- Email: [email protected]
Influence of sphingomyelin and TNF-α release on lethality and local inflammatory reaction induced by Loxosceles gaucho spider venom in mice
Toxicon 42 (2003) 471–479 www.elsevier.com/locate/toxicon
Influence of sphingomyelin and TNF-a release on lethality and local inflammatory reaction induced by Loxosceles gaucho spider venom in mice M.O. Domingosa,*, K.C. Barbarob,1, W. Tynanc, J. Pennyd, D.J.M. Lewisa, R.R.C. Newe a St George’s Medical School, Division of Infectious Diseases, Cranmer Terrace, London SW 17 ORE, UK Laboratory of Immunopathology, Butantan Institute, Av. Vital Brazil 1500, CEP 05503-900, Sa˜o Paulo, Brazil c Department of Pharmacology, Oxford University, Oxford, UK d Department of Medicine, Gastroenterology Section, Manchester University, UK e Proxima Concepts Limited, c/o 2 Royal Street, London NW1 0TU, UK
b
Received 4 March 2003; accepted 4 July 2003
Abstract It is well known that Loxosceles venom induces local dermonecrosis in rabbits, guinea pigs and humans but not in mice, although, depending on the dose, Loxosceles venom can be lethal to mice. In this work we demonstrate that mice injected intradermally in the dorsal area of the back can survive a lethal dose of Loxosceles gaucho venom and also develop an inflammatory reaction (with infiltration of leukocytes shown by histological analysis) at the local injection site when the venom is co-administered with sphingomyelin. It was observed that more venom was retained for a longer period of time at the local injection site when venom was co-administered with sphingomyelin. The presence of exogenous sphingomyelin did not influence significantly the release of TNF-a induced by L. gaucho venom. These results suggest that the action of venom on sphingomyelin, producing ceramide phosphate, causes the development of an inflammatory reaction, which in turn traps the venom in the local area for a long period of time and does not allow it to disperse systemically in a dose sufficient to cause death. Our findings also indicate that the size and availability of the local sphingomyelin pool may be important in determining the outcome of Loxosceles envenoming in different mammalian species. q 2003 Elsevier Ltd. All rights reserved. Keywords: Loxosceles gaucho; Dermonecrosis; Spider venom; TNF-a; Sphingomyelin; Brown spider; Ceramide phosphate
1. Introduction In recent years, envenoming by Loxosceles has been recognised as a public health problem in Brazil, being considered the most serious form of araneism in the country. As a result, hundreds of bites have been reported officially * Corresponding author. Address: Especial Laboratory of Microbiology, Butantan Institute, Av. Vital Brazil 1500, Sa˜o Paulo, SP 05503-900, Brazil. Tel.: þ 55-11-3726-7222; fax: þ 5511-3726-1505. E-mail addresses: [email protected] (M.O. Domingos); [email protected] (K.C. Barbaro). 1 Tel.: þ55-11-37267222x2278; fax: þ 55-11-37261505.
(Ministry of Health, 1998). The most frequent finding after bites from Loxosceles gaucho is a dermonecrotic lesion (Ma´laque et al., 2002), but envenoming can also cause systemic and visceral effects, which may be lethal (Sezerino et al., 1998). The initial process of envenoming is painless. Mild to intense pain starts only after 2 –6 h and is followed by erythema, itching, swelling and tenderness. Three days to 1 week later a very severe scar may form and drop off leaving an ulcer, which after healing within 1 – 2 months, could require plastic surgery to repair the tissue damage. A variety of treatments for loxoscelism have been advocated and implemented (surgical excision, systemic corticosteroids, anti-histamines, dapsone, antivenom therapy) but their
0041-0101/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0041-0101(03)00200-9
472
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
efficacy is uncertain (White et al., 1995). In Brazil, therapies involving specific antiserum and corticoid drugs are used for treatment of Loxosceles spider bite (Ministry of Health, 1998). Dermonecrotic and lethal activities are characteristics common to the venoms of all Loxosceles spiders, and in all the species studied the active components have a similar molecular mass of between 32 and 35 kDa (Futrell, 1992; Barbaro et al., 1996). An interesting feature of this venom is its ability to induce dermonecrosis in rabbits, guinea pigs and humans but not in mice, although, depending on the dose, Loxosceles venom can be lethal to mice. The reason for this discrepancy between species is not clearly understood. The precise pathophysiological action of Loxosceles venom has not been fully elucidated. The cutaneous and visceral effects are probably due to a multifactorial process involving direct tissue damage, secondary vascular injury and release of phamacological agents by polymorphonuclear cells (Futrell, 1992). In addition, it has been reported that, after gel filtration, the fraction of venom responsible for necrotic skin lesions, haemolysis and platelet aggregation has sphingomyelinase D activity (Futrell, 1992; Tambourgi et al., 1998). Loxosceles venom does not affect neutrophils directly, but is completely dependent on the victim’s neutrophils and complement for its effects to be manifested (Patel et al., 1994). This suggests that it probably activates other pathways, which indirectly culminate in complement activation, neutrophil migration and accumulation in the endothelial wall. Ceramides, the breakdown products of sphingomyelin, are known to be powerful regulators of various biological processes (Ballou et al., 1996). Ceramides are also important cellular regulators of TNF-a (Hannun, 1994), which in turn is a potent regulator of neutrophil chemotaxis, adhesion, priming, phagocytosis, inflammatory mediator release and superoxide generation (Ballou et al., 1996). TNF-a also enhances the adhesiveness of neutrophils for activated endothelium and extracellular matrix proteins, such as fibrinogen, fibronectin, and even serum-coated plastic (Gamble et al., 1992). In addition, treatment of monocytes/macrophages with TNF-a causes secretion of various cytokines such as IL-1, IL-6 and IL-8 (Fiers, 1995). Since L. gaucho venom contains an enzyme (a sphingomyelinase D) capable of cleaving sphingomyelin into choline and ceramide phosphate, this study investigated whether the presence of exogenous sphingomyelin and/or its breakdown product ceramide phosphate might be responsible for some of the effects of the venom in vivo. The mouse was chosen for this study since it appears to be less susceptible to the local effects of venom than other species. It is well documented in the literature that the structure and content of sphingomyelin in membranes varies among species. Coleman et al. (1980) found that the differing behaviour of red cells from different species correlated with their differing sphingomyelin content, while Bettger et al. (1998) found that the sphingomyelin of rats and mice
possesses a fatty acid chain that is not present in humans. In light of this, it is possible that such variations between species in the amount and type of sphingomyelin present could account for the observation that Loxosceles venom induces necrosis in rabbits and humans but not in mice (Futrell, 1992; Barbaro et al., 1996). It was hypothesised that this might be precisely because mice lack the appropriate pool of sphingomyelin in sufficient quantity (He et al., 2001), thus permitting studies to elucidate the mechanism by administration of exogenous sphingomyelin. Finally, since TNF-a is a potent chemoattractant for leukocytes to the sites of inflammation, experiments in vivo were conducted to examine whether co-administration of exogenous sphingomyelin would influence TNF-a release induced by L. gaucho venom at the local injection site.
2. Material and methods 2.1. Animals BALB/c female mice (6– 8 weeks old) and BALB/c male mice (18 – 20 g) were obtained from Harlam, UK and Butantan Institute Biological Research Facilities, respectively. 2.2. Venom Loxosceles venom was collected from the fangs of L. gaucho by electrical stimulation as described by Barbaro et al. (1992). Venom from 600 spiders was collected, pooled and lyophilised. Phosphate-buffered saline (6 ml) was added to the lyophilised venom, after which the solution was filtered (0.2 mm), aliquotted and stored at 220 8C until required. The protein content of the venom was determined using the BIO-RAD Protein assay (catalogue number: 5000116) using BSA as protein standard. 2.3. Sphingomyelin and ceramide phosphate Hen egg sphingomyelin was purchased from Sigma (St Louis, MO, USA) and D -erythro-sphingosine-1-phosphate, N-octanoyl (referred to below as ceramide phosphate) was purchased from Calbiochem (San Diego, CA, USA). 2.4. Dispersion of ceramide phosphate in saline One milligram of ceramide phosphate was weighed into a glass vial and dissolved in 1 ml of methanol/dichloromethane (1:2). The solution was dried down under a stream of nitrogen and 1 ml of distilled water was added to the vial. The vial was then capped, sealed with parafilm and sonicated at maximum amplitude in a Model 575D Truesweep Ultrasonic bath (Crest Ultrasonics Corporation,
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
New Jersey, USA) at 40 8C until clarity was achieved. The ceramide phosphate dispersion obtained had a concentration of 1 mg/ml in saline. 2.5. Dispersion of ceramide phosphate in liposomes Five hundred microlitres of dichloromethane/methanol (2:1) containing 10 mg of soya phosphatidyl choline (Lipoid) was added to 1 mg of dry ceramide phosphate (Calbiochem) and the resulting solution was dried down under a stream of nitrogen. Saline (0.5 ml) was then added to the dry residue. The sample was sonicated as described above and had the concentration adjusted to 1 mg/ml of ceramide phosphate in saline. A control preparation of phospholipid liposomes was prepared as described here, except that ceramide phosphate was absent.
473
2.9. Determination of LD50 of L. gaucho venom after subcutaneous injection in the footpad BALB/c male mice were divided in groups of three mice each and injected subcutaneously in the footpad with different concentrations of L. gaucho venom in PBS ranging from 0.06 to 0.36 mg/kg (w/w). The LD50 was determined 48 h after injection by probit analysis (Finney, 1971). LD50 ¼ 0.15 mg/kg of body weight (w/w) (confidence limits 95%, 0.1 – 0.2 mg/kg (w/w)) in 48 h. This experiment was performed twice. 2.10. Effect of sphingomyelin on mortality due to venom To determine whether sphingomyelin had any protective effect against mortality due to L. gaucho venom, the following experiments were performed.
2.6. Dispersion of sphingomyelin in saline One hundred milligrams of sphingomyelin was weighed into a 7 ml glass screw-capped vial and dissolved in either 1 ml or 400 ml of methanol/dichloromethane (1:2). Aliquots (200 ml) of 100 mg/ml sphingomyelin solution were transferred to fresh vials. The 200 ml aliquots containing 20 mg of sphingomyelin or the 100 mg/400 ml solution were dried down under a stream of nitrogen. Two millilitre of saline and 720 ml of saline were added, respectively, to the vials containing 20 and 100 mg of sphingomyelin. The samples were sonicated as described before. The sphingomyelin dispersion obtained was at a concentration of either 10 or 139 mg/ml in saline. 2.7. TNF-a detection An ELISA Kit for murine TNF-a from R & D Systems (Minneapolis, MN, USA) was used to measure TNF-a levels in supernatants of tissue extracts from the local injection site. The tissue extracts were prepared as described in Section 2.13.
2.10.1. Experimental design 1—intradermal injection Twenty eight mice were divided into four groups of seven mice each and injected (0.1 ml) i.d. in the dorsal area of the back with 0.5 mg/kg of the body weight (w/w) of L. gaucho venom either alone or with 1 mg of sphingomyelin were left for 1 h at 37 8C. The controls received either 1 mg/0.1 ml of sphingomyelin alone or 0.1 ml of PBS. Death and other symptoms of envenoming were recorded 24 h after injection. Animals that showed severe symptoms of envenoming prior to this time were sacrificed immediately. 2.10.2. Experimental design 2—subcutaneous injection Twelve mice were divided into four groups of three mice each and injected (30 ml) in the footpad with a LD50 dose of L. gaucho venom alone or with venom incubated for 1 h at 37 8C with 1 mg of sphingomyelin. The controls received either 1 mg of sphingomyelin alone or PBS. Death and other symptoms of envenoming were recorded 24 and 48 h after injection. This experiment was performed twice. Animals that showed severe symptoms of envenoming prior to this time were sacrificed immediately.
2.8. Venom detection 2.11. Induction of wound after intradermal injection Venom levels were determined by ELISA using the method described by Theakston et al. (1977) and modified by Ho et al. (1986). Briefly, 96-well microplates were coated with horse anti-arachnidic serum produced by the Butantan Institute. A standard curve of L. gaucho venom dilutions was used to determine the amount of venom in the test samples. Rabbit anti-L. gaucho venom IgG was used as secondary antibody and goat anti-rabbit IgG labelled with peroxidase was used as conjugate. After washing, substrate solution (ortho-phenylenediamine and H2O2) was added to the wells. The colour development was stopped by addition of 50 ml of diluted hydrochloric acid and the intensity of the colour was measured at 492 nm. The lower limit of sensitivity of the assay was 2 ng/ml.
To investigate whether exogenous sphingomyelin coadministered with L. gaucho venom could induce a wound or a non-specific inflammatory reaction at the local injection site after intradermal injection, mice were treated as described in Section 2.10 (experimental design 1) and examined daily. In addition, two other groups were also included in the experiment to determine the effect of ceramide phosphate (the product of cleavage of sphingomyelin by L. gaucho D -sphingomyelinase) administered locally. These mice were injected (0.1 ml) i.d. at the dorsal area with either 250 mg of ceramide phosphate in aqueous solution (seven mice per group) or 250 mg of ceramide phosphate in liposomes (six mice per group).
474
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
alone or with venom incubated for 1 h at 37 8C with 1 mg of sphingomyelin in 30 ml of PBS. Control groups received either sphingomyelin alone or PBS. At different times after injection, the mice were anaesthetised, bled by axillary plexus scission and sacrificed. Blood samples were placed into eppendorf tubes and centrifuged at 8000g for 5 min in an eppendorf centrifuge. Serum was then collected for assay of venom and TNF-a. Skin samples from the local injection site of mice injected intradermally in the dorsal area (1.5 cm in diameter) were excised after sacrifice and homogenized in 0.2 ml of PBS at room temperature. After homogenisation tissue extracts were centrifuged for 5 min in an eppendorf centrifuge and the supernatants were collected for assay of venom and TNF-a.
Examination for any kind of visible reaction was performed for up to six days post injection. Samples of the skin from the injection site of animals intradermally injected in the dorsal area of the back that died or showed any symptoms of envenoming in 24 h were taken and photographed before fixation with 1% of formaldehyde. The remaining animals were sacrificed either three days (ceramide phosphate and ceramide phosphate in liposome groups) or six days (Loxosceles venom þ sphingomyelin and sphingomyelin only groups) after injection, and, photographs of skin from the injection sites were taken prior to fixation with 1% formaldehyde. Following fixation, skin samples were stained with haematoxilin-eosin and viewed under a light microscope. 2.12. Oedema measurement after footpad injection BALB/c female mice were injected (30 ml) in the footpad with a LD50 dose (0.15 mg/kg of body weight (w/w)) of either L. gaucho venom alone, L. gaucho venom incubated with 1 mg of sphingomyelin, 1 mg of sphingomyelin alone or PBS. Oedema formation was measured with a caliper ruler in length and the landmark was stated as a point in the middle of the footpad. The results express the difference in mm (thickness) in the same animal between treated paw (left footpad) and untreated paw (right footpad). Oedema was measured at different times after injection.
2.13.2. Experimental design 2—subcutaneous injection Mice were injected subcutaneously in the footpad with a LD50 dose (0.15 mg/kg of the body weight (w/w) in 48 h) of L. gaucho venom alone or with venom incubated for 1 h at 37 8C with 1 mg of sphingomyelin in 30 ml PBS. Control groups received either sphingomyelin alone or PBS. At different times after injection, the mice were anaesthetised, bled by axillary plexus scission, sacrificed and footpad samples were collected for subsequent analysis. Blood samples were placed into eppendorf tubes and centrifuged (8000g) for 5 min in an eppendorf centrifuge. Serum was then collected for assay of venom. The footpad samples were processed as follows: after sacrifice, legs of mice treated subcutaneously (right footpad) or untreated (left footpad) were cut-off at the joint of the ankle and homogenized in eppendorf tubes with 0.2 ml of PBS at room temperature. The samples were then centrifuged for 5 min in an eppendorf centrifuge (8000g) and the supernatants were collected for assay of venom and TNF-a by ELISA.
2.13. Detection of venom and TNF-a at the local injection site (skin and footpad) To investigate the rate of venom clearance and TNF-a release at the injection site the following experiments were performed. 2.13.1. Experimental design 1—intradermal injection Mice were injected under light anaesthesia in shaved dorsal skin i.d. with 0.5 mg/kg (w/w) of L. gaucho venom
Table 1 Lethal and dermonecrotic actions of L. gaucho venom in the presence and absence of exogenous sphingomyelin Treatment
Parameters No. of survivors
Wound development
Clinical symptoms
L. gaucho venom 10 mga
0/7
0/7
L. gaucho venom 10 mg plus sphingomyelinb Ceramide phosphatec Ceramide phosphate plus liposomesc Sphingomyelinb PBSb
7/7 7/7 6/6 7/7 7/7
7/7 0/7 5/6 0/7 0/7
Hypothermia, prostration and paralysis of the hind limbs None None None None None
a b c
Parameters measured 24 h after i.d. injection. Parameters measured six days after i.d. injection. Parameters measured three days after i.d. injection.
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
2.14. Statistical analysis Results are presented as the means ^ s.e.m. For statistical analysis, two-way ANOVA followed by the Tukey test was performed using the SigmaStat 3.0
475
software. Values with p , 0:05 were considered statistically significant. Non-parametric tests (Kruskal-Wallis and Mann-Whitney) were employed for analyzing venom clearance at the local injection site and in serum.
Fig. 1. Mice were injected i.d. with either 0.5 mg/kg (w/w) of L. gaucho venom plus sphingomyelin (A—outside of skin, B—inner side of skin), 1.25 mg/kg of the body weight (w/w) of ceramide phosphate plus liposome (C—outside of the skin, D—inner side of the skin) or 0.5 mg/kg (w/w) of L. gaucho venom in 0.1 ml PBS (E—outside of the skin, F—inner side of the skin). The animals were sacrificed either 24 h (L. gaucho venom in PBS), three days (ceramide phosphate plus liposome group) or six days after injection. The black marks appearing in E and F were made by a pen marker to delineate the area of injection.
476
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
3. Results 3.1. Effect of sphingomyelin on toxic actions due to venom The results shown in Table 1 demonstrate that all animals injected i.d. at the dorsal area with venom either died or developed strong symptoms of envenoming such as hypothermia, prostration and paralysis of the hind limbs. Mice injected intradermally in the footpad also develop similar symptoms of envenoming (data not shown). In contrast, all animals from the groups injected with venom co-administered with sphingomyelin survived, and developed at the site of injection a wound on the outer side of the skin and an abscess-like reaction on the inner side (Fig. 1A and B). The liposomal ceramide phosphate group also developed, at the site of injection, a mild wound on the outer side of the skin and a slight haemorrhagic mark on the inner side (Fig. 1C and D). The blood vessels of mice injected with venom alone were more evident but there was no visible sign of lesion (Fig. 1E and F). None of the other groups developed any kind of visible reaction at the injection site. Histological analysis of the skin from mice injected i.d. with venom plus sphingomyelin showed an infiltration of leukocytes at the injection site (data not shown). 3.2. Determination of oedema reaction In order to determine the influence of exogenous sphingomyelin on development of oedema at the local injection site, mice were injected subcutaneously in the footpad with either venom alone, venom co-administered with sphingomyelin, sphingomyelin alone or PBS. The results obtained in this experiment showed that 2 and 6 h after injection, oedema was evident in the groups treated with venom alone or venom co-administered with sphingomyelin, and was significantly more marked ðp , 0:05Þ than in the PBS group (Fig. 2). No statistical difference in size of oedema swelling could be observed 2 and 6 h after injection between the groups treated with venom alone or venom coadministered with sphingomyelin. Twenty four hours after injection, swelling due to oedema was only observed in the group treated with venom co-administered with sphingomyelin ðp , 0:05Þ (Fig. 2). Moreover, six days after injection, oedema was observed only in the group injected with venom co-administered with sphingomyelin (data not shown).
Fig. 2. Influence of exogenous sphingomyelin (Sm) on oedema development. BALB/c female mice were injected (30 ml) in the footpad with a LD50 dose (0.15 mg/kg of body weight (w/w)) of either L. gaucho venom alone (Venom), L. gaucho venom with incubated 1 mg of sphingomyelin (Venom þ Sm), 1 mg of sphingomyelin alone (Sm) or PBS. Oedema reactions were measured with a caliper ruler at different times after injection. The asterisk represents statistically significant differences ðp , 0:05Þ between venom alone, venom plus sphingomyelin and sphingomyelin groups with control (PBS) group.
sphingomyelin. Samples of tissue from the local injection site and blood were collected at different times after injection, and assayed for venom by ELISA. The results showed that in both routes of injection, a higher level of L. gaucho venom was detected at the local injection site when venom was co-administered with sphingomyelin compared with the groups that received venom alone (Fig. 3A and B). Although, 6 h after venom injection, the level of venom at the local injection site decreased markedly in both cases, the level of venom was still two times higher in the group that received venom coadministered with sphingomyelin (Fig. 3A and B). In addition, even 24 h after injection, the level of venom detected locally in the group injected with venom coadministered with sphingomyelin in the footpad was significantly higher than the level obtained in groups that received venom alone. However, in serum, regardless of the route of injection, there was no difference in the levels of venom between the groups treated with venom alone or venom co-administered with sphingomyelin (Fig. 3C and D). 3.4. TNF-a detection at the local injection site
3.3. Determination of venom clearance at the local injection site and serum In order to determine whether exogenous sphingomyelin would influence the kinetics of venom in serum and at the local injection site, mice were injected either intradermally (dorsal skin) or subcutaneously (in the footpad) with L. gaucho venom either alone or co-administered with
In order to determine the influence of exogenous sphingomyelin on TNF-a release at the local injection site, mice were injected intradermally (dorsal skin of the back) or subcutaneously (footpad) with L. gaucho venom either alone or co-administered with sphingomyelin. Two hours after injection, samples of the local injection site were collected and assayed for TNF-a by ELISA after previous
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
477
Fig. 3. Detection of L. gaucho venom at the local injection site and serum after intradermal and subcutaneous injections. BALB/c male mice were injected with L. gaucho venom (alone or incubated with 1 mg of sphingomyelin (Sm) for 1 h at 37 8C) either subcutaneously in the footpad (0.15 mg/kg of body weight (w/w) diluted in 30 ml) or intradermally in the dorsal shaved skin of the back (0.5 mg/kg of body weight (w/w) diluted in 0.1 ml). Samples of the local injection site (A—subcutaneously injected mice; B—intradermally injected mice) and blood (C— subcutaneously injected mice; D—intradermally injected mice) were collected at different times after injection for venom detection by ELISA. # Statistical difference ðp , 0:05Þ between experimental times in the same experimental group. * Statistical difference ðp , 0:05Þ between the groups Venom plus Sm and Venom alone.
observation that local levels of TNF-a drop drastically 6 h after injection. The results show that, at the time-points measured, regardless of the route of injection, there was no significant difference between the levels of TNF-a detected in any of the test groups investigated, compared with controls that received either sphingomyelin or PBS alone (data not shown).
4. Discussion Sphingolipids such as sphingomyelin are most often localized in the outer leaflet of the plasma membrane and are an integral component of signalling pathways used by many types of cells (Boucher et al., 1995; Gulbins et al., 1995). These signals instruct the cells to perform a variety of functions such as cytokine production, differentiation,
proliferation or apoptosis. One of the most important membrane lipids capable of giving rise to metabolically active signal transduction is sphingomyelin. Ceramide, the immediate product of sphingomyelin hydrolysis, and its further metabolites such as sphingosine and sphingosine-1phosphate, play an important role in a variety of signal transduction pathways. It has been demonstrated that the lytic and necrotic action of Loxosceles venom is due to a fraction that has sphingomyelinase D activity (Tambourgi et al., 1998). Natural mammalian sphingomyelinase on the other hand cleaves sphingomyelin to remove not choline, but phosphocholine, giving ceramide as a final product. Although ceramide phosphate can occur naturally, it is generated endogenously not from sphingomyelin, but from ceramide by the action of a calcium-dependent ceramide kinase found predominately in the plasma membrane (Shalini and Kolesnick, 1993; Hinkovska-Galcheva et al., 1998). Ceramide phosphate from either source can be converted back into ceramide by the action of ceramide-1P-phosphatase (Boudker
478
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
and Futerman, 1993; Shinghal et al., 1993). It is conceivable, therefore, that, ceramide phosphate generated exogenously by sphingomyelinase D action could have a profound effect on cells by participating directly in cell signalling pathways. The results presented here show that L. gaucho venom administered i.d. in the dorsal area in the presence of exogenous sphingomyelin generated both a wound and an inflammatory reaction at the injection site and that it also protected adult mice against death. This is the first time it has been observed that the presence of exogenous sphingomyelin can protect mice against a lethal dose of Loxosceles venom and in addition is capable of inducing the venom to generate in mice an inflammatory reaction at the local injection site. L. gaucho venom alone, at the same dose level, induced death in all animals tested without causing any reaction at the injection site. Subcutaneous injection of a LD50 dose of L. gaucho venom alone in the footpad induced an oedema reaction at the local injection site that could be observed 2 and 6 h after injection, but which diminished after 24 h. In contrast, the group treated in the footpad with venom co-administered with sphingomyelin developed an oedema reaction that could be observed up to six days after injection (data not shown). In addition, only two out of six mice survived the LD50 dose injection of L. gaucho venom alone in the footpad (data not shown), while animals all survived without showing any visible symptoms of envenoming when the same dose was administered with sphingomyelin. One hypothesis for the observations described above is that exogenous sphingomyelin helps to retain the venom in the local area, preventing it from dispersing systemically, where it can cause death of the animal. This trapping of venom results in a local inflammatory response that culminates in an internal abscess-like reaction in the skin, and wound formation. This hypothesis is supported by the results obtained in this study, which show that a significantly higher level of venom was retained at the local injection site when venom is co-administered with sphingomyelin either by intradermal injection in the skin or by subcutaneous injection in the footpad. The results obtained by Gomez et al. (2001) also corroborate this hypothesis since they have shown that the extent of dermal inflammation in rabbits is directly correlated with Loxosceles venom diffusion. Furthermore, the results obtained with ceramide phosphate (a small rapidly diffusing molecule) also showed that intradermal injection in saline did not induce any visible reaction at the site of injection nor did it cause death of animals. However, when ceramide phosphate was administered intradermally in liposomal form (to prevent its rapid dispersal from the site), a mild wound was produced on the outer side of the skin, and a haemorrhagic lesion was visible on the inner surface. Likewise, histological observation also showed infiltration of leukocytes in the muscle layer. It is possible that due its small size and high water solubility, ceramide phosphate in saline was dispersed quickly from
the local injection site. However, its incorporation in liposomes may have helped its retention at the site of injection thus inducing a mild inflammatory response. The observation that co-administration of venom with sphingomyelin reduces its systemic toxicity in mice, yet does not appear to reduce the level of venom reaching the bloodstream is surprising, and is worthy of further study. Although work elsewhere (Domingos et al., 2003) has shown that action of L. gaucho venom on sphingomyelin induces TNF-a release by macrophages in vitro, this study demonstrated that no increase in TNF-a levels by L. gaucho is detectable at the local injection site, even in the presence of exogenous sphingomyelin, compared with sphingomyelin alone.
Acknowledgements The authors are also very grateful to Dr Marcelo Larami Santoro for the statistical analysis and Marı´lia Brinati Malta and Sabrina Soares for their excellent technical assistance.
References Ballou, L.R., Laulederkind, S.J.F., Roloniec, E.F., Raghow, R., 1996. Ceramide signalling and the immune response. Biochim. Biophys. Acta 1301, 273–287. Barbaro, K.C., Cardoso, J.L.C., Eickstedt, V.R.D., Mota, I., 1992. Dermonecrotic and lethal components of Loxosceles gaucho spider venom. Toxicon 30, 331–338. Barbaro, K.C., Sousa, M.V., Morhy, L., Eickstedt, V.R.D., Mota, I., 1996. Compared chemical properties of dermonecrotic and lethal toxins from spiders of the genus Loxosceles (Araneae). J. Prot. Chem. 15, 337–343. Bettger, W.J., Blackader, C.B., McCorquodale, M.L., Ewing, R., 1998. The occurrence of iso 24:0 (22 Methyltricosanoic acid) fatty acid in sphingomyelin of rat tissues. Comp. Biochem. Physiol. B. Biochem. Mol. Biol. 119, 229– 304. Boucher, L.M., Wiegmann, K., Futterer, A., Pfeffer, K., Machleidt, T., Schu¨tzer, S., Mak, T.W., Kro¨nke, M., 1995. CD28 signals through acidic sphingomyelinase. J. Exp. Med. 181, 2059–2068. Boudker, O., Futerman, A.H., 1993. Detection and characterization of ceramide-1-phosphate activity in rat liver plasma membrane. J. Biol. Chem. 268, 22150–22155. Coleman, R., Lowe, P.J., Billington, D., 1980. Membrane lipid composition and susceptibility to bile salt damage. Biochim. Biophys. Acta 599, 294–300. Domingos, M.O., Barbaro, K.C., Tynan, W., Penny, J., Lewis, D.J.M., New, R.R.C., 2003. Effect of Loxosceles gaucho venom on cell morphology and behaviour in vitro in the presence and absence of sphingomyelin. Toxicon, in press, Manuscript No. AH/2003/000052. Fiers, W., 1995. Biologic therapy with TNF: preclinical studies. In: DeVita, V.T., Hellman, S., Rosenberg, S.A. (Eds.), Biologic Therapy of Cancer, Second ed., B. Lippincott Co., Philadelphia, Chapter 12.
M.O. Domingos et al. / Toxicon 42 (2003) 471–479 Finney, D.J., 1971. Probit Analysis, Third ed., Cambridge University Press, Cambridge. Futrell, J.M., 1992. Loxoscelism. Am. J. Med. Sci. 304, 261 –267. Gamble, J.R., Smith, W.B., Vadas, M.A., 1992. TNF modulation of endothelial and neutrophil adhesion. In: Beutler, B., (Ed.), Tumor Necrosis Factor. The Molecules and their Emerging Role in Medicine, Raven Press, New York, pp. 65–86. Gomez, H.F., Dian, M., Greefield, M.D., Miller, M.J., Jeffrey, S., Warren, M.D., 2001. Direct correlation between diffusion of Loxosceles reclusa venom and extent of dermal Inflammation. Acad. Emerg. Med. 8, 309–314. Gulbins, E., Bissonnette, R., Mahboubi, A., Martin, S., Nishioka, W., Brunner, T., Baier, G., Bier-Bitterlich, G., Byrd, C., Lang, F., 1995. FAS-induced apoptosis is mediated via a ceramideinitiated RAS signalling pathway. Immunity 2, 341 –351. Hannun, Y.A., 1994. The sphingomyelin cycle and the second messenger function of ceramide. J. Biol. Chem. 269, 3125–3128. He, X., Chen, F., Gatt, S., Shuchman, E.H., 2001. An enzymatic assay for quantifying sphingomyelin in tissues and plasma from humans and mice with Niemann–Pick disease. Anal. Biochem. 293, 204 –211. Hinkovska-Galcheva, V.T.Z., Boxer, L.A., Mansfield, P.J., Harsh, D., Blackwood, A., Shayman, J.A., 1998. The formation of ceramide-1-phosphate during neutrophil phagocytosis and its role in liposome fusion. J. Biol. Chem. 273, 33203–33209. Ho, A., Warrell, M.J., Warrell, D.A., Bidwell, D., Voller, A., 1986. A critical reappraisal of the use of enzyme-linked immunosorbent assays in the study of snake bite. Toxicon 24, 211–221. Ma´laque, C.M., Castro-Valencia, J.E., Cardoso, J.L., Franc¸a, F.O.S., Barbaro, K.C., Fan, H.W., 2002. Clinical and epidemiological features of definitive and presumed loxoscelism in Sa˜o Paulo, Brazil. Rev. Inst. Med. Trop. Sa˜o Paulo 44, 139–143.
479
Ministry of Health, 1998. In: Manual de diagno´stico e tratamento de acidentes por animais pec¸onhentos. Fundac¸a˜o Nacional de Sau´de/Coordenac¸a˜o de Controle de Zoonoses e Animais Pec¸onhentos, CENEPI, Brası´lia,. Patel, K.D., Modur, V., Zimmerman, G.A., Prescott, S.M., McIntyre, T.M., 1994. The necrotic venom of brown recluse spider induces dysregulated endothelial cell-dependent neutrophil activation. J. Clin. Invest. 94, 631–642. Sezerino, U.M., Zannin, M., Coelho, L.K., Gonc¸alves, J. Jr., Grando, M., Mattosinho, S.G., Cardoso, J.L.C., Eickstedt, V.R.D., Franc¸a, F.O.S., Barbaro, K.C., Fan, H.W., 1998. A clinical and epidemiological study of Loxosceles spider envenoming in Santa Catarina, Brazil. Trans. R. Soc. Trop. Med. Hyg. 92, 546– 548. Shalini, M., Kolesnick, R., 1993. Ceramide: A novel second messenger. Adv. Lipid Res. 25, 65–90. Shinghal, R., Scheller, R.H., Bajjalieh, S.M., 1993. Ceramide-1phosphate phosphatase activity in brain. J. Neurochem. 61, 2279–2285. Tambourgi, D.V., Magnoli, F.C., van den Berg, C.W., Morgan, B.P., Araujo, P.S., Alves, E.W., Dias da Silva, W., 1998. Sphingomyelinases in the venom of the spider Loxosceles intermedia are responsible for both dermonecrosis and complement-dependent hemolysis. Biochem. Biophys. Res. Commun. 251, 366–373. Theakston, R.D.G., Lloyd-Jones, M.J., Reid, H.A., 1977. MicroElisa for detecting and assaying snake venom and venom antibody. Lancet 24, 639–641. White, J., Cardoso, J.L., Fan, H.W., 1995. Clinical toxicology of spider bites. In: Meier, J., White, J. (Eds.), Handbook of Clinical Toxicology of Animal Venoms and Poisons, CRC Press, Boca Raton, pp. 261–329.
Influence of sphingomyelin and TNF-a release on lethality and local inflammatory reaction induced by Loxosceles gaucho spider venom in mice M.O. Domingosa,*, K.C. Barbarob,1, W. Tynanc, J. Pennyd, D.J.M. Lewisa, R.R.C. Newe a St George’s Medical School, Division of Infectious Diseases, Cranmer Terrace, London SW 17 ORE, UK Laboratory of Immunopathology, Butantan Institute, Av. Vital Brazil 1500, CEP 05503-900, Sa˜o Paulo, Brazil c Department of Pharmacology, Oxford University, Oxford, UK d Department of Medicine, Gastroenterology Section, Manchester University, UK e Proxima Concepts Limited, c/o 2 Royal Street, London NW1 0TU, UK
b
Received 4 March 2003; accepted 4 July 2003
Abstract It is well known that Loxosceles venom induces local dermonecrosis in rabbits, guinea pigs and humans but not in mice, although, depending on the dose, Loxosceles venom can be lethal to mice. In this work we demonstrate that mice injected intradermally in the dorsal area of the back can survive a lethal dose of Loxosceles gaucho venom and also develop an inflammatory reaction (with infiltration of leukocytes shown by histological analysis) at the local injection site when the venom is co-administered with sphingomyelin. It was observed that more venom was retained for a longer period of time at the local injection site when venom was co-administered with sphingomyelin. The presence of exogenous sphingomyelin did not influence significantly the release of TNF-a induced by L. gaucho venom. These results suggest that the action of venom on sphingomyelin, producing ceramide phosphate, causes the development of an inflammatory reaction, which in turn traps the venom in the local area for a long period of time and does not allow it to disperse systemically in a dose sufficient to cause death. Our findings also indicate that the size and availability of the local sphingomyelin pool may be important in determining the outcome of Loxosceles envenoming in different mammalian species. q 2003 Elsevier Ltd. All rights reserved. Keywords: Loxosceles gaucho; Dermonecrosis; Spider venom; TNF-a; Sphingomyelin; Brown spider; Ceramide phosphate
1. Introduction In recent years, envenoming by Loxosceles has been recognised as a public health problem in Brazil, being considered the most serious form of araneism in the country. As a result, hundreds of bites have been reported officially * Corresponding author. Address: Especial Laboratory of Microbiology, Butantan Institute, Av. Vital Brazil 1500, Sa˜o Paulo, SP 05503-900, Brazil. Tel.: þ 55-11-3726-7222; fax: þ 5511-3726-1505. E-mail addresses: [email protected] (M.O. Domingos); [email protected] (K.C. Barbaro). 1 Tel.: þ55-11-37267222x2278; fax: þ 55-11-37261505.
(Ministry of Health, 1998). The most frequent finding after bites from Loxosceles gaucho is a dermonecrotic lesion (Ma´laque et al., 2002), but envenoming can also cause systemic and visceral effects, which may be lethal (Sezerino et al., 1998). The initial process of envenoming is painless. Mild to intense pain starts only after 2 –6 h and is followed by erythema, itching, swelling and tenderness. Three days to 1 week later a very severe scar may form and drop off leaving an ulcer, which after healing within 1 – 2 months, could require plastic surgery to repair the tissue damage. A variety of treatments for loxoscelism have been advocated and implemented (surgical excision, systemic corticosteroids, anti-histamines, dapsone, antivenom therapy) but their
0041-0101/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0041-0101(03)00200-9
472
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
efficacy is uncertain (White et al., 1995). In Brazil, therapies involving specific antiserum and corticoid drugs are used for treatment of Loxosceles spider bite (Ministry of Health, 1998). Dermonecrotic and lethal activities are characteristics common to the venoms of all Loxosceles spiders, and in all the species studied the active components have a similar molecular mass of between 32 and 35 kDa (Futrell, 1992; Barbaro et al., 1996). An interesting feature of this venom is its ability to induce dermonecrosis in rabbits, guinea pigs and humans but not in mice, although, depending on the dose, Loxosceles venom can be lethal to mice. The reason for this discrepancy between species is not clearly understood. The precise pathophysiological action of Loxosceles venom has not been fully elucidated. The cutaneous and visceral effects are probably due to a multifactorial process involving direct tissue damage, secondary vascular injury and release of phamacological agents by polymorphonuclear cells (Futrell, 1992). In addition, it has been reported that, after gel filtration, the fraction of venom responsible for necrotic skin lesions, haemolysis and platelet aggregation has sphingomyelinase D activity (Futrell, 1992; Tambourgi et al., 1998). Loxosceles venom does not affect neutrophils directly, but is completely dependent on the victim’s neutrophils and complement for its effects to be manifested (Patel et al., 1994). This suggests that it probably activates other pathways, which indirectly culminate in complement activation, neutrophil migration and accumulation in the endothelial wall. Ceramides, the breakdown products of sphingomyelin, are known to be powerful regulators of various biological processes (Ballou et al., 1996). Ceramides are also important cellular regulators of TNF-a (Hannun, 1994), which in turn is a potent regulator of neutrophil chemotaxis, adhesion, priming, phagocytosis, inflammatory mediator release and superoxide generation (Ballou et al., 1996). TNF-a also enhances the adhesiveness of neutrophils for activated endothelium and extracellular matrix proteins, such as fibrinogen, fibronectin, and even serum-coated plastic (Gamble et al., 1992). In addition, treatment of monocytes/macrophages with TNF-a causes secretion of various cytokines such as IL-1, IL-6 and IL-8 (Fiers, 1995). Since L. gaucho venom contains an enzyme (a sphingomyelinase D) capable of cleaving sphingomyelin into choline and ceramide phosphate, this study investigated whether the presence of exogenous sphingomyelin and/or its breakdown product ceramide phosphate might be responsible for some of the effects of the venom in vivo. The mouse was chosen for this study since it appears to be less susceptible to the local effects of venom than other species. It is well documented in the literature that the structure and content of sphingomyelin in membranes varies among species. Coleman et al. (1980) found that the differing behaviour of red cells from different species correlated with their differing sphingomyelin content, while Bettger et al. (1998) found that the sphingomyelin of rats and mice
possesses a fatty acid chain that is not present in humans. In light of this, it is possible that such variations between species in the amount and type of sphingomyelin present could account for the observation that Loxosceles venom induces necrosis in rabbits and humans but not in mice (Futrell, 1992; Barbaro et al., 1996). It was hypothesised that this might be precisely because mice lack the appropriate pool of sphingomyelin in sufficient quantity (He et al., 2001), thus permitting studies to elucidate the mechanism by administration of exogenous sphingomyelin. Finally, since TNF-a is a potent chemoattractant for leukocytes to the sites of inflammation, experiments in vivo were conducted to examine whether co-administration of exogenous sphingomyelin would influence TNF-a release induced by L. gaucho venom at the local injection site.
2. Material and methods 2.1. Animals BALB/c female mice (6– 8 weeks old) and BALB/c male mice (18 – 20 g) were obtained from Harlam, UK and Butantan Institute Biological Research Facilities, respectively. 2.2. Venom Loxosceles venom was collected from the fangs of L. gaucho by electrical stimulation as described by Barbaro et al. (1992). Venom from 600 spiders was collected, pooled and lyophilised. Phosphate-buffered saline (6 ml) was added to the lyophilised venom, after which the solution was filtered (0.2 mm), aliquotted and stored at 220 8C until required. The protein content of the venom was determined using the BIO-RAD Protein assay (catalogue number: 5000116) using BSA as protein standard. 2.3. Sphingomyelin and ceramide phosphate Hen egg sphingomyelin was purchased from Sigma (St Louis, MO, USA) and D -erythro-sphingosine-1-phosphate, N-octanoyl (referred to below as ceramide phosphate) was purchased from Calbiochem (San Diego, CA, USA). 2.4. Dispersion of ceramide phosphate in saline One milligram of ceramide phosphate was weighed into a glass vial and dissolved in 1 ml of methanol/dichloromethane (1:2). The solution was dried down under a stream of nitrogen and 1 ml of distilled water was added to the vial. The vial was then capped, sealed with parafilm and sonicated at maximum amplitude in a Model 575D Truesweep Ultrasonic bath (Crest Ultrasonics Corporation,
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
New Jersey, USA) at 40 8C until clarity was achieved. The ceramide phosphate dispersion obtained had a concentration of 1 mg/ml in saline. 2.5. Dispersion of ceramide phosphate in liposomes Five hundred microlitres of dichloromethane/methanol (2:1) containing 10 mg of soya phosphatidyl choline (Lipoid) was added to 1 mg of dry ceramide phosphate (Calbiochem) and the resulting solution was dried down under a stream of nitrogen. Saline (0.5 ml) was then added to the dry residue. The sample was sonicated as described above and had the concentration adjusted to 1 mg/ml of ceramide phosphate in saline. A control preparation of phospholipid liposomes was prepared as described here, except that ceramide phosphate was absent.
473
2.9. Determination of LD50 of L. gaucho venom after subcutaneous injection in the footpad BALB/c male mice were divided in groups of three mice each and injected subcutaneously in the footpad with different concentrations of L. gaucho venom in PBS ranging from 0.06 to 0.36 mg/kg (w/w). The LD50 was determined 48 h after injection by probit analysis (Finney, 1971). LD50 ¼ 0.15 mg/kg of body weight (w/w) (confidence limits 95%, 0.1 – 0.2 mg/kg (w/w)) in 48 h. This experiment was performed twice. 2.10. Effect of sphingomyelin on mortality due to venom To determine whether sphingomyelin had any protective effect against mortality due to L. gaucho venom, the following experiments were performed.
2.6. Dispersion of sphingomyelin in saline One hundred milligrams of sphingomyelin was weighed into a 7 ml glass screw-capped vial and dissolved in either 1 ml or 400 ml of methanol/dichloromethane (1:2). Aliquots (200 ml) of 100 mg/ml sphingomyelin solution were transferred to fresh vials. The 200 ml aliquots containing 20 mg of sphingomyelin or the 100 mg/400 ml solution were dried down under a stream of nitrogen. Two millilitre of saline and 720 ml of saline were added, respectively, to the vials containing 20 and 100 mg of sphingomyelin. The samples were sonicated as described before. The sphingomyelin dispersion obtained was at a concentration of either 10 or 139 mg/ml in saline. 2.7. TNF-a detection An ELISA Kit for murine TNF-a from R & D Systems (Minneapolis, MN, USA) was used to measure TNF-a levels in supernatants of tissue extracts from the local injection site. The tissue extracts were prepared as described in Section 2.13.
2.10.1. Experimental design 1—intradermal injection Twenty eight mice were divided into four groups of seven mice each and injected (0.1 ml) i.d. in the dorsal area of the back with 0.5 mg/kg of the body weight (w/w) of L. gaucho venom either alone or with 1 mg of sphingomyelin were left for 1 h at 37 8C. The controls received either 1 mg/0.1 ml of sphingomyelin alone or 0.1 ml of PBS. Death and other symptoms of envenoming were recorded 24 h after injection. Animals that showed severe symptoms of envenoming prior to this time were sacrificed immediately. 2.10.2. Experimental design 2—subcutaneous injection Twelve mice were divided into four groups of three mice each and injected (30 ml) in the footpad with a LD50 dose of L. gaucho venom alone or with venom incubated for 1 h at 37 8C with 1 mg of sphingomyelin. The controls received either 1 mg of sphingomyelin alone or PBS. Death and other symptoms of envenoming were recorded 24 and 48 h after injection. This experiment was performed twice. Animals that showed severe symptoms of envenoming prior to this time were sacrificed immediately.
2.8. Venom detection 2.11. Induction of wound after intradermal injection Venom levels were determined by ELISA using the method described by Theakston et al. (1977) and modified by Ho et al. (1986). Briefly, 96-well microplates were coated with horse anti-arachnidic serum produced by the Butantan Institute. A standard curve of L. gaucho venom dilutions was used to determine the amount of venom in the test samples. Rabbit anti-L. gaucho venom IgG was used as secondary antibody and goat anti-rabbit IgG labelled with peroxidase was used as conjugate. After washing, substrate solution (ortho-phenylenediamine and H2O2) was added to the wells. The colour development was stopped by addition of 50 ml of diluted hydrochloric acid and the intensity of the colour was measured at 492 nm. The lower limit of sensitivity of the assay was 2 ng/ml.
To investigate whether exogenous sphingomyelin coadministered with L. gaucho venom could induce a wound or a non-specific inflammatory reaction at the local injection site after intradermal injection, mice were treated as described in Section 2.10 (experimental design 1) and examined daily. In addition, two other groups were also included in the experiment to determine the effect of ceramide phosphate (the product of cleavage of sphingomyelin by L. gaucho D -sphingomyelinase) administered locally. These mice were injected (0.1 ml) i.d. at the dorsal area with either 250 mg of ceramide phosphate in aqueous solution (seven mice per group) or 250 mg of ceramide phosphate in liposomes (six mice per group).
474
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
alone or with venom incubated for 1 h at 37 8C with 1 mg of sphingomyelin in 30 ml of PBS. Control groups received either sphingomyelin alone or PBS. At different times after injection, the mice were anaesthetised, bled by axillary plexus scission and sacrificed. Blood samples were placed into eppendorf tubes and centrifuged at 8000g for 5 min in an eppendorf centrifuge. Serum was then collected for assay of venom and TNF-a. Skin samples from the local injection site of mice injected intradermally in the dorsal area (1.5 cm in diameter) were excised after sacrifice and homogenized in 0.2 ml of PBS at room temperature. After homogenisation tissue extracts were centrifuged for 5 min in an eppendorf centrifuge and the supernatants were collected for assay of venom and TNF-a.
Examination for any kind of visible reaction was performed for up to six days post injection. Samples of the skin from the injection site of animals intradermally injected in the dorsal area of the back that died or showed any symptoms of envenoming in 24 h were taken and photographed before fixation with 1% of formaldehyde. The remaining animals were sacrificed either three days (ceramide phosphate and ceramide phosphate in liposome groups) or six days (Loxosceles venom þ sphingomyelin and sphingomyelin only groups) after injection, and, photographs of skin from the injection sites were taken prior to fixation with 1% formaldehyde. Following fixation, skin samples were stained with haematoxilin-eosin and viewed under a light microscope. 2.12. Oedema measurement after footpad injection BALB/c female mice were injected (30 ml) in the footpad with a LD50 dose (0.15 mg/kg of body weight (w/w)) of either L. gaucho venom alone, L. gaucho venom incubated with 1 mg of sphingomyelin, 1 mg of sphingomyelin alone or PBS. Oedema formation was measured with a caliper ruler in length and the landmark was stated as a point in the middle of the footpad. The results express the difference in mm (thickness) in the same animal between treated paw (left footpad) and untreated paw (right footpad). Oedema was measured at different times after injection.
2.13.2. Experimental design 2—subcutaneous injection Mice were injected subcutaneously in the footpad with a LD50 dose (0.15 mg/kg of the body weight (w/w) in 48 h) of L. gaucho venom alone or with venom incubated for 1 h at 37 8C with 1 mg of sphingomyelin in 30 ml PBS. Control groups received either sphingomyelin alone or PBS. At different times after injection, the mice were anaesthetised, bled by axillary plexus scission, sacrificed and footpad samples were collected for subsequent analysis. Blood samples were placed into eppendorf tubes and centrifuged (8000g) for 5 min in an eppendorf centrifuge. Serum was then collected for assay of venom. The footpad samples were processed as follows: after sacrifice, legs of mice treated subcutaneously (right footpad) or untreated (left footpad) were cut-off at the joint of the ankle and homogenized in eppendorf tubes with 0.2 ml of PBS at room temperature. The samples were then centrifuged for 5 min in an eppendorf centrifuge (8000g) and the supernatants were collected for assay of venom and TNF-a by ELISA.
2.13. Detection of venom and TNF-a at the local injection site (skin and footpad) To investigate the rate of venom clearance and TNF-a release at the injection site the following experiments were performed. 2.13.1. Experimental design 1—intradermal injection Mice were injected under light anaesthesia in shaved dorsal skin i.d. with 0.5 mg/kg (w/w) of L. gaucho venom
Table 1 Lethal and dermonecrotic actions of L. gaucho venom in the presence and absence of exogenous sphingomyelin Treatment
Parameters No. of survivors
Wound development
Clinical symptoms
L. gaucho venom 10 mga
0/7
0/7
L. gaucho venom 10 mg plus sphingomyelinb Ceramide phosphatec Ceramide phosphate plus liposomesc Sphingomyelinb PBSb
7/7 7/7 6/6 7/7 7/7
7/7 0/7 5/6 0/7 0/7
Hypothermia, prostration and paralysis of the hind limbs None None None None None
a b c
Parameters measured 24 h after i.d. injection. Parameters measured six days after i.d. injection. Parameters measured three days after i.d. injection.
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
2.14. Statistical analysis Results are presented as the means ^ s.e.m. For statistical analysis, two-way ANOVA followed by the Tukey test was performed using the SigmaStat 3.0
475
software. Values with p , 0:05 were considered statistically significant. Non-parametric tests (Kruskal-Wallis and Mann-Whitney) were employed for analyzing venom clearance at the local injection site and in serum.
Fig. 1. Mice were injected i.d. with either 0.5 mg/kg (w/w) of L. gaucho venom plus sphingomyelin (A—outside of skin, B—inner side of skin), 1.25 mg/kg of the body weight (w/w) of ceramide phosphate plus liposome (C—outside of the skin, D—inner side of the skin) or 0.5 mg/kg (w/w) of L. gaucho venom in 0.1 ml PBS (E—outside of the skin, F—inner side of the skin). The animals were sacrificed either 24 h (L. gaucho venom in PBS), three days (ceramide phosphate plus liposome group) or six days after injection. The black marks appearing in E and F were made by a pen marker to delineate the area of injection.
476
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
3. Results 3.1. Effect of sphingomyelin on toxic actions due to venom The results shown in Table 1 demonstrate that all animals injected i.d. at the dorsal area with venom either died or developed strong symptoms of envenoming such as hypothermia, prostration and paralysis of the hind limbs. Mice injected intradermally in the footpad also develop similar symptoms of envenoming (data not shown). In contrast, all animals from the groups injected with venom co-administered with sphingomyelin survived, and developed at the site of injection a wound on the outer side of the skin and an abscess-like reaction on the inner side (Fig. 1A and B). The liposomal ceramide phosphate group also developed, at the site of injection, a mild wound on the outer side of the skin and a slight haemorrhagic mark on the inner side (Fig. 1C and D). The blood vessels of mice injected with venom alone were more evident but there was no visible sign of lesion (Fig. 1E and F). None of the other groups developed any kind of visible reaction at the injection site. Histological analysis of the skin from mice injected i.d. with venom plus sphingomyelin showed an infiltration of leukocytes at the injection site (data not shown). 3.2. Determination of oedema reaction In order to determine the influence of exogenous sphingomyelin on development of oedema at the local injection site, mice were injected subcutaneously in the footpad with either venom alone, venom co-administered with sphingomyelin, sphingomyelin alone or PBS. The results obtained in this experiment showed that 2 and 6 h after injection, oedema was evident in the groups treated with venom alone or venom co-administered with sphingomyelin, and was significantly more marked ðp , 0:05Þ than in the PBS group (Fig. 2). No statistical difference in size of oedema swelling could be observed 2 and 6 h after injection between the groups treated with venom alone or venom coadministered with sphingomyelin. Twenty four hours after injection, swelling due to oedema was only observed in the group treated with venom co-administered with sphingomyelin ðp , 0:05Þ (Fig. 2). Moreover, six days after injection, oedema was observed only in the group injected with venom co-administered with sphingomyelin (data not shown).
Fig. 2. Influence of exogenous sphingomyelin (Sm) on oedema development. BALB/c female mice were injected (30 ml) in the footpad with a LD50 dose (0.15 mg/kg of body weight (w/w)) of either L. gaucho venom alone (Venom), L. gaucho venom with incubated 1 mg of sphingomyelin (Venom þ Sm), 1 mg of sphingomyelin alone (Sm) or PBS. Oedema reactions were measured with a caliper ruler at different times after injection. The asterisk represents statistically significant differences ðp , 0:05Þ between venom alone, venom plus sphingomyelin and sphingomyelin groups with control (PBS) group.
sphingomyelin. Samples of tissue from the local injection site and blood were collected at different times after injection, and assayed for venom by ELISA. The results showed that in both routes of injection, a higher level of L. gaucho venom was detected at the local injection site when venom was co-administered with sphingomyelin compared with the groups that received venom alone (Fig. 3A and B). Although, 6 h after venom injection, the level of venom at the local injection site decreased markedly in both cases, the level of venom was still two times higher in the group that received venom coadministered with sphingomyelin (Fig. 3A and B). In addition, even 24 h after injection, the level of venom detected locally in the group injected with venom coadministered with sphingomyelin in the footpad was significantly higher than the level obtained in groups that received venom alone. However, in serum, regardless of the route of injection, there was no difference in the levels of venom between the groups treated with venom alone or venom co-administered with sphingomyelin (Fig. 3C and D). 3.4. TNF-a detection at the local injection site
3.3. Determination of venom clearance at the local injection site and serum In order to determine whether exogenous sphingomyelin would influence the kinetics of venom in serum and at the local injection site, mice were injected either intradermally (dorsal skin) or subcutaneously (in the footpad) with L. gaucho venom either alone or co-administered with
In order to determine the influence of exogenous sphingomyelin on TNF-a release at the local injection site, mice were injected intradermally (dorsal skin of the back) or subcutaneously (footpad) with L. gaucho venom either alone or co-administered with sphingomyelin. Two hours after injection, samples of the local injection site were collected and assayed for TNF-a by ELISA after previous
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
477
Fig. 3. Detection of L. gaucho venom at the local injection site and serum after intradermal and subcutaneous injections. BALB/c male mice were injected with L. gaucho venom (alone or incubated with 1 mg of sphingomyelin (Sm) for 1 h at 37 8C) either subcutaneously in the footpad (0.15 mg/kg of body weight (w/w) diluted in 30 ml) or intradermally in the dorsal shaved skin of the back (0.5 mg/kg of body weight (w/w) diluted in 0.1 ml). Samples of the local injection site (A—subcutaneously injected mice; B—intradermally injected mice) and blood (C— subcutaneously injected mice; D—intradermally injected mice) were collected at different times after injection for venom detection by ELISA. # Statistical difference ðp , 0:05Þ between experimental times in the same experimental group. * Statistical difference ðp , 0:05Þ between the groups Venom plus Sm and Venom alone.
observation that local levels of TNF-a drop drastically 6 h after injection. The results show that, at the time-points measured, regardless of the route of injection, there was no significant difference between the levels of TNF-a detected in any of the test groups investigated, compared with controls that received either sphingomyelin or PBS alone (data not shown).
4. Discussion Sphingolipids such as sphingomyelin are most often localized in the outer leaflet of the plasma membrane and are an integral component of signalling pathways used by many types of cells (Boucher et al., 1995; Gulbins et al., 1995). These signals instruct the cells to perform a variety of functions such as cytokine production, differentiation,
proliferation or apoptosis. One of the most important membrane lipids capable of giving rise to metabolically active signal transduction is sphingomyelin. Ceramide, the immediate product of sphingomyelin hydrolysis, and its further metabolites such as sphingosine and sphingosine-1phosphate, play an important role in a variety of signal transduction pathways. It has been demonstrated that the lytic and necrotic action of Loxosceles venom is due to a fraction that has sphingomyelinase D activity (Tambourgi et al., 1998). Natural mammalian sphingomyelinase on the other hand cleaves sphingomyelin to remove not choline, but phosphocholine, giving ceramide as a final product. Although ceramide phosphate can occur naturally, it is generated endogenously not from sphingomyelin, but from ceramide by the action of a calcium-dependent ceramide kinase found predominately in the plasma membrane (Shalini and Kolesnick, 1993; Hinkovska-Galcheva et al., 1998). Ceramide phosphate from either source can be converted back into ceramide by the action of ceramide-1P-phosphatase (Boudker
478
M.O. Domingos et al. / Toxicon 42 (2003) 471–479
and Futerman, 1993; Shinghal et al., 1993). It is conceivable, therefore, that, ceramide phosphate generated exogenously by sphingomyelinase D action could have a profound effect on cells by participating directly in cell signalling pathways. The results presented here show that L. gaucho venom administered i.d. in the dorsal area in the presence of exogenous sphingomyelin generated both a wound and an inflammatory reaction at the injection site and that it also protected adult mice against death. This is the first time it has been observed that the presence of exogenous sphingomyelin can protect mice against a lethal dose of Loxosceles venom and in addition is capable of inducing the venom to generate in mice an inflammatory reaction at the local injection site. L. gaucho venom alone, at the same dose level, induced death in all animals tested without causing any reaction at the injection site. Subcutaneous injection of a LD50 dose of L. gaucho venom alone in the footpad induced an oedema reaction at the local injection site that could be observed 2 and 6 h after injection, but which diminished after 24 h. In contrast, the group treated in the footpad with venom co-administered with sphingomyelin developed an oedema reaction that could be observed up to six days after injection (data not shown). In addition, only two out of six mice survived the LD50 dose injection of L. gaucho venom alone in the footpad (data not shown), while animals all survived without showing any visible symptoms of envenoming when the same dose was administered with sphingomyelin. One hypothesis for the observations described above is that exogenous sphingomyelin helps to retain the venom in the local area, preventing it from dispersing systemically, where it can cause death of the animal. This trapping of venom results in a local inflammatory response that culminates in an internal abscess-like reaction in the skin, and wound formation. This hypothesis is supported by the results obtained in this study, which show that a significantly higher level of venom was retained at the local injection site when venom is co-administered with sphingomyelin either by intradermal injection in the skin or by subcutaneous injection in the footpad. The results obtained by Gomez et al. (2001) also corroborate this hypothesis since they have shown that the extent of dermal inflammation in rabbits is directly correlated with Loxosceles venom diffusion. Furthermore, the results obtained with ceramide phosphate (a small rapidly diffusing molecule) also showed that intradermal injection in saline did not induce any visible reaction at the site of injection nor did it cause death of animals. However, when ceramide phosphate was administered intradermally in liposomal form (to prevent its rapid dispersal from the site), a mild wound was produced on the outer side of the skin, and a haemorrhagic lesion was visible on the inner surface. Likewise, histological observation also showed infiltration of leukocytes in the muscle layer. It is possible that due its small size and high water solubility, ceramide phosphate in saline was dispersed quickly from
the local injection site. However, its incorporation in liposomes may have helped its retention at the site of injection thus inducing a mild inflammatory response. The observation that co-administration of venom with sphingomyelin reduces its systemic toxicity in mice, yet does not appear to reduce the level of venom reaching the bloodstream is surprising, and is worthy of further study. Although work elsewhere (Domingos et al., 2003) has shown that action of L. gaucho venom on sphingomyelin induces TNF-a release by macrophages in vitro, this study demonstrated that no increase in TNF-a levels by L. gaucho is detectable at the local injection site, even in the presence of exogenous sphingomyelin, compared with sphingomyelin alone.
Acknowledgements The authors are also very grateful to Dr Marcelo Larami Santoro for the statistical analysis and Marı´lia Brinati Malta and Sabrina Soares for their excellent technical assistance.
References Ballou, L.R., Laulederkind, S.J.F., Roloniec, E.F., Raghow, R., 1996. Ceramide signalling and the immune response. Biochim. Biophys. Acta 1301, 273–287. Barbaro, K.C., Cardoso, J.L.C., Eickstedt, V.R.D., Mota, I., 1992. Dermonecrotic and lethal components of Loxosceles gaucho spider venom. Toxicon 30, 331–338. Barbaro, K.C., Sousa, M.V., Morhy, L., Eickstedt, V.R.D., Mota, I., 1996. Compared chemical properties of dermonecrotic and lethal toxins from spiders of the genus Loxosceles (Araneae). J. Prot. Chem. 15, 337–343. Bettger, W.J., Blackader, C.B., McCorquodale, M.L., Ewing, R., 1998. The occurrence of iso 24:0 (22 Methyltricosanoic acid) fatty acid in sphingomyelin of rat tissues. Comp. Biochem. Physiol. B. Biochem. Mol. Biol. 119, 229– 304. Boucher, L.M., Wiegmann, K., Futterer, A., Pfeffer, K., Machleidt, T., Schu¨tzer, S., Mak, T.W., Kro¨nke, M., 1995. CD28 signals through acidic sphingomyelinase. J. Exp. Med. 181, 2059–2068. Boudker, O., Futerman, A.H., 1993. Detection and characterization of ceramide-1-phosphate activity in rat liver plasma membrane. J. Biol. Chem. 268, 22150–22155. Coleman, R., Lowe, P.J., Billington, D., 1980. Membrane lipid composition and susceptibility to bile salt damage. Biochim. Biophys. Acta 599, 294–300. Domingos, M.O., Barbaro, K.C., Tynan, W., Penny, J., Lewis, D.J.M., New, R.R.C., 2003. Effect of Loxosceles gaucho venom on cell morphology and behaviour in vitro in the presence and absence of sphingomyelin. Toxicon, in press, Manuscript No. AH/2003/000052. Fiers, W., 1995. Biologic therapy with TNF: preclinical studies. In: DeVita, V.T., Hellman, S., Rosenberg, S.A. (Eds.), Biologic Therapy of Cancer, Second ed., B. Lippincott Co., Philadelphia, Chapter 12.
M.O. Domingos et al. / Toxicon 42 (2003) 471–479 Finney, D.J., 1971. Probit Analysis, Third ed., Cambridge University Press, Cambridge. Futrell, J.M., 1992. Loxoscelism. Am. J. Med. Sci. 304, 261 –267. Gamble, J.R., Smith, W.B., Vadas, M.A., 1992. TNF modulation of endothelial and neutrophil adhesion. In: Beutler, B., (Ed.), Tumor Necrosis Factor. The Molecules and their Emerging Role in Medicine, Raven Press, New York, pp. 65–86. Gomez, H.F., Dian, M., Greefield, M.D., Miller, M.J., Jeffrey, S., Warren, M.D., 2001. Direct correlation between diffusion of Loxosceles reclusa venom and extent of dermal Inflammation. Acad. Emerg. Med. 8, 309–314. Gulbins, E., Bissonnette, R., Mahboubi, A., Martin, S., Nishioka, W., Brunner, T., Baier, G., Bier-Bitterlich, G., Byrd, C., Lang, F., 1995. FAS-induced apoptosis is mediated via a ceramideinitiated RAS signalling pathway. Immunity 2, 341 –351. Hannun, Y.A., 1994. The sphingomyelin cycle and the second messenger function of ceramide. J. Biol. Chem. 269, 3125–3128. He, X., Chen, F., Gatt, S., Shuchman, E.H., 2001. An enzymatic assay for quantifying sphingomyelin in tissues and plasma from humans and mice with Niemann–Pick disease. Anal. Biochem. 293, 204 –211. Hinkovska-Galcheva, V.T.Z., Boxer, L.A., Mansfield, P.J., Harsh, D., Blackwood, A., Shayman, J.A., 1998. The formation of ceramide-1-phosphate during neutrophil phagocytosis and its role in liposome fusion. J. Biol. Chem. 273, 33203–33209. Ho, A., Warrell, M.J., Warrell, D.A., Bidwell, D., Voller, A., 1986. A critical reappraisal of the use of enzyme-linked immunosorbent assays in the study of snake bite. Toxicon 24, 211–221. Ma´laque, C.M., Castro-Valencia, J.E., Cardoso, J.L., Franc¸a, F.O.S., Barbaro, K.C., Fan, H.W., 2002. Clinical and epidemiological features of definitive and presumed loxoscelism in Sa˜o Paulo, Brazil. Rev. Inst. Med. Trop. Sa˜o Paulo 44, 139–143.
479
Ministry of Health, 1998. In: Manual de diagno´stico e tratamento de acidentes por animais pec¸onhentos. Fundac¸a˜o Nacional de Sau´de/Coordenac¸a˜o de Controle de Zoonoses e Animais Pec¸onhentos, CENEPI, Brası´lia,. Patel, K.D., Modur, V., Zimmerman, G.A., Prescott, S.M., McIntyre, T.M., 1994. The necrotic venom of brown recluse spider induces dysregulated endothelial cell-dependent neutrophil activation. J. Clin. Invest. 94, 631–642. Sezerino, U.M., Zannin, M., Coelho, L.K., Gonc¸alves, J. Jr., Grando, M., Mattosinho, S.G., Cardoso, J.L.C., Eickstedt, V.R.D., Franc¸a, F.O.S., Barbaro, K.C., Fan, H.W., 1998. A clinical and epidemiological study of Loxosceles spider envenoming in Santa Catarina, Brazil. Trans. R. Soc. Trop. Med. Hyg. 92, 546– 548. Shalini, M., Kolesnick, R., 1993. Ceramide: A novel second messenger. Adv. Lipid Res. 25, 65–90. Shinghal, R., Scheller, R.H., Bajjalieh, S.M., 1993. Ceramide-1phosphate phosphatase activity in brain. J. Neurochem. 61, 2279–2285. Tambourgi, D.V., Magnoli, F.C., van den Berg, C.W., Morgan, B.P., Araujo, P.S., Alves, E.W., Dias da Silva, W., 1998. Sphingomyelinases in the venom of the spider Loxosceles intermedia are responsible for both dermonecrosis and complement-dependent hemolysis. Biochem. Biophys. Res. Commun. 251, 366–373. Theakston, R.D.G., Lloyd-Jones, M.J., Reid, H.A., 1977. MicroElisa for detecting and assaying snake venom and venom antibody. Lancet 24, 639–641. White, J., Cardoso, J.L., Fan, H.W., 1995. Clinical toxicology of spider bites. In: Meier, J., White, J. (Eds.), Handbook of Clinical Toxicology of Animal Venoms and Poisons, CRC Press, Boca Raton, pp. 261–329.