Snake Venom Peptides: Promising Molecules with Anti-Tumor Effects

2.1. The Snake Venom Metalloproteinases (SVMP)

Metalloproteinases are among the most abundant toxins in many Viperidae venoms. SVMPs are monozinc endopeptidases varying in size from 20 to 100 kDa. They are phylogenetically most closely related to the mammalian disintegrin and metalloproteinase (ADAM) family of proteins. SVMPs are grouped into several subclasses according to their domain organization [13, 14, 15]. P-I SVMPs are the simplest class of enzymes that contain only a metalloproteinase (M) domain. P-II SVMPs contain a M domain followed by a disintegrin (D) domain. P-III SVMPs contain M, disintegrin-like (D) and cysteine-rich (C) domains. Formally called P-IV, the heterotrimeric class of SVMPs that contain an additional snake C-type lectin-like (snaclec) domain [16] is now included in the P-III group as a subclass (P-IIId).

Most of the functional activities of SVMPs are associated with hemorrhage or the disruption of the hemostatic system, which are primarily mediated by the proteolytic activity of the M domain. SVMPs cause hemorrhage by disturbing the interactions between endothelial cells and the basement membrane through the degradation of endothelial cell membrane proteins (e.g., integrin, cadherin) and basement membrane components (e.g., fibronectin, laminin, nidogen, type IV collagen) [17]. Blood coagulation proteins (e.g., fibrinogen, factor X, prothrombin) are also targets of their proteolytic activities.

Echis carinatus venom contains the specific prothrombin activators, ecarin [18,19] and carinactivase [20]. Adamalysin II, a non-hemorrhagic P-I SVMP isolated from Crotalus adamantus venom, cleaves and inactivates serum proteinase inhibitors including antithrombin III [21]. Kaouthiagin, isolated from the venom of Naja kaouthia specifically binds and cleaves von Willebrand factor (vWF), resulting in loss of both the ristocetin-induced platelet aggregation and collagen-binding activity of vWF [22]. Additionally, a large number of the P-III SVMPs can inhibit platelet aggregation, thus enhancing the hemorrhagic state [23]. The hemorrhagic P-III SVMP jararhagin from the venom of Bothrops jararaca has been shown to degrade platelet collagen receptor α₂β₁ integrin in addition to fibrinogen and vWF, resulting in the inhibition of platelet aggregation [24]. Other platelet receptors are also degraded by SVMPs. GPIbα is cleaved by kistomin; mocarhagin and crotalin [25-27], and GPVI is degraded by alborhagin, crotarhagin and kistomin [28,29].

In the other side, it was reported that several SVMPs inhibited integrin-mediated adhesion of cancer cells on ECM proteins (table 1). BaG, a dimeric PIII class of SVMP from Bothrops alternatus with inactivated enzymatic domain but intact D/C domain, has been reported to inhibit fibronectin-mediated K562 cell adhesion via α5β1 integrin [30].

Several apoptosis-inducing proteins have been purified from hemorrhagic snake venom, such as VAP1 and VAP2 (Crotalus atrox), HV1 (Trimeresurus flavoviridis), halysase (Gloydius halys), and VLAIPs (Vipera lebetina) [31-34], graminelysin [35]. They are members of the SVMP and ADAM family and induce apoptosis of human umbilical vein endothelial cells (HUVECs) [31,36]. The detachment of endothelial cells and resulting apoptosis could be an additional mechanism for the disruption of normal hemostasis by SVMPs. TSV-DM a basic metalloproteinase from Trimeresurus stejnegeri venom inhibits cell proliferation and induces cell morphologic changes transiently of ECV304 cells. However, DNA fragmentation and DNA content analysis demonstrated that this metalloproteinase could not induce ECV304 cells apoptosis.

