Invention of MK-8262, a Cholesteryl Ester Transfer Protein (CETP) Inhibitor Backup to Anacetrapib with Best-in-Class Properties ()

Article abstract of DOI:10.1021/acs.jmedchem.1c00959

Cholesteryl ester transfer protein (CETP) represents one of the key regulators of the homeostasis of lipid particles, including high-density lipoprotein (HDL) and low-density lipoprotein (LDL) particles. Epidemiological evidence correlates increased HDL and decreased LDL to coronary heart disease (CHD) risk reduction. This relationship is consistent with a clinical outcomes trial of a CETP inhibitor (anacetrapib) combined with standard of care (statin), which led to a 9% additional risk reduction compared to standard of care alone. We discuss here the discovery of MK-8262, a CETP inhibitor with the potential for being the best-in-class molecule. Novel in vitro and in vivo paradigms were integrated to drug discovery to guide optimization informed by a critical understanding of key clinical adverse effect profiles. We present preclinical and clinical evidence of MK-8262 safety and efficacy by means of HDL increase and LDL reduction as biomarkers for reduced CHD risk.

Full text of DOI:10.1021/acs.jmedchem.1c00959

pubs.acs.org/jmc
Article
Invention of MK-8262, a Cholesteryl Ester Transfer Protein (CETP)
Inhibitor Backup to Anacetrapib with Best-in-Class Properties
Petr Vachal,* Joseph L. Duffy, Louis-Charles Campeau, Rupesh P. Amin, Kaushik Mitra,
Beth Ann Murphy, Pengcheng P. Shao, Peter J. Sinclair, Feng Ye, Revathi Katipally, Zhijian Lu,
Debra Ondeyka, Yi-Heng Chen, Kake Zhao, Wanying Sun, Sriram Tyagarajan, Jianming Bao,
Sheng-Ping Wang, Josee Cote, Concetta Lipardi, Daniel Metzger, Dennis Leung, Georgy Hartmann,
Gordon K. Wollenberg, Jian Liu, Lushi Tan, Yingju Xu, Qinghao Chen, Guiquan Liu,
Robert O. Blaustein, and Douglas G. Johns
Cite This: https://doi.org/10.1021/acs.jmedchem.1c00959
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Cholesteryl ester transfer protein (CETP) represents one of the key
regulators of the homeostasis of lipid particles, including high-density lipoprotein (HDL)
and low-density lipoprotein (LDL) particles. Epidemiological evidence correlates increased
HDL and decreased LDL to coronary heart disease (CHD) risk reduction. This
relationship is consistent with a clinical outcomes trial of a CETP inhibitor (anacetrapib)
combined with standard of care (statin), which led to a 9% additional risk reduction
compared to standard of care alone. We discuss here the discovery of MK-8262, a CETP
inhibitor with the potential for being the best-in-class molecule. Novel in vitro and in vivo
paradigms were integrated to drug discovery to guide optimization informed by a critical
understanding of key clinical adverse effect profiles. We present preclinical and clinical
evidence of MK-8262 safety and efficacy by means of HDL increase and LDL reduction as
biomarkers for reduced CHD risk.
notion that increasing HDL-C, or increasing the number of
circulating HDL particles, could confer cardiovascular protec-
tion. Epidemiological studies and numerous statin clinical trials
strongly demonstrated that elevated LDL-cholesterol (LDL-C)
was positively associated with elevated cardiovascular risk.
Figure 1 depicts the role of CETP in mediating transfer of
neutral lipids (i.e., cholesteryl ester and triglycerides) between
HDL and LDL particle types. One hypothesized mode of
CETP’s mediation of neutral lipid transfer is that CETP serves as
a “tunnel” through which these particles migrate.12 When this
tunnel is closed, as when CETP is inhibited, cholesteryl ester and
triglycerides can no longer equilibrate. Inhibition of CETP then
results in an increase in the cholesteryl ester content of HDL
particles (manifest as an increase in plasma HDL-C) and in a
concomitant decrease in the level of circulating LDL-C. Since
the emergence of the earliest clinical-stage CETP inhibitors (i.e.,
torcetrapib), further genetic evidence supported the potential
for CETP inhibition, and this may have fueled the development
INTRODUCTION
■
Cardiovascular disease represents a major and persistent global
health risk, accounting for approximately 31% of all deaths
worldwide in 2019. Significant gains have been made toward
reversing this risk, including standard-of-care pharmaceuticals
such as statins and β blockers. Additional successful
interventions include the surgical implantable cardioverter
defibrillator (ICD) deviceswith and without resynchroniza-
tion therapyand improved lifestyle choices due to increased
health awareness. However, cardiovascular disease, mainly
coronary heart disease (CHD) and stroke, remains the leading
cause of death.1 Additional clinically relevant mechanisms
combined with current standard of care bear the potential for
dramatically impacting human mortality caused by cardiovas-
cular disease. One of the key contributors to cardiovascular
disease is CHD caused in part by atherogenic lipoprotein
deposition and inflammatory infiltration. As such, the further
reduction of CHD presents an attractive intervention for the
reduction of overall cardiovascular disease risk.
Received: May 28, 2021
The primary rationale for pursuing cholesteryl ester transfer
protein (CETP) as a target to combat CHD was rooted in
epidemiological evidence associating elevated HDL cholesterol
(HDL-C) with reduced cardiovascular risk.2−6 Studies on HDL
particle biology over the past few decades have demonstrated
anti-atherogenic functionalities of HDL,7−11 supporting the
https://doi.org/10.1021/acs.jmedchem.1c00959
© XXXX American Chemical Society
J. Med. Chem. XXXX, XXX, XXX−XXX
A

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 1. HDL and LDL metabolism; reverse cholesterol transport. Only LDL can deposit into macrophages but both HDL and LDL are cleared
through the liver. CETP facilitates equilibration between HDL and LDL; CETP inhibition leads to a significant increase of circulating HDL-C and
reduction of LDL-C.
Table 1. Summary of Clinical Results for CETP Inhibitors
of additional CETP candidates, including dalcetrapib, evace-
trapib, and anacetrapib.13−15
The clinical outcomes trials of CETP inhibitors were designed
reduction of LDL-C beyond statin therapy would further reduce
the risk of CHD. At the time of the initiation of the backup
molecule for anacetrapib, the CETP outcomes trials had either
failed (torcetrapib) or were in progress. However, we now have
to test whether the elevation of HDL-C and/or the added
B
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
better insight into the beneficial effect of CETP inhibition
beyond standard of care. The results of the REVEAL outcomes
trial demonstrated that the addition of the CETP inhibitor
anacetrapib to atorvastatin for a median duration of 4.1 years
yielded an additional 9% relative reduction in CHD compared to
atorvastatin therapy alone.16,17
that maintained the positive attributes of anacetrapib but are
designed with a lower log D would be devoid of significant
adipose retention, provide a more tractable terminal half-life,
afford similar oral exposure in the fed or fasted state, and be
more amenable to conventional formulation. Finally, the backup
molecule should utilize multiple pathways for in vivo
metabolism and excretion as well as be devoid of pregnane X
receptor (PXR) inhibition to avoid drug−drug interaction
(DDI) potential. It is noteworthy that anacetrapib inhibited
PXR with an IC50 of 3.5 μM in vitro. This finding provided the
rationale for conducting an early safety clinical study aimed at
de-risking the potential for DDI with anacetrapib. This safety
trial, as well as any of the subsequent large clinical trials,
including Phase 3 trials REVEAL and DEFINE, did not reveal
any DDI-related issues for anacetrapib.33
Table 1 summarizes key clinical results for CETP inhibitors18
with two notable time points: row 6 of Table 1 shows current
development status, while row 3 shows the status at the time
when we were initiating a program aimed at improving on
anacetrapib, providing for relevant context to the discovery
program described therein. At the start of this backup program,
Pfizer had terminated Phase 3 clinical trial (ILLUMINATE) due
to statistically significant adverse effects (AEs) as measured by
an excess of cardiovascular (CV) deaths that may have been due
to blood pressure and plasma aldosterone.19 It was not clear at
that time whether the safety findings from the ILLUMINATE
trial resulted from inhibition of CETP and therefore were
mechanism-based or were compound-specific. At the same time,
Roche was conducting a Phase 3 cardiovascular outcomes trial
(dal-OUTCOMES)20 with dalcetrapib, dubbed a CETP
“modulator” rather than “inhibitor” due to presumed covalent,
irreversible modification of one or more cysteine residues of the
CETP target.21 Lilly had just completed a Phase 2b trial for
evacetrapib but the results were not yet publicly disclosed.
Mitsubishi-Tanabe (TA-8995) was holding off further develop-
ment of their clinical candidate, now known as obicetrapib, after
completing Phase 1. The asset TA-8995 has since been acquired,
first by Dezima (DEZ-001), then later by Amgen (AMG-899),
and most recently by New Amsterdam Pharma. Our clinical
candidate anacetrapib was being studied in the first of two Phase
3 trials (DEFINE)22 aimed at demonstrating the safety of CETP
inhibition given the safety signal noted with torcetrapib. A
combination of additional preclinical characterization and data
from the DEFINE trial provided support to the notion that
torcetrapib-related adverse events (AEs) were not CETP
mechanism-based, paving a safe path to a larger Phase 3 trial
for anacetrapib (REVEAL)23 aimed at comparing clinical
outcomes. The Phase 3 results for other CETP inhibitors are
also presented in Table 1.24−27
RESULTS AND DISCUSSION
Figure 2 shows an assay funnel (aka, research operating plan,
ROP) designed to enable triage and optimization of molecules
■
Figure 2. Bioassay hierarchy, aka Research Operating Plan (ROP), for
optimization and triage of molecules.
toward the ideal profile defined above. Of note are the following
features: first, early Tier 2 high throughput in vitro assessment of
physical−chemical properties was demonstrated to correlate
well with more sophisticated, lower throughput in vivo food
effect and tissue accumulation assays (Tier 7). Early de-risking
of the blood pressure signal in a single-dose Tier 5 assay
provided a relatively high-throughput early triage well before the
formal safety dose-dependent risk assessment in Tier 10, saving a
significant amount of time and resources for key molecules. A
gene-signature-based risk assessment for drug-induced liver
injury (DILI)34 was performed for key molecules in Tier 8−9.
Structurally diverse CETP inhibitors have shown clinical
efficacy for elevation of HDL and reduction of LDL (Table 1).
Additional distinct preclinical lead series have been reported
from multiple laboratories, including our own.35,36 We
concurrently advanced the structure−activity relationship
(SAR) study of the oxazolidinone class of CETP inhibitors
characterized by anacetrapib to determine whether further
improvements in this class were attainable toward a best-in-class
profile.
A meaningful backup molecule would need to maintain the
full CETP inhibition observed clinically for anacetrapib
(DEFINE) and evacetrapib (ACCELERATE). The backup
candidate must also differentiate from torcetrapib on safety
(ILUMINATE) as well as from dalcetrapib on both mechanism
of action and absorption, distribution, metabolism and excretion
(ADME). Additionally, evidence was emerging that the long
terminal half-life in plasma of anacetrapib seen after the drug was
discontinued may have resulted from its accumulation into, and
slow washout from adipose tissue.28−32 Despite these
pharmacokinetic concerns, anacetrapib appeared safe in
preclinical toxicity evaluations. In clinical studies, anacetrapib
was well tolerated with no significant AEs in the two phase 3
clinical trials (DEFINE and REVEAL), which together enrolled
over 30 000 patients. Nevertheless, we postulated that a viable
backup molecule should be devoid of extended tissue
accumulation and retention. Additionally, the ideal backup
molecule would show no change in drug exposure between fed
and fasted subjects (food effect). The backup should be
amenable to conventional formulation enabling fixed-dose
combination with a range of statins, which are the standard of
care for CHD and on top of which CETP inhibitors would
ideally be co-administered. We hypothesized that compounds
Anacetrapib has established an outstanding preclinical and
clinical safety profile, and robust translational pharmacokinetics.
In clinical lipid studies, the efficacy profile is consistent with
several other maximally efficacious candidates (Table 1). Our
clinical experience with anacetrapib, however, identified features
of the profile that could be further improved. Anacetrapib is a
C
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
polyfluorinated lipophilic compound with poor aqueous
solubility (<1 μM in pH 2 and pH 7 phosphate buffer) and
very low calculated total polar surface area (TPSA = 44 Å2).37
The log D for the molecule is too high to be accurately measured
by an isocratic high-performance liquid chromatography
(HPLC) method at pH 7.3. Consistent with these properties,
anacetrapib exhibits a positive food effect, with significantly
enhanced oral bioavailability when administered with a high-fat
meal.38 The low aqueous solubility of anacetrapib requires
clinical formulation as a 12% active pharmaceutical ingredient in
a hot-melt extrusion, which could impede a more convenient
fixed-dose combination with statin drugs.39 Anacetrapib is
metabolized primarily by the CYP3A4 enzyme in humans, and
was found to induce CYP3A4 when incubated with human
hepatocytes.40 Therefore, a compound with diminished agonism
of the CYP-inducing PXR would be optimal.41 Finally,
anacetrapib distributes into white adipose tissue in both
preclinical species and in humans.42,43 This biodistribution
leads to a biphasic pharmacokinetic elimination profile with a
long terminal half-life. We hypothesized that the advantageous
profile of the oxazolidinone CETP inhibitor class could be
further improved by an anacetrapib derivative that preserved the
potency of anacetrapib while exhibiting decreased lipophilicity
as measured by TPSA and log D.
Table 2. Modification of Ring A (RA) Substituents on
Oxazolidinone CETP Inhibitors
b
A recent review has summarized the SAR strategy that led to
the discovery of anacetrapib.44 This SAR experience guided our
lipophilicity investigation toward four regions of the molecule as
outlined in Figure 3.45,46 The 3,5-bistrifluoromethyl substituent
Figure 3. Labeled regions of the oxazolidinone lead class investigated
for further optimization to explore improvements on anacetrapib.
a
Unless otherwise noted, IC50 values are from >2 assays with standard
b
deviation ≤60% of the reported mean. Single assay determination.
in Ring C is common to several reversible CETP inhibitors that
have been pursued in the clinic (Table 1), and this substituent
was maintained for the first round of investigation. The
oxazolidinone “Core” constrains the pharmacophore substitu-
ents as a rigid scaffold and was also left intact for the initial
investigation. We reasoned that the Ring B and Ring A regions
could be amenable to incorporation of more hydrophilic
substituents while preserving CETP inhibitory activity.
A broad substitution investigation of the Ring A region was
carried out, and representative compounds from this study are
provided in Table 2. The compounds were tested for their ability
to inhibit CETP-mediated cholesterol ester transfer in 2%
human serum using a fluorogenic inhibition assay (CEF) that has
been described previously.47 The compounds were also
evaluated for their ability to activate PXR, both for their
inflection point (EC50) and maximal activation at 10 μM. As the
HPLC log D for anacetrapib and close analogues are too high to
be accurately measured by isocratic HPLC methods, the
molecules were evaluated for the substituent impact on the
calculated TPSA as a surrogate for log D.
Anacetrapib (1) is a potent inhibitor of CETP but retains
appreciable activation of PXR and low TPSA. We utilized the
previously reported des-F anacetrapib (2) as a surrogate for
anacetrapib (1) for novel experiments during these studies due
to the ongoing clinical evaluation of anacetrapib. Replacement
of the methoxy substituent with Cl (3) diminished PXR activity,
but also CETP potency and TPSA. Replacement of the
isopropyl substituent with a phenyl (4) as substituent “R”
significantly reduced the CETP potency, but also fully ablated
PXR activity. We found that CETP inhibitory activity could be
restored by the addition of an ortho-fluoro substituent (5) or
methyl (6) on the R-phenyl ring without a concomitant increase
in activation of PXR. The TPSA was significantly increased by
the incorporation of an ortho carboxylic acid (7), but with
significant loss of CETP potency. The increased TPSA was
retained and CETP potency was partially restored by trans-
position of the carboxylate to the para position (8). Further
incorporation of the ortho methyl substituent provided 9, which
improved on anacetrapib in several key parameters. This
D
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 3. Solubility, Pharmacokinetics in C57Bl/6 Mice, and Pharmacodynamics in B6-TG Mice
c
d
e
dose IV/PO
(mg/kg)
Cmax
(μM)
oral AUCN
(μM·h·kg/mg)
IV CLb
Δ ↑HDL-C
a
b
Cpd
#
soly pH 7 (μM)
(mL/min/kg)
IV t1/2 (h) F (%)
(mg/dL)
1
2
9
<1
<1
16
43
0.6/2
1/2
1/2
0.28
0.16
0.43
0.39
0.36
0.29
0.86
1.8
11
19
18
3.6
4.9
2.5
6.2
15
17
65
43
35
22
31
30
10
1/2
5.6
a
b
Each value is the mean from at least two mice, with standard deviation (SD) < 30% of the mean. Solubility was measured in pH 7 phosphate
c
d
buffer. IV dose formulated in DMSO:PEG:water (10:55:35); by mouth dose formulated in DMSO:chremophor:saline (2:4:94). Whole blood
total clearance. Mean increase in HDL cholesterol in B6-TG (CETP) mice following BID oral doses at 10 mg/kg.
e
compound exhibited excellent CETP potency, significantly
higher TPSA, and was devoid of PXR activation. Replacement of
the ortho methoxy group in 9 with a chloro-substituent (10)
retained CETP potency in contrast to the decrease observed
between the analogous 1 and 3. The higher calculated TPSA
values for 9 and 10 translated to measurable isocratic HPLC
log D values of 5.5 and 6.2, respectively.
The carboxyl derivatives 9 and 10 were further profiled.
Anacetrapib (1), des-fluoro anacetrapib (2), as well as both 9
and 10 were devoid of significant inhibitory activity against ion
channels Nav 1.5 and Cav1.2 located in the heart and vasculature.
̀
However, both 9 and 10 exhibited concerning human Ether-a-
go-go-Related Gen (hERG) block, with IC50 values of 10 and 5.0
μM, respectively.48 The compounds 1, 2, 9, and 10 have an
undetectable free fraction in plasma from mouse, rat, and
human. The pharmacokinetic properties were evaluated in
C57Bl/6 mice, and the results are shown in Table 3. In these
studies, 10 exhibited the lowest total clearance following an
intravenous (iv) dose and a highest oral dose-normalized area
under the curve (AUCN). The compounds were also evaluated
for their ability to elevate HDL cholesterol in transgenic C57Bl/
6 mice expressing cyno CETP (B6-transgenic, B6-TG mice),
using a BID protocol described previously.46 In this
pharmacodynamic screening assay, both 9 and 10 showed
comparable efficacy to anacetrapib (1) and were slightly
superior to des-fluoro anacetrapib (2). Compound 10 afforded
higher aqueous solubility at neutral pH, which could aid in
eventual formulation studies. Therefore, this analogue was
evaluated in greater depth.
We postulated that 10, with a higher TPSA than anacetrapib
(1), would result in relatively lower tissue accumulation. This
was tested by comparing the terminal half-life of 1 and 10
following 7-day dosing in lean B6-TG mice. The compounds
were administered orally on a once-a-day basis (QD;
anacetrapib at 200 mg/kg/day, and 10 at 100 mg/kg/day to
match final plasma concentrations). The blood levels of the
compounds were measured for 240 h following the final dose.
The results are illustrated in Figure 4. From this experiment, the
t1/2 of anacetrapib (1) was extrapolated to approximately 121
( 36) h, whereas the t1/2 of 10 was 83 ( 46) h. This experiment
indicated that CETP inhibitors with reduced lipophilicity may
provide a simplified clinical elimination profile as compared to
anacetrapib (1).43 However, compounds with even lower
lipophilicity would be required to fully establish this relation-
ship.
Figure 4. Blood levels of 1 and 10 in C57Bl/6 mice following 7 days of
oral dosing at 200 mg/kg for anacetrapib (1) (n = 10) and 100 mg/kg/
day for 10 (n = 12).
aldosterone 60 min after an oral dose of CETP inhibitor (see the
Supporting Materials (SI)). This assay was used to quantify the
exposure window between the no-effect level (NOEL) for
aldosterone release relative to the minimum efficacious dose
(MED) for full plasma CETP inhibition in CETP transgenic
mice. We define this window as the preclinical aldosterone
margin (ALDO margin). Note that 100% CETP inhibition is
not thought to be achievable due to the ex vivo nature of the
CETP activity assays, which rely on measurement of labeled
lipid enrichment in exogenously added lipoprotein fractions.50
Therefore, we identify >75% plasma CETP inhibition in the ex
vivo assay as full inhibition. For control experiments, we
collected results from preclinical studies with torcetrapib and
des-fluoro anacetrapib (2), and the aggregated results are
illustrated graphically in Figure 5. Following a 100 mg/kg dose,
torcetrapib afforded maximal CETP inhibition in mice
expressing human CETP, where expression is driven by the
natural flanking region promoter (NFR-TG mice).51 Following
a 30 mg/kg dose, torcetrapib exhibited significant aldosterone
elevation in B6 mice, providing no margin for efficacy over
aldosterone release. This is consistent with the clinical
observation for torcetrapib from the ILLUMINATE trial.52
On the other hand, 2 exhibited no significant aldosterone release
at plasma levels of 8.2 μM (100 mg/kg oral dose), whereas
maximal plasma CETP inhibition (in cyno-TG mice) could be
achieved at plasma levels of 1.1 μM (1 mg/kg dose), affording an
aldosterone margin of greater than 7.5×.
A significant concern in the clinical development of CETP
inhibitors has been the potential for off-target elevation in
plasma aldosterone as seen with torcetrapib.49 We therefore
evaluated candidate compounds for aldosterone release in a
preclinical screening assay. We optimized a mass-spectrometry
(MS)-based assay of mouse plasma to quantify circulating
The use of cynomolgus transgenic (cyno-TG) mice or NFR-
TG mice for control experiments (Torcetrapib or 2) was
determined by the availability of TG mice at the time of profiling
but does not impact the interpretation of maximum CETP
inhibition. For the investigation of novel compounds, the
aldosterone margin experiment was harmonized to utilize
E
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 5. Definition of Aldosterone (ALDO) margin: Plasma concentration ratio of the no-effect level (NOEL) for compound-induced aldosterone
elevation (μM) divided by the plasma concentration (μM) at which maximal compound-induced CETP inhibition was observed, in the indicated
species (NFR transgenic mice, cyno-tg mice, or C57/B6 mice).
Figure 6. ALDO margin in NFR mice for evacetrapib and 10.
CETP-TG mice for exposure, aldosterone measurements, and
plasma CETP inhibition in a single animal. Using this standard
protocol, evacetrapib exhibited an aldosterone margin of 5.5×
between the NOEL for aldosterone release and maximal CETP
inhibition (Figure 6). These results are consistent with recent
clinical data showing no significant elevation of aldosterone
following fully efficacious doses of evacetrapib in the
ACCELERATE trial.53 Compound 10 exhibits a lower
aldosterone margin than evacetrapib of approximately 4.4×.
During the optimization studies outlined here, we used the
aldosterone margin as a relative measure to prioritize molecules
but not necessarily as an absolute numeral. Given the observed
clinical safety of evacetrapib, however, we sought to significantly
exceed the 5.5× aldosterone margin observed with this inhibitor
in any optimized backup compound to anacetrapib.
The discovery of 9 and 10 established that compounds from
the oxazolidinone class with significantly greater solubility and
TPSA could maintain the favorable CETP potency and improve
the pharmacokinetic (PK) properties of anacetrapib (1). We
sought to further extend these findings by incorporating
heterocycles into the Ring A and Ring B regions of the molecule.
The results of these studies are shown in Table 4. For this phase
of the program, we improved the primary CETP inhibition assay
to the 3[H]-CE transfer method (CET) that has been
described.54 This method allowed the evaluation of compounds
in the presence of 95% human serum, and 1, 2, and 10 are
included in Table 4 with this method for comparison. The hERG
inhibitory potency is also listed, as this activity needed to be
further reduced compared to 10. Finally, the compounds were
evaluated for the elevation of aldosterone in NFR-TG mice 60
min following an oral dose. The NOEL for statistically
significant aldosterone elevation is provided in terms of both
dose and blood level for the compound. In several cases, a
further dose-titration was completed to the point of significant
aldosterone elevation. In these experiments, the aldosterone
percent increase (as compared to vehicle), dose, and exposure
are also provided for the lowest dose affording a significant
aldosterone increase (effect level).
Incorporating the 2-methyl-4-carboxyphenyl substituent “R”
from 10 into the compounds effectively mitigated the PXR
activation for all new entities (EC50 > 30 μM), and that
substituent was included in all compounds in this series. We
discovered that heteroatoms could be incorporated into the RB−
RA region of the molecule without significant loss of intrinsic
potency. Incorporation of a pyridyl nitrogen in the 3-position of
RB provided 11, which maintained CETP potency while
significantly reducing the log D and hERG activity. The
compound also afforded no significant aldosterone release at
higher exposure than the effect level for 10. Transposition of the
pyridyl nitrogen to the 5-position afforded 12, which was a more
potent CETP inhibitor but also significantly more potent for
aldosterone release in mice. The RA pyridyl isomer 13 also
maintained CETP potency but was a potent inhibitor of hERG.
A combination of the pyridyl regioisomers provided 14−16. The
derivative 14 was the most potent CETP inhibitor of the series,
but also maintained activity as an hERG inhibitor. Derivatives 15
and 16 exhibited lower potency as CETP inhibitors but were
devoid of hERG inhibition. The pyrimidine derivatives 17 and
18 were also somewhat less potent CETP inhibitors but were
also devoid of significant hERG inhibition. The amino-
pyrimidine derivatives 19−23 maintained the lower log D
values from this series as well. The most active of these,
dimethylaminopyrimidine 19, is a potent CETP inhibitor and
exhibited the highest ALDO NOEL of this series of compounds
F
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 4. Heterocycle Incorporation in Ring A−B (RA−B
)
Table 4. continued
e
Substituents on Oxazolidinone CETP Inhibitors
c
hERG inhibition was measured by binding displacement of 35S-
d
labeled MK-0499, a well-characterized hERG blocker. Measured in
NFR-TG mice, with standard deviation <40% of the mean (n = 8).
e
Measured in cyno-TG mice (n = 8).
(>91 μM whole blood exposure), albeit with some hERG
activity as well. The methoxypyrimidine 24 was the most
hydrophilic compound of the series, with a log D of 4.2. The
ethylpyrimidine derivative 25 was relatively more potent as a
CETP inhibitor, but also exhibited some hERG inhibition. The
hERG inhibition was mitigated by incorporating the electron-
withdrawing difluoroethyl substituent (26), which maintained
the potency as a CETP inhibitor.
Key compounds from each subseries in Table 4 were profiled
further to determine the concentration required for full CETP
inhibition in NFR-TG mice, and the results are provided in
Table 5. This potency was used with the NOEL for aldosterone
release (Table 4) to provide the aldosterone margin. The 24 h
shake-flask solubility of the compounds was also evaluated in a
pH 7 phosphate buffer to provide an early indication of
formulation feasibility. Additionally, we recognized that the
aliphatic carboxylic acid substituent incorporated from 10 could
be a pharmacophore recognized by the organic anion trans-
porting polypeptides (OATP) transporter. The isoform
OATP1B1 is known to transport atorvastatin, and inhibition
of this isoform by a CETP inhibitor could result in drug−drug
interactions on coadministration.55 Therefore, several com-
pounds were assessed for OATP1B1 inhibition.
From the compounds listed in Table 5, all achieved maximal
CETP inhibition in vivo in NFR-TG mice. The least potent
compounds, 10, 17, and 24, required roughly 1 μM blood
exposure for complete inhibition, and this was similar to des-
fluoro anacetrapib 2. While 12 and 14 exhibited high levels of in
vivo CETP potency, the compounds induced significant
aldosterone elevation at low exposure as well (10 and 12 μM,
respectively, Table 4). Compounds 15, 21, and 26 maintained
potent CETP inhibition in vivo and afforded a wide aldosterone
margin. The methoxypyrimidine derivative 24 afforded the
lowest log D (Table 4) and highest solubility (Table 5) of this
series, and inhibited CETP in vivo. Of these final three
compounds, 15 and 26 were devoid of hERG inhibition as
well. However, we found that all of the carboxyl-containing
compounds tested exhibited similar potency for OATP1B1
inhibition in vitro; the significance of this finding is discussed
further below.
The RB−RA dimethylaminopyrimidine scaffold represented
by 19 was utilized for further investigation, owing to the superior
exposure, solubility, efficacy, and aldosterone margin of this
compound compared to others in the series above. Using this
pharmacophore, we next sought to further reduce the lipophilic
burden caused by the bis(trifluoromethyl)phenyl moiety
present in anacetrapib (Ring C in Figure 3). Substantial effort
was devoted to this SAR, and representative compounds are
shown in Table 6, where the comparable data for 19 are
represented for comparison.
As compared to 19, compounds 27−37 were successful in
reducing lipophilicity while maintaining significant CETP
inhibition. For example, the pyridyl derivative 36 reduced the
HPLC log D by 1.6 as compared to 19, while maintaining CETP
inhibition only 3× below anacetrapib 1 (Table 4). Compounds
27 and 31−33 maintained CETP inhibitory activity comparable
a
Unless otherwise noted, IC50 values are from >2 assays with standard
b
deviation ≤40% of the reported mean. Single assay determination.
G
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 5. Solubility, CETP Inhibition and ALDO Margin in NFR-TG Mice, and OATP1B1 Inhibition for CETP Inhibitors
a
max CETP Inh
Δ ↑HDL-C
mg/dL
b
c
Cpd #
% Inh
μM (mg/kg)
μM
ALDO margin
soly pH 7 (μM)
OATP1B1 % Inh at 500 nM
0
d
2
80
1.1 (1)
1.4 (1)
36
1.0
≥8.2×
4.4×
≥3.2×
85×
≥130×
32×
25×
≥130×
≤30
≥110×
<2
43
29
9
5
6
43
8
52
20
10
12
14
15
17
19
21
24
26
81
86
70
75
76
78
80
≤0.30 (≤0.3)
0.13 (0.15)
≤0.15 (≤0.1)
0.87 (1)
0.24 (0.1)
≤0.28 (≤0.3)
0.90 (0.3)
≤0.40 (≤0.3)
67
65
79
68
83
52
28
36
1.3
1.4
2.0
e
66
80
72
a
b
Measured in B6 cyno-TG mice 4 h after a single 30 mg/kg oral dose, with standard deviation ≤10% of the mean (n = 9). CETP inhibition was
c
determined in NFR-TG mice 60 min following an oral dose, as compared to vehicle, with standard deviation ≤40% of the mean (n = 8). Solubility
d
e
was determined by 24 h shake-flask method in phosphate buffer at 25 °C. Measured in B6 cyno-TG mice (n = 8). Compound CETP inhibition
was only examined at 0.3 mg/kg in vivo.
to 19, but also maintained a comparable block of hERG.
Unfortunately, compared to 19, all derivatives exhibited an
increased stimulation of the aldosterone signal at comparable
plasma concentration. Importantly, the trifluoromethoxy
regioisomer 29 maintained comparable aldosterone stimulation
to 28, despite the complete lack of CETP inhibitory potency of
29. These results narrowed the prospects of being able to replace
the lipophilic bis(trifluoromethyl)phenyl substituent and
refocused the SAR effort to balance the CETP activity,
aldosterone risk and physical−chemical properties to other
parts of the pharmacophore. Therefore, our further evaluation
compared compounds that maintained the bis(trifluoromethyl)-
phenyl substituent present in anacetrapib.
The most advanced molecule, 15, has been extensively
profiled in a panel of preclinical safety assessment assays: no
significant safety signal was detected, and the molecule provided
a wide aldosterone margin (>130×), as well as a clean off-target
profile. As such, compared to close anacetrapib analogues, 15
represents the most advanced preclinical derivative of the series
with improved physical−chemical properties. However, 15
maintained a few potential risks common to both itself and
anacetrapib, which are discussed below.
the off-target activity. Theoretically, as depicted in Figure 7,
there are three ways to rigidify the core of anacetrapib. We
synthesized several examples of each design and found that two
of the three designs (Designs A and B) were detrimental to the
CETP activity, and several corresponding derivatives lost
virtually all inhibitory activity. However, Design C proved to
be tolerated for CETP inhibitory properties. The mixture of
stereoisomers of Design C were determined to be active in the
CETP inhibition assay. Subsequent extensive purification
established that the observed activity was solely due to a single
stereoisomer of Design C while all of the remaining 7
stereoisomers were inactive. Since our computational ap-
proaches were unable to predict the desired stereochemistry in
silico, we resorted to synthesizing and separating all 8 isomers to
identify the single one that was active (38). Isomer 38 inhibited
CETP in vitro at levels comparable to anacetrapib (85 vs 45 nM,
respectively). But unlike anacetrapib, 38 lacked the undesired
PXR inhibition (>30 vs 3 μM, respectively). The first
unoptimized fit-for-purpose synthesis of the complex bicyclic
Design C utilized to prepare each of the 8 isomers is outlined in
Scheme 1.
Encouraged by the initial data on 38, we explored isosteric
derivatives of the novel rigid bicyclic pharmacophore core.
Several designs are represented in Figure 8 (see the Supporting
Information for detailed synthetic procedures for each
derivative). Formal substitution of oxygen for sulfur-generated
molecule 46 with an in vitro CETP inhibition of 170 nM, which
is less than half as potent as 38. Other modifications turned out
to be more detrimental to CETP inhibition. Lactam 47, lactone
50, and thiazolidinone 51 only displayed moderate to low CETP
inhibition in vitro. Both urea and sulfonylurea derivatives 48 and
49 warranted further investigation and are the subject of a
separate communication. However, overall, neither derivative
proved to be advantageous over 38.56 A series of spiro-
analogues, represented here by 52, was designed with the idea of
invoking central core rigidity through conformational rather
than structural constraints. Spiro-analogue 52 displayed good
CETP activity in vitro; however, we deprioritized 52 due to
increased PXR activity in vitro compared to 38 (PXR EC50 for
52 was 2.6 μM while no significant activity was observed for 38
up to 30 μM). We speculate that lesser overall structural rigidity
of 52 compared to 38 was the underlying reason for residual
PXR off-target activity consistent with monocyclic-core series.
In conclusion, none of the central core substitutions led to an
Anacetrapib is a very effective CETP inhibitor suitable for QD
dosing with a clean off-target profile. Only a few potential
liabilities have been identified; physical−chemical properties
were the presumed root cause for several concerns for
anacetrapib. The food effect observed preclinically and clinically
and the potential for tissue accumulation, most notable in
adipose tissue, began to emerge from chronic dosing studies.
Low solubility and crystallinity posed an additional challenge for
development. Ideally, to enable the option of weekly dosing
(QW) without the risk of accumulation, an improved molecule
would have an extended half-life with no tissue accumulation.
Furthermore, anacetrapib exhibited a modest level of PXR
inhibition; it is noteworthy, however, that in vitro PXR activity
of anacetrapib did not translate into any adverse effects in the
clinic. We reasoned that all of these properties may be addressed
by further rigidifying the modestly flexible central core
pharmacophore conformation of anacetrapib. We postulated
that, if the productive conformation responsible for CETP
inhibition is favored, a lower overall dose may be sufficient for
maintaining full intrinsic inhibition. At the same time, off-target
PXR inhibition had been correlated to structural flexibility of the
inhibitor and hence a more rigid core was less likely to display
H
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
e
Table 6. Replacements for Ring C in Oxazolidinone CETP Inhibitors
a
b
Unless otherwise noted, IC50 values are from >2 assays with standard deviation ≤40% of the reported mean. hERG inhibition was measured by
c
binding displacement of 35S-labeled MK-0499, a well-characterized hERG blocker. Percent circulating aldosterone vs vehicle (100%) measured in
C57/B6 mice 1 h after a 100 mg/kg dose. Whole blood exposure measured 1 h after dose. Compound was dosed 2 × 100 mg/kg at t = 0, 1 h,
with aldosterone and exposure measured 1 h after final dose, to achieve comparable exposure.
d
e
improvement over 38, prompting a narrowing of our focus on
this bicyclic oxazolidinone core.
available aldehyde 53, this approach takes advantage of two
stereoselective transformations: Ellman diastereoselective alky-
lation57 yielding the desired amine 55 in 4:1 selectivity, followed
by a highly enantioselective, albeit low-yielding, Shi epoxida-
tion.58 The final cyclization of the synthetic route led to a partial
epimerization of the benzylic stereocenter of 58. While 58 could
easily be isolated as a single isomer, a combination of low-
yielding Shi epoxidation and epimerization in the final step led to
a significant yield erosion. Nevertheless, the second-generation
route provided access to a highly functionalized advanced
intermediate 58 in 6 steps and a 10% overall yield. This route
enabled an initial SAR of the bicyclic pharmacophore as well as a
scale-up of 38 needed for additional profiling in vitro and in vivo
in the ROP.
The fit-for-purpose synthetic route (Scheme 1) utilized for
the preparation of 38 and its isomers was not suitable for broader
SAR investigation. The cumbersome route limited material
supply for 38 and its key analogues, hindering further SAR
optimization and profiling deeper into the ROP. We strive for
chemical synthesis to ensure that we never restrain drug
discovery cycle times. A significant effort was devoted to devise a
flexible, high-yielding synthesis that would be divergent at a late-
stage and therefore allow derivatization of an advanced common
intermediate. Scheme 2 outlines the fruits of this effort. The
second-generation synthesis represents a notable improvement
over the racemic, low-yielding, nonselective, first-generation fit-
for-purpose route (Scheme 1). Starting from commercially
I
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 7. Strategy to increase the rigidity of the anacetrapib pharmacophore aimed at increasing the prevalence of active conformation, reducing the
predicted human dose, and limiting PXR off-target activity.
The third-generation approach that was designed to access
intermediate 58 with high selectivity starts with a racemic
addition of vinyl Grignard to the same commercially available
aldehyde 53, and subsequently to benefit from a catalytic
double-stereodifferentiation kinetic resolution promoted by
Hartwig59 or Helmchen60 iridium catalysts using chiral 60,
yielding a highly functionalized single enantiomer 61 in nearly
quantitative yield. A Schrock-61 or Grubbs-type62 (Zhan 1B63)
catalyst-mediated ring-closing metathesis (RCM) of 61,
followed by hydrogenation of olefin 62 with Wilkinson64
catalyst yielded 58 in just four linear steps, a 40% overall yield,
and with full control of enantio- and stereo-selectivity in all steps.
Versatile bromide 58 can be readily converted to final candidates
under Suzuki conditions with the desired boronic-acid−
coupling partner. The third-generation synthetic approach is
an outstanding example of synthetic chemistry innovation on
the critical path, ultimately enabling drug discovery within a
synthetically challenging but pharmacologically preferred
scaffold. It was also the basis for the first larger-scale delivery
(>10 g) of several key analogues, including 38.
library for modular SAR exploration. Initially, for simplicity of
synthesis and with the knowledge of the modular additivity of
the SAR established for the monocyclic series, we generated the
basic SAR scan in the context of anacetrapib analogues. We
eventually envisioned converting the leading set of molecules
into their respective ring RA carboxylic acids, analogous to the
anacetrapib and 15 pair in the monocyclic series. We found a
reasonable correlation of the CETP activity between the bicyclic
and monocyclic series (Table 7). The bicyclic series was less
sensitive than the monocyclic analogues to changes of the
bottom, highly lipophilic, aromatic ring RC substitutions, as
evidenced by comparing 38 to 63−67. The bicyclic series also
exhibited sensitivity to substitution of the left-hand-side ring RB
with another substituent (68), or with a heterocycle, as 70−72
afforded diminished potency compared to trifluoromethylben-
zene analogue 38. The only exception is the equipotent pyridine
69, which featured the dimethylamino substituent as a potential
metabolic liability. In summary, the initial scan suggested that
the left-hand-side trifluoromethyl benzene pharmacophore may
be preferred but is not unique, while the bottom aromatic ring
SAR might be more flexible compared to the monocyclic series
(Table 7).
Equipped with the highly efficient third-generation synthesis,
we were able to rapidly generate the SAR of the bicyclic
pharmacophore by simply substituting the starting material
aldehyde 53 for the desired substitution pattern of ring RB
(Figure 1) within the third-generation synthetic route (Scheme
2c). This approach enabled us to generate numerous advanced
cyclic oxazolidinone analogues of 58 and to use an aldehyde
We next turned our attention to the terminal, right-hand-side
carboxylic acid moiety of the molecule. We reasoned that with
higher intrinsic CETP inhibition and a better off-target profile of
the bicyclic series, modification of the carboxylic acid portion of
the molecule is a suitable strategy for reducing the overall
J
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Scheme 1. First Fit-for-Purpose Synthesis of Bicyclic Design C (Figure 7) Was Utilized to Prepare Each of the Eight Possible
a
Isomers, Including the Single Active CETP Inhibitor Isomer 38
a
Reagents, conditions, isolated yield: (a) 1-bromo-3,5-bis(trifluoromethyl)benzene, nBuLi, tetrahydrofuran (THF), −78 °C, 5 min; (b) TBSOTf,
43% (2 steps); (c) NaIO4, RuCl3·H2O EtOAc/water, 22 °C, 15 h, 67%; (d) 2-bromo-1-chloro-4-(trifluoromethyl)benzene, iPrMgCl THF at −78
to 22 °C, 15 h, 71%; (e) trifluoroacetic acid (TFA), dichloromethane (DCM), 22 °C, 30 min; (f) NaCNBH4, methanol, 0 °C, 2 h; (g) tetra-n-
butylammonium fluoride (TBAF), 22 °C, 2 h, 39% of mixture of isomers (3 steps); (h) phosgene, toluene, 22 °C, 30 min, 84% (mixture of
isomers); (i) (4-fluoro-5-isopropyl-2-methoxyphenyl)boronic acid, PdOAc2, DtBPF, K3PO4, THF and water, 70 °C, 15 h, 61% (mixture of isomers
separated by chiral chromatography).
Figure 8. SAR of the central rigid bicyclic core; 46−52 are selected derivatives of 38.
lipophilicity. We established that calculated log D (clog D)
correlated well with experimentally measured log D, which in
turn was a good predictor of experimentally measured solubility,
pharmacokinetic profiles, food effect and the propensity to be
retained in adipose tissue. Proposed permutations of key
pharmacophores were assessed computationally for a favorable
predicted clog D. Of these, approximately 100 derivatives were
synthesized for in vitro and in vivo profiling. Selected examples
of the active CETP inhibitors synthesized in this campaign with
key on- and off-target profiles in vitro and in vivo are
summarized in Tables 8 and 9.
Compounds 73−91 represent a general design in which the
established privileged bis-trifluoromethyl phenyl bicyclic
oxazolidinone is substituted with several preferred aryl and
K
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Scheme 2. (b) Second- and (c) Third-Generation Synthetic Strategies for Enabling a Detailed SAR Investigation of the Bicyclic
a
Core Series
a
Reagents, conditions, isolated yield: (a) (R)-(+)-2-methyl-2-propanesulfinamide, Ti(OEt)4, 40 °C, 1 h, 84%; (b) prop-3-en-1-ylmagnesium
bromide, DCM, THF, 22 °C, 1 min, 67%; (c) HCl in dioxane, MeOH, 22 °C, 30 min; (d) N,N-diisopropylethylamine (DIEA), DCM, CBZ-Cl, 0
°C, 30 min, 72% (2 steps); (e) 1,3-bis(trifluoromethyl)-5-vinylbenzene, Zhan catalyst (RC301), DCM, 60 °C, 55 min, 90%; (f) sodium tetraborate
decahydrate, aqueous ethylenediaminetetraacetic acid disodium salt dihydrate solution, Bu4NHSO4 D-Epoxone, CH3CN and EtOAc, Oxone, 0 °C,
1.5 h, 30% conversion, 67% yield based on recovered starting material; (g) DBU, N,N-dimethylformamide (DMF), microwave 200 °C, 3 min, 68%
(mixture of isomers separated by chiral chromatography); (h) vinyl magnesium bromide, THF, 0 °C, 30 min, 67%; (i) (11bS)-N,N-bis((R)-1-
phenylethyl)dinaphtho[2,1-d:1′,2′-f ][1,3,2]dioxaphosphepin-4-amine, [Ir(COD)Cl]2, DBU, THF, 33 °C, 48 h, 48% based on racemic 59 in
kinetic resolution = equivalent of 96% based on 60 (2 steps); (j) 1,3-bis(2,4,6-trimethylphenyl)-4,5-dihydroimidazol-2-ylidene[2-(i-propoxy)-5-
(N,N-dimethylaminosulfonyl)phenyl]methyleneruthenium(II) dichloride, toluene, 60 °C, 2 h, 85%; (k) RhCl(PPh3)3, 40 psi H2, ethanol, 22 °C, 15
h, >99%.
heteroaryl rings (RB, Table 8). These rings are in turn ortho-
substituted by a series of bi-heteroaryl and bi-aryl carboxylic
acids RA. The observed log D (measured by HPLC at pH 7.3),
an important in vitro predictor of desired in vivo properties, is
generally within the predicted range of 5−6, a notable
improvement over log D > 7 for anacetrapib. Log D variability
within the range of molecules 73−91 appears to best correlate
with the overall polarity and pKa properties of the heterocycle.
With the notable exceptions of compounds 80−81, 83, and 86,
derivatives of this design generally exhibited a very clean off-
target profile with no activity against a range of ion channels,
Cytochrome P450 enzymes (CYPs; neither acutely nor in a
time-dependent manner) and PXR (Table 8, “Key off-target
findings” column; unless stated otherwise, no significant off-
target findings in vitro). One notable exception for the entire
structural class was the relatively strong inhibition of the OATP-
family of transporters in vitro (in the absence of plasma), which
correlated to the presence of the carboxylic acid.
shown in Table 8 further reduced the measured log D compared
to their aromatic equivalents, while a high level of CETP
inhibition was still observed. Unfortunately, with added
flexibility of the carboxylic acid side chain, several off-target
findings were noted, including CYP inhibition and ion channel
interference. Additionally, and somewhat surprisingly, three of
the four derivatives displayed poor aqueous solubility at low pH,
reduced further from the observed solubility of the closest
aromatic comparator 73.
The guiding principle for the design of analogues 96−101 was
the desire to de-risk the liability of carboxylic acid for the OATP
inhibition. While nonacidic derivatives 96−101 generally did
not inhibit OATP in vitro, they suffered from a range of
undesirable off-target findings, ranging from CYP inhibition to
strong ion channel liabilities in vitro. Most importantly, the
physical−chemical properties of these molecules deteriorated
significantly, consistent with a significant increase in log D.
These changes ultimately resulted in an overall dramatic loss in
aqueous solubility at both low and high pH. It was reasonable to
assume that in addition to the solubility, 96−101 would also
Compounds 92−95 exemplify an effort to further lower the
log D by reducing the overall polarity and molecular weight. The
hypothesis proved to be correct as each of the four examples
L
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
c
Table 7. Initial SAR Scan of the Bicyclic Pharmacophore of the Leading Molecule 38
a
b
Unless otherwise noted, IC50 values are from >2 assays with standard deviation ≤40% of the reported mean. 20% not significant increase of
c
circulating aldosterone vs vehicle (100%) measured in C57/B6 mice 1 h after a 30 mg/kg dose. Single assay determination.
likely carry a liability for food effect and tissue retention in vivo
and would therefore be unsuitable for further development.
Finally, the most promising permutations of the aryl/
heteroaryl RB (Table 8) ortho-substituted bi-heteroaryl/bi-aryl
carboxylic acids RA were combined with several modifications of
the bottom bis-trifluoromethylphenyl ring in the attempt to
lower the overall lipophilicity, i.e., the variation of RE and RF.
Deleting one or both trifluoromethyl substituents led to a
significant loss of functional activity (107). However, compared
to their bistrifluoromethyl analogues 73, 87, and 91,
respectively, substitution of a single trifluoromethyl with
fluorine or methyl yielded strong CETP inhibition in vitro,
exemplified by 102−106, with the additional benefit of reducing
log D. Despite the improvement in physical−chemical proper-
ties, methyl derivatives suffered from decreased stability in liver
microsome homogenate in vitro as well as decreased aqueous
solubility relatively to their trifluoromethyl analogues. Several of
these analogues also registered a moderate increase in Nav1.5
off-target activity.
acceptable aldosterone margin (Table 8, ALDO column) of at
least 10x above the full CETP inhibition in vivo. While 73 and
104 exhibited improved an aldosterone margin of 10× and 85×,
respectively, a statistically significant increase in blood pressure
(Δ10 mm of mercury) was detected in rats at 15 and 50 mg/kg,
respectively, both of which correlated to an approximate 25×
margin over the projected plasma exposure needed for a full
CETP efficacy in vivo (Table 9). The blood pressure signal for a
chronic indication, disqualified both 73 and 104 from
consideration for further development, especially given the
history of blood pressure signals observed with torcetrapib,
which was withdrawn from clinical development.
Compound 87 represents the single molecule in Table 8
meeting the very stringent criteria for a drug candidate intended
for chronic dosing: high CETP inhibition in vitro and in vivo at
low dose, very clean off-target profile in vitro65 with no
statistically significant blood pressure signal in vivo (Table 9),
relatively high 24× aldosterone margin, and good stability
profile with no significant cross-species variability in vitro.
Compound 87 was selected for a full preclinical safety
assessment and larger quantities of this molecule were required.
An optimized synthesis specifically tailored to preparation of 87
was developed. Scheme 3 shows the approach that allowed
synthesis of 87 in 35% overall yield and seven linear steps from
inexpensive commercially available materials. Convergent
The aggregate of potent CETP inhibition in vitro, lack of off-
target activity and favorable log D guided selection of key
derivatives for in vivo evaluation. These compounds were
profiled for aldosterone margin determination, rodent pharma-
cokinetics and telemetrized rat blood pressure evaluation.
Compounds 74, 75, 90, 91, and 99 all failed to provide an
M
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
h
Table 8. Detailed SAR Investigation of the Bicyclic Pharmacophore
N
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 8. continued
O
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 8. continued
a
b
In vitrounless otherwise noted, IC50 values are from >2 assays with standard deviation ≤40% of the reported mean. In vivoMED, minimum
c
efficacious dose, to maximum CETP inhibition in vivo in NFR-TG-mouse at a dose (mg/kg). Key off-target findings: unless otherwise noted, the
following are all in vitro off-target IC50 > 30 μM: Cav1.2, Nav1.5, MK499 = Ikr, CYP3A4, 2D6, 2C9, 2C8, PXR, OATP; no TDI signal.
d
e
Telemetrized rat blood pressure, IV dosingsee Table 9 for details. LM, liver microsome, stability of parent molecule in LM homogenate in
f
g
vitro. No statistically significant increase in telemetrized rat blood pressure vs vehicle after 50 mg/kg IV dose; see Table 9. Single isomer, absolute
stereochemistry not determined. All four isomers isolated, range of on- and off-target activity for all four shown in the entry, log D determined for
h
the most active isomer.
Table 9. Rodent In Vivo Profile of Key Compounds: Mouse Pharmacokinetics and Telemetrized Rat Blood Pressure
a
b
mouse pharmacokinetics
rat blood pressure
Cpd #
AUCN
MRT
(h)
Cl
Vd
F
dose
(mg/kg)
15
compound vs vehicle
(μM/h)
(mL/min/kg)
(L/kg)
(%)
(Δmm Hg)
73
75
87
91
3.4
0.7
9.2
5.8
2.8
6.0
6.6
33
2.6
2.3
5.6
0.95
60
140
43
10
50
50
not statistically significant
10
a
b
An average of two experiments per arm, dosed at 1 mg/kg IV and 2 mg/kg by mouth, respectively. Three experiments, dosed at 15 or 50 mg/kg
IV, statistically significant blood pressure increase vs vehicle dosed to the same subject over identical time period prior to the administration of the
compound.
Suzuki coupling of 114 and 119 was retained; however, the
synthesis of chiral fragment 114 was revamped to eliminate the
need for chiral separation of building blocks and to reduce cycle
times. An approach using readily available pyroglutamic acid
derivative 108 was developed in which an anti-organometallic
addition to a hemi-aminal 109 proceeded with very high
diastereoselectivity setting the two chiral centers across the
pyrrolidine ring, yielding 110. The final chiral center was also set
via substrate control with a DIBAL reduction of the aryl ketone
111. Deprotection and carbamate formation provided com-
pound 114, which was then ready for end-game coupling.
Conditions for the final coupling were optimized for the Cl-
coupling partner 114, which was necessary, given that the new
approach made it impossible to carry the bromide (58, Scheme
2) through the sequence. This was readily accomplished using a
third-generation XPhos precatalyst.66
The mouse aldosterone margin has served as an effective tool
for prioritizing molecules across diverse chemical series because
it is independent of the absolute value of the on- and off-target
activities. The experiment directly correlates two plasma
concentrations for the same molecule; the minimum concen-
tration needed for full CETP inhibition (MED) vs the maximum
concentration lacking significant aldosterone elevation
(NOEL). However, the translation of these preclinical findings
to the clinical profile of CETP inhibitors (Table 1) was still
hypothetical. Torcetrapib represented a positive control for the
P
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Scheme 3. Optimized Synthetic Strategy for the Scale-Up of Safety Assessment Candidate 87
a
Reagents, conditions, isolated yield. (a) LiBHEt3, pTSA, PhMe-THF, −35 °C, 90%; (b) iPrMgBr, CuBr-Me2S, THF −25 °C, 70%; (c) BuLi,
MTBE, −90 °C 72%; (d) DiBAL-H, THF, −5 °C, 93%; (e) HCl, EtOAc, 20 °C, 84%; (f) triphosgene, DCM, 30 °C, 90%; (g) Boc2O, 4-
dimethylaminopyridine (DMAP), t-BuOH/MTBE, 40 °C, 92%; (h) Pd(dppf)Cl2, B2Pin2, dioxane, 120 °C, 80%; (i) Pd(dppf)Cl2, PhMe/water, 85
°C; (j) XPhos-Palladium Generation 3 precatalyst, B2Pin2, dioxane, 85 °C, 90%; (k) XPhos-Palladium Generation 3 precatalyst, K3PO4, THF/
water, 60 °C, 92%; (l) TFA, toluene, 55 °C, 97%.
Figure 9. Side-by-side comparison of 87 to evacetrapib: a correlation of preclinical margin to clinical results for evacetrapib, which serves as a negative
clinical control of lesser preclinical margin, cf. 87.
correlation, as it exhibited no preclinical aldosterone margin and
clinical aldosterone-related AEs. Anacetrapib could not be used
as a robust negative control for the correlation. The compound
exhibited >8× preclinical aldosterone margin with no clinical
AEs, but anacetrapib was exposure limited at higher potential
margins due to suboptimal physical−chemical properties and
solubility. Fortunately, when 87 was selected as a preclinical
safety assessment candidate, evacetrapib had already completed
Q
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 10. Chronic efficacy in transgenic CETP mice: CETPi and HDL cholesterol efficacy in vivo. Side-by-side comparison of 87 at 10 mg/kg to
evacetrapib and anacetrapib at 100 mg/kg; *** P < 0.001 vs vehicle (one-way analysis of variance (ANOVA), Dunnett’s multiple comparison test);
cholesteryl ester (CE), counts per minute (CPM).
Table 10. Preclinical Pharmacokinetics of 87 in CD1 Mice, WH Rats, Beagle Dogs, and Rhesus Monkeys Following a Single Oral
or IV Dose
mouse
rat
dog
monkey
0.5
IV dose (mg/kg)
Clp (mL/min/kg)
Vdss (L/kg)
1
1
0.5
2.6 0.58
0.95 0.38
4.4 1.1
6.0 1.0
2
3.1 0.27
2.4 0.32
8.1 1.5
13 2.5
2
0.52 0.17
0.54 0.08
15 2.9
18 3.5
2
0.39 0.18
1.3 0.2
50 8.8
60 16
2
t
t
1/2 terminal (h)
1/2 MRT (h)
P.O. (mg/kg)
C
max (μM)
0.43 0.31
2.7 1.2
7.9 1.7
43
0.4 0.04
4.0 0.0
9.7 2.6
65
0.69 0.52
1.7 0.6
13 12
14 13
0.22 0.09
6.7 1.2
11 2.3
8.4 1.8
Tmax (h)
AUC0−∞ (μM·h)
F (%)
the Phase 2 clinical trials with no significant compound-related
effects on aldosterone release or blood pressure and therefore
could serve a viable negative control for the ALDO-margin-to-
clinic correlation. Figure 9 makes a side-by-side comparison of
evacetrapib to 87. Evacetrapib exhibits approximately 5×
preclinical aldosterone margin while 87 shows a 24× margin
over the maximum CETP efficacy in vivo. We postulated that a
24× preclinical aldosterone margin should be sufficient for
advancing 87 into clinical trials with reduced risk of AEs at high
exposure.
To fully profile the comparative efficacy of 87, the compound
was examined in a 2-week chronic treatment of CETP transgenic
mice alongside anacetrapib and evacetrapib, and the results are
provided in Figure 10. The compounds were administered QD
at the minimal dose needed to provide maximal CETP
inhibitions, and this was demonstrated in Figure 10a. Two
hours following the final dose, maximal CETP inhibition
resulted in a significant HDL cholesterol increase in vivo for
87, which was comparable to both positive controls, anacetrapib
and evacetrapib. It is noteworthy that 87 at 10 mg/kg reaches
the same level of maximum efficacy as both evacetrapib and
anacetrapib at 100 mg/kg dose.
Compound 87 displayed low to moderate bioavailability (8.4−
65% across species).
The elimination half-life of 87 from plasma, liver, and adipose
tissue was investigated in several mouse strains to thoroughly
differentiate the compound from anacetrapib.67 In one study, 87
was administered orally to NFR-TG mice in a single dose at 10
mg/kg in 0.5% methocellulose, and serial samples were collected
over a period of 72 h. The plasma half-life of 87 in this study was
12 h (Table 11). Next, the compounds were administered orally
Table 11. Half-Lives of 87 from Plasma, Liver, and Adipose
Tissue in Lean and DIO Wild-Type and Lean NFR-TG Mice
Following Single Dose and 40-Day QD Dose (10 mg/kg)
multiple dose of 87
single dose of 87
lean NFR-
TG
NFR-TG
lean wt
DIO wt
plasma
liver
adipose
12 0.5 h
N/A
N/A
0.8 0.1 d
1.5 1.6 d
0.8 0.2 d
2.8 0.9 d
3.2 1.0 d
3.6 2.0 d
12 5.5 d
2.9 0.9 d
1.5 1.0 d
for 14 days at 10 mg/kg (in 0.5% methocellulose) to examine the
potential for accumulation. Terminal samples were collected on
days 1, 7, 14, 28, and 40, following the last dose. It was found that
plasma half-lives in lean, diet-induced obesity (DIO) and NFR-
TG mice were progressively longer (MD t1/2 0.8, 2.8, and 12
days, respectively), indicating the increased influence of fat
content and CETP-binding to plasma elimination kinetics. A
similar trend was also observed for liver and adipose tissue
(Table 11). Importantly, in lean NFR-TG mice, the plasma T1/2
was approximately 24× longer following the multi-day dosing
than was the single dose. Together, these two studies nicely
The full pharmacokinetic profile of 87 was assessed in several
preclinical species, and the results are presented in Table 10.
Following a single IV dose to wild-type (WT) CD1 mouse,
Wistar-Han (WH) rats, Beagle dogs, and Rhesus monkeys,
plasma clearance and volume of distribution of 87 were low (Clp,
0.39−3.1 mL/min/kg; Vdss, 0.54−2.4 L/Kg). Consistent with
low plasma clearance values, terminal half-lives in all species
were long (ranging from 4.4 h in mouse to 50 h in monkeys).
R
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 11. (a) Time course of parent molecule elimination from white adipose tissue of DIO-NFR-TG-mice after the cessation of 2-week dosing
(Anacetrapib 100 mg/kg/day; Evacetrapib 30 mg/kg/day; 87 10 mg/kg/day; dosed orally once daily in 0.5% methylcellulose vehicle). (b) Calculated
half-life of parent molecule in plasma, liver, and adipose tissues.
demonstrated the effect of multiple dosing, adipose tissue
distribution, and CETP-binding on the half-life of 87. Notably,
in all strains, elimination rates of 87 from adipose tissue were
roughly equal or lesser than that of plasma, suggesting that 87
did not accumulate in the adipose compartment (Table 11).
Using the results from Table 11, we sought to compare the
adipose retention of 87 with the clinical CETP inhibitors
anacetrapib and evacetrapib following multiple-dose 14-day
treatment. For this study, NFR-TG mice were fed to induce the
diet-induced obese (DIO) state. Figure 11 compares the tissue
retention characteristics of the CETP inhibitors. From Figure
11a the time course of parent molecule elimination from white
adipose tissue of DIO-NFR-TG mice is shown after the
cessation of 2-week dosing, which illustrates the comparatively
lower tissue retention of 87 and evacetrapib as compared to
anacetrapib. Figure 11b shows the calculated half-life of the
parent molecule in plasma, liver, and adipose tissues from this
study. Notably, 87 is eliminated from all tissues, including the
adipose, at a rate comparable to evacetrapib and at a significantly
higher rate compared to anacetrapib. It is noteworthy that the
kinetics displayed in Figure 11 only account for the elimination
of the parent molecule and not any metabolites. Evacetrapib
undergoes extensive metabolism in vitro and in vivo, producing
additional entities that might extend drug-related exposure post-
parent elimination due to long-lived circulating and/or
covalently bound metabolites.68
(0.8 d for both plasma and tissue) and DIO (2.8 d from plasma,
3.6 d from tissue) strains. In the NFR-TG mouse, 87 shows
faster elimination from tissue (1.5 d) than from plasma (12 d),
confirming the lack of tissue accumulation in another animal
strain. Preclinical data suggested that 87 would have a shorter
human half-life relative to anacetrapib because of the lack of
tissue accumulation, which represented a clear advantage (Table
11).
The effect of food on the PK parameters of 87 was
investigated next in dogs. The compound was dosed at 1 mg/
kg in fed (high-fat diet) and fasted arms in 0.5% methocel and
Imwitor/Tween formulations. In these studies, no significant
difference in plasma AUC0‑24 h exposure was observed between
the fed and fasted states (Table 12).
Table 12. Food Effect on 87 Pharmacokinetics in Dogs
0.5% methocel
Imwitor/Tween
dosing vehicle fasted/
fed dogs
fasted
fed
fasted
24
fed
normalized AUC0−72 h
(μM·h·kg/mg)
6.5 0.4
4.2 1.7
8
22
7
C
max (μM)
0.5 0.1
14%
0.3 0.1
9%
1.5 0.8
53%
1.3 0.5
48%
estimated %Foral
food effect (ratio
AUC0−72 h) fed/
fasted)
0.6×
0.9×
The mouse served as the primary preclinical species model to
predict a human dose and investigate the PK of 87. The
advantages of this species include the availability of lean WT and
DIO strains that do not express CETP, as well as the NFR
transgenic CD1 strain that does express human CETP.
Following single and multiple oral dosing to WT mice, 87
showed plasma half-life values of 0.18 and 0.8 days, respectively.
The 4x greater half-life at steady state upon multiple dosing was
hypothesized to result from increased adipose tissue distribu-
tion. In comparison, single- and multiple-dose half-life values in
NFR-TG mice were 12 h and 12 days, respectively, suggesting
that binding to CETP significantly increased the compound half-
life as well. A third factor contributing to the elimination of 87 is
the relative fat content of the subject, as evidenced by the 4x
longer half-life in the DIO WT strain (2.8 d) as opposed to lean
WT mice (0.8 d). It is important to note that, although 87 was
distributed to adipose tissue, which extended its half-life
somewhat, it did not accumulate in the adipose compartment.
This is evidenced by the fact that the elimination rate of 87 from
adipose tissue was roughly parallel to that of plasma in both lean
Mindful of the risk stemming from the existence and potential
clinical relevance of the long-lived metabolites observed for
torcetrapib,69 elimination pathways of 87 in WH rats and rhesus
monkeys were investigated following a single oral dose of 3[H]
87 to bile-duct-cannulated (BDC) subjects. Table 13 summa-
rizes the findings for both species. Irrespective of route of
administration, the only species observed in bile, feces, urine,
and plasma were either parent 87, a glucuronide of 87, or in one
case a minor amount of taurine conjugate of 87. The major route
of excretion of 87 included elimination of the parent or its
glucuronide. In plasma, the parent was the major drug-related
material at 2-, 4-, 8-, and 24-h post-dose with trace amounts of
parent glucuronide. No long-lived or circulating metabolites
were observed for 87 in either species, which provided a level of
confidence that the risk of metabolite-related AEs is low.
The effects of 87 on measures of cardiac conduction and
repolarization were assessed both in vitro and in vivo
(anesthetized guinea pigs and conscious telemetered monkey
models). The observed ion channel IC50 values with 87 were
S
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 13. Elimination Pathways of 87 in WH Rats and Rhesus Monkeys Were Investigated Following a Single Oral Dose of [3H]
87 to Bile-Duct-Cannulated Subjects
RAT
relative % of total recovered
dosing
PO
% total radioactivity
87-parent
87-glucoronide
>98
87-taurine c.
0
recovered total bile
recovered total feces composition
recovered total urine
>99%
46%
0%
<2
>98
IV
recovered total bile
recovered total feces
plasma composition T = 2 h
plasma composition T = 24 h
67%
15%
>98
<4
>98
>96
>99
RHESUS MONKEY
% total radioactivity
relative % of total recovered
87-glucoronide
dosing
PO
87-parent
87-taurine c.
3
recovered total bile
recovered total feces composition
recovered total urine
plasma composition T = 2 h
plasma composition T = 24 h
>99%
60%
7%
<1
95
>99
>99
>99
>96
5
>30 μM for hERG, IKs, and ICa. The IC50 for inhibition of INa was
4.1 μM. These values in the context of the unbound plasma
concentration at Cmax of 87 (free fraction of 87 in plasma is
below 0.1%) suggest a low risk for effects on cardiac conduction
or proarrhythmia. Consistent with these data, there was a lack of
any test-article-related effects on PR, QRS, QT, or QTc intervals
with 87 up to 47 μM (∼118-fold above the predicted clinical
The compound had no effect on human CYP protein
concentration or Nrf2 activation in any of the 8 cell lines.
Overall, there was no evidence of 87 undergoing CYP-
dependent bioactivation in HEK293/human CYP cells.
Notably, in contrast to the findings with 87, evacetrapib
demonstrated an increased potential for bioactivation by human
CYP3A4 and dalcetrapib was subject to bioactivation mediated
by human CYP2B6, CYP2C9, and CYP2C19 as well as through
another CYP-independent pathway in the HEK293/human
CYP cell-based assays.
C
max) in an exploratory IV cardiovascular study in anesthetized
guinea pigs or in conscious telemetrized rhesus monkeys at oral
doses up to 25 mg/kg (Cmax = 29 μM) providing an exposure
multiple of ∼73-fold above the predicted clinical Cmax
.
For in vivo studies to assess tolerability and drug-induced
bioactivation in the liver, rats were treated with 87 at 750 mg/
kg/day for 7 days in an exploratory dose-limiting toxicity study,
after which RNA samples from their livers were extracted,
purified, and assessed for the induction of an in vivo rat
bioactivation signature indicative of the generation of drug-
related reactive metabolites.71 Compound 87 did not elicit any
prominent bioactivation signal in rat liver; these transcriptional
analyses indicated no concerning risk for bioactivation and
demonstrated a profile similar to compounds in the clinic that
have no known DILI. In contrast to the findings with 87,
evacetrapib and dalcetrapib were positive for bioactivation based
on analogous studies and their transcriptional profiles, which
supported the conclusion of a notable increase in generation of
reactive metabolites in the rat similar to compounds tested in
rats that are known to have a DILI in clinic.35 Although 750 mg/
kg/day was not tolerated in a subsequent 2-week rat study, 87
was shown to be well tolerated up to 50 mg/kg/day (∼50-fold
over anticipated human clinical exposures) when administered
to rats for 3 months, demonstrating only very slight to slight
changes in a few monitorable biochemical and hematologic
parameters and no adverse histomorphologic changes. Addi-
tionally, doses up to 20 mg/kg/day (∼42-fold the projected
clinically efficacious exposure) for up to 3 months in NHP
showed no adverse antemortem or postmortem findings with
changes limited to expected pharmacologic effects on lipid
parameters, as demonstrated by an approximate 50% increase in
high-density lipoproteins and an approximate 50% decrease in
low-density lipoproteins at the 5 mg/kg/day dose after 2 weeks
of treatment. Further, the potential developmental toxicity of 87
Additionally, there were no test-article-related hemodynamic
changes in non-human primate (NHP) at these exposures. For
all safety species, the safety margin for 87 was either equivalent
to or exceeded those of the remaining comparator clinical-stage
CETP inhibitors, none of which were causing any cardiovascular
AEs in Phase 1−3 clinical trials.70 Compound 87 was also
assessed for potency to a full panel of over 110 known
pharmacology targets in a Eurofins Panlabs panel, and showed
no activity exceeding 50% at 10 μM.
Compound 87 was assessed for its potential to undergo
“bioactivation” to potentially reactive intermediates as an
approach to inform the potential for any increase in the risk of
drug-induced clinical liver injury (DILI). For the in vitro study,
HEK293/CYP cell lines expressing human CYP1A2, CYP2E1,
CYP2D6, CYP2C8, CYP2B6, CYP3A4, CYP2C19, or CYP2C9
were exposed to different concentrations of 87 up to a maximal
concentration of 25 μM limited by effects on cell viability.
