N-Alkylated Monoterpene Indole Alkaloid Derivatives as Chemotherapy Adjuvants That Overcome P-glycoprotein-Mediated Multidrug Resistance in Cancer Cells, Spheroids and Patient-Derived Colorectal Cancer Organoids

Abstract

Purpose

Chemotherapy efficacy is often limited by multidrug resistance (MDR), largely driven by P-glycoprotein (P-gp). This study aimed to evaluate previously prepared N-alkylated dregamine and tabernaemontanine derivatives as potential adjuvants to overcome P-gp–mediated chemoresistance.

Patients and Methods

Twenty-one N-alkylated derivatives (1–21) were tested for cytotoxicity and chemosensitizing activity in sensitive and doxorubicin (DOX)-resistant ovarian (A2780, A2780DOX) and breast cancer cell lines (MCF-7, MCF-7R), as well as in nonmalignant kidney epithelial cells (hRPTEC). P-gp modulation was assessed using rhodamine 123 accumulation and Pgp-Glo Assay Systems kit. The activity of the most promising compound was further evaluated in 3D monocellular and multicellular spheroids, patient-derived colorectal cancer organoids (PDOs), and peripheral blood mononuclear cells (PBMCs). Computational docking was performed to support mechanistic insights.

Results

The derivatives exhibited moderate cytotoxicity toward cancer cells and low toxicity toward hRPTEC. All compounds sensitized resistant cell lines, with the most active ones reducing DOX IC₅₀ by >20-fold at 10 μM and >8-fold at 1 μM. For most derivatives, increased accumulation of rhodamine 123 confirmed modulation of P-gp efflux activity. N-(4-methoxybenzyl) dregamine (compound 11) showed the strongest potential, enhancing DOX sensitivity >7-fold in monocellular and >4-fold in multicellular spheroids. In combination with DOX, compound 11 also increased the sensitivity of colorectal cancer PDOs, thereby reducing their viability from 70% to 30%. No increased toxicity of DOX was observed in PBMCs.

Conclusion

N-alkylated monoterpene–indole alkaloids, particularly compound 11, effectively overcome P-gp–mediated chemoresistance without inducing cytotoxicity in healthy cells and represent promising candidates for further in vivo evaluation.

Plain Language Summary

Cancer cells can become resistant to treatment in several ways. One common mechanism is the increased production of protective proteins. One such protein, P-glycoprotein (P-gp), works like a pump that removes chemotherapy drugs from inside cancer cells, making treatment less effective. A possible solution is to block P-gp activity. However, earlier P-gp inhibitors tested in clinical trials were too toxic and failed. This study therefore focused on new compounds derived from natural substances called dregamine and tabernaemontanine, isolated from the African plant Tabernaemontana elegans. These compounds may modulate P-gp activity while remaining safe for healthy cells. The results showed that these compounds reduced chemotherapy resistance in ovarian and breast cancer cells. Most of them were not toxic to healthy kidney cells. We also confirmed that they decrease P-gp activity. The most effective compound, N-(4-methoxybenzyl) dregamine, was further tested in advanced 3D tumor models called spheroids, which better reflect real tumors. In these models, it increased the sensitivity of resistant cancer cells to chemotherapy without affecting healthy cells. Similar effects were observed in colorectal cancer models derived from colorectal cancer patient samples. Importantly, no increased toxicity was observed in isolated immune cells. Finally, computer modeling suggested how this compound can bind to P-gp and modulate its function. These findings may support the development of safer treatments for drug-resistant cancers.

Graphical Abstract

Introduction

Chemotherapy, alongside surgery and radiotherapy, is one of the fundamental approaches used in the treatment of cancer.1 Doxorubicin (DOX) is among the most widely used agents in both adult and pediatric oncology. DOX inhibits topoisomerase II and creates reactive oxygen species, thereby reducing the growth of cancer cells.2 DOX is also frequently used alongside other chemotherapeutic drugs to enhance treatment effectiveness.3 Nevertheless, an early diagnosis remains crucial for successful therapy.4 However, due to unclear initial symptoms, several types of cancers often goes undetected until an advanced stage, which complicates treatment and contributes to poor outcomes, even when treated with chemotherapy.5 Furthermore, although initial chemotherapy can be effective, disease relapse frequently occurs. Notably, up to 70% of patients with ovarian cancer relapse within 18 months, and subsequent treatment is often only partially successful.5–7

The effectiveness of chemotherapy is often limited by multidrug resistance (MDR), which enables cancer cells to evade the effects of multiple chemotherapeutics.8,9 Several mechanisms contribute to the development of drug-resistant tumors with one of the main factors being the increased expression of ATP-binding cassette (ABC) transporters, particularly P-glycoprotein (P-gp, ABCB1), breast cancer resistance protein (BCRP, ABCG2), and multidrug resistance protein 1 (MRP1, ABCC1).10 Located in the cell membrane, these transporters handle a wide range of substrates and transfer them from the cytoplasm to the extracellular space. Some of these transporters have received considerable attention, particularly because of their strong association with MDR.11

The main ABC transporter involved in MDR is P-gp.12,13 P-gp consists of two transmembrane domains containing 12 transmembrane helices and two nucleotide-binding domains that drive ATP-dependent drug export. Structural studies, including cryo-EM analyses of PDB 8SA1, have identified an extensive polyspecific binding cavity within the transmembrane region, where hydrophobic and aromatic residues mediate interactions with a broad range of substrates and modulators.14,15 Due to that broad substrate specificity, P-gp is capable of transporting many molecules, including steroid hormones, immunosuppressants, antibiotics and chemotherapeutic drugs.16,17 Several generations of inhibitors have been developed in an attempt to suppress MDR by reducing P-gp activity. However, these inhibitors have shown significant side effects, notably high toxicity and low specificity, meaning that none of them have passed a clinical trial yet.18 For example, first-generation P-gp inhibitors such as verapamil and cyclosporine A were limited by cardiac and immunosuppressive side effects, whereas second-generation compounds such as valspodar and dexverapamil increased the risk of drug-drug interactions through inhibition of cytochrome P450 enzymes.19,20 Even newer P-gp inhibitors, such as tariquidar, may show insufficient efficacy at clinically relevant concentrations, particularly in cancer cells exposed to chemotherapy-induced stress.21 Because of this, research has increasingly focused on the development of more clinically useful P-gp inhibitors based on natural product-derived compounds, such as alkaloids.19,22

Dregamine and tabernaemontanine, isolated from the roots of the African medicinal plant Tabernaemontana elegans (Stapf), are two epimeric monoterpene indole alkaloids, whose derivatives represent a great potential as chemotherapy adjuvants capable of modulating P-gp activity without significant adverse effects. Their antitumor properties have previously been demonstrated in the treatment of colon, liver and hepatocellular cancer cells.23 Paterna et al24 described the antitumor effects of hydrazone derivatives on intestinal carcinoma cells. More recently, we reported a dregamine hydrazone derivative (BBIT20) that inhibited homologous DNA repair in triple-negative breast cancer and advanced ovarian cancer cells. Moreover, BBIT20 also demonstrated to be highly effective in overcoming pancreatic ductal adenocarcinoma therapeutic resistance.25,26 To expand our in-house library of compounds sharing the indole alkaloid scaffold, the indole nitrogen was alkylated by reaction with various aliphatic and aromatic halides, yielding a new series of N-alkylated dregamine and tabernaemontanine derivatives. These derivatives were subsequently evaluated for their effects on MDR using resistant human colon adenocarcinoma cells and transfected mouse lymphoma cells, overexpressing P-gp.27 In this work, the evaluation of their anti-MDR effects was extended to resistant clones of breast and ovarian cancer cells, using also advanced cell models.

Based on the mentioned premises, we hypothesized that N-alkylated dregamine and tabernaemontanine derivatives could overcome P-gp-mediated multidrug resistance by reducing drug efflux activity and enhancing chemotherapy efficacy even in advanced cell models. Thus, in this work, the N-alkylated monoterpene indole alkaloids 1–21, previously prepared by Cardoso et al27 were evaluated in terms of their cytotoxicity, chemosensitizing properties, and effects on P-gp activity in resistant ovarian and breast cancer cell lines. We further evaluated the sensitizing effect of the most potent adjuvant, compound 11, using advanced cellular models — monocellular spheroids derived from resistant cancer cells and multicellular spheroids enriched with fibroblasts and adipocytes. We also assessed the impact of this adjuvant therapy on the viability of healthy peripheral blood mononuclear cells (PBMCs). Finally, the significant sensitizing activity of compound 11, combined with its lack of cytotoxicity, was confirmed in organoids derived from intestinal tumors of patients (PDOs). Taken together, using both basic and clinically relevant models, we demonstrated that N-alkylated dregamine and tabernaemontanine derivatives, when applied as adjuvants, markedly reduce tumor cell resistance to chemotherapy while exhibiting no adverse cytotoxic effects in noncancerous cells.

Materials and Methods

Reagents and Chemicals

Dimethylsulfoxide (DMSO), Dulbecco’s Modified Eagle’s medium–high glucose (DMEM), fetal bovine serum (FBS), 100× antibiotic antimycotic solution (penicillin 10000U/mL; streptomycin 10 mg/mL; amphotericin B 25 µg/mL), doxorubicin hydrochloride (DOX), trypsin-2,2’,2”,2”’-(ethane-1,2-diyldinitrilo)tetraacetic acid solution, phosphate-buffered saline (PBS, pH 7.4), poly(2-hydroxethyl methacrylate) (poly-HEMA), 3-isobutyl-1-methylxanthine, dexamethasone, human insulin, resazurin sodium salt, Y-27632 dihydrochloride, nicotinamide, N-acetylcysteine amide, dispase II, A 83–01, human Gastrin I, rhodamine 123, Hoechst 33342, calcein acetoxymethyl ester, prazosin, verapamil, Lymphocyte Separation Medium, Roswell Park Memorial Institute (RPMI) medium and HEPES buffer were purchased from Merck (Germany). ProxUP medium was purchased from Evercyte (Austria). CellTiter-Glo 3D Cell Viability Assay and Pgp-Glo Assay Systems kit was purchased from Promega Corporation (USA). Cultrex UltiMatrix Reduced Growth Factor Basement Membrane Extract was purchased from Bio-Techne (USA). GlutaMAX Supplement, N-Acetyl-L-cysteine, TrypLE Express Enzyme, Advanced DMEM/F-12, B-27 Supplement (50X), Human EGF Recombinant Protein, Human Noggin Recombinant Protein, Human FGF-basic (FGF-2/bFGF) Recombinant Protein, Human R-Spondin 1 Recombinant Protein, N-2 Supplement (100X), Wnt Surrogate-Fc Fusion Recombinant Protein, Collagenase type IV, CellTracker Deep Red, CellTracker Violet BMQC and CellTracker Blue CMF2HC were purchased from Thermofisher (USA). SepMate tubes were purchased from Scintila (Czechia). Primocin was purchased from Baria (Czechia). 40 μm cell strainers were purchased from Starlab (Germany). A lentiviral transduction system for transducing the abcg2 gene was purchased from VectorBuilder (USA).

Preparation of Dregamine and Tabernaemontanine N-Alkylated Derivatives for Biological Assays

The dregamine and tabernaemontanine N-alkylated derivatives were prepared as previously demonstrated.27 The structures of compounds are shown in Figure 1. The compounds were dissolved in dimethylsulfoxid (DMSO) at a stock concentration of 10 mM.

Figure 1.

Compounds 1–21 prepared by derivatization of tabernaemontanine (β-Et) and dregamine (α-Et). The scheme was generated using ChemDraw 22.2.0 software (USA).

Cell Lines and Their Maintenance

Human ovarian cancer cell line (A2780), its DOX-resistant subline (A2780DOX) and human dermal fibroblasts (HDF), Madin-Darby canine kidney cells (MDCK II) were purchased from Merck (Germany). Human renal proximal tubular epithelial cell line (hRPTEC) was purchased from Evercyte (Austria). Human breast carcinoma cell line (MCF-7) and its DOX-resistant subline (MCF-7R) was kindly provided by prof. RNDr. Jan Kovář, DrSc. Human glioblastoma cell line overexpressing MRP1 (U87MG-DOX) was kindly provided by PharmDr. Helena Mertlíková-Kaiserová, Ph.D. Embryonic fibroblast cell line (3T3-L1) was purchased from the American Type Culture Collection (USA). A cell line overexpressing the abcg2 (BCRP) gene (MDCK II_BCRP) was established using a commercially available lentiviral transduction system (VectorBuilder, USA). The use and handling of non-commercial cell lines were approved by the Ethics Committee of UCT Prague (ID: EC 10/26).

The A2780, A2780DOX, MCF-7, MCF-7R, HDF, MDCK II_BCRP and 3T3-L1 cell lines were cultured in DMEM supplemented with 10% FBS and 1× antibiotic-antimycotic solution. The hRPTEC cell line was cultured in ProxUP medium. All cell lines were maintained under standard conditions (humidified atmosphere, 5% CO₂, 37 °C) and were used for experiments at passage numbers between 5 and 20. During the cultivation of resistant sublines, the medium was treated once a week with 80 nM DOX to maintain the resistant properties. In accordance with internationally accepted recommendations for good cell culture practice, all of the cell lines mentioned were regularly authenticated using short tandem repeat (STR) profiling to prevent cell line misidentification or cross-contamination. All cell lines were negative for the mycoplasma presence.

Cytotoxicity Tested in Cell Monolayer

Cytotoxicity of compounds was tested as previously described.16 Briefly, the cells were seeded at a concentration of 10,000 cells/well in 96-well plates. After 24 hours of incubation, the old medium was replaced with new one and the tested concentration of compound was added. The plate was incubated for a further 72 hours and then cell viability was measured by a resazurin-based assay.28 The medium was removed from the plates and after that 100 µL of a resazurin solution (0.03 mg/mL in 1× PBS) was pipetted into each well. After 30–60 min incubation (depending on the cell line), the fluorescence signal (ex./em., 560/590 nm) was measured using a SpectraMax i3× Multi-Mode Microplate reader (Molecular Devices, USA). The concentration of the tested compounds that reduced cell growth by 50% (IC₅₀) was used to express cell growth inhibition. The toxicity of the compounds was evaluated based on their IC₅₀ values. Concentrations that did not reduce cell viability below 80% were considered non-toxic.29 In order to compare the toxic effects of compounds on tumor and non-tumor cell lines, we calculated the selectivity index (SI) using the following formula:30

Similarly, we calculated collateral sensitivity (CS) in order to compare the toxic effects of the compounds on resistant and sensitive tumor cell lines:30

Chemosensitization of a DOX-Resistant Cell Lines

Chemosensitization of a DOX-resistant cell lines was performed as previously described.31 In short, the cells were seeded at a concentration of 10,000 cells/well in 96-well plates. After 24 hours of incubation, the old medium was replaced with new one containing a non-toxic concentration of the tested compound (10 µM, 1 µM or 0.1 µM). The compounds 8 and 12 that showed toxicity at 10 µM were tested at 5 µM instead. For the reference control, only the vehicle solvent was added to the medium. A DOX concentration series was then added to the plate. After a further 72 h of incubation, cell viability was determined using a resazurin-based assay. The IC₅₀ of DOX was then calculated from the viability results. The effect of tested substance was quantified by the following formula:31

Compounds that showed fold equal to or greater than 3 were successively tested at ten times lower concentration. Based on the obtained data, the therapeutic index (TI) was calculated according to the formula:

where IC₅₀ of compound in hRPTEC cell line denotes the concentration of a compound inhibiting the growth of 50% of non-tumor cells (hRPTEC) and tested concentration of compound is the lowest tested concentration showing a fold equal to or greater than 3.

Substrate Accumulation

The modulation of P-gp, BCRP or MRP1 activity was evaluated by monitoring the accumulation of the fluorescent substrates rhodamine 123, Hoechst 33342 or calcein acetoxymethyl ester, respectively.32–34 The A2780DOX, MDCK II_BCRP or U87MG-DOX cells were seeded in a 96-well plate at a concentration of 100,000 cells/well and then left to incubate for 24 hours. The medium was then substituted with a new one containing fluorescent substrate (5 μmol/L rhodamine 123, 5 μmol/L Hoechst 33342, 1 μmol/L calcein acetoxymethyl ester) and the test compound (or vehicle solvent DMSO serving as a reference). The plate was cultured for 1 h and then washed three times with 1× PBS. Finally, the cells were lysed by adding 100 µL of distilled water. The amount of accumulated substrate was determined spectrophotometrically (ex./em., 485/535 nm for rhodamin 123; 352/454 nm for Hoechst 33342 and 485/525 nm for calcein acetoxymethyl ester) by SpectraMax i3× Multi-Mode Microplate reader. The percentages of substrate accumulation in cells were calculated from the fluorescence signal, the background signal (BS, cells without substrate), and non-treated cells (NT) according to the following equation:

Characterization of Compound 11 Interaction with P-Gp Using the Pgp-Glo Assay Systems Kit

The interaction of the compound with P-gp was characterized using the commercial Pgp-Glo Assay Systems kit (Promega Corporation, USA), performed according to the manufacturer’s instructions with minor modifications. Briefly, 20 μL of diluted P-gp membranes was mixed with 20 μL of compound 11 (25 μM), verapamil (25 μM), or Na₃VO₄ (250 μM) solution in a plain, untreated white polystyrene plate and incubated for 5 min at 37 °C. The reaction was then initiated by adding 10 μL of MgATP solution (25 mM) to each well, and the plate was incubated for 60 min at 37 °C. After stopping the reaction with 50 μL of ATP Detection Reagent and incubating the plate for 20 min at room temperature, luminescence was measured using a SpectraMax i3x Multi-Mode Microplate Reader (Molecular Devices, USA) with an integration time of 1000 ms. The percentage of ATP remaining in the solution after the reaction was calculated from the mean of relative luminescence units of the basal membrane activity (RLU_NT) and the sample relative luminescence units (RLU_sample) according to the following equation:

Computational Modeling of Compound 11 Interactions with P-Gp

Two low-energy conformers of N-(4-methoxybenzyl) dregamine (compound 11) consistent with the experimentally determined stereochemistry (all chiral centers S except C20, R) were obtained from the PubChem database. These conformers are in agreement with ^1H NMR data, in particular the unusually upfield-shifted methoxy signal, indicating placement of the ester methoxy group within the shielding cone of the indole ring. Conformations incompatible with these NMR constraints (eg, partially uncurled seven-membered ring geometries) were therefore excluded. For computational modeling, both PubChem conformers were considered with alternative pyramidalization of the non-planar nitrogen atom and with either neutral or monoprotonated states, reflecting possible charge states in hydrophobic versus aqueous environments. Structural models of human P-gp (ABCB1) were selected from cryo-EM entries in the Protein Data Bank. Four inward-open human P-gp structures (PDB IDs: 8SA1, 8SB7, 8SBA, 8SB9), each co-resolved with substrate or inhibitor and two Mg–ATP complexes, were used. These conformations represent a functionally relevant transport cycle state in which inhibitors are expected to bind. All four structures were used for ensemble docking. Docking was performed using AutoDock Vina.35 All prepared ligand variants were docked into each P-gp structure. Resulting poses consistently clustered deep within the transmembrane channel, with the most favorable solutions located at the distal end of the inward-open cavity. Neutral ligand states yielded slightly more favorable scores than the protonated forms. Among the protein models, PDB entry 8SA1 provided the best overall docking results and was selected for subsequent molecular dynamics simulations. The top-ranked docking pose of neutral compound 11 in the 8SA1 structure was used as the starting configuration for all-atom molecular dynamics simulations. Simulations were carried out using OpenMM36 with explicit solvent and counterions. Standard equilibration and production protocols were applied.

Preparation of Monocellular and Multicellular Spheroids

Ultra-low attachment plates were prepared by coating U-bottom plates with poly(2-hydroxyethyl methacrylate) (poly-HEMA), as described by Kuroda et al37 Monocellular spheroids were prepared from resistant tumor cell lines A2780DOX or MCF-7R. The cells were seeded into prepared ultra-low attachment plates at a total concentration of 10,000 cells/well. The plates were then centrifuged at 1000 g for one minute and subsequently incubated at 37 °C in 5% CO₂ for 24 hours, for the spheroids to be formed. Multicellular spheroids were prepared as described by Flörkemeier et al38 Multicellular spheroids were prepared from tumor cell lines (A2780ADR or MCF-7R), fibroblasts (HDF), and adipocytes (dif3T3-L1, differentiated from 3T3-L1 cells). According to Vrzáčková et al,39 3T3-L1 cells were differentiated into mature adipocytes dif3T3-L1 by culturing in DMEM with 6 g/L HEPES buffer, 0.4 μg/mL dexamethasone, 0.11 g/L 3-isobutyl-1-methylxanthine and 10 μg/mL insulin. A2780ADR or MCF-7R cells were mixed with HDF and dif3T3-L1 cells at a ratio of 1:5:5, giving a total concentration of 100,000 cells/mL. Then, 100 μL of this mixture was pipetted into an ultra-low attachment plates, resulting in a total of 10,000 cells/well. The plates were then centrifuged at 200 g for 5 minutes and subsequently incubated at 37 °C in 5% CO₂ for 24 hours.16

Cytotoxicity Tested in Spheroids

Cytotoxicity tested in spheroids was performed as follows: 50 µL of the total 100 µL of medium was carefully removed from the ultra-low attachment plate wells containing spheroids. Then, 50 µL of fresh medium containing the tested compound at double the required concentration was added. Spheroids were then left to incubate for 72 hours before their viability was assessed by the CellTiter-Glo 3D Cell Viability Assay, performed according to the manufacturer’s instructions. Cytotoxicity was determined by comparison with untreated control wells and expressed as IC₅₀ values.

Chemosensitization of DOX-Resistant Spheroids

After the spheroids were formed, 50 µL of the total 100 µL of medium was carefully removed from each well of the ultra-low attachment plate. Subsequently, 50 µL of fresh medium containing DOX and the tested compound (both at twice the final desired concentration) was added. The spheroids were then incubated for 72 hours. After incubation, images of the spheroids were captured using a EUREKAM 5.0 Plus digital camera (KRD, Czechia), and viability was assessed using the CellTiter-Glo 3D Cell Viability Assay, performed according to the manufacturer’s instructions. Enhancement of DOX efficiency in spheroids was determined by comparison with untreated control and expressed as IC₅₀ and fold values.

Analysis of Spheroid Composition by Flow Cytometry

The cell lines used to prepare multicellular spheroids for flow cytometry analysis were first stained with CellTracker fluorescent probes.40 CellTracker Blue CMF2HC (HDF), CellTracker Violet BMQC (dif3T3-L1) and CellTracker Deep Red (A2780DOX, MCF-7R) were used for staining according to the manufacturer’s protocol. The spheroids were then prepared and treated as described above. Prior to flow cytometry analysis, the spheroids (n=12) were dissociated into monocellular suspension by adding TrypLE Express Enzyme and gentle mechanical pipetting. The cells were analyzed using a BD FACSAria™ III flow cytometer. The excitation wavelength was set to 355 nm, and emission was detected in the range of 425–475 nm for CellTracker Blue CMF2HC, 405 nm excitation with 500–550 nm emission for CellTracker Violet BMQC and 640 nm excitation with 655–685 nm emission for CellTracker Deep Red, respectively. The measurement was carried out until 10,000 events were recorded. Spheroids without any staining were used in the analysis to account for cell autofluorescence. Data acquisition and quantification were performed using BD FACSDiva™ 8.0 software (USA). Analysis of spheroid composition was determined as the percentage representation of each cell type.

Cytotoxicity Tested in PBMCs

PBMCs were isolated from peripheral blood obtained from a healthy donor. PBMCs were taken and maintained in accordance with the ethical principles of UCT Prague, established by the Code of Ethics of UCT Prague (Ethics committee approval ID: EC 1/2026) and with the principles of the Helsinki Declaration of 1964 (Ethical principles Research Involving Human Subjects). All samples were obtained with informed consent. Blood samples were collected into EDTA-containing tubes to prevent coagulation and were processed within 2 hours of collection. PBMCs isolation was carried out using density gradient centrifugation. Briefly, 15 mL of lymphocyte separation medium was added to a SepMate tube, and whole blood was carefully layered on top. The samples were centrifuged at 1200 × g for 15 minutes at 20 °C. Following centrifugation, the upper phase containing plasma, the PBMCs fraction, and part of the separation medium was collected and transferred into a new 50 mL conical tube. The collected fraction was diluted with PBS to a final volume of 50 mL and centrifuged at 500 × g for 5 minutes at 20 °C. The resulting cell pellet was resuspended in RPMI culture medium and adjusted to a final concentration of 7.5×10⁵ cells/mL. Cells were seeded into 96-well plates, followed by the addition of the tested compounds. DOX was applied in a concentration range of 1.6 µM to 200 µM, either alone or in combination with compound 11. For combination experiments, compound 11 was added at a fixed, non-toxic effective concentration of 10 µM. Compound 11 was also tested individually at the same concentration. Cells were incubated for 72 hours under standard culture conditions. Cell viability was assessed using a resazurin-based assay. A resazurin stock solution prepared at tenfold concentration was added to the wells together with culture medium, resulting in dilution to the final working concentration (1×). Plates were incubated for 4 hours, and fluorescence was measured at excitation/emission wavelengths of 560/590 nm. Cytotoxicity was determined by comparison with untreated control wells and expressed as IC₅₀ values.

