Glycomimetics: A Versatile de Novo Synthesis of β-1-C-Aryl-deoxymannojirimycin Analogues

Article abstract of DOI:10.1021/jo970585o

The synthesis of 11 β-1-C-aryl-deoxymannojirimycin analogues using a versatile de novo strategy is described. The origins of this report are derived from several recent developments in the area of C₁-substituted glycomimetic polyhydroxylated piperidines. This study is designed as a structure-activity comparison to explore the effects of aryl substitution on the ability of the title compounds to inhibit glycosidase enzymes. The polyhydroxylated piperidine ring was constructed using vinyl bromide 10, which was synthesized in six steps from the bromobenzene microbial oxidation metabolite bromo diol 9. A palladium-catalyzed Suzuki cross-coupling of vinyl bromide 10 and the corresponding arylboronic acid served as the key pseudoanomeric carbon-carbon bond-forming step. Ozonolysis and selective reduction of the resultant carbonyl functions followed by reductive amination served to produce the azasugar ring. The complete NMR analysis of the resultant piperidine ring stereochemistry is also discussed. Fully deprotected β-1-C-aryl-deoxymannojirimycin analogues 8·HCl were obtained upon acidic deprotection.

Full text of DOI:10.1021/jo970585o

6046
J . Org. Chem. 1997, 62, 6046-6050
Glycom im etics: A Ver sa tile d e Novo Syn th esis of
â-1-C-Ar yl-d eoxym a n n ojir im ycin An a logu es
Carl R. J ohnson* and Brian A. J ohns
Department of Chemistry, Wayne State University, Detroit, Michigan 48202-3489
Received April 1, 1997X
The synthesis of 11 â-1-C-aryl-deoxymannojirimycin analogues using a versatile de novo strategy
is described. The origins of this report are derived from several recent developments in the area
of C1-substituted glycomimetic polyhydroxylated piperidines. This study is designed as a structure-
activity comparison to explore the effects of aryl substitution on the ability of the title compounds
to inhibit glycosidase enzymes. The polyhydroxylated piperidine ring was constructed using vinyl
bromide 10, which was synthesized in six steps from the bromobenzene microbial oxidation
metabolite bromo diol 9. A palladium-catalyzed Suzuki cross-coupling of vinyl bromide 10 and the
corresponding arylboronic acid served as the key pseudoanomeric carbon-carbon bond-forming
step. Ozonolysis and selective reduction of the resultant carbonyl functions followed by reductive
amination served to produce the azasugar ring. The complete NMR analysis of the resultant
piperidine ring stereochemistry is also discussed. Fully deprotected â-1-C-aryl-deoxymannojirimycin
analogues 8‚HCl were obtained upon acidic deprotection.
In tr od u ction
The inhibition of carbohydrate processing enzymes
called glycosidases by polyhydroxylated piperidines and
related alkaloids has become an area of vast academic
and industrial interest as attested to by literally hun-
dreds of publications and patents over the past several
years.1 These enzymes are responsible for many biologi-
cal phenomena such as digestion, biosynthesis of glyco-
proteins, as well as the trimming of cell-surface oligosac-
charides and hence play a role in cellular recognition,
an area of enormous implications. Glycosidase inhibitors
have potential in the treatment of viral infections,2
cancer,3 and diabetes and other metabolic disorders.4
These enzymes act upon the glycosidic linkage of oli-
gosaccharides and glycopeptides by stabilizing an inter-
mediate oxonium ion, thus facilitating the lysis and
modification of the anomeric center (Figure 1a).5 Poly-
hydroxylated piperidines, commonly referred to as “aza-
sugars”, display the same dense stereochemical informa-
tion as the common hexoses, but contain an NH in place
F igu r e 1. (a) Proposed transition state of enzymatic glycoside
lysis. (b) Comparison of charged character of oxonium ion (top)
with a protonated azasugar (bottom).
of the pyranose oxygen. The mechanism of inhibition of
glycosidase enzymes by azasugars is believed to be a
result of the protonation of the nitrogen of the piperidine
ring at biological pH, thus providing an electronic mimic
of the transition state oxonium ion involved in enzymatic
glycoside modification (Figure 1b). The pH dependence
of azasugar inhibition is believed to be a result of an ionic
interaction with a negatively charged carboxylate ion
present in the active site of the enzyme which is respon-
sible for the charge stabilization of the positively charged
high energy oxonium ion during normal hydrolysis.6,1a
When the azasugar is in a protonated form, this interac-
tion is possible, whereas the free base is not capable of
anything more than a hydrogen bonding interaction. It
is the optimization of the interaction between the posi-
tively charged substrate and the enzyme active site that
has carried this field of interest.
