The work described in this thesis deals with two pore forming toxins, namely Arcanobacterium pyogenes pyolysin and Streptococcus agalactiae CAMP factor. Pyolysin (PLO) belongs to the homologous group of cholesterol-dependent cytolysins. In chapter 2, it is shown that PLO can form small oligomers in solution, without the requirement for any membranes or membrane lipids. These small oligomers may aggregate into larger ones on membranes; however, in solution, they apparently do not grow by addition of further monomers, as their size is virtually unaffected by variations of incubation time or toxin concentration. The small, solution-derived oligomers retain hemolytic activity. The membrane lesions observed by electron microscopy are similar to those that are formed by monomeric PLO, except that they are mostly incomplete and arc-shaped, as opposed to the predominantly ring-shaped ones formed by monomeric PLO when directly incubated with membranes. This structural difference corresponds to a detectable difference in functional pore size, as determined by marker release experiments. Thus, arc-shaped PLO oligomers may form functional pores of reduced size. In chapter 3, we show that liposomes that contain phosphatidylcholine and ceramide but no cholesterol or other sterol are susceptible to the cholesterol-dependent cytolysin pyolysin. Pyolysin, at a low rate, forms small oligomers in solution. The solution-derived oligomers are more effective on ceramide-containing liposomes than the monomeric toxin. In contrast, they have much lower activity on liposome membranes that contain cholesterol but no ceramide. Our findings therefore show that at least one member of the ‘cholesterol-dependent cytolysins’ is in fact not strictly dependent on the sterol. In addition, in conjunction with previous data, they suggest that the requirement for cholesterol involves early or intermediate stages of oligomer formation, rather than the final event of membrane insertion. Chapter 4 of this thesis concerns Streptococcus agalactiae CAMP factor. CAMP factor has previously been reported to bind the Fc fragments of immunoglobulin G (IgG) and has therefore also been called ‘protein B’, in analogy to protein A of Staphylococcus aureus. We attempted to characterize the interaction of protein B with IgG in more detail. In contrast to protein A, CAMP factor does not inhibit the activation of complement by hemolysin antibodies bound to sheep red cell surfaces. IgG also failed to inhibit the cohemolytic activity of CAMP factor, which it displays on sphingomyelinase-treated sheep red cells; this is in disagreement with previous findings. After co-incubation, CAMP factor and IgG were cleanly separated by gel filtration. Therefore, CAMP factor does not detectably bind to IgG.

The work described in this thesis deals with two pore forming toxins, namely Arcanobacterium pyogenes pyolysin and Streptococcus agalactiae CAMP factor. Pyolysin (PLO) belongs to the homologous group of cholesterol-dependent cytolysins. In chapter 2, it is shown that PLO can form small oligomers in solution, without the requirement for any membranes or membrane lipids. These small oligomers may aggregate into larger ones on membranes; however, in solution, they apparently do not grow by addition of further monomers, as their size is virtually unaffected by variations of incubation time or toxin concentration. The small, solution-derived oligomers retain hemolytic activity. The membrane lesions observed by electron microscopy are similar to those that are formed by monomeric PLO, except that they are mostly incomplete and arc-shaped, as opposed to the predominantly ring-shaped ones formed by monomeric PLO when directly incubated with membranes. This structural difference corresponds to a detectable difference in functional pore size, as determined by marker release experiments. Thus, arc-shaped PLO oligomers may form functional pores of reduced size. In chapter 3, we show that liposomes that contain phosphatidylcholine and ceramide but no cholesterol or other sterol are susceptible to the cholesterol-dependent cytolysin pyolysin. Pyolysin, at a low rate, forms small oligomers in solution. The solution-derived oligomers are more effective on ceramide-containing liposomes than the monomeric toxin. In contrast, they have much lower activity on liposome membranes that contain cholesterol but no ceramide. Our findings therefore show that at least one member of the ‘cholesterol-dependent cytolysins’ is in fact not strictly dependent on the sterol. In addition, in conjunction with previous data, they suggest that the requirement for cholesterol involves early or intermediate stages of oligomer formation, rather than the final event of membrane insertion. Chapter 4 of this thesis concerns Streptococcus agalactiae CAMP factor. CAMP factor has previously been reported to bind the Fc fragments of immunoglobulin G (IgG) and has therefore also been called ‘protein B’, in analogy to protein A of Staphylococcus aureus. We attempted to characterize the interaction of protein B with IgG in more detail. In contrast to protein A, CAMP factor does not inhibit the activation of complement by hemolysin antibodies bound to sheep red cell surfaces. IgG also failed to inhibit the cohemolytic activity of CAMP factor, which it displays on sphingomyelinase-treated sheep red cells; this is in disagreement with previous findings. After co-incubation, CAMP factor and IgG were cleanly separated by gel filtration. Therefore, CAMP factor does not detectably bind to IgG.

The bacterial toxin pyolysin (PLO) is a member of the family of Cholesterol-Dependent Cytolysins, which form large, oligomeric pores in cholesterol-containing membranes. The general CDC structure has an elongated shape and consists of four domains rich in β-sheet structure. Upon binding to a membrane, molecules diffuse laterally on the surface and oligomerize to form a pre-pore complex, then insert into the membrane yielding pores of unusually large size, approximately 30 nm in diameter. In this work, the oligomerization properties of PLO were investigated. In particular, the role of the C-terminal domain in the oligomerization process, the effects of a disulphide-tethered mutant on the activity of the wild type toxin, and the pore-forming ability of oligomers pre-formed in solution were characterized. Chapter 2 characterizes the functional properties of a recombinant fragment that corresponds to the C-terminal domain 4 of PLO. It is shown that this fragment can form hybrid oligomers with intact PLO toxin molecules, and is also capable of self-oligomerization. The fragment has no haemolytic activity of its own; nevertheless, it can to some degree increase the haemolytic activity of the wild type toxin. In addition, in a mixture domain 4 and wild type interact in such a way as to form unusual shapes on cholesterol crystals that have not been previously observed. Chapter 3 describes the effects of a disulphide bond linking domain 2 to a membrane-inserting region of domain 3 on the oligomerization process. The disulphide mutant was not able to oligomerize on its own, and when combined with active PLO toxin, the haemolytic activity of wild type was significantly inhibited. Also, the combination of the disulphide-tethered mutant with intact toxin resulted in the formation of hybrid oligomers. This, in turn, caused an increase in incomplete ring formations on cholesterol surfaces which correlate to a reduction in functional pore size, suggesting that insertion of subunits is partially cooperative. The results of the investigation of the pore-forming ability of solution-derived oligomers (SDO) are described in Chapter 4. Here, the fluorescence emission of an environmentally-sensitive probe on the SDO after membrane insertion was a fraction of that observed with the monomeric control, which was supported by hydrophobic quenching analyses. This suggests that the formation of SDO may block necessary conformational changes in the intact toxin to allow membrane insertion.

The bacterial toxin pyolysin (PLO) is a member of the family of Cholesterol-Dependent Cytolysins, which form large, oligomeric pores in cholesterol-containing membranes. The general CDC structure has an elongated shape and consists of four domains rich in β-sheet structure. Upon binding to a membrane, molecules diffuse laterally on the surface and oligomerize to form a pre-pore complex, then insert into the membrane yielding pores of unusually large size, approximately 30 nm in diameter. In this work, the oligomerization properties of PLO were investigated. In particular, the role of the C-terminal domain in the oligomerization process, the effects of a disulphide-tethered mutant on the activity of the wild type toxin, and the pore-forming ability of oligomers pre-formed in solution were characterized. Chapter 2 characterizes the functional properties of a recombinant fragment that corresponds to the C-terminal domain 4 of PLO. It is shown that this fragment can form hybrid oligomers with intact PLO toxin molecules, and is also capable of self-oligomerization. The fragment has no haemolytic activity of its own; nevertheless, it can to some degree increase the haemolytic activity of the wild type toxin. In addition, in a mixture domain 4 and wild type interact in such a way as to form unusual shapes on cholesterol crystals that have not been previously observed. Chapter 3 describes the effects of a disulphide bond linking domain 2 to a membrane-inserting region of domain 3 on the oligomerization process. The disulphide mutant was not able to oligomerize on its own, and when combined with active PLO toxin, the haemolytic activity of wild type was significantly inhibited. Also, the combination of the disulphide-tethered mutant with intact toxin resulted in the formation of hybrid oligomers. This, in turn, caused an increase in incomplete ring formations on cholesterol surfaces which correlate to a reduction in functional pore size, suggesting that insertion of subunits is partially cooperative. The results of the investigation of the pore-forming ability of solution-derived oligomers (SDO) are described in Chapter 4. Here, the fluorescence emission of an environmentally-sensitive probe on the SDO after membrane insertion was a fraction of that observed with the monomeric control, which was supported by hydrophobic quenching analyses. This suggests that the formation of SDO may block necessary conformational changes in the intact toxin to allow membrane insertion.