Discovery of ML314, a brain penetrant nonpeptidic β-arrestin biased agonist of the neurotensin NTR1 receptor

Article abstract of DOI:10.1021/ml400176n

The neurotensin 1 receptor (NTR1) is an important therapeutic target for a range of disease states including addiction. A high-throughput screening campaign, followed by medicinal chemistry optimization, led to the discovery of a nonpeptidic β-arrestin biased agonist for NTR1. The lead compound, 2-cyclopropyl-6,7-dimethoxy-4-(4-(2-methoxyphenyl)-piperazin-1-yl)quinazoline, 32 (ML314), exhibits full agonist behavior against NTR1 (EC₅₀ = 2.0 μM) in the primary assay and selectivity against NTR2. The effect of 32 is blocked by the NTR1 antagonist SR142948A in a dose-dependent manner. Unlike peptide-based NTR1 agonists, compound 32 has no significant response in a Ca²⁺ mobilization assay and is thus a biased agonist that activates the β-arrestin pathway rather than the traditional G_q coupled pathway. This bias has distinct biochemical and functional consequences that may lead to physiological advantages. Compound 32 displays good brain penetration in rodents, and studies examining its in vivo properties are underway.

Full text of DOI:10.1021/ml400176n

Letter
pubs.acs.org/acsmedchemlett
Discovery of ML314, a Brain Penetrant Nonpeptidic β‑Arrestin Biased
Agonist of the Neurotensin NTR1 Receptor
Satyamaheshwar Peddibhotla,† Michael P. Hedrick,‡ Paul Hershberger,† Patrick R. Maloney,† Yujie Li,‡
Monika Milewski,‡ Palak Gosalia,‡ Wilson Gray,‡ Alka Mehta,† Eliot Sugarman,† Becky Hood,†
Eigo Suyama,† Kevin Nguyen,† Susanne Heynen-Genel,‡ Stefan Vasile,† Sumeet Salaniwal,‡
Derek Stonich,‡ Ying Su,‡ Arianna Mangravita-Novo,† Michael Vicchiarelli,† Gregory P. Roth,†
Layton H. Smith,† Thomas D. Y. Chung,‡ Glen R. Hanson,§ James B. Thomas,⊥ Marc G. Caron,∥
Lawrence S. Barak,*,∥ and Anthony B. Pinkerton*,‡
†Conrad Prebys Center for Chemical Genomics at Sanford-Burnham Medical Research Institute, Orlando, Florida 32827, United
States
‡Conrad Prebys Center for Chemical Genomics at Sanford-Burnham Medical Research Institute, La Jolla, California 92037, United
States
§Department of Pharmacology and Toxicology, University of Utah, 260 S. Campus Drive, Salt Lake City, Utah 84112, United States
∥Duke University Medical Center, Durham, North Carolina 27710, United States
⊥RTI International, 3040 E Cornwallis Road, Durham, North Carolina 27709, United States
S
* Supporting Information
ABSTRACT: The neurotensin 1 receptor (NTR1) is an
important therapeutic target for a range of disease states
including addiction. A high-throughput screening campaign,
followed by medicinal chemistry optimization, led to the
discovery of a nonpeptidic β-arrestin biased agonist for NTR1.
The lead compound, 2-cyclopropyl-6,7-dimethoxy-4-(4-(2-
methoxyphenyl)-piperazin-1-yl)quinazoline, 32 (ML314), ex-
hibits full agonist behavior against NTR1 (EC50 = 2.0 μM) in
the primary assay and selectivity against NTR2. The effect of
32 is blocked by the NTR1 antagonist SR142948A in a dose-
dependent manner. Unlike peptide-based NTR1 agonists,
compound 32 has no significant response in a Ca2+ mobilization assay and is thus a biased agonist that activates the β-arrestin
pathway rather than the traditional Gq coupled pathway. This bias has distinct biochemical and functional consequences that may
lead to physiological advantages. Compound 32 displays good brain penetration in rodents, and studies examining its in vivo
properties are underway.
KEYWORDS: NTR1, neurotensin, GPCR, SR142948A, quinazoline, agonist, β-arrestin bias, addiction
ethamphetamine addiction is a major public health
self-administered, and (c) the effects were associated with
nucleus accumbens dopamine D1 receptors.2
1
Mconcern in the United States. Due to high relapse rates
and the difficulty in breaking addiction, there is an urgent need
for new medications. Neurotensin (NT) receptors are expressed
on dopaminergic neurological pathways associated with reward,
and the neurotensin receptor 1 (NTR1) has been implicated as a
therapeutic target for the treatment of methamphetamine abuse.
In particular, peptide-based NTR1 agonists produce behaviors
that are exactly opposite to the psychostimulant effects observed
with methamphetamine, such as hyperactivity, neurotoxicity,
psychotic episodes, and cognitive deficits. These studies found
that in a methamphetamine self-administration rat model (a) the
substitution of the peptide NT agonist (Lys(CH2NH)lys-
Pro,Trp-tert-Leu-Leu-OEt) for methamphetamine did not
significantly affect motor activity but dramatically reduced lever
pressing associated with the drug, (b) the peptide agonist was not
NTR1 is a G protein coupled receptor (GPCR). Two distinct,
interdependent paradigms are associated with GPCR signaling.
In addition to the well-defined signaling cascades involving
heterotrimeric G proteins, recent advances in receptor
pharmacology have identified the importance of β-arrestins in
regulating alternative biochemical cascades that produce their
own unique biological effects.3 For example, in a mouse model,
Allen et al. developed a series of β-arrestin-2 biased agonists for
the D(2)R with antipsychotic properties and, most importantly, a
Received: May 13, 2013
Accepted: July 22, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ml400176n | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
reduced propensity to induce catalepsy-like standard neuroleptic
antagonists.4,5 These studies illustrate how ligand directed
signaling bias, in this case favoring β-arrestin, can ameliorate
undesirable biological outcomes. Downstream modulators of β-
arrestin/GPCR signaling are less well-characterized than their G
protein counterparts and, due to their potential as targets for
producing new medical therapies, are the subjects of increasing
numbers of investigations. Well-recognized β-arrestin partners
include the proteins Src, ERK, and Jnk. Their agonist-induced
interactions with β-arrestin are associated with clathrin-
compartmentalized signaling and the accumulation of ligand
activated β-arrestin/GPCR complexes in clathrin-coated pits.6
The determination as to whether a GPCR ligand is biased toward
or against β-arrestin may consequently be evaluated by following
these biochemical processes.7−10
Compound 5, a quinazoline derivative, was identified as a
singleton hit with an EC50 of <10 μM against NTR1, with an Emax
of 85% of the control neurotensin peptide. Subsequent medicinal
chemistry optimization relative to the primary β-arrestin assay
resulted in the discovery of 32 (ML314), the first nonpeptidic
small molecule full agonist against NTR1 (EC50 = 2.0 μM) with
>20× selectivity over NTR2 and GPR35. In this report we
discuss the design and synthesis of SAR analogues that led to the
discovery of 32, as well as its in vitro and in vivo pharmacological
profiling.
Analogues were synthesized from substituted methyl
anthranilates in three steps using previously reported methods
(Scheme 1).22,23 Acid-catalyzed condensation of methyl 2-
amino-4,5-dimethoxybenzoate (3a) with cyclobutane carbon-
itrile afforded the 4-hydroxyquinazoline intermediate 4a.
Chlorination of 4a followed by amination under microwave
heating with various substituted aryl piperazines provided the
primary hit compound 5 and quinazoline analogues 14−23
(Scheme 1A).
For a second point of diversity 3a was condensed with a range
of substituted nitriles to afford 4-hydroxyquinazolines (4b−4l),
which were converted to the corresponding quinazolines (25−
36) by sequential chlorination and amination with 1-(2-
methoxyphenyl)-piperazine hydrochloride (Scheme 1B). 2-(2-
Chloroethyl)-6,7-dimethoxyquinazolin-4-ol (4h) under the
above conditions resulted in the 2-vinyl analogue (31) via
elimination and compound 36 via double addition (Scheme 1B).
Using the above three-step route, differentially substituted
methyl anthranilates (3b−3e) were converted to compounds
37−41 via 4-hydroxyquinazolines (4m−q) (Scheme 1C).
Similarly, 2-cyclopropyl-6,7-dimethoxyquinazolin-4-ol (4i) was
subjected to chlorination and then treated with diverse 1-
substituted piperazines to access quinazolines 42−49 (Scheme
1D). All final compounds were obtained in moderate yields and
high purity needed to support SAR studies.
Although there have been reports11,12 of NTR1 antagonists,
there have been few reports of small molecule nonpeptide
agonists for NTR1, even though the tridecapeptide neurotensin
was characterized in the 1970s.13−15 Recently, two known
trisubstituted pyrazole antagonists, SR48692 (1a) and
SR142948A (1b), were modified to identify partial ago-
nists11,12,16,17 such as 2 (Figure 1). Additionally, two simple
Figure 1. Examples of known pyrazole based NTR1 antagonists 1a and
1b and agonist 2.
Compound 5, the singleton screening hit (EC50 = 5.9 μM and
Emax ∼85% of the NT(8−13) peptide response), was
resynthesized (Scheme 1A) and was confirmed equipotent to
the purchased material (Table 1; entry 5). A small set of
commercially available 2-phenylquinazolines (Table 1, entries
6−13) tested inactive except for the 2-methoxyphenylpiperazine
derived analogue 6, which displayed partial agonist activity (EC50
= 20 μM). In general we saw full agonist activity for all
compounds regardless of potency with some exceptions,
although in no case was the Emax below ∼70%. To expand the
SAR around the singleton hit 5, a focused set of analogues was
prepared to investigate the effect of various substituents and their
positions on the aryl piperazine moiety. This revealed that,
irrespective of the electronic nature, all substituents that were
smaller or comparable in size to the 2-OMe group retained
activity within 2−4 fold of compound 5 (Table 2; entries 14−
17). Much larger 2-nitro and basic 2-pyridyl groups were far less
potent than 5 (>6 fold; Table 2; entries 18, 19). Altering the
position of the methoxy group, and disubstitution, also led to a
significant loss of potency (>4−13 fold; Table 2; entries 20−23).
We next investigated the role of substituents at the 2-position of
the quinazoline ring occupied by the cyclobutyl ring in 5 (Table
3, entry 5). 2-H and 2-Me analogs incurred complete loss of
activity (Table 3, entry 24−25); however, the 2-benzyl analog
(Table 3, entry 26) was 2.3× less active than compound 5.
Combined with the fact that 2-phenyl quinazolines (Table 1,
entries 6−13) were mostly inactive, we speculated that a
methylene linker between the 2-substituent and the quinazoline
leucine derivatives unrelated to 1a, 1b, and 2 have shown partial
agonist activity for NTR1.18 However, there have been no
reports of in vivo activity associated with nonpeptide agonists.
Newly published X-ray crystallography data showing the
neurotensin fragment NT (8−13) in a NTR1 binding site may
assist the design of nonpeptide agonists and antagonists in the
future.19
To identify nonpeptide small molecule NTR1 agonists, we
carried out high content screening (HCS) of receptor/β-arrestin-
GFP complexes based on a β-arrestin conjugated green
fluorescent protein (GFP) reporter expressed in a U2OS cell
line (see Supporting Information for details). This assay
measured the quantification of β-arrestin/GFP puncta. A library
of 331,861 compounds from the MLSMR (Molecular Libraries
Small Molecule Repository)20 was tested in this high-content
primary screen. This campaign [PubChem AID 493055]21
identified 559 compounds with ≥40% activity of NT (8−13)
peptide, which is the prototypical reference agonist, at 20 μM.
Liquid samples (499 compounds) were resupplied and tested at
20 μM, and 184 compounds showed >40% activity. The
confirmed compounds were tested in a dose response format
in the HCS primary assay. Dose response testing of reconstituted
dry powders of 122 compounds that had an EC50 < 20 μM was
performed in a three-assay panel: the NTR1 HCS primary assay
to assess potency, and the NTR2 and GPR35 HCS assays to
ascertain selectivity against a related and unrelated GPCR.
B
dx.doi.org/10.1021/ml400176n | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
a
Scheme 1. Synthesis of NTR1 Agonists 32 and Analogues
a
Conditions:22 (i) alkyl carbonitrile, 4 M HCl (1,4-dioxane), 100 °C, 15 h; (ii) POCl3, reflux, 15 h; (iii) piperazine derivative (1.5 equiv), K2CO3 (3
equiv), 1,4-dioxane, microwave, 80 °C, 1.5 h, 1−46% overall yield over three steps.
Table 1. SAR of Quinazoline Agonists of NTR1, Compounds
6−13
Table 2. SAR of Quinazoline-Based Agonists of NTR1,
Compounds 14−23
a
a
EC50
Emax
a
a
compound
R
EC50 (μM)
Emax (%)
compound
R1
R2
R3
(μM)
(%)
NT(8−13)
5
0.001 (>4)
5.9 0.5 (10)
12 1.4
100.0
85.3
NT(8−13)
5
0.001 (>4)
100.0
85.3
2-methoxyphenyl
phenyl
6,7-di-OMe cyclo-
butyl
2-OMe 5.9 0.5
(10)
14
15
16
17
18
19
20
21
22
23
77.0
o-tolyl
15 3.2
916
6
6,7-di-OMe phenyl
6,7-di-OMe phenyl
6,7-di-OMe phenyl
6,7-di-OMe phenyl
6,7-di-OMe phenyl
2-OMe 20 11 (5) 78.0
2-fluorophenyl
2-chlorophenyl
pyridin-2-yl
18 1.0 (3)
12 2.5
97.2
7
H
>80
>80
>80
>80
70.5
8
2-F
4-F
2-Cl
26 2 (6)
75 1 (2)
17 1.0 (2)
46 18 (3)
23 0.4 (6)
61 5.2
91.4
9
2-nitrophenyl
3-methoxyphenyl
4-methoxyphenyl
2,4-dimethoxyphenyl
2,6-dimethylphenyl
100.0
74.9
10
11
6-OEt, 7-
OMe
phenyl
phenyl
phenyl
2-OMe >80
100.0
103.3
100.0
12
13
6-OEt, 7-
OMe
H
>80
>80
6-OEt, 7-
OMe
2-F
a
HCS NTR1 potency measured relative to the EC100 (100 nM) of the
a
NT(8−13) peptide control average SEM (n = 4 unless otherwise
noted); Emax was calculated as the % of the response obtained with
NT(8−13) peptide. None of the compounds from this series showed
activity in the NTR2 (>80 μM) and GPR35 (>40 μM) counterscreens.
HCS NTR1 potency measured relative to the EC100 (100 nM) of the
NT(8−13) peptide control average SEM (n = 4 unless otherwise
noted); Emax was calculated as the % of the response obtained with
NT(8−13) peptide. None of the compounds from this series showed
activity in the NTR2 (>80 μM) and GPR35 (>40 μM) counterscreens.
compound 32 (ML314, Table 3), which is a full agonist (EC50 =
2.0 μM) with 3-fold improvement in activity over 5.
might improve potency and hence generated analogs with alkyl,
branched alkyl and vinyl substituents (Table 3, entries 27−31).
The 2-isopropyl and 2-vinyl derivatives were found to be
equipotent to compound 5 (Table 3, entries 29 and 31). This
result prompted the synthesis of the 2-cyclopropyl analogue,
This met the probe criteria established for the NIH’s MLPCN
(Molecular Libraries Probe Production Centers Network)
program.24 The larger 2-cyclopentyl (Table 3, entry 33) was
equipotent, and the extended 2-methylcyclopropyl analogue
(entry 34) was 2.4× less potent than 5. Furthermore,
C
dx.doi.org/10.1021/ml400176n | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Table 3. SAR of Quinazoline-Based Agonists of NTR1,
Compounds 24−36
Table 4. SAR of Quinazoline-Based Agonists of NTR1,
Compounds 37−41
a
Emax
a
compound
R1
R2
EC50 (μM)
(%)
NT(8−13)
5
0.001 (>4)
100.0
85.3
6,7-di-OMe
cyclobutyl
5.9 0.5
(10)
37
38
39
40
6-OMe
7-OMe
H
cyclobutyl
cyclobutyl
cyclobutyl
cyclobutyl
10 1.6
101.9
111.5
100.0
87.0
30 0 (3)
23 3.9
6,7-OCH2O-
dioxolane
34 17 (3)
41
6-OMe
cyclopropyl 4.1 0.5
95.7
a
HCS NTR1 potency measured relative to the EC100 (100 nM) of the
NT(8−13) peptide control average SEM (n = 4 unless otherwise
noted); Emax was calculated as the % of the response obtained with
NT(8−13) peptide. None of the compounds from this series showed
activity in the NTR2 (>80 μM) and GPR35 (>40 uM) counterscreens.
Table 5. SAR of Quinazoline-Based Agonists of NTR1,
Compounds 42−49
a
a
compound
R
EC50 (μM)
Emax (%)
a
HCS NTR1 potency measured relative to the EC100 (100 nM) of the
NT(8−13)
0.001 (>4)
2.0 0.1 (8)
6.1 0.4
20 2.2
>80
100.0
104.7
98.4
NT(8−13) peptide control average SEM (n = 4 unless otherwise
noted); Emax was calculated as the % of the response obtained with
NT(8−13) peptide. None of the compounds from this series showed
activity in the NTR2 (>80 μM) and GPR35 (>40 μM) counterscreens.
32
42
43
44
45
46
47
48
49
2-methoxyphenyl
2-ethoxyphenyl
2-chlorophenyl
benzenesulfonyl
4-methylbenzyl
2-methoxybenzyl
benzoyl
100.0
25 3.2
35 4.4
45 (1)
100.0
100.0
100.0
100.0
incorporation of basic nitrogen containing substituents (Table 3,
entries 35, 36) resulted in significantly diminished activity.
We also investigated the importance of 6- and 7-OMe groups.
