Discovery of Axelopran (TD-1211): A Peripherally Restricted μ-Opioid Receptor Antagonist

Article abstract of DOI:10.1021/acsmedchemlett.9b00406

The effects of opioids in the central nervous system (CNS) provide significant benefit in the treatment of pain but can also lead to physical dependence and addiction, which has contributed to a growing opioid epidemic in the United States. Gastrointestinal dysfunction is an additional serious consequence of opioid use, and this can be treated with a localized drug distribution of a non-CNS penetrant, peripherally restricted opioid receptor antagonist. Herein, we describe the application of Theravance's multivalent approach to drug discovery coupled with a physicochemical property design strategy by which the N-substituted-endo-3-(8-aza-bicyclo[3.2.1]oct-3-yl)-phenyl carboxamide series of μ-opioid receptor antagonists was optimized to afford the orally absorbed, non-CNS penetrant, Phase 3 ready clinical compound axelopran (TD-1211) 19i as a potential treatment for opioid-induced constipation.

Full text of DOI:10.1021/acsmedchemlett.9b00406

Letter
pubs.acs.org/acsmedchemlett
Cite This: ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX
Discovery of Axelopran (TD-1211): A Peripherally Restricted
μ‑Opioid Receptor Antagonist
Daniel D. Long,* Scott R. Armstrong, David T. Beattie, Christina B. Campbell, Timothy J. Church,
Pierre-Jean Colson, Sean M. Dalziel, John R. Jacobsen, Lan Jiang, Glenmar P. Obedencio,
Miroslav Rapta, Daisuke Saito, Ioanna Stergiades, Pamela R. Tsuruda, Priscilla M. Van Dyke,
and Ross G. Vickery
Theravance Biopharma US, Inc., 901 Gateway Blvd., South San Francisco, California 94080, United States
S
* Supporting Information
ABSTRACT: The effects of opioids in the central nervous
system (CNS) provide significant benefit in the treatment of
pain but can also lead to physical dependence and addiction,
which has contributed to a growing opioid epidemic in the
United States. Gastrointestinal dysfunction is an additional
serious consequence of opioid use, and this can be treated
with a localized drug distribution of a non-CNS penetrant,
peripherally restricted opioid receptor antagonist. Herein, we
describe the application of Theravance’s multivalent approach
to drug discovery coupled with a physicochemical property
design strategy by which the N-substituted-endo-3-(8-aza-bicyclo[3.2.1]oct-3-yl)-phenyl carboxamide series of μ-opioid receptor
antagonists was optimized to afford the orally absorbed, non-CNS penetrant, Phase 3 ready clinical compound axelopran (TD-
1211) 19i as a potential treatment for opioid-induced constipation.
KEYWORDS: μ-Opioid receptor antagonists, multivalent approach, peripherally restricted, opioid-induced constipation
(Movantik) (3),10 approved in 2014 as an oral formulation,
nalgesic opioids such as morphine remain a mainstay of
Atherapeutic intervention to treat chronic malignant and and most recently naldemedine (Symproic) (4)11 approved in
2017 as an oral agent, are all indicated for the treatment of
opioid-induced constipation (OIC) for adult patients with
chronic noncancer pain. In addition, the non-opioid ligand
Amitiza (lubiprostone), a Cl-channel activator, was approved
in 2013 for the treatment of constipation caused by opioids in
patients with chronic, noncancer pain.12
We previously reported the discovery of the N-substituted-
endo-3-(8-aza-bicyclo[3.2.1]oct-3-yl)-phenyl carboxamide ser-
ies A of μ-opioid receptor antagonists, Figure 2.13 Herein, we
describe the design strategy by which this series was optimized
to afford the clinical compound axelopran (TD-1211) 19i,
Figure 2, which balances predictable and moderate oral
bioavailability with a high degree of peripheral selectivity.
The pharmacological profile and optimized synthesis of 19i are
additionally described.
nonmalignant pain as a result of their agonism of the G-protein
coupled μ-opioid receptor, principally within the central
nervous system (CNS).1,2 However, despite their effectiveness,
opioids are associated with a number of centrally mediated side
effects, including severe physical dependence and addiction,
which in recent years has been highlighted as the cause of a
growing opioid epidemic in the United States (US).3 An
additional serious consequence of both therapeutic and illicit
opioid use is the induction of a predominantly peripherally
mediated gastrointestinal (GI) dysmotility, termed opioid
bowel dysfunction (OBD).4,5 The most attractive approach to
combat the often debilitating effects of OBD in patients is by
employing a localized drug distribution of a non-CNS
penetrant, peripherally restricted opioid receptor antagonist
which has no impact on beneficial agonist analgesia but which
can selectively alleviate the GI dysfunction.6 Four peripherally
selective opioid receptor antagonists have been approved by
the FDA for the treatment of opioid-related bowel dysfunction,
Figure 1.
At the time of this work, it was known that oral dosing of
both alvimopan 1 and methyl naltrexone (MNTX) 2 resulted
in minimal CNS exposure, likely a result of the zwitterionic
character14,15 and charged quaternary amine,16 respectively,
which presumably significantly reduce passive cellular perme-
ability. However, this feature was additionally reflected in the
Alvimopan (Entereg) (1) was approved in the US in 2008 as
an oral treatment to accelerate upper and lower gastrointestinal
recovery following partial large or small bowel resection
surgery with primary anastomosis in a hospital setting.7 Methyl
naltrexone (Relistor) (2), approved for subcutaneous dosing
(2008)8 and as an oral formulation (2016),9 naloxegol
Received: September 4, 2019
Accepted: November 6, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acsmedchemlett.9b00406
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
(t1/2 = 6 min), Table 1.25 In contrast, the hydrophilic,
zwitterionic acid compound 7 exhibited weak binding affinity
(pKi = 6.2), an antagonist profile (IA = 4%), and good RLM
stability (t1/2 > 90 min). Combination of these features is
achieved with ADL 08-0011 5, which is both a high affinity μ-
opioid receptor antagonist (pKi = 9.6, IA = −8%) and a high
RLM stability compound (t1/2 > 90 min). A similar result was
obtained with the identical N-substituents applied to our
previously reported endo-3-(8-aza-bicyclo[3.2.1]oct-3-yl)-phe-
nol series of μ-opioid receptor antagonists13 and led us to
conclude that a two-component N-substituent consisting of a
lipophilic and hydrophilic moiety could be tolerated, which
attractively would allow us to effectively modulate the overall
physicochemical properties. Many chemical approaches were
attempted to leverage this insight, but most success was
obtained with the general design of an N-alkyl linker to a
branching N′-amine substituted with a lipophilic component
(R1) and a polar, hydrophilic group (R2), Figure 2. A modular
synthesis was designed to explore this concept, in which a set
of aldehydes 12a−f, 13−15, and 16a and b were prepared,
allowing for variation of the alkyl linker length (n) and the
lipophilic group (R1), Scheme 1.
Figure 1. Peripherally restricted opioid receptor antagonists.
Figure 2. Theravance μ-opioid receptor antagonist core, clinical
compound, and general pharmacophore.
poor oral bioavailability of both of these compounds,
alvimopan 1 %F: 617 and methyl naltrexone 2 %F: < 1.18
Alternate clinical precedent for an orally absorbed, non-CNS
penetrant profile is provided by the second generation of
nonsedating antihistamines such as fexofenadine (Allegra) as
well as the antidiarrheal μ-opioid receptor agonist loperamide
(Imodium). In both cases, the lack of blood−brain barrier
penetration of the compound was determined empirically and
was latterly attributed to their being a substrate for the P-
glycoprotein efflux transporter (P-gp).19,20 Our approach
toward a predictable and balanced orally absorbed, non-CNS
penetrant profile considered the overall physicochemical
properties of the compounds. Many analyses of the optimal
physicochemical properties required for oral absorption have
been reported since Lipinski’s seminal report,21 but fewer have
focused on the subset of CNS active compounds, although this
has recently been addressed.22 In 2006, Hitchcock et al.
analyzed the top 25 CNS drugs and proposed a set of
guidelines for CNS penetration, focused on values for total
polar surface area (tPSA) (<70A2), number of hydrogen bond
donors (HBD) (0−2), lipophilicity clogDpH 7.4 (2−4), and
molecular weight (MW) (<450 Da).23 It appears logical that
inverting these guidelines within a set of limits related to oral
absorption might provide compounds with the desired orally
absorbed, non-CNS penetrant profile. As such, we proposed a
target range of properties, as indicated in Table 1.
Variation of the N-substituent (R) of the novel μ-opioid
receptor antagonist series A, Figure 2, was considered the most
appropriate position to explore and optimize the overall
physicochemical properties in tandem with potency, consistent
with the application of Theravance’s multivalent approach to
drug discovery.13 The core A is considered the primary binding
group and the N-substituent a handle by which secondary
binding interactions and physicochemical properties can be
modulated. Some early work on the nature of the N-
substituent of the reported active metabolite (ADL 08-0011,
5) of alvimopan 1 was instructive, Figure 3.24
Reductive amination of aldehyde 12a with the previously
described13 endo-3-(8-aza-bicyclo[3.2.1]oct-3-yl)-phenyl car-
boxamide series core 17 and subsequent Boc-deprotection
afforded the representative intermediate 18 with an ethyl alkyl
linker and N-cyclohexylmethyl as the lipophilic group (R1),
from which a set of compounds 19a−l exploring the R2 group
were prepared, Scheme 2.
All of the compounds 18 and 19a−l were high affinity μ-
opioid receptor antagonists (pKi = 9.1−10, IA = −3 to 21%),
Table 1. Although the acetamide 19a had reasonable human
liver microsomes (HLM) stability (t1/2 = 63 min), the tert-
butyl analogue 19b and trifluoromethyl compound 19c
exhibited short half-lives in RLM, dog liver microsomes
(DLM), and HLM (t1/2 ≤ 12 min). These amides 19a−c did
not have target properties in our desired range with a tPSA of
only 67 Å2 and 1 HBD. Poor LM stability was also a feature of
the carbamate 19d, urea 19e, sulfonamide 19f, and sulfamate
19g compounds (t1/2 ≤ 20 min across all species). In this
instance, the tPSA of the compounds was increased (76−87
Å2), and specifically for the urea 19e, 2 HBD were introduced.
Revisiting the amide series, introduction of a more polar side
chain with compounds 19h−j resulted in an increase in HLM
stability (t1/2 ≥ 79 min) and tPSA (87−107 Å2) while
maintaining an appropriate clogDpH 7.4 (0.5−0.9) as well as
displaying a moderate to good Caco-2 permeability (Kp 5−40
× 10−6 cm/s). Low DLM stability (t1/2 ≤ 28 min) appeared to
be a species outlier for these otherwise attractive compounds.
The heterocyclic tetrazole amide 19k showed high metabolic
stability across all species with LM t1/2 > 90 min but was of
very low permeability (Kp < 1 × 10−6 cm/s). The furyl amide
19l was a low LM stability compound (t1/2 ≤ 17 min across all
species).
Due to the attractive features of the R2 = HOCH2C(O)-
compound 19h, analogues with varying lipophilic R1 groups
were next examined. Following an analogous synthetic
sequence to Scheme 2, reductive amination of aldehydes
12b−f with core 17 and Boc-deprotection afforded com-
pounds which were reacted with acetoxyacetyl chloride to
provide intermediates 20b−f. The R1 = benzyl and 4,4-diF-c-
hexylmethyl analogues 20a and 20g were prepared directly
from aldehydes 16a and 16b, respectively. Hydrolysis of the
Comparing ADL 08-0011 5 to its constituent parts as the
(CH2)3Ph compound 6 and propionic acid compound 7
highlighted that the lipophilic compound 6 had high binding
affinity (pKi = 10.5), a low intrinsic activity (IA) consistent
with an antagonist effect (IA = −21% of the maximum
response achieved by DAMGO in a GTP binding assay) at the
human recombinant μ-opioid receptor, and poor metabolic
stability in rat liver microsomes (RLM) with a short half-life
B
DOI: 10.1021/acsmedchemlett.9b00406
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
C
DOI: 10.1021/acsmedchemlett.9b00406
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
a
Scheme 4. 'Preparation of Compound 29
Figure 3. Alvimopan metabolite and its constituent parts.
a
Reagents and conditions: (i) (a) (MeO)2CHCHO, NaBH(OAc)3,
DCM, (b) 6 N HCl (aq.), 33% over 2 steps; (ii) (a) 4-aminomethyl-
tetrahydropyran, NaBH(OAc)3, DCM, 19%, (b) AcOCH2COCl,
DIPEA, DCM; (c) LiOH·H2O, EtOH, H2O, 15% over 2 steps.
Scheme 1. Preparation of Aldehyde Intermediates 12a−f,
a
13−15, 16a−b
a
Scheme 5. Preparation of Compounds 32−33
a
Reagents and conditions: (i) For n = 1, R1 = cyclohexylmethyl-,
a
Reagents and conditions: (i) CbzNHCH2 CHO or
phenethyl-, phenpropyl-, cyclobutylmethyl-, and n-hexyl-, and for n =
3, 4, and 5, R1 = 2-ethyl-nbutyl-, then R1Br, EtOH, 70-75°C, 38−94%;
for n = 1, R1 = cyclopentylmethyl-, 4,4-diF-c-hexylmethyl-, then
R1OMs, EtOH, 70−80°C, 64−77%; (ii) Boc2O, DCM, 0°C to RT,
43−100%; (iii) PySO3, DIPEA, DMSO, DCM, 0°C, 49−100%; (iv)
for R1 = Bn, 4,4-diF-c-hexylmethyl-: AcOCH2COCl, DIPEA, DCM,
59−61%.
CbzNHCH2CH2CHO, NaBH(OAc)3, DCM; (ii) (a) H2, Pd(OH)2,
EtOH, (b) 2-ethylbutanal, NaBH(OAc)3, DCM, 42−57% over 3
steps, (c) AcOCH2COCl, DIPEA, DCM; (d) 6N NaOH (aq.),
MeOH, 60−100% over 2 steps.
While all of the compounds 24a−g, 29, and 32 maintained
μ-opioid receptor antagonist profiles, there was a more
pronounced effect on the degree of μ-opioid receptor binding
affinity. Relative to the cyclohexylmethyl compound 19h (pKi
= 10, DLM t1/2 = 18 min), the benzyl 24a, 2-ethyl-nbutyl 32
and 4,4-diF-cyclohexylmethyl 24g analogues exhibited com-
parable μ-opioid receptor binding affinity (pKi = 10, 9.8, and
9.9, respectively) while showing an increase in DLM stability
(t1/2 = 90, 75, and 90 min, respectively) and maintenance of
the favorable target range of properties and Caco-2
permeability values. Extension of the lipophilic phenyl ring of
benzyl compound 24a further from the linking N′-amine with
the phenethyl 24b and phenpropyl 24c analogues resulted in
the observation of a significant reduction in μ-opioid receptor
binding affinity (pKi = 9.2 and 9.1, respectively), and this was
replicated with the straight chain n-hexyl compound 24f (pKi =
9.1). Contraction of the cyclohexyl ring of compound 19h (pKi
= 10, tPSA = 87 Å2) to the cyclopentyl 24d and cyclobutyl 24e
compounds resulted in a progressive reduction of μ-opioid
receptor binding affinity (pKi = 9.6 and 8.8, respectively), while
increasing polarity with the tetrahydropyran analogue 29
(tPSA = 96 Å2) reduced both binding affinity (pKi = 7.9) and
Caco-2 permeability (Kp < 1 × 10−6 cm/s). A further element
of the SAR that was probed was the length of the alkyl linker,
which was done using the constant of R1 = 2-ethyl-nbutyl and
R2 = HOCH2C(O)-. The synthesis detailed in Scheme 3 was
employed for n = 3−5 using aldehydes 13−15 to afford n-butyl
25, n-pentyl 26, and n-hexyl 27 linked compounds,
respectively. An alternate synthetic sequence was used to
prepare the ethyl 32 and n-propyl compounds 33, Scheme 5.
Similar μ-opioid receptor binding affinity was observed for the
ethyl 32 and n-propyl 33 linked compounds (pKi = 9.8 and 10,
respectively; clogDpH 7.4 = 0.6 and 0.7 respectively) and this
was reduced as lipophilicity increased with the n-butyl 25, n-
pentyl 26 and n-hexyl 27 compounds (pKi = 9.2, 9.2 and 9,
respectively; clogDpH 7.4 = 0.8, 1.1, 1.4 respectively).
a
Scheme 2. Preparation of Compounds 18 and 19a−l
a
Reagents and conditions: (i) (a) 12a, NaBH(OAc)3, DCM, (b)
t
TFA, DCM, 57% over 2 steps; (ii) (a) for R2 = Ac-, BuC(O)-,
CF3C(O)-, MeOC(O)-, MeO2S-, Me2NO2S-, AcOCH2C(O)-, 2-
furylC(O)-, then R2Cl, DIPEA, DCM; iPrNCO, DIPEA, DMF; for R2
= 2,2-dimethyl-1,3-dioxolane-4S−C(O)-, MeO2SCH2C(O)-, 1H-
tetrazole-5-CH2(CO)-, then R2OH, HATU, DIPEA, DMF (b) for
R2 = AcOCH2CO-: LiOH·H2O, MeOH; for R2 = 2,2-dimethyl-1,3-
dioxolane-4S−C(O)-: AcOH, H2O, 70°C.
acetyl group of compounds 20a−g yielded ethyl (n = 1) linked
compounds 24a−g, Scheme 3.
An alternate synthesis in which the key linker, R1 and R2
groups were introduced in an alternate sequential order was
utilized to prepare the ethyl linked, R1 = methyl-tetrahy-
dropyran, R2 = HOCH2C(O)-, compound 29, Scheme 4.
Additionally, the R1 = 2-ethyl-nbutyl compound 32 was
prepared according to Scheme 5.
a
Scheme 3. Preparation of Compounds 24a−g, 25−27
a
Reagents and conditions: (i) (a) 12b−f, 13, 14, 15, 16a and b,
NaBH(OAc)3, DCM, (b) for intermediates from 12b−f, 13, 14, 15
only: TFA, DCM, 37−78% over 2 steps, then AcOCH2COCl, DIPEA,
DCM; (ii) LiOH·H2O, EtOH, H2O; from 20a, g, 49% over 2 steps;
from 20b−f, 21, 22, 23, 47−100% over 2 steps.
It should be noted that for all the analogues 18, 19a−l, 24a−
g, 29, 32, 25, 26, 27, and 33, the binding profile at both the
human δ- and guinea pig κ-opioid receptors was also
D
DOI: 10.1021/acsmedchemlett.9b00406
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Table 2. In Vivo Pharmacokinetic and Pharmacodynamic Screening Data
a
b
c
d
e
Cmax
AUC(0−t)
t1/2
F
mouse IT ID50 mouse GE ID50 mouse HP ID50 peripheral selectivity peripheral selectivity
f g
Compound
(μg/mL)
(μg h/mL)
(h)
(%)
(mg/kg)
(mg/kg)
(mg/kg)
0.82
index (IT)
1.4
index (GE)
1.6
naltrexone
MNTX 2
alvimopan 1
ADL
08-0011 5
0.6
67
0.5
25
0.57
0.01
0.71
0.49
0.002
1.20
2.8
3.7
2.6
0.1
53
2.5
0.6
>20
1.1
>8
17
>33
41
0.064
0.027
19h
19i
19j
24g
0.29
0.21
0.88
0.39
0.43
0.47
1.30
1.20
1.3
1.3
1.0
2.0
61
27
60
30
0.10
0.35
0.49
0.23
0.046
0.087
0.33
4.2
14
24
42
40
49
19
91
161
73
0.03
4.4
147
a
Pharmacokinetic properties were evaluated in male Sprague−Dawley rats dosed with test compounds via oral gavage (PO) at a dose of 5 mg/kg, n
b
c
= 3, except for MNTX, which was dosed at 100 mg/kg, and alvimopan, which was dosed at 10 mg/kg. F = oral bioavailability. IT = inhibition of
morphine-induced delays in intestinal transit. GE = inhibition of morphine-induced delays in gastric emptying. HP = inhibition of morphine-
induced hot plate antinociception. Peripheral selectivity index = ratio of hot plate and intestinal transit ID50 values. Peripheral selectivity index =
ratio of gastric emptying and intestinal transit ID50 values.
d
e
f
g
determined, and the selectivity for the three opioid receptor
subtypes was in general consistent. The binding affinity for the
δ-opioid receptor was reduced by 0.9−2 log, and the binding
affinity for the κ-opioid receptor spanned a range of between
0.6 log lower to 0.4 log higher, relative to the affinity at the μ-
opioid receptor.
A representative set of compounds 19h, 19i, 19j, and 24g
with favorable in vitro properties were advanced to subsequent
in vivo pharmacokinetic and pharmacodynamic studies, Table
2. All of the compounds exhibited moderate to good oral
bioavailability in rats (%F: 27−61) with significantly improved
dose normalized Cmax and AUC relative to both alvimopan 1
and MNTX 2.
Pharmacodynamically, following oral administration, 19h,
19i, 19j, and 24g produced a dose-dependent reversal of the
morphine-induced inhibition of intestinal transit (ID50 values
of 0.1, 0.35, 0.49, and 0.23 mg/kg, respectively) and gastric
emptying (ID50 values of 0.046, 0.087, 0.33, and 0.03 mg/kg,
respectively) in mice. Both of these assays were viewed as a
measure of the peripheral GI activity (non-CNS) of the test
compounds. The opioid receptor antagonists naltrexone,
alvimopan, and ADL 08-0011 had a similar range of potencies
in the intestinal transit (ID50 values of 0.6, 2.5, and 0.064 mg/
kg, respectively) and gastric emptying assays (ID50 values of
0.5, 0.6, and 0.027 mg/kg, respectively), while MNTX was
significantly less potent in the intestinal transit (ID50 values 67
mg/kg) and gastric emptying assay (ID50 values of 25 mg/kg),
Table 2.
With respect to the CNS activity of the test compounds
following oral administration, 19h, 19i, 19j, and 24g produced
a dose-dependent reversal of morphine-induced antinocicep-
tion in mice (ID50 values of 4.2, 14, 24, and 4.4 mg/kg,
respectively; Table 2). By comparison, the opioid receptor
antagonists naltrexone, alvimopan, and ADL 08-0011 had ID50
values of 0.82, >20, and 1.1 mg/kg, respectively).
transit), and 1.6, >33, and 41, respectively (gastric emptying).
The data indicated that compounds 19h, 19i, 19j, and 24g all
had a high peripheral selectivity in mice following oral dosing,
clearly exceeding that of the centrally penetrant opioid
receptor antagonist naltrexone as well as being improved
relative to the values for ADL 08-0011, the active metabolite of
alvimopan, which at potentially efficacious concentrations in
humans was shown to have minimal CNS effects.26
Compounds 19i and 24g appeared to have the most favorable
peripheral selectivity indices and each satisfied all the
parameters of our target range of properties; in particular,
19i had the highest tPSA (107 Å2) and number of HBD (3).
Compound 19i was selected for further characterization.27,28
Consistent with its high opioid receptor binding affinity, 19i
had potent opioid receptor antagonist activity of agonist-
induced GTPγS binding in recombinant expression systems
(pKb values = 9.6, 8.8, and 9.5, at μ, δ, and κ, respectively) and
inhibited agonist-induced inhibition of electric field stimulated
contractions of rodent tissue preparations [μ and κ pA2 values
= 10.1 and 8.8, respectively (guinea pig ileum), and δ pKb
value = 8.4 (hamster vas deferens)]. The selectivity of 19i for
opioid receptors was examined; at 1 μM (i.e., a concentration
6300-fold higher than its μ Ki value), 19i had no significant
effect on radioligand binding or functional activity at all 80
receptors, ion channels, and enzymes tested.
Oral pharmacokinetics of 19i in fasted female and male
beagle dogs (%F: 29 and 45, respectively) were consistent with
rat. Additionally, 19i exhibited low plasma protein binding
(23−30%) across species, including human. Glucuronidation
was the major metabolic pathway in rodents while CYP3A4-
mediated hydroxylation was predominant in dog and human
based on in vitro studies. Efflux ratios for 19i of 10 and 24 in a
Caco-2 and MDR1-MDCK cell line, respectively, indicated the
compound is a P-gp substrate which was supported by studies
in MDR1a/1b knockout mice which upon oral dosing of 19i
had increased CNS drug concentrations relative to wild-type
mice. In rats, csf, brain, and plasma concentrations after oral
administration (20 mg/kg) of 19i revealed CNS penetration to
be <2%.
The peripheral selectivity of compounds was defined as the
ratio of the CNS-mediated antinociception ID50 value and the
peripheral intestinal transit or gastric emptying ID50 values.
Compounds 19h, 19i, 19j, and 24g had peripheral selectivities
of 42, 40, 49, and 19, respectively, for potencies in the
antinociception and intestinal transit assays, and 91, 161, 73,
and 147, respectively, upon comparison of potencies in the
antinociception and gastric emptying assays, Table 2. The
corresponding peripheral selectivities of naltrexone, alvimopan,
and ADL 08-0011 were 1.4, >8, and 17, respectively (intestinal
Following oral dosing to rats, 19i reversed loperamide-
induced delays in gastric emptying (ID50 = 0.24 mg/kg) and
castor oil-induced diarrhea (ID50 = 0.01 mg/kg). In dogs, 19i
inhibited nonproductive GI circular smooth muscle contrac-
tility (effective at 3 mg/kg). Orally dosed 19i failed to evoke a
CNS-mediated withdrawal response in morphine-dependent
E
DOI: 10.1021/acsmedchemlett.9b00406
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
a
mice (30 mg/kg), inhibit morphine-induced antinociception in
rat and dog hot plate tests (>60 mg/kg and 3 mg/kg,
respectively), or reverse morphine-induced sedation in dogs
(at 3 mg/kg). It was concluded that 19i has potent in vivo
gastrointestinal activity in mice, rats, and dogs, consistent with
opioid receptor antagonism, and possesses a high degree of
peripheral selectivity in these species.
Subsequently, compound 19i was nominated as a develop-
ment candidate, axelopran (TD-1211), and advanced to
human clinical trials. Accordingly, an alternate, optimized
synthesis of axelopran (TD-1211) 19i was devised to support
IND-enabling studies and early clinical manufacture. The
previously described13 endo-3-(8-aza-bicyclo[3.2.1]oct-3-yl)-
phenyl carboxamide series core 17 was prepared efficiently
from N-benzyl-nortropinone 34. Suzuki reaction of the vinyl
triflate 35 yielded styrene 36. Subsequent styrene reduction
and concomitant benzyl deprotection under acidic conditions
afforded the key intermediate core as a mixture of endo:exo
isomers in a ratio of 93:7, which upon crystallization afforded
the single endo isomer 17 in >99% purity as its HCl salt,
Scheme 6.
Scheme 7. 'Preparation of Aldehyde Intermediate 40
a
Reagents and conditions: (i) (a) H2NCH2CH(OEt)2, H2, Raney-Ni,
Me-THF, 0°C, (b) CbzCl, DIPEA, Me-THF, and (ii) (a) HCl,
AcCN, 30°C, (b) NaHSO3, EtOAc, 75% over 4 steps; (iii) 1M NaOH
(aq.), Me-THF.
a
Scheme 8. Preparation of Axelopran (TD-1211) 19i
a
Reagents and conditions: (i) (a) Me-THF solution of 40,
NaBH(OAc)3, DMF, (b) H2, 10% Pd/C, MeOH, 6M HCl, 65%
over 2 steps; (ii) lithium-(4S)-2,2-dimethyl-1,3-dioxolane-4-carbox-
ylate, PyBOP, DMF, and (iii) 2M H2SO4 (aq.), then crystallization
from 10% H2O in MeOH, 73% over 2 steps.
The aldehyde 40, generated in situ from the bench-stable
sodium bisulfite aldehyde adduct 39 (Scheme 7) was coupled
with the intermediate core 17 via reductive amination, Scheme
8. Deprotection of the Cbz group to the secondary amine 18
and subsequent amide coupling with the lithium salt of the
protected diol chiral acid afforded the acetal protected
compound 41. Removal of the acetal group and crystallization
afforded the monosulfate salt of axelopran (TD-1211) 19i.
A crystal structure of the monosulfate salt of axelopran (TD-
1211) 19i was obtained to unambiguously define its chemical
structure, highlighting the phenyl-carboxamide endocyclic and
hydrophilic group 2S stereochemistry as well as the phenyl
axial conformation, Figure 4.29
Axelopran (TD-1211) 19i has successfully completed Phase
1a and 1b and three Phase 2 studies which were carefully
designed to identify the best dose and regimen for Phase 3
studies.30−32 The key efficacy highlight from the most recent
200 OIC patient Phase 2b study revealed a statistically
significant (p = 0.0001) increase in complete spontaneous
bowel movements (CSBMs) relative to placebo during week 5
of the 5 week treatment with patients transitioning from a
baseline of 0.2 CSBMs to almost 3 at the top dose of 15 mg qd.
Axelopran (TD-1211) 19i has been well-tolerated across all
studies with observation of similar treatment emergent adverse
events relative to placebo. Importantly, in the relevant studies
there has been no impact on the use of opioid analgesics at
efficacious doses, confirming the peripheral selectivity of the
drug. Axelopran (TD-1211) 19i is positioned as Phase 3 ready
for the treatment of OIC.
Figure 4. X-ray crystal structure of the monosulfate salt of axelopran
(TD-1211) 19i.
In summary, Theravance’s multivalent approach to drug
discovery in combination with a physicochemical property
driven design strategy was applied to optimize the N-
substituted-endo-3-(8-aza-bicyclo[3.2.1]oct-3-yl)-phenyl car-
boxamide series A of μ-opioid receptor antagonists to afford
an orally absorbed, non-CNS penetrant profile that resulted in
the clinical compound axelopran (TD-1211) 19i, targeted for
the treatment of OIC. Extensive preclinical characterization of
the pharmacological and DMPK properties of axelopran (TD-
1211) 19i are highlighted along with an optimized synthesis
applied to its clinical manufacturing. Axelopran (TD-1211) 19i
has successfully completed Phase 2 studies in patients with
OIC and is Phase 3 ready.
a
Scheme 6. Preparation of Compound 17
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acsmedchemlett.9b00406.
a
Reagents and conditions: (i) PhNTf2, NaHMDS, THF and (ii) 3-
carbamoylphenyl boronic acid, Pd(OAc)2, dppf, KF, THF, reflux, 78%
over 2 steps; (iii) H2, 10% Pd/C, EtOH, 6M HCl, H2O, 50°C,
endo:exo 93:7, then crystallization from EtOH, 75% and >99% single
endo isomer.
Select assay protocols and representative synthetic
procedures (PDF)
F
DOI: 10.1021/acsmedchemlett.9b00406
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
(15) A zwitterionic strategy was also employed by Pfizer towards a
peripherally-restricted δ-opioid receptor agonist: Middleton et al.,
Highly potent and selective zwitterionic agonists of the δ-opioid
receptor. Part 1. Bioorg. Med. Chem. Lett. 2006, 16, 905−10.
(16) Brown, D. R.; Goldberg, L. I. The use of quaternary narcotic
antagonists in opiate research. Neuropharmacology 1985, 24, 181−91.
(17) Erowele, G. I. Alvimopan (Entereg), a Peripherally Acting mu-
Opioid Receptor Antagonist For Postoperative Ileus. P&T. 2008, 33
(574), 580−3.
AUTHOR INFORMATION
Corresponding Author
*E-mail: dlong@theravance.com.
ORCID
Daniel D. Long: 0000-0002-0742-6915
Notes
The authors declare no competing financial interest.
■
(18) Yuan, C.-S.; Wei, G.; Foss, J. F.; O’Connor, M.; Karrison, T.;
Osinski, J. Effects of Subcutaneous Methylnaltrexone on Morphine
Induced Peripherally Mediated Side Effects: A Double-Blind
Randomized Placebo-Controlled Trial. J. Pharmacol. Exp. Ther.
2002, 300, 118−23.
ACKNOWLEDGMENTS
■
The authors would like to thank Dr. David L. Bourdet, Dr. R.
Murray McKinnell, and Dr. Kris Josephson from Theravance
Biopharma US, Inc. for their assistance in reviewing this article.
(19) Sadeque, A. J.; Wandel, C.; He, H.; Shah, S.; Wood, A. J.
Increased drug delivery to the brain by P-glycoprotein inhibition. Clin.
Pharmacol. Ther. 2000, 68, 231−7.
ABBREVIATIONS
max, maximal concentration; AUC, area under the curve; ID50,
■
C
(20) Tahara, H.; Kusuhara, H.; Fuse, E.; Sugiyama, Y. P-glycoprotein
plays a major role in the efflux of fexofenadine in the small intestine
and blood-brain barrier, but only a limited role in its biliary excretion.
Drug Metab. Dispos. 2005, 33, 963−8.
50% inhibitory dose; GTPγS, guanosine 5′-O-[gamma-thio]-
triphosphate; MDCK, Madin−Darby canine kidney.
(21) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.
Experimental and computational approaches to estimate solubility and
permeability in drug discovery and development settings. Adv. Drug
Delivery Rev. 1997, 23, 3−25.
(22) Travis, T.; Wager, T. T.; Hou, X.; Verhoest, P. R.; Villalobos, A.
Central Nervous System Multiparameter Optimization Desirability:
Application in Drug Discovery. ACS Chem. Neurosci. 2016, 7, 767−
775.
REFERENCES
■
(1) Walsh, T. D. Pharmacological management of cancer pain.
Seminars in Oncology 2000, 27, 45−63.
(2) Chou, R.; Fanciullo, G. J.; Fine, P. G.; Adler, J. A.; Ballantyne, J.
C.; Davies, P.; Donovan, M. I.; Fishbain, D. A.; Foley, K. M.; Fudin,
J.; Gilson, A. M.; Kelter, A.; Mauskop, A.; O’Connor, P. G.; Passik, S.
D.; Pasternak, G. W.; Portenoy, R. K.; Rich, B. A.; Roberts, R. G.;
Todd, K. H.; Miaskowski, C. J. Clinical Guidelines for the Use of
Chronic Opioid Therapy in Chronic Noncancer Pain. J. Pain 2009,
10, 113−30.
(23) Hitchcock, S. A.; Pennington, L. D. Structure-Brain Exposure
Relationships. J. Med. Chem. 2006, 49, 7559−80.
(24) Beattie, D. T.; Cheruvu, M.; Mai, N.; O’Keefe, M.; Johnson-
Rabidoux, S.; Peterson, C.; Kaufman, E.; Vickery, R. The in vitro
pharmacology of the peripherally restricted opioid receptor
antagonists, alvimopan, ADL 08-0011 and methylnaltrexone.
Naunyn-Schmiedeberg's Arch. Pharmacol. 2007, 375, 205−20.
(25) Compounds are classified: −5 < IA < 25 as antagonists, 25 < IA
< 80 as partial agonists, 80 < IA < 120 as full agonists.
(26) Becker, G.; Blum, H. E. Novel opioid antagonists for opioid-
induced bowel dysfunction and postoperative ileus. Lancet 2009, 373,
1198−206.
(3) Skolnick, P. The Opioid Epidemic: Crisis and Solutions. Annu.
Rev. Pharmacol. Toxicol. 2018, 58, 143−59.
(4) Brock, C.; Olesen, S. S.; Olesen, A. E.; Frokjaer, J. B.; Andresen,
T.; Drewes, A. M. Opioid-induced bowel dysfunction: pathophysiol-
ogy and management. Drugs 2012, 72, 1847−65.
(5) Walters, J. B.; Montagnini, M. Current concepts in the
management of opioid-induced constipation. J. Opioid Management
2010, 66, 435−44.
(6) Diego, L.; Atayee, R.; Helmons, P.; Hsiao, G.; von Gunten, C. F.
Novel opioid antagonists for opioid-induced bowel dysfunction.
Expert Opin. Invest. Drugs 2011, 20, 1047−56.
(27) Tsuruda, P. R.; Vickery, R. G.; Long, D. D.; Armstrong Beattie,
D. T. The in vitro pharmacological profile of TD-1211, a neutral
opioid receptor antagonist. Naun. Schmi. Arch. Pharmacol. 2013, 386,
479−491.
(28) Armstrong, S. R.; Campbell, C. B.; Richardson, C. L.; Vickery,
R. G.; Tsuruda, P. R.; Long, D. D.; Hegde, S. S.; Beattie, D. T. The in
vivo pharmacodynamics of the novelopioid receptor antagonist, TD-
1211, in models of opioid-induced gastrointestinal and CNS activity.
Naunyn-Schmiedeberg's Arch. Pharmacol. 2013, 386, 471−8.
(29) CCDC 1948348 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.
uk/structures.
(30) Vickery, R. Pharmacokinetics and constipation relieving effects
of TD-1211 in patients with opioid-induced constipation. J. Pain
2012, 12, P333.
(31) Vickery, R. The safety and pharmacokinetics of TD-1211 in
healthy volunteers: single and multiple ascending oral dose
administration. J. Pain 2012, 12, P334.
(7) Beattie, D. T. Safety and efficacy update: alvimopan in
postoperative ileus. Clin. Med.: Ther. 2009, 1, 199−213.
(8) Rauck, R. L. Treatment of opioid-induced constipation: focus on
the peripheral μ-opioid receptor antagonist methylnaltrexone. Drugs
2013, 73, 1297−306.
(9) Webster, L. R.; Israel, R. J. Oral methylnaltrexone does not
negatively impact analgesia in patients with opioid-induced
constipation and chronic noncancer pain. J. Pain Res. 2018, 11,
1503−10.
(10) Garnock-Jones, K. P. Naloxegol: a review of its use in patients
with opioid-induced constipation. Drugs 2015, 75, 419−25.
(11) Markham, A. Naldemedine: First Global Approval. Drugs 2017,
77, 923−7.
(12) Jamal, M. M.; Adams, A. B.; Jansen, J.-P.; Webster, L. R. A
randomized, placebo-controlled trial of lubiprostone for opioid-
induced constipation in chronic noncancer pain. Am. J. Gastroenterol.
2015, 110, 725.
(13) Jiang, L.; Beattie, D. T.; Jacobsen, J. R.; Kintz, S.; Obedencio,
G. P.; Saito, D.; Stergiades, I.; Vickery, R. G.; Long, D. D. Discovery
of N-substituted-endo-3-(8-aza-bicyclo[3.2.1]oct-3-yl)-phenol and
-phenyl carboxamide series of μ-opioid receptor antagonists. Bioorg.
Med. Chem. Lett. 2017, 27, 2926−30.
(32) Vickery, R.; Schwertschlag, U.; Singla, N.; Webster, L.; Canafax,
D. TD-1211 demonstrates a durable increase in bowel movement
frequency and return toward normal bowel function in a 5-week Ph2b
Opioid-Induced Constipation (OIC) study. J. Pain 2013, 14, S78.
(14) Zimmerman, D. M.; Gidda, J. S.; Cantrell, B. E.; Schoepp, D.
D.; Johnson, B. G.; Leander, J. D. Discovery of a potent, peripherally
selective trans-3,4-dimethyl-4-(3-hydroxyphenyl)piperidine opioid
antagonist for the treatment of gastrointestinal motility disorders. J.
Med. Chem. 1994, 37, 2262−5.
G
DOI: 10.1021/acsmedchemlett.9b00406
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX