Acyl group migration and cleavage in selectively protected β-D-galactopyranosides as studied by NMR spectroscopy and kinetic calculations

Article abstract of DOI:10.1021/ja801177s

The migration of acetyl, pivaloyl, and benzoyl protective groups and their relative stabilities at variable pH for a series of β-D-galactopyranoses were studied by NMR spectroscopy. The clockwise and counterclockwise migration rates for the different este

Full text of DOI:10.1021/ja801177s

Published on Web 06/11/2008
Acyl Group Migration and Cleavage in Selectively Protected
ꢀ-D-Galactopyranosides as Studied by NMR Spectroscopy and
Kinetic Calculations
Mattias U. Roslund,*,† Olli Aitio,‡ Johan Wa¨rnå,# Hannu Maaheimo,§
Dmitry Yu. Murzin,# and Reko Leino*,†
Laboratory of Organic Chemistry and Laboratory of Industrial Chemistry, Åbo Akademi
UniVersity, FI-20500 Åbo, Finland, Program in Structural Biology and Biophysics, Institute of
Biotechnology, UniVersity of Helsinki, FI-00014 Helsinki, Finland, and VTT Technical Research
Centre of Finland, FI-00014 Helsinki, Finland
Received February 16, 2008; E-mail: mattias_roslund@yahoo.com; reko.leino@abo.fi
Abstract: The migration of acetyl, pivaloyl, and benzoyl protective groups and their relative stabilities at
variable pH for a series of ꢀ-D-galactopyranoses were studied by NMR spectroscopy. The clockwise and
counterclockwise migration rates for the different ester groups were accurately determined by use of a
kinetic model. The results presented provide new insights into the acid and base stabilities of commonly
used ester protecting groups and the phenomenon of acyl group migration and may prove useful in the
planning of synthesis strategies.
in the hexose series.6 Undesired migration may severly limit
the types of reactions that can be applied to partially acylated
Introduction
Protective groups play a central role in carbohydrate chemistry
in controlling the reactivity of both glycosyl donors and
acceptors, as well as in the control of regio- and stereoselec-
tivities and decomposition of target molecules and reactants.1
The strategies for the syntheses of partially protected mono-
and oligosaccharides commonly involve the use of orthogonal
protecting groups that should be selectively removed under the
desired reaction conditions. Various esters are commonly used
as protecting groups in carbohydrate syntheses.1a,2 It is well-
known that such groups may, in polyhydroxylic compounds,
be subject to undesired migration and cleavage. Acetyl group
migration was first described by Emil Fischer3 and the intramo-
lecular nature of the phenomenon was confirmed by a radioac-
tive labeling study by Doerschuk.4 Intermolecular rearrangement
for ester groups has likewise been proposed.5
sugars without inducing decomposition. Notably, intramolecular
migration has also been successfully utilized in syntheses and
improvement of yields in peptide chemistry and in carbohydrate
chemistry.7 Anticlockwise migration has been observed for
methyl 6-bromo-2,6-dideoxy-4-O-benzoyl-R-D-ribo-hexopyra-
noside8 and 5-O-acetyl-1,4-anhydro-6-thio-D-glucitol.9 Migration
under acidic conditions has also been reported,7b while being
disputed by some investigators.10
Generally, the migration of acyl groups is considered ir-
reversible11 and a large part of the earlier investigations have
dealt with acetyl group migration in partially acetylated sialic
acid derivatives12 and migration in acyl ꢀ-D-glucopyranosidu-
ronic acid derivatives.13 Mechanistically, a cyclic orthoester-
type intermediate has been suggested. Acyl groups with strongly
electron-attracting R-groups are reported to exist as cyclic
An acyl group may migrate to a neighboring OH under acidic
and neutral, but especially under basic conditions. Such migra-
tion has been reported to take place at all positions except C6
(6) (a) Bonner, W. A. J. Org. Chem. 1959, 24, 1388–1390. (b) Garegg,
P. J. Ark. Kemi 1964, 23, 255–268. (c) Fink, A. L.; Hay, G. W. Can.
J. Chem. 1969, 47, 845–852. (d) Binkley, R. W. Modern carbohydrate
chemistry. Marcel Dekker, Inc.: New York 1988, p 143. (e) Kurahashi,
T.; Mizutani, T.; Yoshida, J.-i J. Chem. Soc., Perkin Trans. 1 1999,
465–473.
