Synthesis and Evaluation of PPARδAgonists That Promote Osteogenesis in a Human Mesenchymal Stem Cell Culture and in a Mouse Model of Human Osteoporosis

pubs.acs.org/jmc

Article

Synthesis and Evaluation of PPARδ Agonists That Promote

Osteogenesis in a Human Mesenchymal Stem Cell Culture and in a

Mouse Model of Human Osteoporosis

Brian J. Kress, Dong Hyun Kim,* Jared R. Mayo, Je ﬀ ery T. Farris, Benjamin Heck, Je ﬀ rey G. Sarver,

Divya Andy, Jill A. Trendel, Bruce E. Heck,* and Paul W. Erhardt *

Cite This: J. Med. Chem. 2021, 64, 6996 −7032

Read Online

sı

*

Metrics & More

Article Recommendations

Supporting Information

ACCESS

ABSTRACT: We synthesized a directed library of compounds to explore

the structure−activity relationships of peroxisome proliferator-activated

receptor δ (PPARδ) activation relative to mesenchymal stem cell (MSC)

osteogenesis. Our scaﬀold used para-substituted cinnamic acids as a polar

headgroup, a heteroatom and heterocycle core connecting units, and

substituted phenyl groups for the lipophilic tail. Compounds were

screened for their ability to increase osteogenesis in MSCs, and the most

promising were examined for subunit speciﬁcity using a quantitative

PPAR transactivation assay. Six compounds were selected for in vivo

studies in an ovariectomized mouse model of human postmenopausal

osteoporosis. Four compounds improved bone density in vivo, with two

(12d and 31a) having activity comparable to that of GW0742, a well-

studied PPARδ-selective agonist. 31a (2-methyl-4-[N-methyl-N-[5-

methylene-4-methyl-2-[4-(triﬂuoromethyl)phenyl]thiazole]]aminocinnamic acid) had the highest selectivity for PPARδ compared

to other subtypes, its selectivity far exceeding that of GW0742. Our results conﬁrm that PPARδ is a new drug target for possible

treatment of osteoporosis via in situ manipulation of MSCs.

fracture,⁷they carry possible long-term side eﬀects including

INTRODUCTION

■

an increased risk of atypical femoral fracture and osteonecrosis

The human skeleton continually renews itself in response to

of the jaw.^8,9While disputed in the literature, these links to

bisphosphonates are thought to be the result of an

accumulation of old bone due to the decreased rate of bone

turnover.⁸

the physical demands of everyday life.¹This renewal process

involves the resorption of old bone by osteoclasts followed by

the formation of new bone by osteoblast proliferation. The

replacement of old bone with new bone is important for

maintaining structural integrity and reducing the risk of

fracture.²In healthy adults, osteoclast function is balanced by

osteoblast activity, resulting in bone homeostasis indicated by

unchanging bone mineral density (BMD).³The bone

remodeling process is controlled by several signaling pathways

and hormones including sex hormones. A reduction in sex

hormones, such as estrogen in postmenopausal women, results

in overactivation of the osteoclasts and the reduction of BMD,

which can lead to the development of osteoporosis.^4,5

Other antiresorptive therapies include hormone replacement

therapy (HRT), which functions in a way that mimics the

eﬀects of estrogen. HRTs include estrogen-only formulations

or estrogen−progesterone formulations and are eﬀective at

preventing bone loss by increasing osteoprotegerin, a

glycoprotein involved in reducing osteoclast activation.^9,10

However, there is evidence that these therapies increase the

risk of blood clots, stroke, and heart attack. The estrogen−

progesterone combination may also increase the risk of

developing breast cancer.^10,11

Treatment options for osteoporosis are categorized

pharmacologically as either antiresorptive or anabolic. This

classiﬁcation depends on whether the pharmaceutical acts

through inhibition of the osteoclast or stimulation of osteoblast

proliferation, respectively. Clinically, the most prominent

front-line therapy involves the inhibition of bone resorption

using the bisphosphonate class of drugs. These drugs act

through encouraging apoptosis of osteoclasts, therefore

decreasing bone turnover and resorption.⁶Although bi-

sphosphonates have been shown to be eﬀective at preventing

Received: March 28, 2021

Published: May 14, 2021

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

6996

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 1. Chemical structures of classical PPAR agonists having selectivity for δ (cardarine and GW0742), α (gemﬁbrozil), and γ (rosiglitazone)

subtypes. Compounds are depicted with their similar acidic “head groups” located on their western edges. In addition to having an acidic head, the

identical connecting portion “core” centrally located in cardarine and GW0742, along with the similar lipophilic “tail” groups located on their

eastern edges, inﬂuenced the design of our initial PPARδ probes.

Osteoclasts play an important role in bone homeostasis by

removing older necrotic bone tissue. Decreasing osteoclast

function with antiresorptive agents can therefore be problem-

atic over time due to an accumulation of old bone.¹²⁻¹⁴

Alternatively, anabolic agents avoid this risk by encouraging

bone formation from an increase in osteoblast proliferation and

function.¹⁵Current anabolic therapies include teriparatide,

abaloparatide, and romosozumab. Romosozumab, the most

recently approved anabolic agent, has a mechanism of action

involving inhibition of sclerostin, a Wnt-β-catenin signaling

inhibitor.^16,17This suggests that increasing Wnt signaling can

have positive eﬀects toward increasing bone density and

treating osteoporosis. This recent ﬁnding provides clinical

relevance for our preliminary research studying peroxisome

proliferator-activated receptor δ (PPARδ) in stem-cell cultures

because others have shown that the activation of PPARδ plays

a central role in bone remodeling though directing an increase

in Wnt signaling.¹⁸Deploying small-molecule agents to

potentially manipulate stem cells in situ for therapeutic

applications was highlighted within the ﬁeld of medicinal

chemistry about 10 years ago.¹⁹Progress in devising these

types of medicinal chemistry strategies has continued,²⁰and

recently it was noted that for stem-cell technologies in general,

“the only constant is change”, while this promising ﬁeld rapidly

evolves in many directions.²¹

PPARs are a family of ligand-activated nuclear receptors that

play a role in regulating cell diﬀerentiation and development in

addition to lipid and glucose metabolism.²²This family

includes three subtypes, PPARα, PPARγ and PPARδ. In

regard to mesenchymal stem cell (MSC) diﬀerentiation, the

activation of PPARγ has been shown to increase adipogenesis

at the expense of osteogenesis,²³whereas the activation of

PPARδ increases osteogenesis and bone formation.^18,24The

negative aspects of PPARγ activation can be seen in the clinic

when considering the long-term side eﬀects of the

thiazolidinedione (TZD) class of drugs. The TZDs are

PPARγ agonists and have been associated with an increased

risk of fracture and an increased risk of developing

osteoporosis.²⁵

rodents demonstrated carcinogenicity at the 2 year junc-

ture.^27,28Despite warnings and its prohibition by the World

Anti-Doping Agency, cardarine continues to be abused by

athletes.²⁹Notoriously known as “endurabol,” it is thought to

enhance physical performance.³⁰There has been considerable

research directed toward understanding the mechanisms for

carcinogenicity associated with certain PPARα and γ agents,

but “PPARδ agonists have not been extensively investigated

and have produced varying and conﬂicting results”.³¹Thus the

“biological function of the PPARδ”³²pathway when activated

with agents like cardarine that could be directly or indirectly

initiating and/or promoting³¹rodent “tumorigenesis remains

complex, conﬂicting”³²and, as of today, is still undeﬁned.

Furthermore, the connection of PPARδ-related in vitro and in

vivo models to human cancer has been questioned³¹⁻³³and,

instead, it has been noted that these particular “adverse eﬀects

may be linked to signiﬁcant oﬀ-target” interactions of the

ligands when present at “very high doses”.³³Speciﬁc studies

into the role of PPAR ligands as potential anticancer agents

have likewise produced contradictory results, and there is no

scientiﬁc consensus on whether this pathway’s activation or

inhibition can be used to eﬀectively promote or prevent human

disease.^32,33For the medicinal chemist, this situation is

accompanied by a paucity of systematically delineated

toxicity-associated structure−activity relationship (SAR) data.

At best, it can be noted “that synthetic ligands that target the

same PPAR subtype may not always possess comparable

eﬃcacy, safety proﬁles, and clinical outcomes”.³²From this

backdrop, the goals of our early-stage drug discovery program

became four-fold: (i) Assess if small-molecule drug-induced

activation of the PPARδ receptor in stem-cell cultures is a

viable mechanistic target to increase bone density; (ii)

Establish preliminary eﬃcacy-related SAR associated with

this novel indication wherein certain molecular features in

our probe compounds have been systematically altered from

those present in cardarine; (iii) Determine if eﬃcacy can be

maintained in situ by using an animal model of human

osteoporosis to study a series of representative compounds;

and, (iv) Identify a promising group of candidate compounds

that can also serve as probes for long-term toxicity SAR. In line

with the guidelines initially stipulated by the U.S. FDA,³⁴it is

our view that long-term toxicity studies will need to be initiated

in the future as an early step toward the preclinical

development of any PPARδ-derived small-molecule drug for

the potential prolonged treatment of any chronic disease.

Currently, there are no selective PPARδ agonists approved

for use in the marketplace. This is due in part to lingering

concerns previously raised by cardarine (Figure 1), a selective

and potent PPARδ agonist²⁶that was entered into clinical

development to treat metabolic diseases. Cardarine’s Phase II

trials were halted in 2007 when long-term toxicity studies in

6997

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 2. Metabolism considerations used during the design of PPARδ probes. (A) Generalized composite of metabolic possibilities for the typical

three-ring scaﬀold. When β-oxidation is prevented, metabolic switching to aromatic hydroxylation followed by glucuronidation and acyl

glucuronidation becomes the most probable biotransformation pathway. (B) Less probable and thus less predominant pathways should they

happen to occur. As shown, however, these pathways could potentially generate reactive metabolites that might be responsible for oﬀ-target toxicity

upon long-term dosing. (C) Stability present in cinnamic acid that circumvents the reactivity of its potential metabolites analogous to those in panel

B. (D) Typical Michael reaction that can sometimes cause oﬀ-target toxicity by this alternate pathway. Note that cinnamic acid resides at the lower

end of reactivity in terms of both R¹and R²adducts. See the additional discussion in text for all panels.

During consideration of our manuscript’s submission, two

extensive review articles were published. The ﬁrst by Takada

and Makishima surveyed PPAR agonist and antagonist ligand

patents (2014−2020).³⁵In addition to the long-standing

possibility of treating metabolic syndromes like dyslipidemia

and type 2 diabetes, many synthetic ligands are being

developed to potentially treat new indications like mitochon-

drial disease, inﬂammatory/autoimmune disease, cancer and

neurological disorders. These authors also oﬀer a brief

discussion about “bone disease” while referring to the earliest³⁶

of our four issued patents in this arena.³⁶⁻³⁹The second paper

by Kadayat et al. provides an excellent overview and

perspective for the medicinal chemistry associated with

PPARδ agonists all the way from drug design and fundamental

SAR, to several drug candidates now undergoing clinical study

for a variety of diseases.⁴⁰They likewise mention our patent for

bone disease relative to new indications.³⁶These reviews will

be discussed further in the section entitled Translational

Consideration of Toxicity Relative to Use in Humans.. At this

point, however, it is interesting to note that both reports clearly

demonstrate that the PPARδ ﬁeld has ﬁnally begun to again

push forward, this time with many new compounds for several

new indications. Furthermore, the summary oﬀered by Kadayat

et al. refers to the ambiguous nature of the historical toxicity,

exactly as we do in terms of the PPARδ pathway’s questionable

involvement and the lack of chemical structural dependency

when relying upon only a single compound’s anecdotal data,⁴¹

namely, what was observed for cardarine several years ago.

In addition to the previously noted increase in Wnt signaling

and bone formation reported by others while using cardarine,¹⁸

members of our group simultaneously reported that GW0742

can attenuate the negative eﬀect that nicotine has on MSC

osteogensis.²⁴GW0742 (Figure 1) is another potent and

selective PPARδ agonist^26,42that has nearly the same structure

6998

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

as cardarine. The three-ring scaﬀold displayed by both of these

structures is signiﬁcantly diﬀerent from that present in either a

classical PPARα or PPARγ agonist (respectively represented in

Figure 1 by gemﬁbrozil⁴³and rosiglitazone⁴⁴). This distinct

scaﬀold has been referred to as having an acidic “head”, a

connecting “core”, and a lipophilic “tail”.^40,45We have used it

as the starting point for the design of our PPARδ-agonist-

directed library. To distinguish the library from cardarine, we

immediately considered replacing the sulfur-connecting atom

with an oxygen or nitrogen atom, both of which have

precedent for retaining eﬃcacy.^26,46−49

all three types of X, we next considered replacing the entire

phenoxyacetic acid with a suitable mimic that can further help

to avoid such pathways. A cinnamic acid moiety has precedent

for aﬀording PPARδ agonist eﬃcacy when used as a

headgroup.^26,45,66,67Of note to us was the display of a

carboxylic acid similarly located two atoms away from a phenyl

ring, wherein its double bond is rich in electron density,

somewhat like the lone electron pairs’ density present on the O

atom involved in the analogous two-atom link within the

phenoxyacetic acid group. Previous modeling studies suggest

that either the E or Z isomers can be deployed because of their

ability to isomerize to the more active form at the PPARδ

binding site, presently thought to be the Z or cis isomer.²⁶In

terms of metabolism, cinnamic acid is similar to phenoxyacetic

acid in that it is not as prone to β-oxidation degradation when

compared with simple alkanoic acids. Opposite to phenoxy-

acetic acid, cinnamic acid withdraws electrons from its aryl

ring, making it less sensitive to aromatic hydroxylation. Finally,

the hypothetical series of conversions that might lead to the

collapse of the phenoxyacetic acid headgroup produces a less

fragile species in the case of cinnamic acid because the latter is

stabilized by direct conjugation with the carboxylic acid moiety

(Figure 2C).

Among these potential beneﬁts, however, we also noted that

as an α,β-unsaturated carbonyl system, our switch to a

cinnamic acid headgroup could itself lead to Michael addition

reaction⁶⁸oﬀ-target toxicity⁶⁹(Figure 2D). Fortunately, when

such carbonyls are part of a carboxylic acid group, Michael

reactivity is signiﬁcantly diminished compared with aldehydes

and ketones, especially so when deployed within biological

systems⁶⁹⁻⁷³where strong bases are not present to activate the

initial step for a donor partner.^68,69Indeed, cinnamic acid and

its derivatives are used extensively in the food and cosmetic

industry without issues stemming from Michael addition

reactions.⁷⁴⁻⁷⁶Nevertheless, to ensure that we were not

swapping out the hypothesized toxic head region in cardarine

for yet another possibly toxic and potentially problematic

pathway, we performed experiments to conﬁrm the safety of

deploying cinnamic acid in our speciﬁc molecular context.

These studies are described in the Preliminary Toxicity

Assessments section. In the end, our ﬁnal library design used

trans-cinnamic acid as a constant headgroup while several

other structural features of the overall scaﬀold were system-

atically examined. Figure 3 summarizes the compounds that

were synthesized to explore the PPARδ agonist eﬃcacy SAR as

it relates to osteogenesis. This series additionally represents a

Part of our directed library compound design was also based

on metabolic and pharmacokinetic (PK) considerations.

Cardarine is reported to have a half life of ca. 10 h in rats

and nearly double that in humans,⁵⁰but its detailed disposition

has not been fully reported, in particular, with regard to the

possibility that it could produce low levels of localized toxic

metabolites that become insidious upon long-term dosing. A

composite of metabolic possibilities for PPARδ ligands is

summarized in Figure 2A.^26,45By analogy to other agents,⁴⁵

cardarine’s aryl-thioether, head-core-linkage region resides in a

metabolic soft spot^51,52such that the aryl moiety may be

susceptible to CYP-450-mediated aromatic hydroxylation,

whereas the sulfur may be subject to CYP-450-mediated S-

dealkylation and to CYP-450 or ﬂavin-containing monoox-

ygenase (FMO)-mediated S-oxidation.⁵³The latter can

complicate PK proﬁling because the S atom becomes

asymmetric, whereas the dealkylated metabolite could

contribute to toxicity because of the resulting aryl-sulfhydryl

group’s inherent reactivity.⁵⁴The dealkylated metabolite also

sets up the possibility for it to form a sulfur version of the even

more reactive para-quinone-type chemical species⁵⁵upon the

loss of acetic acid (Figure 2B, upper arrows) via a pathway that

could initially involve the transient formation of the acid’s

tautomer.⁵⁶Initial rescue from such reactive intermediates in

mammals by glutathione can be compromised by acute

overdose or by continual exposure in sensitive tissues, the

latter potentially making a relevant association with cardarine’s

carcinogenicity being observed only after long-term dos-

ing.^57,58

Although the steps shown subsequent to metabolic deal-

kylation might be considered to occur spontaneously,⁵⁹⁻⁶¹we

have not found any evidence to support this toxic series of

conversions in the case of PPAR ligands like cardarine or even

within the chemical literature associated with any of the

simpler 2-(4-X-phenoxy)acetic acid compounds. Thus if

quinone-like materials are formed at some low level, then

this is likely prompted by some additional biochemical

interaction if not by another complete biotransformation step

such as the previously mentioned possibility for aromatic

hydroxylation leading to catechol-like arrangements wherein

redox chemistry becomes applicable^59,62,63(Figure 2B, lower

arrow). One can imagine that the metabolite resulting from a

common CYP-450-mediated X-dealkylation for any of the

three heteroatom substitutions could also follow the proposed

toxic pathways shown in Figure 2B. In this regard, whereas less

data is available for S-dealkylations, this initial step for the

constructs having X = O and N is known to be highly sensitive

to steric inhibition such that this type of metabolic-driven

toxicity compared with X = S is likely not highly probable

when simultaneously considered against competing metabolic

possibilities.^64,65Nevertheless, because the proposed culprit

quinone-related pathways remain as at least a possibility across

Figure 3. General scaﬀold design outlining the structural

modiﬁcations explored herein.

6999

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Scheme 1. Synthesis of Oxygen Linker Series

a

Reagents and conditions: (a) ethyl 2-chloroacetoacetate, 95% EtOH, reﬂux; (b) LAH, THF, 0 °C; (c) MeSO₂Cl, TEA, DCM, 0 °C; (d) H₂SO₄,

MeOH, reﬂux; (e) 4, Cs₂CO₃, MeCN, rt; (f) NaOH, 95% EtOH, rt.

a

Scheme 2. Synthesis of Sulfur Linker Series

a

Reagents and conditions: (a) ethylene glycol, pTSA, toluene, Dean−Stark reﬂux; (b) t-BuLi, THF, −78 °C; (c) sulfur, THF, −78 °C to rt; (d) 4,

THF, 0 °C; (e) HCl, THF, rt; (f) methyl (triphenylphosporanylidene)acetate, THF, 40 °C; (g) NaOH, 95% EtOH, rt.

family of compounds from which to select probes that can

contribute toward delineating long-term toxicity SAR as may

be useful to us or others in the future.

agonist (Figure 1). These results (Tables 1 and 2) led to the

selection of seven hit compounds (7b,d, 12b,d, 16b, 23b, and

31achemical structures shown on Table 3) that exhibited

varying osteogenic proﬁles while allowing for representative

structural comparisons in the core and lipophilic tail

components. They were tested in a luciferase reporter assay

for each of the three PPAR subunits. This assay generated

dose−response curves and statistically derived EC₅₀values for

quantitative assessments of PPAR selectivity as well as potency

(Table 3). Six of these hit compounds (7d, 12b,d, 16b, 23b,

and 31a) were additionally tested in ovariectomized (OVX)

mice for which the in vivo results (Figures 5 and 6) suggest that

four compounds (7d, 12d, 23b, and 31a) have potential for

treating osteoporosis in mammals. Taken together, the

composite of testing results indicates that two compounds

(12d and 31a) can be considered as the preferred lead agents.

Of these, 31a (2-methyl-4-[N-methyl-N-[5-methylene-4-meth-

yl-2-[4-(triﬂuoromethyl)phenyl]thiazole]]aminocinnamic

Our synthesized library reevaluated some key literature SAR

conclusions, such as the role of sulfur as a linker heteroatom

compared with oxygen and nitrogen, along with an assessment

of more novel structural changes such as trying two diﬀerent

heterocycles in the core region (Figure 3). The ﬁrst series of

synthesized compounds included combinations of diﬀerent

linker heteroatoms with various triﬂuoromethyl and ﬂuorine

substitutions on the terminal phenyl-ring tail. The SAR

conclusions from this series (Table 1) inﬂuenced the structure

design for a second series of compounds while holding the

initially optimized features constant (Table 2). An MSC

diﬀerentiation assay giving comparative values of osteogenesis

or adipogenesis was used as an initial screen to determine

which compounds were the strongest inducers of osteogenesis.

Adipogenesis was prompted for gross comparisons of PPAR

selectivity by administering rosiglitazone, a standard PPARγ

7000

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Scheme 3. Synthesis of N−H Linker Series and Compound 17

a

Reagents and conditions: (a) BOC₂O, THF, 35 °C; (b) 4, NaH, DMF, 0 °C to rt; (c) TFA, DCM, rt; (d) NaOH, 95% EtOH, rt; (e) NaOH, 95%

EtOH, rt.

acid) exhibits extremely high selectivity for PPARδ, much

greater than that of the prototypical standard GW0742.

potential ester prodrug compounds 6 and base hydrolysis to

give the carboxylic acid target compounds 7a−d.

The series of target compounds with an X = sulfur were

synthesized from the commercially available 4-bromobenzal-

dehyde by conversion of the aldehyde using ethylene glycol to

give acetal 8 (Scheme 2). The next step involved a one-pot

reaction⁷⁸using t-BuLi for lithium−halogen exchange of the

bromine followed by the formation of the thiolate using sulfur

powder and lastly, coupling to the alkyl chloride intermediate 4

to give intermediate 9. Despite the multiple reactions occurring

in this one-pot conversion, the overall yields remained

favorable, and no attempts at isolated stepwise improvements

were deemed necessary. The acetal intermediates 9 were

deprotected using HCl to give aldehydes 10 followed by a

Wittig reaction using methyl (triphenylphosphoranylidene)-

acetate to give E-oleﬁns 11. The subsequent base hydrolysis of

esters 11 gave the carboxylic acid targets 12a−d.

The synthesis of target compounds containing an X-position

nitrogen linker began with tert-butyloxycarbonyl (BOC)

protection of the commercially available ethyl 4-amino-

cinnamate to give 13 (Scheme 3). BOC-protected amine 13

was then coupled to alkyl chlorides 4 to give intermediates 14

followed by the removal of the BOC group using triﬂuoro-

acetic acid (TFA) to give the secondary amine prodrugs 15

and base hydrolysis to give target compounds 16a−d

containing the NH linker atom. In one instance, base

hydrolysis was completed without the removal of the BOC

group to give an unanticipated test compound 17 that then

contained an N-BOC linker moiety from the 14d intermediate.

The synthesis of 16a−d was originally attempted without

BOC protection of 4-aminocinnamate by coupling directly to

alkyl chloride 4. This was found to be problematic due to the

formation of a disubstituted side product 18 despite using only

1 equiv of 4b and the weak base CsCO₃. The intermediate 18

RESULTS AND DISCUSSION

■

Chemistry. The ﬁrst series of analogs investigated a SAR

for the linker heteroatom located para- to the E-oleﬁn (X-

position, Figure 3) along with triﬂuoromethyl and ﬂuorine

substitutions on the terminal phenyl ring (R¹and R², Figure

3). The X-position heteroatom had previously been explored

in the literature as either an oxygen or sulfur for a similar

scaﬀold in the development of GW0742.²⁶We decided to

repeat this comparison for our scaﬀold in addition to a

comparison with a nitrogen linker in the form of either N−H

or N-methyl. Our intended use of PPARδ agonists as agents for

controlling MSC diﬀerentiation was also unexplored in the

literature, and we elected to repeat certain SAR conclusions

from other studies to determine whether they remained similar

for our new indication.

The synthesis of the alkyl chloride intermediates 4a−d

(Scheme 1) was completed as previously described²⁶and

involved the formation of ester 2 using a Hantzsch thiazole

synthesis from commercially available thiobenzamides and

ethyl 2-chloroacetoacetate. The resulting ester was reduced

using lithium aluminum hydride (LAH) to give alcohol 3. The

alcohol was chlorinated using methanesulfonyl chloride to give

the desired “eastern-half” intermediates 4a−d as alkyl

chlorides. These key alkyl chloride intermediates were used

in various coupling reactions with the appropriate linker

heteroatom. Target compounds 7a−d containing the X-

position oxygen linker heteroatom were synthesized from

commercially available p-coumaric acid by Fischer esteriﬁca-

tion to give methyl 4-hydroxycinnamate 5⁷⁷followed by

coupling to intermediates 4 using cesium carbonate to give

7001

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Scheme 4. Synthesis of Disubstituted Test Compound 19

a

Reagents and conditions: (a) 4b, NaH, NaI, MeCN, rt; (b) NaOH, 95% EtOH, rt.

a

Scheme 5. Synthesis of N-Me Linker Series and other N-Alkyl Targets

a

Reagents and conditions: (a) R₆-I, NaH, DMF, 0 °C to rt; (b) TFA, 0 °C to rt; (c) 4 (a, b, c, or d), NaI, NaH, DMF, 0 °C to rt; (d) NaOH, 95%

EtOH, rt.

a

Scheme 6. Synthesis of Targets with R³or R⁴Methyl Substitution

a

Reagents and conditions: (a) DIBAL-H, toluene, −78 °C; (b) MeNBOC 26, Pd₂(dba)₃, Cs₂CO₃, 1,4-dioxane; (c) methyl-

(triphenylphosphoranylidene)acetate, THF, 55 °C; (d) TFA, DCM, 0 °C to rt; (e) 4b, NaH or DIPEA, DMF, 0 °C to rt; (f) 3 N NaOH,

95% EtOH, THF, rt.

was later intentionally synthesized using 2 equiv of 4b and

NaH as the base (Scheme 4). This led to the synthesis of

another initially unanticipated test compound 19 after base

hydrolysis.

A similar synthetic strategy involving BOC protection was

utilized in the synthesis of the N-Me linker series, as well as the

later N-alkyl series, to avoid any possible dialkylated side-

product formation. BOC-protected cinnamate 13 was alkylated

using the appropriate alkyl iodide to give series 20 (Scheme 5).

The BOC group was removed using TFA to give intermediates

21 followed by a nucleophilic substitution involving alkyl

chlorides 4 to give ethyl ester 22. Lastly, base hydrolysis of the

esters provided targets 23a−g.

7002

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Scheme 7. Synthesis of R³-Substituted Ethyl and Isopropyl Targets

a

Reagents and conditions: (a) di-tert-butyl dicarbonate, THF, reﬂux; (b) NaH, MeI, DMF, reﬂux; (c) TFA, DCM, 0 °C to rt; (d) 4b, NaH, NaI,

DCM, 0 °C to rt; (e) POCl₃, DMF, 0 °C to rt; (f) H₂O (ice); (g) triethyl phosponoacetate, NaH, THF, 0 °C to rt; (h) 3 N NaOH, 95% EtOH,

THF, rt.

a

Scheme 8. Synthesis of the Thiazole Regioisomer Target

a

Reagents and conditions: (a) Br₂, DCM, rt; (b) 1b, 95% EtOH, reﬂux; (c) LAH, THF, 0 °C to rt; (d) MeSO₂Cl, TEA, DCM, 4 °C; (e) 21a, NaI,

NaH, DMF, 0 °C to rt; (f) NaOH, 95% EtOH, rt.

The second major series of target compounds had structural

features shaped by the SAR results from studying the

aforementioned probes in the MSC diﬀerentiation assay and

selected compounds in the PPAR transactivation assay.

Prominent structural features of interest included the retention

of the terminal phenyl para-triﬂuoromethyl and N-methyl

linker atom (R¹= CF₃, X = N-Me). The length of the alkyl

chain located at the nitrogen X-position was also explored and

was synthesized as previously described (Scheme 5). The other

targets synthesized as part of the second series of targets

contained various alkylations at the western phenyl headgroup

(R⁴and R⁵positions) as well as other targets containing a

thiazole regioisomer or triazole heterocycle in the central core

region.

amine (DIPEA) as base to give 30a,b. Lastly, base hydrolysis of

the esters gave the target compounds 31a,b.

The probes containing either an ethyl or isopropyl

substitution at the R³position were synthesized using a

diﬀerent scheme due to the unavailability of the analogous

commercially available starting materials. These targets were

synthesized from 3-ethylaniline 32a and 3-isopropylaniline 32b

(Scheme 7). The aniline nitrogen was protected using di-tert-

butyl dicarbonate to give the BOC-protected amine

intermediates 33 followed by methylation using methyl iodide

to give carbamates 34 and deprotection of the BOC group

using TFA to give the secondary amines 35. We were unable to

obtain pure secondary amines 35; however, we were able to

eﬀectively use these crude intermediates directly in a coupling

reaction with alkyl chloride 4b to give compounds 36 as pure

materials. An aryl aldehyde was formed para to the tertiary

amine using a Vilsmeier−Haack reaction to give aldehydes 37.

This aldehyde was initially subjected to a Wittig oleﬁnation

similar to that in Scheme 6. These attempts resulted in low

yields, presumably due to the steric hindrance from the larger

R³alkyl groups. Alternatively, a Horner−Wadsworth−

Emmons oleﬁnation using triethyl phosphonoacetate and

NaH was able to give the desired E-oleﬁns 38 in acceptable

yields. Lastly, base hydrolysis of the esters aﬀorded the target

compounds 39a,b.

