Pharmacognosy Magazine

: 2021  |  Volume : 17  |  Issue : 6  |  Page : 268--277

Plectranthus amboinicus (Spreng.) Semi-purified Fractions with Selective β-Glucuronidase Inhibition Elucidated with gas chromatography-mass spectrometry and in silico docking

Michael Russelle S. Alvarez1, Junie Billones2, Chun-Hung Lin3, Francisco Heralde4,  
1 Institute of Chemistry, College of Arts and Sciences, University of the Philippines Los Baños, Laguna; Department of Chemistry, College of Arts and Sciences, Isabela State University, Isabela, Philippines
2 Department of Physical Sciences and Mathematics, College of Arts and Sciences, University of the Philippines Manila, Manila City, Philippines
3 Institute of Biological Chemistry, Academia Sinica, Taipei, Taiwan
4 Department of Biochemistry and Molecular Biology, College of Medicine, University of the Philippines Manila, Manila City, Philippines

Correspondence Address:
Michael Russelle S. Alvarez
Isabela State University, Isabela


Background: Plectranthus amboinicus (Spreng.) is an herb commonly used in folk medicine and food by several Asian countries. The bioactivities of this medicinal plant have so far not been linked to a specific enzyme target. Bacterial beta-glucuronidases (β-GUS) expressed by human gut microbiota affect xenobiotic processing by reactivating toxic substances (e.g., anticancer drugs, nonsteroidal anti-inflammatory drugs, and food carcinogens) in the gut lumen. Objectives: An approach to alleviating the toxic effects of these compounds is by inhibiting bacterial β-GUS. Materials and Methods: We determined the Escherichia coli β-GUS inhibitory activity of P. amboinicus leaves using a bioassay-guided purification approach. The P. amboinicus chloroform extract was purified using normal-phase column chromatography to produce several fractions. The fractions were screened for E. coli β-GUS inhibitory activity using the 4-Methylumbelliferyl glucuronide (4-MUG) assay. Fractions with high activity were assayed further to determine toxicity against E. coli and selectivity compared to human β-GUS. Highly-active and highly-selective fractions were further characterized using gas chromatography-mass spectrometry (GC-MS) and in silico docking to identify specific compounds. Results: Assay-guided purification of the crude chloroform extract with β-GUS inhibitory activity (IC50 = 57.8 μg/mL) yielded four fractions with high activity: Fraction-543W (IC50 = 16.24 μg/mL), Fraction-5231 (IC50 = 3.087 μg/mL), and Fraction-52335A (IC50 = 12.93 μg/mL). The crude extract and fractions exhibited high selectivity for E. coli β-GUS against human β-GUS (P < 0.0001, α =0.05). The antimicrobial assay of fractions showed no toxic effects on E. coli. GC-MS profiling of the active fractions identified the compounds present to be similar to essential oil extracts of P. amboinicus reported previously. Ranking of these compounds by in silico identified the compounds with high binding affinity: Phthalic acid, cyclobutyl tridecyl ester (-7.5 kcal/mol) from Fraction-543W, N-Benzyl-2-allyl-2-tosyl-4-penten-1-amine (-8.0 kcal/mol) from Fraction-5231, and Dehydroabietic Acid (-7.9 kcal/mol) from Fraction-52335A. By comparing to binding modes of reported inhibitors, we show that these compounds also interact with active site residues Tyr469 and Tyr472, and with several residues in the β-GUS bacterial loop. Conclusion: Herein, we identified highly-active and highly-selective E. coli β-GUS inhibitors from P. amboinicus leaf chloroform extracts, utilizing a bioassay-guided purification coupled by metabolomics and in silico docking approach. This is the first report on the potential of P. amboinicus as selective inhibitor of E. coli β-GUSs.

How to cite this article:
S. Alvarez MR, Billones J, Lin CH, Heralde F. Plectranthus amboinicus (Spreng.) Semi-purified Fractions with Selective β-Glucuronidase Inhibition Elucidated with gas chromatography-mass spectrometry and in silico docking.Phcog Mag 2021;17:268-277

How to cite this URL:
S. Alvarez MR, Billones J, Lin CH, Heralde F. Plectranthus amboinicus (Spreng.) Semi-purified Fractions with Selective β-Glucuronidase Inhibition Elucidated with gas chromatography-mass spectrometry and in silico docking. Phcog Mag [serial online] 2021 [cited 2022 Nov 27 ];17:268-277
Available from:

Full Text


  • Plectranthus amboinicus, locally known in the Philippines as oregano, is commonly used in folk medicine and food ingredient by several Asian countries. Although used in several ethnopharmacological applications as treatment to burns, bruises, headaches, and stomachaches, there has so far been no identified functional protein target of the phytochemicals contained in its extract. In this study, we aim to link its ethnopharmacological properties to its inhibition of bacterial beta-glucuronidases (β-GUS), specifically that of Escherichia coli. β-GUS are expressed by gut microbiota that affect xenobiotic processing of certain toxic substances. By the action of these enzymes, glucuronide-conjugated toxic substances that leave the gut lumen are reactivated by release of the glucuronide moiety. As such, through bioassay-guided purification coupled by gas chromatography-mass spectrometry characterization and in silico docking, we identified several compounds that could potentially inhibit bacterial β-GUS. Specifically, these are the compounds Phthalic acid, cyclobutyl tridecyl, N-Benzyl-2-allyl-2-tosyl-4-penten-1-amine, and Dehydroabietic Acid. This is the first report on the potential of P. amboinicus as selective inhibitor of E. coli β-glucuronidase.