† Laboratory of Organic Chemistry, Åbo Akademi University.
‡ University of Helsinki.
# Laboratory of Industrial Chemistry, Åbo Akademi University.
§ VTT Technical Research Centre of Finland.
(7) (a) Sohma, Y.; Sasaki, M.; Hayashi, Y.; Kimura, T.; Kiso, Y. Chem.
Commun. 2004, 124–125. (b) Horrobin, T.; Tran, C. H.; Crout, D.
J. Chem. Soc., Perkin Trans. 1 1998, 1069–1080. (c) Finan, P. A.;
Warren, W. D. J. Chem. Soc. 1962, 4214–4216.
(1) (a) Khan, S. H.; O’Neill, R. A. Modern Methods in Carbohydrate
Synthesis; Harwood Academic Publishers: Amsterdam, 1996. (b)
Hanessian, S. PreparatiVe Carbohydrate Chemistry; Marcel Dekker
Inc.: New York, 1997. (c) Bashkin, J. K. Chem. ReV. 2000, 100, 4265–
4712; Ed. Special Issue. (d) Levy, D. E.; Fu¨gedi, P. The Organic
Chemistry of Sugars; CRC Press: Boca Raton, 2006.
(2) (a) Greene, T. W.; Wuts, P. G. M. ProtectiVe Groups in Organic
Synthesis, 3rd ed.; John Wiley & Sons Inc.: New York, 1999. (b)
Jarowicki, K.; Kocienski, P. J. J. Chem. Soc., Perkin Trans. 1 1999,
1589–1615. (c) Jarowicki, K.; Kocienski, P. J. J. Chem. Soc., Perkin
Trans. 1 2000, 2495–2527. (d) Kocienski, P. J. Protecting Groups,
3rd ed.; Georg Thieme Verlag: Stuttgart, Germany, 2005.
(3) Fischer, E. Ber. Dtsch. Chem. Ges. 1920, 53, 1621–1633.
(4) Doerschuk, A. P. J. Am. Chem. Soc. 1952, 74, 4202–4203.
(5) Kro¨ger, L.; Thiem, J. Carbohydr. Res. 2007, 342, 467–481.
(8) Yasumori, T.; Sato, K.; Hashimoto Yoshimura, H. J. Bull. Chem. Soc.
Jpn. 1984, 57, 2538–2542.
(9) Schulze, O.; Bruns, S.; Voss, J.; Adiwidjaja, G. Carbohydr. Res. 2000,
329, 781–790.
(10) (a) Deferrari, J. O.; Gros, E. G.; Mastronardi, I. O. Carbohydr. Res.
1967, 4, 432–434. (b) Dick, W. E.; Baker, B. G.; Hodge, J. E.
Carbohydr. Res. 1968, 6, 52–62.
(11) Albert, R.; Dax, K.; Stu¨tz, A. E.; Weidmann, H. J. Carbohydr. Chem.
1983, 2, 279–292.
(12) (a) Reinhard, B.; Faillard, H. Liebigs. Ann. Chem. 1994, 193–203.
(b) Kamerling, J. P.; Schauer, R.; Shukla, A. K.; Stoll, S.; van Halbeek,
H.; Vliegenthart, J. F. G. Eur. J. Biochem. 1987, 162, 601–607.
9
10.1021/ja801177s CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 8769–8772 8769

A R T I C L E S
Roslund et al.
Scheme 1
Figure 1. Compounds investigated in the migration studies.
orthoesters, while benzoyl groups have less tendency to migrate
than the corresponding aliphatic acyl groups.4 Reversibly formed
tetrahedral intermediates in acyl transfer reactions have likewise
been studied in detail in terms of theoretical versus kinetic
data.14
As a part of our synthetic and conformational studies on
protected galactopyranosides,15 and owing to the scarce amount
of detailed investigations on the subject, we became interested
in the actual nature of the acyl group migration. Here, we report
our results on the stability of acyl protecting groups and the
migration kinetics of the 2 / 3 / 4 f 6 migration in a series
of benzyl 2-O-acyl-ꢀ-D-galactopyranoside model compounds
1-3 having ester groups of different stabilities, electronic
contributions on the carbonyl carbon, and steric bulk at the 2-O-
positions with free hydroxyls at C3, C4, and C6 (Figure 1).16
The anomeric position was locked in ꢀ-configuration with a
benzyl ether protective group being stable under the migration
conditions applied.
The NMR analyses were performed at 25 °C in D2O solutions
at variable pD values (1.0, 3.0, 6.8, 8.0, and 10.0). The
unbuffered D2O used had a pD ) 6.8 (measured pH* ) 6.4).
A solution with pD ) 8.0 was prepared from 10 mM
Na-phosphate buffer and pD was adjusted to pH* ) 7.6. At
pD ) 1.0 the hydrolyzed acid did not affect the pH* to any
large extent. The migration reactions were followed by 1H NMR
spectroscopy and all compounds formed (2-4) were character-
ized by 1H and 2D NMR techniques (for details, see Supporting
Information).19
Results and Discussion
The induced migrations, ultimately resulting in cleavage of
the ester group from C6 of the galactopyranoside, were followed
by NMR spectroscopy. The equation pD ) pH + 0.4 was used
for converting the pH meter reading (pH*) to pD.17 Under acidic
conditions (pD 1.0-3.0) no migration was observed for any of
the compounds studied, instead a direct hydrolysis of the ester
group took place. At pD ) 5.0 the compounds were still stable
but already at pD ) 6.8 acyl migration occurred.
The intramolecular migration of the acyl group at pD 6.8
was followed from C2 f C3 f C4 f C6 by NMR spectroscopy
after which the protecting group hydrolyzed to form the
respective acid. At pD ) 3.0, a slow cleavage of the acetyl
group without migration was observed (approximately 1%/
month). In earlier studies, polarity of the solvent, pH, and
temperature have been shown to influence the acyl migration
and hydrolysis rates.12a,18 In the present investigation, solvent
polarity was not addressed. A schematic picture of the migration
is given in Scheme 1.
The 1H chemical shifts of the migration products are
remarkably different and can be utilized for determination of
their relative populations by integration of the signal areas in
the 1H NMR spectra. Details on the chemical shifts are provided
in the Supporting Information (Tables S1-S3). First, the
migration and cleavage rate of the acetyl group (1–1c and 4)
were monitored by 1H NMR spectroscopy. The composition of
the reaction mixture versus time at pD ) 8.0 (pH* ) 7.6) is
illustrated in Figure 2. To our knowledge, the present investiga-
tion is the first report on the relative stabilities and reactivities
of the ꢀ-D-galactopyranose OH/OR groups (R ) Ac, Piv, and
Bz) in such compounds.
The pivaloyl migration at pD ) 8.0 was very slow in
comparison to that of the smaller acetyl group, and also slower
than the benzoyl group migration (for clockwise and counter-
clockwise migration rates, see Table 1). The clockwise- and
counterclockwise migration rates observed could not be ac-
curately determined on the basis of the populations determined
from the NMR spectra only. For accurate estimation of the
migration rates, kinetic calculations based on the experimentally
(13) (a) Skordi, E.; Wilson, I. D.; Lindon, J. C.; Nicholson, J. K. Xenobiotica
2005, 35, 715–725. (b) Johnson, C. H.; Wilson, I. D.; Harding, J. R.;
Stachulski, A. V.; Iddon, L.; Nicholson, J. K.; Lindon, J. C. Anal.
Chem. 2007, 79, 8720–8727.
(14) DeTar, D. F. J. Am. Chem. Soc. 1982, 104, 7205–7212.
