With the help of AI, researchers at Chalmers University of Technology, Sweden, have succeeded in designing synthetic DNA that controls the cells’ protein production. The technology can contribute to the development and production of vaccines, medicines for serious diseases and alternative dietary proteins much faster and at significantly lower costs than today.
How our genes are expressed is a process fundamental to the functionality of cells in all living organisms. Put simply, the genetic code in DNA is transferred to the messenger RNA (mRNA) molecule, which tells the cell factory what protein to produce and in what amounts.
Researchers have made great efforts to control gene expression because, among other things, it can contribute to the development of protein-based drugs. A recent example is the Covid-19 mRNA vaccine, which instructed the body’s cells to produce the same protein found on the surface of the coronavirus. The body’s immune system could then learn to form antibodies against the virus. Likewise, it is possible to teach the body’s immune system to defeat cancer cells or other complex diseases by understanding the genetic code behind the production of certain proteins.
Most of today’s new drugs are protein-based, but the techniques used to make them are both expensive and slow because of the difficulty in controlling how the DNA is expressed. Last year a research group at Chalmers led by Associate Professor of Systems Biology Aleksej Zelezniak took an important step towards understanding and controlling how much of a protein is made from a given DNA sequence.
“First of all, it was about being able to ‘read’ the instructions of the DNA molecule completely. Now we have managed to design our own DNA, which contains the precise instructions to control the amount of a specific protein,” says Aleksej Zelezniak about the latest important work of the research group Breakthrough.
DNA molecules made to order
The principle behind the new method is similar to when an AI generates faces that look like real people. By learning what a wide range of faces look like, the AI can then create entirely new but natural-looking faces. It’s then easy to change a face, for example by saying that it should look older or have a different hairstyle. On the other hand, without the use of AI, programming a believable face from scratch would have been much more difficult and time-consuming. Similarly, the researchers’ AI was taught the structure and regulatory code of DNA. The AI then designs synthetic DNA where it is easy to modify its regulatory information in the desired direction of gene expression. Simply put, the AI is told how much of a gene is wanted and then it “prints” the appropriate DNA sequence.
“DNA is an incredibly long and complex molecule. It is therefore experimentally extremely challenging to make changes to it by iteratively reading and changing it and then reading and changing it again. That way, it takes years of research to find something that works. Instead, it is far more effective to have an AI learn the principles of DNA navigation. What used to take years is now reduced to weeks or days,” says first author Jan Zrimec, research associate at the National Institute of Biology in Slovenia and former postdoc in the group of Aleksej Zelezniak.
The researchers developed their method in the yeast Saccharomyces cerevisiae, whose cells resemble mammalian cells. In the next step, human cells are used. Researchers hope their advances will have an impact on the development of new and existing drugs.
“Protein-based medicines for complex diseases or alternative sustainable food proteins can take many years and be extremely expensive to develop. Some are so expensive that it is impossible to generate a return, making them commercially unviable. With our technology, it is much more efficient to develop and manufacture proteins in order to be able to commercialize them,” says Aleksej Zelezniak.
Relation: Zrimec J, Fu X, Muhammad AS, et al. Controlling gene expression through deep generative design of regulatory DNA. Nat Commun. 2022;13(1):5099. doi: 10.1038/s41467-022-32818-8
This article was republished from the following materials. Note: The material may have been edited for length and content. For more information, please refer to the given source.