Background

Infections are a serious global health problem [1]. In Indonesia, 28.1% of the primary causes of death are infections. Infection is caused by various microorganisms, such as bacteria, viruses, fungi, and protozoa [2]. The World Health Organization (WHO) has issued a list of priority pathogens to enhance research efforts in the search for new antibiotics that overcome drug resistance [3]. This necessitates the identification of novel sources of antimicrobial drugs, such as plant-derived natural products [4, 5].

Approximately 80% of the world population uses natural compounds derived from medicinal plants [1]. However, many potentially medicinal plants in Indonesia have not been studied. Of the 9,600 plant species with medicinal properties, only 700–1000 are used for medicinal purposes. One of these species is serai (Cymbopogon spp.), which has numerous uses in Indonesia, including as a spice, drink, and source of essential oils. Cymbopogon spp., which belongs to the family Poaceae, consists of various species, including seasoned lemongrass (Cymbopogon citratus (L.) Rendle) and fragrant lemongrass (citronella; Cymbopogon nardus (DC.) Spatf and Cymbopogon winterianus Jowitt) [6, 7].

The bioactivity of Cymbopogon spp. has been explored extensively. The antimicrobial activity of C. flexuoxus [8], C. citratus [9,10,11,12], C. nardus [13, 14], and C. schoenanthus [15], has been studied. Moreover, the pesticidal activity of C. citratus [16], and C. winterianus [17], as well as the antioxidant activity of C. citratus [18, 19], have been investigated. The essential oil of Cymbopogon spp. plays an important role in some lemongrass bioactivities [8]. This essential oil has shown excellent biocompatibility and few side effects in human studies [20]. Furthermore, the molecular mechanisms of the bioactive compounds of Cymbopogon spp. against some microbes have been evaluated using proteomic analyses [8].

Wahyuni et al., [21] elaborated that metabolite profiles and bioactivities differed with each plant organ. Similarly, antimicrobial and antioxidant activity assays of Cymbopogon spp. organs and computational prediction of the mechanism pathway have been conducted in this study, as there are few reports about antimicrobial and antioxidant activity assays of organ/plant part of Cymbopogon spp. and the mechanism pathway prediction computationally. Moreover, in this study, ethanol was used as the solvent because its structure enables the dissolution of polar compounds such as water, non-polar compounds, hydrophilic compounds such as hexane, and hydrophobic compounds. Ethanol is also used as a medicinal solvent owing to its low toxicity and nonpolar capabilities. Similarly, ethanol extract is more soluble [22].

This study aimed to explore and compare the antioxidant and antimicrobial activities of roots, stems, and leaves derived from three commercial Indonesian Cymbopogon spp., as well as the mechanisms of inhibition. Our findings could enable the pharmaceutical applications of these three Cymbopogon spp. in Indonesia.

Methods

Plant material collection and identification

The three commercials Indonesian Cymbopogon spp. used in this study were Cymbopogon citratus (DC.) (Stapf), Cymbopogon nardus (L.) Rendl, and Cymbopogon winterianus Jowitt. The plants were obtained from the Medicinal Plant Garden (Taman Husada Graha Famili) in Surabaya, East Java, Indonesia (7º18′12.2’’S 112º41′12.7’’7E). Samples were collected and authenticated by the authors in the Plant Systematics Laboratory, Department of Biology, Faculty of Science and Technology, Universitas Airlangga. A voucher specimen was deposited to the Plant Systematics Laboratory, Department of Biology, Faculty of Science and Technology, Universitas Airlangga (No. CC.0110292022; CN.0110292022; and CW.0110292022).

Extraction

The leaves, stems, and roots of Cymbopogon spp. were air-dried and ground into a powder (20 mesh size). Next, 10 g of the ground samples were macerated sequentially in absolute ethanol (pro analysis; Merck, Darmstadt, Germany) at a ratio of 1:10. Each maceration was done thrice for 24 h at room temperature (28 ± 2 °C). The resulting extracts were filtered through filter paper, evaporated on a rotary evaporator at 60ºC to acquire a dry residue, weighed to calculate the yield of each extract, and stored at 4ºC.

Metabolite profiling by gas chromatography-mass spectrometry (GC–MS)

GC–MS was used to establish compound profiles from the ethanolic extracts of Cymbopogon spp. GC–MS analysis was performed using an Agilent GC-MSD (19091S-433UI; Agilent Technologies, Santa Clara, CA, United States) equipped with a capillary column (30 m × 250 µm × 0.25 µm), and a mass detector was operated in electron impact mode with full scan (50,550 amu). Helium was used as the carrier gas at a flow rate of 3 mL/min and a total flow rate of 14 mL/min. The injector was operated at 280 °C, whereas the oven temperature was programmed at an initial temperature of 60 °C and final temperature of 250 °C. Peaks in the chromatograms were identified based on their mass spectra. The compounds were identified through comparisons of their mass spectra with those in the Standard Reference Database (version 02. L; National Institute of Standards and Technology, Gaithersburg, MD, USA). Compounds with similarities > 80% were used in this study. The relative percentage of each component was calculated as the relative percentage of the total peak area in the chromatograph.

In silico analysis of antimicrobial and antioxidant pathways

Ligand retrieval

Compounds from Cymbopogon spp. were identified using GC–MS. The PubChem database (https://pubchem.ncbi.nlm.nih.gov/), accessed October 2023, was used to retrieve information such as cubic inch displacement (CID), formula, and structure data format (SDF) files. Energy minimization of ligands in sdf format was performed using OpenBabel software v2.3.1 for conversion into protein databank format (PDB) and for enhancing the flexibility of the atomic bonds that make up the ligand [23].

Protein preparation

The candidate antimicrobial compounds inhibit the activity of bacteria and fungi, including Bacillus subtilis, Escherichia coli, Staphylococcus aureus, and Candida albicans. For in silico analysis, the protein targets for molecular docking were: filamenting temperature-sensitive mutant Z (FtsZ, PDB ID: 2VAM), a protein encoded by ftsZ that assembles into a ring at the site of future bacterial cell division (also called the Z ring) [24]; Aquaporin Z (PDB ID: 1RC2), a water channel protein in higher and lower organisms [25]; sortase A (SA) (PDB ID: 2MLM), a bacterial transpeptidase of bacterial cell wall proteins [26]; and acetohydroxyacid synthase (AHAS, PDB ID: 6DEK), an enzyme target for antimicrobial drug discovery that is considerably common in microbial synthesis pathways [27]. The 3D structure of the target was obtained from the RCSB PDB database (https://www.rcsb.org/), accessed in October 2023, in pdb format. Target optimization for the docking simulation was performed using the PyMOL software (v.2.5.2; Schrödinger, Inc., Washinton, USA) with an academic license for the removal of water molecules and native ligands. In this study, the targets superoxide dismutase 1 (SOD1) (PDB ID: 5YTU) and catalase (PDB ID: 1DGH) were used to bind candidate antioxidant compounds from Cymbopogon spp. Three-dimensional structures of each target were obtained from the Protein Data Bank (https://www.rcsb.org/), accessed in October 2023, pdb format [28].

Drug-likeness prediction

This test is used to evaluate the characteristics of a query compound as a drug-like molecule by referring to several physicochemical rules using a specific method [29]. Lipinski’s Rule of Five is a drug-likeness prediction method based on molecular mass, high lipophilicity, hydrogen donor–acceptor interactions, and molar refractivity. In the present study, Lipinski's prediction was performed using the SCFBIO server (http://www.scfbio-iitd.res.cn/software/drugdesign/lipinski.jsp), accessed in October 2023, with a query compound sdf file [29].

Virtual screening

Screening the activity of query compounds through the computational simulation of ligand-target binding is known as virtual screening [30]. In this study, we used molecular docking to identify the antimicrobial and antioxidant activities of Cymbopogon spp. by inhibiting FtsZ, Aquaporin Z, SA, and AHAS. Simulation was performed with the position of FtsZ autogrid docking from Bacillus subtilis: Center (Å) X: 28.973, Y: -5.776, Z: -2.052; Dimensions (Å) X: 67.136, Y: 55.679, Z: 64.620; Aquaporin Z from Escherichia coli: Center (Å) X: -33.395, Y: 33.776, Z: 10.766; Dimensions (Å) X: 44.594, Y: 34.944, Z: 53.370; SA from Staphylococcus aureus: Center (Å) X: 19.168, Y: 10.482, Z: 11.952; Dimensions (Å) X: 43.435, Y: 56.922, Z: 39.169; and AHAS from Candida albicans Center (Å) X: 64.379, Y: 247.139, Z: 45.709; Dimensions (Å) X: 63.263, Y: 53.140, Z: 63.241; SOD1 Center (Å) X: -73.660, Y: 25.713, Z: 18.286; Dimensions (Å) X: 69.194, Y: 75.585, Z: 41.691; and catalase Center (Å) X: 21.796, Y: 27.124, Z: 42.396; Dimensions (Å) X: 115.915, Y: 105.681, Z: 110.519 via PyRx software (v1.0.0; Scripps Research, USA) with an academic license. The structure of the ligand–protein complex was visualized using PyMOL (v.2.5.2; Schrödinger, Inc.) with an academic license for structural selection and single coloration [31].

Ligand–protein analysis

Identification of the position and type of chemical bond interactions in ligand–protein molecular complexes was visualized through Discovery Studio Visualizer™ (v.16.1; Dassault Systèmes SE, France), accessed in October 2023. This software can visualize weak bond interactions, such as van der Waals, hydrogen, hydrophobic, pi/alkyl, and electrostatic interactions. Weak bonds in ligand–protein complexes trigger biological activities such as inhibitory responses. In this study, the positions of specific amino acid residues in the ligand-pocket binding domain were referred to as interaction hotspots [32].

Molecular dynamic simulation

Chemical bond interaction stability at interaction hotspots was identified through molecular dynamic simulations using the CABS-flex server (v2.0; http://biocomp.chem.uw.edu.pl/CABSflex2/index), accessed in October 2023. The root-mean-square fluctuation (RMSF) is used to determine the stability of the molecular interactions. In this study, the RMSF refers to the conformational changes in amino acid residues at the interaction hotspot with a query ligand [33]. RMSF values were determined in CABS-flex (v2.0) based on rigidity, restraints, global c-alpha, side chain, number of cycles, cycles between trajectories, temperature range, and random number generator seed [34].

Antimicrobial and antioxidant pathway prediction

The pathways of Cymbopogon spp.-derived compounds with more negative binding affinity values as antimicrobials and antioxidants were predicted using the SwissTargetPrediction server (http://www.swisstargetprediction.ch/), accessed October 2023, for target identification in Homo sapiens. Furthermore, pathway validation was performed through STRINGdb connected to Cytoscape software (v3.9.1). Cytoscape is used to identify interactions between query compounds and target proteins through molecular mechanisms [21].

In vitro antioxidant activity

The 2,2-diphenyl-1-picryl-hydrazyl-hydrate (DPPH) inhibition assay

The DPPH inhibition assay was carried out as described previously [35]. Samples (100 μL) were diluted in methanol to concentrations of 1.075–200 µg/mL and mixed with 100 µL DPPH at 0.2 mM concentration. The mixture was then incubated at room temperature for 30 min. Ascorbic acid and trolox were used as positive controls. A microplate reader (Thermo Fisher Scientific, Waltham, MA, USA) was used to measure DPPH inhibition at 517 nm. The DPPH radical-scavenging capacity (%) was subsequently calculated using the following formula:

$$\frac{{A}_{control}-{A}_{sample}}{{A}_{control}}\times 100\%$$
(1)

where Acontrol represents the absorbance from the DPPH reagent and Asample represents the absorbance of the DPPH reagent and sample mixture. The percentage of inhibition at each concentration was plotted and linearly regressed to obtain the half-maximal inhibitory concentration (IC50). The antioxidant activity of the plant extracts was determined based on IC50 value categorization: very strong (< 50 µg/mL), strong (50–100 µg/mL), moderate (101–250 µg/mL), weak (251–500 µg/mL), and no antioxidant activity (> 500 µg/mL) [36].

