Larvicidal Activity of the Water Extract of Moringa Oleifera Seeds against Aedes Aegypti and Its Toxicity upon Laboratory Animals

In this work, biological effects of the water extract of Moringa oleifera seeds (WEMOS) were assessed on eggs and 3 rd instar larvae of Aedes aegypti and on its toxicity upon laboratory animals (Daphnia magna, mice and rats). Crude WEMOS showed a LC 50 value of 1260μg/mL, causing 99.2 ± 2.9% larvae mortality within 24 h at 5200μg/mL, though this larvicidal activity has been lost completely at 80 • C/10 min. WEMOS did not demonstrate capacity to prevent egg hatching. After extensive dialyses of the crude WEMOS into water-soluble dialyzable (DF) and non-dyalizable (NDF) fractions, only DF maintained its efficacy to kill larvae. Acute toxicity evaluations on daphnids (EC 50 of 188.7μg/mL) and mice (LD 50 of 446.5 mg/kg body weight) pointed out to low toxicity. Despite the thymus hypertrophy, WEMOS revealed to be harmless in orally and subacutely-treated rats. In conclusion, WEMOS has thermostable bioactive compounds against Ae. aegypti larvae with apparent molecular mass lower than 12 kDa and moderately toxic potential.


INTRODUCTION
Dengue is an important human viral disease transmitted by Aedes aegypti Linnaeus, 1762 (Diptera: Culicidae) that is gradually becoming endemic in several Central and South American countries (Gubler 1998).Despite the infection with one or more dengue viruses has reached about 2-5 billion people living in tropical and subtropical countries, there are no specific antiviral drugs to treat it and no vaccines to prevent it (Halstead 2007).
Correspondence to: Dr. Paulo Michel Pinheiro Ferreira Present address: Universidade Federal do Piauí, Campus Senador Helvídio Nunes de Barros (CSHNB), sl.804, Rua Cícero Eduardo s/n, Bairro Junco, 64600-000 Picos, PI, Brasil E-mails: pmifepe@yahoo.com.br/ pmpf@ufpi.edu.br The Northeast region of Brazil has been suffering successive dengue epidemics, and official reports have demonstrated that the State of Ceará has been one of the most affected states at that region.From 1986 to 2008, a total of 243,619 cases were reported and until last December, 178 out of 185 municipalities were infested (SESA 2008).
The selective pressure of conventional insecticides, such as organochloranes, carbamates, pyrethroids and organophosphorates, is enhancing resistance of mosquito populations at an alarming rate, resulting in widespread resurgence, undesirable effects on non-target organisms and environmental and human health concerns (Brown 1986, Mazzari andGeorghiou 1995).This, An Acad Bras Cienc (2009) 81 (2) 208 PAULO M.P. FERREIRA et al. consequently, increases the demand for new products that be environmentally safe, target-specific and easily degradable.
Plant derived products have received much attention due to their natural chemical defenses against insect predators (Carvalho et al. 2003, Omena et al. 2007).However, to be registered as a pesticide, the compounds must be evaluated with respect to their toxicological and ecotoxicological acute and chronic effects under laboratory conditions, in accordance with international standardized procedures (Zucker 1985, OECD 2004).
Besides uncountable pharmacological uses, water extracts obtained from dry seeds have been used due to their excellent turbid water coagulation properties attributable to the presence of cationic electrolytes (Jahn et al. 1988, Gassenschmidt et al. 1995, Ndabigengesere et al. 1995).In Brazil, seed powder suspension has been introduced efficiently into the Northeast Region due to the tree good adaptation to arid areas as an attempt to improve people hygiene habits and life quality, help to reduce child mortality and collaborate with a sustainable development of the region (Morton 1991, Gerdes 1997, Ferreira et al. 2008).
The aim of this work was to assess the activity of the water extract of M. oleifera seeds (WEMOS) upon egg hatching and larval development of Ae. aegypti and its toxicity on Daphnia magna Straus, 1820 (Cladocera: Crustacea) and laboratory mammals.

PLANT MATERIAL AND EXTRACT PREPARATION
Moringa oleifera seeds were collected in Fortaleza, State of Ceará, Northeastern Brazil.A voucher specimen (34,591) was authenticated by Dr. Edson de Paula Nunes and deposited at Prisco Bezerra Herbarium (EAC), Departmento de Biologia, Universidade Federal do Ceará.
Recently collected mature dry seeds were dehulled manually and the kernel was crushed in a powder using a pestle and mortar.Seeds with any kind of visible damage were discarded.Distilled water was added to powdered seeds in the proportion of 1 seed (200 ± 8.2 mg) per 10 mL of distilled water according to the folk uses in Brazil (Gerdes 1997).The whole mixture was stirred for 60 min at room temperature (25 • C) using a magnetic stirrer (Fisatom, Brazil) and then filtered through Whatman No. 1 filter paper (Whatman Inc., Clinfton, NJ).All tests took into consideration the soluble solids concentration of the WEMOS, calculated for the mass present in the water extract, which showed a yield of 26% w/w (26 mg of soluble solids out of 100 mg of powdered seeds).

ANIMALS
Eggs and larvae of Ae. aegypti were obtained from NUVET -SESA (Núcleo de Controle de Endemias Transmissíveis por Vetores -Secretaria de Saúde do Estado do Ceará, Fortaleza, Ceará, Brazil), where a laboratory colony is maintained at 25-30 • C with 80-90% relative humidity under a photoperiod of 12:12 h and free of exposure to pathogens, insecticides or repellents.
Organisms of the D. magna species were cultivated in 2 L glass dishes with natural water and kept in an incubator at 20 ± 2 Total protein, ash and total lipid quantifications were performed according to AOAC (1990).Total carbohydrates (including dietary fiber) were determined by calculating the percentile difference from the total of the other constituents.

EVALUATION ON EGG HATCHING AND LARVAE OF Aedes aegypti
For egg hatching studies, filter paper containing dehydrated eggs were examined using an ordinary stereo microscope, cut in 1 cm 2 pieces with similar quantities of eggs and placed in glass tubes (20 mm × 100 mm) with 10 mL in three different concentrations of crude WEMOS (2600, 5200 and 26000 μg/mL).Control group was treated with distilled water.To evaluate larvicidal activity, tests were run according to methodology described by WHO (2005), with some modifications.Bioassays were performed with Ae. aegypti larvae on 3 rd instar which were collected with a Pasteur pipette, placed on filter paper for removal of excess water and transferred with a tiny brush into 150-mL disposable plastic cups (20 larvae per cup) containing 25 ml of crude WEMOS (375, 750, 1300, 2600 and 5200 μg/mL).Three independent experiments were run in quadruplicate and distilled water was used as negative control.Number of hatched eggs, larvae behavior and mortality were verified after 24 h of treatment at room temperature (25 • C) (Thangam andKathiresan 1991, Carvalho et al. 2003).Larvae were considered dead if they were immobile and unable to reach the water surface.

PARTIAL CHARACTERIZATION OF ACTIVE PRINCIPLE
As an attempt to evidence the nature of substances causing larvae mortality, the WEMOS was submitted to heat treatment (40-80 • C for 10 min) and thus bioassays using 3 rd instar larvae were carried out as described above.
Positive and negative controls were tested with WEMOS not subjected to heat and distilled water, respectively.
Filtered WEMOS was dialyzed extensively (cut off 12 kDa) for 24 h against distilled water at 25 • C in a proportion 1:2 sample: distilled water (non-dialyzable fraction -NDF) with four changes.The water used in the dialysis was frozen by liquid nitrogen, lyophilized and dissolved in distilled water to give 20 mg/mL (dialyzable fraction -DF).All experiments with the DF and NDF were performed with either fresh preparations or lyophilized material.Soon afterwards, the larvicidal capacity of both NDF and DF was evaluated as described above.

ACUTE TOXICITY UPON Daphnia magna
The tests were performed in glass bottles of 30 mL (4.0 cm in diameter and 3.5 cm in height), containing 10 mL of culture water and thirty neonate daphnids of 24 h in age exposed to increasing concentrations of the crude WEMOS (56, 130, 317 and 505μg/mL) that were established by preliminary immobility studies (OECD 2004).Feed was provided using a suspension of Scenedesmus subspicatus at the concentration of 5 × 10 6 cells per organism per day, fermented ration for fish, yeast and a vitamin complex of B1 (7 mg/L), B2 (7 mg)/L, B6 (5 mg/L), B12 (33 mg/L) and H (0.01 mg/L).Immobilized organisms which were incapable of swimming for 15 sec were counted after 24 h of exposure (Arauco et al. 2005).As a reference substance, it was utilized potassium dichromate (Wako Pure Chemical Industries Ltd.) with EC 50 value of 1.42μg/mL.

ACUTE TOXICITY UPON SWISS MICE
The acute toxicity evaluation was performed on male mice with 20-25 g and 6-8 weeks in age.WEMOS prepared just before injection was administered by intraperitoneal route at different doses (150, 250, 400, 550 and 700 mg/kg body weight).This route was chosen as an effort to avoid interfering factors such as absorption, pH of the stomach and intestines, enzymatic activity of the digestive tract and first-pass metabolism (Karalliedde et al. 2003).For each dose, eight animals were used and observed during the following 48 h for detection of any behavioral change, poisonous symptoms and death (Litchfield and Wilcoxon 1949).

SUBACUTE TOXICITY ON WISTAR RATS
Twenty healthy male rats weighing between 180 and 230 g were randomly divided into two groups (n = 10 each).Animals of the experimental group received filtered WEMOS ad libitum (5200 μg/mL of soluble solids) as the only source of drinking water for continuous 30 days while the control group received tap water.At the end of treatment, animals were anesthetized with halothane (Fluothane, Zeneca, São Paulo, Brazil) and blood samples were collected by retro-orbital puncture (Waynforth 1980) using heparin as anticoagulant.Plasma was separated by centrifuging at 2000× g for 10 min and kept at -70 • C until analysis.Liver subacute toxicity was evaluated by measuring enzyme activity of aspartate aminotransferase (AST), alanine aminotransferase (ALT) and alkaline phosphatase (ALP) after 4 weeks of treatment.All estimations were performed using enzymatic diagnostic reagent kits following manufacturer's instructions (Labtest, Brazil).
Immediately after blood collection, all rats were sacrificed by cervical dislocation.Kidneys, spleen, heart, intestines, stomach, thymus, lungs, liver and pancreas were dissected out, weighed and extensively lyophilized.Subsequently, they were weighted again to determine their dry relative weights.To verify the nitrogen incorporation, rat carcasses were dried at 80 • C for 48 h, pulverized with an electric grinder (Broun, Mexico) and submitted to nitrogen quantification (AOAC 1990).

STATISTICAL ANALYSES
The EC 50 , LC 50 and LD 50 values and their 95% confidence intervals were obtained using the Trimmed Sperman-Karber method (Hamilton et al. 1979).In order to determine differences between groups, data (means ± standard deviation) were compared by one-way analysis of variance (ANOVA) followed by Newman-Keuls (P < 0.05) using the GraphPad program (Intuitive Software for Science, San Diego, CA).

LARVICIDAL ACTIVITY
The crude WEMOS at 5200 μg/mL (soluble solids) caused 99.2 ± 2.9% mortality of 3 rd instar larvae within 24 h and toxic effects were still observed in lower con-centrations, showing an increasing progression toward larvae death in a dose-dependent manner, as described in Table I.Working with soluble solids, this extract revealed a LC 50 of 1260 μg/mL.The crude WEMOS was further fractionated and corresponding fractions were analyzed separately.After extensive dialysis for 24 h, the NDF of WEMOS was ineffective to cause larval death.On the other hand, concentration-dependent outcomes were also obtained with the DF of WEMOS, which caused larval mortality that ranged from 10.6 (at 1300 μg/mL) to 97.5% (at 5200 μg/mL).Although the highest concentration has been capable to kill nearly 100% larvae, deaths had begun at higher concentrations (at 1300 μg/mL) than those seen with crude WEMOS, suggesting that some activity had been lost after dialysis.The larvicidal activity was significantly diminished after heat treatment at 60, 70 and 80 • C/10 min, leading to larvae mortality rates of 70.0 ± 8.2, 60.0 ± 11.5 and 0%, respectively (P < 0.05).No deaths occurred in the negative control.Meanwhile, non-heated WEMOS showed 97.3 ± 2.0% mortality.In spite of the larvicidal activity, crude WEMOS showed no capacity to interfere on egg hatching (Ta-ble II) not even when tested at five-fold the concentration that caused 100% mortality (P > 0.05).However, none of the newly hatched larvae was able to reach its second stage, given that all died after hatching at all tested concentrations (P < 0.05) (Table II).

TOXICITY ASSAYS
The evaluation of the WEMOS on Daphnia magna showed, based on the r 2 values (0.99), a high linear relationship (Y = 0.19 X + 5.61) between immobility of the test organisms and concentrations of the WEMOS, which presented an EC 50 of 188.7μg/mL (Table III).
In the studies of acute toxicity on mice, no external manifestation of toxic syndrome was observed within 48 h after injection of a single dose of WEMOS 150 mg/kg body weight.Although no deaths occurred at doses of 250 and 400 mg/kg, some mice (7 out of 16) showed diarrhea after 24 h.On the other hand, deaths were noticed at 550 mg/kg (5 out of 8) and 700 mg/kg (100% mortality) between 10-36 h following inoculation.Subsequent statistic analyses calculated a LD 50 of 446.5 (418.3-476.6)mg/kg body weight.
As described in Table IV, the subacute treatment of rats with WEMOS given ad libitum for 30 days did not cause changes on the AST, ALT and ALP serum values when compared to the control group (P > 0.05).
Seed proximate composition (g/kg dry matter) shows high levels of total proteins (377.5 ± 1.90 g/kg), total lipids (363.2 ± 2.6 g/kg), ash (35.8 ± 2.6 g/kg) and total carbohydrates (223.5 g/kg).This protein level was probably reflected on the nitrogen incorporation by the WEMOS-treated animals (67.5 ± 2.5 g/100g body weight) in comparison with the control group (59.5 ± 3.0 g/100g body weight) (P < 0.05).Nevertheless, there was no statistical difference on body weight gain between oral WEMOS-treated (90.7 ± 16.0 g) and non-treated (89.0 ± 8.4 g) rats (P > 0.05) and neither mortality nor morbidity was recorded during the whole experiment.Most of the organs did not show alterations in their relative weights (Table V), though thymus of the treated group displayed increasing in both wet (0.08 ± 0.02 g) and dry (0.17 ± 0.03 g) relative weights when compared to the controls (0.05 ± 0.01 g and 0.11 ± 0.02 g for wet and dry relative weights, respectively) (P < 0.05).

DISCUSSION
The mosquito Aedes aegypti has domiciliary habits, hiding in dark and closed places, which leads to difficulties in its eradication (Gubler 1998).So, the most efficient way to control dengue resides in preventing the insect breeding through the use of larvicides (Consoli andOliveira 1994, WHO 1999).In this work, we showed that the crude WEMOS and its soluble dialyzable fraction (DF) separated by dialysis membrane (cut off 12 kDa) were toxic to the larvae.These results suggest the participation of low molecular mass compounds in the DF as responsible for these larvicidal effects on Ae. aegypti.Previously, it was confirmed the presence of proteins by the Bradford method and by tricine-sodium dodecyl sulfate polyacrylamide gel electrophoresis with apparent molecular mass ranging from 6 to 30 kDa in both crude WEMOS and DF (Ferreira 2004).These findings are in line with Gassenschmidt et al. (1995), Ndabigengesere et al. (1995) and Gomes (A.S.Gomes, unpublished data) who identified, by different techniques, proteins lower than 13 kDa in the seeds.To evidence facts to support this hypothesis, crude WEMOS was heat-treated prior to exposure to 3 r d instar larvae.WEMOS lost its toxic effects on larvae completely between 70-80 • C, suggesting being reasonable to consider that the protein fraction may be involved in the larvicidal activity.
We also demonstrated that WEMOS was not efficient to interfere in hatching of Ae. aegypti eggs in 24 h exposure.Indeed, after 48 h, the egg embryonic development of Ae. aegypti is usually completed, an event which can explain egg resistance during adverse conditions, and this turns the eggs the most resistant phase of its life cycle (WHO 1999).In order to evaluate environmental impacts of the extract, we performed some acute toxicological tests to understand its potential as a toxicant.Ecotoxicologic assessment was carried out on Daphnia magna, a cladoceran crustacean commonly used for determining the toxicity of pollutants and very recommended for representing aquatic invertebrates (OECD 2004).Moreover, it shows a defined sensitivity to reference substances and has laboratorial advantages in its utilization such as short life cycle, parthenogenesis reproduction and ease of handling (Tatarazako and Oda 2007).
According to the acute toxicity classes proposed by Zucker (1985), moringa seeds were practically nontoxic on daphnids, with a EC 50 of 188.7μg/mL.Recently, using laboratory-cultured Artemia nauplii as an alternative tool to preliminary assessment of WEMOS toxicity, we found similar results and a LC 50 value of 177.8μg/mL (Ferreira et al. 2007). Ali et al. (2004) have already demonstrated that M. oleifera seed extracts are toxic to the green microalga Scenedesmus obliquus.Additionally, the assay of acute toxicity on mice showed that WEMOS has an LD 50 (446.5mg/kg body weight) that is considered only moderately toxic when compared to toxicological human standards (Hodge and Sterner 1944).
Some populations in developing countries do not have access to treated water and so they store water for daily consumption, which facilitates mosquito oviposition and proliferation.Thus, to further assess the safety of moringa seeds, some physiological parameters were monitored in rats receiving WEMOS as the unique source of drinking water during 30 days.The ad libitum intake was in the range of 1300-1670 mg/kg/day and no alterations were observed in the hepatic enzymes (ALT, AST and ALP), which confirms our recent results showing the lack of hepatotoxicity in rats after the administration of WEMOS at 400 mg/kg/day for 30 days (Ferreira et al. 2007).The enzymes ALT, AST and ALP are commonly used as markers of hepatic damage and the degree and type of liver injuries can be evaluated based on the presence or absence of these specific enzymes in the bloodstream (Kumar et al. 2004).On the other hand, Bharali et al. (2003) reported that administration of the hydroalcoholic extract of M. oleifera drumsticks by oral route enhanced levels of some hepatic enzymes involved in detoxification of xenobiotic substances, such as carcinogens and plant venomous compounds.Thereafter, Gupta et al. (2005)  the capacity of moringa seeds to protect animals against oxidant stress caused by arsenic exposure.In truth, several researches attributed this oxidative protection of the seeds to the presence of compounds with antioxidant activity against free radicals (Geervani and Devi 1981, Siddhuraju andBecker 2003) that confer a promising chemo-preventive potential (Guevara et al. 1999, Bharali et al. 2003).Despite various reports have suggested hepatoprotective action of different parts of M. oleifera (Ruckmani et al. 1998, Fakurazi et al. 2008), further investigations must be done to confirm this property in the seeds.Organs of WEMOS-treated rats presented neither macroscopic alterations nor modifications on their relative weight, with exception of the thymus increase.Interestingly, we have seen previously that administration of WEMOS 400 mg/kg promotes spleen hypertrophy (Ferreira et al. 2007).These findings suggest that this extract probably has substances implied in the leukocyte activation and/or immunological amplification (Imboden 1988, Santos et al. 2005).On the other hand, contrasting results were obtained by Oliveira et al. (1999), who, in a 10 day-feeding trial substituting M. oleifera seed flour for egg-white standard protein, showed adverse effects such as decreased body weight and food intake and changes on internal organs, such as hyper-trophy of stomach, small intestine, liver, pancreas, kidneys, heart and lungs and atrophy of thymus and spleen.The seeds possess hemagglutinating activity, glucosinolates (65.5 μmol/g) and phytates (41 g/kg) (Makkar and Becker 1997, Oliveira et al. 1999, Santos et al. 2005) which may be responsible for the observed adverse effects.Nevertheless, the present work did not reveal toxic activity on the rats, suggested by observed normal growth and absence of body and organ weight changes.Quite the opposite, moringa group demonstrated significantly higher body nitrogen levels.
Besides low toxicity, the multiple purpose crude WEMOS has been presented to be advantageous when compared to inorganic or synthetic organic coagulants, being as efficient as aluminum salts for coagulation of raw water (Ndabigengesere and Narasiah 1998), dramatically decreasing clay and bacteria contents (Madsen et al. 1987, Broin et al. 2002, Ghebremichael et al. 2005).It brings advantages in relation to flocculant polymers which are associated with human pathologies, particularly health problems related to residual aluminum like Alzheimer's disease (Martyn et al. 1989) and promotion of carcinogenesis (Mallevialle et al. 1984).Now, this cleaning water property can be combined with its larvicidal action against Ae.aegypti, the dengue urban vector main and a highly domestic tropical mosquito that prefers to lay its eggs in artificial containers commonly found in and around human dwellings, such as flower vases, water storage containers and buckets that collect rainwater (for example, empty bottles, old automobile tires and trash in general) (Consoli and Oliveira 1994), all important places in producing large numbers of larvae which may be exterminated by WEMOS.
Our study clearly demonstrated that water extract of Moringa oleifera seeds have lethal action against Aedes aegypti larvae and low toxic effects on laboratorial animals, which is in agreement with the literature (Berger et al. 1984, Grabow et al. 1985, Ali et al. 2004, Ferreira et al. 2007).These advantages encourage the dispersion of M. oleifera tree around the world as well as the exploitation of its seed as a way to reduce the exorbitant costs of water treatment, mainly in developing countries and rural areas (Jahn et al. 1988).An additional benefit is that M. oleifera seed is available throughout the year, especially when the mosquito population is higher.The plants grow in nature without any extra care or cost and simple technology would be necessary to separate the most suitable fractions to be exploited as a possible chemical to be employed in mosquito control programs.Further investigations are in progress to identify the moringa larval-killing compound(s).
± 3.3 a, bWEMOS is expressed as soluble solids based on the proportion 1 seed: 10 mL(Gerdes 1997).Values are means ± S.D. of three independent experiments performed in quadruplicate.NDF -Non-dialyzable fraction; DF -Dialyzable fraction.a P < 0.05, compared to control by one-way analysis of variance (ANOVA) followed by Newman-Keuls test.b P < 0.05, compared to the immediately lower concentration by one-way analysis of variance (ANOVA) followed by Newman-Keuls test.

TABLE II Evaluation of of the water extract of Moringa oleifera seeds (WEMOS) on egg hatching of Aedes aegypti after 24 h of exposure. Distilled water was used as negative control.
(Gerdes 1997)erdes 1997).Values are means ± S.D. of three independent experiments performed in quadruplicate.a P < 0.05, compared to control by one-way analysis of variance (ANOVA) followed by Newman-Keuls test.

TABLE III Linear regression analysis of immobility tests performed on Daphnia magna organisms and the EC
(50−24h)value

TABLE IV Hepatic enzyme activity of aspartate aminotransferase (AST), alanine aminotransferase (ALT) and alkaline phosphatase (ALP) after the oral treatment of rats with water extract of Moringa oleifera seeds (WEMOS) ad libitum during 30 days. Control group received tap water.
Values are means ± S.D., n = 10 animals/group.

TABLE V Organ relative weights of rats treated with water extract of Moringa oleifera seeds (WEMOS) ad libitum during 30 days. Control group received tap water.
have describedValues are means ± S.D., n = 10 animals/group.a P < 0.05, compared to control by one-way analysis of variance (ANOVA) followed by Newman-Keuls test.