Software-controlled assembly of oligomers

Here’s a brief description of the first experimental project generated by the Ideas Factory. This is not yet an officially announced and approved EPSRC project; various administrative steps remain to be completed, including a more formal costing. However, we anticipate that this project will receive slightly less than half of the £1.5 million available for the ideas factory.

We propose to create a molecular machine that will build new materials under software control. The output of the machine will be chains of building blocks linked by covalent bonds. The machine is modular and is designed to accept many different building blocks, from small molecules to nanoparticles, with a wide range of physical and chemical properties. In order to drive its development we will concentrate on using it to create two target products: a molecular wire, capable of transporting energy and electrical charge, and a catalyst. Software control starts with specification by the end-user of a sequence of building blocks. The target sequence is encoded in an instruction tape which can be read by the machine: the tape is itself a molecule, a synthetic DNA oligomer. The target sequence of building blocks is automatically converted into a control sequence of DNA bases, and the tape is produced by commercial solid-phase synthesis. The job of the machine is to read the instruction tape and to form the bonds between building blocks in the specified sequence. Every component of this molecular factory is itself a molecule: our ambition is to develop the system to the point where it could be distributed to end users as chemicals in plastic vials.

This project was developed by a team from Oxford, Southampton, Cambridge, Exeter Queen’s Belfast and York Universities. The project leader is Andrew Turberfield (Oxford), and the main collaborators are Rachel O’Reilly (Cambridge) and Eugen Stulz (Southampton), with additional contributions from Fred Currell (Queens Belfast), Andy Tyrell (York), and Nicola King (Exeter).

Richard Jones


12 Responses to “Software-controlled assembly of oligomers”

  1. 1 Martin G. Smith January 15, 2007 at 10:16 pm

    On reading this post, two of my crew were seen galloping across the keyboards to find out what an ‘oligomer’. There was a unanimous response of, ‘What can we use them for?’ and ‘When can we get some?’
    It will be interesting to see what comes of all of this, no pressure mind you, is 2008 to soon to order?

  2. 2 Chris Phoenix January 16, 2007 at 3:25 am

    Very, very cool.

    Raises lots of questions:

    How flexible will the linkers be? (Bending and torsion)
    Can the building blocks be aligned (or allowed freedom) so as to permit RNA or DNA conjugation across joints? (Double helix, double linkage, double helix continues)
    How fast will it be?
    What error rate will it have, and is there any chance of error correction or at least rejection?
    Is there any chance for a branched structure, or only linear? (What happens if one of the building blocks contains an extra linker sticking out to the side?)
    Will any of the system’s molecules be buildable by the system from off-the-shelf parts?
    What’s the largest particle size it can use? (Optical tweezer spheres?)
    What’s the chemistry joining the linkers to the building blocks?
    Does it work in any solvents other than water?

    It’s hard to think out a particular application until these questions are answered. But almost any answers will create a VERY useful system. Not just useful for particular purposes, but general-purpose and enabling.

    I especially like that the joints are conductive. That should make it much easier to send signals to actuators. And I’m starting to speculate about a mechanically switched, electrically signaled logic circuit: stack the molecules and current can flow between them, pull on it a certain way and the current stops flowing. Maybe even build an electrostatic actuator (perhaps mediated by charge screening layers)… that could give you memory.

    Oh yes… another very important question… Will the work be patented, or public domain?

    Many thanks,
    An excited Chris

  3. 3 jim moore January 16, 2007 at 4:08 am

    So, you are going to build an Artificial Ribozome, cool.

    Let me throw in some ideas.

    Number of connectors between building two blocks,
    one – allows for rotation
    two – allows for flexing
    three – stiff
    you will want diversity of “monomers” capable of making one bond (for and end cap) up to six bonds for when you need maximum rigidity.

    I would rather have the Artificial Ribozome embedded in a micro-fluidic device. Attach the AR to something like a lipid bilayer. With the reactants, the DNA and the AR being in on one side and the “polymer” emerging on the other side of the barrier.

  4. 4 Chris Phoenix January 16, 2007 at 5:13 am

    It’s not just a ribosome. A ribosome only has a small fixed palette–the amino acids. In theory, you could add novel side chains to the aminos, then bind the new monomers to tRNA, then use a natural ribosome–though I don’t know if that would be compatible with nanoparticles due to steric problems.

    Hm… I suppose that might be how they’re actually doing it. Some of the amino side chains are pretty long; if they could build one where the side chain extended entirely outside the ribosome complex, they could attach anything they wanted to it. This would also fit well with the DNA coded input. Only thing is, I had the impression that this had been tried years ago and nothing much had come of it. But I could be wrong.

    So, one new question and two suggestions:

    How floppy is the connection between the building blocks and the linkers?

    If you are in fact using anything like amino acids, it might be worth thinking about Schafmeister’s spiro-linking technique for rigidifying the bonds post-synthesis.

    It might be nice to have a standard library of building blocks that consist of single-stranded DNA coded in various sequences, so that you can attach anything you want by binding the complementary coded strand to it. And if the library contains complementary code pairs, you could control the folding of the product. Hm… I haven’t seen this, but… if you build a complimentary code pair, and then attached the 3′ end of one strand to a linker and the 5′ end of the other to another linker, you might get a joint that required significantly more force to unzip.