The 2,2’-azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) inhibition assay

The ABTS assay was conducted as described previously [37]. ABTS (7 mM) and 2.4 mM potassium persulfate solutions were mixed to create the ABTS reagent. The mixture was left to sit for 12–16 h in the dark at room temperature. Absorbance of the reagent was measured at 734 nm. Next, 100 μL of samples were combined with 100 μL of the ABTS reagent at concentrations of 1.075–200 µg/mL. They were then incubated for 5 min at room temperature in the dark. The positive controls used were the same as those used for the DPPH assay: ascorbic acid and Trolox. After incubation, the absorbance was measured at 734 nm using a microplate reader (Thermo Fisher Scientific). The percent inhibition and IC50 values for the ABTS assay were calculated using the same formula as for the DPPH assay.

In vitro cytotoxicity of Cymbopogon spp

The cytotoxicity of all Cymbopogon spp. ethanolic extracts was evaluated using a modified version of the 3-[4, 5-dimethylthiazol-2-yl] 2, 5-diphenyl tetrazolium bromide (MTT) assay as described previously [21]. Hepatocyte-derived cellular carcinoma cells (Huh7it-1 cells) were grown at 37 °C in complete Dulbecco's modified Eagle's medium supplemented with 1% (v/v) glutamine (200 mM) at 5% CO2 and 95% humidity. The samples were dissolved in DMSO before dilution to different concentrations (0.1–1000 µg/mL). After 48-h incubation at 37 °C in 5% CO2 and 95% humidity, 5 mg/mL of MTT solution in phosphate buffered saline was added to each well. The cultures were then incubated for 4 h. Next, the solution was withdrawn, and 100 µL DMSO was added to dissolve the formazan crystals as MTT products. Cell viability was measured at 560 and 750 nm using a spectrophotometer. The half-cytotoxic concentration (CC50) was assessed by plotting the cell viability percentage and a series of concentrations to make the regression linear using Microsoft Excel (version 20.0; Microsoft Corporation, Redmond, Washington, USA). The experiment was performed in triplicate.

Antimicrobial activity

Antimicrobial assay: Preparation of media and inoculum

Four microbial strains, B. substilis TISTR 1248, S. aureus ATCC 25923, E. coli ATCC 25922, and C. albicans ATCC 10231, were tested using microbial assays. Solid and liquid media were prepared to conduct antimicrobial assays and maintain microbial cultures. Nutrient Agar (NA) and Potato Dextrose Agar (PDA) were used for the antibacterial and antifungal assays, respectively. In addition, NA and PDA were used to maintain the reference microbial culture in the reaction tubes. Nutrient Broth (NB) and isotonic NaCl solutions were used as liquid media. NB was used as the medium for bacterial strain subculture and preculture, whereas isotonic NaCl was only used for fungal strain preculture. The bacterial strains were precultured in culture bottles filled with sterile NB, incubated for 24 h at 37 °C, and diluted into 10% cultures before inoculation. The absorbance of 10% bacterial cultures was adjusted with sterile water using a spectrophotometer to reach the 0,5 McFarland standard. The fungal strain was precultured by homogenizing one loop of fungi from the reference culture into a culture bottle filled with sterile NaCl solution, and the culture was made immediately before inoculation [38].

