Venomics of the Scorpion Tityus ocelote (Scorpiones, Buthidae): Understanding Venom Evolution in the Subgenus Archaeotityus

The subgenus Tityus (Archaeotityus) includes small variegated species considered until recently, a basal group within Tityus, one of the most diverse and medically relevant genera of scorpions in the New World. Archaeotityus species are distributed in the humid forests of Lower Central America and the Choco, Guiana Shield, and Amazonian regions. Due to their size and habits, Archaeotityus species are not usually considered dangerous, however, there are some clinical reports that show otherwise. To contribute to the toxinological knowledge of these poorly explored species, we characterized the venom of Tityus ocelote from three localities in Costa Rica. In addition, we assessed the evolutionary relationships of putative sodium channel-modulating peptides found in this species with those reported for other members of the genus, through a maximum likelihood phylogenetic analysis based on their amino acid sequences. We observed the presence of homologs of previously identified NaTx from the species Tityus (Archaeotityus) clathratus and some other putative Na+ and K+ channel modulating peptides related to the T. bahiensis group. In addition, we sequenced some peptides related to toxins present in the venom of the subgenus Atreus, such as those reported for T. obscurus, T. pachyurus, and the Costa Rican endemic species T. dedoslargos. Our phylogenetic analysis suggests that the venom of this Archaeotityus species is very complex and that some of the ion channel toxins expressed in it are related to distinct lineages within the genus Tityus, which could represent a plesiomorphic condition conserved in this group of scorpions of the New World.


Introduction
The systematics and taxonomy of scorpions have undergone substantial modifications in the last two decades with significant changes in the organization of families and genera and their historical relationships (Soleglad and Fet, 2003;Sharma et al. 2015;Suranse et al. 2017;Santibáñez-López et al. 2020;Stundlova et al. 2022). Phylogenetic reconstructions have also allowed us to advance in our understanding of the evolution of morphological characters and the toxins that make up their venoms (Santibáñez-López et al. 2018). These advances are more relevant in the case of groups that include medically important species, as is the case of the genus Tityus Koch 1836 in the Neotropical region (Lourenço, 2015;2018;Santibáñez-López et al. 2022). With more than 220 species recognized to date, Tityus is the most diverse genus and one of the most dangerous of the Order. It includes species distributed mainly in South America, although some inhabit Lower Central America and several Antillean islands (Moreno-Gonzalez, 2021).
Several studies show that the species that encompass Tityus probably share a common ancestor (Borges et al. 2020;Moreno-Gonzalez et al. 2021). In his pioneering work, Kraepelin (1911) established three groups to include the species of Tityus known to date, basing his proposal on the pectins' color, body measurements, and external morphology. Lourenço (2002) consolidated this initial classification using similar characters, identifying each group with the first species described: T. clathratus, T. asthenes, and T. bahiensis. Later, Lourenço (2006) reformulated the constitution of the genus, recognizing five subgenera: Tityus (Archaeotityus), Tityus (Atreus), and Tityus (Tityus), to respectively include the three groups of species mentioned above, and adding the subgenera Tityus (Brazilotityus), to include species distributed in the Brazilian forest canopy, and Tityus (Caribetityus), which groups the Antillean species. The relationships among these groups have not been rigorously evaluated since they include many species distributed in very vast regions.
The subgenus T. (Archaeotityus) (= T. clathratus species group) currently includes 24 small scorpions inhabiting the wet environments of Lower Central America and the Chocoan region, the Guyana Shield, and the Amazonian basin in South America (Lourenço, 1992;2002;2006). These spotted scorpions share several morphological features: small size (< 30 mm), yellowish background color covered with variegated brown spots, and a large pyramidal subaculear tubercle (Lourenço, 1992;Moreno-González et al. 2019); a combination of characters that have been considered plesiomorphic (Lourenço, 2005;2006). The discovery of Tityus apozonalli, a Miocene amber fossil preserved in Chiapas, Mexico, further supported this notion since this species has a basal phylogenetic position to other fossil buthid congeners and has a remarkable morphology similarity with existing members of the Archaeotityus group (Riquelme et al. 2015). Therefore, until recently, the general notion was that T. (Archaeotityus) possibly represented an ancient lineage that diverged early during the evolution of the genus (Lourenço, 1999;2002;. Nevertheless, recent phylogenetic molecular analyses challenged this assumption (Ojanguren-Affilastro et al. 2017;Moreno-González et al. 2021).
Only a few venoms of Archaeotityus species have been studied to date (see below), as members of this subgenus are not considered medically relevant (D´Suze et al. 2007;Albuquerque et al. 2009;de Oliveira et al. 2021). This appreciation probably derives from their small size, the minute amount of venom they could inject, and the fact that most species are forest dwellers with limited contact with human populations (Albuquerque et al. 2009;Moreno-González et al. 2019).
In T. (Archaeotityus) clathratus, two toxin transcripts belonging to the β-toxin family of Na+ channel modulating peptides, have been identified (Borges et al. 2012). Two other NaTxs have been isolated from the venom of T. (Archaeotityus) mattogrossensis, a species distributed in the Brazilian Cerrado ecosystem (de Oliveira et al. 2021). These four peptides share amino acid sequence homology with toxins from members of the subgenus T. (Tityus) (in particular, members of the T. bahiensis group sensu Lourenço, 2006Lourenço, , 2015 although they also show close identity with peptides reported in T. (Atreus) species (de Oliveira et al. 2021).
It is clear that members of the Archaeotityus subgenus can inflict dangerous stings, as evidenced by a case report with Tityus silvestris, a species that inhabits the Guiana Shield and northern Amazon Region (Monteiro et al. 2016). Envenomation by this species causes severe clinical manifestations in humans, similar to those reported for T. (Atreus) obscurus, including muscle spasms, ataxia, loss of balance, and dysarthria. Envenoming also causes clinical alterations produced by some species of the subgenus Tityus, such as the ones induced by T. serrulatus, T. bahiensis and T. stigmurus (Monteiro et al. 2016). These findings show that Archaeotityus species could represent a medical problem in some regions (Monteiro et al. 2019;Vasques-Gomez et al. 2020) and open the possibility that the venoms produced by species of this subgenus contain toxins as potent as those found in other species considered medically relevant. Also, it could be possible that Archaeotityus venom gland secretions would express toxins that originated before the separation of the different lineages within Tityus, potentially representing plesiomorphic characters. As would be expected, commercial antivenom against T. (Tityus) serrulatus could not neutralize the Atreus-characteristic symptomatology induced in the patient stung by T. silvestris (Monteiro et al. 2016;Borges et al. 2021).
T. (Archaeotityus) ocelote, is a small scorpion (body length < 35 mm) from Lower Central America, that has a yellowish to reddish-brown background coloration covered with a mottled pattern of darker spots (Francke and Stockwell, 1987;Víquez, 1999) (Fig. 1). The species is distributed in humid lowlands in the Caribbean and Pacific slopes of Costa Rica and the Bocas del Toro region of western Panama (Víquez, 1999;Montoya and Armas, 2002;Teruel and Cozijn, 2011). It is an arboreal species, usually found in the stilt roots of palms and in the buttresses of tropical trees, where they coexist and compete with other scorpion species (Blasco-Aróstegui et al. 2020). Despite being locally abundant, T. ocelote was described only 35 years ago from the Costa Rican Caribbean slope, reflecting the little attention given to this scorpion in the past. Its small size and arboreal habits result in slight contact with humans, although its sting can be excruciating, as confirmed by some personal communications.
In this work, we performed proteomic profiling of the venom of Tityus (Archaeotityus) ocelote Francke and Stockwell 1987. We aimed to contribute to understanding the evolution of venoms from Archaeotityus species and their relationships with the other Tityus subgenera.

