Discovery of novel brain permeable and G protein-biased beta-1 adrenergic receptor partial agonists for the treatment of neurocognitive disorders

Article abstract of DOI:10.1371/journal.pone.0180319

The beta-1 adrenergic receptor (ADRB1) is a promising therapeutic target intrinsically involved in the cognitive deficits and pathological features associated with Alzheimer’s disease (AD). Evidence indicates that ADRB1 plays an important role in regulating neuroinflammatory processes, and activation of ADRB1 may produce neuroprotective effects in neuroinflammatory diseases. Novel small molecule modulators of ADRB1, engineered to be highly brain permeable and functionally selective for the G protein with partial agonistic activity, could have tremendous value both as pharmacological tools and potential lead molecules for further preclinical development. The present study describes our ongoing efforts toward the discovery of functionally selective partial agonists of ADRB1 that have potential therapeutic value for AD and neuroinflammatory disorders, which has led to the identification of the molecule STD-101-D1. As a functionally selective agonist of ADRB1, STD-101-D1 produces partial agonistic activity on G protein signaling with an EC₅₀ value in the low nanomolar range, but engages very little beta-arrestin recruitment compared to the unbiased agonist isoproterenol. STD-101-D1 also inhibits the tumor necrosis factor α (TNFα) response induced by lipopolysaccharide (LPS) both in vitro and in vivo, and shows high brain penetration. Other than the therapeutic role, this newly identified, functionally selective, partial agonist of ADRB1 is an invaluable research tool to study mechanisms of G protein-coupled receptor signal transduction.

Full text of DOI:10.1371/journal.pone.0180319

RESEARCH ARTICLE
Discovery of novel brain permeable and G
protein-biased beta-1 adrenergic receptor
partial agonists for the treatment of
neurocognitive disorders
Bitna Yi1, Alam Jahangir1, Andrew K. Evans1, Denise Briggs1, Kristine Ravina1,
Jacqueline Ernest1, Amir B. Farimani2, Wenchao Sun3, Jayakumar Rajadas3,
Michael Green1, Evan N. Feinberg2, Vijay S. Pande2, Mehrdad Shamloo1*
1 Department of Neurosurgery, Stanford University School of Medicine, Palo Alto, California, United States of
America, 2 Department of Chemistry, Stanford University, Stanford, California, United States of America,
3 Biomaterials and Advanced Drug Delivery Laboratory, Stanford University School of Medicine, Palo Alto,
California, United States of America
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
* mshamloo@stanford.edu
Abstract
OPEN ACCESS
The beta-1 adrenergic receptor (ADRB1) is a promising therapeutic target intrinsically
involved in the cognitive deficits and pathological features associated with Alzheimer’s dis-
ease (AD). Evidence indicates that ADRB1 plays an important role in regulating neuroin-
flammatory processes, and activation of ADRB1 may produce neuroprotective effects in
neuroinflammatory diseases. Novel small molecule modulators of ADRB1, engineered to be
highly brain permeable and functionally selective for the G protein with partial agonistic
activity, could have tremendous value both as pharmacological tools and potential lead mol-
ecules for further preclinical development. The present study describes our ongoing efforts
toward the discovery of functionally selective partial agonists of ADRB1 that have potential
therapeutic value for AD and neuroinflammatory disorders, which has led to the identifica-
tion of the molecule STD-101-D1. As a functionally selective agonist of ADRB1, STD-101-
D1 produces partial agonistic activity on G protein signaling with an EC50 value in the low
nanomolar range, but engages very little beta-arrestin recruitment compared to the unbi-
ased agonist isoproterenol. STD-101-D1 also inhibits the tumor necrosis factor α (TNFα)
response induced by lipopolysaccharide (LPS) both in vitro and in vivo, and shows high
brain penetration. Other than the therapeutic role, this newly identified, functionally selec-
tive, partial agonist of ADRB1 is an invaluable research tool to study mechanisms of G pro-
tein-coupled receptor signal transduction.
Citation: Yi B, Jahangir A, Evans AK, Briggs D,
Ravina K, Ernest J, et al. (2017) Discovery of novel
brain permeable and G protein-biased beta-1
adrenergic receptor partial agonists for the
treatment of neurocognitive disorders. PLoS ONE
12(7): e0180319. https://doi.org/10.1371/journal.
pone.0180319
Editor: Vardan Karamyan, Texas Tech University,
UNITED STATES
Received: February 7, 2017
Accepted: June 14, 2017
Published: July 26, 2017
Copyright: © 2017 Yi et al. This is an open access
article distributed under the terms of the Creative
Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in
any medium, provided the original author and
source are credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files.
Funding: The present study was supported by the
National Institute of Neurological Disorders and
Stroke (NS069375 and NS097945; MS) (https://
www.ninds.nih.gov). The funder had no role in
study design, data collection and analysis, decision
to publish, or preparation of the manuscript.
Introduction
Adrenergic receptors (ADRs) are G protein-coupled receptors (GPCR) that mediate the cen-
tral and peripheral effects of noradrenaline (NA) and adrenaline [1, 2]. They are widely
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
1 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Competing interests: The authors have declared
expressed throughout the body and play an important role in regulating multiple physiological
processes including cardiac muscle contraction, airway reactivity, cognition, arousal, stress-
related behavior, and inflammation [2–5]. Multiple subtypes of ADRs exist. Each subtype is
expressed in distinct patterns and involved in multiple physiological processes [2, 6]. There-
fore, ligands that selectively target one subtype will be valuable both as research tools to iden-
tify the roles of different ADR subtypes and as potential therapeutic agents for multiple
diseases related to dysfunction of the NA and adrenaline systems.
that no competing interests exist.
Among the many subtypes of ADRs, the beta-1 adrenergic receptor (ADRB1) is an impor-
tant target in multiple therapeutic areas. For example, ADRB1 antagonists (beta-blockers)
have been used to treat cardiovascular disease since the 1960s and remain one of the most
commonly used drugs today [7, 8]. Our laboratory has also identified therapeutic potential of
ADRB1 in the treatment of Alzheimer’s disease (AD). For example, we previously demon-
strated that activation of ADRB1 reverses AD-like cognitive deficits in transgenic mice overex-
pressing human amyloid beta protein precursor (APP) [9]. Similarly, we have shown that
acute activation of ADRB1 rescues the contextual memory and spatial memory deficits
observed in the Ts65Dn mouse model of Down syndrome, which displays an accelerated AD-
like pathology [10]. In addition to the cognitive enhancing effects, we also observed that activa-
tion of ADRB1 attenuates pathological features of AD including beta-amyloid burden, tau
pathology, and neuroinflammation [11]. Collectively, our discovery suggests that the ADRB1
agonist may have therapeutic potential for AD as it can address both cognitive symptoms and
AD pathology. In addition to its involvement in AD, accumulating evidence suggests that the
NA system and ADRB1 play a critical role in regulating neuroimmune responses [12–14]. By
regulating the neuroinflammatory process, ADRB1 ligands may produce therapeutic benefits
for the diseases associated with neuroinflammation. Thus, we have focused our attention on
the development of ADRB1 agonists as potential therapeutic agents for AD and neuroinflam-
matory diseases.
In the classical view of GPCR signaling, activation of ADRB1 leads to stimulation of two
ubiquitous and generic mechanisms: G-protein signaling and β-arrestin signaling [15].
Recently, however, it has become clear that agonists can show biased activation of signaling
pathways. The ability of ligands to activate a receptor and produce responses in a pathway-
dependent manner has been termed “signaling bias” or “functional selectivity” [16, 17]. As G
proteins and β-arrestins mediate distinct physiological processes, biased agonists are expected
to provide improved therapeutic selectivity with reduced adverse effects.
In search of compounds having therapeutic potential for AD and neuroinflammatory dis-
eases, we sought to identify G protein-biased partial agonists of ADRB1. As partial agonists,
these compounds would have more subtle effects in the periphery compared to full agonists,
yet be efficacious enough to produce therapeutic benefits for AD and neuroinflammatory dis-
ease. By selectively activating ADRB1 G-protein signaling with minimal to no activity on β-
arrestin signaling, these compounds would also specifically target the disease relevant signaling
pathways without causing the agonist induced tolerance.
Xamoterol (ICI118578) is one of the most selective agonists of ADRB1 reported to date (Fig
1). It is a partial agonist that produces approximately 50% efficacy compared to the full agonist
isoproterenol, and was once used for the treatment of cardiovascular disease. However, xamo-
terol is an exceptionally polar compound (clogP = 0.4) with only 5% oral bioavailability in
humans [18]. In addition, its central nervous system (CNS) penetration is low due to its polar-
ity. Thus, the focus of our efforts was to develop ADRB1 ligands that (1) are partial agonists of
ADRB1, (2) exhibit functional selectivity for the cAMP signaling cascade with minimal or no
β-arrestin signaling, and (3) show improved brain penetration, using xamoterol as a lead
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
2 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 1. Structure of (S)-xamoterol and two sites of modification.
https://doi.org/10.1371/journal.pone.0180319.g001
compound. Here, we report the discovery of a series of compounds, including the highly
potent and selective drug-like ADRB1 partial agonist STD-101-D1.
Materials and methods
General chemistry and procedure for the preparation of compounds
All compounds were obtained at ꢀ 95% purity as determined by HPLC/UV analyses. HPLC
analyses were run on an Agilent 1100HPLC using a reverse phase Hypersil GOLD PFP 3μm
column with a gradient mobile phase of acetonitrile and water containing 0.01% trifluoroacetic
acid. LCMS analyses were run on a Shimadzu LC-10ATvp-API 150EX with an Agilent ZOR-
BAX Eclipse XDB-C18 3.5uM column with a gradient mobile phase of acetonitrile and water
containing 0.02% formic acid. Compounds were fully characterized with 1HNMR. 1HNMR
spectra were recorded on a Varian 400 MHz machine. Samples were run in DMSO-d6 after
removal of the exchangeable protons by D2O exchange with tetramethylsilane as an internal
standard. The chemical shifts are expressed in δ (ppm) values, and coupling constants are
expressed in hertz (Hz). s = singlet, d = doublet, t = triplet, m = multiplet, and brs = broad sin-
glet. HPLC/UV and 1HNMR spectra for all of the compounds are shown in the Supporting
Information.
The synthesis of (R)- and (S)-xamoterol is shown in Fig 2. Reaction of 4-(benzyloxy)phe-
nol with (R)-epichlorohydrin, (R)-5, gave the epoxide (S)-6. Ring opening of this epoxide
with the morpholino amine, 4, and reductive removal of the benzyl group with Pd on char-
coal then gave (S)-xamoterol. In a similar fashion (R)-xamoterol was prepared using (S)-
epichlorohydrin.
Experimental procedure for tert-Butyl 2-(morpholine-4-carboxamido)ethylcarbamate
(3). To a solution of tert-butyl 2-aminoethylcarbamate 1 (50 g, 312.09 mmol) in CH2Cl2
(300mL) was added triethylamine (94.74 g, 936.3 mmol), and morpholine-4-carbonyl chloride
2 (46.7 g, 312.09 mmol) in CH2Cl2 (200mL) dropwise at 0˚C. The reaction mixture was war-
med to room temperature and stirred at ambient temperature overnight. The reaction mixture
was then concentrated under reduced pressure and, after extractive work up, the residue was
purified by silica gel column (CH2Cl2: MeOH = 10:1) to give tert-butyl 2-(morpholine-4-car-
boxamido)ethylcarbamate 3 (75g, 87.9%) as a white solid.
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
3 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 2. Synthesis of (R)- and (S)-xamoterol.
https://doi.org/10.1371/journal.pone.0180319.g002
Experimental procedure for N-(2-aminoethyl) morpholine-4-carboxamide (4). A solu-
tion of tert-butyl 2-(morpholine-4-carboxamido)ethylcarbamate 3 (75 g, 274.39 mmol) in
CH2Cl2 (500mL) was treated dropwise at 0˚C with TFA (156.0 g, 1.37 mol). The reaction mix-
ture was stirred for 1.5 hrs and then concentrated under reduced pressure. The resulting resi-
due was dissolved in CH2Cl2 and 6M NaOH was added to adjust to pH 8. This mixture was
extracted with CH2Cl2 and the organic solution was washed with water, dried over Na2SO4
and evaporated under reduced pressure to give N-(2-aminoethyl) morpholine-4-carboxamide
4 (43 g, 91%) as a yellow oil. Some of the free base was converted to 4-(N-beta-aminoethylcar-
bamoyl)morpholine hydrogen sulfate (mp. 168–169˚C).
Experimental procedure for (S)-2-((4-(benzyloxy)phenoxy)methyl)oxirane ((S)-6).
CsF (113.79 g, 749.14 mmol) was added with stirring to a mixture of 4-(benzyloxy)phenol,
(50 g, 249.71 mmol) and (R)-2-(chloromethyl)oxirane, (R)-5 (69.31 g, 749.14 mmol) in DMF
(500mL). The reaction mixture was stirred at 50˚C for 3 days. The reaction mixture was then
partitioned between water (1 L) and EtOAc (2 L). The organic layer was washed with water
(3x900 mL), brine, and then dried over Na2SO4. The filtered organic extract was concentrated
under reduced pressure and purified by silica column (pet ether:EtOAc, 4:1) to give (S)-2-((4-
(benzyloxy)phenoxy)methyl)oxirane, (S)-6 (47 g, 73.4%) as a white solid.
Experimental procedure for (S)-N-(2-(3-(4-(benzyloxy)phenoxy)-2-hydroxypropyla-
mino)ethyl)morpholine-4-carboxamide ((S)-7). A mixture of N-(2-aminoethyl)morpho-
line-4-carboxamide, 4 (30 g, 104.11 mmol) in isopropanol (120mL) was slowly added to a
solution of (S)-2-((4-(benzyloxy)phenoxy)methyl)oxirane, (S)-6 (19.72 g, 156.66 mmol) in iso-
propanol (230mL) below 50˚C over 5 hrs. The reaction mixture was then cooled to room tem-
perature and thoroughly extracted with ethyl acetate. The organic layer was washed with
water, brine, and then dried over Na2SO4. The ethyl acetate extract was concentrated under
reduced pressure and purified by silica gel column (CH2Cl2: MeOH, 10:1) to give (S)-N-(2-(3-
(4-(benzyloxy)phenoxy)-2-hydroxypropylamino)ethyl)morpholine-4-carboxamide, (S)-7
(20.8 g, 63%) as a white solid.
Experimental procedure for (S)-xamoterol. A mixture of (S)-N-(2-(3-(4-(benzyloxy)
phenoxy)-2-hydroxypropylamino)ethyl)morpholine-4-carboxamide, (S)-7 (20 g, 46.56 mmol)
in 200mL of EtOH containing 2mL of acetic acid was added Pd(OH)2/C (5g). The resulting
reaction mixture was subjected to hydrogenation at 55 psi overnight. The catalyst was removed
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
4 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
by filtration and the filtrate was concentrated under reduced pressure. The residue was dis-
solved in EtOH, and converted to hemifumarate with fumaric acid. The hemifumarate salt was
recrystallized with EtOH to give (S)-xamoterol hemifumarate (12.3 g, 91.7%) as an off-white
solid.
Experimental procedure for (R)-2-((4-(benzyloxy)phenoxy)methyl)oxirane ((R)-6).
This compound was prepared following the procedure as described for (S)-6 from 4-(benzy-
loxy)phenol (8 g, 41.9 5mmol) and (S)-2-(chloromethyl)oxirane, (R)-5, to give (R)-2-((4-(ben-
zyloxy)phenoxy)methyl)oxirane, (R)-6 (6.3 g, 24.6 mmol, 61%) as a white solid.
Experimental procedure for (R)-N-(2-(3-(4-(benzyloxy)phenoxy)-2-hydroxypropyla-
mino)ethyl)morpholine-4-carboxamide ((R)-7). This compound was prepared following
the procedure as described for the preparation of (S)-7 from 6.3 g, (24.6 mmol) of (R)-6 and
12.77 g (73.80 mmol) of 4 to afford 3 g, (6.98 mmol, 30%) of (R)-7.
Experimental procedure for (R)-xamoterol. (R)-7 (3g, 6.98mmol) was converted to (R)-
xamoterol hemifumarate following the procedure as described above to give (R)-xamoterol
hemifumurate (2.5g, 6.8mmol) as an off-white solid.
The general method for the synthesis of compounds STD-101-B1 to B8 is shown in Fig 3.
Reaction of phenols 8-B1 to B8 with (R)-2-(chloromethyl)oxirane and CsF in iPrOH for 48 hrs
at 50˚C gave the coupled epoxides 9-B1 to B8 in good yield. Epoxide opening with an excess of
N-(2-aminoethyl)morpholine-4-carboxamide in iPrOH at 50˚C then provided the benzyl
ethers 10-B1 to B8. Reductive removal of the benzyl protecting group with Pd/C then gave the
free phenols, STD-101-B1 to B8 and E which were isolated either as the hemifumarate salts,
STD-101-B1, B2, B3, B5 and B7 or as the free base, STD-101-B4, B6, B8 and E. Experimental
procedure for the representative example STD-101-B1 is shown below.
Experimental procedure for (S)-2-((4-(benzyloxy)-3-methylphenoxy)methyl)oxirane
(9-B1). To a solution of 4-(benzyloxy)-3-methylphenol (1.95 g, 9.10 mmol) in DMF (20 mL)
was added (R)-2-(chloromethyl)oxirane (2.53 g, 27.30 mmol),CsF (4.15 g, 27.30 mmol). The
reaction mixture was stirred for 48 hrs at 50˚C. The reaction mixture was cooled to room tem-
perature and poured to ice water and extracted with EtOAc (3X20 mL).The combined organic
layers were washed with brine, dried over Na2SO4, filtered, concentrated under reduced pres-
sure and purified by flash silica column chromatography to give (S)-2-((4-(benzyloxy)-
3-methylphenoxy)methyl)oxirane (2.1 g, 84.5%) as a yellow oil.
Experimental procedure for (S)-N-(2-(3-(4-(benzyloxy)-3-methylphenoxy)-2-hydroxy-
propylamino)ethyl)morpholine-4-carboxamide (10-B1). To a solution of (S)-2-((4-(benzy-
loxy)-3-methylphenoxy)methyl)oxirane (0.6 g, 2.22 mmol) in iPrOH (20 mL) was added N-
(2-aminoethyl)morpholine-4-carboxamide (0.77 g, 4.44 mmol). The reaction mixture was stir-
red for 18 hrs at 50˚C. The reaction mixture was cooled to room temperature and poured to
ice water and extracted with EtOAc (3X20 mL). The combined organic layers were washed
Fig 3. General method for the synthesis of compounds STD-101-B1 to B8.
https://doi.org/10.1371/journal.pone.0180319.g003
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
5 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
with brine, dried over Na2SO4, filtered, concentrated under reduced pressure and purified by
flash silica column chromatography to give (S)-N-(2-(3-(4-(benzyloxy)-3-methylphenoxy)-
2-hydroxypropylamino)ethyl)morpholine-4-carboxamide (0.5 g, 50.8%) as a yellow oil.
Experimental procedure for STD-101-B1, hemifumarate. To a solution of (S)-N-(2-(3-
(4-(benzyloxy)-3-methylphenoxy)-2-hydroxypropylamino)ethyl)morpholine-4-carboxamide
(0.5 g, 2.22 mmol) in MeOH(15 mL) was added Pd/C. The reaction mixture was subjected to
H2 for 4 hrs. The reaction mixture was filtered, and the filtrate was concentrated under
reduced pressure. The residue was dissolved in EtOH(15 mL), fumaric acid(0.5eq) was added
and solvent removed under reduced pressure to give STD-101-B1, hemifumarate (0.35 g,
75.5%) as a pink solid.
The synthesis of STD-101-B9 is shown in Fig 4. Reaction of 2-fluoro-5-hydroxybenzalde-
hyde with (R)-2-(chloromethyl)oxirane and CsF in iPrOH gave the coupled epoxide 12, which
was then treated with sodium borohydride to reduce the aldehyde to the hydroxymethylene,
13. Ring opening with N-(2-aminoethyl) morpholine-4-carboxamide gave STD-101-B9,
which was isolated as the hemifumarate salt.
Experimental procedure for (S)-2-fluoro-5-(oxiran-2-ylmethoxy)benzaldehyde (12). A
mixture of 2-fluoro-5-hydroxybenzaldehyde (1.0 g, 7.14 mmol), (R)-2-(chloromethyl) oxirane
(1.98 g, 21.41mmol) and CsF(3.25 g,21.41mmol) in DMF (10 ml) was stirred at 0˚C and then
heated at 50˚C for 48 hrs. The reaction mixture was then portioned between water and ethyla-
cetate (2:1 ratio), the organic layer was washed with water, brine, dried over Na2SO4, filtered,
concentrated under reduced pressure and purified by silica column chromatography, to give
(S)-2-fluoro-5-(oxiran-2-ylmethoxy)benzaldehyde (1.1 g, 78.6%) as a light yellow oil.
Experimental procedure for (S)-(2-fluoro-5-(oxiran-2-ylmethoxy)phenyl)methanol
(13). To a solution of (S)-2-fluoro-5-(oxiran-2-ylmethoxy)benzaldehyde (1.2 g, 6.14 mmol)
in THF (10 ml) was added NaBH4(0.23 g, 6.12 mmol) and the reaction mixture was stirred at
room temperature overnight. Then the reaction mixture was quenched with water and
extracted with EtOAc (3X30mL). The combined organic layers were washed with water, brine,
dried over Na2SO4, filtered, concentrated under reduced pressure and purified by silica col-
umn chromatography, to give (S)-(2-fluoro-5-(oxiran-2-ylmethoxy)phenyl)methanol (0.57 g,
47%) as yellow oil.
Experimental procedure for (S)-N-(2-(2-hydroxy-3-(3-methyl-1H-indazol-4-yloxy)pro-
pylamino)ethyl)morpholine-4 (STD-101-B9). To a solution of (S)-(2-fluoro-5-(oxiran-
2-ylmethoxy) phenyl) methanol (0.57 g, 2.88mmol) in iPrOH (20 ml), was added N-(2-ami-
noethyl) morpholine-4-carboxamide (1.0g, 5.75mmol). The reaction mixture was stirred at
50˚C for 5 hrs, and then evaporated under reduced pressure. Water was added to the residue
and the product was extracted into ethyl acetate. The organic layer was washed with water,
brine, dried over Na2SO4, filtered, concentrated under reduced pressure and purified by silica
column chromatography to give (S)-N-(2-(2-hydroxy-3-(3-methyl-1H-indazol-4-yloxy)propy-
lamino)ethyl)morpholine-4-carboxamide (0.23 g, 18.6%) as a solid.
Fig 4. Synthesis of STD-101-B9.
https://doi.org/10.1371/journal.pone.0180319.g004
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
6 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 5. General method for the synthesis of compounds STD-101-D1 to D6.
https://doi.org/10.1371/journal.pone.0180319.g005
The general method for the synthesis of compounds STD-101-D1 to D6 is shown in Fig 5.
The starting material (S)-1-(4-(benzyloxy)phenoxy)-3-(2-(2-methoxyphenoxy)ethylamino)
propan-2-ol (15) was prepared by treating 4-benzyloxyphenol with (R)-epichlorohydrin as
described above. Opening of the epoxide ring of 15 with 2-(2-methoxyphenoxy)ethanamine
gave the benzyl ether 16 which was then deprotected by catalytic reduction to give the desired
product STD-101-D1. Compounds STD-101-D2 to D6 were prepared in a similar fashion
using the appropriate amine reagent and isolated as the hemifumarate salts. Experimental pro-
cedure for the representative example compound STD-101-D1 is shown below.
Experimental procedure for (S)-1-(4-(benzyloxy)phenoxy)-3-(2-(2-methoxyphenoxy)
ethylamino)propan-2-ol (16). To a solution of (S)-2-((4-(benzyloxy)phenoxy)methyl)oxi-
rane (1.0 g, 3.9mmol) in iPrOH (40 ml), was added 2-(2-methoxyphenoxy)ethanamine (0.98 g,
5.8mmol) in iPrOH dropwise under 50˚C for 5h. The reaction mixture was cooled to room
temperature, and removed the solution, water was added, extracted with EA. The organic layer
was washed with water, brine, dried over Na2SO4, concentrated under reduced pressure and
purified by silica column, to give (S)-1-(4-(benzyloxy)phenoxy)-3-(2-(2-methoxyphenoxy)
ethylamino)propan-2-ol (0.52 g, 1.228mmol).
Experimental procedure for (S)-4-(2-hydroxy-3-(2-(2-methoxyphenoxy)ethylamino)
propoxy)phenol (STD-101-D1). A solution of (S)-1-(4-(benzyloxy)phenoxy)-3-(2-(2-meth-
oxyphenoxy)ethylamino)propan-2-ol (0.52 g, 1.2mmol) and Pd/C (0.1 g) in MeOH (20ml),
was stirred under a H2 atmosphere at RT for 2h. The reaction mixture was filtered, concen-
trated under reduced pressure, and purified by silica column to give (S)-4-(2-hydroxy-3-(2-
(2-methoxyphenoxy)ethylamino)propoxy)phenol (0.35g, 1.05mmol).
The syntheses of the amine components for STD-101-D5 and D6 are shown in Fig 6.
NMR spectral data
NMR spectral results for STD-101-B1 to B9, D1 to D6 and E are shown below. It is important
to highlight that the H-signals for the fumaric acid in the hemifumarate salts are not recorded.
(S)-N-(2-((2-hydroxy-3-(4-hydroxy-3-methylphenoxy)propyl)amino)ethyl)morpholine-
4-carboxamide hemifumarate (STD-101-B1). MS(m/z), 354(M+1); 1HNMR (400 MHz,
DMSO, after D2O): 2.04 (3H, s), 2.91(2H, m), 3.06(1H, m), 3.21 (4H, m), 3.25 (1H,m), 3.50
(4H, m), 3.75 (1H,m), 4.0 (1H, m), 6.5 (1H,s), 6.66(2H, m).
(S)-N-(2-((3-(3-chloro-4-hydroxyphenoxy)-2-hydroxypropyl)amino)ethyl)morpholine-
4-carboxamide hemifumarate (STD-101-B2). MS(m/z), 374(M+1); 1HNMR (400 MHz,
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
7 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 6. Syntheses of the amine components for STD-101-D5 and D6.
https://doi.org/10.1371/journal.pone.0180319.g006
DMSO, after D2O exchange, whenever as salt signals for fumaric acid are not shown): 2.96
(4H, m), 3.25 (4H, m), 3.37 (2H, m), 3.50 (4H, m), 3.84 (2H, m), 4.08 (1H, m), 6.73 (1H, dd,
J = 8,4Hz); 6.76 (1H, d, J = 8Hz), 6.94 (1H, d, J = 4Hz.)
(S)-N-(2-((3-(3-ethyl-4-hydroxyphenoxy)-2-hydroxypropyl)amino)ethyl)morpholine-
4-carboxamide hemifumarate (STD-101-B3). MS(m/z), 368(M+1); 1HNMR (400 MHz,
DMSO, after D2O exchange); 1.05(3H,t J = 7.8 Hz), 2.48 (2H, q, J = 7.8Hz), 2.90(2H, m),2.95–
3.22 (10H, m), 3.48 (4H,m), 4.01(1H, m), 6.41(1H, s), 6.55–6.65(2H,m).
(S)-N-(2-((3-(5-chloro-2-fluoro-4-hydroxyphenoxy)-2-hydroxypropyl)amino)ethyl)
morpholine-4-carboxamide (STD-101-B4). MS(m/z), 392(M+1); 1HNMR (400 MHz,
DMSO, after D2O exchange): 3.19(2H, m),3.25–3.36(6H, m), 3.48 (2H,m), 3.64(4H, m), 3.99
(2H, m) 4.03(1H, m), 6.72(1H, d, J = 12.2Hz), 7.14(1H,d, J = 8.72Hz).
(S)-N-(2-((3-(2,5-difluoro-4-hydroxyphenoxy)-2-hydroxypropyl)amino)ethyl)morpho-
line-4-carboxamide hemifumarate (STD-101-B5). MS(m/z), 376(M+1); 1HNMR (400
MHz, DMSO, after D2O exchange): 2.85–3.10 (4H, m),3.23 (4H, m), 3.48 (2H,m), 3.54–3.70
(6H, m), 3.91(2H, m) 4.13(1H, m), 6.82(1H, d, J = 8Hz), 7.10(1H,d, J = 12Hz).
(S)-N-(2-((3-(3-cyano-4-hydroxyphenoxy)-2-hydroxypropyl)amino)ethyl)morpholine-
4-carboxamide (STD-101-B6). MS(m/z), 365(M+1); 1HNMR (400 MHz, DMSO, after D2O
exchange): 2.48–2.54 (4H, m), 3.07(2H, m), 3.22 (4H, m), 3.50 (4H,m), 3.81 (2H, m), 3.88 (1H,
m), 6.89(1H, d, J = 8.8 Hz), 7.08(1H, d J = 8.8Hz), 7.13(1H,s).
(S)-N-(2-((3-(3,5-difluoro-4-hydroxyphenoxy)-2-hydroxypropyl)amino)ethyl)morpho-
line-4-carboxamide hemifumarate (STD-101-B7). MS(m/z), 376(M+1); 1HNMR (400
MHz, DMSO, after D2O exchange: 2.72–2.86(4H, m), 3.13(4H, m), 3.49 (4H, m), 3.50 (4H,m),
3.60 (2H, m), 3.75 (2H, m), 3.85 (1H,m), 6.68(2H, d, J = 7.2 Hz).
(S)-N-(2-((3-(5-cyano-2-fluoro-4-hydroxyphenoxy)-2-hydroxypropyl)amino)ethyl)
morpholine-4-carboxamide (STD-101-B8). MS(m/z), 383(M+1); 1HNMR (400 MHz,
DMSO, after D2O exchange): 2.85–3.15(4H, m), 3.23–3.34 (4H, m), 3.46(6H, m), 3.98 (2H,m),
4.12(1H, m), 6.86(1H, d, J = 13 Hz), 7.45 (1H, d, J = 9Hz).
(S)-N-(2-((3-(4-fluoro-3-(hydroxymethyl)phenoxy)-2-hydroxypropyl)amino)ethyl)mor-
pholine-4-carboxamide hemifumarate (STD-101-B9). MS(m/z), 372(M+1); 1HNMR (400
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
8 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
MHz, DMSO, after D2O exchange: 2.98 (4H, m), 3.12(4H, m), 3.30 (2H, m), 3.51 (4H, m), 3.89
(2H, m), 4.12(1H,m), 4.47(2H,s), 6.80 (1H,m), 7.02 (2H,m).
(S)-4-(2-hydroxy-3-((2-(2-methoxyphenoxy)ethyl)amino)propoxy)phenol hemifuma-
rate (STD-101-D1). MS(m/z), 334(M+1); 1HNMR (400 MHz, DMSO, after D2O exchange):
2.81 (2H, m), 2.95 (2H, m), 3.07 (2H, t, J = 4Hz), 3.71 (3H, s), 3.96 (1H, m), 4.08 (2H, t,
J = 4Hz), 6.64 (2H, d, J = 8,4Hz); 6.73 (2H, d, J = 8Hz), 6.94 (4H, m)
(S)-4-(3-((2-(3,4-dimethoxyphenethoxy)ethyl)amino)-2-hydroxypropoxy)phenol hemi-
fumarate (STD-101-D2). S(m/z), 392(M+1); 1HNMR (400 MHz, DMSO, after D2O
exchange): 2.70 (2H, m), 2.96 (2H, m), 3.63 (3H, s), 3.66 (3H, s), 3.75 (2H, m), 3.97 (3H, m),
6.64 (2H, d, J = 8,4Hz); 6.73 (2H, d, J = 8Hz), 6.63–6.79 (7H, m).
(S)-4-((2-hydroxy-3-(4-hydroxyphenoxy)propyl)amino)-1-morpholinobutan-1-one
hemifumarate (STD-101-D3). MS(m/z), 339(M+1); 1HNMR (400 MHz, DMSO, after D2O
exchange): 1.73 (2H, m), 2.34 (2H, t, J = 7Hz), 2.75 (2H, m), 2.88 (1H, m), 3.39 (4H, m), 3.55
(4H, m), 3.80 (2H, m), 3.97 (1H, m), 6.65 (2H, d, J = 8Hz), 6.75 (2H, d, J = 8Hz).
(S)-3-((2-hydroxy-3-(4-hydroxyphenoxy)propyl)amino)-1-morpholinopropan-1-one
hemifumarate (STD-101-D4). MS(m/z), 325(M+1); 1HNMR (400 MHz, DMSO, after D2O
exchange): 2.71 (2H, m), 2.88 (1H, m), 3.08 (3H, m), 3.40 (4H, m), 3.60 (4H, m), 3.78 (2H, m),
4.08 (1H, m), 6.66 (2H, d, J = 8Hz), 6.75 (2H, d, J = 8Hz).
(S)-3-hydroxy-N-(2-((2-hydroxy-3-(4-hydroxyphenoxy)propyl)amino)ethyl)azetidine-
1-carboxamide hemifumarate (STD-101-D5). MS(m/z), 326(M+1); 1HNMR (400 MHz,
DMSO, after D2O exchange): 2.98 (4H, m), 3.20 (2H, m), 3.54 (2H, m), 3.81 (2H, m), 3.93
(2H, m), 4.10 (1H, m), 4.37 (1H, m), 6.66 (2H, d, J = 8Hz), 6.75 (2H, d, J = 8Hz).
(S)-N-(2-((2-hydroxy-3-(4-hydroxyphenoxy)propyl)amino)ethyl)-3-(hydroxymethyl)
azetidine-1-carboxamide hemifumarate (STD-101-D6). MS(m/z), 340(M+1); 1HNMR
(400 MHz, DMSO, after D2O exchange): 2.57 (1H,m), 3.00 (3H, m), 3.18 (1H, m), 3.31 (2H,
m), 3.50 (4H, m), 3.80 (4H, m), 4.12 (1H, m), 6.66 (2H, d, J = 8Hz), 6.75 (2H, d, J = 8Hz).
(S)-N-(2-((2-hydroxy-3-((5-hydroxynaphthalen-1-yl)oxy)propyl)amino)ethyl)morpho-
line-4-carboxamide (STD-101-E). MS(m/z), 390(M+1); 1HNMR (400 MHz, DMSO, after
D2O exchange): 2.77–2.84 (4H, m), 3.21 (6H, m), 3.47 (4H, m), 4.08 (3H, m), 6.87(2H, m),
7.23 (2H, m), 7.65 (2H, d, J = 8Hz).
cAMP assay
Pharmacological effects of compounds at the cAMP pathway mediated by ADRB1 were evalu-
ated by measuring cAMP production using the homogenous time-resolved fluorescence detec-
tion method with HEK-293 cells stably expressing human recombinant ADRB1. Briefly, cells
were suspended in HBSS buffer completed with 20 mM HEPES (pH 7.4) and 500 μM IBMX
(3-isobutyl-1-methylxanthine) and distributed at a density of 3x103 cells/well. Subsequently,
cells were incubated with HBSS (basal control), the full agonist isoproterenol hydrochloride
(I5627; Sigma-Aldrich Corp., MO), xamoterol (Santai Labs; http://www.santailabs.com/index.
asp), or test compounds for 30 min. Following incubation, the cells were lysed and the fluores-
cence acceptor (D2-labeled cAMP) and fluorescence donor (anti-cAMP antibody labeled with
europium cryptate) were added. After a 60 min incubation with the fluorescence donor and
acceptor at room temperature, the fluorescence transfer was measured at 337 nm (excitation)
and 620 and 665 nm (emission) using a microplate reader. The cAMP concentration was deter-
mined by dividing the signal measured at 665 nm by that measured at 620 nm (ratio). The
results were expressed as a percent of the maximum efficacy achieved with isoproterenol. Phar-
macological effects of compounds at the cAMP pathway mediated by ADRB2 and ADRB3 were
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
9 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
measured as described above using CHO cells stably expressing human recombinant ADRB2
and human SK-N-MC neurotumor cells endogenously expressing ADRB3, respectively.
β-arrestin assay
Pharmacological activity of compounds in the β-arrestin pathway mediated by ADRB1 were
evaluated using an enzyme fragment complementation method with a β-galactosidase func-
tional reporter. An engineered CHO-K1-ADRB1 PathHunter cell line (DiscoveRx) was used
in the assay. In this cell line, the enzyme acceptor (β-galactosidase fragment) is fused to β-
arrestin and the enzyme donor (β-galactosidase fragment) is fused to the ADRB1. Thus, activa-
tion of the ADRB1 stimulates binding of β-arrestin to the ProLink-tagged ADRB1 and forces
complementation of the two enzyme fragments, resulting in the formation of an active β-galac-
tosidase enzyme. Briefly, CHO-K1-ADRB1 PathHunter cell lines were plated in a total volume
of 20 μL cell plating reagent (DiscoveRx, 93-0563R0A) at a density of 2,500 cells/well into 384
well microplates and incubated overnight at 37˚C in 5% CO2. The following day, 5 μL of the
full agonist isoproterenol, xamoterol (S), or test compounds was added to cells and incubated
at 37˚C for 90 min. After the 90 min incubation, 15 μL of PathHunter Detection reagent cock-
tail (DiscoveRx, 93–0001) was added, followed by a 60 min incubation at room temperature.
Chemiluminescent signal was then read with a PerkinElmer Envision (Perkin Elmer, Inc.,
MA) instrument. The results were expressed as a percent of the maximum efficacy achieved
with isoproterenol.
Crystal structure modeling
Ligands were docked to the binding pocket of PDB: 2YCZ, constrained to a 25 Angstrom
box size surrounding the orthosteric pocket. Docking was performed using Schro¨dinger Glide,
with SP precision and after using LigPrep to prepare all ligands with a maximum of 32 stereo-
isomers and 6 low energy ring conformations. The docking and analysis were facilitated by the
open source software: (https://github.com/evanfeinberg/conformation/blob/master/grids.py)).
In vitro primary microglia tumor necrosis factor α (TNFα) assay
Mixed glial cells were obtained from the cerebral cortex of Sprague Dawley rat pups at postna-
tal days 1–3. Briefly, neonates were euthanized by decapitation and their brain tissues were col-
lected for cortex isolation. The isolated cortex was then trypsinized, triturated, and placed into
tissue culture flasks in DMEM supplemented with 10% fetal bovine serum and 1% Penicillin/
Streptomycin. After 10 days in vitro, microglia were harvested by gentle shaking of the growth
flask, plated in a 96 well plate at a density of 30,000 cells/well, and incubated at 37˚C overnight.
The next day, microglia were stimulated with lipopolysaccharides (LPS) (10 ng/ml) along with
isoproterenol, xamoterol (S), or STD-101-D1 at the concentration of 10 μM for 4 hrs at 37˚C.
Following the 4 hr incubation, cell media was collected and the concentration of TNFα was
measured by ELISA (Invitrogen, KRC3011) according to the manufacturer’s instruction.
Microsomal stability assay
The test compound was pre-incubated with pooled mouse, rat, or human microsomes in 100
mM potassium phosphate buffer (pH 7.4) containing 10 mM MgCl2 for 5 min in a 37˚C shak-
ing waterbath. After the preincubation, the reaction was initiated by adding freshly prepared
NADPH to a final concentration of 1mM. Aliquots of the reaction samples were collected at 0
min, 15 min, 30 min, 45 min, and 60 min after the initiation of the reaction, and quenched
with equal volume of acetonitrile. Samples were then mixed and centrifuged, and supernatants
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
10 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
were diluted with equal volume of water and used for LC-MS/MS analysis to determine the
concentrations of STD-101-D1. Analyte peak areas at different time points were recorded, and
the compound remaining was calculated by comparing the peak area at each time point to
time zero. The half-life was calculated from the slope of the initial linear range of the logarith-
mic curve of compound remaining (%) vs. time, assuming first order kinetics.
Animals
For the in vivo LPS study, a total of 45 male C57Bl/6J mice at the age of 10–12 weeks (Jackson
Laboratory, Bar Harbor, ME, USA) were used. For the pharmacokinetic study, a total of 36
Sprague-Dawley male rats (Charles River, Wilmington, MA, USA) weighing 280–380 g were
used. All animals were kept under a reverse light-dark cycle with lights off at 8:30 AM and on
at 8:30 PM in a temperature- and humidity-controlled environment and given food and water
ad libitum. All animal experiments were conducted in accordance with the U.S. National Insti-
tutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 80–
23). All experimental protocols were reviewed and approved by the Stanford University Insti-
tutional Animal Care and Use Committee (IACUC) (Protocol #—18466).
In vivo LPS assay
C57Bl/6J mice (10–12 weeks old) were injected with xamoterol (3 mg/kg; subcutaneous; n = 11),
STD-101-D1 (3 mg/kg; intraperitoneal; n = 4) or vehicle (5% DMSO/20%PEG400/water; intra-
peritoneal; n = 12) 15 min prior to LPS (50 ug/kg; intraperitoneal). A control group was injected
with vehicle 15 minutes prior to saline (n = 18). Following injections, mice were single-housed.
At 90 min post-LPS/saline, mice were deeply anesthetized with isoflurane and blood was col-
lected from the right ventricle via cardiac puncture (23 g needle) into lithium heparin-containing
vials (BD microtainer plasma tubes). Subsequently, brains were collected after perfusion with
phosphate-buffered saline. Plasma was separated by centrifugation (11,000 rpm for 3 minutes)
within 60 min of collection and stored at -80˚C until analysis. The concentration of TNFα in the
collected plasma was measured by ELISA (Invitrogen, KMC3012) according to the manufactur-
er’s instruction. The LPS-induced inflammatory response in the CNS was assessed by measuring
mRNA expression for genes related to neuroimmune activation in brain tissue according to the
previously reported method [11]. Briefly, total RNA was extracted from hippocampus containing
coronal sections using the RNeasy Lipid Tissue Mini Kit (Qiagen), and transcribed into cDNA
(Superscript III, Invitrogen). PCR was performed in duplicate using TaqMan gene expression
assays, TNFα (Mm00443258_m1), IL1b (Mm00434228_m1), IL6 (Mm00446190_m1), and glyc-
eraldehyde-3-phosphate dehydrogenase (GAPDH; Mm99999915_g1). Fold changes of expres-
sion relative to control were determined after normalization to GAPDH. Relative quantification
and fold change were calculated by the comparative CT method [19]. Brain uptake of STD-
101-D1 was determined in brain tissue. Brain tissues were homogenized in distilled water at a
ratio of 1:3 (weight of tissue:volume of water) and homogenates were analyzed using LC-MS/
MS.
Pharmacokinetic studies
Two cohorts of Sprague-Dawley rats were used in two independent studies. All animals were
fasted overnight before the experiment with free access to water. In study 1, a total of 18 rats
were used for a 4 hr time course pharmacokinetic (PK) study. Rats were anesthetized with 3%
isoflurane and catheters were implanted into the jugular and/or portal veins for compound
administration and/or blood sampling at 1 to 2 days prior to the experiments as previously
described [20, 21]. On the day of the experiment, xamoterol was freshly prepared in saline.
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
11 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
STD-101-D1 was freshly prepared in 5% DMSO, 20% PEG, and 75% distilled water. The pre-
pared xamoterol or STD-101-D1 was administered to the cannulated rats at a dose of 10 mg/
kg intravenously (IV), intraperitoneally (IP), or orally (PO) (n = 3 per route). For IV and IP
groups, approximately 150 μL aliquot of blood samples were collected via jugular vein cathe-
ters before drug administration, and at 1, 5, 10, 45, 60, 90, 120, and 180 min after drug admin-
istration. For the PO group, approximately 150 μL aliquot of blood samples were collected via
jugular and portal vein catheters before drug administration, and at 1, 5, 10, 45, 60, 90, 120,
and 180 min after drug administration. Four hours post-dose, rats were deeply anesthetized
with isoflurane and blood samples were collected by cardiac puncture. In study 2, a total of 18
rats were used for a 20 min post-dose collection study. Xamoterol and STD-101-D1 were
freshly prepared as described above and administered to rats at a dose of 10 mg/kg via IV, IP,
or PO routes (n = 3 per route). At 20 min post-dose, rats were deeply anesthetized with isoflur-
ane and blood samples were collected by cardiac puncture. Brains were collected after perfu-
sion with phosphate buffered saline. All plasma samples were immediately separated after
collection by centrifugation (11,000 rpm for 3 minutes) and stored at −80˚C until analysis. The
brain tissue samples were homogenized in distilled water at a ratio of 1:3 (weight of tissue:vol-
ume of water), and the homogenates were stored at −80˚C until analysis. The concentrations
of xamoterol and STD-101-D1 in plasma and brains were determined using LC-MS/MS (AB
SCIEX QTRAP 4000 mass spectrometer coupled to a Shimadzu UFLC system). For xamoterol,
LC separation was carried out on an Agilent Zorbax SB-Phenyl column (5 μm, 2.1×50 mm)
with isocratic elution using a mobile phase composed of 30% methanol and 70% water with
0.1% of formic acid. Tulobuterol was used as the internal standard. The flow rate was set to
0.45 ml/min. Column temperature was 25˚C. The analysis time was 2.2 min. The injection vol-
ume was 20 μl. The mass spectrometer was operated in the positive mode with multiple-reac-
tion monitoring (MRM). The m/z 340.2!253.2 and 228.1!154.1 transitions were used for
Xamoterol and the internal standard, respectively. For STD-101-D1, LC separation was carried
out on a Phenomenex Synergi Polar-RP column (2.5 μm, 2 mm × 50 mm) with a flow rate of
0.25 ml/min at room temperature. Mobile phase A consisted of 10 mM ammonium acetate
and 0.1% formic acid in LCMS grade water. Mobile phase B consisted of 10 mM ammonium
acetate and 0.1% formic acid in LCMS grade acetonitrile:water 90:10% (v/v). The HPLC elu-
tion program was as follows: 35% B (0.3 min)!85% B (linear increase in 1.2 min)!35% B
(linear decrease in 0.1 min)!35% B (0.9 min). Five μl of the extracted samples were injected.
The mass spectrometer was operated in the positive mode with multiple-reaction monitoring
(MRM) with the transition m/z 334.1!210.1 and 334.1!100.2. Data acquisition and analysis
were performed using the Analyst 1.6.1 software (AB SCIEX, CA).
Cardiovascular studies
Effects of compounds on heart rate and blood pressure were measured in male Sprague-Daw-
ley rats using a fluid filled catheter-transducer system with a disposable blood pressure trans-
ducer MLT 0699 connected to a PowerLab 8/30 recording unit with Quad Bridge Amp (AD
instruments, CO). Briefly, rats were anesthetized with isoflurane (3–4% for induction and
1.5% for maintenance) and a 1 cm longitudinal incision was made on the ventral aspect of the
tail exposing the tail artery. A polyethylene catheter (PE50) filled with heparinized saline was
then inserted into the tail artery and connected to the blood pressure transducer system. After
ensuring the absence of any air bubbles, 5–10 min of baseline systolic and diastolic blood pres-
sure along with heart rate measurements were recorded using LabChart Pro (AD Instruments,
CO). After establishing the baseline, xamoterol or STD-101-D1 was subcutaneously adminis-
tered and changes in blood pressure and heart rate measurements were recorded for an
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
12 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
additional 30 min. The effects of xamoterol or STD-101-D1 on heart rate were calculated as
the difference between average values recorded during the baseline recording and average val-
ues recorded during 5 to 10 min after the administration of compound. The effects of xamo-
terol or STD-101-D1 on blood pressure were calculated as the difference between average
values measured during the baseline recording and the lowest value measured during the 10
min period after administration of the compound.
Calculation and statistics
In vitro pharmacology data for the cAMP pathway represent 2–5 experiments performed sin-
gly or in duplicate. In vitro pharmacology data for the β-arrestin pathway represent technical
replicates within a single experiment. Curve fitting was performed with GraphPad Prism 5.0
software (GraphPad Software, CA) using the equation for a single-site sigmoidal, dose-
response curve with a variable slope. EC50 values were expressed as geometric means (95%
confidence limits). Statistical analyses were performed with GraphPad Prism 5.0. One-way
analyses of variance, followed by Dunnett’s test for post-hoc analyses were performed on in
vitro and in vivo TNFα data. Pharmacokinetic analyses were performed using the Phoenix
WinNonlin Professional Edition computer software version 2.0 (Certara, NJ). Differences
were considered to be significant at a level of p < 0.05. In all cases, outliers were excluded
according to Grubbs’ test and p < 0.05 was considered to be significant.
Results
Structure-activity relationship of compounds at the cAMP pathway
mediated by ADRB1
The β-aminoalcohol linking moiety is a crucial pharmacophoric element of most ADRB1
binding compounds, including xamoterol [22]. Consequently, structure-activity relationship
(SAR) studies and drug discovery programs have focused upon making structural changes at
the two terminal sites of ADRB1 binding molecules. We hypothesize that affinity and selectiv-
ity at ADRB1 can be significantly improved and manipulated by making the appropriate struc-
tural changes to xamoterol at the phenolic and morpholino subsites (Fig 1). As such, we
prepared 18 analogs of xamoterol with the objective of (1) maintaining affinity and efficacy for
ADRB1, and (2) increasing brain penetration. Pharmacological effects of the analogs on
ADRB1 were then evaluated by measuring cAMP production.
The major shortcomings of xamoterol for CNS indications are its poor oral bioavailability
and rapid clearance [18]. Most of the absorbed drug in humans is excreted in the urine
unchanged, with a substantial amount also excreted as the sulfate after first pass by the liver.
We therefore envisioned that structural modifications of the phenol group could enhance PK
properties of the molecules. In the first series of compounds, the phenol group in xamoterol is
substituted with a variety of electron donating and electron withdrawing groups (Fig 7). These
diverse structural modifications also alter the lipophilicity of the compounds and modulate
other molecular properties, such as tPSA, which should enhance oral absorption. As the (S)-
enantiomer of xamoterol was more potent and efficacious than the (R)-enantiomer, all struc-
tural modifications were made on the (S)-enantiomer (Fig 7 and Fig 8A). As demonstrated by
the data shown in Figs 7 and 8B, the introduction of methyl, chlorine, and ethyl residues
(STD-101-B1, B2, and B3) to sterically crowd the phenolic OH abolished the ADRB1 agonist
activity. These analogs produced no pharmacological effects up to concentrations of 100 μM
(Fig 7 and Fig 8B). On the other hand, introduction of cyanide to this position (STD-101-B6)
led to a decrease in both potency and efficacy compared to xamoterol (S) (Fig 7 and Fig 8B).
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
13 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 7. Effects of structural modifications of the phenolic OH moiety of xamoterol on the cAMP pathway mediated by
ADRB1. clogPa, Calculated with ChemDraw Pro Version 16.0 (PerkinElmer Health Sciences, CT; EC50 (nM)b, Geometrical
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
14 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
mean of EC50 values from at least two independent experiments; % Iso maxc, Percent efficacy compared to the maximum
response achieved with isoproterenol; ~d, Could not be determined.
https://doi.org/10.1371/journal.pone.0180319.g007
The addition of fluorines at C2 and C5 with respect to phenolic OH (STD-101-B5) resulted in
an approximately 2-fold efficacy decrease. A fluorine substitution at C3 and C5 (STD-101-B7)
abolished the agonist activity completely (Fig 7 and Fig 8B) as did substitution of the larger
substituents, chlorine or cyano, ortho to the phenolic OH group (STD-101-B4, STD-101-B8).
Compounds (STD-101-B9) and STD-101-E (Fig 7 and Fig 8B) were also inactive.
In the second series of compounds, analogs incorporated structural modifications of the
morpholino urea moiety of xamoterol (Fig 9). Replacement of morpholine-4-carboxamide
moiety in xamoterol (S) by (2-methoxyphenoxy)ethanamine (STD-101-D1) led to an approxi-
mately 7-fold decrease in potency and approximately 30% increase in efficacy (Fig 8C and Fig
9). Substitution with [2-(3,4-dimethoxyphenyl)ethoxy]ethanamie (STD-101-D2) resulted in
an approximately 80-fold decrease in potency and an approximately 10% increase in efficacy.
Similarly, analogs where the NH of the urea was replaced with methylene (STD-101-D3) or
completely excised (STD-101-D4) led to approximately 26- and 73-fold decreases in potency,
respectively. However, these modifications in STD-101-D3 and STD-101-D4 led to approxi-
mately 20% increases in efficacy (Fig 8C and Fig 9). Substitution of the morpholine group in
xamoterol (S) with smaller ring systems, 3-hydroxyazetidine (STD-101-D5) and 3-hydroxy-
methylazetidine (STD-101-D6) led to more than approximately 40-fold decreases in potency.
