A Sustainable One-Pot, Two-Enzyme Synthesis of Naturally Occurring Arylalkyl Glucosides

Article abstract of DOI:10.1002/cssc.201700136

A sustainable, convenient, scalable, one-pot, two-enzyme method for the glucosylation of arylalkyl alcohols was developed. The reaction scheme is based on a transrutinosylation catalyzed by a rutinosidase from A. niger using the cheap commercially available natural flavonoid rutin as glycosyl donor, followed by selective “trimming” of the rutinoside unit catalyzed by a rhamnosidase from A. terreus. The process was validated with the syntheses of several natural bioactive glucosides, which could be isolated in up to 75 % yield without silica-gel chromatography.

Full text of DOI:10.1002/cssc.201700136

Accepted Article
Title: A sustainable one-pot two-enzymes synthesis of naturally
occurring arylalkyl glucosides
Authors: Ivan Bassanini, Jana Krejzova, Walter Panzeri, Daniela Monti,
Vladimir Kren, and Sergio Riva
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: ChemSusChem 10.1002/cssc.201700136
Link to VoR: http://dx.doi.org/10.1002/cssc.201700136
A Journal of
www.chemsuschem.org

10.1002/cssc.201700136
ChemSusChem
FULL PAPER
A sustainable one-pot two-enzymes synthesis of naturally
occurring arylalkyl glucosides
I. Bassanini,a,b J. Krejzová,c W. Panzeri,d D. Monti,a V. Křen,c S. Rivaa,*
Abstract: A sustainable, convenient and scalable one-pot two-
enzymes method for the glucosylation of arylalkyl alcohols has been
developed. The reaction scheme is based on a transrutinosylation
procedures are time-consuming and result in overall mediocre
yields, but they use large amounts of solvents, silica gel and
various toxic chemicals.
catalyzed by
a
rutinosidase from A. niger using the cheap
Therefore, alternative and greener approaches able to
improve selectivity and efficacy under mild reaction conditions
would offer viable solutions.
commercially available natural flavonoid rutin as glycosyl donor,
followed by a selective ‘trimming’ of the rutinoside unit catalyzed by
a rhamnosidase from A. terreus. The process has been validated
with the syntheses of several natural bioactive glucosides, that could
be isolated in up to 75 % yield avoiding silica gel chromatography.
Introduction
Substituted benzyl, phenylethyl and phenylpropenyl glucosides,
such as 1a-8a (Figure 1), are bioactive natural compounds
largely occurring in plants.[1] For example, salidroside (3a) is
associated with a wide number of pharmacological effects,
including antioxidant, anti-inflammatory and anticancer
activities,[2-4] whereas trandirin (5a) and coniferin (also named
citrusin, 7a) are studied for their potential antidepressant[5] and
hypotensive[6] effects, respectively. Those biological effects are
often related to the structures of the aglycones but, importantly,
the physicochemical, pharmacological, and toxicity profiles of
these substances are strongly influenced by the conjugation with
the sugar.[7]
These compounds can be isolated in minor amounts from
natural sources or can be synthesized chemically. For instance,
in a recent work we synthesized the glucoside coniferin (7a).[8]
The synthetic pathway was straightforward, based on well-
defined and reliable protocols, nevertheless the sequence of
protective and deprotective reactions consisted of seven
synthetic steps. Moreover, two chromatographic purifications
were required to obtain the product from glucose and the
respective primary alcohol as starting materials. Not only such
Figure 1. Compounds 1-8 and natural products 1a-8a.
In general, the structural complexity and the ‘fragility’ of
natural bioactive compounds makes them an ideal target for
biocatalysis which, by exploiting the inherent properties of
enzymes in terms of selectivity and efficacy under mild reaction
conditions, offers interesting solutions.
[a]
Dr. S. Riva*, Dr. I. Bassanini, Dr. D. Monti
Istituto di Chimica del Riconoscimento Molecolare,
Consiglio Nazionale delle Ricerche
Via Mario Bianco 9, Milano, 20131, Italy.
E-mail: Sergio.riva@icrm.cnr.it
The formation of stereoselective glycosidic bonds in nature
is mainly catalyzed by glycosyl transferases, a group of
enzymes that however need activated sugars (UDP-glucose or
other nucleosides, glucosyl phosphate). [8-11] Alternatively, it may
be accomplished via the transglycosylation action of
glycosidases, the enzymes that in nature hydrolyze glycosidic
bonds,[12,13] or by their artificial mutants, the so-called
glycosynthases.[12] The latter enzymes are generated by site
directed mutagenesis, depleting natural glycosidases from their
hydrolytic attitude. Due to the special configuration of their active
sites, they can catalyze the formation of a glycosidic bond, but
they are not able to hydrolyze it anymore. Moreover, they also
usually need activated sugar acceptors, such as fluorides or
azides and requires considerable workload.
[b]
[c]
[d]
Dr. I. Bassanini
Dipartimento di Chimica,
Università degli Studi di Milano,
Via Golgi 19, Milano, 20133, Italy.
Prof. V. Křen, Dr. J. Krejzová
Institute of Microbiology, Laboratory of Biotransformation,
Czech Academy of Sciences, Vídeňská 1083, CZ 142 20, Prague,
Czech Republic.
