Enzymatic Synthesis of Therapeutic Nucleosides using a Highly Versatile Purine Nucleoside 2’-DeoxyribosylTransferase from Trypanosoma brucei

Article abstract of DOI:10.1002/cctc.201800775

The use of enzymes for the synthesis of nucleoside analogues offers several advantages over multistep chemical methods, including chemo-, regio- and stereoselectivity as well as milder reaction conditions. Herein, the production, characterization and utilization of a purine nucleoside 2’-deoxyribosyltransferase (PDT) from Trypanosoma brucei are reported. TbPDT is a dimer which displays not only excellent activity and stability over a broad range of temperatures (50–70 °C), pH (4–7) and ionic strength (0–500 mM NaCl) but also an unusual high stability under alkaline conditions (pH 8–10). TbPDT is shown to be proficient in the biosynthesis of numerous therapeutic nucleosides, including didanosine, vidarabine, cladribine, fludarabine and nelarabine. The structure-guided replacement of Val11 with either Ala or Ser resulted in variants with 2.8-fold greater activity. TbPDT was also covalently immobilized on glutaraldehyde-activated magnetic microspheres. MTbPDT3 was selected as the best derivative (4200 IU/g, activity recovery of 22 %), and could be easily recaptured and recycled for >25 reactions with negligible loss of activity. Finally, MTbPDT3 was successfully employed in the expedient synthesis of several nucleoside analogues. Taken together, our results support the notion that TbPDT has good potential as an industrial biocatalyst for the synthesis of a wide range of therapeutic nucleosides through an efficient and environmentally friendly methodology.

Full text of DOI:10.1002/cctc.201800775

Accepted Article
Title: Enzymatic synthesis of therapeutic nucleosides using a highly
versatile purine nucleoside 2'-deoxyribosyltransferase from
Trypanosoma brucei
Authors: Elena Pérez, Sánchez-Murcia Pedro, Justin Jordaan, María
Dolores Blanco, Jose Miguel Mancheño, Federico Gago, and
Jesús Fernández-Lucas
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: ChemCatChem 10.1002/cctc.201800775
Link to VoR: http://dx.doi.org/10.1002/cctc.201800775
A Journal of
www.chemcatchem.org

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
Enzymatic synthesis of therapeutic nucleosides using a highly
versatile purine nucleoside 2’-deoxyribosyl-transferase from
Trypanosoma brucei
Elena Pérez,[a] Pedro A. Sánchez-Murcia,[b] Justin Jordaan,[c,d] María Dolores Blanco,[e] José Miguel
Mancheño,[f] Federico Gago,[g] Jesús Fernández-Lucas*[a,h]
Abstract: The use of enzymes for the synthesis of nucleoside
analogues offers several advantages over multistep chemical
Introduction
methods, including chemo-, regio- and stereoselectivity as well
as milder reaction conditions. Herein, the production, characteri-
Nucleosides are known to be ubiquitous molecules in crucial
biochemical processes including genetic information storage
and transfer. There are many examples in the literature of
nucleoside analogues that are therapeutically used as anti-
microbial, antiviral and anticancer agents or employed as
starting materials for the preparation of antisense oligonucle-
otides.[1] These molecules have been traditionally synthe-
zation and utilization of a purine nucleoside 2’-deoxyribosyl-
transferase (PDT) from Trypanosoma brucei are reported.
TbPDT is a dimer which displays not only excellent activity and
stability over a broad range of temperatures (50−70 ºC), pH
(4−7) and ionic strength (0−500 mM NaCl) but also an unusual
high stability under alkaline conditions (pH 8−10). TbPDT is
shown to be proficient in the biosynthesis of numerous therapeu-
sized by different chemical methods through multistep pro-
tic nucleosides, including didanosine, vidarabine, cladribine,
cesses that require protection and de-protection reactions as
fludarabine and nelarabine. The structure-guided replacement of
well as isolation of intermediates.[2] In contrast, the enzymatic
Val11 with either Ala or Ser resulted in variants with 2.8-fold
synthesis of nucleoside analogues is an alternative that
shows many advantages, such as (i) high chemo-, regio- and
stereoselectivity, (ii) the possibility of being achieved in one-
greater activity. TbPDT was also covalently immobilized on
glutaraldehyde-activated magnetic microspheres. MTbPDT3 was
selected as the best derivative (4200 IU/g, activity recovery of
pot, one-step reactions under mild conditions, and (iii) use of
22%), and could be easily recaptured and recycled for >25 reac-
an environmentally friendly technology.[3]
Nucleoside 2′-deoxyribosyltransferases (NDTs) have been
tions with negligible loss of activity. Finally, MTbPDT3 was suc-
cessfully employed in the expedient synthesis of several nucleo-
generally employed in the synthesis of nucleosides because
side analogues. Taken together, our results support the notion
they catalyze the exchange of 2′-deoxyribose between purine
that TbPDT has good potential as an industrial biocatalyst for the
4]
and/or pyrimidine bases.[3b,
According to their substrate
synthesis of a wide range of therapeutic nucleosides through an
efficient and environmentally friendly methodology.
specificity, NDTs have been classified as type I, which trans-
fer the glycosyl moiety from purine 2’-deoxyribonucleosides
to purine bases (Pur↔Pur) such as adenine 1, guanine 2 or
hypoxanthine 3 (Figure 1), and type II, which also accept
pyrimidines as substrates (Pur↔Pur, Pur↔Pyr, Pyr↔Pyr).[3b,
4-5]
For this reason, type I NDTs are also known as purine-
[a]
Dr. E. Pérez, Dr. J. Fernández-Lucas
Applied Biotechnology Group
Universidad Europea de Madrid
E-28670 Villaviciosa de Odón, Madrid (Spain)
E-mail: jesus.fernandez2@universidadeuropea.es
Dr. P. A. Sánchez-Murcia
Institute of Theoretical Chemistry, Faculty of Chemistry
University of Vienna, 1090 Vienna (Austria)
Dr. J. Jordaan
Biotechnology Innovation Centre, Rhodes University
6140 Grahamstown (South Africa)
Dr. J. Jordaan
ReSyn Biosciences, Meiring Naudé Road, Brummeria
0184 Pretoria (South Africa)
Prof. M. D. Blanco
nucleoside 2′-deoxyribosyltransferases (PDTs). Both type I
and type II NDTs are (i) specific for 2′-deoxyribonucleosides,
(ii) regioselective (N-1 and N-9 glycosylation in pyrimidines
and purines, respectively), and (iii) stereoselective (β-
anomers are exclusively formed). However, some studies
have shown that NDTs can tolerate modifications at C2′ and
C3′ of the 2′-deoxyribose sugar.[3b, 4, 6]
[b]
[c]
[d]
[e]
[f]
Department of Biochemistry and Molecular Biology III, School of
Medicine, Universidad Complutense, E-28040 Madrid (Spain)
Dr. J.M. Mancheño
Department of Crystallography and Structural Biology
Rocasolano Institute (CSIC)
E-28006 Madrid (Spain)
[g]
[h]
Prof. F. Gago
Department of Biomedical Sciences and “U. A. IQM-CSIC”
School of Medicine and Health Sciences
University of Alcalá, E-28805 Alcalá de Henares, Madrid (Spain)
Dr. J. Fernández-Lucas
Grupo de Investigación en Desarrollo Agroindustrial Sostenible
Universidad de la Costa, 080002, Barranquilla (Colombia)
Figure 1. Enzymatic synthesis of nucleosides by 2’-deoxyribosyltransferase
immobilized on magnetic microspheres.
Supporting information for this article is given via a link at the end of the
document.
1
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
Despite the advantages of enzyme-mediated bioconver-
sions, their application in industry is limited by several factors,
including: i) the high cost of recombinant enzymes, ii) the low
stability and short lifespan of the enzyme subjected to the
rather harsh conditions often needed in some industrial pro-
cesses, and iii) difficulties in the isolation of the biocatalyst
from the reaction medium for reutilization. Enzyme immobili-
zation may overcome these problems by enhancing protein
stability, facilitating separation of products, and allowing the
biocatalyst to be reused in various reactor configurations. In
this area, our research group has been largely involved in the
development of immobilized biocatalysts for the synthesis of
therapeutic nucleoside analogues.[3b, 4b, 7] Recently, enzyme
binding to magnetic supports has emerged as an alternative
method that provides immobilized biocatalysts which are
easily separated from the reaction medium by applying a
magnetic field.[7-8] Moreover, these magnetic biocatalysts can
be employed in fluidized bed reactors, which are stabilized by
a magnetic field, or in magnetic rod-driven stirred tank reac-
tors, in which deterioration of the biocatalyst due to mechani-
cal agitation is avoided (Figure 1).
Results and Discussion
Production and purification of TbPDT
The pdt gene of Trypanosoma brucei, annotated as encoding
a putative 2'-deoxyribosyltransferase, was cloned and over-
expressed in E. coli BL21(DE3). The recombinant N-terminal
His6-tagged TbPDT was purified by using two chromato-
graphic steps. SDS-PAGE analysis of the purified enzyme
showed a single protein band with an apparent molecular
mass of 19 kDa (Figure S1, Supporting Information). N-
terminal His6-tagged TbPDT was predicted to contain 169
amino acid residues and to possess a relative molecular
mass
of
19.09
kDa
(Protparam,
http://web.expasy.org/protparam). These results agree with
those described by Bosch et al.[10] who reported the crystal
structure of the same protein but did not carry out any func-
tional characterization that could support its activity as a true
NDT.
Sedimentation velocity experiments reveal TbPDT as a
single species with an experimental sedimentation coefficient
of 3.17 S (s20,w = 3.19) compatible with an obligate homodi-
mer of 39.07 kDa molecular weight. Thus, the single mono-
mer affords a value of 19.53 kDa, which is very close to the
molecular mass calculated from the amino acid sequence
(19.09 kDa). In this respect, different oligomeric states have
been described for various bacterial and protozoan NDTs.
On the one hand, NDTs from Lactobacillus leichmanii, L.
helveticus (LhNDT) and L. reuteri (LrNDT) appear as hex-
amers (displaying a molecular weight ranging from 105 to
120 kDa),[4, 11] whereas a type II NDT from Lactococcus lactis
was characterized as a tetramer of 69 kDa.[12] On the other
hand, we recently reported LmPDT as a dimer[9] whose ac-
tivity is similar to that of its bacterial counterparts (Table 1).
These findings support the view that a higher oligomerization
state for NDTs is not essential for catalysis and/or stabiliza-
tion of the active conformation.
In our search for PDTs suitable as biocatalysts for the syn-
thesis of non-natural therapeutic nucleosides, we found a
good candidate in an uncharacterized protein encoded in the
genome of Trypanosoma brucei (TbPDT) that shares 67%
sequence identity with a previously reported PDT from
Leishmania mexicana (LmPDT).[9] The putative enzyme was
produced and characterized using both biophysical and
biochemical methods. Studying the effects of pH, tempera-
ture and ionic strength on enzyme activity and stability al-
lowed us to select a set of optimal operative conditions for
the synthesis, in aqueous media, of several nucleoside ana-
logues with therapeutic application (4−10 in Figure 2). In
addition, structure-guided site-directed mutagenesis studies
led us to produce two TbPDT variants which showed en-
hanced activity over that of the wild-type enzyme in the syn-
thesis of 2’-deoxyadenosine. Finally, to explore its potential
as a biocatalyst, TbPDT was immobilized onto magnetic
microspheres, which were physically characterized by scan-
ning electron microscopy (SEM) and dynamic light scattering
(DLS).
Biochemical characterization of TbPDT
TbPDT is highly active (>63 %) across a broad temperature
o
o
range (from 50 to 70 C), with an optimum at 60 C (Figure
3A). The pH profile reveals that TbPDT is more than 60 %
active over a broad pH range (from 4 to 7) and displays
maximum activity at a pH between 5 and 6 (Figure 3B). In
addition, negligible loss of activity is observed in the pres-
ence of 500 mM NaCl, and only a slight decrease is ob-
served when this concentration is increased (i.e. 70% activity
remains at 1 M NaCl) (Figure 3C).
The effect of pH on TbPDT stability was evaluated by incu-
o
bating the enzyme for 30 h at 50 C at different pH values
(ranging from 6.5 to 10). The results suggest that alkaline
environments (pH 8-10) do not inactivate the enzyme during
this time period to a significant extent (Figure 3D). This unu-
sual tolerance to alkaline conditions was unexpected for a
mesophilic enzyme and is a very interesting advantage for
enzyme immobilization, as will be described below.
Figure 2. Chemical structures of the most energetically stable tautomeric
forms of (i) purine bases adenine (1), guanine (2) and hypoxanthine (3),
(ii) 2’-deoxyribonucleosides cladribine (4), 2F-dAdo (5) and 6-MPdRib (6),
(iii) arabinosyl nucleosides vidarabine (7), fludarabine (8) and nelarabine
(9), and (iv) 2’,3’-dideoxyinosine (10).
The stability of TbPDT as a biocatalyst was assessed by
showing that the enzyme stored at 4 °C still displays 85%
relative activity after 180 days (data not shown). Nonetheless,
this percentage drops to 23% when TbPDT is stored at -20
oC or -80 oC.
2
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
Table 1. Synthesis of natural nucleosides catalyzed by TbPDT and other PDTs.
Specific activity (IU/mg protein)
LhNDT[a] BbPDT[a]
TbPDT
LmPDT[a]
Acceptor[b]
Ade
Gua
Hyp
Ade Gua Hyp
Ade Gua Hyp
Ade Gua Hyp
Donor
dAdo
dGuo
dIno
-
10.5±2.3 9.5±2.4
-
n.d. 0.06
-
0.09 0.02
-
13.8 75.4
21.4±3.1
-
39.0±2.6
-
0.05
-
0.05
-
0.08
-
0.02
-
52.7
-
68.6
-
38.5±3.3 11.0±2.1
0.08 n.d.
0.04 0.04
73.1 14.5
[a] Activities for LhNDT, BbPDT and LmPDT were taken from refs. 4a, 5 and 9, respectively.
[b]
Reaction conditions were 0.3 μg of enzyme in 40 μL at 50 oC for 5 min. Ade and Hyp concentrations were 10 mM in 50 mM sodium phos-
phate buffer, pH 6.0, whereas that of Gua was 1 mM in 50 mM borate buffer, pH 8.5; n.d.: not detected.
Substrate specificity and structural considerations
Our experimental data show that TbPDT should be classified
as a type I NDT since it exclusively catalyzes the transfer of
LmPDT,[9] because NDTs from Lactobacillus have an aspar-
tic acid at the equivalent position, which means that this side
chain needs to be protonated as a carboxylic acid.[15] The
occurrence of Asn53 in TbPDT (and Asn56 in LmPDT) in a
position usually occupied by Asp in other NDTs may be
advantageous for resilience at high pH values.
2´-deoxyribose between purine bases. In common with other
5, 9]
previously described PDTs,[4a,
TbPDT recognizes 2’-
deoxyinosine (dIno) and hypoxanthine (Hyp) as the best
nucleoside donor and base acceptor, respectively (Table 1).
Because type II NDTs offer several advantages over PDTs,
such as higher activity values under the same experimental
conditions or the possibility of using a broad range of sub-
strates, type II NDTs usually represent the best option for the
enzymatic synthesis of nucleoside analogues. Nevertheless,
the specific activities of TbPDT are higher than those report-
ed for other PDTs from Lactobacillus and Borrelia species
under similar enzymatic assay conditions (Table 1) and com-
parable to those previously described for several type II
NDTs from Lactobacillus reuteri[4b] and Lactobacillus del-
brueckii.[13] TbPDT can also use as substrates the manifold of
purine derivatives and nucleoside analogues displayed in
Tables 2 and 3, and it is to be preferred, in principle, over
LmPDT[9] because it can transglycosylate on a larger number
of purine analogues when dIno is used as the donor nucleo-
side (data not shown). These results encouraged us to con-
sider TbPDT a good candidate for the enzymatic synthesis of
nucleoside analogues.
TbPDT is a homodimer which displays the typical 2′-
deoxyribose binding site motif made up of two acidic residues
from one subunit (Asp76 and Glu82) and one polar residue
from the characteristic NLM motif in the neighboring mono-
mer (Asn126#). The 5’-hydroxyl group of the incoming 2’-
deoxynucleoside hydrogen bonds to the carboxylate of
Asp76 and is hydrogen bonded by the carboxamide of
Asn126#. The nucleophilic attack on C1’ of the sugar is ef-
fected by the side-chain carboxylate of Glu82, which is held
in position by hydrogen bonds donated by the phenol group
of Tyr5 and the 3’-hydroxyl of the sugar (Figure 4). In the
near-attack conformation,[14] one of the lone pairs of the
M128# sulfur is poised to interact with the developing positive
charge on the oxocarbenium ion of the transition state. The
purine ring is sandwiched between the hydrophobic side
chains of Val11 and Phe12, on one side, and those of Ile49
and Leu127# on the other side, whereas the nucleobase’s N-
3 atom is recognized by the carboxamide nitrogen of Asn53.
This is a rather unique feature of PDTs from trypanosomatids
(family Trypanosomatidae), as recently reported for
The results of our molecular dynamics (MD) simulations
reveal that, in the Michaelis complex (Figure 4B), the flexible
39TDNEATEA46 and 116LRVEGF121 loops connecting β2-α3
and β6-α6, respectively, become more ordered because of
the involvement of the side chain of Glu42 in base recogni-
tion (Figure 4A) and the stacking of the phenyl ring of
Phe121# on Ile49 and Leu127#, which results in the approx-
imation of Glu119# towards the N-7 position of the base. In
this “closed” conformation, the substrate is placed within a
preorganized active site in an orientation conducive to reac-
tion. Thereafter, we postulate that glycosydic bond cleavage
by nucleophilic attack of Glu82 on C1’ is accompanied by
proton transfer from Glu119# to N-7 so that this tautomer is
initially generated, in agreement with early data,[16] prior to
departure of the energetically more stable N-9 tautomer.[17]
To understand the structural determinants of TbPDT stabili-
zation, the dimer interface was analyzed with the aid of pro-
grams MM-ISMSA[18] and HBplot,[19] which provided valuable
information about those residues that contribute the most to
stabilizing the dimer. The former dissected, on a per-residue
basis, the monomer-monomer binding energy (Figure 5),
including desolvation, and the latter analyzed the types of
hydrogen bonds (backbone-backbone, side chain-backbone,
and side chain-side chain) between the two monomers (Fig-
ure S2). Our analyses reveal the importance of Phe12, Arg50,
Arg71, Phe81, Leu116, Arg117, and Met128 for dimer stabili-
zation. In this respect, it is noteworthy that Arg71 is part of an
insertion particular to trypanosomatids that is located in a
region which is thought to largely determine the dimeric,
tetrameric or hexameric functional state of NTDs.[9]
In addition to these non-bonded interactions, we note that
Cys77 and Cys77# are spatially close enough to be involved
in a disulfide bridge, despite the lack of electron density
between the sulfur atoms in publicly available X-ray crystal
structures of both TbPDT and LmPDT (PDB entries 2A0K[10]
and 5NBR,[9] respectively). This is most likely due to the fact
3
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
Figure 3. Dependence of TbPDT activity and stability on pH and temperature. A) Effect of temperature on TbPDT activity. B) Effect of pH on TbPDT activity,
(∆) sodium citrate 50 mM (pH 4-6), (●) sodium phosphate 50 mM (pH 6-8), (○) sodium borate 50 mM (pH 8-10). C) Effect of ionic strength on TbPDT activity,
and D) Time course of the thermal inactivation of TbPDT at 50 °C in 10 mM sodium phosphate pH 6.5 (●), and 10 mM sodium borate pH 8 (○) and pH 10
(▲).
Table 2. Synthesis of nucleoside analogues from purine base derivatives
catalyzed by TbPDT.
catalyzes most of the transfer reactions assayed with moder-
ately good conversion rates at short reaction times. In this
regard, the syntheses of therapeutic nucleosides such as 2-
Donor
dIno
Acceptor
Product
Conversion (%)
FdAdo (5, a prodrug used in suicide gene therapy for cancer
treatment)[21] and cladribine (4, an FDA-approved drug for the
treatment of hairy-cell leukemia),[22] or 6MPdRib (6, a precur-
sor of 6-MP, a Hyp analogue used in the treatment of child-
hood acute lymphocytic leukemia)[22] were undertaken as test
cases.
2-FAde
2-ClAde
2,6-DAP
6-ClPur
6-MP
6-MetPur
6-MeOGua
Bzm
2-FdAdo
2-CldAdo
49±1
26±1
40±2
26±0
26±1
31±4
53±3
32±3
n.d.
2,6-DAPdRib
6-ClPurdRib
6-MPdRib
6-MetPurdRib
6-MeOdGuo
BzmdRib
Theo
TheodRib
Over the last decades, C2´- and/or C3´-modified nucleo-
sides have received a great deal of attention in connection
with the quest for new antineoplastic drugs and antiviral
agents (particularly against HIV).[1, 21] Prominent among them
are vidarabine (7), fludarabine (8), nelarabine (9), ddI (10),
zidovudine (3´-azido-2´,3´-dideoxythymidine, AZT), zalcita-
bine (2´,3´-dideoxycytidine, ddC), stavudine (2′,3′-didehydro-
2´,3´-dideoxythymidine, d4T), lamivudine (2′,3′-dideoxy-3′-
Reaction conditions: 0.3 μg of enzyme in 40 μL at 50 oC, 20 min. [Sub-
strates] = 1 mM, 50 mM sodium phosphate buffer, pH 6.0. n.d. not de-
tected
that these highly homologous proteins were expressed in the
cytoplasm of E. coli cells, where the thioredoxin/thioredoxin
reductase and glutathione/glutaredoxin pathways provide a
reducing environment that prevents Cys oxidation.[20] It is
conceivable that, in an oxidizing environment, this intersubu-
nit disulfide would further contribute to dimer stabilization.
thiacytidine,
emtricitabine
(2',3'-dideoxy-5-fluoro-3'-
thiacytidine), and other sugar-modified nucleosides of natural
bases.[1, 21]
Given the relative tolerance of TbPDT to various modifica-
tions at positions C2´ and C3´ of the deoxyribose moiety, we
tested its ability to act as a transglycosylation biocatalyst for
the synthesis of 2´-fluoro-2´-deoxyribosyl, 2´,3´-dideoxy-
ribosyl and arabinosyl nucleosides containing natural purines
(Table 3). To illustrate the therapeutic importance of these
molecules, suffice it to say that ddG is a selective inhibitor of
HIV replication in vitro and an active anti-hepadnavirus nu-
cleoside analogue,[23] ddI (10) is a potent inhibitor of HIV
Enzymatic synthesis of non-natural nucleosides
To explore the potential of TbPDT as a biocatalyst, the en-
zymatic synthesis of several nucleoside analogues was
carried out. A variety of natural and non-natural nucleosides
and bases were employed under different assay conditions.
To determine whether TbPDT can synthesize nucleoside
analogues from natural nucleosides and purine base ana-
logues, the best nucleoside substrate, dIno (Table 1), was
selected as the base donor. As shown in Table 2, TbPDT
4
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
A
B
Figure 4. A. Schematic drawing of the binding of 2’-deoxyinosine (dIno) within the active site of TbPDT. B. Detail of a 3D model of the TbPDT active site (C
atoms colored in green in one monomer and in cyan in the other) with a bound dIno molecule (C atoms in orange) in the near-attack conformation.[14] Enzyme
residues making up the active site and relevant for catalysis are labeled and shown as sticks.
Site-directed mutagenesis and kinetic characterization of
the wild-type TbPDT and engineered variants
Our MD simulations suggested that Val11, a non-catalytic
residue that is positionally close to the interfacial Phe12,
provides a hydrophobic environment for binding of the heter-
ocyclic ring system of both the nucleoside donor and the
acceptor base. In doing so the bulky side chain of Val11
plays a dual role; on the one hand, by stacking on the het-
erobase, it contributes to the binding energy; on the other
hand, its apolar character prevents the approach of incoming
water molecules that might be activated for attack on the
deoxyribosyl-enzyme intermediate to release the bonded
sugar.
Figure 5. Binding energy analysis of the TbPDT dimer interface. The bars
corresponding to residues from the NLM motif are colored in red whereas
that belonging to the catalytic Glu residue is colored in green. Those residues
involved in inter-subunit hydrogen bonding interactions are labeled in blue.
For simplicity, a cutoff of 5 kcal mol-1 was used.
To assess this dual role, two variants (TbPDTV11A and
TbPDTV11S) were generated by single amino acid substitu-
tions at position 11 and their transferase activities were de-
termined using Ade as the acceptor base and either dIno or
dGuo as the nucleoside donor. The specific activity of each
of these two variants was >2-fold that of the wild-type en-
zyme (Table 4). To gain further insight into the factor(s) re-
sponsible for these differences, steady-state kinetic studies
were performed. Our results show that the rates of catalysis
for the two half-reactions, i.e. nucleobase release and
transglycosylation, are enhanced in both variants, despite the
higher KM values, which probably reflect the expected loss in
van der Waals and hydrophobic interactions between en-
zyme and substrates (Table 5). Since enhanced transition
state stabilization associated to increased flexibility of the
active site is the single factor thought to be a requirement for
reducing the free energy barrier and accelerate the enzymat-
ic reaction,[27] this must be the overall effect of replacing
Val11 with the less bulky Ala or Ser residues. Nonetheless,
further experimental and theoretical work will be necessary to
ascertain the origin of the improved catalytic efficiencies of
these variants.
replication,[24] vidarabine (7) is not only an antiviral drug
active against herpes simplex and varicella zoster viruses but
also a potential precursor of fludarabine and clofarabine, both
25]
FDA-approved drugs for cancer treatment,[1a,
dGuo is an anti-influenza virus agent.[26]
and 2´F-
Likewise, the considerable latitude of TbPDT in nucleobase
recognition was advantageously used to produce a further
set of 2’,3’-modified nucleoside analogues containing substi-
tuted purines, such as 2-FAde, 2-ClAde, 2,6-DAP, 6-ClPur,
6-MP, 6-MetPur, 6-MeGua (Table 3). As therapeutically
relevant examples, we highlight nelarabine (9), an ara G
prodrug, and fludarabine (8), both of which are FDA-
approved chemotherapeutic agents for the treatment of sev-
eral malignancies.[22]
Noteworthy, these results demonstrate, for the first time to
the best of our knowledge, that an NDT can successfully
synthesize ara 2-FAde, ara 2-ClAde, ara 6-ClPur, ara 6-
MeOGua, dd2,6-DAPRib, dd6MRib, dd2-FAdo, dd2-ClAdo,
dd6-ClPurRib, dd6-MeOdGuo, 2'F-2,6-DAPdRib, 2'F-6-
MdRib, 2'F-2-FdAdo, 2'F-2-CldAdo, 2'F-6-ClPurdRib, and
2'F-6-MeOdGuo.
5
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
Table 3. Synthesis of nucleoside analogues from 2’,3’-modified
nucleosides catalyzed by TbPDT.
Immobilization of recombinant TbPDT onto glutaralde-
hyde-activated MagReSyn®Amine microspheres Since
immobilization is an essential requirement for the use of
enzymes as industrial biocatalysts, we assayed the stability
of TbPDT under immobilization conditions. Covalent immobi-
lization of proteins is usually achieved via multipoint covalent
attachment through the region with the highest density of
primary amino groups, especially the ε-NH2 of lysine resi-
dues.[28] These covalent immobilization techniques usually
employ long reaction times (4−10 h) and alkaline conditions
(pH 8−10). For this reason, we thought that the exceptional
stability displayed by TbPDT under alkaline conditions could
offer a substantial advantage for efficient immobilization on
magnetic particles. TbPDT immobilization was carried out by
covalent attachment of the enzyme to glutaraldehyde-
activated MagReSyn®Amine microspheres (Figure 6).
Different immobilized biocatalysts (MTbPDT1-MTbPDT5)
were prepared by increasing the enzyme load and MTbPDT3
was selected as the optimum construct (Table 6) because it
showed a remarkably high activity of 4200 IU/g of wet biocat-
alyst (in the synthesis of dAdo from dIno and adenine at
50 °C and 50 mM sodium phosphate pH 6) and an activity
recovery[29] of 22 %. Noteworthy, MTbPDT3 displays the
highest activity ever described for an immobilized NDT. The
calculated specific activities for MTbPDT3 are much higher
(between 64.2 and 1615 times) than those reported for other
NDTs, such as that from LrNDT immobilized on epoxy-
activated Sepabeds® for the enzymatic synthesis of dAdo
(65.4 IU/g of wet biocatalyst)[30] or that from Bacillus psy-
chrosaccharolyticus (BpNDT) immobilized on PEI 600 Da
agarose followed by crosslinking with 70 % oxidized dextran
(20 kDa) (11.3 IU/g of wet biocatalyst).[31] Other similar ex-
amples, such as the enzymatic synthesis of thymidine by
LrNDT immobilized on magnetic chitosan beads (2.10 IU/g of
wet biocatalyst)[7] or the enzymatic synthesis of halogenated
nucleosides by an NDT from Lactobacillus animalis (LaNDT)
immobilized on DEAE-Sepharose (2.6-5.4 IU/g of wet biocat-
alyst)[32] displayed very low activity values compared to
MTbPDT3.
Donor
Acceptor[a]
Product
Conversion
(%)
ddA
Gua
ddG
24±3
27±1
16±2
n.d.
16±2
10±3
2
Hyp
ddI
2-FAde
2-ClAde
2,6-DAP
6-ClPur
6-MP
dd2-FAdo
dd2-ClAdo
dd2,6-DAPRib
dd6-ClPurRib
dd6-MPRib
dd6-MetPurRib
dd6-MeOGuo
ddA
6-MetPur
6-MeOGua
Ade
6±1
18±1
17±1
20
ddI
Gua
ddG
2-FAde
2-ClAde
2,6-DAP
6-ClPur
6-MP
dd2-F-dAdo
dd2-CldAdo
dd2,6-DAPRib
dd6-ClPurRib
dd6-MPRib
dd6-MetPurRib
dd6-MeOGuo
ara G
-
n.d
10
6
2
6-MetPur
6-MeOGua
Gua
4
7
ara A
ara G
2'F-dIno
43±1
55±1
23
Hyp
ara H
2-FAde
2-ClAde
2,6-DAP
6-ClPur
6-MP
ara 2-FAdo
ara 2-ClAdo
ara 2,6-DAP
ara 6-ClPur
ara 6-MP
13
29
7
n.d.
15
6-MetPur
6-MeOGua
Ade
ara 6-MetPur
ara 6-MeOGuo
ara A
38±1
37±1
50±1
37
Hyp
ara H
2-FAde
2-ClAde
2,6-DAP
6-ClPur
6-MP
ara 2-FAdo
ara 2-ClAdo
ara 2,6-DAP
ara 6-ClPur
ara 6-MP
17±1
28
8
21
6-MetPur
6-MeOGua
Ade
ara 6-MetPur
ara 6-MeOGuo
2'F-dAdo
14
52
29
Gua
2'F-dGuo
36
2-FAde
2-ClAde
2,6-DAP
6-ClPur
6-MP
2'F-2-F-dAdo
2'F-2-CldAdo
2'F-2,6-DAPdRib
2'F-6-ClPurdRib
2'F-6-MPdRib
2'F-6-MetPurdRib
2'F-6-MeOdGuo
21
In addition, MTbPDT3 could be recycled after at least 25
consecutive batch reactions with negligible loss of catalytic
activity in the synthesis of dAdo from dIno and Ade.
15
15
4
A physical characterization of magnetic particles can be of
help to establish the optimal conditions in the bioreactor.
SEM micrographs of MTbPDT3 (Figure S3) showed individu-
al particles in a micro-size scale and with a rough surface.
This irregular surface might enhance enzyme attachment and
expand the number of internal spaces available for substrate
binding. Thus, this carrier material might be close to ideal
due to its high accessible surface area and the nanoscaled
intra-pore system that allows both stable enzyme binding and
high flow permeability.[33] Furthermore, DLS analysis of these
systems revealed a Z-average size between 24.6-113.42 µm
and 4.96-19.50 µm for the free MagReSyn®Amine beads and
MTbPDT3, respectively. As Manoel et al. demonstrated with
10±1
4
6-MetPur
6-MeOGua
15
[a] Reaction conditions were 8 μg of enzyme in 40 μL at 50 oC for 4 h.
Substrate concentration was 1 mM in 50 mM sodium phosphate buffer, pH
6.0, except for Gua, in which case it was 10 mM in 50 mM sodium borate
buffer, pH 8.5.
Table 4. Enzymatic activities of wild-type and variant TbNDTs in two
transglycosylation reactions.
Specific activity (IU/mg protein)
dIno + Ade a
dGuo + Ade a
poly(styrene-co-divinylbenzene)-based
supports,[34]
this
Wild type
V11A
38.5 ±4.2
84.7 ±2
90.7 ±3
21.4 ±3
50.3 ±2
59.9 ±2
micrometric size is normally desirable to enable the industrial
implementation of biocatalysts because it avoids the use of
complex methods for separation of the biocatalyst from the
reaction media and the diffusional problems associated to
V11S
a Reaction conditions: 0.3 μg of enzyme in 40 μL at 50 oC, 5 min.
[Substrates] = 10 mM, 50 mM sodium phosphate buffer, pH 6.0.
6
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
Table 5. Kinetic parameters of wild-type and variant TbPDTs for the Glu82-assisted cleavage of the C1‘-N9 glycosidic bond of Ino (first half-
reaction, leading to enzyme glycosylation and release of free Hyp) and for the overall conversion of dIno+Ade to dAdo+Hyp
(transglycosylation reaction).
Parameter
Wild type
V11A
V11S
Wild type
V11A
V11S
Variable dIno [a]
Variable Ade [a]
Nucleobase release
KM (mM)
kcat (s-1)
0.7±0.2
7.9 ±1.3
5.1±1.1
39.6±4.0
6.8±1.6
40.7±3.9
0.6±0.2
8.7±0.95
2.7±0.9
21.5±1.6
1.4±0.4
19.6±1.9
kcat/KM (s-1 M-1)
11314
7765
5985
14500
7962
14000
Vmax
0.21 ±0.01
1.05±0.06
1.08±0.09
0.23±0.04
0.57±0.10
0.52±0.09
Transglycosylation
KM (mM)
kcat (s-1)
0.2±0.1
6.4±1.3
1.2±0.4
14.3±3.4
2.0±0.9
17.0±2.6
1.8±0.2
10.2±1.1
2.9±0.7
23.4±2.3
2.9±0.6
23.8±2.9
kcat/KM (s-1 M-1)
30476
11440
8415
5600
8068
8206
Vmax
0.17±0.01
0.38±0.03
0.45±0.06
0.27±0.04
0.62±0.04
0.63±0.05
a Reaction conditions: 0.3 μg of enzyme in 40 μL at 50 oC, 5 min. [Fixed substrate] = 5 mM, [Variable substrate] = 0.25-15 mM, 50 mM
sodium phosphate buffer, pH 6.0.
Figure 6. Schematic representation of immobilization of recombinant TbPDT on glutaraldehyde-activated MagReSyn®Amine microspheres.
Table 6. Effect of enzyme/support mass ratio on the activity of TbPDT immobilized on glutaraldehyde-activated MagReSyn®Amine microspheres.
Derivative
Biocatalyst load
Immobilization yield
Activity
(IU/gsupport
-
2027±70
2482±55
4200±27
4400±45
3100±30
Specific activity
(IU/mg protein)
38.5±3.3
Recovery
(%)
(mgenz/gsupport
)
(%)c
-
)
TbPDTa
-
-
MTbPDT1b
MTbPDT2b
MTbPDT3b
MTbPDT4b
MTbPDT5b
152
310
620
928
992
100±2
93±4
91±2
80±0
62±2
13.1±0.8
9.6±0.8
8.5±0.5
5.4±0.6
34±2
25±1
22±2
14±4
13±1
5.0±0.5
a Reaction conditions: 0.3 μg of enzyme in 40 μL at 50 oC for 5 min. [dIno]=[Ade]=10 mM in 50 mM sodium phosphate buffer, pH 6.0, at 50 ºC, 5
min, 300 r.p.m Vr = 40 μL (0.4 µmol dIno=A). b Reaction conditions: 2.64 μg enzyme (2.6–26.4 μg beads), [dIno]=[A]=10 mM in 50 mM sodium
phosphate buffer, pH 6.0 at 50 oC, 10 min, 300 r.p.m. Vr = 320 μL (3.2 µmol dIno=A). c Percentage of immobilized enzyme.
very small or very large particles, respectively. These results
Table 7. MTbPDT3-catalyzed synthesis of nucleoside analogues.
reveal that the enzyme displayed on the surface of magnetic
Donor
Acceptor
Product
Activity
(IU/gsupport
Nucleoside
conversion
(%)
)
devices can reduce the presence of aggregates in water disper-
sion and, therefore, promote the stability of individual particles.
This explanation is in accordance with zeta potential measure-
ments.
dGuo
2-FAde
2-ClAde
2,6-DAP
6-MeOGua
2-FdAdo
2-CldAdo
2,6-DAPdRib
6-MeOdGuo
3424±30
3521±35
3057±50
5101±75
40±2
41±2
33±5
46±3
Thus, the zeta potentials of free MagReSyn®Amine beads were
between -0.5 and -0.8 mV and decreased to -64.1 and -67.6 mV
upon enzyme immobilization due to surface charge modification.
The measurement conditions and the immobilized enzyme seem
to reduce the flocculating behavior of free MagReSyn®Amine
beads. The zeta potential is used mainly as an indicator of elec-
trostatic stabilization of colloidal particles, in agreement with the
DLVO theory,[35] according to which large magnitude zeta poten-
tial values, such as those measured for the MTbPDT3 beads,
indicate that an electrostatically stabilized suspension will re-
main stable, unlike free MagReSyn®Amine beads, which have a
zeta potential close to zero.
Finally, as a proof of concept, MTbPDT3 was successfully
employed in the one-pot, one-step synthesis of several nucleo-
side analogues of pharmaceutical interest using dGuo as the
nucleoside donor with conversion rates approaching 50% (Table
7).
Reaction conditions: 2.64 μg μg prot (4.25 μg beads), [dGuo]=[Base]=1 mM
in 50 mM sodium phosphate buffer, pH 6.0 at 50 oC, 20 min, 300 r.p.m.
Vr = 320 μL (3.2 µmol dIno=Base).
Conclusions
The presence of NDTs in Trypanosoma species has not been
ascertained yet, despite a report[10] on the crystal structure of a
protein annotated in bioinformatic databases as a putative NDT
from Trypanosoma brucei, the function of which has not been
described, to the best of our knowledge. We now report the
cloning of the pdt gene, its expression in E. coli, and the purifica-
tion and biochemical characterization of the encoded protein,
TbPDT. Our experimental results reveal TbPDT as a homodi-
meric enzyme endowed with significant activity in a broad range
o
of temperature (50-70 C), pH (4−7) and ionic strength (0-500
7
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
Sedimentation coefficient distributions were calculated by least-squares
boundary modeling of sedimentation velocity using the continuous distri-
bution c(s) Lamm equation model as implemented in SEDFIT 14.7g.[38]
(https://sedfitsedphat.nibib.nih.gov/software/default.aspx). Baseline off-
sets were measured afterwards at 200,000 × g. The experimental sedi-
mentation coefficients were corrected to standard conditions (water, 20
oC, and infinite dilution) using SEDNTERP software to obtain the corre-
sponding standard values (s20,w).[39]
mM) values. Additionally, TbPDT was found to exhibit great
stability in a pH interval from 6.5 to 10.0 and display negligible
loss of activity at 50 oC for 30 h. The potential of TbPDT for
biocatalysis was demonstrated by achieving the enzymatic
synthesis of numerous therapeutic nucleosides, including 2-
ClAdo (4), vidarabine (7), fludarabine (8), nelarabine (9), and ddI
(10), among others (Tables 2 and 3).
The covalent immobilization of TbPDT on glutaraldehyde-
activated MagReSyn®Amine microspheres led to derivative
MTbPDT3, which displayed a remarkable activity of 4200 IU/g of
wet biocatalyst and retained 22% activity in the enzymatic syn-
thesis of dAdo from dIno and adenine. Furthermore, MTbPDT3
was reutilized for at least 25 times without any apparent loss of
catalytic activity, an essential prerequisite for its application in
industrial bioreactors, and allowed the expedient enzymatic
synthesis of several nucleoside analogues at moderate conver-
sion rates.
Enzyme activity assay
The standard activity assay was performed by incubating 0.3 μg of pure
enzyme with 10 mM 2′-deoxyinosine (dIno) and 10 mM adenine in 50 mM
sodium phosphate buffer pH 6.5 in a final volume of 40 μL. The reaction
mixture was incubated at 40 °C for 10 min (300 rpm). The enzyme was
inactivated by adding 40 μL of cold methanol in an ice bath and heating
for 5 min at 100 °C. After centrifugation at 9000 × g for 2 min, the sam-
ples were half-diluted with water and frozen at -20 oC. Nucleoside pro-
duction was analyzed using HPLC to measure the reaction products
quantitatively, as described below. All determinations were carried out in
triplicate and the maximum error was less than 5%. Under such condi-
tions, one international activity unit (IU) was defined as the amount of
enzyme producing 1 μmol/min of 2′-deoxyadenosine under the assay
conditions.
Experimental Section
Materials
Influence of pH and temperature on enzyme activity
Cell culture medium reagents were from Difco (St. Louis, United States).
Trimethyl ammonium acetate buffer was purchased from Sigma-Aldrich
(Madrid, Spain). All other reagents and organic solvents were purchased
from Scharlab (Barcelona, Spain) and Symta (Madrid, Spain). Nucleo-
sides and nucleobases used in this work were provided by Carbosynth
Ltd. (Compton, United Kingdom).
The pH profile of purified recombinant enzyme was initially determined
using the standard assay, as described above, with sodium citrate (pH 4-
6), sodium phosphate (pH 6-8) and sodium borate (pH 8-10) as reaction
buffers (50 mM). The optimum temperature was determined using the
standard assay across a 20-90 oC range. A similar approach was fol-
lowed to characterize the effect of ionic strength on enzyme activity.
TbPDT activity was measured at different concentrations of NaCl, rang-
ing from 0 to 1 M.
Gene Expression and Protein Purification
The pdt gene, which encodes a protein annotated as nucleoside 2′-
deoxyribosyltransferase from Trypanosoma brucei gambiense (European
Nucleotide Archive code: AAX70453.1; UniProtKB UniProtKB Q57VC7,
was purchased from Genscript (USA). The coding sequence appeared
as a NdeI-EcorI fragment subcloned into the expression vector pET28b
(+). The resultant, recombinant vector pET28bTbPDT provided an N-
terminal His6-tagged fusion with a thrombin cleavage site between the
tag and the enzyme. TbPDT was expressed in E. coli BL21(DE3) grown
in LB medium at 37 °C with kanamycin 50 μg/mL. Protein overexpression
was induced by adding 0.5 mM isopropyl β-D-1-thiogalactopyranoside
and the cells were further grown for 3 h. These were harvested via cen-
trifugation at 3500 × g. The resulting pellet was resuspended in 10 mM
sodium phosphate buffer pH 7. Crude extracts were prepared by French
press lysis of cell suspensions. The lysate was centrifuged at 17500 × g
for 30 min and the supernatant was filtered through a 0.22 μm filter
(Millipore). The cleared lysate was loaded onto a 5-mL HisTrap FF col-
umn (GE Healthcare) pre-equilibrated in a binding buffer (20 mM Tris-
HCl buffer, pH 8.0, with 100 mM NaCl and 10 mM imidazole) and the
column was washed. Bound proteins were eluted using a linear gradient
of imidazole (from 10 to 500 mM). Fractions containing TbPDT were
identified by SDS-PAGE, pooled, concentrated and loaded onto a HiLoad
16/60 Superdex 200 prep grade column (GE Healthcare) pre-equilibrated
in 50 mM sodium phosphate, pH 7.0. Fractions with the protein of interest
identified by SDS-PAGE were pooled and the protein was dialyzed
against 10 mM sodium phosphate, pH 7.0, and concentrated and stored
at 4 oC until its use. Electrophoresis was carried out on a 15% poly-
acrylamide slab gel with 25 mM Tris-HCl buffer, pH 8.6, 0.1% SDS.[36]
Protein concentration was determined spectrophotometrically by UV
absorption at 280 nm using 280= 8,940 M-1cm-1.[37]
Thermal and pH stability of TbNDT
TbPDT was stored at 4 °C in 10 mM sodium phosphate, pH 7.0 for 135
days. Samples were taken periodically for enzymatic activity evaluation.
Storage stability was defined as the relative activity between the first and
successive reactions. Moreover, the thermal stability of TbNDT was
assessed by incubating 0.3 μg of pure enzyme in a pH range from 6.5 to
10, at 50 oC for a period of 30 h. Thereafter, the activity was measured
using the standard assay.
Substrate specificity
To explore the substrate specificity of TbPDT, 0.3 μg of recombinant
enzyme were incubated with 10 mM of purine and pyrimidine nucleosides
and bases, in 50 mM sodium phosphate buffer pH 6.0 or 50 mM sodium
borate pH 8.5 at 50 °C and 300 rpm orbital shaking at different reaction
times. Samples were extracted at regular intervals and enzymatic activity
was evaluated as described in the Analytical methods section.
Enzymatic synthesis of nucleoside analogues by soluble TbPDT.
The synthesis of non-natural nucleosides was achieved at variable
amounts of enzyme, using different nucleoside and base analogues.
Reactions were carried out incubating 0.3 − 8 μg of TbPDT with 1 mM
nucleoside and base, in 50 mM sodium phosphate buffer pH 6.0 or
sodium borate pH 8.5 at 50 °C and 300 rpm orbital shaking at different
reaction times. Samples were taken at regular intervals and enzymatic
activity was evaluated as described below.
Kinetic parameters and initial velocity pattern
The steady-state kinetic parameters, KM, kcat and Vmax, were determined
under standard assay conditions at varying concentrations of one sub-
strate (dIno or Ade, 0.25-15 Mm), while the concentration of the other
substrate was fixed at constant saturating level (dIno or Ade, 5 mM).
Apparent KM, kcat and kcat/KM values were determined by non-linear
regression assuming Michaelis–Menten kinetics.
Analytical ultracentrifugation analysis
Sedimentation velocity experiments for TbNDT were carried out in 10
mM sodium phosphate (pH 8.0, 20 °C, 50,000 × g) using an Optima XL-I
analytical ultracentrifuge (Beckman-Coulter Inc.) equipped with UV-VIS
absorbance and Raleigh interference detection systems, using an An-
60Ti rotor and standard (12 mm optical path) double-sector centre pieces
of Epon-charcoal. Sedimentation profiles were recorded at 292 nm.
Enzyme immobilization
8
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
TbPDT immobilization was carried out by covalent attachment of the
enzyme to glutaraldehyde-activated MagReSyn®Amine microspheres
(from ReSyn Biosciences, Pretoria, South Africa) according to the rec-
ommendations of the manufacturer. 25 µL of the bead suspension (20
mg mL−1) were washed and equilibrated in binding buffer (50 mM potas-
sium phosphate buffer, pH 8.5) during 4 h at 25 oC. After the equilibration
procedure, the beads were collected by a magnetic separator and the
supernatant was discarded. Activation of the support was accomplished
by contacting MagReSyn®Amine microspheres with 200.0 µL of 50 mM
sodium phosphate buffer pH 8.5 containing 4 % (w/v) glutaraldehyde
fluoro-2'-deoxyadenosine
(2-F-dAdo),
14.4
min,
2-chloro-2'-
deoxyadenosine (2-CldAdo), 23.2 min; 2,6-diaminopurine-2'-deoxyribose
(2,6-DAPdRib), 16.04 min; 6-chloropurine-2'-deoxyribose (6-ClPurdRib),
22.9 min; 6-mercaptopurine-2'-deoxyribose (6-MPdRib), 13.5 min; 6-
methylpurine-2'-deoxyribose (6-MetPurdRib), 17.7 min; 6-methoxy-2'-
deoxyguanosine (6-MeOdGuo), 21.8 min; benzimidazole-2'-deoxyribose
(BzmdRib), 33.5 min; arabinosyl adenine (ara A), 15.00 min; arabinosyl
guanine (ara G), 11.4 min; arabinosyl hypoxanthine (ara H), 11.6 min;
arabinosyl 2-fluoroadenine (ara 2-FAde), 13.8 min; arabinosyl chloroade-
nine (ara 2-ClAde), 22.6 min; arabinosyl 2,6-diaminopurine (ara 2,6-DAP),
15.8 min; arabinosyl 6-chloropurine (ara 6-ClPur), 22.1 min; arabinosyl 6-
mercaptopurine (ara 6-MP), 13.2 min; arabinosyl 6-methoxyguanine (ara
6-MeOGua), 21.4 min; 2',3'-dideoxyadenosine (ddAdo), 19.7 min; 2',3'-
dideoxyguanosine (ddGuo), 16.4 min; 2',3'-dideoxyinosine (ddIno), 15.9
min; 2-fluoro-2',3'-dideoxyadenosine (dd2-Fado), 18.3 min; 2-chloro-2',3'-
dideoxyadenosine (dd2-Clado), 27 min; 2,6-diaminopurine-2',3'-
o
during 3 h at 25 C, and then washed extensively with distilled water to
remove the excess of the activating agent. Finally, activated micro-
spheres were washed and equilibrated in binding buffer (50 mM potassi-
um phosphate buffer, pH 8.5) shortly prior to use.
Subsequently, different amounts of enzyme (25-125 μg) were mixed
with glutaraldehyde-activated MagReSyn®Amine microspheres. The
enzyme solution with the binding buffer was adjusted to 10× the volume
of the initial beads suspension. The enzyme and bead suspensions were
mixed thoroughly and incubated at room temperature under orbital shak-
ing using a Sunflower 3D Mini–Shaker (BioSan, Latvia) for 10 h. After the
binding procedure, the beads were collected by a magnetic separator
and the supernatant was discarded. The beads were washed with wash-
ing buffer (50 mM potassium phosphate buffer, pH 8.5; 1 M NaCl) for
three times to remove non-covalently bound enzyme. After this, the
beads were also treated with 1 M Tris-HCl, pH 9.0 for 3 h at room tem-
perature to quench any remaining amine reactive residues on the micro-
particles. The beads were washed with 50 mM potassium phosphate
buffer (pH 8.5) for three times and then stored at 4 oC.
dideoxyribose
dideoxyribose
(dd2,6-DAPRib),
(dd6-ClPurRib),
20.9
27
min;
min;
6-chloropurine-2',3'-
6-mercaptopurine-2',3'-
dideoxyribose (dd6MPRib), 19.5 min; 2',3'-dideoxy-6-methoxyguanosine
(dd6-MeOGuo), 25.4 min. 2'-fluoro-2'-deoxyadenosine (2'F-dAdo), 17.8
min; 2'-fluoro-2'-deoxyguanosine (2'F-dGuo), 16.3 min; 2'-fluoro-2'-
deoxyinosine (2'F-dIno), 13.9 min; 2'-fluoro-2-fluoroadenosine (2'F-
2FdAdo), 17.4 min; 2'-fluoro-2-chloroadenosine (2'F-2CldAdo), 33.9 min;
2'-fluoro-2,6-diaminopurine-2'-deoxyribose (2'F-2,6-DAPdRib), 18 min; 2'-
fluoro-6-chloropurine-2'-deoxyribose (2'F-ClPurdRib), 26 min; 2'-fluoro-6-
mercaptopurine-2'-deoxyribose (2'F-6-MPdRib), 19 min; 2'-fluoro-6-
methoxyguanosine (2'F-MeOdGuo), 24.6 min.
Scanning electron microscopy
Enzyme activity assay for immobilized TbPDT
The morphology and particle structure of MagReSyn®Amine and TbPDT-
MagReSyn®Amine beads (MTbPDT beads) were evaluated by SEM. SEM
images were collected using a JEOL JSM-6400F electron microscopy
equipped with a field emission gun at 10.0 kV (Centro de Microscopía
Electrónica, UCM, Madrid). The magnetic devices were dispersed in
distilled water at room temperature (25 ◦C), deposited onto a metal
support and coated with an ultrathin layer of gold.
Particle size distribution and zeta potential were also analyzed by DLS
(Nanopartica, Horiba SZ-100, Braga, Portugal) at 25 oC. For this purpose,
homogeneous dispersions of MagReSyn®Amine beads and MTbPDT
beads in distilled water were prepared inside two different types of dis-
posable cells. Scattered light was collected at a 90° angle for size de-
terminations and data were processed by Nanotrac Flex software. All
quantifications were performed in quintuplicate.
As performed with the soluble enzyme, the enzymatic synthesis of dAdo
from dIno and Ade was established as the standard transglycosidase
reaction. The enzymatic activity of the immobilized enzyme was meas-
ured using 2.6–26.4 μg of different MTbPDT derivatives (each containing
2.64 μg of immobilized TbPDT) which were added to a 320 μL solution
containing 10 mM dIno and 10 mM Ade in 50 mM phosphate buffer, pH
6.0. The reaction mixture was incubated at 50 °C for 10 min under mag-
netic stirring. The reaction was stopped by collecting the beads with a
magnetic separator and the supernatant was treated following the gen-
eral procedure described above. Enzymatic activity (IU/mg of enzyme)
was evaluated as described above to measure specific activity of the
biocatalyst (IU/g derivative).
Enzymatic synthesis of nucleoside analogues by immobilized
TbPDT
Computational Methods
Non-natural nucleosides were synthesized by adding 4.25 μg of
MTbPDT3 (containing 2.64 μg of immobilized TbPDT) to a 320 µL solu-
tion containing 1 mM dGuo and 1 mM purine base analogues in 50 mM
sodium phosphate buffer pH 6.0 at 50 °C for 20 min, under magnetic
shaking. Enzymatic activity (IU/mg of enzyme) was evaluated as de-
scribed above to measure specific activity of the biocatalyst (IU/g deriva-
tive).
The AMBER force field[40] was used for energy refinement and molecular
dynamics simulations[41] of an all-atom representation of PDB entry
2A0K[10] (upon removal of the N-terminal His tag) in explicit solvent,
following a previously described protocol.[42] An ensemble of 10 repre-
sentative low-energy snapshots of TbPDT in complex with 2’-
deoxyinosine (docked as reported earlier for 2’-deoxyadenosine in
LmPDT)[9] was used to calculate the solvent-corrected intersubunit ener-
gy, as well as its decomposition into van der Waals, coulombic, apolar,
and desolvation contributions by means of our in-house tool MM-
ISMSA.[18] The non-electrostatic calculations involve a van der Waals
term, as in molecular mechanics (MM), and the change in solvent-
accessible surface area (SASA) of both subunits brought about by dimer
formation. The former, which accounts for shape complementarity, is
calculated with the standard AMBER 12−6 Lennard-Jones potential
whereas the latter (“apolar” in the Table), representing the nonpolar part
of the desolvation process, is assumed to be linearly related to a combi-
nation of the cavitation term and the van der Waals solute−solvent inter-
actions. The solvent-screened coulombic interactions were estimated by
means of an implicit solvent model (ISM) that makes use of a sigmoidal,
distance-dependent dielectric function.[43] The pattern of interchain hy-
drogen bonds present in the TbPDT dimer was calculated and visualized
using the PHP-based web application HBplot (www.virtuadrug.com/
hbplot)[19], which employs the HBplus algorithm.[44]
Analytical methods
The production of nucleosides was quantitatively measured with an ACE
EXCEL 5 μm CN-ES 250 × 4.6 mm equilibrated with 100 % trimethyl
ammonium acetate at a flow rate of 0.8 mL/min. Retention times for the
reference natural and non-natural bases (hereafter abbreviated accord-
ing to the recommendations of the IUPAC-IUB Commission on Biochem-
ical Nomenclature) were as follows: adenine (Ade), 10.2 min; guanine
(Gua), 7.3 min; hypoxanthine (Hyp), 7.5 min; 2-fluoroadenine (2-FAde),
14.4 min; 2-chloroadenine (2-ClAde), 17.3 min; 2,6-diaminopurine (2,6-
DAP), 11.1 min; 6-chloropurine (6-ClPur), 16.7 min; 6-mercaptopurine
(6-MP), 10.5 min; 6-methylpurine (6-MetPur), 13.8 min; 6-
methoxyguanine (6-MeOGua), 16.4 min; benzimidazole (Bzm), 29.6 min,
and theophylline (Theo), 18.42 min.
Retention times for the nucleosides (hereafter abbreviated according to
the recommendations of the IUPAC-IUB Commission on Biochemical
Nomenclature) were as follows: 2'-deoxyadenosine (dAdo), 15.50 min; 2'-
deoxyguanosine (dGuo), 11.6 min; 2'-deoxyinosine (dIno), 12.0 min; 2-
9
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
[15] R. Anand, P. A. Kaminski, S. E. Ealick, Biochemistry 2004,
43, 2384-2393.
Acknowledgements
[16] J. Holguin, R. Cardinaud, Eur. J. Biochem. 1975, 54, 515-
520.
[17] O. A. Stasyuk, H. Szatyłowicz, T. M. Krygowski, J. Org.
Chem. 2012, 77, 4035-4045.
[18] J. Klett, A. Núñez-Salgado, H. G. Dos Santos, Á. Cortés-
Cabrera, A. Perona, R. Gil-Redondo, D. Abia, F. Gago, A.
Morreale, J. Chem. Theory Comput. 2012, 8, 3395-3408.
[19] Z. Bikadi, L. Demko, E. Hazai, Arch. Biochem. Biophys.
2007, 461, 225-234.
This work was supported by grants SAN151610 from the San-
tander Foundation and 2016/UEM08 from the European Univer-
sity of Madrid to JFL and SAF2015-64629-C2-2-R from the
Spanish Ministerio de Economía y Competitividad to FG.
Conflict of Interest Statement
[20] A. de Marco, Microb Cell Fact 2009, 8, 26.
[21] L. P. Jordheim, D. Durantel, F. Zoulim, C. Dumontet, Nat.
Rev. Drug Discov. 2013, 12, 447-464.
[22] a) C. M. Galmarini, J. R. Mackey, C. Dumontet, Lancet
Oncol. 2002, 3, 415-424; b) W. B. Parker, Chem. Rev.
2009, 109, 2880-2893.
Dr. Justin Jordaan has an interest in ReSyn Biosciences. For the
purpose of this study, he advised on immobilization strategy but
did not directly participate in the experimentation and did not
generate any data.
[23] L. L. Bondoc, Jr., G. Ahluwalia, D. A. Cooney, N. R.
Hartman, D. G. Johns, A. Fridland, Mol. Pharmacol. 1992,
42, 525-530.
Keywords: biocatalysis • nucleoside analogues • 2‘-deoxy-
ribosyltransferase • enzyme immobilization • molecular dynamics
[24] 20th WHO Essential Medicines List (EML),
http://www.who.int/medicines/publications/essentialmedicin
es/en/, March 2017.
References
[25] F. Huguet, T. Leguay, E. Raffoux, P. Rousselot, N. Vey, A.
Pigneux, N. Ifrah, H. Dombret, Leuk. Lymphoma 2015, 56,
847-857.
[26] J. V. Tuttle, M. Tisdale, T. A. Krenitsky, J. Med. Chem.
1993, 36, 119-125.
[27] S. Hammes-Schiffer, S. J. Benkovic, Annu. Rev. Biochem
2006, 75, 519-541.
[28] C. Mateo, J. M. Palomo, G. Fernández-Lorente, J. M.
Guisán, R. Fernández-Lafuente, Enzyme Microb. Technol.
2007, 40, 1451-1463.
[29] R. A. Sheldon, S. van Pelt, Chem. Soc. Rev. 2013, 42,
6223-6235.
[30] J. Fernández-Lucas, A. Fresco-Taboada, C. Acebal, I. de
la Mata, M. Arroyo, Appl. Microbiol. Biotechnol. 2011, 91,
317-327.
[1] a) T. Robak, E. Lech-Maranda, A. Korycka, E. Robak, Curr.
Med. Chem. 2006, 13, 3165-3189; b) M. J. Lapponi, C. W.
Rivero, M. A. Zinni, C. N. Britos, J. A. Trelles, J. Mol. Catal.
B: Enzym. 2016, 133, 218-233.
[2] J. Boryski, Curr. Org. Chem. 2008, 12, 309-325.
[3] a) J. Fernández-Lucas, Appl. Microbiol. Biotechnol. 2015,
99, 4615-4627; b) A. Fresco-Taboada, I. de la Mata, M.
Arroyo, J. Fernández-Lucas, Appl. Microbiol. Biotechnol.
2013, 97, 3773-3785; c) E. Lewkowicz, A. Iribarren, Curr.
Org. Chem. 2006, 10, 1197-1215; d) I. Mikhailopulo, Curr.
Org. Chem. 2007, 11, 317-335.
[4] a) P. A. Kaminski, J. Biol. Chem. 2002, 277, 14400-14407;
b) J. Fernández-Lucas, C. Acebal, J. V. Sinisterra, M.
Arroyo, I. de la Mata, Appl. Environ. Microbiol. 2010, 76,
1462-1470.
[31] a) A. Fresco-Taboada, I. Serra, M. Arroyo, J. Fernández-
Lucas, I. de la Mata, M. Terreni, Catal. Today 2016, 259,
197-204; b) A. Fresco-Taboada, I. Serra, J. Fernandez-
Lucas, C. Acebal, M. Arroyo, M. Terreni, I. de la Mata,
Molecules 2014, 19, 11231-11249.
[32] C. N. Britos, M. J. Lapponi, V. A. Cappa, C. W. Rivero, J. A.
Trelles, J. Fluorine Chem. 2016, 186, 91-96.
[33] G. Hollermann, R. Dhekane, S. Kroll, K. Rezwan, Biochem.
Eng. J. 2017, 126, 30-39.
[34] E. A. Manoel, M. Pinto, J. C. S. dos Santos, V. G. Tacias-
Pascacio, D. M. G. Freire, J. C. Pinto, R. Fernandez-
Lafuente, RSC Adv. 2016, 6, 62814-62824.
[35] H. Ohshima, in Electrical Phenomena at Interfaces and
Biointerfaces: Fundamentals and Applications in Nano-,
Bio-, and Environmental Sciences (Ed.: H. Ohshima), John
Wiley & Sons, Inc., Hoboken, NJ, USA, 2012, pp. 27-34.
[36] U. K. Laemmli, Nature 1970, 227, 680-685.
[37] S. C. Gill, P. H. von Hippel, Anal. Biochem. 1989, 182,
319-326.
[5] K. A. Lawrence, M. W. Jewett, P. A. Rosa, F. C. Gherardini,
Mol. Microbiol. 2009, 72, 1517-1529.
[6] P. A. Kaminski, P. Dacher, L. Dugue, S. Pochet, J. Biol.
Chem. 2008, 283, 20053-20059.
[7] J. Fernández-Lucas, R. Harris, I. Mata-Casar, A. Heras, I.
de la Mata, M. Arroyo, J. Ind. Microbiol. Biotechnol. 2013,
40, 955-966.
[8] a) G. Bayramoglu, M. Yilmaz, M. Yakup Arica, Bioprocess
Biosyst. Eng. 2010, 33, 439-448; b) X. Zhou, I. A.
Mikhailopulo, M. N. Cruz Bournazou, P. Neubauer, J. Mol.
Catal. B: Enzym. 2015, 115, 119-127; c) J. del Arco, S.
Martínez-Pascual, V. Clemente-Suárez, O. Corral, J.
Jordaan, D. Hormigo, A. Perona, J. Fernández-Lucas,
Catalysts 2018, 8, 184.
[9] N. Crespo, P. A. Sánchez-Murcia, F. Gago, J. Cejudo-
Sánches, M. A. Galmes, J. Fernández-Lucas, J. M.
Mancheño, Appl. Microbiol. Biotechnol. 2017, 101, 7187-
7200.
[10] J. Bosch, M. A. Robien, C. Mehlin, E. Boni, A. Riechers, F.
S. Buckner, W. C. Van Voorhis, P. J. Myler, E. A. Worthey,
G. DeTitta, J. R. Luft, A. Lauricella, S. Gulde, L. A.
Anderson, O. Kalyuzhniy, H. M. Neely, J. Ross, T. N.
Earnest, M. Soltis, L. Schoenfeld, F. Zucker, E. A. Merritt,
E. Fan, C. L. Verlinde, W. G. Hol, J. Med. Chem. 2006, 49,
5939-5946.
[38] P. H. Brown, P. Schuck, Biophys. J. 2006, 90, 4651-4661.
[39] K. E. Van Holde, Physical Biochemistry, 2nd ed., Prentice-
Hall, Englewood Cliffs, NJ, 1985.
[40] D. A. Case, J. T. Berryman, R. M. Betz, D. S. Cerutti, I. T.
E. Cheatham, T. A. Darden, R. E. Duke, T. J. Giese, H.
Gohlke, A. W. Götz, N. Homeyer, S. Izadi, P. Janowski, J.
Kaus, A. Kovalenko, T. S. Lee, S. LeGrand, P. Li, T.
Luchko, R. Luo, B. Madej, K. M. Merz, G. Monard, P.
Needham, H. Nguyen, H. T. Nguyen, I. Omelyan, A.
Onufriev, D. R. Roe, A. Roitberg, R. Salomon-Ferrer, C. L.
Simmerling, W. Smith, J. Swails, R. C. Walker, J. Wang, R.
M. Wolf, X. Wu, D. M. York, P. A. Kollman, AMBER 14,
University of California, San Francisco, 2015.
[11] a) J. Becker, M. Brendel, Biol. Chem. Hoppe Seyler 1996,
377, 357-362; b) W. Uerkvitz, Eur. J. Biochem. 1971, 23,
387-395.
[12] Y. Miyamoto, T. Masaki, S. Chohnan, Biochim. Biophys.
Acta 2007, 1774, 1323-1330.
[13] J. Acosta, J. del Arco, S. Martinez-Pascual, V. Clemente-
Suárez, J. Fernández-Lucas, Catalysts 2018, 8, 9.
[14] T. C. Bruice, Acc. Chem. Res. 2002, 35, 139-148.
10
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
[41] R. Salomon-Ferrer, A. W. Götz, D. Poole, S. Le Grand, R.
C. Walker, J. Chem. Theory Comput. 2013, 9, 3878-3888.
[42] P. A. Sánchez-Murcia, A. Cortés-Cabrera, F. Gago, J.
Comput. Aided Mol. Des. 2017, 31, 915-928.
[43] A. Morreale, R. Gil-Redondo, A. R. Ortiz, Proteins 2007,
67, 606-616.
[44] I. K. McDonald, J. M. Thornton, J. Mol. Biol. 1994, 238,
777-793.
11
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800775
ChemCatChem
FULL PAPER
A bioprocess for the enzymatic synthesis of nucleoside analogues has been devel-
oped. A type I NDT from Trypanosoma brucei (TbPDT) was produced and bio-
chemically characterized (pH and T dependence, substrate specificity, site-directed
variants). Covalent immobilization of TbPDT on magnetic microspheres led to
MTbPDT3, which displays 4200 IU/g, an activity recovery of 22% and a reusability
of more than 25 cycles.
Elena Pérez, Pedro A. Sánchez-Murcia,
Justin Jordaan, María Dolores Blanco,
José Miguel Mancheño, Federico Gago,
Jesús Fernández-Lucas*
Page No. – Page No.
Enzymatic synthesis of therapeutic
nucleosides using a highly versatile
purine nucleoside 2’-deoxyribosyl-
transferase from Trypanosoma brucei
12
This article is protected by copyright. All rights reserved.