Analysis of Protein Composition and Bioactivity of Neoponera villosa Venom (Hymenoptera: Formicidae)

Ants cause a series of accidents involving humans. Such accidents generate different reactions in the body, ranging from a mild irritation at the bite site to anaphylactic shock, and these reactions depend on the mechanism of action of the venom. The study of animal venom is a science known as venomics. Through venomics, the composition of the venom of several ant species has already been characterized and their biological activities described. Thus, the aim of this study was to evaluate the protein composition and biological activities (hemolytic and immunostimulatory) of the venom of Neoponera villosa (N. villosa), an ant widely distributed in South America. The protein composition was evaluated by proteomic techniques, such as two-dimensional electrophoresis. To assess the biological activity, hemolysis assay was carried out and cytokines were quantified after exposure of macrophages to the venom. The venom of N. villosa has a profile composed of 145 proteins, including structural and metabolic components (e.g., tubulin and ATPase), allergenic and immunomodulatory proteins (arginine kinase and heat shock proteins (HSPs)), protective proteins of venom (superoxide dismutase (SOD) and catalase) and tissue degradation proteins (hyaluronidase and phospholipase A2). The venom was able to induce hemolysis in human erythrocytes and also induced release of both pro-inflammatory cytokines, as the anti-inflammatory cytokine release by murine macrophages. These results allow better understanding of the composition and complexity of N. villosa venom in the human body, as well as the possible mechanisms of action after the bite.


Introduction
Accidents involving venomous animals, especially insects belonging to the order Hymenoptera, generate a variety of clinical and histopathological manifestations that range from mild irritation at the site of the sting to a severe anaphylactic reaction, often fatal [1,2].
Studies focused on venom, which examine the venom composition and mechanisms of action, have increased and, through these, several bioactive molecules have been isolated and/or identified, often of protein character [3]. We highlight studies on venoms of spiders, snakes, bees, wasps, ants, scorpions, centipedes, and frogs [4][5][6][7][8][9][10][11]. These studies have shown that the bioactivity of venom is related mainly to the release of cytokines as well as inflammatory (e.g., nitric oxide) and lipid (e.g., prostaglandins) mediators [12]. Yi et al. [13] showed that the ant venom from Solenopsis invicta acts on the nitric oxide synthase enzyme of murine macrophages, interfering in the production and release of this mediator. Similar effects were demonstrated in macrophages exposed to the venom from acts on the nitric oxide synthase enzyme of murine macrophages, interfering in the production and release of this mediator. Similar effects were demonstrated in macrophages exposed to the venom from other animals such as scorpions, wasps, bees and spiders [14][15][16][17]. In addition, some venom presents hemolytic and/or cytotoxic activity leading to death, usually by apoptosis in human cells [16,18,19].
In the Neotropical Neoponera genus (Formicidae, Ponerinae), 57 ant species are described [20]. Ants of this genus possess a painful bite, level 2 on a pain scale ranging from 1 to 4 [21] that is attributed to the presence of cyclic dipeptides, as already described for the Neoponera apicalis venom [22]. Although the species of the Ponerinae are very close, several effects inherent to the action of the venom are reported. In insects, the venom initially causes a rapid, reversible, and dose-dependent paralysis, followed by an irreversible paralysis with subsequent permanent paralysis and death. The concentrations that lead to such manifestations vary from 38.7 to 799.2 μg/g, depending on the species of Ponerinae (which includes Neoponera) [23]. The main function of the venom of the genus of these ants is to capture preys and defend the colony, which is the reason why it has a paralytic effect [24].
Protein characterization and biological activity studies have become common for ant venom and, although there are some in this subject focused on the venom from other species of Ponerinae, such as Neoponera goeldii, Brachyponera. chinensis, and Brachyponera. sennaarensis [23,25,26], there are no other studies related to the Neoponera villosa venom, an ant that is found worldwide. In America, it can be found from the southwestern United States of America to southern Brazil [27], and is widely distributed throughout the Brazilian territory and is very common in the state of Bahia [28,29].
Thus, the aim of this study was to evaluate the protein composition and identify the hemolytic activity, cytotoxicity, and cytokine production induced by the N. villosa venom, collected in the south of the state of Bahia, Brazil.

Doses of Protein
The proteins present in venom were quantified using the 2D Quant Kit (GE Healthcare, Little Chalfont, UK) and showed 6.23 μg/μL. By the Bradford method [30], a protein concentration of 2.79 μg/μL was found.   The highest density of protein, shown by the largest number of bands, was observed between 20.1 and 97 kDa, revealing a predominance of higher molecular weight proteins.

Protein Profile Obtained by Two-Dimensional Electrophoresis (2D)
The 2D gels ( Figure 2) demonstrated a protein profile composed of 145 spots, from which 139 were identified (Figure 3 and Table 1).
Int. J. Mol. Sci. 2016, 17,513 3 of 20 The highest density of protein, shown by the largest number of bands, was observed between 20.1 and 97 kDa, revealing a predominance of higher molecular weight proteins.

Protein Profile Obtained by Two-Dimensional Electrophoresis (2D)
The 2D gels ( Figure 2) demonstrated a protein profile composed of 145 spots, from which 139 were identified (Figure 3 and Table 1).   The highest density of protein, shown by the largest number of bands, was observed between 20.1 and 97 kDa, revealing a predominance of higher molecular weight proteins.

Protein Profile Obtained by Two-Dimensional Electrophoresis (2D)
The 2D gels ( Figure 2) demonstrated a protein profile composed of 145 spots, from which 139 were identified (Figure 3 and Table 1).

Hemolytic Activity Induced by Neoponera villosa Venom
Hemolysis was observed up to the fourth largest concentration of venom tested (63 µg/µL). Figure 4 presents the hemolysis rate of the venom in different concentrations. We can see that the hemolytic effect was dose-dependent, and was 50.6%, 40%, 26%, 12.1%, and 2.5% for the concentrations of 500, 250, 125, 63, and 31 µg/mL, respectively. Below these concentrations, it was not possible to observe hemolysis.

Hemolytic Activity Induced by Neoponera villosa Venom
Hemolysis was observed up to the fourth largest concentration of venom tested (63 μg/μL). Figure 4 presents the hemolysis rate of the venom in different concentrations. We can see that the hemolytic effect was dose-dependent, and was 50.6%, 40%, 26%, 12.1%, and 2.5% for the concentrations of 500, 250, 125, 63, and 31 μg/mL, respectively. Below these concentrations, it was not possible to observe hemolysis.

Induction of Cytokines in J774 Macrophages in the Presence of Neoponera villosa Venom
The presence of the venom in cultured murine macrophages was able to stimulate pro-and anti-inflammatory cytokines.
The production of interleukin 6 (IL-6) by macrophages presented significantly higher levels in the presence of venom, especially after 24 h of stimulation. The difference in production was statistically significant at almost all the concentrations of venom used when compared with the negative control (with the exception of the highest and lowest concentrations studied in 2 h). However, within 24 h, the production levels of IL-6 were always lower than the value induced by lipopolysaccharide (LPS) ( Figure 5).

Induction of Cytokines in J774 Macrophages in the Presence of Neoponera villosa Venom
The presence of the venom in cultured murine macrophages was able to stimulate pro-and anti-inflammatory cytokines.
The production of interleukin 6 (IL-6) by macrophages presented significantly higher levels in the presence of venom, especially after 24 h of stimulation. The difference in production was statistically significant at almost all the concentrations of venom used when compared with the negative control (with the exception of the highest and lowest concentrations studied in 2 h). However, within 24 h, the production levels of IL-6 were always lower than the value induced by lipopolysaccharide (LPS) ( Figure 5).

Hemolytic Activity Induced by Neoponera villosa Venom
Hemolysis was observed up to the fourth largest concentration of venom tested (63 μg/μL). Figure 4 presents the hemolysis rate of the venom in different concentrations. We can see that the hemolytic effect was dose-dependent, and was 50.6%, 40%, 26%, 12.1%, and 2.5% for the concentrations of 500, 250, 125, 63, and 31 μg/mL, respectively. Below these concentrations, it was not possible to observe hemolysis.

Induction of Cytokines in J774 Macrophages in the Presence of Neoponera villosa Venom
The presence of the venom in cultured murine macrophages was able to stimulate pro-and anti-inflammatory cytokines.
The production of interleukin 6 (IL-6) by macrophages presented significantly higher levels in the presence of venom, especially after 24 h of stimulation. The difference in production was statistically significant at almost all the concentrations of venom used when compared with the negative control (with the exception of the highest and lowest concentrations studied in 2 h). However, within 24 h, the production levels of IL-6 were always lower than the value induced by lipopolysaccharide (LPS) ( Figure 5).  There was a larger release of interleukin 12 (IL-12) after 24 h of stimulation, which was statistically significant only at the highest concentrations of venom (500 to 62.5 µg/mL). Although the levels of IL-12 were discreet under 2-h stimulation, they were comparable to the amount found under LPS stimulus ( Figure 6). There was a larger release of interleukin 12 (IL-12) after 24 h of stimulation, which was statistically significant only at the highest concentrations of venom (500 to 62.5 μg/mL). Although the levels of IL-12 were discreet under 2-h stimulation, they were comparable to the amount found under LPS stimulus ( Figure 6). Tumor necrosis factor α (TNF-α) presented the highest levels of production when compared with the other pro-inflammatory cytokines evaluated (IL-6 and IL-12). Regardless of the time of cultivation (2 or 24 h) and when compared with the basal condition, macrophages produced significantly greater amounts of TNF-α under stimulus of the Neoponera villosa venom. The production of this cytokine occurred in a dose-dependent way, as the concentration of the venom decreased, the amount of TNF-α detected also decreased. However, the highest levels of this cytokine under venom stimulus were still lower than those found under the LPS stimulus (Figure 7). Regarding interleukin 10 (IL-10), the venom was able to stimulate the production of this cytokine at all concentrations used, not demonstrating any relation to the dose of venom used on the cultures. The difference, statistically significant when compared with the control, was observed at all the Tumor necrosis factor α (TNF-α) presented the highest levels of production when compared with the other pro-inflammatory cytokines evaluated (IL-6 and IL-12). Regardless of the time of cultivation (2 or 24 h) and when compared with the basal condition, macrophages produced significantly greater amounts of TNF-α under stimulus of the Neoponera villosa venom. The production of this cytokine occurred in a dose-dependent way, as the concentration of the venom decreased, the amount of TNF-α detected also decreased. However, the highest levels of this cytokine under venom stimulus were still lower than those found under the LPS stimulus (Figure 7). There was a larger release of interleukin 12 (IL-12) after 24 h of stimulation, which was statistically significant only at the highest concentrations of venom (500 to 62.5 μg/mL). Although the levels of IL-12 were discreet under 2-h stimulation, they were comparable to the amount found under LPS stimulus ( Figure 6). Tumor necrosis factor α (TNF-α) presented the highest levels of production when compared with the other pro-inflammatory cytokines evaluated (IL-6 and IL-12). Regardless of the time of cultivation (2 or 24 h) and when compared with the basal condition, macrophages produced significantly greater amounts of TNF-α under stimulus of the Neoponera villosa venom. The production of this cytokine occurred in a dose-dependent way, as the concentration of the venom decreased, the amount of TNF-α detected also decreased. However, the highest levels of this cytokine under venom stimulus were still lower than those found under the LPS stimulus (Figure 7). Regarding interleukin 10 (IL-10), the venom was able to stimulate the production of this cytokine at all concentrations used, not demonstrating any relation to the dose of venom used on the cultures. The difference, statistically significant when compared with the control, was observed at all the Regarding interleukin 10 (IL-10), the venom was able to stimulate the production of this cytokine at all concentrations used, not demonstrating any relation to the dose of venom used on the cultures. The difference, statistically significant when compared with the control, was observed at all the concentrations used (with the exception of 125 and 62.5 µg/mL). In addition, the levels of IL-10 produced after 24 h of stimulation were slightly higher when compared with the 2-h stimulus, but they were not similar to those observed under LPS stimulus (Figure 8). concentrations used (with the exception of 125 and 62.5 μg/mL). In addition, the levels of IL-10 produced after 24 h of stimulation were slightly higher when compared with the 2-h stimulus, but they were not similar to those observed under LPS stimulus (Figure 8).

Discussion
The protein profile of the N. villosa venom showed predominance of proteins with higher molecular weight (between 20.1 and 97 kDa). Similarly, Lee et al. [25] worked with Brachyponera chinensis venom and observed higher protein density between 12 and 85 kDa. Silva et al. [31] analyzed Odontomachus bauri venom, and isolated proteins with molecular weight ranging from 18 to 160 kDa under nonreducing conditions, with more intense staining for bands above 29 kDa. Under reducing conditions, the electrophoretic profile changed, showing a wider range, from 24 to 160 kDa. The profile obtained by two-dimensional electrophoresis was composed of 145 spots. The distribution of these spots showed a higher density in neutral and alkaline areas and molecular weight above 20.1 kDa. To date, there are no data in the literature about protein profile of N. villosa venom; our results are the first ones. However, similar data were demonstrated in other species of the same genus. Costa Manso et al. [32] described, in the 2D profile of N. goeldii venom, higher density of spots between 10 and 66 kDa, and Lee et al. [25], studied B. chinensis venom and reported prevalence of proteins with pI in the neutral and alkaline zones. For the venom from another species of ant, Solenopsis invicta, all the proteins identified in 2D-SDS-PAGE were located above 14.4 kDa and most presented pI in the alkaline lane [33]. Such features are common to the venom of many hymenoptera as well as to ants, bees and wasps [34].

Discussion
The protein profile of the N. villosa venom showed predominance of proteins with higher molecular weight (between 20.1 and 97 kDa). Similarly, Lee et al. [25] worked with Brachyponera chinensis venom and observed higher protein density between 12 and 85 kDa. Silva et al. [31] analyzed Odontomachus bauri venom, and isolated proteins with molecular weight ranging from 18 to 160 kDa under nonreducing conditions, with more intense staining for bands above 29 kDa. Under reducing conditions, the electrophoretic profile changed, showing a wider range, from 24 to 160 kDa. The profile obtained by two-dimensional electrophoresis was composed of 145 spots. The distribution of these spots showed a higher density in neutral and alkaline areas and molecular weight above 20.1 kDa. To date, there are no data in the literature about protein profile of N. villosa venom; our results are the first ones. However, similar data were demonstrated in other species of the same genus. Costa Manso et al. [32] described, in the 2D profile of N. goeldii venom, higher density of spots between 10 and 66 kDa, and Lee et al. [25], studied B. chinensis venom and reported prevalence of proteins with pI in the neutral and alkaline zones. For the venom from another species of ant, Solenopsis invicta, all the proteins identified in 2D-SDS-PAGE were located above 14.4 kDa and most presented pI in the alkaline lane [33]. Such features are common to the venom of many hymenoptera as well as to ants, bees and wasps [34].
In spots 2, 5, 6, 10, and 13, the following heat shock proteins (HSPs) or stress proteins were identified: HSP 90 (spots 2 and 5), HSP 70 (spots 6 and 10), and HSP 60 (spot 13). These proteins act as chaperones, facilitating the folding of other proteins, and as regulators of signaling, protecting the cell from the event of apoptosis [35]. In addition, the activation of the immune system mediated by HSPs is known. Preparations of purified HSPs, including HSP 60, HSP 70, and HSP 90, from various sources (wild and recombinant bacteria, several mammals, and humans) are potent activators of the innate immune system. These preparations induce the production of pro-inflammatory cytokines, such as TNF-α, IL-1, IL-6, and IL-12, in addition to nitric oxide by monocytes, macrophages and dendritic cells. [36,37]. The presence of such proteins in the venom may be related to the increased release of IL-12 and TNF-α by J774 1.6 macrophages, since, as demonstrated by Asea et al. [38], monocytes derived from human peripheral blood showed high levels of TNF-α, IL-1, and IL-6 release after the exposure to human HSP 70. Similarly, murine peritoneal macrophages and monocytes derived from human peripheral blood showed high expression of mRNA encoder of TNF-α, IL-1, IL-6, and GM-CSF after exposure to HSPs purified from several microorganisms such as Escherichia coli, Mycobacterium tuberculosis, and Legionella pneumophila [39,40].
In spot 23, α-N-acetylgalactosaminidase was identified, which is a hydrolase described for animals and bacteria [41]. It has already been described that this enzyme promotes immunosuppression through the inactivation of macrophages [42].
From spot 27, it was possible to identify an aspartyl protease, an enzyme with cathepsin D activity [43], capable of degrading a variety of substrates, including laminin and fibronectin, constituents of the extracellular matrix, thus contributing to the spread of other venom constituents through the body in which it is inoculated [44].
Several arginine kinase isoforms were identified (spots 33, 37, 38, 49, 50, 51, 52, 53, and 109). This protein has an enzymatic function and is responsible for the energetic metabolism in invertebrates. It is considered a pan-allergen, since it can trigger allergic responses through specific IgE linking [45]. It can be found in many allergenic sources, such as venoms, and has already been isolated from the venom of bees and wasps [35,46], as well as non-venomous animals used as food such as shrimp, octopus, and crab [47][48][49].
In spot 57, a superoxide dismutase (SOD) was identified as well as a catalase in spot 95. Both are responsible for the protecting the venom from the reactive oxygen species produced by the body, such as O 2 and H 2 O 2 [46]. In addition, a form of copper/zinc-dependent SOD was identified with the potential to trigger immune responses activation processes in the body where the venom is inoculated [50].
In spot 119, a hyaluronoglucosaminidase (or hyaluronidase) was identified, an enzyme present in several venoms and found in the venom of snakes, scorpions, bees, and wasps [51]. This enzyme acts on the cellular interstitial, downgrading it and is therefore, known as spreading factor [52]. It maximizes the infiltration of the venom as it digests part of the extracellular matrix, facilitating the action of other components [53].
A phospholipase A2 was identified from spot 123. This enzyme, widely distributed in several venoms, destroys phospholipids in biological membranes, leading to the formation of pores, cell lysis and tissue damage [33], inducing inflammation [54], and inhibiting the activity of macrophages [55]. The presence of the enzyme in the venom can be related to hemolytic activity and to inhibition of NO formation by macrophages.
Spot 132 has been identified as a metalloproteinase that is part of a family of enzymes involved in the occurrence of muscle necrosis, tissue damage, swelling and bleeding at the site of the sting, rare coagulopathies, among other related inflammatory reactions [56,57]. In addition, studies with venom from snakes have shown that the metalloproteinases, similar to hyaluronidases, act on the destruction of the extracellular matrix by interfering in cell adhesion, thus providing tissue disorganization [58].
In spots 87 and 137, forms of peroxiredoxin, belonging to a ubiquitous family of antioxidant proteins were identified [59]. Its presence in the venom is related to its own protection from oxidative stress. This function is also performed by glutathione S-transferase protein, found in spots 129 and 145 [33].
Studies by other authors suggest that the ant cells can rupture when the sting apparatus is removed from the insect with the aid of tweezers, which can cause the addition of these cell proteins to the venom proteins [46].
Although the remaining spots identified in the venom are classified exclusively as intracellular constituents (cytoplasmic or present inside organelles) and their presence in the venom is also related to the lesion of cells surrounding the venom reservoir during the mechanical removal of the glands [46], many of these proteins have been identified in proteomes of several venoms. In Apis mellifera venom, forms of malate dehydrogenase (spots 54 and 122), enolase (spots 35, 45, and 46) and apolipophorine (spot 77) were found [60]. In the venom of the endoparasitoid Pteromalus puparum, forms of vacuolar ATP synthase were found (spots 9, 19, and 21) and glyceraldehyde-3-phosphate dehydrogenase (spots 116, 117, and 118), among others [45]. Furthermore, Torres et al. [61] used transcriptomic approaches, and showed that more than 50% of the genes transcribed in the venom glands of the Dinoponera quadriceps ant were intracellular constituents.
Our study using a proteomic approach showed that N. villosa share many venom proteins with others ant species and even with other hymenopteran venoms. Similar proteomic tools were used by Touchard et al. [62] and they demonstrated that MALDI-TOF (Matrix-Assisted Laser Desorption/Ionization Time-Of-Flight) mass profiles of venom peptides from different species of Neoponera, Pachycondyla and Odontomachus permit taxonomic discrimination in closely related species. This chemotaxonomic tool is important in exploiting the enormous biodiversity of ants and other hymenopteran venoms as sources for novel drugs or biopesticides.
Venoms from various species of hymenoptera also present hemolytic activity, such as Pogonomyrmex barbatus (P. barbatus) [18], Paraponera clavata (P. clavata) [21], Myrmecia pilosula (M. pilosula) [67], Ectatomma tuberculatum [68] ants, and wasps of the genus Vespa and Polistes [69]. The minimum concentrations of hemolysis vary and, within the same genus, there may be a great discrepancy. For example, it takes 500 U/mg dried venom from P. barbatus to provide 50% hemolysis [18], but it only takes 10 U/mg dried venom from P. clavata to obtain the same value of hemolysis [21]. Values close to those of the present study were found by Matuszek et al. [67], who took a concentration of 65 µg/mL venom from the M. pilosula ant to generate 12.5% hemolysis. The Polistes lanio lanio wasp requires a concentration equal to or greater than 50 µg/mL to obtain minimal hemolysis. Hemolysis may be the result of one or more mechanisms of action, and is usually linked to enzyme action. There may be direct hemolysis, in which one or more venom components act on the erythrocytic membrane, yielding pores and leading them to lysis, for example, there is the action of phospholipases that cleave the phospholipids of the membrane forming channels, where the intracellular content pours through [70] and proteins/peptides with surfactant activity [64]. Indirect hemolysis can be mediated by the action of complement, in which venom components, such as metalloproteinases, facilitate the action of cascade-signaling proteins [71].
Regarding N. villosa venom, a synergistic effect is suggested between the phospholipase and the metalloproteinase activities in the venom (spots 123 and 132). As these components are at low concentration in the venom, higher doses are required to obtain hemolysis (in this case, over 31 µg/mL).
The dosage of cytokines (IL-6, IL-10, IL-12, and TNF-α) released in the culture supernatant by macrophages exposed to the venom was performed by ELISA technique and, despite the significant increase in all pro-inflammatory cytokines (IL-6, IL-12, and TNF-α) in 2 and 24 h, there was also an increase of IL-10, a cytokine of anti-inflammatory and immunomodulatory character in both the periods evaluated. The induction pattern of pro-inflammatory cytokine has been observed for venom of several hymenoptera. Lam et al. [72] demonstrated that A. mellifera venom presents pro-inflammatory properties, inducing, in vitro, the release of cytokine IL-1 by macrophages through the activation of the p38 MAPK pathway. Tambourgi et al. [73] showed that, in vivo, the venom from the Loxosceles intermedia spider, is able to increase the IL-6 and TNF-α serum levels. Similar effect was observed by Malaque et al. [74], who stimulated primary culture human keratinocytes with venom from the Loxosceles gaucho spider and noted an increase in TNF-α. The IL-6 cytokine, as well as some chemokines, are also related to inflammation at the site of the sting induced by the venom from the same spider, as presented by Barbaro et al. [75]. The venom from the Tityus serrulatus scorpion generates local inflammation with high serum levels of IL-6 and TNF-α [12]. The same was observed, in vitro, when macrophages were exposed to the venom from the same species of scorpion and showed an increase in the release of IL-6 and TNF-α [76]. This pro-inflammatory pattern of the venom can be explained by the presence of some proteins known to induce macrophages to produce and release cytokines of this character, such as the heat shock proteins (HSPs), phospholipases, and metalloproteinases identified in this study. However, we also observed the production of IL-10, a cytokine that is anti-inflammatory and immunomodulatory in character able to inhibit the secretion of pro-inflammatory cytokines and regulate the differentiation and proliferation of some cells [77]. Some studies with venom also demonstrate increased IL-10 after stimulation. Fialho et al. [12] showed that in addition to the pro-inflammatory cytokines IL-6 and TNF-α, the venom from the scorpion T. serrulatus also induced an increase in serum IL-10. Andrade et al. [78] demonstrated that the stimulus with venom from the same scorpion raises the increased mRNA expression of both pro-(IL-1, IL-6, IFN-γ, and TNF-α) and anti-inflammatory (IL-10) cytokines. Ribeiro et al. [79] noted a similar fact and showed that after stimulating peripheral blood mononuclear cells with the venom from the snake Crotalus durissus collilineatus, there was an increase in the release of IL-10 and TNF-α as well. This fact could be related to the presence of components that stimulate both pro-and anti-inflammatory response in cells stimulated by the venom. Petricevich et al. [80] suggest that the production of IL-10 has a protective character in the process of poisoning by venom from snakes of the Bothrops genus, preventing the systemic shock condition.

Venom Extraction
The ants were collected while foraging on the campus of the State University of Santa Cruz (UESC) and on the Executive Committee of the Cocoa Crop (CEPLAC), in Ilhéus, Bahia, Brazil (14˝47 1 20"S and 39˝02 1 58"W). The insects were killed by freezing at´20˝C, identified in stereoscopic microscope and stored at´20˝C until use. We used 350 individuals (250 for protein characterization and 100 for biological activity tests). The crude venom, obtained by breaking the bag attached to the stinger after the ants' dissection, was diluted in phosphate buffered saline (PBS) 100 mM pH 7.8.

Protein Extraction
To precipitate the proteins from the crude venom, the same amount of 20% (v/v) trichloroacetic acid (TCA) in acetone was added to each microtube containing venom. The microtube was packaged at´20˝C for 60 min and centrifuged at 16,100ˆg for 45 min to form the pellet. The supernatant was discarded. The pellet was washed with 500 µL cold acetone and then sonicated on ice for 21 s in cycles of 7 s with 70% range, and centrifuged for 10 min at 16,100ˆg, forming the pellet again. The supernatant was discarded and the pellet washed four times with acetone. The acetone was evaporated at room temperature and then rehydration buffer was added (7 M urea, 2 M thiourea, 2% CHAPS, and 0.002% bromophenol blue) to solubilize the proteins.

Protein Quantification
A 2D Quant Kit (GE Healthcare) was used to quantify proteins dissolved in the rehydration buffer, according to the manufacturer's recommendations.
For the tests of biological activity (venom dissolved in PBS), dosage was used according to the Bradford protocol [30], using bovine serum albumin as standard.

Electrophoresis (SDS-PAGE)
For electrophoresis, 20 µg protein were subjected to SDS-PAGE (12.5% acrylamide). Aliquots of 7 µL of standard molecular weight (Amersham, Little Chalfont, UK) were also submitted to electrophoresis for comparison. After running, the gels were fixed (40% ethanol and 10% acetic acid) for 10 min, and then the solution was replaced by Coomassie Blue G-50 (Sigma Aldrich, St. Louis, MO, USA) dye, which was maintained for 12 h while stirring. Then, the dye was removed and the gels bleached with distilled water.

Two-Dimensional Electrophoresis (2D-SDS-PAGE)
Gel strips (Immobiline Dry Strip) of 13 cm pH 3-10 NL (GE Healthcare) were rehydrated for 12 h in rehydration buffer containing 350 µg protein. They were focused using a multiple-stage protocol (1 h at 500 V, 1 h at 1000 V and 6 h at 8000 V). Before the second dimension, each strip was balanced for 15 min in buffer containing 50 mM Tris-HCl, pH 8.8, 6 M urea, 30% glycerol, 2% SDS, 0.002% bromophenol blue, supplemented with 10 mg/mL DTT. After this stage, strips were balanced for 15 min in the buffer described above, but the DTT was replaced by 25 mg/mL of iodoacetamide. The second dimension electrophoresis was performed at 150 V using polyacrylamide gels (12.5%) in triplicate.

Protein Visualization and Image Analysis
Polyacrylamide gels were fixed (40% ethanol and 10% acetic acid) for 30 min, and then stained with Coomassie Blue G50 (Sigma) for five days while stirring. Subsequently, the gels were bleached with distilled water. Images of the gels were scanned and analyzed by ImageMaster 2D Platinum 7.0 (GE Healthcare) software to assess the number of spots, and characterize the values of pI and molecular weight.

Spot Excision and Triptych Digestion
The spots were excised from the gels and arranged separately in microtubes. Later, they were added to the gel fragments 200 µL NH 4 HCO 3 (ammonium bicarbonate), 25 mM, pH 8.0 in 50% acetonitrile to discolor and remove SDS. This solution was kept in the microtubes for 24 h. After this period, the spots were washed with 200 µL ultrapure water, 100 µL acetonitrile were added to dehydrate the gel, and the fragments were incubated for 5 min at room temperature. The excess acetonitrile was removed, and the microtubes were vacuum dried for 20 min. Then, the proteins were partially digested with trypsin (5 µL at the concentration of 20 ng/µL) for 10 min at 4˝C. Later, 200 µL NH 4 HCO 3 , 25 mM were added, followed by incubation at 37˝C for 24 h to complete the digestion of the peptides. To extract the peptides, the remaining solution was transferred to a differente and clean microtube, and 50 µL 50% acetonitrile were added to the first microtube containing 0.1% formic acid. These microtubes containing a mixture of acetonitrile, formic acid, and residual peptides were agitated slightly for 30 min. The solution was pipetted to the new microtubes, and this process was repeated. In the end, the microtubes containing fully digested peptides were vacuum dried to the volume of 5-10 µL.

Peptide Sequencing and Protein Identification
The peptides obtained as described in the previous item were separately applied in the nanoAcquity UPLC (Waters, Milford, MA, USA) in two C18 columns ("trapping" of 5 µm, 180 µmˆ20 mm; and the second of 1.7 µm, 100 µmˆ100 mm), under a stream of 0.6 µL/min in a 50-min run. The peptides were separated according to acetonitrile gradient grade, transferred to the Mass Spectrometer (Micromass Q-TOFmicro, Waters) and ionized in a capillary under 3000 V. They were fragmented in the positive mode with selection of minimum relative intensity of 10 counts, and the three most intense ions were analyzed per every 1 s sweep, with collision energy ranging from 20 to 95 V according to the mass/charge (m/z) of peptides. The spectra obtained in each run were analyzed by MASCOT Server Online (Matrix Science: http://www.matrixscience.com/), and compared with the NCBI database. On hydrolysis, by trypsin, the possible loss of a cleavage site was considered. The tolerance of the peptide masses was˘0.3 Da, and the tolerance of the fragment masses was˘0.1 Da.

Hemolytic Activity
To perform the test, a 96-well microplate with a U-shaped bottom was used. The initial venom concentration was 500 µg/mL and was performed a serial dilution to the concentration of 0.2 µg/mL venom. Then, 100 µL red blood cell suspension was added to 1% (v/v). For the reference value of zero, 100 µL saline solution and 100 µL red blood cell suspension were added. For the reference value of 100% hemolysis, 100 µL saline solution containing 1% (v/v) Triton X-100 (Sigma) and 100 µL red blood cell suspension were added. The plates were incubated at 25˝C for 120 min. Then, the microplates were centrifuged for 5 min at 1440ˆg. The supernatant of each well was collected and transferred to the wells of a new plate with a flat bottom. The rate of hemolysis was taken by the absorbance reading at 540 nm and the hemolysis was calculated by the following equation: Haemolysis p%q " Adilution´Aneg.control Aposit.control´Aneg.controlˆ1 00 4.10. Murine Macrophages Culture and Stimulus with Neoponera villosa Venom J774 1.6 murine macrophages (Rio de Janeiro Cell Bank-BCRJ, access No. 0273) were cultured for 24 h until they reached the confluence amid DMEM (Gibco, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco) and 1% antibiotics (penicillin and streptomycin) (Gibco), at 37˝C, 5% CO 2 in the concentration of 1ˆ10 6 cells/mL, and the viability was monitored with trypan blue dye at each bouncing.
For the tests with venom stimulation, the macrophages were cultured for 48 h at the concentration of 1ˆ10 6 cells/mL in 24-well plates containing DMEM, 10% FBS and 1% antibiotics as well as different concentrations of venom (ranging from 500 to 0.03 µg/mL). As positive stimulus control the cells were grown in the presence 2 µg/mL of bacterial lipopolysaccharide (LPS) (Sigma) and as negative control (basal condition) only with the culture supplemented with FBS and antibiotic.

Cytokines Doses
Five hundred milliliter aliquots from culture supernatants described in item 4.10 were removed, in periods of 2 and 24 h after the stimulus. Cytokines IL-6, IL-10, IL-12, and TNF-α were evaluated by ELISA kits (Peprotech, Ribeirão Preto, São Paulo, Brazil), according to the manufacturer's recommendations.
The statistical analysis of cytokine concentration was carried out via the GraphPad Prism 5 software, using the Mann-Whitney non-parametric test, considering p < 0.05 and confidence interval (CI) of 95%.

Conclusions
The venom of Neoponera villosa has a complex protein composition, which consists of 145 proteins in the neutral and alkaline lanes of pH and above 20.1 kDa. In addition, N. villosa venom presents hemolytic activity for concentrations exceeding 31 µg/µL and this fact is related to the presence of phospholipase and metalloproteinase enzymes in the venom (spots 123 and 132, respectively). N. villosa venom induces an increase in the release of pro-inflammatory cytokines (IL-6, IL-8, and TNF-α), and this increase is related to the presence of the following proteins in the venom: HSP 90 (spots 2 and 5), HSP 70 (spots 6 and 10), and HSP 60 (spot 13), since it is known that HSPs are potent activators of the innate immune system, inducing the release of pro-inflammatory cytokines. In addition to the pro-inflammatory cytokines, there was slight increase in IL-10 production by macrophages, and this fact could be related to the presence of components that stimulate both a preand anti-inflammatory response in the venom or with a cellular compensatory mechanism aimed at preventing damages caused by the high concentration of pro-inflammatory cytokines. These results allow a better understanding of the composition and complexity of N. villosa venom in the human body, as well as the venom's possible mechanisms of action after the bite.

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