Improving anticancer activity towards colon cancer cells with a new p53‐activating agent

Abstract

Background and Purpose

Impairment of the tumour suppressor p53 pathway is a major event in human cancers, making p53 activation one of the most attractive therapeutic strategies to halt cancer. Here, we have identified a new selective p53 activator and investigated its potential as an anticancer agent.

Experimental Approach

Anti‐proliferative activity of the (R)‐tryptophanol‐derived bicyclic lactam SYNAP was evaluated in a range of human cancer cells with different p53 status. The anticancer activity and mechanism of action of SYNAP was studied in two‐ and three‐dimensional models of human colon adenocarcinoma HCT116 cells with wild‐type p53 and corresponding p53‐null isogenic derivative cells, alone and in combination with known chemotherapeutic agents.

Key Results

SYNAP showed anti‐proliferative effect in human cancer cells dependent on p53 status. In HCT116 cells, SYNAP caused p53‐dependent growth inhibition, associated with cell cycle arrest and apoptosis, anti‐migratory activity and regulation of the expression of p53 transcriptional targets. Data also indicated that SYNAP targeted p53, inhibiting its interaction with its endogenous inhibitors, murine double minute (MDM)2 and MDMX. Moreover, SYNAP sensitized colon cancer cells to the cytotoxic effect of known chemotherapeutic agents. SYNAP did not induce acquired or cross‐resistance and re‐sensitized doxorubicin‐resistant colon cancer cells to chemotherapy. Additionally, SYNAP was non‐genotoxic and had low cytotoxicity against normal cells.

Conclusion and Implications

SYNAP revealed encouraging anticancer activity, either alone or in combination with known chemotherapeutic agents, in colon cancer cells. Apart from its promising application in cancer therapy, SYNAP may provide a starting point for improved p53 activators.

Abbreviations

5‐FU

5‐fluorouracil

CETSA

cellular thermal shift assay

CI

combination index

Co‐IP

co‐immunoprecipitation

CRC

colorectal cancer

DRI

dose reduction index

MDM

murine double minute

mut

mutant

SRB

sulforhodamine B

wt

wild‐type

Introduction

Colorectal cancer (CRC) is a leading cause of cancer incidence and death worldwide. It has been estimated that about 56% of CRC patients die from the disease, which represents approximately 694,000 deaths every year worldwide (Hammond et al., 2016; Riihimaki et al., 2016; Arnold et al., 2017). The high metastatic potential and ever‐increasing therapeutic resistance of CRC have limited the success of its treatment. Therefore, improved therapeutic alternatives that raise the survival rates of CRC patients, as well as their quality of life by reducing treatment‐associated toxic side effects, are clearly needed (Deben et al., 2015).

The p53 tumour suppressor protein is a crucial regulator of cell proliferation, DNA repair and death. The loss of p53 transcriptional activity may result in uncontrolled cell proliferation and in the accumulation of genomic injuries that culminate in tumour growth and dissemination. In fact, the impairment of p53 has long been recognized as a common event in human cancer. Specifically, about half of cancer patients have mutations in the TP53 gene, while the remaining 50% harbour a defective wild‐type (wt) p53 pathway (Essmann and Schulze‐Osthoff, 2012; Kim and Lozano, 2018). As such, activation of p53 has been recognized as one of the most promising therapeutic strategies against cancer (Essmann and Schulze‐Osthoff, 2012). In cancers that harbour wt p53, this protein is often inactivated, mainly by interactions with the murine double minute (MDM) proteins, MDM2 and MDMX. Therefore, in the last years, the search for inhibitors of the interaction of p53 with MDMs has received great attention. In fact, several inhibitors of the p53‐MDM2 interaction are already under clinical trials, proving the relevance of these molecules in cancer therapy (Burgess et al., 2016; Wang et al., 2017). Nonetheless, given the distinct and cooperative function of both MDMs on p53 inactivation and the resistance of MDMX‐overexpressing cells to MDM2‐only inhibitors (e.g. nutlin‐3a), low MW compounds that block the inhibitory effect of both MDMs, represent the ideal strategy to achieve full p53 activation (Graves et al., 2012; Wang et al., 2017). However, the availability of such compounds is still limited (Burgess et al., 2016).

In the present work, we describe a low MW compound, the (R)‐tryptophanol‐derived bicyclic lactam SYNAP as a novel p53‐activating agent, which acts as an inhibitor of the interaction of p53 with MDM2 and MDMX. Its p53‐dependent, anti‐proliferative, pro‐apoptotic and anti‐migratory effects, as well as its effectiveness in combination therapy with known chemotherapeutic agents, may predict promising therapeutic applications of SYNAP in cancer treatment, particularly of CRC.

Methods

Synthesis of SYNAP

All reagents and solvents were obtained from commercial suppliers and were used without further purification. Melting point was determined using a Kofler camera Bock monoscope M (Frankfurt, Germany). Thin‐layer chromatography was performed using Merck Silica Gel 60 F254 plates (Darmstadt, Germany) and visualized by UV light. For flash column, chromatography was used Merck Silica Gel (230–400 mesh) (Darmstadt). ¹H and ¹³C‐NMR spectra were measured on a Bruker Fourier 300 spectrometer (Fällanden, Zurich). ¹H and ¹³C‐NMR chemical shifts are reported in p.p.m. (δ) referenced to the solvent used. Proton coupling constants ( J) are expressed in hertz (Hz). Microanalysis was performed in a Flash2000 ThermoScientific elemental analyser (Cambridge, UK) and is within ±0.4% of theoretical values.

Chemical synthesis of SYNAP (Figure 1A). To a solution of (R)‐tryptophanol (0.10 g, 0.53 mmol, 1.0 eq) in toluene (Carlo Erba, Val de Reuil Cedex, France) (5 mL) was added 3‐benzoyl propionic acid (Sigma‐Aldrich) (0.10 g, 0.58 mmol, 1.1 eq). The mixture was heated with reflux for 19 h in a Dean‐Stark apparatus. The solvent was removed under reduced pressure. The residue obtained was dissolved in ethyl acetate (LabChem, Santo Antão do Tojal, Portugal) (10 mL), and the organic phase was washed with saturated aqueous NaHCO₃ solution (10 mL) and then with saturated aqueous NaCl solution (10 mL). The organic extract was dried over anhydrous Na₂SO₄ (Carlo Erba). The dried solution was filtered and concentrated under reduced pressure. The resulting crude material was purified by flash column chromatography on silica gel (EtOAc/n‐hexane 1:1) to afford SYNAP (0.166 g, 95.1%). The product was recrystallized with ethyl acetate/n‐hexane (JMS Lda, Odivelas, Portugal). SYNAP: white solid; ${\alpha }_D^{25}$ = −54.7^o (c = 0.002, MeOH); mp: 155–156°C; ¹H‐NMR (300 MHz, CDCl₃) δ 8.00 (s, 1H), 7.52–7.48 (m, 2H), 7.44–7.32 (m, 5H), 7.17 (t, J = 7.5 Hz, 1H), 7.10–7.05 (m, 2H), 4.63–4.53 (m, 1H), 4.16 (dd, J = 8.8, 7.5 Hz, 1H), 3.61 (dd, J = 8.8, 6.9 Hz, 1H), 3.08 (ddd, J = 14.7, 6.3, 0.9 Hz, 1H), 2.92–2.78 (m, 1H), 2.66–2.42 (m, 3H), 2.28–2.20 (m, 1H) p.p.m. (Pereira et al., 2015); ¹³C‐RMN (100 MHz, CDCl₃) δ 180.15 (C=O), 142.67 (Cq), 136.07 (Cq), 128.69 (CH), 128.27 (CH), 127.35 (Cq), 125.07 (CH), 122.08 (CH), 122.03 (CH), 119.39 (CH), 118.75 (CH), 111.62 (Cq), 111.06 (ArC), 102.35 (Cq), 72.69 (CH₂), 55.53 (CH), 35.01 (CH₂), 32.62 (CH₂), 29.62 (CH₂) p.p.m.; Anal. Calcd (C₂₁H₂₀N₂O₂): C, 75.88%; H, 6.06%; N, 8.43%. Found C, 75.87%; H, 5.90%; N, 8.58%.