    I’d guess that for making catalysts, they’d rather have something with more throughput than microfluidics–maybe bead-mounted. (It might be easier or more efficient to fasten down the molecules than make them free-floating, unless they’re actually using a ribosome that does it all in one step.) But however they’re doing it, they could probably find a way to attach it to a chamber in a microfluidic device. That could be nice for reduced reagent use, faster build time (from less diffusion), bigger palettes (after code 17 has had time to flush out of the system, you can change what code 17 means), and probably some other advantages as well.


  5. 5 Phillip Huggan January 16, 2007 at 7:18 am

    Anyone have a rough estimate for the temperature bounds of this machine? 10-50 degrees Centigrade I figure.

  6. 6 Richard Jones January 16, 2007 at 8:41 am

    I have to agree with Chris on this – this is a very, very cool project, exactly the sort of thing we hoped for; a high concept, a practical implementation strategy, and a team of outstandingly talented scientists from a whole range of different disciplines who probably would never have got together without the Ideas Factory format. If (when!) the basic idea is proved, there are a fantastic range of exciting possibilities to take the idea further. Chris has correctly identified some of the particularly beautiful features of the idea – the modularity, which should lend itself to many different types of building blocks. I should say in response to Chris’s second post, no biological ribosomes will be hurt during the making of this machine. For the moment, the team don’t want to be any more specific about the details until they get it to work. As for public domain vs patented, that’s largely up to the inventors. EPSRC policy is that IP generated from its projects is assigned to the Universities where the work is done.

    I hope you like the next project too…

  7. 7 Chris Phoenix January 16, 2007 at 10:26 am

    Wow–they’re doing it without ribosomes? That is extra-impressive. And it implies they may be engineering it de novo, which implies it can be improved as soon as it’s developed. But… *no* more details until they get it to *work*? That could be years! I guess I can’t blame them for not wanting someone else to get there first… but that would imply that a number of different labs could do this R&D.

    Can you tell us whether they’re hoping to reach the chemicals-in-vials stage by the end of the grant, or just the proof-of-concept stage?

    I’m certainly looking forward to hearing about the next project!


  8. 8 Jack Stilgoe January 16, 2007 at 10:56 am

    I’ve just added this to the Demos blog.

    I’m really looking forward to this project getting going. It’s fascinating at a number of levels, scientifically, technologically, intuitively and visually…


    A new soft machine

    As we gear up to tomorrow’s Atlas of Ideas launch, focussing on science in China, India and Korea, I’ve been thinking about some new bits of world-class British science. I spent last week in a Nano-sand-pit, working with 20 of the countries leading nano-scientists on new ways of turning information into stuff (towards a sort of mini 3D printer). The Ideas Factory blog, which over the course of the week climbed into WordPress’s top-ten, attracting over 100 comments, has just announced one of the projects that we funded.

    The concept is a molecular machine that reads a pre-coded bit of DNA, putting together chains of chemicals in whatever format we ask it to. As one participant put it, “I don’t know it it’ll work, but shit it’s beautiful.” The work will be done at Oxford, Cambridge, Southampton, Exeter, Belafst and York, and is truly multi-disciplinary – chemists, physicists and computer people.

    We’ve been following the nanotechnology debate for some time. As government turn their attention to the rather prosaic task of regulating nanoparticles, it’s worth remembering that there are some spectacular things still going on that demand a deeper set of discussions.

  9. 9 Jack Stilgoe January 16, 2007 at 11:06 am

    Minor correction to the above…

    I’ve changed the original Demos post so that it doesn’t say “funded” as there is a few more hoops through which to jump…

  10. 10 Brian Wang January 16, 2007 at 3:54 pm

    This looks like it should be an interesting attempt to create a programmable, modular and step-wise improvable DNA based construction system. It would be molecular system that would prove out a set of molecular components that would work in a construction system which would extend the complexity of what has been done before. It would leverage the improvement of DNA synthesis, nanoparticles and bring in other building blocks.

    The programmable system of pre-coded DNA string looks fairly robust and will improve as DNA synthesis improves. I think DNA synthesis can handle precise string of a thousand or so DNA bases.

    The molecular wire to transport charge and catalyst would be a way to guide site specific reactions in a more general way.

  11. 11 Martin G. Smith January 16, 2007 at 7:58 pm

    One of the dangers in being first out of the gate on a comment page is you never know what to expect next
    To Chris Phoenix – Many thanks for your always inciteful and well founded commentary.
    To Jack Stilgoe – Thank you as well for your diligence and making it out of the pit relatively unscathed.
    And to Brian Wang – The possibilities are limited to what is allowed to grow from it. There is a steep learning curve been created by all of this with my realm. I have rarely seen such focus. This is a positive thing for all.

  12. 12 Kurt9 January 17, 2007 at 11:46 pm

    This is a rather impressive idea. Even if it does not work, it will certainly teach us lots of things that can be used in other approaches. Assuming that this research effort is funded, is this the venue to follow the progress on it?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: