R468A mutation in perfringolysin O destabilizes toxin structure and induces membrane fusion

Abstract

Perfringolysin O (PFO) belongs to the family of cholesterol-dependent cytolysins. Upon binding to a cholesterol-containing membrane, PFO undergoes a series of structural changes that result in the formation of a β-barrel pore and cell lysis. Recognition and binding to cholesterol are mediated by the D4 domain, one of four domains of PFO. The D4 domain contains a conserved tryptophan-rich loop named undecapeptide (ECTGLAWEWWR) in which arginine 468 is essential for retaining allosteric coupling between D4 and other domains during interaction of PFO with the membrane. In this report we studied the impact of R468A mutation on the whole protein structure using hydrogen-deuterium exchange coupled with mass spectrometry. We found that in aqueous solution, compared to wild type (PFO), PFO showed increased deuterium uptake due to exposure of internal toxin regions to the solvent. This change reflected an overall structural destabilization of PFO in solution. Conversely, upon binding to cholesterol-containing membranes, PFO revealed a profound decrease of hydrogen-deuterium exchange when compared to PFO. This block of deuterium uptake resulted from PFO-induced aggregation and fusion of liposomes, as found by dynamic light scattering, microscopic observations and FRET measurements. In the result of liposome aggregation and fusion, the entire PFO molecule became shielded from aqueous solution and thereby was protected against proteolytic digestion and deuteration. We have established that structural changes induced by the R468A mutation lead to exposure of an additional cholesterol-independent liposome-binding site in PFO that confers its fusogenic property, altering the mode of the toxin action.

Interaction of Cholesterol with Perfringolysin O: What Have We Learned from Functional Analysis?

“…The extended binding pose adopted by cholesterol explains the need for the equatorially projected hydroxyl, which is capable of making multiple H-bonding contacts, unlike a largely occluded axial hydroxyl in epicholesterol [ 33 ]. Two other notable events occur during the IFD search to permit the association: (i) the terminal residue of the undecapeptide (R437 in PLY and R468 in PFO), a residue critical for the stability of the toxin structure [ 47 ], becomes solvent exposed and opens up the space between two β-sheets to become available for association with cholesterol; and (ii) the first residue of the undecapeptide, (E427 in PLY and E458 in PFO), having lost its salt-bridge partner while remaining in a largely hydrophobic environment is expected to become protonated with a concomitant pKa shift from ~4.5 in PFO or β-hairpin PLY structures to 6.5 and 7.5 in cholesterol-free and bound 4ZGH-derived models, respectively (Epik, Schrödinger, LCC, New York, NY, USA) [ 48 ], making the binding surface even more hydrophobic and therefore, more welcoming to a ligand as lipophilic as cholesterol. This observation is consistent with the report of low pH (5.5–6) enhancing PFO–membrane association [ 9 ].…”

Insight into the Structural Dynamics of the Lysenin During Prepore-to-Pore Transition Using Hydrogen–Deuterium Exchange Mass Spectrometry

“…Reduction of the disulfide bond leads to disengagement and unfolding of the insertion loop, which completely restores pore-forming activity to the level observed for lysenin WT . As previously shown for another pore-forming protein, perfringolysin O, the introduction of single point mutations induces local changes in the protein structure and may also cause global changes in the protein structure without influencing the secondary structure [25,32]. It has also been shown that the introduction of a point mutation outside of the area involved in membrane binding can alter the interaction between the receptor-binding motif (domain) and the lipid bilayer, indicating both structural and functional coupling of distant and spatially separated domains of PFO [33].…”

“…Samples for hydrogen–deuterium exchange were prepared following a previously described procedure, with some modifications [24,25]. Briefly, lipid-bound samples were prepared by incubation of 4 µM lysenin WT and lysenin V88C/Y131C (with or without 10 mM DTT) with 2 mM SUVs composed of SM/DOPC/cholesterol (molar ratio 1:2:1) at room temperature for 30 min.…”

“…Finally, the samples were injected onto an immobilized pepsin column (Poroszyme; ABI), and the obtained peptides were further separated using the nanoACQUITY ultra-performance liquid chromatography (UPLC) system, followed by mass measurements on the SYNAPT G2 HDMS mass spectrometer (Waters, Milford, MA, USA). Peptide identification was based on a list of peptic peptides obtained for a non-deuterated sample using ProteinLynx Global Server software (Waters, Milford, MA, USA), as previously described [25]. HDX data analysis was performed using the DynamX 2.0 program (Waters, Milford, MA, USA).…”

“…Generally, due to the fact that protection against hydrogen–deuterium exchange may be best correlated with the “order” in a given region of protein, we interpret the results in terms of structural stability in a given region. Therefore, this method provides useful information regarding the dynamics of structural elements of pore-forming proteins, both in solution and upon interaction with a lipid membrane and enables investigation of the conformational changes accompanying transitions between protein states with high reproducibility and precision [24,25,26].…”