A. J. Roecker et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2079–2085
2083
Table 3
Pharmacokinetics of DORA 42 in rat and dog
IVa
Dose (mg/kg)
POb
Tmax (h)
Species
Dose (mg/kg)
Cl (mL/min/kg)
Vdss (L/kg)
T1/2 (h)
AUC (lM h)
Cmax
(
lM)
F (%)
Rat
2
0.5
NA
22
13
NA
0.3
0.7
NA
0.2
0.8
NA
10
3
25
2.8
3.1
6.1
0.3
0.6
0.5
3.8
1.2
2.5
17
34
15
Dog
Ratc
a
b
c
Vehicle = 100% DMSO in rat and dog (n = 2).
Dosed as the hydrochloride salt in 100% PEG200 (n = 2).
Dosed as hydrochloride salt in 20% VitE TPGS in water; satellite rats for sleep study pharmacokinetics (n = 3).
H2N
H2N
S
N
N
O
O
Me
N
N
N
Me
N
(a)
N
N
+
N
N
HN
N
Cl
N
N
S
Me
Me
13
43
42
Scheme 2. Gram scale synthesis of compound 42. Reagents and reaction conditions: (a) triethylamine, DMF, 185 °C, 40 min, microwave, 39% (7.5 mmol scale; 1.3 grams
produced). See Ref. 2c for the synthesis of diazepane core 43.
38 possessed excellent potency profiles, the unsubstituted thio-
phene motif is a known substrate for bioactivation.13 Precedent
suggested that substitution on the thiophene motif can reduce this
liability.14 Since DORA 42 demonstrated the best balance of bind-
ing potency, functional potency, and theoretical bioactivation po-
tential of the final four analogs (39–42), this compound was
selected for additional profiling (Fig. 2).
The 4-amino substituent present in DORA 42 improved free
fraction to measurable levels across species and decreased logD
by 1.8 units compared to compound 34 (Table 2). In addition, the
compound was highly permeable and not a substrate for rat or hu-
man P-glycoprotein efflux. Compound 42 possessed polar surface
area in an acceptable range for brain penetrant small molecules.15
Pharmacokinetic properties were evaluated for compound 42
(dosed as a hydrochloride salt) and these data are displayed in
Table 3. In the rat and the dog, this compound had moderate clear-
ance, a short half-life, and bioavailability of 17% and 34%, respec-
tively. Since this compound was suitable for oral administration,
it was scaled up for additional in vivo characterization in rat sleep
studies. The gram scale synthesis is shown in Scheme 2.
an internal standard (Fig. 4). Although precedent suggested that
substituted thiophene analogs could reduce bioactivation potential
relative to their unsubstituted counterparts,14 the direct compari-
son between phenyl and thiophene isosteres is less well studied.
The bioactivation potential of DORA 42 precluded further develop-
ment of thiophene-containing DORAs.
Expansion of an acetonitrile-assisted, diphosgene-mediated
synthesis of bicyclic 2,4-dichloropyrimidines enabled the expedi-
tious production of analogs of DORA 2. These efforts afforded
DORAs with excellent binding and functional potency on OX1R
and OX2R as well as improved physiochemical properties. DORA
42 was identified as an advanced compound with superb potency
and rat sleep efficacy comparable to DORA 2. Unfortunately, the
bioactivation potential of DORA 42 precluded further development
Figure 3 displays the results of sleep studies from active phase
radiotelemetric recordings in the rat over the two hour period di-
rectly following compound administration using methods previ-
ously described.16 Upon dosing the hydrochloride salt of
compound 42 in rats at 25 mg/kg (p.o. in 20% Vitamin E TPGS),
immediate and significant reduction in active wake and concomi-
tant increases in slow wave sleep (SWS) and REM sleep were ob-
served over the two hour period immediately following
treatment. The sleep efficacy of compound 42 in the rat was
accompanied by a Cmax of 2.5
lM, Tmax of 30 min, and an AUC of
6.1 M h (Table 3). The Cmax data for compound 42 coincided ni-
l
cely with results from ex vivo occupancy studies in transgenic rats
which overexpress human OX2R demonstrating 88% receptor occu-
Figure 3. Sleep architecture effects of DORA 42 and DORA 2 in rat sleep dosed
during their active phase. (a) Mean change in active wake, light sleep, SWS and REM
sleep time relative to vehicle (20% Vitamin E TPGS) in rats monitored for two hours
following p.o. administration of compound 2 (30 mg/kg) and compound 42 (25 mg/
kg). Treatment occurred at Zeitgeber Time 16:00 (4 h after lights off) in a balanced
cross-over design such that each subject received drug and vehicle (3 or 7 days of
consecutive treatment, compounds 2 and 42, respectively). Values represent the
mean of within-subject differences between vehicle and compound treatment
( SEM). Mean times for the compound condition in experiments evaluating
compound 2 and compound 42 experiments, respectively (min): AW: 58, 53; LS:
14, 12; SWS: 41, 46; REM: 7, 9. Data were analyzed using within-subject ANOVA to
determine main effects and one sample t-test to compare to vehicle (N = 14, 8 for
studies evaluating compounds 2 and 42, respectively; ⁄p <0.05, ⁄⁄p <0.01, ⁄⁄⁄p
<0.001).
pancy at a plasma concentration of 2.0 l
M.17 These findings follow
the same occupancy/efficacy relationship as compounds 1 and 2. In
fact, Compound 42 induced a sleep architecture profile nearly
identical to that observed for compound
2
administered at
30 mg/kg (Cmax: 2.0 lM; Tmax: 1.3 h; AUC: 4.5 lM h), as previously
described for this structure as well as compound 1.2c
Finally, DORA 42 and DORA 2 were directly compared for bioac-
tivation potential via our GSH trapping assay in rat and human
microsomal incubations.18 DORA 42 and DORA 2 demonstrated
similar levels of bioactivation in rat and human experiments as
determined by AUC measurements of all GSH adducts relative to