Mechanistic evaluation of MelA α-galactosidase from citrobacter freundii: A family 4 glycosyl hydrolase in which oxidation is rate-limiting

Article abstract of DOI:10.1021/bi101808h

The MelA gene from Citrobacter freundii, which encodes a glycosyl hydrolase family 4 (GH4) α-galactosidase, has been cloned and expressed in Escherichia coli. The recombinant enzyme catalyzes the hydrolysis of phenyl α-galactosides via a redox elimination-addition mechanism involving oxidation of the hydroxyl group at C-3 and elimination of phenol across the C-1-C-2 bond to give an enzyme-bound glycal intermediate. For optimal activity, the MelA enzyme requires two cofactors, NAD⁺ and Mn²⁺, and the addition of a reducing agent, such as mercaptoethanol. To delineate the mechanism of action for this GH4 enzyme, we measured leaving group effects, and the derived β_lg values on V and V/K are indistinguishable from zero (-0.01 ± 0.02 and 0.02 ± 0.04, respectively). Deuterium kinetic isotope effects (KIEs) were measured for the weakly activated substrate phenyl α-d-galactopyranoside in which isotopic substitution was incorporated at C-1, C-2, or C-3. KIEs of 1.06 ± 0.07, 0.91 ± 0.04, and 1.02 ± 0.06 were measured on V for the 1-²H, 2- ²H, and 3-²H isotopic substrates, respectively. The corresponding values on V/K were 1.13 ± 0.07, 1.74 ± 0.06, and 1.74 ± 0.05, respectively. To determine if the KIEs report on a single step or on a virtual transition state, we measured KIEs using doubly deuterated substrates. The measured ^DV/K KIEs for MelA-catalyzed hydrolysis of phenyl α-d-galactopyranoside on the dideuterated substrates, ^DV/K_{(3-D)/(2-D,3-D)} and ^DV/K _{(2-D)/(2-D,3-D)}, are 1.71 ± 0.12 and 1.71 ± 0.13, respectively. In addition, the corresponding values on V, ^DV _{(3-D)/(2-D,3-D)} and ^DV_{(2-D)/(2-D,3-D)}, are 0.91 ± 0.06 and 1.01 ± 0.06, respectively. These observations are consistent with oxidation at C-3, which occurs via the transfer of a hydride to the on-board NAD⁺, being concerted with proton removal at C-2 and the fact that this step is the first irreversible step for the MelA α-galactosidase-catalyzed reactions of aryl substrates. In addition, the rate-limiting step for V_max must come after this irreversible step in the reaction mechanism.

Full text of DOI:10.1021/bi101808h

ARTICLE
pubs.acs.org/biochemistry
Mechanistic Evaluation of MelA r-Galactosidase from Citrobacter
freundii: A Family 4 Glycosyl Hydrolase in Which Oxidation Is
Rate-Limiting
Saswati Chakladar,† Lydia Cheng,‡ Mary Choi,† James Liu,‡ and Andrew J. Bennet*
†Department of Chemistry and ‡Department of Molecular Biology and Biochemistry, Simon Fraser University, 8888 University Drive,
Burnaby, British Columbia V5A 1S6, Canada
S Supporting Information
b
ABSTRACT: The MelA gene from Citrobacter freundii, which encodes a glycosyl hydrolase
family 4 (GH4) R-galactosidase, has been cloned and expressed in Escherichia coli. The
recombinant enzyme catalyzes the hydrolysis of phenyl R-galactosides via a redox eliminationꢀ
addition mechanism involving oxidation of the hydroxyl group at C-3 and elimination of phenol
across the C-1ꢀC-2 bond to give an enzyme-bound glycal intermediate. For optimal activity, the
MelA enzyme requires two cofactors, NADþ and Mn2þ, and the addition ofa reducing agent, such
as mercaptoethanol. To delineate the mechanism of action for this GH4 enzyme, we measured
leaving group effects, and the derived βlg values on V and V/K are indistinguishable from zero
(ꢀ0.01 ( 0.02 and 0.02 ( 0.04, respectively). Deuterium kinetic isotope effects (KIEs) were measured for the weakly activated
substrate phenylR-D-galactopyranoside inwhichisotopicsubstitutionwas incorporatedatC-1, C-2, or C-3. KIEsof1.06( 0.07, 0.91 (
0.04, and 1.02 ( 0.06 were measured on V for the 1-2H, 2-2H, and 3-2H isotopic substrates, respectively. The corresponding values on
V/K were 1.13 ( 0.07, 1.74 ( 0.06, and 1.74 ( 0.05, respectively. To determine if the KIEs report on a single step or on a virtual
transition state, we measured KIEs using doubly deuterated substrates. The measured DV/K KIEs for MelA-catalyzed hydrolysis of
phenyl R-D-galactopyranoside on the dideuterated substrates, DV/K(3-D)/(2-D,3-D) and DV/K(2-D)/(2-D,3-D), are 1.71 ( 0.12 and 1.71 (
0.13, respectively. In addition, the corresponding values on V, DV(3-D)/(2-D,3-D) and DV(2-D)/(2-D,3-D), are 0.91 ( 0.06 and 1.01 ( 0.06,
respectively. These observations are consistent with oxidation at C-3, which occurs via the transfer of a hydride to the on-board
NADþ, being concerted with proton removal at C-2 and the fact that this step is the first irreversible step for the MelA
R-galactosidase-catalyzed reactions of aryl substrates. In addition, the rate-limiting step for Vmax must come after this irreversible
step in the reaction mechanism.
lycosidases (glycosyl hydrolases) make up an important
class of carbohydrate-processing enzymes, which number
a reducing agent such as 2-mercaptoethanol or dithiothreitol for
maximal catalytic activity. Recently, another GH family (GH109)
has been shown to require NADþ for activity.5 In addition to the
requirement for NADþ, the catalytic activity of family 4 GHs is
also dependent on the presence of a divalent metal, typically
Mn2þ, which in the single-crystal X-ray diffraction-based struc-
ture is shown to be chelated to O-2 and O-3 of the sugar ring.6 A
detailed structural analysis of GlvA 6-phospho-R-glucosidase
showed that the Mn2þ ion promotes association of the protein
to give the active tetramer.7 Notably, different members of GH4
possess catalytic activity against R- and β-glycosides, a character-
istic not seen in members of most glycosyl hydrolase families that
directly promote nucleophilic reactions at the anomeric center.
Members of GH4 can hydrolyze both phosphorylated and
nonphosphorylated disaccharide substrates, suggesting the
possible involvement of the phosphoenolpyruvate-dependent
sugar phosphotransferase system (PEP-PTS) in the species
from which the gene has been isolated.8 GH4 members also
G
more than 110 families that are classified on the basis of their
protein sequence.1 Glycosyl hydrolases (GHs) hydrolyze their
respective substrates with either retention or inversion of
anomeric configuration, by using a variety of catalytic strategies.2
Most GH families use a combination of general-acid catalysis,
which assists departure of the aglycone, and either nucleophilic
catalysis (retaining families) or general-base catalysis of water
attack at the anomeric center (inverting families). In contrast,
family 4 (GH4) members, which are solely derived from bacterial
sources, contain a characteristic glycine rich sequence that, in
combination with the topological arrangement of the secondary
structure, is consistent with a Rossmann fold, a structural motif
that is primarily involved in the binding of NAD(H).3 The crystal
structure of R-glucosidase A, AglA, from Thermotoga maritima in
a complex with NADþ and maltose has been determined to a
resolution of 1.9 Å, where Cys-174 and neighboring histidines
have been shown to be key catalytic residues.4 Importantly, this
research group suggested, on the basis of electron density maps,
that Cys-174 is easily oxidized to a sulfinic acid. This observa-
tion is consistent with the noted requirement for the presence of
Received: November 10, 2010
Revised:
April 14, 2011
Published: April 15, 2011
r
2011 American Chemical Society
4298
dx.doi.org/10.1021/bi101808h Biochemistry 2011, 50, 4298–4308
|

Biochemistry
ARTICLE
Scheme 1. Current Proposal for the Mechanism of GH4 Enzymes
are approximately 15% similar in sequence to lactate/malate
dehydrogenases.
Milli-Q water (18.2 MΩ cmꢀ1) was used for all kinetic experi-
ments. All pH values were measured using a standard pH
electrode attached to a VWR pH meter. All NMR spectra were
recorded on either a Bruker 400, 500, or 600 MHz spectrometer.
Chemical shifts are reported in parts per million downfield from
signals for TMS. The signal residues from deuterated chloroform
and external TMS salts (D2O) were used for 1H NMR spectral
references; for 13C NMR spectra, natural abundance signals from
CDCl3 and external TMS salts (D2O) were used as references.
Coupling constants (J) are reported in hertz. Melting points were
determined on a Gallenkamp melting point apparatus and are
not corrected. Optical rotations were measured on a Perkin-
Elmer 341 polarimeter and are reported in units of degrees
square centimeter per gram (concentrations reported in units of
grams per 100 mL). All fitting of kinetic data was performed
using the appropriate nonlinear least-squares equation in Graph-
Pad Prism (version 4.0).
