Activity of steroid 4 and derivatives 4a–4f as inhibitors of the enzyme 5α-reductase 1

Article abstract of DOI:10.1016/j.bmc.2018.06.030

It is known that the growth of prostate metastatic bone tumor depends on androgens, and tumor formation can start from migratory malignant cells produced in that organ. These cells exhibit grater type 1 5α-reductase (5α-R1) activity than type 2 5α-reducta

Full text of DOI:10.1016/j.bmc.2018.06.030

Accepted Manuscript
Activity of steroid 4 and derivatives 4a-4f as inhibitors of the enzyme 5α-re-
ductase1
Yazmín Arellano, Eugene Bratoeff, Yvonne Heuze, Marisol Bravo, Juan
Soriano, Marisa Cabeza
PII:
S0968-0896(18)30453-X
https://doi.org/10.1016/j.bmc.2018.06.030
BMC 14422
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
5 March 2018
20 June 2018
23 June 2018
Please cite this article as: Arellano, Y., Bratoeff, E., Heuze, Y., Bravo, M., Soriano, J., Cabeza, M., Activity of
steroid 4 and derivatives 4a-4f as inhibitors of the enzyme 5α-reductase1, Bioorganic & Medicinal Chemistry (2018),
doi: https://doi.org/10.1016/j.bmc.2018.06.030
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1
Activity of steroid 4 and derivatives 4a-4f as inhibitors of the enzyme 5α-
reductase1
Yazmín Arellanoa, Eugene Bratoeffa,†, Yvonne Heuzeb, Marisol Bravob; Juan Sorianoc and
Marisa Cabezab,*
aDepartamento de Farmacia, Facultad de Química, Universidad Nacional Autónoma de
México, Av. Universidad 3000, Copilco Universidad, Coyoacán, 04510 Ciudad de México,
CDMX, México. bDepartamento de Sistemas Biológicos y de Producción Agrícola y
Animal, Universidad Autónoma Metropolitana-Xochimilco, Calzada del Hueso 1100. Col.
Villa Quietud, 04960 Ciudad de México. México. cDepartamento de Patología, Hospital
General de México, Dr. Balmis 148, CDMX (Department of Pathology of General Hospital
of México (SS)c, CDMX, México).
Running Title: Inhibitors of 5α-reductase
YA: yaz2372@yahoo.com.mx; MB: braru27@gmail.com; YH: ymheuze@gmail.com; JS:
juansorianor@hotmail.com; MC: marisa@correo.xoc.uam.mx.
Address correspondence to:
*Marisa Cabeza, Ph.D.
Departamento de Sistemas Biológicos
Universidad Autónoma Metropolitana-Xochimilco
Calzada del Hueso # 1100
CDMX, CP 04960, México
Phone: +52-555483-7260
Fax: +52-555483-7260
E-mail: marisa@correo.xoc.uam.mx
†Eugene Bratoeffa - Deceased

2
Keywords: Type 1 5α-reductase inhibitors; prostate cancer; 17-(1H-benzimidazol-1-yl)-16-
formylandrosta-5,16-dien)-3β-hydroxy derivatives; flank organs; weight of hamster
prostate.
List of Abbreviations:
Testosterone (T)
Dihydrotestosterone (DHT)
17-(1H-benzimidazol-1-yl)-16-formillandrosta-5,16-dien-3β-yl benzoate (4a-f)
17-(1H-benzimidazol-1-yl)-3-hydroxy-16-formylandrost-5,16-diene (4)
5α-reductase 1 (5α-R 1)
5α-reductase 2 (5α-R 2)
Finasteride (F)
Dimethylsulfoxide (DMSO)
Nicotinamide adenine dinucleotide phosphate reduced (NADPH)
Labeled testosterone ([1,2,6,7-3H]T)

3
Abstract
It is known that the growth of prostate metastatic bone tumor depends on androgens,
and tumor formation can start from migratory malignant cells produced in that organ. These
cells exhibit grater type 1 5α-reductase (5α-R 1) activity than type 2 5α-reductase.
Noteworthy, both isozymes convert testosterone (T) to the more active androgen
dihydrotestosterone (DHT) in the target tissues.
Thus, in order to potentially improve the prognosis of this disease, in this work,
seven derivatives of 17-(1H-benzimidazol-1-yl)-16-formillandrosta-5,16-dien-3β-yl
benzoate (4a-f) and 17-(1H-benzimidazol-1-yl)-3-hydroxy-16-formylandrost-5,16-diene (4)
were synthesized, characterized and identified as inhibitors of type 1 5α-reductase (5αR1).
These derivatives having the advantage of improved plasma half-life.
The inhibitory activity of the compounds towards 5α-R1 isoenzyme was determined
by conversion of T into DHT in the presence or absence of compounds 4, 4a-f. Further, in
vivo experiments were also carried out, treating gonadectomized hamsters with T and/or 4,
4a-f and evaluating their effect on the diameter of hamster flank organs and on the weight
of the prostatic and seminal vesicles. Results indicated that compounds 4, 4b, 4c, served as
in vitro inhibitors of the enzyme 5α-R1 and pharmacological experiments showed that 4
and derivatives 4a-f decreased the diameter of the flank glands, the weight of the prostate
and seminal vesicles of treated hamsters without any appreciable toxicity during
observation. Noteworthy the fact that compound 4 is the product, in all cases, of the
hydrolysis of the series of esters 4a-f, thus they can serve as precursors (prodrugs) of the
active form 4.

