Letter
IC = 93 nM), 20 (R = i-Bu, F877L IC = 102 nM; WT IC
50
5
0
1
50
=
100 nM), and 21 (R = i-Bu, F877L IC = 123 nM; WT
1 50
IC50 = 126 nM). An additional derivatization of R in 22−25
1
did not further enhance the potency. Nevertheless, the
potencies of 18−21 were comparable to or slightly better
than that of 4. Interestingly, chiral (R)-28 (F877L IC = 65
5
0
nM; WT IC50 = 52 nM) and (R)-29 (F877L IC50 = 78 nM;
WT IC50 = 72 nM) were about two times more potent than
the corresponding (S)-26 (F877L IC50 = 112 nM; WT IC =
50
178 nM) and (S)-27 (F877L IC50 = 151 nM; WT IC50 = 170
nM). The assignments of absolute stereochemistry for 26−29
were based on VCD experiments due to the unsuccessful
efforts to obtain single crystals for X-ray studies and the
inability to predict the potency difference in our homology
2
9,30
model.
Significantly, these analogues remained full
antagonists in both AR F877L and the AR WT despite having
the same “B” substituents as 1 (enzalutamide) or 3
(
apalutamide), both of which are agonists in AR F877L.
This observation suggests the intrinsic propensity of scaffold 8
for antagonism against both AR F877L and the AR WT as well
12
as for selectivity over the GR (Figure 4).
To correlate the AR antagonism with an antiproliferative
effect in androgen-dependent tumor cell lines, 12, 16, 18−20,
and 26−29 were also evaluated in a growth inhibition assay in
the AR WT-dependent VCaP cells, again using 1 and 4 as
31
comparators. Luciferase transcription inhibition appeared to
translate into antiproliferative activity in VCaP cells, tracking
well with LNCaP WT potency with IC50 values ranging from
1
2
0 to 2440 nM (Table 3). The potencies of 16, 18−20, (R)-
8, and (R)-29 were comparable to or better than those of 1
and 4.
In mouse single-dose pharmacokinetic (PK) studies, (R)-29,
Figure 5. Hershberger assay: Dosing effect of compounds 16 (30 mg/
kg), 19 (5 mg/kg), (S)-26 (30 mg/kg), and (R)-28 (30 mg/kg)
compared with control group 1 (enzalutamide, 30 mg/kg) and (R)-29
the more potent enantiomer of 19 (Table 1), displayed a
higher area under the receiver operating characteristic (ROC)
curve (AUC) (104 μg·h/mL) after oral dosing and lower
clearance (CL 0.9 mL/min/kg) after IV dosing compared with
the corresponding, less potent (S)-27 (AUC 50.4 μg·h/mL,
CL 3.0 mL/min/kg) (Table 3). The same trend was observed
by comparing (R)-28 and (S)-26 in terms of the exposure and
clearance, indicating the subtle difference between (R)- and
(S)-enantiomers in terms of the PK characteristics. The overall
PK parameters of (R)-29 tracked well with its corresponding
racemate, 19. All analogues displayed favorable PK parameters,
with oral bioavailability ranging from 61 to >100% (Table 4).
Compounds 16, 19, (S)-26, (S)-27, (R)-28, and (R)-29
were evaluated in rats for their inhibitory effect on the growth
of androgen-sensitive organs (ASOs) under stimulation by
testosterone propionate (TP) in the Hershberger assay to
assess their in vivo antiandrogen activities against the WT AR
(Figure 5). Treatment with compounds 16, (S)-26, and (R)-
28 resulted in statistically significant reductions in ASOs versus
the TP control at 30 mg/kg once daily oral dosing for 10 days
(p > 0.0001; Figure 5), comparable to that of positive control
enzalutamide (1) at 30 mg/kg. Importantly, compound 19
showed a similar reduction in ASOs at 5 mg/kg. In a separate
study, treatment with compound (R)-29 resulted in statistically
significant efficacy at 5 mg/kg once daily oral dosing for 10
days (p ≤ 0.0001; Figure 5) compared with that of positive-
control flutamide at 3 mg/kg. In contrast, compound (S)-27
showed minimal effects on ASOs at 5 mg/kg (Figure 5),
consistent with its less robust in vitro AR antagonistic potency
(Tables 1 and 2) and lower in vivo PK exposure compared with
(R)-29 (Table 4).
(
5 mg/kg) and (S)-27 (5 mg/kg) compared with control group
flutamide (3 mg/kg) on ASOs. The ASO development of seminal
vesicles and coagulating glands (SVCG) and the ventral prostate (VP)
is shown. The compound-dependent suppression of ASOs is
significant for each compound tested (p ≤ 0.0001, t-test/Mann−
Whitney). Data are the mean ± SD (n = 6).
observed for spirocyclic compounds under the assay
conditions.
The unsubstituted (R = H) analogues 9 (F877L IC50
5 000 nM; WT IC50 > 15 000 nM) and 10 (F877L IC50
012 nM; WT IC50 = 11 220 nM) were at least 10 times less
>
=
1
1
5
potent compared with the corresponding N-methylated (R =
1
Me) compounds 11 (F877L IC = 676 nM; WT IC = 1380
50
50
nM) and 12 (F877L IC50 = 525 nM; WT IC50 = 646 nM).
However, the potencies of fluorinated (R = F) 13 (F877L
2
IC50 = 759 nM; WT IC50 = 1039 nM) and 14 (F877L IC50
=
3
39 nM; WT IC = 398 nM) were almost equal to or slightly
5
0
better than those of nonfluorinated (R = H) 11 and 12. By
2
contrast, hydantoin 15 (F877L IC50 = 1660 nM; WT IC50
=
5
129 nM) lost significant potency compared with the
corresponding thiohydantoin 12 (F877L IC50 = 525 nM;
WT IC50 = 646 nM), confirming the importance of the
thiocarbonyl moiety. It was notable that analogue 16 (F877L
IC = 178 nM; WT IC = 219 nM) was a full antagonist and
5
0
50
was almost equally potent as benchmark 4 (F877L IC50 = 98
nM; WT IC50 = 191 nM); both compounds possess a
piperidinyloxy substituent on ring “B”. A further potency
increase was achieved in 18 (R = i-Pr, F877L IC = 145 nM;
1
50
WT IC = 112 nM), 19 (R = i-Pr, F877L IC = 93 nM; WT
5
0
1
50
E
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX