How Site-Directed Mutagenesis Boosted Selectivity of a Promiscuous Enzyme

Title: How Site-Directed Mutagenesis Boosted Selectivity of a

Promiscuous Enzyme

Authors: Pavlína Nekvasilová, Natallia Kulik, Nikola Rychlá, Helena

Pelantova, Lucie Petraskova, Zuzana Bosáková, Josef

Cvacka, Kristyna Slamova, Vladimir Kren, and Pavla Bojarová

This manuscript has been accepted after peer review and appears as an

Accepted Article online prior to editing, proofing, and formal publication

of the final Version of Record (VoR). This work is currently citable by

using the Digital Object Identifier (DOI) given below. The VoR will be

published online in Early View as soon as possible and may be different

to this Accepted Article as a result of editing. Readers should obtain

the VoR from the journal website shown below when it is published

to ensure accuracy of information. The authors are responsible for the

content of this Accepted Article.

To be cited as: Adv. Synth. Catal. 10.1002/adsc.202000604

Link to VoR: https://doi.org/10.1002/adsc.202000604

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

FULL PAPER

DOI: 10.1002/adsc.202000604

How Site-Directed Mutagenesis Boosted Selectivity of a

Promiscuous Enzyme

Pavlína Nekvasilová,^a,b,cNatalia Kulik,^dNikola Rychlá,^{a, e}Helena Pelantová,^aLucie

Petrásková,^aZuzana Bosáková,^cJosef Cvačka,^fKristýna Slámová,^aVladimír Křen,^aand

Pavla Bojarová^a,e,*

a

Institute of Microbiology, Czech Academy of Sciences, Vídeňská 1083, CZ-14220 Praha 4, Czech Republic

Phone: (+420) 296442360; Fax: (+420) 244471286; E-mail: bojarova@biomed.cas.cz

Department of Genetics and Microbiology, Faculty of Science, Charles University, Viničná 5, CZ-12843, Praha 2,

b

Czech Republic

c

Department of Analytical Chemistry, Faculty of Science, Charles University, Hlavova 2030/8. CZ-12843 Praha 2,

Czech Republic

d

Center for Nanobiology and Structural Biology, Institute of Microbiology, Czech Academy of Sciences, Zámek 136,

CZ-37333 Nové Hrady, Czech Republic

Department of Health Care Disciplines and Population Protection, Faculty of Biomedical Engineering, Czech

e

Technical University in Prague, Nám. Sítná 3105, CZ-27201 Kladno, Czech Republic

Institute of Organic Chemistry and Biochemistry, Czech Academy of Sciences, Flemingovo nám. 2, CZ-16610 Praha

f

6, Czech Republic

Received: ((will be filled in by the editorial staff))

Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201######.((Please

delete if not appropriate))

Abstract. β-N-Acetylhexosaminidases (GH20; EC 3.2.1.52) expressed in Pichia pastoris, purified, and kinetically

are exo-glycosidases with a dual activity for cleaving both N- characterized. As a result, novel engineered enzymes with an

acetylglucosamine (GlcNAc) and N-acetylgalactosamine up to 7-times higher selectivity for either GalNAc or GlcNAc

(GalNAc) units from glycostructures. This substrate substrates were obtained, preserving the favorable properties

promiscuity is a hurdle in the selective synthesis of N- of the wild type TfHex, mainly its transglycosylation

acetylhexosamine oligosaccharides combining both GlcNAc potential and tolerance to functional groups in the substrate

and GalNAc units since there are hardly any GalNAc molecule. The substrate selectivity and transglycosylation

transferring enzymes available for synthetic applications. We yield were further corroborated by reaction engineering. The

present here site-directed mutagenesis of a synthetically new selective and synthetically capable enzymes were

potent promiscuous -N-acetylhexosaminidase from applied in the preparation of tailored N-acetylhexosamines.

Talaromyces flavus (TfHex), which, as a wild type, exhibits a

GalNAcase/GlcNAcase ratio of 1.2. On the basis of molecular

modeling, we identified crucial amino acid residues

responsible for its GalNAcase/GlcNAcase selectivity. Six

site-directed mutants were prepared, heterologously

Keywords: -N-acetylhexosaminidase; glycosylation;

Talaromyces flavus; Pichia pastoris; site-directed

mutagenesis; substrate specificity

may be fortified by reducing the water activity, adding

organic co-solvents^[3]or inorganic salts^[4]as well as

using a high concentration of the glycosyl acceptor.

-N-Acetylhexosaminidases utilize the substrate-

assisted mechanism with the key catalytic conserved

amino acid pair of Asp (stabilizing residue) and Glu

(catalytic acid/base).^[5]Furthermore, their active center

comprises tyrosine, proposed to stabilize the oxazoline

Introduction

Fungal β-N-acetylhexosaminidases (EC 3.2.1.52,

GH20, http://www.cazy.org/) are exo-glycosidases of

family GH20, which naturally catalyze the cleavage of

N-acetylglucosamine

(GlcNAc)

and

N-

acetylgalactosamine (GalNAc) moieties from

glycostructures.^[1]However, under suitable reaction

conditions, these enzymes can also catalyze the

formation of the glycosidic bond in vitro.^[2]This ability

reaction

intermediate.^[6]

A

model

β-N-

acetylhexosaminidase from Talaromyces flavus

1

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

(TfHex), well known for its broad substrate specificity

and high transglycosylation capability,^[7-9]was studied

for the effect of tyrosine substitution on the

suppression of hydrolytic activity^[10]and on the

GalNAcase/GlcNAcase activity ratio.^[11]This enzyme

as a wild-type (WT) is a typical promiscuous -N-

acetylhexosaminidase,

GalNAcase/GlcNAcase activity ratio of 1.2.^[11]In

general, the specificity of fungal -N-

acetylhexosaminidases varies from relatively selective

GlcNAcases (enzymes from Aspergillus versicolor

(AvHex) with GalNAcase/GlcNAcase ratio 0.09-

0.11^[12]or from Aspergillus oryzae (AoHex) with 0.3-

0.6)^[13]to slightly prevalent GalNAcases (e.g.,

Penicillium oxalicum -N-acetylhexosaminidase

PoHex with GalNAcase/GlcNAcase ratio of up to

1.7).^[14]Importantly, despite these notable changes in

the substrate specificity, all these enzymes share quite

a high homology. For example, TfHex exhibits a 62%

homology with the AoHex^[15]and 60% with PoHex.^[16]

The TfHex protein is a tetramer consisting of two non-

covalently associated subunits (each 65 kDa) and two

non-covalently linked propeptides (each 15 kDa),

accounting for the total of 160 kDa of the native

enzyme.^[17]

enzyme. Based on homology model of TfHex WT,^[27-

^28]we have found that Arg218 and Glu546 residues of

TfHex are involved in direct hydrogen bond interaction

with the substrate C4 hydroxyl (Figure 1), and thus are

mutation hotspots for altering the substrate specificity

of the enzyme both for hydrolysis and

transglycosylation. Prospective point mutations were

proposed based on a combination of in silico saturation

mutagenesis and visualization of enzyme-substrate

interactions in prospective mutants. A total of six

single mutant variants of TfHex were studied

featuring

the

concerning

their

specific

activity,

GalNAcase/GlcNAcase activity ratio and, first and

foremost, their applicability as synthetic tools in the

challenging

preparation

of

combined

N-

acetylhexosamine oligosaccharides.

Whereas β-N-acetylhexosaminidases with

a

naturally increased preference for GlcNAc substrates

occur among glycosidases, pure -GalNAcase activity

is quite rare in this class.^[12]Described WT N-

acetylgalactosaminyltransferases^[18-19]are generally

not applicable for the laboratory synthesis of

oligosaccharides as they are strictly specific for

particular acceptors (e.g., polypeptides), unlike rather

universal glycosidases. A more versatile GalNAc-

transferring activity was found in a mutant

galactosyltransferase from human placenta.^[20]The so

far described selective -GalNAcases, namely from

Paenibacillus,^[21]Clostridium,^[22]and Bacteroides^[23]

species, are of bacterial origin and have no

documented transglycosylation capabilities.^[24]Among

GlcNAcases, many representatives also exhibit purely

hydrolytic activity.^{[12, 25]}Moreover, even in the very

selective GlcNAcases, traces of GalNAcase activity

may be found (e.g., in O-GlcNAcase,^[26]the

GalNAcase/GlcNAcase ratio is 0.03). It is also

noteworthy that although the main source of

specificity of GalNAcase/GlcNAcase is encoded by

the primary amino acid sequence of the -N-

acetylhexosaminidase, other factors, not yet fully

understood, are undeniably involved, e.g., the effect of

posttranslational modifications or of heterologous

expression in a different host.^[11]

Figure 1. Molecular model of the active site of TfHex WT

[27-28]

with docked pNP-GlcNAc (magenta) after 5 ns of

molecular dynamics simulation. For details please see the

Experimental Section. Hydrogen bonds formed in the

snapshot are shown in yellow dashed lines, other possible

hydrogen bonds appearing with Glu546 and Arg218

residues during molecular dynamics are shown in yellow

dotted lines. The inset in the upper left corner shows the

percentage of snapshots during 2-20 ns of molecular

dynamics where hydrogen bonds were formed with amino

acid residues surrounding the C4 hydroxyl of pNP-GlcNAc

and pNP-GalNAc. The C4 atom is encircled in black,

hydrogens are hidden. The catalytic amino acid pair of

TfHex are Asp370 (stabilizing residue) and Glu371

(acid/base).

Results and Discussion

Heterologous Expression, Purification and

Characterization of TfHex Mutants

On the basis of molecular modelling described in

detail in the next section, six single-point mutant

variants of the TfHex enzyme were proposed, which

should feature altered GalNAcase/GlcNAcase activity

ratio. The genes of five TfHex mutant variants, namely

Glu546Tyr, Glu546Gln, Glu546His, Arg218His and

In the present work, original results were obtained

disclosing the impact of a single point mutation in the

amino acid sequence of TfHex on the

GalNAcase/GlcNAcase ratio of the respective mutant

2

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

Arg218Lys, were prepared commercially; the mutant

variant of the Arg218Gln was prepared by PCR

mutagenesis. The synthetic genes were cloned into the

pPICZαA vector, and electroporated into P. pastoris

competent cells KM71H. Expression vectors

contained EcoRI and KpnI restriction sites

downstream of the α-factor that encodes extracellular

targeting of the enzyme and resistance to zeocin. The

expression of enzymes was screened at a small scale.

For each mutant variant, sixteen colonies were

inoculated in the complex medium upon induction by

methanol. The mutant enzyme production by

individual colonies of transformed P. pastoris was

checked by SDS-PAGE and the enzyme activities

were determined by a standard activity assay. Selected

colonies were cryopreserved for later production. Two

of the prepared TfHex mutants, Glu546Tyr and

Arg218Gln, were successfully expressed and purified;

however, they were found to possess no enzymatic

activity. The other four mutant variants, Glu546Gln,

Glu546His, Arg218His, Arg218Lys TfHex, were

produced as active enzymes.

The preparative production of TfHex mutant

variants was performed in minimal cultivation

medium for easier purification. The extracellularly

produced enzymes were purified to homogeneity by

cation-exchange chromatography at pH 3.5 in a single

purification step. The purity of the enzymes was

confirmed by SDS-PAGE (see the Supporting

Information, Figure S2). The purification yield, the

GlcNAcase, and GalNAcase activities and their ratio

were determined and compared to the recombinant

wild-type TfHex (Table 1).

Table 1. Purification and yields of active TfHex mutant variants, and their specific GalNAcase and GlcNAcase activities

TfHex

variant

Yield ^b

[mg]

GlcNAcase

GalNAcase GalNAcase/GlcNAcase

[U mg^-1]

ratio

WT ^a

47 (58 %)

52 (71 %)

61 (66 %)

39 (52 %)

37 (52 %)

37

0.5

0.1

26

45

1.2

3.5

0.8

50

6.5

7.1

Glu546Gln

Glu546His

Arg218His

Arg218Lys

1.9

36

5.3

0.15

^aValues were adopted from Bojarova et. al.^[11]

^bProtein yield from purification of 200 mL of cultivation medium.

The found specific activities clearly show that the

enzyme GalNAcase/GlcNAcase ratio was strongly

influenced by the point mutation introduced. While the

wild-type strain exhibits a ratio of these activities of

1.2, in the Arg218Lys mutant this ratio significantly

decreased to 0.15, affording a GlcNAcase. In other

words, the introduction of Lys into the active center

instead of Arg at position 218 caused an eight-fold

increase in selectivity of the enzyme to pNP-GlcNAc

(1) whereas the specific activity remained comparable

to WT.

In contrast, by substituting Glu at position 546 for

either Gln or His, the selectivity to pNP-GalNAc

increased more than five times. Remarkably, both of

these mutant GalNAcases showed a significant

decrease in specific catalytic activity (13-times in the

case of Glu546Gln, 56-times in the case of Glu546His

TfHex). Apparently, the position of Glu546 is much

more sensitive to changes in the amino acid sequence;

this phenomenon was also observed in other functional

In the case of the last mutant variant, Arg218His

TfHex, the introduction of His into the active center in

position 218 caused only a slightly increased

GalNAcase/GlcNAcase ratio, yielding a weakly

specific GalNAcase. However, the Arg218His point

mutation caused that the enzyme activity decreased

completely during the week. This is in sharp contrast

to the other enzymes used in this work, WT or mutant,

which were active for more than one year at 4 °C.

Basic biochemical parameters were measured for each

of the four active mutant variants. pH Optima profiles

are shown in the Supporting information, Figure S3. In

the case of the selective variants Glu546Gln,

Glu546His, and Arg218Lys TfHex, we analyzed just

their major catalytic activity (GlcNAcase or

GalNAcase). The selective mutant variants had their

activity optimum at pH 5, which falls within the pH

optima plateau found for TfHex WT.^{[11, 17]}For the

Arg218His mutant, a slightly pronounced GalNAcase

(ratio 1.9), the pH optimum of both GalNAcase and

GlcNAcase activities was flat in the range of 4.0-5.5.

Interestingly, the GalNAcase activity of this variant

was relatively stable under basic conditions (pH 7 to

12.5), contrary to its GlcNAcase activity, which

rapidly decreased from pH 7, and from pH 9 it was so

low that it fell behind the detection threshold of the

mutants

of

glycosidases,

namely

mutant

transglycosidases^[11]and glycosynthases.^[29]The

specific catalytic activity in both of these groups of

mutant enzymes is drastically reduced and, to perform

the desired catalytic reaction, high protein amounts

need to be used.

3

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

spectrophotometer. This phenomenon caused a rapid

increase in the GalNAcase/GlcNAcase ratio of

Arg218His TfHex at a higher pH. This phenomenon

was observed before with other TfHex variants^[11]and

probably relates to changed protonation states of the

key amino acid residues in the active site.

(65 °C).^[17]For the Glu546Gln and Arg218Lys TfHex

mutants, temperature optimum was 60 °C, and it was

55 °C for Glu546His TfHex. The Arg218His TfHex

reported a rather flat temperature optimum between

55 °C and 65 °C for GalNAcase and 45-60 °C for

GlcNAcase activities.

The dependence of the hydrolytic activity of all

mutant variants on the temperature is shown in the

Supporting Information, Figure S4. The graphs show

that the temperature optimum of the mutant TfHex

does not differ significantly from TfHex WT

Furthermore, Michaelis-Menten parameters of the

hydrolytic activity of all active mutants were acquired

(Table 2, and Figure S5 in the Supporting

Information).

Table 2. Kinetic parameters of pNP-GlcNAc (1) and pNP-GalNAc (2) hydrolysis by mutant TfHex variants

TfHex

variant

k_cat/K_M

Activity

K_M[mM]

k_cat[s^-1]

[s^-1mM^-1]

GlcNAcase

GalNAcase

GlcNAcase

GalNAcase

GlcNAcase

0.11 ± 0.02

0.69 ± 0.09

0.8 ± 0.2

47 ± 1

104 ± 5

4.9 ± 0.3

0.6 ± 0.1

12 ± 1

434

150

6.1

1.6

3.6

3.1

32

WT ^a

Glu546Gln ^b

Glu546His ^b

0.39 ± 0.02

3.3 ± 0.5

Arg218His

2.1 ± 0.7

6.6 ± 0.9

29 ± 1

Arg218Lys ^b

^aValues were adopted from Bojarová et. al.^[11]

0.9 ± 0.2

^bKinetic parameters were determined only for major activity

The K_Mvalues indicate that in selective

GalNAcases Glu546Gln and Glu546His TfHex, the

affinity to pNP-GalNAc remained either comparable

to WT or even increased. This was not the case of

Arg218 mutants, featuring in all cases a decrease in

affinity to the respective substrates (higher K_M). In the

Arg218Lys GlcNAcase, the catalytic efficiency

(k_cat/K_M) decreased approximately 13-times compared

to WT. In Glu546Gln and Glu546His GalNAcases, the

catalytic efficiency was reduced even more

considerably, by 25-times and 100-times, respectively

In the Arg218His mutants, we found 48-times

decreased catalytic efficiency for pNP-GalNAc. This

drastic reduction in catalytic activity was in all cases

primarily caused by a considerable decrease in the

turnover number (k_cat). Thus, despite a rather strong

binding to the active site, further catalytic steps such

as formation of the oxazoline reaction intermediate

and product release are altered through the mutation.

