Crystal structure of the earthworm toxin reveals the mechanism of assembly and its potential application
Biological membranes are built of sheets of lipid bilayers including various forms of proteins and sugars. They protect cells from their environment as well as enable many important biological processes. Intact membranes are vital to the cells, thus the destruction of the membrane is one of the simplest ways to harm cells. Such destructive processes are used by many organisms, which by using membrane active compounds attack other organisms or defend themselves against them. Targeted cells are usually unable to repair such membrane damage, leading to necrosis or apoptosis and finally to the death of the cell or even organism. One of the most effective ways to damage membranes is to form pores into the lipid bilayer. Pore forming compounds are almost exclusively toxic peptides or proteins, which are expressed by the producing organism as water-soluble monomers, attracted by the membranes via specific receptors. The most studied pore-forming proteins are various families of bacterial toxins, termed pore-forming toxins (PFTs). Bacterial PFT serve as important virulence factors promoting bacterial spread through invading cells and tissues. However, PFTs are not limited to bacteria, and are found in many other organisms.
In this study we have determined the three dimensional atomic structure of the toxin from the earthworm Eisenia fetida, called lysenin. Lysenin is present in the coelomic fluid of earthworms and is produced to act defensively against parasitic microorganisms by forming pores. The toxin is excreted by the defending cells, which are part of the immune system of the earthworm. It then binds to cellular membranes and forms pores in them. Such damage usually leads to the death of the target cell. It is extremely successful in the case of eukaryotic target cells which membranes contain a special compound called sphingomyelin. Lysenin resembles in structure and function some bacterial toxins, especially those serving as crucial virulence factors in infection and food poisoning, and together they form an aerolysin family of pore forming proteins. The first and most studied member of this family is aerolysin from Aeromonas sp. Other members of aerolysin family can be found mainly in bacteria (Clostridium and Bacillus), as well as in plants, fungi and animals. While some of these proteins have been known for more than 20 years, it is still not understood how they damage cellular membranes. This also prevents development of new antibiotics and other therapies against bacterial infections.
Our study has shown that lysenin forms extremely compact and stable pore of 301 kDa on the cell surface of a nanoscale size, built by nine copies of the same molecule (Figure 1).
Figure 1. Crystal structure of the lysenin pore, PDB-ID 5EC5, (a) shown in ribbons, side view, (b) top view. Each of the nine protomoeric units is presented in a different color. Gray lines in (a) indicate the position of the membrane. (c) Electrostatic surface of the pore, side view. Cutoff - 5 kTe-1 was used for negative potential (red) and + 5 kTe-1 for positive potential (blue). We have shown that this pore can translocate DNA. (d) 2Fo-Fc (contoured at 1 σ) electron density map (gray mesh) along the central β-barrel of the pore. For clarity, only Cαatoms are shown (green skeleton) in the electron density.
The structure of a mushroom shaped pore has been solved at 3.3 Å resolution using X-ray crystallography. X-ray diffraction data were collected at Elettra synchrotron XRD1 beamline, and the structure determined at the National Institute of Chemistry, Slovenia. Crystallographic data were supplemented by electron crystallography and mass spectrometry at Oxford University, UK, and atomic force microscopy (AFM) (Riken Institute, Japan). In addition to these methods, molecular biology, biochemistry and molecular dynamics tools (National Institute of Chemistry, Slovenia) have been used to interpret the structure and describe the mechanism of assembly of this nonameric pore. We have shown how the soluble monomer undergoes major conformational changes during the transition into one of the nine protomers building the compact transmembrane pore, via the membrane surface bound intermediate called a prepore (Figure 2). The atomic structure of the lysenin pore as well as the mechanism of its assembly now crucially help to understanding of the action of many related bacterial toxins and will serve as an important basis for development of new approaches towards bacterial infections, where such toxins are being used to infect people and animals.
Figure 2. (a) Ribbon representation of soluble monomeric lysenin (PDB-ID 3ZXG). C-terminal domain is in blue and N-terminal domain in yellow, pink and green. (b) A model, as predicted from AFM imaging and mutagenic analysis, of the intermediate prepore, composed of 9 molecules shown in (a), sitting on a membrane, colors the same as in (a). (c) Ribbon representation of the protomer present in the pore, depicting conformational changes in the N-terminal domain of the lysenin during the transformation from the soluble monomer (a) into the membrane inserted pore (d). The upper gray line depicts the surface of the membrane and the lower (d) the cytosol facing side of the membrane. (d) AFM image experimentally showing (left) the presence of the pore (1) and the prepore (2). Right: AFM image showing the assembly of the lysenin nonamers via arc shaped intermediates.
Moreover, we have shown that this pore translocates DNA molecules, using MinION platforms by Oxford Nanopore Technologies, UK. This company already uses similar pores for determination of DNA sequences in the most modern technologies of the size of the USB key. Due to extreme stability of the lysenin pore, they plan to include it into the future development of the nanosensors for the purpose of sequencing of the whole genomes as well as detection of other substances.
This research was conducted by the following research team:
Marjetka Podobnik1, Nejc Rojko1, Matic Kisovec1, Gregor Anderluh1, Peter Savory2, Neil Wood2, Richard Hambley2, Jonathan Pugh2, E. Jayne Wallace2, Luke McNeill2, Mark Bruce2, Lakmal Jayasinghe2, Idlir Liko3, Timothy M. Allison3, Shahid Mehmood3, Carol V. Robinson3, Neval Yilmaz4, Toshihide Kobayashi4 and Robert J.C. Gilbert5
1 Department for Molecular Biology and Nanobiotechnology, National Institute of Chemistry, Ljubljana, Slovenia.
2 Oxford Nanopore Technologies Ltd., Edmund Cartwright House, Oxford Science Park, Oxford, UK.
3 Department of Chemistry, University of Oxford, Oxford, UK.
4 Lipid Biology Laboratory, RIKEN Institute, Wako, Japan
5 Division of Structural Biology, Wellcome Trust Centre for Human Genetics, University of Oxford, Oxford, UK
Local contact person:
