Navigating CYP1A Induction and Arylhydrocarbon Receptor Agonism in Drug Discovery. A Case History with S1P1 Agonists

Article abstract of DOI:10.1021/acs.jmedchem.5b01102

This article describes the finding of substantial upregulation of mRNA and enzymes of the cytochrome P450 1A family during a lead optimization campaign for small molecule S1P₁ agonists. Fold changes in mRNA up to 10 000-fold for CYP1A1 in vivo

Full text of DOI:10.1021/acs.jmedchem.5b01102

Article
pubs.acs.org/jmc
Navigating CYP1A Induction and Arylhydrocarbon Receptor
Agonism in Drug Discovery. A Case History with S1P1 Agonists
Simon J. Taylor,*,† Emmanuel H. Demont,† James Gray,† Nigel Deeks,† Aarti Patel,‡ Dung Nguyen,§
Maxine Taylor,‡ Steve Hood,‡ Robert J. Watson,† Rino A. Bit,† Fiona McClure,∥ Holly Ashall,∥
and Jason Witherington†
†Immuno-Inflammation Therapy Area Unit, GlaxoSmithKline, Gunnels Wood Road, Stevenage, SG1 2NY, U.K.
‡PTS DMPK, GlaxoSmithKline, Park Road, Ware, SG12 0DP, U.K.
§PTS DMPK, GlaxoSmithKline, Upper Merion, 709 Swedeland Road, King of Prussia, Pennsylvania 19406, United States
∥Safety Assessment, GlaxoSmithKline, Park Road, Ware, SG12 0DP, U.K.
S
* Supporting Information
ABSTRACT: This article describes the finding of substantial
upregulation of mRNA and enzymes of the cytochrome P450
1A family during a lead optimization campaign for small
molecule S1P1 agonists. Fold changes in mRNA up to 10 000-
fold for CYP1A1 in vivo in rat and cynomolgus monkey and
up to 45-fold for CYP1A1 and CYP1A2 in vitro in rat and
human hepatocytes were observed. Challenges observed with
correlating induction in vitro and induction in vivo resulted in
the implementation of a short, 4 day in vivo screening study in
the rat which successfully identified noninducers. Subtle
structure−activity relationships in this series of S1P1 agonists
are described extending beyond planarity and lipophilicity, and
the impact and considerations of AhR and CYP1A induction in the context of drug development are discussed.
CYP1A2 is expressed mainly in the liver across species and is
only weakly expressed in extrahepatic tissues. In human liver
CYP1A2 accounts for about 13% of total CYP content and
metabolizes about 4% of therapeutic drugs including caffeine,
theophylline, and phenacetin.12
INTRODUCTION
■
The aryl hydrocarbon receptor (AhR) is a ligand-dependent
transcription factor regulating the expression of numerous
genes and mediating a diverse range of biological and
toxicological responses following exposure to endogenous and
exogenous chemicals.1−3 These genes include those encoding
xenobiotic metabolizing enzymes such as the cytochrome P450
(CYP) 1A family, and induction of this enzyme family is one of
the most sensitive AhR activation end points.4 The CYP1A
family consists of CYP1A1 and 1A2 in rats, mice, dogs,
monkey, and human with strong amino acid sequence
conservation across species.5
In rats and human, CYP1A1 is expressed only at very low
levels in liver and is present mostly as an extrahepatic enzyme
in tissues notably the small intestine and lung. CYP1A1 levels
are highly variable across human liver and small intestine
preparations where CYP1A1 is predominantly expressed after
induction.5 CYP1A1 can oxidize and metabolically activate
promutagenic or procarcinogenic substances.6−9 These include
polyaromatic hydrocarbons (PAHs) and aromatic amines and
amides, present as environmental contaminants and compo-
nents of cigarette smoke, charbroiled foods, and agricultural
chemicals, respectively. These chemicals are metabolized by
CYP1A and result in electrophillic reactive intermediates
capable of binding to cellular macromolecules, DNA, and
RNA, causing cytotoxic processes.10−14
Details on the mechanisms of AhR activation causing CYP1A
enzyme induction can be found in numerous articles.1,15−17
Differences exist in the relative sensitivities to induction of
CYP1A1 and 1A2 across species with humans reported as being
less responsive and sensitive to CYP1A1 activity and induction
than rats but having a more robust response to CYP1A2.4
In addition to enzyme regulation numerous other physio-
logical functions mediated via the AhR have emerged including
development and regulation of cell differentiation and cycling,
hormonal and nutritional homeostasis, coordination of cell
stress responses (including inflammation and apoptosis),
immune responses and aging and cancer promotion including
regulation of cell proliferation and apoptosis.18,19
The development of drugs that possess off-target agonism of
AhR therefore represents a challenge to the drug discovery
scientist. For example the induction of CYP1A may result in
raised levels of reactive intermediates and increased risk of
carcinogenicity following exposure to environmental contam-
Received: July 14, 2015
Published: September 22, 2015
© 2015 American Chemical Society
8236
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
inants. Additionally, upregulation of CYP1A is considered a
clinically relevant drug−drug interaction liability as has been
demonstrated via the impact of smoking on the increased
metabolism, and therefore reduced exposure to CYP1A2
substrates such as theophylline, caffeine, and clozapine.20,21
Other clinically relevant effects of CYP1A2 regulation include
the effect of rifampicin and carbamazepine to increase caffeine
and clozapine metabolism, respectively. The consumption of
cruciferous vegetables and charbroiled meat is also known to
increase CYP1A2 activity in human. Details on these
interactions are covered in a review by Zhou et al.11 These
drug−drug interactions can often be managed in the clinic
through dosage adjustment, therapeutic monitoring, and
control of co-medications as is routinely applied when
CYP1A2 levels are altered through smoking and sudden
smoking cessation. However, avoiding other consequences of
AhR activation, particularly regulation of cell proliferation,
apoptosis, and potential carcinogenesis could reduce risk in
drug development. Furthermore, the emergence of tumor-
ogenicity may not occur until late in a drug development
campaign, in carcinogenicity studies, or even in clinical
development, where attrition comes after significant invest-
ment.
RESULTS AND DISCUSSION
■
AhR Panel Activation and CYP1A mRNA Upregulation
in Vivo. As part of early safety assessment studies in rats,
toxicogenomic data are frequently collected in addition to
histopathology to highlight cellular pathways and processes
affected by the new chemical entity (NCE). Transcript changes
in the liver can be associated with hepatotoxicity, and through
evaluations of panels of genes, it is possible to elucidate the
mechanism of action of the hepatotoxicity. One example is the
induction of drug metabolizing enzymes, such as the CYP1A
family. The regulation of this enzyme family is classically via the
arylhydrocarbon receptor, and upregulation of this receptor is
associated with increased expression of genes including
CYP1A1, CYP1A2, NAD(P)H quinone oxidoreductase
(NQO1) and epoxide hydrolase (Ephx). This collection of
genes is referred to as the AhR panel. A positive response for
this panel is alerted when upregulation of each gene reaches a
threshold value.
Following 7-day repeat oral administration of S1P1 agonists
3−5 to rats, marked increases in the expression of genes
associated with the AhR panel were observed (Figure 2).
The induction of the mRNA was dose related, and the
magnitude of the changes associated with CYP1A1 was up to
10 000-fold the control value. Interestingly, a retrospective
analysis of the entire GSK data set revealed that these
compounds stood out as the most prolific inducers of this
gene panel.
The details of our lead campaign to discover S1P1 agonists
have been extensively published22−26 with successful optimiza-
tion of parameters such as potency, selectivity, PK, and
predicted human dose illustrated. This article details the finding
of CYP1A induction observed during the optimization and
preclinical development of selective S1P1 agonists. Figure 1
Increases in CYP1A1 mRNA showed the most change and
were considered representative of the entire gene panel.
Additionally, for these compounds, the extent of CYP1A1
mRNA upregulation was found to correlate well with unbound
systemic exposure (Figure 3). Although upregulation of mRNA
was observed, there was no change to the exposure of the test
compound on repeat dosing, suggesting the CYP1A family was
not the primary clearance mechanism of these compounds in
the rat.
Compounds 4 and 5 were terminated for developability
reasons unrelated to AhR or enzyme induction, though
compound 3 was further progressed to non-rodent safety
studies. Following repeat oral 30 mg/kg administration to
cynomolgus monkeys and in contrast to findings in the rat, a
significant reduction in systemic exposure was observed
between day 1 and day 7 (Figure 4). Consistent with the rat,
Taqman data revealed a substantial increase in hepatic CYP1A1
and CYP1A2 mRNA. To further investigate the cause of the
reduced exposure, microsomes were prepared from the livers
obtained from cynomolgus monkeys in the control and treated
groups and the in vitro clearance of compound 3 was
determined. The in vitro clearance in microsomes prepared
from the control animals showed compound 3 to be
metabolically stable, whereas a high in vitro clearance of 11
mL/min/g tissue was observed in microsomes prepared from
treated animals. Taken together these data indicate that
compound 3 induced CYP1A and was also a substrate for
this enzyme in the cynomolgus monkey, an autoinducer. There
was no reduction in exposure observed in the rat on repeat
administration, despite high levels of mRNA induction,
suggesting a species difference in the routes of clearance for
compound 3.
Figure 1. S1P1 agonists progressed to 7-day rat safety studies.
illustrates exemplar compounds from these series for which the
optimization and profile have been previously published.22−25
Consistent with the high in vitro potency and good
pharmacokinetic properties, these molecules showed efficacy
at low oral doses in an acute dosing rat lymphopenia assay and
a chronic dosing rat collagen induced arthritis model. Due to
their promising developability profiles and predicted human PK
and dose, these compounds were further progressed to 7-day
repeat dose safety studies in rats.
CYP1A Upregulation in Vitro. Although upregulation of
genes encoding the AhR panel was substantial in the rat, the
human profile is of ultimate importance. The potential for 3
8237
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
Figure 2. Induction of the AhR gene panel (hepatic mRNA fold induction versus control animals) for compounds 3, 4, and 5 following 7 days of oral
dosing to rats. The response to prototypical inducers β-napthoflavone (BNF) and 3-methylcholanthrene (3MC) after 4 days of administration is
included for reference. Cyp: cytochrome P450. NQO1: NAD(P)H quinone oxidoreductase. Ephx1: epoxide hydrolase.
Figure 3. Correlation between CYP1A1 mRNA fold induction and
exposure (AUC) for compounds 3−5 following 7-day administration
to the rat. Doses administered per day were 30 and 100 mg/kg for
compound 3; 30, 100, and 300 mg/kg for compounds 4 and 5. Part A
shows the correlation with total exposure and part B shows the
correlation with unbound exposure, corrected using the unbound
Figure 4. (A) Exposure data for compound 3 on days 1 and 7 in
fraction in blood.
monkeys following daily dosing at 30 mg/kg. (B) Taqman analysis for
CYP1A1 and CYP1A2 from the livers of treated animals versus
animals receiving vehicle control.
and 4 to interact with the human AhR and cause induction was
studied in human hepatocytes using end points of CYP1A2
mRNA and catalytic enzyme activity (determined using the rate
of CYP1A mediated deethylation of the CYP1A probe substrate
7-ethoxyresorufin) as markers of enzyme induction. As
mentioned previously, because of the low level of hepatic
CYP1A1 expression and the predominant location of CYP1A1
extrahepatically the study of human CYP1A1 induction in vitro
is less straightforward due to the challenges of obtaining
metabolically competent cells from tissues where CYP1A1 is
inducible, such as the lung. We therefore used an available
human hepatocyte assay with CYP1A2 mRNA and activity as
the end point, with the assumption that changes in 1A2 also
represent 1A1.
Compound 3 was shown to cause a 45-fold increase in
CYP1A2 human mRNA levels (Table 1) with an associated
8238
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
Table 1. Comparison of the CYP1A1, CYP1A2 (Rat), and CYP1A2 (Human) Induction Observed across Various in Vivo and in
a
Vitro Assay Formats
b
c
d
rat in vitro
rat in vivo
human in vitro
compd
CYP1A1 mRNA
CYP1A2 mRNA
CYP1A1 mRNA
CYP1A2 mRNA
CYP1A2 mRNA
catalytic activity (EROD)
1
3
4
5
6
7
<1
<1
5
<1
<1
3
3
35
1
2
8
2
2
1
0.2
45
1.1
7.7
1.1
75
0.6
3
2
210
2
43
32
19
8
1
a
b
c
All data expressed as fold-change versus control to enable comparison. Rat in vitro assays conducted at 10 μM. Compounds 1−5: 7-day dosing 30
d
mg/kg. Compounds 6 and 7: 4-day dosing 20 mg/kg. mRNA assay at 10 μM, EROD assay at 5 μM. EROD: 7-ethoxyresorufin O-deethylase
activity.
increase in CYP1A2 catalytic enzyme activity. This finding was
consistent with the CYP1A1 and CYP1A2 mRNA upregulation
observed in rat and cynomolgus monkey substantiating the
evidence for an interaction with the AhR across the species.
However, in contrast to findings in the rat, compound 4 did not
cause upregulation of CYP1A2 mRNA in human hepatocytes,
highlighting a species difference. Although mRNA levels for
CYP1A2 were not increased by compound 4 in human
hepatocytes, the low levels of expression of CYP1A1 in the
human liver complicate the evaluation of CYP1A1 upregula-
tion. The possibility still existed that compound 4 was
interacting with the AhR, but no observation could be made
using human hepatocytes.
and also did not upregulate CYP1A2 in human hepatocytes
(Table 1), we progressed compound 1 to further rodent and
non-rodent safety studies. The program strategy required the
identification of backup compounds that would also need to be
devoid of any potential interaction with the AhR gene panel.
We therefore embarked on a series of experiments with the
objectives of understanding the structure−activity-relationships
and to establish a screening cascade to filter out the AhR gene
panel induction.
Due to the complexities of species differences in expression
and upregulation of enzymes encoded by the AhR, coupled
with the developability risks associated with AhR agonism, a
decision was taken to design out mRNA upregulation and
CYP1A enzyme induction in all species as a way of mitigating
the risk of AhR agonism. The most sensitive marker of this
response for our compounds was CYP1A1 mRNA upregulation
in the rat, and this end point was selected to optimize
compounds.
“Screening” for CYP1A Induction. Utilizing an in vitro
assay system to predict effects in vivo is common approach to
screening in drug discovery. We initially investigated the utility
of rat hepatocytes as a system to predict the in vivo findings.
Compounds 1, 3, 4, and 5 were incubated with rat hepatocytes,
and the increase in mRNA levels for CYP1A1 and CYP1A2 was
quantified using Taqman analysis. The induction of CYP1A1
and CYP1A2 mRNA is presented in Table 1 where it can be
seen that the extent of induction observed in vivo was not
reflected in vitro. Refinements to the in vitro assay system were
conducted and multiple methods of data analysis attempted
(including % change compared to positive control, EC50, and
maximum induction), but an in vitro−in vivo correlation of
CYP1A enzyme induction was not established.
Due to the upregulation of CYP1A2 in human hepatocytes
and substantial reduction in systemic exposure on repeat
administration to the monkey, compound 3 was unable to be
progressed further. Compound 1 was subsequently identified
and entered safety testing in rats. Offering other developability
advantages but having structural similarity with compounds 3−
5, compound 1 was also expected to upregulate the genes of the
AhR panel. However, inspection of the Taqman data following
7-day administration of compound 1 showed levels of gene
expression within the normal range, suggesting a reduced or
absent interaction with the AhR in rat. Compound 1 was
therefore positively differentiated from the previous com-
pounds 3−5 (Figure 5).
As we had now discovered a compound that did not induce
mRNA changes associated with the AhR gene panel in the rat
As the in vitro CYP1A induction assay in rat hepatocytes was
not able to predict the extent of induction in vivo, we
introduced a short-term in vivo screen in the rat. Data from
Baldwin et al.27 demonstrated that hepatic CYP1A mRNA
induction by β-napthoflavone was observed after three daily
administrations. We therefore established an in vivo screening
protocol where test compounds were administered orally to
rats at a single nominal dose level of 30 mg/kg for 4 days. The
dose level was selected to represent the most likely lowest dose
in a 7-day safety assessment study and was manageable with
respect to the availability of drug substance where no greater
than 150 mg of compound was required to dose n = 3 rats. The
systemic exposure was measured on days 1 and 4, the animals
were then culled, the livers were harvested, and the extent of
mRNA induction of genes in the AhR gene panel was
quantified.
Figure 5. Correlation between CYP1A1 mRNA fold induction and
unbound exposure (AUC) for compound 1 (noninducer) compared
to compounds 3−5 (inducers) following 7-day administration to the
rat. Doses administered per day were 1, 30, 100, and 300 mg/kg for
compound 1; 30 and 100 mg/kg for compound 3; 30, 100, and 300
mg/kg for compounds 4 and 5.
8239
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
The first two molecules to enter such studies were the
enantiomeric pair 6 and 7 (Figure 6) for which the in vitro and
Figure 8. Crystallographic data for compounds 1 and 4. The values
represent measured dihedral angles.
Figure 6. Benzoxazepine S1P1 agonists devoid of CYP1A mRNA
upregulation in vivo.
in vivo profile have already been reported.25 Good systemic
exposure was observed for both compounds, and TaqMan
analysis revealed no significant change in hepatic CYP1A
mRNA levels. Interestingly, these two compounds showed
some mRNA upregulation in the in vitro rat hepatocytes assay
(Table 1), further highlighting the discrepancy between in vitro
and in vivo data.
The identification of compounds from the same triaryl series
with a marked difference in their ability to activate the AhR in
the rat prompted us to investigate the impact of subtle
structural changes. A survey of the literature demonstrated that
a significant proportion of compounds described as AhR
agonists are planar hydrophobic polyaromatics such as dioxin 8
(TCDD, Figure 7) or β-benzoflavone 9,28 even if disruption of
the case of the noninducers because of the presence of an
asymmetric carbon. However, it is also possible that the change
from phenyl to pyridine in the distal ring may also impact AhR
binding.
Second Generation S1P1 Agonists. We have presented
the identification of zwitterionic S1P1 agonists (1, 6, and 7)
devoid of CYP1A induction with an overall profile suitable for
progression to preclinical development. In addition to effects of
S1P1 agonists to cause a reduction in peripheral circulating
lymphocytes, emerging data with FTY72032,33 suggested that
efficacy in human multiple sclerosis and its animal model may
also involve direct effects on S1P receptor-mediated signaling in
CNS. This prompted us to explore a subsequent generation of
compounds with CNS penetration properties. We explored this
using the analogous amines to the zwitterions previously
described, aware of the known ability of lipophillic amines to
cross the blood−brain barrier.34
Compound 12 (Table 2) was the first druglike triarylamine
identified with good in vitro potency and selectivity.
Pharmacokinetic properties suitable for oral delivery were
confirmed, and maximum efficacy was demonstrated in the rat
lymphopenia assay at an oral dose of 1 mg/kg. This molecule,
analogous to the AhR negative compound 1, was therefore
progressed to the in vivo rat 4-day induction screen using an
oral dose level of 30 mg/kg. Unfortunately, TaqMan analysis
revealed an average 62-fold increase in CYP1A1 mRNA. This
demonstrated that the lack of planarity per se does not
systemically lead to non AhR agonists.
Without a predictive in vitro assay system and being unable
to demonstrate consistent SAR with AhR gene panel induction,
our path forward was restricted to systematic progression of
compounds with the requisite in vitro pharmacology, PK,
efficacy profile, and human dose prediction into the 4-day
repeat dose rat study. Compounds devoid of AhR gene panel
activation potential were considered as candidates for
preclinical development. The compounds progressed into the
4-day study are highlighted in Figure 9.
Figure 7. Examples of reported AhR agonists.
planarity has been demonstrated to be tolerated in some
occasions.29 “Nonclassical”16 AhR agonists such as benzothia-
zole 5F 203, 10,30,31 or leflunomide 1128 have also been
reported.
Crystallographic data for compounds 1 and 4 highlighted
that one of the aromatic rings in compound 1 (considered AhR
negative) was out of plane when compared with compound 4
(AhR positive). The dihedral angles between the oxadiazole
ring and the phenyl closest to the basic nitrogen were 18°
versus 5°, respectively (Figure 8). We postulated that the ortho-
methylated THIQ (1) was sufficiently out of plane to avoid
binding the AhR. This assumption was in part reinforced by the
fact that the main difference between compounds 4, 6, and 7
was the presence of a carboxylic chain out of the triaryl plain in
Chemistry. The general route to access these compounds
(with the exception of compound 16) is highlighted in Scheme
1 and uses the formation of the oxadiazole (via condensation
and dehydration) as a key step. Once the amines 26 are
deprotected in acidic condition, the final compounds can be
8240
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
rings and their substituents has a significant impact on the
measured pKa (compare 2, 12−14, and 16−17 or 18−23).
When measured, in vivo rat pharmacokinetics demonstrated
oral exposure with distribution into tissues and a moderate
blood clearance (Table 4).