Abbreviations used: 4-MUG assay: 4-methylumbelliferyl assay; ANOVA: Analysis of variance; β-GUS: Beta-glucuronidase; CHCl3 extract: Chloroform extract; CPT-11: Camptothecin-11; DMSO: Dimethyl sulfoxide; E. coli: Escherichia coli; GC-MS: Gas chromatography-mass spectrometry; GPS: Global positioning system; IC50: 50% Inhibitory concentration; LB Broth: Luria-Bertani broth; MPLC: Medium performance liquid chromatography; NaOH: Sodium hydroxide; NSAIDs: Nonsteroidal anti-inflammatory drugs; OD600: Optical density at 600 nm; P. amboinicus: Plectranthus amboinicus; PDB: Protein Data Bank; RMSD: Root mean square deviation; SN-38G: 7-Ethyl-10-hydroxycamptothecin glucuronide; UDP-glucuronyltransferase: Uridine diphosphate-glucuronyltransferase.


Plectranthus amboinicus (Spreng), also called Coleus aromaticus (Benth.), Coleus suganda (Blanco), or Plectranthus aromaticus (Roxb.), is an aromatic succulent herb commonly used in folk medicine and food by several countries, including the Philippines (Oregano, latai, suganda), China (da shou xiang), India (Pashan bhedi, Karpooravalli, Patharchur), and Malaysia (Daun bangun-bangun, Pokkok bangun-bangun), among others.[1] In the Philippines, macerated fresh leaves are applied to burns, bruises, and insect bites. The leaves are also applied to the temples and forehead for headaches. An infusion or syrup from the leaves are used for dyspepsia or asthma. The Chinese also use the juice of the leaves with sugar for coughs. In Indo-China, it is employed as an infusion for asthma, chronic coughs, and epilepsy. Malays use the juice of the plant for stomach aches[2] Previously reported bioactivities of this plant species include antibacterial,[3],[4],[5],[6],[7],[8] antifungal,[9] antiviral,[10],[11] antitumorigenic,[12] anti-inflammatory,[12],[13],[14] antioxidant[7],[15] activities and activity against digestive diseases such as diarrhea.[16],[17],[18]

Escherichia coli beta-glucuronidases (E. coli β-GUS) is a member of the Family 2 glycosyl hydrolases, which also includes human (human β-GUS), mice and rat β-GUSs, as well as Clostridium acetobutylicum, Klebsiella pneumoniae and Lactobacillus bulgaricus β-GUS.[19] These enzymes catalyze the hydrolysis of β-D-glucuronic acid residues from the non-reducing end of carbohydrates.[20] The E. coli β-GUS occurs as a tetramer composed of two asymmetric units, which in turn is composed of two monomers of 597 ordered residues. It contains several domains: The N-terminal domain (residues 1–180) is the sugar binding while the C-terminal domain (residues 274–603) forms an αβ barrel containing the active-site residues Glu413 and Glu504, and the region between the 2 domains show an immunoglobulin-like β-sandwich domain found in other Family 2 glycosyl hydrolases. In addition, it contains a 17-residue “bacterial loop” that is not found in the human form of the enzyme.[21] Bacterial β-GUSs expressed by gut microbiota affect xenobiotic processing in humans, by removing the glucuronic acid that are conjugated to xenobiotics (in a process called glucuronidation) in the liver by UDP-glucuronyltransferases.[22] This process essentially releases the original xenobiotic into the gut lumen, thus affecting the toxic properties of anticancer compounds such as Camptothecin-11 (CPT-11),[23],[24],[25] nonsteroidal anti-inflammatory drugs,[26],[27] and food carcinogens.[28],[29] Specifically, the release of these xenobiotics into the gut results in their toxic properties (e.g., diarrhea, epithelial injury). Thus, targeted inhibition of this bacterial β-GUS could ameliorate these toxic effects. Indeed, several studies have shown that inhibition of bacterial β-GUS alleviated nonsteroidal anti-inflammatory drugs-induced enteropathy in rats[27] and mice,[26] enhanced CPT-11 anticancer activity[30] and modulated CPT-11 toxic side effects.[21],[23],[31],[32],[33]

Herein, we present for the first time, data on the potential of P. amboinicus (Spreng.) crude extract and fractions as selective inhibitor of E. coli β-GUS. We also show that the chloroform extract and fractions obtained do not affect E. coli viability, which is important in preserving the integrity of gut microbiota. Finally, characterization of the extract was undertaken using gas chromatography-mass spectrometry (GC-MS) and the binding modes of the detected putative compounds were determined through in silico modeling.

 Materials and Methods

Materials and reagents

All reagents used were analytical grade. The solvents used for the extraction, gravity column, and medium-pressure liquid chromatography (MPLC) were from Duksan Pure Chemicals (Taipei). Dimethyl sulfoxide used for sample preparation in the assay was analytical grade and obtained from Fisher Chemical (Taipei). 4-Methylumbelliferyl glucuronide (4-MUG) substrate used in the E. coli β-GUS inhibition assay was from Sigma-Aldrich. E. coli cells for the antibacterial assay, purified E. coli and human β-GUS were kindly provided by Dr. Chun-Hung Lin (Academia Sinica, Taiwan).

Sample collection and extraction

P. amboinicus (Spreng.) leaves were obtained from Echague, Isabela, Philippines (GPS: 16.706292, 121.676219). P. amboinicus was authenticated by Michelle Alejado-San Pascual from Botany Division, Museum of Natural History, Laguna, Philippines. Voucher specimens (Accession number: #073652) were kept at the Herbarium of the University of the Philippines Los Baños. Gathered leaves of P. amboinicus were air-dried in the Biochemistry Laboratory, College of Medicine, and University of the Philippines Manila, Philippines, at room temperature. The dried leaves were ground using a mechanical blender to a coarse powder. For preliminary screening, 10 g of the powdered leaves were extracted in each of the following solvents at a ratio of 1 g sample: 10 mL solvent: Hexane, ethyl acetate, dichloromethane, chloroform, ethanol, methanol, and distilled water. After 12 h of soaking with occasional shaking, the resulting mixture was filtered. The extracts were concentrated using a rotary evaporator at 40°C and stored in aluminum foil covered preweighed glass vials and the solvents removed completely using a high-vacuum pump. Dried extracts were stored at 4°C. Upon determining the most active extract was the chloroform extract using the E. coli β-GUS inhibition assay screening, all 200 g of the plant material was extracted using the same procedure.

