Improved synthesis of EM-1745, preparation of its C17-ketone analogue and comparison of their inhibitory potency on 17β-hydroxysteroid dehydrogenase type 1

Abstract

Endocrine therapies are widely used for the treatment of estrogen-sensitive diseases. 17β-hydroxysteroid dehydrogenase type 1 (17β-HSD1) is involved in the last step of the biosynthesis of potent estrogen estradiol (E₂). This enzyme catalyzes the reduction of the C17-ketosteroid estrone (E₁) into the C17β-hydroxy steroid E₂ using the cofactor NAD(P)H. The X-ray analysis of E₂/adenosine bisubstrate inhibitor EM-1745 proven that this compound interacts with both the substrate- and the cofactor-binding sites. However, E₁ is a better substrate of 17β-HSD1 than E₂. Thus, in order to improve the inhibitory potency of EM-1745, the C17-ketone analogue was prepared. During this work, a new and more efficient method for synthesizing EM-1745 was developed using an esterification and a cross-metathesis as key steps. Contrary to what was expected, the C17-ketone analogue of EM-1745 is a less potent inhibitor (IC₅₀ = 12 nM) than the C17-alcohol (IC₅₀ = 4 nM) in homogenated HEK-293 cells overexpressing 17β-HSD1. Our results contribute to the knowledge of an unexpected observation: the C17-ketone steroidal inhibitors of 17β-HSD1 are less potent than their corresponding C17-alcohol derivatives.

Keywords::

• 17β-Hydroxysteroid dehydrogenase
• enzyme
• inhibitor
• steroid
• chemical synthesis
• cross-metathesis

Introduction

Estrogens are well known to be involved in the development of estrogen-sensitive diseases such as breast cancer. In fact, most breast cancers are initially hormone-dependent and estradiol (E₂), the most potent estrogen, plays a crucial role in their development and progression [1,2]. Several endocrine therapies were thus developed since this approach is less toxic than chemotherapy [3]. Consequently, hormonal control should become a predominant choice in the treatment of estrogen-sensitive diseases. Antiestrogens, such as blockers of the estrogen receptor (ER), have been developed and their efficiency has been proved by several clinical trials [4]. Another option in the treatment of ER⁺ breast cancer is the use of an inhibitor of a steroidogenic enzyme involved in the biosynthesis of E₂ 567. In fact, three generations of inhibitors of aromatase, an enzyme catalyzing the conversion of androgens to estrogens, have been developed and the third one is now used for treatment of breast carconima 8910. Inhibitors of steroid sulfatase, an enzyme which hydrolyses the sulfates from circulating estrogens, androgens and hormone precursors, were also prepared, but are not yet available for clinical use 11121314. Another group of steroidogenic enzymes, the 17β-hydroxysteroid dehydrogenases (17β-HSDs), catalyze the last step in the biosynthesis of potent estrogens and androgens. However, even if this family of enzymes was discovered more than fifty years ago [15,16], we have yet to obtain potent inhibitors for clinical use.

Among the 17β-HSDs, we are interested in 17β-HSD1. This reductive enzyme, also called human estradiol dehydrogenase [E.C.1.1.1.62], catalyzes the last step in the biosynthesis of the most potent estrogen, E₂. This enzyme is also a member of the short-chain alcohol dehydrogenase family and exists as a homodimer. It is a protein of 327 amino acids with a subunit mass of 35 kDa [17]. It stereoselectively reduces the C17-ketone of estrone (E₁) preferentially using cofactor NADPH to provide E₂ (Figure 1) 17181920. 17β-HSD1 also catalyzes, at a lower rate, the transformation of dehydroepiandrosterone (DHEA) into 5-androstene-3β,17β-diol (Δ⁵-diol), a weak estrogenic C19 steroid. The reductive activity of 17β-HSDs was found to be present in several steroidogenic and peripheral tissues such as placenta, ovary and breast [21]. Furthermore, a study revealed that the reductive activity (E₁ into E₂) is higher in malignant breast tumours whereas the oxidative pathway (E₂ into E₁) is dominant in normal breast cell [22]. 17β-HSD1 mRNA has been detected in malignant breast tumours by in situ hybridization, immunohistochemistry and reverse transcription-PCR 232425. It was also demonstrated that 17β-HSD1 plays a role in the local biosynthesis of estrogen in the breast, especially in postmenopausal women, after the ovaries have ceased to produce estrogen [26,27]. Thus, 17β-HSD1 is an interesting target for estrogen-sensitive disease such as breast cancer.

Figure 1.  17β-HSD1 catalyzes the reduction of estrone (E₁) into estradiol (E₂) with the cofactor NADH or NADPH.

Several inhibitors have been developed for 17β-HSD1 328293031323334353637. Among these inhibitors E₂/adenosine hybrid compounds are bisubstrate inhibitors interacting with both the substrate-binding (E₂) and the cofactor-binding (adenosine) sites and were developed in part based on the three-dimensional structure of the enzyme 383940. With an alkyl side chain spacer of 8 methylenes in steroidal position C16β to link the two moieties (E₂ and adenosine), EM-1745 (1, Figure 2) is the best of that series of inhibitors with a K_i of 3 nM (E₂ into E₁). In addition, this compound was found to be 13-fold more potent than unlabeled E₁ used as inhibitor in homogenated HEK-293 cells overexpressing 17β-HSD1 [41,42]. Simplified bisubstrate inhibitors were also developed in order to improve the bioavailability of 1 [43]. These are analogues of 1 where the adenosine moiety was replaced by a benzyl group bearing a carboxylic acid function. These inhibitors are, however, less potent than EM-1745.

Figure 2.  Chemical structure of the bisubstrate inhibitors of 17β-HSD1: EM-1745 (1) and its C17-ketone analogue 2. Illustrated only for 1 and 2, the stereogenic centers are the same for all other steroid derivatives reported in this paper.

It is known that E₁, a ketosteroid, is a better substrate for the reductive activity of 17β-HSD1 than the corresponding hydroxy-steroid E₂. In fact, the Michaelis constant (K_m) for E₁ is 0.03 and 0.36 μM when using respectively the cofactors NADPH and NADH while the K_m for E₂ is higher with 4.6 and 1.7 μM when using the cofactors NADP⁺ and NAD⁺[44]. In addition, X-ray analysis of the EM-1745/17β-HSD1 complex confirmed that this inhibitor was interacting with both the substrate- and the cofactor-binding sites. It was also shown that the two moieties of EM-1745 interact with most of the amino acids with which E₂ or E₁ interacts in the substrate-binding site and with which adenosine of NAD(P)H interacts in the cofactor-binding site. Furthermore, a Lineweaver-Burk plot demonstrated clearly that EM-1745 is a typical reversible competitive inhibitor [42]. Taken together the results discussed above suggest that replacing the E₂ nucleus by an E₁ moiety in the structure of EM-1745 should give a better enzyme inhibition. Thus, in order to improve the inhibitory effect of EM-1745 (1), we prepared the C17-ketone analogue 2 (Figure 2). Herein, we report the chemical synthesis of the C17-ketone analogue of EM-1745 (compound 2), as well as a more convenient procedure for synthesizing EM-1745 (1). Moreover, the inhibitory potency of 2 was evaluated on 17β-HSD1 and compared to that of its corresponding C17-hydroxy analogue (EM-1745, 1).

