A one-pot synthesis of 1-α- and 1-β-d-arabinofuranosyl-2- nitroimidazoles: Synthons to the markers of tumor hypoxia

Article abstract of DOI:10.1081/NCN-200055700

1-α- and 1-β-D-Arabinofuranosyl-2-nitroimidazole (α-AZA and β-AZ A) are synthons for a number of potential markers of tissue hypoxia. A one pot synthesis in which 2-nitroimidazole is coupled with a mixture of α-and β-1-O-acetyl-2,3,5-tri-O-benzoyl-D-arabi

Full text of DOI:10.1081/NCN-200055700

Nucleosides, Nucleotides, and Nucleic Acids, 24 (3):173–178, (2005)
Copyright D 2005 Taylor & Francis, Inc.
ISSN: 1525-7770 print/ 1532-2335 online
DOI: 10.1081/NCN-200055700
A ONE-POT SYNTHESIS OF 1-a- AND 1-b-D-ARABINOFURANOSYL-
2-NITROIMIDAZOLES: SYNTHONS TO THE MARKERS OF
TUMOR HYPOXIA
5
Ebrahim Naimi and Piyush Kumar
Department of Oncologic Imaging, Cross Cancer
Institute and Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta,
Edmonton, Canada
5
Alexander J. B. McEwan
Department of Oncologic Imaging, Cross Cancer Institute,
University of Alberta, Edmonton, Canada
5
Leonard I. Wiebe
Department of Oncologic Imaging, Cross Cancer Institute and Faculty
of Pharmacy and Pharmaceutical Sciences, University of Alberta,
Edmonton, Canada
5
1-a- and 1-b-D-Arabinofuranosyl-2-nitroimidazole (a-AZA and b-AZ A) are synthons for a
number of potential markers of tissue hypoxia. A one pot synthesis in which 2-nitroimidazole is coupled
with a mixture of a-and b-1-O-acetyl-2,3,5-tri-O-benzoyl-D-arabinofuranose in the presence of stannic
chloride, followed by deprotection using ammonia/methanol, is described. Previously reported conditions
for coupling 2-nitroimidazole to 1-a-bromoarabinofuranose protected by base-hydrolyzable groups
afforded a-AZA almost exclusively.
Keywords a-AZA, b-AZA, One Pot Synthesis, Hypoxia
INTRODUCTION
Many human solid tumors are less well oxygenated than normal tissues. This
leads to resistance to radiotherapy and anticancer chemotherapy, as well as a
predisposition to increased tumor metastases.[1] An important class of compounds
studied as radiosensitizers are substituted 2-nitroimidazoles and, of these, a num-
ber of iodine-labeled 2-nitroimidazole nucleosides coupled to a radiolabeled sugar
group have been shown to undergo reductive trapping in cells with low oxygen
pressures.[2] Both a-and b-1-D-(5-deoxy-5-iodoarabinofuranosyl)-2-nitroimidazole
Received 1 September 2004, accepted 31 January 2005.
Address correspondence to Leonard I. Wiebe, Department of Oncologic Imaging, Cross Cancer Institute,
University of Alberta, Edmonton T6G 2N8, Canada; Fax: (780) 435-0636; E-mail: leonard.wiebe@ualberta.ca
173
Order reprints of this article at www.copyright.rightslink.com

