S. Boutin, et al.
Bioorganic & Medicinal Chemistry Letters xxx (xxxx) xxxx
1
4
14
Fig. 3. Transformation of [ C]-3α-diol/3β-diol (8:2) into [ C]-5α-DHT in LAPC-4 cells. A) Modifying the number of cells, with a fixed concentration of substrate
1
4
after 24 h of incubation of [ C]-3α-diol/3β-diol (9:1) (non-radiolabeled/radiolabeled 20:1, 1 μM) in LAPC-4 cells. B) Relative % of steroids (5α-DHT, 3α-diol and 3β-
diol) in the preparation of labeled 3α-diol. C) Relative % of steroids (5α-DHT, 3α-diol and 3β-diol) in the preparation of labeled 3α-diol incubated with LAPC-4 cells.
LAPC-4 cells was also never used before in an inhibition assay. There-
fore, different parameters of the assay (time, number of cells, substrate
concentration) needed to be adjusted to obtain a sensitive and re-
producible inhibition assay.
Having established the experimental conditions of the biological
assay, steroid derivatives were selected for a first screening round to
identify the inhibitors of the 5α-DHT formation from 3α-diol in a PC
cell model. These 3β-piperazine androstane derivatives were initially
developed as inhibitors of 17β-HSD10 oxidative activity inactivating
In a first series of experiments, the concentration of substrate (0.2, 1
1
4
32
and 10 μM) of 3α-diol/[4- C]-3α-diol (20:1) and the number of cells
per well (10,000, 50,000, 100,000 and 200,000 cells) were tested,
whereas incubation time was fixed at 24 h. After extraction of the
steroids and elution on TLC plates (see supplementary data for the
complete procedure), the % of transformation was obtained for each
condition. At the lower concentration of substrate (200 nM of a mixture
of radiolabeled/non-radiolabeled 3α-diol) and using the four numbers
of cells mentioned above, percentages of transformation were con-
sidered not suitable for further assay, since the radioactivity measured
on each TLC spot was below the detection limit. At the higher con-
centration of substrate (10 μM) with 100,000 and 200,000 LAPC-4 cells,
the percentages of transformation were also too weak (0–5%). It seems
that a substrate inhibition effect, which was previously observed for
the potent estrogen estradiol, and were therefore reintroduced in our
screening study as the starting point to identify inhibitors of 3α-diol
transformation into 5α-DHT in LAPC-4 cells.
The selected compounds were screened at two concentrations, 3 and
10 μM (Table 1). For this first screening round, the most potent in-
hibitors were compounds 4b, 6, D-3,7 and C-1,12 with inhibition va-
lues ranging from 27.0 to 35.4% at 3 μM, and 44.2 to 55.6% at 10 μM.
The other compounds also inhibited the 5α-DHT formation, but to a
lesser extent; inhibition ranging from 0 to 22.3% at 3 μM and 3.3 to
51.2% at 10 μM. Furthermore, some elements of structure–activity re-
lationship (SAR) were generated from these results. First, among the D-
ring derivatives 1–7, the 17β-OH/17α-C^CH and the 17,17-di-F
functionalities are clearly better than the others (17-ketone, 17β-OH/
3
0,31
other steroid substrates and steroidogenic enzymes,
would be re-
17α-H, 17β-OH/17α-CH
CH ). Second, the presence of 4-CF
chain at C-3β is important, as shown by comparison of the inhibitory
3
, 17β-H/17α-OH and 17-ketone/16,16-di-
sponsible for these results. At a concentration of 1 μM of substrate, the
results showed (Fig. 3A) that 100,000 cells and 24 h were enough to
obtain a good transformation of 3α-diol in LAPC-4 cells (around 18% of
3
3
on the aromatic ring of the side
potencies of 1 (4-H), D-2,4 (4-CH
3
) and D-3,7 (4-CF ), the last one
3
5
α-DHT formation). Furthermore, a dose (cell number)-dependent
being the most potent. Among the last compounds tested in this assay,
with diverse binding functionalities (Y) and aryl moieties (Z), the
thiourea derivative C-1,12 is the best inhibitor.
transformation can be observed, since the transformation increased
with the quantity of cells.
Since the 3β-diol isomer was present in the substrate solution, an
assay was performed to confirm that this minor isomer would not affect
the results of our inhibition screening. Thus, radioactivity of 3α-diol/
For the second screening round (Table 2), several compounds were
chosen to confirm and extend previous results. Inhibition values ob-
tained with D-3,7 and C-1,12 show the reproducibility of the assay.
3
β-diol/5α-DHT was quantified after incubation for 24 h, with or
Also, a 4-CF group on the aromatic ring (Z) provides better inhibition
3
without LAPC-4 cells. When the substrate was incubated without cells,
α-DHT was not present (no significant amount of this metabolite was
than a 4-F moiety, since compound D-3,7 is more potent than D-2,6.
Another SAR element that was studied was the position of the CF
moiety on the phenyl ring. The positioning of CF did not have a sig-
nificant effect on inhibition potency, since D-3,6 (3-CF /5-OCH ), D-
3,7 (4-CF /3-OCH ) and D-3,8 (2-CF /3-OCH ) similarly inhibited the
5α-DHT formation, but were more potent than D-3,5 (2-CF /5-OCH ).
/3-OCH
5
3
detected on the TLC plates) and a 3α-diol/3β-diol ratio of 9:1 was
obtained (Fig. 3B). However, when the substrate was incubated with
LAPC-4 cells, 5α-DHT was formed and a 5α-DHT/3α-diol/3β-diol ratio
of 20/69/11 was obtained (Fig. 3C). Thus, with or without cell in-
cubation, radiolabeled 3β-diol showed no significant change in its re-
lative proportion (11.4 and 10.6%). Therefore, the contribution of 3β-
diol to the amount of 5α-DHT formed was negligible and would not
affect the inhibition results in the screening assay.
3
3
3
3
3
3
3
3
3
3
3
Interestingly, compound 25, a hybrid between D-3,7 (4-CF
)
and 4b (17β-OH/17α-C^CH), inhibited 74.0% of oxidative activity at
3 μM, and 88.8% at 10 μM, similarly to unlabeled substrate 3α-diol.
Another SAR element studied was the impact of the Y functionality
3