Material and methods

General

Reagents were obtained from Sigma-Aldrich Canada Co. (Oakville, ON, Canada) except for benzotriazole-1-yl-oxy-tris-pyrrolidino-phosphonium hexafluorophosphate (PyBOP) which was purchased from NovaBiochem (EMD Biosciences Inc, La Jolla, CA, USA) and tricyclohexylphosphine [1,3-bis(2,4,6-trimethylphenyl)-4,5-dihydroimidazol-2-ylidene] [benzylidine] ruthenium(IV) dichloride (2nd generation Grubbs catalyst) from Strem Chemicals (Newburyport, MA, USA). Usual solvents were obtained from Fisher Scientific (Montréal, Qc, Canada) and VWR (Ville Mont-Royal, Qc, Canada) and were used as received. Anhydrous solvents were purchased from Aldrich and VWR in SureSeal bottles, which were conserved under positive argon pressure. Tetrahydrofuran (THF) was distilled from sodium/benzophenone under argon. All anhydrous reactions were performed under positive argon pressure in oven-dried glassware. Thin-layer chromatography (TLC) was performed on 0.25-mm silica gel 60 F₂₅₄ plates from Whatman (distributed by Fisher Scientific) and compounds were visualized by exposure to UV light (254 nm) and/or with a solution of ammonium molybdate/sulphuric acid/water (with heating). Flash chromatography was performed on Silicycle 60 (Québec, Qc, Canada) 230-400 mesh silica gel. Infrared (IR) spectra were obtained neat, from a thin film of the solubilized compound on NaCl pellets (usually in CH₂Cl₂) or from a KBr pellet. They were recorded on a Perkin-Elmer series 1600 FT-IR spectrometer (Norwalk, CT, USA); only significant bands are reported (in cm¹). ¹H- and ¹³C-NMR spectra were recorded with a Bruker AC/F 300 spectrometer (Billerica, MA, USA) at 300 and 75 MHz, respectively, and a Bruker AVANCE 400 spectrometer at 400 (¹H) and 100 (¹³C) MHz. The chemical shifts (δ) are expressed in ppm and referenced to chloroform (7.26 and 77.0 ppm), acetone (2.07 and 206.0 ppm), or methyl sulfoxide (2.51 and 39.5 ppm) for ¹H and ¹³C respectively. Duplication of NMR signals was generally recorded for THP derivatives. In that case, the presence of two stereoisomers increased the complexity of ¹³C NMR spectra, and additional peaks are written between parentheses. Assignment of NMR signals was made easier using literature data 454647. ¹³C NMR data are reported in Table I. Low-resolution mass spectra (LRMS) were recorded with an LCQ Finnigan apparatus (San Jose, CA, USA) equipped with an atmospheric pressure chemical ionisation (APCI) source on positive or negative mode.

Table I.  ¹³C NMR data for compounds 2, 7-11.

C #	2^*	7^†	8^†	9^†	10^†	11^†
C1′	126.0	126.9	126.9	126.9	126.8	126.8
C2′	112.8	117.9	117.8	117.9	113.4	113.4
C3′	155.0	154.0	153.9	153.9	155.7	155.8
C4′	115.0	120.6	120.5	120.5	115.7	115.8
C5′	137.1	138.4	138.4	138.4	138.3	138.2
C6′	29.1	‡	‡	‡	‡	‡
C7′	28.0	28.0	27.9	28.0	28.2	27.32
C8′	c	40.5	40.5	39.2 (40.5)	41.2	38.7
C9′	47.8	44.6 (44.8)	44.2	44.6 (44.76)	44.8	49.4
C10′	130.0	133.9 (134.0)	134.0	133.9 (134.0)	132.0	131.4
C11′	26.3	26.9 (27.0)	26.9 (27.0)	27.1	27.1	26.5
C12′	37.6	38.7 (39.1)	38.7 (39.0)	39.5	38.6	34.1
C13′	48.0	44.6 (44.8)	44.6 (44.7)	44.2 (44.79)	44.7	48.7
C14′	48.5	49.4 (49.6)	49.3 (49.5)	49.4 (49.6)	49.4	49.6
C15′	28.4	32.4 (32.6)	32.8 (32.9)	33.0 (33.1)	33.2	28.5
C16′	43.6	39.2 (39.4)	39.4 (39.5)	39.06 (39.10)	39.3	44.8
C17′	221.7	86.1 (86.4)	86.0 (86.6)	86.1 (86.7)	82.49	221.5
C18′	13.7	13.6	13.4	13.5	12.9	14.2
(CH3)2Si	–	− 4.2	− 4.1	− 4.4	–	–
(CH3)3C-Si	–	18.6	18.5	18.5	–	–
(CH3)3C-Si	–	25.9	25.8	25.9	–	–
CHO of THP	–	98.1 (99.6)	98.0 (99.4)	98.1 (99.4)	–	–
CH2O of THP	–	62.0 (62.9)	61.9 (62.8)	61.9 (62.8)	–	–
CH2CHO of THP	–	31.5 (31.6)	31.3 (31.6)	31.4 (31.6)	–	–
CH2CH2O of THP	–	26.2 (26.3)	26.2 (26.3)	26.2 (26.3)	–	–
CH2CH2CHO of THP	–	20.0 (20.5)	19.9 (20.4)	20.4	–	–
CH = CH	–	130.9 (131.0) 131.3	–	–	–	–
CHO or COO	172.8	202.6	174.4	173.2	173.2	173.2
CH2CHO or CH2COO	33.3	44.2	34.0	38.8	34.1	32.9
C1′ of ribose	87.7	–	–	91.0	91.0	90.9
C2′ of ribose	72.9	–	–	84.7	84.7	84.7
C3′ of ribose	70.3	–	–	82.6	82.46	82.5
C4′ of ribose	81.4	–	–	85.4	85.3	85.3
C5′ of ribose	63.7	–	–	64.3	64.3	64.3
C(CH3)2	–	–	–	25.41 & 27.3	25.4 & 27.3	25.3 & 27.26
C(CH3)2	–	–	–	114.5	114.4	114.4
C adenine	156.1	–	–	157.1	157.0	157.0
C adenine	152.7	–	–	153.6	153.6	153.6
C adenine	149.3	–	–	150.0	150.0	150.0
C adenine	139.7	–	–	140.5	140.6	140.6
C adenine	119.1	–	–	120.7	120.9	120.6
C alkyl chain	31.6	36.3 (36.5)	33.1	34.2	32.52	32.7
C alkyl chain	28.8	33.0	25.5	30.4	32.48	6C^‡
C alkyl chain	28.7	30.5	5C^‡	25.37	30.22	–
C alkyl chain	28.6	22.5	–	4C^‡	29.1	–
C alkyl chain	27.4	1C^‡	–	–	25.3	–
C alkyl chain	25.5	–	–	–	2C^‡	–
C alkyl chain	24.4	–	–	–	–	–

^* DMSO-d₆ as solvent

^† acetone-d₆ as solvent

^‡ under solvent peak.

Synthesis of 7-octen-1-ol (4)

Halogen exchange (Br-I). To a solution of 6-bromo-1-hexanol (3) (2.90 mL, 22.0 mmol) in acetone (50 mL) was added NaI (13.2 g, 88.1 mmol) and the solution was refluxed and stirred for 16 h. After addition of water, the product was extracted with EtOAc and the organic phase was washed with a saturated aqueous solution of Na₂S₂O₃, washed with brine, dried over MgSO₄, and evaporated to dryness. 6-Iodo-1-hexanol (4.71 g, 94% crude yield) was used for the next step without further purification. Vinylation. CuI (3.58 g, 18.8 mmol) was suspended in anhydrous THF (40 mL) under an argon atmosphere. The mixture was cooled to − 40°C and vinylmagnesium bromide (1 M in THF) (94.4 mL, 94.4 mmol) was added. The reaction was stirred for 15 min at − 40°C. Then, HMPA (6.54 mL, 37.6 mmol) and triethyl phosphite (6.44 mL, 37.6 mmol) were added and the mixture was stirred 5 min at − 40°C. 6-Iodo-1-hexanol (4.28 g, 18.8 mmol) was added and the reaction mixture was stirred 1 h at − 40°C and 4 h at room temperature. The reaction was quenched by addition of a saturated aqueous solution of NH₄Cl. The crude product was extracted with EtOAc, and the organic phase was washed with brine, dried over MgSO₄, and evaporated under reduced pressure. Purification by flash chromatography (hexanes/EtOAc, 85:15) yielded 4 (1.80 g, 75% yield for two steps) as a colourless oil. IR (neat) 3342 (OH), 3077 (H-C = ), 1641 (C = C, alkene); ¹H NMR (400 MHz, CDCl₃) 1.31 (m, 3 × CH₂), 1.52 (m, CH₂CH₂OH), 2.01 (m, CH₂CH = CH₂), 2.49 (s_br, OH), 3.58 (t, J = 6.6 Hz, CH₂OH), 4.93 (m, CH = CH₂), 5.77 (m, CH = CH₂); ¹³C NMR (75 MHz, CDCl₃) 25.5, 28.76, 28.81, 32.5, 33.6, 62.7, 114.1, 138.9; LRMS calculated for C₈H₁₅O [M-H]⁻ 127.1, found 127.0 m/z.

Synthesis of 7-octenal (5)

4-Methylmorpholine-N-oxide (NMO) (3.30 g, 28.2 mmol) and molecular sieves (5 g) were added to a solution of alcohol 4 (1.51 g, 11.8 mmol) in dry DCM (100 mL) under an argon atmosphere at room temperature. The mixture was stirred for 15 min and tetrapropylammonium perruthenate (TPAP) (207 mg, 0.59 mmol) was then added. After the reaction mixture was stirred for 90 min, the solvent was removed under reduced pressure. The crude product was purified by flash chromatography (hexanes/EtOAc, 95:5) to afford aldehyde 5 (765 mg, 51% yield) as a colourless oil. IR (neat) 3077 (H-C = ), 2718 (C-H, aldehyde), 1727 (C = O, aldehyde), 1641 (C = C, alkene); ¹H NMR (400 MHz, CDCl₃) 1.37 (m, 2 × CH₂), 1.64 (m, CH₂CH₂CHO), 2.05 (m, CH₂CH = CH₂), 2.43 (dt, J₁ = 7.4 Hz, J₂ = 1.7 Hz, CH₂CHO), 4.98 (m, CH = CH₂), 5.79 (m, CH = CH₂), 9.76 (t, J = 1.7 Hz, CHO); ¹³C NMR (75 MHz, CDCl₃) 21.9, 28.5 (2x), 33.5, 43.8, 114.5, 138.7, 202.8.

Synthesis of 9-[3′-(tert-butyldimethylsilyloxy)-17′β-(tetrahydro-2H-pyran-2-yl-oxy)-estra-1′,3′,5′(10′)-trien-16′β-yl]-7-nonen-1-al (7)

Diprotected 16β-allyl-estradiol 6 was synthesized in five steps from estrone as previously reported [48]. A mixture of 6 (1.00 g, 1.96 mmol), freshly prepared aldehyde 5 (745 mg, 5.90 mmol) and tricyclohexylphosphine [1,3-bis(2,4,6-trimethylphenyl)-4,5-dihydroimidazol-2-ylidene] [benzylidine] ruthenium(IV) dichloride (Grubbs's catalyst 2nd generation) (166 mg, 0.196mmol) in dry DCM (50 mL) was refluxed for 16 h under an argon atmosphere. The crude mixture was preadsorbed on silica gel and a flash chromatography was performed with hexanes/EtOAc, 95:5 as eluent to afford the desired steroid 7 (597 mg, 50% yield) as a yellowish viscous oil. IR (film) 2714 (C-H, aldehyde), 1727 (C = O, aldehyde); ¹H NMR (400 MHz, acetone-d₆) 0.205 and 0.206 (2s, Si(CH₃)₂), 0.83 and 0.87 (2s, 18′-CH₃), 1.00 (s, SiC(CH₃)₃), 1.05 to 2.35 (m, 28H, CH and CH₂ of steroid skeleton and alkyl chain), 2.44 (t, J = 7.3 Hz, CH₂CHO), 2.81 (m, 6′-CH₂), 3.48 and 3.92 (2m, OCH₂ of THP), 3.76 and 3.81 (2d, J = 9.8 Hz, 17′α-CH), 4.66 and 4.72 (2m, CH of THP), 5.44 (m, CH = CH), 6.58 (d, J = 1.8 Hz, 4′-CH), 6.64 (d, J = 8.5 Hz, 2′-CH), 7.16 (m, 1′-CH), 9.74 (t, J = 1.5 Hz, CHO); ¹³C NMR (75 MHz, acetone-d₆) − 4.2 (2x), 13.6, 18.6, 20.0 (20.5), 22.5, 25.9 (3x), 26.2 (26.3), 26.9 (27.0), 28.0, 29.0 to 30.2 (2C under solvent peaks), 30.5, 31.5 (31.6), 32.4 (32.6), 33.0, 36.3 (36.5), 38.7 (39.1), 39.2 (39.4), 40.5, 44.2, 44.6 (44.8) (2x), 49.4 (49.6), 62.0 (62.9), 86.1 (86.4), 98.1 (99.6), 117.9, 120.6, 126.9, 130.9 (131.0), 131.3, 133.9 (134.0), 138.4, 154.0, 202.6.