(15) (a) Lehtila¨, R. L.; Lehtila¨, J. O.; Roslund, M. U.; Leino, R. Tetrahedron
2004, 60, 3653–3661. (b) Roslund, M. U.; Klika, K. D.; Lehtila¨, R. L.;
Ta¨htinen, P.; Sillanpa¨a¨, R.; Leino, R. J. Org. Chem. 2004, 69, 18–25.
(16) Compounds 1–3 were prepared according to a modified literature
procedure starting from D-galactose (for detailed synthesis and
characterization see ref 15a and Supporting Information): Peracety-
lation of D-galactose followed by transformation of the anomeric acetyl
into benzyl ether, deacetylation, TBDPS protection of the primary
hydroxyl group, introduction of the 3,4-O-acetal and ester protection
of 2-OH and the subsequent silyl ether deprotection by TBAF and
isopropylidene removal using Dowex H+ provided the corresponding
model compounds 1-3 in approximately 40% overall yields over eight
steps.
(18) (a) Dickinson, R. G.; Hooper, W. D.; Eadie, M. J. Drug Metab. Dispos.
1984, 12, 247–252. (b) Smith, P. C.; Hasegawa, J.; Langendijk,
P. N. J.; Benet, L. Z. Drug Metab. Dispos. 1985, 13, 110–112. (c)
Hyneck, M. L.; Munafo, A.; Benet, L. Z. Drug Metab. Dispos. 1988,
16, 322–324. (d) Mortensen, R. W.; Corcoran, O.; Cornett, C.;
Sidelmann, U. G.; Troke, J.; Lindon, J. C.; Nicholson, J. K.; Hansen,
S. H. Drug Metab. Dispos. 2001, 29, 375–380. (e) Mortensen, R. W.;
Sidelmann, U. G.; Tjørnelund, J.; Hansen, S. H. Chirality 2002, 14,
305–312.
(19) The (higher order) 1H NMR spectra for all of the synthesized and
purified compounds were analyzed with PERCH NMR software for
complete spectral analyses, see: Laatikainen, R.; Niemitz, M.; Weber,
U.; Sundelin, J.; Hassinen, T.; Vepsa¨la¨inen, J. J. Magn. Reson. Ser. A
1996, 120, 1–10. The acyl group migration was followed with Varian
500 and 600 MHz (equipped with a cold probe) Inova Spectrometers
and the migration products assigned by standard 2D NMR techniques.
(17) Bastardo, L. A.; Me´sza´ros, R.; Varga, I.; Gila´nyi, T.; Cleasson, P. M.
J. Phys. Chem. B 2005, 109, 16196–16202.
9
8770 J. AM. CHEM. SOC. VOL. 130, NO. 27, 2008

Acyl Group Migration in Galactopyranosides
A R T I C L E S
the rate constants k3, k-3, and k4 could not be accurately
determined (n.d.) for the pivaloyl group migration within this
time frame (Table 1). The C2fC3 migration for the acetyl group
was faster than for the benzoyl (4.91 versus 2.92) and also faster
than the C4 f C6 migration (10.4 versus 1.91). The migration
from C3 f C4 was rather similar for both (17.6 versus 18.8),
whereas the benzoyl group was hydrolyzed much faster than
the acetyl (1.54 versus 0.04) (Table 1). In the literature, such
influence of steric and electronic effects on the reactivity of
esters has been studied frequently.4,21
Summary and Conclusions
To summarize, new data on ester protective groups and their
relative stabilities at variable pH for a series of ꢀ-D-galactopy-
ranoses has been obtained. The clockwise and counterclockwise
migration rates (Table 1) for three different ester groups were
accurately determined from a kinetic model for the first time.
The results presented may prove useful in the planning of
oligosaccharide syntheses and sugar analyses. A significant
increase in the migration rates to and from the axial 4-OH
position (k2 and k-2, Table 1) over the equatorial positions for
the hydroxyl groups in the ꢀ-D-galactopyranose were observed.
