How genomic information is transcribed at high temporal resolution

Mammals like humans or mice use about 25’000 genes to carry out all the cellular functions necessary for life. The information contained in the DNA sequence of a gene is first transcribed into an intermediate molecule called messenger mRNA, which in turn is translated into a protein. Proteins are the actual “nano-machines” – also called “the cells’ work horses” - that carry out the specific functions. Different cell types assume specific functions in the body. For example liver cells specialize in metabolism and white blood cells in immune responses to bacterial or viral infections. The functions conducted by different cells, including also those that are necessary for the multiplication and survival, depend to a large extent on the concentration of each protein present in the cell. When mRNAs and proteins are determined in a large population of cells, for example the hundred million liver cells of a mouse, one finds very precise quantities for each gene product. However, when determined in individual cells, one observes a large variability between the same gene product in different cells and between different gene products in the same cell. This phenomenon, known as stochasticity in gene expression, has attracted considerable attention in recent years.

How is it possible that the gene activity in an entire organ is so precisely regulated when the individual cells and genes are doing their work so sloppily?

In a joint collaboration published in the journal Science, a team of researchers led by Prof. Schibler at the University of Geneva and Felix Naef at the Ecole Polytechnique Fédérale de Lausanne (EPFL) were able to measure in real time how individual genes are transcribed from their DNA template in single cells derived from the mouse skin. They observed that transcription takes place during short episodes of time, lasting only a few minutes. During these periods many copies of mRNA can be synthesized. After a gene has produced a certain number of transcripts, it returns back to a silent state before it initiates the next burst of productivity. The collaborative work of the two teams from the ‘Arc lémanique’ revealed that the durations of the activity and inactivity periods, and the number of mRNAs made during the activity periods, are very characteristic for a given gene. This reconciles the findings that genes can be stochastically active, and nevertheless produce precise and gene-specific quantities of mRNAs, when analyzed in a large number of cells.

To obtain these results, the team had to develop novel experimental methods and mathematical modeling tools. “The combination of genetic engineering with ultra-sensitive bioluminescence microscopy allowed us to monitor transcription in single cells at an unprecedented temporal resolution” says David Suter, a postdoctoral scientist and co-first author of the study. “Working with such precise data was very exciting as it allowed us to develop new computational algorithms to accurately characterize the temporal transcriptional patterns. ” added Nacho Molina, a physicist and the other co-first author of the study. All authors felt that thanks to the interdisciplinary teamwork, the outcome of the study exceeded by far the individual contributions. “It was the ideal collaborative project”, said both Prof. Naef and Prof. Schibler.

The concept that gene transcription occurs in discontinuous bursts is currently the subject of considerable attention, one reason being that the inherent randomness in this process may significantly contribute to phenotypic diversity.

Original text in EN: Prof. Felix Naef, EPFL, School of Life Sciences & Prof. U. Schibler, University of Geneva School of Biology