Enzymatic regioselective and complete deacetylation of two arabinonucleosides

Article abstract of DOI:10.1016/j.molcatb.2009.10.011

Candida antarctica lipase B (CAL-B)-catalysed regioselective deacetylation of 2′,3′,5′-tri- O-acetyl-1-β- d-arabinofuranosyluracil (1) and 2′,3′,5′-tri- O-acetyl-9-β- d-arabinofuranosyladenine (2) was studied. The choice of the reaction medium allowed the regioselective formation of products bearing different degree of acetylation: in isopropanol, CAL-B catalysed the formation of the corresponding 2′- O-acetylated arabinonucleosides, while hydrolyses afforded the 2′,3′-di- O-acetylated products. In particular, the procedure herein described allows a simple and efficient preparation of the reported vidarabine prodrug 2′,3′-di- O-acetyl-9-β- d-arabinofuranosyladenine, avoiding the utilisation of protective groups. Moreover, to achieve full deacetylation of the assayed substrates, a set of commercial hydrolases and fungal keratinases from Doratomyces microsporus (DMK) and Paecilomyces marquandii (PMK) were tested. While only PMK and DMK catalysed the quantitative complete deacetylation of 1, DMK accomplished full deacetylation of 2 in shorter time than the other assayed enzymes.

Full text of DOI:10.1016/j.molcatb.2009.10.011

Journal of Molecular Catalysis B: Enzymatic 62 (2010) 225–229
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Enzymatic regioselective and complete deacetylation of two
arabinonucleosides
María B. Sabainia, María A. Zinnia, Martina Mohorcˇicˇ b, Jozˇefa Friedrichb,
Adolfo M. Iribarrena,c, Luis E. Iglesiasa,∗
a Department of Science and Technology, Universidad Nacional de Quilmes, Roque Sáenz Pe˜na 352, 1876 Bernal, Provincia de Buenos Aires, Argentina
b Department of Biotechnology, National Institute of Chemistry, Hajdrihova 19, SI-1000 Ljubljana, Slovenia
c INGEBI (CONICET), Vuelta de Obligado 2490, 1428 Buenos Aires, Argentina
a r t i c l e i n f o
a b s t r a c t
Candida antarctica lipase B (CAL-B)-catalysed regioselective deacetylation of 2ꢀ,3ꢀ,5ꢀ-tri-O-acetyl-1-␤-d-
arabinofuranosyluracil (1) and 2ꢀ,3ꢀ,5ꢀ-tri-O-acetyl-9-␤-d-arabinofuranosyladenine (2) was studied. The
choice of the reaction medium allowed the regioselective formation of products bearing different degree
of acetylation: in isopropanol, CAL-B catalysed the formation of the corresponding 2ꢀ-O-acetylated arabi-
nonucleosides, while hydrolyses afforded the 2ꢀ,3ꢀ-di-O-acetylated products. In particular, the procedure
herein described allows a simple and efficient preparation of the reported vidarabine prodrug 2ꢀ,3ꢀ-
di-O-acetyl-9-␤-d-arabinofuranosyladenine, avoiding the utilisation of protective groups. Moreover, to
achieve full deacetylation of the assayed substrates, a set of commercial hydrolases and fungal keratinases
from Doratomyces microsporus (DMK) and Paecilomyces marquandii (PMK) were tested. While only PMK
and DMK catalysed the quantitative complete deacetylation of 1, DMK accomplished full deacetylation
of 2 in shorter time than the other assayed enzymes.
Article history:
Received 4 September 2009
Received in revised form 16 October 2009
Accepted 27 October 2009
Available online 31 October 2009
Keywords:
Lipase
Regioselectivity
Nucleosides
Prodrugs
Keratinase
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
tion of modified nucleosides [6,23,24], can be obtained by
CAL-B-catalysed alcoholysis through a simple and regioselec-
Nucleoside analogues are at present well recognised therapeu-
tical compounds [1]; among them, arabinonucleosides have found
an important place as antitumoral [2] and antiviral [3–5] drugs.
