Technical and design issues

Researchers can now design and mass-produce genetic material-a technology that helps build mRNA vaccines. What can it bring us next?

Credit...Illustration by Jaedoo Lee

Ten years ago, when Emily Leproust was the head of research at life science giant Agilent, a pair of scientists and engineers in their 50s—Bill Banyai and Bill · Bill Peck proposed to her the idea of ​​starting a company. These bills came to be known as biotech veterans. Parker is a mechanical engineer with professional training in fluid mechanics; Banyai is a semiconductor expert who has been working in genomics since the mid-2000s, driving the transition from the old-fashioned Sanger sequencing of single DNA fragments at once to the simultaneous processing of millions of DNA fragments. The transformation of next-generation sequencing of fragments. When chemicals are miniaturized and placed on silicon chips, reading DNA becomes fast, cheap, and widespread. The Bills met when Banyai hired Peck for a genomics project. They realized that they had the opportunity to do something similar to writing DNA—making the synthetic gene manufacturing process more scalable and cost-effective.

At the time, DNA synthesis was a slow and difficult process. The reagents—the well-known bases (A, T, C, and G) that make up DNA—are moved to a plastic plate with 96 pits or holes, each having about 50 microliters , Equivalent to a drop of eye drops. "In a 96-well plate, conceptually, what you have to do is put in the liquid, mix it, wait, maybe you can heat it, and then take out the liquid," Leproust said. The bill proposes to put the same process on a silicon chip. The chip has the same footprint as a 96-well plate and will be able to hold 1 million small holes, each with a volume of 10 picoliters, or less than its volume. One millionth of that. The size of the 50 microliter hole.

Because the hole is too small, they cannot simply move the liquid into it. Instead, they use what is essentially an inkjet printer to fill them, distributing A, T, C, and G instead of pigment ink. A catalyst called tetrazole is added to combine bases into single-stranded DNA sequences; advanced optical technology makes perfect alignment possible. As a result, they can now print millions of copies of DNA instead of producing 96 pieces of DNA at the same time.

The concept is simple, but, Leproust said, “engineering is difficult.” She explained that when you synthesize DNA, the yield or success rate decreases with each base added. The combination of A and T is weaker than that of G and C, so DNA sequences with a large number of consecutive A and T are usually unstable. Generally speaking, the longer the DNA strand, the greater the possibility of error. Twist Bioscience, founded by Leproust and Bills, currently synthesizes the longest DNA fragment in the industry, up to 300 base pairs. Known as oligonucleotides, they can then be combined to form genes.

Today, Twist charges 9 cents for DNA base pairs, which is nearly 10 times lower than the industry standard ten years ago. As a customer, you can visit the Twist website, upload a spreadsheet containing the required DNA sequence, select the quantity and pay with a credit card. After a few days, the DNA will be delivered to the door of your laboratory. At that time, you can insert synthetic DNA into the cells and let them start to make (hopefully) the target molecule produced by the DNA code. These molecules eventually become the basis for new drugs, food flavorings, artificial meat, next-generation fertilizers, and industrial products in the petroleum industry. Twist is one of many companies that sell synthetic genes, and they are betting that the future is full of DNA-based bioengineering products.

To some extent, that future has arrived. Gene synthesis was behind the two biggest "products" of the past year: Pfizer and Moderna's mRNA vaccines. Almost after the Chinese Center for Disease Control and Prevention released the genome sequence of SARS-CoV-2 to a public database in January 2020, the two pharmaceutical companies were able to synthesize DNA corresponding to the virus-specific antigen (called the spike protein). This means that their vaccine—unlike traditional analogs, which teach the immune system to recognize the virus by introducing a weakened version of the virus—can transmit genetic instructions to cause the body to produce only the spike protein, so it will actually Viral infection.

Just 10 years ago, this was almost impossible. For researchers, it is a challenge to synthesize a DNA sequence long enough to encode a complete spike protein. But technological advancements in the past few years have allowed vaccine developers to synthesize longer DNA and RNA fragments at a lower cost and at a faster rate. We obtained the vaccine prototype within a few weeks and completed the injection within a year.

