Tracing Monotreme Venom Evolution in the Genomics Era

3.1. The Crural System

The sole extant representative of the derived form of the monotreme venom system is the semi-aquatic, semi-fossorial platypus (O. anatinus) (Figure 2). The venom apparatus is known as the crural system, because it is in the hind limb, and consists of paired crural venom glands connected by venom ducts to an extratarsal spur on each hind leg [16]. The crural glands are derived from modified apocrine sweat glands and migrate during development from the inner surface of each thigh to their final position on the dorsocaudal surface of the pelvis [17,18].

Uniquely, the venom system in the platypus is sexually dimorphic and appears only in males. A small and non-functional spur sheath develops in females, but is lost by the time they reach adulthood [16]. Venom production occurs at maturity in males, with venom production and gland size vastly increasing during the breeding season [17]. In contrast, during the non-reproductive season, the venom gland secretory epithelium is inactive and the gland regresses [17]. Platypus behavior also changes, with males displaying little aggressive use of their spurs outside the breeding season [17].

The male platypus spur is a hollow keratinous sheath with a cornu calcaris core; the sheath develops within a chalky outer conical structure which gradually wears away to reveal a sharp point [16]. The spur is supported by the os calcaris (this bone is absent in females [18]), and is attached to tendon and muscle to allow erection of the spurs prior to envenomation [14,17]. Platypus envenomation is rarely observed; the animals wrap their hind legs around and drive the sharp spurs into the victim. The animals are able to hang by their spurs, requiring manual disengagement [19]. Although platypuses have up to 4 mL of venom available to be injected at any one time [17], in practice, the small diameter of the spur aperture (~0.2 mm) and associated high pressure required for injection mean that injected venom volumes around 100 µL are predicted [5].

3.2. Venom Function

Early naturalists speculated that male platypuses used their spurs to grip the female during mating [20], but there is little to no evidence to support this hypothesis [16]. A prey capture or digestive function can also be ruled out, as the platypus diet consists mainly of benthic invertebrates [16]. A defensive function is possible, but in this case presumably the spurs would be retained in both sexes. Although foxes and dogs kill platypuses, the species has very few native predators (occasional exceptions are crocodiles, Tasmanian devils and raptors) [16,21]. Perhaps, as discussed in Section 2, this was an ancestral function of the venom, which has since been recruited for other uses once the selective pressure of predation was lifted.

The venom system in extant platypuses undoubtedly serves some purpose, as the venom produces strong pharmacological effects (see Section 3.3). As venom is generally metabolically expensive to produce (reviewed in [22]), its retention implies some utility. The restriction of venom to adult male breeding season platypuses suggests that the venom has a reproductive function, such as use in male-male competition and territory defense during the reproductive period. Healed spur marks in wild platypuses suggest that intraspecific envenomation does occur and is not fatal [16]. Envenomated male platypuses are observed to experience temporary limb paralysis [17], which might be sufficient to either deter territory encroachment and/or to temporarily prevent mating by competitors. Platypus venom may thus be an important component of male reproductive fitness.

3.3. Platypus Venom Composition: Pharmacological and Proteomic Characterization

There are few formal clinical reports of platypus envenomation of humans. Envenomation is known to produce the following physiological effects: swelling; a long lasting and excruciating pain that cannot be relieved with conventional painkillers, including morphine [19]; nausea, cold sweats, and lymph node swelling [17]; high erythrocyte sedimentation rates and low total serum protein and albumin levels [19]; and muscle wasting [19]. There have been no recorded human fatalities. Anecdotal evidence suggests that envenomation can kill dogs [20]. There is no antivenin available.

There is a dearth of research into platypus venom compared to that of other animals, despite the potential for drug discovery using newly identified venom compounds [23]. Until recently, platypus venom research was limited to a few biochemical and pharmacological studies, and the venom was poorly characterized due to the difficulty in obtaining samples for study [5]. Early pharmacology studies injecting platypus venom into rabbits produced intravascular coagulation, a drop in blood pressure (likely through vasodilation), hemorrhagic edema, and death [24,25]. Further research identified mild proteolytic activity, histamine release, and cutaneous anaphylaxis in laboratory animals [17]. In vitro experiments revealed that platypus venom produces smooth muscle relaxation [26], feeble hemolysis [25], proteolysis [26], and cation currents in cells [27] or cation channels in artificial lipid bilayers [28]. In vivo, these effects may disrupt ion concentrations and cause edema, nerve firing, and pain.

Subsequent high performance liquid chromatography (HPLC) studies of platypus venom revealed several peptides and proteins of unknown function [26,29,30]. Until the advent of genomic technologies only three types of peptides had been identified and fully sequenced: the particularly biologically active C-type natriuretic peptides (OvCNPs) [31], the abundant defensin-like peptides (OvDLPs, [30]), and nerve growth factor (OvNGF, Torres A and Kuchel PW, unpublished data; described in [32]). A venom peptide isomerase and a hyaluronidase were also found, but not fully sequenced [26]. The peptide isomerase catalyzes the conversion from l- to d-form of the second amino acid from the N-terminus of both an OvDLP and an OvCNP [33]; the d-form of the amino acid may confer prolonged stability and therefore toxin efficacy in the victim [34]. These components of platypus venom have similarity to known toxins in other venomous animals, and have unknown function [5,12].

3.4. Platypus Venom Composition: Genomic Insights

The sequencing of the platypus genome in 2008 [7] provided an unprecedented resource for platypus venom research, and has resulted in a vast increase in our knowledge of platypus toxins. Identification of toxin gene sequences allowed investigations into their tissue expression patterns to facilitate an understanding of their functions and evolution [35,36]. The most notable advance came with the next-generation sequencing of a platypus venom gland transcriptome. This representation of the genes expressed in the tissue reveals a number of predicted new platypus toxins [37]. The putative toxins are categorized into thirteen families: Serine protease, Stonustoxin-like, Kunitz type protease inhibitor, Zinc metalloproteinase, Latrotoxin-like, CRiSP (cysteine rich secretory protein), Sea anemone cytolytic toxin-like, Unknown (IG domains), Mamba intestinal toxin-like, C-type lectin domain-containing, Sarafotoxin-like, VEGF (vascular endothelial growth factor), and DNAse II. These are similar to toxins across a wide range of other venomous animals, including fish, reptiles, spiders, and marine invertebrates. Although the molecules found in platypus venom have not been functionally tested, it is possible to speculate about their effects, ranging from coagulation and inflammation caused by the 26 putative platypus serine proteases, to pain caused by the stonustoxin-like (18 putative toxins) and latrotoxin-like (7 putative toxins) families [37]. This study probably uncovered the majority of platypus venom genes, but the methodology based on similarity searches using known toxins meant that it was likely to miss completely novel venom genes. In order to identify novel venom genes with no similarity to known toxins in other species, it is possible to take advantage of the seasonal nature of platypus venom production by comparing gene expression levels in active versus regressed glands. Genes that are highly expressed in active glands compared to regressed glands are probable toxin genes. This analysis reveals five more putative toxins unique to the platypus: growth differentiation factor 15, nucleobindin-2, complement decay-accelerating factor 55, CXC-chemokine, and corticotropin-releasing factor-binding protein [38]. In the future, functional studies are required to characterize the role of these putative toxins in platypus venom, which may reveal candidates for the discovery of novel drugs (reviewed in [39]).

3.5. Platypus Venom Gene Evolution

As well as identifying a wealth of new putative toxins, genomic studies of platypus venom have enabled characterization of the evolutionary origins of this trait. Ultimately, genetic origins of toxins tend to reflect the tissue origins of venom glands. Snake venom glands are specialized salivary glands, derived from the pancreas, and there are a number of snake toxins that are of pancreatic origin [40] or salivary gland origin [41]. In contrast, the platypus venom glands are thought to be derived sweat glands [17], and many of the platypus venom genes, such as the kallikreins and defensins, are derived from skin-expressed gene families [11,37].

At a proximate level, genomics has uncovered several key mechanisms acting upon platypus venom genes and venom genes in other species. Notably, gene duplication is common, with many toxin families containing multiple members [11,37]. Gene duplication occurs when a gene is replicated within the genome, for example by unequal crossing over during meiosis, to form two copies. One copy performs the original function of the gene; the second copy is then free to vary, and neofunctionalization (the development of new functions) may occur [42]. The process of gene duplication is a significant source of variation for the evolution of novel traits via either positive selection (adaptive evolution) or genetic drift. There is evidence of gene duplication in the evolution of platypus venom genes, where venom peptides have been derived via duplication from genes encoding non-toxins. For example, the OvDLPs have evolved via gene duplication from the antimicrobial beta-defensins thought to protect altricial platypus hatchlings from infection [43], after which they neofunctionalized to become toxins [11]. Gene duplication is also a source of platypus venom diversification, as it has generated large multigene families of toxins, possibly resulting in increased expression levels of a particular toxin type [36]. Gene duplication has been found to be an important process in the evolution of a number of other venoms (reviewed in [44]), particularly those of cone snails, spiders, and snakes (e.g., [45,46,47,48]).

Gene duplication alone does not explain the evolution of the platypus venom genes, revealing that additional mechanisms have played a role in venom evolution. Mechanisms that may also be important sources of new toxins include recruitment of multiple genes from the same family into venom gland expression; mutations in regulatory or coding regions; and alternative splicing [49]. These processes are important in the evolution of venoms in other species, for example in various snake venoms, which exhibit alternative splicing of an acetylcholinesterase gene [50] and modification of existing cysteine rich secretory protein and kallikrein genes [41]. These and further evolutionary processes, such as domain duplication and domain loss (reviewed in [44]), may be identified in platypus venom in the future.

Many platypus toxins have been recruited from existing non-toxin gene families, some of which have been independently co-opted as venom toxins in other species. It thus appears that there are certain protein motifs that are repeatedly selected during the evolution of venom molecules across divergent species (i.e., convergent evolution) [37]. Convergent recruitment of venom toxins has been demonstrated across a wide range of venomous taxa, including hymenoptera, ticks, scorpions, spiders, cephalopods, cnidarians, cone snails, fish, shrews (reviewed in [1]), and now platypus effectively covering all major venomous phyla. These repeated, independent iterations of venom development allow us to determine the features that predispose a protein to evolve into a venom toxin: proteins with stable scaffolds that can be slightly modified to allow diversification into multimember families with differing activities; extensively cysteine cross-linked proteins; and secretory proteins (e.g., [1]). These features are displayed by many of the platypus toxins, including the OvDLPs (cysteine cross-linked) and serine protease inhibitors (a multigene toxin family).