Snakebite envenoming is a significant global health concern and has been recently included in the World Health Organization (WHO) list of Neglected Tropical Diseases [1
]. Over 5 million people are affected every year by snakebites, and over 75% of the estimated 138,000 deaths occur outside of the hospital setting, where antivenom cannot be administered [2
]. In total, there are an estimated 5.8 billion people worldwide living within the geographic range of venomous snakes. A disproportionate number of impoverished populations reside in hot-spots where there is a high risk of receiving a snakebite and are remote from hospital resources and antivenom [3
]. Worldwide, snakebite envenoming is an occupational risk to the majority of people working in manually dominated agricultural settings [5
]. There is a clear, unmet need for effective, heat stable, and economical broad-spectrum snakebite envenoming therapies based on drugs with a high volume of distribution that can be given rapidly and at the time of bite as well as an adjunct therapy to improve the efficacy of antivenoms [7
All snake venoms contain multiple enzymatic and non-enzymatic components, including abundant phospholipases A2
(PLA2), which are major contributors to venom toxicity of the majority of the world’s venomous snakes [9
]. Snake venom secretory sPLA2 and sPLA2-like proteins play a fundamental role in neuromuscular paralysis, coagulopathy, cardiotoxicity, renal toxicity, and skeletal muscle necrosis that can result in rapid death or permanent disability [11
]. Furthermore, there is an increasing body of evidence that mammalian homologs of venom sPLA2 also play a major contributory role in the endogenous inflammatory response, which might be involved in the response to snake venoms, including hemolysis, platelet dysfunction, and possibly fibrinogen depletion [14
]. It is possible that the combined effects of venom and endogenous PLA2s has a critical role in the toxic inflammatory response, morbidity, and mortality resulting from snakebite envenoming.
Because of the central role played by snake venom sPLA2s, they are a logical target for therapeutics with active and peripheral site targets not accessible to antivenom therapies, owing to pharmacokinetic considerations and to the generally low antigenicity of PLA2s [19
]. As a class of therapeutics, some limitations inherent to antivenoms (e.g., poor tissue penetration with a dependence on contact with circulating venom) are compounded by their perishable nature and the development of adverse reactions in a percentage of patients; thus, antivenoms have to be administered in health facilities by medical personnel [21
By focusing on the enzymatic activity of sPLA2s, rather than their antigenicity, we have identified what could be a safe, potent set of compounds that can be readily re-purposed to timely address the cardinal life-threatening complications of snakebite envenoming in the field. LY315920 was first developed as a potent inhibitor of human sPLA2 for indications such as pancreatitis, sepsis, and cardiovascular disease [22
]. Surprisingly, LY315920 and closely related compounds such as the orally-available pro-drug LY333013, are extremely potent inhibitors of 28 medically important venom sPLA2s from snake species of six continents [7
]. These results have been replicated with several venoms both in vitro and in vivo in mice [7
]. Both lead compounds have been through extensive human and animal testing, and although they have excellent safety profiles for short-term use, have never gained FDA approval and are now off patent [22
]. Recently, we demonstrated that LY333013 given orally could rescue mice with treatment delays at time points past which specific Taipan (Oxyuranus scutellatus
) antivenoms are effective [26
]. LY315920 and its pro-drug LY333013, fit the profile of a low-cost, heat stable, easy to administer field antidote for snakebite envenoming, particularly for venoms whose main toxic actions are induced by PLA2s.
