Impact of Aromatic Stacking on Glycoside Reactivity: Balancing CH/πand Cation/πInteractions for the Stabilization of Glycosyl-Oxocarbenium Ions

Article abstract of DOI:10.1021/jacs.9b03285

Carbohydrate/aromatic stacking represents a recurring key motif for the molecular recognition of glycosides, either by protein binding domains, enzymes, or synthetic receptors. Interestingly, it has been proposed that aromatic residues might also assist i

Full text of DOI:10.1021/jacs.9b03285

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC
Article
The impact of aromatic stacking on glycoside reactivity: balancing CH/#
and cation/# interactions for the stabilization of glycosyl-oxocarbenium ions
Laura Montalvillo-Jiménez, Andrés G. Santana, Francisco Corzana, Gonzalo
Jiménez-Osés, Jesús Jiménez-Barbero, Ana M. Gomez, and Juan Luis L. Asensio
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b03285 • Publication Date (Web): 07 Aug 2019
Downloaded from pubs.acs.org on August 7, 2019
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 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
The impact of aromatic stacking on glycoside reactivity: balancing
CH/ and cation/ interactions for the stabilization of glycosyl-
oxocarbenium ions
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
a
a
Laura Montalvillo-Jiménez, ‡ Andrés G. Santana, ‡ Francisco Corzana,b Gonzalo Jiménez-Osés,c
Jesús Jiménez-Barbero,c,d Ana M. Gómez,a Juan Luis Asensioa*
aInstituto de Química Orgánica (IQOG-CSIC), Juan de la Cierva 3, 28006 Madrid, Spain. Fax: +34-91-5644853; Tel:
+34-91-5622900.
bDept. Química and Centro de Investigación en Síntesis Química, Universidad de La Rioja.
cCenter for Cooperative Research in Biosciences (CIC-bioGUNE).
dBasque Foundation for Science, Ikerbasque.
‡ AGS and LM-J are both first authors
* e-mail: juanluis.asensio@csic.es, phone: +34 915622900, Fax: +34 915644853
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 23
ABSTRACT
1
2
Carbohydrate/aromatic stacking represents a recurring key motif for the molecular recognition of
glycosides, either by protein binding domains, enzymes or synthetic receptors. Interestingly, it has
been proposed that aromatic residues might also assist in the formation/cleavage of glycosidic bonds
by stabilizing positively charged oxocarbenium-like intermediates/transition states through cation/
interactions. While the significance of aromatic stacking on glycoside recognition is well stablished,
its impact on the reactivity of glycosyl donors is yet to be explored. Herein, we report the first
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
experimental study on this relevant topic. Our strategy is based on the design, synthesis and reactivity
evaluation of a large number of model systems, comprising a wide range of glycosidic donor/aromatic
complexes. Different stacking geometries and dynamic features, anomeric leaving groups, sugar
configurations and reaction conditions have been explicitly considered. The obtained results
underline the opposing influence exerted by van der Waals and coulombic forces on the reactivity of
the carbohydrate/aromatic complex: depending on the outcome of this balance, aromatic platforms
can indeed exert a variety of effects, stretching from reaction inhibition all the way to rate
enhancements. Although aromatic/glycosyl cation contacts are highly dynamic, the conclusions of
our study suggest that aromatic assistance to glycosylation processes must indeed be feasible, with
far reaching implications for enzyme engineering and organocatalysis.
2
ACS Paragon Plus Environment

Page 3 of 23
Journal of the American Chemical Society
INTRODUCTION
1
2
Aromatic stacking plays a prominent role in the molecular recognition of saccharidic ligands
by proteins, nucleic acids or synthetic receptors.1-3 These contacts result from a combination of
dispersive, hydrophobic and electrostatic forces, involving polarized pyranose CH groups and the
aromatic electronic quadrupole,4,5 and can exhibit a variety of geometries. Their relevance is
highlighted by data-mining studies which show the extraordinary high frequency of aromatic amino
acids, especially tryptophan, in the carbohydrate binding sites of proteins. Of note, tryptophan,
tyrosine and phenylalanine residues are also common within the catalytic sites of glycosidases and
glycosyl transferases, prompting the question of whether they might also assist the cleavage or
formation of glycosidic bonds (Figure 1).
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
Different scenarios can be envisaged to describe the influence of aromatic stacking on the
reactivity of pyranoses. Thus, for concerted SN2 processes, shielding of the anomeric center by the
aromatic platform would be expected to exert a protective influence, potentially leading to a sharp
decrease in reactivity. On the contrary, more dissociative transition states might benefit from
favorable electrostatic forces arising between the aromatic quadrupole and the pyranose developing
positive charge (cation/ interaction), 6,7 especially in low dielectric environments (Figure 1). In
addition, aromatic/glycosyl cation interactions could be expected to drive the stereochemical course
of the reaction by preventing the nucleophilic attack through either the  or  pyranose face. Fittingly,
oxocarbenium-like intermediates or transition states have commonly been invoked to explain the
outcome of glycosylation and glycoside hydrolytic reactions both in the chemical8-12 and enzymatic
contexts.13-15 As a further consideration, the formation of glycosyl cations is usually accompanied by
substantial conformational changes of the reactive pyranose ring.8 Conceivably, this geometric
adjustment could be impeded, or facilitated, by pre-organized CH/ bonds, depending on the
interaction geometry and dynamic behavior of the complex. Additional factors such as the access of
solvent molecules to the catalytic region, its local dielectric enviroment or the presence of counter
ions should also be taken into account, drawing a rather complex scenario. In fact, while the relevance
of aromatic platforms in glycan recognition is undisputable, their possible roles in catalysis remain
uncertain.
Despite the relative lack of experimental information available on this topic, some examples
have been reported in which tyrosine or phenylalanine residues seem to assist the cleavage/formation
of glycosidic linkages by presenting favourable interactions with the carbohydrate transition states.
Thus, a hydrophobic platform comprising a phenylalanine residue, highly conserved in the active
centre of all glycosyl hydrolases, has been proposed as a mechanistically relevant transition state
stabilising factor.16 More recently, a seminal contribution by Davis and col. has put forward a novel
mechanism for a ‘SNi synthase’, engineered from a retaining ‘double-SN2’ hydrolase.17 According to
kinetic, structural and theoretical considerations, this mutant protein makes use of a tyrosine residue,
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 23
instead of a carboxylate, to stabilize the oxocarbenium-like transition state generated during the front-
face attack of the glycosidic acceptor to the acetalic center (Figure 1b). Interestingly, this tyrosine
residue is originally involved in a parallel stacking with the -face of the glycosidic donor and
therefore seems to play a dual role, participating both in substrate recognition and catalysis. These
examples illustrate that under appropriate conditions, aromatic platforms can be employed to stabilize
glycoside transition states. However, the precise structural and chemical requirements for this
function are presently unknown.
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 1.-a) Aromatic interactions in a -glucosidase/DNJ complex as revealed by the
corresponding X-ray diffraction structure (pdb code: 2J77) . b) Aromatic assistance promotes
catalysis in the context of an engineered front-face SNi-synthase.17
Our aim is to understand the influence of carbohydrate/aromatic stacking on the pyranose
reactivity to eventually apply this knowledge for synthesis and catalysis. With this purpose in mind,
and as a first step, we have herein employed a chemical approach based on the design, synthesis and
evaluation of appropriate molecular systems, in conjunction with the extensive use of NMR as a key
13
monitorization and detection tool, facilitated in particular cases by the employment of C-labeled
models. Despite its significance, both from the fundamental and applied perspectives, this work
represents, to the best of our knowledge, the first systematic experimental study reported on this topic.
RESULTS
a) Design of the model glycosidic-donor/aromatic complexes: Basic principles. – In order to
enhance any potential stabilizing electrostatic interaction between the aromatic quadrupole and the
cationic intermediates/transition states, we decided to carry out our studies in a low dielectric
environment such as chloroform (=4.8).18 It is worth mentioning that comparable dielectric constants
have been previously theorized to properly model the inner protein environment in binding sites or
catalytic regions.19 Accordingly, the designed scaffolds encompass pyranose units equipped with
several non-participating MeO- groups, which in turn provide high solubility in organic solvents.
Two alternative families of model systems were conceived (Figure 2a). Compounds in family I
contain a disaccharide scaffold. The reactive pyranoses display alternative leaving groups at the
4
ACS Paragon Plus Environment

