A fresh AIR synthesis

Article abstract of DOI:10.1055/s-1999-3500

Two short synthetic routes to the biosynthetic purine precursor 5- aminoimidazole ribonucleotide (AIR) based on catalytic reduction of 5- nitroimidazoles are described. The synthetic ribonucleoside AIRs, which is also available via one of these routes, is a convenient intermediate for the synthesis of nucleoside analogues and preparations of the novel 'pyridine- stretched' analogues of adenosine and inosine are described.

Full text of DOI:10.1055/s-1999-3500

PAPER
985
A Fresh AIR Synthesis
Mark J. Humphries, Christopher A. Ramsden*
School of Chemistry and Physics, Keele University, Keele, Staffordshire ST5 5BG, UK
Fax +44(1782)712378; E-mail cha33@cc.keele.ac.uk
Received 5 November 1998; revised 1 January1999
which being shorter may have advantages. The relatively
easy access to AIRs provided by our route has enabled us
to begin to explore its chemistry. In particular we describe
C–C bond forming reactions of AIRs which are analogous
Abstract: Two short synthetic routes to the biosynthetic purine pre-
cursor 5-aminoimidazole ribonucleotide (AIR) based on catalytic
reduction of 5-nitroimidazoles are described. The synthetic ribonu-
cleoside AIRs, which is also available via one of these routes, is a
convenient intermediate for the synthesis of nucleoside analogues to that utilised in purine biosynthesis and which provide a
and preparations of the novel "pyridine-stretched" analogues of ad-
enosine and inosine are described.
biomimetic route to novel purine analogues.
Key words: nucleosides, aminoimidazoles, 1-deazapurines, AIR,
biomimetic
An important C–C bond forming step in the biosynthesis
of purine ribonucleotides involves the enzyme mediated
reaction between 5-aminoimidazole ribonucleotide (AIR)
1 and carbon dioxide to give 4-carboxy-5-aminoimid-
azole ribonucleotide (CAIR) 5.1 Recent studies by Davis-
son and co-workers have demonstrated that in E. coli two
carboxylase enzymes are associated with this process.2
AIR 1 and bicarbonate are initially transformed into N5-
carboxyaminoimidazole ribonucleotide (N5-CAIR) 3
which is then converted to the isomer CAIR 5 by a second
enzyme. With this important C–C bond in place, the six-
membered ring of the purine precursor inosine monophos-
phate (IMP) 7 is then constructed in four steps. The 5-ami-
noimidazole AIR 1 is also a precursor of thiamine
(vitamin B1) in prokaryotic organisms, including E. coli.3
Labelling studies have shown that the pyrimidine ring and
its substituents in the biosynthetic intermediate 4 are de-
rived from both the aminoimidazole and the ribofuranosyl
fragments of AIR.4 The mechanistic details of this inter-
esting intramolecular reorganisation (1 → 4) have not yet
been established. The intermediate 4 is subsequently cou-
pled with hydroxyethylthiazole phosphate leading to thia-
mine pyrophosphate 6.
Because of the important role of AIR in these fundamen-
tally important biosynthetic pathways, convenient syn-
thetic routes to AIR are desirable in order to provide
material for enzyme studies and to investigate its chemis-
try, including biomimetic reactions. In 1990 Leonard and
co-workers described a synthesis of AIR 1 in seven steps
from D-ribose.5 This approach involves formation of the
ribonucleoside AIRs 2 by decarboxylation of the corre-
sponding 5-aminoimidazole-4-carboxylic acid and subse-
quent phosphorylation to give AIR 1. We now describe an
alternative and shorter synthesis of AIRs which introduc-
es the 5-aminoimidazole moiety by reduction of the corre-
sponding 5-nitroimidazole.6 Since conversion of AIRs to
AIR (2 → 1) has previously been demonstrated, 5 this ap-
proach provides an alternative synthetic route to AIR
Scheme 1
The route that we have used to synthesise AIRs is shown
in Scheme 2. For many years it was claimed that 5-amino-
4-unsubstituted imidazoles are too unstable to be charac-
terised.7 However, in earlier studies we have shown that
they are formed in good yield by catalytic hydrogenation
of 5-nitroimidazoles and can often be isolated and fully
characterised as crystalline solids.8–10 We therefore antic-
ipated that the reduction of the 5-nitroimidazole 10 would
give satisfactory access to AIRs 2. Our initial attempts to
form the triacetyl intermediate 10 were directed towards a
silyl-Hilbert–Johnson reaction between 1,2,3,5-tetra-O-
acetyl-b-D-ribofuranose and 4-nitro-1-trimethylsilylimi-
Synthesis 1999, No. 6, 985–992 ISSN 0039-7881 © Thieme Stuttgart · New York

986
M. J. Humphries, C. A. Ramsden
PAPER
dazole. This approach is attractive because it is anticipat-
ed to give exclusively the 5-nitro isomer 10. It is well
established that alkylation of 4(5)-nitroimidazole anion
gives only 4-nitroimidazoles11 but all attempts, including
the use of 18-crown-6, to form 4-nitro-1-trimethylsilylim-
idazole by silylation of either the potassium or sodium salt
8a, b using either trimethylsilyl chloride or hexamethyl-
disilazane were unsuccessful. In all cases only unreacted
starting material was recovered.
Attention was then directed to formation of intermediate
10 by direct glycosylation of a heavy metal salt of 4(5)-ni-
troimidazole using a modification of a method previously
described by Imbach.12 Reaction of the mercuric bromide
salt 8c with 2,3,5-tri-O-acetyl-D-ribofuranose bromide 9
in xylene under reflux gave a mixture of the isomers 10
and 11 (1:2) with the undesired 4-nitro isomer as the ma-
jor product. We then investigated the analogous reaction
of the silver salt 8d and obtained a mixture (3:2) with the
5-nitro isomer predominating. All further work was there-
fore carried out using the silver salt. On scaling up this re-
action problems were encountered and on some occasions
none of the desired isomer 10 was obtained. These occa-
sions often coincided with the observation of a misty at-
mosphere above the reaction solution and we deduced that
this was associated with excess HBr in the bromide 9. The
5-nitro isomer 10 appears to be very unstable in the pres-
ence of a trace of acid. To overcome this problem xylene
solutions of the freshly prepared bromide 9 were treated
with activated 4Å molecular sieve before addition to the
reaction mixture. The silver salt 8d was also used in ex-
cess (1.5 equivs) to remove any remaining trace of acid.
Using these conditions and an optimum reaction time of
45 minutes an overall yield of 77% was consistently
achieved and the desired 5-nitroimidazole 10 was reliably
obtained in 48% yield.
Scheme 2
Formation of the b-anomers 10 and 11 was expected on
the basis that 2'-acetyl substituents are well known to sta-
bilise intermediate cations preventing attack of the a face.