Interestingly, these changes in STD-101-D5 and STD-101-D6 only marginally affected their
efficacies. Substitution of the morpholine group in xamoterol (S) with 1,2-dimethoxybenzene
(RO363) led to an approximately 3-fold decrease in potency and an approximately 2-fold
increase in efficacy (Fig 8C and Fig 9). Importantly, none of the second series of compounds
having partial agonist activity on the cAMP pathway via ADRB1 show activity on cAMP path-
ways via ADRB2 or ADRB3, suggesting that they are partial agonists selective for ADRB1 ver-
sus ADRB2 and ADRB3 (data not shown).
Signaling bias of the key compound
To identify G protein-biased agonists of ADRB1, we assessed the pharmacological activity of
the compounds that showed significant agonistic activity at the cAMP pathway (e.g. STD-
101-D1 through STD-101-D6 and RO363) on the ADRB1-mediated β-arrestin pathway. As a
Fig 8. Concentration-response effects of compounds on the cAMP pathway via ADRB1. Data are expressed as a percentage of maximum efficacy
obtained with the full agonist isoproterenol. Values represents means ± S.E.M.s (1–2 experiments with n = 1–2).
https://doi.org/10.1371/journal.pone.0180319.g008
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
15 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 9. Effects of structural modifications of the morpholino urea moiety of xamoterol on the cAMP pathway mediated by ADRB1. clogPa,
Calculated with ChemDraw Pro Version 16.0; EC50 (nM)b, Geometrical mean of EC50 values from at least two independent experiments; % Iso
maxc, Percent efficacy compared to the maximum response achieved with isoproterenol; ~d, Could not be determined.
https://doi.org/10.1371/journal.pone.0180319.g009
classical unbiased agonist, isoproterenol produced concentration-dependent responses at the
β-arrestin pathway via ADRB1 with an EC50 value of 31.3 nM (Fig 10). In contrast, xamoterol
(S) did not produce concentration-dependent responses up to 30 μM, indicating that it has a
very high level of functional selectivity toward the cAMP pathway (Fig 10). The partial agonists
STD-101-D1 through STD-101-D4 displayed very weak partial agonist activity at the β-
arrestin pathway, producing less than 20% efficacy compared to the full agonist isoproterenol
(Figs 9 and 10). The other two compounds STD-101-D5 and STD-101-D6 did not produce
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
16 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 10. Concentration-response effects of compounds on the β-arrestin pathway via ADRB1. Data are
expressed as a percentage of maximum efficacy obtained with the full agonist isoproterenol. Values
represents means ± S.E.M.s (1–2 experiments with n = 1–2). Xamoterola; data has been previously published
[11].
https://doi.org/10.1371/journal.pone.0180319.g010
concentration-dependent responses up to 30 μM (Fig 9). However, RO363, which showed full
agonistic activity on the ADRB1-mediated cAMP pathway, produced partial agonistic activity
on the ADRB1-mediated β-arrestin pathway, achieving 36.5% efficacy with an EC50 value of
32.2 nM. On the basis of potency and partial agonistic activity on the cAMP pathway and func-
tional selectivity for the cAMP pathway over the β-arrestin pathway, the compound STD-
101-D1 was selected for further in vitro and in vivo testing.
Crystal structure modeling
In order to investigate the molecular mechanisms of xamoterol (S), xamoterol (R), and STD-
101-D1, we used molecular docking to predict the binding poses of the three ligands (Fig 11). All
three compounds are predicted to have a common set of characteristic interactions, including: 1)
a salt bridge between the protonated secondary amine of the compound and the carboxylic acid
of D121 of the receptor, 2) a hydrogen bond between the beta amino alcohol of the compound
and D121 of the receptor, and 3) a hydrogen bond between the phenolic hydroxyl group of the
compound and S211 of the receptor. Perceptible differences between the binding poses of the
three compounds are qualitatively subtle and, in lieu of a fully active crystal structure of ARB1,
are difficult to connect with differences in activity of the compounds. The carbon-nitrogen chain
of xamoterol (S) packs much more closely to D121 and to the transmembrane (TM) domain 3
than xamoterol (R), likely owing to the difference in stereochemistry between the two com-
pounds. In addition, whereas the phenolic moiety of all three drugs overlaps nearly exactly in the
predicted pose, the six-membered heterocycle of both xamoterol enantiomers and the phenyl
group of STD-101-D1 are predicted to adopt different positions in the extracellular portion of
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
17 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 11. Crystal structure of ADRB1 with xamoterol (S) (cyan, panel A), xamoterol (R) (blue, panel B), and STD-101-D1 (salmon, panel C) docked
into the ligand-binding site. The transmembrane regions are shown as green ribbons, and. Putative interactions are displayed as yellow dashed lines with
estimated distance in angstroms (Å). The carbon-nitrogen chain of xamoterol (S) is predicted to pack approximately 1.0 Angstroms closer to D121 and
Transmembrane Helix 3 compared to xamoterol (R), and its morpholino ring rests in a rotated pose as well.
https://doi.org/10.1371/journal.pone.0180319.g011
the binding pocket. These differences serve as potential mechanisms for the gap in affinity
between the two chiral forms of xamoterol.
Neuroimmuno-modulatory effects of the key compound
The TNFα signaling pathway has been strongly implicated in AD pathology and neuroinflam-
matory diseases [23–25]. With the aim of identifying compounds that have therapeutic poten-
tial for AD and neuroinflammatory diseases, we assessed whether our key compound STD-
101-D1 could modulate the TNFα signaling pathway. First, the effects of STD-101-D1 on the
TNFα signaling pathway were assessed in vitro by stimulating primary microglia with the bac-
terial endotoxin LPS for 4 hrs in the absence or presence of the test compounds. Previous stud-
ies with this model have shown that the effects of xamoterol on the LPS-induced TNFα
response are dependent on ADRB1; its effects were reversed by the selective ADRB1 antago-
nists CGP 20712A and betaxolol, but not by the selective ADRB2 antagonist ICI-118551 [11].
As shown in Fig 12, stimulation of primary microglia cells with LPS led to a significant
increase in TNFα levels. Treatment with isoproterenol, the unbiased full agonist of ADRB1,
inhibited the LPS-induced TNFα production by approximately 80%, whereas treatment with
xamoterol (S) inhibited LPS-induced TNFα production by approximately 55%. The key com-
pound STD-101-D1 reduced LPS-induced TNFα production by approximately 50%. In order
to investigate whether STD-101-D1 could also inhibit the TNFα response in vivo, the effects of
STD-101-D1 on mice exposed to LPS were also examined. Administration of LPS resulted in a
peripheral inflammatory response as measured by increased levels of TNFα in plasma at 90
min post-LPS (Fig 13A). The LPS-induced systemic TNFα response was markedly inhibited
by pre-treating mice with xamoterol (S) or STD-101-D1 at 3 mg/kg (Fig 13A). Administration
of LPS also led to an inflammatory response in the CNS at 90 min post-LPS. Brain tissue from
LPS-injected mice showed increased gene expression of proinflammatory cytokines TNFα, IL-
1β, and IL-6 (Fig 13B). The LPS-induced inflammatory TNFα response in the brain was atten-
uated by pre-treating mice with STD-101-D1 at 3 mg/kg (Fig 13B). We observed a trend for
attenuation of CNS IL-1β with STD-101-D1 and no effect at IL-6. Significant effects of xamo-
terol were not observed in CNS inflammatory markers at this time point, possibly due to tim-
ing of LPS response and differences in PK properties of xamoterol relative to STD-101-D1.
The finding that 129.9 28.2 ng/g (n = 5) of STD-101-D1 were detected at 105 min post-dose
in brain homogenates from the mice pretreated with STD-101-D1 (3 mg/kg) indicates that it
gets to the brain.
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
18 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 12. Inhibitory effects of ADRB1 ligands on LPS-induced TNFα response in primary microglia. Data
are represented as mean ± S.E.M.s of four independent experiments (n = 3–18 per group, * p < 0.05,
**p < 0.01, *** p < 0.001, one-way ANOVA followed by Dunnett’s multiple comparison against LPS exposure
alone).
https://doi.org/10.1371/journal.pone.0180319.g012
Pharmacokinetic properties of the key compound
The findings that STD-101-D1 has potent partial agonistic activity on ADRB1 with functional
selectivity for the cAMP pathway and produces anti-inflammatory effects suggested that it
could be the promising drug-like lead compound. Toward evaluating this intriguing possibil-
ity, we profiled the PK properties of STD-101-D1 in an in vitro microsomal stability assay and
in vivo pharmacokinetic studies. First, the metabolic stability of STD-101-D1 was assessed
together with the reference compounds verapamil (a calcium channel blocker) and proprano-
lol (a beta blocker) using microsomes from the mouse, rat, and human. In mouse microsomes,
Fig 13. Inhibitory effects of ADRB1 ligands on LPS-induced TNFα response in mice. (A) Plasma TNFα concentrations in control animals and animals
pretreated with xamoterol or STD-101-D1 90 min after LPS injection. (B) TNFα, IL1β, and IL6 mRNA expression in homogenized cortical tissue from control
mice and animals pretreated with xamoterol or STD-101-D1 90 min after LPS injection. Data are represented as mean ± S.E.M.s of three independent
experiments. (n = 4–14 per group, * p < 0.05, **p < 0.01, *** p < 0.001, one-way ANOVA followed by Dunnett’s multiple comparison).
https://doi.org/10.1371/journal.pone.0180319.g013
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
19 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 14. Metabolic stability in mouse, rat, and human microsomes. STD-101-D1 and two reference compounds verapamil and propranolol were
incubated at 0.1 uM in mouse (A), rat (B), or human (C) liver microsomes. Serial samples were removed up until 60 min. All experiments were performed in
duplicate, and data are represented as mean ± S.E.M.
https://doi.org/10.1371/journal.pone.0180319.g014
both verapamil and propranolol were readily metabolized, showing half-lives of less than 30
minutes (Fig 14A). In comparison, STD-101-D1 was stable in mouse liver microsomes, with
62.8% of the compound remaining after 60 min (Fig 14A). The rate of disappearance of verapa-
mil, propranolol, and STD-101-D1 was greater in rat microsomes as compared to mouse micro-
somes (Fig 14B). Verapamil and propranolol were readily metabolized with half-lives of 20.9
min and 7 min, respectively. STD-101-D1 was metabolized with a half-life of 25.2 min in rat
microsomes. In human microsomes, verapamil was metabolized with a half-life of 44.1 min; pro-
pranolol was stable with 72.2% of compound remaining after 60 min (Fig 14C). STD-101-D1
was very stable in human microsomes, with almost no reduction after 60 min. (Fig 14C).
In vivo pharmacokinetic properties of STD-101-D1 were also evaluated in male Sprague-
Dawley rats in 4-hr time-course PK and 20-min post-dose collection studies after IV, IP, and PO
administration of STD-101-D1. In comparison, the pharmacokinetic properties of xamoterol
were also determined. The 4-hr time-course PK study revealed that xamoterol was cleared rap-
idly (Fig 15A, Table 1). Consistent to what has been reported, xamoterol’s oral bioavailability
was low (1.7%). Concentrations of xamoterol in the jugular and portal veins were consistently
low, indicating that the low absolute oral bioavailability of xamoterol is due to poor absorption.
Compared to xamoterol, STD-101-D1 was cleared more slowly and remained in the system for a
longer period of time (Fig 15A, Table 1). Following IP injection, STD-101-D1 was rapidly and
very significantly absorbed, as evidenced by the systemic plasma concentrations (Fig 15A). The
maximum concentration (Cmax) of 762 ng/mL was achieved in systemic plasma after IP adminis-
tration at 90 min post-dose (Table 1). On the other hand, STD-101-D1 was minimally absorbed
after oral administration, resulting in low oral bioavailability (6%) (Fig 15A and 15B). In the sys-
temic (jugular vein) circulation, the Cmax of STD-101-D1 was 116 ng/ml after oral administra-
tion. The corresponding Cmax of STD-101-D1 in portal veins was 946 ng/ml (Table 1).
The 20 min post-dose collection study revealed that systemic exposure to xamoterol was
low after oral administration, in line with its low oral bioavailability. Plasma concentrations of
xamoterol at 20 min after PO administration were 0.7% and 0.2% of the plasma concentrations
achieved via IV and IP administration, respectively (Fig 15C, Table 2). STD-101-D1 adminis-
tration led to plasma concentrations comparables with those obtained from xamoterol admin-
istration. However, STD-101-D1 had higher CNS penetration, with 22-fold higher brain
concentrations after IV treatment of STD-101-D1 compared to xamoterol (Fig 15C, Table 2).
Pharmacological specificity of the key compound
In light of its promising in vitro pharmacology profiles and improved brain penetration, STD-
101-D1 was screened against a panel of CNS relevant targets, including G protein-coupled
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
20 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Fig 15. Pharmacokinetics of xamoterol and STD-101-D1. Systemic (A) and portal vein (B) plasma concentrations of xamoterol and STD-101-D1 as a
function of time after a single injection of xamoterol (10 mg/kg) or STD-101-D1 (10 mg/kg) via intravenous (IV), intraperitoneal (IP) and oral (PO)
administration. Plasma and brain (C) concentrations of xamoterol and STD-101-D1 in rats collected 20 min after a single injection of xamoterol or STD-
101-D1 (10 mg/kg) via IV, IP, and PO administration. Data are represented as mean ± SEM (n = 3 per route).
https://doi.org/10.1371/journal.pone.0180319.g015
receptors and transporters [26]. In comparison, off-target binding of xamoterol (S) was also
determined. Xamoterol (S) displays a distinct preference for ADRB1 and shows low off-target
affinity (Ki > 10 μM) for a broad range of neurotransmitter transporters, ion channels and
other CNS proteins (including opioid, dopamine, serotonin, nicotinic acetylcholine, musca-
rinic acetylcholine, and N-methyl-D-aspartate receptors) (S1 Table). In contrast, STD-101-D1
displays a significant affinity for non-ADRB1 binding including 5-HT1A, 5-HT2B, α1A adren-
ergic, α1D adrenergic, and D3 dopaminergic receptors, while maintaining nominal affinity for
several other CNS proteins including opioid, histamine, and muscarinic acetylcholine recep-
tors (S2 Table). The fact that STD-101-D1 binds to non-ADRB1 CNS targets suggests that its
neuroimmune-modulatory effects shown in in vitro and in vivo LPS studies should be inter-
preted cautiously.
Table 1. Pharmacokinetic parameters of xamoterol and STD-101-D1 determined in the 4-hr time-course study.
Route
C
max (ng/mL)
AUCinf(h*ng/mL)
2446 ± 601
2834 ± 517
42.5 ± 29.5
175.9 ± 53.4
1796 ± 256
2159 ± 310
101 ± 29
Xamoterol
IV
IP
22200 ± 4029
7443 ± 730
23 ± 16
PO–Systemic
PO–Portal
IV
79.8 ± 23.8
3997 ± 1107
762 ± 67
STD-101-D1
IP
PO–Systemic
PO–Portal
116 ± 78
946 ± 867
362 ± 367
https://doi.org/10.1371/journal.pone.0180319.t001
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
21 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Table 2. Pharmacokinetic parameters of xamoterol and STD-101-D1 determined in the 20-min post-dose collection study.
Plasma concentration (ng/ml)
Brain concentration
(ng/g)
Xamoterol
IV
IP
1121 ± 297.1
3113 ± 388.6
8.1 ± 2.0
38.3 ± 9.3
26.5 ± 8.4
15.8 ± 5.7
PO
IV
STD-101-D1
1813 ± 67.41
597.7 ± 149.8
56.9 ± 29.2
838.7 ± 23.6
96.0 ± 22.6
41.7 ± 5.3
IP
PO
https://doi.org/10.1371/journal.pone.0180319.t002
Cardiovascular effects of the key compound
ADRB1 is highly expressed in cardiac tissue and plays an important role in regulating cardiac
function. To determine if xamoterol and STD-101-D1 has peripheral effects, we evaluated the
effects of xamoterol or STD-101-D1 on heart rates and blood pressures in anesthetized rats. At
a dose of 3 mg/kg, xamoterol and STD-101-D1 increased heart rates by 4% and 13%, respec-
tively (Fig 16A). In addition, xamoterol and STD-101-D1 decreased blood pressures by 9%
and 46%, respectively (Fig 16B).
Discussion
In this study we established SARs for a novel chemotype targeting the ADRB1. The key com-
pound, STD-101-D1, produced partial agonistic activity on G protein signaling with an EC50
value in the low nanomolar range, but engaged very little β-arrestin signaling compared to the
unbiased agonist isoproterenol. This biased ligand represents a new mode for ADRB1 activa-
tion and is distinctly different compared to isoproterenol—a full and unbiased agonist.
ADRB1 has been known to play an important role in learning and memory functions. For
example, the ADRB1 selective antagonist betaxolol has been shown to induce contextual mem-
ory impairment in mice, which was reversed by the ADRB1 selective partial agonist xamoterol
in a dose-dependent manner [10]. Similarly, the retrieval deficits exhibited by mice with NA
deficiency have been rescued by the ADRB1 selective partial agonist xamoterol [27]. The
involvement of ADRB1 in learning and memory has important clinical and therapeutic impli-
cations for AD, as severe neurodegeneration of the NA system begins in the early stages of AD.
It is believed that loss of NA signaling and the resulting hypoactivation of ADRB1 may par-
tially contribute to the cognitive symptoms in AD. Therefore, ADRB1 agonists may provide a
Fig 16. Effects of xamoterol and STD-101-D1 on heart rate and blood pressure. Changes in heart rate (A) and blood pressure (B) following
subcutaneous administration of xamoterol or STD-101-D1 at dose of 3 mg/kg. Date are represented as mean ± SEM (n = 3 per compound). (One-sample t-
test vs. 0% theoretical mean, * p < 0.05).
https://doi.org/10.1371/journal.pone.0180319.g016
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
22 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
promising therapeutic strategy to improve cognitive function in AD by restoring the lost NA
signaling. Given the well-characterized role of protein kinase A (PKA)/cAMP response ele-
ment-binding protein (CREB) signaling in learning and memory [28, 29], the cyclic adenosine
monophosphate (cAMP) signaling pathway downstream of ADRB1 is believed to mediate the
cognitive-enhancing effects of ADRB1 agonists. Toward our goal of modulating ADRB1 to
produce therapeutic benefits in AD, we specifically sought to discover partial agonists; these
would have more subtle effects in the periphery yet be efficacious enough to restore the
decreased NA signaling found in AD. Indeed, human trials with ADRB1 partial agonists such
as xamoterol were shown to be very safe, suggesting lack of toxicity with this class of com-
pounds [30, 31]. Accordingly, we chose xamoterol, a known partial agonist of ADRB1, as our
lead compound and exploited SAR around its analogs.
The SAR studies described here showed that, in general, substitution of the phenol ring of
xamoterol was not well tolerated. This was especially true for the ortho position to the phenolic
OH group, where alkyl groups and Cl led to a complete loss of activity. This appeared to be
largely a steric effect, as substitution with the small F atom ortho to the OH group gave a com-
pound with only a small diminution in potency. In contrast, substitutions at the morpholino
urea site were relatively well tolerated although all were less potent than xamoterol. Aside from
the two azetidine urea analogs, all of the compounds showed greater efficacy at the cAMP
pathway via ADRB1 compared to xamoterol (S). This effect did not correlate with lipophilicity
(cLogP), nor any other obvious structural features of this subset of molecules. It is therefore
interesting to speculate that the observed higher efficacy is due to the increased conformational
flexibility of the side chains of the non-urea compounds and their consequent ability to fit
more readily into the binding sites of the receptor.
Biased agonism, the notion that ligands at GPCRs can preferentially stimulate one intracel-
lular signaling pathway over another, is an emerging concept in GPCR signaling [17, 32, 33].
There is growing interest in the development of biased GPCR ligands, because biased ligands
may provide safer and more efficacious therapeutic benefits compared to non-biased ligands
by selectively targeting a subset of the receptor-mediated signaling [34]. As a result, the list of
known biased agonists for GPCRs including the dopamine D2 receptor, serotonin 5-HT2A,
cannabinoid CB1, and κ-opioid receptors, is growing and some have already progressed to
clinical development [35–40]. A key property of xamoterol (S) and its analogs (STD-101-D1
through STD-101-D6) is the functional bias toward the G protein-mediated cAMP pathway. G
protein-mediated signaling of ADRB1 is implicated in cognitive function and neuroinflamma-
tion, whereas β arrestin-mediated signaling is implicated in development of drug-induced tol-
erance. Therefore, the development of a G protein-biased agonist of ADRB1 may provide a
means to optimally tune ADRB1 therapeutics that will ameliorate the cognitive deficits and
pathology underlying AD, as well as other neuroinflammatory diseases, without producing sig-
nificant tachyphylaxis. Biased ligands, such as those reported in this study, can be used as phar-
macological tools to aid in the elucidation of ADRB1-mediated signaling cascades in cellular
systems and in vivo. It will be of great interest to utilize these newly identified functionally
selective G protein-biased ADRB1 agonists in additional in vitro and in vivo assays to deter-
mine the contribution of G-protein signaling to ADRB1-mediated effects.
Computational modeling has enabled us to capture poses of the two enantiomers of xamo-
terol and the ADRB1 agonist STD-101-D1 consistent with experimental data. The docked
poses feature polar contacts between the phenolic hydroxyl of xamoterol (S) and both S211 as
well as with N310. Furthermore, the binding pocket residue D121 dually forms a salt bridge
with the protonated secondary amine of xamoterol’s backbone as well as a hydrogen bond
with the beta amino alcohol of xamoterol. The experimental compound STD-101-D1 is pre-
dicted to form analogous sets of contacts, with the primary difference originating from the
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
23 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
position of its methoxy phenyl ring compared to the heterocyclic ring of xamoterol. Previous
studies have identified the contact with S211 as being important for mediating agonism [41].
Further insights into the structure-activity relationships of xamoterol and its derivatives, such
as STD-101-D1, would benefit greatly from a fully active, G protein-bound or G protein-
mimetic nanobody-bound crystal structure of ADRB1.
Accumulating data suggest a close association between neuroinflammation and AD patho-
genesis. Prominent activation of immune responses characterized by activated microglia,
reactive astrocytes, and increased expression of complement factors and proinflammatory
cytokines associated with Aβ deposits, have been observed in the brains of AD patients as well
as in transgenic mouse models of AD [42, 43]. TNFα is one of the main proinflammatory cyto-
kines known to be elevated in brains from AD patients and animal models of AD, and has
been strongly implicated in AD pathology. For example, elevated levels of TNFα are observed
in serum and cerebrospinal fluid from AD patients compared to age-matched controls [44,
45]. Similarly, overexpression of TNFα has been shown in several animal models of AD,
including the 3xTg-AD and 5XFAD mouse models [11, 46, 47]. More importantly, increased
TNFα in AD models has been correlated with disease progression [48]. At the molecular level,
TNFα has been shown to exacerbate Aβ-induced apoptosis in neurons and increase Aβ pro-
duction by upregulating both β-secretase expression and γ-secretase activity, as well as the
expression of APP [49–53]. It has also been demonstrated that TNFα inhibits phagocytosis of
toxic Aβ species, which might lead to hindering efficient plaque removal by brain resident
microglia [54]. Taken together, these observations suggest that excessive expression of TNFα
may contribute to and accelerate the progression of AD. Targeted inhibition of TNFα signal-
ing in AD may be an effective therapeutic approach to halt or attenuate the progression of AD.
In support of this idea, a recent study demonstrates that inhibition of TNFα signaling prevents
pre-plaque amyloid-associated neuropathology and reduces plaque accumulation and tau
phosphorylation in transgenic mouse models of AD [55, 56]. Relevant to the present study, the
adrenergic system has been shown to be involved in the regulation of TNFα signaling as well
as general peripheral inflammatory responses and CNS inflammatory responses [57–59]. In
our previous in vitro study with primary microglia, we showed that the highly selective
ADRB1 agonist xamoterol inhibited the LPS-induced TNFα response [11]. Its effects were
reversed by the ADRB1 selective antagonists CGP 20712A and betaxolol, but not by the
ADRB2 selective antagonist ICI-118551, suggesting that xamoterol produces its anti-inflam-
matory effects on the TNFα response via ADRB1. When chronically administered to the
5XFAD mouse model of AD, xamoterol also produced anti-inflammatory effects and attenu-
ated increased expressions of proinflammatory markers, including TNFα, shown in the brains
of the transgenic mice. This suggests that ADRB1 is an important player in regulating the
immune response, and modulating ADRB1 activity has therapeutic potential for AD as well as
other neuroinflammatory diseases. Notably, the key compound STD-101-D1 was found to
suppress TNFα production in rat primary microglia challenged with LPS. When administered
to mice prior to the LPS challenge at 3 mg/kg, STD-101-D1 also attenuated the acute periph-
eral and CNS TNFα response induced by LPS. Given the well-established suppressive effects of
cAMP on LPS-induced transcriptional activation of the TNFα gene [60–62], we speculate that
a G protein-biased agonist such as STD-101-D1 acts at the transcription level via a cAMP
dependent mechanism. Our observations that the full ADRB agonist isoproterenol or the par-
tial agonist STD-101-D1 did not produce anti-inflammatory effects when administered after
the LPS challenge (data not shown), and that they require the pretreatment period to produce
inhibitory effects on the LPS-induced inflammatory response, support this hypothesis. How-
ever, it should be noted that the current study measures the acute LPS-induced immune
response. It will be intriguing to determine whether this class of compound produces anti-
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
24 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
inflammatory effects when given after the LPS challenge in the long-term LPS exposure model.
It is also important to note that STD-101-D1 has non-specific activity on other receptors and
binds to several receptors such as 5-HT 1A and 5-HT 2B with comparable affinity to ADRB1.
Thus, it is possible that its anti-inflammatory effects shown in vitro and in vivo could be medi-
ated by non-ADRB1. Additional studies are necessary to validate that STD-101-D1 exerts its
anti-inflammatory effects via ADRB1 and to identify more specific ADRB1 partial agonists.
Regardless, identification of ADRB1 agonists with anti-inflammatory effects such as STD-
101-D1 has important clinical implications.
Xamoterol is a highly selective partial agonist of ADRB1 with functional bias for the cAMP
pathway over the β-arrestin pathway [11]. However, its therapeutic utility as a CNS drug is lim-
ited, as it has very poor oral bioavailability and low CNS penetration (Fig 15). Despite this low
penetration, however, xamoterol was shown to produce biological effects in the brain after sys-
temic administration. For example, a systemic acute treatment with xamoterol led to increases
in phosphorylation of CREB in the brain, an important event involved in learning and mem-
ory [9]. In multiple mouse models of AD, chronic systemic administration of xamoterol was
also shown to enhance cognitive functions and attenuate pathological features of AD [9, 11].
Based on this observation, we hypothesized that we could maximize the therapeutic utility of
this class of drugs for CNS indications by optimizing PK properties of xamoterol. With favor-
able CNS permeability, ADRB1 agonists can be administered at low doses, which will reduce
systemic exposure and lower the peripheral effects, including cardiovascular effects. In order
to evaluate the PK properties of STD-101-D1, we conducted in vitro microsomal stability tests
and in vivo PK studies. In the in vitro microsome stability test, STD-101-D1 was shown to be
very stable, with half-lives greater than 60 min, both in mouse and human microsomes. In rat
microsomes, however, STD-101-D1 was relatively less stable and metabolized with a half-life
of 25.2 min. This observation indicates a marked species difference in the metabolism of STD-
101-D1. Given this faster rate of metabolism in rat liver microsomes compared to mouse liver
microsomes, the in vivo PK study in rat uses a higher dose of STD-101-D1 (i.e. 10 mg/kg) com-
pared to the dose used in the in vivo mouse LPS study (i.e. 3 mg/kg). In the in vivo PK study,
STD-101-D1 was moderately cleared in Sprague-Dawley rat with a half-life of 2.7 hr after oral
administration, in line with its moderate metabolic stability shown in rat microsomes. How-
ever, the oral bioavailability of STD-101-D1 remains low (6%), similar to xamoterol. The low
oral bioavailability of STD-101-D1 could be due to poor absorption through the gut mem-
brane and/or efflux through the P-glycoprotein.
Obtaining compounds with good brain permeability is a major hurdle in CNS drug devel-
opment. Importantly, STD-101-D1 shows greater CNS penetration compared to xamoterol
(Fig 15). For example, when administered intravenously, STD-101-D1 achieved an approxi-
mately 22-fold higher brain concentration compared to xamoterol at 20 min post dose. With
its improved CNS permeability, STD-101-D1 can be administered at lower doses for CNS indi-
cations, reducing systemic exposure and lowering cardiovascular effects. Of note, the brain
concentrations achieved with STD-101-D1 are considered to be within the therapeutically rele-
vant range. For example, when administered intravenously or intraperitoneally, acute single
administration of STD-101-D1 at the dose of 10 mg/kg led to brain concentrations of 2.5 μM
and 288 nM, respectively. As these values are up to 148-fold higher than its EC50 value (16.9
nM) for ADRB1 in vitro, we believe brain concentrations achieved with a single administration
of STD-101-D1, even at lower doses, would be sufficient to induce the CNS target engagement
needed for efficacy. This brain concentration is also in line with other drugs clinically used for
AD, such as memantine and donepezil [63, 64]. Collectively, this suggests a possible beneficial
therapeutic value of STD-101-D1 for CNS indications, although its overall PK properties
remain to be improved. It is important to note that only a trace of STD-101-D1 was detected
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
25 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
in the brain at 4 hrs post dose, indicating that most of the drug is cleared from brain within 4
hrs. Further medicinal chemistry efforts would be needed to identify new compounds that
have improved oral bioavailability and longer half-lives.
ADRB1 is highly expressed in a number of peripheral organs, and plays an important role
in mediating multiple physiological processes. Thus, modulators of ADRB1 could possibly
produce many unwanted peripheral adverse effects. Using a medicinal chemistry approach,
our laboratory is developing partial agonists of ADRB1 with improved brain penetration and,
as a result, we are seeking to decrease peripheral exposure with lower dosing, while achieving
efficacious concentrations in the CNS. As our aim was to develop an ADRB1 partial agonist
with minimal peripheral effects, we have evaluated the peripheral effects of the key compound
STD-101-D1. Notably, at equivalent doses, both xamoterol and STD-101-D1 produced signifi-
cant changes in heart rates and blood pressures of rats after an acute single administration.
The effects of STD-101-D1 were more profound than those of xamoterol, possibility due to its
higher efficacy on ADRB1. Importantly, higher brain permeability suggests that we can reduce
the dose of STD-101-D1, thereby minimizing peripheral exposure. It has been previously
shown that chronic administration of xamoterol (3 mg/kg, subcutaneous administration) does
not lead to significant changes in cardiovascular function or cardiac structure in mice [11]. It
will be important to determine whether chronic administration of STD-101-D1 at a lower
dose leads to changes in cardiovascular function or cardiac structure.
In summary, this study has described the rational design of a novel series of compounds with
ADRB1 agonist activity. Our findings indicate that we have successfully identified biased ligands
with unique pharmacology. Such biased ligands will be invaluable research tools to dissect out
the G protein-coupled receptor signaling transduction mechanisms, and also act as potential
lead compounds for further development to provide safer, more efficacious therapeutics.
Supporting information
S1 Fig. HPLC and NMR spectra for xamoterol (S), xamoterol (R), STD-101-B1 to B9, D1
to D6, and E.
(PDF)
S1 Table. Binding affinity (Ki, nM) of xamoterol (S) at non-ADRB1 binding sites.
(PDF)
S2 Table. Binding affinity (Ki, nM) of STD-101-D1 at non-ADRB1 binding sites.
(PDF)
Acknowledgments
Receptor binding profile was performed by the National Institute of Mental Health’s Psychoac-
tive Drug Screening Program (NIMH PDSP). We thank the director of NIMH PDSP, Bryan L.
Roth MD, PhD, at the University of North Carolina at Chapel Hill and Project Officer Jamie
Driscoll at NIMH, Bethesda MD, USA. We thank Stanford Behavioral and Functional Neuro-
science Laboratory for excellent technical support of in vivo pharmacokinetic and pharmaco-
logical studies.
Author Contributions
Conceptualization: BY AJ AE VP MS MG.
Data curation: BY.
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
26 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
Formal analysis: BY.
Funding acquisition: MS.
Investigation: BY DB KR JE EF AF WS JR.
Methodology: EF AF.
Project administration: BY MS.
Resources: BY.
Supervision: MS.
Visualization: BY EF.
Writing – original draft: BY.
Writing – review & editing: MS.
References
1. Hall RA. Beta-adrenergic receptors and their interacting proteins. Semin Cell Dev Biol. 2004; 15
(3):281–8. https://doi.org/10.1016/j.semcdb.2003.12.017 PMID: 15125891.
2. Bylund DB, Eikenberg DC, Hieble JP, Langer SZ, Lefkowitz RJ, Minneman KP, et al. International
Union of Pharmacology nomenclature of adrenoceptors. Pharmacol Rev. 1994; 46(2):121–36. PMID:
7938162.
3. Sara SJ. The locus coeruleus and noradrenergic modulation of cognition. Nat Rev Neurosci. 2009; 10
(3):211–23. Epub 2009/02/05. nrn2573 10.1038/nrn2573. PMID: 19190638. https://doi.org/10.1038/
nrn2573
4. Feinstein DL, Heneka MT, Gavrilyuk V, Dello Russo C, Weinberg G, Galea E. Noradrenergic regulation
of inflammatory gene expression in brain. Neurochem Int. 2002; 41(5):357–65. Epub 2002/08/15.
S0197018602000499. PMID: 12176079.
5. Gorman AL, Dunn AJ. Beta-adrenergic receptors are involved in stress-related behavioral changes.
Pharmacol Biochem Behav. 1993; 45(1):1–7. PMID: 8100069.
6. Philipp M, Hein L. Adrenergic receptor knockout mice: distinct functions of 9 receptor subtypes. Phar-
macol Ther. 2004; 101(1):65–74. PMID: 14729393.
7. Frishman WH. beta-Adrenergic blockade in cardiovascular disease. J Cardiovasc Pharmacol Ther.
2013; 18(4):310–9. https://doi.org/10.1177/1074248413484986 PMID: 23637119.
8. Cruickshank JM. Beta-blockers and heart failure. Indian Heart J. 62(2):101–10. Epub 2010/12/25.
PMID: 21180298.
9. Coutellier L, Ardestani PM, Shamloo M. beta1-adrenergic receptor activation enhances memory in Alz-
heimer’s disease model. Ann Clin Transl Neurol. 2014; 1(5):348–60. Epub 2014/06/03. https://doi.org/
10.1002/acn3.57 PMID: 24883337; PubMed Central PMCID: PMC4036739.
10. Faizi M, Bader PL, Tun C, Encarnacion A, Kleschevnikov A, Belichenko P, et al. Comprehensive behav-
ioral phenotyping of Ts65Dn mouse model of Down syndrome: activation of beta1-adrenergic receptor
by xamoterol as a potential cognitive enhancer. Neurobiol Dis. 2011; 43(2):397–413. Epub 2011/04/30.
https://doi.org/10.1016/j.nbd.2011.04.011 PMID: 21527343; PubMed Central PMCID: PMC3539757.
11. Ardestani PM, Evans AK, Yi B, Nguyen T, Coutellier L, Shamloo M. Modulation of neuroinflammation
and pathology in the 5XFAD mouse model of Alzheimer’s disease using a biased and selective beta-1
adrenergic receptor partial agonist. Neuropharmacology. 2017; 116:371–86. https://doi.org/10.1016/j.
neuropharm.2017.01.010 PMID: 28089846; PubMed Central PMCID: PMCPMC5385159.
12. Madrigal JL, Feinstein DL, Dello Russo C. Norepinephrine protects cortical neurons against microglial-
induced cell death. J Neurosci Res. 2005; 81(3):390–6. https://doi.org/10.1002/jnr.20481 PMID:
15948176.
13. Madrigal JL, Leza JC, Polak P, Kalinin S, Feinstein DL. Astrocyte-derived MCP-1 mediates neuropro-
tective effects of noradrenaline. J Neurosci. 2009; 29(1):263–7. Epub 2009/01/09. https://doi.org/10.
1523/JNEUROSCI.4926-08.2009 PMID: 19129402.
14. Markus T, Hansson SR, Cronberg T, Cilio C, Wieloch T, Ley D. beta-Adrenoceptor activation depresses
brain inflammation and is neuroprotective in lipopolysaccharide-induced sensitization to oxygen-
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
27 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
glucose deprivation in organotypic hippocampal slices. J Neuroinflammation. 2010; 7:94. https://doi.
org/10.1186/1742-2094-7-94 PMID: 21172031; PubMed Central PMCID: PMCPMC3017519.
15. Violin JD, Lefkowitz RJ. Beta-arrestin-biased ligands at seven-transmembrane receptors. Trends Phar-
macol Sci. 2007; 28(8):416–22. https://doi.org/10.1016/j.tips.2007.06.006 PMID: 17644195.
16. van der Westhuizen ET, Breton B, Christopoulos A, Bouvier M. Quantification of ligand bias for clinically
relevant beta2-adrenergic receptor ligands: implications for drug taxonomy. Mol Pharmacol. 2014; 85
(3):492–509. https://doi.org/10.1124/mol.113.088880 PMID: 24366668.
17. Kenakin T, Christopoulos A. Signalling bias in new drug discovery: detection, quantification and thera-
peutic impact. Nat Rev Drug Discov. 2013; 12(3):205–16. https://doi.org/10.1038/nrd3954 PMID:
23411724.
18. Bastain W, Boyce MJ, Stafford LE, Morton PB, Clarke DA, Marlow HF. Pharmacokinetics of xamoterol
after intravenous and oral administration to volunteers. Eur J Clin Pharmacol. 1988; 34(5):469–73.
Epub 1988/01/01. PMID: 2904884.
19. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc.
2008; 3(6):1101–8. PMID: 18546601.
20. Strubbe JH, Bruggink JE, Steffens AB. Hepatic portal vein cannulation for infusion and blood sampling
in freely moving rats. Physiol Behav. 1999; 65(4–5):885–7. PMID: 10073496.
21. Murakami T, Nakanishi M, Yoshimori T, Okamura N, Norikura R, Mizojiri K. Separate assessment of
intestinal and hepatic first-pass effects using a rat model with double cannulation of the portal and jugu-
lar veins. Drug Metab Pharmacokinet. 2003; 18(4):252–60. PMID: 15618743.
22. Kusiak JW, Heja G, Pitha J. Two beta-adrenergic pharmacophores on the same molecule. A set of ago-
nist-antagonist combinations. Biochem Pharmacol. 1987; 36(2):269–75. Epub 1987/01/15. 0006-2952
(87)90700-3. PMID: 2880591.
23. Ma SL, Tang NL, Lam LC, Chiu HF. Association between tumor necrosis factor-alpha promoter poly-
morphism and Alzheimer’s disease. Neurology. 2004; 62(2):307–9. PMID: 14745077.
24. Fillit H, Ding WH, Buee L, Kalman J, Altstiel L, Lawlor B, et al. Elevated circulating tumor necrosis factor
levels in Alzheimer’s disease. Neurosci Lett. 1991; 129(2):318–20. PMID: 1745413.
25. McCoy MK, Tansey MG. TNF signaling inhibition in the CNS: implications for normal brain function and
neurodegenerative disease. J Neuroinflammation. 2008; 5:45. https://doi.org/10.1186/1742-2094-5-45
PMID: 18925972; PubMed Central PMCID: PMCPMC2577641.
26. Besnard J, Ruda GF, Setola V, Abecassis K, Rodriguiz RM, Huang XP, et al. Automated design of
ligands to polypharmacological profiles. Nature. 2012; 492(7428):215–20. https://doi.org/10.1038/
nature11691 PMID: 23235874; PubMed Central PMCID: PMCPMC3653568.
27. Murchison CF, Zhang XY, Zhang WP, Ouyang M, Lee A, Thomas SA. A distinct role for norepinephrine
in memory retrieval. Cell. 2004; 117(1):131–43. Epub 2004/04/07. S0092867404002594. PMID:
15066288.
28. Suzuki A, Fukushima H, Mukawa T, Toyoda H, Wu LJ, Zhao MG, et al. Upregulation of CREB-mediated
transcription enhances both short- and long-term memory. J Neurosci. 2011; 31(24):8786–802. Epub
2011/06/17. https://doi.org/10.1523/JNEUROSCI.3257-10.2011 PMID: 21677163.
29. De Luca A, Giuditta A. Role of a transcription factor (CREB) in memory processes. Riv Biol. 1997; 90
(3):371–84. Epub 1997/01/01. PMID: 9549403.
30. Waller DG, Webster J, Sykes CA, Bhalla KK, Wray R. Clinical efficacy of xamoterol, a beta 1-adreno-
ceptor partial agonist, in mild to moderate heart failure. U.K. Xamoterol Study Group. Eur Heart J. 1989;
10(11):1003–10. Epub 1989/11/01. PMID: 2574108.
31. Ang EL, Chan WL, Cleland JG, Moore D, Krikler SJ, Alexander ND, et al. Placebo controlled trial of
xamoterol versus digoxin in chronic atrial fibrillation. Br Heart J. 1990; 64(4):256–60. PMID: 1977430;
PubMed Central PMCID: PMCPMC1024416.
32. Kenakin T. Functional selectivity and biased receptor signaling. J Pharmacol Exp Ther. 2011; 336
(2):296–302. https://doi.org/10.1124/jpet.110.173948 PMID: 21030484.
33. Rajagopal S, Rajagopal K, Lefkowitz RJ. Teaching old receptors new tricks: biasing seven-transmem-
brane receptors. Nat Rev Drug Discov. 2010; 9(5):373–86. https://doi.org/10.1038/nrd3024 PMID:
20431569; PubMed Central PMCID: PMCPMC2902265.
34. Whalen EJ, Rajagopal S, Lefkowitz RJ. Therapeutic potential of beta-arrestin- and G protein-biased
agonists. Trends Mol Med. 2011; 17(3):126–39. https://doi.org/10.1016/j.molmed.2010.11.004 PMID:
21183406; PubMed Central PMCID: PMCPMC3628754.
35. Ogawa S, Watanabe T, Moriyuki K, Goto Y, Yamane S, Watanabe A, et al. Structural optimization and
structure-functional selectivity relationship studies of G protein-biased EP2 receptor agonists. Bioorg
Med Chem Lett. 2016; 26(10):2446–9. https://doi.org/10.1016/j.bmcl.2016.03.110 PMID: 27055938.
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
28 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
36. Zhou L, Lovell KM, Frankowski KJ, Slauson SR, Phillips AM, Streicher JM, et al. Development of func-
tionally selective, small molecule agonists at kappa opioid receptors. J Biol Chem. 2013; 288
(51):36703–16. https://doi.org/10.1074/jbc.M113.504381 PMID: 24187130; PubMed Central PMCID:
PMCPMC3868780.
37. Moller D, Kling RC, Skultety M, Leuner K, Hubner H, Gmeiner P. Functionally selective dopamine D(2),
D(3) receptor partial agonists. J Med Chem. 2014; 57(11):4861–75. https://doi.org/10.1021/jm5004039
PMID: 24831693.
38. Allen JA, Yost JM, Setola V, Chen X, Sassano MF, Chen M, et al. Discovery of beta-arrestin-biased
dopamine D2 ligands for probing signal transduction pathways essential for antipsychotic efficacy. Proc
Natl Acad Sci U S A. 2011; 108(45):18488–93. https://doi.org/10.1073/pnas.1104807108 PMID:
22025698; PubMed Central PMCID: PMCPMC3215024.
39. Berg KA, Maayani S, Goldfarb J, Clarke WP. Pleiotropic behavior of 5-HT2A and 5-HT2C receptor ago-
nists. Ann N Y Acad Sci. 1998; 861:104–10. PMID: 9928246.
40. Breivogel CS, Lambert JM, Gerfin S, Huffman JW, Razdan RK. Sensitivity to delta9-tetrahydrocannabi-
nol is selectively enhanced in beta-arrestin2 -/- mice. Behav Pharmacol. 2008; 19(4):298–307. https://
doi.org/10.1097/FBP.0b013e328308f1e6 PMID: 18622177; PubMed Central PMCID:
PMCPMC2751575.
41. Warne T, Moukhametzianov R, Baker JG, Nehme R, Edwards PC, Leslie AG, et al. The structural basis
for agonist and partial agonist action on a beta(1)-adrenergic receptor. Nature. 2011; 469(7329):241–4.
https://doi.org/10.1038/nature09746 PMID: 21228877; PubMed Central PMCID: PMCPMC3023143.
42. Heneka MT, Carson MJ, El Khoury J, Landreth GE, Brosseron F, Feinstein DL, et al. Neuroinflamma-
tion in Alzheimer’s disease. Lancet Neurol. 2015; 14(4):388–405. https://doi.org/10.1016/S1474-4422
(15)70016-5 PMID: 25792098.
43. Heneka MT, O’Banion MK. Inflammatory processes in Alzheimer’s disease. J Neuroimmunol. 2007;
184(1–2):69–91. Epub 2007/01/16. https://doi.org/10.1016/j.jneuroim.2006.11.017 PMID: 17222916.
44. Tarkowski E, Liljeroth AM, Minthon L, Tarkowski A, Wallin A, Blennow K. Cerebral pattern of pro- and
anti-inflammatory cytokines in dementias. Brain Res Bull. 2003; 61(3):255–60. PMID: 12909295.
45. Alvarez A, Cacabelos R, Sanpedro C, Garcia-Fantini M, Aleixandre M. Serum TNF-alpha levels are
increased and correlate negatively with free IGF-I in Alzheimer disease. Neurobiol Aging. 2007; 28
(4):533–6. https://doi.org/10.1016/j.neurobiolaging.2006.02.012 PMID: 16569464.
46. Janelsins MC, Mastrangelo MA, Oddo S, LaFerla FM, Federoff HJ, Bowers WJ. Early correlation of
microglial activation with enhanced tumor necrosis factor-alpha and monocyte chemoattractant protein-
1 expression specifically within the entorhinal cortex of triple transgenic Alzheimer’s disease mice. J
Neuroinflammation. 2005; 2:23. https://doi.org/10.1186/1742-2094-2-23 PMID: 16232318; PubMed
Central PMCID: PMCPMC1276812.
47. Grinan-Ferre C, Sarroca S, Ivanova A, Puigoriol-Illamola D, Aguado F, Camins A, et al. Epigenetic
mechanisms underlying cognitive impairment and Alzheimer disease hallmarks in 5XFAD mice. Aging
(Albany NY). 2016; 8(4):664–84. https://doi.org/10.18632/aging.100906 PMID: 27013617; PubMed
Central PMCID: PMCPMC4925821.
48. Tarkowski E, Andreasen N, Tarkowski A, Blennow K. Intrathecal inflammation precedes development
of Alzheimer’s disease. J Neurol Neurosurg Psychiatry. 2003; 74(9):1200–5. PMID: 12933918; PubMed
Central PMCID: PMCPMC1738668. https://doi.org/10.1136/jnnp.74.9.1200
49. Blasko I, Schmitt TL, Steiner E, Trieb K, Grubeck-Loebenstein B. Tumor necrosis factor alpha aug-
ments amyloid beta protein (25–35) induced apoptosis in human cells. Neurosci Lett. 1997; 238(1–
2):17–20. PMID: 9464644.
50. Yamamoto M, Kiyota T, Horiba M, Buescher JL, Walsh SM, Gendelman HE, et al. Interferon-gamma
and tumor necrosis factor-alpha regulate amyloid-beta plaque deposition and beta-secretase expres-
sion in Swedish mutant APP transgenic mice. Am J Pathol. 2007; 170(2):680–92. https://doi.org/10.
2353/ajpath.2007.060378 PMID: 17255335; PubMed Central PMCID: PMCPMC1851864.
51. Liao YF, Wang BJ, Cheng HT, Kuo LH, Wolfe MS. Tumor necrosis factor-alpha, interleukin-1beta, and
interferon-gamma stimulate gamma-secretase-mediated cleavage of amyloid precursor protein through
a JNK-dependent MAPK pathway. J Biol Chem. 2004; 279(47):49523–32. https://doi.org/10.1074/jbc.
M402034200 PMID: 15347683.
52. Kuo LH, Hu MK, Hsu WM, Tung YT, Wang BJ, Tsai WW, et al. Tumor necrosis factor-alpha-elicited
stimulation of gamma-secretase is mediated by c-Jun N-terminal kinase-dependent phosphorylation of
presenilin and nicastrin. Mol Biol Cell. 2008; 19(10):4201–12. https://doi.org/10.1091/mbc.E07-09-0987
PMID: 18667537; PubMed Central PMCID: PMCPMC2555961.
53. Lahiri DK, Chen D, Vivien D, Ge YW, Greig NH, Rogers JT. Role of cytokines in the gene expression of
amyloid beta-protein precursor: identification of a 5’-UTR-binding nuclear factor and its implications in
Alzheimer’s disease. J Alzheimers Dis. 2003; 5(2):81–90. PMID: 12719626.
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
29 / 30

CNS permeable and G protein-biased beta-1 adrenergic receptor partial agonists
54. Koenigsknecht-Talboo J, Landreth GE. Microglial phagocytosis induced by fibrillar beta-amyloid and
IgGs are differentially regulated by proinflammatory cytokines. J Neurosci. 2005; 25(36):8240–9.
https://doi.org/10.1523/JNEUROSCI.1808-05.2005 PMID: 16148231.
55. Shi JQ, Shen W, Chen J, Wang BR, Zhong LL, Zhu YW, et al. Anti-TNF-alpha reduces amyloid plaques
and tau phosphorylation and induces CD11c-positive dendritic-like cell in the APP/PS1 transgenic
mouse brains. Brain Res. 2011; 1368:239–47. https://doi.org/10.1016/j.brainres.2010.10.053 PMID:
20971085.
56. McAlpine FE, Lee JK, Harms AS, Ruhn KA, Blurton-Jones M, Hong J, et al. Inhibition of soluble TNF
signaling in a mouse model of Alzheimer’s disease prevents pre-plaque amyloid-associated neuropa-
thology. Neurobiol Dis. 2009; 34(1):163–77. PMID: 19320056; PubMed Central PMCID:
PMCPMC2948857.
57. Hetier E, Ayala J, Bousseau A, Prochiantz A. Modulation of interleukin-1 and tumor necrosis factor
expression by beta-adrenergic agonists in mouse ameboid microglial cells. Exp Brain Res. 1991; 86
(2):407–13. PMID: 1684549.
58. Guirao X, Kumar A, Katz J, Smith M, Lin E, Keogh C, et al. Catecholamines increase monocyte TNF
receptors and inhibit TNF through beta 2-adrenoreceptor activation. Am J Physiol. 1997; 273(6 Pt 1):
E1203–8. PMID: 9435537.
59. Yoshimura T, Kurita C, Nagao T, Usami E, Nakao T, Watanabe S, et al. Inhibition of tumor necrosis fac-
tor-alpha and interleukin-1-beta production by beta-adrenoceptor agonists from lipopolysaccharide-
stimulated human peripheral blood mononuclear cells. Pharmacology. 1997; 54(3):144–52. PMID:
9127437.
60. Parry GC, Mackman N. Role of cyclic AMP response element-binding protein in cyclic AMP inhibition of
NF-kappaB-mediated transcription. J Immunol. 1997; 159(11):5450–6. PMID: 9548485.
61. Wall EA, Zavzavadjian JR, Chang MS, Randhawa B, Zhu X, Hsueh RC, et al. Suppression of LPS-
induced TNF-alpha production in macrophages by cAMP is mediated by PKA-AKAP95-p105. Sci Sig-
nal. 2009; 2(75):ra28. https://doi.org/10.1126/scisignal.2000202 PMID: 19531803; PubMed Central
PMCID: PMCPMC2770900.
62. Ollivier V, Parry GC, Cobb RR, de Prost D, Mackman N. Elevated cyclic AMP inhibits NF-kappaB-medi-
ated transcription in human monocytic cells and endothelial cells. J Biol Chem. 1996; 271(34):20828–
35. PMID: 8702838.
63. Parsons CG, Danysz W, Quack G. Memantine is a clinically well tolerated N-methyl-D-aspartate
(NMDA) receptor antagonist—a review of preclinical data. Neuropharmacology. 1999; 38(6):735–67.
PMID: 10465680.
64. Geerts H, Guillaumat PO, Grantham C, Bode W, Anciaux K, Sachak S. Brain levels and acetylcholines-
terase inhibition with galantamine and donepezil in rats, mice, and rabbits. Brain Res. 2005; 1033
(2):186–93. https://doi.org/10.1016/j.brainres.2004.11.042 PMID: 15694923.
PLOS ONE | https://doi.org/10.1371/journal.pone.0180319 July 26, 2017
30 / 30