Table 2. In Vitro and in Vivo Properties of Compound 12
Basic S1P1 Agonists and CYP1A mRNA Upregulation
in Vivo. Compounds 2, 12−24 were progressed into the rat 4-
day in vivo CYP1A induction screening study. All data
associated with these studies are shown in Table 5, and the
CYP1A1 mRNA upregulation is additionally depicted in Figure
10. Doses used were in the range 15−30 mg/kg, selected to be
representative of the lower end of the dose required in safety
testing but also determined by compound availability. The
AUC and Cmax values indicated systemic exposure remained
constant over the 4-day study period, providing no evidence for
autoinduction in the rat. Total exposure was generally within 2-
fold (see compounds 14−24), though it is noted that the
AUC/D values for compounds 2, 12, and 13 are lower. It is
clear from this data set that systemic exposure is not the sole
determinant of the extent of induction, as exemplified by
noninducer compounds 15, 16, and 17 having a higher
exposure than the inducers 12 and 13 within the THIQ
template. Furthermore, the data set confirms the subtle nature
of the SAR with compounds having relatively equivalent
properties, showing marked differences in their ability to cause
CYP1A induction.
a
1 mg/kg iv DMSO/10% (w/v) Kleptose HPB 0.9% saline (aqueous)
b
(2%:98% v/v). 3 mg/kg oral. 1% (w/v) methylcellulose 400
(aqueous).
The extent of induction of CYP1A1 versus CYP1A2 for all
compounds described is shown in Figure 11. A compound was
considered an inducer when CYP1A1 mRNA increases >20-
fold, and CYP1A2 mRNA increases >5-fold were observed.
Looking more broadly across the templates all the amine
benzazepines and benzoxazepines tested showed substantial
CYP1A induction of >40-fold (compounds 18−24). Lower
levels of induction were only observed with some (aza-) THIQ
derivatives. Of the seven (aza-) THIQ compounds tested, only
compounds 12 and 13 showed some upregulation. The greater
induction observed with 12 and 13 is intriguing, given the
structural similarity with compound 2, the only difference
residing in the arrangement of the hydrophilic side chain. The
observation that such a small modification can result in a large
difference in vivo is difficult to interpret when other small
modifications from 2 (aza-THIQ 15 or thiazole 16) or when
more drastic shape modifications (compound 17) have no
effect on CYP1A1 induction.
Similar comments can also be made when comparing the
structure of the benzazepine (compounds 18−23) or the
benzoxazepine (compound 24) with the noninducers THIQs.
Compounds 20 and 24 differ from 2 by one carbon and one
oxygen, respectively, and showed induction. When compared
with the THIQ series, modification of the hydrophilic side
chain in the benzazepines has no notable impact with all
compounds (compare 18 and 19 with 20) considered
substantial inducers. The effect of increasing planarity
(compound 22) appears to lead to increased induction (600-
fold versus 275 for compound 20), but the increased exposure
may contribute to this observation. Lowering lipophilicity on
the distal ring (compound 21) may appear to reduce induction
(80-fold versus 275 for compound 20), but this effect is not
dramatic and other contributing factors may be considered.
One may be tempted to look at these data as a “template
effect”, with the benzazepines and benzoxazepine showing
generally more induction than the THIQs. Moreover, due to
the similarity of intrinsic properties between 2 and 12−24
obtained via reductive amination using optionally protected
dihydroxy aldehyde (see Experimental Section).
The syntheses of the key intermediates to access compounds
2, 12−15, 17, 22, and 23 have already been described. The
syntheses of 27 and 28, key building blocks to access
compounds 18, 21, and 24, are described in Scheme 2 and 3,
respectively.
The synthesis of 27 uses benzoic acid 29 as starting material.
Straightforward modifications of the acid lead to acetal 30
which cyclizes to 31 in acidic conditions. Reduction of the
double bond and the amide functionality of 31 followed by
demethylation of the phenol leads, after protection of the
secondary amine, to 32. The nitrile of 27 is inserted using
palladium chemistry from the phenol triflate of 32. All
transformations are high yielding and can be performed on
multigram quantities.
Compound 28 is obtained from resorcinol 33 which can be
monoformylated. Reductive amination using ethanolamine
leads, after protection of the secondary amine, to 34. The
benzoxapine 35 is then obtained via a Mitsunobu reaction. As
for 27, the nitrile of 28 is introduced using palladium chemistry.
Compound 16 is easily obtained from triaryl 40 (Scheme 4).
The latter is obtained via Suzuki coupling between 37 and 39,
obtained respectively from 36 and 38 whose syntheses have
already been published.37
Profile of Basic S1P1 Agonists. The physicochemical
properties of these compounds are shown in Table 3. As can
been seen from these data, the compounds progressed had
similar properties with respect to molecular weight and polar
surface area, with measured CHILogD at pH 7.4 ranging from
2.15 to 2.84. With the exception of compound 15 (cLogP =
2.35), the cLogP is within a 1.1 unit window. The main
difference could be considered the pKa of these compounds,
mainly driven by the nature of the bicyclic amine (aza-THIQ,
THIQ, benzazepine, and benzoxazepine), as neither the nature
of the amine substituent nor the nature of the other aromatic
8241
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
Figure 9. Structures of compounds 2, 12−24.
a
Scheme 1. General Route To Access Oxadiazole Containing S1P1 Agonists
a
Reagents and conditions: (a) palladium-mediated nitrile formation; (b) addition of hydroxylamine; (c) 25, base, heat; (d) HCl; (e) reductive
amination; (f) alcohol deprotection.
(cLogP, PSA, CHILogD), we may consider that the induction
is related to pKa. Indeed the benzazepines (pKa range 7.8−8.4)
are more basic than the THIQs (pKa range 7.2−7.7, aza-THIQ
15 pKa of 6.2). However, this argument is not supported by the
8242
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Benzazepine Intermediate 27
a
Reagents and conditions: (a) BH3·THF, THF, 0 °C to room temperature, 95%; (b) MsCl, NEt3, CH2Cl2, 0 °C to room temperature, 77%; (c)
NaCN, DMF, 90 °C, 86%; (d) NaOH, EtOH, reflux, 97%; (e) (COCl)2, cat. DMF, CH2Cl2, room temperature, then NEt3, NH2CH2CH(OCH3)2,
98%; (f) conc HCl, AcOH, room temperature, 76%; H2, Pd/C, MeOH, room temperature, 98%; (h) BH3·THF, THF, room temperature, 96%; (i)
aqueous HBr, 100 °C, then Boc2O, NEt3, CH2Cl2/THF, room temperature, 52%; (j) (CF3SO2)2O, pyridine, CH2Cl2, −50 °C to room temperature,
85%; (k) Zn(CN)2, Pd(PPh3)4, DMF, 120 °C, 95%.
a
Scheme 3. Synthesis of Benzoxazepine Intermediate 28
a
Reagents and conditions: (a) POCl3, DMF, 0 °C to room temperature, 40%; (b) NH2(CH2)2OH, EtOH, room temperature; (c) NaHB(OAc)3,
THF, room temperature; (d) Boc2O, NEt3, MeOH, room temperature, 35% (3 steps); (e) PPh3, DIAD, THF, 0 °C, 84%; (f) (CF3SO2)2O, pyridine,
CH2Cl2, 0 °C, 100%; (g) Zn(CN)2, Pd(PPh3)4, DMF, 100 °C, 91%.
a
Scheme 4. Synthesis of Thiadiazole Containing S1P1 Agonists
a
Reagents and conditions: (a) (CF3SO2)2O, pyridine, CH2Cl2, room temperature, 100%; (b) bis(pinacolato)borane, KOAc, PdCl2(dppf), 1,4-
dioxane, reflux, 88%; (c) POCl3, 90 °C, 99%; (d) CuBr2, isoamyl nitrite, CH3CN, room temperature, 63%; (e) Na2CO3, PdCl2(dppf)·CH2Cl2,
DME/H2O, reflux, 71%.
level of induction observed with benzoxazepine 24, with a
measured pKa of 6.3.
subtle structural changes could impact the extent of induction
observed. A robust relationship between in vitro and in vivo
induction in the rat was not established. Due to species
differences in the extent of AhR activation, a decision was taken
to design out evidence of induction in all species, and due to
the lack of in vitro−in vivo correlation, a short-term rat in vivo
model was successfully established as a decision making tool.
This strategy allowed us to identify compounds devoid of
CYP1A induction that were progressed to further development.
It is acknowledged that adopting the strategy to only
progress compounds devoid of induction in both rat and
human could be considered conservative. However, there was
no precedence within GSK for development of a compound
with this type of profile. Additionally, we felt the magnitude of
the upregulation observed justified our conservative approach.
Evaluation of AhR activity in early safety studies is essential to
understand the developability risks associated with a candidate
CONCLUSION
■
Overall, the data presented lead to the conclusion that the SAR
between structure and AhR gene panel upregulation was
complex and undoubtedly multifactorial. The structural
relationships were subtle and extended beyond general
properties such as planarity, basicity, or lipophilicity.
This program of work demonstrated a finding of marked
upregulation of genes encoding the AhR during profiling of
numerous S1P1 agonist compounds. CYP1A1 mRNA upregu-
lation was largely used as a marker for AhR activation, and the
magnitude of upregulation of CYP1A1 mRNA in the rat by
certain compounds was one of the largest observed within the
GSK data set to date. We explored the structure−activity
relationships of this finding in the rat in vivo and demonstrated
8243
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
a
Table 3. Physicochemical Profiles of Amines 2, 12−24
CHILogD,
pH 7.4
measured
pKa
compd MW cLogP
PSA
template
2
449
449
449
449
486
501
471
463
463
463
499
449
449
464
3.75
3.66
3.66
2.97
2.35
3.66
3.55
3.83
3.83
3.95
3.02
3.75
2.82
3.57
2.76
2.70
2.70
2.84
2.15
2.33
2.58
2.36
2.36
2.64
2.44
2.78
2.63
2.86
116
116
116
129
129
102
116
116
116
116
131
116
131
125
THIQ
THIQ
THIQ
THIQ
aza-THIQ
THIQ
THIQ
BZ
7.3
7.7
7.7
7.3
6.2
7.2
7.3
8.4
8.4
8.2
8.3
7.8
8.1
6.3
12
13
14
15
16
17
18
19
20
21
22
23
24
BZ
BZ
BZ
BZ
Figure 10. The upregulation of CYP1A1 mRNA following repeat
administration of basic S1P1 agonists to the male CD rat. The
compounds are categorized by template: THIQ (red), aza-THIQ
(pink), benzazepines (blue), benzoxazepines (green).
BZ
BZO
a
THIQ: tetrahydroisoquinoline. BZ: benzazepine. BZO: benzoxaze-
pine.
Table 4. Pharmacokinetics of Compounds 12, 22, 23, and 24
a
in the Rat
compd
12
22
23
24
b
CLb (mL/min/kg) 53
6
21
3
22 (n = 1) 54 (n = 1)
(n = 3)
6.6 0.2
1.7 0.2
(n = 3)
7.6 0.4
5.3 0.3
68 18
b
Vss (L/kg)
10.9
4.0
b
T1/2 (h)
8.1
1.0
c
oral bioavailability
(%)
56
7
67 14
69 11
(n = 3)
a
b
Data are for n = 1 or are the mean SD at n = 3. 1 mg/kg iv
DMSO/10% (w/v) Kleptose HPB 0.9% saline (aqueous) (2%:98% v/
v). 3 mg/kg oral in 1% (w/v) methylcellulose 400 (aqueous).
c
Figure 11. Upregulation of CYP1A1 and CYP1A2 mRNA following
repeat administration to the male CD rat. The compounds are
categorized by template, THIQ (red), aza-THIQ (pink), indazoles
(yellow), benzazepines (blue), benzoxazepines (green), and shaped by
class, acid (square), amine (circle), zwitterions (diamond). The shaded
area indicates the boundaries considered a noninducer (20-fold
CYP1A1 and 5-fold CYP1A2).
compound. While the drug−drug interaction risks associated
with CYP1A enzyme induction may be considered manageable
in the clinic, there are further potential and emerging risks
associated with AhR agonism and CYP1A enzyme induction,
for example, on cellular proliferation and tumorogenesis.18
Table 5. CYP1A1 mRNA Upregulation and Exposure Data following 4-Day Repeat Administration of Selected Amines to the
a
Male SD Rat
Cmax (μg/mL)
AUC (μg·h/mL)
AUC/D (min·kg/L)
dose
Cyp1A1 mean
fold increase
Cyp1A2 mean
fold increase
compd
template
(mg/kg)
day 1
day 4
day 1
day 4
day 1
0.4
day 4
2
THIQ
THIQ
THIQ
THIQ
aza-THIQ
THIQ
THIQ
BZ
30
30
30
22
17
19
15
19
25
15
24
30
24
25
1.4
62
1.5
4.6
10
2
0.44 0.06 0.79 0.24 4.42 0.13 7.64 0.71
9
16 1.2
12
13
14
15
16
17
18
19
20
21
22
23
24
0.72 0.27 0.71 0.13 7.31 2.1
0.77 0.26 0.50 0.05 4.44 1.1
7.0 1.6
3.6 0.9
28.8 3.7
15
9
4
15
9
3
64
2
2
7
2.1 0.5
2.73 1.0
24.3 1.2
12.9 1.5
66
4
5
9
6
8
79 10
0.9
1.7
1.6
126
215
275
81
2
1.61 0.17 2.17 1.0
16.8 (n = 2) 46
59 (n = 2)
89 10
1
1.75 0.06 2.32 0.54 19.1 2.7
0.90 0.12 0.94 0.01 12.1 1.5
1.29 0.08 1.53 0.45 20.8 2.6
0.73 0.32 0.87 0.11 13.2 5.6
0.55 0.03 0.67 0.05 9.1 0.57
27.8 4.2
13.9 1.8
26 6.3
61
48
65
2
56
6
1.7
2.2
1.9
2
83 21
BZ
15.6 2.2
11.5 1.3
24.9 4.1
23.1 1.7
21.3 1.8
12.6 0.4
31 13 38
5
6
BZ
36
43
48
38
22
2
9
2
1
4
46
BZ
1.14 0.2
1.35 0.21 17 3.9
63 11
BZ
604
100
40
7.3
2
1.60 0.11 1.38 0.12 23 0.4
1.01 0.05 1.22 0.12 15.3 0.9
0.87 0.06 1.14 0.08 8.84 1.7
59
55
30
4
5
2
BZ
BZO
2
a
PK data are the mean SD, n = 3, unless otherwise stated.
8244
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
concentration was added to the bead and membrane suspension.
The bead, membrane, saponin, and GDP suspension was mixed with
[35S]GTPγS (PerkinElmer, 0.3 nM final radioligand concentration)
made in assay buffer (HEPES 20 mM, MgCl2 10 mM, NaCl 100 mM
and pH adjusted to 7.4 using KOH 5 M). The bead, membrane, and
radioligand suspension was dispensed into white Greiner polypropy-
lene 384-well plates (45 μL/well), containing 0.5 μL of a solution of
test compound in 100% DMSO. The final assay cocktail (45.5 μL) was
then sealed, spun on a centrifuge, then read on a Viewlux instrument
following a 3 h incubation of plates at room temperature.
Understanding the consequences of AhR activation is a rapidly
expanding area of scientific research including its utility as a
therapeutic target.35 While we are not advocating to avoid AhR
agonism completely, we suggest caution and the operation of
an appropriate strategy to understand and mitigate potential
risks while the implications of substantial AhR agonism are
more thoroughly understood.
EXPERIMENTAL SECTION
■
In Vitro DMPK. Rat and Human Hepatocyte Induction
Studies. Cryopreserved rat and human hepatocytes were purchased
from Celsius (Baltimore, MD, U.S.) and Invitrogen (Paisley, U.K.).
The hepatocytes were removed from liquid nitrogen storage,
immediately thawed at 37 °C for 90 s, and transferred in prewarmed
cryopreserved hepatocyte recovery medium. The cells were centri-
fuged at 100g for 10 min and resuspended in cryopreserved hepatocyte
plating medium (at 4 °C) before being counted under a microscope
using a hemocytometer. The viability was assessed using the trypan
blue exclusion method, and only cell suspensions with viability of
>80% were used. Cells were seeded according to the supplier’s
protocol and cultured in Williams medium E (WME) supplemented
with fetal bovine serum, dexamethasone, gentamycin, insulin, L-
glutamine, and HEPES (plating supplement pack, Cellzdirect) and
allowed to acclimatize in a humidified CO2 incubator (5% CO2) at 37
°C for 4 h. The medium was then replaced with WME containing ITS
+ (insulin, human transferrin, and selenous acid), dexamethasone,
gentamycin, L-glutamine, and HEPES (cell maintenance supplement
pack, Cellzdirect) and 0.25 mg/mL Matrigel. The hepatocytes were
allowed to acclimatize in a humidified CO2 incubator (5% CO2) at 37
°C for 1 day. Following acclimatization, the medium was removed and
replaced with prewarmed culture medium solutions containing test
compound or positive control (in duplicate). Hepatocytes were also
exposed to solvent control in quadruplicate. All cells were then
incubated for a further 48 h with one change of drug/controls after 24
h. At the end of the incubation period, the medium was removed and
the cells were lysed. Isolation and analysis of mRNA levels of specific
P450 genes were conducted using the methods described by Baldwin
et al.27 The mRNA level for each specific CYP was expressed as a
mean ratio of treated over solvent control. The EROD assays were
performed to assess CYP1A2 activity in human hepatocytes. Cells
were washed with Williams medium E prior to incubation with 0.5 μM
ethoxyresorufin in a serum-free medium. Following incubation for 10
min at 37 °C, the fluorescence was measured, with excitation at 530
nm and emission at 590 nm, using a CytoFluor microplate reader. The
CYP activity was collated as fluorescence units per min in each sample
and expressed as fold increase over control (solvent only).
Microsomal Intrinsic Clearance and mRNA Analysis from
Cynomolgus Monkey Livers. Materials. HPLC-grade DMSO,
phenacetin, furafylline, glucose 6 phosphate, nicotinamide adenine
dinucleotide phosphate, glucose 6-phosphate dehydrogenase, sodium
bicarbonate, sodium chloride, sucrose, Tris-HCl, pH 7.4, were
purchased from Sigma (St. Louis, MO). Potassium phosphate dibasic
and monobasic were obtained from Mallinckrodt (St. Louis, MO). 3
was prepared by GSK. HPLC-grade acetonitrile was purchased from
EMD Chemicals, Inc. (Gibbstown, NJ). The 96-well sample tubes
were purchased from Matrix Technologies (Hudson, NH). Pierce
BCA protein assay reagent kit was purchased from Thermo Scientific
(Waltham, MA).
Methods: Microsome Preparation. Microsomes from control and
treated cynomolgous monkey livers were prepared using the method
described by Jewell et al.36 Control and induced monkey livers were
washed in ice-cold saline (0.9% w/v sodium chloride) and then
homogenized in the presence of sucrose/Tris buffer. Homogenate was
centrifuged at 10 000g for 20 min at approximately 4 °C. The
supernatant was removed and centrifuged at 105 000g for 60 min at 4
°C. The remaining pellet was resuspended in fresh sucrose/Tris buffer
and then centrifuged at 105 000g for 60 min at 4 °C. The supernatant
was discarded and the pellet resuspended in 50 mM potassium
phosphate buffer and stored at −80 °C prior to use. The protein
In Vitro Pharmacology. Materials used were the following:
DMEM F12 catalog no. Sigma D6421 or Invitrogen 11320; charcoal
stripped FBS catalog no. Hyclone SH30068.03. Base buffer was
prepared from Sigma kit H1387, supplemented with 20 mM HEPES
(Sigma H0887) and 4.16 mM NaHCO3 (Sigma S8761), and pH was
adjusted to 7.4. Loading buffer was base buffer supplemented with
0.1% pluronic acid F-68 solution (Gibco/Invitrogen 24040-032) and
0.1% BSA (Calbiochem 126609). Assay buffer/diluent buffer was base
buffer supplemented with 0.1% pluronic acid F-68 solution (Gibco/
Invitrogen 24040-032). Coelenterazine was catalog no. Invitrogen
C6780 or Biotium BT10110-1.
