Final Report Summary - SMILE (Single Molecule Investigations in Living E. coli)
Historically, the interior of the living cell has been a place of mystery, and most of what we have learned about biological processes has been deduced from experiments with purified components in test tubes. With increased microscopy power it eventually became possible to get a snapshot of this tiny organic factory, and during the last decade, technological progress has enabled the study of biochemical reactions as they happen, inside the living organism, one molecule at the time. In this project, we have made progress towards increased resolution in time and space by developing new imaging modalities. For example, we use light polarization – yea, it’s the same phenomenon that is used in polarizing sunglasses – to look at tiny molecular movements.
With these techniques, we can study basic biological phenomena such as replication of the genome and gene regulation. The fundamentality of these processes makes them at the same time scientifically interesting and notoriously hard to study as any intervention inevitably has severe consequences, both in terms of cell viability and secondary effects on the system. The complexity makes computational models really useful. The models can be used to make predictions of what effect different perturbations will have on the system, predictions that we can subsequently test in the experimental set-ups to determine if the models are plausible.
One of the conceptually interesting questions where we have made progress is the genome search problem. Molecules that regulate gene expression or are used for gene editing face a truly prodigious task when locating a specific piece of genetic code among the myriads of similar sequences that constitutes the genome. Imagine trying to find one sentence of text in a document containing 4.5 million characters, or about 1500 pages, and there’s no search function! Not surprisingly, the molecules that specialize in genome search are really optimized for their tasks, sliding along the DNA helix to speed up the process and being just sloppy enough to make the search both fast and accurate. The proteins that have evolved to find only one genetic target can do so in as little as 4 minutes. The search engine of the CRISPR/Cas9 gene-editing tool, on the other hand, has an even more terrifying mission of finding any genetic sequence depending on how it has been programmed. Cas9 cannot slide on the outside of the DNA since it has to open up the helix to compare the DNA sequence with its program code and thus it takes this molecule on average 6 hours to find the target. This is impractical in a gene-editing perspective, and we anticipate that this information can be the starting point to develop more efficient gene-editing tools.
With these techniques, we can study basic biological phenomena such as replication of the genome and gene regulation. The fundamentality of these processes makes them at the same time scientifically interesting and notoriously hard to study as any intervention inevitably has severe consequences, both in terms of cell viability and secondary effects on the system. The complexity makes computational models really useful. The models can be used to make predictions of what effect different perturbations will have on the system, predictions that we can subsequently test in the experimental set-ups to determine if the models are plausible.
One of the conceptually interesting questions where we have made progress is the genome search problem. Molecules that regulate gene expression or are used for gene editing face a truly prodigious task when locating a specific piece of genetic code among the myriads of similar sequences that constitutes the genome. Imagine trying to find one sentence of text in a document containing 4.5 million characters, or about 1500 pages, and there’s no search function! Not surprisingly, the molecules that specialize in genome search are really optimized for their tasks, sliding along the DNA helix to speed up the process and being just sloppy enough to make the search both fast and accurate. The proteins that have evolved to find only one genetic target can do so in as little as 4 minutes. The search engine of the CRISPR/Cas9 gene-editing tool, on the other hand, has an even more terrifying mission of finding any genetic sequence depending on how it has been programmed. Cas9 cannot slide on the outside of the DNA since it has to open up the helix to compare the DNA sequence with its program code and thus it takes this molecule on average 6 hours to find the target. This is impractical in a gene-editing perspective, and we anticipate that this information can be the starting point to develop more efficient gene-editing tools.