Synthesis, biochemical evaluation and rationalisation of the inhibitory activity of a range of 4-substituted phenyl alkyl imidazole-based inhibitors of the enzyme complex 17α-hydroxylase/17,20-lyase (P45017α)

Article abstract of DOI:10.1016/j.bmcl.2006.06.092

We report the preliminary results of the synthesis, biochemical evaluation and rationalisation of the inhibitory activity of a number of phenyl alkyl imidazole-based compounds as inhibitors of the two components of 17α-hydroxylase/17,20-lyase (P450_17α), that is, 17α-hydroxylase (17α-OHase) and 17,20-lyase (lyase). The results show that N-3-(4-bromophenyl) propyl imidazole (12) (IC₅₀ = 2.95 μM against 17α-OHase and IC₅₀ = 0.33 μM against lyase) is the most potent compound within the current study, in comparison to ketoconazole (KTZ) (IC₅₀ = 3.76 μM against 17α-OHase and IC₅₀ = 1.66 μM against lyase). Modelling of these compounds suggests that the length of the alkyl chain enhances the interaction between the inhibitor and the area of the active site corresponding to the C(3) area of the steroid backbone, thereby increasing potency.

Full text of DOI:10.1016/j.bmcl.2006.06.092

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4752–4756
Synthesis, biochemical evaluation and rationalisation
of the inhibitory activity of a range of 4-substituted phenyl
alkyl imidazole-based inhibitors of the enzyme complex
17a-hydroxylase/17,20-lyase (P45017a)
Chirag H. Patel,a Sachin Dhanani,b Caroline P. Owena and Sabbir Ahmeda,*
aDepartment of Pharmacy, School of Pharmacy and Chemistry, Kingston University, Penrhyn Road,
Kingston upon Thames, Surrey, KT1 2EE, UK
bSchool of Life Sciences, Kingston University, Penrhyn Road, Kingston upon Thames, Surrey, KT1 2EE, UK
Received 22 May 2006; revised 30 June 2006; accepted 30 June 2006
Available online 25 July 2006
Abstract—We report the preliminary results of the synthesis, biochemical evaluation and rationalisation of the inhibitory activity of
a number of phenyl alkyl imidazole-based compounds as inhibitors of the two components of 17a-hydroxylase/17,20-lyase
(P45017a), that is, 17a-hydroxylase (17a-OHase) and 17,20-lyase (lyase). The results show that N-3-(4-bromophenyl) propyl imidaz-
ole (12) (IC50 = 2.95 lM against 17a-OHase and IC50 = 0.33 lM against lyase) is the most potent compound within the current
study, in comparison to ketoconazole (KTZ) (IC50 = 3.76 lM against 17a-OHase and IC50 = 1.66 lM against lyase). Modelling
of these compounds suggests that the length of the alkyl chain enhances the interaction between the inhibitor and the area of
the active site corresponding to the C(3) area of the steroid backbone, thereby increasing potency.
ꢀ 2006 Elsevier Ltd. All rights reserved.
The enzyme complex 17a-hydroxylase/17,20-lyase
(P45017a) is a pivotal enzyme in the conversion of prog-
estins (such as progesterone and pregnenolone) to the
androgens1 (androstenedione and dehydroepiandroster-
one, respectively, Fig. 1) and requires both NADPH and
oxygen in the sequential oxidative steps.2
cade leading to the biosynthesis of the sex hormones,
glucocorticoids and mineralocorticoids, the latter two
series of compounds being produced directly from the
17a-hydroxy progestins. This enzyme has become
the focus of attention as it may have a role to play in
the treatment of hormone-dependent prostate cancer
through the overall reduction in the biosynthesis of
androgens. Here, we report: the synthesis of a range of
imidazole-based compounds (which have previously
been evaluated as potential inhibitors of the enzyme aro-
matase by Ahmed et al;4 as well as potential hypolipi-
The enzyme catalyses both the hydroxylation of the pro-
gestin backbone, via 17a-hydroxylase (17a-OHase), fol-
lowed by the cleavage of the C(17) and C(20) bond, via
17,20-lyase (lyase), both of which are believed to be
undertaken within an active site that possesses two bind-
ing sites.3 However, little specific information is known
about the active site of this enzyme as no crystal struc-
ture for it exists, although workers have utilised homol-
ogy modelling in an attempt to discover detailed
information regarding the active site.3 P45017a is there-
fore responsible for the second step in the steroidal cas-
demic
agents5);
their
biochemical
evaluation
[in comparison to ketoconazole (KTZ)], and the ratio-
nalisation of their inhibitory activity using the sub-
strate–haem complex (SHC) approach.
