Functional and Structural Variation among Sticholysins, Pore-Forming Proteins from the Sea Anemone Stichodactyla helianthus

Abstract

Venoms constitute complex mixtures of many different molecules arising from evolution in processes driven by continuous prey-predator interactions. One of the most common compounds in these venomous cocktails are pore-forming proteins, a family of toxins whose activity relies on the disruption of the plasmatic membranes by forming pores. The venom of sea anemones, belonging to the oldest lineage of venomous animals, contains a large amount of a characteristic group of pore-forming proteins known as actinoporins. They bind specifically to sphingomyelin-containing membranes and suffer a conformational metamorphosis that drives them to make pores. This event usually leads cells to death by osmotic shock. Sticholysins are the actinoporins produced by Stichodactyla helianthus. Three different isotoxins are known: Sticholysins I, II, and III. They share very similar amino acid sequence and three-dimensional structure but display different behavior in terms of lytic activity and ability to interact with cholesterol, an important lipid component of vertebrate membranes. In addition, sticholysins can act in synergy when exerting their toxin action. The subtle, but important, molecular nuances that explain their different behavior are described and discussed throughout the text. Improving our knowledge about sticholysins behavior is important for eventually developing them into biotechnological tools.

Keywords: actinoporins; cholesterol; cnidaria; leakage; sphingomyelin; venom.

Figure 1

Phylogenetic tree of cnidaria phylum. Two different groups can be distinguished in cnidaria: Anthozoa, containing Hexacorallia and Octocorallia subclasses, and Medusozoa, including Staurozoa, Cubozoa, Scyphozoa, and Hydrozoa. Figure modified from [103,105].

Figure 2

Overlapping structures of the solved actinoporin three-dimensional structures. Sticholysins I (StnI) (2KS4, red) and II (StnII) (1GWY, blue) from Stichodactyla helianthus, fragaceatoxin C (FraC) (3W9P, green) from Actinia fragacea, and equinatoxin II (EqtII) (1IAZ, yellow) from Actinia equina. They share a common characteristic fold, a β-sandwich core flanked by two α-helices. The RMSD values are shown in the table below. Values are much higher when comparison is made against StnI because it is a structure obtained by NMR [127,129], while the other three are crystalline structures [123,125,128].

Figure 3

Cartoon schematic representation showing most of the steps generally accepted for the pore formation mechanism of actinoporins. In solution, they remain soluble and stably folded (a) [26,110,124,127,136,137,138,139]. Upon interaction with a lipid membrane containing SM (b), their N-terminal α-helix stretch is detached and extended, shortly laying parallel to the membrane (c) [25,26,82,85,87,91,116,140,141,142,143]. Then, monomers oligomerize and insert this N-terminal α-helix, now about 30 residues long, within the hydrophobic membrane core (d) [34,144,145,146,147,148,149]. The existence of such an intermediate is one of the most controversial issues, depending on the degree of acceptance of the evidence about the real existence, or not, of pre-pore assemblies [132,133,150]. Finally, a cation selective channel is established (e) [39,116,128,136,144,151].

Figure 4

Two of the main models representing the pore structure of actinoporins. (A) A tetrameric one in which the membrane adopts a toroidal shape around the pore walls (made using StnII model 1GWY fitted to the structure of FraC from model 4TSY), and (B) an octameric lipid–protein structure (FraC pore, PDB ID: 4TSY) in which lipids (carbon atoms in tan color) are accommodated through pore wall fenestrations. Inserts on the right show a close-up of a lipid on a fenestration (top) and a top view of the octameric complex (bottom). Adapted from [32]. The α-helices are depicted in red and gold, β-sheets in blue, nonperiodic structures in grey. Bulk lipids are depicted in grey (carbon atoms) and red (oxygen atoms).

Figure 5

Four different functionally relevant regions can be distinguished in the water-soluble monomeric actinoporins, as depicted on this representation of the three-dimensional structure of StnII (PDB: 1GWY). The N-terminal stretch (in orange), an array of exposed and basic amino acids (in red and blue), a cluster of aromatic residues (in light blue, except the Tyr residues also belonging to the POC site, which appear in green), and the POC binding site (in yellow or green). Residues taking part in both the cluster of aromatic residues and the POC binding site are highlighted in green. The black arrow on the left panel indicates the point of view of the right panel, which is a close-up of the cluster of aromatic residues and the POC binding site.

Figure 6

(A) Three-dimensional structures of StnI (2KS4) and II (1GWY). (B) Sequence alignment of StnI, II, and III. Identical residues are highlighted in black, those with similar/conserved chemical properties appear grey, and nonconserved residues remain backgrounded in white. Black asterisk (*) indicates StnI Asp9 (Ala8 and Gln11 in StnII and StnIII, respectively). Black hashtag (#) indicates a conserved Lys residue in StnI, StnII, and StnIII, occupying positions 68, 67, and 70, respectively. As indicated in Figure 5, red circles highlight residues that conform the array of basic amino acids. Light blue arrowheads pointing down indicate the residues taking part of the cluster of aromatic residues, yellow arrowheads pointing up mark the amino acids that are part of the POC binding site, and green diamonds indicate residues belonging to both the cluster of aromatic residues and the POC binding site. These residues appear conserved in all three Stns except for those corresponding to StnI Trp111, Lys119, and Arg176 which, only in StnIII, are Leu113, Arg121, and His178. This figure has been modified from [162].

Figure 7

Synergic effect on hemolysis. Maximum hemolytic rate values (expressed as percentage of hemolysis per second) are represented as function of protein concentration (in log scale). (A) StnII (black dots–solid line), StnIII (red dots–solid line), a StnII:StnIII (20:80) mixture (black triangles–solid line). (B) Same as in (A), but now the proteins employed were StnI (orange dots–solid line), StnIII (red dots–solid line), and StnI:StnIII (80:20) mixtures (black triangles–solid line). In both panels, the black triangles–dashed lines were obtained as the arithmetic addition of the rates obtained with the individual proteins for the real concentration of each one in the different mixtures employed. Values are average of n = 3 ± SEM. Hemolysis assays were performed in 96-multiwell plates at 25 °C and exactly as described previously [118]. Briefly, erythrocytes from heparinized sheep blood were washed in 10 mM Tris buffer, pH 7.4, containing 145 mM NaCl, to a final A₆₅₅ of 0.5 when mixing equal volumes of the cell suspension and buffer. The hemolysis was followed as a decrease in A₆₅₅ after addition of the erythrocyte suspension to different final concentrations of protein. An Expert 96 microplate reader (Asys Hitech, GmbH, Eugendorf, Austria) was employed to measure the absorbance. The value obtained with 0.1% (w/v) Na₂CO₃ was considered as 100% hemolysis.

Figure 8

(A) Overlapping structures of StnI (salmon; PDB ID 2KS4) and StnII (cyan; PDB ID 1GWY). The arrow points at residues D76 of StnII and S77 of StnI. (B) Overlapping structures of StnII (cyan) and OlyA (orange; PDB ID 6MYI). The arrow points at residues D76 of StnII and E69 of OlyA. (C) Close up to the compared position, showing the sidechains of the mentioned residues as sticks. Colors as in (A,B).