Proteomic and toxicological analysis of the venom of Micrurus yatesi and its neutralization by an antivenom

Coralsnakes belong to the family Elapidae and possess venoms which are lethal to humans and can be grouped based on the predominance of either three finger toxins (3FTxs) or phospholipases A2 (PLA2s). A proteomic and toxicological analysis of the venom of the coralsnake Micrurus yatesi was performed. This species, distributed in southeastern Costa Rica, was formerly considered a subspecies of M. alleni. Results showed that this venom is PLA2-rich, in contrast with the previously studied venom of Micrurus alleni. Toxicological evaluation of the venom, in accordance with proteomic data, revealed that it has a markedly higher in vitro PLA2 activity upon a synthetic substrate than M. alleni. The evaluation of in vivo myotoxicity in CD-1 mice using histological evaluation and plasma creatine kinase release also showed that M. yatesi venom caused muscle damage. A commercial equine antivenom prepared using the venom of Micrurus nigrocinctus displayed a similar recognition of the venoms of M. yatesi and M. nigrocinctus by enzyme immunoassay. This antivenom also immunorecognized the main fractions of the venom of M. yatesi and was able to neutralize its lethal effect in a murine model.


Introduction
Coralsnakes, genus Micrurus, are small to moderate-sized (from less than 50 cm-150 cm in total length) slender elapid snakes that populate diverse habitats, which include lowland rainforests, deserts, and highland cloud forests from southern United States to central Argentina (Campbell and Lamar, 2004;Roze, 1996). Most coralsnakes have a color pattern of some combination of red, yellow or white, and black rings (Campbell and Lamar, 2004). Although a comprehensive phylogenetic analysis of the species that make up the genus is still pending (Zaher et al., 2021), some monophyletic groups have been identified based on the structure of their hemipenes and molecular characters: the group of monadal black ring coralsnakes, which have slender and strongly bifurcated hemipenes; the group of triad pattern coralsnakes with short, bilobed hemipenes; and the group of bicolored coralsnakes with strongly bilobed and slender hemipenes (Slowinski, 1995). Envenomings by coralsnakes are characterized by paresthesia, local pain, palpebral ptosis, dizziness, blurred vision, weakness, slight local edema, erythema, dysphagia, dyspnea, myalgia, salivation and respiratory failure which may lead to death (Bucaretchi et al., 2016). However, regardless of the toxicity of their venoms, bites by these elapids are far less frequent than those caused by pitvipers, representing less than 2% of snakebite envenomations reported in the Western Hemisphere (Bucaretchi et al., 2016).
Transcriptomic and proteomic analyses have revealed that venoms from coralsnakes are characterized by a predominance of phospholipases A 2 (PLA 2 s) and three-finger toxins (3FTxs), with lower quantities of proteins from other families (Aird et al., 2017;Corrêa-Netto et al., 2011;Lomonte et al., 2016Lomonte et al., , 2021Sanz et al., 2019b). Postsynaptically active 3FTxs in these venoms block nicotinic cholinergic receptors by competing with acetylcholine (Moreira et al., 2010), while presynaptically active PLA 2 s hydrolyze phospholipids at the nerve terminal and impair the release of acetylcholine (Dal Belo et al., 2005). From the proteomic point of view, Micrurus venoms belong to two main groups depending on the predominant components, i.e., 3FTx-rich venoms and PLA 2 -rich venoms (Fernández et al., 2015;Lomonte et al., 2016Lomonte et al., , 2021. Although only a fraction of the total number of coralsnake species has been examined, the expression of these types of venoms might reflect the group's evolutionary history . The selective pressure that mediated this expression pattern is unknown, nor is it clear whether the appearance of PLA 2 -rich venoms has occurred as many independent events within the history of coralsnakes (Lomonte et al., 2021). Integrating more species into the review of venom proteomic profiles, including closely related species, could help elucidate these questions.
Micrurus alleni is a widely distributed species found from eastern Honduras to northwestern Panama (Campbell and Lamar, 2004;Roze, 1996). This is a terrestrial and primarily nocturnal snake that inhabits swamps, the vicinity of creeks and rivers, and is found often under leaf litter in primary and secondary forests (Solórzano, 2004).
Micrurus alleni was first described as a subspecies of Micrurus nigrocinctus by Schmidt (M. n. alleni) from specimens collected in Caribbean Nicaragua (Schmidt, 1936). A related form was soon after described by Dunn (1942) as M. n. yatesi, honoring Thomas Yates, who collected several specimens in Chiriquí, Panama. The status of these two forms and their distinction from M. nigrocinctus was quickly recognized by Taylor (1951) and further supported by Roze (1967) in his early revision of the genus. From these works, M. alleni was considered as a nominal species with at least two distinct allopatric populations: M. a. alleni, distributed from the Honduran Mosquitia to Caribbean Panama, and M. a. yatesi, distributed in the humid forests of the Central and South Pacific of Costa Rica and Chiriquí Province in Panama (Campbell and Lamar, 2004;Roze, 1996).
Although the distinction between these populations is not contested, the use of trinomial nomenclature to distinguish them has not always been accepted (Campbell and Lamar, 2004;Savage and Vial, 1974) and their taxonomic status is currently under review. Previous authors have suggested recognizing M. yatesi as a full species separated from M. alleni based on distinctive external characters (Campbell and Lamar, 2004;Solórzano, 2004) and divergences in molecular characters (Sasa and Smith, 2001). We follow this recommendation here.
The venom of M. alleni from the Caribbean versant of Costa Rica has been previously studied (Fernández et al., 2015) and showed a predominance of 3FTxs. The antivenom used to treat coralsnake envenomings in Central America, prepared against the venom of M. nigrocinctus (a phospholipase A 2 -predominant venom), was able to neutralize the lethality of M. alleni, albeit with a weaker potency (venom/antivenom proportion of 50 μg/mL to protect all mice) compared to the homologous venom (M. nigrocinctus, 300 μg/mL ratio to protect all mice) (Fernández et al., 2015). On the other hand, only few aspects from the venom of M. yatesi have been previously studied. An intravenous (i. v.) median lethal dose (LD 50 ) of 12.0 ± 2.8 μg/mouse (0.7 ± 0.16 μg/g) and an intraperitoneal (i.p.) LD 50 of 12.0 ± 1.8 μg/mouse (0.7 ± 0.11 μg/g) were reported by Bolaños (1972). Neurotoxic and phospholipase A 2 (PLA 2 ) activities were also described previously in M. yatesi venom (Rosso et al., 1996).
The aim of this work is to report the proteomic composition and toxicological characteristics of the venom of M. yatesi, as well as the immunological recognition and neutralization by the antivenom used in Central America to treat coralsnake envenomings.

Venoms and antivenom
Micrurus yatesi specimens were collected in the South Pacific region of Costa Rica and kept at the Laboratory for Dangerous Animals Research (LIAP) at Instituto Clodomiro Picado, Universidad de Costa Rica.
Venom was obtained from one adult specimen of M. yatesi (LIAP 001,

RP-HPLC and SDS-PAGE
Two mg of M. yatesi venom were dissolved in 200 μL of solution A (0.1% trifluoroacetic acid; TFA), centrifuged at 15,000×g for 5 min to remove debris and separated on a C18 column (250 × 4.6 mm, particle size: 5 μm; Teknokroma) using an Agilent 1200 chromatograph with 215 nm monitoring. Elution was performed with a 1 mL/min flow by applying a gradient of solution A (0.1% TFA) to solution B (0.1% TFA in acetronitrile) as follows: 0% B for 5 min, 0-15% B over 10 min, 15-45% B over 60 min, 45-70% B over 10 min and 70% B over 9 min. The venom fractions were collected manually and dried in a vacuum centrifuge (SpeedVac, Thermo). The fractions were redissolved in water, separated by SDS-PAGE in pre-cast 4-20% gels (Sigma-TruPage™) under reducing conditions and later stained with LabSafe GEL Blue™. The protein bands were cut from the gels and subjected to reduction (10 mM dithiothreitol), alkylation (50 mM iodacetamide) and an in-gel digestion with sequencing grade bovine trypsin overnight (in 25 mM ammonium bicarbonate) using an automated digestor (DigestPro MSi, Intavis). Resulting peptides were extracted with a solution of 0.1% TFA and 60% acetonitrile, and concentrated for mass spectrometry analysis.
To evaluate individual venom variation, venoms from 2 other specimens of M. yatesi (LIAP 094, LIAP 704) were also analyzed by RP-HPLC using the same conditions. The RP-HPLC profile of the venom of M. alleni was also obtained for comparison purposes.

MALDI-TOF/TOF and ESI-MS
Tryptic peptides were mixed with an equal volume of a saturated α-cyano-4-hydroxycinnamic acid matrix (α-CHCA; in 50% acetonitrile, 0.1% TFA). One μL of the mix was spotted onto Opti-TOF 384 plates and dried to later be analyzed in positive reflector mode using a Proteomics Analyzer 4800-Plus instrument (Sciex, Washington D.C., USA). Spectra were acquired using a laser intensity of 3000 and 500 shots/spectrum, using CalMix 5 (ABSciex) as external standards spotted on the same plate. Up to 10 precursor ions were chosen from each MS spectrum for automated collision-induced dissociation MS/MS spectra acquisition at 2 kV, in positive mode (500 shots/spectrum, laser intensity 3500). Resulting spectra were searched using the Paragon® algorithm of Pro-teinPilot v.4 software (Sciex) against the UniProt/SwissProt database for Serpentes, at a confidence level of ≥95%, for the assignment of proteins to known families. A few peptides with lower confidence scores were manually searched using BLAST (http://blast.ncbi.nlm.nih.gov), and their sequence was confirmed by manual interpretation of MS/MS spectra.
The monoisotopic mass of proteins from prominent RP-HPLC fractions was determined by direct infusion of the fractions (flow rate 5 μL/ min), dissolved in 50% acetonitrile and 0.1% formic acid, into a Q-Exactive Plus® mass spectrometer (Thermo Fisher Scientific, USA). MS spectra were acquired in positive mode, using 3.9 kV spray voltage, full MS scan range from 800 to 2500 m/z, and an AGC target of 3 × 10 6 ).

Venom protein family abundance
The relative abundance of each venom protein family was estimated by integration of the RP-HPLC peak signals at 215 nm, using Chem Station B.04.01 (Agilent, Santa Clara, California, USA). Densitometry was used for assigning percentual distribution for peak signals with two or more SDS-PAGE bands using Image Lab v.2.0 software (Bio-Rad, Hercules, California, USA). noyloxy-benzoic acid (4-NOBA, 1 mg/mL in acetonitrile) were added (Holzer and Mackessy, 1996). After a 60 min incubation at 37 • C, absorbance was determined at 405 nm by a microplate reader (Thermo). One unit of PLA 2 activity was defined as the change of 1 in absorbance.

In vitro
The assay was performed in triplicates.

Enzyme-linked immunosorbent assay (ELISA)
An ELISA was used to assess the ability of the anticoral antivenom produced at ICP to cross-recognize whole M. yatesi venom or its RP-HPLC fractions. M. yatesi, M. alleni and M. nigrocinctus venoms were dissolved in sodium phosphate buffer (PBS: 0.12 M NaCl, 0.04 M sodium phosphate, pH 7.2) and adsorbed onto a 96-well microplate (1 μg/100 μL/well) overnight at room temperature. After discarding the excess venom samples, wells were blocked by incubation with 100 μL of PBS that contained 3% bovine serum albumin (BSA) during 30 min. The microplates were washed five times with PBS. A volume of 100 μL of various dilutions (from 1:500 to 1:32000) of antivenom (in PBS with 3% BSA) were added to the microplates, which were incubated during 1 h at room temperature. As a negative control, a mock antivenom prepared from plasma from non-immunized horses was run in parallel under identical conditions on wells with M. yatesi venom. After washing the microplates five times with PBS, the antibodies bound to the venoms were detected by the addition of anti-horse IgG/alkaline phosphatase conjugate (Sigma; 1:4000 dilution in PBS with 3% BSA) for 1 h at room temperature, followed by five washes with PBS and the development of final color using p-nitrophenylphosphate (1 mg/mL in 0.1 M glycine, pH 10.4, with 1 mM MgCl 2 and 1 mM ZnCl 2 ). The absorbances were registered at 405 nm by a Multiskan microplate reader (Thermo Scientific). All samples were processed in triplicate wells. A similar procedure was used to evaluate the recognition of the most abundant fractions obtained by 14,15,17,18,20,21,22,23,and 26), using instead 0.4 μg/100 μL/well of each fraction and an antivenom dilution of 1:1000, in PBS-BSA for the binding step of equine antibodies.

In vivo venom activities
Animal experiments were performed following protocols approved by the Institutional Committee for the Care and Use of Laboratory Animals of the University of Costa Rica (CICUA permit 021-17), using CD-1 mice of either sex, bred at Instituto Clodomiro Picado.

