DNA Gets a New — and Bigger — Genetic Alphabet

In 1985, the chemist Steven A. Benner sat down with some colleagues and a notebook and sketched out a way to expand the alphabet of DNA. He has been trying to make those sketches real ever since.

On Thursday, Dr. Benner and a team of scientists reported success: in a paper, published in Science, they said they have in effect doubled the genetic alphabet.

Natural DNA is spelled out with four different letters known as bases — A, C, G and T. Dr. Benner and his colleagues have built DNA with eight bases — four natural, and four unnatural. They named their new system Hachimoji DNA (hachi is Japanese for eight, moji for letter).

Crafting the four new bases that don’t exist in nature was a chemical tour-de-force. They fit neatly into DNA’s double helix, and enzymes can read them as easily as natural bases, in order to make molecules.

“We can do everything here that is necessary for life,” said Dr. Benner, now a distinguished fellow at the Foundation for Applied Molecular Evolution in Florida.

Hachimoji DNA could have many applications, including a far more durable way to store digital data that could last for centuries. “This could be huge that way,” said Dr. Nicholas V. Hud, a biochemist at Georgia Institute of Technology who was not involved in research.

It also raises a profound question about the nature of life elsewhere in the universe, offering the possibility that the four-base DNA we are familiar with may not be the only chemistry that could support life.

The four natural bases of DNA are all anchored to molecular backbones. A pair of backbones can join into a double helix because their bases are attracted to each other. The bases form a bond with their hydrogen atoms.

But bases don’t stick together at random. C can only bond to G, and A can only bond to T. These strict rules help ensure that DNA strands don’t clump together into a jumble. No matter what sequence of bases are contained in natural DNA, it still keeps its shape.

But those four bases are not the only compounds that can attach to DNA’s backbone and link to another base — at least on paper. Dr. Benner and his colleagues thought up a dozen alternatives.

Working at the Swiss university ETH Zurich at the time, Dr. Benner tried to make some of those imaginary bases real.

“Of course, the first thing you discover is your design theory is not terribly good,” said Dr. Benner.

Once Dr. Benner and his colleagues combined real atoms, according to his designs, the artificial bases didn’t work as he had hoped.

Nevertheless, Dr. Benner’s initial forays impressed other chemists. “His work was a real inspiration for me,” said Floyd E. Romesberg, now of the Scripps Research Institute in San Diego. Reading about Dr. Benner’s early experiments, Dr. Romesberg decided to try to create his own bases.

Dr. Romesberg chose not to make bases that linked together with hydrogen bonds; instead, he fashioned a pair of oily compounds that repelled water. That chemistry brought his unnatural pair of bases together. “Oil doesn’t like to mix with water, but it does like to mix with oil,” said Dr. Romesberg.

In the years that followed, Dr. Romesberg and his colleagues fashioned enzymes that could copy DNA made from both natural bases and unnatural, oily ones. In 2014, the scientists engineered bacteria that could make new copies of these hybrid genes.

In recent years, Dr. Romesberg’s team has begun making unnatural proteins from these unnatural genes. He founded a company, Synthorx, to develop some of these proteins as cancer drugs.

At the same time, Dr. Benner continued with his own experiments. He and his colleagues succeeded in creating one pair of new bases.

Like Dr. Romesberg, they found an application for their unnatural DNA. Their six-base DNA became the basis of a new, sensitive test for viruses in blood samples.

They then went on to create a second pair of new bases. Now with eight bases to play with, the researchers started building DNA molecules with a variety of different sequences. The researchers found that no matter which sequence they created, the molecules still formed the standard double helix.

Because Hachimoji DNA held onto this shape, it could act like regular DNA: it could store information, and that information could be read to make a molecule.

For a cell, the first step in making a molecule is to read a gene using special enzymes. They make a copy of the gene in a single-stranded version of DNA, called RNA.

Depending on the gene, the cell will then do one of two things with that RNA. In some cases, it will use the RNA as a guide to build a protein. But in other cases, the RNA molecule floats off to do a job of its own.

Dr. Benner and his colleagues created a Hachimoji gene for an RNA molecule. They predicted that the RNA molecule would be able to grab a molecule called a fluorophore. Cradled by the RNA molecule, the fluorophore would absorb light and release it as a green flash.

Andrew Ellington, an evolutionary engineer at the University of Texas, led the effort to find an enzyme that could read Hachimoji DNA. He and his colleagues found a promising one made by a virus, and they tinkered with it until the enzyme could easily read all eight bases.

They mixed the enzyme in test tubes with the Hachimoji gene. As they had hoped, their test tubes began glowing green.

“Here you have it from start to finish,” said Dr. Benner. “We can store information, we can transfer it to another molecule and that other molecule has a function — and here it is, glowing.”

In the future, Hachimoji DNA may store information of a radically different sort. It might someday encode a movie or a spreadsheet.

Today, movies, spreadsheets and other digital files are typically stored on silicon chips or magnetic tapes. But those kinds of storage have serious shortcomings. For one thing, they can deteriorate in just years.

DNA, by contrast, can remain intact for centuries. Last year, researchers at Microsoft and the University of Washington managed to encode 35 songs, videos, documents, and other files, totaling 200 megabytes, in a batch of DNA molecules.

With eight bases instead of four, Hachimoji DNA could potentially encode far more information. “DNA capable of twice as much storage? That’s pretty amazing in my view,” said Dr. Ellington.

Beyond our current need for storage, Hachimoji DNA also offers some clues about life itself. Scientists have long wondered if our DNA evolved only four bases because they’re the only ones that can work in genes. Could life have taken a different path?

“Steve’s work goes a long way to say that it could have — it just didn’t,” said Dr. Romesberg.