First Insights into the Venom Composition of Two Ecuadorian Coral Snakes

Micrurus is a medically relevant genus of venomous snakes composed of 85 species. Bites caused by coral snakes are rare, but they are usually associated with very severe and life-threatening clinical manifestations. Ecuador is a highly biodiverse country with a complex natural environment, which is home to approximately 20% of identified Micrurus species. Additionally, it is on the list of Latin American countries with the highest number of snakebites. However, there is no local antivenom available against the Ecuadorian snake venoms, and the biochemistry of these venoms has been poorly explored. Only a limited number of samples collected in the country from the Viperidae family were recently characterised. Therefore, this study addressed the compositional patterns of two coral snake venoms from Ecuador, M. helleri and M. mipartitus, using venomics strategies, integrating sample fractionation, gel electrophoresis, and mass spectrometry. Chromatographic and electrophoretic profiles of these snake venoms revealed interspecific variability, which was ascertained by mass spectrometry. The two venoms followed the recently recognised dichotomic toxin expression trends displayed by Micrurus species: M. helleri venom contains a high proportion (72%) of phospholipase A2, whereas M. mipartitus venom is dominated by three-finger toxins (63%). A few additional protein families were also detected in these venoms. Overall, these results provide the first comprehensive views on the composition of two Ecuadorian coral snake venoms and expand the knowledge of Micrurus venom phenotypes. These findings open novel perspectives to further research the functional aspects of these biological cocktails of PLA2s and 3FTxs and stress the need for the preclinical evaluation of the currently used antivenoms for therapeutic purposes in Ecuador.


Introduction
Venoms are multi-component systems with the ability to recognise cellular and molecular targets, resulting in a complex set of alterations in tissue architecture and function that are associated with long-term complications and/or vital consequences [1]. The variable combination and abundance of proteins and peptides in these mixtures play a determining role in the mechanism of action and pattern of destabilising effects caused by the venoms [2,3]. The venom proteome variability has an impact on the toxin-neutralising capabilities of traditional and modern antivenoms [4,5]. Additionally, venom composition studies constitute a valuable platform for an improved understanding of evolutionary mechanisms [6] and their potential in drug discovery [7,8].

Collection of Venoms and Their Fractionation
Venom samples studied here were extracted from two female snakes found in the wild (Ecuador), one from each species (M. helleri and M. mipartitus). Figure 2A details the species, localities, and individual characteristics, and Figure 2B shows the occurrences of these two species in Ecuador based on biodiversity-related datasets available in (GBIF, http://www.gbif.org) and VertNet (http://www.vertnet.org) websites (accessed on 1 September 2022).

Collection of Venoms and Their Fractionation
Venom samples studied here were extracted from two female snakes found in the wild (Ecuador), one from each species (M. helleri and M. mipartitus). Figure 2A details the species, localities, and individual characteristics, and Figure 2B shows the occurrences of these two species in Ecuador based on biodiversity-related datasets available in (GBIF, http://www.gbif.org) and VertNet (http://www.vertnet.org) websites (accessed on 1 September 2022). The fractionation of crude venoms by reverse-phase high-performance liquid chromatography revealed different elution profiles, which suggest variability in the combination of toxins. Comparison of the chromatograms showed variations in the number, peak area, and retention time of proteins. In summary, 25 and 20 venom fractions were obtained from M. helleri ( Figure 3A) and M. mipartitus ( Figure 3B), respectively. The first venom profile showed highest peaks that emerged between 40 and 50 min, while the biochemical M. helleri (red) was extracted from a specimen found in Nangaritza (Zamora Chinchipe), while venom from M. mipartitus (blue) was obtained from a snake found in Quinindé, Jevon Forest Biological Station (Esmeraldas). (B) Spatial distribution of these two Micrurus species in Ecuador. This map was constructed using geospatial data from (GBIF, http://www.gbif.org) and VertNet (http://www. vertnet.org). Both biodiversity-related repositories were accessed on 18 September 2022. The fractionation of crude venoms by reverse-phase high-performance liquid chromatography revealed different elution profiles, which suggest variability in the combination of toxins. Comparison of the chromatograms showed variations in the number, peak area, and retention time of proteins. In summary, 25 and 20 venom fractions were obtained from M. helleri ( Figure 3A) and M. mipartitus ( Figure 3B), respectively. The first venom profile showed highest peaks that emerged between 40 and 50 min, while the biochemical separation of M. mipartitus venom presented major fractions approximately at 20 min. According to the absorbance values and peak areas, fraction 18, which eluted at a retention time of 42 min, had the highest relative abundance in M. helleri, representing~34% of the proteome. In the case of M. mipartitus, peak 6 had a relative abundance of~40% and eluted at approximately 20 min.

Biochemical Characterisation of Toxins
RP-HPLC fractions manually collected were subjected to SDS-PAGE under reducing conditions. The two crude venoms showed a clear predominance of low-molecular weight proteins (<20 KDa), mainly between 5 and 17 KDa ( Figure 4A,B). Nonetheless, higher molecular weight toxins, mainly of approximately 60 KDa, were observed in the last chromatographic fractions. M. helleri venom is characterised by intense bands around 14 KDa ( Figure 4A). However, the electrophoretic separation of the venom components from M. mipartitus showed a high abundance of bands of lower molecular weight, below 11 KDa ( Figure 4B).

Biochemical Characterisation of Toxins
RP-HPLC fractions manually collected were subjected to SDS-PAGE under reducing conditions. The two crude venoms showed a clear predominance of low-molecular weight proteins (<20 KDa), mainly between 5 and 17 KDa ( Figure 4A,B). Nonetheless, higher molecular weight toxins, mainly of approximately 60 KDa, were observed in the last chromatographic fractions. M. helleri venom is characterised by intense bands around 14 KDa ( Figure 4A). However, the electrophoretic separation of the venom components from M. mipartitus showed a high abundance of bands of lower molecular weight, below 11 KDa ( Figure 4B).

Protein Family Assignment by Mass Spectrometry
Bands excised from SDS-PAGE gels and specific chromatography fractions were subjected to proteolytic digestion to identify toxin families using a proteomics approach. The sequences obtained by tandem mass spectrometry were analysed with MASCOT software. Proteins belonging to six families were found in the venom of M. helleri ( Table 1). The first third of the chromatogram (fractions 2-11) was composed of peptides and 3FTxs. The central region showed the presence of PLA2s, which were the most abundant components from this sample. Hyaluronidase and L-amino acid oxidases were detected in the last fractions.

Protein Family Assignment by Mass Spectrometry
Bands excised from SDS-PAGE gels and specific chromatography fractions were subjected to proteolytic digestion to identify toxin families using a proteomics approach. The sequences obtained by tandem mass spectrometry were analysed with MASCOT software. Proteins belonging to six families were found in the venom of M. helleri ( Table 1). The first third of the chromatogram (fractions 2-11) was composed of peptides and 3FTxs. The central region showed the presence of PLA 2 s, which were the most abundant components from this sample. Hyaluronidase and L-amino acid oxidases were detected in the last fractions. The analysis of tryptic-digested peptides from the chromatographic fractions and electrophoretic bands of M. mipartitus venom revealed the presence of four toxins families, with 3FTxs as the most prominent group (Table 2). These low-molecular mass proteins were mainly identified in the first fractions, although some of them were also observed in the central region of the chromatogram, such as fractions 13-15. The last collected toxins correspond to snake venom metalloproteases (SVMPs) and LAOs.  Overall, the proteomic profiles of these two Ecuadorian coral snakes retain the hallmark proteins of Micrurus venom composition, where PLA 2 s or 3FTxs are the major toxins. In both venoms, the sum of the abundance of these two protein families represented more than 80% of the total proteome. Our comparative analysis exposed a 3FTx/PLA 2 venom dichotomy in the country ( Figure 5). M. helleri is a typical PLA 2 -predominant venom, with the PLA 2 isoforms corresponding to 72.1% of the sample composition ( Figure 5A). Conversely, the dominating biomolecules in M. mipartitus are 3FTxs, which constitute 63.4% of the venom proteome ( Figure 5B).

Discussion
Micrurus (coral snake) species belong to the New World elapids, a monophyle group of small, coloured, extremely venomous, and broadly-occurring snakes [34]. Th are considered ideal organisms for ecological, evolutionary, pharmacological, physiolo ical, phylogenetic, and proteomic studies [29]. Their ability to produce small quantities venoms has limited the manufacturing of antivenoms as well as structural, functional, a immunological research [35]. The first integrative venom proteomics using coral sna was developed in 2008 [36], but most of the quantitative estimation of proteome comp sitions were published in the last 10 years [27,30]. Even so, a significant number of M crurus venoms have remained inaccessible to studies [29].
The boom of proteomic approaches has led to a greater understanding of the b chemical arsenal of venoms from New World coral snakes [29,30,37]. The analysis of t distribution and abundance of toxin families has revealed two intriguing divergent co positional patterns: PLA2-rich and 3FTxs rich venoms [24]. In brief, a general trend mark by the polarization of venom phenotypes along Micrurus north-south dispersal has be observed [30]. The abundance of PLA2 is extremely high in venoms from coral snakes habiting the North, while in most southern species, this value is significantly reduced a compensated by a high expression of 3FTxs [25]. So far, proteomic analyses of coral ve oms are still limited, and the challenging 3FTx/PLA2 dichotomy puzzle in Micrurus nee further pieces of information to complete a comprehensive picture.