Thus, systematic removal of 7-OMe, 6-OMe, and both 6- and 7-
OMe groups (Table 4, entries 37−39), revealed that 6-OMe is
key to activity and its removal leads to a ∼4−6 fold loss in activity
(Table 4, entries 38 and 39). This trend was maintained in the
corresponding 2-cyclopropyl-6-methoxy analogue of 32 (Table
4, entry 41). When the 6- and 7-OMe groups were joined as part
of a dioxolane ring (Table 4, entry 40), the potency dropped by
over 6-fold compared to 5.
2-methoxybenzoyl
H
68 5.5 (3)
>80
a
HCS NTR1 potency measured relative to the EC100 (100 nM) of the
NT(8−13) peptide control average SEM (n = 4 unless otherwise
noted); Emax was calculated as the % of the response obtained with
NT(8−13) peptide. None of the compounds from this series showed
activity in the NTR2 (>80 μM) and GPR35 (>40 uM) counterscreens.
Having the more potent and full agonist 32 in hand, we
investigated the aryl piperazine moiety with the goal of further
improving the potency. Replacement of the 2-OMe group with
the larger 2-OEt or an electronegative 2-Cl led to >3−10 fold loss
in potency compared to 32 (Table 5, entries 42 and 43). This was
a more pronounced effect than that observed for the primary hit
compound 5. Replacement of the N-arylpiperazine group in 32
with benzenesulfonyl, substituted benzyl, or benzoyl groups led
to significant erosion of activity (>10−40 fold; Table 5, entries
44−49), demonstrating the importance of a 2-methoxyphenyl
group in this region of the molecule.
Compound 32 was further profiled in an orthogonal NTR1
activation assay based on DiscoveRx β-arrestin complementation
and a downstream functional assay measuring Ca2+ Flux
(ChanTest). The agonist activity of 32 in the primary NTR1
HCS assay was confirmed in the DiscoveRx β-arrestin assay
(EC50 = 3.41 μM; Emax = 86.6%). Compound 32 was inactive
(EC50 = >80 uM, Figure 2A, B) in the NTR1 Ca2+ Flux assay. In
contrast, the known16 partial agonist 2 showed activity (EC50
=
0.75 μM) and only partial agonism (Emax < 80%) in the NTR1
HCS assay. Interestingly, 2 was inactive in the DiscoveRx β-
arrestin assay (EC50 > 33 μM) but showed evidence of traditional
Gq signaling in the Ca2+ flux assay (EC50 < 156 nM, Emax = 63%).
The latter result was consistent with the published data for 2
(EC50 = 67 nM, Emax = 54%).16,17 Thus, 32 appears to be a biased
agonist operating via the β-arrestin pathway rather than the
traditional Gq coupled pathway. Specificity of binding of 32 to
NTR1 was evaluated by the ability of the antagonist 1b to block
the response of 32 in the HCS assay (Figure 2C). Previous
studies have eported the inhibition of binding of [125I]-NT
peptide to the NTR1 receptor by 1b (IC50 = 0.24 nM).25
Preincubation with 1b inhibited compound 32 (EC100 = 10 μM)
D
dx.doi.org/10.1021/ml400176n | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Figure 2. (A) Dose response of the NT(8−13) peptide in the NTR1 HCS, β-arrestin, and Ca2+ Flux assays. (B) Compound 32 (ML314) in the NTR1
HCS, β-arrestin, and Ca2+ Flux assays, and the NTR2 HCS assay. Percent efficacy in each assay is measured relative to the NT peptide control. (C)
Inhibitory dose response of NTR1 antagonist 1b in the presence of an EC100 concentration (10 μM) of compound 32; data points/curves represent the
average of all dose response experiment runs.
mediated NTR1 activation with an IC50 of 50.1 nM,
demonstrating that 32 is inhibited from forming aggregates by
1b in a dose-dependent manner. This result supports the
mechanistic hypothesis that 32 acts via NTR1 binding.
The low metabolic stability may be associated with the presence
of several unsubstituted aryl and alkyl positions and Ar-OMe
ethers susceptible to oxidation, hydrolysis, conjugation, and
other metabolic reactions. Compound 32 had a >15-fold window
for toxicity (LC50 = 30 μM) toward human hepatocytes.