We found no evidence that a-anomers were formed and
NMR evidence supported the assignment of the b-config-
15
urations. Thus, Montgomery has shown that chemical
shifts of acetyl signals of protected ribofuranose sugars in
the range 2.0–2.2 ppm are indicative of b- rather than a-
configurations. The corresponding signals for isomers 10
and 11 all fall within this range. Subsequently the struc-
ture of compound 10 was firmly established by an X-ray
structure determination of its 5'-tert-butyldiphenylsilyl
derivative 13.16 This firm proof of structure provides fur-
ther evidence of the reliability of the 1H and 13C NMR an-
alytical methods described above.
The mixture of isomers 10 and 11 was conveniently sepa-
rated by flash chromatography and the 5-nitro isomer 10
was efficiently deprotected to give the nucleoside 12 us-
ing triethylamine in aqueous methanol. At this stage it was
vital to ensure that the structural assignments 10 and 12
were correct, in particular, that these compounds were
correctly assigned as the 5-nitro regioisomers and have
the b-configuration at the 1'-position. The 5-nitro struc-
ture was assigned on the basis of the 13C NMR spectral
shifts using the method of McKillop and co-workers.13 In
compound 10 the chemical shifts of the imidazole carbon
atoms are 138.7 (C-2), 134.5 (C-4) and 138.8 (C-5)
whereas in the isomer 11 they are 134.6 (C-2), 148.6 (C-
4) and 117.5 (C-5). These values are typical of the consis-
tently different shifts of a series of pairs of 4- and 5-ni-
troimidazole isomers studied by McKillop and which
provide an unambiguous method of assigning the position
of the nitro group. The cross-ring proton–proton coupling
constants (10, J2,4 0.98 Hz and 11 J2,5 1.47 Hz) are also
consistent with the findings of Matthews and Rapoport 14
that the coupling in 1, 5 isomers (e.g. 10) is in the range
0.9–1.0 Hz whereas for 1, 4-isomers (e.g. 11) it is in the
range 1.1–1.5 Hz.
In accord with our previous studies of the reduction of 5-
nitroimidazoles to 5-aminoimidazoles, compound 12 in
8
dioxane solution was cleanly reduced to AIRs 2 (91%) by
catalytic hydrogenation at atmospheric pressure using 5%
Pd/C catalyst. Filtration and concentration gave a grey
solid which was shown by NMR to be a pure sample of
AIRs 2 and which had spectroscopic properties identical
to those reported by Leonard.5 A stable crystalline deriv-
ative of AIRs 2, prepared by the above method, was made
by
reaction
with
ethoxymethylenemalononitrile
[EtOCH=C(CN)2] giving compound 18. The characterisa-
tion of this product 18 and its cyclisation are discussed be-
low. Since AIRs 2 has previously been transformed to
AIR 1 by treatment with pyrophosphoryl chloride, this
completes a formal synthesis of AIR.
5
Synthesis 1999, No. 6, 985–992 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
A Fresh AIR Synthesis
987
We have also investigated an alternative route in which a cal and spectroscopic data and confirmed by X-ray crys-
final catalytic reduction step leads directly to AIR 1 and tallography.16 Subsequent reaction with benzoic
this is summarised in Scheme 3. In particular, the success- anhydride in pyridine gave the 2', 3'-di-O-benzoyl deriva-
ful reduction of the 5-nitroimidazole nucleoside 12 direct- tive 14 as a syrup (87%) after chromatography and this
ed our attention to a route in which the 5-nitro group and material was desilylated using TBAF to give the primary
the benzyl protected phosphate function of the derivative alcohol 15 as a crystalline solid (76%) that was fully char-
17 are simultaneously reduced in the final stage to give acterised. When the protected alcohol 15 was reacted us-
AIR 1. This attractive approach requires the preparation ing the modified Mitsunobu conditions described above
of the 5’-dibenzylphosphate derivative 17 from the readily the dibenzylphosphate ester 16 was obtained in satisfacto-
available alcohol 12 with the constraint that, because of ry yield (67%). The 1H NMR spectrum showed all the ex-
the instability of the 5-nitroimidazole 12, acidic reaction pected features including coupling of the non-equivalent
conditions had to be avoided. For this reason we were par- 5'-CH2 protons at d 4.38 with the phosphorus nucleus as
ticularly attracted to the use of a modified Mitsunobu re- well as with the 4'-proton. Removal of the benzoyl pro-
action to form the P–O bond using conditions described tecting groups using NaOH in aqueous MeOH gave the
by Saady and co-workers17 in which dibenzylphosphate desired derivative 17 (88%) as a gummy syrup that was
reacts with an intermediate alkoxyphosphonium salt gen- pure by tlc and NMR spectroscopy.
erated under basic conditions (pyridine solution). Both
The final stage in the AIR preparation shown in Scheme 3
18
Mitsunobu and Saady and co-workers17 have reported
requires simultaneous reduction and deprotection by cata-
the phosphorylation of unprotected nucleosides, although
the latter group obtained poor yields. Initially, therefore,
lytic hydrogenation. When compound 17 was treated with
5% Pd/C in dioxane under an atmosphere of hydrogen lit-
we investigated the direct 5'-phosphorylation of the alco-
tle reduction took place, even after 24 hours. Investigation
hol 12 using dibenzylphosphate, Ph3P and DEAD in pyri-
of the system showed that the substrate 17 has very low
dine solution but only a poor yield of the desired product
solubility in dioxane and we believe that this contributed
17 was obtained as part of an inseparable mixture. It was
to the low reactivity. Our previous studies of reduction of
clear that in order to obtain satisfactory yields it would be
5-nitroimidazoles have shown that dioxane is the pre-
necessary to selectively protect the secondary alcohols.
ferred solvent giving much cleaner results than ethanol.8
The 2', 3'-di-O-benzoyl derivative 16 was therefore pre-
However, in the light of the solubility problem we then in-
vestigated the same reduction in ethanol solution. Under
pared using the route shown in Scheme 3.
these conditions, the theoretical amount of hydrogen (5
equivalents) had been taken up in eight hours and removal
of the catalyst and solvent gave a colourless solid that was
shown to be predominantly AIR 1 with a 1H NMR spec-
trum virtually identical in all respects with that reported
by Leonard and co-workers.5 The instability of AIR is
well known and this precluded further purification. The
major contaminant was identified by NMR as the partial
reduction product in which the O-benzyl substituents had
been removed but the nitro group remained intact. In spite
of a number of attempts, we did not obtain a sample of
AIR free from this material and further purification of
AIR using this approach would require specialist chroma-
tography.
The synthetic AIRs 2 was further characterised
by
reaction
with
ethoxymethylenemalononitrile
[EtOCH=C(CN)2] to give the 4-(2',2'-dicyanovinyl) de-
rivative 18. This product was isolated as stable yellow
crystals (53%) and fully characterised. In particular the 1H
NMR spectrum of compound 18 showed the presence of
an NH2 function (d 7.66; 2 protons) and the absence of an
imidazole 4-H signal. None of the isomeric N-adduct was
detected. In previous studies 9 we have shown that 4-un-
substituted-5-aminoimidazoles are N,C-ambident nucleo-
philes and that soft electrophiles react on the imidazole
ring to give 4-substituted products. In contrast, hard elec-
trophiles add to the exocyclic nitrogen atom. We have ra-
tionalised this behaviour in terms of the relative sizes of
the HOMO coefficients and the charge distribution in 5-
aminoimidazoles (Figure 1).9 Since the reagent
Scheme 3
The 5'-t-butyldiphenylsilyl derivative 13 was selectively
prepared in 72% yield by reaction of the 5-nitroimidazole
12 with t-butyldiphenylchlorosilane and imidazole in
DMF solution. The structure of this product, isolated as
colourless needles, was fully supported by microanalyti-
Synthesis 1999, No. 6, 985–992 ISSN 0039-7881 © Thieme Stuttgart · New York

988
M. J. Humphries, C. A. Ramsden
PAPER
EtOCH=C(CN)2 is classified as soft, the exclusive forma-
tion of the C-adduct 18 is in agreement with expectation
and provides a useful synthetic procedure.
Figure 1
The formation of the C–C bond at the 4-position of the im-
idazole ring of AIRs in the reaction 2 → 18 is comparable
to the strategically important C–C bond forming sequence
in the biosynthetic steps 1 → 5 (Scheme 1). For this rea-
son the derivative 18 is a useful intermediate for the syn-
thesis of novel nucleoside analogues and access to this
compound has enabled us to develop a general route to
"pyridine stretched" nucleoside analogues (Scheme 4).
The dicyanovinyl compound 18 readily underwent cycli-
sation (80%) to the 5-amino-6-cyanoimidazol[4, 5-b]pyri-
dine 20 using a trace of NaHCO3 in hot ethanol.
Subsequent oxidation of this 1-deazapurine using 30%
H2O2 in ammonium hydroxide solution gave the corre-
sponding amide 22 (67%). Both compounds were fully
characterised and evidence for cyclisation included the
observation of C-7 aromatic ring protons at ca d 8.4 and
strong UV absorption at 340 nm.
Scheme 4
The convenient 1, 2-difunctionality of the 1-deazapurines
20 and 22 facilitates the formation of a third ring. An ini-
tial attempt to react compound 20 with formamidine ace-
tate using conditions described by Taylor and Ehrhart19
gave only recovered starting material. However, reaction
of the nitrile 20 with the more reactive diethoxymethyl ac-
etate [EtO2CH–OCOMe]20 followed by treatment of the
crude product with methanolic ammonia gave the
[4',5':5,6]pyrido[2,3-d]pyrimidine 19. This reaction is be-
lieved to occur via initial formation of an acetyloxime fol-
lowed by amidine formation and cyclisation. The product
19 is the "pyridine stretched" analogue of adenosine and
our approach makes this novel nucleoside analogue readi-
ly accessible. When the nitrile 20 was treated with carbon
disulfide in hot pyridine the dithione derivative 21 was
formed. We believe that this product is formed by CS2 ad-
dition to the amino group, cyclisation to form a 1,3-thiaz-
ine ring followed by a Dimroth rearrangement to give the
thermodynamically more stable pyrimidine derivative 21.
Both of the derivatives 19 and 21 are high melting crystal-
line solids that have 1H NMR spectra fully supporting the
proposed tricyclic structures.
the amide 22 with ethyl formate and sodium ethoxide in
hot ethanol gave the product 23 as colourless needles
(66%). The H NMR spectrum clearly shows three aro-
matic proton singlets at d 8.3, 8.75 and 8.95. An attempt
to form the 6, 8-dione corresponding to the 6, 8-dithione
21 by reaction of the amide 22 with diethyl carbonate gave
the desired product [d 8.9 and 8.95]. However, all at-
tempts to prepare a pure sample free from the starting ma-
terial, which has a very similar Rf value, were
unsuccessful.
1
Mps were determined using a Reichert Kofler Block apparatus and
are uncorrected. Infrared spectra were recorded on a Perkin-Elmer
881 spectrophotometer with only major absorbances being quoted.
1H NMR spectra were recorded at ambient temperatures using a Jeol
GSX270 FT-NMR spectrometer. Chemical shifts are quoted in ppm
and the following abbreviations are used: s = singlet; d = doublet; t
= triplet; q = quartet; m = multiplet; br = broad. Coupling constants
are quoted to the nearest 0.1 Hz. Elemental analyses were deter-
mined using a Perkin-Elmer 240 CHN Elemental Analyser. Low
resolution mass spectra were recorded on an AEI MS12 Mass Spec-
trometer at 70 eV electron impact ionisation. Fast Atom Bombard-
ment (FAB) spectra were recorded by the EPSRC mass
spectrometry service (Swansea) and the abbreviations b and s refer
We have also prepared compound 23, which is the "pyri-
dine stretched" analogue of inosine (cf. 7). Treatment of
Synthesis 1999, No. 6, 985–992 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
A Fresh AIR Synthesis
989
to the nucleoside base and sugar fragments respectively. Flash chro-
matography was performed using silica gel (Janssen Chimica; par-
ticle size 0.035–0.07 mm). Preparative radial (chromatotron)
chromatography was carried out on a Harrison Research Ltd Chro-
matotron 7924 using a 2 mm plate with silica gel 60 PF254 contain-
ing gypsum (Merck). Concentration and evaporation refer to the
removal of volatile materials under reduced pressure on a Büchi Ro-
tovap. All solvents were purified and distilled before use. CH2Cl2,
1,2-dichloroethane, toluene and xylene were distilled from P2O5
and stored in a dark bottle over molecular sieves (4Å). MeCN was
doubly distilled from P2O5. THF was dried by heating under reflux
over sodium until benzophenone held a purple colour. 1, 4-Dioxane
was tested for peroxides with acidified KI solution and, if positive,
the solution was passed through activated alumina and then treated
in the same way as THF.
ed under reflux (with drying tube). 2,3,5-Tri-O-acetyl-D-ribofura-
nose bromide21 9 (2.0 g, 6 mmol), freshly prepared from 1,2,3,5-
tetra-O-acetyl-b-D-ribofuranose, was taken up in xylene (20 mL)
and freshly activated molecular sieve (4Å)(10 g) was added with
swirling. When the take up of any dissolved HBr was complete (ca.
