A mechanistic study on the α-N-acetylgalactosaminidase from E. meningosepticum: A family 109 glycoside hydrolase

Article abstract of DOI:10.1039/c4md00104d

A recombinant glycoside hydrolase family 109 α-N- acetylgalactosaminidase from the pathogenic bacteria E. meningosepticum catalyses the hydrolysis of aryl 2-acetamido-2-deoxy-α-d- galactopyranosides. The sensitivities to leaving group abilities (β_lg values) on V and V/K are -0.08 ± 0.06 and -0.31 ± 0.12, respectively. These results are consistent with an E2 elimination following hydride transfer from C3. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4md00104d

MedChemComm
View Article Online
View Journal | View Issue
CONCISE ARTICLE
A mechanistic study on the a-N-
acetylgalactosaminidase from E. meningosepticum:
a family 109 glycoside hydrolase†
Cite this: Med. Chem. Commun., 2014,
5, 1188
a
b
a
*
Saswati Chakladar, Saeideh Shamsi Kazem Abadi and Andrew J. Bennet
Received 7th March 2014
Accepted 24th April 2014
A recombinant glycoside hydrolase family 109 a-N-acetylgalactosaminidase from the pathogenic bacteria
E. meningosepticum catalyses the hydrolysis of aryl 2-acetamido-2-deoxy-a-D-galactopyranosides. The
sensitivities to leaving group abilities (blg values) on V and V/K are ꢀ0.08 ꢁ 0.06 and ꢀ0.31 ꢁ 0.12,
respectively. These results are consistent with an E2 elimination following hydride transfer from C3.
DOI: 10.1039/c4md00104d
www.rsc.org/medchemcomm
Indeed, three out of the four glycoside hydrolase families that
have a-NAGAL enzymes as members (GH27, GH36 and GH129)
Introduction
are proposed to operate via such nucleophilic substitution
reactions.12,14
There is a growing awareness in the scientic community that
post-translational protein glycosylation is a complex modica-
tion, found throughout nature,1 that individualizes biomole-
cules and imparts physical properties that dene biomolecular
roles in a range of physiological and biological recognition
events. a-N-Acetylgalactosamine (aGalNAc) residues are
commonly found as constituents of mucin glycoproteins,2,3 and
they are critical antigenic components of blood group A4 and of
the sensory nerve structures of humans.5 Enzymatic removal of
the antigenic aGalNAc residue from blood group A cells has
been studied as a potential route for the production of universal
red blood cells (RBCs) for transfusion.6–8 Glycosidases (glycoside
hydrolases: GHs) constitute a superfamily of enzymes that
hydrolyse glycosidic bonds found in a wide range of glyco-
conjugates. Based on sequence alignment, these enzymes are
classied into more than 130 different families.9–11 a-N-Ace-
tylgalactosaminidases (a-NAGALs) are carbohydrate-processing
enzymes that cleave terminal non-reducing aGalNAc residues
from glycoconjugates. a-NAGALs have been classied into GH
families 27, 36, 109 and 129.8,12–14 Of these, GH27 and GH36
share a common ancestral gene and are members of the
glycoside hydrolase clan GH-D, which possess a commonly
found triosephosphate isomerase (TIM) barrel structure.9
Most GHs15 catalyze simple nucleophilic substitution reac-
tions that occur on the anomeric carbon atom of a carbohy-
drate. The enzymatic transition state for this reaction involves
the development of positive charge on the anomeric carbon,
which then can delocalize onto the ring oxygen atom.16,17
In 2007, Liu et al. reported the X-ray crystal structure of an a-
NAGAL enzyme from E. meningosepticum, which is a member of
the exclusively prokaryotic GH109 family.8 The X-ray diffraction
structure revealed several similarities between GH4 glycosi-
dases and GH109 a-NAGALs, as well as a remote relationship to
the sequences of oxidoreductases.8,18 An unusual feature of the
GH4 family is that various cofactors are required for activity, in
particular, NAD+, a divalent metal ion (Mn2+) and a reducing
agent (such as DTT or TCEP).19–22 Scheme 1 shows the currently
accepted mechanism for GH4 NAD+-dependent glycosidases,
using a-galactosidase as an example; the NAD+-dependent
glycosidase reaction mechanism is atypical of glycosidases. In
Scheme 1, the enzyme:ketone complex is shown as a bona de
intermediate,20–22 however, based on kinetic isotope effect
evidence it has been proposed that the E:S complex undergoes a
concerted reaction to give the E:enediolate intermediate
directly.19 Key observations that underpin this mechanism
include: (i) a requirement for NAD+ for activity, (ii) the rst
formed product has the same anomeric conguration as the
substrate (retaining glycoside hydrolase), and (iii) when the
reaction is performed in D2O the hydrolysis product contains a
deuterium on C-2.20–22 Given that Liu et al. reported similar
observations for their E. meningosepticum GH109 a-NAGAL it is
clear that in addition to sharing a structural similarity, GH4 and
GH109 enzymes are mechanistically comparable.8,23 Lastly, the
GH4 family contains enzymes that catalyze the hydrolysis of
both a- and b-glycoside substrates,18,24 an unusual situation that
is not seen in GH109 where all members characterized to date
are a-NAGALs that show strong substrate selectivity.8
aDepartment of Chemistry, Simon Fraser University, Burnaby, BC V5A 1S6, Canada.
E-mail: bennet@sfu.ca; Fax: +1-778-782-3765; Tel: +1-778-782-8814
bDepartment of Molecular Biology and Biochemistry, Simon Fraser University,
Burnaby, BC V5A 1S6, Canada. Fax: +1-778-782-5583
In contrast to GH4 glycosidases, the GH109 family enzyme
E. meningosepticum a-N-acetylgalactosaminidase requires no
external cofactors for activity. That is, the required NAD+ is
embedded in the enzyme environment where it remains tightly
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c4md00104d
1188 | Med. Chem. Commun., 2014, 5, 1188–1192
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
Scheme 1 Proposed mechanism for GH4 glycoside hydrolases showing all possible intermediates in the catalytic cycle. Hydride transfers are
shown in red, while proton transfers to and from C2 are shown in blue.19–22 The required divalent cation (Mn2+) is not shown in all structures in
order to avoid overlap with mechanistic arrows. Also shown is the incorporation of deuterium at C2 in the product when the reaction is per-
formed in D2O.
bound to the protein.8 Because the E. meningosepticum enzyme
does not require a divalent cation, the pKa of the C2 proton of
Substrate synthesis
the intermediate (E:ketone, Scheme 1) is likely to be higher than
that of the corresponding species bound to the GH4 enzyme. In
contrast, the electrophilic stabilization provided by the bound
divalent cation (Mn2+) in GH4 enzymes should provide a
dramatic lowering of the free energy of the E:enediolate inter-
mediate (Scheme 1). This major difference in the GH4 and
GH109 enzyme active sites piqued our interest and provided
incentive for undertaking an in-depth mechanistic study on a
GH109 family member. In this report, we detail the cloning of
the v109 gene from E. meningosepticum (ATCC 51720D) as well
as the expression, purication, and characterization of the
encoded protein (a-NAGAL).
A panel of ve substituted aryl 2-acetamido-2-deoxy-a-D-gal-
actopyranosides were synthesized and used in combination
with the commercially available 4-nitrophenyl substrate (3a) to
assess the effect of leaving group ability on the two catalytic
Enzyme production
We amplied the genomic DNA of E. meningosepticum—using a
different strain than that reported previously in the literature8—
and used this material to generate a truncated a-N-acetylga-
lactosaminidase lacking the rst 17 N-terminal amino acid
residues. In addition, our construct differs in sequence from the
enzyme used by Liu et al.8 by six amino acid residues. Full
experimental details and protein and DNA sequences are given
in the ESI† section. We mapped the six amino acid changes onto
the published structure of the E. meningosepticum a-N-ace-
tylgalactosaminidase (Fig. 1) and found that none are within the
active site of the enzyme.
Fig. 1 The structure for a single monomer of the E. meningosepticum
a-N-acetylgalactosaminidase, where the NAD+ and GalNAc bound to
the active site are shown as stick structures. The positions of the six
amino acid residues that differ between the a-NAGAL used in the
current study and that reported by Liu et al. are shown using space-
filling structures.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 1188–1192 | 1189

View Article Online
MedChemComm
Concise Article
Table 1 Kinetic parameters for GH109 E. meningosepticum a-N-
acetylgalactosaminidase-catalyzed hydrolysis of PNPaGalNAc
Parameter
kcat (sꢀ1
)
kcat/Km (Mꢀ1 sꢀ1
)
Maximal rate constants
19.7 ꢁ 1.0a
7.0 ꢁ 0.1
7.9 ꢁ 0.1
3.4 ꢁ 0.2
2.8 ꢁ 0.2
(1.25 ꢁ 0.05) ꢃ 106a
7.3 ꢁ 0.1
pKa1
pKa2
h1
7.8 ꢁ 0.1
3.7 ꢁ 1.4
h2
2.5 ꢁ 0.8
a
Rate constants reported by Liu et al. kcat ¼ 9.84 ꢁ 0.16 sꢀ1 kcat/Km
¼
(1.28 ꢁ 0.10) ꢃ 106 Mꢀ1 sꢀ1
.
8
of the anomeric protons for both anomers as singlets in the 1H
NMR spectrum.
Scheme 2 Reagents and conditions for the syntheses of aryl 2-acet-
amido-2-deoxy-a-D-galactopyranoside substrates: (i) ArOH, SnCl4,
CH2Cl2, rt; (ii) TBAF, THF, reflux; (iii) Ac2O, pyridine, rt; (vi) NaOMe,
MeOH, 0 ꢂC.
The measured pH versus rate proles for kcat/Km and kcat
(Fig. 2) for the E. meningosepticum GH109 a-NAGAL-catalyzed
hydrolysis of PNPaGalNAc are bell-shaped curves. In theory, a
bell-shaped curve for the pH prole of kcat means that the
enzyme active site has two ionisable groups (one protonated
and the other deprotonated during the catalytic cycle).25 In the
case of our enzyme, which exists in solution as a functional
dimer, the experimental data only t to a pH-rate equation that
incorporated Hill coefficients on both of the ionization events
(Table 1). One possibility for the derived Hill coefficients on the
changes in a-NAGAL activity as a function of pH being greater
than 2 is that the two active sites in the dimeric enzyme—one
from each monomeric unit—are not functionally independent
of one another.25 Our GH109 a-NAGAL enzyme exhibits a
narrow peak of activity in a pH range of 7.0–8.0 (Fig. 2), and the
two apparent pKa values that characterize its activity are around
7.1 and 7.9 (Table 1).
The maximal rate constants measured in our current study
are similar to those reported by Liu et al. (Table 1).8 Thus, we
conclude that the six amino acid differences between the two
enzymes do not signicantly affect their catalytic activities.
Next, we evaluated the sensitivity of the enzymatic rate
constants kcat and kcat/Km to a change in the leaving group
ability. This was accomplished by measuring the change in the
constants (kcat and kcat/Km). We used stannic chloride promoted
glycosylation of methylcarbamate 1 for the synthesis of the
protected aryl a-galactosaminides 2b–f (Scheme 2). Standard
reactions gave access to substrates 3b–f. Full experimental
details are given in the ESI† section.
Kinetic and product studies
In order to verify mechanistic commonality between the
enzymes of glycoside hydrolase families 4 and 109 (GH4 and
GH109) and to conrm that the six amino acid changes to our
enzyme are mechanistically unimportant we performed a
product study for the a-NAGAL-catalyzed hydrolysis of 4-nitro-
phenyl
2-acetamido-2-deoxy-a-D-galactopyranoside
(PNPa
GalNAc) in the presence of methanol. The product of this
reaction contained an anomeric proton with a chemical shi of
5.22 (J1,2 ¼ 3.8 Hz), thereby conrming that the reaction pro-
ceeded with retention of anomeric conguration to give methyl
2-acetamido-2-deoxy-a-D-galactopyranoside. We also conrmed
that a-NAGAL-catalyzed hydrolysis of PNPaGalNAc in D2O gave
the reaction product 2-acetamido-2-deoxy-D-galactose which
was completely deuterated at C-2 as indicated by the appearance
Fig. 3 Brønsted plots. Effect of leaving group ability on kcat/Km (red
circles) and kcat (blue circles) for a-N-acetylgalactosaminidase-cata-
lyzed hydrolyses. All experiments were performed at 37 ꢂC and pH 7.5.
Leaving group abilities represented as pKa (ArOH) are as follows: 4-
Fig. 2 pH activity profiles: kcat/Km (left in red) and kcat (right in blue) for nitrophenol (7.15); 3-nitrophenol (8.36); 3-chlorophenol (9.12); 4-
the hydrolysis of PNPaGalNAc by a-NAGAL.
chlorophenol (9.41); phenol (9.98); and 4-methoxyphenol (10.13).
1190 | Med. Chem. Commun., 2014, 5, 1188–1192
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
logarithm of the rate constant as a function of the pKa of the site promotes binding with the C2 substituent via hydrophobic
leaving group aglycone's conjugate acid.26 Fig. 3 presents and hydrogen-bonding interactions with methionine and tyro-
Brønsted plots of the data; the derived sensitivities (blg values) sine residues, respectively (Fig. 4). We also note that the struc-
on kcat and kcat/Km are ꢀ0.08 ꢁ 0.06 and ꢀ0.31 ꢁ 0.12, ture of a-N-acetylgalactosaminidase does not have an
respectively.
appropriately placed acidic residue to assist in aglycone
The mechanism of action of GH4 and GH109 enzymes departure.8
involves a hydride transfer from C3 to a bound NAD+ co-factor to
Based on the blg value on kcat of close to zero and if the
give a non-covalently bound ketone (Scheme 1). Following the Brønsted blg value on kcat/Km is non-zero, the kinetically
formation of this intermediate,20,22 or during hydride transfer,19 signicant step for kcat occurs aer glycosidic bond cleavage.
a proton is transferred from C2 of the sugar to an enzymatic This step possibly involves either the Michael addition of water
tyrosine residue.8,18,24 This transfer occur either simultaneously to the glycal–ketone intermediate—which would likely occur
with (i.e., an E2 elimination mechanism) or prior to (i.e., an through a network of solvent molecules due to the lack of an
elimination reaction that proceeds via a conjugate base, E1CBirr
aglycone departure to give a bound ketone–glycal intermediate subsequent transfer of either a proton or a hydride.
(Scheme 1, E:glycal–ketone). The measured blg value on kcat/Km Given that both GH4 and GH109 enzymes are only produced
)
acid/base catalyst proximal to the glycosidic oxygen—or the
(ꢀ0.31 ꢁ 0.12), is consistent with cleavage of the anomeric C–O by prokaryotes, the design of selective inhibitors for these two
bond being kinetically signicant, although scatter in the families of glycoside hydrolases is attractive from an antimi-
Brønsted plot makes it difficult to come to a more denitive crobial therapeutics perspective.
conclusion. If glycosidic bond cleavage is partially rate-deter-
mining then breaking this bond is likely to be part of a
concerted E2 elimination reaction which by virtue of the good
Conclusions
leaving groups used in this study, would occur without acid- The E. meningosepticum a-N-acetylgalactosaminidase hydrolyses
catalysis. Mechanistic studies on GH4 enzymes report smaller 2-acetamido-2-deoxy-a-D-galactopyranosides by a NAD+-medi-
blg values20,22 than our values for the GH109 enzyme; the GH4 ated oxidation followed by an a,b-elimination to give a Michael
results are interpreted to mean that the elimination occurs via acceptor intermediate. This intermediate undergoes hydration
an E1CBirr (DHAxh‡ + DN)27 reaction. For the GH4 enzymes along with proton and hydride transfer to generate 2-acetamido-
deprotonation to form the E:enediolate intermediate is partially 2-deoxy-a-D-galactose as the reaction product. The hydration
rate limiting for kcat/Km (Scheme 1).
reaction likely limits kcat, and the E2 elimination is at least
Keeping in mind that GH4 enzymes require a divalent cation partially rate-limiting for kcat/Km. More detailed mechanistic
for activity, we suggest that their active site environment is such conclusions concerning the hydride transfer to NAD+ from C3
that the C-2 proton of the substrate is made more acidic as the will require the development of new synthetic procedures to
enediolate intermediate undergoes electrophilic stabilization make C1, C2 and C3 deuterated substrates.
by the metal cofactor. In contrast, the GH109 a-NAGAL active
References
1 K. W. Moremen, M. Tiemeyer and A. V. Nairn, Nat. Rev. Mol.
Cell Biol., 2012, 13, 448.
2 I. Brockhausen, EMBO Rep., 2006, 7, 599.
3 B. Weissman and D. F. Hinrichs, Biochemistry, 1969, 8, 2034.
4 S. I. Patenaude, N. O. L. Seto, S. N. Borisova, A. Szpacenko,
S. L. Marcus, M. M. Palcic and S. V. Evans, Nat. Struct.
Biol., 2002, 9, 685.
5 H. Wiegandt, Behav. Brain Res., 1995, 66, 85.
6 H. Tuppy and W. l. Staudenb, Biochemistry, 1966, 5, 1742.
7 M. L. Olsson, C. A. Hill, H. de la Vega, Q. Y. P. Liu,
M. R. Stroud, J. Valdinocci, S. Moon, H. Clausen and
M. S. Kruskall, Transfus. Clin. Biol., 2004, 11, 33.
8 Q. Y. P. Liu, G. Sulzenbacher, H. P. Yuan, E. P. Bennett,
G. Pietz, K. Saunders, J. Spence, E. Nudelman, S. B. Levery,
T. White, J. M. Neveu, W. S. Lane, Y. Bourne, M. L. Olsson,
B. Henrissat and H. Clausen, Nat. Biotechnol., 2007, 25, 454.
Fig. 4 Cartoon depiction of the active site of GH109 E. meningo-
septicum a-N-acetylgalactosaminidase (modified from ref. 8 and 23).
Shown is the most likely base for the abstraction of the C2-proton
(Tyr179), presumably activated by the proximal His181). Also, shown are
the amino acids residues that interact with the C2-acetamido group:
Tyr301 and Met375 via hydrogen-bonding and hydrophobic interac-
tions, respectively.
9 P. M. Coutinho, E. Deleury, G. J. Davies and B. Henrissat,
J. Mol. Biol., 2003, 328, 307.
10 G. J. Davies, T. M. Gloster and B. Henrissat, Curr. Opin.
Struct. Biol., 2005, 15, 637.
11 B. L. Cantarel, P. M. Coutinho, C. Rancurel, T. Bernard,
V. Lombard and B. Henrissat, Nucleic Acids Res., 2009, 37, D233.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 1188–1192 | 1191

View Article Online
MedChemComm
Concise Article
12 D. A. Comfort, K. S. Bobrov, D. R. Ivanen, K. A. Shabalin, 19 S. Chakladar, L. Cheng, M. Choi, J. Liu and A. J. Bennet,
J. M. Harris, A. A. Kulminskaya, H. Brumer and
R. M. Kelly, Biochemistry, 2007, 46, 3319.
13 S. C. Garman, L. Hannick, A. Zhu and D. N. Garboczi,
Structure, 2002, 10, 425.
14 M. Kiyohara, T. Nakatomi, S. Kurihara, S. Fushinobu,
H. Suzuki, T. Tanaka, S. Shoda, M. Kitaoka, T. Katayama,
Biochemistry, 2011, 50, 4298.
20 V. L. Y. Yip, J. Thompson and S. G. Withers, Biochemistry,
2007, 46, 9840.
21 V. L. Y. Yip, A. Varrot, G. J. Davies, S. S. Rajan, X. J. Yang,
J. Thompson, W. F. Anderson and S. G. Withers, J. Am.
Chem. Soc., 2004, 126, 8354.
K. Yamamoto and H. Ashida, J. Biol. Chem., 2012, 287, 693. 22 V. L. Y. Yip and S. G. Withers, Biochemistry, 2006, 45, 571.
15 V. Lombard, H. G. Ramulu, E. Drula, P. M. Coutinho and 23 G. Sulzenbacher, Q. P. Liu, E. P. Bennett, S. B. Levery,
B. Henrissat, Nucleic Acids Res., 2014, 42, D490.
Y. Bourne, G. Ponchel, H. Clausen and B. Henrissat,
16 G. Davies, M. L. Sinnott and S. G. Withers, in Comprehensive
Biocatal. Biotransform., 2010, 28, 22.
Biological Catalysis, ed. M. L. Sinnott, Academic Press, San 24 A. Varrot, V. L. Y. Yip, Y. S. Li, S. S. Rajan, X. J. Yang,
Diego, 1st edn, 1998, p. 119.
17 D. J. Vocadlo and G. J. Davies, Curr. Opin. Chem. Biol., 2008,
12, 539.
W. F. Anderson, J. Thompson, S. G. Withers and
G. J. Davies, J. Mol. Biol., 2005, 346, 423.
25 J. R. Knowles, CRC Crit. Rev. Biochem., 1976, 4, 165.
18 S. S. Rajan, X. J. Yang, F. Collart, V. L. Y. Yip, S. G. Withers, 26 T. H. Lowry and K. S. Richardson, Mechanism and theory in
A. Varrot, J. Thompson, G. J. Davies and W. F. Anderson,
organic chemistry, Harper & Row, New York, 1987.
Structure, 2004, 12, 1619.
27 R. Guthrie and W. P. Jencks, Acc. Chem. Res., 1989, 22, 343.
1192 | Med. Chem. Commun., 2014, 5, 1188–1192
This journal is © The Royal Society of Chemistry 2014