Figure 1.

SYNAP has p53‐dependent growth inhibitory effect in human cancer cells. (A) Chemical structure of SYNAP. (B) IC₅₀ values for SYNAP in wt p53‐, p53‐null and mut p53‐expressing cancer cells, after 48 h treatment. (C) Scatter‐plot representation of the relationship between SYNAP IC₅₀ values and p53 status in cancer cells. In (B) and (C), growth obtained with vehicle (DMSO) was set as 100%. Data shown are means ± SEM, n = 5 (two replicates each). In (C), *P < 0.05, significantly different as indicated; two‐way ANOVA followed by Dunnett's test.

The chemical synthesis of the SYNAP enantiomer starting from (S)‐tryptophanol, used the same protocol as described in Pereira et al. (2015).

Yeast‐based screening assay

Saccharomyces cerevisiae cells expressing human wt p53 alone or co‐expressed with human MDM2 or MDMX were used, as previously described (Soares et al., 2015b). Briefly, cells were grown in galactose‐induction selective medium with 0.1–50 μM SYNAP, nutlin‐3a or SJ‐172550 for 42 h, at 30°C. Yeast cell growth was analysed by colony‐forming unit counts.

Human cell lines and growth conditions

Human colon adenocarcinoma HCT116 cell lines expressing wt p53 (HCT116 p53^+/+) and its p53‐null isogenic derivative (HCT116 p53^−/−) were provided by B. Vogelstein (The Johns Hopkins Kimmel Cancer Center, Baltimore, MD, USA); human cervix adenocarcinoma HeLa, breast adenocarcinoma MCF‐7, MDA‐MB‐231 and MDA‐MB‐468, breast ductal carcinoma HCC1419, osteosarcoma SJSA‐1, large‐cell lung NCI‐H460 and non‐small cell lung NCI‐H1299 carcinomas, hepatocarcinoma HepG2, and non‐tumorigenic foreskin fibroblasts HFF‐1 cell lines were purchased from the ATCC (Rockville, MD, USA). HuH‐7 cell lines were purchased from the Japanese Collection of Research Bioresources cell bank (Osaka, Japan). Cancer cells were routinely cultured in RPMI‐1640 medium with UltraGlutamine from Lonza (VWR, Carnaxide, Portugal) supplemented with 10% FBS from Gibco (Alfagene, Lisboa, Portugal). HFF‐1 cells were cultured in DMEM/F‐12 supplemented with 10% FBS. All cells were maintained at 37°C in a humidified atmosphere of 5% CO₂. Cells were routinely tested for mycoplasma infection using the MycoAlert™ PLUS mycoplasma detection kit (Lonza).

Sulforhodamine B (SRB) assay

Human cell lines were seeded in 96‐well plates at a density of 4.0 × 10³ (HepG2, NCI‐H460), 5.0 × 10³ (HCT116, HeLa, SJSA‐1, MCF‐7, HCC1419, NCI‐1299, MDA‐MB‐468, HuH‐7, MDA‐MB‐231) and 1.0 × 10⁴ (HFF‐1) cells per well, for 24 h. Cells were treated with serial dilutions of SYNAP (ranging from 3 to 150 μM), for an additional 48 h. Effects on cell proliferation were measured by SRB assay, as described by Soares et al., (2015b), and IC₅₀ values were determined for each cell line using the GraphPad Prism software version 7.0 (La Jolla, CA, USA).

Colony formation assay

HCT116 p53^+/+ cells were seeded in six‐well plates at a density of 1.0 × 10³ cells per well, followed by incubation with 3.8, 7.8, 15 and 30 μM SYNAP for 11 days. Formed colonies were fixed with 10% methanol and 10% acetic acid for 10 min and then stained with 0.5% crystal violet (Sigma‐Aldrich) in 1:1 methanol/H₂O for 15 min. Colonies containing more than 20 cells were counted.

Cell cycle and apoptosis analyses

The analyses were performed basically as described by Soares et al., (2015b). Briefly, HCT116 cells were seeded in six‐well plates at a density of 1.5 × 10⁵ cells per well for 24 h, followed by treatment with 15 μM SYNAP for an additional 48 h. For cell cycle analysis, cells were stained with propidium iodide (Sigma‐Aldrich), and were analysed by flow cytometry, and cell cycle phases were identified and quantified using the FlowJo X 10.0.7 Software (Treestar, Ashland, OR, USA). For apoptosis, cells were stained using the Annexin V‐FITC Apoptosis Detection Kit I from BD Biosciences (Enzifarma, Porto, Portugal), according to the manufacturer's instructions. The AccuriTM C6 flow cytometer and the BD Accuri C6 software (BD Biosciences) were used.

Western blot analysis

HCT116 cells were seeded in six‐well plates at a density of 1.5 × 10⁵ cells per well for 24 h, followed by treatment with 15 μM SYNAP. Protein extracts were quantified using the Bradford reagent (Sigma‐Aldrich). Proteins were run in SDS‐PAGE and transferred to a Whatman nitrocellulose membrane from Protan (VWR, Carnaxide, Portugal). After blocking, proteins were identified using specific primary antibodies followed by HRP‐conjugated secondary antibodies described in Supporting Information Table S1. GAPDH was used as loading control. The signal was detected with the ECL Amersham kit from GE Healthcare (VWR, Carnaxide, Portugal). Two detection methods were used: the Kodak GBX developer and fixer (Sigma‐Aldrich) or the ChemiDoc™ XRS Imaging System from Bio‐Rad Laboratories (Amadora, Portugal). Band intensities were quantified using Fiji (ImageJ Software for the first method; Laboratory for optical and computational instrumentation, University of Wisconsin‐Madison, USA) as described (Schindelin et al., 2012), or the Image Lab software (version 5.2.1; Bio‐Rad Laboratories). Signal intensity is relative to the respective loading and normalized to control (DMSO), set as 1.

RNA extraction and RT‐qPCR

HCT116 p53^+/+ cells were seeded in six‐well plates at a density of 1.5 × 10⁵ cells per well for 24 h, followed by treatment with 15 μM SYNAP. Total RNA from HCT116 was extracted using the IllustraTM RNAspin Mini RNA Isolation Kit (GE Healthcare). One microgram of RNA was used for cDNA synthesis using the NZY M‐MuLV Reverse Transcriptase from Nzytech (Lisboa, Portugal) in 20 μL final volume, following the manufacturer's instructions. RT‐qPCR assays were performed in a 96‐well plate on a Real‐Time PCR Detection System (Bio‐Rad, version 3.1), starting with 16.5 ng of cDNA. The NZY qPCR Green Master Mix (Nzytech) and specific forward and reverse primers for MDM2 (F‐GGCCTGCTTTACATGTGCAA, R‐GCACAATCATTTGAATTGGTTGTC) and CDKN1A (p21; F‐CTGGAGACTCTCAGGGTCGAA, R‐GATTAGGGCTTCCTCTTGGAG) from Stabvida (Caparica, Portugal), were used; GAPDH was used as a reference gene.

In vitro migration assays

Cell migration was analysed using both the wound‐healing assay and the QCM 24‐Well Fluorimetric Chemotaxis Cell Migration Kit (8 μm) from Merck Millipore (Taper, Sintra, Portugal), as described (Soares et al., 2016). Briefly, in the wound‐healing assay, 5 × 10⁵ (p53^+/+ and p53^−/− HCT116) cells per well were grown to confluence in six‐well plates, and a fixed‐width wound was created in the cell monolayer. Cells were treated with 7 μM SYNAP and photographed at different time points, using an inverted Nikon TE 2000‐U microscope from Nikon Instruments Inc. (Izasa, Carnaxide, Portugal) at 100× magnification with a DXM1200F digital camera (Nikon Instruments Inc.) and an NIS‐Elements microscope imaging software (version 4; Nikon Instruments Inc.). Wound closure was calculated by subtracting the ‘wound’ area (measured using Fiji Software) at the indicated time point of treatment to the ‘wound’ area at the starting point. For Chemotaxis Cell Migration assay, 5 × 10⁵ HCT116 p53^+/+ cells·mL⁻¹ were prepared in serum‐free RPMI‐1640 with 7 μM SYNAP. The prepared cell suspensions were distributed in 24‐well plates (300 μL per insert), followed by an addition of 500 μL medium containing 10% FBS to the lower chamber. After 24 h, cells that migrated through the 8 μm pore membranes were eluted, lysed and stained with a green‐fluorescence dye that binds to cellular nucleic acids. The number of migrated cells is proportional to the fluorescence signal measured using the Bio‐Tek Synergy HT plate reader (Izasa), at 480/520 nm (ex/em).

Co‐immunoprecipitation (Co‐IP) assay

The Pierce Classic Magnetic IP and Co‐IP Kit from Thermo Scientific (Dagma, Carcavelos, Portugal) were used, as described (Soares et al., 2017). Briefly, Co‐IP was performed with anti‐p53 (IP:p53) and anti‐IgG antibodies, after treatment of HCT116 p53^+/+ cells with 15–45 μM SYNAP for 16 h, followed by immunoblotting with mouse monoclonal anti‐MDM2, anti‐MDMX, anti‐p53 and anti‐GAPDH antibodies.

Cellular thermal shift assay (CETSA)

CETSA experiments were performed, as described (Tan et al., 2015; Soares et al., 2017). Briefly, HCT116 p53^+/+ cells were lysed in appropriate buffer (25 mM Tris pH 7.4, 10 mM MgCl₂, 2 mM DTT) by Dounce homogenization. HCT116 p53^+/+ cell lysates were treated with 0.5–25 μM SYNAP for 1 h at room temperature, followed by heating at different temperatures for 3 min, cooled at room temperature for 3 min, and placed on ice. Insoluble protein was separated by centrifugation, and soluble protein was used to detect p53, MDM2, MDMX and GAPDH by Western blot. At 41°C, the increase in non‐denatured p53 of treated lysates was calculated relative to solvent (DMSO), set to unity. In the experiments at different heating temperatures, the signal intensity was normalized to the intensity at 25°C, set to unity.

Comet assay

HCT116 p53^+/+ cells were seeded in six‐well plates at a density of 1.5 × 10⁵ cells per well for 24 h, followed by treatment with 25 μM etoposide (positive control) or 15, 30, 45 μM SYNAP, for additional 48 h. To evaluate DNA damage, an alkaline comet assay was performed using the OxiSelect Comet Assay kit from Cell Biolabs (Meditecno, Carcavelos, Portugal), according to the manufacturer's instructions, as described (Soares et al., 2017). For tail DNA quantification, it was considered the percentage of comet positive cells (cells with more than 5% of DNA in the tail). Tail moment corresponds to the product of the tail length and the % of DNA in the tail. Cells were photographed using a Nikon DS‐5Mc camera and a Nikon Eclipse E400 fluorescence microscope (Izasa), and images processed using a Nikon ACT‐2 U software (Izasa). Images were quantified using Fiji Software (Open Comet/ImageJ). For each condition, five independent experiments were performed, and in each sample, 100 cells were quantified.

Generation of colon cancer spheroids

HCT116 p53^+/+ cells were resuspended in serum‐free stem cell culture media consisting of DMEM/F12 supplemented with 10 ng·mL⁻¹ bFGF, 20 ng·mL⁻¹ EGF from Bio‐techne (Citomed Lda, Lisboa, Portugal), 1 × B27 from Life Technologies (Porto, Portugal) and 5 μg·mL⁻¹ insulin (Sigma‐Aldrich). HCT116 p53^+/+ cells were plated in 24‐well ultra‐low attachment plates (one spheroid per well; Corning Inc., Sigma‐Aldrich), at a density of 1 × 10³ cells per well (Bessa et al., 2018). Cells were treated with 30, 45 and 60 μM SYNAP at the seeding time. Spheroids were then allowed to grow for 96 h in the presence of the compound. To assess the synergistic effect of SYNAP with doxorubicin in spheroids development, colonospheres were allowed to form for 3 days, followed by treatment with 60 μM SYNAP and/or 1 μM DOXO, for additional 8 days. During this period of time, new medium with the drugs (or DMSO only) was added to the wells at days 1, 3 and 6 of treatment. Spheroid formation was monitored using an inverted Nikon TE 2000‐U microscope at 100× magnification, with a DXM1200F digital camera and NIS‐Elements microscope imaging software. Spheroid area was quantified using the Fiji Software (Schindelin et al., 2012).

Combination therapy assays

To assess the synergistic effect of SYNAP with known chemotherapeutic agents, HCT116 p53^+/+ cells were treated with 7 μM SYNAP and/or increasing concentrations of doxrubicin (18.7–150 nM), cisplatin (0.5–4 μM), etoposide (0.37–3 μM), paclitaxel (0.4–3 nM) or 5‐FU (0.6–5 μM) for 48 h. The effect of combined treatments on cell proliferation was analysed by SRB assay. For each combination, the combination index (CI) and the dose reduction index (DRI) values were calculated using the CompuSyn Software version 1.0 (ComboSyn, Inc., Paramus, NJ, USA), according to the following equation: CI = (D)₁/(D_x)₁ + (D)₂/(D_x)₂, where the numerators (D)₁ and (D)₂ are the concentrations of each drug in the combination [(D)₁ + (D)₂] that inhibit x%, and the denominators (D_x)₁ and (D_x)₂ are the concentrations of drug one and two alone that inhibit x%; DRI measures how much the dose of a drug may be reduced in synergistic combination compared to the dose of each drug alone; CI values < 1, 1 < CI < 1.1 and > 1.1 indicate synergistic, additive and antagonistic effects respectively (Chou and Talalay, 1984).

To assess the proliferative capability of HCT116 p53^+/+ cells after removal of drugs, cells treated for 48 h with 7 μM SYNAP and/or increasing concentrations of doxorubicin (18.7–150 nM) and 5‐FU (0.6–5 μM) were allowed to grow in fresh media for 5 days, followed by evaluation of cell proliferation and CI.

