Antibacterial activity of snake, scorpion and bee venoms: a comparison with purified venom phospholipase A 2 enzymes

Transcription

1 Journal of Applied Microbiology ISSN ORIGINAL ARTICLE Antibacterial activity of snake, scorpion and bee venoms: a comparison with purified venom phospholipase A enzymes R. Perumal Samy, P. Gopalakrishnakone, M.M. Thwin, T.K.V. Chow, H. Bow, E.H. Yap and T.W.J. Thong Venom and Toxin Research Programme, Department of Anatomy, Yong Loo Lin School of Medicine, National University of Singapore, Singapore Department of Microbiology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore Keywords L-amino acid oxidase, antibacterial activity, phospholipase A, toxins, venoms. Correspondence P. Gopalakrishnakone, Venom and Toxin Research Programme, Department of Anatomy, National University of Singapore, Singapore antgopal@nus.edu.sg 5/748: received 3 June 5, revised May 6 and accepted 3 July 6 doi:./j x Abstract Aims: Venoms of snakes, scorpions, bees and purified venom phospholipase A (PLA ) enzymes were examined to evaluate the antibacterial activity of purified venom enzymes as compared with that of the crude venoms. Methods and Results: Thirty-four crude venoms, nine purified PLA s and two l-amino acid oxidases (LAAO) were studied for antibacterial activity by disc-diffusion assay ( lg ml ) ). Several snake venoms (Daboia russelli russelli, Crotalus adamanteus, Naja sumatrana, Pseudechis guttata, Agkistrodon halys, Acanthophis praelongus and Daboia russelli siamensis) showed activity against two to four different pathogenic bacteria. Daboia russelli russelli and Pseudechis australis venoms exhibited the most potent activity against Staphylococcus aureus, while the rest showed only a moderate activity against one or more bacteria. The order of susceptibility of the bacteria against viperidae venoms was S. aureus > Proteus mirabilis > Proteus vulgaris > Enterobacter aerogenes > Pseudomonas aeruginosa and Escherichia coli. The minimum inhibitory concentrations (MIC) against S. aureus was studied by dilution method (6 Æ5 lg ml ) ). A stronger effect was noted with the viperidae venoms ( lg ml ) ) as compared with elapidae venoms (4 lg ml ) ). The MIC were comparable with those of the standard drugs (chloramphenicol, streptomycin and penicillin). Conclusion: The present findings indicate that viperidae (D. russelli russelli) and elapidae (P. australis) venoms have significant antibacterial effects against gram (+) and gram ()) bacteria, which may be the result of the primary antibacterial components of laao, and in particular, the PLA enzymes. The results would be useful for further purification and characterization of antibacterial agents from snake venoms. Significance and Impact of the Study: The activity of LAAO and PLA enzymes may be associated with the antibacterial activity of snake venoms. Introduction Antimicrobial peptides are evolutionarily ancient weapons. Their widespread distribution throughout the animal kingdom suggests that they have served a fundamental role in the successful evolution of complex multicellular organisms (Zasloff ). Antimicrobial peptides have been isolated from a variety of natural sources, including antimicrobial secretions and venoms. Snake venoms contain many proteinaceous components, including neurotoxins (pre and postsynaptic), cardiotoxins, myotoxins, cytotoxins, proteases, nucleases (Stiles et al. 99), l-amino acid oxidase (LAAO) (Iwanaga and Suzuki 979) and phospholipase enzymes that are biologically active. Previous reports point towards the association of some venom with antibacterial activity (Stocker and Traynor 986; Theakston et al. 99; Talan et al. 99). A direct lytic factor from the venom of cobra Hemachatus ª 6 The Authors 65 Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

2 R. Perumal Samy et al. Antibacterial activity of phospholipase A enzymes haemachatus has been reported to show antibacterial property (Aloof-Hirsch et al. 968). Various peptides (ahelical, cationic and pore-forming), derived from the venom of scorpions (Opistophtalmus carinatus), also possess broad spectrum of antimicrobial activity against bacteria and fungi (Moerman et al. ). The a-helical polycationic peptides (pandinin and ) from the scorpion venom have been shown to have high antimicrobial activity against a range of gram-positive bacteria (Corzo et al. ). Moreover, CsTX peptides from the venom of the scorpion, Cupiennus salei also exhibit antimicrobial properties (Xu et al. 989). Other than the venoms of snakes and scorpions, the venom of the common honey bee (Apis mellifera) also displays antimicrobial properties (Fennel et al. 968). Its venom component, mellitin, is more active against gram-positive than gram-negative bacteria. Besides, venoms of the wasps, honey bees and various snakes also contain antimicrobial peptides, but their functions have not been investigated (Blaylock ). To date, only few studies have been made on the antimicrobial activities of snake venoms (Stiles et al. 99; Blaylock ). Antimicrobial peptides from other nonvenom sources are considered to kill bacteria by permeabilizing and/or disrupting their membranes (Zhao et al. ). A possible synergistic action between the antimicrobial peptides and the venom enzymes, like phospholipase A (PLA ), has recently been reported (Zhao and Kinnunen 3), thus suggesting some concerted action between the enzymes and antimicrobial peptides of venoms. This potential synergistic property of venom enzymes may be important, particularly at a time when more antibioticresistant bacterial strains are emerging. The enzymatic degradation of phospholipids in the target bacterial membrane may be one of the important factors in the bactericidal property of animal venoms. Hence, antibacterial properties of 34 different venoms from snakes, scorpions, honey bee, and a variety of purified PLA enzymes and laao were investigated against some clinical isolates of pathogenic (gram-positive and gram-negative) bacteria to evaluate the antibacterial activity of venom enzymes as compared with that of the crude venoms. All the venom enzymes used in the experiment have been purified by successive chromatographic steps with the final purity of at least 95%, as assessed by reversedphase HPLC. Crotoxin from Crotalus durissus terrificus (South American rattlesnake) has a molecular weight of 4 35 Da and a pi of approximately 4Æ7. Crotoxin PLA enzyme is very stable and supplied at 4 C in phosphate buffered saline (ph 7Æ4). laao, purified from the venom of Bothrops atrox and Crotalus adamanteus were obtained commercially (Sigma Aldrich, St Louis, MO, USA), reconstituted with 5 mmol l ) of Tris-HCl buffer (ph 7Æ4), and stored at ) C. Chemicals The following antimicrobial agents: chloramphenicol (CHL-3) 3 lg/disc, streptomycin (STR-) lg/disc and penicillin (P-) lg/disc were obtained from (BBL blank disc, 7-mm diameter) Becton Dickinson Labware, Franklin Lakes, MD 5, USA, and included in the antibacterial test as drug controls and blank discs. Mueller Hinton (MH) agar medium was obtained from Oxoids Pte Lte, UK. One mole per litre tris-hcl buffer (ph 7Æ4) was purchased from National University of Medical Institute (NUMI), Singapore. All the buffers were filtered by Æ lm Nalgo nung filter (Apogent Technologies, Rochester, NY, USA) before use. Molecular weight precision protein standards and acrylamide were obtained from Bio-Rad Laboratories, Hercules, CA, USA. All other reagents were of analytical grade. Bacterial strains Six clinical isolates of bacteria were obtained from the Department of Microbiology, Yong Loo Lin School of Medicine, National University of Singapore. Bacteria used in the present investigation include Escherichia coli, Enterobacter aerogenes, Proteus vulgaris, Proteus mirabilis, Pseudomonas aeruginosa and Staphylococcus aureus. The bacterial cultures were spread and allowed to grow overnight at 37 C on ml of MH agar (ph 7Æ4) plates (9-mm diameter), prior to storage at 4 C. Materials and methods Collection of venoms Lyophilized venoms were obtained from commercial sources (Venom Supplies Pte Ltd, Tanunda, South Australia). Venoms were collected in a sterile manner under strict laboratory conditions, and were centrifuged at 4 C, frozen and lyophilized within 6 h of extraction. The dried venom was normally packed and stored in dark at ) C. Antibacterial effects of crude venoms Lyophilized crude venoms ( lg) dissolved in ml of 5 mmol l ) of Tris-HCl buffer (ph 7Æ4), were filtered using Æ lm syringe filter (Millipore, NY, USA), and stored at 4 C for assay. Susceptibility test was performed by disc-diffusion method (Bauer et al. 966) with the following modifications. Bacterial inoculums [ ll of a Æ Æ A 6 culture containing Æ5 6 to 3Æ 8 colony forming units (CFU) ml ) ] were spread by using a ª 6 The Authors Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

3 Antibacterial activity of phospholipase A enzymes R. Perumal Samy et al. sterile cotton swab onto ml of sterile MH agar plates (9-mm diameter). The surface of the medium was allowed to dry for about 3 min. Sterile paper discs (7- mm diameter) were then placed onto the MH agar surface and ll each of venom samples ( lg ml ) ) were added per disc in five replicates. The discs containing the following antimicrobial agents: chloramphenicol (CHL, 3 lg/disc), streptomycin (STR, lg/disc) and penicillin (P, lg/disc) were used as drug controls, and the blank disc containing ll of 5-mmol l ) tris-hcl buffer (ph 7Æ4), served as a normal control. The plates were incubated at 37 C for 4 h. Zones of inhibition were recorded in millimetre diameters and interpreted as sensitive, intermediate or resistant according to the recommendations of the NCCLS (). Antibacterial effects of purified PLA Purified PLA enzymes crotoxin A, crotoxin B, ammodytoxin A, daboiatoxin, mojavetoxin, b-bungarotoxin, taipoxin, mulgatoxin, bee venom PLA and two LAAO (B. atrox and C. adamanteus), were obtained commer- Table Antibacterial effect of different animal venoms, tested by disc-diffusion method, against some clinical isolates of gram-positive and negative bacteria Micro-organism Common name Scientific name Sa (+) Ea ()) Pv ()) Pm ()) Pa ()) Ec ()) Elapidae Death adder Acanthophis augtra ± Æ7 Common death adder Acanthophis antarcticus Æ ± Æ93 4Æ4 ± Æ84 Northern death adder Acanthophis praelongus 8Æ4 ± Æ7 7Æ7 ± Æ85 8Æ ± Æ44 7Æ ± Æ45 Desert death adder Acanthophis pyrrhus Æ4 ± Æ4 5Æ5 ± Æ9 Hector Androctonus australis 7Æ ±Æ84 Malayan krait Bungarus candidus 5Æ ± Æ3 Indian cobra Naja naja naja 7Æ8 ± Æ Spitting cobra Naja sumatrana 4Æ4 ± Æ5 7Æ9 ± Æ7 4Æ4 ± Æ87 King brown snake Pseudechis australis 9Æ9 ± Æ7 Speckled brown snake Pseudechis guttata 5Æ ± Æ83 8Æ8 ± Æ 7Æ7 ± Æ66 Red-bellied black snake Pseudechis porphyriacus Æ5 ± Æ5 Collett s snake Pseudechis colletti 7Æ ± Æ45 Peninsula brown snake Pseudonaja inframaggula 3Æ3 ± Æ46 Eastern brown snake Pseudonaja textilis 5 ± Æ7 Viperidae Pallas Agkistrodon halys 4Æ ±Æ3 7Æ4 ±Æ89 5Æ4 ±Æ74 7Æ ±Æ83 7Æ9 ±Æ74 Diamondback rattlesnake Crotalus adamanteus 5Æ4 ± Æ5 Æ.7 ± Æ3 5Æ6 ± Æ5 Puff adder Bitis arietans 6 ± Æ43 6 ± Æ83 West African gaboon viper Bitis gabonica rhinoceros 7 ± Æ7 7Æ7 ± Æ85 Russell s viper Daboia russelli russelli 9Æ4 ± Æ89 8 ± Æ7 6Æ4 ± Æ98 6Æ8 ± Æ84 7Æ8 ± Æ83 Burmese viper Daboia russelli siamensis 5Æ ± Æ84 4Æ8 ± Æ83 7Æ5 ± Æ5 Saw-scaled viper Echis carinatus 8Æ6 ± Æ8 Wagler s pit viper Trimeresurus wagleri 5Æ ± Æ9 8Æ4 ± Æ89 Apiidae Honey bee venom Apis mellifera 3Æ ± Æ9 Scorpionidae Scorpion Buthotus hottenota 5Æ4 ±Æ89 hottenota Chinese red scorpion Buthus martensii Karsch 6Æ6 ± Æ89 Antibiotics Chloramphenicol (CHL) 3 lg per disc 35Æ ± Æ6 7Æ8 ± Æ9 4Æ6 ± Æ86 33 ± Æ34 5 ± Æ7 3Æ8 ± Æ64 Streptomycin (STR) lg per disc 3Æ3 ± Æ77 5Æ7 ± Æ44 8Æ ± Æ83 7Æ8 ± Æ9 6Æ5 ± Æ5 9Æ4 ± Æ89 Penicillin (P) lg per disc 7Æ4 ± Æ89 8Æ3 ± Æ85 6Æ7 ± Æ98 7Æ8 ± Æ 6Æ6 ± Æ89 5Æ ± Æ84 The values are presented as mean ± SD (n ¼ 5) and represent a venom inhibition zone in mm, including the 7-mm diameter of the disc, after 4- h incubation. The bacterial inoculum per plate contained Æ5 6 to 3Æ 8 CFU, which were spread onto the agar surface with sterile cotton swap. Sterile paper discs (7-mm diameter) were placed onto the agar surface and ll of venom ( lg ml ) ) added. Micro-organisms: Ec, Escherichia coli; Ea, Enterobacter aerogenes; Pv, Proteus vulgaris; Pm, Proteus mirabilis; Pa, Pseudomonas aeruginosa; and Sa, Staphylococcus aureus. Control (); no activity ()). ª 6 The Authors 65 Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

4 R. Perumal Samy et al. Antibacterial activity of phospholipase A enzymes cially. Each enzyme was dissolved in 5 ll of 5 mmol l ) of Tris-HCl (ph 7Æ4) buffer and mixed by vortex (Labnet VX, Labnet International, USA) to a final concentration of Æ5 lmol l ). In vitro antimicrobial activity was determined by the previously described disc-diffusion method (Bauer et al. 966) with some modifications. Minimum inhibitory concentrations The minimum inhibitory concentrations (MIC) were determined by broth dilution method, proposed by Wu and Hancock (999), for the testing of antimicrobial peptides. Twenty-four-hour-old bacterial cultures were harvested from fresh MH agar plates. The strains were grown in Mueller Hinton broth (MHB) to a mid-logarithmic phase with absorbance at A 6 of Æ Æ (Æ5 6 to 3Æ 8 CFU ml ) ). Whole venom was reconstituted at the required test concentrations in the range of 6, 8, 4,,, 5, Æ5 and Æ5 lg ml ) using mol l ) of Tris-HCl buffer (ph 7Æ4). Two hundred microlitre of a mid-logarithmic phase culture of bacteria was added to ll of venom samples in 96-well bottomed plate. Five independent experiments were performed as replicates. One well containing ll of bacterial inoculates served as a bacterial control, while another well containing ll of uninoculated MH broth and ll of mol l ) of Tris-HCl buffer (ph 7Æ4) were used as a negative control. As positive controls, ll of antibiotics (8 Æ5 lg ml ) ) in mol l ) of Tris-HCl buffer were added to ll of bacterial inoculates for the liquid growth inhibition assays. Culture plates were incubated at 37 C for 4 h. The inhibition of bacterial growth was determined by ELISA reader (Molecular Devices E max precision microplate reader; Research Instruments, Singapore) measuring the absorbance at 56 nm. Results were expressed as MIC, the lowest concentration of venom that reduces growth by more than 5% of the strains. Biochemical characterization PLA enzyme activity PLA catalyses the hydrolysis of phospholipids at the sn- position, yielding a free fatty acid and a lysophospholipid. The Cayman Chemical secretory PLA (spla ) assay kit was used for the measurement of spla. This assay uses the,-dithio analogue of diheptanoyl phosphatidylcholine, which serves as a substrate for most PLA (Reynolds et al. 99). Enzymatic activity, expressed as an increase in absorbance per minute, was converted to specific activity (i.e. micro moles of fatty acid released per minute per milligram of protein). Protein assay The protein concentration of the crude venoms was determined using the Bradford (976) protein assay (Bio- Rad protein assay kit; Hercules, CA, USA). Venom samples were prepared at a concentration of mg ml ). Bovine serum albumin ( mg ml ), A 8 ¼ Æ56) was used as the protein standard. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed according to the method of Laemmli (97). Separating gels containing Æ5% acrylamide and stacking gel of 4Æ5% acrylamide were used. The protein sample (Æ Æ36 mg ml ) ) and Laemmli sample buffer (5 mmol l ) of Tris-HCl, ph 7Æ4 containing % SDS, % -mercaptoethanol, % glycerol, Æ% bromophenol blue), were heated for 5 min in a boiling water bath. Twenty microlitre of each protein sample was then loaded onto a gel, and electrophoresis was carried out at a constant voltage ( V for h). The gel was fixed with 5% acetic acid overnight and stained for h in Æ% Coomassie blue R-5 in 5% acetic acid solution. Destaining was carried out Table In vitro antibacterial activity of purified phospholipase A enzymes from snake venoms, tested by disc-diffusion method Phospholipase A enzymes Scientific name Mol. wt. (kda) Conc. lg ml ) Micro-organisms Sa (+) Ea ()) Pm ()) Pa ()) Ec ()) L-Amino acid oxidase Bothrops atrox 7Æ4 ±Æ73 6Æ8 ±Æ6 (LAAO) LAAO Crotalus adamanteus 7Æ ±Æ7 4Æ7 ±Æ9 Crotoxin B Crotalus durissus terrificus 4Æ3 7Æ7 ± Æ Æ ± Æ93 4Æ5 ± Æ86 5 ± Æ7 Mulgatoxin Pseudechis australis 3Æ 8Æ4 ±Æ89 Daboiatoxin (DbTx) Daboia russelli siamensis 3Æ6 4Æ ± Æ84 Bee venom PLA Apis mellifera 9Æ 3Æ3 ±Æ83 The values are presented as mean ± SD (n ¼ 5) and represent a PLA inhibition zone in mm, including the 7-mm diameter of the disc, after 4-h incubation. Micro-organisms: Ec, Escherichia coli; Ea, Enterobacter aerogenes; Pv, Proteus vulgaris; Pm, Proteus mirabilis; Pa, Pseudomonas aeruginosa; and Sa, Staphylococcus aureus. Control (); no activity ()). ª 6 The Authors Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

5 Antibacterial activity of phospholipase A enzymes R. Perumal Samy et al. in a solution containing 35% methanol and 7% acetic acid, until the background became clear. The molecular weights of the protein bands were determined using Precision protein standard markers ( 5 kda). Gels were imaged using a GS-7 calibrated imaging densitometer scanner (Bio-Rad). Statistical analysis The results (mean ± SD, n ¼ 5) were statistically analysed by one way anova with repeated measures used to analyse factors influencing the size of the growth inhibition zone. Results Antibacterial effects of crude venoms Table summarizes the antibacterial activity of 34 different venoms of snake, scorpion and honey bee. The venoms of four viperidaes (Daboia russelli russelli, Echis carinatus, Bitis gabonica rhinoceros and Bitis arietans) and two elapidaes (Pseudechis australis, Naja naja naja) suppress, especially against S. aureus. Among them, P. australis and D. russelli russelli venoms exhibited the maximum inhibitory zones with potency almost equal to that of standard drugs. The standard drugs (chloramphenicol and streptomycin) suppress the S. aureus bacteria, because the activity of chloramphenicol is significantly higher than that of the crude venoms. On the other hand, the venoms of death adder (Acanthophis praelongus), Australian elapid (Pseudonaja textilis) and scorpions (Buthotus hottenota hottenota) showed only moderate antimicrobial effect against S. aureus. Noticeable exceptions were venoms from black scorpion (Acanthophis crasicuda and B. hottentota), sea snake (Hydrophis cyanocinctus), tiger snake (Notechis ater ater), cobras (Naja kaouthia), Australian elapids (Pseudechis nuchalis and Pseudechis affinis), coastal Taipan (Oxyuranus scutellatus) and tiger keel back (Rhabdophis tigrinus), all of which lacked antimicrobial activity. A broad spectrum of activity was seen with some viperidae venoms, including those of Indian Russell s viper (D. russelli russelli), Pallas (Agkistrodon halys), Northern death adder (A. praelongus), Diamondback rattlesnake (C. adamanteus), Speckled brown snake (Pseudechis guttata), Spitting cobra (Naja sumatrana) and Burmese Russell s viper (Daboia russelli siamensis). Among the gram-positive and gram-negative bacteria tested, S. aureus, P. mirabilis and P. vulgaris were more susceptible to elapid