The larger ester group used migrated slower, possibly due to
blocking of the nucleophilic attack of the neighboring hydroxyl
group on the ester carbonyl carbon. The chemical shifts of the
carbonyl carbons in compounds 1-3 (167.6 Bz, 172.3 Ac, and
179.5 Piv) could not directly be correlated with the faster
migration of the benzoyl group.
The partially reversible model employed here, showed a much
better fit with the experimental values than a clockwise migration
solution first screened. In future work, theoretical calculations
could be utilized to estimate the magnitude of energy difference
between the various structures and provide an indication of the
transition state(s) of the reaction. At present this approach is
hampered by difficulties related to the structure, complexation,
and polarization of the reaction intermediates.
Figure 2. Migration of acetyl group (compound 1) at pD ) 8.0 and T )
25 °C in buffered D2O.
Table 1. Migration First Order Forward (k+) and Backward (k-)
Relative Rate Constant for Compounds 1-3 at pD ) 8.0 (n.d. )
Not Determined)
k/h-1
1 Ac
2 Piv
3 Bz
k1
k-1
k2
k-2
k3
k-3
k4
4.91 ( 0.032
2.65 ( 0.068
17.6 ( 0.44
12.5 ( 0.46
10.4 ( 0.14
2.48 ( 0.06
0.04 ( 0.0048
0.06 ( 0.004
0.10 ( 0.04
0.61 ( 0.13
0.26 ( 0.09
n.d.
2.92 ( 0.07
4.11 ( 0.32
18.8 ( 2.1
10.3 ( 1.56
1.91 ( 0.06
0.0
n.d.
n.d.
1.54 ( 0.1
determined populations were utilized. Being more hindered and
sterically shielded toward nucleophilic attack, the pivaloyl and
benzoyl ester groups should be less prone to migration when
compared with the acetyl group (Figure 2 and Supporting
Information). The pivaloyl group is the slowest to migrate while
the Ac and Bz groups migrate with similar rates at pD ) 8.0.
While at pD ) 10.0 the acetyl group migration could still be
determined by NMR spectroscopy, the migration of the 2-OBz
Experimental Section
The detailed experimental procedures for the preparation of
compounds 1–3 are given in the Supporting Information. The NMR
samples for the migration studies were prepared as follows: 1.0-2.4
mg of the benzyl 2-O-ester-ꢀ-D-galactopyranosides were dissolved
in 500 µL of D2O (pD ) 6.8) or 10 mM Na-phosphate buffer in
D2O, pD 8.0. For the higher pD/pH conditions the pH* values were
adjusted to 7.6 or 9.6 (pD ) 8.0 and 10.0) by the addition of NaOD.
The 2-O-benzoyl derivative of benzyl galactose did not dissolve
completely and the undissolved part was removed by filtration
through a 0.45 µm membrane filter directly coupled to the NMR
tube. For the low pD studies, the compounds were dissolved in
500 µL of DCl/D2O, (pH* ) 0.6), all the pD values were verified
by a pH meter (for details, see Supporting Information).
The NMR spectra for benzoyl group migration were recorded
using a Varian Unity INOVA 600 MHz spectrometer equipped with
a 15N/13C/1H triple resonance cold probe for enhanced sensitivity.
NMR experiments for pivaloyl and acetyl group migration were
carried out on a Varian Unity 500 MHz spectrometer at 298 K.
The time evolution of the samples was followed by recording
standard 1H spectra with 16 transients. Presaturation was used for
1
was too fast to be followed by H NMR. At low pH (pD )
1.0), the acetyl group was most readily cleaved in comparison
with the pivaloyl and benzoyl esters. After treatment of 1 at
pD ) 1.0 for 1 week, more than 50% of the hydrolyzed product
[benzyl ꢀ-D-galactopyranose (4)] was formed. The direct
hydrolysis of Ac was very fast at low pH while the Bz hydrolysis
with the highest rate was observed after initial migration to C6
at high pH (Table 1). As shown in Figure 2, the calculations
using the constants reported in Table 1 were in good agreement
with the experimental values.20
The migration reactions described with k1 to k3 were assumed
to be reversible whereas the final hydrolyzation step reaction
k4 was assumed to be irreversible. All the migration rates for
the pivaloyl protected galactopyranoside 2 were much smaller
than for the acetyl and benzoyl protected analogues 1 and 3.