Synthesis of Aryl r-D-Galactosides. All aryl R-D-galactopyrano-
sides, except the 3-nitrophenyl derivative, were synthesized from
1,2,3,4,6-penta-O-acetyl-D-galactopyranose using stannic chloride as
the activator in CH2Cl2. Typically, 1,2,3,4,6-penta-O-acetyl-β-D-
galactose (1.6 g, 4.0 mmol) and the appropriate phenol (1.5 g,
8 mmol) were dissolved in anhydrous CH2Cl2 (50 mL), and then
SnCl4 (0.8 mL, 8 mmol) was added. The reaction mixture was
stirred at ambient temperature under an inert atmosphere for 48 h,
at which time TLC analysis showed that the reaction was complete.
Following the addition of water (35 mL), the reaction mixture
was neutralized via addition of saturated NaHCO3 (20 mL). The
product was extracted from the aqueous layer using CH2Cl2
(3 ꢁ 35 mL), and the combined organic layer was washed with
brine (2 ꢁ 20 mL), dried over Na2SO4, and then concentrated
under reduced pressure to yield the crude product. This material was
purified by column chromatography using an EtOAc/hexane
mixture (1:4) as the eluent to yield the pure R-anomer. Deprotec-
tion was accomplished under Zemplen conditions using catalytic
As a result of these unusual features that are present in family 4
glycosyl hydrolases, in-depth kinetic studies have been per-
formed to delineate the prototypical mechanism of action for
this GH family. Yip et al.6,9,10 reported such mechanistic studies
on two GH4 family members, BglT from Thermotoga maritima
and GlvA from Bacillus subtilis. These authors showed that GH4
members catalyze glycoside hydrolysis via an NADþ-dependent
redox reaction that is coupled to an R,β-elimination reaction
involving the formation of an R,β-unsaturated ketone (glycal)
intermediate.10 Moreover, on the basis of kinetic isotope effect
(KIE) measurements, they concluded that oxidation at C-3 and
the subsequent C-2 deprotonation steps are both partially rate-
limiting.6,9 The proposed mechanistic outline for GH4 enzymes
is depicted in Scheme 1 for an R-galactosidase.
This study details a mechanistic investigation of a GH4
R-galactosidase that is encoded by the MelA gene from the
species Citrobacter freundii. This gene encodes an enzyme that
hydrolyzes the terminal R-galactosyl units from glycolipids and
glycoproteins. Melibiose [R-D-Galp-(1f6)-D-Glcp] utilization
in Escherichia coli is dependent on the melibiose locus (Mel),
which forms an operon consisting of at least two structural genes,
MelA and MelB, which encode an R-galactosidase and a melibiose
carrier protein, respectively.11 Results from identification and
purification studies with MelA have shown it to be a tetrameric
protein with a molecular mass of 200 kDa for the active
enzyme.11 In addition, Anggraeni et al.12 showed that rMel4A
R-galactosidase exhibited the highest kcat values on the natural
substrate melibiose but exhibited the highest affinity for raffinose.
’ MATERIALS AND METHODS
Materials. All chemicals were of analytical grade or better and
were purchased from Sigma-Aldrich unless noted otherwise.
4299
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
NaOMe in MeOH followed by neutralization of the solution with
Amberlite (Hþ) resin. Finally, product purification was achieved by
crystallization from ethanol. The 3-nitrophenyl R-D-galactopyrano-
purification (25% EtOAc/hexane) gave 340 mg of the pure
R-anomer 3a and 600 mg of an anomeric mixture: [R]20D = 165.8
(c= 0.32, CHCl3);1H NMR (400 MHz, CDCl3) δ7.34ꢀ7.27 (m, 2
H, Ar), 7.09ꢀ7.03 (m, 3 H, Ar), 5.58 (dd, 1 H, J = 10.8, 3.4, H-3),
5.53 (dd, 1 H, J = 3.4, 1.2, H-4), 5.28 (d, 1 H, J = 10.8, H-2),
4.38ꢀ4.33 (t, 1 H, H-5), 4.16ꢀ4.03 (m, 2 H, H-6, H-60), 2.17,
2.07, 2.03, 1.94 (4s, 12 H, 4 ꢁ CH3); 13C NMR (151 MHz,
solvent) δ 170.43, 170.33, 170.20, 170.06 (4 ꢁ CdO), 156.26
(Ar), 129.62 (Ar), 122.98 (Ar), 116.75 (Ar), 67.89 (C-4), 67.73
(C-2), 67.54 (C-3), 67.10 (C-5), 61.47 (C-6), 20.74, 20.69,
20.64, 20.57 (4 ꢁ CH3); ESI-MS for C20H23DO10 m/z calcd
for (M þ Naþ) 448.1330, found 448.1336.
side was synthesized using BF3 OEt as the activator by following the
3
reported reaction conditions.13 The detailed characterizations of
phenyl R-D-galactoside (1a) and 4-nitrophenyl R-D-galactoside (2a)
have been previously reported.14 The characterizations of all other
unlabeled substrates used in this study are given below.
3-Nitrophenyl R-D-galactoside (2b): mp 165ꢀ168 °C;
[R]20D = 195.7 (c = 0.15, H2O); 1H NMR (600 MHz, D2O) δ
8.04 (s, 1 H, ArH), 8.00ꢀ7.97 (t, 1 H, ArH), 7.61ꢀ7.56 (m, 2 H,
ArH), 5.79 (d, J = 3.8, 1 H, H-1), 4.12ꢀ4.01 (m, 4 H, H-2, H-3,
H-4, H-5), 3.71 (app dd, J = 8.8, 6.2, 2 H, H-6, H-60); 13C NMR
(151 MHz, D2O) δ 156.42, 130.44, 124.03, 117.80, 112.04,
97.36, 72.02, 69.34, 69.05, 67.91, 60.95; ESI-MS for C12H15NO8
m/z calcd for (M þ Naþ) 324.0695, found 324.0682.
Phenyl r-D-[1-2H]Galactopyranoside (1b). Deprotection of
3b under Zemplen conditions, using catalytic NaOMe in MeOH,
was followed by neutralization using Amberlite (Hþ) resin. The
deprotected compound was further recrystallized from ethanol
to yield 110 mg of 1b (yield, 60%): mp 132ꢀ135 °C; [R]20
=
3,4-Dichlorophenyl R-D-galactoside (2c): mp 163ꢀ166 °C;
1H NMR (400 MHz, D2O) δ 7.50 (d, J = 8.9, 1 H, ArH), 7.40 (d,
J = 2.8, 1 H, ArH), 7.11 (dd, J = 8.9, 2.8, 1 H, ArH), 5.66 (d,
J = 3.7, 1 H, H-1), 4.10ꢀ3.96 (m, 4 H, H-2, H-3, H-4, H-5),
3.74ꢀ3.69 (m, 2 H, H-6, H-60); 13C NMR (126 MHz, D2O) δ
155.32, 130.81, 125.37, 119.03, 117.12, 97.52, 71.83, 69.33,
69.04, 67.85, 60.91; ESI-MS for C12H14Cl2O6 m/z calcd for
(M þ Naþ) 347.0065, found 347.0054.
D
185.6 (c = 0.0.56, H2O); 1H NMR (400 MHz, D2O) δ
7.47ꢀ7.36 (m, 2 H, Ar), 7.25ꢀ7.11 (m, 3 H, Ar), 4.16ꢀ4.05
(m, 3 H, H-3, H-4, H-5), 4.00 (d, 1 H, J = 10.1, H-2), 3.71 (app d,
2 H, J = 6.2, H-6, H-60); 13C NMR (151 MHz, D2O) δ 129.82
(Ar), 123.07 (Ar), 117.34 (Ar), 71.60, 69.41, 69.12, 67.95, 60.95;
ESI-MS for C12H15DO6 m/z calcd for (M þ Naþ) 280.0907,
found 280.0904.
Phenyl2,3,4,6-Tetra-O-acetyl-r-D-[2-2H]galactopyranoside
(3c). Dimeric 3,4,6-tri-O-acetyl-2-deoxy-2-nitroso-R-D-galactopyr-
anosyl chloride 5, made from galactal 4,15 was dissolved in
anhydrous DMF (100 mL), and to this solution was added phenol
(2 g, 2.0 equiv). The resultant mixture was stirred for 70 h at
ambient temperature. At that time, TLC analysis (40% EtOAc/
hexane) indicated the complete disappearance of the starting
material. The reaction mixture was then diluted with Et2O
(50 mL) and extracted with water (3 ꢁ 40 mL), washed with
brine (10 mL), dried (Na2SO4), and concentrated under reduced
pressure to yield a yellow syrup. Purification by flash chromatog-
raphy (35% EtOAc/hexane) gave phenyl 3,4,6-tri-O-acetyl-2-
deoxy-2-oximino-R-D-lyxo-hexopyranoside 6 (1.3 g, 30% over
two steps). The glycosylated product 6 (1.3 g) was dissolved in
CH3CN (15 mL) that contained 1 N HCl (0.5 mL) and
CH3CHO (0.2 mL). The reaction mixture was stirred at ambient
temperature for 5 h, at which time the reaction mixture was
extracted with EtOAc (3 ꢁ 25 mL). The combined organic layer
was then dried (Na2SO4), and volatiles were removed under
reduced pressure to give a pale yellow oil. The crude product was
dissolved in anhydrous THF (30 mL); the solution was cooled to
0 °C, and a solution of NaBD4 (100 mg) in cold D2O (2 mL) was
added. The mixture was then stirred at 0 °C for 1 h and then at
room temperature for 2 h. Cautious addition of glacial AcOH
(5 mL) destroyed the excess borodeuteride, and the resultant
solution was washed with MeOH (20 mL). Removal of the
volatiles under reduced pressure gave a pale yellow syrup that
was directlyacetylated under standard conditions, pyridine (5 mL)
and Ac2O (4 mL) at 25 °C. After the addition of water (30 mL),
the product was extracted into CH2Cl2 (3 ꢁ 30 mL), and the
organic layerwas washed with cold10% H2SO4 (10mL) and brine
(2 ꢁ 10 mL) and dried (Na2SO4). Removal of the volatiles under
reduced pressure gave a pale yellow syrup that was purified by flash
chromatography to give pure phenyl 2,3,4,6-tetra-O-acetyl-R-
D-[2-2H]galactopyranoside 3c (350 mg, 30% over three steps):
[R]20D = 168.3 (c = 0.247, CHCl3); 1H NMR (400 MHz, CDCl3)
δ 7.38ꢀ7.25 (m, 2 H, ArH), 7.14ꢀ7.03 (m, 3 H, ArH), 5.78 (s,
1 H, H-1), 5.58 (d, 1 H, J = 3.4, H-3), 5.53 (d, 1 H, J = 3.3, H-4),
3-Chlorophenyl R-D-galactoside (2d): mp 139ꢀ142 °C;
1
[R]20D = 195.7 (c = 0.499, H2O); H NMR (400 MHz, D2O)
δ 7.34 (t, J = 8.2, 1 H, ArH), 7.25 (t, J = 2.1, 1 H, ArH), 7.18ꢀ7.07
(m, 2 H, ArH), 5.67 (d, J = 3.9, 1 H, H-1), 4.11ꢀ3.93 (m, 4 H,
H-2, H-3, H-4, H-5), 3.74ꢀ3.65 (m, 2 H, H-6, H-60); 13C NMR
(126 MHz, D2O) δ 156.78, 134.25, 130.69, 122.91, 117.47,
115.62, 97.28, 71.79, 69.35, 69.05, 67.95, 60.93; ESI-MS for
C12H15ClO6 m/z calcd for (M þ Naþ) 313.0455, found
313.0441.
4-Chlorophenyl R-D-galactoside (2e): mp 157ꢀ161 °C;
1
[R]20D = 193.5 (c = 0.452, H2O); H NMR (600 MHz, D2O)
δ 7.37 (d, J = 9.0, 2 H, ArH), 7.15 (d, J = 9.0, 2 H, ArH), 5.64 (d,
J = 3.8, 1 H, H-1), 4.05ꢀ3.99 (m, 4 H, H-2, H-3, H-4, H-5), 3.70
(app d, J = 5.9, 2 H, H-6, H-60), 3.34 (d, J = 1.0, OH); 13C NMR
(151 MHz, D2O) δ 154.79, 129.45, 127.27, 118.72, 97.46 (C-1),
71.72, 69.36, 69.08, 67.98, 60.95; ESI-MS for C12H15ClO6 m/z
calcd for (M þ Naþ) 313.0455, found 313.0445.
4-Methylphenyl R-D-galactoside (2f): mp 160ꢀ163 °C;
1
[R]20D = 174.3 (c = 0.562, H2O); H NMR (500 MHz, D2O)
δ 7.22 (d, J = 8.4, 2 H, ArH), 7.08 (d, J = 8.5, 2 H, ArH), 5.60 (d,
J = 3.9, 1 H, H-1), 4.13ꢀ4.03 (m, 3 H, H-3, H-4, H-5), 3.98 (dd, 1
H, H-2), 3.70 (app d, J = 5.7, 2 H, H-6, H-60), δ 2.29 (s, 3 H,
ArCH3); 13C NMR (126 MHz, D2O) δ 153.92, 133.03, 130.13,
117.44, 97.65, 71.55, 69.40, 69.11, 68.05, 60.94, 19.55; ESI-MS
for C13H18O6 m/z calcd for (M þ Naþ) 293.1001, found
293.0990.
Phenyl2,3,4,6-Tetra-O-acetyl-r-D-[1-2H]galactopyranoside
(3b). 1,2,3,4,6-Penta-O-acetyl-D-[1-2H]galactose (Supporting In-
formation) (1.6 g, 4.0 mmol) and phenol (1.5 g, 8 mmol) were
dissolved in anhydrous CH2Cl2 (50 mL) to which SnCl4 (0.8 mL,
8 mmol) was added. This solution was stirred at room tempera-
ture for 48 h, at which time TLC analysis (40% EtOAc/hexane)
showed that the reaction was complete. Following the addition
of water (35 mL) and saturated NaHCO3 (20 mL), the product
was extracted from the aqueous layer using CH2Cl2 (3 ꢁ 35 mL).
The combined organic layer was washed with brine and dried
(Na2SO4). Removal of the volatiles under reduced pressure gave
the crude product in quantitative yield. Flash chromatographic
4300
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
4.36 (t, 1 H, J = 7.1, H-5), 4.09 (m, 1 H, H-6), 4.15 (m, 1 H, H-60),
2.17, 2.08, 2.04, 1.94 (4s, 12 H, 4 ꢁ CH3); 13C NMR (151 MHz,
CDCl3) δ 170.42, 170.33, 170.21, 170.06, 156.26, 129.62 (Ar),
122.99 (Ar), 116.75 (Ar), 94.78 (C-1), 67.90 (C-4), 67.48 (C-3),
67.12 (C-5), 61.47 (C-6), 20.74, 20.68, 20.64, 20.56 (4 ꢁ CH3,
COCH3); ESI-MS for C20H23DO10 m/z calcd for (M þ Naþ)
488.1330, found 448.1321.
three steps): [R]20D = 167.0 (c = 0.15, CHCl3); 1H NMR (400
MHz, CDCl3) δ 7.34ꢀ7.28 (m, 2 H, Ar), 7.09ꢀ7.03 (m, 3 H, Ar),
5.78 (d, 1 H, J = 3.6, H-1), 5.53 (d, 1 H, J = 1.1, H-4), 5.29 (d, 1 H,
J = 3.7, H-2), 4.36 (t, 1 H, J = 6.6, H-5), 4.12 (dd, 1 H, J = 11.3, 6.2,
H-6), 4.06 (dd, 1 H, J = 11.3, 7.1, H-60), 2.17, 2.07, 2.03, 1.94 (4 ꢁ
CH3); 13C NMR (151 MHz, CDCl3) δ 170.41, 170.32, 170.20,
170.05 (4 ꢁ CdO, 4 ꢁ OCOCH3), 156.28 (ArC), 129.62 (ArC),
122.99 (ArCH), 116.77 (ArCH), 94.84 (C-1), 67.86 (C-4), 67.74
(C-2), 67.13 (C-5), 61.47 (C-6), 20.73, 20.67, 20.63, 20.55 (4 ꢁ
CH3, 4 ꢁ OCOCH3); ESI-MS for C20H23DO10 m/z calcd for
(M þ Naþ) 488.1330, found 448.1330.
Complete deprotection of the peracetylated compound was
accomplished under Zemplen conditions using a NaOMe/
MeOH mixture followed by neutralization using Amberlite
(Hþ) resin, to yield 175 mg (90%) of pure phenyl R-
D-[2-2H]galactopyranoside 1c: mp 133ꢀ135 °C; [R]20
=
Complete deprotection of the peracetylated compound was
accomplished under Zemplen conditions using a NaOMe/
MeOH mixture followed by neutralization using Amberlite
(Hþ) resin, to yield pure phenyl R-D-[3-2H]galactopyranoside
1d in quantitative yield: mp 137ꢀ140 °C; [R]20D = 168.3 (c =
0.25, H2O); 1H NMR (600 MHz, D2O) δ 7.46ꢀ7.38 (m, 2 H,
Ar), 7.20 (d, J = 8.2, 2 H, Ar), 7.15 (t, J = 7.4, 1 H, Ar), 5.68 (d,
J = 3.9, 1 H, H-1), 4.09 (t, J = 6.2, 1 H, H-5), 4.06 (s, 1 H, H-4),
3.99 (d, J = 3.9, 1 H, H-2), 3.70 (app d, J = 6.2, 2 H, H-6, H-60);
13C NMR (151 MHz, D2O) δ 129.82 (Ar), 123.07 (Ar), 117.35
(Ar), 97.29 (C-1), 71.63 (C-5), 69.07 (C-4), 67.99 (C-2), 60.95
(C-6); ESI-MS for C12H15DO6 m/z calcd for (M þ Naþ)
280.0907, found 280.0900.
D
1
196.0; H NMR (600 MHz, D2O) δ 7.41 (dd, J = 8.3, 7.5, 2
H, ArH), 7.20 (d, J = 7.9, 2 H, ArH), 7.15 (s, 1 H, ArH), 5.68 (s, 1
H, H-1), 4.06ꢀ4.09 (m, 2 H, H-3, H-4), 4.06 (d, J = 3.3, 1 H,
H-5), 3.70 (app d, J = 6.2, 2 H, H-6, H-60); 13C NMR (151 MHz,
D2O) δ 129.82 (Ar), 123.08 (Ar), 117.35 (Ar), 97.26 (C-1),
71.63 (C-5), 69.35 (C-3), 69.12 (C-4), 60.95 (C-6); ESI-MS
for C12H15DO6 m/z calcd for (M þ Naþ) 280.0907, found
280.0905.
Dimeric 3,4,6-Tri-O-acetyl-2-deoxy-2-nitroso-r-D-[3-2H]-
galactopyranosyl Chloride (8). NOCl gas, which was prepared
by dropwise addition of a saturated solution of sodium nitrite
(13 g) in water (20 mL) to concentrated HCl (50 mL), was
bubbled through a cooled solution (ꢀ45 °C) of thoroughly
dried deuterated galactal 7 (Supporting Information) in EtOAc
(40 mL). After the addition was complete, the sample was heated
to 0 °C over a period of 20 min, and the volatiles were then
removed under reduced pressure to give a bluish-green mass,
which was further dried under vacuum for 2 h. The resultant
crude product was used in the subsequent reaction without
further purification.
Phenyl 2,3,4,6-Tetra-O-acetyl-r-D-[2-2H,3-2H]galactopy-
ranoside (3e). A solution of the ketone made from oxime 9
(285 mg, made as described above) in anhydrous THF (30 mL)
was cooled to 0 °C, and then a solution of NaBD4 (100 mg) in
cold D2O (2 mL) was added dropwise. After the reaction mixture
had been stirred at 0 °C for 1 h and then at room temperature for
2 h, the cautious addition of glacial AcOH (5 mL) destroyed the
excess borodeuteride. The resultant solution was washed with
MeOH (20 mL), and removal of the volatiles under reduced
pressure gave a pale yellow syrup, which was directly acetylated
under standard conditions, pyridine (5 mL) and Ac2O (5 mL) at
25 °C. The crude product was purified by flash chromatography
(25% EtOAc/hexane) to yield pure peracetylated 2,3-dideu-
teriated compound 3e (100 mg, 30% over two steps): 1H NMR
(400 MHz, CDCl3) δ 7.38ꢀ7.24 (m, 2 H, Ar), 7.13ꢀ7.02 (m, 3
H, Ar), 5.78 (s, 1 H, H-1), 5.53 (d, 1 H, J = 1.0, H-4), 4.36 (t, 1 H,
J = 6.5, H-5), 4.18ꢀ3.99 (m, 2 H, H-6, H-60), 2.17, 2.05, 2.03, 1.94
(4 ꢁ CH3); 13C NMR (151 MHz, CDCl3) δ 170.42, 170.33,
170.21, 170.06, 156.26, 129.62 (ArC), 122.99 (ArC) 116.75
(ArC), 94.78 (C-1), 67.86 (C-4), 67.12 (C-5), 61.47 (C-6),
20.74, 20.68, 20.64, 20.57 (4 ꢁ CH3, OCOCH3); ESI-MS for
C20H22D2O10 m/z calcd for (M þ Naþ) 449.1393, found
449.1371.
Phenyl2,3,4,6-Tetra-O-acetyl-r-D-[3-2H]galactopyranoside
(3d). The crude dimeric 3,4,6-tri-O-acetyl-2-deoxy-2-nitroso-R-
D-[3-2H]galactopyranosyl chloride 8 was dissolved in anhydrous
DMF (50 mL), and to this solution was added phenol (800 mg,
2.0 equiv). Stirring was continued for 70 h at ambient tempera-
ture until TLC analysis (40% EtOAc/hexane) showed the pre-
sence of the desired product (Rf = 0.4). Subsequently, the reaction
mixture was diluted with Et2O (50 mL) and extracted with water
(3 ꢁ 30 mL). The combined ethereal layers were pooled together,
washed with brine (20 mL), dried over Na2SO4, and concentrated
under vacuum to give a yellowish syrup that was purified by flash
chromatography (35% EtOAc/hexane) to yield 330 mg (30%
over two steps) of phenyl 3,4,6-tri-O-acetyl-2-deoxy-2-oximino-R-
D-[3-2H]-lyxo-hexopyranoside 9. After 9 (330 mg) had been
dissolved in CH3CN (5 mL) containing 1 N HCl (0.5 mL) and
CH3CHO (0.3 mL), the mixture was stirred at ambient tempera-
ture for 5 h, when water (20 mL) was added. The aqueous layer
was extracted with EtOAc (3 ꢁ 25 mL); the combined organic
layer was dried (Na2SO4), and volatiles were then removed under
reduced pressure to give a pale yellow oil. The resultant material
was dissolved in anhydrous THF (30 mL), and the solution was
cooled to 0 °C. Addition of a solution of NaBH4 (100 mg) in cold
H2O (2 mL) was followed by the reaction being stirred at 0 °C for
1 h and then at room temperature for 2 h. Cautious addition of
glacial AcOH (5 mL) destroyed the excess borohydride, and the
resultant solution was washed with MeOH (20 mL). Removal of
the volatiles under reduced pressure gave a pale yellow syrup that
was directly acetylated under standard conditions, with stirring
in pyridine (5 mL) and Ac2O (4 mL) for 15 h at 25 °C. Flash
chromatographic purification gave pure 3d (80 mg, 40% yield over
Final deprotection of the compound was accomplished under
Zemplen conditions using a NaOMe/MeOH mixture followed
by neutralization using Amberlite (Hþ) resin and gave 40 mg
(80%) of pure phenyl R-D-[2-2H,3-2H]galactopyranoside 1e: mp
1
139ꢀ140 °C; [R]20 = 182.9; H NMR (600 MHz, D2O) δ
D
7.44ꢀ7.38 (m, 2 H, Ar), 7.20 (d, 2 H, J = 8.5, Ar), 7.18ꢀ7.11 (m,
1 H, Ar), 5.68 (s, 1 H, H-1), 4.08 (t, 1 H, J = 6.2, H-5), 4.05 (s, 1
H, H-4), 3.70 (d, 2 H, J = 6.3, H-6, H-60); 13C NMR (151 MHz,
D2O) δ 156.09 (Ar), 129.82 (Ar), 123.07 (Ar), 117.34 (Ar),
97.26 (C-1), 71.63 (C-5), 69.07 (C-4), 60.94 (C-6); ESI-MS for
C12H14D2O6 m/z calcd for (M þ Naþ) 281.0970, found
281.0966.
Cloning of r-Galactosidase. The MelA gene encoding
R-galactosidase was PCR amplified from C. freundii genomic
DNA (ATCC 8090D). The amplification was performed in 5%
4301
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
DMSO using the following primers that introduced NheI and
XhoI restriction sites (underlined) into the forward and reverse
primers, respectively: 50-CGCGGCTAGCATGATGTCTG-
CACCC-30 (forward) and 50-GCGCCTCGAGTTAACGGTG-
CAGCCAG-30 (reverse). The PCR fragment was purified,
digested with NheI and XhoI, inserted into the correspondingly
digested pET28a vector (Novagen), and transformed in E. coli
BL-21(DE3). The plasmid DNA was isolated from a single
colony and verified by restriction digestion and DNA sequencing
by Macrogen using T7 promoter and T7 terminator primers.
MelA Expression and Purification. For expression of the
MelA gene, cells were inoculated in Luria broth and then
expressed in YT medium supplemented with 1% kanamycin at
37 °C to an OD600 of 0.5 before being induced with 0.5 mM
IPTG. The cultures were incubated at 25 °C for 5 h and then
centrifuged (10 min at 8400g), and the pellet from 1 L of culture
was resuspended in binding buffer containing 5 mM imidazole.
The cells were lysed open using 1% lysozyme (from chicken egg
white) and a protease inhibitor cocktail tablet followed by
sonication (20 s onꢀ40 s off cycle at 60% capacity) to ensure
complete lysis of the cells. The lysate was centrifuged at 4 °C
(34000g) to remove the cell debris and intact cells. The clear
supernatant was collected and filtered through a 0.45 μm filter
before being loaded on to a HiTrap Chelating HP column
(5 mL) that had been pre-equilibrated with binding buffer.
The column was washed sequentially with 60, 100, and
150 mM imidazole before the protein was eluted with 250 mM
imidazole. Fractions containing pure His-MelA, as determined by
10% sodium dodecyl sulfateꢀpolyacrylamide gel electrophor-
esis, were pooled together and dialyzed, at 4 °C, three times
against 4 L of Tris buffer (20 mM, pH 7.0) that contained NaCl
(100 mM) and DTT (6 mM). The protein was then concen-
trated by centrifugation through a 10 kDa filter, and its concen-
tration was determined (Bradford Assay or Nanodrop).
Enzyme Kinetics. All kinetic assays were conducted in either
0.2 or 1.0 cm path length quartz cuvettes, unless stated otherwise,
using a Cary 300 UVꢀvis spectrometer equipped with a Cary
temperature controller. For all experiments, the MelA R-galac-
tosidase was preincubated in assay buffer with cofactors at 37 °C
for 15 min prior to the addition of the substrate, which initiated
the enzymatic reaction. The following buffer (buffer A) at pH 7.5
was used for most kinetic experiments: HEPES (50 mM), MnCl2
(1.0 mM), NADþ (100 μM), 2-mercaptoethanol (10 mM), and
BSA (0.1%, w/v). All cofactor solutions were freshly prepared
before each set of kinetic assays. The measurements of all
deuterium kinetic isotope effects were performed in buffer B,
which was identical to buffer A except that the added reducing
agent was TCEP (1 mM) instead of 2-mercaptoethanol (10 mM).
Typical Conditions for the Measurement of Michae-
lisꢀMenten Parameters. The concentration of R-galactosidase
was chosen such that less than 10% of the total substrate was
consumed during the assay. For each assay, the enzyme was
incubated in the appropriate buffer at 37 °C for either 15 or
20 min. Afterward, the reaction was initiated by the addition of
substrate. The initial rate of hydrolysis was followed spectro-
photometrically at the wavelength of the maximal absorbance
change. Typically, the substrate concentration was varied be-
tween 40 μM and 1 mM, and the measured initial rate versus
concentration data were fit to a standard MichaelisꢀMenten
equation.
over a pH range of 4.0ꢀ9.5. The following buffers were used:
NaOAc-HOAc (20 mM, pH 4.0ꢀ4.5), MES (20 mM, MES-
NaOH, pH 6.0ꢀ6.7), HEPES (20 mM, pH 6.5ꢀ8.2), and CHES
(20 mM, pH 8.5ꢀ9.5). Typical assay conditions were as follows. R-
Galactosidase (final concentration of 4.1 μg/mL) was incubated at
37 °C with the appropriate buffer containing NaCl (50 mM),
MnCl2 (1.0 mM), NADþ (100 μM), 2-mercaptoethanol (10 mM),
and BSA (0.1%, w/v) for 10 min prior to the addition of substrate
4NPRG; the hydrolysis reaction was monitored at either 400 nm
(pH 6.5ꢀ9.0) or 340 nm (pH 4.0ꢀ6.0). The difference in the
extinction coefficients (Δε) for 4NPRG and the released pheno-
late/phenol was determined at each pH, and the initial rate
measurements were fit to a standard MichaelisꢀMenten equation.
Kinetic Investigation of NADþ Cofactor Dependence. A
sample of MelA (final concentration of 16.3 μg/mL) in buffer A
was preincubated for 15 min at 37 °C for 15 min. 4NPRG was
added to the solution to initiate the hydrolysis reaction. In
another parallel assay, the enzyme was incubated with all other
cofactors at pH 7.5 in the presence of 10 mM NaBH4, for the
quantitative reduction of NADþ to NADH. The enzyme was
completely inactive after treatment with NaBH4, but approxi-
mately 60% of its activity was restored upon addition of exogenous
NADþ (200 μM).
Determination of Kd Values for NADþ. Solutions containing
MelA (final concentration of 15.8 μg/mL) and NADþ (con-
centration varied from 20 μM to 4 mM) were preincubated in
HEPES buffer (50 mM, pH 7.5) containing MnCl2 (1 mM),
2-mercaptoethanol (10 mM), and BSA (0.1%, w/v) at 37 °C for
15 min, and then 4NPRG (final concentration of 100 μM) was
added to initiate the reaction (final volume of 400 μL). The
measured initial reaction rates were fit to the MichaelisꢀMenten
equation that includes a term for substrate inhibition.
Determination of the Kd Value for Mn2þ. To remove bound
metal ions, MelA was dialyzed three times against 4 L of Tris
buffer (20 mM, pH 7.1). Solutions containing MelA (final
concentration of 15.0 μg/mL) and Mn2þ (concentration varied
from 0 to 400 μM) were preincubated in HEPES buffer (50 mM,
pH 7.5) containing NADþ (100 μM), 2-mercaptoethanol
(10 mM), and BSA (0.1%, w/v) at 37 °C for 15 min, and then
4NPRG (final concentration of 100 μM) was added to initiate
the reaction (final volume of 400 μL). In a separate experiment, a
solution of MelA (200 μL, 3 mg/mL) was dialyzed overnight
against Tris-HCl buffer (500 mL, 20 mM, pH 7.1) containing
EDTA (2 mM) and NaCl (100 mM). The resultant enzyme
solution was then dialyzed twice, for 4 h, against fresh Tris-HCl
buffer (500 mL, 20 mM, pH 7.1) containing NaCl (100 mM).
The enzyme activity was then monitored as a function of Mn2þ
concentration as detailed above. After subtraction of the rate in
the absence of Mn2þ, the resultant initial rates were fit to a
standard MichaelisꢀMenten equation.
Deuterium Kinetic Isotope Effect Measurements. To deter-
mine KIEs (kH/kD) onV, separate measurements of the Michaelisꢀ
Menten parameters were taken for each of the following: phenyl
R-D-galactopyranoside, phenyl R-D-[1-2H]galactopyranoside,
phenyl R-D-[2-2H]galactopyranoside, phenyl R-D-[3-2H]galacto-
pyranoside, and phenyl R-D-[2-2H,3-2H]galactopyranoside. Typi-
cally, for each assay, MelA R-galactosidase (final concentration of
3.1 μg/mL) was incubated in buffer B for 20 min prior to the
initiation of hydrolysis by the addition of substrate. The initial rates
weremeasured bymonitoringthe release ofphenol at 270 nm. The
substrate concentration was varied between 0.04 and 1.0 mM. The
cofactor solution was prepared freshly, and all the data sets were
Determination of the pHꢀRate Profile. To determine the
effect of pH on enzymatic activity, we measured kinetic parameters
4302
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
obtained using the same stock of cofactors and enzyme. Each set of
kinetic experiments was completed within a day to ensure the
smallest possible variation in enzyme activity. All kinetic data were
fit to a standard MichaelisꢀMenten equation.
Conditions for the Measurement of KIEs on V/K. Measure-
ments of individual V/K values were taken by monitoring the
first-order consumption of substrate, whose concentration was
typically e0.1Km, at the wavelength of the maximal absorbance
change (277 nm). Normally, MelA R-galactosidase (final con-
centration in the range of 40ꢀ50 μg/mL) was incubated in buffer
B for 20 min at 37 °C before addition of the appropriate substrate
that initiated the reaction. The absorbance was monitored at
277 nm until the hydrolysis was complete. The absorbance versus
time data were fit to a standard first-order rate equation; for
all cases, it was determined that the fitting residuals were random-
ly scattered around zero. Measurements for the protio and
deuterio substrates were performed in an alternating fashion,
and each measurement was repeated three times. The resulting
DV/K values were calculated using a standard weighted average
method.16
Figure 1. Ligand binding curve for NADþ using PNPRG as the
substrate.
1
Product Studies. H NMR spectroscopy (500 MHz) was
employed to identify the stereochemical course of the enzyme-
catalyzed reaction. The reaction conditions involved incubating,
at 37 °C, the enzyme (4.2 μg/mL) in buffer A containing
CD3OD (5 M). After the addition of 4NPRG (2.5 mg), the
reaction was allowed to proceed at 37 °C until TLC analysis [1:4
(v/v) MeOH:EtOAc] showed no remaining starting material.
Removal of the enzyme by centrifugal ultrafiltration at 4 °C was
followed by stirring of the resultant solution with Chelex resin
(Fluka) for 60 min. Following filtration, the solution was
lyophilized and the resulting solid was dissolved in D2O, and
then the 1H NMR spectrum was recorded.
Brønsted Analysis of the Linear Free Energy Relationship.
A series of substrates with varying leaving groups were synthe-
sized to perform a Brønsted analysis. Full MichaelisꢀMenten
curves were measured, in buffer A, for each substrate using the
protocol listed above.
Table 1. MichaelisꢀMenten Kinetic Parameters for the
MelA-Catalyzed Hydrolysis of 4NPrG as a Function of pH at
37 °C
pH
kcat (sꢀ1
)
10ꢀ3 ꢁ kcat/Km (Mꢀ1 sꢀ1
)
4.5
6.7
7.1
7.5
7.8
8.0
8.4
8.7
9.0
0.12 ( 0.05
0.30 ( 0.04
1.05 ( 0.05
1.68 ( 0.11
2.07 ( 0.21
1.75 ( 0.08
1.37 ( 0.03
0.62 ( 0.03
0.23 ( 0.03
0.091 ( 0.070
0.81 ( 0.26
2.7 ( 0.4
5.8 ( 1.1
9.9 ( 1.9
8.9 ( 1.4
15.6 ( 1.5
5.0 ( 1.2
0.96 ( 0.42
’ RESULTS
of Mn2þ (Kd) (Figure S2 of the Supporting Information). All
other kinetic parameters were evaluated using buffers that
contained NADþ and Mn2þ at concentrations of 100 μM and
1.0 mM, respectively.
To test whether C. freundii MelA galactosidase displays any
catalytic reactivity toward phosphorylated substrates, 4-nitro-
phenyl 6-phospho-R-D-galactoside was synthesized (Supporting
Information). However, it was shown that MelA hydrolyzed
neither 4-nitrophenyl 6-phospho-R-D-galactopyranoside nor
4-nitrophenyl β-D-galactopyranoside, while it was catalytically
active against 4-nitrophenyl R-D-galactopyranoside (data not
shown). The absolute requirement for the catalytic activity of
this GH4 R-galactosidase for NADþ was shown by a standard
borohydride reduction experiment (Figure S1 of the Supporting
Information). The binding constants for the two cofactors,
NADþ and Mn2þ, were determined by monitoring the hydro-
lysis of 4NPRG as a function of cofactor concentration at pH
7.5 and 37 °C. The dependence of enzyme activity on the
concentration of NADþ exhibited marked substrate inhibition
(Figure 1), and the associated kinetic parameters for the MelA
R-galactosidase-catalyzed reaction (Km and Kis) are 240 (
Product Studies. In the absence of a reducing agent, MelA
gave no product even after incubation for 48 h, and when the
reaction was performed with mercaptoethanol in the presence of
methanol (5 M), no methyl galactoside was produced. However,
an additional new anomeric resonance, in addition to galactose,
appeared in the 1H NMR spectrum at a chemical shift of δ 5.8
(J1,2 = 5.5 Hz) that is tentatively assigned to an R-galactoside.
The variations in kinetic parameters kcat/Km and kcat for the
enzyme-catalyzed hydrolysis of 4-nitrophenyl R-D-galactopyra-
noside as a function of pH are listed in Table 1. The kinetic data
for kcat were fit to a classic bell-shaped pHꢀrate curve (Figure 2);
however, the resultant calculated pKa values, for the two cataly-
tically important ionizations, were not well-resolved, with both
values being 7.8 ( 0.5. In addition, the gross shape of the
pHꢀrate curve on kcat/Km is similar, with the maximal activity
occurring at pH ∼8.0 (Figure S3 of the Supporting Information),
but the fit of the kinetic data to a standard bell-shaped profile is
particularly ill-defined.
40 μM and 2.0 ( 0.4 mM, respectively. In the absence of Mn2þ
,
the enzyme exhibited an activity that was approximately 10%
of its maximal value, and after subtraction of this residual acti-
vity from the initial rates measured in the presence of Mn2þ, an
estimate of 220 ( 100 μM was made for the binding constant
4303
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
3,4,6-tri-O-acetyl-1,5-anhydro-2-deoxy-D-lyxo-hex-1-enitol) using
the nitrosyl chloride addition reaction (Scheme 2) pioneered by
Lemieuxand co-workers.17 Deuterated 7 was made by reduction of
a known 3-keto compound18 using sodium borodeuteride in place
of borohydride19 followed by a standard acetylation reaction
(Supporting Information).
Deuterium Kinetic Isotope Effects. To measure KIEs on
V/K, we used the substrate depletion method (Materials and
Methods). Of note, in these experiments, TCEP was used as the
reducing agent because in the presence of 2-mercaptoethanol the
background absorbance at 277 nm exhibited a slow increase over
time, a phenomenon that was traced to the reducing agent, and
this resulted in large errors for the measured rate constants. In
addition, to determine if the measured deuterium KIEs for H-2
and H-3 report on two sequential reactions that involve cleavage
of a CꢀH bond, namely, (i) the transfer of a hydride to NADþ
during oxidation of the 3-OH group and (ii) base-catalyzed
removal of the C-2 proton, or if they originate from a single
transition state, DV/K effects on deuterated substrates
were measured using the dideuterated substrate phenyl R-
D-[2-2H,3-2H]galactopyranoside. For the sake of consistency,
KIEs on Vmax were also measured with TCEP as the reducing
agent. In these experiments, full MichaelisꢀMenten kinetic
profiles were measured and the resultant Vmax values were used
Figure 2. pH dependence of kcat for the MelA-catalyzed hydrolysis of
4NPRG.
D
to determine the V effects. Tables 2 and 3 list the calculated
KIEs for the MelA-catalyzed hydrolysis of PRG.
’ DISCUSSION
The MelA gene from E. coli, which was proposed to encode a
GH4 R-galactosidase, has previously been cloned and sequenced.11
The sequence data showed that the MelA gene encodes a 450 amino
acid-protein that has a predicted molecular mass of 50.6 kDa.11 In this
work, the MelA gene from C. freundii was cloned and recombinantly
expressed in E. coli so a detailed mechanistic study of this GH4
glycosidase could be performed. As with other family 4 glycosyl
hydrolases,6,9,10 the MelA enzyme from C. freundii requires two
cofactors, NADþ and Mn2þ, as well as reducing conditions for
maximal activity. Moreover, this R-galactosidase is incapable of
hydrolyzing 4-nitrophenyl 6-phospho-R-D-galactopyranoside. An-
ggraeni et al. showed that for natural substrates the rMel4A R-
galactosidase from Bacillus halodurans displays higher rates of turn-
over and catalytic efficiencies (kcat and kcat/Km, respectively) for the
disaccharide melibiose [R-D-Galp-(1f6)-D-Glcp] than for the tri-
saccharide raffinose [R-D-Galp-(1f6)-R-D-Glcp-(1T2)β-D-Fruf].12
The dependence of the catalytic activity of C. freundii MelA R-
galactosidase on the presence of NADþ was shown by incubat-
ing the enzyme with NADþ in the presence of NaBH4 (10 mM),
conditions under which NADþ is reduced to NADH. The C.
freundii R-galactosidase was completely inactive under these
conditions, but after addition of a solution of NADþ (200 μM)
to the reaction medium, approximately 60% of the initial enzyme
activity was restored (Figure S1 of the Supporting Information).
Furthermore, the catalytic activity of the MelA R-galactosidase
varied as a function of NADþ concentration, and the activity
profile is consistent with the occurrence of substrate inhibition
[Kd = 240 ( 40 μM, and Kis = 2.0 ( 0.4 mM (Figure 1)]. As a
result, we decided to measure enzyme kinetic parameters at
subsaturating concentrations of NADþ. The variation in enzyme
activity as a function of Mn2þ concentration showed no evidence
of substrate inhibition, and all further kinetic experiments were
performed at saturating Mn2þ levels (1.0 mM; Kd ≈ 0.2 mM).
Figure 3. Brønsted plots for MelA-catalyzed hydrolysis, measured at
37 °C, of a series of aryl R-galactopyranosides and the associated pKa
values for the leaving group phenol. Colored red and blue are the kcat and
k
cat/Km kinetic data, respectively.
A panel of seven aryl R-D-galactosides was synthesized by
following literature procedures using peracetylated galactose as
the starting material (see Materials and Methods for details). Full
MichaelisꢀMenten curves were measured at pH 7.5 for each
substrate (Table S1 of the Supporting Information), and the
associated Brønsted βlg values on kcat and kcat/Km are ꢀ0.01 (
0.02 and 0.02 ( 0.04, respectively (Figure 3).
Synthesis of Deuterated Substrates. Phenyl R-D-[1-2H]-
galactopyranoside was made using peracetylated [1-2H]galactose
(Supporting Information) as the starting material. All other deut-
erated phenyl R-D-galactopyranosides were made by the appropriate
choice of reducing agent and starting tri-O-acetal D-galactal (4 or 7,
4304
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
Scheme 2. Synthetic Reactions Used in the Synthesis of C-2- and C-3-Deuterated PRG Substrates
Table 2. Kinetic Isotope Effects on V/K for the MelA
r-Galactosidase-Catalyzed Hydrolysis of Phenyl
r-D-Galactopyranoside at pH 7.5 and 37 °C
Table 3. Kinetic Isotope Effects on V for the MelA
r-Galactosidase-Catalyzed Hydrolysis of Deuterated Phenyl
r-D-Galactopyranoside at pH 7.5 and 37 °C
isotopologue 1
isotopologue 2
DV/Ka
isotopologue 1
isotopologue 2
DVa
PRG
[1-2H]PRG
[2-2H]PRG
[3-2H]PRG
[2,3-2H2]PRG
[2,3-2H2]PRG
1.13 ( 0.07
1.74 ( 0.06
1.74 ( 0.05
1.71 ( 0.12
1.71 ( 0.13
PRG
[1-2H]PRG
[2-2H]PRG
[3-2H]PRG
[2,3-2H2]PRG
[2,3-2H2]PRG
1.06 ( 0.07
0.91 ( 0.04b
1.02 ( 0.06
0.91 ( 0.06c
1.01 ( 0.06
PRG
PRG
PRG
PRG
[3-2H]PRG
[2-2H]PRG
[3-2H]PRG
[2-2H]PRG
a V/K(isotopologue 1)/V/K(isotopologue 2) ratio.
a Vmax(isotopologue 1)/Vmax(isotopologue 2) ratio. b Repeat measure-
ment gave a value of 0.99 ( 0.08. c Repeat measurement gave a value of
0.98 ( 0.08.
Anggraeni et al. showed, for the Mel4a GH4 R-galactosidase
from B. halodurans, that in the presence of reducing agent DTT
both Co2þ and Fe2þ also gave active enzyme.12 In contrast to
the reported results with the 6-phosphoglucosidases GlvA and
BglT,6,9 which are also members of glycosyl hydrolase family
4 (GH4), no activity was observed with MelA in the absence
of reducing agents. Also, when MelA-catalyzed hydrolysis of
4NPRG was performed in the presence of methanol (5 M), no
trapping by methanol was observed. Consequently, the stereo-
chemical outcome of MelA R-galactosidase-catalyzed reactions
was inferred, by analogy to other previously characterized GH4
hydrolases, to be a retaining enzyme. Also of note, when the
MelA-catalyzed hydrolysis 4NPRG was performed in D2O, the
reaction product, galactose, was completely deuterated at C-2 as
shown by the 1H NMR spectrum in which the anomeric protons
for both anomers were singlets.
experimental error, indistinguishable from zero (Figure 3). There-
fore, it can be concluded that cleavage of the anomeric CꢀObondis
not kinetically significant for the MelA R-galactosidase-catalyzed
hydrolysis of aryl R-D-galactopyranosides.
A detailed pHꢀrate profile for the catalyzed hydrolysis of
4NPRG was measured, and the kinetic data were fit to a standard
bell-shaped curve where the pKa values for the two catalytically
important groups are approximately 7.8ꢀ8.0 (Figure 2 and
Figure S3 of the Supporting Information).
To probe the effects of leaving group ability on catalytic activity,
we synthesized a panel of seven aryl R-D-galactopyranoside substrates
(structures 1a and 2aꢀf). The derived Brønsted βlg values, for these
substrates in which the pKa of the conjugate acid of the aglycone
varied by more than 3 pKa units, on both kcat and kcat/Km are, within
Withers and co-workers have also concluded that glycosidic
bond cleavage was not a kinetically important event for either the
4305
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
kinetic expressions for Scheme 1, in which oxidation (k3) and proton
abstraction (k5) are separate kinetic events. Thus, the formula for
DV/K(C-3) is given in eq 1
ꢀ
ꢁ
k3H k4H
þ
k6H
k7
Dk3 þ
1 þ
DKeq
ꢀ
ꢁ
V
k2 k5H
D
ꢀ
ꢁ
¼
ð1Þ
k3H k4H
þ
k6H
KC-3
C-2-H
1 þ
1 þ
k2 k5H
k7
where Dk3 is the intrinsic isotope effect on k3 and Keq is the
equilibrium isotope effect (EIE) on this step. The ratios of individual
kinetic terms in eq 1 are normally termed commitment factors, either
forward or reverse;21 i.e., cf = k3H/k2, and cr = k4H/k5H(1 þ k6H/k7).
However, to prevent confusion with different commitment factors
for the oxidation and proton removal steps, we used the individual
kinetic terms in the equations.
D
The corresponding expression for DV/K(C-2) is given in eq 2
(Scheme 1). Given the kinetic insignificance of aglycone depar-
ture as shown by the βlg value of zero (k7, Scheme 1), the term
k6H/k7 must be small (k7 . k6H).
ꢀ
ꢁ
k5H
k4H
k3H
k2
k6H
k7
Dk5 þ
1 þ
þ
DKeq
ꢀ
ꢁ
V
Figure 4. More O'FerrallꢀJencks diagram showing the possible transi-
tion state (red circle) for the concerted reaction from the Michaelis
complex to the enediol intermediate that circumvents formation of the
bound ketone (O). Also shown are plausible transition states (blue
circles) for the previously proposed stepwise mechanism.6,9 The enzy-
matic cofactors, NADþ and Mn2þ, have been omitted from the diagram
for the sake of clarity.
D
ꢀ
ꢁ
¼
ð2Þ
k5H
k3H
k6H
KC-2
C-3-H
1 þ
1 þ
þ
k7
k4H
k2
Equation 3 can be derived for the expected V/K KIE on C-3
when C-2 is deuterated, if k6H/k7 ≈ 0 (Supporting Information).
"
#
ꢀ
ꢁ
ꢀ
ꢁ
V
V
D
D
DKeq
ꢀ 1 þ
ꢀ
ꢁ
T. maritima BglT 6-phospho-β-glucosidase6 or the B. subtilis
GlvA 6-phospho-R-glucosidase9 GH4 enzyme. A detailed
KIE study was undertaken to delineate the important transi-
tion states for the MelA enzyme-catalyzed reaction. That is,
several different isotopologues that possessed a weakly acti-
vated leaving group were synthesized (Scheme 2 and the
Supporting Information). Listed in Tables 2 and 3 are the
measured V/K and V KIE values for deuteration at C-1, C-2,
or C-3. Of note, the tabulated DV and DV/K values are
markedly different. At first glance, the magnitudes of the
DV/K effects measured at both the C-2 (1.74 ( 0.06) and C-3
(1.74 ( 0.05) centers suggest that both oxidation and proton
abstraction are kinetically significant in accordance with
previously reported conclusions.6,9 To verify this analysis, we
decided to measure the isotope effects associated with deuterated
substrates. That is, if the C-2 and C-3 KIEs arise from a virtual
transition state resulting from two sequential steps, we would expect
that the KIE on C-2 should decrease if both isotopologues (C-2-H
and C-2-D) are deuterated at C-3. In other words, if deuteration at
C-3 increases the kinetic barrier for hydride transfer, then, for a
stepwise mechanism, the magnitude of the 2H KIE from deuteration
at C-2 must decrease for the C-3-deuterated substrates relative to
that measured with the C-3 nondeuterated galactosides;20 i.e.,
DV/K(C-2) > DV/K(C-2,3-2H).
KC-2
KC-3
V
C-3-H
C-2-H
D
ꢀ
ꢁ
¼
V
KC-3
D
C-2-D
KC-2
C-3-H
ð3Þ
Using the measured DV/K(C-2) and DV/K(C-3) values (both
equal 1.74) in conjunction with the reported EIE (DKeq = 1.18)
for oxidation of cyclohexanol by NADþ,22 the estimated value for
the magnitude of the DV/K(C-3,2-2H) effect is 1.50 (Supporting
Information). Of note, the incorporation of a second deuterium
at C-2 also results in a secondary equilibrium isotope effect for
the hydride transfer reaction (kinetic terms k3H and k4H), and
this (KH/KD) is expected to be approximately 1.054.22 Incor-
poration of this value into eq 3, which results in modification of
the value of DKeq to 1.24 (1.18 ꢁ 1.054), gives an expected effect
for DV/K(C-3,2-2H) of 1.53 (Supporting Information).
A similar formula (eq 4; see the Supporting Information) can
be derived for the value of DV/K(C-2,3-2H). Coincidentally, the
expected value for DV/K(C-2,3-2H), based on eq 4, is either 1.50
or 1.53 depending on whether the secondary EIE correction
noted above is included in the calculations.
"
#
ꢀ
ꢁ
V
D
D
ꢀ 1 Keq
ꢀ
ꢁ
KC-2
V
C-3-H
D
ꢀ
ꢁ
Therefore, phenyl R-D-[2-2H,3-2H]galactoside was synthesized
(Scheme 2), and both the DV/K(C-2,3-2H) and DV/K(C-3,2-2H)
KIEs were measured (Table 2). Of note, neither DV/K(C-2) nor
DV/K(C-3) shows any significant decrease in magnitude upon
incorporation of a remote deuterium (Table 2).
¼
þ 1 ð4Þ
V
KC-2
D
C-3-D
KC-3
C-2-H
Of note in the current example, the calculated values for both
D
DV/K(C-3,2-2H) and V/K(C-2,3-2H) are independent of the
An estimate of the expected reduction in DV/K values upon
incorporation of a second deuterium can be made by deriving the
intrinsic KIEs (Dk3 and Dk5) or any commitment factors. Of
note, given the standard error associated with the measured
4306
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
value for DV/K(C-2,3-2H) of 0.12 (Table 1), there is either
a 4% likelihood, i.e., 100% ꢀ 50% (value of >1.74) ꢀ 46% (value
between 0 and 1.75 standard deviations below 1.71), that the
data are consistent with the stepwise mechanism or a 6.7% prob-
ability when the EIE value is included in the calculations. Thus,
the measured DV/K values on dideuterated substrates of 1.71
(Table 1) are more consistent with a concerted reaction for
oxidation and proton removal.
A single concerted step requires that abstraction of the C-2
proton by the active site base start prior to completion of C-3
oxidation by the transfer of a hydride to the on-board NADþ
cofactor (Figure 4). In other words, during oxidation at C-3 to
form the ketone the C-2 proton becomes progressively more
acidic by virtue of the increasing polarization at C-3 and the
proximal Mn2þ ion until transfer begins, which in this case results
in the circumvention of formation of the proposed ketone
intermediate (Figure 4).
D-galactoside (Schemes S1 and S2) and labeled galactal 7,
derivation of equations for the estimation of expected DV/
K(C-3,2-2H) and DV/K(C-2,3-2H) values for a stepwise me-
chanism, MichaelisꢀMenten kinetic parameters for the MelA
R-galactosidase-catalyzed hydrolysis of the seven aryl R-D-galac-
tosides (Table S1), assay of reduced MelA (Figure S1), ligand
binding curve for Mn2þ (Figure S2), pH dependence of kcat/Km
for the MelA-catalyzed hydrolysis of 4NPRG (Figure S3), and
1H and 13C NMR spectra of labeled compounds 1bꢀe and
substrates 2bꢀf. This material is available free of charge via the
Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: bennet@sfu.ca. Telephone: (778) 782-8814. Fax: (778)
782-3765.
The magnitudes of the DV/K values (on C-2 and C-3)
reported here for the MelA reactions of phenyl R-galactopyrano-
side are similar to, albeit a little larger than, those reported for the
B. subtilis GlvA-catalyzed hydrolysis of 4-nitrophenyl 6-phospho-
R-glucopyranoside,9 a result that implies GH4 enzymatic reac-
tions that involve trans elimination reactions may proceed
through very similar transition state structures. In this study,
the negligible βlg values notwithstanding, no information about
the mechanism for the trans elimination reaction is available, a
situation that contrasts the cis eliminations encountered with
GH4 β-glycosidase where the DV/K values on C-2 are signifi-
cantly higher than those reported here.6,10
Funding Sources
This work was supported by the Natural Sciences and Engineer-
ing Research Council of Canada.
’ ABBREVIATIONS
[1-2H]PRG, phenyl R-D-[1-2H]galactoside; [2-2H]PRG, phenyl
R-D-[2-2H]galactoside; [2,3-2H2]PRG, phenyl R-D-[2-2H,-
3-2H]galactoside; [3-2H]PRG, phenyl R-D-[3-2H]galactoside; 4
NPRG, 4-nitrophenyl R-D-galactoside; BSA, bovine serum albumin;
DMAP, 4-(dimethylamino)pyridine; DTT, dithiothreitol; GH, gly-
cosyl hydrolase; GH4, glycosyl hydrolase family 4; HEPES, 4-(2-
hydroxyethyl)-1-piperazineethanesulfonic acid; IPTG, isopropyl
1-thio-β-D-galactopyranoside; KIE, kinetic isotope effect; NADþ,
β-nicotinamide adenine dinucleotide; NADH, reduced β-nicoti-
namide adenine dinucleotide;PRG, phenyl R-D-galactoside;PNP,
p-nitrophenyl; TBDPS, tert-butyl diphenylsilyl; TCEP, tris(carboxy-
ethyl)phosphine; THF, tetrahydrofuran; TLC, thin layer chromatog-
raphy; Tris, tris(hydroxymethyl)aminomethane; TS, transition state;
UVꢀvis, ultravioletꢀvisible.
The large relative errors associated with the normal secondary
deuterium KIE at the anomeric center, DV/K(C-1), make a
detailed discussion of the origin of the effect difficult at present,
although the minimal Brønsted βlg values are inconsistent with
the occurrence of any TS rehybridization at this position.
The measured KIEs on V for the MelA R-galactosidase-
catalyzed hydrolysis of phenyl R-galactoside are markedly differ-
D
ent from the V/K values (cf. Tables 2 and 3). All effects are
either equal to 1, within experimental error, or slightly inverse
[DV(C-2)]. These KIE measurements are consistent with the TS
for V being associated with a different reaction step. Given that
this step must come after those that limit V/K23 and that the βlg
value on V is approximately zero, the first possible TS for V
involves the hydration of the enzyme-bound glycal intermediate
(k9, Scheme 1). However, at present, given the lack of specific
evidence, other subsequent steps in the mechanism, such as
product release, might be kinetically significant for V.
’ REFERENCES
(1) Henrissat, B. (1991) A classification of glycosyl hydrolases based
on amino acid sequence similarities. Biochem. J. 280, 309–316.
(2) Vocadlo, D. J., and Davies, G. J. (2008) Mechanistic insights into
glycosidase chemistry. Curr. Opin. Chem. Biol. 12, 539–555.
(3) Rao, S. T., and Rossmann, M. G. (1973) Comparison of super-
secondary structures in proteins. J. Mol. Biol. 76, 241–256.
(4) Lodge, J. A., Maier, T., Liebl, W., Hoffmann, V., and Strater, N.
(2003) Crystal structure of Thermotoga maritima R-glucosidase AglA
defines a new clan of NADþ-dependent glycosidases. J. Biol. Chem.
278, 19151–19158.
(5) Liu, Q. Y. P. et al. (2007) Bacterial glycosidases for the produc-
tion of universal red blood cells. Nat. Biotechnol. 25, 454–464.
(6) Yip, V. L. Y., and Withers, S. G. (2006) Mechanistic analysis of
the unusual redox-elimination sequence employed by Thermotoga
maritima BgIT: A 6-phospho-β-glucosidase from glycoside hydrolase
family 4. Biochemistry 45, 571–580.
’ CONCLUSIONS
The MelA R-galactosidase from C. freundii catalyzes hydrolysis
of aryl substrates via a mechanism in which the first irreversible
step (rate-limiting for V/K) involves oxidation of the C-3
hydroxyl group and proton abstraction at C-2, a process that
starts before completion of the NADþ-promoted hydride trans-
fer. Subsequent elimination of the leaving group is kinetically
silent, and this event is followed by hydration of the R,β-
unsaturated enone intermediate. The catalytic cycle is completed
by protonation at C-2 and reduction by the on-board NADH.
(7) Thompson, J., Pikis, A., Ruvinov, S. B., Henrissat, B., Yamamoto,
H., and Sekiguchi, J. (1998) The gene glvA of Bacillus subtilis 168
encodes a metal-requiring, NAD(H)-dependent 6-phospho-R-glucosi-
dase: Assignment to family 4 of the glycosylhydrolase superfamily. J. Biol.
Chem. 273, 27347–27356.
’ ASSOCIATED CONTENT
(8) Coutinho, P. M., Deleury, E., Davies, G. J., and Henrissat, B.
(2003) An evolving hierarchical family classification for glycosyltrans-
ferases. J. Mol. Biol. 328, 307–317.
S
Supporting Information. Synthetic details for 1,2,3,4,6-
b
penta-O-acetyl-D-[1-2H]galactose, 4-nitrophenyl 6-phospho-R-
4307
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308

Biochemistry
ARTICLE
(9) Yip, V. L. Y., Thompson, J., and Withers, S. G. (2007) Mechan-
ism of GlvA from Bacillus subtilis: A detailed kinetic analysis of a
6-phospho-R-glucosidase from glycoside hydrolase family 4. Biochemistry
46, 9840–9852.
(10) Yip, V. L. Y. et al. (2004) An unusual mechanism of glycoside
hydrolysis involving redox and elimination steps by a family 4 β-
glycosidase from Thermotoga maritima. J. Am. Chem. Soc. 126, 8354–8355.
(11) Liljestr€om, P. L., and Liljestr€om, P. (1987) Nucleotide-se-
quence of the MelA gene, coding for R-galactosidase in Escherichia coli
K-12. Nucleic Acids Res. 15, 2213–2220.
(12) Anggraeni, A. A., Sakka, M., Kimura, T., Ratanakhaokchai, K.,
Kitaoka, M., and Sakka, K. (2008) Characterization of Bacillus halodur-
ans R-galactosidase Mel4A encoded by the mel4A gene (BH2228).
Biosci., Biotechnol., Biochem. 72, 2459–2462.
(13) Sokolov, V. M., Zakharov, V. I., and Studentsov, E. P. (2002)
Stereoselectivity of reactions at the glycoside center of carbohydrates:
VII. Synthesis of aryl R- and β-D-glucopyranosides by Helferich,
catalyzed by boron trifluoride etherate. Russ. J. Gen. Chem. 72, 806–811.
(14) Comfort, D. A. et al. (2007) Biochemical analysis of Thermo-
toga maritima GH36 R-galactosidase (TmGalA) confirms the mechan-
istic commonality of Clan GH-D glycoside hydrolases. Biochemistry
46, 3319–3330.
(15) Lemieux, R. U., Nagabhushan, T. L., and O'Neill, I. K. (1964) A
new approach to the synthesis of 2-amino-2-deoxyglycosides. Tetrahe-
dron Lett., 1909–1916.
(16) Taylor, J. R. (1982) An introduction to error analysis: The study of
uncertainties in physical measurements, University Science Books, Mill
Valley, CA.
(17) Lemieux, R. U., Nagabhushan, T. L., and Gunner, S. W. (1968)
Synthesis of 3,4,6-tri-O-acetyl-2-oximino-R-D-hexopyranosides. Can. J.
Chem. 46, 405–411.
(18) Czernecki, S., Vijayakumaran, K., and Ville, G. (1986) Conve-
nient synthesis of hex-1-enopyran-3-uloses: Selective oxidation of allylic
alcohols using pyridinium dichromate. J. Org. Chem. 51, 5472–5475.
(19) Fujiwara, T., and Hayashi, M. (2008) Efficient synthesis of
rare sugar D-allal via reversal of diastereoselection in the reduction of
protected 1,5-anhydrohex-1-en-3-uloses: Protecting group dependence
of the stereoselection. J. Org. Chem. 73, 9161–9163.
(20) Hermes, J. D., Roeske, C. A., O'Leary, M. H., and Cleland,
W. W. (1982) Use of multiple isotope effects to determine enzyme
mechanisms and intrinsic isotope effects. Malic enzyme and glucose-6-
phosphate dehydrogenase. Biochemistry 21, 5106–5114.
(21) Cook, P. F., and Cleland, W. W. (2007) Enzyme kinetics and
mechanism, Garland Science, London.
(22) Cook, P. F., Blanchard, J. S., and Cleland, W. W. (1980)
Primary and secondary deuterium isotope effects on equilibrium con-
stants for enzyme-catalyzed reactions. Biochemistry 19, 4853–4858.
(23) Fersht, A. (1985) Enzyme structure and mechanism, 2nd ed.,
W. H. Freeman, New York.
4308
dx.doi.org/10.1021/bi101808h |Biochemistry 2011, 50, 4298–4308