Synthetic oligonucleotides as therapeutic agents.

Currently there is much excitement in certain quarters about antisense oligonucleotides, both as the drugs of the future for treatment of cancer and viral infections, most notably AIDS, and also as genetic tools for generating mutant phenocopies without genomic mutation in investigations of mammalian gene function. The extent of this excitement is reflected in the proliferation of antisense biotechnology companies, newly formed with venture capital in eager anticipation of the imminent bonanza (Klausner, 1990). It is remarkable how antisense oligonucleotide technology has received publicity in excess of its present achievements, and there are some, not directly involved in this area, who have gained the impression of an established technique, like polymerase chain reaction or DNA fingerprinting, which may be taken off the shelf and used as necessity demands. Without wishing to appear unduly negative, it must be stated that, as yet, antisense oligonucleotides have not arrived (Tidd, 1990). The concept of antisense inhibition of gene expression has gained respectability largely through the elegant efforts of the molecular geneticists. They have demonstrated that production of antisense RNA may be a naturally occurring genetic control mechanism in prokaryotes. Also, antisense RNA was shown to induce sequence specific inhibition of target gene expression when introduced artificially into eukaryotic cells, either by microinjection of preformed antisense molecules, or through transfection with DNA plasmid constructs carrying genes inverted relative to an appropriate transcriptional promoter (Melton, 1988). It is generally assumed that the antisense RNA hybridises with the complementary mRNA to form a double stranded helix analogous to double stranded DNA, and in so doing inhibits nuclear processing of the immature mRNA, transport into the cytoplasm, cytoplasmic translation into protein or, alternatively, stimulates degradation of the message. Despite the impressive results of antisense RNA experiments, the actual intracellular mechanism responsible for the observed effects remains to be determined. The therapeutic implications of antisense technology are compelling, in that inhibition of expression of essential viral genes should selectively curtail viral replication, whereas functional elimination of the appropriate activated oncogenes in a cancer cell would be expected to result in the instant reversal of the process of malignant progession, and might even trigger entry into a terminal differentiation pathway. However, the antisense RNA approach suffers from several shortcomings, namely that the inverted gene constructs are produced in very small amounts by rather laborious molecular genetic techniques, and the antisense RNA is biosynthesised as if it were an mRNA, where long sequences are generally the rule. Long antisense RNA sequences will be insensitive to short mismatches between a normal and mutant allele, and, consequently, will be of little use in discriminating between, for example, an activated ras oncogene and the normal cellular proto-oncogene. In addition, the mere thought of treating patients with recombinant attenuated viruses carrying inverted ras genes may be suffic-