Effect of neutral pyridine leaving groups on the mechanisms of influenza type A viral sialidase-catalyzed and spontaneous hydrolysis reactions of α-D-N-acetylneuraminides

Article abstract of DOI:10.1021/ja001641x

A reagent panel, comprised of five pyridinium salts of α-D-N-acetylneuraminic acid, was synthesized and then used to probe enzymatic (α-sialidase) and nonenzymatic mechanisms of neuraminide hydrolysis. Spontaneous hydrolysis of the pyridinium salts proceeded via two independent pathways, where unassisted C-N bond cleavage was the rate-determining step. Cationic species (i.e., anomeric carboxylate protonated) displayed apparent pK(a) values in the range of 0.4-0.7. However, spontaneous hydrolyses of the cationic and zwitterionic species had similar β(1g) values of -1.22 ± 0.16 and -1.22 ± 0.07, respectively. The results, plus the activation parameters calculated from the hydrolysis of pyridinium α-D-N-acetylneuraminide (ΔH(+) = 112 ± 2 kJ mol^-1 and ΔS(+) = 28 ± 4 J mol^-1 K^-1), strongly suggest that the anomeric carboxylate does not assist in the departure of neutral pyridine leaving groups. Enzymatic hydrolysis was studied using an influenza viral α-sialidase (A/Tokyo/3/67) which was recombinantly expressed using a baculovirus/insect cell expression system. Sialidase protein was purified by a combination of density gradient centrifugation and gel filtration chromatography. Kinetic parameters for the enzymatic hydrolysis of the pyridinium salts were measured at 37 °C and at pH values of 6.0 and 9.5. The β(1g) values derived for k(cat)/K(m) and k(cat) were essentially zero, indicating that chemical transformations/events are not rate-determining. Rather, this observation is consistent with a model for α-sialidase-catalyzed hydrolyses (Guo, X.; Laver, W. G.; Vimr, E.; Sinnott, M. L., J. Am. Chem. Soc. 1994, 116, 5572) in which k(cat)/K(m) is determined by a conformational change of the first-formed Michaelis complex and k(cat) is determined by the virtual transition state made up of two separate conformational events.

Full text of DOI:10.1021/ja001641x

J. Am. Chem. Soc. 2000, 122, 8357-8364
8357
Effect of Neutral Pyridine Leaving Groups on the Mechanisms of
Influenza Type A Viral Sialidase-Catalyzed and Spontaneous
Hydrolysis Reactions of R-D-N-Acetylneuraminides
Doug T. H. Chou,† Jacqueline N. Watson,‡ Andrew A. Scholte,† Thor J. Borgford,*,‡ and
Andrew J. Bennet*,†
Contribution from the Department of Chemistry and the Department of Molecular Biology and
Biochemistry, Simon Fraser UniVersity, 8888 UniVersity DriVe, Burnaby,
British Columbia V5A 1S6, Canada
ReceiVed May 12, 2000
Abstract: A reagent panel, comprised of five pyridinium salts of R-D-N-acetylneuraminic acid, was synthesized
and then used to probe enzymatic (R-sialidase) and nonenzymatic mechanisms of neuraminide hydrolysis.
Spontaneous hydrolysis of the pyridinium salts proceeded via two independent pathways, where unassisted
C-N bond cleavage was the rate-determining step. Cationic species (i.e., anomeric carboxylate protonated)
displayed apparent pKa values in the range of 0.4-0.7. However, spontaneous hydrolyses of the cationic and
zwitterionic species had similar âlg values of -1.22 ( 0.16 and -1.22 ( 0.07, respectively. The results, plus
the activation parameters calculated from the hydrolysis of pyridinium R-D-N-acetylneuraminide (∆Hq ) 112
( 2 kJ mol-1 and ∆Sq ) 28 ( 4 J mol-1 K-1), strongly suggest that the anomeric carboxylate does not assist
in the departure of neutral pyridine leaving groups. Enzymatic hydrolysis was studied using an influenza viral
R-sialidase (A/Tokyo/3/67) which was recombinantly expressed using a baculovirus/insect cell expression
system. Sialidase protein was purified by a combination of density gradient centrifugation and gel filtration
chromatography. Kinetic parameters for the enzymatic hydrolysis of the pyridinium salts were measured at 37
°C and at pH values of 6.0 and 9.5. The âlg values derived for kcat/Km and kcat were essentially zero, indicating
that chemical transformations/events are not rate-determining. Rather, this observation is consistent with a
model for R-sialidase-catalyzed hydrolyses (Guo, X.; Laver, W. G.; Vimr, E.; Sinnott, M. L. J. Am. Chem.
Soc. 1994, 116, 5572) in which kcat/Km is determined by a conformational change of the first-formed Michaelis
complex and kcat is determined by the virtual transition state made up of two separate conformational events.
Introduction
neutral leaving groups react via short-lived nonsolvent equili-
brated oxacarbenium ions (DN*AN).5,6 In the latter case,
nucleophilic attack occurs prior to the complete dissociation of
the intermediate ion-molecule pair.6 Such reactivities are a
consequence of the short lifetime of the glucosyl oxacarbenium
ion, estimated to be on the order of (1-3) × 10-12 s.6,7
In contrast to the body of work on aldopyranoside hydrolysis
reactions, there are comparatively few mechanistic studies
involving N-acetylneuraminides. Mechanistic studies have been
reported for hydrolysis of aryl R-D-N-acetylneuraminides (1)8
and CMP-â-D-N-acetylneuraminide (2).9,10 The presence of an
ionizable carboxylic acid group in the aryl R-N-acetylneur-
aminides increases the number of hydrolytic pathways relative
This report addresses several critical questions concerning
the spontaneous and enzyme-catalyzed hydrolysis reactions of
N-acetylneuraminides. For example, does the C-1 carboxylate
group of the N-acetylneuraminides have a catalytic function?
Is chemistry or conformational change rate-determining in the
enzyme-catalyzed hydrolysis of the N-acetylneuraminides?
Carbohydrates have roles in a variety of biologically important
processes, including receptor/ligand recognition, intracellular
trafficking, and immunity.1 One of the most biologically
important keto-sugars is the nine-carbon N-acetylneuraminic acid
(also known as sialic acid).2 N-Acetylneuraminides are structur-
ally distinct from glucopyranosides in that they contain tertiary
reaction centers (ketose vs aldose) and have an ionizable
carboxylic acid group attached to the anomeric center.
(3) Banait, N. S.; Jencks, W. P. J. Am. Chem. Soc. 1991, 113, 7951-
7958.
(4) Zhang, Y.; Bommuswamy, J.; Sinnott, M. L. J. Am. Chem. Soc. 1994,
116, 7557-7563.
