Communication
ChemComm
0
with the 2 -hydroxyl group, to give an imidate 10 which, in turn,
0
underwent addition, presumably with the 3 -hydroxyl group, to
give the amide acetal 11. The yield of 11 based on 1 was initially
quite poor, but was increased to 32% by inclusion of phosphate
buffer (6.9 eq. 5, pH = 6.9, 100 mM phosphate buffer). This
buffering effect suggests that the hydrolysis of anhydro-
nucleoside 7 occurs at a lower pH than the hydrolysis of the
des-cyano analogue 3. Inclusion of more equivalents of 5 was
dentrimental to the obtention of 11 and products containing two
or three dicyanoethylene groups were additionally observed.
Although the intermediates 7 and 9 in this cascade of reactions
Fig. 1 Crystal structure of the tetracyclic compound 11.
1
were not isolated, reaction time course H NMR spectroscopic
observations were in agreement with our proposed reaction
1
scheme. In the H NMR spectra taken 10 min and 1 h from
The potentially prebiotic reaction of ribose aminooxazoline 1
with dicyanoacetylene 5 was performed at room temperature in
aqueous solutions at pH = 6.5–7.0 and was monitored by
the beginning of the reaction (Fig. 2a and b), we observed a
0
triplet peak which we assign to the 2 -proton of anhydro-
1
nucleoside 7 at d 5.66 (J = 5.4 Hz) in good agreement with data
H NMR spectroscopy. The only product we could isolate from
2
,3
of other anhydronucleoside derivatives. Evidence for the inter-
the reaction mixture was a white crystalline compound, to which
we assigned the structure of the amide acetal 11 on the basis of
NMR data. The new structure was then verified by X-ray crystallo-
graphy (Fig. 1).‡ At first glance, the structure of 11 may appear
surprising, but by drawing an analogy to the reaction of ribose
aminooxazoline 1 and cyanoacetylene 2 (Scheme 1), a cascade of
reactions leading to 11 from 1 and dicyanoacetylene 5 can easily
be envisaged (Scheme 2). Thus, we propose that conjugate
addition of 1 to 5 gives an intermediate 6, which undergoes
spontaneous 6-exo-dig addition to form the anhydronucleoside 7 –
the 6-cyano-analogue of anhydro-a-cytidine 3. Again by comparison
to the reaction of ribose aminooxazoline 1 and cyanoacetylene 2,
the next step in the cascade leading from 1 and dicyanoacetylene 5
is expected to be the hydrolysis of the anhydronucleoside 7 induced
by an increase in the pH of the system due to protonation of the
free base form of the anhydronucleoside. Indeed, the pH of the
reaction mixture after the addition of dicyanoacetylene 5 increased
to 7.8, conditions that should be conducive to the hydrolytic
ring-opening of anhydronucleoside 7 to the lactim 8 and thence,
after tautomerisation, 6-cyano-a-cytidine derivative 9. However,
this latter compound was not isolated and we think it immedi-
mediacy of the hydrolysis product 9 was provided by a multiplet
0
0
0
at d 4.35 for the 2 - and 3 -protons (the 2 -proton signal being
upfield relative to that of the anhydronucleoside precursor 7),
0
along with a doublet at d 6.03 (J = 3.8 Hz) for the 1 -proton.
1
Although the initial H NMR spectra were quite complex, the
intermediates were eventually consumed leaving almost exclusively
1
starting material 1 and the product 11 (Fig. 2c). The H NMR
spectrum of pure 11 confirmed its presence in the crude reaction
products (Fig. 2d).
Given the unusual structure of 11, where a nitrile group is
effectively masked as an amide acetal function, and its
potential relevance to prebiotic chemistry, we were keen to
develop a synthesis which did not involve the starting material
5, because use of the latter would be hazardous on a large scale.
The idea for our alternative synthesis of 11 was to introduce
the nitrile group on the 6-position of a-cytidine 4 (Scheme 3).
A similar synthesis was followed in 1978 for the cyanation of
9
b-cytidine through a 5-bromo intermediate. The starting material,
a-cytidine 4, was easily synthesized following the steps shown in
2
Scheme 1. In the first step of the subsequent conventional synthesis
8
of 11, a-cytidine 4 was selectively acetylated on the amino group,
giving the acetyl derivative 12, by refluxing in a methanolic solution
ately underwent an intramolecular Pinner reaction, presumably
10
containing acetic anhydride. The amide 12 was then perbenzoy-
lated in a second step with benzoyl chloride yielding the tribenzoyl
ester, which was subsequently deacetylated to afford the desired free
amine 13. The latter was used for the reaction with bromine in acetic
acid giving the protected 5-bromo-a-cytidine 14. The key step of this
reaction sequence proved to be the cyanation of 14. It was found that
a slower reaction in a diluted mixture having no excess of sodium
cyanide provided the desired product 15 in the best yield, while more
concentrated reaction mixtures gave rise to many side products, thus
lowering the yield of 15. While cleaving the benzoyl esters with
sodium methoxide, 6-cyano-a-cytidine 9 was formed, but was not
isolated, as the cyano group reacted further under these conditions
to give the amide acetal product 11 identical in all respects with
material synthesised under prebiotic conditions.
In conclusion, the potentially prebiotic reaction of ribose
Scheme 2 Reaction of ribose aminooxazoline 1 with dicyanoacetylene 5. aminooxazoline 1 with dicyanoacetylene 5 was studied and
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2017