α-1-C-Butyl-1,4-Dideoxy-1,4-Imino-L-Arabinitol as a second-Generation iminosugar-based oral α-Glucosidase inhibitor for improving postprandial hyperglycemia

Article abstract of DOI:10.1021/jm301304e

We report on the synthesis and the biological evaluation of a series of α-1-C-alkylated 1,4-dideoxy-1,4-imino-l-arabinitol (LAB) derivatives. The asymmetric synthesis of the derivatives was achieved by asymmetric allylic alkylation, ring-closing metathesis, and Negishi cross-coupling as key reactions. α-1-C-Butyl-LAB is a potent inhibitor of intestinal maltase, isomaltase, and sucrase, with IC₅₀ values of 0.13, 4.7, and 0.032 μM, respectively. Matrix-assisted laser desorption ionization time-of-flight mass spectrometric analysis revealed that this compound differs from miglitol in that it does not influence oligosaccharide processing and the maturation of glycoproteins. A molecular docking study of maltase-glucoamylase suggested that the interaction modes and the orientations of α-1-C-butyl-LAB and miglitol are clearly different. Furthermore, α-1-C-butyl-LAB strongly suppressed postprandial hyperglycemia at an early phase, similar to miglitol in vivo. It is noteworthy that the effective dose was about 10-fold lower than that for miglitol. α-1-C-Butyl-LAB therefore represents a new class of promising compounds that can improve postprandial hyperglycemia.

Full text of DOI:10.1021/jm301304e

Article
pubs.acs.org/jmc
α‑1‑C‑Butyl-1,4-dideoxy-1,4-imino‑L‑arabinitol as a Second-
Generation Iminosugar-Based Oral α‑Glucosidase Inhibitor for
Improving Postprandial Hyperglycemia
Atsushi Kato,*,† Erina Hayashi,† Saori Miyauchi,† Isao Adachi,† Tatsushi Imahori,‡ Yoshihiro Natori,‡
Yuichi Yoshimura,‡ Robert J. Nash,§ Hideyuki Shimaoka,∥ Izumi Nakagome,⊥ Jun Koseki,⊥
Shuichi Hirono,⊥ and Hiroki Takahata*,‡
†Department of Hospital Pharmacy, University of Toyama, Toyama 930-0194, Japan
‡Faculty of Pharmaceutical Sciences, Tohoku Pharmaceutical University, Sendai 981-8558, Japan
§Institute of Biological, Environmental and Rural Sciences/Phytoquest Limited, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB,
United Kingdom
∥S-BIO Business Division, Simitomo Bakelite Company Limited, Tokyo 140-0002, Japan
⊥School of Pharmaceutical Sciences, Kitasato University, Tokyo 108-8641, Japan
S
* Supporting Information
ABSTRACT: We report on the synthesis and the biological
evaluation of a series of α-1-C-alkylated 1,4-dideoxy-1,4-imino-
L-arabinitol (LAB) derivatives. The asymmetric synthesis of the
derivatives was achieved by asymmetric allylic alkylation, ring-
closing metathesis, and Negishi cross-coupling as key reactions.
α-1-C-Butyl-LAB is a potent inhibitor of intestinal maltase,
isomaltase, and sucrase, with IC50 values of 0.13, 4.7, and 0.032
μM, respectively. Matrix-assisted laser desorption ionization
time-of-flight mass spectrometric analysis revealed that this
compound differs from miglitol in that it does not influence
oligosaccharide processing and the maturation of glycopro-
teins. A molecular docking study of maltase-glucoamylase
suggested that the interaction modes and the orientations of α-
1-C-butyl-LAB and miglitol are clearly different. Furthermore, α-1-C-butyl-LAB strongly suppressed postprandial hyperglycemia
at an early phase, similar to miglitol in vivo. It is noteworthy that the effective dose was about 10-fold lower than that for miglitol.
α-1-C-Butyl-LAB therefore represents a new class of promising compounds that can improve postprandial hyperglycemia.
sensitizers such as biguanides and thiazolidinediones, insulin
mimetics such as glucagon-like peptide analogues and agonists,
α-glucosidase inhibitors, and DPP IV inhibitors). However,
some of these drugs have unacceptable side effects in some
patients or lose their effectiveness over time. Therefore, the
search for new antidiabetic drugs has continued to attract
considerable interest.
Among these drugs, α-glucosidase inhibitors are oral
antidiabetic agents which suppress postprandial hyperglycemia
by inhibiting the hydrolysis of disaccharides such as maltose,
isomaltase, and sucrose in the brush border membrane and
delaying the absorption of carbohydrates from the small
intestine. A recent metaregression analysis revealed that a close
relationship exists between 2 h postprandial glucose levels and
cardiovascular risk below the diabetic threshold,10 and there is
growing epidemiological evidence to show that an association
INTRODUCTION
■
Diabetes mellitus, one of the most common chronic metabolic
diseases, occurs when the pancreas produces insufficient levels
of insulin, or when the body cannot effectively use the insulin it
produces.1−3 Type 2 diabetes is induced by a combination of
factors, including lifestyle and genetic factors. Today, the
worldwide prevalence of diabetes is taking on pandemic
dimensions, as changing lifestyles lead to reduced physical
activity and increased obesity. In 2010, 285 million people were
suffering from diabetes, and this number is estimated to
increase to 439 million by 2030.4 In recent years, it has been
proposed that acute blood glucose elevations induce vascular
damage through direct action on the vascular endothelium and
finally cause myocardial infarctions, coronary heart disease, and
cerebral apoplexy.5 Therefore, the careful control of blood
glucose levels can delay or protect at-risk subjects from
developing cardiovascular diseases.6−9 Antidiabetic drugs are
currently classified into several different mechanistic classes
(e.g., insulin secretagogues such as sulfonylureas, insulin
Received: September 11, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
between postprandial hyperglycemia and macrovascular
complications exists in diabetic individuals.11−13 Therefore, α-
glucosidase inhibitors have considerable potential for prevent-
ing these cardiovascular risks. At present, a large number of
compounds mimicking the structures of monosaccharides or
oligosaccharides have been discovered from natural sources,
which include thiosugars.14−16 Among them, N-containing α-
glucosidase inhibitors belong to two different main chemical
classes. The first class is carbasugar-based inhibitors such as
acarbose and voglibose. Carbasugars are carbocyclic analogues
of hexopyranoses in which the ring oxygen has been replaced
by carbon.17 5a-Carba-α-D-glucopyranosylamines such as
validamine,18 valienamine,19 and valiolamine20 in particular
have proved to be important lead compounds for the
development of clinically important antidiabetic agents. The
first commercially available antidiabetic α-glucosidase inhibitor
acarbose was introduced into the market in Germany in 1990
and has since been successfully marketed in Europe and Latin
America. Voglibose was designed from valiolamine. It was
produced by the reductive amination of valiolamine with
dihydroxyacetone and was intended to function as an oral
antidiabetic agent.21 The second class of α-glucosidase
inhibitors is iminosugar-based inhibitors such as miglitol.
Iminosugars are sugar mimics with a nitrogen atom in place
of the ring oxygen of monosaccharides. Miglitol was designed
from 1-deoxynojirimycin (DNJ). DNJ has an excellent α-
glucosidase inhibitory activity in vitro,22 whereas its efficacy in
vivo is only moderate.23 Therefore, a large number of DNJ
derivatives have been prepared in the hopes of increasing in
vivo activity for this class of compounds.24−26 Finally, N-
(hydroxyethyl)-DNJ (miglitol) was reported to be the most
favorable inhibitor out of a large number of in vitro active
agents, and it was introduced into the market in 1999 as a more
potent second-generation α-glucosidase inhibitor.
Although the mechanisms of action against intestinal
enzymes for these two classes of α-glucosidase inhibitors are
similar, each has a distinct clinical strategy. The most important
arguing point is whether absorption from the intestinal tract is
of benefit to the clinical effects. Glycosidases, which are
involved in several important anabolic and catabolic processes,
act not only on intestinal digestion but also on lysosomal
catabolism and the post-translational modification of glyco-
proteins. The inhibition of liver and lysosomal glycosidases is
occasionally responsible for the development of serious hepatic
dysfunctions.27 Thus, carbasugar-based drugs were designed to
not cross the intestinal membranes. Acarbose is poorly
absorbed from the gastrointestinal tract, the mean systemic
bioavailability being only between 0.5% and 1.6%.28 Similarly,
voglibose is designed with a propanediol group from valiol-
amine for the purpose of disrupting its absorption. However,
this poor absorption means that the drug stays in the intestine
and colon for a long period of time and is associated with a high
frequency of side effects such as flatulence, diarrhea, and
hypoglycemia. In contrast, the iminosugar-based drug miglitol
was designed to be nearly completely absorbed from the
intestinal tract and may possess systemic effects in addition to
affecting the intestinal border.29,30 Previous experiments
showed that the plasma half-life of miglitol is 2.4 h with a
plasma peak concentration of 1.1 mg/L and a peak Tmax of 2.3 h
after an oral administration of 2 mg/kg of body weight.31
Moreover, these studies revealed that there are two different
glucose suppression mechanisms associated with miglitol, both
direct and indirect. It directly and reversibly inhibits the α-
glucosidase enzyme, and it induces the secretion of the
glucagon-like peptide-1 (GLP-1). GLP-1 is a gut incretin
hormone that is considered to be a promising therapeutic agent
for type 2 diabetes because it stimulates β-cell proliferation,
insulin secretion, and sensitivity in a glucose-dependent
manner.32,33 However, this absorption activity invites the risk
of potential hepatic dysfunction.
Thus, our strategy has been to design effective iminosugar-
type α-glucosidase inhibitors that affect postprandial hyper-
glycemia that not only show strong inhibitory effects against
maltase, isomaltase, and sucrase but also do not interfere with
the processing of glycoproteins. As part of our research
program aimed at developing new potent and selective α-
glucosidase inhibitors, we reported that the synthetic
enantiomer 1,4-dideoxy-1,4-imino-L-arabinitol (LAB) is a
more active and selective inhibitor of α-glucosidases than the
natural product 1,4-dideoxy-1,4-imino-D-arabinitol (DAB).34 In
addition, the introduction of a branched carbon chain at the C4
position led to an improved sucrase inhibition (IC50 = 0.66
μM).35 Considering this remarkable activity, we envisaged that
a further improvement in activity might be gained in the case of
structures having a longer side chain. To test this hypothesis, a
series of α-1-C-alkylated LAB derivatives were then prepared by
asymmetric allylic alkylation (AAA), ring-closing metathesis
(RCM), and Negishi cross-coupling as key reactions. We herein
describe the synthesis and biological evaluation of a series of α-
1-C-alkylated LAB derivatives, which includes their suppressive
effects on blood glucose levels using a carbohydrate loading test
and their effects on whole cell glycoproteins, as determined by
matrix-assisted laser desorption ionization time-of-flight mass
spectrometric (MALDI-TOF MS) analysis. We also studied the
selectivity of α-1-C-butyl-LAB for intestinal α-glucosidases vs
intestinal β-glycosidases and lysosomal β-glucocerebrosidase.
An additional aim was to demonstrate the molecular docking
properties of α-1-C-butyl-LAB with maltase with the intention
of determining the extent of difference from miglitol.
CHEMISTRY
■
Initially, the bicyclic 2,5-dihydropyrrole 2 was selected as the
common intermediate for the synthesis of 1a−l (Scheme 1).36
The iodide 2 was prepared by iodination of oxazolidinol 3,
obtained by a modified Trost procedure.37 With the iodide 2 in
hand, Csp3−Csp3 bond formation using metal-catalyzed cross-
a
Scheme 1. Synthesis of α-l-C-Alkyl-LAB
a
Reagents: (a) I2, Ph3P, imidazole, CH2C12; (b) bis(l,5-
cyclooctadiene)nickel(0), (S,S)-2,6-bis(4-isopropyl-2-oxazolin-2-yl)-
pyridine, then alkylzinc bromide, N,N-dimethylacetamide; (c) ethyl-
enediaminetetraacetic acid, 1,1,1-trifluoroacetone, Oxone, NaHCO3,
CH3CN, H2O; (d) trifluoroacetic acid, THF/H2O; (e) NaOH, EtOH/
H2O.
B
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
coupling was examined owing to homologation of the alkyl
substituents. Among the several procedures examined, the
Negishi cross-coupling reaction developed by Fu et al.38,39 was
successful and gave the desired alkyl-substituted oxazolidinones
4a−l (Table 1). The epoxidation of olefin in 4a−l with the
with inverse selectivity. These selectivities were previously
observed in a similar example by Trost et al.37 Epoxidation of
11 included the relative amount of two epoxides, 13, which
were cleaved by hydrolysis to give the same triol 14 as a single
diastereomer. Probably, SN2 attack of water occurs from the
sterically less hindered side of both epoxides to produce the
same product 14.34 Finally, hydrogenolysis of 14 afforded the
desired tetrol 7. Similarly, 12 was transformed via 15 and 16
into 8.
Table 1. Negishi Cross-Coupling of 2 with Alkylzinc Halides
Using Catalytic Ni(COD)2 in the Presence of a Tridentate
Nitrogen Ligand [(R,R)-i-Pr-Pybox]
In addition, the N-butyl derivative 17 of 1c and compound
18 with a 3-butyl substituent in place of the 3-hydroxy of 1c
were prepared as exhibited in Scheme 3. Reductive amino
alkylation of 1c with butyraldehyde in the presence of
NaBH3CN gave N-butyl-1-C-butyl-LAB (17). Next the epoxide
of 5c was cleaved by n-butyllithium in the presence of CuI to
provide 19, which was hydrolyzed with base to give 1,3-C-
dibutyl-LAB (18).
a
entry
R′ZnX
product
yield (%)
1
2
MeZnI
EtZnI
4a
4b
4c
4d
4e
4f
69
63
78
63
69
69
68
72
66
71
73
76
3
n-PrZnBr
4
n-BuZnBr
5
n-C5H11ZnBr
n-C6H13ZnBr
n-C7H15ZnBr
n-C8H17ZnBr
n-C9H19ZnBr
n-C10H21ZnBr
Ph(CH2)3ZnBr
(CH3)2CH(CH2)2ZnBr
6
7
4g
4h
4i
8
BIOLOGICAL RESULTS AND DISCUSSION
■
9
Evaluation of N- or 1-C-Butyl-LAB and -DNJ Deriva-
tives as Glycosidase Inhibitors. In connection with our
recent work concerning the synthesis of pyrrolidine-based
iminosugars with a branched carbon chain of biological
interest,35,41 we turned our attention to the design of potent
and selective α-glucosidase inhibitors with the goal of reducing
postprandial hyperglycemia. We first evaluated N-butyl-LAB
and α-1-C-butyl-LAB, since a previous structure−activity
relationship study indicated that the natural product DAB is a
good inhibitor, but the enantiomer LAB is a more potent
inhibitor of α-glucosidases.42 N-Butyl-1-deoxynojirimycin (N-
butyl-DNJ) and α-1-C-butyl-DNJ were also tested to extend
our understanding of the effect of introducing an alkyl
substituent into an iminosugar. We found that the inhibitory
potency of the parent DNJ was slightly better than that of LAB
against α-glucosidases (Table 3). However, DNJ showed a
broad inhibition spectrum against β-glucosidases, β-galactosi-
dases, and trehalases. The introduction of an N-butyl
substituent into LAB and DNJ reduced their effectiveness as
glycosidase inhibitors. In sharp contrast, the introduction of a
butyl chain at C-1 of LAB to give α-1-C-butyl-LAB dramatically
improved its inhibitory potency against sucrase, with an IC50
value of 0.032 μM. We were also intrigued by the fact that α-1-
C-butyl-LAB showed no significant inhibition toward liver and
lysosomal β-glucosidases and β-galactosidases, even at a
concentration of 1000 μM. α-1-C-Butyl-DNJ was a potent
inhibitor of isomaltase, whereas this compound was also a good
inhibitor of human lysosome β-glucocerebrosidase, with an IC50
value of 70 μM. Previous reports pointed out that these
nonspecific inhibitions may cause undesirable side effects in the
treatment of diabetes. In sharp contrast, α-1-C-butyl-LAB
showed no significant inhibition toward β-glucocerebrosidase.
To clarify the structural basis of the interaction of α-1-C-butyl-
DNJ and α-1-C-butyl-LAB with human β-glucocerebrosidase,
we constructed three-dimensional structures of the human β-
glucocerebrosidase complexed with α-1-C-butyl-LAB or α-1-C-
butyl-DNJ. As shown in Figure 1, α-1-C-butyl-DNJ and α-1-C-
butyl-LAB were found to bind to β-glucocerebrosidase with the
same orientation. Both α-1-C-butyl-DNJ and α-1-C-butyl-LAB
interacted with aromatic groups such as Tyr313 and Phe246.
However, the estimated binding energy (ΔE) of α-1-C-butyl-
DNJ (−219.08 kcal mol−1) with β-glucocerebrosidase indicated
that it was more stable than 1-C-butyl-LAB (−185.87 kcal
mol−1) (Table 4). These differences (ΔΔE) in binding energies
10
11
12
4j
4k
4l
a
Isolated yield.
dioxirane, generated in situ from Oxone with 1,1,1-trifluor-
oacetone, resulted in the stereoselective formation of epoxides
5a−l. Since the alkyl substituents of 4 are oriented toward the
axial position of an approximate planar bicycloxazolone in a
molecular model, the stereochemistry of the epoxide was
deduced to be anti with respect to the alkyl substituents. Ring-
opening of the epoxide with aqueous TFA stereoselectively
proceeded to provide diols 6a−l. The stereochemistries of diols
of 6c were determined to be 6α and 7β configurations, because
an NOE correlation was observed between the hydrogen at C-5
and the hydrogen at C-7. Finally, the oxazolinones 6a−l were
transformed by basic hydrolysis into the desired iminosugars
1a−l (Table 2).40
Table 2. Yields of the Three-Step Sequence from 4 to 1
a
entry
R′
product
yield (%)
1
2
Me
1a
1b
1c
1d
1e
1f
51
58
43
52
55
59
35
49
30
45
55
62
Et
3
n-Pr
4
n-Bu
5
n-C5H11
n-C6H13
n-C7H15
n-C8H17
n-C9H19
6
7
1g
1h
1i
8
9
10
11
12
n-C10H21
1j
Ph(CH2)3
1k
1l
(CH3)2CH(CH2)2
a
Isolated yield.
Next the tetrols 7 and 8 were prepared from an enantiomer
(2R,5R configuration), 9, of the known ent-9,34 as shown in
Scheme 2. Treatment of the Weinreb amide 9 with
propylmagnesium bromide gave the ketone 10 in 78% yield,
which was reduced with DIBAL-H to afford the β-alcohol 11
and the α-alcohol 12 in a ratio of 4:1. On the other hand, the
reduction of 10 with NaBH4 provided 12 and 11 in a 2:1 ratio
C
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of 7 and 8
a
Reagents: (a) PrMgBr, THF, 78%; (b) DIBAL-H, Et2O, 65% for 11, 18% for 12; (c) NaBH4, MeOH, 24% for 11, 51% for 12; (d)
ethylenediaminetetraacetic acid, 1,1,1-trifluoroacetone, Oxone, NaHCO3, CH3CN, H2O, 83% for 13, 80% for 15; (e) trifluoroacetic acid, THF/H2O,
93% for 14, 16% for 16; (f) H2, 5% Pd−C, concd HC1, MeOH, 68% for 7, 94% for 8.
a
more potent as an isomaltase inhibitor than α-1-C-octyl-LAB
(1g). These results suggest that the change from LAB to DNJ
in the iminosugar portion of the molecule results in a dramatic
enhancement in enzyme specificity and α-1-C-alkyl-LAB was
clearly different from α-1-C-alkyl-DNJ. For comparison, we
next investigated whether the introduction of different
substituents in the anomeric position had an influence on the
inhibition activities of intestinal maltase, isomaltase, and sucrase
(Table 5). Changing the butyl group to a 4-phenylbutyl group
or a 4-methylpentyl group at the C-1 position results in a
greater specificity for inhibiting maltase; 1k and 1l showed
almost the same inhibition potency against maltase, with IC50
values of 0.22 and 0.19 μM, respectively, but its inhibition
toward isomaltase and sucrose was reduced. It is noteworthy
that the C-1′ OH group dramatically changed the inhibition
potency and spectrum. We previously reported that the 1-C-
hydroxymethyl derivative of LAB, as in L-DMDP, was found to
display inhibitory activity toward intestinal maltase, isomaltase,
and sucrase, with IC50 values of 1.2, 0.77, and 0.62 μM,
respectively.43 These values are comparable to those of LAB. In
contrast, remarkable reductions in isomaltase inhibition were
observed in the case of α-1-C-{(R)-1-hydroxybutyl}-LAB (8),
which is indeed a 1004-fold weaker inhibitor than L-DMDP,
while α-1-C-{(S)-1-hydroxybutyl}-LAB (7) showed a compa-
rable inhibition toward maltase and sucrase. The N-butylation
(17) and 3-C-butylation (18) of α-1-C-butyl-LAB obviously
decreased the affinity for these α-glucosidases. Acarbose
showed inhibitory activity against maltase and sucrase, with
IC50 values of 0.18 and 2.9 μM, respectively. However, acarbose
exhibited less than a 50% inhibition of isomaltase even at
concentrations as high as 1000 μM. Voglibose was found to be
a much stronger inhibitor than acarbose. The IC50 inhibitory
activities toward maltase, isomaltase, and sucrase were 0.12, 5.2,
and 37 μM, respectively. Miglitol showed potent inhibitory
activity against maltase and sucrase, with IC50 values of 1.3 and
1 μM, respectively, but these values were weaker than those of
voglibose. In addition, the most striking difference between
miglitol and α-1-C-butyl-LAB is that miglitol showed an
unfavorable inhibition against rat intestinal cellobiase and
lactase, with IC50 values of 164 and 41 μM, respectively,
Scheme 3. Synthesis of 17 and 18
a
Reagents: (a) NaBH3CN, butyraldehyde, CH3OH, 46%; (b) CuI,
Et2O, then n-BuLi, n-hexane, 23%; (c) NaOH, EtOH/H2O, 38%.
may explain the differences in inhibition potency against β-
glucocerebrosidase.
Optimization of the α-1-C-Alkyl Chain Length. On the
basis of these findings, it clearly appeared that the addition of
an alkyl chain at C-1 of LAB might lead to the production of
highly potent and selective inhibitors of intestinal maltase and
sucrase. Thus, we focused on the anomeric position and
whether extending the alkyl chain at this position could
influence the inhibition activities of LAB (Figure 2). Of these
compounds, the introduction of short (ethyl and propyl) and
longer (decyl and undecyl) groups reduced the inhibition
activities compared to that of LAB (Table 5). Furthermore, we
found that the inhibitory potency of α-1-C-alkyl-LAB against
isomaltase derivatives decreased with increasing length of the
alkyl chain. There is, therefore, a minimum 1-C-alkyl chain
length requirement for achieving a strong inhibition of maltase
and sucrase, with a butyl group being optimal. To check if the
change from LAB to DNJ in the iminosugar part had an
influence on activity toward intestinal α-glucosidases, we
prepared α-1-C-alkyl-DNJs 19−21. This comparison revealed
that the inhibitory activity of α-1-C-alkyl-DNJs against maltase
and sucrase was even weaker than that of α-1-C-alkyl-LABs. For
example, α-1-C-butyl-DNJ (19) is 92-fold and 119-fold less
active on maltase and sucrase than the corresponding α-1-C-
butyl-LAB (1c). In contrast, α-1-C-octyl-DNJ (21) is 625-fold
D
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. Concentration of Iminosugars Giving 50% Inhibition of Various Glycosidases
a
NI = no inhibition (less than 50% inhibition at 1000 μM).
Figure 1. Docking view of (a) α-1-C-butyl-DNJ and (b) α-1-C-butyl-LAB with human β-glucocerebrosidase.
oligomeric assembly of secretory proteins, plasma membrane
proteins, and proteins destined for the various organelles of the
vacuolar system.44,45 The conformational maturations of these
functional glycoproteins are determined not only by the amino
acid sequence but also by post-translational modifications and
by a variety of chaperones and folding enzymes such as ER α-
glucosidases and α-mannosidases.46,47 The N-linked glycosyla-
tion of nascent proteins involves the transfer of
Glc3Man9GlcNAc2 from a dolichol precursor to the asparagine
residue of the Asn-X-Ser/Thr glycosylation sequon of a protein.
The oligosaccharide is subsequently processed by the sequential
action of trimming glycosidases in the ER, prior to the
production of hybrid and complex oligosaccharide structures in
the Golgi.48,49 This process of trimming commences with ER α-
glucosidase I, which removes the outer α-1,2-linked glucose
residue from the oligosaccharide, followed by the removal of
the remaining two glucose residues through the action of an ER
α-glucosidase II. In many cases, misfolded and genetic mutant
Table 4. Inhibitory Activity (IC50), IFDScore, and Binding
Energy of α-1-C-Butyl-DNJ and α-1-C-Butyl-LAB for Human
β-Glucocerebrosidase
IC50
ΔE
ΔΔE
(μM) IFDScore Δ(IFDScore) (kcal/mol) (kcal/mol)
α-1-C-butyl-
70
−890.68
−889.14
−1.53
−219.08
−185.87
−33.21
DNJ
a
α-1-C-butyl-
LAB
NI
0.00
0.00
a
NI = less than 50% inhibition at 1000 μM.
whereas α-1-C-butyl-LAB had no effect on these enzymes (data
not shown). On the basis of a comparison of these
commercially available inhibitors, the α-1-C-butyl-LAB pre-
pared in this study obviously has potential for clinical use.
Structural Modification of Whole Cell N-Linked
Oligosaccharides. The lumen of the endoplasmic reticulum
(ER) is a highly specialized compartment for the folding and
E
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 5. IC50 Values (μM) for α-1-C-Alkyl-LAB against
Intestinal α-Glucosidases, Compared with α-1-C-Alkyl-DNJs
19 and 20, Acarbose, Voglibose, and Miglitol
IC50 (μM)
compound
maltase
isomaltase
sucrase
α-1-C-ethyl-LAB (1a)
α-1-C-propyl-LAB (1b)
α-1-C-butyl-LAB (1c)
α-1-C-pentyl-LAB (1d)
α-1-C-hexyl-LAB (1e)
α-1-C-heptyl-LAB (1f)
α-1-C-octyl-LAB (1g)
α-1-C-nonyl-LAB (1h)
α-1-C-decyl-LAB (1i)
α-1-C-undecyl-LAB (1j)
α-1-C-(4-phenylbutyl)-LAB (1k)
α-1-C-(4-methylpentyl)-LAB (1l)
α-1-C-{(S)-1-hydroxybutyl}-LAB (7)
α-1-C-{(R)-1-hydroxybutyl}-LAB (8)
N-butyl-α-1-C-butyl-LAB (17)
1,3-C-dibutyl-LAB (18)
α-1-C-butyl-DNJ (19)
α-1-C-hexyl-DNJ (20)
α-1-C-octyl-DNJ (21)
acarbose
2.6
11
0.68
0.26
0.032
0.19
0.11
0.24
0.45
1.4
1.7
47
0.13
0.71
0.51
0.38
0.32
0.84
1.2
4.7
18
11
16
75
171
606
NI
14
1.7
3.9
13
0.22
0.19
3.3
0.31
0.24
0.68
5.7
12
96
89
773
NI
NI
0.45
0.31
0.12
NI
5.2
39
25
8.8
a
NI
NI
12
3.8
4.7
2.4
2.1
1.5
0.18
0.12
1.3
2.9
voglibose
0.37
1.0
miglitol
a
NI = less than 50% inhibition at 1000 μM.
2637), and Glc1Man9GlcNAc2 (m/z 2475) being accumulated.
The presence of these glucose-residue-containing oligosacchar-
ide structures is highly indicative of the inhibition of the
processing enzymes α-glucosidase I and II. In sharp contrast,
the control and α-1-C-butyl-LAB had very similar patterns
(Figure 3a,c). These results clearly indicate that there is a risk
associated with the use of miglitol, since it could increase the
expression of immature high-mannose-type N-linked oligosac-
charides, whereas α-1-C-butyl-LAB did not appear to have any
effect on the processing α-glucosidases.
Carbohydrate Loading Test. A recent metaregression
analysis revealed that 2 h postprandial glucose levels and
cardiovascular risk were closely related.10 Furthermore, there is
a substantial body of epidemiological evidence that points to an
association between postprandial hyperglycemia and macro-
vascular complications.12,55 Therefore, α-glucosidase inhibitors
have considerable potential for reducing the risk of these
diseases. The iminosugar-based α-glucosidase inhibitor miglitol
differs from carbasugar-type α-glucosidase inhibitors such as
acarbose and voglibose, because it is rapidly and completely
absorbs in the upper region of the small intestine.31 This
pharmacokinetic property consequently enables the strong and
early phase suppression of postprandial hyperglycemia, with no
or minimal gastrointestinal complications. We then investigated
the influence of α-1-C-butyl-LAB and miglitol, which was used
as a positive control, on blood glucose levels after an in vivo
maltose loading (Figure 4). The control group was loaded with
saline only. When compared to the control, α-1-C-butyl-LAB
caused significantly lower blood glucose levels at 15 and 30 min
compared to a placebo administration (Figure 4b). Moreover,
treatment with α-1-C-butyl-LAB led to a 43.9% decrease of the
area under the curve (AUC) at a dose of 0.25 mg/kg of body
weight. It was interesting to note that the remaining blood
Figure 2. Chemical structures of α-1-C-alkylated LAB derivatives and
sugar-mimic α-glucosidase inhibitors.
proteins are retained in the ER and are eventually degraded by
ER-associated degradation (ERAD).47,50,51 Thus, considering
the fact that α-glucosidase inhibitors could be clinically used as
antidiabetes agents, it is important that they do not disrupt the
folding and maturation processes of glycoproteins. Previous
studies have reported that DNJ, which is a parent compound of
miglitol, inhibited not only intestinal α-glycosidases but also the
processing ER enzymes α-glucosidase I and II.52,53 Therefore,
the use of DNJ could drastically change the glycan structures
on glycoproteins, since it could prevent the transformation of
high mannose- or hybrid-type glycans into complex-type
glycans. To examine the effect of miglitol and α-1-C-butyl-
LAB on whole cell N-linked oligosaccharide structures, we used
the recently established chemoselective glycoblotting technique
and MALDI-TOF MS analysis.54 Human hepatocellular
carcinoma HepG2 cells were treated with either 500 μM
miglitol or 500 μM α-1-C-butyl-LAB for 48 h. All of the N-
linked oligosaccharides were selectively captured onto new
high-density hydrazide beads, which permitted them to be
efficiently purified from cell homogenates. The captured N-
linked oligosaccharides were subjected to on-bead methyl
esterification to stabilize the sialic acids for the simultaneous
quantitation of neutral and sialylated oligosaccharides by
MALDI-TOF MS. As shown in Figure 3b and Table 6, a 48
h treatment with 500 μM miglitol resulted in immature
Glc3Man9GlcNAc2 (m/z 2799), Glc2Man9GlcNAc2 (m/z
F
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 3. MALDI-TOF MS spectra showing the effects of α-glucosidase-inhibiting iminosugars on whole cell glycoproteins. Relative expression
levels of glucose residues containing N-linked oligosaccharides on human hepatocelluar carcinoma HepG2 cells cultured for 48 h with or without 500
μM (b) miglitol and (c) α-1-C-butyl-LAB. Structural identification of glycans was performed by MS analysis and the use of a database for glycan
structures (http://web.expasy.org/glycomod/).
Table 6. Structures and Proportion (%) of Isolated Whole Cell N-Linked Oligosaccharide Peaks from the Control and 500 μM
Miglitol- and 500 μM α-1-C-Butyl-LAB-Treated HepG2 Cells
glucose levels were slightly higher with α-1-C-butyl-LAB than
with the placebo at 60 and 120 min (Figure 4b). This behavior
with strong and early phase suppression of postprandial
hyperglycemia was similar to that observed for miglitol (Figure
4a). For comparison, we next investigated the influence of α-1-
C-butyl-LAB using a sucrose loading test (Figure 5). The
administration of sucrose (2.5 g/kg of body weight, po) to
fasted mice resulted in a rapid increase in blood glucose
concentrations from 80 8 mg/dL to a maximum of 181 16
mg/dL after 30 min. Thereafter blood glucose levels recovered
to the pretreatment level at 120 min. α-1-C-Butyl-LAB resulted
in significantly lower blood glucose levels at 15, 30, and 60 min
than placebo administration (Figure 5b). Moreover, the AUC
for plasma glucose from 0 to 60 min after α-1-C-butyl-LAB
administration was reduced by 49.1% at a dose of 0.1 mg/kg of
body weight and 69.2% at a dose of 0.5 mg/kg of body weight,
compared with the AUC for the control. This behavior was
similar to that observed for miglitol administration (Figure 5a).
G
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
butyl-LAB is absorbed from the upper region of the small
intestine, much the same as miglitol, and it causes less diarrhea
than acarbose and voglibose.
Docking Studies of α-1-C-Butyl-LAB and Miglitol with
Intestinal Maltase. In piperidine pyranoside mimics such as
α-1-C-butyl-DNJ and miglitol, there is usually a good overlap
between the structure of the glycoside and inhibition by the
corresponding iminosugar mimic. However, α-1-C-butyl-LAB is
an unusual L-type pyrrolidine iminosugar, and it does not seem
to overlap extensively with D-type piperidine iminosugars. To
understand the structural basis of the interaction of α-1-C-
butyl-LAB and miglitol with maltase, we first determined the
mode of inhibition and the inhibition constant (Ki) of miglitol
and α-1-C-butyl-LAB from Lineweaver−Burk plots. As shown
in Figure 6, miglitol and α-1-C-butyl-LAB inhibited maltase in a
Figure 4. Effects of miglitol (a) and α-1-C-butyl-LAB (b) on blood
glucose levels. Blood glucose concentrations of male ddY mouse after
an oral load with maltose, 2.5 g/kg of body weight, with (a) 3.0 (light
green) and 6.0 (green) mg/kg of body weight miglitol and (b) 0.25
(orange) and 0.5 (red) mg/kg of body weight α-1-C-butyl-LAB. The
control group was loaded with saline (purple). Each value represents
the mean
SEM (n = 5). Three asterisks indicate a significant
difference (p < 0.001) compared with the control.
Figure 6. Lineweaver−Burk plots of miglitol (a) and α-1-C-butyl-LAB
(b) inhibition of rat intestinal maltase. An increasing concentration of
maltose or sucrose was used to determine the Ki values, and the data
were plotted as 1/V versus 1/[S]. (a) The concentrations of miglitol
were 0 μM (closed circles), 2.0 μM (closed triangles), and 5.0 μM
(closed squares). The calculated Ki value was 0.46 μM. (b) The
concentrations of α-1-C-butyl-LAB were 0 μM (closed circles), 0.2 μM
(closed triangles), and 0.75 μM (closed squares). The calculated Ki
value was 0.085 μM.
Figure 5. Effects of miglitol (a) and α-1-C-butyl-LAB (b) on blood
glucose levels. Blood glucose concentrations of male ddY mouse after
an oral load with sucrose, 2.5 g/kg of body weight, with (a) 2.5 (light
green) and 5.0 (green) mg/kg of body weight miglitol and (b) 0.1
(orange) and 0.5 (red) mg/kg of body weight α-1-C-butyl-LAB. The
control group was loaded with saline (blue). Each value represents the
mean SEM (n = 5). Three asterisks indicate a significant difference
(p < 0.001) compared with the control.
competitive manner, with Ki values of 0.46 and 0.085 μM,
respectively. This result suggests that α-1-C-butyl-LAB occupies
the active site of this enzyme, much the same as miglitol. Thus,
we next attempted to construct three-dimensional structures of
the N-terminal catalytic domain of maltase-glucoamylase
(ntMGAM) of rat complexed with α-1-C-butyl-LAB. For
comparison, we also constructed three-dimensional structures
of rat ntMGAM complexed with α-1-C-butyl-DNJ or miglitol.
As shown in Figure 7, α-1-C-butyl-LAB, α-1-C-butyl-DNJ, and
miglitol were observed to bind to the hydrophobic pocket
consisting of Trp376, Trp516, Trp613, and Phe649 of rat
ntMGAM. However, it was suggested that the orientations of
the butyl groups of α-1-C-butyl-LAB and α-1-C-butyl-DNJ are
different. The butyl group of α-1-C-butyl-LAB has a favorable
interaction with the hydrophobic pocket (Trp376, Leu405, and
Ile441), while the butyl groups of α-1-C-butyl-DNJ bind to the
other hydrophobic pocket (Trp481, Met519) (Figure 7). The
interaction between the butyl group of α-1-C-butyl-LAB and
the hydrophobic pocket of ntMGAM looks important for
ligand binding, which conforms to the study58 by Aguilar-
Moncayo et al. pointing out the critical role of aglyconic
interactions in glycosidase binding. The hydroxyethyl group of
miglitol also binds to the same hydrophobic pocket as α-1-C-
butyl-DNJ, but miglitol has an additional interaction with
hydrophilic residues (Asp616 and Arg600) of rat ntMGAM.
The interactions between rat ntMGAM and α-1-C-butyl-LAB,
The administration of miglitol led to the suppression of the
blood glucose levels in a dose-dependent manner; as a
consequence, the AUC for plasma glucose from 0 to 60 min
after miglitol administration was reduced by 40.5% at a dose of
2.5 mg/kg of body weight and 58.7% at a dose of 5.0 mg/kg of
body weight. Remarkably, the effective doses of α-1-C-butyl-
LAB were about 10 times lower than those of miglitol in both
disaccharide loading tests. The LAB derivative showed a
suppressive effect against postprandial hyperglycemia at a does
of only 0.1−0.5 mg/kg of body weight (Figures 4 and 5). A
recent study suggested that miglitol suppresses glucose levels
via two different mechanisms, namely, the direct inhibition of
α-glucosidase and, indirectly, via affecting the secretion of GLP-
1.56,57 GLP-1 is released from the intestine in response to
nutrients and exerts a potent insulin-releasing effect on
pancreatic β-cells.32,33 The important point to note is that
the GLP-1 release effect appears only when it is administered
orally.13 Further studies will clearly be needed to more
completely understand the mechanisms of α-1-C-butyl-LAB,
but our results could be interpreted to mean that it affects to
GLP-1 secretion because this compound is not only an
iminosugar but also an absorbable α-glucosidase inhibitor
similar to miglitol. On the basis of these findings and by
comparison with our results for miglitol, it appears that α-1-C-
H
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 7. View of interactions of (a) α-1-C-butyl-LAB, (b) α-1-C-
butyl-DNJ, and (c) miglitol with rat ntMGAM.
α-1-C-butyl-DNJ, or miglitol are illustrated in Figure 8. Miglitol
and α-1-C-butyl-DNJ are positioned to form hydrogen bonds
with Asp404, Asp518, Asp616, and His674 of rat ntMGAM,
while α-1-C-butyl-LAB can form hydrogen bonds with Asp518,
Asp616, Trp613, and Asp645 of rat ntMGAM. For the purpose
of clarifying the difference, the binding energies between the
compounds and rat ntMGAM were calculated. The binding
energies (ΔE) for α-1-C-butyl-LAB, α-1-C-butyl-DNJ, and
miglitol were −269.97, −236.53, and −261.36 kcal mol−1
(Table 7). These results were consistent with the experimental
results (IC50) showing that miglitol and α-1-C-butyl-LAB have
higher inhibitory activities against maltase of rat ntMGAM than
α-1-C-butyl-DNJ. For further confirmation, we also constructed
the three-dimensional models of the α-1-C-butyl-LAB−human
ntMGAM complex. The docking calculation was performed
under the same conditions as the docking calculations of rat
ntMGAM using the X-ray structures of human ntMGAM (PDB
IDs 2QLY, 2QMJ, 3CTT, and 3L4W). The binding energy
(ΔE), as estimated from the three-dimensional structures of the
compound with human ntMGAM, suggested that α-1-C-butyl-
LAB would be an inhibitor of human maltase as well as that
from the rat (Table 8). Furthermore, the binding free energy
(ΔG) estimated by the MM-GBSA method using Prime in the
Figure 8. Schematic interaction diagrams between rat ntMGAM and
(a) α-1-C-butyl-LAB, (b) α-1-C-butyl-DNJ, and (c) miglitol. Hydro-
gen bonds are depicted by dashed lines.
Table 7. Inhibitory Activity (IC50) and Binding Energy (ΔE)
of α-1-C-Butyl-LAB, α-1-C-Butyl-DNJ, and Miglitol with Rat
ntMGAM
IC50 (μM)
ΔE (kcal/mol)
Schrodinger Suite strongly supported that α-1-C-butyl-LAB has
̈
inhibition activity in human ntMGAM.
α-1-C-butyl-LAB
α-1-C-butyl-DNJ
miglitol
0.13
12
−269.97
−236.53
−261.36
CONCLUSIONS
1.3
■
In this study, we report on the design and synthesis of a series
of α-1-C-alkylated LAB derivatives by AAA, RCM, and Negishi
cross-coupling as key reactions. Among the produced
compounds, α-1-C-butyl-LAB showed potent inhibitory activity
toward intestinal maltase, isomaltase, and sucrase, with IC50
values of 0.20, 4.7, and 0.032 μM, respectively. The following
are the main features of α-1-C-butyl-LAB: (a) It does not
inhibit intestinal β-glycosidases and lysosomal β-glucocerebro-
sidase even at concentrations as high as 1000 μM. (b) MALDI-
TOF MS analysis revealed that it obviously differs from miglitol
in that it does not disrupt oligosaccharide processing and the
I
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
combined and dried over anhydrous Na2SO4. Filtration and
evaporation in vacuo furnished the crude product, which was purified
by column chromatography (silica gel, 80 g, hexane/EtOAc = 2/1) to
provide 2 (5.16 g, 88%) as a pale yellow oil. [α]2D3 = +141.6 (CHCl3, c
Table 8. Binding Energy (ΔE) and Binding Free Energy
(ΔG) Estimated for α-1-C-Butyl-LAB with Human and Rat
ntMGAM
1
ΔE (kcal/mol)
ΔG (kcal/mol)
= 1.00). H NMR (CDCl3, 400 MHz): δ 3.29−3.37 (m, 2H), 4.24
(dd, J = 8.7, 6.8 Hz, 1H), 4.64 (t, J = 8.7 Hz, 1H), 4.79 (br s, 1H),
4.86−4.90 (m, 1H), 6.03 (d, J = 6.3 Hz, 1H), 6.08 (d, J = 6.8 Hz, 1H).
13C NMR (CHCl3, 100 MHz): δ 8.60, 64.51, 67.20, 68.84, 130.73,
rat ntMGAM
−269.97
−260.51
−42.13
−46.99
human ntMGAM
133.28, 162.11. IR (neat, cm−1): 1749. HRMS (EI): m/z calcd for
C7H8NO2I (M+) 264.9600, found 264.9603.
maturation of glycoproteins. (c) It showed a strong and early
phase suppression of postprandial hyperglycemia, and the
effective dose was 10 times lower than that for miglitol. (d)
Both α-1-C-butyl-LAB and miglitol are competitive inhibitors,
whereas molecular docking studies revealed that the interaction
modes and the orientations of the alkyl chain of α-1-C-butyl-
LAB and miglitol are clearly different. Each alkyl group has a
favorable interaction, but with different hydrophobic pockets.
(e) Docking calculations using X-ray structures of human
ntMGAM (PDB IDs 2QLY, 2QMJ, 3CTT, and 3L4W)
suggested that the binding conformations and orientations of
α-1-C-butyl-LAB and miglitol in human ntMGAM are similar
to those in rat ntMGAM. α-1-C-Butyl-LAB therefore represents
a new class of promising compounds that have the potential for
treating postprandial hyperglycemia.
Preparation of Alkylzinc Bromide (n-Propylzinc Bromide). Iodine
chips (127 mg, 0.50 mmol) were added to a suspension of zinc powder
(981 mg, 15.0 mmol) in N,N-dimethylacetamide (10 mL) at room
temperature. After being stirred for 10 min, the resulting mixture
changed from a dark brown suspension to a colorless suspension. 1-
Bromopropane (0.91 mL, 10.0 mmol) was added by syringe, and the
reaction mixture was stirred at 80 °C for 3 h. The concentration in
N,N-dimethylacetamide was ∼0.75 M.
Typical Procedure for Negishi Coupling (4c). Yellow bis(1,5-
cyclooctadiene)nickel(0) (48.4 mg, 0.176 mmol, 16 mol %) and (S,S)-
2,6-bis(4-isopropyl-2-oxazolin-2-yl)pyridine (106 mg, 0.352 mmol, 32
mol %) were added to N,N-dimethylacetamide (7.0 mL) under an
argon atmosphere, and the resulting mixture was stirred for 30 min at
room temperature. The resulting deep-blue solution was added to a
solution of the n-propylzinc bromide (in N,N-dimethylacetamide, 4.7
mL, 3.52 mmol) and 2 (291 mg, 1.10 mmol). After being stirred for 20
h, the reaction was quenched with iodine chips (440 mg). After being
stirred for 10 min, the dark-brown mixture was passed through a short
pad of silica gel (eluting with AcOEt/hexane = 1/1) (to remove N,N-
dimethylacetamide, inorganic salts, and iodine). The filtrate was then
concentrated, and the residue was purified by flash chromatography
(silica gel, 20 g, eluting with hexane/EtOAc = 3/1) to provide 4c (156
mg, 78%) as a colorless oil.
(5S,7aR)-5-Butyl-1,7a-dihydropyrrolo[1,2-c]oxazol-3(5H)-one
(4c). [α]2D4 = +175.4 (CHCl3, c = 1.08). 1H NMR (CDCl3, 400 MHz):
δ 0.91 (t, J = 6.8 Hz, 3H), 1.34−1.58 (m, 6H), 4.20 (dd, J = 8.2, 5.3
Hz, 1H), 4.54 (br s, 1H), 4.58 (t, J = 8.7 Hz, 1H), 4.72−4.77 (m, 1H),
5.85 (ddd, J = 6.3, 1.4 Hz, 1H), 6.02 (ddd, J = 6.3, 2.4 Hz, 1H). 13C
NMR (CHCl3, 100 MHz): δ 13.94, 22.44, 28.30, 34.02, 63.82, 67.41,
68.72, 127.97, 135.23, 162.98. IR (neat, cm−1): 2930, 1752. HRMS
(EI): m/z calcd for C10H15NO2 (M+) 181.1103, found 181.1099.
Typical Procedure for Epoxidation (5c). To a vial with 4c (158.0
mg, 0.872 mmol) were added 6.4 mL of CH3CN and 4.4 mL of 4 ×
10−4 M ethylenediaminetetraacetic acid in H2O. The solution was
cooled to 0 °C, and 1,1,1-trifluoroacetone (977 mg, 8.72 mmol) was
added. A mixture of solid Oxone (2.68 g, 4.36 mmol) and NaHCO3
(549 mg, 6.54 mmol) was added in four portions over 45 min. The
reaction was stirred for 2 h at 0 °C, then diluted with 5 mL of H2O,
and extracted with CH2Cl2 (3 × 15 mL). The organic layers were
combined and dried over anhydrous Na2SO4. Filtration and
evaporation in vacuo furnished the crude product (231 mg), which
was purified by column chromatography (silica gel, 6 g, Et2O) to
provide epoxide 5c (171 mg, 99%) as a colorless oil.
(1aR,1bS,6S,6aS)-6-Butyltetrahydrooxireno[2′,3′:3,4]pyrrolo[1,2-
c]oxazol-4(1aH)-one (5c). [α]2D2 = +48.2 (CHCl3, c = 0.12). 1H NMR
(CDCl3, 400 MHz): δ 0.90 (t, J = 7.2 Hz, 3H), 1.26−1.58 (m, 6H),
3.54 (d, J = 2.9 Hz, 1H), 3.61 (d, J = 2.9 Hz, 1H), 3.97 (dd, J = 8.7, 4.3
Hz, 1H), 4.03 (dd, J = 9.2, 4.8 Hz, 1H), 4.46 (dd, J = 9.2, 4.3 Hz, 1H),
4.53 (t, J = 8.7 Hz, 1H). 13C NMR (CDCl3, 100 MHz): δ 13.91, 22.39,
28.19, 29.73, 55.27, 57.20, 58.77, 60.42, 64.65, 162.56. IR (neat, cm−1):
2959, 2933, 1754. HRMS (EI): m/z calcd for C10H15NO3 (M+)
197.1052, found 197.1054.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Infrared (IR) spectra were
recorded on a Perkin-Elmer 1600 series FT-IR spectrometer. Mass
spectra were recorded on a JEOL JMN-DX 303/JMA-DA 5000
spectrometer. Microanalyses were performed on a Perkin-Elmer CHN
2400 elemental analyzer. Optical rotations were measured with a
JASCO DIP-360 or JASCO P-1020 digital polarimeter. Proton nuclear
magnetic resonance (1H NMR) spectra were recorded on a JEOL
JNM-EX 270 (270 MHz), Varian Gemini-300 (300 MHz), JEOL
JNM-AL 400 (400 MHz), Varian Unity-500 (500 MHz), or JNM-LA
(600 MHz) spectrometer using tetramethylsilane as the internal
standard. The following abbreviations are used: s = singlet, d =
doublet, t = triplet, q = quartet, m = multiplet, br = broad. Column
chromatography was carried out on Merck silica gel 60 (230−400
mesh) or KANTO silica gel 60N (40−50 mm) for flash
chromatography. The purity of all final compounds was established
to be ≥95% by HPLC analysis (Agilent 1220 Infinity LC system).
HPLC conditions: TSKgel NH2-100 3 μm column (4.6 mm ×15 cm)
using CH3CN/H2O = 90/10 (flow rate 1.0 mL/min) as the eluent and
refractive index detector (retention time 17.2 min for compound 1c).
Preparation of Iminosugars and Their Alkyl Derivatives. 1-
Deoxynojirimycin was isolated from the roots of Adenophora triphylla
according to the literature.59 The 1-C-alkyl-DNJ derivatives 19−21
were prepared according to the method reported previously.60,61 LAB
was prepared from xylitol according to the literature.62 The N-
butylation of LAB and DNJ was performed by treatment with the butyl
bromide and K2CO3 in dimethylformamide. The reaction mixture was
evaporated in vacuo, and the residual syrup was redissolved in MeOH
and applied to an Amberlist 15 column (H+ form), eluted with 0.5 M
NH4OH, and concentrated. The eluate was finally purified by Dowex
+
1-X2 (OH− form) and Amberlite CG-50 (NH4 form) column
chromatography with water as the eluent. Acarbose was purchased
from LKT Laboratories Inc. (St. Paul, MN). Voglibose and miglitol
were purchased from Wako Pure Chemical Industries, Ltd. (Osaka,
Japan).
Typical Procedure for α-1-C-Alkyl-LAB. Preparation of
(5R,7aR)-5-(Iodomethyl)-1,7a-dihydropyrrolo[1,2-c]oxazol-3(5H)-
one (2). Iodine chips (11.3 g, 44.5 mmol) were added to a solution of
alcohol 3 (3.45 g, 22.2 mmol), triphenylphosphine (14.6 g, 55.6
mmol), and imidazole (3.79 g, 55.6 mmol) in CH2Cl2 (89 mL) at 0
°C. After being stirred at room temperature for 6 h, the reaction was
quenched with 1.0 M aq sodium thiosulfate. The water layer was
extracted with CH2Cl2 (2 × 40 mL). The organic layers were
Typical Procedure for the Preparation of a Diol (6c). Trifluoro-
acetic acid (0.78 mL, 10.5 mmol) was added to a solution of the
epoxide 5c (171 mg, 0.867 mmol) in THF/H2O (3/2, 7.5 mL). After
being stirred at 80 °C for 72 h, the reaction mixture was cooled to
room temperature and concentrated. The product was purified by flash
chromatography (silica gel, 10 g, Et2O only → Et2O/CH3OH = 20/1)
to yield diol 6c (136 mg, 74%) as a colorless oil.
J
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(5S,6S,7S,7aS)-5-Butyl-6,7-dihydroxytetrahydropyrrolo[1,2-c]-
1.62−1.72 (m, 1H), 2.84 (dd, J = 12.5, 7.2 Hz, 1H), 2.97 (m, 1H),
oxazol-3(1H)-one (6c). [α]2D3 = +22.2 (CHCl3, c = 0.46). H NMR
1
3.55−3.60 (m, 2H), 3.64−3.68 (m, 1H), 3.74 (t. J = 6.8 Hz, 1H). 13
C
(CDCl3, 400 MHz): δ 0.90−0.95 (m, 3H), 1.25−1.69 (m, 6H), 3.65−
3.67 (m, 1H), 3.84−3.88 (m, 1H), 3.90−3.98 (m, 2H), 4.34 (dd, J =
9.7, 3.9 Hz, 1H), 4.52 (dd, J = 9.2, 7.7 Hz, 1H). 13C NMR (CHCl3,
100 MHz): δ 13.99, 22.42, 28.20, 32.81, 61.99, 65.29, 66.85, 80.93,
83.74, 162.45. IR (neat, cm−1): 3377, 1731. HRMS (EI): m/z calcd for
C10H17NO4 (M+) 215.1158, found 215.1162.
Typical Procedure for Hydrolysis (1c). NaOH (143 mg, 3.58
mmol) was added to a solution of diol 6c (71.0 mg, 0.330 mmol) in
EtOH/H2O (2:1, 4.2 mL), and the mixture was refluxed for 12 h. The
reaction mixture was cooled to room temperature and concentrated.
The product was purified by flash chromatography (silica gel, 10 g,
Et2O/MeOH/25% NH3(aq) = 80/20/1 → Et2O/MeOH/25%
NH3(aq) = 0/100/2) to yield 1c (38.0 mg, 61%) as a white
amorphous solid.
NMR (CD3OD, 100 MHz): δ 14.45, 23.73, 27.77, 30.47, 30.71, 30.89,
33.07, 35.31, 62.69, 63.28, 64.42, 79.61, 83.63. IR (neat, cm−1): 3289,
2923, 2852. HRMS (EI): m/z calcd for C14H29NO3 (M+) 259.2147,
found 259.2151.
α-1-C-Decyl-LAB (1i). [α]2D2 = −36.5 (CH3OH, c = 0.41). 1H NMR
(CD3OD, 400 MHz): δ 0.90 (t, J = 6.5 Hz, 3H), 1.24−1.51 (m, 17H),
1.65−1.75 (m, 1H), 2.86 (dd, J = 12.3, 7.5 Hz, 1H), 2.97 (dd, J = 10.6,
6.3 Hz, 1H), 3.57−3.61 (m, 2H), 3.67 (dd, J = 11.6, 4.3 Hz, 1H), 3.77
(t, J = 6.7 Hz, 1H). 13C NMR (CD3OD, 100 MHz): δ 14.45, 23.73,
27.68, 30.47, 30.67, 30.74, 30.82, 33.07, 34.81, 62.71, 62.94, 64.60,
79.18, 83.15. IR (KBr, cm−1): 3289, 2923, 2852. HRMS (EI): m/z
calcd for C15H31NO3 (M+) 273.2304, found 273.2304.
α-1-C-Undecyl-LAB (1j). [α]2D3 = −37.8 (CH3OH, c = 1.00). H
1
NMR (CD3OD, 400 MHz): δ 0.89 (t, J = 6.3 Hz, 3H), 1.29−1.50 (m,
19H), 1.63−1.75 (m, 1H), 2.86 (dd, J = 12.3, 7.5 Hz, 1H), 2.97 (dd, J
= 10.6, 6.3 Hz, 1H), 3.57−3.61 (m, 2H), 3.65−3.69 (m, 1H), 3.75 (t, J
= 6.5 Hz, 1H). 13C NMR (CD3OD, 100 MHz): δ 14.46, 23.74, 27.77,
30.48, 30.71, 30.75, 30.80, 30.89, 33.08, 35.29, 62.70, 63.25, 64.43,
79.58, 83.61. IR (KBr, cm−1): 3290, 2922. HRMS (EI): m/z calcd for
C16H33NO3 (M+) 287.2460, found 287.2461.
α-1-C-Ethyl-LAB (1a). [α]2D4 = −57.9 (CH3OH, c = 1.37). 1H NMR
(CD3OD, 400 MHz): δ 0.98 (t, J = 7.5 Hz, 3H), 1.39−1.50 (m, 1H),
1.65−1.75 (m, 1H), 2.78 (dt, J = 7.7, 5.3 Hz, 1H), 2.95 (dt, J = 6.3, 4.3
Hz, 1H), 3.56−3.61 (m, 2H), 3.64−3.68 (m, 1H), 3.75 (t, J = 6.8 Hz,
1H). 13C NMR (CD3OD, 100 MHz): δ 11.32, 27.97, 63.29, 64.19,
64.35, 79.67, 83.34. IR (KBr, cm−1): 3297, 2922. HRMS (EI): m/z
calcd for C7H15NO3 (M+) 161.1052, found 161.1056.
α-1-C-(4-Phenylbutyl)-LAB (1k). [α]2D4 = −35.7 (CH3OH, c = 0.88).
1H NMR (CD3OD, 400 MHz): δ 1.39−1.53 (m, 3H), 1.62−1.74 (m,
3H), 2.62 (t, J = 7.7 Hz, 2H), 2.84−2.89 (m, 1H), 2.96−3.00 (m, 1H),
3.56−3.60 (m, 2H), 3.65−3.68 (m, 1H), 3.75 (t, J = 6.5 Hz, 1H). 13C
NMR (CD3OD, 100 MHz): δ 27.33, 32.81, 34.93, 36.80, 62.76, 63.02,
64.53, 79.39, 83.40, 126.68, 129.28, 129.42, 143.78. IR (KBr, cm−1):
3367, 3028, 2931, 2851. HRMS (EI): m/z calcd for C15H23NO3 (M+)
265.1678, found 265.1669.
α-1-C-Propyl-LAB (1b). [α]2D4 = −46.7 (CH3OH, c = 1.18). 1H
NMR (CD3OD, 400 MHz): δ 0.96 (t, J = 7.2 Hz, 3H), 1.35−1.55 (m,
3H), 1.63−1.74 (m, 1H), 2.92−2.97 (m, 1H), 3.04 (dd, J = 6.3, 3.9
Hz, 1H), 3.60−3.65 (m, 2H), 3.68−3.72 (m, 1H), 3.79 (t, J = 6.3 Hz,
1H). 13C NMR (CD3OD, 100 MHz): δ 14.42, 20.77, 36.83, 62.54,
62.65, 64.55, 79.00, 82.97. IR (KBr, cm−1): 3223, 2929. HRMS (EI):
m/z calcd for C8H17NO3 (M+) 175.1208, found 175.1207.
α-1-C-Butyl-LAB (1c). [α]2D7 = −47.7 (CH3OH, c = 1.00). 1H NMR
(CD3OD, 400 MHz): δ 0.93 (t, J = 6.8 Hz, 3H), 1.28−1.45 (m, 5H),
1.64−1.72 (m, 1H), 2.83 (dt, J = 7.7, 5.3 Hz, 1H), 2.94 (dt, J = 6.3, 4.3
Hz, 2H), 3.63−3.67 (m, 1H), 3.74 (t, J = 6.3 Hz, 1H). 13C NMR
(CD3OD, 100 MHz): δ 14.33, 23.77, 29.83, 34.45, 62.67, 62.87, 64.56,
79.13, 83.10. IR (KBr, cm−1): 3285. HRMS (EI): m/z calcd for
C9H19NO3 (M+) 189.1365, found 189.1360.
α-1-C-(4-Methylpentyl)-LAB (1l). [α]2D4 = −44.8 (CH3OH, c =
1
0.75). H NMR (CD3OD, 400 MHz): δ 0.89 (d, J = 6.8 Hz, 6H),
1.19−1.68 (m, 7H), 2.87−2.91 (m, 1H), 2.97−3.02 (m, 1H), 3.57−
3.61 (m, 2H), 3.65−3.69 (m, 1H), 3.76 (t, J = 6.8 Hz, 1H). 13C NMR
(CD3OD, 100 MHz): δ 22.97, 22.99, 25.56, 29.12, 35.33, 40.25, 62.88,
63.00, 64.55, 79.39, 83.40. IR (KBr, cm−1): 3277, 2957, 2928, 2870.
HRMS (EI): m/z calcd for C11H23NO3 (M+) 259.2147, found
259.2151.
α-1-C-Pentyl-LAB (1d). [α]2D4 = −53.5 (CH3OH, c = 1.44). 1H
NMR (CD3OD, 400 MHz): δ 0.91 (t, J = 7.0 Hz, 3H), 1.24−1.50 (m,
7H), 1.60−1.73 (m, 1H), 2.81−2.86 (m, 1H), 2.93−2.97 (m, 1H),
3.55−3.59 (m, 2H), 3.63−3.67 (m, 1H), 3.74 (t, J = 6.5 Hz, 1H). 13C
NMR (CD3OD, 100 MHz): δ 14.39, 23.66, 27.46, 33.14, 35.32, 62.65,
63.34, 64.40, 79.64, 83.68. IR (KBr, cm−1): 3285, 2926. HRMS (EI):
m/z calcd for C10H21NO3 (M+) 203.1521, found 203.1519.
α-1-C-Hexyl-LAB (1e). [α]2D4 = −51.1 (CH3OH, c = 1.18). 1H NMR
(CD3OD, 400 MHz): δ 0.90 (t, J = 6.3 Hz, 3H), 1.25−1.48 (m, 9H),
1.62−1.72 (m, 1H), 2.86 (dd, J = 12.5, 7.2 Hz, 1H), 2.97 (dd, J = 10.9,
6.0 Hz, 1H), 3.56−3.60 (m, 2H), 3.64−3.68 (m, 1H), 3.75 (t, J = 6.5
Hz, 1H). 13C NMR (CD3OD, 100 MHz): δ 14.41, 23.67, 27.70, 30.54,
32.94, 35.26, 62.69, 63.23, 64.41, 79.57, 83.58. IR (KBr, cm−1): 3286,
2924, 2853. HRMS (EI): m/z calcd for C11H23NO3 (M+) 217.1678,
found 217.1685.
Preparation of α-1-C-{(S)-1-Hydroxybutyl}-LAB (7) and α-1-C-
{(R)-1-Hydroxybutyl}-LAB (8). (2R,5R)-Benzyl 2-{(Benzyloxy)methyl}-
5-butyryl-2,5-dihydro-1H-pyrrole-1-carboxylate (10). To a suspen-
sion of the Mg chips (208 mg, 8.54 mmol) in THF (8.5 mL) was
added 1-bromopropane (0.66 mL, 7.32 mmol) dropwise over 5 min at
room temperature, and the reaction mixture was refluxed for 1 h. The
resulting green-black solution was cooled to room temperature. To the
reaction mixture was added a solution of Weinreb amide 9 (1.00 g,
2.44 mmol) in THF (5 mL), and the reaction was stirred for 3 h. The
reaction mixture was poured into a two-layer mixture of Et2O (15 mL)
and saturated aqueous NH4Cl (15 mL) at 0 °C. The aqueous layer was
separated and extracted with EtOAc (30 mL). The combined organic
layers were washed with saturated aqueous NaHCO3 (15 mL) and
brine (15 mL), dried over Na2SO4, and concentrated to yield a yellow
oil. The residue was purified by flash chromatography (silica gel, 40 g,
hexane/EtOAc = 4/1) to yield 10 (720 mg, 75%) as a pale yellow oil.
α-1-C-Heptyl-LAB (1f). [α]2D4 = −46.1 (CH3OH, c = 1.00). 1H NMR
(CD3OD, 400 MHz): δ 0.90 (t, J = 6.8 Hz, 3H), 1.20−1.50 (m, 11H),
1.60−1.75 (m, 1H), 2.84 (dd, J = 12.6, 7.7 Hz, 1H), 2.96 (dd, J = 10.6,
6.3 Hz, 1H), 3.55−3.59 (m, 2H), 3.64−3.68 (m, 1H), 3.74 (t, J = 6.5
Hz, 1H). 13C NMR (CD3OD, 100 MHz): δ 14.43, 23.73, 27.77, 30.39,
30.86, 33.01, 35.32, 62.67, 63.29, 64.41, 79.61, 83.64. IR (KBr, cm−1):
3286, 2923. HRMS (EI): m/z calcd for C12H25NO3 (M+) 231.1834,
found 231.1824.
[α]2D3 = +349.6 (CHCl3, c = 1.96). H NMR (CDCl3, 400 MHz):
1
(major) δ 0.75 (t, J = 7.2 Hz, 3H), 1.33−1.66 (m, 2H), 2.11 (dt, J =
17.4, 7.7 Hz, 1H), 2.23 (dt, J = 17.4, 7.7 Hz, 1H), 3.80 (dd, J = 9.2, 5.8
Hz, 1H), 3.91 (dd, J = 9.7, 2.9 Hz, 1H), 4.52 (d, J = 2.9 Hz, 1H),
4.88−4.91 (m, 1H), 5.10−5.14 (m, 3H), 5.66 (dt, J = 6.3, 1.9 Hz, 1H),
6.06 (dt, J = 6.3, 1.9 Hz, 1H), 7.23−7.38 (m, 10H); (minor) δ 0.89 (t,
J = 7.2 Hz, 3H), 1.33−1.66 (m, 2H), 2.34 (dt, J = 17.4, 7.2 Hz, 1H),
2.46 (dt, J = 17.4, 7.2 Hz, 1H), 3.57 (dd, J = 9.2, 5.8 Hz, 1H), 3.74 (dd,
J = 9.7, 2.9 Hz, 1H), 4.42 (d, J = 6.3 Hz, 1H), 4.78−4.82 (m, 2H),
4.99−5.02 (m, 3H), 5.72 (dt, J = 6.3, 1.9 Hz, 1H), 6.02 (dt, J = 6.3, 1.9
Hz, 1H), 7.23−7.38 (m, 10H). 13C NMR (CDCl3, 100 MHz): δ 13.55,
13.69, 16.48, 16.65, 38.94, 40.15, 64.56, 65.48, 67.18, 67.34, 69.12,
70.23, 73.24, 73.38, 74.29, 74.37, 124.87, 124.96, 126.96, 127.44,
127.46, 127.59, 127.64, 127.69, 128.12, 128.15, 128.20, 128.27, 128.34,
128.39, 128.47, 128.51, 128.55, 132.17, 132.22, 135.94, 136.22, 138.03,
138.20, 153.62, 206.44, 207.07. IR (neat, cm−1): 2962, 2874, 1714.
α-1-C-Octyl-LAB (1g). [α]2D3 = −44.2 (CH3OH, c = 0.70). 1H NMR
(CD3OD, 400 MHz): δ 0.89 (t, J = 7.0 Hz, 3H), 1.30−1.47 (m, 13H),
1.67−1.73 (m, 1H), 2.89−2.94 (m, 1H), 3.01−3.06 (m, 1H), 3.59−
3.63 (m, 2H), 3.69 (m, 1H), 3.77 (t, J = 6.8 Hz, 1H). 13C NMR
(CD3OD, 100 MHz): δ 14.43, 23.73, 27.68, 30.41, 30.64, 30.82, 33.05,
34.79, 62.69, 62.98, 64.65, 79.15, 83.12. IR (KBr, cm−1): 3287, 2924,
2852. HRMS (EI): m/z calcd for C13H27NO3 (M+) 245.1991, found
245.1982.
α-1-C-nonyl-LAB (1h). [α]2D3 = −45.5 (CH3OH, c = 1.33). 1H NMR
(CD3OD, 400 MHz): δ 0.89 (t, J = 6.8 Hz, 3H), 1.22−1.50 (m, 15H),
K
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
HRMS (EI): m/z calcd for C11H23NO4 (M+) 233.1627, found
233.1638.
20/1 →Et2O/MeOH/25% NH3(aq) = 0/100/2) to yield 7 (32.0 mg,
68%) as a white amorphous solid. [α]2D2 = −17.1 (CH3OH, c = 1.32).
1H NMR (CD3OD, 400 MHz): δ 0.97 (t, J = 7.2 Hz, 3H), 1.36−1.64
(m, 4H), 3.37−3.40 (m, 2H), 3.77−3.90 (m, 3H), 3.97 (t, J = 6.8 Hz,
1H), 4.23 (t, J = 6.8 Hz, 1H). 13C NMR (CD3OD, 100 MHz): δ 14.24,
19.96, 36.67, 58.62, 64.77, 66.01, 69.72, 75.09, 76.64. IR (neat, cm−1):
3350, 2962. HRMS (FAB): m/z calcd for C9H20NO4 (M + H+)
206.1392, found 206.1394.
(2R,5R)-Benzyl 2-((Benzyloxy)methyl)-5-{(S)-1-hydroxybutyl}-2,5-
dihydro-1H-pyrrole-1-carboxylate (11). To a solution of ketone 10
(264 mg, 0.668 mmol) in Et2O (4.5 mL) at 0 °C was added DIBAL-H
(0.89 mL, 1.5 M in hexane, 1.34 mmol) dropwise over 5 min. After
being stirred at this temperature for 2 h, the mixture was quenched by
addition of EtOAc (5 mL) and 1 M aqueous potassium sodium
tartrate (5 mL). The resulting mixture was stirred vigorously at room
temperature for 2 h and then was extracted with EtOAc (2 × 15 mL).
The combined organic layers were washed with brine (10 mL) and
dried over anhydrous Na2SO4. Filtration and evaporation in vacuo
furnished the crude product, which was purified by column
chromatography (silica gel, 15 g, hexane/EtOAc = 3/1 → hexane/
EtOAc = 2/1) to give 11 (172 mg, 65%) as a colorless oil and 12 (47.8
(2R,5R)-Benzyl 2-{(Benzyloxy)methyl}-5-{(R)-1-hydroxybutyl}-2,5-
dihydro-1H-pyrrole-1-carboxylate (12). To a solution of ketone 10
(360 mg, 0.915 mmol) in MeOH (6 mL) at 0 °C was added NaBH4
(69.2 mg, 1.83 mmol). After being stirred at this temperature for 1 h,
the mixture was quenched by addition of H2O (5 mL). The resulting
mixture was extracted with EtOAc (2 × 25 mL). The combined
organic layers were washed with brine (15 mL) and dried over
anhydrous Na2SO4. Filtration and evaporation in vacuo furnished the
crude product, which was purified by column chromatography (silica
gel, 15 g, hexane/EtOAc = 3/1 → hexane/EtOAc = 2/1) to give 11
(87.5 mg, 24%) as a colorless oil and 12 (183 mg, 51%) as a colorless
oil. [α]2D2 = +234.6.0 (CHCl3, c = 0.80). 1H NMR (CDCl3, 400 MHz):
(major) δ 0.91 (t, J = 6.8 Hz, 3H), 1.10−1.60 (m, 4H), 3.54 (dd, J =
9.2, 6.3 Hz, 1H), 3.68 (dd, J = 9.2, 2.4 Hz, 1H), 3.94 (br s, 1H), 4.34−
4.43 (m, 3H), 4.51−4.70 (m, 2H), 5.07 (d, J = 6.8 Hz, 2H), 5.82−5.88
(m, 2H), 7.22−7.36 (m, 10H); (minor) δ 0.81 (t, J = 6.8 Hz, 3H),
1.10−1.60 (m, 4H), 3.81−3.91 (m, 2H), 4.16 (br s, 1H), 4.51−4.70
(m, 5H), 5.03−5.22 (m, 2H), 5.82−5.96 (m, 2H), 7.22−7.36 (m,
10H). 13C NMR (CDCl3, 100 MHz): δ 14.11, 18.51, 19.08, 35.13,
64.83, 65.77, 66.88, 67.38, 70.30, 72.07, 73.16, 73.27, 74.11, 126.76,
127.11, 127.39, 127.43, 127.65, 128.22, 128.23, 128.30, 128.39, 128.54,
129.05, 136.08, 138.13, 155.89, 155.90. IR (neat, cm−1): 3435, 2933,
2871, 1682. HRMS (EI): m/z calcd for C24H29NO4 (M+) 395.2097,
found 395.2097.
1
mg, 18%) as a colorless oil. [α]2D3 = +227.2 (CHCl3, c = 0.98). H
NMR (CDCl3, 400 MHz): (major) δ 0.90 (t, J = 7.2 Hz, 3H), 1.20−
1.34 (m, 4H), 3.56 (dd, J = 9.2, 5.8 Hz, 1H), 3.71 (dd, J = 9.2, 2.9 Hz,
1H), 3.84−3.88 (m, 1H), 4.09 (d, J = 9.7 Hz, 1H), 4.34−4.49 (m,
2H), 4.61−4.64 (m, 1H), 4.86−4.87 (m, 1H), 5.08 (d, J = 2.4 Hz,
2H), 5.70 (dt, J = 6.8, 1.9 Hz, 1H), 5.89 (dt, J = 6.8, 1.9 Hz, 1H),
7.22−7.35 (m, 10H); (minor) δ 0.88 (t, J = 7.2 Hz, 3H), 1.20−1.34
(m, 4H), 3.52−3.58 (m, 1H), 3.67−3.721 (m, 1H), 3.84−3.88 (m,
1H), 4.09 (m, 1H), 4.34−4.49 (m, 2H), 4.61−4.64 (m, 1H), 4.86−
4.87 (m, 1H), 5.03−5.09 (m, 2H), 5.70−5.72 (m, 1H), 5.88−6.01 (m,
1H), 7.22−7.35 (m, 10H). 13C NMR (CDCl3, 100 MHz): δ 14.11,
19.46, 33.50, 65.35, 67.38, 70.27, 72.09, 72.33, 73.16, 127.42, 127.52,
127.65, 128.09, 128.16, 128.21, 128.30, 128.38, 128.54, 129.55, 136.12,
138.12, 155.73. IR (neat, cm−1): 3430, 2957, 1683. HRMS (EI): m/z
calcd for C24H29NO4 (M+) 395.2097, found 395.2097.
(2S,3S,4S,5S)-Benzyl-2-{(benzyloxy)methyl}-3,4-dihydroxy-5-{(S)-
1-hydroxybutyl}pyrrolidine-1-carboxylate (14). To a vial containing
11 (122 mg, 0.308 mmol) were added CH3CN (3.1 mL) and 4 × 10−4
M ethylenediaminetetraacetic acid in H2O (2.1 mL). The solution was
cooled to 0 °C, and 1,1,1-trifluoroacetone (345 mg, 3.08 mmol) was
added. A mixture of solid Oxone (947 mg, 1.54 mmol) and NaHCO3
(245 mg, 2.31 mmol) was added in four portions over 45 min. The
reaction was stirred for 2 h at 0 °C, then diluted with 5 mL of H2O,
and extracted with EtOAc (2 × 15 mL). The organic layers were
combined and dried over anhydrous Na2SO4. Filtration and
evaporation in vacuo furnished the crude product, which was purified
by column chromatography (silica gel, 6 g, Et2O) to provide a mixture
of epoxides 13 (105 mg, 83%) as a colorless oil. Trifluoroacetic acid
(0.25 mL, 3.37 mmol) was added to a solution of the epoxides 13 (102
mg, 0.248 mmol) in THF/H2O (3/2, 2.5 mL). After being stirred at
70 °C for 12 h, the reaction mixture was cooled to room temperature,
quenched by addition of saturated aqueous NaHCO3 (5 mL), and
extracted with EtOAc (3 × 5 mL). The combined organic layers were
washed with brine (5 mL) and dried over anhydrous Na2SO4.
Filtration and evaporation in vacuo furnished the crude product, which
was purified by column chromatography (silica gel, 5 g, hexane/EtOAc
= 2/1 → hexane/EtOAc = 1/1) to yield diol 14 (100 mg, 93%) as a
colorless oil. [α]2D2 = −+57.8 (CHCl3, c = 1.21). 1H NMR (CDCl3, 400
MHz, ca.1:1 mixture): δ 0.86 (t, J = 6.8 Hz, 3H), 0.97 (t, J = 6.8 Hz,
3H), 1.25−1.61 (m, 8H), 3.50−3.58 (m, 2H), 3.84−4.15 (m, 10H),
4.28−4.57 (m, 6H), 4.76−5.17 (m, 6H), 7.20−7.35 (m, 10H). 13C
NMR (CDCl3, 100 MHz): δ 14.03, 14.30, 19.46, 19.62, 36.61, 36.69,
67.32, 67.38, 67.76, 67.81, 68.48, 69.26, 70.84, 73.08, 73.65, 73.88,
74.09, 76.03, 77.64, 79.73, 80.69, 128.15, 128.22, 128.26, 128.52,
128.54, 128.55, 128.64, 128.86, 128.92, 128.99, 136.37, 136.45, 136.68,
136. 82, 154.80, 154.83. IR (neat, cm−1): 3390, 2958, 2927, 1683.
HRMS (EI): m/z calcd for C24H31NO6 (M+) 429.2151, found
429.2146.
(2S,3S,4S,5S)-Benzyl 2-{(Benzyloxy)methyl}-3,4-dihydroxy-5-{(R)-
1-hydroxybutyl}pyrrolidine-1-carboxylate (16). To a vial with 12
(155 mg, 0.392 mmol) were added CH3CN (3.9 mL) and 4 × 10−4
M
ethylenediaminetetraacetic acid in H2O (2.6 mL). The solution was
cooled to 0 °C, and 1,1,1-trifluoroacetone (439 mg, 3.08 mmol) was
added. A mixture of solid Oxone (1.21 g, 1.96 mmol) and NaHCO3
(312 mg, 2.94 mmol) was added in four portions over 45 min. The
reaction was stirred for 2 h at 0 °C, then diluted with 5 mL of H2O,
and extracted with EtOAc (2 × 15 mL). The organic layers were
combined and dried over anhydrous Na2SO4. Filtration and
evaporation in vacuo furnished the crude product, which was purified
by column chromatography (silica gel, 6 g, Et2O) to provide a mixture
of epoxides 15 (129 mg, 80%) as a colorless oil. Trifluoroacetic acid
(0.30 mL, 4.04 mmol) was added to a solution of the epoxides 13 (125
mg, 0.304 mmol) in THF/H2O (3/2, 3.0 mL). After being stirred at
70 °C for 12 h, the reaction mixture was cooled to room temperature,
quenched by addition of saturated aqueous NaHCO3 (5 mL), and
extracted with EtOAc (3 × 5 mL). The combined organic layers were
washed with brine (5 mL) and dried over anhydrous Na2SO4.
Filtration and evaporation in vacuo furnished the crude product, which
was purified by column chromatography (silica gel, 5 g, hexane/EtOAc
= 2/1 → hexane/EtOAc = 1/1) to yield diol 14 (21.0 mg, 16%) as a
colorless oil. [α]2D0 = +77.3 (CHCl3, c = 0.19). H NMR (CDCl3, 400
1
MHz): (major) δ 0.94 (t, J = 6.8 Hz, 3H), 1.33−1.88 (m, 4H), 3.45−
3.57 (m, 1H), 3.67−4.09 (m, 5H), 4.29−4.37 (m, 1H), 4.45−4.59 (m,
2H), 5.02−5.17 (m, 2H), 7.18−7.35 (m, 10H); (minor) δ 0.78 (t, J =
6.8 Hz, 3H), 1.33−1.88 (m, 4H), 3.45−3.57 (m, 1H), 3.67−4.09 (m,
5H), 4.29−4.37 (m, 1H), 4.45−4.59 (m, 2H), 5.02−5.17 (m, 2H),
7.18−7.35 (m, 10H). 13C NMR (CDCl3, 100 MHz): δ 13.78, 13.99,
19.08, 19.30, 29.70, 38.19, 67.11, 67.40, 67.59, 68.61, 68.77, 71.18,
72.53, 73.50, 73.74, 74.49, 74.97, 79.62, 80.55, 81.07, 81.68, 127.95,
128.26, 128.35, 128.54, 128.60, 135.99, 136.38, 136.67, 155.16, 156.61.
IR (neat, cm−1): 3390, 2958, 2927, 1674. HRMS (EI): m/z calcd for
C24H31NO6 (M+) 429.2151, found 429.2150.
α-1-C-{(S)-1-Hydroxybutyl}-LAB (7). Pd/C (5% Pd, 9.8 mg, 10 wt
%) was added to a stirred solution of diol 14 (98.4 mg, 0.229 mmol) in
MeOH (2.3 mL) and concentrated HCl (3 drops). The reaction
mixture was stirred for 10 h under a balloon of H2. The reaction
mixture was filtered through Celite (eluting with MeOH) and
concentrated in vacuo. The product was purified by flash
chromatography (silica gel, 3 g, Et2O/MeOH/25% NH3(aq) = 80/
α-1-C-{(R)-1-Hydroxybutyl}-LAB (8). Pd/C (5% Pd, 2.1 mg, 10 wt
%) was added to a stirred solution of diol 16 (20.6 mg, 0.0480 mmol)
in MeOH (1.0 mL) and concentrated HCl (1 drop). The reaction
mixture was stirred for 10 h under a balloon of H2. The reaction
L
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
mixture was filtered through Celite (eluting with MeOH) and
concentrated in vacuo. The product was purified by flash
chromatography (silica gel, 2 g, Et2O/MeOH/25% NH3(aq) = 80/
20/1 → Et2O/MeOH/25% NH3(aq) = 0/100/2) to yield 8 (9.30 mg,
94%) as a white amorphous solid. [α]2D2 = −11.6 (CH3OH, c = 0.10).
1H NMR (CD3OD, 400 MHz): δ 0.97 (t, J = 7.2 Hz, 3H), 1.39−1.57
(m, 4H), 3.20−3.23 (m, 1H), 3.34−3.38 (m, 1H), 3.75−3.88 (m, 3H),
3.97−4.02 (m, 2H), 4.88 (br s, 1H). 13C NMR (CD3OD, 100 MHz):
δ 14.16, 19.64, 37.61, 58.98, 64.92, 67.73, 68.68, 76.16, 76.43. IR (neat,
cm−1): 3367, 2931. HRMS (FAB): m/z calcd for C9H20NO4 (M + H+)
206.1392, found 206.1384.
reaction mixture was added to 3 mL of the glucose CII-test reagent
(Wako Pure Chemical Industries, Osaka, Japan). The absorbance at
505 nm was measured to determine the amount of the released D-
glucose. Kinetic parameters were determined by the double-reciprocal-
plot method of Lineweaver−Burk at increasing concentrations of the
appropriate maltose.
Glycoblotting. The human hepatocelluar carcinoma HepG2 cell
line was obtained from the European Collection of Cell Cultures
(ECACC 85011430). Cells were cultured in low-glucose Dulbecco’s
modified Eagle's medium (Invitrogen, Carlsbad, CA), supplemented
with 10% fetal calf serum (Invitrogen), and maintained in a humidified
incubator at 37 °C with 5% CO2. N-Linked oligosaccharide analysis
was performed on HepG2 cells after treatment for 48 h with miglitol
or α-1-C-butyl-LAB, which gave additional N-linked oligosaccharide in
the MALDI-TOF MS profiles derived from HepG2 homogenates
(Figure 3b,c) in comparison to untreated controls (Figure 3a). Cell
pellets were homogenized in 0.5% Triton X-100 solution. The
solubilized proteinaceous materials were reduced by dithiothreitol
(DTT) at 60 °C for 30 min followed by alkylation with 10 μL of 123
mM iodoacetamide (IAA) by incubation in the dark at 25 °C for 60
min. The mixture was then treated with 10 μL of 40 units/μL trypsin
at 37 °C for 60 min, followed by heat inactivation of the enzyme at 90
°C for 10 min. After being cooled to room temperature, N-glycans
were enzymatically released from trypsin-digested glycopeptides by
incubation with 5 units of PNGase F (Roche Applied Science,
Switzerland) at 37 °C for 16 h. PNGase F-treated samples were
dropped onto BlotGlyco beads (Sumitomo Bakelite Co., Tokyo,
Japan) in a filter plate, and this was followed by addition of 2% acetic
acid/acetonitrile. After incubation at 80 °C for 60 min to covalently
ligate glycans onto the beads, the beads were washed with 2 M
guanidine hydrochloride, distilled water, and 1% triethylamine/
methanol to remove nonspecifically bound impurities. After washing,
the beads were treated with 10% acetic anhydride/methanol for 30
min at ambient temperature to quench the hydrazide groups. The
beads were then washed serially with 10 mM hydrochloric acid,
distilled water, methanol, and dimethyl sulfoxide (DMSO). On-bead
methyl esterification was performed by the reaction with 500 mM 1-
methyl-3-p-tolyltriazene/DMSO solution at 60 °C for 60 min. After
the beads were washed with methanol and distilled water, the trapped
glycans were finally released and recovered as Nα-((aminooxy)acetyl)-
tryptophanylarginine methyl ester (aoWR) derivatives by adding 20
mM aoWR and 2% acetic acid/acetonitrile and incubated at 80 °C for
1 h. The resulting aoWR-labeled glycans were recovered by washing
the beads with distilled water, and the collected solution was further
purified with a solid-phase extraction column (Cleanup column,
Sumitomo Bakelite Co.) to remove the excess aoWR. The purified
solution was mixed with 2,5-dihydroxybenzoic acid solution and
subsequently subjected to MALDI-TOF analysis.
N-Butyl-α-1-C-butyl-LAB (17). Sodium borocyanohydride (9.9 mg,
0.15 mmol) and butyraldehyde (30 μL, 0.32 mmol) were added to the
solution of (2S,3S,4S,5S)-2-butyl-5-(hydroxymethyl)pyrrolidine-3,4-
diol (19 mg, 0.10 mmol) in MeOH (2.0 mL) at 0 °C, and the
resulting solution was stirred for 2 h at the same temperature. Then
the solvent was removed under reduced pressure. The residue was
purified by silica gel SiO2 column chromatography to obtain 17 (11.2
1
mg, 46%). [α]2D0 = +36.9 (CHCl3, c = 1.01). H NMR (400 MHz,
CDCl3): δ 0.87−0.94 (m, 6H), 1.25−1.50 (m, 9H), 1.58−1.67 (m,
1H), 2.57−2.72 (m, 2H), 2.85 (br s, 1H), 3.07 (br d, J = 10.1 Hz, 1H),
3.72−3.81 (m, 2H), 3.86 (br s, 1H), 4.11 (br d, J = 3.4 Hz, 1H), 4.20
(br s, 3H). 13C NMR (100 MHz, CDCl3): δ 13.92, 14.03, 20.58, 22.85,
25.24, 28.36, 29.50, 46.56, 58.90, 67.74, 69.45, 79.97, 80.42. IR (neat)
cm−1 3391, 2959, 2873, 2344, 2173. HRMS (EI): m/z calcd for
C13H27NO3 (M+) 245.1991, found 245.1980.
(5S,6R,7S,7aR)-5,7-Dibutyltetrahydro-6-hydroxypyrrolo[1,2-c]-
oxazol-3(1H)-one (19). CuI (84.6 mg, 0.44 mmol) was dried under
reduced pressure, and Et2O (0.7 mL) was added under an argon
atmosphere. The solution was cooled at 0 °C, and n-BuLi (1.6 M
hexane solution, 0.55 mL, 0.50 mmol) was added. The solution was
stirred until the yellow color disappeared. The resulting solution was
utilized as the ether solution of n-Bu2CuLi. The prepared ether
solution of n-Bu2CuLi was added to the solution of 5c (43.2 mg, 0.22
mmol) in Et2O (0.7 mL) at −20 °C under an argon atmosphere. The
solution was stirred until the starting material disappeared on TLC.
Then a saturated aqueous solution of NH4Cl was added to quench the
reaction. The resulting mixture was extracted with CH2Cl2 and dried
over Na2SO4. The solvent was removed under reduced pressure, and
the residue was purified by silica gel column chromatography (Et2O)
to afford 19 (13 mg, 23%). [α]2D4 = +11.5 (CHCl3, c = 1.30). 1H NMR
(400 MHz, CDCl3): δ 0.88−0.94 (m, 6H), 1.20−1.55 (m, 10H),
1.60−1.69 (m, 2H), 1.83−1.88 (m, 1H), 3.57−3.74 (m, 2H), 3.72 (dd,
J = 7.7, 5.3 Hz, 1H), 4.19 (dd, J = 9.2, 3.9 Hz, 1H), 4.47 (t, J = 8.7 Hz,
1H). 13C NMR (100 MHz, CDCl3): δ 13.85, 13.98, 22.54, 22.94,
28.24, 30,10, 30.50, 33.82, 52.85, 62.13, 66.27, 67.47, 84.92, 161.30. IR
(KBr, cm−1): 3426, 1733, 1398, 1227. HRMS (EI): m/z calcd for
C14H25NO3 (M+) 255.1843, found 255.1832.
(2S,3R,4S,5R)-2,4-Dibutyl-5-(hydroxymethyl)pyrrolidin-3-ol (18).
To the solution of 19 (13 mg, 0.051 mmol) in EtOH (0.4 mL)
were added H2O (0.2 mL) and NaOH (20 mg, 0.5 mmol) at room
temperature. The solution was refluxed at 100 °C for 1 h. Then the
solvent was removed under reduced pressure. The residue was purified
by silica gel SiO2 column chromatography (MeOH) to obtain 18 (4.4
MALDI-TOF MS Analysis. All measurements were performed with
an Autoflex III TOF/TOF mass spectrometer equipped with a
reflector and controlled by the FlexControl 3.0 software package
(Bruker Daltonics GmbH, Bremen, Germany), according to a general
protocol reported previously.51 The peaks were detected generally as a
formula of [M + H]+ ions. In MALDI-TOF MS reflector mode, ions
generated by a Smartbeam (pulsed UV nitrogen laser, λ = 337 nm, 5
Hz) were accelerated to a kinetic energy of 23.5 kV. Masses were
automatically annotated by using FlexAnalysis 2.0. External calibration
of MALDI mass spectra was carried out using singly charged
monoisotopic peaks of a mixture of human angiotensin II (m/z
1046.542), bombesin (m/z 1619.823), adrenocorticotropin (18−39)
(m/z 2465.199), and somatostatin 28 (m/z 3147.472). All measure-
ments were performed as follows: 1 mL of the sample solution was
mixed with 1 mL of 2,5-dihydroxybenzoic acid (10 mg/mL in 30%
ACN), and 1 mL of the resulting mixture was subjected to MALDI-
TOF mass analysis. Structural identification of glycans was performed
by MS analysis and the use of a database for glycan structures64
(http://web.expasy.org/glycomod/).
mg, 38%). [α]2D0 = −15.6 (CH3OH, c = 0.44). H NMR (400 MHz,
1
CD3OD): δ 0.89−0.96 (m, 6H), 1.25−1.80 (m, 13H), 2.90−3.05 (m,
2H), 3.43 (t, J = 8.2 Hz, 1H), 3.55 (dd, J = 11.6, 7.2 Hz, 1H), 3.64 (dd,
J = 11.6, 3.7 Hz, 1H). 13C NMR (100 MHz, CD3OD): δ 14.28, 14.32,
23.81, 24.01, 29.96, 30.79, 30.89, 33.10, 33.28, 35.17, 63.87, 64.04,
64.13, 82.38. IR (KBr, cm−1): 3375, 1561, 1467. HRMS (EI): m/z
calcd for C14H25NO3 (M+) 229.2042, found 229.2049.
Assay of Enzyme Inhibition. Brush border membranes were
prepared from the rat small intestine according to the method of
Kessler et al.63 and were assayed at pH 5.8 for rat intestinal maltase,
isomaltase, sucrase, cellobiase, and lactase using the appropriate
disaccharides as substrates. For rat intestinal maltase activities, the
reaction mixture contained 25 mM maltose and the appropriate
amount of enzyme, and the incubations were performed for 10−30
min at 37 °C. The reaction was stopped by heating at 100 °C for 3
min. After centrifugation (600g, 10 min), 0.05 mL of the resulting
Disaccharide Loading Test. The animal experimental protocols
in this study were approved by the Animal Experiments Committee of
the University of Toyama (S-2010 UH-2). Male ddY mice (29−33 g)
M
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
after an overnight fast were used for acute disaccharide loading tests.
Maltose (2.5 g/kg of body weight) or sucrose (2.5 g/kg of body
weight) as well as the inhibitor, miglitol, was dissolved in 0.9% NaCl
solution and administered to the mice via a stomach tube. A control
group was loaded with saline only. Blood samples for glucose
measurements were obtained from the tail vein at 0, 15, 30, 60, and
120 min after disaccharide loading. The blood glucose levels were
measured by a portable kit, StatStrip Xpress (Nova Biomedical Co.
Ltd., Waltham, MA). The AUC over a period of 60 min after
disaccharide loading was calculated by the trapezoidal method.
Homology Modeling of the Rat N-Terminal Catalytic
Domain of ntMGAM. The α-glucosidase activities are associated
with two small intestinal membrane-bound enzymes: maltase-
glucoamylase and sucrase-isomaltase. The N-terminal catalytic domain
of maltase-glucoamylase (ntMGAM) has a preference for short linear
α-1,4 oligosaccharides such as maltose.65 To investigate the inhibitory
activity of α-1-C-butyl-LAB on the hydrolysis of maltose in rats, we
first constructed a three-dimensional model of rat ntMGAM using
ASSOCIATED CONTENT
■
S
* Supporting Information
General experimental protocols and H and 13C NMR spectra
of the compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
1
AUTHOR INFORMATION
Corresponding Author
*Phone: +81-76-434-7868 (A.K.); +81-22-727-0144 (H.T.). E-
mail: kato@med.u-toyama.ac.jp (A.K.); takahata@tohoku-
pharm.ac.jp (H.T.).
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. Naoki Asano (BioApply Co., Ltd.) for valuable
discussions on the α-1-C-alkyl-DNJ derivatives and Takahito
Kunimatsu and Shinpei Nakagawa (University of Toyama) for
their experimental contribution. This work was supported by a
Grant-in-Aid for Scientific Research (C) (23590127) (A.K.)
from the Japanese Society for the Promotion of Science (JSPS).
homology modeling with Prime 1.5 (Schrodinger, Portland, OR),
̈
because the three-dimensional structure of rat ntMGAM has not been
determined experimentally. The amino acid sequence of rat ntMGAM
was taken from the UniProtKB-Swiss-Prot database (accession number
Q6P7A9). Sequences homologous to the rat ntMGAM against the
Protein Data Bank (PDB) were performed using a BLAST search, and
the three-dimensional structures of human ntMGAM were sub-
sequently found (sequence identity between rat ntMGAM and human
ntMGAM, 45%). Four representative structures of human ntMGAM
were obtained by clustering 12 three-dimensional structures of human
ntMGAM taken from the PDB on the basis of side chain structures
around the ligand (PDB IDs 2QLY, 2QMJ, 3CTT, and 3L4W). Using
four template structures of human ntMGAM, we constructed the
homology models of rat ntMGAM. The stereochemical quality
evaluation using PROCHECK indicated that the four homology
models are reasonable.66
ABBREVIATIONS USED
■
LAB, 1,4-dideoxy-1,4-imino-L-arabinitol; DAB, 1,4-dideoxy-1,4-
imino-D-arabinitol; DNJ, 1-deoxynojirimycin; DMDP, (2R,5R)-
bis(dihydroxymethyl)-(3R,4R)-dihydroxypyrrolidine; IC50, con-
centration of compound required to induce 50% inhibition; ER,
endoplasmic reticulum; MALDI-TOF MS, matrix-assisted laser
desorption ionization time-of-flight mass spectrometry; AUC,
area under the curve; ntMGAM, N-terminal catalytic domain of
maltase-glucoamylase; AAA, asymmetric allylic alkylation;
RCM, ring-closing metathesis
Docking Studies of α-1-C-Butyl-LAB, α-1-C-Butyl-DNJ, and
Miglitol with Rat ntMGAM. Docking calculations of α-1-C-butyl-
LAB, α-1-C-butyl-DNJ, and miglitol with the homology models of rat
ntMGAM were carried out using the standard precision mode for
REFERENCES
■
(1) Tuomilehto, J.; Lindstrom, J.; Eriksson, J. G.; Valle, T. T.;
̈
Hamalainen, H.; Ilanne-Parikka, P.; Keinanen-Kiukaanniemi, S.;
̈
̈
̈
̈
Glide 5.5 in the Schrodinger suite. The top three poses ranked with
̈
Laakso, M.; Louheranta, A.; Rastas, M.; Salminen, V.; Uusitupa, M.
(for the Finnish Diabetes Prevention Study Group). Prevention of
type 2 diabetes mellitus by changes in lifestyle among subjects with
impaired glucose tolerance. N. Engl. J. Med. 2001, 344, 1343−1350.
(2) Lyssenko, V; Jonsson, A.; Almgren, P.; Pulizzi, N.; Isomaa, B.;
Tuomi, T.; Berglung, G.; Alshuler, D.; Nilsson, P.; Groop, L. Clinical
risk factors, DNA variants, and the development of type 2 diabetes. N.
Engl. J. Med. 2008, 359, 2220−2232.
GlideScore67 were minimized by an OPLS2005 force field, and the
docking model that exhibited the most stable interaction energy was
then selected as the best model. The binding free energies between rat
ntMGAM and the compounds were calculated accurately by
combining an OPLS2005 force field with the GB/SA continuum
solvation model to include the effect of solvent.68 The binding energy
was estimated as follows: ΔE = Ecomplex − (Eprotein − Eligand). Here,
Ecomplex, Eprotein, and Eligand are the molecular mechanics energies of the
complex structure, the protein structure, and the ligand structure,
respectively.
(3) McCarthy, M. I. Genomics, type 2 diabetes, and obesity. N. Engl.
J. Med. 2010, 363, 2339−2350.
(4) Shaw, J. E.; Sicree, R. A.; Zimmet, P. Z. Global estimates of the
prevalence of diabetes for 2010 and 2030. Diabetes Res. Clin. Pract.
2010, 87, 4−14.
Docking Studies of α-1-C-Butyl-LAB and α-1-C-Butyl-DNJ
with Human β-Glucocerebrosidase. The docking calculations of α-
1-C-butyl-LAB and α-1-C-butyl-DNJ with β-glucocerebrosidase were
carried out using the X-ray of human β-glucocerebrosidase (PDB ID
(5) Gerich, J. E. Clinical significance, pathogenesis, and management
of postprandial hyperglycemia. Arch. Intern. Med. 2003, 163, 1306−
1316.
(6) UK Prospective Diabetes Study Group.. Intensive blood-glucose
control with sulphonylureas or insulin compared with conventional
treatment and risk of complications in patients with type 2 diabetes
(UKPDS 33). Lancet 1998, 352, 837−853.
(7) UK Prospective Diabetes Study Group.. Effect of intensive blood-
glucose control with metformin on complications in overweight
patients with type 2 diabetes (UKPDS 34). Lancet 1998, 352, 854−
865.
(8) Stratton, I. M.; Adler, A. I.; Neil, H. A. W.; Matthws, D. R.;
Manley, S. E.; Cull, C. A.; Hadden, D.; Turner, R. C.; Holman, R. R.
Association of glycaemia with macrovascular and microvascular
complications of type 2 diabetes (UKPDS 35): prospective
observational study. BMJ [Br. Med. J.] 2000, 321, 405−412.
2NSX) and Induced Fit Docking69 in the Schrodinger package. For
̈
the initial docking process of the Induced Fit Docking, van der Waals
(vdW) radii of both the ligand and the receptor were scaled to 0.5 to
soften the surfaces of the protein and the ligand, and Asn396 was
temporarily mutated to alanine because multiple X-ray structures
showed that the orientation of the side chain of Asn396 changes. The
ligands were then redocked back into the new receptor using the
default vdW radii scaling (1.0 receptor, 0.8 ligand), and Asn396
mutated to alanine was restored in the final docking. The standard
precision mode was used in the docking analysis. IFDScore was the
sum of the docking energy, receptor strain, and solvation terms.
Finally, the top pose ranked by IFDScore was minimized by an
OPLS2005 force field.
N
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(9) Hanefeld, M.; Cagatay, M.; Petrowitsch, T.; Neuse, D.; Petzinna,
D.; Rupp, M. Acarbose reduces the risk for myocardial infarction in
type 2 diabetic patients: meta-analysis of seven long-term studies. Eur.
Heart J. 2004, 25, 10−16.
(10) Coutinho, M.; Gerstein, H. C; Wang, Y.; Yusuf, S. The
relationship between glucose and incident cardiovascular events. A
metaregression analysis of published data from 20 studies of 95,783
individuals followed for 12.4 years. Diabetes Care 1999, 22, 233−240.
(11) Bonora, E.; Muggeo, M. Postprandial blood glucose as a risk
factor for cardiovascular disease in type II diabetes: the epidemio-
logical evidence. Diabetologia 2001, 44, 2107−2114.
(27) Wisselaar, H. A.; van Dongen, J. M.; Reuser, A. J. Effects of N-
hydroxyethyl-1-deoxynojirimycin (BAY m 1099) on the activity of
neutral- and acid α-glucosidases in human fibroblasts and HepG2 cells.
Clin. Chim. Acta 1989, 182, 41−52.
(28) Putter, J. In Enzyme Inhibitors; Brodech, E., Ed.; Verlag Chemie:
Weinheim, Germany, 1980; p 139.
(29) Joubert, P. H.; Venter, H. L.; Foukaridis, G. N. The effect of
miglitol and acarbose after an oral glucose load: a novel hypoglycaemic
mechanism? Br. J. Clin. Pharmacol. 1990, 30, 391−396.
(30) Bischoff, H. Pharmacology of α-glucosidase inhibition. Eur. J.
Clin. Invest. 1994, 24, 3−10.
(31) Arh, H. J.; Boberg, M; Brendei, E.; Krause, H. P.; Steinke, W.
Pharmacokinetics of miglitol. Absorption, distribution, metabolism,
and excretion following administration to rats, dogs, and man.
Arzneimittelforschung 1997, 47, 734−745.
(32) Kreymann, B.; Williams, G.; Ghatei, M. A.; Bloom, S. R.
Glucagon-like peptide-1 7-36: a physiological incretin in man. Lancet
1987, 2, 1300−1304.
(33) Zander, M.; Madsbad, S.; Madsen, J. L.; Holst, J. J. Effect of 6-
week course of glucagon-like peptide 1 on glycaemic control, insulin
sensitivity, and beta-cell function in type 2 diabetes: a parallel-group
study. Lancet 2002, 359, 824−830.
(34) Asano, N.; Ikeda, K.; Yu, Li.; Kato, A.; Takebayashi, K.; Adachi,
I.; Kato, I.; Ouchi, H.; Takahata, H.; Fleet, G. W. J. The L-enantiomers
of D-sugar-mimicking iminosugars are noncompetitive inhibitors of D-
glycohydrolase? Tetrahedron: Asymmetry 2005, 16, 223−229.
(35) da Cruz, F. P.; Newberry, S.; Jenkinson, S. F.; Wormald, M. R.;
Butters, T. D.; Alonzi, D. S.; Nakagawa, S.; Becq, F.; Norez, C.; Nash,
R. J.; Kato, A.; Fleet, G. W. J. 4-C-Me-DAB and 4-C-Me-LAB
enantiomeric alkyl-branched pyrrolidine iminosugarsare specific and
potent α-glucosidase inhibitors; acetone as the sole protecting group.
Tetrahedron Lett. 2011, 52, 219−223.
(36) Natori, Y.; Imahori, T.; Murakami, K.; Yoshimura, Y.;
Nakagawa, S.; Kato, A.; Adachi, I.; Takahata, H. The synthesis and
biological evaluation of 1-C-alkyl-L-arabinoiminofuranoses, a novel
class of α-glucosidase inhibitors. Bioorg. Med. Chem. Lett. 2011, 21,
738−741.
(37) Trost, B. M.; Horne, D. B.; Woltering, M. J. Palladium-catalyzed
DYKAT of butadiene monoepoxide: enantioselective total synthesis of
(+)-DMDP, (−)-bulgecinine, and (+)-broussonetine G. Chem.Eur.
J. 2006, 12, 6607−6620.
(38) Fu, G. C.; Zhou, J. S. Cross-couplings of unactivated secondary
alkyl halides: room-temperature nickel-catalyzed Negishi reactions of
alkyl bromides and iodides. J. Am. Chem. Soc. 2003, 125, 14726−
14727.
(12) Ceriello, A. Postprandial hyperglycemia and diabetes
complications: is it time to treat? Diabetes 2005, 54, 1−7.
(13) Iwasa, M.; Yamada, Y.; Kobayashi, H.; Yasuda, S.; Kawamura, I.;
Sumi, S.; Shiraki, T.; Yamaki, T.; Ushikoshi, H.; Hattori, A.; Aoyama,
T.; Nishigaki, K.; Takemura, G.; Fujiwara, H; Minatoguchi, S. Both
stimulation of GLP-1 recepters and inhibiton of glycogenolysis
additively contribute to a protective effect of oral miglitol against
ishaemia-reperfusion injury in rabbits. Br. J. Pharmacol. 2011, 164,
119−131.
(14) Asano, N.; Nash, R. J.; Molyneux, R. J.; Fleet, G. W. J. Sugar-
mimic glycosidase inhibitors: natural occurrence, biological activity and
prospects for therapeutic application. Tetrahedron: Asymmetry 2000,
11, 1645−1680.
(15) Sim, L.; Jayakanthan, K.; Mohan, S.; Nasi, R.; Johnston, B. D.;
Pinto, B. M.; Rose, D. R. New glucosidase inhibitors from an ayurvedic
herbal treatment for type 2 diabetes: structures and inhibition of
human intestinal maltase-glucoamylase with compounds from Salacia
reticulata. Biochemistry 2010, 49, 443−451.
(16) Eskandari, R.; Jones, K.; Rose, D. R.; Pinto, B. M. Selectivity of
3′-O-methylponkoranol for inhibition of N- and C-terminal maltase
glucoamylase and sucrase isomaltase, potential therapeutics for
digestive disorders or their sequelae. Bioorg. Med. Chem. Lett. 2011,
21, 6491−6494.
(17) Ogawa, S. Synthetic studies on glycosidase inhibitors composed
of 5a-carba-sugars. Carbohydrate Mimics: Concepts and Methods;
Chapleur, Y., Ed.; Wiley-VCH: Weinheim, Germany, 1998; pp 87−
106.
(18) Horii, S.; Iwasa, T.; Mizuta, E.; Kameda, Y. Studies on
validamycins, new antibiotics. VI. Validamine, hydroxyvalidamine and
validatol, new cyclitols. J. Antibiot. 1971, 24, 59−63.
(19) Kameda, Y.; Asano, N.; Yoshikawa, M.; Matsui, K. Valienamine
as an α-glucosidase inhibitor. J. Antibiot. 1980, 33, 1575−1576.
(20) Kameda, Y.; Asano, N.; Yoshikawa, M.; Takeuchi, M.;
Yamaguchi, T.; Matsui, K. Valiolamine, a new α-glucosidase inhibiting
aminocyclitol produced by Streptomyces hygroscopicus. J. Antibiot. 1984,
37, 1301−1307.
(21) Horii, S; Fukase, H.; Matsuo, T.; Asano, N.; Matsui, K.
Synthesis and α-D-glucosidase inhibitory activity of N-substituted
valiolamine derivatives as potential oral antidiabetic agents. J. Med.
Chem. 1986, 29, 1038−1046.
́
(39) Phapale, V. B.; Bunuel, E.; García-Iglesias, M.; Cardenas, D. J.
̃
Ni-catalyzed cascade formation of C(sp3)-C(sp3) bonds by cyclization
and cross-coupling reactions of iodoalkanes with alkyl zinc halides.
Angew. Chem., Int. Ed. 2007, 46, 8790−8795.
(40) Calveras, J.; Egido-Gabas
́ ́
, M.; Gomez, L.; Casas, J.; Parella, T.;
Joglar, J.; Bujons, J.; Clapes, P. Dihydroxyacetone phosphate aldolase
́
catalyzed synthesis of structurally diverse polyhydroxylated pyrrolidine
derivatives and evaluation of their glycosidase inhibitory properties.
Chem.Eur. J. 2009, 15, 7310−7328.
(22) Kato, A.; Kato, N.; Kano, E.; Adachi, I.; Ikeda, K.; Yu, L.;
Okamoto, T.; Banba, Y.; Ouchi, H.; Takahata, H.; Asano, N. Biological
properties of D- and L-1-deoxyazasugars. J. Med. Chem. 2005, 48,
2036−2044.
(41) Best, D.; Jenkinson, S. F.; Saville, A. W.; Alonzi, D. S.; Wormald,
M. R.; Butters, T. D.; Norez, C.; Becq, F.; Bleriot, Y.; Adachi, I.; Kato,
A.; Fleet, G. W. J. Cystic fibrosis and diabetes: isoLAB and isoDAB,
enantiomeric carbon-branched pyrrolidine iminosugars. Tetrahedron
Lett. 2010, 51, 4170−4174.
(23) Junge, B.; Matzke, M.; Stltefuss, J. In Handbook of Experimental
Pharmacology; Kuhlmann, J., Puls, W., Eds.; Springer-Verlag: New
York, 1996; Vol. 119, p 411.
(24) Cauderay, M.; Tappy, L.; Temler, E.; Jequier, E.; Hillebrand, I.;
Felber, J. P. Effect of α-glucohydrolase inhibitors (Bay m1099 and Bay
o1248) on sucrose metabolism in normal men. Metabolism 1986, 35,
472−477.
(42) Scofield, A. M.; Fellows, L. E.; Nash, R. J.; Fleet, G. W. J.
Inhibition of mammalian digestive disaccharides by polyhydroxy
alkaloids. Life Sci. 1986, 39, 645−650.
(43) Best, D.; Wang, C.; Weymouth-Wilson, A. C.; Clarkson, R. A.;
Wilson, F. X.; Nash, R. J.; Miyauchi, S.; Kato, A.; Fleet, G. W. J.
Looking glass inhibitors: scalable syntheses of DNJ, DMDP, and (3R)-
3-hydroxy-L-bulgecinine from D-glucuronolactone and of L-DNJ, L-
DMDP, and (3S)-3-hydroxy-D-bulgecinine from L-glucuronolactone.
DMDP inhibits β-glucosidases and β-galactosidases whereas L-DMDP
(25) Madar, Z.; Olefsky, J. Effect of the α-glucosidase inhibitor Bay-
O-1248 on the metabolic response of nondiabetic and diabetic rats to a
high-carbogydrate diet. Am. J. Clin. Nutr. 1986, 44, 206−211.
(26) Taylor, R. H.; Barker, H. M.; Bowey, E. A.; Canfield, J. E.
Regulation of absorption of dietary carbohydrate in man by two new
glycosidase inhibitors. Gut 1986, 27, 1471−1478.
O
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
is a potent and specific inhibitor of α-glucosidases. Tetrahedron:
Asymmetry 2010, 21, 311−319.
titols: synthesis of 1,4-dideoxy-1,4-imino-D-lyxitol and of both
enantiomer of 1,4-dideoxy-1,4-iminoarabinitol. Tetrahedron Lett.
1985, 26, 3127−3130.
(63) Kessler, M.; Acuto, O.; Strelli, C.; Murer, H.; Semenza, G. A. A
modified procedure for the rapid preparation of efficiently transporting
vesicles from small intestinal brush border membranes. Their use in
investigating some properties of D-glucose and choline transport
systems. Biochem. Biophys. Acta 1978, 506, 136−154.
(64) Cooper, C. A.; Gasteiger, E.; Packer, N. GlycoModa software
tool for determining glycosylation compositions from mass spectro-
metric data. Proteomics 2001, 1, 340−349.
(65) Sim, L.; Willemsma, C.; Mohan, S.; Naim, H. Y.; Pinto, B. M.;
Rose, D. R. Structural basis for substrate selectivity in human maltase-
glucoamylase and sucrase-isomaltase N-terminal domains. J. Biol.
Chem. 2010, 285, 17763−70.
(44) Ellgaard, L.; Molinari, M.; Helenius, A. Setting the standards:
quality control in the secretory pathway. Science 1999, 286, 1882−
1888.
(45) Kopito, R. R. ER quality control: the cytoplasmic connection.
Cell 1997, 88, 427−430.
(46) Hammond, C.; Braakman, I.; Helenius, A. Role of N-linked
oligosaccharide recognition, glucose trimming, and calnexin in
glycoprotein folding and quality control. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 913−917.
(47) Hosokawa, N.; Wada, I.; Hasegawa, K.; Yorifuji, T.; Tremblay,
L. O.; Herscovics, A.; Nagata, K. A novel ER α-mannosidase-like
protein accelerates ER-associated degradation. EMBO Rep. 2001, 2,
415−422.
(66) Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J.
M. PROCHECK: a program to check the stereochemical quality of
protein structures. J. Appl. Crystallogr. 1993, 26, 283−291.
(67) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.;
Klicic, J. J.; Mainz, D. T. Glide: a new approach for rapid, accurate
docking and scoring. 1. Method and assessment of docking accuracy. J.
Med. Chem. 2004, 47, 1739−1749.
(48) Hubbard, S. C.; Ivatt, R. J. Synthesis and processing of
asparagines-linked oligosaccharides. Annu. Rev. Biochem. 1981, 50,
555−583.
(49) Elbein, A. D. Inhibitors of the biosynthesis and processing of N-
linked oligosaccharides. CRC Crit. Rev. Biochem. 1984, 16, 21−49.
(50) Kopito, R. R. ER quality control: the cytoplasmic connection.
Cell 1997, 88, 427−430.
(51) Bonifacino, J. S.; Weissman, A. M. Ubiqitin and the control of
protein in the secretory and the endoplasmic pathways. Annu. Rev. Cell
Dev. Biol. 1998, 14, 19−57.
(52) Elbein, A. D. Glycosidase inhibitors: inhibitors of N-linked
oligosaccharide processing. FASEB J. 1991, 5, 3055−3063.
(53) Asano, N.; Oseki, K.; Kizu, H; Matsui, K. Nitrogen-in-the-ring
pyranoses and furanoses: structural basis of inhibition of mammalian
glycosidases. J. Med. Chem. 1994, 37, 3701−3706.
(68) Still, W. C.; Tempczyk, A.; Hawlely, R. C.; Hendrickson, T.
Semianalytical treatment of solvation for molecular mechanics and
dynamics. J. Am. Chem. Soc. 1990, 112, 6127−6129.
(69) Sherman, W.; Day, T.; Jacobson, M. P.; Friesner, R. A.; Farid, R.
Novel procedure for modeling ligand/receptor induced fit effects. J.
Med. Chem. 2006, 49, 534−553.
(54) Furukawa, J.; Shinohara, Y.; Kuramoto, H.; Miura, Y.; Shimaoka,
H.; Kurogochi, M.; Nakano, M.; Nishimura, S. I. Comprehensive
approach to structural and functional glycomics based on chemo-
selective glycoblotting and sequential tag conversion. Anal. Chem.
2008, 80, 1094−1101.
(55) Dailey, G. Early and intensive therapy for management of
hyperglycemia and cardiovascular risk factors in patients with type 2
diabetes. Clin. Ther. 2011, 33, 665−678.
(56) Arakawa, M.; Ebato, C.; Mita, T.; Fujitani, Y.; Shimizu, T.;
Watada, H.; Kawamori, R.; Hirose, T. Miglitol suppresses the
postprandial increase in interleukin 6 and enhances active glucagon-
like peptide 1 secretion in viscerally obese subjects. Metabolism 2008,
57, 1299−1306.
(57) Aoki, K.; Kamiyama, H.; Yoshimura, K.; Shibuya, M.; Masuda,
K.; Terauchi, Y. Miglitol administered before breakfast increased
plasma active glucagon-like peptide-1 (GLP-1) levels after lunch in
patients with type 2 diseases treated with sitagliptin. Acta Diabetol.
2012, 49, 225−230.
́
(58) Aguilar-Moncayo, M.; Takai, T.; Higaki, K.; Mena-Barragan, T.;
Hirano, Y.; Yura, K.; Li, L.; Yu, Y.; Ninomiya, H.; García-Moreno, M.
I.; Ishii, S.; Sakakibara, Y.; Ohno, K.; Nanba, E.; Ortiz Mellet, C.;
́
García Fernandez, J. M.; Suzuki, Y. Tuning glycosidase inhibition
through aglycone interactions: pharmacological chaperones for Fabry
disease and GM1 gangliosidosis. Chem. Commun. 2012, 48, 6514−
6516.
(59) Kato, A.; Yamashita, Y.; Nakagawa, S.; Koike, Y.; Adachi, I.;
Hollinshead, J.; Nas, R. J.; Ikeda, K.; Asano, N. 2,5-Dideoxy-2,5-imino-
D-aritol as a new class of pharmacological chaperone for Fabry disease.
Bioorg. Med. Chem. 2010, 18, 3790−3794.
(60) Godin, G.; Compain, P.; Masson, G.; Martin, O. R. A general
strategy for the practical synthesis of nojirimycin C-glycosides and
analogues. Extension to the first reported example of an iminosugar 1-
phosphonate. J. Org. Chem. 2002, 67, 6960−6970.
(61) Godin, G.; Compain, P.; Martin, O. R. General access to
iminosugar C-glycoside building blocks by means of cross-metathesis:
a gateway to glycoconjugate mimetics. Org. Lett. 2003, 5, 3269−3272.
(62) Fleet, G. W. J.; Nicholas, S. J.; Smith, P. W.; Evans, S. V.;
Fellows, L. E.; Nash, R. J. Potent competitive inhibition of α-
galactosidase and α-glucosidase activity by 1,4-dideoxy-1,4-iminopen-
P
dx.doi.org/10.1021/jm301304e | J. Med. Chem. XXXX, XXX, XXX−XXX