Nucleic Acids in Chemistry and Biology

sugar moiety. Third, a double disconnection C shows the formation of a purine base onto a preformed imi-
dazole ribonucleoside. We shall now explore each of these three routes in turn.

3.1.1 Formation of the Glycosylic Bond

The synthesis of nucleosides through glycosyl bond formation should ideally address both stereoselectivity
(formation of nucleosides with the natural
-configuration at C-1) and regioselectivity (glycosylation of
pyrimidines at N-1 and purines at N-9). There are essentially three methods that are used: (a) metal salt proce-
dures, (b) silyl base procedures, and (c) fusion reactions together with various modifications of these. The first
two methods are generally more widely applicable and used most frequently. While the following sections
describe methods that may be used for the preparation of nucleosides, these mainly refer to ribonucleosides.
Although such methods may be extended to the syntheses of 2-deoxyribonucleosides, they often give poor
stereoselectivity during glycosyl bond formation. Modifications to these methods that are more suited to the
preparation of 2-deoxyribonucleosides have been developed and are dealt with later in the chapter.

3.1.1.1 Heavy Metal Salts of Bases. Fischer and Helferich,^4 and Koenigs and Knorr introduced the

use of a heavy metal salt (initially silver(I)) of a purine to catalyse the nucleophilic displacement of a halo-
gen substituent from C-1 of a protected sugar. In the late 1940s, Todd’s group adapted this chemistry to
achieve a synthesis of adenosine^5 and guanosine^6 following an initial glycosylation between a protected
1-bromo-ribofuranose derivative and 2,8-dichloroadenine. In a later modification, Davoll and Lowy used
mercury(II) salts to improve the yields of products.^7 Typically, chloromercuri-6-benzamidopurine reacts
with 2,3,5-tri-O-acetyl-D-ribofuranosyl chloride or bromide to give a protected nucleoside from which
adenosine is obtained by removal of the protecting groups (Figure 3.3). These syntheses almost invariably
gave the desired stereoselectivity, predominantly providing the

-anomerat C-1 of ribose owing to the
formation of an intermediate acyloxonium ionby the sugar component (see Section 3.1.1.7). The
chloromercuri salts of a range of purines can be used, provided the nucleophilic substituents are protected.
Thus, amino groups have to be protected by acylation, as shown in a synthesis of guanine nucleosides
using 2-acetamido-6-chloropurine followed by appropriate hydrolysis (Figure 3.3).
The chloromercuri derivatives of suitable pyrimidines can be used in much the same way as illustrated
by a synthesis of cytidine from 4-ethoxypyrimidine-2-one (Figure 3.4).^8 While this type of glycosylation gives

Nucleosides and Nucleotides 79

N

N N

N

O

NHBz

Br

AcO

AcO OAc

HgCl O

AcO

AcO OAc

N

N

N

N

NHBz

O

HO

HO OH

N

N

N

N

NH 2

N

N N

N

Cl

HgCl

NHAc

O

AcO

AcO OAc

N

N

N

N

Cl

NHAc

O

HO

HO OH

N

NH

N

N

O

NH 2

(i)

(i)

(ii)

(iii)

Figure 3.3 Chloromercuri route for synthesis of purine nucleosides. Reagents: (i) xylene, 120°C; (ii) NH 3 , MeOH;
and (iii) NaOH aq