IN THE basement of Imperial College sits the London DNA Foundry. The word “foundry” calls forth images of liquid metal being poured into moulds—of the early phase of the Industrial Revolution, in other words. This foundry is, however, determinedly modern. Liquid is indeed being moved around and poured. But it is in minuscule quantities, and it is not metal. Instead, it is an aqueous suspension of the genetic codes of life.

The laboratory is an example of a wider movement. Similar biofoundries are being set up around the world, from the Broad Institute in Cambridge, Massachusetts, via Silicon Valley, to the National University of Singapore. All offer ways of centralising the donkey work of genetic-engineering research. Instead of biotechnology companies buying and operating their own laboratories, foundries will do it for them.

London DNA Foundry’s operations room is filled with boxy devices, each designed to do one particular operation, such as pipetting, repeatedly and quickly. A robotic arm shuttles small plastic dishes between the machines. Each dish contains up to 1,536 minuscule wells. And in each of those wells sits several microlitres of liquid and a few strands of DNA. Using this arrangement, the foundry can mix, in the precise concentrations required, 150,000 samples of DNA in a morning.

The starting-point for the process is a library of what David McClymont, the foundry’s head of automation, calls “parts”. These are snippets of genetic code from which different genetic “circuits” can be assembled. A circuit, in biotech speak, is a collection of genes that work together towards a common goal—for example, generating a series of enzymes that convert one type of chemical into another. The genes comprising a potential circuit are then assembled into circular DNA molecules called plasmids.

To obtain appropriate plasmids the foundry’s customers may simply order parts from the library. They may also provide their own proprietary snippets. All the required parts are then transferred to bar-coded wells in the dishes and their contents mixed automatically. The whole process is controlled by a piece of computer code, provided by the customer, that describes the experiment.

Once the mixture is correct, the test plates are whisked into a machine which multiplies the number of plasmids in each well using a process called a polymerase chain reaction (PCR). And then, when the PCR has done its work, the plasmids are introduced into living cells—either bacterial or yeast. After that, the cells are incubated, and the result is tested to see which, if any, of the circuits performs as expected.

Such is the London DNA Foundry’s scale that it can build and test about 15,000 different genetic designs in a day. True to its name, the foundry is set up to build and test DNA plasmids only. Some other biofoundries, however, offer a wider range of services. For example Transcriptic, a firm in Silicon Valley, will also allow customers to store entire cell lines that may be tested later, or to grow tissues from them for testing.

One of Transcriptic’s particular specialities is preclinical drug screening. This involves testing vast numbers of compounds that might possibly end up as drugs. A drug might, for example, be intended to alter the operation of a particular protein that shuttles, say, calcium in and out of a cell. In that case, potential drugs might be screened against both the protein in free suspension, and intact cells that have such channels in their membranes.

Little and large

Many of Transcriptic’s customers are small, newly founded firms that cannot afford their own test facilities. One such, Pliant Therapeutics, also in Silicon Valley, is using Transcriptic to test treatments for fibrotic disease—the formation of scar tissue in places like the lungs and the heart. Large, established firms use its facilities, too, though. It is often cheaper for them to do so than to carry out tests in house.

At the moment, each foundry is going its own way, as the industry finds its feet. But Paul Freemont, a director of the London DNA Foundry, hopes that once it is clear what processes are of most interest to customers, the sorts of industrial standards which are commonplace in other industries will start to emerge among biofoundries. That will make it easier for the process of designing new synthetic lifeforms to be scaled up from the bespoke boutique business it is now to something more like a global industry. That day is not yet here. But if there is demand, then biofoundries will surely play their part in the next phase of the Industrial Revolution.