Do translational kinetics influence protein folding?

Prof. Charlotte Deane University of Oxford

In many fields, research is carried out in both the experimental laboratory and on computers, but in some cases divides appear between the researchers in terms of the theories that they use. A case in point is protein structure prediction and protein folding. Biochemists and biologists who work on proteins daily take for granted the idea that the production of a protein will affect its folding and structure. They point to a wealth of structural, biochemical and biophysical evidence. This idea however is almost universally neglected by structural bioinformaticians. Such computational researchers argue that only a protein’s sequence is required in order to predict its structure or to understand how it folds.

The Comparative Analysis of Protein Structure Prediction experiment is a striking example of the divide in action. The experimentalists generating structures consider procedures such as codon harmonisation and various systems of expression in order to aid protein production and folding. The computational prediction community use only the amino acid sequence - no indication of the mRNA or the organism.

Recent research has begun to highlight a link between the coding mRNA and the protein structure. The development of the Coding Sequence and Structure database has provided a detailed mapping between solved protein structures and their mRNA. It has facilitated a comprehensive analysis of codon usage over many organisms. Analysis of this data has found that local protein structure information is encoded in the mRNA sequence. Many codons are found which have structural preferences significantly different from the amino acid that they encode, supporting the premise that codons encode more information than merely amino acids.

Protein structure and mRNA potentially interact at the ribosome via the process of co-translational protein folding. It is theorised that codons affect protein structure via their speed of translation, by modulating the time available for protein folding.

However, despite the clear evidence for a link between codons and local structural features, links between translation speed and structure are less clear. This may be due to the use of relative codon usage measures as a proxy for translation speed. Recent work has shown that experimentally determined tRNA concentrations offer a far better proxy for translation speed and that these concentrations are not correlated with measures of relative codon usage.

When using tRNA concentration as a proxy for speed no evidence was found that domain boundaries are enriched with slow codons. In fact, genes seemingly avoid slow codons around structurally defined domain boundaries. Translation speed, however, does decrease at the transition into secondary structure.

These computational indications of the importance of codons in protein structure formation can now be put alongside the wealth of structural, biochemical and biophysical evidence supporting the theory that co-translational protein folding as a vital mechanism to increase folding efficiency in a crowded cellular environment.

There is still much to be learned about the fundamental mechanisms of co-translational folding but it is clear that there is a link between the kinetics of translation and protein structure formation. Thus computational modellers should look more closely at the biology and kinetics of protein production processes when attempting to model protein structures.


  1. PMID: 21567957