Chemical composition and insecticidal activity of plant essential oils from Benin against Anopheles gambiae (Giles)

Background Insecticide resistance in sub-Saharan Africa and especially in Benin is a major public health issue hindering the control of the malaria vectors. Each Anopheles species has developed a resistance to one or several classes of the insecticides currently in use in the field. Therefore, it is urgent to find alternative compounds to conquer the vector. In this study, the efficacies of essential oils of nine plant species, which are traditionally used to avoid mosquito bites in Benin, were investigated. Methods Essential oils of nine plant species were extracted by hydrodistillation, and their chemical compositions were identified by GC-MS. These oils were tested on susceptible “kisumu” and resistant “ladji-Cotonou” strains of Anopheles gambiae, following WHO test procedures for insecticide resistance monitoring in malaria vector mosquitoes. Results Different chemical compositions were obtained from the essential oils of the plant species. The major constituents identified were as follows: neral and geranial for Cymbopogon citratus, Z-carveol, E-p-mentha-1(7),8-dien-2-ol and E-p-mentha-2,8-dienol for Cymbopogon giganteus, piperitone for Cymbopogon schoenanthus, citronellal and citronellol for Eucalyptus citriodora, p-cymene, caryophyllene oxide and spathulenol for Eucalyptus tereticornis, 3-tetradecanone for Cochlospermum tinctorium and Cochlospermum planchonii, methyl salicylate for Securidaca longepedunculata and ascaridole for Chenopodium ambrosioides. The diagnostic dose was 0.77% for C. citratus, 2.80% for E. tereticornis, 3.37% for E. citriodora, 4.26% for C. ambrosioides, 5.48% for C. schoenanthus and 7.36% for C. giganteus. The highest diagnostic doses were obtained with S. longepedunculata (9.84%), C. tinctorium (11.56%) and C. planchonii (15.22%), compared to permethrin 0.75%. A. gambiae cotonou, which is resistant to pyrethroids, showed significant tolerance to essential oils from C. tinctorium and S. longepedunculata as expected but was highly susceptible to all the other essential oils at the diagnostic dose. Conclusions C. citratus, E. tereticornis, E. citriodora, C. ambrosioides and C. schoenanthus are potential promising plant sources for alternative compounds to pyrethroids, for the control of the Anopheles malaria vector in Benin. The efficacy of their essential oils is possibly based on their chemical compositions in which major and/or minor compounds have reported insecticidal activities on various pests and disease vectors such as Anopheles.


Background
Malaria is a life-threatening disease affecting 3.3 billion people worldwide with 80% of cases and 90% of deaths occurring in sub-Saharan Africa (SSA). Of those affected, children under the age of five and pregnant women are the most vulnerable [1]. The key factor for the reduction of human mortality is their protection against the malaria parasite vectors, the Anopheles mosquito, by the use of long-lasting insecticidal nets (LLIN) and application of indoor residual spraying (IRS) as recommended by the WHO [1]. Pyrethroids are the only insecticides recommended by WHO for the treatment of bednets [1] while in West Africa, Anopheles gambiae Giles is the major vector of malaria [2]. This mosquito species is now subjected to selection pressure, and has become resistant to dichlorodiphenyltrichloroethane (DDT) and pyrethroids because of the use of these insecticides both for agricultural and public health purposes [2,3]. During the last decade, the emergence of resistance to synthetic insecticides in Anopheles populations has been widely described in many African countries such as Benin [4][5][6][7][8][9][10], Ivory Coast [2,11], Niger [12], Burkina Faso [3,13,14], Mali [15], Nigeria [16], Kenya [17], Cameroun [18][19][20], Zanzibar [21], Uganda [22], Equatorial Guinea [23], and Ghana [24]. Because of the widespread occurrence of this resistance, there is a reduction of the efficacy of synthetic insecticides used in the field. Therefore, it is pertinent to explore the pesticidal activity of natural products [25] such as essential oils. Indeed the use of essential oils has been recognized as a potential alternative in the control of vectors of mosquito-borne diseases [26][27][28][29][30] such as A. gambiae. A lot of research on the pesticidal activities of essential oils has been conducted and has proven that essential oils could be considered as potent bioactive compounds against various pests and mosquitoes [28,[31][32][33][34].
The objective of the current study was to investigate the chemical composition and the insecticidal activities of essential oils of nine aromatic plants traditionally used in Benin as a natural method of protection against A. gambiae, in order to confirm the traditional knowledge of Benin's population and to find new valuable sources of active molecules against this malaria vector. Indeed these plant species have been reported to possess repellent and insecticidal activities against various mosquitoes, stored product beetles and other pests such as Anopheles species [30,35], Culex quinquefasciatus [36], Callosobruchus species [37,38], Sitophilus species [39,40] and Tribolium castaneum [41]. These plants include For this purpose, the standard WHO susceptibility test has been used with various concentrations of essential oils under laboratory conditions [42].