Synthesis of 9-[3′-(tert-butyldimethylsilyloxy)-17′β-(tetrahydro-2H-pyran-2-yl-oxy)-estra-1′,3′,5′(10′)-trien-16′β-yl]-nonanoic acid (8)

Hydrogenation. A suspension of 7 (588 mg, 0.965 mmol) and 10% Pd/C (88 mg) in EtOAc (20 mL) was stirred under a hydrogen atmosphere at room temperature. After 16 h, the resulting suspension was filtered through celite, washed with EtOAc and evaporated to dryness to afford 9-[3′-(tert-butyldimethylsilyloxy)-17′β-(tetrahydro-2H-pyran-2-yl-oxy)-estra-1′,3′,5′(10′)-trien-16′β-yl]-nonanal (593 mg, quantitative crude yield) in good purity without purification. Oxidation. This aldehyde (498 mg, 0.815 mmol) was dissolved in a minimum of THF (∼2 mL) followed by addition of t-BuOH (29 mL) and 2-methyl-2-butene (11 mL). An oxidative solution freshly prepared by dissolving NaClO₂ (860 mg) and NaH₂PO₄ (860 mg) in H₂O (8.6 mL) was added and the reaction mixture was allowed to stir for 30 min. The reaction was quenched by addition of water and the extraction was performed with EtOAc. The organic phase was washed with brine, dried over MgSO₄, and evaporated to dryness under reduced pressure. Purification by flash chromatography (hexanes/EtOAc, 8:2) provided carboxylic acid 8 (383 mg, 75% yield for two steps) as a gummy white foam. IR (film) 3600-2300 (OH, carboxylic acid), 1709 (C = O, carboxylic acid); ¹H NMR (400 MHz, acetone-d₆) 0.202 and 0.203 (2s, Si(CH₃)₂), 0.82 and 0.86 (2s, 18′-CH₃), 1.00 (s, SiC(CH₃)₃), 1.00 to 2.35 (m, 32H, CH and CH₂ of steroid skeleton and alkyl chain), 2.30 (t, J = 7.4 Hz, CH₂COOH), 2.81 (m, 6′-CH₂), 3.48 and 3.92 (2m, OCH₂ of THP), 3.74 and 3.79 (2d, J = 10.0 Hz, 17′α-CH), 4.64 and 4.73 (2m, CH of THP), 6.58 (d, J = 2.2 Hz, 4′-CH), 6.64 (d, J = 8.4 Hz, 2′-CH), 7.16 (m, 1′-CH); ¹³C NMR (75 MHz, acetone-d₆) − 4.1 (2x), 13.4, 18.5, 19.9 (20.4), 25.5, 25.8 (3x), 26.2 (26.3), 26.9 (27.0), 27.9, 28.9 to 30.4 (6C under solvent peaks), 31.3 (31.6), 32.8 (32.9), 33.1, 34.0, 38.7 (39.0), 39.4 (39.5), 40.5, 44.2, 44.6 (44.7), 49.3 (49.5), 61.9 (62.8), 86.0 (86.6), 98.0 (99.4), 117.8, 120.5, 126.9, 134.0, 138.4, 153.9, 174.4; LRMS calculated for C₃₈H₆₁O₅Si [M-H]⁻ 625.4, found 625.6 m/z.

Synthesis of 5′-O-{9-[3′-(tert-butyldimethylsilyloxy)-17′β-(tetrahydro-2H-pyran-2-yl-oxy)-estra-1′,3′,5′(10′)-trien-16′β-yl]-nonanoyl}-2′,3′-O-isopropylidene adenosine (9)

To a solution of carboxylic acid 8 (422 mg, 0.673 mmol) in dry DMF (7 mL) under an argon atmosphere at room temperature was added benzotriazole-1-yl-oxy-tris-pyrrolidinophosphonium hexafluorophosphate (PyBOP) (525 mg, 1.01 mmol), hydroxybenzotriazole (HOBt) (136 mg, 1.01 mmol) and diisopropylethylamine (DIPEA) (352 μL, 2.02 mmol). The reaction mixture was stirred for 5 min. Then, 2′,3′-isopropylidene adenosine (414 mg, 1.35 mmol) was added and the mixture was stirred for 16 h at room temperature. Water was added to quench the reaction and the extraction was performed with EtOAc. The organic phase was washed with brine, dried over MgSO₄ and evaporated under reduced pressure. Purification of the crude residue by flash chromatography (hexanes/EtOAc, 5:5 to 4:6) afforded 9 (312 mg, 51% yield) as a white foam. IR (film) 3324 and 3172 (NH₂), 1742 (C = O, ester), 1643 (C = N); ¹H NMR (400 MHz, acetone-d₆) 0.20 (s, Si(CH₃)₂), 0.81 and 0.85 (2s, 18′-CH₃), 0.99 (s, SiC(CH₃)₃), 1.00 to 2.35 (m, 32H, CH and CH₂ of steroid skeleton and alkyl chain), 1.39 and 1.59 (2s, 2 × CH₃ of isopropylidene), 2.26 (dt, J₁ = 7.5 Hz, J₂ = 2.7 Hz, CH₂COO), 2.81 (m, 6′-CH₂), 3.47 and 3.91 (2m, OCH₂ of THP), 3.73 and 3.78 (2d, J = 10.0 Hz, 17′α-CH), 4.24 and 4.31 (2m, 5′-CH₂ of ribose), 4.44 (m, 4′-CH of ribose), 4.63 and 4.72 (2m, CH of THP), 5.17 (m, 3′-CH of ribose), 5.59 (dd, J₁ = 6.2 Hz, J₂ = 2.0 Hz, 2′-CH of ribose), 6.24 (d, J = 2.1 Hz, 1′-CH of ribose), 6.58 (d, J = 2.1 Hz, 4′-CH), 6.63 (d, J = 8.4 Hz, 2′-CH), 6.82 (s_br, NH₂), 7.15 (m, 1′-CH), 8.21 and 8.24 (2s, 2 × CH of adenine); ¹³C NMR (100 MHz, acetone-d₆) − 4.4 (2x), 13.5, 18.5, 20.4, 25.37, 25.41, 25.9 (3x), 26.2 (26.3), 27.1, 27.3, 28.0, 29.1 to 30.2 (5C under solvent peaks), 30.4, 31.4 (31.6), 33.0 (33.1), 34.2, 38.8, 39.06 (39.10), 39.2 (40.5), 39.5, 44.2 (44.79), 44.6 (44.76), 49.4 (49.6), 61.9 (62.8), 64.3, 82.6, 84.7, 85.4, 86.1 (86.7), 91.0, 98.1 (99.4), 114.5, 117.9, 120.5, 120.7, 126.9, 133.9 (134.0), 138.4, 140.5, 150.0, 153.6, 153.9, 157.1, 173.2; LRMS calculated for C₅₁H₇₈N₅O₈Si [M + H]⁺916.6, found 916.4 m/z.

Synthesis of 5′-O-{9-[3′,17′β-dihydroxy-estra-1′,3′,5′(10′)-trien-16′β-yl]-nonanoyl} adenosine (1)

Gaseous hydrogen chloride was bubbled for 3 h at room temperature in a solution of 9 (100 mg, 0.109 mmol) in anhydrous DCM (21 mL). The solvent was removed under vacuum and the residue was preadsorbed on silica gel and purified by flash chromatography (DCM/MeOH, 94:6 to 97:3) to provide EM-1745 (1) (20 mg, 27% yield) as a white solid. The ¹H and ¹³C NMR data agreed with those reported previously in the literature [41].

Synthesis of 5′-O-{9-[3′,17′β-dihydroxy-estra-1′,3′,5′(10′)-trien-16′β-yl]-nonanoyl}-2′,3′-O-isopropylidene adenosine (10)

Deprotection of C17′β-OTHP. Compound 9 (132 mg, 0.144 mmol) was dissolved in a minimum of dry DCM (∼1 mL) under an argon atmosphere. Dry MeOH (3 mL) and pyridinium p-toluenesulfonate (PPTS) (290 mg, 1.15 mmol) were added, and the reaction mixture was refluxed for 5 h. A saturated aqueous solution of NaHCO₃ was added to the cooled mixture to quench the reaction and the crude product was extracted with DCM. The organic phase was washed with an aqueous 1 M CuSO₄ solution, then brine, dried over MgSO₄, and evaporated to dryness. Deprotection of C3′-OTBDMS. To a solution of this crude product (123 mg) in dry THF (3 mL) under an argon atmosphere at 0°C was added TBAF (1 M in THF) (290 μL, 0.290 mmol). The reaction mixture was stirred for 10 min at 0°C and then quenched by addition of a saturated solution of NaHCO₃. The crude product was extracted with EtOAc and the organic phase was washed with brine, dried over MgSO₄, and evaporated to dryness. Purification by flash chromatography (hexanes/EtOAc, 3:7 to 2:8, then DCM/MeOH, 95:5) afforded 10 (75 mg, 72% yield for two steps) as a light yellow solid. IR (film) 3340 and 3188 (OH and NH₂), 1738 (C = O, ester), 1644 (C = N); ¹H NMR (400 MHz, acetone-d₆) 0.80 (s, 18′-CH₃), 0.95 to 2.35 (m, 26H, CH and CH₂ of steroid skeleton and alkyl chain), 1.39 and 1.59 (2s, 2 × CH₃ of isopropylidene), 2.26 (m, CH₂COO), 2.78 (m, 6′-CH₂), 3.72 (d, J = 9.7 Hz, 17′α-CH), 4.25 and 4.32 (2m, 5′-CH₂ of ribose), 4.44 (m, 4′-CH of ribose), 5.16 (m, 3′-CH of ribose), 5.59 (m, 2′-CH of ribose), 6.24 (d, J = 1.8 Hz, 1′-CH of ribose), 6.54 (d, J = 2.0 Hz, 4′-CH), 6.60 (dd, J₁ = 8.4 Hz, J₂ = 2.4 Hz, 2′-CH), 6.69 (s_br, NH₂), 7.11 (d, J = 8.5 Hz, 1′-CH), 7.98 (s, 3′-OH), 8.20 and 8.23 (2s, 2 × CH of adenine); ¹³C NMR (100 MHz, acetone-d₆) 12.9, 25.3, 25.4, 27.1, 27.3, 28.2, 29.1, 29.3 to 30.2 (3C under solvent peaks), 30.22, 32.48, 32.52, 33.2, 34.1, 38.6, 39.3, 41.2, 44.7, 44.8, 49.4, 64.3, 82.46, 82.49, 84.7, 85.3, 91.0, 113.4, 114.4, 115.7, 120.9, 126.8, 132.0, 138.3, 140.6, 150.0, 153.6, 155.7, 157.0, 173.2; LRMS calculated for C₄₀H₅₆N₅O₇ [M + H]⁺718.5, found 718.9 m/z.

Synthesis of 5′-O-{9-[3′-hydroxy-17′-oxo-estra-1′,3′,5′(10′)-trien-16′β-yl]-nonanoyl}-2′,3′-O-isopropylidene adenosine (11)

To a solution of 10 (91 mg, 0.127 mmol) in acetone at 0°C was added Jones' reagent (2.7 M CrO₃ in H₂SO₄/H₂O, 2:8) (56 μL, 0.152 mmol). The mixture was stirred for 8 min. Then, the reaction was quenched by addition of isopropanol and a saturated aqueous solution of NaHCO₃. The crude product was extracted with EtOAc and the organic phase was washed with brine, dried over MgSO₄, and evaporated under reduced pressure. Purification by flash chromatography (DCM/MeOH, 97:3 to 95:5) provided the starting material 10 (20 mg, 22%) and 11 (47 mg, 52% yield) as a white solid. IR (film) 3343 and 3176 (OH and NH₂), 1732 (C = O, ester and ketone), 1644 (C = N); ¹H NMR (400 MHz, acetone-d₆) 0.86 (s, 18′-CH₃), 1.20 to 2.40 (m, 26H, CH and CH₂ of steroid skeleton and alkyl chain), 1.39 and 1.59 (2s, 2 × CH₃ of isopropylidene), 2.26 (m, CH₂COO), 2.82 (m, 6′-CH₂), 4.24 and 4.32 (2m, 5′-CH₂ of ribose), 4.44 (m, 4′-CH of ribose), 5.16 (m, 3′-CH of ribose), 5.59 (m, 2′-CH of ribose), 6.24 (d, J = 2.0 Hz, 1′-CH of ribose), 6.56 (d, J = 2.4 Hz, 4′-CH), 6.62 (dd, J₁ = 8.4 Hz, J₂ = 2.6 Hz, 2′-CH), 6.67 (s_br, NH₂), 7.12 (d, J = 8.4 Hz, 1′-CH), 8.02 (s, 3′-OH), 8.19 and 8.23 (2s, 2 × CH of adenine); ¹³C NMR (75 MHz, acetone-d₆) 14.2, 25.3, 26.5, 27.26, 27.32, 28.5, 28.9 to 30.4 (7C under solvent peaks), 32.7, 32.9, 34.1, 38.7, 44.8, 48.7, 49.4, 49.6, 64.3, 82.5, 84.7, 85.3, 90.9, 113.4, 114.4, 115.8, 120.6, 126.8, 131.4, 138.2, 140.6, 150.0, 153.6, 155.8, 157.0, 173.2, 221.5; LRMS calculated for C₄₀H₅₄N₅O₇ [M + H]⁺716.4, found 716.7 m/z.