Well diffusion assay

The well diffusion assay was used to evaluate antimicrobial activity [39]. The antimicrobial activity of the extracts was determined based on the presence of inhibition zones on the surface of NA and PDA. The first 10 mL of agar medium was poured into sterilized Petri dishes and solidified as the first layer. Next, another 30 mL of agar was poured onto the solidified first layer and sat until it cooled but did not solidify. When the second layer cooled, 1 mL of bacterial culture was inoculated into NA, 1 mL of fungal culture was inoculated into PDA, and the second layer was placed again until it completely solidified. Four wells were then placed on the second solidified layer, into which the samples and controls were applied. The samples used were extracts diluted with 10% DMSO (250 and 500 mg/mL). DMSO (10%) and 1000 µg/mL chloramphenicol were used as the negative and positive controls, respectively, for the antibacterial assay. Similarly, nystatin (1000 µg/mL) was used as the positive control for the antifungal assay. The controls were tested on a separate agar media. Ten microliters of diluted samples were applied four times to each well. The plates were then incubated at 37 °C for 24 h for bacterial strains and 48 h for the fungal strain. After incubation, the plates were examined for the presence of inhibition zones on the agar surfaces. The diameter of the inhibition zone (DIZ, mm) was measured using Vernier calipers, and the mean was calculated and compared to that formed by the positive control. The percentage of inhibition (PI) was used to analyze the DIZ of the sample compared to the positive control and was calculated using the following formula (2):

$$PI \left(\%\right)=\frac{Mean zone of inhibition of the extract}{Zone of inhibition of the positive control}\times 100\%$$
(2)

Statistical analyses

The data of antioxidant, antimicrobial, and cytotoxic activities are expressed as the mean ± standard deviation. The IC50 and CC50 values for in vitro antioxidant activity and cytotoxicity were calculated through linear regression using Microsoft Excel (version 20.0) [40]. Antioxidant, cytotoxicity, and antimicrobial data were analyzed using the Statistical Package for the Social Sciences (SPSS), version 18.0 (IBM Corp. SPSS Inc., Chicago, IL, USA).

Result and discussion

The extract yield and metabolite profile of Cymbopogon spp

The yields of crude ethanolic extracts of Cymbopogon spp. roots, stems, and leaves were approximately 11.40 ± 0.12 – 17.76 ± 0.2% (Additional file 1). GC–MS analysis revealed 53 metabolites from the extracts (Additional files 2 and Fig. 1); these were distributed evenly in every part of the Cymbopogon spp. organs. In contrast, 29, 13, and five metabolites were detected in only one, two, and three extracts, respectively. Additionally, only one metabolite was detected in four, five, six, seven, and nine extracts. 2,5-bis(1,1-dimethylethyl)- phenol (27.87%), alpha-cadinol (26.76%), and 1,2-dimethoxy-4-(1-propenyl)-benzene (20.56%) showed the highest percentage area; therefore, they were predicted to be major compounds. Previous studies have shown the bioactivity of some metabolites. Alpha-cadinol has antifungal, antimicrobial, and antioxidant activities [38, 39, 41, 42]. Similarly, selina-6-en-4-ol and n-hexadecanoic acid exhibit antimicrobial and antioxidant activities [43, 44]. Therefore, in silico and in vitro antimicrobial and antioxidant activities were determined for crude extracts of every Cymbopogon spp. organ. Computational pathway prediction of metabolites with antimicrobial activity is shown in Table 1.

Fig. 1
figure 1

GC–MS chromatogram of Cymbopogon spp. ethanolic extract. A Roots, B stems, C leaves. I. Cymbopogon citratus, II. Cymbopogon nardus, III. Cymbopogon winterianus. a. Tetraethyl silicate; b. Geraniol; c. Methyleugenol; d. Benzene, 1,2-dimethoxy-4-(1-propenyl)-; e. Naphthalene, 1,2,3,4,4a,5,6,8a-octahydro-7-methyl-4-methylene-1-(1-methylethyl)-, (1.alpha.,4a.beta.,8a.alpha.)-; f. gamma.-Muurolene; g. Phenol, 2,5-bis(1,1-dimethylethyl); h. Naphthalene, 1,2,3,5,6,8a-hexahydro-4,7-dimethyl-1-(1-methylethyl)-, (1S-cis)-; i. tau.-Muurolol; j. (1S,4aS,7R,8aS)-1,4a-Dimethyl-7-(prop-1-en-2-yl)decahydronaphthalen-1-ol; k. 1-((1S,3aR,4R,7S,7aS)-4-Hydroxy-7-isopropyl-4-methyloctahydro-1H-inden-1-yl)ethanone; l. Benzenepropanoic acid, 3,5-bis(1,1-dimethylethyl)-4-hydroxy-, methyl ester; m. Phytol. Green arrow: Selin-6-en-4.alpha.-ol; yellow arrow: alpha.-Cadinol; black arrow: Hexadecanoic acid, methyl ester 