However, in order to get agents with better therapeutic efficacy,
the delivery of nucleosides can be improved by the use of prodrugs,
which provide derivatives with better pharmacokinetic properties
and increased metabolical stability [6–12]. To obtain lipophilic pro-
drugs, regioselective derivatisation of the furanosic hydroxyls is
required, a goal traditionally achieved in organic synthesis through
protective groups [6,12]. In contrast to this methodology, hydro-
lases are employed nowadays to obtain regioselectively acylated or
alkoxycarbonylated nucleosides [13,14]; in spite of this, only few
among these works deal with hydrolase-catalysed transformation
of arabinonucleosides [7,8,15–18].
tive procedure. With this in mind, we applied this enzymatic
deacylation to 6-azauridine and 3-deazauridine [25] and keep-
ing this aim, now we extended it to arabinonucleosides. In
the present work we studied CAL-B-catalysed deacetylation of
2ꢀ,3ꢀ,5ꢀ-tri-O-acetyl-1-␤-d-arabinofuranosyluracil (1, Scheme 1), a
model of pyrimidine arabinonucleoside, and the purine arabi-
nonucleoside 2ꢀ,3ꢀ,5ꢀ-tri-O-acetyl-9-␤-d-arabinofuranosyladenine
(2, Scheme 1), the triacetylated derivative of vidarabine (9-␤-d-
arabinofuranosyladenine, Ara-A, 8, Scheme 1). Vidarabine was the
first antiviral nucleoside analogue licensed for the treatment of her-
pes virus in humans. It is also active against a wide variety of viruses,
such as adenovirus causing haemorragic cystitis, poxviruses and
certain rhadboviruses, hepadnarviruses and RNA tumor viruses
[4,5,12]. We report herein the results obtained during enzy-
matic deacetylation of the above mentioned arabinonucleosides 1
and 2.
In our laboratory we have been studying Candida antarctica
B lipase (CAL-B)-catalysed alcoholysis of acylated nucleosides
[19–22], which provides a regioselective access to 2ꢀ,3ꢀ-di-O-
acylribonucleosides [19–21]. Thus, diacylated nucleosides, which
are used as prodrugs and as intermediates in the prepara-
2. Experimental
2.1. General
∗
All employed reagents and solvents were of analytical grade and
obtained from commercial sources.
Corresponding author. Tel.: +54 11 4365 7182; fax: +54 11 4365 7182.
E-mail address: leiglesias@unq.edu.ar (L.E. Iglesias).
1381-1177/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2009.10.011

226
M.B. Sabaini et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 225–229
Scheme 1.
TLC was performed on Silicagel 60 F254 plates (Merck) using
ously described [27], and were used for deacetylation experiments
without further treatment.
dichloromethane/methanol mixtures and column chromatography
was carried out using silica C18 Amicon Corporation (pore diame-
ter: 100 Å, particle size: 20 ␮m).
2.3. Preparation of substrates 1 and 2
HPLC analyses were carried out using a C18 column (length:
150 mm; internal diameter: 4.6 mm; particle size: 5 ␮m) with
detection at 254 nm. For the analysis of aliquots from biotransfor-
mations of 1, a 30 min gradient of water/acetonitrile from 95:5 (v/v)
to 60:40 (v/v) at a flow rate of 0.9 ml min−1 was employed. Samples
involving 2 were analysed by a 26 min gradient of the same solvent
mixture from 90:10 (v/v) to 60:40 (v/v) at the same flow rate.
NMR spectra were recorded on a Bruker AC-500 spectrometer
in DMSO-d6, at 500 MHz for 1H and 125 MHz for 13C.