Compound 32 was profiled for its mouse pharmacokinetics
and showed hepatic clearance and moderate exposure after
intraperitoneal dosing (clearance 81 mL/min/kg, Vdss 6.22 L/kg,
A PubChem analysis of compound 5 (CID1230852) displayed
very low cross reactivity, with NTR1 being the only receptor
activity <10 μM in over 600 assays.26 In addition to NTR2 and
GPR35, compound 32 was tested across a range of GPCRs using
functional high content or β-arrestin based screens and was
found to have no cross reactivity.27 Compound 32 was also
submitted to the Psychoactive Drug Screening Program
(PDSP)28 for testing in a GPCR binding assay panel (∼40
receptors) and was found moderately promiscuous at 10 μM,
with Kis < 10 μM on seven receptors.29 A follow-up study at
Panlabs/Ricerca in their lead profiling panel confirmed activity in
only two of those receptors (MOR, 86% at 10 μM and σ1 69% at
10 μM), In addition, 32 showed moderate binding across a range
of adrenergic receptors (α1a, α1B, α1D, α2A 63−100% at 10 μM) in
the Panlabs panel. Other analogues were not examined for their
binding cross reactivity.
t
1/2 1.93 h after a 2 mg/kg intravenous dose, Cmax 763 ng/mL, t1/2
2.58 h, AUC 1223 ng·h/mL after a 10 mg/kg intraperitoneal
dose) (see Supporting Information for details). However, 32
displayed excellent brain penetration, with brain levels of 924 ng/
mL and 1506 ng/mL at 1 h after a 10 mg/kg or 30 mg/kg
intraperitoneal dose (brain/plasma 1.3 or 1.6, respectively).
In conclusion, a high-throughput screen of the MLSMR
collection using a NTR1 β-arrestin recruitment HCS assay
identified a singleton partial agonist, quinazoline 5, which was
optimized via medicinal chemistry to discover a potent full
agonist 32. Compound 32 was >20× selective over NTR2 and
was inhibited from forming ligand activated β-arrestin/NTR1
complexes by antagonist 1b in a dose-dependent manner.
Compound 32 had no significant activity in a Ca2+ flux assay, in
contrast to previously known agonists, and is therefore an
apparent biased agonist operating via the β-arrestin pathway
rather than the traditional Gq coupled pathway. Compound 32
displayed moderate pharmacokinetics and showed good BBB
penetration in mice. Although 32 displayed some promiscuity in
binding assays to other GPCRs, it showed no functional activity
at other receptors tested. We are currently investigating its
apparent signaling bias and associated physiological consequen-
ces. Current efforts to optimize 32 further and improve its in vivo
profile are underway.
In vitro pharmacology screening (Table 6) was also conducted
for 32. Consistent with its aqueous solubility, 32 exhibited high
Table 6. Summary of in Vitro ADME/T Properties of NTR1
Agonist 32
aqueous solubility (μM) in pION buffer; pH 5.0/6.2/7.4 >297/21.4/1.2
aqueous solubility (μM) in 1× PBS, pH 7.4
1.1
PAMPA permeability, Pe (× 10−6 cm/s), donor pH:
1163/2145/2093
5.0/6.2/7.4, acceptor pH: 7.4
BBB-PAMPA permeability, Pe (× 10−6 cm/s), donor pH: 399
7.4, acceptor pH: 7.4
plasma protein binding
(% bound)
human 1 μM/10 μM 99.5/99.2
mouse 1 μM/10 μM 99.7/98.9
ASSOCIATED CONTENT
* Supporting Information
Detailed experimental procedures, annotated H NMR and
HRMS data for the described compounds, as well as graphical
depictions of the PK and in vivo data. This material is available
free of charge via the Internet at http://pubs.acs.org.
plasma stability (% remaining at 3 h) human/mouse
100/99.6
■
hepatic microsome stability (% remaining at 1 h) human/ 1.4/0.2
mouse
S
1
toxicity toward Fa2N-4 immortalized human hepatocytes 29.6
over 24 h, LC50 (μM)
permeability in the Parallel Artificial Membrane Permeability
Assay (PAMPA) with increasing pH of the donor compart-
ment.30 When incubated with an artificial membrane that models
the blood−brain barrier (BBB), 32 was highly permeable.
Compound 32 was highly plasma protein bound and exhibited
very high plasma stability but was metabolized rapidly when
incubated in vitro with human and mouse liver homogenates.
AUTHOR INFORMATION
Corresponding Author
*L.S.B.: phone, 01-919-684-6245. E-mail: L.Barak@cellbio.duke.
edu. A.B.P.: phone, 01-858-795-5320. E-mail: apinkerton@
sanfordburnham.org.
■
E
dx.doi.org/10.1021/ml400176n | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
(15) Griebel, G.; Holsboer, F. Neuropeptide receptor ligands as drugs
for psychiatric diseases: the end of the beginning? Nat. Rev. Drug
Discovery 2012, 11, 462−478.
(16) Thomas, J. B.; Navarro, H.; Warner, K. R.; Gilmour, B. The
identification of nonpeptide neurotensin receptor partial agonists from
the potent antagonist SR48692 using a calcium mobilization assay.
Bioorg. Med. Chem. Lett. 2009, 19, 1438−1441.
(17) Thomas, J. B.; Runyon, S. P. Compounds Active at the
Neurotensin Receptor. Compounds active at the neurotensin receptor.
World Patent WO 2011/156557 A2, December 15, 2011.
(18) Fan, Y.; Lai, M. H.; Sullivan, K.; Popiolek, M.; Andree, T. H.;
Dollings, P.; Pausch, M. H. The identification of neurotensin NTS1
receptor partial agonists through a ligand-based virtual screening
approach. Bioorg. Med. Chem. Lett. 2008, 18, 5789−5791.
(19) White, J. F.; Noinaj, N.; Shibata, Y.; Love, J.; Kloss, B.; Xu, F.;
Gvozdenovic-Jeremic, J.; Shah, P.; Shiloach, J.; Tate, C. G.;
Grisshammer, R. Structure of the agonist-bound neurotensin receptor.
Nature 2012, 490, 508−513.
(20) NIH’s Molecular Libraries Small Molecule Repository (MLSMR)
collection. http://mli.nih.gov/mli/compound-repository/mlsmr-
compounds (accessed May 3, 2013).
(21) National Center for Biotechnology Information. PubChem
BioAssay Database; 493055, Source = Sanford-Burnham Center for
Chemical Genomics, Sanford, Burnham Medical Research Institute, San
Diego, CA. http://pubchem.ncbi.nlm.nih.gov/assay/assay.cgi?aid=
493055 (accessed May 3, 2013).
(22) Bridges, A. J.; Zhou, H.; Cody, D. R.; Rewcastle, G. W.;
McMichael, A.; Showalter, H. D. H.; Fry, D. W.; Kraker, A. J.; Denny, W.
A. Tyrosine Kinase Inhibitors. 8. An Unusually Steep Structure−Activity
Relationship for Analogues of 4-(3-Bromoanilino)-6,7-dimethoxyqui-
nazoline (PD 153035), a Potent Inhibitor of the Epidermal Growth
Factor Receptor. J. Med. Chem. 1996, 39, 267−276.
(23) Sirisoma, N.; Pervin, A.; Zhang, H.; Jiang, S.; Willardsen, J. A.;
Anderson, M. B.; Mather, G.; Pleiman, C. M.; Kasibhatla, S.; Tseng, B.;
Drewe, J.; Cai, S. X. Discovery of N-(4-Methoxyphenyl)-N,2-
dimethylquinazolin-4-amine, a Potent Apoptosis Inducer and Effica-
cious Anticancer Agent with High Blood Brain Barrier Penetration. J.
Med. Chem. 2009, 52, 2341−2351.
(24) ML314 has been declared a probe via the Molecular Libraries
Probe Production Centers Network (MLPCN) and is available through
the network, see: http://mli.nih.gov/mli/.
(25) Schaeffer, P.; Laplace, M. C.; Bernat, A.; Prabonnaud, V.; Gully,
D.; Lespy, L.; Herbert, J. M. SR142948A is a potent antagonist of the
cardiovascular effects of neurotensin. J. Cardiovasc. Pharmacol. 1998, 31,
545−50.
(26) Other activities <10 μM: inhibition of mycobacterium TB H37Rv
(3.90 μM, AID: 1626), PAX8 cytotoxicity (0.26 μM, AID: 687029),
inhibition of HCV growth (3.16 μM, AID: 651820).
(27) Other GPCRs include GPR55, APJ, k-opiod, CCR6, and CXCR6.
(28) National Institute of Mental Health’s Psychoactive Drug
Screening Program, Contract No. HHSN-271-2008-00025-C (NIMH
PDSP). The NIMH PDSP is directed by Bryan L. Roth MD, Ph.D. at the
University of North Carolina at Chapel Hill and Project Officer Jamie
Driscoll at NIMH, Bethesda, MD, USA.
(29) Follow up studies revealed Kis of 0.4 μM (5-HT1A) 1.0 μM (5-
HT3), 0.93 μM (5-HT7), 1.42 μM (H1) 1.54 μM (MOR), 2.54 μM
(PBR), and 0.41 μM (sigma1).
(30) Avdeef, A.; Nielsen, P. E.; Tsinman, O. PAMPA-a drug absorption
in vitro model 11. Matching the in vivo unstirred water layer thickness by
individual-well stirring in microtitre plates. Eur. J. Pharm. Sci. 2004, 22,
365−374.
