1. Introduction
α-
d-Galactosidases (EC 3.2.1.22) catalyze the hydrolysis of the nonreducing terminal α-
d-galactose (Gal) from α-
d-galactosides, galactooligosaccharides, and polysaccharides, such as galactomannans, galactolipids, and glycoproteins. According to the classification of carbohydrate-active enzymes (CAZy) [
1], α-
d-galactosidases mostly belong to 27, 36, and 110 families of glycoside hydrolases (GH). They are found also in the GH 4, GH 57, and GH 97 families. The GH 27 and GH 36 enzymes, with a common mechanism of catalysis, have the protein structural (β/α)
8-barrel fold in the catalytic domain and similar topology of their active centers, typical for a clan GH D [
2]. The GH 27 and GH 36 family members are classical retaining glycoside hydrolyses in accordance with Koshland’s classification [
3]. These enzymes catalyze the hydrolysis of O-glycosidic bonds by a double displacement mechanism through the galactosyl-enzyme covalent intermediate, as well as the transglycosylation reaction under the specific conditions [
4].
α-
d-Galactosidases are widespread among terrestrial plants, animals, and microorganisms. These enzymes have found many practical uses in different fields from biomedicine to enzymatic synthesis [
4]. The enzymes occur frequently in marine bacteria, especially in γ-Proteobacteria and Bacteroidetes [
5,
6,
7,
8,
9]. Currently, the genes encoding these enzymes can be found in the genomes of marine bacteria, annotated in the National Center for Biotechnology Information NCBI database. For the first time, α-PsGal was isolated from a cold-adapted marine bacterium
Pseudoalteromonas sp. strain KMM 701 inhabiting in the cold water in the Sea of Okhotsk. The enzyme attracted our attention due to its ability to reduce the serological activity of B red blood cells. The marine bacterium’s α-PsGal was more efficient in the model of B-erythrocyte antigen, than a well-known α-
d-galactosidase from green coffee beans, which was usually used in experiments on transformation of donor blood erythrocytes for intravenous injection [
10]. The enzyme also interrupted the adhesion of
Corynebacterium diphtheria to buccal epithelium cells at neutral pH values [
11], as well as stimulated the growth of biofilms of some bacteria [
12]. These properties of the enzyme determined the possible directions for its practical application in biomedicine. According to the structural CAZy classification, α-PsGal belongs to the GH 36 family [
11]. The enzyme is a retaining glycoside hydrolase [
13], cleaving the terminal Gal from melibiose Gal-α(1→6)-Glc, raffinose Gal-α(1→6)-Glc-β(1→4)-(Fru), digalactoside Gal-α(1→3)-Gal, and B-trisaccharide Gal-α(1→3)-(Fuc-α(1→2)-Gal) [
10]. However, the most important glycosynthase properties of the enzyme have yet to be studied.
The present article aimed to compare the properties of recombinant α-d-galactosidase from marine bacterium Pseudoalteromonas sp. KMM 701 (α-PsGal) and its mutants, where the predicted functionally important residues D451 and C494 of the active center were replaced by the less reactive alanine (A) and asparagine (N) residues, respectively. Major attention was focused on the regioselectivity of the transglycosylation reaction.
3. Discussion
The catalytic properties and structure-function relationships for the marine bacterial α-galactosidase from the GH 36 family, whose genes frequently occur in the genomes of marine bacteria, were characterized for the first time for recombinant α-galactosidase from the marine bacterium Pseudoalteromonas sp. KMM 701 (α-PsGal). As a result of our bioinformatic analysis of the amino acid sequence of the enzyme and homologous modeling of the 3D structure, presumably catalytic (D451 and D516) and substrate-binding (C494) residues—extremely important for the functioning of the enzyme—were identified. The predicted nucleophilic residue D451 and substrate-binding residue C494 were replaced with A451 and N494, respectively. Properties of the mutant D451A and C494N were investigated with comparison to wild α-PsGal.
