Menu Search Site

Programming molecular robots

A Q&A with Wyss Core Faculty members William Shih and Peng Yin

In the late 1990s, a small group of bioengineers set out to turn cells into tiny robots. Being bioengineers, they drew ideas from engineering, and envisioned building a set of modular, standard parts akin to the sensors, power source, microprocessor and actuators that enable robots to sense and respond to their surroundings. Those early efforts spurred a wave of optimism about the incredible potential of synthetic biology.

But getting even simple organisms to carry out the right tasks at the right time remains a formidable task, and synthetic biologists are still a long way from making living cells obey their commands. In part that’s because they only partially understand the workings of the cells’ operating systems – their genes and their regulatory networks. So a new contingent of bioengineers is pioneering a different approach. Instead of trying to program living cells, they’re using the cell’s information-carrying molecules — DNA, RNA and protein — to build their own operating systems, their own sensors, their own actuators. Their goal: to build tiny molecular robots.

This past fall the National Science Foundation (NSF) kicked in $10 million from an NSF initiative called Expeditions in Computing to fund 11 scientists, including Wyss Institute Core Faculty member William Shih, Ph.D., Core Faculty member Peng Yin, Ph.D., and former Wyss Institute postdoctoral fellow Shawn Douglas, Ph.D., who is now an Assistant Professor of Cellular and Molecular Pharmacology at the University of California, San Francisco. This prestigious grant is designed to launch the nascent field of molecular programming.

Wyss Institute Core Faculty member, Peng Yin, Ph.D.

We sat down with Shih, who is also an Associate Professor of Biological Chemistry and Molecular Pharmacology at Harvard Medical School and Associate Professor of Cancer Biology at the Dana-Farber Cancer Institute, and Yin, who’s also an Assistant Professor of Systems Biology at Harvard Medical School, to learn the ins and outs of this new branch of engineering, which could revolutionize fields as diverse as information technology, tissue engineering, and manufacturing.

What is molecular programming?

Yin: It means directing information-carrying polymers [DNA, RNA, and protein] to demonstrate prescribed molecular behavior. Using these as tools, we want to develop real-world applications. For example, we are using nanoscale DNA structures as a template to fabricate inorganic materials. We are also building programmable molecular machines made of DNA or RNA to help us study biology.

We have very fast computers these days. Why is it important to build programmable devices out of DNA, RNA and proteins?

Shih: If molecular programs could only do an extremely slow job of what silicon computers can already do very efficiently, they would only be a curiosity. For crunching millions of numbers, they will never compete with these machines [he points to his desktop computer].

I can think of at least two practical reasons for molecular programming. First, sometimes you need your computer to talk to cells, talk to molecules. You need your computer to diagnose a disease. You don’t need to do billions of calculations per second – maybe just ten in the course of an hour. But you need your computer to go to small places in your body and be biocompatible.

The second reason is that you can create massively parallel armies of not very good computers. For example, a traditional strategy for drug discovery entails taking a molecule that works, then tinkering with it and making it work better. A second strategy that has not worked well so far is to make a billion drugs and screen them and build off that. But instead of making a random library of new drugs, what if we had some way to build in logic? You could create a library that is more streamlined, with more of the things you want.

Yin: The critical thing for us is that we’re not trying to use these molecules to merely do computing. We are trying to use them to perform programmable molecular tasks. For example, we can program how they physically interact with each other. As information-carrying molecules, [DNA, RNA and proteins] can specify spatial and dynamic behavior.

One example is digital fabrication. We use a programmable DNA structure to specify the nanoscale 2D and 3D geometric shape. And we can translate from those 2D and 3D shapes to other inorganic substrates. For example, we can translate the DNA shape to a gold particle and later to graphene to create a substrate for nanoelectronics. We’re going from programmable biomolecules and using them to fabricate inorganic devices with a digitally programmable shape.

Wyss Institute Founding Core Faculty member, William Shih, Ph.D.

You and your colleagues write in your grant application that “with molecular programming, chemistry will become the new information technology of the 21st century.” How so?

Shih: It’s not going to be the CPU in smart phones, or here, or there (points to his screen, and to his computer).

But if what you want is ubiquitous computing — trillions of sensors all over the body —maybe with molecular computers, you could do it. Or you could disperse sensors all over a lake, or the atmosphere. Collectively they’re doing something you can’t do with conventional silicon. It’s a new frontier for computing.

What are real-world problems that molecular programming could help solve?

Yin: Molecular programming could lend molecular precision to diagnostics or therapy. Life at its finest scale can be visualized as dynamic, self-assembling molecular systems. To interact with this molecular system, you want instruments on the same scale. You’d have digitally programmed instruments on the molecular scale. For example, there could be a genetically encodable RNA machine to identify cancer cells and kill them. Or you could have tiny fluorescent DNA probes that interact with cellular components to produce super sharp images.

Digital fabrication is also extremely exciting. Using DNA structures as templates, you could specify in a digitally precise fashion the architecture and potentially the function of a nanoelectronic circuit or nanophotonic device.

We’re also collaborating with Pam Silver’s group to assemble RNA bricks into an RNA scaffold to spatially organize enzymes involved in hydrogen production in E. coli. Using an earlier RNA scaffold, Pam’s lab could increase hydrogen production by 50-fold. [Pam Silver is a founding core faculty member of the Wyss Institute and Professor of Systems Biology at Harvard Medical School.]

William, you have developed DNA origami and used it to fold DNA strands into a bridge with handrails and other shapes. Is your molecular programming work a continuation of your DNA origami work?

Shih: One connection is that the way people have done DNA computation so far is to have lots of free-floating strands of DNA in a solution. They sometimes bump into each other, but the process is often slow. To catalyze the reaction, what if you sequester strands in a compartment, or walking along surface. For example, you could make a nanoscale box and put a small number of strands into the box. The effective concentration could be up to a million times higher.

What do you envision as the future of molecular programming?

Yin: My mom was a computer engineer in the early days of punch cards. Now I can just type in Java and C and I can program my computer. We want to do that for these molecules that carry digital information. We want a user-friendly programming interface and an associated “molecular compiler,” so that everyone can do molecular programming.

I don’t see why in the future we couldn’t imagine molecular apps and build the infrastructure to make them programmable, expressive, robust, and functional. Then many other users could really develop their own apps. [Yin pulls out his iPhone and points to some apps.] In 10 or 20 years freshmen at Harvard could learn a molecular programming language to program how molecules behave and generate their own molecular apps. That would be fascinating.

These information-bearing molecules are merging information technology with molecular technology to enable human beings to realize our functional needs in the molecular world. This vision, I think, is an emerging revolution.

Close search results
Close menu