2.1. Insects
Ant workers of the genus Solenopsis were collected using attractive food baits (animal protein) distributed near nests located in a home environment in the municipality of Aracaju, SE, Brazil (10° 53′ S, 37° 06′ W). The ants were collected with the aid of a soft bristle brush, placed in 200 mL plastic pots, covered with organza fabric and sent to the Integrated Pest Management Laboratory located at the Federal University of Sergipe (UFS), São Cristóvão, SE, Brazil (10° 54′ S, 37° 04′ W).
The insects were then placed in plastic containers (11 cm x 7.5 cm), covered with plastic film (G-útil Guarufilme, Guarulhos, SP, Brazil) and kept under controlled conditions (27 ± 2 °C, 70 ± 5% RH and 12 hour photoperiod) for 24h for acclimatization, until the experiments were carried out. During this period, only distilled water was provided to the workers.
2.2. Synthesis of carvacrol derivatives
The derivatives were synthesized at the Pharmaceutical Chemistry Laboratory of the UFS University through nucleophilic nuclear substitution and esterification reactions. Nine compounds were synthesized from the base molecule carvacrol: the esters carvacryl acetate, carvacryl benzoate, carvacryl butyrate, carvacryl hexanoate, carvacryl isobutyrate, carvacryl isovalerate, carvacryl pivaloate, carvacryl ethyl ether and carvacryl trichloroacetate (
Figure 1).
The ether compound was synthesized according to the method described in Collen [
47]. The esters, however, followed the methodologies of Dolly & Barba, Ben Arfa et al., Morais et al. [
48,
49,
50], using THF (tetrahydrofuran) as solvent [
31].
The reactions were followed by thin layer chromatography (TLC), verified in ultraviolet light of 256 ηm and compared with the starting material (carvacrol). Derivatives were purified using sílica gel 60 in the stationary phase and pure hexane in the mobile phase. Melting points were determined on a Logen Scientific melting point apparatus without correction. The Chemicals used in the reactions and the monoterpene carvacrol were obtained from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA).
2.3. Bioassays
The synthetic derivatives were compared with the precursor molecule carvacrol and with the EO of
L. gracilis, rich in carvacrol (50,7%) [
30]. The EO of
L. gracilis was obtained of plants that are maintained at the UFS Active Germplasm Bank – located at the experimental farm – UFS Rural Campus, São Cristóvão, SE, Brazil (10° 55′ S, 37° 11′ W). The extraction, identification and quantification of compounds was performed as described in Santos et al. [
51].
The bioassays were conducted in climate-controlled rooms with controlled conditions (27 ± 2 °C, 70 ± 5% RH and 12-hour photoperiod) at the Integrated Pest Management Laboratory at UFS. The compounds used were diluted in acetone solvent (Panreac, UV-IR-HPLC-GPC PAI-ACS, 99,9% purity) and applied topically (0.5 µL) in the prothorax region of the workers, with the aid of a Hamilton® microsyringe (Reno, NV, USA). In the negative control, only acetone was used.
In order to facilitate the application of the compounds, the ants were immobilized in a freezer at -19 °C (Eletrolux®, Curitiba, PR, Brazil) for 90 s. Preliminary tests have shown that this method does not affect the survival and behavior of ants. To determine the doses to be applied, Fifty workers were weighed on a precision analytical balance (Shimadzu, AUW220D) to obtain the average mass of the insects.
In the acute toxicity bioassays, the treatments used were the EO of L. gracilis, the base molecule carvacrol and the nine synthetic derivatives. In the behavioral bioassays, EO of L. gracilis, carvacrol and its most toxic synthetic derivative (carvacryl benzoate) were used.
2.3.1. Acute toxicity
The bioassays to determine the dose-response curves and lethal doses were conducted in a completely randomized design with eight replications per treatment. The experimental unit consisted of a group of six to seven worker ants of Solenopsis sp. of the same size, totaling 4144 individuals. Initially, tests were carried out with three doses (1, 5 and 10 µg of substance/mg of insect) in order to obtain mortalities between 5 and 95%. Intermediate doses were then tested to establish dose-response curves.
The ants were placed in a Petri dish (Global Trade Technology, Monte Alto, SP, Brazil) (6 x 1.5 cm) lined with filter paper (Unfil) moistened with 0.4 mL of distilled water and covered with plastic film (G-util Guarufilme, Guarulhos, SP, Brazil). Mortality assessment was performed 48h after setting up the bioassays. The criterion was based on the mobility of ants after being stimulated with a soft bristle brush. Insects that did not move after the stimulus were considered dead.
In the bioassays to obtain survival curves and lethal time of the compounds on workers of Solenopsis sp. the LDs90 obtained in the acute toxicity bioassays were used. The procedures were similar to those used in toxicity bioassays, with the exception of doses, number of repetitions and evaluation time. Fifteen repetitions were performed per treatment, totaling 420 individuals. The first mortality assessments were performed every 10 min until completing 2h; then every 30 min until completing 6 h; followed by 1 h assessments during the first 12 h and every 2 h until reaching the first 24 h. Subsequently, evaluations were carried out at intervals of 4 hours in the following 12 hours and every 12 hours until completing 100 hours.
2.3.2. Individual and collective behavior
For this experiment, we observed the individual behavior of Solenopsis sp. treated (individual behavior) (n = 1) and the interaction of a group of ants not exposed to the compounds (6 ants) with a treated one (collective behavior) (n = 7).
The experimental designs were completely randomized, with 60 repetitions per treatment. The workers of Solenopsis sp. were exposed to acetone (negative control) and to the LDs30 of the compounds determined in the acute toxicity bioassays. Before being exposed to the treatments, the ants were placed in a Petri dish (6 x 1.5 cm) lined with filter paper moistened with 0.4 mL of distilled water for 5 min for acclimatization.
In the experiment to evaluate individual behavior, after acclimatization and immobilization the ants were treated and placed in Petri dishes individually, totaling 240 individuals. After 1 min of exposure, self-cleaning behavior was counted. As for the collective behavior experiment, after acclimatization and immobilization, an ant was randomly removed from the group of seven ants and marked with non-toxic yellow paint (Acrilex Tintas Especiais S.A., São Bernardo do Campo, São Paulo, Brazil). After 3 min of marking with the ink, the ant received the treatment and after 1 min it was relocated in the Petri dish with the other untreated ants. One min after the relocation on the board, the counting of the antenation, cleaning and aggregation behaviors began. Preliminarly tests indicated that the paint did not affect the survival and behavior of the ants.
In both bioassays, observations were performed for one continuous minute, with a one-minute interval between them, during a period of 10 minutes, totaling 5 minutes of observation for each insect. Overall, behaviors were recorded for 2400 min [2 bioassays x 4 treatments x 60 repetitions x 5 min of observation].
2.3.3. Walking behavior
The individual walking behavior was evaluated with workers of Solenopsis sp. submitted to the LDs30 of each treatment and acetone (negative control). The experiments were carried out in a completely randomized design, with 60 repetitions per treatment, totaling 240 individuals. The bioassays were performed in Petri dish arenas (6 x 1.5 cm), lined with moistened filter paper and covered with plastic film.
After 1 min of placing the individuals in the arena, recording began for a period of 600 s using a vídeo camera (Panasonic SD5 Superdynamic – model WV-CP504), equipped with a Spacecom lens (1/3” 3-8 mm) attached to a computer. The distance traveled (mm), velocity (mm/s), meander (°/mm) and angular velocity (°/s) were captured in Ethovision XT software (version 8.5; Noldus Integration System, Sterling, VA) and data analyzed using Studio 9 software (Pinnacle Systems, Moutain View, CA).
2.3.4. Horizontal transfer of compounds
The aim of this experiment was to investigate whether the compounds used in this study can be transferred from dead ants to live ants that were not subjected to the treatments. Two proportions were used in the experiments, [1:20] = 1 dead ant to 20 live ants and [1:5] = 4 dead ants to 20 live ants. The design was completely randomized with ten repetitions, totaling 800 live individuals and 200 cadavers. The bioassays were performed in Petri dish arenas (6 x 1.5 cm), lined with filter paper moistened with 0.4 mL of distilled water and covered with film plastic.
Dead ants were obtained after 24 h of exposure of the workers to the LDs90 of the compounds. For the control, cadavers were obtained after application of acetone and freezing in a freezer at -19 °C for 30 min. The number of live and dead ants was evaluated after 1, 2, 4, 6, 10, 14, 26, 38, 50, 62, 74, 86, 98, 110 and 122 hours after setting up the experiments.
2.4. Statistical analysis
Mortality data were submitted to Probit analysis to determine the dose-response curve for each treatment using the PROC PROBIT procedure in SAS (SAS Institute, 2008). From these curves, the lethal doses necessary to cause 30, 50 and 90% mortality (LD30, LD50 and LD90) and their respective confidence limits at 95% probability (LC95) were obtained. The determination of the most toxic synthetic derivative was performed using the criterion of non-overlapping confidence intervals with the origin of the interval in the LDs50.
For analysis of survival, horizontal transfer and their respective lethal times, Kaplan-Meier estimators were used through the Log-Rank test (SigmaPlot, version 14). From this analysis, survival curves and lethal times (LT50) were obtained for each treatment. The Holm-Sidak multiple coparison method was used at a significance level of 0.05 to compare the curves (SigmaPlot, version 14). The LT50 were compared by the criterion of non-overlapping confidence intervals with the origin of the interval.
The data from the behavioral bioassays (individual, collective and walking) (n = 60) were analyzed by parametric tests, considering that the data distribution is close to the Gaussian distribution (Central limit theorem) and, therefore, meets the assumptions of the analysis of variance (ANOVA). Thus, these data were initially submitted to ANOVA followed by Dunnett’s test (p < 0.05) to verify differences between treatments in relation to the control. All behavioral data analyzes were performed using SigmaPlot software, version 14.