Diversity of Micrurus Snake Species Related to Their Venom Toxic Effects and the Prospective of Antivenom Neutralization

Background Micrurus snake bites can cause death by muscle paralysis and respiratory arrest, few hours after envenomation. The specific treatment for coral snake envenomation is the intravenous application of heterologous antivenom and, in Brazil, it is produced by horse immunization with a mixture of M. corallinus and M. frontalis venoms, snakes that inhabit the South and Southeastern regions of the country. However, this antivenom might be inefficient, considering the existence of intra- and inter-specific variations in the composition of the venoms. Therefore, the aim of the present study was to investigate the toxic properties of venoms from nine species of Micrurus: eight present in different geographic regions of Brazil (M. frontalis, M. corallinus, M. hemprichii, M. spixii, M. altirostris, M. surinamensis, M. ibiboboca, M. lemniscatus) and one (M. fulvius) with large distribution in Southeastern United States and Mexico. This study also analyzed the antigenic cross-reactivity and the neutralizing potential of the Brazilian coral snake antivenom against these Micrurus venoms. Methodology/Principal Findings Analysis of protein composition and toxicity revealed a large diversity of venoms from the nine Micrurus species. ELISA and Western blot assays showed a varied capability of the therapeutic antivenom to recognize the diverse species venom components. In vivo and in vitro neutralization assays indicated that the antivenom is not able to fully neutralize the toxic activities of all venoms. Conclusion These results indicate the existence of a large range of both qualitative and quantitative variations in Micrurus venoms, probably reflecting the adaptation of the snakes from this genus to vastly dissimilar habitats. The data also show that the antivenom used for human therapy in Brazil is not fully able to neutralize the main toxic activities present in the venoms from all Micrurus species occurring in the country. It suggests that modifications in the immunization scheme, with the inclusion of other venoms in the antigenic mixture, should occur in order to generate effective therapeutic coral snake antivenom.


Introduction
The Elapidae family has about 250 species, distributed from the Southeastern and Southwestern United States, through Mexico, Central America and South America, and are also found in Asia, Africa and Australia [1]. In the Americas, there is a group of more than 120 species and subspecies, divided into three genera: Micruroides, with one species; Leptomicrurus with three, and Micrurus, with almost 70 species [1,2,3,4].
Coral snakes have a large geographical distribution in Americas, inhabiting extremely diverse environments, from lowland rainforests and deserts to highland cloudy forests [1,5]. Most of the snakes of the Micrurus genus has terrestrial to subfossorial habits, however, some species are semi-aquatic, such as M. surinamensis and M. lemniscatus [1].
Most coral snakes have a color pattern of some combination of the red, yellow or white, and black, usually disposed in rings.
They are proterogliphous animals, presenting the fixed small teeth at the forefront of the mouth. Food is generally composed of small snakes, but may also include lizards and amphisbaenians. Certain species have specialized nutritional habits, feeding on caecilians, swamp eels, and other type of fishes and even onycophorans and other invertebrates [1]. Snakes such as M. lemniscatus and M. surinamensis feed on fish and M. hemprichii of peripatus [6,7,5].
Human envenomations by coral snakes are relatively rare due to their subfossorial habits; however, the case fatality, attributable to respiratory paralysis, may be high [11]. A variety of local and systemic manifestations of envenoming has been described in patients bitten by different species of coral snakes [5,12,11]. The main feature of the venom action is the neurotoxicity, although, experimentally, it has been reported that some Micrurus venoms may induce myotoxicity and local lesions [13,14]. Neurotoxicity can be produced by a post-synaptic action, through blockage of the end-plate receptors by alpha neurotoxins, as determined for M. frontalis venom, or by a pre-synaptic-like activity, which causes inhibition of acetylcholine release at the motor nerve endings, as induced by M. corallinus venom. There are also venom toxins, such as cardiotoxins and myotoxic phospholipases A 2 from M. nigrocinctus and M. fulvius, which block the end-plate receptors and depolarize the muscle fiber membrane [15].
The transcriptomic analysis of a Micrurus snake venom gland (M. corallinus) was recently described [29]. Toxin transcripts represented 46% of the total ESTs and the main toxin classes were neurotoxins, i.e, three-finger toxins (3FTx) and phospholipases A 2 (PLA 2 s). It was also showed that the post-synaptic components (3FTx) were very diverse in terms of sequences, possibly aiming to achieve different types of receptors, whereas the pre-synaptic component (PLA 2 ) was more conserved. The high expression of both types of these neurotoxins is in agreement with the known presence of pre-and post-synaptic activities in the Micrurus venoms. However, eight other classes of toxins were found, including C-type lectins, natriuretic peptide precursors and highmolecular mass components such as metalloproteases and Lamino acid oxidases.
The specific treatment for Micrurus envenomation is the intravenous application of heterologous antivenom. In Brazil, the coral snake therapeutic antivenom produced by Butantan Institute is obtained by the immunization of horses with a mixture containing equivalent amounts of M. corallinus and M. frontalis venoms [30]. In view of the fact that Micrurus venoms can exhibit a diversity of composition and toxicity, the therapeutic antivenom may not be capable to fully recognize all the major components of the distinct venom species occurring in the country. Therefore, the aim of this study was to characterize some biological properties of venoms from nine species of Micrurus, including those used for serum preparation, i.e., M. frontalis and M. corallinus, evaluate their antigenic cross-reactivity, using the Brazilian coral snake antivenom, as well as to test the ability of this antivenom to neutralize the main toxic activities of these venoms.

Electrophoresis and western blot
Samples of 20 mg of Micrurus venoms were solubilised in nonreducing sample buffer and run on 7.5 to 15% SDS-PAGE gradient gels [32]. Gels were stained with silver [33] or blotted onto nitrocellulose [34]. After transfering, the membrane was blocked with PBS containing 5% BSA and incubated with the coral snake antivenom (diluted 1:2,000) for 1 h at room temperature. The membrane was washed 3 times for 10 min with PBS/0.05% Tween 20, and incubated with GAH/IgG-AP (1:7,500) in PBS/1% BSA for 1 h at room temperature. After washing 3 times for 10 min with PBS/0.05% Tween 20, the blot was developed using NBT/BCIP according to the manufacturer's instructions (Promega).

Author Summary
The Elapidae family is represented in America by three genera of coral snakes: Micruroides, Leptomicrurus and Micrurus, the latter being the most abundant and diversified group. Micrurus bites can cause death by muscle paralysis and respiratory arrest few hours after envenomation. The specific treatment for Micrurus envenomation is the application of heterologous antivenom. The aim of this study was to compare the toxicity of venoms from nine species of coral snakes and analyze the neutralization potential of the Brazilian coral snake antivenom. In vitro assays showed that the majority of the Micrurus venoms are endowed with phospholipase and hyaluronidase and low proteolytic activities. These enzymes are not equally neutralized in all venoms by the therapeutic antivenom. Moreover, in vivo assays showed that some of the Micrurus venoms are extremely lethal, such as the ones from M. altirostris, M. corallinus, M. frontalis, M. lemniscatus and M. spixii. Neutralization tests, performed in vivo, showed that the therapeutic antivenom was able to neutralize better the venoms from M. frontalis, M. corallinus, and M. spixii but not from M. altirostris and M. lemniscatus. Taken together, these results suggest that modifications in the immunization antigenic mixture should occur in order to generate more comprehensive therapeutic antivenom.

Determination of LD 50
The lethal potential of Micrurus venoms was assessed in Swiss mice by intraperitoneal injection of different amounts of venoms in 500 mL of PBS. Four animals were used for each venom dose (five doses). The LD 50 was calculated by probit analysis of death occurring within 48 h of venom injection [35]. All animal experiments were approved in advance by the Laboratory Animal Ethics Committee of Butantan Institute.

Phospholipase activity
The phospholipase A 2 activity of Micrurus venoms was determined as described by Price III [36], with some modifications. Samples of the venoms (4 mg) and PBS were added to a final volume of 200 mL. Samples of 180 mL of the mixture containing: 5 mM Triton X-100, 5 mM phosphatidylcholine (Sigma), 2 mM HEPES, 10 mM calcium chloride and 0,124% (wt/vol) bromothymol blue dye in water, at pH 7.5 and at 37uC, were added. After a pre-incubation of 5 min at 37uC, the absorbance of the samples was determined at l 620 nm in a Multiskan spectrophotometer EX (Labsystems, Finland). Results were expressed in nanomoles of acid per minute per mg of venom (compared on pH changes in standard curves of the reaction mixture using HCl).

Proteolytic activity
Samples of the Micrurus venoms (50 mg) were mixed with 5 mM of the Fluorescent Resonance Energy Transfer (FRET) substrate, Abz-FEPFRQ-EDnp, and PBS, for a final volume of 100 mL, and the reactions monitored by measuring the fluorescence (l em 420 nm and l ex 320 nm) in a spectrofluorimeter (Victor 3 TM , Perkin-Elmer, USA) at 37uC, as described by Araújo et al. [37]. The specific proteolytic activity was expressed as units of free fluorescence per minute per mg of venom (UF/ min/mg).

Hyaluronidase activity
Hyaluronidase activity was measured as described [38], with slight modifications. Samples of Micrurus venoms (30 mg) were added to 100 mL of the hyaluronic acid substrate (1 mg/mL) and acetate buffer (pH 6.0) for a final volume of 500 mL. The mixtures were incubated for 15 min at 37uC. After the incubation, it was added to the samples 1 mL of cetyltrimethylammonium bromide 2.5% in NaOH 2%, to develop the turbidity in the mixtures, and the absorbance measured in a spectrophotometer (Multiskan EX) at l em 405 nm. Results were expressed in units of turbidity reduction (UTR) per mg of venom.

Enzyme linked immunosorbent assay (ELISA)
Microtitre plates were coated with 100 mL of Micrurus venoms (10 mg/mL; overnight at 4uC). Plates were blocked with 5% BSA in PBS and increased dilutions of the therapeutic coral snake antivenom were added. After 1 h of incubation at room temperature, plates were washed with PBS/0.05% Tween 20 and incubated with GAH-IgG-HRPO diluted 1:3,000, for 1 h at room temperature. Plates were washed and the reactions developed with OPD substrate according to the manufacturers conditions (Sigma). The absorbances were recorded in an ELISA reader (Multiskan spectrophotometer EX) at l 492 nm. The titer was established as the highest antivenom dilution, in which an absorbance five times greater than that determined for the normal horse serum was measured.

Serum neutralization assays performed in vitro
The ability of the therapeutic Brazilian coral snake antivenom to neutralize the venoms phospholipase, hyaluronidase and proteolytic activities was estimated by incubating Micrurus venoms with the antivenom. The antivenom volume, amount of venoms and the pre-incubation time, for each tested enzymatic activity, was standardized using the immunization pool, composed by 50% of M. corallinus and 50% of M. frontalis venoms. For serum neutralization measurements of the phospholipase activity, samples of 4 mg of the venoms were incubated with the antivenom, diluted 1:10; for the hyaluronidase activity, samples of Micrurus venoms (30 mg) and the antivenom (1:20) were incubated for 20 min at room temperature; for the proteolytic activity, samples of Micrurus venoms (50 mg) and the antivenom (1:4) were incubated for 10 min at room temperature. Venoms residual toxic activities were measured as described above.

Neutralization of the lethal activity
The capacity of the therapeutic coral snake antivenom to neutralize the lethal activity of Micrurus venoms was determined by mixing the venoms, corresponding to 2 LD 50 , with serial dilutions of the horse antivenom. The mixtures were incubated for 30 min at 37uC and the animals received 0.5 mL by the intraperitoneal route. The effective dose (ED 50 ) was calculated from the number of deaths within 48 h of injection of the venom/antivenom mixture using probit analysis, as described above. The ED 50 was expressed as mL of antivenom per mg of venom.

Eletrophoretic characterization of Micrurus venoms
The protein profiles of Micrurus venoms were analyzed by SDS-PAGE followed by silver staining. Figure 1 shows that the venoms from the nine coral species differ in composition, number and intensity of bands. The majority of the components of these venoms present Mr inferior to 64 kDa. Venom from M. surinamensis differs more from the others, by the presence of a few number of components with Mr lower than 20 kDa.

Enzymatic activities of the Micrurus venoms
In order to assess whether the venoms of Micrurus displayed the same biological activities, some functional assays were carried out. The proteolytic activity of the Micrurus venoms was tested using a FRET substrate, Abz-FEPFRQ-EDnp.

Immunochemical cross-reactivity
The coral snake antivenom, produced by Butantan Institute and used in Brazil for human serum therapy, is obtained by the immunization of horses with a mixture of M. corallinus and M. frontalis venoms. In an ELISA, this antivenom was tested for crossreactivity, using Micrurus spp venoms as antigens. Figure 5A  By western blotting, it was demonstrated that the coral snake antivenom could recognize several but not all components present in the Micrurus spp venoms. M. surinamensis venom components were weakly detected. The antivenom was also unable to   (Fig. 5B).

In vitro antivenom neutralization assays
In order to analyze if the Brazilian coral snake antivenom could neutralize the enzymatic activities present in the Micrurus spp venoms, some in vitro assays were performed. Figure 6A shows   (Fig. 6C).

Serum neutralization of the lethal activity
Some coral snake venoms were chosen for further in vivo antivenom neutralization analysis based on their lethal toxicity. Figure 7 shows that the coral snake antivenom was able to neutralize, although with different potencies, the venoms from M.

Discussion
Biochemical studies concerning the Micrurus venoms are very scarce, due to difficulties in the correct identification of the species, extraction of the venom and maintenance of the animals in captivity. Previous studies have demonstrated that snakes venoms from Micrurus genus present individual variations in composition, related to their geographic distribution, age, gender and diet [22,39,40].
In the present study, we have investigated the toxic properties of venoms from nine species of Micrurus, the antigenic cross-reactivity and the neutralizing potential of the Brazilian therapeutic coral snake antivenom against these venoms. Results are summarized in Table 2.
Analysis of the Micrurus spp biological properties, as performed by testing the phospholipase, proteolytic, hyaluronidase and lethal activities showed a great variability in the venoms composition. Thus, data presented here showed that the majority of the venoms present intense PLA 2 activity, although it is lacking in M. surinamensis venom.
Venoms from Micrurus genus have been characterized as possessing low or no proteolytic activity [41]. In the present study, using the FRET substrate Abz-FEPFRQ-EDnp, we could identify proteolytic activity in the majority of Micrurus venoms, being the exception the M. surinamensis venom. Moreover, it was also demonstrated that Micrurus venoms posses varied levels of hyaluronidase activity.
Micrurus venom is primarily neurotoxic, causing little local tissue reaction or pain at the bite site. Once clinical signs of coral snake envenomation appear they progress with alarming rapidity and are difficult to reverse. In a recent clinical report, it was observed unusual features of coral snake (M. lemniscatus helleri) envenomation, with the patient presenting persistent severe local pain, very slow evolution of neurotoxic envenoming, which after 60 h culminated with respiratory failure [11]. These data reinforce the idea that differences in venoms composition may be responsible for the variety of systemic and local manifestations of coral envenoming.
Our results showed that the Brazilian coral snake antivenom presents a variable capability of recognizing venoms antigens, as demonstrated by differences in the antibody titers, as measured by ELISA, being the lowest the one obtained for M. surinamensis venom. By Western blotting, it was revealed that the antivenom recognized components of Mr from 64. Higashi et al. [43] have demonstrated that anti-M. corallinus antivenom was able to neutralize, in vivo, the lethal effect of M. corallinus venom, but not from M. frontalis, M. ibiboboca and M. spixii venoms. In contrast, antivenom against M. frontalis could neutralize the lethal effect of M. frontalis, M. ibiboboca and M. spixii venoms, but not from M. corallinus. Abreu et al. [44] showed that the commercial and experimental coral antivenoms have low efficacy in neutralizing the M. altirostris venom neurotoxicity as measured in in vitro and in vivo (inhibition of the lethality) assays. Table 2 shows that the antivenom antibody titers had no positive correlation with its neutralization potential, indicating that in vitro and in vivo neutralization tests are fundamental to determine the efficacy of the therapeutic antivenom.
Multivalent coral snake antivenom has been also prepared, in horses, against a mixture of venoms from M. nigrocinctus, M.  mipartitus and M. frontalis species [45]. In this study it was suggested that it would be useful in treating bites from most of the important coral snake species in North and South America, such as M. fulvius, M. alleni, M. carinicaudus dumerilii, M. corallinus, M. frontalis, M. lemniscatus, M. mipartitus, M. nigrocinctus and M. spixii. They also note that M. surinamensis venom was not significantly neutralized by the antivenom. Table 2 shows the existence of a large range of both qualitative and quantitative variations in Micrurus venoms, probably reflecting the adaptation of the snakes from this genus, to vastly dissimilar habitats. Thus, the comparative analysis of distinct phenotypes, particularly the venom constituents and their toxic activities, reveals the heterogeneous complexity of the Micrurus venoms ascertaining that both the structural and the ecological evolutions constrain specific characters for adaptive values. The most striking example is given by M. surinamensis, a snake that inhabits an extremely distinct environment, whose venom expresses limited composition. Besides, it was showed that the antivenom used for human therapy in Brazil is not fully able to neutralize the main toxic activities present in all Micrurus spp venoms, indicating that, for the preparation of the Brazilian coral snake antivenom, other venoms should be included in the immunization mixture.
Taking into account the decision made by PAHO/WHO [46] and the countries of the Americas to promote strategies to diminish the health burden of accidents involving poisonous animals in the countries of Latin America, it would be lawful to consider the possibility to prepare a continental coral snake antivenom, thus contributing to countries where national production is insufficient or where it does not have manufacturing laboratories. The appropriate cooperation by scientists in various countries in order to prepare multivalent coral snake antivenom has already been proposed by Bolañ os et al. [46], as early as the 1970's, but until now this relevant aim for the public health of the Americas has not been achieved. Data present in the literature, and results obtained in this study, should encourage PAHO to coordinate a regional cooperative effort to produce multivalent continental Micrurus antivenom that would have an important impact in the treatment of accidents involving coral snakes over the entire continent.