I am a scientist in the
laboratory of professor Erik Winfree at Caltech. It is a wonderful place
to be---we get paid to play with DNA and really, there is no better
company than a bunch of nerdy scientists. A major focus of the lab is
algorithmic self-assembly, which you can learn about from numerous
papers .
Lately, I have developed a method of creating nanoscale
shapes and patterns using DNA. Each of the two smiley faces above, at
right, are actually giant DNA complexes imaged with an atomic force
microscope. Each is about 100 nanometers across (1/1000th the width of
a human hair), 2 nanometers thick, and each is comprised of about
14,000 DNA bases. 7000 of these DNA bases belong to a long single
strand, a DNA molecule that just happens to be the genome of the virus
M13. The other 7000 of these bases belong to about 250 shorter
strands, each about 30 bases long. These short strands fold the long
strand into the smiley face shape. I call the method "scaffolded DNA
origami". (For the record: there is no fundamental significance to the
fact that it is viral DNA; I could buy it and it was cheap and
pure. M13 is a bacteriophage---it can make the bacteria in your
intestine sick but not you!)
While the smiley face shape is somewhat silly DNA artwork, it is
a high technology artifact and there is serious science behind it. We
hope to use the technique of DNA origami (as well as many other
techniques of DNA nanotechnology) to build smaller, faster computers
and many other devices.
The best way to understand how this works is to read the
original article published in Nature .
There are two supplemental files associated with this paper. The
first (82 pages, 6.3 megabytes)
describes the design method, block diagrams for the designs,
sequences for all of the designs, and includes data on control
experiments. This file is likely to be of most interest. The second
(9 pages, 192 kilobytes) includes diagrams for all the
designs that have the full sequence written out in the diagrams. This
file is useful if one wishes to check details of the design or modify
them. It cannot be printed clearly and is best viewed with a PDF
viewer on the screen. Please email me with your PDF viewer name and
version if either of these files do not work for you.
I apologize that my MATLAB code for
the design of origami structures is not ready for primetime. This
should be remedied in "a couple of weeks." To run it you will need to
buy a copy of MATLAB from the
Mathworks . Student versions of Matlab are available for $100. I
apologize for using a commercial programming environment. If you wish
me to send you email when the code is ready then send me an email with
the subject line "ORIGAMI CODE REQUEST".
Once the basic design for a shape has been completed, one can add a
surface pattern on top of the shape. For demonstration purposes in the
Nature paper, the surface pattern was also made out of DNA although in
principle it could be rendered using a chemically, electronically, or
optically interesting material. Because each staple occurs in a
unique location in the origami shape, this location can serve as a
pixel. The normal DNA staple can represent a '0' or a flat pixel. A
modified DNA staple with an extra DNA bump (a DNA hairpin) can be used
to represent a '1'. This is what has been done for the first three
images below, the map, some snowflakes, and the word 'DNA' rendered
above a representation of the double helix. Each of these patterns has
been applied on a roughly 100 nanometer wide rectangle and each pixel
is roughly 6 nanometers in size. Because there are about 200 staples,
each pattern can have 200 pixels. Two copies of the origami have stuck
together in the third image to make the helix appear
continuous. Errors occasionally occur as in the lefthand 'D' in the
first 'DNA' at right.
The structure below is a hexagon that is actually built up of six origami triangles. Each triangle
has a pattern added to it (the green bumps) so that orientation of the triangles in the hexagon can be ascertained.
All of the images shown here are actual atomic force microscope (AFM)
data of DNA molecules (smoothed to remove noise). An AFM measures surface topography. Here,
false color is used to indicate height. The height of a basic shape
(say the smileys) is just the 2 nanometer thickness of DNA. Wherever a
DNA bump has been added in a pattern the surface height is 4
nanometers. The height contrast has been exaggerated (with respect to
the lateral dimensions) to emphasize the topography.
Importantly, there is a long list of people who do work on DNA
nanotechnology. Among them the inventor of DNA
nanotechnology
Ned Seeman at NYU, William Shih at
Harvard, Milan Stojanovic and Darko Stefanovic at Columbia and UNM,
Hao Yan at Arizona State University, Thom LaBean at Duke,
Andrew Turberfield at Oxford
and many others.
Also, DNA nanotechnology encompasses much more than just making shapes
as shown above. It includes extended DNA sheets, DNA tubes, DNA
machines, DNA walkers, etc.
email: pwkr@dna.caltech.edu
Rendered by Nick Papadakis, Copyright P.W.K.R. and N.P.