Compound 87−induced cytotoxicity was assessed +/− the
presence of a particular human CYP enzyme and +/− the
presence of buthionine sulfoximine, a depleter of glutathione.
Cell lysates were also probed via Western blot for the level of
human CYP remaining at the end of the incubation. A reduction
in the amount of human CYP remaining is believed to reflect the
amount of covalent modification of the CYP and its subsequent
elimination in the proteasome. Cell lysates were also evaluated
by Western blot for protein concentration of Nrf2, a biomarker
of oxidative stress. Compound 87 reduced cell viability in a dose-
dependent manner beginning at ∼10 μM in all 8 cell lines
regardless of human CYP expression or glutathione depletion.
T
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 14. Comparison of MK-8262 to Other CETP Inhibitors: A List of Selected Properties That Represent a Differentiation
among the Comparators
MK-8262
anacetrapib
7.1
dalcetrapib
evacetrapib
6.7
log D
5.3
6.5
PXR (μM)
TDI index
clinical dose
dose number
>30 μM
1.0
QD: 1−3 mg (p)
QW: 50 mg (p)
9.1 μM
N/A
100 mg
9.4 μM
N/A
600 mg
4.5 μM
4.6
130 mg
a
b
b
<1 (QD)
133 (QD)
240 (QD)
43 (QD, Ph2)
<5 (QW)
no accumulation
0.6−1.3× dog
1× human
24×
tissue accumulation (NFR-TG-ms)
food effect
accumulation
8× dog
6× human
>8×
N/A
N/A
no accumulation
1.4× dog
aldosterone margin
DILI35
N/A
positive
5×
positive
negative
N/A
a
b
c
TDI, time-dependent inhibition of CYPs. Projected (p) from a Ph1 SAD study. For the full table, see the Supporting Information.
Figure 12. MK-8262 SAD Ph1 clinical results, pharmacokinetics: MK-8262 displayed a biphasic PK profile, with a half-life of approximately 5 days, and
no evidence of tissue accumulation.
was tested in a rat whole embryo culture (WEC) assay using
gestation day 9 embryos.72 The embryos were exposed to
different concentrations of 87 ranging from 3 to 173 μM for 48
h, representing a sensitive window of organogenesis. No effects
were observed in this assay with 87 up to 22 μM, and the WEC-
predictive model ranked developmental toxicity potential lower
for 87 than for evacetrapib and dalcetrapib (dalcetrapib >
evacetrapib > 87).
After a thorough absorption, distribution, metabolism, and
excretion (ADME) process, stability, in vitro counter screening,
in vivo pharmacokinetics and pharmacodynamics, and pre-
clinical safety assessment, 87 was deemed suitable for clinical
development and was designated as MK-8262.
Single-compartment allometric analysis of Clp values from
WH rats, beagle dogs, and rhesus monkeys, derived from single-
dose IV studies, provided a projected human Clp of 0.2 mL/
min/kg. Low turnover of MK-8262 in vitro precluded the use of
in vitro in vivo correlation (IVIVC) methodology for this
assessment. Detailed preclinical work indicated the influence of
dosing frequency, fat content and CETP-binding in determining
the half-life of the compound. Preclinical models are not
available to quantitatively measure the impact of these variables
to accurately estimate a human plasma half-life. To predict a
clinical dose, the half-life was approximated to be 62 h from
allometrically scaled Clb of 0.2 mL/min/kg and mean Vdss of 1.4
L/kg. Assuming an effective steady-state trough concentration
of 0.3 μM, the clinical dose was projected to be 70 mg (with a
conservative bioavailability value of 10%) with steady-state Cmax
and AUC0−24 h values of 0.4 μM and 8.0 μM·h, respectively.
However, accounting for better bioavailability, the human dose
of MK-8262 to achieve maximum CETP inhibition for first-in-
human studies was estimated to be 1−3 mg of QD, or 50 mg of
QW. The improved physical−chemical properties of MK-8262
allowed its dosing as an amorphous suspension in Phase 1
clinical studies rather than the hot-melt extrusion formulation
needed for anacetrapib.
Table 14 shows a side-by-side comparison of key properties of
MK-8262 with those of other CETP inhibitors being tested in
clinical trials while MK-8262 was being developed (see the
Supporting Information for a more comprehensive list).
Acknowledging the different stages of clinical development for
MK-8262 (Ph1) vs anacetrapib, evacetrapib, and dalcetrapib (all
progressed into Ph3), MK-8262 nevertheless stands out for its
potential to be a best-in-class molecule, from several
perspectives. The overall physical−chemical properties of MK-
8262 are superior to the rest of the class, basically represented by
the log D value of the molecule (log D = 5.3). Generally, a
certain level of lipophilicity may be required for strong in vivo
inhibition of a protein like CETP, which interacts with plasma
lipids; excessive lipophilicity can result in additional liabilities,
including prolonged tissue retention. The significantly lower
log D of MK-8262 as compared to all other molecules is
consistent with the de-risked adipose retention outlined in Table
11 and with a range of additional superior properties in Table 14,
including the lack of food effect in the clinic and a good excretion
profile. The projected human dose required for maximum CETP
U
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 13. MK-8262 SAD Ph1 clinical results: percentage of CETP inhibition achieved following administration of a single oral dose of MK-8262 in
healthy young subjects (N = 6/panel); BQL (below quantification limit); limit of quantification (LOQ).
inhibitionand, therefore, full LDL PD effectis approx-
imately 100× lower than the doses required for all other
compounds (the projected efficacious human dose of MK-8262
is 1−3 mg of QD). A low projected dose in combination with
inherit superior physical−chemical properties renders the very
low dose number, a critical indicator of developability of a once-
a-day or once-weekly monotherapy. Additionally, such a
favorable dose number virtually assures the option to combine
MK-8262 with statins, the standard of care on top of which
CETP inhibitors are to be administered, either co-administered
or as a fix-dose-combination single pill with a statin. MK-8262 is
clearly differentiated in the absence of a DILI35 signal for liver
bioactivation as well as exceptional aldosterone margin vs
comparators. In summary, the aforementioned data positioned
MK-8262 for clinical development.
A randomized, placebo-controlled, double-blind phase I
clinical trial was performed in normal healthy volunteers to
determine the safety, tolerability, pharmacokinetics, and
pharmacodynamics (inhibition of CETP activity) of single
oral doses of MK-8262 (1, 3, 6, 12, 24, 45, 90 mg) in human
subjects (Figure 12). MK-8262 was generally well tolerated at
single doses up to 90 mg. MK-8262 displayed a biphasic PK
profile, with a half-life of approximately 120 h, affording the
potential for QW dosing. No evidence of a significant food effect
was seen. In a comparable study, anacetrapib displayed a single-
dose terminal half-life between 9−62 h in normal healthy
volunteers, with a positive food effect that increased exposure 2-
to 3-fold with a low fat meal, and 6- to 8-fold with a high-fat meal,
compared to fasted subjects.73 Importantly, the single-dose PK
of anacetrapib underestimated the extremely prolonged plasma
terminal half-life observed with chronic administration, which
was attributed to distribution into and slow egress of anacetrapib
from adipose tissue,44 an effect which was ruled out for MK-
8262 preclinically. In this single-dose Ph1 study in humans,
doses of MK-8262 above 12 mg were associated with maximum
inhibition of baseline CETP activity (Figure 13). Taken together
with the single-dose PK results, the effect of MK-8262 on CETP
activity suggests a once-daily dose of ≤1 mg. Due to the
progression of the lead molecule anacetrapib through its Phase 3
cardiovascular outcomes trial, a decision was made to suspend
further development of MK-8262. Importantly, the decision was
due to strategic reasons rather than any nonclinical or clinical
safety findings, or any aspect of performance of the molecule.
CONCLUSIONS
■
After decades of CETP-mechanism-related drug discovery
efforts, recent clinical outcomes have demonstrated that a
combination of CETP inhibition and statins has the potential to
reduce CHD risk by as much as additional 9% compared to
statins alone.16,17 Here, we reported the invention of 87, a
unique CETP inhibitor with the potential for being the best-in-
class molecule. We have presented the method by which we
optimized for the desired on-target profile and de-risked the
potential for AEs that had been clinically demonstrated for
several other CETP inhibitors. Our discovery and optimization
methods were based on standard safety de-risking paradigms as
well as novel approaches, balancing on-target efficacy with the
potential for elevations in plasma aldosterone and related
cardiovascular biomarkers. Based on its established preclinical
profile and its potential to be the best-in-class CETP inhibitor,
87 was designated as the development candidate MK-8262. The
single-ascending dose clinical studies of MK-8262 demonstrated
the validity of our approach, rendering the candidate as safe and
fully efficacious at a very low dose, projected at or below 1 mg of
QD, for a full CETP inhibition. The efficacy for CETP inhibition
invoked by MK-8262, along with its safety profile, exhibits a
V
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Steps iii, iv: (4S,5R)-5-[3,5-Bis(trifluoromethyl)phenyl]-3-{[2′-
chloro-5′-iodo-4-(trifluoromethyl)biphenyl-2-yl]methyl}-4-methyl-
l,3-oxazolidin-2-one (Intermediate 6). To a solution of (4S,5R)-5-
[3,5-bis(trifluoromethyl)phenyl]-3-{[2′-chloro-5,-nitro-4-
(trifluoromethyl)biphenyl-2-yl]methyl}-4-methyl-1,3-oxazolidin-2-
one (Intermediate 5, 1.14 g, 1.82 mmol) in EtOH (20 mL) at room
temperature was added SnCl2·H2O (5 equiv). The solution was stirred
at room temperature for 4 h. TLC (CH2Cl2:hexane/l:l) showed that the
reaction was completed. EtOAc (50 mL) was added. The mixture was
washed with water (2 × 20 mL), brine, and dried over sodium sulfate.
The solvent was removed under reduced pressure, and the residue was
diluted with chloroform (30 mL). n-Pentyl nitrite (0.36 mL, 2.73
mmol) and iodine (0.55 g, 2.18 mmol) were added. The mixture was
stirred under refluxing for 1 h. The mixture was diluted with methylene
chloride (30 mL), and the purple solution was washed with saturated
sodium thiosulfate solution, brine, and dried over sodium sulfate. Flash
column using EtOAc:hexane/2:98 as the eluent afforded (4S,5R)-5-
[3,5-bis(trifluoromethyl)phenyl]-3-{[2′-chloro-5′-iodo-4-
(trifluoromethyl)biphenyl-2-yl]methyl}-4-methyl-l,3-oxazolidin-2-one
(Intermediate 6, 650 mg, yield 50%). LCMS (ESI) calcd for
clear potential for reducing the overall cardiovascular risk due to
the modulation of cholesterol and lipoprotein biology.
EXPERIMENTAL SECTION
■
General. All procedures were carried out under a nitrogen
atmosphere using commercially available solvents that were used
without further purification. Starting materials were obtained from
commercial sources and were used without further purification.
Normal-phase column chromatography was carried out using
prepacked silica gel cartridges on a Combiflash Rf separation system
by Teledyne ISCO. Reversed-phase HPLC purification was carried out
using YMC C18 column on a Gilson HPLC system utilizing a gradient
of 0.05% TFA in water/acetonitrile as the eluent. NMR spectra were
recorded using Varian Inova 500 spectrometers. The reported
compounds are of ≥95% purity, unless otherwise noted. Purity analysis
was carried out by LC-Mass analysis using an Agilent 1100 system fitted
with Waters Micromass.
Modular synthesis of 2−37 was achieved by assembling various
building blocks for quick SAR on different parts of target molecules
colored by rings was applied. The synthesis is exemplified by the
preparation of 10.
1
C26H16ClF9INO2 [M + H]+: 707.0, found: 707.0. H NMR (CDCl3,
500 MHz): δ a mixture of 1:1 atopoisomers 7.90 (s, 1H), 7.77 (s, 1H),
7.70−7.75 (m, 4H), 7.65 (d, J = 2.5 Hz, VzH), 7.61(d, J = 2.5 Hz, 1/2
H), 7.40 (m, 1H), 7.28 (m, 1H), 5.66 (d, J = 8Hz, 1/2 H), 5.64 (d, J = 8
Hz, 1/2 H), 4.85 (d, J = 15.5 Hz, 1/2 H), 4.82 (d, J = 14 Hz, 1/2 H),
4.02 (d, J = 16 Hz, 1/2 H), 3.96 (m, 1/2 H), 3.95 (d, J = 15.5 Hz, 1/2
H), 3.79 (m, 1/2 H), 0.64 (d, J = 6.5 Hz, 1.5H), 0.57 (d, J = 6.5 Hz,
1.5H) ppm.
Step i: (4S, 5R)-5-[3,5-Bis(trifluoromethyljpheny1]-3-[2-iodo-5-
(trifluoromethyl)benzyl]-4-methyl-l,3-xazo-lidin-2-one (Intermedi-
ate 3). A solution of (4S,5R)-5-[3,5-bis(trifluoromethylphenyl]-4-
methyl-1,3-oxazolidin-2-one (Intermediate 1, 3.13 g, 10.0 mmol) in
dry THF (93 mL) was added dropwise via cannula to a stirred
suspension of sodium hydride (60% dispersion in mineral oil) (0.317
mL, 9.50 mmol) in dry THF (93 mL) at 0 °C under N2. The reaction
was stirred at 25 °C for 30 min. A solution of 2-(bromomethyl)-L-iodo-
4-(trifluoromethyl)benzene (Int 2, 3.65 g, 10.00 mmol) in dry THF
(46.5 mL) was added dropwise via cannula, and the reaction mixture
was stirred overnight at 25 °C. Water (100 mL) was added, and the
reaction mixture was extracted with EtOAc (3 × 100 mL). The
combined extracts were dried (MgSO4) and concentrated in vacuo to
give the crude product. The crude was purified by flash chromatography
(Biotage Horizon, 65i, Si, ∼85−90 mL/min, 100% hexanes for 360 mL,
gradient to 20% EtOAc in hexanes over 2400 mL, gradient to 50%
EtOAc in hexanes over 1224 mL) to afford desired product (4S,5R)-5-
[3,5-bis(trifluoromethyl)phenyl]-3-[2-iodo-5-(trifluoromethyl)-
benzyl]-4-methyl-l,3-oxazolidin-2-one (Int 3, 5.16 g, 8.64 mmol, 86%
yield), as a colorless solid. LCMS (ESI) calcd for C20H13F9INO2 [M +
H]+: 598.0, found: 598.0. 1H NMR (CDCl3, 500 MHz): δ 8.06 (d, J =
8.2 Hz, 1H), 7.93 (s, 1H), 7.82 (s, 2H), 7.61 (s, 1H), 7.33 (dd, J = 8.2,
1.4 Hz, 1H), 5.79 (d, J = 7.8 Hz, 1H), 4.91 (d, J = 16 Hz, 1H), 4.40 (d, J
= 16 Hz, 1H), 4.16−4.06 (m, 1H), 0.83 (d, J = 6.4 Hz, 3 H) ppm.
Step ii: (4S,5R)-5-[3,5-Bis(trifluoromethyl)phenyl]-3-{[2′-chloro-
5,-nitro-4-(trifluoromethyl) biphenyl-2-yl]methyl}-4-methyl-1,3-ox-
azolidin-2-one (Intermediate 5). A mixture of (4S,5R)-5-[3,5-
bis(trifluoromethyl)phenyl]-3-[2-iodo-5-(trifluoromethyl)-benzyl]-4-
methyl-l,3-oxazolidin-2-one (Intermediate 3, 1.5 g, 2.51 mmol), 2-
chloro-5-nitro phenyl boronic acid (Intermediate 4, 0.740 g, 3.67
mmol), tetrakis(triphenylphosphine) palladium (430 mg, 0.372 mol),
and sodium carbonate (0.532 g, 5.02 mmol) in 50 mL of water/EtOH/
toluene (1:2:4) was heated to reflux for 4 h. Thin-layer chromatography
(TLC) (CH2Cl2:hexane/l:l) showed that the reaction was complete.
The solvents were removed. Water (30 mL) was added. The organic
was extracted with methylene chloride (3 × 40 mL). The combined
methylene chloride layers were washed with brine and dried over
sodium sulfate. Flash column using CH2Cl2:hexane/6:4 as the eluent
afforded (4S,5R)-5-[3,5-bis(trifluoromethyl)phenyl]-3-{[2′-chloro-5,-
nitro-4-(trifluoromethyl)biphenyl-2-yl]methyl}-4-methyl-1,3-oxazoli-
din-2-one (Intermediate 5, 0.96 g, yield 62.2%) as a colorless solid.
LCMS (ESI) calcd for C26H16ClF9N2O4 [M + H]+: 627.1, found:
627.0. 1H NMR (CDCl3, 500 MHz): δ 1:1 mixture of roamers 8.16−
8.31 (m, 1H), 8.21 (d, J = 2.5 Hz, l/zH), 8.16(d, J = 2.5 Hz, VzH), 7.90
(s, 1H), 7.71−7.78 (m, 5H), 7.42−7.46 (m, 1H), 5.66 (d, J = 4.5 Hz, %
H), 5.64 (d, J = 4.5 Hz, VzH), 4.93 (d, J=15.5 Hz, VzH), 4.79 (d, J = 16
Hz, VzH), 4.03 (d, J = 16 Hz, VzH), 3.94 (m, 1H), 3.91 (d, J = 15.5 Hz,
VzH), 0.70 (d, J = 6.5 Hz, 1.5H), 0.64 (d, J = 6.5 Hz, 1.5H) ppm.
Step v. Methyl 2″-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-
methyl-2-oxooxazolidin-3-yl)methyl)-4′-chloro-2-methyl-4″-(tri-
fluoromethyl)-[1,1′:3′,1″-terphenyl]-4-carboxylate. A mixture of
(4S,5R)-5-[3,5-bis(trifluoromethyl)phenyl]-3-{[2′-chloro-5′-iodo-4-
(trifluoromethyl)biphenyl-2-yl]methyl}-4-methyl-l,3-oxazolidin-2-one
(Intermediate 6, 0.490 g, 0.692 mmol), methyl 3-methyl-4-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)benzoate (Intermediate 7, 0.382
g, 1.385 mmol), tetrakis(triphenylphosphine) palladium (40 mg, 5%
mol) and sodium carbonate (0.220 g, 2.077 mmol) in 14 mL of water/
EtOH/toluene (1:2:4) was heated to reflux for 2 h. TLC
(CH2Cl2:hexane/8:2) showed that the reaction was over. The solvents
were removed. Water (10 mL) was added. The organic was extracted
with methylene chloride (3×20 mL). The combined methylene
chloride layers were washed with brine and dried over sodium sulfate.
methyl 2″-(((4S,5R)-5-(3,5-bis(trifluoromethyl)phenyl)-4-methyl-2-
oxooxazolidin-3-yl)methyl)-4′-chloro-2-methyl-4″-(trifluoromethyl)-
[1,1′:3′,1″-terphenyl]-4-carboxylate (0.386 mg, yield 76%) was
obtained after purification with preparative TLC using 18% EtOAc in
hexane as the eluent. MS ESI: calcd for C35H25ClF9NO4 [M + H]+
730.13, found 730.13. 1H NMR (CDCl3, 500 MHz): δ 1:1 mixture of
atropisomers: 8.03 (2H, s), 7.97 (1H, s), 7.90 (1H, d, J = 7.5 Hz), 7.87
(1H, s), 7.69−7.73 (2H, m), 7.48 (1/2 H, d, J = 7.9 Hz), 7.45 (1/2 H, d,
J = 8.0 Hz), 7.37 (1H, m), 7.35 (1H, d, J = 1.9 Hz), 7.31 (1H, d, J = 9.2
Hz), 7.25 (1H, d, J = 1.9 Hz), 7.21 (1H, d, J = 2.1 Hz), 5.63 (1H, m),
5.00 (1/2 H, d, J = 15 Hz), 4.80 (1/2 H, d, J = 16 Hz), 4.08 (1/2 H, d, J
= 16 Hz), 4.00 (1/2 H, d, J = 15.3 Hz), 3.94 (3H, s), 3.90 (1H, m), 2.91
(3H, s), 2.89 (3H, s), 2.37 (1.5H, s), 2.36 (1.5 H, s), 0.62 (1.5 H, d, J =
6.5 Hz), 0.60 (1.5 H, d, J = 6.5 Hz) ppm.
Step vi: 2″-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-
2-oxooxazolidin-3-yl)methyl)-4′-chloro-2-methyl-4″-(trifluoro-
methyl)-[1,1′:3′,1″-terphenyl]-4-carboxylic Acid (10). Methyl 2″-
({(4S,5R)-5-[3,5-bis(trifluoromethyl)phenyl]-4-methyl-2-oxo-l,3-oxa-
zolidin-3-yl}methyl)-4,-chloro-2-methyl-4″-(trifluoromethyl)-1, T:3′,
1-terphenyl-4-carboxylate (25 mg, 0.034 mmol), aqueous potassium
hydroxide (300 mL, 3 M, 0.90 mmol) and ethanol (2 mL) were stirred
at 20 °C overnight. Volatiles were removed under reduced pressure.
The resulting residue was treated with brine followed by extraction with
ethyl acetate. The combined extracts were dried over Na2SO4 followed
by filtration and concentration in vacuo to afford an oil. This oil was
purified by preparative TLC on SiO2 (eluted with 50% EtOAc in
hexanes) to afford a clear glass. The glass was further purified by
reversed-phase preparative HPLC (Kromasil 100-5C18, 100 × 21.1
W
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
mm) eluting by a MeCN (0.1% TFA, v/v)/H2O (0.1% TFA, v/v)
gradient mixture affording 2″-({(4S,5R)-5-[3,5-bis(trifluoromethyl)-
phenyl]-4-methyl-2-oxo-l,3-oxazolidin-3-yl}methyl)-4,-chloro-2-meth-
yl-4″-(trifluoromethyl)-l,l,:3,l″-terphenyl-4-carboxylic acid (10, 22 mg,
yield 90%). MS ESI: calcd for C34H23ClF9NO4 [M + H]+ 716.12, found
= 6.6 Hz). MS ESI: calcd for C33H24F9N3O5 [M + H] 714.16, found
714.10.
4-(2′-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2,6′-dimethoxy-5′-(trifluoromethyl)-[3,3′-
bipyridin]-5-yl)-3-methylbenzoic Acid (15). 1H NMR (500 MHz,
CDCl3) ppm δ 8.30 (d, J = 2.3 Hz, 1H), 8.07 (s, 1H), 8.03 (d, J = 8.0 Hz,
1H), 7.92 (s, 1H), 7.79 (s, 1H), 7.77 (s, 2H), 7.57 (s, 1H), 7.39 (d, J =
7.9 Hz, 1H), 6.0 (br, 1H), 5.73 (d, J = 8.4 Hz, 1H), 4.85 (br, 1H), 4.40
(br, 1H), 4.11 (s, 3H), 4.04 (s, 3H), 3.99 (br, 1H), 2.42 (s, 3H), 0.80
(br, 3H); MS ESI: calcd for C36H33F6N5O6 [M + H] 744.17, found
744.07.
2,2,2-Trifluoroacetic Acid-4-(4′-(((4S,5R)-5-(3,5-bis-
(trifluoromethyl)phenyl)-4-methyl-2-oxooxazolidin-3-yl)methyl)-2-
methoxy-6′-(trifluoromethyl)-[3,3′-bipyridin]-5-yl)-3-methylben-
zoic Acid (16). 1H NMR (500 MHz, CD3OD) ppm δ 8.63 (1H, s), 8.30
(1H, d, J = 2.3 Hz), 7.98−8.00 (2H, m), 7.80−7.92 (5, m), 7.41 (1H, d,
J = 7.9 Hz), 5.80 (1/2 H, br), 5.72 (1/2 H, br), 4.81 (1H, br), 4.40 (2H,
br), 4.03 (3H, s), 2.40 (3H, s), 0.62 (3H, d, J = 6.5 hz); MS ESI: calcd
for C33H24F9N3O5 [M + H] 714.16, found 714.29.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-(trifluoromethyl)pyrimidin-5-yl)-4′-me-
thoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (17). 1H NMR
(500 MHz, DMSO-d6) ppm δ 8.83 (1H, s), 8.01 (1H, s), 7.95 (1H, s),
7.93 (2H, s), 7.89 (1H, d, J = 8.1 Hz), 7.58 (1H, d, J = 8.0 Hz), 7.41
(1H, s), 7.39 (1H, d, J = 7.6 hz), 7.32 (1H, d, J = 7.8 Hz), 5.82 (1H, br),
4.81 (1H, m), 4.38−4.48 (2H, m), 3.93 (3H, s), 2.39 (3H, s), 0.72 (3H,
d, J = 6.3 Hz); MS ESI: calcd for C33H24F9N3O5 [M + H] 714.16, found
714.02.
4-(5-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-
oxooxazolidin-3-yl)methyl)-2-(trifluoromethyl)pyrimidin-5-yl)-6-
methoxypyridin-3-yl)-3-methylbenzoic Acid (18). 1H NMR (500
MHz, CD3OD) ppm δ 8.91 (1H, s), 8.37 (1H, s), 8.03 (1H, s), 8.01
(1H, s), 7.95 (2H, s), 7.93 (1H, m), 7.88 (1H, s), 7.42 (1H, dd, J = 7.9,
1.5 Hz), 5.87 (1H, d, 9.0 Hz), 4.81 (1H, m), 4.40−4.12 (2H, m), 4.03
(3H, s), 2.40 (3H, s), 0.78 (2H, d, J = 6.7 Hz); MS ESI: calcd for
C32H23F9N4O5 [M + H] 715.15, found 715.07.
1
716.12. H NMR signals are doubled because of atropisomerism ‘H
NMR (CDCl3, 500 MHz) δ 7.99 (d, J = 8.0 Hz, 1H), 7.92 (d, J = 8.5 Hz,
1H), 7.85 (d, J = 5 Hz, 1H), 7.72−7.68 (m, 1,5H), 7.67−7.60 (m,
2.5H), 7.47−7.36 (m, 2H), 7.32 (d, J = 8.0 Hz, 0.5H), 7.28 (d, J = 8.0
Hz, 0.5H), 7.15 (s, 1H), 7.08 (dd, J = 8.5, 7.0 Hz, 1H), 5.58 (d, J = 8.0
Hz, 0.5H),5.29 (d, J = 6.5 Hz, 0.5H), 4.97 (d, J = 15.5 Hz, 0.5H), 4.94
(d, J = 14 Hz, 0.5H), 4.16 (d, J = 16 Hz, 0.5H), 3.96 (d, J = 15.5 Hz,
0.5H), 3.87 (s, 3H), 3.83−3.74 (m, 1H), 2.38, 2.33 (s, 3H), 0.55 (d, J =
6.5 Hz, 1.5H), 0.42 (d, J = 7 Hz, 1.5H) ppm.
Using this process with appropriate subunits, the following
compounds were prepared as representative of 2−37.
(4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-3-((6′-methoxy-4-(tri-
fluoromethyl)-[1,1′:3′,1″-terphenyl]-2-yl)methyl)-4-methyloxazoli-
din-2-one (4). 1H NMR (CDCl3, 500 MHz): δ 1:1 mixture of
atropisomers 7.88 (s, IH), 7.74 (s, IH), 7.72 (s, IH), 7.67 (t, J = 7 Hz,
IH), 7.62 (s, IH), 7.54 (m, IH), 7.31−7.34 (m, IH), 6.90−6.93 (m, IH),
5.63 (d, J = 8 Hz, 0.5H), 5.25 (d, J = 8 Hz, 0.5H), 4.98 (d, J = 15.5 Hz,
0.5H), 4.88 (d, J = 16 Hz, 0.5H), 4.12 (d, J = 15.5 Hz, 0.5H), 3.88 (d, J =
16.5 Hz, 0.5H), 3.84 (s, 3H), 3.81 (s, 3H), 3.73 (m, IH), 0.59 (d, J = 6.5
Hz, 1.5H), 0.45 (d, J = 6.5 Hz, 1.5H) ppm. MS ESI: calcd for
C33H24F9NO3 [M + H] 654.16, found 654.2.
2″-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-4′-methoxy-2-methyl-4″-(trifluorometh-
1
yl)-[1,1′:3′,1″-terphenyl]-4-carboxylic Acid (9). H NMR signals are
doubled because of atropoisomerism. 1H NMR (CDCl3, 500 MHz) δ
7.99 (d, J = 8.0 Hz, IH), 7.92 (d, J = 8.5 Hz, IH), 7.85 (d, J = 5 Hz, IH),
7.72−7.68 (m, 1.5H), 7.67−7.60 (m, 2.5H), 7.47−7.36 (m, 2H), 7.32
(d, J = 8.0 Hz, 0.5H), 7.28 (d, J = 8.0 Hz, 0.5H), 7.15 (s, IH), 7.08 (d, J =
8.5, 7.0 Hz, IH), 5.58 (d, J = 8.0 Hz, 0.5H), 5.29 (d, J = 6.5 Hz, 0.5H),
4.97 (d, J = 15.5 Hz, 0.5H), 4.94 (d, J = 14 Hz, 0.5H), 4.16 (d, J = 16 Hz,
0.5H), 3.96 (d, J = 15.5 Hz, 0.5H), 3.87 (s, 3H), 3.83−3.74 (m, IH),
2.38, 2.33 (s, 3H), 0.55 (d, J = 6.5 Hz, 1,5H), 0.42 (d, J = 7 Hz, 1,5H).
MS ESI: calcd for C35H26F9NO5 [M + H]+ 712.16, found 712.10.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-6-(trifluoromethyl)pyridin-3-yl)-4′-me-
thoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (11). 1H NMR
(500 MHz, CD3OD) ppm δ 8.59 (1H, s), 7.99 (1H, d, J = 2.3 Hz),
7.84−7.94 (5H, m), 7.51 (1H, dd, J = 8.4, 2.0 Hz), 7.25−7.36 (3H, m),
5.80 (1/2 H, d, J = 8.4 Hz), 5.55 (1/2 H, d, J = 7.9 Hz), 4.78 (1H, d, J =
16.4 Hz), 4.42 (1/2 H, d, J = 16.2 Hz), 4.24 (1/2H H, d, J = 16.4 Hz),
4.02 (1H, m), 3.90 (3H, s), 2.38 (1.5 H, s), 2.35 (1.5 H, s), 0.56 (1.5 H,
d, J = 7.5 Hz), 0.52 (1.5 H, d, J = 7.5 Hz). MS ESI: calcd for
C34H25F9N2O5 [M + H] 713.16, found 713.24.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-(dimethylamino)pyrimidin-5-yl)-4′-me-
thoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (19). 1H NMR
(600 MHz, CDCl3) ppm δ 8.22 (s, 1H), 8.01 (s, 1H), 7.95 (d, J = 8.9
Hz, 1H), 7.87 (s, 1H), 7.72 (s, 2H), 7.32−7.36 (m, 2H), 7.15 (s, 1H),
7.05 (d, J = 8.7 Hz, 1H), 5.61 (br, 1H), 4.79 (br, 1H), 4.34 (br, 1H),
4.10 (br, 1H), 3.86 (s, 3H), 3.25 (s, 6H), 2.37 (s, 3H), 0.68 (br, 3H);
13C NMR (125 MHz, CDCl3) ppm δ 171.13, 161.59, 161.30, 158.50,
157.61, 146.08, 138.38, 135.69, 133.59, 132.41, 132.19, 131.96, 131.74,
131.44, 130.18, 129.99, 128.23, 127.68, 126.41, 124.25, 123.85, 122.52,
122.05, 120.24, 117.61, 110.70, 77.28, 77.21, 77.00, 76.79, 55.56, 54.29,
44.74, 37.18, 20.55, 14.51; MS ESI: calcd for C34H30F6N4O5 [M + H]
689.21, found 689.34.
3′-(2-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-6-(trifluoromethyl)pyridin-3-yl)-4′-me-
thoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (12). 1H NMR
(500 MHz, CD3OD) ppm δ 7.98 (1H, s), 7.84−7.93 (5H, m), 7.81
(1H, br), 7.47 (1H, dd, J = 18.5, 2.2 hz), 7.34 (1H, d, J = 7.7 Hz), 7.24−
7.30 (2H, m), 4.86 (1/2 H, br), 4.81 (1/2 H, br), 4.81 (3H, s), 4.30−
4.33 (1H, m), 3.89 (3H, s), 2.36 (3H, s), 0.69 (3H, s), MS ESI: calcd for
C34H25F9N2O5 [M + H] 713.16, found 713.12.
4-(5-(2-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-
oxooxazolidin-3-yl)methyl)-4-(trifluoromethyl)phenyl)-6-methoxy-
pyridin-3-yl)-3-methylbenzoic Acid (13). 1H NMR (500 MHz,
CDCl3) ppm δ 8.30 (1H, s), 8.08 (1H, s), 8.01 (1H, d, J = 7.9 Hz),
7.91 (1H, s), 7.74 (2H, d, J = 10.4 Hz), 7.71 (2H, d, J = 7.6 Hz), 7.54
(1H, s), 7.46 (1H, d, J = 8.0 Hz), 7.38 (1H, d, J = 7.0 Hz). MS ESI: calcd
for C34H25F9N2O5 [M + H] 713.16, found 713.10.
4-(2′-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-methoxy-6′-(trifluoromethyl)-[3,3′-bi-
pyridin]-5-yl)-3-methylbenzoic Acid (14). 1H NMR (500 MHz,
CD3OD) ppm δ 8.28 (1H, 2.3 Hz), 7.98 (1H, s), 7.96 (1H, d, J =
8.1 Hz), 7.94 (1H, d, J = 8.1 Hz), 7.90 (3H, m), 7.84 (1H, d, J = 8.0 Hz),
7.77 (1H, d, J = 2.0 Hz), 7.41 (1H, d, J = 7.9 Hz), 5.49 (1H, br), 4.83
(1H, m), 4.32−4.40 (2H, m), 4.02 (3H, s), 2.40 (3H, s), 0.73 (3H, d, J
3′-(2-(Azetidin-1-yl)-4-(((4S,5R)-5-(3,5-bis(trifluoromethyl)-
phenyl)-4-methyl-2-oxooxazolidin-3-yl)methyl)pyrimidin-5-yl)-4′-
methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (20). 1H NMR
(500 MHz, CDCl3) ppm δ 8.22 (s, 1H), 8.01 (s, 1H), 7.95 (d, J = 8.9
Hz, 1H), 7.87 (s, 1H), 7.78 (s, 2H), 7.34−7.38 (m, 2H), 7.19 (s, 1H),
7.05 (d, J = 8.7 Hz, 1H), 5.61 (br, 1H), 4.82 (br, 1H), 4.39 (br, 1H),
4.25 (m, 4H), 4.10 (br, 1H), 3.86 (s, 3H), 2.42 (m, 2H), 2.39 (s, 3H),
0.68 (br, 3H); MS ESI: calcd for C35H30F6N4O5 [M + H] 701.20, found
701.30.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-(3-fluoroazetidin-1-yl)pyrimidin-5-yl)-
4′-methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (21). 1H
NMR (500 MHz, CDCl3) ppm δ 8.22 (s, 1H), 8.01 (s, 1H), 7.95 (d,
J = 8.9 Hz, 1H), 7.89 (s, 1H), 7.78 (s, 2H), 7.32−7.36 (m, 2H), 7.19 (s,
1H), 7.05 (d, J = 8.7 Hz, 1H), 5.61 (br, 1H), 5.41−5.58 (m, 1H), 4.79
(br, 1H), 4.55 (m, 2H), 4.39 (br, 1H), 4.38 (m, 2H), 4.10 (br, 1H), 3.86
(s, 3H), 2.40 (s, 3H), 0.68 (br, 3H); MS ESI: calcd for C35H29F7N4O5
[M + H] 719.20, found 719.30.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-(pyrrolidin-1-yl)pyrimidin-5-yl)-4′-me-
thoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (22). 1H NMR
(500 MHz, CDCl3) ppm δ 8.39 (s, 1H), 7.98 (s, 1H), 7.95 (s, 1H), 7.92
X
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
(d, J = 8.9 Hz, 1H), 7.78 (s, 2H), 7.42 (d, J = 8.5 Hz, 1H), 7.32 (d, J =
8.6 Hz, 1H), 7.22 (s, 1H), 7.10 (d, J = 8.7 Hz, 1H), 5.71 (d, J = 9.0 Hz,
1H), 4.82 (br, 1H), 4.39 (br, 1H), 4.15 (br, 1H), 3.86 (s, 3H), 3.75−
3.85 (br, 4H), 2.15 (br, 4H), 2.39 (s, 3H), 0.78 (br, 3H); MS ESI: calcd
for C36H32F6N4O5 [M + H] 715.23, found 715.22.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-morpholinopyrimidin-5-yl)-4′-me-
thoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (23). 1H NMR
(500 MHz, CDCl3) ppm δ 8.24 (1H, s), 8.03 (1H, s), 7.98 (1H, dd, J =
8.0, 2.1 Hz), 7.88 (1H, s), 7.74 (2H, s), 7.40 (1H, dd, J = 8.0, 2.0 Hz),
7.37 (1H, d, J = 8.1 Hz), 7.18 (1H, s), 7.08 (1H, d, J = 8.2 Hz), 5. 58
(1H, br), 4.83 91H, br), 4.22 (1H, br), 4.0 (1H, br), 3.8−3.93 (7H, m),
2.40 (3H, s), 0.65 (3H, br); MS ESI: calcd for C36H32F6N4O6 [M + H]
731.22, found 731.27.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-morpholinopyrimidin-5-yl)-4′-me-
thoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (24). 1H NMR
(500 MHz, CDCl3) ppm δ 9.0−9.4 (br, 1H), 8.30 (s, 1H), 8.01 (s, 1H),
7.98 (d, J = 8.6 Hz, 1H), 7.89 (s, 1H), 7.72 (s, 2H), 7.42 (d, J = 8.6 Hz,
1H), 7.38 (d, J = 8.7 Hz, 1H), 7.18 (s, 1H), 7.15 (d, J = 8.7 Hz, 1H),
5.61 (br, 1H), 4.85 (br, 1H), 4.25 (br, 1H), 4.18 (br, 1H), 3.91 (br,
3H), 3.8−3.9 (m, 8H), 2.39 (s, 3H), 0.60-0.80 (br, 3H); MS ESI: calcd
for C36H32F6N4O6 [M + H] 731.22, found 731.55.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-ethylpyrimidin-5-yl)-4′-methoxy-2-
methyl-[1,1′-biphenyl]-4-carboxylic Acid (25). 1H NMR (500 MHz,
CDCl3) ppm δ 8.60 (s, 1H), 8.02 (s, 1H), 7.98 (d, J = 8.9 Hz, 1H), 7.89
(s, 1H), 7.78 (s, 2H), 7.45 (d, J = 8.8 Hz, 1H), 7.38 (d, J = 8.7 Hz, 1H),
7.21 (s, 1H), 7.15 (d, J = 9.0 Hz, 1H), 5.65 (br, 1H), 4.95 (br, 1H), 4.40
(br, 1H), 4.20 (br, 1H), 3.91 (s, 3H), 3.16 (q, J = 7.0 Hz, 2H), 2.39 (s,
3H), 1.41 (t, J = 7.1 Hz, 3H), 0.79 (br, 3H); MS ESI: calcd for
C34H29F6N3O5 [M + H] 674.20, found 674.22.
3′-(4-(((4S,5R)-5-(3,5-Bis(trifluoromethyl)phenyl)-4-methyl-2-ox-
ooxazolidin-3-yl)methyl)-2-(1,1-difluoroethyl)pyrimidin-5-yl)-4′-
methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (26). 1H NMR
(500 MHz, CD3OD) ppm δ 8.78 (s, 1H), 8.01 (s, 1H), 7.98 (m, 3H),
7.85 (d, J = 8.7 Hz, 1H), 7.55 (d, J = 8.7 Hz, 1H), 7.39 (m, 2H), 7.28 (d,
J = 8.8 Hz, 1H), 5.82 (br, 1H), 4.80 (br, 1H), 3.92 (s, 3H), 4.40 (br,
2H), 2.39 (s, 3H), 2.10 (t, J = 25 Hz, 3H), 0.72 (d, J = 6.5 Hz, 3H); MS
ESI: calcd for C34H27F8N3O5 [M + H] 710.18, found 710.16.
3′-(2-(Dimethylamino)-4-(((4S,5R)-4-methyl-2-oxo-5-(3-
(trifluoromethyl)phenyl)oxazolidin-3-yl)methyl)pyrimidin-5-yl)-4′-
methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (27). 1H NMR
(500 MHz, CD3OD) ppm δ 8.14 (s, 1H), 7.94 (s, 1H), 7.88 (d, J =
7.8Hz, 1H), 7.68 (d, J = 7.9Hz, 1H), 7.56−7.61 (m, 3H), 7.45−7.47 (d,
J = 8.5 Hz, 1H), 7.36 (d, J = 8.0 Hz, 1H), 7.22−7.26 (m, 2H),5.75 (d, J
= 8.6 Hz, 1H), 4.65 (br, 1H), 4.39 (br, 1H), 4.21 (br, 1H), 3.91 (s, 3H),
3.24 (s, 6H), 2.36 (s, 3H), 0.70 (d, J = 6.4 Hz, 3H); MS ESI: calcd for
C33H31F3N4O5 [M + H] 621.22, found 621.34.
3′-(2-(Dimethylamino)-4-(((4S,5R)-4-methyl-2-oxo-5-(3-
(trifluoromethoxy)phenyl)oxazolidin-3-yl)methyl)pyrimidin-5-yl)-
4′-methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (28). 1H
NMR (500 MHz, CD3OD) ppm δ 8.12 (s, 1H), 7.93 (s, 1H), 7.88
(d, J = 7.8 Hz, 1H), 7.49 (m, 1H), 7.40 (d, J = 8.4 Hz, 1H), 7.35 (d, J =
7.9 Hz, 1H), 7.27 (m, 3H), 7.19 (m, 2H), 5.65 (br, 1H), 4.65 (br, 1H),
4.30 (br, 1H), 4.10 (br, 1H), 3.88 (s, 3H), 3.24 (s, 6H), 2.36 (s, 3H),
0.69 (s, 3H); MS ESI: calcd for C33H31F3N4O6 [M + H] 637.22, found
637.22.
3′-(2-(Dimethylamino)-4-(((4S,5R)-4-methyl-2-oxo-5-(4-
(trifluoromethoxy)phenyl)oxazolidin-3-yl)methyl)pyrimidin-5-yl)-
4′-methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (29). 1H
NMR (500 MHz, CD3OD) ppm δ 8.12 (1H, s), 7.93 (1H, s), 7.87
(1H, d, J = 8.6 Hz), 7.34−7.42 (5H, m), 7.29 (1H, d, J = 8.4 hz), 7.20
(1H, s), 7.19 (1H, d, J = 8.7 Hz), 5.64 (1H, d, J = 7.2 Hz), 4.74 (1H, br),
4.30 (1H, br), 4.09 (1H, br), 3.88 (3H, s), 3.24 (6H, s), 2.36 (3H, s),
0.68 (3H, d, J = 3.1 Hz); MS ESI: calcd for C33H31F3N4O6 [M + H]
637.22, found 637.43.
3′-(4-(((4S,5R)-5-(3-Chlorophenyl)-4-methyl-2-oxooxazolidin-3-
yl)methyl)-2-(dimethylamino)pyrimidin-5-yl)-4′-methoxy-2-meth-
yl-[1,1′-biphenyl]-4-carboxylic Acid (30). 1H NMR (500 MHz,
CD3OD) ppm δ 8.11 (s, 1H), 7.93 (s, 1H), 7.88 (d, J = 7.9 Hz, 1H),
7.33−7.41 (m, 4H), 7.30 (s, 1H), 7.20 (m, 3H), 5.60 (br, 1H), 4.61 (br,
1H), 4.29 (br, 1H), 4.10 (br, 1H), 3.88 (s, 3H), 3.31 (s, 6H), 2.36 (s,
3H), 0.68 (br, 3H); MS ESI: calcd for C32H31ClN4O5 [M + H] 587.20,
found 587.34.
3′-(4-(((4S,5R)-5-(3-Chloro-5-(trifluoromethyl)phenyl)-4-methyl-
2-oxooxazolidin-3-yl)methyl)-2-(dimethylamino)pyrimidin-5-yl)-
4′-methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (31). MS
ESI: calcd for C33H30ClF3N4O5 [M + H] 655.19, found 655.16.
3′-(2-(Dimethylamino)-4-(((4S, 5R)-5-(3-fluoro-5-
(trifluoromethyl)phenyl)-4-methyl-2-oxooxazolidin-3-yl)methyl)-
pyrimidin-5-yl)-4′-methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic
Acid (32). 1H NMR (500 MHz, CD3OD) ppm δ 8.12 (s, 1H), 7.93 (s,
1H), 7.87 (d, J = 7.7 Hz, 1H), 7.42−7.48 (m, 2H), 7.35−7.41 (m, 3H),
7.18−7.20 (m, 2H), 5.73 (br, 1H), 4.66 (br, 1H), 4.33 (br, 1H), 4.08
(br, 1H), 3.90 (s, 3H), 3.24 (s, 6H), 2.36 (s, 3H), 0.69 (s, 3H); MS ESI:
calcd for C33H30F4N4O5 [M + H] 639.22, found 639.39.
3′-(4-(((4S,5R)-5-(3,5-Dichlorophenyl)-4-methyl-2-oxooxazoli-
din-3-yl)methyl)-2-(dimethylamino)pyrimidin-5-yl)-4′-methoxy-2-
methyl-[1,1′-biphenyl]-4-carboxylic Acid (33). 1H NMR (500 MHz,
CD3OD) ppm δ 8.11 (s, 1H), 7.92 (s, 1H), 7.87 (d, J = 7.9 Hz, 1H),
7.44 (s, 1H), 7.40 (d, J = 8.5 Hz, 1H), 7.34 (d, J = 7.9 Hz, 1H), 7.25 (s,
2H), 7.17−7.19 (m, 2H), 5.59 (br, 1H), 4.61 (br, 1H), 4.29 (br, 1H),
4.10 (br, 1H), 3.88 (s, 3H), 3.28 (s, 6H), 2.35 (s, 3H), 0.71 (br, 3H);
MS ESI: calcd for C32H30Cl2N4O5 [M + H] 621.16, found 621.34.
3′-(4-(((4S,5R)-5-(3,5-Difluorophenyl)-4-methyl-2-oxooxazolidin-
3-yl)methyl)-2-(dimethylamino)pyrimidin-5-yl)-4′-methoxy-2-
methyl-[1,1′-biphenyl]-4-carboxylic Acid (34). 1H NMR (500 MHz,
CD3OD) ppm δ 8.14 (s, 1H), 7.95 (s, 1H), 7.88 (d, J = 8.0 Hz, 1H),
7.46 (m, 1H), 7.36 (d, J = 7.9 Hz, 1H), 7.21−7.25 (m, 2H), 6.93−6.99
(m, 3H), 5.65 (d, J = 8.5 Hz, 1H), 4.65 (br, 1H), 4.35 (br, 1H), 4.20 (br,
1H), 3.91 (s, 3H), 3.24 (s, 6H), 2.36 (s, 3H), 0.74 (d, J = 6.4 Hz, 3H);
MS ESI: calcd for C32H30F2N4O5 [M + H] 589.22, found 589.22.
3′-(2-(Dimethylamino)-4-(((4S,5R)-4-methyl-2-oxo-5-(5-
(trifluoromethyl)pyridin-3-yl)oxazolidin-3-yl)methyl)pyrimidin-5-
1
yl)-4′-methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (35). H
NMR (500 MHz, CD3OD) ppm δ 8.92 (s, 1H), 8.80 (s, 1H), 8.15 (s,
1H), 8.08 (s, 1H), 7.95 (s, 1H), 7.88 (d, J = 9.0 Hz, 1H), 7.46 (d, J = 8.5
Hz, 1H), 7.35 (d, J = 7.9 Hz, 1H), 7.24 (m, 2H), 5.76 (br, 1H), 4.66 (br,
1H), 4.40 (br, 1H), 4.10 (br, 1H), 3.89 (s, 3H), 3.28 (s, 6H), 2.37 (s,
3H), 0.75 (d, J = 6.3 Hz, 3H); MS ESI: calcd for C32H30F3N5O5 [M +
H] 622.22, found 622.39.
3′-(2-(Dimethylamino)-4-(((4S,5R)-4-methyl-2-oxo-5-(2-
(trifluoromethyl)pyridin-4-yl)oxazolidin-3-yl)methyl)pyrimidin-5-
1
yl)-4′-methoxy-2-methyl-[1,1′-biphenyl]-4-carboxylic Acid (36). H
NMR (500 MHz, CD3OD) ppm δ 8.74 (d, J = 7.8Hz, 1H), 8.12 (s, 1H),
7.94 (s, 1H), 7.86 (d, J = 8.5Hz, 1H), 7.76 (s, 1H), 7.60 (d, J = 4.7 Hz,
1H), 7.40 (d, J = 8.4 Hz, 1H), 7.35 (d, J = 8.0 Hz, 1H), 7.24 (m, 2H),
5.86 (d, J = 8.1 Hz, 1H), 4.68 (br, 1H), 4.42 (br, 1H), 4.20 (br, 1H),
3.89 (s, 3H), 3.28 (s, 6H), 2.37 (s, 3H), 0.70 (br, 3H); MS ESI: calcd
for C32H30F3N5O5 [M + H] 622.22, found 622.32.
2,2,2-Trifluoroacetic acid--3′-(2-(dimethylamino)-4-(((4S,5S)-4-
methyl-2-oxo-5-(6-(trifluoromethyl)pyridin-2-yl)oxazolidin-3-yl)-
methyl)pyrimidin-5-yl)-4′-methoxy-2-methyl-[1,1′-biphenyl]-4-car-
boxylic Acid (37). 1H NMR (500 MHz, CD3OD) ppm δ 8.13 (1H, s),
8.10 (1H, t, J = 7.9 Hz), 7.94 (1H, s), 7.88 (1H, d, J = 7.4 hz), 7.80 (1H,
d, J = 7.8 Hz), 7.70 (1H, d, J = 8.0 Hz), 7.44 (1H, dd, J = 8.5, 2.1 Hz),
7.36 (1H, d, J = 7.9 Hz), 7.26 (1H, s), 7.22 (1H, d, J = 8.4 Hz), 5.69
(1H, d, J = 8.1 Hz), 4.78 (1H, d, J = 10.2 hz), 4.75 (1H, br), 4.46 (1H,
br), 4.17 (1H, br), 3.90 (3H, s), 3.29 (6H, s), 2.37 (3H, s), 0.71 (3H, d,
J = 6.1 Hz); MS ESI: calcd for C32H30F3N5O5 [M + H] 622.22, found
622.32.
Preparation of 38 via first-generation (Scheme 1) and second-
generation synthesis (Scheme 2b):
(2S)-tert-Butyl 2-((3,5-Bis(trifluoromethyl)phenyl)((tert-
butyldimethylsilyl)oxy)methyl)pyrrolidine-1-carboxylate (40). To a
solution of 1-bromo-3,5-bis(trifluoromethyl)benzene (1.85 g, 6.32
mmol) in 6 mL of THF was added nBuLi (2.4 mL, 6.1 mmol of hexane
solution) dropwise via a syringe at −78 °C. After the resulting reaction
mixture was stirred at −78 °C, 5 min, a solution of (S)-tert-butyl 2-
formylpyrrolidine-1-carboxylate (1.05 g, 5.3 mmol) in 5 mL of THF
was added via a syringe. The resulting reaction mixture was allowed to
Y
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
warm up to ambient temperature slowly. TBSOTf was added. After
stirring at ambient temperature for 30 min, it was diluted with 60 mL of
EtOAc/hexane (1:4) and washed with 60 mL of water. Organics were
separated and dried over sodium sulfate, filtered, and concentrated.
Crude product was purified on Combiflash prepacked silica gel column
eluted with gradient solvents from hexane to 1:9 EtOAc/hexane to give
the desired product (1.2 g, 43% yield) of mixture of two diastereomers
as a yellow liquid. The two diastereomers each show as a pair of
rotamers. It is difficult to assign peaks of H NMR spectra. H NMR
(500 MHz, CDCl3) δ 7.92 (s), 7.82 (s), 7.80 (s), 7.77 (s), 5.3−5.6 (m,
1H), 3.8−4.2 (m, 1H), 1.2−3.6 (m, 6H), 1.54−1.55 (s, 9H), 0.95-0.96
(s, 9H), 0.04 (s, 6H).
(2S)-tert-Butyl 2-((3,5-Bis(trifluoromethyl)phenyl)((tert-
butyldimethylsilyl)oxy)methyl)-5-oxopyrrolidine-1-carboxylate
(41). To a 250 mL round-bottom flask were added 40 (0.60 g, 1.14
mmol), NaIO4 (0.61 g, 2.84 mmol), RuCl3·H2O (13 mg, 0.057 mmol),
30 mL of EtOAc, and 30 mL of water. After flushing with N2, the
reaction mixture was stirred vigorously overnight. TLC indicated
complete conversion. It was diluted with 30 mL of water/30 mL of
hexane. Organics were separated, dried over sodium sulfate, filtered,
and concentrated. Crude product was purified on Combiflash
prepacked silica gel column eluted with hexane to 1:4 EtOAc/hexane
to give the desired product (0.41 g, 67% yield) as a colorless sticky
material. It solidified over time. NMR indicated a pair of diastereomers
at 1:1 ratio. 1H NMR (500 MHz, CDCl3) δ 7.90 (s), 7.86 (s), 7.77 (s),
5.38 (s), 5.38 (d, J = 4.5 Hz), 4.5 (m), 4.3 (d, J = 9.5 Hz), 2.8 (m) 2.4
(m), 1.4−2.2 (m), 1.65 (s, 9H), 1.59 (s, 9H), 0.98 (s, 9H), 0.95 (s, 9H),
0.41 (6H).
tert-Butyl (5S)-5-((S)-(3,5-Bis(trifluoromethyl)phenyl)((tert-
butyldimethylsilyl)oxy)methyl)-2-(2-chloro-5-(trifluoromethyl)-
phenyl)-2-hydroxypyrrolidine-1-carboxylate (42). To a solution of 2-
bromo-1-chloro-4-(trifluoromethyl)benzene (0.38 g, 1.37 mmol) in 5
mL of THF was added a solution of iPrMgCl (0.68 mL, 1.37 mmol) in
THF at 0 °C. After stirring for 10 min, it was transferred to a flask
contains a solution of 41 (0.57 g, 1.05 mmol) in 6 mL of THF at −78
°C. After stirring at −78 °C for 5 min, it was allowed to warm up to
ambient temperature. It was quenched by addition of 20 mL of water
and 20 mL of NH4Cl. The mixture was extracted with 60 mL of E/H
(1:3). Organics were dried over sodium sulfate, filtered, and
concentrated. Crude product was purified on Combiflash prepacked
silica gel column eluted with hexane to 1:9 E/H give desired product
(0.54 g, 71%).
Combiflash prepacked silica gel column, eluted with hexane to 1:4 E/H.
Four fractions were collected. The first fraction contains mainly the
45b. The second fraction contains 45a and 45b (a:b ∼ 3:1 ratio, 45
mg), this material was used for the next step directly without further
separation. The third fraction contains mainly 45c. The fourth fraction
contains mainly 45d. 1H NMR (500 MHz, CDCl3) for 45a: δ 7.93 (s,
1H), 7.87 (s, 2H), 7.74 (s, 1H), 7.53 (s, 2H), 6.14 (d, J = 8.0 Hz, 1H),
5.37 (t, J = 8.1 Hz, 1H), 4.60 (m, 1H), 2.94 (m, 1H), 1.69 (m, 1H), 1.56
(m, 1H), 1.26 (m, 1H); 1H NMR (500 MHz, CDCl3) for 45b: δ 8.50 (s,
1H), 7.94 (s, 2H), 7.83 (s, 1H), 7.52 (s, 2H), 6.03 (br, 1H), 5.6 (br,
1H), 4.2 (br, 1H), 2.88 (br, 1H), 2.4 (br, 1H), 2.3 (br, 1H), 2.0 (br,
1H); 1H NMR (500 MHz, CDCl3) for 45c: δ 7.95 (s, 1H), 7.92 (s, 2H),
7.58 (m, 2H), 7.49 (s, 1H), 5.62 (d, J = 8.0 Hz, 1H), 5.24 (dd, J = 8.9
Hz, J = 2.4 Hz, 1H), 4.28 (m, 1H), 2.79 (m, 1H), 2.35 (m, 1H), 2.24 (m,
1H), 2.06 (m, 1H); 1H NMR (500 MHz, CDCl3) for 45d: δ 7.97 (s,
1H), 7.93 (s, 2H), 7.55 (s, 1H), 7.5 (m, 2H), 5.54 (d, J = 4.1 Hz, 1H),
5.35 (t, J = 7.9 Hz, 1H), 4.16 (m, 1H), 3.07 (m, 1H), 2.40 (m, 1H), 2.00
(m, 1H), 1.85 (m, 1H).
(1R,5S,7aS)-1-(3,5-Bis(trifluoromethyl)phenyl)-5-(4′-fluoro-5′-
isopropyl-2′-methoxy-4-(trifluoromethyl)-[1,1′-biphenyl]-2-yl)-
tetrahydropyrrolo[1,2-c]oxazol-3(1H)-one (38a) and (1R,5R,7aS)-1-
(3,5-bis(trifluoromethyl)phenyl)-5-(4′-fluoro-5′-isopropyl-2′-me-
thoxy-4-(trifluoromethyl)-[1,1′-biphenyl]-2-yl)tetrahydropyrrolo-
[1,2-c]oxazol-3(1H)-one (38b). To a vial contains 45a and 45b (45 mg,
87 μmol, 7a:7b ∼ 3:1 ratio) were added (4-fluoro-5-isopropyl-2-
methoxyphenyl)boronic acid (55 mg, 260 μmol), PdOAc2 (2.0 mg, 8.7
μmol), DtBPF (4.1 mg, 8.7 μmol), K3PO4 (74, 350 μmol), THF (1 mL)
and water (50 μL). The vial was flushed with N2 and was sealed. The
reaction mixture was heated at 70 °C overnight. The reaction mixture
was diluted with dimethyl sulfoxide (DMSO) and water and acidified
with TFA. It was purified on Gilson HPLC with a reversed-phase
column, eluted with water/CH3CN (modified with 0.05% TFA)
gradient solvent. Four fractions were collected. The first fraction
contains mainly unreacted 45b. The second fraction contains 45a (13
mg recovered). The third fraction contains mainly 38b and small
amount impurity from boronic acid fragment. It was repurified by prep-
TLC eluted with 1:4 E/H to give 4 mg of clean 38b as fluffy white solid
after lyophilizing from tBuOH and water. From the fourth fraction, it
afforded 38a (11 mg) as fluffy white solid after lyophilizing from
tBuOH and water. 1H NMR (500 MHz, CDCl3) for 38a major rotamer
(56%): δ 7.88 (s, 1H, overlap with 38a minor rotamer), 7.82 (s, 2H),
7.75 (s, 1H), 7.6 (d, 1H, overlap with 38a minor rotamer), 7.3 (d, 1H,
overlap with 38a minor rotamer), 6.92 (d, JH‑F = 8.6 Hz, 1H), 6.71 (d,
1
1
(3,5-Bis(trifluoromethyl)phenyl)((2S)-5-(2-chloro-5-
(trifluoromethyl)phenyl)pyrrolidin-2-yl)methanol (44). To a solution
42 (250 mg, 0.35 mmol) in 2 mL of DCM was added TFA (0.8 mL,
10.4 mmol). The resulting reaction mixture was stirred at ambient
temperature for 30 min. Volatiles were removed. Residue was dissolved
in 20 mL of EtOAc, washed with 30 mL of 5% KOH, then 10 mL of
brine. Organics were dried over sodium sulfate, filtered, and
concentrated. Residue was dissolved in 2 mL of THF and treated
with TBAF (0.70 mL, 1 M in THF) at room temperature for 30 min.
LC-Mass indicated complete conversion. MeOH (3 mL) was added.
NaBH4 (131 mg, 0.35 mmol) was added in portions at 0 °C. The
resulting reaction mixture was stirred at 0 °C for 2 h and allowed to
warm up to ambient temperature slowly. Volatiles were removed.
Residue was diluted with 20 mL of EtOAc, washed with 15 mL of KOH,
then 10 mL of brine. Organics were dried over sodium sulfate, filtered,
and concentrated. Crude product was purified on Combifalsh
prepacked silica gel column, eluted with hexane to 1:4 E/H. Fractions
contain desired product (mixture of diastereomers) were combined to
give 170 mg of viscous material. NMR indicated it contains a significant
amount of TBAF. It was carried on to the next step without further
purification.
J
H‑F = 12 Hz, 1H), 6.07 (d, J = 7.9 Hz, 1H), 5.08 (t, J = 8.3 Hz, 1H), 4.47
(m, 1H), 3.84 (s, 3H), 3.2 (m, 1H), 1−2.1 (m, 4H), 1.24 (d, J = 7.0 Hz,
3H), 1.19 (d, J = 6.9 Hz, 3H); 1H NMR (500 MHz, CDCl3) for 38a
minor rotamer (44%): δ 7.88 (s, 1H, overlap with 38a major rotamer),
7.81 (s, 2H), 7.72 (s, 1H), 7.6 (d, 1H, overlap with 8a major rotamer),
7.3 (d, 1H, overlap with 38a major rotamer), 7.18 (d, JH‑F = 8.6 Hz,
1H), 6.66 (d, JH‑F = 12 Hz, 1H), 6.04 (d, J = 8.0 Hz, 1H), 4.95 (t, J = 8.3
Hz, 1H), 4.55 (m, 1H), 3.68 (s, 3H), 3.2 (m, 1H, overlap with 8a major
rotamer), 1−2.1 (m, 4H), 1.31 (d, J = 6.9 Hz, 3H), 1.28 (d, J = 6.9 Hz,
3H); MS ESI calcd for C31H25F10NO3 [M + H]: 650.17, found 650.29;
1H NMR (500 MHz, CDCl3) for 38b major rotamer (81%): δ 7.96 (s,
1H, overlap with 38b minor rotamer), 7.83 (s, 2H), 7.57 (d, J = 8.2 Hz,
1H, overlap with 38b minor rotamer), 7.35 (d, J = 8.8 Hz, 1H, overlap
with 38b minor rotamer), 7.33 (d, J = 7.9 Hz, 1H, overlap with 38b
minor rotamer), 7.16 (s, 1H), 6.66 (d, JH‑F = 12.1 Hz, 1H), 5.84 (d, J =
9.0 Hz, 1H), 4.7 (m, 1H), 4.51 (dd, J = 9.4, 2.8 Hz, 1H), 3.70 (s, 3H),
3.23 (m, 1H), 2.45 (m, 1H), 2.2 (m, 1H), 1.8 (m, 1H), 1.2 (m, 1H),
1.28 (d, J = 6.9 Hz, 6H); 1H NMR (500 MHz, CDCl3) for 38b minor
rotamer (19%): δ 7.96 (s, 1H, overlap with 38b major rotamer), 7.9 (s,
2H), 7.55 (d, 1H, overlap with 38b major rotamer), 7.35 (d, 1H,
overlap with 38b major rotamer), 6.85 (d, JH‑F = 8.5 Hz, 1H), 6.70 (d,
(7aS)-1-(3,5-Bis(trifluoromethyl)phenyl)-5-(2-chloro-5-
(trifluoromethyl)phenyl)tetrahydropyrrolo[1,2-c]oxazol-3(1H)-one
(Diastereomers 45a−d). To a solution of 44 in 2 mL of DCM was
added a solution of phosgene in toluene. The resulting reaction mixture
was at ambient temperature for 30 min. It was diluted with 15 mL of
EtOAc and washed with 15 mL of 10% KOH. Organics were dried over
sodium sulfate, filtered, and concentrated. Residue was purified on
J
H‑F = 12.0 Hz, 1H), 5.87 (d, J = 9.0 Hz, 1H), 4.7 (m, 1H), 4.6 (m, 1H),
3.78 (s, 3H), 1.2−2.5 (m, 5H), 1.25 (d, 6H, overlap with 38b major
rotamer). MS ESI calcd for C31H25F10NO3 [M + H]: 650.17, found
650.29.
(1S,5R,7aS)-1-(3,5-Bis(trifluoromethyl)phenyl)-5-(4′-fluoro-5′-
isopropyl-2′-methoxy-4-(trifluoromethyl)-[1,1′-biphenyl]-2-yl)-
Z
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
tetrahydropyrrolo[1,2-c]oxazol-3(1H)-one (38c). 38c was synthe-
sized from 45c with a similar procedure for 38a. 1H NMR (500 MHz,
CDCl3) for 38c major rotamer (74%): δ 7.93 (s, 1H, overlap with 38c
minor rotamer), 7.84 (s, 2H), 7.39 (d, J = 7.9 Hz, 1H) 7.35 (d, JH‑F = 8.8
Hz, 1H), 6.69 (d, JH‑F = 12 Hz, 1H), 5.56 (d, J = 8.5 Hz, 1H), 4.6 (dd, J
= 9.0, 2.6 Hz, 1H), 4.15 (m, 1H), 3.74 (s, 3H), 3.24 (m, 1H), 2.55 (m,
1H), 2.45 (m, 1H), 2.25 (m, 2H), 1.29 (d, J = 6.9 Hz, 3H), 1.28 (d, J =
6.9 Hz, 3H). MS ESI calcd for C31H25F10NO3 [M + H]: 650.17, found
650.30.
sodium sulfate, filtered, and concentrated. Crude product was purified
on Combiflash prepacked silica gel column, eluted with 0−10% EtOAc
in hexane to give 1.65 g desired product as a white solid (72% yield). 1H
NMR (500 MHz, CDCl3): 7.69 (d, J = 8.1 Hz, 1H), 7.53 (d, J = 1.7 Hz,
1H), 7.40 (m, 6H), 5.83 (m, 1H), 5.28 (br, 1H), 5.1 (m, 5H), 2.2 (m,
1H), 2.15 (m, 1H), 1.9 (m, 1H), 1.75 (m, 1H).
(S,E)-Benzyl (5-(3,5-Bis(trifluoromethyl)phenyl)-1-(2-bromo-5-
(trifluoromethyl)phenyl)pent-4-en-1-yl)carbamate (56). To a 25
mL round-bottom flask equipped with a reflux condenser were added
material from prior step (1.65 g, 3.65 mmol), 1,3-bis(trifluoromethyl)-
5-vinylbenzene (1.8 g, 7.5 mmol), and 10 mL of dichloromethane. After
flushing with N2, Zhan catalyst (RC301 0.14 g, 0.19 mmol) was added
under N2. The reaction mixture was heated at 60 °C 30 for min. NMR
indicated 80% conversion. More bis(trifluoromethyl)-5-vinylbenzene
(0.6 g, 2.5 mmol) was added. The reaction mixture was heated at 60 °C
for 25 min. NMR indicated ∼90% conversion. It was diluted with
hexane and loaded on to Combiflash prepacked silica gel column,
eluted with hexane to 15% EtOAc in hexane to give 2.2 g desired
product as a white solid (90% yield). 1H NMR (500 MHz, CDCl3) for
56: 7.73 (s, 3H), 7.56 (s, 1H), 7.40 (m, 7H), 6.47 (d, J = 15.9 Hz, 1H),
6.4 (m, 1H), 5.3 (br, 1H), 5.2 (br, 1H), 5.1 (br, 2H), 2.4 (m, 2H), 2.05
(br, 1H), 1.9 (br, 1H).
Benzyl ((S)-3-((2R,3R)-3-(3,5-Bis(trifluoromethyl)phenyl)oxiran-2-
yl)-1-(2-bromo-5-(trifluoromethyl)phenyl)propyl)carbamate (57).
To a 250 mL round-bottom flask were added 13 (2.2 g, 3.36 mmol),
sodium tetraborate decahydrate (1.3 g, 3.36 mmol), 20 mL of 0.4 mM
ethylenediaminetetraacetic acid disodium salt dihydrate solution,
Bu4NHSO4 (0.11 g, 0.034 mmol), D-Epoxone (0.35 g, 1.35 mmol),
15 mL of CH3CN, and 20 mL of EtOAc. Oxone (6.2 g, 10.1 mmol) in
25 mL of 0.4 mM ethylenediaminetetraacetic acid disodium salt
dihydrate solution and 25 mL of K2CO3 (4.88 g, 35.3 mmol) in water
were added simultaneously into the flask via syringe pump over 1.2 h at
0 °C. NMR indicated only 30% conversion. After aqueous workup,
crude product mixture was subjected the same reaction condition two
time. NMR indicated 75% conversion. The reaction mixture was
diluted with 150 mL of E/H (1:1) and 100 mL of water. Organics were
washed with 50 mL of brine, dried over sodium sulfate, filtered, and
concentrated. Crude product was purified on Combiflash prepacked
silica gel column, eluted with hexane to 1:7 E/H to give 1.5 g desired
product as a white solid (67% yield). 1H NMR (500 MHz, CDCl3) for
57: 7.84 (s, 1H), 7.73 (s, 2H), 7.4 (d, J = 8.0 Hz, 1H), 7.38 (br, 5H), 5.4
(br, 1H), 5.2 (br, 1H), 5.13 (s, 2H), 3.79 (s, 1H), 3.03 (br, 1H), 2.0 (br,
sH), 1.75 (br, 1H).
(1S,5S,7aS)-1-(3,5-Bis(trifluoromethyl)phenyl)-5-(4′-fluoro-5′-iso-
propyl-2′-methoxy-4-(trifluoromethyl)-[1,1′-biphenyl]-2-yl)-
tetrahydropyrrolo[1,2-c]oxazol-3(1H)-one (38d). 38d was synthe-
sized from 45d with a similar procedure for 38a. 1H NMR (500 MHz,
CDCl3) for 38c major rotamer (53%): δ 7.98 (s, 1H, overlap with 38d
minor rotamer), 7.92 (s, 2H), 7.56 (s, 2H, overlop with 38d minor
rotamer), 6.96 (d, JH‑F = 9.6 Hz, 1H), 6.72 (d, JH‑F = 12 Hz, 1H), 5.42
(s, 1H), 5.09 (t, J = 8.0, Hz, 1H), 4.03 (m, 1H), 3.85 (s, 3H), 3.24 (m,
1H, overlop with minor rotamer), 1.6−2.4 (m, 4H), 1.29 (6H, overlap
1
with minor rotamer). H NMR (500 MHz, CDCl3) for 38c minor
rotamer (47%): δ 7.97 (s, 1H, overlap with 38d major rotamer), 7.88 (s,
2H), 7.59 (d, J = 7.3 Hz, 1H, overlop with 38d major rotamer), 7.32 (d,
J = 7.5 Hz, 1H), 7.18 (d, JH‑F = 8.7 Hz, 1H), 6.63 (d, JH‑F = 12 Hz, 1H),
5.42 (s, 1H, overlap with major rotamer), 4.94 (t, J = 8.0, Hz, 1H), 3.94
(m, 1H), 3.72 (s, 3H), 3.24 (m, 1H, overlop with minor rotamer), 1.6−
2.4 (m, 4H), 1.29 (6H, overlap with major rotamer). MS ESI calcd for
C31H25F10NO3 [M + H]: 650.17, found 650.29.
(R,E)-N-(2-Bromo-5-(trifluoromethyl)benzylidene)-2-methylpro-
pane-2-sulfinamide (54). To a THF solution (20 mL) of 53 (2.2 g, 8.7
mmol) and (R)-(+)-2-methyl-2-propanesulfinamide (1.16 g, 9.56
mmol) was added Ti(OEt)4 (3.65 mL, 17.4 mmol) via a syringe
dropwise. After addition is completed, the reaction mixture was stirred
at 40 °C for 1 h. NMR indicated complete conversion. Brine (100 mL)
was added, followed by 100 mL of EtOAc. The resulting mixture was
stirred at room temperature for 15 min. It was filtered. Layers were
separated from the filtrate. Organics were dried over sodium sulfate,
filtered, and concentrated. Residue was purified on Combiflash
prepacked silica gel column, eluted with 0−25% EtOAc in hexane to
give the desire product (2.6 g, 84% yield) as a colorless crystalline solid.
1H NMR (500 MHz, CDCl3) for 54: δ 9.01 (s, 1H), 8.29 (d, J = 1.8 Hz,
1H), 7.82 (d, J = 8.4 Hz, 1H), 7.61 (dd, J = 1.8, 8.3 Hz, 1H), 1.32 (s,
9H).
(R)-N-((S)-1-(2-Bromo-5-(trifluoromethyl)phenyl)pent-4-en-1-yl)-
2-methylpropane-2-sulfinamide (55). To a 250 mL round-bottom
flask cooled with a water bath was added 54 (2.6 g, 7.3 mmol) and 80
mL of DCM. Grignard reagent but-3-en-1-ylmagnesium bromide (17.5
mL, 0.5 M in THF) was added rapidly via a syringe. Right after addition,
the reaction mixture was quenched by addition of 100 mL of NH4Cl.
Hexane (150 mL) was added. Layers were separated. Organics were
dried over sodium sulfate, filtered, and concentrated. Crude product
was purified on Combiflash prepacked silica gel column, eluted with
hexane to 1:2 E/H to give 2 g desired product (66.5% yield). 1H NMR
(500 MHz, CDCl3) for 55: 7.70 (d, J = 8.3 Hz, 1H), 7.68 (d, J = 1.9 Hz,
1H), 7.40 (dd, J = 2.0, 8.3 Hz, 1H), 5.82 (m, 1H), 5.12 (dd, J = 1.6, 17.2
Hz, 1H), 5.06 (dd, J = 1.4, 10.2 Hz, 1H), 5.0 (m, 1H), 3.57 (d, J = 2.2
Hz, 1H), 2.2 (m, 1H), 2.15 (m, 1H), 2.0 (m, 1H), 1.9 (m, 1H), 1.22 (s,
9H).
(S)-Benzyl (1-(2-Bromo-5-(trifluoromethyl)phenyl)pent-4-en-1-
yl)carbamate. To a solution of 55 (2.0 g, 4.85 mmol) in 1.5 mL of
MeOH was added a solution of HCl in dioxane (8.5 mL, 4M). The
reaction mixture was stirred at room temperature for 30 min. Volatiles
were removed. Residue was diluted with 60 mL of E/H (1:3), washed
with 50 mL of 5% KOH, and then 30 mL of brine. Organics were dried
over sodium sulfate, filtered, and concentrated. Crude product was
diluted with 10 mL of DCM. DIEA (2.7 nL, 15.6 mmol) was added,
followed by addition of CBZ-Cl (1.33 mL, 7.8 mmol) dropwise via a
syringe at 0 °C. After 30 min, LC-Mass indicated a still significant
amount of sm. 0.5 mL of cBZ-Cl were added. The reaction mixture was
stirred at 0 °C for another 15 min. The reaction mixture was diluted
with 75 mL of E/H 1:3, washed sequentially with 100 mL of 5% KOH,
60 mL of sat. NH4Cl and 30 mL of brine. Organics were dried over
(1R,5S,7aS)-1-(3,5-Bis(trifluoromethyl)phenyl)-5-(2-bromo-5-
(trifluoromethyl)phenyl)tetrahydropyrrolo[1,2-c]oxazol-3(1H)-one
(58). To a microwave tube were added 57 (1.1 g, 1.64 mmol), DBU (50
mg, 0.33 mmol), and DMF (12 mL). The tube was sealed and flushed
with nitrogen gas. It was heated with microwave at 200 °C for 3 min.
The reaction mixture was diluted with 100 mL of NH4Cl and 100 mL of
E/H (1:2). Organics were separated and washed with 50 mL of NH4Cl,
dried over sodium sulfate, filtered, and concentrated. Crude product
was purified on Combiflash prepacked silica gel column, eluted with
hexane to 1:4 E/H to give 0.63 g desired product as a white solid (68%
1
yield). H NMR (500 MHz, CDCl3) for 58: δ 7.93 (s, 1H), 7.88 (s,
2H), 7.72 (d, J = 8.4 Hz, 1H), 7.7 (d, J = 1.9 Hz, 1H), 7.44 (dd, J = 2.0,
8.2 Hz, 1H), 6.14 (d, J = 8.0 Hz, 1H), 5.34 (t, J = 8.0 Hz, 1H), 4.61 (m,
1H), 3.0 (m, 1H), 1.67 (m, 1H), 1.55 (m, 1H), 1.26 (m, 1H).
Preparation of Bicyclic Derivatives 46−52. (1R,5S,7aS)-1-(3,5-
Bis(trifluoromethyl)phenyl)-5-(4′-fluoro-5′-isopropyl-2′-methoxy-
4-(trifluoromethyl)-[1,1′-biphenyl]-2-yl)tetrahydropyrrolo[1,2-c]-
thiazol-3(1H)-one (46). Step 1. A solution of (2S,5S)-tert-butyl 2-(3,5-
bis(trifluoromethyl)benzoyl)-5-(2-chloro-5-(trifluoromethyl)phenyl)-
pyrrolidine-1-carboxylate (500 mg, 0.848 mmol) in THF (30 mL) was
cooled to −78 °C. K-Selectride (1.695 mL, 1.695 mmol) was added
dropwise. The reaction was stirred at −78 °C for 1 h. The reaction was
diluted with EtOAc, quenched with NH4Cl solution, washed with brine,
dried over Na2SO4, filtered, and concentrated. The residue was purified
by column chromatography to obtain (2S,5S)-tert-butyl 2-((S)-(3,5-
bis(trifluoromethyl)phenyl)(hydroxy)methyl)-5-(2-chloro-5-
(trifluoromethyl)phenyl)pyrrolidine-1-carboxylate. 1H NMR (500
AA
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
MHz, CDCl3): δ 7.87 (s, 2H); 7.85 (s, 1H); 7.45−7.50 (m, 2H); 7.22
(s, 1H); 5.44 (d, J = 3.5 Hz, 1H); 5.24 (d, J = 8.4 Hz, 1H); 5.06 (dd, J =
8.1, 3.5 Hz, 1H); 4.54 (t, J = 8.1 Hz, 1H); 2.04−2.07 (m, 3H, merged
with solvent peak); 1.81−1.85 (m, 1H); 1.62−1.70 (m, 1H);1.55−1.59
(m, 3H, merged with water peak); 1.19−1.20 (m, 9H).
(1R,5S,7aS)-1-[3,5-Bis(trifluoromethyl)phenyl]-5-[4′-fluoro-2′-
methoxy-5′-(1-methylethyl)-4-(trifluoromethyl)biphenyl-2-yl]-
hexahydro-3H-pyrrolizin-3-one (47). Step 1. To a solution of triethyl
phosphonoacetate (3.8 mL, 16.9 mmol) in toluene (40.0 mL) was
added sodium hydride (0.68 g, 16.9 mmol) at 0 °C. The resulting
solution was stirred at room temperature for 0.5 h. The reaction was
cooled back to 0 °C, tert-butyl (2S,5S)-2-(3,5-bis(trifluoromethyl)-
benzoyl)-5-(2-chloro-5-(trifluoromethyl)phenyl)pyrrolidine-1-carbox-
ylate (5.00 g, 8.47 mmol) in toluene (20.0 mL) was added slowly. The
resulting solution was stirred at room temperature for 2.5 h. The
reaction was quenched by the addition of saturated NH4Cl solution
(40.0 mL). It was diluted with ethyl acetate (50 mL), and the layers
were separated. The organic layer was washed with water, dried over
anhydrous Na2SO4, and concentrated to yield tert-butyl (2S,5S)-2-
((Z)-1-(3,5-bis(trifluoromethyl)phenyl)-3-ethoxy-3-oxoprop-1-en-1-
yl)-5-(2-chloro-5-(trifluoromethyl)phenyl)pyrrolidine-1-carboxylate
as a solid that was carried forward to the next step without further
purification. LC-MS APCI calcd for C29H27ClF9NO4 [M-Boc]+ 559.9,
found 560.0.
Step 2. To a nitrogen-purged solution of tert-butyl (2S,5S)-2-((Z)-1-
(3,5-bis(trifluoromethyl)phenyl)-3-ethoxy-3-oxoprop-1-en-1-yl)-5-(2-
chloro-5-(trifluoromethyl)phenyl)pyrrolidine-1-carboxylate (2.6 g,
3.93 mmol) in ethyl acetate (30.0 mL) was added platinum(IV)
oxide (600 mg) and the resulting mixture was stirred at room
temperature under H2 for 16 h. The reaction mixture was filtered
through a bed of diatomaceous earth and the filtrate was concentrated
to yield tert-butyl (2S,5S)-2-((R)-1-(3,5-bis(trifluoromethyl)phenyl)-
3-ethoxy-3-oxopropyl)-5-(2-chloro-5-(trifluoromethyl)phenyl)-
pyrrolidine-1-carboxylate as a solid that was carried forward to the next
step without further purification. LC-MS APCI calcd for
C29H29ClF9NO4 [M + H − Boc]+ 562.2, found 562.2.
Step 3. To a solution of tert-butyl (2S,5S)-2-((R)-1-(3,5-bis-
(trifluoromethyl)phenyl)-3-ethoxy-3-oxopropyl)-5-(2-chloro-5-
(trifluoromethyl)phenyl)pyrrolidine-1-carboxylate (2.3 g, 3.47 mmol)
in dichloromethane (10.0 mL) was added trifluoroacetic acid (20.0
mL). The resulting solution was stirred at room temperature for 16 h.
The reaction solution was concentrated. The residue was partitioned
between ethyl acetate (100.0 mL) and aqueous 10% NaHCO3 solution
(50.0 mL). The organic layer was washed with water, dried over
anhydrous Na2SO4, and concentrated to yield ethyl (R)-3-(3,5-
bis(trifluoromethyl)phenyl)-3-((2S,5S)-5-(2-chloro-5-
(trifluoromethyl)phenyl)pyrrolidin-2-yl)propanoate. The crude prod-
uct was taken up in toluene (25.0 mL) in a 100 mL seal tube.
Triethylamine (1.93 mL, 14.9 mmol) was added, and the reaction
solution was heated at 100 °C for 16 h. The reaction solution was
concentrated. The residue was diluted with ethyl acetate (75.0 mL),
washed with water, dried over anhydrous Na2SO4, and concentrated.
The crude product was purified by flash column chromatography to
yield (1R,5S,7aS)-1-(3,5-bis(trifluoromethyl)phenyl)-5-(2-chloro-5-
(trifluoromethyl)phenyl)hexahydro-3H-pyrrolizin-3-one as a solid.
LC-MS APCI calcd for C22H15ClF9NO [M + H]+ 515.6, found
516.0. 1H NMR (300 MHz, DMSO-d6): δ 8.02 (s, 3H), 7.73−7.65 (m,
3H), 5.04 (t, J = 8.07 Hz, 1H), 4.63−4.55 (m, 1H), 4.30 (q, J = 7.77 Hz,
1H), 3.19−3.10 (m, 1H), 2.85 (q, J = 6.87 Hz, 1H), 2.63−2.56 (m,
1H), 1.60−1.54 (m, 1H), 1.37−1.33 (m, 1H), 1.15−1.00 (m, 1H).
Step 4. In a microwave vial was taken (1R,5S,7aS)-1-(3,5-
bis(trifluoromethyl)phenyl)-5-(2-chloro-5-(trifluoromethyl)phenyl)-
hexahydro-3H-pyrrolizin-3-one (20 mg, 0.037 mmol) in N,N-
dimethylacetamide (0.5 mL). The mixture was degassed with nitrogen.
Chloro(2-dicyclohexylphosphino-2′,4′,6′-triisopropyl-1,1′-biphenyl)-
[2-(2′-amino-1,1′-biphenyl)]palladium(II) (2.95 mg, 3.75 μmol),
aqueous K3PO4 solution (0.037 mL, 0.075 mmol) were added,
followed by (4-fluoro-5-isopropyl-2-methoxyphenyl)boronic acid
(11.92 mg, 0.056 mmol). The resulting mixture was stirred at 85 °C
for 1 h. After cooling, EtOAc was added. The resulting mixture was
washed sequentially with NH4Cl solution and brine, dried over Na2SO4,
filtered, and concentrated. The crude product was purified by reversed-
phase HPLC to yield the title compound 47. Calcd 648.2, found 649.2.
Preparation of Compounds 48a−b and 49. ((1R,5S,7aS)-1-
(3,5-Bis(trifluoromethyl)phenyl)-5-(2-chloro-5-(trifluoromethyl)-
Step 2. To a solution of (2S,5S)-tert-butyl 2-((S)-(3,5-bis-
(trifluoromethyl)phenyl)(hydroxy)methyl)-5-(2-chloro-5-
(trifluoromethyl)phenyl)pyrrolidine-1-carboxylate (100 mg, 0.169
mmol) in DCM (1.5 mL) was added Et3N (0.031 mL, 0.220 mmol).
The resulting solution was kept at 0 °C and methanesulfonyl chloride
(0.014 mL, 0.186 mmol) was added. The reaction mixture was stirred
for 30 min. It was diluted with DCM, washed sequentially with NH4Cl
and brine, dried over Na2SO4, filtered, and concentrated. The residue
was purified by column chromatography to obtain (2S,5S)-tert-butyl 2-
((S)-(3,5-bis(trifluoromethyl)phenyl)((methylsulfonyl)oxy)methyl)-
5-(2-chloro-5-(trifluoromethyl)phenyl)pyrrolidine-1-carboxylate. 1H
NMR (500 MHz, CDCl3): δ 7.96−7.97 (m, 1H); 7.82 (s, 2H); 7.45
(s, 2H); 7.20 (s, 1H); 6.42 (d, J = 4.9 Hz, 1H); 4.74−4.79 (m, 2H); 3.17
(s, 3H); 1.96−2.10 (m, 4H); 1.41−1.45 (m, 1H); 1.21 (s, 8H).
Step 3. To a solution of (2S,5S)-tert-butyl 2-((S)-(3,5-bis-
(trifluoromethyl)phenyl)((methylsulfonyl)oxy)methyl)-5-(2-chloro-
5-(trifluoromethyl)phenyl)pyrrolidine-1-carboxylate (1.5 g, 2.239
mmol) in DCM (3 mL) was added TFA (2.59 mL, 33.6 mmol). The
reaction was stirred for 45 min. The volatiles were removed, and the
residue was dissolved in EtOAc. The resulting mixture was washed with
saturated NaHCO3, then brine, dried over Na2SO4, filtered, and
concentrated to afford (S)-(3,5-bis(trifluoromethyl)phenyl)((2S,5S)-
5-(2-chloro-5-(trifluoromethyl)phenyl)pyrrolidin-2-yl)methyl metha-
nesulfonate. It was used for the next step without further purification.
Step 4. To a solution of (S)-(3,5-bis(trifluoromethyl)phenyl)-
((2S,5S)-5-(2-chloro-5-trifluoromethyl)phenyl)pyrrolidin-2-yl)methyl
methanesulfonate (170 mg, 0.298 mmol) in DCM (3 mL) was added
2,6-di-tert-butyl-4-methylpyridine (123 mg, 0.597 mmol), followed by
SS-isopropyl carbonochlorido(dithioperoxoate) (102 mg, 0.597
mmol) at 0 °C. The reaction was stirred for 30 min. To the resulting
mixture was added MeOH (1.00 mL) followed by triphenylphosphine
(156 mg, 0.597 mmol). The reaction mixture was stirred for 30 min.
The volatiles were removed. The reaction was diluted with DCM,
neutralized with satd. NaHCO3, washed with brine, dried over Na2SO4,
filtered, and concentrated. The residue was purified by column
chromatography to afford (1R,5S,7aS)-1-(3,5-bis(trifluoromethyl)-
phenyl)-5-(2-chloro-5-(trifluoromethyl)phenyl)tetrahydropyrrolo-
1
[1,2-c]thiazol-3(1H)-one. H NMR (500 MHz, CDCl3): δ 7.95 (s,
2H); 7.93 (s, 1H); 7.56 (s, 1H); 7.52 (s, 2H); 5.35 (t, J = 7.9 Hz, 1H);
5.30 (d, J = 8.2 Hz, 1H); 4.86−4.91 (m, 1H); 2.75−2.81 (m, 1H);
1.78−1.86 (m, 1H); 1.54−1.62 (m, 2H); 1.33−1.41 (m, 1H). MS ESI/
APCI calcd for C21H13ClF9NOS [M + H]+ 533.8, found 533.8.
Step 5. In a microwave vial was taken (1R,5S,7aS)-1-(3,5-
bis(trifluoromethyl)phenyl)-5-(2-chloro-5-(trifluoromethyl)phenyl)-
tetrahydropyrrolo[1,2-c]thiazol-3(1H)-one (20 mg, 0.037 mmol) in
N,N-dimethylacetamide (0.5 mL). The mixture was degassed with
nitrogen. Chloro(2-dicyclohexylphosphino-2′,4′,6′-triisopropyl-1,1′-
biphenyl)[2-(2′-amino-1,1′-biphenyl)]palladium(II) (2.95 mg, 3.75
μmol), aqueous K3PO4 solution (0.037 mL, 0.075 mmol) were added,
followed by (4-fluoro-5-isopropyl-2-methoxyphenyl)boronic acid
(11.92 mg, 0.056 mmol). The resulting mixture was stirred at 85 °C
for 1 h. After cooling, EtOAc was added. The resulting mixture was
washed sequentially with NH4Cl solution and brine, dried over Na2SO4,
filtered, and concentrated. The crude product was purified by reversed-
1
phase HPLC to yield the title compound 46. H NMR (500 MHz,
CDCl3, 1:1 mixture of atropisomers): δ 7.81−7.84 (m, 6 H, contains for
two atropisomers protons); 7.53−7.56 (m, 4 H, contains for two
atropisomers protons); 7.28 (s, 2H, contains for two atropisomers
protons); 7.20 (d, J = 8.6 Hz, 1H); 6.88 (d, J = 8.6 Hz, 1H); 6.65 (dd, J
= 23.7, 12.0 Hz, 2H, contains for two atropisomers protons); 5.22 (d, J
= 8.3 Hz, 1H); 5.12−5.14 (m, 1H); 5.02 (t, J = 8.1 Hz, 1H); 4.84 (t, J =
8.1 Hz, 1H); 4.73−4.78 (m, 1H); 4.58−4.63 (m, 1H); 3.80 (s, 3H);
3.65 (d, J = 0.6 Hz, 3H); 3.14−3.24 (m, 2H); 1.17−1.29 (m, 12H). MS
ESI/APCI calcd for C31H25F10NO2S [M + H]+ 665.6, found 666.0.
RTA (95% HS): 2350 nM.
AB
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
phenyl)tetrahydro-1H-pyrrolo[1,2-c]imidazol-3(2H)-one (Inter-
mediate 8-1). Step 1. To a 500 mL three-neck RBF equipped with a
magnetic stirring bar were added, under N2 atmosphere, 3-bromo-4-
chloro-benzotrifluoride (16.5 mL, 110 mmol) and dry THF (150 mL).
The resulting solution was cooled to 0 °C. To this solution was added i-
PrMgBr (110 mL, 110 mmol) and the reaction was warmed to rt and
stirred for 1 h. To another 500 mL RBF containing a solution of CuBr·
Me2S (18.8 g, 91.5 mmol) in THF (100 mL) was added BF3·OEt2 (14
mL, 110 mmol) dropwise at −40 °C. The resulting solution was stirred
at that temperature for 45 min. Then, the Grignard reagent prepared
previously was added dropwise to this solution while maintaining the
internal temperature at −32 to −38 °C. After completion of addition,
the reaction mixture was stirred at that temperature for 1 h. The
reaction mixture was cooled to −78 °C, and a solution of (2S)-1-t-butyl
2-ethyl 5-methoxypyrrolidine-1,2-dicarboxylate (25.0 g, 91.5 mmol) in
THF (50 mL) was added dropwise. The resulting reaction mixture was
allowed to warm up to rt and was stirred for 14 h. The reaction mixture
was cooled to 0 °C, and saturated aqueous NH4Cl solution (100 mL)
was added followed by aqueous NH4OH solution (100 mL) and water
(300 mL). The layers were separated. The aqueous layer was extracted
with EtOAc (3 × 200 mL). The combined organics were dried over
anhydrous Na2SO4 and concentrated. The crude product was purified
by column chromatography eluting with 5% EtOAc in petroleum ether
to yield (2S,5S)-1-tert-butyl 2-ethyl 5-(2-chloro-5-(trifluoromethyl)-
phenyl)pyrrolidine-1,2-dicarboxylate (34.0 g, 79.8 mmol) as a colorless
liquid. 1H NMR (400 MHz, CDCl3) δ 7.32−7.47 (m, 3 H), 5.39−5.50
(m, 1H), 4.58−4.70 (m, 1H), 4.22−4.28 (m, 2H), 2.55 (m, 1H), 2.18
(m, 1H), 2.05 (m, 1H), 1.83 (m, 1H), 1.16−1.43 (m, 12H); LC-MS
ESI calcd for C19H23ClF3NO4 [M-Boc]+ 321.84, found 322.2.
carried forward to the next step without further purification. LC-MS
ESI calcd for C20H16ClF9N2 [M + H]+ 491.1, 493.1, found 491.0, 493.0.
Step 5. To a solution of (R)-(3,5-bis(trifluoromethyl)phenyl)-
((2S,5S)-5-(2-chloro-5-(trifluoromethyl)phenyl)pyrrolidin-2-yl)-
methanamine (525 mg, 1.07 mmol) in CH2Cl2 (10 mL) were added
Et3N (0.45 mL, 3.21 mmol) and triphosgene (476 mg, 1.60 mmol) at 0
°C. The resulting reaction mixture was allowed to warm up to rt and
stirred for 3 h. It was then diluted with CH2Cl2, rinsed with 1.5 N HCl
aqueous solution and water, dried over anhydrous Na2SO4, and
concentrated. The crude product was purified by column chromatog-
raphy eluted with 10% EtOAc in petroleum ether to yield Intermediate
8-1 (180 mg, 0.35 mmol). 1H NMR (400 MHz, CDCl3) δ 7.89 (m, 3
H), 7.80 (s, 1H), 7.48 (d, 2H, J = 2 Hz), 5.51 (s, 1H), 5.43−5.45 (d, 1H,
J = 9 Hz), 5.27−5.31 (t, 1H, J = 8 Hz), 4.41−4.47 (m, 1H), 2.83−2.91
(m, 1H), 1.51−1.59 (m, 1H), 1.41 (m, 2H); LC-MS ESI calcd for
C21H14ClF9N2O [M + H]+ 517.1, 519.1, found 516.8, 518.8.
(1R,5S,7aS)-1-(3,5-Bis(trifluoromethyl)phenyl)-5-(2-chloro-5-
(trifluoromethyl)phenyl)-2-methyltetrahydro-1H-pyrrolo[1,2-c]-
imidazol-3(2H)-one (Intermediate 8-2). To a solution of
(1R,5S,7aS)-1-(3,5-bis(trifluoromethyl)phenyl)-5-(2-chloro-5-
(trifluoromethyl)phenyl)tetrahydro-1H-pyrrolo[1,2-c]imidazol-
3(2H)-one (120 mg, 0.23 mmol) in DMF (5 mL), stirring at 0 °C was
added NaH (11 mg, 0.28 mmol), and the reaction was stirred for 30
min, after which MeI (0.02 mL, 0.35 mmol) was added and the solution
was stirred at rt for 3 h. Ice was added to the reaction mixture, the layers
separated and the aqueous fraction was extracted was with EtOAc. The
combined organic fractions were dried over anhydrous Na2SO4 and
concentrated to yield (1R,5S,7aS)-1-(3,5-bis(trifluoromethyl)phenyl)-
5-(2-chloro-5-(trifluoromethyl)phenyl)-2-methyltetrahydro-1H-
1
Step 2. To a 250 mL RBF containing a solution of (2S,5S)-1-tert-
butyl 2-ethyl 5-(2-chloro-5-(trifluoromethyl)phenyl)pyrrolidine-1,2-
dicarboxylate (4.00 g, 9.48 mmol) and 1-bromo-3,5-bis-
(trifluoromethyl)benzene (2.5 mL, 14.22 mmol) in diethyl ether (40
mL) was added n-BuLi (1.6 M in hexanes, 8.9 mL, 14.2 mmol)
dropwise at −78 °C. The resulting solution was stirred at that
temperature for 2 h and was allowed to warm to rt and stirred for
another 3 h. The reaction was quenched by addition of saturated
NH4Cl solution (20 mL). It was diluted with diethyl ether (100 mL)
and the layers were separated. The organic layer was rinsed with water
(3 × 200 mL), dried over brine and anhydrous Na2SO4, and
concentrated. The crude product was purified by column chromatog-
raphy eluting with 5% EtOAc in petroleum ether to yield (2S,5S)-tert-
butyl 2-(3,5-bis(trifluoromethyl)benzoyl)-5-(2-chloro-5-
(trifluoromethyl)phenyl)pyrrolidine-1-carboxylate (3.80 g, 6.44
pyrrolo[1,2-c]imidazol-3(2H)-one (120 mg, 0.23 mmol). H NMR
(400 MHz, CDCl3): δ 7.92 (s, 1H), 7.80 (s, 1H), 7.77 (s, 2H), 7.49−
7.48 (m, 2H), 5.29 (t, J = 8.28 Hz, 1H), 5.07 (d, J = 8.80 Hz, 1H), 4.31−
3.29 (m, 1H), 2.85 (s, 3H), 2.84−2.79 (m, 1H), 1.61−1.52 (m, 1H),
1.32−1.25 (m, 2H). LC-MS ESI calcd for C22H16ClF9N2O [M + H]+
531.09, found, 531.0.
(3R,3aS,6S)-3-(3,5-Bis(trifluoromethyl)phenyl)-6-(2-chloro-5-
(trifluoromethyl)phenyl)hexahydropyrrolo[1,2-b][1,2,5]thiadiazole
1,1-Dioxide (Intermediate 8-3). To a solution of (R)-(3,5-bis-
(trifluoromethyl)phenyl)((2S,5S)-5-(2-chloro-5-(trifluoromethyl)-
phenyl)pyrrolidin-2-yl)methanamine (300 mg, 0.61 mmol) in pyridine
(5.0 mL) was added sulfuric diamide (176 mg, 1.81 mmol) at room
temperature. The reaction mixture was heated at 105 °C for 16 h. The
reaction mixture was concentrated. The crude product was purified by
flash column chromatography using 10% ethyl acetate in hexanes to
yield (3R,3aS,6S)-3-(3,5-bis(trifluoromethyl)phenyl)-6-(2-chloro-5-
(trifluoromethyl)phenyl)hexahydropyrrolo[1,2-b][1,2,5]thiadiazole
1,1-dioxide (130 mg, 0.23 mmol) as a white solid. 1H NMR (300 MHz,
CDCl3): δ 7.93 (d, J = 3.54 Hz, 2H), 7.88 (s, 2H), 7.50 (d, J = 1.20 Hz,
2H), 5.44−5.36 (m, 2H), 4.75 (d, J = 4.05 Hz, 1H), 4.68−4.61 (m,
1H), 2.63−2.59 (m, 1H), 1.64 (t, J = 11.91 Hz, 2H), 1.58−1.49 (m,
1H). LC-MS ESI calcd for C20H14ClF9N2O2S [M − H]+ 551.8, found
551.84.
4-(5-(2-((1R,5S,7aS)-1-(3,5-Bis(trifluoromethyl)phenyl)-3-oxohex-
ahydro-1H-pyrrolo[1,2-c]imidazol-5-yl)-4-(trifluoromethyl)phenyl)-
6-methoxypyridin-3-yl)-3-methylbenzoic Acid (Exemplar Com-
pound Related to 48a, 48b, 49). Step 1. To a 2−5 mL microwave
reaction vial was charged Intermediate 8-1 (20 mg, 0.039 mmol) along
with methyl 4-(6-methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-
2-yl)pyridin-3-yl)-3-methylbenzoate (19.46 mg, 0.043 mmol), potas-
sium phosphate (18.07 mg, 0.085 mmol) and XPhos PreCat (1.430 mg,
1.935 μmol). The vial was sealed and degassed and refilled with
nitrogen. Solvent THF (1 mL) and Water (0.1 mL) were then added
and the vial was degassed again. The mixture was then exposed to
microwave irradiation at 100 °C for 30 min. The mixture was then
diluted with ethyl acetate (5 mL), filtered through a syringe filter, and
was washed with ethyl acetate (2 × 1 mL). The filtrate was
concentrated, and the residue was purified by MPLC (4 g silica gel, 0
to 25% ethyl acetate in hexanes, 40CV) to afford product methyl 4-(5-
(2-((1R,5S,7aS)-1-(3,5-bis(trifluoromethyl)phenyl)-3-oxohexahydro-
1H-pyrrolo[1,2-c]imidazol-5-yl)-4-(trifluoromethyl)phenyl)-6-me-
1
mmol) as a colorless, gummy solid. H NMR (400 MHz, CDCl3) δ
8.49 (d, 2H, J = 12 Hz), 8.13 (d, 1H, J = 15 Hz), 7.86−7.91 (m, 1H),
7.48−7.55 (m, 2H), 5.55−5.73 (m, 2H), 2.37−2.59 (m, 2H), 1.90−
1.96 (m, 2H), 1.17 (s, 9H); LC-MS ESI calcd for C25H21ClF9NO3 [M-
Boc]+ 489.88, found 490.0.
Step 3. To a solution of (2S,5S)-tert-butyl 2-(3,5-bis-
(trifluoromethyl)benzoyl)-5-(2-chloro-5-(trifluoromethyl)phenyl)-
pyrrolidine-1-carboxylate (2.00 g, 3.39 mmol) in MeOH (20 mL) were
added NH4OAc (2.61 g, 33.9 mmol) and NaCNBH3 (639 mg, 10.2
mmol). The resulting reaction mixture was heated in a 100 mL pressure
tube at 100 °C for 16 h. After cooling to rt, silica gel was added to the
reaction mixture to make a slurry. It was purified by column
chromatography to yield (2S,5S)-tert-butyl 2-((R)-amino(3,5-bis-
(trifluoromethyl)phenyl)methyl)-5-(2-chloro-5-(trifluoromethyl)-
phenyl)pyrrolidine-1-carboxylate (620 mg, 1.05 mmol). LC-MS ESI
calcd for C25H24ClF9N2O2 [M + H]+ 590.1, 592.1, found 491.0, 493.0.
Step 4. To a solution (2S,5S)-tert-butyl 2-((R)-amino(3,5-bis-
(trifluoromethyl)phenyl)methyl)-5-(2-chloro-5-(trifluoromethyl)-
phenyl)pyrrolidine-1-carboxylate (620 mg, 1.05 mmol) in CH2Cl2 (5
mL) was added a solution of HCl in dioxane (10 mL), and the reaction
was stirred at rt for 4 h. The reaction solution was concentrated, the
residue was partitioned between CH2Cl2 and 10% NaHCO3 aqueous
solution. The aqueous layer was extracted with CH2Cl2, dried over
anhydrous Na2SO4, and concentrated to yield (R)-(3,5-bis-
(trifluoromethyl)phenyl)((2S,5S)-5-(2-chloro-5-(trifluoromethyl)-
phenyl)pyrrolidin-2-yl)methanamine (525 mg, 1.05 mmol). It was
AC
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
thoxypyridin-3-yl)-3-methylbenzoate as a white solid. LC-MS (ESI):
[M + H]+ for C36H28F9N3O4: calcd 738.2 found 738.2.
(500 MHz, CDCl3) δ 7.75 (s, 1H), 7.69 (s, 1H), 7.67 (s, 2H), 7.50 (d, J
= 8.3 Hz, 1H), 7.44 (d, J = 8.3 Hz, 1H), 6.50 (d, J = 11.4 Hz, 1H), 5.88
(m, 1H), 3.50 (s, 3H), 3.15 (s, 3H), 2.40 (m, 1H), 2.30 (m, 2H), 1.85
(m, 1H). [M + H]+ 548, found 547.96.
Step 2. A 20 mL sample vial was charged with methyl 4-(5-(2-
((1R,5S,7aS)-1-(3,5-bis(trifluoromethyl)phenyl)-3-oxohexahydro-1H-
pyrrolo[1,2-c]imidazol-5-yl)-4-(trifluoromethyl)phenyl)-6-methoxy-
pyridin-3-yl)-3-methylbenzoate (28 mg, 0.038 mmol) along with THF
(0.5 mL), MeOH (0.500 mL), and water (0.1 mL), followed by
addition of lithium hydroxide monohydrate (12 mg, 0.286 mmol). The
mixture was then stirred at room temperature for 15 h overnight. The
mixture was neutralized with HCl (1 N, 0.286 mL) and concentrated.
The product was purified by reversed-phase HPLC using a column
supplied by YMC 40-80% acetonitrile in water with 0.5% TFA) to
afford a white solid product 4-(5-(2-((1R,5S,7aS)-1-(3,5-bis-
(trifluoromethyl)phenyl)-3-oxohexahydro-1H-pyrrolo[1,2-c]imidazol-
5-yl)-4-(trifluoromethyl)phenyl)-6-methoxypyridin-3-yl)-3-methyl-
benzoic acid. LC-MS (ESI): [M + H]+ for C35H24F9N3O4: calcd 724.2
found 724.1. RTA (95% HS): 335 nM. Exact mass of 48a, 48b, and 49
prepared by this scheme is provided in the SI.
(Z)-4-(3,5-Bis(trifluoromethyl)phenyl)but-3-en-1-yl trifluorome-
thanesulfonate (Intermediate 9-1). Step 1. To a 250 mL RBF were
added 3-butyn-1-ol (500 mg, 7.13 mmol), 3,5-bis(trifluoromethyl)-
bromobenzene (2.3 g, 7.9 mmol), Cs2CO3 (7.0 g, 21.4 mmol), PdOAc2
(160 mg, 0.71 mmol), 2-dicyclohexylphosphino-2′-(N,N-
dimethylamino)biphenyl (561 mg, 1.43 mmol), and acetonitrile (25
mL). The palladium catalyst and ligand are represented as “Pd” in
Scheme A1. The mixture was stirred at 80 °C for 2 h. It was cooled
down to room temperature, diluted with EtOAc (100 mL), and washed
with water and 10% HCl. The organic layer was concentrated and
purified by normal-phase column chromatography to yield 4-(3,5-
bis(trifluoromethyl)phenyl)but-3-yn-1-ol as a yellow solid.). 1H NMR
(500 MHz, CDCl3) δ 7.87 (s, 2H), 7.8 (s, 1H), 3.89 (m, 2H), 2.76 (m,
2H).
Step 2. To a 250 mL RBF were added 4-(3,5-bis(trifluoromethyl)-
phenyl) but-3-yn-1-ol (4.9 g, 17.4 mmol) and Lindlar catalyst (0.18 g,
0.86 mmol), followed by EtOH (20 mL). The reaction mixture was
stirred under an H2 atmosphere from a balloon at room temperature
overnight. NMR showed when the reaction was complete. The reaction
mixture was filtered through a pad of Celite. The filtrate was
concentrated, and the residue was purified by normal-phase column
chromatography to yield (Z)-4-(3,5-bis(trifluoromethyl)phenyl)but-3-
en-1-ol as a yellow oil. 1H NMR (500 MHz, CDCl3) δ 7.78 (m, 3H),
6.64 (d, J = 11.6 Hz, 1H), 5.70 (m, 1H), 3.82 (m, 2H), 2.60 (m, 2H).
Step 3. To a 100 mL RBF was added (Z)-4-(3,5-bis-
(trifluoromethyl)phenyl) but-3-en-1-ol (4.3 g, 15.1 mmol, dissolved
in 50 mL of DCM). Tf2O (3.1 mL, 18.2 mmol) was added at 0 °C,
followed by pyridine (1.22 mL, 15.1 mmol). TLC and NMR showed
when the reaction was complete. Volatiles were removed. The residue
was dissolved in EtOAc, and the product was washed sequentially with
1 N HCl, then saturated NaHCO3. The organic layer was dried over
Na2SO4 and concentrated to give (Z)-4-(3,5-bis(trifluoromethyl)-
phenyl)but-3-en-1-yl trifluoromethanesulfonate as a brown oil. It was
used directly in the next step. 1H NMR (500 MHz, CDCl3) δ 7.83 (s,
1H), 7.70 (s, 2H), 6.78 (d, J = 11.5 Hz, 1H), 5.89 (m, 1H), 4.63 (m,
2H), 2.80 (m, 2H).
3-(3, 5-Bis(trifluoromethyl)phenyl)-6-(2-chloro-5-
(trifluoromethyl)phenyl)-3a,4,5,6-tetrahydro-3H-cyclopenta[c]-
isoxazole (Intermediate 9-2). Step 1. To a 100 mL RBF were added
nBuLi (5.8 mL, 9.3 mmol) and dry THF. Diisopropylamine (1.39 mL,
9.8 mmol) was added at 0 °C under N2. After stirring for 10 min, the
mixture was cooled to −78 °C. A THF solution of 2-(2-chloro-5-
(trifluoromethyl)phenyl)-N-methoxy-N-methylacetamide (2.5 g, 8.9
mmol) was added. After the reaction mixture was stirred at −78 °C for
20 min, (Z)-4-(3,5-bis(trifluoromethyl)phenyl)but-3-en-1-yl trifluor-
omethanesulfonate (3.33 g, 7.99 mmol) was added. The reaction
mixture was allowed to warm to room temperature and was stirred at
room temperature for 1 h. The reaction mixture was diluted with
EtOAc, washed with saturated NH4Cl, and concentrated. The residue
was purified by normal-phase column chromatography to yield (Z)-6-
(3,5-bis(trifluoromethyl)phenyl)-2-(2-chloro-5-(trifluoromethyl)-
phenyl)-N-methoxy-N-methylhex-5-enamide as a yellow oil. 1H NMR
Step 2. To a 100 mL RBF was added (Z)-6-(3,5-bis-
(trifluoromethyl)phenyl)-2-(2-chloro-5-(trifluoromethyl)phenyl)-N-
methoxy-N-methylhex-5-enamide (1.7 g, 3.10 mmol) in THF. DIBAL
(3.72 mL, 3.72 mmol) was added at 0 °C. The mixture was stirred at
room temperature until NMR showed that there was no starting
material left. It was quenched by addition of 1 N HCl, extracted with
EtOAc, concentrated, and purified by normal-phase column
chromatography to yield (Z)-6-(3,5-bis(trifluoromethyl)phenyl)-2-
1
(2-chloro-5-(trifluoromethyl)phenyl)hex-5-enal as a colorless oil. H
NMR (500 MHz, CDCl3) δ 9.79 (s, 1H), 7.8 (s, 1H), 7.76 (s, 1H), 7.60
(s, 2H), 7.58 (d, J = 8.3 Hz, 1H), 7.47 (d, J = 8.3 Hz, 1H), 6.55 (d, J =
11.7 Hz, 1H), 5.82 (m, 1H), 4.18 (m, 1H), 2.58 (m, 1H), 2.38 (m, 2H),
1.95 (m, 1H).
Step 3. To a 100 mL RBF were added (Z)-6-(3,5-bis-
(trifluoromethyl)phenyl)-2-(2-chloro-5-(trifluoromethyl)phenyl)hex-
5-enal (600 mg, 1.23 mmol), hydroxylamine hydrochloride (341 mg,
4.91 mmol), pyridine (0.60 mL, 7.4 mmol), and DCM. The reaction
mixture was stirred at room temperature for 15 min. NMR and LCMS
showed a clean product. Volatiles were removed. The residue was
purified by normal-phase column chromatography to yield (1E,5Z)-6-
(3,5-bis(trifluoromethyl)phenyl)-2-(2-chloro-5-(trifluoromethyl)-
phenyl)hex-5-enal oxime as a white solid. [M + H]+ calcd: 504, found
503.9.
Step 4. To a 20 mL vial were added (1E,5Z)-6-(3,5-bis-
(trifluoromethyl)phenyl)-2-(2-chloro-5-(trifluoromethyl)phenyl)hex-
5-enal oxime (35 mg, 0.069 mmol) and NCS (11.1 mg, 0.083 mmol,
dissolved in DMF). The reaction mixture was stirred at room
temperature. NMR indicated that the starting material was consumed.
TEA (0.25 mL, 1.78 mmol) in DCM was added dropwise. The resulting
reaction mixture was stirred at room temperature for 10 min. LCMS
showed the desired clean product. Volatiles were removed. Residue was
purified by prep-TLC to yield Intermediate 9-2 as a racemic mixture.
1H NMR (500 MHz, CDCl3) δ 7.88 (s, 1H), 7.78 (s, 2H), 7.52 (m,
3H), 5.98 (d, J = 9.9 Hz, 1H), 4.54 (m, 1H), 4.26 (m, 1H), 2.90 (m,
1H), 2.14 (m, 1H), 1.93 (m, 1H), 1.00(m, 1H). [M + H]+ calcd: 501,
found 501.8.
4-(5-(2-((3S,3aR,6S)-3-(3,5-Bis(trifluoromethyl)phenyl)-3a,4,5,6-
tetrahydro-3H-cyclopenta[c]isoxazol-6-yl)-4-(trifluoromethyl)-
phenyl)-6-methoxypyridin-3-yl)-3-methylbenzoic acid (51a), and 4-
(5-(2-((3R,3aS,6R)-3-(3,5-bis(trifluoromethyl)phenyl)-3a,4,5,6-tetra-
hydro-3H-cyclopenta[c]isoxazol-6-yl)-4-(trifluoromethyl)phenyl)-6-
methoxypyridin-3-yl)-3-methylbenzoic Acid (51b) (Exemplar Com-
pounds Related to 51). Step 1. To a 5 mL RBF were added 3-(3,5-
bis(trifluoromethyl)phenyl)-6-(2-chloro-5-(trifluoromethyl)phenyl)-
3a,4,5,6-tetrahydro-3H-cyclopenta[c]isoxazole (Intermediate 10-1)
(50 mg, 0.1 mmol), tert-butyl 4-(6-methoxy-5-(4,4,5,5-tetramethyl-
1,3,2-dioxaborolan-2-yl)pyridin-3-yl)-3-methylbenzoate (51 mg, 0.12
mmol), potassium phosphate (42.3 mg, 0.199 mmol), XPhos
precatalyst (7.84 mg, 9.96 μmol), 1,4-dioxane (2 mL) and water (0.2
mL). The mixture was stirred at 90 °C for 2 h. After cooling down,
volatiles were removed. The residue was purified by silica gel column
chromatography to yield the target compound as a racemic mixture. [M
− H]+ 763.2, found 763.3.
Step 2. To a 10 mL scintillation vial were added tert-butyl 4-(5-(2-
(3-(3,5-bis(trifluoromethyl)phenyl)-3a,4,5,6-tetrahydro-3H-
cyclopenta[c]isoxazol-6-yl)-4-(trifluoromethyl)phenyl)-6-methoxy-
pyridin-3-yl)-3-methylbenzoate (20 mg, 0.026 mmol) and DCM (0.5
mL), followed by TFA (0.81 mL, 10.5 mmol). After the reaction
mixture was stirred at room temperature for 10 min, LCMS showed all
of the starting material has converted. Volatiles were removed. Residue
was purified by reversed-phase HPLC to yield 4-(5-(2-((3S,3aR,6S)-3-
(3,5-bis(trifluoromethyl)phenyl)-3a,4,5,6-tetrahydro-3H-cyclopenta-
[c]isoxazol-6-yl)-4-(trifluoromethyl)phenyl)-6-methoxypyridin-3-yl)-
3-methylbenzoic as a racemic mixture. The racemic mixture was
separated by chiral SFC on Chiral Cel IA (4.6 mm × 250 mm) with 30%
2:1 MeOH:MeCN/CO2 (2.1 mL/min, 100 bar, 35 °C) as the eluent to
AD
https://doi.org/10.1021/acs.jmedchem.1c00959
J. Med. Chem. XXXX, XXX, XXX−XXX