Nadrian C. Seeman
Department of Chemistry
New York University
New York, NY 10003
Control of structure on the nanometer scale relies on intermolecular interactions whose specificity and geometry can be treated on a predictive basis. With this criterion in mind, DNA is an extremely favorable construction medium: the sticky-ended association of DNA molecules occurs with high specificity, and it results in the formation of double-helical DNA, whose structure is well known. The use of stable branched DNA molecules permits one to make stick figures (Seeman 1982). This strategy is illustrated in Figure 9.4: on the left, we show a stable branched DNA molecule, and on the right we show such a molecule with sticky ends; four of these sticky-ended molecules are shown assembled into a quadrilateral.
We have used this strategy to construct a covalently closed DNA molecule whose helix axes have the connectivity of a cube (Chen and Seeman 1991). The cube has been fabricated in solution, which is inefficient. Therefore, we have developed a solid-support-based synthetic methodology that is much more effective. We have used this solid-support-based methodology (Zhang and Seeman 1992) to construct a molecule whose helix axes have the connectivity of a truncated octahedron (Zhang and Seeman 1994). Proof of synthesis relies on digesting target polyhedra with restriction endonucleases to generate target catenanes with characteristic electrophoretic mobilities. The cube and truncated octahedron are shown in Figure 9.5.
Our key aim is the formation of prespecified 2-D and 3-D periodic structures with defined topologies. Applications envisioned include nanomechanical devices, scaffolding for assembly of molecular electronic devices, and assembly of macromolecular-scale zeolites that orient macromolecules for diffraction studies.
The solid support methodology that has been developed (Zhang and Seeman 1992) appears capable of being used to direct the synthesis of any Platonic, Archimedean, or Catalan polyhedron; however, the construction of a closed object such as a polyhedron is a special case of nanoscale construction, because all of the edges can be specified to be made by the ligation of a finite set of sticky ends. Thus, programmability of sticky ends and a predictable local ligation product (B-DNA in the vicinity of the ligated ends) are sufficient to direct the synthesis of a finite object.
However, the goals enumerated above require the ability to construct periodic matter, which has a further requirement, structural rigidity. The assembly (Fig. 9.4) of four branched junctions into a quadrilateral is predicated on the ability of the junctions to retain their cruciform shape. If the angles between the double helical arms were variable, then a quadrilateral would be only one of many polygons that these molecules could form. In fact, branched junctions are flexible. Whereas the formation of periodic matter entails the same contacts from unit cell to unit cell, flexibility can lead to networks of variable content, and hence can destroy the periodicity of the material. For this reason, we have spent the past several years seeking DNA motifs that are more rigid than branched junctions. We have found that another structure related to recombination intermediates (the branched junction is a recombination intermediate analog) provides a rigid species. This structure is the double crossover molecule (Fu and Seeman 1993). There are five isomeric types for this system, but only the two shown below in Figure 9.6 are stable in small systems.
The double helical domains in the two molecules are antiparallel; they differ in their topologies by the number of half-turns (even/odd) between crossovers. For experimental convenience, we have worked with the DAE molecule. Our assay for stiffness is the lack of cyclization when we oligomerize a molecule containing two complementary sticky ends. Branched junctions cyclize as dimers or trimers, but we have yet to detect cyclization in double crossover molecules (Li et al. 1996), even with species as high as 27-mers. We are in the process of attaching these molecules to the sides of DNA triangles and deltahedra so that we can take advantage of the rigidity of those figures (Li et al. 1996). Figure 9.7 shows a 2-D lattice designed from such triangles and an octahedron whose three double crossover edges span 3-space.
One of the attractive features of products containing DNA is the potential of producing the molecules by biological means, either by cloning or by means of the polymerase chain reaction (PCR). Unfortunately, it is not possible to produce branched species in this way, because reproduction of the strands results in heteroduplex molecules rather than branch reproduction. Nevertheless, there is another approach that might work in this case. Figure 9.8 illustrates a pentagonal dodecahedron in a representation known as a Schlegel diagram, where the central pentagon is closest to the viewer, the outer pentagon furthest from the viewer, and the distorted pentagons at intermediate distances back; this is similar to a polar projection of the Earth, with the North Pole at the center and the South Pole at every point on the circumference. Each edge has been overlaid with two turns of DNA, and an exocyclic arm has been added to each pentagon. The exocyclic arms have been connected together to form a long knotted single strand, whose 5' and 3' ends are shown at the top.
