The history of oligonucleotide synthesis goes back over thirty years. The first report on synthesis of dinucleoside phosphate was published by Todd and Michelson in 1955. This synthesis performed by the triester method, however, was ineffective (the yield was 1-2 %) and its historical significance lies in that it is precisely this method (albeit in a new modification) that is the most widely used at present.

From the sixties onward, the most important contribution to DNA synthesis procedures must be credited to Khorana and coworkers who have elaborated the phosphodiester method. From the very outset, the method made it possible to use relatively simple condensation agents (carbodiimide and arylsulfonyl chlorides) and not only produce 10- to 12-unit oligodeoxyribonucleotides but also "assemble" them into genes. In 1970, for example, Khorana synthesized the gene of alanine tRNA. This second gene was incorporated into a bacteriophage, and its activity was demonstrated by gene engineering techniques. This work laid the foundation of DNA synthesis, and Khorana's writings are now used as standard guides both in preparation of the starting material for synthesis and in assembly of double-stranded DNAs from synthetic oligonucleotides. The historical importance of Khorana's work also lies in its having demonstrated the possibility of creating synthetic genes even before gene engineering became routine in molecular biology. In the late seventies it had already become evident that gene engineering might be instrumental in obtaining proteins with synthesized genes. This realization provided a major impetus to the development of synthetic methods. Ever since, organic chemists have been concentrating their efforts in two areas: improvement of techniques for creating internucleotide linkages in solution and solid-phase synthesis. By the end of the seventies, Saran Narang and Reese had come up with a fast and economic phosphotriester method and thereby laid the groundwork for the up-to-date procedures which, in combination with the solid-phase approach, have made it possible to automate synthesis. We should also mention the work by R. L. Letsinger (USA) who developed a solid-phase technique similar to Merrifield's method in peptide chemistry, as well as the phosphite-triester method (M. H. Caruthers, USA) to create internucleotide linkages. In the late seventies, the solid-phase method underwent further improvements proposed by M. J. Gait (UK), Z. A. Shabarova and V K. Potapov (USSR). The early eighties saw a merger of both approaches to synthesis.

This marked a new period which ended, by the mid eighties, with automation of oligonucleotide synthesis, development of several models of automatic gene fragment synthesizers (USSR, USA, UK), and a highly effective modification of the phosphoramidite-triester method of building up oligonucleotide chains. Among the later entries was the hydrophosphoryl method which allows one to automatically synthesize not only natural oligonucleotides but also those with internucleotide linkages not existing in nature.

The two principal methods built into the modern synthesizers are the phosphoramidite-triester and the hydrophosphoryl methods. However, the earlier phosphodiester and -triester methods continue to play a major role in synthetic research. For instance, when Khorana developed the phosphodiester method, techniques were elaborated for introduction of protective groups into and their removal from nucleosides and nucleotides. These techniques are still in wide use. The importance of the triester method of synthesis in solution stems from the fact that it is instrumental in producing preparative amounts (up to hundreds of mg) of oligonucleotides.

Therefore, this chapter will cover all of the above-mentioned methods. In conclusion, we shall offer a list of currently produced synthesizers and a list of genes synthesized by the late eighties.

The chapter begins with discussion of the strategies used in assembling genes from chemically synthesized fragments. The complementarity principle underlying such assembly is also applicable in all other cases where oligonucleotides serve as primers in nucleic acid biosynthesis, probes in RNA isolation, tools for diagnosis of certain sequences in DNA and RNA, gene-targeted agents. genetic engineering tools, and so on.