The synthesis and conformational analysis of a series of phenyl 2,3,6-tri- O-benzyl-β-d-thio galacto- and glucopyranosides and their 6 S-deuterio isotopomers, with systematic variation of the protecting group at the 4-position, are described. For the galactopyranosides, replacement of a 4- O-benzyl ether by a 4- O-alkanoyl or aroyl ester results in a small but measurable shift in side chain population away from the trans, gauche conformation and in favor of the gauche, trans conformer. In the glucopyranoside series on the other hand, replacement of a 4- O-benzyl ether by a 4- O-alkanoyl or aroyl ester results in a small but measurable increase in the population of the trans, gauche conformer at the expense of the gauche, gauche conformer. The possible modulating effect of these conformational changes on the well-known changes in the anomeric reactivity of glycosyl donors as a function of protecting group is discussed, raising the possibility that larger changes may be observed at the transition state for glycosylation. A comparable study with a series of ethyl 2,3,4-tri- O-benzyl-β-d-thioglucopyranosides reveals that no significant influence in side chain population is observed on changing the O6 protecting group.
The synthesis and conformational analysis of a series of phenyl 2,3,6-tri- O-benzyl-β-d-thio galacto- and glucopyranosides and their 6 S-deuterio isotopomers, with systematic variation of the protecting group at the 4-position, are described. For the galactopyranosides, replacement of a 4- O-benzyl ether by a 4- O-alkanoyl or aroyl ester results in a small but measurable shift in side chain population away from the trans, gauche conformation and in favor of the gauche, trans conformer. In the glucopyranoside series on the other hand, replacement of a 4- O-benzyl ether by a 4- O-alkanoyl or aroyl ester results in a small but measurable increase in the population of the trans, gauche conformer at the expense of the gauche, gauche conformer. The possible modulating effect of these conformational changes on the well-known changes in the anomeric reactivity of glycosyl donors as a function of protecting group is discussed, raising the possibility that larger changes may be observed at the transition state for glycosylation. A comparable study with a series of ethyl 2,3,4-tri- O-benzyl-β-d-thioglucopyranosides reveals that no significant influence in side chain population is observed on changing the O6 protecting group.
In carbohydrate chemistry,
it is widely understood that anomeric
reactivity is strongly influenced by the relative configuration of
the complete set of stereogenic centers in the backbone.[1] Thus, for example, galactopyranosides undergo
both acid-catalyzed and spontaneous hydrolysis more rapidly than their
gluco isomers (Figure );[2−5] the same pattern of reactivity is found in glycosylation reactions
with a series of comparably protected thioglycosides.[6] The influence of protecting groups on the anomeric reactivity
of glycosyl donors also is broadly appreciated, with the more electron-withdrawing
(or disarming) esters retarding reaction rates compared to the less
electron-withdrawing (or arming) ethers.[6−11]
Figure 1
Relative
rates of spontaneous hydrolysis of galacto- and glucopyranosides
in water at 37 °C.[5]
Relative
rates of spontaneous hydrolysis of galacto- and glucopyranosides
in water at 37 °C.[5]Ring conformation is another important factor in
anomeric reactivity.
Thus, for any given configuration, ring conformations that maximize
the number of axial (or pseudoaxial) C–O bonds generally exhibit
the greatest anomeric reactivity (Figure ).[12−15] The influence of configuration and ring conformation
on anomeric reactivity is best explained by the ability of axial C–O
bonds to stabilize nascent positive charge at the anomeric center
as compared to their equatorial counterparts.[16−19]
Figure 2
Influence of ring conformation on the
hydrolysis of axial methyl
glucosides in 2 M HCl at 60 °C.[4,15]
Influence of ring conformation on the
hydrolysis of axial methyl
glucosides in 2 M HCl at 60 °C.[4,15]The conformation of the side chain, defined as gauche,gauche (gg), gauche,trans (gt), or trans,gauche (tg) where the
first and second terms refer to the position of O6 relative to O5
and C4, respectively,[20−22] is also increasingly recognized as influencing anomeric
reactivity. Thus, the trans,gauche conformation exerts
maximum retardation due to the strongly electron-withdrawing antiperiplanar
relationship of the C6–O6 and the C5–O5 bonds (Figure ).[23−25] The interplay
between the conformation of the side chain and the glycosidic bond
is further apparent from the work of Vázquez and co-workers
in which it is demonstrated by CD and NMR methods that both the anomeric
configuration and the nature of the aglycone influence the population
of the different side chain conformers.[26−30]
Figure 3
Influence of the gauche,gauche (gg), gauche,trans (gt),
and trans,gauche (tg) side chain
conformations
on the relative rates of spontaneous hydrolysis of 2,4-dinitrophenyl
glycosides in water at 37 °C.[23,24]
Influence of the gauche,gauche (gg), gauche,trans (gt),
and trans,gauche (tg) side chain
conformations
on the relative rates of spontaneous hydrolysis of 2,4-dinitrophenyl
glycosides in water at 37 °C.[23,24]In this Article, we begin to explore the possibility
that protecting
groups, in addition to their well-known influence on glycoside reactivity
as a function of their electron-withdrawing ability,[1,7−11] also exert an indirect influence on anomeric reactivity by modulating
the conformation of the side chain. To this end, we describe the preparation
of a series of galacto and gluco thiopyranosides and, to facilitate
spectral assignment, their 6S-deuterio isotopomers
and study the conformation of the side chain as a function of protecting
group at either the 4- or the 6-position. We show that the side chain
population in a series of phenyl 2,3,6-tri-O-benzyl-β-d-thiogalacto- and glycopyranosides does indeed vary in a systematic
manner on changing the functionality at the 4-position from a hydroxyl
group to an ether and to an ester, albeit in a different manner in
the two configurations. While these protecting-group-induced changes
in conformation are small, they open the possibility that larger changes
might arise at the transition state for glycosylation and thus open
alternative avenues for the explanation of remote protecting group
effects.