Herein, we show results from in vivo rescue studies with LY315920 and its pro-drug LY333013 in juvenile porcine models from lethal envenoming of M. fulvius
venom. In animals, M. fulvius
venom is known to cause neuromuscular paralysis, intravascular hemolysis, and myonecrosis [27
]. M. fulvius
venom was chosen for porcine testing because of its clinical relevance in North America, the fact that there is a scarcity of available antivenoms, and because its main toxic activities depend on the action of PLA2s [27
]. In addition, M. fulvius
venom proved to have high reliability in preliminary porcine lethality tests, reducing the likelihood of unnecessary animal use [30
Continuous infusions or repeated bolus doses of LY315920 and oral LY333013 resulted in 100% survival following severe, experimental, M. fulvius envenoming in pigs. Rescue experiments with a single-dose of LY315920 produced a significant improvement in survival times compared to controls, but after 10+ h from the time of envenoming the symptoms often reappeared, probably due to a delayed venom absorption from a depot in the tissue site where venom was injected. In addition, this suggests that the half-life of the experimental drugs will make single dose therapy for severe envenoming unlikely to result in sustained clinical resolution. However, the ability to treat, rescue, and re-rescue animals at late stages of envenoming, including those already exhibiting severe neurological deficits and clotting disturbances, suggests these drugs might have unexpected versatility in a variety of clinical settings for both human and veterinary applications in M. fulvius envenoming.
We found that drug rescue where antivenom was initially unsuccessful resulted in a lowered drug requirement—i.e., fewer pro re nata (PRN) doses to maintain normal neurological function to the end of the study. This suggests that the circulating antivenom was able to capture drug-neutralized venom diffusing back into circulation from the peripheral tissues. However, confirmation of this requires more detailed studies.
PLA2s are predominant in M. fulvius
venom and are largely responsible for its neurotoxicity [27
]. Hence, in the dichotomic pattern described for Micrurus
venoms from the proteomic standpoint, i.e., PLA2-rich and alpha-neurotoxin-rich venoms [40
], M. fulvius
fits within the first group. In addition, besides neurotoxicity (the primary toxicity in humans) other toxic activities associated with this venom are witnessed in different animal species (e.g., myotoxicity and intravascular hemolysis) and are also caused by the action of PLA2s [37
]. This largely explains the success of these potent PLA2 inhibitors in abrogating lethality, myonecrosis, and hemolysis in our experiments in pigs. Moreover, PLA2s are known to affect coagulation and platelet function [13
], hence explaining the ability of the inhibitors to abrogate these toxic effects as well. Our observations with edrophonium and atropine suggest that post-synaptically acting alpha-neurotoxins are not likely to play a significant role in this venom’s neurotoxicity, therefore supporting the concept that pre-synaptically acting PLA2s are the main neurotoxic components in M. fulvius
venom. The key role played by PLA2s in the overall toxicity of M. fulvius
venom makes it highly amenable to treatment with the PLA2 inhibitors tested in this study, a finding of potential impact in light of the current scarcity of coral snake antivenoms in the US.
Snakebite is an ancient scourge recognized in mythology and history since Biblical times but only recently as a neglected tropical disease by the WHO [1
]. Our findings could represent a first crucial step away from the complete dependence on antivenom in combating the deadly effects of snakebite—especially, with the potential for developing an oral antidote to be taken at the time of a snake’s bite. Use of heat-stable, orally bioavailable, economical small molecules for initial and adjunct treatment for snakebite has the potential to substantially reduce mortality caused by snakebite in rural areas of the developing world. Additionally, this treatment could also reduce the mortality and morbidity associated with settings in which snakebite is an occupational hazard [5
]. Further research is warranted on developing inhibitors to other enzymatic components of snake venom, such as metalloproteases [43
]. Repositioning, the strategy utilized to repurpose LY315920 and LY333013 from their initial intended uses, offers the potential for low costs of development for future inhibitors. Combinations of multiple inhibitors and other antibodies could provide quite effective, safe, and affordable treatment for snakebite envenoming.
The inhibition of enzymatic activity of sPLA2s and other venom components by small molecule inhibitors of specific toxins could address many, but not all, limitations of antivenom. Notably, small molecule antidotes could potentially reduce the volume of antivenom administered, thus increasing the efficacy of antivenoms having a low potency against some venoms [8
] while reducing the cost to the patient and to the healthcare system [45
]. Recent US coral snake guidelines recommend availability of an antidote within one hour—the results of this paper suggest that LY315920 and LY333013 could fulfill these criteria. For example, neurologically intact patients could take the oral antidote while compromised patients could initially be treated with an IV formulation and transition to an oral therapy on recovery. In the US, there is currently no production or availability of previously approved coral snake antivenoms and these bites, while unusual, can be life-threatening, especially to children [46
]. The 3-substituted indoles, such as LY315920 and/or LY333013, could offer a plausible, definitive treatment program for M. fulvius
In conclusion, our findings suggest that LY315920 and/or LY333013 are likely to be effective in abrogating the main clinical manifestations in envenoming by M. fulvius
and could apply to other types of envenoming as well [7
]. Moreover, they support the concept that these drugs may potentiate the therapeutic action of coral snake antivenoms and others [26
]. Owing to the current scarcity of Micrurus
antivenoms in the US, the possibility of using these drugs in the management of these envenomings should lead to the design of clinical trials to assess their efficacy.
5. Materials and Methods Detail
Venom: The venom of M. fulvius was purchased from Medtoxin Venom Laboratories (Deland, FL, USA) (Lot 010918). Venom stock solutions (10 mg/mL) were prepared fresh daily in saline solution, and dilutions were performed in order to reach the desired venom dilutions to be injected.
Antidotes and Excipients
: LY315920 HCl (CAS 172732-42-5; ChemieTek, Indianapolis, Indiana, >99.9% purity by NMR, MS and HPLC) or an LY333013 (CAS 172733-08-3; provided by Ophirex, Inc. of Corte Madera, CA, USA) >99.9% pure by NMR, MS and HPLC) emulsion mixed into Greenies Pill Pockets (Mars Petcare, McLean, VA, USA) were prepared daily for each study and each subsequent day of the study. Briefly, LY315920 was weighed as a powder and mixed w
in 1:1 sodium citrate and 1:2 mannitol followed by dissolution in 8.4% sodium bicarbonate to the final desired concentration of 5mg/mL LY315920. Particulates were filtered using 0.22 µm filters. Control animals received the intravenous (IV) sodium citrate/mannitol solutions without LY315290, and excipients were administered exactly as were the IV drugs. Boluses were administered manually over 15 min and continuous rate infusions (Vet/IV 2.2 Infusion Pump (Heska), Lubland, CO, USA) at 0.67 mg/kg/hour according to the methods for previous Phase II human clinical trials for sepsis [50
]. The oral drug (LY333013) was prepared as a slurry as previously described [26
]. Briefly, the drug was mixed with 8% w
gum Arabic and then mixed into Greenies Pill Pockets. In one instance, the study animal became too weak to swallow on its own and the drug was administered via the orogastic route in gum Arabic alone. Study drugs were prepared by the study sponsor for intravenous administration and by sponsor or Oklahoma State University (OSU) veterinarians for oral administration studies when the sponsor was not present.
: The monospecific Coral-ICP Antivenom produced at Instituto Clodomiro Picado (University of Costa Rica; batch number 5610615ACL) was used. It is prepared from the plasma of horses immunized with the venom of the coral snake, M. nigrocinctus
. It is made of whole IgG molecules purified by caprylic acid precipitation [51
]. The Median Effective Dose (ED50
) of this antivenom against the venom of M. nigrocinctus
is 0.4 mg venom/mL antivenom. This antivenom neutralizes the lethal and myotoxic activities of the venom of M. fulvius
Animals Detail: Swine were housed in approved housing under a 12:12 light–dark cycle with ad libitum water and feed with the exception of an 8 h fast prior to anesthesia induction. All pigs were identified by ear tags. Pigs were acclimated to their new environment for 4–7 days prior to beginning the study. During this period, all animals were visually examined by trained laboratory personnel twice daily for any signs of illness. None of the pigs exhibited signs of illness prior to beginning the project. Pigs were weighed on the morning of the study and given a complete physical examination by Oklahoma State University animal husbandry staff (Control 12.72 ± 2.48 kg, treated 14.82 ± 2.28 kg (p-value: 0.18)). No animals were dropped from the study in either the control or experimental groups. All animals were fasted for a minimum of 8 h prior to anesthesia induction.