2.2. The disintegrins

Disintegrins are a family of non-enzymatic and low molecular weight proteins derived from viper venom [37-39]. They are able to inhibit platelet aggregation and interact with adhesion molecules in particular integrins in a dose-dependent manner. They have a K / RTS sequence which is known as the RGD adhesive loop [37-39]. Their primary structure shows a strong conservation in the arrangement of cysteines [38]. Most disintegrins represent the C-terminal domain of metalloproteinases PIIa, d and e classes and are released into the venom by proteolytic cleavage [40,37,38]. A minority of these proteins exist as D / C domains from the class of SVMPs PIIIb.

Disintegrins can be conveniently divided into five different groups according to their length and the number of disulfide bridges [41]. The first group includes short disintegrins, single polypeptide composed of 49 - 51 amino acids with four disulfide bridges. The second group comprises medium disintegrins containing about 70 amino acids and six disulfide bridges. The third group includes long disintegrins of 83 residues linked by seven disulfide bridges. The disintegrin domains of PIII snake-venom metalloproteinases, containing approx. hundred amino acids with 16 Cysteine residues involved in the formation of eight disulfide bonds, constitute the fourth subgroup of the disintegrin family. Unlike short-, medium- and long-sized disintegrins, which are single-chain molecules, the fifth subgroup is composed of homo and heterodimers. The dimeric disintegrins subunits contain about 67 residues with four disulfide intra-chain bridges and two interchain bridges [42,43].

Although disintegrins are highly homologous, significant differences exist in their affinity and selectivity for integrins, which explains the multitude of effects of these molecules (Table 2).

Disintegrins were first identified as inhibitors of platelet aggregation and were subsequently shown to antagonize fibrinogen binding to platelet integrin αIIbβ3 [44,45]. After that, studies on disintegrins have revealed new uses in the diagnosis of cardiovascular diseases and the design of therapeutic agents in arterial thrombosis, osteoporosis, and angiogenesis-related tumor growth and metastasis (table 2). Triflavin from Trimeresurus flavoviridis venom was one of the first RGD-disintegrins shown to inhibit angiogenesis both in vitro and in vivo [46]. Triflavin strongly inhibited cell migration toward vitronectin and fibronectin nearly thirty orders of magnitude greater than anti-αvβ3 monoclonal antibodies [46]. Triflavin was also more effective in inhibiting TNF-α-induced angiogenesis in the chicken chorioallantoic membrane (CAM) assay. Similar results were obtained with another RGD-disintegrin, rhodostomin, from Agkistrodon rhodostoma venom, which inhibits endothelial cell migration, invasion and tube formation induced by bFGF in Matrigel^TM both in vitro and in vivo [47]. Rhodostomin effects were inhibited by anti-αvβ3 but not by anti-αvβ5 antibodies, thus supporting the hypothesis that the effects of RGD-disintegrins are mediated by blockade of the vitronectin receptor.

Contortrostatin, a disintegrin isolated from the venom of the southern copperhead snake, exhibits anti-cancer activity in a variety of tumor cells [48-50]. It does not display cytotoxic activity in vitro nor animals upon injection. Contortrostatin inhibits adhesion, migration, invasion, metastatic and angiogenesis of tumor and endothelial cells mediated by αvβ3,α5β1 and αvβ5 [48,50-54]. Recently, contortrostatin showed an additive inhibitory effect in combination with docetaxel on the growth of xenograft tumors derived from prostate cancer cells [55].

Lebestatin is an example of a non toxic KTS-disintegrin isolated from Macrovipera lebetina that inhibits migration and VEGF-induced in vivo angiogenesis [56]. The presence of a WGD motif in CC8, a heterodimeric disintegrin from Echis carinatus, increases its inhibitory effect on αvβ3 and α5β1 integrins [57].

There are few reports regarding the effects of ECD-disintegrins on endothelial cell migration. Acurhagin-C, dose-dependently blocked HUVEC migration toward a vitronectin-coated membrane. Furthermore, acurhagin-C elicited endothelial anoïkis via disruption of the αvβ3/FAK/PI3K survival cascade and subsequent initiation of the procaspase-3 apoptotic signaling pathway [58].

Eristostatin, an RGD-disintegrin from Eristocophis macmahoni was tested on individual metastasis steps such as cell arrest, extravasation and migration [59]. Eristostatin treatment did not prevent tumor cell extravasation or migration [60]. However, it was shown later that eristostatin inhibited melanoma cell motility, an effect mediated by fibronectin-binding integrins [61]. Interestingly, this disintegrin, contrary to other RGD-disintegrins, did not inhibit angiogenesis, as stated before [61]. DisBa-01, a αvβ3 integrin-blocking RGD-disintegrin, inhibits not only migration of endothelial cells in vivo [62] but also in vitro migratory ability of fibroblasts and two tumor cell lines.

Since integrin receptors are also quite indiscriminate as they support cell adhesion to several substrates, it seems highly reasonable that the general RGD-disintegrin scaffold of the integrin-binding motif could be employed as a prototype for drug design for new anti-metastatic therapies via blocking both tumor cell adhesion and tumor angiogenesis.

2.3. The snake venom phospholipases

Snake venom is one of the most abundant sources of secretory phospholipases A2 (PLA2), which are one of the potent molecules in snake venoms [63-65].

PLA2 (EC 3.1.1.4)—are enzymes that catalyze the hydrolysis of sn-2-acyl bond of sn-3-phospholipids, generating free fatty acids and lysophospholipids as products [66]. They are currently classified in 15 groups and many subgroups that include five distinct types of enzymes, namely secreted PLA2 (sPLA2), cytosolic PLA2 (cPLA2), Ca²⁺ independent PLA2s (iPLA2), platelet-activating factor acetyl-hydrolases (PAF-AH), lysosomal PLA2, and a recently identified adipose-specific PLA2 [65,67]. PLA2 are low molecular weight proteins with molecular masses ranging from 13-19 kDa that generally require Ca²⁺ for their activities [69,70]. Snake venom sPLA2 are secreted enzymes belonging to only two groups that are based on their primary structure and disulfide bridge pattern [68,71,72]. Those of group I are similar to pancreatic sPLA2 present in mammals, were found in venom of Elapidae snakes, while group II PLA2s belong to the Viperidae and are similar to mammals nonpancreatic, inflammatory sPLA2s [73,74]. The group II can be subdivided mainly in two subgroups, depending on the residue at position 49 in the primary structure: Aspartic acid-49 PLA2s are enzymatically active, while Lysine 49 present low or no enzymatic activity [75]. There are other subgroups, such as Asparagine-49, Serine-49, Glutamine-49 and Arginine -49 [76-83]. Studies have found that catalytic activity is reduced or even abolished when an Aspartic acid of native PLA2 is replaced by another amino acid [80,84].

Despite a high identity of their amino acid sequences, sPLA2 exhibit a wide variety of pharmacological properties such as anticoagulant, haemolytic, neurotoxic, myotoxic, oedema-inducing, hemorrhagic, cytolytic, cardiotoxic, muscarinic inhibitor and antiplatelet activities [63,85-92].

Recently, PLA2s have been shown to possess anti-tumor and anti-angiogenic properties (Table 3). CC-PLA2-1 and CC-PLA2-2 from Cerastes cerastes viper are non-toxic and acidic proteins. They have high inhibitory effects on platelet aggregation and coagulation. In addition, CC-PLA2-1 and CC-PLA2-2 inhibit the adhesion of the human fibrosarcoma (HT1080) and melanoma (IGR39) cells to fibrinogen and fibronectin. In the same direction, CC-PLA2-1 and CC-PLA2-2 potently reduces HT1080 cell migration to fibrinogen and fibronectin with nearly similar IC₅₀ values [93]. This anti-adhesive effect was due to the inhibition of α5β1 and αv-containing integrins [94]. A recent report demonstrated that Bth-A-I, a non-toxic PLA2 isolated from Bothrops jararacussu venom display an anti-tumoral effect upon breast adenocarcinoma as well as upon human leukaemia T and Erlich ascetic tumor [95].