In 3D spheroid model, the area of colonospheres was assessed using the following equation: A = πab, where (a) corresponds to the major axis and (b) to the minor. The synergistic effect was determined using the Additive model, in which a positive drug combination effect occurs when the observed combination effect (E_AB) is greater than the expected additive effect given by the sum of the individual effects (E_A + E_B). The CI was calculated as described (Jonsson et al., 1998; Foucquier and Guedj, 2015): CI= $\frac{\mathrm{EA}+\mathrm{EB}}{\mathrm{EAB}}$.

Establishment of doxorubicin‐resistant cell lines

To generate HCT116 p53^+/+ cells resistant to doxorubicin (DOXO/RES HCT116), cells were exposed to increasing concentrations of this agent as previously reported (Ogawara et al., 2009; Yan et al., 2017). Briefly, HCT116 p53^+/+ cells were exposed to several rounds of selection with increasing concentrations of doxorubicin (0.01, 0.02, 0.04, 0.08, 0.16, 0.32, 0.50 and 1 μM), which were added to cultured medium for 24 h, followed by a recovery period of 48 h in fresh medium without DOXO. To preserve the resistance, the DOXO/RES HCT116 cells were maintained in the presence of 1 μM doxorubicin and grown in medium without doxorubicin for 3 to 4 days before experiments. The same passage number of both control and resistant cells was used in the experiments. The IC₅₀ values of compounds, in DOXO/RES and control HCT116 cells, were determined by SRB assay.

Acquired resistance studies

HCT116 p53^+/+ cells were exposed to five rounds of selection with increasing concentrations of SYNAP (15, 30, 45, 60 and 75 μM), which were added to culture medium for 24 h, followed by a recovery time of two to three days. Cells were harvested, seeded and treated twice with each concentration. At the end of each round, IC₅₀ values were determined by SRB assay after 48 h treatment. Both control and surviving resistant cells were used for each round, with the same passage number.

Data and statistical analysis

Data and statistical analysis in this study comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2018). Data are presented as mean ± SEM of ‘n’ samples, where ‘n’ refers to independent experiments, not replicates. Values of ‘n’ and number of technical replicates, if performed, are given in Figure and Table legends. Where replicates were used, their values were averaged to provide a single value to the data set. Data analyses were carried out using GraphPad Prism Software version 7.0 (La Jolla). All assays with five or more independent repeats were subjected to statistical analysis. In some data sets, log transformation to generate Gaussian‐distributed data set was carried out. Normalization was made for controlling unwanted sources of variation, and data analysis was performed setting controls (DMSO or non‐threated cells) as 100% or as one for comparison purposes. For comparison of two groups, unpaired Student's t‐test was used. For comparison of multiple groups, statistical analysis relative to controls was performed using one‐way or two‐way ANOVA followed by post hoc Sidak's or Dunnet's multiple comparison tests. Statistical significance was set as *P < 0.05. Post hoc tests were run only if F achieved P < 0.05 and when no significant variance inhomogeneity was observed.

Materials

http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=7069, SJ‐172550, and nutlin‐3a were obtained from Sigma‐Aldrich (Sintra, Portugal); http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=6815 was from Calbiochem (VWR, Carnaxide, Portugal); http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5343, http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2770 and http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4789 (5‐FU) were from Enzo Life Science (Taper, Sintra, Portugal). All tested compounds were dissolved in DMSO (Sigma‐Aldrich), except cisplatin which was dissolved in saline. In all experiments, the solvent was included as control in a concentration range that did not affect cell proliferation (0.1–0.25%).

Nomenclature of targets and ligands

Key proteins and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise guide to Pharmacology (Alexander et al., 2017a,b).

Results

SYNAP has p53‐dependent growth inhibitory effect in human cancer cells

In earlier experiments, we identified phenylalaninol‐derived oxazolopyrrolidone lactams (Soares et al., 2015a) and tryptophanol‐derived oxazoloisoindolinones (Soares et al., 2016; Soares et al., 2017) that selectively activated the p53 pathway. Based on these promising results, we developed two enantiomers (starting from (S) or (R)‐tryptophanol) that contain in the same molecule, the tryptophanol and oxazolopyrrolidone moieties. The anti‐proliferative activity of these two compounds was tested in a panel of 13 human cancer cell lines with different p53 status, using the SRB assay. Interestingly, in contrast to the (S)‐tryptophanol‐derivative (data not shown), the (R)‐tryptophanol‐derivative SYNAP (Figure 1A) showed a p53‐dependent growth inhibitory effect in cancer cells (Figure 1B,C). The IC₅₀ values of SYNAP in these 13 cancer cell lines ranged from 8.5 to 20.5 μM in wt p53‐expressing cancer cells (HCT116 p53^+/+, HeLa, SJSA‐1, NCI‐H460, MCF‐7 and HepG2) to 45–65 μM in p53‐null cancer cells (HCT116 p53^−/−, HCC1419 and NCI‐H1299) and 42.5–95.0 μM in mutant (mut) p53‐expressing cancer cells (MDA‐MB‐468, HuH‐7 and MDA‐MB‐231) (Figure 1B). A scatter‐plot representation of the correlation between IC₅₀ values of SYNAP and p53 status in cancer cells clearly showed lower IC₅₀ values for the group of wt p53‐expressing cells compared to the groups of p53‐null and mut p53‐expressing cells (Figure 1C).

Although the lowest IC₅₀ value of SYNAP was obtained in HeLa cells (8.5 μM), the HCT116 colon cancer cells expressing p53 (HCT116 p53^+/+) and respective p53‐knockout (HCT116 p53^−/−) were used to validate the p53‐dependence of the tumour growth inhibitory activity of SYNAP. The growth inhibition of HCT116 p53^+/+ cells by SYNAP was significantly lower than that in HCT116 p53^−/− cells (Figure 2B,C), as shown from the dose–response curves from the SRB assay (Figure 2A), and that from the colony formation assay. Consistent with these results, SYNAP, at 15 μM (IC₅₀ in HCT116 p53^+/+ cells), induced G0/G1‐phase cell cycle arrest (Figure 2D), and apoptosis (annexin‐V positive cells; Figure 2E) with PARP cleavage (Figure 2F‐H) in p53^+/+, but not in p53^−/−, HCT116 cells. Interestingly, in contrast to SYNAP, nutlin‐3a, a known inhibitor of the p53‐MDM2 interaction, did not induce apoptosis in HCT116 p53^+/+ cells, at 3 μM (IC₅₀ value) (Figure 2E). In fact, in these cancer cells, the anti‐proliferative activity of nutlin‐3a was only associated with G0/G1‐phase cell cycle arrest (Figure 2D).

Figure 2.

Anticancer effect of SYNAP in human colon cancer cells. (A) Concentration–response curves for SYNAP in p53^+/+ and p53^−/− HCT116 cells, after 48 h treatment. (B, C) Effect of SYNAP on colony formation of p53^+/+ and p53^−/− HCT116 cells, after 11 days treatment. In (B), a representative experiment is shown. In (C), quantification of colony formation. In (A) and (C), growth obtained with control (DMSO) was set as 100%. Data shown are means ± SEM, n = 5 (two replicates each). *P < 0.05, significantly different from p53^−/−: two‐way ANOVA followed by Sidak's test. (D, E) Effect of 15 μM SYNAP and 3 μM nutlin‐3a in cell cycle (D) and apoptosis (E), in p53^+/+ and p53^−/− HCT116 cells, after 48 h treatment. Data shown are means ± SEM, n = 5. *P < 0.05, significantly different from DMSO; two‐way ANOVA followed by Dunnett's test. (F–H) Effect of 15 μM SYNAP on the protein expression of p53, p53‐target genes and cleaved PARP, in p53^+/+ and p53^−/− HCT116 cells, after 24 h (p53, MDM2) or 48 h (PARP cleavage, p21, Puma, Bcl‐2, survivin) treatment. In (F), representative immunoblots are shown. GAPDH was used as loading control. (G, H) Quantification of protein expression levels in p53^+/+ (G) and p53^−/− (H) HCT116 cells, relative to DMSO (set as 1). Data shown are means ± SEM, n = 3. NS: No signal was detected. (I) Effect of 15 μM SYNAP in mRNA levels of MDM2 and CDKN1A in HCT116 p53^+/+ cells, after 24 h treatment. Data shown are means ± SEM, n = 5 (three replicates each). Fold expression changes relative to DMSO (set as 1). *P < 0.05, significantly different from DMSO; Student's t‐test.

Moreover, by Western blot analysis, we confirmed that 15 μM SYNAP increased the protein levels of p53 and of major p53 transcriptional targets, including MDM2, p21 and http://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=910, while decreasing the protein levels of the anti‐apoptotic p53 transcriptional targets http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2844 and http://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=889#2795, in p53^+/+, but not in p53^−/−, HCT116 cells (Figure 2F–H). Accordingly, 15 μM SYNAP also increased the mRNA levels of p53 transcriptional targets, namely, MDM2 and CDKN1A (p21), in HCT116 p53^+/+ cells (Figure 2I).

SYNAP has p53‐dependent anti‐migratory activity in human colon cancer cells

The effect of SYNAP on the migration ability of HCT116 cells was also studied. In the wound‐healing assay, at 7 μM (a concentration with no significant effect on cell proliferation), SYNAP significantly reduced the wound closure in p53^+/+, but not in p53^−/−, HCT116 cells compared to vehicle (Figure 3A,B). These results were further supported by the chemotaxis cell migration assay, in which 7 μM SYNAP caused a more pronounced reduction of cell migration in HCT116 p53^+/+ cells compared to HCT116 p53^−/− cells, after 24 h treatment (Figure 3C). The p53‐dependent anti‐metastatic effect of 7 μM SYNAP was also reinforced by the higher reduction in the protein expression levels of http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1633 in HCT116 p53^+/+ cells compared to HCT116 p53^−/− cells, after 24 h treatment (Figure 3D,E).

Figure 3.

SYNAP inhibits migration of human colon cancer cells in a p53‐dependent manner. (A, B) HCT116 confluent cells treated with 7 μM SYNAP (or DMSO only) were observed at different time points in the wound‐healing assay. In (A), representative images (scale bar = 50 μm; magnification = 100×) are shown. In (B), quantification of wound closure using randomly selected microscopic fields (six fields per sample). Data shown are means ± SEM, n = 5. *P < 0.05, significantly different from DMSO; two‐way ANOVA followed by Sidak's test. (C) Effect of 7 μM SYNAP on migration of HCT116 cells after 24 h treatment; migratory cells were quantified by fluorescence signal intensity; values obtained with non‐treated cells were set as 1. Data shown are means ± SEM, n = 5 (two replicates each). *P < 0.05, significantly different from DMSO; Student's t‐test. (D, E) Western blot analysis of MMP‐9, in p53^+/+ and p53^−/− HCT116 cells, after 24 h treatment with 7 μM SYNAP (or DMSO only). In (D), representative immunoblots are shown. GAPDH was used as loading control. In (E), quantification of protein expression levels relative to DMSO (set as 1). Data shown are mean ± SEM, n = 3.

SYNAP potentially targets p53, inhibiting its interaction with MDM2 and MDMX, in human colon cancer cells

In order to further understand the possible mechanism(s) underlying p53 activation by SYNAP, we investigated the ability of this compound to disrupt the interaction of p53 with MDM2 and MDMX, using a previously developed yeast‐based assay (Soares et al., 2015b). In this assay, the expression of human wt p53 in yeast induces growth arrest, which is abolished by co‐expressed human MDM2 or MDMX. In this system, inhibitors of the p53 interaction with MDM2 (such as nutlin‐3a; Vassilev et al., 2004) or MDMX (such as SJ‐172550; Reed et al., 2010) abolish the MDM2 or MDMX inhibitory effect, respectively, with the re‐establishment of the p53‐induced growth arrest. Using this cell system, the effect of 0.1–50 μM SYNAP on the p53‐MDM2/MDMX interactions was compared to that of 0.1–50 μM nutlin‐3a and SJ‐172550 (Figure 4A). As expected, while nutlin‐3a abolished the MDM2 inhibitory effect, SJ‐172550 abolished the MDMX inhibitory effect on p53. Unlike nutlin‐3a and SJ‐172550, SYNAP abolished the inhibitory effect of both MDM2 and MDMX, with the re‐establishment of the p53‐induced yeast growth arrest after 48 h treatment. Notably. 0.1–50 μM SYNAP did not interfere with the growth of control yeast (transformed with empty vectors) or yeast expressing p53, MDM2 or MDMX alone (Figure 4B), confirming the selectivity of SYNAP towards the inhibition of the p53‐MDM2/MDMX interactions.

Figure 4.

SYNAP inhibits the p53‐MDM2/MDMX interactions in yeast and human colon cancer cells. (A) Effect of 0.1–50 μM SYNAP, nutlin‐3a and SJ‐172550 on the reversal of MDM2/MDMX effect, by reestablishment of p53‐induced growth inhibition, in yeast cells co‐expressing human p53 and MDM2 or MDMX, after 42 h treatment. (B) Effect of 0.1–50 μM SYNAP on the growth of yeast cells expressing p53, MDM2 or MDMX alone, and yeast transformed with empty vectors, after 42 h treatment. In (A) and (B), data are means ± SEM, n = 6 (two replicates each). The growth of non‐treated yeast cells was set as 100% (control). *P < 0.05, significantly different from DMSO; two‐way ANOVA followed by Dunnett's test. (C, D) Co‐IP was performed in HCT116 p53^+/+ cells treated with SYNAP (or DMSO only) for 16 h. In (C), representative immunoblots are shown; whole‐cell lysate (Input). p53 from IP was used as loading control. In (D), quantification of protein expression levels relative to DMSO (set as 1). Data shown are means ± SEM, n = 3.

The results obtained in yeast led us to check the ability of SYNAP to inhibit the p53‐MDM2/MDMX interactions in human cancer cells, using the Co‐IP assay in HCT116 p53^+/+ cells treated with 15, 30 and 45 μM SYNAP for 16 h (Figure 4C,D). Here, SYNAP caused a visible decrease in the amount of MDM2 and MDMX co‐immunoprecipitated with p53, indicating a disruption of the p53‐MDM2 and p53‐MDMX interactions by this compound. In line with this result, the effects of SYNAP on wt p53‐expressing cancer cells were not affected by the overexpression of MDM2 (in SJSA‐1 and HepG2 cells) or MDMX (in MCF‐7 cells) (Figure 1B).

Using CETSA in HCT116 p53^+/+ cells, we found that SYNAP caused p53 thermal stabilization, measured by the amount of soluble p53 upon heating. Particularly, SYNAP induced a concentration‐dependent p53 thermal stabilization at 41°C (Figure 5A), as well as from 39 to 43°C, at 10 μM (Figure 5B). Notably, the effect of SYNAP on protein thermal stabilization was selective for p53, as SYNAP had no effect on MDM2 and MDMX thermal stabilization, even for the highest concentration tested of SYNAP that caused p53 protein stabilization (25 μM) (Figure 5C,D). These results identified p53 as a putative target of SYNAP.

Figure 5.

SYNAP stabilizes p53 in human colon cancer cells. CETSA experiments were performed in HCT116 p53^+/+ cell lysates. (A) Lysate samples treated with increasing concentrations of SYNAP were heated at 41°C; plot represents the increase of non‐denatured p53 in SYNAP‐treated samples relative to DMSO. Data shown are means ± SEM, n = 3. (B–D) Lysate samples, obtained after treatment with SYNAP (or DMSO only), were heated at different temperatures; plots represent the signal intensity of p53 (B), MDM2 (C) and MDMX (D) normalized to the signal intensity at 25°C. Data shown are means ± SEM, n = 3. In (A–D), representative immunoblots are shown. GAPDH was used as loading control.

SYNAP is non‐genotoxic in human colon cancer cells and shows low growth inhibition in normal cells

The effects of SYNAP on DNA damage were assessed by analysis of comet‐positive cells and of the phosphorylation (Ser¹³⁹) of histone H2AX (γH2AX), in HCT116 p53^+/+ cells, after 48 h treatment. Unlike etoposide (positive control), 15, 30 and 45 μM SYNAP did not increase the percentage of tail DNA (Figure 6A,B) and tail moment (Figure 6A,C). Additionally, unlike etoposide, 30 and 45 μM SYNAP did not induce phosphorylation of H2AX (Figure 6D).

Figure 6.

SYNAP is non‐genotoxic in human colon cancer cells and has low cytotoxicity against normal cells. (A–C) Measurement of DNA damage in HCT116 p53^+/+ cells with etoposide (ETOP; positive control) and SYNAP (or DMSO only), after 48 h treatment. In (A), representative images (scale bar = 50 μm; magnification = 100×) are shown. In (B), quantification of tail DNA percentage. In (C), quantification of tail moment. In (B) and (C), data shown are means ± SEM, n = 5 (100 cells per sample). *P < 0.05, significantly different from DMSO; one‐way ANOVA with Dunnett's test. (D) Analysis of γH2AX expression levels; immunoblots represent one of three independent experiments. GAPDH was used as loading control. (E) Concentration–response curve of SYNAP in HFF‐1 normal human cells, after 48 h treatment; growth obtained with control (DMSO) was set as 100%. Data shown are means ± SEM, n = 5 (two replicates each).

Moreover, because SYNAP was an activator of wt p53, it was essential to check its effects on the growth of normal cells. For that, the effects of SYNAP in HFF‐1 cells were assayed by the SRB method (Figure 6E). We found an IC₅₀ value in HFF‐1 cells (78.5 ± 3.07 μM) five‐fold higher than that obtained in HCT116 p53^+/+ cells, indicating that, despite the expression of wt p53 in normal cells, SYNAP was more effective against cancer cells.

SYNAP sensitizes human colon cancer cells to the effect of known chemotherapeutic agents

To evaluate the ability of SYNAP to enhance the anticancer activity of known anti‐cancer drugs, HCT116 p53^+/+ cells were treated with a single concentration of SYNAP (7 μM; concentration with no significant effect on cancer cell growth) in combination with a range of concentrations of 5‐FU, cisplatin, doxorubicin, etoposide or paclitaxel. The results showed that SYNAP enhanced the growth inhibitory effect of all of these chemotherapeutic agents in cancer cells (Figure 7A). Using the CompuSyn software, a multiple drug‐effect analysis was performed for each combination, with the calculation of CI and DRI (Table 1). This analysis revealed synergistic effects between SYNAP and all chemotherapeutic agents, for at least two of the concentrations tested (Table 1). However, the most promising synergistic effects were obtained with doxorubicin and 5‐FU (Figure 7A; Table 1), which were associated with a significant increase in apoptosis (Figure 7B). Most importantly, after 5 days of recovery in fresh medium (free of drugs), the synergistic effects of SYNAP with doxorubicin and 5‐FU were maintained (Figure 7C; Table 1), indicating no rescue of cell proliferation.

Figure 7.

SYNAP sensitizes human colon cancer cells to known chemotherapeutic agents and inhibits spheroid formation and development, alone and in combination with doxorubicin (DOXO). (A) Cells were treated with a range of concentrations of conventional chemotherapeutic drugs alone and in combination with 7 μM SYNAP; cell proliferation was measured after 48 h treatment; growth obtained with control (DMSO) was set as 100%. Data shown are means ± SEM, n = 6 (two replicates each). *P < 0.05, significantly different from chemotherapeutic drugs alone; two‐way ANOVA followed by Sidak's test. (B) Effect of 7 μM SYNAP in combination with 5‐FU or doxorubicin in apoptosis of HCT116 p53^+/+ cells, after 48 h treatment. Data shown are means ± SEM, n = 5 (two replicates each). *P < 0.05, significantly different from DMSO; two‐way ANOVA followed by Dunnett's test. (C) Cells treated with a range of concentrations of 5‐FU and doxorubicin alone and in combination with 7 μM SYNAP for 48 h followed by 5 days of recovery in fresh medium; growth obtained with control (DMSO) was set as 100%. Data shown are means ± SEM, n = 6 (two replicates each). *P < 0.05, significantly different from 5‐FU and doxorubicin alone; two‐way ANOVA followed by Sidak's test. (D, E) Evaluation of spheroids formation after 96 h treatment with SYNAP (or DMSO only); treatment was performed at the seeding time of HCT116 p53^+/+ cells. (F, G) Effect of 60 μM SYNAP alone and in combination with 1 μM doxorubicin on three‐day‐old HCT116 p53^+/+ spheroids, for up to 8 days treatment. In (D) and (F), representative images are shown. In (E) and (G), determination of spheroids area at the end of treatment. Data shown are means ± SEM, n = 5. *P < 0.05, significantly different from DMSO; Student's t‐test. In (G), CI was determined considering spheroid area. In (D) and (F), scale bar = 50 μm (magnification = 100×).

Table 1.

Effect of SYNAP in combination with known chemotherapeutic agents

	Mutually nonexclusive CI		DRI
Drug combination with SYNAP	CI	Profile	SYNAP	Conventional drug
DOXO (nM)
18.7	0.676	Synergy	2.113	4.926
37.5	0.762	Synergy	2.432	2.847
75	0.701	Synergy	3.762	2.247
150	0.616	Synergy	6.902	2.121
DOXO (nM)a
18.7	0.663	Synergy	3.433	5.975
37.5	0.793	Synergy	5.780	2.522
75	0.893	Synergy	8.881	3.564
150	0.619	Synergy	15.394	2.912
DOXO (nM)b
25	0.360	Synergy	5.182	5.975
62.5	0.585	Synergy	5.282	2.522
125	0.411	Synergy	7.649	3.564
250	0.453	Synergy	9.112	2.912
Cisplatin (μM)
0.5	0.801	Synergy	1.623	5.397
1	0.832	Synergy	1.836	3.467
2	0.987	Additive	1.997	2.056
4	1.08	Additive	2.408	1.053
Paclitaxel (nM)
0.37	1.201	Antagonism	‐‐‐	‐‐‐
0.75	1.216	Antagonism	‐‐‐	‐‐‐
1.5	0.894	Synergy	3.546	1.632
3	0.324	Synergy	10.275	4.399
5‐FU (μM)	–
0.65	0.650	Synergy	1.812	2.957
1.25	0.593	Synergy	2.194	2.606
2.5	0.670	Synergy	2.649	2.192
5	0.726	Synergy	3.378	2.142
5‐FU (μM)a
0.65	0.800	Synergy	3.433	3.582
1.25	0.839	Synergy	5.780	2.091
2.5	0.833	Synergy	8.881	1.458
5	0.762	Synergy	15.394	1.811
ETOP (μM)
0.38	1.099	Additive	1.287	2.577
0.75	0.867	Synergy	1.834	3.110
1.5	0.803	Synergy	2.300	2.711
3	1.099	Additive	2.298	1.351

Effect of 7 μM SYNAP in combination with chemotherapeutic drugs, in HCT116 p53^+/+ cells, was evaluated using CompuSyn software to calculate CI and DRI values for each combined treatment. CI < 1, synergy; 1 < CI < 1.1, additive effect; CI > 1.1, antagonism. Data were calculated using a mean value effect (n = 6). DOXO, doxorubicin; ETOP, etoposide.

^aResults obtained after 5 days of recovery in fresh medium.

^bResults obtained in DOXO/RES HCT116 cells; (‐‐‐): no increase of DRI.

Note that, for paclitaxel, both synergistic and antagonistic effects were obtained (Table 1). This was not an unexpected result as, in wt p53‐expressing cancer cells, inhibitors of the p53‐MDM2 interaction are known to induce a G0/G1‐phase cell cycle arrest, attenuating the activity of S‐phase‐specific anti‐mitotic drugs (Carvajal et al., 2005). Consistent with this finding, antagonistic effects with the anti‐mitotic agent paclitaxel were also observed with nutlin‐3a, in colon cancer cells (Ohnstad et al., 2011).

In an attempt to explore the anticancer activity of SYNAP in a system that more closely resembles the in vivo settings, a 3D colonosphere culture model was generated from HCT116 p53^+/+ cells. The effect of SYNAP in colonosphere formation was thereafter determined by evaluating the sphere area after 96 h treatment. Despite the evident reduction in the efficacy of compounds (SYNAP and doxorubicin) in this culture model, it was possible to confirm that, when added at the seeding time of cancer cells suspension, 45 and 60 μM SYNAP clearly inhibited colonosphere formation, almost abolishing colonosphere formation at 60 μM (Figure 7D,E). The 3D colonosphere model was also used to validate the synergistic effects between SYNAP and doxorubicin. Although doxorubicin is used in many cancer chemotherapy regimens, its clinical application to the treatment of CRC has been hindered by its dose‐limiting toxicity and development of resistance (Yan et al., 2017). This led us to pursue the combination therapy studies with doxorubicin. Therefore, 3‐day‐old spheroids were treated with 60 μM SYNAP alone and in combination with doxorubicin for up to 8 days. In monotherapy, 60 μM SYNAP caused a significant reduction in colonosphere area, compared with the effects of the vehicle (Figure 7F,G). When combined with 1 μM doxorubicin, a synergic reduction in colonosphere area was observed, with the disintegration of colonospheres (Figure 7F,G).

Colon cancer cells develop resistance to doxorubicin but not to SYNAP: doxorubicin‐resistant cancer cells show no cross‐resistance to SYNAP and are re‐sensitized to doxorubicin by SYNAP

We first investigated whether SYNAP might induce resistance in colon cancer cells after several rounds of treatment with increasing concentrations of the compound. In these assays, the emerging resistance was identified by increasing IC₅₀ values in the successive generations. The results showed that colon cancer cells did not develop resistance to SYNAP, as demonstrated by the preservation of its IC₅₀ value in successive generations (Figure 8A). Additionally, as cancer cells that acquire resistance to one chemotherapeutic agent commonly exhibit cross‐resistance to other agents (Gupta et al., 1988), we investigated whether doxorubicin‐resistant colon cancer cells would also exhibit cross‐resistance to SYNAP. To this end, we started by establishing doxorubicin‐resistant HCT116 p53^+/+ cells (DOXO/RES HCT116; Figure 8B), followed by analysis of the effect of increasing concentrations of SYNAP on the growth of the DOXO/RES HCT116 cells (Figure 8C). The results showed that the sensitivity of DOXO/RES HCT116 cells to SYNAP was similar to that of non‐resistant parental HCT116 p53^+/+ cells (control). Altogether, these data revealed that SYNAP did not induce cross‐ or acquired‐resistance in colon cancer cells.

Figure 8.

Human colon cancer cells develop resistance to doxorubicin but not to SYNAP. SYNAP has no cross‐resistance in doxorubicin‐resistant cells and re‐sensitizes these cells to doxorubicin effect. (A) HCT116 p53^+/+ cells were exposed to five rounds of treatment with 15, 30, 45, 60 and 75 μM of SYNAP. IC₅₀ values were determined at the end of each round by SRB assay after 48 h treatment. (B, C) Concentration–response curves for doxorubicin (DOXO; B) and SYNAP (C) in control (non‐resistant) and doxorubicin‐resistant (DOXO/RES) HCT116 cells, after 48 h treatment. (D) DOXO/RES HCT116 cells were treated with a range of concentrations of doxorubicin alone and in combination with 7 μM SYNAP; cell proliferation was measured after 48 h treatment. In (A–D), growth obtained with control (DMSO) was set as 100%. Data shown are means ± SEM, n = 5 (two replicates each). In (A) and (C), values obtained for HCT116 p53^+/+ treated cells (A) or DOXO/RES HCT116 p53^+/+ cells (C) are not significantly different from parental or control cells respectively. In (B), *P < 0.05, significantly different from control; in (D), *P < 0.05, significantly different from doxorubicin alone; two‐way ANOVA followed by Sidak's test.

The ability of SYNAP to re‐sensitize doxorubicin‐resistant colon cancer cells to doxorubicin effect was also evaluated. For this, DOXO/RES HCT116 cells were treated with 7 μM SYNAP in combination with increasing concentrations of doxorubicin (Figure 8D). In these cells and in parental HCT116 p53^+/+ cells, SYNAP enhanced the growth inhibitory effect of doxorubicin. (Figure 8D). In fact, the CI and DRI values confirmed a marked synergistic effect between SYNAP and doxorubicin, with a six‐fold reduction of the effective dose of doxorubicin (DRI value) in doxorubicin‐resistant colon cancer cells (Table 1).

Discussion

The discovery of new pharmacological alternatives with enhanced efficacy against cancer types for which effective therapies are still lacking, such as CRC, has been a priority in cancer research. As a characteristic of human cancer is the suppression of the p53 pathway, the activation of this pathway has become one of the most attractive therapeutic strategies against cancer.

In earlier work, we demonstrated the ability of enantiopure tryptophanol‐derived oxazoloisoindolinones, namely, SLMP53–1 (Soares et al., 2016) and DIMP53–1 (Soares et al., 2017) to activate p53. In the present work, we developed a novel (R)‐tryptophanol‐derived bicyclic lactam, the compound SYNAP, that was found to act as a p53‐activating agent. The growth inhibitory effects of SYNAP on a range of human cancer cell lines with different p53 status was clearly dependent on the p53 status of the cell line. In particular, in colon cancer cells, SYNAP induced cell cycle arrest and apoptosis, increased the p53 protein levels and interfered with the expression levels of major p53 transcriptional targets, increasing MDM2, p21 and Puma, while reducing the expression of the anti‐apoptotic proteins Bcl‐2 and survivin, in a p53‐dependent manner. SYNAP also demonstrated a p53‐dependent anti‐migratory activity in colon cancer cells, which was associated with a marked reduction in the levels of MMP‐9, a prominent player in extracellular matrix homeostasis and metastasis (Said et al., 2014). Interestingly, based on the established correlation between increased MMP‐9 expression and worse prognosis of CRC patients, many studies have claimed that therapies capable of reducing MMP‐9 expression levels would be quite relevant for CRC treatment (Said et al., 2014). Importantly, we also showed that, despite its growth inhibitory activity against cancer cells, SYNAP is non‐genotoxic and had low cytotoxicity against normal cells.

In this work, evidence has emerged for a possible mechanism underlying p53 activation by SYNAP in cancer cells, which seems to involve the disruption of the interaction of p53 with MDM2 and MDMX. In fact, this may explain the similar anti‐proliferative activity of SYNAP in wt p53‐expressing cancer cells overexpressing MDM2 or MDMX, which is not observed with MDM2‐only inhibitors like nutlin‐3a (Graves et al., 2012). Notably, the induction of p53 protein thermal stabilization by SYNAP in cancer cells identified p53 as a putative target of SYNAP. Consistent with these results, a tryptophanol‐derived oxazoloisoindolinone was also recently described as a low MW, p53‐binding compound with the ability to disrupt the p53 interaction with both MDM2 and MDMX (Soares et al., 2017). These results reinforce the notion that tryptophanol derivatives are promising starting points for the development of new p53‐activating agents, particularly those that are inhibitors of the p53‐MDM2/MDMX interactions.

The use of targeted‐therapy agents in combination therapy has been rationalized as a means to overcome developed resistance to drugs, while minimizing the toxic side effects usually observed with known chemotherapeutic agents (Hammond et al., 2016; Lopez and Banerji, 2017). Besides its potential use as monotherapy, SYNAP proved to be also highly effective in combination with several established chemotherapeutic agents in cancer therapy. In fact, SYNAP exhibited synergistic effects with all tested chemotherapeutic drugs, including with that most commonly used in CRC treatment (5‐FU), with a significant reduction of their effective dose. In particular, for doxorubicin, in a 3D colonosphere model, the combination with SYNAP led to the disintegration of the colonospheres.

Moreover, the acquisition of resistance by cancer cells is one of the major problems in cancer treatment that may be overcome using new combination therapies. Interestingly and in contrast to the experience with doxorubicin, colon cancer cells did not develop acquired‐resistance to SYNAP, thus avoiding the drug resistance feature displayed by many chemotherapeutic agents in cancer cells. Also, although cancer cells resistant to one drug can exhibit cross‐resistance to other drugs (Zanjirband et al., 2016; Lopez and Banerji, 2017), in our work, doxorubicin‐resistant cells showed no cross‐resistance to SYNAP. Indeed, SYNAP was able to re‐sensitize doxorubicin‐resistant colon cancer cells to the effect of doxorubicin, leading to a marked reduction in its effective dose. Although doxorubicin has been the therapy of choice for several cancers, its use in CRC therapy has been limited by its cardiotoxicity subsequent to the high doses required to overcome the resistance commonly developed by this type of cancer to doxorubicin (Sonowal et al., 2017). The results obtained in this work revealed that CRC therapy may greatly benefit from the combination of doxorubicin with SYNAP.

A possible explanation for the enhanced sensitization of colon cancer cells to the effects of known chemotherapeutic agents, by SYNAP may be related to the ability of SYNAP to reduce the levels of survivin, a key protein in carcinogenesis almost uniformly overexpressed in these tumours (Hernandez et al., 2011). In fact, survivin overexpression has been associated with cancer progression, chemoresistance and angiogenesis, with apoptosis inhibition and shortened survival in cancer patients (Ling et al., 2010; Hernandez et al., 2011; Souza et al., 2011; Singh et al., 2015; Chen et al., 2017). In addition, inhibition of survivin improved the uptake of drugs by cancer cells with a notable enhancement of their anticancer activity (Singh et al., 2015). As p53 inhibits survivin transcriptional expression, its inactivation has been associated with survivin overexpression in cancer cells (Ling et al., 2010). On this basis, inhibition of survivin by the p53‐activating agent SYNAP may represent a promising approach to the sensitization of CRC to cancer therapy.

It is also interesting to note that, in accordance with earlier work (Tovar et al., 2006; Li et al., 2015), we confirmed that the pronounced anti‐proliferative activity of the MDM2‐only inhibitor nutlin‐3a was not associated with cell death, particularly in human wt p53‐expressing colon cancer cells. Instead, nutlin‐3a induced a marked cell cycle arrest, which has proved to be useful in cyclotherapy, protecting normal cells from undesirable side effects of cytotoxic drugs (Rao et al., 2013). However, in tumours harbouring wt p53, this may reduce the efficacy of cytotoxic drugs (such as nucleoside analogues, like 5‐FU). This may also allow a rescue of cancer progress as, after cessation of treatment, cells can restart growth with only a slight delay (Li et al., 2015). Here, we have provided evidence for the p53‐dependent, cytotoxic effect of SYNAP, either alone or in synergistic combination with other chemotherapeutic drugs (including doxorubicin and 5‐FU). Furthermore, no recovery of cancer cell proliferation was observed after removal of treatment, which may predict a reduced probability of tumour recurrence.

In conclusion, this work reports a new selective activator of p53 and provides relevant insights into its molecular mechanism of action and anticancer activity, both as monotherapy and in combination with known chemotherapeutic agents. Besides its promising application in cancer therapy, particularly of CRC, SYNAP may represent a lead compound for the further development of new and improved p53‐activating agents, particularly those that are inhibitors of the interaction of p53 with MDM2 and MDMX.

Author contributions

L.R. performed experiments, analysed the data and wrote the manuscript; M.E. performed the synthesis and characterization of SYNAP; J.S., J.B.L. and M.G.A. contributed to the experimental work and analysed the data; M.M.M.S. conceived the design and synthesis of SYNAP and analysed the data; L.S. conceived and designed the study, provided financial support for the study, analysed the data and wrote the manuscript. All authors read and approved the final version of the manuscript.

Conflicts of interest

The authors declare no conflicts of interest.

Declaration of transparency and scientific rigour

This http://onlinelibrary.wiley.com/doi/10.1111/bph.13405/abstract acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organisations engaged with supporting research.

Supporting information

Table S1 List of antibodies used in Western blots in this work.

Raimundo, L. , Espadinha, M. , Soares, J. , Loureiro, J. B. , Alves, M. G. , Santos, M. M. M. , and Saraiva, L. (2018) Improving anticancer activity towards colon cancer cells with a new p53‐activating agent. British Journal of Pharmacology, 175: 3947–3962. 10.1111/bph.14468.