Page 5 of 23
Journal of the American Chemical Society
anomeric centre and participate in contacts with a pyrene platform that is attached to a vicinal 3-
amino-allose unit through an amide bond (compounds 1-3, 6 and 8). Alternatively, the compounds in
family II (Figure 2b) include a glycosyl donor involved in a stacking complex with a pyrene unit that
is tethered to a 4,6-O-benzylidene protecting group. Considering the key influence exerted by the
chemical nature and presentation of the substituent at position 2 of the pyranose moiety on the
reactivity of the acetalic center, manno- (10) gluco- (11) and 2-deoxy-gluco- (16) derivatives were
explicitly considered. Moreover, to maximize the influence of the aromatic ring on the pyranose
behavior, a derivative comprising a 2-deoxy-glucose donor sandwiched between two pyrene
platforms was also prepared (18). For comparison purposes, both families include reference
compounds equipped with smaller or non-aromatic units (derivatives 4, 5, 7 and 9 for type I models
and 12-15, 17 and 19 for type II models). Finally, selected type II models were also prepared
incorporating 13C atoms, either at all pyranose positions (11, 14, 15, 16 and 17) or just at the anomeric
center (10 and 13). Thus, relatively fast reactions could be conveniently monitored using 2D HSQC
NMR experiments, also allowing an easy and sensitive detection of the transient species even in
highly overlapped spectra (see the experimental section).
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 2.- Chemical structures of model systems I (a) and II (b). Conformational variability for the
established carbohydrate/aromatic complexes type II is illustrated. c) Chemical reactions selected to
evaluate the impact of aromatic stacking on glycoside reactivity. CIP and SSIP stand for close ion
pair and solvent-separated ion pair, respectively.
There are two important differences between models type I and II that are relevant from a reactivity
perspective and deserve special attention. First, family I scaffolds generate stable
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 23
carbohydrate/aromatic complexes with a defined geometry, involving either the - (1-3, 6) or the -
(8) face of the pyranose (Figure 2a). On the contrary, for type II systems, rotation of the benzylidene
moiety with respect to the pyranose unit is allowed. This fact provides a highly dynamic environment
to form adaptable carbohydrate/aromatic stacking complexes. Indeed, MD simulations confirm that
several interaction geometries are possible for these type II systems. They comprise parallel and
parallel-shifted complexes mediated by the - or  face of the glycoside, together with lateral, edge-
to-face aromatic stacking modes involving the hydroxymethyl moiety (Figure 2b and S1). As a
second consideration, family I derivatives encompass an unconstrained pyranose donor, while for
family II, the pyranose units are fixed by the 4,6-O-benzilidene moiety. Consequently, only a reduced
subset of the pyranose conformational space is available for the oxocarbenium-like transition
states/intermediates. The resulting combinations (fixed stacking/flexible pyranose, in models type I
and adaptable stacking/constrained pyranose, in models type II) allow dissecting the influence of
aromatic stacking on the glycoside reactivity in two complementary scenarios.
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
The anomeric leaving groups for both scaffolds were selected based on chemical considerations
(Figure 2c). Family I derivatives are structurally more complex and more challenging to obtain with
13C-labels. Therefore, they were tested in straightforward acid-catalyzed methanolysis reactions, for
which acid labile p-nitro-phenoxy or acetoxy anomeric substituents were incorporated at the reactive
site. These processes are assumed to proceed through cationic transition states.9 On the other hand,
family II scaffolds afforded less overlapped NMR spectra. The use of 13C-labels for particular models
offered the opportunity to carry out more demanding, low temperature assays with relevant
glycosylation reaction conditions.8 Accordingly, these models displayed phenyl sulfoxide moieties
as leaving groups. Such glycosylations proceed through a continuum of mechanisms spanning the
gap between pure SN2 and SN1 processes, as shown by extensive studies performed by Crich and col.8
These reactions require the participation of several key intermediates (Figure 2c), including glycosyl
triflates20 and/or glycosyl oxocarbenium ions (either solvent equilibrated or forming close ion pairs
with the corresponding counterions).8 Fortunately, more dissociative processes, involving cationic
transition states, can also be favoured by the appropriate choice of the acceptor alcohol, as recently
shown by Codée and col.21 Taking these factors into account, we selected a weakly nucleophilic
acceptor such as trifluoroethanol (TFE), unless stated otherwise.
b) General synthetic strategy. - The synthetic route for the preparation of all systems is represented
in Scheme 1. The glycosylation reaction between a glycosyl donor containing a phenyl 3-azido-3-
deoxy -D-allo-thiopyranoside unit5b and the appropriately protected glycosyl acceptor (2-OH-
galacto-, 3-OH-gluco-, and 4-OH-gluco-pyranosides, containing a leaving group at the anomeric
position from which the oxocarbenium ion could be generated)22 gave the desired (1→2)-, (1→3)-
and (1→4)-linked disaccharides. The subsequent steps included the Staudinger reduction of the
azide moiety followed by amide coupling with the corresponding (aryl)acetic acid to grant access to
6
ACS Paragon Plus Environment

Page 7 of 23
Journal of the American Chemical Society
model sytems type I, 1-9. In a similar manner, 4,6-O-benzylidenation of thioglycosides (gluco-,
manno- and 2-deoxy-gluco-) with the appropriate arylaldehyde dimethylacetal, followed by the
analogous Staudinger reduction/amide coupling sequence provided the target constrained pyranose
type II systems, 10-19. A detailed description of the synthetic protocols, together with the
characterization of products and intermediates, is included in the supplementary material.
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
Scheme 1. General synthetic route for the model systems I and II.
c) Reactivity evaluation of family I systems: methanolysis reactions.- Once the synthesis of the
target models was successfully accomplished, we proceeded to verify the formation of the anticipated
carbohydrate/aromatic complexes in solution. Satisfactorily, obtained results for 1-9 were in
accordance with the design principles previously outlined. As an example, the NMR data for 1 were
consistent with the formation of a well-defined complex involving the pyrene platform and the -
face of the reactive unit. Thus, HSQC signals for Gal H1, H3 and H5 appear extremely up-field shifted
1
in the H dimension (Figure 3a, left and S1), highlighting their direct involvement in strong CH/
bonds with the aromatic unit. To reveal the impact of such interaction pattern on the pyranose
reactivity, the methanolysis of compound 1, and the corresponding references 2 and 4, was monitored
o
using 2D-HSQC experiments in deuterated chloroform at 20 C (Figure 3a-c). Interestingly, the
addition of 1 l of triflic acid to a solution of disaccharide 1 in the absence of acceptor alcohol led to
an extreme broadening of the NMR signals, consistent with the presence of several species in the
intermediate exchange regime on the NMR chemical shift time-scale. Quenching of this sample with
5 L of CD3OD yielded a major product, with no carbohydrate/aromatic stacking, as judged from the
observed chemical shift data. This derivative proved to be extremely unstable, rendering all isolation
efforts fruitless. Fortunately, both NMR and mass spectrometry analyses of the reaction mixture
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 23
provided key structural information, strongly suggesting the formation of the highly strained
macrocyclic C-glycoside shown in Figure 3a (see also Figure S2). Thus, the Gal anomeric carbon
(with a broad peak at C80 ppm) presented non-ambiguous HMBC and NOE correlations with the
1
2
3
4
5
6
7
3
4
pyrene unit. Moreover, the measured JHH coupling constants were still fully consistent with a C1
conformation for both pyranose units. More intriguingly, the final C-glycoside exhibits a -
8
9
configuration as shown by its 3JH1-H2 coupling constant (9.5 Hz). This was totally unanticipated from
the structure of the original carbohydrate/pyrene complex (with the aromatic unit stacked on the -
face of the galactose), suggesting that formation of the glycosyl cation is followed by a significant
conformational rearrangement (Figure 3a).
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.-Family I. a) Top.- Chemical evolution of 1 upon treatment with triflic acid, both in the
absence and presence of 1% CD3OD. Bottom.- Key region of the HSQC NMR spectra recorded for
1, along with the proposed C-glycosidic structure. b) HSQC spectra of a 1(red)/4(black) mixture
before and after triflic acid addition.  values (ppm) for representative signals are indicated. c)
Reactivity assays with a 1(blue)/2(green)/4(red) equimolecular mixture in CDCl3 (1% CD3OD) at 25
oC. Top.- Initial data sets (anomeric region). Middle.- Insets A-D (highlighted with yellow boxes
above) at different times. Bottom.- Reaction profiles derived from the integration of anomeric cross-
peaks. d) Kinetic constant ratios measured for 1-3, 6 and 8 with respect to the corresponding
reference derivatives. The resulting anomeric : ratios are also shown.
In contrast, the addition of 4 L of triflic acid in the presence of 1% CD3OD triggers a smooth
methanolysis process, which can be conveniently monitored by NMR. Interestingly, this protocol
leads to partial protonation of the disaccharides, with opposite behaviors for the pyrene-bearing
compounds and the reference derivatives. As an example, Figure 3b displays the HSQC spectra
acquired for an equimolecular mixture of 1 (red) and 4 (black) in the absence and presence of TfOH
8
ACS Paragon Plus Environment

Page 9 of 23
Journal of the American Chemical Society
(1 L). It can be observed that the acidic conditions induce slight down-field shifts in the NMR signals
of model 4. On the contrary, for 1 those protons already involved in CH/ interactions with the
aromatic platform are shielded even further. As a consequence, the chemical shift perturbations
promoted by aromatic stacking on the reactive pyranose unit (with respect to reference compound 4,
see Figure 3b) are significantly enhanced, reaching unusual values even larger than > 3 ppm. In
conclusion, partial protonation of the reactive pyranose turns the original CH/ interactions into a
much stronger cation/interaction leading to a significant stabilization of the carbohydrate/aromatic
complex. A similar behavior was observed for all the stacking complexes tested regardless of the
pyranose interacting face ( or , as gathered in Figure 3b and S3-S6.
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
The chemical evolution of an equimolecular mixture of 1, 2 and 4, upon the addition of triflic acid
is represented in Figure 3c. The integration of the relevant reagent and product cross-peaks with time
allowed to build a reaction time-course for each of the three derivatives assayed. According to these
data, reference 4, for which no aromatic unit is present, rapidly evolves to yield a mixture of methyl
- and -O-glycosides with a half-reaction time (t1/2) ca. 15 min. A slower methanolysis process is
apparent for 1 and 2, containing a pyrenyl and naphthyl residue (t1/2 values of 353 and 140 minutes,
respectively). Interestingly, in these latter cases a single  methyl glycoside product was detected.
Reaction time-courses were translated into kinetic constants assuming pseudo-first order conditions.
Then, the influence of aromatic stacking on reactivity was expressed in terms of kinetic constant
ratios (kreference/k+aromatic). Employing this general protocol, the rest of family I model systems were
tested. The obtained results are summarized in Figure 3d and reveal a consistent inhibition of the
glycoside reactivity by the presence of aromatic stacking. Indeed, these reactivity falls are in the 2-
to 20-fold range, depending on the strength and geometry of the original carbohydrate/aromatic
interaction. Significantly, this effect is, in some cases, accompanied with an increase in the
stereoselectivity of the process, which is particularly clear for the strongest/less reactive complexes.
For example, methanolysis of the reference compound 4 produces an initial 2:1 : mixture that
equilibrates to the final 1:2 ratio observed under these acidic reaction conditions. On the contrary,
donor 1, which is involved in a strong stacking complex with the pyrene unit, yields only the retention
-product. This observation is in accordance with the behavior recently described for an engineered
SNi-synthase (Figure 1b)17 and might also point to an SNi mechanism for this particular substitution
process.
52
53
54
55
56
57
58
59
60
In summary, according to the obtained data, the aromatic systems in family I models exert a
preferential stabilization of the ground state and partially charged pyranoses over the oxocarbenium-
like transition state, leading to a net increase in the global activation energy of the process. This
conclusion is consistent with the apparent strength of the observed CH/ contacts, specially under
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 23
acidic conditions (as judged by values) and suggests that the geometric and dynamic properties of
the initial complex are of critical importance for reactivity.
1
2
3
4
5
6
7
8
9
d) Reactivity evaluation of family II systems: low-temperature glycosylations. - With these
results in hand, we decided to explore the influence of the pyrene unit in a rather different context
(Figures S7-S21). As previously mentioned, type II molecules combine more dynamic
carbohydrate/aromatic interactions with a conformationally constrained reactive pyranose. Indeed,
the transient dynamic character of the CH/ bonds is now reflected in the reduced chemical shift
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
o
perturbations (< 2 ppm for 10, 11 and 16 even at temperatures below -50 C) promoted by the
aromatic systems at the sugar unit.
Reactivity assays were individually performed for 10-19 at temperatures in the -65/-50 oC range
and monitored by 1D and/or 2D-HSQC NMR experiments, employing regular or 13C-labeled
molecules (Figures 4-6).22 As internal standard for integrations, an equimolecular amount of 4,4,5,5-
tetramethyl-2-(1-naphthalenyl)-1,3-dioxolane was included in the reaction mixtures.23 Considering
the increased complexity of these multi-component reactions, a minimum of three runs, under
identical reaction conditions, were completed for every case. Pairwise competition experiments were
also carried out for qualitative comparison purposes. In all cases, glycosyl donors were activated with
triflic anhydride in the presence of 5-8 equivalents of TFE depending on the model compound, as
described in the experimental section.
The mannose case.- The major conformation of 10 is represented in Figure 4a and displays a
preferred stacking geometry that involves the -face of the mannose unit, as shown by the chemical
shift data (Figure S8). The glycosylation reactions furnished a mixture of products (20and
with a 70-80% overall yield (Figure S9), slightly lower than those obtained from the reference
compounds 12 and 13 (75%-85%). Unexpectedly, the stereoselectivity of the process vanishes in the
presence of the pyrene system (from 1:4 to 1:1).
NMR experiments performed both with individual donors and equimolecular 10/13 13C-labeled
mixtures allowed us to estimate the relative 20 (/ and 22 (/ concentrations at short reaction
times (5-10 min). The obtained values, which are proportional to the relative formation rates, are
represented as a bar chart in Figure 4c (see also Figure S10). According to these data, the aromatic
stacking has an opposite effect on the individual product formation rates: two-fold increase for the -
anomer and 2-fold decrease for the -derivative, consistent with the observed degradation of the
stereoselectivity. Interestingly, even in the presence of five equivalents of TFE, the -glycosyl
triflates formed from both 10 and 13 were detected throughout the reaction and displayed comparable
steady concentrations. This behavior reflects the higher nucleophilic character of the triflate anion
with respect to TFE or to the aromatic unit itself.8,24 It should be noted that the formation the of an -
triflate from 10 requires a rearrangement of the initial carbohydrate/aromatic complex, which is
10
ACS Paragon Plus Environment

