Rutinosidase from Aspergillus niger: crystal structure and insight into the enzymatic activity

Article abstract of DOI:10.1111/febs.15208

Rutinosidases (α-l-rhamnosyl-β-d-glucosidases) catalyze the cleavage of the glycosidic bond between the aglycone and the disaccharide rutinose (α-l-rhamnopyranosyl-(1→6)-β-d-glucopyranose) of specific flavonoid glycosides such as rutin (quercetin 3-O-rutinoside). Microbial rutinosidases are part of the rutin catabolic pathway, enabling the microorganism to utilize rutin and related plant phenolic glycosides. Here, we report the first three-dimensional structure of a rutinosidase determined at 1.27-? resolution. The rutinosidase from Aspergillus?niger K2 (AnRut), a member of glycoside hydrolase family GH-5, subfamily 23, was heterologously produced in Pichia?pastoris. The X-ray structure of AnRut is represented by a distorted (β/α)₈ barrel fold with its closest structural homologue being an exo-β-(1,3)-glucanase from Candida?albicans (CaExg). The catalytic site is located in a deep pocket with a striking structural similarity to CaExg. However, the entrance to the active site of AnRut has been found to be different from that of CaExg – a mostly unstructured section of ~?40 residues present in CaExg is missing in AnRut, whereas an additional loop of 13 amino acids partially covers the active site of AnRut. NMR analysis of reaction products provided clear evidence for a retaining reaction mechanism of AnRut. Unexpectedly, quercetin 3-O-glucoside was found to be a better substrate than rutin, and thus, AnRut cannot be considered a typical diglycosidase. Mutational analysis of conserved active site residues in combination with in?silico modeling allowed identification of essential interactions for enzyme activity and helped to reveal further details of substrate binding. The protein sequence of AnRut has been revised. Databases: The nucleotide sequence of the rutinosidase-encoding gene is available in the GenBank database under the accession number MN393234. Structural data are available in the PDB database under the accession number 6I1A. Enzyme: α-l-Rhamnosyl-β-d-glucosidase (EC?3.2.1.168).

Full text of DOI:10.1111/febs.15208

Rutinosidase from Aspergillus niger: Crystal structure and insight into the
enzymatic activity
1
2
1,3
4
2
Petr Pachl , Jana Kapešová , Jiří Brynda , Lada Biedermannová , Helena Pelantová , Pavla
2
2
1,3
2
Bojarová , Vladimír Křen , Pavlína Řezáčová , Michael Kotik
1
Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences, Prague, Czech
Republic
2
3
Institute of Microbiology of the Czech Academy of Sciences, Prague, Czech Republic
Institute of Molecular Genetics of the Czech Academy of Sciences, Prague, Czech Republic
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
1
0.1111/FEBS.15208
This article is protected by copyright. All rights reserved

4
Institute of Biotechnology of the Czech Academy of Sciences, BIOCEV, Vestec, Czech Republic
Running Title: Crystal structure of rutinosidase from Aspergillus niger
Keywords
diglycosidase; X-ray crystallography; SIRAS; catalytic mechanism; rutin
Abbreviations
AnRut, rutinosidase from Aspergillus niger K2; Exg, exo-β-(1,3)-glucanase; CaExg, exo-β-(1,3)-
glucanase from Candida albicans; ScExg, exo-β-(1,3)-glucanase from Saccharomyces cerevisiae
Correspondence
M. Kotik, Institute of Microbiology of the Czech Academy of Sciences, Vídeňská 1083, Prague 4,
1
4220, Czech Republic
Tel.: +420 2 9644 2569
E-mail: kotik@biomed.cas.cz
and
P. Řezáčová, Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences,
Flemingovo nám. 2, Prague 6, 16610, Czech Republic
Tel.: +420 2 20 183 144
E-mail: rezacova@uochb.cas.cz
Petr Pachl and Jana Kapešová contributed equally to this article.
Enzyme
α-L-Rhamnosyl-β-D-glucosidase (EC 3.2.1.168)
This article is protected by copyright. All rights reserved

Databases
The nucleotide sequence of the rutinosidase-encoding gene is available in the GenBank database
under the accession number MN393234. Structural data are available in the PDB database under the
accession number 6I1A.
ABSTRACT
Rutinosidases (α-L-rhamnosyl-β-D-glucosidases) catalyze the cleavage of the glycosidic bond
between the aglycone and the disaccharide rutinose (α-L-rhamnopyranosyl-(1→6)-β-D-
glucopyranose) of specific flavonoid glycosides such as rutin (quercetin 3-O-rutinoside). Microbial
rutinosidases are part of the rutin catabolic pathway, enabling the microorganism to utilize rutin and
related plant phenolic glycosides. Here we report the first three-dimensional structure of a
rutinosidase determined at 1.27 Å resolution. The rutinosidase from Aspergillus niger K2 (AnRut), a
member of glycoside hydrolase family GH-5, subfamily 23, was heterologously produced in Pichia
pastoris. The X-ray structure of AnRut is represented by a distorted (β/α) barrel fold with its closest
8
structural homologue being an exo-β-(1,3)-glucanase from Candida albicans (CaExg). The catalytic
site is located in a deep pocket with a striking structural similarity to CaExg. However, the entrance
to the active site of AnRut has been found to be different from that of CaExg – a mostly unstructured
section of ~40 residues present in CaExg is missing in AnRut, whereas an additional loop of 13 amino
acids partially covers the active site of AnRut. NMR analysis of reaction products provided clear
evidence for a retaining reaction mechanism of AnRut. Unexpectedly, quercetin 3-O-glucoside was
found to be a better substrate than rutin, and thus AnRut cannot be considered a typical
diglycosidase. Mutational analysis of conserved active site residues in combination with in silico
modelling allowed identification of essential interactions for enzyme activity and helped to reveal
further details of substrate binding. The protein sequence of AnRut has been revised.
This article is protected by copyright. All rights reserved

Introduction
Diglycosidases are disaccharide-specific glycosidases that selectively hydrolyze the glycosidic linkage
between the aglycone and the disaccharide moiety of glycoconjugates [1]. In addition, diglycosidases
that release isoprimeverose units from oligoxyloglucans have been described [2]. A major subgroup
of diglycosidases is formed by α-L-rhamnosyl-β-D-glucosidases (rutinosidases), whose common
feature is their specificity towards diglycosides containing rutinose (α-L-rhamnopyranosyl-(1→6)-β-
D-glucopyranose). Microbial α-L-rhamnosyl-β-D-glucosidases are involved in the catabolic
breakdown of flavonol and flavanone (flavonoid) diglycosides such as rutin and hesperidin [3].
Flavonoid diglycosides are important secondary metabolites in various plants and fruits, many of
which are produced for human nutrition purposes [4,5]. The substrate specificities of α-L-rhamnosyl-
β-D-glucosidases were reported to be either rather narrow – preferentially reacting with 7-O-linked
flavonoid β-rutinosides such as hesperidin [6] – or broad. In the latter case also 3-O-linked flavonoid
β-rutinosides, e.g. rutin and kaempferol-3-O-rutinoside, were accepted as substrates [7], and these
enzymes are usually termed rutinosidases [1].
Rutinosidases catalyze the hydrolysis of rutin by cleaving the heterosidic linkage between the
disaccharide rutinose and the aglycon quercetin (Fig. 1). In addition, owing to their
transglycosylation activity, these enzymes are often capable of establishing a new glycosidic bond
between rutinose and a variety of aglycones [8]. These rutinose-containing transglycosylation
products may exhibit novel biological and biomedical activities for e.g. antiviral applications [9]. In
addition, the terminal rhamnose residue can change the pharmacokinetic properties of compounds
used as therapeutic agents [10], the reason being that rhamnose-capped molecules appear to be
resistant to hydrolysis in human tissues due to the absence of endogenous rutinosidases and α-L-
rhamnosidases, which would normally hydrolyze the respective glycosidic bonds [11]. Rutinosidases
are also interesting from a practical point of view because their substrates such as rutin and
hesperidin are inexpensive byproducts of the agro-processing and fruit juice-producing industry. The
use of the hydrolytic activity of rutinosidases includes taste and flavor improvement of beverages
such as citrus fruit juices and wine [1], and the production of rutinose and the aglycones quercetin
This article is protected by copyright. All rights reserved

and hesperetin for potential cosmetic, nutraceutical or biomedical applications. Furthermore,
rutinosylated compounds are a convenient source of the respective glucosides after trimming the α-
L-rhamnosyl moiety from the rutinoside using α-L-rhamnosidase [12]. The range of possible
applications of rutinosidases has recently been extended. Specifically, the preparation of quercetin
from rutin suspensions has been described using the rutinosidase from Aspergillus niger K2 (AnRut)
[
13]. Moreover, this enzyme has been shown to have a remarkable catalytic activity, affording
glycosyl esters in transglycosylation reactions with cinnamic acid derivatives as rutinose acceptors
14].
[
Most rutinosidases are members of the glycoside hydrolase family GH-5, subfamily 23. In this
work, we present the first X-ray structure of AnRut, the first structurally resolved representative of
this family. Furthermore, substrate specificity data, mutational analysis and structural modelling of
the substrate-enzyme complex have brought insight into the substrate binding interactions and
mechanism of enzyme catalysis. The present results pave the way towards understanding the
structure-function relationships of these enzymes.
Results and discussion
Crystal structure of AnRut
For our structural studies, we employed recombinant AnRut, which has a molecular weight of
approximately 60 kDa including posttranslationally attached carbohydrates or 42 kDa after
deglycosylation [8]. The use of Pichia pastoris as a heterologous expression host enabled the
secretion of AnRut to the growth medium after Kex2/Ste13-mediated cleavage of the engineered N-
terminal α-factor signal peptide from the fusion protein. Recombinant AnRut is a monomeric protein
and after secretion it contains three additional amino acids at the N-terminus compared to wild-type
AnRut without its native signal peptide. The residues in the crystal structure (29–386) were
numbered in accordance with the translated nucleotide sequence of native AnRut from Aspergillus
niger.
This article is protected by copyright. All rights reserved

Both glycosylated and enzymatically deglycosylated AnRut afforded crystals suitable for
diffraction analysis. Glycosylated AnRut crystallized in tetragonal space group and diffracted to
maximal resolution of 1.7 Å (Table 1). These crystals were used for solving the AnRut structure by
the SIRAS method taking advantage of the anomalous signal of iodine after soaking the protein
crystals in sodium iodide. The crystals of deglycosylated recombinant AnRut diffracted to 1.27 Å
resolution, and the dataset of the native protein was used for structure refinement (Table 1). The
crystal lattice system was primitive orthorhombic, and the crystal contained 44% solvent and two
protein molecules in the asymmetric unit. The residues 29–386 of both monomers A and B were
modelled into the electron density map, whereas the residues 17–28 were found to be missing in
both monomers. Non-protein electron density was explained by 10 N-acetyl-β-D-glucosamine
moieties, 690 water molecules and 28 ethane-1,2-diol molecules. The high-resolution electron
density map (see Fig. 6A) revealed that two residues (L90 and H179) were different from the
previously published translated nucleotide sequence [8]. The amino acid sequence derived from the
crystal structure data was confirmed by resequencing of the gene incorporated into the plasmid
pPICZαA-RUT, which was used for the transformation of P. pastoris KM71H [8]. The revised
nucleotide sequence has been deposited in GenBank under the accession number MN393234.
The RMSD for superposition of the 352 C atoms of the two protein monomers in the
α
asymmetric unit was 0.17 Å, a typical value for independently solved structures of identical proteins
[
15]. The monomer A had a better resolved electron density map and a lower overall B-factor and
was therefore used for further detailed structural analysis.
AnRut has a globular structure consisting of eight β-strands and ten α-helices arranged in a
topology that forms an irregular (β/α) barrel fold. There, the residues between β-sheets 5 and 6 are
8
not arranged in an α-helix as in the canonical (β/α) barrel. Another irregularity has been found
8
between β-sheets 1 and 2 where four α-helices are located with short unstructured polypeptide
stretches between them (Fig. 2A,B). We observed electron density indicative of posttranslational
modification by glycans attached to five asparagine residues: N128, N176, N181, N214 and N361.
Because AnRut was treated with endoglycosidase H for partial deglycosylation prior to protein
This article is protected by copyright. All rights reserved

crystallization, only the first N-acetylglucosamine unit was modelled. The N-terminal region of AnRut
is stabilized through one disulfide bond, C72 to C79, connecting α-helix 2 with helix 3. A deep pocket
of highly acidic character is located at the center of the barrel (Fig. 2C). The pocket which forms the
active site of the enzyme is comprised of residues in loop regions that connect the C-terminal ends
of the β-sheets to the following α-helices.
A structural comparison with models deposited in the Protein Data Bank using the program
DALI [20] revealed that the overall architecture of AnRut resembles the structures of GH5 family
glycosidases, which are members of the GH-A clan and as such have a classical (β/α) barrel fold.
8
Despite a low sequence identity of ~26%, a quite high structural similarity between AnRut, CaExg
(PDB code 1CZ1 [21]) and ScExg (PDB code 1H4P [22]) was found with RMSD values of 2.0 Å for the
superposition of 325 C atoms (Fig. 3A).
α
So far, there has been only a single report of a β-diglycosidase structure, the plant β-
primeverosidase from Camellia sinensis, classified as a GH1 family glycosidase (PDB codes 3WG4,
3
WQ5, 3WQ6 [23]). Although both enzymes are functionally similar and share a (β/α) barrel fold,
8
they show no significant structural similarities, including their active sites (the RMSD value for
superposition of 127 C atoms is 16.77 Å).
α
Active site residues and catalysis
In the phylogenetic analysis, AnRut grouped together with other glycosidases of the GH5 family (Fig.
4
A), whose members are known to employ a two-step double-displacement retaining reaction
mechanism [24]. Before the structure of AnRut was determined, the active site residues could be
predicted based on their conservation in the sequence alignment of AnRut with known fungal
rutinosidases and CaExg, whose structure had been determined in the presence of substrate and
inhibitor molecules [21,25] (Fig. 4B). The key catalytic residues were found at invariant positions in
the alignment: E210 as the putative acid/base catalyst and E319 as the putative catalytic
nucleophile. A water molecule which might be involved in catalysis was identified within hydrogen
bonding distance from both catalytic residues (3.0 and 2.8 Å from E210 and E319, respectively).
This article is protected by copyright. All rights reserved

Moreover, its position overlapped with a possible catalytic water molecule in CaExg (Fig. 3B). The
residues in CaExg which are involved in laminaritriose binding interactions at subsite –1, either
directly or via interconnecting water molecules, appear to be conserved in AnRut as well: E50, R108,
H153, N209, and Y284.
The structural alignment of CaExg with our AnRut crystal structure indeed revealed a high
degree of structural similarity of the catalytic sites, including their –1 subsites, which in both cases
bind the glucosyl moieties of their respective substrates (i.e. rutin or laminaritriose; Fig. 3B). A
mutagenesis study was performed to find further evidence for essential interactions of selected
residues during catalysis (Table 2). Five out of the six enzyme variants included amino acid
substitutions of highly conserved sites (Fig. 4B). As a result, these variants exhibited a drastic loss in
specific activity, indicating the importance of these sites for proper substrate binding and catalysis.
Moreover, we demonstrated that the hydrolytic activity of the virtually inactive E210A and
E319A mutants could be partially rescued using the small external nucleophiles azide or formate
(Fig. 5A,B) [26]. This is in accordance with the identification of E210 and E319 as the acid/base
catalyst and the nucleophile, respectively.
The steady-state kinetic parameters of AnRut were determined for rutin and isoquercitrin
hydrolysis (Fig. 1; Table 3). Surprisingly, the glucoside isoquercitrin proved to be a better substrate
than rutin considering its significantly higher specificity constant k /K . On the other hand, AnRut
cat
M
exhibited no or minimal activity towards the monoglycosides pNP β-D-glucopyranoside (pNP-Glc)
and pNP α-L-rhamnopyranoside (pNP-Rha) (Table 4). Moreover, we determined a low specific
activity towards pNP β-rutinoside compared to rutin (Tables 2 and 4). Thus, the substrate specificity
of AnRut is broader than expected with both rutin and isoquercitrin being good substrates; however,
it appears that a good substrate needs to have an aglycone that is identical or structurally similar to
quercetin. In the presence of β-D-glucose, α-L-rhamnose or the hydrolysis product rutinose, no
substantial reduction of enzyme activity was detected using rutin as a substrate. A negligible
inhibitory effect of rutinose on the AnRut activity was also found for enzymatic conversions at very
high rutin concentrations of up to 0.5 M [13]. Thus, rutinose appears to interact only very weakly
with the active site of AnRut in contrast to rutin. Taking all these data together, we can conclude
This article is protected by copyright. All rights reserved

that the particular structure of the aglycone moiety of rutin and isoquercitrin appears to be
important for substrate binding to AnRut. This is in contrast to the above-mentioned glycone-specific
β-primeverosidase from Camellia sinensis with broad aglycone specificity [23].
Catalytic mechanism
The average distance between the carboxyl oxygens of E210 and E319 in the AnRut structure is 5.3
Å. This value is in accordance with typical average distances of ~5.5 Å between the general acid/base
and the nucleophile in retaining glycosidases [27]. Since AnRut is the first rutinosidase whose
structure has been determined, we aimed to verify the reaction mechanism by analyzing the
anomeric configuration of the formed reaction products in situ. The NMR data provided clear
evidence that β-rutinose was formed from the substrate pNP β-rutinoside when following the
reaction time course (Fig. 5C). Similarly, NMR spectroscopy confirmed the formation of β-rutinose
from rutin (possessing β anomeric configuration); however, those spectra are not shown due to their
inferior quality (lower signal-to-noise ratio caused by the very low solubility of rutin in water-based
solutions and partial overlap of anomeric signals with the residual signal of water). Thus, our results
obtained with AnRut are fully consistent with a classical reaction mechanism of retaining
glycosidases [28] (Fig. 5D).
Substrate binding
Initially, we attempted to determine the crystal structure of AnRut in complex with the substrate or
the product rutinose. However, despite extensive efforts we did not succeed in co-crystallizing the
E210A variant neither in complex with rutin nor rutinose. Therefore, bound rutin in the active site of
AnRut was modelled using the available structure of CaExg complexed with the substrate
laminaritriose as a starting point (PDB code 3N9K) [21,25]. The modelling made use of the high
structural similarity between the CaExg and AnRut active sites and of the fact that both bound
substrates, laminaritriose and rutin, have their glucose units bound at the –1 subsites. Taking into
account only the atoms of the glucose units, rutin was superimposed onto laminaritriose with the
coordinates of the corresponding glucose moieties fixed at the same positions. The final
This article is protected by copyright. All rights reserved

conformations of bound rutin and the selected active site residues were obtained by energy
minimization after manually adjusting the dihedral angle between glucose and the aglycone.
In the modelled AnRut-rutin complex (Fig. 6A,C), the catalytic acid/base E210 is positioned
slightly above the glucose ring plane with the shortest distance to the glycosidic oxygen of 3.0 Å
(Oε1). The rutin molecule is deeply buried in the active site pocket, whose entrance is narrowed by
two loops, one comprised of residues 214–227 (loop between β-sheet 4 and α-helix 7) and the other
of residues 284–290 (loop between β-sheet 9 and α-helix 9) as seen in Fig. 6B. Interestingly, these
loops mutually interact through residues V204 and H273 and partially cover the enzyme active site.
As a result, the active site has two openings connecting the catalytic site with the solvent.
Furthermore, we were able to identify polar, hydrophobic and π-stacking interactions in the
modelled complex. As visualized in Fig. 6C, polar interactions were found between the sidechains of
all three subsites and the Glc, Rha and aglycone moieties of bound rutin. On the other hand, it
appears that most hydrophobic and all π-stacking interactions are present only between the
aglycone and the AnRut sidechains. Indeed, the aglycone is clamped by four aromatic sidechains of
F221, F261, Y284 and Y286, building an extensive network of hydrophobic and π-stacking
interactions with all three ring structures of the bound quercetin moiety (Fig. 6D). Additional
hydrophobic interactions were established between L162, M218, and G222 and the aglycone. Thus,
based on the analysis of interactions in the model complex, the aglycone appears to substantially
contribute to the binding of rutin and most likely also isoquercitrin to the AnRut active site through
an extensive number of non-covalent interactions. The importance of the aglycone for a good
substrate conversion can be inferred from the activity data of pNP β-D-glucopyranoside,
isoquercitrin, pNP rutinoside and rutin. The former two compounds are β-D-glucopyranosides
differing only in their aglycone structures, as are the latter two rutinose-containing compounds. Out
of these four compounds only two are good substrates of AnRut: isoquercitrin and rutin, both having
an identical aglycone – quercetin (see Tables 2–4). Thus, from the activity data we can conclude that
the aglycone quercetin is instrumental in the substrate binding and conversion. In addition, the
kinetic data of isoquercitrin clearly indicate that the terminal rhamnosyl unit of rutin is not
important for substrate binding. Interestingly, Tribolo et al. came to a similar conclusion that the
This article is protected by copyright. All rights reserved

hydrophobic environment of the +1 subsite of human β-glucosidase, a member of GH1, significantly
contributes to the binding of flavonoid glucosides [29]. It is also worth noting that another
carbohydrate-processing enzyme – human glycogen phosphorylase – has a specific and high-affinity
binding site for quercetin, which acts as an inhibitor of the enzyme [30].
The determined X-ray structure may also explain the observed activity loss in the AnRut
variants R108L, H153Q and H282S (Fig. 6C) as these sites appear to be in close proximity to the
nucleophile E319 with the shortest distances of 3.3–3.6 Å between the sidechains. We can speculate
that upon mutation at these sites the proper spatial orientation of the nucleophile got distorted,
which led to largely inactive enzyme variants. According to the molecular model, the amino acid
exchange H282S may also affect substrate binding since the nitrogen atom δ of the imidazole ring
appears to be rather close (4.2 Å) to the hydroxyl group of C-2 of the glucosyl moiety, to the oxygen
atom at C-4 and to the hydroxyl at C-5 of the quercetin moiety of bound rutin. According to the X-ray
structure and the molecular model, there is no close interaction of H283 with the residues E210 and
E319 or the bound rutin molecule. This probably accounts for the substantial activity of the H283S
enzyme variant.
At this point, we can only speculate that the special shape of the entrance to the active site
and its specific interactions with the aglycone of the bound substrate may explain the high
transglycosylation capabilities of AnRut observed in various synthetic reactions with rutin as rutinose
donor and a broad range of acceptor molecules, which include also phenolic compounds [8].
Conclusion
Microbial α-L-rhamnosyl-β-D-glucosidases (rutinosidases) play important roles in the degradation of
plant-derived glycosides such as e.g. rutin and hesperidin. Their hydrolytic activities have potential
applications in food-processing and the manufacturing of functional foods and ingredients for
cosmetics. Moreover, their transglycosylation activities may be employed in biomedical applications
for the synthesis of tailored rutinose-containing glycoconjugates. In the present work, we have
determined the first three-dimensional structure of an α-L-rhamnosyl-β-D-glucosidase (rutinosidase
This article is protected by copyright. All rights reserved

from Aspergillus niger; AnRut) and identified the key residues for catalysis and substrate (rutin)
binding through mutational analysis and molecular modelling. The catalytic mechanism of AnRut was
confirmed to be retaining. AnRut exhibited quite a high overall structural similarity to CaExg with
striking similarities in their catalytic sites. However, prominent structural differences were found for
the entries to the corresponding active sites, which may account for the outstanding
transglycosylation capacities of AnRut. In addition, we have shown that the substrate specificity of
AnRut is broader than expected with the enzyme accepting not only the diglycoside rutin as a good
substrate but also the glucoside isoquercitrin. Our data indicate that the aglycone moiety of the
substrate is likely to determine the specificity of AnRut. The present work represents a solid basis for
future investigations into the structure-function relationships of AnRut and other GH5-23 enzymes.
Materials and methods
Activity assays
All activity measurements were performed in the presence of 10% (v/v) dimethyl sulfoxide in
McIlvaine buffer (pH 3.6) at 37 °C. Activity data with pNP rutinoside were determined
spectrophotometrically at 395 nm. The amount of quercetin formed in the hydrolysis reaction was
determined by HPLC [8]. All initial velocity data were acquired in triplicate. The Michaelis-Menten
parameters were determined by nonlinear regression using the web-based tool MyCurveFit. Activity
rescue experiments were conducted with 4 mM rutin and sodium formate and sodium azide
concentrations of 0–5 M. Rutinose was prepared from rutin using AnRut as a catalyst [13].
Isoquercitrin (quercetin 3-O-β-D-glucopyranoside) and p-nitrophenyl rutinoside were prepared as
previously described [31,8]. Protein concentration was determined using the Bradford method.
Protein preparation and crystallization
Recombinant AnRut was prepared by heterologous expression in Pichia pastoris and purified as
described previously [8]. Deglycosylation was performed using the endoglycosidase H (New England
Biolabs) according to the manufacturer's instructions. Deglycosylated AnRut was purified by size-
This article is protected by copyright. All rights reserved

exclusion chromatography on a Superdex 200 10/300 GL column (1.0 × 30 cm; 10 mM MES buffer,
pH 5.0, 150 mM NaCl). Both native and deglycosylated AnRut were transferred to 10 mM MES buffer
–1
(
pH 5.0), concentrated to 35 mg·mL and subjected to crystallization experiments at 18 °C. For
glycosylated AnRut, crystallization screening was performed in vapor diffusion trials using a JSCG+
screen with 400-nL drops (Qiagen) and an ORYX robot (Douglas Instruments). Microcrystals
appeared in the reservoir solution containing 0.2 M zinc acetate (pH 8.0), 0.1 M imidazole, and 20%
PEG 3000. Initial crystallization conditions were further optimized using a capillary method [32].
Crystals used for data collection were grown in single capillaries where protein solution and
precipitating solution (0.2 M zinc acetate (pH 8.0), 0.1 M imidazole, and 20% PEG 3000) were
separated by agarose gel (1.5%). To prepare heavy atom derivatives, crystals were removed from
the capillary and transferred to a drop containing precipitating solution supplemented with 0.1 M
sodium iodide. The crystals were left in the solution for 24 hours under paraffin oil in a microbatch
experiment plate before they were flash-cooled in liquid nitrogen without additional cryoprotection.
For deglycosylated AnRut, the Microseed Matrix Screening approach [33] was employed using
a JSCG+ screen (Qiagen) and an ORYX robot (Douglas Instruments). We added crystallization seed
stock made of microcrystalline precipitate which had been obtained in the first round of screening
with deglycosylated AnRut. Crystals were grown in precipitating solution containing 20 mM
ammonium sulfate, 0.1 M Bis-Tris (pH 5.5), and 25% PEG 3350. Crystallization conditions were
further optimized using the microbatch technique with seeding. Crystals used for data collection
were grown by mixing protein, the previously mentioned precipitating solution and seeds in a ratio
of 5:4:1 in 1.2-µL drops placed under paraffin oil. Crystals were flash-cooled in liquid nitrogen
without any further cryoprotection.
Diffraction data collection and structure determination
All diffraction data were collected at 100 K at the MX14.1 beamline of the BESSY electron storage
ring, Berlin, Germany [34]. The native data sets were collected at a wavelength of 0.918 Å whereas
the anomalous data sets were collected at a wavelength of 1.5 Å. Diffraction data were integrated,
reduced and scaled using XDS [35] with the help of two GUI programs, XDSAPP [36] and XDSGUI [37].
This article is protected by copyright. All rights reserved

Attempts to solve the structure by molecular replacement using the CaExg data (PDB code 1CZ1
21]) were not successful and thus phase information was obtained experimentally. The structure
[
was determined using the SIRAS method by scaling anomalous datasets from three crystals with the
SHELX package and the GUI HKL2MAP [38,39]. Initial model building was carried out using Buccaneer
from the CCP4 package [17] with manual rebuilding in Coot [40]. The structure of deglycosylated
AnRUT was solved by molecular replacement using MOLREP with a model of the previously
determined structure of glycosylated AnRut [41]. Refinement was carried out using the program
REFMAC 5.8 [42]. The quality of the final model was validated with MolProbity [19]. All figures
showing structural representations were prepared in the program PyMOL [43]. The crystal
parameters and data collection and refinement statistics are summarized in Table 1. The atomic
coordinates and experimental structure factors were deposited in the Protein Data Bank under
accession code 6I1A.
Mutagenesis
Site-directed mutagenesis was performed using either the Phusion site-directed mutagenesis kit
(ThermoScientific) or the QuickChange Lightning kit (Agilent Technologies). The design of the
mutations followed the guidelines for substitutions that are unlikely to create disturbances in the
protein fold [44]. The entire protein-encoding sequence was verified by sequencing both DNA
strands. The following forward and reverse mutagenic primers were used.
E319A-F: 5'-CCTGGATTGCCCATGCGCCCACAAACACG;
E319A-R: 5'-CGTGTTTGTGGGCGCATGGGCAATCCAGG;
E210A-F: 5'-GTTGTCGGCGGGTGCGTTGATGGGCTC;
E210A-R: 5'-GAGCCCATCAACGCACCCGCCGACAAC;
R108L-F: 5'-CATAGGTGGTGGGGATCAGCAGGACCGTGATACC;
R108L-R: 5'-GGTATCACGGTCCTGCTGATCCCCACCACCTATG;
H153Q-F: 5'-ATCGACACGCAATCCCTCCC;
H153Q-R: 5'-GATGATGTGCATGTTGTAAGTCTTGATG;
H282S-F: 5'-CATATGCTCGTAGTAATAATGACTCGTATCGAAGACGAGGTTCGTA;
This article is protected by copyright. All rights reserved

H282S-R: 5'-TACGAACCTCGTCTTCGATACGAGTCATTATTACTACGAGCATATG;
H283S-F: 5'-ATCCATATGCTCGTAGTAATAACTATGCGTATCGAAGACGAGGTTC;
H283S-R: 5'-GAACCTCGTCTTCGATACGCATAGTTATTACTACGAGCATATGGAT.
NMR spectroscopy
NMR spectra were recorded on a 600 MHz spectrometer (Bruker Avance III) in 35 mM citrate buffer
(
in D O, pH 3.6) at 30 °C. Hydrolysis reactions were performed in a volume of 600 μL with 1 mg of
2
rutin in the presence of dimethyl sulfoxide-d (33%, v/v) or 4.6 mg of pNP rutinoside. Undissolved
6
1
rutin was removed by centrifugation, and H NMR spectra were recorded over 8 h after adding 650
ng of AnRut. For the hydrolysis of pNP rutinoside, 220 μg of AnRut were used.
Molecular modelling
For the superposition of the AnRut and CaExg proteins as well as for the superposition of their
corresponding substrates – rutin or laminaritriose – the program Chimera was used [45]. For
superposition of the proteins we used the Needleman-Wunsch algorithm [46] with the BLOSUM-62
scoring matrix [47]. The laminaritriose atoms selected for superposition were BGC1001: O5, C1, C2,
C3, C4, C5, C6, O4, O3, O2, and BGC1002: O3; the corresponding rutin atoms were UNK1: O1, C9, C4,
C3, C2, C1, C10, O4, O5, O6, O7. For modelling of the rutin conformers we used the program Molden
[
48]. The entire AnRut-rutin complex was energy-minimized using the program YASARA [49].
This article is protected by copyright. All rights reserved

References
1
2
Mazzaferro LS & Breccia JD (2011) Functional and biotechnological insights into diglycosidases.
Biocatal Biotransform 29, 103–112.
Koseki T, Ishikawa M, Kawasaki M & Shiono Y (2018) β-Diglycosidases from microorganisms as
industrial biocatalysts: biochemical characteristics and potential applications. Appl Microbiol
Biotechnol 102, 8717–8723.
3
4
Tranchimand S, Brouant P & Iacazio G (2010) The rutin catabolic pathway with special emphasis on
quercetinase. Biodegradation 21, 833–859.
Peterson JJ, Beecher GR, Bhagwat SA, Dwyer JT, Gebhardt SE, Haytowitz DB & Holden JM (2006)
Flavanones in grapefruit, lemons, and limes: A compilation and review of the data from the analytical
literature. J Food Compos Anal 19, S74–S80.
5
6
Williamson G (2017) The role of polyphenols in modern nutrition. Nutr Bull 42, 226–235.
Neher BD, Mazzaferro LS, Kotik M, Oyhenart J, Halada P, Křen V & Breccia JD (2016) Bacteria as
source of diglycosidase activity: Actinoplanes missouriensis produces 6-O-α-L-rhamnosyl-β-D-
glucosidase active on flavonoids. Appl Microbiol Biotechnol 100, 3061–3070.
Ishikawa M, Kawasaki M, Shiono Y & Koseki T (2018) A novel Aspergillus oryzae diglycosidase that
hydrolyzes 6-O-α-L-rhamnosyl-β-D-glucoside from flavonoids. Appl Microbiol Biotechnol 102, 3193–
7
8
9
1
1
3
201.
Šimčíková D, Kotik M, Weignerová L, Halada P, Pelantová H, Adamcová K & Křen V (2014) α-L-
Rhamnosyl-β-D-glucosidase (rutinosidase) from Aspergillus niger: characterization and synthetic
potential of a novel diglycosidase. Adv Synth Catal 357, 107–117.
Katayama S, Ohno F, Yamauchi Y, Kato M, Makabe H & Nakamura S (2013) Enzymatic synthesis of
novel phenol acid rutinosides using rutinase and their antiviral activity in vitro. J Agric Food Chem 61,
6
917–9622.
0 Robinson MA, Charlton ST, Garnier P, Wang X, Davis SS, Perkins AC, Frier M, Duncan R, Savage TJ,
Wyatt DA, Watson SA & Davis BG (2004) LEAPT: lectin-directed enzyme-activated prodrug therapy.
Proc Natl Acad Sci USA 101, 14527–14532.
1 Knaup B, Kahle K, Erk T, Anagnostis V, Scheppach W, Schreier P & Richling E (2007) Human intestinal
hydrolysis of phenol glycosides – a study with quercetin and p-nitrophenol glycosides using ileostomy
fluid. Mol Nutr Food Res 51, 1423–1429.
This article is protected by copyright. All rights reserved

