Human cervical cancer (HeLa) cells were used as an in vitro cancer model during this study. HeLa cells were maintained in complete Dulbecco’s modified Eagle medium (DMEM; Sigma-Aldrich; Merck, KGaA) containing 10% FBS (Sigma-Aldrich; Merck KGaA) and penicillin (100U/mL) (Sigma-Aldrich; Merck KGaA), and incubated at 37 °C with 5% CO₂.

Chrysin (powdered, mol wt. 254.241 g/mol) was procured from Sigma-Aldrich (Merck, KGaA), and stock solution was prepared with DMSO (78.67 mM) using DMSO (stock solution). Furthermore, sub-stock (1 mM) and concentrations (5, 10, and 15 µM) of chrysin were prepared using the complete media as a diluent.

The colony-forming assay was performed following the protocol used by Crowley et al. (2016) and Sundaram et al. (2019), with minor modifications (Crowley et al., 2016; Kedhari Sundaram et al., 2019a). Briefly, ∼2.5 × 10⁵ (Ho et al., 2013) cells were dispensed in six-well plates and incubated overnight followed by the treatment with increasing doses of chrysin (5, 10, and 15 µM) for 48 h. DMSO control (i.e., solvent control) and treated cells were collected after 48 h and plated at approximately 500 cells/well, and allowed to grow. The medium was changed time to time as per the requirement. After 14 days, the colonies formed were fixed by using absolute CH₃OH and stained with crystal violet. Olympus inverted microscope (Labomed, United States) was used to obtain the images of the colonies formed. ImageJ software program was used to count the colonies in treated and DMSO control wells.

The scratch-aound assay was performed to examine the inhibitory effect of chrysin on cell migration (Yang et al., 2014; Kedhari Sundaram et al., 2019a). Approximately 2 × 10⁵ (Ho et al., 2013) cells were plated in a six-well plate and incubated at 37°C overnight. On the next day, a “wound” or a “cell-free” line was created on a confluent monolayer of the cells by scratching the monolayer with a pipette tip. The cells were incubated in the presence of different chrysin dilutions (10 and 15 µM) and “healing” of wounds, which ensues through cell migration, and growth toward the cell-free zone was monitored on a regular basis. An inverted microscope was used for capturing the wound images in each well prior and after 24–48 h of the treatment.

The invasion assay was performed to evaluate the migratory and invasive capability of chrysin-treated HeLa cells and DMSO control using the Boyden chamber (Yang et al., 2014). Briefly, 5.0 × 10³ (Aggarwal et al., 2015) cells/well were seeded on the upper side of the insert in separate inserts, and below it in the well of a 24-well plate, the medium with FBS was kept. After 48 h, absolute methanol was used to fix the cells, and 0.1% crystal violet was used for staining purposes. The cells present inside of the chamber were cleared using an ear bud. The assessment of complete migration was performed under the microscope, and any of the five fields were scanned (five fields per filter). The images were captured for each treatment and control using ×200 magnification with an inverted microscope (Olympus Corporation). ImageJ program was used for counting the colonies. The experiment was repeated three times, and mean ± SD was used to plot the graphs considering p-value ≤0.05.

Nuclear extracts from the untreated HeLa cells were prepared using the EpiQuik^TM Nuclear Extraction Kit (Catalog No. OP-0002, Epigentek, United States) following the manufacturer’s protocol. The EpiQuik DNMT Activity Assay Kit (Catalog No. #P-3009, Epigentek, USA) was used to check the effect of chrysin on the DNMT activity. Briefly, chrysin (@ conc. 5, 10 and 15 µM) was added to the untreated nuclear extract, buffer, and Adomet (methyl group donor) to the cytosine-rich DNA substrate–coated assay plate and incubated for 1.5 h at 37 °C. It was followed by incubation with capture and detection antibody. After signal development, the absorbance was read on an ELISA reader at 450 nm. DNMT activity inhibition was calculated by comparing with DMSO controls. The experiment was repeated three times, and the mean ± SD was used to plot the graph. The statistical significance level was calculated using one-way ANOVA, and the p-value was maintained at ≤ 0.05.

