Novel P450(17alpha) inhibitors: 17-(2'-oxazolyl)- and 17-(2'-thiazolyl)-androstene derivatives.

Article abstract of DOI:10.1016/S0039-128X(03)00082-5

Twelve 17-(2'-oxazolyl)- and 17-(2'-thiazolyl)-androsta-5,16-diene derivatives were designed and synthesized from 3 beta-acetoxy-pregna-5,16-dien-20-one (1b) as inhibitors of 17 alpha-hydroxylase-C(17,20)-lyase (P450(17 alpha)). Potent inhibitors of this enzyme could be of value as treatment of prostate cancer. Two substituents (methyl and phenyl) were introduced either at their 4'- or 5'-position in order to investigate their structure-activity relationship. Due to the 16,17-double bond, 17-thiazoles were generally obtained in low yield. The pharmacological results showed that the compounds containing 17-(2'-oxazolyl) (14c) and 17-(2'-thiazolyl) (8c) (41.5%) demonstrated reasonable inhibition against P450(17 alpha). Their 3-acetate (13c and 7c) were less potent than their 3-OH counterparts. The introduction of a phenyl or methyl group generally decreased inhibitory activity. Surprisingly, 17-(5'-methyl-2'-thiazolyl) (12a) was the most potent compound in this series and was almost as potent as L-39, which has good antitumor activity.

Full text of DOI:10.1016/S0039-128X(03)00082-5

Steroids 68 (2003) 603–611
Novel P45017␣ inhibitors: 17-(2ꢀ-oxazolyl)- and
17-(2ꢀ-thiazolyl)-androstene derivatives
Na Zhua, Yangzhi Linga, Xiaoping Leia, Venkatesh Handrattab, Angela M.H. Brodieb,∗
a
Department of Medicinal Chemistry, School of Pharmaceutical Sciences, Peking University Health Center, Peking 100083, PR China
b
Department of Pharmacology & Experimental Therapeutics, School of Medicine, University of Maryland,
685 West Baltimore Street, Room 580G Baltimore, MD 21201, USA
Received 15 November 2002; received in revised form 27 May 2003; accepted 30 May 2003
Abstract
Twelve 17-(2ꢀ-oxazolyl)- and 17-(2ꢀ-thiazolyl)-androsta-5,16-diene derivatives were designed and synthesized from 3␤-acetoxy-pregna-
5,16-dien-20-one (1b) as inhibitors of 17␣-hydroxylase-C17,20-lyase (P45017␣). Potent inhibitors of this enzyme could be of value as
treatment of prostate cancer. Two substituents (methyl and phenyl) were introduced either at their 4ꢀ- or 5ꢀ-position in order to investigate
their structure–activity relationship. Due to the 16,17-double bond, 17-thiazoles were generally obtained in low yield. The pharmacological
results showed that the compounds containing 17-(2ꢀ-oxazolyl) (14c) and 17-(2ꢀ-thiazolyl) (8c) (41.5%) demonstrated reasonable inhibition
against P45017␣. Their 3-acetate (13c and 7c) were less potent than their 3-OH counterparts. The introduction of a phenyl or methyl group
generally decreased inhibitory activity. Surprisingly, 17-(5ꢀ-methyl-2ꢀ-thiazolyl) (12a) was the most potent compound in this series and
was almost as potent as L-39, which has good antitumor activity.
© 2003 Elsevier Inc. All rights reserved.
Keywords: P45017␣ inhibitors; 17-(2ꢀ-Oxazolyl)-androstenes; 17-(2ꢀ-Thiazolyl)-androstenes; Prostatic cancer
1. Introduction
I clinical trial [1]. Jarman et al. also reported that
17-(3ꢀ-pyridyl)-androsta-5,16-dien-3-ol is
potent in-
a
hibitor [6]. These compounds were much more active
than ketoconazole. Although the detailed mechanisms of
these inhibitors are not very clear, the coordination of
17-heterocycles of steroids to the heme-iron at the active
site of the enzyme is thought to be an important factor
related to inhibition [1,4,5]. In addition, the 16,17-double
bond is also important for potent inhibitory activity against
P45017␣ [6]. We reasoned that 17-thiazole and -oxazole,
which are other azoles and the bioisosterism of imidazole,
pyrazole, and isoxazole, would also create potent inhibitors.
Here we report the synthesis of twelve 17-(2ꢀ-thiazolyl)- and
17-(2ꢀ-oxazolyl)-androsta-5,16-dien-3-ol derivatives, with
either a methyl or phenyl group at their 4ꢀ- or 5ꢀ-position.
Their synthetic routes are shown in the Schemes 1 and 2 and
their inhibition activities for P45017␣ are also reported here.
17␣-Hydroxylase-C17,20-lyase (P45017␣) is a key regula-
tory enzyme of the androgen biosynthetic pathway that cat-
alyzes both the 17␣-hydroxylation and the cleavage of the
C17–C20 side chain of 21-carbon steroids in both testes and
adrenals. It is known that androgens play an important role
in the development and progression of several prostatic dis-
eases, most notably, benign prostatic hypertrophy (BPH) and
prostatic cancer. Inhibitors of this enzyme can block andro-
gen synthesis in its early step, and thereby may be useful in
the treatment of prostatic cancer [1]. To date, only ketocona-
zole [2,3], an imidazole antifungal agent, has been used for
this purpose to treat patients with advanced prostatic cancer.
However, this agent is neither selective nor very potent and
has a number of significant side effects. This highlights the
need to design potent and specific inhibitors of P45017␣
.
Recently, we have described a number of inhibitors of
P45017␣, of which 17-imidazolyl, pyrazolyl, and isoxazolyl
androstene are very potent [4,5] and 17-(5ꢀ-isoxazolyl)-
pregna-4,16-diene-3,20-dione (L-39) will soon enter Phase
2. Experimental
2.1. Synthetic methods
∗
Melting points were determined on a XT4 melting point
microscope and are uncorrected. IR spectra were deter-
Corresponding author. Tel.: +1-410-706-3137; fax: +1-410-706-0032.
E-mail address: abrodie@umaryland.edu (A.M.H. Brodie).
0039-128X/$ – see front matter © 2003 Elsevier Inc. All rights reserved.
doi:10.1016/S0039-128X(03)00082-5

604
N. Zhu et al. / Steroids 68 (2003) 603–611
iii or iv
+
N
-
O
I
COCH
2
COOH
R
R
1
R
R
1
R
1
2
2
i
R
ii
2
AcO
1a
1b
AcO
3a
3b
HO
2a
2b
a: R =R =H
1
2
b: R ,R = double bond
1
2
Scheme 1. Regent: (i) I2/pyridine; (ii) NaOH, 50% aqueous EtOH; (iii) I2, NaOH; (iv) Br2, NaOH. a: R1 = R2 = H; b: R1, R2 = double bond.
mined in a Perkin-Elmer 983 spectrometer (wave numbers
in cm−1). 1H NMR data (300 MHz) (internal standard
Me4Si, δ = 0) were recorded in a VARIAN VAN-300, with
samples dissolved in CDCl3 unless otherwise stated. Reac-
tions were monitored by TLC on silica gel plates (GF254)
and visualized by dipping in 4% sulfuric acid in ethanol
followed by heating at ca. 120–150 ◦C. Flash column chro-
matography was carried out on silica gel (230–400 mesh) in
the solvent systems indicated. PE refers to light petroleum
ether, bp 60–90 ◦C.