1
1
2 Bassanini I, Krejzová J, Panzeri W, Křen V, Monti D & Riva S (2017) A sustainable one-pot two-enzymes
synthesis of naturally occurring arylalkyl glucosides. ChemSusChem 10, 2040–2045.
3 Kapešová J, Petrásková L, Markošová K, Rebroš M, Kotik M, Bojarová P & Křen V (2019) Bioproduction
of quercetin and rutinose catalyzed by rutinosidase: novel concept of “solid state biocatalysis”. Int J
Mol Sci 20, 1112.
1
4 Bassanini I, Kapešová J, Petrásková L, Pelantová H, Markošová K, Rebroš M, Valentová K, Kotik M,
Káňová K, Bojarová P, Cvačka J, Turková L, Ferrandi EF, Bayout I, Riva S & Křen V (2019) Glycosidase-
catalyzed synthesis of glycosyl esters and phenolic glycosides of aromatic acids. Adv Synth Catal 361,
2
627–2637.
1
1
1
5 Betts MJ. & Sternberg MJE (1999) An analysis of conformational changes on protein–protein
association: implications for predictive docking. Protein Eng. 12, 271–283.
6 Diederichs K & Karplus PA (1997) Improved R-factors for diffraction data analysis in macromolecular
crystallography. Nat Struct Biol 4, 269–275.
7 Winn MD, Ballard CC, Cowtan KD, Dodson EJ, Emsley P, Evans PR, Keegan RM, Krissinel EB, Leslie
AGW, McCoy A, McNicholas SJ, Murshudov GN, Pannu NS, Potterton EA, Powell HR, Read RJ, Vagin A
&
Wilson KS (2011) Overview of the CCP4 suite and current developments. Acta Crystallogr Sect D-
Biol Crystallogr 67, 235–242.
1
1
8 Brünger AT (1992) Free R value: A novel statistical quantity for assessing the accuracy of crystal
structures. Nature 355, 472–475.
9 Williams CJ, Headd JJ., Moriarty NW, Prisant MG, Videau LL, Deis LN, Verma V, Keedy DA, Hintze BJ,
Chen VB, Jain S, Lewis SM, Arendall WB., Snoeyink J, Adams PD, Lovell SC, Richardson JS & Richardson
DC (2018) MolProbity: More and better reference data for improved all-atom structure validation.
Protein Sci. 27, 293–315.
2
2
0 Holm L & Laakso LM (2016) Dali server update. Nucleic Acids Res 44, W351–W355.
1 Cutfield SM, Davies GJ, Murshudov G, Anderson BF, Moody PCE, Sullivan PA & Cutfield JF (1999) The
structure of the exo-β-(1,3)-glucanase from Candida albicans in native and bound forms: relationship
between a pocket and groove in family 5 glycosyl hydrolases. J Mol Biol 294, 771–783.
2 Taylor SC, FergusonAD, Bergeron JJM & Thomas DY (2004) The ER protein folding sensor UDP-glucose
glycoprotein–glucosyltransferase modifies substrates distant to local changes in glycoprotein
conformation. Nat Struct Mol Biol 11, 128–134.
2
This article is protected by copyright. All rights reserved

2
2
2
3 Saino H, Shimizu T, Hiratake J, Nakatsu T, Kato H, Sakata K & Mizutani M (2014) Crystal structures of
β-primeverosidase in complex with disaccharide amidine inhibitors. J Biol Chem 289, 16826–16834.
4 Aspeborg H, Coutinho PM, Wang Y, Brumer III H & Henrissat B (2012) Evolution, substrate specificity
and subfamily classification of glycoside hydrolase family 5 (GH5). BMC Evol Biol 12, 186.
5 Patrick WM, Nakatani Y, Cutfield SM, Sharpe ML, Ramsay RJ & Cutfield JF (2010) Carbohydrate
binding sites in Candida albicans exo-β-(1,3)-glucanase and the role of the Phe-Phe ‘clamp’ at the
active site entrance. FEBS J 277, 4549–4561.
2
2
6 Ly HD & Withers SG (1999) Mutagenesis of glycosidases. Annu Rev Biochem 68, 487–522.
7 Davies G & Henrissat B (1995) Structures and mechanisms of glycosyl hydrolases. Structure 3, 853–
8
59.
2
2
8 Rye CS & Withers SG (2000) Glycosidase mechanisms. Curr Opin Chem Biol 4, 573–580.
9 Tribolo S, Berrin J-G, Kroon PA, Czjzek M & Juge N (2007) The crystal structure of human cytosolic β-
glucosidase unravels the substrate aglycone specificity of a family 1 glycoside hydrolase. J Mol Biol
3
70, 964–975.
3
3
0 Kantsadi AL, Apostolou A, Theofanous S, Stravodimos GA, Kyriakis E, Gorgogietas VA,
Chatzileontiadou DSM, Pegiou K, Skamnaki VT, Stagos D, Kouretas D, Psarra A-MG, Haroutounian SA,
Leonidas DD (2014) Biochemical and biological assessment of the inhibitory potency of extracts from
vinification byproducts of Vitis vinifera extracts against glycogen phosphorylase. Food Chem Toxicol
6
7, 35–43.
1 Weignerová L, Marhol P, Gerstorferová D & Křen V (2012) Preparatory production of quercetin-3-β-D-
glucopyranoside using alkali-tolerant thermostable α-L-rhamnosidase from Aspergillus terreus.
Bioresour Technol 115, 222–227.
3
3
2 Ng JD, Gavira JA & Garcia-Ruiz JM (2003) Protein crystallization by capillary counterdiffusion for
applied crystallographic structure determination. J Struct Biol 142, 218–231.
3 Stewart PDS, Kolek SA, Briggs RA, Chayen NE & Baldock, PFM (2011) Random microseeding: A
theoretical and practical exploration of seed stability and seeding techniques for successful protein
crystallization. Cryst Growth Des 11, 3432–3441.
3
4 Mueller U, Förster R, Hellmig M, Huschmann FU, Kastner A, Malecki P, Pühringer S, Röwer M, Sparta
K, Steffien M, Ühlein M, Wilk P & Weiss MS (2015) The macromolecular crystallography beamlines at
This article is protected by copyright. All rights reserved

BESSY II of the Helmholtz-Zentrum Berlin: Current status and perspectives. Eur Phys J Plus 130, 141–
50.
1
3
3
5 Kabsch W (2010) XDS. Acta Crystallogr Sect D-Biol Crystallogr 66, 125–132.
6 Krug M, Weiss MS, Heinemann U & Mueller U (2012) XDSAPP: a graphical user interface for the
convenient processing of diffraction data using XDS. J Appl Crystallogr 45, 568–572.
7 Diederichs K (2010) Quantifying instrument errors in macromolecular X-ray data sets. Acta Crystallogr
Sect D-Biol Crystallogr 66, 733–740.
3
3
8 Pape T & Schneider TR (2004) KL2MAP: a graphical user interface for macromolecular phasing with
SHELX programs. J Appl Crystallogr 37, 843–844.
3
4
9 Sheldrick GM (2008) A short history of SHELX. Acta Crystallogr Sect A 64, 112–122.
0 Emsley P, Lohkamp B, Scott WG & Cowtan K (2010) Features and development of Coot. Acta
Crystallogr Sect D-Biol Crystallogr 66, 486–501.
4
4
1 Vagin A & Teplyakov A. (2010) Molecular replacement with MOLREP. Acta Crystallogr Sect D-Biol
Crystallogr 66, 22–25.
2 Murshudov GN, Skubák P, Lebedev AA, Pannu NS, Steiner RA, Nicholls RA, Winn MD, Long F & Vagin
AA (2011) REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallogr Sect D-
Biol Crystallogr 67, 355–367.
4
4
3 DeLano WL (2009) The PyMOL Molecular Graphics System. DeLano Scientific, San Carlos, CA.
4 Bordo D & Argos P (1991) Suggestions for "safe" residue substitutions in site-directed mutagenesis. J
Mol Biol 217, 721–729.
4
5 Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC & Ferrin TE (2004) UCSF
Chimera – A visualization system for exploratory research and analysis. J Comput Chem 25, 1605–
1
612.
4
4
4
6 Needleman SB & Wunsch CD (1970) A general method applicable to the search for similarities in the
amino acid sequence of two proteins. J Mol Biol 48, 443–53.
7 Henikoff S & Henikoff JG (1992) Amino acid substitution matrices from protein blocks. Proc Natl Acad
Sci USA 89, 10915–10919.
8 Schaftenaar G & Noordik JH (2000) Molden: a pre- and post-processing program for molecular and
electronic structures. J Comput-Aided Mol Design 14, 123–134.
4
9 Land H & Humble MS (2018) YASARA: A tool to obtain structural guidance in biocatalytic
investigations. In Methods in Molecular Biology, vol 1685, Protein Engineering (Bornscheuer U &
Höhne M, eds), pp. 43–67. Humana Press, New York, NY.
This article is protected by copyright. All rights reserved

Table 1. Crystal parameters, data collection and refinement statistics
Data collection statistics
For SIRAS phasing
native)
For SIRAS phasing
(NaI soak)
Dataset
For refinement
(
Protein
deglycosylated
glycosylated
glycosylated
P4 2 2
Space group
P2 2 2
P4 2 2
1
1
1
1
1
1 1
Cell parameters [Å; °]
53.0, 97.3, 139.2;
98.2, 98.2, 154.5;
90, 90, 90
2
97.7, 97.7, 155.0;
90, 90, 90
2
9
0, 90, 90
2
Number of molecules in AU
Wavelength [Å]
0.918
1.5
1.5
4
6.60–1.27
49.09–1.74
(1.84–1.74)
147,835 (23,602)
3.9 (3.2)
49.09–2.00
(2.05–2.00)
97,025 (7,051)
18.69 (13.2)
99.9 (98.6)
0.142 (1.953)
18.23 (1.35)
0.999 (0.417)
47.2
Resolution [Å]
(1.35–1.27)
Number of unique reflections
Multiplicity
180,049 (22,781)
2.98 (2.19)
94.6 (74.8)
0.068 (1.446)
9.58 (0.78)
0.999 (0.359)
20.6
Completeness [%]
Rmeasa
99.6 (98.2)
0.067 (69.8)
16.3 (1.96)
0.999 (0.694)
29.2
Average I/(I)
CC1/2
Wilson B [Å2]b
Refinement statistics
4
6.60–1.27
Resolution range [Å]
(1.30–1.27)
No. of reflections in working set
No. of reflections in test set
R value [%]c
178,251 (9,402)
1,801 (95)
15.9 (36.3)
17.8 (39.3)
0.015
Rfree value [%]d
RMSD bond length [Å]
This article is protected by copyright. All rights reserved

RMSD angle [º]
1.925
6,698
5,889
Number of atoms in AU
Number of protein atoms in AU
Number of water molecules in
AU
6
90
1
8.76 / 17.00 /
2.85
Mean all/protein/water [Å2]
3
Ramachandran plot statisticse
Residues in favored regions [%]
Residues in allowed regions [%]
PDB code
97.9
1.8
6I1A
a
b
Rmeas as defined in Diederichs & Karplus [16]. Wilson B determined by the Sfcheck program of the
c
CCP4 suite [17]. R-value = ||F | – |F ||/|F |, where F and F are the observed and calculated
o
c
o
o
c
structure factors, respectively. d Rfree is equivalent to the R value but is calculated for 5% of
e
reflections chosen at random and omitted from the refinement process [18]. As determined by
MolProbity [19].
Table 2. Specific activities of AnRut and its variants for hydrolysis of pNP rutinosidea.
–1
–1
Enzyme
Wild-type
R108L
Spec. activity [μmol min mg ]
0.20 ± 0.03
<0.01
H153Q
E210A
0.01 ± 0.01
N.D.b
H282S
0.01 ± 0.01
0.05 ± 0.02
H283S
This article is protected by copyright. All rights reserved

E319A
<0.01
a
The measurements were performed in triplicate with 4 mM pNP rutinoside.
b
No activity detected.
Table 3. Kinetic parameters of AnRut for hydrolysis of rutin and isoquercitrina.
–
1
–1 –1
Substrate
Rutin
KM [mM]
k [s ]
k /K [mM s ]
cat
cat
M
0.51 ± 0.06
39 ± 1
65 ± 2
77
Isoquercitrin 0.18 ± 0.04
361
a
See Materials and methods for experimental details.
Table 4. Specific hydrolytic activities of AnRut with monoglycosides or rutin in the presence or
absence of β-D-glucose, α-L-rhamnose or rutinosea.
Substrateb
Added
Spec. activity
carbohydrateb [μmol min mg ]
–1
–1
–
50 ± 5
46 ± 6
47 ± 5
48 ± 6
0.4 ± 0.1
N.D.c
β-D-glucose
Rutin
α-L-rhamnose
rutinose
pNP-Glc
–
–
pNP-Rha
a
The measurements were performed in triplicate.
Concentration: 4.0 mM.
b
c
No activity detected.
Figure legends:
This article is protected by copyright. All rights reserved

Fig. 1. Rutin hydrolysis catalyzed by rutinosidase, resulting in the formation of the disaccharide
rutinose and quercetin. AnRut accepts also isoquercitrin (quercetin 3-O-β-D-glucopyranoside) as a
substrate as shown in this work.
Fig. 2. Overall structure of AnRut. (A) Ribbon representation of the AnRut structure showing the
(
β/α) barrel architecture. α-Helices are shown in cyan, β-strands in magenta. (B) Topology diagram;
8
glycosylation sites are indicated by hexagons. (C) Surface representation of AnRut colored according
to the electrostatic potential (red for negative, blue for positive). The positively charged entrance to
the active site is highlighted with a black circle.
Fig. 3. Structural comparison between AnRut and CaExg. (A) Superposition of the AnRut structure
(magenta) onto the structure of CaExg (PDB code 1CZ1, in green). (B) Close-up showing the
superposition of active-site residues of AnRut and CaExg. Two conserved water molecules in AnRut
and CaExg are represented as spheres and labelled Wat1. Residues selected for mutational analysis
are underlined.
Fig. 4. Phylogenetic analysis and sequence alignment of AnRut. The multiple sequence alignment
was performed by using CLUSTALW. The accession numbers of the protein sequences are indicated.
(
A) Phylogenetic relationships between AnRut, known diglycosidases, CaExg and ScExg. The
unrooted phylogenetic tree was constructed in MEGA7 using the Neighbor-Joining method with
000 bootstrap replications. The evolutionary distances were calculated using the p-distance
1
method. The bar represents 0.1 amino acid substitutions per site. (B) Sequence alignment of AnRut
with CaExg and fungal diglycosidases within GH5, subfamily 23. Only segments around conserved
residues that are located in the active site (in red) including the catalytic residues are shown. The
numbering refers to the translated nucleotide sequence of AnRut.
This article is protected by copyright. All rights reserved

Fig. 5. Reaction mechanism of AnRut and catalytic residues. (A) Activity rescue of AnRut variant
E319A with rutin as a substrate using sodium formate (blue circles) or sodium azide (orange squares)
as external nucleophiles. (B) Activity rescue of mutant E210A in the presence of rutin using sodium
formate (blue circles) or sodium azide (orange squares) as external nucleophiles. (C) Hydrolysis of
1
pNP β-rutinoside with AnRut and the time course of rutinose formation monitored by H NMR
spectroscopy. Labelled are the signals of the anomeric protons for the glucose (Glc) and rhamnose
(Rha) units of the substrate (at time zero) and the newly formed α and β isomers of rutinose, which
are marked with superscripts A and B (in red), respectively. The spectra display the enzymatic
formation of the β-anomeric form of rutinose, with a small amount of α-rutinose subsequently
generated by mutarotation. (D) Double-displacement mechanism of retaining AnRut for rutin
hydrolysis with β-rutinose as a product; ROH represents quercetin. The enzymatic reaction takes
place in two main steps: glycosylation and deglycosylation.
Fig. 6. Active site of AnRut with and without bound rutin. (A) Close-up of active site residues of the
crystal structure with the electron density contoured at 1.0 σ. (B) Cross-sectional view of the
entrance to the active site, exposing rutin (represented by space-filling spheres) bound in the active
site. The loops that partially cover the active site are shown in ribbon representation; they are
highlighted in maroon and labelled with residue numbers. (C,D) Molecular model of rutin bound to
the active site. Rutin is depicted by sticks with carbon atoms colored green; residues that interact
with the substrate are shown as sticks with yellow carbon atoms. Polar interactions are represented
by dashed lines. Residues included in the mutational study are underlined. Aromatic sidechains
clamping the aglycone moiety are shown in panel D. The views in panels B and D are rotated by
about 60° along the vertical axis compared to the views in panels A and C.
Acknowledgements
This work was supported by the European Regional Development Fund, OP RDE, project "Chemical
biology for drugging undruggable targets (ChemBioDrug)" (No. CZ.02.1.01/0.0/0.0/16 019/0000729)
This article is protected by copyright. All rights reserved

and the Czech Science Foundation (grant no. 19-00091S). The use of beamline MX14.1 operated by
the Helmholtz-Zentrum Berlin at the BESSY II electron storage ring (Berlin-Adlershof, Germany) for
collecting diffraction data is gratefully acknowledged.
Author contributions
VK, PŘ, MK and PB planned the work. JK, PP, JB, MK and PŘ performed experiments and analyzed
data. LB performed the molecular modelling experiments. HP carried out the NMR experiments. PŘ,
PP and MK wrote the manuscript, which was finally revised by all authors.
Conflicts of Interest: All authors declare no conflict of interest.
This article is protected by copyright. All rights reserved

febs_15208_f1.tif
Thisꢀarticleꢀisꢀprotectedꢀbyꢀcopyright.ꢀAllꢀrightsꢀreserved

febs_15208_f2.tif
Thisꢀarticleꢀisꢀprotectedꢀbyꢀcopyright.ꢀAllꢀrightsꢀreserved

febs_15208_f3.tif
Thisꢀarticleꢀisꢀprotectedꢀbyꢀcopyright.ꢀAllꢀrightsꢀreserved

febs_15208_f4.pdf
A
B
Thisꢀarticleꢀisꢀprotectedꢀbyꢀcopyright.ꢀAllꢀrightsꢀreserved

febs_15208_f5.pdf
A
B
D
acid/base
E210
C
HO
O
O
HO
O
OR
H3C
HO
HO
O
OH
HO
–
O
O
OH
H2O
C
nucleophile
E319
ROH
C
–
O
O
O
H
C
HO
O
O
H3C
HO
HO
O
H
OH
HO
O
O
C
OH
C
HO
O
O
HO
O
H3C
HO
HO
OH
O
OH
HO
–
O
O
OH
C
OH
HO
R:
O
OH
O
OH
Thisꢀarticleꢀisꢀprotectedꢀbyꢀcopyright.ꢀAllꢀrightsꢀreserved