Results and Discussion
Experimental Design and Synthesis
In this investigation,
we focus on the interplay between the protecting groups at O4 and
O6 in the gluco- and galactopyranosides as the strongest candidates
for observation of any changes in side chain conformation due to these
interactions. Judging that the interaction in question could be probed
through the variation of the O4 protecting group in the presence of
a fixed O6 protecting group, or the inverse, we prepared phenyl 2,3,6-tri-O-benzyl-β-d-thiogalactopyranoside (1),[31] phenyl 2,3,6-tri-O-benzyl-β-d-thioglucopyranoside (2),[32] and ethyl 2,3,4-tri-O-benzyl-β-d-thioglucopyranoside (3)[33] by standard means. The use of the ethyl thioglycoside
in 3 as opposed to the phenyl thioglycosides 1 and 2 was a matter of experimental convenience and
was not expected to affect the analysis of side chain conformations,
as is borne out by the subsequent results.As the rigorous assignment of the diastereotopic
pro-R and pro-S hydrogens at the
6-position of 1–3 and their derivatives
is critical to the correct conformational analysis of their side chains,[20] we also prepared 6S-deuterio 1–3 as outlined in Schemes and 2. Thus, preferring
a longer but unambiguous route based on exoselective quenching of
radicals at the 6-position of 1,6-anhydropyranoses[34−38] over shorter routes involving asymmetric reduction
of 6-aldehydo-sugars,[39−41] 2,3,4-tri-O-acetyl-1,6-anhydro-d-galactose (4) was subjected to white-light-mediated
bromination with N-bromosuccinimide to give the exo-bromide 5 in good yield. This transformation follows the literature
description[36] with the exception that the
original solvent, tetrachloromethane, was replaced by the more environmentally
friendly α,α,α-trifluoromethylbenzene[42] as described previously for the gluco series.[43] Reductive debromination with tributyltin deuteride,
prepared according to Neumann,[44] then gave
the 6S-deuterio anhydrogalactose (6)
in 77% yield. A series of standard transformations were then applied
to convert 6 via intermediates 7–10 to the desired 6- uneventfully (Scheme ). The 6S-deuterio analogue
of 2 was prepared by inversion of 6- by triflate formation,
displacement with sodium benzoate, and saponification as reported
in detail in the Experimental Section.
Scheme 1
Synthesis of Phenyl 2,3,6-Tri-O-benzyl-6S-deuterio-β-d-thiogalactopyranoside (6-)
Scheme 2
Synthesis of Ethyl 2,3,4-Tri-O-benzyl-6S-deuterio-β-d-thioglucopyranoside (6-)
The 6S-deuterio-1,6-anhydroglucose derivative
(11)[34,35,43] was the starting material for the preparation of 6-. Thus, 11 was converted to the thioglycoside 12 by cleavage of
the 1,6-anhydro bridge with trimethylsilyl ethanethiol in the presence
of zinc iodide[45] followed by acetylation
(Scheme ). A series
of standard reactions were then employed to convert 12 to the tribenzyl ether 6- as described previously for the nondeuterated
isotopomer[33] (Scheme ).The thiogalactoside 1 and its 6S-monodeuterio
isotopomer were converted to a series of esters 13–20 at the 4-position as well as to the benzyl ether 21 by standard methods as described in the Experimental Section. The 4-O-(2,2,2-trifluoroethyl)
ether (22) and its 6S-deuterio analogue
(6-) were obtained from 2 and 6S-deuterio 2 by triflation followed by displacement with sodium trifluoroethoxide
in DMF (Scheme ).
The thioglucoside 2 and its 6S-monodeuterio
isomer were also converted to the derivatives 23–27 by standard means, as described in the Experimental Section. Similarly, the thioglucoside 3 and its 6S-monodeuterio isotopomer were converted
to the 6-O-esters 28–34, the 6-O-carbamate 35, and the 6-O-benzyl ether 36 by standard means, as described
in the Experimental Section.
Scheme 3
Synthesis
of Phenyl 2,3,6-Tri-O-benzyl-4-O-(2,2,2-trifluoroethyl)-β-d-thiogalactopyranoside
(22) and Its 6S-Deuterio Analogue (6-)
Measurement of NMR Spectra,
Influence of Solvent, and Estimation
of Errors in Coupling Constants
The 1H NMR spectra
of 1–3 and 13–36 were recorded in CDCl3 and deuteriobenzene and
fully assigned by the usual array of 1D- and 2D-NMR methods, with
the distinction between the 6-pro-R and pro-S resonances, herein after H6 and H6, made on the basis of comparison
with the selectively 6S-deuterated analogues. All
first order couplings were analyzed directly. For second order spectra,
including those complicated by the presence of virtual coupling, the
spin simulation tool in the MestReNova 9.0 suite of programs was used
to extract first order coupling constants. The chemical shifts and
so-obtained 3J coupling constants in the
H5, H6, and H6 spin systems are presented in Table for the 4-O-substituted
galacto series of compounds, in Table for the corresponding 4-O-substituted
gluco compounds, and in Table for the 6-O-substituted gluco series of
compounds. For ease of comparison, in each of Tables –3, the alcohols
are listed first, followed by the ethers, and then the esters grouped
according to patterns in the side chain conformations.
