Disentangling the roles of cholesterol and CD59 in intermedilysin pore formation

Abstract

The plasma membrane provides an essential barrier, shielding a cell from the pressures of its external environment. Pore-forming proteins, deployed by both hosts and pathogens alike, breach this barrier to lyse target cells. Intermedilysin is a cholesterol-dependent cytolysin that requires the human immune receptor CD59, in addition to cholesterol, to form giant β-barrel pores in host membranes. Here we integrate biochemical assays with electron microscopy and atomic force microscopy to distinguish the roles of these two receptors in mediating structural transitions of pore formation. CD59 is required for the specific coordination of intermedilysin (ILY) monomers and for triggering collapse of an oligomeric prepore. Movement of Domain 2 with respect to Domain 3 of ILY is essential for forming a late prepore intermediate that releases CD59, while the role of cholesterol may be limited to insertion of the transmembrane segments. Together these data define a structural timeline for ILY pore formation and suggest a mechanism that is relevant to understanding other pore-forming toxins that also require CD59.

Structural basis for tuning activity and membrane specificity of bacterial cytolysins

“…The position of the tryptophan likely affects the efficiency of this mechanism, since substitution at N342 shows a stronger phenotypic change than at N338. Since ILY only requires cholesterol to rupture the membrane and not to bind it 10 , the ILY-CD59 system offers an opportunity to explicitly investigate and potentially modulate the cholesteroldependency of CDC membrane lysis. We therefore measured the lytic activity of our ILY variants on liposomes without cholesterol (Fig.…”

“…ILY from Streptococcus intermedius targets human cells by binding the GPI-anchored complement regulator CD59 9 through an extended β-hairpin of D4 17 . To understand how this interaction initiates oligomerization on a target membrane, we trapped ILY in an early prepore state using a disulfide lock that restricts movement between D2 and the MACPF/CDC domain 10 . This disulfide-locked ILY variant binds CD59 and forms SDS-sensitive loosely associated oligomers 10 , analogous to previously characterized CDC early prepore states 18 .…”

“…Other CDCs, such as intermedilysin (ILY) secreted by Streptococcus intermedius, achieve species-specificity for their hosts by hijacking the cell surface receptor CD59 to initiate membrane binding 9 . For this subgroup of CDCs, the interaction with CD59 is sufficient for attachment and the role of cholesterol is restricted to pore formation 10 . While differences in membrane targeting may reflect diversity in CDC virulence strategies, lipid dependency for membrane penetration remains highly conserved.…”

“…In addition, the supported lipid bilayers generally represent too fluid a support to retain the assemblies in place 3 , except when the assemblies form larger clusters. In this case, the mobility of HB1/HB2-lock prepore assemblies may also be constrained by the dense coverage of CD59 on the membrane surface 10 .…”

Mannose receptor‐derived peptides neutralize pore‐forming toxins and reduce inflammation and development of pneumococcal disease

“…CDCs promote bacterial virulence in many ways such as (i) induction of epithelial barrier dysfunction (Witzenrath et al , 2006), (ii) lysis of phagocytic immune cells (Domon et al , 2016), and (iii) facilitating bacterial invasion of host cells and intracellular survival (Subramanian et al , 2019b) (Birmingham et al , 2008). Prominent examples of bacterial CDCs include pneumolysin (PLY) from Streptococcus pneumoniae , streptolysin O (SLO) from Streptococcus pyogenes , listeriolysin O (LLO) from Listeria monocytogenes , intermedilysin (ILY) from Streptococcus intermedius , anthrolysin O (ALO) from Bacillus anthracis, and perfringolysin O (PFO) from Clostridium perfringens (Gilbert et al , 1999; Park et al , 2004; Boyd et al , 2016; Savinov & Heuck, 2017; Nguyen et al , 2019; Ogasawara et al , 2019; Subramanian et al , 2019a,b; Vogele et al , 2019).…”