Evaluation of the Activity of the Essential Oil from an Ornamental Flower against Aedes aegypti: Electrophysiology, Molecular Dynamics and Behavioral Assays

Dengue fever has spread worldwide and affects millions of people every year in tropical and subtropical regions of Africa, Asia, Europe and America. Since there is no effective vaccine against the dengue virus, prevention of disease transmission depends entirely on regulating the vector (Aedes aegypti) or interrupting human-vector contact. The aim of this study was to assess the oviposition deterrent activity of essential oils of three cultivars of torch ginger (Etlingera elatior, Zingiberaceae) against the dengue mosquito. Analysis of the oils by gas chromatography (GC)—mass spectrometry revealed the presence of 43 constituents, of which α-pinene, dodecanal and n-dodecanol were the major components in all cultivars. Solutions containing 100 ppm of the oils exhibited oviposition deterrent activities against gravid Ae. aegypti females. GC analysis with electroantennographic detection indicated that the oil constituents n-decanol, 2-undecanone, undecanal, dodecanal, trans-caryophyllene, (E)-β-farnesene, α-humulene, n-dodecanol, isodaucene and dodecanoic acid were able to trigger antennal depolarization in Ae. aegypti females. Bioassays confirmed that solutions containing 50 ppm of n-dodecanol or dodecanal exhibited oviposition deterrent activities, while a solution containing the alcohol and aldehyde in admixture at concentrations representative of the oil presented an activity similar to that of the 100 ppm oil solution. Docking and molecular dynamics simulations verified that the interaction energies of the long-chain oil components and Ae. aegypti odorant binding protein 1 were quite favorable, indicating that the protein is a possible oviposition deterrent receptor in the antenna of Ae. aegypti.

Introduction docking and evaluation of system dynamics, have been employed in order to address this issue [23,24].
Numerous plant-derived substances have been screened in order to find natural compounds that could be applied in the control of Ae. aegypti. Such studies have focused mainly on the evaluation of larvicidal and insecticidal properties, but repellent, ovicidal, pupicidal and oviposition deterrent activities have also been assessed [6,[25][26][27][28]. The aims of the present work were: (i) to evaluate the potential deterrent effect of the essential oil of the ornamental flower torch ginger (Etlingera elatior) against Ae. aegypti; (ii) to identify, through gas chromatographic (GC) analysis with electroantennographic detection (EAD), the specific oil constituents responsible for EAD activity; and (iii) to confirm the presence of receptors for these constituents on the antennae of female mosquitoes.

Chemicals
All chemicals and solvents used in the study were of Analytical Grade or of higher purity. Dodecanal and n-dodecanol were purchased from Sigma-Aldrich (Gillingham, UK) and were used as received.

Maintenance of Ae. aegypti
A colony of Ae. aegypti was maintained in the laboratory at 28 ± 1°C under a relative humidity of 70 ± 5% and a photoperiod of 14L:10D. Adult mosquitoes were maintained on a 10% sucrose solution, while females were also blood fed on pigeons (1h) four days after emergence. The sucrose solution was taken off the cage a day prior to the blood feeding. This assay was previously authorized by the Ethical Committee of the Federal University of Pernambuco.

