Unexpected stability of aryl β-N-acetylneuraminides in neutral solution: Biological implications for sialyl transfer reactions

Article abstract of DOI:10.1021/ja042280e

A reagent panel comprised of seven aryl β-D-N-acetylneuraminides was synthesized and then used to probe the mechanisms of nonenzymatic hydrolysis. These reactions proceeded via four independent pathways: (1) acid-catalyzed hydrolysis of the neutral molecule; (2) acid-catalyzed hydrolysis of the anionic form, or its kinetic equivalent spontaneous hydrolysis of the neutral form; (3) spontaneous hydrolysis of the anionic form; and (4) a base-promoted pathway. The pH-independent spontaneous hydrolysis of 4-nitrophenyl α-D-N- acetylneuraminide (5) occurs at a rate that is over 100 times faster than that of the corresponding reaction of 4-nitrophenyl β-D-N-acetylneuraminide (4a). Spontaneous hydrolyses of four aryl β-D-N-acetylneuraminides displayed a β_Ig value of -1.24 ± 0.16 (pH = 8.1, T = 100 °C), and at a pH value of 1.0 (50 °C), all seven panel members gave a β_Ig value of 0.14 ± 0.08. The aqueous ethanolyses of 4a and 5 gave similar products and displayed sensitivity parameters (m) in a standard Winstein-Grunwald analysis of -0.04 ± 0.01 and ± 0.23 ± 0.02, respectively. These results, plus the activation parameters calculated for the spontaneous hydrolyses of the anionic forms of 5 (ΔH? = 116 ± 2 kJ mol^-1 and ΔS? = 27 ± 4 J mol^-1 K^-1) and 4a (ΔH? = 138 ± 3 kJ mol^-1 and ΔS? = 59 ± 8 J mol ^-1 K^-1), are inconsistent with anomeric carboxylate assistance occurring during the hydrolysis reactions, and the likely cause for the enhanced reactivity of 5 in comparison to that of 4a is an increase in ground-state steric strain.

Full text of DOI:10.1021/ja042280e

Published on Web 05/03/2005
Unexpected Stability of Aryl â-N-Acetylneuraminides in
Neutral Solution: Biological Implications for Sialyl Transfer
Reactions
Veedeeta Dookhun and Andrew J. Bennet*
Contribution from the Department of Chemistry, Simon Fraser UniVersity, 8888 UniVersity
DriVe, Burnaby, B.C., Canada V5A 1S6
Received December 22, 2004; E-mail: bennet@sfu.ca
Abstract: A reagent panel comprised of seven aryl â-D-N-acetylneuraminides was synthesized and then
used to probe the mechanisms of nonenzymatic hydrolysis. These reactions proceeded via four independent
pathways: (1) acid-catalyzed hydrolysis of the neutral molecule; (2) acid-catalyzed hydrolysis of the anionic
form, or its kinetic equivalent spontaneous hydrolysis of the neutral form; (3) spontaneous hydrolysis of
the anionic form; and (4) a base-promoted pathway. The pH-independent spontaneous hydrolysis of
4-nitrophenyl R-D-N-acetylneuraminide (5) occurs at a rate that is over 100 times faster than that of the
corresponding reaction of 4-nitrophenyl â-D-N-acetylneuraminide (4a). Spontaneous hydrolyses of four aryl
â-D-N-acetylneuraminides displayed a âlg value of -1.24 ( 0.16 (pH ) 8.1, T ) 100 °C), and at a pH value
of 1.0 (50 °C), all seven panel members gave a âlg value of 0.14 ( 0.08. The aqueous ethanolyses of 4a
and 5 gave similar products and displayed sensitivity parameters (m) in a standard Winstein-Grunwald
analysis of -0.04 ( 0.01 and +0.23 ( 0.02, respectively. These results, plus the activation parameters
calculated for the spontaneous hydrolyses of the anionic forms of 5 (∆Hq ) 116 ( 2 kJ mol-1 and
∆Sq ) 27 ( 4 J mol-1 K-1) and 4a (∆Hq ) 138 ( 3 kJ mol-1 and ∆Sq ) 59 ( 8 J mol-1 K-1), are inconsistent
with anomeric carboxylate assistance occurring during the hydrolysis reactions, and the likely cause for
the enhanced reactivity of 5 in comparison to that of 4a is an increase in ground-state steric strain.
Introduction
sialic acid from an external source onto its own surface
carbohydrates;11 and (c) sialyltransferases are inverting enzymes
Many specific biological processes are mediated by carbo-
hydrate/protein interactions, examples include receptor/ligand
recognition and glycoconjugate intracellular trafficking.1,2 In
many of these biologically important interactions, the sugar
component contains N-acetylneuraminic acid (sialic acid), a
common constituent of the glycolipids and glycoproteins
produced by animal cells.3 Indeed, N-acetylneuraminic acid is
frequently found at the nonreducing termini of cell surface
oligosaccharide moieties,4 and this prominence is in keeping
with its importance in cellular and molecular recognition events.5
Several distinct families of sialyl-transferring enzymes are
found in nature and these include the following: (a) exo-
sialidases (N-acetylneuraminyl glycohydrolases, EC 3.2.1.18)
are retaining glycosidases that remove sialic acid from
glycoconjugates;4,6-10 (b) trans-sialidases are retaining enzymes
expressed by trypanosomes that enable the parasite to transfer
that utilize CMP-â-D-N-acetylneuraminide (1) as the sialyl donor
to make R-sialoside linkages.12 A fourth family, the endo-
sialidases (EC 3.2.1.129),13 has been reported, although at the
present time, it is not known whether these enzymes operate
via a retaining or an inverting mechanism.
Retaining glycosidases generally operate via a process in
which an active site carboxylate group acts as the nucleophile
in a double displacement mechanism.14-18 Such a process
(6) Saito, M.; Yu, R. K. Biochemistry and function of sialidases. In Biology
of the Sialic Acids; Rosenberg, A., Ed.; Plenum Press: New York, 1995;
pp 261-313.
(7) Wilson, J. C.; Angus, D. I.; von Itzstein, M. J. Am. Chem. Soc. 1995, 117,
4214-4217.
(8) Henrissat, B.; Bairoch, A. Biochem. J. 1993, 293, 781-788.
(9) Henrissat, B. Biochem. J. 1991, 280, 309-316.
(10) Henrissat, B.; Bairoch, A. Biochem. J. 1996, 316, 695-696.
(11) Kelly, J. M. Molecular Mechanisms of Pathogenesis in Chagas Disease;
Landes Bioscience: Georgetown, TX, 2003.
(12) Coutinho, P. M.; Deleury, E.; Davies, G. J.; Henrissat, B. J. Mol. Biol.
2003, 328, 307-317.
(13) Muhlenhoff, M.; Stummeyer, K.; Grove, M.; Sauerborn, M.; Gerardy-
Schahn, R. J. Biol. Chem. 2003, 278, 12634-12644.
(14) Sinnott, M. L. Chem. ReV. 1990, 90, 1171-1202.
(15) McCarter, J. D.; Withers, S. G. Curr. Opin. Struct. Biol. 1994, 4, 885-
892.
(1) Rademacher, T. W.; Parekh, R. B.; Dwek, R. A. Annu. ReV. Biochem. 1988,
57, 785-838.
(2) Drickamer, K. Cell 1991, 67, 1029-1032.
(3) Reutter, W.; Eckart, K.; Bauer, C.; Gerok, W. Biological significance of
sialic acids. In Sialic Acids, Chemistry, Metabolism, and Function; Schauer,
R., Ed.; Springer-Verlag: Vienna, 1982; Vol. 10, pp 264-305.
(4) Schauer, R.; Kelm, S.; Reuter, G.; Roggentin, P. Biochemistry and role of
sialic acids. In Biology of the Sialic Acids; Rosenberg, A., Ed.; Plenum
Press: New York, 1995; pp 7-67.
(16) Davies, G.; Sinnott, M. L.; Withers, S. G. Glycosyl transfer. In Compre-
hensiVe Biological Catalysis; Sinnott, M. L., Ed.; Academic Press: San
Diego, 1998; pp 119-209.
(5) Taylor, G. Curr. Opin. Struct. Biol. 1996, 6, 830-837.
(17) Zechel, D. L.; Withers, S. G. Acc. Chem. Res. 2000, 33, 11-18.
9
7458
J. AM. CHEM. SOC. 2005, 127, 7458-7465
10.1021/ja042280e CCC: $30.25 © 2005 American Chemical Society

Stability of Aryl
â-N-Acetylneuraminides in Neutral Solution
A R T I C L E S
Scheme 1. Mechanism of Glycosylation in Sialidases
Experimental Section
generates a covalent acylal enzyme intermediate and thus avoids
formation of a reactive high-energy oxacarbenium ion. In
aqueous solution, the lifetime of the quintessential carbohydrate
oxacarbenium ion, the glucopyranosylium ion 2, has been
estimated to be around (1-3) × 10-12 s.19,20
Materials and Methods. The buffers 2-(N-morpholino)ethane-
sulfonic acid (MES), N-tris[hydroxymethyl]methyl-3-amino-propane-
sulfonic acid (TAPS), and 3-(cyclohexylamino)-1-propanesulfonic acid
(CAPS) were purchased from Sigma and used without further purifica-
tion. N-Acetylneuraminic acid was purchased from Rose Scientific and
used without further purification. Milli-Q water (18.2 MΩ cm-1) was
used for kinetic and product study experiments. Perchloric acid and
sodium hydroxide solutions were made by dilutions of standardized
1.00 M solutions. All other salts used in the hydrolysis runs were of
analytical grade and were used without purification. CH2Cl2 was dried
by distillation from CaH2. NMR spectra were acquired on a Bruker
AMX-400 spectrometer.
The results contained in recent reports point to the conserved
tyrosine residue, rather than the catalytic glutamic acid, as being
the nucleophile for both classes of retaining sialyl transferring
enzymes, that is, exo-sialidases21 and trans-sialidases.22,23 Thus,
the enzymatic mechanism involves formation of an enzyme-
bound aryl â-N-acetylneuraminyl intermediate (Scheme 1,
â-NeuAc-OTyr).21 A central feature of this mechanism is that
formation of an N-acetylneuraminosylium ion (sialyl oxacar-
benium ion; 3) intermediate is avoided, a species that has an
estimated lifetime in aqueous solution of g3 × 10-11 s.24,25
Synthesis of Aryl â-D-N-Acetylneuraminide. The protected â-D-
N-acetylneuraminyl fluoride 626 and 4-nitrophenyl R-D-N-acetyl-
neuraminide 527,28 were synthesized according to published procedures.
Full experimental details for the synthesis of 4-nitrophenyl â-D-N-
acetylneuraminide (4a) are given below, while the experimental
particulars for the synthesis and characterization of 4b-g are given in
the accompanying Supporting Information.
The present report details a series of kinetic and product
studies performed, in aqueous media, on analogues of the
sialosyl-enzyme intermediate (Scheme 1, â-NeuAc-OTyr), that
is, aryl â-D-N-acetylneuraminides (4a-g). In addition, a specific
comparison is made between 4a and 4-nitrophenyl R-D-N-
acetylneuraminide (5), the anomer of 4a. The results from these
studies address critical questions concerning the intrinsic
reactivity of sialosides. For example, does the C-1 carboxylate
group of the N-acetylneuraminides function intramolecularly as
a nucleophile? Are the spontaneous solvolysis reactions of
N-acetylneuraminides associative (SN2) or dissociative (SN1) in
nature?
Methyl [4-Nitrophenyl(5-acetamido-4,7,8,9-tetra-O-acetyl-3,5-
dideoxy-D-glycero-â-D-galacto-non-2-ulopyranosyl)]onate (7a). To
dried powdered 4 Å molecular sieves was added a mixture of
4-nitrophenol (580 mg, 4.17 mmol) and sialosyl fluoride 6 (400 mg,
0.83 mmol), and the resultant solid mixture was maintained under
vacuum (6 mmHg) for 30 min. The solid mixture was then placed under
a N2 atmosphere, and dry CH2Cl2 (30 mL) was added via a syringe;
this mixture was stirred for 1 h. Then a solution of BF3‚OEt2
(0.72 mL, 0.83 mmol) in CH2Cl2 (4 mL) was added, and the mixture
was stirred overnight at room temperature. The mixture was filtered,
and the solid residue was washed thoroughly with CH2Cl2. The
combined filtrates were washed with saturated NaHCO3 (150 mL), water
(150 mL), and brine (150 mL), and the resulting solution was dried
Na2SO4. A pale yellow syrup was obtained after evaporation of the
solvent, and this material was crystallized from Et2O to give a white
powder (196 mg, 39% yield), mp 187-189 °C. [R]20D) -77.8
(c ) 1.26, CH2Cl2). 1H NMR (400 MHz, CDCl3) δ: 1.75, 1.89, 2.06,
2.08, 2.17 (5 × s, 15 H, CH3), 2.03 (m, 1 H, H-3a), 2.68 (dd, 1 H,
J3e,3a ) 12.9 Hz, J3e,4 ) 5.0 Hz, H-3e), 3.76 (s, 3 H, OCH3), 4.04
(dd, 1 H, J6,5 ) 10.4 Hz, J6,7 ) 2.4 Hz, H-6), 4.12 (dd, 1 H, J9a,9b
)
(21) Watson, J. N.; Dookhun, V.; Borgford, T. J.; Bennet, A. J. Biochemistry
2003, 42, 12682-12690.