5 mins), the solution was decanted into a dropping funnel and added
dropwise to the refluxing salt solution. When addition was com-
plete, heating was continued (1 h) and the solution was then cooled
and filtered. The solvent was evaporated to give a syrup which was
dissolved in CHCl3 and washed successively with NaHCO3 and
H2O (20 mL). The organic layer was dried (Na2SO4) and evaporated
to give a yellow syrup (2.0 g). TLC [silica gel: petroleum ether (bp
60–80 °C): EtOAc, 1:1 as eluent] showed this to be a two compo-
nent mixture. Separation was carried out by flash column chroma-
tography. The faster running product was recrystallised from
EtOAc to afford compound 10 as colourless prisms (0.91 g, 48%);
mp 93–95 °C [lit.12 94–96 °C]
4(5)-Nitroimidazole Metal Salts
IR (KBr): n = 1044, 1368, 1466, 1526, 1748, 2398, 3018 cm–1.
Potassium Salt 8a; Typical Procedure
4(5)-Nitroimidazole (1.13 g, 10 mmol) was suspended in MeOH
(25 mL) and powdered KOH (0.56 g, 10 mmol) was added. The
mixture was stirred at 80 °C (10 min). Evaporation gave a bright
yellow solid (1.16 g) that was recrystallised from absolute EtOH to
afford compound 8a as a yellow powder (1.11 g, 73%); mp >300
°C.
1H NMR (CDCl3/TMS): d = 8.13 (1H, d, J = 1.0Hz, 2-H), 8.05 (1H,
d, J = 1.0Hz, 4-H), 6.4 (1H, d J = 2.0Hz, 1’-H), 5.5 (1H, m, 2’-H),
5.3 (1H, m, 3’-H), 4.5 (1H, m, 4’-H), 4.3 (1H, m, 5’-CH2), 2.03 (3H,
s, CH3CO), 2.17 (6H, s, 2 x CH3CO).
13C NMR (CDCl3/TMS): d = 170.5, 170.1, 169.3 (C=O), 138.8 (C-
2), 134.5 (C-4), 90.0 (C-1’), 79.7 (C-4’), 75.2 (C-2’), 68.7 (C-3’),
61.6 (C-5’), 20.7, 20.3 (CH3C=O).
1H NMR (DMSO-d6/TMS): d = 7.98 (1H, s, H-2), 7.39 (1H, s, H-5).
C3H2N3KO2: calc. C 23.84 H 1.33 N 27.80
MS (EI): m/z = 259 (s), 115 (b+3H).
(151.17)found:
C 23.83 H 1.24 N 27.68
C14H17N3O9: calc. C 45.27 H 4.61 N 11.32
Sodium Salt 8b
(371.30)found
C 45.50 H 4.86 N 11.00
Using NaOH in a procedure identical to that described for potassi-
um salt 8a gave compound 8b as a yellow solid (1.11 g, 82%); mp
>300 °C.
1H NMR (DMSO-d6/TMS): d = 7.99 (1H, s, H-2), 7.41 (1H, s, H-5).
The slower running product was recrystallised form EtOAc to af-
ford compound 11 as colourless needles (0.47 g, 29%); mp 100–103
°C [lit.12 103–104 °C].
IR (KBr): n = 1042, 1208, 1420, 1510, 1748, 2402, 3022 cm–1.
C3H2N3NaO2: calc. C 26.63 H 1.49 N 31.11
(135.06)found
C 26.83 H 1.34 N 30.78
1H NMR (CDCl3/TMS): d = 8.05 (1H, d J = 1.5Hz, 2-H), 7.7 (1H,
d J = 1.5Hz, 5-H), 5.8 (1H, d J = 3.0Hz, 1’-H), 5.3 (2H, m, 2’, 3’-H),
4.5 (1H, m, 4’-H), 4.4 (2H, m, 5’-CH2), 2.14 (6H, s, 2 x CH3CO),
2.18 (3H, s, CH3CO).
13C NMR (CDCl3/TMS): d = 170.2, 169.7 (C=O), 134.6 (C-2),
148.6 (C-4), 117.0 (C-5), 89.1 (C-1’), 80.7 (C-4’), 75.2 (C-2’), 69.7
(C-3’), 62.4 (C-5’), 20.6, 20.3 (CH3C=O).
Mercuric Bromide Salt 8c
A mixture of 4(5)-nitroimidazole (1.13 g, 10 mmol) and NaOH (0.4
g, 10 mmol) was dissolved in 50% aq EtOH (70 mL). The yellow
solution was then heated under reflux and mercuric bromide (3.6 g,
10 mmol) in the same solvent (30 mL) was added. A pale precipitate
formed and heating and stirring were maintained (10 min). After
cooling in an ice bath, the product was collected, washed succes-
sively with H2O, EtOH and Et2O, and dried under vacuum at 56 °C
to give compound 8c as a pale cream powder (2.6 g, 66%); mp >280
°C.
MS (EI): m/z = 259 (s), 139 (s-2AcOH), 115 (b+3H).
C14H17N3O9: calc. C 45.27 H 4.61 N 11.32
(371.30)found
C 45.31 H 4.59 N, 11.32
C3H2N3BrHgO2: calc.C 9.18 H 0.51 N 10.70
(392.56)found
C 9.10 H 0.79 N 10.82
5-Nitro-1-(b-D-ribofuranose)imidazole (12)
5-Nitro-1-(2,3,5-tri-O-acetyl-b-D-ribofuranosyl)imidazole
(10)
Silver Salt 8d
(0.25 g, 0.6 mmol) was dissolved in a mixture of MeOH, Et3N and
H2O (8:1:1 v/v) (10 mL). The mixture was stirred overnight at r.t.
Evaporation, followed by addition and evaporation of H2O (2 x 2
mL) gave a pale yellow foam. Analysis by 1H NMR confirmed the
clean formation of compound 12 (0.14 g, 95%) and this was used
without further purification.
A mixture of 4(5)-nitroimidazole (1.13 g, 10 mmol) and ammonia
solution (10 mL, S.G. 0.88) was stirred and to the yellow solution
was added a solution of silver nitrate (1.7 g, 10 mmol) in H2O (10
mL). Immediately a pale yellow precipitate formed. Stirring was
continued (10 min) and the gelatinous precipitate was collected,
washed with portions of EtOH/Et2O and dried under vacuum at 56
°C to give compound 8d as a pale yellow solid (2.02 g, 92%); mp
>300 °C.
1H NMR (DMSO-d6/TMS): d = 8.6 (1H, s, 2-H), 8.1 (1H, s, 4-H),
6.2 (1H, d J = 1.5Hz, 1’-H), 4.2 (1H, br s, 2’-H), 4.0 (1H, br s, 3’-H),
3.9 (1H, m, 4'-H), 3.8–3.6 (2H, br m, 5'-CH2), 5.1–5.6 (3 x OH).