Especially k3 and k4 were very small and due to the low
concentration of benzyl 4-O-pivaloyl-ꢀ-D-galactopyranoside 2b
(20) The model predictions were obtained by solving the differential
equations for all of the components during the parameter estimation.
A stiff ODE-solver was used to solve the system of ODEs, with the
backward difference method implemented in the MODEST software
used in the estimation of the kinetic parameters, see: Haario, H. Modest
6.0-A User’s Guide; ProfMath: Helsinki, 2001. See also: Mastihubova´,
M.; Biely, P. Carbohydr. Res. 2004, 339, 1353–1360.
(21) (a) Jensen, H. H.; Bols, M. Org. Lett. 2003, 5, 3419–3421. (b)
Moitessier, N.; Englebiennea, P.; Chapleur, Y. Tetrahedron 2005, 61,
6839–6853. (c) Regan, C. K.; Craig, S. L.; Brauman, J. I. Science
2002, 295, 2245–2247. (d) Illa, O.; Bagan, X.; Cazorla, A. M.; Lyon,
C.; Baceiredo, A.; Branchadell, V.; Ortun˜o, R. M. J. Org. Chem. 2006,
71, 5320–5327. (e) Uggerud, E. J. Phys. Org. Chem. 2006, 19, 461–
466.
9
J. AM. CHEM. SOC. VOL. 130, NO. 27, 2008 8771

A R T I C L E S
Roslund et al.
the water suppression for the benzyl 2-O-benzoyl-ꢀ-D-galactopy-
ranoside (3) sample. The migration products were identified and
assigned from 1H, DQF-COSY, TOCSY, and HSQC spectra. The
spectra were processed as described in the Supporting Information.
The kinetics of the migration process was described with a first-
order reversible model for reactions 1-3 and an irreversible first
order model for reaction 4.
tions and the concentrations predicted by the model, respectively.
The software Modest used to estimate the rate constants minimizes
the objective function with the Levenberg-Marquardt method and
solves the ordinary differential equations describing the reactor
model by the backward difference method. The fit of the model to
experimental data for the migration process are presented in the
Supporting Information (Figures S8-S10). As seen from the figures,
the model well describes the experimental data. The rate constants
are listed in Tables S4 and S5 (Supporting Information) showing
the initial rates of reaction 1 and initial concentrations of reactant
1. The reaction is the fastest for the acetyl group migration and the
slowest for the pivaloyl group migration.
r1 ) k1c1 - k-1c1a
r2 ) k2c1a - k-2c1b
r3 ) k3c1b - k-3c1c
r4 ) k4c1c-
(1)
(2)
(3)
(4)
Acknowledgment. Financial support from the Glycoscience
Graduate School, Research Foundation of Orion Corporation, and
Stiftelsens fo¨r Åbo Akademi forskningsinstitut are gratefully
acknowledged (M.U.R.). The authors also wish to thank Markku
Reunanen for the EIMS analyses.
The model predictions were obtained by solving the differential
equations for a batch reactor model.
dc
1 ) -r1,
dt
dc
1a ) r1 - r2,
dt
dc1b
dt
)
dc
1c ) r3 - r4,
dt
dc
r2 - r3,
cleaved ) r4
Supporting Information Available: Complete experimental
procedures; product characterization; migration details for the
different ester groups at different pH values; complete kinetic
data; the predicted migration rate constants using experimentally
recorded concentrations. This material is available free of charge
via the Internet at http://pubs.acs.org.
dt
For the parameter estimation the following objective function was
minimized.
2
Q )
ci,t,exp-ci,t,model
( )
∑∑
i
t
where ci,t,exp and ci,t,model are the experimentally recorded concentra-
JA801177S
9
8772 J. AM. CHEM. SOC. VOL. 130, NO. 27, 2008