Biological and molecular properties of yellow venom of the Amazonian coral snake Micrurus surinamensis

Introduction: The coral snake Micrurus surinamensis, which is widely distributed throughout Amazonia, has a neurotoxic venom. It is important to characterize the biological and molecular properties of this venom in order to develop effective antitoxins. Methods: Toxins from the venom of M. surinamensis were analyzed by two-dimensional polyacrylamide gel electrophoresis and their neurotoxic effects in vivo were evaluated. Results and Conclusions: Most proteins in the venom had masses < 14kDa, low phospholipase A2 activity, and no proteolytic activity. The toxins inhibited the coagulation cascade. The venom had neurotoxic effects in mice, with a median lethal dose upon intravenous administration of 700 μg/kg. Immunogenic studies revealed abundant cross-reactivity of antielapidic serum with 14kDa toxins and limited cross-reactivity with toxins < 10kDa. These results indicate that antielapidic serum against M. surinamensis venom has weak potency (0.35mg/ml) in mice.


INTRODUCTION
Approximately 800 species of reptiles have been identified throughout the Brazilian territory; more than 50% of these are snakes including coral snakes, which are members of the Elapidae family 1 comprising approximately 40 genera 2 .Only the genera Micruroides Schmidt, 1928 (North America); Leptomicrurus Schmidt, 1937 (South America); and Micrurus Wagler, 1824 are found in the Americas 3,4 in fossorial or aquatic habitats.Coral snakes are small animals, ranging from 20cm to just over 1m in length, and have fixed venom inoculators of the proteroglyphous type.In general, they do not attack; poisoning by these snakes only occurs when they are handled or trampled 5 .Hence, envenomation is rare although the effects can range from mild to severe, since the venom is highly neurotoxic; symptoms include myasthenia (e.g., weakness and ptosis) that can evolve to paralysis and respiratory failure [6][7][8] .
Six species of coral snake have been identified as medically important in the State of Amazonas in Brazil: Micrurus averyi, Micrurus filiformes, Micrurus hemprichii, Micrurus lemniscatus, Micrurus spixii, and Micrurus surinamensis.These species are associated with low mortality rates in humans 24,25 .Given the diversity and abundance of coral snakes across the northern region of Brazil, biochemical studies of their venom have attracted wide interest.
Micrurus surinamensis has red cephalic scales with black borders and a red/black/yellow ring pattern, with the black rings arranged in perfect triads (Figure 1).Micrurus surinamensis lives in swampy areas (igapós) and streams in the primary forest area and feeds primarily on fish, but has a broad geographic distribution that includes southeastern Venezuela; Guianas; and the Amazonian regions of Colombia, Ecuador, Peru, Bolivia, and Brazil (Amapá, Amazonas, Pará, Maranhão, and Mato Grosso) 4,26,27 .The venom of M. surinamensis is distinct from that of other Micrurus species in terms of enzymatic composition 28,29 , as evidenced from the protein expression profi le.In the present study, we investigated the molecular composition and biological and neurotoxic activities of the venom of M. surinamensis, as well as the capacity of antielapidic serum to neutralize the neurotoxicity and lethal effects of this venom.

Animals and venom
Mice (BALB/cJ) weighing 18-22g were provided by Professor Paulo Friederich Bührnheim of the Experimental Animal Lab of Molecular Toxicology Laboratory Snakebite Center, Tropical Medicine Doctor Heitor Vieira Dourado Foundation, Amazonas, Brazil.The animals were maintained in plastic cages under a controlled temperature ranging from 22°C to 25°C with free access to water and food.The experiments were carried out according to local guidelines for the care and use of laboratory animals.
We extracted venom from snakes by pressing the head against a small glass vial covered with a Parafi lm membrane, with microcapillary tubes under the teeth.Lyophilized crude yellow venom from M. surinamensis (pooled from adult animals) provided by the Molecular Toxicology Laboratory of Venoms bank and developed by the Snakebite Center was stored −20°C.