venoms. The order of the susceptibility of the bacteria tested against viperidae venoms is as follows: S. aureus > P. mirabilis > P. vulgaris > E. aerogenes > P. aeruginosa > E. coli. Table 3 Total and specific activity of phospholipase enzyme (PLA ) activity and protein contents of different venom samples Species PLA activity Antibacterial effects of venom enzymes Total activity Specific activity Yield of protein (lmol min ) ) (lmol min ) mg ) ) (mg ml ) ) Elapidae Acanthophis augtra 3Æ74 74Æ5 ±Æ5 Æ Acanthophis antarcticus 9 487Æ5 ±Æ96 Æ78 Acanthophis praelongus 4Æ 46 ± 3Æ84 Æ34 Acanthophis pyrrhus 38 5 ± Æ98 Æ4 Androctonus australis Æ 85Æ ±Æ7 Æ5 Bungarus candidus 34Æ9 66Æ3 ±Æ56 Æ4 Naja naja naja 93Æ4 333 ± Æ7 Æ44 Naja sumatrana 46Æ5 93Æ5 ±Æ6 Æ9 Pseudechis australis 434Æ ± 3Æ Æ Pseudechis guttata 38Æ5 79 ± Æ Æ78 Pseudechis porphyriacus 76Æ7 333 ± Æ7 Æ44 Pseudechis colletti 5Æ7 Æ8 ±Æ89 Æ8 Pseudonaja inframaggula ± 6Æ6 Æ66 Pseudonaja textilis 46Æ8 83Æ3 ±Æ6 Viperidae Agkistrodon halys 86Æ5 57Æ4 ±Æ Æ Bitis arietans 4Æ 48Æ ±Æ7 Æ54 Bitis gabonica rhinoceros 6Æ5 45Æ4 ±Æ57 Æ56 Bothrops atrox 3Æ8 6Æ36 ± Æ6 Æ (L-amino acid oxidase) Crotalus adamanteus 34Æ Æ3 ±Æ Æ34 (L-amino acid oxidase) Crotalus adamanteus 36Æ4 69Æ4 ±Æ46 Æ56 Echis carinatus 53Æ4 6Æ5 ±Æ49 Æ4 Daboia russelli russelli 39Æ8 785Æ ±Æ4 Æ36 Daboia russelli siamensis 6Æ4 54Æ4 ±Æ44 Æ4 Trimeresurus wagleri 4Æ6 38Æ ±Æ6 Æ4 Apiidae Apis mellifera 3Æ7 Æ5 ±Æ Æ36 Scorpionidae Buthotus hottenota 5Æ 38Æ8 ±Æ Æ6 hottenota Buthus martensii Karsch 4Æ6 9Æ4 ±Æ Æ Total activity of PLA enzyme estimated from the whole venoms (lmol min ) ). PLA enzymatic activity (lmol min ) mg ) ). Values are presented as mean ± SD (n ¼ ) of replicates. The antibacterial activity of the crude venoms was then compared with that of the purified PLA enzymes and LAAO. Among the various purified PLA enzymes examined for antibacterial effects, crotoxin B, daboiatoxin, mulgatoxin and bee venom PLA exhibited significant activity against S. aureus, E. coli, P. aeruginosa and E. aerogenes, with the highest activity noted only for the basic PLA crotoxin B (Table ). LAAO was also found to ª 6 The Authors 654 Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

6 R. Perumal Samy et al. Antibacterial activity of phospholipase A enzymes (a) Daboia russelli russelli (b) Agkistrodon halys (c) - 5 kd 5 kd - 5 kd kd 5 kd - 5 kd - - kd kd - kd - - kd kd - kd kd 5 kd - 5 kd - 37 kd - 37 kd kd 5 kd - 5 kd - Pseudechis australis ( c) Pseudechis australis - 5 kd - kd 5 kd - kd - 5 kd - kd - M M M M (d) Echis carinatus M - 5 kd - 5 kd - kd - kd - 5 kd - 5 kd - 5 kd - kd (e) Crotalus adamanteus M - 5 kd - 5 kd - kd - kd - 5 kd - 5 kd - 5 kd - kd (f) M 3 Bitis guttata kd - 5 kd - kd - kd - 5 kd - 5 kd - 5 kd - kd M (g) Pseudonaja nuchalis (h) Oxyuranus scutellatus (i) Naja kaouthia - 5 kd - 5 kd - 5 kd - 5 kd - kd - kd - kd - kd - 5 kd - 5 kd - 5 kd - 5 kd - 5 kd - kd - 5 kd - kd M M M kd - 5 kd - kd - kd - 5 kd - 5 kd - 5 kd - kd Figure Electrophoretic profile of the crude venoms studied. The venom samples were prepared and run in polyacrylamide gel electrophoresis (SDS-PAGE). Staining was done with coomassie brilliant blue R5. (a) Daboia russelli russelli (Viper), (b) Agkistrodon halys (Pallas), (c) Pseudechis australis (Mulga or Australian king brown snake), (d) Echis carinatus (Saw-scaled viper), (e) Crotalus adamanteus (Diamondback rattlesnake), (f) Bitis guttata (Puff adder), (g) Pseudonaja nuchalis (Western brown), (h) Oxyuranus scutellatus (Coastal taipan), (i) Naja kaouthia (Cobra). have strong effect against S. aureus and P. mirabilis. Five enzymes crotoxin A, ammodytoxin A, mojavetoxin, b-bungarotoxin and taipoxin did not show any effect against all the tested organisms; however, none of the enzymes had any activity against P. vulgaris. Minimum inhibitory concentrations The most promising venoms were further studied for MIC by broth dilution method. The inhibitory effect against S. aureus was found to be stronger with the venoms of P. australis, D. russelli russelli and E. carinatus (MIC lg ml ) ) than that seen with the venoms of N. sumatrana, C. adamanteus and A. halys (MIC 4 lg ml ) ). LAAO from C. adamanteus and B. atrox had the highest inhibition against all the organisms tested. The inhibitory action of the venoms of B. gabonica rhinoceros, C. adamanteus, E. carinatus, D. russelli russelli and LAAO (C. adamanteus) against S. aureus was at least twofold higher than that found against P. vulgaris at all concentrations (MIC 6 Æ5 lg ml ) ). The inhibitory effect shown by these venoms against S. aureus growth was more or less equal to that of the standard drugs. PLA enzyme activity The PLA activity varied depending on the venom of different species. Maximal PLA activity was found for the venoms of D. russelli russelli, A. halys, B. gabonica rhinoceros, E. carinatus, C. adamanteus and P. australis. Viperidae and elapidae venoms exhibited a wide range ª 6 The Authors Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

7 Antibacterial activity of phospholipase A enzymes R. Perumal Samy et al. Figure Comparison of the amino acid sequences of Crotoxin (CB) phospholipase A enzymes (Crotalus durissus terrificus) with other PLA of taipoxin alpha chain (Oxyuranus scutellatus scutellatus), ammodytoxin C [ATXC] (Vipera ammodytes ammodytes), Bothropstoxin [BthTX-I] (Bothrops jararacussu), Mojave toxin basic chain [Mtx-b precursor] (Crotalus scutulatus scutulatus), Ecarpholin S-neutral (Echis carinatus), isozyme PA-C (Pseudechis australis), AGKHP (Gloydius halys), b-bungarotoxin [A-chain precursor] (Bungarus multicinctus), CA [Crotapotin] (Crotalus durissus terrificus) and completely conserved residues in all sequences are bolded and marked by asterisks. The gaps are inserted in the sequences in order to attain maximum homology. CLUSTAL W (Æ83) multiple sequence alignment. of PLA enzyme activities (Table 3). However, D. russelli russelli (activity 785Æ lmol min ) mg ) ), A. halys (activity 86Æ5 lmol min ) mg ) ) and P. australis (activity 3949 lmol min ) mg ) ) venoms showed stronger PLA activity as compared with that of the other venoms. Protein determination The protein content was improved in the antibacterial active crude venoms of C. adamanteus (Æ6 mg), D. russelli russelli (Æ4 mg), E. carinatus (Æ4 mg) when compared with that of nonactive venoms of A. pyrrhus (Æ3 mg) and P. colletti (Æ8 mg), respectively. The yield of venom protein content generally varies with the snake species (Table 3). The protein profile analysis (SDS-PAGE) also supported the view that the venoms with potent antibacterial activity apparently contain the small molecule 5-kDa protein (Fig. a f). Other than small molecule proteins, C. adamanteus venom also expressed the major proteins of 5 and 37 kda. In D. russelli russelli venom, 5-kDa proteins were more prominently expressed than other molecular weight proteins. Whereas the nonactive venoms (Fig. g h) did not show the small molecules between 5 and 5 5 kda) proteins when compared with the active venoms. N-terminal sequencing A comparison of the N-terminal sequences of crotoxin B with other venom PLA sequences shows a moderate degree of similarity (Fig. ). The hydrophobicity profiles of purified PLA from venoms were obtained by placing the hydrophobic indices against the sequence residue numbers (Fig. 3). The sequence of crotoxin B exhibited higher hydrophobicity at the C-terminal region ( ) than that of the other venom PLA enzymes. Discussion The present study provides evidence that several venoms of different snake species have antibacterial effects against both gram-positive and gram-negative bacteria. Among the venoms examined, those from four species of viperidae (D. russelli russelli, E. carinatus, B. gabonica rhinoceros and B. arietans) and two species of elapidae (P. australis and N. naja naja) showed strong antimicrobial effects, especially against S. aureus. These venoms exhibited greater zones of inhibition, equivalent to that shown by the standard drugs, chloramphenicol and streptomycin. Compared with the elapidae (P. australis) venom, which was more specific against S. aureus, the venoms of viperidae (D. russelli russelli and A. halys), on the other hand, exhibited a broader spectrum of antibacterial activity. ª 6 The Authors 656 Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

8 R. Perumal Samy et al. Antibacterial activity of phospholipase A enzymes Crotoxin B (CB) 3 Daboiatoxin (DbTx) Mulgatoxin Taipoxin Ammodytoxin A Bee venom PLA Beta bungarotoxin Mojavetoxin Figure 3 Hydropathic profiles of crotoxin b, daboiatoxin, mulgatoxin, taipoxin, ammodytoxin A, bee venom PLA, b-bungarotoxin and mojavetoxin were calculated by using the kyte-doolittle method. ª 6 The Authors Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

9 Antibacterial activity of phospholipase A enzymes R. Perumal Samy et al. A strong activity was shown against S. aureus by the venoms of viperidae (B. gabonica rhinoceros and B. arietans) and crotalidae (C. adamanteus), while most venoms (A. australis, N. sumatrana, P. guttata, A. halys, B. gabonica rhinoceros, D. russelli russelli) exhibited only a weaker activity against E. coli. In contrast, a stronger activity on E. coli had been reported previously for A. mellifera, Naja sputatrix, V. russellii and C. adamanteus venoms (Stocker and Traynor 986). Noticeable exceptions were venoms from Buthus hottenlota (black scorpion), Buthus martensii (Chinese red scorpion), H. cyanocinctus (sea snake), R. tigrinus (Tiger keel back), N. ater ater (Krefft s), N. kaouthia (Cobras), P. affinis (Dugite), P. nuchalis (Western brown) and Oxyuranus scutellatus (Coastal taipan), which lacked any antimicrobial activity against all the tested bacteria. The venoms with potent antibacterial activity were then chosen for further studies on MIC, as determined by broth dilution method. The MIC against S. aureus was found to be higher with the venoms of P. australis, D. russelli russelli and E. carinatus (MIC lg ml ) )at the lowest dilution than that seen with the venoms of N. sumatrana, C. adamanteus and A. halys (MIC 4 lg ml ) ). Previously, crotalid (C. adamanteus) venoms were reported to exhibit strong inhibitory effect (MIC 8 lg ml ) ) against Staphylococci, P. aeruginosa, Enterobacter, Citrobacter, Proteus and Morganella species (Talan et al. 99). In the present study, bacterial growth was completely inhibited against P. aeruginosa by C. adamanteus, E. carinatus, P. guttata and D. russelli russelli venoms, but the inhibitory effect was lost with prolonged incubation (after 4 h) for many gram-negative species. Venoms from B. arietans, B. candidus, C. adamanteus and N. naja naja were only moderately effective against P. mirabilis at all dilutions (MIC 6 Æ5 lg ml ) ). Comparison of the inhibitory effect of the venoms with that of the purified LAAO showed that the inhibition of the C. adamanteus and B. atrox LAAO against P. mirabilis and S. aureus was significantly higher than that of the crude C. adamanteus venom. Association between an antibacterial property of snake venom and LAAO had previously been reported (Stiles et al. 99). The LAAO from Trimeresurus jerdonii venom inhibited the growth of E. coli, S. aureus, P. aeruginosa and Bacillus megaterium (Lu et al. ), while the LAAO from A. halys (Pallas) exhibited inhibitory activity against bacteria (E. coli) and fungi (Yan et al. ). Besides the antibacterial effect exhibited by venom LAAO, the inhibitory activity seen with the venoms may also be to the result of the multiple biological effects of snake venom PLA (Chioato and Ward 3). PLA enzymes are present in almost all venoms, but their prevalence is particularly higher in viper venoms. The maximum PLA activity was found in the venoms of D. russelli russelli, A. halys, B. gabonica rhinoceros, E. carinatus and C. adamanteus. When the antibacterial activity of purified PLA was compared, crotoxin B from the highly toxic South American rattlesnake (C. durissus terrificus) venom exhibited the most potent activity against S. aureus, E. coli, P. aeruginosa and E. aerogenes. Mammalian spla have been implicated in lipid digestion on host defence mechanisms, including antibacterial defence (Valentin and Lambeau ). A comparison of the N-terminal sequences of crotoxin B with other venom PLA shows a moderate degree of similarity. Crotoxin B is a basic neurotoxic PLA (C. durissus terrificus) containing three chains which enhance its lethal potency (Bouchier et al. 99). It exhibits higher hydrophobicity at the C-terminal region ( ) than those of other PLA enzymes. The C-terminal region of basic PLA Bothrops asper myotoxin-ii has been suggested as the segment responsible for its cytotoxic and myotoxic effects (Lomonte et al. 994, 3). Antimicrobial peptides inhibit microbes by scrambling of the usual distribution of lipids between the leaflets of the bi-layer of the cell wall, thus resulting in the disturbance of membrane function, and the damaging of critical intercellular targets after internalization of the peptides. It appears that the enzyme hydrophobicity might be playing an important role for the antimicrobial action on bacteria. In conclusion, the in vitro screening provides convincing evidence that several venoms have most promising antibacterial effects against gram (+) and gram ()) bacteria. The present findings indicate that viperidae (D. russelli russelli, E. carinatus, P. guttata) and elapidae (P. australis) venoms have significant antibacterial effects, which may be the result of the primary antibacterial components of LAAO, and in particular, the PLA enzymes. The results will be useful for further purification and characterization of antibacterial agents from snake venoms. Acknowledgements The authors are thankful to the Defence Science and Technology Agency (DSTA), Singapore, for the financial support (Grant No R ) to carry out this work. We thank the Department of Microbiology, Yong Loo Lin School of Medicine, National University of Singapore, for their support and permission to use bacterial cultures in this investigation. References Aloof-Hirsch, S., Devries, A. and Berger, A. (968) The direct lytic factor of cobra venom: purification and chemical characterization. Biochem Biophys Acta 54, ª 6 The Authors 658 Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)

10 R. Perumal Samy et al. Antibacterial activity of phospholipase A enzymes Bauer, A.W., Kirby, W.M., Sherris, J.C. and Turck, M. (966) Antibiotic susceptibility testing by a standardized single disk method. Am J Clin Path 45, Blaylock, R. () Antibacterial properties of KwaZulu natal snake venoms. Toxicon 38, Bouchier, C., Boulain, J.C., Bon, C. and Menez, A. (99) Analysis of cdnas encoding the two subunits of crotoxin, a phospholipase A neurotoxin from rattlesnake venom: the acidic non enzymatic subunit derives from a phospholipase A -like precursor. Biochem Biophys Acta 88, Bradford, M.M. (976) A rapid and sensitive method for quantification of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 7, Chioato, L. and Ward, R.J. (3) Mapping structural determinants of biological activities in snake venom phospholipases A by sequence analysis and site directed mutagenesis. Toxicon 4, Corzo, G., Escoubas, P., Villegas, E., Barnham, K.J., He, W., Norton, R.S. and Nakajima, T. () Characterization of unique amphipathic antimicrobial peptides from venom of the scorpion Pandinus imperator. J Biochem 359, Fennel, J.E., Shipman, W.H. and Cole, L.J. (968) Antibacterial action of melittin, a polypeptide from the venom. Proc Soc Exp Biol Med 7, Iwanaga, S. and Suzuki, T. (979) Enzymes in snake venom. In Handbook of Experimental Pharmacology ed. Lee, C.Y., Vol. 5. New York. Springer-Verlag, pp Laemmli, U.K. (97) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 7, Lomonte, B., Angulo, Y. and Santamaria, C. (3) Comparative study of synthetic peptides corresponding to region 5 9 in Lys49 myotoxic phospholipase A from snake venoms. Toxicon 4, Lomonte, B., Tarkowisky, A., Bagge, U. and Hanson, L.A. (994) Neutralization of the cytolytic and myotoxic activities of phospholipase A from Bothrops asper snake venom by glycosaminoglycans of the heparin/heparin sulphate family. Biochem Pharmacol 47, Lu, Q.M., Wei, Q., Jin, Y., Wei, J.F., Wang, W.Y. and Xiong, Y.L. () l-amino acid oxidase from Trimeresurus jerdonii snake venom: purification, characterization, platelet aggregation-inducing and antibacterial effects. J Nat Toxins, Moerman, L., Bosteels, S., Noppe, W., Willems, J., Clynen, E., Schoofs, L., Thevissen, K., Tytgat, J. et al. () Antibacterial and antifungal properties of a-helical, cationic peptides in the venom of scorpions from the South Africa. Eur J Biochem 69, NCCLS () Performance Standards for Antimicrobial Susceptibility Testing, Twelfth Informational Supplement. NCCLS Document no. M-S. Villanova, PA: National Committee for Clinical Laboratory Standards. Reynolds, L.J., Hughes, L.L. and Dennis, E.A. (99) Analysis of human synovial fluid phospholipase A on short chain phosphatidycholine-mixed micelles: development of a spectrophotometric assay suitable for a microtiterplate reader. Ann Biochem 4, Stiles, B.G., Sexton, F.W. and Weinstein, S.A. (99) Antibacterial effects of different snake venoms: purification and characterization of antibacterial proteins from Pseudechis australis (Australian king brown or mulga snake) venom. : Toxicon 9, 9 4. Stocker, J.F. and Traynor, J.R. (986) The action of various venoms on Escherichia coli. J Appl Bacteriol 6, Talan, D.A., Citron, D.M., Overturf, G.D., Singer, B., Froman, P. and Goldstein, E.J. (99) Antibacterial activity of crotalid venoms against oral snake flora and other clinical bacteria. J Infect Dis 64, Theakston, R.G.D., Phillips, R.E., Looareesuwan, S., Escheverria, P., Makin, T. and Warrell, D.A. (99) Bacteriological studies of the venom and mouth cavities of wild Malayan pit viper (Calloselasma rhodostoma) in Southern Thailand. Roy Soc Trop Med Hyg 84, Valentin, E. and Lambeau, G. () What can venom phospholipases A( ) tell us about the functional diversity of mammalian secreted phospholipases A( )? Biochemistry 8, Wu, M. and Hancock, R.E. (999) Interaction of the cyclic antimicrobial cationic peptide bactenecin with the outer and cytoplasmic membrane. J Biol Chem 74, Xu, K., Ji, Y. and Qu, X. (989) Purification and characterisation of an antibacterial peptide from venom of Lycosa singoriensis. Acta Zool Sinica 35, Yan, X.M., Zhang, S.Q., Chang, Q., Liu, P. and Xu, J.S. () Antibacterial and antifungal effects of Agkistrodon halys Pallas: purification of its antibacterial protein-lao. Shi Yan Sheng Wu Xue Bao 33, Zasloff, M. () Antimicrobial peptides of multicellular organism. Nature 45, Zhao, H. and Kinnunen, P.J.K. (3) Modulation of the activity of secretory phospholipase A by antimicrobial peptides. Antimicrob Agents Chemother 47, Zhao, H., Rinaldi, A.C., Rufo, A., Bozzi, A., Kinnunen, P.K.J. and Di Giulio, A. () Structural and charge requirements for antimicrobial peptide insertion into biological and model membranes. In Pore-forming Peptides and Protein Toxins ed. Menestrina, G., Dalla Serra, M. and Lazarovici, P. pp Reading, United Kingdom: Harwood Academic Publishers. ª 6 The Authors Journal compilation ª 6 The Society for Applied Microbiology, Journal of Applied Microbiology (7)