IntroductionWhile the backbone organization of DNA and RNA is normally quite regular, there are many structural forms which these molecules can adapt by undergoing various secondary and tertiary helical and folded interactions, by being subjected to twisting and bending, and by interconversion between various topological and knotted variations. Probably the most common structural form of DNA in nature is a long double helix in which the ends are joined into a circle. The genomes of most lower organisms are organized in this way, 1 and clearly this circular structure lends some important advantage to these organisms or evolution would not have selected for it.Most chemists are chiefly interested in structure and reactivity of molecules smaller than entire bacterial genomes; such large circular DNAs will not be discussed herein for that reason. Of particular interest, then, is the question: what are the smallest cyclic DNAs or RNAs which exist in nature? For DNA, the answer appears to be several hundred to perhaps 1500 nucleotides (nt), which is still rather large. For RNA, probably the smallest known circular structures are the viroid RNAs, which are single-stranded and as small as 246 nucleotides in size. 2 However, this is far from the smallest possible cyclic nucleic acid structure. Duplex DNA can exist in circles at least as small as 125 bp (base pairs); 3 cyclic structures smaller than this are difficult to achieve because of the rigidity of the double helix. 4 Single-stranded DNA and RNAs do not have this problem, and rings as small as two nucleotides are known. 5 Thus, the realm of possible cyclic DNA and RNA structures falls easily into the size range most palatable to chemists.Interestingly, although small synthetic circular DNAs had been reported a number of times as early as in 1968, 6 prior to 1990 there were no reported studies investigating the effect of this quite significant structural modification on DNA's molecular recognition properties. Despite this, there was quite reasonable precedent that such a structural alteration might have a large effect on such recognition properties. Indeed, in recent decades it has become widely recognized that macrocyclic molecular structures can have strong advantages in the formation of noncovalent complexes. 7 Among these advantages (relative to noncyclic molecules of similar structure) are tighter binding affinity and greater specificity for binding the target of interest. The advent of simple synthetic approaches to the construction of oligonucleotides has led to an explosion of studies aimed at studying and modifying their noncovalent binding properties. In our early work we proposed that the principle of macrocyclic recognition could also be applied to nucleic acids in small synthetic well-defined systems. The testing of that hypothesis is the subject of this Account.
Recognition of Nucleic AcidsIn 1990 when we began our studies it was not trivial to synthesize a circular oligonucleotide from a linear precursor; however, the development of new synthetic meth...