The 21st century has been tagged as the Biological Century. Why does it need robots?
The UK has opened the first of its synthetic biology ‘foundries’ – specialised labs that assemble DNA for implantation into bacteria and yeasts that can then be grown to produce advanced drugs and other chemicals.
Sometimes known as biofabs, foundries started operation at centres such as the National University of Singapore and startup Ginkgo Bioworks, a spinout from the Massachusetts Institute of Technology. They have now been joined by the Foundry. Operated by SynbiCITE, a knowledge centre funded by the UK research councils, the Foundry lies in a basement room of the biomedical and startup centre at Imperial College, London.
Following close behind the Foundry is the University of Edinburgh’s Genome Foundry, which will concentrate on assembling large DNA fragments, potentially into complete genomes of up to one million base pairs. The centre expects to take its first orders for DNA assembly this month. The university expects to complete its own flagship project, to create a chromosome for a completely synthetic yeast genome, in 2018.
As well as the UK, the USA is stepping up its investment in this area. The US Department of Energy plans to invest up to $40m by the end of 2017 to develop similar labs.
Professor Susan Rosser of the University of Edinburgh said at the IET’s synthetic biology seminar in April that the use of customised DNA to create customised bacteria and yeasts would help streamline the production of chemicals and increasingly complex pharmaceuticals and medical treatments, many of which use ‘biologics’ – biomolecules such as short protein strands.
“Some 40 per cent of drugs are now biologics,” Rosser claimed. “The ability to produce these efficiently will have a major economic impact.”
Professor Paul Freemont, co-director of Imperial’s Centre for Synthetic Biology and Innovation, claimed at the launch of the Foundry: “It’s not just a landmark for us but a landmark generally for the UK. The UK set a really ambitious target of having a £10bn synthetic biology industry by 2030 and we see SynbiCITE as being a major part of that. The Foundry sits at the interface between academic research and translation at industrial companies. SynbiCITE is already working with close to 30 SMEs, spinouts, entrepreneurships and larger companies as well.”
The key to the rise of the biofoundry is robotic automation. In Imperial’s Foundry, a group of robots prepare and mix chemicals in well plates: trays that contain close to 400 individual samples so that many preparations can be carried out in parallel. The machines assemble fragments of DNA into loops that are inserted into living cells. The new DNA alters the behaviour of the cells, often simple bacteria, so that they produce drugs and other chemicals.
The DNA fragments can be isolated from existing species. For example, sea sponges are being investigated as possible sources of future antibiotics. Or the DNA can be synthesised chemically from scratch by specialist contractors. “You just send off an email with your sequence and get the DNA samples back by post,” said Freemont’s co-director Professor Richard Kitney.
Although the core chemistry behind the process has not changed significantly since the 1980s, according to Freemont, heavy use of automation has allowed the cost of DNA synthesis from scratch to fall over the past couple of decades. The falls are not as rapid as those in gene sequencing. But synthetic biology developers are able to use DNA assembly from ready-made fragments to minimise overall production costs.
Much of the work in synthetic biology revolves around identifying and applying promoters, segments of DNA that sit next to genes within the loop of DNA that makes up a genome. “The promoter is one of the key tools in engineering biology,” Rosser explained. “They switch genes on and off.”
Active genes create proteins, with each gene responsible for the production of a specific type of protein. Certain proteins will bind to specific promoters as conditions within the cell change. Usually, the binding of a protein switches the adjacent gene off. In this way, cells dynamically regulate the production of proteins, many of which process and metabolise other chemicals. The interaction between proteins and promoters also provides a way to build logic circuits that can be harnessed to deliver smart medicines that target specific cells within the body.
The key problem for designers of these biological logic circuits is testing them. Freemont said: “Constructing biology is not plug and play. The parts we are using today are not characterised well enough to do predictive modelling and design.”
Freemont added that the characterisation experiments needed are generally highly laborious. Based on a design developed by researchers working with Geoff Baldwin at Imperial, roughly half of the Foundry is dedicated to characterisation and testing. The robots perform customised experiments for each customer’s collection of samples. Freemont added Imperial researchers are now working on their second-generation characterisation robot, which will help close what they call the “design, build, test, learn” cycle. “This will get us to efficient designs much more quickly,” he said.
Some manual intervention is still needed in the Foundry at Imperial. The well plates currently need to be transferred by hand from the manufacturing robot to the characterisation equipment, but Kitney said the aim is to install conveyors to automate that – an approach similar to that now commonly used in semiconductor fabs.
“In its current configuration, the Foundry can perform around 17,000 DNA assemblies and a thousand characterisation experiments each day,” Freemont claimed. “This is just a start. The Foundry’s design is scalable and modular. Imagine a whole building full of these machines and how many experiments we could do.”
As well as robotics, other concepts from electrical engineering are being harnessed by the biological sciences. Freemont pointed to what he called a breakthrough paper published at the end of March by Christopher Voight’s group at the Massachusetts Institute of Technology (MIT). This uses the Verilog language that was developed for chip design as the basis for a system that automatically synthesises the promoters and genes needed to build DNA sequences that perform logical operations.
Although Verilog underpinned a concept proposed by researchers at the Virginia Institute of Technology in the late 2000s, Doug Densmore and colleagues from the Voight Lab were able to show the technique working in practice for a variety of simple logic circuits based on the results of the generated DNA after being inserted into target cells. “The paper is an absolute tour de force,” said Freemont. “It’s the first time that this idea of automating the design has been shown to work.”
New age, new challenges
Kitney commented: “Engineering biology is a key area for the development of knowledge-based industries. If the 20th century was the digital age, the 21st century will be the biological age.” However, legal and commercial issues threaten to slow down progress. A fight between two US universities over the intellectual property rights for the CRISPR-Cas9 process that performs precisely targeted gene editing could lead to teams avoiding its use for fear of being sued themselves for patent infringement, Freemont warned.
Rick Johnson, CEO of the consultancy Global Helix, pointed to another problem – funding for biological engineering. Commercialisation and scaling to volume production will challenge conventional venture funding models, he said. “There are policy and other issues that are different from other areas of technology. We are looking at new tools to derisk some of these investments.”