Ellipticine was obtained from Santa Cruz Biotechnology (SC-200878; Santa Cruz, CA, USA). NSC311153 and other structural analogs were kindly provided from the U.S. NCI/DTP Open Chemical Repository. All the compounds were dissolved in dimethyl sulfoxide (DMSO) at a final concentration of 10 mg/ml.

The 5′-FAM labelled RET-WT (5′-AGCGGGTAGGGGCGGGGCGGGGCGGGGGCGG-3′) oligonucleotide was purchased from Sigma Genosys (Woodlands, TX, USA). Taq DNA polymerase was purchased from Fermentas (Hanover, MD, USA).

The TT cell line was obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) and maintained in Dulbecco's modified Eagle's medium (DMEM)/F-12 medium (Cellgro, Manassas, VA, USA) supplemented with 15% heat inactivated fetal bovine serum (FBS). The TPC1 cells and another MTC derived cell line, MZ-CRC-1 cells were provided by Dr Rebecca Schweppe (University of Colorado, Denver, CO, USA). These cells were maintained in RPMI-1640 medium supplemented with 9% FBS and DMEM/F-12 medium supplemented with 15% FBS, respectively. The isogenic cell line HEK293-RET, which carries the luciferase reporter gene under the control of RET gene promoter was generated as described in our previous study and grown in DMEM medium supplemented with 9% FBS (23). The normal thyroid cell line, Nthy-ori-3-1 was purchased from Sigma Genosys and cultured in RPMI-1640 medium supplemented with 9% FBS. All the cell lines were maintained in a humidified atmosphere containing 5% CO₂ at 37°C. The stocks for all these cell lines were obtained from the cell bank and utilized within 6 months. These cell lines were also tested for mycoplasma contamination and were further authenticated using STR profiling.

The RET-WT oligonucleotide (5 µM) was allowed to form the G-quadruplex structure in the presence of Tris-HCl buffer (20 mM, pH 7.6) and 25 mM KCl by denaturing at 95°C for 5 min and slowly cooled to room temperature. The CD spectra were recorded using a Jasco J-810 spectrophotometer (Jasco, Inc., Easton, MD, USA) using a quartz cell of 1 mm path length and instrument scanning speed of 100 nm/min with a response time of 1 sec over a wavelength range of 230-330 nm as previously described (23). The T_m was determined by monitoring the molar ellipticity vs. temperature profiles at 262 nm at increasing temperature from 20 to 90°C at a gradient of 1°C/min.

The polymerase stop assay was performed on a DNA template containing the G-quadruplex forming sequence, which is present in the RET promoter region as previously described (39). In brief, the template DNA was annealed with a 5′-γ[³²P] end-labelled primer (P28) and purified on an 8% non-denaturing polyacrylamide gel. The purified primer-DNA template was used in a primer extension assay in the presence of Taq DNA polymerase.

The DMS footprinting was performed on the 5′FAM labelled RET-WT oligonucle-otides as described in our previous studies (23,24). In brief, the oligonucleotides were allowed to form the G-quadruplex structure and were incubated in the absence and presence of NSC311153 (5 equivalents) at room temperature for 1 h. The samples were treated with DMS (0.2%) for 2 min and were resolved on an 8% non-denaturing polyacrylamide gel. Each DNA band was recovered from the gel and treated with piperidine (10%) after ethanol precipitation. The cleaved products were resolved on a 16% denaturing polyacrylamide gel along with purine and pyrimidine specific sequencing markers that were generated according to the published procedure (40). The gel was dried and the fluorescence was read on a Typhoon 8600 scanner (GE Healthcare Life Sciences, Pittsburgh, PA, USA) for analysis.

Total RNA was extracted from the cells using the RNeasy Mini QIAcube kit (Qiagen, Redwood City, CA, USA) according to the manufacturer's protocol. The extracted RNA was subjected to reverse-transcription using the oligo (dT)18 primer with QuantiTect reverse-transcription kit (Qiagen) to generate single-stranded cDNA. The primers used for RT-PCR were as follows: RET forward, (5′-GCAGCATTGTTGGGGGACA-3′) and RET reverse, (5′-CACCGGAAGAGGAGTAGCTG-3′); Rpl9 forward, (5′-CTGAAGGGACGCACAGTTAT-3′) and Rpl9 reverse, (5′-ACGGTAGCCAGTTCCTTTCT-3′). The PCR reactions involved an initial denaturation at 95°C for 3 min followed by 33 and 23 cycles for RET and Rpl9, respectively, at 95°C for 30 sec, 52°C for 30 sec and 72°C for 30 sec on a GeneAmp PCR system 9600 (Perkin-Elmer, Waltham, MA, USA). The PCR products were analyzed on 1.5% agarose gel electrophoresis.

The whole-cell protein extracts were prepared by lysing the cells with 2% CHAPS lysis buffer in the presence of 10 mM Tris-HCl, pH 7.4, 0.15 M NaCl, 5 mM EDTA and Halt Protease Inhibitor Cocktail (Thermo Fisher Scientific, Waltham, MA, USA). The extracted proteins were resolved on a 4-12% polyacrylamide SDS-PAGE, as previously described (23,24). The primary antibodies used were as follows: anti-RET (#3220), anti-RET/PTC1 (#14698), anti-cMYC (#5605), anti-p-mTOR (#2971) and anti-mTOR (#2972) (dilution 1:1,000) were purchased from Cell Signaling Technology (Beverly, MA, USA), anti-Bcl-2 (sc-7382), anti-pERK (sc-7383), anti-ERK (sc-271270), anti-cyclin D1 (sc-20044) and anti-β-actin (sc-47778) (dilution 1:300) were purchased from Santa Cruz Biotechnology. Anti-VEGF antibody was purchased from GeneTex (Irvine, CA, USA; #GTX102643). Mouse and rabbit IgG antibodies tagged with horseradish peroxidase (HRP) (Bio-Rad Laboratories, Hercules, CA, USA) were used as secondary antibodies (dilution 1:1,000). An enhanced chemiluminescence substrate kit (#32106) purchased from Thermo Fischer Scientific was used for detection.

The isogenic cell line HEK293-RET was exposed to different concentrations of NSC311153 up to 24 h. Luciferase expression level is determined using the ONE-Glo Luciferase Assay system (Promega, Madison, WI, USA) following the manufacturer's instruction.

Cells were plated at a concentration of 7,500 cells/well in a 96-well dish and incubated overnight, followed by the treatment with NSC311153 at increasing concentrations up to 96 h. The cell viability was determined by using 0.33 mg/ml MTS dye in the presence of phenazine methosulfate (PMS) (25 µM) as previously described (41). The absorbance was measured at 590 nm using a Synergy HT multi-detection microplate reader (BioTek Instruments, Inc., Winooski, VT, USA).

TT cells were treated with different concentrations of NSC311153 up to 48 h and caspase-3 activity was measured using the ApoAlert Caspase Fluorescent Assay kit (Clontech Laboratories, Inc., Mountain View, CA, USA) by following the manufacturer's protocol.

The in vivo antitumor efficacy of the ellip-ticine derivative was evaluated using 8-10-week old male severe combined immunodeficiency (SCID) mice, xenotrans-planted with MTC derived TT cells. Animal experiments were conducted in accordance with the Institutional Animal Care and Use Committee (IACUC). The experiments were performed in the Experimental Mouse Shared Resource (EMSR) Animal Facility Laboratory (University of Arizona), which is accredited by the International Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). In brief, exponentially growing TT cells (1×10⁷) were subcutaneously injected into the flank of the mice and the tumor growth was monitored every week by measuring the tumor diameters using a vernier caliper. Tumor volume was calculated according to the formula (b²×l)/2 where b and l are the shortest and the longest diameters, respectively. Once the tumor volume reaches 100 mm³, the mice are randomly pair matched to vehicle and treated group (n=8/group). The compound was dissolved in 90% phosphate-buffered saline (PBS) and 10% DMSO and administered intraperitoneally (i.p) at a single dose of 4 mg/kg for 5 days/week up to 2 weeks. The antitumor efficacy was assessed based on the percentage inhibition of tumor growth in treated vs. control group. The toxicity of the compound was evaluated based on the loss of average weight of mice. Tumor tissue from vehicle and treated group were explanted at the end of last dosage to examine the effect of the compound on the expression of proteins that contribute to tumor growth.

The molecular structure of ellipticine consists of an aromatic pyridocarbazole ring, which allows the π-π stacking interactions with the planar guanine residues present in the terminal G-tetrad (Fig. 1) (37). Not surprisingly, previous studies have reported that ellipticine stabilizes different G-quadruplex structures formed on the telomere region and on the promoter region of c-Myc oncogene (37,38). Hence, in the present study we investigated whether this compound binds and stabilizes the RET G-quadruplex structure using CD spectroscopic analysis. First, we examined whether ellipticine alters the structural conformation of the RET G-quadruplex by monitoring the CD spectra of RET G4 sequence (5 µM) in the absence and presence of increasing concentrations of this compound. As shown in Fig. 2A, the positive peak at 262 nm, which corresponds to a parallel G-quadruplex structure (42), was not affected in the presence of ellipticine, suggesting that the parallel configuration of the RET G-quadruplex was not changed in the presence of this molecule. Next, we examined the thermal stability of the G-quadruplex structure by monitoring the CD melting curve in the absence and presence of ellipticine (1 equivalent) at increasing temperatures. As shown in Fig. 2B, the melting temperature (T_m) of the RET G-quadruplex structure in the absence of ellipticine was determined to be 75°C. Based on the melting curves, the binding of ellipticine with the G-quadruplex structure increases its melting temperature up to 85°C and the ∆T_m was found to be 10°C (Fig. 2B). To further understand the structural features required for the interaction of ellipticine with the RET G-quadruplex structure, we examined the binding of 2-hydroxycarbazole with this structure by monitoring the CD melting profiles. As shown in Fig. 2C the T_m of the G-quadruplex structure was not increased even in the presence of 5 equivalents of 2-hydroxycarbazole, suggesting that the presence of pyridine ring in ellipticine allows additional π-π interactions with the guanine residue. Overall these data imply that the pyridocarbazole ring moiety is essential for the stabilization of the RET G-quadruplex structure by ellipticine.

In our previous studies we have clearly demonstrated that the small-molecule mediated stabilization of the G-quadruplex structure formed on the RET promoter region exerts transcriptional inhibitory effect on this gene (23,24). Hence, we determined whether ellipticine downregulates the endogenous RET expression in the TT cell line in which the RET gene transcription is regulated by the promoter region that contains the G-quadruplex forming sequence. As shown in Fig. 2D, a concentration dependent decrease in the RET protein expression was observed in the presence of ellipticine following 48-h exposure. To further confirm that the RET downregulation by ellipticine is mediated through the stabilization of the G-quadruplex structure, we determined the effect of 2-hydroxycarbazole on the RET expression in this cell line. As shown in Fig. 2E the RET protein expression was not decreased in the presence of this compound even at high concentration (5 µg/ml). These data are consistent with our previous studies that the stabilization of the G-quadruplex structure formed within the RET promoter region is responsible for silencing the expression of this gene.

Although we identified ellipticine as a potent RET transcriptional inhibitor, this compound was withdrawn from this study due to its adverse cytotoxic effects in TT cells, which was evaluated by MTS assay. The IC₅₀ of ellipticine was calculated to be 0.3 µg/ml after 96-h treatment of TT cells in the presence of increasing concentration of this compound (Fig. 3A). Hence, we attempted to identify other structural analogs of ellipticine from the NCI/DPT open chemicals repository that could suppress the RET gene transcription at non-toxic concentrations without affecting their ability to stabilize the RET G-quadruplex structure (Figs. 1 and 3A). Based on the CD melting curves the ∆T_m was calculated individually for the ellipticine analogs and the IC₅₀ values for these compounds were also determined by MTS assay (Fig. 3A). The ellipticine derivatives, which have different substituents such as methoxy, chloride and methyl groups at position C-9 of the pyridocarbazole ring stabilizes the RET G-quadruplex structure by increasing the T_m >10°C (Fig. 3A). These compounds also possess RET inhibitory effects in TT cells, which is comparable to their parent molecule, ellipticine (Fig. 3B). However, the structural modifications in these compounds did not improve the cytotoxicity in TT cells as compared to ellipticine (Fig. 3A). Notably, one of the ellip-ticine analogs, NSC311153, which carries a hydroxyl group and a 2-piperidin-1-ylethy moiety at positions C-9 and N-2, respectively showed an increased binding ability with the RET G-quadruplex structure with an estimated ∆T_m as 15°C (Fig. 3A). The IC₅₀ value of this compound was also estimated to be 2.5 µg/ml after 96 h of exposure, which is significantly higher than that of ellipticine (Fig. 3A). Based on these data, we selected NSC311153 as a lead compound to further proceed with the in vitro studies.

