Immobilized Drosophila melanogaster deoxyribonucleoside kinase (DmdNK) as a high performing biocatalyst for the synthesis of purine arabinonucleotides

Article abstract of DOI:10.1002/adsc.201300649

Fruit fly (Drosophila melanogaster) deoxyribonucleoside kinase (DmdNK; EC: 2.7.1.145) was characterized for its substrate specificity towards natural and non-natural nucleosides, confirming its potential in the enzymatic synthesis of modified nucleotides. DmdNK was adsorbed on a solid ion exchange support (bearing primary amino groups) achieving an expressed activity >98%. Upon cross-linking with aldehyde dextran, expressed activity was 30-40%. Both biocatalysts (adsorbed or cross-linked) were stable at pH 10 and room temperature for 24 h (about 70% of retained activity). The cross-linked DmdNK preparation was used for the preparative synthesis of arabinosyladenine monophosphate (araA-MP) and fludarabine monophosphate (FaraAMP). Upon optimization of the reaction conditions (50 mM ammonium acetate, substrate/ATP ratio= 1:1.25, 2 mM MgCl2, 378C, pH 8) immobilized DmdNK afforded the title nucleotides with high conversion (>90%), whereas with the soluble enzyme lower conversions were achieved (78-87%). Arabinosyladenine monophosphate was isolated in 95% yield and high purity (96.5%).

Full text of DOI:10.1002/adsc.201300649

FULL PAPERS
DOI: 10.1002/adsc.201300649
Immobilized Drosophila melanogaster Deoxyribonucleoside
Kinase (DmdNK) as a High Performing Biocatalyst for the
Synthesis of Purine Arabinonucleotides
a,
a,e
b
b
ˇ
Immacolata Serra, * Silvia Conti, Jure Piskur, Anders R. Clausen,
Birgitte Munch-Petersen,c Marco Terreni,a,d and Daniela Ubialia,d,*
Department of Drug Sciences, University of Pavia, via Taramelli 12, 27100 Pavia, Italy
a
Fax : (+39)-0382-422-975; phone: (+39)-0382-987-889; e-mail: immacolata.serra@unipv.it or daniela.ubiali@unipv.it
Department of Biology, Lund University, Soelvegatan 35, 22362 Lund, Sweden
Department of Science, Systems and Models, Roskilde University, P.O. Box 260, DK 4000, Roskilde, Denmark
Italian Biocatalysis Center, via Taramelli 12, 27100 Pavia, Italy
b
c
d
e
Present address: Gedeon Richter Italia srl, via Cassala 16, 20143 Milano, Italy
Received: July 24, 2013; Revised: December 23, 2013; Published online: February 6, 2014
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201300649.
Abstract: Fruit fly (Drosophila melanogaster) deoxy- synthesis of arabinosyladenine monophosphate
ribonucleoside kinase (DmdNK; EC: 2.7.1.145) was (araA-MP) and fludarabine monophosphate (FaraA-
characterized for its substrate specificity towards nat- MP). Upon optimization of the reaction conditions
ural and non-natural nucleosides, confirming its po- (50 mM ammonium acetate, substrate/ATP ratio=
tential in the enzymatic synthesis of modified nucleo- 1:1.25, 2 mM MgCl2, 378C, pH 8) immobilized
tides. DmdNK was adsorbed on a solid ion exchange DmdNK afforded the title nucleotides with high con-
support (bearing primary amino groups) achieving version (>90%), whereas with the soluble enzyme
an expressed activity >98%. Upon cross-linking with lower conversions were achieved (78–87%). Arabi-
aldehyde dextran, expressed activity was 30-40%. nosyladenine monophosphate was isolated in 95%
Both biocatalysts (adsorbed or cross-linked) were yield and high purity (96.5%).
stable at pH 10 and room temperature for 24 h
(about 70% of retained activity). The cross-linked Keywords: biocatalysis; deoxyribonucleoside kinase;
DmdNK preparation was used for the preparative immobilization; nucleotides; phosphorylation
Introduction
(dNDPs) and triphosphates (dNTPs), respectively.
The number and the specificities of dNKs vary from
Natural nucleosides and nucleotides are implicated in organism to organism but, in all cases, dNKs phos-
several physiological pathways, being the constituents phorylate 2’-deoxyribonucleosides much more effi-
of nucleic acids, intracellular mediators and coen- ciently than they phosphorylate ribonucleoside coun-
zymes (i.e., ATP, NADH). Specifically, all organisms terparts.[2]
require 2’-deoxyribonucleoside triphosphates (dNTPs)
Natural ribonucleoside monophosphates are not
for DNA biosynthesis and its regeneration. dNTPs only nucleic acid building blocks but they are also
are synthesized by the de novo or the salvage path- employed as food additives due to their immunosti-
way. The key step in the de novo pathway is through mulant and taste enhancer activity.[3–5]
ribonucleotide reductase, where ribonucleoside di-
Nucleoside analogues are widely used as antiviral
phosphates are reduced to the corresponding deoxyri- and antitumour agents.[6,7] They can be incorporated
bonucleoside diphosphates.[1] The salvage route of in the viral or cellular DNA by DNA polymerases
dNTPs starts from 2’-deoxyribonucleosides (dNs) that after phosphorylation by kinases, or they can inhibit
are phosphorylated by intracellular 2’-deoxyribonu- the enzymes involved in the biosynthesis of DNA pre-
cleoside kinases (dNKs) to 2’-deoxyribonucleoside cursors.[8]
monophosphates (dNMPs) which are then converted
Vidarabine (arabinosyladenine, araA), firstly de-
by deoxyribonucleoside monophosphate and diphos- signed as an antineoplastic drug, is now used as
phate kinases into the corresponding diphosphates a second-line drug for the systemic treatment of
Adv. Synth. Catal. 2014, 356, 563 – 570
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
563

FULL PAPERS
Immacolata Serra et al.
This enzyme is particularly interesting from an in-
dustrial viewpoint for the environmentally sustainable
large-scale production of dNTPs and their ana-
logues.[13] Chemical synthesis of mononucleotides is
routinely performed by treating the precursor nucleo-
side with a large excess of POCl3 in acetic acid in the
presence of pyridine or using short-chain trialkyl
phosphates (i.e., trimethyl phosphate) as solvents
under anhydrous conditions at 08C.[14–16] In both cases,
these synthetic routes usually provide poor or moder-
ate yields, low product purity and are also associated
Figure 1. Chemical structures of araA-MP and FaraA-MP.
Herpex simplex viruses. Upon administration, this with harsh reaction conditions and waste disposal
compound is phosphorylated to the triphosphate issues.
(araA-TP) and competes with 2’-deoxyadenosine tri-
In this frame, the application of white biotech-
phosphate (dAdo-TP) as substrate for the viral DNA nology in the synthesis of nucleotides is an attractive
polymerase leading to the formation of a defective field since biotransformations do not usually require
DNA and resulting in the prevention of viral infec- tedious protection/deprotection steps and frequently
tion.[9] However, the potential usefulness of araA is occur under milder and greener conditions due to the
limited by its poor solubility and the treatment with high regio- and stereospecificity of the enzymes. The
araA requires a continuous intravenous infusion. To chemical production of active pharmaceutical ingredi-
overcome this drawback, araA-5’-monophosphate ents is reported as the worst process for the mass
(araA-MP, Figure 1) has been proposed as a soluble ratio of waste versus product (E-factor).[17–19] The use
prodrug which can be administered intermittently in- of biocatalysts can thus provide novel and straightfor-
tramuscularly or intravenously.[10] The fluorinated ward synthetic schemes for developing sustainable in-
counterpart of araA-MP, fludarabine monophosphate dustrial processes that turn in high yields and purity
(FaraA-MP, Figure 1), is used against hematological of the products.[20–22]
malignancies such as leukemias and lymphomas.[11]
To date, an efficient alternative to dNKs is repre-
In contrast to mammals, the fruit fly Drosophila sented by class A bacterial non-specific acid phospha-
melanogaster was found to have only one multisub- tases (NSAPs, EC: 3.1.3.2) which have been reported
strate deoxyribonucleoside kinase (DmdNK; EC: to catalyze the phosphorylation of nucleosides[23,24]
2.7.1.145). This dNK is able to phosphorylate all four and other compounds (i.e., carbohydrates)[25] using
natural 2’-deoxyribonucleosides with high efficiency. pyrophosphate as a phosphate donor. However, final
The broad specificity and high turnover towards the yields are plagued by the hydrolysis of the products,
naturally occurring 2’-deoxyribonucleosides by thus requiring a laborious reaction optimization or
DmdNK is reflected in a similar broad specificity to- the use of specific inhibitors.[25] For some substrates,
wards nucleoside analogues used in chemotherapy.[12]
in fact, dephosphorylation is much faster than phos-
DmdNK is active as a homodimer with a global phorylation.[26] NSAPs display higher affinities for ri-
molecular weight of 58 kDa. DmdNK is an Mg2+-de- bonucleosides than 2’-deoxyribonucleosides.[27] This
pendent 5’-phosphotransferase and catalyzes the for- feature complements the specificity of dNKs, mostly
mation of a 5’-phosphate nucleoside and ADP from directed versus 2’-deoxyribonucleosides, thus imple-
a nucleoside and a phosphate donor like ATP menting the toolset of phosphorylating enzymes.
(Scheme 1).[2]
The bottleneck in the use of enzymes as biocata-
lysts is often their instability and poor performance
under reaction conditions, their high cost and solubili-
ty in the reaction medium. These issues, which may
affect the economy of a biocatalytic process, can be
frequently overcome by immobilizing the enzyme on
a solid support, especially in the case of multimeric
enzymes. In fact, these enzymes may be easily inacti-
vated by dissociation of the protein subunits.[28–31]
A
convenient strategy to achieve immobilization and
stabilization of multimeric enzymes is the immobiliza-
tion via ionic exchange of the enzyme on a carrier
(epoxy Sepabeads) coated with polyethylenimine
(PEI) followed by cross-linking with a polyaldehyde
macromolecule (oxidized dextran). This process per-
mits one to obtain a multisubunit immobilization.[32,33]
Scheme 1. Phosphorylation reaction of nucleosides catalyzed
by DmdNK.
564
asc.wiley-vch.de
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2014, 356, 563 – 570

FULL PAPERS
Immobilized Drosophila melanogaster Deoxyribonucleoside Kinase
This enzyme efficiently catalyzed the phosphoryla-
tion of natural 2’-deoxyribonucleosides allowing
>98% transformation of these substrates; with dIno
(2’-deoxyinosine, 3) the enzyme displayed a reaction
rate five-fold lower than with dAdo (2’-deoxyadeno-
sine, 1) achieving only 30% conversion after 24 h
(Table 1). DmdNK showed, after 24 h, less than 5%
conversion of Ado (7), Guo (8) and Ino (9). Cytidine
(10) was converted in up to 81% after 24 h with a re-
action rate of 6 IU/mg. A good conversion (64%) was
achieved also with uridine (11) even if the specific ac-
tivity was less than 1 IU/mg (Table 1). In agreement
with previous results,[35] in general DmdNK preferen-
tially phosphorylated pyrimidine derivatives and was
also able to synthesize the nucleotide analogues
araA-MP (12a) and FaraA-MP (13a). In the latter
Figure 2. Substrate specificity of the soluble DmdNK. IU:
mmolminꢀ1.
Furthermore, in some cases, enzyme properties have case final conversions after 24 h were comparable
been enhanced upon immobilization.[34]
with those of the natural substrates (2’-deoxyribonu-
The present work describes the study of substrate cleosides) (Table 1).
specificity of DmdNK from the standpoint of its ap-
plication as a biocatalyst, its immobilization on a solid
support and the use of the resulting biocatalyst for DmdNK Immobilization
the preparative synthesis of araA-MP and FaraA-MP.
With the aim to prepare an active, stable and recycla-
ble biocatalyst, DmdNK was immobilized on a solid
support.
Immobilization was performed through a two-step
procedure, previously reported for immobilization
and stabilization of multimeric enzymes.[32,33] Initially,
Results and Discussion
Substrate Specificity
The soluble DmdNK was tested for its substrate spe- the experiments were performed by loading a low
cificity towards 2’-deoxyribo-, ribo- and arabinonu- amount of enzyme on the carrier (0.1–0.5 mggꢀ1) in
cleosides (Figure 2, Table 1).
order to control diffusion problems. The enzymatic
activity was completely retained after adsorption on
the ionic carrier and after cross-linking in almost all
cases while, as expected, the crucial step was the
ꢀ
Table 1. Phosphorylation catalyzed by DmdNK as depicted
in Scheme 1. In parentheses, the maximum percentages of chemical reduction of imines to stable C N bonds
conversion after 24 h are given.[a]
which resulted in a striking loss of activity of the bio-
catalyst (about 2/3 of initial activity was recovered)
(Table 2).
Nucleoside
R
R’
B
Nucleotide (%)
To minimize the loss of activity upon chemical re-
duction, two strategies were pursued: reduction was
performed by using a lower amount of NaBH4
(entry 3 Table 2) or by replacing NaBH4 with the
same amount (1 mgmLꢀ1) of NaBH3CN (entry 4
Table 2) which is a milder reducing agent. However,
in both cases, no significant improvement of the final
activity was observed. Moreover, not even the addi-
tion of dAdo to the immobilization medium was help-
ful in preserving the enzyme activity during reduction
(entry 5 Table 2). dAdo, in fact, could shield the
enzyme active site by binding to it, thus avoiding pos-
sible distortions occurring during the immobilization
process.[36]
dAdo (1)
dGuo (2)
dIno (3)
dCyt (4)
dUrd (5)
Thd (6)
H
H
H
H
H
H
H
H
H
H
H
H
adenine
guanine
hypoxanthine 3a (30)
cytosine
uracil
1a (>98)
2a (>98)
4a (>98)
5a (>98)
6a (>98)
thymine
Ado (7)
Guo (8)
Ino (9)
Cyt (10)
Urd (11)
H
H
H
H
H
OH adenine
OH guanine
OH hypoxanthine 9a (<5)
OH cytosine
OH uracil
7a (5)
8a (<5)
10a (81)
11a (64)
araA (12)
FaraA (13) OH
araU (14)
OH
H
H
H
adenine
2-F-adenine
uracil
12a (>98)
13a (>98)
14a (>98)
OH
Using the standard immobilization conditions,
a higher amount of protein was also loaded per gram
of carrier (2 mggꢀ1) and 40% of the total activity was
[a]
Experimental conditions: substrate, 1 mM; ATP, 2 mM;
buffer, 50 mM Tris HCl pH 8; MgCl2, 2 mM; tempera-
ture, 378C; volume, 1 mL.
Adv. Synth. Catal. 2014, 356, 563 – 570
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
565

FULL PAPERS
Immacolata Serra et al.
Table 2. Immobilization of DmdNK on Sepabeads PEI 600 cross-linked with aldehyde dextran.
Entry IU/g
(mggꢀ1)
Adsorbed activity Activity after cross-linking
Reducing
agent
Final activity
(IU/g)
Recovered activity
(%)[a]
(%)
IU/g (%)
1
2 (0.1)
>98%
>98%
>98%
>98%
>98%
>98%
2.0 (100)
8.5 (85)
10.0 (100)
11.5 (89)
nd
NaBH4
NaBH4
NaBH4
NaBH3CN
NaBH4
NaBH4
0.6
2.7
2.9
1.5
2.7
16.0
27
27
29
12
28
40
2
10 (0.5)
10 (0.5)
13 (0.6)
10 (0.5)[c]
40 (2.0)
3[b]
4
5[c]
6
nd
[a]
[b]
[c]
Calculated on the basis of the expressed activity considering dAdo as the reference substrate.
[NaBH4] was reduced by half.
2 mM dAdo were added to the immobilization suspension; nd: not determined.
recovered obtaining a solid biocatalyst with 16 IU/g
of activity (entry 6 in Table 2).
Considering the soluble enzyme, in the phosphory-
lation of dAdo (1) using 1.25 or 1 equivalent of ATP,
The stability at pH 10 of immobilized DmdNK the yield slightly decreased, compared with the reac-
before and after cross-linking was investigated. The tion performed with 2 equivalents of ATP (from 99 to
results obtained could not highlight any difference be- 95% and 90%, respectively) and, to a higher extent,
tween the two enzyme preparations under these con- the same trend was observed in the phosphorylation
ditions; in fact, after 24 h of incubation, both prepara- of araA (12) where 52% and 89% of conversion was
tions retained more than 70% of the initial activity achieved using a 1:1 and 1:2 molar ratio of ATP, re-
(Figure 1 in the Supporting Information).
spectively.
Moreover, the possible leakage of the protein from
With the soluble enzyme it is evident that the con-
the support under the reaction conditions was evalu- centration of ATP exerts a remarkable effect on the
ated, because of the reversibility associated to this im- non-natural substrate, i.e., araA, poorly affecting the
mobilization.[37,38] To this aim, the cross-linked and outcome of the phosphorylation of dAdo. On the con-
non-cross-linked preparations were incubated at trary, immobilization resulted in high conversion at all
either 50 mM Tris HCl pH 8 or 50 mM ammonium three molar ratios of nucleoside and ATP (Table 3).
acetate pH 8 and no activity could be detected in so- Consequently, besides the well known advantages of
lution. These results indicate that both the enzyme
preparations can be used indistinctly. However, for
the preparation of active pharmaceutical ingredients,
Table 3. Phosphorylation of nucleosides with different
amounts of ATP.[a]
a covalently bound enzyme (thus cross-linked) is
highly desirable to avoid the possibility of product
contamination. For this reason, the cross-linked im-
mobilized DmdNK was used in this report.
Substrate
DmdNK
ATP
Prod- Conversion
(mM) uct
% (hours)
dAdo (1)
dAdo (1)
dAdo (1)
dAdo (1)
dAdo (1)
dAdo (1)
araA (12)
araA (12)
araA (12)
araA (12)
araA (12)
araA (12)
AraA (12)[b]
soluble
immobilized
soluble
immobilized 1.25
soluble
immobilized
soluble
immobilized
soluble
immobilized 1.25
soluble
2
2
1.25
1a
1a
1a
1a
1a
1a
12a
12a
12a
12a
12a
12a
12a
99 (3)
99 (3)
95 (3)
99 (3)
90 (3)
96 (3)
89 (6)
99 (8)
59 (12)
97 (8)
52 (12)
95 (12)
60 (12)
59 (4)
99 (4)
Preparative Enzymatic Synthesis of araA-MP (12a)
and FaraA-MP (13a)
1
1
2
2
The reaction conditions for the enzymatic synthesis of
nucleotides araA-MP and FaraA-MP have been ex-
tensively explored in order to develop a preparative
process. The reaction parameters that have been con-
sidered are: reactant concentration (stoichiometric
ratio), type of buffer and MgCl2 concentration. The
effect of these parameters was evaluated both with
the soluble and the immobilized enzyme.
1.25
1
1
1
immobilized
immobilized
FaraA (13)[c] soluble
1:1.25 13a
FaraA (13)[c] immobilized 1:1.25 13a
[a]
Experimental conditions: substrate, 1 mM; buffer, 50 mM
Tris·HCl pH 8; MgCl2, 2 mM; temperature, 378C;
volume, 5 mL. Soluble enzyme: 0.2 IU; immobilized
enzyme: 0.16 U.
Addition of K2CO3 50 mM, pH 10.5.
Volume 1 mL; enzyme 0.02 IU.
Role of ATP
The molar ratio between nucleoside and ATP was in-
vestigated (Table 3) aiming at maximizing the conver-
sion with a small excess of ATP.
[b]
[c]
566
asc.wiley-vch.de
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2014, 356, 563 – 570

FULL PAPERS
Immobilized Drosophila melanogaster Deoxyribonucleoside Kinase
Table 4. Influence of reaction conditions on the enzymatic phosphorylation.[a]
Substrate
DmdNK (IU)
Buffer
MgCl2 (mM)
Vs (IU/mg)
Product
Conversion % (hours)
araA (12)
araA (12)
araA (12)
araA (12)
FaraA (13)
FaraA (13)
FaraA (13)
FaraA (13)
FaraA (13)
FaraA (13)
FaraA (13)
soluble (0.4)
soluble (0.4)
Tris
acetate
Tris
acetate
Tris
acetate
Tris
acetate
Tris
Tris
2
2
2
2
2
2
0.2
0.2
0.04
0.02
2
4.4
4.9
0.9
0.8
5.0
3.9
3.7
3.9
3.4
3.1
0.2
12a
12a
12a
12a
13a
13a
13a
13a
13a
13a
13a
90 (6)
86 (6)
96 (6)
98 (6)
>98 (4)
>98 (4)
>98 (4)
>98 (4)
96 (6)
immobilized (0.2)
immobilized (0.2)
soluble (0.02)
soluble (0.02)
soluble (0.02)
soluble (0.02)
soluble (0.02)
soluble (0.02)
immobilized (0.02)
93 (6)
91 (6)
acetate
[a]
Experimental conditions: substrate, 1 mM; ATP, 2 mM; buffer, 50 mM pH 8; temperature, 378C; volume, 5 mL for the
synthesis of araA-MP and 1 mL for the synthesis of FaraA-MP. IU: mmol minꢀ1.