Discussion
Micrurus (coral snake) species belong to the New World elapids, a monophyletic group of small, coloured, extremely venomous, and broadly-occurring snakes [34]. They are considered ideal organisms for ecological, evolutionary, pharmacological, physiological, phylogenetic, and proteomic studies [29]. Their ability to produce small quantities of venoms has limited the manufacturing of antivenoms as well as structural, functional, and immunological research [35]. The first integrative venom proteomics using coral snake was developed in 2008 [36], but most of the quantitative estimation of proteome compositions were published in the last 10 years [27,30]. Even so, a significant number of Micrurus venoms have remained inaccessible to studies [29].
The boom of proteomic approaches has led to a greater understanding of the biochemical arsenal of venoms from New World coral snakes [29,30,37]. The analysis of the distribution and abundance of toxin families has revealed two intriguing divergent compositional patterns: PLA 2 -rich and 3FTxs rich venoms [24]. In brief, a general trend marked by the polarization of venom phenotypes along Micrurus north-south dispersal has been observed [30]. The abundance of PLA 2 is extremely high in venoms from coral snakes inhabiting the North, while in most southern species, this value is significantly reduced and compensated by a high expression of 3FTxs [25]. So far, proteomic analyses of coral venoms are still limited, and the challenging 3FTx/PLA 2 dichotomy puzzle in Micrurus needs further pieces of information to complete a comprehensive picture.
In Ecuador, the Micrurus genus includes 17 species, but their venoms have not been investigated. The country does not produce antivenom and depends on importing these life-saving immunological agents [19,31]. Bothropic and Lachesis bites are covered by the polyvalent antivenom imported from Instituto Clodomiro Picado (Costa Rica) with demonstrated effectiveness, despite the high number of vials needed in some cases [31,38]. However, envenomation by elapids remains a latent problem, despite the reduced number of cases. Most public hospitals lack antivenoms against elapid snakes. To our knowledge, there are no proteomic studies of Ecuadorian coral venoms, and the preclinical efficacy of antivenoms has not been evaluated. Altogether, these antecedents motivated the present proteomic characterization of venoms from Ecuadorian coral snakes.
The recognised venom variability of Micrurus species earlier described [29,39] was confirmed by the comparison of the chromatograms in our study. M. helleri venom separation showed prominent peaks in the central chromatogram region, a higher number of fractions, and a trend of PLA 2 -rich coral venoms observed in samples from the North [40].
In contrast, the HPLC separation using the C18 column of M. mipartitus venom evidenced that the toxins eluted mainly in the first fractions, similar to the 3FTxs-predominant venoms largely reported from venoms from South America [41].
Electrophoresis patterns are similar to the molecular signatures of SDS-PAGE gels of coral snake venoms previously studied, such as Micrurus clarki [42] and Micrurus nigrocinctus [39]. In contrast to the wider molecular weight range of proteins found in viperid snake venoms, the elapid venom samples here evaluated presented a large content of components migrating within the range of 5-17 KDa. These results agree with the typical molecular weight of the two most abundant biomolecules (PLA 2 and 3FTx) expressed in New World snake venoms [26,36].
Tandem mass spectrometry analysis of the tryptic-digested peptides of RP-HPLC fractions and/or gel bands against a database of venom toxins from Micrurus species revealed the existence of the two divergent patterns. Supporting the chromatographic and electrophoretic profiles, our findings demonstrated that M. helleri is a PLA 2 -predominant venom, while M. mipartitus exhibits higher content of 3FTxs. Therefore, in Ecuador, coral snake species with both proteomic profiles coexist, similar to Costa Rica [24], Colombia [43], and Brazil [28]. Previous studies have highlighted the implications of this venomics divergence in terms of the low cross-reactivity of elapid antivenoms across the two divergent phenotypes [20,44,45]. This fact draws attention to the need to guarantee the availability of antivenoms with wide taxonomic coverage, capable of neutralizing both venom phenotypes present in Ecuador.
Micrurus lemniscatus is a complex group of snakes with widely dispersed species (M. carvalloi, M. helleri, M. diutius, and M. frontifasciatus) and controversial taxonomy [46]. Different authors consider them distinct species or subspecies of M. lemniscatus. Some species can be also found in Colombia and Brazil [47]. The distribution of toxins in venoms from these countries was assessed in earlier studies [28], which unveiled some interesting aspects. According to the literature, the venomics profile from samples of M. lemniscatus carvalhoi (Brazil) showed 76.7 and 71.3% of 3FTxs [28], while M. helleri from Colombia has a PLA 2 -predominant phenotype, with 62.5% of the total proteome [28]. Our findings concur with the reported predominance of PLA 2 s in the venom of this subspecies, and indeed indicate higher levels of PLA 2 expression in the Ecuadorian venom of M. helleri (72.1%) compared to the venom proteome of M. helleri from Leticia (Colombia) [28].
The 3FTx predominance characterised in M. mipartitus from Ecuador (63.4% of venom composition) was also reported in a previous study of Colombian and Costa Rican samples [26]. These venoms were rich in 3FTxs: 61.1% and 83%, respectively. PLA 2 family is the second dominant toxin group found in the proteomic analysis of snake venoms from these three countries. These proteins are present in moderate proportions (19.1% of total venom toxins) in Ecuadorian M. mipartitus, while in samples from Colombia and Costa Rica, they account for 29.0% and 8.2% of venom proteome, respectively [26]. Additionally, Ecuadorian venom showed a higher relative abundance of SVMPs (6.0%) and LAOs (8.4%), in contrast to Colombian and Costa Rican venoms, which have a lower proportion of these catalytically active biomolecules [26]. The functional role of these toxins in the symptomatology of elapid envenomation remains unclear. The intriguing expression of haemorrhagic agents in coral snakes was also questioned in a transcriptomic study of M. fulvius [48]. Biological screening of the venom and clinical data reports are crucial for a better understanding of the contributions of these molecules to coral snakebites.
The pathogenesis induced by Micrurus venoms is characterised by neuromuscular paralysis [41]. Ecuadorian Micrurus snakebites have received little attention in the literature. To date, only one case report detailing the main manifestations following a M. helleri envenomation in Ecuador has been documented [33]. Neurotoxic effects and a mild increase in the levels of serological muscle damage biomarker (creatine kinase) were observed. These signs are probably associated with the action of PLA 2 s and 3FTxs, which account for approximately 90% of the total proteins of M. helleri venom. On other hand, we could not find published case reports of envenoming by Ecuadorian M. mipartitus. A case presentation from Colombia evidenced high neurotoxicity with progressive bilateral ptosis and respiratory paralysis [49]. The mechanism behind this characteristic clinical pattern involves the biological properties of pre-synaptically acting PLA 2 s and post-synaptic 3FTxs [29].
Similar to other investigations focused on Micrurus venoms, this study has some limitations. It represents an initial snapshot, since venoms were sampled from just one snake of each species, which precludes a representative picture of the total population and spatial distribution. Other earlier studies have also analysed Micrurus spp. venoms obtained from one specimen [26,30]. Additionally, the low venom volume obtained also hindered further functional assays and anti-venomics approaches.
In summary, new proteomic pieces were added to the current state-of-the-art information on the dichotomy of New World coral snake venoms. This study constitutes an important step toward characterizing the neglected Ecuadorian venoms and motivates further investigations with the other 15 species not yet assessed. Both contrasting venom phenotypes were evidenced in the country. M. helleri is a member of the group of PLA 2 -predominant types, and M. mipartitus is a clear example of a 3FTx-abundant venom phenotype. This finding represents the tip of the iceberg that needs to be uncovered in combination with anti-venomics approaches and preclinical evaluation of therapeutic strategies.

Venoms
The Micrurus venom samples were collected under the research permits MAE-DNB-CM-2015-0017 and MAE-DNB-CM-2019-0115 issued by the Ministry of the Environment of Ecuador. Venom was manually milked from one specimen for each species of snake: M. helleri and M. mipartitus. In this manuscript, we named M. helleri while taking into account the recent findings published by Hurtado-Gomes [46]. The geographical locations of individuals are detailed in Figure 2. Both animals were female adults. Venoms were dried inside a vacuum containing anhydrous calcium sulfate and stored at −20 • C until use.

Venom Fractionation by RP-HPLC
Decomplexation of Ecuadorian coral venoms was performed by RP-HPLC using a C18 according to a routine protocol used for the characterization of venoms from this genus [23,42]. We employed an HPLC equipped with a binary pump system (Waters 1525, Watertown, MA, USA), autosampler (Waters 2707, Watertown, MA, USA) and photodiode array detector (PDA) (Waters 991, Watertown, MA, USA). In the first step, 100-200 µg of venom samples were dissolved in 200 µL of 0.05% trifluoroacetic acid and 5% acetonitrile (v:v) in water. Then, each venom was centrifuged at 13,000× g for 10 min, and the supernatant was injected into a Phenomenex Jupiter C18 stationary phase 250 × 4.6 mm; 5 µm; 30 Å). Mobile phases were prepared as follows: solution A, 0.1% TFA (v/v) in distilled water; solution B, 0.1% TFA (v/v) in acetonitrile. The separation was performed at a constant flow rate of 1 mL/min for 100 min using a linear gradient: 5% of B solution, followed by 5-25% B, 25-45% B, 45-70% B, 70% B, and 45-5% B for 5, 10, 60, 10, 5, and 10 min. Fractions were manually collected based on the absorbance values, which were monitored at an absorbance of 215 nm. The relative abundance of purified molecules was quantified by integration of the peak areas using Empower software. Venom fractions were dried in a vacuum concentrator (Genevac, miVac Duo) for 24 h.

SDS-PAGE
RP-HPLC fractions were submitted to SDS-PAGE performed in a BIO-RAD Mini-PROTEAN Tetra with some modifications [19]. Dried fractions were dissolved in 45 µL of loading buffer (0.075 M Tris-HCl (pH 6.8); 10% (v/v) glycerol; 4% (m/v) SDS; 0.001% (m/v) bromophenol blue). The analysis was carried out under reducing conditions. Five µL of 1 M DL-dithiothreitol was added to the samples, which were heated to 95 • C in a thermoblock (Thermo Scientific, Waltham, MA, USA) for 5 min. Running buffer was prepared using (0.025 M Tris-HCl (pH 8.3); 0.192 M glycine; 0.1% (m/v) SDS). Twenty µL of samples were loaded onto a 5% stacking gel and 12.5% running gel. Five µL of a protein ladder (RunBlue Prestained TriColour) was also loaded onto the first well to estimate the molecular weight size of venom proteins. The electrophoretic separation started at 35 mA for 30 min, ending at 100 mA.

LC-MS/MS Analyses
Fractions obtained by RP-HPLC and protein bands visualised by SDS-PAGE were subjected to proteolytic digestion in similar manner as the protocol described by   [25]. The solutions were freshly prepared. In the case of electrophoresis gel, bands of interest were excised. Coomassie blue dye was extracted with 100 mM ammonium bicarbonate solution (AMBIC) and ACN (1:1) for 1 h. Proteins were reduced with DTT at 56 • C for 30 min. Alkylation was then carried out with iodoacetamide (IAA), in the dark for 20 min and followed by proteolysis with sequencing grade trypsin. AMBIC was added and incubated at 37 • C overnight. After this period, the digested peptides were extracted. The supernatant was concentrated in a vacuum centrifuge for approximately 40 min until almost dry. Finally, it was redissolved in 0.1% formic acid and stored at 4-8 • C until injection into the LC-ESI-QTOF system. The same procedure was carried out for RP-HPLC fractions when the toxin was not visible in the Coomassie blue-stained gel.

MS Data Processing
The resulting tryptic peptides were analysed using an ACQUITY ® Xevo G2-S QTof UPLC ® /MS system fitted with the LockSpray. Mobile phases A and B consisted of 0.1% formic acid in water and 0.1% formic acid in ACN, respectively. The flow rate was set to 0.3 mL/min. The mass spectrometer operated with a capillary voltage of 0.5 kV, cone voltage, of 40 V; 120 • C source temperature; 450 • C desolvation temperature; 30 L/h cone gas flow; and 900 L/h of desolvation gas flow.
The LC gradient went from 1% to 70% B in 30 min. MS data were acquired using the data-dependent acquisition (DDA) method using a mass scan in the m/z region from 400 to 1990 Da. The most abundant ions in the MS1 spectra with charges +2 and +3 were selected for fragmentation (MS2) by collision-induced dissociation (CID). Fragment spectra were interpreted using a licensed version of the MASCOT DISTILLER 2.7 program (https://www.matrixscience.com/distiller.html) against the UniProt/SwissProt database for Serpentes. We also created a database including the primary structure of toxins from Micrurus spp. and other genera of the Elapidae family. The MS/MS tolerance mass was set to 0.6 Da, similar to a previous study with Micrurus pyrrhocryptus venom [39]. Carbamidomethyl cysteine and oxidation of methionine were defined as fixed and variable modifications, respectively.

Conflicts of Interest:
The authors declare no conflict of interest.