Synthesis and biological evaluation of unprecedented ring-expanded nucleosides (RENs) containing the imidazo[4,5-d][1,2,6]oxadiazepine ring system

Article abstract of DOI:10.1039/c2cc33511e

A small collection of ring-expanded nucleosides (RENs), containing the unprecedented bis-alkylated imidazo[4,5-d][1,2,6]oxadiazepine heterocyclic ring system, has been synthesized through a new general approach. Results of preliminary cytotoxicity tests on breast (MCF-7) and lung (A549) cancer cell lines are also reported.

Full text of DOI:10.1039/c2cc33511e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 9310–9312
www.rsc.org/chemcomm
COMMUNICATION
Synthesis and biological evaluation of unprecedented ring-expanded
nucleosides (RENs) containing the imidazo[4,5-d][1,2,6]oxadiazepine ring
systemw
Stefano D’Errico,a Giorgia Oliviero,*a Jussara Amato,a Nicola Borbone,a Vincenzo Cerullo,b
Akseli Hemminki,c Vincenzo Piccialli,d Sabrina Zaccaria,d Luciano Mayola and
Gennaro Picciallia
Received 15th May 2012, Accepted 30th July 2012
DOI: 10.1039/c2cc33511e
A small collection of ring-expanded nucleosides (RENs), containing
the unprecedented bis-alkylated imidazo[4,5-d][1,2,6]oxadiazepine
heterocyclic ring system, has been synthesized through a new
general approach. Results of preliminary cytotoxicity tests on
breast (MCF-7) and lung (A549) cancer cell lines are also reported.
(ddA) and 20-fluoro-20,30-dideoxyadenosine (FddA). Other RENs,
for example those containing the 6-alkylamino-substituted
imidazo[4,5-e][1,3]diazepine or the imidazo[4,5-e][1,2,4]triazepine
moiety, show direct antiviral activity against human hepatitis
B virus and West Nile Virus.8
Due to their noncanonical steric hindrance, RENs have
been inserted in oligonucleotides to study conformational
and stability changes in duplex, triplex and other DNA
secondary structures, as well as the H-bonding profile, base
stacking, endo/exo sugar puckering and the a/b anomeric
ribose configuration.11 From a synthetic point of view, the
construction of a nucleoside containing a 5 : 7-fused ring
base system can be performed using a sugar unit bearing a
functionalized imidazole (or other 5-membered cycles), on
which the seven-membered ring closure can be accomplished
by inserting an appropriate molecular moiety.9,10,12
Nucleosides containing a 5 : 7-fused ring system as a nucleobase,
also known as ring-expanded nucleosides (RENs), are very
interesting molecules from several points of view. They are mimics
of purine nucleosides and most of them possess significant
biological activities due to their interference with the enzymes
involved in the purine metabolism.1 Coformycin and pentostatin
(20-deoxycoformycin) are two naturally occurring RENs
containing an imidazo[4,5-d][1,3]diazepine ring system. They
show a potent antitumor activity due to the strong inhibition
of adenosine deaminase (ADA), acting as transition-state
analogues.2,3 Pentostatin is a drug approved in 1991 by
FDA for the treatment of hairy cell leukemia. However, it is
characterized by a severe toxicity that limits its use in therapy.1
A number of coformycin analogues have been synthesized, both to
investigate the structure–activity relationship and to produce less
toxic derivatives.4–9 Furthermore, imidazo[4,5-c]azepine nucleo-
sides, ring-expanded derivatives of xanthosine, guanosine and
inosine, have been investigated and their cytotoxic activity has
been reported.10 RENs have shown antiviral properties as well.
Coformycin and some of its derivatives are able to enhance the
activity of therapeutic antiviral drugs such as 2,3-dideoxyadenosine
Alternatively, a pre-synthesized 5 : 7-fused heterocyclic
system can be joined to a sugar moiety by reaction with an
activated glycosyl donor.7,13,14 The latter approach is longer
but it does not suffer from problems due to the weakness of the
N-glycosidic bond and it also offers the chance to synthesize
both a/b nucleoside isomers.
