Antitrypanosomal Activity of Anthriscus Nemorosa Essential Oils and Combinations of Their Main Constituents

This study aimed to investigate the susceptibility of Trypanosoma brucei to the Anthriscus nemorosa essential oils (EOs), isolated compounds from these oils, and artificial mixtures of the isolated compounds in their conventional and nanoencapsulated forms. The chemical composition of the essential oils from the aerial parts and roots of Anthriscus nemorosa, obtained from a wild population growing in central Italy, were analyzed by gas chromatography/mass spectrometry (GC/MS). In both cases, the predominant class of compounds was monoterpene hydrocarbons, which were more abundant in the EOs from the roots (81.5%) than the aerial parts (74.0%). The overall results of this work have shed light on the biological properties of A. nemorosa EO from aerial parts (EC50 = 1.17 μg/mL), farnesene (EC50 = 0.84 μg/mL), and artificial mixtures (Mix 3–5, EC50 in the range of 1.27 to 1.58 μg/mL) as relevant sources of antiprotozoal substances. Furthermore, the pool measurements of ADP (adenosine diphosphate) and NTPs (nucleoside triphosphates) in the cultivated bloodstream form of trypanosomes exposed to different concentrations of EOs showed a disturbed energy metabolism, as indicated by increased pools of ADP in comparison to ATP (adenosine triphosphate) and other NTPs. Ultimately, this study highlights the significant efficacy of A. nemorosa EO to develop long-lasting and effective antiprotozoal formulations, including nanoemulsions.


Introduction
African trypanosomiasis (HAT), also called sleeping sickness, is an endemic disease of rural sub-Saharan Africa, caused by two species of Trypanosoma brucei: T. brucei gambiense and T. brucei rhodesiense [1]. The T. brucei gambiense parasite causes more than 98% of human infections, and it is mainly present in Western and Central Africa. In contrast, the T. brucei rhodesiense parasite mostly affects animals and rarely humans, and is most common in eastern and southern Africa [2]. The tsetse fly mediates the transmission of the parasite causing the disease. T. brucei gambiense infections are chronic, and the slow progression of the disease allows the patients to survive for several years. In contrast, T. brucei rhodesiense infections are acute, and the faster progression of the disease leads to death within a few months if untreated [3,4]. Two stages characterize HAT: the first stage symptoms are fever, headache, joint pain, and itching due to the parasite's passage in the blood and lymph system [3,4]. The second stage symptoms are poor coordination, dramatic mood swings, confusion, and convulsions due to the parasite's passage in the central nervous system [3,4]. Gambiense HAT is treated with intramuscular pentamidine once a day for seven days in the early stage and with nifurtimox-eflornithine in the second stage with the hospitalization of the patients. Fexinidazole was introduced in 2018 as an oral treatment of gambiense HAT, and its efficacy was demonstrated in both stages of the disease. Therapy with fexinidazole is not conclusive for patients with severe central nervous system symptoms, and the bioavailability of the drug decreases dramatically if taken without a meal with subsequent failure of the therapy [5]. Nowadays, the development of new therapeutic strategies for HAT treatment represents an enormous challenge in terms of toxicity, costs, and administration, especially in underdeveloped countries. In this regard, natural products, with particular attention to essential oils (EOs), might represent useful tools in discovering new promising drug candidates. In particular, plants from the Apiaceae family have been investigated for their promising activity against T. brucei [6,7]. Within this family, Anthriscus nemorosa (M.Bieb.) Spreng. is an aromatic plant, which is also known with the names of 'gimigimi' or 'peçek' in Turkey [8]. The plant grows in Europe and Asia's temperate areas, in countries like Italy and Japan [9]. The traditional medicine of several countries reserves ample space for A. nemorosa for the treatment of several diseases. For instance, its fruits are traditionally used as a remedy for gastrointestinal disorders [8], and for treating inflammation and rheumatism [10,11]. The plant is also used to prepare herby cheese, called 'otlu peynir', in different areas of Turkey [12]. Moreover, ethnobotanical studies revealed its usefulness to improve memory in Alzheimer's disease. This plant also holds several pharmacological activities, as preventing anxiety and depression [12][13][14]. The root EO and the flower ethyl acetate extract showed a significant antioxidant activity [12]. A. nemorosa also exhibits an antifungal potential, and the hexane extract of its aerial parts was highly efficient against Candida albicans [15]. As far as the volatile fraction is concerned, only a few studies are available on the chemical composition. Because EOs of some Apiaceae plants represent an attractive alternative approach to fighting protozoan parasites, this study aimed to better define the chemical composition of A. nemorosa EOs and to evaluate their antitrypanosomal activity against T. brucei. Being naturally occurring compounds, EOs and their components are always believed to be environmentally safe, user-friendly, and generally not toxic to humans [12]. Unfortunately, EOs can hide unexpected drawbacks (e.g., low stability and water solubility, high volatility, and side-effects). To overcome these issues, we here used encapsulation in nanoemulsions to improve the chemical-physical profile of the EOs.

EO Compositions
Chemical compositions of the EOs hydrodistilled from the A. nemorosa aerial parts (A. nemorosa EO-AP) and roots (A. nemorosa EO-R) are reported in Table 1. Thirty-four volatile components were identified in the two EO samples by GC-MS, accounting for 98.1 and 98.6% of the total compositions, respectively. In both cases, the predominant class of compounds was monoterpene hydrocarbons, which were more abundant in the EOs from the roots (81.5%) than the aerial parts (74.0%). Among the monoterpene hydrocarbons, myrcene dominated over the others, especially in the aerial parts, accounting for 40.1 and 19.9% of the chemical profile in the aerial parts and roots, respectively. In contrast, the β-pinene content in the A. nemorosa EO-AP was only 3.0%, but it represented 22.0% of the total composition in the A. nemorosa EO-R. γ-Terpinene was detected at 11.6 and 9.2% in EOs from aerial parts and roots, respectively, while (Z)-β-ocimene was more abundant in the EO from roots (8.9%) than the aerial parts (4.4%). Another group of compounds occurring in noteworthy levels in the two EOs was the sesquiterpene hydrocarbons. They accounted for only 1% of the whole composition in the case of the root EO, while they represented 18.8% of the chemical profile in the aerial part EO, with germacrene D (10.4%), (E,E)-α-farnesene (4.0%) and δ-cadinene (2.3%) as the main constituents. It is interesting to note that the class of oxygenated monoterpenes was totally absent in both EOs. In contrast, oxygenated sesquiterpenes were detected at 3.0 and 4.4% in the EOs from aerial parts and roots, respectively. Moreover, other compounds, such as alkanes, were found in the EOs from the aerial parts and roots (2.3 and 11.2%, respectively), with nonane being the most significant one (1.7 and 9.5% in the EOs from aerial parts and roots, respectively).

Antitrypanosomal Activity
Testing the EOs hydrodistilled from A. nemorosa aerial parts and roots, we showed that both of them effectively inhibited T. brucei proliferation, especially the EO from aerial parts with an EC 50 value of 1.17 µg/mL (Table 2 and Figure 1) and a selective index (SI) of 4.65. The root EO was two times less active (EC 50 = 2.36 µg/mL) and selective (SI = 2.25). Since the aerial part EO was the most active of the two, it was chosen for further investigations. The major components found in the A. nemorosa EO (aerial parts) belong to two chemical groups, namely monoterpene hydrocarbons and sesquiterpene hydrocarbons. Among the constituents tested, farnesene (EC 50 = 0.84 µg/mL) and β-ocimene (EC 50 = 1.1 µg/mL) were the most potent and showed EC 50 values slightly lower than that of the intact Isocratic HPLC analysis for the simultaneous determination of dNTPs, rNTPs, and ADP in biological samples oil. The β-ocimene tested was then a mixture of Z and E isomers. β-pinene, p-cymene, limonene, myrcene, and γ-terpinene were less active and had EC 50 values higher than 4.5 µg/mL. The sesquiterpene hydrocarbon germacrene D was not tested because it is not commercially available. As reported in Table 2, the antitrypanosomal activity of the major compounds assayed was in the order: farnesene > β-ocimene > p-cymene > limonene > β-pinene > myrcene > γ-terpinene. The chemical structures of the bioactive compounds varied from linear (farnesene, β-ocimene, and myrcene) to monocyclic (limonene, p-cymene, and γ-terpinene) and bicyclic derivatives (β-pinene), and contain one to four unsaturations ( Figure 2).      [17]. b Data from [6]. c Data from [18].

Contribution of Individual Compounds to the Overall Antitrypanosomal Activity and Synergistic Effects
In order to better understand synergistic effects of the different components in the A. nemorosa EO-AP, we carried out a compound elimination assay where omitted compounds were replaced by DMSO (Table 3 and Figure 3). It then emerged that limonene, β-ocimene, p-cymene, and farnesene are the EO's major antitrypanosomal constituents whereas myrcene seemed to have a protective effect ( Table 3). The protective effect was evident from that the removal of myrcene led to a decrease in the EC50 value from 7.25 to 2.25 μg/mL, and the antitrypanosomal effect of limonene, β-ocimene, p-cymene, and farnesene was evident from that if three of these components were removed (Mix 8-11), it led to an increased EC50 value. The strongest effect was if limonene, β-ocimene, and farnesene were simultaneously removed from the mixture leading to an EC50 increase from 7.25 (Mix 1) to 49.2 μg/mL (Mix 9).

Contribution of Individual Compounds to the Overall Antitrypanosomal Activity and Synergistic Effects
In order to better understand synergistic effects of the different components in the A. nemorosa EO-AP, we carried out a compound elimination assay where omitted compounds were replaced by DMSO (Table 3 and Figure 3). It then emerged that limonene, β-ocimene, p-cymene, and farnesene are the EO's major antitrypanosomal constituents whereas myrcene seemed to have a protective effect ( Table 3). The protective effect was evident from that the removal of myrcene led to a decrease in the EC 50 value from 7.25 to 2.25 µg/mL, and the antitrypanosomal effect of limonene, β-ocimene, p-cymene, and farnesene was evident from that if three of these components were removed (Mix 8-11), it led to an increased EC 50 value. The strongest effect was if limonene, β-ocimene, and farnesene were simultaneously removed from the mixture leading to an EC 50 increase from 7.25 (Mix 1) to 49.2 µg/mL (Mix 9).   Table 3 and the curves are colored according to the following scheme: Mix 3 (blue curve), Mix 4 (red curve), Mix 5 (orange curve), and Mix 6 (purple curve). Each EC50 value represents the mean of four independent experiments with standard errors.

Preparation and Characterization of A. nemorosa EO-Based Nanoemulsions
To investigate whether the antitrypanosomal activity and selective index could be improved, nanoemulsions of A. nemorosa EO-AP were formulated and tested against T. brucei and mammalian cells. The surfactant polysorbate 80 at a concentration of 2% (w/w) was used as the stabilizing agent in these nanoemulsions (Table 4). A. nemorosa EO-based nanoemulsions were formulated to evaluate the influence of the nanoencapsulation on the active ingredients' biological activity. An initial screening to assess the A. nemorosa EO-AP total oily phase and the surfactant percentages was carried out to select the final quali-quantitative composition of nanoemulsions named NANO A and NANO B (Table  4). The obtained nanoemulsified systems were assessed by the absence of visible oily droplets through optical microscopy. The droplet size distributions of nanoemulsions were centered between 140 and 170 nm (mean diameter, Z-average) up to two months from the preparation for NANO A and up to four months for NANO B. At a longer storage time, both nanoemulsions showed an increase in the mean droplet diameter, which was more pronounced for NANO A than NANO B. Indeed, the mean droplet diameter reached approximately 700 nm for NANO A (after 9 months from the preparation), while it was below 300 nm for NANO B ( Figure 4A). Besides, the polydispersity Index (PDI) values were lower for NANO B than NANO A ( Figure 4B). PDI is an indicator (ranging from 0 to 1) of the width of the distribution (a smaller index indicates a monodispersed system) [19]. Overall, these results highlight the role of ethyl oleate as co-solvent in the formulation, since it has improved the particle size and stability of nanoemulsions.  Table 3 and the curves are colored according to the following scheme: Mix 3 (blue curve), Mix 4 (red curve), Mix 5 (orange curve), and Mix 6 (purple curve). Each EC 50 value represents the mean of four independent experiments with standard errors.
In Table 3, Wadley's formula was used to analyze synergistic interactions among some of the constituents of the A. nemorosa EO-AP. The different constituents displayed additive, antagonistic, and synergistic relationships. Our results indicated several additive interactions between the six main components of A. nemorosa EO-AP (Mix 1, Mix 8-9, and Mix 11). The simultaneous combination of the six major constituents (Mix 1) produced an EC 50 value (7.25 µg/mL), which is 1.7 fold higher than expected (4.09 µg/mL, R = 0.56, additive). On the other hand, limonene, β-ocimene, p-cymene, and farnesene all showed synergistic effects with each other (Mix 3-7), and this synergy was prevented if myrcene was present (Mix 1). In contrast, there was no negative effect of removing just one of these four active compounds (compare Mix 4-7 with Mix 3), indicating that the synergy is not dependent on that all four components are present.

Preparation and Characterization of A. nemorosa EO-Based Nanoemulsions
To investigate whether the antitrypanosomal activity and selective index could be improved, nanoemulsions of A. nemorosa EO-AP were formulated and tested against T. brucei and mammalian cells. The surfactant polysorbate 80 at a concentration of 2% (w/w) was used as the stabilizing agent in these nanoemulsions (Table 4). A. nemorosa EO-based nanoemulsions were formulated to evaluate the influence of the nanoencapsulation on the active ingredients' biological activity. An initial screening to assess the A. nemorosa EO-AP total oily phase and the surfactant percentages was carried out to select the final quali-quantitative composition of nanoemulsions named NANO A and NANO B (Table 4). The obtained nanoemulsified systems were assessed by the absence of visible oily droplets through optical microscopy. The droplet size distributions of nanoemulsions were centered between 140 and 170 nm (mean diameter, Z-average) up to two months from the preparation for NANO A and up to four months for NANO B. At a longer storage time, both nanoemulsions showed an increase in the mean droplet diameter, which was more pronounced for NANO A than NANO B. Indeed, the mean droplet diameter reached approximately 700 nm for NANO A (after 9 months from the preparation), while it was below 300 nm for NANO B ( Figure 4A). Besides, the polydispersity Index (PDI) values were lower for NANO B than NANO A ( Figure 4B). PDI is an indicator (ranging from 0 to 1) of the width of the distribution (a smaller index indicates a monodispersed system) [19]. Overall, these results highlight the role of ethyl oleate as co-solvent in the formulation, since it has improved the particle size and stability of nanoemulsions.  To determine their in vitro antitrypanosomal activity and selectivity, NANO A and NANO B were tested on T. brucei s427 and Balb/3T3 cells with the same procedure used for EOs, pure compounds, and artificial mixtures. The EC50 values of NANO A and NANO B were much higher than those obtained from testing the pure A. nemorosa EO-AP, mainly because the nanoemulsions only contain 6% w/w of EO and the major fraction of the weight is water. We have therefore also included EC50 values normalized by the percentage of EO in the nanoemulsions in brackets. These normalized values in Table 5 are directly comparable to the EO-AP EC50 value reported in Table 1. The normalized EC50 values are still higher than the EC50 obtained with the EO itself, but on the other hand the nanoemulsions is also less toxic to mammalian cells (Table 5). Thus, an interpretation of the results is that the nanoemulsions prevent the effect of the EO on both the trypanosomes and on the mammalian cells but that this ameliorating effect is stronger on the mammalian cells, especially with NANO A, resulting in a two times higher selectivity index compared to the EO itself.

Determination of NTP and ADP Pools in T. brucei
Finally, we analyzed the effect of A. nemorosa EOs on nucleotide pools in T. brucei grown in culture. Extracted nucleotides from the trypanosomes were separated by HPLC for quantification. The treatment of trypanosomes with different concentrations of A. nemorosa EOs (0−10 μg/mL) led to a concentration-dependent decrease of all NTP pools (Figure 5A). The four NTPs are equally affected at the lower EO concentration, and only some To determine their in vitro antitrypanosomal activity and selectivity, NANO A and NANO B were tested on T. brucei s427 and Balb/3T3 cells with the same procedure used for EOs, pure compounds, and artificial mixtures. The EC 50 values of NANO A and NANO B were much higher than those obtained from testing the pure A. nemorosa EO-AP, mainly because the nanoemulsions only contain 6% w/w of EO and the major fraction of the weight is water. We have therefore also included EC 50 values normalized by the percentage of EO in the nanoemulsions in brackets. These normalized values in Table 5 are directly comparable to the EO-AP EC 50 value reported in Table 1. The normalized EC 50 values are still higher than the EC 50 obtained with the EO itself, but on the other hand the nanoemulsions is also less toxic to mammalian cells (Table 5). Thus, an interpretation of the results is that the nanoemulsions prevent the effect of the EO on both the trypanosomes and on the mammalian cells but that this ameliorating effect is stronger on the mammalian cells, especially with NANO A, resulting in a two times higher selectivity index compared to the EO itself.

Determination of NTP and ADP Pools in T. brucei
Finally, we analyzed the effect of A. nemorosa EOs on nucleotide pools in T. brucei grown in culture. Extracted nucleotides from the trypanosomes were separated by HPLC for quantification. The treatment of trypanosomes with different concentrations of A. nemorosa EOs (0-10 µg/mL) led to a concentration-dependent decrease of all NTP pools ( Figure 5A). The four NTPs are equally affected at the lower EO concentration, and only some deviation is observed at the higher dose. This is most clearly seen in Figure 5B where relative nucleotide pools are plotted as a percentage of the total NTP pool. In contrast, the relative ADP pools are strongly affected by the treatment indicating a disturbed cellular energy charge (ATP: ADP ratio). The effects on the T. brucei nucleotide pools were similar with the EOs from aerial parts and roots.

Discussion
The EOs and their isolated components provide evolutionary validated core structures that can inspire the synthesis of new chemotherapeutic agents. Due to the singular chemical features of their main constituents (e.g., small lipophilic molecules), EOs can cross the cellular membrane and interact with intracellular targets, playing essential roles against pathogen survival, proliferation, and differentiation [20]. The chemical compositions of the EOs obtained from the aerial parts and roots of a central Italian population of A. nemorosa vary considerably from those previously reported for this species. Karakaya et al. found high levels of spathulenol (49.6%) and biclycogermacrene (8.9%) and almost a complete absence of myrcene in the aerial parts of a Turkish population [12]. Nickavar et al. studied an Iranian population and found (E)-nerolidol (41.7%), β-elemene (13%), and α-zingiberene (9.9%) as the major constituents of the aerial parts of A. nemorosa [21]. (E)caryophyllene (23.6%), caryophyllene oxide (12.3%), δ-cadinene (12.1%), and trans-pinocarveol (9.8%) were identified as major constituents of aerial parts EO of A. nemorosa from another Turkish population [13]. Lastly, Kiliç et al. reported that (E)-caryophyllene (15.8%), caryophyllene oxide (14.5%), δ-cadinene (13.4%), germacrene D (8.9%), and transpinocarveol (6.2%) are the primal components of aerial parts EO of A. nemorosa from Turkey [22]. Considering our results, we believe that different factors such as geographic origin, genetics, period of harvesting and environmental conditions, type of processing, and drying may affect the observed variability.
The first important outcome that emerged from our study is linked to the relevant antitrypanosomal activity showed by the A. nemorosa EO with EC50 values of 1.17 μg/mL (SI = 4.65) and 2.36 μg/mL (SI = 2.25) for aerial parts and roots, respectively ( Table 2). Plant EOs frequently exhibit stronger antitrypanosomal activity in comparison to each of their individual constituents. However, in this case, the major components of the EO had only a modest antitrypanosomal effect. Instead, the major activity came from two less abundant components, farnesene and β-ocimene, which when assessed separately showed equal or greater antitrypanosomal effects compared to the natural A. nemorosa EO-AP. Farnesene had a lower EC50 value (0.85 μg/mL) compared both to the intact oil (1.17 μg/mL) and the best artificial mixture (1.06 μg/mL for Mix 6, Table 2). The higher activity of farnesene is logical because it is the strongest antitrypanosomal agent in the mixture and is more concentrated in its pure form. On the contrary, it is remarkable that the EO

Discussion
The EOs and their isolated components provide evolutionary validated core structures that can inspire the synthesis of new chemotherapeutic agents. Due to the singular chemical features of their main constituents (e.g., small lipophilic molecules), EOs can cross the cellular membrane and interact with intracellular targets, playing essential roles against pathogen survival, proliferation, and differentiation [20]. The chemical compositions of the EOs obtained from the aerial parts and roots of a central Italian population of A. nemorosa vary considerably from those previously reported for this species. Karakaya et al. found high levels of spathulenol (49.6%) and biclycogermacrene (8.9%) and almost a complete absence of myrcene in the aerial parts of a Turkish population [12]. Nickavar et al. studied an Iranian population and found (E)-nerolidol (41.7%), β-elemene (13%), and α-zingiberene (9.9%) as the major constituents of the aerial parts of A. nemorosa [21]. (E)-caryophyllene (23.6%), caryophyllene oxide (12.3%), δ-cadinene (12.1%), and trans-pinocarveol (9.8%) were identified as major constituents of aerial parts EO of A. nemorosa from another Turkish population [13]. Lastly, Kiliç et al. reported that (E)-caryophyllene (15.8%), caryophyllene oxide (14.5%), δ-cadinene (13.4%), germacrene D (8.9%), and trans-pinocarveol (6.2%) are the primal components of aerial parts EO of A. nemorosa from Turkey [22]. Considering our results, we believe that different factors such as geographic origin, genetics, period of harvesting and environmental conditions, type of processing, and drying may affect the observed variability.
The first important outcome that emerged from our study is linked to the relevant antitrypanosomal activity showed by the A. nemorosa EO with EC 50 values of 1.17 µg/mL (SI = 4.65) and 2.36 µg/mL (SI = 2.25) for aerial parts and roots, respectively ( Table 2). Plant EOs frequently exhibit stronger antitrypanosomal activity in comparison to each of their individual constituents. However, in this case, the major components of the EO had only a modest antitrypanosomal effect. Instead, the major activity came from two less abundant components, farnesene and β-ocimene, which when assessed separately showed equal or greater antitrypanosomal effects compared to the natural A. nemorosa EO-AP. Farnesene had a lower EC 50 value (0.85 µg/mL) compared both to the intact oil (1.17 µg/mL) and the best artificial mixture (1.06 µg/mL for Mix 6, Table 2). The higher activity of farnesene is logical because it is the strongest antitrypanosomal agent in the mixture and is more concentrated in its pure form. On the contrary, it is remarkable that the EO gives almost as high effect (low EC 50 ) as farnesene itself, despite that the two major antitrypanosomal agents, farnesene and β-ocimene, together make up only 11.9% of the total EO. This is a main reason why we investigated synergistic interactions in the EO.
Our results demonstrate additive, synergistic, and antagonistic interactions among constituents of A. nemorosa EO from aerial parts. Using Wadley's model, we compared the calculated EC 50 values of our artificial EOs with expected EC 50 values. To determine the potential contribution of each component to the overall toxicity of the EO, we made a blend of all major constituents (Mix 1), mimicking the natural EO but excluding components with no antitrypanosomal activity. We also prepared different blends, each lacking one or several major constituents being replaced by DMSO (Mix 2-11), relying on the natural composition of aerial part EO as indicated by GC-MS. When we mixed the six constituents (Mix 1), we found that their observed antitrypanosomal activity was not as high as expected (EC 50 = 7.25 µg/mL of Mix 1 vs. EC 50 = 1.17 µg/mL of the corresponding EO), and the relationship between the constituents of the mixture was additive (R = 0.56). Presumably, the inactive constituents removed from Mix 1 have some additional synergistic effects on the six components. Although not active individually, their presence is necessary to achieve the observed antitrypanosomal activity. As shown in Table 3, the activity of A. nemorosa EO from aerial parts might be attributed to the presence of its bioactive constituents, limonene, β-ocimene, p-cymene, and farnesene (5.7%, 7.2%, 2.2%, and 4.7% in the natural EO and 9.1%, 11.4%, 3.7% and 7.4% in the artificial EO, respectively). Different blends of these components (Mix 3-7 in Table 3) showed strong synergy with R values between 3-6. Interestingly, this synergy was weakened in the presence of β-pinene and disappeared completely in the presence of myrcene. Considering the concentrations of those three components in the essential oil of aerial parts (19.8% combined) and their enhanced effect when combined, the antitrypanosomal activity of A. nemorosa EO-AP might mainly be a consequence of their synergistic interaction. In the EO itself, the antagonistic effects from myrcene and β-pinene seems not to be present and the EC 50 value is comparable to Mix 3 where only the synergistic components are present (EC 50 values: 1.17 µg/mL for the EO vs. 1.43 µg/mL for Mix 3).
The synergistic interactions of EO components might be useful in two different scenarios. First, using artificial mixtures based on EO composition rather than single isolated components might be a more powerful multitarget strategy for tackling trypanosomiasis. Second, understanding the mechanism behind the synergistic interactions between EO components could further help to develop more efficacious antitrypanosomal drugs based on custom blends of monoterpenoids/sesquiterpenoids that do not occur in nature. Various mechanisms are triggering synergistic interactions [23]. The potential synergistic interaction detected in our study may be caused by one or more of these proposed mechanisms because the A. nemorosa EO-AP is a complex mixture of components. For instance, monoterpenes like limonene, characterized by a cycle bearing an exocyclic methylene group, can react with functional groups of proteins, such as trypanothione synthase, producing cell oxidative damage [24]. The aromatic monoterpene p-cymene, being highly hydrophobic, is easily incorporated into the lipid bilayer, facilitating other bioactive components' penetration. A previous study conducted by Monzote et al. has highlighted the antiprotozoal properties of p-cymene [25]. The acyclic monoterpene hydrocarbon β-ocimene, is instead involved in the plant defense against predators and is responsible for evoking specific responses [26]. The first evidence of its effect against T. brucei was reported in 2018 (EC 50 of 1.1 µg/mL and SI > 91) [6]. Lastly, the acyclic olefin (E)-β-farnesene is an efficient and robust alarm pheromone in most aphid species. Various plants also produce (E)-β-Farnesene as aphid repellant since aphids exposed to this chemical become agitated and disperse from the host. Although (E)-β-farnesene is widespread in plants, this is the first report about its effects against T. brucei (EC 50 = 0.84 µg/mL).
Generally, EOs show many biological activities, although their use is often limited by several drawbacks (i.e., polymerization, isomerization, oxidation, and rearrangement), which depend on several environmental parameters. These instability issues may end up in a reduction or loss of activity. Moreover, EOs have poor physicochemical properties (e.g., water insolubility, high volatility, and a short half-life) that make them difficult to handle. In this regard, nanoemulsions can help to overcome the drawbacks of EOs by improving their physicochemical properties and stability. Moreover, due to their subcellular size, nanoemulsions could enhance the bioavailability and active compounds effectiveness through their solubilization into tiny oily droplets, thereby allowing a deeper tissue penetration and easier cellular uptake. Previous studies highlight the importance of vehiculation of EOs against parasites of medical importance, including Plasmodium, Leishmania, and Trypanosoma, as well as against vectors of human diseases. For instance, in a study carried out by Bouyahya et al. in 2017, the nanoemulsions of Lavandula angustifolia L. and Rosmarinus officinalis L. EOs showed antiparasitic effects greater than that obtained with the non-emulsified EOs [27]. As can been seen in Table 5, the EC 50 values obtained with the nanoemulsions are higher than with the A. nemorosa EO aerial parts. On the other hand, the mammalian control cell lines were also less sensitive giving a slightly better selective index. In fact, the selectivity index against the trypanosomes in comparison to the mammalian cells was better with NANO A than with the A. nemorosa EO-AP (SI = 8.42 for NANO A and 4.65 for the A. nemorosa EO-AP). The surfactant itself that was used to make the nanoemulsions (polysorbate 80) was inactive in the concentration range tested.
To hypothesize a potential mechanism of action of A. nemorosa EO-AP and A. nemorosa EO-R, NTP and ADP pools were measured by HPLC. As shown in Figure 5, the treatment of the parasites led to a general decrease in the NTP pools and increased relative ADP levels. The increased ADP levels is a sign of disturbed energy metabolism, indicating that the cell is not able to phosphorylate ADP to ATP efficiently. A consequence of the decreased ATP pools is that other NTP pools also decrease because they are metabolically linked to each other via NDP kinase and other enzymes. NDP kinase uses NTPs to phosphorylate NDPs. In a normal situation, it will most often use ATP as a phosphate donor because it has the highest concentration in the cell, but it can also use other NTPs depending on their abundance. The decreased pools of other NTP pools is therefore also likely to be a consequence of the disturbed energy metabolism. Another factor that can affect the measured nucleotide levels is that the cells are dying and that the number of living cells is, therefore, less in the end of the incubation period as compared to the untreated cells. We have tried to minimize this factor by keeping the incubation period short. However, with the higher dose of EO, the NTP pools were decreased even more, but with no further increase in the ADP pools indicating that cell death can be a contributing factor also in this short time frame if the EO concentration is high enough. A disturbed energy metabolism is very harmful to the cell, which needs ATP to keep intracellular ion concentrations intact.

Hydrodistillation
Hydrodistillation (HD) was performed on dried samples of aerial parts and roots of A. nemorosa. In both cases, we used 2.5 kg of plant material soaked in a 20 L glass flask filled with 12 L of water, heated by a mantle system Falc MA (Falc Instruments, Treviglio, Italy) for 5 h. The EOs were recovered by a glass Clevenger-type apparatus and stored on the fridge before chemical characterization and biological evaluation. The EO yield was calculated on a dry-weight basis.

GC-MS Analysis
The chemical composition of A. nemorosa EOs was analyzed through an Agilent 6890N GC-MS system coupled to a 5973N single quadrupole detector mass spectrometer. The separation was provided by an HP-5MS capillary column (5% phenylmethylpolysiloxane, 30 m × 0.25 mm i.d., 0.1 µm f.t., Agilent, Santa Clara, CA, USA). The temperature program was as follows: 60 • C for 5 min, then 4 • C min −1 up to 220 • C, finally 11 • C min −1 to 280 • C, maintained for 15 min, for a total run time of 65 min. The temperature of the injector and detector was 280 • C. Helium (He) was the carrier gas, with a flow rate of 1 mL min −1 and a 1:50 split ratio. The chromatograms were acquired in full scan in the range 29.0-400.0 uma, using electron-impact (EI, 70 eV) mode. Dilution 1:100 of essential oils in n-hexane was injected (2 µL) into the GC-MS system. The MSD ChemStation (Agilent, Santa Clara, CA, USA, Version G1701DA D.01.00) and the NIST Mass Spectral Search Program were employed for the data analysis. Identifying the principal compounds was achieved by the correspondence of retention indices and mass spectra to those of ADAMS, NIST 17, FFNSC2, and WILEY 275 libraries. Furthermore, the analytical standards available in the laboratory (Sigma-Aldrich, Milan, Italy) were used for further confirmation. The relative peak area percentages were obtained by area normalization without using correction factors [28].

Growth Inhibition Assay on T. brucei and Balb/3T3 Cells
All the compounds that were tested were dissolved in dimethyl sulfoxide (DMSO). The final DMSO concentration in each well was less than 1% in all experiments and has no effect on the trypanosomes in this concentration range. They were serially diluted in 96-wells microtiter plates with growth medium to obtain a range of concentrations from 2 × 10 −5 µg/mL to 200 µg/mL (100 µL/well). In each well 100 µL of T. brucei or mammalian fibroblasts cell culture were added to have 20,000 and 2000 cells per well, respectively. The plates were incubated for 48 h at 37 • C with 5% CO 2 , treated with 20 µL of 0.5 mM resazurine (Sigma-Aldrich, St Louis, MO, USA) and treated for 24 h at the incubation conditions. Subsequently, the surviving cells were quantified by fluorescence (540 nm excitation and 590 nm emission) with an Infinite M200 microplate reader (Tecan Group, Ltd. Männedorf, Switzerland). Data were analyzed using the GraphPad Prism 5.04 software to obtain EC 50 values by fitting the data to a log inhibitor vs. response curve (variable slope, four parameters) and calculating the selectivity index (SI) from the comparison of T. brucei and Balb/3T3 EC 50 values.

Comparative Activities
To evaluate individual constituents' contribution to overall activity, a compoundelimination assay was conducted as previously described by Miresmailli [29]. A series of artificial EOs were prepared using the constituents in their natural proportions, either as the full mixture or with one, two, or three constituents omitted. Each artificial EO had the same amounts in weight of individual compounds as in the complete EO, with the amount of missing compounds replaced by DMSO.

Synergistic Interactions among Four Major Constituents
To evaluate potential synergies between the six major constituents (farnesene, βocimene, p-cymene, limonene, β-pinene, myrcene), mixtures were prepared following the actual constituent ratio based on chemical analysis of the EO. These artificial oils were applied to T. brucei or mammalian cell culture, and their EC 50 values were calculated after 48 h. To determine the mixtures' relationships, we used a statistical model, named Wadley's model [30], to compare expected and observed EC 50 values. Based on Wadley's calculation, the expected EC 50 values (assuming additive interaction) were determined from the equation: where a is the proportion of compound A in the mixture, and EC 50(a) is the EC 50 of compound A. The interaction between the observed and theoretical EC 50 values was compared as: The relationship between the constituents of the mixture was defined as either synergistic (when R > 1.5), additive (1.5 ≥R > 0.5), or antagonistic (R ≤ 0.5) based on this model.

Preparation of Anthriscus nemorosa EO-Based Nanoemulsions
For the preparation of NANO A, A. nemorosa EO (from aerial parts) was added dropwise to a Polysorbate 80 aqueous solution under high-speed stirring (Ultraturrax T25 basic, IKA ® Werke GmbH & Co.KG, Staufen, Germany) for 5 min at 9500 rpm to form an emulsion with final concentrations of 6% (w/w) EO and 2% (w/w) polysorbate. The obtained emulsion was then homogenised by using a French Pressure Cell Press (American Instrument Company, Aminco, MY, USA) for four cycles at the pressure of 130 MPa. NANO B was formulated in a similar way but a 1:2 (w/w) mixture of ethyl oleate and EO was used in the first step instead of the pure EO.

Determination of NTP and ADP Pools in T. brucei
Blood-stream forms of T. b. brucei s427 were maintained at 37 • C and 5% CO 2 in Hirumi's modified Iscove's Medium (HMI)-9 medium lacking thymidine and Serum Plus but containing 10% fetal bovine serum. The omission of thymidine does not affect the growth of the parasites [32]. Trypanosomes (50 mL, 10 6 cells/mL), harvested in late logarithmic phase, were chilled on ice for 5 min before being collected and centrifuged at 3000× g for 5 min at 4 • C. Subsequently, the liquid was decanted and the pellet resuspended in the remaining liquid in the tube by tapping the bottom of the tube. The resuspended slurry was transferred to an Eppendorf tube, and centrifuged at 16,000× g for 1 min at 4 • C. The supernatant was discarded and the collected trypanosomes were disintegrated by pipetting them up and down in 500 µL of ice-cold 0.6 M trichloroacetic acid containing 15 mM MgCl 2 . The resultant solution was centrifuged at 16,000× g for 1 min at 4 • C, and the supernatant was extracted with 720 µL chloroform (78% v/v)-trioctylamine (22% v/v). Each sample was purified by an OASIS WAX cartridge as described previously [33] and the collected solution was evaporated to dryness in a Speedvac (Thermo Fisher Scientific Savant, Waltham, MA, USA) and dissolved in 200 µL of water. This fraction was analyzed by HPLC using the Fast Protocol [33] to quantify ADP and NTPs with a 150 × 3 mm C18-WP HPLC column (ChromaNik Technologies Inc., Osaka, Japan). The mobile phase was a ternary mixture of 43% solution A (23 g/L KH 2 PO 4 pH 5.6 in 5.8% acetonitrile); 37% solution B (5.8% acetonitrile) and 20% solution C (3.52 g/L tetrabutylammonium bromide in 5.8% acetonitrile) and the column temperature was 30 • C. The nucleotides were quantified by measuring areas and comparing them to the areas of a 1 µM standard nucleotide mixture.

Conclusions
In conclusion, the overall results of our work have shed light on the biological properties of A. nemorosa EOs, farnesene and artificial mixtures (Mix 3-5 in Table 3) as relevant sources of bioactive substances and highlighted the opportunity to synthesize other farnesene-based derivatives as prototypes for future antiprotozoal drug development.
To our knowledge, this is the first report on the antiprotozoal activity of A. nemorosa EO and the sesquiterpene hydrocarbon farnesene. The EO had a strong effect on the T. brucei energy metabolism and caused a major decrease in NTP pools and a concomitant increase in ADP. However, further studies are needed to understand the more exact mode of action and the possible molecular targets related to the antitrypanosomal activity of A. nemorosa EOs, the artificial mixtures of EO components, and farnesene.
This study also highlights the significant efficacy of A. nemorosa EO nanoemulsions to develop long-lasting and effective prototype formulations active against T. brucei. The poor hygiene conditions of the affected geographic areas and the lack of proper facilities require straightforward and safe administration methods, and nanoemulsions represent one of these.

Author Contributions:
The contributions of the respective authors are as follows: F.M., R.P., L.C. and A.H., conceptualization as well as interpreting the data to prepare the manuscript; E.O., E.M. and L.A., collecting plant material, hydrodistillation, and GC analyses; D.R.P. and G.B., preparation and characterization of Anthriscus nemorosa EO-Based Nanoemulsions; G.F., C.B., F.R. and A.H., conducted the tests for antitrypanosomal activity as well as interpreting the relative data; R.P., S.D., F.M., A.H., L.C. and C.B., drafting, reviewing and editing. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.