Bjorn Hogberg, Phd is a research fellow at the Shih Lab, Dana-Farber Cancer Institute Department of Cancer Biology at Harvard Medical School. He is also a Shapeways Community member. I was intrigued when I saw his post on our "it arrived" forum mentioning his DNA slotted cross. The "DNA origami" model that he had printed with us looked nice enough and referenced an article in Nature.
I checked out his website and asked Bjorn what it all meant, and he completely blew my mind. Seriously and without hyperbole. Bjorn and his fellow researchers are currently working on 3D printing with DNA. They are attempting to let DNA self-assemble using a technique called DNA origami. They hope to use this technique to make drug delivery systems & molecular machines.
Joris: What exactly is "building with DNA?"
Bjorn: Well, DNA in our cells is used mostly as an information storage device.
Almost 30 years ago Ned Seeman at NYU came up with the idea to use DNA
as a building material. Today, DNA nanotechnology is all about just
that. Ignoring the biological function of DNA and just treating it as
How can something self-assemble?
This is where DNA excels compared to many other molecules. Im sure
you have heard about the A,C,G and T bases that every DNA chain is
built up of. In a DNA double helix A pairs with T and G pairs with C so
if you mix for example a AAAA molecule with a TTTT molecule they will
find each-other in the solution and self assemble to form a double
A few years ago, Paul Rothemund at
Caltech, discovered that if you mix a long DNA molecule with some
200 cleverly designed short DNA molecules, these short DNA molecules
can attach to the long molecule in such a way that they fold the long
molecule to whatever shape you want. Among a bunch of other designs, he
self-assembled a nano smiley that has become quite famous. The
technology developed in our lab is an extension of this technique that
allows us to self-assemble complex three dimensional shapes.
What shapes are you able to make?
The shapes we can make are limited in size by the DNA molecules we use.
Today most object we make are around 40-50 nm long in all dimensions.
The "voxels" we use are basically small chunks of DNA double helices
about 6x2x2 nm big so the structures are also quite "pixelated". This
paper contains a small gallery of the shapes we are able to make,
some structures with curvature has also been made by Hendrik Dietz,
another post doc in our lab. They have not been published yet but his
webpage contains a preview of those curved things.
What are these structures useful for now?
Now, there is basically only one 'useful' application and that is to
make small stiff rods of DNA that provides a way to determine the
structure of membrane proteins by nuclear magnetic resonance. The
structure of this particular class of proteins is notoriously difficult
to solve so the technology fills an important gap.
What could they be used for in the future?
printing at the nanoscale for hobbyists (although you would need a
pretty expensive electron microscope to actually see what you built).
Rational design of molecular machines that could be used for example as
smart drug delivery vehicles. Circuit construction for nanoelectronics.
Is this really 3D printing with DNA?
I think it is. You could argue that anything that self-assembles isn't
really printing because there is no machine that is doing the actual
printing, its sort of a 3D powder printer in which the powder just
sticks itself together in the way you want when you shake it in a test tube.
In the Nature piece you say:"We anticipate that our strategy for self-assembling custom three-dimensional shapes will provide a general route to the manufacture of sophisticated devices bearing features on the nanometre scale." What do you mean by devices?
In that case we are talking about molecular machines. Our body is filled with wonderful little molecular nanomachines called proteins. They are very good at what they do but we basically have no idea how to re-design them to do something else. DNA nanotechnology provides us with the means to rationally design new molecular machines.
What is a molecular machine and how could it be used?
For example the ribosome, it reads the sequence of a gene, and produces a protein according to the sequence it gets. This little machine that exist in many copies in every cell on the planet is an excellent example of a molecular machine. It took some billion years to evolve by evolution. In the future, we might be able to build our own molecular machines.
Will these things have applications in biotech, medicine, beyond?
First I think we will see applications in biotech and medicine. Molecular machines that we might produce with DNA include a drug delivery vehicle, it could for example be programmed to release chemotherapeutic drugs only when it enters cells of a growing tumor. On a longer time frame, I personally believe that building schemes such as this will be used to build circuits for nanoscale molecular electronics and quantum computers.
This model I designed to use at presentations when I describe my cross design. Its basically a 3D schematic that in a stylized way shows how the long DNA chain winds its way through the entire design. I also thought it would be cool to have a macroscale version of the cross design, a 2 million times enlarged version.
Why did you print it with us?
I've been interested in 3D printing for more than 2 years. I wanted to build myself a CandyFab (http://www.candyfab.org/), but the move to a smaller apartment in Boston with a new job as a post-doc slowed those plans down. When I fund your site I though "This is exactly what I have been looking for". The fact that you upload your file and immediately get a yes or now for printable along with a price, thats almost like having your own 3D printer.
What is caDNAno?
When I first came to this lab we all designed our structures by drawing a large schematic in Adobe Illustrator, then manually transferring all the coordinates to a custom python script that would calculate the DNA sequences we needed to order. My cross design probably took like 3 weeks to design if you include the rigorous error checking we all had to do on our designs. It was a pretty error prone and time-consuming process to design a new shape. Shawn Douglas in our lab wanted a better way to design stuff so he wrote caDNAno, now using his software the design process is much less painful and the probability of making small errors is decreased. Its a great step forward in spreading the use of this technology.
In the Nature article you guys say that "we seek to demonstrate actuation potential and mechanical controllability of prestressed DNA tensegrities, to create DNA-protein chimeras that integrate defined 3D DNA nanostructures into natural extracellular matrices, and to use these artificial matrices to control mammalian cell behavior and multicellular organization by mechanically actuating physical changes in the internal DNA nanostructures." This to me reads very sci fi. Does it mean that you will 'print' a structure that will make changes in the body?
Yes, that project has an ambitious goal and it is not clear that this is exactly how it will work, but in principle it should be possible to make a DNA nanostructure that react to a chemical signal in the cells and in turn make its own changes to the cell. In fact this is already how cells work, what we could contribute is a way to make it easier to 're-program' certain cell functions. This probably lies many years ahead though so I guess you are right when you call it sci fi.