Plant material
Nine plant species belonging to five botanical families were collected in several regions of Benin. Collected plant material was dried, free from light, at room temperature. All these plants have been the subject of classification and botanic description at the National Herbarium of the University of Abomey-Calavi, Benin [43]. Locations of collection, organs extracted and nature of soil are summarized in Table 1.

Essential oils
Essential oils from all plants were extracted by hydrodistillation using a Clevenger-type apparatus for two to four hours until essential oils were extracted completely from the organs. The extracted oils were dried over anhydrous magnesium sulfate and stored at 4°C, away from UV rays, before use. Each essential oil was diluted tenfold with diethyl ether. One milliliter of each diluted solution was charged into a sampler flask for GC-MS analysis.
Chemical analyses of essential oils by gas chromatography coupled with mass spectrometry The GC-MS analysis of the essential oils was performed on an Agilent 6890 GC Plus automatic sampler system, coupled to a quadruple mass spectrometer 5973 MSD (Agilent Technologies, Diegem, Belgium) and equipped with a capillary HP-5MS column fused with silica (length: 30 m; diameter: 0.25 mm, film thickness: 0.25 microns) in the split mode 1:100. The oven temperature was programmed at 60°C for 3 min and to 350°C at a rate of 5°C/min. The injector was kept at 250°C programmed with a rate of 10°C/s. Helium was used as carrier gas at a flow rate of 1 ml/min. All analyses were performed at constant flow. The mass detector conditions were: transfer line temperature 260°C; ionization mode electron impact: 70 eV. The identification of sample compounds was carried out in single runs. The Kovats retention indices were calculated for all volatile constituents using a series of n-alkanes C 7 -C 17 [44]. Quantification of each compound was performed using percentage peak area calculations. The identification of oil volatile compounds was done by comparing their retention indices with those of reference compounds in the literature and confirmed by GC-MS by comparison of their mass spectra with those of reference compounds [44][45][46][47]. The relative concentration of each compound in the essential oil was quantified according to the peak area integrated by the analysis program (Chemstation data analysis).

Mosquito collection
Larvae of the resistant strain of A. gambiae were collected three times a week, from March 2012 to June 2012, from the natural breeding site in "Ladji" in Cotonou. Ladji is a neighborhood of the outskirts of Cotonou, the capital of Benin. Larvae were brought and reared in the insectary of the Cotonou Research Center of Entomology (CREC). Larvae were fed with honey juice and the adults were placed in cages and fed in the same way. Emerging adult female mosquitoes of the resistant strain were used to carry out the susceptibility tests, whereas a susceptible strain of A. gambiae "Kisumu" originating from Kenya and maintained at the insectary, were used as a reference.

Bioassay on adult mosquitoes
Susceptibility tests were carried out using WHO insecticide susceptibility test-kits and standard procedures [42].
Four replicates of batches of 20-25 unfed females, aged 2-5 days old were exposed to filter papers (Whatman N°1) 12 cm × 15 cm, impregnated with 2 ml of various doses of the essential oil, diluted in acetone. The susceptible strain A. gambiae "Kisumu" was used as reference to determine the diagnostic doses. The filter papers of the control holding tubes were impregnated with acetone only. Tests were carried out at 25°C (± 2°C) and 70-80% relative humidity. Six doses were tested for each essential oil: 0.25% (w/v), 0.50% (w/v), 1% (w/v), 2% (w/v), 4% (w/v) and 8% (w/v). The number of mosquitoes knocked down was recorded every 5 min. Permethrin 0.75% was used as reference insecticide. Permethrin impregnated papers were obtained from the WHO reference center of "the Vector Control Research Unit of The University Sains Malaysia". After the exposure time, mosquitoes were transferred to holding tubes and were fed with 10% honey juice for 24 hours. Subsequently the mortality was recorded.
Data were plotted to determine the diagnostic doses to which the resistant strain from Ladji was exposed.

Data analysis for bioassay
Times at which 50% or 95% of mosquitoes fell down on their back or on their side, i.e. knockdown time (KDT 50 and KDT 95 ) was calculated by means of the log timeprobit using SPSS 17.0 software. The relation between the KDT, the mortality and the doses were assessed by probit regression. The diagnostic concentration which is the double of the minimum concentration at which 100% of the susceptible strain died [48] was also taken into account. All the results obtained for KDT (KDT 50 and KDT 95 ) and lethal doses (LC 50 and LC 99 ), were expressed with 95% confidence limits.

Results and discussion
Chemical composition of essential oils Chemical compositions and essential oil yields, expressed as oil wt./wt. of dried organ extracted, showed a large variation (Tables 2, 3, 4, 5 and 6). These yields varied from 0.2% for Cochlospermum species to 4.6% for E. citriodora.

C. planchonii
The essential oil of C. planchonii was extracted with a yield of 0.20% (w/w) lower than in the tubercle from Burkina Faso (0.30%) [66]. In the current study the composition of    [66]. The sample characterized by Ouattara et al. [67] was similar to the previous one with major components involving 3-tetradecanone, 3-tetradecenone, dodecyl acetate, tetradecyl acetate, 2-tridecanone and β-elemene.

S. longepedunculata
The essential oil of S. longepedunculata was extracted with a yield of 0.7% (w/w). This yield is higher than the result obtained (0.30%-0.52%) by Alitonou et al. [68] in Benin and Adebayo et al. in Nigeria [69]. The S. longepedunculata essential oil was characterized by only one major constituent, namely methyl salicylate (99.4%). Indeed Nebie et al. [70] has shown that the essential oil of S. longepedunculata from Burkina Faso contains only one compound which is methyl salicylate. The same compound was also found in the methanol extract of S. longepedunculata from Ghana [71] and in essential oil extracted in Nigeria and Benin [68,69].

Adult bioassay on susceptible strains of Anopheles gambiae
The resistant status of mosquito samples was determined according to the WHO criteria summarized as follows [42]: -98-100% mortality indicates susceptibility of the mosquito strain to the tested essential oil -Mortality less than 98% is suggestive of the existence of a resistance to the essential oil that needs to be confirmed by two additional tests -Mortality less than 90% suggests resistance in the mosquito population

KDT 50 and KDT 95
KDT 50 and KDT 95 calculated with 95% confidence limits are summarized in Table 7. The lowest KDT 50 and KDT 95 values were obtained with C. citratus and are

Mortality rates
Mortality rates to different essential oils are shown in Table 7. At 0.25%, the mortality rate of A. gambiae "Kisumu" varies from 0.0% to 72.5%. However, the mortality rates increased with the dosage. At 0.50%, mortality has reached 100% for C. citratus, whereas it was at 5.6% for S. longepedunculata. At 1%, mortality was still 100% for C. citratus whereas it was 6.7% for C. schoenanthus. At 2%, the mortality rate was 29.6% for S. longepedunculata and 100% for C. citratus, E. citriodora, E. tereticornis and C. ambrosioides. At 4% and 8%, the mortality rates varied from 79.2% to 100% for C. tinctorium and C. planchonii. To summarize these results, the susceptibility tests on the sensitive strain A. gambiae "kisumu" have demonstrated its susceptibility status on essential oils tested. The most efficient essential oil was C. citratus at 0.50%, followed by E. tereticornis at 1%, E. citriodora and C. ambrosioides at 2%, C. schoenanthus, C. giganteus, C. planchonii and S. longepedunculata at 4%.

Diagnostic concentrations
The diagnostic concentration is defined as twice the lethal concentration (LC) for 99% mortality (LC 99 ) on sensitive strains [42]. Lethal concentration for 50% mortality (LC 50 ), lethal concentration for 99% mortality (LC 99 ) expressed with 95% confidence limits and diagnostic concentrations for all essential oils tested are summarized in Table 8.
All diagnostic doses obtained above were tested on the resistant strain of A. gambiae and results obtained were        summarized in Table 9. Concerning the diagnostic doses tested, the lowest KDT 50 and KDT 95 were observed with C. schoenanthus (2.58 min, 3.84 min), followed by S. longepedunculata, C. giganteus, E. tereticornis, C. ambrosioides and E. citriodora for which the KDT 50 and KDT 95 were lower than 5 min and 6 min, respectively. A moderate knock down effect (KDT 50 ≥ 12 min and KDT 95 ≥ 20 min) was observed with C. planchonii, C. tinctorium and C. citratus. The KDT 50 observed was 6 to 35 fold lower than for permethrin, the positive control, and the KDT 95 was 5 to 38 fold lower than permethrin. This observation demonstrates the promising insecticidal properties of these plants species on the A. gambiae resistant strain used.
Apart from the essential oil from C. tinctorium and S. longepedunculata for which resistance was suspected because the mortality was less than 97%, the resistant strain of A. gambiae was susceptible to all essential oils at diagnostic doses tested ( Table 9). The resistance to permethrin in southern Benin has been demonstrated and was explained by the massive use of DDT in house spraying and agriculture, during the WHO malaria eradication program, which has permitted the apparition and the increase of kdr mutation in A. gambiae populations in Benin [7,10].
We have noticed that the knock-down times, the mortalities, and the resistance status of A. gambiae did not only depend on the values of doses used but mainly on the chemical composition of essential oils used.
Larvicidal activity of C. citratus essential oil on Aedes aegypti at lower concentrations (LC 50 = 0.28 μl/ml and LC 90 = 0.56 μl/ml) and its major component citral (neral and geranial) has been demonstrated by Freitas et al. [73]. C. citratus has also demonstrated a very good repellency against A. aegypti [30]. C. citratus has shown in the current study the best insecticidal activity against A. gambiae, but its KDT 50 and KDT 90 were higher than  some of the other essential oils. The same conclusion has been found by Phasomkusolsil et al. [33] when the essential oil was tested on A. aegypti, C. quinquefasciatus and Anopheles dirus. The topical application of C. citratus showed high toxicity against Sitophilus oryzae [50]. C. citratus essential oil and citral have been shown to be potential anti-Leishmania agents [51]. Insecticidal and larvicidal activities of C. citratus have been attributed to citral that has demonstrated 100% mortality against A. aegypti, at 2.5 μl/ml with LC 50 = 0.02 μl/ml and LC 90 = 0.28 μl/ml respectively [73], and its repellent effect at 15% (v/v) is comparable to 5% C. citratus essential oil [74]. In conclusion, the presence of geranial and neral is potentially responsible for the insecticidal activity of the essential oil of C. citratus, as demonstrated in the current work. C. giganteus essential oil, rich in limonene and (Z and E)-p-mentha-1(7),8-dien-2-ol such as the current sample from Benin, has proven to be toxic by fumigation to Callosobruchus species [28]. The insecticidal properties noticed in this study against A. gambiae might also be attributed to these main compounds.
Several studies on essential oils, rich in piperitone such as the essential oil from C. schoenanthus have demonstrated insecticidal activity against some pests. This is the case for Cymbopogon olivieri, which demonstrated good larvicidal activity against A. stephensi with LD 50 = 321.9 mg/l [75]. Exposure of Callosobruchus maculatus to C. schoenanthus essential oil for 24 hours resulted in 90% of adult mortality at 6.7 μl/l [56]. Piperitone has been reported to be powerful against ants of Crematogaster spp [76]. Adults, newly laid eggs and neonate larvae of C. maculatus with an LC 50 recorded at 1.6 ± 0.14 μl/l and all eggs were aborted at 6.7 μl/l with a total inhibition of the neonate larvae penetration in the seed [38]. Also in 2008, piperitone isolated from Artemisia judaica L., was studied against the third larvae of Spodoptera littoralis (Boisd) and has revealed a high insecticidal and antifeedant activity against this pathogen, with a LD 50 = 0.68 μg/larvae [77]. Following this previous research we could attribute the insecticidal activity of C. schoenanthus to its main compound, i.e. piperitone.
The essential oil from E. citriodora, rich in citronellal, citronellol and isopulegol, has been revealed to be repellent against Tribolium castaneum at 0.084 ml/l and was more active than the commercial product IR3535 at 0.686 ml/l [41]. The insecticidal activity of E. citriodora has been demonstrated at 5 mg/ml against Lutzomyia longipalpis [27]. E. citriodora has also demonstrated larvicidal activity against C. quinquefasciatus [36] and acaricidal activity against larvae of Amblyomma cajennense and Anocentor nitens [78]. The presence of citronellal, citronellol and isopulegol could well explain the insecticidal activity of E. citriodora against A. gambiae.
The essential oil from E. tereticornis leaf extract has shown a larvicidal activity against A. stephensi at 160 ppm which has provoked 100% of oviposition deterrence [35]. The sensitivity of adults of A. aegypti has been shown, resulting from the presence of 1,8-cineole, α-pinene and p-cymene and is correlated to the amount of 1,8-cineole in the extract [26,60]. The insecticidal activity of the essential oil of E. tereticornis observed in the current work, might be explained by the presence of one of its major components (p-cymene) but also by a minor compound (1,8-cineole), which both have demonstrated insecticidal activity.
In essential oils from Cochlospermum species one minor compound (2-tridecanone) has been found to have insect repellent properties. Indeed, 2-tridecanone has demonstrated repellent activity against the granary weevil Sitophilus granarius and S. zeamais at 100 ppm and 500 ppm on wheat [79]. Its repellent activity was confirmed against ticks since 0.63 mg/cm 2 was repellent to 87% of Amblyomma americanum after 12 hours and to 72% of Dermacentor variabilis after 15 hours [80]. The weak insecticidal activity of the essential oils of these two Cochlospermum species could be due to the low abundancy of 2-tridecanone.
Root powder, the methanol extract, and the main volatile component of S. longepedunculata (methyl salicylate) have proven to exhibit repellent and toxic effects against S. zeamais. In the same study, methyl salicylate has demonstrated a dose dependent fumigant effect with an LD 100 of 60 μl in a 1-l container after 24 hours exposure on S. zeamais, Rhyzopertha dominica and Prostephanus truncates and after 6 days exposure, 100% mortality could be recorded with 30 μl in a 1-l container [31].
The C. ambrosioides essential oil has demonstrated a larvicidal activity against A. arabiensis and A. aegypti after 24 hours exposure with LC 50 and LC 90 equal to 17.5 ppm and 33.2 ppm for A. arabiensis and 9.1 ppm and 14.3 ppm for A. aegypti under laboratory conditions [81]. Contact and fumigant toxicity of isolated compounds from this plant species have shown that ascaridole (LC 50 = 0.84 mg/l) followed by isoascaridole (LC 50 = 2.45 mg/l) were the most efficient insecticidal compounds by fumigation and contact with LC 50 = 0.86 mg/l (ascaridole) and 2.16 mg/l (isoascaridole). The crude oil was less active with LC 50 = 3.08 mg/l by fumigation and 2.12 mg/l by contact [39]. The insecticidal activity of C. ambrosioides, noticed in the study, might be explained by the presence of ascaridole and isoascaridole, which were among its major constituents.

Conclusions
The current study has dealt with the insecticidal properties of essential oils of nine plant species traditionally used in Benin for their repellency against A. gambiae bites. This research has shown that the essential oils from all the plant species studied, have insecticidal properties against this vector of malaria. The most promising was C. citratus followed in order of effectiveness by E. tereticornis, E. citriodora, C. ambrosioides, C. schoenanthus, C. giganteus and C. planchonii. The chemical composition of each plant essential oil has been elucidated by GC-MS and correlated with the insecticidal properties of these plant species. To our knowledge, it was the first time that diagnostic doses of essential oils on A. gambiae were determined, using the WHO susceptibility test protocol. These doses were presented in %, mg/ml and mg/cm 2 to facilitate further research on these plant species. KDT 50 and KDT 95 , LC 50 and LC 99 and results obtained have proven that all essential oils from these plant species are more effective against the resistant strain of A. gambiae than permethrin at the diagnostic doses tested. C. citratus, E. tereticornis, E. citriodora and C. ambrosioides as well as essential oil isolated components, such as citral, piperitone, 1,8-cineole, citronellal, 2-tridecanone, methyl salicylate, which possess demonstrated insecticidal properties, may be included in malaria vector control programs. These plants, occurring in the natural environment of local populations, could be obtained at lower cost and represent today a valuable source of bioactive compounds for the protection of the population against malaria.