Most of the biologically interesting members of this
class of compounds are 1-deoxy analogues or diastereo-
mers of 1-deoxynojirimycin (DNJ ) (1), the stereochemical
equivalent of D-glucose (2). The presence of substitution
at the C1 position of azasugars is known to dramatically
effect the selectivity and potency of inhibition.1a Due to
the hydrolytic lability of the N/O aminal function, aza-
O-glycosides have not emerged as a viable class of
glycomimetics for studying glycosidase activity.7 This is
also evident in the relative stabilities of DNJ (1) and the
X Abstract published in Advance ACS Abstracts, August 1, 1997.
(1) (a) Winchester, B.; Fleet, G. W. J . Glycobiology 1992, 2, 199. (b)
Elbein, A. D. Annu. Rev. Biochem. 1987, 56, 497. (c) Fleet, G. W. J .;
Karpas, A.; Dwek, R. A.; Fellows, L. E.; Tyms, A. S.; Petursson, S.;
Namgoong, S. K.; Ramsden, N. G.; Smith, P. W.; Son, J . C.; Wilson, F.
X.; Witty, D. R.; J acob, G. S.; Rademacher, T. W. FEBS Lett. 1988,
237, 128. (d) Look, G. C.; Fotsch, C. H.; Wong, C.-H. Acc. Chem. Res.
1993, 26, 182. (e) Hughes, A. B.; Rudge, A. J . Nat. Prod. Rep. 1994,
135.
(2) (a) Sunkara, P. S.; Bowlin, T. L.; Liu, P. S.; Sjoerdsma, A.
Biochem. Biophys. Res. Commun. 1987, 148, 206. (b) Gruters, R. A.;
Neefjes, J . J .; Tersmette, M.; de Goede, R. E. Y.; Tulp, A.; Huisman,
H. G.; Miedema, F.; Ploegh, H. L. Nature 1987, 330, 74. (c) Walker,
B. D.; Kowalski, M.; Goh, W. C.; Kozarsky, K.; Krieger, M.; Rosen, C.;
Rohrschneider, L.; Haseltine, W. A.; Sodroski, J . Proc. Natl. Acad. Sci.
U.S.A. 1987, 84, 8120. (d) Karpas, A.; Fleet, G. W. J .; Dwek, R. A.;
Petursson, S.; Namgoong, S. K.; Ramsden, N. G.; J acob, G. S.;
Rademacher, T. W. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 9229.
(3) (a) Humphries, M. J .; Matsumoto, K.; White, S. L.; Olden, K.
Cancer Res. 1986, 46, 5215. (b) Spearman, M. A.; J amieson, J . C.;
Wright, J . A. Exp. Cell Res. 1987, 168, 116.
(4) (a) Truscheit, E.; Frommer, W.; J unge, B.; Mu¨ller, L; Schmidt,
D. D.; Wingender, W. Angew. Chem., Int. Ed. Engl. 1981, 20, 744. (b)
Horii, S.; Fukase, H.; Matsuo, T.; Kameda, Y.; Asano, N.; Matsui, K.
J . Med. Chem. 1986, 29, 1038. (c) Anzeveno, P. B.; Creemer, L. J .;
Daniel, J . K.; King, C.-H. R.; Liu, P. S. J . Org. Chem. 1989, 54, 2539.
(5) (a) J ensen, J . L.; Tsuang, S. C.; Uslan, A. H. J . Org. Chem. 1986,
51, 816. (b) Liu, P. S. J . Org. Chem. 1987, 52, 4717. (c) Cogoli, A.;
Semenza, G. J . Biol. Chem. 1975, 250, 7802. (d) Sinnot, M. L. Chem.
Rev. 1990, 90, 1171.
(6) Winchester, B.; Barker, C.; Baines, S.; J acob, G. S.; Namgoong,
S. K.; Fleet, G. W. J . Biochem. J . 1990, 265, 277.
S0022-3263(97)00585-9 CCC: $14.00 © 1997 American Chemical Society

Synthesis of â-1-C-Aryl-deoxymannojirimycin Analogues
J . Org. Chem., Vol. 62, No. 17, 1997 6047
parent nojirimycin (3) containing an anomeric hydroxyl.
