Quinazoline Ligands Induce Cancer Cell Death through Selective STAT3 Inhibition and G-Quadruplex Stabilization

The signal transducer and activator of transcription 3 (STAT3) protein is a master regulator of most key hallmarks and enablers of cancer, including cell proliferation and the response to DNA damage. G-Quadruplex (G4) structures are four-stranded noncanonical DNA structures enriched at telomeres and oncogenes’ promoters. In cancer cells, stabilization of G4 DNAs leads to replication stress and DNA damage accumulation and is therefore considered a promising target for oncotherapy. Here, we designed and synthesized novel quinazoline-based compounds that simultaneously and selectively affect these two well-recognized cancer targets, G4 DNA structures and the STAT3 protein. Using a combination of in vitro assays, NMR, and molecular dynamics simulations, we show that these small, uncharged compounds not only bind to the STAT3 protein but also stabilize G4 structures. In human cultured cells, the compounds inhibit phosphorylation-dependent activation of STAT3 without affecting the antiapoptotic factor STAT1 and cause increased formation of G4 structures, as revealed by the use of a G4 DNA-specific antibody. As a result, treated cells show slower DNA replication, DNA damage checkpoint activation, and an increased apoptotic rate. Importantly, cancer cells are more sensitive to these molecules compared to noncancerous cell lines. This is the first report of a promising class of compounds that not only targets the DNA damage cancer response machinery but also simultaneously inhibits the STAT3-induced cancer cell proliferation, demonstrating a novel approach in cancer therapy.

. Primary screening of the synthesized compounds in a Taq-polymerase stop assay using S. pombe with (a) parallel ribosomal G4 DNA, (b) hybrid telomeric G4 DNA, (c) antiparallel cdc13 + promoter G4 DNA, and (d) nonG4 DNA as a control. All graphs represent the mean of two independent experiments ± absolute error. Original representative sequencing gels for a-d are shown in Figures S1 and S2. 25 µM compound was used, and the names of the compounds are indicated below each graph. Arrows indicate the compounds selected for further study.  Figure S6. Dose response of Taq-polymerase stop assay using 5b as the treatment. The concentration of 5b is indicated above each well.      Table S2.      Table S1: Binding energies and simulation times of the clusters obtained from MD simulations. The average binding energy (kJ/mol) between 4f and c-MYC Pu24T was calculated for the first 50 frames from each cluster with the MM/PBSA method using the g_mmpbsa tool. Standard errors were calculated using the block-averaging method.

Cluster
No.

Simulation
Time (  Supplementary table S3. Compound permeability was measured using a Caco-2 cell assay which indicate that both compounds are able to pass the cell membranes. The cut-off Papp values used corresponds to the fraction absorbed in the gut of <20% (Papp value of 0.2 x10-6 cm/s for low permeable compounds), and >80% (Papp of 1.6 x 10-6 cm/s for high permeable compounds). Thus, 4f has a high permeability and 8g a medium permeability and a high efflux ratio. Therefore, 8g will most likely have limited absorption after oral administration and further optimization of this property may be needed. Compound

Compounds synthesis
To improve the hit compounds ability to bind and stabilize G4 DNA structures and to understand which factors that control selectivity and potency, we designed synthetic routes to broadly explore the compounds structure-activity and structure-selectivity relationships.
This resulted in a library of forty-seven derivatives that are all based on the initial hit compound 5b, as outlined in Figure 1 and scheme 1-4. The key intermediates 3a-h were synthesized form commercially available substituted anilines (1a-h) in two steps. In the first step, a modified Skraup synthesis 1 was used to generate the substituted 2,2,4-trimethyl-1,2-dihydroquinolines (2a-h) in 53-78 % yield. In the second step, the 1,2-dihydroquinolines (2a-h) were reacted with 2-cyanoguanidine to give N- treatment of the intermediates 3a-h with acetylacetone yielded the desired quinazolinepyrimidine derivatives 4(a, c-f) in 50-65% yield (Scheme 1). Intermediate 3(b, g-h) was not compatible with this method and was therefore synthesized using a different approach starting with the synthesis of 4,6-dimethyl-pyrimidin-2-yl-cyanamide (6)  S28 yield (Scheme 2). In addition to these derivatives, the condensation of 3a-b with mesityl oxide in DMSO at 100 °C also yielded quinazoline-dihydropyrimidine derivatives (5a-b) (Scheme 1) in 21-26% yield.
The temperature was measured with an IR sensor.

General procedure for the synthesis of 2,2,4-trimethyl-1,2-dihydroquinoline derivatives 2(a-h):
The mixture of different anilines 1(a-h) (12.17 mmol) with anhydrous magnesium sulphate (7.3 g, 60.67 mmol) in anhydrous acetone (50 ml) was added to iodine (154mg, 5 mol%) and tert-butylcatechol (61 mg, 3 mol%) and heated to reflux for 12hrs. The progress of reaction was monitored on TLC till consumption of aniline and then reaction mixture was allowed to cool and filtered through a bed of celite. The filtrate solution so obtained was concentrated under reduced pressure to give brown coloured semi-solid material which was S32 purified through column chromatography over silica in EtOAc (0.5-5%) in heptane to give desired 2,4-trimethyl-1,2-dihydroquinoline derivatives 2(a-h).
Finally, the crude reaction mixture is purified through HPLC in Gilson instrument with acetonitrile (10-70%) and TFA (0.1%) in water system to give required derivatives in 23-31% yield.