channels and the CYP2C9, CYP2C19, CYP2D6, and CYP3A4
enzymes (Table 3). Compound 36 showed no significant
inhibition (IC50>5 M) of the hERG channel, a measure often used
to assess potential cardiovascular risks, as well as moderate to no
inhibition (IC50>1.8 M) of the CYP450 enzymes. Finally, this
compound was administered orally to Sprague-Dawley male rats,
exhibiting an oral bioavailability F=14% with an in vivo half-life
of 3.4 hours.
11-(S)
8.5 ± 0.1
3.2
N
N
H
(S)
(S)
12
13
14
8.2 ± 0.3
8.2 ± 0.0
7.3 ± 0.2
4.8
6.1
50
N
N
N
N
H
N
(S)
(S)
H
To evaluate the in vivo activity of this class of molecules, a
single dose (0.1, 1.0, or 10 mg/kg) of compound 36 (rat sst2
EC50=4.5 nM) or vehicle were subcutaneously administered to
Sprague-Dawley male rats. Growth hormone-releasing hormone
(GHRH) challenges (3 g GHRH/rat) were administered
intravenously to induce secretion of growth hormone (GH) at 1 h,
3 h, and 8 h post-dose. Five minutes after each GHRH challenge,
plasma was collected to measure GH levels. GH was assayed using
an in-house sandwich ELISA and normalized to the GHRH-
induced GH secretion in the presence of vehicle at each time point.
Dose-dependent suppression of GHRH-induced GH secretion was
observed at each time point (Figure 2). The plasma concentration
of compound 36 at 1 h, 3 h and 8 h time points was also measured,
and the results are summarized in Figure 3. These results
demonstrated that a nonpeptide sst2 agonist with good plasma
stability can effectively suppress the secretion of GH over longer
periods.
NH2
O
15
6.4 ± 0.2
383
N
N
H
aPurity of compounds is more than 95% as determined by
LCMS or 1H NMR analysis. bpEC50 is average of two or more
independent measurements. For EC50 values >1000 nM, the
c
pEC50 is reported as < 6.
Once the preferred basic side chain was identified, we explored
the modification of the phenol ring, and the results are summarized
in Table 2. Compared to compound 12, the non-substituted phenyl
analog 16 showed a dramatic decrease in sst2 activity, which
suggests a proton acceptor/donor could be required in this part of
binding. We examined more diverse functional groups as
demonstrated by analogs 17-22, however, such variations led to
reduced activity. In addition, shifting the hydroxyl group from the
meta position to para position also reduced sst2 activity (23). We
next added methyl (24), fluoro (25), and CF3 (26) at the meta
position of hydroxyl group to reduce the possibility of
reactive/toxic metabolite formation.13 This variation was well
tolerated, and the latter two compounds showed improved sst2
activities with EC50 values of 1.2 nM and 0.4 nM, respectively.
120
100
80
60
40
20
0
We next studied the impact of altering the 3, 5-dimethylphenyl
group (Table 2). Initially, we tried to replace both methyl groups
with metabolically stable CF3, OCF3, or F (27-30), but none of
these compounds maintained sst2 activity comparable to the 3,5-
dimethylphenyl group. However, when only one of the two methyl
groups were replaced with halogen, as exemplified by 31-32, the
sst2 activity was fully restored. In addition, the chloro-substituent
was found to be more potent. We also prepared compound 33,
bearing a 3,5-dichlorophenyl group, which resulted in a less potent
compound than 32. Interestingly, polar substituents like a
hydroxymethyl group can also be tolerated (34). Introducing F
substituent to the ortho position of this aromatic ring substantially
decreased sst2 functional activity, probably due to the disruption
of planar molecular structure (35). We also attempted to replace
this entire carbocyclic ring with 5-membered or 6-membered
heterocyclic rings (i.e. pyrazole and pyridine) but such alterations
resulted in complete loss of sst2 activity (data not shown).
1h
3h
8h
Time Post-Dose
GHRH+Vehicle
GHRH+0.1 mg/kg 36
GHRH+1 mg/kg 36
GHRH+10 mg/kg 36
Figure 2. GH suppression study of compound 36.
The optimized fragments evolved from previous rounds of SAR
were combined to furnish a small set of pyrido-pyrimidin-4-one
compounds and the representative example 36 is shown in Table
3. Compounds 36 exhibited sub-nanomolar sst2 activity (EC50=
0.45 nM), excellent selectivity against human sst1 and sst5
receptors, and moderate selectivity against human sst3 receptor.
However, it is also a potent sst4 agonist. Although compound 25,
26 and 32 showed human sst2 potency similar to 36, their
selectivity against human sst3 and sst4 receptors is worse. We also
evaluated the capability of this compound to induce receptor
internalization, which could relate with the loss of efficacy
observed in acromegaly patients treated with somatostatin peptide
analogs. Compound 36 is biased for Gi signaling over
internalization by >135-fold (Table 3). As part of our screens for
drug-like characteristics, we measured the inhibition of hERG ion
Figure 3. Plasma concentration of compound 36 in GH suppression
study.