The former does not contain any reactive functionality
at the anomeric center and hence cannot undergo any
deleterious reactions, yet still retains much of its inhibi-
tory properties. C-Glycosides have emerged as one pos-
sible answer to the hydrolytic stability problem of many
carbohydrates and are well studied for furanose, and
pyranose systems.8 Several studies concerning C1 carbon
substituted azasugars have appeared recently in the
literature.9 Included among these are the homoazasug-
ars such as homonojirimycin (HNJ ) (4). HNJ was iso-
lated from Omphaela diandra and shown to be an inhibi-
tor of several R-glycosidases;9a since then several syn-
theses of homoazasugars have also appeared. A potent
and specific glycosidase inhibitor, and an antidiabetic
drug candidate has been produced by glycosylation of the
C1 hydroxymethyl substituent of HNJ with the synthesis
of MDL 25,637 (5).5b Several reports of aza-C-disaccha-
rides such as 6 from our laboratories and others have
recently appeared.10 Biological results support the sug-
gestion that differing C1 substituents can dramatically
alter the potency and specificity of the azasugar.10f,9b
desired to construct C1-aryl-substituted azasugars. To
our knowledge only two such compounds have been re-
ported. J unge reported the synthesis of R-1-C-phenyl-
deoxynojirimycin (7a ),11 and Fleet reported the synthesis
of â-1-C-phenyl-deoxymannojirimycin (8a ).6,9b Of the two
reports, only Fleet presents biological studies. As is
common with most C1-substituted deoxymannojirimycin
derivatives, the â-phenyl derivative 8 shows no inhibition
of mannosidases, but is a potent inhibitor of R-L-fucosi-
dase. It has been suggested that the 4, 3, and 2 positions
of the fuco stereochemistry are mimicked by the 2, 3, and
4 hydroxyl groups in the manno configuration of the
inhibitor (Figure 2).6 The 1-â substituent can also likely
occupy the same stereochemical region as the methyl
group at the 5 position of fucose. We desired to further
explore the effects of aryl substitution on the ability of
the â-1-C-aryl-deoxymannojirimycin system to inhibit
glycosidases. We now report the synthesis of the parent
â-1-C-phenyl-deoxymannojirimycin (8a ) and 10 addi-
tional analogues (8b-k ) using a versatile route based on
a palladium-catalyzed Suzuki cross-coupling13 as the key
carbon-carbon pseudoglycosidic bond forming step.
F igu r e 2. (a) R-L-Fucose. (b) â-1-C-Substituted-deoxymanno-
jirimycin.
Resu lts a n d Discu ssion
The recent development of the microbial oxidation of
halobenzenes to produce enantiopure halogen-substituted
cyclohexadiene diols into a commercial process has at-
tracted our attention.12 Specifically, the bromo derivative
9 derived from bromobenzene was envisioned to be a very
useful intermediate due to its ability to be functionalized
at every position, and eventual use of the vinyl bromide
moiety as a partner in a palladium-catalyzed Suzuki
cross-coupling,13 an area of current interest in our
group.14,10f,17 Vinyl bromide 10 was prepared in six steps
from bromo diol 9 in an overall yield of 55-60% on a
multigram scale as previously reported (eq 1).10f
In order to further examine the effects of C1 substit-
uents on the biological properties of azasugars, we
(7) For a recent example of an azasugar thioglycoside see Suzuki,
K.; Hashimoto, H. Tetrahedron Lett. 1994, 35, 4119.
(8) (a) Postema, M. H. D. C-Glycoside Synthesis; CRC Press: Boca
Raton. FL, 1995. (b) Levy, D. E.; Tang, C. The Chemistry of C-
Glycosides; Elsevier Science: Oxford, 1995. (c) Postema, M. H. D.
Tetrahedron 1992, 48, 8545.
(9) (a) Kite, G. C.; Fellows, L. E.; Fleet, G. W. J .; Liu, P. S.; Scofield,
A. M.; Smith, N. G. Tetrahedron Lett. 1988, 29, 6483. (b) Fleet, G. W.
J .; Namgoong, S. K.; Barker, C.; Baines, S.; J acob, G. S.; Winchester,
B. Tetrahedron Lett. 1989, 30, 4439. (c) Liotta, L. J .; Lee, J .; Ganem,
B. Tetrahedron 1991, 47, 2433. (d) Bruce, I.; Fleet, G. W. J .; di Bello,
C.; Wincheser, B. Tetrahedron 1992, 48, 10191. (e) Henderson, I.;
Laslo, K.; Wong, C.-H. Tetrahedron Lett. 1994, 35, 359. (f) Holt, K.
E.; Leeper, F. J .; Handa, S. J . Chem. Soc., Perkin Trans. 1 1994, 231.
(g) Martin, O. R.; Saavedra, O. M. Tetrahedron Lett. 1995, 36, 799. (h)
Martin, O. R.; Xie, F.; Liu, L. Tetrahedron Lett. 1995, 36, 4027. (i)
Wong, C.-H.; Provencher, L.; Porco, J . A., J r.; J ung, S.-H.; Wang, Y.-
F.; Chen, L.; Wang, R.; Steensma, D. H. J . Org. Chem. 1995, 60, 1492.
