Conformationally Controlled Reactivity of Carbasugars Uncovers the Choreography of Glycoside Hydrolase Catalysis

Article abstract of DOI:10.1021/acs.joc.9b03152

Glycoside hydrolases (GHs) catalyze hydrolyses of glycoconjugates in which the enzyme choreographs a series of conformational changes during the catalytic cycle. As a result, some GH families, including α-amylases (GH13), have their chemical steps conceal

Full text of DOI:10.1021/acs.joc.9b03152

Subscriber access provided by GRIFFITH UNIVERSITY
Article
The Conformationally Controlled Reactivity of Carbasugars
Uncovers the Choreography of Glycoside Hydrolase Catalysis
Pal John Pal Adabala, Saeideh Shamsi Kazem Abadi,
Oluwafemi Akintola, Sandeep Bhosale, and Andrew J. Bennet
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.9b03152 • Publication Date (Web): 29 Jan 2020
Downloaded from pubs.acs.org on February 2, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
The Conformationally Controlled Reactivity of
Carbasugars Uncovers the Choreography of Glycoside
Hydrolase Catalysis
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Pal John Pal Adabala,1 Saeideh Shamsi Kazem Abadi,2 Oluwafemi Akintola,1 Sandeep
Bhosale,1 and Andrew J. Bennet1,*
1. Department of Chemistry, Simon Fraser University, 8888 University Drive, Burnaby, British
Columbia, V5A 1S6, Canada.
2. Department of Molecular Biology and Biochemistry, Simon Fraser University, 8888
University Drive, Burnaby, British Columbia, V5A 1S6, Canada
* To whom all correspondence should be addressed. e-mail: bennet@sfu.ca, Tel: (778)782–
8814, Fax: (778)782–3765.
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 44
1
2
3
4
ABSTRACT
5
6
7
8
Glycoside hydrolases (GHs) catalyze hydrolyses of glycoconjugates in which the enzyme
choreographs a series of conformational changes during the catalytic cycle. As a result, some
glycoside hydrolase families, including the -amylases (GH13), have their chemical steps
concealed kinetically. To address this issue for a GH13 enzyme we made seven cyclohexenyl-
based carbasugars of -D-glucopyranoside that we show are good covalent inhibitors of a GH13
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
yeast -glucosidase. The linear free energy relationships between rate constants and pKa of the
leaving group is curved upwards, which is indicative of a change in mechanism, with the better
leaving groups reacting by a SN1 mechanism, while reaction rates for the worse leaving groups
are limited by a conformational change of the Michaelis complex prior to a rapid SN2 reaction
with the enzymatic nucleophile. Five bicyclo[4.1.0]heptyl-based carbaglucoses were tested with
this enzyme, and our results are consistent with pseudo-glycosidic bond cleavage occurs via SN1
transition states that include non-productive binding of the leaving group to the enzyme. In total
we show that the conformationally orthogonal reactions of these two carbasugars reveal
mechanistic details hidden by conformational changes that the Michaelis complex of enzyme and
natural substrate undergoes that aligns the nucleophile for efficient catalysis.
2
ACS Paragon Plus Environment

Page 3 of 44
The Journal of Organic Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
An essential metabolic process involves the hydrolytic cleavage of carbohydrate residues from
glycoconjugates with the major enzyme family that performs this catabolic task being the
glycoside hydrolases (GHs).1 Glycoside hydrolases have been classified into GH families based
on their amino acid sequences, and currently there are over 160 such families.2 The majority of
GH families, with a few notable exceptions,3-7 catalyze the hydrolysis of glycosidic bonds by
standard nucleophilic substitution reactions where the nucleophile attacks the anomeric center.1
Such enzymatic mechanisms can be sub-divided into two sets, those that occur with a single
chemical step (inverting GHs) and those that use two sequential substitutions (retaining GHs).1,8
For most retaining GHs, the first substitution results in the formation of a covalent glycosyl-
enzyme intermediate, while the subsequent catalytic step involves hydrolysis of this intermediate.
The transition states for these two enzyme-catalyzed inversions of configuration are characterized
by their pyranosylium ion-like character, which involves delocalization of the nascent positive
charge onto the ring oxygen and this enforces the pyran ring into a non-chair conformation during
transition state traversal.8-10
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Currently, the mechanisms for many of these retaining glycoside hydrolases are proposed to
involve a glycopyranosylium ion-like TS in which the ring is in a 4H3 conformation (Figure 1).9,10
For example, retaining -glycosidases (GH families 18, 26, and 39) operate, during
glycosylation, via a 1S3 → [4H3]‡ → 4C1 conformational itinerary (CI).11-13 Therefore, a simple
microscopic reversibility argument suggests that the corresponding retaining -glycosidases
should traverse the reverse CI (4C1 → [4H3]‡ → 1S3), an analysis that is consistent with the
retaining -glucosidases from GH3114 and -galactosidases from GH2715 and GH37.16 However,
both kinetic and structural evidence suggests that -amylases, which are -glucosidases from
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 44
1
2
3
4
5
6
7
8
9
GH13, do not operate via the same CI as do GH31 glycoside hydrolases.17-19 Specifically, Hosie
and Sinnott argued that based on Brønsted lg values of 0 [plots of log(kcat/Km) versus aglycone
pKa(ArOH)] that a conformational change, prior to glycosidic bond cleavage, was rate limiting for
the yeast GH13 -glucosidase-catalyzed hydrolysis of aryl -D-glucopyranosides.17
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A
O
O
4C1 enzyme
intermediate
HO
O
H
general acid
O
O
O
O
H
-
OR

OH
O
HO
HO
OR
HO
HO
H
HO
OH
O
O
O
-
O
OH
O
1S3 Michaelis
complex
O
O
4H3 glucosylation
transition state
nucleophile
B
HO
O
HO
O
HO
O
O
HO
HO
O
OH
OH OH HO
OH
OH
N
OH
OH
H
OH
1
HO
HO
HO
HO
NO2
HO
HO
NO2
O
O
OH
3
2
Figure 1: Panel A: Mechanism of action and conformational itinerary for retaining -
glucosidases; Panel B: Structures of the natural product -glucosidase inhibitor acarbose (1) and
the mechanism-based GH13 inhibitors (2 and 3).
Also, these authors noted that the large -secondary deuterium KIE values measured for the
catalyzed hydrolysis of -D-glucopyranosyl pyridinium salts requires efficient hyperconjugative
overlap between the C2–H bonding electrons and the nascent electron-deficient p-orbital at the
glucopyranosylium ion-like TS; conformers with a nascent double bond, which is forming
4
ACS Paragon Plus Environment

Page 5 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
between the anomeric carbon and the ring oxygen, that fit this TS criterion include 2,5B and 4H3.17
Thus, it appears that the two major GH families (GH13 and 31) that hydrolyze -D-
glucopyranoside linkages have similar pyranosylium ion transition states, but have antithetical
conformational itineraries. In other words, GH13 and GH31 enzymes have similar glycosylation
TSs with their respective CIs being the converse of each other.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Glycoside hydrolase family 13 (GH13) is the principal enzyme family that breakdown
polysaccharides containing -D-glucopyranoside linkages, which are important dietary
carbohydrates.20 Of note, in addition to other activities GH13 family members display a wide
range of substrate specificities,21 with many of these enzymes being closely related.
Consequently, GH13 has been further classified into several subfamilies.22 Indeed, due to the
importance of GH13 enzymes as the entry point for glycolysis, by virtue of their catalytic
production of glucose, it is not surprising that numerous natural product inhibitors are known for
this enzyme family.23-26 For instance, the valienamine class of inhibitors have been well studied,
however, whether acarbose (1, Figure 1b),27 a slow binding inhibitor which binds with its
cyclohexenyl ring in a 2H3 half-chair conformation,28,29 is a transition state analogue inhibitor of
GH13 enzymes is a controversial issue.30,31
Previously, we have shown that the cyclohexenyl-based carbasugar analogue 2, a structural
analogue of acarbose (1), is a mechanism-based covalent inhibitor of a retaining GH13 enzyme.32
That is, the mechanism for the GH-catalyzed "pseudo-glycosylation" reaction entails covalent
bond formation between the pseudo-anomeric carbon of the carbasugar 2 and the active site
nucleophile.33 This nucleophilic substitution reaction results in the time-dependent loss of
enzymatic activity, however, the covalently bound intermediate undergoes slow hydrolytic
cleavage to regenerate active enzyme.32 In addition, we have shown that the bicyclo[4.1.0]heptyl
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 44
1
2
3
4
5
6
7
8
9
glucose analogue 3 is also a mechanism-based covalent inhibitor of the yeast GH13 enzyme.32 To
investigate the finer mechanistic details of action for these covalent inhibitors, and to gain insight
into the conformational itinerary of GH13 enzymes, we synthesized two panels of carbaglucose
covalent inhibitors (4a–f and 5a–d, Figure 2). Then using linear free energy relationships
(LFERs), we delineated the dependence of leaving group ability on the enzymatic rate constants
for covalent labeling, so as to investigate whether conformational changes are also implicated for
these GH13-catalyzed events.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HO
HO
HO
HO
HO
HO
HO
HO
O HO
NO2
F
NO2
F
O
O
O
O
OH
NO2
F
OH
F
OH
4c
4a
4b
HO
HO
HO
HO
HO
HO
HO
HO
HO
O
F
O
NO2
O
OH
OH
OH
4d
4e
4f
HO
HO
HO
HO
HO
F
HO
HO
NO2
HO
HO
NO2
O
O
F
O
OH
NO2
Cl
OH
5c
OH
5a
5b
HO
HO
HO
HO
Cl
HO
HO
O
OH
OH
5d
OH
4-OH
Figure 2: Covalent inhibitor panels used in the current study; 4a–f and 5a–d and the natural
product streptol 4-OH.
6
ACS Paragon Plus Environment

Page 7 of 44
The Journal of Organic Chemistry
1
2
3
4
RESULTS
5
6
7
8
In order to make panels of cyclohexenyl (4a–f) and bicyclo[4.1.0]heptyl (5a–d) covalent
inhibitors we used two main approaches, which involved simplifying our previous syntheses that
gave the two carbasugar skeletons. Specifically, for the syntheses of 4b, 4c, 4d, and 4f we
changed our protecting group strategy32 to be similar to that reported for the synthesis of the
parent natural product streptol 4-OH, but we used a metathesis reaction rather than the reported
Horner-Wadsworth-Emmons olefination for ring closure.34 We started with benzyl -D-
glucopyranoside (6) and protected the C2 and C3 hydroxyl groups using butane-2,3-dione in the
presence of trimethyl orthoformate, and this material, after separation from the contaminating
3,4-isomer, was reacted to protect the C4 and C6-OH groups with an acetonide giving 7 (Scheme
1). Hydrogenolysis of the benzyl group, aldose reduction and protection of the primary alcohol
gave 8, which underwent oxidation of the secondary alcohol, Wittig olefination and removal of
the silyl protecting group yielded 9. We found that a TEMPO mediated oxidation followed by
vinyl Grignard addition gave a major diastereomer that after ring closing metathesis we
characterized to be the R-diastereomer 10 (Scheme 1). We completed the syntheses of four new
cyclohexenyl covalent inhibitors (4b-d, f) by performing Mitsunobu reactions followed by the
acid-catalyzed removal of the protecting groups.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 44
1
2
3
4
5
6
O
O
O
HO
HO
OH
O
O
O
a, b
c-e
f-h
O
OBn
OTBDPS
OBn
7
8
O
O
O
HO
OH
9
MeO
OMe
MeO
OMe
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6
7
8
O
O
O
O
HO
HO
i-k
l, m
OH
OH
O Ar
O
O
HO
OH
O
O
MeO
OMe
MeO
OMe
4b-d 4f
,
9
10
Scheme 1. Synthesis of Cyclohexenyl Carbaglucose Covalent Inhibitors 4b–d and 4f. Reagents
a) CH3COCOCH3, HC(OCH3)3, cat H+, 57%; b) (CH3)2C(OCH3)2, acetone, cat H+, 73%; c)
H2/Pd, EtOAc; d) NaBH4, MeOH; e) TBDPS-Cl, imidazole, CH2Cl2 78% steps c–e; f) Dess-
Martin periodinane, CH2Cl2, 92%; g) Julia-Kocienski olefination, 83%; h) (Bu)4NF, THF, 96%;
i) TEMPO, BIAB, CH2Cl2; j) CH2=CHMgBr, THF, 64% two steps; k) i) Grubbs' 2nd generation
catalyst ii) Ac2O, Et3N, iii) NaOMe/MeOH, 78% three steps; l) ArOH, Ph3P, DIAD; m) TFA,
CH2Cl2, H2O; yield range 6–43% over two steps.
For the synthesis of 4a and 4e as well as three of the bicyclo[4.1.0]heptyl carbaglucose
inhibitors we used the tetra-O-ethoxymethyl cyclohexenol 11 (Scheme 2), which we made by the
route of Shing et al.35 For the two remaining cyclohexenyl inhibitors we installed the aromatic
leaving groups by SNAr reactions (Scheme 2). Whereas, for the bicyclic inhibitors, we used the
pseudo-anomeric hydroxyl group in (11) to direct the cyclopropanation to give the -D-
glucopyranose analogue 12 as the only isolable diastereomer. Following installation of a variety
of aromatic groups, via SNAr reactions, acid-catalyzed deprotection yielded three of the desired
bicyclo[4.1.0]heptyl analogues of -D-glucopyranoside. Finally, we made 5c by using the tetra-
8
ACS Paragon Plus Environment

Page 9 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
O-benzyl protected bicyclo[4.1.0]heptyl alcohol (13), which we used in the synthesis of 3,32 by
performing an SNAr reaction with 1,3,5-trifluorobenzene followed by hydrogenolysis of the
benzyl ethers (Supporting Information).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EtO
O
EtO
O
HO
HO
HO
O
O
EtO
EtO
a
b
or
c
EtO
EtO
Ar
O
Ar
O
O
O
OH
OH
EtO
O
EtO
O
4a
4e
and
11
d
EtO
O
HO
HO
HO
O
BnO
HO
HO
HO
e
then
f
or ,
EtO
EtO
BnO
BnO
f h
g
,
Ar
Ar
O
O
OH
OH
O
OH
5a, 5b, 5d
EtO
O
BnO
13
HO
12
5c
Scheme 2. Synthesis of Cyclohexenyl Carbaglucose Analogue Covalent Inhibitors 5a–d and
Cyclohexenyl Analogues 4a and 4e. Reagents: a) DABCO, 1-F,2,4-NO2Ph, DMF, 84%; b) NaH,
(2,4,6-OMe3Ph)I+(3-NO2Ph) OTs–, TBME, 45%; c) CH3OH, cat H+, yield range 63–84% over
two steps; d) ZnEt2, CH2I2, CH2Cl2, 0 C 72%; e) DABCO, 1-F,2,4-NO2Ph, DMF, 15%; f) NaH,
ArF, DMF, yield range 27–77%; g) TFA, CH2Cl2, H2O, yield range 56–77%; h) H2, Pd(OH)2,
MeOH, 71%.
Shown in Figure 3 are the measured first-order rate constants (kobs) for the loss of yeast GH13
-glucosidase activity as a function of the concentration of five cyclohexenyl carbasugar
analogues (4b–f), and Figure 4 contains the corresponding kinetic data for the covalent labeling
by bicyclo[4.1.0]heptyl inhibitors 5a–d. Of note, due to the high reactivity of 4a, we monitored
the loss of GH13 activity in the presence of high concentrations of substrate (full details in
Supporting Information).
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 44
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Inactivation kinetics for covalent inhibition of yeast GH13 -glucosidase by
cyclohexenyl inhibitors. Leaving group ability represented as pKa (BH+); Panel A: 6,8-difluoro-4-
methylumbelliferone (4b, 4.81, solid red); 3,5-difluorophenol (4d, 8.26, solid blue); and phenol
(4f, 9.92, solid black); Panel B: 3-fluoro-4-nitrophenol (4c, 5.86, hollow blue); and 3-nitrophenol
(4e, 8.26, hollow red). The solid and dashed lines are the best non-linear fits to a standard
Michaelis–Menten equation. Conditions for all experiments were T = 25 C in sodium phosphate
buffer (50 mM, pH 6.84, [BSA] = 1 mg/mL).
10
ACS Paragon Plus Environment

Page 11 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Inactivation kinetics for covalent inhibition of yeast GH13 -glucosidase by
bicyclo[4.1.0]heptyl inhibitors. Leaving group ability represented as pKa (BH+); Panel A: 2,4-
dinitrophenol (5a, 3.96, solid red); and 2-chloro-4-nitrophenol (5b, 5.45, solid blue); Panel B:
3,5-difluorophenol (5c, 8.26, hollow blue); and 4-chlorophenol (5d, 9.41, hollow red). The solid
and dashed lines are the best non-linear fits to a standard Michaelis–Menten equation. Conditions
for all experiments were T = 25 C in sodium phosphate buffer (50 mM, pH 6.84, [BSA] = 1
mg/mL).
We then fit the kinetic data for inhibition of the yeast enzyme by both covalent inhibitor panels
(Figures 3 and 4) to a standard Michaelis-Menten expression to obtain values for the two
inhibition constants kinact and kinact/Ki (the kinetic data is compiled in Tables S1 and S2;
Supporting Information). Shown in Figure 5 are the Brønsted LFER plots for the covalent
labeling of GH13 yeast -glucosidase by these cyclohexenyl and bicyclo[4.1.0]heptyl
carbasugars.
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 44
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Brønsted plots for the reactions of yeast GH13 -glucosidase. Panel A, second-order
rate constant data (kinact/Ki) for covalent labeling by 2 and 4a–f (red circles) and by 3 and 5a–d
(blue circles). Conditions for all experiments are T = 25 C in sodium phosphate buffer (50 mM,
pH = 6.84, [BSA] = 1 mg/mL). The solid red lines are the best linear fits for the inhibitors 4a–c
and 4c–f & 2 (kinact/Ki values), while the solid blue line is the best linear fit for inhibitors 3 and
5a–d. Panel B, first-order rate constant data (kinact) for covalent labeling by 2 and 4a–f (red
circles) and by 3 and 5a–d (blue circles). The red line is the best linear fit for the Brønsted plot
for the covalent labeling by 4b–f & 2 (point for 4a is excluded), while the blue line is the linear
fit for inhibitors and 5a–d (point for 3 is excluded).. . Leaving group ability represented as pKa
(BH+): 2,4-dinitrophenol (4.10); 6,8-difluoro-4-methylumbelliferone (4.81); 2-chloro-4-
nitrophenol (5.45); 3-fluoro-4-nitrophenol (5.86); 4-nitrophenol (7.15); 3,5-difluorophenol (8.26);
4-chlorophenol (9.41); and phenol (9.92). Where error bar are not shown that are encompassed
within the data point symbol.
The Brønsted plots for covalent labeling of yeast GH13 -glucosidase by the panel of
cyclohexenyl-carbasugars display evidence of a change in mechanism as the leaving group ability
is modified. That is, the slope of the logarithm of the second-order (kinact/Ki) rate constants and
12
ACS Paragon Plus Environment

Page 13 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
pKa(ROH) at high pKa values is 0 while for good leaving groups it is –1.3 (Figure 5A, red
circles). A similar trend, although less pronounced, is noticeable in the corresponding plot for the
first-order rate constants for covalent labeling (kinact; Figure 5B, red circles).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The derived lg values on the first- (kinact, the point for 3 was excluded from the fit) and second-
order rate constants (kinact/Ki) for the panel of bicyclo[4.1.0]heptyl covalent inhibitors are –0.22 
0.05 (r2 = 0.92) and –0.56  0.07 (r2 = 0.95), respectively (Figure 5, blue data points). The good
correlation on kinact/Ki for covalent inhibition by the bicyclic carbasugars covers an approximately
~103–fold change in kinact/Ki values that spans a pKa range of 5 units (Supporting Information
Tables S2).
DISCUSSION
The interest in carbasugars analogues, such as acarbose (1), as GH inhibitors is based, in part,
on their presumed status as transition state analogues (TSAs) due to their tight competitive
binding. Of note, researchers have tried to test whether such compounds are TSAs by using
Bartlett's linear free energy relationship (LFER) approach.36 Withers and co-workers concluded
that acarbose is a TSA inhibitor of cyclodextrin glycosyltransferase, a GH13 family member,
using the truncated substrates -D-glucopyranosyl and -D-maltotriosyl fluoride.31 In contrast,
based on pre-steady state kinetics with extended natural substrates, using a different GH13
enzyme, Berland et al concluded that acarbose is not a TSA inhibitor for the hydrolysis of
oligosaccharide natural substrates.30 Of note, these two group reported, markedly, different slopes
for their respective correlations of log(Ki) against log(Km/kcat), which based on TS theory should
be 1,36 with Withers and co-workers recording a slope of 0.42 [reported as a slope of 2.2 for the
inverse correlation of log(Km/kcat) against log(Ki)].31 While Berland et al disclosed a slope of 2
[reported as 2 for a log(Ka) versus log(kcat/Km) LFER].30
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 44
1
2
3
4
5
6
7
8
9
Given these divergent conclusions we decided to use an alternative approach for TS and
conformational itinerary characterization of GH13-catalysis by delineating the detailed
mechanism of covalent inhibition by a panel of allylic and bicyclic carbaglucose analogues. We
note that different patterns of reactivity are apparent for the covalent inhibition of the yeast GH13
-glucosidase by the cyclohexenyl- and the bicyclo[4.1.0]-carbasugars (Figure 5).
We first analyze our results with the cyclohexenyl covalent inhibitors (4a–f and 2) that are
close mimics of the valienamine class of natural products. We show that, for the worse leaving
groups (pKa(ArOH) > 6.0), no obvious correlation is apparent in the Brønsted plots for either of the
two kinetic constants kinact/Ki (Figure 5A) and kinact (Figure 5B). Moreover, given that we are
monitoring enzymatic covalent inhibition only steps up to, and including, the covalent event itself
contribute to the measured rate constants. Unexpectedly, our observations are reminiscent of
those reported by Hosie and Sinnott who noted that no Brønsted correlation was apparent for the
-glucosidase-catalyzed hydrolysis of aryl -D-glucopyranosides (Scheme 3), results that formed
the basis of their conclusion that an essential and rate-limiting, conformational change occurs
prior to glycosidic bond cleavage.17
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
k3
k1
k2
k4
E + GluOH
E:Sc
E:Ssb
E + S
E-G
k-1
Scheme 3: Mechanism for the yeast GH13 -glucosidase-catalyzed hydrolysis of aryl -D-
glucopyranosides in which the reaction of substrate proceeds through chair (Sc) and skew-boat
(Ssb) conformations.17
In other words, they proposed that the initial Michaelis complex (E:Sc, Scheme 3) undergoes a
conformational change, to give the substrate in a non-chair, likely a skew boat, conformation
(E:Ssb), and this event orientates the anomeric center in a position that facilitates a rapid SN2-type
14
ACS Paragon Plus Environment

Page 15 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
nucleophilic displacement reaction.17,18 However, in contrast to the Brønsted plots of Hosie and
Sinnott, Figures 5A and B display upward curvatures for the best leaving groups (weakest
phenoxide bases). Such a concave upward plot is consistent with a change in mechanism
occurring as the leaving group ability increases. For GH-catalyzed reactions, such plots have
been interpreted to indicate a change in mechanism from phenoxide departure (with no
protonation) to one where glycosidic bond cleavage requires acid catalysis.37 However, we
propose that for this panel of cyclohexenyl inhibitors, with the inferior leaving groups, that the
initial Michaelis complex (E:IOArhc; Scheme 4) requires a conformational change in order to
undergo a catalyzed SN2-type reaction with the enzymatic nucleophile. Whereas, for the better
leaving groups no such assistance is required and the mechanism involves a unimolecular SN1
reaction that occurs directly from the initial Michaelis complex (Scheme 4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
kreact
k1
k3
k2
E + IOAr
E-I
E:IOArhc
kinact
E + IOHhc
E:IOAr
-ArO(H)
SN2
k-1
SN1
-
-ArO
kreact
E-I
E + IOHhc
Scheme 4: Proposed mechanism for the yeast GH13 -glucosidase-catalyzed covalent
inhibition by aryl cyclohexenyl carbaglucose analogues 4a–f and 2 in which the covalent
inhibitor (IOArhc) and reaction product (IOHhc) have half-chair (hc) ground states.
For the bicyclo[4.1.0]heptyl covalent inhibitors (3, and 5a-d), we note that the second-order
rate constants (kinact/Ki) for covalent inhibition of -glucosidase exhibits a Brønsted correlation
with a lg value of –0.56  0.07 (Figure 5B). This value is consistent with the enzyme-catalyzed
cleavage of the pseudo-anomeric C–O bond being the major kinetically significant step.
Unexpectedly, a less negative lg value of –0.22 (without the data point for the 4-nitrophenyl
derivative 3) is evident for the Brønsted correlation on the first-order pseudo-glycosylation rate
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 44
1
2
3
4
5
6
7
8
constant (kinact) by these bicyclic inhibitors. This result requires, in the simplest kinetic scheme
(Scheme 5), that either a step following C–O bond cleavage is partially rate-determining for kinact
or that the equilibrium constant for dissociation of the Michaelis complex (Ki = k2/k1) is weakly
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
correlated with the pKa of the aglycone phenol. Given that the first-order rate constant (kinact
)
reports on the transformation of the E:I Michaelis complex to the covalently labeled enzyme (E–
I; Scheme 5) the only step that occurs after covalent bond formation is aglycone dissociation
from the active site, a process that should be orders of magnitude more rapid, and therefore
kinetically silent, than the rate constant for covalent labeling (kinact < 2  10–3 s–1).
k1
k2
kinact
kreact
E:IOH
E + IOAr
E:IOAr
E-I
-
-ArO
Scheme 5: Proposed mechanism for the yeast GH13 -glucosidase-catalyzed covalent
inhibition by aryl bicyclo[4.1.0]heptyl carbaglucose analogues 5a–d and 3.
We therefore conclude that interactions of the +1 site on the enzyme that accommodates the
aglycone during hydrolysis of -D-glucopyranosides with the aromatic leaving group are
consistent with non-productive binding for inhibitors 5a–d, an effect that is expressed at the
Michaelis complex but not at the pseudo-glycosylation TS.38 That is, a rough correlation exists
between Ki and pKa(ArOH) with the more electron deficient aromatic containing inhibitors having
higher binding affinities (lower Ki).38
We note that the large difference in lg values on kinact/Ki for the SN1-type reactions of the
allylic (lg  –1.3) and cyclopropyl (lg = –0.56) for covalent inhibition of the -glucosidase
requires that the TS for cleavage of the allylic C–O bond is later (more advanced) than for the
corresponding cyclopropyl inhibitors. For the more reactive cyclohexene-base inhibitors (4a–c)
the lg value is close to the maximally expected limit and is thus consistent with a late transition
16
ACS Paragon Plus Environment

Page 17 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
state with no acid-catalysis being provided by the enzyme, as would be expected for such good
leaving groups. In contrast, the significantly lower lg value for the bicyclic inhibitors (5a–d)
indicates that the enzyme-catalyzed reaction for these covalent inhibitors, which also occurs
without acid catalysis, has an earlier TS with less C–O cleavage. We conclude that an earlier TS
for the bicyclic inhibitors is a result of strain energy release as the bicyclobutonium ion
intermediate is formed by C–C -bond participation. We also suggest that these covalent
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
inhibitors, by virtue that the mechanism involves backside C–C participation into the *-orbital
of the pseudo-glycosidic C–O bond, are unable to undergo direct SN2 reactions within the
enzyme active site and that reactivity of such inhibitors is determined by the intrinsic reactivity of
the bound conformation rather than by conformational changes that promote SN2-type reactions
as noted with the cyclohexenyl compounds with the worse leaving groups.
B2,5
OS2
1S5
E5
4H5
1,4B
3,OB
OH5
4E
4H3‡
E3
OE
OH1
1S3
4C1
3S1
E1
2H3
2E
B1,4
B3,O
2H1
5S1
2SO
2,5B
Figure 6. Proposed reaction coordinates for GH13 enzyme-catalyzed hydrolysis of natural
substrates and covalent inhibition by carbasugar analogues shown on a Northern polar Stoddart
diagram projection. Substrate hydrolysis occurs via an initial distortion of the Michaelis complex
via a 2H3 half-chair toward a 1S3 skew boat (dashed red arrow) so that the anomeric carbon is
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 44
1
2
3
4
5
6
7
8
9
aligned for nucleophilic attack to give the covalent enzyme intermediate (E–G, 4C1) via the 4H3
TS (solid red arrow). Bicyclo[4.1.0]heptyl and activated cyclohexenyl inhibitors react from 1,4B
(teal lettering) and 2H3 (blue lettering) conformations, respectively. Less activated cyclohexenyl
inhibitors require a conformational distortion to allow a rapid SN2-type reaction (blue dashed
arrow).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Finally, we propose a unified scheme (Figure 6) that is consistent with previously published
protein structures,19,39 and the reactivity of GH13 enzymes with substrates17 and our covalent
inhibitors. Specifically, we suggest that substrate hydrolysis involves distortion of the initial
Michaelis complex towards a 2H3 half-chair,19 which is driven, in part, by an interaction between
the CH2OH of the glucose residue in the –1 site and the catalytic nucleophile, and as this distance
decreases the distortion increases. A deduction that is based on the structure of a GH13 enzyme
with acarbose bound (Figure S33, Supporting Information). As this rate-limiting ring
conformational change17 occurs (Figure 6, dashed red arrow) the anomeric carbon, and the
glycosidic oxygen, move into position for rapid nucleophilic substitution that proceeds via TS
traversal (4H3) to give the covalent enzyme intermediate (E–G, 4C1; solid red arrow). The less
activated cyclohexenyl inhibitors require a rate-determining conformational change to permit a
rapid SN2 reaction (Figure 6, blue dashed arrow), an event that results in a covalent labeled
enzyme in which the carbocycle relaxes back to a 2H3 conformation (arrow not shown on Figure
6). In contrast, covalent inhibition by the bicyclo[4.1.0]heptyl and the more activated
cyclohexenyl compounds occurs from their ground state bound conformations that are 1,4B (teal
lettering) and 2H3 (blue lettering), respectively (Figure 6). In these SN1 mechanism the resulting
cationic intermediates are trapped by the enzymatic nucleophile (after rearrangement in the
18
ACS Paragon Plus Environment

Page 19 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
bicyclobutonium ion case) to give covalent intermediates in the same conformations as the
Michaelis complexes.
7
8
CONCLUSIONS
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The mechanism of action for these two classes of carbasugar analogues have been shown to
display significant differences as the aglycone leaving group ability changes. Such carbasugar
covalent inhibitors, unlike other inactivators including cyclophellitol and cyclophellitol
aziridine,40-42 can have their reactivity turned by changing the leaving group. Moreover, we have
shown that the cyclohexenyl carbasugar inhibitors undergo a change in mechanism from SN1
(good leaving groups) to an obligatory conformational change followed by a rapid SN2
displacement (worse phenoxide leaving groups). In contrast, the bicyclo[4.1.0]heptyl
carbaglucose covalent inhibitors react via SN1 mechanisms in which the lg value (kinact/Ki) of –
0.55 is indicative of an early TS with little delocalization of positive charge that results from C–C
-bond participation. While the corresponding SN1 reactions for the allylic inhibitors with good
leaving groups proceed via late TSs in which charge delocalization is essentially complete.
EXPERIMENTAL METHODS
General. All reactions described were performed under an atmosphere of dry nitrogen using
oven dried glassware unless otherwise specified. All reactions that required heating were
performed using a silicone oil bath. Analytical thin-layer chromatography (TLC) was performed
on aluminum plates pre-coated with silica gel 60F-254 as the adsorbent. The developed TLC
plates were exposed to UV light and/or sprayed with a solution containing 1% Ce(SO4)2 and
1.5% molybdic acid in 10% aqueous H2SO4, and heated for visualization of spots. Column
chromatography was performed with an automated flash chromatography system using pre-
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 44
1
2
3
4
5
6
7
8
packed high performance RediSep Rf Gold flash columns packed with 20–40 µm spherical silica
gel. All reagents, solvents and starting materials were purchased from Sigma Aldrich, Alfa Aesar,
TCI America, Strem, EMD, Anachemia, Caledon, Fisher or ACP and were used without further
9
1
purification unless otherwise specified. H and 13C NMR spectra were recorded on either a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bruker Avance 600 equipped with a QNP or QCI cryoprobe (600 MHz) instrument using CDCl3,
CD3COCD3 and CD3OD as solvents. Chemical shifts are reported in ppm downfield to
1
13
tetramethylsilane. Coupling constants are reported and expressed in hertz. All H and C NMR
1
1
peaks were assigned with the aid of H-1H (COSY) and/or H-13C (HSQC) experiments using
standard pulse programs. High resolution mass spectra were acquired on a Bruker maXis Impact
TOF LC/MS/MS system. The reagent (3-nitrophenyl)(2,4,6-trimethoxyphenyl)iodonium tosylate
was made by the method that Seidl et al. used for their synthesis of (3-nitro-4-
methylphenyl)(2,4,6-trimethoxyphenyl)iodonium tosylate.43
Benzyl 2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-diyl]-β-D-glucopyranoside: To a stirred
solution of 1,2,3,4,6-penta-O-acetylglucopyranose (100 g, 256 mmol) and benzyl alcohol (26.6
mL, 256 mmol) in anhydrous CH2Cl2 (375 mL) was added BF3·Et2O (31.6 mL, 256 mmol)
dropwise. The reaction mixture was stirred at room temperature (rt) overnight and quenched by
the slow addition of saturated aqueous NaHCO3 (500 mL) at 0 C. The organic layer was
separated and aqueous layer was washed with CH2Cl2 (2 × 600 mL). The combined organic
layers were washed with brine and dried (Na2SO4). Evaporation of the solvent followed by
purification of the crude product by column chromatography (40% to 50% EtOAc in hexanes)
gave benzyl 2,3,4,6-tetra-O-acetyl-β-D-glucopyranoside as a white solid (44 g, 39%). To a stirred
solution of this compound (44 g, 100 mmol) in MeOH (150 mL) a catalytic amount of a 1 M
NaOMe solution (5.2 mL) was added and this reaction mixture was stirred at rt for 24 h, followed
20
ACS Paragon Plus Environment

Page 21 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
by the addition of IR120 ion-exchange resin (H+–form) so as to lower the pH of the reaction
mixture to 7.0. The reaction mixture was filtered and the solvent was evaporated to give crude
benzyl β-D-glucopyranoside 6 (27 g). This crude product was used directly in the next step. That
is, dimethoxybutane-2,3-diyl protecting group was added by following a literature procedure44 to
give a white foam (23 g, 57% yield) that was used in the next step.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Benzyl 2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-diyl]-4,6-O-isopropylidene-β-D-glucopyr-
anoside (7): 2,2-Dimethoxypropane (20.3 mL, 166 mmol) and a catalytic amount of
camphorsulfonic acid (1.75 g, 7.5 mmol) were added to a stirred solution of benzyl -glucoside
from the previous step (29 g, 75.4 mmol) in acetone (125 mL). The reaction mixture was stirred
at rt for 24 h, and then powdered NaHCO3 was added. The resulting mixture was filtered through
a Celite pad and the filtrate was concentrated under reduced pressure to give crude product that
was purified by column chromatography (10% to 20% EtOAc in hexanes) to give compound 7 as
a white solid (23.5 g, 73%).
1H NMR (600 MHz, CDCl3): δ 7.36 (d, J = 7.2 Hz, 2H, Ar-H), 7.33 (t, J = 7.5 Hz, 2H, Ar-H),
7.28 (d, J = 7.2 Hz, 1H, Ar-H), 4.91 (d, J = 12.2 Hz, 1H, −CHPh), 4.67 (d, J = 12.2 Hz, 1H,
−CHPh), 4.65 (d, J1,2 = 8.0 Hz, 1H, H-1), 3.90 (dd, J6a,6b = 10.8, J5,6a = 5.3 Hz, 1H, H-6a), 3.84-
3.78 (m, 2H, H-4, H-6b), 3.75 (t, J2,3 = J3,4 = 9.3 Hz, 1H, H-3), 3.67 (dd, J2,3 = 9.5, J1,2 = 8.0 Hz,
1H, H-2), 3.29 (s, 3H, -OCH3), 3.28 (s, 3H, -OCH3), 3.28–3.24 (m, 1H, H-5), 1.49 (s, 3H, -CH3),
1.41 (s, 3H, -CH3), 1.33 (s, 3H, -CH3), 1.33 (s, 3H, -CH3); 13C{1H} NMR (151 MHz, CDCl3): δ
137.7 (Ar-C), 128.4 (Ar-C), 127.6 (Ar-C), 127.3 (Ar-C), 100.7 (C-1), 99.7, 99.6, 99.4, 71.2 (-
CH2Ph), 70.6 (C-3), 70.5 (C-2), 70.0 (C-4), 68.5 (C-5), 62.2 (C-6), 48.0 (-OCH3), 47.9 (-OCH3),
29.1 (-CH3), 19.1 (-CH3), 17.6 (-CH3), 17.6 (-CH3); HRMS (ESI-UHR-Qq-TOF) m/z: (M + Na)+
Calcd for C22H32O8Na 447.1995; Found: 447.1992.
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 44
1
2
3
4
1-O-tert-Butyldiphenylsilyl-2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-diyl]-4,6-O-
5
6
7
8
isopropylidene-D-glucitol (8): To a solution of benzyl glucoside 7 (23.5 g, 55.4 mmol) in EtOAc
(300 mL) was added 20% Pd(OH)2/C (1.2 g). The reaction mixture was stirred for 5–6 h under
H2 (1 atm) at rt. The catalyst was removed by filtration through a Celite pad, which was washed
with EtOAc (700 mL) and the combined filtrate was concentrated under reduced pressure. The
crude hemiacetal (20.7 g, 61.9 mmol) was dissolved in MeOH (350 mL) to which NaBH4 (5.4 g,
142 mmol) was added portionwise at 0 C. This reaction mixture was stirred at rt for 1 h. The
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
solvent was evaporated and the residue was dissolved in EtOAc (400 mL) to which saturated
aqueous NaHCO3 (400 mL) was added and then stirred for 10 min. The organic layer was
separated and the aqueous layer was washed with EtOAc (2 × 400 mL). The combined organic
layers were washed with brine, dried (Na2SO4), and the solvent was evaporated to give crude
diol. To a solution of this crude product (18.5 g, 55 mmol) in dry CH2Cl2 (225 mL) was added
imidazole (8.24 g, 120 mmol) and TBDPSCl (17.2 mL, 66 mmol). The reaction mixture was
stirred at rt for 2 h, quenched with MeOH followed by the addition of water (400 mL). The
organic layer was separated and the aqueous layer was washed with CH2Cl2 (2 × 400 mL). The
combined organic layers were washed with brine, dried (Na2SO4), and the solvent was
evaporated. The resulting crude residue was purified by column chromatography (30% to 40%
EtOAc in hexanes) to give the 8 as a colorless liquid (25 g, 78% three steps).
1H NMR (600 MHz, CDCl3): δ 7.76–7.72 (m, 4H, Ar-H), 7.42–7.40 (m, 2H, Ar-H), 7.38–7.35
(m, 4H, Ar-H), 4.03–3.98 (m, 2H), 3.93–3.90 (m, 1H), 3.88–3.83 (m, 2H), 3.76–3.73 (m, 1H),
3.70 (dd, J = 8.9, 2.1 Hz, 1H), 3.57–3.52 (m, 1H), 3.28 (s, 3H, -OCH3), 3.25 (s, 3H, -OCH3), 2.58
(s, 1H, -OH), 1.34 (s, 3H, -CH3), 1.33 (s, 3H, -CH3), 1.30 (s, 3H, -CH3), 1.23 (s, 3H, -CH3), 1.05
(s, 9H, -C(CH3)3); 13C{1H} NMR (151 MHz, CDCl3): δ 135.8 (Ar-C), 135.7 (Ar-C), 133.9 (Ar-
22
ACS Paragon Plus Environment

Page 23 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
C), 133.8 (Ar-C), 129.5 (Ar-C), 129.5 (Ar-C), 127.6 (Ar-C), 127.6 (Ar-C), 99.0, 99.0, 98.4, 73.0,
69.5, 68.0, 64.6, 63.4, 63.2, 48.0 (-OCH3), 47.8 (-OCH3), 28.2 (-CH3), 26.7 (-C(CH3)3), 19.4 (-
C(CH3)3), 19.12 (-CH3), 17.7 (2x-CH3); HRMS (ESI-UHR-Qq-TOF) m/z: (M + Na)+ Calcd for
C31H46O8SiNa 597.2860; Found: 597.2868.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1-O-tert-Butyldiphenylsilyl-2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-diyl]-4,6-O-
isopropylidene-5-oxo-D-xylo-hexitol: To a stirred solution of alcohol 8 (25 g, 43.5 mmol) Dess-
Martin periodinane (DMP, 22.14 g, 52.2 mmol) was added and the reaction mixture was stirred at
rt for 3 h. Then a mixture of saturated aqueous sodium thiosulfate (200 mL) and saturated sodium
bicarbonate (200 mL) were added, and this mixture was stirred for 30 min. After separating the
layers, the aqueous phase was extracted with CH2Cl2 (2 × 500 mL), and the combined organic
layers were washed with brine, dried (Na2SO4), and concentrated under reduced pressure. The
resulting crude residue was purified by column chromatography (20% to 30% EtOAc in hexanes)
to afford a ketone, which was used directly in the following reaction, as a colorless liquid (23.2 g,
93%).
1H NMR (600 MHz, CDCl3): δ 7.75 (dd, J = 8.0, 1.4 Hz, 2H, Ar-H), 7.71–7.68 (m, 2H, Ar-H),
7.44–7.40 (m, 2H, Ar-H), 7.38–7.35 (m, 4H, Ar-H), 4.55 (dd, J2,3 = 9.7, J3,4 =2.4 Hz, 1H, H-3),
4.34–4.33 (m, 1H, H-4), 4.30 (dd, J6a,6b = 15.8, J6a,4 = 1.4 Hz, 1H, H-6a), 4.01 (dt, J2,3 = 9.7, J2,1a
= J2,1b = 3.6 Hz, 1H, H-2), 3.91 (d, J6a,6b = 15.8 Hz, 1H, H-6b), 3.77–3.71 (m, 2H, H-1a, H-1b),
3.25 (s, 3H, -OCH3), 3.22 (s, 3H, -OCH3), 1.48 (s, 3H, -CH3), 1.30 (s, 3H, -CH3), 1.25 (s, 3H, -
13
CH3), 1.24 (s, 3H, -CH3), 1.06 (s, 9H, -C(CH3)3); C{1H} NMR (151 MHz, CDCl3): δ 207.0 (-
CO), 136.0 (Ar-C), 135.7 (Ar-C), 133.8 (Ar-C), 133.4 (Ar-C), 129.8 (Ar-C), 129.8 (Ar-C), 127.8
(Ar-C), 127.7 (Ar-C), 100.7, 99.3, 99.1, 74.8 (C-4), 68.5 (C-2), 67.4 (C-3), 67.4 (C-6), 63.7 (C-
1), 48.2 (-OCH3), 47.8 (-OCH3), 26.9 (-C(CH3)3), 24.9 (-CH3), 23.4 (-CH3), 19.5 (-C(CH3)3),
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 44
1
2
3
4
5
17.9 (-CH3), 17.6 (-CH3); HRMS (ESI-UHR-Qq-TOF) m/z: (M + Na)+ Calcd for C31H44O8SiNa
595.2703; Found: 595.2705.
6
7
8
1-O-tert-Butyldiphenylsilyl-5-deoxy-2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-diyl]-4,6-O-
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
isopropylidene-5-methylene-D-xylo-hexitol: To
a
cooled (–78 C) solution of 5-
(methanesulfonyl)-1-phenyl-1H-tetrazole (11.3 g, 50.3 mmol) in 80 mL of dry THF was added
dropwise a 1M solution of NaHMDS in THF (47.2 mL, 47.1 mmol) and the resulting pale yellow
mixture was stirred for 30 min at –78 C. A solution of ketone from previous step (9 g, 15.7
mmol) in THF (80 mL) was added dropwise and the resultant mixture was stirred at –78 C for a
further 1 h. After the addition of water (300 mL) the reaction mixture was extracted with EtOAc
(3 × 300 mL) and combined the organic layers were washed with brine, dried (Na2SO4), and
concentrated under reduced pressure. The resulting crude product was purified by column
chromatography (8% to 12% EtOAc in hexanes) to afford the olefin, which was used directly in
the following reaction, as a colorless liquid (7.5 g, 83%).
1H NMR (600 MHz, CD3COCD3): δ 7.82–7.77 (m, 4H, Ar-H), 7.47–7.41 (m, 6H, Ar-H), 5.14
(s, 1H, -C=CH), 4.98 (s, 1H, -C=CH), 4.59 (s, 1H, H-4), 4.26 (d, J6a,6b = 12.9 Hz, 1H, H-6a),
4.17–4.12 (m, 2H, H-2, H-3), 4.03 (d, J6a,6b = 12.9 Hz, 1H, H-6b), 3.94 (dd, J1a,1b = 11.4, J1a,2
=
1.7 Hz, 1H, H-1a), 3.81 (dd, J1a,1b = 11.4, J1b,2 = 5.0 Hz, 1H, H-1b), 3.23 (s, 3H, -OCH3), 3.20 (s,
3H, -OCH3), 1.31 (s, 3H, -CH3), 1.28 (s, 3H, -CH3), 1.25 (s, 6H, 2 x -CH3), 1.06 (s, 9H, -
C(CH3)3); 13C{1H} NMR (151 MHz, CD3COCD3): δ 145.2 (-C=CH2), 136.7 (Ar-C), 136.5 (Ar-
C), 134.7 (Ar-C), 134.5 (Ar-C), 130.7 (Ar-C), 130.7 (Ar-C), 128.7 (Ar-C), 128.6 (Ar-C), 107.9 (-
C=CH2), 100.2, 99.9, 99.5, 71.7 (C-4), 71.0 (C-2), 70.0 (C-3), 65.0 (C-6), 64.6 (C-1), 48.3 (-
OCH3), 47.9 (-OCH3), 27.3 (-C(CH3)3), 26.8 (-CH3), 24.1 (-CH3), 20.1 (-C(CH3)3), 18.2 (-CH3),
24
ACS Paragon Plus Environment

Page 25 of 44
The Journal of Organic Chemistry
1
2
3
4
5
18.1 (-CH3); HRMS (ESI-UHR-Qq-TOF) Calcd for C32H46O7SiNa m/z: (M + Na)+ 593.2911;
Found: 593.2906.
6
7
8
9
5-deoxy-2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-diyl]-4,6-O-isopropylidene-5-methylene-D-
xylo-hexitol (9): To a stirred solution of the above mentioned olefin (18.5 g, 32.4 mmol) in THF
(325 mL) was added a 1.0 M solution of TBAF in THF (48.6 mL, 48.6 mmol) and this reaction
mixture was stirred at rt overnight. Removal of the volatiles under reduced pressure gave a crude
product that was purified by column chromatography (40% to 50% EtOAc in hexanes) to afford
primary alcohol 9 as a colorless liquid (10.4 g, 96%); HRMS (ESI-UHR-Qq-TOF) m/z: (M +
Na)+ Calcd for C16H28O7Na 355.1733; Found: 355.1738. Compound 9 was used directly in the
next step due to facile double bond rearrangement, Figures S31 and S32 (Supporting
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
Information) show the H NMR spectra of 9, directly after isolation, in CDCl3 and acetone-d6
(where a slower rearrangement occurs).
5,5a-Didehydro-2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-diyl]-4,6-O-isopropylidene-5a-
carba--D-xylo-hexopyranosyl acetate): To a solution of alcohol 9 (2.3 g, 6.9 mmol) in CH2Cl2
(40.0 mL), (diacetoxyiodo)benzene (2.9 g, 9.0 mmol) and TEMPO (108 mg, 0.69 mmol) were
added. After the reaction mixture had stirred at rt overnight, the solvent was evaporated and the
resulting crude aldehyde was used directly in the next step. To a cooled (–78 C) solution of this
crude aldehyde in dry THF (30 mL), a 1.0 M solution of vinylmagnesium bromide in THF (20.8
mL, 20.8 mmol) was added dropwise. The reaction mixture was allowed to warm to –40 C and
then stirred for 2h at this temperature. Addition of saturated aqueous NH4Cl (80 mL) quenched
the reaction. After separating the layers, the aqueous phase was extracted with ethyl acetate (3 ×
100 mL), and the combined organic extracts were washed with brine and dried (Na2SO4). After
removal of the solvent, the crude product was purified by column chromatography (40% to 50%
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 44
1
2
3
4
5
6
7
8
EtOAc in hexanes) to afford a diastereomeric mixtures of dienes (1.6 g, 64% two steps). The
compound was used directly for the next reaction without further analysis. HRMS (ESI-UHR-
Qq-TOF) m/z: (M + Na)+ Calcd for C18H30O7Na 381.1889; Found: 381.1893.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Following addition of a second-generation Grubbs' catalyst (Zhan-1B, 65 mg, 0.089 mmol) to a
stirred solution of diastereomeric dienes (1.06 g, 2.96 mmol) in dry toluene (30 mL) at rt, the
1
resultant mixture was stirred at 110 C for 2 h. After completion of reaction (monitoring by H-
NMR spectroscopy), the solvent was evaporated under reduced pressure to give crude product
that was purified by column chromatography (60% to 80% EtOAc in hexanes) to give an
inseparable mixture of allylic alcohols (783 mg, 80%). To a solution of these alcohols (400 mg,
1.2 mmol) in CH2Cl2 (5 mL) was added trimethylamine (0.5 mL, 3.6 mmol), acetic anhydride
(0.35 mL, 3.6 mmol) and a catalytic quantity of DMAP (15 mg, 0.12 mmol). The resulting
mixture was stirred at rt overnight. Following evaporation of the solvent, the crude material was
purified (30% to 40% EtOAc in hexanes) to give colorless β-acetate as a solid (388 mg, 86%) and
13
an oil of the α-acetate (39 mg, 9 %). The spectral data (1H and C{1H} NMR) of both the β-
acetate and α-acetate exactly matched that reported in the literature.45
5,5a-Didehydro-2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-diyl]-4,6-O-isopropylidene-5a-
carba--D-xylo-hexopyranose (10): To a stirred solution of β-acetate (350 mg, 0.94 mmol) in
methanol (10 mL) was added a catalytic amount NaOMe (1 M in MeOH, 0.1 mL, 0.1 mmol) and
this mixture was stirred overnight. After the solvent was evaporated, the residue was dissolved in
EtOAc (70 mL), saturated aqueous NH4Cl (70 mL) was added, and the organic layer separated.
The aqueous layer was then washed with EtOAc (2 × 70 mL), and the combined organic layers
were washed with brine, dried (Na2SO4), and evaporated under reduced pressure. The resulting
crude product was purified by column chromatography (60% to 80% EtOAc in hexanes) to afford
26
ACS Paragon Plus Environment

Page 27 of 44
The Journal of Organic Chemistry
1
2
3
4
5
alcohol 10 as a solid (283 mg, 91%). The spectral data (1H and 13C NMR) of alcohol 10 matched
that reported in the literature.45
6
7
8
9
6,8-Difluoro-4-methylumbelliferyl 5,5a-didehydro-2,3-O-[(2R,3R)-2,3-dimethoxybutane-2,3-
diyl]-4,6-O-isopropylidene-5a-carba--D-xylo-hexopyranoside: To a cooled (0 C) solution of
alcohol 10 (45 mg, 0.136 mmol) in dry THF (2.0 mL) was added triphenylphosphine (54 mg,
0.204 mmol) and 6,8-difluoro-4-methylumbelliferone (43 mg, 0.204 mmol). The resulting
mixture was stirred at 0 C for 20 min, when DIAD (0.04 mL, 0.204 mmol) was added and the
reaction mixture was warmed to rt and stirring was continued for 24 h. The reaction mixture was
dissolved in water (20 mL) and extracted with EtOAc (3 × 20 mL). The combined organic layers
were washed with brine, dried (Na2SO4), and concentrated. The resulting crude residue was
purified by column chromatography (40% to 50% EtOAc in hexanes) to afford the protected 8-
fluoroumbelliferyl inhibitor as a white solid (25 mg, 35%).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (600 MHz, CDCl3): δ 7.05 (dd, J = 10.7, 1.8 Hz, 1H, Ar-H), 6.26 (s, 1H, Ar-H), 5.71
(d, J1,6 = 4.2 Hz, 1H, H-5a), 5.03 (t, J1,2 = J1,5a = 3.9 Hz, 1H, H-1), 4.47–4.42 (m, 2H, H-4, H-6a)
4.38 (dd, J2,3 = 10.9, J3,4 = 8.0 Hz, 1H, H-3), 4.20 (d, J6a,6b = 14.2 Hz, 1H, H-6b), 3.76 (dd, J2,3
=
10.9, J1,2 = 3.5 Hz, 1H, H-2), 3.33 (s, 3H, -OCH3), 3.17 (s, 3H, -OCH3), 2.38 (s, 3H, -CH3), 1.52
13
(s, 3H, -CH3), 1.43 (s, 3H, -CH3), 1.29 (s, 3H, -CH3), 1.03 (s, 3H, -CH3); C{1H} NMR (151
MHz, CDCl3): δ 159.6 (CO), 152.0 (dd, J = 246.5, 4.4, Hz, Ar-C), 151.6 (appt, Ar-C), 143.4 (dd,
J = 253.4, 5.8 Hz, Ar-C), 139.9 (-C=CH), 139.7 (dd, J = 10.2, 2.0 Hz, Ar-C), 139.3 (dd, J = 15.6,
10.7 Hz, Ar-C), 116.3 (-C=CH), 114.7 (Ar-C), 104.5 (d, J = 8.6 Hz, Ar-C), 105.1 (dd, J = 22.1,
3.1 Hz, Ar-C), 99.6, 99.5, 98.9, 76.5 (appt, C-1), 70.3 (C-4), 68.4 (C-2), 67.0 (C-3), 62.8 (C-6),
48.2 (-OCH3), 48.0 (-OCH3), 27.9 (-CH3), 20.7 (-CH3), 18.9 (-CH3), 17.9 (-CH3), 17.3 (-CH3);
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 44
1
2
3
4
5
HRMS (ESI-UHR-Qq-TOF) m/z: (M + Na)+ Calcd for C26H30F2O9Na 547.1756; Found:
547.1752.
6
7
8
9
6,8-Difluoro-4-methylumbelliferyl
5,5a-didehydro-5a-carba--D-xylo-hexopyranoside
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(4b). To a stirred solution of 4-methylumbelliferyl derivative (20 mg, 0.038 mmol) in CH2Cl2
(1.5 mL) TFA (0.18 mL) and water (0.018 mL) were added and the reaction was stirred at rt for 2
h. Evaporation of the solvent followed by purification of the crude material by column
chromatography (10% to 15% MeOH in CH2Cl2) gave the compound 4b as a white solid (5 mg,
35%).
1H NMR (600 MHz, CD3OD) δ 7.42 (dd, J = 11.4, 2.1 Hz, 1H, Ar-H), 6.36 (d, J = 1.4 Hz, 1H,
Ar-H), 5.97 (dq, J1,5a = 5.5, J5a,6a = J5a,6b = J4,5a = 1.7 Hz, 1H, H-5a), 5.11 (t, J1,2 = J1,5a = 4.6 Hz,
1H, H-1), 4.25 (ABq, J6a,6b = 15.3 Hz, 2H, H-6a, H-6b), 4.13 (dd, J2,3 = 10.3, J3,4 = 7.6 Hz, 1H,
H-3), 4.06 (dd, J3,4 = 7.7, J4,5a = 1.7 Hz, 1H, H-4), 3.71 (dd, J2,3 = 10.2, J1,2 = 3.6 Hz, 1H, H-2),
2.46 (d, J = 1.3 Hz, 3H, -CH3); 13C{1H} NMR (151 MHz, CD3OD) δ 161.3 (CO) 154.5 (appt, J
= 2.4 Hz, Ar-C), 153.2 (dd, J = 244.6, 4.3 Hz, Ar-C), 148.8 (-C=CH), 144.4 (dd, J = 250.6, 6.0
Hz, Ar-C), 140.9 (dd, J = 9.9, 2.1 Hz, Ar-C), 139.9 (dd, J = 15.8, 10.6 Hz, Ar-C), 118.2 (-
C=CH), 116.3 (d, J = 9.5 Hz, Ar-C), 115.0 (Ar-C), 107.2 (dd, J = 22.8, 3.2 Hz, Ar-C), 80.5 (appt,
C-1), 74.3 (C-4), 74.1 (C-3), 72.9 (C-2), 62.7 (C-6), 18.7 (-CH3); HRMS (ESI-UHR-Qq-TOF)
m/z: (M + Na)+ Calcd for C17H16F2O7Na 393.0762; Found: 393.0756.
3-Fluoro-4-nitrophenyl 5,5a-didehydro-5a-carba--D-xylo-hexopyranoside (4c): To a
cooled (0 C) solution of compound 10 (59 mg, 0.179 mmol) in dry THF (2.5 mL) was added
triphenylphosphine (70 mg, 0.268 mmol) and 3-fluoro-4-nitrophenol (42 mg, 0.268 mmol) with
the resulting mixture being stirred at 0 C for 20 min. DIAD (0.05 mL, 0.268 mmol) was added
and the mixture was stirred at rt for 24 h. After the addition of water (30 mL) the mixture was
28
ACS Paragon Plus Environment

Page 29 of 44
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
extracted with EtOAc (3 × 30 mL). The combined organic layers were washed with brine, dried
(Na2SO4), and concentrated. The resulting crude product was purified by column chromatography
(40% to 50% EtOAc in hexanes) to afford the protected 3-fluoro-4-nitrophenyl derivative (15
mg, 18%). This material was deprotected directly by adding it (13 mg, 0.028 mmol) to a solution
of CH2Cl2 (1 mL), TFA (0.14 mL), and water (0.014 mL). The resultant mixture was stirred at rt
for 2 h, when evaporation of the volatiles, followed by flash column chromatography (10% to
15% MeOH in CH2Cl2) gave 4c as a white solid (3 mg, 34%).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (400 MHz, CD3OD) δ 8.11 (t, J = 9.1 Hz, 1H, Ar-H), 7.07 (dd, J = 13.4, 2.6 Hz, 1H,
Ar-H), 6.99 (ddd, J = 9.2, 2.7, 1.0 Hz, 1H, Ar-H), 6.02–5.98 (m, 1H, H-5a), 5.12 (t, J1,2 = J1,5a
=
4.6 Hz, 1H, H-1), 4.27–4.16 (m, 2H, H-6a, H-6b), 4.04 (d, J3,4 = 7.2 Hz, 1H, H-4), 3.95 (dd, J2,3 =
10.2, J3,4 = 7.4 Hz, 1H, H-3), 3.75 (dd, J2,3 = 10.2, J1,2 = 3.9 Hz, 1H, H-2); 13C NMR (151 MHz,
CD3OD) δ 166.1 (dd, J = 11.4 Hz, Ar-C), 158.6 (dd, J = 262.3 Hz, Ar-C), 148.2 (-C=CH), 128.8
(Ar-C), 117.5 (-C=CH), 113.0 (dd, J = 2.9 Hz, Ar-C), 105.7 (dd, J = 24.3 Hz, Ar-C), 75.8 (C-1),
74.5 (C-3), 74.1 (C-4), 71.9 (C-2), 62.7 (C-6). HRMS (ESI-UHR-Qq-TOF) m/z: (M + Na)+ Calcd
for C13H14FNO7Na 338.0652; Found: 338.0658.
3,5-Difluorophenyl 5,5a-didehydro-5a-carba--D-xylo-hexopyranoside (4d): To a cooled
(0 C) solution of alcohol 10 (40.0 mg, 0.121 mmol) in dry THF (2.0 mL) was added
triphenylphosphine (48 mg, 0.182 mmol) and 3,5-difluorophenol (24 mg, 0.182 mmol) with the
resulting mixture being stirred at 0 C for 20 min. DIAD (0.04 mL, 0.182 mmol) was added and
the mixture was stirred at rt for 24 h. After addition of water (20 mL), the mixture was extracted
with EtOAc (3 × 20 mL) and the combined organic layers were washed with brine, dried
(Na2SO4), and concentrated. The resulting crude residue was purified by column chromatography
(30% to 40% EtOAc in hexanes) to afford the protected 3,5-difluorophenyl derivative (52 mg,
29
ACS Paragon Plus Environment