Synthetic biology is the next generation of genomics, bringing a more engineered, design-driven approach to the development of therapeutics and diagnostics. As synthetic biology matures, scientists are finding that they can program cells rather like computers, make DNA at scale, deliver RNA to non-liver targets with relative ease, and much more.

Just as important, synthetic biology is beginning to produce returns for investors. “We’re in version 2.0 now, where the technology is applied to the highest value products, [which means] that companies have a chance of making profits,” declares Emily Leproust, PhD, CEO of Twist Bioscience. Next, we will see organizations turbocharging a few areas for big changes.”

Before those big changes can be achieved, however, synthetic biology must resolve a few mechanical issues. “We need to do a better job inferring the design principles for biology and understanding how biomolecular components interact with each other in synthetic gene circuits and with the host,” admits James J. Collins, PhD, Termeer Professor of Medical Engineering and Science, Massachusetts Institute of Technology (MIT). “We also remain in significant need of a greater number of molecular parts to create synthetic biology devices.”

RNA switches for programmable cells

Collins’ laboratory is working to increase the number of available parts by using machine learning to design RNA switches for diagnostic platforms, and by engineering synthetic gene circuits and programmable cells in bacteria. The benefits of a machine learning/artificial intelligence approach to molecular parts discovery and development include a better understanding of the function of natural biological systems and components, and the recognition that many of these systems and components can be repurposed for biomedical and biotechnological applications.

“CRISPR is an exciting example,” Collins points out. “It is exciting to think there are additional related and unrelated mechanisms to be discovered.” His immediate goals for the development of RNA switches are to advance toward biomedical applications and to better integrate artificial intelligence with synthetic biology.

Collins also is working on synthetic gene circuits and programmable cells. For example, he is engineering bacteria to allow them to “produce an enzyme to break down a drug, or a molecule to keep a disease in check.” Work of this kind is being pursued at Collins’ MIT laboratory, as well as at Synlogic, a company that Collins co-founded. Synlogic develops “Synthetic Biotics.” Several are currently in clinical trials, including Synthetic Biotics to treat solid tumors.

“Biology is not yet an engineering discipline,” Collins points out. “That makes it exciting, because a lot of work remains to expand the necessary tools and parts lists.” He suggests that scientists will eventually read intracellular activity in real time and use desktop machines to synthesize DNA rapidly and economically.

Treating DNA code like computer code

Ginkgo Bioworks develops cellular programs like software companies develop computer programs—except that Ginkgo works with DNA bases rather than zeros and ones. The idea is to give cells modular, reusable code so that they can produce anything from therapeutics to synthetic food.

“DNA is a digital code that can be written and then printed out using DNA synthesis,” stresses Jason Kelly, PhD, co-founder and CEO of Ginkgo. “The tools are the same to program a cell, regardless of whether the result is a therapeutic, chemical, or food.”

Kelly and Ginkgo’s other co-founders—including Tom Knight, a senior researcher at MIT and the “godfather of synthetic biology”—intend to realize a vision: a synthetic biology company that works like an information technology (IT) company. Accordingly, Ginkgo uses a horizontal platform, an operating system, and languages for cell programming. The company’s co-founders believe that, just as one needn’t be a computer engineer to program a computer, one needn’t be a biologist to program a cell.

Two projects are of particular interest to Kelly. One bridges the gaps in the nucleic acid vaccine supply chain and is vital to the COVID-19 response. “The key enzymes to make nucleic acid–based drugs are in short supply,” he notes. “Until the pandemic, they were mainly used for R&D.” Scarce enzymes, he adds, include capping enzymes.

The other project is Ginkgo’s antibiotic discovery program with Roche. In this program, Ginkgo’s work involves using a genome mining program (acquired from Revolutions Medicine’s Warp Drive Bio division) to dig through codebases spanning more than 135,000 bacterial strains while “looking for interesting sets of genes,” Kelly says. “Screening the genomes is a new way to hunt compounds.”

The codebases include genetic codes developed specifically for Ginkgo’s clients. “We retain the right to reuse the genetic code,” Kelly points out. Reuse is practical because the genetic code is modular (just like IT code), reducing project times and costs for each partner.

Twice the DNA in half the time

Twist Bioscience is halving the time needed to make DNA so that large quantities can be synthesized in about a week. By synthesizing DNA more quickly, Twist hopes to contribute to the development of personalized medicine—a kind of medicine in which “everyone gets their own drug for their own mutations,” Leproust says. She anticipates that personalized medicine will “make diseases like cancer chronic diseases.”

A patient’s cancer could be controlled by treating it with a new drug each time it mutates. This approach is feasible, Leproust points out, because “we’ll be able to sequence the cancer as it evolves.”

To turbocharge DNA synthesis yet further, Twist Bioscience continues to miniaturize its silicon chip. “We can make 1 million oligos per chip now,” Leproust asserts. “In a few years, we will be able to make trillions of oligos, skipping the billion-oligo stage entirely.

Twist Bioscience lab
The silicon chip shown in this image is used by Twist Bioscience to “write” DNA at commercial scale. One chip of this size makes about 1 million oligonucleotides. The company indicates that it continues to miniaturize DNA synthesis, and that in a few years, it will be able to make trillions of oligonucleotides per chip.

“We have 1 million features on a 50-micron-well chip now. That’s the size of a human hair, but it is the Grand Canyon in ‘silicon land.’ We are shrinking the dimension of that feature by 1,000 times to create 1 million times more oligos. Decreasing the feature size by a factor of 10 increases the number of features by a square, so making a well 10 times smaller creates 100 times more features.”

The challenge is in the testing. “If I make one piece of DNA, it’s easy to say whether it’s good or bad,” Leproust notes. “Now that we’re going to 1 trillion, however….” Twist Bioscience’s solution is to sequence the trillion pieces of DNA created in this new chip. This solution will allow Twist Bioscience to fill bigger orders faster, which will, Leproust says, allow the company’s customers to “turbocharge their discovery programs.”

Twist Bioscience is developing this particular chip to use DNA as a storage medium for archived data files. “About 60% of the world’s data is archived,” Leproust says. “Often, it’s written once and read never.” With traditional storage technologies, maintaining the ability to access the data becomes painful and expensive.

“We can convert the 0s and 1s to A-C-G-T, make the DNA on our silicon chip, and store it for a thousand years or more,” Leproust states. “If it’s needed, technicians/robots can extract the DNA from the chip and then perform PCR on the DNA to extract a specific file, sequence it, and get the file back to the user within 24 hours.” DNA storage can be the same price as hard drive storage, without the cost of subsequent transfers as storage methods become outdated.

Delivering RNA outside the liver

At the Georgia Institute of Technology, James Dahlman, PhD, associate professor of biomedical engineering, has developed a system that can deliver RNA and DNA to targets other than the liver. The system uses nanoparticles that incorporate DNA barcodes to verify that therapeutic payloads go where they are needed.

“Although it’s clear that RNA and DNA therapeutics can do a lot of good, we can’t deliver them precisely where we want,” Dahlman says. Therefore, many potential therapeutics can’t be developed.

“All the RNA drugs that have worked are delivered to the liver, or they are, like the mRNA vaccines, jabbed into the arm and taken up by surrounding tissue,” Dahlman notes. But if the payload—the mRNA or siRNA—could be delivered to specific targets outside the liver, the therapies could be developed for a wider range of diseases. A better targeting approach could also apply to cell and gene therapies generally.

“A lot of scientists are working on non-liver RNA delivery,” he observes. “What sets us apart is the ability to run thousands of experiments simultaneously.” This approach was developed at Guide Therapeutics, a company that Dahlman co-founded and that has since been acquired by Beam Therapeutics.

DNA barcoding, Dahlman asserts, can help scientists determine which delivery vehicles are best for a specific tissue or payload. “There are lots of options,” he continues, “and you need a way to narrow them down, hence the barcoding.”

Dahlman tags the delivery vehicle with a known sequence of DNA rather than a fluorescent label. The benefit, he says, is that “you can sequence many with exquisite sensitivity.” This capability should make it possible to accelerate the development of drugs that have non-liver targets.

Dahlman is also finding better ways to translate data from mouse models to larger animals. He has 10 journal papers in various stages of preparation or review and says that he expects one in particular to have high impact. This paper, he hints, is about “one of the new technologies we’re developing that could be used to study species-dependent changes in delivery.”