We showed that α-PsGal and its mutant C494N are cold-active enzymes characterized by their neutral pH-optima (6.5–8.0) and low thermostability of 20 to 30 °C among the known α-galactosidases. The wild enzyme exhibited about 30% activity at 5 °C. No data on temperature and pH effects on the activity were available in the literature for the prototype α-galactosidases from the mesophiles
L. acidophilus. The α-galactosidases from different mesophilic lactobacilli showed an acidic optimum activity, in the pH range from 5.2 to 5.9, and maximum activity between higher temperatures of 38 to 42 °C [
23] compared with α-PsGal. The optimal temperature for the activity of the AgaA enzyme from psychrophilic lactic acid bacterium
Carnobacterium piscicola was 32 to 37 °C [
24]. The optimum temperature of the enzyme from
Lactobacillus fermentum was found to be 45 °C. The enzyme was inactivated at temperatures higher than 55 °C and stable in wide ranges of temperatures and pH [
25,
26]. As for thermophilic enzymes from bacteria-thermophiles and hyperthermophiles
Bifidobacterium adolescentis DSM 20083,
B. stearothermophilus,
Thermus brockianus [
27],
Thermus sp. T2 [
28],
Thermoanaerobacterium polysaccharolyticum [
29], and
Thermotoga maritime, their temperature optimums were 75 to 100 °C. The last enzyme was inactive at 30 °C [
30].
Capability of catalyzing a transglycosylation reaction is an inherent property of all members of the retaining α-
d-galactosidases of the GH 27 and GH 36 families [
4]. The inverting α-
d-galactosidases of the GH 110 family [
31], as well as NAD
+- and Mn
2+-dependent α-
d-galactosidases found in the family GH 4 [
32], have lost their transglycosylation properties. To date, there is no information about the transglycosylation ability of α-
d-galactosidases from the GH 97 and GH 57 families.
It is known that the first step in the catalytic reaction is cleavage of the glycosidic bond of the melibiose or pNP-α-Gal molecules, as well as the formation of the covalent galactosyl-enzyme intermediate. The molecules of Glc and pNP are leaving groups. In the second step, water or some carbohydrate molecules attack the covalent galactosyl-enzyme intermediate, and then hydrolysis or transglycosylation, respectively, can be observed. In the case where the substrate is an attacking molecule, we can observe an autocondensation reaction (
Figure S5).
α-PsGal catalyzed synthesis with a total yield of transglycosylation products ranging from 6.0% to 12% (
Table 6), similar to the known retaining bacterial galactosidases of the GH 36 family. It was difficult to identify the structures of the transglycosylation products without appropriate standards. However, the use of three methods (TLC, MALDI MS in conjunction with ESIMS/MS, and NMR) provided a suggestion of the relationships between the sugars’ molecular weights and the type of O-glycoside bonds in the synthesized oligosaccharides.
TLC is commonly used to analyze low-molecular-weight sugars and their derivatives that differ in the number of carbon atoms, configurations, and molecule sizes. If two carbohydrates have one of these three different characteristics, they can be separated [
33]. Based on the results, we assumed that the spot with R
f of 0.35 corresponds to bihexoses, distinguishable from melibiose by the configuration of the stereocenter, but the spot with R
f of 0.26 corresponds to sugars distinct from melibiose by the degree of polymerization (
Figure 7, lane 1). The
1H NMR signals of the anomeric atoms of the trisaccharides were not detected. However, the signal of [Hex
3 + Na]
+ at
m/
z 527 was observed in the MALDI MS (
Figure S2a). The structure of the trisaccharide Hex-(1→4)-Hex-(1→6)-Hex was established by electrospray ionization tandem mass spectrometry as Gal-(1→4)-Gal-(1→6)-Glc (
Figure S2b). The kinetic of accumulation and consumption of Gal-(1→4)-Gal-(1→6)-Glc was registered with the use of heavy-oxygen water (
Figure 8b). In the NMR spectrum of the reaction mixture, we found the 1H signals of the anomer atoms (1→6)-α- and (1→4)-α-linked bigalactosides only. In this connection, we think that the spot with an R
f of 0.35 corresponds to two poorly shared (1→6)-α- and (1→4)-α-linked bigalactosides Gal-α(1→6)-Galα,β (
3) and Gal-α(1→4)-Galα,β (
4), respectively. This assumption was confirmed by EISMS-MS (
Figure S2c).
Thus, when melibiose was used as the substrate, the enzyme synthesized the (1→6)-α-linked bigalactosides (
Figure S2c), similar to all known melibiases of the GH 36 family [
16,
34,
35,
36,
37,
38,
39,
40,
41,
42] and to their closely-related GH 27 representatives [
43,
44,
45,
46,
47,
48,
49,
50,
51,
52]. Furthermore, α-PsGal formed the (1→4)-α-linked bigalactosides as described for mesophilic terrestrial α-
d-galactosidases from
Bifidobacterium breve 203 [
35],
Lactobacillus reuteri [
16], and the acidic GH 27 family α-
d-galactosidases (AgaBf3S) from the bacterium
Bacteroides fragilis. The latter was able to transfer galactosyl residues from pNP-α-Gal in lactose Gal-β(1→4)-Glc with the efficiency and strict (1→4)-α-regioselectivity [
52], whereas α-PsGal synthesized both (1→6)-α- and (1→4)-α-
O-glycoside bonds in the bigalactosides from melibiose in the ratio of 9:1 at 20 °C and 5:1 at 8 °C. It is interesting to note that glucose, which is released from melibiose, did not participate in the transglycosylation reaction as an acceptor because its content in the mixtures was almost half of all products without any change in the course of the reaction (
Table 3).
Similarly, we established the structure of the autocondensation products in the mixtures of α-PsGal and pNP-α-Gal (
Figure 7, lanes 2 and 4, respectively). α-PsGal was able to produce novel compounds by catalyzing the autocondensation reaction of pNP-α-Gal. Both the substituted Gal
2-pNP with (1→6)-α- and (1→3)-α-O-glycoside bonds and unsubstituted Gal
2 with (1→6)-α- and (1→4)-α-
O-glycoside bonds were found in the reaction mixture. The ratio of (1→6)-α-:(1→3)-α-linked Gal
2-pNP was 7:1, but the ratio for unsubstituted (1→6)-α-:(1→4)-α-linked bigalactosides was 3:1 at 20 °C. The ratio of (1→6)-α-:(1→4)-α-linked bigalactosides reached up to 2:1 at 8 °C (
Table 6).
The transglycosylation properties are well-studied for the highly thermoresistant GH 36 α-
d-galactosidase from the hyperthermophilic bacterium
Thermotoga maritima (TmGal36A). This enzyme catalyzes an autocondensation reaction with pNP-α-Gal as a substrate, forming substituted (1→2)-α-, (1→3)-α- and (1→6)-α-linked Gal
2-pNP [
22]. In total, the wild TmGal36A can produce up to 5.5% transglycosylation products. The mechanism of the hydrolysis and synthesis in TmGal36A is not favorable for the formation and breaking of the (1→4)-α-O-glycosidic linkage [
22], unlike α-
d-galactosidases from human intestine [
34,
35,
36,
37] and α-PsGal from marine bacterium.