Two-dimensional (2D) gel electrophoresis
Nonlinearly immobilized pH gradient (IPG) Immobiline DryStrips (24cm in length with a pH gradient of 3.0-10.0and 13cm in length with a pH gradient of 4.0-7.0 or 7.0-11.0)(GE Healthcare, Little Chalfont, UK) were hydrated for 12h with 500µg of M. surinamensis venom in sample buffer.Isoelectric focusing was carried out using an IPGphor III system (GE Healthcare).The IPG strips were separated according to mass by PAGE.
Discontinuous gels were obtained for the 24-cm strips with a pH gradient of 3.0-10.0.The 4% stacking gel and 15% separating gel were prepared from a 30% acrylamide solution and 0.8% N, N methylene-bisacrylamide dissolved in ultrapure water.The separating gel was prepared using 1.5M Tris-HCl buffer (pH 8.8) containing 0.1% SDS to solubilize the samples and 0.7% APS and 0.7% TEMED to polymerize the gel.The stacking gel was prepared after polymerization using 0.5M Tris-HCl buffer (pH 6.8), 0.1% SDS, 0.7% APS, and 0.7% TEMED.
Gradient gels (5%-20% polyacrylamide) were prepared using 13-cm IPG strips with pH gradients of 4.0-7.0 and 7.0-11.0.The resolving gel was prepared using acrylamide stock solution and 3.0 M Tris-HCl buffer (pH 8.8), to which 400µl of sucrose (1% by volume) were added.The spacer gel consisted of acrylamide stock solution and 3.0M Tris-HCl buffer (pH 8.8).The stacking gel was prepared from acrylamide stock solution and 0.5M Tris-HCl buffer (pH 6.8).B. atrox venom toxins (10µg) were used as molecular mass markers (50, 23, and 14kDa).All gels were scanned and analyzed using the Image Master 2D Platinum 6.0 system (GE Healthcare) according to the manufacturer's instructions.

Intracranial injection of mice
The neurotoxic effects of M. surinamensis venom on the mammalian central nervous system (CNS) were evaluated.Four groups of three mice were intracranial (i.c.) injected with 0.2, 1, 2, or 4µg venom diluted in physiological saline (0.15M NaCl).Samples were preincubated in a water bath at 37°C for 10 min prior to injection.Physiological saline was used as a control.A 1-ml insulin syringe equipped with a 0.3-mm needle (Ultra-Fine II; BD Biosciences, Franklin Lakes, NJ, USA) was used for injection.The needle size was modifi ed to a length of 3mm to accommodate the depth of insertion into the rat cranium.The animals were immobilized with their heads in a horizontal position, and the venom in a fi nal volume of 20µl was injected into the dorsal region of the frontal lobe at the longitudinal fi ssure.After inoculation, the animals were observed for 48h and neurological symptoms and mortality were documented.

Calculation of 50% lethal dose
Eight groups of four mice each were intravenously (i.v.) injected with 1, 2.5, 5, 7.5, 10, 12.5, 15, or 20µg of M. surinamensis venom diluted in saline solution (0.15M NaCl) for a final volume of 200µl.The animals were observed for 48h after inoculation of the venom, and their symptoms and mortality were documented.The lethal dose (LD50) was calculated using the Probit analysis method 31 based on the number of mice that died within 48h at each dose of venom.

Venom neutralization by antielapidic serum
The potency of the antielapidic serum was predetermined by calculating the 50% effective dose, defined as the dose at which 50% of the lethal effect exerted by a particular concentration of venom was neutralized with 1ml of the serum.Different dilutions of antielapidic serum produced at the Butantan Institute were mixed with 5× LD50 venom according to the recommendations of the World Health Organization 32 .Specifically, 1:1, 1:2, and 1:3 venom:serum dilutions were prepared and incubated for 30 min at 37°C in a water bath; the mixture (final volume of 200µl) was i.v.injected into four groups of three mice each.M. surinamensis venom was administered at 5× LD50 as a control.The animals were observed for 48 h, and the number of animals that died was noted.The potency of venom neutralization by the serum was determined by Probit analysis 31 .

Evaluation of defibrinogenating and hemorrhagic activity
The defibrinogenating activity of M. surinamensis venom was evaluated according to a previously described method 33 , with some modifications.Three groups of four mice were i.v.injected with 10, 15, or 20µg of venom diluted in saline solution (0.15M NaCl) for a final volume of 200µl per sample.The animals were anesthetized with ether and exsanguinated via cardiac puncture 1h after injection.Blood samples (1ml) in hemolysis tubes were allowed to stand at 26 °C for evaluation of blood coagulability.One animal per group was selected for assessment of hemorrhagic activity; the animals were anesthetized and sacrificed 2h after venom injection, and their thoracic cavities were opened to determine if bleeding was present.

Evaluation of PLA2 activity
PLA2 activity was detected by 1% agarose gel electrophoresis using 3% egg yolk phosphatidylcholine in 0.04 PBS buffer (pH 8.1) as a substrate.The agarose solution and egg yolk were homogenized, and the mixture was applied to a plastic plate and allowed to gel at room temperature.Eight wells (4mm in diameter) were created in the gel to form four columns.Column 1 served as a control for PLA2 activity and contained B. atrox venom (10µg venom/20µl physiological saline solution); and columns 2-4 contained M. surinamensis venom at concentrations of 10, 20, and 40µg/20µl physiological saline, respectively.The samples were incubated at 37°C in a humid chamber for 24h.After incubation, the area forming a clear halo -reflecting PLA2 activity -was measured in millimeters using digital calipers.

Immunoneutralization of PLA2 activity
Inhibition of PLA2 activity was determined on plates containing 1% agarose gel using 3% egg yolk phosphatidylcholine in 0.04M PBS buffer (pH 8.1) as a substrate.Different amounts of antielapidic serum were used at M. surinamensis venom volume ratios of 1:1, 1:0.75, 1:0.50, 1:0.25, 1:0.10, and 1:0.05 (µg of venom: µl of serum) to verify the efficacy of the serum in neutralizing PLA2 enzymatic activity.The venom/serum mixtures were incubated in a water bath at 37°C for 30 min.Venom from M. surinamensis (20µg/µl) and B. atrox (10µg/ µl) without antivenom served as controls.A 20-μl volume of the solution (venom + serum) was applied to each well of the gel, and the plates were incubated for 24h at 37°C in a humid chamber.Gel areas were analyzed for the presence or absence of PLA2 activity.Specifically, the diameters of the halos (mm) were measured with a digital caliper.The inhibitory activity of the serum was quantified as the percent difference in PLA2 activity in the experimental group relative to the control group according to the following formula: % inhibition = (100 − PLA2 activity-related halo diameter) × 100/(control PLA2 activityrelated halo diameter).

Evaluation of proteolytic activity by zymography
Proteolytic activity was assessed by zymography using a previously described method 34 .A 15% SDS-PAGE gel was prepared, and 1% (w/v) casein, fibrinogen, and gelatin substrates were separately added to each gel before polymerization.After electrophoresis, gels were washed with Triton X-100 for 1h at room temperature followed by ultrapure water for several minutes, then placed in glycine buffer (pH 8.3) (gels with casein or fibrinogen) or collagenase buffer (pH 7.5) (gels with gelatin) in an oven at 37°C for 24h.B. atrox venom (30µg) was used as a control for proteolytic activity and as a molecular mass marker.

In vitro evaluation of anticoagulant activity (recalcification time)
The anticoagulant activity of the venom was evaluated as human plasma recalcification time 35 .Blood mixed with 3.8% sodium citrate at a 1:9 ratio was centrifuged at 1610 g for min, and the plasma was separated and refrigerated at 4°C.A 200-μl volume of human plasma was added to hemolysis tubes and maintained in a water bath at 37°C.Samples with 20, 40, or 80µg of M. surinamensis venom diluted in physiological saline were prepared and 20μl were separately transferred to the human plasma-containing tubes.At the predetermined concentration, each mixture was homogenized and then combined with 20µl of 0.4M calcium chloride (CaCl 2 ).Triplicate samples were evaluated by observing the recalcification (clot formation) start time.The control consisted of plasma alone combined with 20µl of 0.4M CaCl 2 .

Western blot analysis of competitive interaction
Western blotting was performed as previously described 36 , with some modifications.Samples were subjected to 5%-20% SDS-PAGE and the proteins were transferred to a nitrocellulose membrane in transfer buffer with a constant current of 260mA and 60V.The membrane was stained with Ponceau S to verify the efficiency of the transfer and then cut into strips that were washed thoroughly with distilled water to remove excess dye and treated with blocking solution composed of 5g skim milk dissolved in 100ml Tris-buffered saline (TBS) with Tween 20 (TBS-T) for 2h.The strips were washed three times for 5 min each with TBS-T solution.To quantify the number of antibodies capable of binding to venom toxins, samples were pre-incubated in a water bath at 37°C for 30 min in mixtures containing 5, 20, or 80µg of M. surinamensis venom with 4µl of antielapidic serum in 5ml TBS.Each preparation was applied to a nitrocellulose membrane strip and incubated for 1.5h.As a control, a strip treated with antielapidic serum only (diluted 1:1,000 in TBS) was incubated for 1.5h at room temperature.Nitrocellulose membranes were washed three times with TBS-T for 5 min each and then incubated for 1.5h at room temperature with the immunoenzyme conjugate [peroxidase-labeled antihorse immunoglobulin G (IgG)] diluted 1:2,000 in TBS.After washing three times with TBS-T and twice with TBS for 5 min each, protein bands were detected by adding the peroxidase substrate (1.5mg of 4-α-chloro-1-naphthol in 24µl of H 2 O 2 , 0.5ml methanol, and 17.5ml TBS).Antigen-antibody binding ability was determined by competition assays.

Molecular profi les of Micrurus surinamensis venom proteins
Two-dimensional gel electrophoresis of venom samples in gels with a pH range of 3-10 revealed 43 spots, including 22 and 21 in the acidic and basic regions, respectively.A total of 26 spots had a mass below 23kDa, and 17 had approximately the same mass above 50kDa but had different pH ranges (Figure 2A).Gels with a pH range of 4-7 had 42 spots, most of which were in the pH range of 5-6.Fourteen spots had a mass below 23kDa, and 28 had a mass above 50kDa; the majority of spots in this acidic pH range had similarly high masses (Figure 2B).Spots detected in gels with a pH range of 7-11 had low masses below 23kDa (Figure 2C).

Micrurus surinamensis venom
Micrurus surinamensis venom at doses greater than 20µg was highly toxic, causing immediate muscle and respiratory paralysis and death in mice within a few minutes of administration.The following symptoms were observed following i.c.injection of 0.2, 1, 2, and 4µg of venom: total muscle paralysis and difficulty breathing (< 1 min); leaping, increased energy, hypersensitivity to touch and sound, compulsive scratching (30 min); and convulsions followed by death at the higher doses.Symptoms caused by i.c.injection of 20µl of physiological saline solution (as a control for activity) were wheezing and apathy that ceased within 10 min of injection without leading to death.The following symptoms were observed following i.v.injection of 1, 10, 15, or 20µg of M. surinamensis venom: diffi culty breathing; muscle hypotonia; unilateral and bilateral ptosis (30 min); compulsive scratching (legs, genitalia, tail, head, nose, and eyes); muscle stretching in the hind legs (1:30h); and death due to respiratory arrest at the higher doses.
Probit analysis revealed that the LD50 of i.v.administered venom was 14µg (16.8-11.3)per 20g body weight (700μg/kg).The neutralizing capacity of the antielapidic serum was very  low, with a potency of 0.3 (0.5-0.02)mg/ml (venom/serum).In animals i.v.injected with 7.5, 10, or 13µg of venom there was no hemorrhage in the thoracic cavity, nasal cavity, or genitalia.Blood collected from these animals coagulated within a normal time interval for mouse blood (< 60s).However, in vitro experiments revealed a delay in the coagulation time of venom-treated plasma as compared to that of control plasma.Specifically, human plasma clotted after 10 and 30 min in the presence of the lowest (20µg) and highest (80µg) doses of venom, respectively.Animals that received 15 or 20µg of venom were autopsied immediately after death and exhibited no hemorrhaging in the thoracic cavity or in other visible areas of the body, such as the external genitalia and nose (Table 1).

Evaluation of serum-mediated neutralization of Micrurus surinamensis venom enzymatic activity
Micrurus surinamensis venom exhibited low PLA2 activity compared to the control (10µg of B. atrox venom) at doses of 10, 20, and 40µg (Figure 3A).The antielapidic serum inhibited 100% of the PLA2 activity of M. surinamensis venom when administered at a ratio between 1:1 and 1:0.05 (Figure 3B).The antielapidic serum (maximum and minim doses of 20 and 1µl, respectively) also inhibited 100% of the PLA2 activity of M. surinamensis venom.The venom had no caseinolytic, fibrinolytic, or gelatinolytic protease activity according to the zymography results (data not shown).

Western blot analysis of competitive interaction
A western blotting analysis revealed that the antielapidic serum had low potency (Figure 4).Antivenom (4µl) + venom (5, 20, and 80µg) mixtures preincubated at 37°C for 30 min showed several low-intensity bands at 7-20kDa and especially around 14kDa, although these exhibited variable signal intensity.The proteins were detected after just 1h of reaction with peroxidase-specific substrates.

Biological activities
In the present study   venom of M. surinamensis differs from that of other Micrurus species owing to its abundance of peptides between 6-7kDa, which was confirmed in the present work.
The 2D SDS-PAGE profiles of M. surinamensis venom revealed differences in venom composition.Previous studies using this approach identified approximately 30 proteins within the pH range 3-10, with the most intense spots observed in the range of 11-25kDa and only one spot in the range of 30-150kDa that showed high similarity to l-aspartate oxidase 41 .In the present study, we identified 43 spots within the pH range of 3-10.All of the spots in the basic pH range had a low mass, whereas more than 60% of the spots in the acidic pH range had a high mass.A total of 17 spots had similar masses above 50kDa and an acidic pH, suggesting that they were isoenzymes.In some species, isoenzymes with basic pH and masses of 8-13kDa have been isolated that exhibit various biological activities (e.g., cardiotoxic, neurotoxic, PLA2, and hemorrhage-inducing) and, in some cases, lethality in association with hemoptysis and hemoglobinuria [42][43][44] .
Despite their similar profiles, each type of snake venom has peculiarities in terms of protein composition and abundance 45,46 .Additionally, differences in snake venom proteomes within the same species inhabiting different areas are not uncommon 47 .Proteomic variations in the venom of Micrurus species may be associated with differences in their geographic location and/or habitat as well as seasonal variations, dietary factors, age, sexual dimorphism, and even evolutionary history.
In this study, toxins from M. surinamensis did not cause bleeding and produced symptoms typical of venom from this genus, including uni-and/or bilateral ptosis, muscular hypotonia, spasms, respiratory failure, and exophthalmos.M. surinamensis venom exhibited high neurotoxicity, causing death via respiratory paralysis within minutes of injection.Toxins present in Micrurus venoms generally do not cause localized or systemic bleeding and induce mild myonecrosis 48,11 , while M. surinamensis venom is not known to induce myotoxicity, hemorrhage, or edema 49 .At high concentrations, M. altirostris venom does not cause bleeding, dermonecrosis, or coagulant activity 37 ; venoms of M. averyi and M. fulvius cause hemorrhaging, renal damage, and severe inflammation 49,50 .In some cases, PLA2 is responsible for inducing severe bleeding and myoglobinuria 51 , but owing to the lethality of the neurotoxins present in the venom, these symptoms are often not manifested.
In general, snake toxins are associated with less bleeding and more neurotoxic effects; however, the venoms of snakes in the Viperidae and Crotalidae families cause serious bodily injury due to bleeding within vital organs and blood circulation impairment caused by clot formation, edema, and necrosis.These symptoms are mainly attributed to serine and metalloproteases responsible for proteolytic degradation in mammalian blood plasma and tissues, which may function as blood coagulation activators or inhibitors 15,[52][53][54] .
The venoms of some species of the genera Naja, Bungarus, and Micrurus lack proteolytic activity against gelatin, casein, and fibrinogen 55,56 ; this was also true of the venom of the species examined in the present study.However, high concentrations of M. surinamensis venom caused an increase in human plasma coagulation time, suggesting that it has anticoagulant activity.Proteases are known to inhibit platelet aggregation 57 and PLA2 inactivates blood clotting factors to block the coagulation cascade 58 .Thus, PLA2 concentration may also be related to anticoagulant activity 44,59 .
Venom toxins of most coral snakes exhibit PLA2 activity along with neurotoxic and myotoxic effects 60 .Both α-and β-type neurotoxins are responsible for the lethality of the venom in mice and are 10 times more potent when directly injected into the CNS.Thus, although these neurotoxins have no effects in the peripheral nervous system (PNS), even small amounts are highly toxic in the CNS 61 , causing damage in the cerebral cortex, rostrocaudal region of the brain, hemispheric white matter, corpus callosum, fornix, and hippocampus 14 .Similar damage is caused by some three-finger neurotoxins that bind to rat hippocampus-specific muscarinic acetylcholine and A-type γ-aminobutyric acid receptors following jaw paralysis and intense seizures induced by intracerebroventricular injection of venom 61,62 .
Four fractions (7-22kDa) exhibiting neurotoxic PLA2 activity isolated from M. lemniscatus venom caused severe symptoms of envenomation in the mouse CNS 63 .Small doses (1-2.1μg/μl)induced a variety of detectable symptoms, including spasms, breathing difficulty, limbic seizures, cortical and hippocampal epileptic discharges, episodes of convulsion, and ultimately death (in 80% of animals); high doses (4.5μg/μl) caused symptoms consistent with severe intoxication such as reduced motor activity, hypersensitivity to touch and sound, aggression, and death.Histological analysis revealed massive hippocampal neuronal loss.These results demonstrate that M. surinamensis venom is highly neurotoxic to both the PNS and CNS.Clarifying the molecular mechanisms underlying these effects can lead to the development of new drugs for the treatment of neurological diseases based on neuropeptides produced by these snakes 64 .
The species examined in the present work showed low PLA2 activity that was effectively neutralized by antielapidic serum from the Butantan Institute.A previous study found no phospholipase activity in M. surinamensis venom, and antivenom produced by the Butantan Institute was minimally effective in detecting venom components 65 .In the present study, M. surinamensis venom components larger than 10kDa were only weakly detected by western blotting.Our in vitro tests revealed the presence of antibodies in the antielapidic serum that recognized and neutralized proteins of approximately 14kDa in M. surinamensis venom, but not bands corresponding to neurotoxins smaller than 10kDa.These results demonstrate that individual snakes of the same species have unique venom protein components and immunological profiles.
Our in vivo analysis showed that the antielapidic serum had low efficacy in neutralizing M. surinamensis venom components that exhibited high neurotoxicity by i.v.injection (14µg/20g venom/mouse), indicating that high doses of serum are required for complete neutralization 66,67 .The low recognition of these neurotoxins is a significant limitation given their abundance and toxicity 68 .
The immune profile of M. surinamensis was distinct from that of other species.In particular, the toxic components of its venom showed limited reactivity with the antielapidic serum 65,69 produced in Brazil by hyperimmunizing horses with Micrurus corallinus and Micrurus frontalis venoms obtained from the Butantan Institute.Despite their small size and corresponding low venom inoculation capacity, high doses of serum are recommended for envenomations involving coral snakes 7 .The intra-species variability of coral snake venom raises concerns regarding the effectiveness of sera developed using venoms produced by a limited number of coral snake species 70 .A monoclonal antibody against M. nigrocintus nigrocintus reacted with M. surinamensis venom, indicating that the former belongs to a group with a different antigenic profile than other Micrurus species of medical importance in Brazil 39 .Geographic and dietary variations as well as phylogenetic factors and mutations account for the diversity in snake venom composition.Although diet is the main contributing factor 71 , other evolutionary forces driven by natural selection such as allelic mutations in an enzyme that leads to structural changes in the protein may be equally important 72 .
A recent study evaluated the immunogenic potential of venoms of different Micrurus species to ensure good crossreactivity with antielapidic sera produced in experimental animals and to establish an antigenic mixture for generating polyvalent antivenoms with higher efficacy than those used to treat envenomation in humans 69 .The results were positive for most but not all of the species investigated; in fact, the M. surinamensis venom was less efficiently recognized by the antivenom, underscoring the difficulty of this task.Additional basic and clinical studies are needed to generate a monovalent serum or venom that includes these species in the pool used to produce commercial antivenom.Snake bites remain a serious public health problem that have generally been neglected by the authorities and the general population 73 .Thus, there are many outstanding challenges and issues that must be addressed to promote snake bite prevention and to develop effective treatments for envenomation.

TABLE 1
Comparative biological activities of toxins in Micrurus surinamensis venom.