Structurally Robust and Functionally Highly Versatile-C-Type Lectin (-Related) Proteins in Snake Venoms

Abstract

Snake venoms contain an astounding variety of different proteins. Among them are numerous C-type lectin family members, which are grouped into classical Ca²⁺- and sugar-binding lectins and the non-sugar-binding snake venom C-type lectin-related proteins (SV-CLRPs), also called snaclecs. Both groups share the robust C-type lectin domain (CTLD) fold but differ in a long loop, which either contributes to a sugar-binding site or is expanded into a loop-swapping heterodimerization domain between two CLRP subunits. Most C-type lectin (-related) proteins assemble in ordered supramolecular complexes with a high versatility of subunit numbers and geometric arrays. Similarly versatile is their ability to inhibit or block their target molecules as well as to agonistically stimulate or antagonistically blunt a cellular reaction triggered by their target receptor. By utilizing distinct interaction sites differentially, SV-CLRPs target a plethora of molecules, such as distinct coagulation factors and receptors of platelets and endothelial cells that are involved in hemostasis, thrombus formation, inflammation and hematogenous metastasis. Because of their robust structure and their high affinity towards their clinically relevant targets, SV-CLRPs are and will potentially be valuable prototypes to develop new diagnostic and therapeutic tools in medicine, provided that the molecular mechanisms underlying their versatility are disclosed.

Keywords: C-type lectin-related protein (CLRP); C-type lectins; adhesion receptor; coagulation; hematogenous metastasis; hemostasis; platelet; snaclecs; snake venom.

Figure 1

Scheme of (a) the molecular structures of canonical C-type lectins and of (b) non-carbohydrate-binding C-type lectin-related proteins (CLRPs) from snake venoms. Both protein types share a similar set of secondary structure elements. The two β-sheets, consisting of β-strands, β0–β1–β5 and β2–β3–β4. The former brings the N- and C-termini of the protein in close proximity at the N-/C-terminal pole of molecule. The latter contributes to the hydrophobic core and includes the C-type lectin-consensus sequence, W–I–G–L, within the β2 strand. The hydrophobic core is flanked by two amphipolar α-helices, α1 and α2. The highly conserved disulfide bridges are indicated. A long loop is inserted between β-strands, β2 and β3, which clearly distinguish the canonical C-type lectins from CLRPs. (a) In the classical C-type lectins, the long loop folds back to the β2–β3–β4 sheet, together with which it complexes a Ca²⁺ ion and thus shapes the binding site for sugar residues, preferentially galactose residues. The characteristic motifs of Ca²⁺-complexing residues are highly conserved and denoted as E/Q–P–D/N and W–N–D in the long loop and β4 strand, respectively. The W-residue of the latter is part of the hydrophobic core. Additional residues, a basic R and two acidic E residues within the helices, α1 and α2, respectively, as well as an R residue and a less conserved cysteine residue within the long loop are responsible of the assembly of this C-type lectin subunit in higher order aggregates. (b) In CLRPs, the long loop is expanded to an index finger loop domain, which, together with a less conserved cysteine residue, mediates the association of two different CLRP subunits into the typical CLRP heterodimers. Via this index finger loop-swap domain, the two subunits are tilted against each other along their longitudinal axis, resulting in a concave face, also called the bay region. Within the short loop, a structure-stabilizing Ca²⁺ ion complexed by glutamate, serine, and tyrosine residues of the loop connecting the two α-helices, helix α2 and β-strands, and β5 and β1. These residues are shown in subunit α and omitted in subunit β.

Figure 2

Supramolecular structures of canonical C-type lectins and C-type lectin-related proteins (CLRPs) from snake venoms. (a) The snake venom C-type lectins exclusively form homooligomeric structures. Ten subunits of the galactose-binding CTLD subunits from Crotalus atrox assemble into a double pentameric star. Each star consists of five CTLD subunits, whose N-/C-terminal pole points towards the center of the star. The pentamer is stabilized by salt bridges between glutamate and arginine residues (dashed lines). Turned around by 180° along an axis within the plain of the star, the second pentameric ring associates with the first ring and is stabilized by disulfide bridges (-SS-) between the five pairs of homodimers. The galactose-binding domains points outwards. (b) As a basic unit, SV-CLRPs consist of heterodimers, which dimerize via their characteristic index finger loop-swap domain in a slightly tilted manner. This results in a banana-like dumbbell shape of the heterodimeric molecule with a concave face, called the bay region. The N-/C-termini of the two subunits point in opposite directions and constitutes the two ends of the heterodimeric molecule. Such SV-CLRPs assemble into higher aggregates. (c) In rhodocetin, the two heterodimeric subunits form a cruciform tetrahedral molecule. The binding site for α2β1 integrin is shaped by a lateral bay region and is fully activated through conformational changes. (d) and (e) In rhodocytin/aggretin, the two heterodimers associate laterally (d), whereby two (αβ)₂ aggregates even bundle up into a heterooctameric (αβ)₄ complex (e). The binding sites for the CLEC-2 ligands are located at the N-/C-terminal pole of the rhodocytin α subunit. (f) In convulxin and flavocetin, four heterodimeric units join each other into a ring-like structure via a disulfide-stabilized head-to-tail connection at their N-/C-terminal poles. For convulxin, even a double ring assembly with a quaternary structure of (αβ)₈ has been reported.

Figure 3

Platelet receptors are targeted by various snake venom CLRPs. Five different platelet receptors have so far been identified as targets of SV-CLRP: the glycoprotein (GP) Ib, integrin α2β1, von Willebrand factor (vWF) A-domain (which interacts with GPIb), as well as GPVI and CLEC-2. SV-CLRPs that activate platelets and agonistically cause their aggregation are indicated in green; those ones which only aggluinate platelets are shown in gray; and inhibitory and antagonistically platelet-blocking SV-CLRPs are indicated in red. Overlapping receptor specificities were observed for several SV-CLRPs. For some of them, such as rhodocetin and alboaggregin, the binding sites for different receptors are located on different heterodimeric subunits. For others, the mechanism of recognizing different receptors has remained elusive. Depending on the binding partner, SV-CLRPs employ different interaction sites; e.g., on one hand, their concave face for binding of integrin α2β1 and vWF-A-domain (rhodocetin-γδ, EMS16, botrocetin, bitiscetin, albeit with different orientation with respect to the receptor), on the other hand, their N-/C-terminal pole for CLEC-2 binding (rhodocytin/aggretin).