Molecular determinants of self-assembly of the pore forming toxin Cytolysin A

Abstract

Pore forming toxins (PFTs) belong to a class of bacterial exotoxins that disrupt the biological membrane barrier by formation of nanopores. These toxins are central to the virulence of pathogens such as S. pneumoniae, V. cholerae, L. monocytogenes and B. anthracis to name a few. Formation of pores lead to the killing of cells due to cellular ion imbalance and efflux of important biomolecules such as ATP, from the cell. Unique to this class of proteins is their ability to exist in bi-stable states i.e., they are produced in soluble forms and upon exposure to membrane, undergo conformational changes, oligomerization and form a stable membrane integrated structure. This spontaneous transformation occurs without the aid of any extrinsic processes such as post-translation modifi cation, tethering to membrane anchors and translocon mediated membrane insertion. Self-assembly and bi-stability being intrinsic properties of PFTs, are embedded in sequence and structure. However, barring a few PFTs, molecular details of regions in the protein that are essential for the aforementioned processes have not been well characterized. These questions are explored in an -PFT, Cytolysin A (ClyA) from E. coli. ClyA is produced as a monomer and upon membrane exposure, undergoes conformational changes and nally oligomerizes to form a dodecameric pore. It is one of the few PFTs for which X-ray crystal structures are available, for both the soluble and membrane inserted pore form. This information is valuable to correlate structure, sequence, PFT self-assembly pathway and lytic action. In the rst part of the work, the role of solvent-exposed C-terminal loop, comprising of 11 amino acids, is examined. This region is devoid of any intra-protein contacts in the monomer and neither is it speculated to have any inter-protein contacts in the pore structure, however its deletion renders the protein inactive. Absence of any interactions in the initial and final states implies that it is essential in the intermediary steps of pore assembly. To identify the mechanistic details of modulation of activity, the different steps of pore formation were monitored in a mutant devoid of the C-terminal loop. Studies on binding and detergent-induced oligomerization did not indicate any anomaly in comparison to the wild type protein. Nevertheless mutant protein was found to be inactive. Thermal unfolding experiments revealed that deletion of the C-terminal loop destabilized the structure of the monomer. Furthermore, by using novel methods to reconstitute detergent-induced oligomers into vesicles, and examining oligomer stability, it became clear that the instability in the monomer was propagated to the oligomers as well. Hemolytic activity of the wild type protein reduced signi ficantly when co-incubated with the C-terminal deletion mutant, possibly due to sequestration of wild type units into inactive hetero-oligomers with the mutant protein. Thus, instability in the monomer due to deletion of the C-terminal loop, altered the delity of the conformational transition upon membrane exposure. This resulted in formation of aberrant oligomers on the membrane surface. Therefore, regions of the PFTs, not directly involved in the pore structure, appear to assist in transitioning through intermediary steps of assembly, leading to successful pore formation in a membrane environment. PFT assembly on membrane is dependent on lateral motion of the toxin monomers, as they directly impinge on collision frequency and hence oligomerization. However, all studies to examine ClyA assembly have been carried out in solution, with detergent, wherein micellar aggregation is the dominant driving force rather than toxin self-assembly. Toxin self-assembly on the other hand is contingent on establishing inter-protomer contacts and is influenced by lateral motion of the individual protein molecules on the surface. However, little is known about PFT mobility on the membrane surface and its implication in the formation of higher-order structures. In the third Chapter, the lateral motion of ClyA on membrane was analyzed by using single-molecule fluorescence and spectroscopy. Single-particle tracking of ClyA was carried out on polymer-cushioned, arti ficial membrane systems. Diffusional analysis of particle trajectories revealed the existence of two discrete mobility states exhibiting fast and slow dynamics. In the presence of membrane cholesterol, the population of the slower moving species increased, with the concomitant decrease in the fast mobility population. Analysis of transition probabilities between the mobility states revealed that the conversion from fast to slower mobility state was due to the conformational transition from a peripherally-associated protein conformation to a membrane-inserted conformation respectively. Furthermore, a cholesterol consensus and recognition (CRAC) motif was discovered in the transmembrane helix of ClyA. Upon disruption of this motif, the fraction of fast moving species increased signi - ficantly and activity of the mutant protein was signi ficantly compromised. This validated that the conformational transition to form the membrane bound structure was stabilized by the interaction with cholesterol. The presence of membrane cholesterol triggered the insertion of the transmembrane helices leading to the formation of a functional pore, thus providing a strategy for selective pore formation, in eukaryotic membranes. Additionally, it serves as evidence that heterogeneous motion on the membrane need not necessarily arise from the presence of `lipid-microdomains', and conformational plasticity is also an important determinant of protein lateral motion. In summary, this study has provided novel molecular insights into hitherto unknown regulatory mechanisms of ClyA pore formation such as the C-terminal loop and CRAC motif. Albeit seemingly disparate, the solvent exposed and transmembrane domains appear to regulate different aspects of the assembly pathway for successful pore formation. The results presented in the thesis allow reconciliation of biologically relevant mechanism of virulence factor dissemination and activation on target cells. In addition to enhancing the understanding of mechanisms in uencing toxin function, results presented here can enable design of novel `anti-toxin' therapies which can used in case of pathogenesis.