(22) Watts, A. G.; Damager, I.; Amaya, M. L.; Buschiazzo, A.; Alzari, P.; Frasch,
A. C.; Withers, S. G. J. Am. Chem. Soc. 2003, 125, 7532-7533.
(23) Amaya, M. F.; Watts, A. G.; Damager, T.; Wehenkel, A.; Nguyen, T.;
Buschiazzo, A.; Paris, G.; Frasch, A. C.; Withers, S. G.; Alzari, P. M.
Structure 2004, 12, 775-784.
(24) Horenstein, B. A.; Bruner, M. J. Am. Chem. Soc. 1996, 118, 10371-10379.
(25) Horenstein, B. A.; Bruner, M. J. Am. Chem. Soc. 1998, 120, 1357-1362.
(26) Sharma, M. N.; Eby, R. Carbohydr. Res. 1984, 127, 201-210.
(27) Eschenfelder, V.; Brossmer, R. Carbohydr. Res. 1987, 162, 294-297.
(28) Rothermel, J.; Faillard, H. Carbohydr. Res. 1990, 196, 29-40.
(18) Vasella, A.; Davies, G. J.; Bo¨hm, M. Curr. Opin. Chem. Biol. 2002, 6,
619-629.
(19) Amyes, T. L.; Jencks, W. P. J. Am. Chem. Soc. 1989, 111, 7888-7900.
(20) Zhu, J.; Bennet, A. J. J. Am. Chem. Soc. 1998, 120, 3887-3893.
9
J. AM. CHEM. SOC. VOL. 127, NO. 20, 2005 7459

A R T I C L E S
Dookhun and Bennet
12.5 Hz, J9a,8 ) 6.8 Hz, H-9a), 4.20 (q, 1 H, J5,4 + J5,6 + J5,NH ) 31.2
was used. Ionic strength was maintained at 0.3 M with sodium
perchlorate except when [H+] or [OH-] > 0.3 M.
Hz, H-5), 4.64 (ddd, 1 H, J9b,8 ) 2.7 Hz, H-9b), 4.90 (ddd, 1 H, J8,7
)
8.7, H-8), 5.26 (d, 1 H, JNH,5 ) 10.3 Hz, NH), 5.35 (dd, 1 H, H-7),
5.48 (td, 1 H, J4,3a + J4,5 ) 21.8 Hz, J4,3e ) 5.0 Hz, H-4), 7.08-7.15
(m, 2 H, Ar-H), 8.12-8.22 (m, 2 H, Ar-H). 13C NMR (100 MHz,
CDCl3) δ: 20.6. 20.7, 20.8, 23.1 (CH3), 38.5 (C-3), 49.1 (C-5), 53.5
(OCH3), 61.8 (C-9), 67.8 (C-4, C-7), 71.5 (C-8), 72.9 (C-6), 99.7
(C-2), 116.8, 125.9, 143.1, 158.8 (Ar-C), 170.0, 170.2, 170.3, 170.5,
170.9 (CdO, C-1). Anal. Calcd for C26H32N2O15: C, 50.98; H, 5.27;
N, 4.57. Found: C, 50.75; H, 5.44; N, 4.34.
The hydrolysis reactions of 4a were monitored using three different
protocols. For pH values less than 4.50 and greater than 11.20, a
methanol stock of 4a (50 µL, 2.3 µM) was injected into a cuvette
containing 3 mL of the required buffer that had been pre-equilibrated
for 10 min at 50 °C. The absorbance versus time data were monitored
for 3 half-lives at 340 and 408 nm for the low and high pH ranges,
respectively. Rate constants were calculated by fitting the absorbance
versus time data to a standard first-order rate equation using the
nonlinear least-squares routine computer program, Grafit.
4-Nitrophenyl (5-Acetamido-3,5-dideoxy-D-glycero-â-D-galacto-
non-2-ulopyranosylonic Acid) (4a). To a solution of 7a (50 mg,
0.08 mmol) in anhydrous methanol (5 mL) was added a methanolic
sodium methoxide solution (5 equiv), and this mixture solution was
stirred at 0 °C for 15 min. Dowex 50W HCR-W2(H+) cation exchange
resin (prewashed with methanol) was added to neutralize the solution.
After removal of the resin by filtration, it was washed several times
with methanol. The combined solvent was evaporated under reduced
pressure, and the resultant residue was dissolved in 3:1 v/v THF/water
(2 mL) at 0 °C. To this solution was added LiOH‚H2O (16 mg,
2.5 equiv). After stirring at 0 °C for 25 min, the mixture was neutralized
with Dowex 50W HCR-W2(H+) resin. Following removal of the resin,
the filtrate was concentrated under reduced pressure. The remaining
aqueous solution was then lyophilized to give a white solid (33 mg,
92%). [R]20D) -86.9 (c ) 1.22, H2O). 1H NMR (400 MHz, D2O)
δ: 1.85 (t, 1 H, J3e,3a + J3a,4 ) 24.5 Hz, H-3a), 2.04 (s, 3 H, CH3),
2.60 (dd, 1 H, J3e,3a ) 12.9 Hz, J3e,4 ) 5.0 Hz, H-3e), 3.45 (d, 1 H,
J7,8 ) 9.0 Hz, H-7), 3.60 (dd, 1 H, J9a,9b ) 11.8 Hz, J9a,8 ) 5.2 Hz,
For pH values between 4.50 and 6.33, a discontinuous assay method
was used. A solution of 4a (4 mg) in 20 mL of the required buffer was
aliquoted into glass ampules (1 mL of solution) which were then sealed.
