A Nucleus-Imaging Probe That Selectively Stabilizes a Minor Conformation of c-MYC G-quadruplex and Down-regulates c-MYC Transcription in Human Cancer Cells

The c-MYC proto-oncogene is a regulator of fundamental cellular processes such as cell cycle progression and apoptosis. The development of novel c-MYC inhibitors that can act by targeting the c-MYC DNA G-quadruplex at the level of transcription would provide potential insight into structure-based design of small molecules and lead to a promising arena for cancer therapy. Herein we report our finding that two simple bis-triazolylcarbazole derivatives can inhibit c-MYC transcription, possibly by stabilizing the c-MYC G-quadruplex. These compounds are prepared using a facile and modular approach based on Cu(I) catalysed azide and alkyne cycloaddition. A carbazole ligand with carboxamide side chains is found to be microenvironment-sensitive and highly selective for “turn-on” detection of c-MYC quadruplex over duplex DNA. This fluorescent probe is applicable to visualize the cellular nucleus in living cells. Interestingly, the ligand binds to c-MYC in an asymmetric fashion and selects the minor-populated conformer via conformational selection.

The c-MYC oncogene is overexpressed in a broad spectrum of human malignancies and emerged as a potential therapeutic target for cancer treatment 1 . The overexpression of c-MYC in hepatic cells is frequently associated with the development of hepatocellular carcinoma 2 . Small molecule mediated inhibition of c-MYC imparts growth arrest in liver cancer cells and even down-regulates the human telomerase reverse transcriptase (hTERT) activity 3,4 . Therefore c-MYC is an attractive target in developing new therapies for hepatocellular carcinoma. Transcription of c-MYC is primarily regulated by a 27 base guanine-rich sequence present within the nuclease hypersensitivity element III1 (NHE III1) 5 . This sequence, located -142 to -115 bp upstream of the P1 promoter of c-MYC oncogene, exists in equilibrium between transcriptionally active forms (double helical and single stranded) and a silenced form, which is able to fold into a G-quadruplex structure 6 . The 27-mer sequence responsible of c-MYC regulation contains five guanine runs and it has been shown that in K + -containing solution only the four consecutive 3′ G-runs are involved in the formation of the major G-quadruplex structure which causes the gene transcriptional silencing [7][8] . However, the major G-quadruplex formed under superhelicity conditions involves the four λ max = 260, 290 nm), when excited at 290 nm in 100 mM Tris•HCl buffer, at pH 7.4 containing 100 mM KCl. Upon titration with the promoter quadruplexes (0-6 equiv.), BTC f (1.0 μM) showed an increase in fluorescence (2-3.3 fold) along with a blue shift (14-20 nm). The degree of fluorescence enhancement was the highest for c-MYC (3.3 fold) compared to the c-KIT1 (2.0 fold) and c-KIT2 (2.0 fold) quadruplexes. The blue shift was the largest for the c-MYC (20 nm) compared to the c-KIT1 (15 nm) and c-KIT2 (14 nm). Almost negligible fluorescence enhancement was observed (Fig. 3c) when BTC f was titrated with the ds DNA (0-6 equiv.). The dissociation constants (K D ) calculated from the fluorescence spectra were found to be in the low micromolar range (0.3-1.38 μM). Ligand BTC f shows a 5-fold selectivity for c-MYC with a K D value of (K D = 300 ± 15 nM), over c-KIT1 (K D = 1.38 ± 0.07 μM) and c-KIT2 (K D = 1.37 ± 0.06 μM) quadruplexes. These results are consistent with the FRET melting data that BTC f shows selectivity for the c-MYC compared to the c-KIT1 and c-KIT2 quadruplexes. It is worth mentioning that BTC f exhibited a strong blue shift of 25 nm with a 5.5 fold increase in fluorescence upon titration with the c-MYC in the absence of K + (Fig. 3b).
Next, the binding behaviour of BTC c with the c-MYC quadruplex was investigated ( Supplementary  Fig. S3a-b). The emission spectra of BTC c showed an emission maximum at 402 nm with a comparatively lower quantum yield (quantum yield, Φ = 0.044, abs, λ max = 256, 342 nm) along with two weak shoulders, when excited at 330 nm in 100 mM KCl and 100 mM Tris•HCl buffer, at pH 7.4. When the fluorimetric titration of BTC c (1 μM) was carried out with increasing concentration of the c-MYC (0-6equiv.), only a slight increase in fluorescence intensity was observed. The difference in the fluorescence spectra between BTC c and BTC f may be attributed to the presence of an extended π -electron delocalization between the lone pair of electrons on triazole nitrogen atom of the donor and the acceptor carbonyl group in BTC f. These results indicate that BTC f is an effective light-up probe for the c-MYC quadruplex. In good agreement with the FRET melting results, BTC f that exhibited a higher stabilization potential showed several fold higher binding affinity for the c-MYC (K D = 300 ± 15 nM) compared to BTC c (K D > 20 μ M).
In order to understand the observed blue shift upon interaction of BTC f with quadruplexes, the effect of environment polarity was analysed by recording the emission spectra of BTC f in various solvent environments such as ethyl acetate, DMF, DMSO, ethanol, methanol and water ( Supplementary Fig. S4a). The observed emissions (λ max em ) were plotted against Reichardt's E T (30) polarity parameter 34 ( Supplementary  Fig. S4b) to give a linear correlation (y = 4.03x + 267.3, r 2 = 0.9292) over the range of the solvents. The Stokes shift (S) was calculated from the maximum wavelength (λ max ) value of the excitation and emission spectra (Supplementary Table S4). Both λ max em and S values showed their strong dependence on the polarity of the medium. We observed that the emission maximum of BTC f is blue-shifted from 522 to 410 nm and the fluorescence intensity is simultaneously enhanced as the solvent polarity decreases from water to ethylacetate. These observations suggest that the enhancement in the fluorescence intensity along with the blue-shift in the emission spectra of the BTC f with the quadruplexes is a consequence of the proximity of BTC f to the non-polar hydrophobic regions 35 of the G-quadruplexes thereby manifesting the high affinity of BTC f towards the G-quadruplex structures.
BTC acquires an induced CD signal upon interaction with c-MYC G-quadruplex. The circular dichroism (CD) spectrum of c-MYC quadruplex sequence showed a positive peak at 260 nm and a negative peak around 240 nm, in the absence of K + , which is characteristic of a parallel conformation [36][37][38] ( Supplementary Fig. S5). As potassium stabilizes G-quadruplex structure, the molar ellipticity of both positive and negative peaks of c-MYC was enhanced in K + containing buffer ( Supplementary Fig. S5). The incremental addition of BTC f to the c-MYC resulted in a slight decrease in the positive band at 260 nm in the presence and absence of K + . This suggests that BTC f binds to the parallel quadruplex structure of c-MYC and does not disrupt the structure. In addition, BTC f exhibited a positive induced circular dichroism (ICD) in the absorbing 290-350 nm range (centered at 322 nm) in the presence and absence of K + . It is worth mentioning that, no such ICD signal was detected upon binding of BTC f to c-KIT 1 and c-KIT 2 quadruplexes (Supplementary Fig. S5c-d), which indicates that BTC f can discriminate between the promoter quadruplexes.
UV/Vis spectroscopy indicated that BTC f exhibits hypochromism upon addition of the c-MYC (0-0.4 equiv.) and the peaks at 260-290 nm overlap with the c-MYC at higher concentration ( Supplementary  Fig. S6). The observed hypochromism in UV spectra suggests strong stacking interaction between aromatic chromophores of BTC f with the G-quadruplex base pairs and indicates possible binding of these ligands into the nonpolar site of the c-MYC quadruplex. The red shifted ICD band at 322 nm in CD spectra may arise due to the binding of the optically inactive BTC f inside the asymmetric microenvironment of the c-MYC quadruplex 39 . CD spectroscopic analysis further reveals that BTC f bound c-MYC structure is similar in the presence and absence of K + .
Bis-triazolylcarbazoles selectively bind to a minor conformation of c-MYC quadruplex. NMR titrations show that the ligand BTC f is strongly interacting with the c-MYC sequence, as suggested by the significant changes of the imino pattern profile upon addition of the ligand (Fig. 4). Figure 4a shows that upon addition of BTC f, a set of minor peaks (already observable in the DNA free form and indicated with arrows) becomes more intense, while the chemical shift and the line width of the imino protons of the major conformation are not perturbed. At a [ligand]:[DNA] molar ratio of 2, the intensity of the peaks of the major conformation is reduced, compared to the intensity of the peaks of the ligand-bound conformation. We propose that BTC f binds to the c-MYC via conformational selection: out of an ensemble of DNA conformers in equilibrium with each other (major conformation- Fig. 4d and one or several minor conformations), the ligand selects the minor-populated conformer and binds to it in a specific manner. Interestingly, BTC f binds to the minor conformer, most likely a parallel stranded structure with different capping structures 13 , and upon ligand binding, the equilibrium of the free form DNA major /DNA minor conformational equilibrium is accordingly shifted. Moreover, upon ligand addition we also observe the appearance of new signals in the imino region (e.g., at 10.5 ppm) suggesting that the ligand BTC f can induce conformational changes in the structure of the binding-competent minor conformation. The proposed binding mechanism is summarized in Fig. 4c. The major conformation (M) is not binding-competent and the binding-competent minor conformation (m) undergoes small conformational changes upon ligand binding (m* Ligand ). NMR titration of BTC f in the absence of KCl (Fig. 4b) shows that the ligand bound conformation of c-MYC resembles the one observed in the presence of 100 mM KCl. These results support FRET melting, fluorescence and CD studies that BTC f can stabilize the c-MYC quadruplex in the absence and presence of potassium. NMR titration of c-MYC with ligand BTC c, which lacks the two cationic side chains showed line broadening with no significant chemical shift perturbation of the imino proton signals of the DNA major conformation ( Supplementary Fig. S7a). Similar to the [BTC f]:[c-MYC] complex, upon addition of BTC c, a new set of peaks was observed, which corresponds to one or more minor conformers of the c-MYC quadruplex. The NMR data suggest that these ligands interact with and stabilize a minor conformation of c-MYC quadruplex.

Bis-triazolyl carbazoles inhibit the growth of cancer cells.
To characterize the activity in biological systems, the antiproliferative activities of this class of compounds BTC a-g were determined in human hepatocellular liver carcinoma HepG2 cells using MTT assays 40 (Supplementary Fig. S8 and Table  S5). The results indicated that BTC f potently inhibits the growth of cancer cells at low micromolar concentrations with an IC 50 value of ∼ 4.3 ± 0.69 μM. BTC c (IC 50 = 11.38 ± 1.54 μM) displays a somewhat lower antiproliferative activity than BTC f. However, BTC d-e and BTC g do not show strong activity against HepG2 cells (Supplementary Table S5). We then evaluated the growth inhibitory activity of BTC f in two other human cancer cell lines such as breast carcinoma (MCF-7) and colon carcinoma (HCT  Fig. S8), but importantly it is less toxic to the model normal cell line (IC 50 > 45 μ M for C2C12, data not shown). It is worth noting that the main core carbazoledialkyne 5 does not show cytotoxicity towards normal as well as cancer cells (data not shown). These results indicate that the triazole ring and the side chains contribute towards the antiproliferative activities of the carbazole derivatives in cancer cells. Among the cancers cells (HepG2, MCF-7 and HCT 116), BTC f is the most potent in HepG2 cells. It has been reported that the expression c-MYC gene is elevated in liver carcinoma HepG2 cells compared to the normal human liver cells 2 . Biophysical analysis suggested that BTC f shows high specificity for the c-MYC G-quadruplex over duplex DNA. This encouraged us to evaluate the molecular mechanism underlying the anti-proliferative effect of BTC f in HepG2 cells. For a comparison, the effect of the less potent c-MYC quadruplex binding ligand BTC c with a reduced antiproliferative activity towards HepG2 cells was also investigated.

Bis-triazolylcarbazoles down-regulate c-MYC expression.
Ligands BTC c and BTC f were evaluated for their effect on transcriptional regulation of c-MYC in HepG2 cells (Fig. 5a). Total mRNA was isolated from HepG2 cells after treatment with varying concentrations (1.0, 2.5 and 5.0 μM) of BTC c and BTC f for 24 h. The level of c-MYC mRNA was quantified using quantitative real-time polymerase chain reaction (qRT-PCR) and the gene expression was normalized relative to the expression of a constitutively expressed house-keeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Analysis of the qRT-PCR data revealed that both BTC c and BTC f are able to reduce the level of c-MYC mRNA in a dose dependent manner (Fig. 5b). At 5.0 μ M concentration, BTC f reduces the level of c-MYC mRNA by 75 ± 3% relative to the control. Compared to that, BTC c at 5.0 μM leads to a suppression of c-MYC mRNA level by 48 ± 3% of the control suggesting its relatively lower efficacy than BTC f. The GAPDH mRNA is equally expressed in control, BTC c and BTC f treated HepG2 cells, which confirms that the reduction of mRNA level due to BTC c and BTC f is c-MYC gene-specific. To investigate that the observed reduction in c-MYC mRNA levels leads to an inhibition of MYC protein, we have employed western blotting using anti-MYC antibody (Fig. 5c,d). Protein levels were measured for the c-MYC and the housekeeping gene GAPDH in HepG2 cells after treatment with BTC c and BTC f for 24 h at final concentrations of 1.0, 2.5 and 5.0 μ M. Both BTC c and BTC f exhibit reduced expression of MYC protein in a dose dependent manner compared to the untreated HepG2 cells (Fig. 5c,d), which is in good agreement with the qRT-PCR analysis data. The densitometric analysis of the blots revealed that the decrease of MYC expression in cancer cells is in the range of 85 ± 3% and 52 ± 2% at 5.0 μM concentration of BTC f and BTC c, respectively. The concentration required for 50% inhibition in c-MYC expression ( ) IC 50 MYC is approximately 2 μM for BTC f and 4.6 μM for BTC c. Almost negligible reduction in GAPDH expression was observed in both treated and untreated cells. Together these results suggest that the BTC ligands can reduce the expression of MYC protein in HepG2 cancer cells at mRNA and protein levels (Fig. 5a).

BTC f inhibits cell cycle progression and induces apoptosis.
To determine whether BTC f mediated inhibition of HepG2 cell proliferation was associated with cell cycle arrest, we performed cell cycle analysis by propidiumiodide (PI) staining using flow cytometer. Flow cytometry analysis of BTC f (1.0-5.0 μM) treated HepG2 cells showed an increase in the SubG1 population (2.7% to 32.7%) with the increasing concentration of BTC f (Fig. 6a). Interestingly, at lower concentrations of BTC f, the cell population in G2 phase was increased (18.2% to 24.8%), and as the concentration of BTC f is increased, both S and G2 phase populations were decreased (7.3% to 3.9% and 18.2% to 14.8%, respectively) with a subsequent increase in the population of SubG1 phase.
To gain insight into the mechanism by which BTC f induced cell death of HepG2 cells, flow cytometry was employed to investigate the mode of cell death using Annexin-V and PI dual staining assay (Fig. 6b). HepG2 cells were incubated with BTC f (1.0-5.0 μM) for 24 h and the untreated cells were used as control. The flow cytometry analysis revealed that the apoptotic cell population was increased significantly (0.2% to 59.9%) in a dose-dependent manner. However, the extent of necrotic death was only 1.9% under the experimental conditions. This is in well agreement with the reduced activity of c-MYC in liver carcinoma cells after treatment with BTC f.  (Fig. 6d) establishes the localization of BTC f within the nucleus. These data suggest that BTC f is cell-permeable and binds to the cellular DNA. As BTC f shows higher affinity towards the G-quadruplex over duplex DNA, its localization in the nucleus area indicates that BTC f may induce apoptosis by stabilizing the G-quadruplex in HepG2 cells.

Discussion
A modular access to bis-triazolyl carbazole derivatives has been devised, where a simple carbazole dialkyne precursor 5 was prepared from the commercially available carbazole. The carbazole dialkyne 5 was treated with a variety of azides 6a-g using Cu(I) catalyzed azide-alkyne cycloaddition to preapre the corresponding bis-triazoloylcarbazole derivatives BTC a-g in high yields (Fig. 1). FRET melting analysis (Fig. 2) revealed that the carbzoledialkyne 5 does not change the melting temperature of any of the three investigated promoter quadruplexes (c-MYC, c-KIT1 and c-KIT2) or the duplex DNA, while the bis-triazoloyl carbazole derivative BTC f with two carboxamide side chains exhibits high stabilization potential for the quadruplexes over the duplex DNA ( Fig. 2 and Fig. 7). The diamino bis-triazolylcarbazole BTC c displayed moderate stabilization for the G-quadruplex sequences. Competitive FRET-melting experiments (Fig. 2b) clearly showed that the presence of 100 mol equivalent excess ds DNA did not significantly interfere with the stabilization of quadruplexes induced by the ligand BTC f, indicating its high selectivity for the quadruplexes over duplex DNA. FRET melting experiments at various concentrations of BTC f revealed that higher concentrations (5-7 fold) of BTC f are required to achieve the maximum stabilization potential (Δ T m ) for c-KIT1 and c-KIT2 quadruplexes compared to the c-MYC quadruplex. These results indicated that BTC f shows a preference for the c-MYC over the c-KIT1 and c-KIT2 quadruplexes. BTC f can also stabilize the quadruplexes to attain the maximum stabilization potential in the absence of K + ion. Despite showing high stabilization potential for the quadruplexes, BTC f did not exhibit any detectable stabilisation for the ds DNA.
Fluorescence studies showed that BTC f is a microenvironment-sensitive fluorescence "turn-on" sensor that selectively detects the c-MYC quadruplex over the duplex DNA with an enhancement in the fluorescence intensity with a blue shift (Fig. 3). However, the carbazole derivative BTC c in the presence of c-MYC neither triggered a significant fluorescence "turn-on" response nor a blue shift. The difference in fluorescence properties between BTC c and BTC f may be attributed to the lack of extended π -electron delocalization in BTC c. In agreement with the FRET melting results, the fluorescence spectroscopy studies showed high specificity of BTC f for the quadruplexes over duplex DNA, and in particular, BTC f showed a 5-fold preference for c-MYC (K D = 300 nM) over c-KIT1 and c-KIT2 promoter quadruplexes. Moreover, the confocal laser scanning microscopy (CLSM) shows that BTC f can be successfully applied to visualize the live-cell nucleus of HepG2 cells with high selectivity (Fig. 6). NMR and Circular Dichroism (CD) spectroscopic analyses were employed to gain insight into the structural basis for the recognition of BTC c and BTC f to c-MYC quadruplex. CD binding titrations revealed that the binding of BTC f to c-MYC gives rise to a CD signal (ICD peak) from the bound ligand ( Supplementary Fig. S5a-b). No such ICD bands are observed upon interaction of BTC f with the c-KIT1 and c-KIT2 quadruplexes, which indicates that BTC f binds to the c-MYC in an asymmetric fashion. It has been reported that multiple conformations of 3′ -and 5′ -flanking capping structures of c-MYC quadruplex co-exist at equilibrium 13 . NMR analyses (Fig. 4) suggested that BTC f selects and binds specifically to one of these minor conformations of the c-MYC via conformational selection. The [BTC f]:[c-MYC] interaction process is similar with or without K + . The binding of BTC f could also induce a conformational change in the binding-competent minor conformation of the c-MYC, which may be useful to specifically alter the biological function of the c-MYC.
Among the synthesized carbazole analogues, compound BTC f was found to be the most potent molecule that inhibited the growth of human hepatocellular liver carcinoma HepG2 cells (which highly expresses MYC) at a low micromolar concentration (IC 50 = 4.3 ± 0.69 μ M) without affecting the normal mouse myoblast C2C12 cells. The less potent quadruplex stabilizer BTC c also showed good antiproliferative activity with an IC 50 value of 11.38 ± 1.54 μ M in HepG2 cells. It is worth mentioning that the starting carbazoledialkyne 5 neither stabilized the G-quadruplex DNA (FRET melting data) nor effectively inhibited the growth of cancer cells (Fig. 7). Since BTC f exhibits high specificity for the c-MYC quadruplex and the c-MYC gene is overexpressed in liver carcinoma HepG2 cells, the effect of BTC f and the less potent BTC c on transcriptional regulation c-MYC gene was studied in HepG2 cells (Fig. 5). BTC f was able to inhibit c-MYC expression in both transcriptional and translational level as suggested by qRT-PCR and Western blotting analysis (Fig. 5). Analysis of the cell cycle data revealed that cells were unable to transit from G1 phase to S phase with increasing concentration of BTC f. The increase in concentration of BTC f triggered cell cycle arrest in SubG1 phase and subsequent apoptosis in cancer cells, probably by down-regulating c-MYC gene expression [41][42][43] as suggested by qRT-PCR and Western blot analysis (Figs 5 and 6). Our results collectively suggest that the reduction in c-MYC expression is probably due to the binding of BTC f to the c-MYC promoter-quadruplex. However, the exact molecular mechanism of BTC f mediated c-MYC down-regulation 44 is currently under investigation.
This work highlights that a simple synthetic protocol can be devised to synthesize fluorescent bis-triazolylcarbazole derivatives, which can effectively inhibit c-MYC transcriptional activity (Fig. 7). The carbazole derivative with carboxamide side chains shows pronounced environment-sensitive fluorescence and selectively detects the c-MYC G-quadruplex over the duplex DNA via "turn-on" fluorescence. The ligand binds to the c-MYC with nM binding affinity and shows a five-fold preference for the c-MYC over the c-KIT promoter quadruplexes. The ligand can bind and stabilize the c-MYC quadruplex in the presence and absence of added K + . It is intriguing that the ligand can be used to 'programme' the c-MYC to adopt a specific conformation, which is stable with or without K + . The specific nucleic acid structure may find applications in the fields of nanobiotechnology and biomedical technology. This small molecule probe is an attractive probe for bio-imaging as it can rapidly and selectively stain the nucleus in living cells. Further this small molecule can induce cell cycle arrest and promote cancer cell death by apoptosis. These results collectively suggest that the carbazole derivative is a potent anticancer agent and a viable lead for further development of anticancer drugs.

Synthetic Protocols. Synthesis of 3-(3,6-diiodo-9H-carbazol-9-yl)-N,N-dimethylpropan-1-amine 3.
A mixture of carbazole 1 (15.0 g, 89.82 mmol), KI (19.36 g, 116.62 mmol), KIO 3 (19.20 g, 89.82 mmol) in acetic acid (100 mL) and deionized water (10 mL) was stirred at 80 °C for 48 h under N 2 atmosphere. After cooling to room temperature, the mixture was filtered, washed with deionized water and saturated Na 2 CO 3 solution to afford 3, 6-diiodocarbazole as a colorless solid (24.4 g, 65%). 1  A mixture of 3, 6-diiodocarbazole (500 mg, 1.193 mmol) and NaH (114.24 mg, 4.77 mmol) in 20 mL THF was stirred at room temperature for 2 h under nitrogen atmosphere. In another flask, 3-dimethylaminopropyl chloride hydrochloride 2 (435.33 mg, 3.58 mmol) and NaOH (1.0 g) were diluted in 5 mL water and cooled to room temperature, then the upper layer of this solution was added drop-wise to the above mixture. The reaction mixture was refluxed for an additional 24 h. After the removal of the solvent under reduced pressure, the residue was purified by column chromatography on silica gel using DCM/methanol (20:1) as eluent. Re-crystallization from ethanol gave the compound 3 (529 mg, 88%) as a colorless solid, mp 129-131 °C. 1  atmosphere for 12 h, concentrated, washed with brine and dried over anhydrous Na 2 SO 4 . The crude product was purified by column chromatography. Then the resulting crude alcohol was refluxed with 5 equiv. KOH in toluene under an argon atmosphere for 12 h. The reaction mixture was concentrated, washed with brine and dried over anhydrous Na 2 SO 4 . The crude product was purified by column chromatography to give the dialkyne 5 (161 mg, 90% yield) as a yellow liquid. 1  General procedure for the synthesis of bis-triazole derivatives BTC a-g using click chemistry (GP1). Dialkyne 5 (50 mg, 0.166 mmol) was dissolved in a 1:2 mixture of t-BuOH/H 2 O (4 mL). Copper(II) sulphate pentahydrate (4.14 mg, 0.0166 mmol) and sodium ascorbate (3.2 mg, 0.0166 mmol) were added and the solution was stirred for 10 min. The corresponding azide 6 (2.5 × 0.166 mol equiv.) was added and the mixture was then heated for 4 h at 70 °C under microwave irradiation. After cooling to room temperature, the reaction mixture was concentrated. The crude product was purified by flash column chromatography (from CH 2 Cl 2 (100%) to CH 2 Cl 2 /MeOH (10:1) to CH 2 Cl 2 /MeOH/NH 4 OH (10:1:0.5) to give the corresponding bis-traizole derivatives BTC a-g.  Fig. 4d) 13 . The proton resonance assignment of the c-MYC DNA in 90% H 2 O/10% D 2 O was performed on the basis of the assignment reported in literature. 13   The PCR mixture (25 μL) contained 15 pmol of each primer, 7 μL of water, 5 μ L of cDNA, 12.5 μL 2X JumpStart Taq ReadyMix. The samples were placed in 96-well plates (Roche), and PCR amplification was performed using Light Cycler 480 II real-time PCR detection system (Roche). The thermal cycling conditions were 2 min at 94 °C and then 40 cycles of 15 s at 94 °C, followed 60 s at 60 °C. We used the comparative cycle threshold method (C T method) for relative quantification of gene expression 49 . The C T for the target and the C T for the internal control (GAPDH) were determined for BTC c or BTC f or untreated (control) samples. Finally, the arithmetic calibrator (2 −ΔΔCT ) was used to calculate the relative mRNA level expression of c-MYC. Difference in c-MYC expression was expressed as fold changes.
Western blot analysis. Hepatocellular carcinoma HepG2 cells were treated with different concentration (1.0, 2.5 and 5.0 μM) of BTC c and BTC f for 24 h at 37 °C in humidified CO 2 incubator. After the incubation period, cells were washed once with PBS (pH 7.4) and lysed with cold cell lysis buffer (20 mM Tris, 100 mM NaCl, 1 mM EDTA in 0.5% Triton X-100). Cell lysate were collected from the treated and untreated cells, and the total protein content was estimated by Lowry method 50 . Equal amount of proteins (60 μg) from the cell lysate were separated by 10% SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked, washed and probed using antibodies directed against c-MYC and GAPDH (as loading control) for overnight at room temperature. The blots were washed and immunoreactive bands were incubated with a 1:2000 dilution of ALKP conjugated secondary antibody for 2 h at room temperature. Binding signals were visualized with NBD/BCIP substrate. Relative band intensities were determined by using ImageJ software.
Flow cytometric determination of the cell cycle histogram by PI staining. Cell cycle histogram was analysed using propidium iodide (PI) staining by Flowcytometry. HepG2 cells (1 × 10 6 ) per 60 mm petridish (~80% confluence) were treated with various concentration of BTC f (1.0, 2.5, 5.0 μM) for 24 h in fresh growth medium. Cells were harvested by trypsinization, resuspended in PBS and fixed with 2 mL of ice-cold 70% ethanol for overnight at 4 °C. The pellets were collected by centrifugation and resuspended in PBS solution, containing 10 μg/mL PI (Sigma) and 10 μg/mL RNaseA (Sigma). After incubation for 30 minutes in the dark at 37 °C, cells were analyzed for DNA content using a FACS flow cytometer (BD Biosciences). Cell distribution among cell cycle phases and the percentage of apoptotic cells were evaluated using Cell-Quest Pro software (BD).
Flowcytometric assay of apoptosis. Annexin V-FITC and propidium iodide (PI) stains were used to determine the percentage of cells undergoing apoptosis and necrosis. An apoptosis assay was conducted using the protocol supplied by the manufacturer. Briefly, 1 × 10 6 HepG2 cells per 60 mm petridish (~90% confluence) were treated with different concentration of BTC f (1.0, 2.5, 5.0 μ M) for 24 h in fresh growth medium. Cells were then harvested with trypsinization. After centrifugation at 700 rpm for 5 min at 4 °C, cell pellet was suspended in 500 μ L 1 × binding buffer and then treated with 5 μ L Annexin V-FITC and 5 μ L PI. After incubation for 5 min on ice, each sample was analysed immediately using fluorescenceactivated cell sorting (FACS) analysis (BD Biosciences, Mountain View, CA, USA). Approximately 10,000 HepG2 cells were detected for each sample. Cytogram analysis was done using the Cell Quest software.
Confocal microscopy. Cellular localization of BTC f was monitored by live cell imaging. HepG2 cells (1 × 10 3 ) were cultured on 35 mm diameter glass-bottomed cover slips for 24 h followed by incubation with BTC f (5.0 μ M) for 4 h inside CO 2 (5%) incubator at 37 °C. After incubation, cells were washed with PBS three times to remove the excess ligand and bathed in DMEM (2 mL) before imaging. Localization of BTC f was viewed under confocal microscope (Zeiss LSM 510 laser scanning microscope, Standort Göttingen, Germany). At least 5 fields per slide and three independent sets were examined. Statistical analysis. All data were given as mean ± S.D. Differences between two groups were compared by unpaired Student's t-test. For multi-group comparisons, analysis of variance was determined by ANOVA. A value of P < 0.05 was considered statistically significant. The statistical analysis was done by using GraphPad Instant Software (Graph-Pad, La Jolla, CA, USA).