Escherichia coli beta-glucuronidases inhibition assay

E. coli β-GUS inhibition assay was performed after each round of purification of the extract, to constitute bioassay-guided fractionation. The assay was modified from the 4-MUG assay.[34] For the enzyme reaction, 5 μL of the extract (prepared in DMSO) or DMSO (negative control) was added to 75 μL HEPES buffer and 10 μL enzyme (50 ng/mL). These were incubated at 37°C for 20 min. Then, 10 μL of 4-MUG substrate (1.5 mM) were added and incubated for another 15 min in 37°C. Afterward, 100 μL stop buffer (glycerin, 1.0 M NaOH) was added to terminate the reaction. The percent inhibition was estimated using fluorescence measurements (Ex: 365 nm, Em: 455 nm) and calculated using the following equation:


Equation 1: Equation for calculating average percent inhibition.

The dose-response assay was performed based on this protocol, after optimizing the concentration ranges (seven 2-fold dilutions) for each crude extract and fraction. IC50 was determined using GraphPad PRISM 7 software.

Purification of fraction 5231

The crude chloroform extract was loaded into a glass column (27.6 cm × 4.8 cm) packed with silica then eluted using the following gradient: 100% hexane, 1:1 hexane-chloroform, 100% chloroform, 1:49 methanol-chloroform, 1:39 methanol-chloroform, 1:29 methanol-chloroform, 1:19 methanol-chloroform, 1:14 methanol-chloroform, 1:9 methanol-chloroform, 1:4 methanol-chloroform, 1:2 methanol-chloroform, 1:1 methanol-chloroform, 100% methanol. The eluents were collected every 15 mL and pooled using 1:19 methanol-chloroform and 1:10 methanol-chloroform. From this, fraction 5 was collected and then purified further using another glass column (27.6 cm × 4.8 cm) packed with silica, which was eluted using 100% hexane, 1:1 hexane-chloroform, 100% chloroform, 1:49 methanol-chloroform, 1:44 methanol-chloroform, 1:39 methanol-chloroform, 1:34 methanol-chloroform, 1:29 methanol-chloroform, 1:24 methanol-chloroform, 1:19 methanol-chloroform, 1:14 methanol-chloroform, 1:9 methanol-chloroform, 1:4 methanol-chloroform, 1:1 methanol-chloroform, 100% methanol. The eluents were collected every 15 mL and pooled using 1:19 methanol-chloroform and 1:10 methanol-chloroform. This time, fraction 2 was collected for final purification using MPLC (CombiFlash Rf 200® [Teledyne, ISCO]). Here, a 4 g column and 18 mL/min flow rate used. The solvent gradient was: 0% ethyl acetate (5 min), 0% ethyl acetate-30% ethyl acetate (10 min), 30% ethyl acetate (5 min), 30% ethyl acetate-50% ethyl acetate (5 min), 50% ethyl acetate (5 min), 50%–75% ethyl acetate (10 min), 75% ethyl acetate (5 min), 75%–100% ethyl acetate (10 min), change solvent ethyl acetate to chloroform (5 min), 100% chloroform (5 min), 100%–70% chloroform-methanol (3 min), 70% chloroform-methanol (5 min). From here, fraction 1 was collected (yield = 0.5 mg) and labeled Fraction 5231.

Purification of fraction 52335A

For the purification of Fraction 52335A, the same protocol was followed for the purification of Fraction 5231, until the MPLC step. Here, after the purification run, fraction 3 was collected and then purified further using gravity column using the following solvent gradient: 100% hexane, 5% ethyl acetate, 7.5% ethyl acetate, 10% ethyl acetate, 12.5% ethyl acetate, 15% ethyl acetate, 17.5% ethyl acetate, 20% ethyl acetate, 100% ethyl acetate, 10% methanol, 30% methanol, 100% methanol. Eluates were collected every 5 mL and then pooled using 1:19 MeOH-CHCl3 solvent system. From here, fraction 5 was collected. Solvent-solvent extraction using equal parts hexane and acetonitrile was used to obtain Fraction 52335A (yield = 0.5 mg).

Purification of fraction 543W

For the purification of Fraction 543W, the fourth fraction of the second purification step in Fraction 5231 was collected instead. Then, purification by medium-phase column chromatography: 0% ethyl acetate (5 min), 0% ethyl acetate-25% ethyl acetate (5 min), 25% ethyl acetate (5 min), 25%–50% ethyl acetate (5 min), 50% ethyl acetate (5 min), 50%–75% ethyl acetate (5 min), 75% ethyl acetate (5 min), 75%–100% ethyl acetate (5 min), change solvent ethyl acetate-chloroform (5 min), 100% chloroform (5 min), 100%–85% chloroform-methanol (5 min), 85%–70% chloroform-methanol (2 min), 70% chloroform-methanol (2 min), 70%–60% chloroform-methanol (2 min), 60% chloroform-methanol (2 min), 60%–50% chloroform-methanol (5 min), 50% chloroform-methanol (5 min), 50%–0% chloroform-methanol (3 min), 100% methanol (10 min); pooled using 1:19 MeOH-CHCl3. From this, the third fraction was collected and labeled Fraction 543W (yield = 37.9 mg).

Selectivity assay compared to human beta-glucuronidases

The selectivity assay was based on the E. coli β-GUS inhibitory activity screening assay, but with modifications. Five microliters of the active fractions and crude extract (prepared at the highest possible concentration in DMSO) and negative control (DMSO) were added to 75 μL HEPES buffer and 10 μL enzyme (10 μg/mL). These were incubated for 15 min; after which, 10 μL 4-MUG (10 mM) were added. These were further incubated for 20 min, and then 100 μL stop buffer was added. The fluorescence intensity was measured using Excitation: 365 nm and Emission: 445 nm. Average percent inhibition of the fraction against the human GUS was calculated using Equation 1. In a parallel experiment, the same fractions were tested against the E. coli B-GUS. The average percent inhibitions against either β-GUS were compared and analyzed using GraphPad PRISM 7 (Student's t-test, α =0.05).

Antibacterial assay against Escherichia coli

Antimicrobial activity screening was done using 96-well plate time-kill assay. E. coli cells were first inoculated into LB broth overnight (37°C, 150 RPM). Afterward, the cells were diluted until it made OD600 = 0.5 and total volume of 200 mL (LB broth). This was then plated in 96 well-plates (190 μL/well). Active fractions were added into the wells at 10 μL each (four replicates, 0.5 mg/mL) to make a total concentration of 25 μg/mL, in addition to the broth only (background), untreated cells, vehicle control (DMSO), and positive control ampicillin (0.1 mg/mL or 5 μg/mL total concentration). The plates were first read for OD600 at time 0 for a baseline measurement. Then, it was incubated at 37°C (150 RPM) and taken out to be read after 10, 30, 60, 120, 300, 420, 600, 720, and 1440 min. A growth curve was plotted for each treatment and then compared against the vehicle control using One-way analysis of variance and Dunnett's post hoc test.

Gas chromatography-mass spectrometry identification of compounds in fractions 5231, 52335A, and 543W

Fractions collected were analyzed at the Metabolomics Core Facility of the Agricultural Biotechnology Research Center, Academia Sinica. To the dried sample, 20 μL of 20 mg/mL methoxyamine was added and then incubated at 30°C for 90 min. Afterward, 100 μL of BSTFA was added and then incubated again at 70°C for 120 min. These derivatized samples (1 μL) were injected in triplicates to an Rtx-5MS (30 m × 0.25 mm × 0.25 μm) column at 1 mL/min flow rate in helium and injector temperature of 250°C. The temperature program was set as follows: 40°C for 1 min, 10°C/min to 310°C, and finally 310°C for 8 min. The ion source temperature was 200°C and detector voltage used was 1800 V. Data were acquired at 10 spectra/s, targeting the mass range of 50–600 m/z. Data processing and acquisition was performed using LECO Chroma TOF software ( Compounds were identified by mass spectra matching against the NIST, LECO/Fiehn, and Wiley Registry 9th Edition MS libraries.

In silico docking analysis of compounds in fractions 5231, 52335A, and 543W in Escherichia coli beta-glucuronidases

From the Protein Databank,[35] the three-dimensional (3D) structure of E. coli β-GUS bound with a novel potent inhibitor (1-((6,8-dimethyl-2-oxo-1,2-dihydroquinolin-3-yl)methyl)-1-(2-hydroxyethyl)-3-(4 hydroxyphenyl) thiourea) (PDB ID: 5czk; resolution = 2.39 Å)[33] was obtained. Using Chimera,[36] the solvents and other ligands were removed, and the inhibitor saved as a separate pdb file. To the protein structure, it was prepared for docking using the “Dock Prep” function of Chimera. Here, the modified residues were corrected (specifically, selenomethionine to methionine), hydrogens and charges were added, protonation states were corrected and the whole structure minimized. After preparing for docking, the E. coli β-GUS protein structure was imported as a receptor for docking into the PyRx platform.[37] Subsequently, the inhibitor (1-((6,8-dimethyl-2-oxo-1,2-dihydroquinolin-3-yl) methyl)-1-(2-hydroxyethyl)-3-(4-hydroxyphenyl)thiourea) was also imported as a ligand for docking protocol validation with AutoDock Vina.[38] The binding region (i.e., gridbox) was selected to cover the region wherein the inhibitor was also bound. After optimizing the gridbox's center (−11.9250, −32.2327, 58.2878) and dimensions (8.5115 Å, 9.8500 Å, 11.9868 Å), the RMSD of the docked ligand (comparison to the experimental pdb file) was calculated and found to be satisfactory (RMSD = 0.444, 28–28 atoms). The analysis of the two-dimensional (2D)-interaction diagram of the redocked ligand shows similar amino acid interactions compared to the experimental pdb file. To analyze the binding affinities of the compounds identified by GC-MS, the compounds were first drawn, then 3D conformations were predicted using PyRx software, which were subsequently imported to be used as AutoDock ligands. The gridbox optimized in the docking validation step was used for the in silico docking analyses. After docking, the binding affinities of each compound (reported as-kcal/mol) as well as docking conformations were exported. 2D- and 3D-interaction diagrams between the top inhibitors and E. coli β-GUS were subsequently generated using Discovery Studio Visualizer.[39]

 Results and Discussion

Plectranthus amboinicus chloroform extract and fractions inhibit Escherichia coli beta-glucuronidases

Prior to fractionation, preliminary experiments were performed wherein different extracts of P. amboinicus – hexane, ethyl acetate, dichloromethane, chloroform, ethanol, and water – were prepared and the IC50 of each compared (data not shown). From these preliminary assays, we found that the chloroform extract showed the highest inhibitory activity with an IC50 of 57.8 μg/mL (43.05–77.84 μg/mL). The chloroform extract was subjected to further bioassay-guided fractionation to yield three fractions with significant increase in bioactivity relative to the crude extract [Table 1].{Table 1}

Natural product compounds as a source of β-GUS inhibitors are unexplored, with only few reported inhibitors. Silymarin components were found to inhibit E. coli β-GUS with the component silybin having IC50 value of 120 μg/mL.[40] Four fractions from Chondria crassicualis crude methanol extract were found to inhibit E. coli β-GUS with IC50 values of 7.26–18.9 μg/mL.[41] Twelve plant species screened by Molan and Saleh Mahdy[42] showed Mentha piperita to maximally inhibit E. coli β-GUS with an IC50 of 140 μg/mL. Clearly, the P. amboinicus fractions isolated in this study showed higher activity than those natural products previously reported.

Plectranthus amboinicus chloroform extract and fractions selectively inhibit Escherichia coli beta-glucuronidases over human beta-glucuronidases

The crude extract and fractions also had their β-GUS inhibitory activity compared against the human, in order to test their selectivity against the human enzyme [Figure 1]. A specific bacterial β-GUS inhibitor should have high activity against the bacterial enzyme without affecting human β-GUS activity.[43] In humans, the role of β-GUS is in degrading glucuronate-containing glycosaminoglycans.[44] Previous reports show increased human β-GUS expression in the tumor microenvironment.[45],[46] Furthermore, cancer cells with increased human β-GUS expression was shown to be more sensitive to SN-38G (CPT-11 therapy).[47],[48] This suggests that nonselective inhibition of human β-GUS by a bacterial β-GUS inhibitor could have detrimental effect on anticancer therapy.{Figure 1}

All of the purified fractions had significantly higher activity against the E. coli enzyme compared to the human β-GUS, with Fraction 5231 having the highest difference in percent inhibition. This result demonstrates the selectivity of the P. amboinicus extract and fractions for E. coli β-GUS, compared to the human β-GUS. The high selectivity of the bacterial β-GUS inhibitor could be due to the structural differences between the E. coli and human β-GUS, wherein a unique bacterial loop present in the bacterial β-GUS (and absent in the human β-GUS) could provide additional interactions with the inhibitor, resulting in tighter binding and inhibition.[49],[50]

Plectranthus amboinicus chloroform extract and fractions are not toxic against gut microbiota (Escherichia coli)

In addition to being active against the E. coli β-GUS and relatively inactive against the human β-GUS, an appropriate bacterial β-GUS inhibitor must also be non-toxic to E. coli cells, i.e., it must have low bactericidal effect. The reason being that bacterial β-GUS inhibitors with bactericidal activities also eliminate gut microbiota, which are essential in the metabolism and processing of carbohydrates, vitamins, bile acids, sterols and xenobiotics.[51],[52] Furthermore, unnecessary intake of bactericidal compounds could do more harm, by increasing risk of infections (by suppression of indigenous microflora) or contributing to antibiotic resistance accumulation.[53],[54]

To demonstrate non-toxicity of the P. amboinicus compounds against E. coli, we did a time-kill assay of the crude extract and fractions, compared against a positive (0.5 μg/mL ampicillin), negative (DMSO) and untreated control [Figure 2]. As expected, the DMSO solvent exhibited a small bactericidal effect, with lower OD600 absorbance values compared to the untreated cells. Thus, the effects of the fractions were compared instead against the vehicle control. At all the time points, the OD600 absorbance values of the positive control was lower than that of the vehicle control (P < 0.05), while none of the P. amboinicus extract nor fractions had any significant bactericidal effect. This suggests that at the concentration of 25 μg/mL, which is at least 2-fold higher than the IC50 values did not exhibit significant bactericidal activity. Thus, the active fractions and isolates purified in the study satisfy the criteria of selectivity (compared to the human β-GUS) and non-toxicity (to E. coli cells).{Figure 2}

Gas chromatography-mass spectrometry identified the compounds present in each Plectranthus amboinicus fraction and in silico docking predicted binding affinity to Escherichia coli β-GUS

GC coupled with MS was performed to determine the compounds present in each P. amboinicus fraction [Table 2], [Table 3], [Table 4]. Fraction 543W contained at least 29 compounds, with 3 compounds exhibiting high abundance (i.e., Phthalic acid, 2-ethylhexyl isohexyl ester, Dodecanoic acid, methyl ester, 1,2-Benzenedicarboxylic acid, dihexyl ester). Fraction 5231 contained at least 35 compounds, with 2 compounds exhibiting high abundance (i.e., 1,2-Benzenedicarboxylic acid, diisooctyl ester and 2-Undecyltetrahydrofuran). Fraction 52335A contained at least 36 compounds, with 4 compounds exhibiting high abundance (i.e., 2-Methyl-4-(2-thienyl) quinoline, 1,2-Benzenedicarboxylic acid, diisooctyl ester, hexadecanoic acid, and Decanedioic acid, bis (2-ethylhexyl) ester). GC-MS analysis showed high relative abundances of volatile components in each P. amboinicus fraction, which were similar to the compounds present in leaf essential oil extracts.[55],[56],[57],[58]{Table 2}{Table 3}{Table 4}

In silico docking using the AutoDock Vina algorithm through the PyRx platform, was able to rank the compounds identified by GC-MS in the P. amboinicus fractions according to their predicted binding affinities against E. coli β-GUS [Table 2], [Table 3], [Table 4]. For the docking simulations, the crystal structure by Wallace et al.[33] (PDB ID: 5czk) was used, particularly due to the presence of an inhibitor. Using the inhibitor as a template, the region surrounding, it was selected as the binding grid for the docking studies. Based on the analyses, the compounds with the highest binding affinities from each fraction were: Phthalic acid, cyclobutyl tridecyl ester (-7.5 kcal/mol) from Fraction 543W, N-Benzyl-2-allyl-2-tosyl-4-penten-1-amine (-8.0 kcal/mol) from Fraction 5231, and Dehydroabietic Acid (-7.9 kcal/mol) from Fraction 52335A. The compounds with high abundance are not the same compounds with the highest predicted binding affinities against E. coli β-GUS although some ranked second third or fourth in the list [Table 2], [Table 3], [Table 4].

