Crystal and electron microscopy structures of Sticholysin II actinoporin reveal insights into the mechanism of membrane pore-formation

José M. Mancheño, Martín Martínez-Ripoll & Juan A. Hermoso

Departamento de Cristalografía y Biología Estructural, Instituto "Rocasolano", CSIC, Serrano 119, 28006-Madrid, Spain.

For more information, contact xjosemi@iqfr.csic.es or xjuan@iqfr.csic.es


 

 



Overview

Pore-forming proteins (PFPs) show the remarkable property of existing in at least two definite states. They are synthesised as stable water-soluble monomeric molecules that assemble into oligomeric pores to lyse target cells upon interaction with a membrane. The mechanism of action of PFPs involves protein-membrane interactions, protein-protein interactions, and a special protein folding pathway underlying the conformational transition from the water-soluble state to the membrane-bound form. PFPs are nowadays considered suitable experimental systems for elucidating the mechanisms of membrane binding, assembly and insertion. StnII is an actinoporin produced by the anemone Stichodactyla helianthus (order Actiniaria), currently classified as a transmembrane solute transporter from the pore-forming equinatoxin family 1.C.38. Actinoporins are PFPs displaying similar molecular masses (~20kDa) and high isoelectric points (pI´s > 9.0), related to each other in amino acid sequence (~66-85% identity), and presumably sharing a common mechanism of pore-formation

Water-soluble Stn2

The resulting structure is based on a beta-sandwich fold composed of ten beta-strands, flanked on each side by two short alpha-helices (Figure 1).The analysis of the Stn II structure revealed the striking presence of an exposed cluster of aromatic amino acids composed of Tyr106, Trp110, Tyr111, and Trp114 (coming from the loop comprised between strands b6 and b7), and Tyr131, Tyr135 and Tyr136 from helix a2. These residues are known to have affinity for the membrane interface.

Phosphocholine binding-site

Cocrystals of StnII with POC revealed for the first time the existence of a POC-binding site in a member of the actinoporins family (Figure 2). Phosphocholine binds to a cavity with overall dimensions of 9 x 11 x 13 Å. This cavity is partly hydrophilic due to the phenolic hydroxyl groups of Tyr131, Tyr135, and Tyr136 and the side chains of Ser52 and Ser103, and partly hydrophobic because it contains the side chains of Val85, Pro105 and the aromatic rings of Tyr111 and Tyr135. The positive charge of choline moiety is stabilized by cation-p interactions between the electron-rich systems of the aromatic rings of Tyr111 and Tyr135. Besides, the phosphate moiety interacts with the phenolic hydroxyl groups of Tyr111 and Tyr136 and presumably would be further stabilised by the cationic side chain of Arg51 (Figure 2). Interestingly, comparison of the structures of free StnII (backbone: orange; side chains:green) and that of the complex StnII:POC (backbone: dark blue; side chains: yellow) revealed some backbone modifications in the loop between strands b6 and b7 (rmsd 0.71 Å versus an overall value of 0.30 Å). However, the structural differences are mostly due to side chain rearrangements to facilitate POC-binding (Figure 3).

 

 

Stn2 on lipid monolayers

Two-dimensional crystals of StnII were grown on lipid monolayers (Figure 4A). The 2D crystals belong to the P4212 space group, with unit cell dimensions of 163 x 163 Å. The projection map obtained without symmetry imposition (Figure 4B) is essentially identical to that obtained after symmetry imposition (Figure 4C). Each cell contained eight molecules of StnII arranged in two tetrameric assemblies. A total of 52 images were selected, processed and combined in the P4212 space group to generate a 3D map of the structure of StnII in the lipid interface. The final 3D map was filtered at 18 Å resolution. As suggested by the 2D projection maps the 3D reconstruction of StnII in the lipid monolayer showed tetrameric assemblies that defined an inner pore (Figure 5).The tetrameric assembly showed distinct wide and narrow regions. The maximum outer diameter in the narrow region is 95 Å, and 110 Å in the wide one, and their pores have an inner diameter of 50 Å, with a height of 43 Å. The Mr of the tetramer calculated from the 3D map is ~76 kDa, assuming an average protein density of 0.75 Da Å-3 ( the Mr of the StnII monomer is 19.25 kDa), which strongly suggests that StnII does not significantly insert into the lipid monolayer.