(10) (a) J ohnson, C. R.; Miller, M. W.; Golebiowski, A.; Sundram,
H.; Ksebati, M. B. Tetrahedron Lett. 1994, 35, 8991. (b) Baudat, A.;
Vogel, P. Tetrahedron Lett. 1996, 37, 483. (c) Martin, O. R.; Liu, L.;
Yang, F. Tetrahedron Lett. 1996, 37, 1991. (d) Fre´rot, E.; Marquis,
C.; Vogel, P. Tetrahedron Lett. 1996, 37, 2023. (e) Saavedra, O. M.;
Martin, O. R. J . Org. Chem. 1996, 61, 6987. (f) J ohns, B. A.; Pan, Y.
T.; Elbein, A. D.; J ohnson, C. R. J . Am. Chem. Soc. 1997, 119, 4856.
(11) Bo¨shagen, H.; Geiger, W.; J unge, B. Angew. Chem., Int. Ed.
Engl. 1981, 20, 806.
(12) The diol derived from bromobenzene is now prepared in
crystalline form on a multikilogram scale by Genecor International,
Inc., Rochester, NY. For a nice overview of applications of such cis-
cyclohexadienediols see Hudlickly, T.; Thorpe, A. J . J . Chem. Soc.,
Chem. Commun. 1996, 1993.
(13) (a) Miyaura, N.; Yanagi, T.; Susuki, A. Synth. Commun. 1981,
11, 513. (b) Suzuki, A. Pure Appl. Chem. 1985, 57, 1749. (c) Miyaura,
N.; Suzuki, A. Chem. Rev. 1995, 95, 2457.

6048 J . Org. Chem., Vol. 62, No. 17, 1997
J ohnson and J ohns
Ta ble 1. Su zu k i Cr oss-Cou p lin g
Sch em e 1
entry
boronic acid
time, h
product
% yield
1
2
3
4
5
6
7
8
9
11a
11b
11c
11d
11e
11l
11m
11n
11o
11p
11q
22
12a
12b
12c
12d
12e
12l
12m
12n
12o
12p
12q
89
95
88
75
73
97
93
84
79
92
74
4.75
4.75
2
23
4.5
24
24
5.5
5.5
14
10
11
Ta ble 2. Th r ee-Step Aza su ga r Rin g Con str u ction
(Sch em e 1)
P a lla d iu m -Ca ta lyzed Su zu k i Cr oss-Cou p lin g. As
eluded to above, our synthetic strategy centers on the
palladium-catalyzed Suzuki cross-coupling of vinyl bro-
mide 10 with the requisite arylboronic acids to produce
trisubstituted olefins 12 (a -e, l-q) (Table 1). The cou-
plings proceeded smoothly using 10 mol % PdCl2(PPh3)2
as the catalyst and 2 M (aq) Na2CO3 as the base in
refluxing THF (Table 1). Under these conditions com-
plete consumption of the starting vinyl bromide was
accomplished without significant aryl-aryl homocoupling
or reduction products as is sometimes an issue during
such cross-coupling reactions.13 The choice of catalyst
and base were extremely important. The use of tetrakis-
(triphenylphosphine)palladium(0) as the catalyst and
Ba(OH)2‚8H2O as the base resulted in a 65% yield for
the coupling of p-methoxyphenylboronic acid (11e) with
vinyl bromide 10, but an attempt to couple phenylboric
acid (11a ) with 10 under similar conditions resulted in
the complete recovery of the vinyl bromide starting
material. The use of Na2CO3 as the base gave similar
results. The more reactive PdCl2(PPh3)2 catalyst con-
taining only two triphenylphosphine ligands proved to
be the superior choice and worked sufficiently for all
cases except for 2-nitrophenylboronic acid.15 The ex-
tremely electron deficient 2-nitrophenylboronic acid re-
sulted in only partial conversion so an alternative
coupling partner was sought (entry 9). The boronic acid
coupling partners ranged from electron rich (entries 3-8)
to slightly electron deficient (entry 10). It should also
be noted that the synthesis of boronic acids in entries 6,
9, and 11 resulted in products that contained several
boronic species by 1H NMR.16 This was likely due to
dimeric and cyclic trimers of the boronic acid derivatives
and was not a problem for the coupling step. The crude
boronic acid mixtures underwent coupling smoothly using
1.5 to 2 equiv of the requisite boronic acid to produce only
the desired trisubstituted olefinic products. It is known
that the Suzuki conditions will easily tolerate unpro-
entry
olefin
protected azasugar
% yield
1
2
3
4
5
6
7
8
9
12a
12b
12c
12d
12e
12l
12m
12n
12o
12p
12q
15a
15b
15c
15d
15e
15f
15g
15h
15o
15j
56
67
41
56
74
61
55
66
67
49
66
10
11
15k
tected alcohols and amines,13 but our synthesis required
that the phenolic and aniline functionality be protected
during the subsequent ozonolysis step discussed below.