In silico analysis shows possible binding modes of predicted Escherichia coli beta-glucuronidases inhibitors from Plectranthus amboinicus

From the top-binding compounds in each fraction-Phthalic acid, cyclobutyl tridecyl ester from Fraction 543W, N-Benzyl-2-allyl-2-tosyl-4-penten-1-amine from Fraction 5231, and Dehydroabietic Acid from Fraction 52335A–the 2D interaction were constructed to show the E. coli β-GUS amino acid residues binding to these compounds [Figure 3]. Roberts et al.[32] previously identified inhibitors that bind to the amino acid residues in the active site cleft of E. coli β-GUS– Asp163, Val446, Phe448, Tyr472, Arg562, the catalytic residue Glu413, and Leu361 from the bacterial loop (residues 360–376)–by protein crystallization. The inhibitor used in the docking validation, (1-((6,8-dimethyl-2-oxo-1,2-dihydroquinolin-3-yl) methyl)-1-(2-hydroxy ethyl)-3-(4-hydroxyphenyl) thiourea) was designed by Wallace et al.[33] to specifically inhibit E. coli β-GUS and was found to interact with the residues Tyr472, Tyr 469, Phe448, Ile363, and Glu413. A computational-guided screening approach by Cheng et al. identified several E. coli β-GUS inhibitors that also interact with active site residues Tyr472 and Tyr469, as well as bacterial loop residues Leu361 and Ile363, among others.[50]{Figure 3}

In our models, we found that the top-binding compounds from each fraction also interact with the active site and bacterial residues. Phthalic acid, cyclobutyl tridecyl ester from Fraction 543W [Figure 3]a, [Figure 3]b formed alkyl interactions with Leu361, Ile363, and Ile560, and Pi-Pi stacked interactions with Tyr 469 and Tyr 472. On the other hand, N-Benzyl-2-allyl-2-tosyl-4-penten-1-amine from Fraction 5231 [Figure 3]c, [Figure 3]d, also formed alkyl interaction with Leu361, contacts with Phe448 and Met447, Pi-alkyl interactions with Ile363, as well as Pi-Pi stacked interactions with Tyr 469 and Tyr 472. Finally, Dehydroabietic Acid from Fraction 52335A [Figure 3]e, [Figure 3]f formed mostly Pi-Pi stacked interactions with Tyr 469 and Tyr 472, and Pi-alkyl interactions with Ile363 and Trp471. These three compounds formed interactions with active site amino acids Tyr 469 and Tyr472, as well as with bacterial loop residues such as Leu361 and/or Ile363. Roberts et al.[32] deduced that active bacterial β-GUS inhibitors could bind effectively to the bacterial β-GUS by interacting with the inner loop unique to bacterial β-GUS.

The assay-guided purification of the P. amboinicus crude chloroform extract yielded fractions with β-GUS inhibitory activity and high selectivity for E. coli β-GUS compared to human β-GUS with no toxic effects on E. coli. The GC-MS profiling of the active fractions identified compounds that are similar to essential oil extracts of P. amboinicus reported in literature. The in silico docking identified the compounds: Phthalic acid, cyclobutyl tridecyl ester, N-Benzyl-2-allyl-2-tosyl-4-penten-1-amine, and Dehydroabietic Acid, to be highly binding with modes suggesting interaction with active site residues Tyr 469 and Tyr 472, as well as with several residues in the β-GUS unique bacterial loop. The GC-MS and in silico modeling enabled elucidation of the β-GUS inhibitory activity of P. amboinicus semi-purified fractions and may provide an alternative approach to support drug discovery initiative.


In this study, we identified several highly-active and highly-selective E. coli β-GUS inhibitors from the partially-purified fractions of the P. amboinicus leaf chloroform extracts, utilizing a bioassay-guided purification coupled by metabolomics and in silico docking approach. Furthermore, these compounds were shown to be highly selective against E. coli β-GUS compared to human β-GUS. Using antibacterial assays, we have shown that these are non-toxic to E. coli bacteria, preserving gut microbiota. These compounds were identified using GC-MS metabolomics, and subsequently modelled for binding affinities against E. coli β-GUS to study their interactions with active site residues. This is the first report on the potential of P. amboinicus as selective inhibitor of E. coli β-GUSs.


The authors would like to acknowledge the Metabolomics Core Facility of the Agricultural Biotechnology Research Center (ABRC), Academia Sinica, Taipei, Taiwan, for the use of their GC-MS facility in identifying the compounds. The authors would like to thank Nitish Verma, Punsaldulam Dashnyam, and Nadia Morales for their valuable input during experimentation and data collection.

Financial support and sponsorship

This project was supported by the Philippine Department of Science and Technology's ASTHRDP research grant.

Conflicts of interest

There are no conflicts of interest.