4
1. Introduction
Nowadays cancer has become a global health problem. Among the different types of
cancer, prostate cancer is one of the main causes of death worldwide in men and thus
improved therapies are a current topic of research around the world. In this sense, our
research group and others have been involved in the synthesis of steroid derivatives for
improved therapies of metastatic prostate cancer. [1-7] Metastatic bone tumors start from
malignant prostate cancer cells, leading eventually to death of patients that develop it. With
over 300,000 deaths and more than 1.1 million new cases annually, this disease is the
second most common cancer among men worldwide. [8]
Cells of metastatic tumors show higher preference for 5α-reductase 1 (5α-R 1)
activity rather than that of the predominant isoenzyme 5α-R 2 in normal prostate tissue. [9]
Both isozymes reduce testosterone (T) to the more active androgen dihydrotestosterone
(DHT), that binds to the androgen receptor in the cell cytoplasm and forms a dimer that
translocates to the cell nuclei inducing synthesis of new proteins that affect different
cellular functions. [10-12] It has been reported that 5α-R 1 activity levels may be correlated
with risk of prostate cancer, thus, specific 5α-R 1 steroidal inhibitors may lead to a better
therapy for this metastatic cancer, with fewer adverse effects than current treatments for
this disease. These steroids act as natural substrates to block the enzyme, [3, 14] increasing
their activities when a bulky benzyl group is included in the main backbone of the steroid,
reducing the probability of hydrolysis and thus increasing its half-life.
Nevertheless, esterases present in the liver or plasma will ultimately disrupt this
ester linkage, producing a highly active compound as inhibitor of the 5α-R 1 enzyme in the
target cells. Thus, the present research focus on the synthesis of a series of 17-(1H-
benzimidazole-1-yl)-16-formylandrosta-5,16-dien-3β-yl benzoate derivatives (4 and 4a-4f)

5
as potential precursors (prodrugs) of 17-(1H-benzimidazol-1-yl)-3-hydroxy-16-
formylandrosta-5,16-diene of the true active inhibitor of type 1 5α-reductase and the
exploration their pharmacological effect.
2. Results
2.1. Synthesis of compounds 4a-4f.
Compounds 4 [14], 2a-2f and 3a-3f [3] were synthesized according to previously
reported procedures. Further functionalization of 2a-2f and 3a-3f was performed to include
a benzimidazole moiety at the 17 position, however keeping the formyl group with the
double bond at C-16 to achieve a better selectivity to inhibit type 1 5α-R without affecting
the action of the isoenzyme type 2. It is noteworthy the fact that derivatives 4c, 4e and 4f
required double molar amounts of benzimidazole to give place to the desired products. All
reactions proceeded in short reaction times, producing the series of complexes in good
yields and purities after a simple recrystallization.
In addition, we believe that the series of compounds 4a-4f slowly hydrolyze in
aqueous solution (Figure 1). Moreover, this process certainly can proceed faster in plasma,
catalyzed by the presence of esterases. Thus, in order to probe this hypothesis, biological
tests were performed showing compounds 4b and 4c to only exhibit weak in vitro effects.
Conversely, steroids 4a-4f exhibited significant pharmacological activity (Section 2.2).
Moreover, the bond cleaved during ester hydrolysis (pH 7.5) of the series of
synthesized steroids 4a-4f does not affect the C-3 stereochemistry of the A ring of the
steroid backbone (Figure 1).

6
N
N
N
N
CHO
CHO
O
O
O
O
HO
OH
X
X
tetrahedral intermediate
X= H(4a), F(4b), Cl(4c), Br(4d), I(4e), CH3(4f)
N
N
CHO
O
O H
O
Alkoxide ion
X
Carboxylic acid
-H+
N
N
CHO
O
O
+
HO
Alcohol
X
Carboxylate ion
Figure 1. Possible mechanism for the hydrolysis of the synthesized esters 4a-4f. In a series
of consecutive steps, the hydroxide attacks the carbonyl group to form a tetrahedral
intermediate, that further produces the corresponding alkoxide by elimination of the
corresponding carboxylic acid, which upon further rapid proton transfer gives place to the
corresponding carboxylate and the alcohol steroid derivative 4. This exothermic transfer of
protons leads to complete saponification and in the process a mole of the base is consumed
to deprotonate the acid.
2.2. Activity of steroid 4 and the derivatives 4a-4f as inhibitors of the enzyme 5α-R1
The results of the inhibitory effect of compounds 4 and 4a-4f on the 5α-R1 activity
are shown in Table 1 (experiments were performed with 5α-R2, showing all the compounds
presented in this work are selective inhibitors of isoenzyme 5α-R1 alone). These results

7
indicate that steroids 4, 4b and 4c are able to block this enzymatic action in vitro, whereas
compounds 4a, 4d-4f does not exhibit biological activity. Moreover, derivative 4 was the
best steroid of the group, with an IC50 value of 0.291 μM, compared to higher IC50 values
for the less active compounds 4b (40 µM) and 4c (83 µM). In this series of experiments,
the IC50 value of steroid 4 was lower than previously reported by our group. [3] This
variation being probably due to the fact that tested animals were different. Since, despite
the same procedure was strictly followed, there is genetic variability among individuals of
the same species. The values of log P (Table 1) also indicate that steroid 4 is very soluble in
the cell membrane, enabling it to efficiently inhibit the activity of 5α-R1, while, steroids
4a-4f, having higher log P values, show low solubility in the membrane. The data also
suggest that these steroids could have a higher plasma half-life.
Table 1. Structure and biological activity of synthesized derivatives.
Structure
Steroid
X
Log P
IC50 (5α-R1)
[μM]
N
N
O
2.96
0.291±0.03
4
H
H
H
HO
21(2-oxo-ethyl-benzimidazol)-3-ceto,
pregnen-16,17-diene
5.08
5.24
5.64
5.91
6.46
5.34
NA
4ª
-H
N
40±5.0
83±12
NA
4b
4c
4d
4e
-F
N
O
-Cl
-Br
-I
H
O
H
H
O
NA
X
NA
X= H(4a), F(4b), Cl(4c), Br(4d), I(4e), CH3(4f) 4f
-CH3

8
2.3 In vivo experiments
2.3.1 Flank organ test
Hamster flank organs are two pigmented spots located in the dorsal skin surface of
hamsters. In female hamsters, the diameter of the pigmented spot measures 2 mm, whereas
in males about 8 mm. These nodules shrink after castration until they resemble the nodules
of the females. However, daily injections or topical applications of testosterone 1 (T) or 2
(DHT) restore their original size [3]. These glands metabolize T to DHT due to 5α-R in this
tissue, [3] thus they are useful as model to test different compounds for 5α-R inhibition. [3,
14]
In this series of experiments, the diameter of the pigmented spot of gonadectomized
male hamster flank organs decreased in size (p < 0.005) when compared to those of the
normal glands. In addition, treatment with the vehicle alone did not produce a change,
however s.c. injections of 1 mg/kg of T for 6 days significantly increased (p < 0.005) the
diameter of the pigmented spot in castrated male hamster flank organs (Figure 1). When
T/finasteride (1 mg/kg) or T/steroids 4, 4a-f (2 mg/kg) were injected, the diameter of the
pigmented spot decreased significantly (p< 0.005) compared to those observed in
testosterone-treated animals (Figure 2). In this case, steroid 4b exhibited the best
pharmacological activity in this gland (the smallest diameter of all those treated, 1.7 ± 0.5
mm).
2.3.2 Effect of steroids 4, 4a-f on the weight of prostate and seminal vesicles
Results from these experiments showed that the weight of the hamster prostate and
seminal vesicles decreased, (p < 0.05) compared to that of normal glands (data not shown).
As was the previous case, treatment with vehicle alone did not produce any change.

9
However, daily subcutaneous injections of 1 mg/kg of T for 6 days significantly increased
(p < 0.05) the weight of these glands in castrated male hamsters (Figure 3). Analogously
when T/finasteride (1 mg/kg) or T/steroids 4, 4a-f (2 mg/kg) were injected, the weight of
the prostate and seminal vesicles decreased significantly (p < 0.05) compared to those of T-
treated animals (Figure 3). In this case also compound 4b exhibited the highest effect,
decreasing the weight of the prostate; however steroid 4e produced the best results in
decreasing the weight of the seminal vesicles.
2.4. Statistical test
ANOVA test produced averages and standard deviations shown in Table 1 and
Figures 2 and 3. Comparison of these averages was made by Dunnett test, indicating
significant differences in activity between the treatment with T alone and the combined
treatments with T/steroids, showing in all cases smaller diameters of the pigmented spot of
gonadectomized male hamster flank organs.
6
5
4
*
*
*
*
3
2
1
0
*
*
*
*
C
T
T+F
T+4
T+4a
T+4b
T+4c
T+4d
T+4e
T+4f
Treatment

10
Figure 2. Flank organ tests. Effect of tested steroids 4, 4a-4f on diameter size of pigmented
spot of gonadectomized male hamster flank organs ± standard deviation. C- Control; T,
testosterone, F finasteride. *highlights the better results compared to T alone.
180
Seminal vesicle
Prostate
160
140
120
100
80
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
60
40
*
20
0
C
T
T+F
T+4
T+4a
T+4b
T+4c
T+4d
T+4e
T+4f
Treatment
Figure 3. Effect of steroids 4, 4a-4f on weight of prostate and seminal vesicles ± standard
deviation. C- Control; T, testosterone, F finasteride. *highlights the better results
compared to T alone.
2.5 Identification of steroidal derivatives in treated hamster plasma
Half hour after compounds 4b-e have been injected, the plasma extract
chromatography plates revealed a single spot with an identical chromatographic behavior to
that observed for steroid 4 (Rf 0.42) with no additional stains due to the ester derivatives,
these results being consistent with those obtained from the in vivo assays. But, other stains,
that were not identified, were observed in each lane of the chromatogram (Rf= 0.086), other
than that, no more traces were observed. Analogously, injection of plasma extracts obtained
at 60 and 120 minutes from hamsters treated with compounds 4a and 4d, 4f, exhibited a

11
plate showing stains due to steroid 4, leading us to conclude that compounds 4a-4f were
hydrolyzed in the in vivo experiments.
3. Discussion
From the series of new steroid esters, only compounds 17-(1H-benzimidazol-1-yl)-
16-formylandrosta-5,16-dien-3-yl-p-fluorobenzoate (4b) and 17-(1H-benzimidazol-1-yl)
formilandrosta-5,16-dien-3β-yl-p-chlorobenzoate (4c) were found to inhibit the activity of
the isoenzyme 1 5α-reductase on in vitro experiments. Moreover, steroid (4) (17-(1H-
benzimidazol-1-yl)-3-hydroxy-16-formylandrosta-5,16-diene) exhibited the smaller IC50
value, proving to be the best in vitro inhibitor of this enzyme. These results being in line
with our previous observations, that compound 4 is a very active steroid in vivo. [3]
The above results also explain the pharmacological effect of 4a, 4d-f in decreasing
the diameter of the hamster flank organs and the weight of the prostate and seminal vesicles
of gonadectomized hamsters treated with T. Since presence of esterase in the plasma,
kidney and liver of hamsters, [3] hydrolyzes steroids 4a, 4d-f in biological media (in vivo
experiments) thus rendering compounds 4a, 4d-f as efficient substrates to inhibit the
enzyme activity after being hydrolyzed producing steroid 4. [3] These observations being
supported by the presence of steroid 4 in plasma of gonadectomized hamsters treated with
T/4a-f. In these experiments compounds 4b-e were hydrolyzed faster than derivatives 4a
and 4d, 4f.
Thus, the results of these experiments show that steroids 4b and 4c slowly
hydrolyze in water, thus having shorter half-life than steroids 4a, 4d-f in the biological
system. On the other hand, steroid 4e resulted to be the most active to decrease the weight
of the seminal vesicle but not as efficient to reduce the weight of prostate. This could well