174
E. Naimi et al.
(a-IAZA; b-IAZA) have been developed as diagnostic radiopharmaceuticals to
detect and monitor regional hypoxia in disease.[3–6]
The present study reports the synthesis of b-AZA, the precursor for b-IAZA,
by simple coupling of azomycin (2-nitroimidazole) with arabinofuranose.
Initial attempts to synthesize b-azomycin arabinofuranoside (b-AZA) by coupling
1-a-bromo-2,3,5-tri-O-benzoyl arabinofuranose with 1-trimethylsilyl-2-nitroimidazole
resulted in the exclusive formation of a-coupled product,[3,7] as expected according
to the trans rule, whereby the benzoyl carbonyl group in the 2’-ara configuration
facilitates this mechanism.[8]
Intramolecular involvement of the b-plane electrons with electropositive C-1
creates steric hindrance to nucleophilic (2-nitroimidazole) approach from the b-face
and results in the exclusive formation of a-azomycin arabinofuranoside (a-AZA).
Efforts to avoid this retention of configuration have included replacing the benzoyl
protective groups with benzyl-or substituted silyl-(non-carbonylated) groups.[9] An
alternative approach to synthesize b-AZA via inversion of configuration at C-2’ of
the corresponding b-azomycin riboside has been reported.[6]
To date there have been no reports of a one pot synthesis to provide useful
yields of both a-and b-arabinofuranosyl azomycin nucleosides in reasonable yields.
We now report a procedure where both a-and b-anomers are synthesized in good
yield in a single reaction.
DISCUSSION
Interaction of the protective group at C-2’ hydroxyl plays a significant role in
determining the final conformation of the coupled base. The commonly used base-
labile (benzoyl and acetyl) protective groups interact electronically with the C-1’
halogen during coupling at C-1’, resulting in exclusive formation of a-AZA,[7] even
when azomycin is coupled with the trans (a-)arabinofuranosyl bromide. The
introduction of alternate, non-carbonyl protecting groups such as benzyl-or
substituted silyl-moieties either proceeded sluggishly or these groups were cleaved
under the acidic reaction conditions that develop during displacement of the C-1’
halogen. It was observed that nucleosidic bond in azomycin nucleosides cleaves
very rapidly under acidic conditions (pH = 4 and below) to generate free sugar and
2-nitroimidazole. The use of HBr gas for bromination generated strong acidic
conditions that caused deprotection when tert-butyldiphenylsilyl or benzyl groups
were used as protective functions.[10] The fact that catalytic de-benzylation is also
capable of reducing the nitro substituent on the imidazole ring further limits the
effectiveness of protection by benzylation. p-Methoxybenzyl protection of the
arabinofuranose hydroxyl groups, which can theoretically be readily removed by
DDQ oxidation[11] was also not effective because chlorination or bromination at C-
1’ to form the respective arabinose halide led to de-benzylation prior to coupling.
a-AZA (6) has previously been obtained by coupling either 1-a-or 1-b-
bromo-2,3,5-tri-O-benzoylarabinofuranose and 1-trimethysilyl-2-nitroimidazole in the

One-Pot Synthesis of 1-a- and 1-b-D-Arabinofuranosyl-2-Nitroimidazoles
175
presence of Hg(CN)2.[3,7,12] b-AZA (5) was prepared via an unconventional route
starting from 1-b-D-(ribofuranosyl)-2-nitroimidazole (AZR), with a change of con-
figuration at the C-2’-position.[6] Trace amounts of b-AZA were reported as a side-
product in the a-AZA synthesis from 1-a-bromo-2,3,5-tri-O-benzoylarabinofuranose,
but no chemical characterization was provided.[12]
In the present study, both a-AZA and b-AZA were obtained by reacting
2-nitroimidazole with a mixture of a-and b-1-O-acetyl-2,3,5-tri-O-benzoyl-D-arabino-
furanose (2) in the presence of stannic chloride, followed by deprotection in the
presence of ammonia/methanol (Scheme 1). The acetoxy group located at the
anomeric C-1’ does not create as strong a nucleophilic center at this carbon as
would bromine, and therefore the inductive effect of the benzoyl carbonyl function
at C-2’ is not strong enough to participate in the orientation of the incoming
nucleophile (nitroimidazole). Consequently, when the coupling is performed
SCHEME 1 Reagents and conditions: (i) AcCl, MeOH, 0°C!25°C, 3 h; (ii) pyridine, C6H5COCl, 0°C!25°C,
overnight; (iii) AcOH, Ac2O, H2SO4, 0°C!25°C, 3 h; (iv) CH3CN, Hg(CN)2, SnCl4, 60°C, 1.5 h; (v) NH3/MeOH,
0°C (10 h), 25°C (6 h).

176
E. Naimi et al.
starting from the anomeric mixture of 1-O-acetyl substituted precursors 2 (the
1
2a:2b isomeric ratio is 4:1 as observed by their H and 13C NMR spectra), the
process affords both a-and b-azomycin coupled nucleosides in satisfactory yields.
In theory, SN1 nucleophilc substitutions occur with the inversion of
configuration and, therefore, the expected isomeric ratio of coupled product 4
according to this concept would be 4:1 for 4b:4a, respectively. Nonetheless, the
coupling reaction yielded an isomeric preference for a-product 4 (b:a = 2:1). This
implies that the trans rule, although suppressed due to the weaker electronic effect
of the acetyl C-1’ carbonyl, still applies under these conditions. The precise function
of stannic chloride remains obscure but the formation of 4b can be loosely based
upon a proposal by Lemieux and Morgan[13] that involves reduced formation of
1,2-benzoxonium ion due to trans location of 1-O-acetyl and 2-O-benzoyl groups in
2a in comparison to 2b. Therefore, approach of a possible tautomer of the
intermediate imidazole nitronate ester[14] at 1’-b-position progresses, albeit not in the
same isomeric proportions. It is also reported that strong Lewis acids such as stannic
chloride may play significant role in partial anomerization leading to the mixture of
the isomers.[15]
EXPERIMENTAL
a- and b-1-O-Acetyl-2,3,5-tri-O-benzoyl-D-arabinofuranosides
(2). D-arabinose (16 g, 32.8 mmol) upon treatment with acetyl chloride followed
by methanol was converted to 1-O-methyl arabinofuranose and benzoylated to
afford 1-a-O-methyl-2,3,5-tri-O-benzoyl arabinofuranose 1 in satisfactory yield
(ꢀ52%). Upon acetylation using a mixture of glacial acetic acid (169 mL) and
acetic anhydride (34 mL), and conc. H2SO4 (10 mL) as a Lewis acid, at 0°C, 1 was
converted to a mixture of a-and b-1-O-acetyl-2,3,5-tri-O-benzoyl-D-arabinofuranoses,
2. After completion of the reaction (3 h at 22°C) the mixture was poured into ice
water (300 mL) and extracted with dichloromethane (2 ꢁ 300 mL). The extracts
were combined, washed successively with saturated aqueous sodium bicarbonate
(2 ꢁ 100 mL) and water (2 ꢁ 100 mL), dried (anhydrous sodium sulfate), filtered
and evaporated under reduced pressure to give an isomeric mixture of 2 as a white
foam (16.9 g, 99.76%). The isomeric ratio for 2-a:2-b at this stage was found to be
1
79:21. The H and 13C NMR spectra of the 2-a-and 2-b-anomers (1-O-acetyl-2,3,5-
tri-O-benzoyl-D-arabinofuranoses) confirmed their chemical identities.[7,12,16]
1-b- and 1-a-D-(2,3,5-Tri-O-benzoylarabinofuranosyl)-2-nitro-
imidazole (3 and 4, respectively). SnCl4 (0.78 mL, 6.21 mmol) was added
to a solution of a-and b-1-O-acetyl-2,3,5-tri-O-benzoyl-D-arabinofuranosides, 2, in
dry acetonitrile (25 mL) in the presence of 2-nitroimidazole (0.42 g, 3.72 mmol) and
mercuric cyanide (1.55 g, 6.21 mmol) under argon. After stirring at 60°C for 90 min,
the dark reaction solution was cooled and the solvent was removed. The impure
mixture was redissolved in dichloromethane (300 mL), filtered, and the filtrate was

