Something startling has happened at the world's most prestigious engineering school. A vote among staff just under a month ago ushered in a brand new undergraduate course. This might be an everyday thing at a lesser institution but, at the Massachusetts Institute of Technology (MIT) in Cambridge, Boston, new courses are a distinction reserved only for the most cutting edge science or technology. The last time the institute felt moved to do this was nearly 30 years ago, in the burgeoning area of linguistics and philosophy.
The new course is, true to form, at the frontier of technology. It is a discipline that promises to open up new understanding of how life functions and, along the way, harness the stuff of life itself - DNA, cells, proteins and amino acids - to make better drugs, extract minerals from rocks, convert sunlight to hydrogen, and even design entirely new organisms that behave in pre-programmed ways. The course is in biological engineering.
Do not be fooled into thinking that this is a just a new era of biotechnology. Biological engineers will take apart the molecules of life and re-arrange them for their own purposes with a mathematical precision that has so far been unavailable to biologists. They could even programme the DNA of an organism so that it grows into your next home. In comparison, modern biotechnology, which relies heavily on trial and error and is difficult to control, is a shot in the dark.
The subject has been growing for almost a decade at MIT, its roots lying in a group of engineers who realised the floods of data coming from the genome sequencers around the world presented some very interesting opportunities.
MIT researcher Angela Belcher, for example, has been working out how to use viruses to make tiny wires for microelectronic circuits.
"Right now, the way you make microelectronic circuits or small magnetic storage devices requires very toxic chemicals, very high temperatures and pressures. There's concern about environmental effects," says Douglas Lauffenburger, head of the biological engineering division.
Belcher's viruses are engineered to express proteins and interact with the right organic and metal compounds to then turn them into long thin wires or rings; no toxic chemicals, no high temperatures - environmental issues sorted at a stroke.
"Biological engineering is not necessarily understanding systems but rather, I want to be able to design and build biological systems to perform particular applications," says Drew Endy, assistant professor in biological engineering and one of the subject's leading lights. "The scope of material I can work with is not limited to the set of things that we inherit from nature."
One of the things Endy wants to create, for example, is a counter to process information. This raises a question - are electronic counters not already plentiful and cheap? Biological engineers have no plans to simply replace existing technology. They want to come up with new uses that people might never have thought of before.
"The point of using biological [engineering] to do information processing isn't in order to replace your laptop computer," says Endy. "Instead, we can use biology-based computing to implement modest amounts of memory and logic in places where we don't have any - like the cells in your liver."
Imagine building a biological counter into a liver cell that was triggered every time the cell divided. Another biological device monitors the counter so that, if the cell has divided more than 200 times (in other words, it has lost control of cell division and might become a tumour), it is killed. This could be a very effective way to beat cancer - with none of the suffering of chemotherapy or inconvenience of surgery.
Even this simple idea is some way off, but it outlines the potential that engineers see in biology. "To a mechanical and electrical engineer, a counter is a trivial system," says Endy. "To a biological engineer, making a counter is a very challenging dream."
Slowly but surely, this dream is coming true.
MIT established the biological engineering division in 1998 as a way of taking on what it saw as tremendous opportunities in the field of biology. Other universities around the world might have been doing similar research but few focused their efforts in this way.
Lauffenburger says that ever since the discovery of the structure of DNA in 1953, scientists have known they ought to be able to develop technologies to exploit the knowledge. But it has been a long time coming.
The nascent biological engineers were disappointed that, after decades of research, most biotechnology - everything from gene manipulation to drug discovery - was imprecise.
"The technologies about manipulating biological systems to make them work better exist but they've been created out of trial and error," says Lauffenburger. "That's why they are so inefficient, why so few drugs succeed and why they're so costly. You're paying for all your mistakes along the way."
Even the relatively familiar techniques of genetic engineering are almost random in most cases. It takes lots of experiments to learn what effect a genetic manipulation has and, even then, definitive answers may not be forthcoming.
Biological engineering will be different from existing fields such as biotechnology in two crucial ways. It will use precise mathematical models to predict behaviour; and second, the applications would extend beyond medicine and health.
"I wouldn't equate genetic engineering and biological engineering in any hue," says Linda Griffiths, a professor of mechanical and biological engineering and the woman behind the new undergraduate course. "Genetic engineering came into common use as a term really just to mean any genetic manipulation of a cell. Biological engineering is way beyond that."
At the moment, biotechnologists identify interesting proteins or compounds and then chemical engineers figure out how to make them on a large scale. Just like the misnomer that is genetic engineering, there is actually no engineering of the biology going on.
Proper engineering involves starting with a precise mathematical model of a system. That model must then be tested to work out how it behaves in a range of conditions. Finally, the system must be built to exact specifications from raw materials, so that it will behave exactly as predicted.
"If you design an automobile engine, you can predict what will happen if I change the air-fuel ratio or the ignition temperature," says Lauffenburger. That capability has not been available in biology, which lacked sufficient quantitative data to build decent mathematical models.
Genome sequencing projects began to change all that and the much-needed data began flooding in.
"The science of biology is still reeling from the advent of DNA sequencing that basically started to produce tonnes of data about components," says Endy. "It became obvious that we didn't have the systematic understanding for how to put those things together."
Where biologists struggled, engineers got excited. They saw that, like chemistry and electronics decades before, the new data could plug straight into building mathematical models.
The information helped biologists to begin unpicking how biological systems worked. But for engineers like Endy, understanding is just a means to an end.
Engineers are interested solely in interacting with, modifying or applying the biology.
"Ultimately, you want to be able to manipulate those biological systems just like they manipulated the complex physics and chemistry systems. You want to be able to make a biological system operate differently than it did," says Lauffenburger.
