Synthesis and anticancer properties of a hybrid molecule with the testosterone and estradiol head-groups

Abstract

This is the first report on a unique hybrid molecule made of estradiol and testosterone (TS). This distinctive hybrid molecule (1) was designed to interact with both the estrogen receptor (ER) and the androgen receptor (AR) found in hormone-dependent female and male cancer cells, and was synthesized using ethynylestradiol (17EE) as the estrogenic component and 7α-(4-azido-but-2-enyl)-4-androsten-17β-ol-3-one as the androgenic counterpart in a seven-step reaction with ~ 26 % overall yield. We reasoned that the dual receptor binding ability could allow 1 to act as an antihormone. This was tested on hormone-dependent and hormone-independent breast cancer (BCa) and prostate cancer (PCa) cells. The antiproliferative activity was also assessed on colon and skin cancer cells. We found that 1 was active against MCF7 (ER + ) BCa cells (IC₅₀ of 4.9 μM), had lower inhibitory potency on LNCaP (AR + ) PCa cells (IC₅₀ > 5 μM) and no effect on PC3 and DU145 (AR-) PCa cells. This suggests that the estrogenic component of 1 can interact with the ER on MCF7 cells more effectively than the androgenic component with the AR on LNCaP PCa cells, possibly due to a suboptimal spacer or linkage site(s). Nonetheless, the hybrid 1 was active against colon (HT-29) and melanoma (M21) cancer cells (IC₅₀ of 3.5 μM and 2.3 μM, respectively), and had low cross-reactivity with the drug- and androgen-metabolizing cytochrome P450 3A4 (CYP3A4, IC₅₀ ≫ 5 µM). These findings demonstrate the anticancer potential of 1 and warrant further explorations on this new type of hybrids.

1. Introduction

Cancer is the leading cause of death amongst the Canadian population, with estimated 247,100 new cancer cases and 88,100 cancer deaths in 2024 [1]. Prostate cancer (PCa) is the most prevalent cancer in men, while breast cancer (BCa) is the most common in women, with estimated 25,900 and 29,400 new cases, respectively [1]. Based on data from 2015 to 2017, the predicted five-year net survival rate from diagnosis is high for both PCa (94 %) and BCa (91 %). Despite these encouraging statistics, several academic and pharmaceutical research teams are still developing novel anticancer agents for BCa and PCa, pushing further the boundaries of drug discovery. Both types of cancer are usually hormone-dependent at the onset of the disease but ultimately progress to a hormone-independent cancer [2]. The sex hormones exert their action by binding to and triggering dimerization of the respective receptors, estrogen receptors (ERα and ERβ) or androgen receptor (AR), which is essential for their function as transcription factors. Thus, one strategy for the disruption of oncogenic signaling would be the prevention of ligand binding to the receptor and/or its dimerization.

Toward this goal, our laboratory utilized the concept of bivalent ligands and developed several dimeric estrogenic [3–5] and androgenic [6–9] molecules with antiestrogenic and antiandrogenic activities. Noteworthy, dimerization of biologically active drugs and natural products is currently the subject of important research activities [10,11]. As a logical continuation of our studies, we have now constructed a mixed steroid hybrid made of estradiol and testosterone (TS) (see Scheme 1). To our knowledge, this is the first report on this type of hybrid molecule.

Scheme 1. Reagents and conditions:

a) Ac₂O, AcCl, pyridine, reflux, 4 h (100 %); b) NBS, Li₂CO₃. LiBr, DMF/H₂O, 0 ^◦C, 6 h (76 %); c) Allyltrimethylsilane, TiCl₄, pyridine, CH₂Cl₂, −78 ^◦C to −30 ^◦C, 2.75 h (68 %); d) Allylchloride, 5 % Hoveyda-Grubbs cat. 2nd gen., CH₂Cl₂, reflux, 6 h (100 %); e) MeOH, HCl, r.t. 48 h (94 %); f) NaN₃, MeOH, reflux, 2 h (95 %); g) Ethynylestradiol, CuSO₄⋅5H₂O, Sodium ascorbate, THF/H₂O, r.t. 16 h (56 %).

Contrary to the conventional C2-symmetric estradiol and TS dimers, this conceptually innovative hybrid could interact with cancer cells containing both female and male hormone receptors. Hence, such a hybrid could disrupt cell signaling pathways in a distinct fashion as compared to known ER and AR antagonists and provide alternative therapies useful for the treatment of cancer. As mentioned, the ER regulates the growth of hormone-dependent BCa and the AR controls the growth of hormone-dependent PCa. Furthermore, there is emerging evidence that both types of receptors (AR, ERα and ERβ) are present in these cells and regulate cancer cell survival and progression [12]. The AR is expressed not only in PCa cells but also in many other types of cancer (Table 1) [13]. Therefore, we hypothesized that a chimeric compound with two different hormone functionalities would largely impact the growth of hormone-dependent cancers by simultaneously targeting multiple key receptors implicated in cell proliferation. The ERα and AR are without doubt key targets for cancer therapy [12–14]. However, considering the complexity of cancer evolution, scientists are persistently looking to improve treatments [15]. Along with the development of natural product-based hybrids as anticancer agents [16], unconventional approaches and innovative molecules must also be considered as a way to discover novel cancer treatments. In the present study, we demonstrate biological potential of one such fascinating hybrid molecule.

Table 1.

Androgen receptor expression (>50 %) in various cancers [13].

Cancer type	Subtype	AR+ % (positive/tested)
Astrocytoma	Anaplastic	75 % (3/4)
Basal cell carcinoma	Not specified	65 % (20/31) – 78 % (25/32)
BCa	Her-2 adenocarcinoma	85.6 % (89/104)
BCa	BRCA negative	76 % (56/74 %)
Desmoid tumors	Not specified	52.9 % (14/27)
Juvenile nasopharyngeal fibroma	Not specified	38.4 % (5/13) – 75 % (18/24)
Non-small cell lung cancer	Not specified	31 % (20/64) – 70.5 % (18/24)
Ovarian cancer	Steroid cell tumors	64 % (9/14)
Peritoneal mesothelioma	Not specified	65 %
PCa	Not specified	95 %
Thymoma	B3 thymoma	58.8 % (10/17)
Thyroid carcinoma	Papillary	80 % (4/5)
Salivary gland	Pleomorphic adenoma	4.8 % (2/41) – 100 % (14/14)
Salivary gland	Salivary duct carcinoma	92 % (11/12) – 100 % (6/6)
Sarcoma	Osteocarcoma	28.5 % (8/28) – 50.8 % 33/65)
Uterine carcinoma	Endometrial carcinoma	16 % (4/25) – 88.6 % (39/44)

2. Experimental protocols

2.1. Biological methods

2.1.1. Cell lines culture

MCF7 (ER + ) and MDA-MB-231 (ER-) human breast carcinomas, LNCaP (AR + ), PC3 (AR-) and DU145 (AR-) human prostate adenocarcinomas, HT-29 human colon carcinoma and T84 wt + human colon adenocarcinoma cell lines were purchased from the American Type Culture Collection (ATCC, Manassa, VA, USA). HaCaT epidermal keratinocytes were purchased from Thermo Fisher Scientific (Saint-Laurent, QC, Canada). M21 human melanoma cells were kindly provided by Dr. David Cheresh (University of California, San Diego School of Medicine, CA, USA).

MCF7 and DU145 cells were cultured in Eagle’s Minimum Essential Medium (EMEM, Wisent Bioproducts, St-Bruno, QC, Canada). MCF7 cells were supplemented with 0.01 mg/mL recombinant human insulin (Gibco, Thermo Fisher Scientific, Saint-Laurent, QC, Canada). DU145 cells were also cultured in EMEM. MDA-MB-231 cells were cultured in Leibovitz Medium (L-15, Wisent Bioproducts, St-Bruno, QC, Canada). LNCaP cells were cultured in RPMI 1640 medium (Wisent Bioproducts, St-Bruno, QC, Canada). PC3 was cultured in Kaighn’s Modification of Ham’s F-12 Medium (F12K, Corning Life Science, Fisher Scientific, Montreal, QC, Canada). HT-29, M21, HaCaT cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Wisent Bioproducts, St-Bruno, QC, Canada). T84 wt cells were cultured in DMEM/F12 medium. All culture cells were supplemented with 10 % fetal bovine serum (FBS, Gibco, Thermo Fisher Scientific, Saint-Laurent, QC, Canada) and 1 % penicillin–streptomycin (Wisent Bioproducts, St-Bruno, QC, Canada). Finally, except MDA-MB-231, all cells were maintained at 37 ^◦C in moisture saturated atmosphere containing 5 % CO₂. MDA-MB-231 was cultured at 37 ^◦C in moisture saturated atmosphere without CO₂.

2.1.2. Antiproliferative activity assay

The antiproliferative activity (IC₅₀) was evaluated following the procedure described by the National Cancer Institute (NCI) Developmental Therapeutics Program for its drug screening program (NCI-60) with minor modifications [17]. The ninety-six-well Costar microtiter clear plates (Fisher Scientific, Montreal, QC, Canada) were seeded with 75 μL of a suspension of either MCF7 (3 × 10³), MDA-MB-231 (2.5 × 10³), LNCaP (3.5 × 10³), PC3 (3 × 10³), DU145 (3 × 10³), T84 wt (2 × 10³), HT-29 (2 × 10³), M21 (2 × 10³) or HaCaT (2 × 10³) cells per well in their corresponding medium for 24 h. The tested molecules were freshly solubilized in DMSO (40 mM), diluted in each medium, and the aliquots of serially diluted solutions (75 μL) were added to the plate. Final chemical concentrations were ranged from 100 µM or the maximal solubility concentration to 0.01 µM. DMSO concentration was kept constant at 0.5 % (v/v) to prevent toxicity. After 48 h, cell growth was halted by adding 50 % (w/v) cold CF₃CO₂H (10 % w/v, final concentration) and the plates were laid at 4 ^◦C for 30 min. Afterwards, plates were washed 4-times with ultrapure water (Milli-Q) and dried at room temperature. The residual cells were stained by incubating with 75 μL of a Sulforhodamine B solution (0.1 % w/v) in 1 % acetic acid for 15 min. After staining, unbound dye was discarded and the plates were rinsed 4-times with 1 % CH₃CO₂H and air-dried. Cell-bound dye was then solubilized with 200 μL of a 20 mM Tris base solution for at least 2 h. Finally, plates were read for absorbance at an optimal wavelength (between 530 and 570 nm) with a SpectraMax® i3x spectrometer (Molecular Devices, San Jose, CA, USA). The percentage of cell growth for each compound was determined by comparing the absorbance data of treated cells to that of control cells fixed on the day of the treatment. IC₅₀ and standard error of the mean (SEM) were calculated using GraphPad Prism 10.1.2 software (GraphPad Software, San Jose, CA, USA). The experiments were done at least thrice in triplicate.

2.2. In vitro studies on CYP3A4

2.2.1. Preparation of human CYP3A4

Codon-optimized full-length human CYP3A4 was co-expressed with GroESL in E. coli C41 strain. After inoculation into TB media (0.6 L in 2.8 L flasks) supplemented with ampicillin (100 mg/L) and chloramphenicol (34 mg/L), cells were grown at 37 ^◦C and 220 rpm until OD at 600 nm reached 0.6, then temperature was decreased to 30 ^◦C and 0.5 mM 5-aminolevulenic acid and 0.5 mM isopropyl β-D-1-thiogalactopyranoside (final concentration) were added to induce protein production. Cells were grown at 100 rpm for 48 h and, after harvesting, resuspended in lysis buffer (0.1 M potassium phosphate pH 7.4, 0.1 M NaCl, 20 % glycerol, 5 mM mercaptoethanol, and 1 mg/L leupeptin) and broken by passing once through a microfluidizer (Microfluidics, Newton, MA, USA). Cell lysate was incubated with 0.4 % Nonidet P-40 (Sigma-Aldrich, St. Louis, MO, USA) for 60 min at 4 ^◦C to solubilize membranes and centrifuged at 32,000 × g for 60 min to remove cell debris. The supernatant fraction was loaded on HisPur™ Ni-NTA resin (Thermo-Fisher, Waltham, MA, USA) and washed with 3 volumes of lysis buffer with 0.2 % Nonidet P-40 and 2 volumes of the latter buffer supplemented with 5 mM histidine. Protein was eluted with 0.1 M potassium phosphate pH 7.4, 20 % glycerol, 0.2 % Nonidet P-40, and 30 mM histidine. The peak fractions were concentrated and loaded on a CM Sepharose Fast Flow column (GE Healthcare, Chicago, Il, USA) equilibrated with 20 mM potassium phosphate pH 7.4, 20 % glycerol, and 2 mM DTT. The column was washed overnight with 1 L of CM loading buffer. Protein was eluted with a linear gradient of CM loading buffer vs. 0.125 M potassium phosphate pH 7.4, 0.2 M NaCl, 20 % glycerol, and 2 mM DTT. Fractions with A_{417/280 nm} > 1.5 were combined, concentrated, and stored at − 80 ^◦C. P450 concentration was determined according to Omura and Sato [18].

2.2.2. Ligand binding to CYP3A4

Ligand binding to CYP3A4 was monitored at ambient temperature in 0.1 M phosphate buffer, pH 7.4, containing 20 % glycerol and 1 mM dithiothreitol. Protein (1.5 μM) was titrated with small aliquots of compound 1 dissolved in DMSO, and spectral changes were recorded using a Cary 300 spectrophotometer. The final DMSO concentration in the experimental sample did not exceed 2 %. The titration plot was built based on the difference absorbance spectra and fitted to a hyperbolic equation to derive the K_d value: ΔA = ΔA_max x [ligand] /(K_d + [ligand]) where ΔA_max and [ligand] are the maximum absorbance change and the total ligand concentration after each titrant addition, respectively. Inhibitory potency of 1 for the BFC debenzylation activity of CYP3A4 was assessed at 37 ^◦C according to the previously described procedure [19]. The [% activity] vs. [compound 1] plot was fitted to a four-parameter logistic nonlinear regression equation: y = D + ((A-D)/(1 + 10^(X-logC)•B)), where A and D are the maximal and minimal activity values, respectively, B – a slope factor, and C is IC₅₀.

2.3. Chemistry

All reactions were done under an inert atmosphere of nitrogen and with anhydrous solvents. The starting material, reactant and solvents were obtained commercially and used per se or purified and dried by standard methods [20]. The organic solutions were dried over magnesium sulfate (MgSO₄), filtered and evaporated on a rotary evaporator under reduced pressure. All reactions were checked by UV fluorescence. Commercial TLC plates were Sigma T 6145 (polyester silica gel 60Ǻ, 0.25 mm). Flash column chromatography was performed on Merck grade 60 silica gel, 230–400 mesh [21]. All solvents used in flash chromatography were distilled before use.

The IR were taken using a Cary 660 FTIR, Agilent Technologies, USA. Mass spectral assays were acquired using a MS model 6210, Agilent technology instrument. The high-resolution mass spectra (HRMS) were obtained by TOF (time of flight) using ESI (electrospray ionization) via positive mode (ESI+, Université du Québec à Montréal). Nuclear magnetic resonance (NMR) spectra were recorded either on a Varian 200 MHz, on a BrukerAM-300 MHz or on a Bruker 400 MHz NMR apparatus. Samples were dissolved in deuterated chloroform (CDCl₃) or acetone (acetone‑d₆) for data acquisition using the residual solvent signal as internal standard (CHCl₃ δ 7.26 ppm for ¹H NMR and δ 77.23 ppm for ¹³C NMR; (CH₃)₂C = O δ 2.05 ppm for ¹H NMR and δ 206.26 ppm for ¹³C NMR. Chemical shifts (δ) are indicated in parts per million (ppm), the coupling constants (J) are expressed in hertz (Hz). Multiplicities are described by the following abbreviations: s for singlet, d for doublet, t for triplet and m for multiplet, and bs for broad singlet.

Note: The compounds are named using the androgen (4-androsten-17β-ol-3-one) and the estrogen (estradiol) skeleton for clarity to the readers.

2.3.1. Synthesis of (E)-17β-Hydroxy-7α-{4-[4-(estradiol-17α-yl)-[1,2,3] triazol-1-yl]-but-2-enyl}−4-androsten-3-one (1)

2.3.1.1. Preparation of (E)-7α-(4-azido-but-2-enyl)-4-androsten-17β-ol-3-one.

This product was synthesized following the six-step synthetic path from TS as reported previously (Scheme 1) [6,22,23]. The final azide was obtained with 46.1 % overall yield. The spectral data of the intermediates as well as the azide were in agreement with previously reported data. Proton and carbon NMR spectra of the azide as reported [22]: IR ν (cm⁻¹): 3432 (O—H), 2945 and 2873 (C—H), 2094 and 2073 (N═N = N), 1653 (C═O), 1617 (C═C). ¹H NMR (200 MHz, CDCl₃, δ ppm): 5.72 (1H, s, 4-CH), 5.56 (2H, m, 21-CH and 22-CH), 3.67 (3H, m, 23-CH₂ and 17-CH), 1.21 (3H, s, 19-CH₃), 0.80 (3H, s, 18-CH₃). ¹³C NMR (200 MHz, CDCl₃, δ ppm): 199.1, 169.1, 134.9, 126.2, 125.1, 81.6, 52.6, 47.2, 46.2, 42.9, 38.9, 38.8, 36.9, 36.5, 36.4, 36.2, 34.2, 30.5, 28.8, 23.0, 21.1, 18.2, 11.2.

2.3.1.2. Preparation of (E)-17β-hydroxy-7α-{4-[4-(estradiol-17α-yl)-[1,2,3]triazol-1-yl]-but-2-enyl}−4-androsten-3-one (1).

Commercially available ethynylestradiol (78.1 mg, 0.262 mmol, 1.7 eq.) and 7α-(4-azido-but-2-enyl)-4-androsten-17β-ol-3-one (59.0 mg, 0.154 mmol, 1 eq.) are added to a mixture of THF/water (3:1, 3 mL) and stirred. Then, sodium ascorbate (76.2 mg, 0.385 mmol, 2.5 eq.) and CuSO₄ • 5 H₂O (7.3 mg, 0.046 mmol, 0.3 eq.) are added and the mixture is stirred at room temperature for 16 h. The product is extracted with a mixture of EtOAc and diethylether (1:1, 3 × 25 mL) and washed with water (3 × 15 mL). The organic phases are combined and dried over MgSO₄, filtered and evaporated under reduced pressure. The crude product is purified by flash chromatography (acetone/hexane, 2:3) to give 59.1 mg of 1. Yield, 56 %; white powder. Pf: 175–177 ^◦C. IR ν (cm⁻¹): 3400 (O—H), 2923 and 2869 (C—H), 1652 (C═O), 1612 (C═C). ¹H NMR (400 MHz, acetone‑d₆, δ ppm): 7.89 (1H, s, 17′-OH), 7.78 (1H, s, 24-CH-triazole), 7.01 (1H, d, J = 8.4 Hz, 1′-CH), 6.55 (1H, dd, J = 8.4 Hz and 2.5 Hz, 2′-CH), 6.50 (1H, s, 4′-CH), 5.70 (2H, bs, 4-CH and 22-CH), 5.35 (1H, s, 21-CH), 4.99 (2H, m, 20-CH₂), 4.22 (1H, s, 17-OH) 3.60 (1H, t, J = 8.7 Hz, 17-CH), 1.20 (3H, s, 19-CH₃), 1.04 (3H, s, 18′-CH₃), 0.78 (3H, s, 18-CH₃). ¹³C NMR (300 MHz, acetone‑d₆, δ ppm): 198.4, 169.3, 155.0, 154.9, 137.8, 135.0, 131.4, 126.3, 126.2, 125.8, 121.7, 115.3, 112.9, 82.0, 81.1, 51.8, 48.4, 47.4, 47.3, 46.3, 43.7, 43.1, 39.8, 38.8, 38.6, 37.8, 36.5, 36.2, 36.1, 36.0, 34.0, 33.2, 29.8, 28.8, 27.6, 26.5, 23.8, 22.8, 21.0, 17.9, 14.3, 11.0. HRMS (ESI + ): (M + H)⁺ calculated for C₄₃H₅₈N₃O₄= 680.4422; found: 680.4400.

3. Results and discussion

3.1. Design and chemistry

Hybrid 1 was engineered with a 1,2,3-triazole-based linker between positions 7α of TS and 17α of estradiol. Functionalization deep within the B ring of TS was chosen to limit potential perturbations in the binding to the AR, which occurs mainly through hydrogen bonding interactions with carbonyl at C4 and hydroxyl at 17β [24]. Ethynylestradiol was chosen as the estrogenic component based on its high binding affinity towards the ERs, increased resistance toward metabolism, commercial availability, and the presence of the 17α-alkyne functional group that could be easily linked to the TS unit [25]. A 1,2,3-triazole-based linker was selected to increase the solubility of the dimer and to improve resistance to cleavage in biological environment. The length of the linker was chosen to allow enough space between the two steroid units, as steric hindrance could prevent their binding to the respective receptors.

The synthetic path for the preparation of hybrid 1 is depicted in Scheme 1. 7α-(4-Azido-but-2-enyl)-4-androsten-17β-ol-3-one was obtained in about 46 % overall yield from TS through a six-step reaction sequence developed in our laboratory [22]. The reaction conditions are detailed in Scheme 1 legend. Initially, TS was transformed into 7α-allyl-4-androsten-17β-ol-3-one acetate intermediate with 52 % overall yield using a three-step reaction sequence described by Bastien et al. [23]. Another three-step reaction sequence led to the azide precursor of hybrid molecule 1 [22]. The azide was obtained with 89.3 % yield from the acetate intermediate and used for the 1,3-dipolar Huisgen cycloaddition with commercially available ethynylestradiol (Scheme 1) [26,27]. So, upon treatment of the azide with ethynylestradiol, sodium carbonate and copper sulfate in a mixture of tetrahydrofuran and water, the final 1,2,3-triazole-linked estradiol-TS hybrid molecule 1 was obtained with 56 % yield. The overall yield for the seven-step reaction was 26 %. All intermediates and the final hybrid molecule 1 were fully characterized by infrared (IR) and nuclear magnetic resonance spectroscopy (¹H and ¹³C NMR) and by high-resolution mass spectrometry (HRMS).

3.2. Antiproliferative activity of hybrid 1 on cancer cell lines

The antiproliferative activity of the hybrid compound 1 was measured using the Sulforhodamine B colorimetric assay on two BCa cell lines: estrogen-dependent MCF7 (ER + ) and estrogen-independent MDA-MB-231 (ER-). The antiestrogen tamoxifen and its active metabolite 4-hydroxy-tamoxifen were used as positive controls [17]. Compound 1 was also tested on three PCa cell lines: androgen-dependent LNCaP (AR + ) and androgen-independent PC3 (AR-) and DU145 (AR-). The antiandrogen drug, bicalutamide, was used as a positive control. The results from the antiproliferative activity assay are shown in Table 2, where IC₅₀ is the compound concentration inhibiting cell growth by 50 %. It is noteworthy that the maximal concentration of hybrid 1 was limited to 5 μM to ensure its full solubility in cell culture medium. At higher concentrations, hybrid 1 has tendency to precipitate out of aqueous solutions. With BCa cells, we were able to observe an IC₅₀ of 4.9 μM on MCF7 (ER + ) and IC₅₀ > 5 μM on MDA-MB-231 (ER-). The respective IC₅₀ values for tamoxifen were 1.1 μM and 4.0 μM versus 0.013 μM and 7.0 μM for its metabolite. Thus, on MCF7 cells, 1 is considerably less potent than 4-hydroxytamoxifen but displays a comparable activity with tamoxifen (IC₅₀ of 4.9 μM vs 0.013 μM and 1.1 μM, respectively). These results suggest that the hybrid molecule 1 interacts with ER + BCa MCF7 cells more favorably than with ER-MDA-MB-231 cells. Thus, the estrogenic component of the hybrid seems to retain some affinity for the ER. On AR + PCa LNCaP cells, compound 1 showed an IC₅₀ > 5 μM, while bicalutamide exhibited an IC₅₀ of 3.1 μM. On AR-PC3 and DU145 PCa cells, the hybrid 1 had no effect at the maximum concentration tested. This suggests that the androgenic component of 1 is not interacting very strongly at this concentration with the AR in LNCaP cells.

Table 2.

Antiproliferative activity of compound 1, bicalutamide, tamoxifen and 4-hydroxy-tamoxifen on human breast and prostate human prostate adenocarcinoma cancer cell lines.

Compound	IC50 (μM)a ± SEM
	MCF7 (ER + )	MDA-MB-231 (ER−)	LNCaP (AR + )	PC3 (AR−)	DU145 (AR−)
1	4.9 ± 0.4	> 5	> 5	> 5	> 5
Bicalutamide	3.1 ± 0.6	44 ± 4	10 ± 1	62 ± 3	52 ± 1
Tamoxifen	1.1 ± 0.4	4 ± 1	17 ± 2	14 ± 5	13.2 ± 0.2
4-OHT b	0.013 ± 0.005	7 ± 1	7 ± 1	13 ± 1	13 ± 1

^aInhibitory concentration (IC₅₀) is concentration of drug inhibiting cell growth by 50%.

^b4-OHT: 4-hydroxytamoxifen.

There is substantial evidence that the ERs and AR are implicated in cell growth and tumorigenesis of colon [28–32] and melanoma cancer cells [33–36]. Therefore, the hybrid 1 was tested on colon cancer cells (T84 wt and HT-29), melanoma cells (M21), as well as normal immortalized human keratinocytes (HaCaT), to further verify its potential as an anticancer agent. The results are presented in Table 3. On colon T84 wt cells and normal immortalized human keratinocytes HaCaT cells, compound 1 was inactive at the maximal concentration tested but displayed an IC₅₀ of 3.5 μM against colon HT-29 cells. On the latter cancer cells, bicalutamide, tamoxifen and 4-hydroxytamoxifen showed an IC₅₀ of 42 μM, 6.4 μM and 0.9 μM, respectively. On melanoma cells M21, the hybrid 1 was active, displaying an IC₅₀ of 2.3 μM. In comparison, bicalutamide, tamoxifen and 4-hydroxytamoxifen had IC₅₀ of 49 μM, 5.7 μM and 4.2 μM, respectively. These results confirm the inhibitory potential of 1 on other types of cancers. This activity could be due to the presence of the AR and ERs in these cells [13,28–36].

Table 3.

Antiproliferative activity of compound 1, bicalutamide, tamoxifen and 4-hydroxytamoxifen on colon (T84 wt, HT-29) and skin (M21) cancer cells and on immortalized human keratinocytes (HaCaT).

Compound	IC50 (μM)a ± SEM
	T84 wt	HT-29	M21	HaCaT
1	> 5	3.5 ± 0.1	2.3 ± 0.1	> 5
Bicalutamide	56 ± 2	42 ± 2	49 ± 1	95 ± 5
Tamoxifen	12.4 ± 0.1	6.4 ± 0.1	5.7 ± 0.3	13 ± 1
4-OHT b	6.9 ± 0.3	0.9 ± 0.2	4.2 ± 0.2	10.4 ± 0.6

^aInhibitory concentration (IC₅₀) is concentration of drug inhibiting cell growth by 50%.

^b4-OHT: 4-hydroxytamoxifen.

3.3. Interaction of hybrid 1 with CYP3A4

Our previous studies showed that TS dimers can bind to the active site and inhibit the activity of cytochrome P450 3A4 (CYP3A4), the major and most clinically relevant drug-metabolizing enzyme in humans [7,9,37]. This cross-reactivity is undesired, because it could affect clearance of drugs, contribute to drug-drug interactions, and alter systemic levels of androgens predominantly metabolized by CYP3A4 [38,39]. As a result, CYP3A4 inhibition could alter the AR signaling, prostate growth and cancer pathogenesis. Therefore, it was important to assess the inhibitory potential of the new hybrid 1 against CYP3A4.

Whether the chemical compound is capable of binding to the CYP3A4 active site can be determined spectrally by following changes in the heme Soret band. As seen from Fig. 1A, addition of 1 to the recombinant CYP3A4 led to a blue shift in the Soret band, indicative of a low-to-high spin transition. Although the hybrid could approach the heme and perturb its environment, the high-spin shift was relatively low, only 23 %. The dissociation constant (K_d) derived from the titration plot (right inset in Fig. 1A) was 2.0 ± 0.1 μM. For comparison, for the 7α- and 17α-linked TS dimers, the high-spin shift and K_d values were in the 33–99 % and 0.37–9.6 μM range, respectively [7,9]. The inhibitory potency of 1 for CYP3A4 was assessed in a soluble reconstituted system using 7-benzyloxy-4-(trifluoromethyl)coumarin (BFC) as a substrate. As seen from the inhibitory plot (Fig. 1B), the 50 % inhibition level could not be reached under studied conditions. Thus, compound 1 is a poor inhibitor of CYP3A4, with IC₅₀ ≫ 5 μM. In accord with the spectral data, this suggests that, due to structural distinctions, the hybrid cannot block the active site as effectively as some TS dimers do [7].

Fig.1.

A – Spectral changes caused by compound 1 in recombinant CYP3A4. Absorbance spectra of 1.5 μM CYP3A4 at different ligand concentrations were recorded at room temperature in 0.1 M phosphate buffer, pH 7.4, supplemented with 20 % glycerol and 1 mM dithiothreitol. Solid black and red spectra correspond to the ligand-free and 1-bound CYP3A4, respectively. Arrows indicate the direction of absorbance changes. Left inset shows the difference spectra; right inset is a titration plot with hyperbolic fitting. The derived K_d is indicated. B – Inhibition of the BFC debenzylase activity of CYP3A4 by compound 1. The assay was conducted at 37 ^◦C in a soluble reconstituted system containing 0.2 μM CYP3A4 and 0.3 μM cytochrome P450 reductase. Formation of a fluorescent product was monitored at increasing concentrations of 1. The remaining activity was calculated relative to the DMSO-containing sample used as a control (100 % activity).

4. Conclusion

To explore new avenues for cancer drug discovery, we utilized an unconventional strategy and designed a novel hybrid molecule made of estradiol and TS which could potentially interact with both hormone receptors present in female and male cancer cells. The new hybrid 1 was tested on hormone-dependent and hormone-independent BCa and PCa cells, as well as on colon and skin cancer cells. The hybrid 1 was active against MCF7 (ER + ) BCa cells, displaying an IC₅₀ comparable to that of tamoxifen, but had no effect on any of the PCa cells tested at its maximum evaluated concentration. This suggests that the estrogenic component of 1 interacts with the ER on MCF7 cells better than the androgenic component interacts with the AR on LNCaP PCa cells, possibly due to a suboptimal linkage. Nonetheless, the hybrid was active against colon (HT-29) and melanoma (M21) cancer cells and had low cross-reactivity with the drug- and androgen-metabolizing CYP3A4. In-depth studies will help to elucidate the precise mechanism of action of hybrid 1. The current findings demonstrate the anticancer potential of a unique type of bi-functional hybrid and warrant further explorations on these molecules, which can be constructed using natural hormones, as exemplified in this manuscript, with a mix of natural and synthetic hormones or simply by a combination of two synthetic hormones. Computer modeling approach and various head-groups, including dihydrotestosterone, will be utilized in our future studies for hybrid 1 optimization. Such hybrid compounds might be useful in cancer therapy, hormone replacement therapy or even as molecules allowing a perfect balance of desired hormonal/anti-hormonal activities for a patient.

Abbreviations:

BCa

breast cancer

BFC

7-benzyloxy-4-(trifluoromethyl)coumarin

CYP3A4

cytochrome P450 3A4

17EE

17α-ethynylestradiol

PCa

prostate cancer

TS

testosterone