One-Pot Synthesis of 1-a- and 1-b-D-Arabinofuranosyl-2-Nitroimidazoles
177
washed successively with solutions of saturated aqueous sodium bicarbonate
(2 ꢁ 100 mL), 40% KI (100 mL), and H2O (2 ꢁ 100 mL). The organic phase was
dried (anhydrous sodium sulfate) and filtered. Evaporation of the solvent followed
by flash column chromatography on silica gel using benzene–ethyl acetate (10:1, v/
v) as eluent afforded 3 (0.74 g, 43%) and 4 (0.35 g, 20%) as white foams, which were
1
crystallized from hexanes-ethyl acetate (2:1, v/v, 20 mL). 3: mp = 123–124°C, H
NMR (CDCl3) d 7.08–8.13 (m, 17 H, H-4, 5, and Ar-H), 7.06 (d, J = 4.27 Hz, 1 H,
H-1’), 6.13 (d, J = 4.27 Hz, 1 H, H-2’), 5.67 (s, 1 H, H-3’), 4.98 (dd, J = 11.9, 6.41 Hz,
1 H, H-5’a), 4.86 (dd, J = 11.9, 3.66 Hz, 1 H, H-5’b), 4.69 (m, 1 H, H-4’). 13C NMR
(CDCl3) d 166.13, 165.23, 164.09, 144.10, 154.76, 134.5, 133.88, 133.41, 129.96,
129.72, 129.60, 129.29, 128.70, 128.61, 128.49, 128.40, 128.27, 127.65, 123.37, 88.42,
82.03, 75.39, 63.06. Anal. calcd. for C29H23N3O9: C, 62.48; H, 4.16; N, 7.54. Found:
C, 61.94; H, 4.04; N, 7.31. The melting point, 1H and 13C NMR spectra for 4 were
identical to those reported previously.[12]
1-b- and a-D-Arabinofuranosyl-2-nitroimidazole (5 and 6,
respectively). To 3 (0.04 g, 0.72 mmol) or 4 (0.03 g, 0.54 mmol) was added a
solution of NH3 in CH3OH (4 mL of 2M solution) at 0°C under argon and the
resulting solution was stirred 10 h at 0°C and 6 h at room temperature. The solvent
was then removed under vacuum (rotovap). Purification of the residue via flash
column chromatography on silica gel using 15% CH3OH in CH2Cl2 gave 5 (0.16 g,
91%) or 6 (0.12 g, 94%) as white solids, which were recrystallized from ethyl acetate-
hexane (1:1, v/v, 5 mL). The melting points, 1H and 13C NMR spectra for 5 and 6
were identical to those reported previously.[6,7]
ABBREVIATIONS
AZA
IAZA
DDQ
AZR
1-D-arabinofuranosyl-2-nitroimidazole
1-D-(5-deoxy-5-iodoarabinofuranosyl)-2-nitroimidazole
2,3-Dichloro-5,6-dicyano-1,4-benzoquinone
1-b-D-(ribofuranosyl)-2-nitroimidazole
Hg(CN)2 mercuric cyanide
ACKNOWLEDGMENTS
This work was supported in part by a grant (RI-14) from the Alberta Cancer
Board. We gratefully acknowledge the gift of azomycin (2-nitroimidazole) from
Professor Koh-Ichi Seki, Hokkaido University, Japan.
REFERENCES
1. Brown, J.M. Exploiting the hypoxic cancer cell: mechanisms and therapeutic strategies. Mol. Med. Today
2000, 6, 157–162.