Due to the high yield and cost-effectiveness of the

recombinant enzyme production in P. pastoris and its

facile downstream processing compared to other

available eukaryotic hosts, the decrease in specific

activity is not a major problem for any synthetic

applications since larger amounts of enzyme can easily

be added to the reaction to compensate for this

problem. As already mentioned above, this

phenomenon is quite usual in functional glycosidase

mutants

such

as

transglycosidases

or

glycosynthases.^[20]Supression of the undesired activity

rather than enhancement of the desired activity is a

general weakness of the rational engineering. Possibly,

an alternative may be dosed random mutagenesis

around the binding site to optimize beneficial

structure-function relations.

Docking and Molecular Dynamics Simulations of

TfHex Mutants

To investigate the structural properties and binding

preferences of the prepared TfHex mutant variants, we

performed docking of pNP-GlcNAc or pNP-GalNAc

substrates into the active sites of energetically

minimized models of all variants. The respective

molecular models were prepared on the basis from our

established molecular model of TfHex WT^[27]

corrected with respect of the crystal structure of β-N-

acetylhexosaminidase from A. oryzae ^[28](for details

see the Experimental Section). Averaged binding

scores calculated for equilibrated enzyme-substrate

complexes from molecular dynamics simulations are

shown in Table 3.

4

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

Table 3. Averaged binding scores of substrates in the active site of prepared mutant variants (kJ mol^-1), calculated for

equilibrated complexes after molecular dynamics simulation.

Variant ^a

pNP-GlcNAc

pNP-GalNAc

Score ratio

pNP-GalNAc/ pNP-GlcNAc

WT ^b

-38.6 (-23.5)

-35.4

-34.9 (-29.5)

-38.2

0.9

1.08

1.17

1.06

0.88

1.09

1.18

Glu546Gln

Glu546His

Arg218His

Arg218Lys

Glu546Tyr ^c

Arg218Gln

-34.5

-40.3

-30.6

-32.3

-36.4

-32.0

-35.2

-38.4

-27.0

-31.9

^aMutant GalNAcases are in green, mutant GlcNAcases in blue, inactive mutants in grey.

^bValues were adopted from Bojarová et. al.^[11]The numbers in italics indicate the binding score of the respective

hydrolytic product (GlcNAc or GalNAc, respectively).

^cAfter structure minimization, the Tyr546 residue was forced to move away from the ligand to fit it in a position that enabled

docking by induced fit. However, this movement led to the destruction of many interactions found in WT. In vivo, docking

is likely to be impossible as could be seen from molecular dynamics simulation of mutated enzyme without ligand.

The binding score calculated by Glide XP

algorithm characterizes the affinity of enzyme to the

substrate. The score ratio may be used as an indication

of enzyme discrimination between the substrates.

From the ratio we could see that WT and Arg218Lys

TfHex have a higher affinity (i.e., lower binding score)

to pNP-GlcNAc, which is in agreement with lower K_M

value for pNP-GlcNAc in WT. In contrast, the other

active variants (Arg218His, Glu546Gln, and

Glu546His TfHex) prefer pNP-GalNAc, which also

correlates with the kinetic results (Table 2, and Figure

S5 in the Supporting Information). To fully assess the

structure-function relationship, the geometry of

substrate-enzyme complex should also be considered.

In the WT, Arg218 in the active site forms

hydrogen bonds with C3/C4 hydroxyl of pNP-GalNAc

and pNP-GlcNAc (Figure 1). In the Arg218Lys TfHex

variant (GlcNAcase), interaction of C4 hydroxyl of

substrates with Lys218 is not possible due to the sp³

hybridization of nitrogen of Lys (Figure 2A, B). While

Lys218 in the Arg218Lys TfHex-pNP-GlcNAc

complex interacts with the C3 hydroxyl of the

substrate (Figure 2A), the axial C4 hydroxyl in pNP-

GalNAc sterically hinders the formation of this

interaction. As a result, Lys218 interacts with the

neighboring active site residue Glu546, which leads to

the decrease in the number of hydrogen bonds formed

with pNP-GalNAc in the Arg218Lys variant

(Supporting Information, Figure S6A) and increase in

the binding score (Table 3).

molecular dynamics simulation time). However, the

interaction of both substrates with Glu332 is

completely lost, leading to a higher binding score

(Table 3, Figure 2C). Even though pNP-GalNAc does

not form hydrogen bonding with His218 (Figure 2D),

pNP-GalNAc is better stabilized in the active site of

Arg218His variant due to the improved interaction

with Glu546. The observed instability of Arg218His

variant could result from a highly flexible interaction

with catalytic residues. The distance between Glu371

and the oxygen of the glycosidic bond often changed

(increased) during molecular dynamics simulation

(Supporting Information, Figure S6B), and catalytic

Asp370 formed hydrogen bond with the N-acetyl

group of the substrate in less than 30% of the

simulation time.

The docking of pNP-GalNAc and pNP-GlcNAc

in Arg218Gln variant (inactive) showed a slight

increase in binding scores for both substrates and a

decrease in the number of hydrogen bonds (Supporting

Information, Figure S6A). Moreover, the binding

score of substrates is comparable to the score of the

product of hydrolysis (Table 3).^[11-12]This may lead to

fast substrate unbinding. In the Arg218Gln complex

with both substrates (Supporting Information, Figure

S7), the substrates changed orientation with respect to

WT, and unusual interactions were formed with

Glu218. The main reason for enzyme inactivity may

be the uncompensated negative charge located close to

Gln218 in the active site, which leads to reorientation

of substrates and of catalytic Glu371, and to the

formation of unusual hydrogen bonding by protonated

Glu371 and Glu332 (Supporting Information, Figure

S7). The change in charge distribution in the active site

might influence deprotonation of the catalytic Glu371,

and impair proton donation to the glycosidic oxygen.

In the case of Arg218His TfHex variant

(unpronounced GalNAcase), the conformation of the

His218 side chain is different (nitrogen in sp²

hybridization). Thus, His218 can still make hydrogen

bonding with C4 of pNP-GlcNAc in the enzyme-

substrate complex (present during 52% of the

5

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

Figure 2. Orientation of docked substrates after 10 ns of molecular dynamics simulation. Active site amino acid residues,

which could form hydrogen bond interaction with the substrate, are shown. Tyr470 and Trp509 residues are hidden for

clarity. Amino acids and substrates are labeled in black; the probability of hydrogen bond formation with respective amino

acid residues during molecular dynamics is indicated in magenta. Position of C4 carbon in substrates was marked in a black

circle. (A), Arg218Lys TfHex-pNP-GlcNAc; (B), Arg218Lys TfHex-pNP-GalNAc; (C), Arg218His TfHex-pNP-GlcNAc;

(D), Arg218His TfHex-pNP-GalNAc; (E), Glu546His TfHex-pNP-GlcNAc; (F), Glu546His TfHex-pNP-GalNAc; (G),

Glu546Gln TfHex-pNP-GlcNAc; (H), Glu546Gln TfHex-pNP-GalNAc.

6

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

Another active-site residue close to the C4

only by induced fit because sterical hindrance with

Tyr546 disallowed substrate binding.

hydroxyl, prospective for influencing the enzyme

axial/equatorial selectivity at C4, is Glu546. In the WT

this residue interacts with C4 or C6 hydroxyls of pNP-

GlcNAc and with the C4 hydroxyl of pNP-GalNAc

(Figure 1). Substitution of Glu546 residue by His or

Gln led to a more favorable binding score compared to

WT with a clear preference for pNP-GalNAc substrate.

In the Glu546His variant (GalNAcase), His546

residue forms a hydrogen bond with pNP-GalNAc and

reinforces interaction with Glu332 (higher probability

of hydrogen bonding) by stabilizing the position of

Glu332 closer to the substrate (Figure 2F). In the

complex with pNP-GlcNAc, we observed interaction

between Glu371 and Glu332 similar to the Arg218Gln

variant (Supporting Information, Figure S7). This

interaction is not so frequent here but still occurs in

68 % of molecular dynamics snapshots and leads to a

partial inactivation of Glu371 for the glycosidic bond

cleavage. It occurred as a result of changes in

hydrogen bonding network: while Glu332 rotated and

started to interact with the substrate C6 hydroxyl,

His546 maintained its interaction with the substrate C4

hydroxyl but Arg218 completely lost interaction with

the substrate (Figure 2E).

Notably, in all mutants catalytic Asn370 interacts

not only with the N-acetyl group of the substrate but

also with its C3 hydroxyl (Figure 2B, E, G), which

might also result in decreased catalytic activity.^[30-31]

Transglycosylation Capability of Mutant

Enzymes

Besides the biochemical characterization of

prepared mutant enzymes, we aimed to address the

question of their transglycosylation potential. As

already mentioned, many native enzymes exhibiting a

high selectivity to either GlcNAc^[12]or GalNAc^[21-23]

moieties are useless for synthetic application due to the

lack of transglycosylating activity. Moreover, we

wanted to test if the remarkable substrate promiscuity

valued in TfHex WT, i.e. its ability to process

structurally modified substrates, had been maintained

in the mutants. For this aim, we chose the unique

transglycosylation reaction using GlcNAc azide

substrate, which required the enzymatic hydrolysis of

the C-N bond (Scheme 1). The Arg218Lys TfHex

mutant easily cleaved the azido substrate and

transferred the GlcNAc moiety to afford the

In the complex of Glu546Gln TfHex

(GalNAcase) with pNP-GalNAc, the substrate is

oriented closer to the catalytic Glu371 and it is

stabilized by a great number of hydrogen bonds,

including the hydrogen bond between the C4 hydroxyl

and Gln546. Notably, pNP-GalNAc changed

conformation from boat to semichair during molecular

dynamics (Figure 2H). In contrast, pNP-GlcNAc lost

interaction between its C4 hydroxyl and Gln546 and

with Arg218 due to the bulkier side chain in position

546 upon Glu546Gln mutation (Figure 2G). The

differences in hydrogen bond network influenced the

orientation pNP-GlcNAc with respect to Glu371,

leading to the formation of the inactivating hydrogen

bond between Glu371 and Glu332 in 3 % of molecular

dynamics snapshots, analogous to the Glu546His

variant (Figure 2G).

The modeling of the Glu546Tyr mutation

(inactive mutant variant) in the protein without

substrate followed by molecular dynamics simulation

showed that in the equilibrated model, Tyr546 could

be in two distinct binding modes. Either it interacted

with Trp509, occupying the place of the substrate C6

hydroxyl (mode A; Supporting Information, Figure

S8A), or it interacted with Glu332 (mode B;

Supporting Information, Figure S8B). This is probably

caused by the high flexibility of Glu546Tyr TfHex and

of its active site amino acid residues during molecular

dynamics (Supporting Information, Figure S9). The

orientation of Tyr546 before and after molecular

dynamics simulation allowed docking of substrate

corresponding disaccharide

in

27%

yield.

Interestingly, the reaction was fully regioselective in

contrast to the analogous reaction with TfHex WT,^[32]

with a comparable yield of the desired product (cf. the

previously published 32% yield with the WT).^[33]

Scheme 1. Enzymatic synthesis of functionalized

disaccharide

GlcNAcβ4GlcNAc-N₃

catalyzed

by

Arg218Lys TfHex.

We further hypothesized that the introduced

mutation may have impact on the transglycosylation

capability of the enzyme. For this aim, we analyzed the

formation of pNP-functionalized chitooligomers

(N,N´-diacetylchitobioside

to

N,N´,N´´,N´´´-

tetraacetylchitotetraoside) under the catalysis of

Arg218Lys TfHex mutant (Scheme 2) and compared it

to an analogous reaction catalyzed by TfHex WT. We

found that under the same reaction conditions, the

Arg218Lys TfHex mutant forms longer chains of pNP-

chitooligomers in higher yields. In total, we obtained

53 mg of pNP-chitooligomers in a ratio of 37: 23: 16

(5: 6: 7). The longer chain formation was previously

observed only with mutant transglycosidases.^[33]In

contrast, TfHex WT yielded small amounts of

chitobioside 5 and chitotrioside 6 (Supporting

7

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

Information, Fig. S10). Structures of all prepared

compounds were confirmed by NMR and MS

(Supporting information, Figs. S12-14, Tables S3-5).

to WT, the residual GlcNAcase activity still caused a

partial cleavage of the N,N'-diacetylchitobiose

acceptor and the formation of a mixture of products.

This is mainly because in glycosidase-catalyzed

synthesis, the acceptor is used in a large molar excess

(usually 5-10 times). Thus, to accomplish the desired

synthesis of a complex oligosaccharide, we adopted a

combined strategy employing the mutant enzyme

along with a straightforward selective protection of

N,N'-diacetylchitobiose acceptor at the C6 hydroxyl

with an acetyl moiety. Though the C6 acetylated

acceptor is still cleaved by the WT enzyme (with ca 6-

fold reduced affinity), which is not sufficient for a

selective transglycosylation reaction, the combination

of substrate protection and the mutant enzyme

Glu546Gln TfHex enabled to perform selective

GalNAc transfer to the intact acceptor in a good yield.

The influence of the 6-O-acetylation of the substrate

on the affinity of mutant Glu546Gln TfHex and the

WT enzyme is shown in the Supporting Information,

Table S10 and Figure S19.

The C6 protected glycosyl donor pNP-6-O-Ac-β-

D-GlcNAc (8, 99 % isolated yield) was prepared under

the catalysis by Novozym 435 lipase.^[34]It was then

used as a glycosyl donor for the preparation of a

selectively protected disaccharide 10 under the

catalysis by the Arg218Lys TfHex mutant GlcNAcase

(23 % isolated yield). Then, disaccharide 10 was used

as an acceptor in the glycosylation by pNP-GalNAc

donor under the catalysis by Glu546Gln TfHex,

affording the 6ˈ-O-acetylated trisaccharide 11 (28 %

isolated yield), which was finally deprotected by

ammonium hydroxide in methanol (1:10) to yield the

Scheme 2. Enzymatic synthesis of functionalized

chitooligomers: p-nitrophenyl N,N´-diacetylchitobioside

(GlcNAcβ4GlcNAc-OpNP),

N,N´,N´´

-

triacetylchitotrioside

(GlcNAcβ4GlcNAcβ4GlcNAc-

OpNP), and N,N´,N´´,N´´´ -tetraacetylchitotetraoside

(GlcNAcβ4GlcNAcβ4GlcNAcβ4GlcNAc-OpNP)

catalyzed by Arg218Lys TfHex.

Chemoenzymatic

Trisaccharides

Synthesis

of

Complex

We employed the prepared selective mutant

enzyme variants for the synthesis of complex N-

acetylhexosamine oligosaccharides to demonstrate

their utility in this challenging synthetic task. p-

Nitrophenyl glycosides are quite popular donors in

preparative transglycosylation reactions. First, they

are much cheaper than respective natural substrates

(e.g., N,N´-diacetylchitobiose). Secondly, they enable

a more efficient separation of transglycosylation

products from the remaining donor, e.g., by extraction

of the aromatic portion on XAD-resin or into ethyl

acetate. Our original intention to perform a selective

GalNAc

transfer

to

unprotected

N,N'-

diacetylchitobiose (GlcNAcβ4GlcNAc) with one of

the new mutant GalNAcases, however, proved to be

too challenging. Although the GalNAcase activity was

strongly (more than 6-times) increased in both

Glu546Gln and Glu546His mutant variants compared

desired

complex

trisaccharide

GalNAc4GlcNAc4GlcNAc (12, 99 % isolated

yield). The synthetic route is depicted in Scheme 3.

8

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

Scheme 3. Chemoenzymatic synthesis of the complex trisaccharide GalNAcβ4GlcNAcβ4GlcNAc catalyzed by a cascade

reaction employing Novozym 435 lipase and two selective mutant TfHex variants.

Conclusion

Preparation and Cloning of Mutant Genes

By means of site-directed mutagenesis based on

The pPICZαA expression constructs carrying the genes

of mutant TfHex (Glu546Tyr, Glu546Gln, Glu546His,

rational design following from in silico modeling, we

were able to considerably enhance the substrate

Arg218His,

Arg218Lys

TfHex)

were

prepared

specificity of TfHex, a typically promiscuous -N-

acetylhexosaminidase with a comparable affinity to

both GlcNAc- and GalNAc-terminated carbohydrates.

Out of the six single-point mutants prepared, we

identified two quite selective mutant GalNAcases and

one GlcNAcase, with ca 7-times higher selectivity for

either substrate. The enzymes were purified to

homogeneity, biochemically characterized, and

applied in a series of transglycosylation reactions to

demonstrate their high synthetic potential. The

enhanced selectivity of the prepared mutants, in

combination with their synthetic potential, afford

promising synthetic tools in the production of complex

-N-acetylhexosamines combining both GlcNAc and

GalNAc units. These compounds are otherwise

impossible to prepare by wild-type glycosidases.

commercially (Generay, China). The genes were cloned

downstream of the α-factor-encoding DNA segment for

extracellular protein targeting and zeocin resistance using

EcoRI and KpnI restriction sites. The construct containing

the gene of Arg218Gln TfHex was prepared by site-directed

mutagenesis using TfHex WT^[17]construct as a template

(GenBank ID: JN601495), and the pair of primers (Generi

Biotech,

Czech

Republic):

Fw:

5’-

gattgacacgggccaaaactttattactgt-3’; and

Re: 5’-acagtaataaagttttggcccgtgtcaatc-3’.

The PCR reaction was performed using NZYProof DNA

polymerase (NZYTech, Portugal) in a T-Personal Thermal

Cycler (Biometra, Germany) under the following

conditions: initial denaturation at 95 °C for 1 min, followed

by 35 cycles of denaturation at 95 °C for 30 s, primer

annealing at 59 °C for 1 min, elongation at 72 °C for 6 min

and final elongation at 72 °C for 10 min.

Then, 1 µL of DpnI (New England Biolabs, USA) was

added into the PCR mixture (50 µL), and the reaction

mixture was incubated at 37 °C and 300 rpm overnight to

cleave the methylated template. The PCR product was then

transformed into Escherichia coli Top10 competent cells

(Thermo Fisher Scientific, USA) by heat shock method,

under the selection pressure of zeocin (100 µg mL^-1). The

plasmids isolated from sixteen selected E. coli colonies

using High Pure Plasmid Isolation Kit (Roche, Switzerland)

were sequenced using the sequencing primer 5’-

gactggttccaattgacaagc-3’. The plasmids containing the gene

of Arg218Gln TfHex were isolated at a semi-preparative

Experimental Section

Materials

pNP-GlcNAc (1) and pNP-GalNAc (2) were obtained from

Gold Biotechnology, USA. GlcNAc was obtained from

Acros Organics, USA. Vinyl acetate was supplied by

Sigma-Aldrich, Czech Republic, and Novozym 435 by

Novozymes (Denmark). If not stated otherwise, all other

chemicals were from VWR Chemicals or Lach-Ner (Czech

Republic).

9

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

scale and purified using the High Pure Plasmid Midi

Isolation kit (Roche, DE) for transformation into P. pastoris

as described further.

purity of the enzyme was verified by SDS-PAGE, the

enzyme was concentrated by ultrafiltration in McIlvaine

buffer (50 mM citric acid/100 mM Na₂HPO₄, pH 5.0), and

stored at 4 °C.

Transformation of Mutant Genes

Protein Characterization

The yeast expression vectors pPICZαA carrying individual

genes of six TfHex mutant variants (Glu546Tyr, Glu546Gln,

Glu546Tyr, Arg218His, Arg218Lys, Arg218Gln TfHex)

were linearized (15 µg) using restriction endonuclease SacI

(New England Biolabs, USA). The vectors were then

electroporated into Pichia pastoris KM71H competent cells

(Invitrogen, Life Technologies, USA) using the standard

protocol by Invitrogen (EasySelect Pichia Expression Kit;

Invitrogen, Life Technologies, USA). The protein

expression and enzyme activity were screened at small-scale

production as described in the next section. Two to three

colonies showing a good production of mutant enzyme were

cryopreserved for long-term storage at -80 °C in 15% (v/v)

glycerol.

The purity of native and deglycosylated TfHex mutants was

verified by SDS-PAGE in 10% polyacrylamide gel. The

protein concentration was determined by Bradford

method^[35]using Protein Assay Dye Reagent Concentrate

(Bio-Rad, UK). Calibration was performed for bovine

plasma γ-globulin (Bio-Rad, UK). Deglycosylation of

TfHex variants was performed according to the

manufacturer´s protocol using endo-glycosidase H (EndoH;

New England Biolabs, USA). Each of the TfHex variants

(20 µg) was combined with 10× Glycoprotein Denaturating

Buffer (1 µl) supplied by the manufacturer and diluted with

water to a total reaction volume of 10 µL. Then the protein

was denatured by heating (100 °C; 10 min), mixed with 10×

GlycoBuffer 3 (2 µL), EndoH (2 µL; 1000 U) and water up

to a final volume of 20 µL and incubated overnight (37 °C,

300 rpm).

Heterologous Expression of β-N-Acetylhexosaminidase

in Pichia pastoris

Mutant TfHex variants were produced basically according

to the manufacturer’s instructions (EasySelect Pichia

Expression Kit; Invitrogen, Life Technologies, USA). For a

screening of the enzyme production in Pichia pastoris at a

small-scale, we used a combination of BMGY medium

(Buffered Glycerol complex Medium) and BMMY medium

(Buffered Methanol complex Medium). The transformants

(sixteen selected colonies from electroporation) were

inoculated into 100 mL of BMGY medium and incubated at

28 °C and 220 rpm overnight. Cultures with grown cells

were centrifuged (5000 rpm, 10 min, 12 °C) and the pellets

were resuspended in 30 mL of BMMY medium, where the

expression of the recombinant protein was induced by

methanol (0.5 % v/v). The cultures were shaken at 28 °C and

220 rpm for three days. Another batch of methanol was

added every 24 hours. On day 5, the cultures were tested for

the presence of TfHex by SDS-PAGE and for enzyme

activity.

Preparative production was performed in a combination

of BMGH medium (Buffered Glycerol Minimal Medium)

and BMMH medium (Buffered Methanol Minimal

Medium). The cryopreserved cultures (100 µL) were

inoculated into YPD medium (15 mL; 1 % yeast extract, 2 %

peptone, 2 % D-glucose) and incubated at 28 °C and 220

rpm for 5 hours. This preculture was inoculated into 1 L of

BMGH medium in 3 L Erlenmeyer flasks. On day 2, the

cells were collected by centrifugation (5000 rpm, 10 min,

and 12 °C) and resuspended in BMMH medium (200 mL)

in 1 L Erlenmeyer flask. The culture was shaken at 28 °C

and 220 rpm, the expression of the enzyme was induced by

methanol (0.5 % v/v) every 24 h.

Standard Assay for Enzyme Activity

The determination of the hydrolytic activity of β-N-

acetylhexosaminidases was performed using an end-point

assay with spectrophotometric detection. The substrates p-

nitrophenyl 2-acetamido-2-deoxy-β-D-glucopyranoside (1;

pNP-GlcNAc) and p-nitrophenyl 2-acetamido-2-deoxy-β-

D-galactopyranoside (2; pNP-GalNAc) were used. The

reaction mixture in microtubes containing 2 mM pNP-

GlcNAc (1) or pNP-GalNAc (2) substrate in McIlvaine

buffer (50 mM citric acid/100 mM Na₂HPO₄, pH 5.0) was

incubated at 35 °C in a thermoshaker, and reaction was

started by adding 10 µL of diluted enzyme. The reaction ran

for 10 min at 850 rpm and was quenched by adding 1 mL of

0.1 M sodium carbonate. The liberated p-nitrophenol,

cleaved off the respective substrate, occurs in the basic

medium in the form of the p-nitrophenolate anion that is

yellow and absorbance can be measured at 420 nm. The

measurement was performed in triplicate against a reference

sample containing only substrate and buffer. One unit of

enzymatic activity was defined as the amount of enzyme

that cleaved 1 mol of the respective substrate (GlcNAcase

activity for cleaving pNP-GlcNAc, GalNAcase activity for

cleaving pNP-GalNAc) per minute under the above

conditions.

The activity of TfHex WT and mutant Glu546Gln

TfHex with the protected donor p-nitrophenyl 6-O-acetyl-2-

acetamido-2-deoxy-β-D-glucopyranoside

(8)

was

performed analogously to the standard assay determination

as described above and was evaluated as a per cent value

related to the standard substrate pNP-GlcNAc (1); see also

Table S10 in the Supporting Information.

The pH optimum of TfHex was determined in the

universal Britton-Robinson buffer (0.04 M H₃PO₄, 0.04 M

phenylacetic acid, 0.04 M H₃BO₃/ 0.2 M NaOH), which is

used in the pH range of 2 to 12. The same buffer was used

for respective enzyme dilutions. The enzyme activity was

determined after 10 min at 35 °C and 850 rpm.

On day 5, the culture was centrifuged (5000 rpm, 10 min,

12 °C) and supernatant was purified to homogeneity by

-

cation exchange chromatography (Fractogel EMD SO₃,

Merck, D) on an Äkta Purifier system (Amersham

Biosciences, Uppsala, Sweden). The column was

equilibrated with 10 mM sodium citrate-phosphate buffer

pH 3.5. The elution of the enzyme was performed with a

linear gradient of 0-1 M NaCl (60 mL, 2 mL min^-1). The

10

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

The temperature optimum was determined in McIlvaine

product release and possible sterical conflicts in the active

site, which resulted in the final selection of mutations.