Sampling and Venom Extraction
The Biodiversity Commission from the University of Costa Rica approved the procedures and protocol for this study (N° Page 3 of 17 2 293-2021). We collected specimens of T. ocelote in the secondary forest cover of three locations in Costa Rica: Carara National Park (9°46′53.04″N, 84°36′16.09″W) and Hacienda Barú (9°16′17.00″N, 83°52′52.27″W), both in Puntarenas Province, and Pueblo Nuevo de Sarapiquí (10°28′26.11″N, 84°4′29.54″W), Heredia Province. This last location is near (8.0 km) to the type locality for the species and is isolated from the other two sites by the central mountainous axis crossing the country (see map in Fig. 1).
The scorpions were transferred to the Instituto Clodomiro Picado in San José, Costa Rica, and kept individually in plastic boxes with litter as a substrate. Mealworms and crickets were offered as prey weekly, and a soaked cotton swab provided water ad libitum.
Briefly, we extracted venom by electro-stimulating the telson and collected it into capillary tubes. The procedure involved placing the telson between two electrodes and applying a 40 to 50 V (0.7 to 0.8 A) discharge. Venom samples were pooled in a 1.5 mL plastic tube, lyophilized, and stored at − 70 °C until the different analyses were performed. The venom extracted from adult males and females (around 40 individuals) was combined in the same pool.

Gelatin and Hyaluronic acid Electrophoresis and Zymography
The venom (20 µg) was separated by SDS-PAGE (15%) under reducing conditions (5% 2-mercaptoethanol, at 100 °C for 10 min) in a Mini-Protean system (Bio-Rad) at 180 V, and the proteins were stained with Coomassie blue R-250. The same procedure was followed to separate the compounds found in the RP-HPLC fractions (see below).
We determined hyaluronidase activity in the venom using the method proposed by Cevallos et al. (1992)  on SDS-PAGE on a 12% gel containing 0.5 mg/mL cockscomb hyaluronic acid (Sigma Chemical Co.). After washing for one hour with 1% Triton X-100, the gels were incubated at 37 °C for 16 h in 0.1 M NaCl, 0.1 M sodium phosphate buffer, pH 6.6. Carbohydrates were stained with Alcian Blue 8GX (Sigma Chemical Co.).
We performed zymography to determine the gelatinolytic activity in the venom. Briefly, unreduced venoms (15 µg) were subjected to SDS-PAGE on 12% gels containing type A gelatin (Sigma Chemical Co.) at a 0.25 mg/ml concentration. After washing for one hour with 1% Triton X-100, the gels were incubated at 37 °C for 16 h in 50 mM Tris-HCl buffer, pH 8.0, containing 5 mM CaCl2, and stained with Coomassie. Blue R-250.
We determined phospholipase A 2 (PLA 2 ) activity on micellar phosphatidylcholine (Sigma Chemical Co.) using the phenol red colorimetric method (de Araujo and Radvanyi, 1987). Twenty µg of venom were added to 1 mL of substrate (0.25% w/v sn-3-phosphatidylcholine, 0.4% v/v Triton X-100, 0.004% w/v phenol red) at 30 °C, and the change in absorbance at 558 nm was monitored for 1 min. We used a purified PLA 2 from Bothrops asper snake venom as a positive control. PLA 2 activity was also assayed using the synthetic monodisperse substrate 4-nitro-3-octanoyloxybenzoic acid (4-NOBA) following previously described methods (Mora-Obando et al. 2014). We also tested this enzymatic activity using egg yolk and erythrocyte-containing agarose plates, according to Gutiérrez et al. (1986).
Human erythrocytes were obtained by centrifugation of citrated blood, washed three times with phosphate-buffered saline (PBS), and used for the direct hemolytic assay. We tested various venom concentrations on a 10% erythrocyte suspension for 30 min at 37 °C and read absorbances at 545 nm after centrifugation.