C3H2N3AgO2: calc.C 16.38 H 0.92 N 19.11
(219.94)found
C 16.40 H 0.80 N 18.60
5-Amino-1-(b-D-ribofuranosyl)imidazole (AIRs) (2)
5-Nitro-1-(b-D-ribofuranose)imidazole 12 (0.2 g, 0.8 mmol) was
dissolved in pure anhyd. 1, 4-dioxane (10 mL) [freshly prepared by
passing through activated alumina and distillation from sodium]. To
this solution was added 5% palladium on charcoal (0.2 g, 100% wt/
wt). The mixture was hydrogenated at atmospheric pressure with
5-Nitro-1-(2,3,5-tri-O-acetyl-b-D-ribofuranosyl)imidazole (10)
4(5)-Nitroimidazole silver salt 8d (1.58 g, 7.2 mmol) was suspend-
ed in xylene (100 mL) and solvent was removed by distillation until
pure xylene (bp 138 °C) was collected. The solution was then heat-
Synthesis 1999, No. 6, 985–992 ISSN 0039-7881 © Thieme Stuttgart · New York

990
M. J. Humphries, C. A. Ramsden
PAPER
vigorous shaking for 5 h after which the mixture had taken up three
equivalents of hydrogen (53 mL). The mixture was then filtered
through Celite to remove charcoal giving a bright pink solution.
Evaporation gave a grey solid that was collected, dried under high
vacuum, and stored in a desiccator at 0 °C. This product was iden-
tified as 5-amino-1-(b-D-ribofuranosyl)imidazole (2) (0.16 g, 91%).
1H NMR (D2O/TMS): d = 7.5 (1H, s, 2-H), 6.3 (1H, s, 4-H), 5.5 (1H,
d J = 6.0Hz, 1'-H), 4.4 (1H, pseudo-t J = 6.0Hz, 2'-H), 4.2 (1H,
pseudo-t, J = 5.0Hz, 3'-H), 4.0 (1H, m, 4'-H), 3.7 (2H, m, 5'-CH2).
lised from EtOH to afford compound 22 as colourless needles (13
mg, 67%); mp 255–7 °C.
IR (KBr): n = 1064, 1396, 1560, 1588, 1622, 1656, 1764, 2908,
3094, 3290, 3368, 3432, 3464 cm–1.
1H NMR (DMSO-d6/TMS): d = 8.36 (1H, s, 7-H), 8.31 (1H, s, 2-H),
7.95 (1H, br s, CONH2’), 7.29 (3H, br s, NH2 + CONH2"), 5.87 (1H,
d J = 5.8Hz, 1'-H), 5.0–5.4 (3 x OH), 4.53 (1H, m, 2'-H), 4.12 (1H,
m, 3'-H), 3.91 (1H, m, 4'H), 3.60 (2H, br m, 5'-CH2).
UV (EtOH): lmax = 213 (e 17200), 239 (e 8150), 261 (e 2950) and
342 (e 5500) nm.
MS (FAB): m/z = 309 (M+), 310 (M+H), 175 (b-H), 205 (b+CHO).
13C NMR (D2O/TMS): d = 137.2 (C-5), 128.3 (C-2), 118.1 (C-4),
89.0 (C-1'), 85.0 (C-4'), 74.1 (C-3'), 69.7 (C-2'), 60.5 (C-5')(identi-
cal to the literature values5).
C12H15N5O5: calc. C 46.60 H 4.89 N 22.64
5-Amino-4-(2, 2-dicyanovinyl)-1-(b-D-ribofuranosyl)imidazole
(18)
(309.28)found
C 46.44 H 4.82 N 22.39
Ethoxymethylenemalonitrile (0.1 g, 0.8 mmol) was added to a solu-
tion of 2 (0.17 g, 0.8 mmol) in dioxane and the mixture stirred vig-
orously overnight. The dark solution was analysed by TLC (silica
gel: EtOAc–MeOH 9:1 as eluent) and showed one major product as
a yellow spot (Rf 0.3). The mixture was filtered through Celite and
purification by flash chromatography gave a bright yellow solid.
This material was recrystallised from EtOH to afford compound 18
as yellow needles (123 mg, 53%); mp 181–183 °C.
8-Amino-3-(b-D-ribofuranosyl)[4’,5’:5,6]pyrido[2,3-d]pyrimi-
dine (19)
A suspension of 20 (50 mg, 0.17 mmol) in excess diethoxymethyl
acetate (1.5 mL) was heated under reflux (2 h). After removal of the
excess reagent, the resulting dark syrup was dissolved in methanolic
ammonia (10 mL) and allowed to stand at r.t. (24 h). TLC (silica gel:
EtOAc as eluent) showed the formation of a product at Rf 0.2 and,
after evaporation, the yellow syrup was dissolved in CHCl3 (10
mL). Filtration and evaporation gave a pale yellow foam which was
dissolved in 25% w/w HOAc (10 mL) and stirred overnight. Fol-
lowing the addition and evaporation of H2O (3 x 7.5 mL) the acidity
of the solution was adjusted to pH 6 using 1M NaOH. The resulting
mixture was then chilled and the solid which formed was collected
and recrystallised from H2O to afford compound 19 as colourless
needles (30 mg, 55%); mp 276–9 °C (decomp).
IR (KBr): n = 1078, 1342, 1450, 1554, 1650, 2218, 2924,
3384 cm–1.
1H NMR (DMSO-d6/TMS): d = 7.82 (1H, s, CH=C), 7.68 (1H, s, 2-
H), 7.66 (2H, br s, NH2), 5.5 (1H, d J = 6.0Hz, 1’-H), 4.3 (1H, m, 2’-
H), 4.0 (1H, m, 3’-H), 3.9 (1H, m, 4’-H), 3.6 (2H, m, 5’-CH2).
UV (EtOH): lmax = 391 (e 17900) nm.
MS (FAB): m/z = 291 (M+), 292 (M+H).
C12H13N5O4: calc. C 49.48 H 4.50 N 24.04
IR (KBr): u = 1048, 1184, 1280, 1332, 1356, 1420, 1502, 1582,
1622, 1690, 3298 cm–1.
1H NMR (DMSO-d6/TMS): d = 9.08 (1H, s, 9-H), 8.92 (1H, s, 6-H),
8.50 (1H, s, 2-H), 8.05 (2H, br s, NH2), 6.11 (1H, d J = 5.9Hz, 1’-
H), 4.71 (1H, pseudo-t, 2’-H), 4.22 (1H, m, 3’-H), 4.00 (1H, m, 4’-
H), 3.67 (2H, br m, 5’CH2), 5.2–5.4 (3 x OH).
(291.27)found
C 49.42 H 4.42 N 24.09
5-Amino-6-cyano-3-(b-D-ribofuranosyl)imidazo[4,5-b]pyridine
(20)
To a hot solution of 18 (0.1 g, 0.35 mmol) in EtOH (10 mL) was
added 5% aq NaHCO3 (5 drops). The mixture was then heated under
reflux with stirring (15 min). After this time the colour of the solu-
tion had faded and TLC (silica gel: EtOAc, MeOH, 9:1 as eluent)
showed only one very intense spot (Rf 0.3). After cooling, the solu-
tion was concentrated and chilled (1 h). The off-white solid which
formed was collected (78 mg) and recrystallised from EtOH to af-
ford compound 20 as colourless needles (68 mg, 68%); mp 219–220
°C.
UV (EtOH): lmax = 220 (e 26900), 246 (e 16950) and 336 (e 9750)
nm.
MS (FAB): m/z = 318(M+), 319.1138 (MH+ requires 319.1150),
187(b+2H), 215 (b+CH2O)
C13H14N6O4: calc. C 49.06 H 4.43 N 26.40
(318.29)found
C 48.76 H 4.36 N 26.18
3-(b-D-Ribofuranosyl)[4’,5’:5,6]pyrido[2,3-d]pyrimidine-6,8-
dithione (21)
IR (KBr): n = 1114, 1234, 1308, 1430, 1508, 1574, 1636, 2220,
2914, 3230, 3440 cm–1.
1H NMR (DMSO-d6/TMS): d = 8.44 (1H, s, 7-H), 8.26 (1H, s, 2-
H)(slowly exchanges), 6.77 (2H, br s, NH2), 5.88 (1H, d J = 6.0Hz,
H-1’), 4.48 (1H, m, H-2’), 4.12 (1H, m, H-3’), 3.90 (1H, m, H4’),
3.5–3.7 (2H, m, 5'-CH2), 5.0–5.4 (3 x OH).
5-Amino-6-cyano-3-(b-D-ribofuranosyl)imidazo[4,5-b]pyridine 20
(30 mg, 0.1mmol) was dissolved in anhyd pyridine (5 mL) and heat-
ed under reflux. To this solution, CS2 (freshly distilled) (2 mL) was
added dropwise over 5 min and heating was then maintained (18 h).
Cooling and evaporation gave a bright yellow powder that was re-
crystallised from aq EtOH to afford compound 21 as a fine orange
powder (20 mg, 52%); mp > 280°C.
UV (EtOH): lmax = 240 (e 22900), 340 (e 11700) nm.
MS (FAB): m/z = 291 (M+), 292 (M+H), 160 (b+2H), 188
(b+CH2O).
IR (KBr): u = 1236, 1304, 1368, 1398, 1572, 1582, 1620, 2856,
2922, 3414 cm–1.
C12H13N5O4: calc. C 49.48 H 4.50 N 24.04
1H NMR (DMSO-d6/TMS): d = 13.67 (2H, br s, NH), 8.86 (1H, s,
6-H), 8.82 (1H, s, 2-H), 6.01 (1H, d J = 5.9 Hz, 1’-H), 4.68 (1H, m,
2’-H), 4.22 (1H, m, 3’-H), 3.95 (1H, m, 4’-H), 3.65 (2H, br m, 5’-
CH2), 5.0–5.5 (3 x OH).
(291.27)found
C 49.24 H 4.29 N 24.18.
5-Amino-3-(b-D-ribofuranosyl)imidazo[4,5-b]pyridine-6-car-
boxamide (22)
A mixture of 20 (20 mg), NH4OH (1 mL) and 30% H2O2 (100 µL)
was stirred at r.t. (30 min). The mixture was then refrigerated over-
night and the colourless solid product was collected and recrystal-
MS (FAB): m/z = 367 (M+), 368 (M+H).
C13H13N5O4S2: calc.C 42.50 H 3.57 N 19.06
(367.40)found
C 42.79 H 3.30 N 18.83
Synthesis 1999, No. 6, 985–992 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
A Fresh AIR Synthesis
991
3-(b-D-ribofuranosyl)[4’,5’:5,6]pyrido[2,3-d]pyrimidin-8-one
(23)
C24H29N3O6Si: calc.C 59.61 H 6.04 N 8.69
(483.60)found C 60.28 H 6.18 N 8.67
5-Amino-3-(b-D-ribofuranosyl)imidazo[4,5-b]pyridine-6-carboxa-
mide 22 (25 mg, 0.08 mmol) was dissolved in ethanolic sodium
ethoxide solution, freshly made from sodium (75 mg) and EtOH (5
mL). To this mixture was added ethyl formate (150 mL) and the so-
lution was heated under reflux (2 h). After cooling, H2O was added
(5 mL) and the acidity of the clear solution adjusted to pH 5–6 using
2M HCl. Concentration and cooling gave a solid which was collect-
ed and recrystallised from 5% aq MeOH to afford compound 23 as
colourless needles (17.5 mg, 66%); mp 230 °C (dec).
1-(b-D-5’-tert-Butyldiphenylsilyloxy-2’,3’-di-O-benzoylribo-
furanose)-5-nitroimidazole (14)
A mixture of 13 (0.6 g, 1.2 mmol), DMAP (0.25 eq) and benzoic an-
hydride (2.8 g, 10 eq) was dissolved in anhyd. pyridine (15 mL) at
0 °C and stirred overnight at r.t. The mixture was then poured onto
ice (50 mL) and extracted with CHCl3 (3 x 20 mL). The organic
phase was dried (Na2SO4) and evaporated to give a dark syrup
which was purified by chromatography (silica gel: CHCl3–MeOH,
9:1 as the eluent). Collection of the component at Rf 0.8 gave a pale
syrup which was identified as compound 14 (0.7 g, 87%) and used
without further purification.
IR (KBr): n = 1076, 1118, 1398, 1506, 1570, 1632, 2936, 3424,
3668, 3698 cm–1.
1H NMR (DMSO-d6/TMS): d = 8.95 (1H, s, 9-H), 8.75 (1H, s, 6-H),
8.35 (1H, br s, NH), 8.3 (1H, s, 2-H), 6.12 (1H, d J = 5.8Hz, 1’-H),
5.20–5.56 (3 x OH), 4.69 (1H, m, 2'-H), 4.22 (1H, m, 3'-H), 4.01
(1H, m, 4'-H), 3.68 (2H, br m, 5'-CH2).
IR (CHCl3): n = 894, 1180, 1376, 1418, 1464, 1726, 1816, 2932,
2998, 3152, 3410 cm–1.
1H NMR (CDCl3/TMS): d = 8.46 (1H, s, 2-H), 7.3–8.19 (21H, br m,
4-H, Ar), 6.85 (1H, d J 2.9Hz, 1'-H), 5.96 (2H, m, 2', 3'-H), 4.55
(1H, m, 4'-H), 3.8–4.2 (2H, br m, 5'-CH2), 1.10 (9H, s, (CH3)3Si).
UV (EtOH): lmax = 219 (e 27600), 254 (e 9200), 319 (e 10800) and
328 (e 10850) nm.
MS (FAB): m/z = 319 (M+), found 320.0979 (M+H requires
320.0995); C13H13N5O5 (M).
1-(b-D-2’,3’-Di-O-benzoylribofuranose)-5-nitroimidazole (15)
To a solution of 14 (0.1 g, 0.14 mmol) in anhyd THF (5 mL) was
added TBAF (1M in THF, 0.14 mL, 1 eq.) dropwise. After 2 h the
crude reaction material was purified by flash chromatography (sili-
ca gel: petroleum ether (bp 60–80 °C)–EtOAc 70:30 as the eluent)
and the main product (Rf 0.6) was recrystallised from petroleum
ether (bp 60–80 °C)–EtOAc to afford compound 15 as colourless
needles (45 mg, 76%); mp 168–170 °C.
3-(b-D-Ribofuranosyl)[4’,5’:5,6]pyrido[2,3-d]pyrimidine-6,8-di-
one
5-Amino-3-(b-D-ribofuranosyl)imidazo[4, 5-b]pyridine-6-carboxa-
mide 22 (25 mg, 0.08 mmol) was dissolved in ethanolic sodium
ethoxide solution, freshly made from sodium (75 mg) and EtOH (5
mL). To this stirred solution was added diethyl carbonate (freshly
distilled) (54 mL, 0.44 mmol) and the mixture was heated under re-
flux (5 h). The resulting gelatinous material was then dissolved in
H2O (5 mL) and the acidity of the solution adjusted to pH 5–6 using
2M HCl. Evaporation to dryness afforded a solid which was extract-
ed with hot EtOH (3 x 10 mL). The combined extracts were analy-
sed by tlc (silica gel: EtOAc–MeOH 9:1 as the eluent) which
showed two closely running intense spots corresponding to unreact-
ed starting material and a product. Despite all efforts, the separation
of the mixture by chromatography proved to be incomplete but
analysis of the higher running component showed a mixture en-
riched in a product identified as 3-(b-D-ribofuranosyl)[4',5':5,6]py-
rido[2,3-d]pyrimidine-6, 8-dione.
IR (CHCl3): n = 796, 1016, 1094, 1120, 1204, 1370, 1424, 1478,
1506, 1728, 2400, 3038, 3686 cm–1
1H NMR (CDCl3/TMS): d = 8.64 (1H, s, 2-H), 7.29–8.06 (11H, br
m, Ar, 4-H), 6.78 (1H, d J = 2.6Hz, 1'-H), 5.92 (1H, m, 2'-H), 5.81
(1H, m, 3'-H), 4.51 (1H, m, 4'-H), 3.89–4.14 (2H, br m, 5'-CH2).
MS (FAB): m/z = 454 (M+H), 341(M+-b), 113 (b+H).
C22H19N3O8: calc. C 58.28 H 4.22 N 9.27
(453.41)found
C 58.23 H 4.17 N 9.01
1-(b-D-5’-Dibenzylphosphoroyloxy-2’,3’-di-O-benzoylribofura-
nose)-5-nitroimidazole (16)
1H NMR (DMSO-d6/TMS): d = 11.7 (1H, br s, NH), 11.4 (1H, br,
NH), 8.9 (1H, s, 2-H), 8.7 (1H, s, 6-H), 6.2 (1H, d J = 5.4 Hz, 1'-H),
5.25–5.56 (3 x OH), 4.74 (1H, m, 2'-H), 4.31 (1H, m, 3'-H), 4.1 (1H,
m, 4'-H), 3.7 (2H, br m, 5'-CH2).
To a mixture of 15 (227 mg, 5 mmol), dibenzyl phosphate (139 mg,
5 mmol) and freshly recrystallised PPh3 (328 mg, 2.5 eq) in anhyd
pyridine (5 mL) was added DEAD (218 mg, 196 mL, 2.5 eq) drop-
wise with stirring. The mixture was stirred for a further 3 h. The sol-
vent was removed and the residue subjected to chromatotron
chromatography (EtOAc–petroleum ether (bp 60–80 °C) 40:60 as
the eluent). Four fractions were collected and the component col-
lected at Rf 0.4 was identified as compound 16 (240 mg, 67%) as a
pale syrup and used without further purification.
1H NMR (CDCl3/TMS): d = 8.24 (1H, s, 2-H), 8.05 (1H, s, 4-H),
7.96–7.26 (20H, br m, Ar), 6.81 (1H, d J 2.2 Hz, 1'-H), 5.73 (2H, m,
2', 3'-H), 5.12 (4H, m, ArCH2), 4.58 (1H, m, 4'-H), 4.38 (2H, br m,
5'-CH2).
1-(b-D-5’-tert-Butyldiphenylsilyloxyribofuranose)-5-nitroimi-
dazole (13)
A mixture of 12 (0.5 g, 2.04 mmol), TBDPSCl (1.2 eq, 634 µL) and
imidazole (2.4 eq, 0.33 g) in anhyd DMF (5 mL) was stirred at r.t.
overnight. Removal of the DMF followed by extraction with CHCl3
gave after evaporation a pale syrup. Separation by chromatography
(silica gel: CHCl3–MeOH 9:1 as eluent) gave a major component
(Rf 0.6) which after recrystallisation from petroleum ether (bp 60–
80 °C): EtOAc afforded compound 13 as colourless needles (0.71 g,
72%); mp 172–5 °C.
1-(b-D-5’-Dibenzylphosphoroyloxyribofuranose)-5-nitroimida-
zole (17)
IR (KBr): n = 708, 796, 820, 878, 996, 1058, 1108, 1216, 1370,
1416, 1430, 1474, 1530, 2858, 2934, 3322, 3552 cm–1
1-(b-D-5'-Dibenzylphosphoroyloxy-2',3'-di-O-benzoylribofura-
nose)-5-nitroimidazole (16) (240 mg, 3.3 mmol) was dissolved in
1% solution of NaOH in MeOH (10 mL) [freshly prepared by dis-
solving 0.1g of NaOH in H2O (10 mL) and dilution with MeOH (90
mL)]. After stirring (45 min), evaporation and extraction of the res-
idue with CHCl3 gave a pale syrup. This material was purified by
chromatography (silica gel: CHCl3–EtOAc 1:1 as the eluent) to give
compound 17 as a gummy syrup (150 mg, 88%).
