Sculpting With Biomolecules

With the help of computer-aided design and a little chemical know-how, researchers have devised new ways to make biomolecules bend to their will, creating nanoscale pictures and containers out of DNA and proteins, respectively. The techniques they developed to build these biobased nanostructures could lead to advanced functional materials.

Peng Yin, Bryan Wei, and Mingjie Dai of Wyss Institute for Biologically Inspired Engineering at Harvard University chose DNA as their biomolecule medium. They report in Nature (DOI: 10.1038/nature11075) that their basic building block, or pixel, is a 42-base single strand of the nucleic acid. Each floppy strand associates with four other strands in such a manner that the pieces of DNA bend and then stack like rigid bricks when mixed.

To create pictures, the researchers designed a self-assembling 310-pixel canvas. Depending on the shape they want to make, they simply leave out DNA from certain regions. For example, to create a ring, they omit DNA strands that make up the canvas’s center portion.

This modular pick-and-mix approach allows the researchers to create nanostructures much faster than with other DNA sculpting methods, such as DNA origami, note DNA nanotechnology experts Paul W. K. Rothemund of Caltech and Ebbe Sloth Andersen of Denmark’s Aarhus University in a commentary that accompanies the report. “This advance truly brings DNA nanotechnology into the rapid-prototyping age and enables DNA shapes to be tailored to every experiment,” they point out.

Constructing new self-assembling, nanoscale objects with DNA may be years ahead of efforts to use proteins as similar building blocks, but last week chemists working in the latter field achieved two major milestones.

Reporting in back-to-back papers in Science, two teams of researchers describe different strategies to build the first self-assembling containers made of protein subunits (DOI: 10.1126/science.1219351; 10.1126/science.1219364). Viruses commonly construct protein cages through self-assembly to house their pathogenic parts. The two teams’ work marks the first time humans have copied nature’s self-assembly strategy using synthetic protein analogs.

The 16-nm-wide, tetrahedral cage built by Yen-Ting Lai, Duilio Cascio, and Todd O. Yeates at UCLA self-assembles from 12 subunits, each consisting of two fused proteins. This container has been under development since 2001, but researchers only recently optimized the amino acid sequence so that the surfaces of the proteins could successfully interact to form a cage structure and then be visualized with X-ray crystallography.

Neil P. King and David Baker at the University of Washington, Seattle, and coworkers took a different approach. They developed a computational program to design proteins that can self-assemble into a variety of cage architectures. Then they synthesized nearly 50 of the computer program’s suggested protein sequences. Using X-ray crystallography, they observed that two of these designed proteins self-assemble into tetrahedral and octahedral cages that fit the prediction to within an angstrom.

A program that predicts protein sequences that self-assemble is a “major advance,” comments Martin Noble, a chemist at Newcastle University, in England. Most predictive software for protein design is “hit and miss, but Baker’s group has really made these proteins sit up and beg,” Noble says. “We’ve been playing catch-up with DNA. Now we can start to exploit the unique aspects of proteins.”

“Proteins are both structurally and functionally significantly more versatile than DNA and RNA as building blocks, because they are composed of 20 different chemical functionalities rather than DNA’s four,” comments F. Akif Tezcan, who studies metal-directed self-assembly at the University of California, San Diego. “That is also precisely the reason why proteins are harder to assemble into desired supramolecular structures.”