Table 1 The selected metabolites of Cymbopogon spp. from GCMS as ligand information with PubChem database for computational analysis

In silico analysis of antioxidant and antimicrobial activity

Ligand retrieval

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 analysis

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.

Table 2 The results of drug-likeness prediction

Virtual analysis

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.

Fig. 2
figure 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.

Table 3 Binding affinity of Cymbopogon spp. compounds
Fig. 3
figure 3

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).

Fig. 4
figure 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).

Fig. 5
figure 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].

Fig. 6
figure 6

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

In vitro antioxidant activity

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].

Table 4 Half-maximum inhibition concentration (IC50) of three commercial Cymbopogon spp. Extracts

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].

In vitro cytotoxicity of Cymbopogon spp ethanolic extracts

The toxicity of Cymbopogon spp. ethanolic extracts to Huh7it-1 cells was assessed using an MTT assay (Table 5). C. citratus leaf extract had the lowest cytotoxic effect on the human hepatocyte cells (1421.07 ± 20.88 µg/mL), followed by C. citratus root (822.18 ± 36.45 µg/mL), C. winterianus leaf (768.79 ± 74.19 µg/mL), C. nardus root (753.81 ± 40.44 µg/mL), C. winterianus root (621.81 ± 10.59 µg/mL), C. winterianus stem (555.71 ± 44.72 µg/mL), C. citratus stem (477.75 ± 4.01 µg/mL), C. nardus leaf (437.39 ± 7.47 µg/mL), and C. nardus stem (426.74 ± 26.05 µg/mL) extracts. The 50 percent cytotoxic concentration (CC50) is defined as the concentration of the sample compound required to reduce cell viability by 50% [60]. United States National Cancer Institute (US NCI) plant screening program, crude extract is considered cytotoxic when CC50 is lower than 30 µg/mL. This study showed that the CC50 of all extracts were higher than 30 µg/mL, which means that all extracts were low toxicity [61].

Table 5 The toxicity result of three commercial Cymbopogon spp. extracts

Antimicrobial activity

Antimicrobial tests against bacteria and yeasts were performed using all crude extracts. All C. citratus extracts, except the root extract at 250 mg/mL and the leaf extract at both concentrations, showed active antibacterial activity against B. subtilis (Table 6), with a PI of 19.58 ± 1.72%, 17.27 ± 0.77%, and 17.75 ± 0.48% for the root extract at 500 mg/mL and the stem extract at 250 and 500 mg/mL, respectively. All C. nardus extracts exhibited antibacterial activity against B. subtilis (Table 6) (50.57 ± 2.92% and 52.33 ± 6.91% inhibition by 250 and 500 mg/mL root extracts, 34.99 ± 3.00% and 45.93 ± 7.62% by 250 and 500 mg/mL stem extracts, and 39.43 ± 4.37% and 51.93 ± 2.19% by 25% and 50% leaf extracts, respectively). Similarly, C. winterianus extracts showed antibacterial activity against B. subtilis (Table 6) (22.45 ± 1.28% and 28.28 ± 2.59% inhibition at 250 and 500 mg/mL root extracts, 48.09 ± 1.42% and 64.35 ± 1.50% at 25% and 50% stem extracts, and 19.97 ± 3.41% and 27.69 ± 3.94% at 250 and 500 mg/mL leaf extracts, respectively). The C. nardus root (89.38 ± 2.42% and 96.41 ± 2.61% at 250 and 500 mg/mL), stem (28.98 ± 0.30% and 32.03 ± 1.08% at 250 and 500 mg/mL), and leaf (76.70 ± 3.55% and 86.72 ± 3.15% at 250 and 500 mg/mL) extracts showed excellent antibacterial activity against S. aureus. Additionally, the stem extracts of C. citratus (28.98 ± 0.30% and 32.03 ± 1.08% at 250 and 500 mg/mL) and C. winterianus (48.08 ± 5.74% and 64.46 ± 5.40% at 250 and 500 µg/mL) showed antibacterial activity. Only C. nardus extracts showed active antibacterial activity against E. coli, with a PI of 88.97 ± 5.71% and 127.24 ± 5.03% at 25% and 50% root extracts, 47.10 ± 4.01% and 68.01 ± 5.84% at 250 and 500 mg/mL stem extracts, and 105.56 ± 7.41% and 124.85 ± 6.09% at 250 and 500 mg/mL leaf extracts. Of all extracts tested against C. albicans, only the root extracts of C. nardus showed active antifungal activity, with a PI of 23.61 ± 1.26% and 24.33 ± 1.23% at 250 and 500 mg/mL, respectively.

Table 6 Diameter of inhibition zone (DIZ) and percentage of inhibition (PI) of C. citratus ethanolic extract against Bacillus subtilis, Staphylococcus aureus, Escherichia coli, and Candida albicans

C. citratus extracts showed no significant inhibition zones against E. coli and C. albicans. In contrast, they had a considerable effect against B. subtilis and S. aureus. In a previous study, C. citratus essential oil exhibited stronger antibacterial activity against gram-positive bacteria (B. subtilis and S. aureus) than against gram-negative bacteria (E. coli) [59]. Additionally, S. aureus was less resistant to antimicrobials than B. subtilis in the present study, which is consistent with previous findings [9]. The contrast in activity between C. citratus against E. coli and S. aureus is also comparable to a similar test carried out by Subramaniam et al., in which C. citratus root and leaf extracts exhibited stronger antimicrobial activity against S. aureus than against E. coli [9]. However, only the stems exhibited antibacterial activity against S. aureus in the present study, and there was no activity against E. coli. Moreover, antifungal activity against C. albicans was insignificant. This is contrary to the findings of Pratama and Permana [13], who reported that C. citratus essential oils showed significant inhibition zones at 200, 300, and 500 mg/mL.

All C. nardus extracts showed substantial inhibition zones against B. subtilis, S. aureus, and E. coli, whereas only the root extract showed inhibition zones against C. albicans. The root extracts exhibited higher antibacterial activity against gram-positive bacteria, followed by the leaf and stem extracts. In contrast, the leaf extracts were more effective against gram-negative bacteria, followed by the root and stem extracts. The antimicrobial activity of C. nardus extracts is related to the presence of terpenoids (especially monoterpenes) and phenolic compounds. Kamal et al. [13] reported that C. nardus essential oil has antibacterial activity against both gram-positive and gram-negative bacteria because of the presence of citronellal and z-citral, two of the 13 monoterpenes identified in the essential oil. In the present study, all C. nardus extracts exhibited strong antibacterial activity against both gram-positive and gram-negative bacteria. Antibacterial activity was comparable to that reported by Yunita et al [62], in which the leaf extract of C. nardus showed a high response against S. aureus at 20 mg/mL. Similarly, Shintawati et al. [63] showed that S. aureus is less susceptible to the C. nardus oil obtained from leaves than E. coli, which was also reflected in our study. Antifungal activity was due to the presence of citronellal [13]. However, C. nardus root extract exhibited weaker antifungal activity in the present study, as demonstrated by the smaller inhibition zone observed against C. albicans.

C. winterianus extracts showed no significant antimicrobial activity against E. coli and C. albicans. Rather, they showed activity against B. subtilis and S. aureus, similar to C. citratus extracts. Although the C. citratus stem and root extracts only exhibited antibacterial activity against B. subtilis at 500 mg/mL, all C. winterianus extracts were effective at both 250 and 500 mg/mL. Additionally, only the stem extract of C. winterianus showed antibacterial activity against S. aureus, similar to C. citratus. However, its activity against S. aureus was weaker than that against B. subtilis. This result is consistent with that of Munda et al. [64] in which C. winterianus essential oil showed wider inhibition zones against B. subtilis than those against S. aureus. It also exhibited significant antibacterial activity against E. coli and antifungal activity against C. albicans, as demonstrated by wider inhibition zones; this contradicted our study findings. The percentage of the inhibition zones showed that the C. nardus root, stem, and leaf extracts were significantly effective antimicrobial agents, with the root extract showing the highest inhibition (127%).

Cymbopogon essential oil, also known as citronella oil, mainly consists of three compounds: citronellal, citronellol, and geraniol [65]. Citronellal is an aldehyde monoterpene with a fresh scent and desiccant characteristics. This compound causes cell death through dehydration. Citronellol and geraniol are alcoholic monoterpenes that inhibit bacterial growth by disrupting cell division or cellular membranes [66]. Kumala et al. [66] and Munda et al. [64] have demonstrated that C. nardus and C. winterianus essential oils contain high concentrations of these three compounds. Citral, another main component of these two essential oils, is a mixture of trans-isomers known as neural isomers [67]. Its content is used to determine the quality of essential oil. Moreover, high contents of this compound, geranial, and neral have been reported in C. citratus essential oil [68]. Citral is also abundant in the root stalk and shoot parts of C. winterianus. It inhibits bacterial growth by impairing cell membrane integrity, leading to membrane rupture [69]. Similarly, citronellal confers antifungal activity by inhibiting ergosterol biosynthesis, resulting in cell membrane damage [70].

The antibacterial and antifungal activities exhibited in the present study, especially by C. nardus extracts, were likely due to the abundance of these compounds. However, the abundance of these compounds was not the sole factor in their antimicrobial activity. Other factors, such as the resistance of gram-positive and gram-negative bacteria, may also play a role. The antimicrobial resistance of gram-positive bacteria is likely due to the presence of thick peptidoglycan in the cell wall, which makes it difficult for antimicrobial agents to penetrate. The antimicrobial resistance of gram-negative bacteria is due to porin channels in the outer membrane, where lipophilic drugs have difficulty penetrating [71]. This indicates that antagonistic and synergistic effects are caused by multiple components of the extracts [68].

In thi study, alpha-cadinol was predicted as a key compound in the mechanism underlying the antioxidant and antimicrobial activity of Cymbopogon spp. using in silico analysis (Fig. 6). In vitro antioxidant and antimicrobial assays also showed that C. winterianus and C. nardus leaf extracts had the best antioxidant and antimicrobial activities, with low cytotoxicity observed. Although Cymbopogon spp. generally demonstrated antioxidant and antimicrobial activities in the present study, C. winterianus and C. nardus leaf extracts were the most effective. These results suggest that Cymbopogon spp. extracts can be used to develop new antimicrobial drugs. However, bioassay-guided isolation is required to isolate pure compounds, such as alpha-cadinol.

Conclusion

This study highlights the antioxidant and antimicrobial activities of C. citratus L., C. nardus (L.) Rendle., and C. winterianus Jowitt, as well as their low toxicity. We conclude that alpha-cadinol is a key Cymbopogon spp. compound that plays an important role in antioxidant and antibacterial activity pathways. Cymbopogon spp. leaves exhibited the highest antioxidant activity; C. nardus leaf extract exhibited the highest antimicrobial activity against E. coli and S. aereus; and C. winterianus stem extract showed the highest antimicrobial activity against B. substilis. These findings provide a theoretical basis for the clinical application of Cymbopogon spp.-derived compounds. However, bioassay-guided isolation of alpha-cadinol and clinical safety assessment of this extract and important compound are required for further pharmaceutical application.