Page 11 of 23
Journal of the American Chemical Society
greatly facilitated by the adaptable nature of the designed scaffold. Indeed, the pyrene unit swings
from the mannose -face to a lateral disposition, establishing CH/ bonds with the hydroxymethyl
moiety, whose protons are now shielded up to = -1.7 ppm. These intermediates were also
characterized by NMR employing a pre-activation protocol, as shown in Figure S8. A similar
outcome was obtained employing 10/12 mixtures (Figure S11). These unexpected results prompted
us to further interrogate alternative aromatic/donor complexes also belonging to family II.
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.- Family II. a) The mannose case. Glycosylation of 10 with TFE. Yields and
stereoselectivities obtained from 10, 12 and 13 are shown below. Representative HSQC regions of
the NMR-based competition experiment (-60 oC) carried out with an equimolecular 13C-labeled 10/13
mixture at three different reaction times are displayed. b) The glucose case. Glycosylation of 11 with
TFE. Yields and stereoselectivities obtained from 11, 14 and 15 are shown below. Representative
HSQC regions of glycosylations (-60 oC) performed with13C-labeled 11 (top), 14 (middle) and a 11/14
mixture (bottom) at three different reaction times are displayed. c) Bar charts represent the observed
relative populations for the corresponding products at short reaction times.
The glucose case.- -glucopyranosyl sulfoxide 11 exhibits a highly dynamic conformational
1
behavior, as deduced from the significant chemical shift perturbations observed for the H NMR
resonances for both pyranose faces. MD simulations also supported the co-existence of - and -
stacking complexes in equilibrium (Figures S1 and S12-S14). Glycosylations with donors 11, 14 and
15 (see also Figures S15-S17) revealed a significant effect of the pyrene unit on the observed
stereoselectivity, which decreased from >20:1 to 2:1, although with an overall improvement of the
reaction yield. The use of 13C-labeled donor molecules allowed to pinpoint further reactivity
differences between 11 and the reference derivatives 14 and 15. Thus, for the pyrene-stacked donor
11, the process proceeds with a substantial accumulation of three transient activated species: an -
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 23
glycosyl triflate and two unidentified -species, marked as S+ in Figure 4b. Their 13C NMR chemical
1
2
3
1
shifts and heteronuclear JHC constant values (168 Hz) strongly point at sulfonium species. On the
4
5
6
7
8
9
contrary, analogous intermediates were barely detectable in the absence of a stacked aromatic unit
(14 and 15). Of note, the triflate formed from 11 presents clear CH/ interactions with the pyrene
platform, which in this case are mediated by the pyranose -face CH groups. This transient triflate
species was also prepared following a pre-activation protocol, which allowed its unambiguous
characterization (Figure S14). Regarding the reaction kinetics, both individual assays and competition
experiments revealed that the formation rate of the major -stereoisomer is marginally increased by
the presence of the pyrene unit (Figure 4b, 23 vs 24), while that of the minor -anomer is much
more enhanced, as represented in Figure 4c (bar chart). This fact accounts for the observed
stereoselectivity.
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
As a final point, the larger accumulation of intermediate species formed from 11 indicates that
their formation and/or consumption rates must be significantly altered by the existence of the
carbohydrate/aromatic stacking. Fittingly, pre-activation experiments (Figure S18-S19) confirmed
that triflate formation is significantly accelerated in the presence of the pyrene unit, strongly
suggesting the presence of a transition state with a significant cationic character for this step.
The 2-deoxy-glucose case.- Among common glycoside donors, 2-deoxy-derivatives, lacking an
electron-withdrawing substituent at the neighboring position to the anomeric center, present the
highest tendency to evolve through dissociative, more cationic transition states.25 This fact ensures a
SN1-like behavior even when nucleophilic alcohols are considered. Accordingly, for the reactivity
analysis of these models, monofluoroethanol (MFE) was incorporated as an additional acceptor.

NMR reactivity experiments performed with C-labeled models revealed, in all cases, a clean
evolution from reagents to products, with no detectable intermediates, which is consistent with the
higher reactivity expected for any potential transient species involved. Glycosylation yields, either
with TFE or with MFE, were improved in the presence of the pyrene platform (Figure 5a and S20).
On the contrary, the / stereoselectivity of these processes, already marginal for the reference
compounds (2:1 and 1:2 with TFE and MFE, respectively) was further reduced in these cases.
Interestingly, competition experiments monitored by HSQC spectra showed that this effect relies on
a substantial acceleration in the production of the minor anomer, regardless of their stereochemistry
or the alcohol employed in each case. That is, while the presence of the pyrene moiety significantly
accelerates the formation of the -linked product with TFE (26 vs 28), it promotes the same effect
but for the -anomer with MFE (27 vs 29Figure 5a).
1
Regarding glycosylations with TFE, the presence of well resolved H NMR signals for reagents
and products prompted us to carry out additional experiments to monitor the individual reaction
kinetics through sequential low-temperature 1D-NMR experiments. The representative time courses
12
ACS Paragon Plus Environment

Page 13 of 23
Journal of the American Chemical Society
independently obtained from the evolution of 16 and 17 are shown in Figure 5c, along with the
relative initial reaction rates. The analysis of these data revealed a modest 2-fold increase for the
formation rate of the -anomer (26), and a concomitant 4-fold increase for the -anomer (26, thus
equating both formation rates and explaining the observed 1:1 stereoselectivity. Overall, the
glycosylation reaction is 3-fold faster ( +  products) in the presence of the pyrene platform (Figure
5b, bar chart).
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.- Family II. The 2-deoxy-glucose case. a) Reactions with 16 (up) and 17 (down) at -65 oC
employing TFE or MFE. Yields and stereoselectivities are shown. b) Competition experiments with
equimolecular 13C-labeled 16/17 mixtures with TFE and MFE monitored by HSQCs. Data sets
measured at different reaction times are displayed. Intensity ratios for the reaction products are
indicated. c) Reactions performed either with 16 (left) or 17 (right) employing TFE, monitored
o
through sequential 1D-NMR experiments at -65 C. Kinetics of formation for the corresponding 
and  products, derived from two independent experiments, are superimposed above. Bar chart
represents the relative populations for the corresponding products at short reaction times.
Encouraged by these promising results, we decided to push these latter models one step further
into the SN1 realm by the incorporation of a second pyrene moiety. To our delight, unprecedented
shieldings (with 1H NMR signals appearing with negative chemical shifts up to = -1.3 ppm) were
induced by this optimized covalent scaffold for 18 and its corresponding glycosylation products 30
/ (Figure 6a). Of note, MD simulations showed that this effect results from the additive influence
of both aromatic platforms, but does not imply a significantly tighter, more rigid geometry of the
CH/ complex.
Reactions of both 18 and the reference derivative 19 with TFE yielded the final glycosylation
products (30 / and 31 / ) with identical yields (in the 70-80% range) and stereoselectivities (1:2).
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 23
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 6.- Family II. The 2-deoxy-glucose case. The influence of double stacking. a) Pyranose
chemical shift perturbations measured for 18 and 30 at -65 C. CPK models for relevant
o
conformations are shown below, together with the assigned COSY NMR experiments. b) The
reactions of 18 (left, red) and 19 (right, green) with TFE were monitored by 1D-NMR experiments at
-65 oC. Yields and stereoselectivities are displayed above. Formation kinetics of the corresponding
 and  products are shown on the right. Bar chart represents the relative populations deduced for
the glycosylation products at short reaction times. c) Assays carried out in CD2Cl2. For 18 the
glycosylation was already complete in the first 1H-NMR spectrum. Yields and stereoselectivities are
indicated below. d) Reactions with 18 and 19 employing a synthetically relevant carbohydrate
acceptor. Yields and stereoselectivities are displayed below. Bar chart represents the relative
populations deduced for the glycosylation products at short reaction times.
Satisfactorily, a careful monitorization of the glycosylation progress through sequential low-
temperature 1D-NMR experiments was greatly facilitated by the presence of several proton signals
with negative  values. Product conversions at the shortest reaction time tested (already showing a
40% evolution for 18) were consistent with a 7-fold increase in the glycosylation rates for 30- and
30-anomers, as displayed in Figure 6b (see also Figure S21). Moreover, these effects seem further
noticeable in the presence of extremely poor O-nucleophiles, such as hexafluoroisopropanol (HFIP)
(see Figure S22).
14
ACS Paragon Plus Environment

Page 15 of 23
Journal of the American Chemical Society
As expected, increasing the solvent dielectric from chloroform (=. 4.81) to dichloromethane (=
1
2
8.93) translates in faster glycosylation rates. However, reference model 19 still reacts significantly
slower than the sandwiched donor 18 (whose reaction is now too fast to be followed by NMR. Figure
6c and S23). Of note, the acceleration observed for 18 is accompanied by a significant increase in the
reaction yield.
3
4
5
6
7
8
9
Finally, models 18 and 19 were assayed in the context of a more conventional glycosylation
reaction, employing the galactose acceptor shown in figure 6d (see also Figure S24). These
experiments showed that the favorable influence exerted by aromatic platforms on the reaction rates
are not limited to small, weakly nucleophilic alcohols but, to some extent, holds also for carbohydrate
acceptors used in a slight stoichiometric excess. Interestingly, while reference derivative 19 reacts
with significant stereoselectivity to yield a major -glycosidation product, sandwiched donor 18
furnishes an almost equimolecular : mixture. According to these results, the employed galactose
acceptor would react with reference 19 through a more associative process, probably involving an -
triflate intermediate, and with 18 in a more dissociative SN1-like manner.
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
In summary, the combination of a 2-deoxy-glucopyranose donor with the appropriately
positioned aromatic platforms has permitted, for the first time, the observation of glycosylation rate
enhancements associated to carbohydrate/aromatic interactions.
DISCUSSION
The results described in the previous sections for families I and II reveal a stricking behavior,
where the pyrene unit exerts a gradient of effects on the pyranose reactivity, ranging from significant
inhibitions to moderate rate enhancements, as summarized in Figure 7. These data can be globally
understood by considering the major driving forces that stabilize aromatic/carbohydrate complexes
and how they evolve throughout the chemical reactions. On the one hand, the development of positive
charge at the anomeric position reinforces attractive electrostatic interactions with the aromatic
electronic quadrupole. This effect would be expected to be most significant under low dielectric
environments and for highly dissociative SN1-like processes. On the other hand, as the anomeric
leaving group departs, the reactive pyranose distorts. In the simplest scenario, anomeric sp3 to sp2
rehybridation would imply that shape complementarity between the pyranose and the aromatic
platform, optimal for a 4C1 chair, is decreased, determining the disruption of at least one CH/ bond
(the one involving the anomeric CH). Furthermore, for unconstrained pyranoses, more dramatic
conformational changes could take place, thus leading to an overall deterioration of the van der Waals
interactions with the aromatic platform. According to this view, the net influence of aromatic stacking
on the pyranose reactivity depends on a delicate balance between two conflicting energy
contributions: electrostatic and van der Waals, favoring and opposing chemical evolution,
respectively.
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 23
4
Regarding family I models, they encompass an unconstrained reactive C1 pyranose
participating in strong CH/ interactions with the aromatic unit, in both the ground and activated
(protonated) states. The experimentally observed decrease in reactivity for the carbohydrate/aromatic
complexes indicates that the weakening of the van der Waals interactions at the transition state is not
compensated by electrostatic forces. Indeed, CH/ and electrostatic contributions seem to be much
more stabilizing for the activated species than for the distorted transition state, totalling in a net
increase in the activation energy. This view is further supported by MD simulations (Figure 7a),
which predict that a protonation of the anomeric substituent reinforces the CH/ bonds, leading to
well-defined interaction pose, in agreement with the experimental chemical shift perturbations
(Figure 2b). Significantly, the departure of the leaving group increases the internal
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 7.- a) Summary of the results obtained for family I (left, yellow) and family II (right, cyan).
The influence of the aromatic platform is expressed as rate inhibition (downwards) or enhancement
(upwards) factors. Numeric values are shown in red and blue, respectively. Views of the
conformational ensembles for the ground, activated and glycosyl cation states derived from solvated
MD simulations (explicit CHCl3) with a modified version of model 6 (bottom-right corner) and with
18 (top-left corner). b) Free energy penalty associated with the protonation of models a, d, f (top to
bottom) and the subsequent glycosyl cation formation, in the presence (red) and absence (blue) of a
complexed pyrene platform calculated by quantum mechanics with PCM/M06-2X/6-31g(d,p) in
chloroform. The geometry of the complex ( or ) is also indicated. Net stabilizations promoted by
the pyrene unit are expressed as G (kcal/mol). The most stable conformations of the glycosyl
cations were considered for the free and complexed states.
mobility of the complex and frequent disruptions of the carbohydrate/aromatic contacts take place,
due to the lack of an appropriate shape complementarity between the aromatic platform and the newly
formed glycosyl cation. Indeed, the presence of axially oriented substituents in several of the
conformational states of these species is not compatible with the geometric requirements for
16
ACS Paragon Plus Environment

Page 17 of 23
Journal of the American Chemical Society
establishing optimal CH/interactions. On the contrary, for family II models, the presence of weaker,
more dynamic CH/ bonds in the ground state of the constrained pyranose analogues determine their
distinct behavior. Three common trends are apparent from the obtained data: first, the larger the SN1
character of the reaction, the more favorable the influence exerted by the aromatic platform.
Accordingly, major enhancements in the glycosylation rates are observed only for 2-deoxy-gluco-
donors. Second, the conformational restriction imposed on the donor pyranose allows for a more
optimized glycosyl cation/aromatic complex, as deduced by the analysis of the MD simulations
(Figure 7a). Three, despite the improved shape complementarity between the pyrene unit and the
nascent glycosyl oxocarbenium intermediate/transition state, the interactions are still highly mobile.
Thus, the aromatic system provides a nurturing, more polar local environment for the developing
pyranose charge, rather than fixing a preferred geometry of the complex, which translates into rate
enhancements that are not accompanied by any particular stereocontrol. For example, for 18 no
variation in the stereoselectivity of the reaction was detected with respect to reference model 19 when
TFE was used as an acceptor (Figure 6a,b). On the contrary, the bulkier, more nucleophilic diacetone
galactose yielded an opposite outcome (Figure 6d and S24): a preferential increase in the formation
rate of the minor anomer, which determines a reduction in the reaction stereoselectivity. A similar
behavior was also observed for donors 10, 11 and 16, which overall can be interpreted in terms of a
shift in the reaction mechanism toward a more dissociative SN1-like process. Interestingly, rate
accelerations do come with moderate-to-significant improvements in the reaction yields for most
analyzed derivatives (for example, 11 vs 14 or 15, or models 16 vs 17). This effect is clearly
observable in the reaction of pair 18/19 with TFE in dichloromethane (Figure 6c and S23) and
suggests that the aromatic platforms exert a certain protective influence, limiting the negative impact
of undesired parasitic reactions on the glycosylation outcome. Consistently, we have also shown that
these trends are maintained in other solvents, or mixtures of solvents with ranging polarities (Figure
S23).
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
In summary, while the presence of aromatic quadrupole platforms facilitates the formation of
glycosyl cations by electrostatic forces, pyranose distortion is detrimental for the establishment of
proper CH/ contacts. Significantly, quantum mechanics calculations (PCM/M06-2X/6-31g(d,p) in
chloroform) performed with simplified 1-O-acetyl model glycosides (Figure 7b), either constrained
(f) or unconstrained (a and d), predict an enhanced stabilization of the glycosyl cation by the pyrene
unit in the former complex f (G of 5 kcal/mol vs 2.9-3.0 kcal/mol), in accordance with our
experimental observations. Nevertheless, the theoretical stabilizations are, in all cases, overestimated,
which might reflect the limitations inherent to these theoretical methods (implicit solvation,
simplified intermolecular model complexes, absence of triflate counterions).
Overall, these data show that aromatic stacking promotes and facilitates SN1 anomeric
substitutions with all the expected consequences from a synthetic perspective. The fundamental
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 23
information derived from our studies could now be extrapolated to a more general chemical or
enzymatic context. Hence, while measured accelerations are moderate, much larger enhancements
would conceivably result from the additive influence of several, properly positioned aromatic
platforms. According to our data, geometrical considerations are key to this goal. In particular,
conventional parallel stacking leads to conflicting electrostatic and van der Waals contributions into
the activation energy, which can limit or even, overturn the favorable influence exerted by the
aromatic platform. Alternative stacking geometries, either lateral or parallel-shifted, could alleviate
this conflict. This view is also supported by the unusual stacking modes observed in the active site of
several glycosidases in complex with ligands or inhibitors (pdb codes 3WY4 or 2J77. Figure S25),
which usually involve the formation of only one or two CH/bonds per aromatic interaction, in
contrast with the typical parallel stacking geometries (fully involving either the  or - face of the
interacting pyranose) that take place in lectins and CBMs. This phenotypic recurring feature probably
reflects the need to maintain the sugar ring relatively unconstrained by the aromatic residue, allowing
the penalty-free pyranose distortions required for the reaction to proceed. Interestingly, quantum
mechanics calculations performed on a simplified inhibitor/active site complex strongly suggest that
interacting aromatic side-chains also contribute to stabilize the substrate positive charge, while
keeping the active site free of water molecules (Figure S26).
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
It is worth mentioning that conflicting electrostatic and dispersive contributions to the
activation energy could also be tilted in favor of the former by the incorporation of electron-donating
substituents on the aromatic platform, which provides an alternative strategy to further reduce the
transition state energy. Accordingly, previous studies by our group have shown that this modification
indeed enhances carbohydrate/aromatic electrostatic forces, while having a lesser influence on the
dispersive component of the stacking complexes.5a Carefully designed systems based on these
concepts might eventually lead to the development of synthetic carbohydrate receptors with catalytic
properties or engineered enzymes carrying unnatural aminoacids.
In view of the results presented in this work, aromatic moieties can indeed contribute to the
catalysis of highly dissociative anomeric substitutions, provided that the electrostatics outcompete
the van der Waals contributions to the activation energy.
CONCLUSIONS
While the participation of aromatic platforms in the molecular recognition of carbohydrates
is undeniable, its potential involvement in catalysis remains an open question. In the work presented
herein, we have tackled this issue by answering a simple question of fundamental relevance in
glycoscience: can glycosyl cations be stabilized by aromatic platforms? Accordingly, we have
dissected, in a systematic and extensive manner, the influence exerted by aromatic stacking on
glycosylation or glycoside solvolysis reactions, whereby the potential of aromatic platforms as key
18
ACS Paragon Plus Environment

Page 19 of 23
Journal of the American Chemical Society
elements to sense and stabilize transient positive charge development was evaluated. The results
