Purification and structural characterisation of phospholipase A1 (Vespapase, Ves a 1) from Thai banded tiger wasp (Vespa affinis) venom

Toxicon 61 (2013) 151–164

Contents lists available at SciVerse ScienceDirect

Toxicon journal homepage: www.elsevier.com/locate/toxicon

Purification and structural characterisation of phospholipase A1 (Vespapase, Ves a 1) from Thai banded tiger wasp (Vespa affinis) venom Sophida Sukprasert a, Prapenpuksiri Rungsa a, Nunthawun Uawonggul c, Paroonkorn Incamnoi a, Sompong Thammasirirak a, Jureerut Daduang b, Sakda Daduang a, * a

Protein and Proteomics Research Group, Department of Biochemistry, Faculty of Science, Khon Kaen University, Khon Kaen 40002, Thailand Department of Clinical Chemistry, Faculty of Associated Medical Sciences, Khon Kaen University, Khon Kaen 40002, Thailand c Faculty of Liberal Arts and Science, Nakhon Phanom University, Nakhon Phanom 48000, Thailand b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 June 2012 Received in revised form 24 September 2012 Accepted 30 October 2012 Available online 15 November 2012

The Thai banded tiger wasp (Vespa affinis) is one of the most dangerous vespid species in Southeast Asia, and stinging accidents involving this species still cause fatalities. In the present study, four forms of V. affinis phospholipase A1 were identified through a proteomics approach. Two of these enzymes were purified by reverse-phase chromatography, and their biochemical properties were characterised. These enzymes, designated Ves a 1s, are not glycoproteins and exist as 33441.5 and 33474.4 Da proteins, which corresponded with the 34-kDa band observed via SDS-PAGE. The thermal stabilities of these enzymes were stronger than snake venom. Using an in vivo assay, no difference was found in the toxicities of the different isoforms. Furthermore, the toxicity of these enzymes does not appear to be correlated with their PLA1 activity. The cDNAs of the full-length version of Ves a 1s revealed that the Ves a 1 gene consists of a 1005-bp ORF, which encodes 334 amino acid residues, and 67- and 227-bp 50 and 30 UTRs, respectively. The two isoforms are different by three nucleotide substitutions, resulting in the replacement of two amino acids. Through sequence alignment, these enzymes were classified as members of the pancreatic lipase family. The structural modelling of Ves a 1 used the rat pancreatic lipaserelated protein 2 (1bu8A) as a template because it has PLA1 activity, which demonstrated that this enzyme belongs to the a/b hydrolase fold family. The Ves a 1 structure, which is composed of seven a-helixes and eleven b-strands, contains the b-strand/εSer/a-helix structural motif, which contains the Gly-X-Ser-X-Gly consensus sequence. The typical surface structures that play important roles in substrate selectivity (the lid domain and the b9 loop) were shortened in the Ves a 1 structure, which suggests that this enzyme may only exhibit phospholipase activity. Moreover, the observed insertion of proline into the lid domain of the Ves a 1 structure is rare. We therefore propose that this proline residue might be involved in the stability and activity of Ves a 1s. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Allergen Hymenoptera Molecular modelling Phospholipase A1 Wasp venom Vespapase

1. Introduction Hymenoptera venoms are well-known to contain a variety of major protein allergens, including antigen 5,

* Corresponding author. Tel./fax: þ66 43 342 911. E-mail address: [email protected] (S. Daduang). 0041-0101/$ – see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.toxicon.2012.10.024

phospholipase A1B, hyaluronidase and protease (Abe et al., 2000; Hoffman, 1978; King et al., 1984). These components are reported to be life-threatening and accelerate fatal anaphylactic reactions in allergic patients. Phospholipase A1s (PLA1s) hydrolyse the sn-1 position of the phospholipid packing of several types of biological membranes, thereby releasing lysophospholipids and free fatty acids as reaction products. Vespid PLA1s cause local inflammatory reactions

152

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

(Ho et al., 1993; King et al., 2003) and act as allergens. In animals, these PLA1s can also cause severe haemolysis with consequent cardiac dysfunction and death (Ho and Hwang, 1991; Ho et al., 1993; Ho and Ko, 1988). In addition, these enzymes have been reported to activate platelet aggregation and induce thrombosis in vivo (Yang et al., 2008). However, until now, the characterisation of these proteins was still incomplete, and their biological functions remained unclear. All of the extracellular PLA1s, such as human pancreatic lipase (HPL), lipoprotein lipase (LPL), endothelial lipase (EL), pancreatic lipase-related protein 2 (PLRP2), phosphatidylserine (PS)-specific PLA1 (PS-PLA1), and the PLA1 in vespid venoms, belong to the pancreatic lipase gene family. Carriere et al. (1998) divided the pancreatic lipase gene family into eight subfamilies. Vespid PLA1s form part of subgroup number eight. PLA1s from different sources exhibit significant conserved region similarity to HPL and PLRP2. These enzymes all contain the typical catalytic triad residues Ser-His-Asp (Aoki et al., 2007). The activity of these enzymes significantly increases in the presence of Ca2þ. Crystallographic studies of HPL show that each lipase is composed of a large N-terminal domain and a smaller Cterminal domain. The large N-terminal domain belongs to the a/b hydrolase fold and is essential for catalytic activity. The Gly-X-Ser-X-Gly region is a consensus sequence that, to date, has only been found in lipases and esterases. The tertiary structure dominants, the b5 and b9 loops and the lid domain, are believed to play an important role in catalysis. The b5 loop is involved in the formation of an oxyanion hole, whereas the b9 loop and the lid domain are involved in substrate recognition and substrate selectivity (Carriere et al., 1998; Winkler et al., 1990). The Vespinae subfamily of wasps in Thailand includes 18 species. However, the biochemical, pharmacological and immunological characteristics of their venoms are still unclear. Vespa affinis, the banded tiger wasp, is mostly distributed in forests throughout Thailand. V. affinis is considered to be one of the most dangerous species among vespids. Land abuse contributes to the invasion of these species in human habitats, which results in the many outof-record stinging accidents that occur every year. In addition, two Asian wasp envenomation cases of V. affinis in Nepal have been reported. The symptoms found in these cases were very serious and showed evidence of haemolysis, hepatic dysfunction, oligoanuria and azotaemia (Das and Mukherjee, 2008). A better understanding of PLA1 wasp venom would prove valuable for both therapeutics and commercial applications. Therefore, in the present work, we have identified, purified, biochemically characterised, sequenced and modelled the PLA1 isoforms present in the venom of the Thai banded tiger wasp, V. affinis. 2. Materials and methods 2.1. Venom collection Worker wasps were collected from Siang Sao Village (Sri Songkram District, Nakornpanom Province, Thailand) and immediately shocked in ice. The venom reservoirs of the wasps were pulled out and gently squeezed. The droplet of

venom that appeared at the tip of the sting was immediately collected in a 1.5-mL microcentrifuge tube. The crude venom was then maintained at 80  C until use. The protein contents were quantitatively determined by the Bradford method (1976) using bovine serum albumin as the standard. 2.2. Polyacrylamide gel electrophoresis (PAGE) One-dimensional SDS-PAGE was performed following the standard method using a 13% (w/v) separating gel and a 4% (w/v) stacking gel. Phosphorylase B (97 kDa), bovine serum albumin (66 kDa), chicken ovalbumin (45 kDa), carbonic anhydrase (30 kDa), trypsin inhibitor (20.1 kDa) and a-lactalbumin (14.4 kDa) were used as the standards. After the samples were applied to the gel, the proteins were resolved at 150 V for 1 h. The gels were stained with silver staining solution. For the two-dimensional polyacrylamide gel electrophoresis (2D-PAGE), the crude venom protein (80 mg) or purified protein (20 mg) were mixed with a rehydration solution containing 7 M urea, 2 M thiourea, 2% CHAPS, 7 mg/2.5 mL DTT, 2% IPG buffer and 1% bromophenol blue to obtain a total volume of 125 mL. This solution was applied to IPG dry strips (pH 3-11 NL, GE Healthcare). After rehydration for 12 h, isoelectric focussing (IEF) was performed through the application of 500 V for 30 min, 1000 V for 30 min and 5000 V for 1.4 h at a charge of 50 mA per strip. After the IEF, reduction and alkylation steps were completed, the IPG strips were placed on a 13% (w/v) SDSPAGE gel, which was used for the second dimension. The electrophoresis was performed for 15 min at 10 mA/gel and then at 20 mA/gel until the dye front reached the bottom of the gel. After staining with colloidal Coomassie Brilliant Blue G-250 or silver nitrate, the gels were scanned using a flatbed scanner. The data were analysed using ImageMaster 2D Platinum, version 5.0 (GE Healthcare, Sweden). The identities of the proteins were determined by mass spectrometry. 2.3. Protein identification by liquid chromatography coupled with mass spectrometry (LC–MS/MS) This procedure was previously described by Sukprasert et al. (2012). Briefly, the excised spots were washed and trypsinised overnight at 37  C. After extraction and washing, the digested peptides were separated by nanoscale LC using a NanoAcquity system (Waters Corp., USA), which was equipped with a Symmetry C18 (5 mm, 180mm  200-mm) Trap column and a BEH130C18 (1.7 mm, 100mm  100-mm) analytical reversed-phase column (Waters Corp., USA). A flow rate of 350 mL/min was used in this separation. Water and acetonitrile, both of which contained 0.1% formic acid, were used as solvents A and B, respectively. All of the samples were analysed in triplicate. The analysis of the tryptic peptides was performed using a SYNAPTÔ HDMS mass spectrometer (Waters Corp., UK). All of the analyses were performed using the positive nanoelectrospray ion mode. The time-of-flight analyser of the mass spectrometer was externally calibrated with [Glu1] fibrinopeptide B from m/z 50 to 1600, and the

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

acquisition lock mass was corrected using the monoisotopic mass of the doubly charged precursor of [Glu1] fibrinopeptide B. The quadrupole mass analyser was adjusted such that the ions from m/z 200 to 1990 were efficiently transmitted. The BioWorks 3.2 software (Thermo Electron, USA) was used to process and convert the data into a Mascot Generic File. 2.4. Database searching The ion spectra of the peptides were searched using an MS/MS ion search (http://www.matrixscience.com/cgi/ search_form.pl?FORMVER¼2&SEARCH¼MIS) of the SWISSPROT databases. The following search parameters were used: a specified trypsin enzymatic cleavage with one possible missed cleavage; þ/ 0.6 Da mass tolerances for MS/ MS; a peptide tolerance of 1.2 Da; 1þ, 2þ, or 3 þ ions; methionine oxidation variable modification; carbamidomethyl (C) fixed modification; monoisotopic mass; and 20 responses. A theoretical pI determination was performed using the free online software provided by ExPASy (http://br. expasy.org/cgi-bin/peptide-mass.pl?P14790). 2.5. Protein purification The crude venom was diluted in 50 mM Tris–HCl (pH 8.0) before it was loaded onto a C4 reverse-phase HPLC column (Jupiter, 300A, 250  4.6 mm, Phenomenex, USA) at a flow rate of 1 mL/min. The elution used a linear gradient of 20–70% acetonitrile in 0.1% trifluoroacetic acid (Chou and Hou, 2008). The ODs at 220 and 280 nm were monitored. The purity of the protein was confirmed by 13% (w/v) SDSPAGE under reducing and non-reducing conditions. 2.6. Enzymatic assay The phospholipase A1 activity was assayed using the EnzChekÒ Phospholipase A1 Assay Kit (Invitrogen, Life technologies, USA) according to the manufacturer’s protocol. LecitaseÒ Ultra was used as the standard. All of the measurements were performed in quadruplicate. The phospholipase A1 reaction buffer in the absence of enzyme was used as the negative control. 2.7. Molecular mass determination The matrix-assisted laser desorption/ionisation time-offlight mass spectrometry (MALDI-TOF MS) technique was used to obtain the molecular masses of the purified proteins. The MALDI mass spectra were recorded on a 4800 MALDI-TOF mass spectrometer (Applied Biosystems, USA) equipped with a 2 GHz LeCroy digitiser and a 337 nm N2 laser in linear mode. The spectrometer was calibrated for a mass range of 10–50 kDa using carbonic anhydrase as the standard. Fifty shots were used per position, and 3000 total shots were combined for the analysis. The purified proteins were spotted at 1:10 and 1:1000 dilutions with sinapinic acid at 10 mg/mL in 30% acetonitrile with 0.1% trifluoroacetic acid. The ProteoMassÔ Peptide & Protein MALDI-MS Calibration Kit (Sigma, USA) was used as the control standard. The instrumental parameters were the following:

153

positive polarity; acceleration voltage, 25.13 kV; IS/2, 23.58; focussing lens voltage, 6.03 kV; and extraction delay, 100 ns. The program used to generate the mass lists was Compass 1.2 for flexSeries (FC/FA 3.0 Service Release 1, Bruker Daltonik, Germany). 2.8. Thermostability screening The thermostability of the enzymes was determined through the modified PCY agarose plate method using soy bean lecithin as the substrate (Kim and Rhee, 1994; Song and Rhee, 2000). The PCY agar mixture (1% agarose gel, 0.8% soy bean lecithin, 1% NaCl, 0.25% taurocholic acid and 20 mM CaCl2) was prepared on Petri dishes. The samples were diluted in 50 mM Tris–HCl (pH 8.0) or heated at 100  C for 5 min and then loaded. After a 30-min incubation, the halo sizes were observed. 2.9. Pro-QÒ Emerald 300 glycoprotein gel stain After electrophoresis, the gel was immediately fixed through the addition of the fixing solution and incubated at room temperature with gentle agitation for 30 min to ensure that the SDS was fully removed. The carbohydrates were oxidised in an oxidising solution and then washed again. The gels were then stained with Pro-QÒ Emerald 300 staining solution (Molecular ProbesÔ, Invitrogen Detection Technologies, USA) for 120 min (protected from light) and washed. The gels were then scanned using the 300-nm UV transilluminator in the Gel DocÔ 2000 system (Bio-Rad, USA). 2.10. SYPROÒ Ruby protein gel stain After staining with the Pro-QÒ Emerald 300 staining solution, the gel was rinsed twice in deionised water for at least 5 min. The gel was then incubated directly in the SYPROÒ Ruby gel stain solution (Molecular ProbesÔ, Invitrogen Detection Technologies, USA) overnight and in the dark to obtain maximum signal strength. The gel was then transferred to a clean container and washed with washing solution to reduce the fluorescence background and increase the sensitivity. The gel was viewed and photographed using the UV imaging system (optimal with UV trans-illumination at 300 nm) in the Gel DocÔ 2000 system (Bio-Rad, USA). 2.11. Paralytic dose 50 (PD50) assay Crickets (Gryllus sp.) were used to determine the PD50. The venom activity was quantified by injecting PBS buffer (pH 7.4), crude venom or purified protein into the abdomen of crickets. Each test used six crickets and was performed in triplicate. The non-paralysed crickets were counted after injection for 30 min; the crickets that were unable to turn over from the overturned position were considered to be paralysed. The PD50 value, which was defined as the amount of venom that was able to paralyse 50% of the venom-injected crickets, was then estimated (Uawonggul et al., 2007).

154

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

2.12. Gene-specific primer design

2.14. DNA sequencing and data analysis

PLA1 gene-specific primers (GSPs) were designed based on the sequence similarities of the conserved region of the PLA1s from several sources: lipase, pancreatic lipase gene family and vespid PLA1s (Table 1). The nucleotide sequences used were the following: gij124015212j PS-PLA1 (Homo sapiens); gij190140j lipase; gij126318j HPL (H. sapiens); gij1708841j PLRP2 (H. sapiens); gij548449jDol m 1.01 (Dolichovespula maculata); gij1709542j Dol m 1.02 (D. maculata); gij1709545j Ves m 1 (Vespula maculifrons); gij1352699j Ves v 1 (Vespula vulgaris); gij122112895j Ves g 1 (Vespula germanica); gij14423833j Pol a 1 (Polistes annularis); gij75007953j Pol d 1.01 (Polistes dominulus); and gij166216292j Poly p 1 (Polybia paulista). The PLA1-GSPs were designed for the rapid amplification of the cDNA ends (50 and 30 RACE) using the Primer-BLAST software provided by NCBI (http://www.ncbi.nlm.nih.gov/tools/primer-blast/ index.cgi?LINK_LOC¼BlastHome) and the free Oligo Analyser 1.1.2 software (www.genelink.com).

The sequencing was performed with the DYEnamic E Dye Terminator Cycle sequencing kit using the MegaBACEÔ 1000 instrument (GE Healthcare, UK) and the M13F and M13R universal primers. The sequence was confirmed through the use of the BigDye Terminator v3.1 cycle sequencing kit on ABI (Applied Biosystems Applera, 1st BASE, Malaysia). The data processing relied on the Cimarron 3.12 software (GE Healthcare, Thailand). The basic characterisations of the gene and protein sequences were analysed using NCBI (http:// www.ncbi.nlm.nih.gov/Database/index.html) and the basic local alignment search tool (BLAST, http://www.ncbi.nlm. nih.gov/BLAST/). The ClustalW2 (http://www.ebi.ac.uk/ Tools/msa/clustalw2/) and PSI-BLAST programs (http:// www.ebi.ac.uk/Tools/sss/psiblast/) were used to analyse the homology and the conserved domains of the PLA1 group against the EMBL and SWISSPROT databases. The mass and pI of the proteins were computed using the pI/MW tool of the ExPASy Proteomics Server (http://au.expasy.org/tools/pi_ tool.html). The DiANNA 1.1 web server (http://clavius.bc. edu/clotelab/DiANNA/) was used to predict the formation of disulfide bonds. An evolutionary tree was constructed using MEGA5 (Tamura et al., 2011). The evolutionary history was inferred using the Neighbour-Joining method (Saitou and Nei, 1987).

2.13. RT-PCR and rapid amplification of cDNA ends (50 -RACE and 30 -RACE) The total RNA was isolated from the venom apparatus (gland and reservoir) using the Illustra RNAspin Mini RNA Isolation Kit (GE Healthcare, UK). RT-PCR was performed using RevertAidÔ First Strand cDNA Synthesis Kit (Fermentas, Singapore), as described in the instruction manual. VaPLA1F2, VaPLA1F3 and VaPLA1R3 were used as the GSPs for the RT-PCR. The RACE procedures were performed using the RACE System (Invitrogen, Life Technologies). The firststrand cDNAs for the 30 RACE were synthesised from the total RNAs (5 mg), which were used as templates for the Adapter primer (provided in the kit). The 30 RACE was performed using VaPLA1F1 and VaPLA1F2 as the GSPs. For the 50 RACE, the single strands were synthesised from the total RNA using VaPLA1R3 and cDNAR2 as the GSPs. The nested PCR of the 50 RACE was performed using VaPLA1R1 and VaPLA1R2 as the nested GSPs. The RACE (30 or 50 ) was performed using the PCR master mix reagent kit (Fermentas, Singapore) with the Taq DNA polymerase. The first round of PCR was performed for 32 cycles, each of which consisted of 30 s at 94  C, 1 min at 50  C and 2 min at 72  C. In the second round of PCR, or the nested PCR, the amplicon from the first round was used as the template. The PCR products, which contained a single A overhang at the 30 end, were cloned into a vector with compatible T overhangs (pGEM-T Easy vector, Promega, USA) for sequencing.

2.15. Three-dimensional molecular modelling Three-dimensional models were constructed using the Swiss-Model System, the automated protein homologymodelling server at ExPASy (Switzerland) (Arnold et al., 2006; Kiefer et al., 2009; Peitsch, 1995) using rat PLRP2 (PDB code: 1bu8A), which exhibits PLA1 activity, as a template (Roussel et al., 1998). The full cDNA sequence was blasted against protein databases using PSI-BLAST to determine the template. The three-dimensional models were visualised, analysed and compared using DeepView (Swiss PDB Viewer, http://spdbv.vital-it.ch/) and PyMOL version 1.3r1 (http://www.pymol.org/). 3. Results 3.1. Protein identification by mass spectrometry In this study, a proteomics approach was used to identify the PLA1s in the V. affinis venom. Approximately 17 protein spots were clearly visualised after colloidal Coomassie Brilliant Blue G-250 staining (Fig. 1). Based on the

Table 1 Gene-specific primers and PCR product sizes. Forward primers

Reverse primers

Product size (bp)

VaPLA1F2: 50 - TGGTGGGACACAGTTTGGGCG-30 (GSP for RT-PCR) VaPLA1F3: 50 - CATGGTTTTACTTCGTCTGC-30 (GSP for RT-PCR) VaPLA1F1: 50 -GGTCGATTGGCGGATGGCTGC-30 (GSP for 30 RACE) VaPLA1F2: 50 - TGGTGGGACACAGTTTGGGCG-30 (GSP for 30 RACE) VaPLA1R3: 50 -CACGGCTCTCGTATGAGAGC-30 (GSP1 for 50 RACE cDNA synthesis) cDNAR2: 50 -ATTTGAACATAATTTGCGTC-30 (GSP2 for first round PCR of 50 RACE) VaPLA1R1: 50 -GGCCCAGCAGGATCAAGCCC-30 (nested GSP primer for 50 RACE) VaPLA1R2: 50 -GCGCCCAAACTGTGTCCCACC-30 (nested GSP primer for 50 RACE)

VaPLA1R3: 50 -CACGGCTCTCGTATGAGAGC-30 VaPLA1R3: 50 -CACGGCTCTCGTATGAGAGC-30 Abridged universal amplification primer (AUAP) Abridged universal amplification primer (AUAP)

301 548 900 700

Abridged anchor primer (AAP) Abridged universal amplification primer (AUAP) Abridged universal amplification primer (AUAP)

800 750 650

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

155

and 13) are isoenzyme PLA1s in the V. affinis venom. Therefore, we propose that there are at least 4 isoforms of PLA1 in the V. affinis venom, which are designated “Ves a 1” or “Vespapase”. 3.2. Purification of Ves a 1 The crude V. affinis venom was fractionated through reverse-phase HPLC into eleven fractions (Fig. 2A). All of the fractions were tested for phospholipase activity through a modified PCY assay. Fractions VIII and IX showed relatively high activity because these exhibited a cloudy halo through the catalysis of the degradation of lecithin on the plate assay (Fig. 2B). Therefore, the purity of both fractions was then examined by 13% SDS-PAGE. A single major band, which was approximately 34 kDa in size was found under reducing and non-reducing conditions (Fig. 2C) which was related with molecular masses of 33441.50 and 33474.20 Da for Fractions VIII and IX, respectively (Fig. 2D and E). Therefore, Fractions VIII and IX were designated “Ves a 1.01” and “Ves a 1.02”, respectively, to discriminate between the two isoforms of “Ves a 1”. The Coomassie-stained PVDF membrane did not show any terminal residues, which assumes an N-terminal block. In addition, the position of the two purified Ves a 1s was investigated by 2D-PAGE. As shown in Fig. 3, the Ves a 1.01 (Fig. 3A) and Ves a 1.02 (Fig. 3B) appeared at the same position as spots 9 and 10 (Fig. 1), respectively. This result indicated that spot 9 was Ves a 1.01, whereas spot 10 was Ves a 1.02.

Fig. 1. Representative 2D-PAGE profile of the Vespa affinis venom. In the first dimension (IEF), approximately 80 mg of crude venom protein was loaded onto a 7-cm IPG strip with a non-linear gradient at a pH range of 3–11. After the second dimension, which involves electrophoresis with 13% SDS-PAGE, the protein spots were visualised through the colloidal Coomassie Brilliant Blue G-250 staining solution. The four spots numbered 9, 10, 11, and 13 (indicated by arrows) were identified by LC–MS/MS analysis. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

MW and pI of the vespid PLA1s reported in the literature, the four major V. affinis PLA1 spots, which are indicated by arrows in Fig. 1, are the following: spots numbered 9, 10, 11 and 13, which have pI and MW values of approximately 10, 9.58, 9.96, and 8.85 and 35, 34, 33, and 35 kDa, respectively. These spots were trypsinised and subsequently identified by LC-MS/MS. As shown in Table 2, all of the spots were similar to the Dol m 1 allergen, which is a PLA1 found in white-face hornet (D. maculata) venom (Accession No. P53357). This result indicated that the four spots (9, 10, 11

3.3. PLA1 activity of Ves a 1 A fluorometric commercial substrate was used to verify the position specificity of these isoenzymes. The enzymatic activity of Ves a 1.01 and 1.02 was 3.6 and 6.3 U/mL, respectively. The crude venom exhibited an activity of 10 U/mL at a protein concentration of 5 mg/mL compared with the 5 U/mL activity of lecitase (Fig. 4). Ves a 1.02 exhibited a higher activity than the other proteins tested.

Table 2 Identification of proteins in the venom of the Thai banded tiger wasp (V. affinis). Spot no.

Matched protein

Accession no.

Theoretical MW/pIa

Experimental MW/pIb

Peptide massc

Peptide sequencesd

Score XCe

Species

9

Phospholipase A1

P53357

33781.87/8.92

35/10.00

Dolichovespula maculata

Phospholipase A1

P53357

33781.87/8.92

34/9.58

30.22

Dolichovespula maculata

11

Phospholipase A1

P53357

33781.87/8.92

33/9.96

30.21

Dolichovespula maculata

13

Phospholipase A1

P53357

33781.87/8.92

35/8.85

RNECVCVGLNAKE (4) RAVKYLTECIRR (3) RLIGHSLGAQIAGFAGKE (16) KLVPEEISFVLSTRE (17) RNECVCVGLNAKE RLIGHSLGAQIAGFAGKE KLVPEEISFVLSTRE RNECVCVGLNAKE (4) RLIGHSLGAQIAGFAGKE (4) KLVPEEISFVLSTRE (2) RNECVCVGLNAKE (3) KLVPEEISFVLSTRE

38.27

10

1264.46 1253.50 1540.79 1490.73 1264.46 1540.79 1490.73 1264.46 1540.79 1490.73 1264.46 1490.73

20.19

Dolichovespula maculata

a Theoretical molecular weight (MW) obtained after the LC–MS/MS analysis. The pI values were calculated using the ExPASy Peptide Mass program (http://br.expasy.org/cgi-bin/peptide-mass.pl?P14790). b The experimental MW and pI of the protein spots were determined by the 2D platinum software (GE Healthcare). c Theoretical molecular mass of the peptide fragment. d Deduced peptide sequence obtained after LC–MS/MS (the number of matching peptides is indicated in parentheses). e Score XC obtained after LC–MS/MS analysis.

156

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

Fig. 2. Purification and characterisation profile of the V. affinis PLA1s. A, Crude venom was injected into a C4 Jupiter column (250  4.6 mm) reverse-phase HPLC and eluted with a gradient of 20–70% acetonitrile in 0.1% TFA at a flow rate of 1.0 mL/min. B, fractions VIII and IX, which were designated Ves a 1.01 and Ves a 1.02, displayed phospholipase activity through the PCY plate assay. C, purity assay of fractions VIII and IX by 13% SDS-PAGE under reducing conditions with silver staining. M, low molecular weight marker (GE Healthcare). C, crude venom. D and E, MALDI-TOF mass spectra of Ves a 1.01 and Ves a 1.02, respectively.

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

157

Fig. 3. 2D-PAGE analysis of purified Ves a 1s (V. affinis PLA1s). Approximately 20 mg of Ves a 1.01 (A) and Ves a 1.02 (B) were loaded onto a 7-cm IPG strip with a non-linear gradient at the pH range of 3–11. After electrophoresis (second dimension) using 13% SDS-PAGE, the gels were stained with silver nitrate solution. The PLA1 spots are indicated with arrows.

This result indicated that the cleavage of both enzymes occurred at the sn-1 position, which confirms that these are both PLA1 enzymes (E.C. 3.1.1.32).

Ves a 1s may exhibit stability only when they are found together in the venom. 3.5. Glycoprotein detection

3.4. Thermostability assay The thermostability of the Ves a 1s was investigated using the lecithin plate assay, in which the enzymatic activity is evoked as a halo. After it was exposed to 100  C for 5 min, the activity of the V. affinis venom was still partially retained, as observed through the cloudy halo. In contrast, the transparent halo produced by the cobra venom (Naja naja siamensis) was completely diminished by the heat treatment (Fig. 5). A 10-min exposure to the high temperature caused the activity of the crude venom to be completely lost (data not shown). This result suggested that the venom exhibits high thermostability, depending on the isoforms that are present. The thermal stability of the purified Ves a 1s was also investigated. The results showed that these appear to completely lose their activity after the heat treatment (data not shown), which implies that the

Fig. 4. Enzymatic activity of the Ves a 1s (V. affinis PLA1s). The assay, which used a fluorometric substrate (Invitrogen), was performed in quadruplicate. A protein concentration of 5 mg/mL was used. A lecitase concentration of 5 U/mL was used as the positive control (P).

The presence of the attached carbohydrate in the Ves a1s structure was explored using the CandyCaneÔ Glycoprotein molecular weight standard and lysozyme as the positive and negative controls, respectively. After staining with the fluorescent probe (Pro-QÒ Emerald 300 solution), V. affinis PLA1s did not exhibit any carbohydrate attachment (Fig. 6A). This gel was post-stained with SYPROÒ Ruby solution to

Fig. 5. Thermostability assay of the Ves a 1s (V. affinis PLA1s). A PCY gel containing lecithin was used. Cobra venom, Naja naja siamensis; wasp venom, Vespa affinis; buffer (50 mM Tris–HCl, pH 8.0, negative control). The reaction mixture was heated at 100  C for 5 min before it was loaded into the wells.

158

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

Fig. 6. Glycoprotein detection of the Ves a 1s (V. affinis PLA1s). After electrophoresis using 13% SDS-PAGE under reducing conditions, the gel was stained using the Pro-QÒ Emerald 300 glycoprotein gel staining solution (A) and the SYPROÒ Ruby staining solution (B). Lane 1, low molecular weight marker (GE Healthcare); Lane 2, crude venom protein; Lanes 3-4, purified Ves a 1.01 and Ves a 1.02, respectively; Lane 5, lysozyme; and Lane 6, CandyCaneÔ glycoprotein molecular weight standard (Molecular Probes, Invitrogen, USA). A 0.1 mg/mL protein concentration of each sample was loaded.

evoke the total protein (Fig. 6B). The results of this assay confirmed that these enzymes are not glycoproteins. 3.6. In vivo toxicity assay To determine the toxicity of the V. affinis PLA1s, the in vivo PD50 was assayed using crickets. The crickets were injected with PBS as a mock control. The PD50 of the crude venom in the crickets was approximately 12.5 mg/g body weight (Table 3). If the amount of crude venom was increased to 100 mg/g body weight, all of the injected crickets died (data not shown). Furthermore, a 120 mg/g body weight concentration of purified Ves a 1.02 was able to paralyse 67% of the crickets (4/6). However, if a concentration of 60 mg/g body weight of the mixture of both forms was used, only two of the six crickets were paralysed. These results indicated that the toxicities of both isoforms of Ves a 1 were equal. 3.7. cDNA cloning and sequence analysis of Ves a 1 The cDNA encoding the PLA1s found in the venom gland of V. affinis was cloned and sequenced. RT-PCR and 4 rounds of RACE were used to determine the complete nucleotide sequence of the Ves a 1s. The Ves a 1 precursor contained a 67-bp 50 -untranslated region (UTR), a 1005-bp open

Table 3 Viability of crickets (Gryllus sp.) after injection with crude venom. Concentration of V. affinis venom injected (mg/g body weight)

Number of paralysed crickets after venom injection/total number of cricketsa

1.25 2.5 7.5 12.5 25 50

0/6(3) 1/6(2), 0/6 2/6(3) 3/6(3) 5/6(2), 4/6 6/6(3)

a The number of replicates is indicated in parentheses. The PD50 is defined as the amount of venom that was able to paralyse 50% of the venom-injected crickets; the crickets that were unable to turn over from the overturned position were considered paralysed.

reading frame (ORF) and a 227-bp 30 -UTR. The ORFs consisted of a 99-bp predicted leader sequence, which corresponds to 33 amino acid (aa) residues, and a 906-bp mature coding region encoding 301 aa residues (Fig. 7). As shown in Fig. 7, nucleotide variants between isoforms were found at three positions: A to G substitution at the position 499; C to G at 501 and G to C at 895. These nucleotide substitutions resulted in amino acid substitutions of isoleucine to valine and glutamic acid to glutamine at positions 167 and 299, respectively. The two resulting complete pla1 genes were designated vesa 1.01 and vesa 1.02, which corresponded to Ves a 1.01 and Ves a 1.02, respectively. The sequence alignment exhibited higher homology similarity (53–77%) with the vespid PLA1s than with the related proteins in the lipase superfamily (24–28%) (Fig. 8A). Three amino acid residues of the catalytic triad  serine, aspartate and histidine  are identical to those observed in the members of the pancreatic lipase family. Through alignment analysis, the b-strand/εSer/ahelix structural motif was found to include the Gly-X1-SerX2-Gly sequence, which is a lipase-specific consensus sequence; this sequence was also found in Ves a 1. Thus, Ves a 1 belongs to the eighth subfamily of the pancreatic lipase gene family (Carriere et al., 1998). The alignment revealed that the thirteen cysteine positions were highly conserved among all of the vespid venom PLA1s. As shown through the Ves a 1 sequences, twelve of the cysteines, Cys4-Cys245, Cys87-Cys228, Cys176-Cys219, Cys181-Cys262, Cys240-Cys246, Cys269-Cys294, are linked through six disulfide bridges, whereas Cys267 is a free sulfhydryl group (Fig. 8B). The tertiary structure elements, such as the b5, b9 loops and lid domain, that are typically found in the pancreatic lipase family were also found in Ves a 1. Interestingly, a shortening of the b9 loop and lid domain of Ves a 1 was common among the vespid venom PLA1s, which suggests that these enzymes only retain the phospholipase activity. The phylogenetic relationships between Ves a 1 and its related proteins were constructed using vitellogenin-1, a non-enzymatic yolk protein from Drosophila melanogaster, as the out group. The results of this study demonstrated that Ves a 1 is found in the same cluster as Ves m 1, Ves v 1, Ves g 1, and Dol m 1. However, due to their

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

159

Fig. 7. cDNA and deduced amino acid sequences of the Ves a 1s (V. affinis PLA1s). The numbering corresponds to the Ves a 1.01 sequence. The 50 and 30 UTRs are indicated by small letters. The nucleotide substitutions are indicated by small letters, and the three nucleotide substitution positions that result in the two amino acid substitutions are underlined. The signal peptide, which consists of 33 amino acid residues, is shown in italics. The catalytic triad (Ser, Asp and His) is bolded and italicised. The stop codon is indicated with an asterisk.

different subfamilies, the Ves a 1 cluster is different from the cluster in which Poly p 1, Pol a 1 and Pol d 1.01 are found. All vespid PLA1s were monophyletic group, whereas the related proteins were found in non-monophyletic groups (Fig. 9). Ves a 1 showed closed evolution with Dol m 1, which is in accordance with the identity analysis obtained through sequence alignment (Fig. 8A).

