Henry Ford is still best known for his 1908 Model T, an inexpensive automobile that almost anyone could own. But his real innovation was integrating the conveyor belt into assembly line production. The inspiration came from the meat packing industry. There, factory workers bled, skinned, and sliced ugly carcasses into smooth chops and filets. Ford adapted the process for cars. With the conveyor belt, he took Model T production from a 12 hour process down to roughly 90 minutes. The efficiency allowed Ford to drop the price of his Model T and led to incredible sales. It also sparked a revolution involving industrial production of all sorts.
In the 2019 documentary Human Nature, Stanford law professor Hank Greely asserts that CRISPR is the biotech industry’s Model T. If that’s true, a startup called Inscripta is rolling out an automated assembly line for the bioscience era. The biotechnology company is launching Onyx, a compact gene editor that can make thousands of changes to genomes in millions of cells for $347,000. This initial device will edit yeast and E. Coli, but Inscripta has promised to make large scale gene-altering a reality in mammalian cells with a forthcoming device.
Officially, CRISPR refers to a DNA sequence found in bacteria called “clustered regularly interspaced short palindromic repeats.” But the term has come to represent a process that lets scientists snip out genes in DNA as if they were snags in a strand of yarn. Scientists use a kind of RNA to find offending genetic code and a protein called Cas-9 cuts it out. Guide RNA then puts in a new set of code where the snag was, changing the original DNA sequence. The discovery has given scientists a way to reprogram life.
What could gene editing yield? The list of possibilities is truly endless, but at a minimum it could pave the way for new types of energy and food, as well as unique ways to fight disease. Last year, Jennifer Doudna and Emmanuele Charpentier were awarded the Nobel for discovering these golden scissors. But CRISPR still has limitations. It can make the same repair in lots of different places or it can make several different repairs in one place, but it cannot do both at the same time. CRISPR can also be expensive to license commercially.
For these reasons, Inscripta doesn’t use CRISPR Cas-9. Its scientists use an enzyme that acts in similar ways, but is proprietary to the company. In certain contexts, commercial users will have to pay a licensing fee to use the enzyme, but in general Inscripta is trying to keep these kinds of additional charges to a minimum. The goal of the technology is to make biological research as easy as doing a Google search.
In the 1980s, computer users had to become experts in MS-DOS to get things done, says Kevin Ness, Inscripta’s former CEO and cofounder of genetic sequencing company 10xGenomics, by way of comparison. “But now we can go in and make a PowerPoint slide or work in Word and Excel, because computers have gotten so much easier to use—that is the jump that Inscripta is giving the world.”
The gene editing market, which is expected to be worth $9.2 billion by 2026, according to Acumen Research and Consulting, has gotten the most attention over its potential for editing human life. However, that isn’t what it’s being used for at first. CRISPR based editing is being tested as a remedy for cancer predisposition and genetic blood diseases, but the technology is not ready for clinical use, according to a panel of scientists from ten countries.
Gene editing is also being used to develop novel therapeutics, but those too will take time. Drugs must go through a long approval process and they don’t always make it out the other end. In the meantime, gene editing is far more likely to deliver a better meatless burger in the next few years than a child with blue eyes.
Ness, who remains an advisor, to Inscripta, says just as a wave of computer coders gave birth to the Internet economy, the company’s easy to use “labtop” editor will pave the way for a new biotech workforce. He sees biologists as the computer programmers of the biotech economy.
The comparison raises potentially fraught aspects of the gene-editing revolution. While the rise of computer coding and the internet led to greater sharing of information, it also gave way to malware and disinformation. The introduction of easy-to-use gene editors will no doubt have negative consequences as well. As synthesizing new life gets easier, it lowers the bar for creating bi threats.
The biotech industry and the government will have to safeguard against intentional and accidental biological threats. Inscripta is already doing this by working with biosecurity experts and agencies within the federal government. However, given the failures of the pandemic response, there’s a concern the U.S. may not be adequately prepared for the next biological threat.
What is Onyx?
The idea for Onyx came out of graduate research. Andrew Garst joined Ryan T. Gill’s Lab at University of Colorado as a postdoc in 2012, hoping to embed himself in the CRISPR craze. He spent two years banging his head against a wall trying to figure out a way to scale gene editing with CRISPR Cas-9.
“Getting it to work for more than 10,000 sites across the genome was a real technical challenge that my lab and many other very prestigious labs were working on at the time,” he says. CRISPR Cas-9 made editing genes dramatically easier than it had been, but Garst wanted to be able to do more.
In October 2014, while on a retreat at Chautauqua State Park, a lush wildland near Boulder, Colorado, Garst was looking over slides from a colleague’s project. The goal was to integrate genes into E. coli that would make it produce isobutanol, an alcohol frequently used in manufacturing processes. The experiment seemed to be working.
