Synthesis of 2,6-dihalogenated purine nucleosides by thermostable nucleoside phosphorylases

Article abstract of DOI:10.1002/adsc.201400966

The enzymatic transglycosylation of 2,6-dichloropurine (26DCP) and 6-chloro-2-fluoropurine (6C2FP) with uridine, thymidine and 1-(β-D-arabinofuranosyl)-uracil as the pentofuranose donors and recombinant thermostable nucleoside phosphorylases from G. thermoglucosidasius or T. thermophilus as biocatalysts was studied. Selection of 26DCP and 6C2FP as substrates is determined by their higher solubility in aqueous buffer solutions compared to most natural and modified purines and, furthermore, synthesized nucleosides are valuable precursors for the preparation of a large number of biologically important nucleosides. The substrate activity of 26DCP and 6C2FP in the synthesis of their ribo- and 2′-deoxyribo-nucleosides was closely similar to that of related 2-amino- (DAP), 2-chloro- and 2-fluoroadenines; the efficiency of the synthesis of β-D-arabinofuranosides of 26DCP and 6C2FP was lower vs. that of DAP under similar reaction conditions. For a convenient and easier recovery of the biocatalysts, the thermostable enzymes were immobilized on MagReSyn epoxide beads and the biocatalyst showed high catalytic efficiency in a number of reactions. As an example, 6-chloro-2-fluoro-(β-D-ribofuranosyl)-purine (9), a precursor of various antiviral and antitumour drugs, was synthesized by the immobilized enzymes at 60°C under high substrate concentrations (uridine:purine ratio of 2:1, mol). The synthesis was successfully scaled-up [uridine (2.5 mmol), base (1.25 mmol); reaction mixture 50 mL] to afford 9 in 60% yield. The reaction reveals the great practical potential of this enzymatic method for the efficient production of modified purine nucleosides of pharmaceutical interest.

Full text of DOI:10.1002/adsc.201400966

FULL PAPERS
DOI: 10.1002/adsc.201400966
Synthesis of 2,6-Dihalogenated Purine Nucleosides by
Thermostable Nucleoside Phosphorylases
Xinrui Zhou,a Kathleen Szeker,a Lin-Yu Jiao,b Martin Oestreich,b
Igor A. Mikhailopulo,c,* and Peter Neubauera,*
a
Chair of Bioprocess Engineering, Institute of Biotechnology, TechnischeUniversitꢀt Berlin, Ackerstraße 76 ACK24, 13355
Berlin, Germany
Fax : (+49)-030-3142-7577; phone: (+49)-030-3147-2269; e-mail: peter.neubauer@tu-berlin.de
Institute of Chemistry, Technische Universitꢀt Berlin, Straße des 17. Juni 115, 10623 Berlin, Germany
Institute of Bioorganic Chemistry, National Academy of Sciences of Belarus, Acad. Kuprevicha 5/2, 220141 Minsk,
b
c
Belarus
Fax : (+375)-17267-8161; phone: (+375)-17267-8148; e-mail: imikhailopulo@gmail.com
Received: October 8, 2014; Revised: January 29, 2015; Published online: && &&, 0000
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201400966.
Abstract: The enzymatic transglycosylation of 2,6-di- venient and easier recovery of the biocatalysts, the
chloropurine (26DCP) and 6-chloro-2-fluoropurine thermostable enzymes were immobilized on MagRe-
(6C2FP) with uridine, thymidine and 1-(b-d-arabino- Synꢁ epoxide beads and the biocatalyst showed high
furanosyl)-uracil as the pentofuranose donors and re- catalytic efficiency in a number of reactions. As an
combinant thermostable nucleoside phosphorylases example,
6-chloro-2-fluoro-(b-d-ribofuranosyl)-
from G. thermoglucosidasius or T. thermophilus as purine (9), a precursor of various antiviral and anti-
biocatalysts was studied. Selection of 26DCP and tumour drugs, was synthesized by the immobilized
6C2FP as substrates is determined by their higher enzymes at 608C under high substrate concentrations
solubility in aqueous buffer solutions compared to (uriACTHUNTGRNEUNGdine:purine ratio of 2:1, mol). The synthesis was
most natural and modified purines and, furthermore, successfully scaled-up [uridine (2.5 mmol), base
synthesized nucleosides are valuable precursors for (1.25 mmol); reaction mixture 50 mL] to afford 9 in
the preparation of a large number of biologically im- 60% yield. The reaction reveals the great practical
portant nucleosides. The substrate activity of 26DCP potential of this enzymatic method for the efficient
and 6C2FP in the synthesis of their ribo- and 2’- production of modified purine nucleosides of phar-
deoxyACHTUNGTRENNUNGribo-nucleosides was closely similar to that of maceutical interest.
related 2-amino- (DAP), 2-chloro- and 2-fluoroade-
nines; the efficiency of the synthesis of b-d-arabino-
furanosides of 26DCP and 6C2FP was lower vs. that Keywords: biocatalysis; immobilization; nucleoside
of DAP under similar reaction conditions. For a con- phosphorylases; purine nucleosides
Introduction
ribo- (7 and 9) and 2’-deoxyribo-nucleosides (8 and
10) are of particular interest. Indeed, (i) cladribine
Nucleoside analogues are broadly used as antitumour (1), clofarabine (3) and fludarabine (6) have attracted
and antiviral agents. In the living cell, they manifest much attention during the last two decades as drugs
biological effects via diverse metabolic transforma- with great therapeutic potentials, in particular for the
tions, primarily as a result of conversion into deriva- treatment of leukaemia;[2,4,7] (ii) 2-chloroadenosine
tives of phosphoric acid, or themselves by inhibiting (2Cl-Ado, 2) was shown to be an agonist of the ade-
the metabolic pathways or as blockers of the synthesis nosine receptor that induces apoptosis in several cell
of nucleic acids.[1]
lines;[3] (iii) 2’-deoxy-2-fluoroadenosine (2F-dAdo, 4)
Among the great diversity of biologically important and 2-fluoroadenosine (2F-Ado, 5) are the first mem-
nucleosides, adenosine analogues containing a chlorine bers of the adenosine analogues and were first synthe-
(1–3) or fluorine (4–6) atom at the C-2 carbon atom sized in the middle of the last century; they show high
as well as 2,6-dichloro- and 6-chloro-2-fluoropurine cytostatic activity against tumour cells that, unfortu-
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
ÞÞ
These are not the final page numbers!

FULL PAPERS
Xinrui Zhou et al.
Table 1. Structure (see Scheme 1) and bioactivity of halogenated purine nucleosides.
Nucleoside
Cladribine
X
Y
R
Bioactivity
1
2
3
4
5
6
7
8
9
10
NH2
NH2
NH2
NH2
NH2
NH2
Cl
Cl
Cl
Cl
Cl
Cl
Cl
F
F
F
Cl
Cl
F
H
antileukemia; used in clinic[2]
induces apoptosis in several cells[3]
antileukemia; used in clinic[4]
cytotoxic[5]
2Cl-Ado
Clofarabine
2F-dAdo
OH (ribo)
F (arabino)
H
OH (ribo)
OH (arabino)
OH (ribo)
H
2F-Ado
highly cytotoxic[5,6]
Fludarabine
26DCP-R
26DCP-dR
6C2FP-R
6C2FP-dR
antileukemia; used in clinic[7]
smooth muscle relaxant[8]
unknown
OH (ribo)
H
unknown
unknown
F
Scheme 1. Enzymatic synthesis of modified nucleosides.
nately, is accompanied by the high toxicity for normal step, a lack of selectivity of chemical reactions, and
cells.[5.6] Dihalogenated purine nucleosides 7–10 (see the problems associated with the poor regio- and ste-
Table 1) are of great value as precursors for the syn- reoselectivity of chemical reactions (see, e.g.[12] and
thesis of numerous purine nucleosides in the frame- the papers cited therein).
work of the R&D of efficient routes to the prepara-
tion of known substances as well as new nucleosides by the transglycosylation reaction proceeds with strict
of potential biological importance. stereo- and regioselectivity (except for a few specific
In contrast, the enzymatic synthesis of nucleosides
Until now, the vast majority of nucleoside ana- substrates[13]) and requires organic solvents (ethanol,
logues has been synthesized by chemical methods. acetonitrile) only for the isolation of the individual
Two approaches to the synthesis of purine nucleosides desired compound. Obviously, the substrate specificity
have been widely and extensively studied. The first, of the enzyme and the physicochemical properties of
convergent approach consists in the condensation of the substrates are critically important factors of this
a purine base with a suitable pentofuranose derivative reaction. From the viewpoint of the practical synthe-
and subsequent deprotection (see, e.g.,[5,6,9] and the re- sis, nucleoside phosphorylases (NPs) of diverse origin
views[10]). The second comprises the chemical transfor- are presently very useful biocatalysts for the synthesis
mation of the purine base or/and the pentofuranose of base- and pentofuranose-modified nucleosides,[13,14]
fragment of the natural commercially available which include pyrimidine NPs (PyNP, EC 2.4.2.2; UP,
ribo- or 2’-deoxyribo-nucleoside in the desired EC 2.4.2.3; TP, EC 2.4.2.4) and purine NPs (PNP,
compound.[11] Despite the very impressive progress EC 2.4.2.1; MTAP, EC 2.4.2.28). These enzymes cata-
achieved in the development of chemical methods, lyse the reversible phosphorolysis of a nucleoside in
the preparation of purine nucleosides remains a chal- the presence of inorganic phosphate to afford a-d-ri-
lenging problem. Without going into details, it should bofuranose-1-phosphate (a-d-Rib-1P) or 2-deoxy-a-d-
be noted that a common drawback of both ap- ribofuranose-1-phosphate (a-d-dRib-1P) and the nu-
proaches is the indispensable introduction and subse- cleobase. The basic synthetic strategy is illustrated in
quent removal of protective groups, which is associat- Scheme 1.
ed with chromatographic purification in almost every
2
asc.wiley-vch.de
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Synthesis of 2,6-Dihalogenated Purine Nucleosides
Table 2. Temperature dependence of the substrate solubility
The rationality of the transglycosylation reaction
consists in the transfer of the pentofuranose moiety of
a donor, which is a commercially available pyrimidine
nucleoside or its pentofuranose-modified derivative,
to the purine base; the direction of the reaction is de-
termined by the differences in kcat vs. Km of the
purine or pyrimidine phosphorylases, respectively, to-
wards the different substrates. The interplay of two
reversible phosphorolysis reactions results usually in
the formation of the desired purine nucleoside if the
respective PyNP and PNP are combined in a one-pot
reaction. Moreover, it is effective to use pyrimidine
nucleosides as the pentofuranose donor because the
product (a pyrimidine base) from the first step will
not become the competitor of the reactant (the
purine base) for the second step (Scheme 1).
in aqueous solution.[a]
Temp. 2CA 2FA Ade 26DCP 6C2FP 6C2FP-R
[8C]
[mM][b]
25
40
55
65
80
0.1
0.2
0.3
0.4
0.5
0.9
1.4
2.1
2.2
4.0
5.7[c] 15.8[c]
29.0[c]
49.3
53.4
81.8
150.2
79.9
211.5
ND[e]
ND
9.9[d] 17.6
16.5
21.7
36.9
19.8
34.0
38.4
ND
[a]
Solubility was measured by UV spectroscopy at the max-
imum wavelength of each compounds (284–288 nm) in
2 mM potassium phosphate buffer (pH 7.0). 2CA=2Cl-
Ade; 2FA=2F-Ade; Ade=adenine; 26DCP=2,6-di-
chloropurine; 6C2FP=6-chloro-2-fluoropurine; 6C2FP-
R=6-chloro-2-fluoropurineriboside.
The largest relative standard deviation is less than 3%
(n=4).
Determined at 228C.
[b]
A number of halogenated nucleosides has been syn-
thesised by NPs from different sources. For instance,
2-Cl (or F) substituted adenine is accepted as sub-
strate of PNPs from E. coli,[15] Geobacillus stearother-
mophilus[16] and Sulfolobus solfataricus[17] for the syn-
thesis of cladribine 1,[15a–d,16] Fludarabine 6[15e,16] and
[c]
[d]
[e]
Determined at 438C.
ND=not determined.
their analogues; 6-Cl substituted purine is accepted ing points. Enzymes from thermophiles are advanta-
by PNPs from E. coli[18] and Aeromonas hydrophila.[19] geous over the conventional bacterial enzymes for the
The reported conversions resulted in satisfactory synthesis of modified nucleosides. Aside from the ad-
yields of the desired nucleosides (mainly HPLC data), vantage that biocatalytic reactions can be performed
and some of them gave yields in the range from 50% at higher temperatures, also it is likely that these en-
to 80% for the pure individual nucleosides. The use zymes provide a different promiscuity to substrate
of whole bacterial cells as biocatalysts is well docu- modifications that is, due to their evolutionary dis-
mented[14,15] and the synthesis of a number of biologi- tance to E. coli.
cally important nucleosides (AraA, AraG, cladribine,
Secondly, the purine bases 2-chloroadenine (2Cl-
fludarabine, 2-fluorocordycepin etc.) has been de- Ade) and 2-fluoroadenine (2F-Ade) have extremely
scribed. Whole bacterial cells represent a kind of nat- low solubility in aqueous solutions which are most
urally immobilised nucleoside phosphorylases that are suitable for the enzymatic reactions. We tested the
necessary to realise the transglycosylation reactions in effect of temperature on the solubility of 2Cl-Ade
combination with further enzymes which provide pro- and 2F-Ade as well as for the other relevant com-
substrates like arabinose and additional enzymes ena- pounds in aqueous solution in the range from 258C to
bling us to employ pro-substrates like cytosine arabi- 808C and found that the use of the halogenated
noside (AraC) which is used as the arabinofuranose 26DCP and 6C2FP as substrates of the enzymatic syn-
donor due to the presence of cytidine deaminase.[14b,15] thesis is much more promising than 2Cl-Ade and 2F-
However, the wider application of whole cell extracts Ade from the viewpoint of preparative synthesis be-
is restricted due to the following limitations.
cause of their greater solubility (Table 2). Neverthe-
Firstly, broadly used bacterial whole cells as biocat- less, the use of 2Cl-Ade and 2F-Ade in the enzymatic
alysts possess a number of enzymes that can (i) con- synthesis of nucleosides was described in a number of
sume substrates; (ii) catalyse undesired transforma- publications (e.g., refs.[15b,16]).
tions of substrates or formed nucleosides; moreover,
Enzyme immobilisation is considered as an advant-
(iii) secrete cell metabolite(s) in the buffer, which age when enzymes are applied under the relatively
makes the further purification challenging. Purified harsh conditions required in industrial processes. Im-
NPs (e.g., E. coli enzymes) typically exhibit lower ac- mobilisation also allows for a convenient and easier
tivity towards the analogues of the natural substrates, recovery of the biocatalysts and simplifies the design
and they will become inactive under the harsh condi- of a reactor.[22] Considering the multimeric nature of
tions including high temperature, high substrate con- the enzyme,[23] and some examples where NPs were
centration and organic solvents. Even though today successfully immobilised on a hydrophilic support
molecular evolution techniques are available to im- (e.g., epoxide),[24] we chose MagReSyn microspheres
prove the E. coli enzymes[20] or even for adapting the as an immobilisation carrier. These microspheres are
cell to organic solvents,[21] the genetic reservoir of the made of magnetite particles combined with a hydro-
variety of microorganisms probably offers better start- philic polymer with epoxide-functional groups,[25]
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3
ÞÞ
These are not the final page numbers!

FULL PAPERS
Xinrui Zhou et al.
Table 3. Apparent activity of PNPs towards modified purine
base.
which provides high functional group density through-
out the flexible fibre network allowing for stabilisa-
tion of multimeric enzymes. Besides, the magnetic
property makes it easy to remove the biocatalyst from
the reaction medium. The detailed immobilisation
process and the evaluation are presented else-
where.[26]
Purine base[b]
Apparent activity[a] [UmgÀ1]
DgPNP 558C GtPNP 708C ApMTAP 808C
DAP
6C2FP
26DCP
98.96
0.33
0.91
152.76
9.81
4.30
10.62
0.25
0.33
[a]
The apparent activity was determined under the condi-
tions such that the ratio of a-d-Rib-1P to purine base
was 0.6 mM:1 mM and the base conversion rate was in
the linear range. 1 U is the amount of the enzyme con-
verting 1 mmol of purine base per minute under the de-
fined conditions.
In 2 mM phosphate buffer (pH 7.0). DAP=2,6-diamino-
purine; 6C2FP=6-chloro-2-fluoropurine; 26DCP=2,6-
dichloropurine.
Results and Discussion
Scrutiny of the published methods for the chemo-en-
zymatic synthesis of base- and pentofuranose-modi-
fied purine nucleosides prompted us to formulate
a new versatile approach to the synthesis of this class
of biologically important compounds. Our approach
consists in the enzymatic synthesis of 2,6-dihalogenat-
ed purine nucleosides, in particular 2,6-dichloropurine
[b]
(26DCP) and 6-chloro-2-fluoropurine (6C2FP), as val- investigated, and the results are summarised in
uable scaffolds for the enzymatic and/or chemical Table 4. The choice of combinations of the enzymes
preparation of a broad palette of modified nucleo- as well as the reaction temperatures is based on our
sides. The motivation for this work was connected to previous study of the substrate specificities of the
a previous screening study which provided us access used enzymes.[27] At the indicated temperatures,
to a number of PyNPs and PNPs from thermophilic TtPyNP and GtPyNP are interchangeable for the
microorganisms (TtPyNP from Thermus thermophilus, phosphorolysis of uridine and thymidine, however,
GtPyNP and GtPNP from Geobacillus thermoglucosi- AraU showed no little substrate activity for GtPyNP,
dasius, DgPNP from Deinococcus geothermalis, and but is a substrate for TtPyNP (Table 4).
ApMTAP from Aeropyrum pernix) and the study of
The data of Table 4 confirm the excellent substrate
their substrate properties,[27] which allowed us to properties of the bases studied in the synthesis of
choose the most efficient biocatalysts for the synthesis ribo- and 2’-deoxyribo nucleosides. Indeed, the con-
of halogenated purine nucleosides. It is noteworthy version of 26DCP and 6C2FP reached 54–69% with-
that so far dihalogenated purines were not applied as out prior optimisation of the experimental conditions
substrates for the enzymatic synthesis of their nucleo- in regard to the donor:acceptor ratio and the reaction
sides due to their very poor substrate activity for the time (especially in the case of halogenated bases) and
so far described PNPs.
thus in our opinion the results demonstrate the great
The present study was initiated by comparatively potential of the suggested approach for the synthesis
testing the activity of DgPNP, GtPNP and ApMTAP of the corresponding nucleosides on a preparative
under the optimal reaction conditions described earli- scale. The excellent substrate activity of DAP towards
er.[27] With this aim in view, a two-step analysis was PNPs of diverse origin is well known,[1b,13,14] however,
used, that is, phosphorolysis of uridine by TtPyNP or a high conversion rate of the 2Cl-Ade and 2F-Ade
GtPyNP to generate a-d-Rib-1P in step 1. After this into the corresponding ribo- and 2’-deoxyribo nucleo-
a purine base and PNP were added to the reaction sides was unexpected taking into account the ex-
mixture (step 2) and the formation of the correspond- tremely poor solubility of the bases in aqueous media.
ing purine nucleoside was monitored by HPLC. The These data on the whole suggest an unusually high ef-
activity of 2,6-dihalogenated purines was compared ficiency of the GtPNP-catalyzed practically irreversi-
with 2,6-diaminopurine (DAP) that is known to be ble coupling of a base with the intermediates a-d-
one of the best purine substrates for PNPs from di- Rib-1P or a-d-dRib-1P, respectively, promoting the
verse sources. The results are shown in Table 3. As ex- solution of 2Cl-Ade and 2F-Ade in the reaction mix-
pected, DAP showed the best substrate properties for ture.
all PNPs tested; GtPNP showed superior activities vs.
In the case of the synthesis of ara-nucleosides, the
those of DgPNP and ApMTAP towards the tested low phosphorolytic efficiency of TtPyNP along with
purine bases and thus it was selected for the further the lower catalytic power of GtPNP toward a-d-ara-
study.
1P vs. that in regard of the natural a-d-Rib-1P and a-
In the next series of experiments, the enzymatic d-dRib-1P are obviously the main reason for the low
syntheses of ribosides (entries 1–5), 2’-deoxyribosides reaction rate. A careful optimisation of the reaction
(entries 6–10), and arabinosides (entries 11–13) using conditions is the basis to improve the formation rate
GtPNP and TtPyNP or GtPyNP as biocatalysts were and yield of the desired nucleosides further.
4
asc.wiley-vch.de
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Synthesis of 2,6-Dihalogenated Purine Nucleosides
Table 4. Synthesis of modified purine ribonucleosides via enzymatic transglycosylation.[a]
Entry
Pentofuranosyl
Purine base[c]
Synthesized nucleoside
Time/Temp [h/8C]
Conversion [%]
Donor[b]
Urd
R
X
Y
Donor[d]
Base[e]
1
2
3
4
5
6
7
8
OH (ribo)
NH2
NH2
NH2
Cl
NH2
Cl
F
Cl
F
DAP-R
2Cl-Ado (2)
2F-Ado (5)
1/70
1/70
1/70
0.5/65
0.5/65
1/70
1/70
1/70
0.5/65
0.5/65
3/70
3/70
3/70
66.9
45.8
73.3
52.8
54.1
66.8
55.0
83.6
58.6
59.7
54.6
40.8
40.0
95.3
93.3
91.6
53.9
63.0
80.0
80.0
65.9
67.4
69.3
39.5
9.9
26DCP-R (7)
6C2FP-R (9)
DAP-dR
cladribine (1)
2F-dAdo (4)
26DCP-dR (8)
6C2FP-dR (10)
DAP-araR
Cl
Thd
H
NH2
NH2
NH2
Cl
NH2
Cl
F
Cl
F
9
10
11
12
13
Cl
AraU
OH (arabino)
NH2
Cl
Cl
NH2
Cl
F
26DCP-araR (11)
6C2FP-araR (12)
7.2
[a]
Enzymes: TtPyNP 0.1 mgmLÀ1 (5.3 UmLÀ1 towards Urd at 708C) and GtPNP 0.2 mgmLÀ1 (2.0 UmLÀ1 towards 6C2FP
at 708C) (entries 1–3, 6–8, 11–13); GtPyNP (5.1 UmLÀ1 towards Urd at 658C) and GtPNP (0.9 UmLÀ1 towards 6C2FP at
658C), 0.1 mgmLÀ1 each (entries 4–5, 9–10); medium: 2 mM K-phosphate buffer (pH 7.0).
Pentofuranosyl donor: 2 mM. Urd=uridine, Thd=thymidine, AraU=1-(b-d-arabinofuranosyl)uracil.
Purine base: 1 mM. 2Cl-Ade (entries 2 and 7) can only reach 0.6 mM.
[b]
[c]
[d]
[e]
Yield (according to HPLC) based on pyrimidine nucleosides (donor) conversion.
Yield (according to HPLC) based on purine base conversion.
With this aim in view, TtPyNP and GtPNP were
In small-scale reactions (0.4 mL), under optimal
immobilised on MagReSynꢁ epoxide beads [polyethy- conditions, the conversion of 26DCP and 6C2FP into
leneimine-based iron oxide magnetic microspheres the corresponding ribo-nucleosides 7 and 9 was veri-
with epoxide functional groups (from ReSyn Biosci- fied by HPLC to be 78% and 85%, respectively (see
ences, Pretoria, South Africa] for a convenient and the Experimental Section for details).
easier recovery of the biocatalysts. The activities of
To validate the efficiency of the immobilised en-
the immobilized enzymes were measured by the activ- zymes for the preparation of pure nucleosides on the
ity assay as previously described[27] and the following multimilligram scale, the synthesis of 6-chloro-2-
data were obtained: TtPyNP – 4.8 mg enzyme were fluoro-9-(b-d-ribofuranosyl)-purine (9) was studied.
loaded per mL beads, 41% residual activity; GtPNP – The reaction mixture (in 2 mM K-phosphate buffer,
16.5 mg enzyme were loaded per mL beads, 83% re- pH 7.0; total volume 50 mL) containing 50 mM Urd,
sidual activity. It should be stressed that a high residu- 25 mM 6C2FP was kept at 608C for 20 h and the reac-
al activity of immobilised enzymes points to the pres- tion progress was monitored by HPLC. During this
ence of all subunits in the biocatalyst. Noteworthy, period, the formation of nucleoside 9 reached 75%
the enzymatic activity of E. coli PNP that is a trimer and the reaction mixture was worked-up (removing
of dimers was recently investigated in detail and it the biocatalyst by a magnet, silica gel column chroma-
was convincingly proven that only the hexameric tography) and resulted in a 60% yield of the product
structure displays activity.[23] It was found that Urd which had a purity of 98% (HPLC).
and 6C2FP in concentrations up to 80 mM and
The results of the synthesis of nucleoside 9 prompt-
55 mM, respectively, do not inhibit the immobilised ed us to extend this work to the preparation of nu-
enzymes.
cleosides 7, 8 and 10 aiming to exploit the new effi-
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
5
ÞÞ
These are not the final page numbers!

FULL PAPERS
Xinrui Zhou et al.
Scheme 2. Chemo-enzymatic routes for the synthesis of cladribine and fludarabine.
cient chemo-enzymatic routes to known and new Experimental Section
modified purine nucleosides. This work is still in prog-
ress. Emphasis is made on the role of the successive General
treatments for the efficient synthesis of purine nucleo-
All chemicals and solvents were of analytical grade or
higher and purchased, if not stated otherwise, from Sigma–
Aldrich (Steinheim, Germany), Carl Roth (Karlsruhe, Ger-
many), TCI Deutschland (Eschborn, Germany), and VWR
(Darmstadt, Germany). Water was purified by a purification
system from Merck Millipore (Schwalbach, Germany). Ma-
gReSynꢁ Epoxide beads were a kind gift from ReSyn Bio-
sciences (Pretoria, South Africa). The UV spectra and ab-
sorption of purine bases were recorded with a Biochrom Ul-
trospec 3300 pro UV/Visible Spectrophotometer (Cam-
bridge, England). HPLC analyses were carried out with an
Agilent 1200 series system equipped with an Agilent DAD
detector using a Phenomenex (Torrance, United States) re-
versed phase C18 column (150ꢃ4.60 mm). Column chroma-
tography was performed on silica gel 60 (0.043–0.06 mm,
230–400 mesh, ASTM) from Grace GmbH. NMR data were
recorded on Bruker AV 500 instruments. High resolution
mass spectrometry (HR-MS) analyses were performed by
the Analytical Facility at the Institut fꢄr Chemie of the
Technische Universitꢀt Berlin.
sides, e.g., cladribine (1) and clofarabine (3), using
either ribonucleosides 7 and 9 (first B then A, and
first D then C) or their corresponding 2-chloro- and
2-fluoroadenine derivatives (first A then B, and first
C then D) as donors for the relevant purine bases in
the transglcosylation reactions (Scheme 2) (cf. ref.[28]).
Conclusions
In summary, we have studied the possibility of appli-
cation of different enzyme pairs (GtPNP and TtPyNP,
GtPNP and GtPyNP) as biocatalysts for the synthesis
of various ribo-, 2’-deoxyribo- and arabino-purine nu-
cleosides. We found that DAP, 2Cl-Ade, 2F-Ade,
26DC-P and 6C2FP are effectively transformed into
the corresponding ribo- and 2’-deoxyribo-nucleosides;
while the formation of the arabino-nucleosides of
DAP, 26DC-P and 6C2FP had a lower efficiency. Im-
mobilisation of GtPNP and TtPyNP on MagReSynꢁ
epoxide beads resulted in efficient transformation of
26DC-P and 6C2FP into the corresponding ribo-nu-
cleosides as it was exemplary shown by the synthesis
Solubility Test
Calibration curves were made by UV spectroscopy for each
of 6-chloro-2-fluoro-9-(b-d-ribofuranosyl)-purine (9) compound to be tested. The wavelengths of maximum ab-
sorption are: 285 nm (2Cl-Ade, e=624.6 MÀ1 cmÀ1), 284 nm
on a hundred milligram scale. The results of this study
(2F-Ade,
261.5 MÀ1 cmÀ1), 288 nm (26DCP, e=3215 MÀ1 cmÀ1), 285 nm
(6C2FP, e=2335 MÀ1 cmÀ1 6C2FP-R, e=1627 MÀ1 cmÀ1).
e=284.9 MÀ1 cmÀ1),
284 nm
(Ade,
e=
in toto represent an interesting example of the oppor-
tunities that are offered by the exploitation of new
enzymes from thermophilic organisms for the efficient
production of base and pentofuranose modified
purine nucleosides.
;
The saturated solutions suspended in 2 mM phosphate
buffer (pH 7.0) were incubated at different temperatures in
a water bath or on a thermomixer for 30 min. The superna-
tant was withdrawn and cooled down shortly, and then dilut-
ed to the suitable concentration for the absorption measure-
ment. Solubility was calculated according to the saturation
concentration at the defined temperature.
6
asc.wiley-vch.de
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Synthesis of 2,6-Dihalogenated Purine Nucleosides
30 min at 708C and HPLC analysis showed a 78% base con-
version for the nucleoside 7.
Expression and Preparation of Enzymes
TtPyNP, GtPyNP, GtPNP, DgPNP and ApMTAP were ex-
pressed in E. coli BL21 and purified by heat treatment and
standard Ni-NTA affinity chromatography as previously de-
The optimal conditions for the synthesis of nucleoside 9
were: 0.4 mL reaction containing 50 mM Urd, 20 mM
6C2FP, 2 mM K-phosphate (pH 7.0), 4.1 U TtPyNP (towards
Urd at 708C) and 5.4 U GtPNP (towards 6C2FP at 708C).
The reaction mixture was shaken at 1000 rpm for 180 min at
708C and HPLC analysis showed an 85% base conversion
for the nucleoside 9.
scribed.[27] Purified TtPyNP has
a specific activity of
53 UmgÀ1 at 708C and GtPyNP 51 UmgÀ1 at 658C towards
Urd in 50 mM phosphate buffer (pH 7).
Determination of PNP Apparent Activity towards
Purine Base
Synthesis and Purification of 6-Chloro-2-fluoro-9-(b-
d-ribofuranosyl)-purine (9)
The procedure contains two steps: (i) 0.5 mL 4 mMUrd was
incubated with 0.1 mg GtPyNP or TtPyNP in 2 mM phos-
phate buffer (pH 7) for 5 min at the temperature of the
second step. At the end of the reaction, the equilibrium was
reached (30% Urd was converted to Ura and a-d-Rib-1P).
(ii) 0.5 mL 2 mM purine base in 2 mM phosphate buffer
(pH 7) with PNP (the concentration was selected in such
a way that the reaction rate was in the linear range) was
added into the first step reaction for 2 min at 558C
(DgPNP), 658C (GtPNP) or 808C (ApMTAP). The reaction
mixtures were analysed by HPLC (see the Supporting Infor-
mation). One unit PNP apparent activity was defined as the
amount of PNP which catalyses the conversion of 1 mmol
purine base per minute under the aforementioned condi-
tions.
The reaction mixture (50 mL) containing uridine (0.61 g,
2.5 mmol), 6-chloro-2-fluoropurine (0.216 g, 1.25 mmol), po-
tassium phosphate (0.1 mmol, pH 7.0) and immobilized bio-
catalysts (TtPyNP: 8 mg, ca. 67 U, towards Urd at 608C;
GtPNP: 15 mg, ca. 87 U, towards 6C2FP at 608C) was pre-
pared in a 50 mL-falcon tube. This tube was closed and ro-
tated in an end-over-end rotator at 55–608C in a water bath
for 20 h. A 75% conversion for the base was determined by
HPLC. The biocatalyst was removed by a permanent
magnet and the product was isolated by silica gel column
chromatography (dichloromethane/ethanol 15:1 and then
ethyl acetate/acetone 7:3) to afford the desired product 9 as
a white powder; yield: 0.230 g (0.75 mmol, 60%; HPLC
purity 98%) (see the Supporting Information for more de-
tails).
Enzymatic Synthesis of Halogenated Purine
Nucleosides
Acknowledgements
The reaction mixture of 0.4 mL contained 2 mM pentofura-
nosyl donor (Urd, Thd or AraU), 1 mM purine base, 2 mM
phosphate buffer (pH 7.0), 0.1 mgmLÀ1TtPyNP (or
GtPyNP) and 0.1–0.2 mgmLÀ1GtPNP. Reactions were per-
formed on at thermomixer at defined temperatures for 0.5
to 3 h (Table 4) and stopped by adding ice-cold water at
a 1:1 dilution. Controls (Table 4, entries 2–3, 7–8, 11–13)
without enzyme were performed in parallel. The conversion
rates were calculated from the integration of the corre-
sponding HPLC peaks and are presented as the percentage
of converted donor or base, respectively.
We are grateful to Dr. Justin Jordaan (ReSyn Biosciences,
Pretoria, South Africa) for the kind donation of the MagRe-
Synꢀ Epoxide beads as well as useful suggestions for the im-
mobilization experiments. This work is part of the Cluster of
Excellence “Unifying Concepts in Catalysis” coordinated by
the Technische Universitꢁt Berlin. Financial support by the
Deutsche Forschungsgemeinschaft (DFG) within the frame-
work of the German Initiative for Excellence is gratefully ac-
knowledged (EXC 314). X.Z. gratefully acknowledges finan-
cial support by an ESF-grant of the TU Berlin. I.A.M. is
deeply thankful to the A. von Humboldt-Stiftung (Bonn-Bad-
Godesberg, Germany) and the Byelorussian Republican
Foundation for Fundamental Research (www.fond.bas-net.by;
project #X13MC-027) for partial financial support. L.-Y.J.
thanks the China Scholarship Council (CSC) for a predoctor-
al fellowship (2011–2015).
Immobilisation of TtPyNP and GtPNP
TtPyNP and GtPNP were immobilised on MagReSynꢁ ep-
oxide beads (ReSyn Biosciences, Pretoria, South Africa) ac-
cording to ReSynꢅs manual with a slight modification. Brief-
ly, 0.04 mL suspension beads (1 mg dry beads) were loaded
with TtPyNP (0.19 mg; 4.1 U towards Urd at 708C) or
GtPNP (0.66 mg; 2.4 U and 5.4 U towards 26DCP and
6C2FP at 708C, respectively) in a total volume of 0.40 mL References
binding buffer (50 mM potassium phosphate buffer, pH 6.0)
and incubated for 4 h on the thermomixer at 950 rpm, 508C.
The quenching step was skipped. Beads were washed and
stored at 48C in 50 mM potassium phosphate buffer
(pH 7.0) until further use.
The optimal conditions for the synthesis of the nucleoside
7 were: 0.4 mL reaction containing 20 mM Urd, 7 mM
26DCP, 1.4 mM K-phosphate (pH 7.0), 4.1 U TtPyNP (to-
wards Urd at 708C) and 2.4 U GtPNP (towards 26DCP at
708C). The reaction mixture was shaken at 1000 rpm for
[1] a) P. Herdewijin, Modified Nucleosides: in Biochemistry
Biotechnology and Medicine, Wiley-VCH, Weinheim,
2008; b) I. A. Mikhailopulo, A. I. Miroshnikov, Acta
Naturae 2010, 2, 36-59; c) W. Sneader, Drug Discovery:
A History, John Wiley & Sons, London, 2005.
[2] T. Robak, P. Robak, Curr. Pharm. Des. 2012, 18, 3373–
3388.
[3] a) I. Bellezza, A. Tucci, A. Minelli, Anticancer Agents
Med. Chem. 2008, 8, 783–789; b) L. Bastin-Coyette, C.
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
7
ÞÞ
These are not the final page numbers!

FULL PAPERS
Xinrui Zhou et al.
Smal, S. Cardoen, P. Saussoy, E. Van Den Neste, F.
Bontemps, Biochem. Pharmacol. 2008, 75, 1451–1460.
[4] P. L. Bonate, L. Arthaud, W. R. Cantrell Jr, K. Stephen-
son, J. A. Secrist III, S. Weitman, Nat. Rev. Drug Dis-
covery 2006, 5, 855–863.
[17] R. M. Arevalo, V. M. D. Thomas, C. L. Gꢆmez, M. P.
Gilabert, C. E. Company, J. C. Boliart, (Institut Univ.
de Ciꢇncia
i
Tecnologia), European Patent EP
2338985A1, 2011.
[18] T. A. Krenitsky, G. W. Koszalka, J. V. Tuttle, Biochem-
[5] J. A. Montgomery, K. Hewson, J. Med. Chem. 1969, 12,
498–504.
[6] J. A. Montgomery, K. Hewson, J. Am. Chem. Soc. 1957,
79, 4559.
[7] H. Wallen, J. A. Thompson, J. Z. Reilly, R. M. Rod-
myre, J. Cao, C. Yee, PLoS One 2009, 4, e4749.
[8] S. G. McKenzie, R. Frew, H. P. Bar, Eur. J. Pharmacol.
1977, 41, 183–192.
[9] G. M. Blackburn, M. J. Gait, D. Loakes, D. M. Williams,
Nucleic Acids in Chemistry and Biology, 3rd edn., Royal
Society of Chemistry, Cambridge, 2006.
istry 1981, 20, 3615–3621.
[19] D. Ubiali, C. D. Serra, I. Serra, C. F. Morelli, M. Terre-
ni, A. M. Albertini, P. Manitto, G. Speranza, Adv.
Synth. Catal. 2012, 354, 96–104.
[20] a) D. F. Visser, F. Hennessy, J. Rashamuse, B. Pletschke,
D. Brady, J. Mol. Catal. B: Enzym. 2011, 68, 279–285;
b) D. V. Chebotaev, L. B. Gul’ko, V. P. Veiko, Russ. J.
Bioorg. Chem. 2001, 27, 160–166.
[21] R. Aono, K. Aibe, A. Inoue, K. Horikoshi, Agr. Biol.
Chem. Tokyo. 1991, 55, 1935–1938.
[22] C. Mateo, J. M. Palomo, G. Fernandez-Lorente, J. M.
Guisan, R. Fernandez-Lafuente, Enzyme Microb. Tech-
nol. 2007, 40, 1451–1463.
[10] a) H. Vorbrꢄggen, C. Ruh-Pohlenz, in: Organic Reac-
tions, John Wiley & Sons, Inc., New York, 2004;
b) J. K. Watts, M. J. Damha, Can. J. Chem. 2008, 86,
641–656.
