Neutrophil Immunomodulatory Activity of Farnesene, a Component of Artemisia dracunculus Essential Oils

Despite their reported therapeutic properties, not much is known about the immunomodulatory activity of essential oils present in Artemisia species. We isolated essential oils from the flowers and leaves of five Artemisia species: A. tridentata, A. ludoviciana, A. dracunculus, A. frigida, and A. cana. The chemical composition of the Artemisia essential oil samples had similarities and differences as compared to those previously reported in the literature. The main components of essential oils obtained from A. tridentata, A. ludoviciana, A. frigida, and A. cana were camphor (23.0–51.3%), 1,8-cineole (5.7–30.0%), camphene (1.6–7.7%), borneol (2.3–14.6%), artemisiole (1.2–7.5%), terpinen-4-ol (2.0–6.9%), α-pinene (0.8–3.9%), and santolinatriene (0.7–3.5%). Essential oils from A. dracunculus were enriched in methyl chavicol (38.8–42.9%), methyl eugenol (26.1–26.4%), terpinolene (5.5–8.8%), (E/Z)-β-ocimene (7.3–16.0%), β-phellandrene (1.3–2.2%), p-cymen-8-ol (0.9–2.3%), and xanthoxylin (1.2–2.2%). A comparison across species also demonstrated that some compounds were present in only one Artemisia species. Although Artemisia essential oils were weak activators of human neutrophils, they were relatively more potent in inhibiting subsequent neutrophil Ca2+ mobilization with N-formyl peptide receptor 1 (FPR1) agonist fMLF- and FPR2 agonist WKYMVM, with the most potent being essential oils from A. dracunculus. Further analysis of unique compounds found in A. dracunculus showed that farnesene, a compound with a similar hydrocarbon structure as lipoxin A4, inhibited Ca2+ influx induced in human neutrophils by fMLF (IC50 = 1.2 μM), WKYMVM (IC50 = 1.4 μM), or interleukin 8 (IC50 = 2.6 μM). Pretreatment with A. dracunculus essential oils and farnesene also inhibited human neutrophil chemotaxis induced by fMLF, suggesting these treatments down-regulated human neutrophil responses to inflammatory chemoattractants. Thus, our studies have identified farnesene as a potential anti-inflammatory modulator of human neutrophils.


Introduction
The genus Artemisia is one of the most broadly distributed groups in the Asteraceae family [1]. Within this family, Artemisia is in the tribe Anthemideae and comprises about 530 species of plants that are found on all continents except for Antarctica. These plants are found mainly in the Northern Hemisphere, with only 25 species in the Southern Hemisphere [2]. According to the latest taxonomic revisions, there are 50 Artemisia species in North America, which belong to Artemisia (Miller) Less, Absinthium (Miller) Less, Dracunculus Besser, and Tridentatae (Rydb.) subgenera [3]. Artemisia is most prevalent in the western half of the United States, with 11 species in or near the Rocky Mountain region [4].

Essential Oil Composition
Essential oils were obtained from the flowers (designated by Fl) and leaves (designated by Lv) of five Artemisia species for subsequent phytochemical and biological characterization. Plant material was collected from wild plants in 2019 around Bozeman and Three Forks (MT, USA) ( Table 1). After botanical identification of the plant material, the flowers and leaves were dried for 7-10 days at room temperature but protected from direct sunlight.
The yields (v/w) of Artemisia spp. essential oils ranged from 0.3 to 3.3% (Table 1). The composition of the extracted essential oils was determined using simultaneous GC-FID and GC/MS. Twenty-seven major compounds found in the essential oils are shown in Table 2 and Supplementary Table S1 shows the total of all compounds identified in these samples. The main classes of compounds in all Artemisia species were oxygenated monoterpenes, ranging from 1.9 (AD Fl ) to 84.8% (AC Lv ), and monoterpene hydrocarbons, which ranged from 6.5 (AC Lv ) to 25.8% (AD Fl ). Essential oils from A. tridentata, A. ludoviciana, A. frigida, and A. cana contained high amounts of oxygenated monoterpenes (77.4-84.8%), with 1,8-cineole and camphor being the most representative, which is consistent with previous reports for A. ludoviciana, A. frigida, and A. cana essential oils [26]. In addition, essential oils from the A. dracunculus were enriched in phenylpropanoids (65.3-69.0%), including methyl chavicol (estragole) and methyl eugenol, also supporting previous reports analyzing essential oils from aerial parts of this plant [26,27]. Note that this is the first report comparing individual leaf and flower essential oils from these Artemisia species. Artemisia exhibits significant intraspecies variations in terpenes present in their essential oils [16,26,. For example, the variation in volatile components of these plants may occur during plant maturation at different altitudes. Our studies confirmed these findings and extended this analysis by providing the first report comparing essential oils isolated from preparations of A. tridentata flowers and leaves prepared from the same plant samples. Interestingly, we found that some compounds exhibited remarkable differences in terms of their presence in different parts of the same Artemisia species. For example, santolinatriene and artemiseole were found only in the flowers but not in the leaves of A. ludoviciana. Likewise, santolina epoxide and the monoterpene alcohol grandisol, a pheromone, were detected only in A. cana flowers. Similarly, trans-chrysanthenol and its ester trans-chrysanthenyl acetate were detected only in A. frigida leaves, and fragranol was found only in A. cana leaf essential oils and not those isolated from flowers. Therefore, it is important to evaluate different parts of plants in terms of target compounds. A comparison across species also demonstrated some differences in the presence of compounds, with some compounds (e.g., (Z)-β-ocimene, methyl chavicol, (E,E)-α-farnesene, xanthoxylin), being present in only one Artemisia species. It is also interesting that A. dracunculus essential oils had more unique compounds that were not present in essential oils from the other Artemisia species. Neutrophils are the most common leukocyte in human blood and are required for innate immune responses [59]. These cells respond rapidly to infection and injury in various tissues. They represent the first line of defense and utilize multiple mechanisms of oxygen-dependent and oxygen-independent processes to protect against infection, including phagocytosis, cytokine secretion, and reactive oxygen species production [60,61]. Thus, neutrophils represent a potential target for the development of novel anti-inflammatory therapeutics. Indeed, natural products, such as essential oils, have exhibited neutrophil immunomodulatory activity [20][21][22][23][24].
Artemisia essential oils were evaluated for their immunomodulatory effects on human neutrophils. Initially, we measured the effects of the essential oils on [Ca 2+ ] i , which is a key component of neutrophil activation and function [62]. We found that treatment of neutrophils with Artemisia essential oils only modestly enhanced [Ca 2+ ] i at relatively high concentrations, with EC 50 values ranging from in the range of 15.8 µg/mL (AL Lv ) to 48.7 µg/mL (AC Lv ). We also evaluated some of the individual compounds present in the various Artemisia essential oils to determine if they were responsible for neutrophil activation, including 1,8-cineole, (+)-camphor, (−)-camphor, (±)-bornyl acetate, farnesene, piperitone, and xanthoxylin. As shown in Table 3, 1,8-cineole, (+)-camphor, (−)-camphor, and piperitone did not affect human neutrophil Ca 2+ flux, whereas farnesene, xanthoxylin, and (±)-bornyl acetate were weak activators, but only at relatively high concentrations. Note that previous analysis of some of the other major compounds that we also found here to be present in Artemisia essential oils (i.e., α-pinene, camphene, (E/Z)-β-ocimene, p-cymene, terpinolene, linalool, terpinen-4-ol, methyl chavicol, p-cymen-8-ol, and methyl eugenol) showed that they had no effect on human neutrophil [Ca 2+ ] i [24,25].  Since Artemisia essential oils and individual compound components exhibited only weak neutrophil activation, we considered whether they might have some effect on neutrophils activated by strong agonists, such as N-formyl peptide receptor 1 (FPR1) agonist f MLF and FPR2 agonist WKYMVM, as ligands can downregulate or alter neutrophil responses to subsequent treatment with heterologous or homologous agonists [63]. As shown in Table 3, Artemisia essential oils were relatively more potent inhibitors of [Ca 2+ ] i in f MLFand WKYMVM-stimulated neutrophils, with some IC 50 Figure 1A.

Effect of A. dracunculus Essential Oils and Pure Compounds on Neutrophil Chemotaxis
Since A. dracunculus essential oils and their unique component, farnesene, were effective inhibitors of agonist-induced neutrophil [Ca 2+ ] i , we evaluated their effects on neutrophil chemotaxis. Pretreatment of neutrophils with AD Fl for 30 min dose-dependently inhibited f MLF-induced human neutrophil chemotaxis toward f MLF, with an IC 50 of 2.4 ± 1.1 µg/mL (Figure 2A). Likewise, pretreatment of neutrophils with farnesene also effectively inhibited neutrophil migration toward f MLF, with an IC 50 of 3.3 ± 1.4 µM. A representative dose-dependent response for the inhibition of f MLF-induced chemotaxis in human neutrophils by farnesene is shown in Figure 2B. Analysis of (±)-bornyl acetate showed that this compound also inhibited f MLF-induced neutrophil chemotaxis, although with much lower potency (IC 50 = 11.9 ± 1.9 µM). In contrast, xanthoxylin did not affect this response.
trophil chemotaxis. Pretreatment of neutrophils with ADFl for 30 min dose-dependently inhibited fMLF-induced human neutrophil chemotaxis toward fMLF, with an IC50 of 2.4 ± 1.1 µg/mL (Figure 2A). Likewise, pretreatment of neutrophils with farnesene also effectively inhibited neutrophil migration toward fMLF, with an IC50 of 3.3 ± 1.4 µM. A representative dose-dependent response for the inhibition of fMLF-induced chemotaxis in human neutrophils by farnesene is shown in Figure 2B. Analysis of (±)-bornyl acetate showed that this compound also inhibited fMLF-induced neutrophil chemotaxis, although with much lower potency (IC50 = 11.9 ± 1.9 M). In contrast, xanthoxylin did not affect this response. To ensure that Artemisia essential oils or active pure compounds were not cytotoxic, we evaluated the cytotoxicity of the essential oil samples (up to 55 µg/mL) and test compounds at various concentrations (up to 50 µM) in human neutrophils during a 1.5 h incubation period, which covers the times used to measure Ca 2+ influx (up to 30 min) and cell migration (up to 1.5 h). As shown in Table 3, essential oils from A. tridentata, A. ludoviciana, and A. cana had no cytotoxicity in neutrophils, although essential oils from A. dracunculus and A. frigida had a little cytotoxicity at the highest concentrations (25-50 M), To ensure that Artemisia essential oils or active pure compounds were not cytotoxic, we evaluated the cytotoxicity of the essential oil samples (up to 55 µg/mL) and test compounds at various concentrations (up to 50 µM) in human neutrophils during a 1.5 h incubation period, which covers the times used to measure Ca 2+ influx (up to 30 min) and cell migration (up to 1.5 h). As shown in Table 3, essential oils from A. tridentata, A. ludoviciana, and A. cana had no cytotoxicity in neutrophils, although essential oils from A. dracunculus and A. frigida had a little cytotoxicity at the highest concentrations (25-50 µM), which were >10-fold higher than those required to induce biological effects. The pure compounds (±)-bornyl acetate, farnesene, and xanthoxylin were nontoxic in human neutrophils during the 90 min incubation. In addition, these compounds did not affect the viability of human THP-1 monocytic cells after a 24 h incubation period (Figure 3). Thus, the biological effects of these compounds on neutrophil function are not due to compound cytotoxicity and support the conclusion that farnesene is a novel innate immunomodulator.

Identification of Potential Protein Targets for Selected Compounds
Both farnesene and bornyl acetate inhibited neutrophil [Ca 2+ ] i and chemotaxis, although farnesene was the most effective by far. Farnesene encompasses a set of six closely related sesquiterpenes, as the α-form can exist as four stereoisomers, and the β-isomer exists as two stereoisomers. We identified both (Z,E)-α-farnesene and (E,E)-α-farnesene in A. dracunculus, although (E,E)-α-farnesene was present at a higher level (Table 2). Both αand βforms of farnesene have been reported previously in essential oils from Artemisia spp. For example, (Z,E)-α-farnesene was found in A. maderaspatana L. [64] and (E,E)-α-farnesene was present in A. ordosica Krasch. and A. dracunculus [26,65,66]. (E)-β-farnesene was re- ported to be present in A. annua L. [67], A. magellanica Sch. Bip. [68], A. absinthium L. [69], A. lavandulifolia DC. [70], and A. biennis Willd. essential oils [26]. Finally, (Z)-β-farnesene was found in essential oils from A. absinthium [26], A. annua [71], and A. tournefortiana Reichb. [72]. Although the αand β-forms of farnesene are well-known pheromones [73,74], we did not find any reports on the biological activity of these compounds in human cells. which were >10-fold higher than those required to induce biological effects. The pure compounds (±)-bornyl acetate, farnesene, and xanthoxylin were nontoxic in human neutrophils during the 90 min incubation. In addition, these compounds did not affect the viability of human THP-1 monocytic cells after a 24 h incubation period (Figure 3). Thus, the biological effects of these compounds on neutrophil function are not due to compound cytotoxicity and support the conclusion that farnesene is a novel innate immunomodulator. The data presented are the mean ± SD of triplicate samples from one experiment. Representative of two independent experiments with similar results.

Identification of Potential Protein Targets for Selected Compounds
Both farnesene and bornyl acetate inhibited neutrophil [Ca 2+ ]i and chemotaxis, although farnesene was the most effective by far. Farnesene encompasses a set of six closely related sesquiterpenes, as the α-form can exist as four stereoisomers, and the β-isomer exists as two stereoisomers. We identified both (Z,E)-α-farnesene and (E,E)-α-farnesene in A. dracunculus, although (E,E)-α-farnesene was present at a higher level ( Table 2). Both α-and β-forms of farnesene have been reported previously in essential oils from Artemisia spp. For example, (Z,E)-α-farnesene was found in A. maderaspatana L. [64] and (E,E)-αfarnesene was present in A. ordosica Krasch. and A. dracunculus [26,65,66]. (E)-β-farnesene was reported to be present in A. annua L. [67], A. magellanica Sch. Bip. [68], A. absinthium L. [69], A. lavandulifolia DC. [70], and A. biennis Willd. essential oils [26]. Finally, (Z)-βfarnesene was found in essential oils from A. absinthium [26], A. annua [71], and A. Bornyl acetate has been reported previously to be a major component of essential oils isolated from Artemisia spp, including A. ludoviciana and A. frigida [17,53]. Bornyl acetate is also a volatile constituent present in numerous conifer essential oils and essential oils from Valeriana officinalis [75]. Previous studies have reported that bornyl acetate has anti-inflammatory properties in different experimental models. For example, bornyl acetate downregulated the levels of proinflammatory cytokines in vitro and in vivo and reduced the number of total cells, neutrophils, and macrophages in murine bronchoalveolar lavage fluid after intranasally injected lipopolysaccharide. Moreover, bornyl acetate inhibited the expression and production of IL-1β and TNF in human umbilical vein endothelial cells (HUVECs) and RAW 264.7 macrophages [76,77]. Bornyl acetate has also been reported to have analgesic effects [78].
To better understand the properties of these compounds, we calculated their most important physicochemical parameters using SwissADME [79]. The LogP values estimated using ALOGPS 2.1 program [80] and tPSA values allowed us to predict that (−) bornyl acetate, (+) bornyl acetate, (Z,E)-α-farnesene, and (E,E)-α-farnesene can all permeate the blood-brain barrier (BBB), according to the binary classification tree model [81] (Table 4).  Farnesene is very lipophilic (Table 4). Thus, we propose that the neutrophil signaling inhibitory mechanisms may be based on the allosteric interaction of the farnesene chain with the membrane portion of the receptor, and we are currently investigating this mechanism. Indeed, other lipophilic compounds, such as the bile acids deoxycholic acid and chenodeoxycholic acid, have been reported to antagonize FPR1 at high concentrations (>100 µM) [82][83][84]. Moreover, lipoxin A 4 was also reported as an allosteric modulator of a CB1 cannabinoid receptor and FPR2 [85,86]. To further investigate this issue, we aligned (E,E)-α-farnesene, (Z,E)-α-farnesene, and lipoxin A 4 using FieldTemplater software ( Figure 4) and found that the alignment was determined mainly by the hydrophobic hydrocarbon skeletons of the compounds. The combined similarity measure of the superimposition was relatively high (S = 0.690), suggesting that the farnesene molecules could mimic lipoxin A 4 and maybe other related specialized pro-resolving mediators (known as SPMs), including resolvins, maresins, and protectins [87]. Indeed, many of these molecules have been demonstrated to act allosterically on a number of GPCRs (reviewed in [88]). Interestingly, we found previously that 6-methyl-3,5-heptadien-2-one (MHDO), a compound structurally similar to farnesene, also inhibited neutrophil activity [25], although its molecular targets were not identified. structurally similar to farnesene, also inhibited neutrophil activity [25], although its molecular targets were not identified. Not much is known about the specific cellular targets of farnesene or bornyl acetate. Thus, we performed reverse-pharmacophore mapping on the molecular structures of (Z,E)-α-farnesene, (E,E)-α-farnesene, and the (−) and (+) forms of bornyl acetate to identify potential biological targets. PharmMapper compared a database of pharmacophore patterns with these compounds and generated target information, such as pharmacophoric characteristics and normalized fitness scores. Note, however, that PharmMapper depends on the availability of structures for pharmacophore mapping, and most GPCRs are not represented. The proper optical isomers of the compounds were submitted to the Phar-mMapper server as mapping explicitly accounts for the three-dimensional structure of a Not much is known about the specific cellular targets of farnesene or bornyl acetate. Thus, we performed reverse-pharmacophore mapping on the molecular structures of (Z,E)α-farnesene, (E,E)-α-farnesene, and the (−) and (+) forms of bornyl acetate to identify potential biological targets. PharmMapper compared a database of pharmacophore patterns with these compounds and generated target information, such as pharmacophoric characteristics and normalized fitness scores. Note, however, that PharmMapper depends on the availability of structures for pharmacophore mapping, and most GPCRs are not represented. The proper optical isomers of the compounds were submitted to the Phar-mMapper server as mapping explicitly accounts for the three-dimensional structure of a molecule. The ten top-ranked potential targets found by PharmMapper are shown in Table 5.  Legend: AKR1C2, aldo-keto reductase family 1 member C2; Apo A-II, apolipoprotein A-II; BMP2, bone morphogenetic protein 2; CA2, carbonic anhydrase 2; ERK2, extracellular signal-regulated kinase 2; JNK1, c-Jun N-terminal kinase 1; KIF11, kinesin family member 11; LCE1, liver carboxylesterase 1; MAO-B, monoamine oxidase B; NR1A2, thyroid hormone receptor β; NR3C4, androgen receptor; p38α, p38α mitogen-activated protein kinase; PDPK1, 3-phosphoinositide-dependent protein kinase 1; PgR, progesterone receptor; Pim-1, proto-oncogene serine/threonine-protein kinase; PPIase A, peptidyl-prolyl cis-trans isomerase A; TGFBR1, TGF-β receptor type-1.

(E,E)-α-Farnesene (Z,E)-α-Farnesene
PharmMapper analysis indicated that kinases could be among the potential targets for bornyl acetate and farnesene. Among the top ten ranked targets for the (±)enantiomers of bornyl acetate were c-Jun N-terminal kinase 1 (JNK1), 3-phosphoinositidedependent protein kinase 1 (PDPK1), and transforming growth factor-β receptor type-1 (TGFBR1), a serine/threonine-protein kinase. Likewise, JNK1, p38α mitogen-activated protein kinase (MAPK), extracellular signal-regulated kinase 2 (ERK2), and proto-oncogene serine/threonine-protein kinase (Pim-1) were kinases identified among the ten top-ranked targets for (E,E)-and (Z,E)-forms of α-farnesene (Table 5). MAPK signaling plays an important role in neutrophil signal transduction cascades [89], and studies have shown that members of the MAPK, JNK, and the p38 MAPK families of proteins are activated in response to neutrophil priming/activation (reviewed in [90]). It is also clear from previous studies that one or more of these MAPK pathways is induced by FPR agonists [91,92] and IL-8 [93]. Thus, these compounds and especially farnesene may be general inhibitors of neutrophil activation through GPCRs, and further studies are in progress to evaluate this idea and identify their specific molecular targets.

Essential Oil Extraction
Essential oils were obtained by hydrodistillation of air-dried plant material. Hydrodistillation was performed with a Clevenger-type apparatus, as previously described [25]. To avoid artifacts, we used conditions accepted by the European Pharmacopoeia to avoid artifacts. Essential oil yields (w/v) were calculated based on the amount of air-dried plant material used. We prepared stock solutions of the essential oils in 10 mg/mL DMSO for biological evaluation. For gas-chromatographic (GC) analysis, samples were prepared in 10% w/v n-hexane.

GC-Flame Ionization Detector (GC-FID) and GC-Mass Spectrometry (GC-MS) Analysis
We performed GC-MS analysis using an Agilent 5975 GC-MSD system, as previously described [94]. An Agilent Innowax FSC column (60 m × 0.25 mm, 0.25 µm film thickness) was used with He as the carrier gas (0.8 mL/min). The GC oven temperature was kept at 60 • C for 10 min, increased to 220 • C at a rate of 4 • C/min, kept constant at 220 • C for 10 min, and then increased to 240 • C at a rate of 1 • C/min. The split ratio was adjusted to 40:1, and the injector temperature was 250 • C. MS spectra were monitored at 70 eV with a mass range of 35 to 450 m/z. GC analysis was carried out using an Agilent 6890 N GC system. To obtain the same elution order as with GC-MS, the line was split for FID and MS detectors, and a single injection was performed using the same column and appropriate operational conditions. The flame ionization detector (FID) temperature was 300 • C. The essential oil components were identified by co-injection with standards (whenever possible), which were purchased from commercial sources or isolated from natural sources. The identities of compounds were also using the MassFinder software 4.0 (Dr. Hochmuth Scientific Consulting, Hamburg, Germany), Adams Library, Wiley GC/MS Library (Wiley, Hoboken, NJ, USA), and NIST Library. Confirmation was also achieved using the in-house "Başer Library of Essential Oil Constituents" database. The database was created using chromatographic analysis of pure compounds run under identical conditions. Samples were spiked with a C 8 -C 40 n-alkane standard solution (Fluka, Buchs, Switzerland) to determine relative retention indices (RRI). The FID chromatograms were used to calculate the relative amounts (%) of the separated compounds.

Isolation of Human Neutrophils
We obtained human neutrophils using blood collected from healthy donors. Blood collection was approved by the Institutional Review Board at Montana State University (Protocol #MQ041017). Neutrophils were isolated as described previously [95]. The isolated neutrophils were resuspended in HBSS -. Neutrophil preparations were routinely >95% pure and >98% viable, as determined by light microscopy and trypan blue exclusion, respectively.

Ca 2+ Mobilization Assay
Changes in intracellular Ca 2+ concentrations ([Ca 2+ ] i ) were monitored with a FlexStation 3 (Molecular Devices, Sunnyvale, CA, USA). For these assays, neutrophils were loaded with 1.25 µg/mL Fluo-4AM and incubated in the dark at 37 • C for 30 min. The cells were then washed with HBSS -. The dye-loaded cells were resuspended in HBSS + and pipetted into the wells of black microtiter plates at 2 × 10 5 cells/well. To measure the direct effects of samples on [Ca 2+ ] i , the test samples were added to the wells (final concentration of DMSO was 1%), and fluorescence was monitored (λ ex = 485 nm, λ em = 538 nm). Changes in fluorescence were monitored every 5 s at room temperature for 240 s. To evaluate the inhibitory effects of the test samples, the samples were added to the wells and incubated for 10 min, with the subsequent addition of 5 nM f MLF, 5 nM WKYMVM, or 25 nM IL-8. Responses were normalized to the response induced by control 5 nM f MLF, 5 nM WKYMVM, or 25 nM IL-8 alone without pretreatment. These responses were assigned as 100%. To calculate median effective concentrations (EC 50 or IC 50 ), we used curve fitting (at least five or six points) and nonlinear regression analysis of the dose-response curves. Curve fitting was performed with Prism 9 (GraphPad Software, Inc., San Diego, CA, USA).

Chemotaxis Assay
To evaluate effects of samples on neutrophil migration, we resuspended neutrophils in HBSS + containing 2% (v/v) heat-inactivated FBS (2 × 10 6 cells/mL). We analyzed chemotaxis using 96-well ChemoTx chambers (Neuroprobe, Gaithersburg, MD). The neutrophils were first preincubated with the indicated concentrations of test samples at room temperature for 30 min. The pretreated cells were then pipetted into the chamber upper wells (4 × 10 4 cells/well). The lower wells contained 30 µL of HBSS + with 2% (v/v) heatinactivated FBS, the indicated test samples or control 1% DMSO, and 1 nM f MLF as the chemoattractant. Three lower wells were reserved for background controls (DMSO treated cells in the upper wells and DMSO without f MLF in the lower wells). We allowed the cells to migrate through the polycarbonate membrane filter for 60 min at 37 • C/5% CO 2 . We determined the number of migrated cells by measuring ATP in lysates of transmigrated cells and comparing this to a standard curve obtained with known neutrophil numbers, as described previously [23]. Calculation of median effective concentrations (IC 50 ) was performed by nonlinear regression analysis of the dose-response curves.

Cytotoxicity Assay
We analyze cytotoxicity in human neutrophils or THP-1 monocytic cells. Cytotoxicity was analyzed using a CellTiter-Glo Luminescent Cell Viability Assay Kit (Promega, Madison, WI, USA). Briefly, the cells were incubated (10 4 cells/well) with the indicated concentrations of essential oil or compound for 90 min (for neutrophils) or 24 h (for THP-1 cells) at 37 • C/5% CO 2 . After incubation, we added substrate. The samples were analyzed using a Fluoroscan Ascent FL microplate reader.

Molecular Modeling
The PharmMapper Server [96] was used to identify potential protein targets for (E,E)α-farnesene (CID: 5281516) and (Z,E)-α-farnesene (CID: 5362889) (CIDs are compound identifiers in PubChem). PharmMapper recognizes potential targets based on "invert" pharmacophore mapping. The protein biotargets are represented by sets of pharmacophore points in reference databases incorporated in the software. We used PubChem (https: //pubchem.ncbi.nlm.nih.gov; accessed on 20 February 2022) as a source of initial 3D structures for our compounds. These structures were downloaded from PubChem in SDF format. We then uploaded the structures into PharmMapper. The system automatically generated up to 300 conformers of each compound based on the software option. We performed pharmacophore mapping using the "Human Protein Targets Only" database, which contained 2241 targets. We retrieved the top 250 potential targets for each compound evaluated. The potential targets were sorted by normalized fit score. We calculated compound physicochemical properties using SwissADME (http://www.swissadme.ch; accessed on 20 March 2022). Properties were calculated for the structures of (−)-bornyl acetate (CID: 93009), (+)-bornyl acetate (CID: 6950274), (E,E)-α-farnesene (CID: 5281516), and (Z,E)-α-farnesene (CID: 5362889). The alignment of lipoxin-A 4 , (E,E)-α-farnesene, and (Z,E)-α-farnesene was made with the use of FieldTemplater software (Cresset Group, England, UK).

Statistical Analysis
For statistical analysis, we performed a one-way analysis of variance (ANOVA), followed by Tukey's pair-wise comparisons. We considered differences at p < 0.05 to be statistically significant.

Conclusions
Our results demonstrate that: (1) Artemisia spp. essential oils contain compounds that exhibit neutrophil immunomodulatory activity, which might contribute to the reported pharmacological properties of extracts from these plants; (2) the biological effects of these compounds on neutrophil function are not due to compound cytotoxicity; (3) essential oils from A. dracunculus flowers (AD Fl ) contain farnesene isomers that may have anti-inflammatory activity due to their ability to inhibit neutrophil responses to inflammatory chemoattractants, establishing farnesene as a novel innate immunomodulator, and (4) synergetic effects may be possible with other essential oil constituents, such as bornyl acetate or xanthoxylin. Further work is in progress to define the mechanisms of farnesene action and evaluate its therapeutic potential.