Scientists reprogrammed bacteria to be immune to viruses

Scientists created a synthetic genome for a bacterium by stringing together building blocks of DNA — and the new genome made the microbe immune to viral infection.

Even when exposed to a cocktail of bacteriophages — viruses that infect bacteria — the designer Escherichia coli remained unscathed, while an unmodified version of the bacterium quickly succumbed to the viral attack and died, the research team reported in their new study, published Thursday (June 3) in the journal Science. That's because viruses usually hijack a cell's internal machinery to make new copies of themselves, but in the designer E. coli, that machinery no longer existed.

"Our understanding of the genetic code allowed us to hypothesize that viruses shouldn't be able to infect and propagate" in the modified E. coli, and that turned out to be true, said first author Wesley Robertson, a postdoctoral researcher in synthetic biology at the MRC Laboratory of Molecular Biology (MRC-LMB) in the U.K. Making bacteria resistant to viral infection could be useful in drug development, since drugs like insulin and some vaccine ingredients are grown in bacteria, for instance, the authors wrote in their study. 

Related: Genetics by the numbers: 10 tantalizing tales 

But while a nice perk, making E. coli invulnerable to viruses wasn't the main goal of the research, Robertson said. The team wanted to replace the genes and cellular machinery they'd removed with reprogrammed machinery of their own design, so the microbe would produce proteins according to their instructions.

Cells normally only use 20 building blocks, called amino acids, to build all their proteins, but now, scientists can introduce "unnatural amino acids" for use in protein construction, which have the same basic backbone as all amino acids, but novel side chains. In this way, the team prompted their modified microbes to build macrocycles — a class of molecules used in various drugs, including antibiotics — with unnatural amino acids incorporated in their structures. In the future, the same system could potentially be adapted to make plastic-like materials, without the need for crude oil, Robertson said. 

"This was unthinkable ten years ago," said Abhishek Chatterjee, an associate professor of chemistry at Boston College, who was not involved in the study. Assuming the method can be adopted easily by other labs, it could be used for a wide range of purposes, from drug development to the production of never-before-seen materials, he said.

"You can actually create a class of polymers that are completely unheard of," Chatterjee said. "When this [technology] becomes really efficient and all the kinks are ironed out, it could become an engine for developing new classes of biomaterials," which could be used in medical devices that get implanted in the human body, for example, he said.

Building genomes from scratch

To create their programmable E. coli, the team took advantage of a quirk in the process of how genetic information gets translated into proteins. 

Just like human DNA, E. coli chromosomes contain four bases, adenine (A), thymine (T), cytosine (C) and guanine (G). A set of three bases — such as TCG or AGC, for example — is known as a codon, and each codon corresponds to one amino acid, or protein building block. In addition, some codons tell the cell when to stop building a protein; these are named "stop codons."

When a cell needs a particular protein built, an enzyme swoops in and copies down all the relevant codons for that protein and stores that information in a new molecule called messenger RNA (mRNA). The mRNA then gets shipped to the cell's protein-building factory, the ribosome, where another molecule called transfer RNA (tRNA) reads off those copied instructions. The tRNA then fetches all the necessary amino acids to build the desired protein, up to the stop codon.  

DNA bases can be arranged in 64 different three-base codons, with three of these being stop codons. That said, cells actually only have 20 amino acids to work with, meaning several different codons code for the same amino acids.

Related: 10 amazing things scientists did with CRISPR

"There is this inherent redundancy in the genetic code, where you have 64 codons, but only 20 building blocks," Robertson said. Robertson and his colleagues wondered if, by replacing redundant codons with their "synonyms," they could then reassign some of these redundant codons to code for new amino acids without killing the cell.

In a previous study, published in 2019 in the journal Nature, the team overcame the first hurdle in this challenge by creating a new strain of E. coli with a pared-down genome. Led by Jason Chin, a program leader at MRC-LMB and head of the Centre for Chemical & Synthetic Biology, the group swapped out all TCG and TCA codons for AGC and AGT, which all code for the amino acid serine. 

They did this using a technique called "replicon excision for enhanced genome engineering through programmed recombination," or just REXER for short. REXER can cut out large portions of the E. coli genome in a single step and replace the excised chunk with synthetic DNA, which in this case, used AGC and AGT in place of TCG and TCA. This process can be applied in a stepwise fashion, inching down the genome so that chunk after chunk gets replaced with synthetic DNA; in this way, the team expunged all instances of TCG and TCA from their E. coli strain.      

"If you're going to make a bunch of changes, it's actually more efficient to start from scratch and just build it bottom-up," rather than swapping codons one-by-one from the natural genome, Robertson said. The team also swapped the stop codon TAG for TAA, a synonymous stop codon, and thus freed up three codons for them to reprogram, since the cell no longer contained TCG, TCA or TAG. 

And despite having these three codons removed, the new strain of E. coli survived well in the lab environment, and the team selected for those cells that grew fastest in the cell culture. Cells that underwent this directed evolution grew reliably in lab dishes, although the modified E. coli would quickly die if placed outside the controlled lab environment, Robertson noted. 

A 'plug-and-play' system 

Now, in their most recent study, the team made one final tweak to their E. coli by deleting genes that code for two specific tRNA molecules — the molecules that read the codons and collect all the appropriate amino acids. These tRNAs would usually recognize TCG and TCA codons. The team also deleted genes for a so-called release factor that normally recognizes the TAG stop codon. These changes made the new bacterial strain invulnerable to viruses, the team found.

Virus genomes contain TCG, TCA and TAG codons, but without the right tRNA and release factors, the designer E. coli can't read these viral genes and therefore can't fall prey to the pathogens. "When the virus infects, it doesn't have the same genetic code as our [modified E. coli] cells, and then it can't make its own proteins and it can't propagate," Robertson said. 

But again, the main goal of the study was to reprogram the freed codons in order to generate new proteins. To do so, the team generated tRNA molecules that paired with unnatural amino acids of their own design; these tRNAs were programmed to recognize the TCG, TCA and TAG codons now missing from the modified E. coli strain. The team reintroduced the missing codons by placing them within small loops of DNA, called plasmids, which can be inserted into the bacterium without altering its genome.

The plasmids, tRNA and unnatural amino acids provided all the blueprints, tools and materials the cells needed to build designer proteins for the researchers. "So you can make proteins in a cell in a programmable fashion, based upon the DNA we provide to the cell, with 23 building blocks," rather than 20, Robertson said. "It's quite a plug-and-play system."

Other research groups have attempted to introduce unnatural amino acids into proteins in the past, but these strategies were not very efficient, Chatterjee and Delilah Jewel, a graduate student in Chatterjee's lab, wrote in a commentary published in the same issue of Science. For example, Chatterjee's lab successfully paired unnatural amino acids with the stop codons in E. coli, but this method only allowed them to insert these unnatural amino acids at a single site in the final protein, they reported in a 2019 study in the Journal of the American Chemical Society. 

Now, with the new method, scientists can begin pushing the boundaries of what proteins and polymers they can build, Chatterjee told Live Science. "It's kind of up for imagination. What could those amino acids look like?" he said. "What kind of chemistry could they have, functionalities could they have, that nature never had access to?" 

Looking into the future, scientists could potentially remove even more codons from the E. coli genome, freeing up even more channels for designer protein construction, Robertson said. But for now, three open channels are likely plenty to work with, he said. "Do we need seven open channels? Or is three open channels enough to really expand what we can do, in terms of providing new applications?" he said. "It's beneficial to just focus on the applications now." 

Originally published on Live Science.