COOH
COCl
CONH
2
ii
v
i
4
3d
AcO
3C
iii
iv
CONHCH COR
2
CSNH
CN
2
(
)
9a,9b
iii
vii
N
5
6
CH
Ph
3
v
N
R
O
R
R
O
S
N
N
N
S
SH
(
)
S
15
11
13a,13b,13c
vi
10a,10b
vi
7a,7b,7c
vi
CH
R
3
R
R
N
vi
N
N
O
N
S
S
O
HO
HO
HO
HO
16
12a,12b
14a,14b,14c
8a,8b,8c
a = CH , b = Ph , c = H
R:
3
Scheme 2. Regent: (i) SOCl2; (ii) NH4OH; (iii) P2S5; (iv) RCOCH2NH2·HCl/Et3N; (v) RCOCH2Br or ClCH2CHClOEt; (vi) KOH/CH3OH; (vii)
PPh3/I2/NEt3. R: a = CH3; b = Ph; c = H.

N. Zhu et al. / Steroids 68 (2003) 603–611
605
2.2. 3β-Hydroxy-androsta-5,16-dien-17-carboxylic acid
(1a)
2.3. 3β-Acetoxy-androsta-5,16-dien-17-carboxylic
acid (3c)
2.2.1. Method 1: the hydrolysis of 21-pyridinium iodide
salt (2b)
17-Acid 3b (0.75 g, 2.37 mmol) was dissolved by warm-
ing in 3 ml of pyridine. After the solution had cooled to room
temperature, 1 ml of acetic anhydride was added and the
mixture was allowed to stand overnight. It was then treated
with 4 ml of water and heated to a boil until the precipitate
had dissolved. Another 10 ml of water was added and the
mixture was cooled. The crystalline product was collected
and recrystalized from glacial acetic acid to give 3␤-acetate
3c 0.78 g (91%) as white crystals, mp 240–244 ◦C (lit. [9]:
255 ◦C). IR (cm−1): 1726 (AcO).
3␤-Acetoxy-pregna-5,16-dien-20-one (DPA, 1b) (0.213 g,
0.6 mmol) was dissolved in 0.5 ml of pyridine and heated
under reflux, iodine (0.163 g, 0.64 mmol) was then added
within 0.5 h and the solution was heated under reflux for 2 h.
After cooling naturally, water was added and the precipitate
was filtered, washed with water, and dried. Crude product
was recrystallized from CH2Cl2/ethyl acetate to give pure
pyridinium iodide 2b (0.31 g, 91.7%), mp 200–202 ◦C. H
1
NMR: δ 8.92 (d, J = 5.4, 2H, 2ꢀ,6ꢀ-H of Py), 8.69 (t, 1H,
4ꢀ-H of Py), 8.23 (t, 2H, 3ꢀ,5ꢀ-H of Py), 7.26 (ws, 1H, 16-H),
6.10 (q, 2H, 17-CH2), 5.39 (d, J = 4.8, 1H, 6-H), 4.46
(m, 1H, 3␣-H), 1.99 (s, 3H, AcO), 1.02 (s, 3H, 19-CH3),
0.90 (s, 3H, 18-CH3). Anal. C28H36O3NI·1/2(H2O), C
58.94%, H 6.53%, N 2.45%; found C 59.23%, H 6.54%,
N 1.83%.
Sodium hydroxide (175 mg, 4.375 mmol) was added to
a suspension of the above pyridinium iodide 2b (190 mg,
0.34 mmol) in 50% aqueous ethanol. The mixture was heated
under reflux for 2 h and acidified with 3N HCl to pH 3. The
acidic fraction was collected by filtration. The crude product
(60 mg, 26%) was crystallized from methanol to give a dark
yellow product 3b (48 mg, 18%) with mp 249–257 ◦C (lit.
[8]: 255–257 ◦C).
2.3.1. 3β-Acetoxy-androsta-5,16-dien-17-carboxylic acid
chloride (3d) and 17-carboxamide (4)
Compound 3c (2.5 g, 6.98 mmol) was mixed with SOCl2
(10 ml) at 0 ◦C, and then stirred for 6 h at room tempera-
ture. The solvent was evaporated under reduced pressure af-
ter addition of C6H6 to the mixture in order to get rid of
any remaining SOCl2. This gave the acid chloride 3d. This
compound was stirred vigorously with concentrated aqueous
NH4OH (10 ml), the precipitate was collected and dried, and
yielded the amide 4 (2.345 g, 94.1%), mp 217–220 ◦C (from
ethanol). IR (cm−1): 3465, 3326 (NH2), 1668 (CONH2),
1726 (AcO). 1H NMR: δ 6.49 (ws, 1H, 16-H), 5.70 (ws, 2H,
NH2), 5.39 (d, J = 5.1, 1H, 6-H), 4.60 (m, 1H, 3␣-H), 2.04
(s, 3H, AcO), 1.07 (s, 3H, 19-CH3), 1.02 (s, 3H, 18-CH3).
Anal. C22H37O3N, C 73.92%, H 8.74%, N 3.92%; found C
74.20%, H 8.39%, N 3.64%.
2.2.2. Method 2: bromoform reaction
A solution of sodium hydroxide (1.46 g, 36.5 mmol) in
12.5 ml water was cooled to −5 ◦C in an ice-salt bath at a rate
that maintained the reaction temperature below 0 ◦C. While
stirring, bromine (1.5 g, 9 mmol, 0.48 ml) was added from
a separatory funnel at a rate that maintained the reaction
temperature below 0 ◦C. The ice-cold solution was diluted
with 8.3 ml cold dioxane. This solution was kept at 0 ◦C
until required.
A solution of DPA 1b (1 g, 2.8 mmol) in dioxane (38.2 ml)
and water (11 ml) was cooled in an ice bath. When the in-
ternal temperature had fallen to 8 ◦C, the above cold hypo-
bromite solution was added. The temperature of the mixture
was maintained at 8–10 ◦C throughout the reaction. After
the solution became colorless, the mixture was stirred for
an additional 2 h. The excess sodium hypobromite was de-
stroyed by the addition of 10% aqueous sodium sulfite. The
mixture was refluxed for 15 min, and the solution, while still
hot (90 ◦C), was acidified to pH 6. The solution was kept in
water for 24 h. The precipitate was collected by suction fil-
tration, washed with water, and dried. The white product was
crystallized from ethanol to give pure 3b, mp 255–257 ◦C
(0.91 g, 81%, lit. [8]: 255–257 ◦C). IR (cm−1): 3173, 1710
(COOH, OH). 1H NMR: δ 12.03 (s, 1H, COOH), 6.66 (ws,
1H, 16-H), 5.28 (d, J = 4.2, 1H, 6-H), 4.62 (s, 1H, OH),
3.26 (m, 1H, 3␣-H), 0.98 (s, 3H, 19-CH3), 0.87 (s, 3H,
18-CH3).