Nucleophilic substitution reactions on aldopyranosides con-
taining anionic leaving groups, such as fluoride, react with
anionic nucleophiles3 and water4 via an “exploded” ANDN (SN2)
transition state (TS). Alternatively, aldopyranosides that contain
(5) Bennet, A. J.; Sinnott, M. L. J. Am. Chem. Soc. 1986, 108, 7287-
7294.
(6) (a) Huang, X.; Surry, C.; Hiebert, T.; Bennet, A. J. J. Am. Chem.
Soc. 1995, 117, 10614-10621. (b) Zhu, J.; Bennet, A. J. J. Am. Chem.
Soc. 1998, 120, 3887-3893.
† Department of Chemistry.
(7) Amyes, T. L.; Jencks, W. P. J. Am. Chem. Soc. 1989, 111, 7888-
‡ Department of Molecular Biology and Biochemistry.
(1) (a) Rademacher, T. W.; Parekh, R. B.; Dwek, R. A. Annu. ReV.
Biochem. 1988, 57, 785-838. (b) Drickamer, K. Cell 1991, 67, 1029-
1032.
7900.
(8) Ashwell, M.; Guo, X.; Sinnott, M. L. J. Am. Chem. Soc. 1992, 114,
10158-10166.
(9) Horenstein, B. A.; Bruner, M. J. Am. Chem. Soc. 1996, 118, 10371-
(2) Schauer, R.; Kelm, S.; Reuter, G.; Roggentin, P.; Shaw, L. In Biology
of the Sialic Acids; Rosenburg, A., Ed.; Plenum Press: New York, 1995;
Chapter 2.
10379.
(10) Horenstein, B. A.; Bruner, M. J. Am. Chem. Soc. 1998, 120, 1357-
1362.
10.1021/ja001641x CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/18/2000

8358 J. Am. Chem. Soc., Vol. 122, No. 35, 2000
Chou et al.
Scheme 1
Scheme 2
to aryl aldopyranosides. Potential reaction pathways are shown
in Scheme 1 as follows: (a) spontaneous reaction of the car-
boxylate (k1), (b) acid catalysis of the carboxylate (k1H ) or the
+
kinetically equivalent spontaneous reaction of the carboxylic
+
acid (k2), (c) acid catalysis of the carboxylic acid (k2H ), and
-
(d) base-catalyzed reaction of the carboxylate (kOH ). Notably,
the existence of a second ionic functionality in CMP â-D-N-
acetylneuraminide (2) further increases the number of potential
hydrolytic pathways.
Conflicting evidence has been reported concerning the role
of the carboxylate group during nonenzymatic hydrolysis reac-
tions. Ashwell et al. proposed that the C-1 carboxylate group
assists nucleophilically in both the general-acid-catalyzed depar-
ture of a phenol and the spontaneous departure of a phenolate
anion during the hydrolysis reactions of aryl R-D-N-acetylneur-
aminides (1).8 In contrast, Horenstein and co-workers propose
that the carboxylate group is coplanar to the oxacarbenium ion
at the transition state for hydrolysis of CMP-â-D-N-acetyl-
neuraminide (i.e., nucleophilic participation cannot occur).9,11
Analysis of enzyme-catalyzed reactions is naturally more
complex than that of the corresponding solution-phase reactions.
For example, there are two acidic residues (Asp/Glu) present
in the active site of the majority of glycosidases; these groups
participate during the hydrolysis of carbohydrates. In 1953,
Koshland12 proposed a mechanism for retaining glycosidases
which is still generally accepted.13 One of the two acidic amino
acid residues acts as a general-acid catalyst, while the conjugate
base of the second residue reacts as a nucleophile.
Sialidases (N-acetylneuraminyl glycohydrolase, EC 3.2.1.18)
are retaining glycosidases that catalyze the hydrolysis of
N-acetylneuraminic acid residues R-linked to glycoproteins,
glycolipids, and polysaccharides.14-16 Different hydrolytic
mechanisms have been proposed for the sialidase family of
enzymes, also involving two acidic active-site residues.17,18 One
mechanism, based largely on X-ray structure determinations on
an influenza type A enzyme,19,20 is proposed in which the R-N-
acetylneuraminide is bound in a boat conformation (ESB). Strong
electrostatic binding occurs between the anionic carboxylate of
the N-acetylneuraminide and an arginine residue on the enzyme.
According to this mechanism, an aspartic acid residue then
functions as a general-acid catalyst to assist departure of the
aglycon (Scheme 2). Consistent with this role of the aspartic
acid, evidence for proton transfer exists from a kinetic study
on the influenza virus sialidase21 and from a computational
study.22
An alternative mechanism for enzyme-catalyzed hydrolysis
has been proposed by von Itzstein and co-workers.18 According
to this proposal, water directly attacks an oxacarbenium ion via
an internal return mechanism.
Notably, the catalytic roles of essential glutamate and tyrosine
residues are not fully described,23 though it appears that one or
both residues stabilizes the transition state either electrostatically
or nucleophilically. Thomas et al. concluded that an oxacarbe-
nium ion is a potential intermediate in sialidase-catalyzed
hydrolyses; their conclusion was based on a modeling study of
the cation bound to the enzyme active site (B/Beijing/1/87).24
We have also initiated studies to unravel the complex nature
of these hydrolyses. The current investigation simplifies the
hydrolysis kinetics through the use of neutral leaving groups
that obviate the need for acid catalysis. Five pyridinium R-D-
N-acetylneuraminides were synthesized and their hydrolyses
characterized in both enzymatic (influenza type A/Tokyo/3/67)
and nonenzymatic systems.
(11) Horenstein, B. A. J. Am. Chem. Soc. 1997, 119, 1101-1107.
(12) Koshland, D. E. Biol. ReV. 1953, 28, 416-436.
(13) (a) Sinnott, M. L. Chem. ReV. 1990, 90, 1171-1202. (b) Legler, G.
AdV. Carbohydr. Chem. Biochem. 1990, 48, 319-384. (c) McCarter, J. D.;
Withers, S. G. Curr. Opin. Struct. Biol. 1994, 4, 885-892. (d) Davies, G.;
Withers, S. G.; Sinnott, M. L. Glycosyl Transfer. In ComprehensiVe
Biological Catalysis: A Mechanistic Reference; Sinnott, M. L., Ed.;
Academic Press: San Diego, 1997; Vol. 1, Chapter 3, pp 119-208.
(14) Saito, M.; Yu, R. K. In Biology of the Sialic Acids; Rosenberg, A.,
Ed.; Plenum Press: New York, 1995; Chapter 8.
Materials and Methods
The buffers used in this study, 2-(N-morpholino)ethanesulfonic acid
(MES), 3-(N-tris(hydroxymethyl)methyl)amino)propanesulfonic acid
(15) Wilson, J. C.; Angus, D. I.; von Itzstein, M. J. Am. Chem. Soc.
