Steroids 76 (2011) 1377–1382

Contents lists available at ScienceDirect

Steroids

journal homepage: www.elsevier .com/locate /s teroids

Synthesis of steroid–ferrocene conjugates of steroidal 17-carboxamidesvia a palladium-catalyzed aminocarbonylation – Copper-catalyzed azide–alkynecycloaddition reaction sequence

Eszter Szánti-Pintér a, János Balogh a, Zsolt Csók b, László Kollár b, Ágnes Gömöry c, Rita Skoda-Földes a,⇑a University of Veszprém, Institute of Chemistry, Department of Organic Chemistry, Egyetem u. 10. (P.O. Box 158) H-8200 Veszprém, Hungaryb University of Pécs, Department of Inorganic Chemistry, Ifjúság u. 6. (P.O. Box 266) H-7624 Pécs, Hungaryc Chemical Research Center of the Hungarian Academy of Sciences, Institute of Structural Chemistry, Department of Mass Spectrometry, Pusztaszeri u. 59-67. (P.O. Box 17) H-1525Budapest, Hungary

a r t i c l e i n f o a b s t r a c t

Article history:Received 21 April 2011Received in revised form 30 May 2011Accepted 12 July 2011Available online 20 July 2011

Keywords:AminocarbonylationPalladiumFerroceneCopperCycloaddition

0039-128X/$ - see front matter � 2011 Elsevier Inc. Adoi:10.1016/j.steroids.2011.07.006

⇑ Corresponding author. Address: University of VeszDepartment of Organic Chemistry, Egyetem u. 8. (P.OHungary. Tel.: 36-88-624719; fax: 36-88-624469.

E-mail address: skodane@almos.uni-pannon.hu (R

Steroids with the 17-iodo-16-ene functionality were converted to ferrocene labeled steroidal 17-carbox-amides via a two step reaction sequence. The first step involved the palladium-catalyzed aminocarbony-lation of the alkenyl iodides with prop-2-yn-1-amine as the nucleophile in the presence of the Pd(OAc)2/PPh3 catalyst system. In the second step, the product N-(prop-2-ynyl)-carboxamides underwent a facileazide–alkyne cycloaddition with ferrocenyl azides in the presence of CuSO4/sodium ascorbate to producethe steroid–ferrocene conjugates. The new compounds were obtained in good yield and were character-ized by 1H and 13C NMR, IR, MS and elemental analysis.

� 2011 Elsevier Inc. All rights reserved.

1. Introduction

Several representatives of steroidal 17-carboxamides are effi-cient inhibitors [1] of 5a-reductase, the enzyme converting testos-terone to dihydrotestosterone (DHT). Elevated levels of DHT maycause benign prostatic hyperplasia, that can efficiently be treatedwith 4-aza-steroid drugs such as finasteride [2] or dutasteride[3]. Other steroidal 17-carboxamides, e.g. epristeride [4] were alsofound to exhibit 5a-reductase inhibitory effect. Palladium-cata-lyzed aminocarbonylation was shown to be a very efficient methodto attach the carboxamide moiety to C-17 of androstanes [5–7] andestra-1,3,5(10)-trienes [8–10] as well as to other positions of thesteroidal skeleton [11,12]. The corresponding enol triflates [5,8–10] or alkenyl iodides [6,7,11,12] were used as substrates.

According to structure–activity relationship studies, the inhibi-tory effect of steroidal 17-carboxamides is enhanced by the pres-ence of lipophilic amido or keto groups in position C-17 [1] andthe space available to groups about this position is not restricted[13]. As in some cases, the incorporation of the lipophilic ferrocene

ll rights reserved.

prém, Institute of Chemistry,. Box 158) H-8200 Veszprém,

. Skoda-Földes).

moiety into molecules with pharmacological activity was found toenhance the biological effect of the parent compound [14], wedecided to develop a new methodology to introduce the ferrocenylgroup into the N-alkyl side chain of steroid 17-carboxamides. Asanother advantage, the presence of the ferrocene moiety may allowthe electrochemical detection of the steroid with a detection limitsuitable for quantitative pharmacological analysis as it was shownfor ferrocene–ethynylestradiol [15] or ferrocene–testosterone con-jugates [16].

In this work, steroidal alkenyl iodides, that had been proved tobe good starting materials for palladium-catalyzed aminocarbony-lation [6,7] and can easily be obtained from readily available ste-roidal ketones [17], were converted into 17-carboxamides with aC–C triple bond in the side chain in the first step.

In the second step, the ferrocene moiety was attached to thesteroidal backbone by a copper-catalyzed azide–alkyne cycloaddi-tion reaction using a ferrocenyl azide as the reaction partner.Although this method has effectively been used for the attachmentof ferrocene to various biomolecules, e.g. amino acids [18] or carbo-hydrates [19], to the best of our knowledge, there is only oneexample for a similar functionalization of a steroid in the synthesisof a tetrameric estrone-based macrocycle. In this compound, twosteroidal dimers are connected with spacers containing ferrocenylgroups to form a tetrameric structure [20].

1378 E. Szánti-Pintér et al. / Steroids 76 (2011) 1377–1382

2. Experimental

1H and 13C NMR spectra were recorded in CDCl3 on a VarianInova 400 spectrometer at 400.13 and 100.62 MHz, respectively.Chemical shifts d are reported in ppm relative to CHCl3 (7.26 and77.00 ppm for 1H and 13C, respectively). GLC analyses were carriedout with a HP-5890/II gas chromatograph using a 15 m HP-5 col-umn. Infrared (IR) spectra were recorded in KBr pellets using anAvatar330 FT-IR instrument. Elemental analyses were measuredon a 1108 Carlo Erba apparatus. GC–MS of compounds 1b–5b wererecorded on a HP-5971A MSD connected to a HP-5890/II gas chro-matograph. MS spectra of 2c, 4c, 5c were obtained on an Autoflex IITOF/TOF (Bruker Daltonics, Bremen, Germany) spectrometer in2,5-dihydroxybenzoic acid matrix, MS spectra of 1c, 3c, 2d weremeasured on a Waters Micromass Quattro micro API spectrometerusing electrospray ionization.

Ferrocenyl azides 9 [21] and 10 [22] were synthesized byknown procedures in 75% and 35% yields, respectively. The physi-cal properties and the spectra of the azides corresponded well toliterature data (9 [23], 10 [24]).

2.1. General procedure for the carbonylation of steroidal alkenyliodides 1a–5a

In a typical procedure, the steroidal alkenyl iodide (0.6 mmol),Cs2CO3 (1.2 mmol, 391 mg), Pd(OAc)2 (0.03 mmol, 6.7 mg) andPPh3 (0.06 mmol, 15.7 mg) were placed under carbon monoxidein a Schlenk-tube equipped with a magnetic stirrer, a septum inlet,and a reflux condenser with a balloon on the top. 1,4-Dioxane(10 ml) and prop-2-yn-1-amine (3 mmol, 192 ll) was addedthrough the septum inlet and the reaction mixture was heated at80 �C for 8 h. The reaction was monitored by TLC. Cs2CO3 was fil-tered and the solvent was removed in vacuo. The products wereisolated by column chromatography (silica gel, eluent: n-hexane/ethyl acetate = 1/1 (1b, 3b), chloroform/methanol = 25/1 (2b), andchloroform/methanol = 20/1 (4b, 5b)).

2.2. General procedure for the azide–alkyne cycloaddition ofcarboxamides 1b–5b

A mixture of the steroidal carboxamide (0.1 mmol), 1-azido-ethylferrocene (racemic mixture) 9 (0.1 mmol, 25.5 mg) or ethyl2-azido-3-ferrocenyl-propenoate 10 ((0.1 mmol, 32.5 mg), Cu-SO4�5H2O (0.015 mmol, 3.8 mg), sodium ascorbate (0.038 mmol,7.5 mg), CH2Cl2 (1 ml) and water (1 ml) was stirred at room tem-perature for 5 days. The product was extracted with CH2Cl2

(3 � 2 ml). The combined organic phase was dried over Na2SO4

and concentrated. The product was purified by column chromatog-raphy (silica, eluent: toluene/methanol = 6/1 (1c–3c, 5c, 2d) andchloroform/methanol = 20/1 (4c)).

2.3. Characterization of the products

2.3.1. 17-(N-(prop-2-ynyl)-carbamoyl)-5a-androst-16-ene (1b)1H NMR(d,CDCl3): 6.28–6.30 (m, 1H, 16-H); 5.68(t, 5.2 Hz, 1H,

NH); 4.03 (dd, 5.2 Hz, 2.5 Hz, 2H, NH-CH2); 2.15(t, 2.5 Hz, 1H,CCH); 0.66–2.28 (m, 22H, ring protons); 0.91(s, 3H, 18-H3);0.75(s, 3H, 19-H3). 13C NMR(d,CDCl3): 165.4; 150.0; 136.7; 79.8;71.5; 56.8; 55.1; 47.2; 46.6; 38.4; 36.4; 34.9; 33.7; 31.9; 31.7;29.0; 28.9 (2C); 26.8; 22.1; 20.6; 16.5; 12.2. MS (m/z/rel. int.):339 (M+)/24; 324/26; 281/12; 257/72; 55/100. IR (KBr, m (cm�1)):3310, 1644, 1591. Analysis calculated for C23H33NO (339.52): C,81.37; H, 9.80; N, 4.13; Found: C, 81.55; H,10.05; N, 3.98. Rf (hex-ane/EtOAc = 1/1): 0.82. Yield: 79%.

2.3.2. 17-(N-(prop-2-ynyl)-carbamoyl)-6b-hydroxy-3a,5a-cycloandrost-16-ene (2b)

1H NMR (d,CDCl3): 6.32–6.34(m, 1H, 16-H); 5.80(t, 5.2 Hz, 1H,NH); 4.07(dd, 5.2 Hz, 2.5 Hz, 2H, NH-CH2); 3.26(m, 1H, 6-H); 2.20(t, 2.5 Hz, 1H, CCH); 0.80–2.23 (m, 17H, ring protons); 1.07(s, 3H,18-H3); 1.02(s, 3H, 19-H3); 0.52(m, 1H, 4-Ha); 0.28(m, 1H, 4-Hb).13C NMR(d,CDCl3): 165.3; 150.0; 136.3; 79.7; 73.5; 71.5; 56.7;48.1; 46.8; 43.1; 39.1; 36.9; 35.0; 32.9; 31.7; 28.9; 28.2; 24.9;24.2; 22.2; 20.1; 16.6; 11.5. MS (m/z/rel. int.): 353 (M+)/2, 335/9,320/6, 186/57, 105/59, 91/100, 79/53. IR (KBr, m (cm�1)): 3307,1636, 1589. Analysis calculated for C23H31NO2 (353.50): C, 78.15;H, 8.84; N, 3.96; Found: C, 77.97; H, 8.91; N, 4.08. Rf (CHCl3/MeOH = 25/1): 0.77. Yield: 64%.

2.3.3. 17-(N-(prop-2-ynyl)-carbamoyl)-3-methoxy-estra-1,3,5(10),16-tetraene (3b)

1H NMR(d,CDCl3): 7.18(d, 8.7 Hz, 1H, 1-H); 6.70(dd, 8.7 Hz,2.7 Hz, 1H, 2-H); 6.62(d, 2.7 Hz, 1H, 4-H); 6.35–6.37(m, 1H, 16-H); 5.77(t, 5.2 Hz, 1H, NH); 4.10(dd, 5.2 Hz, 2.5 Hz, 2H, NH-CH2);3.76(s, 3H, OCH3); 2.23(t, 2.5 Hz, 1H, CCH); 0.83–3.01 (m, 13H, ringprotons); 1.00(s, 3H, 18-H3). 13C NMR(d,CDCl3): 165.3; 157.5;150.1; 137.7; 136.2; 132.6; 126.0; 113.9; 111.4; 79.7; 71.6; 55.9;55.2; 46.9; 44.2; 37.0; 34.8; 31.5; 29.6; 28.9; 27.7; 26.4; 16.5. MS(m/z/rel. int.): 349 (M+)/100, 334/11, 173/36, 160/33. IR (KBr, m(cm�1)): 3301, 1641, 1592. Analysis calculated for C23H27NO2

(349.47): C, 79.05; H, 7.79; N, 4.01; Found: C, 79.31; H, 7.64; N,4.17. Rf (hexane/EtOAc = 1/1): 0.69. Yield: 59%.

2.3.4. 4-Methyl-17-(N-(prop-2-ynyl)-carbamoyl)-4-aza-5a-androst-16-en-3-one (4b)

1H NMR(d,CDCl3): 6.35–6.37(m, 1H, 16-H); 5.86(t, 5.2 Hz, 1H,NH); 4.08(dd, 5.2 Hz, 2.5 Hz, 2H, NH-CH2); 3.02–3.07(m, 1H, 5-H); 2.93(s, 3H, N-Me); 2.21(t, 2.5 Hz, 1H, CCH); 0.81–2.48(m,17H, ring protons); 0.92(s, 3H, 18-H3); 0.88(s, 3H, 19-H3). 13CNMR(d,CDCl3): 170.7; 165.2; 149.9; 135.9; 79.9; 71.5; 65.7; 56.0;52.3; 46.8; 36.6; 34.6; 32.8; 32.7; 31.5; 30.0; 29.1; 29.0; 28.9;25.3; 20.9; 16.4; 12.3. MS (m/z/rel. int.): 368 (M+)/20, 353/15,281/37, 133/60, 96/93, 57/100. IR (KBr, m (cm�1)): 3313, 1658,1633, 1605, 1593. Analysis calculated for C23H32N2O2 (368.52): C,74.96; H, 8.75; N, 7.60; Found: C, 74.79; H, 8.91; N, 7.77. Rf

(CHCl3/MeOH = 20/1): 0.54. Yield: 70%.

2.3.5. 17-(N-(prop-2-ynyl)-carbamoyl)-4-aza-5a-androst-16-en-3-one (5b)

1H NMR(d,CDCl3): 6.35–6.37(m, 1H, 16-H); 5.91(t, 5.2 Hz, 1H,CONH); 5.8(brs, 1H, 4-NH); 4.08(dd, 5.2 Hz; 2.5 Hz, 2H, NH-CH2);3.02–3.07(m, 1H, 5H); 2.21(t, 2.5 Hz, 1H, CCH); 0.81–2.48 (m,17H, ring protons); 0.92(s, 3H, 18-H3); 0.88(s, 3H, 19-H3). 13CNMR(d,CDCl3): 172.4; 165.2; 149.9; 135.8; 79.7; 71.6; 60.8; 56.0;51.6; 46.9; 35.9; 34.4; 33.4; 33.1; 31.6; 29.3; 28.9; 28.4; 27.3;20.9; 16.4; 11.3. MS (m/z/rel. int.): 353(M+)/16, 336/16, 272/42,105/53, 91/74, 55/100. IR (KBr, m (cm�1)): 3358, 3296, 1657,1588. Analysis calculated for C22H30N2O2 (354.49): C, 74.54; H,8.53; N, 7.90; Found: C, 74.43; H,8.67 ; N, 7.74. Rf (CHCl3/MeOH = 20/1): 0.40. Yield: 56%.

2.3.6. 17-(N-(1-(1-ferrocenyl-ethyl)-1,2,3-triazol-4-yl)-methyl-carbamoyl)-5a-androst-16-ene (1c)

1H NMR(d,CDCl3): 7.34(s, 1H, CH(triazole)); 6.21–6.39(m, 2H,NH, 16-H); 5.58–5.69(m, 1H, CH-CH3); 4.40–4.51(m, 2H, NH-CH2); 4.24–4.28(m, 1H, substituted Cp); 4.21–4.15(m, 3H, substi-tuted Cp); 4.13(s, 5H, unsubstituted Cp) 1.85(d, 6.5 Hz, 3H,CH-CH3); 0.63–2.18 (m, 22H, ring protons); 0.91(s, 3H, 18-H3);0.78(s, 3H, 19-H3). 13C NMR(d,CDCl3): 165.8; 150.1; 144.2; 136.3;119.9; 87.5; 69.3; 68.5; 68.1; 66.5; 56.6; 56.5; 55.0; 47.1; 46.5;38.3; 36.3; 34.8; 34.6; 33.6; 31.8; 31.5; 28.9; 28.8; 26.7; 22.0;

E. Szánti-Pintér et al. / Steroids 76 (2011) 1377–1382 1379

21.3; 20.5; 16.4; 12.0. MS (m/z/rel. int.): 1211 (2M + Na)+/17, 1189(2M + H)+/3, 595 (M + H)+/62, 213/100. IR (KBr, m (cm�1)): 3420,1648, 1637. Analysis calculated for C35H46FeN4O (594.62): C,70.70; H, 7.80; N, 9.42; Found: C, 70.82; H, 7.69; N, 9.57. Rf (tolu-ene/MeOH = 6/1): 0.56. Yield: 67%.

2.3.7. 17-(N-(1-(1-ferrocenyl-ethyl)-1,2,3-triazol-4-yl)-methyl-carbamoyl)-6b-hydroxy-3a,5a-cycloandrost-16-ene (2c)

1H NMR(d,CDCl3): 7.44(s, 1H, CH(triazole)); 6.33–6.38(m, 2H,NH, 16-H); 5.65–5.75(m, 1H, CH-CH3); 4.47–4.61(m, 2H, NH-CH2); 4.22–4.26(m, 1H, substituted Cp); 4.18–4.20(m, 1H,substituted Cp); 4.16–4.13(m, 2H, substituted Cp); 4.13(s, 5H,unsubstituted Cp); 3.25–3.28(m, 1H, 6-H); 1.89(d, 6.5 Hz, 3H, CH-CH3); 0.79–2.25(m, 17H, ring protons); 1.07(s, 3H, 18-H3); 1.00(s,3H, 19-H3); 0.50–0.54(m, 1H, 4-Ha); 0.26–0.30(m, 1H, 4-Hb). 13CNMR(d,CDCl3): 165.8; 150.2; 144.8; 136.3; 120.1; 87.7; 73.5;69.2; 68.4; 68.0; 66.3; 56.9; 56.7; 48.0; 46.7; 43.1; 39.0; 36.9;35.1; 34.8; 32.9; 31.7; 28.2; 24.9; 24.2; 22.2; 21.3; 20.1; 16.7;11.6. MS (m/z/rel. int.): 631 (M + Na)+/20, 608 M+/3, 213/100. IR(KBr, m (cm�1)):. 3429, 1651, 1645. Analysis calculated forC35H44FeN4O2 (608.61): C, 69.07; H, 7.29; N, 9.21; Found: C,68.89; H, 7.41; N, 9.13. Rf (toluene/MeOH = 6/1): 0.40. Yield: 78%.

2.3.8. 17-(N-(1-(1-ferrocenyl-ethyl)-1,2,3-triazol-4-yl)-methyl-carbamoyl)-3-methoxy-estra-1,3,5(10),16-tetraene (3c)

1H NMR(d,CDCl3): 7.35(s, 1H, CH(triazole)); 7.17(d, 8.5 Hz, 1H,1-H); 6.69(dd, 8.5 Hz; 2.6 Hz, 1H, 2-H); 6.61(d, 2.6 Hz, 1H, 4-H);6.28–6.38(m, 2H, NH, 16-H); 5.65–5.68(m, 1H, CH-CH3); 4.46–4.52(m, 2H, NH-CH2); 4.28–4.33(m, 1H, substituted Cp); 4.23–4.18(m, 3H, substituted Cp); 4.17(s, 5H, unsubstituted Cp);3.75(s, 3H, OCH3); 0.81–2.89(m, 13H, ring protons); 1.87 (d,6.6 Hz, 3H, CH-CH3); 0.96(s, 3H, 18-H3). 13C NMR(d,CDCl3): 165.7;157.3; 150.2; 144.2; 137.7; 136.0; 132.5; 126.0; 120.0; 113.7;111.3; 87.3; 69.1; 68.3; 68.0; 66.2; 56.6; 55.7; 55.1; 46.7; 44.1;36.9; 34.7; 31.4; 29.6; 29.5; 27.6; 26.3; 21.3; 16.4. MS (m/z/rel.int.): 1231 (2 M + Na)+/9, 1209 (2 M + H)+/22, 605 (M + H)+/95,213/100. IR (KBr, m (cm�1)): 3430, 1646, 1634. Analysis calculatedfor C35H40FeN4O2 (604.57): C, 69.53; H, 6.67; N, 9.27; Found: C,69.61; H, 6.75; N, 9.10. Rf (toluene/MeOH = 6/1): 0.54. Yield: 63%.

2.3.9. 17-(N-(1-(1-ferrocenyl-ethyl)-1,2,3-triazol-4-yl)-methyl-carbamoyl)-4-methyl-4-aza-5a-androst-16-en-3-one (4c)

1H NMR(d,CDCl3): 7.38(s, 1H, CH(triazole)); 6.41(brs, 1H, NH);6.29–6.33(m, 1H, 16-H); 5.63–5.73(m, 1H, CH-CH3); 4.42–4.63(m,2H, NH-CH2); 4.26–4.56(m, 1H, substituted Cp); 4.34–4.17(m, 3H,substituted Cp); 4.16(s, 5H, unsubstituted Cp); 3.06–3.15(m, 1H,5-H); 2.95(s, 3H, N-CH3); 1.87(d, 6.6 Hz, 3H, CH-CH3); 0.81–2.58(m, 17H, ring protons); 0.96(s, 3H, 18-H3); 0.91(s, 3H, 19-H3).13C NMR(d,CDCl3): 170.8; 165.7; 150.1; 144.2; 135.8; 119.9; 87.5;69.2; 68.4; 68.0; 66.3; 65.7; 56.7; 56.0; 52.3; 46.7; 36.6; 34.7;34.6; 32.8; 32.7; 31.5; 30.0; 29.1; 29.0; 25.3; 21.4; 20.9; 16.5;12.3. MS (m/z/rel. int.): 646 (M + Na)+/10, 623 M+/8, 213/100. IR(KBr, m (cm�1)): 3431, 1655, 1647, 1636. Analysis calculated forC35H45FeN5O2 (623.62): C, 67.41; H, 7.27; N, 11.23; Found: C,67.55; H, 7.39; N, 11.31. Rf (CHCl3/MeOH = 20/1): 0.39. Yield: 88%.

2.3.10. 17-(N-(1-(1-ferrocenyl-ethyl)-1,2,3-triazol-4-yl)-methyl-carbamoyl)-4-aza-5a-androst-16-en-3-one (5c)

1H NMR(d,CDCl3): 7.35(s, 1H, CH(triazol)); 6.47(brs, 1H, NH);6.23–6.36 (m, 1H, 16-H); 5.49–5.64(m, 1H, CH-CH3); 4.42–4.52(m, 2H, NH-CH2); 4.24–4.36(m, 4H, substituted Cp); 4.26(s,5H, unsubsituted Cp); 3.06–3.22(m, 1H, 5-H); 1.87(d, 6.5 Hz, 3H,CH-CH3); 0.75–2.50 (m, 17H, ring protons); 0.96(s, 3H, 18-H3);0.92(s, 3H, 19-H3). 13C NMR(d,CDCl3): 172.7; 165.6; 150.0; 144.1;135.6; 119.9; 87.4; 69.1; 68.4; 68.0; 66.3; 60.6; 56.6; 55.9; 51.6;46.7; 35.9; 34.5; 34.4; 33.4; 33.3; 31.5; 29.3; 29.2; 27.3; 21.3;

20.9; 16.4; 11.2. MS (m/z/rel. int.): 632 (M + Na)+/6, 609M+/5,213/100. IR (KBr, m (cm�1)):3420, 1650, 1646, 1636. Analysis calcu-lated for C34H43FeN5O2 (609.59): C, 66.99; H, 7.11; N, 11.49;Found: C, 67.11; H, 7.20; N, 11.31. Rf (toluene/MeOH = 6/1): 0.30.Yield: 64%.

2.3.11. 17-(N-((1-(1-ethoxycarbonyl-2-ferrocenyl)-ethen-1-yl)-1,2,3-triazol-4-yl)-methyl-carbamoyl)-6b-hydroxy-3a,5a-cycloandrost-16-ene (2d)

1H NMR(d,CDCl3): 7.85(s, 1H, Fc-CH=); 7.54(s, 1H, CH(triazole));6.49(brs, 1H, NH); 6.38(brs, 1H, 16-H); 4.61–4.69(m, 2H, substi-tuted Cp); 4.43–4.53(m, 2H, substituted Cp); 4.32(s, 5H, unsubsti-tuted Cp); 4.22(q, 6.5 Hz, 2H, CH2CH3); 3.75–3.86(m, 2H, NHCH2);3.26–3.32(m, 1H, 6-H); 1.25(t, 6.5 Hz, 3H, OCH2CH3); 0.79–2.25(m, 17H, ring protons); 1.08(s, 3H, 18-H3); 1.03(s, 3H, 19-H3); 0.53(m, 1H, 4-Ha); 0.29(m, 1H, 4-Hb). 13C NMR(d,CDCl3):165.8; 163.1; 150.2; 144.9; 143.8; 136.2; 124.4; 121.0; 73.5;72.8; 72.6; 70.9; 70.8; 70.3; 61.8; 56.7; 48.0; 46.7; 43.1; 39.1;36.9; 35.1; 34.7; 32.9; 31.7; 28.2; 24.9; 24.2; 22.3; 20.1; 16.7;14.2; 11.6. MS (m/z/rel. int.): 1379 (2M + Na)+/3, 701 (M + Na)+/36, 679 (M + H)+/100. IR (KBr, m (cm�1)):. 3422, 1717, 1637, 1511,1263, 1038. Analysis calculated for C38H46FeN4O4 (678.65): C,67.25; H, 6.83; N, 8.26; Found: C, 67.16; H, 6.92; N, 8.37. Rf (tolu-ene/MeOH = 6/1): 0.31. Yield: 77%.

3. Results and discussion

3.1. Synthesis of steroidal 17-carboxamides with a C–C triple bond inthe side chain

In the past few years, several new steroidal carboxamides weresynthesized in our groups via a palladium-catalyzed aminocarb-onylation of the corresponding alkenyl-iodides, using amines[25], hydrazines [7], and amino acids [26] as nucleophilic reagents.In these reactions, the active catalyst is a Pd(0) complex formedin situ from Pd(OAc)2 and triphenylphosphane [27] The main stepsof the catalytic cycle involve oxidative addition of iodoalkene, COinsertion leading to an acyl-palladium complex, nucleophilic at-tack of the amine, formation of the product, and regeneration ofthe catalytically active Pd(0) complex via reductive eliminationof hydrogen iodide. The latter step is facilitated by a base, e.g.Et3N. Usually a large excess of the amine is needed, and the useof polar solvents (DMF or CH3CN) is favorable [28].

In order to introduce a side chain with a C–C triple bond thatcan undergo the cycloaddition in the next step, prop-2-yn-1-aminewas used as nucleophile in the aminocarbonylation reaction. Fivesteroidal alkenyl iodides (Fig. 1, 1a–5a) served as reaction partners.17-Iodo-5a-androst-16-ene (1a) and 17-iodo-6b-hydroxy-3a,5a-cycloandrost-16-ene (2a) can be considered as model compoundsproducing reaction mixtures that can easily be analyzed by GCand GC–MS. Conversion of 3a–5a leads to carboxamides that areclose analogs of the intermediates in the synthesis of estranes[8–10] and 4-azaandrostanes [2,3] with 5a-reductase inhibitoryeffect.

Aminocarbonylation of 1a under the usual reaction conditions,using Et3N as the base and DMF as the solvent, led to carboxamide1b (Scheme 1) in poor yield (Table 1, entry 1). This can be partlyexplained by the lower nucleophilicity of prop-2-yn-1-amine (6)compared to the secondary amines usually used as reaction part-ners. Side reaction leading to 8 has been observed before in car-bonylations with the relatively poor organometallic nucleophileNaBPh4, too [29]. As it was reported previously, dimeric anhydrides(e.g. 8) were formed as primary isolable products of the carbonyl-ation of steroidal alkenyl iodides (e.g. 1a), when the reaction wascarried out in DMF containing water impurities [30]. The

Scheme 1. Palladium catalyzed aminocarbonylation of steroid 1a with amine 6 as anucleophile.

Fig. 1. Steroidal alkenyl iodides 1a–5a, used as substrates.

1380 E. Szánti-Pintér et al. / Steroids 76 (2011) 1377–1382

anhydrides precipitated from DMF and hydrolyzed slowly to pro-duce carboxylic acids upon prolonged heating, when a sufficientamount of water was present. Although the exact mechanism ofthe reaction is not known, it can be assumed that the acyl-palla-dium complex is hydrolyzed in the presence of a small amountof water to the carboxylic acid that reacts with another acyl-palla-dium intermediate to form the anhydride. When water was pres-ent in higher concentration, the formation of the carboxylic acidwas favored to that of the corresponding anhydride.

The other side product, the steroidal primary amide 7 was iden-tified by comparing its MS and 1H NMR spectra with those of anauthentic sample obtained by carbonylation of 1a using ammo-nium carbamate as the nucleophile [31]. Primary amide 7 wasshown to be formed via the N–C cleavage of the N-propargyl moi-ety of 1b under carbonylation conditions. When amide 1b washeated in the presence of Pd(OAc)2, PPh3 and Et3N in toluene incarbon monoxide atmosphere at 80 �C for 4 h, it was convertedinto 7 in 15%.

In order to minimize the amount of water impurity and to hin-der formation of 8, solvents (toluene or 1,4-dioxane) freshly dis-tilled from sodium were used in the further experiments. Thischange in the reaction conditions led to the carboxamide 1b inhigher yield (entries 2–4). An increase in the CO pressure led to aslight increase both in the conversion and in the yield of 1b (entry3), however, the best results were obtained in 1,4-dioxane at atmo-spheric CO pressure and in the presence of Cs2CO3 as the baseinstead of Et3N (entry 4).

Table 1Aminocarbonylation of 17-iodo-5a-androstan-16-ene (1a) with prop-2-yn-1-amine (6)a.

Entry Solvent Base CO pressu

1c DMF Et3N 12 Toluene Et3N 13 Toluene Et3N 64 1,4-dioxane Cs2CO3 1

a Reaction conditions: 1a/6/Pd(OAc)2/PPh3/base = 1/5/0.05/0.1/2, 8 h, 80 �C.b Determined by GC using androst-16-ene as internal standard.c 5a-Androstan-16-ene-17-carboxylic anhydride (8) was also formed.

The same conditions were successfully used for the synthesis ofcarboxamides 2b–5b (Scheme 2). Although the reactivity of ste-roids 2a–5a was somewhat lower than that of 1a, an increase inthe reaction time did not lead to a considerable increase in theyields of carboxamides 2b–5b. Although almost complete conver-sions of 2a–5a were obtained after 24 h, the yields of 2b–5b didnot change considerably and the formation of an unidentified,probably polymeric or oligomeric material as side product wasobserved.

3.2. Copper-catalyzed azide–alkyne cycloaddition of steroids 1b–5b

Copper-catalyzed azide–alkyne cycloaddition, one of the typical‘click’ reactions [32], has attracted wide attention for its highreaction efficiency, mild reaction conditions and stability of theproducts. The reaction is regiospecific and leads exclusively to1,4-disubstituted 1,2,3-triazoles. Besides, it is not significantlyaffected by the steric and electronic properties of the groupsattached to the azide and alkyne centers, so it can efficiently beused for labeling of biological molecules [33].

Heterocyclic systems have potent receptor binding properties,so the incorporation of such moieties into steroidal side chainsmay alter or enhance biological activity. Accordingly, the synthesis

re (bar) Conversion (%)b Yield (%)b

1b 7

85 47 2378 87 1393 92 898 95 5

Scheme 2. Aminocarbonylation of steroids 1a–5a (isolated yields are presented inbrackets).

Scheme 3. Copper-catalyzed azide–alkyne cycloaddition of 1b-5b and 9 (isolatedyields are presented in brackets).

Scheme 4. Copper-catalyzed azide–alkyne cycloaddition of 2b and 10 (isolatedyield is presented in brackets).

E. Szánti-Pintér et al. / Steroids 76 (2011) 1377–1382 1381

of several steroidal heterocycles with valuable pharmacologicalactivities have been reported [34]. Various 17-imidazolyl- and17-triazolyl-androstanes were shown to be efficient inhibitors ofhuman 17a-hydroxylase-C17,20-lyase as well as 5a-reductase[35,36]. A 17-triazolyl steroid was found to have anti-androgenicactivity and to inhibit the growth of human prostate cancer celllines [37]. In spite of the evident significance of the incorporationof the triazolyl moiety into the side chain of steroids, there are onlya few examples for the functionalization of steroids by the copper-catalyzed azide–alkyne cycloaddition. Recently, Banday et al. re-ported the synthesis of 21-triazolyl pregnenolone derivatives bythis methodology. The new compounds showed promising anti-cancer activity [38].

In the cycloaddition reaction, the catalytically active species is aCu(I) complex. Although various Cu(I) salts and complexes can effi-ciently be used as catalyst, the simplest catalytic system consists ofCuSO4 and sodium ascorbate. The catalytically active Cu(I) speciesis formed in situ from the Cu(II) salt in the presence of the ascor-bate as the reducing agent. In this case, it is not necessary to useabsolutely oxygen-free conditions and the reaction usually takesplace smoothly at atmospheric pressure, at room temperature inan organic solvent/water mixture [39].

These reaction conditions were applied in the synthesis of ste-roid-ferrocene conjugates 1c–5c (Scheme 3) using ferrocenyl azide9 as the reaction partner. The reactions were checked by thin-layerchromatography. Steroidal carboxamides 1b–5b could be con-verted to the products 1c–5c selectively, no formation of side prod-ucts was observed. However, good yields could be obtained onlywhen freshly made ferrocenyl azide 9 was used. A slow decompo-sition of 9 was observed by TLC even when the azide was stored inan inert atmosphere.

As racemic ferrocenyl azide 9 was used in all of the reactions,cycloadditions led undoubtedly to epimers of products 1c–5c.However, the epimers could not be separated by column chroma-tography and gave identical Rf values when reaction mixtures wereanalyzed by TLC. Also, the two epimers gave identical signals bothin the 1H and 13C NMR spectra. This phenomenon might be ex-plained by the fact that the chiral center of the side chain is rela-tively far from those of the steroid skeleton. Besides, it isconnected to the steroidal core by a flexible linker containing amethylene group.

The ferrocenyl alkenyl azide (10) also readily underwent theazide–alkyne cycloaddition with carboxamide 2b resulting in highyields comparable to those obtained with azide 9 (Scheme 4).

4. Conclusion

Ferrocene labeled steroidal 17-carboxamides can easily beobtained via a two-step reaction sequence starting from thecorresponding alkenyl iodides. The first step involves the palla-dium-catalyzed aminocarbonylation of the alkenyl iodides withprop-2-yn-1-amine as a nucleophile in order to introduce a C–Ctriple bond into the steroidal side chain. Side reactions can be

1382 E. Szánti-Pintér et al. / Steroids 76 (2011) 1377–1382

suppressed by the proper choice of reaction conditions and the 17-(N-(prop-2-ynyl)-carbamoyl)-steroids can be obtained in goodyield. These latter compounds undergo a facile azide–alkyne cyclo-addition with ferrocenyl azides in the presence of CuSO4/sodiumascorbate to produce the steroid–ferrocene conjugates in good toexcellent yield. Both the palladium- and copper-catalyzed reactionstolerate various functional groups in the steroid framework.

Recently, the efficient synthesis of several steroidal alkenyl io-dides, incorporating the iodo-ene functionality in various positionsof a steroid, has been reported [11,12,40]. So the reaction sequencedescribed in the present paper can serve as a useful methodologyfor the introduction of both the triazolyl and ferrocenyl moietiesinto the 11-, 12- or 17 positions of steroids of diverse structure.

Acknowledgments

The authors thank the Hungarian National Science Foundationfor the financial support (OTKA NK71906, CK78553). The help ofLászló Márk in the measurement of some mass spectra is alsoacknowledged.

References

[1] Aggarwal S, Thareja S, Verma A, Bhardwaj TR, Kumar M. An overview on 5a-reductase inhibitors. Steroids 2010;75:109–53.

[2] Uygur MC, Arik AI, Altug U, Erol D. Effects of the 5a-reductase inhibitorfinasteride on serum levels of gonadal, adrenal, and hypophyseal hormonesand its clinical significance: a prospective clinical study. Steroids1998;63:208–13.

[3] Rittmaster R, Hahn RG, Ray P, Shannon JB, Wurzel R. Effect of dutasteride onintraprostatic androgen levels in men with benign prostatic hyperplasia orprostate cancer. Urology 2008;72:808–12.

[4] Levy MA, Brandt M, Sheedy KM, Dinh JT, Holt DA, Garrison LM, Bergsma DJ,Metcalf BW. Epristeride is a selective and specific uncompetitive inhibitor ofhuman steroid 5a-reductase isoform 2. J Steroid Biochem Mol Biol1994;48:197–206.

[5] Morera E, Ortar G. A palladium-catalyzed route to primary amides.Tetrahedron Lett 1998;39:2835–8.

[6] Tuba Z, Horváth J, Széles J, Kollár L, Balogh G, inventors; Chemical Works ofGedeon Richter Ltd, assignee. Novel process for preparing 17b-substituted 4-azaandrostane derivatives. WO 95/00531. 5 January 1995.

[7] Skoda-Földes R, Szarka Z, Kollár L, Dinya Z, Horváth J, Tuba Z. Synthesis of N-substituted steroidal hydrazides in homogeneous catalytichydrazinocarbonylation reaction. J Org Chem 1999;64:2134–6.

[8] Holt DA, Levy MA, Ladd DL, Oh H, Erb JM, Heaslip JI, Brandt M, Metcalf BW.Steroidal A ring aryl carboxylic acids: a new class of steroid 5a-reductaseinhibitors. J Med Chem 1990;33:937–42.

[9] Holt DA, Oh H, Levy MA, Metcalf BW. Synthesis of a steroidal A ring aromaticsulfonic acid as an inhibitor of steroid 5a-reductase. Steroids 1991;56:4–7.

[10] Holt DA, Levy MA, Metcalf BW, inventors; Smithkline Beecham Co, assignee.Phosphonic acid substituted aromatic steroids as inhibitors of steroid 5a-reductase EP Patent 0375344 A1, 27 June 1990.

[11] Ács P, Müller E, Czira G, Mahó S, Pereira M, Kollár L. Facile synthesis of 12-carboxamido-11-spirostenes via palladium-catalyzed carbonylation reactions.Steroids 2006;71:875–9.

[12] Ács P, Jakab B, Takács A, Kollár L. Facile synthesis of 11-carboxamidoandrost-4,9(11)-dienes via palladium-catalyzed aminocarbonylation. Steroids2007;72:627–32.

[13] Ahmed S, Denison S. Structure activity relationship study of known inhibitorsof the enzyme 5a-reductase (5AR). Bioorg Med Chem Lett 1998;8:409–14.

[14] Fouda MFR, Abd-Elzaher MM, Abdelsamaia RA, Labib AA. On the medicinalchemistry of ferrocene. Appl Organomet Chem 2007;21:613–25.

[15] Osella D, Nervi C, Galeotti F, Cavigiolio G, Vessieres A, Jaouen G. Theferrocenylethynyl unit: a stable hormone tag. Helv Chim Acta2001;84:3289–98.

[16] Higashi T, Shimada K. Derivatization of neutral steroids to enhance theirdetection characteristics in liquid chromatography – mass spectrometry. AnalBioanal Chem 2004;378:875–82.

[17] Barton DHR, Bashiardes B, Fourrey JL. An improved preparation of vinyliodides. Tetrahedron Lett 1983;24:1605–8.

[18] (a) Sudhir VS, Venkateswarlu C, Musthafa OTM, Sampath S, Chandrasekaran S.Click chemistry inspired synthesis of novel ferrocenyl-substituted amino acidsor peptides. Eur J Org Chem 2009:2120–9;(b) Holub JM, Jang H, Kirshenbaum K. Clickety-click: highly functionalizedpeptoid oligomers generated by sequential conjugation reactions on solid-phase support. Org Biomol Chem 2006;4:1497–502;(c) Köster SD, Dittrich J, Gasser G, Hüsken N, Castañeda ICH, Jios JL, VédovaCOD, Metzler-Nolte N. Spectroscopic and electrochemical studies of ferrocenyl

triazole amino acid and peptide bioconjugates synthesized by click chemistry.Organometallics 2008;27:6326–32.

[19] (a) Casas-Solvas JM, Ortiz-Salmerón E, Giménez-Martínez JJ, García-Fuentes L,Capitán-Vallvey LF, Santoyo-González F, Vargas-Berenguel A. Ferrocene–carbohydrate conjugates as electrochemical probes for molecular recognitionstudies. Chem Eur J 2009;15:710–25;(b) Casas-Solvas JM, Ortiz-Salmerón E, Fernández I, García-Fuentes L, Santoyo-González F, Vargas-Berenguel A. Ferrocene–b-cyclodextrin conjugates:synthesis, supramolecular behavior, and use as electrochemical sensors.Chem Eur J 2009;15:8146–62;(c) Hasegawa T, Umeda M, Numata M, Fujisawa T, Haraguchi S, Sakuraib K,Shinkai S. Click chemistry on curdlan: a regioselective and quantitativeapproach to develop artificial b-1,3-glucans with various functionalappendages. Chem Lett 2006;35:82–3;(d) Hasegawa T, Umeda M, Numata M, Li C, Bae AH, Fujisawa T, Haraguchi S,Sakurai K, Shinkai S. Click chemistry on polysaccharides: a convenient, general,and monitorable approach to develop 1,3-b-D-glucans with various functionalappendages. Carbohydrate Res 2006;341:35–40.

[20] Ramírez-López P, de la Torre MC, Montenegro HE, Asenjo M, Sierra MA. Astraightforward synthesis of tetrameric estrone-based macrocycles. Org Lett2008;10:3555–8.

[21] Casas-Solvas JM, Vargas-Berenguel A, Capitán-Vallvey LF, Santoyo-González F.Convenient methods for the synthesis of ferrocene-carbohydrate conjugates.Org Lett 2004;6:3687–90.

[22] Murakami Y, Watanabe T, Suzuki H, Kotake N, Takahashi T, Toyonari K, OhnoM, Takase K, Suzuki T, Kondo K. Synthetic studies on indoles and relatedcompounds. 43. New findings on the Hemetsberger-Knittel reaction. ChemPharm Bull 1997;45:1739–44.

[23] Vicennati P, Cozzi PG. Facile access to optically active ferrocenyl derivativeswith direct substitution of the hydroxy group catalyzed by indium tribromide.Eur J Org Chem 2007:2248–53.

[24] Molina P, Pastor A, Vilaplana MJ, Desamparados Velasco M. Structuralcharacterization and electrochemical study of novel ferrocene derivativesprepared from [(b-ferrocenylvinyl)imino]phosphorane by aza Wittig reactions.Organometallics 1997;16:5836–43.

[25] Skoda-Földes R, Takács E, Horváth J, Tuba Z, Kollár L. Palladium-catalysedaminocarbonylation of steroidal 17-iodo-androst-16-ene derivatives in N,N’-dialkyl-imidazolium-type ionic liquids. Green Chem 2003;5:643–5.

[26] Müller E, Péczely G, Skoda-Földes R, Takács E, Kokotos G, Bellis E, Kollár L.Homogeneous catalytic aminocarbonylation of iodoalkenes and iodobenzenewith amino acid esters under conventional conditions and in ionic liquids.Tetrahedron 2005;61:797–802.

[27] Jutand A. Contribution of electrochemistry to organometallic catalysis. ChemRev 2008;108:2300–47.

[28] Arcadi A. Carbonylation of enolizable ketones (enol triflates) and iodoalkenes.In: Kollar L, editor. Modern carbonylation methods. Weinheim: Wiley-VCH;2008. p. 223–50.

[29] Skoda-Földes R, Székvölgyi Z, Kollár L, Berente Z, Horváth J, Tuba Z. Facilesynthesis of steroidal phenyl ketones via homogeneous catalyticcarbonylation. Tetrahedron 2000;56:3415–8.

[30] Skoda-Földes R, Csákai Z, Kollár L, Szalontai G, Horváth J, Tuba Z. Homogeneouscatalytic dehalodimerization of 17-iodo-D16 steroids. Steroids1995;60:786–90.

[31] Balogh J, Mahó S, Háda V, Kollár L, Skoda-Földes R. Facile synthesis of steroidalprimary amides via palladium-catalyzed aminocarbonylation of steroidalalkenyl halides. Synthesis 2008:3040–2.

[32] Kolb HC, Finn MG, Sharpless KB. Click chemistry: diverse chemical functionfrom a few good reactions. Angew Chem Int Ed 2001;40:2004–21.

[33] Best MD. Click chemistry and bioorthogonal reactions: unprecedentedselectivity in the labeling of biological molecules. Biochemistry2009;48:6571–84.

[34] Schneider G, Wölfling J. Synthetic cardenolides and related compounds. CurrOrg Chem 2004;8:1381–403.

[35] Njar VCO, Kato K, Nnane IP, Grigoryev DN, Long BJ, Brodie AMH. Novel 17-Azolyl steroids, potent inhibitors of human cytochrome 17a-hydroxylase-C17,20-lyase (P45017a): Potential agents for the treatment of prostate cancerJ. Med Chem 1998;41:902–12.

[36] Handratta VD, Vasaitis TS, Njar VCO, Gediya LK, Kataria R, Chopra P, NewmanJr D, Farquhar R, Guo Z, Yiiu Y, Brodie AMH. Novel C-17-heteroaryl steroidalCYP17 inhibitors/antiandrogens: synthesis, in vitro biological activity,pharmacokinetics, and antitumor activity in the LAPC4 human prostatecancer xenograft model. J Med Chem 2005;48:2972–84.

[37] Nnan LP, Njar VCO, Brodie AMH. Pharmacokinetic profile of 3b-hydroxy-17-(1H-1, 2, 3-triazol-1-yl)androsta-5,16-diene (VN/87–1), a potent androgensynthesis inhibitor, in mice. J Ster Biochem Mol Biol 2001;78:241–6.

[38] Banday AH, Verma M, Srikakulam S, Gupta BD, Kumar HMS. D-ring substituted1,2,3-triazolyl 20-keto pregnenanes as potential anticancer agents: Synthesisand biological evaluation. Steroids 2010;75:801–4.

[39] Hein CD, Liu XM, Wang D. Click chemistry, a powerful tool for pharmaceuticalsciences. Pharm Res 2008;25:2216–30.

[40] Ács P, Takács A, Szilágyi A, Wölfling J, Schneider G, Kollár L. The synthesis of13a-androsta-5,16-diene derivatives with carboxylic acid, ester andcarboxamido functionalities at position-17 via palladium-catalyzedcarbonylation. Steroids 2009;74:419–23.