Substrates 1 and 2 (Scheme 1) were prepared by reaction of
the corresponding free arabinonucleoside 7 and 8 (1 mmol) with
acetic anhydride (3.6 mmol) in acetonitrile (15 ml) containing tri-
ethylamine (4 mmol) and 4-dimethylaminopyridine (0.1 mmol),
according to
a
previously reported protocol [28]. 2ꢀ,3ꢀ,5ꢀ-
Tri-O-acetyl-1-␤-d-arabinofuranosyluracil (1) was purified by
recrystallisation from isopropanol and 2ꢀ,3ꢀ,5ꢀ-tri-O-acetyl-9-␤-d-
arabinofuranosyladenine (2) by silicagel column chromatography
eluting with dichloromethane/methanol 90:10 (v/v). Both sub-
strates afforded satisfactory NMR data.
2.2. Enzymes
Enzymatic reactions were carried out in
a temperature-
2.4. Preparation of reference samples of
2ꢀ-O-acetyl-1-ˇ-d-arabinofuranosyluracil (3) and
2ꢀ-O-acetyl-9-ˇ-d-arabinofuranosyladenine (4)
controlled incubator shaker (Sontec OS 11, Argentina) at 200 rpm
and 30 ◦C. Control experiments carried out in the absence of
enzyme showed no conversion of the substrates.
Reference samples of 3 and 4 were prepared through respective
PLE-catalysed hydrolysis of 1 and 2 according to [7] and isolating
the resulting products using C18 flash chromatography. Follow-
ing the HPLC conditions for 1 reported in Section 2.1, 3 gave
tr = 7.17 min, while a tr = 5.05 min was obtained for 4 according to
the conditions described for 2.
2.2.1. Assayed commercial hydrolases
The following enzymes were purchased from Sigma–Aldrich
Co.: lipases from Candida rugosa (CRL, Sigma Type VII, 875 U/mg
solid), Pseudomonas sp. (PSL, Sigma Type XIII, 21 U/mg solid) and
Pseudomonas cepacia (PCL, 90 U/g solid); esterase from pig liver
(PLE, 19 U/mg solid); proteases from Tritirachium album (proteinase
K, PK, 6.1 U/mg solid) and Bacillus licheniformis (subtilisin, Sigma
Type VIII, 11 U/mg solid) was purchased from ICN. Lipase B from
Candida antarctica (CAL-B, Novozym 435, 10,000 PLU/mg solid;
PLU: Propyl Laurate Units) was a generous gift from Novozymes
(Brazil).
2.5. General procedure for the enzymatic alcoholysis of 1 and 2
Typically, experiments of enzymatic alcoholyses were per-
formed by adding CAL-B (12 mg) to a solution of the substrate
(0.04 mmol) in the assayed alcohol at the indicated A/S (alco-
hol/substrate ratio) (10.4 mmol for A/S = 260 or 40 mmol for
A/S = 1000) and shaking the resulting mixtures at 200 rpm and
30 ◦C. Aliquots from the biotransformations were withdrawn at dif-
ferent times and after removal of the enzyme, analysed by TLC and
HPLC.
Enzymes were used straight without any further treatment or
purification.
2.2.2. Keratinolytic enzymes from Paecilomyces marquandii
(PMK) and Doratomyces microsporus (DMK)
For preparation of monoacetylated nucleosides 3 and 4, the
above described protocol for alcoholysis was employed as follows.
Keratinases of the filamentous fungi Doratomyces microsporus
(DMK) and Paecilomyces marquandii (PMK) were prepared by aer-
obic submerged fermentation followed by partial purification as
reported by Gradisˇar et al. [26] and stored as freeze-dried powder
at −20 ◦C. The obtained enzyme powders of DMK and PMK showed
activities of 7.8 U mg−1 and 23.0 U mg−1, respectively, measured
on keratin from human stratum corneum as a substrate as previ-
2.5.1. 2ꢀ-O-Acetyl-1-ˇ-d-arabinofuranosyluracil (3)
A mixture of 1 (0.24 mmol), isopropanol (18.6 ml, A/S = 1000)
and CAL-B (72 mg) were shaken for 16 days. The biotransformations
were then stopped by filtering off the enzyme and washing it with

M.B. Sabaini et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 225–229
227
methanol and dichloromethane. Vacuum evaporation of the filtrate
gave 3, which provided satisfactory spectral data: 1H NMR (DMSO-
d6, 500 MHz): ı 1.95 (s, 3H, –CH3), 3.63 (m, 2H, H-5ꢀ), 3.80 (m, 1H,
H-4ꢀ), 4.12 (t, J = 5.0 Hz, 1H, H-3ꢀ), 5.16 (t, J = 5.0 Hz, 1H, H-2ꢀ), 5.64
(d, J = 8.1 Hz, 1H, H-5), 6.15 (d, J = 5.0 Hz, 1H, H-1ꢀ), 7.71 (d, J = 8.1 Hz,
1H, H-6), 11.35 (s, 1H, H-3). 13C NMR (DMSO-d6, 125 MHz): ı 20.71
(–CH3), 62.51 (C-5ꢀ), 72.76 (C-3ꢀ), 77.46 (C-2ꢀ), 82.85 (C-4ꢀ), 84.02
(C-1ꢀ), 101.30 (C-5), 141.60 (C-6), 150.51 (C-2), 163.45 (C-4), 169.51
(CO).
reaction was then stopped, the buffer evaporated in vacuo and the
resulting crude subsequently purified by C18 column chromatog-
raphy, eluting with water/acetonitrile 85:15 (v/v), to give 6: 1H
NMR (DMSO-d6, 500 MHz): ı 1.75 (s, 3H, –CH3), 2.11 (s, 3H, –CH3),
3.69 (dd, J1 = 4.3 Hz, J2 = 12.2 Hz, 1H, H-5ꢀa), 3.76 (dd, J1 = 4.3 Hz,
J2 = 12.1 Hz, 1H, H-5ꢀb), 4.11 (m, 1H, H-4ꢀ), 5.52 (m, 2H, H-2ꢀ and
H-3ꢀ), 6.51 (d, J = 4.6 Hz, 1H, H-1ꢀ), 7.35 (s, 2H, NH2), 8.16 (s, 1H, H-
2), 8.31 (s, 1H, H-8). 13C NMR (DMSO-d6, 125 MHz): ı 20.42, 21.11
(–CH3s), 60.73 (C-5ꢀ), 74.93 (C-3ꢀ), 75.52 (C-2ꢀ), 81.59 (C-4ꢀ), 82.29
(C-1ꢀ), 118.67 (C-5), 140.08 (C-8), 149.57 (C-2), 153.21 (C-4), 156.45
(C-6), 169.21, 170.14 (COs).
2.5.2. 2ꢀ-O-Acetyl-9-ˇ-d-arabinofuranosyladenine (4)
Analogously, 2 (0.24 mmol), isopropanol (18.6 ml, A/S = 1000)
and CAL-B (72 mg) were shaken for 12 days. The enzyme was fil-
tered off, washed with methanol and dicholoromethane and the
filtrates evaporated in vacuo. C18 flash chromatography of the
crude (elution solvent: water/acetonitrile 85:15 v/v) gave 4, afford-
ing satisfactory spectral data: 1H NMR (DMSO-d6, 500 MHz): ı 1.71
(s, 3H, –CH3), 3.74 (m, 2H, H-5ꢀ), 3.88 (m, 1H, H-4ꢀ), 4.46 (t, J = 5.8 Hz,
1H, H-3ꢀ), 5.30 (t, J = 5.8 Hz, 1H, H-2ꢀ), 6.45 (d, J = 5.7 Hz, 1H, H-1ꢀ),
7.33 (s, 2H, NH2), 8.15 (s, 1H, H-2), 8.28 (s, 1H, H-8). 13C NMR
(DMSO-d6, 125 MHz): ı 20.43 (–CH3), 62.51 (C-5ꢀ), 71.93 (C-3ꢀ),
77.93 (C-2ꢀ), 81.66 (C-4ꢀ), 83.40 (C-1ꢀ), 118.73 (C-5), 140.07 (C-8),
149.63 (C-2), 153.14 (C-4), 156.41 (C-6), 169.59 (CO).
