Protein import through the E. coli cell envelope

Lead Research Organisation: University of Oxford
Department Name: Biochemistry


We are rapidly running out of antibiotics for use in medicine and agriculture due to the inexorable rise of antibiotic resistant bacteria. The failure of drug discovery programmes in the pharmaceutical industry to furnish any new classes of antibiotics in over 30 years means our society is rapidly approaching a time akin to the pre-antibiotic era where routine operations carried a high degree of risk from potentially lethal infections, a view emphasized by recent reports on antimicrobial resistance from across the globe. Moreover, protecting our food against bacterial pathogens is an increasing concern; for example, contamination of food products by E. coli O157:H7 in 2011 across Europe resulted in multiple deaths.

Clearly, new approaches are needed to respond to this problem. A group of molecules that have yet to be exploited in the treatment of bacterial infections are protein bacteriocins. Bacteriocins are protein antibiotics produced by bacteria to kill their close related neighbours during competition for nutrients. This proposal focuses on one class of bacteriocins known as colicins. Colicins are made by the bacterium Escherichia coli to kill other E. coli, and have recently been shown to be highly effective at controlling E. coli O157:H7 infections in food crops.

My proposal is focused on understanding how these toxic proteins, which have no activity against human cells, circumvent the defences of bacteria (known as the 'cell envelope') in order to kill them. As our model, we use the non-pathogenic organism E. coli K-12. We have recently made two major advances in understanding how colicins translocate into E. coli (translocation is the process by which they move from the outside of the cell to the inside). First, we have isolated complexes of colicins bound to their cell surface targets, which primes them for entry into the cell, and have obtained preliminary data to suggest that we will be able to determine three dimensional structures for such translocation-competent states. Second, we have devised new fluorescence-based microscopy tools for visualizing the entry of these molecules into bacteria in real time, allowing us to address questions that have until now been impossible to investigate.

The proposal has three specific objectives:

1. To use crystallographic methods to determine the three dimensional structures of colicins bound to proteins that constitute the translocation machinery in E. coli;
2. To use colicins impregnated with photoreactive cross-linking groups (which can be engineered genetically) as a means of trapping colicins as they pass through the cell envelope when cells are illuminated with UV light. This approach will allow us to follow the path taken by these molecules through the cell envelope and identify proteins they come into contact with;
3. To exploit the microscopy tools we have developed to track colicins as they pass through the cell envelope and dissect their mechanism of entry using protein engineering.

Technical Summary

The cell envelope of Gram-negative bacteria is a robust, adaptive ~30 nm structure that supports the colonization of a broad range of environments. Escherichia coli, for example, can grow in open water and soil as well as colonizing humans, animals and plants as a commensal or pathogen. The importance of the cell envelope is underscored by the fact that it is the site of action of many common antibiotics. Yet our understanding of cell envelope organisation and in particular how active processes at the outer membrane are driven is poor. I aim to address these issues through colicin import. Colicins are E. coli-specific bacteriocins (protein antibiotics), released by E. coli populations to kill competing bacteria, which translocate a cytotoxic domain across one or both membranes of the cell envelope.

We have found ways of trapping and crystallizing colicin translocons, which are partially-translocated states containing both outer membrane and periplasmic proteins. In addition, using fluorescence microscopy, we have, for the first time, imaged single colicin molecules entering bacteria. These assays have also demonstrated that import across the outer membrane is dependent on the proton-motive force across the inner membrane, much speculated on in the literature but never before demonstrated. These new methods are the foundation of the present proposal which aims to address the central question of how colicins translocate through the E. coli cell envelope. The proposal has three objectives:

1. Determining the architecture and structure of outer membrane colicin translocons and/or receptor complexes using X-ray crystallography or chemical cross-linking.
2. Following the translocation paths of colicins by photoactivated cross-linking using colicins impregnated with non-native amino acids and synchronised for cell entry using disulfide bonds.
3. Using fluorescence microscopy to dissect the colicin import mechanism at the single molecule level

Planned Impact

This research will benefit two main constituents:

1. The biotechnology and agricultural industries
While the research will determine how colicin molecules translocate toxic domains into bacteria it will also show how they can be used as Trojan horses to deliver other molecule types into the cell. Thus far we have shown that fluorophores can be delivered by this route into E. coli. The research will establish whether, as I postulate, our colicins of choice are delivering these molecules to distinct cellular compartments (periplasm or cytoplasm). Through this research we will be exploring what variety of organic fluorophores can be translocated via colicins. Longer term, we will evaluate the viability of this approach for the delivery of other types of molecule (e.g. antibiotics) into bacteria. A long-term vision for this type of work is the development of a protein-based transformation system in which colicins can be used to deliver molecules to different cellular compartments of a Gram-negative bacterium. In addition, the recent demonstration that colicins can reduce the burden of E. coli O157:H7 in a variety of cultivated crop species (Schulz et al (2015) PNAS) puts a spotlight on their mode of action. This programme of research will elucidate how they enter and kill bacteria, opening up the possibility of engineering these molecules so that, for example, resistance can be avoided in plant-based antimicrobials.

2. Society
While the present study is a fundamental investigation into how colicins are imported across the cell envelope of E. coli it lays the foundations for the use of these molecules as protein antibiotics in the future. It is notable that major pharmaceutical companies (e.g. GSK) have switched their antibiotic discovery programmes entirely to protein-based therapeutics (mostly antibodies) since small molecule target-led drug discovery has failed to deliver new antibiotics (at a cost of £billions). In order to exploit colicins (or colicin-like proteins) we need to understand how they work. This study is designed to reveal these mechanisms. Society needs new approaches (and new molecules) with which to treat multidrug resistant bacteria, especially Gram-negative bacteria for which there are few options if an organism is resistant to major classes of antibiotics such as the carbapenems. This is a significant problem in nosocomial infections where even resistance to colistin, a last resort polymyxin used in the treatment of Gram-negative infections, is beginning to emerge. Society needs to have a range of approaches and strategies in development. This study provides one such an approach for the future.
Description There have been two major outcomes, both currently being prepared for publication. The submission of these outputs has been delayed by the pandemic. The first relates to our use of fluorescence microscopy to follow import of bacteriocins (colicins specifically) into bacteria. We have developed methods to follow two different colicins into their specific compartments in an E. coli cell, colicin B into the periplasm and colicin D into the cytoplasm. The latter activity is associated with a structured domain known as the pyocin S domain. We have also shown that colicins can be used to transport ssDNA across the outer membrane of E. coli. The second discovery relates to the structure of the colicin E9 translocon. This is an assembly of colicin e9 bound to two outer membrane proteins (OmpF and BtuB) and a periplasmic protein (TolB). We have solved the structure for this complex by cryo-EM, which is the first for any colicin in a partially translocated state. The structure validates our past biochemical work (showing how colicins exploits porins) but remarkably shows that porin threading (the process by which a disorder domain threads two domains of OmpF) causes the colicin to dissociate from its primary receptor, BtuB. We have also shown that large fluorophores can be transported through E. coli porins by colicins. Both papers have now been published, in mBio and EMBO J.
Exploitation Route Both pieces of work show that colicins can be used to transport large molecules across the Gram-negative outer membrane. This has implications for delivering large antibiotics that are normally excluded from such organisms.
Sectors Agriculture, Food and Drink,Healthcare

Description Molecular dynamics simulations of outer membrane proteins 
Organisation University of Southampton
Country United Kingdom 
Sector Academic/University 
PI Contribution My laboratory have been investigating two aspects of outer membrane biology that are relevant to this collaboration. First, the transport of bacteriocins through the outer membranes of Gram-negative bacteria. Second, the mobility of outer membrane lipoproteins. We exploit the simulations from the Southampton group to design new experiments. Two publications thus far have resulted from this collaboration.
Collaborator Contribution I established a new collaboration with Prof Syma Khalid in the Department of Chemistry in Southampton. Prof Khalid's groupd have been conducting molecular dynamics simulations on bacterial outer membrane proteins and lipoproteins for comparison to experimental data collected in my laboratory.
Impact Housden et al (2018) Biochemistry 57, 4374 Szczepaniak et al (2020) Nature Communications In Press
Start Year 2017