Synthesis and Cytotoxicity Evaluation of Dehydroepiandrosterone Derivatives by Iron-Catalyzed Stereoselective Hydroamination

Abstract

The direct modification of structurally complex natural product dehydroepiandrosterone (DHEA) through iron-catalyzed direct hydroamination of DHEA with various nitro(hetero)arenes was carried out to afford 5α-arylamino-DHEAs (1–25) in good yields (53–72%). Though as a radical reaction, it features high stereoselectivity, and only the 5α-substituted derivatives were produced. The in vitro antiproliferative activity of these synthesized compounds against the human breast cancer MCF-7 cell was evaluated, showing that most of DHEA analogues possessed the moderate cytotoxic activity. The preliminary structure–activity relationship analysis revealed that the electron-withdrawing groups installed at the para-position of arylamine ring had a great contribution to the improvement of the DHEA’s cytotoxic potency. Among them, (4-trifluoromethylaniline)-DHEA (4) displayed the most potent cytotoxicity, with an IC₅₀ value of 19.3 µM, which was 2.3-fold more active than DHEA.

Introduction

Steroids as a branch of polycyclic natural products are widely existed in plants, animals, and humans.¹⁾ Dehydroepiandrosterone (DHEA), one of the naturally occurring endogenous steroid hormones, also known as androstenolone or prasterone, is a pharmacologically active molecule associated with various potential biological activities, such as anti-inflammatory,²⁾ antiviral,³⁾ immunomodulatory,⁴⁾ and moderate antitumor activity.^5–7) Many DHEA analogues, such as Abiraterone and Zytiga (Fig. 1), exhibit the excellent clinical therapeutics to defend the human health.⁸⁾ Given the double advantages of being both easily accessible and economical, DHEA as the suitable starting material has attracted considerable interest to develop new drugs in medicinal chemistry community. Many efforts have been devoted to search the potential DHEAs candidate drugs in recent years.^9–11) However, the previously reported traditional synthetic methods used in the DHEA’s structure modification have been complicated. Developing simple and convenient strategies to construct complex DHEA analogues remains an important direction in pharmaceutical synthesis.

Fig. 1. Structures of DHEA and Its Analogues

As the integral parts of biomolecules are present in the majority of top-selling anticancer drugs (such as Vemurafenib, Bosutinib, and Caprelsa), amines play the significant roles in their activities.^12,13) Hence, introducing containing-nitrogen units into the steroids to generate aminosteroids is an effective strategy for developing medicines. For example, the steroid derivative Galeterone contained a benzimidazole was developed as a candidate drug for the treatment of advanced prostate cancers, and enters the Phase III clinical trial stage at present¹⁴⁾ (Fig. 1). 3β-(1H-Imidazole-1-carboxylate)-17-(1H-benzimidazol-1-yl)androsta-5,16-diene was selected as the strong candidate drug, possessing the same effect as that of Galeterone.¹⁵⁾ Due to the ubiquitous application in pharmaceutical science, amines synthesis is one of the hotspots in the current study.

Conventional methods for synthesis of amine like amine-carbonyl reductive amination, amine N-alkylation by hydrogen auto-transfer, and transition-metal-catalyzed C–N cross-coupling show low functional group tolerance.¹⁶⁾ In 2015, Baran and colleagues reported a practical olefin hydroamination with nitroarenes using an abundant Fe(acac)₃ as a catalyst, and inexpensive phenylsilane and zinc metal as the reductants¹⁷⁾ (Chart 1). The reaction was proposed through a radical pathway. An alkyl radical derived from the donor olefins adducts a nitrosoarene which was generated in situ by hydrido-iron-mediated nitroarene reduction. Cleavage of the resultant N–O bond then liberates the desired hindered secondary amine. The reaction shows good functional group tolerance including substrates bearing halide, ketone, amide, and alcohol functionalities, and is an efficient tool for the construction of highly functionalized amines. The site-selective synthetic methodologies provide the efficient strategies for the diversification of complex natural products.¹⁸⁾ The efficient olefin hydroamination developed by Baran and colleagues has not been used in any synthesis or modification of natural products until now. As part of our researches toward development of active molecules deriving from fascinating natural products,^19–23) this work applied the distinct and convenient method mentioned above to prepare 5α-arylamino-DHEAs which were further explored their potential anti-tumor activity. It is worth noting that when the reaction was used to structurally diversify the DHEA, only the 5α-substituted derivatives were produced, showing the good stereoselectivity (Chart 1). Moreover, most of modified derivatives present the better cytotoxicity against MCF-7 cells than the parent compound DHEA, indicating introduction of arylamino group at C-5 position is beneficial to the cytotoxicity activity of DHEA.

Chart 1. Iron-Catalyzed Olefin Hydroamination

Results and Discussion

Chemistry

The work commenced with the direct preparation of derivatives 1–25 which were obtained through Iron-catalyzed radical hydroamination of DHEA Δ^5,6 double bond with various nitro(hetero)arenes (1a–25a) (Chart 2). At the initial stage of the investigation, based on the olefin hydroamination with nitroarenes under the Iron catalyst system reported by Baran,¹⁷⁾ the reaction of DHEA and nitrobenzene (1a) using PhSiH₃ and Zn/HCl as reductants was carried out. When extended the time to 2 h, the desired product 1 was obtained in 70% yield. To our delight, the reaction was very clean, only a product spot was found in TLC plate. Combining with the NMR spectra of isolated product, the result indicated the reaction showed the good stereoselectivity, which was rare in the radical reactions. Then, the scopes of nitroarenes and nitroheteroarenes were investigated. As exemplified in Chart 2, all of the investigated nitro(hetero)arenes direct provided the corresponding arylamino-DHEA products in moderate to good yields (53–72%). Both electron-withdrawing groups (fluoro, chloro, bromo, cyano, and trifluoromethyl groups) and electron-donating groups (methyl, ethyl, methoxy, and sulfurmethyl groups) at meta- or para-position of nitroarenes were well tolerated. The ortho-substituents of nitroarenes were suitable as well in this process to furnish products in yields of 56% (9), 53% (10), 70% (12), and 67% (13). Nitroheteroarenes including 2-methoxyl-5-nitropyridine, 3-methoxyl-2-nitropyridine, and 2-methoxyl-3-nitropyridine also successfully afforded the corresponding coupling products 23–25 in 69, 53 and 60% yields, respectively.

Chart 2. Synthesis of 5a-Arylamino-DHEA Derivatives 1–25

The structures of all synthetic compounds were confirmed by NMR and high resolution electrospray ionization (HR-ESI)MS spectra. All the obtained spectroscopic data supported the structures assigned. ¹H-NMR spectra showed the disappearance of downfield proton signals of ∆^{5, 6} cyclic olefinic bond in parent compound DHEA, and two corresponding carbon signals were also no longer observed in the ¹³C-NMR spectra. The reaction takes place on the more hindered site of double bond in DHEA. The expected signals relating to the various amine groups joined at C-5 of steroid framework were also exhibited in both ¹H- and ¹³C-NMR spectra. As a radical reaction, two products could be got when chiral substrate was applied. However, as mentioned above, only one stereoisomer was obtained when DHEA participated the reaction. It was supposed that the stereoselectivity was induced by the substrate. The 18-Me of DHEA was in β-orientation, leading that the intermediates nitrosoarenes could only attack from the α-orientation of alkyl radical to give 5α-arylamino-DHEAs. To confirm the supposition and the stereo-structures of these compounds, derivative 9 was selected as a representative compound and further characterized by X-ray single-crystal diffraction (CCDC 2201684). The single crystal of 9 was cultivated by slow evaporation at room temperature using petroleum ether and acetone (10 : 1, V : V) as solvents. As shown in Fig. 2, compound 9 formed eutectic with acetone. The central rings A, B and C of the steroid nucleus in compound 9 are chair conformations, and ring D shows a slightly distorted half-chair conformation. The configurations of known chiral centers in DHEA should remain unchanged during the present process of synthesis. The 18-Me and amino group are located at the opposite spatial orientations, demonstrating α-orientation of amino moiety in 9 with 5R configuration. The C-5-arylamino groups in this series of derivatives is unequivocally possess the same configuration as that of 9.

Fig. 2. Crystal Structure of Compound 9 (CCDC 2201684)

Cytotoxicity

The cytotoxicity against MCF-7 human breast cancer cells of twenty-five synthesized DHEA derivatives (1–25) was evaluated using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, employing DHEA and 5-fluorouracil as the positive controls. The primary screening test was conducted at a concentration of 40 µM. The inhibition rates of proliferation are shown in Table 1. Most of the synthesized DHEA analogues (2–4, 6–12, 15, 18, and 21) showed favorable antiproliferation efficacity superior to parent compound DHEA. Derivatives 3 and 4 bearing fluoro or trifluoromethyl group at para-position of benzene rings showed the most potential cytotoxicity with inhibition rates of 80.1 and 84.7% respectively at 40 µM, while DHEA only had 40.8% inhibition rate. Compound 2 with meta-chloro-arylamine group also behaved good antiproliferative activity with inhibition rate of 66.9%. Among derivatives owned the electron-donating groups in aromatic ring, compound 21 showed the best cytotoxic selectivity with inhibition rate of 66.1% than others. In general, the substituents on arylamine functionalities have significant impact on the activity. The introduction of electron-withdrawing groups in the benzene ring generally exhibited better activity than electron-donating groups. In addition, the substituents position in phenylamino moieties is one of the main influential factors of activity. The compounds installed substituents at para-position showed better activity than those having meta-substituents. For example, compound 3 possessing fluorine-substituted on the para-position of phenylamino exhibited good cytotoxic potency, while compound 5 having the meta-fluorine showed obvious decreased antiproliferative effect. Similarly, derivatives 15 and 18 with para-substitution displayed more effectively cytotoxicity against MCF-7 cells than compounds 16 and 19 with meta-substitution. For different ortho-substituted arylamine groups (9,10 and 12,13), the order of cytotoxicity is -F > -OMe, showing the halide atom substituted on ortho-position was better than the electron donating group. When DHEA was modified by introducing aminopyridine units to produced compound 23–25, but cytotoxic activities of them were below that of DHEA, indicating that the aminopyridine groups was not conducive to improving DHEAs’ cytotoxicity. Whereafter, the compounds with inhibition rates over 50% were subjected to IC₅₀ examination. As shown in Table 2, the IC₅₀ values of all the selected derivatives (IC₅₀ ranging from 19.3 to 35.1 µM) are better than the parent compound DHEA (IC₅₀ = 44.5 µM). Notably, 5α-(4-trifluoromethylaniline) DHEA (4) displayed the most potent cytotoxicity on MCF-7 cells with IC₅₀ value of 19.3 µM, which was 2.3-fold higher than that of parent compound.

Table 1. Antiproliferative Activity of Compounds 1–25 against MCF-7 Cells^a)

Compounds	Inhibition rate (%)b)	Compounds	Inhibition rate (%)b)
1	21.6 ± 0.9	15	43.1 ± 2.8
2	66.9 ± 2.6	16	32.4 ± 4.1
3	80.1 ± 2.1	17	10.4 ± 3.0
4	84.7 ± 2.1	18	52.3 ± 1.9
5	18.3 ± 2.3	19	37.7 ± 0.1
6	51.7 ± 2.1	20	12.1 ± 0.2
7	56.3 ± 0.1	21	66.1 ± 0.1
8	57.7 ± 2.0	22	25.3 ± 3.3
9	54.4 ± 0.1	23	24.3 ± 2.3
10	57.5 ± 2.9	24	24.7 ± 2.3
11	41.1 ± 0.5	25	9.2 ± 1.7
12	42.4 ± 6.9	DHEA	40.8 ± 3.3
13	15.2 ± 5.0	5-Fluorouracil	92.1 ± 2.7
14	23.9 ± 1.1

a) Concentration of 40 µM. b) Data are expressed as the means ± standard deviation (n = 3).

Table 2. IC₅₀ Values of Sselected Ccompounds for Antiproliferative Activity against MCF-7 Cells

Compounds	IC50 (µM)a)
2	22.5 ± 1.5
3	22.0 ± 2.5
4	19.3 ± 0.5
6	32.7 ± 0.7
7	27.7 ± 0.9
8	26.4 ± 1.3
9	27.7 ± 0.3
10	29.2 ± 1.3
18	35.1 ± 2.6
21	22.9 ± 0.3
DHEA	44.5 ± 0.4
5-Fluorouracil	4.9 ± 0.3

a) Data are expressed as the means ± S.D. (n = 3).

Conclusion

In summary, twenty-five novel 5α-arylamino-DHEAs were designed and synthesized by direct modification using Iron catalyzed stereoselective radical olefin hydroamination of DHEA. The reaction was firstly applied to the structural modification of complex natural product, introducing the arylanimo units into DHEA to enhance its cytotoxicity against MCF-7 cell lines. Only 5α-coupling products were obtained, indicating that the radical reaction showed the good stereoselectivity. Most of the synthesized compounds displayed potent antiproliferative activity in vitro against MCF-7 cells superior to parent compound DHEA. Moreover, 5α-(4-trifluoromethylaniline)-DHEA (4) with an IC₅₀ value of 19.3 µM exhibited the best cytotoxicity among the derivatives, and could be further explored as a promising lead compound.

Experimental

General Experimental Procedures

DHEA was purchased from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China), with a purity of >98%. All of the reagents and solvents were used directly as obtained commercially unless otherwise noted. Analytical TLC was performed on silica gel plates (GF 245) and visualized by UV irradiation (254 nm) or by staining with iodine. Column chromatography was carried out using silica gel 200–300 mesh (Qingdao Sea Chemical Factory, Qingdao, People’s Republic of China) under pressure. ¹H- and ¹³C-NMR spectra were recorded in CDCl₃ or CD₃OD at ambient temperature on Bruker AV 400 or 600 NMR instruments. Chemical shifts were recorded in ppm relative to tetramethylsilane as the internal standard. HRESI-MS spectra were taken on Waters Acquity UPLC/Xevo G2-S QT mass spectrometer. Absorbance was recorded on microplate reader (SpectraMax CMax Plus, Molecular Devices, Sunnyvale, CA, U.S.A.). Optical rotations were measured using a PerkinElmer, Inc. 341 polarimeter.

General Synthesis of Compounds 1–25

To a solution of the DHEA (0.4 mmol, 2 equivalent (equiv.)) and Fe(acac)₃ (30 mol%) in EtOH (5.0 mL) was added nitroarenes (0.2 mmol, 1 equiv.), and PhSiH₃ (0.4 mmol, 2 equiv.) at room temperature. The resulting mixture was stirred at 60 °C for 1 h. Then, the reaction mixture was cooled to room temperature, 2N HCl (1 mL) was added, followed by careful addition of zinc powder (4.0 mmol, 20 equiv). After heating at 60 °C for 1 h, the reaction mixture was cooled to room temperature. Reaction suspension was filtered over a pad of Celite^®. The mixture was added to a saturated solution of NaHCO₃ (aq.) to adjust the mixture to pH 7 and then the mixture was extracted with EtOAc three times. The organic layer was washed with H₂O and brine, dried with Na₂SO₄, filtered, and concentrated. The crude mixture was purified by silica gel column chromatography (PE/EtOAc, about 6 : 1) to afford the desired derivatives 1–25 in 53–72% yields. The ¹H-NMR, ¹³C-NMR, and HRESIMS data of the synthesized compounds are listed in the Supplementary Materials.

X-Ray Diffraction Analysis of Compound 9

Colorless crystals of compound 9 was recrystallized from Petroleum ether/acetone (10 : 1, V : V) at room temperature. X-Ray data were collected on Oxford Xcalibur Eos diffractometer with Mo Kα radiation (λ = 0.71073 Å). Each structure was solved by direct methods using SHELXL-97, and all atoms were refined anisotropically using full-matrix least-squares difference Fourier techniques. Crystallographic data for the structure of 9 have been deposited with the Cambridge Crystallographic Data Centre as supplementary publication CCDC 2201684. Copies of these data can be obtained, free of charge, on application to the CCDC via www.ccdc.cam.ac.uk/conts/retrieving.html (or 12 Union Road, Cambridge CB2 1EZ, UK, fax: + 441223 336033, e-mail: [e-mail]).

Crystallographic Data of 9. C₂₈H₃₉BrFNO₃, mass (M) = 536.51 g/mol, monoclinic, a = 10.0122(7) Å, b = 12.8687(10) Å, c = 20.3135(14) Å, α = 90°, β = 90°, γ = 90°, V = 2617.3(3) ų, Z = 4, T = 293.15 K, μ(MoKα) = 1.607 mm⁻¹, F (000) = 1128.0, D_calc = 1.362 g/cm³. A total of 8357 reflections measured (6.332° ≤ 2θ ≤ 52.738°), containing 5255 unique reflections (R_int = 0.0264, R_sigma = 0.0696), which were used in all calculations. The final R₁ was 0.0523 (I > 2σ(I)) and wR₂ was 0.1076 (all data). The goodness of fit on F² was 1.003. Flack parameter = 0.007(8).

Cell Culture and Cytotoxic Assay

MCF-7 (human breast cancer cell line) cells were obtained from ATCC, and the cells were cultured in Dulbecco’s Modified Eagle’s Medium-High (DMEM) medium (containing 10% fetal bovine serum (FBS), 100 µg/mL penicillin, and 0.03% L-glutamine) and incubated at 37 °C with a humidified 5% CO₂ air atmosphere. The cytotoxic activities of compounds 1–25 were determined using an MTT assay. MTT (M2128) was purchased from Sigma-Aldrich Co., Ltd. (St. Louis, MO, U.S.A.). The MCF-7 cells were seeded into 96-well microplates at a density of 1 × 10⁴ cells per well and supplemented with culture medium. After incubation for 24 h, tested compounds (dissolved in dimethyl sulfoxide (DMSO)) were added to each well at various concentrations and incubated for 24 h. Then, 10 µL of MTT (5 mg/mL) was added into each well and incubated for another 4 h in the dark. The absorbance was detected at a wavelength of 492 nm using a microplate reader (Thermo Fisher MK3, U.S.A.).

Acknowledgments

The research was financially supported by Grants from the National Natural Science Foundation of China (31870329 and 82003634) and the Fundamental Research Funds for the Central Universities of China (Grant Nos. 2682020ZT88 and 2682021CX089).

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References

• 1) Shagufta, Ahmad I., Panda G., Eur. J. Med. Chem., 133, 139–151 (2017).
• 2) Maciel-Baron L. A., Morales-Rosales S. L., Rodriguez-Barrera R. H., Garcia-Alvarez J. A., Luna-Lopez A., Perez V. I., Torres C., Konigsberg M., Biogerontology, 19, 415–433 (2018).
• 3) Torres N. I., Castilla V., Bruttomesso A. C., Eiras J., Galagovsky L. R., Wachsman M. B., Antiviral Res., 95, 37–48 (2012).
• 4) Zhang N., Gui Y. Y., Qiu X. M., Tang W., Li L. S., Gober H. J., Li D. J., Wang L., BioScience Trends, 10, 277–287 (2016).
• 5) Girón R. A., Montano L. F., Escobar M. L., Lopez-Marure R., FEBS J., 276, 5598–5609 (2009).
• 6) Liu L., Wang D., Li L. L., Ding X., Ma H. T., Mol. Med. Rep., 14, 705–714 (2016).
• 7) Lopez-Marure R., Contreras P. G., Dillon J. S., Eur. J. Pharmacol., 660, 268–274 (2011).
• 8) Watson P. A., Arora V. K., Sawyers C. L., Nat. Rev. Cancer, 15, 701–711 (2015).
• 9) Çapan I., Shehu A., Sert Y., Celik I., Erol M., Koca I., Servi S., J. Biomol. Struct. Dyn., 1–12 (2022).
• 10) Romero-Hernández L. L., Merino-Montiel P., Meza-Reyes S., Vega-Baez J. L., Lopez O., Padron J. M., Montiel-Smith S., Eur. J. Med. Chem., 143, 21–32 (2018).
• 11) Huang X., Shen Q. K., Zhang H. J., Li J. L., Tian Y. S., Quan Z. S., Molecules, 23, 2243 (2018).
• 12) Magano J., Dunetz J. R., Chem. Rev., 111, 2177–2250 (2011).
• 13) Akhtar J., Khan A. A., Ali Z., Haider R., Yar M. S., Eur. J. Med. Chem., 125, 143–189 (2017).
• 14) Alyamani M., Li Z. F., Berk M., Li J. N., Tang T. T., Upadhyay S., Auchus R. J., Sharifi N., Cell. Chem. Biol., 24, 825–832.e6 (2017).
• 15) Purushottamachar P., Godbole A. M., Gediya L. K., Martin M. S., Vasaitis T. S., Kwegyir-Afful A. K., Ramalingam S., Ates-Alagoz Z., Njar V. C. O., J. Med. Chem., 56, 4880–4898 (2013).
• 16) Zhu K. L., Shaver M. P., Thomas S. P., Chem. Asian J., 11, 977–980 (2016).
• 17) Gui J. H., Pan C. M., Jin Y., Qin T., Lo J. C., Lee B. J., Spergel S. H., Mertzman M. E., Pitts W. J., La Cruz T. E., Schmidt M. A., Darvatkar N., Natarajan S. R., Baran P. S., Science, 348, 886–891 (2015).
• 18) Hong B., Luo T., Lei X., ACS Cent. Sci., 6, 622–635 (2020).
• 19) Miao S. X., Wan L. X., He Z. X., Zhou X. L., Li X. H., Gao F., J. Nat. Prod., 84, 2374–2379 (2021).
• 20) Wan L. X., Zhen Y. Q., He Z. X., Zhang Y., Zhang L., Li X. H., Gao F., Zhou X. L., ACS Omega, 6, 9960–9972 (2021).
• 21) Wan L. X., Miao S. X., He Z. X., Li X. H., Zhou X. L., Gao F., ACS Omega, 6, 23347–23354 (2021).
• 22) Du Y. F., Huang L. Y., Wang N., Li X. H., Zhou X.-L., Zhang L., Huang S., Walsh P. J., Gao F., Adv. Synth. Catal., 364, 1002–1008 (2022).
• 23) Huang L. Y., Xu J. B., Li X. Y., Song H. N., Chen L., Zhou X. L., Gao F., J. Nat. Prod., 85, 2026–2034 (2022).