Synthesis of 5′-O-{9-[3′-hydroxy-17′-oxo-estra-1′,3′,5′(10′)-trien-16′β-yl]-nonanoyl} adenosine (2)

Compound 11 (34 mg, 0.048 mmol) in dry THF (500 μL) was treated with a TFA/H₂O (9/1, v/v) solution (2.4 mL). The reaction mixture was stirred for 30 min at room temperature and quenched by addition of a saturated aqueous NaHCO₃ solution. The crude product was extracted with EtOAc and the organic phase was washed with brine, dried over MgSO₄, and evaporated under reduced pressure. The crude residue was purified by flash chromatography (DCM/MeOH, 95:5) to afford 2 (18 mg, 56% yield) as a white solid. IR (KBr) 3338 and 3213 (OH and NH₂), 1737 (C = O, ester), 1725 (C = O, ketone), 1655 (C = N); ¹H NMR (400 MHz, DMSO-d₆) 0.77 (s, 18′-CH₃), 1.15 to 2.35 (m, 26H, CH and CH₂ of steroid skeleton and alkyl chain), 2.28 (t, J = 7.1 Hz, CH₂COO), 2.74 (m, 6′-CH₂), 4.07 (m, 4′-CH of ribose), 4.19 and 4.32 (2m, 5′-CH₂ of ribose), 4.25 (m, 3′-CH of ribose), 4.66 (m, 2′-CH of ribose), 5.39 (d, J = 5.5 Hz, OH of ribose), 5.60 (d, J = 4.8 Hz, OH of ribose), 5.90 (d, J = 4.8 Hz, 1′-CH of ribose), 6.45 (d, J = 2.0 Hz, 4′-CH), 6.51 (dd, J₁ = 8.3 Hz, J₂ = 2.2 Hz, 2′-CH), 7.05 (d, J = 8.5 Hz, 1′-CH), 7.33 (s, NH₂), 8.14 and 8.31 (2s, 2 × CH of adenine), 9.04 (s, 3′-OH); ¹³C NMR (75 MHz, DMSO-d₆) 13.7, 24.4, 25.5, 26.3, 27.4, 28.0, 28.4, 28.6, 28.7, 28.8, 29.1, 31.6, 33.3, 37.6, 38.7 to 40.3 (1C under solvent peaks), 43.6, 47.8, 48.0, 48.5, 63.7, 70.3, 72.9, 81.4, 87.7, 112.8, 115.0, 119.1, 126.0, 130.0, 137.1, 139.7, 149.3, 152.7, 155.0, 156.1, 172.8, 221.7; LRMS calculated for C₃₇H₅₀N₅O₇ [M + H]⁺676.4, found 676.7 m/z.

Enzymatic assay: Inhibition of 17β-HSD1 in homogenated cells

This enzymatic assay on 17β-HSD1 was performed as previously described [49]. Briefly, HEK-293 cells transfected with 17β-HSD1 cDNA fragment were sonicated in 50 mM sodium phosphate buffer (pH 7.4), containing 20% glycerol and 1 mM EDTA to obtain cellular fragmentation. The cytosol fraction containing the enzyme was isolated as the supernatant after centrifugation (100 000 × g, 5 min, 4°C). The enzymatic reaction was performed at 37°C for 2 h in 1 mL of a solution which included 980 μL of a stock solution containing 50 mM sodium phosphate buffer (pH 7.4, 20% glycerol and 1 mM EDTA), 0.1 mM NADH and 0.1 μM [¹⁴C]-estrone (54 mCi/mmol, American Radiolabeled Chemicals Inc., St-Louis, MO, USA), 10 μL of the indicated inhibitor dissolved in ethanol and 10 μL of diluted enzymatic source in phosphate buffer. Each inhibitor was assessed in duplicate. Afterward, radiolabeled steroids were extracted twice from the reaction mixture by 1 mL of diethyl ether. The organic phases were pooled and evaporated to dryness with nitrogen. Residues were dissolved in 50 μL of DCM, applied on silica gel 60 F₂₅₄ thin layer chromatography plates (EMD Chemicals Inc., Gibbstown, NJ, USA) and eluted with a mixture of toluene/acetone (4:1). Substrate ([¹⁴C]-E₁) and metabolite ([¹⁴C]-E₂) were identified by comparison with reference steroids and quantified using the Storm 860 system (Molecular Dynamics, Sunnyvale, CA, USA). The percentage of transformation of [¹⁴C]-E₁ into [¹⁴C]-E₂ was calculated as follows: % transformation = 100 × ([¹⁴C]-E₂/([¹⁴C]-E₂+[¹⁴C]-E₁)), and subsequently, % inhibition = 100 × ((% transformation without inhibitor − % transformation with inhibitor)/% transformation without inhibitor).