As Garst looked at the slides, he made a broader realization. Through engineering E. coli to produce isobutanol he realized that instead of bombarding cells with separate editing tools and instructions, he could synthesize all the changes he wanted into a single package before inserting them into a cell, leading to more efficient integration. This epiphany could make CRISPR useful at a much larger scale.
Two years later, he published a paper detailing a way to use CRISPR to introduce as many as 500,000 changes to the structure of a genome at the same time. His design made these changes trackable. The paper also included computer code to help researchers reverse engineer the genetic coding in order to get a cell to do what they wanted it to do. The technique was called CREATE, which is shorthand for CRISPR-enabled Trackable Genome Engineering Using Homologous Recombination.
The results allowed scientists to run a large number of experimental edits and learn more about what certain genetic changes would do in a cell. The novel methodology had particularly interesting implications for the biotech community, which alters cells to produce a range of products including compounds that serve as the basis for medicines. Seeing the opportunity, Garst co-founded Inscripta around the time the paper was released, hoping to refine the CREATE technology into a more automated system.
Over the last four years, Inscripta has company raised $300 million. In October, it hired a new CEO, Sri Kosaraju, who spent 16 years overseeing JPMorgan’s healthcare and technology portfolio, to scale the business. It also launched the Onyx benchtop: a black box roughly the size of a small fish tank with a computer screen front. It’s as slick looking as any digital consumer product, but made for scientists, with an interface that lets them create their experiments. Onyx’s software is designed to then optimize tests for the best possible results.
Inscripta has shifted away from using CRISPR Cas-9 to develop its own proprietary DNA cutting nucleases. (As the name “Cas-9” implies, the Cas protein is one of many.) Working with other nucleases gives Inscripta a palette of tools to make edits with. The company has developed a man-made nuclease called MAD-7. Garst said that identifying novel nucleases is a part of how the company intends to keep costs low and the technology accessible.
MAD-7 is generally free. Businesses have to pay a fee only if it’s used in a therapeutic or reocurring manufacturing.
The company is in the process of engineering other nucleases. Garst explains that part of the reason for developing them was to avoid the massive royalties associated with Cas-9. “Some groups are now deciding to go ahead and just sign up for those licensing fees,” he says. “But over time we believe that as they see the performance of the systems we’re building and that any royalties attached are going to be way lower than what they’d be looking at anywhere else—they’re going to be pushed to consider Onyx and Inscripta tools.”
The new class of biotech
Inscripta is part of a growing class of biotech companies that are driving down the cost of gene reading, writing, and editing. It includes old guard genetics companies such as Illumina, but also newer entrants like DNA printer Ginkgo Bioworks, CRISPR-based gene editing company Mammoth Biosciences, therapeutics company Editas, and gene therapy company Bluebird Bio.
For biotech businesses, being able to easily make complex changes in the genomes of single-celled organisms could be a huge cost saver. Yeast and E. Coli are used to produce compounds that serve as the foundation for basic consumer products such as beauty products and more complex ones like medicines.
Take lysine, for example. It’s an amino acid, something human bodies use to make proteins that play a role in a variety of functions like growth and hormone production. We typically get it from food: Apricots, avocados, and beets are all rich in lysine. It’s also used in animal feed and multivitamins.
Onyx helps researchers identify the right design and materials to run their experiments.
Instead of breeding, Inscripta’s platform suggests that bio manufacturers can edit in their desired traits and potentially get an even higher lysine output. Last year, the company came up with a combination of edits that led to a strain of E. Coli that had a 10,000-fold increase in production versus the wild type. To get there, the company tested 200,000 edits to figure out what would get more lysine out of E. Coli. The stunt aimed to show off what its hardware and software could do for lysine producers.
Onyx helps researchers identify the right design and materials to run their experiments. Then it lets scientists see where their edits take them. For example, researcher Shelley Copley at Colorado University has been toiling the last few years on a project aimed at better understanding the role of synonymous mutations, often called silent mutations for their lack of observable effect in E. Coli growth. In order to have a meaningful sample, she concluded she needed to make an enormous number of edits—some 300,000 of them—in a few hundred E. Coli genomes.
“When I first came up with the number of 300,000, it was partly because I had no idea what percentage of these mutations would have any effect at all,” explains Copley. “So I was shooting for a very big number so we would at least find something.”
She had applied to the National Institutes of Health for grant funding and got it. She planned to use Garst’s CREATE method, which was open source. The process fell short for her (Garst admits that the pre-commercialized version of his methodology was a bit oversold). It turned out to be an inefficient process and she couldn’t get enough cells to take the edits.
Copley ended up getting an early version of Inscripta’s gene editing technology and created an initial library of 50,000 mutants. When she ran her experiment, all the mutations she wanted represented in her population of cells came through. Garst says the process has a near 70% efficiency, which is about 10% below the most state-of-the-art controlled, molecular biology systems.
In the end, 5% of these so-called silent mutations affected growth. That’s such a statistically significant number that Copley may not need to do the full 300,000 cuts she planned.
“With the first 50,000, we’ve already come up with enough mutants to keep us busy doing interesting science,” she says.
The risks of democratizing biology
Onyx’s ability to run large scale experiments is unprecedented for evolutionary biology research as well as the biotech industry at large. Never before have scientists been able to edit this many genomes at once, in so many locations, and see the results using a compact desktop device.
This scale of editing also raises questions about what kinds of mutants this technology may ultimately bear. What does democratizing this kind of technology mean for the safety of human health? Just as the automated assembly line can be made to produce cars, it can also be made to produce guns.
This issue was the subject of a joint hearing on biosecurity convened by both the House Armed Services subcommittee and House Foreign Affairs subcommittee in October. Six committee members sat more than six feet apart at their elevated wooden desk. Ami Bera, a silver haired doctor and representative for California’s 7th District, chaired the meeting.
“We’ve not had an aircraft carrier brought to its knee by kinetic force of missile, but what we just saw in this past year was an aircraft carrier brought to port because of a virus—that really underscores what I worry about,” said Bera. “The availability of technologies to alter viruses and do genetic editing, the know-how, and the capabilities are rapidly increasing and that is something that keeps me awake at night.”
COVID-19 has been a painful example of how biology can cause staggering damage. If the goal of companies such as Inscripta is to make doing complex biology more accessible, that means that the technology could be available to anyone who might want to get their hands on it.
Inscripta does a background check on its customers to ensure the company they’re working with is running a legitimate biological operation.
Andrew Weber, a fellow at the Center on Strategics Risks, says that what makes illegal bioweapon operations more complex to combat is that they’re hard to track down. “It’s not like the Soviet program which had huge factories, massive quantities of Anthrax that you can see on satellite,” he says. “You could produce bioweapons in a kitchen.”
Weber spent 20 years on nuclear, chemical, and biological defense programs at government agencies and now advises Gingko Bioworks on security strategy. He says that biosurveillance is only properly funded at the government level when there’s an immediate threat at hand such as COVID-19. But in order to catch upcoming threats, the government needs to more consistently fund agencies that monitor for biothreats.
“Having good early warning and good disease surveillance, and rapid counter measures and stockpiling counter measures, when we have them can go a long way to reducing the threat,” he says.
Weber says partnerships between private industry and government could help to shore up some of these gaps. He points to two programs by the Intelligence Advanced Research Projects Activity (IARPA): Fun GCat and Felix. Both programs are designed to create tools for the private sector to help prevent the creation of biothreats.
The biotech community also has asked for inter-agency departments and multilateral organizations aimed specifically at managing biothreats. Agencies have the capacity to work together, as the response to the Ebola outbreak has shown. But there’s a desire for a more permanent group that might work across health, commerce, and defense departments, so that the private sector can more easily coordinate with the government.
For its part, Inscripta is working with IARPA broadly on developing new tools to help mediate some of the government’s current gaps in biosecurity. In particular, the company is working on screening databases. It also hired Tom Slezak, a bioinformatics expert who spent 13 years at the Lawrence Livermore National Laboratory, to build out databases of DNA sequences of interest and advise it on its biosecurity efforts.
Inscripta does a background check on its customers to ensure the company they’re working with is running a legitimate biological operation. The company also screens customers’ genomes, even if they’re proprietary, to see if they contain anything that might be considered dangerous. It also looks at proposed edits that might have hazardous results.
Some disease research labs are authorized to work with risky material—such as super flu—to better understand it. In that work, there is always the possibility that a virus might escape the lab in what’s known as an accidental release scenario. Furthermore, companies such as Inscripta cannot stave off copycat devices that may not come with the kinds of safeguards they build into their platforms.
“Inscripta has taken to heart is that when you’re launching a new industry—which is really what Inscripta is doing with this platform—you really have a responsibility to set a very high bar for biosecurity and bring the rest of this nascent industry up to that level,” says Slezak.
The industry is nascent indeed. So far Inscripta has sent Onyx to one commercial partner through a beta program. By the end of March, it will deliver its device to its first commercial customer, a company in Europe. Its labtop gene editor may be a breakthrough—but what will matter most are the breakthroughs it could help deliver in the years ahead.