Table 1
Pertinent 1H Chemical Shifts, 3J Coupling Constants, and Side Chain Populations
for 1 and 13–22 in CDCl3 and C6D6
chem
shifta (ppm)
3J5,6a (Hz)
populationa (%)
4-O-Subs
H6R
H6S
J5,6R
J5,6S
gg
gt
tg
1
H
3.78 (3.72)
3.81 (3.75)
5.8 (6.1)
5.7 (5.8)
25.6 (21.9)
31.8 (34.4)
42.6 (43.7)
21
PhCH2
3.65 (3.57)
3.67 (3.66)
nd (5.7)
nd (7.4)
nd (13.2)
nd (22.6)
nd (64.2)
22
CF3CH2
3.69 (3.55)
3.74 (3.68)
5.5 (5.7)
7.7 (7.6)
12.8 (11.7)
19.1 (21.7)
68.0 (66.7)
13
Ac
3.64 (3.47)
3.55 (3.42)
6.0 (6.1)
6.7 (6.5)
15.8 (16.4)
29.0 (31.0)
55.2 (52.6)
14
Me3CCO
3.62 (3.49)
3.49 (3.42)
6.1 (6.0)
6.7 (6.8)
14.8 (15.0)
30.1 (28.6)
55.1 (56.4)
17
PhCO
3.69 (3.53)
3.59 (3.49)
5.8 (6.1)
6.8 (6.6)
17.0 (15.6)
26.5 (30.5)
56.5 (53.9)
18
p-MeC6H4CO
3.68 (3.56)
3.57 (3.52)
5.9 (6.2)
6.8 (6.5)
16.0 (15.4)
27.5 (32.0)
56.5 (52.5)
19
p-MeOC6H4CO
3.69 (3.57)
3.59 (3.54)
6.1 (6.2)
6.5 (6.4)
16.4 (16.2)
31.0 (32.5)
52.6 (51.3)
20
p-O2NC6H4CO
3.69 (3.47)
3.56 (3.42)
5.7 (5.6)
7.2 (7.0)
14.8 (17.3)
23.6 (23.5)
61.6 (59.1)
15
CF3CO
3.67 (3.39)
3.49 (3.36)
5.6 (5.7)
8.3 (8.1)
7.1 (7.7)
17.3 (19.2)
75.6 (73.0)
16
Cl3CCO
3.71 (3.43)
3.59 (3.46)
5.7 (nd)
8.2 (nd)
6.9 (nd)
18.8 (nd)
74.3 (nd)
Chemical shift, coupling constants,
and populations in CDCl3 (chemical shift, coupling constants,
and populations in C6D6).
Table 2
Pertinent 1H Chemical Shifts, 3J Coupling Constants,
and Side Chain Populations
for 2 and 23–27 in CDCl3 and C6D6
chem
shifta (ppm)
3J5,6a (Hz)
populationa (%)
4-O-Subs
H6R
H6S
J5,6R
J5,6S
gg
gt
tg
2
H
3.76 (3.60)
3.80 (3.64)
5.3 (5.3)
4.1 (3.7)
43.1 (46.2)
34.5 (36.4)
22.4 (17.4)
27
PhCH2
3.73 (3.63)
3.80 (3.63)
4.8 (nd)
1.9 (nd)
65.2 (nd)
40.0 (nd)
–5.3 (nd)
23
Ac
3.58 (3.54)
3.58 (3.58)
nd (5.8)
nd (3.2)
nd (45.3)
nd (43.9)
nd (10.8)
25
PhCO
3.65 (3.58)
3.65 (3.62)
nd (6.0)
nd (2.9)
nd (45.7)
nd (47.3)
nd (7.0)
24
CF3CO
3.58 (3.30)
3.61 (3.37)
4.7 (4.3)
3.6 (3.5)
52.9 (57.6)
30.8 (27.2)
16.3 (15.2)
26
PhSO2
3.51 (3.69)
3.71 (3.81)
5.6 (5.3)
2.1 (2.0)
55.9 (59.6)
47.1 (44.6)
–3.0 (−4.2)
Chemical shift, coupling constants,
and populations in CDCl3 (chemical shift, coupling constants,
and populations in C6D6).
Table 3
Pertinent 1H Chemical Shifts
and 3J Coupling Constants for 3 and 28–36 in CDCl3 and
C6D6
chem
shifta (ppm)
3J5,6a (Hz)
populationa (%)
6-O-Subs
H6R
H6S
J5,6R
J5,6S
gg
gt
tg
3
H
3.71 (3.58)
3.87 (3.69)
4.8 (nd)
2.7 (nd)
59.0 (nd)
36.1 (nd)
4.9 (nd)
36
PhCH2
3.69 (3.61)
3.76 (3.61)
5.0 (nd)
1.9 (nd)
63.3 (nd)
42.0 (nd)
–5.3 (nd)
28
Ac
4.20 (4.26)
4.33 (4.43)
5.4 (5.4)
2.4 (2.2)
55.5 (57.0)
43.7 (44.6)
0.9 (−1.7)
29
Me3CCO
4.12 (4.23)
4.44 (4.55)
5.6 (5.7)
1.8 (2.1)
58.2 (54.9)
48.6 (48.2)
–6.8 (−3.0)
31
PhCO
4.46 (4.51)
4.65 (4.65)
nd (5.6)
nd (2.3)
nd (54.3)
nd (46.2)
nd (−0.5)
32
p-MeC6H4CO
4.40 (4.68)
4.59 (4.54)
5.5 (5.7)
2.2 (2.2)
56.1 (54.1)
45.6 (47.7)
–1.7 (−1.8)
33
p-MeOC6H4CO
4.58 (4.56)
4.41 (4.69)
5.5 (5.6)
2.3 (2.2)
55.3 (55.1)
45.2 (46.7)
–0.4 (−1.7)
34
p-O2NC6H4CO
4.44 (4.40)
4.62 (4.56)
5.4 (5.6)
2.2 (2.3)
57.0 (54.3)
44.6 (46.2)
–1.7 (−0.5)
30
CF3CO
4.32 (4.05)
4.54 (4.24)
6.3 (6.3)
2.1 (2.2)
49.0 (48.2)
54.2 (53.8)
–3.3 (−2.0)
35
PhNHCO
4.37 (4.46)
4.40 (4.43)
nd (4.6)
nd (2.5)
nd (62.5)
nd (35.1)
nd (2.4)
Chemical shift, coupling constants,
and populations in CDCl3 (chemical shift, coupling constants,
and populations in C6D6).
Chemical shift, coupling constants,
and populations in CDCl3 (chemical shift, coupling constants,
and populations in C6D6).Chemical shift, coupling constants,
and populations in CDCl3 (chemical shift, coupling constants,
and populations in C6D6).Chemical shift, coupling constants,
and populations in CDCl3 (chemical shift, coupling constants,
and populations in C6D6).Inspection of the coupling constant
data in Tables –3 reveals
no systematic difference in coupling constants for any given system
on changing from CDCl3 to C6D6 consistent
with an earlier study with rigid bicyclic models of galacto- and glucopyranosides,[46] from which we conclude that the side chain conformation
is unaffected by the nature of the solvent (CDCl3 or C6D6). It follows that data acquired in one of the
two solvents can be extrapolated to the other solvent in cases for
which resolution was insufficient to enable the determination of coupling
constants in both solvents. Small nonsystematic differences between
solvents are found and permit the estimation of errors in the 3J5,6 and 3J5,6 coupling
constants. For the 18 compounds (and hence 36 sets of coupling constants)
in which both values could be measured in both solvents, differences
in a given coupling constant are 0.4 Hz or less consistent with the
digital resolution of 0.38 Hz, leading us to adopt 0.4 Hz as the error
limit in these measurements. The differences in the 3J5,6 and 3J5,6 coupling constants of
the phenyl and ethyl glycosides of 2,3,4,6-tetra-O-benzyl-β-d-thioglucopyranoside (27 and 36) (Tables and 3, respectively) are less than the error
limit and so substantiate the use of the different aglycones in the
two series.
Calculation of Side Chain Populations and
Estimation of Errors
With the aid of limiting 3J–3J coupling constants
for the pure gg, gt, and tg conformers (Table ), determined using a series of rigid bicyclic models,[46] the side chain populations (fgg–ftg) of all compounds
were determined with the aid of eqs –3 in the usual manner
and reported in Tables –3.[47−49] Further inspection of Tables –3 reveals that an error of 0.4 Hz in one of the coupling
constants results in a maximal change of 5% in the population of any
given conformer. Therefore, in the discussion that follows, we adopt
5% as the error limit for any given conformer. In two series of compounds
(Tables and 3), small negative populations of the tg conformer are computed for some derivatives, which have no physical
significance. In view of the 5% error, these negative populations
are either indistinguishable from zero or so close to it as not to
warrant further discussion.
Table 4
Limiting Coupling Constants for the
Pure gg, gt, and tg Staggered Conformers
3JH5,H6R
3JH5,H6S
JR,gg
JR,gt
JR,tg
JS,gg
JS,gt
JS,tg
1.0
11.0
4.8
2.2
2.5
10.2
Corrections for the Influence of Electronegative
Groups on the
Magnitude of Coupling Constants
The derivation of side chain
populations from experimental NMR vicinal coupling constants requires
that the magnitude of the coupling constants not be affected by differences
in electronegativity of the substituents across the series. It is
well-known that vicinal coupling constants are reduced by the presence
of electronegative substituents in the coupled system,[50−53] but more subtle differences on replacement of ether groups by esters
are less appreciated.[51,54,55] Consideration of the 3J3,4 coupling constants in the series of gluco- and galactopyranoside
derivatives in Figure indicates that replacement of a single ether group in a vicinal
diether by an ester results in an increase of approximately 0.5 Hz
in the coupling constant, whether the coupled spins have a fixed trans or gauche relationship irrespective
of the solvent, CDCl3 or C6D6. However,
it is known that in freely rotating systems that more closely approximate
the C5–C6 bond in the 6-O-substituted series 3 and 28–36 the difference
in coupling constants on replacing an ether by an ester substituent
is only 0.1 Hz.[54] As this is within the
error limit, no correction for the change in substituents is required
for the coupling constants in Table .
Figure 4
Differing influences of ether and ester substituents on
vicinal
coupling constants.
Differing influences of ether and ester substituents on
vicinal
coupling constants.The influence of replacing
an ether substituent by an ester in
the β-position to the coupled spins is expected to be smaller,
as is borne out by the constant magnitude of the 3J2,3 coupling constants in the gluco- and galactopyranosides
of Figure regardless
of the substituent at the 4-position. The 3J5,6 coupling constants in the galactopyranosides 1 and 13–22 (Table ) and the glucopyranosides 2 and 23–27 (Table ) therefore do not require correction
for the nature of the substituent at the 4-position.Comparison
of the 3J2,3 coupling
constants in the galactopyranosides reveals them to be ∼0.4
Hz larger than the corresponding coupling constants in the glucopyranosides
(Figure ), as we have
noted previously in a series of rigid bicyclic models.[46] This is a manifestation of the Altona and Haasnoot
β-effect wherein the vicinal coupling constant between a pair
of coupled antiperiplanar spins is increased by approximately 0.5
Hz when one of the coupled spins is also antiperiplanar to an electron-withdrawing
substituent at the β-position.[56] A
comparable relationship exists between H6, H5, and O4 in the gt conformer and
between H6, H5, and O4 in
the tg conformer of the galactopyranosides (Figure ) but not in the
glucopyranosides, albeit in a series of rigid bicyclic models, no
significant difference was found in the H5–H6 coupling constants between the galacto and gluco configurations,[46] suggesting that the effect does not extend to
this spin system. Moreover, as the magnitude of the β-effect
is not influenced by the switch from an ether to an ester (Figure ), any residual influence
will be of a systematic nature and affect all derivatives to a similar
extent. The result of any small systematic β-effect simply will
be to overestimate the population of the gt and tg conformers in the galactopyranoside series (Table ) and underestimate
that of the gg conformer correspondingly, with respect
to the glucopyranosides (Table ).
