Scheme 1. Hydrolysis of the N-Acylcytosine Derivatives 1
observed only in basic solution (pH ∼12). Amino-hydrolysis
h. This presumably resulted from the increased steric
hindrance and/or hydrophobic effect of the 4-N-pivalyl group.
The starting material and product pairs were soluble in H2O/
DME (2:3) solutions at 125 °C. However, the product esters
derived from the arylcarboxylic acids, 2a,b,d,e, precipitated
upon cooling and were isolated directly from the reaction
mixtures in high yields (entries 1, 2, 4, and 5). Both the
starting materials 1 and products 2 derived from the aliphatic
acids were soluble in H2O/DME solutions at ambient
temperature. The latter products 2 were isolated by evapora-
tion of volatiles and chromatography of the residues.
of cytosine to uracil derivatives proceeds slowly at most pH
values,7,8 and a patent report9 of hydrolytic deamination of
cytidine in 1 M NaOH/H2O at 100 °C attests to the severity
of the conditions required.
Bisulfite-mediated hydrolytic deaminations are accelerated
by prior addition of the sulfite nucleophile at C6 of the 5,6-
double bond.10 The latter method is useful for small-scale
deaminations but impractical for gram-scale synthesis.
The need arose in our laboratory to convert multigram
quantities of 2′-deoxycytidine into 2′-deoxyuridine. At-
tempted application of Holy’s method to 2′-deoxy-4-N-3′,5′-
di-O-tri(4-methylbenzoyl)cytidine (1b) resulted in formation
of complex mixtures. We then substituted a neutral organic
cosolvent (1,4-dioxane or DME) for Holy’s acetic acid
component and heated the solution in a pressure vessel at
125 °C. We were delighted that 2′-deoxy-3′,5′-di-O-(4-
methylbenzoyl)uridine (2b) was produced in high yield and
that 2b crystallized from the reaction mixture upon cooling.
We then investigated the generality of this hydrothermal
deamination with other cytosine nucleoside derivatives 1
(Scheme 1, Table 1).
Because the O-acetyl and O-propionyl esters underwent
slow hydrolysis under our standard conditions, the reaction
time was shortened from 12 to 9 h (entries 8 and 9). The
amides derived from aliphatic acids, 1f-i, underwent
hydrolytic attack at both C4 and the amide carbonyl carbon
to give uridine 2 and cytidine 3 derivatives (2/3, 3:2-2:1)
(entries 6, 7, 8, and 9). Compound 1j (entry 10) with the
strongly electron-withdrawing trifluoroacetyl group at N4
underwent rapid hydrolysis of the amide bond. A minor
amount of uridine derivative 2c (16%) was formed plus the
cytidine compound 3j (66%, as a trifluoroacetate salt).
Two sites are susceptible to nucleophilic attack by water.
Addition at C4 followed by elimination of an amide results
in formation of uridine derivatives 2 (Scheme 2, path 1).
Attack at the amide carbonyl carbon results in hydrolysis of
the amide linkage and formation of a carboxylic acid and
cytidine products 3 (path 2). The preference for path 1 or 2
depends primarily on the electronic effects of the group
attached to the carbonyl carbon. However, steric and/or
hydrophobic effects might also be involved. Conjugation of
the carbonyl group with an aromatic ring results in clean
elimination of an aryl amide from C4. By contrast, amides
derived from aliphatic acids undergo partitioning between
paths 1 and 2, and in the case of the trifluoroacetamide 1j,
pathway 2 is clearly preferred.
Table 1. Hydrolysis of 1 (H2O/DME/∆)a To Give 2 and 3
entryb
R
R′
X
h
2c
3c
1 (a)
2 (b)
3 (c)
4 (d)
5 (e)
6 (f)
7 (g)
8 (h)
9 (i)
Ph
MePh
Ph
MePh
ClPh
Ph
iPr
Et
Ph
MePh
Ph
MePh
ClPh
t-Bu
iPr
Et
Me
CF3
H
H
PhCO2
MePhCO2
ClPhCO2
PhCO2
iPrCO2
EtCO2
12
12
12
12
12
12
12
9
82
84
e
87
85
53
54
49
55
16
d
d
d
d
25
38
30
25
66
Me
Ph
MeCO2
PhCO2
9
1
10 (j)
a Reactions were performed by the general procedure (see Supporting
Information). b Starting materials in parentheses. c Isolated % yield relative
to starting nucleoside (acylated and then subjected to hydrolysis). d Traces
of 3 were detected in the polar residue from the mother liquor. e Reaction
was complete (1H NMR), but 2c and benzamide were not separated by
chromatography or recrystallization.
We also considered the possibility that compounds 3 might
undergo hydrolytic deamination under our reaction condi-
(8) (a) Frick, L.; Mac Neela, J. P.; Wolfenden, R. Bioorg. Chem. 1987,
15, 100-108. (b) Snider, M. J.; Gaunitz, S.; Ridgway, C.; Short, S. A.;
Wolfenden, R. Biochemistry 2000, 39, 9746-9753.
In almost all cases, reactions were complete (or near
complete) within 12 h. Starting material (18%) remained after
heating the 4-N-(trimethylacetyl) compound (entry 6) for 12
(9) Fujishima, T.; Uchida, K.; Yoshino, H. Jpn. Kokai Tokkyo Koho 1973;
Chem. Abst. 1974, 80, 3772.
(10) (a) Shapiro, R.; Servis, R. E.; Welcher, M. J. Am. Chem. Soc. 1970,
92, 422-424. (b) Hayatsu, H.; Wataya, Y.; Kai, K. J. Am. Chem. Soc. 1970,
92, 724-726. (c) Shapiro, R.; DiFate, V.; Welcher, M. J. Am. Chem. Soc.
1974, 96, 906-912.
(7) Garret, E. R.; Tsau, J. J. Pharm. Sci. 1972, 61, 1052-1061.
4904
Org. Lett., Vol. 7, No. 22, 2005