The coupling of 2-(tert-butoxycarbonylamino)phenylbo-
ronic acid (11o) with vinyl bromide 10 (entry 9) resulted
in two products, the desired coupled product in 62% yield
along with the N-deprotected product in 26% yield
presumably derived from the basic hydrolysis of the BOC
carbamate under the Suzuki conditions. The N-depro-
tected byproduct was converted to the desired product
in 64% yield upon refluxing in THF with BOC2O.
Aza su ga r Rin g Con str u ction . The trisubstituted
olefins 12 underwent smooth ozonolysis to produce the
corresponding keto-aldehyde intermediate 13 (Scheme 1).
1H NMR of the crude product mixture suggested that the
open chain form of intermediate 13 was the major isomer
present of the several possible equilibrium forms. This
was confirmed by the presence of an NH signal between
5.5 and 6 ppm, as well as a deshielding in the aromatic
protons in conjugation with the ketone carbonyl. The
success of the ozonolysis was dependent on having the
aromatic nitrogen substituents protected with electron-
withdrawing groups (carbamates), and the oxygen sub-
stituents protected as their corresponding ethers in order
to minimize the possibility for ring oxidation and N-oxide
formation. Only the 2-methoxy derivative 12c led to
unidentified overoxidation products; this was minimized
by short reaction times, and the impurities were removed
by chromatography. The aldehyde carbonyl group was
selectively reduced in the presence of the equilibrating
ketone/aminal function using NaBH3CN in conjunction
(14) (a) J ohnson, C. R.; Braun, M. P. J . Am. Chem. Soc. 1993, 115,
11014. (b) Braun, M. P. Ph.D. Dissertation, Wayne State University,
1995.
(15) (2-Nitrophenyl)boronic acid has been cited as
a sluggish
coupling partner under similar conditions see Thompson, W. J ;
Gaudino, J . J . Org. Chem. 1984, 49, 5237.
1
with a AcOH/NaOAc (aq) pH 4 buffer in THF. The H
(16) Complex mixtures of several boronic acid/anhydride species
have also been noted by Hoye; see Hoye, T. R; Chen, M. J . Org. Chem.
1996, 61, 7940. These mixtures are difficult to characterize but
fortunately present no problems when subjected in crude form to the
Suzuki cross-coupling conditions resulting in only the desired coupled
products.
NMR of the crude corresponding keto-alcohols 14 were
not as clean as their precursors due to the presence of
several equilibrating species. Finally, the keto-alcohols
14 were subjected to reductive amination conditions (10%

Synthesis of â-1-C-Aryl-deoxymannojirimycin Analogues
J . Org. Chem., Vol. 62, No. 17, 1997 6049
along with NaBH3CN was prepared as a stock solution by
dissolving 1.5 g of sodium acetate along with 5.7 mL of glacial
acetic acid in 6 mL of H2O. Boronic acids 11a and 11p were
purchased from Aldrich Chemical Co. Boronic acids 11b, 11l,
and 11q were prepared via lithium-halogen exchange from
the corresponding aryl halide. Boronic acids 11l and 11q
resulted in complex mixtures of monomers along with dimeric
and trimeric anhydrides and were used without further
purification.16 The remaining boronic acids were prepared
according to the literature methods: 11c,15,18 11d ,e,19 11m ,20
11n ,21 11o.22
Gen er a l P r oced u r e A: Su zu k i Cr oss-Cou p lin g of Vin yl
Br om id e 10 w ith Ar ylbor on ic Acid s. Vinyl bromide 10 (1
mol equiv) was added to a 25-mL flask. The flask was fitted
with a reflux condenser and septum and flushed with argon.
THF was added to produce a 0.1 M solution, followed by the
arylboronic acid (1.5 to 2.0 mol equiv), and PdCl2 (PPh3)2 (0.1
mol equiv). To this yellow solution was added 2 M aqueous
Na2CO3 (4 mol equiv), and the mixture was heated to reflux
with vigorous stirring for 2-24 h. The reactions typically
turned dark within the first 5 min of heating. When the
reaction was deemed complete (absence of starting vinyl
bromide by TLC), it was cooled to rt and poured into H2O and
extracted with Et2O. The organic layers were washed with
brine and dried over Na2SO4. Filtration, removal of solvent,
and flash chromatography on silica gel gave the desired aryl-
substituted olefins 12.