 

 

Oligomerization at the lipid interface

A high-resolution model of the StnII tetramer was constructed by docking the atomic model of the soluble StnII into the 3D reconstruction (Figures 6A and 6B). This model predicts that the beta-sandwich core of the StnII structure would not suffer significant conformational changes upon oligomerization as deduced by its perfect fitting into the EM map. Moreover, the only regions clearly out of the density envelope, and thus presumably involved in conformational changes, are the N-terminal region (from Ala1 to Val27), the highly basic loop between strand b7 and helix a2 (121Ser-Gly-Lys-Arg-Arg-Ala-Asp-Gln128) and part of this helix (from Gly129 to Asp133). Furthermore, this model suggests that a simple pseudo-rigid body movement of the N-terminal region about the loop between helix a1 and strand b2, particularly a rotation between Ser28 and Arg29, would be sufficient to fill up the EM density in between monomers (Figure 6C). Consequently, the N-terminal region would be initially involved in oligomerization through interactions with the C-terminal region of the adyacent monomer. Thus, the loop between helix a1 and strand b2 would be essential for this conformational change, acting as a hinge-segment.

Pore-formation mechanism

Considering the above results, a model for the mechanism of actinoporin pore-formation can be envisaged. The mechanism would proceed through at least three steps: monomer binding to the membrane interface, assembly of four monomers, and final formation of the functional pore. Binding of soluble monomers to the membrane surface would be mainly driven by the affinity they exhibit for the phosphocholine group of lipids (POC binding-site) and also for the water/lipid interface (additional exposed aromatic residues), although other structural details of the lipids must be recognized for effective binding. After adsorption to the membrane surface, four protein molecules would associate forming an oligomeric assembly similar to that found in the 2D crystals. This step would involve changes in the conformation of the protein molecules mainly affecting both the N-terminal segment (Ala1 to Arg29) and a flexible and basic loop (Ser121 to Gln128). While the first segment would be initially involved in protein-protein contacts, the second would further anchor the protein to the membrane. Regarding to the final step of the mechanism, i.e. formation of the functional pore, it has been recently proposed that StnII originates toroidal pores in the membrane, similar to those formed by other toxic peptides, and particularly melittin which shows structural homology with the N-terminal end of actinoporins. According to this, both protein molecules and lipids would form the walls of the pore. Considering the high-resolution model of the tetrameric assembly, a tentative hypothesis for this last step can then be proposed which would involved the extension of the N-terminal region of each protomer which would adopt a helical conformation (Figure 7). In this regard, very recently it has been proposed that the N-terminal segment of EqtII (from Asp10 to Asn28) adopts an a-helical conformation in the functional pore that forms a tilting angle of around 21º with respect to the plane of the membrane, what agrees well with our proposed model (tilting angle ~30º). These helical segments together with lipid molecules would line the pore which has a diameter < 3 nm, close to the diameter of the functional pore (diameter ~2 nm).An artistic picture showing the main features of the molecular mechanism of StnII pore-formation is shown in Figure 8.

 

 

References

  1. Mancheño, J.M., Martínez-Ripoll, M., Gavilanes, J.G., Hermoso, J.A.
    Crystallization and preliminary X-ray diffraction studies of the pore-forming toxin Sticholysin II from the sea anemone Stichodactyla helianthus
    Acta Crystallogr.(2002) D58, 1229-1231

  2. Mancheño, J.M., Martín-Benito, J., Martínez-Ripoll, M., Gavilanes, J.G., Hermoso, J.A.
    Crystal and electron microscopy structures of Sticholysin II actinoporin reveal insights into the mechanism of membrane pore-formation
    Structure (2003) 11, 1319-1328