Figure 5
Vicinal coupling constants subject to the Altona–Haasnoot
β-effect.
Vicinal coupling constants subject to the Altona–Haasnoot
β-effect.In the final analysis,
no corrections to the diagnostic coupling
constants used for conformational analysis of the side chain arising
from changes in electronegativity of the substituents or the Altona
and Haasnoot β-effect were deemed necessary.
Influence of
Substituents at the Galactopyranose 4-Position
Comparison
of alcohol 1 with ethers 21 and 22 in Table reveals
that, while the two ethers have the same population
distribution for the side chain given the 5% error, converting the
4-hydroxy group to a benzyl or trifluoroethyl ether has a significant
influence on the side chain conformation. Thus, the side chain population
of the two ethers of 21 and 22 consists
of ∼13% of the gg conformer, ∼21% of
the gt conformer, and ∼66% of the tg conformer, whereas the alcohol 1 has a much
greater population of the gg conformer (∼24%)
and a greater population of the gt conformer (∼33%),
which are balanced by a significantly lower occupancy of the tg conformer (∼43%). These differences reflect either
the destabilizing influence of the steric bulk at the 4-position on
the gg conformer or the stabilization of the gg conformer in the alcohol 1 by a favorable
intramolecular hydrogen bond to O6. Because of the differences in
conformation between the alcohol 1 and the ethers 21 and 22 and because glycosyl donors typically
do not have unprotected hydroxy groups, we retain the benzyl ether 21 as the standard for further comparisons.Esterification
of 1 gives a series of 4-O-esters, alkanoyl 13 and 14, or aroyl 17–20, that all adopt the same side chain population, but one
that differs significantly from that of the benzyl ether 21. With similar proportions of the gg conformer in
the ether 21 and the esters 13, 14, and 17–20, the change in the overall
side chain population can be described as one of an approximately
10% decrease in the population of the tg conformer
in favor of the gt conformer on going from the ethers
to the esters. Installation of the more electron-withdrawing trifluoroacetyl
and trichloroacetyl groups affords a separate set of two esters 15 and 16, respectively, whose side chain populations
exhibit a pronounced shift away from the gg conformer
toward the tg conformer. This change in population
is also accompanied by a reduction in the population of the gt conformer with respect to the more standard alkanoyl
and aroyl esters such that the tg conformer dominates
the equilibrium and accounts for ∼74% of the population. Although
the effect is smaller, the p-nitrobenzoate 20 exhibits a shift in side chain population away from that
of the more electron-rich benzoates 17–19 in the same direction as that observed with the trifluoro- and trichloroacetates 15 and 16, suggesting that this change is a function
of the electron-withdrawing nature of these esters.
Influence of
Substituents at the Glucopyranose 4-Position
In the glucopyranose
series, there is also a significant change
in the side chain conformation when the alcohol 2 is
converted to the benzyl ether 27. Specifically, benzylation
results in a drop in the population of the tg conformer
that is compensated by an increase in the population of the gg conformer and a smaller one in that of the gt conformer (Table ). This effect parallels that seen in the galactopyranose series,
in that it is the conformer in which O4 and O6 have a syn-pentane-type relationship (gg in galactose and tg in glucose) whose population is reduced on benzylation
(Figure ), suggesting
that this change in conformation is caused by the loss of a favorable
hydrogen bond or increased steric interactions in both cases.
Figure 6
syn-Pentane conformations of the galacto- and
glucopyranoses destabilized on replacement of a hydroxy group (X =
H) by an ether (X = R).
syn-Pentane conformations of the galacto- and
glucopyranoses destabilized on replacement of a hydroxy group (X =
H) by an ether (X = R).As in the galactopyranose series, we adopt the benzyl ether
as
the standard for the subsequent comparisons with the influence of
alternative protecting groups at the 4-position. The 4-O-acetate 23 and the benzoate 25 adopt very
similar conformations in which the gg conformer is
populated to a noticeably lower extent than in the benzyl ether 27, while the population of the tg conformer
increases. With the more electron-withdrawing trifluoroacetate 24, the population of the tg conformer increases
further, but it is balanced by a reduction in the population of both
the gg and gt conformers when compared
to the benzyl ether 27. The population of the side chain
conformation of the benzenesulfonate 26 is anomalous
insofar as, unlike the other esters, the tg conformation
is not occupied, presumably for steric reasons arising from an increased syn-pentane interaction. This minimal population of the tg conformation in 26 results in a higher population
of the gg conformer than in the acetal and benzoate
esters 23 and 25.
Influence of Substituents
at the Glucopyranose 6-Position
In contrast to the differences
in side chain population brought
about by changing protecting groups at the 4-position in the galacto-
and glucopyranose series, changes in the protecting group at the 6-position
of the 2,3,4-tri-O-benzyl glucopyranosides have a
minimal influence on the side chain conformation (Table ). It is noteworthy, however,
that two derivatives, the 6-alcohol 3 and the 6-carbamate 35, have a small population of the tg conformation,
with the syn-pentane conformation (Figure ), suggesting that this conformer
is stabilized by hydrogen bonding. The only other noteworthy feature
from this series of compounds is the increased population of the gt conformer at the expense of the gg conformer
on the installation of the strongly electron-withdrawing trifluoroacetyl
group. In view of the relatively small changes in side chain conformation
observed in the glucopyranose series with variation in the O6 protecting
group, we did not undertake a parallel study in the galactopyranose
series.
