Limoniastrum guyonianum samples were collected from El Hamâ at Gabbes (a region situated in southern Tunisia). Dr. Fethia Skhiri (Department of Botany, Higher Institute of Biotechnology, University of Monastir) performed sample identification and verification according to the Tunisian Guide on Flora 30 . A voucher specimen (#L.g-10.09) was preserved for future reference. Luteolin (> 90% of purity) was purchased from Extrasynthese (Genay, France). 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) was from Euromedex (Mundolsheim, France), propidium iodide (PI), Tris Buffered Saline with tween 20 (TBST) and dimethylsulfoxide (DMSO) from Sigma-Aldrich (St. Quentin Fallavier, France). Dulbecco’s Modified Eagle’s Medium (DMEM), fetal calf serum (FCS), trypsin and L-glutamine were purchased from Invitrogen Life Technologies (Cergy Pontoise, France). Folin-Ciocalteu phenol reagent was obtained from BDH laboratory (Poole, England). Sodium carbonate (Na₂CO₃) was purchased from Acros Organics (Geel, Belgium). Nitrite sodium (NaNO₂) and aluminum chloride (AlCl₃) were procured from Aldrich (Steinheim, Germany).

The collected gall samples were shade-dried, powdered, and then stored in a tightly closed container for further use. When needed, powdered gall (100 g) was extracted in boiling water (1 L) for 15–20 min and after filtration, the aqueous extract was frozen and then lyophilized and kept at 4°C. The total aqueous extract concentrate yield (per gram dried plant material) was determined using the formula: 100 x weight (g) of dried extract/dry-weight (g) of plant material. The actual percentage yield in this study was 17.8%. From this material, extract solutions containing different concentrations from 100 to 300 μg/ml were then prepared for use in the evaluation of their cytotoxic and pro-apoptotic effects on HeLa cells. The polyphenol content of L. guyonianum gall aqueous extract was quantified by the Folin-Ciocalteau method 31 32 and was expressed as gallic acid equivalent. Aliquots of test sample (100 μl) were mixed with 2.0 ml of 2% Na₂CO₃ and incubated at room temperature for 2 min. After the addition of 100 μl of 50% Folin-Ciocalteau phenol reagent, the reaction tube was incubated for 30 min at room temperature, and finally absorbance was read at 720 nm. A known volume of the extract was placed in a 10 ml volumetric flask to estimate flavonoid content 33 . After addition of 75 μl of NaNO₂ (5%), 150 μl of freshly prepared AlCl₃ (10%), and 500 μl of NaOH (1 N), the volume was adjusted with distilled water until 2.5 ml. After 5 min incubation, the total absorbance was measured at 510 nm. Quercetin was used as a standard for constructing a calibration curve.

The method described by 34 was used for the determination of tannin content of samples. Extraction of tannins was achieved by dissolving 5 g of sample in 50 ml of distilled water in a conical flask, allowing the mixture to stand for 30 min with shaking the flask at 10 min intervals, and then centrifuging at 5000 g to obtain a supernatant (tannin extract). The extract was diluted to 100 ml in a standard flask using distilled water. Five milliliters of the diluted extract and 5 ml of standard tannic acid (0.1 g/L) were measured into different 50 ml volumetric flasks. One milliliter of Folin-Denis reagent was added to each flask followed by 2.5 ml of saturated sodium carbonate solution. The solutions were made up to the 50 ml mark with distilled water and incubated at room temperature (20–30°C) for 90 min. The absorption of these solutions was measured against the reagent blank (containing 5 ml distilled water in the place of the extract or the standard tannic acid solution) at 760 nm wavelength. Tannin content was calculated in triplicates as: sample reading/standard reading × 20 35 .

The human cervical cancer cell line HeLa was obtained from the American Type Culture Collection (Rockville, Maryland, USA) and maintained in a humidified incubator with 5% CO2 at 37°C, and grown in DMEM (Dulbecco’s Modified Eagle’s Medium). The medium was supplemented with 10% (v/v) fetal calf serum (FCS, Biowhitaker, Lonza, Belgium), 2 mM glutamine, 100 U/ml penicillin and 50 mg/ml streptomycin (Sigma St. Louis, MO).

Cells were seeded in 96-well cell culture plates at a density of 0.5 × 10⁴ cells/well, grown for 24 hours and exposed to different concentrations of G extract or luteolin for 24 hours. Cell proliferation rate was then assessed by colorimetric assay using the CellTiter 961 Aqueous One Solution Cell Proliferation Assay (MTT), following the manufacturer’s recommendations. Early and late apoptosis were monitored by flow cytometry (Guava PCA-96 Merck/Millipore, Molsheim, France). To discriminate between negative and positive events in the analysis, a non-stained control sample from each culture condition always accompanied acquisition of the stained cells to define their cut off. Gates were drawn around the appropriate cell populations using a forward scatter (FSC) versus side scatter (SSC) acquisition dot plot. Late apoptotic cells are double labelled by Annexin V and 7-AAD (Guava Nexin Reagent kit Merck/Millipore). Cytometers performances are checked weekly using the Guava easyCheck Kit 4500–0025 (Merck/Millipore/Guava Hayward, CA, USA).

Cells were seeded in 96-well cell culture plates at a density of 0.5 × 10⁴ cells/well and grown for 24 hours, then exposed to different concentrations of G extract or luteolin for 24 hours. Cells were then washed with phosphate-buffered saline (PBS), trypsin detached then fixed with paraformaldehyde (1%). After 30 min incubation at room temperature, 5 μl of propidium iodide was added in each well (1 μg/ml). Cellular DNA content was assessed by capillary cytometry (Guava EasyCyte 96 Plus). Data were analyzed on the Guava CytoSoft™ Express Pro software (Merck/Milli pore/Guava Tech). CytoSoft Express Pro was used to identify the three cell cycle phases and calculate relevant statistics, including population percentages (subG1, G0/G1, S and G2/M phases).

HeLa cells were treated with G extract (200 μg/ml) or luteolin (25 μM) for 48 hours. DNA was purified using QIAamp® DNA Kit. The content of methylated DNA was determined using 200 ng of DNA from untreated cells, treated cells with G extract or luteolin, as described by the manufacturer; Sigma’s Imprint® Methylated DNA Quantification Kit.

HeLa cells (6 × 10⁵) were seeded into 6-well cell culture plates and grown for 24 hours. Cells were treated with different concentrations of G extract or luteolin for 24 and 48 hours. The cells were then harvested, centrifuged to discard the DMEM medium, washed with cold PBS (phosphate buffered saline), resuspended in RIPA buffer (25 mM Tris, pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate and 0.1% SDS; Sigma–Aldrich, USA) containing protease inhibitors. Equal amounts of total protein were separated on 10–12% polyacrylamide gel and electrophoretically transferred to a nitrocellulose membrane. After blocking with 5% non-fat dry milk or 3% BSA (Bovine Serum Albumin) and tween 20 in PBS, the nitrocellulose membranes were incubated with either a mouse monoclonal anti-UHRF1 antibody (Proteogenix, Oberhausbergen, France), a mouse monoclonal anti-DNMT1 (clone 60B1220.1, Proteogenix), and a rabbit polyclonal anti-p16^INK4A antibody (DeltaBiolabs, Gilroy, CA) according to the manufacturer’s instructions (4°C, overnight). Membranes were thereafter incubated with the appropriate horseradish peroxidase-conjugated secondary antibody (diluted to 1:10,000 for anti-mouse antibodies and 2: 10,000 for anti-rabbit antibody) at room temperature for 45 minutes. The membranes were then washed with TPBS five times. Signals were detected by chemiluminescence using the ECL Plus detection system (Amersham, GE Healthcare UK Limited).

Data were analyzed with student’s t-test and presented as mean value ± S.E.M of three independent measurements in separate experiments.

Aqueous gall extract from L. guyonianum was the subject of a chemical study with the aim of having a global idea in their composition. The metabolites contents of the tested extract are presented in Table 1. Quantitative phytochemical analysis showed that the extract contained an important quantity of flavonoids, polyphenols, and tannins. In fact, 1 mg of G extract was equivalent to 85 μg of gallic acid and 460 μg of quercetin. The higher content of tannins was also recorded in the extract with 77.5 mg/100 g.

Values are means ± S.E.M. of three independent experiments.

The present study was undertaken to investigate the effect of G extract on the expression of UHRF1/DNMT1 tandem known to be involved in gene expression regulation via DNA methylation 9 11 . HeLa cells were treated with different concentrations (100, 200 and 300 μg/ml) of G extract for 24 and 48 hours. As shown in Figure 1A, treating the cells with 300 μg/ml of G extract for 24 hours induced a significant decrease in the expression of UHRF1, DNMT1 and this expression was abolished after 48 hours of treatment. Cells treatment with 200 μg/ml of G extract also induced a significant decrease of UHRF1 and DNMT1 expressions but only after exposure for 48 hours whereas at 100 μg/ml there was no effect. Several studies have been shown that UHRF1 negatively regulates the expression of the p16 ^INK4A tumor suppressor gene 19 36 . Thus, we aimed to know whether G extract and luteolin could affect the expression of p16^INK4A in HeLa cell line. Our results showed that G extract induced a dose dependently up-regulation of p16^INK4A expression (Figure 1A). This effect was associated with the G extract-induced down-regulation of UHRF1 and DNMT1 expression (Figure 1A). Quantitative phytochemical analysis of G extract showed that flavonoids are the major compounds present in this extract, which suggest that G extract-induced effect on UHRF1 and DNMT1 expression could be attributed, at least in part to these compounds. In order to obtain evidence for this hypothesis, the effect of luteolin, a dietary flavonoid on the expression of UHRF1, DNMT1 and p16^INK4A proteins has been investigated. As shown in Figure 1B, treating cells with luteolin induced a dose and time down-regulation of UHRF1. Indeed, UHRF1 expression was significantly decreased after 24 hours treatments and approximately disappeared at 50 μM after 48 hours (Figure 1B). For DNMT1, only 50 μM induced a significant decrease of DNMT1 expressions after incubation for 24 hours. After treatment of cells for 48 hours, DNMT1 expression was significantly decreased at 25 μM and totally abolished at 50 μM whereas at 12.5 μM there was no effect (Figure 1B).

Next, we wanted to determine whether the reduced UHRF1 and DNMT1 levels in G extract and luteolin-treated HeLa cells are correlated with an effect on genomic DNA methylation. We assessed global genomic DNA methylation by Imprint® Methylated DNA Quantification assay. As shown in Table 2, a general decrease in genomic DNA methylation was evidenced by both natural products. Indeed, our results demonstrate that G extract and luteolin inhibited DNA methylation as compared to untreated cells (Table 2) with percent inhibition of 42.4% ± 1.6% and 46.5% ± 1.1% in the presence of G extract and luteolin, respectively. Altogether, these findings showed that both G extract and luteolin were able to decrease UHRF1 and DNMT1 expression leading to a reduced genomic DNA methylation which could induce the re-expression of the p16 ^INK4A tumor suppressor gene.

Total DNA was isolated from HeLa cancer cells using QIAamp^® DNA Kit. the content of methylated DNA was determined using 200 ng of DNA from untreated cells (C), treated cells with 200 μg/ml of G extract (G200 or with 25 μM of luteolin(L25) for 48 hours and the commercial methylated control (MC) (Imprint Methylated DNA Quantification Kit) Values are means ± S.E.M. of three independent experiments. Statistically significant, *P < 0.001 (versus the untreated cells).