178
E. Naimi et al.
2. Sorger, D.; Patt, M.; Kumar, P.; Wiebe, L.I.; Barthel, H.; Seese, A.; Dannenberg, C.; Tannapfel, A.; Kluge, R.;
Sabri, O. [18F]Fluoroazomycinarabinofuranoside (18FAZA) and [18F]Fluoromisonidazole (18FMISO): a
comparative study of their selective uptake in hypoxic cells and PET imaging in experimental rat tumors.
Nucl. Med. Biol. 2003, 30, 317–326.
3. Mannan, R.H.; Somayaji, V.V.; Lee, J.; Mercer, J.R.; Chapman, J.D.; Wiebe, L.I. Radioiodinated 1-(5-Iodo-5-
deoxy-b-D-arabinofuranosyl)-2-nitroimidazole (Iodoazomycin Arabinoside:IAZA): a novel marker of tissue
hypoxia. J. Nucl. Med. 1991, 32, 1764–1770.
4. Urtasun, R.C.; Parliament, M.B.; McEwan, A.J.; Mercer, J.R.; Mannan, R.H.; Wiebe, L.I.; Morin, C.;
Chapman, J.D. Measurement of hypoxia in human tumors by non-invasive SPECT imaging of azomycin
arabinoside. Br. J. Cancer., Suppl. 1996, 27, S209–S212.
5. Vinjamuri, S.; O’Driscol, K.; Maltby, P.; McEwan, A.J.; Wiebe, L.I.; Critchley, M. Identification of hypoxic
regions in traumatic brain injury. Clin. Nucl. Med. 1999, 24, 891–892.
6. Kumar, P.; Ohkura, K.; Beiki, D.; Wiebe, L.I.; Seki, K. Synthesis of 1-b-D-(5-deoxy-5-iodoarabinofuranosyl)-
2-nitoimidazole (b-IAZA): A novel marker of tissue hypoxia. Chem. Pharm. Bull. (Tokyo) 2003, 51, 399–
403.
7. Kumar, P.; Wiebe, L.I.; Atrazheva, E.; Tandon, M. An improved synthesis of a-AZA, a-AZP and a-AZG, the
precursors to clinical markers of tissue hypoxia. Tetrahedron Lett. 2001, 42, 2077–2078.
8. Baker, R.B.; Joseph, J.P.; Schraub, R.E. Puromycin. Synthetic studies. XIV. Use of the N-phthalyl blocking
group for synthesis of aminonucleosides. J. Am. Chem. Soc. 1955, 77, 5905–5910.
9. Glaudemans, C.P.J.; Fletcher, H.G., Jr. Syntheses with partially benzylated sugars. III. 1A simple pathway to a
‘‘cis-nucleoside,’’ 9-b-D-arabinofuranosyladenine (Spongoadenosine). J. Org. Chem. 1963, 28, 3004–3006.
10. Kumar, P.; Wiebe, L.I. Unpublished.
11. Nakajima, N.; Horita, K.; Abe, R.; Yonemitsu, O. MPM (4-methoxybenzyl) protection of hydroxyl functions
under mild acidic conditions. Tetrahedron Lett. 1988, 29, 4139–4142.
12. Schneider, R.F.; Engelhardt, E.L.; Stobbe, C.X.C.; Fenning, M.C.; Chapman, J.D. The synthesis and
radiolabeling of novel markers of tissue hypoxia of the iodinated azomycin nucleoside class. J. Labelled
Compd Radiopharm. 1997, 39, 541–557.
13. Lemieux, R.U.; Morgan, A.R. Mechanisms for reactions of anomeric tetra-O-acetyl-D-glucopyranosyl
bromides with pyridine and 2-pyridones. Can. J. Chem. 1965, 43, 2214–2218.
14. Prisbe, E.J.; Verheyden, J.P.H.; Moffat, J.G. 5-Aza-7-deazapurine Nucleosides. 2. 1Synthesis of some 8-(D-
ribofuranosyl)imidazo[1,2-a]-1,3,5-triazine derivatives. J. Org. Chem. 1978, 43(25), 4784–4794.
15. Baker, B.R.; Schaub, R.E.; Kissman, H.M. Puromycin. Synthetic Studies XV. 3’-Amino-3’-deoxyadenosine.
J. Am. Chem. Soc. 1955, 77, 5911–5915.
16. Fletcher, H.G. In Methods in Carbohydrate Chemistry. Whistler, R.L., Wolfrom, M.L., Eds.; Academia Press:
New York, 1992, 228.