The ampules were placed in a water bath held at 50 °C. Ampules were
taken out of the water bath at regular intervals, and the reaction was
quenched by cooling to 0 °C. To determine the reaction end point, two
ampules were heated at 100 °C for 2 days. A portion of the solution
from each ampule (200 µL) was added to pH 10.0 CAPS buffer
(0.3 M, 400 µL), and the absorbance of the resultant mixture was
measured at 408 nM. These reactions were monitored in this fashion
until they were about 5% complete.
For pH values between 6.40 and 11.20, the hydrolysis of 4a was
followed by observing the change in absorbance at 408 nm at 50 °C
until about 3-4% of the hydrolytic reaction had occurred. Subsequently,
the reaction end-points were determined by increasing the cell-block
temperature to 75 °C and monitoring the absorbance until the reaction
was complete. Rate constants for the reactions performed between pH
values of 4.50 and 11.20 were calculated using the equation
ln([A]t/[A]0) ) -k × t. The hydrolysis reactions of 5 were monitored
using identical protocols.
H-9a), 3.64-3.77 (m, 4 H, NH, H-6, H-8, H-9b), 4.00 (t, 1 H, J5,4
J5,6 ) 20.8 Hz, H-5), 4.25 (td, 1 H, J4,3a + J4,5 ) 21.8 Hz, J4,3e
+
)
5.0 Hz, H-4), 6.85-7.21 (m, 2 H, Ar-H), 8.20-8.38 (m, 2 H, Ar-
H). 13C NMR (100 MHz, D2O) δ: 19.0 (CH3), 37.5 (C-3), 48.6 (C-5),
60.3 (C-9), 63.4 (C-4), 65.2 (C-7), 66.8 (C-8), 68.5 (C-6), 98.1 (C-2),
113.8, 122.7, 138.9, 156.8 (Ar-C), 170.3, 171.7. HRMS(FAB) m/z
(M - H+), C17H21N2O11 requires 419.1145, found 429.1156.
4-Nitrophenyl ((3R)-5-Acetamido-3-deuterio-3,5-dideoxy-D-gly-
cero-â-D-galacto-non-2-ulopyranosylonic Acid) (4a-2H). Deuterium
was incorporated stereospecifically on C-3 in N-acetylneuraminic acid
according to a published procedure.29 Following the base-catalyzed
exchange reaction, 4-nitrophenyl ((3R)-5-acetamido-3-deuterio-3,
5-dideoxy-D-glycero-â-D-galacto-non-2-ulopyranosylonic acid) was
made using the same procedure as outlined above. The NMR spectrum
of 4a-2H was identical to that of the unlabeled compound (4a) except
for the absence of the apparent triplet at 1.85 ppm and the presence of
a doublet at 2.60 ppm instead of a pair of doublets.
Solvolysis Kinetics. The aqueous ethanolyses of 4a and 5 were
monitored in a manner similar to that of the corresponding hydrolysis
reactions. Specifically, reactions were followed at 408 nm in 15 mM
phosphate buffer (Na2HPO4:NaH2PO4 ) 2:1; I ) 35 mM) at 75 °C.
Brønsted Plots. For the acid-catalyzed region (pH ) 1.00), the
reactions were monitored directly by following the change in absorbance
versus time. Absorbance values were recorded at wavelengths of 340
and 270 nm for 4b and 4e, respectively, while for all other substrates
(4c, 4d, 4f, and 4g), the wavelength utilized was 280 nm. In the
spontaneous hydrolysis regime, the hydrolyses of 4b-d were monitored
at a pH of 8.08 using ampules as outlined above. The ampules were
maintained at a temperature of 100 °C by placing them in a “boiler” in
which they were immersed in the vapor generated from a boiling water
bath. Ampules were taken out at regular intervals, and the reactions
were quenched by cooling the ampule in an ice/water bath. These
reactions were generally monitored for a time period that corresponds
to about 2-3 half-lives for hydrolysis. The rate constant for hydrolysis
of 4a under these conditions (pH 8.08, T ) 100 °C) was estimated by
extrapolation of rate constant data measured at 50, 65, 75, and 85 °C.
Product Studies. The hydrolyses of 4a and 5 (3 mg) in ethanol/
water mixtures (2 mL) containing 3 equiv of N-methylmorpholine were
performed in sealed glass ampules at 75 °C. The reactions were allowed
to proceed for about 9 half-lives. The resultant solution was lyophilized,
4-Nitrophenyl (5-Acetamido-3,5-dideoxy-D-glycero-R- and
â-D-galacto-non-2-ulopyranosylonic (18O1)-Acid) (4a-18O and 5-18O).
Oxygen-18 was incorporated into the respective carboxylate groups
by performing the final hydrolytic deprotection of the respective methyl
esters in THF/H218O (3:1 v/v, 5 mL) containing LiOH‚H2O (3 equiv)
at 0 °C for 30 min.
Hydrolysis Kinetics. The hydrolysis reactions were monitored by
measuring absorbance versus time data using a Cary-3E UV-vis
spectrophotometer equipped with the Cary Six-Cell Peltier constant
temperature accessory. The buffers and their associated pH ranges used
in this study were: malonic acid, 2.10-3.19; succinic acid, 3.76-4.02;
acetic acid, 4.46-5.36; N-methylmorpholinoethanesulfoninc acid (MES),
5.50-6.30; N-[tris(hydroxymethyl)methyl]-3-aminopropane-1-sulfonic
acid (TAPS), 7.00-8.20; and 3-(cyclohexylamino)propane-1-sulfonic
acid (CAPS), 10.00-11.13. Buffer pH values were measured at room
temperature, and no temperature corrections were applied. For solution
where [H+] > 0.01 M, perchloric acid-sodium perchlorate was used,
and when [OH-] > 0.01 M, sodium hydroxide-sodium perchlorate
1
and the solid residue was analyzed using H NMR spectroscopy.
18O Exchange Experiments. 18O-Incorporation into the 4-nitro-
phenol product was monitored for the base-promoted reactions of 4a
and 5 at [OH-] ) 1.0 M. Specifically, an approximate 1:1 H216O/H218O
mixture was made by adding equal volumes of H218O (95.1 atom %
18O; Marshall Isotopes Ltd., batch number 020414nw) and freshly made-
up aqueous NaOH (2.0 M). To this aqueous media (0.4 mL) was added
either 4a or 5 (2 mg, 4.6 µmol), and the resulting solutions were
maintained at 50 °C for approximately 5 half-lives for hydrolysis.
Following neutralization with an excess of malonate buffer (0.3 M,
5 mL, pH 2.69), the resultant aqueous media were extracted with
distilled dichloromethane (2 × 2 mL). The combined organic extracts
(29) Ashwell, M.; Guo, X.; Sinnott, M. L. J. Am. Chem. Soc. 1992, 114, 10158-
10166.
9
7460 J. AM. CHEM. SOC. VOL. 127, NO. 20, 2005

Stability of Aryl
â-N-Acetylneuraminides in Neutral Solution
A R T I C L E S
Table 1. Calculated pKa Values and Rate Constants for the
Hydrolyses of 4-nitrophenyl R- and â-D-N-acetylneuraminides
(5 and 4a, respectively) at 50 °C (µ ) 0.3, NaClO4)
parameter
4aa
5a
pKa
kH
k0
ksp
1.58 ( 0.07
2.85b
b
(7.0 ( 2.6) × 10-4 M-1 s-1 (1.18 × 10-2 M-1 s-1
)
+
(6.7 ( 0.8) × 10-4 s-1
(2.27 × 10-4 s-1
)
b
(2.72 ( 0.17) × 10-7 s-1
(3.08 ( 0.06) × 10-5 s-1
(1.74 × 10-5 s-1
)
b
k
OH-KW (9.7 ( 0.8) × 10-19 M s-1
(2.38 ( 0.09) × 10-18 M s-1
a This work unless stated otherwise. b Values taken from ref 29.
(iii) the spontaneous reaction of the anion; and (iv) a base-
promoted reaction, respectively.
Figure 1. Plot of log(kobs) versus -log[H3O+] for the hydrolyses of
4-nitrophenyl R-D-N-acetylneuraminide (O) and 4-nitrophenyl â-D-N-
acetylneuraminide (b), T ) 50 °C. The included solid lines are the best
nonlinear fits of the kinetic data to eq 1 or eq 2. The dashed line is generated
from kinetic rate constants reported for the hydrolysis of 4-nitrophenyl R-D-
N-acetylneuraminides, T ) 50 °C (ref 29).
log(kobs) ) log{kH+[H+]/(1 + K /[H+]) +
a
k0/(1 + Ka/[H+]) + ksp/(1 + [H+]/Ka) +
kOH-K /[H+](1 + [H+]/K )} (1)
W
a
Also, shown in Figure 1 are recently measured rate constants
for pH values between 7.3 and 14.0 and the best-fit line for the
previously reported kinetic data (pH values 0.8-10.0) for
hydrolysis of 4-nitrophenyl R-D-N-acetylneuraminide (5). The
newly measured rate constants for 5 (Table S2, Supporting
Information) were fit to eq 2, where ksp and kOH-Kw represent
the same rate constants detailed above.
Scheme 2
log(kobs) ) log{ksp + kOH-K /[H+]}
(2)
W
Listed in Table 1 are the so-derived values for the kinetic
parameters.
Figure 2 displays the Brønsted plot, log(kobs) versus the pKa
of the conjugate acid of the leaving group for the spontaneous
reactions of 4a-d measured in the pH-independent domain at
100 °C (Table S3, Supporting Information). The derived âlg
value for this reaction is -1.24 ( 0.16. Similarly, a Brønsted
plot of log(kobs) for the hydrolyses of 4a-g measured at a pH
value of 1.0 and a temperature of 50 °C gave a slope of
0.14 ( 0.08 (Figure 2; Table S4, Supporting Information).
The pseudo-first-order rate constants for the spontaneous
reactions of the anions of 4a and 5 were also measured as a
function of temperature (Tables S5 and S6, Supporting Informa-
tion), and the corresponding Eyring plot is shown in Figure 3.
The activation parameters derived for the spontaneous hydrolysis
of 4a and 5 are ∆Hq ) 138 ( 3 kJ mol-1 and ∆Sq ) 59 (
8 J mol-1 K-1, and ∆Hq ) 116 ( 2 kJ mol-1 and ∆Sq ) 27 (
4 J mol-1 K-1, respectively.
To probe for intramolecular nucleophilic attack on the ipso-
carbon of the aromatic ring either by the anomeric carboxylate
or directly by hydroxide ion, oxygen-18 exchange experiments
were performed. The 4-nitrophenol isolated from the base-
promoted reactions of 4a and 5 in 18O-water (≈50 atom %)
contained approximately 2.1 atom % 18O after time periods of
20 and 6 h at 50 °C, respectively. While the 4-nitrophenol
isolated from the spontaneous ethanolysis and base-promoted
reactions of 4a-18O1 and 5-18O1, it showed no incorporation of
18O into the phenol.
were then dried (Na2SO4), and following removal of the solvent under
reduced pressure, the resulting solid residues were analyzed by standard
EI mass spectrometry. The atom % 18O in the isolated 4-nitrophenol
was calculated by the least-squares method of Brauman30 using the
isotopic peak distribution of the molecular ion observed in the mass
spectrum.
Similarly, any 18O-incorporation into the 4-nitrophenol product from
the anomeric carboxylate group for the reactions of 4a-18O and 5-18O
in water at a pH value of 14.0 ([OH-] ) 1.0 M) and in aqueous ethanol
(0, 50, and 100% v/v containing 3 equiv of N-methylmorpholine) was
monitored using an analogous protocol to that reported above.
Results
Aryl â-D-N-acetylneuraminides were prepared from the
readily made and fully protected â-D-N-acetylneuraminyl fluo-
ride (6)26 according to the synthetic route outlined in Scheme
2. The coupling reactions that give the protected aryl â-sialosides
(7a-g) also yielded an appreciable quantity of an elimination
product (the R,â-unsaturated ester).
pH Rate Profile. Figure 1 presents the logarithms of the
measured rate constants (kobs) as a function of pH for the
hydrolyses of 4-nitrophenyl â-D-N-acetylneuraminide (4a) at
50 °C. Individual rate constants are given in Table S1 of
Supporting Information. The observed rate constants for 4a were
fit to eq 1, where Ka is the ionization constant of 4a, while
+
-
kH , k0, ksp, and kOH Kw are the rate constants for (i) the
acid-catalyzed reaction of the neutral molecule; (ii) the acid-
catalyzed reaction of the carboxylate form, or its kinetic
equivalent, the spontaneous reaction of the neutral molecule;
To check for the possibility of a base-promoted E2 elimination
mechanism, the â-secondary deuterium kinetic isotope effect
(â-SDKIE) for the reactions of 4a and its axial 3-deuterio-
labeled isotopomer (4a-2H) were evaluated in NaOH (1 M) at
(30) Brauman, J. I. Anal. Chem. 1967, 38, 607-610.
9
J. AM. CHEM. SOC. VOL. 127, NO. 20, 2005 7461

A R T I C L E S
Dookhun and Bennet
log(kobs/k0) ) mYX
(3)
Thus, when the kinetic data for the aqueous ethanolyses of
4a and 5, which are given in Tables S7 and S8 (Supporting
Information), are fit to eq 3 using Ypic values,32-34 the derived
m values for the reactions of 4a and 5 are -0.04 ( 0.01 and
+0.23 ( 0.02, respectively. Figure 4 displays these Grunwald-
Winstein plots for the solvolyses of 4a and 5. To minimize
differential solvation effects on the leaving group,31 the Ypic scale
was chosen because the picrate leaving group in the reference
reaction, the solvolysis of 1-adamantyl picrate (25 °C),33 is
structurally similar to the nucleofuge for the spontaneous
solvolysis reactions under study (i.e., 4-nitrophenolate).
To ensure that the ethyl N-acetylneuraminide products did
not undergo any subsequent acid-catalyzed reactions, all solvo-
lytic product studies were performed in the presence of 3 molar
equiv of N-methylmorpholine. Given in Table 2 are the
measured percentages of the products, which were identified
by standard 1H NMR spectroscopic analysis, that were formed
during the solvolysis of 4a and 5.
Discussion
First, a comment is warranted on the difference between the
rate constant for spontaneous hydrolysis of 5 reported by
Ashwell et al.29 and that obtained in the current study (ksp; Table
1). In the current study, no temperature corrections were applied
to the buffer pH values (measured at 25 °C), yet all of the
computed rate constants for hydrolysis of 5 (up to [NaOH] )
0.01 M) were identical within experimental error (Table S2,
Supporting Information). Also, the rate constants measured in
TAPS buffer (0.3 M, pH ∼ 7.9) with and without ionic strength
adjustment are identical within experimental error (Table S2).
Therefore, neither solution acidity nor ionic strength differences
can account for the variation in spontaneous rate constant noted
above. The most likely explanation is that a temperature
difference existed between the two data sets. Indeed, given the
activation parameters reported above, an imbalance of 4 °C,
between the two studies, is sufficient to rationalize a rate
difference of 1.7-fold (Table 1). It is, nevertheless, logical to
compare the newly acquired rate data for hydrolysis of 5 with
that evaluated for 4a using the same instrumental setup.
Spontaneous Hydrolytic Reactions. Remarkably, the spon-
taneous hydrolysis of 5, the thermodynamically less stable
diastereomer, occurs at a rate that is over 100 times faster than
that of the corresponding reaction of 4a. The enhanced reactivity
of the R-anomer is not caused by intramolecular nucleophilic
catalysis by the C-1 carboxylate group occurring during its
spontaneous hydrolysis for the following reasons: (i) the
respective âlg values on ksp for the R- and â-anomers of -1.32
(60 °C)29 and -1.24 ( 0.16 (100 °C) are indistinguishable, thus
Figure 2. Dependence of the observed rate constants (kobs) for hydrolysis
of aryl â-D-N-acetylneuraminides on the pKa of the parent phenol: pH 1.0
at 50 °C (O) and pH 8.08 at 100 °C (b). The lines shown are the best
linear least-squares fits through the data points.
Figure 3. Eyring plot for the spontaneous hydrolyses of the anions of
4-nitrophenyl R-D-N-acetylneuraminide (O) and 4-nitrophenyl â-D-N-
acetylneuraminide (b). The lines shown are the best linear least-squares
fits through the data points.
50 °C. Under these conditions, the measured â-SDKIE (kH/kD)
was 1.096 ( 0.025.
To probe the effect of solvent polarity on the spontaneous
rate constant (ksp), 4a and 5 were solvolyzed in various aqueous
ethanol mixtures. The standard Grunwald-Winstein eq 3 is
commonly used to analyze solvent effects on DN + AN (SN1)
substitution reactions, where the YX parameter is a measure of
the solvent’s ionizing power (leaving group ) X) and m is a
sensitivity variable.31
(32) Bentley, T. W.; Carter, G. E. J. Am. Chem. Soc. 1982, 104, 5741-5747.
(33) Bentley, T. W.; Roberts, K. J. Org. Chem. 1985, 50, 4821-4828.
(34) The Ypic values for water, 10% EtOH, and 60% EtOH were estimated from
the linear correlation (r ) 0.9981) between Ypic values (ref 33) and YCl (ref
32).
(35) Horenstein, B. A. J. Am. Chem. Soc. 1997, 119, 1101-1107.
(36) Chou, D. T. H.; Watson, J. N.; Scholte, A. A.; Borgford, T. J.; Bennet, A.
J. J. Am. Chem. Soc. 2000, 122, 8357-8364.
(37) Chandrasekhar, S.; Kirby, A. J. J. Chem. Soc., Chem. Commun. 1978, 171-
172.
(38) Chandrasekhar, S.; Kirby, A. J.; Martin, R. J. J. Chem. Soc., Perkin Trans.
2 1983, 1619-1626.
(31) Bentley, T. W.; Llewellyn, G. Prog. Phys. Org. Chem. 1990, 17, 121-
(39) Wolfenden, R.; Lu, X.; Young, G. J. Am. Chem. Soc. 1998, 120, 6814-
6815.
158.
9
7462 J. AM. CHEM. SOC. VOL. 127, NO. 20, 2005

Stability of Aryl
â-N-Acetylneuraminides in Neutral Solution
A R T I C L E S
Scheme 4
carboxylate groups in both 4a and 5 were labeled, it is also
possible to rule out a mechanism involving an intramolecular
nucleophilic attack on the ipso-carbon that would result in
formation of a transient 4-nitrophenyl ester (Scheme 4).