12
be explained by the presence of another enzyme in the seminal vesicles, which could have
been targeted by 4e hindering the expected activity. [13]
Furthermore, 5α-R 1 is a membrane enzyme, [14] so it is reasonable to think that
lipophilic steroids could be more active in inhibiting its activity. In line with this, steroid 4,
having the lowest log P value (2.96), was indeed the most effective of all in the inhibiting
of 5α-R1. In addition, derivatives 4b and 4c, produced IC50 values of 135.5 and 285 times
greater than that observed for 4, as well as a higher log P value, surprisingly these species
were also effective in blocking the activity of the enzyme in vitro, albeit to be the less
active species. This probably being due to the presence of fluorine and chlorine
substituents, that can promote the hydrolysis process to 4 of esters 4b and 4c in the absence
of esterase due their greater electronegativity.
The advantages of using these ester derivatives is their longer plasma half-life,
which upon presence of esterase in plasma and liver ensures their conversion to their more
active metabolite compound 4 that efficiently inhibits 5α-R1 activity, thus these species
become effective precursors of 4. This being true for steroids 4a, 4d-f, according to the
observed pharmacological effects. Conversely, compounds 4b and 4c are readily
hydrolysable in water, thus although they may serve as precursors of 4, they shorter half-
life leaves steroids 4a, 4d-f to be the best species of the series in biological system.
4. Conclusions
5α-reductase type 1 is the predominant isoenzyme in metastatic prostate bone
tumor, so the activity of steroid 4 and derivatives 4a-4f were tested as inhibitors of type 1
5α-R.

13
In vitro assays showed that steroids 4, 4b and 4c inhibited the activity of type 1 5α-
R, and derivatives 4, 4a-f were very active pharmacologically, being steroid 4e the one
displaying the greatest effect in the seminal vesicles.
The presence of plasma esterases indicated that derivatives 4a, 4d-f having a C-3
ester residue hydrolyze to produce compound 4. This steroid being very active
pharmacologically, thus turning the series of synthesized steroids in this work as efficient
precursors (prodrugs) for the production of steroid 4 increasing their half-life in plasma.
5. Material and methods
5.1. Chemical and radioactive materials
The solvents used were laboratory grade or better. Melting points were determined
on a Fisher-Johns melting point apparatus and are reported without correction, and 1H-
NMR and 13C{1H}-NMR were measured using a Varian Gemini 400 and VRX-300
respectively. Chemical shifts are given in ppm relative to Me4Si (δ =0) in CDCl3 (the
abbreviations of signal patterns are as follows: s, singlet; d, doublet, t, triplet, m, multiplet).
Mass spectra were measured with an HP5985-B spectrometer.
The (1, 2, 6, 7-3H) testosterone [3H] T with specific activity 95 Ci/mmol were
purchased from Perkin Elmer (Boston, MA). Steraloids (Wilton, NH, U.S.A.) supplied
radioinert T and 5α-dihydrotestosterone, and Sigma Chemical Co. (ST. Louis, Mo)
provided NADPH and Lubrol PX.
5.2. Chemical Synthesis
3β-Hydroxy-17-(1H-benzimidazol-1-al)-16-formyl androsta-5,16-diene, 4, was
prepared by the procedure previously reported. [3]

14
Synthesis of steroids 4a-f was performed according to the reaction scheme shown in
Figure 4, using commercially available trans-dehydroandrosterone of ≥ 99% purity (Sigma-
Aldrich®, Mexico City) as starting material. Compounds 2a-f and 3a-f (Figure 4) were
prepared according to the procedure previously reported by our group. [15] The synthesis
of 4a-f is described below.
5.2.1 General procedure for preparation of 17-chloro-16-formyl-androst-5,16-dien-3β-yl
benzoate derivatives containing –H (4a), -F (4b), and –Br (4d) atoms
In a round-bottomed flask fitted with a magnetic stirrer, a solution containing 1.0
mmol of the corresponding ester 17-chloro-16-formyl-androst-5,16-dien-3β-yl benzoate or
17-chloro-16-formyl-androst-5,16-dien-3β-yl p-fluorobenzoate or 17-chloro-16-formyl-
androst-5,16-dien-3β-yl p-bromobenzoate, 3 mmol of benzimidazole and 6 mmol of
sodium carbonate in 5 ml of dimethylformamide was heated to reflux under a nitrogen
atmosphere (flask saturated with nitrogen gas) for about two hours. At the end of the
reaction time, the resulting reaction mixture was cool down to room temperature and
poured into a beaker containing 20 ml of ice to precipitate the ester. This mixture was
filtered under vacuum and the solid obtained was washed three times with cold water to
remove excess sodium carbonate. This product was then purified by flash column
chromatography using Florisil as support (FCC, Florisil® <200 mesh, fine powder from
Sigma-Aldrich) and using hexane–ethyl acetate (96:4) as an elution solvent. The structure
and purity of these compounds were confirmed by spectroscopic and spectrometric
methods (See supplementary material).

15
5.2.2. General procedure for preparation of 17-chloro-16-formyl-androst-5,16-
dien-3β-yl benzoate derivatives containing –Cl (4c), -I (4e), and –CH3 (4f) groups
These compounds were synthesized and purified in a similar manner as compounds 4a, 4b
and 4d using 1.0 mmol of the corresponding ester 17-chloro-16-formyl-androst-5,16-dien-
3β-yl p-chlorobenzoate or 17- chloro-16-formyl-androst-5,16-dien-3β-yl p-iodobenzoate or
17-chloro-16-formyl-androst-5,16-dien-3β-yl p-methylbenzoate, 6 mmol of benzimidazole
and 12 mmol of sodium carbonate in 5 ml of dimethylformamide.
O
O
H
H
X(Ar)COOH,
DCC-DMAP/CHCl3
1h, ta
O
H
H
H
H
HO
O
(1)
X
X= H(2a), F(2b), Cl(2c), Br(2d), I(2e), CH3(2f)
Vilsmeier-Haack
POCl3-DMF/CHCl3
Reflux, 40oC, atm N2
Cl
H
O
H
H
H
H
N
N
O
O
Cl
+
X
H
CHO
H
CHO
Bencimidazole
Na2CO3, DMF
O
O
H
H
O
X
X= H(4a), F(4b), Cl(4c), Br(4d), I(4e), CH3(4f)
X
X= H(3a), F(3b), Cl(3c), Br(3d), I(3e), CH3(3f)
Figure 4. Synthesis route of (3S,8R,9S,10R,13S,14S)-17-(benzimidazol-1-yl)-3-hydroxy-
10,13-dimethyl-2,3,4,7,8,9,11,12,14,15-decahydro-1H-cyclopenta[a]phenanthrene-16-
carbaldehyde benzoate derivatives. [14]

16
17-(1H-Benzimidazole-1-yl)-16-formylandrosta-5,16-dien-3β-yl benzoate (4a). Yield:
77% (240.2 mg) white solid; m.p. 262–63 ºC; IR (KBr): 2,962 (18-CH3, 19-CH3), 719
and 698 (aromatic monosubstitution), 990 and 590 (benzimidazole), 1709 (C=O,
ester), 1,661 (C=O, aldehyde), 1H NMR (400 MHz, CDCl3): δ 1.12 (s, H-19, 3H), 1.22
(s, H-18, 3H), 4.88 (m, H-3, 1H), 5.48 (d, J=5.2 Hz, H-6, 1H), 7.35 (m, 2H benzimidazole),
7.43 (t, Jortho=8.0 Hz, H-meta, 2H), 7.55 (t, Jortho=7.6 Hz, H-para, 1H), 7.86 (m, H-23 and
H-26, 2H benzimidazole), 7.99 (s, H-21, 1H benzimidazole), 8.05 (dd, Jortho=8.4 Hz,
Jmeta=1.2 Hz, H-ortho, 2H), 9.59 (s, -CHO, 1H); 13C{1H} NMR (300 MHz, CDCl3): δ 19.45
(C19), 20.57 (C18), 74.33 (C3), 120.78 (C26 benzimidazole), 122.04 (C23 benzimidazole),
123.57 (C25 benzimidazole), 124.81 (C24 benzimidazole), 128.43 (Cortho), 129.67 (Cmeta),
140.22 (C5), 130.84 (Cipso), 132.93 (Cpara), 140.17 (C21 benzimidazole), 141.58 (C17),
166.11 (C=O ester), 187.93 (CHO);FAB+-MS (M+ + 1) 521 m/z.
17-(1H-Benzimidazole-1-yl)-16-formylandrosta-5,16-dien-3β-yl p-fluorobenzoate (4b).
Yield: 64% (172.2 mg) white solid; m.p. 242–43 ºC; IR (KBr): 1,376 (18-CH3, 19-CH3),
744 and 688 (aromatic disubstitution), 1,238 (Ph-F), 985 (benzimidazole), 1,714
(C=O, ester), 1,674 (C=O, aldehyde, 1H NMR (400 MHz, CDCl3): δ 1.11 (s, H-19, 3H),
1.24 (s, H-18, 3H), 4.86 (m, H-3, 1H), 5.47 (d, J=4.8 Hz, H-6, 1H), 7.10 (t, Jortho=8.0 Hz,
JH-F=8.0 Hz H-meta, 2H), 7.35 (m, 2H benzimidazole), 7.86 (m, H-23 and H-26, 2H
benzimidazole), 7.99 (s, H-21, 1H benzimidazole), 8.05 (dd, Jortho=9.0 Hz, JH-F=5.6 Hz, H-
ortho, 2H), 9.59 (s, -CHO, 1H); 13C{1H} NMR (300 MHz, CDCl3): δ 19.44 (C19), 20.57
(C18), 74.52 (C3), 115.44, 115.66 (Cmeta), 120.79 (C26 benzimidazole), 122.12 (C23
benzimidazole), 123.57 (C25 benzimidazole), 124.81 (C24 benzimidazole), 127.05 (Cipso),