1995, 117, 4214-4217.
(16) (a) Henrissat, B. Biochem. J. 1991, 280, 309-316. (b) Henrissat,
B.; Bairoch, A. Biochem. J. 1993, 293, 781-788. (c) Henrissat, B.; Bairoch,
A. Biochem. J. 1996, 316, 695-696.
(21) Guo, X.; Laver, W. G.; Vimr, E.; Sinnott, M. L. J. Am. Chem. Soc.
1994, 116, 5572-5578.
(22) Barnes, J. A.; Williams, I. H. Biochem. Soc. Trans. 1996, 24, 263-
268.
(23) Lentz, M. R.; Webster, R. G.; Air, G. M. Biochemistry 1987, 26,
5351-5358.
(24) Thomas, A.; Jourand, D.; Bret, C.; Amara, P.; Field, M. J. J. Am.
Chem. Soc. 1999, 121, 9693-9702.
(17) Chong, A. K. J.; Pegg, M. S.; Taylor, N. R.; von Itzstein, M. Eur.
J. Biochem. 1992, 207, 335-343.
(18) Taylor, N. R.; von Itzstein, M. J. Med. Chem. 1994, 37, 616-624.
(19) Varghese, J. N.; Colman, P. M. J. Mol. Biol. 1991, 221, 473-486.
(20) Varghese, J. N.; McKimm-Breschkin, J.; Caldwell, J. B.; Kortt, A.
A.; Colman, P. M. Proteins: Struct., Funct. Genet. 1992, 14, 327-332.

Hydrolysis Reactions of N-Acetylneuraminides
J. Am. Chem. Soc., Vol. 122, No. 35, 2000 8359
(TAPS), and 2-(N-cyclohexyamino)ethanesulfonic acid (CHES) were
purchased from Sigma and used without further purification. Pyridine,
3-methoxypyridine, 3-methylpyridine, 4-methylpyridine, and 3,4-di-
methylpyridine were purchased from Aldrich and used without further
purification. N-Acetylneuraminic acid was purchased from Rose
Scientific and used without further purification. Hydrochloric acid
solutions were made by dilutions of a standard 1.00 M solution (Fischer
Scientific). All other salts used in the hydrolysis studies were of
analytical grade and were used without further purification. Milli-Q
water (18.2 MΩ cm-1) was used for kinetic experiments. NMR spectra
were acquired on a Bruker AMX-400 spectrometer. The conditions used
for HPLC purification were as follows: C-18 reverse-phase column,
25 × 10 mm, using a flow rate of 5 mL/min.
fractions were collected and lyophilized to give a white solid (26 mg,
1
30%): [R]20 ) -33.3 (c 0.36, H2O). H NMR (400 MHz, D2O): δ
D
1.98 (t, 1 H, J3a,3e ) J3a,4 ) 12 Hz, H-3a), 2.32 (s, 3 H, CH3), 2.36 (dd,
1 H, J3e,3a ) 12 Hz, J3e,4 ) 4 Hz, H-3e), 3.61-3.68 (m, 2 H, H-7,
H-9a), 3.85 (dd, 1 H, J9b,9a ) 12 Hz, J9b,8 ) 3 Hz, H-9b), 3.94 (ddd, 1
H, J8,7 ) 10 Hz, J8,9a ) 6 Hz, J8,9b ) 3 Hz, H-8), 3.97-4.11 (m, 3 H,
H-4, H-5, H-6), 8.08-8.12 (m, 2 H, Ar-H), 8.55-8.61 (m, 1 H, Ar-
H), 9.10-9.15 (m, 2 H, Ar-H). 13C NMR (100 MHz, D2O): δ 24.8
(CH3), 43.4 (C3), 53.9 (C5), 65.9 (C9), 70.6 (C7), 70.7 (C4), 73.7 (C8),
77.5 (C6), 98.2 (C2), 130.8, 142.5, 149.8 (Ar-C), 171.4 (C1), 177.8
(CdO). HRMS (EI): calcd for C16H23N2O8 [M + H+], 371.1454; found,
371.1462.
Hydrolysis Kinetics. The hydrolysis reactions of 3a-e were
followed by monitoring absorbance versus time using a Cary-3E UV-
vis spectrophotometer equipped with the Cary six-cell Peltier constant-
temperature accessory. Hydrolysis reactions were initiated by injection
of an aqueous stock solution of the pyridinium salt (25 µL, 45-60
mM) into a cuvette containing 2.0 mL of the required buffer that had
been preequilibrated for 20 min at 65 °C.
The protected â-D-N-acetylneuraminyl chloride (5),25 (1-pyridinio)-
acetate,26 and 4-nitrophenyl N-acetylneuraminide25a,27 were synthe-
sized according to published procedures. Full experimental details for
the synthesis of the parent pyridinium N-acetylneuraminide (3b) are
given below, while the experimental particulars for the synthesis and
characterization of 3a,c,d,e are given in the Supporting Information.28
N-[Methyl (5-Acetamido-4,7,8,9-tetra-O-acetyl-3,5-dideoxy-D-
glycero-r-D-galacto-non-2-ulopyranosyl)onate]pyridinium Tetraflu-
oroborate (4b). N-Acetylneuraminyl chloride 5 (300 mg, 0.59 mmol)
and silver tetrafluoroborate (0.14 g, 0.71 mmol) were added with stirring
to a cooled solution (-10 °C) of pyridine (1 mL) in anhydrous THF
(8 mL). After the solution was stirred at 5 °C for 48 h, the solvent was
evaporated under reduced pressure. Methanol (15 mL) was added to
the residue, and the resulting mixture was sonicated for 10 min. After
sonication the solution was filtered and the methanol removed under
reduced pressure to give a pale yellow syrup. The crude product was
purified by flash column chromatography (silica gel, 6:1 CH2Cl2/MeOH)
to give a white solid (0.17 g, 45%). Rf ) 0.49 (7:1 CH2Cl2/MeOH). 1H
NMR (400 MHz, CDCl3): δ 1.90 (s, 3 H, CH3), 1.93 (s, 3 H, CH3),
2.06 (s, 3 H, CH3), 2.16 (s, 3 H, CH3), 2.18 (s, 3 H, CH3), 2.45 (dd, 1
H, J3a,3e ) 13.0 Hz, J3a,4 ) 9.0 Hz, H-3a), 3.77 (dd, 1 H, J3e,3a ) 13.0
At all [H3O+] the reactions of 3a and 3c were monitored at 298 and
275 nm, respectively, while the hydrolysis of 3b was followed at 268
and 271 nm for reactions at pH values of >4.0 and <4.0, respectively.
