Abstract Artemisia argyi H. Lév. and Vaniot is a variety of Chinese mugwort widely cultured in central China. A. verlotorum Lamotte, another variety of Chinese mugwort, has been used in the southern region of China since ancient times. Despite their similar uses in traditional medicine, little is known about the differences in their active ingredients and potential benefits. Herein, the chemical compositions of the essential oils (EOs) from both varieties were analyzed using chromatography-mass spectrometry (GC-MS). A series of databases, such as the Traditional Chinese Medicine Systems Pharmacology database (TCMSP), SuperPred database and R tool, were applied to build a networking of the EOs. Our results revealed significant differences in the chemical compositions of the two Artemisia EOs. However, we found that they shared similar ingredient–target–pathway networking with diverse bioactivities, such as neuroprotective, anti-cancer and anti-inflammatory. Furthermore, our protein connection networking analysis showed that transcription factor p65 (RELA), phosphatidylinositol 3-kinase regulatory subunit alpha (PIK3R1) and mitogen-activated protein kinase 1 (MAPK1) are crucial for the biological activity of Artemisia EOs. Our findings provided evidence for the use of A. verlotorum as Chinese mugwort in southern China. Keywords: Artemisia argyi, Artemisia verlotorum, essential oils, chemical composition, networking pharmacology 1. Introduction Artemisia spps. are grown in Asia, North America and Europe [[36]1]. They are widely used as aromatic ethnic medicines in China, with their essential oils (EOs) being the major active ingredients frequently used in pharmaceuticals and cosmetics [[37]2]. Due to the synergistic effects of a number of their active components, Artemisia EOs have pleiotropic bioactivity [[38]3] and have attracted considerable attention for their biological diversities and bioactivities. Chinese mugwort, made from the leaves of A. argyi, has a long history of use in China for controlling dysmenorrhea, bleeding and eczema [[39]4,[40]5]. The essential oils extracted from A. argyi (AAEOs) are one of the major active ingredients [[41]6]. AAEOs have been shown to possess anti-oxidant [[42]7,[43]8], anti-inflammatory [[44]9], anti-microbial [[45]10], anti-fungal [[46]11], anti-histamine [[47]12], analgesic [[48]13], anti-tumor and immunomodulatory effects [[49]14]. It is noteworthy that A. verlotorum, a perennial plant widely distributed in the northern hemisphere [[50]15], has similar folk medicinal uses as A. argyi, and is also widely used to make Chinese mugwort in southern China, especially in Guangdong province [[51]16,[52]17]. However, the chemical constituents and activities of the essential oils of A. verlotorum (AVEOs) are still unclear. Herein, we analyzed the chemical ingredients of the EOs from different parts of A. argyi and A. verlotorum using gas chromatography-mass spectrometry (GC-MS) as the basis for potential activity exploration. We applied a networking pharmacology approach to study the interrelationships between the compounds, targets and related pathways for the EOs extracted from the leaves of A. argyi and A. verlotorum. The study involved a complex networking approach to investigate the interrelationships between the ingredients, target proteins and related signaling pathways of the EOs from both Artemisia. We further studied the statistical characteristics of the networking to explore the main active compounds and potential activities of the Artemisia EOs. Our results indicated that the ingredients of the EOs from the two Artemisia were different. The contents of monoterpenoids and sesquiterpenes could be used as markers to distinguish AAEOs from AVEOs. However, the networking pharmacology analysis suggested that the AAEOs and AVEOs have similar potential targets and pathways, indicating that they have similar biological activities. Our results provided evidence for the use of A. verlotorum as Chinese mugwort in southern China. 2. Results and Discussion 2.1. Chemical Compositions of EOs from Different parts of A. argyi and A. verlotorum A. argyi and A. verlotorum have been used as Ay Tsao or Chinese mugwort in China [[53]18]. A. argyi is known as northern Ay Tsao or northern Chinese mugwort, whereas A. verlotorum is called southern Chinese mugwort and is widely used by local people in southern China. Although many studies have reported the chemical constituents of A. argyi, few have reported on the chemical compositions of A. verlotorum. EOs are one of the major active ingredients of aromatic plants [[54]19]. Previous studies have suggested that the yield of AAEOs was higher than AVEOs and the EOs of their leaves were higher than that of the aerial part [[55]20]. However, systematic comparative studies of the constitutional differences between two Ay Tsao were unavailable. Herein, we analyzed the differences in the chemical compositions of EOs extracted from the whole grass of A. argyi (GAAEOs, YP-1), the leaves of A. argyi (LAAEOs, YP-2), the whole grass of A. verlotorum (GAVEOs, YP-3) and the leaves of A. verlotorum (LAVEOs, YP-4) using GC-MS ([56]Table 1, [57]Table S1, [58]Figure S5). Table 1. Yields of the EOs from different parts of two Artemisia spps. No. Latin Name Plant Parts Abbreviation Appearance of Essential Oils Collecting Locations Yield (v/w, %) YP-1 A. argyi whole grass GAAEOs blue Nanyang, Henan 0.15 YP-2 A. argyi leaves LAAEOs blue Nanyang, Henan 0.34 YP-3 A. verlotorum whole grass GAVEOs yellow-green Luofo mountain, Guangdong 0.015 YP-4 A. verlotorum leaves LAVEOs yellow-green Luofo mountain, Guangdong 0.032 [59]Open in a new tab A total of 132 volatile compounds of the Artemisia essential oils were identified using GC-MS, representing 75–86% of the total compounds ([60]Table S1, [61]Figure 1). The Artemisia EOs were abundant in terpenoids, including 70 sesquiterpenoids (45 sesquiterpenes, 18 sesquiterpenic alcohols, three sesquiterpenic ethers, two sesquiterpenic ketones and two sesquiterpenic esters), 37 monoterpenoids (nine monoterpenes, 15 monoterpenic alcohols, seven monoterpenic ketones, three monoterpenic esters, two monoterpenic ethers and one monoterpenic aldehyde) and three diterpenoids. Additionally, these EOs also contained five aromatics and 16 aliphatics. Among them, caryophyllene, caryophyllene oxide, neointermedeol and seven other ingredients were common components in the Artemisia essential oils. Diverse sesquiterpenoids accounted for the highest percentage of all Artemisia oils. The total contents of the sesquiterpeoids in the GAAEOs (38.42%), LAAEOs (52.01%), GAVEOs (63.56%) and LAVEOs (77.72%) were greater than 35%. Figure 1. [62]Figure 1 [63]Open in a new tab The chemical constituents of EOs from Artemisia. 2.1.1. Comparison and Analysis of the Chemical Composition of AAEOs and AVEOs The chemical compositions of the Vs were more abundant than the AVEOs and differed significantly. Among them, neointermedeol and eucalyptol were the main ingredients of the AAEOs ([64]Figure 2), while caryophyllene oxide was the primary component of the AVEOs ([65]Figure 3). Furthermore, the content of monoterpenoids from the AAEOs (>30%) was higher than that of the AVEOs (<5%). However, sesquiterpenes had a higher proportion in the AVEOs (>60%) than in the AAEOs (<52%). Moreover, diterpenoids were only found in the whole grass of both AAEOs and AVEOs. The contents of monoterpenoids and sesquiterpenes may be used as markers to distinguish AAEOs from AVEOs. Figure 2. [66]Figure 2 [67]Open in a new tab The chemical composition content of AAEOs. (A) The chemical composition content of essential oils whole grass from A. argyi. (B) The chemical composition content of essential oils leaves from A. argyi. Figure 3. [68]Figure 3 [69]Open in a new tab The chemical composition content of AVEOs. (A) The chemical composition content of essential oils whole grass from A. verlotorum. (B) The chemical composition content of essential oils leaves from A. verlotorum. 2.1.2. Comparison and Analysis of the EOs from Whole Grass and Leaves of A. Argyi The A. argyi used in this study was collected in Henan province, China. For the EOs of the whole grass of A. argyi (GAAEOs, YP-1), 65 volatile compounds were identified. Among them, 22 monoterpenoids (38.71%), 27 sesquiterpenoids (38.42%), one diterpenoids (0.11%), four aromatics (2.03%) and 11 aliphatics (3.11%) were detected. Neointermedeol (9.69%) was the major compound, followed by caryophyllene (8.73%), caryophyllene oxide (5.93%), cis-chrysanthenol (5.92%) and eucalyptol (5.77%) ([70]Figure 2A, [71]Table S1). It is noteworthy that the chemical compositions of the GAAEOs showed less consistency across different locations. For instance, artemisia alcohol (44.52%) and yomogi alcohol (24.08%) were the principal components of GAAEOs originating from Korea reported in the literature [[72]21]. The compounds and content of GAAEOs originating from different provinces in China also showed significant variation [[73]3,[74]18,[75]22]. For the leaves of A. argyi (LAAEOs, YP-2) ([76]Figure 2B, [77]Table S1), we identified a total of 59 compounds, including 26 monoterpenoids (31.29%), 26 sesquiterpenoids (52.01%), three aromatic compounds (1.58%) and four fatty compounds (1.03%). Among them, eucalyptol (11.51%), 2-borneol (10.09%), (−)-4-α-terpinenol (5.68%) and (+)-2-bornanone (5.14%) were the major monoterpenes. Sesquiterpenoids accounted for a higher proportion and contain more abundant structural skeletons, of which caryophyllene (9.24%), neointermedeol (8.02%) and caryophyllene oxide (3.42%) were the principal components. Moreover, a few aromatic compounds, such as eugenol (0.84%), were observed, as well as aliphatics, such as 1-octen-3-ol (0.34%). The chemical composition of the LAAEOs was relatively consistent with previous reports [[78]7,[79]18,[80]23]. Our results were consistent with a previous study on the constitute of LAAEOs, of which eucalyptol (9–33%) was considered a crucial constitute [[81]18,[82]24], as well as caryophyllene 2-borneol, (−)-4-α-terpinenol and humulene [[83]9,[84]25]. Additionally, artemisia alcohol (>30%) was reported to be a major component of LAAEOs collected from Korea [[85]21]. However, artemisia alcohol had a lower content, of 2.64%, in LAAEOs from China. The cultivation location was considered to be an important factor in affecting the secondary metabolites of AAEOs. Some studies have established the GC-MS fingerprint of LAAEOs collected from Qichun county in Hubei province, identifying 23 common components as characteristic compositions [[86]26]. However, only nine components, including eucalyptol, caryophyllene and borneol, were observed in our LAAEOs. Interestingly, all the principal components involved in the fingerprint, with the exception of thujone and alcanfor, were also detected in our EOs. Thujone, a neurotoxic substance commonly found in absinthe, was detected in multiple batches of the A. argyi samples [[87]27]. Many important factors, such as the cultivation conditions, harvesting time, storage, extraction method and detecting conditions, influence the chemical constituents of the EOs of A. argyi. The whole grasses of A. argyi were more chemically diverse than the leaves. The main chemical substances of the EOs from the two parts of A. argyi were similar, while the contents had some differences. There were 42 identical components, including neointermedeol, caryophyllene, caryophyllene oxide, etc., ([88]Table S2) in both parts of the EOs. Neointermedeol and caryophyllene were the principal components of the whole grass (9.69% and 8.73%, respectively). They were also present at a high content in the leaves (8.02% and 9.24%, respectively). The main constituents of the LAAEOs were eucalyptol (11.51%) and 2-borneol (10.09%). However, the content of these two ingredients in the GAAEOs was only 5.77% and 4.54%, respectively. Notably, the LAAEOs had fewer compounds, but a higher content of major components than the GAAEOs, which may be one of the reasons why the leaves have been commonly used as medicinal herbs. 2.1.3. Comparison and Analysis of the EOs from Whole grass and Leaves of A. verlotorum The A. verlotorum used in this study was collected in Guangdong province, China. For the whole grass of A. verlotorum (GAVEOs, YP-3) ([89]Figure 3A, [90]Table S1), we detected 47 ingredients, including two monoterpenoids (1.05%), 37 sesquiterpenoids (63.56%), two diterpenoids, one aromatic compounds (3.36%), four fatty compounds (6.68%) and one phosphate (0.42%). Among them, caryophyllene oxide (17.06%), (+)-β-eudesmol (6.67%), neointermedeol (4.72%), α-gurjunene (3.82%) and isospathulenol (3.77%) were the major sesquiterpenoids. Similarly to the LAVEOs, the GAVEOs had a higher content and diverse structure of sesquiterpenes compared to monoterpenes. It is clear that the yield and constitutes of Artemisia oils may be influenced by the cultivation location, harvest time, vegetative cycle stage, extraction method and selection of plant parts, based on preliminary studies [[91]18,[92]28]. A comparison of the variance of terpenoid biosynthesis among different parts of Artemisia from the gene perspective suggested that there was a significant difference in the expression pattern of genes [[93]29]. For the leaves of A. verlotorum (LAVEOs, YP-4) ([94]Figure 3B, [95]Table S1), we detected 55 volatile constituents. Among them, nine monoterpenoids (4.4%), 42 sesquiterpenoids (77.72%) and five aliphatics (2.83 %) were identified in the LAVEOs. Caryophyllene oxide was the major sesquiterpenoid in the LAVEOs, with more than 23% contents, followed by other sesquiterpenes, including humulene (5.87%), caryophyllene (3.80%) and δ-cadinene (3.55%). The sesquiterpenoids with significant structural diversity had a higher proportion (>65%) than those of monoterpenoids (<5%). The chemical compositions of the LAVEOs were completely different from those in previous reports. For example, germacrene D (23.6%) was the major component of AVEOs steam from Mauritius [[96]30], which was not detected in our EOs. The major compounds α-thujone (47.0%), β-thujone (10.0%), eucalyptol (21.0%) and caryophyllene (3.4%) were identified from LAVEOs harvested in France [[97]31]. However, eucalyptol, α-thujone (47.0%) and β-thujone were not detected in our EOs. Interestingly, the content of caryophyllene (3.8%) in our study was consistent with the report in this study [[98]31]. The chemical composition of the GAVEOs was more abundant than that of the LAVEOs. Diterpenoids and aromatics were detected only in the whole grass. Interestingly, caryophyllene oxide (>17%) was the main substance in the EOs from different parts of A. verlotorum. Thirty common constituents were identified in the LAVEOs and GAVEOs, such as himbaccol, (+)-β-eudesmol, isospathulenol, etc. ([99]Table S2). 