Author Contributions
S.P., M.P.H., and P.H. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by NIH grants 1 R03 MH089653-01 to
L.S.B., 5P30DA029925 (M.G.C., L.S.B.), R01 DA031833 to
G.R.H., and an NIH Molecular Libraries grant (U54 HG005033-
03) to the Conrad Prebys Center for Chemical Genomics at the
Sanford-Burnham Medical Research Institute, one of the
comprehensive centers of the NIH Molecular Libraries Probe
Production Centers Network (MLPCN).
REFERENCES
■
(1) Results from the 2011 National Survey on Drug Use and Health:
Summary of National Findings; NSDUH Series H-44, HHS Publication
No. (SMA) 12−4713. Substance Abuse and Mental Health Services
Administration: Rockville, MD, 2012. http://www.samhsa.gov/data/
NSDUH/2011SummNatFindDetTables/ Index.aspx (accessed Nov 5,
2012).
(2) Hanson, G. R.; Hoonakker, A. J.; Alburges, M. E.; McFadden, L.
M.; Robson, C. M.; Frankel, P. S. Response of limbic neurotensin
systems to methamphetamine self-administration. Neuroscience 2012,
203, 99−107.
(3) DeWire, S. M.; Ahn, S.; Lefkowitz, R. J.; Shenoy, S. K. Beta-
arrestins and cell signaling. Annu. Rev. Physiol. 2007, 69, 483−510.
(4) Allen, J. A.; Yost, J. M.; Setola, V.; Chen, X.; Sassano, M. F.; Chen,
M.; Peterson, S.; Yadav, P. N.; Huang, X.; Feng, B.; Jensen, N. H.; Che,
X.; Bai, X.; Frye, S. V.; Wetsel, W. C.; Caron, M. G.; Javitch, J. A.; Roth,
B. L.; Jin, J. Discovery of β-Arrestin-Biased Dopamine D2 Ligands for
Probing Signal Transduction Pathways Essential for Antipsychotic
Efficacy. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 18488−18493.
(5) Rajagopal, S.; Rajagopal, K.; Lefkowitz, R. J. Teaching old receptors
new tricks: biasing seven-transmembrane receptors. Nat. Rev. Drug
Discovery 2010, 9, 373−386.
(6) Lefkowitz, R. J.; Shenoy, S. K. Transduction of Receptor Signals by
β-Arrestins. Science 2005, 308, 512−517.
(7) Kenakin, T. The potential for selective pharmacological therapies
through biased receptor signalling. BMC Pharmacol. Toxicol. 2012, 13, 3.
(8) Barak, L. S.; Peterson, S. Modeling of bias for the analysis of
receptor signaling in biochemical systems. Biochemistry 2012, 51, 1114−
25.
(9) Barak, L. S.; Ferguson, S. S.; Zhang, J.; Caron, M. G. A beta-
arrestin/green fluorescent protein biosensor for detecting G protein-
coupled receptor activation. J. Biol. Chem. 1997, 272, 27497−500.
(10) Whalen, E. J.; Rajagopal, S.; Lefkowitz, R. J. Therapeutic potential
of β-arrestin and G protein-biased agonists. Trends Mol. Med. 2011, 17,
126−139.
(11) Gully, D.; Labeeuw, B.; Boigegraiin, R.; Oury-Donat, F.; Bachy,
A.; Poncelet, M.; Steinberg, R.; Suaud-Chagny, M. F.; Santucci, V.; Vita,
N.; Pecceu, F.; Labbe-Jullio, C.; Kitabgi, P.; Soubrie, P.; Le Fur, G.;
Maffrand, J. P. Biochemical and Pharmacological Activities of SR
142948A, a New Potent Neurotensin Receptor Antagonist. J. Pharmacol.
Exp. Ther. 1997, 280, 802−812.
(12) Gully, D.; Canton, M.; Boigegraiin, R.; Jeanjean, F.; Molimard, J.-
C.; Poncelet, M.; Gueudet, C.; Heaulme, M.; Leyris, R.; Brouard, A.;
Pelaprat, D.; Labbe-Jullio, C.; Mazellai, J.; Soubrie, P.; Rostene, W.;
Kitabgi, P.; Le Fur, G. Biochemical and pharmacological profile of a
potent and selective nonpeptide antagonist of neurotensin receptor.
Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 65−69.
(13) Carraway, R.; Leeman, S. E. The isolation of a new hypotensive
peptide, neurotensin, from bovine hypothalamus. J. Biol. Chem. 1973,
238, 6854−6861.
(14) Carraway, R.; Leeman, S. E. Characterization and radio-
immunoassayable neurotensin in the rat. J. Biol. Chem. 1976, 251,
7045−7052.
F
dx.doi.org/10.1021/ml400176n | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX