*3.1. LiRecTCTP Expression and Purification*

LiRecTCTP expression was performed using the same heterologous system previously described, using *E. coli* and a His-tag [2], but we used an improved protocol of purification. In the former protocol, recombinant toxin was purified in native conditions in a 2-step chromatographic approach: Ni+2-NTA affinity chromatography using, and ion-exchange chromatography using DEAE agarose [2]. Herein, LiRecTCTP was purified under native conditions in just one step of chromatography (Ni+2-affinity chromatography). This new protocol resulted in highly purified recombinant toxin and the yield was 16 mg/<sup>L</sup> of *E. coli* culture (Figure 1A). Purified LiRecTCTP toxin was submitted to circular dichroism to analyze protein folding. Deconvolution results show 43% of defined secondary structures as alpha-helix and beta-sheets (Figure 1B,C). These results are in agreemen<sup>t</sup> with previous data obtained for LiRecTCTP [2].

**Figure 1.** Heterologous expression, purification, and circular dichroism spectroscopy CD analyses of recombinant LiTCTP. (**A**) SDS-PAGE (12.5%) analysis of recombinant LiTCTP toxin expression stained with Coomassie blue dye. Lane 1 show *E. coli* BL21 (DE3) pLysS cells before induction with IPTG. Lane 2 shows *E. coli* BL21(DE3) pLysS after induction for 4h with 0.1 mM isopropyl-d-thiogalactoside (IPTG) (supernatant from cell lysates obtained by freezing and thawing in extraction buffer before incubation with Ni2<sup>+</sup>-NTA beads). Lane 3 depicts the void from Ni2<sup>+</sup>-NTA chromatography. Lane 4 shows eluted recombinant protein from Ni2<sup>+</sup>-NTA beads. Molecular mass markers are shown on the left. (**B**) The UV-CD spectrum was obtained in a Jasco J-810 spectropolarimeter (Jasco Corporation, Tokyo, Japan) by diluting the sample at 0.5 mg/mL in phosphate buffer, pH 7.4 at 20 ◦C. Graphic representation was plotted using GraphPad Prism 6. (**C**) The deconvolution of data, α-helix and β-sheet percentages of LiRecTCTP structure was calculated using K2D3 tool.

### *3.2. LiRecTCTP Activity on RBL-2H3 Cells*

LiRecTCTP activity on mast cells degranulation was evaluated using RBL-2H3, a mast cell–like cell line, originally a rat basophilic leukemia cell. Initially, a cytotoxic effect of LiRecTCTP (100 μg/mL) on these cells was disregarded by evaluating cells viability (MTT assay) and morphology (SEM) (Figure 2A,C) after 2h-treatment with LiRecTCTP (10, 50, and 100 μg/mL). LiRecDT1H12A, a mutated and almost inactive toxin (only residual levels of activity) produced in the same way (heterologous system and chromatographic purification protocols), was included in the experiment (as negative control), as well as a compound that triggers degranulation (48/80, a positive control). Crude venom activity was also evaluated by MTT assay, and the resulting absorbance did not differ from controls. We did not observe deleterious effects of LiRecTCTP (100 μg/mL) in the viability/activity measured by MTT metabolization in formazan salt (Figure 2A) or in the cellular morphology shown in SEM (Figure 2C). Only 200 μg/mL of LiRecTCTP induced alteration in RBL-2H3 cells morphology (Figure 2C), and this concentration was not used for further functional characterization of LiRecTCTP. RBL-2H3 cells degranulation was measured by the beta-hexosaminidase activity assay (Figure 2B), a widely used test, mainly for research purposes. Beta-hexosaminidase is an acid hydrolase that characterizes lysosomal-derived secretory granules that are released during degranulation; the enzyme activity was measured by using p-nitrophenyl *<sup>N</sup>*-acetyl-beta-d-glucosamine as colorimetric substrate. The degranulation effect is evident when LiRecTCTP is incubated for 2 h with the cells and this activity was dependent of toxin concentration. It is important to highlight that the mutated toxin (LiRecDT1 H12A), produced following the same methodological procedures as LiRecTCTP, was not able to induce beta-hexosaminidase release, ruling out the possibility of contaminants being involved in the effect (Figure 2B). The activity of beta-hexosaminidase after 50 and 100 μg/mL LiRecTCTP treatments was increased two-fold and three-fold, respectively, when compared to the negative control (LiRecDT1 H12A). As shown, 100 μg/mL of LiRecTCTP had a higher degranulation effect than the positive control 48/80, a well-known polymer which triggers mast cell activation. *L. intermedia* crude venom also activates RBL-2H3 cells degranulation in a concentration-dependent manner. It is important to mention the cromolyn inhibitory effect on LiRecTCTP induced beta-hexosaminidase activity. Cromolyn blocks or reduces the mediators released from mast cells, suggesting a pro-inflammatory mechanism of histamine release induced by the LiRecTCTP toxin.

**Figure 2.** Effects of LiRecTCTP in mast cells viability and degranulation in vitro. RBL-2H3 cells (mast-like cells) were incubated with LiRecTCTP (10, 50, and 100 μg/mL), total venom from *L. intermedia* (10, 50, and 100 μg/mL), compound 48/80 (100 μg/mL) (positive control), LiRecDT1 H12A (100 μg/mL) (negative control) or PBS (negative control). After 2 h viability, morphology and activity of the granular enzyme Beta-hexosaminidase were measured. Inhibition of degranulation was performed using cromolyn (CROM) (20 μM). (**A**) Cell viability was evaluated using MTT assay. The values represent the average of the three independent experiments ± s.d. (performed in pentaplicate). (**B**) Beta-hexosaminidase activity assay. Results are expressed as the percentage of the total beta-hexosaminidase activity present in the cells, after subtracting the activity in the supernatant of unstimulated cells. The values represent the average of the three independent experiments ± s.d. (performed in pentaplicate) (\* *p* < 0.1; \*\* *p* < 0.01; \*\*\* *p* < 0.001 and *\*\*\*\* p* < 0.0001 compared with control; ## *p* < 0.01, ### *p* < 0.001 compared to LiRecTCTP treatment). (**C**) Scanning electron microscopy (SEM) of RBL-2H3 cells after 2 h treatment with LiRecTCTP (100 and 200 μg/mL). Images of each sample represent different fields and magnification. Scale bars indicate 10 μm, magnification 1000× (left) and 2000× (right). Arrow: apoptotic cell.

### *3.3. E*ff*ects of LiRecTCTP on the Ca2*+ *Signaling and Cytokines Expression*

As changes in the cytosolic Ca2+ are central for mast cells activation, we performed an assay to measure the Ca2+ influx in RBL-2H3 following a LiRecTCTP treatment (Figure 3A). We could observe a dose-dependent positive effect of LiRecTCTP in the Ca2+ influx. Cromolyn abrogated LiRecTCTP effects on Ca2+ levels. We also analyzed the cytokine production evoked by LiRecTCTP treatment in RBL-2H3 cells, by measuring the relative mRNA levels for IL-3 (Figure 3B), IL-4 (Figure 3C) and IL-13 (Figure 3D) by means of RT-PCR. LiRecTCTP increased the expression of IL-3, IL-4 and IL-13 in a dose-dependent manner, when compared to the negative controls (PBS and 100 μg/mL of GFP recombinant protein).

**Figure 3.** Effect of treatment of RBL-2H3 cells with LiRecTCTP on calcium influx and expression of interleukins. (**A**) RBL-2H3 cells were incubated with LiRecTCTP (50 and 100 μg/mL) in the presence of Fluo-4 AM in buffer containing calcium. The fluorescence of Fluo-4 was measured after 0, 5, 15, 30, 60, and 90 min. As a negative control, cells were incubated without LiRecTCTP. Cromolyn (CROM) (20 μM) inhibitory effect was evaluated in the presence of LiRecTCTP (100 μg/mL). The values represent the average of the three experiments ± s.d. (\*\* *p* < 0.01 and \*\*\*\* *p* < 0.0001). (**B**) Quantitative real-time PCR of IL-3 mRNA levels in RBL-2H3 exposed or not to LiRecTCTP (50 and 100 μg/mL) and GFP (100 μg/mL). (**C**) Quantitative real-time PCR of IL-4 mRNA levels in RBL-2H3 exposed or not to LiRecTCTP (50 and 100 μg/mL). (**D**) Quantitative real-time PCR of IL-13 mRNA levels in RBL-2H3 exposed or not to LiRecTCTP (50 and 100 μg/mL) and GFP (100 μg/mL). For quantification, we used the ΔΔCt method with GAPDH as an endogenous control for each sample (\* *p* < 0.1 and \*\*\*\* *p* < 0.0001, compared with controls, PBS and GFP recombinant protein). Data represent mean ± s.d. of three independent experiments.

### *3.4. LiRecTCTP In Vivo E*ff*ects—Vascular Permeability and Edema*

In order to evaluate the e ffect of di fferent histamine receptor blockers on the histaminergic response triggered by LiRecTCTP, we performed two animal studies in which vascular permeability and edema formation were assessed. Well-established antihistaminic drugs with di fferent targets were used in these experiments: promethazine (PRO), an H1 receptor antagonist; cimetidine (CIM), an H2 receptor antagonist; thioperamide (THIO), acts on H3 and H4 receptors; and cromolyn (CROM), a mast cell degranulation blocker. Vascular permeability was measured by Evans blue extravasation (Figure 4A) after intradermic inoculation of LiRecTCTP in mice, previously treated with an antihistaminic or not. Quantification was performed after Evans blue elution (Figure 4B). Images from mice skin and amount of dye eluted show that cromolyn was the most e ffective drug to reduce LiRecTCTP e ffects on vascular permeability (absorbance of the eluted dye was very similar to the negative control, PBS). Promethazine and thioperamide could inhibit about 30% of the LiRecTCTP e ffect in increasing vascular permeability. Cimetidine did not alter the increase in vascular leakage of Evans dye caused by LiRecTCTP.

**Figure 4.** Effect of treatment with the mast cell degranulation inhibitor or histamine receptor antagonists on vascular permeability induced by LiRecTCTP. Mice (*n* = 5) were pre-treated with promethazine (PRO), cimetidine (CIM), thioperamide (THIO), cromolyn (CROM), or PBS (control). Animals received solution of Evans blue dye in PBS intravenously prior to intradermal injection of LiRecTCTP (10 μg) or PBS (control). ( **A**) Representative images of mice dorsal skin at the point of sample inoculation. (**B**) Quantitative measurement of dye leakage induced by LiRecTCTP and inhibition with CROM, PRO, CIM, and THIO. Data represent mean ± s.e.m of one representative experiment from three independent biological replicates (\* *p* < 0.1 and \*\* *p* < 0.01).

We also used mice to evaluate if LiRecTCTP edematogenic e ffects could be inhibited by the anti-histaminic drugs; the e ffect of these inhibitors is shown on Figure 5A compared to LiRecTCTP by itself. Cimetidine did not present a significant e ffect on the paw edema caused by LiRecTCTP; a small inhibition of edema is observed in the first 10 min after toxin administration (Figure 5D). As shown for vascular permeability, promethazine, thioperamide, and cromolyn prevented LiRecTCTP e ffects on mice paw (Figure 5B,C,E). The time-courses for promethazine and thioperamide had the same profile (Figure 5A). When we compared the e ffects of these drugs on the paw edema generated by LiRecTCTP, we observed a decrease along the time (5–1440 min). Among tested drugs, cromolyn showed the highest inhibition of LiRecTCTP-induced edema (Figure 5A,E). Figure 5B–E show the anti-histaminic effects and their comparison with respective controls: (i) LiRecTCTP or PBS inoculation in animals previously treated with the drug and (ii) preliminary treatment with PBS and following inoculation

of PBS or LiRecTCTP in mice paw. These graphs show there was no significant swelling of the paw after the inoculation of the same volume of PBS and also that the anti-histaminic drugs did not cause unspecific edema. When we compare LiRecTCTP edema curves in the presence of cromolyn and the negative control (PBS), they are very similar (Figure 5E). As observed in permeability assay, cromolyn abrogates LiRecTCTP edematogenic effects.

**Figure 5.** Effect of the treatment with the mast cell degranulation inhibitor or histamine receptor antagonists on edema induced by LiRecTCTP. Mice (*n* = 5) were pre-treated with promethazine (PRO), cimetidine (CIM), thioperamide (THIO), cromolyn (CROM), or PBS (control), and thereafter, animals were injected with LiRecTCTP (10 μg) or PBS (control) into footpads for edema. (**A**) Paw edema observed after injection with LiRecTCTP, in animals previously treated with PBS (LiRecTCTP), mast cell degranulation inhibitor (CRO), or histamine receptor antagonists (PRO, CIM, THIO). (**B**) Paw edema observed after injection with LiRecTCTP or PBS, in animals previously treated with PBS or promethazine (PRO). (**C**) Paw edema observed after injection with LiRecTCTP or PBS, in animals previously treated with PBS or thioperamide (THIO). (**D**) Paw edema observed after injection with LiRecTCTP or PBS, in animals previously treated with PBS or cimetidine (CIM). (**E**) Paw edema observed after injection with LiRecTCTP or PBS, in animals previously treated with PBS or cromolyn (CROM). Values represent the thickness difference between the edema after injection with LiRecTCTP and initial before injections. Each point represents the mean ± s.e.m of five animals from one representative experiment from three independent biological replicates (\* *p* < 0.1; \*\* *p* < 0.01; \*\*\* *p* < 0.001 and \*\*\*\* *p* < 0.0001).

## *3.5. LiRecTCTP In Vivo E*ff*ects—Dermonecrotic Lesion*

We analyzed the role of LiRecTCTP in the dermonecrotic lesion provoked by *Loxosceles* spiders bite accidents. These necrotic lesions are the hallmark of cutaneous manifestation of *Loxosceles* envenomation. The most studied class of *Loxosceles* toxins is phospholipases-D (also called dermonecrotic toxins) whose biological effects can reproduce the ones observed by crude venom inoculation in rabbit skin [7]. In these experiments, we used LiRecTCTP together with an isoform of *L. intermedia* dermonecrotic toxin, LiRecDT1 [28], to evaluate the synergistic action of these toxins. As a negative control, we used an inactive recombinant protein, GFP, which was produced and purified using the same conditions as LiRecTCTP and LiRecDT1. After the subcutaneous inoculation of toxins in rabbit skin, the site was photographed at the time of inoculation and after 24 h for macroscopic evaluation (Figure 6A). After this time, skin patches were collected and processed for microscopic evaluation by histology analyses (Figure 7). Negative controls show that inoculation (PBS or GFP protein) did not trigger any macroscopic (Figure 6A) or microscopic (Figure 7A,A1) alterations in rabbit skin during our experiment. PBS control did not alter normal skin histology (data not show). Dermonecrotic toxin (LiRecDT1) triggered the hallmark lesion, presenting gravitational spreading (Figure 6A) and a marked inflammatory response, which can be observed in the histological analyses by a grea<sup>t</sup> number of neutrophils around blood vessels and diffused to connective tissue surrounding the inoculating site, and development of edema (Figure 7B,B1). As expected, LiRecTCTP alone did not cause a skin lesion, but dose-dependent erythema and edema at inoculation site were observed (Figure 6A). The swelling caused by the toxin is visualized in the histology sections, where the width of the skin is greater in LiRecTCTP samples (Figure 7C,D), when compared to GFP recombinant protein (Figure 7A, negative control) and LiRecDT1 (Figure 7B). Image analysis of histological sections revealed that LiRecDT1 promoted an increase of 41% in the tissue area (from epidermis to adipose) when compared to GFP. LiRecTCTP triggered increased edema of 87% and 91% (compared to GFP), when 10 and 20 μg were used, respectively. The combination of LiRecDT1 and LiRecTCTP 20 μg provoked an increased edema area of 340% when compared to GFP.

**Figure 6.** Inflammatory response of combined recombinant toxins (LiRecTCTP and LiRecDT1) in vivo. (**A**) Macroscopic evaluation of rabbit skin exposed to recombinant toxins (LiRecTCTP, LiRecDT1, or combined toxins LiRecTCTP/LiRecDT1). Rabbits were subcutaneously injected with dermonecrotic toxin LiRecDT1 (1 μg, as positive control), LiRecTCTP (10 and 20 μg), LiRecDT1 (1 μg) combined with LiRecTCTP (10 and 20 μg), GFP (20 μg), a recombinant inactive protein (negative control), or PBS (negative control) (<sup>+</sup>, present; −, absent). Animal skins were photographed just after inoculation (0 h) and 24 h following injection. The same animal received the seven samples for adequate comparison, experiment was repeated twice, using 2 and 4 rabbits respectively. (**B**) Inflammatory reactions induced by toxins and controls were estimated by measurement of myeloperoxidase activity from neutrophils infiltrate at dermis. Values are expressed as mean ± s.e.m of absorbance at 610 nm. Each point represents the average of three replicates from the inoculation site on rabbit skin at the end of experiment (24 h) (\*\* *p* < 0.01 and \*\*\*\* *p* < 0.0001). (**C**) Effect of LiRecTCTP and LiRecDT1 on vascular permeability of skin vessels. Mice were injected intradermally with of LiRecTCTP (5 or 10 μg), LiRecDT1 (1 μg), or recombinant GFP (10 μg) (negative control). PBS was used as a vehicle control. Experiment was performed three times using groups of five mice for each condition. Dye leakage induced by LiRecTCTP combined with LiRecDT1 is higher than the leakage observed with each toxin alone. Scale bar points 0.2 cm.

**Figure 7.** Microscopic evaluation of rabbit skin exposed to recombinant toxins (LiRecTCTP, LiRecDT1, or combined toxins LiRecTCTP/LiRecDT1). Light microscopic analysis of tissue sections was performed on rabbit skin after 24 h of injection. The tissue sections were stained with hematoxylin and eosin. Edema triggered in rabbit skin by (**A**) GFP, (**B**) LiRecDT1 (1 μg), (**C**) LiRecTCTP (10 μg), (**D**) LiRecTCTP (20 μg), and (**E**) the combination of LiRecDT1 (1 μg) and LiRecTCTP (20 μg), as visualized by skin thickness. Skin structures are compared via scanning of images from epidermal (on the right of figure) to muscular tissues (on the left of figure) under the same laboratory conditions (Scale bars indicate 500 μm). The width of the tissue (**E**) points to a deep edema after LiRecTCTP and LiRecDT1 combination when compared to toxins alone (**B**, **D**). Isolated LiRecTCTP (**C**, **D**) induced a higher edema compared to LiRecDT1 alone (**B**) or negative control GFP (**A**), which shows a normal skin histology (**A1**). An intense inflammatory response with the presence of neutrophils and fibrinoid exudates into the dermis is shown when both toxins were administered (**E1**, **E2**) compared to isolated LiRecTCTP (**C1**, **D1**) or LiRecDT1 (**B1**) (Scale bars indicate 100 μm). Closed arrows indicate disorganization of collagen fibers and dermal edema, closed arrowheads indicate a massive inflammatory response with the presence of neutrophils, and open arrows indicate fibrin network deposition. Thickness of the skin tissue (**A**–**E**) is shown in the bottom of each tissue section (μm).

Furthermore, dermal edema and disorganization and separation of collagen fibers provide evidence for this swelling (Figure 7C1,D1). The combined use of LiRecDT1 and LiRecTCTP toxins has a synergistic effect on dermonecrosis development—a higher gravitational spreading (Figure 6A) and an increased inflammatory response (Figure 7E) when compared to LiRecDT1 alone (Figure 7B). There are more leucocytes recruited at the lesion site, notably in the papillary dermis, when LiRecTCTP is injected together with the dermonecrotic toxin (Figure 7E), and this number is directly dependent on the dose. Capillary changes with increased permeability, resulting in the passage of plasma into the connective tissue, evidenced by fibrin network formation, is also observed in the presence of both toxins (Figure 7E,E1).

The edema observed in the presence of both toxins was enormous; although macroscopic images do not reveal clearly, this aspect can be verified by respective histology sections, e.g., massive increase in the width of the skin, disorganization of collagen fibers and dermal edema (Figure 7E). In order to quantify the inflammatory response triggered by the toxins in rabbit skin, we evaluate myeloperoxidase activity in the skin patches that were collected (Figure 6B). We can observe an increase of 25% in myeloperoxidase activity when LiRecDT1 was combined with 10 μg LiRecTCTP when compared to LiRecDT1 alone. When 20 μg of LiRecTCTP was used, the increase reaches 65%. LiRecTCTP alone has no significant effect on myeloperoxidase activity; these results were similar to the negative controls (PBS and GFP). We also investigated the synergy of LiRecDT1 and LiRecTCTP in the vascular permeability using mice. Figure 6C shows representative images from Evans Blue assay; dye leakage can be observed after the inoculation of both toxins. This permeability is increased when toxins are administered together, and this was dependent of LiRecTCTP dose. Combination of 10 μg of LiRecTCTP and 1 μg LiRecDT1 administration resulted in a huge and intense blue spot at the inoculation site. There was no relevant extravasation of Evans blue in negative controls (PBS and GFP). A recombinant GFP obtained in the same heterologous system as the *Loxosceles* toxins was used to rule out an unspecific effect due to bacterial contaminants.
