CR (17.1 µg) and NK (3.4 µg) venoms produced clear zone diameters of 21.1 ± 0.3 mm and 20.6 ± 0.3 mm, respectively, on agarose-erythrocyte egg yolk gel plate and these clear zones were selected as the minimum indirect hemolytic dose for each venom. From the indirect PLA₂ assay, the results revealed that the enzyme from both venoms exhibited indirect hemolytic activity. This was probably due to the activity of PLA₂s in the hydrolysis of lecithins to lysolecithins and lysis of the cell membranes of red blood cells [19]. Figure 1 shows the log dose-response curves of MSKE and its phenolic principles (GA, MG and PGG) on the inhibition of PLA₂s of CR and NK venoms. It can be seen that MSKE and its phenolic principles clearly showed a dose-dependent inhibitory activity against PLA₂s of both venoms. The order of potency as judged by the half-effective dose (ED₅₀) was PGG (25.6 ± 1.3 µg) > MG (42.9 ± 2.3 µg) > MSKE (44.4 ± 0.4 µg) > GA (66.0 ± 2.5 µg) for CR venom, and PGG (30.2 ± 2.4 µg) > MSKE (57.9 ± 0.5 µg) > MG (108.7 ± 2.0 µg) > GA (123.5 ± 2.5 µg) for NK venom (Table 1). The extract, GA, MG and PGG alone and solvent systems had no effect on this study. Of these results implied that the inhibition of PLA₂s of both venoms by MSKE may result in the prevention of red blood cell rupture.

MSKE and its phenolic principles (GA, MG and PGG) inhibited the activity of hyaluronidase in CR and NK venoms in a dose-dependent manner (Figure 2). The order of potency as judged by the half-effective dose (ED₅₀) was PGG (3.4 ± 2.4 µg) > MG (5.9 ± 2.4 µg) > MSKE (6.0 ± 1.9 µg) > GA (16.0 ± 2.5 µg) for CR venom, and PGG (2.6 ± 1.7 µg) > MSKE (5.0 ± 2.6 µg) > MG (8.6 ± 2.1 µg) > GA (14.8 ± 1.4 µg) for NK venom (Table 1). The inhibition of this enzymatic activity by MSKE may contribute to the prevention of extracellular matrix degradation, and thereby decrease the diffusion of toxins through tissues.

MSKE and its phenolic principles (GA, MG and PGG) inhibited the LAAO activity of both venoms in a dose-dependent manner (Figure 3). The order of potency as judged by the half-effective dose (ED₅₀) was PGG (0.9 ± 2.4 µg) > MSKE (1.5 ± 3.2 µg) > GA (1.7 ± 3.4 µg) > MG (2.3 ± 2.4 µg) for CR venom, and PGG (2.2 ± 2.0 µg) > MSKE (4.0 ± 1.7 µg) > MG (4.4 ± 2.0 µg) > GA (4.5 ± 2.8 µg) for NK venom (Table 1). In this assay, H₂O₂ generated by the enzyme of both venoms was used by horseradish peroxidase to oxidize o-dianisidine to the radical cation, which was monitored at 436 nm using a spectrophotometer. Therefore, the inhibition of LAAO activity of CR and NK venoms by the MSKE might presumably play a role in the prevention of highly localized concentrations of generated H₂O₂ in the victim’s tissues, since this snake venom enzyme can induce necrosis and apoptosis in mammalian endothelial cells and consequently cause hemorrhage, possibly through the production of highly localized concentrations of H₂O₂[10]. The inhibition of LAAO activity by the MSKE may contribute to the prevention of hemorrhage and necrosis.

It is seen from Figure 1, Figure 2 and Figure 3 and Table 1 that MSKE and its phenolic principles (PGG, MG and GA) exhibited dose-dependent inhibitory effects on PLA₂, hyaluronidase and LAAO activities of both venoms. These results indicated that the anti-enzymatic potency of MSKE may be attributed to its major principle (PGG) and other unidentified constituents, since PGG exerted its effect at the lowest ED₅₀ value and had the highest percentage content (61.28%) within the MSKE compared with GA (0.44%) and MG (0.68%). It could be noted that although GA and MG inhibited these enzymatic activities of both venoms, their percentage content within the MSKE were very low. Therefore, GA and MG may only possess a negligible effect compared with PGG. It has been reported that other plant extracts also contain compounds that capable of neutralizing PLA₂, hyaluronidase and LAAO activities of snake venoms [20,21,22]. The neutralization of these enzymes not only prevents the early effects of envenomation (local tissue damage, paralysis and hypotension), but also limits the systemic effects by reducing the distribution of toxin.