The likely cause for the large difference in the rate of
spontaneous hydrolysis of 4a and 5 is steric relief of ground
state strain caused by the anomeric carboxylate group, an effect
that is absent in the hydrolysis reactions of comparable acetals.
Specifically, the rate ratios for the spontaneous hydrolyses of
the two 4-nitrophenyl trans-decalin acetals 12 and 13 (3.6-fold
at 39 °C37,38) and the two anomers of methyl glucopyranoside
14 and 15 (2.5-fold at 25 °C39) are considerably smaller than
the value measured for the ketals 4a and 5 (∼110-fold at
50 °C). In these two cases, the axial acetal (12 or 14), which is
the thermodynamically more stable diastereomer, reacts slower
than the corresponding equatorial acetal (13 or 15). That is,
replacement of an equatorial hydrogen atom by a carboxylate
group (cf. 12 and 4a) introduces less ground-state steric strain
into the molecule than does the same substitution in an axial
position (cf. 13 and 5).
Figure 4. Plot of Grunwald-Winstein correlation for the aqueous
ethanolysis of 4a (b) and 5 (O) versus Ypic at T ) 75 °C. The lines shown
are the best linear least-squares fits through the data points.
Table 2. Observed Product Percentages Formed during the
Solvolysis of 4a and 5 in Aqueous Ethanol (v/v) Solvent Mixtures
at 75 °Ca,b
NeuAc
Et-
RNeuAc
Et-
âNeuAc
glycal
substrate
EtOH%
(8)
(9)
(10)
(11)
unknown
4ac
5
5
100
100
50
9
10
6
73
90
19
9
10
trd
7
trd
3
65
a All solvents contained 3 molar equiv of N-methylmorpholine. b Esti-
mated errors (5%. c No values given for solvolysis of 4a in 50% v/v EtOH:
H2O because extensive decomposition occurs under these conditions.
d tr ) trace.
Scheme 3
Another consequence of the greater ground state steric
compression in these compounds is that the carboxylic acid
group in â-N-acetylneuraminides is more acidic than the same
group in R-N-acetylneuraminides; this presumably results from
the greater ease of solvation of an equatorial carboxylate in
comparison to an axial carboxylate. Thus, the calculated pKa
value of the equatorial carboxylic acid in 4a is 1.58 (Table 1),
a value that is 1.27 units more acidic than the corresponding
ionization associated with the axial CO2H group in 5.29 A similar
trend in acidity constants has been reported for the cis- and trans-
isomers of 4-tert-butyl cyclohexanecarboxylic acid, where in
50% aqueous ethanol, the equatorial carboxylic acid (trans) is
more acidic than the axial isomer by 0.46 pKa units.40
suggesting similarly late transition states for both anomers; and
(ii) both retained and inverted substitution products are formed
during the aqueous ethanolyses of 4a and 5 (Table 2), an
observation that is inconsistent with intramolecular nucleophilic
participation because such reactions should yield retained ethyl
N-acetylneuraminide as the only glycosidic product (Scheme
3). Therefore, at the present time, no evidence exists for the
occurrence of intramolecular assistance during the hydrolysis
reactions of N-acetylneuraminides, a conclusion that other
workers have also made.35,36 Furthermore, given that no isotopic
enrichment was observed in the 4-nitrophenol product when the
(40) Hoefnagel, A. J.; Wepster, B. M. Recl. TraV. Chim. Pays-Bas 1990, 109,
455-462.
9
J. AM. CHEM. SOC. VOL. 127, NO. 20, 2005 7463

A R T I C L E S
Scheme 5
Dookhun and Bennet
Scheme 6
recently reported value of +0.66 (YOTs scale) for the spontaneous
solvolysis of 2-(4-nitrophenoxy)tetrahydropyran.45
Base-Promoted Reaction. The results from the 18O-
incorporation experiments (in 18O-H2O) rule out the possibility
that the base-promoted reactions of 4a and 5 occur by nucleo-
philic attack of hydroxide ion on the ipso-carbon of the
4-nitrophenyl ring (Scheme 5).
Indeed, the small amount of 18O-exchange measured in the
current system is consistent with hydroxide attacking the
liberated 4-nitrophenoxide.46 In addition, the possibility that the
base-promoted reaction involves an E2 elimination mechanism
is precluded, at least in the case of 4a, because the observed
deuterium KIE (kH/kD ) 1.097) for this reaction is too small to
be associated with a primary KIE for a standard a trans-diaxial
elimination reaction (Scheme 6).
It appears that the base-promoted reactions of 4a and 5 do
not involve initial reaction at either the anomeric carbon or the
ipso-carbon, but rather probably involve reaction at a remote
center, as proposed by Sinnott and co-workers.29
Low pH Regime. In the low pH domain, the rate constants
for hydrolysis of 4a display an inflection point, which presum-
ably is associated with protonation of its carboxylate group. The
calculated pKa value of this equatorial carboxylic acid is 1.58
(Table 1), a value that is 1.27 units more acidic than the
corresponding ionization associated with the axial CO2H group
in 5.29 A similar trend in acidity constants has been reported
for the cis- and trans-isomers of 4-tert-butyl cyclohexane-
carboxylic acid, where in 50% aqueous ethanol, the equatorial
carboxylic acid (trans) is more acidic than the axial isomer by
0.46 pKa units.40
The second-order rate constant, kH+, which refers to the acid-
catalyzed hydrolysis of the neutral N-acetylneuraminide, is
approximately 17-fold larger for the reactions of 5 than for the
corresponding reactions of 4a (Table 1). Of note, this ratio of
second-order rate constants is larger than that normally associ-
ated with the specific acid-catalyzed hydrolysis of glycosides,
an observation that is consistent with relief of additional steric
crowding of the axial carboxylic acid group in the R-anomer
(vide supra).
Furthermore, the activation parameters for the spontaneous
hydrolysis of 4a and 5 are ∆Hq ) 138 ( 3 kJ mol-1 and
∆Sq ) 59 ( 8 J mol-1 K-1, and ∆Hq ) 116 ( 2 kJ mol-1 and
∆Sq ) 27 ( 4 J mol-1 K-1, respectively. The large positive
activation entropies are consistent with these spontaneous
hydrolysis reactions proceeding via dissociative transition states.
Moreover, the more positive activation entropy for the â-anomer
(4a) is consistent with the less sterically congested equatorial
carboxylate being solvated to a greater extent in the ground state
than is the case for the R-anomer (5), and that solvation of these
two anionic groups becomes similar at their respective hydrolytic
transition states.
Sialosyl Oxacarbenium Ion Lifetime. It is likely that the
sialosyl oxacarbenium ion (3) possesses a lifetime close to that
required for a cationic intermediate to become solvent-
equilibrated for the following reasons: (1) both anomers
solvolyze in 100% ethanol to give very similar products (Table
2), thus suggesting that these reactions proceed via similar
oxacarbenium ion-like intermediates; and (2) more inverted
product is obtained during the aqueous methanolysis of CMP
â-D-N-acetylneuraminide (1),24,25 a compound in which the
leaving group is more basic than 4-nitrophenoxide. It is well-
known that reactions of glycosides in which the leaving group
is a nondelocalized anion, fluoride,41,42 are more prone to react
via an ANDN (SN2) mechanism than are the corresponding
spontaneous reactions when the leaving group is a resonance-
stabilized anion, such as 2,4-nitrophenoxide.43
Given that intramolecular nucleophilic participation by the
anomeric carboxylate group has been ruled out, the low m values
obtained for the solvolyses of 4a and 5 (-0.04 and +0.23,
respectively) are consistent with the occurrence of charge
delocalization at their individual zwitterionic transition states.
Indeed, the interaction between positive and negative charges
at the zwitterionic transition states is the probable origin for
the observed low m values,44 which are much lower than the
The kinetic process (k0, Table 1) that is associated with either
the acid-catalyzed reaction of the carboxylate form or the kinetic
equivalent the spontaneous reaction of the neutral molecule is
calculated to be faster for the â-anomer. However, in the pH
rate profiles for both 4a and 5, it is evident that for both
compounds, only small deviations from linearity are discernible,
and this observation makes a thorough mechanistic analysis of
this difference impracticable.
Biological Implications. Clearly, the natural leaving group
for sialyl transferases, cytidine monophosphate (CMP), is a good
(41) Banait, N. S.; Jencks, W. P. J. Am. Chem. Soc. 1991, 113, 7951-7958.
(42) Zhang, Y.; Bommuswamy, J.; Sinnott, M. L. J. Am. Chem. Soc. 1994, 116,
7557-7563.
(43) Namchuk, M. N.; McCarter, J. D.; Becalski, A.; Andrews, T.; Withers, S.
G. J. Am. Chem. Soc. 2000, 122, 1270-1277.
(45) Ahmad, I. A.; Birkby, S. L.; Bullen, C. A.; Groves, P. D.; Lankau, T.;
Lee, W. H.; Maskill, H.; Miatt, P. C.; Menneer, I. D.; Shaw, K. J. Phys.
Org. Chem. 2004, 17, 560-566.
(44) The m values obtained in fits against YOTs for the solvolyses of 4a and 5
are -0.02 and +0.11, respectively.
(46) Hengge, A. C. J. Am. Chem. Soc. 1992, 114, 2747-2748.
9
7464 J. AM. CHEM. SOC. VOL. 127, NO. 20, 2005

Stability of Aryl
â-N-Acetylneuraminides in Neutral Solution
A R T I C L E S
leaving group. Given the above analysis, one would expect CMP
R-D-N-acetylneuraminide to react about 100-fold faster than the
natural substrate CMP â-D-N-acetylneuraminide (1), a com-
pound for which the pseudo-first-order rate constant for hy-
drolysis (khyd) at pH values of 5.0, 6.0, and 7.0 is calculated to
be 1.51 × 10-4, 2.48 × 10-5, and 1.15 × 10-5 s-1, respec-
tively.24 Therefore, assuming a factor of 100 for the increased
reactivity of the R-anomer relative to that of the natural
â-anomer, the expected half-time (t1/2) for the spontaneous
hydrolysis of CMP R-D-N-acetylneuraminide at pH values of
5.0, 6.0, and 7.0 are around 0.75, 4.5, and 10 min, respectively.
This time interval is likely too short to allow for efficient
transport of CMP R-D-N-acetylneuraminide from where it would
be synthesized, if it were the donor sugar instead of its â-anomer,
(nucleus) to the site where sialyl transfer occurs (golgi).47
acetylneuraminide given above, it can be calculated that CMP
â-D-N-acetylneuraminide would be expected to hydrolyze about
8-10-fold faster than 4a. The discrepancy between the experi-
mental and the calculated values likely results from an elec-
trostatic destabilization of the CMP â-D-N-acetylneuraminide
ground state that is induced by the close proximity of two
negative charges.
Acknowledgment. The authors are grateful to the Natural
Sciences and Engineering Research Council of Canada for
financial support. In addition, we thank the three anonymous
referees for their thorough critical reviews of this manuscript
and for their many helpful comments.
Supporting Information Available: Experimental details for
the synthesis of aryl â-D-N-acetylneuraminides: 3-nitrophenyl
(4b), 4-chlorophenyl (4c), 3-chlorophenyl (4d), phenyl (4e),
4-methylphenyl (4f), and 3-methylphenyl (4g). Tables of
observed rate constants for the solvolysis reactions of 4-nitro-
phenyl R- and â-D-N-acetylneuraminide (5 and 4a). Also
included are tables of observed rate constants for the hydrolysis
reactions of the seven aryl â-D-N-acetylneuraminides at pH
values of 1.00 and 8.08. This material is available free of charge
via the Internet at http://pubs.acs.org.
In addition, the rate constant for the spontaneous hydrolysis
of CMP â-D-N-acetylneuraminide (1) at 37 °C is 1 × 10-5 s-1 24
,
a value that is approximately 270 times larger than that
extrapolated for 4-nitrophenyl â-D-N-acetylneuraminide (4a) of
3.7 × 10-8 s-1. On the basis of the pKa for CMP of 6.4348 and
the âlg value for the spontaneous hydrolyses of aryl â-D-N-
(47) Murrell, M. F.; Yarema, K. J.; Levchenko, A. ChemBioChem 2004, 5,
1334-1347.
(48) Chen, H. M.; Parkinson, J. A.; Morris, R. E.; Sadler, P. J. J. Am. Chem.
Soc. 2003, 125, 173-186.
JA042280E
9
J. AM. CHEM. SOC. VOL. 127, NO. 20, 2005 7465