Quercetin abrogates chemoresistance in melanoma cells by modulating ΔNp73

Cell lysates were electrophoresed in 7 and 10% NUPAGE gels (Invitrogen Corp., CA, USA). Separated proteins were electrophoretically transferred to Hybond PVDF membrane (Amersham Pharmacia Biotech, UK) and the membrane was blocked for 1 hr by incubating the membrane in I-block (Tropix kit, Applied Biosystems, CA, USA). Primary antibodies were used at the dilutions which the manufacturers suggested. ALP conjugated goat anti-rabbit IgG was used at a dilution of 1:10000 for antibodies for phospho-p53, DNApk, Bax and PARP whereas, anti-mouse IgG was used for Total p53, Total p73, phospho-ATM and GAPDH at a dilution of 1:10000. Western detection was carried out using CDP star from Tropix kit, Applied Biosystems, CA, USA.

The p53 wild type and mutant cell lines were grown up to 50% confluency and were treated as mentioned above. Apoptosis was determined using Fluorescein isothiocyanate-conjugated Annexin V (Annexin V-FITC)/Propidium Iodide (PI) apoptosis detection kit (R&D systems, Minneapolis, MN, USA) as per manufacturer's instructions. Approximately 5 × 10⁵ cells were resuspended in 100 μl of 1× binding buffer, 1 μl of Annexin V-FITC and 10 μl of propidium iodide. After 15 min incubation at room temperature in the dark, 400 μl of 1× binding buffer was added and the cells positive for Annexin V-FITC and/or PI were analyzed using a BD FACS flow cytometer.

Homogenization of cells and isolation of RNA were performed using QIAshredder spin columns and an RNeasy Kit as instructed by the manufacturer (Qiagen, Valencia, CA). 1 μg of RNA was reverse transcribed using a Super Script III Kit as instructed by the manufacturer (Invitrogen, Carlsbad, CA) and diluted 1:5 for subsequent analysis. The following PCR reaction mix was used: 5 ul of diluted cDNA, 1 ul of mixed forward and reverse primers (10 uM each), 12.5 ul SYBR Green (Qiagen), and nuclease-free water to a final volume of 25 ul. For non-quantitative PCR, cDNA was amplified for thirty cycles. Forty cycles of quantitative PCR were performed (95°C for 15 seconds, 54°C for 30 seconds, 72°C for 30 seconds) using an iQ5 Real-Time PCR Detection System (BioRad, Hercules, CA) and run on a 1% agarose gel. Real-time PCRs were run in triplicate for each cDNA sample using an iQ5 Real-Time PCR Detection System. Forty cycles of PCR were performed (as described above) with fluorescence detection during the 72°C step at each cycle. The data were analyzed using the 2^-ΔΔCt method [18], and results were normalized to S15, which remains unchanged in response to treatment. Normalized values were plotted as relative fold over untreated. The following primers were purchased from Integrated DNA Technologies (Coralville, IA, USA): S15 [19] transcriptionally active p73 (TAp73) [20] and ΔNp73 [21].

The cells were grown on cover slips in 100 mm tissue culture dishes and were treated with TMZ and Qct as mentioned above. The cover slips were placed in 6 well dishes and washed with PBS and fixed with 95% ethanol and 5% glacial acetic acid for 5 min. The slides were rinsed with PBS and were incubated with 0.5% of Triton X-100 in PBS for 10 min to permeabilize the membranes and rinsed again. After blocking the endogenous peroxidase with 3% hydrogen peroxide (H₂O₂) in PBS for 20 min the cover slips were processed according to staining procedure of the manufacturer's protocol for Histostain plus kits, Zymed Laboratories (Invitrogen, CA, USA). Total p73 antibody was used in the dilution of 1:250.

siRNA transfection was carried out according to manufacturer's protocol (Invitrogen, CA, USA). Cells were grown up to 50% confluence in antibiotic free medium in 100 mm dishes. Stealth RNAi for p73 at varying concentrations (1 nM-50 nM) was diluted in 1.5 ml OPTI-MEM I reduced serum. 30 μl of Lipofectamine™ 2000 was diluted in 1.5 ml OPTI-MEM I reduced serum medium, mixed gently. After 5 min incubation at RT, diluted oligomer was combined with diluted Lipofectamine™ 2000, mixed gently and incubated at room temp for 20 min. The oligomer-Lipofectamine™ 2000 complexes were added to each plate in OPTI-MEM I reduced serum medium by mixing gently and by rocking the plate back and forth. After 6 hrs incubation in a 5% CO₂ incubator at 37°C, the plates were subjected to TMZ treatment for 48 hrs without removing the complexes. BLOCK-iT Fluorescent oligomer was used as positive control.

DB-1 cells were transiently transfected with 6 μg of tyrosinase or pcDNA3 DNA using Lipofectamine™ 2000 in serum free OPTI-MEM medium for 5 hr followed by leaving the complex in Neomycin containing α MEM complete medium for at least 18 hrs as demonstrated previously [12].

To establish a dose response to TMZ, DB-1 cells were treated with varying concentrations of TMZ (0-400 μM) for 48 hrs and the amount of cell death (apoptosis and necrosis) was determined by Annexin V-FITC/PI double staining and flow cytometry analysis. Consistent with a chemoresistant melanoma phenotype, DB-1 cells were resistant to TMZ-induced cell death at concentrations lower than 100 μM (Figure 1A). In contrast, Qct treatment (0-100 μM) resulted in dose dependent cell death at all concentrations studied (Figure 1B).

Bild vergrößern

For subsequent combination studies, a concentration of 400 μM TMZ plus 75 μM Qct was used as those concentrations individually resulted in a distinguishable degree of apoptotic cell death (Figure 1C). To treat cells in combination TMZ was administered for 48 hrs followed by 24 hrs of Qct. In DB-1 cells treatment with TMZ or Qct induced 9.2% (± 0.4%) or 11.4% (± 0.4%) apoptosis (lower right quadrant of flow-cytometeric scatter plot, Figure 1E), respectively (Figure 1C). However, the combination treatment of TMZ plus Qct induced 34.25% (± 2.6%) apoptosis, which was greater than additive in these cell lines. Late apoptosis or necrosis (upper right quadrant of flow-cytometeric scatter plot, Figure 1E) was also increased in a more than additive manner. Apoptosis analysis was also carried out in p53 mutant SK Mel 28 cells and there was no significant increase in cell death across treatments (Figure 1D). Representative flow cytometeric scatter plots for DB-1 cells can be seen in Figure 1E.

We next investigated the effect of Qct and TMZ on the DNA damage response. ATM is a protein kinase that is rapidly activated by DNA double strand breaks (DSBs) and is known to phosphorylate downstream target substrates such as p53. An increase in ATM phosphorylation at Ser 1981 (p-ATM) was observed in DB-1 cell lines treated with TMZ and a slight increase in p-ATM was observed by Qct treatment (Figure 2). The greatest increase in p-ATM was observed in the combination treatment. A similar increase in p-ATM was observed in SK Mel 28 cells, but Qct did not measurably activate this protein. Induction of DNApk, another DNA repair protein, was also observed in all the treatment groups (Figure 2).

Bild vergrößern

Because ATM was activated we investigated the effect of treatment on post-translational modification of p53 by evaluating phosphorylation of its serine moieties. In the DB-1 cells phosphorylation of p53 at serine 15, 37 and 392 was increased following treatment with TMZ or TMZ plus Qct (Figure 3A). The expression of total p53 also increased accordingly. We confirmed the sensitization effect in SK Mel 5, another p53 wild type cell line (Figure 3B). However, there was a minimal activation of p53 at serine 15 in SK Mel 28 cells (Figure 3C) and phosphorylation at other sites could not be detected (data not shown). No changes in the levels of total p53 were detected in the SK Mel 28 cells.

Bild vergrößern

The effect of the treatments on the proteins downstream of p53, including Bax and cleaved PARP were also investigated (Figure 4). The levels of apoptosis (Figure 1C), cleaved PARP and caspase 3 (Figure 4A and 4B) did not correlate with the increased levels of p-ATM and phospho-p53 (Figures 2 and 3), indicating that signaling between p53 and its downstream targets was attenuated and thus blocked apoptosis. In the SK Mel 28 mutant cell lines, there were no changes in total p53 levels (data not shown) and levels of cleaved PARP (Figure 4C).

Bild vergrößern

To investigate the abrogation of apoptosis in the cells treated with TMZ we examined the role of p53 family members, which have been shown to have inhibitory effects on p53 function. The p73 gene is a homologue of full length p53 and is involved in the transactivation of p53 target genes and thereby causes an induction in apoptosis and inhibition of cell proliferation [22]. Transcriptionally active or TAp73 is the active isoform of p73 frequently expressed in human tumors [23] and inhibited by either N terminally truncated p63 (ΔNp63) or ΔNp73. ΔNp73 can also antagonize p53. We performed RT-PCR for TAp73 and ΔNp73 to determine the isoforms expressed in DB-1 and SK-Mel28 cells. The only detectable isoform in DB-1 cells was ΔNp73 (Figure 5A) so DB-1 cells were used to further investigate the role of the p73 deletion mutant on apoptosis and chemoresistance. Following treatment with TMZ or the TMZ plus Qct combination there was an increase in the transcription of only ΔNp73 (Figure 5B). Western blot analysis also revealed a single band of increased protein levels of p73 in the DB-1 cells treated with TMZ alone or in combination with Qct and there was no change in protein levels of p73 in the SK Mel 28 cells (Figure 5C).

Bild vergrößern

The localization of p73 plays a major role in its activity so the effect of Qct on localization of p73 was investigated. The full length transcriptionally active form of p73 (TAp73) is normally found in the nucleus and can induce the transcription of downstream p53 target genes. However, when the ΔN form is in the nucleus, it can antagonize p53, which could be the reason why the levels of Bax (Figure 4) did not correlate with the increased levels of phospho-p53 (Figure 3) and the abrogation of apoptosis in the TMZ-treated cells. Consistent with previous reports of p73 localization, immunocytochemistry (ICC) analysis revealed the nuclear staining in the untreated cells that increased following treatment with TMZ (Figure 5D). In contrast, Qct treatment caused a re-distribution of ΔNp73 from the nucleus into cytoplasm as well as in the nucleus. The re-distribution was also observed with the combination treatment in DB-1 cell lines (Figure 5D) and indicates that the redistribution induced by Qct can reduce the antagonist effect of ΔNp73.

To confirm the role of ΔNp73 in the attenuation of the apoptotic response siRNA experiments were performed. Naumann et al. [24] identified PARP cleavage as a hallmark of TMZ activity, so we investigated the effect of p73 siRNA on PARP levels. Treatment with TMZ resulted in no significant cleavage of PARP compared to control (Figure 6B). However, incubation with p73 siRNA in TMZ-treated cells knocked down the ΔNp73 protein levels and resulted in an increase in the ratio of cleaved to uncleaved PARP (Figure 6C). Taken together, these results confirmed the role of ΔNp73 as a p53 antagonist in the response to TMZ and suggests that Qct promotes apoptosis by inducing the translocation of ΔNp73 out of the nucleus.

Bild vergrößern

Tyrosinase activity is increased as melanoma tumors develop and we have previously shown that tyrosinase activity is associated with an increase in p53 [11, 12]. Because p53 can transcriptionally activate p73, we investigated if tyrosinase overexpression could induce the expression of p73. We transiently overexpressed tyrosinase (Figure 6D) and observed an increase in the expression of p53 and its transcriptional target p73 (Figure 6D). Therefore, tumorigenesis in melanoma, which coincides with the overexpression of tyrosinase, could also be associated with an increase in transcription of p53 antagonists and inducing resistance to chemotherapies. Targeting these antagonists through compounds such as quercetin could serve as an effective therapeutic or treatment modifier by restoring p53 activity.

We have shown for the first time that Qct can affect the cellular distribution of ΔNp73 and abrogate its anti-apoptotic effects. We also demonstrated in melanoma that Qct can affect the phosphorylation of p53, which is a key factor in mediating TMZ-induced apoptosis and a major determinant of cancer cell response to TMZ [25]. TMZ and DTIC methylate DNA at O⁶ methyl guanine residues and cause mispairing with thymine and activation of MutSa-dependent mismatch repair resulting in apoptosis [26, 27]. TMZ treatment during the present study caused an induction in total p53 and phosphorylation of its serine residues at 15, 37 and 392. Studies by Mhaidat NM et al. [28], illustrate that the sensitivity of melanoma cells to TMZ was dependent on their p53 status and associated G2/M arrest. We did confirm the central role of p53 with TMZ treatment and further demonstrated TMZ is a strong activator of ATM. ATM activation was enhanced by Qct, which provides a mechanism for the activation and stabilization of p53 while modulation of p73 likely contributes to increased p53 activity.

Even though apoptosis is considered the major mechanism of death by TMZ, [29] Qct also increased necrosis and indicates that apoptosis might not be the only therapeutic outcome when using combination treatments. Cell death reported here consists of an apoptotic fraction and a second fraction of cells that consisted of late apoptosis or necrosis. Necrosis as a mechanism of cell death can not be ruled out and is line with another study in malignant melanoma cells [30] that concluded TMZ-induced O⁶-methyl guanine triggers the apoptotic as well as necrotic pathway through the formation of DSBs. However, these pathways appear in part mediated by p53 as the SK Mel 28 cells were relatively void of treatment-induced cell death.

Another mechanism of action of Qct appears to be through the change in distribution of p73, allowing for an increase in p53 nuclear activity. p73 belongs to the p53 family of proteins that exhibit sequence homology [31] and they include p53, p63 and p73. Endogenous TAp73 is upregulated in response to DNA damage or treatment with chemotherapeutic drugs and provides antitumor activity, while the upregulation of ΔNp73 promotes resistance to these drugs [32]. The functionality of the p53 family members primarily depends on the nuclear localization of p73. Studies demonstrate that the export of the truncated form, ΔNp73, to the cytoplasm is a major inducer of p53 functionality [33]. In earlier studies with hepatocellular and cholangiocellular carcinomas, p73 was reported to be confined to the nucleus [34, 35]. It was also found to be localized mainly in the nucleus of undifferentiating neuroblastomas [36].

Our ICC study demonstrates that expression of ΔNp73 was confined to the nucleus in control melanoma cells and this expression was even increased in the nucleus following TMZ treatment. Re-localization to the cytoplasm could indicate a shift in the balance of p53 to p73 in the nucleus and allow for transcription of p53 target genes, or may indicate a functional change in p73 itself. Inoue et al. [37] demonstrated that p73 undergoes active export which in part is mediated by nuclear export signal (NES) localized in the C terminus. Qct can modify phosphorylation which is required for a functional NES [38]. The change in localization could be therefore be related to changes in nuclear import and export signals and related changes in the overall distribution of p73.

The notion that p53 upregulates its own antagonist provides a unique mechanism for averting functional p53 which is commonly found in melanoma. We have previously demonstrated that tyrosinase overexpression could induce ROS which may induce p53 [11, 12]. The upregulation of p53 in a tumor would not be beneficial unless it was transcribing an antagonistic protein such as ΔNp73. Box et al [14] have also proposed that p53 and antagonists are increased with melanogenesis. Our results with overexpression of tyrosinase are consistent with this notion. Abrogation of the protective effect of p53 antagonists would provide a valuable target for tumor therapy, especially since relatively non toxic molecules such as Qct can affect its translocation and activity.