Gen er a l P r oced u r e B: Ozon olysis, Ald eh yd e Red u c-
tion , a n d Su bsequ en t Red u ctive Am in a tion . A 50-mL
flask equipped with a side-arm gas bubbling inlet, and a gas
outlet was charged with the trisubstituted olefin 12 (1 mol
equiv) in CH2Cl2/MeOH (1:1) (0.2 M) and cooled to -78 °C.
An ozone/oxygen mixture was bubbled through the reaction
mixture at a rate of 2 L/min until no starting material
remained as judged by TLC (5-15 min). Dimethyl sulfide (1.5
mL) was added at -78 °C, and the reaction mixture was
allowed to warm to rt. The resultant solution was stirred for
1.5-2 h and then concentrated in vacuo. 1H NMR of the crude
product was obtained showing clean conversion to the keto-
aldhyde unless otherwise stated. The crude material was
taken up in a sufficient amount of THF to form a 0.07 to 0.1
M solution and cooled to 0 °C. NaBH3CN (1.3 mol equiv) was
added in one portion followed by 1-2 drops of a AcOH/NaOAc
pH 4 buffer solution; the resultant solution was allowed to
warm to rt and stirred for 2 h. The reaction mixture was
poured into Et2O and washed with NaHCO3 and brine. The
aqueous layers were extracted three times with Et2O, and the
combined organic layers were dried over Na2SO4. Filtration
and concentration in vacuo provided the corresponding crude
keto-alcohol. A high pressure tube was charged with the crude
keto-alcohol in MeOH (0.03 M), 10% Pd/C (Degussa type 50
wt % based on substrate), and H2 (40 psi). The mixture was
stirred vigorously for 12 h. Filtration through Celite and
concentration in vacuo, followed by flash chromatography on
silica gel, gave the desired protected azasugar 15.
F igu r e 3. (a) Preferred approach of the hydrogenation cata-
lyst. (b) Results of 1H NMR nuclear Overhauser enhancement
(NOE) studies of 15.
Pd/C Degussa type, H2, MeOH/EtOAc) to yield exclusively
one diastereomer of the corresponding protected azasug-
ars 15 (Table 2). Only one chromatographic purification
was necessary for the three-step conversion.
The stereochemical outcome of the reductive amination
was determined to be a â-oriented aryl group, corre-
sponding to hydrogen delivery from the R face (Figure
3a). The stereochemistry was confirmed by positive
nuclear Overhauser enhancements (NOE) between H1
and H5 which reside in a 1,3-diaxial relationship below
the piperidine ring, along with a coupling constant of 2.5
Hz between H1 and H2 consistent with an axial/equatorial
relationship as depicted in Figure 3b. Also worth noting
is the removal of the benzyl O-protecting groups and Cbz
N-protecting group in entries 6-8 and 11 respectively.
The nitro group in entry 10 was also reduced to an amine
under these conditions to yield the m-aniline derivative
15j. The hydrogenation conditions were initially a
concern due to the potential for hydrogenolysis of the
newly formed benzylic amine functionality at the pseudo-
anomeric center. This fortunately was not a problem in
any of the examples that we encountered.
Dep r otection . The remaining protecting groups on
15 were removed under acidic conditions (6 N HCl, THF,
12 h, rt) to produce the fully deprotected â-1-C-aryl-
deoxymannojirimycin azasugars 8 in high yield as their
HCl salts (Scheme 2).
Sch em e 2
Gen er a l P r oced u r e C: Acid Dep r otection of Aza su g-
a r s. A 25-mL flask was charged with the protected azasugar
derivative (10-30 mg) and 2 mL of THF was added followed
by the dropwise addition of 1 mL of 6 N HCl. The resultant
solution was stirred for 12 h and then concentrated in vacuo.
The residue was dissolved in a minimal amount of MeOH and
the deprotected sugar was precipitated with Et2O as the
corresponding HCl salt. After the solids settled, the excess
solvent was removed by pipette and the precipitate was
washed twice with Et2O. The white solids were dried in vacuo
providing the desired azasugar hydrochlorides (8‚HCl).
Con clu sion
In summary, an efficient and versatile method for the
construction of C1 aryl substituted azasugars containing
the manno configuration has been developed. This
method takes advantage of the biocatalytic synthesis of
enantiopure starting materials, as well as a mild and
contemporary method of carbon-carbon bond formation
in the use of the versatile Suzuki cross-coupling reaction.
Biological studies will be reported in due course. Current
work is in progress on methodology to access the corre-
sponding R-1-C-aryl-deoxymannojirimycin analogues.17
(17) J ohnson, C. R.; J ohns, B. A. Manuscript in preparation.
(18) Eggers, C. A.; Kettle, S. F. A. Inorg. Chem. 1967, 6, 160.
(19) Diorazio, L. J .; Widdowson, D. A.; Clough, J . M. Tetrahedron
1992, 48, 8073.
(20) Gleave, C. A.; Sutherland, I. O.; J . Chem. Soc., Chem. Commun.
1994, 1873.
(21) Mu¨ller, W.; Kipfer, P.; Lowe, D. A.; Urwyler, S. Helv. Chim.
Acta 1995, 78, 2026.
(22) Sharp, M. J .; Cheng, W.; Snieckus, V. Tetrahedron Lett. 1987,
28, 5093.
Exp er im en ta l Section
Gen er a l. Ozonolysis was performed using a OREC Model
03V5-0 ozonator. The pH 4 buffer used for selective reductions

6050 J . Org. Chem., Vol. 62, No. 17, 1997
J ohnson and J ohns
Typ ica l P r oced u r e for t h e P r ep a r a t ion of Bor on ic
Acid s. 4-ter t-Bu tylp h en ylbor on ic a cid (11b). A 100-mL
flask was charged with 1-bromo-4-tert-butylbenzene (1.0 g, 4.69
mmol) in 20 mL of THF and cooled to -78 °C. To this solution
was added tert-butyllithium (1.7 M/hexanes, 5.8 mL, 9.85
mmol) dropwise over 10 min. The resultant solution was
stirred at -78 °C for an additional 30 min. Trimethyl borate
was added rapidly via syringe, and the mixture was stirred
for 15 min at -78 °C followed by stirring at rt for 4 h. The
reaction mixture was then cooled to 0 °C, 15 mL of 2 N HCl
was added, and the mixture was stirred for 16 h. The mixture
was poured into Et2O, the layers were separated, and the
aqueous layer was washed three times with Et2O. The
combined organic layers were washed with brine and dried
over Na2SO4. Filtration, removal of solvent, and flash chro-
matography on silica gel (4:1 to 2:1 to 1:1 hexanes:EtOAc) gave
the desired boronic acid 11b in trimeric anhydride form as a
white powder: mp 182-185 °C; Rf 0.17 (4:1, hexanes:EtOAc);
CHCl3:hexane, 43.2 mg (67%) of azasugar derivative 15b as a
white solid: mp 272-275 °C dec; Rf 0.53 (2:1, hexanes:EtOAc);
IR (neat) 3397, 2956, 2930, 2857, 1568, 1245, 1221, 1135, 1115,
1
1071, 1051, 835, 778 cm-1; [R]20 +27.8 (c 1.08, CH3OH); H
D
NMR (500 MHz, CD3OD) δ 0.20 (s, 3 H), 0.24 (s, 3 H), 0.94 (s,
9 H), 1.30 (s, 3 H), 1.33 (s, 9 H), 1.61 (s, 3 H), 3.34 (ddd, 1 H,
J ) 10.0, 6.5, 3.5 Hz), 3.82 (dd, 1 H, J ) 11.5, 6.5 Hz), 4.02
(dd, 1 H, J ) 11.5, 3.5 Hz), 4.06 (dd, 1 H, J ) 10.0, 6.5 Hz),
4.22 (dd, 1 H, J ) 6.5, 5.5 Hz), 4.56 (dd, 1 H, J ) 5.5, 2.5 Hz),
4.85 (d, 1 H, J ) 2.5 Hz), 7.50 (s, 4 H); 13C NMR (125 MHz,
CD3OD) δ -4.75, -3.64, 19.17, 26.36, 26.50, 28.54, 31.78,
35.70, 59.65, 60.59, 63.08, 71.28, 76.62, 81.06, 111.78, 126.97,
129.40, 131.52, 153.95; HRMS (EI) calcd for C25H43NO4Si: (M+)
449.2961, found 449.2951.
â-1-C-(4-ter t-Bu tylp h en yl)-d eoxym a n n ojir im ycin Hy-
d r och lor id e (8b). Protected azasugar derivative 15b (36.5
mg, 0.081 mmol) was treated with HCl according to general
procedure C to provide 26.9 mg (99%) of azasugar 8b as its
1
IR (neat) 3077, 3032, 2964, 2904, 1608, 1405, 1350, 1312 cm-1
;
HCl salt: mp 199-201 °C; [R]20 +32.3 (c 1.15, CH3OH); H
D
1H NMR (300 MHz, CDCl3) δ 1.38 (s, 27 H), 7.54 (d, 6 H, J )
8.4 Hz), 8.17 (d, 6 H, J ) 8.4 Hz); 13C NMR (75 MHz, CDCl3)
δ 31.20, 35.05, 124.92, 135.57, 155.95; HRMS (EI) calcd for
C30H39O3B3: (M+) 480.3178, found 480.3172.
NMR (500 MHz, CD3OD) δ 1.32 (s, 9 H), 3.23 (ddd, 1 H, J )
10.5, 7.5, 3.0 Hz), 3.73 (dd, 1 H, J ) 9.5, 2.5 Hz), 3.87 (dd, 1
H, J ) 12.0, 7.5 Hz), 3.89 (dd, 1 H, J ) 10.5, 9.5 Hz), 4.01 (d,
1 H, J ) 2.5 Hz), 4.08 (dd, 1 H, J ) 12.0, 3.0 Hz), 4.49 (s, 1 H),
7.49 (s, 4 H); 13C NMR (125 MHz, CD3OD) δ 31.80, 35.63,
60.22, 63.34, 63.59, 68.01, 72.74, 75.65, 126.85, 128.80, 133.17,
153.40; HRMS (EI) calcd for (-HCl) C16H25NO4: (M+) 295.1784,
(M+ - CH3O) 264.1599, found 264.1595.
(3R,4R,5S,6R)-3-[(Ben zyloxycar bon yl)am in o)]-4-O-[(ter t-
b u t yld im et h ylsilyl)oxy]-5,6-O-(isop r op ylid en ed ioxy)-1-
(4-ter t-bu tylp h en yl)cycloh ex-1-en e (12b). Vinyl bromide
10 (164 mg, 0.32 mmol) was coupled to 4-tert-butylphenylbo-
ronic acid (114 mg, 0.64 mmol) following general procedure A
to provide, after flash chromatography on silica (8:1, hexanes:
EtOAc), 172 mg (95%) of olefin 12b as a clear oil: Rf 0.54 (4:
1, hexanes:EtOAc); IR (neat) 3449, 3333, 3085, 3062, 3027,
Ack n ow led gm en t. This work was supported by a
grant from the National Science Foundation (CHE-
9223011). We would like to thank Dr. Adam Golebio-
wski and Dr. Michael W. Miller for their valuable input
to this program. We acknowledge the generous as-
sistance of Genencor International for providing the key
intermediate (1S,2S)-3-bromo-3,5-cyclohexadiene-1,2-
diol and Dr. Gregg Whited of that company for helpful
discussions.
2954, 2930, 2857, 1728, 1501, 1215, 1070, 837, 777 cm-1; [R]20
D
-38.9 (c 0.92, CHCl3); 1H NMR (500 MHz, CDCl3) δ 0.08 (s, 3
H), 0.12 (s, 3 H), 0.86 (s, 9 H), 1.32 (s, 9 H), 1.34 (s, 3 H), 1.40
(s, 3 H), 4.29 (m, 1 H), 4.33 (t, 1 H, J ) 5.0 Hz), 4.69 (d, 1 H,
J ) 9.0 Hz), 5.01 (dd, 1 H, J ) 5.5, 1.0 Hz), 5.07 (d, 1 H, J )
10.0 Hz), 5.14 (s, 2 H), 5.81 (d, 1 H, J ) 1.0 Hz), 7.31-7.37
(m, 7 H), 7.42 (d, 2 H, J ) 8.5 Hz); 13C NMR (125 MHz, CDCl3)
δ -4.99, -4.63, 17.94, 25.68, 26.59, 27.76, 31.29, 34.51, 48.73,
66.77, 70.03, 72.62, 76.07, 109.76, 124.37, 125.43, 125.80,
128.08, 128.50, 135.77, 136.56, 136.93, 150.56, 155.74; HRMS
(EI) calcd for C33H47NO5Si: (M+) 565.3223 (M+ - C4H9)
508.2519, found 508.2522.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
1
tails and characterization data for 12, 15, and 8, and H NMR
and 13C NMR for 11b, 12, 15, and 8 (84 pages). This material
is contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
4-O-(ter t-Bu tyldim eth ylsilyl)-â-1-C-(4-ter t-bu tylph en yl)-
2,3-O-isop r op ylid en e-d eoxym a n n ojir im ycin (15b). Olefin
12b (80.0 mg, 0.14 mmol) was subjected to the conditions of
general procedure B to provide, after recrystallization from
J O970585O