3. Results and discussion
Based on our previous results of CAL-B-catalysed deacetyla-
tion of peracetylated nucleosides [19–22,25,29], first experiments
of enzymatic deacetylation of 1 and 2 were assayed by applying
alcoholysis and using a high excess of alcohol (alcohol/nucleoside
ratio, A/N). In this way, several alcohols (ethanol, butanol and iso-
propanol) were tested at an A/N = 260 and 1000.
Under the assayed experimental conditions, 1 and 2 did not
afford diacetylated products in good yields because at the stage
of monodeacetylation, aliquots from biotransformations contained
also unreacted and further deacetylated substrate. However, 2ꢀ-
monoacetylated products 3 and 4 were respectively obtained in
excellent yields (Scheme 1; Entries 1 and 2, Table 1). The struc-
ture of each product was determined by NMR and HPLC, comparing
with reference samples of 3 and 4; it can also be mentioned that
by HPLC analysis the different monoacetylated regioisomers can
be clearly differentiated. The CAL-B-catalysed alcoholyses herein
reported afforded the 2ꢀ-monoacetylated products involving an
easier work up of the reaction than by applying PLE [7], since CAL-B
is an immobilised biocatalyst.
CAL-B-catalysed hydrolysis of 1 and 2 at pH 7 and 8 was
then tested and the 2ꢀ,3ꢀ-di-O-acetylated arabinonucleosides 5
and 6 could be obtained in good yields (Scheme 1); Table 1
(Entries 3 and 4) reports the best experimental conditions found
to obtain these products. Compound 6 has been described as a
useful topical prodrug of vidarabine for treatment of herpes virus
infections [23,30]. Its reported three-steps synthesis [23] involved
protection of the 5ꢀ hydroxyl group with a tert-butyldimethylsilyl
derivative, subsequent acetylation and removal of the silyl pro-
tecting group. In this way, the enzymatic procedure herein
reported provides a shorter and more efficient preparation of this
product.
The regioselectivity observed in the monodeacetylation of 1 and
2 agrees with the preference towards 5ꢀ displayed by CAL-B in
acylation and deacetylation of ribo- [10,13,14,19,21,25], 2ꢀ-deoxy
[13,14,22] and arabinonucleosides [8,15–17]. However, and inter-
estingly, the behaviour of 1 and 2 contrasts to the performance
of acetylated ribonucleosides in CAL-B-catalysed deacylations
[19–21,25]: the hydrolysis of the latter gave non-selective mix-
tures of partially acylated products. In the case of herein studied
arabinonucleosides 1 and 2, different products could be regioselec-
2.6. General procedure for the enzymatic hydrolysis of 1 and 2
Experiments of lipase-catalysed hydrolysis were carried out
by adding the assayed hydrolase (300 mg mmol−1 substrate) to a
mixture of the substrate (0.04 mmol) in sodium phosphate buffer
(30 mM, pH 7 and 8; 4 ml). The resulting reaction mixtures were
shaken at 200 rpm at 30 ◦C. Samples were taken at different times
and the enzyme removed by centrifugation, except for CAL-B,
which was separated by decantation. The resulting aliquots were
monitored by TLC and HPLC.