The general approach for the construction of the SHC
has been described previously6–9 and will therefore not
be detailed here. In general, however, the structures of
progesterone, 17a-hydroxyprogesterone, the haem and
potential hydrogen bonding groups were all constructed
and refined by performing a pre-optimization calcula-
tion in mechanics using Augmented MM2 as imple-
mented in the molecular modelling program CaChe.10
Keywords: Imidazole; 17a-Hydroxylase/17,20-lyase; Inhibitors; Pros-
tate cancer.
*
Corresponding author. Tel.: +44 20 8547 2000; fax: +44 20 8547
7562; e-mail: S.AHMED@KINGSTON.AC.UK
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.06.092

C. H. Patel et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4752–4756
4753
O
O
O
OH
lyase
17α-OHase
O
O
O
Figure 1. Action of the overall enzyme complex on progesterone resulting in the subsequent biosynthesis of androstenedione.
The atoms of the SHC were then ‘locked’ and the pro-
posed inhibitors attached to the SHC representing the
active site of the overall enzyme complex resulting in
the production of the enzyme–inhibitor complex, in
the attachment of the inhibitors, interactions at both
the haem (N–Fe dative covalent bond) and potential po-
lar–polar interactions with groups at the active site were
mimicked. The enzyme–inhibitor complex was mini-
mised and optimised so as to produce the final binding
conformer of the inhibitor within the representation of
Similarly, in the synthesis of 4-halogen-substituted phenyl
propyl bromide, only the appropriate cinnamic acid
derivatives were commercially available and the reactions
outlined in Scheme 3 were undertaken in the synthesis of
the range of 4-halogen substituted phenyl propyl bro-
mides. No substituted phenyl alkyl bromides (or deriva-
tives) were found to be commercially available for
compounds containing spacer groups (between the substi-
tuted phenyl ring and the azole moiety) greater than C3,
and the synthesis of these compounds is currently under
investigation within our laboratories.
the overall active site of P45017a
.
In the synthesis of the proposed inhibitors, the azole
functionality was reacted with a phenyl alkyl halide in
the presence of a suitable base (Scheme 1); the synthesis
of N-1-phenylethyl-imidazole (2) is given as an exam-
ple.11 In general, the compounds were synthesised in
good yield (ranging from 45% to 80% yield) and without
any major problems.
The synthesised compounds were then initially screened
and IC50 values determined against both 17a-OHase and
lyase using modified literature methods,12–15 where a
modified mobile phase to that used by Li et al.12 in
the separation and identification of the radiolabelled
substrate from the radiolabelled products was used.
Table 1 shows the IC50 values obtained for the inhibi-
tion of both 17a-OHase and lyase by the synthesised
compounds; in general, most of the compounds pos-
sessed weaker inhibitory activity than KTZ. In particu-
lar, consideration of the inhibitory data shows that the
compounds based upon the propyl chain possessed good
inhibitory activity in comparison to the shorter chain-
containing compounds, such as those based on the ben-
zyl backbone (e.g., 1, 4, 7 and 10). The most potent
inhibitor was found to be 4-bromophenyl propyl imid-
azole (12) (IC50 = 2.95 lM against 17a-OHase and
IC50 = 0.33 lM against lyase) which is slightly more
potent against 17a-OHase and ꢀ5 times more potent
against lyase, in comparison to the standard compound
KTZ (IC50 = 3.76 lM against 17a-OHase and
IC50 = 1.66 lM against lyase).
However, in the synthesis of the majority of the long
chain-containing compounds (n > 1; X = F, Cl, and Br),
the 4-substituted phenyl alkyl bromides were not readily
available, and as such, were required to be synthesised.
For example, 2-(4-bromophenyl)-ethyl bromide was syn-
thesised from 2-(4-bromophenyl)-acetic acid (Scheme 2).
N
Br
N
(CH2)n
(CH2)n
a
X
X
Scheme 1. Synthesis of potential inhibitors of P45017a. Reagents: (a)
imidazole/K2CO3/THF/D. (n = 1–3; X = H, F, Cl and Br).
Further consideration of the inhibitory activity (Table 1)
shows that within the series of compounds considered, a
trend is observed where the IC50 value decreases with an
increase in the alkyl spacer group. Consideration of the
fluoro derivatives shows that compound 4 (n = 1, Table
1) is found to possess IC50 values of 96.46 and 11.26 lM
against 17a-OHase and lyase, respectively, whereas
compound 6 (n = 3, Table 1) has IC50 values of 27.81
and 1.96 lM against 17a-OHase and lyase, respectively.
