Efficient Synthesis of an Indinavir Precursor from Biomass-Derived (-)-Levoglucosenone

Article abstract of DOI:10.1071/CH17227

Lignocellulosic biomass pyrolysis with acid catalysis selectively produces the useful chiral synthon 6,8-dioxabicyclo[3.2.1]oct-2-ene-4-one ((-)-levoglucosenone, LGO). In this report, LGO was used to prepare (3R,5S)-3-benzyl-5-(hydroxymethyl)-4,5-dihydrofuran-2(3H)-one, which is an intermediate used in the construction of antivirals including the protease inhibitor indinavir. To achieve the synthesis, the hydrogenated derivative of LGO was functionalised using aldol chemistry and various aromatic aldehydes were used to show the scope of the reaction. Choice of base affected reaction times and the best yields were obtained using 1,1,3,3-tetramethylguanidine. Hydrogenation of the α-benzylidene-substituted bicyclic system afforded a 4:3 equatorial/axial mixture of isomers, which was equilibrated to a 97:3 mixture under basic conditions. Subsequent Baeyer-Villiger reaction afforded the target lactone in 57% overall yield for four steps, a route that avoids the protection and strong base required in the traditional approach. The aldol route is contrasted with the α-alkylation and a Baylis-Hillman approach that also both start with LGO.

Full text of DOI:10.1071/CH17227

CSIRO PUBLISHING
Aust. J. Chem. 2017, 70, 1146–1150
Full Paper
https://doi.org/10.1071/CH17227
Efficient Synthesis of an Indinavir Precursor from
Biomass-Derived (2)-Levoglucosenone
A
A
A B
,
Edward T. Ledingham, Kieran P. Stockton, and Ben W. Greatrex
A
School of Science and Technology, University of New England, Armidale, NSW 2351,
Australia.
B
Corresponding author. Email: ben.greatrex@une.edu.au
Lignocellulosic biomass pyrolysis with acid catalysis selectively produces the useful chiral synthon 6,8-dioxabicyclo
[3.2.1]oct-2-ene-4-one ((ꢀ)-levoglucosenone, LGO). In this report, LGO was used to prepare (3R,5S)-3-benzyl-
5-(hydroxymethyl)-4,5-dihydrofuran-2(3H)-one, which is an intermediate used in the construction of antivirals including
the protease inhibitor indinavir. To achieve the synthesis, the hydrogenated derivative of LGO was functionalised using
aldol chemistry and various aromatic aldehydes were used to show the scope of the reaction. Choice of base affected
reaction times and the best yields were obtained using 1,1,3,3-tetramethylguanidine. Hydrogenation of the a-benzylidene-
substituted bicyclic system afforded a 4 : 3 equatorial/axial mixture of isomers, which was equilibrated to a 97 : 3 mixture
under basic conditions. Subsequent Baeyer–Villiger reaction afforded the target lactone in 57 % overall yield for four
steps, a route that avoids the protection and strong base required in the traditional approach. The aldol route is contrasted
with the a-alkylation and a Baylis–Hillman approach that also both start with LGO.
Manuscript received: 26 April 2017.
Manuscript accepted: 13 June 2017.
Published online: 7 July 2017.
Introduction
the treatment of human immunodeficiency virus (HIV) infec-
tion, which was until recently on the World Health Organisation
(WHO) list of essential medicines.[18] The core pentanamide
motif contains two chiral centres and can be constructed
from (S)-5-(hydroxymethyl)-3-benzyldihydrofuran-2(3H)-one
(7a) (Scheme 1).[17] In the original description of the synthesis
The chiral synthon levoglucosenone[1] (1, 6,8-dioxabicyclo
[3.2.1]oct-2-en-4-one, LGO) has recently become available in
kilogram quantities owing to advances in reaction processing and
separation of the product from impurities.[2] LGO can be syn-
thesised in a single step from cellulose or lignocellulosic biomass
such as bagasse or sawdust, making it attractive as a chiral pool
material.[3–5] It has been used to prepare a variety of bioactive
compounds,[6] carbohydrate derivatives,[7,8] and precursors for
pharmaceuticals.[9] Reduction of LGO yields 6,8-bicyclo[3.2.1]
octan-4-one 2 (Scheme 1), which is under development as a
solvent (CyreneTM) owing to the scalability of its preparation and
unique solvent properties.[10,11] The current availability of 2 and
its functionality have led us to investigate the conversion of 2 into
chiral g-butyrolactones, which are important motifs in natural
products and bioactive molecules.[12]
The Baeyer–Villiger reaction can be used to generate
(S)-5-(hydroxymethyl)-4,5-dihydrofuran-2(3H)-one (4), a widely
used chiral synthon, from the 6,8-dioxabicyclo[3.2.1]octan-4-one
ring system.[9,13,14] The rigid bicyclic ring system in 1 and 2
can be employed for diastereoselective reactions[15] and, com-
bined with the Baeyer–Villiger reaction, allows the synthesis of
g-lactones with a variety of substitution patterns.[16] The high
migratory aptitude of the a-ketal means that excellent yields and
selectivity are obtained for lactones even if a0-substitution is
present.[16] The synthesis of 4 from LGO can be performed using
high-yielding steps and requires no water-sensitive reagents or
chromatography, which are clear advantages over the traditional
route that starts with glutamic acid.[17]
O
4
O
5
H2, Pd/C
neat
6
O
O
3
8
O
O
7
99 %
2
1
2
1
CH3CO3H, H2O
90–95 %
RO
HO2C
H2N
1. HNO2
2. BH3
LHMDS,
RO
O
O
SMe2
BnBr, THF
O
O
86 %
Ref. 19, 68 %
Ph
CO2H
3
HF
pyridine
97 %
TBSCl
pyridine
96 %
4 R ϭ H
6 R ϭ TBS
7a R ϭ H
Ref. 17
5 R ϭ TBS
OH
Ref. 17
O
HN
Ph
7a
N
OH
N
O
N
Indinavir, 8
N
H
Scheme 1. Preparation of lactone 7a from glutamic acid 3 via lactone 4 and
showing its incorporation into indinavir (8)[17] and preparation of lactone 4
from LGO 1.[9]
This chemistry is applicable to the synthesis of indinavir 8, a
highly active protease inhibitor developed by Merck and Co. for
Journal compilation Ó CSIRO 2017
www.publish.csiro.au/journals/ajc

Synthesis of an Indinavir Precursor
1147
of indinavir by Dorsey et al., the pentanamide was constructed
via the stereoselective benzylation of tert-butyldimethylsilyl
(TBS)-protected g-lactone 5.[17] This procedure required a five-
step synthesis starting with glutamic acid, a borane reduction to
obtain the alcohol 4,[19] and anhydrous conditions with strong
base for the benzylation reaction. The commercial route to
indinavir now avoids the use of 7a, and instead constructs the
pentanamide using the allylation of a 3-phenylpropanamide
derivative.[20,21] Recent patents describing bioactives synthe-
sised incorporating 7a have used the strategy of Dorsey et al. to
construct 7a, which indicates that this is still the best-known
route to this lactone.[22] Lactone 7a has also been used for the
synthesis of protease inhibitors active against indinavir-resistant
virus,[23] vasoactive intestinal peptide receptor inhibitors,[22]
and allylamides useful for the treatment of Alzheimer’s dis-
ease.[24] The importance of lactone 7a in the medicinal chemistry
literature led us to investigate methods to synthesise it from LGO,
but our previous attempts to insert benzyl groups on the 3-iodide
derivative of LGO using Suzuki–Miyaura chemistry were not
successful.[16] We now report on the efficiency of three different
approaches for the synthesis of lactone 7a using LGO, and
the development of a general strategy for the preparation of
3-(arylmethyl)-6,8-dioxabicyclo[3.2.1]octanones.
Table 1. Alkylation reactions of 2 with benzylbromide
O
O
O
O
Ph
Ph
BnBr
Ph
Ph
O
O
O
O
O
O
THF, Base
O
O
2
9a
10a
11
O
O
O
Ph
O
O
12
BnBr equiv. Base
Temp. [8C] Time [h] YieldA (9a : 10a : 11)
1.1
1.0
2.4
LDAC
tBuOK 25
tBuOK 25
ꢀ78 to 25
48
24
7 (1 : 3 : 26)B
31 (3 : 2 : 95)
88 (0 : 0 : 100)
1.5
AIsolated yield with ratio determined by gas chromatography–mass spec-
trometry (GC-MS).
B16 % of adduct 12 was also isolated.
CLDA ¼ lithium diisopropylamide.
Results and Discussion
We envisaged that the synthesis of lactone 7a could be
approached using direct alkylation of an enolate, aldol chem-
istry to introduce a benzylidene at the a-position of 2, or Baylis–
Hillman chemistry starting with LGO, the key step being the
C–C bond forming reaction at the a-centre. Watson et al.
recently reported that 2 undergoes aldol reactions to form a
dimer in the presence of base, providing support for an aldol
approach to generate 7a.[11] Furthermore, LGO is known to
undergo a stereoselective domino oxa-Michael–aldol reaction
with water and aromatic aldehydes, which would involve similar
reactivity.[25,26]
The conceptually simplest approach to the a-benzylated
derivatives via a-alkylation of ketone 2 consistently failed to
give appreciable amounts of monoalkylated products 9a or 10a
and reactions were dominated by the dibenzylated product 11
(Table 1). Lowering the temperature and using a strong base
afforded mainly dibenzylated adduct 11 as well as compound
12, which presumably resulted from a dimerisation of 2, subse-
quent g-deprotonation, and alkylation. The propensity for dia-
lkylation led us to increase the number of equivalents of halide; a
very good yield of the dialkylated adduct 11 was obtained. Thus,
the a-alkylation reaction was an excellent way to dibenzylate the
carbon adjacent to the carbonyl in 2, but efficient mono-
a-benzylation of the ketone using benzyl bromide and base
was not possible.
Turning our attention to the aldol reaction, an optimisation
study for the reaction of 2 with benzaldehyde was performed and
representative reactions are shown in Table 2. Initial attempts
using K2CO3 and NEt3 as base gave either none or very little
aldol adduct (Table 2, entries 1–3). Switching the base to
1,1,3,3-tetramethylguanidine (TMG), an excellent yield of ben-
zylidene was obtained; however, reactions took 24–48 h to reach
completion. Eliminating the solvent and running the reactions
neat gave a better yield of 13a and reduced the reaction time to
just 1 h. The geometry of the alkene in 13a was assigned as E on
the basis of a cross-peak in the 2D NOESY spectrum between
the aromatic ring protons and the methylene at C2. It should also
be noted that photochemical isomerisation of the E-enone to the
Z-enone was observed over time, requiring the products to be
kept in the dark.
With the success of the aldol reaction of 2 with benzalde-
hyde, we attempted to extend the derivatisation to other types of
aldehydes. Not only could these materials be used to construct
analogues of indinavir or the other bioactives prepared from
lactone 7a, the preparation of new chiral materials from such an
abundant precursor could lead to new chiral ligands or catalysts.
Thus, ketone 2 was reacted with a series of aromatic aldehydes
as shown in Table 2 to give the enones 13b–13i in moderate to
excellent yield. Although successful for the parent aldol reac-
tion, the neat reaction conditions were found to be unsatisfactory
for other aldehydes examined both in yield and product
distribution.
In addition to the aldol adducts, some double-bond migration
was observed, giving endocyclic enones 14b, d, f–h, which were
evident in the 1H NMR spectra of crude reaction mixtures. These
products lacked the H2a and H2b signals normally seen
1
between d 2.5 and 3.5 ppm in the H NMR spectrum, which
were replaced with a benzylic AB quartet at d 3.60–3.70 ppm.
The migration was more pronounced in reactions with aromatic
aldehydes substituted with electron-withdrawing groups
(Table 2, entries 10 and 19). We found that the isomerisation
was virtually eliminated when the reaction was performed in
acetonitrile at 508C and the amount of base was also limited. The
installation of an alkyl group using butyraldehyde was also
attempted; however, the yield of alkylidene 13j was poor,
indicating that an aldol approach was not feasible for alkyl
substituents.
