Chemical Modifications of PhTX-I Myotoxin from Porthidium hyoprora Snake Venom: Effects on Structural, Enzymatic, and Pharmacological Properties

We recently described the isolation of a basic PLA2 (PhTX-I) from Porthidium hyoprora snake venom. This toxin exhibits high catalytic activity, induces in vivo myotoxicity, moderates footpad edema, and causes in vitro neuromuscular blockade. Here, we describe the chemical modifications of specific amino acid residues (His, Tyr, Lys, and Trp), performed in PhTX-I, to study their effects on the structural, enzymatic, and pharmacological properties of this myotoxin. After chemical treatment, a single His, 4 Tyr, 7 Lys, and one Trp residues were modified. The secondary structure of the protein remained unchanged as measured by circular dichroism; however other results indicated the critical role played by Lys and Tyr residues in myotoxic, neurotoxic activities and mainly in the cytotoxicity displayed by PhTX-I. His residue and therefore catalytic activity of PhTX-I are relevant for edematogenic, neurotoxic, and myotoxic effects, but not for its cytotoxic activity. This dissociation observed between enzymatic activity and some pharmacological effects suggests that other molecular regions distinct from the catalytic site may also play a role in the toxic activities exerted by this myotoxin. Our observations supported the hypothesis that both the catalytic sites as the hypothetical pharmacological sites are relevant to the pharmacological profile of PhTX-I.

Within the family Viperidae, two distinct types of venom PLA 2 molecules have been described, all of them sharing a high degree of homology both in primary and threedimensional structure [13]: the "classical" PLA 2 , that present an invariant Asp49 residue that plays a key role in catalysis; and the "PLA 2 homologues, " devoid of enzymatic activity, that present the substitution of Asp49 by Lys49 or, less frequently, by Ser, Arg, Gln, or Asn [14,15]. Despite their difference in catalysis, both the Asp49 and the Lys49 proteins are able to induce various pharmacological effects [16]. us the structure-function relationship among this group of proteins is subtle and complicated [1]. In some cases, the pharmacological effects result from their enzymatic activities, probably through the action of the products of hydrolysis, lysophospholipids, and fatty acids, that alter cell membrane shape and permeability [17,18] but for many of them, the pharmacological effects are independent of their enzymatic activities, such as Lys49 PLA 2 myotoxins, which lack hydrolytic activity and therefore act via another mechanism, which is only partially understood. A site close to the C-terminus, comprising a variable combination of basic and hydrophobic amino acids, has been identi�ed as being responsible for toxicity [19].
In a previous work, we showed that Porthidium hyoprora snake venom is a rich source of PLA 2 enzymes. Additionally, we puri�ed a myotoxic PLA 2 (PhTX-I) to homogeneity in reverse-phase HPLC, which constitutes of a single polypeptidic chain, has a molecular mass of 14.249 Da, and whose amino acid sequence exhibits high identity with other myotoxic Asp49 PLA 2 [20]. We also demonstrated that PhTX-I (20 g/mL) caused edema, in vivo creatine kinase release, C2C12 skeletal muscle myoblasts cytotoxicity, and neuromuscular blockade of chick biventer cervicis muscle preparations. However, it is still unknown whether these pharmacological effects were mediated by the phospholipase catalytic activity of PhTX-I or not. One of the strategies employed for the elucidation of the relationship between catalytic activity and pharmacological effects of PLA 2 is based on the chemical modi�cation of speci�c residues in these enzymes [21]. Using this approach, a dissociation of pharmacological effects and enzymatic activity for various PLA 2 has been observed, suggesting the presence of separate enzymatic and pharmacological active site(s) contained within their amino acid sequences [22,23]. In the present study, we investigated the effects of chemical modi�cations of speci�c amino acid residues (His, Tyr, Lys, and Trp), performed in PhTX-I, on their enzymatic, structural, and pharmacological properties.  [20]. Brie�y, 5 mg of whole venom was dissolved in 200 L of buffer A (0.1% TFA) and centrifuged at 4500 g; the supernatant was then applied to a -Bondapak C18 column (0.78 × 30 cm; Waters 991-PDA system), previously equilibrated with buffer A for 15 min. e elution of the protein was then conducted using a linear gradient (0%-100%, v/v) of buffer B (66.5% Acetonitrile in buffer A) at a constant �ow rate of 1.0 mL/min. e chromatographic run was monitored at 280 nm of absorbance, and aer elution the fraction was lyophilized and stored at −40 ∘ C.

