Flu viruses showing the protein coat surrounding their genetic material.
Synthetic biology has become a catch-all term for attempts to get organisms to do things they normally wouldn’t do. Efforts to date have ranged from assembling logic circuits in bacteria to replacing an entire bacterial genome with one synthesized from scratch. So far, however, the field has largely yielded some very impressive proofs-of-concept. Not much progress has been made with obvious practical applications.
That might change. Researchers have taken a technique that has been used a number of times before — manipulating cells to use an artificial amino acid — and applied it to create a flu virus that acts as a vaccine. The vaccine is very effective and because it relies on an amino acid that our cells don’t use, it can’t cause infections in us. The best yet: If the vaccine gets into cells with a normal flu virus, it interferes with its ability to generate an infection.
All our proteins are made with different combinations of the same 20 amino acids. While many additional amino acids exist, those 20 seem to be the standard toolkit shared by all life. There are some exceptional organisms that use a 21st, but these eccentric amino acids are usually close chemical relatives of existing ones.
Revision of the genetic code
In recent years, researchers have discovered how to get cells to use some of those extra amino acids, which may be chemically different from the existing 20. These artificial amino acids open up the possibility of proteins with dramatically different chemistry, which can catalyze different reactions or interact with chemicals that life normally does not have to deal with. So far, the results have been mostly potential.
However, a team from Peking University has been working on another use for artificial amino acids: making viruses that don’t work in normal cells. The idea is to make a vaccine from these semi-artificial viruses.
The logic here is extremely clever. Many vaccines contain only one or a few proteins of an infectious agent. But these are not always effective, as they lack the complexity and context of an intact virus or bacteria. The same may be true for a vaccine made from an inactivated virus. The alternative, using a weakened virus, carries the risk of people with weak immune systems contracting a full-blown infection.
How can artificial amino acids help with this? If you make a virus that depends on it, it can only reproduce in cells that supply that amino acid. Since that would exclude all our cells, using this virus as a vaccine poses no risk of developing an infection. To the immune system, it should look a lot like a normal virus, so it should be an effective vaccine.
Instead of focusing on an artificial amino acid with a radically different chemistry, the people behind the new work decided to use one that has a close chemical connection to an existing amino acid. And to make it easy, they chose one that has already been used by a microbe. This meant that all the genes needed to put the amino acid into proteins existed; they just had to be taken out of the microbe and put into cells that a flu virus could infect.
(The artificial amino acid in question is Nϵ-2-azidoethyloxycarbonyl-L-lysine. It is closely related to the normal amino acid lysine and chemically resembles another normal amino acid. Technically it is not artificial as it is naturally used by some microbes But it’s artificial in human cells, so we’ll keep using that term.)
The system is also effective in ensuring that genes that use the artificial amino acid do not function without it. The three-base code for the artificial one is UAG. Humans and most other organisms interpret that code as telling the cell to stop making the protein. So any gene with UAG in the middle will normally be made into a protein by cells with the microbial genes, but will be stopped early and produce a severely truncated version in normal cells.
The genes of the microbe were placed in a human kidney cell line. Tests with a fluorescent protein indicated that as long as it was supplied in the media used to feed these cells, the artificial amino acid would be incorporated into proteins.
Engineering the virus
The system was then tested with a flu virus. A single amino acid code in one of his genes was changed to UAG. When this version of the virus was put into normal cells, they didn’t produce any functional virus because that gene’s translation was prematurely terminated. But when it was placed in the kidney cells that carried the microbial system, the virus was produced normally. The resulting virus could infect other cells, but if they didn’t also have the microbial system, the infection would stop there.
That is, for the most part. Mutations occur at a constant rate, and some of these changed the UAG so that it coded for a different amino acid. If that change doesn’t deactivate the virus, it can infect normal cells again. The researchers saw exactly this happen: normally infectious viruses turned up at a low frequency during their tests.
The authors went back and tested 21 other different sites that they changed to UAG, targeting any amino acid that was chemically similar to the artificial one. Some of these completely disabled the virus; the artificial replacement was close, but not close enough. But seven of the changes produced a viable virus. And several of these can be combined, making the gene very resistant to these kinds of evolutionary changes. Gradually, they engineered UAGs into genes on each of the eight different RNA segments that make up the flu virus.
This virus would grow just fine in the kidney cells designed to carry the microbial system, and the resulting virus could infect other cells. But unless the cells it infected also carried the microbial system, the virus stopped there. Mature viruses have never been produced. And because it would take so many different mutations to return the virus to its original state, the virus remained dependent on the microbial genes to reproduce.
This was true when they tested it in animals. While a given amount of normal virus would kill half the mice it was injected into, they found that they could inject 100,000 times as much of the engineered virus and there would be no evidence of any health problems. However, the mice showed a robust immune response against the virus, one that was broader than that against a normal flu vaccine. The virus also worked as a vaccine in ferrets and guinea pigs.
The authors also tested what happened when cells were infected with both the engineered virus and a normal flu virus. It turned out that the manipulated version suppressed the normal version’s infection. Remember where we mentioned that the flu virus has a genome made of eight different RNA molecules? In cells infected with both viruses, the progeny was a mixture of segments taken at random from both sources. Thus, the vast majority of viruses produced contained at least one engineered segment and failed to successfully infect normal cells.
Much of the promise of synthetic biology seems a bit hand-waving – we can probably do something useful with this at some point. This included the use of artificial amino acids. Yes, they could potentially expand the chemistry of life, but it wasn’t clear that they would allow us to do things that normal amino acids couldn’t. Still, this is clearly a case where the artificial are at the center of biotechnology, and the applications are obvious.
Science2016. DOI: 10.1126/science.aah5869 (About DOIs).