The hydrolysis reactions of 3d were monitored at 229 and 265 nm for
[H3O+] g 0.2 M and [H3O+] < 0.2 M, respectively, and the reactions
of 3e were monitored at 235 and 272 nm for [H3O+] concentrations
g0.1 M and <0.1 M, respectively. Clean isosbestic points were
observed during the hydrolyses of both 3a and 3b at 260.0 and 289.5
nm, and 243.0 and 257.5 nm, respectively. Hydrolysis rate constants
for reactions of 3a and 3b at all pH values, and those of 3c-e at pH’s
greater than 4.0, were calculated by using a standard nonlinear least-
squares fit of the absorbance versus time data (to 4 halftimes for
hydrolysis). Hydrolyses of 3c-e in acidic solutions ([H3O+] g 0.01
M) were monitored using an initial rate methodology. Specifically, the
change in absorbance was monitored until about 2-3% of the starting
material had reacted. The reaction endpoint absorbance was measured,
under identical reaction conditions, using an authentic mixture of
N-acetylneuraminic acid and the corresponding pyridine. Rate constants
were then calculated using a standard first-order rate expression.
Product Studies. The UV-vis spectrum (220-350 nm) of a
completely hydrolyzed (>10 half-lives) sample of 3a-e (5.0 × 10-5
M) in TAPS buffer (µ ) 1.00, KCl) at 65 °C was identical within
0.010-0.025 au with the spectrum of a solution containing N-
acetylneuraminic acid and the corresponding freshly distilled pyridine
(5.0 × 10-5 M).
Hz, J3e,4 ) 4.6 Hz, H-3e), 3.85 (s, 3 H, OCH3), 4.06 (dd, 1 H, J9a,9b
)
12.0 Hz, J9a,8 ) 6.5 Hz, H-9a), 4.17 (q, 1 H, J5,4 ) J5,6 ) J5,NH ) 10.0
Hz, H-5), 4.32 (dd, 1 H, J9b,9a ) 12.0 Hz, J9b,8 ) 3.0 Hz, H-9b), 4.62
(dd, 1 H, J6,7 ) 2.0 Hz, J6,5 ) 11.0 Hz, H-6), 5.4-5.48 (m, 2 H, H-4,
H-7), 5.51 (ddd, J8,9a ) 6.5 Hz, J8,9b ) 3.0 Hz, J8,7 ) 9.0 Hz, H-8),
6.44 (d, 1 H, JNH,H5 ) 10.0 Hz, NH), 8.28-8.34 (m, 2 H, Ar-H),
8.66-8.71 (m, 1 H, Ar-H), 9.30-9.34 (m, 2 H, Ar-H). 13C NMR
(100 MHz, CDCl3): δ 20.6, 20.8, 20.9, 21.0, 22.9 (CH3), 36.0 (C3),
47.8 (C5), 55.4 (OCH3), 62.5 (C9), 66.9 (C7), 67.5 (C8), 68.5 (C4),
94.2 (C2), 128.7 (Ar-C), 141.2 (Ar-C), 148.0 (Ar-C), 164.7, 169.6,
170.0, 170.6, 170.8, 170.9 (C1, CdO).
Identification of the carbohydrate-based hydrolysis product was
accomplished by 1H NMR analysis. Specifically, the N-acetyl-
neuraminide was dissolved in phosphate buffer (10 mM, µ ) 1.00 M
KCl, pH 8.0), and the solution was then heated at 65 °C for a period
of time corresponding to 10 half-lives for hydrolysis. The reaction
mixture was then evaporated to dryness, and the resultant residue was
subjected to two rounds of D2O exchange (1.5 mL) followed by
N-[(5-Acetamido-3,5-dideoxy-D-glycero-r-D-galacto-non-2-ulopy-
ranosyl)onate]pyridinium (3b). Compound 4b (150 mg, 0.23 mmol)
was dissolved in an ice-cold solution of sodium methoxide (10-15
equiv) in anhydrous methanol (5 mL). After the solution had stirred at
0 °C for 2 h, Dowex 50W HCR-W2 resin (H+ form) was added to
neutralized the solution. Following filtration the resin was washed twice
with methanol. The combined filtrate and methanol washings were
evaporated under reduced pressure, and the resulting residue was
dissolved in aqueous 0.1 M NaOH solution (5 mL). After being stirred
for 30 min at 0 °C, the mixture was neutralized with Dowex 50W HCR-
W2 (H+) resin, and the resin was removed by filtration. The resin was
then washed with water, and the combined aqueous layers were
evaporated to dryness under reduced pressure to give a pale yellow
solid. The crude product was purified by HPLC using 1.0% (v/v) acetic
acid in water as the eluent (retention time ) 9.52 min). The combined
1
lyophilization to afford a colorless solid. The H NMR spectrum of
the hydrolysis product was identical to a spectrum of authentic
N-acetylneuraminic acid.
The hydrolyses of 3a and 3b afford a small amount of an unidentified
product (∼2%), as shown by the appearance of a broad singlet peak at
1
5.88 ppm in the H NMR spectra. The hydrolyses of 3c-e also give
rise to a second minor product (∼5%), as shown by the appearance in
the 1H NMR spectra of a singlet at a chemical shift of 5.65. However,
neither of these two minor products was the corresponding unsaturated
sugar (i.e., glycal 6) that would result from an elimination reaction.
Enzymatic Materials and MethodssWild-Type Sialidase Gene.
All restriction endonucleases and DNA modification enzymes were
purchased from Gibco BRL or New England BioLabs Inc. (Beverly,
MA). All DNA manipulations were carried out according to standard
protocols.29 The BS-TokNA plasmid, containing the wild-type sialidase
gene, was a generous gift from Dr. G. Air (University of Oklahoma).
As described by Lentz et al., the gene was cloned into the SalI site of
(25) (a) Rothermal, J.; Faillard, H. Carbohydr. Res. 1990, 196, 29-40.
(b) Kuhn, R.; Lutz, P.; MacDonald, D. C. Chem. Ber. 1966, 99, 611-617.
(c) Roy, R.; Laferriere, C. Can. J. Chem. 1990, 68, 2045-2054.
(26) (a) Champa, R. A.; Fishel, D. L. Can. J. Chem. 1973, 51, 2750-
2758. (b) Bapat, J. B.; Epsztajn, J.; Katritzky, A. R.; Plau, B. J. Chem.
Soc., Perkin Trans. 1 1977, 1692-1697.
(27) Eschenfelder, V.; Brossmer, R. Carbohydr. Res. 1987, 162, 294-
297.
(28) We were unable to acquire a 13C NMR spectrum for 3a because of
this compound’s significant rate of hydrolysis at ambient temperatures. A
consequence of this reactivity can be seen in the emergence of peaks
(29) Maniatis, T.; Fritsch, E. F.; Sambrook, J. Molecular Cloning: A
Laboratory Manual, 2nd ed.; Cold Spring Harbor Laboratory Press: Cold
Spring Harbor, NY, 1989.
1
corresponding to 3-methoxypyridine in the H NMR spectrum (δ).

8360 J. Am. Chem. Soc., Vol. 122, No. 35, 2000
Chou et al.
the vector Bluescript KS (Stratagene, La Jolla, CA) in an orienta-
tion to allow expression from the T7 promoter.30,31 The NA gene was
excised by SalI digestion from the BS-TokNA plasmid, gel-purified,
and treated with T4 DNA polymerase to form blunt ends. The 1.5 kb
fragment containing the NA gene was then ligated into SmaI-digested
pUC19 (New England BioLabs Inc.) The new plasmid construct, pJW1,
was propagated in Escherichia coli DH5R before further manipulation.32
Insertion of NA into pVL1392 Shuttle Vector. The sialidase gene
was excised from pJW1 by EcoRI/BamHI digestion. The 1.5 kb
fragment was gel-purified and ligated into EcoRI/BamHI-digested vector
pVL1392 (PharMingen, San Diego, CA) to create the 11.1 kb plasmid
pJW2. DNA sequence analysis (University of Calgary) using the
dideoxy chain termination method33 was used to verify accurate cloning
and to ensure that spurious mutations had not occurred during DNA
manipulation.
Cells and Viruses. Trichoplusia ni insect cells were cultured in
EX-CELL 405 medium (JRH Biosciences, Lenexa, KS) supplemented
with 10% fetal bovine serum (JRH Biosciences) according to standard
procedures.34 Recombinant baculovirus stocks were produced by
cotransfecting 5 µg of the pJW2 recombinant shuttle vector with 1 µg
of BaculoGold AcNPV DNA (PharMingen). Single clones were isolated
by plaque assays.35 High-titer virus stocks were produced by three
rounds of virus amplification34 at a multiplicity of infection (MOI) of
0.1.
Figure 1. Plots of log(kobs) versus -log [H3O+] for the hydrolysis of
five R-D-N-acetylneuraminyl pyridinium salts, T ) 65 °C: 9, 3a; 1,
3b; 2, 3c; [, 3d; and b, 3e. Error limits are encompassed within the
symbol diameter. The included lines are the best nonlinear fits of the
kinetic data to eq 1.
Scheme 3
Expression of NA. T. ni cells (2 × 106 cells mL-1) were infected
with recombinant baculovirus at a MOI of 10 using serum-free medium,
supplemented with 1% antibiotic/antimycotic (Gibco BRL). Infected
cells were incubated at 27 °C in 1-L spinner flasks until levels of active
NA had reached a plateau, typically 60-65 h postinfection.
Sialidase Assay. The fluorometric substrate 4-methylumbelliferyl
R-D-N-acetylneuraminide (MUNANA) (Sigma Chemical Co.) was used
to monitor sialidase activity. Based on the method of Potier et al.,36
1-10 µL of sample was typically used in a 50-µL reaction volume
containing 200 mM NaAc, pH 6.0, 2 mM CaCl2, and 0.4 mM
MUNANA. Following a 30-min incubation at 37 °C, the enzymatic
reaction was stopped by the addition of 250 µL of 133 mM glycine,
pH 10.7, 60 mM NaCl, and 83 mM NaHCO3. A standard curve for the
measurement of product was prepared using 4-methylumbelliferone.
The fluorescence of standards and unknown samples was determined
using an excitation wavelength of 365 nm and an emission wavelength
of 450 nm.
Purification of Sialidase. T. ni cells were removed from the culture
medium by centrifugation. Eight-hundred-milliliter volumes of medium
were centrifuged at 10000g for 15 min. Recombinant virus particles
(decorated with NA) were isolated from the culture supernatant by a
second centrifugation at 40000g for 30 min at room temperature. Virus
pellets were resuspended in 50 mL of buffer A (50 mM HEPES, pH
7.3, 150 mM NaCl, and 2 mM CaCl2). Active sialidase heads were
cleaved from the baculovirus particles by Pronase digestion as de-
scribed by Laver,37 using 0.05 mg/mL Pronase (Calbiochem, La Jolla,
CA) at 37 °C for 16 h. Virus particles were separated from NA by
centrifugation as before. To separate NA and Pronase, supernatants
were loaded onto a 10-25% sucrose gradient and centrifuged in an
SW28 rotor at 28 000 rpm for 29 h at 5 °C. Fractions containing NA
but not containing Pronase were pooled, concentrated, and exchanged
into buffer B (50 mM HEPES, pH 8, 150 mM NaCl, and 2 mM CaCl2)
using ultrafiltration (30 kDa cutoff, Amicon). Concentrated 1-mL
samples were then loaded onto a Sephacryl-200-HR (Sigma Chemical
Co.) gel filtration column, equilibrated with buffer B. The purity of
the active fractions was assessed by SDS-PAGE and silver staining.
Aliquots of the pure enzyme were then stored frozen at -80 °C until
needed.
Enzyme Kinetics. Michaelis-Menten parameters were measured
by standardizing enzyme activity on 4-nitrophenyl R-D-N-acetyl-
neuraminide to obtain relative kcat values for the pyridinium N-
acetylneuraminide substrates. Kinetic measurements at pH 6.0 were
performed in 50 mM sodium acetate, 0.1 mM CaCl2, 0.32 mM MgCl2,
and 60 mM NaCl.21 Measurements at pH 9.5 were performed in 50
mM glycine, 0.1 mM CaCl2, 0.32 mM MgCl2, and 60 mM NaCl.21
Initial rate measurements were made using a Varian Cary 3E UV-vis
spectrophotometer equipped with a Peltier temperature controller. The
progress of the reactions was continuously monitored for 10-20 min
at 37 °C, and the derived initial velocities were corrected for background
hydrolysis. Each 400-µL reaction was performed by equilibrating the
buffer and substrate in the cell block for 3-5 min prior to the addi-
tion of enzyme (50 µL). Kinetic parameters were determined from eight
initial rate measurements at substrate concentrations ranging from Km/3
to 3Km. Extinction coefficient differences were calculated by measuring
the absorbances of the pyridinium N-acetylneuraminide substrate and
those of the products, N-acetylneuraminic acid and the corresponding
pyridine. The kinetic rate data were fitted to the Michaelis-Menten
equation using a standard nonlinear least-squares program (Grafit).
Results
(30) Lentz, M. R.; Air, G. M.; Laver, W. G.; Webster, R. G. Virology
1984, 135, 257-265.
The N-(R-D-N-acetylneuraminyl)pyridinium compounds were
prepared from a readily made and fully protected â-D-N-
acetylneuraminyl chloride (5)25 according to the synthetic route
outlined in Scheme 3. To some extent dehydrohalogenation of
the tertiary â-D-N-acetylneuraminyl chloride (5) was anticipated,
generating an R,â-unsaturated ester product. Nevertheless, the
silver-promoted reaction of 5 with various pyridines generated
the desired salts in reasonable yields (40-60%).