3.8. Three-dimensional models of Ves a 1 The X-ray crystal structure of rPLRP2, which has a 1.80A resolution, was used as a template for computational homology modelling (Fig. 10A). The sequence alignment of Ves a 1 with the template showed a 24.7% sequence identity with an E value of 0.00e-1. The superimposition of Ves

160

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

Fig. 8. Sequence analysis of Ves a 1 (V. affinis PLA1). A, comparison of the structural feature dominants, the lid domain and the b9 loop of Ves a 1, with those of related proteins that belong to the lipase family. The catalytic triad (Ser, Asp and His) is indicated with an asterisk. The lid domain and the b9 loop are shown with thick lines. The conserved cysteine positions among the vespid PLA1s are indicated with black spots. The analysed sequences are the following: spjP16233j HPL (Homo sapiens); spjP54315j 2PPL (H. sapiens); spjP81139j GPLRP2 (Cavia porcellus); spjQ53H76j PS-PLA1 (H. sapiens); spjP06858j LPL (H. sapiens); spjP51528j Ves m 1 (Vespula maculifrons); spjP49369j Ves v 1 (Vespula vulgaris); spjQ3ZU95j Ves g 1 (Vespula germanica); spjQ9U6W0j Pol a 1 (Polistes annularis); spjQ6Q252j Pol d 1.01 (Polistes dominulus); spjA2VBC4j Poly p 1 (Polybia paulista); spjQ06478j Dol m 1.01 (Dolichovespula maculata); and Ves a 1 (this study). B, disulfide linkages of Ves a 1. The twelve cysteine residues that form disulfide bridges (Cys4-Cys245, Cys87-Cys228, Cys176-Cys219, Cys181-Cys262, Cys240-Cys246, Cys269-Cys294) are linked by solid lines.

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

Fig. 9. Evolutionary relationships of the Ves a 1 (V. affinis PLA1) with related proteins. The analysis involved 14 amino acid sequences. All of the positions that contained gaps or missing data were eliminated. The analysed sequences are detailed in Fig. 8. The vitellogenin-1 (spjP02843j Drosophila melanogaster) sequence was used as an out group.

a 1 with the template revealed that the core region around the active site was highly conserved. However, Ves a 1 completely lacked the C-terminal domain, which is always present in the pancreatic lipase family (Fig. 10B). The sequence alignment of Ves a 1 and the template revealed that Ves a 1 is composed of seven a-helixes and eleven bstrands (Fig. 10C). The modelling revealed that the full Ves a 1 sequence corresponded only to the N-terminal domain, which is important for catalytic activity. The dominant structural features, the b9 loop and the lid domain, which are important for substrate selectivity, are shortened in the Ves a 1 structure (Fig. 11A). The comparison of the dominants with related proteins showed the presence of shortened b9 loops and lid domains in Ves a 1 and Dol m 1, which are believed to explain why these enzymes only display PLA1 activity (Fig. 11B). 4. Discussion Vespid venoms are still widely investigated, and their components, activity and applications are interesting topics among researchers in toxicology and other related fields of study. This study, for the first time, identified the biochemical and bioactivity characteristics of the venom of Thai wasps. Although these enzymes showed PLA1 activity, which is commonly found in other vespid venoms, the Ves a 1s from the V. affinis (Thai wasps) venom were characterised as novel molecules. In this study, some of the characteristics of Ves a 1 were found to be significantly different from those of other PLA1s in the pancreatic lipase and Vespidae families. First, this study describes the existence of at least four Ves a 1 isoforms in the venom. Second, this is the first report that demonstrates the thermal stability of the PLA1 activity of vespid venom through the use of the PCY assay at a high temperature. Third, the observed proline insertions in the lid domain of Ves a 1s, which are related to their function, are structurally different from other PLA1s.

161

The isoforms of the PLA1s in vespid venom are not surprising, although the importance of the substitutions of only a few amino acid residues remains unknown. Although the isoforms of PLA1s in vespid venom were discovered recently (Ho et al., 1999; Moawad et al., 2005; Santos et al., 2010, 2011; Soldatova et al., 1993), the functional characterisations of these have not yet been completely determined. In the present investigation, four isoforms of Ves a 1 were identified and partially matched with Dol m 1. The proteomics approach revealed the four Ves a 1 PLA1s (spots 9, 10, 11, and 13) that are present in the V. affinis venom. After a one-step purification, the Ves a 1s were separated into two distinct peaks with relatively close retention times, which indicates the high similarity of their primary structures. After the positions of the two purified Ves a 1s were confirmed by 2D-PAGE, the Ves a 1.01 corresponds to spot 9 and appears to mix with spot 11, whereas Ves a 1.02 was found in spot 10. Spot 13 could not separate by RP-HPLC due to its very low abundance in the venom. The presence of the Ves a 1 PLA1 isoforms in the V. affinis venom may provide these species an advantage for their adaptation and survival in the natural environment. Ves a 1.01 and 1.02 were isoenzymes with slightly different molecular masses (33441.50 and 33474.20 Da, respectively) that differed by only 32.7 Da. Similarly, the isolated verutoxin 2a and 2b, which are the isotoxins with PLA1 activity found in the Vespa verutina venom, exhibit only a 14-Da difference in their masses and a similar toxicity (Ho et al., 1999). We thus hypothesised that the Ves a 1 isoforms might have a similar structure and function. The toxicities of the different isoforms of Ves a 1 were investigated through a PD50 assay. We found that the PD50 of the mixture of the two isoforms was decreased by 50% compared with the PD50 of one of the purified isoforms. This result confirmed that, although Ves a 1.02 has higher PLA1 activity than Ves a 1.01, the two isoforms exhibit no difference in their toxicity. The data imply that the toxicity of the Ves a 1s does not correlate with their PLA1 activity. The Ves a 1s appear to be blocked at the N-terminus through an a-amino group. This blocking was predicted to be due to post-translational acetylation processes (Ribeiro et al., 2004). Nevertheless, we attempted to predict the N-terminus of Ves a 1 through sequence alignment. The deduced amino acids of Ves a 1 (from the phenylalanine residue at position 34 to the phenylalanine residue at position 58) showed 100% identity with the N-termini of the isotoxins from the V. verutina venom (Ho et al., 1999), as shown in Table 4. Moreover, the phenylalanine of Dol m 1 is its first N-terminal residue (Soldatova et al., 1993). Thus, the phenylalanine at position 34 can likely be considered an Nterminal residue of the mature Ves a 1 sequence. Furthermore, the theoretical/experimental (determined by MS) molecular masses of Ves a 1.01 and 1.02 were 33381.73/ 33441.50 and 33366.73/33474.20 Da, respectively. The mass shifts were therefore 59.77 and 107.47 Da, respectively (Table 4), and are presumably caused by the occurrence of acetylation and cysteine oxidation within the molecules. The thermostabilities of the Ves a 1 isoforms were also investigated. The cystine-rich isoforms improved the structural stabilisation of Ves a 1s. The result clearly

162

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

Fig. 10. Structural analysis of Ves a 1 using rat pancreatic lipase-related protein 2 (rPLRP2, PDB code: 1bu8A) as the template. The model was obtained using the automated Swiss PDB system and visualised using the PyMOL software. A, structural feature dominants, b5 loop, b9 loop and lid domain of the template model. B, superimposition of Ves a 1 and the template model. C, amino acid sequence alignment between Ves a 1 and the template.

showed that the V. affinis venom displays thermal stability, whereas snake venom, which is known to contain many forms of phospholipase, is not thermally stable. Interestingly, the Ves a 1s contain a proline in the lid domain, which is peculiar in vespid PLA1s. Therefore, we hypothesised that this proline residue might play an important role in the enzyme-substrate stabilisation during hydrolysis or might protect the enzyme from exposure to high temperatures. A high proline content has been previously

reported in the a and b phospholipases found in the V. mandarina venom (Abe et al., 2000). These enzymes showed a broad optimal temperature range of 40–60  C, which implies that proline might be involved in the structural stabilisation of these proteins. The lecithin plate assay showed a difference in the halos produced by wasp and snake venoms. This difference is due to the different types of phospholipases that are present in the two venoms, which produce lysolecithins and free fatty

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

163

Fig. 11. Structural features of the Ves a 1s (V. affinis PLA1s). A, three-dimensional modelling of Ves a 1. B, comparison of the structural dominants of Ves a 1 and related proteins (HPL, human pancreatic lipase; rPLRP2, rat pancreatic lipase-related protein 2; GPLRP2, guinea pig pancreatic lipase-related protein 2; and Dol m 1, PLA1 from D. maculata venom). The amino acid residues are numbered according to the HPL sequence.

acids, respectively; the free fatty acids then form complexes with calcium, thereby leading to the formation of a white zone (Ramrakhiani and Chand, 2011). Vespid venoms are defined as A1B, whereas snake and bee venoms are type A2 (King et al., 1984). The A1B venom is known to catalyse the hydrolysis of acyl-ester bonds at both the sn-1 and the sn-2 positions of phosphatidylcholine, whereas the A2 type catalyses the hydrolysis of the ester bond at the sn-2 position only, thereby releasing fatty acids and 1lysophospholipids. The PLA1 found in the Thai tropical red fire ant (Solenopsis geminata) venom also produces a cloudy halo on the PCY plate, which is similar to that produced by the V. affinis venom (data not shown). Our findings clearly show that the phospholipases in hymenopteran venom produce a cloudy halo on a plate assay that contains soy bean lecithin as the substrate. The sequence alignment and modelling of Ves a 1 with related proteins revealed the different features of Ves a 1. (1) The first 24 residues, which are always present in the large catalytic N-terminal domain of pancreatic lipases, were absent in the Ves a 1 structure. (2) The small C-terminal

domain (residues 336 through 446), which is the colipase binding domain, was completely absent in the Ves a 1 and Dol m 1 sequences. As a consequence, these enzymes are unable to hydrolyse triglycerides (Aoki et al., 2007; Carriere et al., 1998; Roussel et al., 1998; Winkler et al., 1990; Withers-Martinez et al.,1996). (3) The b9 loop of the Ves a 1s was shortened through the deletion of residues 206 through 213. (4) The lid domain of the Ves a 1s was shortened through the deletion of residues 242 through 258. (5) As shown in Fig.11B, the glutamine at position 82 of the b5 loop was replaced by a threonine residue in the Ves a 1 structure. This substitution is predicted to eliminate the salt bridge that stabilises the interaction between the b5 loop and the core of the protein, as described previously in the open forms of the HPL and GPLRP2/HPL proteins (WithersMartinez et al., 1996). Therefore, the b5 loop of the Ves a 1s may adopt a different conformation. (6) The shortening of the b9 loop near the active site of Ves a 1s may contribute to the increase in the accessibility of the Ves a 1 active site. The Leu215 in Ves a 1 and Dol m 1 superimposes with the Phe215 of HPL, rPLRP2 and GPLRP2, which suggests that this

Table 4 N-terminal partial sequences and molecular masses of the Ves a 1s (V. affinis) and the VT-2a and 2b verutoxins (V. verutina). Protein

N-terminal sequences

Molecular mass (Da)

Theoretical mass (Da)e

Ves a 1.01 Ves a 1.02 VT-2a VT-2b

34

33441.50c 33474.20c 33360d 33374d

33381.73 (þ59.77) 33366.73(þ107.47) – –

a

FNPCPYSDDTVKMIITLRENKKHDF58 a FNPCPYSDDTVKMIITLRENKKHDF58 a 1 FNPCPYSDDTVKMIILTRENKKHDF25 b 1 FNPCPYSDDTVKMIILTRENKKHDF25 b 34

Deduced amino acid sequences from the cDNA template. Amino acid sequences from the automated Edman degradation (Ho et al., 1999). Molecular mass determined with the MALDI-TOF mass spectrometer. d Molecular mass determined with the ESI mass spectrometer (Ho et al., 1999). e Theoretical mass calculated using the ExPASy Peptide Mass program (http://br.expasy.org/cgi-bin/peptide-mass.pl?P14790). The mass shifts are indicated in parentheses. b c

164

S. Sukprasert et al. / Toxicon 61 (2013) 151–164

residue may be useful in the acyl chain stabilisation of the substrate. (7) The lid domain of Ves a 1 showed a 16-residue deletion. The observed insertion of Pro255 in the lid domain of Ves a 1 is rare. We therefore propose that this residue might enhance the hydrophobic properties of the lid domain, which is putatively stabilised for substrate binding, and might exhibit a weak triglyceride lipase activity. Similarly, although it has a short lid and a short b9 loop, Dol m 1 exhibits weak lipase activity (Soldatova et al., 1993; Withers-Martinez et al., 1996). This is the first report of the PLA1 isoforms that are found in the venom of Thai wasps (V. affinis). The immunoreactive properties of Ves a 1s should therefore be investigated through additional structural and functional analyses. Acknowledgements This work was supported by a CHE Grant, which is funded by the program for “Strategic Scholarships for Frontier Research Network for the Ph.D. Program Thai Doctoral Degree” (CHE-PhD-THA-SUPV), awarded to S.S. During the 2007-2010 fiscal years, this work was also supported by the TRF-CHE Research Grant for Mid-Career University Faculty, which is jointly funded by the Thailand Research Fund (TRF) and the Office of the Higher Education Commission (CHE) of the Thai Ministry of Education. Additional support was obtained from “The Khon Kaen University (KKU) Research Fund” during the 2007–2009 fiscal years. We would like to thank Ms. Sirinee Poonchaisri for the identification of the Thai wasp species. The authors also thank Prof. Yukifumi Nawa and the KKU Publication Clinic of Khon Kaen University for their critical review of the manuscript. Conflict of interest The authors declare that there are no conflicts of interest. References Abe, T., Sugita, M., Fugigura, T., Hiyoshi, J., Akasu, M., 2000. Giant hornet (Vespa mandarinia) venomous phospholipase: the purification, characterization and inhibitory properties by biscoclaurine alkaloids. Toxicon 38, 1803–1806. Aoki, J., Inoue, A., Makide, K., Saiki, N., Arai, H., 2007. Structure and function of extracellular phospholipase A1 belonging to the pancreatic lipase gene family. Biochimie 89, 197–204. Arnold, K., Bordoli, L., Kopp, J., Schwede, T., 2006. The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22, 195–201. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 248–254. Carriere, F., Withers-Martinez, C., Tilbeurgh, H.V., Roussel, A., Cambillau, C., Verger, R., 1998. Structural basis for the substrate selectivity of pancreatic lipases and some related proteins. Biochimica et Biophysica Acta 1376, 417–432. Chou, C.-C., Hou, M.-H., 2008. Crystallization and preliminary X-ray diffraction analysis of phospholipase A1 isolated from hornet (Vespa basalis) venom. Acta Cryst F64, 1118–1120. Das, R.N., Mukherjee, K., 2008. Asian wasp envenomation and acute renal failure: a report of two cases. MJM 11, 25–28. Ho, C.L., Hwang, L.L., 1991. Local edema induced by the black-bellied hornet (Vespa basalis) venom and its components. Toxicon 29, 1033–1042.

Ho, C.L., Hwang, L.L., Chen, C.T., 1993. Edema-inducing activity of a lethal protein with phospholipase A1 activity isolated from the black-bellied hornet (Vespa basalis) venom. Toxicon 31, 605–613. Ho, C.L., Ko, J.L., 1988. Purification and characterization of a lethal protein with phospholipase A1 activity from the hornet (Vespa basalis). Biochim. Biophys. Acta 963, 414–422. Ho, C.L., Lin, Y.L., Li, S.F., 1999. Three toxins with phospholipase activity isolated from the yellow-legged hornet (Vespa verutina) venom. Toxicon 37, 1015–1024. Hoffman, D.R., 1978. Allergens in hymenoptera venom. V: identification of some of the enzymes and demonstration of multiple allergens in yellow jacket venom. Annals of Allergy 40, 171–176. Kiefer, F., Arnold, K., Künzli, M., Bordoli, L., Schwede, T., 2009. The SWISSMODEL repository and associated resources. Nucleic Acids Research 37, D387–D392. Kim, M.K., Rhee, J.S., 1994. Simple and rapid screening methods for microorganisms with phospholipase A1, A2 and C activities. Biotechnology Techniques 8, 635–638. King, T.P., Jim, S.Y., Wittkowski, K.M., 2003. Inflammatory role of two venom components of yellow jackets (Vespula vulgaris): a mast cell degranulating peptide mastoparan and phospholipase A1. International Archives of Allergy and Immunology 131, 25–32. King, T.P., Kochoumian, L., Joslyn, A., 1984. Wasp venom proteins: phospholipase A1 and B. Archives of Biochemistry and Biophysics 230, 1–12. Moawad, T.I., Hoffman, D.R., Zalat, S., 2005. Isolation, cloning and characterization of Polistes dominulus venom phospholipasee A1 and its isoforms. Acta Biologica Hungarica 56, 261–274. Peitsch, M.C., 1995. Protein modeling by e-mail. Bio/Technology 13 (13), 658–660. Ramrakhiani, L., Chand, S., 2011. Recent progress on phospholipases: different sources, assay methods, industrial potential and pathogenicity. Applied Biochemistry and Biotechnology 164, 991–1022. Ribeiro, S.P., Mendes, M.A., Santos, L.D., Souza, B.M., Marques, M.R., Azevedo Jr., W.F., Palma, M.S., 2004. Structural and functional characterization of N-terminally blocked peptides isolated from the venom of the social wasp Polybia paulista. Peptides 25, 2069–2078. Roussel, A., Yang, Y., Ferrato, F., Verger, R., Cambillau, C., Lowe, M., 1998. Structure and activity of rat pancreatic lipase-related protein 2. The Journal of Biological Chemistry 273, 32121–32128. Saitou, N., Nei, M., 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4, 406–425. Santos, L.D., Santos, K.S., Pinto, J.R.A., Dias, N.B., Souza, B.M., Santos, M.F., Perales, J., Domont, G.B., Castro, F.M., Kalil, J.E., Palma, M.S., 2010. Profiling the proteome of the venom from the social wasp Polybia paulista: a clue to understand the envenoming mechanism. Journal of Proteome Research 9, 3867–3877. Santos, L.D., Silva Menegas, A.R., Santos Pinto, R.A., Santos, K.S., Castro, F.M., Kalil, J.E., Palma, M.S., 2011. Proteomic characterization of the multiple forms of the PLAs from the venom of the social wasp Polybia paulista. Proteomics 11, 1403–1412. Soldatova, L., Kochoumian, L., King, T.P., 1993. Sequence similarity of a hornet (D. maculata) venom allergen phospholipase A1 with mammalian lipases. FEBS Letters 320, 145–149. Song, J.K., Rhee, J.S., 2000. Simultaneous enhancement of thermostability and catalytic activity of phospholipase A1 by evolutionary molecular engineering. Applied and Environmental Microbiology 66, 890–894. Sukprasert, S., Uawonggul, N., Jamjanya, T., Thammasirirak, S., Daduang, J., Daduang, S., 2012. Characterization of the allergen Sol gem 2 from the fire ant venom, Solenopsis geminata. Journal of Venomous Animals and Toxins Including Tropical Diseases 18, 334–343. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and Evolution 28, 2731–2739. Uawonggul, N., Thammasirirak, S., Chaveerach, A., Arkaravichien, T., Bunyatratchata, W., Ruangjirachuporn, W., Jearranaiprepame, P., Nakamura, T., Matsuda, M., Kobayashi, M., Hattori, S., Daduang, S., 2007. Purification and characterization of heteroscorpine-1 (HS-1) toxin from Heterometrus laoticus scorpion venom. Toxicon 49, 19–29. Winkler, F.K., D’Arcy, A., Hunziker, W., 1990. Structure of human pancreatic lipase. Nature 343, 771–774. Withers-Martinez, C., Carriere, F., Verger, R., Bourgeois, D., Cambillau, C., 1996. A pancreatic lipase with a phospholipase A1 activity: crystal structure of a chimeric pancreatic lipase-related protein 2 from guinea pig. Structure 4, 1363–1374. Yang, H., Xu, X., Ma, D., Zhang, K., Lai, R., 2008. A phospholiapse A1 platelet activator from the wasp venom of Vespa magnifica (Smith). Toxicon 51, 289–296.