an immobilized enzyme (stability, recyclability, recov- ported in Table 4 reveal that both Tris HCl and ace-
ery), in this case the immobilization appeared to exert tate buffer can be used: the possibility of using ace-
a dramatic influence on the catalytic properties of the tate buffer is of outstanding importance, in fact, due
enzyme allowing the reaction to achieve almost to its volatility, it can be easily removed by lyophiliza-
a quantitative conversion even when using a equimo- tion. The reduction of the concentration of Mg2+,
lar ratio of nucleoside/ATP. The improved performan- which is essential for catalysis,[2] at a concentration
ces obtained with the immobilized DmdNK, com- lower than 0.2mM delivered a slight reduction of the
pared to the soluble enzyme, might be ascribed to the reaction rate and yields. The threshold concentration
microenvironment generated by the carrier surface. In of MgCl2 was thus set to 2mM.
fact, the strongly cationic surface, resulting after coat-
ing the carrier with PEI, could create a driving force
to direct the negatively charged ATP to the immobi- Scale-Up and Purification
lized enzyme. The absorption of ATP on the carrier
was experimentally confirmed by detecting (HPLC By using the immobilized enzyme prepared by load-
monitoring) the decrease of ATP concentration (but ing 2 mg of DmdNK per gram of carrier (final activity
not of the monophosphate) in the supernatant upon 16 IU/g) the phosphorylation of 12 and 13 in acetate
incubation with the support.
To further corroborate this hypothesis, the percent- centration and using a 1:1.25 ratio of nucleoside/ATP.
age of conversion of araA (12) using 1 equivalent of Quantitative conversions were obtained at all con-
buffer was then developed at increased substrate con-
phosphate donor was evaluated in conditions that centrations of substrates (Table 5) using a small
avoid the adsorption of ATP (i.e., pH 10.5). At this amount of immobilized enzyme in 10 mL of reaction
pH, ATP was not adsorbed on the carrier and the im- volume (about 1% weight/volume). When the soluble
mobilized enzyme was stable and active. After 12 h, kinase was used the reactions were slower; for exam-
the conversion obtained was similar to that of the ple, starting from 12 mM of araA, the reaction
soluble enzyme (60%), thus indicating that the ad- reached only 78% of conversion in 24 h, while the im-
sorption of ATP on the carrier is a key point to ach- mobilized derivative achieved the almost complete
ieve high conversions with the immobilized enzyme.
We concluded that a molar ratio of 1:1.25 was the
phosphorylation of the substrate after only 12 h.
Arabinosyladenine monophosphate (25 mM) was
optimal condition in terms of percentage of conver- purified by preparative HPLC. The collected fractions
sion, consumption of reagents and impurity profile for were pooled together affording, upon lyophilization,
the synthesis of araA-MP. These conditions were used 95% of final yield. The identity of the isolated prod-
for the synthesis of the fluorinated counterpart uct was confirmed by NMR and MS analysis; chroma-
FaraA-MP (13a) as well confirming the trend ob- tographic analysis revealed the presence of 3.5% of
served for araA and resulting in 99% conversion.
ADP (see the Supporting Information for details).
Optimization of Experimental Conditions
Conclusions
A further optimization concerned the use of different Fruit fly deoxyribonucleoside kinase (DmdNK) can
buffers and concentrations of MgCl2. The results re- be used as biocatalyst for the synthesis of a wide
Adv. Synth. Catal. 2014, 356, 563 – 570
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
567

FULL PAPERS
Immacolata Serra et al.
Table 5. Scale-up of araA-MP and FaraA-MP.[a]
Protein concentration assay was performed on a Shimadzu
spectrophotometer UV 1601 by the Bradford method using
bovine serum albumin as standard.[40] 1H NMR spectra were
recorded at 400.13 MHz, on a Bruker AVANCE 400 spec-
trometer equipped with the XWIN-NMR software package
Substrate
ATP
(mM)
Enzyme
Product Conversion
% (hours)
12 (6.5 mM)
12 (12 mM)
12 (12 mM)
12 (25 mM)
12 (25 mM)
13 (6.5 mM)
13 (12 mM)
8
immobilized 12a
soluble 12a
immobilized 12a
soluble 12a
immobilized 12a
immobilized 13a
immobilized 13a
>98 (12)
78 (24)
>98 (12)
87 (24)
97 (12)
93 (4)
1
(Bruker, Karlsruhe, Germany) at 300 K. H chemical shifts
15
15
30
30
8
(d) were referenced to the solvent signal (dH =3.31 for
CD3OD). Electrospray mass spectroscopy (ESI-MS) was
performed in MeOH/H2O on an LTQ linear ion trap mass
spectrometer (Thermo Finnigan) in negative ion mode.
All experiments were performed at least as duplicate.
15
91 (4)
[a]
Experimental conditions: buffer, 50 mM ammonium ace-
tate pH 8; MgCl2, 2 mM; temperature, 378C; volume,
10 mL; soluble enzyme, 2.8 IU; immobilized enzyme,
1 IU. In the case of 25 mM araA, soluble enzyme, 5.6 IU;
immobilized enzyme, 2.5 IU.
Enzyme Production and Purification
Recombinant DmdNK was expressed and purified as previ-
ously described.[41] The purity of the enzyme in the collected
fractions was assessed by SDS-PAGE (>90%) and the final
enzyme concentration was about 3 mgmLꢀ1.
Preparation of Sepabeads-PEI
range of nucleoside monophosphates and it can be
thus a convenient alternative to POCl3-based phos-
phorylation. Upon immobilization, productivity and
rate of synthesis of DmdNK in the bioconversion of
araA (12) were dramatically enhanced with respect to
the non-immobilized enzyme, affording the title com-
pound in 95% yield and 96.5% of purity.
For the synthesis of high-added value products,
such as the monophophates of araA 12a and fludara-
bine 13a, the use of stoichiometric ATP could be jus-
tified. Otherwise, for the synthesis of other mono-
phosphates, the recycle of ADP produced during the
Activation of epoxy Sepabeads with PEI was performed as
previously reported.[33]
Preparation of Aldehyde Dextran
Dextran (1.67 g) was suspended in deionized water (50 mL)
and NaIO4 (0.43 g) was added to obtain 10% of oxidation
degree. The reaction was carried out for 2 h at room temper-
ature and then the solution was dialyzed (3500 Da MWCO)
against deionized water. The oxidized dextran was stored at
ꢀ208C.[32]
reaction should be considered using, for instance, pyr- Immobilization on Sepabeads-PEI and Cross-Linking
uvate, acetate or polyphosphate kinases.[39]
with Aldehyde Dextran
Immobilization on Sepabeads-PEI and stabilization by
cross-linking with 10% oxidized dextran was performed by
slightly modifying the procedure previously described.[32]
Briefly, the activated carrier (1.0 g) was suspended in 5 mM
phosphate buffer pH 7.5 containing the soluble DmdNK
(0.1–2.0 mg) in a total volume of 14 mL. The suspension was
kept under mechanical stirring at room temperature and
after 1 hour aldehyde dextran (1.4 mL) was added and al-
lowed to stir for an additional hour. The temperature was
then lowered to 48C, the pH was adjusted to 10.05 and
15.4 mg of NaBH4 were added (1 mgmLꢀ1 of suspension);
the reaction was carried out for 30 min under mechanical
stirring. The immobilized preparation was then filtered and
washed with 10 mM potassium phosphate buffer pH 5 and
deionized water.
Experimental Section
General
Arabinosyladenine (12) and thymidine 5’-monophosphate
(6a) were purchased from Jena Biosciences (Jena, Germa-
ny). Arabinosyladenine 5’-monophosphate (12a) was from
Acros Organics (Geel, Belgium). 2’-Deoxyuridine (dUrd, 5)
was a gift from Pro.Bio.Sint. (Varese, Italy). Fludarabine
(13) and fludarabine 5’-monophosphate (13a) were from
Adorkem Technologies (Costa Volpino, Bergamo, Italy). All
the other nucleosides and nucleotides were from Sigma Al-
drich and/or Alfa Aesar (Milano, Italy). Solvents, polyethy-
lenimine (PEI) MW 600, dl-dithiothreitol and dextran MW
6000 were purchased from Sigma Aldrich or VWR Interna-
tional (Milano, Italy). Sepabeads EC-EP was kindly sup-
plied by Resindion (Mitsubishi, Binasco, Italy). Regenerated
cellulose tubular membrane for dialysis was from Membrane
Filtration Products, Inc. (Seguin, TX, USA). 10 kDa Nano-
sepꢁ centrifugal devices were from Pall (Buccinasco,
Milano, Italy). All solvents were HPLC grade. Enzymatic
reactions were monitored by using HPLC Merck Hitachi L-
7100 equipped with a UV detector L-7400 and column oven
L-7300 (Darmstadt, Germany).
Standard Enzymatic Assay
The standard enzymatic assay was carried out by adapting
the previously reported radiometric assay.[12]
In a final volume of 1 mL a solution of Tris-HCl 50 mM
pH 8.0 containing dAdo (1 mM), ATP (2 mM), MgCl2
(2 mM) was prepared. The reaction was started by the addi-
tion of the enzyme (soluble or immobilized) and the mixture
was stirred at 378C. At different times (5, 10, and 15 min),
samples were withdrawn, filtered (either by centrifugation
with 10 kDa MWCO Nanosep at 48C and 13,200 rpm for
568
asc.wiley-vch.de
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2014, 356, 563 – 570

FULL PAPERS
Immobilized Drosophila melanogaster Deoxyribonucleoside Kinase
2 min for the soluble enzymes, or by using pipette filter de-
vices for the immobilized enzymes) and analyzed by HPLC
(see below for chromatographic conditions and tR). The ac-
tivity of DmdNK was expressed in international units (IU).
One IU corresponds to the amount of enzyme that produces
1 mmol of dAdo-MP (1a) per min.
cleosides, the conversion was determined using the mono-
phosphates as external standards.