To further validate that NSC311153 stabilizes the RET G-quadruplex structure formed on the promoter region of this gene, a DNA polymerase stop assay was performed as previously described (23). In this assay, the ligand mediated stabilization of the G-quadruplex structure that arises on the DNA template prevents the progression of the Taq DNA polymerase during primer extension. As shown in Fig. 4A, in the presence of NSC311153 at increasing concentrations, a dose-dependent increase in the amount of arrested product was observed, indicating the potential stabilization of the G-quadruplex structure by this compound.

Next, we investigated whether the interaction of NSC311153 with the parallel RET G-quadruplex structure changes the guanine residues that are involved in the G-tetrad formation using dimethyl sulfate (DMS) footprinting. DMS footprinting is a well-established technique to determine the guanine nucleotides that are involved in the formation of G-quadruplex structures (40,43). The N7 position of each of the guanine residues that are involved in Hoogsteen base pairing to form G-tetrads is inaccessible to methylation by DMS, which attacks this position. As shown in Fig. 4B, the pattern of N7-guanine methylation produced by the RETG4 sequence in the presence of 100 mM K⁺ is consistent with two parallel G-quadruplexes (lane C) in which either the guanines (G19-G21) or (G20-G22) is involved in the G-tetrad formation (Fig. 4C, models 1 and 2). As shown in Fig. 4B, the binding of NSC311153 (5 equivalents) changes the pattern of N7 guanine methylation (lane D) in which the guanines (G14-G16) or (G15-G17) is involved in the G-tetrad formation (Fig. 4C, models 1 and 3). The change in the G-quadruplex structure in the presence of this compound clearly suggest that NSC311153 binds with this secondary structure.

Since in our previous studies we reported that the G-quadruplex structure formed on the promoter region of RET gene acts as transcriptional silencer element, we next examined whether the stabilization of RET G-quadruplex structure by NSC311153 interferes with the transcriptional activation of this gene in TT cell line. As shown in Fig. 5A, NSC311153 decreased the RET mRNA expression by >50 and 90% at a non-toxic concentration of 2.5 µg/ml following the exposure up to 24 and 48 h, respectively. We also utilized the MZ-CRC-1 cell line, which harbors an M918T mutation in the tyrosine kinase domain of the RET protein. Moreover, the RET gene expression in this cell line is regulated by the same promoter region as found in the TT cells, which harbors the G-quadruplex forming motif (44). As shown in Fig. 5B, a dose-dependent decrease in the RET protein expression was observed in the presence of NSC311153 after 48-h incubation.

To determine whether the decrease in RET mRNA expression by NSC311153 is a direct effect of the promoter-specific transcriptional inhibition of this gene, a bioluminescent reporter gene assay was performed using an isogenic cell line HEK293-RET in which the luciferase expression is under the control of the RET promoter region as described in our previous study (23). As shown in Fig. 5C, in the presence of NSC311153 a dose-dependent decrease in the basal luciferase expression was observed in the HEK293-RET cell line following 48-h incubation. To further demonstrate that the mechanism of action of NSC311153 is through stabilizing the G-quadruplex structure present on the RET promoter region, we utilized PTC1 derived TPC1 cells as a control. Chromosomal rearrangement between the RET kinase domain coding region and the CCD6 gene results in a chimeric RET/PTC1 expression, whose transcriptional activation is controlled by the CCD6 promoter region that lacks the G-quadruplex forming motif (45-49). Notably, NSC311153 did not suppress the RET/PTC1 expression in this cell line even after 48-h incubation (Fig. 5D). Overall, these data provide clear evidence to support that NSC311153 intervenes in the transcription mechanism of RET gene by targeting the intracellular G-quadruplex structure formed on its promoter region.

The oncogenic RET activation is mainly involved in mediating the cell proliferation and survival and hence the effect of NSC311153 on the viability of TT cell line was determined using the MTS assay. As shown in Fig. 6A the proliferation of the TT cells decreased with increasing concentrations of NSC311153 and the IC₅₀ was found to be ~2.5 µg/ml after 96-h exposure. The TPC1 cell line in which the RET/PTC1 expression was not inhibited by NSC311153 was less sensitive to this compound and the IC₅₀ was 10-fold more than the TT cells (Fig. 6A). This clearly shows that the inhibition of cell growth by NSC311153 is mediated through the downregulation of the RET gene. To further determine whether the anti-proliferative effect of NSC311153 is cancer cell-specific, we included a normal thyroid cell line, Nthy-ori-3-1 in the present study. Based on the MTS data, the IC₅₀ of NSC311153 in the Nthy-ori-3-1 cell line was estimated to be 10 µg/ml (Fig. 6A), which is 4-fold higher than that of TT cells, suggesting that this compound is selectively sensitive to the mutant RET driven thyroid cancer.

To further characterize the mechanism through which NSC311153 suppresses the TT cell proliferation, we investigated whether this compound inhibits the RET mediated downstream signal transduction pathways. Previous studies have revealed that RET activates the Raf/MEK and PI3K/Akt downstream signaling pathways, which in turn phosphorylate and activate ERK1/2 and mammalian target of rapamycin (mTOR) proteins, respectively (10-12). As shown in Fig. 6B, the phosphorylation status of ERK1/2 and mTOR were decreased in a dose-dependent manner following the exposure of TT cells in the presence of NSC311153 up to 48 h. However, in TPC1 cells the phosphorylation levels of these proteins were not altered by NSC311153 (Fig. 6C). These data clearly suggest that the inhibitory effect of NSC311153 on this pathway is a consequence of RET downregulation.

The Raf/MEK/ERK and PI3K/Akt/mTOR pathways are known to promote cancer cell survival through the activation of cyclin D1 and Bcl-2, which are involved in enhancing cell-cycle progression and inhibiting the apoptosis mechanism, respectively (50-52). As shown in Fig. 6D, NSC311153 also decreased the expression of cyclin D1 and Bcl-2 in TT cells in a concentration-dependent manner. Furthermore, the down-regulation of Bcl-2 by NSC311153 is further accompanied by the increase in the caspase-3 activity, which is also a known indicator of apoptosis (Fig. 6E).

In the final step of the present study, we evaluated the in vivo effect of the ellipticine derivative on the tumor growth of MTC xenografts through subcutaneous injection of the TT cells into SCID mice. Although NSC311153 showed potential anti-proliferative effects in TT cells in vitro, the poor solubility of this compound hindered its in vivo antitumor examination. However, a previous study reported that a water soluble ellipticine analog, NSC311152 (datelliptium) was well tolerated in different cancer patients in a phase-I clinical trial (53). As shown in Fig. 1, the structure of NSC311152 is very similar to NSC311153 with the presence of a diethylaminoethyl-moiety at position N-2 that improves the solubility of this compound compared to that of NSC311153. Based on our preliminary data, the inhibitory effect of NSC311152 on RET expression and its IC₅₀ value is the same as that of NSC311153, suggesting that the slight structural modification does not alter the efficacy of this compound. Hence, we decided to pursue with NSC311152 to test the in vivo antitumor activity. Dosing regimen and treatment schedules were determined based on our preliminary dose optimization studies. Upon systemic administration via i.p, NSC311152 (4 mg/kg) was well tolerated without any significant decrease in the average body weight (Fig. 7A). As shown in Fig. 7B, ~60% inhibition of the tumor growth was observed in mice treated with NSC311152 compared to the vehicle treated group. To further validate whether the inhibition of tumor growth by NSC311152 in vivo is mediated through target specific effect, we determined the RET expression in tumor tissues of vehicle- and NSC311152-treated mice using western blotting. As shown in Fig. 7C, the RET protein expression was decreased in the presence of NSC311152 compared to that of vehicle. Furthermore, NSC311152 reduced the expression of other proteins such as c-Myc, Bcl-2 and cyclin D1 that contribute to cell proliferation, which is in accordance with the in vitro studies. Overall, these data suggest that NSC311152 possesses potent in vivo antitumor activity through RET downregulation.