Venom lethality
Various amounts of M. yatesi venom (from 3 to 23 μg) dissolved in 100 μL of PBS were injected intravenously (caudal vein) in groups of four mice (body weight between 16 and 18 g). The deaths were recorded after 24 h and the median lethal dose (LD 50 ) was calculated by Probit analysis using the BioStat 2008 Professional program (Finney, 1971).

Myotoxic activity
Groups of five mice (18-20 g) received an intramuscular injection of either M. yatesi or M. nigrocinctus venom (5 μg in 50 μL of PBS) in the right gastrocnemius muscle. A control group received an injection of 50 μL of PBS alone. After 3 h, blood samples were obtained from the tail of each mouse into heparinized capillary tubes. After centrifugation, 4 μL of plasma was used to determine the creatine kinase (CK) activity using a UV-kinetic assay (Wiener Lab, Argentina). CK activity was expressed in Units/L. Myotoxic activity was confirmed by extracting the injected gastrocnemius of mice 24 h after injection, subsequent to their euthanasia by carbon dioxide inhalation. Muscle tissue was fixed with formalin (3.7%) overnight and routinely processed for embedding in paraffin. Sections of 4 μm thickness were cut and stained with hematoxylin-eosin for histological observation.

Neutralization of lethality
The capacity of the SAC-ICP anticoral antivenom produced in ICP to neutralize the lethal activity of M. yatesi venom was assessed by injecting intravenously groups of five mice (16-18 g) with 200 μL of a solution that contained 30 μg of M. yatesi venom (equivalent to 3 × LD 50 ), which was previously mixed and incubated for 30 min at 37 • C with different dilutions of antivenom to obtain the following venom/ antivenom ratios: 100, 200 and 400 μg of venom/mL of antivenom. The control group of mice received the same dose of venom but was incubated with PBS instead of antivenom. Deaths were recorded after 24 h and the median effective dose (ED 50 ) was determined using Probit analysis (Finney, 1971).

Statistical analyses
The significance of differences between means of two groups was assessed by Student's t-test, or between means of three groups by ANOVA with post-hoc Tukey HSD. Differences were considered significant if p < 0.05.

Individual venom variability and comparison with M. alleni venom
Individual venoms of three M. yatesi specimens showed clear differences in the RP-HPLC profile, run under identical conditions (Fig. 3). However, prominent fractions of all M. yatesi venoms were located in a segment of the chromatogram characterized by the elution of PLA 2 s (35-55 min). The RP-HPLC profile of M. alleni venom, a 3FTx-rich venom, showed major fractions at the 20-30 min period (Fig. 4), in contrast with all M. yatesi individual venoms.

In vitro and in vivo activities of M. yatesi venom
Venoms of M. yatesi and M. nigrocinctus had similar PLA 2 activities in vitro (Fig. 5), which were higher than the activity of the venom of M. alleni. The intramuscular injection of M. yatesi venom in the gastrocnemius of mice significantly increased plasma CK activity, compared to controls injected only with the vehicle (Fig. 6). The in vivo myotoxic activity of M. yatesi venom was confirmed by histological analysis of injected muscles which showed widespread distribution of necrotic fibers characterized by hypercontraction and disorganization of the myofibrillar material, as well as edema (Fig. 6). The intravenous (i. v.) LD 50 of Micrurus yatesi venom in mice was 10.1 μg (95% confidence limits: 5.9-14.3 μg) per 16-18 g mouse, or 0.59 μg/g (95% confidence limits: 0.35-0.84 μg/g).

Immunorecognition and neutralization of M. yatesi venom by antivenom
The venom of M. yatesi was recognized by SAC-ICP antibodies with a similar ELISA signal to the one obtained with M. nigrocinctus venom, and a higher signal than the one obtained for M. alleni (Fig. 7). The immunorecognition by the antivenom of the most abundant RP-HPLC fractions of the venom was also assessed (Fig. 8). The fraction with the highest signal contained proteins from the metalloproteinase and serine proteinase families, while the least recognized major fraction contained a 3FTx. Fractions that contained PLA 2 s and Vespryn/Ohanin were recognized with a relatively moderate signal by the antivenom. The SAC-ICP antivenom was able to neutralize the lethal activity of M. yatesi venom with an ED 50 of 262 μg of venom/mL of antivenom (95% confidence limits: 187-419 μg/mL).

Discussion
Several factors including distribution in remote locations, low abundance, venom yield, and limited survival in captivity, have historically precluded a thorough analysis of the venom of M. yatesi. Very few specimens of this species have been kept at the serpentarium of ICP along the years. However, the recently collected venom from this species allowed the determination of the proteomic and toxicological characteristics, as well as the immunorecognition and neutralization by an antivenom.
A marked difference in venom composition was observed when the venom of M. yatesi was compared with that of M. alleni. The former has a predominance of toxins from the PLA 2 family while the latter has a predominance of toxins from the 3FTx family. This PLA 2 -3FTx dichotomy constitutes a general trend that has been observed in other coralsnake venoms (Fernández et al., 2015;Lomonte et al., 2016Lomonte et al., , 2021Lomonte et al., , 2016Sanz et al., 2016Sanz et al., , 2019a. Toxins from both protein families are able to exert neurotoxicity using different mechanisms. The 3FTxs compete with acetylcholine, blocking nicotinic cholinergic receptors at the motor end-plate (Moreira et al., 2010). On the other hand, toxins from the PLA 2 family impair the release of acetylcholine (Dal Belo et al., 2005) by hydrolyzing phospholipids of the nerve terminal plasma membrane. Differences in other less abundant components were also noted, since Kunitz-type inhibitors and serine proteinases were detected in M. yatesi but not in M. alleni, while nerve growth factor was reported only in the venom of M. alleni. Therefore, the venoms of these closely related species show significant variation in a number of venom protein families.
The toxicological analysis of the venom of M. yatesi estimated the intravenous LD 50 of this venom at 10.1 μg/mouse (0.59 μg/g), whereas the LD 50 of M. alleni was previously estimated in 6.3 μg/mouse (0.37 μg/ g) (Fernández et al., 2015). The 95% confidence limits of these determinations overlapped, thus indicating non-significant differences between the toxicity of these venoms. Previously, an i.v. LD 50 of 12.0 ± 2.8 μg/mouse (0.7 ± 0.16 μg/g) was reported for the venom of M. yatesi (Bolaños, 1972). The LD 50 values of M. alleni and M. yatesi venoms suggest that they are able to induce lethality in mice through different neurotoxic mechanisms based on the proteomic profiles, i.e., predominantly presynaptically in the case of M. yatesi and postsynaptically in the case of M. alleni, a hypothesis that deserves further pharmacological

Fig. 1. Micrurus yatesi specimen (A, photo by Andrés Vega) and elution profile of its venom proteins by RP-HPLC (B), followed by SDS-PAGE analysis (C). A C 18
column with an acetonitrile gradient was used for venom fractionation, as described in Methods. Further fractionation of proteins was performed using SDS-PAGE under reducing conditions. Molecular weight markers (M) are labeled in kDa. SDS-PAGE bands were excised, in-gel digested with trypsin, and analyzed by MALDI-TOF/TOF for protein family assignment, as shown in Table 1.   studies.
In agreement with proteomic results, the venom of M. yatesi displayed significant PLA 2 activity upon the synthetic substrate 4-NOBA, in a similar fashion as the PLA 2 -rich venom of M. nigrocinctus. The 3FTxrich venom of M. alleni, also in agreement with its composition, exhibited very low PLA 2 activity. When injected in CD-1 mice, the venom of M. yatesi exerted muscle damage, evidenced by the increase of plasma CK activity and by histological evaluation of injected muscles. Since M. yatesi is a PLA 2 -rich venom, in similarity to M. nigrocinctus (Fernández et al., 2011), such myotoxic activity was expected since PLA 2 s are the main myotoxic components in Micrurus venoms (Alape-Girón et al., 1999). The 3FTx-rich venom of M. alleni has been reported to induce a low (Fernández et al., 2015) to moderate (Gutiérrez et al., 1983) myotoxic effect. Mild myotoxicity has been described in  some human cases of envenomings by coral snakes (Bucaretchi et al., 2016), but this effect is not clinically significant. Experimentally, venoms of several species of coralsnakes have been shown to induce prominent myonecrosis in mice (de Roodt et al., 2012;Gutiérrez et al., 1983;Rey-Suárez et al., 2016).
When the immunorecognition of M. yatesi venom was evaluated using a commercial equine antivenom, prepared by the immunization of horses with the venom of M. nigrocinctus, the PLA 2 -rich venom of M. yatesi was recognized to a similar extent as M. nigrocinctus venom. In contrast, the 3FTx-rich venom of M. alleni was recognized to a lower extent. It has been previously noted that coralsnake venoms with PLA 2 predominance are better recognized and neutralized by this antivenom than 3FTx predominant venoms (Fernández et al., 2015). Thus, results are in line with the proposal that the compositional 3FTx/PLA 2 dichotomy of coralsnake venoms is linked with a divergence in their immunological characteristics .
The most abundant fractions of M. yatesi venom, which contained mostly toxins from the PLA 2 family, but also from vespryn, metalloproteinase, serine proteinase, and 3FTx families, were recognized by the SAC-ICP antivenom. The only exception was fraction 21, which contained a 3FTx and displayed a lower signal in the ELISA assay. Larger venom proteins, such as metalloproteinases, are generally better recognized by this coralsnake antivenom than proteins and peptides with a lower molecular mass, such as 3FTxs (Lippa et al., 2019;Rey--Suárez and Lomonte, 2020). This also explains why the fraction that contained a metalloproteinase and a serine proteinase showed the highest signal.
The SAC-ICP antivenom was able to neutralize the lethal activity of the venom of M. yatesi in a murine model. This preclinical assay predicts that, in case of envenomings by M. yatesi, treatment using this antivenom is likely to be effective. A lower neutralization capacity of the 3FTx predominant M. alleni venom was previously described using this antivenom (Fernández et al., 2015). Neutralization assays performed with PLA 2 -rich coralsnake venoms reveal an effective neutralization by this antivenom, while 3FTx-rich venoms are poorly neutralized, or in the case of the venom from M. mipartitus, not neutralized (Rey-Suárez et al., 2011).
The close relationship between M. yatesi and M. alleni has been pointed out in a previous analysis based on mitochondrial DNA (Sasa and Smith, 2001). Although there is scarce knowledge on the natural history of these two coralsnake species, they are likely to share similar ecological niches, and therefore the present findings on their contrasting venom proteomic profiles are intriguing. A handful of stomach records indicate that M. alleni often consumes swamp eels (Synbranchus marmoratus) and small fossorial colubrids (Solórzano, 2004). Less information is available on the diet of M. yatesi, but they have been seen preying on caecilians and small colubrids. Whether these observations reflect differences in the ecological contexts that allowed the selection of uneven venom patterns in different settings is unknown. However, the potential adaptive role of these venom types in immobilizing different prey species deserves further consideration.

Concluding remarks
The study of venom from M. yatesi allowed to determine its venom composition and to compare it with the venom from M. alleni. Toxicological analyses, in accordance with the proteomic profile, showed that this PLA 2 -predominant venom possessed significant PLA 2 activity in vitro and caused muscle damage in a murine model. The Micrurus antivenom prepared at Instituto Clodomiro Picado recognizes the different fractions of M. yatesi venom and neutralizes its lethal activity, hence implying that it is likely to be effective in envenoming by this species.

Ethical statement
Animal experiments were performed following protocols approved by the Institutional Committee for the Care and Use of Laboratory Animals of the University of Costa Rica (CICUA permit 021-17).

Declaration of interests
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Stephanie Chaves-Araya, Fabián Bonilla, Mahmood Sasa, José María Gutiérrez, Bruno Lomonte and Julián Fernández work at Instituto Clodomiro Picado, where the antivenom used in this study is produced.    , evaluated by ELISA. Venoms were adsorbed to microplates and incubated with various dilutions of antivenom or a mock antivenom prepared using normal horse serum. An anti-horse IgG/alkaline phosphatase conjugate was used to detect bound antibodies, as described in Methods. Each point represents mean ± SD of triplicates. *Differences among all means are statistically significant (p < 0.01) except when the means of M. yatesi and M. nigrocinctus are compared with each other (no statistical difference). **Differences among all means are statistically significant (p < 0.01 or p < 0.05) except when the means of M. yatesi and M. alleni are compared with each other or when the means of M. alleni and mock antivenom are compared (no statistical difference). Statistical analyses of the other two dilutions are not shown. Fig. 8. Immunorecognition of major RP-HPLC Micrurus yatesi venom fractions by a commercial equine antivenom prepared using M. nigrocinctus venom (SAC-ICP). An ELISA assay in which venom fractions were adsorbed onto microplates and bound antivenom antibodies were detected using antiequine immunoglobulins conjugated to alkaline phosphatase, followed by color development using pnitrophenylphosphate substrate was performed. A mock antivenom was used as a negative control. Each bar represents mean ± SD of triplicate wells. Colored circles above the bars indicate the protein family identified in each chromatographic fraction: Phospholipase A 2 (PLA 2 ), Vespryn/Ohanin, three-finger toxin (3FTx), metalloproteinase (MP), serine proteinase (SP).