2.2. Analysis of the Main Active Ingredients and the Potential Effect of EOs from A. argyi and A. verlotorum Using the Ingredient-Target-Pathway Networking We analyzed the active ingredients, key targets and pathways of the different Artemisia EOs and created an “ingredient–target-pathway” map with a merge function. The networking was visualized using Cytoscape 3.9.1 software ([100]Figure S2). The ingredients were represented by the purple V, the targets were represented by the yellow ellipse and the pathways were represented by the blue rectangle. The active ingredients were represented by the node, and the edge links represented the targets and the active ingredients. The high number of linkages demonstrated the importance of networking the active component or aim. Among them, the nodes of the GAAEOs (YP-1) included 55 components, 315 targets and 287 pathways, where neointermedeol had the highest number of networking edges. The nodes of the LAAEOs (YP-2) included 54 components, 329 targets and 287 pathways, where trimethylenenorbornane had the highest number of networking edges. The nodes of the GAVEOs (YP-3) included 39 components, 278 targets and 275 pathways, where pentamethylcyclopentadiene had the highest number of networking edges. The nodes of the LAVEOs (YP-4) included 50 components, 284 targets and 280 pathways, where 9-(1-methylethylidene)-1,5-cycloundecadiene had the highest number of networking edges. The results showed that the ten components with the highest connectivity in these EOs differed. In contrast, the targets of those ten components were quite similar. Moreover, they corresponded to exactly the same pathways ([101]Figure 4, [102]Table S3). The results suggest that although the components are different, they have similar biological activities. Figure 4. [103]Figure 4 [104]Open in a new tab The top 10 components, targets and pathways of the number of networking edges (in [105]Figure S2). (A) The number of networking edges from GAAEOs. (B) The number of networking edges from LAAEOs. (C) The number of networking edges from GAVEOs. (D) The number of networking edges from LAAEOs. 2.2.1. Comparison and Analysis of Main Active Ingredients of AAEOs and AVEOs In order to focus on the possible active components, we investigated the drug-like characteristics of these molecules. The discovered compounds in the AAEOs and AVEOs were evaluated for their drug-like characteristics using Lipinski’s five guidelines. R was applied to analyze the correlation between the nodes. The nodes with a strong correlation were displayed with the same/similar color ([106]Figure S3). The active ingredients contained in the top 50 scores are shown in [107]Table 2, and the results showed that the ingredients with the highest scores among the four EOs were not the same. Table 2. Ingredients included in the top 50 of the score by R cluster analysis. GAAEOs NO ^a Scores ^b LAAEOs NO ^a Scores ^b GAVEOs NO ^a Scores ^b LAVEOs NO ^a Scores ^b methyleugenol C37 8462.68 alloaromadendrene oxide C44 8986.16 m-anisalcohol C34 11053.29 9-(1-methylethylidene)-1,5-cycloundecadiene C14 6343.27 1-octen-3-ol C9 7899.98 methyleugenol C26 8884.00 alloaromadendrene oxide C32 8097.39 4-methylene-5-hexenal C13 5064.08 o-cymene C40 7509.06 1-octen-3-ol C6 8458.30 pentamethylcyclopentadiene C35 7295.15 β-elemene C46 5060.33 chamazulene C22 6276.07 chamazulene C15 6509.30 β-elemene C29 6874.15 β-calacorene C44 4812.89 eugenol C32 5706.63 p-cymene C49 6502.40 β-calacorene C21 5972.17 (+)-α-calacorene C5 4325.45 bornyl isovalerate C16 4615.64 trimethylenenorbornane C50 5701.88 neointermedeol C17 5612.14 bornyl acetate C16 4218.15 cis-carveol C23 4358.01 eugenol C22 5568.67 chlorpyrifos C33 5080.56 caryophyllene oxide C20 4076.81 junenol C35 4222.82 β-elemene C53 5117.62 (+)-α-calacorene C1 4997.34 2,5-dimethyl-3-methylene-1,5-heptadiene C11 4066.86 neointermedeol C38 3900.09 junenol C25 4547.44 2-borneol C4 4197.71 (−)-verbenone C9 3884.48 γ-pironene C53 3777.84 trans-4-thujanol C31 4372.10 caryophyllene oxide C9 4072.41 neointermedeol C36 3875.57 eucalyptol C31 3512.37 cis-carveol C16 4218.90 mustakone C16 3428.25 ledane C32 3635.38 copaene C28 3422.12 2-ethylidene-6-methyl-3,5-heptadienal C43 4017.18 β-selinene C23 3130.88 isoaromadendrene epoxide C28 3631.91 camphor C8 3298.98 caryophyllene oxide C14 3864.40 (−)-xanthorrhizol C27 2908.70 2-borneol C12 3330.03 methyl isocostate C36 3249.50 α-himachalene C51 3523.03 β-costol C22 2801.19 (+)-α-cyperone C6 3159.83 γ-costol C51 3240.85 γ-pironene C37 3487.45 himbaccol C12 2651.97 (−)-calamenene C1 2998.45 caryophyllene oxide C21 3158.27 neointermedeol C27 3457.32 ledol C15 2651.97 δ-cadinene C49 2985.95 4-thujanol C12 2986.18 eucalyptol C21 3336.46 palustrol C18 2397.18 cis-carveol C22 2899.12 trans-4-thujanol C44 2986.18 bornyl acetate C45 3168.03 4(15),5,10(14)-germacratrien-1-ol C11 2333.46 linalool C34 2869.81 (−)-β-bourbonene C4 2968.32 γ-terpinene C38 3009.81 isoledene C13 2331.30 mustakone C35 2747.32 2-borneol C10 2951.02 chrysanthenone C46 2935.24 α-gurjunene C19 2318.41 α-himachalene C42 2652.59 camphene C17 2884.04 1-(1-butynyl) cyclopentanol C41 2756.51 (±)-α-curcumene C3 2280.32 neoisolongifolene C23 2606.73 cis-p-menth-2-en-1-ol C27 2594.42 copaene C19 2752.69 τ-muurolol C25 1888.10 β-selinene C47 2491.54 (±)-piperitone C6 2490.96 cis-p-menth-2-en-1-ol C18 2519.15 α-neocallitropsene C20 1815.24 α-copaene C40 2418.91 (−)-carvone C2 2489.47 (−)-β-bourbonene C3 2434.89 α-muurolene C38 1762.92 himbaccol C27 2395.18 selina-4,11-dien C42 2458.10 yogomi alcohol C32 2385.97 ledol C33 2395.18 (+)-δ-cadinene C11 2187.37 selina-4,11-dien C29 2376.64 camphor C31 2272.53 [108]Open in a new tab ^a Numbering of compounds in the R cluster analysis diagram (Figure S3). ^b Scores of components from R cluster analysis. Methyleugenol, the highest scoring compound in the GAAEOs, was a phenylpropanoid used as a flavoring agent, a fragrance and an anesthetic in rodents [[109]32,[110]33]. However, the content of methyleugenol was only 0.17% in the GAAEOs, and it is unclear whether it is related to the activity of the EOs. In addition, no clear bioactivity studies have been reported for alloaromadendrene oxide (the highest-scoring compound in LAAEOs), m-anisalcohol (the highest-scoring compound in GAVEOs) and 9-(1-methylethylidene)-1,5-cycloundecadiene (the highest-scoring compound in LAVEOs). The caryophyllene oxide and neointermedeol were the common components of these four Eos, as shown in [111]Table 2 using Venny analysis ([112]Figure 5). Caryophyllene oxide has broad activities that have attracted much attention. It exhibited cytotoxicity against multiple cancer cells, including MG-63 (human osteosarcoma cells), HepG2 (human leukemia cancer cells), AGS (human lung cancer cells) and SNU-1 (human gastric cancer cell) [[113]34,[114]35]. Another study showed that caryophyllene oxide is a potent CNS depressant [[115]36]. In addition, the CYP3A enzyme activity was markedly decreased by caryophyllene oxide, which could generally have an impact on the pharmacokinetics of the active compounds [[116]37]. However, few studies have reported the biological activity of neointermedeol. Figure 5. [117]Figure 5 [118]Open in a new tab (A) Venny diagram of the number and ratio of EOs ingredients in [119]Table 2. Venny analysis found that the EOs from different parts of two Artemisia share the same ingredients of caryophyllene oxide and neointermedeol. (B) Chemical structures of caryophyllene oxide and neointermedeol. 2.2.2. Comparison and Analysis of Key Proteins of AAEOs and AVEOs Compounds show bioactivities by binding with particular proteins. [[120]38]. [121]Table 3 shows the target proteins included in the top 50 scores after clustering analysis by R. Among them, NFKB1 has the highest score in all of the essential oils. NFKB1 is a member of the NF-κB family and an important regulator of NF-κB activity in vivo. It has been shown that NFKB1 is closely associated with inflammation, aging and cancer in the body [[122]39,[123]40,[124]41,[125]42]. Thus, NFKB1 is one of the key targets of Artemisia essential oils with similar biological activity. Table 3. Proteins are included in the top 50 of the score by R cluster analysis. GAAEOs NO. ^c Score ^d LAAEOs NO. ^c Score ^d GAVEOs NO. ^c Score ^d LAVEOs NO. ^c Score ^d NFKB1 T26 14,010.38 NFKB1 T26 14,260.94 NFKB1 T35 13,462.24 NFKB1 T40 13,974.02 PIK3CD T118 9959.54 MAPK1 T70 10,134.16 MAPK1 T171 10,188.79 MAPK1 T139 10,667.82 MAPK1 T70 9895.40 PIK3CD T105 9055.98 PIK3CD T161 6665.46 PIK3CD T110 6385.01 PRKCA T175 6076.31 PIK3CA T134 6661.89 PRKCA T120 6410.28 PIK3CA T260 4726.91 PIK3CA T144 5979.37 PRKCA T150 5812.37 PIK3CA T180 5693.44 PIK3CB T261 4431.85 PIK3R1 T156 5327.63 PIK3CB T263 5567.77 PIK3CB T181 5693.44 PIK3R1 T186 4424.04 RELA T215 5264.89 RELA T190 5363.08 RELA T46 4034.71 NOS2 T42 4065.17 PIK3CB T281 5001.83 NOS2 T47 4740.83 PIK3R1 T144 3792.55 PRKCA T197 3704.80 NOS2 T47 4484.77 PIK3R1 T201 3986.61 NOS2 T36 3742.43 RELA T120 3315.37 PRKCB T99 3924.41 PRKCB T245 3899.08 CASP9 T57 2926.28 PRKCB T164 3205.22 CASP9 T137 3217.58 CASP9 T127 3476.93 ITGB3 T219 2724.65 CASP9 T58 2908.34 NOS3 T27 2945.04 NOS3 T27 3296.31 RAC1 T266 2605.33 NOS3 T43 2799.68 STAT3 T125 2873.77 IDH1 T286 2816.61 ADAM10 T178 2482.65 ACACA T132 2548.51 SLC9A1 T122 2457.80 CYP1A2 T116 2592.61 NOS3 T37 2260.63 ITGB1 T135 2498.96 TP53 T230 2279.98 ACACA T63 2570.09 PLA2G1B T94 2167.69 FBP1 T283 2494.99 CYP1A2 T86 2186.07 PLA2G1B T144 2496.26 MAOA T32 2087.59 PRKCG T166 2431.40 GUSB T245 2182.93 ENPP1 T271 2386.76 MAPK14 T6 1975.62 STAT3 T86 2402.65 MTOR T201 2178.04 TP53 T197 2281.28 CYP1A2 T68 1920.10 UGT2B7 T255 2334.90 ENPP1 T290 2166.02 MAOA T66 2247.70 MTOR T158 1832.80 CYP1A2 T90 2251.76 STAT3 T126 1802.37 TP53 T239 2194.28 UGT2B7 T172 1774.58 