Polyamide struts for DNA architectures.

Owing to its well-established synthetic access, its well-known structural properties, and its size, DNA is a very interesting material for building nanoarchitectures in a bottom-up approach. This has already been proven in a number of studies. For example, cubes, octahedra, Borromean rings, as well as tubes (nanowires) have been built with DNA. However, DNA can not only be used as a structural material on a nanometer scale but also as a functional material, for example, to measure temperature force. However, as DNA is a linear polymer, it is difficult to build stable, branched three-dimensional architectures. Branched DNA duplexes offer a possible solution, but complex “tiles” have to be used for stable structures. Other approaches, for example, rely on the use of synthetic DNA derivatives to build trisoligonucleotidyl junctions. Polyamides derived from nonproteinogenic amino acids containing N-methlypyrrole, N-methylimidazole, or N-methylhydroxypyrrole groups can bind in a hairpin motif to the minor groove of double-stranded (ds) DNA. The base sequence to which they can bind selectively and with high affinity can be programmed by the sequence of the respective residues in the two antiparallel strands of the hairpin polyamide. A complete set of pairing rules is available for the straight-forward recognition of the majority of possible DNA sequences. DNA-binding polyamides of this type can be used for a variety of applications, like, for example, gene regulation or as sequence-selective nuclear stains. As most of the names for this class of compound are either lengthy or imprecise, we propose the term “Dervan-type polyamide” in honor of P. B. Dervan who made the major contributions in this field. Herein, we report on the use of Dervan-type polyamides as a possible second orthogonal structural element—besides Watson–Crick base pairing—for DNA-based architectures. Our rationale is that by connecting two hairpin polyamides with a long and flexible linker, we generate “DNA struts” that target their matching binding sites on different DNA duplex strands and can thus function as “sequence-selective glue”, which will make it easier to build up complex stable architectures. Only once before has a Dervan polyamide been shown to bind across twoDNAduplexes, but in this case, the two duplexes were preorganized next to each other in the nucleosome core particle. To test whether Dervan polyamides alone can be used to hold DNA objects together, we synthesized the heterobifunctional strut S-AB shown in Figure 1. It consists of the two