(31) Nuss, J. M.; Air, G. M.Virology 1991, 183, 496-504.
(32) A gift from Dr. R. Kay, Terry Fox Laboratories, Vancouver, Canada.
(33) Sanger, F.; Nicklen, S.; Coulson, A. R. Proc. Natl. Acad. Sci. U.S.A.
1977, 74, 5463-5467.
(34) Summers, M. D.; Smith, G. E. Texas Agricultural Experiment Station
Bulletin No. 1555, 1987.
(35) O’Reilly, D. R.; Miller, L. K.; Luckow, V. A. BaculoVirus
Expression Vectors: A Laboratory Manual; W. H. Freeman and Co.: New
York, 1994.
(36) Potier, M.; Mameli, L.; Belisle, M.; Dallaire, L.; Melancon, S. B.
Anal. Biochem. 1979, 94, 287-296.
(37) Laver, W. G. Virology 1978, 86, 78-87.
Spontaneous Hydrolysis. Figure 1 presents the observed rate
constants (kobs) obtained as a function of [H3O+] for the

Hydrolysis Reactions of N-Acetylneuraminides
J. Am. Chem. Soc., Vol. 122, No. 35, 2000 8361
Scheme 4
The absolute values for kcat and kcat/Km on the enzyme-
catalyzed hydrolyses of pyridinium R-D-N-acetylneuraminides
are shown in Figure 4. The assumption used in deriving these
values is that the recombinant and nonrecombinant forms of
the enzyme have the same absolute kcat value.
Despite several attempts, it was not possible to derive kinetic
parameters for the hydrolysis of 3,4-dimethylpyridinium (3e)
at pH 6.0. Hydrolysis of this compound produced too small a
change in absorbance (∆ꢀ, Table 2) to allow the reproducible
measurement of R-sialidase-catalyzed kinetic parameters. Also,
accurate parameters could not be derived for enzyme-catalyzed
hydrolysis of 3a because the rate of spontaneous hydrolysis was
significantly faster than the rate of enzymatic hydrolysis. At a
pH of 9.5, the âlg values for sialidase-catalyzed hydrolyses of
the series 3b-e were +0.17 ( 0.27 and 0.00 ( 0.12 for kcat
and kcat/Km, respectively.
hydrolyses of the five salts (3a-e). Individual rate constants
are given in Table S1 of the Supporting Information. The
observed rate constants for 3a-e were fit to eq 1, where k1 and
k2 are the first-order rate constants for the hydrolysis of the
zwitterion and cation, respectively, and Ka is the apparent
ionization constant of the carboxylic acid group (Scheme 4).
k1Ka + k2[H3O+]
Ka + [H3O+]
kobs
)
(1)
Discussion
Aryl N-acetylneuraminides exist in different protonation
states. Therefore, these carbohydrates are potentially hydrolyzed
by different independent pathways (Scheme 1),8 and conse-
quently the catalytic function of the anomeric carboxylic acid
group is obscured.
Typically, the accurate description of catalyzed and uncata-
lyzed reaction mechanisms necessitates kinetic analysis at
many different pH’s. However, substrates with positively
charged leaving groups are able to simplify many aspects of
the hydrolysis kinetics, because acid-catalyzed pathways are no
longer relevant. Furthermore, such substrates may be used to
dissect the catalytic machinery of sialidases by negating the
catalytic role of one of the two active-site carboxylic acid
residues.
Acidity of Pyridinium N-Acetylneuraminides. Although the
calculated pKa’s for the cations of 3a-e (Table 1) might appear
to be low, these values are not unexpected. The measured pKa
of (1-pyridino)acetic acid (Py+CH2CO2H) is 1.72 ( 0.01 (µ )
1, KCl; T ) 65 °C). When this value is adjusted to account for
the inductive effect of the ring oxygen atom (methoxyacetic
acid is 1.18 pKa units more acidic than acetic acid),39 the
pKa for 3b is expected to be ∼0.5 (µ ) 1, KCl; T ) 65 °C),
which is in good agreement with the calculated value of 0.46
(Table 1).
Spontaneous Hydrolysis ReactionssRole of the C-1 Car-
boxylate Group. The observed rate constants for hydrolysis of
pyridinium R-D-N-acetylneuraminides (3a-e) are consistent
with spontaneous hydrolyses of both the zwitterion, k1, and the
cation, k2; the pKa of the cation is in the range of 0.4-0.7
(Scheme 4).
In principle, the carboxylic acid group could exert an effect
on the reactivity of R-D-N-acetylneuraminides via an intramo-
lecular nucleophilic attack to form a transient R-lactone.8
However, the carboxylate group does not appear to play the
role of a nucleophile during the C-N bond cleavage reac-
tions for the following reasons: (1) the rate differences be-
tween the reactions of the carboxylate (k1) and carboxylic acid
(k2) only differ by approximately a factor of 3 (Table 1),
whereas, with the comparable acetate/acetic acid pair, a rate
enhancement of at least 104-fold is anticipated;41 (2) the âlg
values for the two processes (k1 and k2) are identical, and
therefore consistent with late TSs having similar degrees of
C-N bond cleavage, (i.e., both processes display characteristics
The observed rate constants were determined up to an acid
concentration of 1.0 M; beyond this concentration of acid, (1)
the ionic strength is no longer constant and (2) the sialic acid
product becomes increasingly unstable, thus making it difficult
to measure accurate kinetics. Therefore, the values for k2 were
extrapolated since the plateau at high [H3O+] was not achieved.
At one extreme k2 is indistinguishable from zero and the
observed fall-off in rate constant is caused solely by titration
to an inactive form. To discount this possibility, the data were
also fit to a model in which k2 ) 0 (eq 1). Shown in Figures
S1-S5 (Supporting Information) are the residuals of the fits to
the unconstrained and the constrained model (k2 ) 0). It is
evident that the residuals are unevenly distributed about zero
for the constrained fit. The data-fitting exercise allows us to
conclude that there are two independent pathways for the
hydrolysis of the pyridinium N-acetylneuraminides, and these
pathways are reasonably described by the parameters (k1, k2,
and pKa) listed in Table 1. Also given in Table 1 are the pKa’s
for the conjugate acids of the pyridine leaving groups.38
Figure 2 displays the Brønsted plots, log(k1) and log(k2) versus
the pKa of the conjugate acid of the leaving group for the
spontaneous reactions of 3a-e. The derived âlg values for these
reactions are -1.22 ( 0.07 and -1.22 ( 0.16 for the
zwitterionic (k1) and the cationic (k2) forms, respectively.
Analysis of the hydrolytic products for 3a-e showed that
the major carbohydrate formed (g95%) is N-acetylneuraminic
acid, thus indicating that these hydrolysis reactions involve
heterolytic C-N bond cleavages.
The rate of 3b hydrolysis (Table S2, Supporting Information)
was shown to be independent of acetate concentration; therefore,
buffer catalysis is not a factor. The first-order rate constants
for the hydrolysis of the zwitterionic form of 3b were also
measured as a function of temperature (Table S3, Supporting
Information), and the corresponding Eyring plot is shown in
Figure 3. The activation parameters derived for the spontaneous
hydrolysis of 3b are ∆Hq ) 112 ( 2 kJ mol-1 and ∆Sq ) 28
( 4 J mol-1 K-1
.
Enzyme-Catalyzed Hydrolysis. Sialidase concentrations
were normalized using a 4-nitrophenyl R-D-N-acetylneuraminide
substrate, thereby making possible a direct comparison of the
enzyme-catalyzed hydrolyses of pyridinium R-D-N-acetyl-
neuraminides with the published values for aryl R-D-N-acetyl-
neuraminides.21 The kinetic parameters for viral sialidase-
catalyzed hydrolyses are listed in Tables 2 and 3 in terms of
relative kcat and absolute Km values.
(38) Perrin, D. D. Dissociation Constants of Organic Bases in Aqueous
Solution; Butterworth: London, 1965.
(39) pKa(AcOH) ) 4.75; pKa(MeOCH2CO2H) ) 3.57, ref 40.

8362 J. Am. Chem. Soc., Vol. 122, No. 35, 2000
Chou et al.
Table 1. Calculated Rate Constants for the Spontaneous Hydrolysis of the Zwitterionic (k1) and Cationic (k2) Forms of the Pyridinium Salts
(3a-e) at 65 °C (µ ) 1.0, KCl)
leaving group
pKa (Py-H+)a
105k1 (s-1
)
105k2 (s-1
)
pKa (app)
3-methoxypyridine
pyridine
3-methylpyridine
4-methylpyridine
3,4-dimethylpyridine
4.91
5.22
5.63
5.98
6.46
291 ( 1
69 ( 9
0.36 ( 0.04
0.46 ( 0.04
0.47 ( 0.04
0.55 ( 0.05
0.74 ( 0.05
89.8 ( 0.4
31.8 ( 0.2
9.79 ( 0.08
3.68 ( 0.03
32.9 ( 2.4
6.4 ( 1.0
1.86 ( 0.37
1.23 ( 0.10
a Values taken from ref 38.
Table 3. Michaelis-Menten Parameters for the Influenza Type A
Sialidase-Catalyzed Hydrolysis of Pyridinium
R-D-N-Acetylneuraminides, at pH 9.5 and 37 °C.
∆ꢀ
wavelength
(nm)
(cm-1 M-1
)
rel kcat
Km (µM)
a
3b
3c
3d
3e
-3470
-3610
-3690
-2180
261
275
256
264
0.151 ( 0.012
0.0675 ( 0.0067
0.215 ( 0.023
0.178 ( 0.026
216 ( 44
138 ( 37
450 ( 120
261 ( 87
a Reported kcat values are relative to that measured for 4-nitrophenyl
R-D-N-acetylneuraminide, pH 6.0, at 37 °C.
Figure 2. Brønsted plots for the hydrolyses of the five pyridinium
R-D-N-acetylneuraminides, T ) 65 °C: [ (k1), hydrolysis of the
zwitterion; and b (k2), hydrolysis of the cation. The included lines are
the best linear fits to the data.
Figure 3. Eyring plot for the hydrolysis of the zwitterionic form of
pyridinium R-D-N-acetylneuraminide.
Table 2. Michaelis-Menten Parameters for the Influenza Type A
Sialidase-Catalyzed Hydrolysis of Pyridinium
R-D-N-Acetylneuraminides, at pH 6.0 and 37 °C.
∆ꢀ
wavelength
(nm)
(cm-1 M-1
)
rel kcat
Km (µM)
a
Figure 4. Plots of log(kcat) and log(kcat/Km) versus pKa(B-H+) for
R-sialidase-catalyzed hydrolysis of four pyridinium R-D-N-acetyl-
neuraminides, pH 6.0 (b) and pH 9.5 (O), at T ) 37 °C. The included
lines are the best linear fits to the pH 9.5 data.
3b
3c
3d
3e
-3450
-2600
-3640
-1400
261
275
256
275
0.140 ( 0.007
0.0214( 0.0020
0.385 ( 0.022
72 ( 12
16 ( 6
234 ( 28
oxacarbenium ion can be estimated by a kinetic comparison of
the N-acetylneuraminyl and the 2-deoxyglucopyranosyl sys-
tems.6 Specifically, the uncatalyzed hydrolysis for the zwitter-
ionic form of 3b was compared to the 2-deoxyglucosyl iso-
quinolinium salt 7. The comparison between 3b and 7 is valid
a Reported kcat values are relative to that measured for 4-nitrophenyl
R-D-N-acetylneuraminide, pH 6.0, at 37 °C.
of SN1 reactions); and (3) the large positive ∆Sq value for the
hydrolysis of the zwitterion (k1) is not consistent with a highly
ordered TS such as would be required for an intramolecular
nucleophilic attack.
Lifetime of the N-Acetylneuraminyl Oxacarbenium Ion.
The influence of the carboxylate group on the lifetime of an
(40) Serjeant, E. P.; Dempsey, B. Ionisation Constants of Organic Acids
in Aqueous Solution; Pergamon Press: Oxford, 1979.
(41) Acetic acid is about 100-fold less nucleophilic than water on the
basis of NOTs values, while water is about 500-fold less nucleophilic than
acetate on the basis of Swain-Scott n values, ref 42.
because (1) the pyridinium ring in both compounds is located
in an equatorial position, (2) both compounds have ground-
state chair conformations, (3) the pKa’s for the conjugate acids

Hydrolysis Reactions of N-Acetylneuraminides
J. Am. Chem. Soc., Vol. 122, No. 35, 2000 8363
of pyridine and isoquinoline are similar,38 and (4) any structural
differences are remote from the anomeric center.