The EpiQuik HDAC Activity Assay Kit (Catalog Number P-4002, Epigentek, United States) was used for evaluating the effect of chrysin on the HDAC activity. The acetylated histone substrate–coated assay plate was prepared by adding 50 μL of the biotinylated HDAC substrate diluted in wash buffer to all wells. The assay plate was incubated at room temperature for 45 min and washed with wash buffer. HDAC assay buffer was dispensed to the wells in chrysin (5, 10, and 15 µM) treated and untreated nuclear extract placed. The plate was kept at 37°C for an hour followed by incubation with capture and detection antibody. Developing solution was used to develop the signal, stop solution stopped the reaction, and OD was read at 450 nm. The percentage inhibition of the HDAC activity against chrysin treatment was calculated by comparing with DMSO control. The experiment was repeated three times, and mean ± SD was used to plot a graph. One-way ANOVA was used to check the statistical significance at the p-value ≤ 0.05.

The assay was done by using the EpiQuik HMT H3K9 Activity Assay Kit (Catalog No. P-3003, Epigentek, USA) following the protocol given by the manufacturer. Briefly, to the histone three lysine substrate–coated assay plate, chrysin (@ conc. of 5, 10, and 15 µM) was added in separate wells, with untreated nuclear extract, buffer biotinylated substrate, and Adomet (methyl group donor), and incubated at 37 °C for 1.5 h. Following this step, capture antibody and detection antibody were added to the wells, and incubated at room temp for 30 min. Finally, a color was developed, and absorbance was measured using an ELISA reader at 450 nm. The percentage inhibition was calculated by comparing with DMSO control. The experiment was repeated three times, and the mean ± SD was calculated to plot a graph. One-way ANOVA was used for checking the statistical significance, and the p-value was maintained as ≤ 0.05.

The assay was done using the EpiQuik™ HAT Activity Assay Kit (Catalog No. P-4003 Epigenetek, United States) following the manufacturer’s protocol. Briefly, the nuclear extract and the substrate for HAT were incubated for 1 h followed by washing with the wash buffer. An inhibitor was added in the test samples, and signals were captured and detected by the capture and detection antibody, respectively. After color development, the plate was read at 450 nm using the ELISA plate reader. HAT activity and percentage inhibition were calculated by comparing with DMSO control samples. A graph was plotted by taking the mean of three experiments ± SD. One-way ANOVA was used to check the statistical significance at fixed p-value ≤ 0.05.

DNA isolation was done by using GenElute Mammalian Genomic DNA Miniprep Kit (Catalog No. G1N70, Sigma-Aldrich, Merck, KGaA) following the manufacturer’s protocol. About 1.5 × 10⁶ cells were treated with chrysin (@ conc. 5, 10, and 15 µM for 48 h) and DMSO control. DNA was isolated from chrysin treated and the DMSO control samples, and its quality was checked by gel electrophoresis using 1% agarose gel (Catalog No. A9539, Sigma-Aldrich, Merck, KGaA) in 0.5XTBE buffer with ethidium bromide. The quantitation of DNA samples was completed by spectrophotometry using NanoDrop 2000 (Thermo Scientific™, USA) and stored at-80°C.

MethylFlash™ Methylated DNA Quantification Kit (Catalog No. P-1034, Epigentek, USA) was used to analyze the methylated DNA in treated cells (@ chrysin conc. 5, 10, and 15 µm for 48 h) and DMSO controls. The kit was used for the detection of methylated DNA using antibodies against 5-mC (cytosine) that can be analyzed calorimetrically. Optical density was measured using the ELISA reader at 450 nm wavelength. As established, the extent of methylation on the gene is directly related to the optical density. The levels of methylation were calculated in comparison with the DMSO control. The experiment was performed three times at a significance level of p-value ≤0.05.

CpG island DNA methylation quantification was conducted for estimating the percentage of methylated DNA in the total DNA content of HeLa cells by using the EpiTect Methyl II PCR System (Catalog No. 335452, Qiagen, United States). The method employed calculates the methylation of promoter regions in the input DNA after cleavage with methylation-dependent restriction enzymes and methylation-sensitive restriction enzymes that digest methylated and unmethylated DNA, respectively. Following restriction digestion, the cleaved DNA from each reaction was computed by using it as a template for Human Tumor Suppressor Genes EpiTect Methyl II Signature PCR Array (Qiagen, United States) real-time PCR in an assay plate with primers that border the promoter region of the anticipated genes. The relative amounts of unmethylated and methylated DNA were calculated by comparing the amounts of each reaction with that of a control (no enzymes added) reaction using the ^∆∆CT method. The gene panel (with predesigned primers) consisted of tumor-suppressor genes comprising TP73, MGMT, APC, CDKN2A, BRCA1, PTEN, CDH1, DAPK1, CDH13, SOC51, RARB, ESR1, FHIT, RASSF1, WIF1, GSTP1, RUNX3, MLH1, NEUROG1, VHL, PDLIM4, and TIMP3.