ˇ
ˇ
´
[23] a) B. Bertosa, G. Mikleusevic, B. Wielgus-Kutrowska,
ˇˇ ´
´
ˇ
´
M. Narczyk, M. Hajnic, I. Lescic Asler, S. Tomic, M.
[11] a) K. A. Watanabe, S. E. Patterson, in: Frontiers in Nu-
cleosides and Nucleic Acids, (Eds.: R. F. Schinazi, D. C.
Liotta), IHL Press, Tucker, GA 2005, pp 3–56; b) K. W.
Pankiewicz, Carbohydr. Res. 2000, 327, 87–105.
[12] a) B. G. Anderson, W. E. Bauta, J. W. R. Cantrell, T.
Engles, D. P. Lovett, Org. Process Res. Dev. 2008, 12,
1229–1237; b) Y. Cen, A. A. Sauve, Nucleos. Nucleot.
Nucl. 2010, 29, 113–122; c) T. Tennilꢀ, E. Azhayeva, J.
Vepsꢀlꢀinen, R. Laatikainen, A. Azhayev, I. A. Mikhai-
lopulo, Nucleos. Nucleot. Nucl. 2000, 19, 1861–1884.
[13] I. A. Mikhailopulo, Curr. Org. Chem. 2007, 11, 317–
335.
´
Luic, A. Bzowska, FEBS J. 2014, 281, 1860–1871;
b) M. J. Pugmire, S. E. Ealick, Biochem. J. 2002, 361, 1–
25.
ˇ
[24] a) I. Serra, D. Ubiali, J. Piskur, S. Christoffersen, E. S.
Lewkowicz, A. M. Iribarren, A. M. Albertini, M. Terre-
ni, ChemPlusChem 2013, 78, 157–165; b) I. Serra, T.
Bavaro, D. A. Cecchini, S. Daly, A. M. Albertini, M.
Terreni, D. Ubiali, J. Mol. Catal. B: Enzym. 2013, 95,
16–22; c) I. Serra, C. D. Serra, S. Rocchietti, D. Ubiali,
M. Terreni, Enzyme Microb. Technol. 2011, 49, 52–58;
d) D. Ubiali, S. Rocchietti, F. Scaramozzino, M. Terreni,
A. M. Albertini, R. Fernꢈndez-Lafuente, J. M. Guisꢈn,
M. Pregnolato, Adv. Synth. Catal. 2004, 346, 1361–1366.
[25] D. Brady, J. Jordaan, Biotechnol. Lett. 2009, 31, 1639–
1650.
[26] X. Zhou, I. A. Mikhailopulo, M. N. Cruz Bournazou, P.
Neubauer, J. Mol. Catal. B: Enzym. 2015, 115, 119–127.
[27] a) X. Zhou, K. Szeker, B. Janocha, T. Bçhme, D. Al-
brecht, I. A. Mikhailopulo, P. Neubauer, FEBS J. 2013,
280, 1475–1490; b) K. Szeker, X. Zhou, T. Schwab, A.
Casanueva, D. Cowan, I. A. Mikhailopulo, P. Neubauer,
J. Mol. Catal. B: Enzym. 2012, 84, 27–34.
[14] I. A. Mikhailopulo, A. I. Miroshnikov, Mendeleev
Commun. 2011, 21, 57–68.
[15] a) I. A. Mikhailopulo, A. I. Zinchenko, Z. Kazimierc-
zuk, V. N. Barai, S. B. Bokut, E. N. Kalinichenko, Nu-
cleos. Nucleot. 1993, 12, 417–422; b) V. N. Barai, A. I.
Zinchenko, L. A. Eroshevskaya, E. N. Kalinichenko,
T. I. Kulak, I. A. Mikhailopulo, Helv. Chim. Acta 2002,
85, 1901–1908; c) H. Komatsu, T. Araki, Nucleos. Nu-
cleot. Nucl. 2005, 24, 1127–1130; d) G. Zuffi, S. Mon-
ciardini, (Explora Laboratories SA, Switzerland), Euro-
pean Patent EP1932918A1, 2008; e) T. A. Krenitsky,
G. B. Elion, J. E. Rideout, (The Wellcome Foundation
Limited), European Patent EP0002192B1, 1981; f) J. V.
Tuttle, M. Tisdale, T. A. Krenitsky, J. Med. Chem. 1993,
36, 119–125.
[28] a) A. I. Zinchenko, V. N. Barai, S. B. Bokut, E. I. Kva-
syuk, I. A. Mikhailopulo, Appl. Microbiol. Biotechnol.
1990, 32, 658–661; b) V. B. Berzin, E. V. Dorofeeva,
V. N. Leonov, A. I. Miroshnikov, Russ. J. Bioorg. Chem.
2009, 35, 193–196.
[16] S. Taran, K. Verevkina, S. Feofanov, A. Miroshnikov,
Russ. J. Bioorg. Chem. 2009, 35, 739–745.
8
asc.wiley-vch.de
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 0000, 000, 0 – 0
ÝÝ
These are not the final page numbers!

FULL PAPERS
Synthesis of 2,6-Dihalogenated Purine Nucleosides by
Thermostable Nucleoside Phosphorylases
9
Adv. Synth. Catal. 2015, 357, 1 – 9
Xinrui Zhou, Kathleen Szeker, Lin-Yu Jiao,
Martin Oestreich, Igor A. Mikhailopulo,* Peter Neubauer*
Adv. Synth. Catal. 0000, 000, 0 – 0
ꢂ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
9
ÞÞ
These are not the final page numbers!