Mr. W. Panzeri
Dipartimento di Chimica, Materiali ed Ingegneria Chimica ‘Giulio
Natta’, Politecnico di Milano
P.zza Leonardo da Vinci 32, Milano, 20133, Italy.
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201700136
ChemSusChem
FULL PAPER
to the presence of rhamnose-recognizing receptors on their
surface. [20]
Scheme 1. Two-enzymatic synthesis of coniferin (7a).
Often, however, the corresponding -glucosides are
required. For this purpose, -L-rhamnosidases are perfectly
suitable to catalyze the selective de-rhamnosylation. We have in
our hands a highly efficient thermo- and alkali-stable -L-
rhamnosidase from Aspergillus terreus, which is now available
as a heterologous protein produced by a recombinant strain of P.
pastoris.[21] This enzyme preparation, devoid of -glucosidase
activity, has been shown to efficiently and selectively cleave the
(61) rhamnosyl unit of rutin to yield isoquercitrin (11). [15,22]
We present here a sustainable, convenient and scalable
telescoping two-enzymes glucosylation of arylalkyl alcohols,
exploiting rutin as a glycosyl donor.
Among the number of papers published on this topic, a few
years ago Akita et al. described the synthesis of a series of
glucosides of primary alcohols, including compounds 3a-7a,
exploiting a cheap and available -glucosidase from almonds.[14]
Products were obtained from the unprotected glucose and
respective alcohol in a single step.
However, isolated yields were moderate to low and
purification steps by silica gel chromatography (where typically
losses of polar compounds occur due to partial irreversible
adsorption) were always required.
Rutin (9) is a glycosyl derivative of the flavonoid quercetin
(10;
glycon:
rutinose,
-L-rhamnopyranosyl--D-
glucopyranoside). [15] It is an important phytochemical contained
in various plants and fruits (buckwheat, tobacco, apples,
tomatoes, citrus and grapes), which was named after one of its
primary natural sources, the common rue Ruta graveolens. [16]
Rutin is now produced as a multi-ton commodity chemical from
the Brasilian tree Fava d’anta (Dimorphandra mollis); it is also
accumulated as a waste byproduct in different manufactory
processes (e.g., the production of fruit juices). Rutin has a
GRAS status (cf. www.hc-sc.gc.ca) and it is used in plethora of
Results and Discussion
Transglycosylations are kinetically controlled reactions where
the product may become
a substrate for a subsequent
hydrolysis catalyzed by the same enzyme, lowering thus the
yields. Therefore, the first critical point of our investigation was
the thorough optimization of the rutinosidase-catalyzed reaction.
As shown in Scheme 1, coniferyl alcohol (7), the aglycone of the
natural hypotensive agent coniferin (7a), was chosen as a model
glycosyl acceptor. When 7 was subjected to the rutinosidase
action following the published transglycosylation protocol (a 50
mM 1:1 mixture of rutin and glycosyl acceptor dissolved in a
50 % v/v DMSO solution in citrate-phosphate buffer at pH 5,
nutraceuticals and pharmaceutical preparation as
a
capillary/blood vessel protectant and as an antiviral agent.
Therefore, rutin is a highly suitable, inexpensive (presently it
costs less than 50 €/Kg) and non-toxic starting material for the
synthesis of rutinosylated conjugates. All these properties make
rutin superior to the typically used glycosyl donors for
glycosidase-catalyzed synthesis, e.g. expensive and toxic
nitrophenyl glycosides.
We have recently shown that the diglycosidase
rutinosidase from Aspergillus niger (cloned and expressed in
Pichia pastoris[17]) has very strong transglycosylation activity
both for aliphatic alcohols and phenols. This diglycosidase is
much more efficient in transglycosylation than other commonly
used (mono)glycosidases. This reaction has another advantage:
all the reactants and products are non-toxic (natural) flavonoids
and thus they do not compromise potential pharmaceutical
and/or nutraceutical applications. Rutinosylated substances can
incubated at 35 °C for 7 h with 0.026 U mmol-1
of
substrate
rutinosidase), only 10
% isolated yield of the desired
disaccharide derivative 7b was obtained since the major
rutinosidase-catalyzed reaction was rutin hydrolysis.
To optimize the reaction, we have focused on three key
parameters: the enzyme amount, the glycosyl donor/acceptor
stoichiometric ratio and the amount of cosolvent added.
After
a whole set of experiments (for details see
Supplementary Material, sections S1 and S2), we found that the
use of a lower amount of rutinosidase (optimum value ca 0.013
U mmol-1alcohol) and a reaction time of 7-8 h strongly improved
transglycosylation over hydrolysis. Moreover, an initial excess of
alcohol acceptor over the glycosyl donor (a starting mixture of
rutin/alcohol 0.5 : 1 eq) gave better results. To limit a significant
substrate inhibition caused by rutin, a lower amount of the
be used also as food additives, ingredients in cosmetics and
[18,19]
antiviral agents.
Besides this, rutinose-containing
glycopolymers may have entrapment abilities for fibroblasts due
This article is protected by copyright. All rights reserved.