Housing: Swine were housed in pens (14 square feet) with raised plastic flooring to minimize contact with urine and feces. The pens allowed for individual or group housing depending on the phase of the study. During the acclimation period, pigs were group housed. Once intravenous catheters were placed, the pigs were individually housed to avoid the destruction of the catheters. Fresh water and feed were available at all times to the pigs during the studies with the exception of an 8-h fast prior to the day of catheter placement. Daily care of the pigs was provided by the centralized animal care unit, Animal Resources, as part of the Center for Veterinary Health Sciences’ animal care program accredited by AAALAC International. Environment: Swine were housed in a temperature-controlled environment with a 12/12 light–dark cycle. Animals were identified by ear tags provided by the USDA.
Anesthesia, Instrumentation, and Monitoring: Pigs were anesthetized using the following protocol: midazolam (5 mg/mL) was administered intramuscularly using a 20-gauge, 1-inch needle in the semimembranosus/semitendinosus region at a dose of 0.25 mg/kg. Pigs were allowed to rest quietly in their pen and become sedated for 15 min. Pigs were then anesthetized utilizing inhalant isoflurane given via an anesthetic mask. Once pigs were anesthetized they were endotracheally intubated and placed on maintenance isoflurane gas. Pigs were instrumented with a temperature probe and continuous cardiac rhythm monitoring (ECG). Heart rate, respiratory rate, and temperature were monitored every 15 min to measure anesthetic depth. Once the pigs were at an acceptable depth of anesthesia, intravenous and intra-arterial catheters were placed according to the following protocols.
Intravenous catheter placement
: With pigs in dorsal recumbency, the front legs were retracted caudally until they were close to parallel with and secured to the surgery table. A triangle was visualized utilizing the caudal ramus of the mandible, the lateral manubrium and medial portion of the point of the shoulder and was sterilely prepared. Initially, catheters were blindly placed utilizing these landmarks and based on the procedure described by Fluornoy and Mani [53
]. However, ultrasound guidance proved to increase the efficiency and safety of the procedure and so was employed. Briefly, a 18–5 mHz linear probe was placed inside a sterile glove to enable the use of ultrasound while maintaining sterility and the jugular vein was visualized. A 5 fr 13 cm double lumen central line catheter made by Arrow was inserted using standard Seldinger technique with the guide needle for the over the wire catheter was placed under ultrasound guidance. The catheter was then placed according to standard procedure over the wire catheter placement technique. Catheters were secured in place using 0 PDS suture and the catheters were wrapped with 4-inch Elastikon bandages.
Intra-arterial catheter placement: Arterial catheters (20 g, 1.88 in) were placed in the femoral artery and secured using 0 PDS suture, super glue, and 2-inch Elastikon bandages. Invasive blood pressure was measured using the arterial catheter every 15 min during anesthesia.
Post-envenoming analgesia: Analgesia (Buprenorphine 0.05 mg/kg IM) was given to all pigs prior to venom administration, at 4 h post envenoming, and then as needed every 4 h for a behavior indicating pain as judged by the attending veterinarian (e.g., limping or guarding of envenomed forelimb independent of overall neurological examination—e.g., generalized weakness clinically indicative of systemic neurotoxicity). For notation purposes, the assessment of pain was modified from the Obel laminitis grading system and the AAEP Lameness Scale.
Euthanasia: Pigs were euthanized or given rescue treatment if they reached a clinical score of 5 for two consecutive evaluations or if they reached a clinical score of 6 at any one evaluation. If they did not respond to rescue treatment they were humanely euthanized. All pigs were submitted for postmortem evaluation.
Study Period Detail: The pigs were monitored throughout the study period as follows:
Pigs were given a clinical score to record specific and general neurological status of the animals in a predetermined manner, see Table 3
of the manuscript and Figure 5
Every 15 min post venom administration for the first 4 h following recovery from general anesthesia.
From 4 h to 8 h post venom administration, pigs were given a clinical and lameness (veterinary surrogate for pain) score every 30 min.
From 8 h to 48 h post venom administration pigs were given a clinical and lameness score every hour.
From 48 h to 96 h pigs were given a clinical and lameness score every 6 h.
All monitoring was done in person until 48 h. After 48 h, if pigs were asymptomatic they were observed in person twice daily and monitored by video for the other time points. If pigs were noted to be showing signs of pain (lameness score of 3 or greater) or distress (down and unable to rise, dragging hind legs, visible signs of dyspnea such as open mouth breathing or abdominal press when breathing, or a clinical score of 4 or greater) one of the study veterinarians examined the pig in person within 1 h of noting abnormal signs. Pigs that experienced life-threatening toxicity defined as a clinical score of 5 for two consecutive observations or a clinical score of 6 once either received immediate treatment with the study drug or antivenom or were humanely euthanized using 39% sodium pentobarbital given intravenously.
In lethality-dose-finding studies, to deduce if the weakness was due to toxicity from alpha-neurotoxins which bind to the nicotinic cholinergic receptor at the motor end-plate, animals were challenged with edrophonium and atropine. Pigs were given a combination of edrophonium (1 mg/kg) and atropine (0.02 mg/kg) combined in the same syringe and given slowly intravenously over one minute and did not exhibit any changes in condition (n = 2 animals), thus indicating that neurotoxicity in this model is based on presynaptic activity of neurotoxins. This protocol was therefore discontinued.
Laboratory Test Details: An arterial blood sample was collected and assayed to measure PaO2 and PaCO2 during anesthesia at baseline and post venom just prior to recovery to ensure animals adequately oxygenated. There were no hypoxic episodes unrelated to envenoming and oxygenation/ventilation did not reverse venom-induced weakness (data not shown). Venous blood samples were drawn through the catheter immediately following placement for baseline blood chemistry, complete blood count, prothrombin time (PT), and dynamic viscoelastic coagulometry (TEG [54
] and Sonoclot®
). All laboratory testing, with the exception of the Sonoclot and TEG, was performed by Antech Diagnostics (Stillwater, OK, USA). In addition, after centrifugation of blood, the color appearance of plasma was observed in order to judge the presence of gross hemolysis. Sonoclot and TEG were performed bedside at OSU. Following blood draws, catheters were flushed with 3 mL of heparinized saline. Subsequent blood samples were collected at the following time points utilizing a two-syringe technique to avoid heparin contamination of the samples; 0, 30min, 60 min, 4 h, 8 h, 24 h, 48 h, 100 h, and 120 h [54
]. Briefly, 3 mLs of blood were drawn from the catheter and discarded prior to drawing the sample. The sample was drawn and then the catheter was again flushed with 2 mLs of heparinized saline.
Sonoclot: Dynamic coagulation testing was utilized to demonstrate a more complete picture of the coagulopathy caused by M. fulvius
utilizes viscoelastic coagulometry to asses the time until clot formation, the strength of the clot that is formed, clot retraction, and clot lysis [55
]. Values provided by Sonoclot®
testing are platelet function, activated clotting time (ACT = time to initiate fibrin formation), clot rate (rate at which fibrinogen is converted to fibrin), and time to peak (time to reach peak clot strength).
Thromboelastography (TEG): TEG 5000 Thromboelastograph Hemostasis System (Haemoscope Corporation, Niles, IL, USA) was used to collect TEG measurements. Citrated blood (1 mL) was transferred to a kaolin tube (Kaolin activator, Haemonetics Corporation, Braintree, MI, USA). A sample of kaolin-activated blood was transferred to a cup with calcium chloride. The TEG was initiated at 37 °C. Paired samples were taken from corresponding citrated blood and kaolin-activated tubes and then placed into two separate disposable cups. Several TEG parameters were collected, including R-time (or clotting time; normal values: 2–8 min), alpha angle (normal values: 55–788), maximal amplitude (normal values: 51–69 mm), and LY30 (normal <8%) [54
]. An example of the technique compared to Sonoclot is shown as Figure 3