The minimum hemorrhagic dose (MHD) of CR venom was determined to be 20.4 µg/mouse, while the venom at three MHDs-induced hemorrhagic lesions with a diameter of 13.4 ± 0.4 mm. The MSKE significantly (P < 0.05) diminished the diameters of hemorrhagic lesions induced by three MHDs of CR venom in a dose-dependent manner, and the hemorrhagic activity could be completely inhibited at the extract amount of 67 mg/kg mouse (Figure 4A). The extract and solvent systems used did not induce hemorrhagic lesion in mice. The minimum necrotizing dose (MND) of NK venom was determined to be 63 µg/rat, which induced a necrotic lesion with a diameter of 4.7 ± 0.1 mm. Figure 4B shows the percentage inhibition of dermonecrotic activity of NK venom by MSKE which was dose-dependent. The extract with an amount of 0.60 mg/kg rat was effective in completely protecting rats from the dermonecrotic activity of one MND of NK venom (P < 0.05). Solvent and the extract alone did not induce necrotic lesion in rats.

The results of molecular modeling using a molecular docking method revealed the possible molecular orientation of MSKE constituents (GA, MG and PGG) in the PLA₂ binding pocket of CR and NK venoms (Figure 5 and and Table 2). GoldScores revealed that PGG was bound tighter than GA and MG in their binding pockets of CR and NK venoms. The binding orientations of GA, MG and PGG in the PLA₂-CR structure are shown in Figure 5A. Strong H-bonding interactions from GA, MG and PGG were found between a hydrogen atom of the hydroxyl group of ligands and an oxygen atom of the carboxylate group of Asp49. Moreover, common H-bonding interactions with Phe6, Gly30 and Lys61 were shown in the docked GA, MG and PGG. GA had a GoldScore of 37.55 and other H-bonding interactions were also formed with Ser2, Leu3 and Tyr52 ( Figure 5B). In Figure 5C, docked MG, with a GoldScore of 37.49, revealed other H-bonding interactions with Leu3, Glu7, Ile10, Tyr28, Cys29 and Trp31. Docked PGG had a GoldScore of 77.26 and revealed H-bonding interactions with Phe6, Glu7, Ile10, Gly18, Phe19, Tyr22, Ser23, Phe24, Tyr28, Gly30, Trp31, Gly32, Cys45, His48, Asp49, Lys61 and Phe97 (Figure 5D). The docked conformations of GA, MG and PGG in PLA₂-NK structure are shown in Figure 6. Their GoldScores were 35.52, 34.75 and 64.92, respectively. The docked conformation of GA, MG and PGG showed common H-bonding interactions with Tyr28, Gly30, Arg31, Gly32, Gly33, Cys45, His48, Asp49 and Tyr64. Additional H-bonding with Ser34, Glu56, Trp62, Pro63 and Phe65 were also formed with the docked conformation of PGG. It can be seen from Figure 5 and Figure 6 that the tannin principles of MSKE could form hydrogen bonds with active site residue His48 N^δ1 and Asp49 O^δ1 of the PLA₂ from CR and NK venoms. His48 and Asp49 amino acids are essential for enzymatic activity in the calcium-dependent PLA₂s; if the inhibitors are bonded to these amino acids the PLA₂ catalytic efficiency can be changed [7,23]. Furthermore, the tannins of MSKE also form important hydrogen bonds with the carbonyl oxygen atoms of the residues involved in calcium coordination (Tyr28, Gly30, Gly32 and Asp49) [23]. Therefore, the results of molecular docking indicated that the phenolic molecules of MSKE could selectively bind to the active sites or their proximity, or to modify conserved residues that are critical to the catalysis of PLA₂ enzymes. In addition, a previous study found that MSKE was a natural chelator [17] and it is well known that PLA₂s depend on Ca²⁺ to exert its action. This confirmed the presence of a selective mechanism for MSKE against PLA₂ activity in CR and NK venoms.

The possible molecular orientation of the MSKE constituents (GA, MG and PGG) in the LAAO binding pocket of CR venom is shown in Figure 7A. Their GoldScores (Table 2) were 38.95, 39.71 and 69.58, respectively. It was proposed that PGG was bound in the LAAO binding pocket of CR venom tighter than GA and MG. From their docked conformation in LAAO-CR structure, common H-bonding interaction of GA, MG and PGG were found with Arg90 and His463. GA also formed H-bonding interaction with Tyr372, Ile430, Gly464 and Trp465, as shown in Figure 7B. Moreover, a H-π interaction was found between GA and phenyl ring of Trp465 with a distance about 3 Å. In the case of docked MG, it revealed H-bonding interactions with Gly464, Trp465, and FAD and a π-π interaction with Trp465 (Figure 7C). The docked PGG showed H-bonding interactions with Leu207, Asn208, Glu209, Glu219, His223, Asp224, Phe227, Ala228, Arg322, Lys345, Phe354, Tyr356, Tyr372, Ile374, Thr431, Thr432 and FAD (Figure 7D). The interactions found between the amino acid residues of LAAO and tannins in this study agreed well with the crystal structures of LAAO complexed with citrate and with o-aminobenzoate [12]. Pawelek et al. [12] reported the amino acid residues of LAAO were involved in the binding of citrate and o-aminobenzoate; i.e. Leu51, Arg90, Asn208, His223, Asp224, Arg283, Gly313, Tyr372, Ile430 and Gly464. It can be seen from Figure 7 that the tannins of MSKE could form hydrogen bonds with amino acid residues in the substrate-binding domain (Arg90, His223, Asp224, Phe227, Ala228, Lys345, Phe354, Tyr356, Tyr372 and Ile374) of the LAAO from CR venom. These results indicated the presence of a selective mechanism of extract against LAAO activity in CR venom, since a protein database search revealed that the N-terminal sequence of LAAO from NK venom had the highest similarity score to LAAO from CR venom and had similarities to several other snake venom LAAOs [11]. Similar to CR venom, the phenolic compounds of the MSKE could therefore possibly bind to the LAAO binding pocket of NK venom and thereby inhibited the LAAO activity.

The in vivo inhibition of viper venom-induced hemorrhagic and cobra venom-induced dermonecrotic actions by MSKE is proposed. The inhibition was due firstly to a non-selective mechanism as a result of complexation between the venom proteins and tannins in the extract at the injection site [24] and to a selective mechanism where the active sites of the enzymes (PLA₂ and LAAO) of CR and NK venoms were blocked by the phenolic principles of MSKE as demonstrated in the molecular modeling study (Figure 5, Figure 6, and Figure 7). In addition, the MSKE has been reported to demonstrate chelating property with ferrous metal ions [17]; therefore, the authors will investigate further on the potential of MSKE in chelating the venom’s hemorrhagic Zn²⁺-metalloproteinases since these enzymes presumably act by destroying the collagenous basement membrane and other connective tissue collagen with consequent weakening of blood vessel stability causing the hemorrhagic effect [25]. The molecular docking method of the phenolic principles of the MSKE with the crystal structure of venom metalloproteinases will also be evaluated further.

PLA₂ activity in CR and NK venoms was measured using an indirect hemolytic assay on agarose-erythrocyte-egg yolk gel plates to define the minimum indirect hemolytic dose (MIHD) [26], the dose of venom that induced a hemolysis halo of 20 mm diameter after incubation for 20 h at 37 °C. The inhibition of PLA₂ activity by MSKE and its isolated phenolic principles was determined against one MIHD of CR and NK venoms. Test solutions and the venoms (0.05 mL each) were pre-incubated for 1 h at 37 °C. After centrifugation at 10,000 rpm for 10 min, the supernatant (20 µL) was tested for PLA₂ activity. The anti-PLA₂ potential of the extract and its phenolic principles was presented as percent inhibition of the enzymatic activity versus dose, in which 100% inhibition should produce no clear zone. ED₅₀, the effective dose that PLA₂ activity of the venoms was reduced by 50%, was calculated.