2.�. �hemical Modi�cations.
Modi�cation of His residues with 2,4 � -Dibromoacetophenone (BPB) was carried out as previously described [21]. Brie�y, 3 mg of PhTX-I were dissolved in 1 mL of 0.1 M Tris-HCl containing 0.7 mM EDTA (pH 8.0) and 150 L of BPB (1.5 mg/mL, in ethanol), and the mixture incubated for 24 h at 25 ∘ C. Modi�cation of Lys residues with acetic anhydride (AA) was performed at a protein : reagent molar ratio of 1 : 50 [21]. PhTX-I (3 mg) was dissolved in 1.5 mL of 0.2 M Tris-HCl buffer at pH 8.0, and 10 L of AA was added and the mixture was incubated for 1 h at 25 ∘ C. Tyr residues were modi�ed by treatment with 4-nitrobenzenesulphonyl �uoride (NBSF) as previously described [24]. Brie�y, 1 mol of PhTX-I (10 mol of Tyr) was dissolved in 14 mL of 0.1 M Tris-HCl (pH 8.0) and incubated with 10 mol, of NBSF for 20 h at 25 ∘ C. Modi�cation of Trp residues was performed according to Takasaki et al. [25]. Brie�y, 9 mg of PhTX-I were dissolved in 4 mL 50% acetic acid containing 1 mg of 2-nitrobenzenesulfonyl chloride (NPSC) and incubated for 1 h at 25 ∘ C. In all cases, excess reagent was removed by ultra�ltration through a Millipore's Amicon Ultra-15 membrane and washed with distilled water, followed by lyophilization.

Amino Acid Analysis.
Amino acid analysis was performed on a Pico-Tag Analyzer (Waters Systems) as described by Heinrikson and Meredith [26]. Native PhTX-I PLA 2 and their modi�ed derived samples (30 g) were hydrolyzed at 105 ∘ C for 24 h, in 6 M HCl (Pierce sequencing grade) containing 1% phenol (w/v). e hydrolysates were reacted with 20 L of derivatization solution (ethanol : triethylamine : water : phenylisothiocyanate, 7 : 1 : 1 : 1, v/v) for 1 h at room temperature, aer which the PTC amino acids were identi�ed and quanti�ed by HPLC, by comparing their retention times and peak areas with those from a standard amino acid mixture.

Mass
Spectrometry. An aliquot (4.5 L) of the modi�ed proteins was injected in C18 (100 m × 100 mm) RP-UPLC (nanoAcquity UPLC, Waters) coupled with nanoelectrospray tandem mass spectrometry on a Q-Tof Ultima API mass spectrometer (MicroMass/Waters) at a �ow rate of 600 nL/min. e gradient was 0%-50% acetonitrile in 0.1% formic acid over 45 min. e instrument was operated in MS continuum mode, and the data acquisition was from m/z 100-3.000 at a scan rate of 1 s and an interscan delay of 0.1 s. e spectra were accumulated over about 300 scans and the multiple charged data by the mass spectrometer on the m/z scale were converted to the mass (molecular weight) scale using maximum entropy-based soware supplied with Masslynx 4.1 soware package. e processing parameters were output mass range 6.000-20.000 Da at a "resolution" of 0.1 Da/channel; the simulated isotope pattern model was used with the spectrum blur width parameter set to 0.2 Da; the minimum intensity ratios between successive peaks were 20% (le and right). e deconvoluted spectrum was then smoothed (2 × 3 channels, Savitzky Golay smooth) and the mass centroid values obtained using 80% of the peak top and a minimum peak width at half height of 4 channels.

Circular
Dichroism. Circular dichroism (CD) spectra of native PhTX-I PLA 2 and their modi�ed derivatives were recorded with a JASCO model J-720-ORD 306 spectropolarimeter equipped with a thermoelectric sample temperature controller (Peltier system) following standard procedures previously described [28]. Aer centrifugation at 4000 g for 5 min, samples (1-4 M protein in 10 mM sodium phosphate, pH 8) were transferred to a 10-mm path-length quartz cuvette. Circular dichroism spectra in the wavelength range 260 to 200 nm were collected, using a bandwidth of 1 nm and a response time of 1 s. Data collection was performed at 25 ∘ C with 50 nm/min scanning speed. At least ten scans were accumulated for each sample, and all spectra were corrected by subtraction of buffer blanks. e estimation of secondary structure elements was performed using the CDNN Deconvolution soware (version 2.1), and Origin 7.5 (OriginLab) was used for graphics and analysis.
2.8. PLA 2 Activity. PLA 2 activity was measured using the assay described by Cho and Kezdy [29] and Holzer and Mackessy [30] modi�ed for 96-well plates. e standard assay mixture contained 200 L of buffer (10 mM Tris-HCl, 10 mM CaCl 2 , and 100 mM NaCl, pH 8.0), 20 L of substrate 4nitro-3-(octanoyloxy) benzoic acid (3 mM), 20 L of water, and 20 L of PhTX-I PLA 2 or their modi�ed derivatives (1 mg/mL) in a �nal volume of 260 L. Aer adding proteins (20 g) the mixture was incubated for up to 40 min at 37 ∘ C, measuring absorbance at intervals of 10 min. e enzyme activity, expressed as the initial velocity of the reaction ( ), was calculated based on the increase of absorbance aer 20 min. All assays were done in triplicate, and the absorbances at 425 nm were measured with a VersaMax 190 multiwell plate reader (Molecular Devices, Sunnyvale, CA).

2.9.
Inhibition. e inhibitory effects of EDTA or low molecular weight heparin from porcine intestinal (Mr 6.000 Da) on pharmacological and enzymatic activities of PhTX-I were assessed by incubating the enzyme with 1 mM solution of this chelating agent or a heparin : toxin molar ratio of 2 : 1 for 30 min at 37 ∘ C. e inhibition of PLA 2 activity of PhTX-I by crotapotins F2 and F3 from Crotalus durissus collilinetaus also was evaluated by incubating the two proteins (1 : 1, w/w) for 30 min at 37 ∘ C and then assaying the residual enzyme activity.

Chick Biventer Cervicis Muscle Preparation (BCP).
Animals were anesthetized with halothane and sacri�ced by exsanguination. e biventer cervicis muscles were removed and mounted under a tension of 0.5 g, in a 5 mL organ bath (automatic organ multiple-bath LE01 Letica Scienti�c Instruments. Barcelona  . ree hours aer injection, blood was collected from the tail into heparinized capillary tubes, and the plasma creatine kinase (CK; EC 2.7.3.2) activity was determined by a kinetic assay (Sigma 47-UV). Activity was expressed in U/L, one unit de�ned as the phosphorylation of 1 mol of creatine/min at 25 ∘ C.

Edema-Forming
Activity. e ability of PhTX-I PLA 2 and their modi�ed derivatives to induce edema was studied in groups of �ve Swiss mice (18-20 g). Fiy L of phosphatebuffered saline (PBS; 0.12 M NaCl, 0.04 M sodium phosphate, pH 7.2) with toxins (1 g/paw) were injected in the subplantar region of the right footpad. e le footpad received 50 L of PBS, as a control. e paw volume was evaluated plethysmographically (Model 7140 Plethysmometer, Ugo Basile, USA), immediately before the injection (basal) and aer 1 h. Edema-forming activity was expressed as the percentage of the increase in volume of the right foot pad in comparison to the le foot pad (control). e equation for calculation of the percentage of edema in toxins injected paw was where is the edema (volume) measured at each time interval and 0 is the volume of the paw (intact, zero time before toxins injection). e percentage of edema calculated was subtracted from the matched values at each time point in the saline injected hind paw (control).  (7) 6.56 (7) 7.15 (7) 7.42 (7) r 6 6.34 (6) 6.09 (6) 6.35 (6) 6. 34 (6) Table 1 shows the result of analysis of amino acid composition of the modi�ed proteins of PhTX-I PLA 2 compared to the amino acid sequence of native PhTX-I. Aer chemical treatment a single His, 4 Tyr, and 7 Lys residues were modi�ed by BPB, NBSF, and AA, respectively. With the methodology employed it was not possible to determine the changes in Trp residues. In all these cases the exact position of the groups modi�ed is unknown. e homogeneity of the modi�ed proteins was evaluated by SDS-PAGE, as shown in Figure 1. To determine whether the chemical modi�cations in PhTX-I caused changes in protein secondary structure, far-UV circular dichroism (CD) was used. Figure 3 shows CD spectra of native PhTX-I and their modi�ed forms which demonstrate two negative bands of similar magnitude (−11,000 to −10,500 deg⋅cm 2 ⋅dmol −1 ) at 208 and 222 nm and a positive one at ∼190 nm (data not shown), indicating a consistent content of -helical structures. e exception was NBSF modi�ed PhTX-I that lad lower signal at 222 nm than the unmodi�ed protein. Based on the CDNN program analysis of the native PhTX-I spectrum, the contents ofhelices, -sheets, and -turns were 32%, 18%, and 17%, respectively. e contents of secondary structure of PhTX-I chemically modi�ed by BPB, NPSC, or AA were very similar to native PhTX-I, with no signi�cant changes in the CD spectrum between them, suggesting that secondary structure of protein remains practically unchanged. However, NBSF treatment resulted in a signi�cant change (−10,500 to −9,000 deg⋅cm 2 ⋅dmol −1 ) in molar ellipticity of the negative band at 222 nm, which by CDNN soware analysis indicates an altered content of -helices, -sheets, and -turns (30%, 19% and 18%, resp.). e catalytic activity of native PhTX-I and their modi�cations were studied using the chromogenic substrate 4-nitro-3-(octanoyloxy) benzoic acid. e catalytic activity of native PhTX-I was almost completely abolished by BPB, but only partially reduced aer modi�cation of Tyr or Lys residues; NPSC did not cause a signi�cant decrease in this activity ( Figure 4). e incubation of native PhTX-I with crotapotins F2 and F3 from C. d. collilinetaus and EDTA diminished the activity; heparin did not signi�cantly inhibit the catalytic activity ( Figure 4). Figure 5 shows the graphical representation of the blockade of the contractile response in the neuromuscular transmission (BCP) of native PhTX-I and modi�ed derivatives. e native PhTX-I at a concentration of 1.4 M blocked the indirectly evoked contractions reaching 50% of the block in about 20 min. is activity was markedly diminished for all the PhTX-I chemically modi�ed derivatives, except for modi�cation of Trp where no signi�cant change in force of contraction was produced by the modi�cation. Figure 6 shows the effect of the different chemical modi�cations on the myotoxic activities of PhTX-I, by time-course measurement of plasma levels CK aer intramuscular injection of proteins. It can be seen that even when BPB and AA were the most effective reagents altering this activity, most of the treatments had some effect. ree hours aer treatment started, the myotoxic effect was reduced by 85%, 80%, and 72% by treatment with AA, BPB, and NBSF, respectively. EDTA and heparin inhibited around 74% and 50% of this activity.

Results
Regarding the edema-inducing effect, aer alkylation with BPB, PhTX-I lost around 70% of its activity one hour aer treatment, while modi�cation of Lys and Tyr residues caused only a partial decrease of this activity (55% and 40%, resp.); however, NPSC did not inhibit this activity, as can be seen in Figure 7.
Native PhTX-I was cytotoxic to NIH-3T3 �broblast cell line and to a lesser extent for the NG97 cell line derived from a human astrocytoma grade III (Figure 8). e cytotoxic activity of PhTX-I was independent of enzymatic activity, since BPB-treated PhTX-I was able to produce cytotoxicity in both cell types. Aer treatment with NPSC the cytotoxic activity was partially reduced. On the other hand, acetylation and sulfonylation of Lys and Tyr residues, respectively, reduced strongly the cytotoxic activity in both cell types.

Discussion
We recently described the isolation of a basic PLA 2 (PhTX-I) from P. hyoprora using reverse phase HPLC [20]. is toxin exhibits high catalytic activity, shares various structural similarities with other "bothropics" PLA 2 , and has the conserved and essential Asp residue in position 49. PhTX-I induces in vivo myotoxicity, moderates footpad edema (at concentrations up to 10 g/mL and 0.5 g/mL, resp.), and causes in vitro neuromuscular blockade in chick biventer cervicis muscle preparations (at concentrations of 1.4 M). Here, we have described the chemical modi�cations of speci�c amino acid residues (His, Tyr, Lys, and Trp), performed in PhTX-I and how theses modi�cations affected the structural, enzymatic, and pharmacological properties of this myotoxin.
A �rst important observation was that aer chemical treatment a single His, 4 Tyr, 7 Lys, and one Trp residues were modi�ed by BPB, NBSF, AA, and NPSC, respectively (Table 1). ese results were con�rmed by mass spectrometry (Figure 2). e mass of native PhTX-I, 14249.22 Da [20], aer treatment with NBSF (15068.96 Da) increased 819.74 Da (Figure 2(d)), demonstrating modi�cation of fourth Tyr residues (up to 187.16 Da in each residue corresponding to reagent NBS that would be incorporated). Similarly, only three Tyr residues (7, 70, and 77) with the highest exposed surface areas in the notexin were modi�ed by NBSF, suggesting that the seven remaining residues are "buried" within the molecule [32]. Acetylation of Lys residues of basic PLA 2 myotoxins three (PrTX-I,-III and BnSP-7) caused a complete loss of basicity, being demonstrated by electrophoretic analysis [22,33]; however AA-treated PhTX-I (14537.9 Da) ( Figure  2(b)) shows that only seven Lys residues were modi�ed, due to which there is an increase of 288.60 Da, equivalent to seven times the mass of acetyl radical (42 Da) incorporated into the amino group of K residues, showing similar behavior to the native protein in electrophoresis gel (Figure 1). Similarly, an increase of 221.39 Da in NPSC-treated PhTX-I (Figure 2(c)) indicated a modi�cation of a single Trp residue. By analogy with other results using PrTX-I and -III from B. pirajai [22], we would expect that, with the three Trp residues present in the structure of PhTX-I, this reagent should modify the residue with a larger area of exposure; it would be more easily attacked by the NPSC. Alkylation of His by BPB has been widely used to assess the role of enzymatic activity in the pharmacological actions of PLA 2 [22,23,[33][34][35][36]. PhTX alkylated with BPB had a molecular mass of 14440.7 Da (Figure 2(a)), which con�rmed the modi�cation of only one residue of His. His48 is a highly conserved amino acid residue in PLA 2 , which has a vital role in catalysis [37]. Since the enzymatic activity of the PhTX-I was almost completely abolished a�er this modi�cation, His48 was likely the residue modi�ed, because this amino acid is part of the catalytic triad of this protein family.
Examination of native PhTX-I by CD spectroscopy indicated that the predominant secondary structure of this PLA 2 consisted of alpha-helices (Figure 3), in agreement with the results obtained for others as PLA 2 from Taiwan cobra [38], PLA2A from C. d. ruruima [39], and BnIV from B. newidii [40]. e secondary structure of PhTX-I chemically modi�ed derivatives did not alter signi�cantly a�er modi�cations as evidenced by the CD spectra, which exhibited almost the same pro�le as that of native toxin. �nly NBSF-treated toxin revealed detectable changes in secondary structure composition when compared to native toxin (Figure 3). is difference could be due to partial helix unfolding but the small change suggested that such unfolding was minimal. Kini [1] described that the alkylation by BPB does not affect the three-dimensional structure of PLA 2 or its ability to bind phospholipids, but may alter the ability to interact with speci�c proteins or ligands. us, it is suggested that the chemical modi�cations performed in this study mainly affected the speci�c residues involved in such modi�cations and did not result in drastic conformational changes in the molecule.
e PhTX-I PLA 2 is a Ca 2+ -dependent enzyme, with maximum activity at pH 8 and 40 ∘ C, reaching max and of 11.76 nmoles/min and 1.96 mM, respectively [20]. Heparin slightly decreased enzymatic activity of PhTX-I, (25%) (Figure 4); similarly, this polyanionic compound resulted acting as negative allosteric modulator of PLA 2 C. d. cascavella [41]. On the other hand, EDTA greatly decreased catalytic activity of PhTX-I (88%), as -bungarotoxin and notexin, which were inhibited by EDTA even in the presence of an excess of Ca 2+ [42]. e F2 and F3 crotapotins from C. d. collilineatus signi�cantly inhibited the enzymatic activity of PhTX-I at approximately 55% (Figure 4), in agreement with BjIV PLA 2 of B. jararacussu, which was inhibited by 50% in its catalytic activity by F7, F3, and F4 crotapotins from C. d. terri�cus, C. d. collilineatus, and C. d. cascavella [43]. ese results suggest that crotapotins can bind to bothropics PLA 2 a manner similar to that of crotalics PLA 2 , and raise the possibility that bothropic venoms may contain crotapotin-like proteins which inhibit the catalytic activity of PLA 2 .
Acetylation of Lys residues signi�cantly reduced the enzymatic activity; however, a residual activity was detected, corresponding to 26% (Figure 4), similarly to MT-III (B. asper) and MT-I (B. godmani) [21]. e mode of speci�c acetylation is not clear, but there is some evidence of reduction of the calcium-binding capacity, thereby reducing the enzymatic activity of PLA 2 [44]. In contrast, both myotoxic and cytotoxic effects were totally abolished, whereas a residual edematogenic effect remained (Figures 6, 7, and 8). ese observations agree with previous studies in which the Lys of PrTX-I, PrTX-III, and BnSP-7 PLA 2 were modi�ed by acetylation [22,33] and drastically decreased myotoxic, edema-inducing, and bactericidal activities. Acetylation of Lys residues in PhTX-I also decreased the blockade of the contractile response in the neuromuscular transmission, more than enzymatic activity ( Figure 5). is greater effect on some pharmacological properties than on the enzymatic activity of Lys-modi�ed PhTX-I demonstrates the dissociation between enzymatic and pharmacological activities and evidence of the existence of molecular regions, distinct from the catalytic site, which may be responsible for at least some of the pharmacological properties of these toxins, in agreement with previous studies [23,34,35].
Soares et al. [22] suggested that the bactericidal effect of PrTX-III and myotoxin III from B. pirajai and B. asper, respectively, might be related to overall basicity of the protein from the N-terminal helix and residues the 115-129 in the C-terminal region that is rich in aromatic and basic residues. is C-terminal region was identi�ed as a structural determinant of this effect [45]. Short synthetic peptides representing the C-terminal region of PLA 2 myotoxins showed cytolytic and muscle damaging activities similar to their parent proteins, although they display a lower potency [46][47][48]. Signi�cant decrease of myotoxic activity aer incubation with heparin con�rms the involvement of C-terminal region of PhTX-I which is a basic myotoxin with high content of Lys residues [20].
His48 is a highly conserved residue in PLA 2 , which has an important role in catalysis [37]. Alkylation of His by BPB has been widely used to assess the role of enzymatic activity in the pharmacological actions of PLA 2 [21,35,36,49]. Here, alkylation of His at the active site of PhTX-I markedly abolished enzymatic activity (<4% residual activity) ( Figure  4). Others have reported residual enzymatic activity following alkylation of His48, as BuTX and notexin from N. nigricollis and N. n. atra PLA 2 , with values close to that found for BPB-PhTX-I (5%-8% of residual activity) [34]. Myotoxic, neurotoxic, and edema-forming activities of PhTX-I, were drastically reduced by this modi�cation (Figures 5, 6, and 7); EDTA by treatment also affected myotoxicity and edematogenic activity (Figures 6 and 7), strengthening the hypothesis that phospholipid enzymatic hydrolysis is involved in these effects. Similarly, neurotoxic and myotoxic activities were inhibited almost completely aer alkylation of His48 of PLA 2 Basp-III (B. asper), PrTX-III (B. pirajai), BthTX-II (B. jararacussu), Cdc-9, and Cdc-10 (C. d. cumanensis) [21,22,35,49], showing that these pharmacological effects are dependent on the catalytic activity. Since alkylation of the active site His48 completely abolished the catalytic activity and strongly attenuated these three pharmacological effects, Kini [1] suggested the hypothesis that PLA 2 activity potentiates these pharmacological effects induced by Bothrops and Crotalus myotoxins.
In contrast, cytotoxic activity upon NG97 and NCIH-3T3 cells not was affected by His modi�cation (Figure 8), suggesting that enzymatic activity is not required for this effects and that there are other molecular regions involved in cellular membrane perturbation. Our results agree with MTX-I and II PLA 2 from B. brazili, which displayed cytotoxic activity against Jurkat lines independently of catalytic activity [50]. Some authors propose that cytotoxic activity on tumor cell lines is associated with apoptosis induction, considering the fact that PLA 2 enzymes have been proposed to play a role in mediating apoptosis in various models, including cell lines [51]. e PLA 2 activity is proposed to accelerate turnover of phospholipids, which may in�uence membrane changes that occur during apoptosis [52]. We suggested important role of Lys residues in cytotoxic effect, because PhTX-I treatment with AA abolished this activity.
Studies have been directed trying to understand the mechanisms involved in the in�ammatory response induced by myotoxic PLA 2 from several snake venoms [53][54][55]. However, the relationship between enzymatic activity and edema is contradictory [56]. It is assumed that myotoxic and edematogenic activities can be induced by different structural domains in these PLA 2 , or that a partial overlapping of these domains occurs [55,57].
Residual enzymatic activity aer sulfonylation of Tyr residues was 38% (Figure 4). Tyr52 and Tyr73 are part of the catalytic site of PLA 2 ; giving structural support to stabilize the catalytic system [37], changes in this system would affect the enzymatic activity. Tyr-modi�ed PhTX-I decreased myotoxic and neurotoxic activities more than the enzymatic activity (Figures 4, 5, and 6), once again indicating the dissociation between enzymatic and pharmacological activities. Cytotoxic activity of PhTX-I also was reduced drastically aer modi�cation by NBSF (Figure 8), indicating that Tyr residues would also be involved in this process. Zhao et al. [58] observed that myotoxic PLA 2 have a set of Tyr residues located at the C-terminal region of the molecule. ese Tyr may contribute to the hydrophobic-cationic combination proposed to play a role in myotoxicity and cytotoxicity [19,46,59]. PhTX-I show a Tyr residue in Cterminal region; alterations of this acid amino in this region of the molecule are causing reduction of toxicity of PhTX-I. However, the in�uence of conformational changes induced by NBSF in these effects cannot be discarded ( Figure 3). NPSC-treated PhTX-I did not signi�cantly decrease enzymatic activity (Figure 4), suggesting that the modi�ed Trp residue is not related to the catalytic system. In the same way, this modi�cation showed no changes as compared to native PhTX-I, in edematogenic and cytotoxic activities (Figures 7 and 8); the myotoxic effect was minimally affected ( Figure 6). In this sense Trp residues of PhTX-I have little or no direct action on the muscle. e Trp modi�cations of PhTX-I also maintained the action upon blockade of the contractile response in chick biventer cervicis muscle preparation ( Figure 5). In contrast, modi�ed Trp affected only the neurotoxic effect caused by MjTX PLA2-II [60], indicating the relevant role of this residue in this activity and suggesting that chemical modi�cation could be interfering with the stability of the interaction between the monomers of this dimeric toxin, since Trp77 helps to maintain the homodimeric interaction. is suggests that the shi from dimeric to monomeric form of myotoxin may reduce the ability to affect the plasma membrane [60]. Because PhTX-I is a monomeric toxin, modi�ed Trp would not affect the pharmacological activity of this toxin.

Conclusions
Our results indicate the critical role played by Lys and Tyr residues in myotoxic, neurotoxic activities and mainly in the cytotoxicity displayed by PhTX-I. His residue and, therefore, catalytic activity of PhTX-I are relevant for edematogenic, neurotoxic, and myotoxic effects, but not for its cytotoxic activity. Our observations supported the existence of pharmacological sites, distinct from the catalytic site, that contribute to the development of toxicity of these toxins and the hypothesis that the catalytic activity would potentiate the myotoxic and neurotoxic effects induced by snake venom PLA 2 . Finally, although a partial dissociation is shown, both the catalytic sites as the hypothetical pharmacological sites are relevant to the pharmacological pro�le of PhTX-I.

Ethical Approval
e animals and research protocols used in this study followed the guidelines of the Ethical Committee for use of animals of ECAE-IB-UNICAMP SP, Brazil (protocol number 1860-1) and international law and policies. All efforts were made to minimize the number of animals used and their suffering.