With a souped up synthetic genome, scientists make bacteria impervious to viral attack

Two years ago, scientists in Britain swapped out the DNA of the bacteria Escerichia coli for a genetic coding program that was entirely human-made. At the time, it was the largest and most complex synthetic genome ever created.

On Thursday, that same group of researchers at the Medical Research Council Laboratory of Molecular Biology reported in Science that with continued tinkering, they’ve made their artificial life form virtually invincible to viral infection. Other adjustments to the bacteria’s designer genome endowed the bug with the ability to string together non-natural amino acids to produce proteins never before seen inside a living cell.

In an accompanying editorial, scientists not involved in the work said the achievement promises to unlock the potential for making all sorts of new classes of chemicals — for medicine, food production, and industrial manufacturing.

“It’s brilliant work,” Tom Ellis, director of the Center for Synthetic Biology at Imperial College London, who was not involved in the new study, told STAT via email. “It shows two really exciting applications that can be achieved by having cells with synthetic genomes that have been genetically recoded.”

DNA, fundamentally, is an information storage system; your DNA stores all the information required to build the molecules and cells and tissues that make you, you. But to actually build those things, you also need a storage-retrieval and decoding system. And on Earth, all of life — from pond scum to peregrine falcons — evolved to use a single system. Every organism decodes DNA using the same 64 codons (three-letter combinations of the DNA bases known as A’s, T’s, C’s, and G’s)  to specify which of the 20 amino acids are going into the proteins it makes. Which means there are redundancies.

So, TCA, for example, tells the cell to grab an amino acid called serine and stick it next in the chain of whatever protein it’s building. But so does TCT. And AGT. In fact, there are six codons that all specify serine.

For years, Jason Chin, a molecular biologist at the MRC lab who led the new study, worked to whittle down some of those redundant codons and assign them new functions. Previously he and his team had systematically eliminated three codons across their E. coli’s genome, creating a bacteria, dubbed Syn61, that lived and squirmed and formed colonies despite using 61 codons rather than nature’s 64. But Syn61 still possessed the genetic program for creating all the molecules that actually read the codons and shuttle the correct amino acid to the site of protein production. These critical decoding molecules are called transfer ribonucleic acids, or tRNAs.

In their latest work, Chin and his colleagues went through and deleted the tRNAs that recognize the codons they had already removed. The exercise wasn’t just about further streamlining their synthetic genome. Viruses don’t have any of their own decoding or protein production machinery. They rely on a host cell to do that. And in theory, if a virus invaded a bacteria that didn’t have the ability to read all of the codons in the viral genome, it would stymie the would-be hijacker’s plans for self-replication.

“The hypothesis was that the cells should be completely resistant to viruses,” said Chin. And indeed, that’s exactly what his team found. When mixed together with a cocktail of five different bacteriophages — viruses that prey on bacteria — their genetically reformulated E. coli rebuffed the viral onslaught. Eliminating the tRNAs was crucial. Colonies of their original synthetic strain, Syn61, were blasted apart by the phage invaders.

This proof of concept could be quite useful to huge swaths of the biotech industry, which use E. coli as a workhorse organism to produce enzymes, drugs, and biochemicals. Phage contamination can set back a production line for months. “An E. coli that can grow and produce with almost zero chance of phage infection is the dream for a lot of industry people,” wrote Ellis.”It will be really interesting to see how this strain could end up being used by companies and startups.”

Floyd Romesberg, a research scientist at Sanofi agreed that phage-proofing the production of biomedicines is a legitimate need. Before being brought on to the company as part of its acquisition of Synthorx, Romesberg had spent decades at the Scripps Institute developing expanded synthetic DNA alphabets. (Sanofi is currently testing its first drug made with one of these novel genetic letters in a Phase 1 trial for solid tumor patients.) But he cautions that replacing today’s E. coli workforces with synthetic ones isn’t so straightforward.

“The lab is a generous growing condition compared to the high density environment inside an industrial bioreactor,” said Romesberg. In addition to deleting the tRNAs, Chin’s group made some other adjustments, and put the bacteria through a forced evolution process to make it grow better. That added 482 random changes to its synthetic genome. In a lab, that might not matter. But in a resource-constrained environment, such alterations could prove problematic. “Four-hundred and eight two mutations makes me a little nervous,” said Romesberg.

Then there’s the issue of strain optimization. Companies often have to comb through massive strain libraries to find a version of E. coli that efficiently produces the protein they want it to under the environmental constraints they have. “Jason has given us this virus-resistant strain, which is great, but it’s just one. If you needed to go to a different host strain you’d have to recode that as well, which is no small task.”

Beyond neutralizing viral threats, the new study offers another exciting advance for the field of synthetic biology. Chin’s group showed that by freeing up some of the codons in their phage-proof E. coli, they could give them a new function: coding for two amino acids beyond the natural 20. Other research groups have shown it’s possible to incorporate a single non-natural amino acid into a protein. But Chin’s bacterial colonies successfully strung together eight in a row, a new record.

It’s still a long way from making a full protein, which usually weighs in between 100 and 400 amino acids. But it shows that making polymers out of fully unnatural components might really be possible. And that would be the true test of whether nature’s way is the only way.

“We weren’t trying to show that you can make a specific, existing molecule, but rather to just explore the idea that you can write DNA sequences in cells and program the synthesis of completely new polymers,” said Chin. He hopes to combine this work with other projects he has for recoding ribosomes — where all this protein production takes place — to one day get away from petroleum-based technologies for making chemicals and other materials and usher in a new paradigm of regenerative manufacturing.

“If we think about plastics and other materials built from crude oil, they are made with no sense of how we are ever going to get rid of them,” said Chin. He and other genome recoders envision an alternative reality where materials are engineered with their whole life cycle in mind from the get-go. That starts by programming bacteria to produce polymers built in a way that the chemical linkages between them are readily snapped by bespoke enzymes. So that when they are no longer useful, they get sent into a vat to be broken down into their original molecules to be reused again by other bacterial factories.

“Being able to build in this element of sequence control from the beginning, is a real opportunity to get us out of the plastics and microplastics problem,” Chin said.