An Isoflavone from Dipteryx alata Vogel is Active against the in Vitro Neuromuscular Paralysis of Bothrops jararacussu Snake Venom and Bothropstoxin I, and Prevents Venom-Induced Myonecrosis

Snakebite is a neglected disease and serious health problem in Brazil, with most bites being caused by snakes of the genus Bothrops. Although serum therapy is the primary treatment for systemic envenomation, it is generally ineffective in neutralizing the local effects of these venoms. In this work, we examined the ability of 7,8,3'-trihydroxy-4'-methoxyisoflavone (TM), an isoflavone from Dipteryx alata, to neutralize the neurotoxicity (in mouse phrenic nerve-diaphragm preparations) and myotoxicity (assessed by light microscopy) of Bothrops jararacussu snake venom in vitro. The toxicity of TM was assessed using the Salmonella microsome assay (Ames test). Incubation with TM alone (200 μg/mL) did not alter the muscle twitch tension whereas incubation with venom (40 μg/mL) caused irreversible paralysis. Preincubation of TM (200 μg/mL) with venom attenuated the venom-induced neuromuscular blockade by 84% ± 5% (mean ± SEM; n = 4). The neuromuscular blockade caused by bothropstoxin-I (BthTX-I), the major myotoxic PLA2 of this venom, was also attenuated by TM. Histological analysis of diaphragm muscle incubated with TM showed that most fibers were preserved (only 9.2% ± 1.7% were damaged; n = 4) compared to venom alone (50.3% ± 5.4% of fibers damaged; n = 3), and preincubation of TM with venom significantly attenuated the venom-induced damage (only 17% ± 3.4% of fibers damaged; n = 3; p < 0.05 compared to venom alone). TM showed no mutagenicity in the Ames test using Salmonella strains TA98 and TA97a with (+S9) and without (−S9) metabolic activation. These findings indicate that TM is a potentially useful compound for antagonizing the neuromuscular effects (neurotoxicity and myotoxicity) of B. jararacussu venom.


Introduction
Envenomation by Bothrops snakes is characterized by local (pain, edema, inflammation, blistering, hemorrhage and necrosis) and systemic (coagulopathy, systemic hemorrhage, acute kidney injury and circulatory shock) manifestations [1]. Bothrops jararacussu is a large pit viper found in southeastern Brazil and northern Argentina [2]. Envenomation by this species shares many of the foregoing features with other Bothrops species [3], with most of the clinical manifestations of envenoming being mediated predominantly by snake venom metalloproteases (SVMPs), serine proteases, phospholipases (PLA 2 ) and C-type lectins. Transcriptomic analysis has confirmed that these are indeed the major protein classes in this venom, with PLA 2 being particularly abundant: Lys49-PLA 2 homologs accounted for 83.2% of PLA 2 transcripts, acidic Asp49-for 0.6% and basic Asp49-PLA 2 for 0.1% [4].
In addition to the features indicated above, clinical studies in the early 1900s suggested that bites by B. jararacussu also involved systemic manifestations reminiscent of envenoming by the South American rattlesnake, Crotalus durissus terrificus, viz., neurotoxicity, blindness, blurred vision, difficulty in swallowing and paralysis [5], in addition to other unspecified signs of neurotoxicity [6][7][8].
Serum therapy, the treatment of choice for snakebites, efficiently neutralizes the systemic manifestations but is generally ineffective against the local effects (edema, inflammation, hemorrhage and necrosis) of venoms [17]. Extensive local tissue damage leading to tissue loss after Bothrops bites can result in permanent disability and amputations [18,19]. Hence, there is a clinical and therapeutic impetus to develop alternatives for treating these local manifestations, with plant-derived bioactive products providing important candidate or lead molecules [20,21].
To increase our understanding of the components involved in the protective action of D. alata extracts, in this study we examined the ability of 7,8,3'-trihydroxy-4'-methoxyisoflavone (TM), an isoflavone from D. alata [32], to attenuate the neurotoxicity and myotoxicity of B. jararacussu venom and its main myotoxin, bothropstoxin-I (BthTX-I), in mouse phrenic nerve-diaphragm preparations in vitro. The mutagenicity (toxicity) of TM, as an indicator of it potential use as a clinical agent, was assessed in the Ames test using Salmonella strains TA 98 and TA 97a.