Scheme 5
In a previous study, Huang et al. correlated the lifetimes of
the 2-deoxyglucosyl and the glucosyl cations with their rates
of formation.6a To wit, a 680-fold change in reactivity was
associated with a 4-fold change in cation lifetime.6a In the
-
present study, addition of an anomeric CO2 group enhances
the reactivity of the pyridinium salt by approximately 200-fold
relative to that of the 2-deoxy compound 7. This rate enhance-
ment was estimated from the pH-independent rate constants for
the hydrolyses of 7 (µ ) 2.0 M, NaClO4)6a and 3b (µ ) 1.0 M,
catalyzed hydrolyses of aryl R-D-N-acetylneuraminides (k5,
Scheme 5). However, both the initial conformational change
(k2) and cleavage of the glycosidic C-O bond (k3) are partially
rate-determining for kcat/Km in these sialidase-catalyzed reactions
(Scheme 5).21
1
KCl), adjusted by a factor of /3 to account for the differences
in ionic strength and added salt.6a Hence, the zwitterionic form
of the N-acetylneuraminyl oxacarbenium ion is estimated to have
a lifetime of approximately 3 times longer than that of the
corresponding 2-deoxyglucosyl cation (i.e., 3 × 10-11 s).6 This
figure is in agreement with the estimate of g3 × 10-11 s made
by Horenstein and Bruner for the lifetime of the N-acetyl-
neuraminyl oxacarbenium ion.10
The cationic form of 3b reacts more rapidly than 7 despite
the presence of the electron-withdrawing carboxylic acid group.
For example, Amyes et al. noted that addition of an ester group
in 4-methoxybenzyl derivatives caused a 40-fold decrease in
rate.43 The approximately 60-fold faster reaction of the proto-
nated form of 3b relative to that of 7 may be explained by a
steric argument. Specifically, the ground state for the pyridin-
ium N-acetylneuraminides contains severe 1,3-diaxial interac-
tions that are relieved upon approach to the reaction TSs;
therefore, in the current example, any electronic factors are
outweighed.
Influenza Virus Type A (Tokyo 3/67) r-Sialidase-Cata-
lyzed Hydrolyses. A long-term objective is to produce a
complete description of the enzyme-catalyzed hydrolysis of
N-acetylneuraminides, dissecting the contributions to catalysis
of individual residues in the sialidase active site. This objective
will be accomplished by a combination of studies involving site-
directed mutagenesis on a viral sialidase and the use of novel
substrates.
As a first step in dissecting the individual roles of catalytic
residues, hydrolysis of the pyridinium salts (3a-e) was studied.
There is no evidence in the current study for rate-determining
glycosidic C-N cleavage at either pH 6.0 or 9.5 (Figure 4).
Given that the absolute values for kcat/Km are around 104 M-1
s-1, it is unlikely that substrate binding is rate-determining for
the catalyzed hydrolyses of 3a-e. However, a rate-determining
conformational change is consistent with previously observed
rates for conformational changes on other enzyme systems.44,45
In the current example a conjoint enzyme-substrate confor-
mational change will result in the pyridinium moiety being
placed in a thermodynamically unfavorable pseudoaxial orienta-
tion. Also, it is reasonable to expect that the conjoint change
will be slower for the pyridinium substrates than for the aryl
neuraminides. However, once the Michaelis complex com-
pletes its transition (k2, Scheme 5), then C-N bond cleavage is
facile and not rate-determining for kcat/Km. The results shown
in Figure 4 for the correlation of kcat and the pyridinium pKa’s
suggest that chemistry is not rate-limiting; thus, it is pro-
posed that kcat is determined by a virtual transition state
comprised of the two required conformational changes (k2 and
k5, Scheme 5).
We conclude that there is a fine balance between the rates of
conformational changes on the enzyme and the rate of C-X
bond cleavage reactions. Therefore, subtle changes in leaving
group chemistry can alter the various rate-determining steps.
Recombinant influenza virus sialidase was expressed using
a baculovirus/insect cell system and then purified. The use of a
recombinant enzyme will facilitate future studies of catalysis
because the enzyme is amenable to site-directed mutagenesis.
Purification of the recombinant enzyme was similar to that of
the natural sialidase; however, it should be noted that the enzyme
is a glycoprotein and the insect cell expression system does not
glycosylate proteins in an identical manner to that of animal
cells. A Km value of 93 ( 13 µM was observed for the
recombinant sialidase-catalyzed hydrolysis of 4-nitrophenyl R-D-
N-acetylneuraminide. This value is lower than that reported for
a similar enzyme isolated directly from a viral strain.21 It is not
known whether the difference reflects a strain-to-strain variation
in the activity of the sialidases or rather reflects the difference
in protein glycosylation.
In 1994, Guo et al. proposed a mechanistic scheme for
sialidase-catalyzed hydrolyses which contains two important
features.21 First, the Michaelis complex undergoes a conforma-
tional change that results in a chair-to-boat transition of the
substrate, and second a similar conformational change from a
boat to a chair occurs for the product Michaelis complex
(Scheme 5).
Conclusions
In conclusion, the pyridinium R-D-N-acetylneuraminides
effectively simplified interpretation of spontaneous and enzy-
matic hydrolysis kinetics. Therefore, spontaneous hydrolyses
were observed to proceed via two independent pathways
where C-N bond cleavage was unassisted and rate-deter-
mining. Moreover, the anomeric carboxylate does not nucleo-
philically assist in the departure of the neutral leaving group
during hydrolysis of zwitterionic species, as previously con-
jectured.8
In the case of influenza R-sialidase, the pyridinium R-D-N-
acetylneuraminides were used to show that kcat/Km is determined
by a conformational change of the first-formed Michaelis
complex, whereas kcat is a virtual transition state made up of
two separate conformational events.
Acknowledgment. The authors gratefully acknowledge the
Natural Sciences and Engineering Research Council of Canada
for financial support of this work. The authors also thank Dr.
David McGillivray for acquiring the HRMS. J.N.W. thanks BC
(43) Amyes, T. L.; Stevens, I. W.; Richard, J. P. J. Org. Chem. 1993,
Guo et al. propose that a conjoint change of the initially
formed enzyme-product complex limits kcat for the sialidase-
58, 6057-6066.
(44) Sawicki, C. A.; Gibson, Q. H. J. Biol. Chem. 1977, 252, 5783-
5788.
(42) Connors, K. A. Chemical Kinetics: The Study of Reaction Rates in
Solution; VCH Publishers Inc.: New York, 1990.
(45) Eckfeldt, J.; Hammes, G. G.; Mohr, S. C.; Wu, C. Biochemistry
1970, 9, 3353-3362.

8364 J. Am. Chem. Soc., Vol. 122, No. 35, 2000
Chou et al.
Science Council for receipt of a GREAT scholarship. We also
thank Professor Derek Horton for his help with IUPAC
nomenclature of carbohydrates.
function of both temperature and sodium acetate concentration;
plots of the residuals from the constrained and unconstrained
fits to eq 1 of the kobs versus -log [H3O+] data for the hy-
drolyses of 3a-e (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
Supporting Information Available: Experimental details
1
for the synthesis of 3a,c,d,e; H NMR spectra for 3a-e, and
13C NMR spectra for 3b-e; tables of observed rate constants
for the hydrolysis reactions of 3a-e at 65 °C, and 3b as a
JA001641X