Spiders are the most successful venomous animals with an estimated 100,000 extant species [1]. The vast majority of spiders employ a lethal cocktail to rapidly subdue their prey, which are often many times their own size. However, despite their fearsome reputation, less than a handful of these insect assassins are harmful to humans [2,3]. Nevertheless, it is this small group of medically important species that first prompted scientists more than half a century ago to begin exploring the remarkable pharmacological diversity of spider venoms.

Amongst the ranks of animals that employ venom for their survival, spiders are the most successful, the most geographically widespread, and arguably consume the most diverse range of prey. Although the predominant items on a spider’s dinner menu are other arthropods, larger species will readily kill and feed on small fish, reptiles, amphibians, birds, and mammals. Thus, spider venoms contain a wealth of toxins that target a diverse range of receptors, channels, and enzymes in a wide range of vertebrate and invertebrate species.

Spider venoms are complex cocktails composed of a variety of compounds, including salts, small organic molecules, peptides, and proteins [4,5,6,7,8,9]. However, peptides are the primary components of spider venoms, and some species produce venom containing >1000 unique peptides of mass 2–8 kDa [10]. Based on the number of described spider species and a relatively conservative estimate of the complexity of their venom it has been estimated that the potential number of unique spider venom peptides could be upwards of 12 million [11]. In recent years there has been an exponential increase in the number of spider-toxin sequences being reported [12] due to the application of high-throughput proteomic [13,14] and transcriptomic [15,16,17] approaches, or a combination of these methods [10,18,19]. In the last 18 months alone the number of toxins in the ArachnoServer spider-toxin database [20,21] has more than doubled, and is now excess of 900 (see http://www.arachnoserver.org/). Nevertheless, our knowledge of the diversity of spider-venom peptides is still rudimentary, with less than 0.01% of potential peptides having been isolated and studied.

Although only a small number of spider venom peptides have been pharmacologically characterized, the array of known biological activities is impressive [9]. In addition to the well known neurotoxic effects of spider venoms, they contain peptides with antiarrhythmic, antimicrobial, analgesic, antiparasitic, cytolytic, haemolytic, and enzyme inhibitory activity. Furthermore, the crude venom of Macrothele raveni has antitumor activity, for which the responsible component has not yet been identified [22,23]. Finally, larger toxins such as the latrotoxins from the infamous black widow spider (Latrodectus mactans) and related species induce neurotransmitter release and they have played an important role in dissecting the process of synaptic vesicle exocytosis [24].

Since spiders employ their venom primarily to paralyse prey, it is no surprise that these venoms contain an abundance of peptides that modulate the activity of neuronal ion channels and receptors. Indeed, the majority of characterized spider-venom peptides target voltage-gated potassium (K_V) [25], calcium (Ca_V) [26,27], or sodium (Na_V) [26,28] channels. More recently, novel spider-venom peptides have been found that interact with ligand-gated channels (e.g., purinergic receptors [29]) and recently discovered families of channels such as acid sensing ion channels [30], mechanosensitive channels [31], and transient receptor potential channels [32]. Not only do most of these peptides have selectivity for a given class of ion channel, they can have anything from mild preference to exquisite selectivity for a given channel subtype. This potential for high target affinity and selectivity makes spider-venom peptides an ideal natural source for the discovery of novel therapeutic leads [33].

Despite the advent of automation and the rise of high-throughput and high-content screening in the pharmaceutical industry there has been a sharp decline in the rate of discovery and development of novel chemical entities [34,35]. We recently reviewed the emerging role that venom-derived components can play in addressing this decline with an emphasis on technical advances that can aid the discovery process [36]. It is worth noting that, as of 2008, two of the 20 FDA-approved peptide pharmaceuticals were derived from animal venoms (i.e., ziconitide and exendin-4) [37]. In this review we specifically examine the structure, targets, and mechanisms of action of spider-venom peptides with potential therapeutic applications.

All peptide names are based on the rational nomenclature for naming peptide toxins that has been adopted by ArachnoServer and UniProtKB [12]. Common synonyms are also provided.

Peptides have generally been considered poor candidates for human therapeutics because of their susceptibility to proteolytic degradation in vivo and their limited penetration of intestinal mucosa [37,38]. However, in contrast with most peptides, the presence of an inhibitor cystine knot (ICK) in most spider-venom toxins provides these peptides with extraordinary stability. The inhibitor cystine knot (ICK) is defined as an antiparallel β sheet stabilized by a cystine knot [39,40,41]. In spider toxins, the β sheet typically comprises only two β strands although a third N-terminal strand is sometimes present (Figure 1A) [42]. The cystine knot comprises a ring formed by two disulfides and the intervening sections of polypeptide backbone, with a third disulfide piercing the ring to create a pseudo-knot (Figure 1B). The compact hydrophobic core of the ICK motif consists primarily of the two central disulfide bridges that emanate from the two β strands that characterize the ICK fold [43]. Except for the special case of cyclic ICK peptides, cystine knots are not true knots in the mathematical sense as they can be untied by a non-bond-breaking geometrical transformation [44]. Nevertheless, the cystine knot converts ICK toxins into hyperstable mini-proteins with tremendous chemical, thermal, and biological stability. ICK toxins are typically resistant to extremes of pH, organic solvents, and high temperatures [45]. However, from a therapeutic perspective, their most important property is their resistance to proteases; ICK peptides are typically stable in human serum for several days and have half-lives in simulated gastric fluid [46] of >12 hours (GFK and VH, unpublished). It was recently demonstrated that stabilization of a 16-residue α-conotoxin through cyclization dramatically increased its oral activity [47], and it is therefore possible that the inherent stability of ICK peptides might impart them with oral activity without the need to introduce exotic modifications.

ICK toxins have proliferated in spider venoms to the point where they now dominate most spider-venom peptidomes. The marked insensitivity of this structural scaffold to changes in intercystine residues has enabled spiders to develop diverse pharmacologies using the same disulfide framework [48]. Moreover, many of these ICK peptides not only have high affinity but also exquisite selectivity for their cognate targets. With the exception of those with antibacterial/antifungal activity, all of the spider-venom peptides to be discussed in this review contain an ICK motif.

Normal nociceptive pain is a key adaptive response that limits our exposure to potentially damaging or life-threatening events. In contrast, aberrant long-lasting pain transforms this adaptive response into a debilitating and often poorly managed disease. About 20% of adults suffer from chronic pain, a figure that increases to 50% for those older than 65 [49]. In 2007, global sales of pain medications totaled $34 billion [50], highlighting the pervasive nature of this condition. Nevertheless, there are few drugs available for treatment of chronic pain, and many of these have limited efficacy and significant side-effects. Recently, a number of ion channels have been shown to be critical players in the pathophysiology of pain, and in many cases the most potent and selective blockers of these channels are spider-venom peptides. Here we review some of these peptides with promise as drug leads or as analgesics in their own right.

Mechanosensitive channels (MSCs), sometimes referred to as stretch-activated channels, are found in all cells [90], but mechanosensitivity is best viewed as a phenotype rather than a genotype [91]. Only two selective inhibitors of MSCs have been isolated, namely M-TRTX-Gr1a (GsMTx4) and κ-TRTX-Gr2a from the venom of the tarantula Grammostola rosea [92,93]; the latter peptide is identical to κ-TRTX-Ps1b (Phrixotoxin-2) isolated from venom of the Chilean copper tarantula Paraphysa scrofa. κ-TRTX-Gr2a/κ-TRTX-Ps1b is a low affinity blocker of MSCs, with a K_d of 6 μM in rat astrocytes, but it potently inhibits K_V4.2 and K_V4.3 channels, with IC₅₀ values of 34 and 71 nM, respectively. In contrast, M-TRTX-Gr1a is a significantly more potent inhibitor of MSCs, with a K_d of 630 nM in rat astrocytes [92], and it has proved to be a valuable tool for study of MSCs [31].

Atrial fibrillation results from stretching of the atrial chamber, which has been associated with the activity of mechanosensitive channels. Block of these channels presumably explains the remarkable observation that M-TRTX-Gr1a suppresses atrial fibrillation in dilatated rabbit heart [94]. This suggests that MSCs might be a novel target for antiarrhythmic agents. M-TRTX-Gr1a itself is unlikely to be a useful therapeutic agent because of its unusual mode of action. It does not interact directly with MSCs, since an enantiomer comprised entirely of D-amino acids is equipotent with the native peptide [31]. Rather, the peptide perturbs the channel-bilayer boundary by partitioning into the membrane [95,96], and this membrane-disrupting activity presumably also underlies its antimicrobial activity [97]. Nevertheless, M-TRTX-Gr1a is likely to be a useful tool for determining the potential of MSCs as a therapeutic target for the treatment of pathologies as diverse as cardiac arrhythmias, spinal cord damage, muscular dystrophy, and gliomas [31].

Penile erection is a complex process initiated by activation of parasympathetic pelvic nerves, resulting in arterial dilatation followed by relaxation of corpora cavernosa [98]. Nitric oxide (NO) plays a major role in the generation and maintenance of intracavernous pressure and penile erection [99]. NO, which is released from nitrergic nerves within the trabecular and arterial tissues as well as by the endothelial tissue of penile arteries, exerts its relaxing action by activating soluble guanylyl cyclase. This causes an increase in intracellular cGMP which relaxes the smooth muscles of the cavernous body and results in penile erection.

There are several drugs on the market today for treatment of erectile dysfuncion (ED), including sildenafil (Viagra®), tadalafil (Cialis®), and vardenafil (Levitra®) [100]. These drugs all affect phosphodiesterase type 5 (PDE5), which is present in large amounts in the penis. Inhibition of PDE5 leads to increased levels of cGMP and hence increased blood flow to the penis [101]. The aforementioned PDE5-blocking drugs have similar side effects including headache, flushing, dyspepsia, nasal congestion, impaired vision, photophobia and blurred vision [102]. Hence, there is a need for better drugs with fewer side effects for the treatment of ED.

In South America, humans bitten by the “armed-spider” Phoneutria nigriventer experience a variety of symptoms including priapism. The toxin responsible for this effect, δ-ctenitoxin-Pn2a (δ-CNTX-Pn2a; Tx2-6) was isolated in 1992 [103] and subsequently found to modulate the activity of Na_V channels [104]. δ-CNTX-Pn2a is a 48-residue toxin with five disulfide bonds; it is unclear whether it contains an ICK motif. The toxin has a complex pharmacology that results in inhibition of Na_V channel inactivation and a hyperpolarizing shift in the channel activation potential [104]. Intracerebroventricular injection of δ-CNTX-Pn2a into mice results in scratching, hypersalivation, lachrymation, sweating, and agitation followed by spastic paralysis of the anterior and posterior extremities and death [103]. However, subcutaneous injection of δ-CNTX-Pn2a into rats induced erection, and rats with severely depressed erectile function could be normalized by subcutaneous administration of δ-CNTX-Pn2a [105]. A minimum dose of only 0.006 μg/kg was required to cause an erection in mice when injected directly into the corpus cavernosum; at this dose, no local and systemic collateral toxic effects were observed and the erection was lost after 120–140 min [106].

The potency and specificity of δ-CNTX-Pn2a makes it an attractive lead molecule for the development of new therapeutics for ED treatment that might have fewer side effects than current drugs. The pharmacology inherent in δ-CNTX-Pn2a may be widespread in spider venoms, at least amongst members of the Ctenidae family (which currently comprises 475 species [107]), since orthologous toxins are found in several other Phoneutria species as well as a related ctenid spider (see ArachnoServer entry http://www.arachnoserver.org/toxincard.html?id=16).

The introduction of antibiotics in the 1930s and 1940s was in large part responsible for the dramatic decline in the mortality rate from communicable diseases in developed countries [108]. However, bacteria are remarkably proficient at adapting to environmental stresses, and they have evolved at least one mechanism of resistance for all 17 classes of antibiotics that have been developed to date [109]. The recent widespread emergence of antibiotic resistance in clinically important bacterial pathogens such as Staphylococcus aureus, Streptococcus pneumoniae, and Enterococcus faecalis, combined with a dramatic decrease in the rate of development of new antibiotics, has led some to suggest that we may be approaching the post-antibiotic era [109]. While this may overstate the problem, there is nevertheless an urgent need to develop new antimicrobials with novel mechanisms of action.

The recent successful introduction of the lipopeptide antibiotic daptomycin [110] has rekindled interest in antimicrobial peptides [111]. To date, 40 membrane-acting antimicrobial peptides (MAMPs) have been isolated from the venom of four different families of araneomorphs, suggesting that antimicrobial activity is widespread in this infraorder of spiders. These MAMPs often have a wide range of antimicrobial activities, with some toxins active against Gram-positive and Gram-negative bacteria as well as fungal pathogens such as Candida albicans [112]. Some MAMPs also have anti-trypanosomal activity [113]. Curiously, no MAMPs have thus far been isolated from the venom of mygalomorph spiders. It is possible that MAMPs were recruited into araneomorph venoms following the split from mygalomorphs around 280 million years ago [114].

MAMPs differ from most other spider-venom peptides in their structure and mode of action. Rather than utilizing the ICK fold common to most spider toxins, MAMPs are α-helical amphipathic peptides that interact with and perturb cell membranes to yield their antimicrobial effects [115,116,117,118]. This mode of action can be potentially problematic from a therapeutic perspective since MAMPS that interact nonspecifically with cell membranes are cytolytic [119]. Indeed, this property is likely to be the basis of their biological function in spider venoms. Although it has been proposed that MAMPS might protect the spider’s venom apparatus against infection [115] their primary role is more likely to be as membrane disrupting agents that augment the activity of the disulfide-rich neurotoxic peptides by facilitating their spread [120,121].

It has been shown that the cytolytic activity of at least some spider-venom MAMPs can be minimized by truncation without significantly disrupting their antimicrobial activity [122]. Nevertheless, the therapeutic use of these peptides is likely to be limited by their inherent susceptibility to proteolysis, which is likely to result in short gut and plasma half-lives. Whether this problem can be solved by strategies such as cyclization or grafting key sequence elements onto more stable ICK scaffolds [123] remains to be seen.

The antimicrobial action of spider toxins is not limited to bacteria and fungi, but also extends to the malaria parasite. There were 243 million cases of malaria in 2008, resulting in a death every 35 seconds, and most of these were children under the age of five [124]. Malaria is caused by Plasmodium infections spread by female anopheline mosquitoes. There are five genera of Plasmodium that cause malaria, with Plasmodium falciparum being the most virulent. Widespread resistance to chloroquine has made this drug largely ineffective for treating Plasmodium falciparum in high-transmission areas and few cheap alternatives are available [125].

U₁-TRTX-Pc1a (Psalmopeotoxin I) and U₂-TRTX-Pc1a (Psalmopeotoxin II) are ICK peptides isolated from the venom of the Trinidad chevron tarantula Psalmopoeus cambridgei that are effective against the intra-erythrocyte stage of Plasmodium falciparum. Interestingly, this is the same spider from which π-TRTX-Pc1a, the most potent known blocker of ASIC1a, was isolated, indicating that a single spider can provide multiple therapeutic leads. U₁-TRTX-Pc1a and U₂-TRTX-Pc1a are unrelated peptides that comprise 33 and 28 residues, respectively. They inhibit intra-erythrocyte development of Plasmodium falciparum with ED₅₀ values of 1.1–1.6 μM but, unlike most MAMPS, they do not have hemolytic, antibacterial or antifungal activity. The mode of action of these peptides is unknown. It seems unlikely that they directly target the malaria parasite since this would require the peptides to traverse both the erythrocyte membrane as well as the parasitophorous vacuolar membrane that encapsulates the parasite. One possibility is that these toxins target the new permeability pathways that are established in the erythrocyte membrane following parasite invasion [126]. Thus, in addition to being useful therapeutic leads, these peptide toxins might help validate a new anti-malarial drug target.

In the preceding sections we surveyed a number of spider-venom peptides with potential therapeutic applications (Figure 2). With the exception of the MAMPs that have antimicrobial/antifungal activity, all of these peptides are stabilized by an inhibitor cystine knot motif. Thus, they are unlikely to suffer from one of the major problems typically associated with peptide drugs, namely rapid proteolytic degradation. As any surviving bite victim will attest, the devastating in vivo efficacy of the 42-residue lethal Na_V channel toxin from the Australian funnel-web spider [42,127] plainly demonstrates how minute quantities of ICK peptides can be effective in humans. Should rapid proteolysis prove to be an issue for a spider-venom peptide of therapeutic interest, strategies such as D-amino acid substitution of susceptible residues, cyclization to reduce conformational flexibility, and protection of the termini via C-terminal amidation or use of N-terminal pyroglutamate could be employed to improve proteolytic resistance [37,128,129]. Indeed, spiders routinely employ C-terminal amidation as a protection strategy, with ~12% of all known spider toxins containing this posttranslational modification [21].

Because of their inherent proteolytic resistance, the plasma half-life of ICK peptides is likely to be determined by the rate at which they are cleared by glomerular filtration, which efficiently removes small proteins and peptides that are not bound to carrier proteins such as serum albumin [130]. However, there are a variety of strategies that can be employed to reduce peptide clearance rates, such as increasing the peptide mass by PEGylation, conjugation to carrier proteins, or by making peptides more hydrophobic in order to enhance their association with serum albumin; the latter approach can also cause hydrophobic depoting, which provides sustained release from a subcutaneous injection [37].

For acute life-threatening conditions as well as chronic conditions such as persistent pain, subcutaneous injection is likely to be an acceptable route of peptide-drug administration. In certain pathologies where quality of life is dramatically reduced, intrathecal administration may even be a viable option if the difficulties associated with peptide delivery are outweighed by the benefits of treatment. Generally, however, oral delivery is likely to be desirable. The intrinsic stability of ICK peptides is likely to facilitate the development of oral delivery strategies since they will presumably have much longer gut and plasma residence times than typical peptides. Moreover, spider-venom ICK peptides are small enough to consider alternative routes of administration such as intranasal, transdermal, and pulmonary [131].

An alternative but complementary approach is to develop small-molecule mimetics of spider-venom peptides. The epitope (pharmacophore) that mediates the interaction of these peptides with their cognate receptors or ion channels can be remarkably small. For example, the interaction between the spider-venom peptide ω-hexatoxin-Hv1a and invertebrate Ca_V channels is mediated by a pharmacophore comprising only three spatially contiguous residues with a solvent-accessible surface area of ~200 Å [132,133], which approximates the typical solvent-accessible surface area of a small drug [134]. As long as a high-quality structure of the peptide is available, this enables ab initio design of nonpeptide mimetics [135,136], identification of small molecule mimetics via in silico screening of chemical libraries, or a combination of these approaches.

Over a period of more than 300 million years, spiders have evolved an extensive library of bioactive peptides. Moreover, in contrast with man-made combinatorial peptide libraries, spider-venom peptides have been pre-optimized for high affinity and selectivity against a diverse range of molecular targets. It is therefore not surprising that numerous spider-venom peptides have been characterized that potently and selectively modulate the activity of a diverse range of therapeutic targets. These include peptides that target Na_V channels, ASICs, MSCs, and purinergic receptors as well as peptides with antimalarial and antimicrobial activity. Most of these peptides contain an ICK motif, and their extraordinary stability provides a variety of delivery options for therapeutic administration. Only a small fraction of spider-venom peptides have been characterized, and continued technical advances in the venoms-based drug discovery process are likely to uncover many new therapeutic leads from spider venoms.