Considering that p16 ^INK4A tumor suppressor gene is a downstream target of UHRF1 and a negative regulator of cell proliferation 17 36 , we then wanted to determine whether G extract- or luteolin-induced up-regulation of p16^INK4A leads to cell proliferation inhibition and cell cycle arrest. As illustrated in Figure 2, exposure of HeLa cells to G extract (A) or luteolin (B) inhibited cell proliferation in a dose- and time-dependent manner. The IC₅₀ values were determined graphically and the inhibition percentages were calculated. Inhibition of proliferation of HeLa cells, by G extract, reached a maximum of 79.6% and 59.7% at a concentration of 300 μg/ml after 48 and 24 hours of incubation, respectively (Figure 2A). IC₅₀ values were 170 μg/ml and 140 μg/ml of G extract after 24 and 48 hours treatment, respectively. Interestingly, G extract had no effect on normal human keratinocytes when cells were treated with similar concentrations for 24 and 48 hours (Figure 2C). This suggests that G extract specifically targets cancer cells.

Luteolin was also able to induce cytotoxicity in HeLa cells (Figure 2B) with an IC₅₀ value of 21.8 μM after 24 hours. At 50 μM, luteolin decreased proliferation of HeLa cells by 83.8% and 85.9% after 24 hours and 48 hours of incubation, respectively. These results indicate that both natural products induce a dose-dependent cell growth inhibition of HeLa cells.

Because cell proliferation is a consequence of the progression of the cells through the different phases of the cell cycle, we next determined the effects of G extract and luteolin on the cell cycle distribution (Figure 3). HeLa cells were incubated in the presence and/or absence of different concentrations of G extract (A) or luteolin (B) for 24 hours. Treatment of HeLa cells with the extract caused an increase in G2/M peaks and a decrease in the S and G0/G1-phases fraction in a concentration-dependent manner (Figure 3A). Indeed, the percentage of cells in the G0/G1 phase was decreased from 50.1% (control) to 32.3% at 300 μg/ml whereas an accumulation of the cell population was observed in the the G2/M from 7.5% in untreated cells to 19.6% at the same concentration. Similarly to G extract, treatment of HeLa cells with luteolin caused an increase in G2/M phase and a decrease in the G0/G1-phase fraction in a concentration-dependent manner (Figure 3B). It appears therefore that G extract is able to inhibit the proliferation of Hela cells by promoting cell cycle arrest at the G2/M phase.

UHRF1 down-regulation has been shown to induce apoptosis in cancer cells 37 . Moreover, it has recently been demonstrated that UHRF1 down-regulation inhibits cell growth and induces apoptosis of colorectal cancer through p16^INK4A up-regulation 17 . Thus, we next investigated whether G extract- or luteolin-induced UHRF1 down-regulation and p16^INK4A up-regulation could induce apoptosis in HeLa cells. As shown in Figure 4, increasing concentrations of both products are associated with increasing number of apoptotic cells. Indeed, the percentage of apoptotic cells significantly increased from 8.6 (control) to 46.4% at 200 μg/ml and 54.4% at 300 μg/ml of G extract (Figure 4A), and to 37.2% at 25 μM and 49.2% at 50 μM of luteolin (Figure 4B). As expected, the calculated half-maximal effect of G extract on apoptosis was 170 μg/ml for 24 hours of treatment. Hence, these data were consistent with those obtained from cell proliferation assays (Figure 2). As a next step, cell cycle phase distribution analysis was focused on the detection of specific G0/G1 apoptotic cells; as shown in Figure 4, increasing concentrations of G extract led to increasing number of hypodiploid sub-G0/G1 cells (Figure 4C). Thus, G extract induced an increase in sub-G1 peak in a concentration-dependent manner ranged from 10.2% to 27.6% at concentrations of 100 and 300 μg/ml, respectively. At 50 μM of luteolin, an increment from 8.1% (control) 23.5% was observed in subG1 phase (Figure 4D). Finally, all these results suggest the occurrence of apoptosis in HeLa cells related to UHRF1 down-regulation and p16^INK4A up-regulation when exposed to G extract or luteolin.