So, if biology is to truly turn into a technology, engineers need to develop standard biological components that they can simply plug into their applications - something available in all other forms of engineering.
Some biological engineers will need to understand the atomic-level interactions between an amino acid and a part of DNA. Others just need to take it for granted that this works. Endy's answer is a system of standardisation that hides complexity from those that do not need to see it.
"We can probably engineer biology in such a way to produce components that are insulated from one another, that are designed to be easy to put together and then behave in ways that you expect," he says.
Where electrical engineers have transistors, capacitors and resistors, biological engineers have biobricks.
First thought up in 2001 by Tom Knight, an MIT computer scientist with an interest in biology, biobricks are bits of information specifying a piece of DNA and the function encoded by it.
"Biobricks are the first example of standard biological parts," says Endy. "You will be able to use biobricks to program systems that do whatever biological systems do."
Biobricks come in three flavours. "Parts" encode basic biological functions; "devices" are made from a collection of parts and encode some human-defined functions (such as logic gates in electronic circuits); and "systems" perform tasks (such as counting).
Knight made the first six biobricks but Endy has been instrumental in moving the work forward. Working with teams of students over the last few years, his biobrick database has grown to 500 basic parts. In the next few years, he expects that figure to rise to more than 10,000.
In the last few years, Endy's students have come up with cells that can flash a sequence of lights and organisms that grow in multicoloured pre-programmed colonies.
Biobricks are a natural extension of something Endy began back in his PhD days. He had created a computer model of a virus called T7, which attacks the E.coli bacterium. He modelled how the virus attacked its prey and, in the process, in which order its genes were switched on and off. The model accurately predicted the virus's known interactions but came a cropper when Endy tried to work out what would happen in unexpected scenarios - when the virus had mutated, for example, or if the genes were switched on and off in the wrong sequence.
Endy decided that the only way was to take the virus apart, look at its DNA sequence in detail, and then try and build his own version of it. He reasoned that if he could build a virus from scratch, he would fully understand the organism.
Endy not only wanted to create a version of the virus that worked in the right way but was also simple enough to modify at will. The T7 genome is a complex thing - many of the genes overlap, so carrying out an experiment on just one of them is impossible. For his synthetic virus, Endy decided to throw out anything that, by his reckoning, added complexity but not functionality.
There was no guarantee that it would work. But Endy's T7 virus, despite being heavily altered, behaved as normal. It was a crucial step.
This ability to redesign and build genomes would have been very difficult a few years ago. Reading DNA is now well-established and the genome sequences published in recent years are testament to that. Writing the sequences is harder. A recent estimate put the technology behind sequencing behind by about 10 years.
This means that synthesising strands of DNA is relatively expensive - about $2 per base pair of nucleotides (the letters of the genetic alphabet). But the price is dropping. In a few years, it could be a few cents per base pair. At 3bn base pairs, producing a full human genome would still be costly but bacterial and viral genomes (the T7 has about 60,000 base pairs) are well within reach.
For his part, Endy can order DNA sequences from specialist companies such as Blue Heron as if he were buying a book from Amazon.
The importance of DNA synthesis cannot be overestimated. "Synthesis technology is interesting because it enables biological engineering. It means you can programme DNA," says Endy. "Imagine what the science around the origin of the universe might be like if physicists could construct universes. It just so happens that in biology, the technology of synthesis [allows you to] instantly take your hypothesis and compile it into a physical instance and then test it."
There are plenty of challenges ahead. Biological engineers want to design viruses, for example, that keep the organism's useful properties (such as getting genetic material into a cell, for example) but engineer out the bad parts (such as inflammation).
Griffiths' research is a prime example. Given the shortage of human liver donors, she began looking at ways of building new livers in the lab.
But about seven years ago, she has what she called her "epiphany". Realising that the leading cause of liver failure in the western world is hepatitis C, she considered using the emerging work by biological engineers to study the virus in unprecedented detail. Maybe she could work out a way to neutralise it.
"If you look at other problems with the liver, you start to develop this whole list of other things that contribute to the need for transplant," she says. "If you pick them off one by one, by understanding liver disease better and having accessible models that can be used to test therapies, it might chop off the need for so many transplants."
This sounds fine in principle but new levels of understanding, into how viruses make their way into cells and are then transported through the intracellular traffic, would be needed.
This poses more questions. How is the activity of hundreds of simultaneous cellular processes measured? And which are the most important?
"By the laborious methods of biotechnology in the last few decades, you may be able to measure [up to] six proteins under a small set of conditions that will take you months to do," says Lauffenburger.
Again, the engineers have had to start from scratch when developing the bespoke tools for their new field.
Lauffenburger's team has developed a microscopic cantilever system, for example, where each arm is set up to detect a specific protein. Dump a batch of mixed proteins from a cell reaction onto a series of these cantilevers and, using a laser to watch the cantilevers going up and down like piano keys, precise measurements of hundreds of different molecules are disclosed.
In a field so loaded with possibilities, it is difficult for the researchers to map out the future. Lauffenburger is certain that within 50 years, the entire pharmaceutical industry will operate on an engineered basis, eliminating the need for messy trial and error methods of drug discovery.
For its part, MIT is pinning its hopes on the students on its new course. In its first year there will be just 20 places, but that number is sure to rise as universities the world over follow MIT's lead.
Further reading
mit.edu/be/
MIT's biological engineering division
mit.edu/be/people/endy.htm
Drew Endy's work in synthetic biology
parts.mit.edu/
Biobricks database
www.blueheronbio.com
Synthesise your own strands of DNA