derived from our study provide an affirmative answer to the aforementioned question: aromatic
platforms can indeed accelerate and improve SN1-like glycosylations, presumably by stabilizing
positively charged transition states through cation/interactions. The obtained data also lent some
meaningful insight into the nature of glycosyl cation/aromatic interaction and its influence on
glycosylations. First, the glycosylations with 2-deoxy-donors, even in highly apolar environments
and at very low temperatures, must involve highly cationic transition states, whereas glucose and
mannose derivatives seem to proceed through either close ion pairs or highly reactive covalent
intermediates. Secondly, glycosyl cation/aromatic interactions tend to be highly dynamic, thus
reflecting the lack of a proper shape complementarity between both species. As a consequence, for
flexible systems, they can exert a negligible influence on the stereoselectivity of the process. For
more associative processes, the ability of aromatic units to selectively hinder bi-molecular
substitutions, while still promoting the development of positive charge at the anomeric center, seems
to determine a shift in the reaction mechanism towards the SN1, which has a concomitant negative
impact on the overall stereoselectivity. This effect is especially clear when using bulky/nucleophilic
acceptors and it is maintained for alternative solvents.
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
Finally, glycoside/aromatic complexes stabilized by strong CH/interactions already present
in the ground state can indeed undergo SN1-like reactions, such as solvolysis, with a greatly improved
stereoselective outcome. Yet, this effect is accompanied by an increase in the activation energy of the
process, which translates into a slower reaction rate, as exemplified by the methanolysis of model
system 1.
To the best of our knowledge, this represents the first experimental study on the influence of
cation/interactions in the formation and cleavage of glycosidic linkages, whose potential
implications branch out into general catalysis and CAZyme engineering.
METHODS
A detailed description of the experimental and synthetic protocols together with the
characterization of products and intermediates is included in the supplementary material.
SUPPLEMENTARY MATERIALS
A detailed description of the experimental methods and synthetic protocols together with the
characterization of products and intermediates. Figures S1-S26 showing details of the NMR and
reactivity experiments.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 23
ACKNOWLEDGMENTS
1
2
This investigation was supported by research grants of the Spanish “Plan Nacional”
CTQ2016-79255-P/CTQ2015-66702-R and the Mizutani Foundation for Glycoscience 17-0045.
A.G.S acknowledges the Ministerio de Ciencia, Innovación y Universidades for a Juan de la Cierva
incorporation contract. L. M-J thanks Ministerio de Economia y Competitividad for a FPI contract
(BES-2014-070232).
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
REFERENCES
1.- a) Asensio, J.L.; Ardá, A.; Cañada, F.J.; Jiménez-Barbero, J. Carbohydrate-Aromatic
Interactions. Acc. Chem. Res. 2013, 46, 946-954. b) Terraneo, G.; Potenza, D.; Canales, A.; Jiménez-
Barbero, J.; Baldridge, K.K.; Bernardi, A. A Simple Model System for the Study of
Carbohydrate−Aromatic Interactions. J. Am. Chem. Soc. 2007, 129, 2890-2900. (c) Fernández-
Alonso, M.C.; Cañada, F.J.; Jimenez-Barbero, J.; Cuevas, G. Molecular Recognition of Saccharides
by Proteins. Insights on the Origin of the Carbohydrate−Aromatic Interactions. J. Am. Chem. Soc.
2005, 127, 7379-7386. (d) Ramírez-Gualito, K.; Alonso-Ríos, R.; Quiroz-García, B.; Rojas-Aguilar,
A.; Díaz, D.; Jiménez-Barbero, J.; Cuevas, G. Enthalpic Nature of the CH/π Interaction Involved in
the Recognition of Carbohydrates by Aromatic Compounds, Confirmed by a Novel Interplay of
NMR, Calorimetry, and Theoretical Calculations. J. Am. Chem. Soc. 2009, 131, 18129-18138. (e)
Chavez, M.; Andreu, C.; Vidal, P.; Aboitiz, N.; Freire, F.; Groves, P.; Asensio, J.L.; Asensio, G.;
Muraki, M.; Cañada, F.J.; Jimenez-Barbero, J. On the Importance of Carbohydrate–Aromatic
Interactions for the Molecular Recognition of Oligosaccharides by Proteins: NMR Studies of the
Structure and Binding Affinity of AcAMP2‐like Peptides with Non‐Natural Naphthyl and
Fluoroaromatic Residues. Chem.-Eur. J. 2005, 11,7060-7074.
2.- a) Phillips, D. C. The hen hegg-white lysozyme molecule. Proc. Nat. Acad. Sci. USA 1967,
57, 483-495. b) Glickson, J.D.; Phillips, W.D.; Rupley, J.A. Proton magnetic resonance study of the
indole NH resonances of lysozyme. Assignment, deuterium exchange kinetics, and inhibitor binding.
J. Am. Chem. Soc. 1971, 93, 4031-4035. c) Quiocho, F.A. Carbohydrate-binding proteins-tertiary
structures and protein-sugar interactions. Ann. Rev. Biochem. 1986, 55, 287-315. d) Lutteke, T.;
Frank, M.; von der Lieth, C.W. Carbohydrate Structure Suite (CSS): analysis of carbohydrate 3D
structures derived from the PDB. Nucl. Acids Res. 2005, 33, D242-D246.
3.- a) Tromans, R.A.; Carter, T.S.; Cabanne, L.; Crump, M.P.; Li, H.; Matlock, J.V.; Orchard,
M.G.; Davis, A.P. A biomimetic receptor for glucose. Nat. Chem. 2019, 11, 52-56. b) Rios, P.;
Mooibroek, T.J.; Carter, T.S.; William, C.; Wilson, M.R.; Crump, M.P.; Davis, A. P. Enantioselective
carbohydrate recognition by synthetic lectins in water. Chem. Sci. 2017, 5, 4056-4061. c) Rios, P.;
Carter, T.S.; Mooibroek, T.J.; Crump, M.P.; Lisbjerg, M.; Pittelkow, M.; Superkar, N.T.; Boons, G.J.;
Davis, A. P. Synthetic Receptors for the High-Affinity Recognition of O-GlcNAc Derivatives. Angew.
Chem. Int. Ed. 2016, 55, 3387-3392. d) Mooibroek, T.J.; Casas-Solvas, J.M.; Harniman, R.L.;
Renney, C.M.; Carter, T.S.; Crump, M.P.; Davis, A.P. A threading receptor for polysaccharides. Nat.
Chem. 2016, 8, 69-74. e) Ke, C.; Destecroix, H.; Crump, M.P. Davis, A.P. A simple and accessible
synthetic lectin for glucose recognition and sensing. Nat. Chem. 2012, 4, 718-723. f) Davis, A.P.
Supramolecular chemistry: Sticking to sugars. Nature 2010, 464, 169-170. g) Ferrand, Y.; Crump,
M.P.; Davis, A.P. A synthetic lectin analog for biomimetic disaccharide recognition. Science 2007,
318, 619-622.
4.- a) Hudson, K.L.; Bartlett, G.J.; Diehl, R.C.; Agirre, J.; Gallagher, T.; Kiessling, L.L.
Carbohydrate–Aromatic Interactions in Proteins. J. Am. Chem. Soc. 2015, 137, 15152-15116. b)
Laughrey, Z.R.; Kiehna, S.E.; Riemen, A.J.; Waters, M.L. Carbohydrate−π Interactions: What Are
They Worth?. J. Am. Chem. Soc., 2008, 130, 14625-14633. c) Kiehna, S.E.; Laughrey, Z.R.; Waters,
M.L. Evaluation of a carbohydrate-π interaction in a peptide model system. Chem. Commun., 2007,
39, 4026-4028. d) Che-Hsiung, H.; Park, S.; Mortenson, D.E.; Foley, B.L.; Wang, X.; Woods, R.J.;
Case, D.A.; Powers, E.T.; Wong, C.H.; Dyson, H.J.; Kelly, J.W. The Dependence of Carbohydrate-
20
ACS Paragon Plus Environment

Page 21 of 23
Journal of the American Chemical Society
Aromatic Interaction Strengths on the Structure of the Carbohydrate. J. Am. Chem. Soc. 2016, 138,
7636-7648.
1
2
3
4
5
6
7
8
5.- a) Jiménez-Moreno, E.; Jiménez-Osés, G.; Gómez, A.M.; Santana, A.G.; Corzana, F.;
Bastida, A.; Jiménez-Barbero, J.; Asensio, J.L. A thorough experimental study of CH/π interactions
in water: quantitative structure-stability relationships for carbohydrate/aromatic complexes.
Chemical Science 2015, 6, 11, 6076-6085. b) Jiménez-Moreno, E.; Gómez, A.M.; Bastida, A.;
Corzana, F.; Jiménez-Oses, G.; Jiménez-Barbero, J.; Asensio, J.L. Modulating weak interactions for
molecular recognition: A dynamic combinatorial analysis for assessing the contribution of
electrostatics to the stability of CH-π bonds in water. Angew. Chem. Int. Edit. 2015, 54, 4344-4348.
c) Santana, A.G.; Jiménez-Moreno, E.; Gómez, A.M.; Corzana, F.; González, C.; Jiménez-Oses, G.;
Jiménez-Barbero, J.; Asensio, J.L. A dynamic combinatorial approach for the analysis of weak
carbohydrate/aromatic complexes: Dissecting facial selectivity in CH/π stacking interactions. J. Am.
Chem. Soc. 2013, 135, 3347-3350.
6.- a) Ma, J.C.; Dougherty, D.A. The cation-π interaction. Chem. Rev. 1997, 97, 1303-1324.
b) Dougherty, D.A. Cation-π interactions in chemistry and biology: A new view of benzene, Phe, Tyr,
and Trp. Science 1996, 271, 163-168.
7.- Neel, A.J.; Hilton, M.J.; Sigman, M.S.; Toste, F.D. Exploiting non-covalent π interactions
for catalyst design. Nature 2017, 543, 636-646.
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
8.- a) Adero, P.O.; Amaraseka, H.; Wen, P.; Bohe, L.; Crich, D. The Experimental Evidence
in Support of Glycosylation Mechanisms at the SN1-SN2 Interface. Chem. Rev. 2018, 118, 8242-
8284. b) Adero, P.O.; Furukawa, T.; Huang, M.; Mukherjee, D.; Retailleau, P.; Bohé, L.; Crich, D.
Cation clock reactions for the determination of relative reaction kinetics in glycosylation reactions:
Applications to gluco- and mannopyranosyl sulfoxide and trichloroacetimidate type donors. J. Am.
Chem. Soc. 2015, 137, 10336-10345. c) Huang, M.; Retailleau, P.; Bohé, L.; Crich, D. Cation clock
permits distinction between the mechanisms of α- And β-O- and β-C-glycosylation in the
mannopyranose series: Evidence for the existence of a mannopyranosyl oxocarbenium ion. J. Am.
Chem. Soc. 2012, 134, 14746-14749. d) Huang, M.; Garrett, G.E.; Birlirakis, N.; Bohé, L.; Pratt,
13
D.A.; Crich, D. Dissecting the mechanisms of a class of chemical glycosylation using primary C
kinetic isotope effects. Nat. Chem. 2012, 4, 663-667. e) Crich, D. Mechanism of a chemical
glycosylation reaction. Acc. Chem. Res. 2010, 43, 1144-1153. f) Crich, D.; Chandrasekera, N.S.
Mechanism of 4,6-O-benzylidene-directed β-mannosylation as determined by α-deuterium kinetic
isotope effects. Angew. Chem. Int. Ed. 2004, 43, 5386-5389.
9.- a) Bennet, A.J.; Sinnot, M.L. Complete kinetic isotope effect description of transition states
for acid-catalyzed hydrolyses of methyl .alpha.- and .beta.-glucopyranosides. J. Am. Chem. Soc. 1986,
108, 7287-7294. b) Namchuk, M.N.; McCarter, J.D.; Becalski, A.; Andrews, T.; Withers, S.G. The
Role of Sugar Substituents in Glycoside Hydrolysis. J. Am. Chem. Soc. 2000, 122, 1270-1277. c)
Chan, J.; Tang, A.; Bennet, A.J. A Stepwise Solvent-Promoted SNi Reaction of α-d-Glucopyranosyl
Fluoride: Mechanistic Implications for Retaining Glycosyltransferases. J. Am. Chem. Soc. 2012, 134,
1212-1220. d) Beaver, M.G.; Buscagan, T.M.; Lavinda, O.; Woerpel, K.A. Stereoelectronic Model to
Explain Highly Stereoselective Reactions of Seven-Membered-Ring Oxocarbenium-Ion
Intermediates. Angew. Chem. Int. Ed. 2016, 55, 1816-1819. e) Smith, D.M.; Woerpel, K.A.
Electrostatic interactions in cations and their importance in biology and chemistry. Org. Biomol.
Chem. 2006, 4, 1195-1201. f) van Rijssel, E.R.; van Delft, P.; Lodder, G.; Overkleeft, H.S.; van der
Marel, G.A.; Filippov, D.V.; Codꢀe, J.D.C. Furanosyl oxocarbenium ion stability and
stereoselectivity. Angew. Chem. Int. Ed. 2014, 53, 10381-10385. g) Ayala, L.; Lucero, C.G.; Romero,
J.A.C.; Tabacco, S.A.; Woerpel, K.A. Stereochemistry of Nucleophilic Substitution Reactions
Depending upon Substituent: Evidence for Electrostatic Stabilization of Pseudoaxial Conformers of
Oxocarbenium Ions by Heteroatom Substituents. J.Am. Chem. Soc. 2003, 125, 15521-15528.
10.- a) Chatterjee, S.; Moos, S.; Hentschel, F.; Gilmore, K.; Seeberger, P.H. An Empirical
Understanding of the Glycosylation Reaction. J. Am. Chem. Soc. 2018, 140, 11942-11953. b)
Heuckendorff, M.; Bendix, J.; Pedersen, C.M.; Bols, M. β-selective mannosylation with a 4,6-
silylene-tethered thiomannosyl donor. Org. Lett. 2014, 16, 1116-1119.
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 23
11.- Martin, A.; Arda, A.; Désiré, J.; Martin-Mingot, A.; Probst, N.; Sinaÿ, P.; Jiménez-
Barbero, J.; Thibaudeau, S.; Blériot, Y. Catching elusive glycosyl cations in a condensed phase with
HF/SbF5 superacid. Nat. Chem. 2016, 8, 186-191.
12.- a) Mucha, E.; Marianski, M.; Xu, F.F.; Thomas, D.A.; Meijer, G.; von Helden, G.;
Seeberger, P.H.; Pagel, K. Unravelling the structure of glycosyl cations via cold-ion infrared
spectroscopy. Nat. Commun. 2018, 9, 1-5. b) Elferink, H.; Severijnen, M.E.; Martens, J.; Mensink,
R.A.; Berden, G.; Oomens, J.; Rutjens, F.P.J.T.; Rijs, A.M.; Boltje, T.J. Direct Experimental
Characterization of Glycosyl Cations by Infrared Ion Spectroscopy. J. Am. Chem. Soc. 2018, 140,
6034-6038.
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
13.- Lillelund, V.H.; Jensen, H.H.; Liang, X.; Bols, M. Recent Developments of Transition-
State Analogue Glycosidase Inhibitors of Non-Natural Product Origin. Chem. Rev. 2002, 102, 515-
553.
14.- a) Danby, P.M.; Withers, S.G. Glycosyl Cations versus Allylic Cations in Spontaneous
and Enzymatic Hydrolysis. J. Am. Chem. Soc. 2017, 139, 10629-10632. b) Zechel, D.L.; Withers,
S.G. Glycosidase Mechanisms:ꢁ Anatomy of a Finely Tuned Catalyst. Acc. Chem. Res. 2000, 33, 11-
18. c) Vasella, A.; Davies, G.J.; Bohm, M. Glycosidase mechanisms. Curr. Opin. Chem. Biol. 2002,
6, 619-629.
15.- a) Ren, W.; Pengelly, R.; Farren-Dai, M.; Shansi Kazem Abadi, S.; Oehler, V.; Akintola,
O.; Draper, J.; Meanwell, M.; Chakladar, S.; Świderek, K; Moliner, V.; Britton, R.; Gloster, T.M.;
Bennet, A.J. Revealing the mechanism for covalent inhibition of glycoside hydrolases by carbasugars
at an atomic level. Nat. Commun. 2018, 9, 1-12. b) Shamsi Kazem Abadi, S.; Tran, M.; Yadav, A.K.;
Adabala, P.J.P.; Chakladar, S.; Bennet, A.J. New Class of Glycoside Hydrolase Mechanism-Based
Covalent Inhibitors: Glycosylation Transition State Conformations. J. Am. Chem. Soc. 2017, 139,
10625-10628.
16.- Nerinckx, W.; Desmet, T.; Claeyssens, M. A hydrophobic platform as a mechanistically
relevant transition state stabilising factor appears to be present in the active centre of all glycoside
hydrolases. Febs. Lett. 2003, 538, 1-7.
17.- Iglesias-Fernández. J.; Hanckock, S.M.; Lee, S.S.; Khan, M.; Kirkpatrick, J.; Oldham,
N.J.; McAuley, K.; Fordham-Skelton, A.; Rovira, C.; Davis, B.G. A front-face 'SNi synthase'
engineered from a retaining 'double-SN2' hydrolase. Nat. Chem. Biol. 2017, 13, 874-886.
18.- Vacas, T.; Corzana, F.; Jiménez-Osés, G.; González, C.; Gómez, A.M.; Bastida, A.;
Revuelta, J.; Asensio, J.L. Role of aromatic rings in the molecular recognition of aminoglycoside
antibiotics: Implications for drug design. J. Am. Chem. Soc. 2010, 132, 12074-12090.
19.- Lin, L.; Li, C.; Alexov, E. On the dielectric "constant" of proteins: Smooth dielectric
function for macromolecular modeling and its implementation in DelPhi. J. Chem. Theory Comput.
2013, 9, 2126-2136.
20.- Crich, D.; Sun, S. Are glycosyl triflates intermediates in the sulfoxide glycosylation
method? A chemical and 1H, 13C, and 19F NMR spectroscopic investigation. J. Am. Chem. Soc. 1997,
119, 11217-11223.
21.- a) van der Vorm, S.; Hansen, T.; Overkleeft, H.S.; van der Marel, G.A.; Codꢀe, J.D.C.
The influence of acceptor nucleophilicity on the glycosylation reaction mechanism. Chem. Sci. 2017,
8, 1867-1875. b) van der Vorm, S.; van Hengst, J.M.A.; Bakker, M.; Overkleeft, H.S.; van der Marel,
G.A.; Codée, J.D.C. Mapping the Relationship between Glycosyl Acceptor Reactivity and
Glycosylation Stereoselectivity. Angew. Chem. Int. Ed. 2018, 57, 8240-8244.
22.-Reactivity assays, were carried out with enantiomerically pure sulfoxides unless explicitly
stated. This structural parameter was shown to have no influence on the obtained conclusions.
23.- Hon, D.; Taha, H.A.; Lowary, T.L. 2,3-Anhydrosugars in Glycoside Bond Synthesis:
Mechanism of 2-Deoxy-2-thioaryl Glycoside Formation. J. Am. Chem. Soc. 2009, 131, 12937-12948.
24.- Kiyooka, S.I.; Kaneno, D.; Fujiyama, R. Intrinsic reactivity index as a single scale
directed toward both electrophilicity and nucleophilicity using frontier molecular orbitals.
Tetrahedron 2013, 69, 4247-4258.
25.-Bennet, C.S.; Galan, M.C. Methods for 2-Deoxyglycoside Synthesis. Chem. Rev. 2018,
118, 7931-7985.
22
ACS Paragon Plus Environment

Page 23 of 23
Journal of the American Chemical Society
TOC
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
23
ACS Paragon Plus Environment