Discussion
The observed changes
in side chain conformation with protecting
groups at the 4-position for both the galacto- and glucopyranose systems
are summarized in Figure . These changes are small, worth <1 kcal·mol–1, difficult to compute accurately with electronic structure calculations,[57−60] and insufficient to account alone for the changes in anomeric reactivity
and selectivity seen with such comparable changes in protecting group.[6,61−66] As the effects are small, we make no attempt here to rationalize
them in terms of stereoelectronic or other phenomena, other than to
note that they are certainly related among other things to the distinct
conformational preferences of esters and ethers.[67−69] Nevertheless,
such changes can be considered to modulate the larger effects on anomeric
reactivity arising from replacing an arming with a disarming protecting
group.
Figure 7
Summary of changes in side chain conformation with protecting groups
at the 4-position of (a) galactopyranosides and (b) glucopyranosides.
Summary of changes in side chain conformation with protecting groups
at the 4-position of (a) galactopyranosides and (b) glucopyranosides.In both the galacto- and glucopyranosyl
systems, the replacement
of an arming[8] ether protecting group at
the 4-position by an ester group results in diminished anomeric reactivity,
whether under standard conditions for the activation of thioglycosides[6] or in SN2-displacements of anomeric
bromides by chloride.[9] This change in reactivity
is usually understood in terms of the increased electron-withdrawing
ability of the ester destabilizing nascent positive or partial positive
charge at the reaction center (Scheme ).[1,5,6,8,17,70] The results presented in Table and summarized in Figure a indicate that this effect will be moderated
by the change in side chain conformation in the galactopyranoside
series. Thus, the increase in the gt conformation
with its intermediate reactivity at the expense of the less reactive tg conformation on replacement of a benzyl ether by an alkanoyl
or aroyl ester will partially offset the added electron-withdrawing
effect of the ester group. In the glucopyranosyl series on the other
hand (Table and Figure b), the main effect
of the replacement of a benzyl ether by an alkanoyl or aroyl ester
is to reduce the population of the most reactive gg conformer in favor of population of the least reactive tg conformer, thereby complementing the increased electron-withdrawing
effect of the ester.
Scheme 4
Abbreviated General Glycosylation Mechanism
and Influence of Protecting
Groups and Side Chain Conformation
In both the galacto-
and glucopyranosyl series (Tables and 2, Figure ), on replacement of a benzyl ether at the
4-position by the strongly electron-withdrawing trifluoroacetyl group
the trichloroacetyl group investigated in the galactose series, Table , the population of
the most strongly electron-withdrawing tg conformer
is increased. Thus, in both the galacto- and glucopyranosides, the
change in side chain conformation on installation of a trifluoroacetyl
group will reinforce the diminution of anomeric reactivity occasioned
by the increased electron-withdrawing effect.The sulfonyl protecting
group, initially explored as a strongly
electron-withdrawing group at the 2-position capable of stabilizing
manno- and rhamnopyranosyl triflates and other sulfonates,[71−73] and subsequently employed with varying degrees of success at the
4-position of 2,6-dideoxyglycopyranosyl donors,[74] and at the 3-, 4-, and 6-positions of other pyranosyl donors,[75,76] does not change the population of the side chain when replacing
a benzyl ether at the 4-position of a glucosyl donor (Table , Figure b). The influence of the 4-O-sulfonyl group on glucopyranosylation can therefore be interpreted
solely in terms of the change in electron-withdrawing ability.The small changes in side chain conformation in a series of 4-O-alkanoyl and aroyl esters in both the galacto- and glucopyranosyl
series do not provide strong support for a protecting-group-induced
change in side chain population as the basis for the changes in anomeric
selectivity and previously seen in the series of 17–20[77] and related systems,[78] previously explained by the controversial[61−66] concept of stereodirecting participation by remote groups. Similarly,
the consistent side chain conformation observed with numerous protecting
groups at the 6-position in the 2,3,4-tri-O-benzyl
glucopyranosides (Table ) does not support a role of the side chain conformation in the changes
in anomeric reactivity and glycosylation selectivity reported in such
series of compounds.[79−81]
Conclusion
Replacement of a benzyl
ether at the 4-position of phenyl 2,3,6-tri-O-benzyl-β-d-thio-galactopyranosides by either
an alkanoyl or aroyl ester results in a small but consistent change
in the population of the three staggered conformers of the side chain
in which the proportion of the less reactive tg conformer
is reduced in favor of the gt conformer. This suggests
that the reduction in anomeric reactivity occasioned by the benzyl
ether–ester switch seen with glycosyl donors, and attributed
to the increased electron-withdrawing ability of the ester, is moderated
by a change in side chain conformation in the galactose series. In
the corresponding glucopyranosyl series, on the other hand, the same
ether–ester change results in an increased population of the
less reactive tg conformer, indicating that the change
in conformation reinforces the effect of the increased electron-withdrawing
ability of the ester group. In both the galacto- and glucopyranosyl
series, the installation of a trifluoroacetyl group at the 4-position
results in an enhanced population of the less reactive tg conformer. While the variations in side chain conformation with
changes in protecting group recorded here are small, it must be understood
that they are for the unactivated glycosyl donor. In view of the partial
positive charge on the ring oxygen during glycosylation, it is likely
that such changes are accentuated at the transition state for glycosylation
reactions.. This possibility is under active investigation in our
laboratory.
Experimental Section
General Experimental Section
Commercial reagents were
used without further purification unless otherwise stated. NMR spectra
were recorded in CDCl3 solution unless otherwise stated
at 400, 500, or 600 MHz. 13C NMR spectra were recorded
in CDCl3 solution unless otherwise stated at 100, 125,
or 150 MHz. Mass spectra were recorded in the +ve ion mode using electrospray
ionization (ESI-TOF). ESI-HRMS were recorded with a Waters LCT Premier
Xe time-of-flight mass spectrometer. Specific rotations were recorded
in dichloromethane solution at room temperature.