Results and Discussion
This work was based on three premises: (1) an understanding of the principal clinical manifestations of B. jararacussu envenomation [1]; (2) the low efficacy of antivenom against the local effects of Bothrops venoms [18,19]; and (3) the availability of a molecule, 7,8,3'-trihydroxy-4'-methoxyisoflavone (TM) from D. alata, with potentially interesting activity against B. jararacussu snake venom. Figure 1 shows the chemical structure of TM from D. alata [32], a compound also found in Xanthocercis zambesiaca (Baker) plant extract [34]. In preliminary screening for anti-snake venom activity, fraction F7 of a dichloromethane extract of D. alata bark [32] showed the best activity out of 11 fractions when tested against the neuromuscular activity of B. jararacussu venom [31]. Isoflavonoids are a large and very distinctive subclass of flavonoids, but only a few plants, including D. alata, have been reported to contain isoflavonoids [35]. Isoflavones are biologically active plant-food constituents that are common in the human diet and food industry, and are widely used as preservatives in pharmaceutical products [36]. This range of activities and uses suggested the possibility that isoflavones could be potentially useful as an adjuvant treatment for snakebites alongside standard serum therapy, particularly since antiserum shows little or no neutralizing activity towards the local effects of Bothrops venoms [18,19].

Pharmacological Assays
The ability of TM to inhibit the neuromuscular blockade caused by B. jararacussu venom was investigated in mouse phrenic nerve-diaphragm (PND) preparations. Figure 2 shows the characteristic irreversible neuromuscular blockade induced by B. jararacussu venom in vitro (Bjssu, 40 µg/mL, n = 6, * p < 0.05 compared to control), in agreement with that originally reported by Rodrigues-Simioni et al. [9]. Incubation with TM alone (200 µg/mL, n = 4) did not change the muscle twitch-tension responses, which were similar to those of preparations incubated with Tyrode solution (negative control; n = 6). Preincubation of TM (200 µg/mL) with venom (40 µg/mL) for 30 min prior to testing totally abolished the venom-induced neuromuscular blockade (n = 4, * p < 0.05 compared to venom alone).
Bothrops jararacussu venom is rich in PLA 2 , most of which are Lys49-PLA 2 , with considerably fewer basic and acidic Asp49-PLA 2 [4]. Bothropstoxin-I (BthTX-I) is the major Lys49-PLA 2 myotoxin and totally reproduces the in vitro neurotoxicity and myotoxicity of B. jararacussu venom [11]. This toxin acts presynaptically before causing membrane depolarization [15,37]. As shown in Figure 3A, incubation of PND preparations with BthTX-I (20 µg/mL) mimicked the characteristic progressive, irreversible decrease in muscle twitch-tension caused by the venom. As with the venom, preincubation of BthTX-I (20 µg/mL) with TM (200 µg/mL) fully protected the PND against toxin-induced neuromuscular blockade ( Figure 3B). When TM was added 10 min after BthTX-I there was still attenuation of the neuromuscular blockade, although this was much less marked than with the preincubation protocol ( Figure 3C). These findings suggest that the protective action of TM against venom-induced neuromuscular blockade most likely resulted from the inhibition of BthTX-I, presumably by attenuating the presynaptic activity of the toxin [15,37].

Quantitative Histological Analysis
To assess the anti-myotoxic activity of TM, diaphragm muscles exposed to different treatments (TM, venom or TM + venom) were analyzed by light microscopy, as described by Queiroz et al. [38] for mouse tibialis anterior muscle. Figure 4A shows a characteristic myographic recording of PND contractile activity during incubation with TM (200 µg/mL) for 120 min; similar responses were observed in preparations incubated with Tyrode solution alone (results not shown). The percentage of damage cells in these two protocols was <10% [9.7% ± 1.6% (n = 6) for Tyrode solution and 9.2% ± 1.7% (n = 4) for TM-treated muscles]. In indirectly stimulated PND, B. jararacussu venom (40 μg/mL, n = 6) induced a transient contracture that was followed by progressive neuromuscular blockade of muscle twitches during the following 120 min ( Figure 5A). In contrast to the foregoing controls, there was a marked increase in the percentage of damaged fibers (to 50.3% ± 5.4%; p < 0.05 compared to the Tyrode control) in diaphragms incubated with venom alone (40 μg/mL, n = 3) for 120 min ( Figure 5B). In diaphragm muscle incubated with a mixture of TM (200 μg/mL) + venom (40 μg/mL) ( Figure 5C, n = 3) there was a significant decrease in the percentage of damaged fibers (17% ± 3.4%; p < 0.05 compared to venom alone). Although the venom produced complete neuromuscular blockade, quantitative analysis of the fibers showed that only 50% were damaged. This discrepancy may reflect the fact that muscle cross-sections do not allow observation of the entire fiber length. Although diaphragm muscle has a high resistance to fatigue and a safety margin of 10% [39], the venom nevertheless had a striking effect on the physiological excitation-contraction coupling mechanism. As shown here, the preincubation of venom with TM was effective in protecting the muscle against venom-induced myotoxicity. In this work, we focused on the ability of TM to antagonize the neuromuscular action (neurotoxicity and myotoxicity) of B. jararacussu venom and its main PLA 2 myotoxin, BthTX-I. However, B. jararacussu venom has a variety of other biological activities (edema-formation, hemorrhage and lethality) against which it would be interesting to assess the neutralizing capacity of TM. Nevertheless, with the isolation procedure used in this study, the amount of isoflavone obtained was low, which severely limited the number of protocols and activities that could be tested. This limitation (low yield/availability of purified TM) is common to other phytochemical compounds with biological activity [40]. Nonetheless, as shown here, TM faithfully mimicked the protection against the venom-induced neuromuscular blockade seen with a crude extract of D. alata [31].
Several mechanisms have been proposed to explain the anti-snake venom activity of phytochemical compounds [21], with the specific interactions depending on the plant compound (tannins, coumarins, flavonoids, etc.) and the snake genus involved. In the case of TM, we suggest that its interaction with B. jararacussu venom and BthTX-I involves the hydroxyl groups at positions C-8 and C-7 of the isoflavone (see Figure 1). Since the neighboring hydroxyls are susceptible to radical formation [35], this chemical interaction would ultimately lead to a very reactive 7,8-quinone moiety that could modify specific sites in enzymes responsible for the local and systemic effects of the venom.

Salmonella Mutagenicity Assay
The toxicology of TM was assessed in the Salmonella mutagenicity assay, using test strains TA97a and TA98 that can be reverted by frameshift mutagens [41,42]. Although the Organisation for Economic Cooperation and Development (OECD) [41] recommends the use of at least five Salmonella strains to meet regulatory demands, the low yield of isoflavonoid limited the use of five strains in this work. On the other hand, satisfactory toxicological results have been obtained using only TA97a and TA98 [43] or TA97a and TA100 [44], and were considered as valid mutagenic screening. In our experimental conditions a GC frameshift was evaluated. The mutagenicity assays showed that TM did not increase the number of revertant colonies relative to the negative control, indicating no mutagenicity for this compound (Table 1). The genes affected in Salmonella strains TA98 and TA97a were hisD6610 and hisD3052, respectively, although these strains also contained the mutations ∆urvB, rfa and pKM101. The former was designed to enhance the mutagenicity of compounds, presumably through the nucleotide excision repair system [42] that extended into a gene for biotin synthesis and required the addition of biotin to the culture medium because of the deletion in this region. The rfa mutation changes the properties of the bacterial cell wall and results in partial loss of the lipopolysaccharide (LPS) barrier, thereby increasing the permeability of cells to certain chemicals; this mutation is generally detected based on the altered sensitivity to crystal violet. The R factor plasmid (pKM101) makes the strains more responsive to a variety of mutagens [45].

Plant Material and Extraction
Bark samples were collected from an adult D. alata Vogel tree in Pedro Afonso (Tocantins, Brazil) in December 2007 and identified by Dr. Roseli B. Torres (Institute of Agronomy of Campinas-IAC). A voucher specimen was deposited in the herbarium of the IAC (IAC 50629). The bark (1.269 kg) was dried at 37 C for 48 h and then powdered, ground in a mill, macerated for 5 days (200 g in 2 L of 70% ethanol) and the suspension then percolated (protected from light) at 20 drops/min, resulting in a 20% (m/v) hydroalcoholic extract. The extract was concentrated under reduced pressure and lyophilized to yield a final residue of 170 g that corresponded to an extraction efficiency of 85% [46].

Isoflavone Solubilization
For the pharmacological assays, fresh solutions of TM were prepared daily by dissolving in 30 µL of DMSO (CAQ -Casa da Química Ind. E Com. Ltd, Diadema, SP, Brazil); this volume of solvent did not alter the basal response of the neuromuscular preparation [47,48].

Venom and Purification of BthTX-I
Bothrops jararacussu venom was collected from two adult specimens kept in the Serpentário do Centro de Estudos da Natureza (CEN). The venom was lyophilized and certified by Dr. José Carlos Cogo (Vale do Paraiba University-UNIVAP, São José dos Campos, SP, Brazil). BthTX-I was purified and its identity confirmed as described by Homsi-Brandeburgo et al. [11].

Animals
Male Swiss white mice (26-32 g) were supplied by Anilab (Animais de Laboratório, Paulínia, SP, Brazil). The animals were housed at 25 ± 3 C on a 12 h light/dark cycle and had access to food and water ad libitum. This project was approved by the institutional Committee for Ethics in Animal Research of Vale do Paraiba University (protocol no. A013/CEUA/2011) and the experiments were done in accordance with the general guidelines of the Brazilian Society for Laboratory Animal Science (SBCAL).
Control preparations were incubated with Tyrode solution alone (n = 6), whereas other preparations were incubated with TM (200 µg/mL, n = 4), venom (40 µg/mL, n = 6) or a mixture of TM + venom preincubated for 30 min at 37 °C prior addition to the organ bath (n = 4) [33]. BthTX-I (20 µg/mL) was assayed in two protocols: in one, PND preparations were incubated with toxin followed by the addition of TM (200 µg/mL) while in the other the toxin was preincubated with TM (200 µg/mL) prior to testing in PND.

Quantitative Histological Analysis
At the end of the incubations (120 min), the preparations from the various protocols (control, TM, venom and venom + TM) were processed for histological analysis. The preparations (n = 3 for each treatment) were fixed in Bouin solution and processed by routine morphological techniques. Cross-sections (5 µm thick) of diaphragm muscle were stained with 0.5% (w/v) hematoxylin-eosin for examination by light microscopy. Tissue damage (edema, intense myonecrosis characterized by muscle fiber atrophy, a hyaline aspect, sarcolemmal disruption and myofibril lysis) was expressed as a myotoxicity index (MI), defined as (the number of damaged muscle cells/the total number of cells) × 100 in three non-overlapping, non-adjacent areas of each preparation [28].

In vitro Mutagenicity Assay
Mutagenic activity was evaluated by the Salmonella microsome assay, using the Salmonella typhimurium test strains TA98 and TA97a, kindly provided by Dr. B.N. Ames (Berkeley, CA, USA), with (+S9) and without (−S9) metabolization, by the preincubation method [50]. As discussed in the Results section, the OECD [41] recommends the use of five strains (S. typhimurium 1535, TA97a, TA98, TA100, TA102), but this was not feasible in this study because of the low yield of isoflavonoid during extraction. The strains were grown from frozen cultures overnight for 12-14 h in Oxoid Nutrient Broth No. 2. The metabolic activation mixture (S9 fraction), prepared from livers of Sprague-Dawley rats treated with the polychlorinated biphenyl mixture Aroclor 1254 (500 mg/kg), was purchased from Molecular Toxicology Inc. (Boone, NC, USA) and freshly prepared before each test. The metabolic activation system consisted of 4% S9 fraction, 1% 0.4 M MgCl 2 , 1% 1.65 M KCl, 0.5% 1 M D-glucose-6-phosphate disodium, 4% 0.1 M NADP (nicotinamide adenine dinucleotide phosphate), 50% 0.2 M phosphate buffer and 39.5% sterile distilled water. The isoflavone was dissolved in DMSO (Sigma Chemical Co., St. Louis, MO, USA) to provide a non-toxic concentration of 0.19 mg/plate. This concentration was added to 0.5 mL of 0.2 M phosphate buffer, or to 0.5 mL of 4% S9 mixture, with 0.1 mL of bacterial culture and then incubated at 37 °C for 20-30 min. Subsequently, 2 mL of top agar [0.6% agar, histidine and biotin (0.5 mM each) and 0.5% NaCl] was added and the mixture poured onto a plate containing minimal agar. The plates were incubated at 37 °C for 48 h and the His+ revertant colonies were counted manually. All experiments were done only in duplicate because of the low amount of isoflavone available. The use of duplicate plating is acceptable when scientifically justified [41]. The results were analyzed with the statistical software package Salanal 1.0 (U.S. Environmental Protection Agency, Monitoring Systems Laboratory, Las Vegas, NV, from Research Triangle Institute, Research Triangle Park, NC, USA), using the model of Bernstein et al. [51]. The mutagenic index (MI), defined as the average number of revertants per plate with the test compound divided by the average number of revertants per plate with the negative (solvent) control, was also calculated. A sample was considered positive when the MI was equal to or greater than two for at least one of the concentrations, and if it had a reproducible dose-response curve [52]. The standard mutagen used as a positive control in experiments without the S9 mix was 4-nitro-o-phenylenediamine (NOPD, 10 μg/plate). In experiments with S9 activation, 2-anthramine (2-AA, 1.25 μg/plate) was used. DMSO (50 μL/plate) served as the negative (solvent) control.

Statistical Analysis
Each pharmacological protocol was repeated at least four times and the results are shown as the mean ± SEM. The number of experiments (n) is indicated in the corresponding figure legend. Student's t-test was used for statistical comparison of the data and the confidence level was set as 5% (α = 0.05).

Conclusions
In conclusion, the bioactive isoflavone, 7,8,3'-trihydroxy-4'-methoxyisoflavone (TM) from D. alata Vogel efficiently counteracted the myotoxicity and neuromuscular activity of B. jararacussu venom and its major myotoxin, BthTX-I, in vitro. TM was not mutagenic in S. typhimurium strains TA 97a and TA 98, indicating its safety. The results of this study reinforce the potential of TM for therapeutic use in the treatment of venomous snakebites.