Hyaluronidase activity was determined according to the methods of Ferrante [27] and Yingprasertchai et al. [28] with slight modification. CR or NK venom (25-1,000 µg) in PBS (pH 7.4) was incubated with hyaluronic acid (50 μg) in 0.2 M sodium acetate buffer (1 mL, pH 5) that contained 0.15 M NaCl for 25 min at 37 °C. The reaction was stopped by adding 2.5% cetyltrimethylammonium bromide in 2% NaOH (2 mL). The optical density at 400 nm of each sample was read after 30 min with a Shimadzu UV-VIS spectrophotometer UV-160A against blank. Turbidity-reducing activity was expressed as the percentage of the remaining hyaluronic acid, taking the absorbance of a tube to which no enzyme was added as 100%. The venom concentration that produced a reduction in turbidity of ~50% was used in inhibition experiments. The inhibition of venom hyaluronidase activity by MSKE and its isolated phenolic principles was determined after pre-incubating an equal volume (100 µL) of test sample (with different amounts) in acetate buffer with CR or NK venom in PBS (2,500 µg/mL) for 30 min at 37 °C, after which hyaluronidase activity was measured. Blanks containing only the buffer and test sample were run in parallel. The anti-hyaluronidase potential of the extract and its phenolic principles was presented as percent inhibition of the enzymatic activity versus dose. ED_50, the effective dose that hyaluronidase activity of the venoms was reduced by 50%, was calculated.

LAAO activity of CR and NK venoms was determined by using an enzyme-coupled assay [29]. The assay mixture contained 0.025 mL 0.014% peroxidase and 0.2 M triethanolamine buffer (1 mL, pH 7.6) that contained 0.1% L-leucine and 0.0065% o-dianisidine. After incubation at 25 °C for 3 min, 0.1 mL CR or NK venom in PBS (1 mg/mL) was added, and an increase in absorbance at 436 nm (Shimadzu UV-VIS spectrophotometer UV-160A) was measured against blank. Enzymatic activity was calculated from the linear portion of the time–absorbance plot, where one unit of enzymatic activity was defined as the amount of enzyme that caused an increase of 0.001 absorbance units per minute. The percent inhibition of LAAO activity by MSKE and its isolated phenolic principles was evaluated by pre-incubating CR or NK venom (0.1 mL, 1 mg/mL) with test sample in sterile water for injection (SWI) (0.1 mL) for 1 h at 37 °C. After pre-incubation, the mixture was centrifuged at 10,000 rpm for 5 min, and supernatant (0.1 mL) was evaluated for LAAO activity. Venoms, test sample, and solvents alone were used as controls. The anti-LAAO potential of the extract and its phenolic principles was presented as percent inhibition of the enzymatic activity versus dose. ED_50, the effective dose that LAAO activity of the venoms was reduced by 50%, was calculated.

Hemorrhagic activity of CR venom was determined according to the method of Kondo et al. [30]. The minimum hemorrhagic dose (MHD) of CR venom was established by injecting 40 µL of different amounts of venom in sterile PBS (pH 7.2) intradermally (i.d.), into two separately marked positions on the shaved dorsal skin of unanesthesized mice (n = 8). After 90 minutes, mice were killed with an overdose of ether, the dorsal skin was removed and the diameter of the hemorrhagic lesion on the inner surface of the skin was measured using calipers and background illumination. The mean diameter of the hemorrhagic lesion was calculated for each venom dose. A dose-response curve for venom dose and hemorrhagic lesion diameter was plotted, and the MHD was recorded as the dose that caused a lesion diameter of 10 mm. The anti-hemorrhagic potential of MSKE was determined against three MHDs of CR venom in mice (n = 8) according to the modified methods of Yingprasertchai et al. [28] and Gutiérrez et al. [31]. Twenty microliters of venom in sterile PBS (3.06 μg/μL) was injected (i.d.) into the shaved dorsal skin of mice, immediately followed by test sample in SWI (20 µL) at the same injection site. After 90 minutes, mice were killed by ether overdose. Dorsal skin was removed and the diameter of the hemorrhagic lesion on the inner surface of the skin was measured. Sterile PBS, SWI, venom and MSKE alone were used as controls.

The methods of Kondo et al. [30] and Theakston and Reid [32] were used to determine the minimum necrotizing dose (MND) of NK venom in rats (n = 4) which is defined as the smallest amount of venom that could cause a necrotic lesion of 5 mm in diameter on the inner dorsal skin of the rat 72 h after intradermal injection. The anti-dermonecrotic potential of MSKE was determined against one MND of NK venom. The venom in sterile PBS (40 µL, 1.57 μg/μL) was injected i.d. into the shaved dorsal skin of rats (n = 4), immediately followed by test sample (40 µL) at the same injection site. Seventy-two hours after injection, rats were killed by ether overdose. The dorsal skin was removed, and the necrotic lesion on the inner surface of the skin was measured. Sterile PBS, venom, and MSKE alone were used as controls.