Anti-Inflammatory and Cancer-Preventive Potential of Chamomile (Matricaria chamomilla L.): A Comprehensive In Silico and In Vitro Study

Abstract

Background and aim: Chamomile tea, renowned for its exquisite taste, has been appreciated for centuries not only for its flavor but also for its myriad health benefits. In this study, we investigated the preventive potential of chamomile (Matricaria chamomilla L.) towards cancer by focusing on its anti-inflammatory activity.

Methods and results: A virtual drug screening of 212 phytochemicals from chamomile revealed β-amyrin, β-eudesmol, β-sitosterol, apigenin, daucosterol, and myricetin as potent NF-κB inhibitors. The in silico results were verified through microscale thermophoresis, reporter cell line experiments, and flow cytometric determination of reactive oxygen species and mitochondrial membrane potential. An oncobiogram generated through comparison of 91 anticancer agents with known modes of action using the NCI tumor cell line panel revealed significant relationships of cytotoxic chamomile compounds, lupeol, and quercetin to microtubule inhibitors. This hypothesis was verified by confocal microscopy using α-tubulin-GFP-transfected U2OS cells and molecular docking of lupeol and quercetin to tubulins. Both compounds induced G2/M cell cycle arrest and necrosis rather than apoptosis. Interestingly, lupeol and quercetin were not involved in major mechanisms of resistance to established anticancer drugs (ABC transporters, TP53, or EGFR). Performing hierarchical cluster analyses of proteomic expression data of the NCI cell line panel identified two sets of 40 proteins determining sensitivity and resistance to lupeol and quercetin, further pointing to the multi-specific nature of chamomile compounds. Furthermore, lupeol, quercetin, and β-amyrin inhibited the mRNA expression of the proinflammatory cytokines IL-1β and IL6 in NF-κB reporter cells (HEK-Blue Null1). Moreover, Kaplan-Meier-based survival analyses with NF-κB as the target protein of these compounds were performed by mining the TCGA-based KM-Plotter repository with 7489 cancer patients. Renal clear cell carcinomas (grade 3, low mutational rate, low neoantigen load) were significantly associated with shorter survival of patients, indicating that these subgroups of tumors might benefit from NF-κB inhibition by chamomile compounds.

Conclusion: This study revealed the potential of chamomile, positioning it as a promising preventive agent against inflammation and cancer. Further research and clinical studies are recommended.

Keywords: Kaplan–Meier survival analysis; anti-inflammatory; carcinogenesis; cytokines; flavonoids; microscale thermophoresis; natural products; prevention; proteomics.

Figure 1

In silico binding of selected phytochemicals extracted from chamomile (Matricaria chamomilla) and triptolide (positive control) to NF-κB. Molecular docking analyses have been performed with NF-κB -RelA (PDB ID: 1NFI). (A) The lowest binding energies (LBE, kcal/mol) of the top 28/212 compounds (=10.4%) significantly correlated with the predicted inhibition constants (pKi, µM) (p = 2.61 × 10⁻⁶; r = 0.76). (B) The top 6/212 compounds were bound to different pockets within two domains. The interactions of these six compounds with the amino acids of NF-κB are displayed as 2D and 3D figures: (C) β-amyrin, (D) β-sitosterol, (E) myricetin, (F) daucosterol, (G) β-eudesmol, (H) apigenin, and (I) triptolide (positive control).

Figure 2

In vitro binding to and inhibition of NF-κB for selected phytochemicals extracted from chamomile (Matricaria chamomilla) and triptolide (positive control). (A) Binding of β-amyrin to NF-κB as determined through microscale thermophoresis (MST). The resulting binding kinetics is shown as normalized fluorescence (LED power: 40%; MST power: 10%). (B) Inhibition of NF-κB activity using an NF-κB reporter assay. The percentages of the NF-κB rest activity are shown after 24 h treatment with β-amyrin, β-sitosterol, β-eudesmol, daucosterol, myricetin, and apigenin at concentrations of 0.1 µM, 1 µM, and 10 µM followed by 100 ng/mL of TNF-α for 24 h (* p < 0.05). (C,D) Pearson correlation of NF-κB rest activity (%) vs. lowest binding energy (kcal/mol) of the six selected compounds at concentrations of (C) 0.1 µM and (D) 1 µM. The correlation using a concentration of 10 µM was statistically not significant. The mean values ± SD of three independent experiments are shown. (* p < 0.05, ** p < 0.01, *** p < 0.001).

Figure 3

Effect of β-amyrin on the generation of reactive oxygen species (ROS) in HEK-Blue Null 1 (HBN1). The cells were treated with 100 µM of H₂O₂ (15 min) and 100 ng/mL of TNF-α (3 h) with and without β-amyrin at concentrations of 0.1 µM, 1 mM, and 10 µM (24 h). The statistical analysis was performed by using the paired student’s t-test. * p = 0.05 (1 µM) and ** p = 0.01 (10 µM) compared with TNF-α- and H₂O₂-treated control cells, ^# p = 0.02 (cells treated with H₂O₂), and ^## p = 0.005 (cells treated with TNF-α and H₂O₂) compared with untreated cells. β-amyrin significantly reduced ROS generation. DMSO treatment served as the solvent control. The mean values ± SD of three independent experiments are shown.

Figure 4

Flow cytometric determination of mitochondrial membrane potential (MMP) in HEK-Blue Null 1 Cells through JC–1 staining. Cells were left untreated (control) or treated with 10 µM of β-amyrin or vinblastine for 24 h, followed by 100 ng/mL of TNF-α for 3 h. (A) Representative histograms; (B) statistical analysis cells with disrupted MMP (dead cells) or intact MMP (healthy cells). Mean values ± SD of three independent experiments are shown. The results are significant at ** p < 0.01 for death cells and * p < 0.05 for living cells if compared to the control DMSO untreated cells (paired two-tailed t-test).

Figure 5

Cytotoxicity and oncobiogram analyses. Chemical structures of (A) lupeol and (B) quercetin. (C) Cytotoxicity of six selected phytochemicals from chamomile to the NCI tumor cell line panel plotted as a mean log₁₀IC₅₀ for each tumor type. (D) The cytotoxicity of lupeol and quercetin in cell lines of different tumor types compared with the established anticancer drug chlorambucil as the positive control. (E) Cross-resistance profiling of lupeol and quercetin to 91 standard drugs with known modes of action against tumor cells. (F) Oncobiogram for lupeol and quercetin. The correlation coefficients for lupeol and quercetin to 10 known tubulin inhibitors are plotted.

Figure 6

Confocal immunofluorescence microscopy of the microtubule network in U2OS cells upon treatment with (A) quercetin and (B) lupeol at concentrations of 0.1 µM, 1 µM, and 10 µM for 24 h. Vincristine (1 µM) and paclitaxel (1 µM) served as positive controls and DMSO as the negative control. The cells were imaged using a Thunder Imager Live Cell microscope with a 63×/1.40 NA objective lens (HC PL APO CS2 63×/1.40 OIL UV). The microtubules were visualized using green fluorescence for GFP (green), and the images were merged with DAPI (blue) to highlight the nucleus.

Figure 7

Molecular docking analysis of lupeol and quercetin to tubulin (5N5N). (A) Illustrates the binding sites of vincristine, paclitaxel, and colchicine to α- and β-tubulin. On the right, a zoomed-in view shows lupeol and quercetin binding to the same pocket as vincristine. (B) The correlation of the predicted inhibition constants (pKi, mM) vs. the lowest binding energies (LBE, kcal/mol) (p = 0.004, r = 0.85). (C,D) Presents 3D and 2D illustrations of the interaction of lupeol, quercetin, and vincristine with the amino acids of α-tubulin docked at vincristine’s binding site. (C) Vincristine, (D) quercetin, and (E) lupeol.

Figure 8

Cell cycle arrest of U2OS cells by quercetin and lupeol. (A) Debris was gated out (SSC-A vs. FSC-A) with the first gate. (B) With the second gate (FSC-H vs. FSC-A), only single cells of normal morphology were gated. Duplets were gated out. (C,D) Three-dimensional representation of DNA histograms of U2OS cells exposed to 1 × IC₅₀ and 4 × IC₅₀ quercetin and lupeol for 72 h. DMSO was used as the negative control, and 1 × IC₅₀ vincristine was used as the positive control. (C) Cells treated with lupeol and (D) cells treated with quercetin. The histograms were obtained through flow cytometry using an excitation of 488 nm and an emission wavelength of 530 nm. (E,F) Bar diagrams showing the distinct phases of cell cycle upon treatment with quercetin and lupeol for 72 h. (E) Cells treated with lupeol and (F) cells treated with quercetin. The bar diagrams were created through the calculation of the mean values ± SD of three independent experiments. *** p < 0.001, ** p < 0.01, and * p < 0.05 compared to the negative control using paired two-tailed t-test.

Figure 9

Dose–response curves of quercetin and lupeol as determined through resazurin assay. The mean values and standard deviation values are from three independent experiments. The tumor cells were subjected to treatment with each compound at concentrations of 10 µM, 25 µM, 50 µM, and 100 µM for 72 h. (A) Sensitive CCRF-CEM and the drug-resistant P-glycoprotein overexpressing CEM/ADR5000 leukemia cells. (B) U87/ΔEGFR transfected with a deletion-activated cDNA of EGFR and its wild-type U87MG glioblastoma cells. (C) HCT116 p53^+/+ and knockout HCT116 p53^−/− colorectal cancer cells. (D) U2OS osteosarcoma cells.

Figure 10

Detection of cell death in CCRF-CEM cells using flow cytometry and annexin-V/PI staining to measure apoptosis using a flow cytometer. (A,B) Cells treated with 0.25 × IC₅₀, 0.5 × IC₅₀, 1 × IC₅₀, 2 × IC₅₀, and 4 × IC₅₀ of quercetin and lupeol, for 72 h. DMSO was used as negative control. (A) Cells treated with lupeol and (B) cells treated with quercetin. Q1 represents necrotic cells (−) annexin V/(+) PI; Q2 represents late apoptotic cells exhibiting annexin V (+)/PI (+); Q3 represents early apoptotic cells (+) annexin V/(−) PI; Q4 represents viable cells (−) annexin V/(−) PI. (C,D) Bar diagrams representing the percentages of cells in the different quadrants. (C) Effects of lupeol and (D) Effects of quercetin. The treatment of both compounds at increasing concentrations significantly enhanced the percentage of necrotic cells. *** p < 0.001, ** p < 0.01, and * p < 0.05 compared to the negative control using paired two-tailed t-test. The bar diagrams were created based on the calculation of the mean values ± SD of three independent experiments.

Figure 11

Expression of NF-κB and downstream genes. (A) Western blotting analysis of NF-κB in HEK-Blue Null1 cells treated with 10 µM or 50 µM of β-amyrin, quercetin, and lupeol for 24 h, followed by 24 h of TNF-α at 100 ng/mol. (B) The percentages of NF-κB expression in the cell. (C) qRT-PCR analysis of IL-1β and IL-6 gene expression in HEK-Blue Null 1 cells treated with 10 µM or 50 µM of β-amyrin, quercetin, and lupeol for 24 h with TNF-α 100 ng/mL for another 24 h. The statistical analysis was performed through the paired one-tailed t-test (** p ≤ 0.01) (* p ≤ 0.05) from three independent trials.

Figure 12

Cluster analysis in a 2D colored heat map of proteins’ expression in NCI tumor cell lines responding to lupeol. Clusters A and B represent the top 40 proteins, and clusters 1–3 show the tumor cell lines. The cell lines were clustered according to their degrees of relatedness to each other based on their protein expression included in the analysis. Color code: red, 0–25% quartile; orange, 26–50% quartile; grey, median value; light green, 50–75% quartile; and dark green, 76–100% quartile. Depending on individual log₁₀IC₅₀ values, the responsiveness of the cell lines to lupeol was classified as sensitive if their log₁₀IC₅₀ values were lower than the median value of all cell lines (marked in yellow) and as resistant if their log₁₀IC₅₀ values were higher than the median value (marked in red). The χ2 test shows a statistical significance (p = 1.44 × 10⁻⁴) upon comparing the two clusters of protein expression in the cell lines, where clusters 1 and 2 contained mainly resistant cell lines to both lupeol and quercetin, respectively, and cluster 3 contained mainly sensitive cell lines.

Figure 13

Cluster analysis in a 2D colored heat map of proteins’ expression in NCI tumor cell lines responding to quercetin. The χ2 test shows a statistical significance (p = 0.044). For further details, see Figure 12.

Figure 14

Kaplan–Meier analysis from the KM-Plotter database of overall survival time (months) for renal clear cell carcinoma cells correlating with the expression of NFKB2 mRNA. (A–F) Different patient profiles where NF-κB expression correlated significantly with the overall survival time.