This protocol was applied to prepare diacetylated products 5
and 6, as follows:
2.6.1. 2ꢀ,3ꢀ-Di-O-acetyl-1-ˇ-d-arabinofuranosyluracil (5)
A mixture of 1 (0.20 mmol), sodium phosphate buffer 30 mM
pH 8 (20 ml) and CAL-B (60 mg) were shaken for 12 h at 30 ◦C. The
reaction was then stopped, the buffer evaporated in vacuo and the
resulting crude purified by C18 flash chromatography, employing
water/acetonitrile 95:5 v/v as the elution solvent, to give 5: 1H NMR
(DMSO-d6, 500 MHz): ı 1.95 (s, 3H, –CH3), 2.09 (s, 3H, –CH3), 3.63
(m, 2H, H-5ꢀ), 4.10 (m, 1H, H-4ꢀ), 5.16 (t, J = 4.9 Hz, 1H, H-2ꢀ), 5.34 (dd,
J1 = 4.9 Hz, J2 = 3.3 Hz, 1H, H-3ꢀ), 5.66 (d, J = 8.0 Hz, 1H, H-5), 6.19 (d,
J = 5.0 Hz, 1H, H-1ꢀ), 7.75 (d, J = 8.1 Hz, 1H, H-6). 13C NMR (DMSO-d6,
125 MHz): ı 20.64, 21.01 (–CH3s), 63.63 (C-5ꢀ), 72.65 (C-3ꢀ), 78.67
(C-2ꢀ), 79.97 (C-4ꢀ), 85.96 (C-1ꢀ), 100.84 (C-5), 142.48 (C-6), 151.08
(C-2), 164.03 (C-4), 170.20, 170.71 (COs).
2.6.2. 2ꢀ,3ꢀ-Di-O-acetyl-9-ˇ-d-arabinofuranosyladenine (6)
Analogously, 2 (0.20 mmol), sodium phosphate buffer 30 mM
pH 7 (20 ml) and CAL-B (72 mg) were shaken for 24 h at 30 ◦C. The
Table 1
Enzymatic deacetylation of arabinonucleosides 1 and 2 (Scheme 1).
Entry
Substrate
Product
Yield (%)a
t (h)
Enzyme
Deacetylation mediumb
1
2
3
4
5
6
1
2
1
2
1
2
3
4
5
6
7
8
100
96
72
83
100
100
384
288
12
24
216
168
CAL B
CAL B
CAL B
CAL B
PMK
Isopropanol
Isopropanol
Phosphate buffer pH 8
Phosphate buffer pH 7
Phosphate buffer pH 8
Phosphate buffer pH 8
DMK
a
Determined by HPLC (see Section 2.1).
See Sections 2.5 and 2.6.
b

228
M.B. Sabaini et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 225–229
Chart 1.
tively and satisfactorily obtained in both reaction media (alcohol
or buffer). These results show the influence of nucleoside furanosic
moiety on CAL-B recognition.
Besides, the regioselectivity provided by CAL-B on 2 is com-
plementary to that recently obtained in Candida rugosa lipase
(CRL)-catalysed hydrolysis of 2, which gave the product deacety-
lated at 3ꢀ [18].
Quantitative complete deacylation of nucleosides under mild
reaction conditions is a valuable tool in modified nucleoside
synthesis, since it provides a useful deprotection procedure for
base-labile nucleoside derivatives. With this aim, we have studied
model acylated ribonucleosides [19,20] and applied it to two chlo-
rinated base-labile nucleosides [20] and to nucleoside carbonates
and carbamates [31]. Following with this idea, we studied quantita-
tive full deacetylation of 1 and 2, as models of arabinonucleosides.
At reaction times longer than those reported in Entries 1 and 2
(Table 1), no quantitative full deacetylation of 1 and 2 was observed
in CAL-B-catalysed alcoholyses. Under the hydrolysis conditions,
CAL-B catalysed the complete full deacetylation of 2 after 14 days at
pH 7, but 1 was not quantitatively fully deacetylated at the assayed
pHs.
In summary, in this work we reported CAL-B-catalysed regios-
elective deacetylation of arabinonucleosides 1 and 2 through good
or high yield reactions. The choice of the reaction media allowed
the regioselective formation of products bearing different degree of
acetylation: in isopropanol, CAL-B-catalysed alcoholyses gave the
2ꢀ-O-acetylated derivatives 3 and 4, while hydrolyses afforded the
2ꢀ,3ꢀ-di-O-acetylated products 5 and 6. In particular, the procedure
herein described allows a simple and efficient preparation of the
reported vidarabine prodrug 6, avoiding the utilisation of protec-
tive groups. Moreover, the application of two keratinases allowed
the quantitative complete deacetylation of 1 and 2.