Single point mutations were introduced in YASARA.^[38]

Protonation states of active-site residues was assigned based

on Delphi^[39]and Propka^[40](from Schrödinger Suite 2018-

4, Build 12) prediction at pH 7; protonation state for

catalytic residues was assigned manually. The conformation

of Gln residue in mutants was selected based on the analysis

of rotameric library during initial docking by Glide. More

favorable conformation for substrate binding was selected.

The models of mutant variants were minimized in

YASARA to allow the exchanged residue to acquire

favorable orientation and to optimize contacts with

neighboring residues. Constrained minimization was done

in vacuo using NOVA force field,^[38]the last minimization

run was done in TIP3P water model, periodic boundary

condition, YASARA2 force field,^[41]NPT ensemble.

Substrate structures were built in YASARA and used for

docking with Glide (part of Schrödinger suite).^{[40, 42-43]}

Molecular dynamics simulations were run using the

substrate-enzyme complexes minimized in YASARA;

molecular dynamics was done in explicit TIP3P water at a

temperature of 298 K and pH 7 for 20 ns, using YASARA2

force field. Sodium ions were used for system neutralization,

Particle Mesh Ewald algorithm^[41]was employed for long-

range interaction modeling, and Berendsen thermostat was

applied for temperature control. Molecular dynamics

simulations were analyzed with YASARA tools.^[38]The

number of hydrogen bonds formed between substrate and

enzyme was calculated for a stable period of molecular

dynamics (2-20 ns), defined from root-mean square

deviations of the C atoms of proteins (data not shown). The

probability of hydrogen bond formation during molecular

dynamics was calculated as a number of snapshots with

formed hydrogen bonding versus total number of snapshots

and expressed as per cent value. The complexes for

calculation of docking scores were selected from

equilibrated molecular dynamics simulations (2-20 ns) with

the best parameters for hydrolysis of the glycosidic bond,

namely the smallest distance between Glu371 (OE2 atom)

and the substrate glycosidic oxygen, and the highest number

of hydrogen bonds. Five selected structures from molecular

dynamics simulation of each enzyme-substrate complex

were used for short minimization with Protein Preparation

Wizard (part of Schrödinger suite^[40]) with OPLS3 force

field^[42]in vacuo. Minimized structures were used for rigid

docking to calculate Glide XP binding scores.^[43]Final

reported binding score is calculated as an average value.

buffer (50 mM citric acid/100 mM Na₂HPO₄, pH 5.0) in

5 °C intervals from 20 °C to 80 °C or until the enzyme was

completely inactivated.

Michaelis-Menten parameters of hydrolysis of

substrates 1 and 2 (0.1-3.0 mM of 2 for both GalNAcases,

0.1-7.0 mM of 1 for Arg218His TfHex and 0.1-4.5 mM of 1

for Arg218Lys TfHex) were measured by discontinuous

kinetic assay. With fungal -N-acetylhexosaminidases,

which are generally acidophilic, the continuous assay shows

a low sensitivity due to the low proportion of the ionized p-

nitrophenolate in the reaction mixture The reaction mixture

(400 µL), consisting of substrate 1 or 2, McIlvaine buffer

pH 5.0 and TfHex mutant enzyme, was incubated at 25 °C

and 850 rpm for 7 min. Aliquots (50 µL) were taken every

1 min and the reaction was stopped by mixing with 1 M

sodium carbonate (100 µL) in a microplate. Then

absorbance at 420 nm was determined by Sunrise

Microplate Reader (Tecan, Switzerland). All data were

measured in triplicate. The kinetic parameters (K_M, k_cat)

were extracted by non-linear regression using GraphPad

Prism 7 (GraphPad Software, USA). Respective fitting

curves are depicted in Figure S5 in the Supporting

Information.

Molecular Modeling, Docking, and Molecular Dynamics

The molecular modeling of mutant variants was

accomplished starting from the molecular model of TfHex

WT^[27]and modified according to the resolved structure of

AoHex.^[28]Briefly,^[36]loops HL1 (residues 365-373), HL2

(residues 416-508) and the N-terminal propeptide were

remodeled by restrained modeling in MODELLER 9.16^[37]

and minimized in YASARA^[36]by standard protocol using

YASARA2 force field in TIP3P water. The dimer was

constructed by structural alignment in YASARA^[36]based

on the crystal of β-N-acetylhexosaminidase from A.

oryzae^[15]– (pdb code: 5oar). This model was used for

docking and molecular dynamics simulations previously.^[11,

36]

The mutation hotspots (Glu546 and Arg218) were

identified on the basis of the difference between interactions

of the WT enzyme with pNP-GalNAc and pNP-GlcNAc. In

order to predict prospective mutations at these positions, we

used in silico saturation mutagenesis. Both residues were

substituted with all possible amino acids and the change in

the binding energy for both pNP-GalNAc and pNP-GlcNAc

substrates upon the respective mutation was calculated in

YASARA (see the Supporting Information, Table S1).

Before energy calculation, the models of the mutants were

minimized in YASARA in water (by standard protocol,

YASARA2 force field, with fixed residues further than 0.8

nm from the mutated residue). The binding energy was

calculated by YASARA macro as in the MM/PBSA method

(without the entropy cost) as the energy of separated

compounds minus energy of their complex. As a parameter

for selecting the prospective mutant variants we used the

difference between the binding energies (similar to G) of

pNP-GalNAc and pNP-GlcNAc. Lower values correspond

to the increase in affinity to pNP-GalNAc over pNP-

GlcNAc compared to the WT. This pre-selection was

followed by analyzing the substrate interaction with the

theoretical mutants including the ability of glycan binding,

Analytical Methods

HPLC Analysis: HPLC analysis was used to monitor the

course of transglycosylation reactions. The Shimadzu

Prominence LC analytical system comprised Shimadzu

CBM-20A system controller, Shimadzu LC-20AD binary

HPLC pump, Shimadzu CTO-10AS column oven,

Shimadzu SIL-20ACHT cooling autosampler, and

Shimadzu SPD-20MA diode array detector (Shimadzu, JP).

Analyses were performed on a TSK gel Amide-80 column

(250  4.6 mm, 5 m) preceded by TSKgel Amide-80

Guardgel (3.2  15 mm, Tosoh corp., Japan) in acetonitrile/

water (4/1, v/v), with gradient elution as follows (A =

acetonitrile, B = water): 22 % B for 0-7 min, 22-31 % B for

11

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

7-16 min, 31-22 % B for 16-17 min, and 22 % B for 17-22

min for column equilibration; flow rate of 1 mL min⁻¹at

25 °C; injection volume 1 μL: Detection was performed at

200 nm.

mg) were incubated in 30% acetonitrile in McIlvaine buffer

pH 5 (total reaction volume 6 mL) at 35 °C and 1000 rpm.

After 2 h another portion of substrate 1 (180 mg, 0.5 mmol)

was added. The reaction progress was monitored by TLC

(propane-2-ol/water/NH₄OH aq., 7/2/1) and by HPLC. The

reaction was stopped after 5.5 h by boiling for 2 min.

Denatured enzyme was removed by centrifugation (13,500

rpm, 10 min) and the supernatant was loaded onto Biogel P-

2 column (26 × 1000 mm, Bio-Rad, UK) with water as a

mobile phase at an elution rate of 7.3 mL h^-1, ambient

temperature. Fractions containing the products were

collected and lyophilized. Traces of p-nitrophenol were

removed by extraction into ethyl acetate. The fractions of

products 5-7 still contained traces of non-substituted

oligosaccharides, which were removed by extraction on

Amberlite XAD-4 resin (BDH Chemicals, Ltd, UK). The

samples were dissolved in water, loaded onto XAD-resin

and after washing with water, pure pNP chitooligomers

were eluted with MeOH. Compounds 5-7 were obtained as

white solids: p-nitrophenyl 2-acetamido-2-deoxy-β-D-

ESI-MS Analysis: Mass spectra were measured using LTQ

Orbitrap XL hybrid mass spectrometer (Thermo Fisher

Scientific, USA) equipped with an electrospray ion source.

The mobile phase was methanol/water (4:1, v/v) at a flow

rate of 100 μL min^-1. The samples were dissolved in

methanol or methanol/ water and injected using a 5-μL loop

into the mobile phase flow. For the negative ion mode, spray

voltage, capillary voltage, tube lens voltage, and capillary

temperature were 5.0 kV, -25 V, -125 V, and 275 °C,

respectively. For the positive ion mode, spray voltage,

capillary voltage, tube lens voltage, and capillary

temperature were 5.0 kV, 9 V, 150 V, and 275 °C,

respectively. The spectra were recorded at a resolution of

100,000.

NMR Analysis: NMR spectra were acquired on a Bruker

Avance III 400 MHz (compound 8), 600 MHz (compounds

4, 10, 11, and 12), and 700 MHz (compounds 5, 6, and 7)

spectrometers (Bruker BioSpin, Rheinstetten, Germany) in

D₂O (100 atom % D, Sigma-Aldrich, Steinheim, Germany)

at 30 °C. Residual signal of D₂O (δ_H4.732 ppm) was used

for the proton spectra reference; carbon spectra were

referenced to the signal of acetone (δ_C30.50 ppm). The

individual monosaccharide units were assigned using

COSY, HSQC, 1d-TOCSY, and HSQC-TOCSY

experiments; their acetylation was proved by HMBC

experiment. Glycosidic linkage (1→4) was confirmed by

the HMBC correlation of carbon C4 with the adjacent

anomeric proton.

glucopyranosyl-

(1→4)-2-acetamido-2-deoxy-β-D-

glucopyranoside (5; 20 mg, 37 µmol), p-nitrophenyl 2-

acetamido-2-deoxy-β-D-glucopyranosyl-

(1→4)-2-

acetamido-2-deoxy-β-D-glucopyranoside (6; 17 mg, 23

µmol) and p-nitrophenyl 2-acetamido-2-deoxy-β-D-

glucopyranosyl-

(1→4)-

2-acetamido-2-deoxy-β-D-

(1→4)-2-acetamido-2-deoxy-β-D-

glucopyranoside (7; 16 mg, 16 µmol). For the structural

analysis, see the Supporting information (Tables S3-5,

Figures S12-14). For compound 5, []_D²⁰= −20.7 (c 0.237,

20

H₂O); for compound 6, []_D= −20.2 (c 0.233, H₂O); for

compound 7, []_D²⁰= −15.4 (c 0.136, H₂O/CH₃CN, 3/2).

Optical Rotation: Specific rotation of prepared compounds

was measured with an Autopol IV polarimeter (Rudolph

Research Analytical, USA, 2001). The []_Dvalues were

p-Nitrophenyl

6-O-acetyl-2-acetamido-2-deoxy-β-D-

20

glucopyranoside (8): The selective protection of the C6

hydroxyl was carried out according to Simerská et al.^[34]

pNP-GlcNAc (1, 200 mg, 0.6 mmol) was dissolved in

pyridine (15 mL) and then, acetone (21 mL) was added,

followed by vinyl acetate (7.1 mL) and a catalytic amount

of Novozym 435 (lipase from Candida antarctica

immobilized on acrylic resin). The reaction mixture was

filtered through a filter paper and co-evaporated with

toluene in vacuo to dryness. The title compound 8 was

obtained as a light brown solid (230 mg, 0.6 mmol, 99 %

recorded in water, water/ acetonitrile or DMSO according

to the solubility of compounds as specified further; the

concentration units are g mL^-1.

Preparative Synthetic Reactions

2-Acetamido-2-deoxy-β-D-glucopyranosyl-

acetamido-2-deoxy-β-D-glucopyranosyl

(1→4)-2-

azide (4):

GlcNAc-N₃(3, 183 mg, 0.7 mmol) and Arg218Lys TfHex

(6.5 µL, 4.5 U, 0.2 mg) were incubated in McIlvaine buffer

pH 5 (total reaction volume 1.5 mL) at 35 °C and 850

rpm. The reaction was monitored by TLC (propane-2-

ol/water/NH₄OH aq., 7/2/1) and HPLC. After 5 h the

reaction was stopped by boiling for 2 min. The reaction

mixture was centrifuged (13,500 rpm, 10 min) and the

supernatant was separated by gel permeation

chromatography using Biogel P-2 column (26 × 1000 mm,

Bio-Rad, UK) in water as a mobile phase at an elution rate

of 7.2 mL h^-1, ambient temperature. The fractions containing

the product were collected and lyophilized. The title

compound 4 was obtained as a white solid (24 mg, 53.4

µmol, 27 % isolated yield). For structural data of compound

4, see the Supporting information (Table S2, Figure S11).

[]_D²⁰= −20.6 (c 0.325, H₂O).

isolated yield). For the structural analysis, see the

20

Supporting information (Table S6, Figure S15). []_D

−29.8 (c 0.225, DMSO).

=

6'-O-Acetyl-2-acetamido-2-deoxy-β-D-glucopyranosyl-

(1→4)-2-acetamido-2-deoxy-D-glucopyranose

(10):

Compound 8 (48 mg, 0.1 mmol) as a donor, GlcNAc (9, 444

mg, 2.0 mmol) as an acceptor and Arg218Lys TfHex (98 µL,

40 U, 2.7 mg) were incubated in 30% acetonitrile in

McIlvaine buffer pH 5 (total reaction volume 4 mL) at 35 °C

and 850 rpm. The progress of the reaction was monitored by

TLC (propane-2-ol/water/NH₄OH aq., 7/2/1) and HPLC.

The reaction was stopped after 5 h by boiling for 2 min. The

mixture was centrifuged (13,500 rpm, 10 min) and the

supernatant

was

separated

by size

exclusion

pNP-Functionalized Chitooligomers 5-7: pNP-GlcNAc (1,

390 mg, 1.1 mmol) and TfHex Arg218Lys (13 µL, 6 U, 0.2

chromatography using Biogel P-2 column (26 × 1000 mm,

12

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

Bio-Rad, UK) and water as a mobile phase at an elution rate

of 7.3 mL h^-1, ambient temperature. The fractions containing

the product were combined and lyophilized. The title

compound 10 was obtained as a white solid (14 mg,

30 µmol, 23 % isolated yield). For the structural analysis,

see the Supporting information (Table S7, Figure S16).

[]_D²⁰= +11.1 (c 0.215, H₂O).

[7]

[8]

K. Slámová, R. Gažák, P. Bojarová, N. Kulik, R.

Ettrich, H. Pelantová, P. Sedmera, V. Křen,

Glycobiology 2010, 20, 1002-1009.

P. Bojarová, K. Slámová, K. Křenek, R. Gažák,

N. Kulik, R. Ettrich, H. Pelantová, M. Kuzma, S.

Riva, D. Adámek, K. Bezouška, V. Křen, Adv.

Synth. Catal. 2011, 353, 2409-2420.

P. Bojarová, K. Křenek, M. Kuzma, L.

Petrásková, K. Bezouška, D.-J. Namdjou, L.

Elling, V. Křen, J. Mol. Catal. B: Enzym. 2008,

50, 69-73.

[9]

2-Acetamido-2-deoxy-β-D-galactopyranosyl-(1→4)-2-

acetamido-2-deoxy-β-D-glucopyranosyl-(1→4)-2-

acetamido-2-deoxy-D-glucopyranose

(11):

pNP-

[10]

[11]

[12]

[13]

[14]

[15]

K. Slámová, J. Krejzová, P. Marhol, L.

Kalachova, N. Kulik, H. Pelantová, J. Cvačka, V.

Křen, Adv. Synth. Catal. 2015, 357, 1941-1950.

P. Bojarová, N. Kulik, M. Hovorková, K.

Slámová, H. Pelantová, V. Křen, Molecules 2019,

24, 599.

P. Bojarová, N. Kulik, K. Slámová, M. Hubálek,

M. Kotik, J. Cvačka, H. Pelantová, V. Křen, Appl.

Microbiol. Biotechnol. 2019, 103, 1737-1753.

L. Weignerová, P. Vavrušková, A. Pišvejcová, J.

Thiem, V. r. Křen, Carbohydr. Res. 2003, 338,

1003-1008.

P. Nekvasilová, I. Andreasová, L. Petrásková, H.

Pelantová, V. Křen, P. Bojarová, BBA - Proteins

Proteom. 2020, 1868, 140319.

R. Ettrich, V. Kopecký, K. Hofbauerová, V.

Baumruk, P. Novák, P. Pompach, P. Man, O.

Plíhal, M. Kutý, N. Kulik, J. Sklenář, H. Ryšlavá,

V. Křen, K. Bezouška, BMC Struct. Biol. 2007, 7,

32.

GalNAc (2, 12 mg, 35 µmol) as a donor, compound 10 (52

mg, 0.1 mmol) as an acceptor and Glu546Gln TfHex (18 µL,

2.8 U, 0.7 mg) were incubated in McIlvaine buffer pH 5

(total reaction volume 1.1 mL) at 35 °C and 850 rpm. The

reaction progress was monitored by TLC (propane-2-

ol/water/NH₄OH aq., 7/2/1) and HPLC. The reaction was

stopped after 8 hours by boiling for 2 min. The denatured

enzyme was removed by centrifugation (13,500 rpm, 10

min) and the supernatant was separated by size exclusion

chromatography. Biogel P-2 (26 × 1000 mm; Bio-Rad, UK)

was used as a stationary phase, and water as a mobile phase

at an elution rate of 7.5 mL h^-1. The 6´-O-acetylated

compound 11 was obtained as a white solid (6.6 mg, 10

µmol, 28 % isolated yield). For the structural analysis, see

20

the Supporting information (Table S8, Figure S17). []_D

=

+20.9 (c 0.215, H₂O). In the next step, it was deprotected by

ammonium hydroxide in methanol (1:10) to afford the title

compound 12 as a white solid (6.2 mg, 10 µmol, 99 %

isolated yield). For the structural analysis, see the

Supporting information (Table S9, Figure S18).

[16]

[17]

H. Ryšlavá, A. Kalendová, V. Doubnerová, P.

Skočdopol, V. Kumar, Z. Kukačka, P. Pompach,

O. Vaněk, K. Slámová, P. Bojarová, N. Kulik, R.

Ettrich, V. Křen, K. Bezouška, FEBS J. 2011,

278, 2469-2484.

K. Slámová, P. Bojarová, D. Gerstorferová, B.

Fliedrová, J. Hofmeisterová, M. Fiala, P.

Pompach, V. Křen, Protein Expression Purif.

2012, 82, 212-217.

K. Furukawa, K. Takamiya, K. Furukawa, BBA -

Gen. Subjects 2002, 1573, 356-362.

P. L. Deangelis, A. J. Padgett-Mccue, J. Biol.

Chem. 2000, 275, 24124-24129.

A. Šimonová, C. E. Kupper, S. Böcker, A. Müller,

K. Hofbauerová, H. Pelantová, L. Elling, V. Křen,

P. Bojarová, J. Mol. Catal. B: Enzym. 2014, 101,

47-55.

T. Sumida, K. Fujimoto, M. Ito, J. Biol. Chem.

2011, 286, 14065-14072.

I. Noach, B. Pluvinage, C. Laurie, K. T. Abe, M.

G. Alteen, D. J. Vocadlo, A. B. Boraston, J. Mol.

Biol. 2016, 428, 3253-3265.

C. Roth, M. Petricevic, A. John, E. D. Goddard-

Borger, G. J. Davies, S. J. Williams, Chem.

Commun. 2016, 52, 11096-11099.

Acknowledgments

Support by the project 20-00215S by the Czech Science

Foundation and by the mobility projects LTC18038 and LTC19035

by the Ministry of Education, Youth and Sports of the Czech

Republic and is gratefully acknowledged. Computational

resources are provided by the project "e-Infrastruktura CZ" (e-

INFRA LM2018140), a part of the program Projects of Large

Research, Development and Innovations Infrastructures. We thank

the group of Dr. S. Matějková, Institute of Organic Chemistry and

Biochemistry, Czech Academy of Sciences, for determining

specific rotations.

[18]

[19]

[20]

References

[1]

[2]

[3]

K. Slámová, P. Bojarová, L. Petrásková, V. Křen,

Biotechnol. Adv. 2010, 28, 682-693.

P. Bojarová, J. Bruthans, V. Křen, Appl.

Microbiol. Biotechnol. 2019, 103, 7869-7881.

P. Fialová, L. Weignerová, J. Rauvolfová, V.

Přikrylová, A. Pišvejcová, R. Ettrich, M. Kuzma,

P. Sedmera, V. Křen, Tetrahedron 2004, 60, 693-

701.

[21]

[22]

[4]

J. Rauvolfová, M. Kuzma, L. Weignerová, P.

Fialová, V. Přikrylová, A. Pišvejcová, M.

Macková, V. r. Křen, J. Mol. Catal. B: Enzym.

2004, 29, 233-239.

[23]

[24]

[25]

K. Slámová, P. Bojarová, BBA - Gen. Subjects

2017, 1861, 2070-2087.

M. S. Macauley, G. E. Whitworth, A. W.

Debowski, D. Chin, D. J. Vocadlo, J. Biol. Chem.

2005, 280, 25313-25322.

[5]

[6]

B. L. Mark, M. N. G. James, Can. J. Chem. 2002,

80, 1064-1074.

B. L. Mark, D. J. Vocadlo, S. Knapp, B. L.

Triggs-Raine, S. G. Withers, M. N. G. James, J.

Biol. Chem. 2001, 276, 10330-10337.

13

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

[26]

J. Krejzová, P. Šimon, L. Kalachova, N. Kulik, P.

[36]

C. Garcia-Oliva, P. Hoyos, L. Petrásková, N.

Kulik, H. Pelantová, A. H. Cabanillas, Á.

Rumbero, V. Křen, M. J. Hernáiz, P. Bojarová,

Int. J. Mol. Sci. 2019, 20, 6181.

Bojarová, P. Marhol, H. Pelantová, J. Cvačka, R.

Ettrich, K. Slámová, V. Křen, Molecules 2014,

19.

[27]

[28]

N. Kulik, K. Slámová, R. Ettrich, V. Křen, BMC

Bioinformatics 2015, 16, 28.

J. Škerlová, J. Bláha, P. Pachl, K. Hofbauerová,

Z. Kukačka, P. Man, P. Pompach, P. Novák, Z.

Otwinowski, J. Brynda, O. Vaněk, P. Řezáčová,

FEBS J. 2018, 285, 580-598.

Y.-W. Kim, H. Chen, S. G. Withers, Carbohydr.

Res. 2005, 340, 2735-2741.

S. J. Williams, B. L. Mark, D. J. Vocadlo, M. N.

G. James, S. G. Withers, J. Biol. Chem. 2002,

277, 40055-40065.

E. S. Feldblum, I. T. Arkin, Proc. Natl. Acad. Sci.

2014, 111, 4085.

P. Fialová, A. T. Carmona, I. Robina, R. Ettrich,

P. Sedmera, V. Přikrylová, L. Petrásková-

Hušáková, V. Křen, Tetrahedron Lett. 2005, 46,

8715-8718.

D. Laaf, P. Bojarová, B. Mikulová, H. Pelantová,

V. Křen, L. Elling, Adv. Synth. Catal. 2017, 359,

2101-2108.

P. Simerská, A. Pišvejcová, M. Kuzma, P.

Sedmera, V. r. Křen, S. Nicotra, S. Riva, J. Mol.

Catal. B: Enzym. 2004, 29, 219-225.

M. M. Bradford, Anal. Biochem. 1976, 72, 248-

254.

[37]

[38]

[39]

[40]

A. Šali, T. L. Blundell, J. Mol. Biol. 1993, 234,

779-815.

E. Krieger, G. Koraimann, G. Vriend, Proteins:

Struct., Funct., Bioinf. 2002, 47, 393-402.

L. Wang, M. Zhang, E. Alexov, Bioinformatics

2015, 32, 614-615.

G. Madhavi Sastry, M. Adzhigirey, T. Day, R.

Annabhimoju, W. Sherman, J. Comput. Aided

Mol. Des. 2013, 27, 221-234.

U. Essmann, L. Perera, M. L. Berkowitz, T.

Darden, H. Lee, L. G. Pedersen, J. Chem. Phys.

1995, 103, 8577-8593.

E. Harder, W. Damm, J. Maple, C. Wu, M.

Reboul, J. Y. Xiang, L. Wang, D. Lupyan, M. K.

Dahlgren, J. L. Knight, J. W. Kaus, D. S. Cerutti,

G. Krilov, W. L. Jorgensen, R. Abel, R. A.

Friesner, J. Chem. Theory Comput. 2016, 12, 281-

296.

R. A. Friesner, R. B. Murphy, M. P. Repasky, L.

L. Frye, J. R. Greenwood, T. A. Halgren, P. C.

Sanschagrin, D. T. Mainz, J. Med. Chem. 2006,

49, 6177-6196.

[29]

[30]

[41]

[42]

[31]

[32]

[33]

[34]

[35]

[43]

14

10.1002/adsc.202000604

Advanced Synthesis & Catalysis

FULL PAPER

How Site-Directed Mutagenesis Boosted Selectivity

of a Promiscuous Enzyme.

Adv. Synth. Catal. Year, Volume, Page – Page

Pavlína Nekvasilová, Natalia Kulik, Nikola Rychlá,

Helena Pelantová, Lucie Petrásková, Zuzana

Bosáková, Josef Cvačka, Kristýna Slámová,

Vladimír Křen, Pavla Bojarová*

15

Article Doi

How Site-Directed Mutagenesis Boosted Selectivity of a Promiscuous Enzyme

DOI: 10.1002/adsc.202000604

Source and publish data:

Authors:

Article abstract of DOI:10.1002/adsc.202000604

Full text of DOI:10.1002/adsc.202000604

Products guided by the article

R&D Labs maybe for 7284-16-4

Relevant to this article

Hot Product