The idea is to make the entire structure as a long strand, get it to fold, and then trim off the excess connecting DNA by the use of restriction endonucleases, leaving a DNA polyhedron ready to ligate into periodic matter (Seeman 1991).
It is clear that this is a complicated and somewhat speculative proposal, so we have tried to see how effectively we can get DNA to follow our design in folding. To this end, we have produced trefoil knots with negative nodes from DNA (Mueller, Du, and Seeman 1991) and RNA (Wang, Di Gate, and Seeman 1996). We have also produced trefoil knots with positive nodes (Du, Stollar, and Seeman 1995) and figure-8 knots (Du and Seeman 1992) (with half positive and half negative nodes) from DNA. As a further test of our control of topology, we have recently constructed Borromean rings from DNA (Mao, Sun, and Seeman 1997). These species are shown in Figure 9.9.
The last issue is the ability to produce DNA products in industrial quantities. DNA synthetic capabilities have increased substantially over the years due to solid-support-based chemistry (Caruthers 1985). There are no obstacles to the production of 100-mers in small quantities (( 200 nm). Scale-up may be a more substantial problem. Recently, it has been reported that NeXstar has developed a new methodology for the large-scale production of DNA that combines both solid-support and solution synthetic steps (Liszewski 1997).
The only other comparably oriented effort of which we are aware is the one initiated by Donald Bergstrom and colleagues at Purdue (Shi and Bergstrom 1997). They have taken two strands of DNA and hooked them together with rigid organic linkers, which are designed to take the place of the DNA at the bends of conventional DNA branched junctions. When these strands are hybridized, they form a series of cyclic molecules. There is no indication that the rigidity of the organic linkers is able to direct the angle between DNA double helices.
The work from our laboratory discussed here has been supported by grants from the Office of Naval Research and the National Institute of General Medical Sciences (NIH).
Caruthers, M.H. 1985. Gene synthesis machines. Science 230: 281-285.
Chen, J., and N.C. Seeman. 1991. Synthesis from DNA of a molecule with the connectivity of a cube. Nature 350: 631-3.
Du, S.M., and N.C. Seeman. 1992. The synthesis of a DNA knot containing both positive and negative nodes. J. Am. Chem. Soc. 114: 9652-9655.
Du, S.M., B.D. Stollar, and N.C. Seeman. 1995. A synthetic DNA molecule in three knotted topologies. J. Am. Chem. Soc. 117: 1194-1200.
Fu, T.J., and N.C. Seeman. 1993. DNA double crossover structures. Biochemistry 32: 3211-3220.
Li, X.J., X.P. Yang, J. Qi, and N.C. Seeman. 1996. Antiparallel DNA double crossover molecules as components for nanoconstruction. 1996. J. Am. Chem. Soc. 118: 6131-40.
Liszewski, K. 1997. NeXstar unveils technique for synthesizing oligonucleotides. 1997. Gen. Eng. News 17(5): 1, 10, 34.
Mao, C., W. Sun, and N.C. Seeman. Construction of Borromean rings from DNA. 1997. Nature 386: 137-8.
Mueller, J.E., S.M. Du, and N.C. Seeman. 1991. The design and synthesis of a knot from single-stranded DNA. J. Am. Chem. Soc. 113: 6306-6308.
Seeman, N.C. 1982. Nucleic acid junctions and lattices. J. Theor. Biol. 99: 237-247.
_____. 1991. The construction of 3-D stick figures from branched DNA. DNA and Cell Biology 10: 475-6.
Shi, J., and D.E. Bergstrom. 1997. Assembly of novel DNA cycles with rigid tetrahedral linkers. Angew Chem. (Int. Ed., Engl.) 36: 111-113.
Wang, H., R.J. Di Gate, and N.C. Seeman. 1996. An RNA topoisomerase. Proc. Nat. Acad. Sci. (USA) 93: 9477-9482.
Zhang, Y., and N.C. Seeman. 1992. A solid-support methodology for the construction of geometrical objects from DNA. J. Am. Chem. Soc. 114: 2656-2663.
_____. 1994. The construction of a DNA truncated octahedron. J. Am. Chem. Soc. 116: 1661-1669.