Acknowledgements
We thank UNQ and SECyT (Scientific Cooperation Project
ES/PA05E02 Argentina-Slovenia), as well as the Slovenian Research
Agency, for partial financial support. AMI and LEI are research mem-
bers of CONICET. We are grateful to Novozymes (Brazil) for the
generous gift of CAL-B.
References
Several commercial hydrolases were then tested: lipases (CRL,
PSL, PCL), an esterase (PLE) and the protease subtilisin. Kerati-
nases are a group of mostly extracellular serine-proteases that
hydrolyse keratins, highly resistant scleroproteins, and are pro-
duced by some insects and mainly by microorganisms [32].
Two nonpathogenic fungi, Doratomyces microsporus and Pae-
cilomyces marquandii were selected for the production of the
active keratinolytic serine-proteases DMK [26,27,33,34] and PMK
[26], respectively, which hydrolyse different keratinous-, non-
keratinous proteins [26,27,33,34] and synthetic peptidic substrates
[26,27]. In addition of the set of commercial hydrolases above men-
tioned for the hydrolysis of 1 and 2, we tested then PMK, DMK and
also proteinase K (PK), a commercially available serine-protease
displaying keratinolytic activity.
Assays of 1 and 2 hydrolysis were carried out at pH 8 for PMK
and DMK and at pH 7 and 8 for all the other hydrolases. Except for
PMK and DMK, none of the above listed assayed enzymes allowed
1 quantitative complete deacetylation; PMK (Entry 5, Table 1)
required shorter reaction time than DMK (12 days). Full complete
deacetylation of 2 could be reached by DMK in 7 days (Entry 6);
CRL and PCL required longer reaction times (13 days) and the other
assayed enzymes, including PMK, did not give quantitative deacety-
lation. None of the tested hydrolases catalysed the formation of
regioselectively acetylated products, except for PLE, which gave
the 2ꢀ-monoacetylated derivatives 3 and 4, as previously reported
[7].
[1] E. De Clercq, Antiviral Res. 67 (2005) 56–75.
[2] A. Matsuda, T. Sasaki, Cancer Sci. 95 (2004) 105–111.
[3] E. De Clercq, J. Clin. Virol. 22 (2001) 73–89.
[4] H.J. Field, E. De Clercq, Microbiol. Today 31 (2004) 58–61.
[5] K. Kurosaki, N. Miwa, Y. Yoshida, M. Kurokawa, M. Kurimoto, S. Endo, K. Shiraki,
Antiviral Chem. Chemother. 15 (2004) 281–285.
[6] L.A. Jones, A.R. Moorman, S.D. Chamberlain, P. de Miranda, D.J. Reynolds, C.L.
Burns, T.A. Krenitsky, J. Med. Chem. 35 (1992) 56–63.
[7] S. Manfredini, P.G. Baraldi, R. Bazzanini, F. Bortolotti, S. Vertuani, N. Ashida, H.
Machida, Antiviral Chem. Chemother. 9 (1998) 25–31.
[8] M. Mahmoudian, J. Eaddy, M. Dawson, Biotechnol. Appl. Biochem. 29 (1999)
229–233.
[9] S. Manfredini, P.G. Baraldi, E. Durini, L. Porcu, A. Angusti, S. Vertuani, N. Solaroli,
E. De Clercq, A. Karlsson, J. Balzarini, Bioorg. Med. Chem. Lett. 11 (2001)
1329–1332.
[10] M. Tamarez, B. Morgan, G.S.K. Wong, W. Tong, F. Bennett, R. Lovey,
J.L. McCormick, A. Zaks, Org. Process. Res. Dev.
953.
[11] N. Ciliberti, S. Manfredini, A. Angusti, E. Durini, N. Solaroli, S. Vertuani, L. Buz-
zoni, M.C. Bonache, E. Ben-Shalom, A. Karlsson, A. Saada, J. Balzarini, Bioorg.
Med. Chem. 15 (2007) 3065–3081.
[12] W. Shen, J.S. Kim, P.E. Kish, J. Zhang, S. Mitchell, B.G. Gentry, J.M. Breit-
enbach, J.C. Drach, J. Hilfinger, Bioorg. Med. Chem. Lett. 19 (2009) 792–
796.
[13] M. Ferrero, V. Gotor, Chem. Rev. 100 (2000) 4319–4347.
[14] D. Kadereit, H. Waldmann, Chem. Rev. 101 (2001) 3367–3396.
[15] F. Morís, V. Gotor, Tetrahedron 49 (1993) 10089–10098.
[16] X.F. Li, M.H. Zong, H. Wu, W.Y. Lou, J. Biotechnol. 124 (2006) 552–
560.
[17] X.F. Li, M.H. Zong, R.D. Yang, J. Mol. Catal. B: Enzym. 38 (2006) 48–53.
[18] T. Bavaro, S. Rocchietti, D. Ubiali, M. Filice, M. Terreni, M. Pregnolato, Eur. J. Org.
Chem. (2009) 1967–1975.
[19] L.E. Iglesias, M.A. Zinni, M. Gallo, A.M. Iribarren, Biotechnol. Lett. 22 (2000)
361–365.
[20] D.I. Roncaglia, A.M. Schmidt, L.E. Iglesias, A.M. Iribarren, Biotechnol. Lett. 23
(2001) 1439–1443.
[21] M.A. Zinni, L.E. Iglesias, A.M. Iribarren, Biotechnol. Lett. 24 (2002)
979–983.
7
(2003) 951–
DMK and PMK were also assayed in the hydrolysis of 9 and 11
(Chart 1), the corresponding ribo counterparts of 1 and 2. Although
9 was not quantitatively full deacetylated to 10 by these kerati-
nases, 11 was completely deacetylated to adenosine (12) by PMK, in
much shorter time (48 h) than the arabino substrates. These results
also show the influence of nucleoside structure on these kerati-
nases recognition; as far as we know, DMK and PMK had not been
previously tested with non-peptidic substrates.
[22] M.A. Zinni, S.D. Rodríguez, R.M. Pontiggia, J.M. Montserrat, L.E. Iglesias, A.M.
Iribarren, J. Mol. Catal. B: Enzym. 29 (2004) 129–132.
[23] W.H. Shannon, G. Arnett, D.C. Baker, S.D. Kumar, W.I. Higuchi, Antimicrob.
Agents Chemother. 24 (1983) 706–712.
[24] T.F. Wu, K.K. Ogilvie, J. Org. Chem. 55 (1990) 4717–4724.

M.B. Sabaini et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 225–229
229
[25] M.A. Zinni, L.E. Iglesias, A.M. Iribarren, J. Mol. Catal. B: Enzym. 47 (2007) 86–90.
[30] D.C. Baker, A.K. Hebbler, Nucleic Acid Chem. (1991) 206–210.
[26] H. Gradisˇar, J. Friedrich, I. Krizˇaj, R. Jerala, Appl. Environ. Microbiol. 71 (2005)
[31] M. Capello, M. González, S.D. Rodríguez, L.E. Iglesias, A.M. Iribarren, J. Mol. Catal.
B: Enzym. 36 (2005) 36–39.
3420–3426.
[27] H. Gradisˇar, S. Kern, J. Friedrich, J. Appl. Microbiol. Biotechnol. 53 (2000)
[32] R. Gupta, P. Ramnani, Appl. Microbiol. Biotechnol. 70 (2006) 21–33.
[33] J. Friedrich, S. Kern, J. Mol. Catal. B: Enzym. 21 (2003) 35–37.
[34] J. Friedrich, H. Gradisˇar, M. Vrecl, A. Pogacˇnik, Enzyme Microb. Technol. 36
(2005) 455–460.
196–200.
[28] A. Matzuda, M. Shinozaki, M. Suzuki, K. Watanabe, T. Miyasaka, Synthesis
(1986) 385–386.
[29] M.A. Zinni, E. Aljinovic, L.E. Iglesias, A.M. Iribarren, Quim. Nova 27 (2004)
496–498.