A
solution of compound 12 (0.105 g, 0.27 mmol) in anhydrous
methanol (1.0 mL) was cooled to
0 °C, treated with Na (cat), and stirred under an argon atmosphere
at room temperature for 1.5 h. Then, the reaction was quenched with
Amberlyst-15 resin (pH ∼ 7) and filtered through cotton before
it was concentrated to dryness. The crude residue was dissolved in
anhydrous pyridine (0.8 mL) and treated with trityl chloride (84 μL,
0.32 mmol) at room temperature. The reaction mixture was stirred in
the dark at room temperature for 3 days. The reaction was then quenched
with methanol (∼70 μL), stirred for 1 h, and concentrated
to dryness. The crude residue was dissolved in chloroform and washed
with cold aqueous saturated NaHCO3 and cold brine. The
organic layer was dried over anhydrous Na2SO4 and concentrated to dryness. The crude residue was dissolved in
anhydrous DMF (0.9 mL) and cooled to 0 °C before it was treated
with 60% NaH in mineral oil (48 mg, 1.2 mmol) followed by benzyl bromide
(141 μL, 1.2 mmol). The reaction mixture was stirred under an
argon atmosphere at room temperature for 6 h. The reaction mixture
was cooled to 0 °C before it was quenched with water. The aqueous
layer was extracted with ethyl acetate. The combined organic layers
were washed with brine, dried over anhydrous Na2SO4, and concentrated to dryness. The crude residue was passed
through a short pad of silica, eluting with hexane/ethyl acetate (7:3),
and concentrated. The residue was dissolved in a mixture of glacial
acetic acid and water (4:1, 1.3 mL) and heated at 80 °C for 2.5
h. The reaction mixture was concentrated to dryness and purified using
silica gel column chromatography, eluting with hexane/ethyl acetate
(4:1 to 7:2), obtaining compound 6- (0.045 g, 35%). 1H
NMR (400 MHz, CDCl3) δ 7.40–7.25 (m, 15H),
4.94 (d, J = 10.9 Hz, 1H), 4.92 (d, J = 10.2 Hz, 1H), 4.87 (d, J = 10.9 Hz, 1H), 4.87
(d, J = 10.9 Hz, 1H), 4.75 (d, J = 10.2 Hz, 1H), 4.66 (d, J = 10.9 Hz, 1H), 4.51
(d, J = 9.8 Hz, 1H), 3.72 (t, J =
9.0 Hz, 1H), 3.68 (d, J = 4.9 Hz, 1H), 3.58 (dd, J = 9.7, 9.0 Hz, 1H), 3.41 (dd, J = 9.8,
8.8 Hz, 1H), 3.38 (dd, J = 9.7, 4.9 Hz, 1H), 2.84–2.69
(m, 2H), 1.33 (t, J = 7.4 Hz, 3H). HRMS (ESI) m/z calcd for C29H33DO5SNa [M + Na]+, 518.2087; found, 518.2066.
A mixture of l,6-anhydro-2,3,4-tri-O-acetyl-β-d-galactopyranose 4(83) (0.77 g, 2.67 mmol) and N-bromosuccinimide (1.9
g, 10.6 mmol)) in trifluorotoluene (40 mL) was refluxed over a 300
W heat lamp for 8 h. After 8 h, the solvent was evaporated under reduced
pressure and the crude product dissolved in EtOAc (50 mL). The solution
was successively washed with aqueous saturated Na2S2O3, aqueous saturated NaHCO3, and dried
over Na2SO4. Evaporation of the solvent under
reduced pressure gave a crude product which was purified through silica
gel column chromatography (eluent: 20% EtOAc in hexane) to give 5 (0.73g, 75%) as a yellowish oil. [α]D22 = −63.6 (c 1.50, CH2Cl2). 1H NMR (400 MHz, CDCl3) δ 6.60
(s, 1H), 5.81 (s, 1H), 5.29–5.20 (m, 2H), 4.74 (dd, J = 4.0, 1.3 Hz, 1H), 4.71 (br s, 1H), 2.14 (s, 3H), 2.13
(s, 3H), 2.05 (s, 3H). 13C NMR (100 MHz, CDCl3) δ 169.2, 169.1, 168.9, 101.7, 82.6, 79.6, 69.3, 66.9, 64.4,
20.7 (2), 20.5. HRMS (ESI) m/z calcd
for C12H1579BrO8Na [M
+ Na]+, 388.9848; found, 388.9849; calcd for C12H1581BrO8Na [M + Na]+, 390.9828; found, 390.9828.
To a solution of
compound 6 (1.4 g, 4.84
mmol) in MeOH (20 mL) was added Na (catalytic amount) slowly and the
reaction mixture was stirred for 7 h, then neutralized with Amberlite
IR120, filtered, and concentrated under reduced pressure to give (6S)-[6-2H1]-1,6-anhydrogalactose.
A solution of (6S)-[6-2H1]-1,6-anhydrogalactose (0.73 g, 4.47 mmol) in Ac2O (20 mL) was treated with conc. H2SO4 (0.35 mL) at 0 °C and stirred for 3 h. The reaction mixture
was poured into a saturated aqueous NaOAc solution (50 mL) and extracted
with CHCl3 (3 × 20 mL), washed with brine (10 mL),
dried over Na2SO4, and concentrated under a
high vacuum to give the crude product. Column chromatography (eluent:
30% EtOAc in hexane) on silica gel yielded 7 (1.5 g,
81% over two steps) as a colorless syrup with spectral data consistent
with the literature.[36]1H NMR
(400 MHz, CDCl3) δ 6.29 (d, J =
2.2 Hz, 1H), 5.42 (s, 1H), 5.29–5.19 (m, 2H), 4.27 (d, J = 6.4 Hz, 1H), 3.99 (d, J = 6.5 Hz, 1H),
2.09 (s, 3H), 2.08 (s, 3H), 1.96 (d, J = 2.5 Hz,
3H), 1.95 (d, J = 3.9 Hz, 3H), 1.93 (s, 3H).
To a stirred
solution of 8 (0.9 g, 2.04 mmol) in MeOH (10.0 mL) was
added Na metal (catalytic amount) slowly. The reaction mixture was
stirred for 2 h, then neutralized with Amberlite IR120, filtered,
and concentrated under reduced pressure to give a crude thiogalactoside.
The crude residue was dissolved in dry CH3CN (30 mL) and
treated with benzaldehyde dimethylacetal (0.41 mL, 2.75 mmol) followed
by camphorsulfonic acid (4.2 mg, 0.18 mmol) at room temperature. The
reaction mixture was stirred for 2.5 h, then neutralized with triethylamine
(0.5 mL), and concentrated under reduced pressure to give a crude
compound 9. After filtration through a short silica gel
column, compound 9 (0.52 g) was dissolved in dry DMF
(10 mL) and then treated with NaH (0.17 g, 4.15 mmol) and BnBr (2.27
g, 1.6 mL, 13.3 mmol) at 0 °C. The resulting solution was warmed
to room temperature and stirred for 0.5 h. Upon completion of the
reaction (TLC), the excess NaH was quenched using sat. NH4Cl solution. The product was extracted with EtOAc (3 × 25 mL),
and the combined organic layer was washed with brine (10 mL), dried
over Na2SO4, and concentrated under a high vacuum.
The crude product was purified via silica gel column chromatography
(30% EtOAc in hexane) to give 10 (0.63 g, 81%) as a colorless
oil. 1H NMR (400 MHz, CDCl3) δ 7.72–7.71
(m, 2H), 7.55–7.53 (m, 2H), 7.43–7.18 (m, 16H), 5.50
(s, 1H), 4.74–4.68 (m, 4H), 4.62 (d, J = 9.5
Hz, 1H), 4.16 (d, J = 3.2 Hz, 1H), 3.98 (br s, 1H),
3.91 (t, J = 9.4 Hz, 1H), 3.63 (dd, J = 9.2, 3.3 Hz, 1H), 3.42 (br s, 1H). 13C NMR (150 MHz,
CDCl3) δ 138.6, 138.2, 138.0, 132.9, 132.8, 129.2,
129.0, 128.6, 128.5, 128.4, 128.3, 128.0, 127.9, 127.6, 126.8, 101.5,
86.7, 81.5, 75.6, 75.5, 73.8, 72.0, 69.9, 69.3 (t, J = 21.7 Hz). HRMS (ESI) m/z calcd
for C33H31DO5SNa [M + Na]+, 564.1931; found, 564.1927.
Compound 11(43) (0.1 g, 0.346 mmol) was dissolved
in anhydrous dichloroethane (10.5
mL) and treated with (ethylthio)trimethylsilane (167 μL, 1.03
mmol) followed by ZnI2 (332 mg, 1.03 mmol). The reaction
mixture was stirred under an argon atmosphere at room temperature
for 3 h before it was diluted with ethyl acetate and filtered through
Celite. The organic layer was washed with aqueous saturated NaHCO3 and brine, dried over anhydrous Na2SO4, and concentrated to dryness. The crude residue was dissolved in
a mixture of THF:H2O (1:1, 20 mL), treated with K2CO3 (200 mg, 1.44 mmol), and stirred for 15 min. The reaction
mixture was then diluted with ethyl acetate, washed with water and
brine, and treated with anhydrous Na2SO4 before
it was concentrated to dryness. Then, the crude residue was passed
through a short pad of silica gel, eluting with hexane/ethyl acetate
(1:1), and the eluent was concentrated to dryness. The residue (60
mg) was dissolved in pyridine (1.4 mL), treated with acetic anhydride
(0.7 mL), and stirred overnight under an argon atmosphere at room
temperature. The reaction mixture was then concentrated to dryness,
and the crude reaction mixture was purified by silica gel column chromatography,
eluting with hexane/ethyl acetate (7:3), to obtain 12 (0.051 g, 40%). 1H NMR (400 MHz, CDCl3) δ
5.22 (t, J = 9.4 Hz, 1H), 5.08 (dd, J = 10.0, 9.4 Hz, 1H), 5.03 (dd, J = 10.0, 9.4 Hz,
1H), 4.49 (d, J = 10.0 Hz, 1H), 4.22 (d, J = 5.0 Hz, 1H), 3.70 (dd, J = 10.0, 5.0
Hz, 1H), 2.78–2.63 (m, 2H), 2.07 (s, 3H), 2.06 (s, 3H), 2.02
(s, 3H), 2.00 (s, 3H), 1.27 (t, J = 7.5 Hz, 3H). 13C NMR (100 MHz, CDCl3) δ 170.6, 170.2, 169.4,
83.5, 75.9, 73.9, 69.8, 68.3, 61.9 (t, J = 22.8 Hz),
24.2, 20.7, 20.6, 20.6, 14.8. HRMS (ESI) m/z calcd for C16H23DO9SNa
[M + Na]+, 416.1102; found, 416.1098.
Authors: Ciaran McDonnell; Oscar López; Paul Murphy; José G Fernández Bolaños; Rita Hazell; Mikael Bols Journal: J Am Chem Soc Date: 2004-10-06 Impact factor: 15.419
Authors: Michael G Pirrone; Marina Gysin; Klara Haldimann; Sven N Hobbie; Andrea Vasella; David Crich Journal: J Org Chem Date: 2020-09-23 Impact factor: 4.354
Figure 1
Figure 2
Figure 3
Scheme 1
Scheme 2
Scheme 3
Figure 4
Figure 5
Figure 6
Figure 7
Scheme 4