1Arumugam G, Swamy MK, Sinniah UR. Plectranthus amboinicus (Lour.) Spreng: Botanical, phytochemical, pharmacological and nutritional significance. Molecules 2016;21:369.
2Quisumbing EA. Medicinal Plants of the Philippines. Quezon City, Philippines: Katha Publishing Co.; 1978.
3Aguiar JJ, Sousa CP, Araruna MK, Silva MK, Portelo AC, Lopes JC, et al. Antibacterial and modifying-antibiotic activities of the essential oils of Ocimum gratissimum L. and Plectranthus amboinicus L. Eur J Integr Med 2015;7:151-6.
4Akagawa M, Shigemitsu T, Suyama K. Production of hydrogen peroxide by polyphenols and polyphenol-rich beverages under quasi-physiological conditions. Biosci Biotechnol Biochem 2003;67:2632-40.
5Frame AD, Ríos-Olivares E, De Jesús L, Ortiz D, Pagán J, Méndez S. Plants from Puerto Rico with anti-Mycobacterium tuberculosis properties. PR Health Sci J 1998;17:243-52.
6Ikigai H, Nakae T, Hara Y, Shimamura T. Bactericidal catechins damage the lipid bilayer. Biochim Biophys Acta 1993;1147:132-6.
7Shubha JR, Bhatt P. Plectranthus amboinicus leaves stimulate growth of probiotic L. plantarum: Evidence for ethnobotanical use in diarrhea. J Ethnopharmacol 2015;166:220-7.
8Vijayakumar S, Vinoj G, Malaikozhundan B, Shanthi S, Vaseeharan B. Plectranthus amboinicus leaf extract mediated synthesis of zinc oxide nanoparticles and its control of methicillin resistant Staphylococcus aureus biofilm and blood sucking mosquito larvae. Spectrochim Acta A Mol Biomol Spectrosc 2015;137:886-91.
9Murthy PS, Ramalakshmi K, Srinivas P. Fungitoxic activity of Indian borage (Plectranthus amboinicus) volatiles. Food Chem 2009;114:1014-8.
10Hattori M, Nakabayashi T, Lim YA, Miyashiro H, Kurokawa M, Shiraki K, et al. Inhibitory effects of various ayurvedic and Panamanian medicinal plants on the infection of herpes simplex virus-1 in vitro and in vivo. Phytother Res 1995;9:270-6.
11Kusumoto IT, Nakabayashi T, Kida H, Miyashiro H, Hattori M, Namba T, et al. Screening of various plant extracts used in ayurvedic medicine for inhibitory effects on human immunodeficiency virus type 1 (HIV-1) protease. Phytother Res 1995;9:180-4.
12Gurgel AP, da Silva JG, Grangeiro AR, Oliveira DC, Lima CM, da Silva AC, et al. In vivo study of the anti-inflammatory and antitumor activities of leaves from Plectranthus amboinicus (Lour.) Spreng (Lamiaceae). J Ethnopharmacol 2009;125:361-3.
13Chen YS, Yu HM, Shie JJ, Cheng TJ, Wu CY, Fang JM, et al. Chemical constituents of Plectranthus amboinicus and the synthetic analogs possessing anti-inflammatory activity. Bioorg Med Chem 2014;22:1766-72.
14Chiu YJ, Huang TH, Chiu CS, Lu TC, Chen YW, Peng WH, et al. Analgesic and Antiinflammatory Activities of the Aqueous Extract from Plectranthus amboinicus (Lour.) Spreng. Both in vitro and in vivo. Evid Based Complement Alternat Med 2012;2012:508137.
15Kumaran A, Joel Karunakaran R. Antioxidant and free radical scavenging activity of an aqueous extract of Coleus aromaticus. Food Chem 2006;97:109-14.
16Gurib-Fakim A, Sewraj MD, Narod F, Menut C. Aromatic plants of Mauritius: Volatile constituents of the essential oils of Coleus aromaticus Benth., Triphasia trifolia (Burm.f.) and Eucalyptus kitsoniana F. Muell. J Essent Oil Res 1995;7:215-8.
17Morton JF. Country borage (Coleus amboinicus Lour.). J Herbs Spices Med Plants 1992;1:77-90.
18Ong HC, Nordiana M. Malay ethno-medico botany in Machang, Kelantan, Malaysia. Fitoterapia 1999;70:502-13.
19Henrissat B. A classification of glycosyl hydrolases based on amino acid sequence similarities. Biochem J 1991;280 (Pt 2):309-16.
20Sinnott M. Comprehensive Biological Catalysis: A Mechanistic Reference. San Diego: Academic Press; 1998.
21Wallace BD, Wang H, Lane KT, Scott JE, Orans J, Koo JS, et al. Alleviating cancer drug toxicity by inhibiting a bacterial enzyme. Science 2010;330:831-5.
22Pellock SJ, Redinbo MR. Glucuronides in the gut: Sugar-driven symbioses between microbe and host. J Biol Chem 2017;292:8569-76.
23Fittkau M, Voigt W, Holzhausen HJ, Schmoll HJ. Saccharic acid 1.4-lactone protects against CPT-11-induced mucosa damage in rats. J Cancer Res Clin Oncol 2004;130:388-94.
24Stein A, Voigt W, Jordan K. Chemotherapy-induced diarrhea: Pathophysiology, frequency and guideline-based management. Ther Adv Med Oncol 2010;2:51-63.
25Takasuna K, Hagiwara T, Hirohashi M, Kato M, Nomura M, Nagai E, et al. Involvement of β-glucuronidase in intestinal microflora in the intestinal toxicity of the antitumor camptothecin derivative irinotecan hydrochloride (CPT-11) in rats. Cancer Res 1996;56:3752-7.
26LoGuidice A, Wallace BD, Bendel L, Redinbo MR, Boelsterli UA. Pharmacologic targeting of bacterial β-glucuronidase alleviates nonsteroidal anti-inflammatory drug-induced enteropathy in mice. J Pharmacol Exp Ther 2012;341:447-54.
27Zhong ZY, Sun BB, Shu N, Xie QS, Tang XG, Ling ZL, et al. Ciprofloxacin blocked enterohepatic circulation of diclofenac and alleviated NSAID-induced enteropathy in rats partly by inhibiting intestinal β-glucuronidase activity. Acta Pharmacol Sin 2016;37:1002-12.
28Dietrich CG, Ottenhoff R, De Waart DR, Oude-Elferink RP. Lack of UGT1 isoforms in Gunn rats changes metabolic ratio and facilitates excretion of the food-derived carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b] pyridine. Toxicol Appl Pharmacol 2001;170:137-43.