RP-HPLC
Four mg of venoms were dissolved in 200 µL of 0.1% trifluoroacetic acid (TFA; solution A), centrifuged for 5 min at 10,000 g to remove insoluble material, and separated by reverse-phase HPLC on a C18 column (250 × 4.6 mm, 5 µm particle size; Aqua Phenomenex) using an Agilent 1220 chromatograph with monitoring at 215 nm. Elution was performed at 1 mL/min by applying the following gradient with solution B (acetonitrile, containing 0.1% TFA): 0-60% B over 60 min.

Mass Spectrometry (MS)
Major protein bands selected from RP-HPLC venom fractions were excised from gels and subjected to reduction with dithiothreitol (10 mM) and alkylation with iodoacetamide (50 mM), followed by overnight in-gel digestion with sequencing grade bovine trypsin (Sigma Chemical Co.) or chymotrypsin (G-Biosciences). Some RP-HPLC fractions were directly reduced, alkylated, and digested in solution. The resulting proteolytic peptides were analyzed by MALDI-TOF-TOF on an Applied Biosystems instrument (ABSciex 4800 Plus). Mixtures of 0.5 µL of α-cyano-4hydroxycinnamic acid and 0.5 µL of each sample were spotted onto an Opti-TOF 384 plate, dried, and analyzed in positive reflector mode. Spectra were acquired after 500 shots at a laser intensity of 3000, using as external standards Cal-Mix-5 (ABSciex) spotted on the same plate. Up to 10 precursor ions from each MS spectrum were selected for automatic collision-induced dissociation MS/MS spectra acquisition at 2 kV, in positive mode (500 shots/spectrum, laser intensity of 3900). The resulting spectra were searched using Protein-Pilot v.4 and the Paragon® algorithm (ABSciex) against the UniProt/SwissProt Scorpiones database (Uniprot Scorpions, June 6, 2022; 13,242 entries), at a confidence level of ≥ 95%, for protein identification or protein family assignment by similarity. Few fragmentation spectra of sufficient quality were also manually interpreted to obtain de novo sequences or sequence tags. To find the most similar proteins in those databases, we performed a matching search for the obtained sequences using BLAST (https:// blast. ncbi. nlm. nih. gov).
To identify the new Na+ channel toxins found in this venom, we used the nomenclature proposed by Becerril et al. (1996) and followed by Guerrero-Vargas et al. (2012). Accordingly, the name assigned to each toxin would come from an abbreviation that combines the first letter of the genus in capital letters followed by the first two letters of the specific epithet in the lower case.
The whole venom of T. ocelote was analyzed by a "shotgun" MS approach. A venom sample of 15 μg was reduced in solution with 10 mM dithiothreitol for 30 min at 56 °C, alkylated with 50 mM iodoacetamide for 20 min in the dark, and digested with sequencing grade trypsin at 37 °C overnight, in a total volume of 40 μL. After adding 0.5 μL formic acid, the resulting tryptic peptide mixture was centrifuged and subjected to RP-HPLC on a nano-Easy 1200® chromatograph (Thermo) in-line with a Q-Exactive Plus® mass spectrometer (Thermo). Ten μL of peptide mixture, containing 0.7 μg, were loaded on a C18 trap column (75 μm × 2 cm, 3 μm particle; PepMap, Thermo), washed with 0.1% formic acid (solution A), and separated at 200 nL/min on a C18 Easyspray® column (75 μm × 15 cm, 3 μm particle; Thermo). A gradient toward solution B (80% acetonitrile, 0.1% formic acid) was developed in a total of 120 min (1-5% B in 1 min, 5-26% B in 84 min, 26-80% B in 30 min, 80-99% B in 1 min, and 99% B for 4 min). MS spectra were acquired in positive mode at 2.0 kV, with a capillary temperature of 200 °C, using 1 μscan in the range 400-1600 m/z, maximum injection time of 50 ms, AGC target of 1 × 10 6 , and resolution of 70,000.
Page 5 of 17 2 The top 10 ions with 2-5 positive charges were fragmented with an AGC target of 3 × 10 6 , minimum AGC 2 × 10 3 , maximum injection time 110 ms, dynamic exclusion time 5 s, and resolution 17,500. MS/MS spectra were processed against protein sequences contained in the UniProt/SwissProt Scorpiones database using Peaks X® (Bioinformatics Solutions), and matches were assigned to known protein families by similarity. Cysteine carbamidomethylation was set as a fixed modification, while deamidation of asparagine or glutamine and methionine oxidation were set as variable modifications, allowing up to 3 missed cleavages by trypsin. Parameters for match acceptance were set to FDR < 0.1%, detection of at least one unique peptide, and − 10lgP protein score ≥ 30.