In this communication we report the synthesis of new
ring-expanded nucleosides 7a–c (Scheme 1) containing the
unprecedented imidazo[4,5-d][1,2,6]oxadiazepine heterocyclic
system as the base moiety through a new approach featuring
the opening/reclosure of the pyrimidine ring system of 20,30,50-
tri-O-(tert-butyldimethylsilyl) nebularine N1-oxide 1. Our
synthetic strategy allowed us to prepare a small set of 5,8-dialkyl-
(aryl)-substituted imidazo[4,5-d][1,2,6]oxadiazepine nucleosides
(7a–c), in high yields, exploiting the previously reported reactivity
of 1 towards Grignard reagents.15a
a Dipartimento di Chimica delle Sostanze Naturali,
`
Universita degli Studi di Napoli Federico II, Via D. Montesano 49,
80131, Napoli, Italy. E-mail: golivier@unina.it
b Laboratory of Immunovirotherapy, Division of Biopharmaceutics and
Pharmacokinetics, Faculty of Pharmacy, University of Helsinki,
Helsinki, Finland
The reactivity of a-carbons towards Grignard reagents in
aromatic nitrones has previously been observed for pyridine
N-oxide16–18 which results in cleavage of the N–C a bond with
the concomitant formation of oxime and trans-alkene functions. In
our previous studies15a we observed that reaction of nitrone 1 with
a Grignard reagent leads to addition at C6 furnishing the adducts
2, as a mixture of C6 diastereomers, which slowly decomposed on
standing, partially dehydrating to C6-alkyl(aryl)nebularines 3.
c Cancer Gene Therapy Group, University of Helsinki & Helsinki
University Central Hospital, Haartmaninkatu 8, Helsinki, Finland
d Dipartimento di Scienze Chimiche, Universita` degli Studi di Napoli
Federico II, Via Cintia 21, 80126, Napoli, Italy
w Electronic supplementary information (ESI) available: Experimental
procedures, compound characterization data, copies of 1H- and
13C-NMR spectra of compounds 4, 6 and 7, copies of UV spectra of
compounds 4 and 6, biological assays of compounds 7. See DOI:
10.1039/c2cc33511e
c
9310 Chem. Commun., 2012, 48, 9310–9312
This journal is The Royal Society of Chemistry 2012

View Article Online
Scheme 2 Proposed mechanism for the formation of compounds 5a–c.
Scheme 1 Reagents and conditions: (i) R0MgBr (2 equiv.), THF, 2 h,
rt; (ii) Ac2O–pyridine (6 : 4), 1 h, 50 1C; (iii) MeReO3–H2O2 (30%
aq.), MeOH, 15 h, rt; (iv) pyridine, 16 h, 70 1C; (v) R00MgBr (4 equiv.),
t
THF, 2 h, rt; (vi) BuOOH (10 equiv.), CCl4, reflux, 1 h; (vii) NH4F
(10 equiv.), MeOH, 5 h, reflux. *Reaction yields (%) from 1.
This process was pushed towards aromatization by treatment
of the crude adduct 2 with Ac2O in pyridine at 50 1C, to give 3
in nearly quantitative yields.15
Scheme 3 Proposed mechanism for the formation of compounds 6a–c.
We reasoned that the introduction of a N1-nitrone function
on 3 could induce an electrophilic behaviour of C2 as well, thus
allowing us to obtain, at least in principle, 2,6-dialkylpurine. To
probe this hypothesis the representative compound 3a was treated
with catalytic amounts of methyltrioxorhenium (MeReO3) in the
presence of H2O2, the same oxidizing mixture used to obtain
nitrone 1 from silylated nebularine. Under these conditions, 3a
produced only traces of N1-oxide 4a. Alternatively, the oxidation
of the N1-hydroxy derivative 2a with air in pyridine at 70 1C
afforded the C6-ethylnebularine N1-oxide 4a in 78% yield.
The similar procedure using DMF as a solvent was reported
for 2,6-alkyl(aryl)-substituted N-hydroxypyridine derivatives.18
The successive reaction of 4a with Grignard reagents confirmed
the electrophilic reactivity of C2 but, in this case, the addition was
followed by the breaking of the N1–C2 bond leading to
compound 5a (60% yield) as a 1:1 mixture of (E/Z) isomers at
the imine double bond (Scheme 2). In particular, 2D NMR
experiments and HRMS data confirmed the structure of 5a.
The HMBC experiment supported the presence of the aldimine
moiety showing clear correlations between the aldimine proton
and both the aldimine carbon and the Grignard residue R00.