2.4. 3β-Acetoxy-androsta-5,16-dien-
17-thiocarboxamide (5)
Phosphorus pentasulfide (62 mg, 0.28 mmol) was stirred
vigorously in anhydrous dioxane (4 ml) to become a pow-
dered solid, to which the amide 4 (100 mg, 0.28 mmol)
was added. The mixture was stirred at 14 ◦C for 2 h and
then filtered. The solution was adjusted to neutral by
adding 50% aqueous KOH. Then the solvent was evap-
orated under reduced pressure and the residue purified
by flash chromatography. Elution with acetone/PE (1:4)
yielded thiocarboxamide 5 (50 mg, 48.3%), mp 173–175 ◦C
(from acetone/PE). IR (cm−1): 3314, 3191 (NH2), 1382
1
(CSNH2), 1707 (AcO). H NMR: δ 7.46 and 6.59 (s, each
=
1H, –CSNH2↔–HS–C NH–, exchangeable with D2O),
5.38 (d, J = 5.1, 1H, 6-H), 4.62 (m, 1H, 3␣-H), 4.15 (m,
1H, 16-H), 2.04 (s, 3H, AcO), 1.02 (s, 3H, 19-CH3), 0.79
(s, 3H, 18-CH3). Anal. C22H31ONS·2(H2O), C 64.51%, H
8.61%, N 3.42%; found C 64.56%, H 8.24%, N 3.10%.
2.5. 3β-Acetoxy-17-cyano- (6) and -17-(4ꢀ-methyl-
2ꢀ-thiazolyl)-androsta-5,16-diene (7a)
A suspension of powdered P2S5 (23.2 mg, 0.10 mmol)
and the amide 4 (200 mg, 0.5 mmol) in anhydrous diox-

606
N. Zhu et al. / Steroids 68 (2003) 603–611
ane (10 ml) was stirred at 14 ◦C for 2 h. The bromoacetone
(43.2 mg, 0.32 mmol, bp 68–75 ◦C, prepared from the bromi-
nation of acetone in acetic acid [10]) was added and the mix-
ture was refluxed for 2 h, 50% KOH was added to neutralize
the solution. Solvent was removed under reduced pressure
and the residue was flash chromatographed. Elution with
PE/ethyl acetate (20:1) yielded a white product (7a) (20 mg,
8.8%), mp 199–200 ◦C (from acetone). IR (cm−1): 1717,
1243, 1034 (AcO), 1406 (thiazolyl). 1H NMR: δ 6.75 (s, 1H,
5ꢀ-H), 6.50 (ws, 1H, 16-H), 5.40 (d, J = 5.1, 1H, 6-H), 4.63
(m, 1H, 3␣-H), 2.45 (s, 3H, 4ꢀ-CH3), 2.04 (s, 3H, AcO), 1.08
(s, 3H, 19-CH3), 1.07 (s, 3H, 18-CH3). Anal. C25H33O2NS,
C 72.95%, H 8.08%, N 3.40%; found C 72.74%, H 8.32%,
N 3.11%.
Further elution with PE/ethyl acetate (9:1) yielded
17-cyanide 6 (20 mg, 14%), mp 200–202 ◦C (from ace-
tone). IR (cm−1): 2209 (CN), 1720 (AcO). 1H NMR: δ
6.64 (m, 1H, 16-H), 5.38 (d, J = 4.8, 1H, 6-H), 4.60 (m,
3H, 3␣-H), 2.04 (s, 3H, AcO), 1.06 (s, 3H, 19-CH3), 0.95
(s, 3H, 18-CH3). Anal. C22H29O2N, C 77.55%, H 8.85%,
N 4.14%; found C 77.84%, H 8.61%, N 4.13%.
1H, 4ꢀ-H), 7.20 (d, J = 3.3, 1H, 5ꢀ-H), 6.54 (ws, 1H, 16-H),
5.36 (d, J = 5.1, 1H, 6-H), 4.58 (m, 1H, 3␣-H), 2.09 (s,
3H, AcO), 1.07 (s, 3H, 19-CH3), 1.04 (s, 3H, 18-CH3).
2.6. 3β-Hydroxy-17-(4ꢀ-methyl-2ꢀ-thiazolyl)-
(8a), -17-(4ꢀ-phenyl-2ꢀ-thiazolyl)- (8b), and
-17-(2ꢀ-thiazolyl)-androsta-5,16-diene (8c)
Compound 7a (45 mg, 0.11 mmol) was dissolved by
warming in methanol (10 ml), 0.86N KOH–CH3OH (2 ml)
was added, and the reaction solution was heated under
reflux for 0.5 h. The solvent was removed under reduced
pressure and the residue was chromatographed. Elution
with PE/ethyl acetate (4:1) yielded 8a (28 mg, 70%), mp
236–237 ◦C (from acetone). 1H NMR: δ 6.74 (s, 1H, 5ꢀ-H),
6.48 (ws, 1H, 16-H), 5.38 (d, J = 4.8, 1H, 6-H), 3.52 (m,
1H, 3␣-H), 2.44 (s, 3H, 4ꢀ-CH3), 1.07 (s, 6H, 18,19-CH3).
Anal. C23H31ONS, C 74.75%, H 8.45%, N 3.79%; found
C 74.95%, H 8.70%, N 3.54%.
Following the same procedure described above for 8a,
compound 7b (20 mg, 0.04 mmol) gave 3␤-ol 8b (18 mg,
99%), mp 205–207 ◦C (from acetone). IR (cm−1): 3344
(OH), 1437, 1372 (thiazolyl). 1H NMR: δ 7.87 (d, J = 7.2,
2H, Ph–H), 7.24–7.43 (m, 4H, Ph–H and 5ꢀ-H), 6.45 (ws, 1H,
16-H), 5.32 (d, J = 5.1, 1H, 6-H), 3.50 (m, 1H, 3␣-H), 1.08
(s, 3H, 19-CH3), 1.05 (s, 3H, 18-CH3). Anal. C28H33ONS,
C 77.91%, H 7.71%, N 3.24%; found C 77.64%, H 7.88%,
N 3.18%.
Similarly, compound 7c (14 mg, 0.03 mmol) gave 3␤-ol 8c
(10 mg, 80%), mp 185–187 ◦C (from acetone). IR (cm−1):
3267 (OH), 1595, 1497 (thiazolyl). 1H NMR: δ 7.80 (d,
J = 3.3, 1H, 4ꢀ-H), 7.22 (d, J = 3.3, 1H, 5ꢀ-H), 6.61
(ws, 1H, 16-H), 5.38 (d, J = 5.1, 1H, 6-H), 3.53 (m, 1H,
3␣-H), 1.09 (s, 3H, 19-CH3), 1.08 (s, 3H, 18-CH3). Anal.
