Tobias von der Haar, Professor of Systems Biology at the University of Kent discussed his work on sequence design for efficient and manufacturable RNA therapeutics. Before RNA is introduced into a patient it must be made into a therapeutic protein and then protein should have the desired effect in a patient.
Von der Haar said: “First, the RNA has to engage the translational machinery and make a protein, then the protein has to engage the immune system and generate a therapeutic effect. You need to get both parts right in order for the RNA therapeutic to work.”
Getting both parts right requires very specific expertise, von der Haar’s focus is on synthesising the protein from the RNA rather than generating therapeutic effects from the protein. The principal challenge of designing efficient RNA sequences is that for every individual protein, there is an insurmountable number of RNA sequences available that could be used to encode that protein. Depending on the length of the protein sequences the number of possible proteins that can be generated is large, for example, SARS-CoV-2-spike has 1273 amino acids which equates to 10632 possible DNA or RNA sequences. Different RNA sequences can result in varying levels of protein production, which impacts the effectiveness of RNA therapeutics.
To tackle this challenge, one must aim to have sequences that are decoded rapidly and efficiently. In order to ensure optimal decoding, von der Haar emphasised the importance of selecting the right codons and adjusting the GC content of the sequences which is a critical parameter for DNA synthesis. The parameters are usually interdependent meaning it is complicated to control one parameter without the others being affected.
Von der Haar proposed using nano luciferase to generate sequences by individual transcription. Changing coding choices in the open reading frame affects expression. Manufacturability of RNA therapeutics can be controlled through sequence design, for instance, T7 RNA polymerase is used to generate in vitro transcribed RNAs. RNA sequence design can also control off-pathway effects.
Genetic code functions because ribosomes read RNAs in three nucleotide codons, and if this happens correctly then the intended protein sequence will be produced. However, ribosomes can malfunction and lose this reading frame causing it to generate a different peptide from the intended peptide which could trigger an unwanted immune response, this phenomenon is called frameshift peptides.
Von der Haar’s group is working on building software that automates sequence optimisation for RNA vaccines and therapeutics. He stated: “Our aim now is to try and encode this knowledge in software that allows anyone to make use of it and to design good and efficient RNAs for a specified cargo protein that they want to deliver.”
Looking ahead the goals are to develop tools to predict therapeutic outcomes from RNA designs, beyond maximizing expression levels and bridge the gap between in vitro optimization and in vivo efficacy. This research represents a major step in using RNA sequence design in advancing RNA therapeutics.