Off-the-shelf iPSCs for conformity

Fate Therapeutics is developing off-the-shelf induced pluripotent stem cells (iPSCs). The company creates renewable master cell banks by expanding multiplexed engineered clonal iPSC lines from a single cell. “You can launch each laboratory experiment and each manufacturing campaign with the same starting material,” says Bob Valamehr, PhD, the company’s chief R&D officer. “It’s uniform and consistent.” Accordingly, it’s capable of yielding higher quality products.

Using the “master cell bank from a single cell” concept, Fate created FT596, a multitargeting off-the-shelf natural killer (NK) cell therapy that expresses a CD19-directed, NK cell–optimized chimeric antigen receptor (CAR) and a high-affinity, noncleavable CD16 Fc receptor. Next, an IL15-IL15Ra fusion receptor was added, allowing the cells to expand without external cytokine support and to extend cell survival and efficacy.

“This is the first NK cell product to uniformly contain … distinct engineered entities that can be routinely manufactured from a uniform starting material and at large scale,” Valamehr asserts. “This off-the-shelf approach can slash per treatment costs by a factor of 100, and the patient doesn’t have to wait for an autologous therapy to be manufactured because it is delivered on demand.”

The approach, which is called logic gating, suggests that many synthetic modalities can be combined to attack, in concert, each defense a cancer cell mounts after an attack. This way, Valamehr explains, “we can start peeling the onion to get to the heart of the cancer and eliminate it for good.” He predicts that in the next generation of iPSC products, the inserted transgenes may communicate with each other.

Fate’s goal, Valamehr states, is “to expand manufacturing to the 10,000-L scale so that, one day, iPSC therapies may be stocked by pharmacists and filled by prescription.”

“Rewiring” unknown biosensors

At Rice University, Jeffrey Tabor, PhD, associate professor of bioengineering, is developing next-generation biosensors. His approach repurposes two-component signaling systems in bacteria to produce new medical diagnostics and therapeutics and, perhaps, overcome multidrug resistance.

“The challenge,” he explains, “is that of the roughly 50,000 two-component systems that have been identified, approximately 99% are uncharacterized.” Their basic activity is the same, though. A sensor kinase on the bacterial cell surface reacts to a stimulus and initiates phosphorylation to activate a response regulator inside the cell, which turns a specific gene on or off.

two-component signaling systems
Although thousands of two-component signaling systems in bacteria have been identified, relatively few of them have been characterized. To study these systems, and “hack” bacterial sensing, Rice University researchers led by Jeffrey Tabor, PhD, have developed a system to mix and match tens of thousands of sensory inputs and genetic outputs. The rewiring of signaling systems could lead to new applications in basic science, medicine, and industry. [Jeff Fitlow, Rice University]

“We can see them,” he says, “but we don’t know what they are sensing and what genes they are regulating.” Very little is known about these systems. Most of the bacteria containing them can’t be grown in a laboratory.

Tabor and his team are developing technology to discover what a a two-component system is sensing. It is based upon sequencing the genome, finding the system, and remaking the system using gene synthesis. Then the system can be re-expressed in Escherichia coli or other convenient bacteria.

The second part of his research involves “rewiring the system” using molecular scissors or in silico methods to identify and edit the DNA-binding domain in the response receptor. Essentially, a segment of DNA is replaced with a well-characterized alternative that can then be expressed and, therefore, measured.

“Ultimately, we want to use this system to control therapeutic production in the body,” Tabor states. “Think of it as ‘Dr. Robot’ in the body. It hangs out, senses disease, and then produces a reporter gene to turn on production of a specific therapeutic.

“We’re transferring the technology to a startup company—PanaBio—to produce diagnostic and therapeutic bacteria. We also are collaborating with a biomaterials lab to produce hydrogel materials” to encapsulate the bacteria in a semipermeable membrane through which therapeutic molecules—but not the bacteria—can cross. “Our goal,” he continues, “is to have this in the clinic within four years.”

Synthetic biology constitutes the next revolution in the life sciences, and it is moving from the research laboratory to the bedside now. In the not-too-distant future, even more concisely engineered therapeutics will become available that will be even more effective, better targeted, and safer for patients.