L. amazonensis promastigotes (strain MHOM/BR/77/LTB0016) were maintained at 26°C in Schneider’s medium (Sigma-Aldrich, St. Louis, USA) supplemented with 5% fetal bovine serum (FBS), 100 μg/ml streptomycin and 100 U/ml penicillin. Alternatively, the parasites were grown in a nutritionally restricted medium with defined chemical composition, characterized by the absence of lipids and FBS. This medium consisted of a mixture of RPMI media (Sigma- Aldrich, St. Louis, USA) and DMEM (Invitrogen Corporation, GIBCO, Leiden, Netherlands) in the proportion of 1:1 (v/v). The composition of the nutritionally restricted medium was adapted from the protocol elaborated by Merlen et al. (Merlen et al., 1999), which has been described as suitable for the cultivation of several strains of Leishmania spp. and also of Trypanosoma cruzi epimastigotes. The main modification concerning the original protocol consisted of withdrawal of cholesterol supplementation. Regarding the other supplements used, the concentrations were maintained similar to those established by the authors.

The isoprenoid t, t-FOH (96% purity) was purchased from Sigma-Aldrich. After the bottle was opened, all contents were aliquoted, in the dark environment, for small bottles of amber glass. Immediately before sealing the flasks, the internal atmosphere was saturated with nitrogen gas (N₂). The aliquots’ vials were stored in a container with silica at low temperatures (-20°C) until use.

To analyze the activity of t, t-FOH on the growth of L. amazonensis, promastigotes were cultured in Schneider’s medium (5% FBS) or nutritionally restricted medium containing different concentrations of t, t-FOH. The initial inoculum was 5x10⁵ parasites/mL, and the concentration of promastigotes in the culture was determined after 24, 48, and 72 h by counting the parasites diluted in formalin (4%) using a hemocytometer. Graphs and IC₅₀ values were obtained from the GraphPad Prism software (version 7; GraphPad Software, Inc., La Jolla, CA, USA).

To assess whether t, t-FOH interfere with the cell division of L. amazonensis, carboxyfluorescein succinimidyl ester (CFSE) labeled promastigotes were cultured in Schneider’s medium or in the nutritionally restricted medium in the presence of t, t-FOH. Labeling of the parasites was performed by incubating a suspension of 2.5 x 10⁷ parasites/ml in PBS (0.1% bovine albumin) with 10 μM CFSE (Kit Cell Trace, Molecular Probes/Life Technologies) for 15 minutes at 26° C and protected from light. After this time, the reaction was stopped using cold Schneider’s medium (5% FBS). After labeling, the promastigotes were incubated with 37 µM in Schneider’s medium or 7.5 µM in the nutritionally restricted medium of t, t-FOH. The fluorescence intensity of the CFSE was analyzed 24, 48, and 72 h after labeling with CFSE. The data were acquired using a FACSCalibur flow cytometer equipped with the Cell Quest program. The obtained data were analyzed using the Summit v4.3 computer program.

Promastigotes treated with 46 µM of t, t-FOH, corresponding to the IC₅₀, were fixed with 70% ethanol, and maintained at -20° C for one hour. Next, the cells were centrifuged, and the pellet was resuspended and incubated for one hour at 26° C in 500μl of RNAse (200 μg/mL) and solubilized in PBS. Subsequently, 20 μL of propidium iodide (PI, 40μg/mL) were added, and the cells were incubated in the dark for 20 minutes at room temperature (Sen et al., 2007). The data were acquired using a FACSCalibur flow cytometer equipped with the Cell Quest program. The data were analyzed using the Summit v4.3 computer program.

DNA fragmentation in promastigotes was analyzed using a terminal deoxyribonucleotide transferase-mediated dUTP nick-end labeling (TUNEL) apoptosis detection system (Promega, Madison, WI, USA) to the manufacturer’s recommendations. Briefly, 5×10⁶ promastigotes were collected after a 48 h treatment with 46 µM of t, t-FOH, washed twice with PBS, and fixed with fixation/permeabilization solution (eBioscience, San Diego, CA, USA) for 10 minutes at 25°C. The fixed cells were incubated with a TdT reaction mixture containing FITC-labelled dUTP for 1 h at 26°C. Cells were washed and resuspended in 0.5 mL of PBS (pH 7.4) containing 0.5 mg/mL PI (BD Biosciences) before analyzing using a FACSCalibur flow cytometer. DNAse was utilized as a positive control.

The mitochondrial functionality of L. amazonensis promastigotes were evaluated by flow cytometry using the rhodamine 123 fluorescent probe (Rho 123). In these analyzes, the promastigotes (1x10⁶ parasites/mL) treated with 46 µM of t, t-FOH for 24 hours were incubated with 10 μg/mL of Rho 123 (Sigma-Aldrich, St. Louis, USA) for 20 minutes at room temperature and protected from light. In addition, promastigotes incubated under the same conditions with miltefosine (20 μM) were used as a positive control of the assay. The data were acquired using a FACSCalibur flow cytometer equipped with the Cell Quest program. The analysis of the data obtained by the cytometer was performed using the Summit v4.3 computer program.

L. amazonensis promastigotes were cultivated in the nutritionally restricted medium using glass bottles and subjected to slight agitation. Initially, the promastigotes in the culture were removed by centrifugation (3000 rpm/15 minutes), followed by sterilization by a filtration membrane filter of 0.22 μM pore size (Merck Millipore, Brazil). Next, an aliquot of 0.5 μg of progesterone was added to the sterile supernatant, which was used as the internal control of the extraction technique. For lipid extraction, we used ethyl acetate (99.9% purity, Sigma-Aldrich). Subsequently, the ethyl acetate in the sample was evaporated using a rotary evaporator, and to remove residual solvent, nitrogen gas (N₂) was used. Finally, the dried samples were stored at -20°C until analysis of the compounds.

Lipids extracted from the secretome were analyzed by chromatography gas and mass spectrometry (GC/MS). The dried samples were resuspended in 100 μL of ethyl acetate immediately before being injected into the equipment GCMS-QP2010 Ultra (Shimadzu Scientific Instruments, Tokyo, Japan). After injection, the column temperature was maintained at 50° C for 1 minute and then increased to 270°C in a ratio of 10° C/min and finally to 300°C in a ratio of 1° C/min. The helium gas flow was kept constant at 1.1 ml/min. The injector and detector temperatures were 250°C and 280°C, respectively (Torres-Santos et al., 2009).

To perform in silico analysis and identify putative farnesene synthase sequences, we gathered annotated sequences of the enzyme. A total of 109 protein sequences from plants and bacteria were obtained from RefSeq (NCBI). Given the need to focus on distant homologs, we emplyed the Hidden Markov Model (HMM profile) approach (Eddy, 1996). Initially, the sequences were aligned using MAFFT 7 software (Katoh and Standley, 2013). Subsequently, we constructed an HMM protein model using hmmbuild from HMMER 3.2.1 (Finn et al., 2011) and utilized hmmsearch to search the profile against the predicted proteins of Leishmania amazonensis, obtained from the Laboratory of Computational Biology and Systems. Concurrently, an OrthoMCL analysis (Li et al., 2003) was conducted, involving seven species of the Leishmania genus (L. panamensis, L. infantum, L. braziliensis, L. amazonensis, L. major, L. donovani, and L. mexicana), to identify orthologs within these species. Following this, with the validation of InertPro (Mitchell et al., 2019), we successfully identified putative farnesene synthase protein sequences in each Leishmania species.

The effect of t, t-FOH on L. amazonensis promastigotes growth was evaluated using a nutritionally rich medium composed of Schneider´s medium supplemented with 5% FBS (Figure 1A) or a nutritionally restricted medium (Figure 1B) containing different concentrations of t, t-FOH. The results indicate that the concentration of t, t-FOH required to inhibit 50% of the growth of L. amazonensis is significantly higher in a nutritionally rich medium compared to a nutritionally restricted medium, with the respective IC₅₀ values (µM) as follows: 24 h (46.2 ± 2.3 vs 7.2 ± 0.6), 48 h (33.0 ± 1.2 vs 4.5 ± 0.3) and 72 h (36.4 ± 1.4 vs 5.7 ± 0.7).

To evaluate the cell division of L. amazonensis promastigotes, the parasites were labeled with CFSE and then cultured in a nutritionally rich or restricted medium containing the IC₅₀ of t, t-FOH, and the fluorescence was measured after 24, 48, and 72 hours.

Figures 2A-D show the cell division of L. amazonensis promastigotes grown in a nutritionally rich medium containing 37 μM of t, t-FOH. By analyzing the histogram obtained after 48 hours of incubation, it is observed that there was a decrease in the number of cell divisions of the parasite.

Figures 2E-H show the cell division profile of the parasites grown in a nutritionally restricted medium with 7.5 μM of t, t-FOH. Interestingly, t, t-FOH had a very distinct effect on cell division of the promastigotes in the poor medium. We can observe that the parasites are distributed in two different populations, with a particular pattern of cell division. One group divides faster than the untreated control, while the other divides slower. It is more evident in 48 hours of culture when 40.6% of the treated cells divided faster than the median of the untreated parasites, while the remaining 59.4% divided slower. This is partially reversed in 72 h, possibly because the faster group reached the stationary phase, and the control continued to divide.

In this assay, we evaluated whether t, t-FOH reduces the proliferative capacity of the promastigotes by interfering in the cell cycle of the parasite. The promastigotes were cultured in the nutritionally rich medium with the concentration of t, t-FOH corresponding to the IC₅₀ (46 µM). After 24 hours, a culture sample was labeled with PI and analyzed by flow cytometry. Figure 3 shows that after the promastigotes remained incubated with the IC₅₀ value of t, t-FOH, there was a significant decrease in cells in the G₂ phase, compatible with a stop in the cell cycle. At the same time, we also observed an increase in cells in the region corresponding to Sub-G1, that is, with less DNA than G1. The occurrence of these hypodiploid cells may be suggestive of apoptosis.

To investigate the increase of promastigotes in the sub-G₀/G₁ phase after incubation with t, t-FOH, we looked for DNA fragmentation in the parasites. Analysis of the effect of t, t-FOH on the promastigotes cultured in a nutritionally rich medium was done after 48 hours of incubation, and the concentration of t, t-FOH used was equal to the IC₅₀ (46 µM). However, despite the hypodiploidy observed in the cell cycle analysis, there was no DNA fragmentation in parasites incubated with t, t-FOH (Figure 4).

Mitochondrial membrane potential (ΔΨ_m), another hallmark of apoptosis, was also investigated in cells incubated with t, t-FOH. Promastigotes cultured in a nutritionally rich medium (Schneider´s medium supplemented with 5% FBS) were treated with t, t-FOH corresponding to the IC₅₀ (46 µM). After 24 hours, the ΔΨ_m was assessed by flow cytometry. Figure 5 shows that t, t-FOH does not interfere with the mitochondrial membrane potential of the parasites.

To extract lipophilic metabolites in the culture supernatant, promastigotes of L. amazonensis were cultivated for ten days. The lipids were extracted on days 1, 4, 7, and 10 of the culture supernatants. The days 1 and 4 represent the logarithmic phase, and the days 7 and 10 illustrate the stationary phase of the cultures.

GC-MS analysis was conducted to examine the neutral lipids in the supernatant, and the representative chromatograms are depicted in Figure 6. The compound β-farnesene was detected starting from day four, while α-farnesene was observed in increasing concentrations on days 7 and 10 (Table 1).

The analysis of the chromatograms and mass spectra obtained by GC-MS, with assistance from the NIST11 spectral library, revealed the presence of both α-farnesene and β-farnesene isomers. Although the spectra are highly similar, a notable distinction lies in the intensity of the peak at m/z 81, which constitutes 50% of the total intensity in the mass spectrum of α-farnesene. This distinction can be explained through the proposed fragmentation mechanisms, which indicate potential differences in the stability of the m/z 81 fragments between the two isomers (Figure 7). Specifically, the fragmentation of β-farnesene, through sigma cleavage between the C5-C6 carbons, leads to a fragment at m/z 81 with resonance structures alternating between a primary and a secondary carbocation. In contrast, the same fragmentation mechanism for α-farnesene results in resonance structures alternating between a primary and a tertiary carbocation, with the latter offering greater stability to this fragment. This accounts for higher intensity of the m/z 81 peaks in the mass spectrum of α-farnesene. Another contributing factor to the lower stability of the m/z 81 fragments obtained from β-farnesene is the ring tension within the proposed fragment structure.

α- and β-farnesene were found in the secretome of L. amazonensis, prompting an investigation into the enzyme responsible for this reaction, which is interconnected with the mevalonate pathway. This enzyme, previously described in plants (Pechous and Whitaker, 2004) and annotated in RefSeq-NCBI as farnesene synthase, was sought after. Using these sequences, we initiated a search for distant homologs to identify a corresponding sequence in Leishmania spp. Through this in silico analysis, we identified a putative farnesene synthase for Leishmania spp. (Figure 8). This search allowed us to identify a putative homologous sequence of farnesene synthase in L. amazonensis, based on a profile constructed using previously annotated sequences from different species. Utilizing the OrthoMCL software, which included seven other species of Leishmania, including L. amazonensis, we determined the homologous group to which it belonged based on its sequence identifier. Subsequently, we verified the classification of the enzymes in Leishmania and the bacteria used to construct the HMM profile with InterPro, aiming to validate the putative homolog. In this in silico validation, we found that all sequences exhibited similar characteristics to the farnesene synthase found in bacteria. These characteristics include protein family membership, involvement in the same biological processes, and molecular functions (Table 2).