29Humblot C, Murkovic M, Rigottier-Gois L, Bensaada M, Bouclet A, Andrieux C, et al. β-glucuronidase in human intestinal microbiota is necessary for the colonic genotoxicity of the food-borne carcinogen 2-amino-3-methylimidazo[4,5-f]quinoline in rats. Carcinogenesis 2007;28:2419-25.
30Bhatt AP, Pellock SJ, Biernat KA, Walton WG, Wallace BD, Creekmore BC, et al. Targeted inhibition of gut bacterial β-glucuronidase activity enhances anticancer drug efficacy. Proc Natl Acad Sci U S A 2020;117:7374-81.
31Kehrer DF, Sparreboom A, Verweij J, de Bruijn P, Nierop CA, van de Schraaf J, et al. Modulation of irinotecan-induced diarrhea by cotreatment with neomycin in cancer patients. Clin Cancer Res 2001;7:1136-41.
32Roberts AB, Wallace BD, Venkatesh MK, Mani S, Redinbo MR. Molecular insights into microbial β-glucuronidase inhibition to abrogate CPT-11 toxicity. Mol Pharmacol 2013;84:208-17.
33Wallace BD, Roberts AB, Pollet RM, Ingle JD, Biernat KA, Pellock SJ, et al. Structure and inhibition of microbiome β-glucuronidases essential to the alleviation of cancer drug toxicity. Chem Biol 2015;22:1238-49.
34Dashnyam P, Mudududdla R, Hsieh TJ, Lin TC, Lin HY, Chen PY, et al. β-Glucuronidases of opportunistic bacteria are the major contributors to xenobiotic-induced toxicity in the gut. Sci Rep 2018;8:16372.
35Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The protein data bank. Nucleic Acids Res 2000;28:235-42.
36Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, et al. UCSF Chimera – A visualization system for exploratory research and analysis. J Comput Chem 2004;25:1605-12.
37Dallakyan S, Olson AJ. Small-molecule library screening by docking with PyRx. Methods Mol Biol 2015;1263:243-50.
38Trott O, Olson AJ. AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem 2010;31:455-61.
39BIOVIA. Discovery Studio 2019. San Diego: Dassault Systèmes; 2019.
40Kim DH, Jin YH, Park JB, Kobashi K. Silymarin and its components are inhibitors of β-glucuronidase. Biol Pharm Bull 1994;17:443-5.
41Sekikawa C, Kurihara H, Goto K, Takahashi K. Inhibition of β-glucuronidase by extracts of Chondria crassicaulis. Bull Fish Sci Hokkaido Univ 2002;53:27-30.
42Molan AL, Saleh Mahdy A. Iraqi medicinal plants: Total flavonoid contents, free-radical scavenging and bacterial beta-glucuronidase inhibition activities. IOSR JDMS 2014;13:72-7.
43Cheng KW, Tseng CH, Tzeng CC, Leu YL, Cheng TC, Wang JY, et al. Pharmacological inhibition of bacterial B-glucuronidase prevents irinotecan-induced diarrhea without impairing its antitumor efficacy in vivo. Pharmacol Res 2019;139:41-9.
44Naz H, Islam A, Waheed A, Sly WS, Ahmad F, Hassan I. Human β-glucuronidase: Structure, function, and application in enzyme replacement therapy. Rejuvenation Res 2013;16:352-63.
45Feng S, Song JD. Determination of β-glucuronidase in human colorectal carcinoma cell lines. World J Gastroenterol 1997;3:251-2.
46Sperker B, Werner U, Mürdter TE, Tekkaya C, Fritz P, Wacke R, et al. Expression and function of beta-glucuronidase in pancreatic cancer: Potential role in drug targeting. Naunyn Schmiedebergs Arch Pharmacol 2000;362:110-5.
47Huang PT, Chen KC, Prijovich ZM, Cheng TL, Leu YL, Roffler SR. Enhancement of CPT-11 antitumor activity by adenovirus-mediated expression of β–glucuronidase in tumors. Cancer Gene Ther 2011;18:381-9.
48Prijovich ZM, Chen KC, Roffler SR. Local enzymatic hydrolysis of an endogenously generated metabolite can enhance CPT-11 anticancer efficacy. Mol Cancer Ther 2009;8:940-6.
49Cheng KW, Tseng CH, Yang CN, Tzeng CC, Cheng TC, Leu YL, et al. Specific inhibition of bacterial β-glucuronidase by pyrazolo[4,3-c]quinoline derivatives via a pH-dependent manner to suppress chemotherapy-induced intestinal toxicity. J Med Chem 2017;60:9222-38.
50Cheng TC, Chuang KH, Roffler SR, Cheng KW, Leu YL, Chuang CH, et al. Discovery of specific inhibitors for intestinal E. coli β-glucuronidase through in silico virtual screening. ScientificWorldJournal 2015;2015:740815.
51Flieger D, Klassert C, Hainke S, Keller R, Kleinschmidt R, Fischbach W. Phase II clinical trial for prevention of delayed diarrhea with cholestyramine/levofloxacin in the second-line treatment with irinotecan biweekly in patients with metastatic colorectal carcinoma. Oncology 2007;72:10-6.
52Guarner F, Malagelada JR. Gut flora in health and disease. Lancet 2003;361:512-9.
53Levy SB, Marshall B. Antibacterial resistance worldwide: Causes, challenges and responses. Nat Med 2004;10:S122-9.
54Nord CE, Kager L, Heimdahl A. Impact of antimicrobial agents on the gastrointestinal microflora and the risk of infections. Am J Med 1984;76:99-106.
55Asiimwe S, Borg-Karlsson AK, Azeem M, Maud Mugisha K, Namutebi A, James Gakunga N. Chemical composition and toxicological evaluation of the aqueous leaf extracts of Plectranthus amboinicus Lour. Spreng. Int J Pharm Sci Invent 2014;3:19-27.
56Kweka EJ, Senthilkumar A, Venkatesalu V. Toxicity of essential oil from Indian borage on the larvae of the African malaria vector mosquito, Anopheles gambiae. Parasit Vectors 2012;5:277.
57Prudent D, Perineau F, Bessiere JM, Michel GM, Baccou JC. Analysis of the essential oil of wild oregano from Martinique (Coleus aromaticus Benth.) – evaluation of its bacteriostatic and fungistatic properties. J Essent Oil Res 1995;7:165-73.
58Senthilkumar A, Venkatesalu V. Chemical composition and larvicidal activity of the essential oil of Plectranthus amboinicus (Lour.) Spreng against Anopheles stephensi: A malarial vector mosquito. Parasitol Res 2010;107:1275-8.