The replacement of the predictive nucleophilic residue D451 to A451 in the active center led to complete loss of the ability of α-PGal to catalyze the hydrolysis. For unknown reasons, the rescue strategy, with an addition of the external nucleophilic sodium azide, proved to be ineffective in this case. Molar concentrations of sodium azide or sodium formate were unable to restore or increase the activity of the mutant D425G of α-
d-galactosidase from archaeon
Sulfolobus solfataricus [
53]. Sodium azide did not inhibit any activity of the wild enzyme α-PsGal [
10], but it did not restore the activity in its mutant D451A. Galactosyl-β-azide was not found both either of the reaction products of mutant D451A and pNP-α-Gal substrate, as occurred in the experiment with TmGal36A [
54]. The D327G mutant of TmGal36A lost hydrolytic properties but retained glycosynthase properties and became an effective α-galactosynthase, which could produce various galactosylated disaccharides from galactosyl-β-azide as a donor and pNP-α(β)-galactosides as acceptors [
55].
The mutation C494N changed the specificity for α-PsGal in the synthesis of O-glycoside bonds. Under the action of the C494N mutant on pNP-α-Gal, the yield of pNP-Gal-α-(1→6)-Gal (
6) decreased, whereas pNP-Gal-(1→3)-α-Gal was not observed. In addition, the content of Gal-(1→4)-α-Gal (
4) increased two-fold (
Table 6). In the literature, it has been reported that the substitution of some bulk residues in the active site of α-
d-galactosidase Aga A from
Bacillus stearothermophilus KVE39 resulted in a 4.5-fold increase in the yield of substituted (1→3)-α-linked compared with pNP-Gal-α-(1→6)-Gal [
37]. A number of single and double substitutions of protruded residues in the active site of α-
d-galactosidase from
Bifidobacterium adolescentis DSM 20083 led to an increase in the yield of the total transglycosylation products but they did not change the regioselectivity of the reaction [
22,
38].
4. Materials and Methods
4.1. Materials
The 4-nitrophenyl-α-d-galactopyranoside (pNP-α-Gal), melibiose (Gal-α-(1→6)-Glc), galactose (Gal), glucose (Glc), Bovine serum albumin (BSA), NaN3, and 2,5-dihydroxybenzoic acid were purchased from Sigma Chemical Company (St. Louis, MO, USA). Encyclo DNA-polymerase and enterokinase were purchased from Evrogen (Moscow, Russian Federation). Sodium phosphates, one- and two-substituted, were purchased from PanReac AppliChem GmbH (Darmstadt, Germany). IMAC Ni2+ Sepharose, Q-Sepharose, Mono-Q, and Superdex-200 PG were purchased from GE Healthcare (Uppsala, Sweden). Heavy-oxygen water was purchased from Component Reactive (Moscow, Russia).
4.2. Homology Model of α-PsGal 3D Structure
The target-template alignment customization of the modeling process and 3D model building of α-PsGalA (GenBank: ABF72189.2) were carried out using the Molecular Operating Environment version 2018.01 [
14] package (Chemical Computing Group ULC: 1010 Sherbrooke St. West, Suite #910, Montreal, QC, Canada, H3A 2R7, 2018) using the forcefield Amber12: EHT. The α-
d-galactosidase from
Lactobacillus acidophilus NCFM (PDB code: 2XN2) with a high-resolution crystal structure was used as a template. The evaluation of structural parameters, contact structure analysis, physicochemical properties, molecular docking, and visualization of the results were carried out with the Ligand interaction and Dock modules in the MOE 2018.01 program (Chemical Computing Group ULC: 1010 Sherbrooke St. West, Suite #910, Montreal, QC, Canada, H3A 2R7, 2018). The results were obtained using the equipment of Shared Resource Center Far Eastern Computing Resource of Institute of Automation and Control Processes Far Eastern Branch of the Russian Academy of Sciences (IACP FEB RAS) [
56]
4.3. Production of Recombinant Enzymes
The recombinant wild α-
d-galactosidase α-PsGal was produced as described earlier [
13]. The D451A and C494N mutants were produced by polymerase chain reaction (PCR)-mediated site-directed mutagenesis using the full-length wild gene of α-PsGal. The mutations were inserted in the sequences of synthetic oligonucleotides for each DNA chain of the wild gene:
- (1)
D451A dir 5′-TTAAGTACATTAAATGGGCTATGAACCGCGA-3′ D451A rev 5′-GTTAATATCGCGGTTCATAGCCCATTTAATG-3′
- (2)
C494N dir 5′-AGGGCTTGAAATAGAAAGCAATTCGTCAGGTGG-3′ C494N rev 5′-ACGTGCACCACCTGACGAATTGCTTTCTATTTC-3′
The plasmid DNA pET40 containing an insertion of the α-
d-galactosidase gene of the marine bacterium
Pseudoalteromonas sp. KMM 701 (α-PsGal) or its D451A and C494N mutants were transformed in the
E. coli strain Rosetta (DE3). Heterological expression was carried out at optimal conditions into
E. coli, as described previously [
57]. Purification of the recombinant α-PaGal and its D451A and C494N mutant forms were performed according to the procedures described previously [
13].
4.4. Enzyme and Protein Essays
To determine the activity, 0.02 mL of an enzyme solution were mixed with 0.38 mL of the pNP-α-Gal solution (1 mg/mL in 0.05 M sodium phosphate buffer, pH 7.0). The reaction mixture was incubated at 20 °C for 10 min. The reaction was stopped by addition of 0.6 mL of 1 M Na
2CO
3. One unit of activity (U) was determined as the amount of enzyme that releases 1 μmol of pNP per 1 min at 20 °C in 0.05 M sodium phosphate buffer at pH 7.0. The amount of released pNP was determined spectrophotometrically (ε
400 = 18300 M
−1 cm
−1). The specific activity was calculated as U/mg of protein. Protein concentration was determined by the Bradford method and calibrated with BSA as a standard [
58].
4.4.1. Circular Dichroism Spectra
The CD spectra were recorded in the ultraviolet (UV) region of 190 to 250 nm with Chirascan plus CD spectrometers (Applied Photophysics Ltd., Leatherhead, UK), equipped with an optional Peltier temperature controller for rapid and precise temperature control of the sample cell (Quantum North West, 22910 E Appleway Avenue, Suite 4 Liberty Lake, WA, USA), in 0.01 M sodium phosphate buffer (pH 7.0) and 20 °C. The average molecular weight of the amino acid residue for calculation of molar ellipticity [Ѳ] (degree cm
2 dmol
−1) was assumed to be 112 Da. The secondary structure elements were calculated by the Provencher–Glöcker method CONTIN/LL modified by Sreerama N. of the CDPro software package, 2000 (Colorado State University, Fort Collins, Colorado, USA,
http://lamar.colostate.edu/sreeram/CDPro) [
59,
60]
4.4.2. UV Absorption Spectra
Absorption spectra of proteins were recorded with a UV-Visible spectrophotometer UV-1601 PC (Shimadzu, Kyoto, Japan) in quartz cells with an optical path length of 1 cm, 0.1 cm, and 0.01 cm in the range of 190 to 400 nm. The molar extinction coefficient of enzyme ε
280 = 100,770 M
−1 cm
−1 was calculated from the content of aromatic amino acids using the ExPASy server [
61].
4.5. Effect of pH and Temperature
The pH optimums of purified enzymes were determined with pNP-α-Gal as the substrate in the pH range of 5.2 to 6.5 using 0.1 M sodium citrate buffer, in the pH range of 6.2 to 8.0 using 0.1 M sodium phosphate buffer, and in the pH range of 7.8 to 9.0 with 0.1 M Tris-HCl buffer. The temperature optimums for the purified enzymes were determined at pH 7.0 in the temperature range of 5 to 40 °C. The temperature stabilities of the enzymes were investigated after incubation for 60 min at 10 to 40 °C.
4.6. Determination of Kinetic Parameters
All kinetic studies were performed in 0.1 M sodium phosphate buffer, pH 7.0, at 20 °C. The Michaelis–Menthen constants, Km and Vmax, were determined from the coefficients of linear regression of the Lineweaver–Burk plot. The substrate concentrations (mM) were 3.24, 2.59, 1.62, 1.29, 0.81, 0.65, 0.40, and 0.32 for wild α-PsGal and 2.50, 2.0, 1.25, 1.0, 0.62, 0.50, 0.31, and 0.25 for the C494N mutant.
4.7. Transglycosylation
4.7.1. Acceptor Specificity of Transglycosylation
For preliminary determination of acceptor specificity of transglycosylation, the synthesis reactions were performed at 20 °C for 24 h in a mixture (10 μL) containing 0.01 U of an enzyme, 10 mM of pNP-α-Gal or melibiose as the substrate, and 20 mM of glucose, galactose, fructose, fucose, or xylose as acceptor in 0.05 M sodium phosphate buffer (pH 7.0). The reaction was stopped by heating at 100 °C for 5 min and the reaction mixture was centrifuged at 14,000 rpm. Sugars were analyzed by TLC, mass spectrometry, and NMR spectroscopy methods.
4.7.2. Transglycosylation Using Heavy-Oxygen Water (H218O)
An experiment using mass spectrometry and heavy-oxygen water was prepared similarly as described in
Section 4.7.1. above; but the concentrations were significantly lowered. Seven identical reaction mixtures were created. Each mixture contained 1.4 mg melibiose, 10 μL enzyme (1 U), and 70 μL H
2O
18 (0.02 M sodium phosphate buffer, 0.05 M NaCl, pH 7.0). The mixtures were incubated for 1, 2, 4, 6, 8, 12, and 24 h. Each reaction mixture was dissolved in 1 mL methanol and introduced into the mass spectrometer. For ESIMS and ESIMS-MS experiments, direct injection was performed using a syringe pump (KD Scientific, Hollison City, MA, USA) at a flow rate 5 µL/min. For LC-ESIMS experiments, samples were further diluted 10 times in methanol.
4.7.3. Identification of Transglycosylation Products
The recombinant α-PsGal or D451A and C494N mutants, in an aqueous solution of 0.05 M sodium phosphate buffer (pH 7.0), were added to the preweighed dry samples of substrates melibiose or pNP-α-Gal or their mixture and were incubated for a certain time (
Table 3) at 20 °C or 8 °C. The standard units of activity (U) or milligrams enzyme added (mg), the incubation time (τ) and reaction temperature are shown in
Table 2. The reaction was stopped by heating at 100 °C for 5 min. The samples were centrifuged to remove the denatured protein and dried on Refrigerated CentriVap Concentrater (Labconco, Kansas City, MO, USA). The qualitative composition of the hydrolysis and transglycosylation products were analyzed by TLC and MALDI-MS without their separation from the reaction mixtures. Identification of oligosaccharides in the mixture and their output was performed via NMR spectroscopy.
4.7.4. Thin-Layer Chromatography Analysis
Mono- and oligosaccharide composition of the hydrolysis and transglycosylation products were analyzed on silica gel TLC plates on aluminum foil (Sigma-Aldrich, St. Louis, MO, USA) with a 254 nm fluorescent indicator. The pore diameter was 60 A. Rf was calculated for every stain.
Weighed freeze-dried reaction product mixtures were placed in 0.5 mL Eppendorf, dissolved in distilled water to a concentration of 5 mg/mL, and centrifuged at 10,000 rpm to remove the denatured protein. A small spot of the analyzed mixture and standard compounds were applied at the start line of the TLC plate and chromatographed over 30 minutes in a sealed chamber Latch-Lid ChromatoTank (General Glass Blowing Co. Inc., Richmond, CA, USA), containing 100 mL of the mobile phase butanol/acetic acid/water (3:1:1; v/v/v). For visualization of stains, the plate was treated three times with 5% sulfuric acid solution, drying by warm air after each spraying.
4.7.5. Mass Spectrometry Analysis
The molecular weights of the oligosaccharide ions were recorded as sodium adducts [M + Na]+ using MALDI time-of-flight mass spectrometer, Ultra Flex III (Bruker BioSpin GmbH, Rheinstetten/Karlsruhe, Germany) equipped with a smartbeam laser (355 nm, Bruker Daltonik GmbH, Bremen, Germany) in reflector mode at an accelerating voltage of 21 kV, using the saturated solution (acetonitrile-water, 1:1) of 2,5-dihydroxybenzoynoic acid as a matrix.
The composition of the oligosaccharide mixture after enzymatic transformation was performed using an Ultimate 3000 rapid separation liquid chromatography (RSLC) nano system (Dionex, Thermo Fisher Scientific, Waltham City, MA, USA) connected to a Bruker Impact II quadrupole time-of-flight (Q-TOF) mass spectrometer (Bruker Daltonics, Bremen, Germany). An Acclaim (Thermo Fisher Scientific, Waltham City, MA, USA) PepMap RSLC column (75 µm × 150 mm, C18, 2 μm, 100 A) was used for chromatographic separation. The mobile phases were 0.1% formic acid in H2O (eluent A) and 0.1% formic acid in acetonitrile (eluent B). The gradient program was: isocratic at 1% of eluent B from start to 5 min, from 1% to 10% eluent B from 5 to 10 min, from 10% to 95% eluent B from 10 to 11 min, and isocratic at 95% of eluent B to 15 min. After returning to the initial conditions, equilibration was achieved after 10 min. Chromatographic separation was performed at a 0.4 µL/min flow rate at 40 °C. Injection volume was 0.2 μL. The mass spectrometry detection was performed using CaptiveSpray (Bruker Daltonics, Bremen, Germany) ionization source at a capillary voltage of 1.3 kV. Collision induced dissociation (CID)-produced ion mass spectra were recorded in auto-MS/MS mode with collision energy 43 eV. The precursor ions were isolated with an isolation width of 1 mass unit.
The mass spectrometer was calibrated using the ESI-L Low Concentration Tuning Mix (Agilent Technologies, Santa Clara, CA, USA). The instrument was operated using the OTOFControl software (version 4.0, Bruker Daltonics, Bremen, Germany) and data were analyzed using Data Analysis software (version 4.3, Bruker Daltonics, Bremen, Germany).
4.7.6. NMR Spectroscopy Analysis
The structure of disaccharides was characterized by NMR spectroscopy. Signals in the NMR spectra of sugars were assigned by two-dimensional correlated spectroscopy (H,H-COSY) and two-dimensional heteronuclear multiple bond correlation spectroscopy (HSQC) experiments. One-dimensional 1H-NMR and 13C-NMR, and two-dimensional H,H-COSY and HSQC spectra were recorded with a Bruker Avance III 500 HD (Bruker BioSpin GmbH, Rheinstetten/Karlsruhe, Germany) spectrometer in D2O at 50 °C with acetone as internal standard (δ = 31.45 and 2.20 ppm for 13C NMR and 1H NMR spectra, respectively). 1H NMR anomer signals of α-pNP-galactopyranose, β-galactopyranose, melibiose, and bigalactosides, as well as proton signals of the free 4-nitrophenol ring were analyzed and integrated by the standard software TopSpin 3.2.
The depth of the pNP-α-Gal conversion (H
pNPαGal, %) was calculated by
where I
Gal is integrated intensities of all 1H signals of liberated Galα,β, I
pNPαGal is the integrated intensity of 1H proton signals in substrate pNP-α-Gal, and I
Gal2pNP and I
Gal2 are the integrated intensity of 1H proton signals in transglycosylation products.
The depth of melibiose conversion (H
Mel, %) was calculated by
where I
Mel is the integrated intensities of signals 1H of initial mixture of melibiose.
Yield of oligosaccharides in the total reaction mixture (Y, %) was calculated by
where I
product is the integrated intensity of all 1H signals of a particular oligosaccharide.