Atomic Force Microscopy Study Of Perfringolysin O Pore Forming Process

The correct assembly of protein complexes is essential for their function, especially formacromolecular protein complexes such as ribosomes or multi-subunit membrane channels.However, the precise assembly mechanisms of multimeric protein complexes are generally poorlyunderstood owing to the difficulty in reconstituting the process in-vitro. Many pore formingtoxins (PFTs) are well known to self-assemble from water soluble monomers intohomo-oligomers upon contact with a lipid bilayer, and are thus excellent model systems toprovide insight into this process.Cholesterol dependent cytolysins (CDCs) are a large family of PFTs that are secreted bymany Gram-positive bacteria, many of which are pathogenic. Perfringolysin O (PFO) secreted byClostridium perfringens is the prototypical member of the CDC family. Pore formation of PFOproceeds according to a three-step process common to many pore-forming proteins. Water-solublemonomers first bind to membranes via cholesterol as a receptor, then membrane-bound monomersself-assemble into non-perforating prepore complexes, and finally prepore complexes insert intothe bilayer and form the pore. It is also known that the PFO protein forms a wide range of porestoichiometries, including some containing over50subunits. A central question of this process ishow so many subunits can undergo the same transition from the prepore to the pore at the sametime. For example, with a50-member ring complex and2.5nm width per monomer, subunits onopposite sides of the ring, separated by24subunits or60nm, apparently undergo exactly the sametransition at exactly the same time. Understanding how this could possibly happen is the focus ofmy project.Recognizing that the size distribution of the pore complexes could provide information aboutthe self-assembling process, we investigated the size of the membrane bound complexes with highresolution Atomic Force Microscopy (AFM) in solution. From images in which thestoichiometries can be determined simply by directly counting the number of subunits in eachcomplex, we found unexpectedly that the complexes consist most frequently of integral multiplesof six subunits and the smallest size is hexamer. We thus suggest that assembly and/or insertionactually occurs in terms of predominantly hexameric sub-complexes/intermediates that differ instructure from smaller sub-complexes. With this, the difficult task of coordinating the behavior ofso many subunits in one complex is shown to be achieved by breaking the problem down tocoordinating the conformational changes first within a hexamer and then between differenthexamers. Such an hierarchical organization might prove generally applicable for othermulti-component toxins and more generally to large protein assemblies and self-assembledbiotechnological complexes.Finally, we note that we also performed high speed AFM in an attempt to directly observethe PFO pore forming process. While these experiments were not successful, the images obtainedunambiguously showed that complexes that are incomplete rings (or arcs) can indeed form a porein the membrane, which is a subject of recent considerable debate.