17
132.15, 132.24 (Cortho),140.08 (C17), 140.22 (C5), 141.01 (C21 benzimidazole), 165.15
(Cpara-F), 167.09 (C=O ester), 187.94 (CHO); FAB+-MS (M+ + 1) 539 m/z.
17-(1H-Benzimidazole-1-yl)-16-formylandrosta-5,16-dien-3β-yl p-chlorobenzoate (4c).
Yield: 64% (177.4 mg) white solid; m.p. 242–43 ºC; IR (KBr): 1,372 (18-CH3, 19-CH3),
738 and 685 (aromatic disubstitution), 815 and 1,089 (Ph-Cl), 591 and 472
(benzimidazole), 1,716 (C=O, ester), 1,665 (C=O, aldehyde), 1H NMR (400 MHz,
CDCl3): δ 1.11 (s, H-19, 3H), 1.22 (s, H-18, 3H), 4.86 (m, H-3, 1H), 5.47 (d, J=5.2 Hz, H-
6, 1H), 7.35 (m, 2H benzimidazole), 7.40 (d, Jortho=8.4 Hz, H-meta, 2H), 7.85 (m, H-23 and
H-26, 2H benzimidazole), 7.97 (d, Jortho=8.8 Hz, H-orto, 2H), 7.99 (s, H-21, 1H
benzimidazole), 9.59 (s, -CHO, 1H); 13C{1H} NMR (300 MHz, CDCl3): δ 19.43 (C19),
20.56 (C18), 74.65 (C3), 120.78 (C6), 122.16 (C23 benzimidazole), 123.57 (C25
benzimidazole), 124.80 (C24 benzimidazole), 128.76 (Cipso), 129.27 (Cmeta), 131.07
(Cortho), 139.37 (C16), 140.02 (C5), 141.57 (C21 benzimidazole), 165.23 (C=O ester),
187.92 (CHO). FAB-MS (M+ + 1) 555 m/z, (M+ + 3) 557 m/z.
17-(1H-Benzimidazole-1-yl)-16-formylandrosta-5,16-dien-3β-yl p-bromobenzoate (4d).
Yield: 67% (200.3 mg) white solid; m.p. 223–25 ºC; IR (KBr): 1,375 (18-CH3, 19-CH3),
758 and 683 (aromatic disubstitution), 1,009 (Ph-Br), 590 and 501 (benzimidazole),
1,716 (C=O, ester), 1,664 (C=O, aldehyde), 1H NMR (400 MHz, CDCl3): δ 1.11 (s, H-
19, 3H), 1.22 (s, H-18, 3H), 4.86 (m, H-3, 1H), 5.48 (d, J=4.4 Hz, H-6, 1H), 7.35 (m, 2H
benzimidazole), 7.57 (d, Jortho=8.8 Hz, H-meta, 2H), 7.86 (m, H-23 and H-26, 2H
benzimidazole), 7.89 (d, Jortho=8.4 Hz, H-ortho, 2H), 7.99 (s, H-21, 1H benzimidazole),

18
9.59 (s, -CHO, 1H); 13C{1H} NMR (300 MHz, CDCl3): δ 19.43 (C19), 20.56 (C18), 74.68
(C3), 120.79 (C6), 122.17 (C23 benzimidazole), 123.56 (C25 benzimidazole), 124.77 (C24
benzimidazole), 128.02 (Cipso), 129.73 (C16), 131.22 (Cmeta), 131.76 (Cortho), 140.01 (C5),
141.57 (C21 benzimidazole), 165.59 (C=O ester), 187.91 (CHO); FAB+-MS (M+ + 1) 599
m/z.
17-(1H-Benzimidazole-1-yl)-16-formylandrosta-5,16-dien-3β-yl p-yodobenzoate (4e).
Yield: 60% (200.3 mg) white solid; m.p. 228–30 ºC; IR (KBr): 755 and 685 (aromatic
disubstitution), 532, 475 and 459 (Ph-I), 985 and 600 (benzimidazole), 1,731 (C=O,
ester), 1,713 (C=O, aldehyde), 1H NMR (400 MHz, CDCl3): δ 1.11 (s, H-19, 3H), 1.22
(s, H-18, 3H), 4.86 (m, H-3, 1H), 5.48 (d, J=4.0 Hz, H-6, 1H), 7.35 (m, 2H benzimidazole),
7.76 (d, Jortho=8.6 Hz, H-ortho, 2H), 7.86 (m, H-23 and H-26, 2H benzimidazole), 7.99 (s,
H-21, 1H benzimidazole), 9.59 (s, -CHO, 1H); 13C{1H} NMR (300 MHz, CDCl3): δ 19.43
(C19), 20.56 (C18), 74.65 (C3), 120.78 (C6), 122.16 (C23 benzimidazole), 124.82 (C24
benzimidazole), 130.30 (C16), 131.15 (Cmeta), 137.78 (Cortho), 140.01 (C5), 141.57 (C21
benzimidazole), 165.59 (C=O ester), 187.91 (CHO); FAB+-MS (M+ + 1) 647 m/z.
17-(1H-Benzimidazole-1-yl)-16-formylandrosta-5,16-dien-3β-yl p-methylbenzo-ate
(4f). Yield: 80% (213.6 mg) white solid; m.p. 217-18 °C; IR (KBr): 2,955 and 1,452 (18-
CH3, 19-CH3), 742 and 691 (aromatic disubstitution), 988 and 588 (benzimidazole),
1,705 (C=O, ester), 1H NMR (400 MHz, CDCl3): δ 1.11 (s, H-19, 3H), 1.26 (s, H-18, 3H),
2.40 (s, Ar-CH3, 3H), 4.86 (m, H-3, 1H), 5.48 (d, J=4.4 Hz, H-6, 1H), 7.23 (d, Jortho=8.8
Hz, H-meta, 2H), 7.35 (m, 2H benzimidazole), 7.85 (m, H-23 and H-26, 2H

19
benzimidazole), 7.93 (d, Jorto=8.4 Hz, H-ortho, 2H), 7.99 (s, H-21, 1H benzimidazole),
9.59 (s, -CHO, 1H); 13C{1H} NMR (300 MHz, CDCl3): δ 19.45 (C19), 20.56 (CH3-Ar),
21.78 (C18), 74.10 (C3), 120.76 (C26 benzimidazole), 121.96 (C23 benzimidazole), 123.57
(C25 benzimidazole), 124.80 (C24 benzimidazole), 128.10 (Cipso), 129.13 (Cmeta), 129.69
(Cortho), 140.24 (C5), 141.58 (C21 benzimidazole), 143.56 (C17), 166.18 (C=O ester),
187.94 (CHO); FAB+-MS (M+ + 1) 535 m/z.
5.3 Biological Activity
5.3.1 Animals and tissues
For the in vitro experiments, two adult 8-month-old rats were sacrificed with CO2
and their livers removed in order to prepare microsomes as the source of the 5α-R1
isoenzyme. The rats were previously fasted overnight to decrease glycogen levels before
removal of the livers as indicated in Levy et al. [16]. For the in vivo experiments, 2.5-
month-old adult male golden hamsters (150–200 g) were obtained from the Metropolitan
University–Xochimilco, Mexico. Gonadectomies were performed on the hamsters under
pentobarbital anesthesia and the castrated hamsters were kept in a room with controlled
temperature (22 ºC) with 12-hour light–dark periods. Food and water were provided ad
libitum. After 36 days under these conditions, the hamsters were sacrificed with CO2. Both
protocols were approved by the Institutional Care and Use Committee of the Metropolitan
University of Mexico (UAM). The experiments with the gonadectomized animals was
carried out on 8 groups of 4 animals/experiment.
Rat livers were minced (15–30 g) in one volume of buffer A (20 mM sodium
phosphate, pH 7.5 containing 0.32 M sucrose, 0.1 mM dithiothreitol; Sigma-Aldrich) using
an IKA A11 basic tissue mill. Unless otherwise specified, the following procedures were

20
carried out at 4 ºC. The tissue was homogenized, and the suspension centrifuged at 10,000
× g for 30 minutes (Beckman L70K ultracentrifuge); the pellet was discarded. The
supernatant liquid was filtered through a strainer and centrifuged at 100,000 × g for 60
minutes. The microsomal pellet was washed by resuspension in five volumes of buffer A
with a homogenizer, and the proteins were determined by the Bradford method. 17 The
suspension was centrifuged at 100,000 ×g and the pellet suspended in buffer A to a
concentration of 10–20 mg of protein/mL; the microsomal solution was stored at −80 °C
prior to preparation of the solubilized steroid type 1 5α-reductase [16].
5.3.2 Solubilization of type 1 5α-reductase from rat liver microsomes
The solubilization of type 1 5α-reductase isozyme from rat liver microsomes was
carried out according to Levy et al. [16]. Briefly, the microsomes were suspended in five
volumes of buffer B (100 mM sodium citrate, 100 mM potassium chloride, 1.0 mM
dithiothreitol (DTT), 0.1% nonaethylene glycol monododecyl ether (Lubrol PX), 20%
glycerol, and 0.1 mM NADPH adjusted to pH 7.5). The blend was mixed, and the solution
incubated on ice for 30 minutes and then centrifuged at 100,000 × g for 60 min. Aliquots
from the supernatant liquid which contained more than 80% of the microsomal activity
were stored at −80 °C. The protein concentration was determined by the Bradford method.
17
5.3.3 Determination of type 1 5α-reductase activity

21
The activity of the type 1 5α-reductase isoenzyme was determined following the
conversion of T into DHT, as previously reported. We used solubilized liver microsomes as
the source of this enzyme. [15, 16]
5.3.4 Determination of the 50% of the inhibitory concentration of steroids 4, 4a-f in type 1
5α-reductase activity.
The concentration of steroids 4, 4a-f required to inhibit 50% of 5α-R1 activity (IC50
values) was determined using seven series of tubes containing increasing concentrations of
these steroids (10−11–10−3 M solubilized in DMSO). The tubes were incubated in duplicate
with 1 mM of dithiothreitol, 20 mM sodium phosphate buffer of pH 7.5, 2 mM NADPH, 2
nM [1,2,6,7-3H]T and 6.31 µM of non-labeled T. The reaction was started by adding 60 µg
of soluble type 1 5α-reductase isozyme and the final incubation volume was 1 mL. [15]
The reactions were carried out in duplicate at 37 °C for 60 min; 1 mL of
dichloromethane was added to stop the reaction. The extraction and chromatographic
procedures were carried out as we have described previously [15] and the radioactivity
contained in the fraction corresponding to DHT standard was determined.
The concentrations of steroids 4, 4a-f required to inhibit the 5-R1 activity by 50%
(IC50) were determined from the inhibition plots using SigmaPlot Software.
5.3.5 Pharmacological Experiments
For the daily subcutaneous injections, 2 mg/kg of [18-19] steroids 4, 4a-f dissolved
in 200 L of sesame oil were administered to the gonadectomized hamsters together with 1
mg/kg of testosterone for 6 days. Three groups of gonadectomized hamsters were kept as
controls; one group was injected with 200 L of sesame oil, the second with 1 mg/kg of T

22
and the third with T plus 1 mg/kg of finasteride for 6 days. After treatment, the animals
were sacrificed with CO2. The flank organ diameter size was measured using a Vernier
caliper and the prostate and seminal vesicles of each animal were dissected and weighed.
Two separate experiments were performed for each group of steroid-treated animals. The
results were analyzed using one-way analysis of variance and Dunnett’s method to compare
means, using JMP IN 5.1 software.
5.4 Determination of the steroidal derivatives in hamster’s plasma
In order to identify the presence of hydrolysis products of compounds 4a-f to the
highly in vivo active compound 4 (due to kidney and plasma esterases [20, 21]), six groups
of gonadectomized hamsters were treated with a single sc injection of mixtures of T
(1mg/kg) and each of the steroidal esters (4 mg/Kg). After this treatment, two hamsters
from each group were slaughtered after periods of 30, 60, 120 minutes.
Blood samples from the hamsters obtained by cardiac puncture, were centrifuged at
500 g for 5 minutes and steroids present in the plasma were extracted with benzene (1: 1).
The solvent was evaporated until dryness in vacuum, and the resulting extract dissolved in
5 l of ethyl acetate. Samples were dripped on different lanes of a TLC and 5µL of 4 and
4a-f were spotted to both sides of this plate as reference standards. Plates were developed in
a mixture of toluene/ethyl acetate/methanol= 10/5/1.5 and plates were exposed to UV
radiation at a wavelength of 274 nm for revealing.
5.5. Calculation of theoretical log P value.
The calculations of the logP parameter were carried out using the ACD/LABS
software version 9. These values are calculated on the base of the molecular structure of the

23
tested steroid and describes the lipophilicity of the drug. The values correlate the partition
coefficient (log P) and the minimum concentration of the drug (log 1/c) necessary to
produce a biological activity. Thus, the higher the value of log P indicates a greater
solubility of the compound in the lipoprotein cell membrane. For the antiandrogenic effect
the optimum log P value is around 5.

24
6. ACKNOWLEDGEMENT
We would like to thank CONACYT project No 165049 for generous financial support.
7. CONFLICT OF INTEREST
The authors declare that there are no real or perceived conflicts of interest arising
from intellectual, personal, or financial circumstances of the research. Additionally, all
authors are aware and approve of the contents and order of authorship of the manuscript.

25
8. References
[1]
[2]
[3]
A. Chávez-Riveros, E. Bratoeff, Y. Heuze, J. Soriano, I. Moreno, A. Sánchez-
Márquez, M. Cabeza, Archiv der Pharmazie 348(11) (2015) 808-816.
M. Garrido, M. Cabeza, F. Cortés, J. Gutiérrez, E. Bratoeff, European Journal of
Medicinal Chemistry 68(0) (2013) 301-311.
Y. Arellano, E. Bratoeff, T. Segura, M.E. Mendoza, A. Sánchez-Márquez, Y.
Medina, Y. Heuze, J. Soriano, M. Cabeza, Journal of Enzyme Inhibition and
Medicinal Chemistry 31(6) (2016) 908-914.
[4]
[5]
[6]
[7]
[8]
F. Cortés-Benítez, M. Cabeza, M.T. Ramírez-Apan, B. Alvarez-Manrique, E.
Bratoeff, European Journal of Medicinal Chemistry 121 (2016) 737-746.
G.A. Elmegeed, W.K.B. Khalil, R.M. Mohareb, H.H. Ahmed, M.M. Abd-Elhalim,
G.H. Elsayed, Bioorganic & Medicinal Chemistry 19(22) (2011) 6860-6872.
K. Li, L. Li, S. Wang, X. Li, T. Ma, D. Liu, Y. Jing, L. Zhao, European Journal of
Medicinal Chemistry 126(Supplement C) (2017) 910-919.
Y.F. Mui, J. Fernández-Gallardo, B.T. Elie, A. Gubran, I. Maluenda, M. Sanaú, O.
Navarro, M. Contel, Organometallics 35(9) (2016) 1218-1227.
J. Ferlay, I. Soerjomataram, R. Dikshit, S. Eser, C. Mathers, M. Rebelo, D.M.
Parkin, D. Forman, F. Bray, International Journal of Cancer 136(5) (2015) E359-
E386.
[9]
L.N. Thomas, C.B. Lazier, R. Gupta, R.W. Norman, D.A. Troyer, S.P. O'Brien, R.S.
Rittmaster, The Prostate 63(3) (2005) 231-239.
[10] I.U. Agoulnik, A. Vaid, M. Nakka, M. Alvarado, W.E. Bingman, H. Erdem, Cancer
Res 66 (2006) 10594-10602.

26
[11] C. Cai, H.H. He, S. Chen, I. Coleman, H. Wang, Z. Fang, Cancer Cell 20 (2011)
457-471.
[12] C.E. Massie, A. Lynch, A. Ramos-Montoya, J. Boren, R. Stark, L. Fazli, EMBO J
30 (2011) 2719-2733.
[13] E. Koh, T. Noda, J. Kanaya, M. Namiki. The Prostate 53(2)(2002):154-159.
[14] D.W. Russell, J.D. Wilson, Annual Review of Biochemistry 63(1) (1994) 25-61.
[15] E. Bratoeff, A. Sánchez, Y. Arellano, Y. Heuze, J. Soriano, M. Cabeza, Journal of
Enzyme Inhibition and Medicinal Chemistry 28(6) (2013) 1247-1254.
[16] M.A. Levy, M. Brandt, A.T. Greway, Biochemistry 29(11) (1990) 2808-2815.
[17] M.M. Bradford, Analytical Biochemistry 72(1–2) (1976) 248-254.
[18] M. Cabeza, Y. Heuze, H. Quintana, E. Bratoeff. Asian Journal of Animal and
Veterinary Advances, 5 (2010) 202-209.
[19] M. Cabeza, E. Bratoeff, E. Flores, E. Ramírez, J. Calleros, D. Montes, A. Quiroz, I.
Heuze. Chemical and Pharmaceutical Bulletin, 50 (2002) 1447-1452.
[20] J.J. Li, H. Kirkman, R.L. Hunter. Journal of Histochemistry & Cytochemistry, 17
(1969) 386-393.
[21] B.C. Allen, D.J. Moore. Endocrinology, 78 (1966) 655-658.

27
Graphical abstract

28
Highlights
Highlights
 5α-Reductase type1 is the predominant isoenzyme in prostate metastatic bone tumor
 7 steroid derivatives were synthesized in a simple, high yield manner.
 This species selective inhibit the activity of the isoenzyme 5α-R1
 Their inhibitory activity is associated to their stability to hydrolysis.
 These derivatives exhibit improved plasma half-life.