Purification of AraA-MP (12a)
Purification of 12a was accomplished by preparative HPLC
using an AKTA Basic100 instrument (Pharmacia, Uppsala,
Sweden); chromatographic conditions were as follows:
column, Phenomenex Jupiter RP-18 (10 mm, 250ꢂ10 mm,
Merck); flow rate, 5 mLminꢀ1; detector, l=260 nm and
Enzyme Stability Assay
The residual activity was determined by modifying the gen-
eral procedure previously reported.[33]
280 nm; mobile phase MeOH/ACHTNUGTRNEUNG(NH4)HCO3, isocratic elution
at 3% MeOH for 2 column volumes, then gradient elution
from 3% to 40% MeOH in 2 column volumes. After lyophi-
lization the product appeared as a white solid.
Synthesis of AraA-MP and FaraA-MP
The suitable amount of nucleoside and ATP was dissolved
in 50 mM ammonium acetate containing 2 mM MgCl2. The
final pH was set to 8.0 with NaOH. DmdNK (soluble or im-
mobilized) was then added and the mixture was kept under
mechanical stirring at 378C. The reaction was stopped by fil-
tering the enzyme. The filtered solution was analyzed by
HPLC (see below for chromatographic conditions and tR).
Acknowledgements
We thank Dr. P. Francescato for his assistance in the prepara-
tive purification of araA-MP (12a) and Prof. G. Speranza
(University of Milano, Italy) for NMR spectra and helpful
discussions.
Chromatographic Conditions
The reaction was monitored by HPLC (l=260 nm), identi-
fying the products by comparing their retention times with
those of authentic samples. The column was a Kromasil References
RP18 (5 mm, 250ꢂ4.6 mm) kept at 358C; flow rate was
1 mLminꢀ1.
[1] P. Reichard, Annu. Rev. Biochem. 1988, 57, 349–374.
Mobile phase: eluent A=2 g (NH4)H2PO4 +0.5 g
(NH4)2HPO4/L+30 mL MeOH; eluent B=90% CH3CN.
Gradient: 0 (100% A)!8 min. (100% A)!22 min. (70%
A–30% B)!23 min. (100% A).
[2] S. Eriksson, B. Munch-Petersen, K. Johansson, H. Eck-
lund, Cell. Mol. Life Sci. 2002, 59, 1327–1346.
[3] C. T. Van Buren, A. D. Kulkarni, F. B. Rudolph, J. Nutr.
1994, 124, 160S–164S.
[4] S. Yamaguchi, K. Ninomiya, Food Rev. Int. 1998, 14,
123–138.
[5] M. Behrens, W. Meyerhof, C. Hellfritsch, T. Hofmann,
Angew. Chem. 2011, 123, 2268–2291; Angew. Chem. Int.
Ed. 2011, 50, 2220–2242.
[6] E. De Clercq, J. Clin. Virol. 2004, 30, 115–133.
[7] C. M. Galmarini, Electron. J. Oncol. 2002, 22–32.
[8] D. B. Longley, D. P. Harkin, P. G. Johnston, Nat. Rev.
Cancer 2003, 3, 330–338.
Retention times (nd=not detected): ATP (adenosine tri-
phosphate) 3.2 min; ADP (adenosine diphosphate) 3.6 min;
dAdo (2’-deoxyadenosine, 1) 16.5 min, dAdo-MP (2’-deoxy-
adenosine monophosphate, 1a) 10 min; dGuo (2’-deoxygua-
nosine, 2) 15 min, dGuo-MP (2’-deoxyguanosine monophos-
phate, 2a) 5.5 min; dIno (2’-deoxyinosine, 3) 14.1 min, dIno-
MP (2’-deoxyinosine monophosphate, 3a) 5.1 min; dCyt (2’-
deoxycytidine, 4) 6.9 min, dCyt-MP (2’-deoxycytidine mono-
phosphate, 4a) 2.9 min; dUrd (2’-deoxyuridine, 5) 9.4 min,
dUrd-MP (2’-deoxyuridine monophosphate, 5a) 5.1 min;
Thd (thymidine, 6) 15.4 min, Thd-MP (thymidine monophos-
phate, 6a) 5.6 min; Ado (adenosine, 7) 17.73 min, Ado-MP
(adenosine monophosphate, 7a) 5.3 min; Guo (guanosine, 8)
12.2 min, Guo-MP (guanosine monophosphate, 8a) nd; Ino
(inosine, 9) 12.7 min, Ino-MP (inosine monophosphate, 9a)
nd; Cyt (cytidine, 10) 4.9 min, Cyt-MP (cytidine monophos-
phate, 10a) 2.5 min; Urd (uridine, 11) 6.7 min, Urd-MP (uri-
dine monophosphate, 11a) 2.7 min; araA (arabinosylade-
nine, 12) 15.5 min, araA-MP (arabinosyladenine monophos-
phate, 12a) 5.6 min; FaraA (2-fluoro-arabinosyladenine, 13)
21.31 min, FaraA-MP (2-fluoroarabinosyladenine mono-
phosphate, 13a) 9.09 min; araU (arabinosyluracil, 14)
8.6 min, araU-MP (arabinosyluracil monophosphate, 14a)
2.8 min.
[9] F. Superti, M. G. Ammendolia, M. Marchetti, Curr.
Med. Chem. 2008, 15, 900–911.
[10] R. J. Whitley, B. C. Tucker, A. W. Kinkel, N. H. Barton,
R. F. Pass, J. D. Whelchel, C. G. Cobbs, A. G. Diethelm,
R. A. Buchanan, Antimicrob. Agents Chemother. 1980,
18, 709–715.
[11] W. Plunkett, P. Huang, V. Gandhi, Semin. Oncol. 1990,
17, 3–17.
[12] B. Munch-Petersen, J. Piskur, L. Sondergaard, J. Biol.
Chem. 1998, 273, 3926–3931.
[13] H. Ihlenfeldt, B. Munch-Petersen, J. Piskur, L. Sonder-
gaard, (Roche Diagnostics GmbH), EP Patent
EP0999275A2, 2000.
[14] T. Ikemoto, A. Haze, H. Hatano, Y. Kitamoto, M.
Ishida, K. Nara, Chem. Farm. Bull. 1995, 43, 210–215.
[15] G. Cotticelli, B. Verzola, (Adorkem Technology SPA),
WO Patent WO2005040183, 2005.
[16] P. Blumbergs, M. S. Khan, R. L. Kalamas, A. Patel,
M. P. Lamontagne, WO Patent WO9200312, 1992.
[17] R. A. Sheldon, Green Chem. 2007, 9, 1273–1283.
[18] R. A. Sheldon, Chem. Commun. 2008, 3352–3365.
The percentage of conversion was calculated on the basis
of the depletion of nucleoside compound and monitoring
the formation of the nucleotide product: conversion (%)=
[product area/(product area+nucleoside area)]ꢂ100.
In the case of araA-MP (12a) and FaraA-MP (13a) syn-
thesis scale-up, due to the low solubility of the starting nu-
Adv. Synth. Catal. 2014, 356, 563 – 570
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
569

FULL PAPERS
Immacolata Serra et al.
[19] R. A. Sheldon, Chem. Soc. Rev. 2012, 41, 1437–1451.
[20] W. Cabri, Catal. Today 2009, 140, 2–10.
[21] J. Tao, J. H. Xu, Curr. Opin. Chem. Biol. 2009, 13, 43–
50.
[32] S. Rocchietti, D. Ubiali, M. Terreni, A. M. Albertini, R.
Fernꢄndez-Lafuente, J. M. Guisꢄn, M. Pregnolato, Bio-
macromolecules 2004, 5, 2195–2200.
[33] I. Serra, C. D. Serra, S. Rocchietti, D. Ubiali, M. Terre-
[22] J. M. Woodley, Trends Biotechnol. 2008, 26, 321–327.
[23] Y. Asano, Y. Mihara, H. Yamada, J. Mol. Catal. B:
Enzym. 1999, 6, 271–277.
ni, Enzyme Microb. Technol. 2011, 49, 52–58.
[34] R. C. Rodrigues, C. Ortiz, A. Berenguer-Murcia, R.
Torres, R. Fernandez-Lafuente, Chem. Soc. Rev. 2013,
42, 6290–6307.
[35] W. Knecht, M. P. B. Sandrini, K. Johansson, H. Eklund,
B. Munch-Petersen, J. Piskur, EMBO J. 2002, 21, 1873–
1880.
[24] Y. Mihara, T. Utagawa, H. Yamada, Y. Asano, J.
Biosci. Bioeng. 2001, 92, 50–54.
[25] N. Tanaka, Z. Hasan, A. F. Hartog, T. van Herk, R.
Wever, Org. Biomol. Chem. 2003, 1, 2833–2839.
[26] T. van Herk, A. F. Hartog, A. M. van der Burg, R.
Wever, Adv. Synth. Catal. 2005, 347, 1155–1162.
[27] R. Mꢃdici, J. I. Garaycoechea, A. L. Valino, C. A. Per-
eira, E. S. Lewkowicz, A. M. Iribarren, Appl. Micro-
biol. Biotechnol. 2013, in press, doi: 10.1007s00253-013-
5194-1.
[28] D. Brady, J. Jordaan, Biotechnol. Lett. 2009, 31, 1639–
1650.
[29] D. N. Tran, K. J. Balkus, ACS Catal. 2011, 1, 956–968.
[30] R. Fernandez-Lafuente, Enzyme Microb. Technol.
2009, 45, 405–418.
[36] C. M. Rosell, R. Fernandez-Lafuente, J. M. Guisan, Bi-
ocatal. Biotransform. 1995, 12, 67–76.
[37] C. Mateo, O. Abian, R. Fernandez-Lafuente, J. M.
Guisan, Biotechnol. Bioeng. 2000, 68, 98–105.
[38] R. Torres, C. Mateo, M. Fuentes, J. M. Palomo, C.
Ortiz, R. Fernandez-Lafuente, J. M. Guisan, A. Tam,
M. Daminati, Biotechnol. Prog. 2002, 18, 1221–1226.
[39] Z. Zou, Q. Ding, L. Ou, B. Yan, Appl. Microbiol. Bio-
technol. 2013, 97, 9389–9395.
[40] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
[41] B. Munch-Petersen, W. Knecht, C. Lenz, L. Sonder-
gaard, J. Piskur, J. Biol. Chem. 2000, 275, 6673–6679.
[31] C. Garcia-Galan, A. Berenguer-Murcia, R. Fernandez-
Lafuente, R. C. Rodrigues, Adv. Synth. Catal. 2011,
353, 2885–2904.
570
asc.wiley-vch.de
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2014, 356, 563 – 570