Bacterial proteins that knock out cell defences: Cruise missiles or cannonballs?

A problem presented at the UK MMSG Oxford 2005.

Presented by:
Dr Gail Preston (Department of Plant Sciences, University of Oxford)
A Fletcher, G Mirams, G Preston, J Rougier, I Shvartsman, A Walter

Problem Description

Bacteria that cause diseases of plants and animals, such as the plague bacterium Yersinia pestis, are able to deliver proteins across membranes and into host cells. These "effector" proteins disable or modulate host defence mechanisms by interacting with specific host proteins. Effectors are structurally and functionally diverse proteins that act on a wide range of cellular targets. These include targets that are localized to different compartments within the cell such as the nucleus, outer membrane, cytoplasm and specific membrane-bound organelles. Bacteria can use effectors to disable a cell that is more than 1000x their own size. This raises the question of how a limited number of bacterial proteins injected at a specific location at the periphery of the cell are able to encounter target proteins with sufficient speed and efficiency to disable surveillance and defence mechanisms that are activated within minutes of contact between the invading bacterium and the cell under attack.

The MMSG investigation will examine the effect of cell architecture, spatial distribution and concentration of effectors and targets, and passive vs. active transport processes on the likelihood of effector-target contact. In particular, we would like to answer the following questions: (1) How important are active spatial localization mechanisms in enabling bacterial proteins to connect with host targets? (2) What experiments do we need to do to describe and interpret protein targeting mechanisms in bacteria-host interactions?

Download the full problem description

Study Group Report

In this report we have developed continuum and probabilistic models of the invasion of bacterial effector proteins into a host plant cell. The continuum model considered the simplified two species case of effector proteins and target proteins binding in a reversible reaction, with binding representing the subversion of target proteins. The model was used to suggest that cytoplasmic streaming aides effector proteins in their conquest of the host cell.

The first probabilistic model also considered the host cell\u2019s defence proteins and considered the case of a 'well-mixed' cell. We were able to show that the assumption of random collisions between effector proteins and their targets leads to an estimate of several weeks for the time for complete subversion of the host cell to occur. Since this far exceeds the observed time scale, the model provides further evidence in favour of the existence of a co-localisation mechanism exploited by effector proteins. Further evidence for this was provided by the second probabilistic model.

Download the full report