Abstract

ORIGINAL PAPER

Insects as biological models to assay spider and scorpion venom toxicity

M. F. Manzoli-Palma^I; N. Gobbi^{I, II}; M. S. Palma^{II, III}

^ICenter of Environmental Studies - CEA, UNESP, São Paulo, Brazil

^IICenter for the Study of Venoms and Venomous Animals - CEVAP, UNESP, São Paulo, Brazil

^IIIDepartment of Biology - Institute of Biosciences of Rio Claro, UNESP, São Paulo, Brazil

^{Correspondence} Correspondence to N. Gobbi Centro de Estudos Ambientais (CEA) Av. 24 A, 1515, Bela Vista 13506-900, Rio Claro, SP, Brasil Phone/Fax: 55 19 3523 2605 Email: ngobbi@rc.unesp.br

ABSTRACT

This study was undertaken to develop an experimental protocol using insects as biological models to assay venom toxicity of the following spiders Loxosceles gaucho, Phoneutria nigriventer, Nephilengys cruentata and Tityus serrulatus scorpion. Three different insect species were bioassayed: Apis mellifera (Hymenoptera), Grillus assimilis (Orthoptera), and Diatraea saccharalis (Lepidoptera). Venoms were injected into the hemocele of insects with a microsyringe at concentrations that caused dose/weight-dependent effects; doses causing either paralysis (ED50) or death (LD50) were recorded for each venom and insect test-species. T. serrulatus and L. gaucho venoms were lethal to all tested species, while P. nigriventer venom caused paralysis and death, and N. cruentata venom caused only paralysis at the doses assayed. A comparison between the insect test species described above revealed that A. mellifera was highly sensitive to all venoms tested; even a tiny amount of N. cruentata non-lethal venom caused a change in the walking pattern leading to transient paralysis. D. saccharalis larvae were very resistant to all four venoms.

Keywords: bioassay, venom, insect model, toxicity, Apis mellifera.

INTRODUCTION

The nervous systems of insects serve as a major target for conventional insect control, which relies currently on hazardous chemicals of four groups of pesticides: organochlorines, organophosphates, carbamates, and pyrethroids. The main problem with the currently used insecticides is that they are non-selective chemicals used in very large amounts and insects develop resistance to them. The chemicals are polluting our environment and killing other animals along with pest insects. Presently, one of the aims of the chemical industry is to develop new safer means for pest control; this goal may be achieved by using natural anti-insect agents, such as those developed by scorpions and spiders during millions of years of evolution.

Spider and scorpion venoms are rich sources of low molecular weight compounds, which are generally neurotoxic. Some of these compounds are insect-specific, others are mammal-specific, while others are directed at both insect and mammal (10). In particular, the search for new insect-specific neurotoxins is becoming an important area of investigation by agrochemical companies to be used as starting points for the development of highly selective bioinsecticides (8,17).

Different screening strategies have led to the identification of potent neurotoxins, which act either on central or peripheral nervous systems (6). The assay protocols for spider and scorpion venoms in general use insects of different orders: Blattaria (Blatta orientalis); Lepidoptera (Manduca sexta), Coleoptera (Tenebrio molitor); Orthoptera (Acheta domesticus, Spodoptera litoralis, Locusta migratoria, and Grillus bimaculatus); and Diptera (Drosophila melanogaster, Sarcophaga, argyrostoma, and Musca domestica). These insects may present very different sensitivity to each venom type. Thus, it is often difficult to compare different results for the same venom, when assayed in different test-insects, mainly under different experimental conditions.

The purpose of this investigation was to develop a simple, reliable, and reproducible comparison protocol, which may be applied to the selection of a suitable insect model for toxicity evaluation of each type of venom in bioprospective investigations. In this study, we assayed the venoms of four common venomous Arthropoda in Southeast Brazil against honeybees, crickets, and Lepidoptera larvae.

MATERIAL AND METHODS

Venom - The venoms investigated were from the spiders Loxosceles gaucho, Phoneutria nigriventer, and Nephilengys cruentata and the scorpion Tityus serrulatus collected in the wild. Crude venom was extracted from frozen spiders and scorpions; their glands were dissected, homogeneized, and centrifuged always maintaining the temperature at 4°C. The clear viscous supernatant obtained by centrifugation was then lyophilized in serum bottle with a HETO freeze-dryer. This lyophilized material, henceforth referred to as venom, was weighed, stored in a dessecator, and kept at –15°C.

Insects - The effects of spider and scorpion venoms were tested in three insect species: just-emerged adult Apis mellifera (Hymenoptera, Apidae) weighing between 70 and 80 mg were collected in the apiary of UNESP at Rio Claro, SP; Gryllus assimilis (Orthoptera, Gryllidae) were reared in laboratory conditions at 26ºC (14); 3^rd instar-nymphs of 65-75 mg body weight were used in all experiments. Larvae of Diatraea saccharalis (Lepidoptera, Pyralidae) at the 6^th instar (with 100.0 ± 10.0mg. body weight) were obtained from São João Sugar and Alcohol Plant in Araras –São Paulo State, and kept in cylindrical glass vials using artificial diet (7).

Injections - The insects were not anesthetized but carefully immobilized with the hands. Venom extracts were injected in the pronotum of honeybees and Diatraea saccharalis larvae, while in the crickets the injections were performed intrathoracically between the second and third pair of legs. Venoms (1µl per insect) were injected with a 10 µl Hamilton syringe. The LD50 or ED50 values for each situation were determined by using four different concentrations of each venom and compared against a control (physiological solution); N=15 per concentration. The number of dead or paralyzed insects was determined up to 5 and 24 h after venom application, according to the characteristic of each venom. Toxicity levels were calculated according to the method of Probit (2) and expressed as lethality (LD50) or paralysis (ED50).

RESULTS AND DISCUSSION

A large percentage of Arthropod venom composition is generally represented by non-protein/non-peptide components that contribute significantly to venom toxicity (11). Crude venoms must be extracted and quantified by their dry-weight. Protein quantification cannot be performed because it produce under-represented results.

Table 1 shows toxicity data from the above insect tests for venoms of some venomous Arthropoda. Controls receiving injections of physiological solution showed no locomotive effects in all species tested.

Bioassay - Apis mellifera

Just-emerged adult Apis mellifera showed to be excellent toxicity indicators to all venoms assayed. Preliminary tests determined that recently emerged bees are better manipulated than older ones and allow the use of workers of exactly the same age (in hours); young workers still do not have the sting hardened, so they cannot sting the researcher making their manipulation reasonably safe. Thus, it is not necessary to anesthetize them. Honeybees are very active insects, and the effect caused by venom injections are clearly observed since any abnormal behavior is easily perceptible. The normal behavior pattern of this insect considers the control group as always active, feeding itself on glucose solution, walking in circles, climbing on Petri dish covers, or sometimes touching each other.

The group of workers assayed with Tityus serrulatus venom showed an excitatory phase with the bees walking fast on Petri dish, until the toxicity symptoms started to appear with a decrease in locomotion. Three hours after venom injection, the bees dropped dorsally and presented difficulties to return to the normal standing position. They were dizzy and their methatoraxic legs did not fold. After falling dorsally, the bees did not return to the normal position in spite of being able to move their legs; at this stage spasmic abdomen contractions occurred until insect death.

According to Bucherl (1), the average content of T. serrulatus venom is 15.37mg, which is 1052 times higher than the total amount of venom needed to kill a honeybee worker with a LD50 = 15.58 ng/mg. Toxicity of T. serrulatus venom was also studied by other researchers, considering either the contractile effects on the body of test larvae, expressing the results in "paralytic contractile unit" (PCU) and paralysis of adult insects expressed as ED50; these values generally changed from 2.60 to 30.38 ng/mg (3,4,9,18,19,21,22). LD50 determined in this study corresponds to a rate value in this interval. However, it is important to emphasize that the use of Apis mellifera as a model insect for bioassays is very convenient because this insect is more sensitive to scorpion venom than mostinsects previously used as models in literature (3,4,21), since the same concentrations of T. serrulatus venom that caused only paralytic effects on Diptera and Orthoptera were lethal to honeybees.

P. nigriventer venom injection in honeybees caused flaccidity and loss of body balance, also the walking pattern became lethargic and after lateral dropping the honeybee workers remained paralyzed in this position until death. Paralysis occurred slowly and was an intermediary state before insect death. Phoneutria nigriventer has about 2.15 mg of venom in its glands, which is 1529 fold higher than the amount needed to reach LD50 (15.20 ng/mg ) for honeybee workers.

After L. gaucho venom injection, the honeybees began staggering, dropping, and bending the body due to a sustained contraction of the ventral musculature; their death occurred one hour later. Lethal effects are fast and the bees become rigid, dropping without any reflex. The toxic effects of this venom are faster and more intense than Phoneutria nigriventer and Tityus serrulatus venoms. According to Norment and Smith (15), Loxosceles reclusa venom causes hemocyte lysis in Acheta domesticus.

LD50 (LD50=3.76ng/mg) (Table 1) is similar to that described for Loxosceles deserta when Blatta orientalis was used as insect test (1.8 ng/mg) (13). If compared with other toxicity data available in literature (5,12,13,17,20), it could be considered one of the seven most lethal ones against Arthropoda in general in spite of the different sensitivity of each insect test. According to Bucherl (1), spiders of the Loxosceles genus possess between 0.13 and 0.27 mg of venom in their glands, which represents on average about 583 times the LD50 needed value observed in this study.

When assayed in honeybees, toxicity value found for Nephilengys cruentata venom was similar (ED50 = 37.00ng/mg) (Table 1) to that of Argiope bruennichi venom (also a web-spider) assayed in Tenebrio molitor (5), and lower than in Blatta orientalis (5). Aerial web-spider venoms are rich in acilpolyaminetoxins, which are blockers of glutamate receptors; these toxins cause locomotive paralysis in preys (11). Paralysis generally requires relatively high doses of crude venom, as observed in this study. Even pure toxins isolated from Hololena curta caused some effects at high doses. Thus, when assayed in Manduca sexta, the toxins HO-489 and HO-505 presented ED50 equal to 54 and 81 ng/mg, respectively (17). Although these doses are relatively high, specifically in the case of Nephilengys cruentata, this represents the need of 3.422 ng of venom to reach ED50 for a bee worker. Considering the 700 mg of venom existing in its glands the total venom content of this spider is about 204 times higher than necessary to cause paralysis with the ED50 described above.

Paralysis caused in honeybee workers can last from hours to days, depending on the amount of venom injected. This is a characteristic of most web-spider venoms. According to Quicke (16), when the injected insects are left to rest for some time and then stimulated by touch, they can move a little in Petri dish, returning then to the paralyzed state.

Bioassay - Grillus assimilis

After T. serrulatus venom injection in Gryllus assimilis, an abrupt interruption of movements is observed, and in some cases, a lengthening of the abdomen, almost separating it from the thorax. Sometimes, the beginning of the ecdyse process was seen, probably associated with an increase in hemolymph pressure caused by scorpion venom toxins, which facilitates the rupture of the ecdyse line; 1 hour after venom injection the crickets died.

The crickets became paralyzed after injection of P. nigriventer venom (ED50 = 43.53 ng/mg), but most individuals recovered after 5 hours; no lethal effect was observed under the assayed conditions. When paralyzed, some crickets still moved their legs while others trembled. The cerci was rigid and when touched some crickets jumped a little bit or dragged themselves with their abdomen touching the Petri dish; immediately after this, they became paralyzed again.

The actions of P. nigriventer venom are better understood in vertebrates; mice seem to provide the better physiological and pharmacological systems for assaying this venom (19). Despite this venom not being lethal to mammals, it has impressive toxicity to mice, LD50=200µg/kg (0.2ng/mg) is about 218 times lower than that necessary to cause paralysis in Gryllus assimilis.

Immediately after L. gaucho venom injection, the crickets showed important toxicity symptoms: they dragged themselves with abdomen touching the Petri dish; they were not able to sustain the bodies on the legs, and the antennae drooped (crickets of the control group maintained them elevated all time). One hour after injection, the crickets fell over on their sides, became completely rigid, and died (LD50 = 20.61ng/mg). When the value of lethal dose is compared to that obtained in bees (LD50 = 3.76ng/mg), a larger amount is necessary for bioassay in crickets than Apis mellifera, indicating a higher sensitivity from honeybee to the venom.

Injection of N. cruentata venom caused immediate paralysis in crickets similar to that already described above for honeybees, ED50=24.48ng/mg. In spite of paralysis, the crickets were able to move their legs; from time to time they jumped a little or crawled; however, they were not able to maintain their bodies on the legs. These symptoms characterize the effect of this venom as incomplete paralysis.

In spite of being less sensitive than honeybees, its facility of breeding and active behavior make crickets a good insect-model to study toxic effects of scorpion and spider venoms.

Bioassay - Diatraea saccharalis

Diatraea saccharalis larvae are not active and since toxicity evaluation depend at least in part on behavioral observations, it is difficult to perform an accurate analysis by using these larvae. Thus, to quantify and analyze them, the larvae were placed in Petri dishes in the presence of small blocks of artificial diet as described by Guerra (7). This creates a new and important behavioral parameter. Thus, if toxic effects of venom appear after injection, some larvae will not feed, remaining outside the diet blocks. Meanwhile, those walking frequently towards the diet block (or entering it) and feeding regularly may be considered non-intoxicated and presenting normal behavior, just as the control group in which only physiological solution was injected.

Actions of T. serrulatus venom on Diatraea saccharalis larvae were very slow; the larvae decreased their activity becoming flaccid and progressively paralyzed. After regurgitation of a dark liquid they stopped feeding and when touched only lift their head. This condition continued for 5 hours, after which they died contorted and rigid. Toxicity value observed for scorpion venom in Diatraea saccharalis larvae (LD50=1567.00 ng/mg) is very high when compared to other insect tests (Table 1).

During bioassay of P. nigriventer venom in Diatraea saccharalis, the larvae also regurgitated a dark liquid and defecated immediately after venom injection. They remained still for 1 hour, but when touched moved their head; after this time, they died (LD50=1493.00ng/mg) (Table 1). Their resistance to the toxic effects of Phoneutria nigriventer and Tityus serrulatus venoms was remarkable, as they were lethal only at high doses.

Diatraea saccharalis larvae were also very resistant to L. gaucho venom (LD50= 29.97ng/mg ), since the lethal dose is higher than that obtained in bees and crickets (Table1). After injection, the larvae eliminated a dark liquid and the region around the injection site became blackish, probably due to necrosis; 30 minutes after lethal dose injection they became rigid, contorted, and died.

After N. cruentata venom injection, the larvae became flaccid, slow, and paralyzed (not walking towards the diet block); 1 to 2 hours later, they recovered their normal activity. Considering that the action of N. cruentata venom is immediate paralysis of preys, the ED50 found for Diatraea saccharalis (ED50 = 608.00 ng/mg) demands the consumption of 79 ng of venom to induce this effect in each larvae. Thus, the total content of N. cruentata venom is sufficient to paralyze nine larvae despite the fact that Lepidopteran larvae are not natural web-spider preys. Nevertheless, this toxicity value can be considered very high, as it is 16 and 24 times higher than those of A. mellifera and G. assimilis, respectively.

CONCLUSIONS

From the three insects bioassayed, Apis mellifera showed to be more sensitive than crickets and Lepidopteran larvae to the toxicity test of Arthropoda venom. Honeybees are very active insects, and thus, the toxic effects are clearly seen, in addition to demanding low venom consumption. It is important to emphasize the remarkable resistance of Diatraea saccharalis larvae to the toxic effects of all assayed venoms. Thus, the use this larvae as a bioassay system is not reliable and convenient for simple screening purposes, since it demands large venom amounts and toxic effects are not clearly seen. However, considering that Lepidopteran larvae are common pests in agriculture, the use of Diatraea saccharalis provided a clear understanding about this insect’s resistance to toxic effects of some Arthropoda venoms. If this Lepidoptera was the target for the development of a new selective bioinsecticide, the venom of Loxosceles gaucho could be used as a starting point to search for a lead compound due to its high toxicity against this insect.

To perform toxicity bioassays of Arthropoda venoms using live insects, it is necessary to consider some aspects such as: sensitivity of insect tests to specific venoms; ease of breeding and maintaining each insect; appropriate body weight (in order to consume low amount of venom); proper description of paralysis, and the occurrence of easily observable behavioral patterns to help characterize the toxic effects of each venom.

ACKNOWLEDGEMENT

This work was supported by FAPESP from Brazil.

Received October 30, 2001

Accepted December 5, 2002

Publication Dates

History