Chapter 5 - Armed and Dangerous
5. Armed and Dangerous
The poison fangs are artfully contrived, by some diabolical freak of nature, as pointed tubes, through which the poison is injected into the base of the wound inflicted. The extreme point of the fang is solid, and is so finely sharpened that beneath a powerful microscope it is perfectly smooth, although the point of the finest needle is rough!
Sir Samuel Baker, 1866
A late night drive up the Arima Valley in Trinidad’s Northern Range had been very unproductive. I was alone and tired. It had not rained in several days, so the frogs were silent, and other than the occasional large bat and a few tarantulas there was little to see. Close to the field station I stopped for a Terciopelo, a common and widespread pitviper. It was not large, perhaps 70 cm. As I approached it did not move, and when I attempted to slide a snake hook under the snake’s body it responded unexpectedly. It struck at me, and in the process lifted its entire body off the pavement so that it was level with the headlight on the car. With a few more maneuvers, the snake was in the bag. Seeing the small snake strike that distance was un-nerving. The rule of thumb for pitviper striking distance is about a third of their body length. This animal launched a strike that was equal to its body length. The following morning I removed the snake from its bag, only to find it dead. Apparently a car had hit the snake before I found it. A surge of adrenalin in the injured snake may explain the spectacular strike.
The first person to recognize tooth structure could be used to distinguish between dangerous and harmless snakes appears to be John Ray in 1693. But, dentition characteristics did not enter into the Linnean system for the classification of snakes and Ray’s observations were lost in history. In 1787, Patrick Russell wrote a memorandum to physicians working for the East Indian Trading Company detailing the fact that harmless snakes had a complete row of maxillary teeth while venomous snakes had the maxillary teeth replaced with one- or two fangs. Russell’s work became well known through his two volumes, An Account of Indian Serpentes Collected on the Coast of Coromandel. He failed to detect the venomous nature of rear-fanged snakes through his experiments with Ahaetulla nasuta (what he called coluber Mycterizans). He suggested rear-fanged species were harmless, an idea that Schlegel supported and expanded (see Chapter 4).
In the first decade of the 21st century, it has become clear that virtually all snakes have a venom gland and teeth modified into fangs do not have to be present for snakes to produce venom. Keep in mind, however, that the presence of venom does not necessarily translate to “dangerous to humans.” Of the 3500 species of known snakes, less than 5% pose any serious threat to human health.
The venom apparatus of snakes usually consists of a venom gland and associated muscles, a duct to transport venom, an accessory gland that may be a distinct gland or part of the venom duct, a compartmentalized sheath around the fang, and, of course, the fang. That said, venom glands vary widely and may be small and virtually non-functioning or huge and extend from the head well into the body cavity. Venom ducts may lead from the gland to the fang, but some ducts are branched and open into the mouth near the teeth and are not directly connected to a fang. Fangs come in a variety of shapes and sizes and may have hollow channels; deep, open grooves; shallow grooves; or be smooth with no grooves at all. They may be quite long or virtually the same size as the other teeth and, while many snakes have been traditionally thought to lack them completely, all snakes have them. A tooth on the posterior dental lamina of the snake embryo is the homologue of the tubular fang. Intermediate states between the posterior tooth and tubular fang exist throughout living snakes. This definition renders the term fang virtually meaningless. Here, I will apply the term to any maxillary tooth that is longer than the others, any grooved tooth, or any tooth that has a venom conducting canal.
Sir Samuel Baker recognized the elegant, yet deadly nature of viper fangs when he described the venom conducting teeth of the African Puff Adder. His comment characterizes the hypodermic-like structure of a viper’s fang but, as noted, not all fangs have the same structure. Variations in fang structure represent adaptations to prey and predators, as well as the snake’s ancestry. The variety of fangs is quite impressive and yet not surprising when one closely examines the diversity of snakes.
When George Haas took a close look at the teeth of the limbed Cretaceous fossil snake Pachyrhachis, he found three of the exposed teeth were, “…grooved along their anterior slope.” It is doubtful that these parallel the deep grooves seen in many of the rear-fanged snakes, but more likely, something similar to the shallow furrows present on the surface of many fish-eating species. Evolution has been tinkering with snake fangs for at least 100 MY. The fossil record also supports the idea that the hypodermic-like snake fangs of those of modern snakes in their present form have been around for a while. Fangs virtually identical to modern viper and elapid fangs were recovered from a German quarry, and date to 23 MYA. Furthermore, the DNA clock suggests vipers are much older than this.
Snakes usually have teeth located on six different bones in the mouth. This led one 19th-century author to describe snakes as having six jaws. In most snakes, teeth are found on the right and left upper jaws (the maxillary bones), the right and left lower jaws (dentary bones), and on the two pterygoid bones (located on the roof of the mouth behind the palatine bone). The prefrontal and palatine bones may also sport teeth, and at least two species are known to carry their longest teeth on the palatine bone. All of the bones with their associated teeth are capable of independent movement, making it entirely appropriate to describe snakes as six-jawed creatures. So what do they do with all of these teeth? Snakes can’t chew in the sense that they do not grind up their food. Additionally, they lack limbs for prey handling. As a result, some teeth have adapted for holding food while others transport it to the back of the mouth where it is swallowed whole. And, of course, other teeth deliver venom. Here we will focus on teeth located on the maxillary bone closely associated with the venom glands.
Conflicting hypotheses on fang origin and the snakes that have them are largely the results of the idea that snakes can be categorized as advanced, transitional, or primitive. At the end of the 19th century, Edward Cope at the Philadelphia Academy of Sciences and George Boulenger at the British Museum of Natural History believed colubrid snakes without fangs were ancestral to the snakes with open-grooved fangs. Cope, and Boulenger also proposed the hypothesis that colubrids gave rise to a second lineage of snakes, the fixed-front fanged elapids. According to their thinking, elapids then gave rise to the vipers with their hypodermic-like fangs capable of folding back on a rotating maxillary bone [Figure 5–1]. Elapid fangs were considered ancestral to viper fangs because a visible seam found along the front surface in many elapid species. The seam was thought to result from the tooth folding during development and, for this reason, elapid fangs were considered to be slightly incomplete. Also, elapids usually have fixed, immovable fangs with smaller, open-grooved teeth behind the fangs on the maxillary. Vipers, on the other hand, have fangs with a central, enclosed canal for venom delivery, a smooth surface, and the maxillary bone lacking any other teeth. By the mid-20th century, species with grooved rear fangs were thought to represent the ancestral condition to the front fangs of elapids and vipers.
INSERT FIGURE 5-1.
Figure 5–1. Edward D. Cope and George Boulenger’s concept of snake phylogeny in the late 19th century is illustrated above. The ancestral snakes (Peropoda) were the boas, pythons, and other snakes that had remnant limbs (except the scolecophidians). The Tortricina were the small-gape pipe snakes. The Aglyphodonta were snakes without fangs and mostly considered colubrids. The Opisthoglypha were snakes with rear-fangs. The Proteroglypha were snakes with fixed fangs, mostly elapids. The Solenoglypha were the vipers and the pit vipers. Redrawn and modified from Cope (1900, The Crocodilians, Lizards, and Snakes of North America, Washington: U. S. National Museum).
Fang Development
Hypotheses about the evolutionary origin of fangs were not in short supply. Front-fanged snakes were thought to have shared a common, rear-fanged ancestor whose posterior teeth gave rise to more modern front fanged snakes. Others thought elapid fangs developed from front teeth, and that viper fangs developed from rear teeth. And yet still others thought both elapid and viper fangs evolved from rear fangs. Solid fangs, grooved fangs, and fangs with central venom canals were seen as a continuum in a variety of combinations. Add to this, the facts that most elapid fangs have a seam running along the front of the fang, while viper fangs tend to be solid and not show a seam. This is anatomy that could be interpreted to suggest steps from ungrooved to grooved (partially folded fang), to an almost closed canal (an almost completely folded fang), to a completely closed venom canal (a completely closed folded fang).
In the mid 1960’s, Elazar Kochva compared the development of venom glands in the embryos of the Palestinian viper, the European cat snake, and the eastern garter snake. What he found was remarkable. The front-fanged viper, the rear-fanged cat snake, and the solid-toothed garter snake all had venom glands developed from the same embryonic tissue in the same basic pattern. This discovery was later verified again using modern molecular studies. Interestingly, front-fanged venomous snakes (like the Palestine Viper used in Kochva’s study) develop their fangs at the rear of the mouth with the venom gland. The fangs then move forward during embryonic development, while the venom gland stays behind or slightly under the eye.
Kate Jackson examined viper and elapid fang development using a scanning electron microscope and a series of developing fangs from embryos, and replacement fangs from adult cobras and puff adders. In both cases, she found a series of developing teeth in the tissue near the fang. These teeth were not attached to bone, though the smallest of the developing series had a discharge orifice, or venom exit. The entrance to the venom canal, the point where the venom enters the fang, was not detectable until much later in development, proving strong support for what Charles Bogert referred to as the brick chimney hypothesis. Bogert’s hypothesis suggested that fangs were not folded but grew from the base with new mineral material being added where the tooth and the bone were in close association. Using this idea, the venom canal was not necessarily formed by folding a tooth to form the venom canal of the fang. The story does not end here.
Oldrich Zahradnicek and colleagues examined the question of how the venom canal in the fang actually forms. They used newborn white-lipped pitvipers (Trimeresurus albolabris) and the sonic hedgehog gene (Shh) as a marker to study fang development. This is the same gene involved with limb suppression in snakes from Chapter 3. As we’ve already discussed, genes multitask and Shh had been previously linked with tooth formation in mice and pythons. The researchers found the inner enamel epithelium does, in fact, in-fold. The Shh gene is turned on at the tip of the fan; a wave of Shh travels up the fang triggering an increase in the number of cells in the area that will become the venom canal. Later, the cells die and leave behind a hollow canal. Fangs with a venom-conducting canal go through the same stages as other teeth, but they undergo the folding process. Enamel may never develop within the venom canal of some snakes, but it has been reported in the canal of some sea snakes and vipers― an issue that still needs to be resolved. Therefore, fangs form by a combination of the brick chimney and folding processes.
Evidence that the tubular fanged vipers were basal to many modern snakes and that elapids (cobras, coral snakes, kraits etc.) evolved relatively recently revolutionized hypotheses regarding snake evolution. Hypodermic needle-like fangs apparently evolved more than once. Freek Vonk and colleagues examined 96 snake embryos of eight species and used the sonic hedgehog gene (Shh) to mark activity. The team’s 3D imaging provides clear evidence that front fangs develop at the rear of the maxillary bone and are very similar in development to the rear fangs found in many snake lineages. This was verification of the observations made by Elazar Kochva and support for the idea that fangs, front and rear, are homologous. Vonk and co-workers found Shh was not active in the front portion of the front-fanged snakes’ maxillary, and growth and development moved the developing front fang forward. The front-fanged snakes had at least two different ancestors. Yet the data suggest a striking similarity between fang development and independent loss of the tooth-producing tissue in the front of the maxillary. They suggested the posterior tooth-forming epithelial uncoupled from the rest of the dentition, allowing the independent evolution of fangs.
Venom, the mechanisms to produce it, and the ability to deliver it to prey were major innovations in squamate evolution that allowed the advanced snakes to evolve and fill many ecological niches as small to large predators for at least the last 65 MYA, probably longer.
Vipers
Vipers have curved fangs attached to small, mobile maxillary bones that contain no other teeth. The maxillary bone forms a rotating joint with the prefrontal bone, and the viper’s snout is very flexible so that all bones can move independent of the braincase. The fangs are often paired with one functional fang and a replacement fang. Vipers can erect and fold the fangs independently of each other and, when the mouth is open, the fangs do not have to be erected. The fangs can remain folded against the roof of the mouth, covered by the soft tissue of the sheath when the snake opens its mouth; therefore, viper fangs can be longer than expected [Figure 5–2].
INSERT FIGURE 5-2.
Figure 5–2. A: A cleared and stained Terciopelo (Bothrops asper) with the maxillary (M) bone and attached fangs visible and folded back. B: An Eastern Diamondback Rattlesnake (Crotalus adamanteus) skull with the maxillary (M) and fang folded back and pterygoid teeth in the background.
Fang size closely correlates to body and head length, and as we will see to feeding behavior and diet. Carl Ernst studied the Russell’s Viper (Daboia russelii) and found fangs averaged 5.47 mm with a range of 1.5–13.0 mm in length. Similar results were found in four genera of pitvipers: Agkistrodon, Calloselasma, Deinagkistrodon and Hypnale. The longest fangs occur in the Gaboon Viper (Bitis gabonica); a 1.8 m specimen had 50 mm fangs, nine times longer than the average Russell’s viper fang.
Elazar Kochva and Carl Gans examined the gross morphology of the venom apparatus in the vipers and found it to consist of four primary parts; the main gland, the primary duct, a two- part accessory gland, and a secondary duct that carries the venom to the sheath and fang. The main venom gland is composed of branching tubules covered with an epithelium that contains several types of cells and opens into a large lumen or cavity. This cavity acts as a venom reservoir. The most common kinds of cells are columnar cells that actively produce venom molecules by protein synthesis. Venom production ceases when the lumen becomes full. When they are secreting venom the cells are tall, but when they are empty they become cube-shaped or flat, and are apparently inactive. When venom is used during a bite, or when it is manually removed by a researcher, the cells again begin to produce venom. The accessory gland is a source of neuraminic acid (a 9-carbon simple sugar), a mucus-like molecule likely to play an important role because many snakes produce them, even species that lack an accessory gland. Most of the venom components are believed to originate in the main gland.
Viper venom glands are variable in size and larger species can be expected to have larger glands than smaller species. Major parts of the venom apparatus are shown in Figure 5–3.
INSERT FIGURE 5-3.
Figure 5–3. A diagrammatic overview of the venom apparatus of a viper.
Species that feed on soft-bodied invertebrates, such as earthworms, may be expected to have smaller glands. On the other hand, at least two of the six species of night adders (genus Causus) have venom glands that extend well past the head into the body cavity. At birth, the Rhombic Night Adder (Causus rhombeatus) has venom glands extending only slightly beyond the posterior border of the skull, but adults have a huge gland extending well into the body cavity, and it is clear that this gland continues to grow long after birth.
Vipers tend to be ambush predators with heavy posterior bodies for anchoring the snake as the forebody strikes. They often strike, deliver venom under high pressure, and then release prey, trailing it until the venom has incapacitated the animal. The bite-and-release envenomation strategy likely evolved among vipers for its ability to reduce injury to the snake. However, it may be costly, in that it increases the probability of losing the prey.
Elapids
Hypodermic-like fangs that deliver venom under high pressure are also found in the elapids. The fangs of these snakes do not move to the degree seen in vipers, and the limited mobility has resulted in the fangs being described as “fixed.” Despite this description, there are several species, including the mambas and death adders, that have an articulating maxillary bone. In these species, the prefrontal sits approximately half way along the maxillary, resulting in anterior fangs which can rock back and forth. Instead of folding the fangs back into the mouth like the vipers, the elapids slide theirs into a slot in the lower jaw. Species with exceptionally long fangs may have the tips protrude just slightly through the tissue of the lower jaw. Elapid fangs tend to be much shorter in length than those found in viper and stiletto snakes of similar size [Figure 5–4].
INSERT FIGURE 5-4.
Figure 5–4. Top: A cleared and stained specimen of Spectacled Cobra (Naja naja). The relatively small fang (arrow) is visible and attached to the maxillary. Bottom: The fang and maxillary bone of the King Cobra (Ophiophagus hannah).
The lengths of elapid fangs are quite variable and mambas appear to have the longest, and perhaps the most slender, fangs. Charles Bogert reported the mean body length of the Green Mamba (Dendroaspis viridis) as 2045 mm, and the mean fang length as 6.7 mm, while Jameson’s Mamba (Dendroaspis jamesoni) has a mean body length of 2166 mm, and a mean fang length of 7.3 mm. He also found the Hamadryad (Ophiophagus hannah) to have a mean body length of 3055 mm, and a mean fang length of 8 mm. Thus, while these snakes have relatively large fangs for elapids, they are large animals. The largest elapid fangs per unit of body length were found in a Thailand population of the Monocellate Cobra (Naja kaouthia). Not included in Bogert’s work were Australian elapids, the taipans (Oxyuranus), and the death adders (Acanthophis), which may have fang lengths that are 10 mm and 7 mm respectively.
Some elapids have evolved exceptionally small fangs. Seven poorly known New Guinea species in the genus Toxicocalamus have mean fang lengths of 0.7–0.9 mm. At least one of these snakes is known to feed on earthworms, and it is likely that all seven species may have evolved shorter fangs in response to their docile, soft-bodied, thin-skinned prey. Similarly, the Arizona Coral Snake, Micrurus euryxanthus, has a mean fang length of 0.8 mm and is known to feed on small snakes. The elapid with the shortest fang is, in all likelihood, the Turtle-headed Sea Snake (Emydocephalus annulatus) with a reported fang length of only 0.15 mm. It feeds exclusively on fish eggs which require no venom and, as might be predicted, the venom apparatus has dramatically degenerated [Figure 5–5].
INSERT FIGURE 5-5.
Figure 5–5. The skull of Eydoux’s Sea Snake (Aipysurus eydouxii), a fish egg-eating specialist that has greatly reduced fangs (arrow) and venom apparatus.
The elapid maxillary bone frequently has smaller teeth behind the fang; of particular interest is that these teeth are grooved, again suggesting that tubular elapid fangs evolved from a grooved, rear-fanged ancestor quite independently from the vipers.
Fangs, like other snake teeth, are replaceable: Charles Bogert examined the skulls of more than 100 cobras of several species, and most had two full-sized fangs located on each maxillary bone. One fang was usually well anchored and functional while the other, a replacement, sat ready to move into position when the older fang was shed. On occasion both fangs may be well anchored and functional for a short time. Besides the functional fang and its immediate replacement, cobras may have as many as six or seven developing fangs on each maxillary. These are reserve fangs and sit in two close, alternating rows, with the next largest fang ready to displace the current functional fang.
Some elapids have modified the fangs for spraying venom, a behavior that is often described as “spitting.” Spitting cobras range from Africa to the Philippines and have a modified fang opening which is smaller and more circular than that found in their non-spitting relatives. Spitting venom will be discussed in more detail in Chapter 9.
The elapid venom gland is oval, covered with a fibrous sheath and, like the vipers, has an accessory gland composed of mucus-secreting cells. The elongated accessory gland surrounds the venom duct and is thought to produce molecules that activate the venom. The tubular structure is not as complexly folded as that of the vipers, and the gland’s lumen or reservoir for storing venom is relatively small. Instead, the venom is stored within the cells, in granules. Elapids often have a chewing response, giving them time to release the venom from the cells and get it into the prey. Elapid venom glands are located behind the eye and usually do not extend beyond the angle of the jaw. There is an elongated venom duct that transports the venom toward the opening at the base of the fang. But, like some night adders, the Blue Malaysian Coral Snake (Calliophis bivirgata) has an exceptionally long gland extending well past the angle of the jaw and into the body, sometimes reaching one-third to one-half of the body’s length.
Elapids tend to be slender-bodied, actively foraging snakes, and most species seem to depend upon vision as well as the vomero-nasal system (VNS) to find prey. Elapids typically crawl up to their prey, seize it, hold it in their mouth, and chew to envenomate. Once the prey is subdued, it can be swallowed. There are exceptions to this behavior. The chewing envenomation behavior has been replaced by a viper-like bite and release strategy in the taipans (Oxyuranus) and a few other species.
Stiletto Snakes
Stiletto snakes (Atractaspis) fangs are long, seamless, and anchored on a very short maxillary bone with no other teeth. Like the vipers and elapids, they inject venom under high pressure. The arrangement of mouth bones is similar to that seen in vipers and accounts for why the two groups were long thought to be related. However, stiletto snakes have some very un-viper like morphology. The maxilla forms a complex, saddle-like joint with the prefrontal which is tightly joined to the braincase by ligaments that allow only limited movement. This is related to the burrow-hunting habits of these snakes (discussed in Chapter 13). The long fangs and mobile maxilla have confused herpetologists for decades because the fangs swivel to the side, anatomy that does not conform well to any other clade of venomous snakes [Figure 5–6]. Considered vipers by some, colubrids by others, and placed in their own family by yet others, their phylogeny to other snakes remained unclear. Monique Bourgeois examined the skulls of numerous African snakes and realized that the long-fanged, front-fanged stiletto snakes (Atractaspis) were most closely related to the rear-fanged purple glossed snakes (Amblyodipsas) and the centipede-eating snakes (Aparallactus). Sam McDowell suggested the harlequin snakes (Homoroselaps) also be added to this group. These relationships were later confirmed with molecular studies, and it was demonstrated they were nested in an African radiation of snakes that includes six families recognized by Christopher Kelly and colleagues. Stiletto snakes are now placed in the African Burrowing Snake family (Atractaspididae).
INSERT FIGURE 5-6.
Figure 5–6. A: a cleared and stained specimen of the Variable Stiletto Snake (Atractaspis irregularis) with a pair of fangs (one functional, one replacement). B: The replacement fangs of the Small-scaled Stiletto Snake, (Atractaspis microlepidota), the functional fang can be seen behind and above the two replacement fangs. It is exceptionally long and its tip is not in the photo.
Stiletto snakes have also been called burrowing asps or mole vipers because of their subterranean lifestyle and long fangs. The moniker, stiletto snakes, is used here as a way to recognize their lineage as distinct and separate from the vipers. Stiletto snakes have a bad reputation in herpetological circles. The time honored method of venomous snake capture has been to secure the head so the snake does not have the ability to turn, bite, and deliver a dose of venom. Bibron’s Stiletto Snake (Atractaspis bibroni) was described by Vesey-FitzGerald as a small, shiny black snake of whose fangs he wrote,
These are extremely long and project backwards from the corner of the mouth. Therefore the snake can give ones finger an injection without even opening its mouth. The result is nasty and there are reports of fatalities...
African stiletto snakes have surprised more than one snake handler with their backward, side-stabbing. Except for the single, giant fang on the maxillary bone other teeth are usually absent, making it difficult to explain how the snake transports prey into its mouth and throat for swallowing.
INSERT FIGURE 5-7.
Figure 5–7. A Stiletto Snake, Atractaspis sp., from the Shimba Hills, Tanzania. Photo by Gernot Vogel.
The stiletto snake venom gland forms a cylinder with a central cavity and un-branched tubules radiating out from the cavity, and, they lack the distinctive accessory glands found in vipers. Exceptionally elongated venom glands occur in Atractaspis microlepidota and A. engaddensis. Both have glands that extend well beyond the skull, and may be 15–30% of the body length. The glands on the left side of these snakes are usually longer than those on the right, and they are often twisted along their length.
Rear-Fanged Snakes
Rear-fangs have a highly variable morphology. These enlarged teeth may have deep, open grooves that range from less than half the length of the fang to nearly its entire length. Some have only shallow grooves, and others have entirely smooth surfaces. Many rear-fanged species have knife-like ridges on the surface of the fang used to slice through tissue. These blade-like ridges are often found near the tip or on the bottom half of the fang. Interestingly, snakes with modified, grooved rear-fangs have a gap, an area of the maxillary bone that has no teeth. This gap is also found in front-fanged elapids just behind the fang and in front of the smaller rear teeth. The gap’s function is uncertain, but it may allow the soft sheath of tissue to cover the fangs without snagging. The Asian mud snakes (homalopsids) have grooved fangs on each side of their mouth on the back of the maxillary bone and very close to the corner of the mouth. The fangs are long and slender in most fish-eating species, but they are similar in size to the other maxillary teeth in the Tentacled Snake (Erpeton tentaculatus). The Tentacled Snake frequently swallows small fish with the strike and has little need for venom. The Crab-eating Snake (Fordonia leucobalia) has thick, robust fangs for penetrating the exoskeletons of crustaceans. The Boomslang (Dispholidus typus) and twig snakes (Thelotornis) discussed in Chapter 4 have grooved fangs on a shortened maxillary bone, which places the fangs just under the eye, in a more forward position than occurs in most other rear-fanged species.
Rear-fanged snakes usually lack compression muscles directly attached to the venom gland. Instead, the muscles that close the jaw also compress the venom gland to expel its contents. The Boomslang and its relatives are the exceptions. These species have a muscular ring surrounding the gland and connections between them so that they can pressurize the venom, adaptations that increase their danger to humans. Hognose snakes (Heterodon) discussed in the last chapter have fangs in the back of their mouth that are solid. Since hognose snakes feed mostly on toads that inflate their body in defense, some hypothesize that the fangs act as a mechanism to deflate a toad that is being swallowed.
Rear-fanged and non-fanged species with venom must hold and chew their prey to introduce the venom. In species with open-grooved fangs, venom is likely under low pressure, and the chewing motion provides multiple opportunities to expose prey to the toxin. It is also possible, however, that while the fang is embedded in the prey the tissue around the fang acts to seal the open groove. In effect, this allows venom to be pressurized. Some rear-fang species have modified their fangs and feeding behavior for specialized diets. The Asian kukri snakes (Oligodon) are specialized for feeding on reptile eggs. For this feeding behavior, the fangs are blade-like and are used to hold and cut reptile egg shells so that the snake can insert its head into the egg, drink the contents and swallow the embryo. For this feeding behavior, the kukri snake’s maxillary bone has some mobility. While holding an Oligodon behind the head, it was able to rub its teeth in a sawing-like motion on my index finger.
Venom injection was long suspected to be the function of the rear teeth, but only a quarter of the species have grooves for the transfer of venom. Enlarged rear fangs may have some other function. Kenneth Kardong proposed three mechanical advantages to rear fangs: (1) rear teeth are the first teeth to contact and engage the prey; (2) rear teeth make the longest arc of any of the teeth, and have the mechanical advantage of being able to move the prey closer to the snake’s stomach than teeth located in a more forward position on the maxillary bone and; (3) the rear teeth act as spikes that secure the prey in the snake’s mouth. This last mechanism may be particularly important if the prey is slippery, wet, or has an exceptionally smooth, hard surface. Some rear-fanged snakes kill their prey by constriction but some do not. Prey may be alive and struggling as it is being swallowed, and escape is possible if the snake does not have the prey secured. By quickly moving the prey farther into the throat, the snake imprisons the animal so that struggling is less likely to result in escape. Long rear teeth have the advantage of aiding the snake in swallowing prey. Kardong came to this conclusion by analyzing film of rear-fanged snakes feeding and swallowing prey.
Exceptions in nature are ubiquitous, and the Asiatic Mock Viper (Psammodynastes pulverulentus) is unusual because it has grooved rear fangs and front fangs with shallow grooves. This is the only snake known to have fangs in both positions on the maxillary [Figure 5–8]. It is not a true viper, but mimics them in its general appearance and behavior. Harry Greene suggested the front fangs are used to snare hard-scaled, slippery skinks (a group of lizards with a boney plate under each very smooth scale). When Kate Jackson and Thomas Fritts looked at the Asiatic mock viper’s fangs, however, they found grooves in the front fangs. The grooves were not the deep folds with rounded edges that run the length of the fang as in some rear fanged species. These teeth contained multiple shallow grooves that could allow liquids from the mouth to enter a wound, increase the ability of the fang to penetrate the prey, or facilitate pulling the teeth out of the prey. They asked the question, is the mock viper a protoelapid? That is, is the mock viper a remnant of an ancestral group that gave rise to elapid snakes? Recent DNA studies suggest that the Asiatic Mock Viper is, in fact, nested within an African radiation of snakes within the Elapoidea. Its relationships remain unclear, though one study considers Psammodynastes the sister to the African Forest Snake (Buhoma procterae). The Asiatic Mock Viper continues to be of interest because it has evolved the unusual combination of front and back fangs independently of other snakes.
INSERT FIGURE 5-8.
Figure 5–8. A: The Asiatic Mock Viper (Psammodynastes pulverulentus) from Boloven Highlands, Laos. Photo by Gernot Vogel. B: The maxillary bone from the Mock Viper (Psammodynastes pulverulentus) with both front and rear fangs. Note the front fang is solid, and the rear fang is grooved for venom delivery.
Glands, Ducts, and Fangs
Felice Fontana performed more than 6,000 experiments using more than 3,000 vipers and gave us some of the first scientific knowledge of the venom delivery system in snakes in his 1781 Treatise on the Venom of the Viper. He was also the first to provide a detailed and accurate description of the venom gland in vipers.
In a survey of snake head glands, Garth Underwood noted that glands have different positions and relationships to teeth in snakes, and that all of those investigated to date produce toxins. Some of the glands are located below and behind the eye as in the Boomslang, while others occur at the corner of the mouth. Some are very small while others are quite extensive, and may extend well back into the body. While some glands consist of a single duct connecting to a pocket surrounding a tooth, others are composed of multiple ducts.
For the last two centuries, it has been common to distinguish between venom glands in front-fanged lineages and Duvernoy’s glands in rear-fanged lineages. Recently, Brian Fry and colleagues suggested the distinction between these glands is artificial and the structures should be considered homologous. Fry and colleagues refer to both glands as venom glands, and here I follow that suggestion.
As we have discussed, snakes have a great architectural diversity in their venom apparatus. An 18th century study of snakes suggested the venom duct was directly connected to the fang, but most of the 20th century work suggests the venom duct stops just short of the fang. At the time of publication, exactly how the venom was transported from the end of the duct to the fang opening was speculative and various authors proposed hypothetical mechanisms for preventing the venom from leaking out around the base of the fang.
As we will see, species with a venom duct-fang sheath connection form tension between the thickened portions of the dental sheath and the fang, and act with ligaments from the pterygoid muscle to seal the dental sheath around the fang, preventing venom leakage. As the mouth is opened the tension created by the dental sheath and the ligaments align the venom duct and the fang opening.
Venom
Francesco Redi was born and raised in the hills of Tuscany in the early 17th century. His father was a physician, and Francesco became a pharmacist and a member of the Accademia del Cimento, a scientific organization. Redi investigated chemistry and physics, as well as biological problems using the experimental method. In June of 1663, he turned his attention to vipers and discovered, among other things, that the yellowish fluid that flows from the fangs can be fatal if placed under the skin. He also found liquid venom could be dried to a solid and then rehydrated, and that it was still deadly. Redi’s findings were not left unchallenged.
The French apothecary, Möyse Charas, insisted the fluid was saliva and the true poison was the spirit formed in the angry serpent. The motive behind Charas’s statement is unclear. Though it may have been linked to the fact he had been previously jailed as a heretic for stating that Spanish vipers were poisonous. At the time, a local archbishop had allegedly exorcised venom from all Spanish vipers, thus rendering them harmless.
Charas did his best to defend the widespread belief that biological events were controlled by spiritual forces, possibly for political motives. However, jealousy of Redi’s experimental methods may have also motivated the challenge. Redi responded to Charas’ claims with further experimentation. He enraged a viper, allowed it to expend its venom, and then bite a chicken. The bird was unharmed. He followed up this experiment by removing the venom glands from a live snake which he then allowed to bite a bird. Again, the bird lived. And, in a final series of experiments, Redi collected venom and placed it in wounds of animals. Not surprisingly, the animals died. Clearly, enraged snake spirits were not involved in killing prey for snakes. There were chemical and biological mechanisms at work.
Richard Mead was physician to King George II, but he investigated the nature of snake venom in his early years. His work included the 1702 publication A Mechanical Account of Poisons in Several Essays. Two important misbeliefs resulting from Mead’s work had long-term consequences: the idea that venom was composed of salts and that venom was an acid. Mead retracted these hypotheses in a later edition of his book, published in 1756, but the damage had been done. As so often is the case, once an idea gains popularity, it becomes difficult to replace. In 1747, it was suggested that snake poison could be neutralized with the alkali liquid ammonia, a treatment that assumes venom is acidic and can be neutralized by a base. Treating snakebites with ammonia lasted well into the 20th century and was as harmful as the venom.
Felice Fontana studied both Redi and Meads’ work and started his own experiments in 1764. After an animal died from envenomation, a detailed autopsy followed. Fontana discovered blood leaked from the vessels at some locations because it became more fluid while in other parts of the body, it coagulated. He wrote,
“Every advance I made, in this new career of experiments, presented me either with something paradoxical, or with novel and unexpected circumstances.”
Fontana is often regarded as the father of toxinology, and his thoughtful experiments did much to further the scientific study of venoms. Fontana was confused by the observation that viper venom could cause both blood clotting and excessive bleeding. It was not until the late 19th century that Cesaire-Auguste Phisalix demonstrated viper venom caused blood clotting at high doses, and excessive bleeding at low doses. In 1843 Lucien Bonaparte (Napoleon’s brother) isolated “viperine,” molecules from viper venom that acted like a digestive enzyme and discovered venoms’ protein nature, noting that it was similar to albumin. About 20 years later, American physician and scientist Silas Weir Mitchell isolated “crotaline” by boiling venom and treating it with alcohol. This lead Mitchell to an important conclusion; venom is composed of more than just one toxin.
Our understanding of venom has come a long way since Redi, Mead, and Fontana’s attempts to understand snake poisons and how they work. Today we know that, of the hundreds of molecules found in venom glands, most are proteins. And, while diverse in nature, these complex macromolelcules have been adapted from a small number of ancestral protein families. Active molecules in venom have traditionally been divided into two categories: enzymes and toxins. Enzymes tend to be relatively large molecules and are perhaps best known as the agents of food digestion, their names usually end in ― ase. Enzymes have a specific shape that allows them to bind to other molecules of a specific shape and break the chemical bonds holding the substrate molecule together. For example, cell membranes can have holes cut in them by enzymes. This allows other molecules to enter the cell or the contents of the cell to spill out. Yet, other enzymes interfere with nerve impulses, keeping the chemical message from being transferred from one nerve cell to another. Since enzymes are re-usable, a small number of them can cause considerable damage when they are turned loose in an organism. Enzymes are found in all snake venoms, but vary in type and concentration from species to species.
More than 26 enzymes have been found in snake venoms and many have relatively low toxicity. The role of enzymes in venom has been suspected to be digestion and to facilitate the spread of toxins. A discussion of a few of these enzymes follow. Hyaluronidase has been found in almost all venoms tested; it is an enzyme that breaks down connective tissue. Nuclease enzymes breakdown DNA and RNA and are known to occur in most venoms. Other enzymes work to degrade the energy transport molecule ADP, and yet others are involved in preventing blood clotting. Mucleases inhibit the repair of blood vessels. Phospholipases A2 (PLA2) are enzymes that act on fatty acids particular phospholipase; the most abundant component of cell membranes. The PLA2 enzymes are found in many snake venoms as well as insects and mammal tissues and, they often trigger inflammation and pain. Sea krait (Laticauda) venom contains a considerable amount of PLA2 while other elapids have far lesser amounts. The enzyme L-amino acid oxidase (LAAO) occurs in most pitvipers and many elapids and is a molecule that multitasks. LAAO degrades amino acids, releasing hydrogen peroxide which triggers apoptosis – cell death. LAAO also induces swelling, sometimes promoting blood clots, and at other times prevents blood from clotting. Additionally, it has anti-bacterial and anti-viral properties. Metalloproteinases are found in most all snake venoms and produce excessive bleeding, prevent platelets from aggregating to form a clot, cause necrosis in muscles, and cause inflammation. Most of these metalloproteinases have a zinc-binding site and they can be very large molecules with multiple biological activities. Arginine esterases inhibit blood clotting, kininogenase stimulates a rapid drop in blood pressure, and serine proteinases cleave fibrinogen and disrupt blood clotting.
Jay Nicholson and colleagues tested the venom of the Coastal Taipan (Oxyuranus scutellatus) for digestive activity in a simulation. They used mouse hind legs and injected them with 0.1 ml of taipan venom and simulated the acidic environment of a snake’s stomach. The mixture was sampled every 24 hours and assayed for soluble protein. Nicholson’s analysis, when compared with a control not exposed to the same enzymes, suggested the rate of digestion increased in the presence of venom. Other studies have suggested venom is not involved in digestion.
Marshall McCue tested the digestive effectiveness of Western Diamondback Rattlesnake (Crotalus atrox) venom on the energetic costs of digestion, assimilation time, gut passage time, and assimilation efficiency. Captive rattlesnakes were fed artificially envenomated mice as well as control mice injected with saline solution. The experiments showed envenomation had no significant influence on any of the digestive variables tested. Gut passage time averaged six days, regardless of the presence of venom, and assimilation efficiency averaged 79%. These results were surprising, given that Western Diamondbacks have venom with exceptionally high enzymatic and necrotic properties.
Similar results were found in a study of two Taiwanese pitvipers of the genus Trimeresurus, the Gracile Green Pitviper (T. gracilis) and Stejneger’s Green Pitviper, (T. stejnegeri), though this study also took temperature into consideration. The Gracile Green Pitviper inhabits montane habitats above 2000 m, a relatively cool environment. Stejneger’s Pitviper, on the other hand, is a lowland species using warmer habitats. Chia-Wei Chu and colleagues tested the hypothesis that venom was used to accelerate digestion, and that its role may be especially important to species in a cool climate. Experimental mice were injected with a 1.2 mg dose of venom, while the control mice were injected with saline. The Gracile Green Pitvipers from the cool climate were kept at 14ºC while the lowland Stejneger’s Pitvipers were maintained at 24º C. Gut-passage time, cost of digestion, and digestive efficiency were measured, but the results, again, did not support the hypothesis that envenomation accelerated digestion in any of the snakes tested. Given these results, venom enzymes do not appear to play a significant role in digestion of prey; however, further research may alter this view.
An alternative role for enzymes in venom is they increase the ability of toxins to spread throughout the prey’s body. Enzymes are damaging tissues, breaking open blood vessels and rupturing cell membranes, and allowing the toxins to move from the site of the bite into the blood stream and throughout the prey. It seems likely that enzymes are indeed digesting tissues before the snake swallows the prey, but the degree of digestion may be small and not readily detectable at the whole-body level.
Toxins tend to be peptides, chains of amino acids smaller than enzymes, with distinctive shapes that allow them to attach to specific receptors on cells and interfere with cellular activities. The volume of toxins in the venom has a role in determining how toxic the venom is. Smaller toxins can usually disperse from the site of the bite quickly, probably with the help of enzymes. These may stop nerves cells from communicating normally by attaching to and blocking specific receptors, constrict blood vessels, damaging blood cells, altering blood pressure, and damaging kidneys. Elapids (coral snakes, cobras, kraits, sea snakes) have venom rich in toxins, but the vipers also have these, just as elapids have enzymes in their venoms. Some toxins that have been studied follow.
Waglerins are found in Wagler’s pit viper (Tropidolaemus wagleri) venom and bind to muscle-type nicotinic acetylcholine receptors where they induce respiratory failure in mice by blocking acetylcholine at neuromuscular junctions. Crotamine-like myotoxins are known from rattlesnake venoms and bind to Na+ channels on skeletal muscles and lipid membranes; they also depolarize skeletal muscles, cause muscle damage, and induce pain. They also enter the cell nucleus. CRISP toxins are found in most all venoms and are Ca2+ channel blockers that cause paralysis of smooth muscles and stimulate hypothermia. Natriuretic peptides occur in elapid venoms and they increase guanosine monophospahte (GMP), a molecule that causes blood vessels to relax smooth muscle. Bradykinin-potentiating peptides from pit viper venoms bind to the angiotensin-converting enzyme (ACE) that lead dilation of the arteries and veins and a decrease in arterial blood pressure. Dendrotoxins from mambas (Dendroaspis) block potassium channels and cause muscle twitches, and may cause convulsions and death. C-type Lectins and lectin-like proteins are found in many venoms and bind to receptors on platelets and clotting factors. They can inhibit and activate platelets, and cause blood clotting as well as prevent blood from clotting. Safrotoxins are found only in the Stiletto snakes (Atractaspis) and bind to endothelial receptors to constrict blood vessels and cause heart failure. Disintegrins in vipers prevent blood platelets from sticking together, preventing blood from clotting. They also increase inflammation, and inhibit cell adhesion. Disintegrins also inhibit the integrins, molecules that allow cells to communicate. So where did all of these molecules in venom come from?
Antibacterial Venom
Hypotheses suggesting that venom was derived from salivary molecules were popular. After all, saliva is made in the mouth and has an enzyme or two that start food digestion in many vertebrates. Others suggested the venom gland was tissue similar to that found in the pancreas and that venom enzymes were similar to digestive enzymes produced in the pancreas. As discussed in Chapter 4, there is strong evidence that venom is older than snakes― possibly 100 MY older than snakes. Furthermore, it appears few molecules used in venom were actually derived from saliva, more of them appear to have been derived from proteins that do a variety of other miscellaneous jobs in organisms. Some of the molecules came from ancestors found in the brain, blood, and liver.
The literature on snake venom frequently reports anti-bacterial or anti-microbial properties. In fact, anti-microbial action has been reported in the venom of all major clades of snakes. John Shivik has expanded the hypothesis that venom may have its origin in molecules that neutralize the chemical defenses of bacteria and fungi that are growing on carcasses eaten by an early snake or snake ancestor. The Brown Treesnake (Boiga irregularis) is an exceptional predator, but Shivik also found it to be an enthusiastic scavenger. Like many rear-fanged species, predatory behavior in the Brown Treesnake relies on constriction leaving its potent venom unused. This apparent paradox may be the result of another function for venom molecules. Venom, or at least some of its components, may be the way snakes compete for energy and nutrients with microorganisms. Bacteria and fungi are well known for producing chemicals that make decomposing food inedible. Savengers have to deal with these chemicals. Many snakes will eat dead food, but, even if the prey is alive, bacteria are abundant in the digestive system as well as, in the blood and on the skin of live animals. Eating an entire animal means the snake is taking in large amounts of bacteria. Venom components may work to render the otherwise dangerous microorganisms harmless to the snake. Therefore, the precursors of venom molecules may have been doing dramatically different jobs in the Toxicoferean ancestor. Antibiotic-like functions make sense give that snakes eat whole animals and as we will see, may act as scavengers more often that previously thought.
Evolution Tinkers
With the ability to compare genes and proteins from species to species, it becomes possible to see evolution tinkering. Duplicating genes or sets of genes and then modifying the copies allows evolution to tinker without causing damage to the organism. It is much like making a copy of computer software and changing it without altering the original copy. You can change the program code and make it better (or worse), or give it a new function. If the results are positive, you have a new program. If the results are negative, the copy is discarded. Snakes, like all other organisms, have used gene duplication techniques to produce novel adaptations.
Most molecues in venom are proteins, and making proteins demands energy and, a source of amino acids. Therefore, venom is apparently expensive to make. Marshall McCue examined the metabolic costs associated with venom production in three North American pitvipers: the Timber Rattlesnake, the Western Diamondback Rattlesnake, and the Copperhead. Venom was collected and weighed from 33 snakes while they were anesthetized. Afterward, the metabolic rates for each snake were measured for 72 consecutive hours using oxygen consumption. The snakes replenished their venom during the 72 hours by increasing their metabolic rates about 11% compared to controls where no venom was extracted. Venom production reached a peak level within a few days after expenditure. McCue suggested the 11% is probably an underestimate of the actual cost to manufacture the venom. The cost of venom production may be the reason why snakes meter venom during a bite, a topic covered in the next chapter.
Min Li and colleagues studied the Marbled-headed Sea Snake (Aipysurus eydouxii), a fish-egg-eating specialist with reduced fangs and a degenerated venom gland. The venom it does produce is 40–100 times less toxic than that of other sea snakes. A similar situation has evolved in the Australian shovel-nosed snakes (Brachyurophis), a terrestrial elapid adapted to a diet of reptile eggs. The PLA2 enzymes in snake venom discussed above are quite toxic, but Li and colleagues found the Marbled-headed Sea Snake PLA2 genes had more mutations that produced altered PLA2 molecules, making it less toxic. Li and colleagues attributed the numerous mutations to the lack of selection pressure on the venom to improve performance.
Vincent Lynch examined the PLA2 genes from a variety of elapids, including sea snakes and vipers and identified mutational paths that lead from non-toxic to highly toxic PLA2 molecules. The results supported the idea that duplication of genes is, indeed, the mechanism used to produce the novel molecules in venom. He suggests shifts in prey preferences test the PLA2 genes through a selective sieve, genes that are effective at subduing and killing prey are born and genes that are no longer effective are not expressed or lost.
Three-fingered toxins (TFTs) have 60–74 amino acids, all of them share four specific disulfide bonds, and they show a similar pattern of folding that accounts for their name. Bryan Fry and colleagues used molecular techniques to examine TFTs in elapid snakes. Comparing amino acid sequences from about 41 different elapids, they discovered some are shared between many elapids clades while others are restricted to specific species. TFTs appear to have evolved from a set of genes responsible for making the three-fingered peptides, molecules widespread in vertebrates that function in cell adhesion, occur in the immune system, and are also found in the central nervous system. Further studies demonstrated TFTs to be older than the elapid snakes, with one recovered from the Asian ratsnake, Coelognathus radiatus. Ratsnakes are powerful constrictors and, despite the absence of fangs, the presence of a TFT in the snake’s venom gland suggests it may not be using constriction alone to kill its prey.
Three-fingered toxins come in numerous forms. Curare-like toxins may compose up to 70% of sea krait (Laticauda) venom, and 1 to 4 millionths of a gram can kill a 20 g mouse. The TFTs work by blocking receptors on skeletal muscles. When the diaphragm muscle receptors are blocked, breathing ceases. Muscarinic toxins, known from mamba venom, are short chained TFTs that bind to receptors in the brain and blood vessels interfering with heart rate, blood pressure, and cognitive functions. Fasciculins are short TFTs known only from mambas and, while they are weakly toxic they appear to cause death by respiratory paralysis. Cytotoxins, also called cardiotoxins, poison cells and breakdown red blood cells, they do this by working with an enzyme and they disrupt the cell’s water balance. Cells fill with water and bursts. Calciseptine is a mamba toxin that works on calcium channels in cell membranes. Mambin, also found in mambas, inhibits blood clotting.
Venoms of vipers are often described as haemotoxic and venoms of elapids are often described as neurotoxic. Critics of the distinction argue that this is an over simplification. However, there are two studies that suggest this classification may not be too unreasonable as long as the exceptions are acknowledged.
In a comparison of venom enzymes in true vipers (Viperinae), pitvipers (Crotalinae) and elapids (Elapidae), Marshall McCue found they did not differ statistically in PLA2 activity, but viper venoms did have proteolytic activities that were markedly higher than elapid venoms. He developed a measure for toxic risk that includes toxicity and venom yield and found the three groups of snakes do not differ significantly from each other despite their distinct evolutionary histories.
Chromatography can be used to produce patterns for specific venoms, and Rob ert Graham and colleagues used gel filtration chromatography to build patterns for 30 snake venoms. They found distinct venom profiles that supported the notion that elapid venoms were primarily neurotoxic and viper venoms were mostly haemotoxic, but exceptions were common. For example, the New Guinea Taipan (Oxyuranus scutellatus canni), an elapid, had a venom profile that looked more like a viper while the South American Rattlesnake (Crotalus durissus terrificus) had a profile that was similar to an elapid because of its neurotoxic components.
The Terciopelo (Bothrops asper) is widely distributed, ranging from northeast Mexico to Argentina, and it occurs on the continental island of Trinidad. It is a large, highly venomous snake responsible for a large percentage of the serious snakebites in the Neotropics. Widespread species show geographic variation in their anatomical, physiological, and behavioral characteristics, and venomous snakes show geographical variation in their venom. Variation in venom can be important in understanding the symptoms of envenomation, and it also provides information about the evolution of the species. Alberto Alape-Girón and co-workers examined the venom from two isolated Terciopelo populations in Costa Rica, one from the Caribbean side and one from the Pacific side. The two populations have been separated for the last 8–5 MY (late Miocene to early Pliocene) by a series of mountain ranges. Alape-Girón found that proteins in the two populations were 52% similar. One major difference between the two populations was the distinctive PLA2 molecules found in each population. Secondly, the researchers found individual variation in all of the families of proteins, but did not find any trends related to gender. Venom patterns in the two populations were distinct enough that snakes could be identified as to their geographic origin by their venom profiles. Third, the authors examined changes in venom composition between neonates and adults and found obvious shifts in composition between the two age groups. While they stated the reason for such shifts remains obscure, they wrote,
...venoms from neonate snakes are more toxic to lizards and inbred mice than adult venoms. Young Bothrops snakes preferentially eat amphibians, lizards, birds, and shift to mammals when they become adults. The qualitative and quantitative adjustments in the composition of the venom proteome linked to the development of B. asper are likely related to the survival of the snake by prey adaptation.
Additionally, they suggest that the shift in venom seen in neonates to adults may be related to digestion, and the histolytic enzymes in the neonates may work to breakdown the bolus so that the prey does not decompose in the snake’s digestive system.
Toxicity and Co-Evolutionary Arms Races
The Coastal Taipan (Oxyuranus s. scutellatus) has highly toxic venom; 120 mg is said to be able to kill 12,000 adult guinea pigs. The Inland Taipan (Oxyuranus microlepidotus) is often cited as the most toxic snake, and said to carry enough venom to kill 100 humans, or 250,000 mice, in its venom glands. Why would a snake evolve such highly toxic venom?
Toxicity of snake venom has long been measured using 20 gram white mice in LD50 tests (lethal dose, 50%). The LD50 is the dose of a toxin that kills 50% of the test animals. Toxicity varies greatly with the route of delivery and venom can be delivered under the skin (subcutaneously), into a muscle (intramuscular), intraperitoneal (into the body cavity), intravenous (into a vein), or intracerebro-ventricular (into the brain). The mouse LD50 for Eastern Diamondback Rattlesnake (Crotalus adamanteus) venom injected into a vein is 34 micrograms (mcg), while the same venom injected under the skin has to be 274 mcg for an LD50. Eastern Diamondback Rattlesnake venom is eight times more toxic when it is delivered directly into the blood stream than when it is placed under the skin. However, the mouse LD50 of Jameson’s Mamba (Dendroaspis jamesoni) venom is 8.4 mcg regardless of whether it is placed under the skin or directly into a blood vessel. The intravenous injection route produces the most consistent results in LD50 experiments. The route of venom delivery has a dramatic impact on the test results and suggests another reason why there is great variation in the outcomes of snakebites. The mouse LD50 has little practical application. Although it might alert a person to the potential danger of a given species of snake, it will contribute nothing to saving a person’s life.
Returning to the question, why is it that some snakes have evolved such toxic venom? Snakes did not evolve to eat humans or lab mice. The Inland Taipan evolved to hunt prey with the capacity to retaliate with a bite using their sharp teeth, including the Plague Rat (Rattus villosissimus), and small marsupials, including the bandicoots (Peramelemorpha) and the Kultarr (Antechinomys laniger). It should be noted that the Plague Rat, like the marsupials, is an Australian endemic rodent that has shared a long evolutionary history with the taipan. Taipans have a rapid bite-and-release strategy, quickly biting and releasing the prey, and biting it again and again if necessary to immobilize the mammal. Afterwards, the the snake trails the prey until it is unable to resist. Co-evolution of the Plague Rat and Inland Taipan is likely responsible for the highly toxic venom. As rats and other mammalian prey became resistant to the venom (those that survived the bite reproduce leaving resistant offspring) the individual snakes with the most effective venom became the most successful hunters. Predator-prey arms races are responsible for the highly toxic venoms of snakes. Therefore, the high or low toxicity of snake venom to humans is purely coincidental.
Moray eels (Gymnothorax) are associated with coral reefs in oceans around the world. These carnivorous eels hide in crevices waiting for prey to approach. In the Pacific, moray eels are hunted by the Yellow-lipped Sea Krait (Laticauda colubrina). The snakes swim along the reef’s edge and explore crevices to locate hiding morays. Harold Heatwole and Naomie Poran found New Guinea morays were remarkably resistant to Yellow-lipped Sea Krait venom, being able to tolerate massive doses ranging from 42.5 and 75 mg/kg. Garden eels (Heteroconger cobra) and freshwater eels (Anguillidae) not preyed upon by Yellow-lipped Sea Krait were not resistant to the snake’s venom. In a follow-up experiment, Heatwole and Powell tested moray eels from the Caribbean to see if the eels were also resistant to Yellow-lipped Sea Krait. Sea kraits don’t live in the Caribbean and the eels were found to be very sensitive to the snake venom, responding fatally to doses of 0.1 mg/kg. These experiments demonstrate that predator and prey co-evolve. The snakes and their eel prey are locked in an arms race. As the eels evolve more defenses against the venom, the snakes evolve more potent venom.
Jenny Daltry and colleagues collected venom samples from the Malayan Pitviper, (Calloselasma rhodostoma) at 36 different locations in Southeast Asia. The venom showed considerable variation over the snake’s Southeast Asian distribution. In an attempt to explain this variation, they tested three hypotheses and found support for only one― venom varied with diet. Malayan Pitvipers are dietary generalists, feeding on frogs, reptiles, birds and mammals but in different proportions at different geographic localities. When venom similarities were tested, geographic proximity and phylogenetic proximity did not correlate to venom variation. However, diet did.
California Ground Squirrels (Spermophilus beecheyi) have evolved to resist the venom of rattlesnakes. James Biardi and colleagues used blood plasma from four California Ground Squirrel populations to test the rodent’s resistance to rattlesnake venom. At one location where Northern Pacific Rattlesnakes (Crotalus oreganus) were common, the ground squirrels were able to resist the rattlesnake’s metalloprotease, a venom enzyme that breaks down the lining of blood vessels. Squirrels from other localities seemed able to withstand the impact of the snake venom’s to burst red blood cells. Again, evidence suggests prey and predator are in the midst of a co-evolutionary arms race, where the prey becomes resistant and the snake becomes more toxic to overcome the resistance at the molecular level.
Yet another example of venom adapted to prey has been documented in the saw-scaled vipers (Echis). About nine species of saw-scaled vipers inhabit Africa, the Middle East, and Asia. Three of the known species groups feed primarily on scorpions and centipedes, hunting vertebrates far less frequently than the fourth group which specializes on the larger prey. Venom protein profiles did not show a correlation between venom and diet, but the toxicity of the venom to scorpions was telling. The mean scorpion LD50 of saw-scaled viper venom was much lower in the scorpion-eating species at only 10–40 mcg per gram of body weight, while the vertebrate-eating species had a scorpion LD50 of 140 mcg per gram of body weight. Comparing the mouse LD50 for these same species, the values were all between 0.7 and 1.25 mcg per gram of body weight and the differences were not a matter of relative lethality of the venoms―some saw-scaled vipers have specialized their venom for killing scorpions.
The rattlesnakes of the genus Sistrurus tend to feed on mice, lizards, and frogs, with the Eastern Massasaugua (S. c. catenatus) and Western Massasaugua (S. c. tergeminus) specializing in mice. The Pygmy Rattlesnake (S. barbouri) feeds on lizards and frogs. H. Lisle Gibbs and Stephan Mackessy found Sistrurus venom highly toxic to mammals, but the venom toxicity to other prey had not evolved significantly. Of course, lizards and frogs can do little mechanical damage to a rattlesnake that is trying to eat it.
Using venom from15 species of coral snakes (Micrurus), Nelson Jore de Siliva and Steven Aird compared the effects of venom in lab mice and natural prey. They found coral snake venoms to be significantly more toxic to their natural prey than to lab mice, except in one case where the natural prey was unknown.
Not surprisingly, snakes specializing in insects have venom that is more effective against the arthropods. Vladislav Starkov and colleagues demonstrated snakes with insects in their diet have venom that is more toxic to crickets than other prey. The insect-eating North Caucasus viper (Vipera lotievi) had the most effective venom against crickets.
Randy Powell and Carl Leib proposed larger prey species like woodrats (Neotoma) and ground squirrels (Spermophilus) would benefit the most from venom resistance. Smaller species would have little use for the development of venom resistance because they may be seized and swallowed with little opportunity to resist or escape. Selection for venom resistance may only benefit larger prey species capable of fighting back or by snakes with a bite-and-release envenomation strategy. Resistant animals that are bitten and released stand a chance of escape and survival. These authors also note that the American opossum’s (Didelphis marsupialis) resistance to venom is puzzling because it is an unlikely prey species as a result of its size and sharp teeth. They hypothesize that it may have inherited the resistance from an ancestor with a distinctly different natural history. However, the metalloproteinase inhibitor found in the opossum Didelphis (marsupialis) aurita has been found in the blood of newborns as well as the milk of the mother. It seems likely the Jararaca (Bothrops jararaca), is feeding on the young marsupials rather than the adults.
Molecules that make prey immune to elements in snake venom have been found in possums of the genera Didelphis, Lutreolina, and Philander; the Hedgehog (Erinaceus eurapaeus); the Wood Rat (Neotoma micropus); the Cotton Rat (Sigmodon hispidus); the Mexican Ground Squirrel (Spermophilus mexicanus); the Grey Mongoose (Herpestes edwardsii); and numerous venomous and non-venomous snake species. Ana Gisele C. Neves-Ferreira and colleagues have summarized the literature on these molecules and reported metalloproteinase inhibitors, and PLA2 inhibitors that occur as antimytotoxics and antineurotics. The presence of these molecules further strengthens the idea of the co-evolutionary arms race between predators and prey.
The Extratoxic Phenomenon
A 1999 Southern Pacific Rattlesnake (Crotalus helleri) bite in San Bernardino County, California, resulted in neurotoxic symptoms including double vision, facial twitches, and an inability to swallow or talk. Symptoms also included the usual hemorrhaging and swelling associated with the venom of this species. The victim recovered with the administration of 35 vials of antivenom. The case was the first instance of neurotoxicity associated with the Southern Pacific Rattlesnake envenomation. Many Arizona and California populations of Mojave Rattlesnake (Crotalus scutulatus) were found to contain neurotoxins in the late 1970’s and speculation that rattlesnake venoms were becoming super-toxic or extra-toxic because of the addition of neurotoxins spread through the media. The proposed mechanism for this was the highly improbable idea that various rattlesnake species were hybridizing with the neurotoxic Mojave rattlesnakes and the neurotoxic gene, or genes, were spreading through the country, coast to coast.
Wendy French and colleagues examined 25 Southern Pacific Rattlesnakes for Mojave toxin (the neurotoxin) from the Mt. San Jacinto area of Riverside County, California. All of the snakes examined had the neurotoxins. Because this population was isolated from any Mojave Rattlesnake populations, the possibility of hybridization could be eliminated. The Mojave Rattlesnake had several forms of the neurotoxin while the Southern Pacific Rattlesnake had only one type of neurotoxin. However, like the Mojave Rattlesnake, they found the Southern Pacific Rattlesnake showed venom with three patterns; (A) venom with neurotoxin and no protease activity, (B) venom with protease activity and no neurotoxin, and (A+B) venom with both neurotoxin and protease activity. Interestingly, snakes with the neurotoxin were not found in San Bernardino County where the 1999 bite occurred.
Snake venom variability is often related to geography, and populations in one locality have venom distinct from a nearby population of the same species. The variability of venoms may be the result of snake adaptations to prey and adaptations by prey to snakes. Complexity increases by prey species adapting differently to venom from the same predator at different geographical locations. But snake venom can also vary in composition from young snakes to older snakes. This variation in venom has led to the suggestion that human predation on adult snakes has selected rattlesnakes for retaining their juvenile venom composition into adulthood. Whether this could account for the variation in venom remains untested. However, the A, B and A+B venom types also suggest this may be a normal polymorphism in snake venom. Individuals are born in any given population with one of these combinations. Depending on the local prey, individuals with one venom type will be more successful predators than individuals with another type. Populations tend to express one of the venom morphs more frequently than the others because of the advantages it confers on the individual snake.
Self Protection and Venom Storage
While working on the Trinidad herpetofauna, I collected two Rouleas (Pseudoboa neuwiedii), a rear-fanged snake known to have venom capable of killing a domesticated cat (Felis catus). One snake struck and bit the second. Within a few hours the bitten snake was dead. The observation that a snake could kill another snake of the same species with its venom was not a new one. Before we procede with this discussion, a word of caution; it is possible that some of these deaths could have resulted from mechanical damage done by the fang. While this seems an unlikely explanation for all of them, it may explain some.
F. W. FitzSimons experimented with venoms in an effort to document their effect on snakes of the same species and different species. He observed one Puff Adder (Bitis arietans) biting another. A postmortem dissection showed that the snake had hemorrhaged into the tissues and the body cavity. The Puff Adder had been carrying embryos which were also killed by the venom. FitzSimons collected six drops of Puff Adder venom and injected it into the snake from which it was collected. The snake died five hours later from its own venom. Five drops of Cape Cobra (Naja flava) venom were injected subcutaneously into a mamba, which died in 45 minutes. Day-old Puff Adders were fed to a carp (Cyprinidae). The fish died an hour later, and the postmortem showed a large hemolytic patch on the lining of the stomach. Mongooses were also tested with the venom from the cape cobra, the mammals died within 16–23 minutes.
Snake suicide? Probably not. It seems likely that the cobra detected the odor of prey on its own body and bit itself.
Being venomous poses problems. First, how do snakes keep from poisoning themselves via an accidental introduction of venom? Secondly, if venoms contain enzymes which tend to be unstable molecules, how do snakes store them for weeks or months at a time without having them degrade?
Snakes have multiple mechanisms for protecting themselves from their own venom. They have antibodies in their blood to neutralize venom molecules, and it is likely that the receptors the venom toxins bind to may have a modified shape in the snake so that the toxin cannot bind to the receptors on the snake’s own cells. Moreover, snakes have the same generalized protections from alien enzymes that other animals have, alpha2-macroglobulins. These are large molecules that protect against enzymes that breakdown protein. All of these mechanisms reduce the chances that a snake will poison itself, but are by no means 100% effective.
Zoltan Takacs and colleagues have shown that the Egyptian Cobra (Naja haje) has slightly modified its nicotinic acetylcholine receptor so the alpha-neurotoxin will not bind to it. Therefore, the snake is protected from at least some elements of its own venom and the venom of other species that use this toxin. However, snakes have clearly not protected themselves from all elements of their own venom, given the number of reports of fatalities linked to self-bites from their own venom, or the venom of other individuals of the same species.
Snakes can go for weeks or months without eating and, therefore, need to store venom long-term so it will be instantly available and ready to use when prey is located. Enzymes are notoriously unstable. High temperatures cause them to unfold and lose their shape, making them useless. Light waves that penetrate the tissues and enter the gland may also degrade venom enzymes. Additionally, one enzyme may breakdown another, and they readily combine with heavy metals that jam their active sites, again, rendering them useless.
Venom glands in all groups of snakes contain melanin, the pigment that gives color to the iris of the eye and skin. Interestingly, of the species studied, the small mountain-dwelling Ridge-nosed Rattlesnake (Crotalus willardi) had the most pigmented venom gland of the snakes examined. The pigment probably has the effect of absorbing harmful light waves that may damage the venom. This little rattlesnake has the dense pigment because its high elevation habitat receives considerable amounts of ultraviolet radiation.
But, long-term storage of enzymes cannot just be solved by shutting out the light. How are enzymes kept from degrading each other and the glands that contains them? Stephen Mackessy and Louise Baxter studied the storage problem in rattlesnakes using electron microscopy to examine venom gland and accessory gland morphology as well as the structure of the cells that compose them. They found a folded epithelium in the main gland area at the rear of the venom apparatus. The cells were mostly protein-secreting cells, but about 2% of the cells were rich in mitochondria and changed shape and height during the venom production cycle like the protein producing cells. The mitochondria-rich cells resembled the cells found in the lining of the mammalian stomach, and the authors hypothesize that they produce acid, like the stomach cells. Snakes appear to use citrate to lower the venom pH to 5.4, a pH low enough to inhibit enzyme action but not low enough to denature the enzymes. When the venom reaches the prey’s tissue, it encounters an environment with a pH of 7.2 to 7.4, something much closer to the neutral pH of 7.0 at which most enzymes work best. In addition to the lowered pH, other molecules in the venom gland act to stabilize and inhibit enzyme activity. It appears that snakes have co-opted some enzymes and enzyme-inhibition mechanisms for their venom and venom injection system from genes normally used in the digestive system. This, again, demonstrates how evolution tinkers with what is available in order to produce new innovations that contribute to the survival and diversity of organisms.
Constriction
Venom, and the apparatus used to deliver it, are an evolutionary innovation of snakes that have allowed them to diversify and fill many different niches. However, there is a behavior that preceded front-fangs which made snakes highly effective predators, constriction. Wrapping the prey in loops of its body allows the snake to hold and kill with unexpected efficiency. Constriction has been reported in almost all snake families. The exceptions are the vipers and the shield-tailed snakes, and it seems likely that these lineages had ancestors that were constrictors. Despite its effectiveness, constriction has been replaced by other specialized feeding habits in some lineages. Venomous snakes that strike and release prey have no need for constriction (vipers and some elapids), nor do snakes that feed on earthworms, small arthropods, or are specialist egg predators. Although, there are of course exceptions, such as the mussarana of the genus Clelia, that has potent venom and uses constriction.
Constriction requires the snake to get close to the prey and hold it while its body is wrapped around the animal. This can pose risks to the snake because prey with sharp teeth may bite, and prey with sharp claws or spines may damage the snake’s skin or digestive system. But snakes are adept in prey handling and injury is usually avoided.
Frank Wall may have been the first to propose that death by constriction results from circulatory arrest. Constricting snakes are able to exert enough force on an animal’s body to prevent the return of venous blood to the heart. Others had proposed death from constriction resulted from suffocation, but death by suffocation takes longer and may increase the chances the prey will injure the snake. Circulatory arrest that results in rapid heart failure or spinal fracture kill the prey more quickly and probably reduce the risk of injury to the snake. The amount of force exerted by constricting snakes was measured in 12 species by Brad Moon and Rita Metha. They found constricting pressures between 5–175 kPa (kilopascals) in snakes that had trunk diameters with constriction regions that ranged between 0.85 and 12.5 cm. [Note: a kilopascal is a measure of force, and is equal to about 1% of atmopsheric pressure at sea level.]
Not surprisingly, larger snakes exerted more force than smaller snakes and were found to exert a force directly related to the diameter of the trunk and the number of loops. Moon and Metha proposed that snakes that exert a high amount of force probably kill prey by spinal fracture and by dislocation of vertebral joints. Snakes that exert more moderate forces may kill prey using circulatory arrest, and small snakes that exert less force may kill prey with suffocation. Of interest is that the snake doing the constriction may also cause blood flow reduction in its own muscles while it is constricting prey. Snakes have low arterial blood pressure but are tolerant of anoxia, and can use anaerobic metabolism during constriction, so the anoxia does not result in adverse effects. The authors predict that large constricting boas and pythons may be able to exert forces that approach 900kPa.
Snakes use their epaxial muscles (muscles that are dorsal to the transverse processes of the vertebrae) for constriction, and Olivier Lourdais and colleagues examined these muscles in the Columbian Rainbow Boa (Epicrates maurus) using MRI (magnetic resonance imaging) and caliper measurements. They measured the epaxial musculature in snakes, their performance in escaping a predator, and their ability to handle prey. Snakes were fed after a fasting period and had their musculature and performances measured. Lourdais and colleagues found epaxial muscles were a determinant of strength intensity. Snakes were fed for six months after the fast and showed an increase in width of the epaxial muscles along the length of the body, but muscle gain was greatest in the posterior body. During a fast this region lost the most protein to fuel metabolism.
Venom and constriction have been responsible for much of the success snakes have had as predators. These two predatory mechanisms are not mutually exclusive, however. Many snakes employ both and the reasons for this remain unclear, though unraveling them promises an interesting new insight into the natural history of snakes. With the discovery of Toxicofera comes the implication that venom evolved before constriction. However, constriction is used by some of the oldest living snake lineages (Tropidophiidae, Bolyeriidae, Acrochordidae) and all of these lack an efficient method of delivering venom. In fact, based on living snake lineages, it appears snakes did not evolve an efficient venom delivery system until the origin of the vipers, about 54.4 MYA. This is despite the fact that snakes have been around for 160 MYA, based upon the DNA clock. In other words, snakes evolved the genes for venom production 100 MY before they ever used them to make venom molecules. This suggests to me that the ancestral venom molecules had a dramatically different purpose in the early snakes than they do today.
Comments
Post a Comment