S1P1 β-Arrestin Recruitment Assay. β-Arrestin recruitment
assays were carried out using the PathHunter CHO-K1 EDG1 β-
arrestin cell line (DiscoveRx Corporation) in a chemiluminescence
detection assay. This cell line stably expresses β-arrestin 2 and S1P1
fused to complementing portions of β-galactosidase (“EA” and
“prolink”, respectively) which associate upon arrestin recruitment to
form functional β-galactosidase enzyme. Cells were grown to 80%
confluency in growth medium (F12 nutrient HAMS supplemented
with 10% heat-inactivated USA FBS, 1% L-Glutamax, 800 μg/mL
Geneticin, and 300 μg/mL hygromycin). Cells were harvested from
the flask using enzyme free cell dissociation buffer (Gibco) and washed
from flasks with Optimem solution (Gibco). Cells were then
centrifuged at 1000 rpm for 2−3 min and resuspended in assay buffer
(prepared from Sigma kit H1387 supplemented with 20 mL/L
HEPES, 4.7 mL/L NaHCO3, 0.1% pluronic acid F-68 solution, 0.1%
BSA and adjusted to pH 7.4 using sodium hydroxide at 1 × 106 cells/
mL). Cells were dispensed into assay plates containing compounds
(100 nL/well of a solution of test compound in 100% DMSO) at 1 ×
104 cells/well and incubated at 37 °C/5% CO2 for 90 min followed by
15 min at room temperature. An amount of 5 μL of detection mix (1
part Galacton Star, 5 parts Emerald II, 19 parts assay buffer;
DiscoveRx) was added per well, and the plates were incubated at room
temperature for 60 min. Luminescence was quantified using a Viewlux
plate reader.
Membrane Preparation for S1P3 GTPγS Assay. All steps were
performed at 4 °C. Cells were homogenized within a glass Waring
blender for two bursts of 15 s in 200 mL of buffer (50 mM HEPES, 1
mM leupeptin, 25 μg/mL bacitracin, 1 mM EDTA, 1 mM PMSF, 2
μM pepstatin A). The blender was plunged into ice for 5 min after the
first burst and 10−40 min after the final burst to allow foam to
dissipate. The material was then spun at 500g for 20 min and the
supernatant spun for 3 min at 48 000g. The resultant pellet was
resuspended in the same buffer without PMSF and pepstatin A but
containing 10% w/v sucrose. The membrane suspension was then
layered on top of buffer without PMSF and pepstatin A containing
40% w/v sucrose and spun at 100 000g for 60 min. The cloudy
interface between the two sucrose layers was removed and
resuspended in buffer without PMSF and pepstatin A. The material
was spun at 48 000g for 45 min. The resultant cell pellet was
resuspended in the required volume in buffer without PMSF and
pepstatin A (usually ×4 the volume of the original cell pellet),
aliquoted, and stored frozen at −80 °C.
S1P3 GTPγS Assay. S1P3 expressing RBL membranes (0.44 μg/
well) purified through a sucrose gradient were homogenized by
passing through a 23G needle. These were then adhered to WGA-
coated SPA beads (GE Healthcare 0.5 mg/well) in assay buffer
(HEPES 20 mM, MgCl2 10 mM, NaCl 100 mM and pH adjusted to
7.4 using KOH 5 M). An amount of 2 μg/well of saponin was added.
After 30 min of precoupling on ice, 5 μM GDP final assay
8245
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
concentration was determined using the method described by Lowry
et al.37
Rat 4 Day Oral CYP1A mRNA Induction Study. Male CD rats
(226−325 g, supplied by Charles River UK Ltd.) were selected for
study based on health. Rats had free access to food and water
throughout. Rat PK studies were conducted as a repeat dose design,
with once daily oral administration for 4 days. On each study day, n =
3 male rats received an oral gavage of the test compound formulated in
a 1% (w/v) methylcellulose aq suspension at a concentration of 3−6
mg/mL administered by gavage at 5 mL/kg to achieve a target dose of
15−30 mg/kg. Tail vein blood sampling was conducted on days 1 and
4 up to 24 h postdose and frozen prior to sample analysis. At the end
of the study the rats were exsanguinated via abdominal aorta under
isofluorane anesthesia. At necropsy, a liver section was removed from
the left lateral lobe and cut into six approximately equal pieces and
mixed with RNALater and placed on crushed wet ice for RNA/
Hepatotaq analysis.
Blood Sample Analysis. Diluted blood samples (1:1 with water)
were extracted using protein precipitation with acetonitrile containing
an analytical internal standard. An aliquot of the supernatant was
analyzed by reverse phase LC−MS/MS using a turbo ionspray
interface in positive ion mode. Samples were assayed against
calibration standards prepared in diluted (1:1 with water) control
blood.
Metabolic Stability. The compound was incubated with isolated liver
microsomes of both control and induced cynomolgous monkey livers
at 37 °C for 30 min in 50 mM potassium phosphate buffer (pH 7.4)
and 50 μL of NADPH generating system cofactor. At 0, 3, 6, 9, 12, 18,
and 30 min, 50 μL of the incubation mix was removed and quenched
with 200 μL of solvent containing an internal standard. Appropriate
control incubations were also conducted simultaneously. The samples
were analyzed by LC−MS/MS, and the intrinsic clearance values were
determined according to Clarke et al.38
In Vitro Metabolic Stability in Hepatocytes. Materials.The
metabolic stability of test compounds (0.5 μM) was determined in
pooled human, Sprague−Dawley rat, beagle dog, and CD-1 mouse
cryopreserved primary hepatocytes (CellzDirect Invitrogen Corpo-
ration, Durham, NC, USA). Hepatocytes were a pool from at least 50
donors.
Methods. A 5 mM stock solution of test compound was prepared in
DMSO and diluted to 100 μM with DMSO. An aliquot was taken and
further diluted to 1 μM with Williams medium E (WME) and
prewarmed at 37 °C. The cryopreserved hepatocytes for each species
were thawed in the manner recommended by the supplier. Cell
concentration and viability were ascertained using the Trypan blue
exclusion method. The cell suspension was then diluted to 1.4 million
cells per mL with prewarmed (37 °C) WME. An amount of 300 μL of
cell suspension was added to each well of a 12-well plate to which 300
μL of 1 μM test compound solution was also added to give a final
incubation concentration of 0.5 μM test compound and 0.7 million
cells/mL. The incubations were maintained at 37 °C for the duration
of the experiment. At set intervals up to 120 min 25 μL of cell
suspension was transferred to a 96-well block containing 100 μL of
stopping solution containing internal standard. The samples were
centrifuged, and the resulting supernatant was analyzed by LC−MS/
MS. Intrinsic clearance was calculated by nonlinear regression analysis
of peak area ratio vs time using Grafit, version 5.0.8 (Erithacus
software, U.K.) to determine the first order elimination rate constant
(k). To scale the hepatocyte clearance, a hepatocyte yield of 1.2 × 10 8
cells per gram of liver in human, rat, monkey, and mouse and 2.4 × 10
Hepatotaq Gene Expression Analysis. Gene expression analysis
of liver samples was conducted according to the methods described in
Swain et al.40 Liver samples on ice were homogenized with TRIzol
before RNA purification, DNasing of extract, and conversion to cDNA
took place. The prepared samples were analyzed on low density arrays
to provide Cyp1a1 data.
PK Data Analysis. PK parameters were obtained from the blood
concentration−time profiles using noncompartmental analysis with
WinNonlin Professional 4.1a (Pharsight, Mountain View, CA).
Chemistry. General Methods. All solvents were purchased from
Sigma-Aldrich (Hy-Dry anhydrous solvents), and commercially
available reagents were used as received. All reactions were followed
by TLC analysis (TLC plates GF254, Merck) or LCMS (liquid
chromatography−mass spectrometry) using a Waters ZQ instrument.
NMR spectra were recorded on a Bruker nanobay 400 MHz or a
Bruker AVII+ 600 MHz spectrometers and are referenced as follows:
1H NMR (400 or 600 MHz), internal standard TMS at δ = 0.00; 13C
NMR (100.6 or 150.9 MHz), internal standard CDCl3 at δ = 77.23 or
DMSO-d6 at δ = 39.70. Column chromatography was performed on
prepacked silica gel columns (30−90 mesh, IST) using a Biotage SP4.
Mass spectra were recorded on Waters ZQ (ESI-MS) and Q-Tof 2
(HRMS) spectrometers. Mass directed autoprep was performed on a
Waters 2767 with a MicroMass ZQ mass spectrometer using Supelco
LCABZ++ column.
39
8
cells per gram of liver in the dog were used.
In Vivo Studies. All animal studies were ethically reviewed and
carried out in accordance with Animals (Scientific Procedures) Act
1986 and the GSK Policy on the Care, Welfare and Treatment of
Laboratory Animals. For all studies, the temperature and humidity
were nominally maintained at 21 °C
2 °C and 55%
10%,
respectively. The diet for rodents was 5LF2 Eurodent diet 14% (PMI
Labdiet, Richmond, IN). There were no known contaminants in the
diet or water at concentrations that could interfere with the outcome
of the studies.
Synthetic Methods and Characterization of Compounds.
Abbreviations for multiplicities observed in NMR spectra are the
following: s; singlet; br s, broad singlet; d, doublet; t, triplet; q,
quadruplet; p, pentuplet; spt, septuplet; m, multiplet. The purity of all
Rat PK Studies. Male CD rats were supplied by Charles River UK
Ltd. and were surgically prepared under anesthesia at GSK with
implanted cannulae in the femoral vein (for drug administration) and
jugular vein (for blood sampling). Each rat was allowed to recover
from surgery for at least 2 days prior to dosing. Rats had free access to
food and water throughout. Where definitive PK was required studies
were conducted using a crossover design with n = 3 animals over two
dosing occasions, with 3 days between dose administrations. Other
data presented are from single animals only. On dosing day 1, n = 3
male rats each received a 1 h intravenous infusion of the test
compound formulated in DMSO and 10% (w/v) Kleptose in saline
(≤5%:≥95%) at a concentration of 0.2 mg/mL. If necessary the pH of
the dose solution was adjusted to 8.0 using 0.1 mM NaOH, and the
dose was filtered using a ∼0.2 μm syringe filter unit. Test compounds
were administered as a 1 h iv infusion at 5 mL kg−1 h−1 to achieve a
target dose of 1 mg/kg. On dosing day 2, the same three rats each
received an oral administration of test compound suspended in 1%
(w/v) methylcellulose aq at a concentration of 0.6 mg/mL
administered by gavage at 5 mL/kg to achieve a target dose of 3
mg/kg. At the end of the study the rats were euthanised by
administration of sodium pentobarbital (Euthatal) through the jugular
vein cannula.
1
compounds was determined by LCMS and H NMR and was always
>95%.
LCMS. UPLC analysis was conducted on an Acquity UPLC BEH
C18 column (50 mm × 2.1 mm, i.d. 1.7 μm packing diameter) at 40
°C. Flow rate was 1 mL/min. For the formate method, solvents
employed were A = 0.1% v/v solution of formic acid in water and B =
0.1% v/v solution of formic acid in acetonitrile. For the high pH
method, the solvents employed were A = 10 mM ammonium
hydrogen carbonate in water adjusted to pH 10 with ammonia
solution and B = acetonitrile. For both methods the gradient employed
is indicated in Table 6.
The UV detection was a summed signal from wavelength of 210 to
350 nm. Mass spectra were obtained on a Waters ZQ instrument:
ionization mode, alternate-scan positive and negative electrospray;
scan range 100−1000 amu; scan time 0.27 s; interscan delay 0.10 s.
MDAP (Mass-Directed Automatic Preparative Purification).
For the formate method, HPLC analysis was conducted on either a
Sunfire C18 column (100 mm × 19 mm, i.d. 5 μm packing diameter)
or a Sunfire C18 column (150 mm × 30 mm, i.d. 5 μm packing
diameter) at ambient temperature. The solvents employed were A =
8246
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
1,1-dimethylethyl 5-methyl-6-oxo-3,4,6,7,8,8a-hexahydro-2(1H)-
isoquinolinecarboxylate, 1,1-dimethylethyl 4-oxo-1-piperidine-
carboxylate (70 g, 350 mmol) and pyrrolidine (43.6 mL, 530 mmol)
were dissolved in toluene (310 mL), and the resulting mixture was
refluxed under Dean−Stark conditions for 24 h and then concentrated
in vacuo. The residue was dissolved in anhydrous toluene (270 mL)
and treated with hydroquinone (0.40 g) and 1-penten-3-one (29.6 g,
350 mmol). The resulting solution was refluxed for 24 h and then
diluted with EtOAc (300 mL). The mixture was washed with HCl (0.5
N in water, 500 mL) and the aqueous phase extracted with EtOAc
(300 mL). The combined organic phases were dried (MgSO4) and
concentrated. Purification of the residue by flash chromatography on a
silica cartridge (1.5 kg) gave 1,1-dimethylethyl 5-methyl-6-oxo-
3,4,6,7,8,8a-hexahydro-2(1H)-isoquinolinecarboxylate (55.2 g, 59.2%)
as pale yellow oil which crystallized on standing. LCMS (method
formate): retention time 1.04 min, [M + H]+ = 266.24. 1H NMR (400
MHz, CDCl3) δ ppm 4.16−4.02 (m, 2H), 3.08−3.01 (m, 1H), 2.77−
2.71 (m, 1H), 2.58−2.49 (m, 3H), 2.39−2.26 (m, 2H), 2.06−2.00 (m,
1H), 1.79 (s, 3H), 1.59−1.52 (m, 1H), 1.49 (s, 9H).
Table 6. Gradient Used in UPLC Analysis
time (min)
flow rate (mL/min)
% A
% B
0
1
1
1
1
99
3
1
97
1.5
1.9
2.0
3
97
0
100
0.1% v/v solution of formic acid in water and B = 0.1% v/v solution of
formic acid in acetonitrile. Run as a gradient over either 15 or 25 min
(extended run) with a flow rate of 20 mL/min (100 mm × 19 mm, i.d.
5 μm packing diameter) or 40 mL/min (150 mm × 30 mm, i.d. 5 μm
packing diameter).
For the high pH method, HPLC analysis was conducted on either
an Xbridge C18 column (100 mm × 19 mm, i.d. 5 μm packing
diameter) or a Xbridge C18 column (100 mm × 30 mm, i.d. 5 μm
packing diameter) at ambient temperature. The solvents employed
were A = 10 mM ammonium bicarbonate in water, adjusted to pH 10
with ammonia solution and B = acetonitrile. Run as a gradient over
either 15 or 25 min (extended run) with a flow rate of 20 mL/min
(100 mm × 19 mm, i.d. 5 μm packing diameter) or 40 mL/min (100
mm × 30 mm, i.d. 5 μm packing diameter).
Step 2. Lithium bis(trimethylsilyl)amide (1 M in THF, 246 mL, 246
mmol) was added dropwise to a solution of 1,1-dimethylethyl 5-
methyl-6-oxo-3,4,6,7,8,8a-hexahydro-2(1H)-isoquinolinecarboxylate
(54.4 g, 210 mmol) in THF (200 mL) at −63 °C, and the mixture was
stirred for an additional 30 min. Chloro(trimethyl)silane (31.4 mL,
250 mmol) was added dropwise, and the resulting mixture was stirred
for 2 h at −70 °C. The reaction was warmed to room temperature
over 20 min and diluted with diethyl ether (800 mL). The reaction
was added to a saturated sodium carbonate solution, and the phases
were separated. The aqueous phase was extracted with diethyl ether
(300 mL), and the combined organic phases were dried (Na2SO4) and
concentrated in vacuo. The residue was dissolved in acetonitrile (200
mL), and palladium(II) acetate (46.0 g, 210 mmol) was added. The
resulting mixture was cooled (water bath) to maintain a reaction
temperature below 35 °C and stirred overnight. The reaction was
filtered through Celite and the residue rinsed with EtOAc (3 × 300
mL). The filtrate was further filtered through a 1 in. pad of silica gel
and concentrated. The residue was dissolved in EtOAc (500 mL),
treated with tetrabutylammonium fluoride (1 M in THF, 200 mL).
The resulting mixture was allowed to stand for 30 min, washed with
HCl (0.5 N in water, 300 mL) and a 10% sodium thiosulfate solution,
dried (MgSO4), and concentrated. Purification of the residue by flash
chromatography on silica gel (300 g), eluting with an EtOAc/
cyclohexane gradient (0−60%), gave 1,1-dimethylethyl 6-hydroxy-5-
methyl-3,4-dihydro-2(1H)-isoquinolinecarboxylate (29.9 g, 55%) as a
white solid. LCMS (method formate): retention time 1.07 min, [M +
For both methods the UV detection was a summed signal from
wavelength of 210 to 350 nm. Mass spectra were obtained on a Waters
ZQ instrument: ionization mode, alternate-scan positive and negative
electrospray; scan range 100−1000 amu; scan time 0.50 s; interscan
delay 0.20 s.
(R)-5-(3-(2-(2,3-Dihydroxypropyl)-5-methyl-1,2,3,4-
tetrahydroisoquinolin-6-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxybenzonitrile (12). To a suspension of L-(−)-glycer-
aldehyde (127 mg, 1.41 mmol) and 2-[(1-methylethyl)oxy]-5-[3-(5-
methyl-1,2,3,4-tetrahydro-6-isoquinolinyl)-1,2,4-oxadiazol-5-yl]-
benzonitrile hydrochloride (as obtained in ref 36, 187 mg, 0.46 mmol)
in CH2Cl2 (10 mL) under nitrogen, was added AcOH (0.039 mL, 0.68
mmol) followed by sodium triacetoxyborohydride (436 mg, 2.06
mmol). The resulting mixture was stirred at room temperature for 20
h. Further sodium triacetoxyborohydride (193 mg, 0.91 mmol) was
added to the mixture, and stirring was continued for a further 72 h.
The solution was then treated with a saturated NaHCO3 aqueous
solution (5 mL), and the resulting biphasic mixture was vigorously
stirred for 30 min and then was concentrated in vacuo. The residue
was partitioned between a saturated NaHCO3 aqueous solution (5
mL) and AcOEt (5 mL), and the layers were separated. The aqueous
phase was extracted with EtOAc (3 × 4 mL). The combined organic
phases were dried using a hydrophobic frit and concentrated in vacuo.
Purification of the residue by mass-directed autopreparative reverse-
phase HPLC gave (R)-5-(3-(2-(2,3-dihydroxypropyl)-5-methyl-
1,2,3,4-tetrahydroisoquinolin-6-yl)-1,2,4-oxadiazol-5-yl)-2-isopropoxy-
benzonitrile (147 mg, 72%) as a pale yellow gum. LCMS (method
1
H]+ = 264.12. H NMR (400 MHz, DMSO-d6) δ ppm 9.03 (s, 1H),
6.75 (d, J = 8.1 Hz, 1H), 6.64 (d, J = 8.1 Hz, 1H), 4.36 (s, 2H), 3.52 (t,
J = 6.1 Hz, 2H), 2.62 (t, J = 6.1 Hz, 2H), 2.01 (s, 3H), 1.41 (s, 9H).
HRMS calculated for C15H22NO3: 264.1600. Found: 264.1605.
Step 3. To a solution of 1,1-dimethylethyl 6-hydroxy-5-methyl-3,4-
dihydro-2(1H)-isoquinolinecarboxylate (3.16 g, 12 mmol) in CH2Cl2
(50 mL) at room temperature under nitrogen was added pyridine
(1.94 mL, 24 mmol), and the resulting solution was cooled to −30 °C
before trifluoromethanesulfonic anhydride (2.23 mL, 13.20 mmol) was
added dropwise. The resulting mixture was stirred for 40 min at this
temperature, then warmed to room temperature and concentrated in
vacuo. The residue was dissolved in EtOAc, and the organic phase was
washed sequentially with HCl (1 N in water), a saturated NaHCO3
aqueous solution, and brine. The solution was then dried (MgSO4)
and concentrated in vacuo to give 1,1-dimethylethyl 5-methyl-6-
{[(trifluoromethyl)sulfonyl]oxy}-3,4-dihydro-2(1H)-isoquinoline-
carboxylate (4.85 g, 102%) as a red oil which was used in the next step
without further purification. LCMS (method high pH): retention time
formate): retention time 0.86 min, [M + H]+ = 449.3. H NMR (400
1
MHz, CDCl3) δ ppm 8.41 (d, J = 2.0 Hz, 1H), 8.34 (dd, J = 9.0, 2.3
Hz, 1H), 7.86 (d, J = 8.0 Hz, 1H), 7.12−7.20 (m, 2H), 4.84 (spt, J =
6.0 Hz, 1H), 4.62−4.77 (m, 3H), 4.48−4.61 (m, 2H), 4.43 (br s, 1H),
3.59−3.87 (m, 3H), 3.40 (d, J = 10.3 Hz, 1H), 3.17−3.32 (m, 3H),
2.51−2.61 (m, 3H), 1.51 (d, J = 6.0 Hz, 6H).
(S)-5-(3-(2-(2,3-Dihydroxypropyl)-5-methyl-1,2,3,4-
tetrahydroisoquinolin-6-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxybenzonitrile (13). (S)-5-(3-(2-(2,3-Dihydroxypropyl)-5-
methyl-1,2,3,4-tetrahydroisoquinolin-6-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxybenzonitrile was obtained as a cream colored gum in an
analogous manner to compound 12, using D-glyceraldehyde: 96 mg,
44% yield. LCMS (method formate): retention time 0.87 min, [M +
H]+ = 449.3. 1H NMR (400 MHz, CDCl3) δ ppm 8.40 (d, J = 2.0 Hz,
1H), 8.33 (dd, J = 8.9, 2.1 Hz, 1H), 7.85 (d, J = 8.0 Hz, 1H), 7.13−
7.22 (m, 2H), 4.83 (spt, J = 6.1 Hz, 1H), 4.51−4.65 (m, 1H), 4.45 (br
s, 1H), 3.59−4.04 (m, 7H), 3.42 (br s, 1H), 3.17−3.35 (m, 3H), 2.55
(s, 3H), 1.48−1.55 (d, J = 6.1 Hz, 6H).
1
1.46 min, [M − H]− = 394.2. H NMR (400 MHz, CDCl3) δ ppm
7.10 (d, J = 8.1 Hz, 1H), 7.02 (d, J = 8.1 Hz, 1H), 4.58 (s, 2H), 3.68 (t,
J = 5.8 Hz, 2H), 2.76 (t, J = 5.8 Hz, 2H), 2.25 (s, 3H), 1.50 (s, 9H).
Step 4.
A solution of 1,1-dimethylethyl 5-methyl-6-
{[(trifluoromethyl)sulfonyl]oxy}-3,4-dihydro-2(1H)-isoquinoline-
carboxylate (26.1 g, 66 mmol) in DMF (200 mL) was degassed for 10
min under vacuum and then flushed with nitrogen. The solution was
5-(3-(2-(1,3-Dihydroxypropan-2-yl)-5-methyl-1,2,3,4-
tetrahydroisoquinolin-6-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxynicotinonitrile Hydrochloride (14). Step 1. To make
8247
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
treated with tetrakis(triphenylphosphine)palladium (7.6 g, 6.6 mmol)
and zinc cyanide (10.1 g, 86 mmol), and the resulting mixture was
stirred at 100 °C under nitrogen for 6 h and cooled to room
temperature. The mixture was filtered, the residue washed with EtOAc,
and most of the solvent evaporated in vacuo. The residue was
dissolved in EtOAc and washed with a saturated sodium hydrogen
carbonate aqueous solution (×2). The combined aqueous phases were
extracted with EtOAc (×2), and the combined organic phases were
washed with brine, dried (MgSO4), and concentrated in vacuo.
Purification of the residue by flash chromatography, eluting with an
EtOAc/cyclohexane gradient, gave 1,1-dimethylethyl 6-cyano-5-
methyl-3,4-dihydro-2(1H)-isoquinolinecarboxylate (16.6 g, 92%) as a
white solid. LCMS (method high pH): retention time 1.24 min, [M +
H]+ = 273.2. 1H NMR (400 MHz, DMSO-d6) δ ppm 7.57 (d, J = 8.1
Hz, 1H), 7.22 (d, J = 8.1 Hz, 1H), 4.56 (s, 2H), 3.59 (t, J = 6.1 Hz,
2H), 2.72 (t, J = 6.1 Hz, 2H), 2.39 (s, 3H), 1.44 (s, 9H). HRMS
calculated for C16H21N2O2: 273.1603. Found: 273.1608.
Step 5. A mixture of 1,1-dimethylethyl 6-cyano-5-methyl-3,4-
dihydro-2(1H)-isoquinolinecarboxylate (16.6 g, 61 mmol), sodium
hydrogen carbonate (30.7 g, 370 mmol), and hydroxylamine
hydrochloride (25.4 g, 370 mmol) in ethanol (250 mL) was refluxed
for 28.5 h and allowed to cool to room temperature. The reaction was
filtered and the residue washed with ethanol. The combined filtrate
and washings were concentrated in vacuo. The residue was poured
into water (100 mL) and stirred at room temperature for 20 min. The
precipitated solid was isolated by filtration and dried under vacuum at
40 °C for 16 h to give 1,1-dimethylethyl 6-[(hydroxyamino)(imino)-
methyl]-5-methyl-3,4-dihydro-2(1H)-isoquinolinecarboxylate (16 g,
86%) as a white solid. LCMS (method high pH): retention time
0.93 min, [M + H]+ = 306.2. 1H NMR (400 MHz, DMSO-d6) δ ppm
9.21 (s, 1H), 7.07 (d, J = 8.9 Hz, 1H), 6.99 (d, J = 8.9 Hz, 1H), 5.65
(s, 2H), 4.48 (s, 2H), 3.58 (t, J = 5.9 Hz, 2H), 2.67 (t, J = 5.9 Hz, 2H),
2.20 (s, 3H), 1.43 (m, 9H). HRMS calculated for C16H24N3O3:
306.1818. Found: 306.1817.
Step 9. 5-Cyano-6-[(1-methylethyl)oxy]-3-pyridinecarbonyl chloride
(660 mg, 2.94 mmol) was added portionwise to a suspension of 1,1-
dimethylethyl 6-[(hydroxyamino)(imino)methyl]-5-methyl-3,4-dihy-
dro-2(1H)-isoquinolinecarboxylate (855 mg, 2.80 mmol) in toluene
(10 mL) and pyridine (10 mL) at room temperature under nitrogen,
and the resulting mixture was stirred at this temperature for 20 min
and then was stirred at 120 °C for 2 h before being cooled to room
temperature and concentrated in vacuo. Purification of the residue via
flash chromatography on silica gel (0−50% EtOAc in cyclohexane)
gave tert-butyl 6-(5-(5-cyano-6-isopropoxypyridin-3-yl)-1,2,4-oxadia-
zol-3-yl)-5-methyl-3,4-dihydroisoquinoline-2(1H)-carboxylate (1.24 g,
86%) as colorless foam. LCMS (method formate): retention time 1.54
1
min, [2M + H]+ = 951; [M − C4H8]+ = 420. H NMR (400 MHz,
CDCl3) δ ppm 9.16 (d, J = 2.3 Hz, 1H), 8.65 (d, J = 2.5 Hz, 1H), 7.79
(d, J = 8.1 Hz, 1H), 7.13 (d, J = 8.1 Hz, 1H), 5.52−5.64 (m, 1H), 4.66
(s, 2H), 3.74 (t, J = 5.8 Hz, 2H), 2.87 (t, J = 5.7 Hz, 2H), 2.54 (s, 3H),
1.53 (s, 9H), 1.49 (d, J = 6.3 Hz, 6H).
Step 10. A solution of 1,1-dimethylethyl 6-(5-{5-cyano-6-[(1-
methylethyl)oxy]-3-pyridinyl}-1,2,4-oxadiazol-3-yl)-5-methyl-3,4-dihy-
dro-2(1H)-isoquinolinecarboxylate (1.12 g, 2.35 mmol) in CH2Cl2 (5
mL) at 0 °C was treated with trifuoroacetic acid (0.91 mL, 11.8 mmol)
dropwise. The resulting mixture was allowed to warm to room
temperature and was stirred at this temperature for 16 h. Most of the
solvent was removed in vacuo and the residue was triturated with Et2O
(2 × 5 mL) and then filtered off to give 2-[(1-methylethyl)oxy]-5-[3-
(5-methyl-1,2,3,4-tetrahydro-6-isoquinolinyl)-1,2,4-oxadiazol-5-yl]-3-
pyridinecarbonitrile trifluoroacetate (1.1 g, 96%) as a colorless solid.
LCMS (method formate): retention time 0.96 min, [M + H]+ = 376.
1H NMR (400 MHz, DMSO-d6) δ ppm 9.21 (d, J = 2.5 Hz, 1H), 9.13
(br s, 2H), 8.99 (d, J = 2.5 Hz, 1H), 7.79 (d, J = 8.1 Hz, 1H), 7.30 (d, J
= 8.1 Hz, 1H), 5.51 (spt, J = 6.2 Hz, 1H), 4.38 (br s, 2H), 3.49 (br s,
2H), 2.99 (t, J = 6.2 Hz, 2H), 2.48 (s, 3H), 1.42 (d, J = 6.1 Hz, 6H).
Step 11. 2,2-Dimethyl-1,3-dioxan-5-one (199 mg, 1.53 mmol) was
added to a stirred solution of 2-[(1-methylethyl)oxy]-5-[3-(5-methyl-
1,2,3,4-tetrahydro-6-isoquinolinyl)-1,2,4-oxadiazol-5-yl]-3-pyridine-
carbonitrile trifluoroacetate (150 mg, 0.31 mmol) in 1,2-dichloro-
ethane (2 mL) and THF (2 mL). The reaction mixture was stirred at
room temperature for 30 min, and then sodium triacetoxyborohydride
(325 mg, 1.53 mmol) was added. Stirring at room temperature was
continued for 4 h. Further portions of 2,2-dimethyl-1,3-dioxan-5-one
(199 mg, 1.53 mmol) and sodium triacetoxyborohydride (325 mg,
1.53 mmol) were added, and stirring continued for 16 h at room
temperature. Further portions of 2,2-dimethyl-1,3-dioxan-5-one (199
mg, 1.53 mmol) and sodium triacetoxyborohydride (325 mg, 1.53
mmol) were added, and stirring was continued for 6 h. A saturated
NaHCO3 aqueous solution (10 mL) was added and the mixture
extracted with EtOAc (2 × 10 mL). The combined extracts were
washed with water and brine, dried over MgSO4, and concentrated in
vacuo. Purification of the residue by flash chromatography on silica gel
(10−40% EtOAc in isohexane] gave 5-(3-(2-(2,2-dimethyl-1,3-dioxan-
5-yl)-5-methyl-1,2,3,4-tetrahydroisoquinolin-6-yl)-1,2,4-oxadiazol-5-
yl)-2-isopropoxynicotinonitrile (30 mg, 20%) as a colorless solid after
trituration with Et2O. LCMS (method formate): retention time 1.04
min, [M + H]+ = 490. 1H NMR (400 MHz, DMSO-d6) δ ppm 9.20 (s,
1H), 8.99 (s, 1H), 7.66 (d, J = 3.0 Hz, 1H), 7.11 (d, J = 3.0 Hz, 1H),
5.50−5.53 (m, 1H), 3.97−4.00 (m, 2H), 3.81−3.85 (m, 3H), 2.89−
2.91 (m, 2H), 2.73−2.76 (m, 2H), 2.66−2.68 (m, 2H), 2.43 (s, 3H),
1.41 (d, J = 7.8 Hz, 6H), 1.28−1.32 (m, 6H).
Step 6. A mixture of methyl 5-cyano-6-oxo-1,6-dihydro-3-pyridine-
carboxylate (2.98 g, 16.7 mmol), isopropyl iodide (6.68 mL, 66.9
mmol), and silver carbonate (9.23 g, 33.5 mmol) in CHCl3 (100 mL)
was stirred at 60 °C for 4 h under nitrogen and then was cooled to
room temperature and filtered using a pad of Celite. Most of the
solvent was removed in vacuo. Purification of the residue by flash
chromatography on silica gel (0−30% EtOAc in cyclohexane) gave
methyl 5-cyano-6-isopropoxynicotinate (3.2 g, 87%) as a colorless
solid. LCMS (method formate): retention time 1.08 min, [M + H]+ =
1
220.0. H NMR (600 MHz, CDCl3) δ ppm 8.97 (d, J = 2.2 Hz, 1H),
8.47 (d, J = 2.2 Hz, 1H), 5.54 (spt, J = 6.2 Hz, 1H), 3.96 (s, 3H), 1.45
(d, J = 6.2 Hz, 6H).
Step 7. A solution of LiOH (1.74 g, 72.7 mmol) in water (10 mL)
was added to a solution of methyl 5-cyano-6-[(1-methylethyl)oxy]-3-
pyridinecarboxylate (3.2 g, 14.5 mmol) in MeOH (30 mL), and the
resulting mixture was stirred at room temperature for 4 h. Most of
MeOH was removed in vacuo, and the residual aqueous phase was
acidified to pH 1 with a 2 N HCl aqueous solution and then was
extracted with EtOAc (3 × 50 mL). The combined organic phases
were washed with brine (100 mL), dried over MgSO4, and
concentrated in vacuo to give 5-cyano-6-isopropoxynicotinic acid
(3.05 g, 95%) as a colorless solid. LCMS (method formate): retention
1
time 0.90 min, [M + H]+ = 206. H NMR (400 MHz, DMSO-d6) δ
ppm 13.48 (br s, 1H), 8.92 (d, J = 2.3 Hz, 1H), 8.61 (d, J = 2.5 Hz,
1H), 5.45 (spt, J = 6.2 Hz, 1H), 1.37 (d, J = 6.2 Hz, 6H).
Step 12. A 2 N HCl aqueous solution (2 mL) was added to a
solution of 5-{3-[2-(2,2-dimethyl-1,3-dioxan-5-yl)-5-methyl-1,2,3,4-
tetrahydro-6-isoquinolinyl]-1,2,4-oxadiazol-5-yl}-2-[(1-methylethyl)-
oxy]-3-pyridinecarbonitrile (30 mg, 0.06 mmol) in THF (2 mL). The
resulting mixture was stirred at room temperature for 2 h and then was
diluted with EtOAc (10 mL) and basified by the addition of a
saturated NaHCO3 aqueous solution. The aqueous phase was
separated and extracted with EtOAc (10 mL). The combined organics
were dried over MgSO4 and concentrated in vacuo. The residue was
dissolved in THF and treated with a 4 N HCl solution in 1,4-dioxan (1
mL, excess). The solvent was concentrated in vacuo. Trituration of the
residue with Et2O gave 14 (20 mg, 73%) as a colorless solid. LCMS
Step 8. Oxalyl chloride (1.94 mL, 22.1 mmol) was added at room
temperature to a suspension of 5-cyano-6-[(1-methylethyl)oxy]-3-
pyridinecarboxylic acid (3.04 g, 14.7 mmol) in CH2Cl2 (50 mL),
followed by DMF (0.011 mL, 0.15 mmol), and the resulting mixture
was stirred at this temperature for 3 h. The resulting homogeneous,
pale yellow solution was concentrated in vacuo and the residue was
azeotroped with toluene (3 × 10 mL) to give 5-cyano-6-isopropoxy-
nicotinoyl chloride (3.5 g, 95%) as a pale yellow oil which solidified on
standing under vacuum. 1H NMR (400 MHz, CDCl3) δ ppm 9.06 (d,
J = 2.5 Hz, 1H), 8.53 (d, J = 2.5 Hz, 1H), 5.59 (spt, J = 6.1 Hz, 1H),
1.47 (d, J = 6.3 Hz, 6H).
8248
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
(method formate): retention time 0.91 min, [M + H]+ = 450. H
NMR (400 MHz, DMSO-d6) δ ppm 10.32−10.45 (m, 1H), 9.21 (d, J
= 2.0 Hz, 1H), 8.99 (d, J = 2.0 Hz, 1H), 7.79 (d, J = 8.1 Hz, 1H), 7.29
(d, J = 8.1 Hz, 1H), 5.41−5.57 (m, 3H), 4.68−4.81 (m, 1H), 4.52−
4.65 (m, 1H), 3.90 (br s, 5H), 3.49−3.63 (m, 1H), 3.14 (br s, 2H),
2.47 (s, 3H), 1.42 (d, J = 6.1 Hz, 6H).
2H), 7.37 (d, J = 7.8 Hz, 1H), 7.35 (br s, 2H), 4.83−4.89 (m, 1H),
1.35 (d, J = 8.0 Hz, 6H).
1
Step 3. A solution of 5-(5-amino-1,3,4-thiadiazol-2-yl)-2-[(1-
methylethyl)oxy]benzonitrile (6.36 g, 24.4 mmol) in CH3CN (60
mL) was treated with copper(II) bromide (10.9 g, 48.9 mmol) and
isoamyl nitrite (6.53 mL, 48.9 mmol), and the resulting mixture was
stirred at room temperature for 1 h and then was concentrated in
vacuo. The residue was dissolved in EtOAc, and the organic phase was
filtered through a pad of Celite. The insoluble material was washed
with EtOAc and the combined organics were washed with a 2 N HCl
aqueous solution (100 mL), then with brine (100 mL), dried over
MgSO4, and concentrated in vacuo to give 5-(5-bromo-1,3,4-
thiadiazol-2-yl)-2-isopropoxybenzonitrile (4.98 g, 63%) as a yellow
solid. LCMS (method formate): retention time 1.14 min, [M + H]+ =
324, 326 (1 Br). 1H NMR (400 MHz, DMSO-d6) δ ppm 8.33 (s, 1H),
8.22 (dd, J = 8.0, 3.0 Hz, 1H), 7.48 (d, J = 7.8 Hz, 1H), 4.91−4.96 (m,
1H), 1.35 (d, J = 7.8 Hz, 6H).
Step 4. A solution of 1,1-dimethylethyl 5-methyl-6-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)-3,4-dihydro-2(1H)-isoquinoline-
carboxylate (300 mg, 0.80 mmol) and 5-(5-bromo-1,3,4-thiadiazol-2-
yl)-2-[(1-methylethyl)oxy]benzonitrile (313 mg, 0.96 mmol) in 1,2-
dimethoxyethane (6 mL) and water (2 mL) was treated with Na2CO3
(426 mg, 4.02 mmol) and Pd(dppf)Cl2·CH2Cl2 (1.13 g, 1.39 mmol).
The resulting mixture was refluxed under nitrogen for 3 h and then
was cooled to room temperature, and most of the solvent was removed
in vacuo. The residue was partitioned between EtOAc (100 mL) and
water (100 mL), and the biphasic mixture was filtered through a pad of
Celite and the insoluble washed with EtOAc. The combined layers
were separated, and the organic phase was washed with water (100
mL), dried over MgSO4, and concentrated in vacuo. Purification of the
residue via flash chromatography on silica gel (5−30% EtOAc in
hexanes) gave tert-butyl 6-(5-(3-cyano-4-isopropoxyphenyl)-1,3,4-
thiadiazol-2-yl)-5-methyl-3,4-dihydroisoquinoline-2(1H)-carboxylate
(394 mg, 71%) as a pale yellow foam. LCMS (method formate):
retention time 1.46 min, [M + H]+ = 491. 1H NMR (400 MHz,
CDCl3) δ ppm 8.25 (d, J = 8.3 Hz, 1H), 8.18 (s, 1H), 7.46 (d, J = 8.1
Hz, 1H), 7.30−7.35 (m, 1H), 7.11 (d, J = 8.6 Hz, 1H), 4.74−4.84 (m,
1H), 4.66 (s, 2H), 3.74 (t, J = 5.6 Hz, 2H), 2.86 (br s, 2H), 2.48 (s,
3H), 1.52 (s, 9H), 1.48 (d, J = 6.1 Hz, 6H).
5-(3-(7-(2,2-Dimethyl-1,3-dioxan-5-yl)-4-methyl-5,6,7,8-tet-
rahydro-1,7-naphthyridin-3-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxybenzonitrile (15). Step 1. 5-(3-(7-(2,2-Dimethyl-1,3-
dioxan-5-yl)-4-methyl-5,6,7,8-tetrahydro-1,7-naphthyridin-3-yl)-1,2,4-
oxadiazol-5-yl)-2-isopropoxybenzonitrile was obtained from 2-isopro-
poxy-5-(3-(4-methyl-5,6,7,8-tetrahydro-1,7-naphthyridin-3-yl)-1,2,4-
oxadiazol-5-yl)benzonitrile (synthesis published in ref 38) using a
similar procedure as described in step 11 for compound 14. LCMS
1
(method formate): retention time 0.91 min, [M + H]+ = 490.2. H
NMR (400 MHz, DMSO-d6) δ ppm 8.80 (s, 1H), 8.52 (s, 1H), 8.41
(dd, J = 9.0, 2.3 Hz, 1H), 7.56 (d, J = 9.0 Hz, 1H), 4.96−5.01 (m,1H),
4.01−4.04 (m, 2H), 3.83−3.87 (m, 4H), 3.35 (s, 1H), 2.92−2.95 (m,
2H), 2.74−2.80 (m, 2H), 2.49 (s, 3H), 1.34−1.41 (m, 12H).
Step 2. 15 was obtained as a white foam from 5-(3-(7-(2,2-dimethyl-
1,3-dioxan-5-yl)-4-methyl-5,6,7,8-tetrahydro-1,7-naphthyridin-3-yl)-
1,2,4-oxadiazol-5-yl)-2-isopropoxybenzonitrile using a similar proce-
dure to the one described in step 12 for compound 14: 62 mg, 18%
yield (2 steps). LCMS (method formate): retention time 0.80 min, [M
1
+ H]+ = 450.1. H NMR (400 MHz, DMSO-d6) δ ppm 10.80 (br s,
1H), 8.93 (s, 1H), 8.56 (d, J = 2.3 Hz, 1H), 8.44 (dd, J = 9.0, 2.3 Hz,
1H), 7.59 (d, J = 9.0 Hz, 1H), 5.92−6.91 (br s, 2H), 5.00 (spt, J = 6.0
Hz, 1H), 4.85 (dd, J = 16.1, 8.0 Hz, 1H), 4.56 (d, J = 15.8 Hz, 1H),
3.86−4.07 (m, 4H), 3.47−3.70 (m, 2H), 3.08−3.33 (m, 2H), 2.56 (s,
3H), 1.40 (d, J = 6.0 Hz, 6H).
5-(5-(2-(1,3-Dihydroxypropan-2-yl)-5-methyl-1,2,3,4-
tetrahydroisoquinolin-6-yl)-1,3,4-thiadiazol-2-yl)-2-
isopropoxybenzonitrile Hydrochloride (16). Step 1. To make tert-
butyl 5-methyl-6-(4,4,5-trimethyl-1,3,2-dioxaborolan-2-yl)-3,4-dihy-
droisoquinoline-2(1H)-carboxylate, a solution of 1,1-dimethylethyl 5-
methyl-6-{[(trifluoromethyl)sulfonyl]oxy}-3,4-dihydro-2(1H)-
isoquinolinecarboxylate (step 3, compound 14, 0.791 g, 2 mmol) in
1,4-dioxane (10 mL) was degassed under vacuum for 10 min and then
was treated with PdCl2(dppf) (0.146 g, 0.200 mmol), potassium
acetate (0.785 g, 8.00 mmol), and bis(pinacolato)diboron (0.609 g,
2.400 mmol). The resulting mixture was refluxed for 5 h and then was
cooled to room temperature, and most of the solvent was removed in
vacuo. The residue was partitioned between EtOAc and water. The
biphasic mixture was filtered through a pad of Celite, and then the
layers were separated. The aqueous phase was extracted with EtOAc,
and the combined organics were washed with brine, dried over
MgSO4, and concentrated in vacuo. Purification of the residue by flash
chromatography on silica gel (0−10% MeOH in CH2Cl2) gave tert-
butyl 5-methyl-6-(4,4,5-trimethyl-1,3,2-dioxaborolan-2-yl)-3,4-
dihydroisoquinoline-2(1H)-carboxylate (660 mg, 88%) as a pale
yellow solid. LCMS (method formate): retention time 1.56 min, [M +
H]+ = 374 (weak), [2M + H]+ = 747. 1H NMR (400 MHz, CDCl3) δ
ppm 7.62 (d, J = 7.8 Hz, 1H), 6.97 (d, J = 7.8 Hz, 1H), 4.54−4.58 (br
s, 2H), 3.61−3.70 (m, 2H), 2.70−2.80 (m, 2H), 2.48 (s, 3H), 1.50 (s,
9H), 1.38 (s, 12H).
Step 5. A solution of 1,1-dimethylethyl 6-(5-{3-cyano-4-[(1-
methylethyl)oxy]phenyl}-1,3,4-thiadiazol-2-yl)-5-methyl-3,4-dihydro-
2(1H)-isoquinolinecarboxylate (912 mg, 1.86 mmol) in CH2Cl2 (10
mL) was treated with trifluoroacetic acid (5.00 mL, 64.9 mmol) at
room temperature. The resulting mixture was stirred for 30 min and
then was concentrated in vacuo. The residue obtained was
subsequently taken up twice in toluene and concentrated in vacuo,
then was triturated with Et2O and filtered off to give 2-isopropoxy-5-
(5-(5-methyl-1,2,3,4-tetrahydroisoquinolin-6-yl)-1,3,4-thiadiazol-2-yl)-
benzonitrile trifluoroacetate (808 mg, 86%) as a brown solid which
was used in the next step without further purification. LCMS (method
1
formate): retention time 0.82 min, [M + H]+ = 391. H NMR (400
MHz, DMSO-d6) δ ppm 9.20 (br s, 2H), 8.38 (s, 1H), 8.30 (br s, 1H),
7.56 (d, J = 8.1 Hz, 1H), 7.50 (d, J = 9.1 Hz, 1H), 7.28 (d, J = 8.1 Hz,
1H), 4.87−5.03 (m, 1H), 4.38 (br s, 2H), 3.48 (br s, 2H), 2.99 (br s,
2H), 2.39 (s, 3H), 1.38 (d, J = 5.8 Hz, 6H).
Step 2. Phosphoric trichloride (90 g, 587 mmol) was added to a
mixture of 3-cyano-4-[(1-methylethyl)oxy]benzoic acid (20.9 g, 102
mmol) and hydrazinecarbothioamide (13.9 g, 153 mmol), and the
resulting mixture was stirred at 90 °C for 3 h, then was cooled to room
temperature and slowly poured into a ice-cooled 5 M NaOH aqueous
solution (600 mL), making sure that the temperature was maintained
below 35 °C. The solution was cooled to 0 °C, and concentrated
aqueous HCl was slowly added to adjust the pH to 10. The beige solid
formed was filtered off and then dissolved in CH2Cl2 (1 L) and MeOH
(50 mL). The organic phase was washed with water (500 mL), dried
over MgSO4, and concentrated in vacuo to give 5-(5-amino-1,3,4-
thiadiazol-2-yl)-2-isopropoxybenzonitrile (26.5 g, 99% yield) as a pale
yellow solid which was used in the next step without further
purification. LCMS (method formate): retention time 0.82 min, [M +
Step 6. A solution of 2-[(1-methylethyl)oxy]-5-[5-(5-methyl-1,2,3,4-
tetrahydro-6-isoquinolinyl)-1,3,4-thiadiazol-2-yl]benzonitrile trifluor-
oacetate (795 mg, 1.58 mmol) and 2,2-dimethyl-1,3-dioxan-5-one
(0.56 mL, 4.73 mmol) in CH2Cl2 (20 mL) was treated with sodium
triacetoxyborohydride (1.67 g, 7.88 mmol) portionwise. The resulting
mixture was stirred for 18 h at room temperature and then was treated
with a saturated NaHCO3 aqueous solution. The resulting mixture was
stirred for 15 min, and then the layers were separated. The aqueous
layer was extracted three times with CH2Cl2. The combined organic
layers were washed with brine, dried over MgSO4, and concentrated in
vacuo. Purification of the residue via flash chromatography on silica gel
(2−5% MeOH in CH2Cl2) gave 5-(5-(2-(2,2-dimethyl-1,3-dioxan-5-
yl)-5-methyl-1,2,3,4-tetrahydroisoquinolin-6-yl)-1,3,4-thiadiazol-2-yl)-
2-isopropoxybenzonitrile (824 mg, 99%) as a yellow solid. LCMS
1
1
H]+ = 261. H NMR (400 MHz, DMSO-d6) δ ppm 7.99−8.04 (m,
(method formate): retention time 0.89 min, [M + H]+ = 505. H
8249
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
NMR (400 MHz, CDCl3) δ ppm 8.08−8.14 (m, 2H), 7.33 (d, J = 8.1
Hz, 1H), 7.05 (d, J = 9.1 Hz, 1H), 6.94 (d, J = 8.1 Hz, 1H), 4.71 (spt, J
= 6.1 Hz, 1H), 4.03 (dd, J = 11.9, 5.1 Hz, 1H), 3.98 (dd, J = 12.1, 2.5
Hz, 1H), 3.88 (dd, J = 11.9, 7.8 Hz, 1H), 3.79 (s, 2H), 3.67 (dd, J =
12.0, 3.9 Hz, 1H), 3.49−3.55 (m, 1H), 2.87−2.93 (m, 2H), 2.77 (t, J =
6.1 Hz, 2H), 2.36 (s, 3H), 1.35−1.42 (m, 12H).
CH2Cl2 (1 L), and the resulting mixture was allowed to warm to room
temperature over 2 h and then was washed with water (500 mL) and
brine (200 mL), dried over MgSO4, and concentrated in vacuo.
Purification of the residue by flash chromatography on silica gel (750 g
column, 0−30% EtOAc in cyclohexane) gave 1-(chloromethyl)-3-
methoxy-2-methylbenzene (35.2 g, 77%) as pale yellow oil. LCMS
1
(method formate): retention time 1.17 min, [M + H]+ = 171. H
Step 7. A solution of 5-{5-[2-(2,2-dimethyl-1,3-dioxan-5-yl)-5-
methyl-1,2,3,4-tetrahydro-6-isoquinolinyl]-1,3,4-thiadiazol-2-yl}-2-[(1-
methylethyl)oxy]benzonitrile (824 mg, 1.63 mmol) in THF (15 mL)
was treated with a 2 N HCl aqueous solution (15 mL), and the
resulting mixture was stirred at room temperature for 2 h and then was
concentrated in vacuo and the residue coevaporated twice with
toluene. The residue obtained was triturated with Et2O and the
precipitate formed was filtered off and dried under vacuum to give 16
(502 mg, 60%) as a light yellow solid. LCMS (method formate):
NMR (400 MHz, CDCl3) δ ppm 7.19 (t, J = 8.0 Hz, 1H), 6.97 (d, J =
7.6 Hz, 1H), 6.88 (d, J = 8.1 Hz, 1H), 4.64 (s, 2H), 3.79−3.93 (m,
3H), 2.32 (s, 3H).
Step 3. A solution of 1-(chloromethyl)-2-methyl-3-(methyloxy)-
benzene (35 g, 205 mmol) in DMF (200 mL) at room temperature
was treated with sodium cyanide (17 g, 347 mmol), and the resulting
mixture was stirred at 90 °C for 4 days, then was cooled to room
temperature and diluted with water (1 L). The aqueous phase was
extracted with Et2O (1 L) and the organic phase was washed with
water (500 mL), dried over MgSO4, and concentrated in vacuo to give
2-(3-methoxy-2-methylphenyl)acetonitrile (28.5 g, 86%) as a pale
yellow oil which was used in the next step without further purification.
LCMS (method formate): retention time 0.96 min, [M + H]+ = 162.
1H NMR (400 MHz, CDCl3) δ ppm 7.21 (t, J = 8.0 Hz, 1H), 6.99 (d,
J = 7.6 Hz, 1H), 6.87 (d, J = 8.3 Hz, 1H), 3.86 (s, 3H) 3.69 (s, 2H)
2.13−2.31 (m, 3H).
1
retention time 0.80 min, [M + H]+ = 465. H NMR (400 MHz,
DMSO-d6) δ ppm 10.29 (br s, 1H), 8.39 (d, J = 2.3 Hz, 1H), 8.31 (dd,
J = 8.8, 2.3 Hz, 1H), 7.57 (d, J = 8.1 Hz, 1H), 7.50 (d, J = 9.3 Hz, 1H),
7.28 (d, J = 8.1 Hz, 1H), 5.43−5.56 (m, 2H), 4.94 (s, 1H), 4.67−4.82
(m, 1H), 4.53−4.65 (m, 1H), 3.90 (br s, 4H), 3.52−3.64 (m, 1H),
3.30 (d, J = 5.6 Hz, 2H), 3.13 (br s, 2H), 2.39 (s, 3H), 1.38 (d, J = 6.1
Hz, 6H).
5-(3-(2-(1,3-Dihydroxypropan-2-yl)-1,2,3,4-tetrahydro-
isoquinolin-5-yl)-1,2,4-oxadiazol-5-yl)-2-isopropoxy-
benzonitrile Hydrochloride (17). Step 1. 5-(3-(2-(2,2-Dimethyl-1,3-
dioxan-5-yl)-1,2,3,4-tetrahydroisoquinolin-5-yl)-1,2,4-oxadiazol-5-yl)-
2-isopropoxybenzonitrile was obtained as a yellow solid from 2-
isopropoxy-5-(3-(1,2,3,4-tetrahydroisoquinolin-5-yl)-1,2,4-oxadiazol-5-
yl)benzonitrile (synthesis described in ref 39) using a similar
procedure as described in step 11 for compound 14. LCMS (method
Step 4. A solution of [2-methyl-3-(methyloxy)phenyl]acetonitrile
(28 g, 174 mmol) in EtOH (200 mL) was treated with a 5 N NaOH
aqueous solution (174 mL, 868 mmol), and the resulting mixture was
heated at reflux for 2 days and then was cooled to room temperature
and concentrated in vacuo to half its volume. The residue was diluted
with water (100 mL), and the aqueous phase was washed with Et2O
(200 mL) and then was acidified with a 5 N HCl aqueous solution to
pH 2. The resulting suspension was extracted with CH2Cl2 (2 × 200
mL). The combined organics were washed with water (100 mL), dried
over MgSO4, and concentrated in vacuo to give 2-(3-methoxy-2-
methylphenyl)acetic acid (30.3 g, 97% yield) as a white solid. LCMS
1
formate): retention time 0.94 min, [M + H]+ = 475.2. H NMR (400
MHz, CDCl3) δ ppm 8.41 (br s, 1H), 8.32 (d, J = 8.1 Hz, 1H), 7.98
(d, J = 6.8 Hz, 1H), 7.25−7.33 (m, 1H), 7.20 (d, J = 6.6 Hz, 1H), 7.12
(d, J = 9.1 Hz, 1H), 4.72−4.85 (m, 1H), 4.01−4.15 (m, 1H), 3.91−
3.99 (m, 1H), 3.88 (br s, 2H), 3.75 (d, J = 9.3 Hz, 1H), 3.26 (br s,
2H), 2.94 (br s, 2H), 2.83 (br s, 2H), 1.36−1.54 (m, 12H).
Step 2. 17 was obtained as a white solid from 5-(3-(2-(2,2-dimethyl-
1,3-dioxan-5-yl)-1,2,3,4-tetrahydroisoquinolin-5-yl)-1,2,4-oxadiazol-5-
yl)-2-isopropoxybenzonitrile using a similar procedure as the one
described in step 12 for compound 14: 233 mg, 51% yield (2 steps).
LCMS (method formate): retention time 0.84 min, [M + H]+ = 435.2.
1H NMR (400 MHz, DMSO-d6) δ ppm 10.37 (br s, 1H), 8.51−8.55
1
(method formate): retention time 0.83 min, [M + H]+ = 181. H
NMR (400 MHz, CDCl3) δ ppm 7.16 (t, J = 8.0 Hz, 1H), 6.84 (t, J =
7.6 Hz, 2H), 3.79−3.92 (m, 3H), 3.70 (s, 2H), 2.20 (s, 3H). COOH
was not observed.
Step 5. A solution of [2-methyl-3-(methyloxy)phenyl]acetic acid
(30.2 g, 168 mmol) in CH2Cl2 (300 mL) was treated with oxalyl
chloride (29.3 mL, 335 mmol) and then with DMF (0.2 mL, catalytic),
and the resulting mixture was stirred at room temperature for 2 h until
the effervescence ceased. Most of the solvent was evaporated in vacuo
to give a dark purple liquid which was dissolved in CH2Cl2 (100 mL)
and added dropwise at room temperature to a solution of 2,2-
dimethoxyethanamine (19.38 g, 184 mmol) and NEt3 (30.4 mL, 218
mmol) in CH2Cl2 (200 mL). The resulting mixture was stirred at this
temperature for 1 h, then was washed successively with water (200
mL), a 10% aqueous AcOH solution (200 mL), and a saturated
NaHCO3 aqueous solution (300 mL), then was dried over MgSO4 and
concentrated in vacuo to give N-(2,2-dimethoxyethyl)-2-(3-methoxy-
2-methylphenyl)acetamide (43.9 g, 98% yield) as a beige crystalline
solid. LCMS (method formate): retention time 0.81 min, [M + H]+ =
(m, 1H), 8.39−8.44 (m, 1H), 8.03−8.09 (m, 1H), 7.55−7.59 (m, 1H),
7.47−7.55 (m, 2H), 5.45−5.58 (m, 2H), 4.93−5.04 (m, 1H), 4.72−
4.83 (m, 1H), 4.58−4.69 (m, 1H), 3.83−3.99 (m, 5H), 3.45−3.60 (m,
4H), 1.39 (d, J = 6.1 Hz, 6H).
(R)-5-(3-(3-(2,3-Dihydroxypropyl)-6-methyl-2,3,4,5-tetrahy-
dro-1H-benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxybenzonitrile Hydrochloride (18). Step 1. To make (3-
methoxy-2-methylphenyl)methanol, to a solution of 2-methyl-3-
(methyloxy)benzoic acid (24.9 g, 150 mmol) in THF (300 mL) at
0 °C under nitrogen was slowly added borane−tetrahydrofuran
complex (1 M in THF, 308 mL, 308 mmol) via canula over 15 min.
After 10 min, the mixture was allowed to warm to room temperature
and stirred for 3 h. The solution was then cooled to 0 °C and slowly
treated with a 2 N HCl aqueous solution (200 mL) so that the
temperature was kept below 20 °C (CAUTION: formation of
hydrogen!). Most of the THF was removed in vacuo, and the residue
was partitioned between EtOAc and brine. The layers were separated,
and the aqueous phase was extracted twice with EtOAc. The combined
organic phases were washed with brine, dried over MgSO4, and
concentrated in vacuo to give (3-methoxy-2-methylphenyl)methanol
(22.8 g, 95%) as a brown solid used as it was in the next step without
further purification. LCMS (method formate): retention time 0.79
min, [M + H]+ = 153. 1H NMR (400 MHz, CDCl3) δ ppm 7.15−7.24
(m, 1H), 7.01 (d, J = 7.6 Hz, 1H), 6.85 (d, J = 8.3 Hz, 1H), 4.72 (s,
2H), 3.85 (s, 3H), 2.24 (s, 3H). OH was not observed.
1
268. H NMR (400 MHz, CDCl3) δ ppm 7.17 (t, J = 8.0 Hz, 1H),
6.82 (t, J = 8.2 Hz, 2H), 5.49−5.63 (m, 1H), 4.31 (t, J = 5.4 Hz, 1H),
3.85 (s, 3H), 3.61 (s, 2H), 3.28−3.39 (m, 8H), 2.16 (s, 3H).
Step 6. N-[2,2-Bis(methyloxy)ethyl]-2-[2-methyl-3-(methyloxy)-
phenyl]acetamide (44 g, 165 mmol) was dissolved in a mixture of
concentrated aqueous HCl (100 mL, 3.29 mol) and acetic acid (100
mL, 1.75 mol), and the resulting solution was stirred at room
temperature for 3 h, then added to ice (500 g). The aqueous phase was
extracted with CH2Cl2 (1 L). The organic phase was washed with
water and brine, dried over MgSO4, and concentrated in vacuo to give
8-methoxy-9-methyl-1H-benzo[d]azepin-2(3H)-one (25.4 g, 76%) as
a dark brown solid which was used in the next step without further
purification. LCMS (method formate): retention time 0.84 min, [M +
1
H]+ = 204. H NMR (400 MHz, DMSO-d6) δ ppm 9.52 (br s, 1H),
Step 2. Methanesulfonyl chloride (23.0 mL, 295 mmol) was added
dropwise at 0 °C to a solution of [2-methyl-3-(methyloxy)phenyl]-
methanol (51 g, 268 mmol) and NEt3 (44.8 mL, 322 mmol) in
7.11 (d, J = 8.6 Hz, 1H), 6.93 (d, J = 8.6 Hz, 1H), 6.25−6.34 (m, 1H),
6.16−6.25 (m, 1H), 3.79 (s, 3H), 3.35 (s, 2H), 2.23 (s, 3H).
8250
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
Step 7. A solution of 9-methyl-8-(methyloxy)-1,3-dihydro-2H-3-
benzazepin-2-one (24 g, 118 mmol) in MeOH (1.2 L) was treated
with palladium on carbon (10% w/w, 50% wet; 2 g), and the resulting
mixture was stirred under an atmosphere of hydrogen for 16 h. The
catalyst was filtered off under nitrogen and washed with MeOH. The
combined organics were concentrated in vacuo to give 8-methoxy-9-
methyl-4,5-dihydro-1H-benzo[d]azepin-2(3H)-one (23.8 g, 98%) as
pale green solid which was used in the next step without further
purification. LCMS (method formate): retention time 0.76 min, [M +
eluting with 0−50% EtOAc in cyclohexane) gave tert-butyl 6-methyl-7-
(((trifluoromethyl)sulfonyl)oxy)-4,5-dihydro-1H-benzo[d]azepine-
3(2H)-carboxylate (15.0 g, 85%) as a colorless solid. LCMS (method
1
formate): retention time 1.45 min, [M + H]+ = 410. H NMR (400
MHz, CDCl3) δ ppm 7.05 (s, 2H), 3.59 (br s, 4H), 2.84−3.06 (m,
4H), 2.33 (s, 3H), 1.45 (s, 9H).
Step 11.
A solution of 1,1-dimethylethyl 6-methyl-7-
{[(trifluoromethyl)sulfonyl]oxy}-1,2,4,5-tetrahydro-3H-3-benzaze-
pine-3-carboxylate (13 g, 32 mmol) in DMF (150 mL) was degassed
using vacuum using nitrogen (3 purges), and then zinc cyanide (4.85
g, 41.3 mmol) and palladium tetrakis (3.67 g, 3.18 mmol) were added
and the resulting mixture was stirred under nitrogen at 120 °C for 6 h
and then was cooled to room temperature. The mixture was diluted
with EtOAc (400 mL) and filtered through a pad of Celite. The
insolubles were rinsed with EtOAc, and the combined organics were
washed with water (2 × 200 mL), dried over MgSO4, and
concentrated in vacuo. Purification of the residue by flash
chromatography on silica gel (330 g silica column, eluting with 0−
50% EtOAc/cyclohexane) gave tert-butyl 7-cyano-6-methyl-4,5-dihy-
dro-1H-benzo[d]azepine-3(2H)-carboxylate (8.6 g, 95% yield) as a
white solid. LCMS (method formate): retention time 1.21 min, [M +
1
H]+ = 206. H NMR (400 MHz, CDCl3) δ ppm 6.99 (d, J = 8.2 Hz,
1H), 6.73 (d, J = 8.2 Hz, 1H), 6.01 (br s, 1H), 3.89 (s, 2H), 3.82 (s,
3H), 3.43−3.53 (m, 2H), 3.05−3.17 (m, 2H), 2.29 (s, 3H).
Step 8. A solution of 9-methyl-8-(methyloxy)-1,3,4,5-tetrahydro-2H-
3-benzazepin-2-one (21.1 g, 103 mmol) in THF (100 mL) at room
temperature under nitrogen was treated dropwise with borane−
tetrahydrofuran complex (206 mL, 206 mmol), and the resulting
mixture was stirred at 60 °C for 16 h and then was cooled to 0 °C. The
mixture was carefully treated with a 5 N HCl aqueous solution (20
mL), and then most of the solvent was removed in vacuo. The
aqueous residue was diluted with water (40 mL), and the aqueous
phase was washed with Et2O (2 × 50 mL). The aqueous layer was
then basified with a 10 N NaOH aqueous solution to pH 10 and
extracted first with CH2Cl2 and then with EtOAc. The combined
organics were dried over MgSO4 and concentrated in vacuo to give 7-
methoxy-6-methyl-2,3,4,5-tetrahydro-1H-benzo[d]azepine (18.9 g,
96%) as a pale yellow solid. LCMS (method formate): retention
time 0.80 min, [M + H]+ = 192. 1H NMR (400 MHz, CDCl3) δ ppm
6.93 (d, J = 8.1 Hz, 1H), 6.64 (d, J = 8.1 Hz, 1H), 3.82 (s, 3H), 2.82−
3.03 (m, 8H), 2.21 (s, 3H), 1.85 (br s, 1H).
1
H]+ = 287. H NMR (400 MHz, CDCl3) δ ppm 7.42 (d, J = 7.8 Hz,
1H), 7.08 (d, J = 7.6 Hz, 1H), 3.58 (d, J = 4.0 Hz, 4H), 2.86−3.08 (m,
4H), 2.56 (s, 3H), 1.44 (s, 9H).
Step 12. Oxalyl chloride (6.4 mL, 73 mmol) was added to a solution
of 3-cyano-4-[(1-methylethyl)oxy]benzoic acid (10.7 g, 52 mmol) in
CH2Cl2 (100 mL) followed by the addition of DMF (0.044 mL, 0.57
mmol), and the resulting mixture was stirred at room temperature for
4 h. The reaction mixture was concentrated in vacuo. The residue was
coevaporated with cyclohexane (2 × 50 mL) to give 3-cyano-4-[(1-
methylethyl)oxy]benzoyl chloride (11.7 g, 100%) as pale yellow oil
Step 9. A solution of 6-methyl-7-(methyloxy)-2,3,4,5-tetrahydro-1H-
3-benzazepine (18.9 g, 99 mmol) in aqueous HBr (48% w/w, 100 mL)
was stirred at 100 °C for 3 h, then was cooled to room temperature.
Most of the water was removed in vacuo to give a beige solid. This
residue was suspended in THF (100 mL) and CH2Cl2 (100 mL), and
the resulting solution was treated with NEt3 (41.3 mL, 296 mmol),
then with di-tert-butyl dicarbonate (31.0 mL, 133 mmol). The
resulting mixture was stirred at room temperature for 2 h, then was
concentrated in vacuo. The residue was dissolved in CH2Cl2 (200 mL)
and the organic phase was washed with 0.5 N HCl aqueous solution
(200 mL), then brine (200 mL), dried over MgSO4, and concentrated
in vacuo to give a mixture of the desired phenol and the byproduct
tert-butyl carbonate. The residue was dissolved in MeOH (200 mL)
and treated at room temperature with a 2 N NaOH aqueous solution
(200 mL). The resulting mixture was stirred at room temperature for 2
h. Half of the solvent was then removed in vacuo, and the residual
aqueous phase was washed with Et2O (200 mL). The aqueous layer
was then acidified with a 2 N HCl aqueous solution and then was
extracted with EtOAc (2 × 200 mL). The combined organic phases
were dried over MgSO4 and concentrated in vacuo to give a brown
gum. Purification of this residue by flash chromatography on silica gel
(330 g column, 0−50% EtOAc in cyclohexane) gave tert-butyl 7-
hydroxy-6-methyl-4,5-dihydro-1H-benzo[d]azepine-3(2H)-carboxylate
(12.0 g, 44%) as a white solid. The Et2O layer previously kept was
washed with a 2 N NaOH aqueous solution (100 mL), and the
aqueous phase was acidified with a 5 M HCl aqueous solution before
being extracted with CH2Cl2 (2 × 50 mL). The combined organics
were dried over MgSO4 and concentrated in vacuo to give a second
batch of tert-butyl 7-hydroxy-6-methyl-4,5-dihydro-1H-benzo[d]-
azepine-3(2H)-carboxylate (2.1 g, 7.7%). LCMS (method formate):
1
which solidified on standing. H NMR (400 MHz, DMSO-d6) δ ppm
8.37 (s, 1H), 8.26−8.29 (m, 1H), 7.07 (d, J = 7.7 Hz, 1H), 4.79−4.85
(m, 1H), 1.49 (t, J = 8 Hz, 6H).
Step 13. A mixture of tert-butyl 7-cyano-6-methyl-4,5-dihydro-1H-
benzo[d]azepine-3(2H)-carboxylate (3.0 g, 10.5 mmol), hydroxyl-
amine hydrochloride (2.18 g, 31.4 mmol), and NaHCO3 (2.64 g, 31.4
mmol) in EtOH (50 mL) was refluxed for 24 h. Hydroxylamine
hydrochloride (1 g) and NaHCO3 (1 g) were further added, and the
reflux was continued for 24 h. The reaction mixture was then cooled to
room temperature, and the insolubles were filtered off. The solvent
was evaporated in vacuo, and the residue was dissolved in water (50
mL). The aqueous phase was extracted with EtOAc (2 × 50 mL) and
the combined organics were washed with water and brine, dried over
MgSO4, and concentrated in vacuo to give tert-butyl 7-(N′-
hydroxycarbamimidoyl)-6-methyl-4,5-dihydro-1H-benzo[d]azepine-
3(2H)-carboxylate (3.3 g, 99%) as colorless foam. LCMS (method
1
formate): retention time 0.94 min, [M + H]+ = 320. H NMR (400
MHz, DMSO-d6) δ ppm 9.21 (s, 1H), 7.61 (br s, 1H), 7.30 (br s, 1H),
6.92−7.10 (m, 2H), 3.46 (br s, 4H), 2.81−2.98 (m, 4H), 2.27 (s, 3H),
1.37 (s, 9H).
Step 14. 3-Cyano-4-[(1-methylethyl)oxy]benzoyl chloride (1.84 g,
8.22 mmol) was added portionwise to a stirred solution of 1,1-
dimethylethyl 7-[(hydroxyamino)(imino)methyl]-6-methyl-1,2,4,5-tet-
rahydro-3H-3-benzazepine-3-carboxylate (2.5 g, 7.83 mmol) in toluene
(20 mL) and pyridine (10 mL). 10 min after the addition was
complete, the resulting mixture was warmed and stirred at reflux for 3
h and then was cooled to room temperature and concentrated in
vacuo. Purification of the residue by flash chromatography on silica gel
(5−25% EtOAc in isohexane) gave tert-butyl 7-(5-(3-cyano-4-
isopropoxyphenyl)-1,2,4-oxadiazol-3-yl)-6-methyl-4,5-dihydro-1H-
benzo[d]azepine-3(2H)-carboxylate (2.7 g, 71%) as a colorless solid.
LCMS (method formate): retention time 1.49 min, [M + H]+ = 489.
1H NMR (400 MHz, DMSO-d6) δ ppm 8.49 (d, J = 2.3 Hz, 1H), 8.39
(m, 1H), 7.52−7.70 (m, 2H), 7.27 (d, J = 7.8 Hz, 1H), 4.90−5.08 (m,
1H), 3.43−3.59 (m, 4H), 2.97 (m, 4H), 2.48 (s, 3H), 1.39 (d, J = 5.8
Hz, 6H), 1.30 (s, 9H).
1
retention time 1.10 min, [M + H]+ = 278. H NMR (400 MHz,
CDCl3) δ ppm 6.84 (d, J = 8.1 Hz, 1H), 6.58 (d, J = 8.1 Hz, 1H), 4.76
(s, 1H), 3.56 (m, 4H), 2.80−2.99 (m, 4H), 2.23 (s, 3H), 1.47 (s, 9H).
Step 10. A solution of 1,1-dimethylethyl 7-hydroxy-6-methyl-1,2,4,5-
tetrahydro-3H-3-benzazepine-3-carboxylate (12 g, 43.3 mmol) and
pyridine (7.00 mL, 87 mmol) in CH2Cl2 (200 mL) at −50 °C under
nitrogen was treated with triflic anhydride (9.50 mL, 56.2 mmol). The
resulting mixture was allowed to warm to room temperature over 1 h
and then was washed with water and a 1 N HCl aqueous solution,
dried over MgSO4, and concentrated in vacuo. Purification of the
residue by flash chromatography on silica gel (330 g silica column,
Step 15. A solution of tert-butyl 7-(5-(3-cyano-4-isopropoxyphenyl)-
1,2,4-oxadiazol-3-yl)-6-methyl-4,5-dihydro-1H-benzo[d]azepine-
3(2H)-carboxylate (1.9 g, 3.9 mmol) in CH2Cl2 (15 mL) was slowly
8251
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
treated with trifluoroacetic acid (5 mL), and the resulting mixture was
stirred at room temperature for 2 h and then was concentrated in
vacuo. The residue was coevaporated twice with toluene and triturated
with Et2O to give 2-isopropoxy-5-(3-(6-methyl-2,3,4,5-tetrahydro-1H-
benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)benzonitrile trifluoroacetate
(1.7 g, 87%) as a colorless solid. LCMS (method formate): retention
time 0.98 min, [M + H]+ = 389.2. 1H NMR (400 MHz, DMSO-d6) δ
ppm 8.90 (br s, 2H), 8.49 (d, J = 2.3 Hz, 1H), 8.39 (dd, J = 8.8, 2.3
Hz, 1H), 7.52−7.70 (m, 2H), 7.27 (d, J = 7.8 Hz, 1H), 4.98 (dt, J =
12.1, 6.1 Hz, 1H), 3.02−3.32 (m, 8H), 2.48 (s, 3H), 1.39 (d, J = 6.1
Hz, 6H).
filled under nitrogen with phosphorus oxychloride (50 mL, 536
mmol), and the resulting suspension was stirred at 70 °C for 20 h and
then was cooled to room temperature and carefully added to a mixture
of ice and water (800 g). The precipitate formed was filtered off and
dried under vacuum at 40 °C to give a brown solid. This residue was
dissolved in CH2Cl2 and the organic phase was washed with water,
dried using a phase separator, and concentrated in vacuo to give ethyl
6-chloro-5-cyanonicotinate (9.75 g, 89%) as a very pale gray solid.
LCMS (method formate): retention time 0.98 min, [M + H]+ = 211.
1H NMR (400 MHz, DMSO-d6) δ ppm 9.13 (d, J = 2.3 Hz, 1H), 8.92
(d, J = 2.3 Hz, 1H), 4.38 (q, J = 7.1 Hz, 2H), 1.35 (t, J = 7.1 Hz, 3H).
Step 2. A flask was charged with ethyl 6-chloro-5-cyano-3-
pyridinecarboxylate (2.11 g, 10 mmol), isopropylamine (1.8 mL,
21.01 mmol), and triethylamine (2.8 mL, 20.09 mmol) and then filled
with EtOH (14 mL), and the resulting solution was stirred under
nitrogen at 120 °C for 25 min under microwave irradiation and then
was cooled to room temperature and concentrated in vacuo, The
residue was partitioned between EtOAc and a saturated NaHCO3
aqueous solution, and the layers were separated. The aqueous phase
was extracted with EtOAc and the combined organic phases were
washed with brine, dried over MgSO4, and concentrated in vacuo to
give ethyl 5-cyano-6-(isopropylamino)nicotinate (2.2 g, 95%) as a very
pale gray solid which was used in the next step without any further
purification. LCMS (method formate): retention time 1.13 min, [M +
Step 16. 18 was obtained as a colourless solid from 2-isopropoxy-5-
(3-(6-methyl-2,3,4,5-tetrahydro-1H-benzo[d]azepin-7-yl)-1,2,4-oxadia-
zol-5-yl)benzonitrile using L-glyceraldehyde in an procedure analogous
to the one described for compound 12: 214 mg, 43% yield. LCMS
1
(method formate): retention time 0.85 min, [M + H]+ = 463.2. H
NMR (400 MHz, DMSO-d6) δ ppm 10.11 (br s, 1H), 8.50 (d, J = 2.3
Hz, 1H), 8.39 (dd, J = 8.8, 2.3 Hz, 1H), 7.65 (d, J = 7.8 Hz, 1H), 7.56
(d, J = 9.1 Hz, 1H), 7.27 (d, J = 7.8 Hz, 1H), 5.62 (br s, 1H), 4.92−
5.07 (m, 2H), 3.94−4.07 (m, 1H), 3.75 (br s, 2H), 3.40−3.53 (m,
2H), 3.21−3.39 (m, 4H), 2.97−3.20 (m, 4H), 2.50 (s, 3H), 1.39 (d, J
= 6.1 Hz, 6H).
(S)-5-(3-(3-(2,3-Dihydroxypropyl)-6-methyl-2,3,4,5-tetrahy-
dro-1H-benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxybenzonitrile Hydrochloride (19). This compound was
obtained as a colorless solid from 2-isopropoxy-5-(3-(6-methyl-2,3,4,5-
tetrahydro-1H-benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)benzonitrile
using D-glyceraldehyde in an procedure analogous to the one described
for compound 13: 112 mg, 47% yield. LCMS (method formate):
1
H]+ = 233. H NMR (400 MHz, CDCl3) δ ppm 8.15 (d, J = 2.0 Hz,
1H), 7.50 (d, J = 2.3 Hz, 1H) 6.52 (s, 1H) 4.60 (m, 1H) 3.53−3.78
(m, 3H) 0.48−0.69 (m, 8H).
Step 3. A solution of ethyl 5-cyano-6-[(1-methylethyl)amino]-3-
pyridinecarboxylate (9.85 g, 42.2 mmol) in MeOH (130 mL) at room
temperature was treated with 1 N LiOH aqueous solution (127 mL,
127 mmol), and the resulting mixture was stirred at this temperature
for 2.5 h. Most of MeOH was then removed in vacuo. The aqueous
layer was cooled at 0 °C, and the solution was treated with AcOH
(12.1 mL, 211 mmol). The precipitate formed was filtered off and
dried at 40 °C under vacuum to give 5-cyano-6-(isopropylamino)-
nicotinic acid (7.6 g, 88%) as a pale yellow solid. LCMS (method
1
retention time 0.89 min, [M + H]+ = 463.3. H NMR (400 MHz,
DMSO-d6) δ ppm 10.11 (br s, 1H), 8.49 (s, 1H), 8.39 (d, J = 8.8, 1H),
7.57−7.64 (m, 2H), 7.26 (d, J = 8.8 Hz, 1H), 4.95−5.01 (m,1H),
3.99−4.15 (m, 1H), 3.02−3.43 (m, 14H), 2.50 (s, 3H), 1.39 (d, J = 6.1
Hz, 6H).
5-(3-(3-(1,3-Dihydroxypropan-2-yl)-6-methyl-2,3,4,5-tetra-
hydro-1H-benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxybenzonitrile Hydrochloride (20). Sodium triacetox-
yborohydride (1.05 g, 4.98 mmol) was added portionwise to a stirred
solution of 2-isopropoxy-5-(3-(6-methyl-2,3,4,5-tetrahydro-1H-benzo-
[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)benzonitrile (500 mg, 1 mmol)
and 2,2-dimethyl-1,3-dioxan-5-one (388 mg, 3 mmol) in CH2Cl2 (10
mL). The reaction mixture was stirred at room temperature for 16 h
and then was treated with a saturated NaHCO3 aqueous solution (20
mL), and the resulting mixture was vigorously stirred for 15 min. The
layers were separated. The aqueous phase was extracted with CH2Cl2
(10 mL), and the combined organic phases were dried using a phase
separator and concentrated in vacuo. The residue was dissolved in
THF (3 mL), and the solution was treated with a 2 N HCl aqueous
solution (3 mL). The resulting mixture was stirred at room
temperature for 5 h and then was treated with a saturated NaHCO3
aqueous solution (3 mL). The resulting mixture was stirred for 10 min,
and the layers were separated. The aqueous phase was extracted with
CH2Cl2 (10 mL), and the combined organics were dried using a phase
separator and concentrated in vacuo. Purification of the residue via
flash chromatography on silica gel (0−6% MeOH in CH2Cl2) gave a
compound which was treated with 1 N HCl solution in Et2O (2 mL).
The solvent was then evaporated and the residue triturated with Et2O
to give 20 (203 mg, 41%) as a colorless solid. LCMS (method
1
formate): retention time 0.47 min, [M + H]+ = 206. H NMR (400
MHz, DMSO-d6) δ ppm 12.13−13.65 (br s, 1H), 8.73 (d, J = 2.3 Hz,
1H), 8.23 (d, J = 2.3 Hz, 1H), 7.44 (d, J = 8.1 Hz, 1 H), 4.40 (dspt, J =
6.3, 8.1 Hz, 1H), 1.15−1.25 (d, J = 6.3 Hz, 6H).
Step 4. A mixture of 5-cyano-6-(isopropylamino)nicotinic acid (1 g,
4.87 mmol), N-ethylmorpholine (1.12 g, 1.23 mL, 9.75 mmol), HATU
(2.22 g, 5.85 mmol), and tert-butyl 7-(N′-hydroxycarbamimidoyl)-6-
methyl-4,5-dihydro-1H-benzo[d]azepine-3(2H)-carboxylate (1.87 g,
5.85 mmol) in DMF (10 mL) was stirred at room temperature for
4 h and then was diluted with EtOAc (50 mL). The organic phase was
washed successively with a saturated NaHCO3 aqueous solution,
water, and brine, then was dried over MgSO4 and concentrated in
vacuo. The residue was dissolved in toluene (40 mL) and pyridine (10
mL), and the resulting mixture was refluxed for 2 h and then was
cooled to room temperature and concentrated in vacuo. Purification of
the residue via flash chromatography on silica gel (20% EtOAc in
hexanes) gave tert-butyl 7-(5-(5-cyano-6-(isopropylamino)pyridin-3-
yl)-1,2,4-oxadiazol-3-yl)-6-methyl-4,5-dihydro-1H-benzo[d]azepine-
3(2H)-carboxylate (1.31 g, 55%) as a colorless solid. LCMS (method
1
formate): retention time 1.51 min, [M + H]+ = 489. H NMR (400
MHz, DMSO-d6) δ ppm 8.99 (d, J = 2.3 Hz, 1H), 8.57 (d, J = 2.3 Hz,
1H), 7.81 (d, J = 8.1 Hz, 1H), 7.48−7.60 (m, 1H), 7.17 (d, J = 8.1 Hz,
1H), 4.46 (m, 1H), 3.41−3.59 (m, 4H), 2.90−3.03 (m, 4H), 2.47 (s,
3H), 1.31 (br s, 9H), 1.24 (d, J = 6.6 Hz, 6H).
1
formate): retention time 0.92 min, [M + H]+ = 463. H NMR (400
MHz, DMSO-d6) δ ppm 9.95−10.10 (br s, 1H), 8.50 (d, J = 2.3 Hz,
1H), 8.39 (dd, J = 9.1, 2.3 Hz, 1H), 7.65 (d, J = 7.8 Hz, 1H), 7.56 (d, J
= 9.3 Hz, 1H), 7.27 (d, J = 7.8 Hz, 1H), 5.41 (br s, 2H), 4.93−5.03
(m, 1H), 3.80 (d, J = 4.5 Hz, 6H), 3.51−3.62 (m, 1H), 3.40 (br s, 2H),
3.24 (d, J = 10.6 Hz, 2H), 3.02−3.12 (m, 1H), 2.51 (s, 3H), 1.39 (d, J
= 6.1 Hz, 6H).
5-(3-(3-(1,3-Dihydroxypropan-2-yl)-6-methyl-2,3,4,5-tetra-
hydro-1H-benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-
(isopropylamino)nicotinonitrile (21). Step 1. To make ethyl 6-
chloro-5-cyanonicotinate, a flask was charged with ethyl 5-cyano-6-
oxo-1,6-dihydro-3-pyridinecarboxylate (10 g, 52.0 mmol) and then
Step 5.
A suspension of tert-butyl 7-(5-(5-cyano-6-
(isopropylamino)pyridin-3-yl)-1,2,4-oxadiazol-3-yl)-6-methyl-4,5-dihy-
dro-1H-benzo[d]azepine-3(2H)-carboxylate (1.3 g, 2.66 mmol) in 1,4-
dioxane (5 mL) at room temperature was treated with a 4 N HCl
solution in 1,4-dioxane (10 mL), and the resulting mixture was stirred
at this temperature for 7 h before being diluted with Et2O (70 mL).
The solid formed was filtered off, washed with Et2O, and dried under
vacuum to give 2-[(1-methylethyl)amino]-5-[3-(6-methyl-2,3,4,5-
tetrahydro-1H-3-benzazepin-7-yl)-1,2,4-oxadiazol-5-yl]-3-pyridine-
carbonitrile hydrochloride (1.05 g, 93%) as a colorless solid. LCMS
8252
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
(method formate): retention time 0.89 min, [M + H]+ = 389. H
NMR (400 MHz, DMSO-d6) δ ppm 9.28 (br s, 2H), 9.00 (d, J = 2.3
Hz, 1H), 8.58 (d, J = 2.3 Hz, 1H), 7.84 (d, J = 7.8 Hz, 1H), 7.54−7.68
(m, 1H), 7.18−7.34 (m, 1H), 4.48 (spt, J = 6.6 Hz, 1H), 3.08−3.31
(m, 8H), 2.52 (s, 3H), 1.24 (d, J = 6.6 Hz, 6H).
7.42 (m, 1H), 4.42−4.60 (m, 1H), 3.42−4.53 (m, 4H), 2.80−3.02 (m,
4H), 1.46 (s, 9H), 1.25 (d, J = 6.5 Hz, 6H).
1
Step 2. 2-(Isopropylamino)-5-(3-(2,3,4,5-tetrahydro-1H-benzo[d]-
azepin-7-yl)-1,2,4-oxadiazol-5-yl)nicotinonitrile was obtained as a
white solid from tert-butyl 7-(5-(5-cyano-6-(isopropylamino)pyridin-
3-yl)-1,2,4-oxadiazol-3-yl)-4,5-dihydro-1H-benzo[d]azepine-3(2H)-
carboxylate using a similar procedure as described for compound 21
(step 5): 887 mg, 91% yield. LCMS (method formate): retention time
Step 6. Sodium triacetoxyborohydride (748 mg, 3.53 mmol) was
added portionwise to a stirred solution of 2-[(1-methylethyl)amino]-5-
[3-(6-methyl-2,3,4,5-tetrahydro-1H-3-benzazepin-7-yl)-1,2,4-oxadia-
zol-5-yl]-3-pyridinecarbonitrile hydrochloride (300 mg, 0.71 mmol)
and 2,2-dimethyl-1,3-dioxan-5-one (276 mg, 253 μL, 2.12 mmol) in
CH2Cl2 (10 mL). The reaction mixture was stirred at room
temperature for 16 h and then was treated with a saturated
NaHCO3 aqueous solution (20 mL), and the resulting mixture was
vigorously stirred for 15 min. The layers were separated. The aqueous
phase was extracted with CH2Cl2 (2 × 10 mL), and the combined
organic phases were dried using a phase separator and concentrated in
vacuo. The residue was dissolved in THF (5 mL), and the solution was
treated with a 2 N HCl aqueous solution (5 mL). The resulting
mixture was allowed to stand at room temperature for 60 h and then
was treated with a saturated NaHCO3 aqueous solution (10 mL). The
resulting mixture was stirred for 10 min, and the layers were separated.
The aqueous phase was extracted with CH2Cl2 (3 × 10 mL), and the
combined organics were dried using a phase separator and
concentrated in vacuo. Purification of the residue via flash
chromatography on silica gel (0−8% MeOH in CH2Cl2) gave a
compound which was treated with 1 N HCl solution in Et2O (5 mL).
The solvent was then evaporated and the residue triturated with Et2O
to give 21 (181 mg, 51%) as a colorless solid. LCMS (method
1
0.83 min, [M + H]+ = 375. H NMR (400 MHz, DMSO-d6) δ ppm
9.32 (br s, 2H), 9.00 (d, J = 2.3 Hz, 1H), 8.58 (d, J = 2.3 Hz, 1H), 7.92
(s, 1H), 7.87−7.91 (m, 1H), 7.85 (d, J = 8.1 Hz, 1H), 4.47 (spt, J = 6.0
Hz, 1H), 3.22 (m, 8H), 1.24 (d, J = 6.6 Hz, 6H).
Step 3. 5-(3-(3-(2,2-Dimethyl-1,3-dioxan-5-yl)-2,3,4,5-tetrahydro-
1H-benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-(isopropylamino)-
nicotinonitrile was obtained as a yellow oil from 2-(isopropylamino)-5-
(3-(2,3,4,5-tetrahydro-1H-benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-
nicotinonitrile using a similar procedure as described for compound 21
(step 6). LCMS (method formate): retention time 0.98 min, [M +
H]+ = 488.
Step 4. 23 was obtained as a white foam from 5-(3-(3-(2,2-dimethyl-
1,3-dioxan-5-yl)-2,3,4,5-tetrahydro-1H-benzo[d]azepin-7-yl)-1,2,4-oxa-
diazol-5-yl)-2-(isopropylamino)nicotinonitrile using a similar proce-
dure as described for compound 21 (step 6): 29 mg, 13% yield (2
steps). LCMS (method formate): retention time 0.67 min, [M + H]+
1
= 448. H NMR (400 MHz, DMSO-d6) δ ppm 9.00 (d, J = 2.5 Hz,
1H), 8.58 (d, J = 2.5 Hz, 1H), 7.76−7.85 (m, 3H), 7.31 (d, J = 7.8 Hz,
1H), 4.47 (spt, J = 6.6 Hz, 1H), 4.24 (dd, J = 5.9, 4.4 Hz, 2H), 3.36−
3.52 (m, 4H), 2.88−3.00 (m, 4H), 2.80 (br s, 4H), 2.71 (qt, J = 6.5
Hz, 1H), 1.24 (d, J = 6.6 Hz, 6H).
1
formate): retention time 0.89 min, [M + H]+ = 463. H NMR (400
5-(3-{4-[2-Hydroxy-1-(hydroxymethyl)ethyl]-9-methyl-
2,3,4,5-tetrahydro-1,4-benzoxazepin-8-yl}-1,2,4-oxadiazol-5-
yl)-2-[(1-methylethyl)oxy]benzonitrile (24). Step 1. To make 2,4-
dihydroxy-3-methylbenzaldehyde, POCl3 (8.02 mL, 86 mmol) was
slowly added over 20 min to DMF (26 mL, 336 mmol) in a two-neck
flask with internal thermometer, using an ice bath to make sure the
temperature remained between 10 and 20 °C. Once the addition was
completed, the resulting mixture was added dropwise to a solution of
2-methyl-1,3-benzenediol (4.84 g, 39 mmol) in DMF (26 mL),
making sure using an ice bath that the temperature did not exceed 30
°C. The resulting yellow solution was stirred at room temperature for
16 h and then was cautiously added to an ice cold solution of 5 N
NaOH aqueous solution (15 mL of 5 M NaOH aqueous solution were
used per mL of POCl3), making sure the temperature did not rise
above 30 °C. The aqueous phase was extracted with Et2O and then
slowly acidified with concentrated aqueous HCl. An ice bath was used
to make sure that the temperature was kept below 50 °C. The aqueous
phase was extracted 3 times with EtOAc. The combined organic
phases were washed slowly with a saturated NaHCO3 aqueous
solution (warning: CO2 evolution), then brine, dried over MgSO4, and
concentrated in vacuo. NMR analysis of the crude material showed
mainly a ∼2/1 mixture of product and starting material. Trituration of
this residue with CH2Cl2 gave 2,4-dihydroxy-3-methylbenzaldehyde
(2.37 g, 40%) as a white solid. LCMS (method formate): retention
time 0.43 min, [M + H]+ = 153.1. 1H NMR (400 MHz, DMSO-d6) δ
ppm 11.60 (s, 1H), 10.78 (br s, 1H), 9.71 (s, 1H), 7.43 (d, J = 8.0 Hz,
1H), 6.56 (d, J = 8.0 Hz, 1H), 1.97 (s, 3H).
MHz, DMSO-d6) δ ppm 10.04 (br s, 1H), 9.00 (d, J = 2.3 Hz, 1H),
8.58 (d, J = 2.3 Hz, 1H), 7.83 (d, J = 7.8 Hz, 1H), 7.62 (d, J = 7.8 Hz,
1H), 7.26 (d, J = 8.1 Hz, 1H), 5.40 (br s, 2H), 4.46 (dd, J = 14.3, 6.7
Hz, 1H), 3.69−3.88 (m, 6H), 3.51−3.62 (m, 1H), 3.38 (d, J = 6.8 Hz,
2H), 3.34 (s, 3H), 3.18−3.31 (m, 2H), 3.01−3.10 (m, 1H), 1.24 (d, J
= 6.6 Hz, 6H).
5-(3-(3-(1,3-Dihydroxypropan-2-yl)-2,3,4,5-tetrahydro-1H-
benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-isopropoxy-
benzonitrile (22). Step 1. 5-(3-(3-(2,2-Dimethyl-1,3-dioxan-5-yl)-
2,3,4,5-tetrahydro-1H-benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-
isopropoxybenzonitrile was obtained as a pale yellow gum from 2-
isopropoxy-5-(3-(2,3,4,5-tetrahydro-1H-benzo[d]azepin-7-yl)-1,2,4-ox-
adiazol-5-yl)benzonitrile (synthesis described in ref 39) using a similar
procedure that the one described for compound 14 (step 11). LCMS
1
(method formate): retention time 1.09 min, [M + H]+ = 489. H
NMR (400 MHz, CDCl3) δ ppm 8.45 (d, J = 2.0 Hz, 1H), 8.35 (dd, J
= 9.0, 2.1 Hz, 1H), 7.87−7.99 (m, 2H), 7.22−7.32 (m, 1H), 7.14 (d, J
= 9.1 Hz, 1H), 4.73−4.87 (m, 1H), 4.05−4.14 (m, 4H), 3.73−3.81 (m,
5H), 3.50−3.59 (m, 2H), 2.82 (d, J = 10.1 Hz, 2H), 1.32−1.56 (m,
12H).
Step 2. 22 was obtained as a cream colored crunchy foam from 5-(3-
(3-(2,2-dimethyl-1,3-dioxan-5-yl)-2,3,4,5-tetrahydro-1H-benzo[d]-
azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-isopropoxybenzonitrile using a
similar procedure as the one described for compound 14 (step 12):
66 mg, 33% yield (2 steps). LCMS (method formate): retention time
0.88 min, [M + H]+ = 448. 1H NMR (400 MHz, CDCl3) δ ppm 8.46
(d, J = 2.0 Hz, 1H), 8.37 (dd, J = 8.8, 2.3 Hz, 1H), 7.92−7.98 (m, 2H),
7.28−7.32 (m, 1H), 7.16 (d, J = 9.0 Hz, 1H), 4.84 (spt, J = 6.1 Hz,
1H), 3.78 (d, J = 6.3 Hz, 4H), 3.53 (s, 2H), 2.84−3.33 (m, 9H), 1.52
(d, J = 6.0 Hz, 6H).
Step 2. To a solution of 2,4-dihydroxy-3-methylbenzaldehyde (2.37
g, 15.58 mmol) in EtOH (50 mL) at room temperature was added
ethanolamine (1.0 mL, 16.4 mmol), and the resulting mixture was
stirred at this temperature for 1 h. The mixture was concentrated in
vacuo to give 4-{[(2-hydroxyethyl)imino]methyl}-2-methyl-1,3-benze-
nediol (3.04 g, 100%) as a yellow foam which was used in the next step
without further purification. LCMS (method formate): retention time
0.39 min, [M + H]+ = 196.2. 1H NMR (400 MHz, MeOD-d4) δ ppm
8.10 (s, 1H), 6.94 (d, J = 8.0 Hz, 1H), 6.24 (d, J = 8.0 Hz, 1H), 3.59−
3.79 (m, 6H), 2.76−2.79 (m, 1H), 2.02 (s, 3H).
5-(3-(3-(1,3-Dihydroxypropan-2-yl)-2,3,4,5-tetrahydro-1H-
benzo[d]azepin-7-yl)-1,2,4-oxadiazol-5-yl)-2-(isopropylamino)-
nicotinonitrile (23). Step 1. tert-Butyl 7-(5-(5-cyano-6-
(isopropylamino)pyridin-3-yl)-1,2,4-oxadiazol-3-yl)-4,5-dihydro-1H-
benzo[d]azepine-3(2H)-carboxylate was obtained as a pale yellow
solid from 5-cyano-6-(isopropylamino)nicotinic acid and tert-butyl 7-
(N′-hydroxycarbamimidoyl)-4,5-dihydro-1H-benzo[d]azepine-3(2H)-
carboxylate using a similar procedure as described for compound21
(step 4): 1.24 g, 73% yield. LCMS (method formate): retention time
Step 3. To a solution of 4-{[(2-hydroxyethyl)imino]methyl}-2-
methyl-1,3-benzenediol (3.04 g, 15.58 mmol) in THF (50 mL) at
room temperature was added sodium triacetoxyborohydride (6.60 g,
31.2 mmol), and the resulting yellow suspension was stirred at this
1
1.11 min, [M + H]+ = 475. H NMR (400 MHz, DMSO-d6) δ ppm
8.97−9.04 (m, 1H), 8.55−8.62 (m, 1H), 7.79−7.92 (m, 3H), 7.34−
8253
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
temperature for 1 h. Sodium triacetoxyborohydride (500 mg, 2.35
mmol) was added, and the resulting mixture was stirred at room
temperature for 16 h. The mixture was then treated with 4 N HCl in
1,4-dioxane (15.6 mL, 62.3 mmol) and then was concentrated in
vacuo. The residue was dissolved in MeOH and loaded on a prewet 70
g SCX column. The column was eluted with MeOH followed by a 2 N
NH3 solution in MeOH. The ammoniac fractions were combined and
concentrated in vacuo to give 4-(((2-hydroxyethyl)amino)methyl)-2-
methylbenzene-1,3-diol (3.07 g, 100%) which was used in the next
step without further purification. The reaction was considered as
quantitative at this stage. LCMS (method formate): retention time
0.46 min, [M + H]+ = 198.2.
mmol). The resulting mixture was stirred at 100 °C under nitrogen for
7 h, then was cooled to room temperature. The insoluble material was
filtered off and washed with EtOAc. The combined organic phases
were concentrated in vacuo. The residue was dissolved in EtOAc, and
the organic phase was washed with a saturated NaHCO3 aqueous
solution. The aqueous phase was extracted with EtOAc, and the
combined organics were washed with brine, dried over MgSO4, and
concentrated in vacuo. Purification of the residue by flash
chromatography on silica gel (50 g column, 5−25% EtOAc in
hexanes) gave tert-butyl 8-cyano-9-methyl-2,3-dihydrobenzo[f ][1,4]-
oxazepine-4(5H)-carboxylate (1.16 g, 91%) as a white solid. LCMS
1
(method formate): retention time 1.21 min, [M + H]+ = 289.2. H
Step 4. To a solution of 4-{[(2-hydroxyethyl)amino]methyl}-2-
methyl-1,3-benzenediol (3.07 g, 15.6 mmol) in MeOH (30 mL) was
added triethylamine (6.51 mL, 46.7 mmol) and then di-tert-butyl
dicarbonate (3.80 mL, 16.4 mmol), and the resulting mixture was
stirred at room temperature for 1 h and then was concentrated in
vacuo. The residue was purified using a 70 g SCX column (eluting first
with MeOH and then with a 2 N NH3 solution in MeOH), and the
ethanolic fractions were combined and concentrated in vacuo.
Purification of the residue via flash chromatography on silica gel
(10−60%, EtOAc in hexanes) gave a white solid which was further
purified. This residue was dissolved in EtOAc and the organic phase
extracted three times with a 2 N NaOH aqueous solution. The
combined aqueous phases were acidified with a 5 N HCl aqueous
solution and then extracted twice with EtOAc. The combined organics
were washed with brine, dried over MgSO4, and concentrated in vacuo
to give tert-butyl 2,4-dihydroxy-3-methylbenzyl(2-hydroxyethyl)-
carbamate (1.61 g, 35%) as a pale pink solid. LCMS (method
NMR (400 MHz, CDCl3) δ ppm 7.28−7.32 (m, 1H), 7.20−7.30 (m,
1H), 4.40−4.51 (m, 2H), 4.05−4.07 (m, 2H), 3.83−3.85 (m, 2H),
2.46 (s, 3H), 1.43 (s, 9H).
Step 8. Oxalyl chloride (6.4 mL, 73 mmol) was added to a solution
of 3-cyano-4-[(1-methylethyl)oxy]benzoic acid (10.7 g, 52 mmol) in
CH2Cl2 (100 mL) followed by the addition of DMF (0.044 mL, 0.57
mmol), and the mixture was stirred at room temperature for 4 h. The
reaction mixture was filtered and concentrated in vacuo. The residue
was coevaporated with cyclohexane (2 × 50 mL) to give 3-cyano-4-
[(1-methylethyl)oxy]benzoyl chloride (11.7 g, 100%) as a pale yellow
1
oil which solidified on standing. H NMR (400 MHz, CDCl3) δ ppm
8.35 (d, J = 2.3 Hz, 1H), 8.26 (dd, J = 2.3, 9.1 Hz, 1H), 7.06 (d, J = 9.1
Hz, 1H), 4.81 (spt, J = 6.1 Hz, 1H), 1.47 (d, J = 6.1 Hz, 6H).
Step 9. To a stirred solution of 1,1-dimethylethyl 6-
[(hydroxyamino)(imino)methyl]-5-methyl-3,4-dihydro-2(1H)-
isoquinolinecarboxylate (4.7 g, 15.39 mmol) in toluene (35 mL) and
pyridine (26 mL) at ambient temperature was slowly added 3-cyano-4-
[(1-methylethyl)oxy]benzoyl chloride (3.8 g, 17.1 mmol) in toluene
(35 mL). After 20 min, the mixture was warmed to 80 °C and stirred
at this temperature for 15 h, then was cooled to room temperature and
concentrated in vacuo. The residue was suspended in EtOAc (50 mL),
and the organic phase was washed sequentially with 0.5 N HCl
aqueous solution (2 × 40 mL), a saturated NaHCO3 aqueous solution
(2 × 40 mL), and water (2 × 40 mL). The organic layer was dried
through a hydrophobic frit and the solvent removed in vacuo.
Purification of the residue via flash chromatography on silica gel (330 g
column, 0−60% EtOAc in cyclohexane) gave tert-butyl 8-(5-(3-cyano-
4-isopropoxyphenyl)-1,2,4-oxadiazol-3-yl)-9-methyl-2,3-dihydrobenzo-
[f ][1,4]oxazepine-4(5H)-carboxylate (2.85 g, 35%) as an off white
solid. LCMS (method formate): retention time 1.43 min, [M + H]+ =
1
formate): retention time 0.93 min, [M + H]+ = 298.2. H NMR (400
MHz, MeOD-d4) δ ppm 6.82 (d, J = 8.1 Hz, 1H), 6.31 (d, J = 8.1 Hz,
1H), 4.38 (br s, 2H), 3.64 (t, J = 6.30 Hz, 2H), 2.07 (s, 3H), (1.50,
9H). Some peaks were obscured by residual NMR solvent.
Step 5. To a solution of 1,1-dimethylethyl [(2,4-dihydroxy-3-
methylphenyl)methyl](2-hydroxyethyl)carbamate (1.61 g, 5.41 mmol)
and triphenylphosphine (1.56 g, 5.96 mmol) in THF (40 mL) at 0 °C
was added DIAD (1.16 mL, 5.96 mmol) dropwise, and the resulting
yellow mixture was stirred at this temperature for 30 min and then was
concentrated in vacuo. The residue was dissolved in EtOAc, and the
organic phase was washed twice with brine, dried over MgSO4, and
concentrated in vacuo. Purification of the residue by flash
chromatography on silica gel (100 g silica, 10−40% EtOAc in
hexanes) gave tert-butyl 8-hydroxy-9-methyl-2,3-dihydrobenzo[f ]-
[1,4]oxazepine-4(5H)-carboxylate (1.26 g, 84%) as white foam.
1
491.0. H NMR (250 MHz, DMSO-d6) δ ppm 8.50 (d, J = 2.1 Hz,
1H) 8.36−8.45 (m, 1H) 7.61−7.72 (m, 1H) 7.56 (d, J = 9.0 Hz, 1H)
7.28 (d, J = 7.9 Hz, 1H) 4.99 (s, 1H) 4.48 (br s, 2H) 4.07 (br s, 2H)
3.77 (br s, 2H) 2.47 (s, 3H) 1.40 (d, J = 6.1 Hz, 6H) 1.36 (s, 9H).
Step 10. A solution of 1,1-dimethylethyl 8-(5-{3-cyano-4-[(1-
methylethyl)oxy]phenyl}-1,2,4-oxadiazol-3-yl)-9-methyl-2,3-dihydro-
1,4-benzoxazepine-4(5H)-carboxylate (2.91 g, 5.93 mmol) in CH2Cl2
(25 mL) at room temperature was treated with trifluoroacetic acid (6
mL, 78 mmol), and the resulting mixture was stirred at this
temperature for 2 h and then was concentrated in vacuo, and the
residue coevaporated with toluene. Trituration of the residue with
Et2O gave 2-isopropoxy-5-(3-(9-methyl-2,3,4,5-tetrahydrobenzo[f ]-
[1,4]oxazepin-8-yl)-1,2,4-oxadiazol-5-yl)benzonitrile trifluoroacetate
(2.9 g, 97%) as a white solid. LCMS (method formate): retention
time 0.86 min, [M + H]+ = 391.1. 1H NMR (400 MHz, DMSO-d6) δ
ppm 9.45 (br s, 2H), 8.49 (d, J = 2.3 Hz, 1H), 8.38 (dd, J = 9.0, 2.1
Hz, 1H), 7.73 (d, J = 7.8 Hz, 1H), 7.55 (d, J = 9.1 Hz, 1H), 7.49 (d, J
= 7.8 Hz, 1H), 4.98 (spt, J = 6.0 Hz, 1H), 4.41 (s, 2H), 4.26 (d, J = 2.3
Hz, 2H), 3.56 (br s, 2H), 2.50 (s, 3H), 1.38 (d, J = 6.1 Hz, 6H).
Step 11. A stirred mixture of 2,2-dimethyl-1,3-dioxan-5-one (0.75 g,
5.76 mmol) and 2,2-dimethyl-1,3-dioxan-5-one (0.75 g, 5.76 mmol) in
CH2Cl2 (20 mL) at room temperature under nitrogen was treated with
sodium triacetoxyborohydride (2.036 g, 9.60 mmol) added in two
portions over 5 min. The mixture was stirred for 15 h, then was
evaporated under vacuum and the white solid suspended in EtOAc (20
mL). The organic layer was washed sequentially with a saturated
NaHCO3 aqueous solution (20 mL) and then water (20 mL). The
organic layer was separated, dried through a hydrophobic frit and the
LCMS (method formate): retention time 1.05 min, [M + H]+
=
280.3. 1H NMR (400 MHz, CDCl3) δ ppm 6.84−7.02 (m, 1H), 6.34−
6.58 (m, 1H), 5.06 (s, 1H), 4.35 (s, 2H), 3.98−4.08 (m, 2H), 3.74−
3.88 (m, 2H), 2.17 (s, 3H), 1.43 (s, 9H).
Step 6. To a solution of 1,1-dimethylethyl 8-hydroxy-9-methyl-2,3-
dihydro-1,4-benzoxazepine-4(5H)-carboxylate (1.23 g, 4.42 mmol) in
CH2Cl2 (40 mL) at 0 °C under nitrogen was added pyridine (0.715
mL, 8.84 mmol), then triflic anhydride (0.822 mL, 4.86 mmol)
dropwise. The yellow solution was stirred at this temperature for 10
min, then was concentrated in vacuo and the residue dissolved in
EtOAc. The organic phase was washed with a 1 N HCl aqueous
solution, a saturated NaHCO3 aqueous solution, then brine, dried over
MgSO4, and concentrated in vacuo to give tert-butyl 9-methyl-8-
(((trifluoromethyl)sulfonyl)oxy)-2,3-dihydrobenzo[f ][1,4]oxazepine-
4(5H)-carboxylate (1.82 g, 100%) as an orange oil which was used in
the next step without further purification. LCMS (method formate):
1
retention time 1.43 min, [M + H2O]+ = 429.2. H NMR (400 MHz,
CDCl3) δ ppm 7.12−7.22 (m, 1H), 6.90−7.00 (m, 1H), 4.40−4.43
(m, 2H), 4.07−4.10 (m, 2H), 3.84−3.86 (m, 2H), 2.29 (s, 3H), 1.43
(s, 9H).
Step 7. A solution of crude 1,1-dimethylethyl 9-methyl-8-
{[(trifluoromethyl)sulfonyl]oxy}-2,3-dihydro-1,4-benzoxazepine-
4(5H)-carboxylate (1.82 g, 4.42 mmol) in DMF (15 mL) was
degassed for 10 min under house vacuum and quenched several times
with nitrogen, then was treated with tetrakis(triphenylphosphine)-
palladium(0) (0.511 g, 0.442 mmol) and zinc cyanide (0.675 g, 5.75
8254
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
solvent removed in vacuo. Purification of the residue via flash
chromatography on silica gel (100 g column, 0−100% AcOEt in
CH2Cl2) gave 5-{3-[4-(2,2-dimethyl-1,3-dioxan-5-yl)-9-methyl-2,3,4,5-
tetrahydro-1,4-benzoxazepin-8-yl]-1,2,4-oxadiazol-5-yl}-2-[(1-
methylethyl)oxy]benzonitrile (0.74 g, 1.47 mmol, 76%) as a white
solid. LCMS (method formate): retention time 0.99 min, [M + H]+ =
ans 2,3,4,7,8-pentachlorodibenzofuran. Toxicol. Sci. 2010, 118 (1),
224−235.
(5) Martignoni, M.; Groothuis, G.; de Kanter, R. Species differences
between mouse, rat, dog, monkey and human CYP-mediated drug
metabolism, inhibition and induction. Expert Opin. Drug Metab.
Toxicol. 2006, 2 (6), 875−894.
1
505.2. H NMR (400 MHz, DMSO-d6) δ ppm 8.49 (d, J = 2.3 Hz,
(6) Butler, M. A.; Iwasaki, M.; Guengerich, F. P.; Kadlubar, F. F.
Human cytochrome P-450PA (P-450IA2), the phenacetin O-
deethylase, is primarily responsible for the hepatic 3-demethylation
of caffeine and N-oxidation of carcinogenic arylamines. Proc. Natl.
Acad. Sci. U. S. A. 1989, 86, 7696−7700.
1H), 8.39 (dd, J = 9.1, 2.3 Hz, 1H), 7.60 (d, J = 7.8 Hz, 1H), 7.55 (d, J
= 9.1 Hz, 1H), 7.32 (d, J = 7.8 Hz, 1H), 4.92−5.03 (m, 1H), 3.99−
4.06 (m, 2H), 3.86−3.94 (m, 4H), 3.71 (dd, J = 11.2, 9.0 Hz, 2H),
3.10−3.17 (m, 2H), 2.64−2.74 (m, 1H), 2.46 (s, 3H), 1.38 (d, J = 6.1
Hz, 6H), 1.36 (s, 3H), 1.25 (s, 3H).
(7) Fuhr, U.; Doehmer, J.; Battula, N.; Wolfel, C.; Kudla, C.; Keita, Y.
̈
Step 12. A solution of 5-{3-[4-(2,2-dimethyl-1,3-dioxan-5-yl)-9-
methyl-2,3,4,5-tetrahydro-1,4-benzoxazepin-8-yl]-1,2,4-oxadiazol-5-
yl}-2-[(1-methylethyl)oxy]benzonitrile (0.72 g, 1.43 mmol) in THF
(10 mL) was treated with a 2 N HCl aqueous solution (8 mL, 16.0
mmol) and allowed to stand in an open vessel at room temperature for
17 h. The reaction mixture was evaporated to dryness under vacuum.
The solid residue was twice suspended in toluene and coevaporated
under vacuum. The solid was then triturated with Et2O and dried
under vacuum to give 24 (400 mg, 56%) as a white solid. LCMS
Biotransformation of caffein and theophylline in mammalian cell lines
genetically engineered for expression of single cytochrome P450
isoforms. Biochem. Pharmacol. 1992, 43, 225−235.
(8) Ioannides, C.; Lewis, D. F. V. Cytochromes P450 in the
bioactivation of chemicals. Curr. Top. Med. Chem. 2004, 4, 1767−1788.
(9) Kim, D.; Guengerich, F. P. Cytochrome P450 activation of
arylamines and heterocyclic amines. Annu. Rev. Pharmacol. Toxicol.
2005, 45, 27−49.
(10) Ma, Q.; Lu, A. CYP1A induction and human risk assessment: an
evolving tale of in vitro and in vivo studies. Drug Metab. Dispos. 2007,
35, 1009−1016.
(11) Zhou, S.; Wang, B.; Yang, L.; Liu, J. Structure, function,
regulation and polymorphism and the clinical significance of human
cytochrome P450 1A2. Drug Metab. Rev. 2010, 42, 268−354.
(12) Wang, B.; Zhou, S. Synthetic and natural compounds that
interact with human cytochrom P450 1A2 and implications in drug
development. Curr. Med. Chem. 2009, 16, 4066−4218.
(13) Miller, E.; Miller, J. Mechanisms of chemical carcinogenesis.
Cancer 1981, 47, 1055−1064.
(14) Gelboin, H. Benzo[α]pyrene metabolism, activation and
carcinogenesis: role and regulation of mixed function oxidases and
related systems. Physiol. Rev. 1980, 60, 1107−1166.
(15) Nebert, D.; Dalton, T.; Okey, A.; Gonzalez, F. Role of aryl
hydrocarbon receptor-mediated induction of the CYP1 enzymes in
environmental toxicity and cancer. J. Biol. Chem. 2004, 279, 23847−
23850.
1
(method formate): retention time 0.90 min, [M + H]+ = 464.9. H
NMR (400 MHz, DMSO-d6) δ ppm 8.50 (d, J = 2.3 Hz, 1H), 8.36−
8.42 (m, 1H), 7.58 (d, J = 7.8 Hz, 1H), 7.55 (d, J = 9.1 Hz, 1H), 7.24
(d, J = 8.1 Hz, 1H), 4.92−5.03 (m, 1H), 4.36 (t, J = 5.2 Hz, 2H),
3.99−4.08 (m, 2H), 3.90 (s, 2H), 3.45−3.62 (m, 4H), 3.16 (br s, 2H),
2.69−2.80 (m, 1H), 2.45 (s, 3H), 1.38 (d, J = 6.1 Hz, 6H).
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.5b01102.
Molecular formula strings (CSV)
AUTHOR INFORMATION
Corresponding Author
*E-mail: Simon.5.Taylor@GSK.com. Phone: +44 (0) 1438 745
■
(16) Denison, M.; Nagy, S. Activation of the aryl hydrocarbon
receptor by structurally diverse exogenous and endogenous chemicals.
Annu. Rev. Pharmacol. Toxicol. 2003, 43, 309−334.
(17) Westwell, A. The therapeutic potential of aryl hydrocarbon
receptor (AhR) agonists in anticancer drug development. Drugs Future
2004, 29, 479−491.
745.
Notes
The authors declare no competing financial interest.
(18) Matsumura, F.; Puga, A.; Tohyama, C. Biological functions of
the arylhydrocarbon receptor: Beyond induction of cytochrome P450s.
Biochem. Pharmacol. 2009, 77, 473−760.
ACKNOWLEDGMENTS
■
The authors acknowledge Andy Ayrton, Gail Nolan, and Mike
Kelly for insightful discussions and contributions to the strategy
of this program
(19) Guyot, E.; Chevallier, A.; Barouki, R.; Coumoul, X. The AhR
twist: ligand-dependent AhR signaling and pharmaco-toxicological
implications. Drug Discovery Today 2013, 18, 479−486.
(20) Kroon, L. Drug interactions with smoking. Am. J. Health-Syst.
Pharm. 2007, 64, 1917−1921.
ABBREVIATIONS USED
AhR, arylhydrocarbon receptor; THIQ, tetrahydroisoquinoline;
BZ, benzazepine; BZO, benzoxazepine; CYP, cytochrome P450
■
(21) Zevin, S.; Benowitz, N. Drug interactions with tobacco smoking.
An update. Clin. Pharmacokinet. 1999, 36, 425−438.
(22) Demont, E. H.; Andrews, B. I.; Bit, R. A.; Campbell, C. A.;
Cooke, J. W. B.; Deeks, N.; Desai, S.; Dowell, S. J.; Gaskin, P.; Gray, J.
R. J.; Haynes, A.; Holmes, D. S.; Kumar, U.; Morse, M. A.; Osborne,
G. J.; Panchal, T.; Patel, B.; Perboni, A.; Taylor, S.; Watson, R.;
Witherington, J.; Willis, R. Discovery of a selective S1P1 receptor
agonist efficacious at low oral dose and devoid of effects on heart rate.
ACS Med. Chem. Lett. 2011, 2, 444−449.
(23) Harris, R. M.; Andrews, B. I.; Clark, S.; Cooke, J. W. B.; Gray, J.
C. S.; Ng, S. Q. Q. The Fit for Purpose Development of S1P1 Receptor
Agonist GSK2263167 Using a Robinson Annulation and Saegusa
Oxidation to Access an Advanced Phenol Intermediate. Org. Process
Res. Dev. 2013, 17, 1239−1246.
(24) Demont, E. H.; Arpino, S.; Bit, R. A.; Deeks, N.; Desai, S.;
Dowell, S. J.; Gaskin, P.; Gray, J. R. J.; Harrison, L. A.; Haynes, A.;
Heightman, T. D.; Holmes, D. S.; Humphreys, P. G.; Kumar, U.;
REFERENCES
■
(1) Denison, R.; Soshilov, A.; He, G.; DeGroot, D.; Zhao, B. Exactly
the same but different: promiscuity and diversity in the molecular
mechanisms of action of the aryl hydrocarbon (dioxin) receptor.
Toxicol. Sci. 2011, 124 (1), 1−22.
(2) Furness, S. G.; Whelan, F. The pleiotropy of dioxin toxicity-
xenobiotic misappropriation of the aryl hydrocarbon receptor’s
alternative physiological roles. Pharmacol. Ther. 2009, 124, 336−353.
(3) Hankinson, O. Role of co-activators in transcriptional activation
by the aryl-hydrocarbon receptor. Arch. Biochem. Biophys. 2005, 433,
379−386.
(4) Budinsky, R.; LeCluyse, E.; Ferguson, S.; Rowlands, C.; Simon, T.
Human and rat primary hepatocytes CYP1A1 and 1A2 induction with
2,3,7,8-tetrachlorodibenzo-p-dioxin, 2,3,7,8-tetrachlorodibenzofuran,
8255
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256

Journal of Medicinal Chemistry
Article
Morse, M. A.; Osborne, G. J.; Panchal, T.; Philpott, K. L.; Taylor, S.;
Watson, R.; Willis, R.; Witherington, J. Discovery of a brain-penetrant
S1P3-sparing Direct Agonist of the S1P1 and S1P5 receptors Efficacious
at Low Oral Dose. J. Med. Chem. 2011, 54, 6724−6733.
(25) Skidmore, J.; Heer, J.; Johnson, C. N.; Norton, D.; Redshaw, S.;
Sweeting, J.; Hurst, D.; Wall, I.; Ahmed, M.; Rivers, D.; Myatt, J.;
Giblin, G.; Philpott, K.; Kumar, U.; Bit, R. A.; Haynes, A.; Taylor, S.;
Watson, R.; Witherington, J.; Demont, E. H.; Heightman, T. D.
Optimisation of S1P1 receptor agonists: effects of chemotype on
ADMET and efficacy. J. Med. Chem. 2014, 57 (24), 10424−42.
(26) Taylor, S.; Gray, J. R. J.; Willis, R.; Deeks, N.; Haynes, A.;
Campbell, C.; Gaskin, P.; Leavens, K.; Demont, E.; Dowell, S.; Cryan,
J.; Morse, M.; Patel, A.; Garden, H.; Witherington, J. The utility of
pharmacokinetic−pharmacodynamic modeling in the discovery and
optimization of selective S1P1 agonists. Xenobiotica 2012, 42, 671.
(27) Baldwin, S.; Bramhall, J.; Ashby, C.; Yue, L.; Murdock, P.;
Hood, S.; Ayrton, A.; Clarke, S. Cytochrome p450 gene induction in
rats ex vivo assessed by quantitative real-time reverse transcriptase-
polymerase chain reaction (taqman). Drug Metab. Dispos. 2006, 34,
1063−1069.
glutathione S-transferase, albumin and kidney injury molecule-1 with
histopathology and gene expression. J. Appl. Toxicol. 2011, 31 (4),
366−377.
(28) Hu, W.; Sorrentino, C.; Denison, M.; Kolaja, K.; Fielden, M.
Induction of Cyp1a1 Is a Nonspecific Biomarker of Aryl Hydrocarbon
Receptor Activation: Results of Large Scale Screening of Pharmaceut-
icals and Toxicants in Vivo and in Vitro. Mol. Pharmacol. 2007, 71,
1475−1486.
(29) Fujita, Y.; Yonehara, M.; Tetsuhashi, M.; Noguchi-Yachide, T.;
Hashimoto, Y.; Ishikawa, M. B-Naphthoflavone analogs as potent and
soluble aryl hydrocarbon receptor agonists: Improvement of solubility
by disruption of molecular planarity. Bioorg. Med. Chem. 2010, 18,
1194−1203.
(30) Trapani, V.; Patel, V.; Leong, C.-O.; Ciolino, H. P.; Yeh, G. C.;
Trepel, J. B.; Stevens, M. F. G.; Sausville, E. A.; Loaiza-Perez, A. L.
DNA damage and cell cycle arrest induced by 2-(4-amino-3-
methylphenyl)-5-fluorobenzothiazole (5F 203, NSC 703786) is
attenuated in aryl hydrocarcbon receptor deficient MCF-7 cells. Br.
J. Cancer 2003, 88, 599−605.
(31) Aiello, S.; Wells, G.; Stone, E. L.; Kadri, H.; Bazzi, R.; Bell, D. R.;
Stevens, M. F. G.; Matthews, C. S.; Bradshaw, T. D.; Westwell, A. D.
Synthesis and Biological Properties of Benzothiazole, Benzoxazole, and
Chromen-4-one Analogues of the Potent Antitumor Agent 2-(3,4-
Dimethoxyphenyl)-5-fluorobenzothiazole (PMX 610, NSC 721648). J.
Med. Chem. 2008, 51, 5135−5139.
(32) Brinkmann, V. FTY720 (fingolimod) in Multiple Sclerosis:
therapeutic effects in the immune and the central nervous system. Br. J.
Pharmacol. 2009, 158, 1173−1182.
(33) Noguchi, K.; Chun, J. Roles for lysophospholipid S1P receptors
in multiple sclerosis. Crit. Rev. Biochem. Mol. Biol. 2011, 46, 2−10.
(34) Didziapetris, R.; Japertas, P.; Avdeef, A.; Petrauskas, A.
Classification Analysis of P-Glycoprotein Substrate Specificity. J.
Drug Target. 2003, 11, 391−406.
(35) Guyot, E.; Chevalier, A.; Barouki, R.; Coumoul, X. The AhR
twist: ligand-dependent AhR signaling and pharmaco-toxicological
implications. Drug Discovery Today 2013, 18 (9−10), 479−86.
(36) Jewell, W.; Miller, M. Comparison of human and rat metabolism
of Molinate in liver microsomes and slides. Drug Metab. Dispos. 1999,
27 (7), 842−847.
(37) Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J.
Protein measurement with the Folin-Phenol reagents. J. Biol. Chem.
1951, 193, 265−275.
(38) Clarke, S. E.; Jeffrey, P. Utility of Metabolic Stability Screening:
Comparison of In Vitro and In Vivo Clearance. Xenobiotica 2001, 31,
591−598.
(39) Bayliss, M. K.; Bell, J. A.; Jenner, W. N.; Wilson, K. Prediction of
intrinsic clearance of Loxtidine from kinetic studies in rat, dog and
human hepatocytes. Biochem. Soc. Trans. 1990, 18 (6), 1198−9.
(40) Swain, A.; Turton, J.; Scudamore, C. L.; Pereira, I.; Viswanathan,
N.; Smyth, R.; Munday, M.; McClure, F.; Gandhi, M.; Sondh, S.; York,
M. Urinary biomarkers in hexachloro-1:3-butadiene-induced acute
kidney injury in the female Hanover Wistar rat; correlation of α-
8256
DOI: 10.1021/acs.jmedchem.5b01102
J. Med. Chem. 2015, 58, 8236−8256