Preparation of essential oil
Floral bracts (ca. 500 g of each cultivar) were comminuted in a blender and hydrodistilled for 3 h in a Clevenger-type apparatus. The essential oil layer was separated, dried over anhydrous sodium sulfate and stored in a hermetically sealed amber-glass vial at -5°C until required for assay. The yield of oil was reported as the quotient of the mass of oil collected and the fresh weight of plant material extracted [29].
Analysis of essential oil by gas chromatography-mass spectrometry (GC-MS) Essential oil constituents were identified by GC-MS analysis performed on an Agilent Technologies (Palo Alto, CA, USA) 5975C Series GC/MSD quadrupole instrument equipped with an Agilent J&W non-polar DB-5 fused silica capillary column (30 m × 0.25 mm i.d.; film thickness 0.25 μm). The analytical conditions were: sample (1 μL) injected in split mode (50:1) with injector temperature at 250°C; oven temperature held initially at 40°C for 2 min, then increased to 230°C at 4°C/min and held at 230°C for 5 min; helium carrier gas flow maintained at 1 mL/ min at a constant pressure of 7.0 psi; mass selective detector source and quadrupole temperatures set to 230°C and 150°C, respectively; MS obtained at 70 eV and recorded in the range 35-350 m/z at 1.0 scan/s. Individual components of the essential oil were tentatively identified by comparison of retention indices, obtained by co-injection of sample with C 9 -C 30 linear hydrocarbons and calculated according to the Van den Dool and Kratz equation [30], [29] with those reported in the literature. The MS data acquired for each component were matched with those stored in the mass spectral library of the GC-MS system (MassFinder 4, NIST08 and Wiley Registry™ 9th Edition) and with published spectra [31] in order to confirm identity.

Electrophysiological analysis
Electrophysiological analyses were performed with the aim of detecting constituents of the essential oil of E. elatior with the potential to be perceived by Ae. aegypti females. The instrumentation comprised a Thermo Scientific (Milan, Italy) Trace™ Ultra GC equipped with a flame ionization detector (FID), a Syntech (Kirchzarten, Germany) EAD with heated transfer line and two-channel universal serial bus acquisition controller, and a VICI Metronics (Poulsbo, WA, USA) VB-5-ValcoBond 1 capillary column (30 m x 0.25 mm i.d.; 0.25 μm film thickness) [32]. The column outlet was coupled to two lengths of deactivated capillary (40 cm x 0.25 mm i.d.) via an SGE Analytical Science (Trajan, Melbourne, VIC, Australia) splitter tee. One capillary led to the FID while the other was routed outside the GC oven and into a glass tube where the effluent was mixed with a clean and humidified airflow and directed over the head preparation. The flow of helium carrier gas through the column was maintained at 1 mL/ min by electronic control and nitrogen make-up gas was added immediately before the splitter. The injector temperature was set at 200°C, and an aliquot (0.5 μl) of a headspace sample of essential oil was injected in splitless mode with the oven temperature of 60°C. After 1 min, the injector split valve was opened and the oven temperature was increased to 200°C at 7°C/min and held at 200°C for 5 min.
Ten female mosquitoes (10 to 20 days old) were used in the analyses three days after a blood meal. The head of each individual was excised from the thorax with a scalpel and the tips of the antennae were cut. The base of the head and the tip of both antennae were then mounted between two glass capillary electrodes filled with insect ringer solution (8.0 g/L sodium chloride, 0.4 g/L potassium chloride, 0.4 g/L calcium chloride), and two silver wire electrodes were inserted into each capillary, thereby closing the electrical circuit with the head preparation. An essential oil constituent was considered EAD-active when it elicited a depolarization response in at least four individual head preparations.

Oviposition deterrent activity assay
In order to evaluate the oviposition deterrent activity of the essential oils, female Ae. aegypti mosquitoes were subjected to standard choice assays. Test solutions containing 100 ppm of essential oil or 50 ppm of n-dodecanol or dodecanal were prepared by dissolving 0.02 g of oil or 0.01 g of the standard compound in six drops of Tween 1 80 and diluting to a final volume of 200 mL with water. Test solutions were diluted further with water where necessary. Negative control solutions were prepared in the same manner but without oil or standard compound. Two disposable cups were placed at diagonally opposite corners of a cage measuring 33 × 21 × 30 cm, with one cup containing 25 mL of test solution and the other 25 mL of negative control solution. For paired assays, one cup contained 100 ppm of oil solution and the other a mixture of the two standard compounds in amounts equivalent to those present in the oil solution, i.e. n-dodecanol (25 ppm) and dodecanal (50 ppm) Filter papers were placed on the internal surface of each cup in order to provide support for oviposition. Ten gravid insects were transferred to the cage, which was then maintained at 28 ± 1°C and 70 ± 5% relative humidity for 16 h in the dark. Oviposition response was determined at the end of the assay period by counting the numbers of eggs laid on each of the filter papers. Each assay was replicated eight times and the mean values obtained for each of the test samples were compared using Student's t-test at an alpha level of 0.05 [33].
Molecular docking and dynamics of constituents of E. elatior oil to Ae. aegypti OBP 1 Molecular formulae and initial structural coordinates of the long chain components present in the essential oils of E. elatior were obtained with the aid of BIOVIA Draw software (Biovia, San Diego, CA, USA; http://accelrys.com/products/informatics/cheminformatics/draw/) and VEGA ZZ molecular modeling toolkit (http://nova.disfarm.unimi.it/cms/index.php?Software_ projects:VEGA_ZZ) [34]. Tripos mol2 format (http://www.tripos.com/data/support/mol2.pdf) files with coordinates were submitted to the antechamber module [35] of the AMBER package [36] in order to estimate AM1-BCC charges [37] and assign general AMBER force field (GAFF) atom types [38], each of which were subsequently inspected individually. Structures were energy minimized in a periodical TIP3 water box with edges at least 12 Å from any nonwater atoms using the sander module of the AMBER package [36]. Docking was performed using the Autodock Vina program (http://vina.scripps.edu/index.html) [39] in which the Ae. aegypti odorant binding protein 1 [20] (AaegOBP1, PDB ID: 3K1E) was the receptor, immersed in a box with dimensions 120 × 120 × 120 Å. Sites for molecular interaction were searched throughout the whole protein surface in blind docking mode with search exhaustiveness set to 128. Molecular dynamics simulations were applied to the five best poses for each ligand in order to estimate binding energy. In these calculations, the GAFF force field [38] was applied to the ligands, the Amber ff99SB protein force field [40] was applied to the protein receptor and the TIP3 model was used for water molecules [41]. Initially, each docked complex was immersed in a water box with edges at least 8 Å from any non-water atoms and sufficient counter-ions were put in place of solvent molecules in order to neutralize the total charge of the system. Subsequently, the following steps were carried out: i) minimization of system energy, with strong restraints on the receptor + ligand positions, until convergence; ii) minimization of complete system until convergence; iii) heating of system from 0 to 300 K over 50 ps with weak restraints on the receptor + ligand positions; iv) short dynamics for 50 ps with weak restraints on the receptor + ligand positions; and v) exploration dynamics for 500 ps. In all cases, a cut-off of 12 Å was applied to all non-bonded interactions. One thousand snapshots of each trajectory were taken at equal intervals and one hundered of them were subjected to MM/ GBSA methodology [42] in order to estimate the binding free energy ΔG.

Results and Discussion
The yields of essential oils obtained by hydrodistillation of inflorescences of red torch, pink torch and porcelain cultivars of E. elatior were 0.094, 0.052 and 0.049% (w/w), respectively. Similar oil yields have been reported previously from inflorescences of E. elatior [43] and from inflorescences of another member of the Zingiberaceae, namely Alpinia purpurata [44]. Fortythree components were identified in E. elatior oil by GC-MS analysis (Table 1), of which fifteen were present in the oils of all three cultivars, three were detected only in the red torch cultivar, eleven in the pink torch cultivar and two in the porcelain cultivar.
The volatiles of the inflorescences of E. elatior were rich in long chain alcohols and aldehydes and contained a range of mono-and sesquiterpenes. Qualitatively, the oils of the three cultivars were not substantially different, but there were significant variations in the relative proportions of the constituents even though dodecanal, n-dodecanol and α-pinene were the major components in all cases. The overall oil composition shown in Table 1 is in accordance with that described by Zoghbi and Andrade [45] for this species, although these authors did not specify the cultivar employed and the specimens were collected in the northern region of Brazil. In contrast, Jaafar et al. [46] reported that 1,1-dodecanediol diacetate, cyclododecane and α-pinene were the major constituents of E. elatior collected in Penang, Malaysia. This difference could be related to dissimilar edaphic conditions or to variations in the populations studied. Relationships between such conditions and oil yield or composition have been described for the essential oils of various species including Salvia officinalis [47], Cymbopogon winterianus [48] and Coriandrum sativum [49]. We have reported previously that the ornamental flowers of two cultivars of the Zingiberaceous species A. purpurata also presented differences in their essential oils [44]. As shown in Fig 1, the essential oils of E. elatior cultivars showed oviposition deterrent effects at 100 ppm (P < 0.05) against Ae. aegypti in that the numbers of eggs laid in cups containing oil solutions were significantly lower (30% or less) than those oviposited in control cups. The choice of oviposition site by a gravid female mosquito is determined by several factors. Potential sites are initially identified on the basis of visual and olfactory cues [16], following which short-range cues become increasingly important in subsequent selection. Such cues include temperature and chemical signals received by contact chemoreceptors distributed along the body of the mosquito [16]. Electrophysiological studies on Ae. aegypti have revealed that, while the blood meal is being digested, neurons susceptible to host-produced cues, such as lactic acid, become less sensitive, while neurons susceptible to oviposition site attractants, such as methyl butyrate, become more sensitive [50]. Accordingly, when oviposition deterrents are detected, few, if any, eggs are laid at that site [16]. In the present study, eggs laid in cups containing oil solution were placed preferentially on the outside of the support, thereby signifying the reluctance of females to lay eggs adjacent to the solution or the source of volatiles.
Few reports are available concerning the effectiveness of essential oils, or their individual components, as mosquito oviposition deterrents [27,[51][52][53][54]. However, the oviposition  activities of the oils of E. elatior against Ae. aegypti established in the present study are similar to those previously reported for two Zingiberaceous species, namely A. purpurata [44] and A. galanga [55]. Regarding the activities of individual constituents of the oil, it is important to note that the process of isolation and purification of specific components can be time consuming and expensive by virtue of the complex nature of the oil and the low concentrations of many of the constituents. For this reason, the development of techniques that would allow active components to be identified without prior isolation is of considerable importance [56].
In the present study, we have integrated for the first time the technique of GC-EAD with bioassays in order to identify the individual compounds within a complex oil that are responsible for oviposition deterrent effects against Ae. aegypti. The essential oil from the red torch cultivar of E. elatior was selected for GC-EAD analysis since it presented the highest oviposition deterrent activity. Ten components of the oil triggered antennal depolarization in Ae. aegypti females, and these were identified as n-decanol, 2-undecanone, undecanal, dodecanal, transcaryophyllene, (E)-β-Farnesene, α-humulene, n-dodecanol, isodaucene and dodecanoic acid (Fig 2). However, dodecanal and n-dodecanol elicited the most pronounced antennal depolarization and, since these two compounds accounted for almost 80% of the total oil content, they were submitted to bioassay in order to test their significance as oviposition deterrents.
Choice assays performed with solutions containing commercial standards of dodecanal and n-dodecanol clearly showed that both compounds possessed oviposition deterrent activity ( Fig  3). Thus, in comparison with the controls, gravid females laid fewer eggs in cups containing 50 ppm dodecanal (28.5% of total laid) or 50 ppm n-dodecanol (23.9% of total laid). However, in order to confirm the role of these components in the overall deterrent activity of the essential oil, paired tests were performed in which one cup was filled with a solution containing 100 ppm of oil while the second held a mixed solution containing dodecanal (50 ppm) and ndodecanol (25 ppm), such concentrations representing their respective proportions in the oil of red torch cultivar. No significant differences were observed between the numbers of eggs laid in the cups containing the oil and the standard compounds in admixture (Fig 3). These findings provide strong evidence that the two components of the essential oil are mainly responsible for its oviposition deterrent activity.   17,18,20,[25][26][27][28][29]33 and 38 were identified as n-decanol, 2-undecanone, undecanal, n-dodecanal, trans-caryophyllene, (E)-β-Farnesene, α-humulene, n-dodecanol, isodaucene and dodecanoic acid, respectively (see Table 1).
doi:10.1371/journal.pone.0150008.g002 Previous studies involving the GC-EAD technique, but without associated bioassays, have demonstrated that a range of compounds are capable of eliciting responses in the antennae of mosquitoes. Thus, Campbell et al. [56] reported that a number of mono-and sesquiterpenes, including trans-caryophyllene, α-terpineol, β-pinene, germacrene-D, limonene and α-zingiberene, present in essential oils were able to stimulate the antennae of Ae. aegypti. Additionally, the sesquiterpenes α-curcumene, β-sesquiphellandrene, zingiberene and β-bisabolene from the essential oil of ginger (Zingiber officinale, Zingiberaceae) [57], trans-caryophyllene from Ocimum forskolei [58] and limonene from Panicum maximum [59] reportedly stimulate the antennae of Ae. aegypti females. Long chain compounds present in the oil of E. elatior, particularly the aldehydes undecanal and dodecanal that have also been found in human odor exhalation [15,60], have been reported to stimulate the antenna of other mosquito species including Culex quinquefasciatus [61]. Although their capacities to repel or deter oviposition have not been determined, earlier studies have shown that unsaturated fatty acids [62] and aliphatic carboxylic acids [63] exhibit deterrent activity against C. quinquefasciatus, whereas tetradecanoic acid is repellent to both Ae. aegypti and C. quinquefasciatus [64].
Although only dodecanal and n-dodecanol were assayed for oviposition deterrent activity in the present study, other long chain oxygenated compounds, such as dodecanoic acid, undecanal, 2-undecanone and n-decanol, elicited responses in the antennae of Ae. aegypti females. Such responses may well be due to OBPs present in the sensillar of the antennae, a possibility supported by the computational models studied in this work.
A number of three dimensional structures of OBPs bound to long chain compounds have already been determined including PDB ID: 3K1E [20], 3OGN [65], 3R1O [66] and 3V2L [67]. The blind docking calculations relating to the binding of dodecanal, n-dodecanol, dodecanoic acid, undecanal, 2-undecanone and n-decanol to Ae. aegypti OBP 1 performed in the present study showed the clear preference of these ligands for the long tunnel present in the protein structure. Fig 4A and 4B show only the first best (for clarity) docked pose for these ligands, but all five best-scored poses for each of the ligands occupied this tunnel. Molecular dynamics simulations indicated the permanence of the ligands in this site, and mean ΔG values (calculated according to MM/GBSA methodology) of -2463, -2430, -2431, -2437, -2459 and -2469 kcal/ mol for the ligands in the order listed above were quite favorable for binding. These results suggest that OBP 1 may be one of the receptors in the antennae of Ae. aegypti that binds these volatiles when the female is at the stage of oviposition site selection.

Conclusions
The essential oils of three cultivars of E. elatior exhibited oviposition deterrent activity against Ae. aegypti females. Although ten components of the complex oil elicited responses in female antennae, the major constituents, namely dodecanal and n-dodecanol, produced the most pronounced responses. Commercial standards of the long chain aldehyde and alcohol were themselves active in oviposition deterrent bioassays, and a mixture containing these compounds in proportions equivalent to those found in the essential oil was found to be as active as the oil itself. Docking and molecular dynamics calculations showed that these compounds bind to OBP 1 and may play an important role in oviposition site-seeking behavior, thus indicating that the protein is a possible oviposition deterrent receptor in the antenna of Ae. aegypti.