How bacteria replicate their DNA in spite of barriers, one molecule at a time

If the information necessary to make and maintain a living organism were contained inside a book, then DNA would represent the book's letters - it is, in effect, the alphabet of life. A key feature of any organism is the ability to replicate these letters, either to make more of itself or to create progeny by creating daughter cells. This crucial process of copying DNA is "DNA replication". It involves molecular nanomachines, which run along DNA like a car runs along a road, forcing open its double-helix, then making copies of the separate strands.

However, many molecular barriers exist to the efficient running of these nanomachines potentially causing collisions that if the cell did not correct could be lethal. But what happens with such a replication nanomachine at collisions? Do they drop off, fall apart, are they helped to push through the barrier, or are new nanomachines built the other side of the barrier that can then move on unimpeded?

Cells have evolved a suite of remarkable strategies that allow these nanoscale collisions to be repaired so that DNA replication can still occur. Other scientists have done great work previously in helping us to understand how DNA repair processes are achieved, but most of them have studied populations of many cells looking at an average of many molecules, instead of just individual nanomachines in single cells, mainly because the technology available to look at molecules in cells has not been good enough - until now! Here we will use the bacterial model organism Escherichia coli in which we can visualise and track individual nanomachines. We will also purify DNA and the replication nanomachines and barriers in the test tube and look at these at a single-molecule in the microscope. These two complementary approaches will really allow us to piece together observations of the molecular scale processes of just the purified components as well as what actually happens inside real, complex cell environments. This will allow us to see in exquisite detail when replication nanomachines collide, and how other molecules then respond to repair the collision. This revolutionary approach will be combined with new, exciting analysis of our real time "collision movies", with Artificial Intelligence or "AI" - this software consists of complex layers of interacting code, similar to the ways that nerve cells in the brain link together in the visual cortex. Each such nerve performs specific maths operations on inputs and passes the outputs to nerve cells in the next layer. In doing so a network of nerve cells can be "trained" to recognise key features and patterns from images, which can be really useful for the relatively noisy image data that we have in single-molecule microscopy both in test tubes and in living cells, to tell us where different molecules are and how they interact with each other.

Our work will tell us what helps the replication nanomachines back on the DNA road if they have been blocked by an obstacle or even pushed off. Also, very importantly, it will allow us to understand how antibiotics which target DNA replication and repair actually work in cells. This will be important information, since many so-called "super-bugs" are emerging which no longer respond to antibiotics, and so this may aid other researchers in being able to design new types of better antibiotics. Following single molecules and establishing better techniques, as we aim to do here, will enable basically any scientist involved in DNA replication in any cells or organism to improve their work.

Activities of DNA polymerases in replication result in collisions which, if remaining unresolved, can be lethal and therefore must be resolved efficiently. Extensive knowledge exists from genetics and biochemistry about the enzymes involved, but we know little of how this is achieved at the level of single molecules. Here we integrate interdisciplinary expertise, using E. coli as a model to study replication collisions and subsequent crucial repair, modifying cells to label key proteins used in replication used in replication and repair. We will use artificial blocks to controllably mimic collisions both in vitro and in living cells and native blocks involving RNAP from transcription that will allow us to study collisions in defined areas of the chromosome in any orientation. Some replisomes remain following collisions while others disassemble. By using labelled replisomes and blocks probed with super-resolved single-molecule microscopy, microfluidics and novel AI biocomputation tools, we will be able to define the conditions when forks disassemble under physiological conditions. We will then visualise which repair proteins are recruited and, finally, be able to characterise how restart proteins can re-recruit active replisomes to continue synthesis. We will also search for hitherto undiscovered factors that assist in vital replication collision resolution.

Our analyses will address fundamental questions concerning the resolution of collisions between replication molecular machinery and nucleoprotein blocks. DNA replication and repair offer key antibiotic targets, and we anticipate our findings will have longer term societal benefit in addressing how poisons targeting these processes can be tolerated by cells and lead to antibacterial resistance, aiding development of new antibiotics in addition to substantive development of new microscopy instrumentation, bioinformatics and high-precision analytical software tools of wide benefit to the biosciences.