The amount of DNA left after the restriction digestion was calculated by using qPCR array results. This was used for constructing the methylation profile of each gene with the ^∆ΔCT method. The methylation and unmethylation fraction of the promoter of tested tumor-suppressor genes in chrysin-treated and untreated HeLa cells was estimated as per the protocol available with the kit. The levels of methylation were presented in the form of a graph. Statistical significance was calculated by taking the mean of three experiments by one-way ANOVA using the SPSS program with p-value ≤ 0.05.

The RNA from chrysin-treated (conc. 10 and 15 µM for 48 h) and DMSO control HeLa cells were extracted by using the GenElute Mammalian Genomic Total RNA Kit (Catalog No. RTN70 Sigma-Aldrich, Merck KGaA) and further quantified with the help of NanoDrop. The RNA (2 μg was used as a template) was then subjected to cDNA synthesis by using Applied Biosystems™ High-Capacity cDNA Reverse Transcription Kit (Catalog No. 4368814, ABI-Thermo Fisher, United States). This kit supports random primers’ scheme for initiating the synthesis of cDNA. The expression of genes related to various pathways of migration, inflammation, and TSGs was analyzed with the help of TaqMan-based custom array (4391524 and 4369514 master mix). The PCR array was run on QuantStudio3 and analyzed with the ^ΔΔ C _T method using the DataAssist™ program (Thermo Fisher, United States). GAPDH (housekeeping gene) was used for normalizing the data. Relative expression was calculated in comparison with the DMSO control. The statistical significance was calculated by maintaining p-value <0.05.

The expression analysis of TSGs, migration, and inflammation-related proteins was performed by the Proteome Profiler Array (Catalog No. ARY026, R&D, USA). The relative expression levels of 84 oncogenes were investigated with the help of this array. Briefly, 1.5×10⁶ HeLa cells were plated in 25 cm² flasks, and four such flasks were treated with 10 and 15 µM of chrysin for 48 h. The treated and DMSO control cells were collected and suspended in lysis buffer 17 (1 ml per 10⁷ cells) containing 10 μg/ml each of aprotinin (Catalog No. A6279; Sigma, USA), Leupeptin (Catalog No. 1167/25, Tocris, USA), and pepstatin (Catalog No. 1190/10, Tocris, United States) and shaken gently at 2–8°C for 30 min. The lysate produced was quantitated by the Pierce BCA Assay Kit (Catalog No. 23225; Thermo-Fisher Scientific, United States) following the manufacturer’s protocol. For this assay, 400 µg of a protein in 250 µL volume of the diluted cell lysate treated with chrysin (10 and 15 µM for 48 h) and the DMSO control was used for each membrane. The signal produced was then quantified by the chemiluminescent detector gel-doc system (Bio-Rad Laboratories, United States). The expression of proteins was analyzed by the intensity of proteins in the blot using Image Lab software (version 6.1). The fold change after normalization with the reference spot was calculated by comparing the treated (chrysin) values with the DMSO control values (mean of three experiments ±SD at p-value ≤ 0.05).

Chromatin-modifying enzymes like writers—DNA and histone methyl transferases, histone acetyl transferases, and erasers like histone deacetylases and histone demethylases—help in dynamically sustaining cell metabolism and controlling processes such as cell growth propagation and gene expression by recognition of specific “marks” on histone proteins and DNA (Kouzarides, 2007). RNA was extracted from the DMSO control and chrysin-treated HeLa cells (10 and 15 µM for 48 h). RT² Profiler™ PCR Array Human Epigenetic Chromatin Modification Enzymes (Catalog No. PAHS-085Z, Qiagen, United States) were used to check the expression of genes responsible for the modulation of DNA and histones including DNA methyl transferases, histone methyl transferases, histone acetyl transferases, and demethylases. RNA at a concentration of 1 µg was used to synthesize cDNA, and it was diluted to 1,350 µL with nuclease-free water and an equal amount of RT2 SYBR^® Green qPCR Master mix (Catalog No. 330504; Qiagen, United States) was added to this. From this mixture, 25 µL was poured into each well of the array plate having predefined primers, and the plate was run on ABI Quant Studio 3. The normalization was performed using GAPDH housekeeping gene and the fold change was calculated by comparing the chrysin-treated samples with the DMSO control. The statistical significance was calculated at p-value ≤ 0.5.

In order to understand the role of chrysin as an epigenetic modifier, the Histone Extraction kit (Catalog No. ab113476, Abcam, Cambridge, UK), and Histone H3 and H4 Modification Multiplex Assay kits (Catalog Nos. ab185910 and ab185914) were procured from Abcam, Cambridge, UK. Following the extraction of histone using the Histone Extraction Kit, ∼100 ng of histone protein was used per well, and the protocol given by the manufacturer was followed. OD was measured at 405 nm, and graphs were plotted for reflecting the effect of chrysin compared to the DMSO control. The experiments were performed in triplicates, one-way ANOVA was used to determine the significance of the experiments, and p-value was maintained at ≤0.05.

Statistical analysis was performed by SPSS program (version 21). The data were examined by using one-way ANOVA followed by Tukey’s HSD post hoc analysis. All experiments were performed in triplicate. The results are expressed as the mean ± SD of three distinct experimentations. The statistical significance was set at p-value ≤0.05.

The colony-forming assay was performed to understand the long-term effect of chrysin on the growth and division of HeLa cells. After calculating the survival factor (SF), it was observed that the DMSO control had plating efficiency (PE) of 95%, whereas the survival factor for 5 and 10 µM of chrysin showed only 120 and 30 colonies, respectively. At 15 µM chrysin conc., very few colonies were formed. Hence, it can be inferred that chrysin restrained the capability of cells to form colonies. These results suggest that chrysin is not only capable of causing cell death but also leads to cytostatic state (Figure 1A and Figure 1B).

Likewise, chrysin decreased the migration capacity of HeLa cells as demonstrated by scratch-wound and invasion assay using trans-well. Chrysin increased the wound width by 8 and 14% at 10 µM conc. treatment for 24 and 48 h and 17 and 25% at 15 µM conc. treatment for 24 and 48 h, respectively, whereas in DMSO control cells, there was almost complete wound closure after 48 h (Figure 2A and Figure 2B). This was further corroborated by significant decrease in the number of migrating cells after chrysin treatment using the trans-well assay. Only 15 and 2.5% migration at 10 and 15 µM chrysin treatment for 48 h was observed, respectively, in comparison with the DMSO control cells (Figure 2C and Figure 2D).

qPCR was done to understand the effect of decreased methylation of various tumor-suppressor genes following the treatment of chrysin. It was observed that chrysin treatment increased the expression of various TSGs (such as TIMP3, TIMP4, RARB, RASIF1, TP53, PTEN, CDH1, and SOCS1) and reduced the expression of genes responsible for metastasis (viz. MMP 2, MMP 9, MMP 14, SNAIL1, SMAD3, SMAD4, and MTA1, 2) and the genes involved in the inflammatory process (viz. IL2, IL1A, IL6, and CxCL8). Chrysin treatment also decreased the expression of oncogenes like FOS, JUN, MYC, ESR1, and TWIST1 (Figure 3A and Table 1). Relative quantification (RQ) derived from the 2^–∆∆Ct method specifies the fold change in gene expression against the DMSO control after normalization with the selected endogenous gene (GAPDH). The upregulation is documented at RQ ≥ 1.5 and downregulation at RQ ≤ 0.5.

Proteome profiler–based quantitation of proteins that are involved in proliferation and migration and other cellular events revealed chrysin-supported modulation was consistent with mRNA expression. The treatment of HeLa cells with 10 and 15 µM of chrysin resulted in the downregulation of the expression of various proteins related to migration viz MMP2, MMP9, and MMP3, mesothelin, MUC1, leptin, and M_CSF; inflammatory proteins like CCL8/MCP-2, CCL7/MCP-3, IL-18 BPa, CXCL8/IL-8, and IL-2 Rα; and oncogenes like HER1, 2, 3, and 4, and ICAM-1/CD54. Proteins related to cell proliferation, growth, and apoptosis like BCL-X, HIF-1α, HNF-3β, and HO-1/HMOX1 were also downregulated, that is, related to tumor progression, whereas upregulation of E-cadherin (CDH1) was observed after chrysin treatment (Figure 3B, Figure 3C and Table 2). Fold changes in protein expression were calculated by comparing the treated cells with those of the DMSO control. The upregulation was fixed at ≥1.5 fold and downregulation at ≤0.5 fold.

Chrysin inhibited DNMT activity in HeLa cells in a dose-dependent manner. The incubation of the nuclear extract with 5, 10, and 15 µM of chrysin resulted in the inhibition of DNMT activity by 35, 54, and 61%, respectively, compared to the DMSO control (Figure 4A).

Nuclear extracts were incubated with increasing concentrations (5, 10, and 15 µM) of chrysin; it was found that it inhibited the activity of HDACs by 30, 36, and 42% in a dose-dependent response compared with the DMSO control (Figure 4B).

Histone acetyl transferases cause acetylation at N-terminal tails of histone proteins. The incubation of nuclear extract with varying concentrations of chrysin (5, 10, and 15 µM) showed decline in HAT activity in the treated cells compared to the DMSO control. A decrease of 22, 34, and 52% was observed at 5, 10, and 15 µM conc. of chrysin treatment, respectively (Figure 4C).

HMT H3K9 can add methyl groups at histone three and lysine 9. All the methylation marks—mono, di, and trimethylation—are repressive marks. The incubation of HeLa cell nuclear extracts with 5, 10, and 15 µM conc. of chrysin reduced the activity of the enzyme by 25, 38, and 45%, respectively (Figure 4D).

RT (Rahman et al., 2016) Profiler™ PCR Array Human Epigenetic Chromatin Modification Enzymes (Catalog No. PAHS-085Z; Qiagen, USA) were used to check the expression of various chromatin-modifying enzymes following the treatment of chrysin (@ conc. 10 and 15 µM) for 48 h compared to the DMSO control. Chrysin treatment down-regulated the expression of DNA methyltransferases like DNMT1, 3A, and 3B significantly. HDAC1, 2, 3, 4, and 11 also showed a steep decline after the above-stated chrysin treatment. Remarkably, downregulation of WHSC1, AURKA, AURKB, and AURKX. EHM2, PRMT8, and HAT1 were also observed after 10 and 15 µM of chrysin treatment. However, enhanced expression of SETD2, ESC O 2, and CIITA was found after the same chrysin treatment (RQ in Figure 5A and Table 3).

Chrysin modulated the expression of methylation, acetylation, and phosphorylation H3 and H4 marks. H3K9me1, H3K9me2, H3K9me3, H3K27me1, H3K27me2, H3K27me3, H3K36me1, H3K36me3, H3K79me1, H3K79me2, and H3K79me3 marks were reduced after the treatment of HeLa cells with 15 µM of chrysin for 48 h; similarly, H3 acetylation marks were diminished after treatment with 15 µM of chrysin (Figure 5B). The expression of H3K9ac, H3K18ac, H3K14ac, and H3K56ac was reduced after chrysin treatment. Likewise, the acetylation marks at H4 were also modulated against chrysin treatment including H4K5ac, H4K8ac, H4K12ac, and H4K16ac. H4 methylation marks, like H4K20me1, H4K20me2, and H4K20me3, showed decreased expression after chrysin treatment (Figure 5C). Phosphorylation marks of H3ser28p, H4ser10 p, H4R3m2a, and H4Rm2s were also decreasingly expressed after 15 µM chrysin treatment of HeLa cells for 48 h (Figure 5B and 5C).

An obvious decrease in global methylation was observed after 48 h treatment of 5, 10, and 15 µM of chrysin against HeLa cells. Global DNA methylation was studied by comparing with the DMSO control. This was reduced to 61, 44, and 30% against 5, 10, and 15 µM chrysin treatment of HeLa cells, respectively (Figure 6A).

Methylation-sensitive restriction enzyme PCR revealed that chrysin decreased the promoter methylation of crucial tumor-suppressor genes of HeLa cells. These TSGs included APC, BRCA1, CDH1, PTEN, GSTP1, FHIT, DAPK1, CDH13, CDKN2A, MGMT, MLH1, RARB, RASSF1, SOCS1, VHL, WIFI, and TIMP3. The methylation percentage of the abovesaid genes decreased significantly as APC (8%, 3%), BRCA1 (18%, 2%), CDH1 (55%, 43%), CDH13 (39%, 3%), CDKN2A (14%, 4%), DAPK1 (17%, 11%), FHIT (5%, 3%), GSTP1 (7%, 3%), MGMT (14%, 4%) MLH1 (22%, 15%), PTEN (14%, 11%), RARB (22%, 11%), RASSF1 (19%, 15%), SOCS1 (73%, 56%), TIMP3 (13%, 3%), VHL (15%, 12%), and WIFI (82%, 61%) at 10 and 15 µM chrysin, respectively, compared to the DMSO control, wherein the methylation percentage was much higher (Figure 6B and Table 4).