Target compounds containing methyl substitutions at either

the R³or R⁴position were synthesized from the appropriate

commercially available 4-bromobenzonitrile 24a or 24b

(Scheme 6). Stephen aldehyde synthesis of bromides 24

using diisobutylaluminum hydride aﬀorded aldehydes 25.⁷⁹

The BOC-protected methylamine 26 was used in a Buchwald−

Hartwig amination of aldehydes 25 to give carbamates 27.⁷⁹

A

Wittig oleﬁnation of aldehydes 27 was used to exclusively form

the E-oleﬁns 28. The BOC group was then removed using

TFA to give secondary amines 29 followed by a nucleophilic

substitution of 4b using either NaH or N,N-diisopropylethyl-

7003

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Scheme 9. Synthesis of Triazole Targets

a

Reagents and conditions: (a) AcOH, H₂O, HCl, NaNO₂, 0 °C; (b) ethyl acetoacetate, 95% EtOH, NaOAc, Na₂CO₃, 0 °C; (c) CuCl₂, NH₄OAc,

95% EtOH, reﬂux; (d) 2 M LiAlH₄, THF, 0 °C to rt; (e) MeSO₂Cl, Et₃N, DCM; (f) 21a, NaI, NaH, DMF, 0 °C to rt; (g) 3 N NaOH, 95% EtOH,

THF, rt; (h) 13, NaI, NaH, DMF, 0 °C to rt; (i) TFA, DCM, 0 °C to rt; (j) NaOH, 95% EtOH, THF, rt

Figure 4. Treatment of MSCs with the PPARδ agonist, GW0742, and the PPARγ agonist, rosiglitazone, results in osteogenesis and adipogenesis,

respectively. Typical results showing a robust osteogenic response (+++) to 1 μM GW0742 when visualized using Alizarin Red stain. Alternatively,

the moderate (++) to robust (+++) adipogenic responses to 1 and 10 μM rosiglitazone, respectively, are shown by the formation of white lipid

droplets.

A regioisomer of the thiazole was also synthesized based on

a literature example that suggested that this change within a

diﬀerent overall scaﬀold can result in an increase in the PPARδ

potency and selectivity.⁴⁶Synthesis of the thiazole regioisomer

was completed, as described in the literature,⁴⁶using bromine

and 4-(triﬂuoromethyl)thiobenzamide 1b to give intermediate

40 (Scheme 8). The subsequent steps were completed

analogously to those used in the synthesis of the N-methyl

series. The ester 40 was reduced to the primary alcohol 41

using LAH. The alcohol 41 was chlorinated using meth-

anesulfonyl chloride to give the alkyl chloride 42 followed by a

coupling reaction using NaH with 21a to give 43. Lastly, ester

43 was hydrolyzed to give the carboxylic acid target 44.

Two targets having a 1,2,3-triazole heterocycle as part of the

core were synthesized by the formation of the triazole ring in

the ﬁrst step using a one-pot reaction (Scheme 9).⁸⁰This

reaction utilized in situ nitrous acid to form a diazonium salt

from the 4-(triﬂuoromethyl)aniline starting material, the

addition of ethyl acetoacetate, and triazole formation using

CuCl₂and excess ammonium acetate to give 45. The targets,

containing either N−H or N-methyl at the X-position (49 and

52), were synthesized using methods analogous to the

previously shown schemes for 16 and 23, respectively.

In Vitro Testing. All target compounds were screened for

their ability to promote the osteogenic diﬀerentiation of

human-bone-marrow-derived MSCs by using Alizarin Red,

which stains osteocytes red followed by microscopic

examination. This action is indicative of PPARδ activation,

whereas observation of adipogenesis is a sign of PPARγ

activation. The latter was ascertained before and after staining

by microscopic examination wherein the formation of white

droplets is readily discernible without the need for staining

methods. The divergent roles of PPARγ and PPARδ in MSC

diﬀerentiation are illustrated in Figure 4, showing the results

from treating MSCs with either the PPARγ agonist,

rosiglitazone, or the PPARδ agonist, GW0742. This com-

parative diﬀerentiation assay was utilized as an initial screen for

our library of compounds, and the degree of osteogenesis or

7004

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Table 1. Results of the First Series of Target Compounds Using a Comparative MSC Diﬀerentiation Assay

compd

X

R¹

R²

R⁵

AD(1 μM)

AD (10 μM)

AD (20 μM)

OS (1 μM)

OS (10 μM)

OS (20 μM)

GW0742

6b

6c

7a

7b

0

+++

++

+

0

+

0

++

0

+

0

toxic

+++

toxic

+

++

0

+

+++

0

++

toxic

0

toxic

+++

0

++++

0

++

0

+++

0

O

S

CF₃

H

CF₃

H

CF₃

H

CF₃

H

CF₃

H

CF₃

H

CF₃

H

CF₃

H

CF₃

F

H

CF₃

F

H

CF₃

F

H

CF₃

F

Me

H

0

++

+++

+

++

+

0

++

+

++

+

7c

7d

0

+++

12a

12b

12c

12d

16a

16b

16c

16d

17

++

0

+++

0

N−H

N-BOC

N-ThPh

N-Me

+++

++

+++

+

0

++

+

+++

++++

+++

toxic

+++

++

++++

0

toxic

0

++++

0

19

23a

23b

23c

23d

0

+++

0

CF₃

H

CF₃

0

++

+

++

++++

0

a

AD, adipogenesis; OS, osteogenesis. Number of plus signs (“+”) indicates an increasing OS response relative to GW0742 or vehicle control or an

increasing AD response relative to rosiglitazone or vehicle control. “Toxic” response indicates a signiﬁcant loss of overall cell numbers; >20%

compared with controls upon microscopic examination.

adipogenesis was scored in gradients relative to the osteogenic

response from treatment with GW0742 or the adipogenic

response elicited by rosiglitazone. For example, the dark-red

robust response of GW0742 was assigned “+++”, whereas if

microscopic examination appeared to be approximately half as

red or only very weakly red, scores of “++” or “+” were

assigned. Compounds unable to generate any red beyond the

nontreated control were assigned a zero value. The objective of

this preliminary assay was to identify compounds with

promising osteogenic properties and minimal adipogenesis

activity. Compounds displaying this proﬁle were further tested

in a quantitative manner by using a panel of PPAR-transfected

cells that can also assess the subtype speciﬁcity. The combined

results from these assays allows for SAR correlations between

MSC osteogenesis and PPARδ activation.

MSC Diﬀerentiation Assay. As previously noted, the results

from the MSC diﬀerentiation assay were based on microscopic

visualization of Alizarin Red staining or lipid droplet formation

for osteogenesis (OS) or adipogenesis (AD), respectively. The

degree of osteogenesis or adipogenesis is indicated by an

increased number of “plus” (+) symbols or by a “0”, indicating

no response. Figure 4 outlines the qualitative scoring relative to

the robust response from standards GW0742 and rosiglitazone.

All of the ﬁrst series of compounds are listed in Table 1 along

with their MSC diﬀerentiation data, including those for

GW0742 as a standard.

targets 7b and 7c resulted in osteogenesis at 1 μM. This

result highlighted the need for the incorporation of the

carboxylic acid functional group in the target compound

design, and additional ester compounds were not tested in the

in vitro assays, even though they can still be considered as

potential ester prodrugs if later administered in vivo.

The para-triﬂuoromethyl (R¹= CF₃) seems to be desirable

for osteogenesis, as shown by the response from 12b, 16b, and

23b. A meta-substituted ﬂuorine (R²= F) in combination with

the para-triﬂuoromethyl also seems to increase osteogenesis at

lower doses, as seen in 7d and 23d. These two phenyl ring

arrangements were later further explored by the selection of

compounds 6b, 6c, 12b, and 12d for use in the quantitative

PPAR transactivation assay. Because of the promising results

from compounds containing the R¹-CF₃, we incorporated this

structural feature into the design of further targets.

The speciﬁc heteroatom used for the X-position seems to

have the least pronounced eﬀect. Each heteroatom (O, S, N−

H, or N-Me) had at least one example of a target compound

with a score of “+++” or better at one or more concentrations.

The inclusion of N-Me at the X-position seemed to have the

greatest eﬀect of increasing osteogenesis with compounds 23b

and 23d. The success of a methylated nitrogen led to our

exploration of other alkylations at this position as part of the

second series of target compounds (23e−g). However, results

from compounds 17 and 19 showed little osteogenesis,

indicating that additionally larger groups at this position

were unfavorable. The impact of the X-position was further

explored in the quantitative PPAR transactivation assay by the

selection of 7b, 12b, 16b, and 23b.

To evaluate the importance of the carboxylic acid functional

group, intermediates 6b and 6c were tested for their ability to

increase osteogenesis. Treatment of the MSCs with the acid-

masked methyl esters 6b or 6c strictly resulted in adipogenesis.

Treatment with the analogous unmasked carboxylic acid

7005

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Table 2. Results of the Second Series of Target Compounds Using a Comparative MSC Diﬀerentiation Assay

a

AD, adipogenesis; OS, osteogenesis. Number of plus signs (“+”) indicates an increasing response relative to GW0742 or vehicle control or an

increasing AD response relative to rosiglitazone or vehicle control.

The second series of compounds explored the eﬀect of

incorporating alkyl groups at the R³position or lengthening

them at the N-alkyl X-position. This series also examined the

eﬀect that R⁴methylation has compared with R³methylation

along with the impact of two diﬀerent heterocycles. All of the

second series of compounds are listed in Table 2 along with

their MSC diﬀerentiation data, including those for GW0742 as

a standard.

The target compounds 23e,f, containing bulkier N-

alkylations than those of the previous 23b, showed an overall

decrease in osteogenesis, indicating that a simple methyl group

at this position is preferred. Similar alkylations were also

explored at the R³position. Methylation at this location (31a)

resulted in a slight decrease in osteogenesis when compared

with 23b of the previous series. However, it showed a very

signiﬁcant increase in the selectivity by completely avoiding

adipogenesis when assessed by microscopic visualization. This

observation was further supported later by the PPAR

transactivation assay, where no PPARα or PPARγ activity

was shown for compound 31a. Increasing the size of the alkyl

substitution at R³to ethyl (39a) or isopropyl (39b) resulted in

similar osteogenesis and therefore was not considered to be

more favorable than the methyl substitution. The substitution

of a methyl group at the R⁴position, compared with R³,

resulted in a decrease in osteogenesis and can be seen when

comparing 31b and 31a.

Thiazole regioisomer 44 showed an increase in osteogenesis

at 1 and 10 μM but a drop-oﬀ in activity at 20 μM when

7006

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Table 3. Transactivation Activity of Selected Compounds

a

b

a

b

a

b

X

R²

R³

PPARα EC₅₀(μM) (% activation)

PPARδ EC₅₀(μM) (% activation)

PPARγ EC₅₀(μM) (% activation)

d

e

Gemﬁbrozil

11.8 [8.93−15.6]

inactive

68.4

c

(110%)

(>52%)

9.53 [2.37−38.3]

(179%)

0.144 [0.0551−0.377]

(100%)

f

e

GW0742

2.50

0.0269 [0.00940−0.0767]

c

(>108%)

(135%)

e

Rosiglitazone

3.96

8.89 [4.17−18.9]

(159%)

0.274 [0.116−0.652]

(171%)

0.0959 [0.0533−0.173]

(151%)

0.535 [0.0472−6.07]

(114%)

0.134 [0.0730−0.246]

(160%)

1.22 [0.190−7.85]

(172%)

0.199 [0.114−0.347]

(107%)

0.335 [0.197−0.568]

(59%)

c

(>120%)

2.23 [0.138−36.1]

(135%)

51.3 [16.5−159]

(346%)

17.7 [10.0−31.4]

(338%)

0.915 [0.609−1.37]

(163%)

6.01 [2.89−12.5]

(377%)

f

d

7b

O

S

H

F

H

inactive

e

7d

13.2

(>74%)

17.0 [6.24−46.1]

(83%)

1.30 [0.497−3.40]

(114%)

5.28 [3.34−8.30]

(76%)

24.7 [2.78−220]

(80%)

inactive

12b

12d

16b

H

F

H

S

H

N−H

N-Me

H

f

23b

H

0.947 [0.650−1.38]

(44%)

f

d

31a

Me

inactive

a

Activity of reference and test compounds in the cell-based transient transfection assay with the receptor response quantiﬁed using a luciferase

reporter. Data are representative of three or more independent experiments performed in triplicate (N = 9 or more). EC₅₀values were derived from

b

a four-parameter variable slope model best ﬁt, with EC₅₀95% conﬁdence interval shown in brackets. % activation: the highest activation achieved

c

when compared with the reference compound normalized to 100%. Reference compound response normalized to 100% at the concentration used

d

e

in the test compound assays. Inactive: compound is considered inactive if EC₅₀is >100 μM or if the % activation is <40%. Maximum value of

f

normalized response could not be determined, preventing a reliable estimation of the EC₅₀95% conﬁdence interval. Potential toxicity observed at

doses signiﬁcantly higher than eﬃcacious range. See the discussion in Preliminary Toxicity Assessments section.

compared with 23b. Compound 44 also showed no adipo-

genesis, indicating a likely increase in selectivity versus PPARγ.

The other heterocycle explored was a 1,2,3-triazole that

showed a decrease in osteogenesis when comparing compound

49 with the thiazole of 23b. Interestingly, this same

substitution showed an increase in osteogenesis at lower

concentrations when the X-position was N−H, as seen when

comparing compounds 52 and 16b.

The results generated from this assay (Table 3) showed that

substitution of the R²position with ﬂuorine results in roughly a

3- to 4-fold increase in PPARδ potency when comparing 7b

with 7d or 12b with 12d. This substitution also shows an

increase in selectivity toward PPARα but a 13-fold or greater

decrease in selectivity toward PPARγ. We considered this

decrease in selectivity important to consider for the design of

an osteogenic compound due to the contrasting adipogenic

function of PPARγ activation. When considering the optimal

X-position heteroatom, a comparison between 7b, 12b, 16b,

and 23b showed that the N-Me-containing compound 23b had

the greatest PPARδ potency. Compound 23b also showed 124-

fold selectivity toward PPARδ versus PPARγ compared with

31-fold selectivity for 12b, 4-fold selectivity for 16b, and

undeﬁned selectivity for 7b. This result in combination with an

excellent osteogenic proﬁle for 23b in the MSC diﬀerentiation

assay led us to favor the N-Me X-position when designing the

structure of later compounds. Lastly, methyl substitution at the

R³position (31a) resulted in a 1.7-fold decrease in PPARδ

activity when compared with 23b. However, this substitution

resulted in a further increase in the selectivity, and 31a had

essentially no activity on PPARα or PPARγ. Despite the

decrease in the PPARδ activity, the strong increase in the

overall selectivity was considered beneﬁcial, and additional

compounds were synthesized to explore larger alkyl sub-

stitutions at the R³position.

PPAR Transactivation Assay. We selected seven com-

pounds for the quantitative measurement of PPAR potency

and subunit selectivity using a luciferase reporter assay utilizing

COS-7 cells transfected with human PPARα, PPARγ, or

PPARδ.^81,82Our selection was based on results from the MSC

diﬀerentiation assay (Tables 1 and 2) and included

compounds 7b,d, 12b,d, 16b, 23b, and 31a (chemical

structures are shown in Table 3). This series exhibited

signiﬁcant osteogenic activities while allowing for representa-

tive structural comparisons for substitutions on the head,

varying cores, and substitutions on the lipophilic tail

components. Speciﬁcally, we selected structurally diverse

compounds for maximizing possible SAR comparisons at the

R¹, R², R³, and X positions (Figure 3). Eﬀective concentration

at 50% response (EC₅₀) values were obtained for each of the

three PPAR subunits for each selected target compound. The

percent maximum activation (E_max) was measured relative to

the positive control gemﬁbrozil for PPARα, rosiglitazone for

PPARγ, and GW0742 for PPARδ (structures are shown in

Figure 1).

In vivo Testing. Six compounds, as described later, were

selected for the in vivo study involving OVX mice, a mouse

7007

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 5. 3-D μCT images of the distal femur.

Figure 6. Analysis of the femur microarchitecture of sham, OVX, or test-compound-treated mice after 6 weeks of treatment. *P < 0.0332, **P <

0.0021, ***P < 0.0002, ****P < 0.0001 versus OVX group (ordinary one-way analysis of variance (ANOVA) followed by Tukey multiple

comparisons test; Brown−Forsythe and Welch ANOVA were used for the Bone surface area (BSA)/Bone volume (BV)).

model for human postmenopausal osteoporosis.⁸³The study

used female C57BL/6 mice that were ovariectomized at 12

weeks of age. A sham surgery was also performed on a control

group of mice, leaving the ovaries intact. After a recovery

period, each group of mice was dosed with an assigned

compound or vehicle by oral gavage each day for 6 weeks. This

was followed by sacriﬁce at 22 weeks of age and micro-

computer tomography analysis of the femur trabecular

structure. The test compounds used in this study were the

same as those used in the PPAR transactivation assay, apart

from 7b. Therefore, compounds 7d, 12b, 12d, 16b, 23b, and

31a were all selected based on the same SAR rationale

previously explained for the transactivation assay. Compound

7b was excluded due to the potential for toxicity (Tables 1 and

3; see the later discussion) and for its lower PPARδ potency in

the transactivation assay (Table 3). Importantly, these six

compounds still allowed for a systematic SAR comparison with

all four linker heteroatoms included in the study. These

7008

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

compounds were tested for their ability to increase the bone

density and trabecular structure in OVX mice compared with

the sham group of mice. Comparisons were also made toward

the same positive control used in the in vitro experiments,

GW0742.

A visual inspection of the microcomputer tomography

(μCT) images (Figure 5) of the distal femur showed an

expected decrease in the trabecular structure of untreated OVX

mice, indicating the development of osteoporosis and

decreased bone density. OVX mice treated with test

compounds showed varying degrees of improvement in the

trabecular structure. Compounds 7d, 12d, 23b, and 31a

improved the bone density at least back to normal (sham

control) levels. Those treated with 12d or 31a showed an even

greater improvement in the trabecular structure, comparable to

or slightly better than GW0742.

constructs of our molecular scaﬀolds, we conducted bio-

chemical studies under more physiologic conditions.

Biochemical Studies. Compounds were incubated in

triplicate with a 10-fold excess of glutathione for up to 24 h

at 25 °C in pH 7.4 phosphate buﬀer with 50% methanol

present to ensure uniform solubility across all substrates.

Analyses were performed at several time points using LC-MS/

MS to monitor the disappearance of substrate and to conﬁrm

the structure of any glutathione-conjugated products. Target

compounds 23b, 31a, 7d, and 12d were studied to examine the

possibility that diﬀerent para-substituents on the cinnamic acid

portion of our scaﬀold (X units) may have a diﬀering inﬂuence

on the reactivity. All four of the test agents remained stable for

24 h compared with a positive-control compound, benzylide-

neacetone, which underwent 60−70% of the reaction within 4

h. Cinnamic acid was also tested and, not unexpectedly, shown

to be nonreactive. Further details of these experiments will be

published as part of a larger study that intends to convey a

useful SAR across a broad range of scaﬀolds and appendages

associated with α,β-unsaturated carbonyl systems in general.

Cell-based Observations. From the MSC diﬀerentiation

assay previously described, observations regarding cell toxicity

were also made. Cell toxicity was generally not observed for

the test compounds within relevant dose ranges of 1 to 20 μM.

However, higher levels in the 0.1 to 1 mM range did cause

cellular death (>25% overall loss of cells compared with

controls) that we deﬁned as toxicity. Interestingly, there was

one family of compounds wherein X = O did appear to

contribute to cell loss and death in a consistent manner at

lower doses, namely, beginning at 10 μM. Thus, these agents

have been noted in Table 1 as “toxic” and are likewise deﬁned

by a footnote. Whereas this may or may not translate to actual

toxicity in humans (see the later discussion), this level of cell-

based toxicity in our primary eﬃcacy model interferes with

eﬀectively assessing osteogenesis or adipogenesis. Hence we

noted these observations as “toxic” and consider them as

potential ﬂags for such compounds should they need to be

dosed at high levels that could lead to systemic and thus

cellular concentrations in the so-noted range.

Similarly, the previously described PPAR transfection assay

allows for some speculation regarding the potential toxicity of

the seven test compounds. Eﬃcacy results are shown in Table

3 with the possible toxicity noted at higher doses for a few of

the compounds by footnote f. Cells exhibiting a signiﬁcant

reduction in constitutive Renilla luciferase activity may be

considered to be experiencing toxicity. Using a light micro-

scope to monitor cell health revealed changes in the cell

morphology for PPARδ-transfected cells when treated with 50

μM of GW0742, 7b, 23b, and 31a. However, lower treatment

concentrations in the range relevant for eﬃcacious activity did

not show changes in cell morphology. This is in general

agreement with the toxicity pattern previously noted.

In Vivo Studies. During our 6 week mouse eﬃcacy studies,

none of the compounds demonstrated signs of gross toxicity,

for example, alteration of weight, behavior, eating or other

habits, compared with controls.

Selection of Potential Lead Compounds and Speciﬁc

Assessment of their Cytotoxicity. From the several

compounds showing signiﬁcant osteogenesis activity in the

preliminary human MSC assay, 12d and 31a exhibited

promising in vivo activity in the mouse model that was

comparable to or better than that of GW0742, a prototypical

PPARδ selective agonist. Compound 31a proved to be the

An assessment of bone microarchitecture also highlighted

promising improvements in bone density and structure. The

ratio of bone volume to total volume (Figure 6) was increased

the most for mice treated with compounds 12d and 31a,

reaching a ratio even higher than those in the sham group. As

previously indicated, this indicates that the bone density of the

OVX mouse was corrected and increased beyond that of the

sham mice. Mice treated with 7d and 23b also showed bone

volume ratios that returned to sham levels but were not as high

as those displayed by 12d and 31a. Other assessments of bone

density and microarchitecture were also favorable, as shown in

Figure 6. Reductions in both the bone surface area/volume

ratio and trabecular spacing further reﬂect enhanced bone

density and integrity. Likewise, signiﬁcant increases in the

trabecular number accompanied by moderate increases in the

trabecular thickness are indicative of healthy bone remodeling.

Finally, the microarchitectures for 12d, 23b, and 31a are

distinctive in the orientations of the newly forming trabeculae

along the vectors needed to best accommodate stress in this

bone’s particular gravitational orientation and physical use.

This can be discerned by comparing the test agents with the

sham control in Figure 5, where it can also be noted that the

positive-control standard GW0742 does not appear to be as

eﬀective in this regard. The composite of these assessments

indicates that treatment with compounds 12d and 31a, in

particular, results in signiﬁcant increases in bone density and is

able to return measured values to sham levels and beyond, with

the results being either comparable to or better than GW0742

in terms of both bone density and healthy microarchitecture.

Preliminary Toxicity Assessments. A speciﬁc study was

performed at the biochemical level, and general observations

were made during the eﬃcacy-related cell culture and in vivo

studies. In addition to the possibility of toxicity after long-term

dosing, of potential immediate concern was the presence of an

α,β-unsaturated carbonyl system. The latter may be subject to

spurious Michael addition reactions that could lead to oﬀ-

target toxicity.^68,69However, when the carbonyl is part of a

carboxylic acid or ester functional group, Michael acceptor

reactivity is signiﬁcantly reduced⁶⁸⁻⁷³(Figure 2D). For

example, cinnamic acid and its simple esters have been

extensively deployed within the food and perfume industry for

many years without incident.⁷⁴⁻⁷⁶We saw no evidence of

Michael addition side products during our synthetic reactions

when basic amines capable of serving as donor partners were

also present. Nevertheless, to conﬁrm that this type of

reactivity is not insidiously present within the speciﬁc

7009

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 7. Cytotoxicity studies using MTT analyses of potential lead compounds 12d and 31a and of a standard PPARδ agonist GW0742. The top

two graphs and the lower left graph are summaries of our studies using HEK 293T/17 cells. The lower right graph is a summary of the results from

independent studies conducted by NAMSA (see the accompanying text) using L-929 mouse ﬁbroblast cells. Our experiments were conducted in

quadruplicate (N = 4), and graphs of the mean with standard deviations were produced using Prism software. Points at 0 μM have only cell

medium. Points at “Vehicle” have no drug while containing the drug vehicle at its highest DMSO concentration (0.1%), equivalent to what would

be present after a ﬁrst dilution of drug stock solution with the cell medium (i.e., a 100 μM drug dose). Points thereafter are increasing drug doses in

the vehicle obtained after serial dilutions with additional cell medium. Asterisks (*) denote statistically signiﬁcant diﬀerent data points compared

with their 0 data points using Graphpad to calculate a two-tailed P value. The upper left plot is for our toxic positive control (H₂O₂), which

provides a classic dose-toxicity−response relationship across the concentration range of primary interest for our compounds’ eﬃcacies and for what

we considered as acceptable toxicity levels (namely, ca. 100 times or more higher than the eﬃcacious dose). Translation to potential toxicity in

humans is generally considered to be relevant when the cell viability becomes <70% after 24 h of exposure of the cell culture to a given drug

concentration (upper right graph). To further prompt toxicity and assess possible SAR aspects, we also ran our studies for exposure times of 72 h

(lower left graph). The NAMSA experiments were conducted in triplicate (N = 3), and their data are routinely reported as the tabulated averages

for each concentration tested. These tabulated data are provided in the Supporting Information. For a ready comparison to our summary graphs,

we also manually plotted the NAMSA averages using ChemDraw. In this case, the plus symbols (+) denote that the falloﬀ in cell viability at that

concentration of drug may be relevant for predicting potential toxicity in humans.

most selective for PPARδ among all of the compounds studied

in our transfection assay. Alternatively, 12d’s selectivity proﬁle

was nearly evenly balanced as an agonist across all three PPAR

subtypes. Because they represented the most promising

candidates from our in vivo studies, we decided to pursue

12d and 31a as potential lead compounds while also assessing

the potential impact of their diﬀerences in PPAR selectivity

with regard to a toxicity-related SAR. Our ﬁrst step was to

initiate validated Good Laboratory Practice (GLP)-compliant

bioanalytical assays for each lead compound using LC-MS/MS

instrumentation. These methods are provided in the

Experimental Section, and the results are added to each

compound’s chemical synthesis details. Our next step was to

evaluate their cytotoxicity in a more speciﬁc manner to

determine if their signiﬁcant diﬀerence in PPAR selectivity

would have an impact on this critical aspect of their ADMET

(absorption, distribution, metabolism, excretion, and toxicity)

proﬁles. Toward this end, we performed cell viability tests

using human embryonic kidney cell (HEK 293T/17) cultures

assessed by standard MTT analyses.^84,85These studies were

initially conducted for the typical 24 h exposure of the test

agent to the HEK 293T/17 cultures. Because our compounds

demonstrated only minimal losses in cell viability, these studies

were also repeated for 72 h exposure times to further prompt

more pronounced toxicity end points. As in all of our prior

assays, GW0742 was included as a prototypical PPARδ agonist.

Hydrogen peroxide was used as a toxic positive control,⁸⁶non-

drug-containing medium was used as a negative control, and

our drug-vehicle solution without drug was studied at its

highest concentration of dimethyl sulfoxide (DMSO) (0.1% in

cell culture medium prior to the addition to the cell culture

milieu) that would be given to cells when used to deliver the

highest concentration of drug. The results for all tests are

summarized in Figure 7 as graphical representations. Tables of

the raw data and individual plots for all tests are provided in

the Supporting Information.

As expected, cell media and vehicle controls⁸⁷had essentially

no eﬀect on the cell viability at 24 or 72 h of exposure. Also, as

expected, hydrogen peroxide exhibited a typical dose−response

curve, making a transition from >70% cell viability to ca. 60%

viability at ca. 100 μM after either 24 or 72 h of exposure.⁸⁶

This transition percent range is thought to be relevant when

assessing the appropriateness of translating the cell-based

toxicity to potential toxicity within humans during clinical

studies.⁸⁸After the 24 h of exposure, GW0742 and 12d had

little eﬀect on the cell viability, whereas 31a began to

demonstrate modest but statistically signiﬁcant toxicity at 10

and 100 μM, despite it having the highest PPAR selectivity

7010

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 8. Chemical structures for comparison. GW0742, L165041, cardarine, and 31a are PPARδ agonists. INT131 is a PPARγ agonist. The latter’s

acidic sulfonamide group has been situated toward the “west” to be at least somewhat similar to the other structures. All of these compounds have

been reported to be osteogenic in vitro. Whereas cardarine decreased bone density in rat studies, 31a maintained its osteogenic properties in mouse

studies as reported herein. Elaﬁbranor and seladelpar are PPARδ agonists that have undergone advanced clinical testing. Although cardarine proved

to be carcinogenic during 2 year toxicity studies in rodents, both elaﬁbranor and seladelpar have more recently been found to not be problematic.

proﬁle. Importantly, none of the test agents dipped into the

relevant toxicity range for potentially ﬂagging the possibility of

displaying human toxicity. We then pushed the assay’s typical

time frame further by extending the test agent exposure to 72

h. All three agents proved to be toxic to cells in the key loss of

viability range at doses somewhere higher than 10 μM and

lower than 100 μM. In addition to our own cytotoxicity study,

we also contracted to have these three compounds

independently examined in a similar assay conducted by

NAMSA (Northwood, OH). NAMSA is an FDA-certiﬁed

contract research organization that specializes in GLP toxicity

assessments of devices and bioimplantable materials.⁸⁹One of

their cytotoxicity assays also lends itself to testing solid and

liquid samples/formulations of potential pharmaceutical

agents. For comparison, the NAMSA experimental procedure

is recorded herein within the Experimental Section immedi-

ately after that of our own study’s method. A hand-plot

summarized from their data is also provided in Figure 7 for

ready comparison to our summarized data. Raw data are

tabulated in the Supporting Information. The NAMSA results

are in close agreement with our ﬁndings. Taken together, the

data support the overall conclusion that GW0742 and 31a

demonstrate “cytotoxic potential” at 100 μM, with 31a

possibly doing so at a somewhat lower dose as well.

Alternatively, 12d remains nontoxic unless exposure times

are elongated past the typical assessment time frame generally

utilized for these types of assays. Importantly, even for 31a, the

lower threshold for potential toxicity is ∼50 times higher than

its eﬃcacious dose range in the MSC assay. Interestingly,

PPAR selectivity does not appear to be correlated with less

cytotoxicity. If the PPARδ potency taken from the trans-

activation assay is also taken into consideration analogous to a

therapeutic index (TI) by calculating an approximated “cell

toxicity index” (CTI = CT₅₀/EC₅₀, wherein the denominator

can be taken directly from the transactivation assay and the

numerator can be represented by the average from

extrapolated estimates derived from the plots of both of the

cytotoxicity assays), then all three compounds demonstrate at

least 100-fold margins of “cell safety”. The calculated CTI

values are approximately 1700 for GW0742, 1350 for 12d, and

250 for 31a. Although these numbers reﬂect approximations

and their individual values should not be misconstrued with

actual TI calculations, together their relative rankings can be

informative from a toxicity SAR perspective. For example, it

appears that their sequence of relative cell-based “safeties” is

directly proportional to the PPARδ potency (and toxicity is

indirectly proportional to eﬃcacy). When this observation is

then taken together with the noted PPAR selectivity

assessment, it seems reasonable to speculate that net

cytotoxicity does not derive from a simple overstimulation of

the PPARδ pathway and likely does not derive from the

overstimulation of either of the other two PPAR pathways

either. Similarly, our results begin to hint that toxicity may

occur in a stepwise manner relative to increasing dose. This is

suggested most clearly for 31a in our 72 h toxicity−SAR

experiment. It then becomes quite apparent in the NAMSA

study for all three compounds because their experiments

involved an additional three dose levels in the key range

between 1 (or 3) μM and 10 (or 12) to 100 μM, namely,

additional NAMSA doses at 6, 25, and 50 μM. Taken with the

7011

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

previously noted suggestions, these stepped curves imply that

there may be multiple pathways leading to net toxicity that are

somewhat similar among all three compounds and for which

none of the mechanisms are likely linked in a proportional

manner to any of the PPAR pathways.

authors ﬁnally conclude “that the tumorigenicity of GW501516

(also known as cardarine) may be a compound-speciﬁc not

PPARδ-speciﬁc phenomenon”.⁴⁰This further underscores

what we ﬁrst raised as a general question in the Introduction

and then were able to discern above as an even more distinct

possibility from the Takeda and Makishima patent report when

their general assessments were coupled to our speciﬁc results,

as reported herein.

Translational Consideration of Toxicity Relative to

Use in Humans. Our preliminary toxicity assessments and

speciﬁc cytotoxicity experiments have not raised any red ﬂags

with regard to moving 12d or 31a into preclinical develop-

ment; however, because of the unfavorable history still looming

from cardarine, long-term toxicity studies in rodents (up to at

least 2 years) will need to be initiated as an early step toward

the preclinical development of any PPARδ drug intended as a

prolonged treatment for a chronic illness. Such studies are

beyond the scope of early drug discovery eﬀorts, which in our

case focused ﬁrst on simply assessing the possibility to impact

favorably upon bone disease and, if so, then secondarily toward

identifying potential candidate compounds for future testing.

Alternatively, it is noteworthy that commentary applicable to

addressing short- and long-term toxicity versus eﬃcacy is

aﬀorded in the two recent reviews previously mentioned in the

Introduction. In their patent review, Takada and Makishima³⁵

ﬁrst refer to favorable eﬀects on bone resulting from treatments

with two diﬀerent PPARδ agonists, namely, GW0742 (our

prior work)²⁴and L165041.^33,90,91Likewise, we referred to

mechanistic studies by others that also showed cardarine to

have favorable eﬀects on bone.¹⁸On the contrary, Takeda and

Makishima next cite studies in rats using cardarine that caused

a decrease in bone density.⁹²Our studies reported herein used

mice wherein we identiﬁed several cardarine analogs to have

very positive eﬀects on bone. Thus we fully agree with Takeda

and Makishima’s concluding remarks that the “eﬀects of novel

PPARδ agonists in the treatment of bone disorders require

careful evaluation”, and we further add that our work clearly

demonstrates that certain analogs can favorably diverge from

cardarine in terms of osteogenic activity.³⁵This, in turn, raises

the distinct possibility that analogs may also be able to diverge

from cardarine in terms of the latter’s long-term toxicity.

Finally, these authors also discuss the PPARγ agonist,

INT131,⁹³which can increase bone density⁹⁴by a

“mechanism” that “remains unclear”.³⁵A side-by-side compar-

ison of all of the aforementioned compounds is provided in

Figure 8 along with one of our more preferred agents, 31a. The

diﬀerent structure of the PPARγ compound is immediately

evident even after similarly aligning the orientation of its

weakly acidic sulfonamide.

In their medicinal chemistry review, Kadayat et al. cite nine

PPARδ agonists that have recently “reached early stage clinical

trials, with some of them under active development.”⁴⁰Several

of these are intended for shorter term treatments. This range of

compounds moving into the clinic suggests that, in general, the

PPARδ agonists as a family are typically nonproblematic in

terms of acute or short-term toxicity. As previously discussed,

our preliminary results are in accord with this overall

assessment. Similarly, for PPAR agonists, in general, “it is

well established that drug candidates are nongenotoxic in vitro

and in vivo”.³¹Alternatively, as we have previously emphasized,

it is their potential for long-term toxicity, likely resulting from

oﬀ-target actions, that remains to be clariﬁed. Importantly in

this regard, Kadayat et al. also note that from their listing, two

of the more advanced compounds “have cleared the 2 year

carcinogenicity studies in rodents”,⁴⁰namely, elaﬁbranor⁹⁵and

seladelpar⁹⁶(structures also shown in Figure 8). Thus these

CONCLUSIONS

■

We synthesized a directed library of about 30 small-molecule

probes designed to explore the selective activation of PPARδ

and the possibility of promoting osteogenesis and bone

formation. SARs were examined in the middle and at each

end of a classical PPAR scaﬀold consisting of three

independent rings. Probe features were designed by analogy

to the structural requirements of known PPARδ agonists as

well as with the intention of exploring certain novel

modiﬁcations. Several compounds showed promising osteo-

genic activity in human MSC cultures. Six of these compounds

were then tested in a PPAR transactivation assay and in an in

vivo mouse model of human osteoporosis that required the in

situ activation of inherent stem cells. Results from the

transactivation assay provided quantitative assessments of

potency and PPAR selectivity, with compound 7d being the

most potent and compound 31a standing out as the most

selective toward PPARδ. Compounds 7d, 12d, 23b, and 31a

promoted improvements in bone density and microarchitec-

ture in vivo, with compounds 12d and 31a, in particular,

showing activity comparable to or better than GW0742, a

classic PPARδ agonist. Because of their signiﬁcant in vivo

activity, 12d and 31a are regarded as our more preferred lead

compounds. The extremely high selectivity for PPARδ

demonstrated by 31a is also noteworthy. Although an α,β-

unsaturated carbonyl is present, the compounds do not

undergo Michael addition reactions under simulated physio-

logic conditions. Most of the compounds do not display any

signiﬁcant cell-based toxicity within their eﬃcacious dose

ranges. No gross toxicity was observed during the in vivo

studies, which involved drug treatments for up to 6 weeks.

Speciﬁc cytotoxicity assessments of 12d and 31a did not reveal

any ﬂags for further consideration as preclinical development

compounds. Interestingly, 12d proved to be less toxic than 31a

despite the latter’s high degree of selectivity. Cytotoxicity does

not appear to be associated with a simple overstimulation of

any of the PPAR pathways and, instead, likely derives from a

composite of more than one other type of mechanism.

Importantly, because of the clinical history for this class of

compound, long-term (2-year) toxicity studies in rodents will

need to be initiated early as a key step during preclinical

development. Because most osteoporosis treatments currently

rely upon the inhibition of osteoclast function, our results

support PPARδ as a possible novel mechanistic target for

treating osteoporosis or fracture healing through the in situ

promotion of bone cell proliferation. Thus, continued success

of prompting this mechanism through drug development and

clinical studies without encountering signiﬁcant unwanted side

eﬀects or toxicity could eventually result in a major paradigm

shift for the treatment of osteoporosis, fractures, and other

metabolic bone diseases.

7012

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

was added concentrated H₂SO₄(1 mL), and the solution was heated

to reﬂux temperature for 20 h. The mixture was cooled to room

temperature and concentrated under reduced pressure. The residue

was dissolved in ethyl acetate, washed with water, dried with Na₂SO₄,

and concentrated. The residue was puriﬁed by ﬂash chromatography

to give 5 as a white solid (0.574 g, 3.221 mmol, 75.1%). TLC R_f(25%

EXPERIMENTAL SECTION

■

Chemistry. General Methods. Chemical reactions were con-

ducted using reagents purchased commercially from Sigma-Aldrich or

Fisher Scientiﬁc International with the exception of the starting

materials 4-(triﬂuoromethyl)thiobenzamide 1b and 3-ﬂuoro-4-

(triﬂuoromethyl)thiobenzamide 1d, which were purchased from

Matrix Scientiﬁc. TLC was completed using Baker-ﬂex precoated

ﬂexible TLC sheets purchased from VWR International (cat. no.

JT4463-2) and visualized using shortwave ultraviolet light or by

iodine chamber. Flash chromatography was completed using

technical-grade silica gel (230−400 mesh particle size, 40−63 μm

particle size, 60 Å pore size) purchased from Sigma-Aldrich. Melting

points (Mp) were recorded using a Mel-Temp II melting point

apparatus equipped with a 250 °C thermometer. Nuclear magnetic

resonance (NMR) spectra were recorded using a Bruker Avance III

600 MHz spectrometer with a cryoprobe. Peak locations were

referenced using a residual nondeuterated solvent and recorded as the

chemical shift (δ) measured in parts per million (ppm). All

compounds had >95% purity, as assessed by NMR and C,H,N

elemental analysis. Elemental analysis was performed by Atlantic

Microlab, and experimental values within 0.4% of the calculated

values were considered acceptable. A few compounds contain traces

of solvent that were likewise well-deﬁned by elemental analyses and

accounted for as formula weights when conducting biological studies.

4-Methyl-2-phenylthiazole-5-carboxylic Acid Ethyl Ester (2a). To

a suspension of thiobenzamide 1a (6.05 g, 0.044 mol) in 95% ethanol

was added ethyl 2-chloroacetoacetate (6.10 mL, 0.044 mol), and the

mixture was stirred at reﬂux temperature for 26 h. The reaction

mixture was concentrated under reduced pressure, and the resulting

residue was suspended in ice-cold hexane and stirred for 20 min. The

suspension was ﬁltered, and 2a was collected as a cream-colored solid

(7.434 g, 0.030 mol, 68.3%). TLC R_f(25% EtOAc/hexane) = 0.63.

Mp 84−87 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.19 (2H, d, J = 7.32

Hz), 7.56 (1H, t, J = 7.32 Hz), 7.53 (2H, t, J = 7.08 Hz), 4.41 (2H, q,

J = 7.14 Hz), 2.94 (3H, s), 1.41 (3H, t, J = 7.14 Hz). ¹³C NMR

(CDCl₃, 150 MHz): δ 171.25, 161.07, 157.83, 133.21, 129.82, 128.10,

122.55, 62.37, 16.29, 14.52.

1

EtOAc/hexane) = 0.18. Mp 137−138 °C. H NMR (CDCl₃, 600

MHz): δ 7.65 (1H, d, J = 16.02 Hz), 7.44 (2H, d, J = 8.58), 6.86 (2H,

d, J = 8.58), 6.31 (1H, d, J = 15.98), 5.57 (1H, s), 3.81 (3H, s). ¹³C

NMR (CDCl₃, 150 MHz): δ 168.22, 157.87, 144.85, 130.14, 127.29,

116.01, 115.26, 51.87.

Methyl 4-[[4-Methyl-2-phenylthiazol-5-yl]methyl]-

methoxycinnamate (6a). To a stirred solution of methyl 4-

hydroxycinnamate 5 (0.142 g, 0.797 mmol) and chloromethyl 4a

(0.150 g, 0.670 mmol) in anhydrous acetonitrile (5 mL) was added

cesium carbonate with partial solubility. The resulting mixture was

stirred for 24 h at room temperature at which TLC showed that the

chloromethyl 4a had been consumed. The reaction mixture was

concentrated, and the residue was dissolved in ethyl acetate and

washed with water and brine, dried with Na₂SO₄, and concentrated.

Flash chromatography failed to give a pure product and the crude

white solid 6a (0.202 g, 0.553 mmol, 82.6%) collected was moved to

the next step without further puriﬁcation.

4-[[4-Methyl-2-phenylthiazol-5-yl]methyl]methoxycinnamic

Acid (7a). To a stirred solution of methyl ester 6a was added

dropwise 3 N NaOH. After 20 h, the mixture was acidiﬁed with 1 N

HCl to pH 1 to 2 and concentrated. The residue was suspended in

ethyl acetate and washed with water and brine. The aqueous phase

was extracted with a separate portion of ethyl acetate, and the organic

phases were combined, dried with Na₂SO₄, and concentrated. The

residue was puriﬁed by ﬂash chromatography using a 25% EtOAc/

hexane to 100% EtOAc mobile-phase gradient to give 7a as a white

solid (0.061 g, 0.173 mmol, 41.7%). TLC R_f(50% EtOAc/hexane) =

0.17. Mp 209−211 °C. ¹H NMR (DMSO-d₆, 600 MHz): δ 7.91 (2H,

m), 7.67 (2H, d, J = 8.76 Hz), 7.55 (1H, d, J = 15.96 Hz), 7.49 (3H,

m), 7.09 (2H, d, J = 8.82 Hz), 6.41 (1H, d, J = 15.96 Hz), 5.38 (2H,

s), 2.46 (4H, s). ¹³C NMR (DMSO-d₆, 150 MHz): δ 159.85, 130.43,

129.72, 126.40, 115.83, 79.21, 15.43. Anal. Calcd for C₂₀H₁₇NO₃S

(with 0.3 EtOH mol per target): C, 67.74; H, 5.54; N, 3.83. Found:

C, 67.97; H, 5.54; N, 3.76.

4-Methyl-2-phenyl-5-thiazolemethanol (3a). To a stirred sol-

ution of ethyl ester 2a (0.304 g, 1.237 mmol) in anhydrous THF (1

mL) at 0 °C was added 2 M lithium aluminum hydride solution in

THF (1.24 mL, 2.48 mmol). The resulting mixture was stirred under

inert gas at 0 °C for 1.5 h. The reaction mixture was quenched by the

careful addition of water (0.5 mL) followed by dilution with ethyl

acetate (2.5 mL), and drying with anhydrous sodium sulfate (0.92 g).

The mixture was stirred for 15 min and was ﬁltered and concentrated

under reduced pressure to give 3a as a light-yellow solid (0.215 g,

1.053 mmol, 85.1%). TLC R_f(25% EtOAc/hexane) = 0.11. Mp 101−

Ethyl 4-Methyl-2-[4-(triﬂuoromethyl)phenyl]-thiazole-5-carbox-

ylate (2b). By analogy to the procedure described in example 2a, 4-

(triﬂuoromethyl)thiobenzamide 1a (1.065 g, 5.190 mmol) was

treated with ethyl-2-chloroacetoacetate (0.75 mL, 5.42 mmol) in

95% ethanol (30 mL) to give 2b as a cream-colored solid (1.148 g,

3.644 mmol, 70.2%). TLC R_f(25% EtOAc/hexane) = 0.69. Mp 89−

1

89.5 °C. H NMR (CDCl₃, 600 MHz): δ 8.10 (2H, d, J = 8.04 Hz),

7.73 (2H, d, J = 8.16 Hz), 4.39 (2H, q, J = 7.14 Hz), 2.81 (3H, s),

1.42 (3H, t, J = 7.08 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 167.87,

162.17, 161.38, 136.16, 132.59 (q, ²J_FC= 33 Hz), 127.18, 126.22 (m),

123.90 (q, ¹J_FC= 270 Hz), 123.13, 61.61, 17.67, 14.47. Anal. Calcd for

C₁₄H₁₂F₃NO₂S (with 0.2 H₂O mol per target): C, 52.73; H, 3.92; N,

4.39. Found: C, 52.59; H, 3.85; N, 4.71.

1

102 °C. H NMR (CDCl₃, 600 MHz): δ 7.88 (2H, d, J = 7.92 Hz),

7.41 (3H, m), 4.79 (2H, s), 2.94 (1H, s), 2.41 (3H, s). ¹³C NMR

(CDCl₃, 150 MHz): δ 166.39, 150.22, 133.59, 131.49, 130.07, 129.03,

126.46, 56.79, 15.13.

5-Chloromethyl-4-methyl-2-phenyl-thiazole (4a). To a stirred

solution of alcohol 3a (4.095 g, 0.019 mol) in anhydrous

dichloromethane (100 mL) was added triethylamine (5.50 mL,

0.039 mol). The resulting mixture was cooled to 4 °C, and

methanesulfonyl chloride (2.30 mL, 0.029 mol) was slowly added.

The mixture was stirred at 4 °C for 24 h and then diluted with

dichloromethane (100 mL), washed with saturated NaHCO₃

solution, water, and brine, dried with Na₂SO₄, and concentrated.

The residue was puriﬁed by ﬂash chromatography using 10% ethyl

acetate/hexane to give 4a as a light-yellow solid (2.850 g, 0.013 mol,

4-Methyl-2-[4-(triﬂuoromethyl)phenyl]-thiazole-5-methanol

(3b). By analogy to the procedure described in example 3a, ethyl ester

2b (1.320 g, 4.190 mmol) in solution with anhydrous THF (15 mL)

was treated with 2 M LiAlH₄solution in THF (2.0 mL, 4.0 mmol) to

give 3b as a yellow solid (0.904 g, 3.308 mmol, 79.0%). TLC R_f(25%

1

EtOAc/hexane) = 0.16. Mp 121.5−122 °C. H NMR (CDCl₃, 600

MHz): δ 8.02 (2H, d, J = 8.04 Hz), 7.69 (2H, d, J = 8.16 Hz), 4.86

(2H, s), 2.48 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 164.33,

150.91, 136.74, 132.57, 131.63 (q, ²J_FC= 33 Hz), 126.66, 126.08 (m),

1

64.0%). TLC R_f(25% EtOAc/hexane) = 0.57. Mp 89−90 °C. H

124.03 (q, J_FC= 270 Hz), 57.04, 15.27.

NMR (CDCl₃, 600 MHz): δ 7.90 (2H, m), 7.43 (3H, m), 4.80 (2H,

s), 2.50 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 166.99, 152.68,

133.34, 130.27, 128.97, 127.55, 126.46, 37.53, 15.10. Anal. Calcd for

C₁₁H₁₀NSCl: C, 59.06; H, 4.51; N, 6.26. Found: C, 59.32; H, 4.58; N,

6.33.

Methyl 4-Hydroxycinnamate (5). To a stirred solution of p-

coumaric acid (0.704 g, 4.288 mmol) in anhydrous methanol (10 mL)

5-Chloromethyl-4-methyl-2-[4-(triﬂuoromethyl)phenyl]-thiazole

(4b). By analogy to the procedure described in example 4a, alcohol 3b

(0.883 g, 3.231 mmol) was treated with methanesulfonyl chloride

(0.40 mL, 5.168 mmol) and triethylamine (0.90 mL, 6.462 mmol) in

anhydrous DCM (25 mL) to give 4b as a light-yellow solid (0.790 g,

2.708 mmol, 83.8%). TLC R_f(25% EtOAc/hexane) = 0.53. Mp 68.5−

1

69 °C. H NMR (CDCl₃, 600 MHz): δ 8.03 (2H, d, J = 8.10 Hz),

7013

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

7.70 (2H, d, J = 8.16 Hz), 4.81 (2H, s), 2.52 (3H, s). ¹³C NMR

(1.295 g, 4.739 mmol) was treated with methanesulfonyl chloride

(0.55 mL, 7.11 mmol) and triethylamine (1.3 mL, 9.33 mmol) in

anhydrous DCM (25 mL) to give 4c as a light-yellow solid (0.830 g,

2.846 mmol, 60.1%). TLC R_f(25% EtOAc/hexane) = 0.61. Mp 43−

44 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.19 (1H, s), 8.07 (1H, d, J =

7.8 Hz), 7.68 (1H, d, J = 7.8 Hz), 7.57 (1H, t, J = 7.8 Hz), 4.81 (2H,

s), 2.51 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 165.09, 153.16,

2

(CDCl₃, 150 MHz): δ 165.02, 153.25, 136.51, 131.92 (q, J_FC= 33

1

Hz), 129.06, 126.77, 126.12 (m), 123.99 (q, J_FC= 270 Hz), 37.32,

15.20. Anal. Calcd for C₁₂H₉ClF₃NS: C, 49.41; H, 3.11; N, 4.80.

Found: C, 49.43; H, 3.22; N, 4.75.

Methyl 4-[[4-Methyl-2-[4-(triﬂuoromethyl)phenyl]-thiazol-5-yl]-

methyl]methoxy]cinnamate (6b). By analogy to the procedure

described in example 6a, chloromethyl 4b (0.331 g, 1.135 mmol) and

methyl 4-hydroxycinnamate 5 (0.212 g, 1.133 mmol) were treated

with cesium carbonate (0.570 g, 1.749 mmol) in anhydrous

acetonitrile (6 mL) to give 6b as a light-yellow solid (0.365 g,

0.842 mmol, 74.3%). TLC R_f(25% EtOAc/hexane) = 0.32. Mp 153−

2

134.21, 131.68 (q, J_FC= 33 Hz), 129.70, 129.67 128.75, 126.78 (q,

³J_FC= 3.6 Hz), 123.90 (q, ¹J_FC= 271 Hz), 123.30 (q, ³J_FC= 3.6 Hz),

37.37, 15.21. Anal. Calcd for C₁₂H₉ClF₃NS: C, 49.41; H, 3.11; N,

4.80. Found: C, 49.46; H, 3.16; N, 4.91.

Ester 4-[[4-Methyl-2-[3-(triﬂuoromethyl)phenyl]-thiazol-5-yl]-

methyl]methoxycinnamate (6c). By analogy to the procedure

described in example 6a, chloromethyl 4c (0.472 g, 1.618 mmol)

and methyl 4-hydroxycinnamate 5 (0.281 g, 1.577 mmol) were

treated with cesium carbonate (0.773 g, 2.372 mmol) in anhydrous

acetonitrile (10 mL) to give 6c as a white solid (0.551 g, 1.270 mmol,

1

155 °C. H NMR (CDCl₃, 600 MHz): δ 8.04 (2H, d, J = 8.1 Hz),

7.70 (2H, d, J = 8.2 Hz), 7.68 (1H, d, J = 16.0 Hz), 7.52 (2H, d, J =

8.6 Hz), 7.00 (2H, d, J = 8.7 Hz), 6.36 (1H, d, J = 16.0 Hz), 5.24 (2H,

s), 3.8 (3H, s), 2.54 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 167.79,

2

165.13, 159.81, 152.49, 144.33, 136.67, 131.80 (q, J_FC= 33 Hz),

1

129.96, 128.20, 127.61, 126.75, 126.10 (m), 124.02 (q, J_FC= 270

80.5%). TLC R_f(25% EtOAc/hexane) = 0.30. Mp 125−127 °C. H

Hz), 116.08, 115.38, 62.29, 51.82, 15.57. Anal. Calcd for

C₂₂H₁₈F₃NO₃S: C, 60.96; H, 4.19; N, 3.23. Found: C, 60.79; H,

4.36; N, 3.15.

NMR (CDCl₃, 600 MHz): δ 8.19 (1H, s), 8.09 (1H, d, J = 7.7 Hz),

7.68 (1H, d, J = 7.8 Hz), 7.67 (1H, d, J = 16.0 Hz), 7.56 (1H, t, J =

7.8 Hz), 7.51 (2H, d, J = 8.6 Hz), 6.99 (2H, d, J = 8.7 Hz), 5.24 (2H,

s), 3.81 (3H, s), 2.54 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ

4-[[4-Methyl-2-[4-(triﬂuoromethyl)phenyl]-thiazol-5-yl]methyl]-

methoxycinnamic Acid (7b). Methyl ester 6b (0.117 g, 0.270 mmol)

was dissolved in 95% ethanol (5 mL) and THF (5 mL). 3 N NaOH

(1.0 mL) was slowly added, and the reaction mixture was stirred

overnight followed by neutralization with 1 N HCl to pH 2. The

solvent was evaporated under reduced pressure, and the residue was

taken up into EtOAc, washed with saturated NaHCO₃and brine,

dried with Na₂SO₄, ﬁltered, and concentrated. The crude material was

puriﬁed with ﬂash chromatography using a 25−50% EtOAc/hexane

mobile-phase gradient supplemented with dropwise acetic acid. The

recovered product was recrystallized using 95% ethanol to give 7b as a

white solid (0.064 g, 0.152 mmol, 56.5%). TLC R_f(50% EtOAc/

hexane) = 0.15. Mp 227−228 °C. ¹H NMR (acetone-d₆, 600 MHz): δ

8.18 (2H, d, J = 8.1 Hz), 7.85 (2H, d, J = 8.2 Hz), 7.67 (2H, d, J = 8.8

Hz), 7.63 (1H, d, J = 16.0 Hz), 7.13 (2H, d, J = 8.8 Hz), 6.41 (1H, d,

J = 16.0 Hz), 5.44 (2H, s), 2.53 (3H, s). ¹³C NMR (acetone-d₆, 150

MHz): δ 167.94, 164.82, 160.83, 153.28, 144.93, 137.90, 131.71 (q,

²J_FC= 33 Hz), 130.75, 129.46, 128.87, 127.50, 126.93 (m), 125.11 (q,

2

167.81, 165.26, 159.79, 152.23, 144.34, 134.20, 131.66 (q, J_FC= 32

3

Hz), 129.96, 129.72, 129.67, 128.19, 127.36, 126.72 (q, J_FC= 3.9

1

3

Hz), 125.71 (q, J_FC= 270 Hz), 123.31 (q, J_FC= 3.5 Hz), 116.06,

115.39, 62.28, 51.81, 15.51. Anal. Calcd for C₂₂H₁₈F₃NO₃S (with 0.4

H₂O mol per target): C, 59.97; H, 4.30; N, 3.18. Found: C, 59.66; H,

4.29; N, 3.07.

4-[[4-Methyl-2-[3-(triﬂuoromethyl)phenyl]-thiazol-5-yl]methyl]-

methoxycinnamic Acid (7c). By analogy to the procedure described

in example 7a, methyl ester 6c (0.207 g, 0.477 mmol) was treated

with 3 N NaOH (1 mL) in 95% ethanol (16 mL) to give 7c as a white

solid (0.138 g, 0.329 mmol, 69.0%). TLC R_f(50% EtOAc/hexane) =

1

0.40. Mp 179.5−181 °C. H NMR (acetone-d₆, 600 MHz): δ 8.29

(1H, s), 8.23 (1H, d, J = 7.8 Hz), 7.83 (1H, d, J = 7.8 Hz), 7.76 (1H,

t, J = 7.8 Hz), 7.69 (2H, d, J = 8.7 Hz), 7.65 (1H, d, J = 16.0 Hz), 7.14

(2H, d, J = 8.7 Hz), 6.43 (1H, d, J = 16.0 Hz), 5.46 (2H, s), 2.54 (3H,

s). ¹³C NMR (acetone-d₆, 150 MHz): δ 167.93, 164.81, 160.85,

2

153.12, 145.02, 135.37, 131.73 (q, J_FC= 32 Hz), 131.14, 130.77,

3

1

¹J_FC= 270 Hz), 117.15, 116.24, 62.85, 15.42. ¹⁹F NMR (acetone-d₆,

376 MHz): δ −63.67 (3F, s). Anal. Calcd for C₂₁H₁₆F₃NO₃S: C,

60.14; H, 3.85; N, 3.34. Found: C, 60.12; H, 3.85; N, 3.31.

130.70, 129.09, 128.85, 127.27 (q, J_FC= 3.3 Hz), 125.01 (q, J_FC

=

F

270 Hz), 123.11 (q, J_FC= 3.5 Hz), 117.05, 116.24, 62.86, 15.42. ¹⁹

3

NMR (acetone-d₆, 376 MHz): δ −63.76 (s, 3F). Anal. Calcd for

C₂₁H₁₆F₃NO₃S: C, 60.14; H, 3.85; N, 3.34. Found: C, 59.96; H, 3.89;

N, 3.30.

Ethyl 4-Methyl-2-[3-(triﬂuoromethyl)phenyl]-thiazole-5-carbox-

ylate (2c). By analogy to the procedure described in example 2a, 3-

(triﬂuoromethyl)thiobenzamide 1c (2.00 g, 9.75 mmol) was treated

with ethyl-2-chloroacetoacetate (1.35 mL) in 95% ethanol (55 mL) to

give 2c as a cream-colored solid (2.103 g, 6.677 mmol, 68.5%). TLC

Ethyl 4-Methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)phenyl]thiazole-5-

carboxylate (2d). By analogy to the procedure described in example

2a, 3-ﬂuoro-4-(triﬂuoromethyl)thiobenzamide 1d (1.606 g, 7.195

mmol) was treated with ethyl-2-chloroacetoacetate (1.00 mL, 7.23

mmol) in 95% ethanol (40 mL) to give 2d as a cream-colored solid

(1.657 g, 4.971 mmol, 69.1%). TLC R_f(25% EtOAc/hexane) = 0.72.

1

R_f(25% EtOAc/hexane) = 0.63. Mp 90−91 °C. H NMR (CDCl₃,

600 MHz): δ 8.26 (1H, s), 8.14 (1H, d, J = 7.86 Hz), 7.73 (1H, d, J =

7.80 Hz), 7.60 (1H, dd, J₁= 7.86 Hz, J₂= 7.80 Hz), 4.38 (2H, q, J =

7.14 Hz), 2.81 (3H, s), 1.41 (3H, t, J = 7.14 Hz). ¹³C NMR (CDCl₃,

1

Mp 101−102 °C. H NMR (CDCl₃, 600 MHz): δ 7.86 (1H, d, J =

2

10.9 Hz), 7.82 (1H, d, J = 7.7 Hz), 7.70 (1H, t, J = 7.7 Hz), 4.40 (2H,

150 MHz): δ 167.94, 162.18, 161.34, 133.83, 131.80 (q, J_FC= 32.7

q, J = 7.1 Hz), 2.80 (3H, s), 1.41 (3H, t, J = 7.1 Hz). ¹³C NMR

Hz), 130.03, 129.78, 127.48 (q, ³J_FC= 3.5 Hz), 123.81 (q, ³J_FC= 271

1

3

(CDCl₃, 150 MHz): δ 166.23, 161.97, 161.43, 160.16 (d, J_FC= 254

Hz), 123.69 (q, J_FC= 3.5 Hz), 122.84, 61.59, 17.65, 14.46. Anal.

4

3

Hz), 138.56 (d, J_FC= 8.2 Hz), 128.16 (q, J_FC= 3.9 Hz), 123.75,

122.39 (q, ¹J_FC= 271 Hz), 122.34 (d, ³J_FC= 3.9 Hz), 120.20 (dq, ²J_FC

= 33 Hz, ²J_FC= 12 Hz), 115.07 (d, ²J_FC= 23 Hz), 67.74, 17.62, 14.45.

¹⁹F NMR (CDCl₃, 376 MHz): δ −61.74 (3F, s), −113.07 (1F, s).

Calcd for C₁₄H₁₂F₃NO₂S: C, 53.33; H, 3.84; N, 4.44. Found: C,

53.05; H, 3.93; N, 4.54.

4-Methyl-2-[3-(triﬂuoromethyl)phenyl]-thiazole-5-methanol

(3c). By analogy to the procedure described in example 3a, ethyl ester

2c (2.053 g, 6.511 mmol) in solution with anhydrous THF (20 mL)

was treated with 2 M LiAlH₄solution in THF (3.3 mL, 6.60 mmol) to

give 3c as a yellow solid (1.273 g, 4.658 mmol, 71.5%). TLC R_f(25%

EtOAc/hexane) = 0.16. Mp 58−60 °C. ¹H NMR (CDCl₃, 600 MHz):

δ 8.19 (1H, s), 8.07 (1H, d, J = 7.8 Hz), 7.66 (1H, d, J = 7.8 Hz), 7.56

(1H, t, J = 7.8 Hz), 4.87 (2H, s), 2.49 (3H, s). ¹³C NMR (CDCl₃, 150

Anal. Calcd for C₁₄H₁₁F₄NO₂S: C, 50.45; H, 3.33; N, 4.20. Found: C,

50.47; H, 3.32; N, 4.24.

5-Hydroxymethyl-4-methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)-

phenyl]thiazole (3d). By analogy to the procedure described in

example 3a, ethyl ester 2d (4.640 g, 13.92 mmol) was dissolved in

THF (50 mL) and treated with a 2 M LiAlH₄solution in THF (7.2

mL, 14.4 mmol) to give 3d as a yellow solid (2.988 g, 10.26 mmol,

2

MHz): δ 164.43, 150.67, 134.31, 132.36, 131.61 (q, J_FC= 32.5 Hz),

1

3

1

73.7%). TLC R_f(25% EtOAc/hexane) = 0.21. Mp 158−159 °C. H

129.63, 126.52 (q, J_FC= 3.5 Hz), 123.92 (q, J_FC= 271 Hz), 123.19

3

NMR (CDCl₃, 600 MHz): δ 7.79 (1H, d, J = 11.2 Hz), 7.75 (1H, d, J

= 8.3 Hz), 7.65 (1H, t, J₁= 7.6 Hz, J₂= 7.7 Hz), 4.87 (2H, d, J = 5.2

Hz), 2.48 (3H, s), 1.93 (1H, t, J = 5.5 Hz). ¹³C NMR (CDCl₃, 150

(q, J_FC= 3.5 Hz), 56.95, 15.20. Anal. Calcd for C₁₂H₁₀F₃NOS: C,

52.74; H, 3.69; N, 5.13. Found: C, 52.55; H, 3.64; N, 5.21.

5-Chloromethyl-4-methyl-2-[3-(triﬂuoromethyl)phenyl]thiazole

(4c). By analogy to the procedure described in example 4a, alcohol 3c

MHz): δ 162.73, 160.16 (d, ¹J_FC= 255 Hz), 151.11, 139.20 (d, ⁴J_FC

=

7014

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

1

8.2 Hz), 133.37, 127.98 (m), 122.53 (q, J_FC= 271 Hz), 121.83 (d,

2-[4-Thio-[5-methylene-4-methyl-2-phenylthiazole]phenyl]-1,3-

dioxolane (9a). Acetal 8 (0.934 g, 4.077 mmol) was dissolved in

anhydrous THF (15 mL) under inert gas and cooled to −78 °C. A t-

BuLi solution in THF (1.7 M, 4.8 mL, 8.55 mmol) was slowly added,

and the reaction was stirred for 2 h. Sulfur (0.131 g, 4.078 mmol) was

suspended in anhydrous THF (9 mL) and added to the reaction at

−78 °C. The mixture was stirred at rt for 1.5 h and then cooled to 0

°C. Thiazole 4a (0.918 g, 4.103 mmol) was dissolved in THF (15

mL) and added to the reaction mixture. The reaction mixture was

allowed to warm to rt and stirred for an additional 2 h. The mixture

was quenched with aqueous NH₄Cl, and the organic phase was

separated. The aqueous phase was extracted using two portions of

EtOAc, and the organic phases were combined, washed with water

and brine, and dried over Na₂SO₄. After evaporation of the solvent, a

crude product 9a was isolated as a sticky yellow solid. Attempts to

purify the product using ﬂash chromatography were unsuccessful, and

the material was moved to the next step (0.941 g, 2.546 mmol,

62.5%). TLC R_f(25% EtOAc/hexane) = 0.23.

4-Thio-[5-methylene-4-methyl-2-phenylthiazole]benzaldehyde

(10a). Crude acetal 9a (0.476 g, 1.288 mmol) was dissolved in THF

(15 mL), and 3 N HCl (5 mL) was added. The reaction mixture was

stirred at rt for 2 h and was concentrated under reduced pressure. The

residue was neutralized with 1 N NaOH and extracted with EtOAc.

The organic extract was washed with water and brine and dried with

Na₂SO₄. After evaporation of the solvent, the product was puriﬁed

with ﬂash chromatography using a gradient of 10−20% EtOAc/

hexane. The puriﬁed product 10a was collected as a yellow solid

(0.302 g, 0.928 mmol, 72.0%). TLC R_f(25% EtOAc/hexane) = 0.35.

Mp 116−117 °C. ¹H NMR (CDCl₃, 600 MHz): δ 9.95 (1H, s), 7.86

(2H, m), 7.80 (2H, d, J = 8.3 Hz), 7.43 (2H, d, J = 8.3 Hz), 7.41 (3H,

m), 4.37 (2H, s), 2.43 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ

191.34, 165.94, 151.23, 144.55, 134.30, 133.51, 130.28, 130.17,

129.04, 128.34, 127.31, 126.42, 29.28, 15.34. Anal. Calcd for

C₁₈H₁₅NOS₂: C, 66.43; H, 4.65; N, 4.30. Found: C, 66.29; H,

4.53; N, 4.19.

Methyl 4-[Thio-[5-methylene-4-methyl-2-phenylthiazole]]-

cinnamate (11a). Aldehyde 10a (0.251 g, 0.771 mmol) and methyl

(triphenylphosphoranylidene)acetate (0.291 g, 0.870 mmol) were

dissolved in anhydrous THF (6 mL) under inert gas and stirred at 60

°C for 2 days. The reaction mixture was concentrated, and the

resulting residue was taken up into EtOAc, washed with water and

brine, and dried with Na₂SO₄. After evaporation of the solvent, the

product was puriﬁed with ﬂash chromatography using a gradient of

5−15% EtOAc/hexane. The product collected from the column was

recrystallized using 95% EtOH to give 11a as a light-yellow solid

(0.244 g, 0.640 mmol, 83.1%). TLC R_f(25% EtOAc/hexane) = 0.36.

Mp 102−103 °C. ¹H NMR (CDCl₃, 600 MHz): δ 7.87 (2H, m), 7.63

(1H, d, J = 16.0 Hz), 7.44 (2H, d, J = 8.2 Hz), 7.41 (3H, m), 7.34

(2H, d, J = 8.3 Hz), 6.41 (1H, d, J = 16.0 Hz), 4.27 (2H, s), 3.80 (3H,

s), 2.35 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 167.47, 143.99,

137.99, 133.19, 130.51, 130.32, 129.10, 128.86, 128.74, 128.48,

127.35, 126.62, 126.53, 118.09, 51.92, 30.43, 15.07. Anal. Calcd for

C₁₈H₁₅NOS₂: C, 66.11; H, 5.02; N, 3.67. Found: C, 66.05; H, 4.87;

N, 3.56.

³J_FC= 3.3 Hz), 119.17 (dq, J_FC= 33 Hz, J_FC= 12 Hz), 114.48 (d,

2

²J_FC= 23 Hz), 57.05, 15.27. ¹⁹F NMR (CDCl₃, 376 MHz): δ −61.81

(3F, s), −113.84 (1F, s). Anal. Calcd for C₁₂H₉F₄NOS: C, 49.49; H,

3.11; N, 4.81. Found: C, 49.52; H, 3.09; N, 4.79.

5-Chloromethyl-4-methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)-

phenyl]thiazole (4d). By analogy to the procedure described in

example 4a, alcohol 3d (1.502 g, 5.157 mmol) was treated with

methanesulfonyl chloride (0.60 mL, 7.74 mmol) and triethylamine

(1.40 mL, 10.314 mmol) in anhydrous DCM (50 mL) to give 4d as a

yellow oil (1.357 g, 4.382 mmol, 85.0%). TLC R_f(25% EtOAc/

1

hexane) = 0.68. H NMR (CDCl₃, 600 MHz): δ 7.75 (2H, m), 7.65

(1H, m), 4.79 (2H, s), 2.50 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ

1

4

163.38, 161.02 (d, J_FC= 255 Hz), 159.32, 153.52, 139.00 (d, J_FC

=

3

1

8.2 Hz), 129.83, 128.02 (q, J_FC= 4 Hz), 123.39 (q, J_FC= 270 Hz),

121.91 (m), 119.47 (dq, ²J_FC= 33 Hz, ²J_FC= 12 Hz), 114.60 (d, ²J_FC

= 22 Hz), 37.15, 15.21. ¹⁹F NMR (CDCl₃, 376 MHz): δ −61.82 (3F,

s), −113.62 (1F, s). Anal. Calcd for C₁₂H₈ClF₄NS: C, 46.54; H, 2.60;

N, 4.52. Found: C, 46.61; H, 2.37; N, 4.44.

Methyl 4-[Oxo-[5-methylene-4-methyl-2-[3-ﬂuoro-4-

(triﬂuoromethyl)phenyl]-thiazole]]cinnamate (6d). By analogy to

the procedure described in example 6a, chloromethyl 4d (0.494 g,

1.595 mmol) and methyl 4-hydroxycinnamate 5 (0.287 g, 1.611

mmol) were treated with cesium carbonate (1.037 g, 3.183 mmol) in

anhydrous acetonitrile (10 mL) to give 6d as a white solid (0.614 g,

1.360 mmol, 85.3%). TLC R_f(25% EtOAc/hexane) = 0.46. Mp 180

1

°C. H NMR (CDCl₃, 600 MHz): δ 7.78 (1H, d, J = 11.6 Hz), 7.75

(1H, d, J = 8.3 Hz), 7.66 (2H, m,), 7.51 (2H, d, J = 8.7 Hz), 6.98

(2H, d, J = 8.8 Hz), 6.34 (1H, d, J = 16.0 Hz), 5.23 (2H, s), 3.80 (3H,

s), 2.52 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 167.78, 163.52,

1

4

160.17 (d, J_FC= 255 Hz), 159.73, 152.66, 144.29, 139.12 (d, J_FC

=

1

7.5 Hz), 129.97, 128.43, 128.27, 128.00 (m), 122.51 (q, J_FC= 270

Hz), 121.89 (d, ³J_FC= 3 Hz), 119.48 (dq, ²J_FC= 33 Hz, ²J_FC= 12 Hz),

2

116.14, 115.36, 114.58 (d, J_FC= 22 Hz), 62.24, 51.82, 15.56. Anal.

Calcd for C₂₂H₁₇F₄NO₃S: C, 58.53; H, 3.80; N, 3.10. Found: C,

58.56; H, 3.83; N, 3.15.

4-[Oxo-[5-methylene-4-methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)-

phenyl]thiazole]]cinnamic Acid (7d). By analogy to the procedure

described in example 7a, methyl ester 6d (0.562 g, 1.245 mmol) was

treated with 3 N NaOH (2 mL) in 95% EtOH (20 mL) and THF (10

mL) to give 7d as a white solid (0.403 g, 0.921 mmol, 74.0%). TLC R_f

1

(50% EtOAc/hexane) = 0.29. Mp 192−194 °C. H NMR (acetone-

d₆, 600 MHz): δ 7.97 (2H, t, J = 11 Hz), 7.87 (1H, t, J = 7.9 Hz), 7.68

(2H, d, J = 8.7 Hz), 7.65 (1H, d, J = 16.0 Hz), 7.14 (2H, d, J = 8.7

Hz), 6.43 (1H, d, J = 16.0 Hz), 5.46 (2H, s), 2.53 (3H, s). ¹³C NMR

(acetone-d₆, 150 MHz): δ 167.89, 163.30, 161.58 (d, ¹J_FC= 254 Hz),

4

160.79, 153.41, 144.98, 140.57 (d, J_FC= 8.4 Hz), 130.77, 130.46,

129.14 (q, J_FC= 3.9 Hz), 128.91, 124.51 (q, J_FC= 270 Hz), 123.06

(d, J_FC= 4 Hz), 119.11 (dq, J_FC= 33 Hz, J_FC= 13 Hz), 117.10,

116.24, 114.89 (d, J_FC= 23 Hz), 62.86, 15.43. Anal. Calcd for

3

1

3

2

C₂₁H₁₅F₄NO₃S: C, 57.66; H, 3.46; N, 3.20. Found: C, 57.95; H, 3.48;

N, 3.20.

2-(4-Bromophenyl)-1,3-dioxolane (8). 4-Bromobenzaldehyde

(4.303 g, 23.26 mmol) was dissolved in anhydrous toluene (50

mL), and ethylene glycol (3.80 mL, 69.18 mmol) was added followed

by p-toluenesulfonic acid monohydrate (0.303 g, 1.59 mmol). The

mixture was heated to a vigorous reﬂux in a Dean−Stark apparatus

and stirred at that temperature for 2 days. The mixture was allowed to

cool to rt and was poured in a 75 mL saturated NaHCO₃solution and

extracted with 40 mL of toluene. The organic phase was collected and

washed with water twice, dried with Na₂SO₄, and concentrated under

reduced pressure. The resulting crude mixture was puriﬁed using ﬂash

chromatography to give 8 as a white solid (2.787 g, 12.17 mmol,

4-[Thio-[5-methylene-4-methyl-2-phenylthiazole]]cinnamic Acid

(12a). Methyl ester 11a (0.200 g, 0.524 mmol) was dissolved in 95%

EtOH (15 mL) using heat. The mixture was cooled to rt, and 3 N

NaOH was added. The reaction mixture was stirred for 24 h and was

concentrated under reduced pressure. The resulting residue was taken

up into EtOAc and washed with acidiﬁed water (HCl, pH 3) and

brine and dried with Na₂SO₄. After evaporation of the solvent, the

product was puriﬁed by ﬂash chromatography using a gradient of 25−

100% EtOAc/hexane. The product collected from the column was

recrystallized using 95% EtOH to give 12a as an oﬀ-white solid (0.089

g, 0.243 mmol, 46.4%). TLC R_f(50% EtOAc/hexane) = 0.33. Mp

1

52.3%). TLC R_f(25% EtOAc/hexane) = 0.55. Mp 34−35 °C. H

1

NMR (CDCl₃, 600 MHz): δ 7.52 (2H, d, J = 8.4 Hz), 7.36 (2H, d, J =

8.4 Hz), 5.78 (1H, s), 4.10 (2H, m), 4.02 (2H, m). ¹³C NMR

(CDCl₃, 150 MHz): δ 137.29, 131.63, 128.30, 123.33, 103.19, 65.43.

Anal. Calcd for C₉H₉O₂Br: C, 47.19; H, 3.96; N, 0.00. Found: C,

47.00; H, 3.97; N, 0.00.

192.5−193.5 °C. H NMR (CDCl₃, 600 MHz): δ 7.90 (2H, m),

7.65−7.61 (3H, m), 7.45 (5H, m), 6.51 (1H, d, J = 16.0 Hz), 4.52

(2H, s), 2.35 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 167.71,

165.36, 151.94, 144.51, 139.16, 134.56, 133.79, 130.76, 130.55,

129.86, 129.65, 129.56, 126.76, 119.24, 15.18. Anal. Calcd for

7015

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

C₁₈H₁₅NOS₂(0.2 H₂O mol per target): C, 65.37; H, 4.66; N, 3.81.

Found: C, 64.64; H, 4.57; N, 3.72.

g, 4.765 mmol). The crude product 9c was isolated as an orange oil,

and attempts to purify the product using ﬂash chromatography were

unsuccessful. The material was moved to the next step (0.604 g, 1.381

mmol, 29.3%). TLC R_f(25% EtOAc/hexane) = 0.31.

4-Thio-[5-methylene-4-methyl-2-[3-(triﬂuoromethyl)phenyl]-

thiazole]benzaldehyde (10c). By analogy to the procedure described

in example 10a, crude acetal 9c (0.604 g, 1.381 mmol) was treated

with 3 N HCl (5 mL), and the crude product was puriﬁed with

column chromatography using a gradient of 10−20% EtOAc/hexane

to give 10c as a yellow solid (0.275 g, 0.698 mmol, 50.6%). Elemental

analysis was completed after NMR experiments, leading to CDCl₃

contaminant. TLC R_f(25% EtOAc/hexane) = 0.40. Mp 89−91 °C.

¹H NMR (CDCl₃, 600 MHz): δ 9.95 (1H, s), 8.15 (1H, s), 8.02 (1H,

2-[4-Thio-[5-methylene-4-methyl-2-[4-(triﬂuoromethyl)phenyl]-

thiazole]phenyl]-1,3-dioxolane (9b). By analogy to the procedure

described in example 9a, acetal 8 (0.909 g, 3.968 mmol) was reacted

in a one-pot reaction involving a t-BuLi solution in THF (1.7 M, 4.7

mL, 7.99 mmol), sulfur (0.130 g, 4.053 mmol), and thiazole 4b (1.150

g, 3.942 mmol). The crude product was isolated as a sticky yellow

solid and was puriﬁed by ﬂash chromatography using a 5−20%

EtOAc/hexane gradient to give 9b as a yellow solid (0.658g, 1.504

mmol, 38.2%). TLC R_f(25% EtOAc/hexane) = 0.33. Mp 57−61 °C.

¹H NMR (CDCl₃, 600 MHz): δ 7.98 (2H, d, J = 8.1 Hz), 7.66 (2H, d,

J = 8.1 Hz), 7.42 (2H, d, J = 8.2 Hz), 7.38 (2H, d, J = 8.4 Hz), 5.78

(1H, s), 4.24 (2H, s), 4.12 (2H, m), 4.04 (2H, m), 2.31 (3H, m). ¹³C

NMR (CDCl₃, 150 MHz): δ 163.50, 151.43, 137.36, 136.75, 135.85,

131.48 (q, ²J_FC= 32 Hz), 131.10, 130.36, 127.40, 126.57, 126.01 (m),

124.05 (q, ¹J_FC= 270 Hz), 103.34, 65.46, 30.98, 15.12. Anal. Calcd for

C₂₁H₁₈NO₂S₂(with 0.4 H₂O mol per target): C, 56.72; H, 4.26; N,

3.15. Found: C, 56.50; H, 4.00; N, 3.32.

d, J = 7.8 Hz), 7.81 (2H, d, J = 8.5 Hz), 7.65 (1H, d, J = 7.8 Hz), 7.54

(1H, d, J = 7.8 Hz), 7.44 (2H, d, J = 8.3 Hz), 4.37 (2H, s), 2.44 (3H,

s). ¹³C NMR (CDCl₃, 150 MHz): δ 191.32, 163.96, 151.64, 144.24,

2

134.40, 134.27, 131.60 (q, J_FC= 33 Hz), 130.33, 129.60, 129.54,

128.65, 128.39, 126.54 (m), 123.90 (q, ¹J_FC= 271 Hz), 123.13, 29.22,

15.34. Anal. Calcd for C₁₉H₁₄F₃NOS₂(with 1.0 H₂O and 0.1 CDCl₃

mol per target): C, 54.17; H, 3.81; N, 3.31. Found: C, 54.11; H, 3.57;

N, 3.41.

4-Thio-[5-methylene-4-methyl-2-[4-(triﬂuoromethyl)phenyl]-

thiazole]benzaldehyde (10b). By analogy to the procedure described

in example 10a, acetal 9b (0.589 g, 1.346 mmol) was treated with 3 N

HCl (5 mL), and the crude product was puriﬁed with ﬂash

chromatography using a gradient of 5−20% EtOAc/hexane to give

10b as a yellow solid (0.297 g, 0.754 mmol, 56.0%). TLC R_f(25%

EtOAc/hexane) = 0.36. Mp 77−79 °C. ¹H NMR (CDCl₃, 600 MHz):

δ 9.94 (1H, s), 7.96 (2H, d, J = 8.1 Hz), 7.79 (2H, d, J = 8.5 Hz), 7.65

(2H, d, J = 8.2 Hz), 7.43 (2H, d, J = 8.3 Hz), 4.36 (2H, s), 2.43 (3H,

s). ¹³C NMR (CDCl₃, 150 MHz): δ 191.28, 163.85, 151.76, 144.21,

Methyl 4-[Thio-[5-methylene-4-methyl-2-[3-(triﬂuoromethyl)-

phenyl]-thiazole]]cinnamate (11c). By analogy to the procedure

described in example 11a, aldehyde 10c (0.239 g, 0.607 mmol) was

treated with methyl (triphenylphosphoranylidene)acetate (0.228 g,

0.682 mmol), and the crude product was puriﬁed with ﬂash

chromatography using 10% EtOAc/hexane to give 11c as a light-

yellow solid (0.130 g, 0.289 mmol, 47.6%). TLC R_f(25% EtOAc/

1

hexane) = 0.53. Mp 142−143 °C. H NMR (CDCl₃, 600 MHz): δ

2

8.14 (1H, s), 8.01 (1H, d, J = 7.9 Hz), 7.65 (2H, m), 7.53 (1H, t, J =

7.9 Hz), 7.45 (2H, d, J = 8.2 Hz), 7.34 (2H, d, J = 8.3 Hz), 6.41 (1H,

d, J = 16.0 Hz), 4.27 (2H, s), 3.80 (3H, s), 2.34 (3H, s). ¹³C NMR

(CDCl₃, 150 MHz): δ 167.45, 163.73, 151.48, 143.95, 137.88, 134.41,

136.58, 134.39, 131.65 (q, J_FC= 32 Hz), 130.30, 128.94, 128.37,

1

126.57, 126.04 (m), 123.99 (q, J_FC= 271 Hz), 29.19, 15.32. Anal.

Calcd for C₁₈H₁₅NOS₂: C, 58.00; H, 3.59; N, 3.56. Found: C, 57.83;

H, 3.67; N, 3.63.

2

133.25, 131.59 (q, J_FC= 33 Hz), 130.52, 129.57, 129.55, 129.53,

Methyl 4-[Thio-[5-methylene-4-methyl-2-[4-(triﬂuoromethyl)-

phenyl]-thiazole]]cinnamate (11b). By analogy to the procedure

described in example 11a, aldehyde 10b (0.248 g, 0.629 mmol) was

treated with methyl (triphenylphosphoranylidene)acetate (0.238 g,

0.712 mmol), and the crude product was puriﬁed with ﬂash

chromatography using a gradient of 5−15% EtOAc/hexane to give

11b as a light-yellow solid (0.201 g, 0.447 mmol, 71.1%). TLC R_f

1

128.75, 126.44 (m), 123.94 (q, J_FC= 271 Hz), 123.11 (m), 118.14,

51.92, 30.41, 15.23. Anal. Calcd for C₁₈H₁₅NOS₂: C, 58.78; H, 4.04;

N, 3.12. Found: C, 58.49; H, 3.90; N, 3.18.

4-Thio-[5-methylene-4-methyl-2-[3-(triﬂuoromethyl)phenyl]-

thiazole]cinnamic Acid (12c). By analogy to the procedure described

in example 12a, methyl ester 11c (0.120 g, 0.267 mmol) was treated

with 3 N NaOH (1 mL), and the crude product was puriﬁed with

ﬂash chromatography using 25% EtOAc/hexane supplemented with

dropwise acetic acid. The product was recrystallized using 95% EtOH

to give 12c as a light-yellow solid (0.082 g, 0.188 mmol, 70.7%). TLC

R_f(50% EtOAc/hexane) = 0.51. Mp 150−151 °C. ¹H NMR

(acetone-d₆, 600 MHz): δ 8.21 (1H, s), 8.13 (1H, d. J = 7.8 Hz), 7.78

(2H, d, J = 7.8 Hz), 7.71 (1H, t, J = 7.8 Hz), 7.64 (3H, m), 7.46 (2H,

d, J = 8.3 Hz), 6.52 (1H, d, J = 16.0 Hz), 4.54 (2H, s), 2.38 (3H, s).

¹³C NMR (acetone-d₆, 150 MHz): δ 167.64, 163.37, 152.37, 144.54,

1

(25% EtOAc/hexane) = 0.44. Mp 121−122 °C. H NMR (CDCl₃,

600 MHz): δ 7.98 (2H, d, J = 8.1 Hz), 7.65 (3H, m), 7.45 (2H, d, J =

8.2 Hz), 7.35 (2H, d, J = 8.3 Hz), 6.42 (1H, d, J = 16.0 Hz), 4.28 (2H,

s), 3.81 (3H, s), 2.36 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ

167.43, 163.68, 151.50, 143.91, 137.80, 136.62, 133.27, 131.61 (q,

²J_FC= 32.5 Hz), 130.53, 129.90, 128.74, 126.58, 126.03 (m), 124.01

1

(q, J_FC= 270 Hz), 118.16, 51.93, 30.39, 15.19. Anal. Calcd for

C₁₈H₁₅NOS₂: C, 58.78; H, 4.04; N, 3.12. Found: C, 58.98; H, 4.05;

N, 3.13.

2

138.95, 135.43, 133.89, 131.69 (q, J_FC= 33 Hz), 131.22, 131.08,

4-Thio-[5-methylene-4-methyl-2-[4-(triﬂuoromethyl)phenyl]-

thiazole]cinnamic Acid (12b). By analogy to the procedure described

in example 12a, methyl ester 11b (0.178 g, 0.396 mmol) was treated

with 3 N NaOH (1 mL), and the crude product was puriﬁed by ﬂash

chromatography and recrystallized with 95% EtOH to give the

product 12b as a light-yellow solid (0.119 g, 0.273 mmol, 68.9%).

TLC R_f(50% EtOAc/hexane) = 0.54. Mp 217 °C. ¹H NMR

(acetone-d₆, 600 MHz): δ 10.79 (1H, bs), 8.11 (2H, d. J = 8.1 Hz),

7.81 (2H, d, J = 8.1 Hz), 7.64 (3H, m), 7.47 (2H, d, J = 8.5 Hz), 6.52

(1H, d, J = 16.1 Hz), 4.55 (2H, s), 2.39 (3H, s). ¹³C NMR (acetone-

d₆, 150 MHz): δ 167.62, 163.35, 152.54, 144.54, 138.94, 137.98,

1

130.65, 130.54, 129.60, 127.05 (m), 125.00 (q, J_FC= 270 Hz),

122.92 (m), 119.25, 15.18. ¹⁹F NMR (acetone-d₆, 376 MHz): δ

−63.79 (3F, s). Anal. Calcd for C₂₁H₁₆F₃NO₂S₂: C, 57.92; H, 3.70; N,

3.22. Found: C, 57.81; H, 3.72; N, 3.23.

2-[4-Thio-[5-methylene-4-methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)-

phenyl]thiazole]phenyl-1,3-dioxolane (9d). By analogy to the

procedure described in example 9a, acetal 8 (0.915g, 3.99 mmol)

was reacted in a one-pot reaction involving a t-BuLi solution in THF

(1.7 M, 4.7 mL), sulfur (0.129g, 4.022 mmol), and thiazole 4d (0.959

g, 3.29 mmol). The crude product was puriﬁed with column

chromatography using a 0−20% EtOAc/hexane gradient to give the

product 9d as an oﬀ-white solid (0.354 g, 0.777 mmol, 19.5%). TLC

R_f(25% EtOAc/hexane) = 0.40. Mp 105−107 °C. ¹H NMR (CDCl₃,

600 MHz): δ 7.72 (1H, m), 7.69 (1H, m), 7.62 (1H, t, J = 7.5 Hz),

7.40 (2H, m), 7.37 (2H, m), 5.78 (1H, s), 4.23 (2H, s), 4.11 (2H, m),

4.03 (2H, m), 2.30 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 161.91,

160.14 (d, ¹J_FC= 255 Hz), 151.67, 139.21 (d, ⁴J_FC= 7.5 Hz), 137.47,

135.66, 131.31, 131.19, 127.90 (q, ³J_FC= 4.5 Hz), 127.45, 122.55 (q,

2

133.89, 131.60, 131.51 (q, J_FC= 32 Hz), 130.64, 129.60, 127.31,

3

1

126.88 (q, J_FC= 4.5 Hz), 125.13 (q, J_FC= 270 Hz), 119.25, 15.20.

¹⁹F NMR (acetone-d₆, 376 MHz): δ −63.67 (s, 3F). Anal. Calcd for

C₂₁H₁₆F₃NO₂S₂: C, 57.92; H, 3.70; N, 3.22. Found: C, 57.80; H,

3.66; N, 3.29.

2-[4-Thio-[5-methylene-4-methyl-2-[3-(triﬂuoromethyl)phenyl]-

thiazole]phenyl]-1,3-dioxolane (9c). By analogy to the procedure

described in example 9a, acetal 8 (1.079 g, 4.710 mmol) was reacted

in a one-pot reaction involving a t-BuLi solution in THF (1.7 M, 5.6

mL, 9.52 mmol), sulfur (0.176 g, 5.488 mmol), and thiazole 4c (1.390

2

¹J_FC= 270 Hz), 121.75 (d, ³J_FC= 4.5 Hz), 119.04 (dq, J_FC= 33 Hz,

²J_FC= 13 Hz), 114.47, 114.32, 103.33, 65.49, 30.96, 15.11. Anal.

7016

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Calcd for C₂₁H₁₇F₄NO₂S₂: C, 55.38; H, 3.76; N, 3.08. Found: C,

55.41; H, 3.65; N, 3.20.

C₁₆H₂₁NO₄: C, 65.96; H, 7.27; N, 4.81. Found: C, 65.81; H, 7.42;

N, 4.93.

4-Thio-[5-methylene-4-methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)-

phenyl]thiazole]benzaldehyde (10d). By analogy to the procedure

described in example 10a, acetal 9d (0.354g, 0.7770 mmol) was

treated with 3 N HCl (3 mL), and the crude product was puriﬁed

with ﬂash chromatography using a 5−20% EtOAc/hexane gradient to

give the desired product 10d as an oﬀ-white solid (0.244 g, 0.593

mmol, 76.3%). TLC R_f(25% EtOAc/hexane) = 0.33. Mp 122−123

°C. ¹H NMR (CDCl₃, 600 MHz): δ 9.91 (1H, s), 7.76 (2H, d, J = 8.4

Hz), 7.69 (1H, d, J = 11.2 Hz), 7.65 (1H, d, J = 8.3 Hz), 7.59 (1H, t, J

= 7.6 Hz), 7.39 (2H, d, J = 8.4 Hz), 4.32 (2H, s), 2.39 (3H, s), 1.50

(3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 191.28, 162.30, 160.15 (d,

Ethyl 4-[N-(tert-Butoxycarbonyl)-N-[(5-methylene-4-methyl-2-

phenyl)thiazole]]aminocinnamate (14a). Ethyl 4-(N-tert-

butoxycarbonyl)aminocinnamate 13 (1.336 g, 4.586 mmol), thiazole

4a (1.027 g, 4.590 mmol), and NaI (0.690 g, 4.603 mmol) were

dissolved in anhydrous DMF (30 mL) under inert gas and cooled in

an ice bath. NaH (60% dispersion in mineral oil, 0.281 g, 7.025

mmol) was carefully added to the reaction mixture by brieﬂy exposing

the system to air. The reaction mixture was stirred for 3 h at room

temperature and quenched with a 50% dilution of sat. NaHCO₃. The

mixture was extracted with three portions of ether that were

combined, washed with brine, dried with Na₂SO₄, and concentrated.

The crude material was puriﬁed by column chromatography using a

gradient of 100% hexane to 25% EtOAc/hexane to give the desired

product 14a as a yellow oil (1.488 g, 3.109 mmol, 67.8%). TLC R_f

4

¹J_FC= 255 Hz), 152.00, 143.99, 139.02 (d, J_FC= 7.5 Hz), 134.50,

3

130.36, 129.92, 128.48, 127.97 (dq, J_FC= 4.5 Hz, J_FC= 3.0 Hz),

122.50 (q, ¹J_FC= 270 Hz), 121.77, 119.30 (dq, ²J_FC= 33 Hz, ²J_FC= 13

Hz), 114.52, 114.36, 29.19, 15.33. Anal. Calcd for C₁₉H₁₃F₄NOS₂: C,

55.47; H, 3.18; N, 3.40. Found: C, 55.64; H, 3.35; N, 3.42.

1

(25% EtOAc/hexane) = 0.39. H NMR (CDCl₃, 600 MHz): δ 7.87

(2H, m), 7.64 (1H, d, J = 16.0 Hz), 7.46 (2H, d, J = 8.5 Hz), 7.39

(3H, m), 7.15 (2H, bd, J = 8.0 Hz), 6.39 (1H, d, J = 16.0 Hz), 4.94

(2H, s), 4.25 (2H, q, J = 7.1 Hz), 2.19 (3H, s), 1.47 (9H, s), 1.32

(3H, t, J = 7.1 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 167.00, 166.24,

154.12, 150.93, 143.69, 143.49, 133.67, 132.70, 130.02, 128.99,

128.71, 128.16, 127.49, 126.41, 118.57, 81.61, 60.65, 45.87, 28.42,

15.13, 14.43. Anal. Calcd for C₂₇H₃₀N₂O₄S (with 0.2 H₂O mol per

target): C, 67.25; H, 6.35; N, 5.81. Found: C, 67.02; H, 6.55; N, 5.58.

Ethyl 4-[N-[5-Methylene-4-methyl-2-phenylthiazole]]-

aminocinnamate (15a). BOC-protected amine 14a (1.354 g, 2.749

mmol) was dissolved in anhydrous DCM (15 mL) and cooled in an

ice bath. Triﬂuoroacetic acid (3 mL) was slowly added, and the

mixture was stirred at room temperature for 3 h. The reaction mixture

was then washed with a chilled saturated NaHCO₃solution, water,

and brine and dried with Na₂SO₄. After ﬁltration and evaporation of

the solvent, the crude product was puriﬁed with ﬂash chromatography

using a gradient of 10−25% EtOAc/hexane. The puriﬁed product 15a

was collected as a yellow solid (0.810 g, 2.140 mmol, 77.8%). TLC R_f

Methyl 4-[Thio-[5-methylene-4-methyl-2-[3-ﬂuoro-4-

(triﬂuoromethyl)phenyl]-thiazole]]cinnamate (11d). By analogy to

the procedure described in example 11a, aldehyde 10d (0.244 g,

0.594 mmol) was treated with methyl (triphenylphosphoranylidene)-

acetate (0.216 g, 0.646 mmol), and the crude product was puriﬁed

with ﬂash chromatography using a 5−20% gradient to give the desired

product 11d as a white solid (0.056 g, 0.12 mmol, 20.2%). TLC R_f

1

(25% EtOAc/hexane) = 0.40. Mp 121−123 °C. H NMR (CDCl₃,

600 MHz): δ 7.72 (1H, d, J = 11.2 Hz), 7.69 (1H, d, J = 8.3 Hz), 7.63

(2H, m), 7.44 (2H, d, J = 8.4 Hz), 7.33 (2H, d, J = 8.3 Hz), 6.40 (1H,

d, J = 16.0 Hz), 4.27 (2H, s), 3.80 (3H, s), 2.34 (3H, s). ¹³C NMR

1

(CDCl₃, 150 MHz): δ 167.41, 162.08, 160.14 (d, J_FC= 255 Hz),

3

151.77, 143.87, 139.08 (d, J_FC= 9.0 Hz), 137.61, 133.37, 130.84,

130.59, 128.77, 127.94 (d, ³J_FC= 4.0 Hz), 122.52 (q, ¹J_FC= 271 Hz),

3

2

121.75 (d, J_FC= 4.5 Hz), 119.15 (dq, J_FC= 33 Hz, J_FC= 13 Hz),

118.24, 114.48, 114.33, 51.94, 30.35, 15.18. Anal. Calcd for

C₂₂H₁₇F₄NO₂S₂: C, 56.62; H, 3.67; N, 3.00. Found: C, 56.54; H,

3.73; N, 3.02.

1

(25% EtOAc/hexane) = 0.23. Mp 112−113 °C. H NMR (CDCl₃,

4-Thio-[5-methylene-4-methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)-

phenyl]thiazole]cinnamic Acid (12d). Methyl ester 11d (0.053 g,

0.113 mmol) was treated with 3 N NaOH (1 mL), and the crude

product was puriﬁed with ﬂash chromatography using a 25% EtOAc/

hexane mobile phase followed by recrystallization of the column

product with 95% ethanol to give 12d as a white solid (0.020 g, 0.044

mmol, 39.0%). TLC R_f(25% EtOAc/hexane) = 0.20. Mp 180−182

°C. ¹H NMR (CDCl₃, 600 MHz): δ 7.88 (2H, m), 7.83 (1H, m), 7.64

(3H, m), 7.47 (2H, m), 6.52 (1H, d, J = 15.9 Hz), 4.56 (2H, s), 2.39

(3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 167.58, 161.85, 160.73 (d,

600 MHz): δ 7.86 (2H, m), 7.60 (1H, d, J = 15.9 Hz), 7.39 (5H, m),

6.63 (2H, d, J = 8.6 Hz), 6.23 (1H, d, J = 15.8 Hz), 4.47 (2H, s), 4.37

(1H, bs), 4.23 (2H, q, J = 7.1 Hz), 2.48 (3H, s), 1.32 (3H, t, J = 7.1

Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 167.83, 165.86, 150.16, 149.10,

144.89, 133.63, 130.02, 129.49, 129.01, 126.38, 124.65, 113.75,

113.00, 60.30, 40.19, 15.44, 14.50. Anal. Calcd for C₂₂H₂₂N₂O₂S: C,

69.81; H, 5.86; N, 7.40. Found: C, 69.57; H, 5.88; N, 7.53.

4-[N-[5-Methylene-4-methyl-2-phenylthiazole]]aminocinnamic

Acid (16a). Methyl ester 15a (0.721 g, 1.905 mmol) was dissolved in

a mixture of THF (20 mL) and 95% EtOH (10 mL). A 3 N NaOH

solution (3 mL) was added, and the reaction mixture was stirred for

24 h. The reaction was found to be incomplete, as determined by

TLC, and an additional 3 N NaOH (3 mL) was added. The mixture

was stirred for an additional 24 h (48 h total) and was neutralized and

then acidiﬁed with 1 N HCl (to pH 3). Organic solvents were

removed under reduced pressure, and the residue was extracted with

EtOAc. The organic extract was washed with brine, dried with

Na₂SO₄, and concentrated. The crude material was puriﬁed by ﬂash

chromatography using a 25−30% EtOAc/hexane gradient supple-

mented with dropwise amounts of acetic acid to give the desired

product 16a as a yellow solid (0.308 g, 0.879 mmol, 46.1%). TLC R_f

4

¹J_FC= 254 Hz), 152.76, 144.53, 140.61 (d, J_FC= 8.8 Hz), 138.78,

133.95, 132.75, 130.67, 129.63, 129.08 (q, ³J_FC= 4.5 Hz), 123.63 (q,

¹J_FC= 269 Hz), 122.87 (d, ³J_FC= 3 Hz), 119.26, 118.84 (dq, ²J_FC= 33

2

Hz, J_FC= 12 Hz), 114.66, 114.51, 30.34, 15.19. Anal. Calcd for

C₂₁H₁₅F₄NO₂S₂: C, 55.62; H, 3.33; N, 3.09. Found: C, 55.35; H,

3.44; N, 3.04. LC-MS/MS analysis shows >98% chromatographic

purity (total ion current) with retention time = 6.13 min and with m/

z = 454 (for M + 1).

Ethyl 4-(N-tert-Butoxycarbonyl)aminocinnamate (13). Ethyl 4-

aminocinnamate (1.270 g, 6.641 mmol) and di-tert-butyl dicarbonate

(1.480 g, 6.781 mmol) were dissolved in THF (20 mL) under inert

gas and heated to reﬂux at 65 °C overnight. The mixture was cooled

to room temperature and concentrated under reduced pressure. The

resulting residue was taken up into EtOAc and washed with sat.

NaHCO₃, water, and brine, dried with Na₂SO₄, and concentrated.

The crude product was puriﬁed with ﬂash chromatography using a

10−20% EtOAc/hexane mobile-phase gradient to give the desired

product 13 a light-orange solid (1.200 g, 4.122 mmol, 62.1%). TLC R_f

1

(50% EtOAc/hexane) = 0.15. Mp 205−207 °C. H NMR (acetone-

d₆, 600 MHz): δ 10.41 (1H, bs), 7.90 (2H, m), 7.56 (2H, d, J = 15.9

Hz), 7.45 (5H, m), 6.76 (2H, d, J = 8.6 Hz), 6.24 (1H, d, J = 15.9

Hz), 6.16 (1H, bs), 4.61 (2H, m), 2.47 (3H, s). ¹³C NMR (acetone-

d₆, 150 MHz): δ 168.36, 165.08, 151.07, 150.59, 146.12, 134.77,

132.01, 130.75, 130.58, 129.83, 126.71, 124.33, 113.57, 113.46, 40.12,

15.40. Anal. Calcd for C₂₀H₁₈N₂O₂S: C, 68.55; H, 5.18; N, 7.99.

Found: C, 68.46; H, 5.20; N, 8.06.

1

(25% EtOAc/hexane) = 0.46. Mp 93−96 °C. H NMR (CDCl₃, 600

MHz): δ 7.63 (1H, d, J = 16.0 Hz), 7.46 (2H, d, J = 8.6 Hz), 7.39

(2H, d, J = 8.5 Hz), 6.58 (1H, bs), 6.34 (1H, d, J = 16.0 Hz), 4.25

(2H, q, J = 7.1 Hz), 1.52 (9H, s), 1.33 (3H, t, J = 7.1 Hz). ¹³C NMR

(CDCl₃, 150 MHz): δ 167.39, 152.46, 144.20, 140.40, 129.27, 129.22,

118.42, 116.65, 81.19, 60.54, 28.43, 14.49. Anal. Calcd for

Ethyl 4-[N-(tert-Butoxycarbonyl)-N-[(5-methylene-4-methyl-2-

[4-(triﬂuoromethyl)phenyl]thiazole]]aminocinnamate (14b). By

analogy to the procedure described in example 14a, ethyl ester 13

(1.342 g, 4.606 mmol) and thiazole 4b (1.345 g, 4.611 mmol) were

treated with NaI (0.693 g, 4.623 mmol) and NaH (60% dispersion in

7017

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

mineral oil, 0.288 g, 7.20 mmol). The crude product was puriﬁed by

ﬂash chromatography using a gradient of 100% hexane to 25%

EtOAc/hexane to give the desired product 14b as a yellow solid

(1.302 g, 2.384 mmol, 51.8%). TLC R_f(25% EtOAc/hexane) = 0.45.

EtOAc/hexane gradient to give 15c as a light-yellow solid (0.811 g,

1.817 mmol, 76.4%). TLC R_f(25% EtOAc/hexane) = 0.23. Mp 133−

135 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.14 (1H, s), 8.02 (1H, d, J

= 7.9 Hz), 7.63 (1H, d, J = 7.8 Hz), 7.60 (1H, d, J = 15.9 Hz), 7.52

(1H, m), 7.39 (2H, d, J = 8.6 Hz), 6.64 (2H, d, J = 8.6 Hz), 6.24 (2H,

d, J = 15.9 Hz), 4.51 (2H, s), 4.23 (2H, q, J = 7.1 Hz), 2.51 (3H, s),

1.32 (3H, t, J = 7.1 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 167.80,

1

Mp 51−54 °C. H NMR (CDCl₃, 600 MHz): δ 7.99 (2H, d, J = 8.2

Hz), 7.65 (3H, m), 7.48 (2H, d, J = 8.5 Hz), 7.16 (2H, bd, J = 8.1

Hz), 6.40 (1H, d, J = 16.0 Hz), 4.95 (2H, s), 4.26 (2H, q, J = 7.1 Hz),

2.21 (3H, s), 1.47 (9H, s), 1.33 (3H, t, J = 7.1 Hz). ¹³C NMR

(CDCl₃, 150 MHz): δ 167.00, 164.25, 154.13, 151.53, 143.62, 143.41,

2

163.89, 150.44, 148.92, 134.40, 131.57 (q, J_FC= 32 Hz), 130.94,

1

130.05, 129.57, 129.52, 126.42 (m), 124.90, 123.92 (q, J_FC= 271

2

136.80, 132.81, 131.53 (q, J_FC= 33 Hz), 129.60, 128.76, 127.43,

Hz), 123.10 (m), 113.97, 113.08, 60.33, 40.29, 15.46, 14.51. Anal.

Calcd for C₂₃H₂₁F₃N₂O₂S: C, 61.87; H, 4.74; N, 6.27. Found: C,

61.87; H, 4.79; N, 6.26.

1

126.60, 126.01 (m), 124.03 (q, J_FC= 270 Hz), 118.69, 81.75, 60.70,

45.87, 28.41, 15.15, 14.44. Anal. Calcd for C₂₈H₂₉F₃N₂O₄S: C, 61.53;

H, 5.35; N, 5.13. Found: C, 61.63; H, 5.53; N, 5.27.

4-[N-[5-Methylene-4-methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)-

phenyl]thiazole]]aminocinnamic Acid (16c). By analogy to the

procedure described in example 16a, ethyl ester 15c (0.629 g, 1.409

mmol) was treated with 3 N NaOH (2 mL), and the crude product

was puriﬁed using ﬂash chromatography with a 25% EtOAc/hexane

mobile phase that was supplemented with dropwise amounts of glacial

acetic acid. A yellow−white solid was collected (0.474 g) and

recrystallized using 95% ethanol to give the desired product 16c as a

light-yellow solid (0.122 g, 0.292 mmol, 20.7%). TLC R_f(50%

Ethyl 4-[N-[5-Methylene-4-methyl-2-[4-(triﬂuoromethyl)phenyl]-

thiazole]]aminocinnamate (15b). By analogy to the procedure

described in example 15a, BOC-protected amine 14b (1.140 g, 2.085

mmol) was treated with triﬂuoroacetic acid (3 mL), and the crude

product was puriﬁed with ﬂash chromatography using a gradient of

10−25% EtOAc/hexane. The puriﬁed product 15b was collected as

an ivory solid (0.793 g, 1.776 mmol, 85.2%). TLC R_f(25% EtOAc/

1

hexane) = 0.20. Mp 120−122 °C. H NMR (CDCl₃, 600 MHz): δ

1

7.98 (2H, d, J = 8.1 Hz), 7.65 (2H, d, J = 8.2 Hz), 7.60 (1H, d, J =

15.9 Hz), 7.39 (2H, d, J = 8.6 Hz), 6.65 (2H, d, J = 8.6 Hz), 6.24 (1H,

d, J = 15.9 Hz), 4.52 (2H, s), 4.24 (2H, q, J = 7.1 Hz), 2.51 (3H, s),

1.32 (3H, t, J = 7.1 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 167.79,

EtOAc/hexane) = 0.33. Mp 171−173 °C. H NMR (acetone-d₆, 600

MHz): δ 8.22 (1H, s), 8.13 (1H, d, J = 7.9 Hz), 7.77 (1H, d, J = 7.8

Hz), 7.70 (1H, t, J = 7.8 Hz), 7.56 (1H, d, J = 15.8 Hz), 7.47 (2H, d, J

= 8.6 Hz), 6.76 (2H, d, J = 8.7 Hz), 6.24 (2H, d, J = 15.8 Hz), 4.64

(2H, s), 2.50 (3H, s). ¹³C NMR (acetone-d₆, 150 MHz): δ 168.41,

2

163.90, 150.59, 148.79, 144.77, 136.67, 131.61 (q, J_FC= 32.5 Hz),

1

2

131.22, 130.05, 126.60, 126.05 (m), 125.05, 124.02 (q, J_FC= 270

163.08, 150.95, 150.93, 146.09, 135.61, 133.69, 131.67 (q, J_FC= 32

1

Hz), 114.08, 113.18, 60.36, 40.37, 15.47, 14.53. Anal. Calcd for

C₂₃H₂₁F₃N₂O₂S: C, 61.87; H, 4.74; N, 6.27. Found: C, 61.73; H,

4.77; N, 6.37.

Hz), 131.05, 130.76, 130.47, 126.88 (m), 125.02 (q, J_FC= 270 Hz),

124.45, 113.61, 113.58, 40.17, 15.38. ¹⁹F NMR (acetone-d₆, 150

MHz): δ −63.76 (3F, s). Anal. Calcd for C₂₁H₁₇F₃N₂O₂S: C, 60.28;

H, 4.10; N, 6.69. Found: C, 60.16; H, 4.15; N, 6.65.

4-[N-[5-Methylene-4-methyl-2-[4-(triﬂuoromethyl)phenyl]-

thiazole]]aminocinnamic Acid (16b). By analogy to the procedure

described in example 16a, methyl ester 15b (0.595 g, 1.333 mmol)

was treated with 3 N NaOH (2 mL), and the crude material was

puriﬁed by ﬂash chromatography using a 25−30% EtOAc/hexane

gradient supplemented with dropwise amounts of acetic acid to give

the desired product 16b as a yellow solid (0.153 g, 0.365 mmol,

Ethyl 4-(N-tert-Butoxycarbonyl)-[N-[5-methylene-4-methyl-2-[3-

ﬂuoro-4-(triﬂuoromethyl)phenyl]thiazole]]cinnamate (14d). By

analogy to the procedure described in example 14a, ethyl ester 13

(1.274 g, 4.373 mmol) and thiazole 4d (1.314 g, 4.250 mmol) were

treated with NaI (0.677 g, 4.520 mmol) and NaH (60% dispersion in

oil, 0.222 g, 5.55 mmol) to give a crude yellow−orange oil that was

puriﬁed using ﬂash chromatography with a 100% hexane to 15%

EtOAc/hexane gradient, yielding the desired product 14d as a yellow

solid (1.075 g, 1.90 mmol, 44.8%). TLC R_f(25% EtOAc/hexane) =

1

27.4%). TLC R_f(50% EtOAc/hexane) = 0.35. Mp 224−226 °C. H

NMR (acetone-d₆, 600 MHz): δ 10.42 (1H, bs), 8.12 (2H, d, J = 8.2

Hz), 7.79 (2H, d, J = 8.3 Hz), 7.56 (1H, d, J = 15.8), 7.48 (2H, d, J =

8.6 Hz), 6.77 (2H, d, J = 8.7 Hz), 6.24 (2H, m), 4.65 (2H, s), 2.50

(3H, s). ¹³C NMR (acetone-d₆, 150 MHz): δ 168.35, 163.07, 151.11,

1

0.43. Mp 64−66 °C. H NMR (CDCl₃, 600 MHz): δ 7.73 (2H, m),

7.64 (2H, m), 7.49 (2H, d, J = 8.4 Hz), 7.16 (2H, d, J = 8.2 Hz), 6.41

(1H, d, J = 15.9 Hz), 4.97 (2H, s), 4.27 (2H, q, J = 7.14 Hz), 2.22

(3H, s), 1.47 (9H, s), 1.34 (3H, t, J = 7.14 Hz). ¹³C NMR (CDCl₃,

2

150.95, 146.07, 138.17, 134.09, 131.35 (q, J_FC= 31.8 Hz), 130.76,

1

127.22, 126.85 (m), 125.15 (q, J_FC= 270 Hz), 124.46, 113.61,

113.58, 40.17, 15.40. ¹⁹F NMR (acetone-d₆, 376 MHz): δ −63.64 (s,

3F). Anal. Calcd for C₂₁H₁₇F₃N₂O₂S (with 0.4 mol H₂O per target):

C, 59.26; H, 4.22; N, 6.58. Found: C, 59.41; H, 4.58; N, 6.36.

Ethyl 4-(N-tert-Butoxycarbonyl)-[N-[5-methylene-4-methyl-2-[3-

(triﬂuoromethyl)phenyl]thiazole]]aminocinnamate (14c). By anal-

ogy to the procedure described in example 14a, ethyl ester 14 (1.313

g, 4.507 mmol) and thiazole 4c (1.315 g, 4.509 mmol) were treated

with NaI (0.680 g, 4.538 mmol) and NaH (60% dispersion in mineral

oil, 0.277 g, 6.925 mmol). The crude product was puriﬁed with ﬂash

chromatography using a 100% hexane to 15% EtOAc/hexane gradient

to give 14c as a viscous yellow oil (1.621 g, 2.966 mmol, 65.8%). TLC

R_f(25% EtOAc/hexane) = 0.35. ¹H NMR (CDCl₃, 600 MHz): δ 8.18

(1H, s), 8.05 (1H, d, J = 7.8 Hz), 7.65 (2H, m), 7.55 (1H, t, J = 7.8

Hz), 7.49 (2H, d, J = 8.5 Hz), 7.17 (2H, d, J = 8.1 Hz), 6.41 (1H, d, J

= 16.0 Hz), 4.98 (2H, s), 4.28 (2H, q, J = 7.1 Hz), 2.23 (3H, s), 1.49

(9H, s), 1.35 (3H, t, J = 7.1 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ

167.00, 164.36, 154.15, 151.38, 143.63, 143.41, 134.40, 132.82,

131.56 (q, ²J_FC= 32 Hz), 129.62, 129.56, 128.77, 126.44 (m), 123.93

1

150 MHz): δ 166.85, 162.60, 159.52 (d, J_FC= 254 Hz), 153.99,

1

151.66, 143.45, 143.22, 139.10 (d, J_FC= 8.4 Hz), 132.75, 130.29,

1

128.67, 127.79, 127.78 (d, J_FC= 3.3 Hz), 127.28, 122.42 (q, J_FC

=

2

269 Hz) 121.67, 119.00 (dq, J_FC= 33 Hz,, J_FC= 13 Hz) 118.65,

114.32 (d, J_FC= 22 Hz), 81.72, 60.60, 28.30, 15.02, 14.33. Anal.

1

Calcd for C₂₈H₂₈F₄N₂O₄S: C, 59.57; H, 5.00; N, 4.96. Found: C,

59.30; H, 5.13; N, 4.95.

Ethyl 4-[N-[5-Methylene-4-methyl-2-[3-ﬂuoro-4-

(triﬂuoromethyl)phenyl]thiazole]] Aminocinnamate (15d). By

analogy to the procedure described in example 15a, BOC-protected

amine 14d (0.684 g, 1.473 mmol) was treated with triﬂuoroacetic acid

(5 mL), and the solid yellow crude product was puriﬁed using ﬂash

chromatography with a 100% hexane to 35% EtOAc/hexane gradient,

yielding the desired product 15d as a yellow solid (0.296 g, 0.637

mmol, 43.3%). TLC R_f(25% EtOAc/hexane) = 0.20. Mp 138−140

1

°C. H NMR (CDCl₃, 600 MHz): δ 7.77 (1H, d, J = 11.2 Hz), 7.72

(1H, d, J = 8.3 Hz), 7.63 (2H, m), 7.42 (2H, d, J = 8.6 Hz), 6.68 (2H,

d, J = 8.6 Hz), 6.28 (1H, d, J = 15.9 Hz), 4.51 (2H, s), 4.26 (2H, m),

2.52 (3H, s), 1.34 (3H, t, J = 7.14 Hz). ¹³C NMR (CDCl₃, 150

1

(q, J_FC= 271 Hz), 123.15 (m), 118.70, 81.75, 60.70, 45.85, 28.43,

15.13, 14.45. Anal. Calcd for C₂₈H₂₉F₃N₂O₄S: C, 61.53; H, 5.35; N,

5.13. Found: C, 61.39; H, 5.48; N, 5.01.

Ethyl 4-[N-[5-Methylene-4-methyl-2-[3-(triﬂuoromethyl)phenyl]-

thiazole]]aminocinnamate (15c). By analogy to the procedure

described in example 15a, BOC-protected amine 14c (1.30 g, 2.378

mmol) was treated with triﬂuoroacetic acid (3 mL), and the crude

material was puriﬁed using ﬂash chromatography and a 10% to 25%

1

MHz): δ 169.86, 162.15, 160.01 (d, J_FC= 255 Hz), 150.74, 148.53,

1

144.60, 139.8 (d, J_FC= 8.3 Hz), 132.02, 129.93, 129.21, 127.80 (d,

¹J_FC= 4.2 Hz), 125.02, 122.39 (q, ¹J_FC= 270 Hz), 121.62, 118.97 (dq,

²J_FC= 33 Hz, ²J_FC= 13 Hz), 117.96, 114.25 (d, ¹J_FC= 22 Hz), 60.25,

40.28, 15.36, 14.39. Anal. Calcd for C₂₃H₂₀F₄N₂O₂S: C, 59.48; H,

4.34; N, 6.03. Found: C, 59.48; H, 4.24; N, 6.00.

7018

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

4-[N-[5-Methylene-4-methyl-2-[3-ﬂuoro-4-(triﬂuoromethyl)-

phenyl]thiazole]]aminocinnamic Acid (16d). By analogy to the

procedure described in example 16a, ethyl ester 15d (0.098 g, 0.211

mmol) was treated with 3 N NaOH (1.0 mL), and the crude material

was puriﬁed by ﬂash chromatography using a 25−40% EtOAc/hexane

gradient supplemented with dropwise amounts of acetic acid to give

the desired product as a yellow solid (90.6 mg). The product was

recrystallized using 95% ethanol to give 16d as a cream-colored

crystalline solid (0.024 g, 0.055 mmol, 26.1%). TLC R_f(50% EtOAc/

114.22, 113.51, 46.95, 15.65. Anal. Calcd for C₃₃H₂₅F₆N₃O₂S₂: C,

58.83; H, 3.74; N, 6.24. Found: C, 58.57; H, 3.88; N, 5.96.

Ethyl 4-[N-(tert-Butoxycarbonyl)methyl]aminocinnamate (20).

Ethyl 4-(N-tert-butoxycarbonyl)aminocinnamate 13 (0.935 g, 3.211

mmol) was dissolved in anhydrous DMF (17 mL) under inert gas and

cooled in an ice bath. NaH (60% dispersion in oil, 0.198 g, 4.95

mmol) was added, and the mixture was stirred for 60 min followed by

the addition of iodomethane (0.60 mL, 9.63 mmol). The reaction

mixture was stirred overnight at room temperature and quenched with

a 50% aqueous solution of NaHCO₃and extracted twice with ether.

The ether extracts were combined and washed with brine, dried with

Na₂SO₄, and concentrated. The crude material was puriﬁed by ﬂash

chromatography using a 10% EtOAc/hexane mobile phase, giving the

product 20 as an oﬀ-white solid (0.897 g, 2.937 mmol, 91.5%). TLC

1

hexane) = 0.30. Mp 213−214 °C. H NMR (CDCl₃, 600 MHz): δ

7.89 (2H, d, J = 9.8 Hz), 7.82 (1H, t, J = 7.71 Hz), 7.57 (1H, d, J =

15.9 Hz), 7.48 (2H, d, J = 8.6 Hz), 6.76 (2H, d, J = 8.6 Hz), 6.24 (1H,

d, J = 15.9 Hz), 4.66 (2H, s), 2.51 (3H, s). ¹³C NMR (CDCl₃, 150

MHz): δ 168.34, 161.56, 159.88, 151.28, 150.83, 146.08, 140.81 (d,

1

3

⁴J_FC= 8.5 Hz), 135.33, 134.01, 130.77, 129.03 (d, J_FC= 4.1 Hz),

R_f(25% EtOAc/hexane) = 0.53. Mp 44−45 °C. H NMR (CDCl₃,

124.52, 123.64 (q, ¹J_FC= 270 Hz), 122.81, 122.79 118.66 (dq, ²J_FC

33 Hz, J_FC= 13 Hz), 114.56, 114.41, 113.61, 40.13, 15.39. Anal.

Calcd for C₂₁H₁₆F₄N₂O₂S: C, 57.79; H, 3.70; N, 6.42. Found: C,

57.74; H, 3.72; N, 6.28.

=

600 MHz): δ 7.65 (1H, d, J = 16.0 Hz), 7.47 (2H, d, J = 8.5 Hz), 7.27

(2H, d, J = 8.5 Hz), 6.38 (1H, d, J = 16.0 Hz), 4.25 (2H, q, J = 7.1

Hz), 3.27 (3H, s), 1.46 (9H, s), 1.33 (3H, t, J = 7.1 Hz). ¹³C NMR

(CDCl₃, 150 MHz): δ 167.19, 154.48, 145.60, 143.97, 131.21, 128.40,

125.34, 117.84, 80.97, 60.62, 37.10, 28.43, 14.46. Anal. Calcd for

C₁₆H₂₁NO₄: C, 66.86; H, 7.59; N, 4.59. Found: C, 66.60; H, 7.52; N,

4.45.

2

4-[N-(tert-Butoxycarbonyl)-N-[5-methylene-4-methyl-2-[3-ﬂuo-

ro-4-(triﬂuoromethyl)phenyl]thiazole]]aminocinnamic Acid. (17).

By analogy to the procedure described in example 16a, ethyl ester 14d

(0.206 g, 0.365 mmol) was treated with 3 N NaOH (1.0 mL), and the

crude product was puriﬁed by ﬂash chromatography using a 20−35%

EtOAc/hexane gradient supplemented with dropwise amounts of

acetic acid to give the desired product as a white solid. The solid was

recrystallized with 95% EtOH giving 17 as a crystalline white solid

product (0.107 g, 0.199 mmol, 54.6%). TLC R_f(50% EtOAc/hexane)

Ethyl 4-(N-Methyl)aminocinnamate (21). Ethyl 4-[N-(tert-

butoxycarbonyl)methyl]aminocinnamate 20 (0.775 g, 2.538 mmol)

was dissolved in anhydrous DCM (20 mL) and cooled in an ice bath.

Triﬂuoroacetic acid (5 mL) was slowly added, and the mixture was

allowed to warm to room temperature. The mixture was stirred for 90

min, and the solvent was removed under reduced pressure. The

residue was taken up into EtOAc and washed with chilled saturated

NaHCO₃and brine, dried with Na₂SO₄, ﬁltered, and concentrated.

The crude material was puriﬁed by ﬂash chromatography using 10%

EtOAc/hexane to give the product 21 as a yellow solid (0.474 g,

2.307 mmol, 90.9%). TLC R_f(25% EtOAc/hexane) = 0.37. Mp 49−

1

= 0.46. Mp 212−214 °C. H NMR (CDCl₃, 600 MHz): δ 7.92 (2H,

m), 7.84 (1H, t, J = 7.8 Hz), 7.68 (2H, d, J = 8.4 Hz), 7.66 (1H, d, J =

15.9 Hz), 7.34 (2H, d, J = 8.4 Hz), 6.51 (1H, d, J = 15.9 Hz), 5.12

(2H, s), 2.23 (3H, s), 1.47 (9H, s). ¹³C NMR (CDCl₃, 150 MHz): δ

1

167.62, 162.52, 160.73 (d, J_FC= 253 Hz), 154.52, 152.70, 144.50,

1

50 °C. H NMR (CDCl₃, 600 MHz): δ 7.61 (1H, d, J = 15.8 Hz),

144.42, 140.69 (d, ³J_FC= 8.6 Hz), 133.43, 132.15, 129.53, 129.06 (d,

7.38 (2H, d, J = 8.5 Hz), 6.56 (2H, d, J = 8.6 Hz), 6.22 (1H, d, J =

15.8 Hz), 4.24 (2H, q, J = 7.1 Hz), 4.08 (1H, bs), 2.87 (3H, s), 1.32

(3H, t, J = 7.1 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 168.01, 151.18,

145.26, 130.01, 123.47, 112.86, 112.20, 60.21, 30.41, 14.53. Anal.

Calcd for C₁₂H₁₅NO₂: C, 70.22; H, 7.37; N, 6.82. Found: C, 70.03;

H, 7.25; N, 6.71.

1

³J_FC= 4.1 Hz), 128.29, 123.64 (q, J_FC= 269 Hz), 122.96, 119.38,

118.91 (dq, ²J_FC= 33 Hz, ²J_FC= 13 Hz), 114.74, 114.58, 81.67, 46.12,

28.39, 15.15. Anal. Calcd for C₂₆H₂₄F₄N₂O₄S: C, 58.20; H, 4.51; N,

5.22. Found: C, 57.93; H, 4.63; N, 5.08.

Ethyl (E)-3-[4-[Bis[(4-methyl-2-(4-(triﬂuoromethyl)phenyl)-

thiazol-5-yl)methyl]amino]phenyl]acrylate (18). Ethyl 4-amino-

cinnamate (0.315 g, 1.647 mmol), thiazole 4b (0.969 g, 3.321

mmol), and NaI (0.556 g, 3.709 mmol) were dissolved in anhydrous

DMF (20 mL) under inert gas and cooled in an ice bath. NaH (60%

in mineral oil, 0.130 g, 3.250 mmol) was added by brieﬂy exposing the

reaction to air. The reaction mixture was stirred at rt for 24 h and

subsequently quenched with water and diluted with ether and

saturated NaHCO₃solution. Three ether extracts were obtained,

combined, dried with Na₂SO₄, ﬁltered, and concentrated. Puriﬁcation

of the material failed using ﬂash chromatography, and the crude

material 18 (0.240 g, 0.342 mmol, 20.8%) was moved to the next step.

(E)-3-[4-[Bis[(4-methyl-2-(4-(triﬂuoromethyl)phenyl)thiazol-5-yl)-

methyl]amino]phenyl]acrylic Acid (19). Ethyl ester 18 (0.238 g,

0.339 mmol) was dissolved in 95% EtOH (5 mL) and THF (5 mL).

The mixture was cooled in an ice bath, and 3 N NaOH (1 mL) was

added. The reaction mixture was stirred at rt overnight and was

concentrated under reduced pressure. The resulting residue was taken

up into EtOAc and washed with acidic water (HCl, pH 3) and brine,

dried with Na₂SO₄, ﬁltered, and concentrated. The subsequent crude

material was puriﬁed using ﬂash chromatography with a 25% EtOAc/

hexane mobile phase supplemented with dropwise amounts of acetic

acid. A yellow solid was collected and recrystallized using a mixture of

EtOAc and EtOH to give the product 19 as a light-yellow solid (0.062

g, 0.092 mmol, 27.1%). TLC R_f(50% EtOAc/hexane) = 0.50. Mp

Ethyl 4-[N-Methyl-N-[5-methylene-4-methyl-2-phenylthiazole]]-

aminocinnamate (22a). Ethyl 4-(N-methyl)aminocinnamate 21

(0.200 g, 0.974 mmol), thiazole 4a (0.238 g, 1.063 mmol), and NaI

(0.180 g, 1.200 mmol) were dissolved in anhydrous DMF (15 mL)

and cooled in an ice bath. NaH (60% dispersion in oil, 0.080 g, 2.000

mmol) was added, and the reaction mixture was stirred at room

temperature for 3.5 h. The reaction mixture was diluted with ether

and washed with a 10% aqueous NaHCO₃solution. The organic

phase was collected, and the aqueous phase was extracted with ether

three times. The combined organic extracts were washed with brine,

dried with Na₂SO₄, ﬁltered, and concentrated. Puriﬁcation by silica

gel plug was unsuccessful, and the yellow oil crude product 22a

(0.346 g, 0.881 mmol, 90.5%) was moved to the next step. TLC R_f

(25% EtOAc/hexane) = 0.35.

4-[N-Methyl-N-[5-methylene-4-methyl-2-phenylthiazole]]-

aminocinnamic Acid (23a). Ethyl ester 17a (0.205 g, 0.522 mmol)

was dissolved in 95% ethanol (5 mL) and THF (5 mL). The solution

was cooled in an ice bath, and 3 N NaOH (1 mL) was added and the

mixture was stirred at room temperature overnight. The reaction

mixture was concentrated, and the residue was taken up into EtOAc

and acidiﬁed water (HCl, pH 3). The organic phase was collected and

washed with brine, dried with Na₂SO₄, ﬁltered, and concentrated. The

crude product was puriﬁed by column chromatography using a 25%

EtOAc/hexane mobile phase containing dropwise amounts of acetic

acid. The collected product was recrystallized using a mixture of

EtOH and EtOAc to give the product 23a as a light-yellow solid

(0.061 g, 0.167 mmol, 32.0%). TLC R_f(50% EtOAc/hexane) = 0.36.

Mp 196−197 °C. ¹H NMR (CDCl₃, 600 MHz): δ 7.88 (2H, m), 7.58

(1H, d, J = 15.8 Hz), 7.54 (2H, d, J = 8.8 Hz), 7.42 (3H, m), 6.90

(2H, d, J = 8.9 Hz), 6.27 (1H, d, J = 15.9 Hz), 4.85 (2H, s), 3.10 (3H,

1

199−200 °C. H NMR (CDCl₃, 600 MHz): δ 7.97 (4H, d, J = 8.2

Hz), 7.72 (1H, d, J = 15.8 Hz), 7.65 (4H, d, J = 8.3 Hz), 7.51 (2H, d,

J = 8.9 Hz), 6.90 (2H, d, J = 8.9 Hz), 6.30 (1H, d, J = 15.8 Hz), 4.73

(4H, s), 2.47 (6H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 171.32,

2

164.07, 150.93, 149.21, 146.67, 136.64, 131.71 (q, J_FC= 32 Hz),

3

130.50, 130.18, 126.61, 126.07 (q, J_FC= 3 Hz), 125.11, 123.10,

7019

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

s), 2.48 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 168.38, 165.15,

151.46, 150.95, 145.89, 134.74, 130.75, 130.71, 129.91, 126.82,

124.24, 114.07, 113.71, 49.09, 38.46, 15.57. Anal. Calcd for

C₂₁H₂₀N₂O₂S: C, 69.12; H, 5.53; N, 7.69. Found: C, 69.03; H,

5.69; N, 7.76.

Ethyl 4-[N-Methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl]thiazole]]aminocinnamate (22b). By anal-

ogy to the procedure described in example 22a, ethyl 4-(N-

methyl)aminocinnamate 21 (0.430 g, 2.094 mmol) and thiazole 4b

(0.661 g, 2.265 mmol) were treated with NaI (0.336 g, 2.241 mmol)

and NaH (60% dispersion in oil, 0.132 g, 3.300 mmol). Separation of

the starting material and the desired product by ﬂash chromatography

was unsuccessful, and the yellow−orange oil 22b (0.721 g, 1.565

mmol, 74.7%) was moved to the next step. TLC R_f(25% EtOAc/

hexane) = 0.36.

8.8 Hz), 6.76 (2H, d, J = 8.8 Hz), 6.26 (1H, d, J = 16 Hz), 4.68 (2H,

s), 4.26 (2H, q, J = 7.14 Hz), 3.05 (3H, s), 2.50 (3H, s), 1.33 (3H, t, J

= 7.14 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 167.81, 161.97, 160.97

(d, ¹J_FC= 254 Hz), 150.57, 150.09, 144.69, 139.29 (d, ⁴J_FC= 8.6 Hz),

131.88, 129.99, 127.85, 124.07, 123.42 (q, ¹J_FC= 270 Hz), 121.70 (d,

³J_FC= 3.5), 119.16 (dq, ²J_FC= 14 Hz, ²J_FC= 33 Hz), 114.40 (d, ²J_FC

=

23 Hz), 113.97, 112.85, 60.33, 49.15, 38.42, 15.57, 14.53. Anal. Calcd

for C₂₄H₂₂F₄N₂O₂S (with 0.3 water mol per target): C, 59.57; H,

4.71; N, 5.79. Found: C, 59.19; H, 4.55; N, 5.74.

4-[N-Methyl-N-[5-methylene-4-methyl-2-[3-ﬂuoro-4-

(triﬂuormethyl)phenyl]thiazole]]aminocinnamic Acid (23d). By

analogy to the procedure described in example 23a, ethyl ester 22d

(0.157 g, 0.328 mmol) was treated with 3 N NaOH (1 mL), and the

crude product was puriﬁed by ﬂash chromatography with a gradient of

25−50% EtOAc/hexane containing dropwise amounts of acetic acid

to give the product 23d as an orange solid (0.045 g, 0.100 mmol,

30.5%). TLC R_f(50% EtOAc/hexane) = 0.36. Mp 214 °C. ¹H NMR

(acetone-d₆, 600 MHz): δ 7.86 (2H, m), 7.81 (1H, m), 7.60 (1H, d, J

= 16 Hz), 7.56 (2H, d, J = 8.9 Hz), 6.91 (2H, d, J = 8.9 Hz), 6.29

(1H, d, J = 16 Hz), 4.89 (2H, s), 3.12 (3H, s), 2.52 (3H, s). ¹³C NMR

(acetone-d₆, 150 MHz): δ 168.30, 161.56, 159.86 (d, ¹J_FC= 255 Hz),

4-[N-Methyl-N-[5-methylene-4-methyl-2-[4-(triﬂuoromethyl)-

phenyl]thiazole]]aminocinnamic Acid (23b). By analogy to the

procedure described in example 23a, ethyl ester 22b (0.555 g, 1.205

mmol) was treated with 3 N NaOH (1 mL), and the crude product

was puriﬁed by ﬂash chromatography using a 25% EtOAc/hexane

mobile phase containing dropwise amounts of acetic acid. The

collected product was recrystallized using EtOAc to give the product

23b as a yellow solid (0.338 g, 0.782 mmol, 64.9%). TLC R_f(50%

4

151.61, 151.21, 145.74, 140.72 (d, J_FC= 9 Hz), 133.73, 130.69,

3

129.04 (m), 124.52 (q, ¹J_FC= 270 Hz), 124.33, 122.87 (d, J_FC= 3.3

1

2

EtOAc/hexane) = 0.40. Mp 210−212 °C. H NMR (CDCl₃, 600

Hz), 118.85 (dq, J_FC= 33 Hz, J_FC= 12 Hz), 114.60 (d, J_FC= 23

Hz), 114.14, 113.63, 49.16, 38.54, 15.47. Anal. Calcd for

C₂₂H₁₈F₄N₂O₂S: C, 58.66; H, 4.03; N, 6.22. Found: C, 58.43; H,

4.09; N, 6.21.

MHz): δ 10.46 (1H, bs), 8.10 (2H, d, J = 8.1 Hz), 7.79 (2H, d, J = 8.3

Hz), 7.59 (1H, d, J = 15.9 Hz), 7.55 (2H, d, J = 8.9 Hz), 6.90 (2H, d,

8.9 Hz), 6.29 (1H, d, J = 15.9 Hz), 4.88 (2H, s), 3.11 (3H, s), 2.51

(3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 186.32, 163.07, 151.42,

Ethyl 4-[N-Ethyl-N-(tert-butoxycarbonyl)]aminocinnamate

(20e). To a stirred solution of ethyl 4-(N-tert-butoxycarbonyl)-

aminocinnamate 13 (1.459 g, 5.01 mmol) in anhydrous DMF (20

mL) under inert gas at 0 °C was added NaH (60% dispersion in

mineral oil, 0.317 g, 13.21 mmol). The mixture was stirred for 1 h,

iodoethane (1.2 mL, 14.9 mmol) was then added, and the mixture

continued to stir at room temperature overnight. The mixture was

then diluted with 50% sodium bicarbonate solution and extracted

with ether three times. The ether extracts were combined and washed

with brine, dried with Na₂SO₄, ﬁltered, and concentrated. The residue

was concentrated then puriﬁed by ﬂash chromatography using a

gradient of 10−15% EtOAc/hexane to give the desired product 20e as

a yellow solid (1.253 g, 3.923 mmol, 78.3%). TLC R_f(25% EtOAc/

hexane) = 0.56. Mp 58−60 °C. ¹H NMR (CDCl₃, 600 MHz): δ 7.67

(1H, d, J = 16.0 Hz), 7.49 (2H, d, J = 8.4 Hz), 7.23 (2H, d, J = 8.4

Hz), 6.40 (1H, d, J = 16.0 Hz), 4.28 (2H, q, J = 7.1 Hz), 3.71 (2H, q,

J = 7.0 Hz), 1.44 (9H, s), 1.35 (3H, t, J = 7.1 Hz), 1.17 (3H, t, J = 7.0

Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 167.19, 154.30, 144.44, 143.99,

131.82, 128.54, 126.93, 118.04, 80.62, 60.65, 44.93, 28.46, 14.47,

14.10. Anal. Calcd for C₁₈H₂₅NO₄: C, 67.69; H, 7.89; N, 4.39. Found:

C, 67.62; H, 7.90; N, 4.42.

Ethyl 4-(N-Ethyl)-aminocinnamate (21e). By analogy to the

procedure described in example 21, ethyl ester 20e (1.143 g, 3.579

mmol) was treated with triﬂuoroacetic acid (6 mL), and the crude

product was puriﬁed by ﬂash chromatography using a 5−10% EtOAc/

hexane mobile phase to give the desired product 21e as a light-yellow

solid (0.741 g, 3.380 mmol, 94.4%). TLC R_f(25% EtOAc/hexane) =

0.43. Mp 65−67 °C. ¹H NMR (CDCl₃, 600 MHz): δ 7.61 (1H, d, J =

16.0 Hz), 7.39 (2H, d, J = 8.5 Hz), 6.61 (2H, d, J = 8.5), 6.22 (1H, d,

J = 16.0 Hz), 4.25 (2H, q, J = 7.1 Hz), 3.22 (2H, q, J = 7.2 Hz), 1.33

(3H, t, J = 7.1 Hz), 1.28 (3H, t, J = 7.2 Hz). ¹³C NMR (CDCl₃, 150

MHz): δ 167.89, 145.00, 143.07, 132.70, 130.03, 129.23, 113.58,

60.29, 39.11, 14.54, 14.49. Anal. Calcd for C₁₃H₁₇NO₂: C, 71.21; H,

7.81; N, 6.39. Found: C, 71.01; H, 7.75; N, 6.33.

2

151.28, 145.80, 138.05, 132.54, 131.38 (q, J_FC= 33 Hz), 130.69,

1

127.27, 126.84 (m), 125.14 (q, J_FC= 270 Hz), 124.27, 114.06,

113.64, 49.09, 38.48, 15.49. Anal. Calcd for C₂₂H₁₉F₃N₂O₂S: C,

61.10; H, 4.43; N, 6.48. Found: C, 61.11; H, 4.41; N, 6.49.

Ethyl 4-[N-Methyl-N-[5-methylene-4-methyl-2-[3-

(triﬂuoromethyl)phenyl]thiazole]]aminocinnamate (22c). By anal-

ogy to the procedure described in example 22a, ethyl 4-(N-

methyl)aminocinnamate 21 (0.247 g, 1.203 mmol) and thiazole 4c

(0.407 g, 1.395 mmol) were treated with NaI (0.213 g, 1.421 mmol)

and NaH (60% dispersion in oil, 0.085 g, 2.125 mmol). Separation of

the starting material and the desired product by ﬂash column

chromatography was unsuccessful, and the yellow solid 22c (0.250 g,

0.543 mmol, 45.1%) was moved to the next step. TLC R_f(25%

EtOAc/hexane) = 0.45.

4-[N-Methyl-N-[5-methylene-4-methyl-2-[3-(triﬂuoromethyl)-

phenyl]thiazole]]aminocinnamic Acid (23c). By analogy to the

procedure described in example 23a, ethyl ester 22c (0.240 g, 0.521

mmol) was treated with 3 N NaOH (1 mL), and the crude product

was puriﬁed by ﬂash chromatography using a 25% EtOAc/hexane

mobile phase containing dropwise amounts of acetic acid. The

collected product was recrystallized using EtOH to give the product

23c as a yellow solid (0.098 g, 0.226 mmol, 43.5%). TLC R_f(50%

EtOAc/hexane) = 0.35. Mp 217 °C. ¹H NMR (CDCl₃, 600 MHz): δ

8.21 (1H, s), 8.12 (1H, d, J = 7.8 Hz), 7.76 (1H, d, J = 7.8 Hz), 7.69

(1H, t, J = 7.8 Hz), 7.59 (1H, d, J = 15.9 Hz), 7.55 (2H, d, J = 8.6

Hz), 6.91 (2H, d, J = 8.8 Hz), 6.28 (1H, d, J = 15.8 Hz), 4.88 (2H, s),

3.12 (3H, s), 2.51 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 168.28,

2

163.09, 151.29, 151.24, 145.79, 135.51, 132.20, 131.66 (q, J_FC= 33

3

1

Hz), 131.05, 130.69, 130.52, 126.94 (q, J_FC= 4 Hz), 125.01 (q, J_FC

3

= 270 Hz), 124.26, 122.89 (q, J_FC= 4 Hz), 114.05, 113.62, 49.08,

38.47, 15.47. Anal. Calcd for C₂₂H₁₉F₃N₂O₂S: C, 61.10; H, 4.43; N,

6.48. Found: C, 60.92; H, 4.45; N, 6.45.

Ethyl 4-[N-Methyl-N-[5-methylene-4-methyl-2-[3-ﬂuoro-4-

(triﬂuoromethyl)phenyl] thiazole]]aminocinnamate (22d). By

analogy to the procedure described in example 22a, ethyl ester 21

(0.2885 g, 1.405 mmol) and thiazole 4d (0.4754 g, 1.535 mmol) were

treated with NaI (0.2384 g, 1.590 mmol) and NaH (0.0697, 2.904

mmol). The crude product was puriﬁed by ﬂash chromatography with

a 5−20% EtOAc/hexane mobile-phase gradient, giving the desired

product 22d as an orange solid (0.192 g, 0.401 mmol, 28.5%). TLC R_f

Ethyl 4-[N-Ethyl-N-[5-methylene-4-methyl-2-[4-(triﬂuormethyl)-

phenyl]thiazole]]amino Cinnamate (22e). Ethyl ester 21e (0.362

g, 1.651 mmol), thiazole (0.593 g, 2.032 mmol), and NaI (0.392 g,

2.615 mmol) were combined in a two-necked ﬂask ﬁtted with a reﬂux

condenser. The system was ﬂushed with nitrogen, and anhydrous

acetonitrile (30 mL) was added. The reaction mixture was cooled in

an ice bath, and N,N-diisopropylethylamine (0.43 mL, 2.469 mmol)

was added. The reaction mixture was heated to reﬂux temperature

overnight. The mixture was allowed to cool, and the solvent was

1

(25% EtOAc/hexane) = 0.43. Mp 142 °C. H NMR (CDCl₃, 600

MHz): δ 7.73 (1H, m), 7.67 (1H, m), 7.63 (2H, m), 7.45 (2H, d, J =

7020

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

removed under reduced pressure. The residue was taken up into

DCM, washed with water and brine, and dried with Na₂SO₄. After

ﬁltration, the ﬁltrate was concentrated, and the resulting crude

product was puriﬁed by column chromatography with a 5−25%

EtOAc/hexane mobile-phase gradient to give the product 22e as a

yellow solid (0.704 g, 1.483 mmol, 89.8%). TLC R_f(25%EtOAc/

(2H, d, J = 8.9 Hz), 6.22 (1H, d, J = 15.8 Hz), 4.49 (2H, s), 4.28 (1H,

sept., J = 6.3 Hz), 4.23 (2H, q, J = 8.14 Hz), 1.32−1.29 (9H, m). ¹³C

NMR (CDCl₃, 150 MHz): δ 167.86, 163.27, 149.91, 148.34, 144.69,

2

136.93, 134.83, 131.27 (q, J_FC= 32 Hz), 129.91, 127.57, 126.38,

125.95 (q, ³J_FC= 3.5 Hz), 124.06 (q, ¹J_FC= 270 Hz), 113.81, 113.61,

60.31, 49.14, 41.65, 20.12, 15.71, 14.51.

1

hexane) = 0.40. Mp 108−109 °C. H NMR (CDCl₃, 600 MHz): δ

4 - [ N - I s o p r o p y l - N - [ 5 - m e t h y l e n e - 4 - m e t h y l - 2 - ( 4 -

triﬂuoromethylphenyl)thiazole]]aminocinnamic Acid (23f). By

analogy to the procedure described in example 23a, N-isopropyl

ester 22f (0.168 g, 0.344 mmol) was treated with 3 N NaOH (1 mL),

and the crude product was puriﬁed by ﬂash chromatography with a

gradient of 5−30% EtOAc/hexane to give the desired product 23f as a

white solid (0.122 g, 0.266 mmol, 77.1%). TLC R_f(50% EtOAc/

hexane) = 0.52. Mp 208−210 °C. ¹H NMR (acetone-d₆, 600 MHz): δ

10.28 (1H, bs), 8.09 (2H, d, J = 8.1 Hz), 7.77 (2H, d, J = 8.3 Hz),

7.56 (1H, d, J = 15.9 Hz), 7.51 (2H, d, J = 8.9 Hz), 6.89 (2H, d, J =

9.0 Hz), 6.26 (1H, d, J = 15.8 Hz), 4.68 (2H, s), 4.41 (1H, sept., J =

6.6 Hz), 2.52 (3H, s), 1.33 (6H, d, J = 6.6 Hz). ¹³C NMR (acetone-

d₆, 150 MHz): δ 168.24, 162.68, 151.03, 149.56, 145.67, 138.24,

136.40, 131.20 (q, ²J_FC= 32 Hz), 130.61, 127.10, 126.79 (q, ³J_FC= 3.5

Hz), 125.17 (q, ¹J_FC= 270 Hz), 124.48, 114.57, 114.15, 49.93, 41.98,

20.04, 15.60. Anal. Calcd for C₂₄H₂₃F₃N₂O₂S: C, 62.60; H, 5.03; N,

6.08. Found: C, 62.30; H, 5.09; N, 6.03.

Ethyl 4-[N-Propyl-N-(tert-butoxycarbonyl)]aminocinnamate

(20g). By analogy to the procedure described in example 20e, ethyl

4-aminocinnamate-BOC 13 (2.147 g, 7.374 mmol) was treated with

NaH (60% dispersion in mineral oil, 0.314 g, 7.850 mmol) and 1-

iodopropane (2.2 mL, 22.558 mmol). The crude material was puriﬁed

by ﬂash chromatography with a gradient of 100% hexane to 10%

EtOAc/hexane to give the desired product 20g as a yellow solid

(2.290 g, 6.868 mmol, 93.2%). TLC R_f(25% EtOAc/hexane) = 0.59.

Mp 56 °C. ¹H NMR (CDCl₃, 600 MHz): δ 7.66 (1H, d, J = 16 Hz),

7.48 (2H, d, J = 8.4 Hz), 7.22 (2H, d, J = 8.4 Hz), 6.39 (1H, d, J = 16

Hz), 4.26 (2H, q, J = 7.14 Hz), 3.60 (2H, t, J = 7.4 Hz), 1.56 (2H,

sex., J = 7.4 Hz), 1.43 (9H, s), 1.34 (3H, t, J = 7.14 Hz), 0.87 (3H, t, J

= 7.4 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 167.13, 154.51, 144.53,

143.95, 131.81, 128.51, 127.03, 118.02, 80.54, 60.60, 51.53, 28.41,

21.92, 14.44, 11.26. Anal. Calcd for C₁₉H₂₇NO₄: C, 68.44; H, 8.16; N,

4.20. Found: C, 68.46; H, 8.22; N, 4.22.

7.92 (2H, d, J = 8.2 Hz), 7.59 (3H, m), 7.40 (2H, d, J = 8.9 Hz), 6.69

(2H, d, J = 8.9 Hz), 6.22 (1H, d, J = 16 Hz), 4.58 (2H, s), 4.22 (2H,

q, J = 7.1 Hz), 3.48 (2H, q, J = 7.1), 2.49 (3H, s), 1.30 (3H, t, J = 7.1

Hz), 1.21 (3H, t, J = 7.1 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ

167.90, 163.49, 149.74, 149.02, 144.78, 136.84, 132.18, 131.42 (²J_FC

=

32 Hz), 130.09, 126.49, 125.99 (³J_FC= 3.7 Hz), 124.05 (¹J_FC= 270

Hz), 123.54, 113.55, 112.55, 60.31, 46.95, 45.39, 15.62, 14.54, 12.38.

Anal. Calcd for C₂₅H₂₅F₃N₂O₂S: C, 63.28; H, 5.31; N, 5.90. Found:

C, 63.02; H, 5.35; N, 5.90.

4-[N-Ethyl-N-[5-methylene-4-methyl-2-[4-(triﬂuormethyl)-

phenyl]thiazole]]aminocinnamic Acid (23e). By analogy to the

procedure described in example 23a, ethyl ester 22e (0.132 g, 0.278

mmol) was treated with 3 N NaOH (1 mL). The crude product was

puriﬁed by ﬂash chromatography with a gradient of 15−45% EtOAc/

hexane with dropwise amounts of acetic acid to give the desired

product 23e as an orange solid (0.061 g, 0.137 mmol, 49.3%). TLC R_f

(50% EtOAc/hexane) = 0.43. Mp 198 °C. ¹H NMR (acetone-d₆, 600

MHz): δ 8.09 (2H, d, J = 8.9 Hz), 7.78 (2H, d, J = 8.2 Hz), 7.58 (1H,

d, J = 16 Hz), 7.52 (2H, d, J = 8.9 Hz), 6.86 (2H, d, J = 8.9 Hz), 6.27

(1H, d, J = 16 Hz), 4.82 (2H, s), 3.62 (2H, q, J = 7 Hz), 2.51 (3H, s),

1.24 (3H, t, J = 7 Hz). ¹³C NMR (acetone-d₆, 150 MHz): δ 168.45,

2

162.97, 150.80, 150.05, 145.83, 138.08, 133.86, 131.43 (q, J_FC= 32

3

1

Hz), 130.80, 127.21, 126.82 (q, J_FC= 4 Hz), 126.03 (q, J_FC= 270

Hz), 123.88, 113.76, 113.36, 47.22, 45.92, 15.49, 12.52. Anal. Calcd

for C₂₃H₂₁F₃N₂O₂S with 0.17 mol per target CDCl₃: C, 59.60; H,

4.53; N, 6.00. Found: C, 59.56; H, 4.75; N, 5.96.

Ethyl 4-[N-Isopropyl-N-(tert-butoxycarbonyl)]aminocinnamate

(20f). By analogy to the procedure described in example 20e, ethyl

4-aminocinnamate-BOC 13 (0.840 g, 2.885 mmol) was treated with

NaH (60% dispersion in mineral oil, 0.262 g, 6.550 mmol) and 2-

iodopropane (1 mL, 10.1 mmol). The crude material was puriﬁed by

ﬂash chromatography with a gradient of 5−10% EtOAc/hexane to

give the desired product 20f as a yellow oil (0.488 g, 1.464 mmol,

50.7%). TLC R_f(25% EtOAc/hexane) = 0.58. ¹H NMR (CDCl₃, 600

MHz): δ 7.68 (1H, d, J = 15.9 Hz), 7.50 (2H, d, J = 8.3 Hz), 7.10

(2H, d, J = 8.3 Hz), 6.42 (1H, d, J = 16.0 Hz), 4.50 (1H, sept, J = 6.8

Hz), 4.27 (2H, q, J = 7.14 Hz), 1.38 (9H, s), 1.34 (3H, t, J = 7.14

Hz), 1.13 (6H, d, J = 6.8 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ

167.11, 154.70, 144.00, 141.47, 133.11, 130.37, 128.35, 118.54, 80.16,

60.69, 48.93, 28.49, 21.69, 14.48.

Ethyl 4-(N-Isopropyl)aminocinnamate (21f). By analogy to the

procedure described in example 21, ethyl ester 20f (1.05 g, 3.149

mmol) was treated with TFA (8 mL). The crude product was puriﬁed

by ﬂash chromatography to give the desired product 21f as a white

solid (0.609 g, 2.613 mmol, 83.0%). TLC R_f= 0.11 (25% EtOAc/

hexane). Mp 112−115 °C. ¹H NMR (CDCl₃, 600 MHz): δ 7.59 (1H,

d, J = 15.8 Hz), 7.35 (2H, d, J = 8.5 Hz), 6.54 (2H, d, J = 8.6 Hz),

6.20 (1H, d, J = 15.8 Hz), 4.23 (2H, q, J = 7.1 Hz), 3.95 (1H, bs),

3.67 (1H, sept., J = 6.3 Hz), 1.32 (3H, t, J = 7.1 Hz), 1.22 (6H, d, J =

6.3 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 168.03, 149.32, 145.23,

130.12, 123.27, 112.94, 112.73, 60.21, 44.24, 22.96, 14.55. Anal. Calcd

for C₁₄H₁₉NO₂: C, 72.07; H, 8.21; N, 6.00. Found: C, 72.03; H, 8.30;

N, 5.98.

Ethyl 4-(N-Propyl)aminocinnamate (21g). By analogy to the

procedure described in example 21, ethyl ester 20g (2.080 g, 6.238

mmol) was treated with triﬂuoroacetic acid (8 mL), and the crude

material was puriﬁed by ﬂash chromatography with a gradient of 10−

20% EtOAc/hexane to give an orange solid 21g (0.573 g, 2.455

1

mmol, 39.4%). TLC R_f(25% EtOAc/hexane) = 0.43. Mp 78 °C. H

NMR (CDCl₃, 600 MHz): δ 7.61 (1H, d, J = 16 Hz), 7.38 (2H, d, J =

8 Hz), 6.61 (2H, d, J = 8 Hz), 6.23 (1H, d, J = 16 Hz), 4.25 (2H, q, J

= 7.14 Hz), 3.13 (2H, t, J = 7.3 Hz), 1.68 (2H, sex., J = 7.3 Hz), 1.33

(3H, t, J = 7.14 Hz), 1.00 (3H, t, J = 7.3 Hz). ¹³C NMR (CDCl₃, 150

MHz): δ 167.97, 149.60, 145.11, 130.06, 129.20, 124.13, 133.22,

60.28, 46.01, 22.50, 14.55, 11.67. Anal. Calcd for C₁₃H₁₇F₄NO₂(with

0.13 mol CDCl₃per target): C, 68.17; H, 7.69; N, 5.63. Found: C,

68.09; H, 7.50; N, 5.53.

Ethyl 4-[N-Propyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuormethyl)phenyl]aminocinnamate (22g). By analogy to the

procedure described in example 22a, ethyl ester 21g (0.237 g, 1.014

mmol) and thiazole 4b (0.325 g, 1.115 mmol) were treated with NaI

(0.231 g, 1.542 mmol) and NaH (0.045, 1.896 mmol). The crude

material was puriﬁed by ﬂash chromatography with a gradient of 10−

30% EtOAc/hexane to give 22g as an orange solid (0.294 g, 0.603

mmol, 59.5%). TLC R_f(25% EtOAc/hexane) = 0.35. Mp 128−129

Ethyl 4-[N-Isopropyl-N-[5-methylene-4-methyl-2-(4-

triﬂuoromethylphenyl)thiazole]]aminocinnamate (22f). By analogy

to the procedure described in example 22a, N-isopropyl ester 21f

(0.490 g, 2.100 mmol) and thiazole 4b (0.745 g, 2.554 mmol) were

treated with NaI (0.41g, 2.735 mmol) and NaH (0.107 g, 2.675

mmol). Attempts to purify the crude sample by ﬂash chromatography

failed, and the crude product 22f was moved to the next step without

further puriﬁcation (0.181 g, 0.370 mmol, 17.8%). TLC R_f(25%

EtOAc/hexane) = 0.35. Mp 80−82 °C. ¹H NMR (CDCl₃, 600 MHz):

δ 7.94 (2H, d, J = 7.9 Hz), 7.60 (3H, m), 7.40 (H, d, J = 8.9 Hz), 6.76

1

°C. H NMR (CDCl₃, 600 MHz): δ 7.97 (2H, d, J = 8.2 Hz), 7.64

(2H, d, J = 8.2 Hz), 7.60 (1H, d, J = 16 Hz), 7.43 (2H, d, J = 8.9 Hz),

6.75 (2H, m), 6.24 (1H, d, J = 16 Hz), 4.66 (2H, s), 4.23 (2H, q, J =

7.14 Hz), 3.39 (2H, t, J = 7.4 Hz), 2.51 (3H, s), 1.71 (2H, sex., J = 7.6

Hz), 1.32 (3H, t, J = 7.14 Hz), 0.98 (3H, t, J = 7.4 Hz). ¹³C NMR

(CDCl₃, 150 MHz): δ 167.89, 163.61, 149.79, 148.99, 144.73, 136.68,

131.92, 131.60 (²J_FC= 32 Hz), 130.05, 126.58, 126.01 (³J_FC= 3.9

Hz), 124.92 (¹J_FC= 270 Hz), 123.67, 113.64, 112.69, 60.32, 53.18,

7021

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

47.66, 20.62, 15.57, 14.53, 11.56. Anal. Calcd for C₂₆H₂₇F₃N₂O₂S: C,

63.92; H, 5.57; N, 5.73. Found: C, 63.71; H, 5.55; N, 5.65.

122.29, 81.41, 36.92, 28.47, 19.76. Anal. Calcd for C₁₄H₁₉NO₃(with

0.1 H₂O and 0.1 mol DCM per target): C, 65.24; H, 7.53; N, 5.40.

Found: C, 65.27; H, 7.18; N, 5.06.

4-[N-Propyl-N-[5-methylene-4-methyl-2-[4-(triﬂuormethyl)-

phenyl]]aminocinnamic Acid (23g). By analogy to the procedure

described in example 23a, ethyl ester 20c (0.165 g, 0.337 mmol) was

treated with 3 N NaOH (1 mL). The crude material was puriﬁed by

ﬂash chromatography with a gradient of 15−50% EtOAc/hexane

supplemented with dropwise amounts of acetic acid to give the

desired compound 23g as an orange solid (0.134 g, 0.290 mmol,

86.1%). TLC R_f(50% EtOAc/hexane) = 0.36. Mp 205 °C. ¹H NMR

(acetone-d₆, 600 MHz): δ 10.41 (1H, bs), 8.09 (2H, d, J = 8.2 Hz),

7.78 (2H, d, J = 8.2 Hz), 7.57 (1H, d, J = 16 Hz), 7.52 (2H, d, J = 8.9

Hz), 6.85 (2H, d, J = 8.9 Hz), 6.25 (1H, d, J = 16 Hz), 4.85 (2H, s),

3.50 (2H, t, J = 7.5 Hz), 2.51 (3H, s), 1.73 (2H, sex., J = 7.5 Hz), 0.98

(3H, J = 7.5 Hz). ¹³C NMR (acetone-d₆, 150 MHz): δ 168.36,

162.95, 150.85, 150.25, 145.83, 138.09, 133.72, 131.43 (²J_FC= 32

Hz), 130.76, 127.21, 126.83 (³J_FC= 4 Hz), 126.04 (¹J_FC= 270 Hz),

123.82, 113.71, 113.37, 53.49, 47.80, 21.17, 15.51, 11.49. Anal. Calcd

for C₂₄H₂₃F₃N₂O₂S: C, 62.60; H, 5.03; N, 6.08. Found: C, 62.20; H,

5.04; N, 6.03.

Methyl 2-Methyl-4-[N-(tert-butoxycarbonyl)methyl]-

aminocinnamate (28a). Aldehyde 27a (0.591 g, 2.371 mmol) and

methyl(triphenylphosphoranylidene)acetate (0.876 g, 2.620 mmol)

were dissolved in anhydrous THF (10 mL) under nitrogen gas and

stirred at 55 °C for 2 days. The reaction mixture was concentrated,

and the resulting crude solid was taken up into EtOAc, washed with

water and brine, and dried with Na₂SO₄. After ﬁltration, the ﬁltrate

was concentrated and puriﬁed using ﬂash chromatography with a 5−

15% EtOAc/hexane gradient to give the desired product 28a as a

yellow oil (0.548 g, 1.794 mmol, 75.7%). TLC R_f(25%EtOAc/

hexane) = 0.53. ¹H NMR (CDCl₃, 600 MHz): δ 7.94 (1H, d, J = 15.9

Hz), 7.51 (1H, d, J = 9.1 Hz), 7.11 (2H, m), 6.33 (1H, d, J = 15.9

Hz), 3.81 (3H, s), 3.26 (3H, s), 2.42 (3H, s), 1.47 (9H, s). ¹³C NMR

(CDCl₃, 150 MHz): δ 167.70, 154.56, 145.36, 141.95, 138.32, 130.16,

127.01, 126.78, 123.11, 118.34, 80.87, 51.84, 37.14, 28.46, 20.06.

Anal. Calcd for C₁₇H₂₃NO₄(with 0.1 mol H₂O per target): C, 66.47;

H, 7.61; N, 4.56. Found: C, 66.26; H, 7.44; N, 4.56.

4-Bromo-2-methylbenzaldehyde (25a). 4-Bromo-2-methylbenzo-

nitrile 24a (1.859 g, 9.483 mmol) was dissolved in anhydrous toluene

(40 mL) under nitrogen and was cooled to −78 °C. Reagent grade

DIBAL-H (1.70 mL, 9.538 mmol) was slowly added, and mixture was

stirred for 30 min. Anhydrous methanol (3 mL) was carefully added

followed by 2 M H₂SO₄(9 mL) dropwise. The mixture was stirred

overnight at rt. The mixture was diluted with ethyl acetate (50 mL),

and the organic phase was collected, dried with Na₂SO₄, and

concentrated. The crude material was puriﬁed with ﬂash chromatog-

raphy using 5% EtOAc/hexane to give 25a as an oﬀ-white solid

(1.590 g, 7.988 mmol, 84.2%). TLC R_f(25% EtOAc/hexane) = 0.80.

Methyl 2-Methyl-4-(N-methyl)aminocinnamate (29a). Methyl 2-

Methyl-4-[N-(tert-butoxycarbonyl)methyl]aminocinnamate 28a

(0.516 g, 1.689 mmol) was taken up into anhydrous DCM (10

mL) under nitrogen gas and cooled in an ice bath. Triﬂuoroacetic acid

(4 mL) was slowly added, and the mixture was stirred at room

temperature for 1.5 h. The reaction mixture was concentrated, and the

residue was taken up into EtOAc, washed with chilled sat. NaHCO₃

and brine, and dried with Na₂SO₄. After ﬁltration, the ﬁltrate was

concentrated, and the ﬁltrate was puriﬁed using ﬂash chromatography

and a 10−20% EtOAc/hexane mobile-phase gradient to give the

product 29a as a yellow solid (0.229 g, 1.116 mmol, 66.1%). TLC

1

Mp 29−30 °C. H NMR (CDCl₃, 600 MHz): δ 10.21 (1H, s), 7.66

(25% EtOAc/hexane) = 0.31. Mp 79−83 °C. H NMR (CDCl₃, 600

(1H, d, J = 8.2), 7.51 (1H, dd, J₁= 8.2 Hz, J₂= 1.9 Hz), 7.44 (1H, d, J

= 1.1 Hz), 2.65 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 191.79,

142.53, 134.88, 133.37, 133.04, 129.85, 129.03, 19.42.

MHz): δ 7.92 (1H, d, J = 15.7 Hz), 7.47 (1H, d, J = 8.5 Hz), 6.44

(1H, dd, J₁= 8.5 Hz, J₂= 2.5 Hz), 6.19 (1H, d, J = 15.8 Hz), 4.11

(1H, bs), 3.78 (3H, s), 2.86 (3H, s), 2.39 (3H, s). ¹³C NMR (CDCl₃,

150 MHz): δ 168.51, 150.94, 142.66, 139.97, 128.13, 122.18, 113.62,

113.10, 110.61, 51.52, 30.41, 20.26. Anal. Calcd for C₁₂H₁₅NO₂: C,

70.22; H, 7.37; N, 6.82. Found: C, 70.00; H, 7.26; N, 6.68.

tert-Butyl-N-methylcarbamate (26). Di-tert-butyl dicarbonate

(10.91 g, 49.98 mmol) was dissolved in anhydrous THF (30 mL)

under nitrogen and was cooled in an ice bath. Methylamine (2 M) in

THF (50.0 mL, 0.1 mol) was slowly added, and mixture was allowed

to warm to rt and was stirred overnight. The solvent was removed

using rotary evaporation, and the resulting residue was taken up into 1

M HCl and DCM. The organic phase was collected, and the aqueous

phase was extracted with a separate portion of DCM. The organic

phases were combined and washed with water, dried with sodium

sulfate, and concentrated. Residual solvents were removed using

Kugelrohr distillation to give the title compound 26 as a light-yellow

Methyl 2-Methyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl]-thiazole]]aminocinnamate (30a). Methyl

ester 29a (0.316 g, 1.540 mmol), thiazole 4b (0.520 g, 1.782 mmol),

and NaI (0.328 g, 2.188 mmol) were dried in vacuum desiccator

overnight to ensure dry starting reagents. These starting reagents were

combined in a two-necked ﬂask under nitrogen, and anhydrous DMF

(15 mL) was added. The reaction mixture was cooled in an ice bath,

and NaH (60% dispersion in mineral oil, 0.101 g, 2.525 mmol) was

added by brieﬂy exposing the system to air. The mixture was warmed

to room temperature and stirred for 2.5 h. The reaction mixture was

carefully quenched with water, diluted with ether, and neutralized

with NaHCO₃. Additional water was added to obtain a clear two-

phase solution. The organic phase was collected, and the aqueous

phase was extracted with two portions of ether. The organic phases

were combined, washed with water and brine, dried with Na₂SO₄.

After ﬁltration, the ﬁltrate was concentrated, and the residue was

puriﬁed by ﬂash chromatography with a 5−20% EtOAc/hexane

mobile-phase gradient to give the product 30a as a yellow−orange

solid (0.523 g, 1.136 mmol, 73.8%). TLC (25% EtOAc/hexane) =

0.33. Mp 151−152 °C. ¹H NMR (CDCl₃, 600 MHz): δ 7.96 (2H, d, J

= 8.2 Hz), 7.93 (1H, d, J = 15.8 Hz), 7.64 (2H, d, J = 8.2 Hz), 7.53

(1H, d, J = 8.8 Hz), 6.66 (1H, dd, J₁= 8.8 Hz, J₂= 2.7 Hz), 6.58 (1H,

d, J = 2.5 Hz), 6.24 (1H, d, 8.8 Hz), 4.67 (2H, s), 3.79 (3H, s), 3.04

(3H, s), 2.51 (3H, s), 2.43 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ

168.35, 163.65, 150.31, 149.90, 142.29, 139.88, 136.73, 131.51 (q,

1

oil (6.476 g, 49.37 mmol, 98.7%). H NMR (CDCl₃, 600 MHz): δ

4.43 (1H, bs), 2.73 (3H, d, J = 4.92 Hz), 1.44 (9H, s). ¹³C NMR

(CDCl₃, 150 MHz): δ 156.76, 79.27, 28.59, 27.28. Anal. Calcd for

C₆H₁₃NO₂(with 0.1 mol H₂O and 0.1 mol DCM per target): C,

51.79; H, 9.55; N, 9.90. Found: C, 51.70; H, 9.34; N, 9.73.

2-Methyl-4-[N-(tert-butoxycarbonyl)methyl]benzaldehyde

(27a). Aldehyde 25a (1.286 g, 6.461 mmol), tert-butyl-N-methyl-

carbamate 26 (1.015 g, 7.737 mmol), and Cs₂CO₃(2.958 g, 9.078

mmol) were dissolved in anhydrous dioxane (40 mL) under nitrogen.

Xantphos (0.175 g, 0.302 mmol) and Pd₂(dba)₃(0.136 g, 0.148

mmol) were measured out in a nitrogen bag and added to the reaction

ﬂask. The reaction mixture was reﬂuxed with stirring overnight and

was then cooled to room temperature before removal of the solvent

under reduced pressure. The resulting residue was taken up into

EtOAc and water followed by ﬁltration through Celite. The EtOAc

phase was collected, washed with water and brine, and dried with

Na₂SO₄. After ﬁltration, the ﬁltrate was concentrated, and the residue

was puriﬁed using ﬂash chromatography and a 100% hexane to 5%

EtOAc/hexane mobile-phase gradient to give the desired product 27a

as a yellow oil (0.976 g, 3.914 mmol, 60.6%). TLC R_f(25% EtOAc/

3

²J_FC= 32.4 Hz), 131.02, 128.18, 126.57, 126.02 (q, J_FC= 3.5 Hz),

1

124.03 (q, J_FC= 271 Hz), 122.88, 114.52, 114.21, 111.08, 51.64,

49.06, 38.33, 20.65, 15.56. Anal. Calcd for C₂₄H₂₃N₂F₃O₂S: C, 62.60;

H, 5.03; N, 6.08. Found: C, 62.36; H, 5.06; N, 6.05.

1

hexane) = 0.53. H NMR (CDCl₃, 600 MHz): δ 10.20 (1H, s), 7.75

(1H, d, J = 8.4 Hz), 7.28 (1H, d, J = 2.2 Hz), 7.19 (1H, d, J = 1.7 Hz),

3.30 (3H, s), 2.66 (3H, s), 1.50 (9H, s). ¹³C NMR (CDCl₃, 150

MHz): δ 191.66, 154.22, 148.55, 141.43, 132.73, 130.96, 127.39,

2-Methyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl]thiazole]]aminocinnamic Acid (31a).

Methyl ester 30a (0.400 g, 0.868 mmol) was dissolved in anhydrous

7022

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

THF (10 mL) and 95% ethanol (5 mL) and was cooled in an ice bath.

Upon cooling, a precipitate formed but then returned to solution after

the addition of 3 N NaOH (1 mL). The reaction mixture was stirred

overnight at room temperature and was then concentrated under

reduced pressure. The resulting residue was taken up into EtOAc and

washed with acidic water (pH 3, HCl) and brine and dried with

Na₂SO₄. After ﬁltration, the ﬁltrate was concentrated, and the residue

was puriﬁed by ﬂash chromatography using a 25% EtOAc/hexane

mobile phase that was supplemented with dropwise amounts of acetic

acid. The column-puriﬁed product was collected and recrystallized

using 95% ethanol to give 31a as a yellow solid (0.081 g, 0.181 mmol,

unsuccessful, and the crude yellow solid 29b (0.758 g, 3.693 mmol,

83.0%) was moved to the next step. TLC (25% EtOAc/hexane) =

0.36. ¹H NMR (CDCl₃, 400 MHz): δ 7.63 (1H, d, J = 16.0 Hz), 7.35

(1H, dd, J₁= 8.4 Hz, J₂= 2.0 Hz), 6.57 (1H, d, J = 8.4 Hz), 6.24 (1H,

d, J = 15.6 Hz), 3.77 (3H, s), 2.93 (3H, s), 2.13 (3H, s).

Methyl 2-Methyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl]-thiazole]]aminocinnamate (30b). By anal-

ogy to the procedure described in example 30a, methyl ester 29b

(0.200 g, 0.974 mmol) and thiazole 4b (0.340 g, 1.166 mmol) were

treated with NaI (0.215 g, 1.434 mmol) and DIPEA (0.255 g, 1.464

mmol). The crude product was puriﬁed by ﬂash chromatography with

a 5−13% EtOAc/hexane mobile-phase gradient to give the product

30b as a yellow solid (0.341 g, 0.740 mmol, 76.0%). TLC (25%

EtOAc/hexane) = 0.50. Mp 92−94 °C. ¹H NMR (CDCl₃, 600 MHz):

δ 8.00 (2H, d, J = 8.1 Hz), 7.65 (3H, m), 7.40 (1H, d, J = 1.6 Hz),

7.35 (1H, dd, J₁= 8.2 Hz, J₂= 2.0 Hz), 7.04 (1H, d, J = 8.2 Hz), 6.36

(1H, d, J = 16.0 Hz), 4.23 (2H, s), 3.80 (3H, s), 2.72 (3H, s), 2.43

(6H, m). ¹³C NMR (CDCl₃, 150 MHz): δ 167.85, 164.05, 153.14,

1

20.9%). TLC (50% EtOAc/hexane) = 0.46. Mp 228−230 °C. H

NMR (acetone-d₆, 600 MHz): δ 10.36 (1H, bs), 8.10 (2H, d, J = 8.1

Hz), 7.90 (1H, d, J = 15.8 Hz), 7.79 (2H, d, J = 8.3 Hz), 7.64 (1H, d,

J = 8.7 Hz), 6.78 (1H, dd, J₁= 9.0 Hz, J₂= 2.8 Hz), 6.75 (1H, d, J =

2.6 Hz), 6.25 (1H, d, J = 15.8 Hz), 4.86 (2H, s), 3.10 (3H, s), 2.52

(3H, s), 2.41 (3H, s). ¹³C NMR (acetone-d₆, 150 MHz): δ 168.33,

163.07, 151.40, 151.08, 142.79, 140.12, 138.07, 132.66, 131.38 (q,

²J_FC= 31.8 Hz), 128.82, 127.27, 126.84 (q, ³J_FC= 3.9 Hz), 125.14 (q,

¹J_FC= 271 Hz), 122.88, 115.10, 114.68, 111.88, 48.97, 38.38, 20.31,

15.49. Anal. Calcd for C₂₃H₂₁F₃N₂O₂S (with 0.2 H₂O mol per target):

C, 61.38; H, 4.79; N, 6.22. Found: C, 61.15; H, 4.86; N, 6.18. LC-

MS/MS analysis shows >98% chromatographic purity (total ion

current) with retention time = 5.21 min and m/z = 447 (for M + 1).

4-Bromo-3-methylbenzaldehyde (25b). By analogy to the

procedure described in example 25a, 4-Bromo-3-methylbenzonitrile

24b (2.100 g, 10.71 mmol) was treated with DIBAL-H, and the crude

material was puriﬁed by column chromatography using a 100%

hexane to 5% EtOAc/hexane mobile-phase gradient. The desired

product 25b was collected as a white solid (1.801 g, 9.051 mmol,

2

151.00, 144.71, 136.93, 133.26, 131.45 (q, J_FC= 33 Hz), 131.31,

3

2

129.73, 126.91, 126.57, 126.02 (q, J_FC= 3.5 Hz), 124.07 (q, J_FC

=

270 Hz), 120.39, 116.30, 51.97, 51.79, 41.05, 18.78, 15.50. Anal.

Calcd for C₂₄H₂₃F₃N₂O₂S (with 0.2 H₂O mol per target): C, 62.11;

H, 5.08; N, 6.04. Found: C, 62.08; H, 5.06; N, 5.93.

2-Methyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl]thiazole]]aminocinnamic Acid (31b). By

analogy to the procedure described in example 31a, methyl ester 30b

(0.220 g, 0.478 mmol) was treated with 3 N NaOH (2 mL), and the

crude product was puriﬁed using ﬂash chromatography and a 25%

EtOAc/hexane mobile phase that was supplemented with dropwise

amounts of acetic acid. The column product was recrystallized using

95% ethanol to give 31b as a pale-yellow solid (0.170 g, 0.382 mmol,

1

84.5%). TLC (25% EtOAc/hexane) = 0.70. Mp 120−122 °C. H

1

79.8%). H NMR (acetone-d₆, 600 MHz): δ 10.66 (1H, bs), 8.14

NMR (CDCl₃, 600 MHz): δ 9.95 (1H, s), 7.72 (1H, s), 7.70 (1H, d, J

= 8.1 Hz), 7.55 (1H, d, J = 8.0 Hz), 2.48 (3H, d). ¹³C NMR (CDCl₃,

600 MHz): δ 191.56, 139.29, 135.54, 133.38, 132.37, 131.59, 128.41,

23.05. Anal. Calcd for C₈H₇OBr: C, 48.27; H, 3.54; N, 0.00. Found:

C, 48.01; H, 3.56; N, 0.13.

(2H, d, J = 8.1 Hz), 7.81 (2H, d, J = 8.2 Hz), 7.61 (1H, d, J = 16.0

Hz), 7.56 (1H, s), 7.48 (1H, dd, J₁= 8.2 Hz, J₂= 1.7 Hz), 7.18 (1H,

d, J = 8.3 Hz), 6.43 (1H, d, J = 16.0 Hz), 4.37 (2H, s), 2.75 (3H, s),

2.45 (3H, s), 2.42 (3H, s). ¹³C NMR (acetone-d₆, 150 MHz): δ

167.93, 163.64, 154.06, 151.91, 145.33, 138.21, 134.06, 132.68,

3-Methyl-4-[N-(tert-butoxycarbonyl)methyl]benzaldehyde

(27b). By analogy to the procedure described in example 27a, the

reaction with aldehyde 25b (1.423 g, 7.149 mmol) and tert-butyl-N-

methylcarbamate 26 (1.123 g, 8.561 mmol) using Pd₂(dba)₃(0.149 g,

0.163 mmol) gave a crude product that was puriﬁed with ﬂash

chromatography using a 100% hexane to 5% EtOAc/hexane mobile-

phase gradient to give the desired product 27b as a light-yellow,

almost colorless oil (0.837 g, 3.357 mmol, 47.0%). TLC (25%

EtOAc/hexane) = 0.41. ¹H NMR (CDCl₃, 600 MHz): δ 9.98 (1H, s),

7.75 (1H, s), 7.71 (1H, d, J = 7.7 Hz), 3.17 (3H, s), 2.30 (3H, s), 1.53

(3H, s), 1.34 (7H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 191.85,

154.47, 148.25, 136.77, 135.08, 132.16, 128.46, 128.18, 80.54, 36.69,

28.36, 17.78. Anal. Calcd for C₁₄H₁₉O₃N: C, 67.45; H, 7.68; N, 5.62.

Found: C, 67.26; H, 7.73; N, 5.67.

Methyl 3-Methyl-4-[N-(tert-butoxycarbonyl)methyl]-

aminocinnamate (28b). By analogy to the procedure described in

example 28a, reaction with aldehyde 27b (0.736 g, 2.953 mmol) and

methyl(triphenylphosphoranylidene)acetate (1.087 g, 3.251 mmol)

gave a crude material that was puriﬁed using ﬂash chromatography

and a 100% hexane to 7% EtOAc/hexane mobile-phase gradient to

give the desired product 28b as a light-yellow oil (0.561 g, 1.837

mmol, 62.2%). TLC (25% EtOAc/hexane) = 0.43. ¹H NMR (CDCl₃,

600 MHz): δ 7.66 (1H, d, J = 16 Hz), 7.37 (1H, s), 7.35 (1H, d, J =

7.9 Hz), 7.09 (1H, d, J = 7.9 Hz), 6.42 (1H, d, J = 15.8 Hz), 3.80 (3H,

s), 3.14 (3H, s), 2.23 (3H, s), 1.51 (3H, s), 1.33 (6H, s). ¹³C NMR

(CDCl₃, 150 MHz): δ 167.57, 154.86, 144.47, 144.38, 136.17, 133.16,

130.57, 127.92, 126.50, 117.91, 80.17, 51.87, 36.79, 28.39, 17.75.

Anal. Calcd for C₁₇H₂₃O₄N: C, 66.86; H, 7.59; N, 4.59. Found: C,

66.57; H, 7.57; N, 4.55.

2

131.85, 131.37 (q, J_FC= 32 Hz), 130.50, 127.73, 127.27, 126.85 (q,

1

³J_FC= 3.8 Hz), 125.16 (q, J_FC= 270 Hz), 121.40, 117.38, 52.07,

41.45, 18.66, 15.45. ¹⁹F NMR (acetone-d₆, 376 MHz): δ −63.62 (3F,

s). Anal. Calcd for C₂₃H₂₁F₃N₂O₂S: C, 61.87; H, 4.74; N, 6.27.

Found: C, 61.69; H, 4.58; N, 6.14.

3-Ethyl-N-(tert-butoxycarbonyl)aniline (33a). 3-Ethylaniline 32a

(3.0 mL, 0.024 mol) and di-tert-butyl dicarbonate (5.840 g, 0.026

mol) were dissolved in anhydrous THF (40 mL) under inert gas. The

mixture was reﬂuxed at 65 °C overnight. The reaction mixture was

then concentrated under reduced pressure, and the resulting residue

was taken up into EtOAc, washed with saturated NaHCO₃solution,

water, and brine, and dried with Na₂SO₄. After ﬁltration and

evaporation of the solvent, the crude product was puriﬁed with ﬂash

chromatography using 10% EtOAc/hexane. The puriﬁed product 33a

was collected as an orange oil containing EtOAc (5.837 g, 0.026 mol,

110%). TLC R_f(25% EtOAc/hexane) = 0.64. ¹H NMR (CDCl₃, 600

MHz): δ 7.20 (1H, m), 7.12 (1H, d, J = 7.7 Hz), 6.88 (1H, d, J = 7.5

Hz), 6.43 (1H, bs), 2.62 (2H, q, J = 7.6 Hz), 1.52 (9H, s), 1.22 (3H,

t, J = 7.6 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 152.91, 145.47,

138.42, 129.00, 122.77, 118.15, 115.97, 80.53, 29.06, 28.50, 15.70.

Anal. Calcd for C₁₃H₁₉NO₂: C, 70.56; H, 8.65; N, 6.33. Found: C,

70.27; H, 8.52; N, 6.10.

3-Ethyl-[N-(tert-butoxycarbonyl)methyl]aniline (34a). 3-Ethyl-N-

tert-butoxycarbonyl)aniline 33a (4.547 g, 20.55 mmol) was dissolved

in anhydrous DMF (50 mL) and cooled in an ice bath. NaH (60%

dispersion in mineral oil, 1.234 g, 30.85 mmol) was added, and the

mixture was stirred for 45 min followed by the addition of

iodomethane (3.8 mL, 61.04 mmol). The reaction mixture was

stirred overnight at room temperature and quenched with 50%

saturated NaHCO₃and extracted with two portions of ether. The

ether extracts were combined and washed with brine, dried with

Na₂SO₄, and concentrated. The crude material was puriﬁed by ﬂash

chromatography using 5% EtOAc/hexane to give the desired product

Methyl 3-Methyl-4-(N-methyl)aminocinnamate (29b). By anal-

ogy to the procedure described in example 29a, methyl ester 28b

(1.359 g, 4.450 mmol) was treated with triﬂuoroacetic acid (7 mL,

91.48 mmol). Flash chromatography was attempted to purify the

crude material along with recrystallization. Puriﬁcation methods were

7023

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

1

34a as a yellow−orange oil (4.318 g, 18.35 mmol, 89.3%). TLC R_f

130.33, 126.58, 126.02 (m), 123.99 (q, J_FC= 271 Hz), 123.96,

112.81, 109.78, 48.65, 38.30, 26.68, 16.42, 15.56. Anal. Calcd for

C₂₂H₂₁F₃N₂OS: C, 63.14; H, 5.06; N, 6.69. Found: C, 62.85; H, 5.19;

N, 6.58.

1

(25% EtOAc/hexane) = 0.62. H NMR (CDCl₃, 600 MHz): δ 7.23

(1H, m), 7.07 (1H, s), 7.04 (1H, d, J = 7.9 Hz), 7.01 (1H, d, J = 7.6

Hz), 3.25 (3H, s), 2.66 (2H, q, J = 7.6 Hz), 1.45 (9H, s), 1.23 (3H, t,

J = 7.6 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 155.00, 144.85, 143.94,

128.53, 125.26, 125.07, 122.80, 80.26, 37.53, 28.90, 28.50, 15.65.

Anal. Calcd for C₁₄H₂₁NO₂: C, 71.46; H, 9.00; N, 5.95. Found: C,

71.17; H, 8.89; N, 5.85.

Ethyl 2-Ethyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl] thiazole]]aminocinnamate (38a). Trie-

thylphosphonoacetate (0.24 mL, 1.21 mmol) was taken up into

anhydrous THF (5 mL) in an ice bath and under inert gas. NaH (60%

dispersion in mineral oil, 0.055 g, 1.38 mmol) was added by brieﬂy

exposing the system to air. In a separate ﬂask, aldehyde 37a (0.385 g,

0.920 mmol) was dissolved in anhydrous THF (5 mL) under inert

gas. After the triethylphosphonoacetate mixture had stirred for 30

min, the aldehyde solution was slowly added. The reaction mixture

was stirred for 3 h at room temperature and upon completion was

quenched with sat. NH₄Cl and was extracted with three portions of

EtOAc. The combined organic layers were washed with brine, dried

with Na₂SO₄, and concentrated. The crude material was puriﬁed by

column chromatography using a gradient of 10−20% EtOAc/hexane

to give the desired product 38a as a light-yellow solid (0.300 g, 0.614

mmol, 66.7%). TLC R_f(25% EtOAc/hexane) = 0.38. Mp 127−128

3-Ethyl-(N-methyl)aniline (35a). 3-Ethyl-[N-(tert-

butoxycarbonyl)methyl]aniline 34a (3.582 g, 15.22 mmol) was

dissolved in anhydrous DCM (40 mL) under inert gas and cooled

in an ice bath. Triﬂuoroacetic acid (10 mL, 0.13 mol) was slowly

added, and the mixture was stirred at rt for 1.5 h. The reaction

mixture was diluted with DCM and neutralized with chilled saturated

NaHCO₃. The organic phase was collected and washed with water

and brine, dried with Na₂SO₄, and concentrated. The crude material

was puriﬁed by column chromatography using a gradient of 5−50%

EtOAc/hexane and appeared to oxidize upon collection to give the

desired product 35a as a crude black oil with trace EtOAc (2.845 g).

This material was used in the next step without further puriﬁcation.

TLC R_f(25% EtOAc/hexane) = 0.61. ¹H NMR (CDCl₃, 600 MHz):

δ 9.66 (1H, bs), 7.35 (1H, m), 7.28 (2H, m), 7.24 (1H, d, J = 7.6

Hz), 2.99 (3H, s), 2.67 (2H, q, J = 7.6 Hz), 1.22 (3H, t, J = 7.6 Hz).

¹³C NMR (CDCl₃, 150 MHz): δ 147.25, 137.64, 130.31, 129.07,

1

°C. H NMR (CDCl₃, 600 MHz): δ 7.95 (3H, m), 7.64 (2H, d, J =

8.3 Hz), 7.56 (1H, d, J = 8.8 Hz), 6.65 (1H, dd, J₁= 8.8 Hz J₂= 2.7

Hz), 6.60 (1H, d, J = 2.7 Hz), 6.24 (1H, d, J = 15.7 Hz), 4.67 (2H, s),

4.25 (2H, q, J = 7.1 Hz), 3.04 (3H, s), 2.77 (2H, q, J = 7.6 Hz), 2.52

(3H, s), 1.33 (3H, t, J = 7.1 Hz), 1.23 (3H, t, J = 7.6 Hz). ¹³C NMR

(CDCl₃, 150 MHz): δ 167.96, 163.57, 150.34, 150.20, 146.00, 141.78,

136.84, 131.45 (q, J_FC= 32.6 Hz), 131.15, 128.23, 126.53, 125.98

(m), 124.05 (q, J_FC= 270 Hz), 122.02, 114.81, 112.97, 111.11,

60.31, 49.03, 38.26, 27.10, 16.28, 15.59, 14.55. Anal. Calcd for

C₂₆H₂₇F₃N₂O₂S: C, 63.92; H, 5.57; N, 5.73. Found: C, 63.66; H,

5.61; N, 5.58.

121.51, 119.34, 38.08, 28.73, 15.25.

3-Ethyl-[N-methyl-N-[(5-methylene-4-methyl-2-(4-

(triﬂuoromethyl)phenyl)]thiazole]aniline (36a). 3-Ethyl-(N-

methyl)aniline 35a (0.693 g, 5.125 mmol), thiazole 4b (1.478 g,

5.069 mmol), and NaI (0.790 g, 5.271 mmol) were dissolved in

anhydrous DMF (20 mL) under inert gas and cooled in an ice bath.

NaH (60% dispersion in mineral oil, 0.314 g, 7.85 mmol) was

carefully added to the reaction by brieﬂy exposing the system to air.

The reaction mixture was stirred for 3 h at room temperature and

quenched with a 50% dilution of sat. NaHCO₃. The mixture was

extracted with three portions of ether that were combined, washed

with brine, dried with Na₂SO₄, and concentrated. The crude material

was puriﬁed by ﬂash chromatography using a gradient of 5−30%

EtOAc/hexane to give the desired product 36a as an orange oil (0.650

g, 1.66 mmol, 32.8%). TLC R_f(25% EtOAc/hexane) = 0.56. ¹H

NMR (CDCl₃, 600 MHz): δ 7.96 (2H, d, J = 8.1 Hz), 7.63 (2H, d, J =

8.3 Hz), 7.19 (1H, m), 6.68 (3H, m), 4.61 (2H, s), 2.97 (3H, s), 2.63

(2H, q, J = 7.6 Hz), 2.50 (3H, s), 1.24 (3H, t, J = 7.6 Hz). ¹³C NMR

(CDCl₃, 150 MHz): δ 163.46, 150.19, 149.15, 145.66, 136.99, 131.30

2

1

2-Ethyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl]thiazole]]aminocinnamic Acid (39a). Ethyl

ester 38a (0.279 g, 0.571 mmol) was dissolved in 95% EtOH (8 mL)

and THF (8 mL). 3 N NaOH (2 mL) was added, and the mixture was

stirred overnight at room temperature. The reaction was determined

to be incomplete, and an additional 3 N NaOH (2 mL) was added.

The mixture was stirred for an additional 24 h and was neutralized

with 1 N HCl to pH 3. The solvent was removed under reduced

pressure, and the residue was taken up into EtOAc, washed with sat.

NaHCO₃and brine, dried with Na₂SO₄, and concentrated. The crude

material was puriﬁed by ﬂash chromatography using a 25−50%

EtOAc/hexane gradient supplemented with dropwise amounts of

acetic acid to give the desired product 39a as a bright-yellow solid

(0.181 g, 0.393 mmol, 68.8%). TLC R_f(50% EtOAc/hexane) = 0.45.

Mp 216−218 °C. ¹H NMR (acetone-d₆, 600 MHz): δ 10.47 (1H, bs),

8.10 (2H, d, J = 8.1 Hz), 7.94 (1H, d, J = 15.7 Hz), 7.78 (2H, d, J =

8.3 Hz), 7.66 (1H, d, J = 8.8 Hz), 6.78 (2H, m), 6.26 (1H, d, J = 15.7

Hz), 4.87 (2H, s), 3.10 (3H, s), 2.77 (2H, q, J = 7.6 Hz), 2.52 (3H, s),

1.20 (3H, t, J = 7.6 Hz). ¹³C NMR (acetone-d₆, 150 MHz): δ 168.37,

163.09, 151.43, 151.30, 146.39, 142.50, 138.07, 132.64, 131.38 (q,

²J_FC= 32 Hz), 128.90, 127.27, 126.85 (m), 125.14 (q, ¹J_FC= 271 Hz),

1

(q, ²J_FC= 33 Hz), 129.40, 126.49, 125.96 (m), 124.08 (q, J_FC= 270

Hz), 118.11, 113.57, 111.36, 50.01, 38.53, 29.44, 15.81, 15.56. Anal.

Calcd for C₂₁H₂₁F₃N₂S: C, 64.60; H, 5.42; N, 7.17. Found: C, 64.31;

H, 5.48; N, 7.01.

2-Ethyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl]thiazole]] Aminobenzaldehyde (37a). An-

hydrous DMF (1.0 mL, 12.92 mmol) was added to reaction ﬂask and

cooled in an ice bath. POCl₃(0.5 mL, 5.364 mmol) was slowly added,

and the mixture was stirred for 15 min. Aniline product 36a (0.612 g,

1.567 mmol) from the previous reaction was dissolved with

anhydrous DMF (1 mL) in a separate ﬂask. The aniline product

solution was carefully transferred to the reaction ﬂask containing the

Vilsmeier reagent. The reaction mixture was stirred for 2.5 h at 70 °C

and then allowed to cool before being poured onto crushed ice. This

mixture was neutralized with 3 N NaOH until a pH of 10 was

reached. This mixture produced no precipitate and was extracted with

two portions of ether. The ether extracts were combined, washed with

brine, dried with Na₂SO₄, and concentrated. The crude material was

puriﬁed by ﬂash chromatography using a gradient of 5−20% EtOAc/

hexane to give the desired product 37a as a yellow−orange solid

(0.432 g, 1.032 mmol, 65.9%). TLC R_f(25% EtOAc/hexane) = 0.28.

121.95, 114.81, 113.80, 112.04, 48.99, 38.39, 27.43, 16.63, 15.51. ¹⁹F

NMR (acetone-d₆, 376 MHz): δ −63.66. Anal. Calcd for

C₂₄H₂₃F₃N₂O₂S: C, 62.60; H, 5.03; N, 6.08. Found: C, 62.45; H,

5.07; N, 6.24.

3-Isopropyl-N-(tert-butoxycarbonyl)aniline (33b). By analogy to

the procedure described in example 33a, 3-isopropylaniline 32b

(1.087 g, 8.039 mmol) was treated with di-tert-butyl dicarbonate

(1.988 g, 9.109 mmol), and the crude product was puriﬁed with ﬂash

chromatography using a 5% EtOAc/hexane mobile phase to give the

product 33b as a light-orange solid (1.660 g, 7.054 mmol, 87.7%).

TLC R_f(25% EtOAc/hexane) = 0.70. Mp 44−45 °C. ¹H NMR

(CDCl₃, 600 MHz): δ 7.25 (1H, bs), 7.21 (1H, m), 7.15 (1H, d, J =

7.4 Hz), 6.91 (1H, d, J = 7.5 Hz), 6.47 (1H, br-s), 2.88 (1H, sept., J =

7.0 Hz), 1.52 (9H, s), 1.24 (6H, d, J = 7.0 Hz). ¹³C NMR (CDCl₃,

150 MHz): δ 152.92, 150.08, 138.40, 128.99, 121.28, 116.85, 116.17,

80.48, 34.31, 28.49, 24.07. Anal. Calcd for C₁₄H₂₁NO₂: C, 71.46; H,

9.00; N, 5.95. Found: C, 71.72; H, 9.03; N, 5.86.

1

Mp 97−98 °C. H NMR (CDCl₃, 600 MHz): δ 10.02 (1H, s), 7.96

(2H, d, J = 8.2 Hz), 7.74 (1H, d, J = 8.7 Hz), 7.64 (2H, d, J = 8.3 Hz),

6.70 (1H, dd, J₁= 8.7 Hz J₂= 2.6 Hz), 6.58 (1H, d, J = 2.5 Hz), 4.74

(2H, s), 3.12 (3H, s), 3.03 (2H, q, J = 7.5 Hz), 2.53 (3H, s), 1.27

(3H, t, J = 7.5 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 190.26, 163.76,

2

152.40, 150.51, 149.64, 136.59, 134.89, 131.60 (q, J_FC= 33 Hz),

7024

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

3-Isopropyl-[N-methyl-N-(tert-butoxycarbonyl)]aniline (34b). By

analogy to the procedure described in example 34a, 3-isopropyl-N-

(tert-butoxycarbonyl)aniline 33b (3.148 g, 13.38 mmol) was treated

with NaH (60% dispersion in oil, 0.805 g, 20.13 mmol) and

iodomethane (2.5 mL, 40.16 mmol). The crude material was puriﬁed

by ﬂash chromatography using a 100% hexane to 5% EtOAc/hexane

gradient giving the desired compound 34b as an orange oil (2.769 g,

11.10 mmol, 83.0%). TLC R_f(25% EtOAc/hexane) = 0.73. ¹H NMR

(CDCl₃, 600 MHz): δ 7.24 (1H, t, J = 7.8 Hz), 7.09 (1H, bs), 7.03

(2H, dd, J₁= 7.7 Hz, J₂= 1.6 Hz), 3.26 (3H, s), 2.89 (1H, sept., J =

6.9 Hz), 1.45 (9H, s), 1.25 (6H, d, J = 6.9 Hz). ¹³C NMR (CDCl₃,

150 MHz): δ 155.00, 149.45, 143.91, 128.50, 124.03, 123.61, 122.77,

80.23, 37.54, 34.18, 28.51, 24.09. Anal. Calcd for C₁₅H₂₃NO₂: C,

72.25; H, 9.30; N, 5.62. Found: C, 72.53; H, 9.27; N, 5.71.

3-Isopropyl-(N-methyl)aniline (35b). By analogy to the procedure

described in example 35a, 3-isopropyl-[N-methyl-N-(tert-

butoxycarbonyl)]aniline 34b (2.409 g, 9.661 mmol) was treated

with triﬂuoroacetic acid (7.4 mL, 96.70 mmol). The crude material

was puriﬁed by ﬂash chromatography using a gradient of 5−50%

EtOAc/hexane and appeared to oxidize upon collection to give the

desired product 35b as a crude black oil with trace EtOAc (1.733 g).

This material was used in the next step without further puriﬁcation.

TLC R_f(25% EtOAc/hexane) = 0.51. ¹H NMR (CDCl₃, 600 MHz):

δ 7.36 (1H, m), 7.31 (1H, s), 7.27 (1H, m), 2.99 (3H, s), 2.92 (1H,

sept., J = 6.9 Hz), 1.24 (6H, t, J = 6.9 Hz). ¹³C NMR (CDCl₃, 150

MHz): δ 151.94, 137.93, 130.31, 127.44, 120.05, 119.33, 37.97, 34.15,

23.78.

15.7 Hz), 7.95 (2H, d, J = 8.2 Hz), 7.63 (2H, d, J = 8.3 Hz), 7.53 (1H,

d, J = 8.8 Hz), 6.69 (1H, d, J = 2.6 Hz), 6.65 (1H, dd, J₁= 8.7 Hz J₂=

2.6 Hz), 6.23 (1H, d, J = 15.6 Hz), 4.67 (2H, s), 4.25 (2H, q, J = 7.1

Hz), 3.38 (1H, sept., J = 6.8 Hz), 3.05 (3H, s), 2.52 (3H, s), 1.33

(3H, t, J = 7.1 Hz), 1.25 (6H, d, J = 6.8 Hz). ¹³C NMR (CDCl₃, 150

MHz): δ 167.93, 163.62, 150.36, 150.22, 149.93, 141.78, 136.80,

2

131.45 (q, J_FC= 32.7 Hz), 128.31, 126.54, 126.00 (m), 124.04 (q,

¹J_FC= 271 Hz), 121.88, 115.40, 111.04, 109.44, 60.33, 49.18, 38.37,

29.22, 23.87, 15.60, 14.55. Anal. Calcd for C₂₇H₂₉F₃N₂O₂S: C, 64.52;

H, 5.82; N, 5.57. Found: C, 64.45; H, 5.83; N, 5.63.

2-Isopropyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl] thiazole]]aminocinnamic Acid (39b). By

analogy to the procedure described in example 39a, ethyl ester 38b

(0.327 g, 0.651 mmol) was treated with 3 N NaOH. The crude

material was puriﬁed by column chromatography using a 15−50%

EtOAc/hexane gradient supplemented with dropwise amounts of

acetic acid to give the desired product 39b as a yellow solid (0.179 g,

0.377 mmol, 57.9%). TLC R_f(50% EtOAc/hexane) = 0.46. Mp 218−

1

220 °C. H NMR (acetone-d₆, 600 MHz): δ 10.50 (1H, bs), 8.10

(2H, d, J = 8.2 Hz), 8.04 (1H, d, J = 15.8 Hz), 7.78 (2H, d, J = 8.3

Hz), 7.64 (1H, d, J = 8.8 Hz), 6.81 (1H, d, J = 2.6 Hz), 6.77 (1H, dd,

J₁= 8.8 Hz, J₂= 2.6 Hz), 6.25 (1H, d, J = 15.6 Hz), 4.88 (2H, s), 3.37

(1H, sept., J = 6.8 Hz), 3.12 (3H, s), 2.53 (3H, s), 1.26 (3H, s) 1.25

(3H, s). ¹³C NMR (acetone-d₆, 150 MHz): δ 168.36, 163.12, 151.38,

2

150.21, 142.45, 138.07, 132.73, 131.39 (q, J_FC= 32.3 Hz), 128.97,

1

127.27, 126.84 (m), 125.14 (q, J_FC= 270 Hz), 121.73, 115.37,

111.89, 110.09, 49.13, 38.51, 23.94, 15.54. ¹⁹F NMR (acetone-d₆, 376

MHz): δ −63.66. Anal. Calcd for C₂₅H₂₅F₃N₂O₂S: C, 63.28; H, 5.31;

N, 5.90. Found: C, 63.18; H, 5.33; N, 5.98.

3-Isopropyl-[N-methyl-N-[(5-methylene-4-methyl-2-(4-

(triﬂuoromethyl)phenyl)]thiazole]aniline (36b). By analogy to the

procedure described in example 36a, 3-isopropyl-(N-methyl)aniline

35b (1.352 g, 9.059 mmol) and thiazole 4b (1.755 g, 6.016 mmol)

were treated with NaI (0.990 g, 6.604 mmol) and NaH (60%

dispersion in mineral oil, 0.363 g, 9.075 mmol). The crude material

was puriﬁed by ﬂash chromatography using a gradient of 100% hexane

to 5% EtOAc/hexane to give the desired product 36b as an orange oil

(1.657 g, 4.096 mmol, 68.1%). TLC R_f(25% EtOAc/hexane) = 0.58.

¹H NMR (CDCl₃, 600 MHz): δ 7.97 (2H, d, J = 8.2 Hz), 7.64 (2H, d,

Ethyl 5-Methyl-2-[4-(triﬂuoromethyl)phenyl]-1,3-thiazole-4-car-

boxylate (40). 2-Ketobutyric acid (3.245 g, 0.032 mol) was dissolved

in anhydrous DCM (30 mL) under nitrogen gas. Bromine (1.7 mL,

0.033 mol) was slowly added, and the reaction mixture was stirred at

room temperature for 15 min followed by evaporation of the solvent

under reduced pressure. Toluene was added and immediately

removed under reduced pressure. 4-(Triﬂuoromethyl)thiobenzamide

(6.224 g, 0.303 mol) was added to the residue, dissolved in 95%

ethanol, and reﬂuxed overnight. The solvent was then removed, and

the residue was puriﬁed by ﬂash chromatography using a DCM

mobile phase to give 40 as a white solid (4.140 g, 13.13 mmol,

J = 8.3 Hz), 7.21 (1H, t, J = 8.1 Hz), 6.72 (2H, m), 6.68 (1H, m),

4.61 (2H, s), 2.98 (3H, s), 2.87 (1H, sept., J = 6.9 Hz), 2.49 (3H, s),

1.27 (6H, m). ¹³C NMR (CDCl₃, 150 MHz): δ 163.55, 162.66,

1

2

43.3%). TLC (25% EtOAc/hexane) = 0.52. Mp 91 °C. H NMR

150.33, 136.97, 133.67, 131.31 (q, J_FC= 32.4 Hz), 129.40, 127.58,

1

(CDCl₃, 600 MHz): δ 8.03 (2H, d, J = 8.1 Hz), 7.67 (2H, d, J = 8.16

Hz), 4.44 (2H, q, J = 7.14 Hz), 2.80 (3H, s), 1.43 (3H, s). ¹³C NMR

(CDCl₃, 150 MHz): δ 162.51, 161.90, 145.57, 142.91, 136.08, 131.93

126.51, 125.97(m), 124.08 (q, J_FC= 271 Hz), 116.74, 112.41,

111.69, 50.17, 38.62, 34.65, 24.18, 15.57. Anal. Calcd for

C₂₂H₂₃F₃N₂S: C, 65.33; H, 5.73; N, 6.93. Found: C, 65.06; H,

5.67; N, 6.83.

(q, ²J_FC= 33 Hz), 126.98, 126.00 (q, ³J_FC= 4.1 Hz), 123.93 (q, ¹J_FC

=

270 Hz), 61.44, 14.47, 13.54. Anal. Calcd for C₁₄H₁₂F₃NO₂S: C,

53.33; H, 3.84; N, 4.44. Found: C, 53.19; H, 3.93; N, 4.35.

2-Isopropyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl]thiazole]] Aminobenzaldehyde (37b). By

analogy to the procedure described in example 37a, aniline product

36b (1.156 g, 2.858 mmol) was treated with Vilsmeier reagent.

Puriﬁcation of the crude material was attempted using ﬂash

chromatography with a gradient of 5−10% EtOAc/hexane. The

yellow oil 37b (0.414 g, 0.957 mmol, 33.5%) was moved to the next

step without further puriﬁcation. TLC R_f(25% EtOAc/hexane) =

0.28. ¹H NMR (CDCl₃, 600 MHz): δ 10.08 (1H, s), 7.96 (2H, d, J =

8.2 Hz), 7.74 (1H, d, J = 8.8 Hz), 7.64 (2H, d, J = 8.3 Hz), 6.72 (1H,

d, J = 2.5 Hz), 6.68 (1H, dd, J₁= 8.7 Hz J₂= 2.6 Hz), 4.74 (2H, s),

4.01 (1H, sept. J = 6.8 Hz), 3.13 (3H, s), 2.54 (3H, s), 1.28 (6H, d, J

= 6.8 Hz). ¹³C NMR (CDCl₃, 150 MHz): δ 190.30, 163.81, 154.01,

4-Hydroxymethyl-5-methyl-2-[4-(triﬂuoromethyl)phenyl]-

thiazole (41). Ethyl ester 40 (3.284 g, 10.42 mmol) was dissolved in

anhydrous THF (30 mL) under nitrogen gas. The mixture was cooled

in an ice bath, and a chilled 2.0 M solution of LiAlH₄in THF (5.2

mL, 10.4 mmol) was slowly added. The reaction mixture was stirred

in an ice bath for 15 min before removal from the ice bath to allow

warming to room temperature. The reaction mixture stirred for 90

min and was quenched with water and dried with Na₂SO₄. After

ﬁltration, the ﬁltrate was concentrated, and the residue was puriﬁed by

ﬂash chromatography using a 10−50% EtOAc/hexane mobile-phase

gradient to give 41 as a yellow solid (1.663 g, 6.086 mmol, 58.5%).

1

2

TLC (25% EtOAc/hexane) = 0.16. Mp 112 °C. H NMR (CDCl₃,

152.53, 150.52, 136.59, 135.09, 131.62 (q, J_FC= 32.5 Hz), 130.39,

1

600 MHz): δ 7.99 (2H, d, J = 8.2 Hz), 7.68 (2H, d, J = 8.3 Hz), 4.74

(2H, s), 2.50 (3H, s). ¹³C NMR (CDCl₃, 150 MHz): δ 162.99,

152.58, 136.67, 131.50 (q, ²J_FC= 32.5 Hz), 130.70, 126.52, 126.07 (q,

126.60, 126.04 (m), 124.00 (q, J_FC= 270 Hz), 123.49, 109.64,

108.85, 48.74, 38.37, 28.07, 23.91, 15.58.

Ethyl 2-Isopropyl-4-[N-methyl-N-[5-methylene-4-methyl-2-[4-

(triﬂuoromethyl)phenyl] thiazole]]aminocinnamate (38b). By

analogy to the procedure described in example 38a, aldehyde 37b

was treated with triethylphosphonoacetate (0.28 mL, 1.41 mmol) and

NaH (60% dispersion in mineral oil, 0.065 g, 1.63 mmol). The crude

material was puriﬁed by ﬂash chromatography using a gradient of 10−

15% EtOAc/hexane to give the desired product 38b as a yellow solid

(0.371 g, 0.739 mmol, 68.0%). TLC R_f(25% EtOAc/hexane) = 0.35.

1

³J_FC= 3.5 Hz), 124.05 (q, J_FC= 271 Hz), 58.64, 11.32. Anal. Calcd

for C₁₂H₁₀F₃NOS: C, 52.74; H, 3.69; N, 5.13. Found: C, 52.67; H,

3.77; N, 5.11.

4-Chloromethyl-5-methyl-2-[4-(triﬂuoromethyl)phenyl]thiazole

(42). Alcohol 41 (1.606 g, 5.877 mmol) was dissolved in anhydrous

DCM (40 mL) under nitrogen gas, and triethylamine (1.65 mL,

11.838 mmol) was added at room temperature. The mixture was

cooled in an ice bath, methanesulfonyl chloride (0.69 mL, 8.915

1

Mp 99−100 °C. H NMR (CDCl₃, 600 MHz): δ 8.05 (1H, d, J =

7025

https://doi.org/10.1021/acs.jmedchem.1c00560

J. Med. Chem. 2021, 64, 6996−7032

Article Doi

Synthesis and Evaluation of PPARδAgonists That Promote Osteogenesis in a Human Mesenchymal Stem Cell Culture and in a Mouse Model of Human Osteoporosis

DOI: 10.1021/acs.jmedchem.1c00560

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.jmedchem.1c00560

Full text of DOI:10.1021/acs.jmedchem.1c00560

Products guided by the article

R&D Labs maybe for 870788-49-1

Relevant to this article

Hot Product