Furthermore, the absence of correlation between aldimine
proton and C6 of the purine ring strongly suggested the opening
of the cycle at the N1–C2 bond. Next, we tested the reactivity of
the open system of 5a for the possible ring reclosure. In
particular, treatment of 5a with tBuOOH (10 equiv.) in refluxing
CCl4 gave nucleoside 6a, containing the imidazo[4,5-d][1,2,6]-
oxadiazepine framework (Scheme 3) in a 92% yield.
It is reported that nucleophiles usually open the oxaziridine
ring by attack at the nitrogen or oxygen20 atom rather than on
the carbon atom. However, it is known that oxaziridines
containing aromatic substituents could easily rearrange to
nitrones upon heating,21 making the carbon the preferred
attack site. Accordingly, nitrone 9a, derived from 8a, would
undergo a fast ring reclosure by nucleophilic attack of the
oxime hydroxyl functionz,22 on the nitrone carbon,23 to give
intermediate 10a (not isolable) which would lead to the
oxadiazepine cycle 6a by water loss.
Previous mechanistic studies on the reactivity of the 2-halo-
quinazoline 3-oxide system with alkali24 showed that the opening
of the pyrimidine ring by nucleophilic attack of the hydroxyl ion
at C2 afforded an anti-oxime. In our case, 2D-NMR studies of
compound 5a performed in pyridine-d5 did not give conclusive
information on the oxime configuration, although the (Z)
configuration is needed for the ring closure (9a - 10a).
The same reaction sequence leading to 6a furnished compounds
6b–c in comparable yields to 6a (Scheme 1). Compounds 5a–c are
rather unstable and decompose on standing within a few hours.
Therefore, reactions of 5b and 5c were performed in the crude
reaction mixture without observing lowering of the yields.
Finally, nucleosides 6a–c were deprotected at ribose using
NH4F in MeOH25 to give, after HPLC purification, compounds
7a–c (94–97% yield, Scheme 1).
The structure of nucleosides 6a–c and 7a–c was confirmed
by 2D-NMR experiments and HRMS data. Moreover, the
disappearance of the characteristic purine band around 260 nm
in the UV spectra of 6a–b (see ESIw) provided further evidence
for the 7-membered ring formation, excluding the presence of
the isomeric 2,6-disubstituted purine N1-oxide system.
The accomplishment of the seven-membered ring could be
explained by invoking the formation of the oxaziridine 8a
(Scheme 3) through oxidation of the aldimine functionality.19
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 9310–9312 9311

View Article Online
Considering that RENs significantly inhibit the growth of
several cancer cell lines,1 we tested the cytotoxicity of the novel
RENs by treating lung A549 and breast MCF-7 cancer cell
lines with different concentrations (0.1, 1.0, 10 and 100 mM) of
7a–c. At different time points for each concentration the cell
viability was determined by measuring the mitochondrial activity.26
Cisplatin was used as the control since its activity in these cell lines
has been extensively studied.27,28 We found that at day 7 all tested
compounds significantly reduced the cell viability of breast cancer
derived MCF-7 cells at the highest tested concentration (also at
10 mM in the case of 7a). An explanation for such a delayed activity
could be the slow formation of an active metabolite during the
assay. At the same time point the inhibition of A549 cells growth,
albeit observed, was only marginal (Fig. S1, MCF-7 cell line and
Fig. S2, A549 cell line in ESIw).
In conclusion, we have proposed new reactions of the purine
nucleobase enlarging the synthetic toolbox to obtain novel
nucleoside analogues. The reactivity of the herein proposed
C6-alkyl(aryl)purine-N1-oxides towards Grignard reagents
afforded the new 4,5-disubstituted imidazo-nucleosides 5 from
which the unprecedented imidazo[4,5-d][1,2,6]oxadiazepine
nucleosides were built in good yields.
R. K. Robins and L. B. Townsend, Academic Press, New York,
1978, p. 159.
3 P. C. Tyler, E. A. Taylor, R. F. G. Frohlich and V. L. Schramm,
¨
J. Am. Chem. Soc., 2007, 129, 6872.
4 G. A. LePage, L. S. Worth and A. P. C. R. Kimball, Cancer Res.,
1976, 36, 1481.
5 H. D. H. Showalter, S. R. Putt, P. E. Borondy and J. L. Shillis,
J. Med. Chem., 1983, 26, 1478.
6 R. S. Hosmane and M. Hong, Biochem. Biophys. Res. Commun.,
1997, 236, 88.
7 M. Hong and R. S. Hosmane, Nucleosides Nucleotides, 1997,
16, 1053.
8 V. S. R. K. Yedavalli, N. Zhang, H. Cai, P. Zhang, M. F. Starost,
R. S. Hosmane and K -T Jeang, J. Med. Chem., 2008, 51, 5043.
9 P. Zhang, N. Zhang, V. E. Buckwold and R. S. Hosmane, Bioorg.
Med. Chem., 2007, 15, 4933.
10 N. Minakawa, T. Sasaki and A. Matsuda, Tetrahedron, 1998,
54, 13517.
11 R. S. Hosmane, Prog. Heterocycl. Chem., 2009, 21, 35.
12 E. Chan, S. R. Putt and H. D. H. Showalter, J. Org. Chem., 1982,
47, 3457.
13 R. S. Hosmane, A. Bhan and R. L. Karpel, J. Org. Chem, 1990,
55, 5882.
14 L. Wang, A. Bhan and R. S. Hosmane, Nucleosides Nucleotides,
1994, 13, 2307.
15 (a) S. D’ Errico, V. Piccialli, G. Oliviero, N. Borbone and J. Amato
V. D’ Atri and G. Piccialli, Tetrahedron, 2011, 67, 6138; (b) A. E. A.
Hassan, R. A. I. Abou-Elkhair, J. N. Riordan, P. W. Allan,
W. B. Parker, R. Khare, W. R. Waud, J. A. Montgomery and
J. A. Secrist III, Eur. J. Med. Chem., 2012, 47, 167. Compound 3a
was desilylated at the ribose moiety as for 6a–c and the spectro-
scopic data of the obtained product were in agreement with those
published in ref. 15b.
16 T. J. van Bergen and R. M. Kellogg, J. Org. Chem., 1971, 36, 1705.
17 H. Andersson, X. Wang, M. Bjorklund, R. Olsson and
F. Almqvist, Tetrahedron Lett., 2007, 48, 6941.
18 H. Andersson, F. Almqvist and R. Olsson, Org. Lett., 2007,
9, 1335.
Moreover, it is to be noted that this heterocyclic system has
not been so far reported and the proposed synthetic strategy
allows the introduction of C-substituents on C5 and C8 of the
oxadiazepine cycle. Finally, we have reported that new RENs
are able to inhibit the cell growth of breast cancer derived
MCF-7 cells. The observed cytotoxic effects of 7a–c were dose-
dependent and comparable to those of cisplatin.
Further studies to enlarge the collection of imidazo[4,5-d]-
[1,2,6]oxadiazepine nucleosides, to investigate the reactivity of
nebularine N1-oxide toward other C-nucleophiles, including
organo-lithium reagents, as well as of the open intermediates
5a–c will be carried out.
19 F. A. Davis and A. C. Sheppard, Tetrahedron, 1989, 45, 5703.
20 J. Vidal, S. Damestoy, L. Guy, J. C. Hannachi, A. Aubry and
A. Collet, Chem.–Eur. J., 1997, 3, 1691.
21 J. Hamer and A. Macaluso, Chem. Rev., 1964, 4, 473.
22 S. Ostrowski, Molecules, 1999, 4, 287.
23 A. Pohjakallio and P. M. Pihko, Chem.–Eur. J., 2009, 15, 3960.
24 A. Stempel, E. Reeder and L. H. Sternbach, J. Org. Chem., 1965,
30, 4267.
Notes and references
25 W. Zhang and M. J. Robins, Tetrahedron Lett., 1992, 33, 1177.
26 A. H. Cory, T. C. Owen, J. A. Barltrop and J. G. Cory, Cancer
Commun., 1991, 3, 207.
z The nitrogen of the oxime requires harsh conditions to react, as
outlined in ref. 22.
1 R. S. Hosmane, Curr. Top. Med. Chem., 2002, 2, 1093.
2 R. P. Argawal, S. Cha, G. W. Crabtree and R. E. Parks, Chemistry
and Biology of Nucleosides and Nucleotides, ed. R. E. Harmon,
27 C. W. Yde and O. G. Issinger, Int. J. Oncol., 2006, 29, 1397.
28 G. Telford, J. Bradford, W. Godfrey, R. W. Robey and S. E. Bates,
Stem Cells, 2007, 24, 1029.
c
9312 Chem. Commun., 2012, 48, 9310–9312
This journal is The Royal Society of Chemistry 2012