Now, the company and scientists are looking forward to the post-Covid future, when gene synthesis will be deployed to solve various other problems. If the first phase of the genomics revolution focused on reading genes through gene sequencing, then the second phase would be writing genes. The gene editing technology Crispr, whose inventor won the Nobel Prize last year, has received more attention, but the rise of gene synthesis is expected to be an equally powerful development. Crispr is like editing an article, allowing us to make precise changes to the text at a specific location; gene synthesis is like writing an article from scratch.

Like many technologies in their infancy, gene synthesis (and the field it supports, synthetic biology) has triggered a lot of speculation and startup activity. Most companies-except for those researching the coronavirus-are in the experimental stage; their applications have not yet come to a conclusive result. Nonetheless, these possibilities still attract investors and scientists, whether they are manufacturing microbes to produce industrial chemicals or designing human cells to treat medical diseases. Even if a small part of these efforts succeed, they could lead to a trillion-dollar market. The analogy often used by biotech venture capitalists is that we are in the Apple II era of synthetic biology, and products comparable to iMac and iPhone are still to come. This is a grand proposition-but it is not unbelievable, especially now that Covid has already tested some basic technologies. Personal computing has created our digital lives; reading and writing DNA may mean controlling our bodies.

One of the maxims of synthetic biology is: Nature is the best innovator. For example, the "cutting" enzyme CaS-9 used in Crispr was originally developed by bacteria to defend against viruses. But this adage obscures the fact that for most of human history, nature is also opaque, requiring humans to discover its invention entirely by accident. Penicillin, quinine-many of our medicine cabinet staple foods are found by placing food for a long time or by searching for active ingredients in herbs. It was not until the advent of modern chemistry that we were able to write formulas that are common in physics and mathematics.

Then came the genomics revolution. The first phase was marked by milestones such as human genome sequencing and the emergence of companies such as 23andMe, and the focus was on reading genes. The second phase underway is about writing genes. We can now use our understanding of molecular biology—how DNA specifies the sequence of RNA, which in turn specifies the production of proteins—and use Crispr and DNA synthesis to design genetic formulas that produce the output we want. So what does this look like in practice?

One of Twist's largest customers is Ginkgo Bioworks, a cell engineering company that went public in September and was valued at $25 billion by mid-November. Ginkgo's main office is located in a converted warehouse in the Boston Harbor District. A few months ago, when I visited the executives of Ginkgo, Patrick Boyle took me to see their five "foundries"-named after the microchip manufacturing plants. We passed a machine that uses microfluidics to mix reagents and cells, and another that uses mass spectrometry to quickly analyze the chemical composition of liquids.

For decades, the basic labor unit for biological research has been low-level graduate students, who diligently suck liquids, make measurements, check the results, and if they are lucky, they may conduct experiments several times a month. In contrast, Ginkgo brings the efficiency of the assembly line to the laboratory, using machines that can pipette, mix, and analyze more accurately than anyone, so thousands of different experiments can be performed at the same time.

Ginkgo is a "platform" company-it does not produce final products for itself, but designs cells for customers. The process is roughly like this: A customer called Ginkgo and said: “We want to produce a rose fragrance for our perfume that is cheaper than distilling from flowers.” The designer of Ginkgo combed a gene bank, from which Pick out those genes known from previous observations or sequencing to produce the properties of rose oil. After arranging these sequences on the computer, Ginkgo orders DNA from Twist or other suppliers who are responsible for the synthesis of most of the base pairs.

In Ginkgo, the synthesized DNA is then inserted into the host cell, perhaps yeast, and it begins to produce enzymes and peptides. The trial and error are as follows. Maybe the output of the first gene sequence was too floral and not spicy enough; maybe those from the second gene sequence had the correct smell, but the cells did not produce enough smell. Once a valid prototype is found, Ginkgo will increase yield by growing yeast in a vat and simplifying the process of extracting the required molecules from the soup. What Ginkgo provides is the formula and ingredients—the winning genetic code, the host cells and the conditions under which the cells must be cultured—and then customers can use them by themselves.

