Background
Snake venoms are chemically complex mixtures of pharmacologically active proteins or peptides, which serve not only as a source of digestive enzymes, but also play an important role in immobilizing the prey and acting as offensive weapons. They can target multiple tissues, causing simultaneous damage of multiple physiological systems. One of the components which contribute significantly to the lethality of snake venoms is phospholipase A
2 (PLA
2) (EC 3.1.1.4) [
1]. PLA
2 constitutes a family of structurally related proteins hydrolyze phospholipids at the sn-2 position in a calcium-dependent manner, releasing fatty acids and lysophospholipids [
2]. Snake venom PLA
2s are low-molecular weight (13,000–14,000 Da), secretory phospholipases containing seven disulfide bonds. Based on their amino acid sequence and disulfide bond pattern, snake venom PLA
2s are classified into group I PLA
2 (from Elapidae/Hydrophidae) or group II PLA
2 (from Crotalidae/Viperidae) [
3]. Usually, the group II PLA
2s are further subdivided into two major subgroups: the Asp-49 PLA
2s (D49 PLA
2s), which have an aspartic acid at position 49 and high catalytic activity towards artificial phospholipid substrates; and Lys-49 PLA
2s (K49 PLA
2s), which have a lysine substitutes at position 49 and very low or no hydrolytic activity towards artificial phospholipid substrates [
4,
5]. Recently, a unique subgroup of snake venom group II PLA
2, named N49 PLA
2 subgroup was identified from several Asiatic snake venoms [
6‐
8]. The N49 PLA
2 was found to differ from the other subgroups in its structure and biological activities.
Besides the digestive function, snake PLA
2s exhibit severalother pharmacological properties including antiplatelet [
9,
10], anticoagulant [
11], hemolytic [
9], neurotoxic (presynaptic) [
12], myotoxic [
13‐
15]. They have also been employed widely as pharmacological tools to investigate the roles of these enzymes in diverse models of experimental inflammatory processes such as edema, inflammatory cell infiltration and mast cells activation [
15‐
20]. Mast cells are primarily located in mucosal and perivascular areas of various tissues, which play an important role in body defense processes. Recent studies found that mast cells can be activated by snake venom and release carboxypeptidase A and possibly other proteases, which can degrade venom components [
21,
22]. Our former study also showed that atrahagin, a metalloprotienase purified from
Naja atra snake, could potently activate human colon, lung and tonsil mast cells to release histamine [
23]. Several snake venom PLA
2s were reported to be able to activate the rat mast cells, to induce microvascular leakage and inflammatory cell accumulation at the sites of inflammation [
15‐
20]. However, little is known of the action of N49 PLA
2s on human mast cells, and the mechanisms through which N49 PLA
2 induces microvascular leakage and inflammatory cell accumulation still remain obscure. Therefore, we investigated the mechanisms of TM-N49 [
6] in induction of microvascular leakage and mast cell accumulation and activation in the present study.
Discussion
As a novel subgroup of snake venom group II PLA2, the proinflammatory activities of N49 PLA2 have not yet been studied. Therefore, this is the first study to investigate the potential actions of N49 PLA2 in inflammation. TM-N49 was found to be a potent stimulus of microvascular leakage in the skin of rats. The potency of this unique, inactive PLA2 in induction of microvascular leakage is similar to the potency of bradykinin and histamine on a molar basis. Therefore it is not difficult to anticipate that TM-N49 by itself should be able to induce severe inflammatory reactions upon snake bite.
Since histamine/5-HT antagonist cyproheptadine abolished the TM-N49-induced microvascular leakage, but selective histamine H
1 receptor antagonist terfenadine did not, the action of TM-N49 on skin edema seems to be histamine independent. It is likely that 5-HT and other mast cell products, such as PAF, serine proteases, leukotrienes, prostaglandins are involved in the process [
24]. Indeed, it was observed that ginkgolide B, a PAF receptor antagonist inhibited microvascular leakage induced by TM-N49, suggesting that PAF was involved in the microvascular leakage process. Mast cell mediated edema formation in response to different snake venom PLA
2 has been reported previously. Thus, PLA
2 from
Vipera russeli,
Naja mocambique and honey bee produced a rapid but transient oedematous response in rat paw [
25]. Bothropstoxin-I and bothropstoxin-II, two K49 PLA
2s isolated from
Bothrops jararacussu snake venom caused dose-dependent rat paw and skin oedema formation [
17]. PLA
2s from
Crotalus durissus cascavella,
Crotalus durissus collilineatus and
Crotalus durissus terrificus venoms increased the vascular permeability in the rat skin [
26]. However, the observed reduction of mast cell numbers by injection of compound 48/80 hardly affects the inflammatory potential of TM-N49: even a 10-fold reduction in detectable mast cells does not result in a significant reduction of extravasation. This implied that mast cell activation may not be the relevant mechanism for TM-N49 induced skin edema. It is difficult to exclude the possibility that TM-N49 bound to a receptor of PLA
2 and caused microvascular leakage, though the latter has not yet been identified. Recently, it was found that snake venom catalytically inactive PLA
2 homologue could binds to vascular endothelial growth factor receptor-2 via a C-terminal loop region, exhibiting mast cell unrelated mechanism for the increased the vascular permeability [
27].
In order to study the activation of mast cells by TM-N49, mast cell challenge experiments were performed. TM-N49 can activate human colon, lung and tonsil mast cells in a dose dependent manner. It was discovered further that TM-N49 induced mast cell degranulation is an energy consuming process and involved a G protein coupled receptor in mast cells [
28] as both metabolic inhibitors and pertussis toxin were capable of inhibiting TM-N49 elicited histamine release from mast cells. It was reported previously that PLA
2s were able to directly activate rat mast cells
in vitro [
17,
20,
29]. The action of D49 PLA
2s on rat mast cells was dependent on their catalytic activity [
16,
30], whereas the influence of K49 PLA
2s, which possess negligible catalytic activity, on rat mast cells was through a heparin-sensitive mechanism [
17]. Removal of Ca
2+ and Mg
2+ from challenge buffer greatly reduced the cell response to the stimulation of TM-N49 indicated that exogenous calcium and magnesium ions were crucial for mast cell degranulation induced by TM-N49. The increased serotonin level in mouse peritoneal lavage fluid supplied further evidence on TM-N49 causing mast cell degranulation in rats as mast cells are major source (if not the only source) of serotonin in mouse peritoneum. The most relevant mast cell population in regard to snake bites would be skin mast cells which differ in phenotype and activating stimuli from other mast cell populations. From those populations investigated, tonsil mast cells most closely resemble skin mast cells [
31]. Interestingly, as shown in figure
4, tonsil mast cells only responded to rather high concentrations of TM-N49, and the overall release (~7%) is very little compared to other venom components (e.g. atrahagin, induces 26% histamine release) [
23]. TM-N49 at 30 μg appears to be toxic, as the observed release is later shown to be Ca
2+-independent. This, again, indicates that mast cell activation may not be the relevant mechanism for TM-N49 induced inflammation.
With a mouse peritoneal model, it was found for the first time that snake venom PLA
2 could induce mast cell accumulation even at 10 min following injection. This appeared not a mast cell-specific effect since the relative number of mast cells in total nucleated cells was not significantly altered. The action of TM-N49 in induction of mast cell accumulation could be inhibited by co-injection of TM-N49 with ginkgolide B, cyproheptadine and terfenadine, suggesting that histamine, 5-HT and PAF may play a role in mast cell accumulation induced by TM-N49. Antibodies specific for CD18, CD11a and ICAM-1 blocked TM-N49 induced mast cell accumulation, while L-selectin specific antibody failed to do so. These observations suggested that LFA-1 (CD11a/CD18) and ICAM-1, but not L-selectin were involved in the TM-N49 induced mast cell migration process. Involvement of CD11a/CD18 and ICAM-1 in mast cell migration was reported previously with Saban et al. [
32,
33]. Since mast cells and endothelial cells express both LFA-1 and ICAM-1 [
32,
33], we believe that CD11/CD18 and ICAM-1 are important adhesion molecules for TM-N49 induced mast cell migration. The increase in mast cell numbers could be regarded as a beneficial host response since mast cells were shown to be able to detoxify venom components [
21].
Methods
Reagents
Protobothrops mucrosquamatus crude venom was obtained from the stock of the Kunming Institute of Zoology, the Chinese Academy of Sciences. SP-sephadex C-25, heparin sepharose (FF) and superdex 75 were from LKB Pharmacia (Uppsala, Sweden). The following compounds were purchased from Sigma (St. Louis, USA): egg phosphatidyl choline, Triton X-100, trifluoroacetic acid, honey bee venom phospholipase A2, platelet activating factor (PAF), cyproheptadine, ginkgolide B, heparin, collagenase (type I), hyaluronidase (type I), soybean trypsin inhibitor (SBTI), bovine serum albumin (BSA, fraction V), penicillin and streptomycin, calcium ionophore A23187, antimycin A, 2-deoxy-D-glucose, pertussis toxin. Quinacrine was from Calbiochem (San Diego, CA, USA). Reagents for sodium dodecyl- sulphate-polyacrylamine gel electrophoresis (SDS-PAGE) were from Bio-Rad Laboratories Inc (Hercules, USA). Coomassie Plus™ assay kit was from Pierce Chemical Co (Rockford, IL, USA). PolyATract system 1000 kit and Reverse transcription system kit were from Promega Biotech (Madison, WI, USA). Goat anti-human IgE (inactivated) was from Serotec (Kidlington, Oxford, UK). Fetal calf serum (FCS) and minimum essential medium (MEM) containing 25 mM N-2-hydroxylethylpiperazine-N'-2 -ethane sulphonic acid (HEPES) were from Gibco (Paisley, Renfrewshire, UK). Rat monoclonal antibodies, anti-mouse CD 11a [lymphocyte function-associated antigen 1(LFA-1) α chain], clone M17/4; anti-mouse CD 62L (L-selectin), clone MEL-14; anti-mouse CD18 (integrin β2 chain), clone M18/2; rat IgG2a isotype standard, clone R35-95; hamster anti-mouse CD54 [intercellular adhesion molecule 1 (ICAM-1)] monoclonal antibody, clone 3E2; hamster IgG1 isotype standard, clone A19-3 were from BD Biosciences Pharmigen (CA, USA). Hepes and all other chemicals were of analytical grade. BALB/c mice (20–25 g) and Wistar rat (180–220 g) were bred and reared under strict ethical conditions according to international recommendation.
Purification of TM-N49
TM-N49 was isolated from
Protobothrops mucrosquamatus crude venom following the procedures described previously [
6]. Briefly, the lyophilized venom (1.5 g) was dissolved in 20 ml of 50 mM sodium phosphate buffer (pH 5.8) and loaded on a SP-sephadex C-25 column equilibrated with the same buffer. The absorbed proteins were eluted with a linear gradient of NaCl (0–0.8 M) in 50 mM sodium phosphate (pH 5.8). The fractions in peak 9 were collected, lyophilized, and then dissolved in a running buffer containing 25 mM sodium phosphate (pH 5.8) and 0.15 M NaCl. Superdex 75 column was used to further isolate TM-N49 and the fractions 50–56 in the main protein peak was collected and loaded on a heparin-sepharose (FF) column equilibrated with a 25 mM sodium phosphate buffer (pH 5.8). The absorbed proteins were eluted with a linear gradient of NaCl (0–0.8 M) in 25 mM sodium phosphate (pH 5.8). The fractions 40–47 in the main protein peak eluted from heparin agarose were collected, lyophilized, dissolved in 0.1% trifluoroacetic acid (v/v) and loaded on a reverse-phase C
18 high performance liquid chromatography colomn (Symmetry 300™, 5 μM; Waters Corporation, Milford, Massachusetts, USA). The elution was performed with 0.1% trifluoroacetic acid and a gradient of 15–75% buffer B (containing 100% acetonitrile (v/v) and 0.1% trifluoroacetic acid) over 60 min at a flow rate of 1 ml/min. The fractions in main protein peak representing the purified TM-N49 were pooled, lyophilized and analyzed with mass spectrometer and stored at -20°C until use. Protein concentration was determined by a modified Bradford method [
34] using a Coomassie Plus™ assay kit with BSA as standard. The PLA
2 activity was routinely assayed by a titration method using egg yolk as substrate according to Ishimaru et al. [
35], and by a colorimetric assay using L-phosphatidylcholine as substrate according to de Ajaujo and Radvanyi [
36]. Honey bee PLA
2 was employed as positive control.
Induction of microvascular leakage
Microvascular leakage experiments were performed mainly following the procedures described previously by He et al. [
37]. Briefly, Wistar rats (180–220 g) were anaesthetized by intra-peritoneal injection of sodium pentobarbitone at a dose of 21 mg·kg
-1. After shaving the back, 1% (w/v) Evan's blue dye in normal saline was injected intravenously into the tail vein, and then various doses of TM-N49, bradykinin, histamine and PAF all at 5 μg, BSA at 5 μg and 50 μg or normal saline were intradermally injected, randomized sites placed 2–3 cm apart, 50 μl per injection point. Animals were kept warm, and at 20 min following the final injection, were killed by inhalation of carbon dioxide and the skin removed. Two perpendicular diameters were recorded for the blue area on the inside of the skin, and multiplied to give a measure of the relative size of the area of cutaneous edema.
For compound 48/80 pretreatment experiments, groups of rats (including 6 h, day 1, day 2, day 3, day 4 groups and their corresponding control groups) were treated with compound 48/80 to deplete mast cells, as described by Di Rosa et al. [
38]. Briefly, compound 48/80 (0.1% solution in 0.9% sterile saline) was injected intravenously at a dose of 0.6 mg·kg
-1, twice a day for 3 days, and doubled dose in day 4. At 5 h following the final injection, the test substances in the skin oedema experiments were administered as described above. The control groups received a normal saline injection. Both skins and peritoneal lavages of the testing rats were collected and analyzed.
To investigate potential mechanisms involved in TM-N49 induced microvascular leakage, several anti-inflammatory compounds including cyproheptadine (5 μg), a 5-HT
2/5-HT
1Cserotonin receptor antagonist and H
1 histamine receptor antagonist; terfenadine (5 μg), a selective H
1 histamine receptor antagonist; ginkgolide B (5 μg), a PAF receptor antagonist [
39] and quinacrine (5 μg), an inhibitor of PLA
2 [
40] were co-injected into the skin of rats with TM-N49 PLA
2 (5 μg), bradykinin (5 μg), histamine (5 μg), or normal saline respectively for 20 min.
Mast cell challenge and analysis of histamine release
Macroscopically normal lung and colon tissues were collected at bronchial or colon resection from patients with lung or colon cancer, respectively, and tonsil tissue was obtained at tonsillectomy. The procedure for mast cell dispersion was similar to that described previously [
23,
28]. Briefly, tissue was chopped finely with scissors into fragments of 0.5–2.0 mm
3, and incubated with 1.5 mg/ml collagenase and 0.75 mg/ml hyaluronidase in MEM containing 2% FCS (1 g lung/10 ml buffer) for 70 min at 37°C. Dispersed cells were separated from undigested tissue by filtration through nylon gauze (pore size 100 μm diameter), and were maintained in MEM (containing 10% FCS, 200 U/ml penicillin, 200 μg/ml streptomycin) on a roller overnight at room temperature. Mast cell numbers were determined by light microscopy after staining with alcine blue staining solution, and represented 2.5 to 5.3%, 4.2 to 5.7%, and 1.2 to 3.1% of nucleated cells in lung, colon or tonsil suspensions, respectively.
Prior to challenge with stimulus the cells were washed with HBSS (pH 7.4) without added calcium or magnesium (500 g, 10 min, 25°C), and then resuspended in HBSS with 1.8 mM CaCl2 and 0.5 mM MgCl2. Aliquots of 100 μl containing 4–6 × 103 mast cells were added to a 50 μl aliquot of purified TM-N49, heparin, or control secretagogue in complete HBSS and incubated for 15 min at 37°C. The reaction was terminated by the addition of 150 μl ice cold HBSS and the tubes centrifuged immediately (500 g, 10 min, 4°C). All experiments were performed in duplicate. For the measurement of total histamine concentration the suspension in some tubes was boiled for 6 min. Supernatants were stored at -20°C until histamine concentrations were determined. Where added, TM-N49 at 30 μg/ml was preincubated with heparin on ice for 10 min before being added to cells.
For the experiments with pertussis toxin, cells were incubated with 0.1 or 1.0 μg/ml pertussis toxin for 4 h at 37°C, and then washed with HBSS before adding stimulus. Similarly, for the experiments with metabolic inhibitors, cells were incubated with 2-deoxy-D-glucose (10 mM) and antimycin A (1 μM) for 40 min at 37°C before challenged with stimulus.
A glass fibre-based, fluorometric assay was employed to determine histamine levels in supernatants, as previously described [
23,
28]. Histamine bound to a glass-fibre matrix (RefLab, Copenhagen, Denmark) was detected by addition of o-phthaldialdehyde (OPT) and the color change measured using a spectrophotofluorometer (Perkin-Elmer LS 2, Denmark). Histamine release was expressed as a percentage of total cellular histamine levels, and corrected for the spontaneous release measured in tubes in which cells had been incubated with the HBSS diluent alone. For the time course study, data were presented as percentage of maximum histamine release, which equals (actual net histamine release/maximum net histamine release induced by the testing compound) × 100%.
Induction of mast cell accumulation
Various doses of TM-N49, BSA or normal saline were injected in 0.5 ml volumes into the peritoneum of male BALB/c mice, whose abdominal skin was swabbed with 70% ethanol, a group of 6 mice for each dose. This model was adapted from that described by Thomas et al. [
41], which complied with the European Community guidelines for use of experimental animals. At 10 min, 2 h, 6 h or 16 h following injection, animals were sacrificed by inhalation of carbon dioxide, and their peritoneal lavages were collected following a standardized procedure with 5 ml normal saline using heparinised tubes. After centrifugation at 500 g for 10 min at 4°C, supernatants were collected and stored at -40°C until use, and cells were resuspended in 1 ml of MEM. The total cell numbers were determined by enumerating them with an Improved Neubauer haemocytometer after being stained with 0.1% trypan blue. For the differential cell counting, cytocentrifuge preparations were made, air dried and stained with modified Wright's stain. Differential cell counts were performed for a minimum of 500 cells. The results were expressed as absolute numbers of each cell type per mouse peritoneum.
For the experiments investigating mast cell migration mechanisms, groups of mice were pretreated intravenously (tail vein injection) with monoclonal antibodies against the adhesion molecules L-selectin, CD11a, CD18 and ICAM-1 (all at a dose of 1 mg·kg
-1) [
42‐
44], respectively for 30 min before intra-peritoneal injection of 5 μg TM-N49. Control animals received an equivalent dose of the corresponding normal rat or hamster IgG isotype control. At 6 h following injection, the mice were sacrificed and their peritoneal lavages were processed as described above.
To investigate potential mechanisms involved in TM-N49 induced inflammatory cell accumulation, several anti-inflammatory compounds including cyproheptadine (2 mg·kg
-1) [
17], terfenadine (3 mg·kg
-1) [
45,
46], ginkgolide B (5 mg·kg
-1) [
39] and quinacrine (10 mg·kg
-1) [
40] were co-injected into the peritoneum of mice with TM-N49 (5 μg per mouse). Control animals received an injection of drug alone. At 6 h following injection, mice were sacrificed and their peritoneal lavages were processed as described above. Certain mice were pretreated with compound 48/80 for 4 days as described above before peritoneal injection of 5 μg of TM-N49.
Statistical analysis
Statistical analyses were performed using SPSS software (version 12.0). For in vitro experiments, because data were normally distributed, they are shown as the mean ± SEM. for the number of experiments indicated. Where analysis of variance indicated significant differences between groups, for the preplanned comparisons of interest, paired Student's t test was applied. For in vivo experiments, since the data were not normally distributed, they were presented as the median and range, for the numbers of animals indicated. Where Kruskal-Wallis analysis indicated significant differences between groups, for the preplanned comparisons of interest the Mann-Whitney U-test was employed. For all analyses, P < 0.05 was taken as significant.
Authors' contributions
JFW and SHH conceived and designed the study. JFW, XLW and YZM performed the experiments. JFW drafted the manuscript and SHH revised the manuscript. SHH supervised and coordinated the whole project. All authors have read and approved the final manuscript.