Hijacking the Sec machinery in bacterial warfare

Lead Research Organisation: University of Bristol
Department Name: Biochemistry

Abstract

All cells are surrounded by membranes, made up from a double layer of fatty molecules called phospholipids. Cell membranes act as a molecular "skin", keeping the cell's insides in, and separating different biochemical reactions. The barrier needs to be breached in a controlled manner to allow transport of nutrients, waste products and for communication with the outside world; this is achieved by a wide range of membrane-inserted proteins. We understand a great deal about the diverse biological functions that membrane proteins bestow, such as transport, respiration, photosynthesis. However, we know much less about how membranes are formed, or about the how new proteins are transported across or into membranes.

Our lab aims to understand more about how proteins are able to get in and out of the cell. Proteins, such as hormones and antibodies, are normally exported by cells via the secretory ('Sec' for short) machinery, which is essential for life. Simple bacterial cells also secrete proteins for many purposes: to form a protective cell wall, for survival and antibiotic resistance (AMR); to stick to surfaces; and to cause disease. This proposal concerns the discovery that bacteria also produce proteins that can enter into other bacterial cells, bypassing their membranes and cell wall. This activity is particularly important during bacterial competition, and helps determine which bacteria survive in bacterial communities such as the gut, and how they respond to external factors, e.g. the arrival of disease-causing bacteria. One of the weapons that bacteria deploy to gain the upper hand are Contact-Dependent growth Inhibitor (CDI) toxins. This project relates to a recent discovery that CDI toxins hijack the Sec system for import into rival cells.

The work will build on a current study analysing how proteins are exported. Most clinically relevant bacteria are surrounded by two membranes, each of which has their own export machinery. We have recently discovered that these machineries -Sec in the inner membrane and BAM in the outer membrane- interact directly with one another. These same two complexes are also hijacked by CDI toxins, so our hypothesis is that this interaction is important both for import as well as export.

The objectives of the project are to understand how this assembly is co-opted for toxin import through an analysis of its architecture and measurement of reversed protein transport (import) activity. To do so the project will harness complementary expertise in biochemistry and new breakthrough technologies in imaging by high-resolution electron cryo-microscopy. These scientific methods will illuminate how the CDI toxin hijacks the secretion machinery for its passage into the cytoplasm.

The results of the project will be important in terms of delivering new understanding of a fundamental process -protein trafficking- that spans the breadth of biology. Moreover, the information we gain could be further exploited. First of all, if we were able to copy and adapt the mechanism deployed by CDI toxins this would allow the delivery of bespoke proteins into bacteria -a feat that is currently very difficult to achieve. This could be useful, for example, for the delivery of toxic proteins as a strategy to kill specific bacteria for the development of next generation antibiotics. Additionally, a more benign application could be for the import of proteins designed to bestow new synthetic activities for technical innovation useful in academic research and for commercialisation.

Technical Summary

Protein transport is essential for life. Bacteria secrete proteins for a wide range of purposes, including cell wall biogenesis, pathogenicity and antibiotic resistance. Less well known is their ability to import proteins. In order to survive, bacteria deploy an array of different weapons, including the delivery of toxic proteins into rival cells, to gain the upper hand. One example is the import of toxins to confer Contact-Dependent growth Inhibition (CDI), which is very poorly understood. This phenomenon is important for bacterial competition and is a major force underlying the organisation and composition of communities, such as the gut microbiome, and how they respond to external factors, e.g. pathogens.

We have discovered the bacterial secretory machinery of the inner membrane makes contact with the barrel assembly machinery (BAM) of the outer membrane for efficient outer membrane protein delivery. Remarkably, this inter-membrane assembly appears to be hijacked for the import of CDI toxins through the bacterial cell wall and into the cytosol. The project will harness a powerful combination of biochemistry and cryoEM: we will deploy high-tech protein transport assays to monitor the import process in vivo and through membranes of in vitro reconstituted systems; alongside a structural analysis of the machinery associated with specialised import factors and clients. Our goal is to understand the structural dynamics of the interacting translocons of the inner and outer membranes, including their distinct action during import and export.

The results will reveal fundamental details of this extraordinary process, with far reaching consequences for our understanding of protein transport through the bacterial cell wall, as well as other interconnected membranes of eukaryotes, e.g. mitochondrial and ER membranes. The results could also suggest new strategies for protein delivery for synthetic biology and biomedical applications, such as antibiotic development.

Planned Impact

The overarching aim of the proposal is to understand important fundamental aspects of bacterial biology: protein transport during bacterial contact dependent growth inhibition (CDI). The immediate impact in terms of the current project will lie in scientific advancement and the generation of new knowledge. The project will also present new hypothetical concepts that if proven to be true will have a major impact in our understanding of the bacterial toxin delivery, as well as revealing new avenues for the exploitation of this activity.

The main areas of impact are:
1. Application and exploitation. While the proposed project is at a "pre-competitive" stage in terms of commercial exploitation, the knowledge generated will have an immediate benefit to both the National and International bioscience community (academic and commercial) in terms of understanding a fundamental process that spans the breadth of biology. The specific process of CDI is of fundamental importance for bacterial competition underlying the organisation and composition of bacterial communities such as the gut microbiome, and how they respond to external factors, e.g. pathogens. Thus, our understanding and exploitation of this machinery could have far reaching implications for improvement of human health, particularly for defence against bacterial infections. Moreover, these studies could lead to the exploitation of the protein import machinery for the delivery of toxins or new synthetic activities for technical innovation for academic research and commercialisation. Thus, the new knowledge gained could support an ongoing drug discovery programme (funded by the Wellcome Trust - University of Bristol institutional Translation Partnership award) seeking new strategies against AMR. Bristol has mechanisms in place to increase the impact of research and to exploit any commercialisation (see main impact plan).

2. Engagement. The benefits to the bioscience community are summarised above. The standard routes to information dissemination (e.g. pre-print submissions, papers in journals and presentations at conferences) will be used throughout the project. A more general benefit of our work to the UK stems from our commitment to public engagement. The PI and researcher Co-I routinely participate in public engagement activities, from school children to politicians, and for the promotion of science to women and girls. The group will continue with public engagement activities throughout the course of the project, using work generated from the project to exemplify the importance of research.

3. Environment and climate change: We are conscience of the impact our research on the environment, which we aim to minimise and encourage others to follow suit. We are recent recipients of a silver Lab Efficiency Assessment Framework (LEAF) award, recognising our continuing effort to adhere to excellent levels of environmental sustainability and research practice. The co-investigator researcher of this proposal is lobbying the head of School to implement a policy to offset the carbon for flights of Biochemistry staff, and the Collinson group has committed to this measure and to explore lab practices which reduce their overall CO2 footprint.

4. Staff training. The project will ultimately generate trained staff with desirable expertise in complex biochemical and biophysical analysis of membrane protein complexes involved in important bacterial activities. The researcher co-investigator will be in demand in both the academic and commercial sectors. During the project, further development will be encouraged through additional technical training, attending courses in areas directly and indirectly related to their role as research scientists (e.g. project management and leadership). By the end of the project we anticipate highly competitive applications to senior fellowships and/ or University staff positions completing his journey towards a fully independent researcher.
 
Description We have made a progress on all aspects of the project. The major challenge has been that toxin import appears to be a very low probability event, hence it is difficult to capture it in the process of transport. Nonetheless, we have some promising preliminary results: given that there is a year left on the project, and we now have established protocols for the major parts of the project, we are confident that we are on track to produce important novel insights into this system within the timeframe.

To start off with, we established a purification protocol for the C-terminal domain of the GN05224 CDI toxin (demarcated by the VENN cleavage motif), comprising the inner membrane import domain (IM), PG-rich linker and C-terminal toxin domain. Based on discussion with our collaborators in California, we settled on a construct with a his-tag at the N-terminus followed by a TEV protease cleavage site which, when cut, reveals the same N-terminal residue as would be present after the physiological cleavage event. When provided to non-immune bacterial cells with the outer membrane disrupted by polymyxin B, the purified construct results in rapid cleavage activity against the tRNAs in the cell.

Next, we looked at association between purified holo-translocon (the proposed inner membrane entry site) and CDI toxin. Glycerol gradient centrifugation suggests that the two do indeed associate, however we were unable to pull down the complex by chromatographically: most likely the affinity is too low. We have therefore pursued a cross-linking strategy. For this, we were aided by the release of Alphafold2, which provided a structural model both for the IM domain and toxin domain (the PG-rich linker is, unsurprisingly, unstructured). Based on the inner membrane domain structure and conserved features of other CDI toxins that utilise SecYEG for entry, we developed a working model for CDI toxin entry. Effectively, the inner membrane import domain binds into the periplasmic exit to SecY, bringing a conserved stretch of amino acids resembling a reversed signal sequence into the lateral gate of SecY, wedging it open. Presumably, this leads to passage across the membrane in a PMF-dependent manner, although the exact details of how this might work are not yet clear. We have modelled this interaction, and it remains stable in a molecular dynamics simulation, indicating that it is at least plausible.

In order for the CDI-SecY interaction to take place, the plug domain of SecY must first be displaced from the periplasmic cavity. This would explain the low binding affinity of CDI toxin for SecY. It also provides an explanation for the very specific ability of the S281F mutation of SecY to resist intoxication. We have shown by molecular dynamics simulations that the phenylanaline introduced in S281F forms a specific pi-stacking interaction with a phenylalanine in the plug domain of SecY, which prevents the plug from opening spontaneously. We have also confirmed by FRET and biochemical methods that the mutation does indeed prevent the plug from opening spontaneously; however it can still be opened from inside the cell, hence S281F does not prevent transport. Overall, we therefore have a very plausible entry mechanism for CDI toxin, and an explanation why a single, very specific point mutation can prevent it from gaining access.

Based on the model above, we attempted to crosslink the inner membrane import domain of the CDI toxin into SecY at a range of different sites using site-specific crosslinks (disulphide bonds between fixed positions in the model) and non-specific crosslinks (the photo-reactive crosslinker p-azido-phenylalanine, incorporated into CDI toxin by amber suppression). One of the disulphide bonds gave a clear single band on a western blot that reacts both to anti-SecY antibody and an anti-v5 tag antibody (when a v5 tag is added to the C-terminus of the CDI toxin), suggesting that this interaction does indeed take place. However, it has so far not proved possible to scale this up to the point where it can be purified for further biochemical and structural analyses. We are now looking at ways to increase the yield of this product, including non-specific labelling at the same site, and using SecY variants where the plug is pre-opened (which according to our hypothesis should promote the interaction).

Another major goal of the grant was to develop an assay to measure import. So far we have tried a number of constructs in different experimental setups. The most promising for producing a positive assay for import would be the development of a split luciferase assay (NanoLuc), already deployed in our laboratory for measuring protein transport. Thus, replacing the toxin domain with the small fragment of NanoLuc (pep86) and providing it to cells expressing the large complementary fragment (11S) in the cytosol, in the presence of polymyxin B nonapeptide (PMBN, which partially permeabilises the outer membrane without killing the cells). In this case we get a luminescent signal that requires both the addition of PMBN and the presence of the inner membrane domain on the Cdi toxin. However, as it stands the assay is very challenging to perform and has a very poor signal to noise ratio. It also only works with growing cells (at 37 °C) in a fairly specific window, which makes the reproducibility and throughput very poor. Furthermore, under the conditions it works, the kinetics of import appear to be limited by outer membrane permeabilisation - there is no obvious difference between wild-type SecY, S281F and S281F+?ppiD (which should completely obviate import). Thus, we are currently looking at alternative ways to perform the assay to get a more meaningful import data.

To get a positive measure of import that can be screened against to identify components that assist import (one key goal of the grant), we are looking to use the entire CDI toxin system expressed in host cells, with the toxin domain substituted for a pep86 (of the split luciferase) or split GPF complementation domain. Cloning of constructs for this is now nearly complete, but we have not yet performed the assay. To measure the kinetics of import in a purified system, we now also intend to switch to a FRET-based assay using an RNA substrate of the toxin domain labelled with two dyes at either end. We have so far produced the RNA substrate, and show that it provides a good signal when provided to the toxin domain. A variant of the toxin with a point mutation in one of the catalytic histidines (identified from the Alphafold2 structure) does not produce a signal, indicating that it is an effective, real-time measure of CDI toxin activity. We next intend to encapsulate the fluorescent substrate inside proteo-liposomes incorporating the HTL (possibly with other Sec components) to get an assay running with the native toxin domain. Ideally this would be at the single proteoliposome level (together with collaborators) to minimise material usage and maximise the kinetic information obtainable.

Finally, we are also looking at the other components of the SecYEG interactome implicated in assisting Cdi toxin import - PpiD, YfgM and SecDF. Since this grant started a paper has been published suggesting that these four proteins act together in a single pathway, interacting with some proteins as they emerge from the SecY channel; and thus presumably also with CDI toxin as it enters into the channel. These interactions can be rationalised with Alphafold2 structures of the entire complex. We have cloned these proteins and are currently working to optimise expression, with the aim of using them in the above assays.
Exploitation Route The results have revealed fundamental details of the extraordinary process of bacterial protein import. The new knowledge could suggest new strategies for protein delivery for synthetic biology and biomedical applications, such as antibiotic development.
Sectors Healthcare

Pharmaceuticals and Medical Biotechnology

 
Description Analysis of the mechanism of protein translocation by single molecule fluorescence with Profs Sheena Radford and Roman Tuma 
Organisation University of Leeds
Country United Kingdom 
Sector Academic/University 
PI Contribution Provision of expertise and material. Conducting in parallel ensemble analysis of protein transport machinery See joint BBSRC grants: Recently awarded: BB/T006889/1 (joint with BB/T008059/1) BB/N017307/1 (joint with BB/N015126/1) BB/I006737/1 (joint with BB/I008675/1)
Collaborator Contribution Single molecule expertise, experimental set up and data collection
Impact Yes, publications: Joel Crossley, Matthew A. Watson, Tomas Fessl, Daniel Watkins, Robin A. Corey, Tara Sabir, Sheena E. Radford, Ian Collinson, Roman Tuma. Energy landscape steering in SecYEG mediates dynamic coupling in ATP driven protein translocation. bioRxiv 793943; doi: https://doi.org/10.1101/793943. Submitted to JACS. Fessl T., Watkins D., Oatley P., Allen W.J., Corey R.A., Horne J., Baldwin S.A., Radford S.E., Collinson I. & Tuma R. (2018) Dynamic action of the Sec machinery during initiation, protein translocation and termination. eLife: 10.7554/eLife.35112 Allen, W. J., Corey, R. A., Oatley, P., Sessions, R. B., Baldwin, S. A., Radford, S. E., Tuma, R., and Collinson, I. (2016) Two-way communication between SecY and SecA suggests a Brownian ratchet mechanism for protein translocation. eLife. 10.7554/eLife.15598 Deville, K., Gold, V. A. M., Robson, A., Whitehouse, S., Sessions, R. B., Baldwin, S. A., Radford, S. E., and Collinson, I. (2011) The oligomeric state and arrangement of the active bacterial translocon. J. Biol. Chem. 286, 4659-4669
 
Description Analysis of the mechanism of protein translocation by single molecule fluorescence with Profs Sheena Radford and Roman Tuma 
Organisation University of South Bohemia
Country Czech Republic 
Sector Academic/University 
PI Contribution Provision of expertise and material. Conducting in parallel ensemble analysis of protein transport machinery See joint BBSRC grants: Recently awarded: BB/T006889/1 (joint with BB/T008059/1) BB/N017307/1 (joint with BB/N015126/1) BB/I006737/1 (joint with BB/I008675/1)
Collaborator Contribution Single molecule expertise, experimental set up and data collection
Impact Yes, publications: Joel Crossley, Matthew A. Watson, Tomas Fessl, Daniel Watkins, Robin A. Corey, Tara Sabir, Sheena E. Radford, Ian Collinson, Roman Tuma. Energy landscape steering in SecYEG mediates dynamic coupling in ATP driven protein translocation. bioRxiv 793943; doi: https://doi.org/10.1101/793943. Submitted to JACS. Fessl T., Watkins D., Oatley P., Allen W.J., Corey R.A., Horne J., Baldwin S.A., Radford S.E., Collinson I. & Tuma R. (2018) Dynamic action of the Sec machinery during initiation, protein translocation and termination. eLife: 10.7554/eLife.35112 Allen, W. J., Corey, R. A., Oatley, P., Sessions, R. B., Baldwin, S. A., Radford, S. E., Tuma, R., and Collinson, I. (2016) Two-way communication between SecY and SecA suggests a Brownian ratchet mechanism for protein translocation. eLife. 10.7554/eLife.15598 Deville, K., Gold, V. A. M., Robson, A., Whitehouse, S., Sessions, R. B., Baldwin, S. A., Radford, S. E., and Collinson, I. (2011) The oligomeric state and arrangement of the active bacterial translocon. J. Biol. Chem. 286, 4659-4669
 
Description CDI toxin import with Profs David Low and Dr Sanna Koskiniemi 
Organisation University of California, Santa Barbara
Country United States 
Sector Academic/University 
PI Contribution Cooperation on the mechanism of CDI toxin import
Collaborator Contribution Discussions and sharing of reagents
Impact Joint paper submitted to PNAS Named collaborator (Prof David Low) on pending BBSRC grant application (BB/V001531/1)
Start Year 2019
 
Description CDI toxin import with Profs David Low and Dr Sanna Koskiniemi 
Organisation Uppsala University
Country Sweden 
Sector Academic/University 
PI Contribution Cooperation on the mechanism of CDI toxin import
Collaborator Contribution Discussions and sharing of reagents
Impact Joint paper submitted to PNAS Named collaborator (Prof David Low) on pending BBSRC grant application (BB/V001531/1)
Start Year 2019
 
Description Computational analysis of the bacterial translocon with Dr R. Corey 
Organisation University of Oxford
Country United Kingdom 
Sector Academic/University 
PI Contribution Expertise in empirical analysis of the bacterial secretion machinery
Collaborator Contribution Expertise for computational analysis of the bacterial secretion machinery
Impact None so far
Start Year 2018
 
Description Mass spectrometry of the bacterial Sec machinery with Dr A. Politis 
Organisation King's College London
Department Department of Chemistry
Country United Kingdom 
Sector Academic/University 
PI Contribution We are conducting experiments using mass spectrometry to characterise the properties of the bacterial secretion machinery. We provide the material for analysis.
Collaborator Contribution Mass spec equipment and expertise
Impact In progress
Start Year 2017
 
Description Protein Biophysics of protein transport apparatus with Dr T. Fessl and Prof. R. Tuma 
Organisation University of South Bohemia
Country Czech Republic 
Sector Academic/University 
PI Contribution Provision of samples for biophysical analysis, especially for single molecule applications
Collaborator Contribution Biophysical analysis of protein transport apparatus, including single molecule applications
Impact Joel Crossley, Matthew A. Watson, Tomas Fessl, Daniel Watkins, Robin A. Corey, Tara Sabir, Sheena E. Radford, Ian Collinson, Roman Tuma. Energy landscape steering in SecYEG mediates dynamic coupling in ATP driven protein translocation. bioRxiv 793943; doi: https://doi.org/10.1101/793943. Submitted to JACS. Corey, R. A., Ahdash, Z., Shah, A., Pyle, E., Allen, W.J., Fessl, T., Lovett, J.E., Politis, A. and Collinson, I. (2019) ATP-induced asymmetric pre-protein folding as a driver of protein translocation through the Sec machinery. eLife: 10.7554/eLife.41803 Fessl T., Watkins D., Oatley P., Allen W.J., Corey R.A., Horne J., Baldwin S.A., Radford S.E., Collinson I. & Tuma R. (2018) Dynamic action of the Sec machinery during initiation, protein translocation and termination. eLife: 10.7554/eLife.35112
Start Year 2018