Ginkgo's platform first attracted customers in the perfume industry, but for the past two years, it has been working with pharmaceutical companies to find new therapies. One of the projects is seeking to discover the next generation of antibiotics to combat antibiotic resistance. Lucy Foulston, who has a background in molecular microbiology, is leading the work; chemist Tom Keating is working with her. Together, they emphasized to me a beautiful and twisted paradox—most antibiotics and most antibiotic resistance come from the bacteria themselves. Bacteria carry genetic fragments with instructions to produce antibacterial molecules that kill other bacteria. Usually, they are also self-resistant, so the bacteria that make specific antibiotics will not kill themselves, but this resistance can be transferred between bacteria and become widespread.

Historically, people have adopted two approaches to develop new antibiotics. The first, celebrated in the story of Alexander Fleming and moldy bread, is to find them in nature: scientists go out, get a little soil from a geyser or coral reef, put what they find in a petri dish, and see if It will kill any interesting bacteria. The second method is to comb through the chemical library to find molecules with antibacterial activity. Together, these two methods provided us with a stable supply of new antibiotics until the 1980s and 1990s, when discovery began to dry up.

"There is a lot of speculation," Keating said. "Have we found everything that is useful? Have we found everything that is easy to find? Have we encountered bacteria that are hard to kill now, so that the new bacteria we found don't work on them?" Regardless of the reason, The reality is that in the face of increasing antibiotic resistance, our new antibiotics have been used up.

Ginkgo's antibiotic project is looking for fragments that encode new antibacterial agents through the bacterial genome. Sequencing efforts in the 1990s and 2000s produced large public and private bacterial genome databases, giving scientists a deeper and deeper understanding of which genes produce which molecules. As Foulston said, the scientists also developed the necessary technology to "take these genes out and put them into another bacterial strain"—a strain they know how to use—"and then induce that specific strain to produce interest. Numerator."

Keating continued: "We don't need the organism anymore. We don't need it to grow on the plate. We don't need it to kill anything else. All we need is code."

No matter how many programming metaphors you use, DNA is more confusing than code. If you type "print'hello world'", you want the computer to return "hello world". If you synthesize a DNA sequence ACTCAG and put it into a cell, you may be able to predict with certainty what will happen in the cell, but you will never really know.

Nevertheless, biotechnology has ushered in a unique new moment-software, hardware, data science, and laboratory science are finally mature enough to work together and reinforce each other. An mRNA vaccine that had not been approved by the Food and Drug Administration before the pandemic is a good example. Ginkgo's antibiotic program is another. Further advances in machine learning and computer modeling will only increase the possibilities. The same is true for semiconductors: Although Twist's 10 picoliter well looks small, Leproust points out that from the perspective of the semiconductor industry in the 21st century, it is "a grand canyon, almost like it was in the Stone Age." The company is already experimenting with chips with holes more than 300 times smaller and 150 nanometers in diameter. (For reference, Intel is now manufacturing 7-nanometer silicon chips for computers.) This advancement is expected to reduce the cost of gene synthesis by a million times and allow more researchers to use it and conduct more experiments. And play a role in the application.

For synthetic biology, the next frontier is to go to places that even nature has never been. Rather than trying to replicate the fragrance of roses, can we combine genes to produce a more intoxicating fragrance? Can we convert DNA into circuits that enable cells to function as living computers? "So far, we just take what nature has invented, copy it, maybe optimize it," Keating said. But he aspires to obtain the kind of command and creativity that chemists now enjoy, and they can synthesize anything that can be represented by diagrams. "I think we just scratched the surface, whether we can program biology to do what chemists traditionally do," he said. "If you can draw a molecule on a piece of paper, can we design an organism to produce that molecule, even if it is something that nature has never seen before? We are still far away—but, you know, little step."

Yiren Lu is a writer and software engineer living in New York. She ended up writing an article for the magazine about startups trying to fix virtual meetings.