Ribosome traffic flow on the mRNA as a regulator of cellular protein production: an integrated modelling and experimental analysis

In this proposal, an interdisciplinary team of biologists and physicists will establish novel technologies to predict the protein composition of a cell. Proteins are used by the cells within all organisms to carry out the essential biochemical processes that constitute life. Knowing which proteins and in what quantities are being made by a cell, defines the properties of that particular cell. Being able to predict the protein composition of a cell therefore represents a very powerful tool to understand cell biology. Proteins themselves are made of a string of chemical building blocks called amino acids, of which there are twenty different types. It is the distinct sequence of the amino acids in the protein chain that gives the protein its biochemical and catalytic properties. Even a relatively simple organism such as baker's yeast, the subject of this proposal, can have about 6,000 different varieties of protein, each with its own specific amino acid sequence. The cell makes proteins of the correct amino acid sequence using information encoded in its genes. Each gene codes for a single protein type, so baker's yeast has 6,000 genes encoding the same number of distinct proteins. To make a protein, the coding information in a gene is first copied into a short linear molecule termed a messenger RNA, or mRNA. Then an assembly of bio-molecules called ribosome reads the information within the mRNA, a process called translation. The ribosome moves along the mRNA from one end to the other, reading the information coded in the mRNA, and translating it by sequentially adding the amino acids to make a protein chain. The amino acids are brought to the ribosomes by transfer RNA molecules (tRNAs). The protein is then released to carry out its function in the cell. In fact, the mRNA can by translated by multiple ribosomes at the same time, with ribosomes following each other like cars down a road. This traffic analogy is rather apt; sometimes, just as cars get stuck in a traffic jam, so ribosomes can slow down or even pause completely as they translate the mRNA, usually in response to a section of the mRNA that is difficult to translate. When this happens, queues of ribosomes can build up, reducing the rate at which that protein is produced. All mRNAs are comprised of many different slowly and rapidly translated regions, for instance, caused by different abundances of distinct tRNA species. Ribosome queues can then begin to merge, sometimes extending back to the beginning of the mRNA and preventing ribosomes from joining the mRNA. This will reduce the amount of protein synthesis directed by that mRNA. Ribosomal traffic flow on mRNAs is therefore a key regulator of the quantities of the different proteins being made. To understand which population of proteins a cell will express, and in which quantities, therefore requires an ability to predict ribosomal traffic flow on the mRNA, and how whole populations of ribosomes interact with each of the 6,000 mRNAs in yeast. Predicting exactly how ribosomes interact and queue as they translate is a challenging task that requires joint application of both mathematical and biological techniques. In work leading up to this proposal, we have developed a mathematical model to simulate ribosome traffic on mRNAs. This model makes a number of important predictions about how ribosome traffic flow affects the translation of mRNAs, predictions that will be tested in this proposal. The proposed research will also develop the model much further, incorporating detailed mathematical descriptions of the translation process. The model will be tested and validated by experimentally analysing translation reactions in yeast. Overall, the interdisciplinary approach will not only provide genuine insight into the fundamental mechanisms a cell uses to express its genes, but will have implications for the study of many other traffic flow systems in Biology and Physics.

The relationship between cellular mRNA abundance and protein concentration is not fixed but mRNA-specific for all model organisms where comprehensive quantitative studies have been carried out. Major properties that govern this intricate relationship are the nature of the individual mRNAs, their codon composition, secondary structures, as well as the lengths and sequences of their leader regions where the ribosomes begin translation. All these properties give rise to complex dynamical interactions among the ribosomes translating any given mRNA, and profoundly influence the translational efficiency. In order to accurately predict the translational efficiency of the mRNA population, this project will for the first time develop and validate a complete stochastic model of translation. It will make use of computationally efficient and novel analytic methods, to model translation in the yeast Saccharomyces cerevisiae. The model will incorporate predictive descriptions of the initiation, elongation and termination stages, and include the key effects of mRNA structure and tRNA supply on the ribosome transit process. Experimental testing of engineered mRNAs will be used to validate the model predictions of the interplay between translating and queuing ribosomes in response to a complex pattern of slow and rapidly translated mRNA sequences. Finally, network modelling approaches will be applied to determine how the global cell translation apparatus responds to perturbations in the tRNA supply and demand ratios as the transcriptome is translated into proteome.