Chim. Acta, 1996, 79, 560.
11. (a) Davis, B., Brandstetter, T. W., Smith, C., Hackett, L., Winchester, B. G., Fleet, G. W. J., Tetrahedron Lett., 1995, 36,
7507; (b) Estevez, J. C., Saunders, J., Besra, G. S., Brennan, P. J., Nash, R. J., Fleet, G. W. J., Tetrahedron: Asymmetry,
1996, 7, 383; (c) Estevez, J. C., Smith, M. D., Wormald, M. R., Besra, G. S., Brennan, P. J., Nash, R. J., Fleet, G. W. J.,
Tetrahedron: Asymmetry, 1996, 7, 391; (d) Davis, B., Bell, A. A., Nash, R. J., Watson, A. A., Griffiths, R. C., Jones, M. G.,
Smith, C., Fleet, G. W. J., Tetrahedron Lett., 1996, 37, 8565; (e) Fleet, G. W. J., Estevez, J. C., Smith, M. D., Bleriot, Y.,
De la Fuente, C., Krülle, T. M., Besra, G. S., Brennan, P. J., Nash, R. J., Johnson, L. N., Oikonomakos, N. G., Stalmans,
W., Pure Appl. Chem., 1998, 70, 279.
12. Fellows, L. E., Bell, E. A., Lynn, D. G., Pilkiewicz, F., Miura, I., Nakanishi, K., J. Chem. Soc., Chem. Commun., 1979,
977.
13. (a) Fuhrmann, U., Bause, E., Legler, G., Ploegh, H., Nature, 1984, 307, 755; (b) Elbein, A. D., Legler, G., Tlutsy, A.,
McDowell, W., Schwarz, R., Arch. Biochem. Biophys., 1984, 235, 579; (c) Leglich, G., Jülich, E., Carbohydr. Res., 1984,
128, 61; (d) Myerscough, P. M., Fairbanks, A. J., Jones, A. H., Bruce, I., Choi, S. S., Fleet, G. W. J., Al-Daher, S. S., Cenci
di Bello, I., Winchester, B., Tetrahedron, 1992, 48, 10177.
14. Evans, S. V., Fellows, L. E., Shing, T. K. M., Fleet, G. W. J., Phytochem., 1985, 24, 1953.
15. (a) Leontein, K., Lindberg, B., Lönngren, J., Acta Chem. Scand. B, 1982, 36, 515; (b) Legler, G., Jülich, E., Carbohydr.
Res., 1984, 128, 61; (b) Poitout, L., Le Merrer, Y., Depezay, J.-C., Tetrahedron Lett., 1994, 35, 3293; (d) Poitout, L., Le
Merrer, Y., Depezay, J.-C., Dosbaa, I., Geoffroy, S., Foglietti, M.-J., Bioorg, Med. Chem., 1997, 5, 519.
16. (a) Fleet, G. W. J., Ramsden, N. G., Dwek, R. A., Rademacher, T. W., Fellows, L. E., Nash, R. J., Green, D. S., Winchester,
B., J. Chem. Soc., Chem. Commun., 1988, 483; (b) Fleet, G. W. J., Ramsden, N. G., Witty, D. R., Tetrahedron Lett., 1988,
29, 2871; (c) Fleet, G. W. J., Ramsden, N. G., Witty, D. R., Tetrahedron, 1989, 45, 319; (d) Fleet, G. W. J., Carpenter,
N. M., Petursson, S., Ramsden, N. G., Tetrahedron Lett., 1990, 31, 409; (e) Ermert, P., Vasella, A., Helv. Chim. Acta,
1991, 74, 2043; (f) Ermert, P., Vasella, A., Weber, M., Rupitz, K., Withers, S. G., Carbohydr. Res., 1993, 250, 113; (g)
Horsch, M., Hoesch, L., Fleet, G. W. J., Rast, D. M., J. Enzym. Inhibit., 1993, 7, 47; (h) Heightman, T. D., Ermert, P.,
Klein, D., Vasella, A., Helv. Chim. Acta, 1995, 78, 514; (i) Brandstetter, T. W., Davis, B., Hyett, D., Smith, C., Hackett,
L., Winchester, B. G., Fleet, G. W. J., Tetrahedron Lett., 1995, 36, 7511; (j) Hoos, R., Huixin, J., Vasella, A., Weiss, P.,
Helv. Chim. Acta, 1996, 79, 1757; (k) Heightman, T. D., Vasella, A., Tsitsanou, K. E., Zographos, S. E., Skamnaki, V. T.,
Oikonomakos, N. G., Helv. Chim. Acta, 1998, 81, 853.
17. (a) Dominguez, J. N., Owen, L. N., Carbohydr. Res., 1979, 75, 101; (b) Mantell, S. J., Fleet, G. W. J., Brown, D., J. Chem.
Soc., Chem. Commun., 1991, 1563.
18. (a) Marin, R. B., Hedrick, R., Parcell, A., J. Org. Chem., 1964, 29, 158; (b) Austin, P. W., Buchanan, J. G., Saunders, R.
M., J. Chem. Soc. C, 1967, 372; (c) Buchanan, J. G., MacLean, K. A., Wightman, R. H., Paulsen, H., J. Chem. Soc., Perkin
Trans. 1, 1985, 1463; (d) Bird, P., Dolphin, D. H., Withers, S. G., Can. J. Chem., 1990, 68, 317.
19. Overkleeft, H. S., van Wiltenburg, J., Pandit, U. K., Tetrahedron, 1994, 50, 4215.
20. Fleet, G. W. J., Namgoong, S. K., Barker, C., Baines, S., Jacob, G. S., Winchester, B., Tetrahedron Lett., 1989, 30, 4439.
21. Schneider, M. J., Ungemach, F. S., Broquist, H. P., Harris, T. M., Tetrahedron, 1983, 39, 29.
22. Dorling, P. R., Huxtable, C. R., Colegate, S. M., Biochem. J., 1980, 191, 649.
23. Winkler, D. A., Holan, G. J. Med. Chem., 1989, 32, 2084.

2960
B. G. Davis et al. / Tetrahedron: Asymmetry 9 (1998) 2947–2960
24. Winkler, D. A., J. Med. Chem., 1996, 39, 4332.
25. Sinnott, M. L., Biochem. J., 1984, 224, 817.
26. Dale, M. P., Ensley, H. E., Kern, K., Sastry, K. A. R., Byers, L. D., Biochem., 1985, 24, 3530.
27. Van den Broek, L. A. G. M., Vermaas, D. J., Heskamp, B. M., van Boeckel, C. A. A., Tan, M. C. A. A., Bolscher, J. G. M.,
Ploegh, H. L., Kemenade, F. J., de Goede, R. E. Y., Miedema, F., Recl. Trav. Chim. Pays Bas, 1993, 112, 82.
28. Oxford Molecular Limited, Oxford, UK (http://www.oxmol.co.uk).
29. Csonka, G. I., Elias, K., Csizmadia, I. G., J. Comput. Chem. 1996, 18, 330.
30. Tripos Associates, St. Louis (http://www.tripos.com).
31. Killean, R. C. G., Lawrence, J. L., Sharma, V. C., Acta Crystallogr. B, 1971, 27, 1707.
32. Clark, M., Cramer, R. D., Van Opdenbosch, N., J. Comput. Chem., 1989, 10, 982.