Arresting and releasing Staphylococcal alpha-hemolysin at intermediate stages of pore formation by engineered disulfide bonds

Abstract

alpha-Hemolysin (alphaHL) is secreted by Staphylococcus aureus as a water-soluble monomer that assembles into a heptamer to form a transmembrane pore on a target membrane. The crystal structures of the LukF water-soluble monomer and the membrane-bound alpha-hemolysin heptamer show that large conformational changes occur during assembly. However, the mechanism of assembly and pore formation is still unclear, primarily because of the difficulty in obtaining structural information on assembly intermediates. Our goal is to use disulfide bonds to selectively arrest and release alphaHL from intermediate stages of the assembly process and to use these mutants to test mechanistic hypotheses. To accomplish this, we created four double cysteine mutants, D108C/K154C (alphaHL-A), M113C/K147C (alphaHL-B), H48C/ N121C (alphaHL-C), I5C/G130C (alphaHL-D), in which disulfide bonds may form between the pre-stem domain and the beta-sandwich domain to prevent pre-stem rearrangement and membrane insertion. Among the four mutants, alphaHL-A is remarkably stable, is produced at a level at least 10-fold greater than that of the wild-type protein, is monomeric in aqueous solution, and has hemolytic activity that can be regulated by the presence or absence of reducing agents. Cross-linking analysis showed that alphaHL-A assembles on a membrane into an oligomer, which is likely to be a heptamer, in the absence of a reducing agent, suggesting that oxidized alphaHL-A is halted at a heptameric prepore state. Therefore, conformational rearrangements at positions 108 and 154 are critical for the completion of alphaHL assembly but are not essential for membrane binding or for formation of an oligomeric prepore intermediate.

Figure 1.

Structures of a homology model of the water-soluble αHL monomer and a protomer from the assembled heptamer. (A) Location of the double cysteine mutation sites in αHL. The model of water-soluble αHL monomer was created using Swiss-Model program (Peitsch and Tschopp 1995; Peitsch 1996; Guex and Peitsch 1997) based on the structure of water-soluble LukF monomer. Because the amino terminus of αHL is not homologous to that of LukF, the first 16 residues are not shown. Three putative disulfide bridges in the mutants αHL-A (D108C/K154C), αHL-B (M113C/K147C), and αHL-C (H48C/N121C) are depicted for illustration purposes only. For mutant αHL-D (I5C/G130C), the locations of Gly130 and the amino terminus of this model (Asn17) are illustrated. In parts (A) and (B) of this figure, the pre-stem is in green, the β-sandwich is in blue, the rim is in red, and the triangle region is in yellow. (B) A schematic presentation of the αHL conformational change during pore formation. The water-soluble monomer was modeled as indicated in (A). A protomer of the heptameric pore is depicted on the right-hand side of the panel. Here, the amino-latch is in pink.

Figure 2.

SDS-PAGE of purified mutant proteins in the presence and absence of a reducing agent. Recombinant proteins were purified by Ni-NTA chromatography, and 4 μg of each protein was analyzed +/− β-mercaptoethanol. The lanes are composed of wild-type, untagged αHL (WT), wild-type His-tagged αHL, and mutants αHL-A (A), αHL-B (B), αHL-C (C), and αHL-D (D).

Figure 3.

Hemolytic activity of mutant proteins, defined as in Figure 2 ▶, under reducing and nonreducing conditions. RBC in hemolysis (HL) buffer were incubated with wild-type and mutant proteins at 20°C for 20 min. The hemolytic activity was monitored by following the increase of absorbance at 475 nm. The reactions were normalized to the extent of lysis caused by diluting an equivalent volume of erythrocytes into 8 volumes of Milli Q water. The reported value is the mean and the error bars define standard deviations (n = 3). The symbol “b” indicates buffer, the symbol “WT” indicates His-tagged wild-type protein, and the symbol “+/−” indicates +/− DTT.

Figure 4.

Lysis and binding experiments using wild-type toxin and αHL-A. (A) Time course hemolytic activity of αHL-A. RBC in hemolysis (HL) buffer were incubated with wild-type and αHL-A protein either in the absence (left) or presence (right) of DTT at 20°C for 0, 2, 4, 6, 8, and 10 min. The hemolytic activity was monitored by following the increase of absorbance at 475 nm. The reactions were normalized to the extent of lysis caused by diluting an equivalent volume of erythrocytes into 8 volumes of Milli Q water. The reported value is the mean, and the error bars define one standard deviation (n = 3). (B) αHL-A binding to RBC. The recombinant proteins were incubated with RBC in the presence or absence of DTT on ice for 1 h. The supernatant and pellet fractions were separated by centrifugation, and the amount of recombinant protein was detected by Western blotting using anti-His₆ antibody. The symbol “S” indicates supernatant and the symbol “P” indicates pellet.

Figure 5.

SDS-PAGE of trypsin-digested αHL-A protein under reducing and nonreducing conditions. (A) The two most sensitive trypsin digestion sites (K8 and K131) in αHL-A are shown. (B) The digestion mixtures were analyzed in the absence of β-mercaptoethanol (lanes 2–6) or in the presence of β-mercaptoethanol (lanes 7–11). Lanes 2–6 and 7–11 are digestion mixtures with trypsin/protein ratios of 0, 1/5000, 1/500, 1/250, 1/50 (w/w), respectively.

Figure 6.

Wild-type and αHL-A toxin assembly in the absence or presence of β-mercaptoethanol. The recombinant proteins (0.5 mg/mL) were incubated with 8 mM DiC₈PC at room temperature for 1 h. Oligomerization was detected by SDS-PAGE +/− 0.7 M β-mercaptoethanol. The positions of molecular weight standards are shown on the left, and the numbers are in units of kD. The symbol “WT” indicates His-tagged wild-type protein.

Figure 7.

SDS-PAGE of crosslinked αHL-A mutant oligomer under reducing and nonreducing conditions. αHL-A protein (1.3 mg/mL) was incubated with 8 mM DiC₈PC (lanes 8–17) or without DiC₈PC (lanes 3–7) at room temperature for 1 h with DTT (B) or without DTT (A). The resulting reaction mixtures were incubated with glutaraldehyde (GA) at room temperature either for 1 h (lanes 8–12) or 24 h (lanes 3–7, 13–17). Five different GA concentrations (0, 0.25, 0.5, 2.5, and 25 μM) were used for each set of lanes 3–7, 8–12, and 13–17, respectively. The triangle depicts the increase in GA concentrations.

Figure 8.

A schematic showing the arrest and the release of αHL-A. Water-soluble αHL-A monomer (α₁) binds to a target membrane through the rim-domain (α₁*) followed by prepore formation (α₇*). αHL-A is arrested from forming a pore at this stage by the disulfide bond formed between the β-sandwich domain and the rim domain under oxidizing conditions. Under reducing conditions, however, the disulfide bridge between residues 108 and 154 is reduced, the pre-stem can rearrange, and αHL-A forms a heptameric pore (α₇). In this schematic, only four protomers are depicted in the heptameric prepore (α₇*) and the heptameric pore (α₇) states.