Results

Chemistry

The C17-ketone analogue of EM-1745, compound 2, was prepared by oxidation of the C17-alcohol of an advanced intermediate in the synthesis of EM-1745. However, the chemical procedure initially developed for the synthesis of 1 was time-consuming and difficult [41]. In fact, the addition of a long alkyl chain in the α position of the C17 ketone gave a mixture of mono (C16α- and C16β-alkylated) and C16-dialkylated E₁ derivatives. Moreover, it was impossible to separate the C16β-alkylated-E₁ derivatives from the mixture using standard purification by flash chromatography. After stereoselective reduction of the C17-ketone, they applied a four-step sequence to isolate the pure β-alkylated-E₂. This sequence was based on selective protection of the C17β-alcohol derivatives with TBDMS. The C17β-alcohol of the 16α-isomer could easily be protected as a TBDMS using classical condition (TBDMSCl, imidazole). However, the C17β-alcohol of the 16β-isomer or the dialkylated derivatives needed more drastic conditions (TBDMSOTf, lutidine). The other weakness of this former procedure was the final deprotection. Even if the three protective groups (two TBDMS and one isopropylidene) can normally be removed in one step under acid conditions, a two-step sequence was necessary to hydrolyse the C17β-OTBDMS. Thus, the protective groups were removed by bubbling hydrogen chloride in dry DCM, except for the C17β-OTBDMS that was taken off by a subsequent treatment with TBAF at 60°C. A new procedure was then developed for preparing EM-1745 (1). The strategy that is presented here is roughly the same one developed to prepare the androstenedione/adenosine hybrid inhibitor of 17β-HSD3 [45]. Briefly, an alkyl side chain was added on 16β-allyl-E₂ by a cross-metathesis methodology. Furthermore, a THP ether, more easily removed under acid conditions than a C17-OTBDMS, was chosen as the C17β-protective group.

The chemical synthesis of 1 and 2 is illustrated in Scheme 1. Starting from 6-bromo-1-hexanol (3), a Br-I exchange using Finkelstein conditions followed by addition of a vinyl cuprate using HMPA and triethylphosphite as additives afforded vinyl alcohol 4 in 75% yield for two steps. This alcohol was next oxidized into aldehyde 5 with TPAP. The low boiling point of 5 explains the low 51% yield for this oxidation that seemed complete on TLC. Diprotected 16β-allyl-E₂ (6) was prepared in five steps from commercially available E₁ as previously reported [48]. A cross-metathesis [50] between steroid 6 and olefin 5 using Grubbs' catalyst afforded 7 in 50% yield. The formation of a side product resulting from the metathesis of two molecules of steroid 6 was also observed. It is noteworthy that these kinds of aldehydes are very unstable and are rapidly oxidized into carboxylic acid. Thus, a freshly prepared olefinic aldehyde 5 is necessary since the Grubbs catalyst is not compatible with carboxylic acids, resulting in a lower cross-metathesis yield. Olefin 7 was next reduced with hydrogen catalyzed by palladium followed by oxidation of the aldehyde under mild conditions (NaClO₂, NaH₂PO₄, 2-methyl-2-butene in tert-butanol) to provide carboxylic acid 8 in 75% yield.

Scheme 1.  A new procedure for preparing EM-1745 (1) and synthesizing the C17-ketone analogue of EM-1745 (2). Reagents, conditions and yields: (a) NaI, acetone, reflux, 16 h; (b) i. CuI, vinylMgBr, THF, − 40°C, 15 min; ii. HMPA, P(OEt)₃, THF, − 40°C to rt, 5 h (75% for two steps); (c) TPAP, NMO, molecular sieves, DCM, rt, 90 min (51%); (d) 5, 2nd generation Grubbs' catalyst, DCM, reflux, 16 h (50%); (e) 10% Pd/C, H₂, EtOAc, rt, 16 h; (f) NaClO₂, NaH₂PO₄, THF, 2-methyl-2-butene, t-BuOH, rt, 30 min (75% for two steps); (g) 2′,3′-isopropylidene adenosine, PyBOP, HOBt, DIPEA, DMF, rt, 16 h (51%); (h) HCl_g, DCM, rt, 3 h (27%); (i) PPTS, DCM, MeOH, reflux, 5 h; (j) TBAF, THF, 0°C, 10 min (72% for two steps); (k) Jones' reagent, acetone, 0°C, 8 min (52%); (l) TFA/H₂O, 9:1, THF, rt, 30 min (56%).

The adenosine and the steroid moieties were next linked together. Commercially available 2′,3′-isopropylidene-adenosine was used to direct the addition on the C5′-OH of adenosine. A classical esterification procedure (EDCI and DMAP in DMF) resulted in a low 11% coupling yield, and thus, mainly starting materials were recovered after the reaction. However, reagents primarily used for coupling amino acids on solid phase synthesis (PyBOP, HOBt and DIPEA in DMF) gave a better 51% esterification yield and provided the E₂/adenosine hybrid compound 9. Then, EM-1745 (1) was obtained in one deprotection step. Unfortunately, the deprotection conditions developed for the synthesis of the hybrid inhibitors of 17β-HSD3 (TFA/H₂O, 9:1) [45], did not give the expected product. Therefore, the three protective groups were removed by bubbling hydrogen chloride in dry DCM, which afforded EM-1745 (1) in a non-optimized 27% yield. Starting from 6, EM-1745 (1) was obtained in five steps and 5% overall yield (19% overall yield before the final trideprotection).

In order to obtain the C17-ketone analogue 2 of EM-1745, the C17-OTHP protective group needs to be removed from 9 without hydrolysis of the isopropylidene group. In addition, one must take care to avoid hydrolysis of the ester bond. All the following reagents that we tried in various conditions: HCl/MeOH, p-TSA/DCM, HCl/dioxane or acid resins DOWEX or Amberlyst, removed both the C17β-OTHP and the C3-OTBDMS protective groups, but unfortunately some hydrolysis of the ester bond was observed. C17-OTHP and C3-OTBDMS were then removed in two steps. First, the THP protective group was removed with pyridinium p-toluenesulfonate (PPTS) in refluxed MeOH. With these conditions, part of the C3-OTBDMS was hydrolysed. Then, the remaining C3-OTBDMS group was removed carefully with TBAF in THF at 0°C to afford steroid 10 in 72% yield for two steps.

The last step was to oxidize the C17-alcohol of 10 into the C17-ketone. However, this step proved to be more difficult than expected. Oxidation with TPAP, Dess-Martin or PCC resulted in the decomposition of 10. Only oxidation with Jones' reagent gave 11 in 52% yield with 22% of the starting material recovered after chromatography. In order to avoid formation of side reactions such as ester hydrolysis and oxidation at C6 steroidal position, the reaction time was limited to 8 minutes. Final deprotection of 11 with TFA/H₂O, 9:1 provided the C17-ketone analogue of EM-1745, compound 2, in 56% yield. From steroid intermediate 6, compound 2 was obtained in eight steps (4% overall yield) and was fully characterized by IR, ¹H NMR, ¹³C NMR and LRMS to validate its structure.

Inhibition of 17β-HSD1

The enzymatic assay was performed with homogenated human embryonic kidney (HEK)-293 cells transfected with a vector encoding for 17β-HSD1 as previously described [49]. This test was carried out at 37°C for 2 h using NADH as cofactor to promote the reductive activity of the enzyme. Compounds 1 and 2 were evaluated for their ability to inhibit the transformation of [¹⁴C]-E₁ into [¹⁴C]-E₂. For this study, the inhibitors were tested in duplicate at four concentrations: 0.001, 0.01, 0.1 and 1 μM (Figure 3 and Table II). From these data, we estimated the IC₅₀ values of 4 and 12 nM for 1 and 2, respectively.

Figure 3.  Inhibition of the transformation of [¹⁴C]-E₁ into [¹⁴C]-E₂ by EM-1745 (1) and its C17-ketone analogue 2 in homogenated HEK-293 cells everexpressing 17β-HSD1. See experimental section for more details. Data with the same symbol (* or **) are significantly different (P < 0.01).

Table II.  Inhibition of 17β-HSD1 by different kinds of inhibitors.

			Inhibition of 17β-HSD1
Entry	Compounds	C17 (X)	% at 0.01 mM	% at 0.1 mM	% at 1 mM	IC50 (nM)
1^*	1 (EM-1745)	β-OH	73	95	98	4
	2	O	44	84	96	12
2^* [51]	12	β-OH	–	15	46	–
	13	O	–	12	23	–
3^† [53]	14	β-OH	–	77	94	44
	15	O	–	51	88	171
4^‡ [54]	16	β-OH	–	–	–	460
	17	O	–	–	–	310
5^‡ [54]	Unlabeled-E2	β-OH	–	–	–	>1000
	Unlabeled-E1	O	–	–	–	300

^* Inhibition of the transformation of [¹⁴C]-E₁ (100 nM) into [¹⁴C]-E₂ in homogenated HEK-293 cells everexpressing 17β-HSD1. See experimental section for more details

^† Inhibition of the transformation of [¹⁴C]-E₁ (60 nM) into [¹⁴C]-E₂ by 17β-HSD1 in breast cancer T-47D cells

^‡ Inhibition of the transformation of [³H]-E₁ (6 nM) into [³H]-E₂ in partially purified human placenta cytosolic 17β-HSD1.

Discussion

Several families of inhibitors have been developed for 17β-HSD1, such as competitive reversible inhibitors, competitive irreversible inhibitors and bisubstrate inhibitors 2829303132. EM-1745 (1), the bisubstrate inhibitor we previously synthesized [41,42], contains an E₂ moiety. The K_m values reported in the literature for E₁ (0.36 μM) and E₂ (1.7 μM) with 17β-HSD1 indicated that E₁ is a much better substrate of 17β-HSD1 [44]. Furthermore, competitive inhibitors are known to interact with the substrate-binding site of the enzyme. We then hypothesized that modifying compound 1 by replacing E₂ with E₁ (a C17-ketone instead of a C17-alcohol) should give a better inhibition of 17β-HSD1. A C17-ketone derivative of EM-1745, compound 2, was prepared in order to confirm our working hypothesis and we evaluated the ability of 1 and 2 to inhibit the transformation of [¹⁴C]-E₁ into [¹⁴C]-E₂ on homogenated HEK-293 cells over-expressing 17β-HSD1. However, the C17-ketone (E₁) derivative 2 gave a lower inhibition than the C17-alcohol (E₂) derivative 1. This result was surprising. In fact, X-ray analysis of the EM-1745/17β-HSD1 complex clearly showed that the substrate moiety of EM-1745 interacts with the same site as E₁ does and kinetic studies demonstrated clearly that it is a reversible competitive inhibitor [42].

After obtaining this unexpected result, we looked the inhibitory potency of some of our 17β-HSD1 steroidal inhibitors available under the C17β-hydroxy and the C17-ketone forms (Figure 4, Table II). For 6β-alkylamide-E₂ and -E₁ derivatives (compounds 12 and 13) [51], which are analogue to EM-678 inhibitor [52], a better inhibition of 17β-HSD1 was obtained with the C17β-hydroxy analogue, as for the bisubstrate inhibitors 1 and 2. In fact, the C6β-alkylamide-E₂ (12) gave a better inhibition with 46% at 1 μM whereas the C6β-alkylamide-E₁ (13) gave almost no inhibition with 23% at 1 μM. The same tendency was also observed for the 16β-m-carbamoylbenzyl derivative of E₂, compounds 14 and 15, recently reported as inhibitors of 17β-HSD1 [53]. At a concentration of 0.1 μM, the C17β-hydroxy compound 14 inhibited 77% the transformation of E₁ into E₂, but only 51% was obtained for the C17-ketone analogue (IC₅₀ = 44 and 171 nM for 14 and 15, respectively).

Figure 4.  A selection of our previously reported inhibitors of 17β-HSD1.

For the competitive irreversible inhibitors 16 and 17 a similar inhibition was reported for the 16α-(bromopropyl)-E₂ (16) and the 16α-(bromopropyl)-E₁ (17), with a slightly better inhibition (less than 1.5 fold) for the C17-ketone analogue (17) [54], whereas unlabeled E₁ gave a better inhibition than unlabeled E₂ with IC₅₀ values of 300 nM and >1000 nM, respectively.

In summary, a more efficient procedure to prepare EM-1745 (1) was developed. Because the C17-ketosteroid E₁ is a better substrate for the reductive activity of 17β-HSD1 than the corresponding C17-hydroxy E₂, the C17-ketone analogue of EM-1745 (compound 2) was also synthesized using this new procedure. Unexpectedly, the C17-ketone derivative 2 gave a lower inhibition of the enzyme than C17-hydroxy derivative 1. In fact, the presence of a C17-ketone increases the binding affinity of a C18 steroid for 17β-HSD1, but it can also modify the orientation of the 16β-side chain and thus the positioning of the adenosine group interacting with the cofactor binding site of the enzyme. In that case, the potency or number of these key interactions could be reduced with as consequence a diminution of the inhibitory activity of 2. A lower inhibitory activity of ketone derivatives also holds true for two series of reversible inhibitors (12, 13 and 14, 15), and similar inhibition was obtained for two competitive irreversible inhibitors (16 and 17). Nonetheless, the inhibitory potency of a larger number of inhibitors should be evaluated in a same enzymatic assay in order to confirm this observation (the C17-alcohol analogues are better inhibitors than their corresponding C17-ketone analogues). That knowledge could be applied to the development of new and more potent competitive inhibitors of 17β-HSD1.

Acknowledgements

We thank the Canadian Institutes of Health Research for financial support (DP) and a scholarship (MB). We are also grateful to Dr. Van Luu-The for providing HEK-293 cells everexpressing 17β-HSD1 and Fatima Kamal for chemical synthesis of intermediate compound 6. Careful reading of the manuscript by Sylvie Méthot is also greatly appreciated.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.