Sequence Alignment and Phylogenetic Reconstruction of Toxins´ Tree
To determine the correspondence between toxins found in T. ocelote venom and those reported in databases for other species of the genus, we compared the putative mature amino acid sequences retrieved in our analysis with those deposited in the UniProt database http:// www. unipr ot. org/. Sequence alignment was performed in MEGA 5, using manual adjustment and multiple alignments with the MUSCLE algorithm. The matrix of aligned sequences was used to build the phylogenetic analysis, implementing it in IQTREE (version 2) under the maximum likelihood (ML) criterion (Minh et al. 2020). Before selecting the phylogenetic tree, we determined the best-fit substitution model using the ModelFinder model selection procedure implemented in IQTREE (under the BIC value criterion (Kalyaanamoorthy et al. 2017). Ultrafast bootstrap values were calculated in IQTREE after 1000 replicates. We edited the tree files with Figtree v1.4.4 (https:// github. com/ ramba ut/ figtr ee/).
For the phylogenetic reconstruction of the sodium channel toxins, in addition to the Tityus sequences, we incorporated the sequence of the major toxic components of the buthids Parabuthus tansvaalicus (Birtoxina, UniProtKB P58752, Inceoglu et al. 2001) and Centruroides tecomanus (Ct16, Uni-ProtKB PUD0I1, Martin et al. 1988) as outgroups (Borges and Graham, 2016).

General Venom Composition and Enzymatic Activities
Tityus ocelote venom (pooled from adult males and females) from the three populations analyzed (Fig. 1), presented similar physical characteristics with comparable electrophoretic patterns, although some slight variations were noted in terms of band intensities. In all venoms, the electrophoretic pattern showed the presence of small and medium-sized proteins and some poorly-separated thick bands containing peptides between 4 and 13 kDa (Figs. 2A and 3).
In contrast, gelatinolytic activity evaluated by zymography showed some geographical differences in the molecular masses of the expressed proteases (Fig. 2B). However, the activity was very weak and difficult to measure using this technique. We also detected similar hyaluronidase activity in the venom of all tested populations (Fig. 2C). Conversely, we did not detect direct hemolytic nor PLA 2 activities in the venom of any of the studied populations (results not shown).

Proteomic Profiling
At least 30 fractions were recovered from T. ocelote (Carara) venom after separation by RP-HPLC (Fig. 3). Most venom components were peptides of molecular weights below 10 kDa, while medium and large-sized proteins eluted at the end of the RP-HPLC chromatogram, after ~ 42 min. The complexity of the venom makes it very difficult to separate individual components, as shown in the SDS-PAGE analysis of the chromatographic fractions (Fig. 3). We were able to identify at least 41 different proteins following shotgun-MS analysis. These include metalloproteinases, serine proteases, CRISPs, SCP domain-containing proteins, hyaluronidases, and monooxygenases. We also found sequences matching putative phosphodiesterases (acid ceramidase), phospholipase A 2 , glycosidases, and nucleotidases (Table 1).

Identification and Evolutionary Relationships of Venom Ion Channel Modulating Peptides
Most peptides in T. ocelote venom matched putative Na+ and K+ channel toxins (Figs. 4,5,6,7). The phylogenetic relationships of the eight putative Na+ -toxins found in this venom were evaluated by maximum likelihood phylogenetic reconstruction based on their complete or partial amino acid sequences. For this reconstruction, we also included 72 sequences of Na+ -channel modulating toxins reported for 21 other species within the genus (see Supplementary  Table 1 for UniProt accession numbers). About 80% of the sites were informative. The recovered tree was based on a selected WAG + I + G substitution model (Log-likelihood: -3717.831, BIC score 8233.199). The consensus tree, constructed from 1000 bootstrap trees, shows 18 clusters with reasonably good support (Ultrafast bootstrap values > 75) (Fig. 8). Putative sodium channel toxic peptides expressed in T. ocelote venom have homology with T. (Atreus) or T. (Tityus) neuropeptides and occupy different positions in the tree. Clades that do not include T. ocelote toxins also show segregation between toxins from those subgenera. We did not obtain good support for the internal nodes of our phylogenetic tree, using neither the maximum likelihood criterion nor under a Bayesian analysis (data not shown), precluding inferences about relationships between clusters.
We identified two homologs of the Na+ channel-modulating toxins Tcl1 and Tcl2 from Tityus (Archaeotityus) clathratus (Borges et al. 2012) (Fig. 4). We call these peptides Toc1 and Toc2, and although they share 79% sequence identity between them, they belong to two separated clusters in our phylogenetic tree (Fig. 8). Both homologs also show similarities to β-toxins found in members of the T. (Tityus) bahiensis group (Fig. 4). Toc1 sequence has 97% identity to Tcl1 and shares 82% identity with Ts1 from T. serrulatus and Tf1 from T. fasciolatus, and 79% identity with Tb1, Tst1, and Tco1 from T. bahiensis, T stigmurus, and T. costatus, respectively. Likewise, Toc2 has 87% sequence identity to Tcl2 and also shows homology (> 80% identity) to the same sodium channel modulating β-toxins mentioned above (Fig. 4). These toxins have a marked specificity for mammals and are included in the NaTx6 group of Guerrero-Vargas et al. (2012).
We also retrieved the complete sequences of two other peptides from T. ocelote venom, Toc3 and Toc4. They are homologs to Na+ -channel toxins found in representatives of the subgenus Tityus (Atreus) (Fig. 4). Each of these two peptides shows variants with a few minor amino  acid substitutions, possibly representing isoforms of the same toxin (data not shown). Toc3 is related to To6 and To7 from T. obscurus, with sequences sharing 87% identity. It is also related to TdNa9 (84%) and TdNa10 (75%), expressed in T. discrepans venom (Fig. 4). These toxins present sequence similarity to α-NaScTxs and are included in the NaTx7 cluster of Guerrero-Vargas et al. (2012). On the other hand, Toc4 shares 95% and 90% identity with T.
obscurus toxins To11 and To3. It also shares more than 92% identity with the sequences of Tfe2, Tfal2, and Tz1 toxins identified in T. festae, T. falconensis, and T. zulianus, respectively. Other toxins that show similarity to Toc4 are Td4, Td5, and Bactridin-2, isolated from T. discrepans, and Tde1 isolated from T. dedoslargos, an endemic species from Costa Rica Borges et al. 2020). These toxins show a similar mechanism of action to α-NaScTxs but have We also obtained a complete sequence of the NaTx homolog to T. obscurus To15 (84% identity), which is also 100% identical to TpaCR2, a toxin identified in the transcriptome of the Costa Rican scorpion T. pachyurus (currently classified as T. jaimei) by Borges et al. (2020) (Fig. 4). We named it Toc5 (Fig. 4). In addition, this peptide is related to the toxins of T. zulianus (Tz2, 74%), T. trinitatus (Ttr5, 72%), and T discrepans (TdNa6, 72%). These species are distributed in northern South America and classified within the subgenus T. (Atreus) (Lourenço, 2006). Therefore, Toc5 has a relationship with the NaTx1 cluster (Guerrero-Vargas et al. 2012), including peptides considered insect-specific β-toxins.
We also identified and sequenced a potential α-NaTx, which we named Toc8 (Fig. 4), which is a peptide with high homology to Ts3 and Ts17 from T. serrulatus and Tb3 from  Moreover, three partial sequences corresponding to putative NaTx peptides were also obtained (Fig. 5). One was named Toc6, and its N-terminal 30 amino acid residues were identified. This sequence shows > 84% identity with other β-NaTx found in scorpions of the subgenus Atreus: To4 (T. obscurus), TpaP7, and Tpa7 (T. pachyurus from Panama and Colombia, respectively) and TdNa7, Td3 and Td11 (T. discrepans). These β-toxins make up the NaTx13 cluster (Guerrero-Vargas et al. 2012) and are specific for insects.
In addition, we were also able to identify two peptides whose partial sequences show homology to Na+ channel modulating toxins found in T. serrulatus, and associated with α-subunits of lipolysis-activating peptides Ts28 and Ts41 (Fig. 6) (Kalapothakis et al. 2021). We named them Toc9 and Toc10 and even when it is suggested they probably act on Na+ channels, they were not included in the phylogenetic analysis.
Interestingly, all the above-mentioned putative Na+ channel toxins were identified in the venoms of T. ocelote from the two Pacific regions of Costa Rica. In the case of venom from Sarapiquí specimens (Caribbean region), we could only identify peptides with homology to the T. bahiensis group (not from Atreus subgenus), together with the two Archaeotityuspreviously identified toxins. In this venom, we identified the presence of Toc1, Toc2, Toc8 and Toc9.
Besides Na+ channel toxins, in T. ocelote venom we found two sequences matching putative K+ channel-modulating toxins. We call these peptides TocKTx1 and TocKTx2 (Fig. 7), and they are homologs to the β-KTx that have been found in T. discrepans venom, as well as other species of both, the T. (Atreus) and T. (Tityus) groups. The sequences of these putative toxins are not similar to each other (35%  . TocKTx1 sequence has 54 amino acid residues and shows great similarity to ToKTx from T. obscurus (UniProt A0A1E1WVV4; 96% identity) and TdiKIK (Q0GY43, 92%) from T. discrepans. It also shows correspondence with members of the T. (Tityus) group: T. serrulatus (Ts19, P86822; 77%), T. stigmurus (TstKMK, P0C8W4; 76%), and T. bahiensis (TX, 75%) venom toxins. Likewise, TocKTx2 sequence resulted in 48 amino acid residues and complete similarity with a KTx from T. obscurus (ToKTX, A0A1E1WVU0). It also matches with TdiβKTX (Q0GY44, 81% identity) from T. discrepans, as well as with toxins from several members of the T. (Tityus) clade (Fig. 7). Both putative K+ channel toxins were found in the venom of T. ocelote specimens of the Central Pacific and Caribbean. Several small peptides corresponding to other putative KTx were also identified, which match with Tc30 and To32 from T. obscurus and Ts6 and Ts7 from T. serrulatus, among others (results not shown).

Identification of Other Venom Peptides
Within the group of Cys-containing peptides that do not modulate ion channels, in T. ocelote venom we identified a potential protease inhibitor with the partial sequence CKTY-DDCKDVCKAR. This peptide is identical to the protease inhibitor TopI1 from T. obscurus venom .
Regarding the presence of non-containing disulfide bridges peptides (NDBP), a partial sequence of a putative hypotensin (KIKETNAKPPAR) was retrieved, with similarity to Ts14 (former TsHpT-I) from T. serrulatus (Verano-Braga et al. 2008;, and a partial sequence of a putative antimicrobial peptide, FFSLI(L)PSLIGGLVSAIK, was also identified. Also, a partial sequence of a putative bradykininpotentiating peptide (EMLKDYANR) present in T. ocelote venom is similar to venom peptides from the Old World scorpions.

Venom Composition
Scorpions belonging to the subgenus Archaeotityus have been portrayed as being less relevant from a medical perspective (Albuquerque et al. 2009;Borges et al. 2012). However, our results show that T. ocelote venom is as diverse and has a composition comparable to that found in other Tityus species. It combines a variety of proteins and a group of cysteine-containing peptides that potentially act as modulators of ion channels, inducing neurotoxic effects (Guerrero-Vargas et al. 2012;de Oliveira et al. 2015;2018;Ortiz and Possani, 2018;Kalapothakis et al. 2021). It also has particularities that make it worth mentioning.
Enzymatic effects were almost exclusively associated with proteolytic and hyaluronidase activities. Proteinases are the leading group of medium-sized proteins found in Tityus species (Ortiz et al. 2014). Among them, serine proteases are frequently present, but in a low proportion, in the toxic secretions of these and other buthid scorpions . This situation contrasts with their predominant presence in the venom of non-buthid species (Carmo et al. 2014;Lok So et al. 2021).
The presence of metalloproteinases in T. ocelote venom confirms their ubiquity in Buthidae secretions, as well as in other scorpion families (Cid-Uribe et al. 2019). Multiple paralogs are expressed in different Tityus species (Carmo et al. 2014;Ortiz et al. 2014;de Oliveira et al. 2015;2018). Despite representation, the proteolytic effect attributable to these enzymes on gelatin was weak in T. ocelote venom. However, this could be the result of the substrate commonly utilized for testing the activity, since invertebrate venom proteolytic enzymes could probably be related to toxin processing and their target would not be necessarily the extracellular matrix, as in the case of vertebrate venom proteases (Langenegger et al. 2018). In any case, the low proteolytic activity of Tityus venoms has been previously reported (Carmo et al. 2014).
Scorpion venom metalloproteinases are relatively shorter and do not have the multiple cysteine-rich domains found in snake venoms, and it has been postulated that this type of proteinase evolved from an ancestor of the ADAM Arachnida type, which lost the disintegrin and cysteine-rich domains (Carmo et al. 2014;de Oliveira et al. 2018).
Although we could not detect any enzymatic activity, partial sequences identified as belonging to PLA 2 s were expressed at the protein level in T ocelote. The consistent records of PLA 2 in the venom gland transcriptome of several scorpion species (Rodríguez de la Vega et al. 2010;Ma et al. 2012) suggest that these enzymes are also ubiquitous in the different scorpion lineages, although in several members of the Tityus genus they represent less than 1% of the transcripts (de . Thus, most of these phospholipases may have lost their enzymatic activity or the remaining scaffolds could have developed new functions associated with, for instance, the induction of inflammation with reactions that often accompany envenomings induced by scorpions and other arthropods (Lok So et al. 2021).
As in all members of the Buthidae family, cysteine-containing peptides dominate in T. ocelote venom. These molecules are responsible for the neurotoxic effects commonly associated with the sting of buthids, modifying the function of ion channels in excitable and non-excitable tissues (Lok So et al. 2021). A diversity of peptides affects different ion channels, but those that modulate sodium (NaTxs) and potassium (KTx) channels, prevail in Buthidae (Lok So et al. 2021).

Page 1of 17
At least six of the β-NaTx type and two of the α-NaTx type, are displayed in T. ocelote venom. These two types of sodium-channel modulators differ in their effects on membrane potentials and channel binding sites (Cid-Uribe et al. 2020). There is a differential distribution of these NaTxtypes in Buthidae, with α-NaTxs predominating in the venom of Old World species, while the β-NaTx prevailing in the New World forms (Cid-Uribe et al. 2020). β-NaTxs are remarkably diverse in Tityus and other American buthids and probably precede the appearance of the α-NaTxs, whose origin could have occurred before the separation of South America from Africa (Guerrero-Vargas et al. 2012). The loss and subsequent modifications of the genes that code for these peptides would explain the existence, and lower prevalence, of α-toxins in American buthids, including in T. ocelote. Zhu and Gao (2006) suggested that the ancestor of the NaTxs was probably a β-toxin, a variant of a lipolysis activating peptide (LAP) α-subunit, which displays 7 cysteines instead of 8, contrary to most of the NaTxs. Interestingly, two peptides with homology to LAP were present in T. ocelote venom, a finding that, in the case of New World scorpions, has only been reported for species from the T. (Tityus) subgenus and Centruroides (Zhu and Gao, 2006;de Oliveira et al. 2018;Kalapothakis et al. 2021). On the other hand, the presence of potassium channel toxins in T. ocelote venom is expected, considering that transcripts of these peptides have been found in every studied scorpion species to date, including non-buthid families (Zhu et al. 2011;Cid-Uribe et al. 2020).
In terms of the Na+ channel modulating peptides, the proteomic analysis of T. ocelote venoms from three populations shows some geographic variation between Sarapiquí and the other two locations. Conversely, the profiles of Carara and Barú, both on the Central Pacific coast of the country, did not register significant variation. In general terms, T. ocelote from Sarapiquí only showed Tityus subgenus Na+ channel modulating homologs in its venoms, whereas specimens from the Costa Rican Central Pacific coast showed peptides with similarity to both subgenera, Atreus and Tityus. The same two main K+ modulating toxins identified in this study appear in all the analyzed venoms.
The intraspecific divergence in venoms between populations on both sides of the Costa Rican central mountain axis has been recorded in vipers (Alape-Girón et al. 2008;Madrigal et al. 2012), elapids (Mena et al. 2022) and dartfrogs (Mebs et al. 2014). In these cases, quantitative differences in venom components could translate into variation at the level of their pathophysiological effects (Gutiérrez et al. 1980). The final uplift of the central mountainous axis, particularly the Cordillera de Talamanca, during the Pliocene (MacMillan et al. 2004) constitutes one of the most recognized cladogenic events that shaped the biogeography of the region (Daza et al. 2010). This final uplift promoted the population structure and restricted gene flow between populations that, like T. ocelote, are distributed in the humid lowlands of Lower Central America (Wang et al. 2008;Saldarriaga-Córdoba et al. 2017).

Evolutionary Significance of T. ocelote Venom Composition
Until recently, subgenera of Tityus were considered natural groups (Lourenço 2006). Furthermore, the basal position of members of the subgenus Archaeotityus had been assumed by most researchers mainly based on a few putative plesiomorphic morphological characters (Lourenço 1999;2002;Borges et al. 2010). However, this view is changing due to new phylogenetic evidence based on the analysis of molecular characters at various substitution rates (Borges et al. 2010;Ojanguren-Afilastro et al. 2017;Román et al. 2018;Moreno-González et al. 2021).
In independent studies, Ojanguren-Afilastro et al. (2017) and Moreno-González et al. (2021) presented their hypotheses of the evolutionary relationships among species that make up the most influential groups within Tityus. The most important point for our discussion, of these two studies, is the finding that T. (Archaeotityus) turns out to be the sister clade of a subgroup of Tityus (Tityus), composed of the T. bahiensis, T. stigmurus, and T. trivitattus species complex (the "T. (Tityus) bahiensis clade," sensu Ojanguren-Affilastro et al. 2017). The analyses of Borges and Graham (2016) and Román et al. (2018) also support these findings, although none of the presented reconstructions show high support for this branch.
According to Ojanguren-Affilastro et al. (2017), deep divergence within Tityus occurred about 30 Mya ago (95% CI 22.9-37.6), rendering the mentioned major division within the genus, while the T. (Archaeotityus) and the T. (Tityus) bahiensis clades might have split from a common ancestor around 24.2 Mya (95% CI 16.6-30.6) during the Oligocene. Following this proposition, members of the T. (Tityus) bahiensis clade and those included in T (Archaeotityus) would exhibit closer relationships in gene expression with each other than with other Tityus clades. In support of this prediction, Borges et al. (2012)  identity with the toxins expressed in those Brazilian scorpions and that -in the case of the first two-also share significant sequence similarities to the T. (Archaeotityus) toxins previously reported.
The aforementioned phylogenetic hypothesis would also predict the segregation between NaTxs from T. (Tityus) bahiensis and those from the subgenus T. (Atreus). In that regard, Guerrero-Vargas et al. (2012) analyzed 75 peptides from 10 nominal species of both groups and, in their reconstruction, found 13 well-supported NaTxs clusters. Except for one (NaTx3), each cluster included toxins from members of only one subgenus. However, even in NaTx3 (composed of toxins from both groups), they occupy different positions in the cluster, highlighting the segregation between toxin homologs produced in these phylogenetically distant groups.
The extent of divergences translates to the immunochemical level since Borges et al. (2020) found that scorpion venoms belonging to subgenera T. (Atreus) and T. (Tityus) display high antigenic diversity against several commercial antivenoms. Based on the differences in immunoreactivity and peptide venom composition, these authors recognize the existence of four "toxinological regions": (1) Lower Central America-Amazonia, (2) Venezuela, (3) Southeastern South America, and (4) a region seemingly centered in the Andean foothills. The first two regions comprise T. (Atreus) species, while the third includes those of the T. (Tityus) bahiensis clade. The fourth region contains species of T. (Tityus) bolivianus clade (sensu Moreno-Gonzáles et al. 2021) together with T. cerroazul from Lower Central America, suggesting a previously unrecognized similarity. Interestingly, the immunogenicity of the T. (Archeaotityus) group was previously evaluated through the type species T. clathratus, showing poor recognition by commercial antivenoms (Borges et al. 2010), which suggests some evolutive separation, either from regions II (T. discrepans) and III (T. serrulatus) groups. Based on clinical evidence, Borges and collaborators (2020) even included a member of the Archaeotityus subgenus, Tityus silvestris, in region I, suggesting close similarity of Archaeotityus with some Atreus species such as T. obscurus, for instance.
Our phylogenetic analyses also show strong homologies between T. (Archaeotityus) NaTxs and those from the T. (Atreus) clade. Five T. ocelote toxins (Toc3, Toc4, Tc5, Toc6, and Toc7) shared well-supported phylogenetic proximity with several species included in T. (Atreus) group. Toc5 sequence is identical to that of TpaCR2, a toxin first identified in T. (Atreus) jaimei (formerly, T. pachyurus) from Costa Rica (Borges et al. 2020). In addition, in our phylogenetic analysis, and the one presented by Borges et al. (2012, their consensus tree), Tcl1 is included in a clade composed of toxins of the subgenus T. (Atreus). Likewise, we were not able to resolve the relationship of T. mattogrossensis Tm1, and Tm2 with toxin homologs from either the T. (Tityus) bahiensis clade or T. (Atreus), a situation also evidenced in the de Oliveira et al. (2021) analysis (their Fig. 7). Thus, the claim that these same toxins are more closely related to the T. (Tityus) bahiensis clade is, at best, weak. Increasing the number of terminal groups in a phylogenetic analysis can reduce the effects of long branches, which might otherwise "attract" and erroneously group, terminal nodes (Wiens, 2005). However, this issue is commonly ignored during reconstructions of phylogenetic relationships instead of increasing the sequence length or the total number of characters in the analysis. Taxon sampling also improves estimates of evolutionary parameters derived from phylogenetic trees and is therefore crucial for optimizing applications of phylogenetic analyses (Heath et al. 2008).
Since, according to previous molecular phylogenetic evidence, Archaeotityus and Atreus are not considered sister groups, the closeness we found of some NaTxs from these subgenera homologs is intriguing. One possible explanation is that these are paralogs that conceivably arose by gene duplication before the divergence of the most recent ancestor between these subgenera. These paralogs must have been lost later, during the divergence of the T. (Tityus) bahiensis clade. How ancestral toxins were maintained in Archaeotityus is not entirely clear. Both positive and purifying selection has allowed the development and maintenance of venoms adapted to their ecological functions: selecting toxins that are more sensitive to specific groups of predators or toxins that allow arthropods to be immobilized (Sunagar et al. al., 2013;Van der Meijden et al. 2017). The evolution of α-NaTx in Old World scorpions probably resulted from defensive pressure, as predator's sodium channels show greater variability than their insect counterparts (Zhang et al. 2015). Indeed, multiple independent origins of mammalianspecific NaTxs occurred in separate lineages of medically important buthids, and this diversification chronologically coincides with the appearance of their mammalian predators (Santibáñez-López et al. 2022). Thus, increased toxicity in some Tityus lineages (i.e., T. discrepans, T. bahiensis, T. pachyurus) may result from greater selective pressure. Unlike other members of the genus, Archaeotityus scorpions have a small body size, reduced telson, and are primarily of reserved arboreal habits. These characteristics suggest that these animals are probably not frequently targeted by mammalian or other vertebrate predators (but see Gabriel et al. 2021 for a documented predation case from a lizard). Therefore, they might experience less selective pressure on toxic components targeting mammals than the more conspicuous and toxic species of the genus. In a relaxed selection scenario, the pressure to maintain a trait is removed or reduced, but the result does not necessarily lead to a clear predictive pattern (Lahti et al. 2009). Thus, a trait maintained by several factors of selection (such as the joint action of trophic and defensive pressures) could retain its advantage even when one of the sources of selection declines. In the present case, relaxed selection could have supported the maintenance of toxins derived from different ancestral lineages in scorpions of the genus Archaeotityus by reducing the pressure to maintain toxins with a specific function.

Conclusions
Our analysis of T. ocelote venom proteome provides evidence that venoms of species in the subgenera Archaeotityus are as complex as those of other members of the genus, recognized for their medical relevance. The presence of toxin homologies to the Atreus and T. bahiensis groups suggests the possibility that some of these characters could be plesiomorphic, but their origin remains to be resolved. Venoms from only approximately 16% of T. (Archaeotityus) species, 19% of T. (Tityus, sensu lato), 25% of T. (Atreus), and none of those in (Caribetityus) and T. (Brazilotityus), have been analyzed to date, so the portrayal of their expressed toxins remains fragmented.
In addition, despite technical improvements in toxin purification and sequencing, several low molecular weight compounds in scorpion venoms have evaded identification in proteomic and transcriptomic studies. These factors combined make our understanding of the toxins expressed in Archaeotityus, only partial. Without a doubt, as the number of species and toxins analyzed increases, it will improve our understanding of the evolutionary history of these scorpions.