To access these allylic rearrangement products, the isomer-
isation of 13b to endo-alkene 14b was attempted. No isomerisa-
tion was observed when 13b was heated in neat Hunig’s base at
¨
808C; however, isomerisation was observed by heating in TMG
at 1008C, although the reaction was slow and significant
decomposition occurred, makingisolation difficult. Thus, heating
13b at 1008C for 48h with TMG gave 13b/14b with a ratio of

1148
E. T. Ledingham, K. P. Stockton, and B. W. Greatrex
Table 2. Aldol condensations with 2 with aldehydes
O
O
O
O
O
R
R
R
O
O
O
Base, solvent
O
O
2
13a–j
14a–j
Entry
R
Equiv. RCHO
Solvent
Temp. [8C]
Base (equiv.)
Time [h]
Product
Yield (13 : 14)A
1
Ph
1.5
1.5
1.5
1.5
1.5
1.5
1.5
2.0
1.3
1.2
1.1
Dioxane
Dioxane
MeCN
THF
25
70
K2CO3 (1.0)
K2CO3 (1.0)
NEt3 (1.0)
TMG (2.0)
TMG (1.0)
TMG (1.0)
TMG (1.0)
TMG (2.0)
TMG (0.2)
TMG (1.0)
TMG (0.1)
72
24
72
24
45
45
18
1
13a/14a
13a/14a
13a/14a
13a/14a
13a/14a
13a/14a
13a/14a
13a/14a
13a/14a
13b/14b
13b/14b
Trace (NDE)
Trace (ND)
11 (ND)
2
3
70
4
70
75 (ND)
5
Dioxane
MeCN
DMSO
–
70
78 (ND)
6
70
74 (ND)
7
70
84 (99 : 1)
82 (99 : 1)
84 (99 : 1)
32 (8 : 25)
82 (.99 : 1)
8
100
100
100
50
9
–
9
10
11
–
2.5
17.5
MeCN
Cl
12
13
1.2
1.1
–
100
50
TMG (1.0)
TMG (1.0)B
1.5
96
13c/14c
13c/14c
58 (.99 : 1)
57 (93 : 7)
MeCN
O
14
15
1.2
1.1
–
100
50
TMG (1.0)
TMG (1.0)B
20
13d/14d
13d/14d
60 (10 : 1)
O
MeCN
140
58 (.99 : 1)
O
O2N
16
1.2
MeCN
50
TMG (0.1)
24
13e/14e
90 (.99 : 1)
17
18
1.2
1.1
–
100
50
TMG (1.0)
TMG (1.0)B
43
13f/14f
13f/14f
30 (25 : 6)
MeCN
100
53 (.99 : 1)
19
20
1.2
1.1
–
100
50
TMG (1.0)
TMG (0.1)
3.5
13g/14g
13g/14g
11 (20 : 11)
MeCN
16.5
87 (.99 : 1)
N
21
22
1.2
1.1
–
100
50
TMG (1.0)
TMG (0.1)
1.5
13h/14h
13h/14h
61 (.99 : 1)
15 (50 : 3)
O
O
MeCN
110
H
N
23
24
1.1
1.0
–
100
50
TMG (1.0)
TMG (0.1)
6.5
16
13i/14i
13j
48C (.99 : 1)
nBu
MeCN
22D
AIsolated yields with ratios determined using GC-MS.
BAdditional TMG was added during the course of the reaction to bring the number of equiv. from 0.1 to 1.0.
C(E)-13i : (Z)-13i ¼ 50 : 11 by GC-MS.
D13j was not separable from unidentified impurities.
END ¼ not determined.
1 : 4, which was comparable with that obtained using the neat
reaction conditions for 2.5h (Table 2, entry 10). Transition metal-
catalysed processes are known for isomerisation of an exocyclic
double bond into a ring giving cyclic a,b-unsaturated ketones;[27]
however, further examination of this isomerisation was beyond
the scope of the present work.
the reaction conditions, some E to Z isomerisation of the alkene
in 13k presumably occurred and subsequent cyclisation by
addition of the phenol to the ketone resulted in a stable hemi-
ketal. The stereochemistry of the alcohol is suggested owing to
the preferred a-facial addition of nucleophiles to the ketone in
2.[28] Furthermore, the alcohol formed by the addition to the
a-face can hydrogen bond with the neighbouring axial O-6,
which would further stabilise hemiketal 15.
The reaction of 2 with salicylaldehyde afforded hemiketal 15
containing a Z-double bond in excellent yield (Scheme 2). Under

Synthesis of an Indinavir Precursor
1149
R
O
R
O
O
O
HO
CH3CO3H
CH2Cl2
1. Pd/C, H2,
EtOAc
13f 2. (iPr)2NEt
HO
O
13a
or
O
O
R
O
HO
HO
O
O
67 %
O
O
2
O
O
TMG, 100ЊC
O
9
10
7a, R ϭ Ph, 87 %
7f, R ϭ 2-Napthyl
51 %
13k
9a/10a 97:3, R ϭ Ph, 80 %
15
9f/10f 93:7, R ϭ 2-Napthyl 97 %
Scheme 2. Aldol reaction of 2 with salicylaldehyde.
HO
PhCHO, DABCO
O
7αH
7βH
7αH
7βH
Ph
NH2CHO
O
O
Ph
1
3H
25 %
The conversion of 13a through to the lactone 7a was
investigated using hydrogenation conditions and is contrasted
with a Baylis–Hillman strategy starting with LGO in Scheme 3.
The Pd/C-mediated hydrogenation of benzylidene 13a afforded
a mixture with significant amounts of 10a consistent with the
hindrance encountered on the b-face of the bicyclic ring system.
The isomerisation of 10a to the more stable equatorially oriented
9a was achieved by stirring the mixture in Hunig’s base and gave
a 97 : 3 ratio of isomers at equilibrium. This two-step approach
gave consistently higher yields than the hydrogenation of 13a
using PdCl2/H2, which afforded 45 % of 9a with the isomerisa-
tion of 10a to 9a promoted by the HCl liberated on catalyst
activation. Conversion through to lactone 7a was achieved using
peracetic acid in 87 % yield and traces of the diastereomer were
removed by crystallisation. Matching the spectroscopic data
obtained for 7a with the literature[17] confirmed that the assigned
stereochemistry for the intermediate 9a was correct. Similarly,
the synthesis of 2-napthyl-substituted lactone 7f proceeded
through the mixture of 9f and 10f, although the isolated yield
of 7f was low following Baeyer–Villiger reaction owing to
difficulties removing the minor isomer.
O
O
O
H
O
H
2β
2β
O
O
2αH
16
2αH
3H
9a
10a
J2α-3 ϭ 9.8 Hz
Ph
2α-3 ϭ 11.6 Hz
PdCl2, H2
THF
J
9a/10a, 97:3, 49 %
Scheme 3. Preparation of 7a and 7f from 9a.
into 7a via the mixture of diastereomers 16 was thus achieved
in three steps when 16 was hydrogenated using PdCl2/H2
(Scheme 3). The yields obtained for the Baylis–Hillman reac-
tion were low, and the hydrogenation gave a mixture requiring
chromatography; however, a variety of catalysts are available
for this reaction and if optimised, it could be the preferred
route.
Conclusions
This report described the a-functionalisation of the 6,8-
dioxabicyclo[3.2.1]octan-3-one ring system using aldol reactions.
Subsequent reduction of the products and then Baeyer–Villiger
reaction afforded the valuable lactone 7a in an overall 57 % yield
from LGO. The preparation of 7a could also be achieved in only
three steps from LGO, albeit in low yield, using a Baylis–Hillman
approach. These compounds have previously been used to syn-
thesise the antiviral compound indinavir and a range of other
protease inhibitors, which demonstrates the utility of this
chemistry.
The NMR spectra, and in particular the coupling constants
for the ring protons of the reduction products 9a and 10a,
warrant discussion. Cross-peaks between H3 (2.85 ppm) and
H7b (3.90 ppm) in the 2D NOESY NMR were consistent with
the assigned stereochemistry of 9a. The C2 methylene in 9a
appeared as a non-first-order multiplet; however, NOESY
interactions of H2b with H7b allowed the differentiation of
H2a (downfield) and H2b (upfield) in the multiplet. As
expected for a chair-like conformation, a large JH3–H2a coupling
(11.4 Hz) was observed indicating a trans-diaxial relationship
Supplementary Material
4
for these protons. In the isomer 10a, a W-path J coupling
All experimental procedures, spectral data and copies of 1H and
13C NMR spectra for all new compounds are available on the
Journal’s website.
between H7a (3.73 ppm) and H2a (2.27 ppm) was used to assign
H2a. Surprisingly, 10a also exhibited large coupling constants
between H3 and H2a (JH3–H2a 9.8 Hz). A chair-like conforma-
tion for 10a should have given a small coupling constant
(4–7 Hz) between H3 (2.75 ppm) and H2a and the observed
coupling of 9.8 Hz indicates a distorted ring for 10a. Steric
interactions of the benzyl substituent at C3 and the bridging
methylene ether may cause the six-membered ring to flatten,
which closes the H2a–C2–C3–H3 dihedral angle, explaining the
observed coupling constants.
It was envisaged that the sequence from LGO to 7a could be
shortened if a Baylis–Hillman reaction was used to construct
the a-C–C bond. LGO has been employed in Baylis–Hillman
reactions crossed with aliphatic aldehydes promoted by diethy-
laluminium iodide;[29,30] however, there are no examples in the
literature of successful reactions of LGO with aromatic alde-
hydes. A survey of Baylis–Hillman conditions for the reaction
of 1 with benzaldehyde resulted in either no reaction or low
yields of product, with the best yield of 25 % obtained when the
reaction was promoted by formamide/1,4-diazabicyclo[2.2.2]
octane (DABCO).[31] A proof-of-principle conversion of LGO
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgements
We thank Mr Tony Duncan and Dr Warwick Rafferty from Circa Australia
for kindly supplying levoglucosenone 1 and ketone 2 (CyreneTM). E.T.L.
thanks the Australian government for an Australian Postgraduate Award
scholarship.
References
[1] Y. Halpern, R. Riffer, A. Broido, J. Org. Chem. 1973, 38, 204.
doi:10.1021/JO00942A005
[2] C. Lawrence, W. Raverty, A. Duncan, WO Patent Application WO/
2011/000030 2011.
[3] S. T. Walter, M. O. Jason, W. Chen, D. R. Kevin, B. A. Kirk, W. Yu,
Z. Huiyan, C. Steven, C. Synthia, in Carbohydrate Synthons in Natural
Products Chemistry (Eds Z. J. Witczak, K. Tatsuda) 2002, ACS

1150
E. T. Ledingham, K. P. Stockton, and B. W. Greatrex
Symposium Series, Vol. 841, pp. 21–31 (American Chemical Society:
Washington, DC).
[18] M. Boyd, Expert Opin. Pharmacother. 2007, 8, 957. doi:10.1517/
14656566.8.7.957
[4] F. Cao, T. J. Schwartz, D. J. McClelland, S. H. Krishna, J. A. Dumesic,
G. W. Huber, Energy Environ. Sci. 2015, 8, 1808. doi:10.1039/
C5EE00353A
[5] M. Sarotti, M. M. Zanardi, R. A. Spanevello, A. G. Suarez, Curr. Org.
Synth. 2012, 9, 439. doi:10.2174/157017912802651401
[6] D. Urabe, T. Nishikawa, M. Isobe, Chem. Asian J. 2006, 1, 125.
doi:10.1002/ASIA.200600038
[19] K. J. Stone, R. D. Little, J. Am. Chem. Soc. 1985, 107, 2495.
doi:10.1021/JA00294A049
[20] P. E. Maligres, S. A. Weissman, V. Upadhyay, S. J. Cianciosi, R. A.
Reamer, R. M. Purick, J. Sager, K. Rossen, K. K. Eng, D. Askin, R. P.
Volante, P. J. Reider, Tetrahedron 1996, 52, 3327. doi:10.1016/0040-
4020(95)01114-5
[21] K. Rossen, R. A. Reamer, R. P. Volante, P. J. Reider, Tetrahedron Lett.
1996, 37, 6843. doi:10.1016/0040-4039(96)01529-8
[22] D. W. Landry, M. Karayiorgou, S. Deng, U.S. Patent Application US
2015/0239831 2015.
[7] Z. J. Witczak, R. Chhabra, H. Chen, X.-Q. Xie, Carbohydr. Res. 1997,
301, 167. doi:10.1016/S0008-6215(97)00100-6
[8] E. D. Giordano, A. Frinchaboy, A. G. Sua´rez, R. A. Spanevello, Org.
Lett. 2012, 14, 4602. doi:10.1021/OL302061A
[9] A. Flourat, A. Peru, F. Teixeira, F. Brunissen, Allais, Green Chem.
[23] Z. Lu, S. Raghavan, J. Bohn, M. Charest, M. W. Stahlhut, C. A.
Rutkowski, A. L. Simcoe, D. B. Olsen, W. A. Schleif, A. Carella,
L. Gabryelski, L. Jin, J. H. Lin, E. Emini, K. Chapman, J. R. Tata,
Bioorg. Med. Chem. Lett. 2003, 13, 1821. doi:10.1016/S0960-894X(03)
00262-2
[24] R. Home, WO Patent Application WO2003030886A2 2003.
[25] V. Samet, D. N. Lutov, S. I. Firgang, K. A. Lyssenko, V. V. Semenov,
Tetrahedron Lett. 2011, 52, 3026. doi:10.1016/J.TETLET.2011.
04.004
2015, 17, 404. doi:10.1039/C4GC01231C
[10] J. Sherwood, A. Constantinou, L. Moity, C. R. McElroy, T. J. Farmer,
T. Duncan, W. Raverty, A. J. Hunt, J. H. Clark, Chem. Commun. 2014,
9650. doi:10.1039/C4CC04133J
[11] K. L. Wilson, A. R. Kennedy, J. Murray, B. Greatrex, C. Jamieson,
A. J. Watson, Beilstein J. Org. Chem. 2016, 12, 2005. doi:10.3762/
BJOC.12.187
[12] R. R. A. Kitson, A. Millemaggi, R. J. K. Taylor, Angew. Chem. Int. Ed.
2009, 48, 9426. doi:10.1002/ANIE.200903108
[26] T. Nishikawa, H. Araki, M. Isobe, Biosci. Biotechnol. Biochem. 1998,
62, 190. doi:10.1271/BBB.62.190
[13] K. Tomioka, T. Ishiguro, K. Koga, J. Chem. Soc. Chem. Commun.
1979, 652. doi:10.1039/C39790000652
[27] M. Tanaka, H. Mitsuhashi, M. Maruno, T. Wakamatsu, Chem. Lett.
1994, 23, 1455. doi:10.1246/CL.1994.1455
[14] F. Shafizadeh, R. H. Furneaux, T. T. Stevenson, Carbohydr. Res. 1979,
71, 169. doi:10.1016/S0008-6215(00)86069-3
[15] F. Shafizadeh, D. D. Ward, D. Pang, Carbohydr. Res. 1982, 102, 217.
doi:10.1016/S0008-6215(00)88064-7
[16] K. P. Stockton, C. J. Merritt, C. J. Sumby, B. W. Greatrex, Eur. J. Org.
[28] P. Tsypysheva, F. A. Valeev, E. V. Vasil’eva, L. V. Spirikhin, G. A.
Tolstikov, Russ. Chem. Bull. 2000, 49, 1237. doi:10.1007/BF02495766
[29] C. Marquis, F. Cardona, I. Robina, G. Wurth, P. Vogel, Heterocycles
2002, 56, 181. doi:10.3987/COM-01-S(K)16
[30] R. Demange, C. Buhlmann, P. Vogel, Helv. Chim. Acta 2003, 86, 361.
doi:10.1002/HLCA.200390037
Chem. 2015, 6999. doi:10.1002/EJOC.201501083
[17] B. D. Dorsey, R. B. Levin, S. L. McDaniel, J. P. Vacca, J. P. Guare,
P. L. Darke, J. A. Zugay, E. A. Emini, W. A. Schleif, J. Med. Chem.
1994, 37, 3443. doi:10.1021/JM00047A001
[31] V. K. Aggarwal, D. K. Dean, A. Mereu, R. Williams, J. Org. Chem.
2002, 67, 510. doi:10.1021/JO016073Y