PLA2G1B T76 2169.10 SLC9A1 T51 1773.40 [126]Open in a new tab ^c Numbering of targets in the R cluster analysis diagram ([127]Figure S3). ^d Scores of targets from R cluster analysis. A protein association network was constructed using STRING databases to screen the key target proteins with high interactions ([128]Figure 6). The nodes encoded the networking of the target proteins. Furthermore, the protein–protein connection was defined as the connectivity degree. Genes with a high connectivity degree were defined as hub genes. The study found that RELA (transcription factor p65), PIK3R1 (phosphatidylinositol 3-kinase regulatory subunit α) and MAPK1 (mitogen-activated protein kinase 1) were the key targets and played a crucial role in various functions of the EOs from Artemisia in therapy. RELA (RELA Proto-Oncogene, NF-κB Subunit) is a pleiotropic transcription factor in practically all cell types. It is the endpoint of a series of signal transduction events that are sparked by various stimuli related to numerous biological processes, including tumorigenesis, differentiation, cell growth and apoptosis [[129]43,[130]44,[131]45,[132]46]. PIK3R1 functions as an adapter, mediating the association of the p110 catalytic unit to the plasma membrane. It binds to activated (phosphorylated) protein-Tyr kinases through its SH2 domain. This is required for the insulin-stimulated increase in glucose uptake and glycogen synthesis in insulin-sensitive tissues [[133]47,[134]48]. These targets are of great significance and deserve further investigation based on the relevant biological activity research before AAEOs, including their anti-inflammatory [[135]9,[136]49], neuroprotection [[137]50], hypoglycemic and anti-oxidant [[138]51], as well as anti-cancer [[139]52,[140]53] and other beneficial pharmacological, effects [[141]18,[142]54]. Figure 6. [143]Figure 6 [144]Open in a new tab Protein-protein interaction (PPI) networking. Protein-protein interactions (set confidence level p > 0.9) targets from R cluster analysis. (A) Protein interaction network diagram of GAAEOs. (B) Protein interaction network diagram of LAAEOs. (C) Protein interaction network diagram of GAVEOs. (D) Protein interaction network diagram of LAVEOs. Additionally, the study found that the two MAPKs that are crucial to the MAPK/ERK cascade are MAPK1/ERK2 and MAPK3/ERK1. The regulation of transcription, translation and cytoskeletal rearrangements by the MAPK/ERK cascade regulates a variety of biological tasks, including cell growth, adhesion, survival and differentiation, depending on the cellular context [[145]55,[146]56]. Our results showed that although the ingredients were different from AVEOs and AAEOs, the targets of those ingredients were the same proteins. The results suggested that AVEOs and AAEOs have similar biological activities. 2.2.3. GO Enrichment Analysis and KEGG Pathway Annotation The biological processes, cellular components and molecular functions were the three major functional categories identified from the GO term enrichment analysis of the EOs. The top ten GO terms of each category are illustrated in [147]Figure S4. The results showed that the cellular components had the most significantly enriched terms, with protein binding being the most significant cellular function among the four EOs. Protein binding can improve or impair a drug’s effectiveness [[148]57]. To identify the pathways and targets that were involved directly, the pathway information related to the targets was obtained through KEGG analysis. [149]Figure 7 shows that the associated pathways mainly included metabolic pathways, the neuroactive ligand–receptor interaction and pathways in cancer. Substantial experimental evidence showed that AAEOs had good activities in neuroprotection [[150]50], anti-inflammatory, analgesic and anti-tumor effects [[151]18,[152]54], which were consistent with the prediction of networking pharmacology. Further analysis based on GO and KEGG showed that the enrichment results of the AAEOs and AVEOs were highly consistent, indicating that AVEOs and AAEOs might have the same action pathway and similar pharmacological activities. Figure 7. [153]Figure 7 [154]Open in a new tab Bubble diagram of KEGG enrichment analysis of EOs from Artemisia. (A) KEGG pathway of GAAEOs. (B) KEGG pathway of LAAEOs. (C) KEGG pathway of GAVEOs. (D) KEGG pathway of LAVEOs. Y-axis label represents the pathway, and X-axis label represents the gene ratio. Each bubble represents an enriched function, and the size of the bubble represents the number of genes enriched in the pathway. The bubble is colored according to its −log (p value). 2.2.4. Network Analysis of the Unique Components of the Four Artemisia Essential Oils Furthermore, we constructed a network between the unique composition of Artemisia essential oils (GAAEOs, LAAEOs, GAVEOs, LAVEOs) and its targets and pathways ([155]Figure 8). The key targets and pathways of the top five (red nodes) revealed by the enrichment results were highly consistent with the results of the GO and KEGG analyses ([156]Figure 6 and [157]Figure 7), further illustrating that the two different Artemisia essential oils have similar pharmacological activities despite significantly differing in their compositions. In addition, the key targets have been proved to mediate the relevant signaling pathways to exert immunomodulatory, anti-inflammatory and neuroprotective activities, etc. [[158]18,[159]58,[160]59,[161]60]. Figure 8. [162]Figure 8 [163]Open in a new tab Network diagram between the unique components of Artemisia EOs (GAAEOs, LAAEOs, GAVEOs, LAVEOs) and targets, and pathways. Red nodes represent the targets and pathways with the highest degree of association. 2.3. In Vitro and In Vivo Toxicity of Artemisia Essential Oils As a processing of Chinese materia medica widely used in clinic, the toxicological evaluation of the Artemisia essential oils was particularly important. In order to better ensure the drug safety, it can be assumed that the biological activity of Artemisia essential oils was assessed through in vitro and in vivo toxicity tests. The results of the cellular level assay showed that 100 μg/mL of Artemisia essential oils did not show cytotoxic activity against both HEK-293T cell lines treated for 24 h ([164]Figure 9A,B). Furthermore, the in vivo assay in zebrafish showed that 10 μg/mL of Artemisia essential oils did not affect the growth and survival of zebrafish ([165]Figure 9C). This study presents the first systematic assessment of the toxicity of Artemisia essential oils using an in vitro human normal cell line (HEK-293T cells) assay in combination with an in vivo zebrafish assay. In this regard, Artemisia essential oils are highly safe. Figure 9. [166]Figure 9 [167]Open in a new tab Toxicity test results for Artemisia essential oils. (A) Effect of Artemisia essential oils on the viability of HepG2 cells. (*** p < 0.001 positive group vs. control group). (B) Effect of Artemisia essential oils on the viability of A549 cells. (C) Effects of Artemisia essential oils (10 µg/mL) on the development and growth of zebrafish. All data were presented as means ± SD (n = 5), calculated by student t test. 3. Conclusions Aromatic chemical components from folkloric medicinal plants, such as EOs, have been claimed to be useful in treating or preventing a variety of illnesses [[168]61]. The fragrant medicinal plants of the Artemisia species have complicated locations and origins [[169]1]. Although the compounds and activity of A. argyi have been investigated, the mechanism by which these components act on human health at the cell level has remained largely unknown. Few studies have focused on the biological activity of the essential oils of A. verlotorum. Herein, we analyzed the compositional differences and potential bioactivities of EOs from two Chinese mugworts. The results showed that the chemical composition of the EOs from A. argyi and A. verlotorum were quite different. However, networking pharmacology and R cluster analysis studies showed that they share similar key target proteins, as well as highly consistent protein interactions and signaling pathways. This indicated that these two Artemisia essential oils could be substituted for each other in aromatic therapy. Our study provides evidence to better understand the development and application of A. argyi and A. verlotorum in Chinese traditional medicine and lays a foundation for the clinical safety and scientific medication of Artemisia essential oils. 4. Materials and Methods 4.1. Plant Materials and Reagent The A. argyi H. Lév. and Vaniot and A. verlotorum Lamotte were derived from different species of the same genus and have extremely similar morphological characteristics. These samples were (n = 3) collected from Nanyang, Henan Province, and Luofo mountain, Guangdong Province (specific medicinal plant cultivation sites) [[170]62,[171]63], People’s Republic of China, respectively, and were collected strictly according to the medicinal age (1 year) and time of harvest. Both plants were identified by Prof. Chong-Ren Yang. The materials were dried in the shade at 25 °C until the humidity was lower than 5%, and stored in the refrigerator at 4 °C. Voucher specimens (more than 100 mg) were deposited at the Center for Pharmaceutical Sciences, Faculty of Life Science and Technology, Kunming University of Science and Technology, Kunming, P. R. China. 4.2. Extraction of EOs A. verlotorum Steam distillation was used to extract the EOs from different parts of the whole grass and leaves of A. argyi (YP-1 and YP-2) and A. verlotorum (YP-3 and YP-4). The dried Artemisia (100 g), in a round bottom flask, was added to distilled water (1 L). The extractions of Eos, in detail, followed previous studies [[172]64]. Detailed information of all the EOs is presented in [173]Table 1. 4.3. GC-MS Analysis GC-MS was used to analyze the compounds of AAEOs and AVEOs. The methods, in detail, followed previous studies [[174]64]. 4.4. Chemical Ingredients Database Building of EOs The discovered chemicals in the EOs were screened for drug-like characteristics using Lipinski’s five rules [[175]65]. The Traditional Chinese Medicine Systems Pharmacology database (TCMSP) ([176]http://tcmspw.com/tcmsp.php, 18 February 2022) [[177]66] and PubChem database ([178]https://pubchem.ncbi.nlm.nih.gov/, 18 February 2022) were used to assess these crucial pharmaceutical features. 4.5. Collection of Target Proteins and Pathways of the EOs Potential targets of the compounds in the AAEOs and AVEOs were collected via networking databases (probability > 0.7), including SuperPred ([179]https://prediction.charite.de, 20 February 2022), SwissTargetPrediction ([180]http://www.swisstargetprediction.ch/, 20 February 2022) and TargetNet ([181]http://targetnet.scbdd.com/, 20 February 2022). The ingredients’ absent target proteins were eliminated. In addition, the information on all the collected proteins was made uniform using Uniprot ([182]http://www.uniprot.org/, 3 March 2022) [[183]67]. Using the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database ([184]http://www.kegg.jp/kegg/, 4 March 2022) and the Database for Annotation, Visualization and Integrated Discovery (DAVID) ([185]https://david.ncifcrf.gov/summary.jsp, 4 March 2022), the analysis of pathways was carried out on the chosen targets. The database is an encyclopedia of genes and genomes and includes information on signal transduction, cellular biology and homologous conservative route [[186]68]. The study uses Human sapiens as model species. 4.6. Networking Construction Cytoscape 3.9.1 was used to analyze and visualize the data gathered to create complicated networks [[187]69]. In this network, nodes stood in for components, targets or pathways, while edges denoted their connections. Subsequently, the tight lines and complexities of the connections between important chemical components, targets and pathways were considered to explore the underlying mechanism of action. Therefore, cluster analysis of the relevant collected information was performed by R (cluster_louvain). The related ingredients, targets and pathways information resulted in a data set that was converted to an igraph graph using the “igraph” software package. A function of graph_from_incidence_matrix creates a bipartite igraph graph from the incidence matrix of the data for targets and pathways. Based on the bipartite igraph graph, a function of igraph_cluster_louvain implements the multi-level modularity optimization algorithm for finding community structure, and different communities were marked in different colors by a R_rainbow, while their relationships were analyzed using the igraph_deg function [[188]70]. Based on the top targets in the R processing results, protein–protein interaction networking analysis (PPI) was also carried out to assess the targets. The association between the targets was evaluated as follows. PPI: To show how the target proteins interact, the target proteins were uploaded to the STRING databases platform ([189]http://string-db.org, 6 April 2022). In this study, we eliminated the isolated targets and constructed a PPI networking by screening them with a confidence score > 0.90. 4.7. Gene Ontology and Pathway Enrichment Analysis Gene ontology (GO) enrichment analysis was performed on the candidate targets using the online tool DAVID and KEGG, as well as KEGG pathway annotation. p values < 0.05 were considered statistically significant. 4.8. Toxicity Analysis Cytotoxicity assay: The toxicity of Artemisia essential oils on HEK-293T cells (human embryonic kidney 293 cells, were obtained from the Kunming Cell Bank of Chinese Academic of Sciences) was detected using the MTT method, consistent with the previous study [[190]71]. Cells were treated with Artemisia essential oils (25, 50, 100 μg/mL) for 24 h. The cells incubated with 3 µM paclitaxel (PTX) (Adamas Reagent Co., Ltd., 48803A, Shanghai, China) was used as the positive control (paclitaxel was drugs widely used clinically for the treatment of cancer). Zebrafish toxicity assay: Healthy 24 h post-fertilization (hpf) embryos (transparent fertilized eggs) were randomly transferred to different concentrations of methanol extract of Artemisia essential oils (10 μg/mL) in the sterile 12 well plate. Each group was provided with 20 embryos (repeated experiment, n = 5). The development and morphological changes of the embryos after drug exposure were observed, photographed and recorded using an inverted optical microscope at 18 and 36 hpf, and recorded the hatching rate of the embryos at 36 hpf. 4.9. Statistical Analysis All the data were statistically analyzed using the GraphPad Prism 9 software, using a two-tailed Student’s t-test. p values < 0.05 were considered statistically significant. Acknowledgments