Based on the GC–MS results of Cymbopogon spp., 8 compounds with the highest percent area > 3.26% were used as ligands. Information on the CID and formula of each compound was obtained from PubChem (Table 1), for drug-likeness prediction analysis before conducting in silico analysis of the antioxidant and antimicrobial activity assays.

Drug likeness assesses the physicochemical properties of query compounds containing drug molecules. Several parameters, including molecular mass, high lipophilicity, hydrogen bond donors, hydrogen bond acceptors, and molar refractivity, were used for drug-like molecule determination in this study. Query compounds with positive predictions as drug-like molecules must fulfill at least two Lipinski rules. In the present study, all compounds from Cymbopogon spp. were drug-like molecules (Table 2), as predicted based on Lupinski’s Rule of Five.

Virtual analysis or molecular docking is used to identify bonding interaction patterns and ligand activity on a target. Ligand activity is indicated by the binding affinity value of the ligand–protein complex [28]. Moreover, antioxidant and antimicrobial activities are related to pathway mechanisms regulated by proteins, such as the filamenting mutant Z (FtsZ) protein, the MciZ synthetic peptide [45], aquaporin Z [46], SA [47], acetohydroxy acid or AHAS [48], SOD1, and catalase [49]. The structures of FtsZ, aquaporin Z, AHAS, SA, SOD1, and catalase are displayed as transparent surfaces, cartoons, and single colors in Fig. 2.

Three-dimensional structure visualization of target proteins. A Bacillus subtilis-FtsZ, B Escherichia coli-Aquaporin Z, C Candida albicans-AHAS, D Staphylococcus aureus-SA, E SOD1, and F Catalase

Binding affinity is the negative binding energy formed when interactions between molecules refer to the thermodynamic rule. When the value of binding affinity is more negative, the ligand activity increases; this makes it possible to trigger biological responses such as inhibition [50]. In the present study, molecular docking was used to screen for the antibacterial activity of Cymbopogon spp. through the binding of FtsZ, Aquaporin Z, AHAS, SA, SOD, and catalase. The results of docking simulation showed that alpha-cadinol has the most negative binding affinity to FtsZ, aquaporin Z, AHAS, SA, SOD1, and catalase (Table 3). This metabolite is predicted to act as an antibacterial agent by inhibiting three targets, namely FtsZ, aquaporin Z, and SA from B. subtilis, E. coli, and S. aureus. Moreover, it acts as an antifungal agent by inhibiting AHAS in C. albicans. The inhibition of SOD and catalase by alpha-cadinol showed better antioxidant activity than other compounds from Cymbopogon spp. Figure 3 shows the three-dimensional structures of the following ligand–protein complexes: alpha cadinol-FtsZ, alpha cadinol-aquaporin Z, alpha cadinol-AHAS, alpha cadinol-SA, alpha cadinol-SOD1, and alpha cadinol-catalase.

Ligand–protein visualization from molecular docking simulation. A Alpha Cadinol-FtsZ, B Alpha Cadinol-Aquaporin Z, C Alpha Cadinol-AHAS, D Alpha Cadinol-SA, E Alpha Cadinol-SOD1, and F Alpha Cadinol-catalase

Weak bond interactions trigger target-specific biological activities such as inhibition. For example, van der Waals, hydrogen, hydrophobic, pi/alkyl, and electrostatic bonds can play a role in the inhibitory activity of the target [34]. In this study, alpha-cadinol interacted with FtsZ via van der Waals interactions at Gly107, Glu139, Gly106, Arg143, Met105, Pro135, Gly104, Thr133, Asn166, Gly22, Asp187, and Leu190 and via alkyl/pi interactions at Phe183 and Ala186. Moreover, alpha-cadinol interacted with Aquaporin Z via van der Waals interactions at Asn182, Val24, Ser118, Phe116, Gly115, Ser114, and Gly28; hydrogen interactions at Ala117; and alkyl/pi interactions at Ala117, Ala23, Val39, Ala27, Ile178, and Phe36. This metabolite further interacted with AHAS via van der Waals interactions at Arg340, Gln481, Thr507, Lys485, Glu486, Thr511, Gln508, Thr505, and Ser477 and via alkyl/pi interactions at Ala480, Phe504, Val489, Trp506, and Val487. Additionally, alpha-cadinol interacted with SA via van der Waals interactions at Ser58, Thr122, Glu113, Lys117, Leu111, Gln120, Val108, and Arg139, and via alkyl/pi interactions at Ala46, Ile124, and Ile141. Alpha-cadinol interacted with SOD1 via van der Waals interactions at Glu78, His71, Gly72, Gly127, Thr135, Ile99, and Glu100 and via alkyl/pi interactions at Lys128, Lys75, and Pro74. Finally, alpha-cadinol interacted with catalase via van der Waals interactions at Glu330, Asn171, Tyr325, Asn324, Asn397, Asp389, Asn403, His166, Lys169, and Arg170, and via alkyl/pi/sigma interactions at Phe326, His175, and Pro172 (Fig. 4).

