Advances in DNA synthesis, which have made it possible to widely use manmade DNAs in molecular biology, genetics, virology, and other fields, represent the most important contribution to the chemistry of nucleic acids in recent years. Along with achievements in molecular cloning and nucleic acid sequencing, DNA synthesis opens up broad possibilities for future progress in biology. Synthetic DNAs and their fragments are already extensively employed in the synthesis and isolation of genes, in protein engineering, as probes in the diagnosis of infectious and genetic diseases, and, last but not least, as gene-targeted agents.

The availability of synthetic DNA fragments with a predetermined monomer sequence, which has made the above advances in molecular biology, genetic engineering and other related fields possible, was ensured by automation of their chemical synthesis and developments in purification of the preparations. The entire process of synthesizing gene fragments, assembling them into genes, cloning, selection of clones and structural analysis is now a matter of just a few weeks. Armed with powerful tools, the organic chemist is entering a new era. It has become possible to study the structure-activity relationship in proteins as well as to shape their three-dimensional structures in the course of DNA-dependent synthesis in vitro. According to predictions, new enzymes with predetermined functions will be designed and synthesized by the turn of the century.

Emphasis in this chapter will be placed on the chemical principles underlying synthesis of oligodeoxyribonucleotides.

Currently, synthesis of oligonucleotides in automatic synthesizers is becoming routine even in laboratories without specialists in synthetic organic chemistry. This places special importance on reviews and textbooks summarizing the latest knowledge about DNA synthesis not only for the benefit of students but also scientists involved in synthetic oligo(poly)nueleotide research.

Chemical synthesis of oligo- and polynucleotides represents the most exciting and challenging field of synthetic chemistry of nucleic acids. In establishing an internucleotide bond (chemical linkage of monomer units), the problem boils down to using the phosphate of one monomer unit (nucleotide component) to acylate the hydroxyl group of another unit (nucleoside component):

Enzymatic linking of monomer units involves a "directed" reaction or, to be more precise, activation of the phosphate and formation of a natural 3'-5'internucleotide bond. The enzyme "takes care" of directed activation and linkage of groups capable of intermonomer bonding. To achieve the same end by chemical synthesis requires accomplishment of at least two tasks:

(1) protection of all functional groups of the monomer, not subject to phosphorylation;

(2) activation of the phosphate group in the nucleotide or, in other words, transforming the nucleotide to a phosphorylating agent.

If these tasks are compared with the similar ones arising during peptide bond synthesis, it becomes evident that the synthesis of oligonucleotides gives rise to many more difficulties. There are basically two reasons for that:

(a) a great number of functional groups in the monomer units of nucleic acids and presence of sufficiently labile bonds in the latter; and

(b) much lower activity of the reacting groups (the hydroxyls of ribose make a much weaker nucleophile in comparison with the amino group of amino acids, whereas the phosphate group in nucleotides is a weaker electrophile than the carbonyl of amino acids).

Once an oligonucleotide has been synthesized, the internucleotide phosphate does not remain neutral. It may react, for example, with the activated phosphate present in the reaction mixture, which significantly affects the oligonucleotide yields. To circumvent this difficulty, methods have been developed to ensure internucleotide linkage not of the phosphodiester type (3) (phosphodiester method of synthesis) but of the phosphotriester type (4) (phosphotriester method).

To create natural 3'-5'-internucleotide bonds one can, in principle, resort to phosphorylation of the corresponding nucleoside components with both 5'phosphates (path I) and 3'-phosphates (path II):