Chemosensitization of PDOs

A PDOs were prepared using the procedure described by Sato et al,41 with some minor modifications. Surgically resected colorectal cancer tissues were obtained from three patients at the General University Hospital in Prague. Collection of patient tissue was taken in accordance with the ethical principles of UCT Prague (Ethics committee approval ID: EC 1/2026), the ethical principles of the General University Hospital in Prague (Ethics committee approval ID: 60/24 S-IV) and with the principles of the Helsinki Declaration of 1964 (Ethical principles Research Involving Human Subjects). All samples were obtained with informed consent. Immediately after resection, the tissue sample was transported to the laboratory in cold DMEM/F-12 medium supplemented with HEPES (10 mM), primocin (100 μg/mL) and Y-27632 (10 μM). In the laboratory, the sample was first washed in PBS containing 1× antibiotic–antimycotic solution, then rinsed in PBS alone, and finally washed in PBS supplemented with primocin (100 μg/mL). Fatty and necrotic tissue was removed, and the remaining material was cut into small pieces (< 1 mm²) using a scalpel. The tissue fragments were then incubated for 1 h in a digestion solution prepared from DMEM/F-12 medium, HEPES (10 mM), collagenase type IV (2 mg/mL), dispase (125 μg/mL), and primocin (100 μg/mL), and subsequently filtered through a 40 μm cell strainer. The cells were centrifuged (200 g, 5 min), resuspended in cold Matrigel (1000 cells per 25 μL), and pipetted into a 24-well plate. The Matrigel domes were allowed to solidify for several minutes in an incubator at 37 °C, after which organoid culture medium prepared according to Sato et al41 was added to each well. For treatment experiments, organoid culture medium containing DOX (10 μM), compound 11 (10 μM), or their combination was added to the established PDOs. Vehicle solvent (DMSO) served as the control. The PDOs were incubated for 72 h, and their viability was assessed using the CellTiter-Glo 3D Cell Viability Assay performed according to the manufacturer’s instructions.

Data Processing and Statistical Analysis

The experiments were carried out in a respective number (n) of repetitions. The IC₅₀ values were calculated by non-linear regression using Graph Pad Prism 7 software (GraphPad Software, USA) according to the formula:

Data are presented as average with a standard error of the mean (SEM). The statistical significance of the data was evaluated using a t-test (two-tailed distribution, heteroscedastic type). The differences between the groups were analysed using one-way analysis of variance (ANOVA) and Tukey’s post hoc test (P < 0.05) processed by Graph Pad Prism 7 software.

Cropping of spheroid images and addition of scale bars were performed using Fiji (ImageJ, version 1.54, USA).

Results

Cytotoxicity and Anticancer Potential of Dregamine and Tabernaemontanine N-Alkylated Derivatives

We evaluated the anticancer activity and toxic potential of the previously prepared dregamine and tabernaemontanine derivatives27 (compounds 1–21) in human ovarian carcinoma cells (A2780), breast carcinoma cells (MCF-7), their resistant sublines (A2780DOX, MCF-7R), and non-cancerous human renal proximal tubular epithelial cells (hRPTEC). The effect of each compound was assessed after 72 hours of exposure and expressed as the concentration halving the cell growth (IC₅₀). The resulting values for each compound and cell line are summarized in Table 1. Most of the compounds exhibited moderate cytotoxicity toward ovarian and breast cancer cells, with IC₅₀ values in the tens of micromolar range. In most cases, we found lower IC₅₀ values in the resistant A2780DOX and MCF-7R sublines compared with their parental cell lines. In contrast, only six compounds showed IC₅₀ values below 100 μM in the non-cancerous hRPTEC line, indicating generally low toxicity toward healthy kidney cells and supporting the potential of these compounds to preferentially target cancer cells.

Table 1.

Concentrations of Compounds 1–21 Halving the Viability (IC₅₀, μM) of Human Renal Proximal Tubular Epithelial Cells (hRPTEC), Ovarian Cancer Cell Line (A2780) and Their Resistant Subline (A2780DOX) and Breast Carcinoma Cell Line (MCF-7) and Their Resistant Subline (MCF-7R)

Compound	IC50 [μM]
	hRPTEC	A2780	A2780DOX	MCF-7	MCF-7R
1	>100	>100	>100	>100	41.6 ± 5.1 e-g
2	46.3 ± 4.2 d	>100	33.1 ± 0.6 f-g	76.1 ± 1.7 g	44.7 ± 2.1 f-h
3	>100	>100	54.4 ± 1.9 k	38.7 ± 1.9 c	54.6 ± 7.0 g-i
4	>100	>100	44.4 ± 3.9 h-i	64.6 ± 3.5 e-f	38.7 ± 4.0 d-f
5	>100	>100	67.2 ± 3.4 L	>100	42.1 ± 4.8 f-g
6	12.6 ± 0.6 b	30.5 ± 1.8 b-c	13.8 ± 1.3 b-c	>100	>100
7	27.3 ± 1.8 c	49.9 ± 0.1 d	13.8 ± 0.5 b-c	51.8 ± 2.0 d	28.1 ± 1.8 a-e
8	6.2 ± 0.1 a-b	25.1 ± 0.7 b	10.9 ± 0.1 b	>100	58.0 ± 4.7 h-i
9	28.6 ± 1.5 c	34.1 ± 2.3 c	16.0 ± 0.1 c	29.7 ± 0.2 b	28.0 ± 0.9 a-e
10	>100	81.8 ± 4.7 f	35.9 ± 0.5 g	64.1 ± 1.3 e-f	44.7 ± 2.1 f-h
11	>100	68.7 ± 5.3 e	30.9 ± 1.4 f	48.7 ± 2.3 d	22.0 ± 2.6 a-b
12	>100	76.1 ± 4.0 f	10.8 ± 0.5 b	>100	>100
13	>100	51.8 ± 1.3 d	32.5 ± 0.4 f-g	62.3 ± 1.5 e	19.1 ± 2.6 a
14	>100	49.9 ± 0.2 d	25.2 ± 1.0 d-e	>100	>100
15	>100	30.9 ± 1.1 b-c	42.3 ± 0.2 h	>100	77.6 ± 9.8 j-k
16	40.7 ± 4.2 d	45.0 ± 1.1 d	17.1 ± 0.5 c	42.4 ± 0.9 c	26.8 ± 3.6 a-d
17	>100	97.5 ± 0.3 g	29.5 ± 0.9 e-f	83.2 ± 3.1 h	65.4 ± 0.3 i-j
18	>100	>100	14.1 ± 0.9 b-c	>100	>100
19	>100	48.1 ± 0.8 d	48.5 ± 3.3 i-j	>100	84.1 ± 6.7 k
20	>100	>100	50.7 ± 0.2 j-k	>100	82.0 ± 3.7 k
21	>100	34.2 ± 1.0 c	22.1 ± 0.6 d	68.8 ± 0.5 f	35.9 ± 6.8 c-f
VER	>100	>100	41.6 ± 1.2 h	65.9 ± 1.5 e-f	35.4 ± 0.5 b-f
DOX	0.069 ± 0.006 a	0.017 ± 0.002 a	1.4 ± 0.2 a	0.12 ± 0.02 a	23.0 ± 2.0 a-c

Notes: VER (verapamil), DOX (doxorubicin). Values represent the mean of IC₅₀ ± SEM calculated from three replicates. Data were analyzed using one-way ANOVA followed by Tukey’s post hoc test; P < 0.05. Statistically different levels of significance within the cell lines are indicated by superscript letters. Compounds sharing the same superscript letter are not significantly different, whereas compounds with different superscript letters differ significantly (P < 0.05).

To compare the selectivity of the compounds toward cancer cell lines, we calculated Selectivity Index (SI) values as the ratio of toxicity against control and cancer cells, presented in Table 2. Higher SI values indicate greater selectivity for inhibiting the growth of cancer cells. Among the tested derivatives, compounds 11, 12, 13, 14, 18 and 21 demonstrated significant selectivity, however, their anticancer activity was observed only in the resistant cancer cell lines. To further compare toxicity between resistant and sensitive tumor cells, we calculated collateral sensitivity (CS) values. The CS values show that compounds 12 and 18 exhibit markedly higher toxicity toward the A2780DOX resistant line compared with the parental A2780 cells, whereas they did not show toxicity in the MCF-7 and MCF-7R cells.

Table 2.

Selectivity Index (SI) and Collateral Sensitivity (CS) Calculated from the Concentration of Compounds Halving the Viability (IC₅₀, μM) of Ovarian (A2780) and Breast (MCF-7) Carcinoma Cell Lines, Their Resistant Sublines (A2780DOX, MCF-7R), and Human Renal Proximal Tubular Epithelial Cells (hRPTEC)

Compound	Ovarian Cancer Cell Line A2780					Breast Cancer Cell Line MCF-7
	SI		SI		CS	SI	SI	CS
	hRPTEC/A2780		hRPTEC/A2780DOX		A2780/A2780DOX	hRPTEC/MCF-7	hRPTEC/MCF-7R	MCF-7/MCF-7R
1	–		–		–	–	>2.4	>2.4
2	<0.5		1.4 ± 0.1 c		>3.0	0.6 ± 0.1 b	1.0 ± 0.1 b	1.7 ± 0.1 b-d
3	–		>1.8		>1.8	>2.6	>1.8	0.7 ± 0.1 e-f
4	–		>2.3		>2.3	>1.5	>2.6	1.7 ± 0.2 b-d
5	–		>1.5		>1.5	–	>2.4	>2.4
6	0.4 ± 0.03 c-d		0.9 ± 0.1 d		2.2 ± 0.2 c-d	<0.1	<0.1	–
7	0.5 ± 0.04 b-d		2.0 ± 0.1 b		3.6 ± 0.1 b	0.5 ± 0.1 b	1.0 ± 0.1 b	1.8 ± 0.1 b-d
8	0.2 ± 0.01 d		0.6 ± 0.01 d		2.3 ± 0.1 c-d	<0.1	0.11 ± 0.01 c	>1.7
9	0.8 ± 0.1 b-c		1.8 ± 0.1 b		2.1 ± 0.1 c-e	1.0 ± 0.1 a	1.0 ± 0.1 b	1.1 ± 0.1 d-e
10	>1.2		>2.8		2.3 ± 0.1 c-d	>1.6	> 2.2	1.4 ± 0.1 c-e
11	>1.5		>3.2		2.2 ± 0.2 c-d	>2.1	> 4.5	2.2 ± 0.3 b
12	>1.3		>9.3		7.0 ± 0.5 a	–	–	–
13	>1.9		>3.1		1.6 ± 0.04 e-f	>1.6	>5.2	3.3 ± 0.5 a
14	>2.0		>4.0		2.0 ± 0.1 d-f	–	–	–
15	>3.2		>2.4		0.7 ± 0.03 h	–	>1.3	>1.3
16	0.9 ± 0.1 b		2.4 ± 0.3 a		2.6 ± 0.1 c	1.0 ± 0.1 a	1.5 ± 0.3 a	1.6 ± 0.2 b-d
17	>1.0		>3.4		3.3 ± 0.1 b	>1.2	>1.5	1.3 ± 0.1 c-e
18	–		>7.1		>7.1	–	–	–
19	>2.1		>2.1		1.0 ± 0.1 g-h	–	>1.2	>1.2
20	–		>2.0		>2.0	–	>1.2	>1.2
21	>2.9		>4.5		1.5 ± 0.1 f-g	>1.5	>2.8	1.9 ± 0.4 b-c
VER	–		>2.4		>2.4	>1.5	>2.8	1.9 ± 0.4 b-c
DOX	4.1 ± 0.4 a		0.05 ± 0.004 e		0.01 ± 0.001 i	0.6 ± 0.1 b	0.003 ± 0.0004 c	0.005 ± 0.001 f

Notes: VER (verapamil), DOX (doxorubicin). Data are presented as the mean value ± SEM from three replicates. The SI was calculated as a ratio of IC₅₀ against tumor cells and IC₅₀ against control cells. The CS was calculated as a ratio of IC₅₀ against resistant and IC₅₀ against sensitive tumor cell line. SI and CS values > 4 were considered highly effective and are marked in red. The symbol “–” indicates values that could not be calculated. Data were analyzed using one-way ANOVA followed by Tukey’s post hoc test; P < 0.05. Statistically different levels of significance within the cell lines are indicated by superscript letters. Compounds sharing the same superscript letter are not significantly different, whereas compounds with different superscript letters differ significantly (P < 0.05).

Chemosensitization of DOX-Resistant Cancer Cells by Dregamine and Tabernaemontanine N-Alkylated Derivatives Adjuvants

We evaluated the ability of the compounds to chemosensitize DOX-resistant cell lines (A2780DOX and MCF-7R). The resistance of these lines had been confirmed previously in our cytotoxicity measurements (see Tables 1 and 2). To reach the IC₅₀ in A2780DOX cells, a 100× higher concentration of DOX was required compared with A2780, and in MCF-7R cells an approximately 200× higher concentration was needed relative to the parental MCF-7 line.

For the combination experiments, the resistant cell lines were incubated for 72 hours with a DOX concentration range and a single selected concentration of the derivative. To ensure that the intrinsic toxicity of the compounds did not influence the measurements, we used a maximum concentration of 10 μM (and 5 μM for compounds 8 and 12 in A2780DOX due to their higher toxicity). To evaluate how the compounds influenced DOX efficacy, we calculated a fold value defined as IC₅₀ of DOX divided by IC₅₀ of DOX in the presence of derivative. These fold values are presented in Table 3. Compounds enhancing the cytotoxicity of DOX exhibited fold values higher than 1. We considered fold values > 3 to be highly effective.

Table 3.

The Increase in DOX Cytotoxicity in the Presence of a Nontoxic Concentration of the Compounds 1–21 (10, 1, or 0.1 μM) Expressed as the Ratio of the DOX IC₅₀ Values with and without the Derivative (Fold)

Comp.	A2780DOX			MCF-7R
	10 μM	1 μM	0.1 μM	10 μM	1 μM	0.1 μM
1	2.5 ± 0.2 i	–	–	1.6 ± 0.1 j-k	–	–
2	6.5 ± 0.4 g-h	0.9 ± 0.1 j	–	4.9 ± 0.6 e-g	0.9 ± 0.1 e-f	–
3	6.0 ± 0.4 g-i	1.3 ± 0.1 i-j	–	4.1 ± 0.4 f-h	1.0 ± 0.1 d-f	–
4	15.3 ± 0.4 c-f	1.7 ± 0.1 h-j	–	4.5 ± 0.3 f-h	0.6 ± 0.1 f	–
5	3.2 ± 0.3 h-i	2.4 ± 0.2 h-i	–	3.1 ± 0.1 g-k	0.7 ± 0.1 f	–
6	26.4 ± 1.3 a	8.2 ± 0.9 a	1.1 ± 0.1 b-c	2.6 ± 0.2 h-k	–	–
7	13.1 ± 1.9 e-f	1.6 ± 0.1 h-j	–	3.9 ± 0.5 g-I	2.1 ± 0.2 b-c	–
8#	8.0 ± 0.5 g	4.0 ± 0.4 f-g	1.1 ± 0.1 b-c	2.8 ± 0.1 h-k	–	–
9	19.2 ± 3.2 b-c	2.8 ± 0.2 g-h	–	15.6 ± 1.2 a	1.8 ± 0.2 c-d	–
10	14.5 ± 1.0 d-f	2.4 ± 0.2 h-i	–	7.6 ± 0.7 c-d	0.9 ± 0.1 e-f	–
11	22.6 ± 1.2 a-b	7.3 ± 0.3 a-b	2.1 ± 0.3 a	16.7 ± 0.8 a	9.4 ± 0.9 a	3.8 ± 0.4
12#	17.5 ± 2.8 c-d	5.8 ± 0.3 c-d	1.1 ± 0.1 b-c	3.4 ± 0.2 g-j	1.0 ± 0.1 d-f	–
13	16.3 ± 0.5 c-e	1.2 ± 0.2 i-j	–	6.8 ± 0.6 c-e	2.0 ± 0.3 c	–
14	16.8 ± 0.8 c-e	1.2 ± 0.2 i-j	–	1.2 ± 0.2 k	–	–
15	8.1 ± 1.1 g	2.6 ± 0.2 h	–	2.0 ± 0.1 i-k	–	–
16	12.3 ± 1.8 f	6.9 ± 1.0 b-c	0.9 ± 0.1 b-c	8.4 ± 1.6 b-c	1.7 ± 0.1 c-e	–
17	4.8 ± 0.5 g-i	4.2 ± 0.1 e-f	0.9 ± 0.1 b-c	7.4 ± 0.7 c-d	1.0 ± 0.1 d-f	–
18	18.5 ± 1.1 c	5.3 ± 0.4 d-e	0.9 ± 0.1 b-c	2.8 ± 0.5 h-k	–	–
19	3.3 ± 0.1 h-i	2.4 ± 0.4 h-i	–	6.0 ± 0.7 d-f	0.8 ± 0.1 f	–
20	2.3 ± 0.1 i	–	–	4.5 ± 0.2 f-h	1.9 ± 0.2 c	–
21	16.0 ± 0.4 c-f	6.3 ± 0.7 b-d	0.8 ± 0.1 c	10.1 ± 0.7 b	2.3 ± 0.4 b-c	–
VER	17.4 ± 1.4 c-d	5.4 ± 0.5 d-e	1.2 ± 0.1 b	6.5 ± 0.6 c-e	2.9 ± 0.3 b	0.7 ± 0.1

Notes: VER (verapamil), DOX (doxorubicin). # The highest tested concentration was 5 μM. The data are presented as mean values ± SEM from three replicates. Fold > 3 was considered highly effective and is marked in red. The symbol “–” indicates that the value was not measured. Data were analyzed using one-way ANOVA followed by Tukey’s post hoc test; P < 0.05. Statistically different levels of significance within the cell lines are indicated by superscript letters. Compounds sharing the same superscript letter are not significantly different, whereas compounds with different superscript letters differ significantly (P < 0.05).

All compounds at the highest tested concentrations (10 or 5 μM) increased DOX cytotoxicity in both resistant lines, and most achieved fold values >3. We subsequently tested compounds with fold value > 3 at a concentration of 1 μM. As expected, their activity decreased at this concentration. At a concentration of 1 μM, we observed fold values >3 for eight derivatives (6, 8, 11, 12, 16, 17, 18 and 21) in the A2780DOX line, and for only compound 11 in the MCF-7R line. Testing an even lower concentration of 0.1 μM revealed a notable effect for derivative 11, which, at this concentration, enhanced DOX sensitivity 2.8-fold in A2780DOX cells and 3.8-fold in MCF-7R cells. This derivative also shows strong potential for safe use in adjuvant therapy, as it demonstrated no toxicity toward the hRPTEC kidney cell line and therefore exhibited a therapeutic index (TI) greater than 1000 (Table 4).

Table 4.

The Therapeutic Index (TI) Calculated as the Ratio of the Compound Concentration Halving the Viability of Human Renal Proximal Tubular Epithelial Cells (hRPTEC) and the Lowest Concentration That Significantly (Fold Value > 3) Modulated DOX Efficiency in the Resistant Ovarian (A2780DOX) and Breast Cancer (MCF-7R) Cell Lines

Compound	TI
	A2780DOX	MCF-7R
1	–	–
2	4.6	4.6
3	>10	>10
4	>10	>10
5	>10	>10
6	12.6	<0.1
7	2.7	2.7
8	6.2	<0.1
9	2.9	2.9
10	>10	>10
11	>100	>1000
12	>100	>10
13	>10	>10
14	>10	–
15	>10	–
16	40.7	4.1
17	>100	>10
18	>100	–
19	>10	>10
20	>10	>10
21	>100	>10
VER	>100	>10

Notes: VER (verapamil), DOX (doxorubicin). The symbol “–” indicates values that could not be calculated.

Modulation of P-Gp Activity Monitored Through Rhodamine 123 Accumulation

We next focused on determining whether the compounds could modulate the efflux activity of the P-gp, which contributes to DOX resistance in the tested DOX-resistant cell line A2780DOX. Increased expression of the P-gp in that cell line had been confirmed in previous study.31 For measurements, we used rhodamine 123, a fluorescent P-gp substrate that accumulates inside cells but is actively exported by functional P-gp. Modulating P-gp activity therefore increases intracellular rhodamine accumulation. The accumulation values are shown in Figure 2a. In total, fourteen of the tested derivatives, at a concentration of 1 μM, increased rhodamine accumulation in A2780DOX cells, with compounds 3, 11, 13, 17 and 21 showing increases of more than 30% compared with the untreated control. A similarly elevated accumulation was also observed for verapamil, a well-known P-gp modulator.

Figure 2.

Accumulation of the P-gp substrate rhodamine 123 in A2780DOX cells (a) treated with a 1 μM concentration of compounds or the P-gp modulator verapamil (VER) and (b) treated with various concentrations of compound 11. The data represent the mean ± SEM of (a) four or (b) five replicates. Data were analysed using Student’s t-test against non-treated control (NT); P < 0.05 (*), P < 0.01 (**), P < 0.001 (***).

Taken together, the results— low toxicity toward hRPTEC cells, strong enhancement of DOX efficacy in both resistant lines, and significant ability to modulate P-gp activity—led us to select compound 11 [N-(4-methoxybenzyl) dregamine] for further investigation. Across all tested concentrations, derivative 11 significantly increased rhodamine accumulation, as shown in Figure 2b.

Modulation of BCRP and MRP1 Activity by Compound 11 Monitored Through Fluorescent Substrate Accumulation

To determine whether compound 11 also modulates the activity of the efflux pumps BCRP and MRP1, we used cell lines with increased expression of these transporters: a canine kidney cell line transduced with the BCRP gene (MDCKII-BCRP) and a DOX-resistant human glioblastoma cell line overproducing MRP1 protein. The effect of compound 11 at 10, 5, and 2.5 μM on transporter activity was assessed using the accumulation of the fluorescent substrates Hoechst 33342 for BCRP and calcein acetoxymethyl ester for MRP1. A decrease in efflux pump activity leads to increased intracellular accumulation of the fluorescent substrate. The results are shown in Figure 3. Compound 11 did not exert any modulatory effect on BCRP at any of the tested concentrations. In the case of MRP1, compound 11 reduced the accumulation of the fluorescent substrate. Known modulators of the tested efflux pumps, prazosin for BCRP and verapamil for MRP1, were used for activity comparison. Characterization of compound 11 interaction with P-gpTo further characterize the mechanism by which compound 11 modulates P-gp activity, we used a commercial Pgp-Glo Assay Systems kit based on isolated membranes containing the P-gp transporter. Incubation of these membranes with the tested compound allows discrimination between P-gp substrates and inhibitors based on ATP consumption. Substrates activate the transport function of P-gp and thereby increase ATP consumption. Inhibitors, on the other hand, reduce ATP consumption in the reaction mixture. The results shown in Figure 4 indicate that compound 11 activates the transport function of P-gp and accelerates ATP consumption. For comparison, P-gp substrate verapamil at 10 μM produced a similar effect, whereas ATP levels in membranes treated with the P-gp inhibitor Na₃VO₄ at 100 μM were higher than the basal ATP consumption observed in the untreated control (NT).

Figure 3.

Accumulation of (a) the BCRP substrate Hoechst 33342 in MDCKII-BCRP cells and (b) the MRP1 substrate calcein acetoxymethyl ester in U87MG-DOX cells after treatment with 10 μM compound 11, 200 μM prazosin as a BCRP modulator, or 10 μM verapamil as an MRP1 modulator. The data represent the mean ± SEM of three replicates. Data were analysed using Student’s t-test against non-treated control (NT); P < 0.001 (***).

Figure 4.

Relative ATP level after incubation of P-gp-containing membranes with compound 11 at 10 μM, the substrate verapamil at 10 μM, or the inhibitor Na₃VO₄ at 100 μM. The untreated control (NT) represents basal ATP consumption. Substrates decrease the amount of ATP in the reaction mixture, whereas inhibitors prevent ATP consumption. Data represent the mean ± SEM of three replicates. Statistical significance was analysed using Student’s t-test against the untreated control (NT); P < 0.001 (***).

Computational Analysis of Compound 11 Interactions

Docking results (Figure 5) indicate that compound 11 binds deeply within the P-gp transmembrane channel, occupying a well-defined hydrophobic pocket at the terminal region of the inward-open conformation. The ligand geometry closely matches the shape of this cavity and is primarily stabilized by extensive hydrophobic contacts. Additional stabilization through potential π–π stacking interactions with aromatic side chains is possible in multiple closely related poses.

Figure 5.

Computational model of compound 11 bound to human P-gp. MD simulation (200ns) starting point was a structure from PDB entry 8SA1 with ligand replaced by compound 11 in the best pose suggested by docking (Vina docking score 9.6 kcal mol⁻¹). All the residues named in the picture are in close proximity of the ligand, the most prominent being hydrophobic interactions of Phe978, Val982 and especially Phe983 (on one of the two Orange helices), Phe728 and Phe732 (green helix), Phe336 and Phe343 (green-ish cyan helix), and Phe303 and Tyr307, accompanied by Tyr310’s hydrogen bond with the carbonyl oxygen of the ligand (cyan helix). (a) Overall view of P-gp in the inward-open conformation with compound 11 shown in pink. The N-terminal domain is shown in blue and the C-terminal domain in red; ATP-binding sites are indicated. (b) Close-up view of the hydrophobic binding pocket highlighting interacting side chains. (c) View of the binding pocket from the cytoplasmic side. (d) View parallel to the membrane plane.

During molecular dynamics simulations, compound 11 remained stably bound in the predicted pocket without signs of spontaneous dissociation. The surrounding protein residues adapted to the ligand, leading to a tighter hydrophobic enclosure and effective displacement of water molecules from the cavity. Multiple low-energy positional variants were observed, all maintaining strong hydrophobic complementarity and potential aromatic interactions.

Chemosensitization of DOX-Resistant Monocellular and Multicellular Spheroids by Compound 11

The ability of compound 11 to enhance the cytotoxicity of DOX chemotherapy through modulation of tumor cell resistance via P-gp activity prompted us to further evaluate this compound in advanced cellular models. First, we prepared monocellular spheroids from the DOX-resistant ovarian carcinoma cell line A2780DOX and the breast carcinoma cell line MCF-7R, which better mimic conditions in vivo due to their three-dimensional architecture. As shown in Figure 6, compound 11 at a concentration of 10 μM did not exhibit any cytotoxic effects in the spheroids; a decrease in viability below 80% was observed only at concentrations 100 μM and higher. Moreover, treatment with 10 μM derivative 11 significantly enhanced DOX cytotoxicity in both A2780DOX and MCF-7R spheroids by approximately six- to seven-fold (Figure 7a and b). This effect was also reflected in altered spheroid growth dynamics and increased spheroid disintegration following treatment with DOX alone or in combination with compound 11. Although the fold increase observed in spheroids was lower than that obtained in monolayer cultures, it remained substantial.

Figure 6.

Cytotoxicity of compound 11 in monocellular spheroids of the DOX-resistant (a) ovarian cancer (A2780DOX) and (b) breast cancer cell line (MCF-7R). The values represent the mean ± SEM of three replicates.

Figure 7.

Chemosensitization of doxorubicin (DOX)-resistant spheroids by compound 11. Graphs, together with their corresponding spheroid images (a–d) shows reduction in DOX IC₅₀ values after the addition of 10 μM of compound 11 either in (a and b) monocellular or (c and d) multicellular spheroids. Monocellular spheroids were prepared from (a) the resistant ovarian carcinoma cell line A2780DOX or (b) the resistant breast carcinoma cell line MCF-7R. Multicellular spheroids were prepared from the fibroblast cell line HDF, differentiated adipocytes dif3T3-L1, and either (c) the resistant ovarian carcinoma cell line A2780DOX or (d) the resistant breast carcinoma cell line MCF-7R. (e and f) Percentage of individual cell types within the multicellular spheroids 72 h after treatment with 5 μM DOX and 10 μM of compound 11 measured by flow cytometry. IC₅₀ and fold values are presented as mean ± SEM from two or three replicates. Data were analyzed using Student’s t-test; P < 0.05 (*), P < 0.01 (**). Scale bar 200 μm.

We further evaluated the effect of compound 11 in multicellular spheroids composed of DOX-resistant cancer cells (A2780DOX or MCF-7R), fibroblasts, and adipocytes. Derivative 11, at a concentration of 10 μM, did not affect the viability of the multicellular spheroids (viability > 90%). However, in combination with DOX, the addition of 10 μM of derivative 11 led to an approximately ten-fold reduction in the DOX IC₅₀ value (Figure 7c) in multicellular spheroids derived from A2780DOX cells. In multicellular spheroids containing MCF-7R cells, the IC₅₀ shift was nearly four-fold (Figure 7d). Similarly, to monocellular spheroids, increasing DOX concentrations resulted in reduced multicellular spheroid size, with spheroid disintegration observed at higher concentrations.

We also examined changes in the cellular composition of the multicellular spheroids after the treatment with DOX alone or in combination with compound 11 (Figure 7e and f). Exposure of A2780DOX multicellular spheroids to 5 μM DOX reduced the proportion of cancer cells from 52% to 32%, accompanied by an increase in fibroblast and adipocyte representation. Combined treatment with 5 μM DOX and 10 μM of compound 11 further shifted the cellular balance toward non-cancerous cells, reducing the cancer cell population to 20%, while fibroblasts and adipocytes were present in equivalent amounts. Similarly, multicellular spheroids derived from MCF-7R cells also responded to 5 μM DOX with a reduction in the proportion of cancer cells relative to non-cancerous cells. However, upon treatment with a combination of 5 μM DOX and 10 μM derivative 11, the proportion of cancer cells was further reduced to 29%.

Overall, compound 11 demonstrated low toxicity and an ability to enhance DOX cytotoxicity in both monocellular and multicellular spheroids, while simultaneously promoting increased survival of fibroblasts and adipocytes.

Chemosensitization Effects of Compound 11 on Patient-Derived Models – PDOs and PBMCs

To evaluate the efficacy and safety of compound 11 in patient-relevant models, we tested this derivative in PDOs from resected colorectal cancers and isolated PBMCs.

Firstly, we examined whether compound 11 could enhance DOX cytotoxicity in PDOs derived from colorectal cancers (Figure 8a). In this experiment, both DOX and the derivative were tested at 10 μM. Compound 11 alone did not significantly reduce PDO viability, whereas DOX exhibited a moderate cytotoxic effect, lowering viability to approximately 70%. When combined with DOX, compound 11 substantially increased its toxicity against PDOs, reducing viability to approximately 30%.

Figure 8.

Effect of compound 11 in patient-derived models—peripheral blood mononuclear cells (PBMCs) and patient-derived colon cancer organoids (PDOs). (a) PDOs treated with 10 μM compound 11 (comp. 11), 10 μM DOX, or their combination (red). (b) Toxicity of DOX (black) and the combination of DOX with 10 μM compound 11 (red) in PBMCs. Compound 11 (10 μM) alone showed no toxicity in PBMCs (viability = 105.1 ± 2.5%). Viability values are expressed as mean ± SEM from 7–13 replicates. Data were analyzed using Student’s t-test; P > 0.05 (ns); P < 0.001 (***).

Next, we isolated PBMCs from the blood of healthy donors as a model of non-cancerous immune cells and assessed the toxicity of compound 11, DOX, and their combination (Figure 8b). Compound 11 alone at 10 μM showed no toxic effect on PBMCs (viability = 105.1 ± 2.5%), whereas DOX markedly reduced cell viability. Notably, the combination of DOX with compound 11 (10 μM) reproduced the same toxicity profile as DOX alone. The absence of increased toxicity in the combination highlights the safety of compound 11 as a potential chemotherapy adjuvant.

In summary, testing in patient-derived models demonstrated preserved chemosensitization effects and a favorable safety profile of compound 11.

Discussion

Building on the work of Cardoso et al,27 we investigated a set of 21 dregamine and tabernaemontanine N-alkylated derivatives for their anticancer properties and their ability to modulate the activity of the transmembrane transporter P-gp, a key efflux pump responsible for the extrusion of chemotherapeutics such as DOX from cancer cells, thereby reducing treatment efficacy.42 Several of the tested compounds exhibited notable anticancer activity; which may be relevant in the context of collateral sensitivity, where tumor cells that acquire resistance to one drug can develop collateral vulnerabilities to other agents through altered glutathione metabolism,43,44 changes in reactive oxygen species levels,45 disruption of redox homeostasis,46 or abnormal ATP depletion.47 However, application of tested compounds as standalone chemotherapeutics would be limited due to the relatively high concentrations required to achieve cytotoxic effects. On the other hand, the derivatives demonstrated substantial potential to enhance DOX cytotoxicity in resistant cell lines overexpressing P-gp, likely through modulation of this efflux transporter in MDR cancer cells.

Based on our results, we identified N-(4-methoxybenzyl) dregamine (compound 11) as the most promising candidate for adjuvant therapy in the treatment of DOX-resistant tumors. This derivative markedly enhanced DOX cytotoxicity in resistant cancer cells not only in standard monolayer assays but also in advanced models, including monocellular and multicellular spheroids and PDOs generated from colorectal cancers. The potential of compound 11 is further supported by the absence of toxicity toward healthy cells, both when administered alone and in combination with DOX, as demonstrated in assays using PBMCs. Taken together, compound 11 emerges as a potential P-gp inhibitor suitable for use as an adjuvant in the treatment of DOX-resistant cancers.

We have previously described the ability of N-alkylated dregamine and tabernaemontanine derivatives to sensitize resistant cancer cell lines. Strong synergistic effects between such derivatives and DOX have been previously observed in human colon adenocarcinoma (COLO 320) cells and ABCB1-transfected mouse T-lymphoma (L5178Y) cells, largely attributed to inhibition of P-gp transport activity.27 P-gp modulation was demonstrated through increased intracellular accumulation of rhodamine 123 and alterations in ATPase activity. In further studies, using as model drug resistant KB-ChR-8-5 cells, N-(para-bromobenzyl) tabernaemontanine not only inhibited P-gp but also downregulated its expression by interfering with NF-кB-mediated signaling. Combination of this derivative with DOX induced apoptosis by downregulating Bcl-2 and activating BAX, caspase-3, and p53.48 We have also explored the effect of structural modification at C-3 on the P-gp function and also on other major MDR efflux transporters including multidrug resistance protein 1 (MRP1) and breast cancer resistance protein (BCRP). Certain azine derivatives also selectively inhibited P-gp or MRP1.23,49–51

Previous studies are consistent with the findings of this work, in which we likewise confirmed the modulation effect of dregamine and tabernaemontanine derivatives on P-gp activity through increased accumulation of rhodamine 123. Among the most active, compounds 11, 17, 21 bear a methoxy group in the benzene ring. Besides being a hydrogen bond acceptor, the methoxy group increases the electronic density of the benzene ring, mostly in para-position, thus allowing to improve the interactions with P-gp. When comparing compounds with a benzyl moiety (10, 11, 16), the difference of activity seems to be related to the position of the methoxy group, being also influenced by the stereochemistry at C-20 (11, α-ethyl; 10, and 16 β-ethyl). Conversely, between compounds 19 and 21, with a phenethyl moiety and thus higher conformational flexibility due to the additional CH₂, compound 21, with the methoxy group in meta, is among the most active compounds while compound 19 was inactive, at the concentration tested. Therefore, the obtained results suggest that with the methoxy group in meta-position, compound 21 probably adopts a better conformation, allowing to increase π–π staking and hydrophobic interactions with P-gp.

However, some of these results do not directly correlate with sensitizing activity. This discrepancy may be attributed to differences in incubation times of compounds with cells. Moreover, rhodamine 123 accumulation reflects only the modulation of P-gp activity, whereas chemosensitization may also involve additional resistance mechanisms affected by derivatives. In the case of compound 11, measurement of fluorescent substrate accumulation suggests that activity modulation occurs only for P-gp, but not for BCRP or MRP1. The modulation of P-gp activity was further supported by Pgp-Glo Assay Systems kit and computational analysis.

Characterization of the interaction with P-gp using the Pgp-Glo Assay Systems kit suggests that the sensitizing effect of compound 11 may be mediated primarily by saturation of the transport capacity of P-gp rather than by direct inhibition of its function. Docking followed by all-atom MD simulations captured a timescale sufficient to evaluate local protein-ligand interactions, although it was too short to resolve the complete P-gp transport cycle. Rather than ligand dissociation, the simulations indicated rearrangements of nearby residues that strengthened binding affinity. The predicted binding mode suggests that compound 11 may interfere with the P-gp transport cycle by occupying a deep transmembrane site. The ligand may prevent proper closure of the inward-open conformation upon Mg–ATP binding during conformational transitions, thereby slowing transport to the extracellular space. This interpretation was guided by the complementary experimental findings, whereas future integration with coarse-grained simulations could provide further insight into larger conformational changes associated with Mg-ATP binding and NBD dimerization. Despite its promising potential, this derivative have not previously been evaluated in advanced culture models that more accurately recapitulate tumor architecture and are critical for assessing translational potential.

This study also validated the effect of compound 11 in spheroid models. Unlike monolayer cultures, spheroids reflect the natural tumor architecture, including limited drug penetration, cell–cell and cell–matrix interactions, and the formation of hypoxic cores that promote the upregulation of efflux transporters such as P-gp.52–54 Multicellular spheroids also enable incorporation of additional cell types present in the tumor microenvironment.55 In this study, we likewise observed increased resistance of monocellular spheroids compared with cell monolayers. The reduced penetration of compounds into the spheroids is likely responsible for the weaker sensitizing effect of compound 11. In multicellular spheroids containing the resistant breast cancer cell line MCF-7R, the sensitizing effect of compound 11 was further diminished, most likely due to the presence of healthy fibroblasts and adipocytes, which increase spheroid compactness and thereby affect molecular penetration. In contrast, ovarian carcinoma A2780DOX formed less compact multicellular spheroids and was therefore probably more sensitive to treatment. As previously reported by Flörkemeier et al,38 spheroids generated by co-culturing A2780 ovarian cancer cells with dermal fibroblasts exhibited increased sensitivity to chemotherapeutic agents such as cisplatin. A similar effect was observed in this study with respect to the sensitizing activity of compound 11. In addition, our experimental findings suggest that the combination of compound 11 with DOX may have reduced the cancer cell fraction within the spheroid, potentially decreasing competition for oxygen and nutrients and thereby indirectly supporting the survival of healthy cells.

We further examined the effects of compound 11 in PDOs. PDOs generated from primary tumor samples represent clinically relevant models that preserve tumor architecture, cellular heterogeneity, and key physiological characteristics of the original tissue. For these reasons, PDOs are increasingly used in personalized medicine and cancer research.55 Importantly, PDOs prepared from colorectal cancer tissue frequently exhibit elevated expression of P-gp, which contributes to reduced intracellular accumulation of chemotherapeutic agents and consequent drug resistance.56 The high P-gp activity in colorectal cancer PDOs likely underlies their decreased sensitivity to DOX, as also reported by Deng et al,57 who observed that 9 μM DOX reduced PDO viability to approximately 75% of control levels. Consistent with these findings, our results demonstrate limited DOX efficacy when administered alone. However, co-treatment with compound 11 significantly enhanced DOX-induced cytotoxicity. Mechanistically, inhibition of P-gp by compound 11 likely increased intracellular retention of DOX, thereby restoring its ability to induce cytotoxic effects in tumor cells. These findings highlight the considerable therapeutic potential of dregamine and tabernaemontanine derivatives acting as P-gp inhibitors in patient-relevant models.

P-gp inhibitors investigated to date have failed in clinical trials mainly because of high toxicity.42,58,59 There we also showed that combination of compound 11 combined with DOX was not toxic to PBMCs and that the derivatives alone were also non-toxic to kidney epithelial cells, indicating that these compounds may represent a safer class of adjuvants for overcoming MDR of cancer cells.

Although this study provides insights into the activity of dregamine and tabernaemontanine N-alkylated derivatives in patient-relevant models, several limitations should be considered. This work focuses on enhancing cancer cells sensitivity via P-gp inhibition, but only one chemotherapeutic agent was examined. Because P-gp contributes to resistance against multiple chemotherapeutics,60 future studies should evaluate derivatives in combination with additional clinically relevant cytotoxic drugs. Moreover, the effects of N-alkylated dregamine and tabernaemontanine derivatives on other efflux transporters involved in MDR, such as BCRP or MRP1, should be investigated.61 Furthermore, this study does not address the selectivity of the derivatives toward P-gp. Finally, this study does not include experiments in vivo, which are currently necessary for translation toward clinical application and are essential for confirming the safety of compounds.

Overall, our study demonstrates the significant potential of dregamine and tabernaemontanine derivatives as chemosensitizing adjuvants for the treatment of drug-resistant cancers, expands experimental validation to advanced cellular models, and provides better assessment of their toxicity.

Conclusion

MDR and the associated overexpression of P-gp represent major limitations in the effectiveness of cancer chemotherapy. Our study demonstrates that N-alkylated dregamine and tabernaemontanine derivatives possess potential for overcoming MDR in cancer through modulation of P-gp activity. Among the tested compounds, N-(4-methoxybenzyl) dregamine exhibited the most favorable profile, combining low intrinsic toxicity with strong chemosensitizing activity in resistant cancer cell lines. These derivatives may therefore represent promising candidates for in vivo studies, and, if favorable efficacy and safety profiles are confirmed, they could potentially be further developed as chemotherapy adjuvants for resistant tumors. They may also provide a basis for the development of a new generation of P-gp inhibitors with lower toxic effects on healthy cells. Further studies on N-alkylated dregamine and tabernaemontanine derivatives should investigate their ability to modulate activity of additional efflux transporters involved in MDR, as well as the molecular mechanisms underlying their effects on transporter proteins. The development of efflux transporter inhibitors that do not induce adverse toxic effects is essential for ensuring the effectiveness of chemotherapy in patients with drug-resistant cancers.

Funding Statement

This article is based upon work from COST Action CA23110<INFOGUT>, supported by COST (European Cooperation in Science and Technology, http://www.cost.eu). This research was supported by the Ministry of Education, Youth and Sports of the Czech Republic (LUC25122), by the UCT Specific university research, grant number A1_FPBT_2026_002, and Fundação para a Ciência e Tecnologia (FCT), Portugal (projects: 2022.05718.PTDC and PTDC/MED-QUI/30591/2017; and PhD grant PD/BD/135291/2017). Computational resources were provided by the e-INFRA CZ project (ID:90254), supported by the Ministry of Education, Youth and Sports of the Czech Republic.

Abbreviations

ABC, ATP-binding cassette; BCRP, breast cancer resistance protein; CS, collateral sensitivity; DMEM, Dulbecco’s Modified Eagle’s medium; EDTA, 2,2’,2”,2”’-(ethane-1,2-diyldinitrilo)tetraacetic acid; FBS, fetal bovine serum; MDR, multidrug resistance; MRP1, multidrug resistance-associated protein 1; MRP2, multidrug resistance-associated protein 2; PBS, phosphate-buffered saline; P-gp, P-glycoprotein; SEM, standard error of the mean; STR, short tandem repeat; SI, selectivity index; poly(2-hydroxyethyl methacrylate), HEMA; peripheral blood mononuclear cells, PBMC, Peripheral Blood Mononuclear Cells.

Data Sharing Statement

The data that support the findings of this study are available from the corresponding author, [JV], upon reasonable request.

Ethics Approval and Informed Consent

Ethical approval for this study was granted by the Ethics Committee of the UCT Prague (EC 1/2026), and the Ethics Committee of the General University Hospital in Prague (60/24 S-IV) and written informed consent was obtained from all patients.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest in this work.