MVL-PLA2 is a snake venom phospholipase purified from Macrovipera lebetina venom that inhibited adhesion and migration of human microvascular endothelial cells (HMEC-1) without being cytotoxic. Using Matrigel^TM and chick chorioallantoic membrane assays, MVL-PLA2, as well as its catalytically inactivated form, significantly inhibited angiogenesis both in vitro and in vivo. Also, the actin cytoskeleton and the distribution of αvβ3 integrin, a critical regulator of angiogenesis and a major component of focal adhesions, were disturbed after MVL-PLA2 treatment. The enhancement of microtubule dynamics of HMEC-1 cells, in consequence of treatments by MVL-PLA2, may explain the alterations in the formation of focal adhesions, leading to inhibition of cell adhesion and migration [96].

A cell death activity was discovered in Lysine 49-PLA2 called BPII. It induces caspase-independent cell death in human leukaemia cells regardless of its depressed enzymatic activity [97].

2.4. The C-type lectins

The C-type lectins are abundant components of snake venom with various function. Typically, these proteins bind calcium and sugar residues. However, the C-type lectin like proteins from snake venom (termed actually snaclec) does not contain the classic calcium/sugar binding loop and have evolved to bind a wide range of physiologically important proteins and receptors [98].

Snaclecs have a basic heterodimeric structure with two subunits, nearly always linked covalently, via a disulphide bond. The heterodimers are often further multimerized either non-covalently or covalently via additional disulphide bonds, to form larger structures [99]. The two subunits form a concave surface between them [100] thus constituting the main site of ligand binding [101,102]. The subunits have a high structural degree of homology between them and with other snaclecs [103]. Despite their highly conserved primary structure, the snaclecs are characterized by various biological activities. They were and are still considered as modulators of platelet aggregation by targeting vWF, GPIb-IX-V, GPVI and possibly other platelet receptors.

Recently, novel activities of snaclecs were highlighted. They were described for their potential anti-tumor effect by blocking adhesion, migration, proliferation and invasion of different cancer cell lines (Table 4). Among these proteins, EMS16, a heterodimer isolated from the venom of Echis multisquamatus, inhibits the adhesion of HUVECs cells on ECM proteins and their migration by inhibiting the binding of integrin α2β1 to collagen [104].

Lebecetin and lebectin, purified from Macrovipera lebetina venom, are the only snaclecs, until today, with an evident anti-tumor effect in addition to their anti-aggregation activity on platelets. Indeed, these two non cytotoxic proteins inhibit the adhesion of various cancer cell lines: melanoma (IGR39), adenocarcinoma (HT29-D4), fibrosarcoma (HT1080) and leukemia cells (K562) on different ECM proteins. They also inhibit the proliferation, migration and invasion of HT1080 cells [105,106]. Lebectin also displays anti-angiogenic activity at very low concentrations both in vitro and in vivo [107]. Thus, lebectin presents the best anti-angiogenic efficacy yet described for snake venom-derived peptides [108,109]. These observed effects are mediated by α5β1 and αv integrins [107].

Extensive researches have been shown that cell adhesion activities in cancer disease are deregulated. According to this idea, it was also reported that lebectin inhibits these alterations by promoting N-cadherin/catenin complex reorganisation at cell-cell contacts, inducing a strengthening of intercellular adhesion [110].

Another snaclec, BJcuL isolated from Bothrops jararacussa venom, was also described for its anti-tumor, but the receptor or integrin implicated has not been determined yet. This homodimeric protein inhibits proliferation of several cell lines of renal, pancreatic, prostate and melanoma origin, but no effect was observed on colon or breast cancer cells [111]. BJcuL also affects the viability of some tumor cell lines of different origins, but has no effect on the growth of K562 and T24 cells, suggesting that these cells do not express the receptor recognized by the lectin. BJcuL induces apoptosis in human gastric carcinoma cells accompanied by inhibition of cell adhesion and actin cytoskeleton disassembly [112].