C22H29ONS·1/2(H2O), calculated C 72.48%, H 8.29%, N
3.84%; found C 72.26%, H 8.06%, N 3.63%; MS (EI): 355
(M+).
2.5.1. 3β-Acetoxy-17-(4ꢀ-phenyl-2ꢀ-thiazolyl)-androsta-
5,16-diene (7b)
A suspension of powdered P2S5 (100 mg, 0.45 mmol) and
the amide 4 (200 mg, 0.5 mmol) in anhydrous dioxane (6 ml)
was stirred at 14 ◦C for 2 h. The P2O5 and other precipitates
were removed by filtration and then 2-bromoacetophenone
(100 mg, 0.53 mmol) was added into the filtrate, and then
refluxed for 0.5 h. The solution was concentrated under re-
duced pressure and the residue was diluted with CH2Cl2,
then 50% KOH was added to make the solution neutral. The
whole mixture was washed with water to remove the resid-
ual dioxane. The water layer was extracted with CH2Cl2
and the combined organic layer was dried over sodium sul-
fate. After evaporation, the residue was purified by flash
chromatography. Elution with PE/ethyl acetate yielded 7b
1
(30 mg, 10%), mp 173–174 ◦C (from acetone). H NMR: δ
7.92 (d, J = 7.5, 2H, Ph–H), 7.24–7.43 (m, 4H, Ph–H and
5ꢀ-H), 6.54 (ws, 1H, 16-H), 5.39 (d, J = 5.1, 1H, 6-H), 4.58
(m, 1H, 3␣-H), 2.03 (s, 3H, AcO), 1.08 (s, 3H, 19-CH3),
1.05 (s, 3H, 18-CH3).
2.7. N-Acetylmethyl-3β-acetoxy-androsta-5,16-diene-
17-carboxamide (9a)
Aminoacetone hydrochloride (16.4 mg, 0.15 mmol, pre-
pared from glycine and acetic anhydride [11]) and tri-
ethylamine (35 ul, 0.25 mmol) were added to a solu-
tion of 17-acid chloride 3 (50 mg, 0.12 mmol) in dry
dichloromethane at 0 ◦C. The mixture was stirred at room
temperature for 0.5 h. Then, it was washed successively
with 5% hydrochloric acid, 5% sodium hydrogen carbon-
ate, water, and then dried over sodium sulfate. The solvent
was evaporated and the residue was chromatographed. Elu-
tion with PE/acetone (4:1) gave ␤-methylketo amide 9a
(50 mg, 92%), bp 190–192 ◦C (from acetone). IR (cm−1):
2.5.2. 3β-Acetoxy-17-(2ꢀ-thiazolyl)-androsta-
5,16-diene (7c)
The 17-thiamide 5 (60 mg, 0.14 mmol) was dissolved in
warm anhydrous DMF, ␣,␤-dichloroether (20 ul, 0.14 mmol)
added and the mixture heated under reflux for 0.5 h. The
solution was concentrated under reduced pressure and the
residue was diluted with CH2Cl2 and washed with water to
remove the residual DMF. The water layer was extracted
with CH2Cl2, the combined organic layer was dried over
sodium sulfate. The solvent was removed and the residue
was purified by chromatography. Elution with PE/ethyl ac-
etate (15:1) yielded 17-thiazole 7c (10 mg, 15.75%), mp
1
1726, 1246, 1037 (AcO), 3375, 3326, 1649 (CONH). H
NMR: δ 6.45 (ws, 2H, 16-H and NH), 5.39 (d, J = 5.1,
1H, 16-H), 4.60 (m, 1H, 3␣-H), 4.27 (s, 2H, NHCH2CO),
2.23 (s, 3H, CH2COCH3), 2.04 (s, 3H, AcO), 1.06 (s,
160–162 ◦C (from acetone). H NMR: δ 7.77 (d, J = 3.3,
1

N. Zhu et al. / Steroids 68 (2003) 603–611
607
3H, 19-CH3), 1.02 (s, 3H, 18-CH3). Anal. C25H35O4N, C
72.61%, H 8.53%, N 3.39%; found C 72.51%, H 8.56%,
N 3.30%.
(o-Ph–C), 122 (6-C), 73 (3-C), 67 (16-C), 54 (17-C), 49
(9-C), 46 (13-C), 38 (14-C), 31 (8-C), 21 (AcO–C), 19
(19-C), 13 (18-C); MS (EI): 507 (M+), 474 (M+–SH), 414
(M+–SH–AcOH) NOESY: 17-H has no NOE with both
18-CH3 and 16-H. This suggests the configuration was
17␣-H, 16␤-H.
2.7.1. N-Benzoylmethyl-3β-acetoxy-androsta-
5,16-diene-17-carboxamide (9b)
Following the same procedure described above, acylation
of 2-aminoacetophenone hydrochloride (25 mg, 0.14 mmol)
with 17-acid chloride 3 yielded 9b (50 mg, 78%), mp
163–165 ◦C (from acetone). IR (cm−1): 3327 (NH), 1686
(CONH), 1727 (AcO). 1H NMR: δ 7.5–8.0 (m, 5H, Ph–H),
6.80 (s, 1H, NH), 6.55 (ws, 1H, 16-H), 5.41 (d, J = 5.1, 1H,
6-H), 4.84 (s, 2H, NHCH2CO), 4.61 (m, 1H, 3␣-H), 2.04
(s, 3H, AcO), 1.09 (s, 3H, 19-CH3), 1.08 (s, 3H, 18-CH3).
Anal. C30H37O4N, C 75.76%, H 7.84%, N 2.94%; found C
75.35%, H 7.88%, N 2.88%.
2.8.2. 3β-Hydroxy-17-(5ꢀ-methyl-2ꢀ-thiazolyl)- (12a) and
-17-(5ꢀ-phenyl-2ꢀ-thiazolyl)-androsta-5,16-diene (12b)
Following the same procedure described above for the
preparation of 8a, 3␤-acetate 10a (20 mg, 0.044 mmol) was
hydrolyzed in KOH/methanol to give 3␤-ol 12a (17 mg,
94%), mp 163–167 ◦C (from acetone/PE). IR (cm−1): 3370
1
(OH), 1458 (thiazole). H NMR: δ 7.39 (s, 1H, 4ꢀ-H), 6.41
(ws, 1H, 16-H), 5.37 (d, J = 5.1, 1H, 6-H), 3.50 (m, 1H,
3␣-H), 2.43 (s, 3H, 5ꢀ-CH3), 1.08 (s, 3H, 19-CH3), 1.05 (s,
3H, 18-CH3). Anal. C23H31ONS, C 74.75%, H 8.45%, N
3.79%; found C 74.84%, H 8.32%, N 3.62%.
And from 3␤-acetate 10b (20 mg, 0.042 mmol) gave 12b
(18 mg, 99%), mp 218–218.8 ◦C (from acetone). IR (cm−1):
3314 (OH), 1591, 1478 (aromatic ring). 1H NMR: δ 7.90 (s,
1H, 4ꢀ-H), 7.30–7.54 (m, 5H, Ph–H), 6.48 (ws, 1H, 16-H),
5.37 (d, J = 5.1, 1H, 6-H), 3.56 (m, 1H, 3␣-H), 1.10 (s, 3H,
19-CH3), 1.07 (s, 3H, 18-CH3). Anal. C28H33ONS·(H2O),
C 74.79%, H 7.85%, N 3.11%; found C 75.17%, H 7.53%,
N 2.91%.
2.8. 3β-Acetoxy-17-(5ꢀ-methyl-2ꢀ-thiazolyl)-androsta-
5,16-diene (10a)
Compound 9a (50 mg, 0.11 mmol) and phosphorus pen-
tasulfide (33.3 mg, 0.15 mmol) in dioxane (3 ml) was stirred
at room temperature for 3.5 h, and then refluxed for 1 h. The
solution was concentrated and washed with water. The water
layer was extracted with CH2Cl2 and the combined CH2Cl2
layer was dried over sodium sulfate. The solvent was evap-
orated and the residue was chromatographed. Elution with
PE/ethyl acetate (15:1) yielded product 9a (14 mg, 28%), mp
162–163 ◦C (from acetone). 1H NMR: δ 7.38 (s, 1H, 4ꢀ-H),
6.38 (ws, 1H, 16-H), 5.38 (d, J = 4.8, 1H, 6-H), 4.60 (m,
1H, 3␣-H), 2.42 (s, 3H, 5ꢀ-CH3), 2.05 (s, 3H, AcO), 1.06
(s, 3H, 19-CH3), 1.05 (s, 3H, 18-CH3).
2.9. 3β-Acetoxy-17-(4ꢀ-methyl-2ꢀ-oxazolyl)-androsta-
5,16-diene (13a)
The 17-carboxamide 4 (100 mg, 0.25 mmol) and bro-
moacetone (34 mg, 0.25 mmol, prepared from the bromi-
nation of acetone [10]) in DMF (2 ml) was refluxed for
0.5 h, the solution was concentrated under reduced pres-
sure. Water (10 ml) was added and extracted with CH2Cl2
(3 × 10 ml), the combined CH2Cl2 was evaporated, and
the residue was purified by chromatography. Elution with
PE/ethyl acetate (9:1) gave 4ꢀ-methyloxazole 13a (44 mg,
40.2%), mp 196–198 ◦C (from acetone). IR (cm−1): 1728,
1242, 1035 (AcO), 1615, 1507, 1370 (oxazolyl). 1H NMR:
δ 7.28 (s, 1H, 5ꢀ-H), 6.56 (ws, 1H, 16-H), 5.40 (d, J = 4.8,
1H, 6-H), 4.60 (m, 1H, 3␣-H), 2.19 (s, 3H, 4ꢀ-CH3), 2.04
(s, 3H, AcO), 1.08 (s, 3H, 19-CH3), 1.02 (s, 3H, 18-CH3).
2.8.1. 3β-Acetoxy-17-(5ꢀ-phenyl-2ꢀ-thiazolyl)- (10b)
and 3β-acetoxy-16α-thiol-17-(5ꢀ-phenyl-2ꢀ-thiazolyl)-
androsta-5,16-diene (11)
Compound 9a (100 mg, 0.2 mmol) and phosphorus pen-
tasulfide (67 mg, 0.3 mmol) in toluene (2 ml) were heated
at 80 ◦C for 2 h. The solvent was removed under reduced
pressure and the residue was purified by flash chromatog-
raphy. Elution with PE/ethyl acetate (15:1) yielded product
10b (20 mg, 21.1%), mp 155–158 ◦C (from acetone). IR
1
(cm−1): 1717 (AcO), 1457 (aromatic rings). H NMR: δ
7.94 (s, 1H, 4ꢀ-H), 7.20–7.59 (m, 5H, Ph–H), 6.65 (ws, 1H,
16-H), 5.39 (d, J = 4.8, 1H, 6-H), 4.60 (m, 1H, 3␣-H),
2.02 (s, 3H, AcO), 1.09 (s, 3H, 19-CH3), 1.08 (s, 3H,
18-CH3).
2.9.1. 3β-Acetoxy-17-(4ꢀ-phenyl-2ꢀ-oxazolyl)-androsta-
5,16-diene (13b)
The same procedure described above for 13a, was fol-
lowed except 2-bromoacetophenone (50 mg, 0.27 mmol)
replaced bromoacetone, to obtain 4ꢀ-phenyloxazole 13b
(68 mg, 67.9%), mp 196–197.8 ◦C (from acetone). IR
(cm−1): 1720, 1253 (AcO), 1618, 1564, 1482, 1372 (oxa-
Following the same procedure described above except
dioxane was used to replace toluene, 16␣-thiol 11 (13 mg,
1
12.7%) was obtained, mp 173–175 ◦C (from acetone). H
NMR: δ 7.89 (s, 1H, 4ꢀ-H), 7.27–7.54 (m, 5H, Ph–H), 5.37
(d, J = 5.1, 1H, 6-H), 4.60 (m, 1H, 3-H), 4.07 (m, 1H,
16-H), 3.04 (d, 1H, 17-H), 2.02 (s, 3H, AcO), 1.08 (s, 3H,
19-CH3), 0.63 (s, 3H, 18-CH3). 13C NMR: δ 170 (CO),
167 (2ꢀ-C), 139 (Ph–C linked with thiazole), 138 (5-C),
137 (4ꢀ-C), 131 (5ꢀ-C), 129 (m-Ph–C), 128 (p-Ph–C), 126
1
zolyl). H NMR: δ 7.83 (s, 1H, 5ꢀ-H), 7.26–7.78 (m, 5H,
Ph–H), 6.61 (m, 1H, 16-H), 5.41 (d, J = 5.4, 1H, 6-H), 4.62
(m, 1H, 3␣-H), 2.04 (s, 3H, AcO), 1.10 (s, 3H, 19-CH3),
1.07 (s, 3H, 18-CH3). Anal. C30H35O3N, C 78.74%, H
7.71%, N 3.06%; found C 78.80%, H 7.70%, N 2.97%.

608
N. Zhu et al. / Steroids 68 (2003) 603–611
2.9.2. 3β-Acetoxyl-17-(2ꢀ-oxazolyl)-androsta-
2.9.5. 3β-Hydroxy-17-(5ꢀ-methyl-2ꢀ-oxazolyl)-androsta-
5,16-diene (13c)
5,16-diene (16)
The same procedure described above for 13a was followed
but using ␣,␤-dichloroether (30 ul, 0.21 mmol) instead of
bromoacetone. The mixture was refluxed for 1 h and gave
oxazole 13c (16 mg, 15.1%), mp 178–179.5 ◦C (from ace-
tone/PE). IR (cm−1): 1726, 1238, 1031 (AcO), 1615, 1529,
1370 (oxazolyl). 1H NMR: δ 7.60 (d, 1H, 5ꢀ-H), 7.10 (d, 1H,
4ꢀ-H), 6.60 (ws, 1H, 16-H), 5.40 (d, J = 4.8, 1H, 6-H), 4.60
(m, 1H, 3␣-H), 2.04 (s, 3H, AcO), 1.09 (s, 3H, 19-CH3), 1.03
(s, 3H, 18-CH3). Anal. C24H31O3N, C 75.56%, H 8.19%,
N 3.67%; found C 75.37%, H 8.22%, N 3.43%.
Following the same procedure described above for 8a,
acetate 15 (20 mg, 0.046 mmol) gave the 3␤-ol 16 (10 mg,
50%), mp 200–202 ◦C (from acetone). 1H NMR: δ 6.72
(s, 1H, 4ꢀ-H), 6.47 (ws, 1H, 16-H), 5.37 (d, J = 4.8, 1H,
6-H), 3.54 (m, 1H, 3␣-H), 2.31 (s, 3H, 5ꢀ-CH3), 1.07 (s, 3H,
19-CH3), 1.02 (s, 3H, 18-CH3); MS (EI): 353 (M+).
3. P45017␣ (lyase) assay [17]
Abbreviations used: IPTG, isopropyl ␤-d-thiogalacto-
pyranoside; ALA, ␦-aminolevulenic acid; LBAG, LB
medium with ampicillin and glucose; TBA, terrific
broth containing ampicillin; DTT, dithiothretiol; MOPS,
3-(N-morpholino)propanesulfonic acid; ACN, acetonitrile.
2.9.3. 3β-Hydroxy-17-(4ꢀ-methyl-2ꢀ-oxazolyl)-
(14a), -17-(4ꢀ-phenyl-2ꢀ-oxazolyl)- (14b), and
-17-(2ꢀ-oxazolyl)-androsta-5,16-diene (14c)
Following the same procedure described above for 8a,
3␤-acetate 13a (40 mg) gave product 14a (10 mg, 56%), mp
249–250 ◦C (from acetone). IR (cm−1): 3280 (OH), 1601,
3.1. Materials
1
1529 (oxazolyl). H NMR: δ 7.24 (s, 1H, 5ꢀ-H), 6.52 (ws,
The vector pCWH17modHis was a generous gift of
Dr. M.R. Waterman (University of Texas, Southern Medi-
cal Center, Dallas, TX). JM109 bacterial strain was from
Gibco-BRL (Grand Island, NY). Terrific Broth, LB Agar,
IPTG, Ampicillin were from Gibco-BRL. ALA and EDTA
were from Sigma (St. Louis, MO). Protease Inhibitor
Cocktail was from Boehringer-Mannheim (Mannheim,
Germany). [21-3H]Hydroxypregnenolone (specific activ-
ity 13.61 mCi/mmol) was prepared in our laboratory as
described previously [5]. HPLC grade acetonitrile was pur-
chased from Sigma-Aldrich. The reference standards for
the inhibitors of androgen synthesis were ketoconazole, and
L-39 (1).
1H, 16-H), 5.33 (d, J = 5.4, 1H, 6-H), 3.50 (m, 1H, 3␣-H),
2.14 (s, 3H, 4ꢀ-CH3), 1.10 (s, 3H, 19-CH3), 1.07 (s, 3H,
18-CH3). Anal. C23H31O2N, C 78.15%, H 8.84%, N 3.96%;
found C 78.40%, H 8.87%, N 3.77%.
Acetate 13b (30 mg, 0.065 mmol) yielded 14b (20 mg,
73.5%), mp 142–144 ◦C (from acetone). 1H NMR: δ
7.75–7.83 (m, 2H, 5ꢀ-H, Ph–H), 7.30–7.48 (m, 3H, Ph–H),
6.65 (ws, 1H, 16-H), 5.38 (d, J = 5.1, 1H, 6-H), 3.55 (m,
1H, 3␣-H), 1.09 (s, 3H, 19-CH3), 1.08 (s, 3H, 18-CH3).
Anal. C28H33O2N, calculated C 80.93%, H 8%, N 3.37%;
found C 80.90%, H 8.20%, N 3.09%; MS (EI): 415 (M+).
The acetate 13c (30 mg, 0.078 mmol) gave the product 14c
(20 mg, 77%), mp 193.8–195 ◦C (from acetone). IR (cm−1):
1
3422 (OH), 1616, 1530 (oxazole). H NMR: δ 7.56 (ws,
1H, 5ꢀ-H), 7.13 (ws, 1H, 4ꢀ-H), 6.58 (ws, 1H, 16-H), 5.37
(d, J = 4.8, 1H, 6-H), 3.52 (m, 1H, 3␣-H), 1.08 (s, 3H,
19-CH3), 1.04 (s, 3H, 18-CH3). Anal. C24H31O3N, calcu-
lated C 77.84%, H 8.61%, N 4.12%; found C 77.45%, H
8.61%, N 3.90%; MS (EI): 355 (M+).
3.2. Bacterial stock preparation
JM109/H mod17His was incubated overnight in LBAG
medium (containing 20 mM glucose and 1 mg/ml of ampi-
cillin) and grown overnight at 37 ◦C with agitation at
250 rpm. Overnight cultures were diluted 1:100 in terrific
broth containing 100 mg/ml ampicillin (TBA) and allowed
to grow to OD600 = 0.8–0.9 at 37 ◦C for 4–5 h. Cytochrome
P45017␣ synthesis was stimulated by adding 1 mM of IPTG,
and heme precursor ALA. After incubation for 16–18 h at
27 ◦C with agitation at 150 rpm, bacteria were collected by
centrifugation at 2000 × g for 20 min at 4 ◦C. The pellet
was resuspended and washed with 10 volumes of buffer A
(20 mM glucose, 2 mM DTT, 1 mM EDTA, and 100 mM
MOPS). After centrifugation at 2000×g for 20 min at 4 ◦C,
the pellet was resuspended in buffer B (buffer A and protease
inhibitor cocktail) to 2 × 109 cells in 1 ml (OD600 = 0.6).
2.9.4. 3β-Acetoxy-17-(5ꢀ-methyl-2ꢀ-oxazolyl)-androsta-
5,16-diene (15)
The ␤-keto amide 9a (100 mg, 0.23 mmol) was added
into a flask containing a freshly prepared solution of
triphenylphosphine (118 mg, 0.46 mmol), iodine (117 mg,
0.46 mmol) and triethylamine (130 ul) in CH2Cl2 (5 ml).
The mixture was stirred at room temperature for 2 days.
The solution was diluted with CH2Cl2 (10 ml). The or-
ganic layer was washed with saturated aqueous NaHCO3
and dried over anhydrous Na2SO4. The solvent was evap-
orated and the residue was chromatographed. Elution with
PE/ethyl acetate (9:1) yielded 5ꢀ-methyloxazole 15 (20 mg,
21%), mp 183–184 ◦C (from ethyl acetate/PE). H NMR:
3.3. Measuring P45017α activity in cultured bacteria
1
δ 6.70 (s, 1H, 4ꢀ-H), 6.50 (t, 1H, 16-H), 5.40 (d, J = 4.8,
1H, 6-H), 4.60 (m, 1H, 3␣-H), 2.31 (s, 3H, 5ꢀ-CH3), 2.04
(s, 3H, AcO), 1.08 (s, 3H, 19-CH3), 1.02 (s, 3H, 18-CH3).
Substrate ([21-3H]hydroxypregnenolone (specific activ-
ity 13.61 mCi/mmol)) and different concentrations of in-

N. Zhu et al. / Steroids 68 (2003) 603–611
609
hibitors were mixed in the sample tube (13 mm × 100 mm
polystyrene tubes, Becton Dickinson, Lincoln Park, NJ) and
the solvent (ethanol) was evaporated under airflow. The con-
trol used in this experiment was the substrate and bacteria
without the presence of inhibitor.
by 17-acid chloride 3c affording the ␤-keto amide 9a
and 9b. Refluxing 9a with P2S5 in dioxane furnished the
5ꢀ-methylthiazole 10a in 30% yield. However, the same pro-
cedure only afforded a small amount of 5ꢀ-phenylthiazole
10b, and mainly resulted in the side-product 11. Its 1H
NMR showed a five-proton signal of 5ꢀ-phenyl around
7–8 ppm, the disappearance of the 16-ethylenic proton
signal at 6–7 ppm as well as the carbon signal peak
of ꢀ16 double bond in 13C NMR suggesting that the
ꢀ16 double bond had been saturated. Mass spectrome-
try showed the molecular-ion peak 507 and a fragment
474 (M+–SH), confirming that the structure was the
16-thiol-17-(5ꢀ-phenyl-2ꢀ-thiazolyl)-androstene derivative.
NOESY spectrometry showed that there were no NOE
between 17-H and 13␤-CH3, 16-H separately, confirm-
ing the configuration of 16␣-SH 11. The Michael addi-
tion of the SH− group at C-16 occurred preferentially at
the less steric crowded ␣ face owing to the hindrance of
13␤-methyl. When toluene was used to replace dioxane,
5ꢀ-phenylthiazole 10b was obtained albeit in lower (21%)
yield. This is probably because the SH anion barely exists
in the aprotic solvent toluene.
The reaction was started by adding 1 ml of prepared bac-
terial culture to each tube. Tubes were sealed and mixed
in a Dynal Sample Mixer (Dynal, Inc., Long Island, NY)
at 40 rpm for 5–6 h at room temperature. The reaction was
stopped by adding 1 ml of ACN, which precipitated bacte-
rial and medium proteins. After centrifugation at 2000 rpm
for 15 min at 4 ◦C, the supernatant was collected (∼1.9 ml),
transferred into borosilicate glass tubes, and the steroids
were extracted with 2 ml of chloroform at 4 ◦C for 30 min.
After 30 min, the tubes were centrifuged at 2000 rpm for
15 min at 4 ◦C. The aqueous phase which contains the
[3H]acetic acid, was collected (∼0.9 ml), and this was
mixed with 300 ␮l of 8.5% charcoal suspension for 30 min
at 4 ◦C. After centrifugation at 2000 rpm for 15 min at 4 ◦C,
the supernatant, which contained [3H]acetic acid, was re-
moved and radioactivity measured in a Liquid Scintillation
Analyzer (Packard Instrument Co., Meriden, CT).
For the preparation of 4ꢀ-substituted thiazoles, the conve-
nient one step procedure of Hantzsch [14] was employed,
i.e. 17-thiamide 5 was prepared in situ without separation.
Reaction of 17-amide 4 with P2S5 and then condensed with
␣-bromoacetophenone in dioxane gave 4ꢀ-methylthiazole 7a
(9% yield), together with the dehydrated product 17-cyanide
6. However, the same procedure did not give 4ꢀ-phenyl
7b and thiazole 7c. We finally found that the solution of
17-thiamide 5 formed in situ had to be filtered first to re-
move P2O5 and other precipitates, and then condensed with
2-bromoacetophenone. In this way, 4ꢀ-phenylthiazole 7b
was obtained in 10% yield.
Owing to the instability of ␣-bromoacetaldehyde which
always exists in polymer form, its more stable derivative
␣,␤-dichloroether [14] was used as reactant for the prepa-
ration of thiazole 7c. However, using the procedure for 7a
or 7b gave a very poor yield. Here, pure 17-thiamide 5 was
separated and then refluxed with ␣,␤-dichlorether in DMF,
to give the thiazole 7c (16% yield). It should be pointed out
that the preparation of 7a and 7b with pure thiamide 5 did
not increase the yield.
In summary, because of the existence of 16,17-double
bond, the activity of the conjugated 17-carbonyl group de-
creased, and the Michael addition reaction took place. This
made the reaction for synthesis of 17-thiazole more compli-
cated and the yield was generally low.
On the other hand, the synthesis of 17-oxazole deriva-
tives were much easier. 4ꢀ-Methyloxazole 13a was obtained
in 40% yield by the direct cyclization of 17-amide 4
with bromoacetone in boiling DMF. In the same manner,
2-bromoacetophenone gave 4ꢀ-phenyloxazole 13b in 68%
yield, and oxazole 13c was obtained with ␣,␤-dichloroether.
For the preparation of 5ꢀ-methyloxazole 15, the classic
Robinson–Gabriel cyclodehydration of ␤-keto amide 9a was
4. Results and discussion
4.1. Chemistry
The key starting material ꢀ16-17-carboxylic acid 3b could
be obtained by three synthetic routes shown in the Scheme 1,
as reported in the literature. The first route was the nu-
cleophilic addition of cyano anion to the 17-keto group
of epiandrostene, followed by dehydration and hydrolysis
[9,12]. This was not tried due to the difficulty of handling
toxic cyanide in the lab. The second route was the hydrol-
ysis of 21-pyridinium iodide (2a) [13]. Practically, 17-acid
3a, which was the starting material for finasteride, was pre-
pared in kg quantities by this route. We tried this proce-
dure and ꢀ16-21-pyridinium 2b was obtained in 90% yield,
but the hydrolysis of 2b to 3b only gave ꢀ16-17-acid 3b in
poor (20%) yield. The third method was the iodoform re-
action: the DPA 1b was treated with I2 in strong base and
gave 17-acid 3b directly, but the yield was less than 10%
[8]. When we used bromine to replace iodine to prepare hy-
pobromide instead of hypoiodide, the yield was increased
to 80%. This important improvement enabled us to obtain a
sufficient supply of 1a for the following steps. After 3-OH
was protected with the acetate group to give 3c, this was fur-
ther converted to key intermediates 17-acid chloride 3 and
17-amide 4.
Although the syntheses of 17␤-(2ꢀ-thiazolyl) derivatives
have been reported by Urbansky and Drasar [7], the synthesis
of their ꢀ16 counterpart was achieved with difficulty.
The 5ꢀ-methyl 10a and the 5ꢀ-phenylthiazole 10b
were synthesized by the Gabriel method. In the pres-
ence of triethylamine, the hydrochlorides of aminoace-
tone, 2-aminoacetophenone were acylated, respectively,

610
N. Zhu et al. / Steroids 68 (2003) 603–611
Table 1
<10%) were also less potent than the 3-OH counterparts,
as P45017␣ preferred 3-OH derivatives as substrate. The
introduction of a phenyl group at either the 4ꢀ-ring posi-
tion (8b, 26.7% and 14b, 26.2%) or 5ꢀ-ring position (12b,
<10%) decreased inhibition substantially. Surprisingly, al-
though the introduction of a methyl group at the 4ꢀ-ring
position decreased inhibition (8a, 13.2% and 14a, 19.4%),
17-(5ꢀ-methyl-2ꢀ-oxazolyl) (16, 45.0%) still retained the ac-
tivity to its parent compound 17-(2ꢀ-oxazolyl) (14c, 56.0%).
The 17-(5ꢀ-methyl-2ꢀ-thiazolyl) (12a, 72.1%) had similar
activity to ketoconazole and was the most potent inhibitor
among this series. Compound 12a was almost as potent
as L-39, which has good antitumor activity in preclinical
models and is scheduled for Phase I clinical trials.
Percent inhibition of recombinant human P45017␣-expressed in E. coli
by the steroids (␮M)
Compound
Inhibition %
7a
8a
7b
8b
7c
8c
10a
10b
12a
12b
13a
14a
13b
14b
13c
14c
15
27.6
13.2
<2
26.7
<10
41.5
67.0
0
72.1
<10
<10
19.4
28.6
26.2
25.8
56.0
19.9
45.0
100.0
116.7
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (No. 30171110) and NIH Grant CA
2774.
16
Ketoconazole
L-39
The assay was performed in triplicate as described in Section 2 and was
repeated on two occasions. Results are expressed as percent of inhibition
relative to ketoconazole. Ketoconazole was used as an assay control and
all values for the test compounds expressed relative to the activity of
ketoconazole.
References
[1] Njar VCO, Brodie AMH. Inhibitors of 17␣-hydroxylase/17,20-lyase
(CYP17): potential agents for the treatment of prostate cancer. Curr
Pharm Des 1999;5:163–80.
employed. However, all the dehydrating agents including
concentrated sulfuric acid, PCl5, P2O5, polyphosphoric acid
and acetic acid anhydride [15] failed to give 9a. A newer
method [16] with mild conditions, i.e. the use of triph-
enylphorphine and iodine in triethylamine as dehydrating
reagent was tried resulting in 21% yield of 5ꢀ-methyloxazole
15. The freshly distilled triethylamine was crucial for the
success of this reaction. However, the same method failed
to give 5ꢀ-phenyloxazole.
[2] Trachtenberg J. Ketoconazole therapy in advanced prostatic cancer.
J Urol 1984;132:61–3.
[3] Williams G, Kerle DJ, Doble A, Dunlop H, Smith C, Allen J, et al.
Objective responses to ketoconazole therapy in patients with relapsed
progressive prostatic cancer. Br J Urol 1986;58:45–51.
[4] Ling YZ, Li JS, Liu Y, Kato K, Klus GT, Brodie AMH.
17-Imidazolyl, pyrazolyl, and isoxazolyl androstene derivatives.
Novel steroidal inhibitors of human cytochrome
(P45017␣). J Med Chem 1997;40:3297–304.
C17,20-lyase
[5] Njar VCO, Kato K, Nnane IP, Grigoryev DM, Long BJ, Brodie AMH.
Novel 17-azolyl steroids, potent inhibitors of human cytochrome
17␣-hydroxylase-C17,20-lyase (P45017␣): potential agents for the
treatment of prostate cancer. J Med Chem 1998;41:902–12.
[6] Jarman M, Barrie SE, Llera JM. The 16,17-double bond is
needed for irreversible inhibition of human cytochrome p45017␣
by abiraterone (17-(3-pyridyl)androsta-5,16-diene-3␤-ol) and related
steroidal inhibitors. J Med Chem 1998;41:5375–81.
All of the above 3␤-acetate compounds were hydrolyzed
in methanolic potassium hydroxide affording their 3␤-ol
counterparts. These were more potent inhibitors of P45017␣
.
4.2. Structure–activity relationship
[7] Urbansky M, Drasar P. Simple syntheses of steroidal 17␤-
(2ꢀ-thiazolyl) derivatives. Syn Commun 1993;23:829–45.
The percentage inhibition of P45017␣ by the different
compounds synthesized relative to ketoconazole is shown
in Table 1. Since the latter is currently the only com-
pound used clinically to inhibit this enzyme, we have
compared the activities of our compounds to that of ke-
toconazole. The activity of the compounds as inhibitors
of P45017␣ was determined using recombinant human
P45017␣ expressed in E. coli [17]. The results showed that,
the compounds containing 17-(2ꢀ-oxazolyl) (14c, 56.0%
inhibition) and 17-(2ꢀ-thiazolyl) (8c, 23.7%) demonstrated
inhibition against P45017␣, although they were less po-
tent than ketoconazole. They were also less potent than
L-39 which had greater activity than ketoconazole in the
assay. The 3-acetate derivatives (e.g. 13c, 25.8% and 7c,
[8] Marker RE, Wagner RB. The hypoiodite oxidation of pregnanolones
and pregnenolones. J Am Chem Soc 1942;64:1842–3.
[9] Ruzicka L, Hardegger E, Kauter C. Uber die
ꢀ
5,16-3␤-oxy-
ätiocholadiensäure und einige ihrer umwandlungsprodukte. Helv
Chim Acta 1944;27:1164.
[10] Preparation of bromoacetone. Org Syn Coll 2:88.
[11] Hepworth JD. Aminoacetone semicarbazone hydrochloride. Org
Synth Ol 45:1–3.
[12] Butenandt A, Schmidt-Thomé J. Überführung von dehydro-
androsteron in 3-aceoxy-ꢀ5-aetiocholensäure: ein Beitrag zur
verknüpfung der Androsteron mit der corticosteron-Gruppe. Ber
1938;71:1487–9.
[13] Carroll L. Preparation of 21-pyridinium-3-␤-hydroxy-5-pregnene-20-
one halides and 3-␤-hydroxy-5-androstene-17-carboxylic acid. J Am
Chem Soc 1944;66:1612.

N. Zhu et al. / Steroids 68 (2003) 603–611
611
[14] Vernin G. General synthetic methods for thiazole and thiazolium
salts. In: Metzger JV, editor. Thiazole and its derivatives. New York:
Interscience Publication; 1979. p. 169–209.
[15] Turchi IJ. Synthesis and reactions of functionalized oxazoles.
In: Metzger JV, editor. Oxazole and its derivatives. New York:
Interscience Publication; 1979. p. 3.
[16] Wipf P, Miller CP. A new synthesis of highly functionalized oxazoles.
J Org Chem 1993;58:3604–6.
[17] Grigoryev DN, Kato K, Njar VCO, Long BJ, Ling YZ, Wang X, et
al. Cytochrome P450c 17-expressing Escherichia coli as a first-step
screening system for 17␣-hydroxylase-C17,20-lyase inhibitors. Anal
Biochem 1999;267:319–30.