Proteins from Mucuna pruriens and Enzymes fromEchis carinatus Venom

Mucuna pruriens seeds have been widely used against snakebite in traditional medicine. The antivenin property of a water extract of seeds was assessed in vivoin mice. The serum of mice treated with extract was tested for its immunological properties. Two proteins of Echis carinatusvenom with apparent molecular masses of 25 and 16 kDa were detected by Western blot analysis carried out using IgG of mice immunized with extract or its partially purified protein fractions. By enzymatic in-gel digestion and electrospray ionization-mass spectrometry/mass spectrometry analysis of immunoreactive venom proteins, phospholipase A2, the most toxic enzyme of snake venom, was identified. These results demonstrate that the observed antivenin activity has an immune mechanism. Antibodies of mice treated with non-lethal doses of venom reacted against some proteins ofM. pruriens extract. Proteins of E. carinatusvenom and M. pruriens extract have at least one epitope in common as confirmed by immunodiffusion assay.

Snakebite is a considerable problem in certain tropical and subtropical countries. According to World Health Organization estimates, 40,000 of 5 million cases of snakebite are fatal. Antivenins obtained from horses treated with snake venom are one of the principal remedies against snakebite. This therapy has the disadvantage that antivenins must be given immediately, and snakebite victims may develop an adverse reaction including anaphylactic shock (1). The use of endogenous plants with a reputation against snakebite is therefore worth considering (2).
In preliminary experiments (3,4) we demonstrated that extract of M. pruriens (MPE), 1 a medicinal plant widely used in Nigeria for its chemical and pharmacological properties, protects mice against the lethal effect of Echis carinatus venom (EV). Both MPE and EV are heterogeneous mixtures, their interaction represents a complex phenomenon, and there is no information about its biochemical mechanism. EV contains proteins with different toxic properties including opposite effects on blood clotting. Well known proteins are: disintegrins EC3 (5), EC6 (6), and echistatin (7), which inhibit the interaction of fibrinogen with the glycoprotein IIb-IIIa receptor on the platelet surface; echicetin (8) and ECLVIX/Xbp (9) with an opposite effect on platelet aggregation; two metalloproteases, ecarin (10,11) and carinactivase (12), which are prothrombin activators and act as procoagulant enzymes; and phospholipase A 2 (PLA 2 ) (13)(14)(15), the most abundant enzyme, which has many effects including inhibition of prothrombin activation by ecarin and carinactivase (16). When injected into mice, this complex mixture of proteins induces disseminated intravascular coagulation leading to death in less than a day. The composition of the M. pruriens seed is also complex and variable, with 20 -30% protein (lectins, globulins, protease inhibitors), 1-10% fat, 4 -5% ash, 4 -9% water, 4 -7% fiber (17)(18)(19)(20), and L-DOPA (21), an interesting non-protein component. The aim of the present study was to study the mechanism, the factors of MPE, and the proteins of EV involved in the observed phenomenon.
Plant Material and Animals-M. pruriens (family: Fabaceae; subfamily: Papilionoideae; genus: Mucuna; species: pruriens) seeds were collected in the Rukuba area in Jos, Nigeria with the aid of a traditional healer. They were authenticated by Prof. S. W. H. Hussini of the Department of Botany, University of Jos. Voucher specimen number A102 is deposited in the Pharmacy Herbarium of the University of Jos. CDI-ICR mice (30 g) from Nossan were kept at a temperature of 22 Ϯ 1°C with a relative humidity of 60 Ϯ 5% in a 12-h light/dark cycle with a standard diet and water ad libitum.
Preparation and Partial Purification of M. pruriens Seed Extract-Sundried seeds of M. pruriens were ground to a paste of uniform consistency, 50 g of which was soaked in 100 ml of H 2 O, extracted for 24 h at 4°C, and centrifuged at 10,000 ϫ g for 20 min. The supernatant lyophilized to a powder (24% protein), which was stored at Ϫ4°C. Separation of the protein (P) and non-protein (NP) fractions was * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ To whom correspondence should be addressed. achieved by gel filtration on a HiPrep 26/10 column: 5 ml of MPE solution (5.4 mg/ml) was loaded and eluted with 50 mM Tris buffer, pH 7 (flow rate of 5 ml/min). E 280 nm , E 254 nm , and conductivity were monitored. For partial purification of MPE proteins, 5 ml of MPE solution (5.4 mg/ml) was applied to a Sephacryl S-200 HR column (2.6 ϫ 80 cm) eluted with 50 mM Na 2 HPO 4 , pH 7.5, at a flow rate of 1.5 ml/min and read at 280 nm. P1, P2, and P3 were obtained.
In Vivo Protective Effect of MPE, P, and NP against EV-Groups of eight mice were injected with MPE, P, or NP fractions or saline (control). At different times afterward, the mice were injected with a minimum lethal dose of EV (minimal lethal dose, 2 mg/kg). The percentage of survivors was assessed 24 h later. The control group was injected with saline before EV. All fractions were injected intraperitoneally at doses proportional to body weight, calculating dilution after separation.
Preparation of Antisera and Purification of IgG-Six groups of eight mice were treated once a week for 3 weeks with MPE, P1, P2, P3, EV (non-lethal dose), and saline. After 28 days they were sacrificed, blood was withdrawn, and anti-MPE, anti-P1, anti-P2, anti-P3, anti-EV, and preimmune sera were obtained. All antisera were purified by affinity chromatography with an AKTA liquid chromatography system. 10 ml of each anti-serum diluted in binding buffer (20 mM Na 2 HPO 4 , pH 7) and filtered on a 0.22-m membrane was adsorbed on a 5-ml protein G column (1.6 ϫ 2.5 cm) equilibrated in binding buffer until all unbound material was washed out. IgG fractions were then eluted at a flow rate of 2.5 ml/min with 0.1 M glycine-HCl, pH 2.7 (elution buffer). The fractions were neutralized with 1 M Tris, pH 9, and concentrated with Centriplus membrane (final concentration of 1.5 mg/ml).

SDS-PAGE and Western Blot
Analysis-Proteins in all fresh samples were determined by a Bio-Rad assay (22) and separated by 12% SDS-PAGE according to Laemmli (23). They were transferred to a nitrocellulose membrane (0.45 m) at 100 V for 1 h at 4°C and stained with Ponceau S. The membrane was blocked with 0.3% non-fat powdered milk in 1ϫ PBS containing 0.1% Tween 20 and incubated overnight at 37°C with treated mouse IgG. IgG of untreated mice was used as the negative control. The membrane was washed in 1ϫ PBS containing 0.1% (v/v) Tween 20, incubated with peroxidase-conjugated goat antimouse IgG (1:2000), and developed by ECL. Detection of total proteins after 12% SDS-PAGE was achieved by silver staining using PlusOne silver staining kit protein.
Neutralization of Lethal Potency of EV-The minimal lethal dose of EV was preincubated with 100 l each of anti-MPE, anti-EV, anti-P1, anti-P2, anti-P3, and preimmune IgG fractions at 37°C for 1 h. The preincubated mixtures were then injected into six groups of eight mice. Control groups were injected with EV mixed with saline or purified IgG from serum of preimmune mice. The number of deaths in the subsequent 24 h was recorded.
Immunodiffusion Assay-Antigenic relationships between the various antigens were studied by double diffusion test, according to Ouchterlony and Nilsson (24). Holes 5 mm in diameter were punched in Enzymatic In-gel Digestion-Coomassie Blue-stained EV and MPE protein bands separated by SDS-PAGE were excised from the gel and digested with trypsin according to known procedures (25) with slight modifications. Briefly, gel slices were washed for at least 1 h in 100 mM NH 4 HCO 3 , pH 8.0, and then for 1 h with 50% acetonitrile,100 mM NH 4 HCO 3 , pH 8.0, under shaking. Acetonitrile was added to shrink the gel pieces, and after 10 -15 min of incubation, the solvent was removed and the samples were dried in a Speed Vac. Gel slices were reswollen with 25 mM NH 4 HCO 3 , pH 8.0, containing modified trypsin (Promega) and incubated for 4 h at 37°C. The supernatant was acidified with trifluoroacetic acid to a final concentration of 1%. Peptides were extracted from the gel slices twice with 60% acetonitrile and 0.1% trifluoroacetic acid for 20 min. All supernatants were combined, and after evaporation to near dryness, peptide fragments were reconstituted in 20 l of 0.1% trifluoroacetic acid.
