BoaγPLI: Structural and functional characterization of the gamma phospholipase A2 plasma inhibitor from the non-venomous Brazilian snake Boa constrictor

Plasma in several organisms has components that promote resistance to envenomation by inhibiting specific proteins from snake venoms, such as phospholipases A2 (PLA2s). The major hypothesis for inhibitor’s presence would be the protection against self-envenomation in venomous snakes, but the occurrence of inhibitors in non-venomous snakes and other animals has opened new perspectives for this molecule. Thus, this study showed for the first time the structural and functional characterization of the PLA2 inhibitor from the Boa constrictor serum (BoaγPLI), a non-venomous snake that dwells extensively the Brazilian territory. Therefore, the inhibitor was isolated from B. constrictor serum, with 0.63% of recovery. SDS-PAGE showed a band at ~25 kDa under reducing conditions and ~20 kDa under non-reducing conditions. Chromatographic analyses showed the presence of oligomers formed by BoaγPLI. Primary structure of BoaγPLI suggested an estimated molecular mass of 22 kDa. When BoaγPLI was incubated with Asp-49 and Lys-49 PLA2 there was no severe change in its dichroism spectrum, suggesting a non-covalent interaction. The enzymatic assay showed a dose-dependent inhibition, up to 48.2%, when BoaγPLI was incubated with Asp-49 PLA2, since Lys-49 PLA2 has a lack of enzymatic activity. The edematogenic and myotoxic effects of PLA2s were also inhibited by BoaγPLI. In summary, the present work provides new insights into inhibitors from non-venomous snakes, which possess PLIs in their plasma, although the contact with venom is unlikely.


Introduction
Snake envenomation, reclassified as a neglected tropical disease by the World Health Organization (WHO), can have serious pathophysiological consequences [1][2][3] use of experimental animals and with the ethical principles adopted by the Brazilian College of Animal Experimentation (COBEA).

Snake venom
Snake venoms from Crotalus durissus terrificus (three captivite animals) and Bothrops jararacussu (three captivite animals) were provided by the Laboratory of Herpetology at Instituto Butantan. Venoms were obtained by manual extraction, centrifugated at 1700 g for 15 minutes, lyophilized and stored at -20˚C.

Serum
The blood of five Boa constrictor from captivity was collected by puncture of the paravertebral vein using plastic syringes and pooled. The volume of blood collected corresponded to 1% of snake total weight. Blood was maintained for 18 hours at 4˚C for coagulation, prior to centrifugation (1200 g, 15 minutes) and stored at -20˚C.

Purification of PLI
The purification of PLI from Boa constrictor serum was performed in two chromatographic steps according to Serino-Silva et al., (2018) [30], with some modifications. In the first step, the Boa constrictor serum (5 mL) was diluted in 5 mL of 25 mM Tris buffer pH 7.5 (buffer A) and applied to an anion exchange column (HiTrap DEAE FF 5 mL, GE Healthcare), previously equilibrated with 95% buffer A (25 mM Tris, pH 7.5) and 5% buffer B (25 mM Tris, 1 M NaCl pH 7.5). Elution was performed maintaining 10% buffer B (100 mM NaCl) and then with a gradient of buffer B up to 50% (500 mM NaCl). The run was maintained at a flow rate of 1 mL/min, monitored at 280 nm, and the samples were fractionated every 5 mL (Akta purifier, GE Healthcare). After the chromatography, the protein fractions were pooled. Pool desalination was done by dialysis in PBS (140 mM NaCl, 2.6 mM KCl, 10 mM Na 2 HPO 4 , 1.7 mM KH 2 PO 4 , pH 7.4) for 24 hours at 4˚C in a 10000 MWCO membrane. Then, the D2 fraction from DEAE chromatography was applied to an affinity column, previously prepared by the coupling of crotoxin into a CNBr-activated Sepharose matrix (GE Healthcare) and equilibrated with PBS. The non-adsorbed material was removed by extensive washing with PBS.
Finally, the proteins adsorbed to the resin were eluted with 0.1 M glycine pH 2.7, fractionated in microtubes, and the pH of the fractions was neutralized by the addition of 1 M Tris buffer pH 8.8 (9:1 v/v). The elution was manually measured in a spectrophotometer (Spectramax, Molecular Device) at 280 nm.

SDS-containing polyacrylamide gel electrophoresis (SDS-PAGE)
PLI samples were subjected to 12% SDS-PAGE, according to Laemmli, (1970) [31] under reducing by β-mercaptoethanol or non-reducing conditions. By lane, 20 μg of protein was applied. The molecular marker used was Dual Color Precision Plus, BioRad. The gels were stained using Coomassie Blue R350 (GE Healthcare).

Mass spectrometry
Protein bands were excised from SDS-PAGE (under reducing conditions) which were dehydrated with acetonitrile addition and subjected to reduction with 5 mM dithiothreitol for 30 min at 60˚C, alkylation with 15 mM iodoacetamide for 30 min under light protection at room temperature and overnight in-gel digestion with sequencing grade trypsin (Sigma), in 50 mM ammonium bicarbonate at 37˚C. Digested peptides were analyzed by a Synapt G2 mass spectrometer coupled to a nanoAcquity UPLC system (Waters). Samples were injected into a trap column (C18 nanoAcquity trap Symmetry column 180 μm x 20 mm, Waters) with 0.1% (v / v) formic acid. Peptides were eluted on a capillary analytical column (C18 nanoAcquity BEH 75 μm x 150 mm, 1.7 μm column) using a gradient of 93% A (0.1% formic acid) and 7% B (99.9% ACN 0.1% formic acid) to 35% B over 30 min in a flow of 275 nl / min. Data were acquired in MS E mode [32,33] in duplicate. Protein identification, PTM and homology searches were performed in PEAKS Studio 7.5 software (Bioinformatics Solutions Inc.) by MS/MS search against the Serpentes databases obtained from Uniprot (2844 sequences, downloaded in October 25 th , 2018). Analyses were carried out with precursor mass tolerance of 10 ppm, fragment mass tolerance of 0.025 Da and peptide cleavage by trypsin. Carbamidomethylation of the cysteines was considered as a fixed modification and the oxidation of methionines, N-terminal acetylation, and deamidation of asparagines and glutamines as variable modifications. Assignments for peptides and proteins were accepted at a false discovery rate < 1%. As the inhibitor was indentified with γPLIs, it was named as BoaγPLI.

Circular dichroism
The Asp-49 and Lys-49 PLA 2 (30 μg), BoaγPLI (30 μg) or the incubated mixture (PLA 2 -PLI, 1:1) (w/w) was subjected to circular dichroism assay (J815 spectropolarimeter, Jasco). Proteins diluted in 0.002 M Tris 0.015 M NaCl 0.1 mM CaCl 2, pH 8 were run at wavelengths of 190 to 260 nm (1 nm/s) under 8 convolutions. All analyses were subtracted from the buffer used. The data were exported by Spectra manager and the relative percentages of secondary protein structures were determined by Circular Dichroism analysis using Neural Networks (CDNN) software.

Inhibition of the PLA 2 enzymatic activity
The PLA 2 activity was performed according to Holzer and Mackessy, (1996) [34]. The enzyme (C.d. terrificus Asp-49 PLA 2, 1 mg/mL) and the inhibitor (with a range of concentrations of 2 mg/mL, 1.5 mg/mL, 1 mg/mL and 0.5 mg/mL) were incubated for 10 minutes. Then, 20 μL of each solution were applied to the plate. The chromogenic substrate 4-nitro-3-octanoyloxy benzoic acid (NOB) (Enzo Life Sciences) (solubilized in 3 mM acetonitrile PA) was used (20 μL). Afterwards, the buffer 0.01 M Tris-HCl, 0.10 M NaCl and 0.01 M CaCl 2, pH 8 was added to the samples (200 μL). The controls were composed by 20 μL of saline 0.85%, 20 μL of inhibitor (2 mg/mL) or 20 μL of PLA 2 (1 mg/mL) . The activity was analyzed by spectrophotometer Spectramax (Molecular Devices) at 425 nm, and the readings occurred over 90 minutes, with 5-minute intervals between readings. The absorbances of the last reading were transformed into the PLA 2 specific activity as nmol/mg/min per unit. Then, the percentage of inhibition was determined. The experiment was done in triplicate.

Paw edema inhibition
The paw edema was induced by a subplantar injection in male Swiss mice (18 to 21 g; n = 5) of 10 μg of Asp-49 or Lys-49 PLA 2 , previously incubated for 30 min with 20 μg of BoaγPLI with a final volume injection of 20 μL. The contralateral paw was injected with the same volume of sterile 0.85% NaCl solution. Control groups were inoculated with 20 μg of BoaγPLI, 0.85% NaCl, or 10 μg of Asp-49 or Lys-49. Paw thickness was measured using a caliper reading to 0.01 mm at 0, 0.5, 1, 2, 4, 6 and 24 h after injection. Results were expressed as the difference in thickness of both paws and represented as the percentage increase in paw thickness.

Myotoxicity inhibition
The myotoxicity inhibition by BoaγPLI was determined according to Belchor et al. (2017) [35], by the injection of 40 μL of Asp-49 or Lys-49 PLA 2 (10 μg) incubated for 30 min with BoaγPLI (20 μg) on the right gastrocnemius muscle (18 to 21 g male Swiss mice, n = 5). Control groups received 20 μg of Asp-49 or Lys-49 PLA 2 or 20 μg of BoaγPLI or 40 μL of 0.85% saline. Mice blood samples were collected from the tip of the tail [36] into tubes containing citrate as anticoagulant, centrifuged at 1200 g for 15 minutes and the plasma was separated. The amount of Creatine Kinase (CK) present in the samples was estimated with a commercial CK kit (Sigma), according to the manufacture's instructions.

Statistical analyses
The data were expressed as mean ± standard deviation (SD). The enzymatic and myotoxic assays were analyzed with one-way ANOVA, with Tukey as a posteriori test, while the edematogenic test was analyzed with a two-way ANOVA, with Tukey as a posteriori test. The inhibition of the edema activity was specified by the area under the curve analysis. Values of p<0.05 was considered significant.

Results and discussion
This work shows for the first time the structural and functional characterization of the Boaγ-PLI, a PLI from Boa constrictor, a non-venomous Brazilian snake. Contrasting with the number of PLIs isolated from venomous snakes [14,16], only a small number of this molecule were isolated from non-venomous snakes [19,22], suggesting that the PLI is not restricted to protection against self-envenomation, but it may be involved in other unknown physiological mechanisms that has yet to be determined [23].
In this study, was performed using two chromatographic steps (Fig 1) and the percentage of recovery was 0.63% (Table 1). a reasonable value considering that the inhibitor is a minor protein of the serum, especially in a non-venomous species, whose contact with venom is unlikely.  However, the value was superior to the percentage of recovery from Phyton reticulatus' PLI, which was 0.25% [37], although the distinct methodology applied could have affected the recovery.
In comparison, the purification of the γPLI from Bothrops jararaca (γBjPLI) presented 1% of recovery, using the same methodology applied in this case [30], while under different approaches the γPLI from Crotalus durissus collilineatus (γCdcPLI) presented 2.69% [38]. The higher recovery of PLIs from venomous snakes may be related to the constant contact with the venom, which may stimulate the presence of inhibitors in the plasma [39].
In fact, a difference in the PLIs recovery percentage between two non-venomous snakes of the same genus (Elaphe quadrivirgata and Elaphe climacophora) was observed. This behavior can be related to its dietary habits, since E. quadrivirgata may be ophiophagus, which would result in an indirect contact with venom. Consequently, this animal can have its inhibitors positively regulated [19]. Nevertheless, contact with venom may not be the only mechanism related to the presence of PLIs in snake's plasma.
The SDS-PAGE of BoaγPLI after elution of affinity chromatography, showed a band at 25 kDa, under reducing conditions, and 20 kDa, under non-reducing conditions. Actually, data in the literature suggest a monomerization when inhibitors had contact with PLA 2 [40]. Thus, since affinity chromatography has crotoxin coupled in the resin, the contact of the D2 fraction with it probably monomerizes the inhibitor, explaining the result found in the electrophoresis (Fig 1, insert). BoaγPLI had an anomalous migration under reducing conditions, in which it has a higher molecular mass than under non-reducing conditions. Molecular interactions of PLI, such as non-covalent bonds and disulfide bonds, might facilitate migration under nonreducing conditions. When the molecule is linearized, in turn, a slower migration occurs [40]. In addition, the same phenomenon was observed with γBjPLI [30].
When the inhibitor was identified by mass spectrometry, the peptides identified presented similarities with 6 previously described PLIs and it is possible to observe that the identified sequence of BoaγPLI is well conserved compared to other PLIs (Table 2). In addition, the γPLI sequence of Lachesis muta was identified through its transcript [41], having its signal peptide of 19 amino acids, which is absent in a purified plasma protein. Disregarding the signal peptide, the coverage increases to 64%.
The oligopeptide 104 QPFPGLPLSRPNGYY 118 was suggested as a site of interaction between γPLIs and PLA 2 s in Bothrops sp [42] and this amino acid sequence was partially identified in BoaγPLI (part of this sequence shown in Fig 2), which reinforces the conserved primary structure of these inhibitors [30,40]. Some authors also suggest that there may be three phosphorylation sites, at position 21 S, 22 S and 111 T, and that these sites may be involved in other physiological roles, beyond the inactivation with PLA 2 , which would reinforce the hypothesis about their presence in plasma of non-venomous snakes. So far, no other role has been assigned to PLIs, and these phosphorylation sites were not found in BoaγPLI partial sequence [30,40,42].
The size-exclusion chromatography showed two peaks corresponding to oligomer and monomer, respectively (Fig 3A). This result is corroborated by the γPLI homologue from Python sabae, that also showed oligomers and monomers in its spectrum [43]. On the other hand, γBjussuMIP [40] showed only the oligomer, as well as the purified γPLI of Macropisthodon rudis [20,40]. Our group has found only the monomeric form of γBjPLI. Moreover, the interaction of γBjPLI with Asp-49 PLA 2 was also evidenced in this work [30].
BoaγPLI secondary structure was investigated by circular dichroism (Fig 3B), presenting 21.4% of alpha-helices, 39.4% of beta-sheets and 38.7% of random coils. Besides that, incubation of BoaγPLI with Asp-49 and Lys-49 PLA 2 s does not significantly alter the spectrum shown in circular dichroism, which was also evidenced in γCdcPLI and γBjussuMIP [38,40]. This observation suggests a weak interaction of BoaγPLI with PLA 2 , most likely non-covalently. The PLI of Macropisthodon rudis also showed non-covalent binding characteristics evidenced by SDS-PAGE when incubated with PLA 2 under non-reducing conditions [20].
Once the possible interaction between BoaγPLI and PLA 2 s was evidenced, an inhibition assay was performed (Fig 4A), resulting in a dose-dependent inhibition, reaching 48.7% when  40 μg of inhibitor were used, values similar to found for γBjPLI [30]. In contrast, a γPLI inhibitor from Python sabae showed a low enzymatic activity inhibition of PLA 2 from bee venom. In turn, PIP, the γPLI isolated from Python reticulatus, showed a high inhibition at 1:1 molar ratio, reaching 90%, measured by egg yolk acidimetric method, with PLA 2 of D. r. russelli [37]. The inhibitory potential of BoaγPLI was also verified considering edematogenic and myotoxic effects of Asp-49 and Lys-49 PLA 2 s. Wherein, when incubated with the inhibitor, edema was significantly lower compared to isolated PLA 2 s. In order to highlight the differences between the curves of treatments, the area under the curve was estimated. As expected, edema profile induced by isolated PLA 2 s resulted in a higher area under the curve when compared to the groups that received PLA 2 s pre-incubated with BoaγPLI (Control: 58.66; BoaγPLI: 43.51; Asp-49 PLA 2 : 206.0; BoaγPLI + Asp-49 PLA 2 : 134.8; Lys-49 PLA 2 : 126.5; BoaγPLI + Lys-49 PLA 2 : 105.7). (Fig 4B). Creatine kinase levels, which reflects myotoxicity, were also reduced when PLA 2 s were incubated with the inhibitor (Fig 4C). Other γPLIs from venomous and non-venomous snakes also demonstrated the inhibitory potential in the pharmacological activities caused by PLA 2 [30,38,40]. In addition, our data showed no statistical differences of  the inhibition potential of BoaγPLI against Asp-49 and Lys-49 PLA 2 s, unlike γBjussuMIP, which possess higher affinity for Asp-49 PLA 2 [40] . An interesting point of the study is that, even being a non-venomous snake, the functional and structural characteristics of BoaγPLI were well conserved when compared with γPLIs from venomous snakes.

Concluding remarks
In summary, the study showed the purification of a γPLI of the non-venomous snake Boa constrictor (BoaγPLI), which has a molecular mass of approximately 22 kDa, oligomerization capacity, and a primary structure similar to the PLI of Lachesis muta. The direct interaction of BoaγPLI with Asp-49 PLA 2 of C. d. terrificus and Lys-49 PLA 2 from Bothrops jararacussu was evidenced by circular dichroism, and the molecule displayed inhibition upon enzymatic, edematogenic and myotoxic activities of PLA 2 s.
An interesting point of this study is that our data add another piece of evidence pointing the wide distribution of these inhibitors, that appears in venomous and non-venomous snakes, and, beyond that, their structure and inhibitory activity seems to be well conserved between them. That evidence amplifies the primary hypothesis for the PLI presence, that was the protection against the self-envenomation. Therefore, the inhibitor is not restricted to such function.
In addition, since γPLIs inhibit PLA 2 s from IIA group, which includes proinflammatory PLA 2 s from mammalian and from viperids, the isolation and characterization of distinct γPLIs may contribute to the enrichment of information for the bioprospection that can be useful, not only in antivenom therapy, but also in other inflammatory processes triggered by human PLA 2 s, whereas their catalytic site is well conserved [8,42]. In this context, the present study may contribute to the elucidation of the presence of PLA 2 inhibitors in non-venomous snakes and to provide new perspectives for PLA 2 inhibitors from snake plasma by characterizing and comparing the similarities and differences between PLI from venomous and non-venomous snakes.