On modelling these compounds within the SHC, we ob-
served that the increase in chain length of the spacer
group allows for increased interaction between the
substituted phenyl moiety and the area of the enzyme
active site corresponding to the C(3) area of the steroid
backbone (Fig. 2), thereby presumably leading to the
formation of a more stable enzyme–inhibitor complex,
and more potent inhibitory activity.
OH
OH
a
O
X
X
b
Br
X
Scheme 2. Synthesis of 2-(4-substituted phenyl)-ethyl bromide (X = F,
Cl and Br). Reagents: (a) LiAlH4/THF/D; (b) PBr3/THF/D.

4754
C. H. Patel et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4752–4756
O
O
O
a
b
OH
OEt
OEt
X
X
X
c
d
Br
OH
X
X
Scheme 3. Synthesis of 3-(4-substituted phenyl)-propyl bromide (X = H, F, Cl and Br). Reagents: (a) EtOH/D/H+; (b) H2/Pd/C; (c) LiAlH4/THF/D;
(d) PBr3/THF/D.
Table 1. Table to show the percentage inhibition and the IC50 values obtained for a range of imidazole-based compounds against 17a-OHase and
lyase (the values are means of three determinations, n = 9; ND = not determined)
(CH2)n
N
N
X
X
n
Compound % inhibition ([I] = 10 lM) 17a-OHase IC50 values (lM) % inhibition ([I] = 10 lM) Lyase IC50 values (lM)
H
H
H
F
1
2
3
1
2
3
1
2
3
1
2
3
—
1
2
13.51 0.65
10.56 1.76
23.35 0.97
20.47 1.79
22.48 1.84
40.77 0.53
45.70 0.98
28.42 2.92
71.00 0.53
49.29 1.97
37.54 0.60
70.15 0.94
61.54 1.53
154.20 7.93
ND
12.64 0.65
6.36 0.65
39.95 0.89
50.98 1.36
45.07 1.21
73.15 0.33
73.65 0.99
57.38 2.50
80.85 1.28
76.25 0.65
61.52 2.49
86.29 0.3
78.69 2.44
50.90 0.86
ND
3
30.95 0.68
96.46 0.20
120.20 7.74
27.81 1.44
29.84 0.27
49.64 1.48
5.85 0.19
16.55 0.23
30.66 0.13
2.95 0.03
3.76 0.01
6.14 1.21
11.26 0.22
ND
4
F
5
F
6
1.96 0.01
4.94 0.17
ND
Cl
Cl
Cl
Br
Br
Br
—
7
8
9
0.55 0.07
2.85 0.08
ND
10
11
12
KTZ
0.33 0.02
1.66 0.15
With regard to the ethyl spacer group-containing com-
pounds, it should be noted that these compounds were
synthesised and fully characterised [the synthesis of N-
1-phenylethyl-imidazole (2) is given as an example11],
however, upon evaluating these compounds for inhibito-
ry activity, we discovered that they possessed weaker
inhibitory activity than the compounds based on the
benzyl backbone. We undertook analysis of the com-
pounds (involving GC–MS and thin-layer chromatogra-
phy) and discovered that these compounds, although
initially showing good purity (as evidenced by GC–MS
analysis) prior to initial screening, had undergone
decomposition prior to the determination of their IC50
values. As such, the IC50 values for compounds 2, 5, 8
and 11 against the lyase component were not
determined.
hydrogen-bonding group at the active site [which would
normally undergo H-bond interaction with the C(3)@O
moiety within the natural substrate]. This potential
interaction between the heteroatom on the phenyl ring
and groups within the active site of P45017a has previ-
ously been suggested by other workers16,17 and by
Ahmed (involving the molecular modelling study of a
range of imidazole-based antimycotic compounds as
inhibitors of P45017a).18 We therefore hypothesise that
similar interaction is possible, however, we propose that
the bromine atom is able to interact with the group(s) at
the active site more effectively than either Cl or F due to
the increased electronegativity of the latter two atoms,
thereby resulting in weaker binding of the inhibitor
within the active site of the enzyme and poorer inhibito-
ry activity. As such, compound 12 is found to possess
not only the appropriate alkyl spacer group (n = 3) but
also a 4-bromo-substituted aromatic ring system allow-
ing this compound to interact favourably within the
enzyme active site. As such, these compounds are excel-
lent lead compounds in the search for more potent and
specific inhibitors of P45017a. Furthermore, the ability of
these compounds to inhibit the lyase step in preference
to the hydroxylation step is also a major advantage since
the inhibition of the 17a-OHase activity may result in
From the inhibitory activity data outlined in Table 1, it
can be seen that within the substituted series of com-
pounds, the bromo-derivatives are more potent than
the chloro-derivatives, which are in turn more potent
than the fluoro-derivatives and the non-substituted
derivatives. We hypothesise that the trend observed is
presumably due to the ability of the substituted atom
being able to undergo polar–polar interaction with a

C. H. Patel et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4752–4756
4755
3. Laughton, C. A.; Neidle, S.; Zvelebil, M. J. J. M.;
Sternberg, M. J. E. A. Biochem. Biophys. Res. Commun.
1990, 171, 1160.
4. Ahmed, S.; Smith, J. H.; Nicholls, P. J.; Whomsley, R.;
Cariuk, P. Drug Des. Discovery 1995, 13, 27.
5. Baggaley, K. H.; Heald, M.; Hindley, R. M.; Morgan, B.;
Tee, J. L.; Green, J. J. Med. Chem. 1975, 18, 833.
6. Ahmed, S. Bioorg. Med. Chem. Lett. 1995, 5, 2795.
7. Ahmed, S.; Davis, P. J. Bioorg. Med. Chem. Lett. 1995, 5,
2789.
8. Ahmed, S. J. Enzyme Inhib. 1997, 12, 59.
9. Ahmed, S. Biochem. Biophys. Res. Commun. 2000, 275, 75.
10. CaChe is a trademark of Oxford Molecular Ltd, Oxford
Science Park, Oxford, UK.
11. N-1-Phenylethyl-imidazole (2): imidazole (2 g, 29.4 mmol)
was added to anhydrous potassium carbonate (K2CO3)
(1.02 g, 7.34 mmol) and anhydrous tetrahydrofuran
(THF) (50 mL). The mixture was stirred at room temper-
ature for 10 min prior to the addition of phenylethyl
bromide (2.72 g, 14.7 mmol). The mixture was then stirred
under reflux for 24 h. After filtration, the THF was
removed under vacuum to leave a yellow solid which was
dissolved in dichloromethane (DCM) (40 mL) and washed
with water (3· 50 mL). The organic layer was then
extracted using hydrochloric acid (HCl) (2 M, 3· 30 mL)
followed by water (2· 50 mL). The combined acid layer
was neutralised with solid saturated sodium bicarbonate
(NaHCO3) and then extracted into DCM (2· 40 mL). The
combined DCM layer was washed with water (3· 50 mL),
dried over anhydrous magnesium sulfate (MgSO4) and
filtered. Removal of DCM under vacuum gave 2 as a
yellow oil (0.99 g, 39%). m(max) (film) cmÀ1: 3387.3 (NCN
imidazole), 3031.7 (CH aromatic), 1605.1 (C@C aromat-
ic); dH (300 MHz, CDCl3): 7.31 (1H, s, NCHN imidazole),
7.21 (5H, m, Ph-H), 6.82 (1H, s, NCH imidazole), 4.16
(2H, t, J = 7 Hz, Ph-CH2), 3.13 (2H, t, J = 7Hz, NCH2);
dC (75 MHz, CDCl3): 137.45 (NCN), 135.00 (ImC),
129.43, 128.77, 128.59, 126.99, (ArC), 118.77 (ImC),
48.52 (Ph-CH2), 37.86 (NCH2); GCMS tR 8.21 min m/z
172 (M+), 54 (base peak).
Figure 2. Binding of compound 6 to the overall SHC for P45017a
.
significant side-effects due to an effect on corticosteroid
biosynthesis.19
In conclusion, the SHC approach has allowed the design
and subsequent synthesis of a small range of potent
inhibitors of this enzyme complex. Whilst the com-
pounds within the current study have been shown, in
general, to possess weaker inhibitory activity against
the lyase components of the overall enzyme complex
of P45017a in comparison to the standard compound
KTZ, all are significantly weaker inhibitors against the
17a-OHase component. The greater selectivity is seen
for compounds 6 and 9 which show over ꢀ14- and
ꢀ10-fold differences, respectively, in activity between
the two components. They are therefore extremely good
lead compounds in the design and synthesis of more
12. Li, J. S.; Li, Y.; Son, C.; Brodie, A. M. H. J. Med. Chem.
1996, 39, 4335.
13. 17a-OHase assay. Rat testicular microsomal suspension
was thawed under cold running water and vortexed. The
final incubation assay mixture (1 mL) consisted of sodium
phosphate buffer (50mM, pH 7.4, 905 ll), radiolabelled
progesterone as substrate (1.5 and 15 lL), NADPH
generating system (50 lL) and inhibitor (10 and 20 lL).
Tubes were warmed to 37 ꢁC for 5 min and the assay
initiated by the addition of microsomal enzyme (final
concentration 0.16 mg/mL, 10 lL). The assay mixture was
incubated for 15 min. The reaction was quenched by the
addition of ether (2 mL), vortexed and placed on ice.
The organic layer was then placed into a separate tube.
The assay mixture was further extracted with ether
(2· 2 mL), and the organic layers were combined. The
solvent was removed under a stream of nitrogen, acetone
(30 lL) was added to each tube and the solution spotted
onto silica-based TLC plates along with carrier steroids
(progesterone, 17a-hydroxyprogesterone, testosterone and
androstenedione, 5 mg/mL). The TLC plates were devel-
oped using the mobile phase DCM/ethylacetate (7:3). The
separated spots were identified under UV light and each
spot cut out and placed into scintillation vials. Acetone
(1 mL) and scintillation fluid (HiSafe) (3 mL) were added
potent inhibitors of P45017a
.
Acknowledgments
The authors thank the EPSRC National Mass Spec-
trometry service at the University of Wales College
Swansea (UK) and the elemental analysis service at
the School of Pharmacy, University of London (UK),
for the provision of high resolution and elemental anal-
ysis data, respectively.
References and notes
1. Ortiz de Montellano, P. R. In Cytochrome P-450: Struc-
ture Mechanism and Biochemistry; Ortiz de Montellano, P.
R., Ed.; Plenum Press: New York, 1986; pp 217–272.
2. Robichaud, P.; Wright, J. N.; Akhtar, M. J. Chem. Soc.,
Chem. Commun 1994, 12, 1501.
3
to each vial, vortexed and counted for 3 min for H.
14. Lyase assay. Rat testicular microsomal suspension was
thawed under cold running water, and vortexed. The final
incubation assay mixture (1 mL) consisted of sodium

4756
C. H. Patel et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4752–4756
phosphate buffer (50 mM, pH 7.4, 905 lL), radiolabelled
and placed into scintillation vials. Acetone (1 mL) and
scintillation fluid (Cocktail T) (3 mL) were added to each
vial, vortexed and counted for 3 min for H.
17a-hydroxyprogesterone as substrate (1 and 10 lL),
NADPH generating system (50 lL) and inhibitor (10
and 20 lL). Tubes were warmed to 37 ꢁC for 5 min and the
assay initiated by the addition of microsomal enzyme
(final concentration 0.23 mg/mL, 15 lL). The assay mix-
ture was incubated for 30 min. The reaction was quenched
by the addition of ether (2 mL), vortexed and placed on
ice. The organic layer was then removed and placed into a
separate tube. The assay mixture was further extracted
with ether (2· 2mL), and the organic layers combined. The
solvent was removed under a stream of nitrogen, acetone
(30 lL) was added to each tube and the solution spotted
onto silica-based TLC plates along with carrier steroids
(17a-hydroxyprogesterone, testosterone and androstene-
dione, 5 mg/mL). The TLC plates were developed using
the mobile phase DCM/ethylacetate (4:1). The separated
spots were identified under UV light and each spot cut out
3
15. IC50 determination. In determining the IC50 values for the
most potent compounds, the inhibitory activity was
determined using the method outlined above, however,
for each compound, five or more inhibitor concentrations
were used and the inhibitory activity determined at each
concentration (in triplicate); the IC50 was then determined
from a graph (using linear regression analysis) of the
inhibitory activity versus log[I].
16. Wachall, B. G.; Hector, M.; Zhuang, Y.; Hartmann, R.
W. Bioorg. Med. Chem. 1999, 7, 1913.
17. Leroux, F.; Hutschenreuter, T. U.; Charriere, C.; Scopel-
liti, R.; Hartmann, R. W. Helv. Chim. Acta 2003, 86, 2671.
18. Ahmed, S. Drug Des. Discovery 1994, 12, 77.
19. Njar, V. C. O.; Brodie, A. M. H. Curr. Pharm. Des. 1999,
5, 163.