Liquid Chromatography/Electrospray Ionization-Tandem Mass Spectrometry (LC/ESI-MS/MS) Analyses-The mixture of peptides was separated using a Nucleosil C 8 column (4.6 ϫ 250 mm, 5-m particle size, 300 A) and analyzed with a Finnigan LCQ ion trap mass spectrometer (San Jose, CA) with an ESI source. A detailed scheme of the experimental setup for this type of analyses is described elsewhere (26). Briefly, a positive voltage of 3 kV was applied to the electrospray needle, and a N 2 sheath flow was applied to stabilize the ESI signal. The LC/MS analysis was conducted using a PerkinElmer high pressure liquid chromatography system coupled to the LCQ. The mobile phase from the column (flow rate of 0.5 ml/min) was split before the mass injector by a Tee-connector. The enzymatically digested peptides were eluted from the column using 0.5% formic acid in water (mobile phase A) and 0.5% formic acid in acetonitrile (mobile phase B) with a threestep linear gradient of 5-10% B in the first 10 min, 10 -35% B in the next 40 min, and 35-40% B in the last 5 min. The LC/ESI-MS/MS analysis was accomplished using an automated data acquisition procedure in which a cyclic series of three different scan modes was performed. Data acquisition was conducted using the full scan mode (m/z 300 -2000) to obtain the most intense peak (signal Ͼ 1.5 ϫ 10 5 counts) as the precursor ion, followed by a high resolution zoom scan mode to determine the charge state of the precursor ion and MS/MS scan mode to determine the structural fragment ions of the precursor ion. The resulting MS/MS spectra were then matched against a protein data base (Owl) by Sequest software to confirm the sequence of tryptic peptides.

RESULTS
Antivenin Activity of MPE-We first found that the protective effect of MPE against the lethal effect of EV was exerted at a dose of 21 g/g and was evident 24 h and 1-4 weeks after administration. To understand the chemical nature of substances responsible for the protection, an in vivo test was set up with two fractions obtained by HiPrep separation of MPE, one containing P and the other one NP compounds from MPE (Fig.   1). NP fractions contain small molecules like L-DOPA responsible for E 280 nm absorbance, free amino acids, ions, and fatty acids as already reported (18,19).
The in vivo test showed that P and NP fractions exerted protection in different ways. As shown in Table I, the NP fraction conferred short term protection (1 day), whereas the P fraction conferred long term protection (1, 2, and 3 weeks after administration). When we comparatively injected mice with MPE, P, and NP fractions once a week for 3 weeks, the protective effect of the P fraction specifically increased. Some compounds in the NP fraction may be adjuvants in the long term protective effect because the P fraction was less active when used alone. Only with a booster dose of P fraction was the total effect restored.
Three well resolved protein peaks, P1, P2, P3, and NP fraction were obtained by further purification of MPE by gel filtration on Sephacryl S-200 HR. They are shown in Fig. 2.
Antibodies against EV Induced in Mice by Extracts of M. pruriens-The protective activities of MPE, P1, P2, and P3 were tested for their capacities to raise antibodies in mice against EV proteins. IgG were purified from anti-MPE, anti-P1, anti-P2, anti-P3, and preimmune mice serum with protein G affinity separation and used in Western blot experiments.
When anti-MPE IgG were tested against EV proteins, two protein bands with apparent molecular masses of about 25 kDa (EV25) and 16 kDa (EV16) were detected. The signal was only visible under reducing conditions implying that the epitope on the native EV protein was in a cryptic state. When MPE proteins were incubated with the specific anti-MPE IgG, as positive control, a pattern similar to that obtained with silver staining was achieved, indicating that almost all MPE proteins were highly antigenic (Fig. 3A).
When the protein fractions obtained by partial purification of MPE on Sephacryl S-200 were injected into mice, we also obtained antibodies against EV. Antibodies raised in mice by injecting P2 fraction gave similar results to those obtained with anti-MPE IgG, whereas when anti-P3 IgG was used, only EV16 was detected. No signal against any EV proteins was obtained using IgG of mice treated with P1 fraction (anti-P1 IgG). Positive controls of anti-P1, -P2, and -P3 IgG were the corresponding P1, P2, and P3 fractions (Fig. 3B).
Antibodies against Proteins of M. pruriens Induced in Mice by Administration of EV-Under reducing conditions, some anti-EV IgG reacted with at least three bands of MPE proteins with molecular masses in the range of 22-28 kDa (MPE 22-28) as shown in Fig. 3C. These proteins mainly belong to the P2 fraction. EV proteins reacted strongly against the specific anti-EV IgG under non-reducing conditions, and this sample was used as positive control.
Neutralization of Lethal Potency of EV-The capacity of anti-MPE, anti-P1, anti-P2, and anti-P3 IgG to neutralize the toxicity in vivo of EV was tested after incubation of the fractions with venom, and results are reported in Table II. No neutralization was observed in control groups 1 and 2 (negative controls); lethal potency of EV was neutralized by anti-EV IgG obtained from mice treated with a non-lethal dose of EV (group 3, positive control). Anti-MPE (group 4) and anti-P2 (group 6) showed neutralizing effects similar to that of group 3; less neutralization was obtained with anti-P3 (group 7), and no neutralizing effect was obtained with anti-P1 (group 5). A 50% survival percentage was considered a satisfactory neutralizing effect.
Immunodiffusion Test-To confirm the results of Western blot experiments and to ascertain the presence of one or more common antigenic epitopes in MPE and EV proteins, the double diffusion test was used. When anti-MPE IgG was tested FIG. 5. Identification of EV16 band by mass spectrometry. The most abundant peaks of the tryptic digested mixture of EV16 band were selected, tandem mass spectrum was performed, and the sequence stretch, together with its starting mass, its end mass, and the molecular weight of the peptide were entered in the data base search program (Sequest) where they were converted to a peptide sequence tag. Two peptide were partially sequenced. A, NLFQFAEMIVK corresponds to the fragmentation of ion with m/z 670.5. B, DNLNTYDKK corresponds to the fragmentation of a ion with m/z 982.2.
against MPE and EV proteins, a pattern shown in Fig. 4a was obtained indicating coalescence of antigens. When anti-EV IgG was tested against MPE, P2, P3, and EV one precipitin line was formed in all cases (Fig. 4b).
Protein Identification by ESI-MS/MS-For the identification of proteins involved in this phenomenon, enzymatic in-gel digestion and ESI-MS/MS analysis of MPE and EV immunoreactive proteins were performed. The most intense peaks in the mass spectrum were automatically selected using Sequest software for the data base search. The proteins were identified by correlation of the experimentally obtained MS/MS spectra to the theoretically predicted peptide MS/MS spectra of proteins present in the data base. Two peptides of the tryptic digested mixture of EV16 protein were found with sequences NLFQ-FAEMIVK (Fig. 5A) and DNLNTYDKK (Fig. 5B) matching that of Russell's viper phospholipase A 2 .

DISCUSSION
The present results demonstrate that extracts of M. pruriens seeds protect mice against the toxic effects of EV. It does so by an immunological mechanism based on a series of specific epitopes common to some vegetal and venom proteins.
The in vivo test results indicate that administration of MPE proteins raised antibodies responsible for the protection observed. The long term protection was in fact more complete when the P fraction was administered according to an immunization protocol (once a week for 3 weeks). Several proteins in the P2 fraction are involved in raising antibodies, while the most purified P3 fraction is less active. We demonstrated by Western blot analysis that certain antibodies induced in mice by injection of vegetal extracts reacted directly with certain EV proteins, and immunodiffusion experiments pointed out the cross-reaction between MPE and EV proteins.
Western blot analysis carried out using anti-MPE and anti-P2 IgG showed that only EV proteins with molecular masses around 25 and 16 kDa were targets of the antibodies raised in mice after injection of MPE or P2; the toxic effect in vivo of EV was neutralized when the venom was preincubated with these IgG before injection. EV25 and EV16 may be the most toxic components of EV acting in concert, and their neutralization seems to impair the toxic mixture making it nonlethal. Anti-P3 IgG did not totally neutralize the venom, presumably because it only reacted with a single protein of EV (EV16).
Electrospray mass spectrometry combined with a peptide sequence tag search lead to the identification of the proteins with an estimated molecular mass of 16 kDa in the EV16 band as PLA 2 . The sequences of two peptides obtained by trypsin digestion matched the sequence of Russell's viper PLA 2 (Daboia russellii), while the venom we used was certified to be from E. carinatus sochureki. This may be explained by the fact that only one sequence of E. carinatus PLA 2 is deposited in the data base, whereas various isoenzymes are known in literature (13)(14)(15)27), and EV varies with country of origin and other factors (28) as shown in our experiments and declared by our supplier. Since the genera Echis and Daboia are very similar and both belong to the subfamily Viperinae, the E. carinatus PLA 2 that we analyzed may be a genetic variant, the sequence of which is closest to that of Russell's viper. In our recent paper (29) we demonstrated the in vitro effects of anti-MPE on EV enzymes catalyzing prothrombin transformation. PLA 2 is known to be implicated in that reaction, and our present results confirmed that PLA 2 , the enzyme largely responsible for the lethal effects of snake venom, is the putative target of the antibodies induced by MPE. Analysis of in-gel trypsin digestion of the EV25 band gave us no significant conclusions because no sequences of EV proteins with an estimated molecular mass of 25 kDa has been yet reported in the data bases.
We also tried to identify the specific MPE protein(s) responsible for the observed phenomenon. Our results suggest the presence of similar epitopes in EV and MPE proteins. This enabled us to use anti-EV IgG to detect specific proteins of MPE, limiting the numbers of animals necessary for in vivo testing during protein purification. We succeeded in identifying a group of MPE immunoreactive proteins (at least three bands) very likely due to isoforms. When tryptic digested mixtures of these bands were analyzed by ESI-MS/MS no significant matchings were obtained considering that the genome and proteome of M. pruriens is still unexplored and that no protein sequences are reported in the data bases. A molecular biology approach is required to identify these proteins, and it will be a topic for future researches.
It seems surprising that proteins of EV and MPE could have common epitopes; however, other examples of sequences shared by plant and snake venom proteins have been reported. The lectin domain, contained in all plant chitin-binding proteins, shows a sequence similarity to disintegrins of crotalid and viperid snake venoms (30). Another example is the similar overall folding pattern of the three-dimensional structures of snake venom postsynaptic neurotoxins and the domains of wheat germ agglutinin (31). All venom proteins have different functional activities but they show structural and evolutionary  in vivo effect of MPE, P, and NP fractions against EV is shown. Groups of eight mice were injected with the indicated fractions and treated with a minimum lethal dose of EV (2 mg/Kg) 24 h, 1 week, and 3 weeks later. The control group was injected with saline and then EV. Survivors were counted 24 h after EV injection. All fractions were injected intraperitoneally with doses proportional to body weight (g/g). All groups received one injection except group A in the last column, which was immunized with one injection a week for 3 weeks. relationships; they are derived from a common precursor and share conserved domains (32)(33)(34), some of which are common to plant proteins. The C-type lectin domain (CTL or carbohydraterecognition domain, CRD) has a highly conserved structure (35,36), is responsible for carbohydrate binding, and has been found in all calcium-dependent type lectin-related proteins, echicetin, ECLVIX/XBp, and carinactivase. Like many other Leguminosae plants, M. pruriens seeds contain legume lectins (37) that belong to the same family of C-type animal lectins and may contain the same CTL domain of calcium-dependent type lectin-related proteins of EV. We can conclude that when MPE proteins are injected into mice in such a way as to induce abundant antibody production, a polyclonal serum against epitopes present on one or more EV proteins is obtained. If MPE extract and some of its proteins protect mice against EV PLA 2 or other snake venom proteins that show procoagulant and anticoagulant activities, very likely they could interfere in the coagulation process.
The present findings open new perspectives in the field of vaccine by natural products and may be useful in the therapy of snakebite and other coagulation disorders.