December 5, 2020

I, Science

The science magazine of Imperial College

Using analogue computation could ramp up the processing power of synthetic life ...


Every year teams of scientists gather in Cambridge, Massachusetts, to build machines from synthetic life. So far the International Genetically Engineered Machine (iGEM) competition has produced tens of different synthetic cells that can do useful jobs, such as sniffing out arsenic in drinking water. Yet new research suggests these biosynthetic implements are not as smart as they could be.

Even the most impressive examples of synthetic biology currently process information digitally. In the arsenic-sensing cells, for example, a protein that pumps out a molecular alarm signal is switched on when the toxin concentration reaches a certain limit. Above the threshold the toxin is a ‘1’ and below it is a ‘0’. But this digitisation ignores lots of the information in biology; rather like knowing the radio is on but not the position of the volume dial.

Now scientists have built synthetic cells that can calculate using analogue information processing, harnessing that extra information. The team led by Tim Lu, an Assistant Professor in biological engineering at Massachusetts Institute of Technology, began by adding circular loops of DNA to engineered E. coli cells. These loops contained instructions for making a light-emitting protein, among other things. The amount of protein (and thus light) produced is modulated by the interactions between the loops, as determined by the concentration of a simple sugar called arabinose. The intensity of the light could be notched up or down by altering the sugar concentration.

The team then tweaked these prototype genetic circuits in order to program the cells to do maths. They could add up chemical concentrations, multiply them and even find the ratio of between them. In each case the answer to the calculation could be read out from the intensity of the light the cells produced.

The secret to the group’s success is working in harmony with biology, says Guy-Bart Stan, who lectures in synthetic biology at Imperial College London. “We have to remember there is no digital in the real world,” he says. That means synthetic biologists who have developed digital cellular calculators had to shoehorn them into it; laboriously encoding information about chemical concentrations in binary language before the cells could understand it. A recently reported synthetic cell that could digitally calculate square roots needed 130 strings of DNA. Dr Lu’s system has just two pieces.

The simplicity of the system will mean a wider range of people, such as citizen scientists, will be able to experiment with synthetic biology. The building blocks the team used are widely available and easy to put together. Dr Lu has made his DNA sequence information available online and says a high school student who has access to a shared DNA synthesiser could replicate his experiments. “This is a case of plug and play,” he says.

Now there is a way to perform calculations with just a few pieces of DNA, more complex decision-making circuits could be programmed into cells. By changing the DNA sequence the output protein could be varied. One potential example is a diabetes management system. This would measure the ratio of the sugar-controlling hormones insulin and glucagon and then synthesise more of whichever is in short supply. Electrical implants that do this job are available, but – tellingly – they are plagued by a build-up of biological debris on their surfaces and need replacing regularly. Synthetic biological devices wouldn’t suffer this drawback.

Whatever their use, the ultimate benefit of synthetic cells is that they would be self-replicating, meaning that once perfected, they could be produced at negligible cost. “We are only limited by the number of different proteins we can code for – and our imaginations,” says Dr Stan. At any rate, this summer’s iGEM competitors will have another programming tool in their arsenal.


IMAGE: Flattop341, flickr