10.1002/cssc.201700136
ChemSusChem
FULL PAPER
cosolvent DMSO (15 % v/v instead of 50 %) was used to
decrease the active concentration of rutin in the reaction. In this
system, the donor was just suspended and only partially
dissolved.
coniferyl alcohol was achieved by sequentially combining the
transglycosylation grafting and the rhamnose trimming in the
same reactor. Only, rutinosidase had to be thermally inactivated
prior adding the rhamnosidase (0.5 U mmol-1substrate) and stirring
the final reaction mixture overnight. Following the previously
described purification procedure using XAD4 resin, the target
coniferin (7a) was obtained in 75 % isolated yield. In comparison,
the chemical synthesis of 7a required seven reactions
(protection of carbohydrate hydroxyl groups, deprotection of the
anomeric OH, activation of the anomeric OH, protection of the
phenolic hydroxyl group, condensation between activated sugar
and the aglycone, deprotection of the phenolic moiety,
deprotection of the sugar hydroxyl groups) and two silica gel
chromatography. Therefore, this enzymatic approach is superior
to the chemical one in terms of better atom economy. [8]
To make this procedure fully scalable and green, a
purification strategy avoiding the use of silica gel
chromatography was developed. By fully exploiting the chemo-
physical differences between the starting alcohol 7, the product
7b, the residual rutin 9 (if any), the produced quercetin 10 and
the free disaccharide rutinose always formed as a byproduct, it
was possible to purify the rutinoside 7b by means of i)
centrifugation, to recover the produced quercetin (10); ii)
extraction with organic solvents, to recover the unreacted
alcohol and; iii) solid phase extraction on Amberlite®XAD4 non-
[23]
ionic resin.
Accordingly, the produced quercetin (10) was
removed by centrifugation, the unreacted coniferyl alcohol was
extracted with AcOEt, the hydrolyzed rutinose, the buffer salts
and the residual co-solvent DMSO were easily removed in the
eluate from the product 7b that was adsorbed to the XAD4 resin.
The DMSO used in the procedure does not compromise the
green concept: its toxicity is very low (this compound is currently
used in cosmetics and medications to increase the skin
permeability) and its residual presence could not be detected –
by NMR – in the isolated products.
Both enzymes used in this work, i.e., rutinosidase and -L-
rhamnosidase, do not need to be recycled. In fact, they are both
produced heterologously by P. pastoris and their production
costs are very low. Typical activity is ca 0.15 U/mL cultivation
broth. This means that less than 1 mL of the medium needs to
be used for 1 g of coniferyl alcohol. The previously described
experiments were run in the presence of purified rutinosidase,
but we have verified that even crude enzyme (just fermentation
medium) can be used equivalently. By rough estimation the
substrate used for our optimization study (coniferyl alcohol) is
significantly (ca 100 – 1000 times) more expensive than the
enzymes used, thus making non-economic the additional work
required to regenerate the biocatalysts. Moreover, the use of
immobilized enzymes is not compatible with the reaction
protocol, as we work with a suspension of substrates and
therefore we need soluble enzymes.
The preparative scale transrutinosylation of
7 was
conducted with 1 g of alcohol. Rutin (a total of 1.5 eq) was
portion-wise added (0.5 equivalent each time) during 8 h (the
disappearance of rutin was monitored by TLC).
The product 7b was obtained in 70 % isolated yield and
fully characterized by MS ([M + Na+] m/z 511.3) and NMR. In
addition to the expected signals due to the coniferyl moiety, the
presence of a disaccharide unit was clearly demonstrated by the
respective signals in the 1H-NMR spectrum. Specifically, the
anomeric protons of the rhamnopyranose and glucopyranose
units were doublets (J = 1.6 Hz and J = 7.8 Hz, respectively)
centered at 4.80 and 4.37 ppm. The presence of the
rhamnopyranose unit was further confirmed by the signal of its
methyl group (δH 1.30, d, J = 6.3 Hz).
On the contrary, polystyrene microporous resin Amberlite
[23]
XAD4
can be well reused in tens of cycles and refreshed by
simple washing with MeOH or acetone. When the resin starts to
lose capacity, well assessed regeneration protocols are
available (see Supplementary Information, S6). This means that
the resin can be technically used for tens to hundreds of cycles.
The same protocol as above was then applied to the
alcohols 1-8 and the results are summarized in Table 1. All the
compounds were accepted by the rutinosidase, except for
sinapyl alcohol (8). Compounds 4-6 behaved analogously to 7
and with coumaryl alcohol (5) the transglycosylation reaction
was also fully regioselective. Compounds 1-3 were also
glycosylated, but apparently, they were less reactive, as the
isolated yield were lower. A more careful investigation of the
reaction kinetics by HPLC (see Supplementary Information,
sections S1 and S3) showed that the rutinosides 1b-3b were
formed, but they were hydrolyzed much faster than the
rutinosides of the phenylpropenyl alcohols 4-7. Therefore, the
reactions needed to be carefully monitored and optimized.
Thus, the initial transglycosylation reaction was also highly
regioselective, as no formation of phenolic glycoside at C-1 OH
was observed.
After the optimization of the first step, the consecutive step,
i.e., the enzymatic trimming of the rhamnose unit, was
straightforward. The citrate-phosphate solution (50 mM, pH 5) of
rutinoside 7b (200 mg, 100 mM) incubated at 35 °C for 3 h with 2
U mmol-1
of enzyme, followed by XAD4 solid phase
rutinoside
extraction, gave the target coniferin (7a) in 90 % isolated yield.
1
7a was characterized by MS ([M + Na+] m/z 365.2 Da) and H-
and 13C-NMR, which showed the presence of both the coniferyl
aglycone and the glucopyranosyl moiety. These data agreed
with literature data. [24]
As a further improvement of the efficiency of the synthetic
process, we investigated the possibility of a one-pot approach by
merging the two enzymatic reactions. We used the so-called
‘telescoping synthesis’: the reactants were added sequentially to
the reactor without any intermediate purification or work-up and
the reaction ran in one pot. Two-enzymes glucosylation of
This article is protected by copyright. All rights reserved.

10.1002/cssc.201700136
ChemSusChem
FULL PAPER
Table 1. Telescoping two-enzymes synthesis of glucosides.
Substrate
Product
Isolated Yield [%]
1[a]
2[a]
3[a]
1a
2a
12
24
3a
3c
44.5[b]
4.5[b]
4
5
6
7
8
4a
5a
70
68
58
75
--
6a
7a
NO
[a] Reactions with substrates 1, 2 and 3 were run for 24 h instead of 8 h;
[b] Relative ratio estimated by 1H NMR.
Tyrosol (3) behaved differently in terms of regioselectivity
(Scheme 2). The MS spectra of the isolated product ([M + Na+]
m/z 232.2) confirmed the introduction of a glucose moiety.
However, the 1H-NMR spectrum clearly showed the presence of
two products in a 10 : 1 ratio. The signal due to the -anomeric
proton (H-10) of the major one, a doublet with J = 7.8 Hz
resonating at 4.29 ppm could be overlapped to the
corresponding signal of the glucoside of phenylethanol (2a),
therefore it could be possible to assign the structure 3a to this
compound.
The significant downfield shift of the aromatic protons of
the minor one (the two coupled doublets (J = 8.4 Hz) due to the
AA’BB’ aromatic system resonating at 7.18 and 7.03 ppm,
whereas they were recorded at 7.08 and 6.71 ppm in 3a),
indicated the glucosylation of the phenolic OH (3c).
Scheme 2. Glycosides of tyrosol (3). A) Two-enzymatic synthesis of tyrosol
glucosides 3a and 3c; B) Rutinosidase-catalyzed synthesis of tyrosol
rutinosides 3b and 3d.
Conclusions
The proposed synthetic protocols allowed the preparation,
in good isolated yields, of two types of glycosylated products,
rutinosides and -glucopyranosides, and specifically of valuable
natural glucosides. This goal was achieved using the glycosyl
donor rutin, presently a byproduct in various manufactory
processes and
a
cheap and biocompatible food-grade
To further investigate the glycosylation of this interesting
compound, we decided to isolate the intermediate tyrosol
rutinosides 3b and 3d. Following the previously described
protocol, a mixture of the expected regioisomers 3b and 3d (MS
commodity chemical, and avoiding the use of protective groups,
toxic activating agents, expensive reagents, environment-
unfriendly experimental procedures requiring dry solvents and
harsh conditions and, last but not least, avoiding
chromatographic purifications. Moreover, the flavonoid quercetin
(10), the main by-product of our synthesis has a high value per
se and is easily recovered (centrifugation) during our established
purification method.
All these aspects (the use of a renewable feedstock, the
use of biocatalysts, the use of less hazardous chemical
synthesis, the use of safer solvents and chemicals, the reduced
number of synthetic derivatives) perfectly fit with most of the so-
called ’12 principles of Green Chemistry’ proposed by Anastas
and Warner a few years ago. [25]
1
and H-NMR in accordance with the proposed structures), was
again obtained, this time in 3 : 1 ratio. The different observed
regioisomeric ratio between 3a and 3c vs. 3b and 3d could be
explained taking in account the different reaction time allowed
for the transrutinosylation step in the two protocols, that was
longer in the telescopic synthesis. A detailed kinetic study
performed by monitoring the reactions by HPLC (see
Supplementary Material, sections SI, Figure S2), clearly showed
that the phenolic rutinoside 3d could be hydrolyzed faster than
the alkyl one 3b.
The separation of the two regioisomeric glycosides (3a
from 3c, 3b from 3d) proved to be quite troublesome and
several techniques failed, i.a. silica gel flash chromatography
and reverse phase preparative HPLC. Separation of the
regioisomeric rutinosides 3b and 3d could be achieved by gel-
filtration using a Biogel P2 (BioRad), and the major product 3b
could be obtained in a 94% purity, as determined by 1H-NMR.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201700136
ChemSusChem
FULL PAPER
13C-NMR (101 MHz; MeOD, 25 °C):  147.7, 146.3, 133.1, 128.9, 122.1,
119.7, 114.8, 109.2, 101.6, 100.9, 76.7, 75.5, 73.7, 72.6, 70.97, 70.83,
70.3, 69.5, 68.4, 66.8, 54.9, 29.3, 16.7
Experimental Section
3.1. Materials, chemicals and equipment
EI-MS, m/z 511.3 [M+Na]+.
NMR spectra were recorded on Bruker AC400 spectrometer (400 or 500
MHz) in MeOH-d4, DMSO-d6 or D2O, MS spectra were recorded on
Bruker Esquire 3000 Plus spectrometer. NMR and the MS spectra in
extenso are available in the Supplementary Material, sect. SI.
Biotranformations were carried out using a G24 Environmental Incubator
New Brunswick Scientific Shaker (Edison, USA) or a Thermomixer
Comfort (Eppendorf, DE).
b) Rhamnosidase from A. terreus (0.82 U) was dissolved in citrate-
phosphate buffer (50 mM, pH 5) solution of 7a (200 mg, 0.41 mmol, 2.5
mL). The resulting mixture was incubated at 35 °C and 750 rpm for 3 h.
After checking the complete conversion of the starting rutinoside into the
target glucoside by TLC analysis, heated to 100 °C for 10 minutes and
then cooled again to room temperature to be subjected to an Amberlite
XAD4 solid phase extraction (loaded and washed with H2O; eluted with
MeOH), affording pure coniferin, 7a (yellow solid, 126 mg, 90%).
Phenylpropanoid alcohols were prepared from their corresponding acids,
following standard procedures (Supplementary Material S1).
1H-NMR (400 MHz; MeOD, 25 °C):  7.03 (d, J = 1.9 Hz, 1H: H-5), 6.87
(dd, J = 8.2, 1.9 Hz, 1H: H-3), 6.76 (d, J = 8.1 Hz, 1H: H-2), 6.59 (d, J =
15.9 Hz, 1H: H-7), 6.21 (ddd, J = 15.9, 6.7, 6.0 Hz, 1H: H-8), 4.51 (ddd, J
= 12.4, 5.9, 1.4 Hz, 1H: H-9A), 4.39 (d, J = 7.8 Hz, 1H: H-10), 4.32 (ddd, J
= 12.4, 6.8, 1.2 Hz, 1H: H-9B), 3.90 (dd, J = 11.9, 2.0 Hz, 1H: H-15A),
3.88 (s, 3H: OCH3), 3.70 (dd, J = 11.9, 5.4 Hz, 1H: H-15B), 3.41-3.23 (m,
4H: H-11, H-12, H-13, H-14).
The enzymatic glucosylations were monitored by TLC [precoated silica
gel 60 F254 plates (Merck, DE); mobile phase = AcOEt : MeOH : HCOOH ,
4 : 1 : 0.5; development = UV lamp, Komarovsky reagent (1 mL of 50 %
ethanolic H2SO4 with 10 mL of 2% methanolic 4-hydroxybenzaldehyde)].
HPLC analysis were conducted using a Jasco 880-PU pump equipped
with a Jasco 875-UV/Vis detector Kinetex 5µm EVO C18 100Å 150 × 4.6
mm column, gradient of A = aqueous TFA (1 ppm, pH 2.5-3) and B =
CH3CN as a mobile phase (0 min 80% A; 1-15 min 70 % A; 15-20 min
20 % A; 20-25 min 80 % A), flow rate 0.8 mL/min at 25 °C, detection at
270 nm. -L-Rhamnosyl--D-glucosidase (rutinosidase) from Aspergillus
niger and -L-rhamnosidase from Aspergillus terreus were prepared
according to the literature and were used based on their respective
activities determined by the previously published assays. [14,15]
13C-NMR (101 MHz; MeOD, 25 °C):  147.7, 146.3, 132.9, 129.0, 122.4,
119.8, 114.8, 109.3, 101.8, 76.74, 76.57, 73.7, 70.3, 69.6, 61.4, 55.0.
EI-MS, m/z 365.2 [M+Na]+.
All other reagents were of the highest purity grade from commercial
suppliers.
3.3. Telescoping two-enzymatic glucosylation
General procedure: arylalkyl alcohol (1 eq, 128 mM) and rutinosidase
from A. niger (0.013 U mmol-1substrate) were dissolved in a 15 % solution of
DMSO in citrate-phosphate buffer (50 mM pH 5) and incubated at 35 °C
and 750 rpm. The glycosylation reaction was started by adding 0.5 eq of
rutin. Monitoring the consumption of the glycosyl donor by TLC analysis,
two portions of rutin, 0.5 eq each, were added. The transglycosylations
were conducted for 8 or 16 h, depending on the substrate used. When all
the rutin was consumed the suspension was heated to 100 °C for 10
minutes and then cooled again to 35 °C and rhamnosidase from A.
terreus (0.5 U mmol-1substrate) was added to the mixture, which was
incubated overnight at 35 °C and 750 rpm. The complete conversion of
the intermediate disaccharide to the target -glucopyranoside was
checked by TLC analysis and then the resulting yellow suspension was
cooled to room temperature, diluted with ten volumes of citrate-
phosphate buffer and centrifuged. The solid proved to be the expected
quercetin (10) containing ca 20-25 % of unreacted rutin. The supernatant
water layer, after shifting the pH from 5 to 7.5 – 7.7, was extracted twice
with half the volume of AcOEt and then, after removing the residual
AcOEt in vacuo, subjected to an Amberlite XAD4 solid phase extraction
(mobile phase H2O : MeOH, from 100 : 0 to 0 : 100), affording the pure
arylalkylglucoside.
3.2. Two-step two-enzymatic synthesis of coniferin (7a)
a) A suspension of rutin (1.70 g, 2.78 mmol) in a 15 % solution of DMSO
in citrate-phosphate buffer (50 mM pH 5) containing coniferyl alcohol
(1.00 g, 5.56 mmol) and rutinosidase from A. niger (0.073 U) were
prepared and incubated at 35 °C and 750 rpm. The consumption of the
glycosyl donor was checked by TLC analysis and 1.70 g of rutin was
added again to the reacting mixture after 3 and 5 h. After 8 h the
obtained yellow suspension was heated to 100 °C for 10 minutes and
then cooled again to room temperature, diluted with ten 400 mL of
citrate-phosphate buffer and centrifuged, recovering only the supernatant.
The pH was changed from 5 to 7.5 – 7.7 and the water layer was
extracted with AcOEt (2 × 200 mL) and then, after removing the residual
AcOEt via concentration in vacuo, it was loaded loaded to a column (5 ×
30 cm) with Amberlite XAD4 resin (300 g) in water (elution with slow
gradient H2O to MeOH), affording the pure rutinoside 7b (yellow solid,
755 mg, 70 %). The resin can be simply reused after washing with pure
MeOH followed by extensive washing with water.
1H-NMR (400 MHz; MeOD, 25 °C):  7.04 (d, J = 1.9 Hz, 1H: H-5), 6.89
(dd, J = 8.2, 1.9 Hz, 1H: H-3), 6.76 (d, J = 8.1 Hz, 1H: H-3), 6.60 (d, J =
15.9 Hz, 1H: H-7), 6.21 (ddd, J = 15.8, 6.8, 5.9 Hz, 1H: H-8), 4.80 (d, J =
1.6 Hz, 1H: H-16), 4.48 (ddd, J = 12.3, 5.8, 1.3 Hz, 1H: H-9A), 4.37 (d, J =
7.8 Hz, 1H: H-10), 4.28 (ddd, J = 12.3, 7.0, 1.1 Hz, 1H: H-9B), 4.01 (dd, J
= 11.3, 1.7 Hz, 1H: H-15A), 3.88 (s, 3H: OCH3), 3.73-3.64 (m, 3H: H-15B,
H-18, H-20), 3.44-3.21 (m, 6H: H-11, H-12, H-13, H-14, H-17, H-19), 1.30
(d, J = 6.3 Hz, 3H: H-21).
Compound 1a: According to the General Procedure, compound 1a (30
mg, 0.11 mmol, isolated yield: 12 %, yellow oil, 24 h) was obtained from
100 mg of benzyl alcohol (1) and 846 mg of rutin (9).
1H-NMR (500 MHz; MeOD, 25 °C):  7.44 (d, J = 7.3 Hz, 2H: H-2, H-6),
7.35 (t, J = 7.4 Hz, 2H: H-3, H-5), 7.29 (t, J = 7.3 Hz, 1H: H-1), 4.95 (d, J
This article is protected by copyright. All rights reserved.

10.1002/cssc.201700136
ChemSusChem
FULL PAPER
= 11.8 Hz, 1H: H-7A), 4.69 (d, J = 11.8 Hz, 1H: H-7B), 4.38 (d, J = 7.7 Hz,
1H: H-10), 3.92 (d, J = 10.2 Hz, 1H: H-15A), 3.72 (dd, J = 11.8, 5.5 Hz,
1H: H-15B), 3.39-3.26 (m, 6H: H-11, H-12, H-13, H-14).
Compound 5a, triandrin: According to the General Procedure,
compound 5a (141 mg, 0.45 mmol, isolated yield: 68 %, yellow solid, 8 h)
was obtained from 100 mg of 4-hydroxycinnamyl alcohol (5) and 603 mg
of rutin (9).
13C-NMR (101 MHz; MeOD, 25 °C):  132.4, 122.55, 122.48, 122.0, 96.6,
71.41, 71.36, 71.33, 71.32, 68.49, 68.44, 65.04, 65.03, 56.1, 33.7.
1H-NMR (400 MHz; D2O, 25 °C):  7.32 (d, J = 8.0 Hz, 2H: H-3, H-5),
6.83 (d, J = 7.8 Hz, 2H: H-2, H-6), 6.58 (d, J = 15.9 Hz, 1H: H-7), 6.16 (dt,
J = 15.0, 7.1 Hz, 1H: H-7), 4.47-4.40 (m, 2H: H-9A, H-10), 4.31-4.27 (m,
1H: H-9B), 3.86 (d, J = 12.6 Hz, 1H: H-15A), 3.69-3.66 (m, 1H: H-15B),
3.44-3.24 (m, 4H: H-11, H-12, H-13, H-14).
EI-MS, m/z 293.1 [M+Na]+.
13C-NMR (101 MHz; D2O, 25 °C):  215.4, 155.6, 133.6, 128.9, 128.2,
122.3, 115.7, 101.1, 75.92, 75.90, 73.2, 70.5, 69.7, 60.8, 38.8, 30.3.
Compound 2a: According to the General Procedure, compound 2a (36
mg, 0.20 mmol, isolated yield: 24 %, yellow solid, 24 h) was obtained
from 100 mg of 1-phenyl ethanol (2) and 750mg of rutin (9).
EI-MS, m/z 335.1 [M+Na]+.
1H-NMR (500 MHz; MeOD, 25 °C):  7.27 (d, J = 4.5 Hz, 4H: H-2, H-3 H-
5, H-6), 7.19 (q, J = 4.4 Hz, 1H: H-1), 4.32 (d, J = 7.8 Hz, 1H: H-10), 4.11
(dt, J = 9.5, 7.5 Hz, 1H: H-8A), 3.88 (dd, J = 11.7, 1.5 Hz, 1H: H-15A),
3.81-3.76 (m, 1H: H-8B), 3.69 (dd, J = 11.8, 5.2 Hz, 1H: H-15B), 3.39-
3.19(m, 4H: H-11, H-12, H-13, H-14), 2.96 (td, J = 7.3, 1.6 Hz, 2H: H-7).
Compound 6a, vimalin: According to the General Procedure, compound
6a (115 mg, 0.35 mmol, isolated yield: 58 %, yellow solid, 8 h) was
obtained from 100 mg of 4-methoxycinnamyl alcohol (6) and 558 mg of
rutin (9).
13C-NMR (101 MHz; MeOD, 25 °C):  133.4, 123.3, 122.67, 122.64,
120.53, 120.50, 97.7, 71.43, 71.29, 68.4, 65.0, 56.1, 33.7, 30.6.
1H-NMR (500 MHz; MeOD, 25 °C):  7.35 (d, J = 8.7 Hz, 2H: H-3, H-5),
6.87 (d, J = 8.8 Hz, 2H: H-2, H-6), 6.61 (d, J = 16.0 Hz, 1H: H-7), 6.22 (dt,
J = 15.9, 6.3 Hz, 1H: : H-8), 4.51 (ddd, J = 12.5, 5.8, 1.3 Hz, 1H: H-9A),
4.38 (d, J = 7.8 Hz, 1H: H-10), 4.30 (ddd, J = 12.4, 6.8, 1.1 Hz, 1H: H-9B),
3.90 (dd, J = 11.9, 2.2 Hz, 1H: H-15A), 3.78 (s, 3H: OCH3), 3.71 (dd, J =
11.9, 5.5 Hz, 1H: H-15B), 3.41-3.24 (m, 4H: H-11, H-12, H-13, H-14).
EI-MS, m/z 307.0 [M+Na]+.
Compound 3a, salidroside: According to the General Procedure,
compound 3a (106 mg, 0.35 mmol, isolated yield: 49 %, yellow solid, 24
h) was obtained from 100 mg of tyrosol (3) and 663 mg of rutin (9).
13C-NMR (101 MHz; MeOD,25 °C):  154.1, 127.0, 124.1, 122.0, 117.5,
108.3, 96.5, 71.36, 71.20, 68.4, 64.9, 64.3, 56.1, 49.0, 33.7.
1H-NMR (500 MHz; D2O, 25 °C):  7.05 (d, J = 8.3 Hz, 2H: H-3, H-5),
6.69 (d, J = 8.3 Hz, 2H: H-2, H-6), 4.29 (d, J = 7.8 Hz, 1H: H-10), 4.04-
3.99 (m, 1H: H-8A), 3.85 (d, J = 10.7 Hz, 1H: H-15A), 3.68 (td, J = 13.2,
5.8 Hz, 3H: H-7A, H-8B, H-15B), 3.36 (t, J = 8.7 Hz, 1H: H-11), 3.31-3.24
(m, 3H: H-7A, H-12, H-14), 3.18 (t, J = 8.4 Hz, 1H: H-13).
EI-MS, m/z 349.2 [M+Na]+.
Compound 7a, coniferin: According to the General Procedure,
compound 7a (508 mg, isolated yield: 75 %, yellow solid, 8 h) was
obtained from 250 mg of coniferyl alcohol (7) and 1.30 g of rutin. The
NMR and mass spectra were in accordance with those described above.
13C-NMR (101 MHz; D2O, 25 °C): 154.3, 128.45, 128.26, 113.7, 101.8,
75.59, 75.41, 72.6, 69.6, 69.1, 60.3, 38.0, 33.9.
EI-MS, m/z 323.2 [M+Na]+.
Compound 4a, rosin: According to the General Procedure, compound
4a (140 mg, 0.47 mmol, isolated yield: 70 %, yellow solid, 8 h) was
obtained from 100 mg of cinnamyl alcohol (4) and 684 mg of rutin (9).
Acknowledgements
The authors acknowledge the Czech-Italian academic
collaborative project between CNR and ASCR No. CNR-16-30
(2016-2018) and the COST Action “Systems Biocatalysis”
CM1303 (MSMT LD15085).
1H-NMR (400 MHz; MeOD, 25 °C):  7.44-7.42 (m, 2H: H-2, H-6), 7.31
(dd, J = 8.2, 6.7 Hz, 2H: H-3, H-5), 7.23 (t, J = 7.3 Hz, 1H: 1), 6.70 (d, J =
16.0 Hz, 1H: H-7), 6.39 (ddd, J = 16.0, 6.4, 5.8 Hz, 1H: H-8), 4.55 (ddd, J
= 12.8, 5.7, 1.5 Hz, 1H: H-9A), 4.39 (d, J = 7.8 Hz, 1H: H-10), 4.35 (ddd, J
= 12.8, 6.5, 1.4 Hz, 1H: H-9B), 3.91 (dd, J = 11.9, 2.0 Hz, 1H: H-15A),
3.71 (dd, J = 11.9, 5.3 Hz, 1H: H-15B), 3.42-3.24 (m, 4H:H-11, H-12, H-
13, H-14).
Keywords: Enzymatic glycosylation • Rutinosidase •
Rhamnosidase • Rutin • Coniferyl alcohol
[1]
J. Crouzet, D. Chassagne, in Naturally Occurring Glycosides, ed. R.
Ikan, Wiley, Baffins Lane, Chichester, Wet Sussex PO19 1UD, England,
1st edn. 1999, pp. 225–274.
13C-NMR (101 MHz; MeOD, 25 °C):  136.8, 132.4, 128.2, 127.3, 126.1,
125.3, 102.0, 76.72, 76.59, 73.7, 70.3, 69.4, 61.4.
[2]
[3]
J. Zhang, A. Liu, R. Hou, J. Zhang, X. Jia, W. Jiang, J. Chen, Eur. J.
Pharmacol. 2009, 607, 6–14.
EI-MS, m/z 319.2 [M+Na]+.
C. Sun, Z. Wang, Q. Zheng, H. Zhang, Phytomedicine. 2012, 19, 355–
363.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201700136
ChemSusChem
FULL PAPER
[4]
[5]
[6]
L. Zhang, H. Yu, X. Zhao, X. Lin, C. Tan, G. Cao, Z. Wang, Neurochem.
Int. 2010, 57, 547–555.
[16] L. S. Mazzaferro, J. D. Breccia, Biocatal. Biotrans 2011, 29, 103–109.
[17] D. Šimčíková, M. Kotik, L. Weignerová, P. Halada, H. Pelantová, K.
Adamcová, V. Křen, Adv. Synth. Catal. 2014, 357, 107–117.
[18] A. Chanet, D. Milenkovic, C. Manach, A. Mazur, C. Morand, J. Agric.
Food Chem. 2012, 60, 8809–8822.
V. A. Kurkin, A. V. Dubishchev, V. N. Ezhkov, I. N. Titova, E. V.
Avdeeva, Pharm. Chem. J. 2006, 40, 614–619.
Y. Matsubara, T. Yusa, A. Sawabe, Y. Iizuka, K. Okamoto, Agric. Biol.
Chem. 1991, 55, 647–650.
[19] S. Katayama, F. Ohno, Y. Yamauchi, M. Kato, H. Makabe, S.
Nakamura, J. Agric. Food Chem. 2013, 61, 9617–9622.
[7]
[8]
V. Křen, L. Martínková, Curr. Med. Chem. Rev. 2001, 8, 1313–1338.
I. Bassanini, P. Gavezzotti, D. Monti, J. Krejzová, V. Křen, S. Riva, J.
Mol. Catal. B Enzym. 2016, 134, 295-301.
[20] G. Faury, J. Molinari, E. Rusova, B. Mariko, S. Raveaud, P. Huber, V.
Velebny, A. M. Rober, L. Robert, Arch. Gerontol. Geriatr. 2011, 53,
106–112.
[9]
K. Schmölzer, A. Gutmann, M. Diricks, T. Desmet, B. Nidetzky,
Biotechnol. Adv. 2016, 34, 88–111.
[21] D. Gerstorferová, B. Fliedrová, P. Halada, P. Marhol, V. Křen, L.
Weignerová, Process Biochem. 2012, 47, 828–835.
[10] D.-M. Liang, J.-H. Liu, H. Wu, B.-B. Wang, H.-J. Zhu, J.-J. Qiao, Chem.
Soc. Rev. 2015, 44, 8350–74.
[22] L. Weignerová, P. Marhol, D. Gerstorferová, V. Křen, Bioresour.
Technol. 2012, 115, 222–227.
[11] D. Aerts, T. F. Verhaeghe, B. I. Roman, C. V. Stevens, T. Desmet, W.
Soetaert, Carbohydr. Res. 2011, 346, 1860–1867.
[23] F. A. Tomas-Barberan, M. A. Blazquezt, C. Garcia-Viguera, F. Ferreres,
T.-L. Francisco. Phytochem. Anal. 1992, 3, 178–181.
[12] P. Bojarová, V. Křen, Trends Biotechnol. 2009, 27, 199–209.
[13] T. Desmet, W. Soetaert, P. Bojarová, V. Křen, L. Dijkhuizen, V.
Eastwick-Field, A. Schiller, Chem. - A Eur. J. 2012, 18, 10786–10801.
[14] H. Akita, E. Kawahara, M. Kishida, K. Kato, J. Mol. Catal. B Enzym.
2006, 40, 8–15.
[24] M. Kishida, H. Akita, Tetrahedron. 2005, 61, 10559–10568.
[25] P. T. Anastas, J. C. Warner, Green Chem. Theory Pract. Oxford Univ.
Press. New York. 1998, 30.
[15] J. B. Harborne, C. A. Williams, in The Flavonoids, eds. T. J. Mabry and
H. Mabry, Chapman and Hall Ltd, 11 New Fetter Lane London EC4P
4EE, 1st edn. 1975, pp. 420–427.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201700136
ChemSusChem
FULL PAPER
Entry for the Table of Contents
FULL PAPER
A one-pot two-enzymes synthesis of
arylalkyl glucosides exploiting the
I. Bassanini, J. Krejzová, W. Panzeri, D.
Monti, V. Křen, S. Riva*
activity of
a rutinosidase and a
rhamnosidase is here proposed. The
cheap, non-toxic, natural flavonoid
rutin was selected as a convenient
Page No. – Page No.
A sustainable one-pot two-enzymes
synthesis of naturally occurring
arylalkyl glucosides.
glycosyl-donor
and
the
target
glucosides were obtained in isolated
yield up to the 75%, avoiding the use
of PGs, activating agents and flash
column chromatography on silica gel,
producing the valuable quercetin as
the only recovered by-product.
This article is protected by copyright. All rights reserved.