1H NMR (DMSO-d6/TMS): d = 8.43 (1H, s, 2-H), 8.12 (1H, s, 4-H),
7.96–7.40 (10H, br m, Ph2Si), 6.22 (1H, s, 1'-H), 5.78–5.23 (2 x
OH), 4.26 (2H, m, 2', 3'-H), 3.99 (3H, br m, 4'-H, 5'-CH2), 1.01 (9H,
s, (CH3)3CSi).
UV (EtOH): lmax = 296 (e 48900) nm.
MS (EI): m/z = 227 (M+ 255), 115 (b+3H).
Synthesis 1999, No. 6, 985–992 ISSN 0039-7881 © Thieme Stuttgart · New York

992
M. J. Humphries, C. A. Ramsden
PAPER
IR (CHCl3): n = 998, 1112, 1254, 1368, 1450, 1602, 1726, 2988,
(2) Mueller, E.J.; Meyer, E.; Rudolph, J.; Davisson, V.J.; Stubbe,
J. Biochemistry 1994, 33, 2269.
3370, 3608, 3694 cm–1
Firestine, S.M.; Davisson, V.J. Biochemistry 1994, 33, 11917.
Firestine, S.M.; Poon, S-W.; Mueller, E.J.; Stubbe, J.;
Davisson, V.J. Biochemistry 1994, 33, 11927.
(3) Begley, T.P. Natural Product Reports, 1996, 13, 177.
Estramareix, B.; David, S. Biochem. Biophys. Res. Commun.
1986, 134, 1136.
1H NMR (CDCl3/TMS): d = 8.13 (1H, s, 2-H), 8.02 (1H, s, 4-H),
7.43 (10H, d J = 8.7 Hz, Ph), 6.3 (1H, s, 1’-H), 5.06 (4H, m, PhCH2),
4.28 (5H, br m, 2’, 3’, 4’-H, 5’-CH2).
MS (FAB): m/z = 505 (M+), found: 506.1335 (M+H requires
506.1328) C22H25N3O9P = M+H), 393 (M+-b).
(4) Estramareix, B.; Thérisod, M. J. Am. Chem. Soc. 1984, 106,
3857.
(5) Bhat, B.; Groziak, M.P.; Leonard, N.J. J. Am. Chem. Soc.
1990, 112, 4891.
(6) Prelimimary communication: Ramsden, C.A.; Humphries,
M.J. Synlett 1995, 203.
(7) Lythgoe, D.J.; Ramsden, C.A. Adv. Heterocycl. Chem. 1994,
61, 1.
(8) Ramsden, C.A. Chem. Heterocyl. Compounds 1995, 1323;
Jones, R.H.; Lothian, A.P.; Ramsden, C.A. Acta Cryst. 1996,
C52, 982.
(9) Al-Shaar, A.H.M.; Gilmour, D.W.; Lythgoe, D.J.;
McClenaghan, I.; Ramsden, C.A. J. Chem. Soc., Perkin Trans.
1 1992, 2779.
(10) Al-Shaar, A.H.M.; Chambers, R.K.; Gilmour, D.W.; Lythgoe,
D.J.; McClenaghan, I.; Ramsden, C.A. J. Chem. Soc., Perkin
Trans. 1 1992, 2789.
(11) Grimmett, M.R. Adv. Heterocycl. Chem. 1980, 27, 241.
Begtrup, M.; Larsen, P. Acta Chem. Scand. 1990, 44, 1050.
(12) Chavis, C.; Grodenic, F.; Imbach, J-L. Eur. J. Med. Chem.
1979, 14, 123.
5-Aminoimidazole ribonucleotide AIR 1
1-(b-D-5’-Dibenzylphosphoroyloxyribofuranose)-5-nitroimidazole
(17) (100 mg) was dissolved in EtOH (10 mL) and 5% Pd/C (100%
wt/wt) was added. This mixture was hydrogenated at atmospheric
pressure with vigorous shaking (8 h). After this time the theoretical
uptake of hydrogen (5 equiv.) was achieved. The solution was
quickly filtered through Celite and the solvent evaporated to give a
colourless solid. Analysis by NMR, and comparison with the spec-
troscopic data published by Leonard and co-workers, 5 showed that
this was mainly the required 5-aminoimidazole ribonucleotide
(AIR) 17 along with a small amount of a compound which was at-
tributed to the non-reduced nitro compound. All further attempts to
improve the procedure led to either decomposition or incomplete re-
duction.
1H NMR (D2O/TMS): d = 8.56 (1H, s, 2-H), 6.66 (1H, s, 4-H), 5.80
(1H, d J = 4.9Hz, 1’-H), 4.58 (1H, pseudo t, 2’-H), 4.42 (1H, pseudo-
t, 3’-H), 4.39 (1H, m, 4’-H), 4.00 (2H, br m, 5’-CH2); (literature
values5) (D2O) 8.59 (1H, s, 2-H), 6.78 (1H, s, 4-H), 5.89 (1H, d J =
4.8Hz, 1’-H), 4.58 (1H, m, 2’-H), 4.43 (1H, m, 3’-H), 4.37 (1H, m,
4’-H), 4.14 (2H, m, 5’-CH2).
(13) McKillop, A.; Wright, D.E.; Podmore, M.L.; Chambers, R.K.
Tetrahedron 1983, 39, 3797.
(14) Matthews, H.R.; Rapoport, H. J. Am. Chem. Soc. 1973, 95,
2297.
(15) Montgomery, J. Carbohydr. Res. 1974, 33, 184.
(16) Jones, R. H., personal communication.
(17) Saady, M.; Lebeau, L.; Mioskowski, C. Tetrahedron Lett.
1995, 36, 2239.
Acknowledgement
We wish to thank the EPSRC for the award of a studentship (to
MJH), the EPSRC Crystallography Service (Cardiff) for collection
of X-ray data, Dr Richard H. Jones (Keele) for interpretation of X-
ray data and the EPSRC Mass Spectrometry Service (Swansea) for
FAB and High Resolution spectra.
(18) Mitsunobu, O. Synthesis 1981, 1.
(19) Taylor, E.C.; Ehrhart, W.A., J. Am. Chem. Soc. 1960, 82,
3138.
(20) Chung, F.L.; Schram, K.H.; Panzica, R.P; Earl, R.A.;
Wotring, L.L.; Townsend, L.B. J. Med. Chem. 980, 23, 1158
References
(1) Schrimsher, J.L.; Schendel, F.J.; Stubbe, J. Biochemistry
1986, 25, 4356.
Schrimsher, J.L.; Schendel, F.J.; Stubbe, J.; Smith, J.M.
Biochemistry 1986, 25, 4366.
Article Identifier:
1437-210X,E;1999,0,06,0985,0992,ftx,en;P05898SS.pdf.
Synthesis 1999, No. 6, 985–992 ISSN 0039-7881 © Thieme Stuttgart · New York