Two-dimensional visualization of ligand–protein interactions. A Alpha Cadinol-FtsZ, B Alpha Cadinol-Aquaporin Z, C Alpha Cadinol-AHAS, D Alpha Cadinol-SA, E Alpha Cadinol-SOD1, and F Alpha Cadinol-catalase

The RMSF values at the ligand–protein complex hotspots in this study consisted of van der Waals forces (1.485, 2.188, 0.797, 0.627, 0.432, 0.578, 0.489, 0.133, 0.372, 0.392, 0.194, and 0.128) and alkyl/pi interactions (0.183 and 0.186) in the FtsZ domain (MD plot link: https://biocomp.chem.uw.edu.pl/CABSflex2/job/4cbe0f320ffa1fa/), accessed April 2024. RMSF on the Aquaporin Z domain (MD plot link: https://biocomp.chem.uw.edu.pl/CABSflex2/job/c7edfbf8a9bc31/), accessed April 2024, occurred via van der Waals forces (0.725, 0.707, 1.478, 1.020, 1.276, 1.110, and 0.328), hydrogen bonds (1.196), and alkyl/pi interactions (1.196, 0.523, 0.142, 1.442, 0.514, and 0.365). RMSF in the AHAS domain (MD plot link: https://biocomp.chem.uw.edu.pl/CABSflex2/job/9d455305023de11/), accessed April 2024, occurred via van der Waals (1.471, 0.641, 1.326, 1.021, 0.715, 0.678, 1.971, 0.907, and 0.400) and alkyl/pi (0.480, 0.687, 0.489, 1.025, and 0.4871) interactions. The RMSF in the SA domain (MD plot link: https://biocomp.chem.uw.edu.pl/CABSflex2/job/ea22bce9fb5421d/), accessed April 2024, occurred via van der Waals (0.310, 0.153, 0.709, 1.161, 1.221, 0.175, 2.011, and 0.351) and alkyl/pi (0.356, 0.193, and 0.337) interactions. RMSF in the SOD1 domain (MD plot link: https://biocomp.chem.uw.edu.pl/ CABSflex2/job/5eadcf318a44b3/), accessed April 2024, occurred via van der Waals (2.863, 2.596, 1.440, 0.239, 2.439, 0.509, and 0.485) and alkyl/pi (2.997, 0.540, and 2.156) interactions. Finally, RMSF in the catalase domain (MD plot link: https://biocomp.chem.uw.edu.pl/CABSflex2/job/d89e8a25c3981f4/), accessed April 2024, occurred through van der Waals (0.829, 0.914, 1.137, 1.277, 1.452, 1.686, 2.990, 0.686, 0.703, and 0.768) and alkyl/pi (1.161, 1.310, and 1.639) interactions (Fig. 5).

Root Mean Square Fluctuation (RMSF) dynamic plot and conformational structures of target proteins. A Alpha Cadinol-FtsZ, B Alpha Cadinol-Aquaporin Z, C Alpha Cadinol-AHAS, D Alpha Cadinol-SA, E Alpha Cadinol-SOD1, and F Alpha Cadinol-catalase

The stability of molecular interactions at the hotspots was identified using molecular dynamics simulations. The stability of bonding interactions at hotspots is indicated by RMSF values < 3 (Å). RMSF refers to the deviation in the interaction distance formed by the ligand in the target domain [30]. The alpha cadinol interactions of Cymbopogon spp. formed stable interactions on FtsZ, Aquaporin Z, SA, AHAS, SOD1, and catalase domains with RMSF values < 3 (Å). This indicates that the compound has antibacterial, antifungal, and antioxidant properties.

Target prediction showed that alpha cadinol from C. citratus has other targets for antioxidant activity, namely cyclooxygenase-2 (PTGS2), cyclooxygenase-1 (PTGS1), nitric-oxide synthase (NOS1), monoamine oxidase B (MAOA), aldehyde dehydrogenase (ALDH2), and carboxylesterase 1 (CES1). Alpha-cadinol exhibited inhibitory activity against PTGS1, PTGS2, NOS1, MAOA, and ALDH2. CES1 can also be activated by alpha-cadinol as an antioxidant pathway, additionally annotating the pathway with FtsZ, Aquaporin Z, SA, and AHAS inhibitory targets for antibacterial and antifungal activity (Fig. 6A). PTGS1, PTGS2, NOS1, MAOA, and ALDH2 increase ROS production [51,52,53,54,55]. In contrast, CES1 reduces ROS production under oxidative stress [56]. The target proteins of alpha-cadinol include PTGS1, PTGS2, NOS1, MAOA, and ALDH2. In the biological pathways of Homo sapiens, nodes with hexagonal, pentagonal, and elliptical shapes were the targets obtained during docking analysis. Additional pathway prediction targets from the database are shown as nodes with rounded rectangular shapes (Fig. 6B). The pathway consisting of antibacterial, antifungal, and antioxidant drugs has a confidence value of 0.6 (medium confidence) or 70–80%, which is highly accurate [54,55,56].

Prediction of antioxidant and antimicrobial activity pathways of alpha cadinol. A Target position of alpha cadinol from the pathway database. B Antibacterial, antifungal, and antioxidant pathway of alpha cadinol compounds. The T shape on the target indicates inhibitory activity and arrows for activation. The yellow color of the nodes indicates the target position. FtsZ: Filamenting temperature-sensitive mutant Z; SA: Sortase A; AHAS: Acetohydroxyacid synthase; SOD: Superoxide dismutase 1; PTGS2: Cyclooxygenase-2; PTGES: Prostaglandin E synthase; PTGIS: Prostaglandin I2 synthase; TBXAS1: Thromboxane A synthase 1; PTGES3: Prostaglandin E synthase 3; PTGDS: Prostaglandin D2 synthase; HPGDS: Hematopoietic prostaglandin D synthase; CYP2C19: Cytochrome P450 family 2 subfamily C member 19; CYP2E1: Cytochrome P450 family 2 subfamily E member 1; ALOX15B: Arachidonate 15-lipoxygenase type B; CYP2C8: Cytochrome P450 family 2 subfamily C member 8; CYP2J2: Cytochrome P450 family 2 subfamily J member 2; MPO: Myeloperoxidase; CYP2C9: Cytochrome P450 family 2 subfamily C member 9; ALOX12: Arachidonate 12-lipoxygenase; CYP4F3: Cytochrome P450 family 4 subfamily F member 3; CYP4F2: Cytochrome P450 family 4 subfamily F member 2; PTGS1: Cyclooxygenase-1; NOS1: Nitric-oxide synthase; ADH4: Alcohol dehydrogenase 4; ADH1B: Alcohol dehydrogenase 1B; ADH5: Alcohol dehydrogenase 5; CAT: Catalase; MAOA: Monoamine oxidase B; AOC3: Amine oxidase copper containing 3; AOC2: Amine oxidase copper containing 2; COMT: Catechol-O-methyltransferase; CYP3A4: Cytochrome P450 family 3 subfamily A member 4; ALDH2: Aldehyde dehydrogenase; CES1: Carboxylesterase 1

DPPH and ABTS assays were performed to assess the antioxidant activities of the extracts. The IC50 values of the crude extracts prepared using each solvent are listed in Table 4. The leaves of Cymbopogon spp. had strong antioxidant activity, which was determined based on the Prieto criteria [28]. IC50 values of ABTS were 40.90 ± 0.94, 44.86 ± 1.23, 47.04 ± 1.03, 75.93 ± 1.48, 77.44 ± 1.20, 80.53 ± 1.38, 85.46 ± 21.17, 103.45 ± 6.69, and 128.93 ± 18.49 µg/mL for C. nardus leaves, C. citratus leaves, C. winterianus leaves, C. citratus stems, C. winterianus stems, C. winterianus roots, C. nardus roots, C. nardus stems, and C. citratus roots, respectively. Contrastingly, the IC50 values of DPPH were 61.30 ± 1.04, 86.36 ± 1.09, 92.87 ± 1.54, 101 ± 13.01, 131.54 ± 10.74, 152.46 ± 10.79, 162.97 ± 18.83, 309.74 ± 6.48, and 346.19 ± 13.36 µg/mL for C. winterianus leaves, C. citratus leaves, C. nardus leaves, C. winterianus stems, C. nardus stems, C. citratus stems, C. winterianus roots, C. nardus roots, and C. citratus roots, respectively. The IC50 of ABTS was lower than that of DPPH. Overall, the ABTS and DPPH IC50 values of the leaf extract (strong) were the lowest, followed by those of the stem (moderate) and root (moderate-weak) extracts [28]. The potent antioxidant activity of the ethanolic extract of Cymbopogon spp. was likely due to the presence of bioactive compounds (Table 1, Additional File 2). Compared to other studies that have previously reported high antioxidant compounds, the antioxidant activity of ethanolic extracts from Cymbopogon spp. was higher than that of Sonchus arvensis L. leave [30] and Pterocarpus macrocarpus Kurz. bark extracts [29].

The IC50 values for C. citratus. leaf ethanolic extract in the present study were lower against ABTS than those previously reported for C. citratus leaf methanolic extract and fractions using n-hexane, chloroform, and ethyl acetate [57]. This indicates the higher antioxidant activity of C. citratus extracts in the present study. Compared to previously reported C. nardus essential oil [58], C. nardus leaf ethanolic extracts showed higher IC50 values against DPPH in the present study. Moreover, the IC50 values for all C. winterianus leaf ethanolic extracts against ABTS and DPPH were lower than those of the essential oil extracted from C. winterianus leaves [59]. In contrast, the IC50 values of Cymbopogon spp. root and stem ethanolic extracts were higher than those previously reported for Cymbopogon spp. leaf extracts [57,58,59].