Discovery of 2-[3,5-dichloro-4-(5-isopropyl-6-oxo-1,6-dihydropyridazin-3- yloxy)phenyl]-3,5-dioxo-2,3,4,5-tetrahydro[1,2,4]triazine-6-carbonitrile (MGL-3196), a highly selective thyroid hormone receptor β agonist in clinical trials for the treatment of dyslipidemia

Article abstract of DOI:10.1021/jm4019299

The beneficial effects of thyroid hormone (TH) on lipid levels are primarily due to its action at the thyroid hormone receptor β (THR-β) in the liver, while adverse effects, including cardiac effects, are mediated by thyroid hormone receptor α (THR-α). A pyridazinone series has been identified that is significantly more THR-β selective than earlier analogues. Optimization of this series by the addition of a cyanoazauracil substituent improved both the potency and selectivity and led to MGL-3196 (53), which is 28-fold selective for THR-β over THR-α in a functional assay. Compound 53 showed outstanding safety in a rat heart model and was efficacious in a preclinical model at doses that showed no impact on the central thyroid axis. In reported studies in healthy volunteers, 53 exhibited an excellent safety profile and decreased LDL cholesterol (LDL-C) and triglycerides (TG) at once daily oral doses of 50 mg or higher given for 2 weeks.

Full text of DOI:10.1021/jm4019299

Drug Annotation
pubs.acs.org/jmc
Discovery of 2‑[3,5-Dichloro-4-(5-isopropyl-6-oxo-1,6-
dihydropyridazin-3-yloxy)phenyl]-3,5-dioxo-2,3,4,5-
tetrahydro[1,2,4]triazine-6-carbonitrile (MGL-3196), a Highly
Selective Thyroid Hormone Receptor β Agonist in Clinical Trials for
the Treatment of Dyslipidemia
Martha J. Kelly,*,† Sherrie Pietranico-Cole,‡ J. Douglas Larigan,‡ Nancy-Ellen Haynes,‡
Charles H. Reynolds,§ Nathan Scott,‡ John Vermeulen,‡ Mark Dvorozniak,‡ Karin Conde-Knape,‡
Kuo-Sen Huang,‡ Sung-Sau So,‡ Kshitij Thakkar,‡ Yimin Qian,‡ Bruce Banner,‡ Frank Mennona,‡
Sara Danzi,∥ Irwin Klein,⊥ Rebecca Taub,† and Jefferson Tilley‡
†Madrigal Pharmaceuticals, Inc., Fort Washington, Pennsylvania 19034, United States
‡Hoffmann-La Roche, Inc., Nutley, New Jersey 07110, United States
§Gfree Bio LLC, Doylestown, Pennsylvania 18902, United States
∥Queensborough Community College, Bayside, New York 11364, United States
⊥New York University School of Medicine, New York, New York 10016, United States
S
* Supporting Information
ABSTRACT: The beneficial effects of thyroid hormone (TH)
on lipid levels are primarily due to its action at the thyroid
hormone receptor β (THR-β) in the liver, while adverse
effects, including cardiac effects, are mediated by thyroid
hormone receptor α (THR-α). A pyridazinone series has been
identified that is significantly more THR-β selective than
earlier analogues. Optimization of this series by the addition of
a cyanoazauracil substituent improved both the potency and
selectivity and led to MGL-3196 (53), which is 28-fold
selective for THR-β over THR-α in a functional assay.
Compound 53 showed outstanding safety in a rat heart model
and was efficacious in a preclinical model at doses that showed no impact on the central thyroid axis. In reported studies in
healthy volunteers, 53 exhibited an excellent safety profile and decreased LDL cholesterol (LDL-C) and triglycerides (TG) at
once daily oral doses of 50 mg or higher given for 2 weeks.
increased rates of atherosclerosis, while excessive levels of
thyroid hormone (TH) can lead to adverse effects, particularly
in heart and bone. The beneficial metabolic effects of TH are
mediated by the thyroid hormone receptor β isoform (THR-β),
the predominant liver TH receptor.5−8 The adverse heart and
bone effects are primarily due to the interaction of TH with the
thyroid hormone receptor α isoform (THR-α).9−11 Thus, the
search for therapeutic agents has focused on finding agonists
with selectivity for both the THR-β and the liver. The ligand
binding domains of the TH-α and -β receptors differ by only a
single amino acid, Asn331 (THR-β)/Ser277 (THR-α), that is
capable of making a hydrogen bond with the acidic moiety of 1.
As a result, much of the work on THR-β selective compounds
reported to date has focused on close analogues of 1 bearing
INTRODUCTION
■
Despite the success of statins in lowering LDL cholesterol
(LDL-C) and thereby reducing cardiovascular disease, there is a
need for new drugs for the treatment of dyslipidemia. Up to
10% of hypercholesterolemic patients do not tolerate statins,
and roughly 70% of high risk cardiovascular patients do not
achieve LDL-C goals.1,2 Diabetes is characterized by a particular
form of dyslipidemia including elevated triglycerides (TG), low
HDL cholesterol (HDL-C), and nonalcoholic fatty liver disease
(NAFLD), and 60−75% of diabetics die of macrovascular
cardiovascular disease.3 Drugs that lower fatty acid accumu-
lation and degeneration in the liver may also be useful in the
treatment of nonalcoholic steatohepatitis (NASH).4
The beneficial metabolic effects on cholesterol and TG levels
of triiodothyronine (T3) 1 and its prohormone thyroxine (T4)
have been known for many years. In part because of associated
hypercholesterolemia, hypothyroidism is associated with
Received: December 16, 2013
Published: April 8, 2014
© 2014 American Chemical Society
3912
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
variations in the acidic region capable of selectively interacting
with these residues.12 Chart 1 shows a number of compounds
modifications that we investigated, the pyridazinone analogues
represented by 8 proved to be the most promising, and this
novel heterocycle provided compounds with improved
selectivity for THR-β that form the basis of this paper. This
work culminated in the identification of MGL-3196 (53) which
is currently in clinical development.20
Chart 1
CHEMISTRY
■
In general, the core bicyclic framework of the compounds
shown in Table 1 was prepared by the condensation of the
appropriate aryl oxygen, carbon, or sulfur nucleophile with the
isopropyldichloropyridazine 1121 followed by functional group
transformations as shown in Schemes 1−7. The phenyl acetates
10a and 10b were prepared from the corresponding phenols 9a
and 9b in four steps as shown in Scheme 1 using the general
method described by Gardner et al.22 Condensation of the
phenols 10a and 10b with 11 was effected in the presence of
potassium carbonate as the base and a catalytic amount of
cuprous iodide in DMSO at 90 °C as generally described by
Chen et al.23 Finally, ester hydrolysis was accomplished under
basic conditions and conversion of the chloropyridazines 12a
and 12b to the corresponding pyridazinones 13 and 14 was
carried out by treatment with sodium acetate in acetic acid at
100 °C.24
As shown in Scheme 2, phenyl acetate 15 was chlorinated
with sulfuryl chloride followed by reduction with LAH to give
dichlorophenol 18a. The corresponding 2,6-dibromophenols
18b and 18c were prepared by bromination of the phenols 16
and 17 with bromine. The key condensation of these
compounds with 11 was performed by reaction with potassium
tert-butoxide in dimethylacetamide at 140 °C overnight in 20−
49% yields. Oxidation of the resulting alcohols using Jones
reagent (19a and 19c) or 2,2,6,6-tetramethylpiperidin-1-yl)oxy
(TEMPO) (19b)25 followed by hydrolysis of the chloropyr-
idazine under the conditions noted above provided the 4-
substituted 3,5-dihalophenylalkanoic acids 20−22. For the
synthesis of N-methylpyridazinone 23, hydrolysis of chloropyr-
idazine 19a was accompanied by acylation of the primary
alcohol. The pyridazinone was methylated with iodomethane in
the presence of potassium carbonate, and the acetate was
hydrolyzed with sodium hydroxide in methanol. Jones
oxidation completed the synthesis of 23.
For the synthesis of the glycine and oxamic acid derivatives
26 and 27, aniline 24 was condensed with 11 under the
conditions described in Scheme 1 to give a 50% yield of the
ether 25 (Scheme 3). Hydrolysis of the chloropyridazine was
accompanied by partial acetylation of the aniline, necessitating a
hydrolysis step prior to reductive amination of glyoxylic acid
mediated by commercially available resin based MP-cyanobor-
ohydride in the presence of acetic acid to provide 26 directly.
Acylation of 25 with methyloxalyl chloride, ester hydrolysis,
and hydrolysis of the chloropyridazine provided the oxamic
acid 27.
illustrating the range of TH mimetic structures. GC-1 (2)6,7
and KB-1415 (3) represent examples whose pharmacology has
been studied extensively. The azauracil 4 represents a further,
unusual variation on the acid replacements in the arena of
thyroid hormone receptor (THR) mimetics.13
A number of THR-β agonists, including 2, 5 (eprotirome),
and a prodrug of 6, have been evaluated in clinical trials
providing validation of their effects on LDL-C and TG
lowering. In human clinical studies, once daily oral doses of
up to 200 μg of 5 (eprotirome, KB2115) lowered LDL-C by up
to 40%,14 while daily doses of 25−100 μg of 5 given for 12
weeks to patients already receiving statins lowered LDL-C by
up to 32%, with no change in thyroid stimulating hormone
(TSH) or T3 reported.15 In a 2-week dose-ascending phase 1
study, a prodrug of 6, at doses of 2.5 and 5 mg, reduced TG and
LDL-C levels by 30% and 15−41%, respectively.16 However,
clinical development of these analogues was discontinued.
Some safety issues were noted that included low margins
relative to doses that suppress the central thyroid axis, elevated
liver enzymes in phase 1 and 2 studies, and in the case of 5,
cartilage damage in preclinical toxicology studies.
In the area of nuclear hormone receptor agonists, it is
difficult to predict how subtle alterations in a ligand structure
might impact its biological profile. In an effort to identify
compounds that were more THR-β selective and potentially
safer, we were interested in profiling analogues of T3 that did
not contain the very electron rich phenol ring, which was a
potential metabolic and safety liability. Indeed, it is reported
that in early clinical trials a potentially mutagenic nitrated
derivative of 5, which is thought to be formed by the free
radical nitration of the phenolic ring with nitrites at gastric pH,
was observed in plasma.17 A series of patent applications from
Bayer describing compounds such as the indole 7 as potent TH
agonists illustrated that the T3 phenolic ring could be replaced
by a heterocycle.18,19 Of several electron deficient heterocyclic
Schemes 4 and 5 show the preparation of the carbon linked
analogues 32 and 38. Selective displacement of the para-
chlorine of nitrobenzene 28 by the anion of tert-butyl
cyanoacetate provided the alkylated intermediate. Reduction
of the nitro group with stannous chloride in ethanolic
hydrochloric acid was accompanied by ester hydrolysis and
decarboxylation to give aniline 29.26 Condensation of 29 with
11 followed by hydrolysis provided the pyridazinone 30.
Diazotization of 30 followed by treatment with copper bromide
gave the intermediate aryl bromide, which was subjected to
3913
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
a
Table 1. Structure−Activity Relationship of Pyridazinone Based TH Agonists
a
N/D is not determined. MNT is in vitro micronucleus test. Superscript b indicates average of triplicate determinations. Superscript c indicates that
selectivity is normalized for the selectivity of T3 in the same assay. Superscript d indicates that T3 was run in every assay; the values for THR-β range
from 0.12 to 0.024 μM, and the values for THR-α range from 0.003 to 0.10 μM.
3914
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
a
a
Scheme 1
Scheme 3
a
a
Reagents and conditions: (a) 11, K2CO3, CuI, DMSO, 90 °C, 24 h,
50%; (b) NaOAc, HOAc, 100 °C, 24 h; (c) 1 N NaOH/MeOH, 100
°C, 24 h, 57% for the two steps; (d) glyoxylic acid, MP-(CN)BH3,
HOAc/MeOH/CH2Cl2, MgSO4, 50 °C, 24 h, 17%; (e) methyloxalyl
chloride, DIPEA, THF, 76%; (f) 1 N NaOH, MeOH, 98%; (g)
NaOAc, HOAc, 100 °C, 24 h, 56%.
Reagents and conditions: (a) 40% aq dimethylamine, formaldehyde,
EtOH, reflux, 46−66%; (b) (1) iodomethane, ether, (2) NaCN,
EtOH, reflux, 55−60% for the two steps; (c) KOH, H2O, ethylene
glycol dimethyl ether, reflux, X = CH3 5%, X = Cl, 97%; (d) MeOH,
H2SO4, 70 °C, 86−91%; (e) 11, K2CO3, CuI, DMSO, 90 °C, 24 h,
67−78% ; (f) NaOH, MeOH, 100%; (g) NaOAc, HOAc, 100 °C, 24−
48 h, X = CH3 65%, X = Cl, 32%.
a
Scheme 4
a
Scheme 2
a
Reagents and conditions: (a) tert-butyl cyanoacetate, K2CO3, DMF,
a
50 °C, 18 h; (b) SnCl2, EtOH, HCl, 110 °C, 58% (two steps); (c) 11,
t-BuOK, THF, 60 °C, 1 h, 77%; (d) HCl/HOAc/H2O, 120 °C, 24 h,
76%; (e) NaNO2, H2SO4, HOAc, then CuBr/HBr, 100 °C, 51%; (f)
Pd(OAc)2, dppp, Et3N, CO, MeOH, acetonitrile, 45 psi, 90 °C, 90%;
(g) DIBALH, THF, 24 h; (h) CBr4, PPh3, CH2Cl2, 0 °C, 36% (two
steps); (i) NaCN, H2SO4, DMSO, 50 °C; (j) HCl, reflux, 13% for the
two steps.
Reagents and conditions: (a) SO2Cl2, diisobutylamine, toluene, 83%;
(b) LAH, THF, 100%; (c) (1) Br2, HOAc; (2) THF, 4 N NaOH, 82−
99%; (d) 11, t-BuOK, dimethylacetamide, 140 °C, 24 h, 20−49%; (e)
Jones reagent, acetone, 0−10 °C, 95% for 20 and 22, 19% for 23; (f)
NaOAc, HOAc, 114−120 °C, 24 h, 65−100%, 73% over two steps for
21; (g) TEMPO, NaClO2, NaOCl, H2O, acetonitrile, CH2Cl2,
Na3PO4, pH 6.7, 73%; (h) K2CO3, CH3I, 40 °C, 4 h, then 25 °C,
24 h, 72%; (i) NaOH, MeOH, 25 °C, 24 h, 97%.
of sodium hydride in DMF at 95 °C gave the carbon linked
THP-derivative 36. A related condensation has been reported
previously.26 This material was subjected to the standard
chloropyridazine hydrolysis conditions followed by treatment
with hydrochloric acid in acetic acid at reflux to simultaneously
remove the protecting tetrahydropyranyl group and convert the
resulting primary alcohol to the corresponding benzylic
chloride 37 in 36% yield for the two steps. Conversion of 37
to 38 was carried out in the same manner as the conversion of
the benzyl bromide to 32 (Scheme 4).
The sulfur linked analogues 42−44 were available from
thiophenol 40, which was prepared from 18a (Scheme 6).
Coupling of 40 with 11 mediated by potassium carbonate in
DMSO gave the pyridazine 41. Further conversations involved
palladium-catalyzed carbonylation in methanol followed by
treatment with DIBALH in tetrahydrofuran to yield alcohol 31.
Carbon tetrabromide and triphenylphosphine in methylene
chloride was used to convert 31 to the corresponding benzyl
bromide. Bromide displacement with sodium cyanide27
followed by hydrolysis of the intermediate nitrile gave 32.
Free radical bromination of methyl 3,5-dibromo-4-methyl-
benzoate 33 followed by DIBALH reduction of the ester gave
the bromomethyl derivative 34. Displacement of the newly
introduced bromine atom with sodium cyanide in DMSO in
the presence of sulfuric acid and protection of the primary
alcohol as the THP ether afforded the phenylacetonitrile 35.
Condensation of 35 with dichloropyridazine 11 in the presence
3915
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
a
give the azauracils 57 and 58 was accomplished in two steps by
hydrolysis to the corresponding acid mediated by hydrochloric
acid in acetic acid at reflux and decarboxylation by means of
treatment with mercaptoacetic acid at elevated temperature.29
Scheme 5
RESULTS AND DISCUSSION
■
Activity in in Vitro Functional Assays. Most of the
selectivity reported for thyromimetics in the literature is derived
from assessment of the compound’s relative binding affinities
for THR-β and THR-α. While this information is useful, it does
not provide information on their ability to act as agonists. Thus,
we decided to implement an in vitro functional assay as a
potentially more relevant indicator of selectivity. A cell free
coactivator recruitment assay was developed for both THR-β
and THR-α, providing information on the functional potency
and selectivity of analogues. In this assay, T3 or compound is
allowed to bind to his-tagged THR in the presence of RXR.
The ability of this liganded heterodimeric complex to recruit
biotin-linked coactivator peptide GRIP1 (SRC-2) is measured
with FRET. The activity of the compounds described above for
THR-β and THR-α is reported in Table 1. The selectivity
reported for each compound is normalized for the selectivity of
T3 run in the same assay.
Two well-studied compounds, 2 and 3, have been reported
to be 10-fold30 and 14-fold31 β selective, respectively, in binding
assays. These compounds were unselective in the coactivator
recruitment assay. In a reporter cell assay,31 3 was only 3-fold β
selective, which illustrates that the selectivity in binding and
functional assays can be quite different. While 5 and 6 were not
tested in our assay, 5 is reported to have “modestly higher”
THR-β selectivity,15 and 6 was reported to be 12-fold selective
in a binding assay,32 suggesting that these analogues did not
have high THR-β selectivity.
a
Reagents and conditions: (a) NBS, AIBN, CCl4, reflux, 99%; (b)
DIBAL-H, THF, 0 °C, 94%; (c) NaCN, H2SO4, DMSO, 78%; (d)
dihydropyran, p-toluenesulfonic acid, CH2Cl2, 85%; (e) 11, NaH,
DMF, 95 °C, 65%; (f) (1) NaOAc, HOAc, reflux, (2) HCl, AcOH,
reflux, 36% for the two steps; (g) NaCN, H2SO4, DMSO, 60 °C, 67%;
(h) HCl, reflux, 49%.
a
Scheme 6
Replacement of the phenol of 3 with a pyridazinone gave 20
which was 10-fold more selective for THR-β than 3. While 20
was 100-fold less potent than 3 (2.38 μM versus 0.023 μM), the
potency was good enough to provide a reasonable starting
point for optimization of this series directed toward improving
potency and selectivity. Replacing one or both of the chlorines
with methyl (13, 14) resulted in a slight loss of both potency
and selectivity. On the other hand, the dibrominated analogue
21 was 5-fold more potent and equally as selective as 20.
Replacing the oxygen linker of analogues 20 and 21 with a
methylene linker gave compounds 32 and 38 that were
comparable in potency but somewhat less selective. The sulfur,
sulfoxide, and sulfone linkers (42−44) were all less selective
than the oxygen linker. While the dibrominated analogues are
more potent than the dichlorinated analogues, concerns about
the general safety of aromatic brominated compounds led us to
focus on the dichlorinated analogues.
Modification of the acetic acid substituent potentiates the
potency and selectivity of the compounds. The propionic acid
analogue 22 was 6-fold less selective than the acetic acid
analogue 21. The amino acetic acid and amino oxalic acid
analogues 26 and 27 were more potent, but less selective, than
20.
A large increase in potency was obtained when the acyclic
acidic groups were replaced with an azauracil. Azauracil 57 was
20-fold more potent than 20 and slightly more selective. A
further increase in selectivity was obtained by substituting the
azauracil with a cyano group. This substituent increased the
selectivity from 12-fold (57) to 28-fold (53) with minimal
impact on potency (53, 0.21 μM; 57, 0.12 μM). As with the
a
Reagents and conditions: (a) t-Bu(Ph)2SiCl, Et3N, CH2Cl2, 60%; (b)
39, DABCO, DMF, 90%; (c) 198 °C, 70%; (d) 3 N KOH, EtOH, 95
°C, 76%; (e) 11, K2CO3, DMSO, 90 °C, 52%; (f) Jones reagent,
acetone, 0 °C, 89%; (g) HOAc, NaOAc, 100 °C, 17%; (h) 30% H2O2,
HCO2H, 0 °C, 41% 43 and 27% 44.
hydrolysis and Jones oxidation to provide 42. Oxidation of 42
with 30% hydrogen peroxide in formic acid at 0 °C gave a
separable mixture of the sulfoxide 43 and sulfone 44.
Scheme 7 outlines the preparation of the oxygen and carbon
linked azauracils 53−58. For the synthesis of the intermediate
N-methylpyridazinones 46 and 47, it was necessary to protect
the starting anilines as their phthalimides prior to methylation
with dimethylformamide dimethyl acetal using a procedure
described by Sotelo et al.28 The general procedure described by
Carroll et al29 was employed to prepare intermediates 49−52
by sequential diazotization of anilines 30 and 45−47 followed
by treatment of the resulting diazonium salt with cyanoacetylur-
ethane 48. Cyclization of the crude product was effected by
treatment with sodium acetate in acetic acid at 120 °C to
provide the targeted cyanoazauracils. Removal of the nitrile to
3916
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
a
Scheme 7
a
Reagents and conditions: (a) (1) NaOAc, HOAc, H2O, 100 °C, 1.5 h, (2) NaOH/MeOH, 57%; (b) (1) phthalic anhydride, toluene, reflux, (2)
NaOAc, HOAc, H2O, 120 °C, 68%; (c) DMF dimethyl acetal, 100 °C, 52−73%; (d) butylamine, methanol, reflux, 1.5 h, 74%; (e) phthalic
anhydride, HOAc, 130 °C, 61%; (f) HOAc, 110 °C, 69%; (g) (1) NaNO2, HCl, 0 °C, (2) 48, pyridine, H2O, 0 °C, 76−95%; (h) NaOAc, HOAc, or
KOAc, DMA 120 °C, 1.5−3.5 h, 44−86%; (i) HCl, HOAc, 120 °C, 48 h, 78%; (j) mercaptoacetic acid, 170 °C, 1 h, 23−40%; (k)
dimethylacetamide, KOAc, 120 °C, 2 h, 86%.
acyclic acid substituted compounds, the methylene linked
azauracils 54 and 58 were roughly comparable in potency but
slightly less selective than the corresponding analogues with an
oxygen linker (53 and 57).
For T3 and other TH agonists, a key structural feature has
been the presence of phenol or an alternative hydrogen bond
donor to interact with histidine 435. Methylation of the
pyridazinone nitrogen provides derivatives with no possibility
of acting as hydrogen bond donors. Remarkably, the
methylated analogues 23, 55, and 56 are more potent, although
less selective, than the corresponding protonated analogues 20,
53, and 54. For the methylated analogues, and possibly for all
the pyridazinone compounds, histidine 435 may be serving as a
hydrogen bond donor to the pyridazinone carbonyl.
(4) that incorporates a bound azauracil in place of the
carboxylic acid moiety found in most ligands for this target. In
order to construct a model for the pyridazinones bound to
THR-β, we modified 4 to construct 53. This structure was then
submitted to limited optimization, providing a very con-
servative model for 53 bound to THR-β (Figure 1), shown with
T3 (from 3GWS)33 superimposed.
The cyano group in 53 was easily accommodated in the
1N46 active site with little induced movement of the active site
residues (Figure 1). Indeed the cyano group appears well
In addition to the IC50 values, the level of maximum activity
is reported in Table 1 as the percentage of the maximum
activity of each analogue relative to the maximum efficacy of
T3. It is interesting to note that while 2 and 3 achieved higher
activity than T3, all of the pyridazinone agonists had less than
100% activity. This is another indication of the differences
between the pyridazinone series and earlier analogues.
Early pyridazinone compounds 13, 14, and 21 substituted
with an acetic acid side chain were positive in an in vitro
micronucleus assay, as were 3 and T3 (1). Eliminating this
activity was one of the goals of the program. We were pleased
to find that the cyanoazauracils 53 and 54 were not only more
potent and selective than the acetic acid analogues but were
also negative in this assay.
Modeling. We modeled the binding of the cyanoazauracil
analogues with THR-β. Of the many published cocrystal
structures of THR-β, the structure with the most similar bound
ligand is PDB structure 1N46.13 This structure includes the
THR-β ligand binding domain together with the TH mimetic
Figure 1. Model of 53 (magenta) bound to THR-β (1N46) with the
T3 geometry (cyan) from 3GWS superimposed. Polar interactions of
53 in the anion binding site are highlighted.
3917
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
Figure 2. (A) 2D description of the binding site for T3 (PDB code 3GWS). (B) 2D description of the binding site for the 53 model (MOE).
Figure 3. Left panel: cardiac α-MHC hnRNA relative levels (arbitrary units) in untreated thyroidectomized rats (control), euthyroid rats, and
thyroidectomized rats 6 h after exposure to 53 dosed intraperitoneally at the specified doses.35 Right panel: activities of tested compounds relative to
full activity (euthyroid or T3-treated) and exposure of the compound 6 h after dose. ND is not determined.
positioned to pick up an additional interaction with Arg316
(Figures 1 and 2).
rats with T3 restores the level of gene transcription to
euthyroid-like levels.
Modeled 53 binds in a fashion that is very similar to T3. As
can be seen from the 2D projections (Figure 2), the T3
interactions with the positively charge ARG320 are mediated by
a series of bridging waters. In the model, as in the azauracil
analogue crystal structure (1N46), the cyanoazauracil of 53
interacts directly with ARG320. The large more diffuse
negatively charged heterocycle in 53 may convey advantages
over the conventional acids, since it displaces several waters and
would be expected to suffer a smaller binding desolvation
penalty relative to a carboxylic acid.34
Cardiac-Specific Effects of a Single Dose of 53 in
Adult Male Sprague−Dawley Rats. An important feature of
a selective TH mimetic is avoiding cardiac effects, which are
largely mediated by THR-α receptors. The purpose of this
study was to determine the effect of 53 on the expression of the
α-myosin heavy chain (α-MHC) gene in the hearts of rats that
are hypothyroid because of surgical removal of the thyroid. α-
MHC expression is up-regulated by the activation of the THR-
α. Since hypothyroid rats express very low levels of α-MHC
compared to euthyroidic rats, α-MHC expression in hypo-
thyroidic rats is a sensitive marker for TH regulated gene
expression in the heart. Treatment of these thyroidectomized
Expression of α-MHC hnRNA after treatment with low,
medium, and high doses of 53 ip is shown in Figure 3.
Compound 53 was given at 5, 20, and 37.5 mg/kg, ip, to yield
six hour postdose plasma concentrations of 15.4, 57, and 94
μM, respectively, and remarkably showed little to no induction
of α-MHC hnRNA. The data suggest that 53, compared to T3
or literature analogues such as 2, has little effect in the rat heart
as measured by changes in α-MHC levels (Figure 3, right
panel).35
The acetic acid analogue 21, dosed orally, had 35% of the T3
activity at reasonably low exposures. The methylene linked
cyanoazauracil analogue 54 (dosed ip) had no effect at the two
lowest concentrations tested but did demonstrate a small but
significant increase in α-MHC transcription to 17% of
euthyroid levels at the highest concentration tested (37.5
mg/kg).35 By contrast, the N-methylated analogue 55 (dosed
ip) induced full agonist activation of α-MHC transcription of
the cardiac-specific gene α-MHC at all doses tested. Even at the
lowest dose tested (5 mg/kg), transcription was measured at
euthyroid levels.
Several factors, including heart uptake and THR-α activity,
impact the performance of compounds in this assay.
3918
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
Table 2. Selected Property Data for 53
% free
(human)
human hepatocyte, CLint
hERG, IC20
(μM)
CYP inhibition,
IC50 (μM)
CYP TDI 3A4/5,
2C9, 2C19
solubility, pH
7.04 (μM)
Caco-2 A−B
Caco-2
efflux ratio
((μL/min)/106 cells)
(1 × 10−6 cm/s)
0.6
1.04
∼30
3A4/5: >50
2C19: >50
2C9: ∼22
none detected
1.0
1.37
18.4
Compound 53 was the only analogue with no activity at high
exposure and was selected for further evaluation.
cholesterol lowering, there was no effect on liver size. A
significant increase in heart and kidney size (data not shown)
accompanied by significant decreases in BMD of 10−13% was
observed at all doses of T3 (Figure 4).
Compound 53 was evaluated in a number of additional
animal models.36 In both hypercholesterolemic rat and rabbit
models, 53 significantly lowered non-HDL-C and liver TG at
doses where, when assessed, there was no effect on TSH levels
or the central thyroid axis. In the rabbit model, there were
indications of an additive lipid lowering effect when 53 was
coadministered with atorvastatin. Additionally, in a DIO mouse
study, 53 improved insulin sensitivity and lowered glucose to a
near normal level.
First in Human Studies. The safety profile and tolerability
of 53 were assessed in humans in a single ascending dose study
followed by a 2-week multiple dose study.20 In the multiple
ascending dose study, healthy subjects with mildly elevated
LDL-C (>110 mg/dL) received once daily oral doses of 53
ranging from 5 to 200 mg. Compound 53 was safe and well-
tolerated at all doses tested with no drug-related adverse events.
There was no evidence of central thyroid axis dysfunction at
any dose, no effect on heart rate, QT interval, or vital signs, and
no effect on liver enzymes.
At the highest doses of 53 reversible dose-related reductions
in free T4 of ∼20% (200 mg) and ∼10% (100 mg) were
observed. Free T3 and TSH showed no meaningful or dose-
dependent change compared with placebo, indicating that this
is not an effect on the central thyroid axis. This change, which is
not of a magnitude that is clinically meaningful, may be
attributed to increased hepatic THR-β agonism leading to
increases in the level of hepatic deiodinase 1, which reduces the
level of the prohormone T4 and has no effect on the
metabolism of the active T3.20 Similar observations were
made in preclinical studies.36
In Vitro and in Vivo Profiling of 53. Compound 53
showed low clearance and high stability in cryopreserved
human hepatocytes (Table 2). In Caco-2 cells, 53 displayed low
to medium-low permeability in the apical (A) to basolateral (B)
direction and high permeability in the B to A direction. In
addition, a liver to plasma ratio of 8:1 was observed for 53 in
diet induced obese (DIO) mice.36 Since the metabolic efficacy
of 53 is due to its effects in the liver, its concentration in the
liver coupled with the reasonably low permeability should
contribute to the safety of the compound by reducing systemic
exposure. This is especially important for limiting uptake to the
pituitary, where THR-β is responsible for central thyroid axis
regulation.
Compound 53 showed an IC20 of roughly 30 μM for
blockage of the hERG channel (Table 2). The IC50 for
CYP3A4/5 and for CYP2C19 was >50 μM, the highest
concentration tested, and there was only weak inhibition
(roughly 22 μM) of CYP2C9. In the presence or absence of
NADPH, no time dependent inhibition (TDI) of any of these
CYP enzymes was observed.
When dosed as a suspension, 53 exhibited good exposures
and reasonable oral bioavailability in rats (Table 3). The
volume of distribution and clearance were both low. Dose
proportional increases in exposure were observed for a
suspension of 53 given orally to DIO mice.36
Table 3. Pharmacokinetic Parameters of 53 after iv and po
Administration to Rats (Mean Values, N = 3, SD in
Parentheses)
rat
route
iv
po
Drug exposure in the plasma increased with increasing dose.
The increase in exposure was fairly linear through 80 mg, with
higher variability at the higher doses.
dose (mg/kg)
AUC (μg·h/mL)
Tmax (h)
5
5
38.1(11.4)
17.4(8.6)
6(0)
While the primary purpose of these studies was assessment of
safety, effects on lipids were observed in the multiple ascending
dose study. Statistically significant reductions relative to placebo
of up to 30% for LDL-C and statistical trends of up to 60%
reduction in TG were observed for doses ranging from 50 to
200 mg. The near maximal lipid effects were observed at a dose
of 80 mg daily.
Cmax (μg/h)
CL (mL min−1 kg−1
Vss (L/kg)
T1/2 (h)
1.71(0.85)
)
2.35(0.85)
0.422(0.057)
3.4(0.26)
4.08(1.16)
45(22.5)
F (%)
In a 23-day study of 53 in C57Bl/6J-diet-induced obese
(DIO) mice, animals were treated with daily oral doses of either
53 vehicle control or 0.3, 1.0, 3.0, or 10 mg/kg 53. In a parallel
24-day study using identical age DIO mice and identical
protocol design, mice were treated with either T3-vehicle
control or 10, 30, or 100 μg/kg T3. At the end of the study,
laboratory assessments and body composition measurements
were made. In animals treated with 53 there was a reduction in
cholesterol and in liver size, which is secondary to reduction of
liver TG.36 There was no effect on bone mineral density
(BMD) or heart or kidney size (data not shown) in 53 treated
animals. By contrast, while T3 treated animals demonstrated
CONCLUSION
■
Our efforts to identify selective THR-β agonists that did not
contain the highly electron-rich phenolic moiety characteristic
of other TH mimetics led to the identification of a pyridazinone
series. The initial compounds were significantly less potent and
more selective than previously reported compounds 2 and 3.
Replacement of the acetic acid substituent with a cyanoazauracil
improved both the potency and selectivity of the compounds,
leading to 53 which is 28-fold selective for THR-β over THR-α
in an in vitro functional coactivator recruitment assay.
3919
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
Figure 4. Effects of 53 vs T3 on cholesterol and BMD in DIO mice.
spectrometer with electrospray ionization (ESI); Chromegabond
WR C18 3 μm, 120 Å, 3.2 mm × 30 mm column; solvent A,
H2O−0.02% TFA; solvent B, MeCN−0.02% TFA; flow rate, 2 mL/
min; start 2% B, final 98% B in 4 min, linear gradient. Method 2
involved the following: Waters 2795 pump and Waters 2996
photodiode array detector set at 214 nm with the MS detection
performed with a Waters ZQ mass spectrometer (ESI); Epic Polar
hydrophilic 3 μm, 120 Å, 3.2 mm × 30 mm column; solvent A, H2O−
0.03% HCO2H; solvent B, MeCN−0.03% HCO2H; flow rate, 2 mL/
min; start 10% B, final 100% B in 3 min linear gradient, remaining for
1 min. All final compounds reported were analyzed using one of these
analytical HPLC methods and were at least 95% pure.
Compared with all other compounds tested, 53 showed no
THR-α mediated gene expression in a rat heart model. In vitro
assays are predictive of low potential for off target actions.
Despite the low to medium permeability and potential for efflux
observed in the Caco-2 assay, 53 has demonstrated reasonable
plasma exposures in preclinical animal models and in humans.
The permeability and potential for efflux coupled with selective
uptake into the liver may contribute to the safety of 53 by
minimizing thyroid actions outside the liver. When dosed orally
to DIO mice, 53 reduced cholesterol levels and liver size
without the side effects observed with T3 of decreased BMD
and increased heart and kidney size. In healthy volunteers, 53,
at once daily oral doses of 50 mg or higher for 2 weeks,
exhibited an excellent safety profile, significantly decreased
LDL-C by up to 30%, and showed a statistical trend to reduce
TG up to 60%.20 Future investigations, including clinical trials,
will assess the safety and test the ability of 53 to treat patients,
including diabetics, with mixed dyslipidemias and/or NAFLD
and NASH both as a single agent and in combination with
statins.
Flash column chromatography was performed on EM Science silica
gel 60 or with prepacked Biotage or Isco silica gel columns. Preparative
reverse-phase high-pressure liquid chromatography (RP-HPLC) used
a Waters Delta prep 4000 pump/controller, a 486 detector set at 215
nm, and a LKB Ultrorac fraction collector on a C-18 column using
1
acetonitrile/water with 0.1% TFA. H NMR spectra were recorded
using a Varian Mercury 300 MHz or Varian Inova 400 MHz
spectrometer. The chemical shifts are in parts per million (δ)
referenced to Me4Si (0.00 ppm) or CHCl3 (7.26 ppm). High-
resolution mass spectra were recorded on a Bruker Apex II FTICR
mass spectrometers with a 4.7 T magnet (ES) or Micromass AutoSpec
(EI) mass spectrometers.
6-(4-Amino-2,6-dichlorophenoxy)-4-isopropyl-2H-pyrida-
zin-3-one 45. A mixture of glacial acetic acid (30 mL), sodium
acetate (860 mg, 10.5 mmol), and 25 (1.0 g, 3.0 mmol) was heated to
100 °C for 24 h. The reaction mixture was cooled to room
temperature, stirred for 2 days, and concentrated. The residue was
diluted with water (200 mL) and was made basic to pH 9 by the
addition of 1 N sodium hydroxide solution. The suspension formed
was extracted with ethyl acetate (1 × 250 mL). The water layer was
acidified to pH 5 by the addition of concentrated hydrochloric acid.
The water layer was extracted with ethyl acetate (1 × 250 mL). The
organic layers were combined, dried with magnesium sulfate, filtered,
and concentrated under vacuum. The resulting oil was diluted with
methanol (20 mL) and was treated with a 1 N aqueous sodium
hydroxide solution (20 mL, 20 mmol). The reaction mixture was
EXPERIMENTAL SECTION
■
All nonaqueous reactions were carried out under an argon or nitrogen
atmosphere at room temperature unless otherwise noted. All reagents
and anhydrous solvents were used as obtained commercially without
further purification or distillation unless otherwise stated. Analytical
thin-layer chromatography (TLC) was performed on EMD Chemicals
silica gel 60 F254 precoated plates (0.25 mm). Compounds were
visualized by UV light and/or stained with either p-anisaldehyde,
iodine, or phosphomolybdic acid or KMnO4 solutions followed by
heating. Analytical high-pressure liquid chromatography (HPLC) and
LC−MS analyses were conducted using the following two instruments
and conditions. Method 1 involved the following: Hewlett-Packard
HP-1090 pump and HP-1090 PDA detector set at 215 nm with the
MS detection performed with a Micromass Platform II mass
3920
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
heated to 120 °C for 24 h. The reaction mixture was cooled to room
temperature, and the solvent was concentrated under vacuum. The
residue was diluted with water (100 mL) and was extracted with ethyl
acetate (200 mL). The ethyl acetate layer was washed with water
containing a 1 N aqueous hydrochloric acid solution (to pH 5) and
saturated sodium chloride solution, dried with magnesium sulfate,
filtered, and concentrated under vacuum. The residue was dissolved in
chloroform and purified by flash chromatography (Biotage 40L) using
silica gel, eluting with a 1:1 ethyl acetate/hexanes solution with 0.5%
glacial acetic acid. The desired fractions were collected and
concentrated under vacuum and dried under high vacuum at 37 °C.
The solid was slurried in diethyl ether (∼10 mL) and petroleum ether
(10 mL). The solid was stirred for 20 min at room temperature,
filtered, and rinsed well with petroleum ether. The solid was dried
under high vacuum to afford 45 (538 mg, 57%) as an off-white solid.
LRMS for C13H13Cl2N3O2 (M + H)+ m/z = 314.
Ethyl (2-Cyano-2-(2-(3,5-dichloro-4-((5-isopropyl-6-oxo-1,6-
dihydropyridazin-3-yl)oxy)phenyl)hydrazono)acetyl)-
carbamate 49. A suspension of 45 (134 mg, 0.42 mmol) in water
(5.6 mL) was treated with concentrated hydrochloric acid (2.8 mL).
The reaction mixture was cooled to 0 °C and was treated with a
solution of sodium nitrate (36.5 mg, 0.529 mmol) in water (0.2 mL)
under the surface of the reaction mixture followed by a water (0.2 mL)
rinse. The reaction mixture was stirred at 0 °C for 30 min, and a
solution formed. In a separate flask, equipped with a magnetic stirrer,
were added 48 (73 mg, 0.46 mmol), water (9.4 mL), and pyridine (2.8
mL). This mixture was cooled to 0 °C, and the solution from the first
reaction was quickly filtered and poured into the second reaction
mixture. An orange precipitate formed, and the suspension was stirred
at 0 °C for 30 min. The solid was filtered and rinsed with water
followed by petroleum ether. Drying in a vacuum oven overnight at 80
°C afforded 49 (156 mg, 76%) as an orange solid. EI(+)-HRMS m/z
calcd for C19H18Cl2N6O5 (M + H)+ 481.0789, found 481.0790.
2-[3,5-Dichloro-4-(5-isopropyl-6-oxo-1,6-dihydropyridazin-
3-yloxy)phenyl]-3,5-dioxo-2,3,4,5-tetrahydro[1,2,4]triazine-6-
carbonitrile 53. A mixture of 49 (3.49 g, 7.17 mmol) in glacial acetic
acid (72 mL) was treated with sodium acetate (2.94 g, 35.8 mmol) at
room temperature. The reaction mixture was heated to 120 °C for 1.5
h. At this time, the reaction mixture was cooled to 0 °C, diluted with
water (220 mL), and stirred for 30 min. The resulting solid was filtered
and rinsed with water (3 × 100 mL) followed by petroleum ether (3 ×
100 mL). The solid was air-dried for 30 min. The solid was dissolved
in hot acetonitrile (250 mL). The resulting red mixture was treated
with neutral decolorizing carbon, filtered through Celite, and rinsed
with acetonitrile (1 L) until no further UV active material eluted. The
yellow filtrate was concentrated under reduced pressure. The resulting
solid was triturated with hot acetonitrile (50 mL), cooled for 15 min,
diluted with water (100 mL), and filtered. The solid was triturated
again with hot acetonitrile (10 mL), filtered, and rinsed with
acetonitrile, water, and petroleum ether. The solids were collected
and dried under high vacuum overnight and then dried in a vacuum
oven at 80 °C overnight to afford 53 (1.91 g, 61%) as a yellow solid.
1H NMR (400 MHz, DMSO-d6) δ 13.28 (br s, 1H), 12.24 (s, 1H),
4 °C. The suspension was centrifuged in a Ti45 Beckmann rotor for 2
h at 127 300 RCF, and the supernatant was loaded onto NI_ NTA
agarose (Quigen 30210) column. After a washing with buffer A, H6-
TRβ or H6-TRα was eluted with buffer A containing 0.25 M
imidazole.
The ligand binding domain of human retinoid X receptor (amino
acids 225−462) (RxRα) was engineered with N-terminal His6 and EE
(EFMPME) tags, a thrombin cleavage site between the His6 and EE
sequences, and cloned into pACYC vector. The resulting His6-EE-
tagged protein was produced in E. coli cells. Cells were grown using
shake flasks with an 18 h induction in 0.1 mM IPTG at 18 °C,
harvested, and suspended with five volumes of buffer B (0.025 M Tris,
0.3 M NaCl, 0.02 M imidazole, 0.01 M β-mercaptoethanol, pH 8.0).
Lysozyme (0.2 mg/mL, Sigma) and Complete protease inhibitor
cocktail (Roche Diagnostics Gmbh) were added and stirred for 30 min
at 4 °C. The suspension was sonicated for 30 s, five times, at 4 °C. The
suspension was centrifuged for 20 min at 12 000 RCF. The
supernatant was filtered by 0.45 μm pore size membrane, and 0.5%
NP-40 was added. The His6-tagged protein was bound to and eluted
from NiNTA metal-affinity resin (QIAGEN, Valencia, CA). The
protein was concentrated and dialyzed.
The His6 tag was removed from EE-RxRα by thrombin digestion,
using 10 units of thrombin (Pharmacia, Piscataway, NJ) per milligram
of protein and incubating for 2 h at 25 °C. Removal of thrombin was
done batchwise using benzamidine-Sepharose 6B (Pharmacia, Piscat-
away, NJ). The protein was concentrated and dialyzed. This protein
was used in the coactivator peptide recruitment assay.
Europium-conjugated anti hexa His antibody and APC-conjugated
streptavidin were purchased from PerkinElmer Life and Analytical
Sciences.
THR-β/RXR/GRIP1 Coactivator Peptide Recruitment Assay.
An amount of 30 μL of H6-THR-β (50 nM) in 50 mM Hepes, pH 7.0,
1 mM DTT, 0.05% NP40, and 0.2 mg/mL BSA (binding buffer) was
mixed with an equal volume of EE-RxRα (50 nM) in binding buffer.
An amount of 6 μL of T3 (0−14.8 μM) or test compound (0−1.2
mM) in DMSO was then added and the solution incubated at 37 °C
for 30 min. Then 30 μL of biotin-GRIP1 peptide (biotin-Aca-
HGTSLKEKHKILHRLLQDSSSPVDL-CONH2) (100 nM) in 30 μL
of binding buffer plus 5% DMSO was added and the solution
incubated at 37 °C for 30 min. An amount of 30 μL of solution
containing 12 nM europium-conjugated anti hexa His antibody and
160 nM APC-conjugated streptavidin in 50 mM Tris, pH 7.4, 100 mM
NaCl, and 0.2 mg/mL BSA was added, and the solution was incubated
at 4 °C overnight. An aliquot (35 μL/sample) was transferred to 384-
well black microtiter plates. The HTRF signal was read on the Victor 5
reader (PerkinElmer Life and Analytical Sciences).
THR-α/RXR/GRIP1 Coactivator Peptide Recruitment Assay.
The assay protocol is essentially the same as that of THR-β/RXR/
GRIP1 coactivator peptide recruitment assay as described above
except that 125 nM H6-THR-α, 125 nM EE-RxRα, and 250 nM
biotin-GRIP1 were used.
Computational Procedure. Analyses of the crystal structures and
computer graphics were done using PyMol.37 The structure for 53 was
constructed by manually modifying the crystal structure of the
azauracil derivative 4. The resulting structure was then submitted to
limited optimization using the MMFF force field38 with the protein
structure frozen, followed by additional optimization using the LigX
capability in MOE.39 The ligand was given a formal charge of −1, and
the protonation states for the protein were assigned using MOE. The
protein atoms were restrained during optimization with all atoms
falling beyond 6 Å being frozen. This resulted in a model for 53 bound
to the receptor where the protein structure is little changed from PDB
code 1N46.13
7.78 (s, 2H), 7.45 (s, 1H), 2.94−3.17 (m, 1H), 1.20 (d, J = 6.82 Hz,
6H). EI(+)-HRMS m/z calcd for C17H12Cl2N6O4 (M + H)+ 435.0370,
found 435.0368.
THR/RXR/GRIP1 Assay. The ligand binding domain (amino acids
148−410) of THR-β (H6-THR-β) and the ligand binding domain
(amino acids 202−461) of THR-α (H6-THR-α) were cloned into an
E. coli expression vector pET28a (Novagen, Milwaukee, WI) that
contained a N-terminal hexa His sequence. The resulting recombinant
hexa His tagged proteins were produced in E. coli BL21(DE3) cells.
Cells were grown in Terrific Broth (in-house prepared medium of
Bacto tryptone (3.3%, w/v), Difico yeast extract (2.0%, w/v), and
NaCl (0.5%, w/v)) using shake flasks with a 24 h induction in 0.2 mM
IPTG at 25 °C, harvested, and lysed with five volumes of buffer A
(0.05 M Tris, 0.3 M NaCL, 1%W/V betaine, 0.01 M imidazole, 0.02 M
β-mercaptoethanol, pH 8.0). Lysozyme (1.0 mg/mL, Sigma) and
Complete protease inhibitor cocktail (Roche Diagnostics Gmbh) were
added to slurry, and the solution was sonicated for 1 min five times at
Thyroidectomized Rat Cardiac Myocyte Assay. These assays
were conducted under the direction of Irwin Klein using the previously
described procedures.35 Compounds 53, 54, and 55 were formulated
in 4% DMSO, 15% PEG-400, and 81% of 30% HPBCD in phosphate
buffer and were administered intraperitoneally. For 53 and 54, 4 rats
per group were tested at 5, 20, and 37.5 mg/kg. For 55, 3 rats per
group were tested at 5 and 15 mg/kg and 4 rats were tested at 50 mg/
3921
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
kg. Compound 21 was formulated in 2% Klucel LF, 0.1% Tween 80,
water and dosed orally (n = 3 per group) at 0.0625, 0.25, and 1.25 mg/
kg.
(5) Grover, G. J.; Mellstrom, K.; Ye, L.; Malm, J.; Li, Y.-L.; Bladh, L.-
̈
G.; Sleph, P. G.; Smith, M. A.; George, R.; Vennstrom, B.; Mookhtiar,
̈
K.; Horvath, R.; Speelman, J.; Egan, D.; Bater, J. D. Selective thyroid
hormone receptor-beta activation: a strategy for reduction of weight,
cholesterol, and lipoprotein (a) with reduced cardiovascular liability.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 10067−10072.
C57Bl/6J-Diet-Induced Obese (DIO) Mice Study. All animal
studies were done according to IACUC approved protocols. Six week
old C57Bl/6J mice were placed on a high fat diet for 34 weeks. At day
0, 9 mice per group were treated daily doses by gavage with vehicle
(2% Klucel LF, 0.1% Tween 80 in water) or 0.3, 1, 3, or 10 mg/kg 53
for 23 days. In a parallel study, at day 0, 9 mice per group were treated
with daily doses of vehicle (Dulbecco’s phosphate buffered saline, pH
adjusted to 9.0 with 1 N NaOH) or 10, 30, or 100 μg/kg T3. Body
weight and food intake were monitored during the study. BMD and
body composition assessments were made on day 22. On day 23 at
necropsy, organ weights were obtained for determination of organ
weight and blood samples were assessed for cholesterol and other
chemistry parameters.
(6) Grover, G. J.; Egan, D. M.; Sleph, P. G.; Seehler, B. C.; Chiellini,
G.; Nguyen, N.-H.; Baxter, J. D.; Scanlan, T. S. Effects of the thyroid
hormone receptor agonist GC-1 on metabolic rate and cholesterol in
rats and primates: selective actions relative to 3,5,3′-triiodo-L-
thyronine. Endocrinology 2004, 145, 1656−1661.
(7) Johansson, L.; Rudling, M.; Scanlan, T. S.; Lundasen, T.; Webb,
̊
P.; Baxter, J.; Angelin, B.; Parini, P. Selective thyroid receptor
modulation by GC-1 reduces serum lipids and stimulates steps of
reverse cholesterol transport in euthyroid mice. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102, 10297−10302.
(8) Morkin, E.; Pennock, G. D.; Spooner, P. H.; Bahl, J. J.; Goldman,
S. Clinical and experimental studies on the use of 3,5-diiodothyr-
opropionic acid, a thyroid hormone analogue, in heart failure. Thyroid
2002, 12, 527−533.
(9) Bassett, J. H; Williams, G. R. The molecular actions of thyroid
hormone in bone. Trends Endocrinol. Metab. 2003, 14, 356−364.
(10) Simonides, W. S.; Thelen, M. H. M.; van der Linden, C. G.;
Muller, A.; van Hardeveld, C. Mechanism of thyroid-hormone
regulated expression of the SERCA genes in skeletal muscle:
implications for thermogenesis. Biosci. Rep. 2001, 21, 139−154.
(11) Klein, I.; Danzi, S. Thyroid disease and the heart. Circulation
2007, 116, 1725−1735.
ASSOCIATED CONTENT
* Supporting Information
Synthetic procedures and analytical data for all compounds
except for 53; procedures for in vitro tests. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*Phone: 267-327-4448. E-mail: martha@madrigalpharma.com.
Notes
■
(12) Joharapurkar, A. A.; Dhote, V. V.; Jain, M. R. Selective
thyromimetics using receptor and tissue selectivity approaches:
prospects for dyslipidemia. J. Med. Chem. 2012, 55, 5649−5675.
(13) Dow, R. L.; Schneider, S. R.; Paight, E. S.; Hank, R. F.; Chiang,
P.; Cornelius, P.; Lee, E.; Newsome, W. P.; Swick, A. G.; Spitzer, J.;
Hargrove, D. M.; Patterson, T. A.; Pandit, J.; Chrunyk, B. A.; LeMotte,
P. K.; Danley, D. E.; Rosner, M. H.; Ammirate, M. J.; Simons, S. P.;
Schulte, G. K.; Tate, B. F.; DaSilve-Jardine, P. Discovery of a novel
series of 6-azauracil-based thyroid hormone receptor ligands: potent,
TRβ subtype-selective thyromimetics. Bioorg. Med. Chem. Lett. 2003,
13, 379−382.
The authors declare the following competing financial
interest(s): M. Kelly and R. Taub are employed by Madrigal
Pharmaceuticals.
ACKNOWLEDGMENTS
■
We thank Joseph Grimsby and Dennis Kertesz for helpful
discussions, Lianhe Shu and Ping Wang for assistance in scale
up, and Anjula Pamidimukkala for providing PK data.
ABBREVIATIONS USED
■
(14) Berkenstam, A.; Kristensen, J.; Mellstrom, K.; Carlsson, B.;
̈
APC, allophycocyanin; BMD, bone mineral density; DIO, diet
induced obese; EE-RxRα, ligand binding domain of retinoid X
receptor with EE tag; H6-TR-α, ligand binding domain of
thyroid hormone receptor α with hexa His tag; H6-THR-β,
ligand binding domain of thyroid hormone receptor β with
hexa His tag; LXR, liver X receptor; α-MHC, α-myosin heavy
chain; NAFLD, nonalcoholic fatty liver disease; NASH,
nonalcoholic steatohepatitis; T3, triiodothyronine; T4, thyro-
xine; TDI, time dependent inhibition; TEMPO, 2,2,6,6-
tetramethylpiperidin-1-yl)oxy; TG, triglyceride; TH, thyroid
hormone; THR, thyroid hormone receptor; THR-α, thyroid
hormone receptor α; THR-β, thyroid hormone receptor β;
TSH, thyroid stimulating hormone
Malm, J.; Rehnmark, S.; Garg, N.; Andersson, C. M.; Rudling, M.;
Sjoberg, F.; Angelin, B.; Baxter, J. D. The thyroid hormone mimetic
̈
compound KB2115 lowers plasma LDL cholesterol and stimulates bile
acid synthesis without cardiac effects in humans. Proc. Natl. Acad. Sci.
U.S.A. 2008, 105, 663−667.
(15) Ladenson, P. W.; Kristensen, J. D.; Ridgway, E. C.; Olsson, A.
G.; Carlsson, B.; Klein, I.; Baxter, J. D.; Angelin, B. Use of the thyroid
hormone analogue eprotirome in statin-treated dyslipidemia. N. Engl. J.
Med. 2010, 362, 906−916.
(16) Erion, M. D.; Cable, E. E.; Ito, B. R.; Jiang, H.; Fujitaki, J. M.;
Finn, P. D.; Zhang, B.-H.; Hou, J.; Boyer, S. H.; van Poelje, P. D.;
Linemeyer, D. L. Targeting thyroid hormone receptor-β agonists to
the liver reduces cholesterol and triglycerides and improves the
therapeutic index. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 15490−
15495.
REFERENCES
(17) Garg, N.; Rao, V. M.; Gandhi, R. B.; Washburn, W. N.; Koehler,
K.; Malm, J. Stable oral pharmaceutical composition containing
thyroid hormone receptor agonists. US 20090220598, 2009.
(18) Woltering, M.; Haning, H.; Schmidt, G.; Pernerstorfer, J.;
■
(1) Waters, D. D. Lipid treatment assessment project 2: a
multinational survey to evaluate the proportion of patients achieving
low-density lipoprotein cholesterol goals. Circulation 2009, 120, 28−
34.
(2) Joy, T. R.; Hegele, R. A. Narrative review: statin-related
myopathy. Ann. Intern. Med. 2009, 150, 858−868.
(3) DeFronzo, R. A.; Ferrannini, E. Insulin resistance. A multifaceted
syndrome responsible for NIDDM, obesity, hypertension, dyslipide-
mia, and atherosclerotic cardiovascular disease. Diabetes Care 1991, 14,
173−194.
(4) Cazanave, S. C; Gores, G. J. Mechanisms and clinical implications
of hepatocyte lipoapoptosis. Clin. Lipidol. 2010, 5, 71−85.
Bischoff, H.; Kretschmer, A.; Vohringer, V.; Faeste, C. Indazoles
̈
having an action similar to that of a thyroid hormone. Method for the
production thereof and their use in medicaments. WO200222586,
2002.
(19) Haning, H.; Schmidt, G.; Pernerstorfer, J.; Schmeck, C.; Muller,
̈
U.; Bischoff, H.; Vohringer, V.; Reinemer, P.; Apeler, H.; Schmidt, D.;
̈
Jonghaus, W.; Faeste, C.; Zoche, M.; Hauswald, M.; Woltering, M.;
Kretschmer, A. Indoles for treating diseases that can be treated using
thyroid hormones. WO200170687, 2001.
3922
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923

Journal of Medicinal Chemistry
Drug Annotation
(20) Taub, R.; Chiang, E.; Chabot-Blanchet, M.; Kelly, M. J.; Reeves,
R. A.; Guertin, M.-C.; Tardif, J.-C. Lipid lowering in healthy volunteers
treated with multiple doses of MGL-3196, a liver-targeted thyroid
hormone receptor-β agonist. Atherosclerosis 2013, 230, 373−380.
(21) Samaritoni, J. G. Homolytic alkylation of 3,6-dichloropyridazine.
Org. Prep. Proced. Int. 1988, 20 (1−2), 117−121.
(38) Halgren, T. A. Merck molecular force field. I. Basis, form, scope,
parameterization, and performance of MMFF94. J. Comput. Chem.
1996, 17, 490−519.
(39) Molecular Operating Environment (MOE), version 2012.10;
Chemical Computing Group Inc. (1010 Sherbooke St. West, Suite
910, Montreal, QC, Canada, H3A 2R7), 2012.
(22) Gardner, P. D.; Rafsanjani, H. S.; Rand, L. Reaction of phenolic
Mannich base methiodides and oxides with various nucleophiles. J. Am.
Chem. Soc. 1959, 81, 3364−3367.
(23) Chen, Y. L.; Ruggeri, S. The process for converting 2,4-
dichloropyridines into 2-aryloxy-4-chloropyridines. WO 199639388,
1996.
(24) Hickey, D. M. B.; Leeson, P. D.; Novelli, R.; Shah, V. P.; Burpitt,
B. E.; Crawford, L. P.; Davies, B. J.; Mitchell, M. B.; Pancholi, K. D.;
Tuddenham, D.; Lewis, N. J.; O’Farrell, C. Synthesis of thyroid
hormone analogues. Part 3. Iodonium salt approaches to SK&F L-
94901. J. Chem. Soc., Perkin Trans. 1 1988, 3103−3111.
(25) Anelli, P. L.; Biffi, C.; Montanari, F.; Quici, S. Fast and selective
oxidation of primary alcohols to aldehydes or to carboxylic acids and of
secondary alcohols to ketones mediated by oxoammonium salts under
two-phase conditions. J. Org. Chem. 1987, 52, 2559−2562.
(26) Salturo, F.; Bemis, G.; Gao, H. Inhibitors of p38. WO
200017204, 2000.
(27) Law, H.; Dukat, M.; Teitler, M.; Lee, D. K. H.; Mazzocco, L.;
Kamboj, R.; Rampersad, V.; Prisinzano, T.; Glennon, R. A.
Benzylimidazolines as h5-HT1B/1D serotonin receptor ligands: a
structure−affinity investigation. J. Med. Chem. 1998, 41, 2243−2251.
(28) Sotelo, E.; Ravina, E. Pyridazine derivatives. XXIV. Efficient N-
methylation of diversely substituted 3(2H)-pyridazinones using N,N-
dimethylformamide dimethyl acetal. Synth. Commun. 2002, 32 (11),
1675−1680.
(29) Carroll, R. D.; Miller, M. W.; Mylari, B. L.; Chappel, L. R.;
Howes, H. L., Jr.; Lynch, M. J.; Gupta, S. K.; Rash, J. J.; Koch, R. C.
Anticoccidial derivatives of 6-azauracil. 5. Potentiation by benzophe-
none side chains. J. Med. Chem. 1983, 26, 96−100.
(30) Yoshihara, H. A. I.; Apriletti, J. W.; Baxter, J. D.; Scanlan, T. S.
Structural determinants of selective thyromimetics. J. Med. Chem.
2003, 46, 3152−3161.
(31) Ye, L.; Li, Y.-L.; Mellstrom, K.; Mellin, C.; Bladh, L.-G.; Koehler,
̈
K.; Garg, N.; Collazo, A. M. G.; Litten, C.; Husman, B.; Persson, K.;
Ljunggren, J.; Grover, G.; Sleph, P. G.; George, R.; Malm, J. Thyroid
receptor ligands. 1. Agonist ligands selective for the thyroid receptor
β1. J. Med. Chem. 2003, 46, 1580−1588.
(32) Boyer, S. H.; Jiang, H.; Jacintho, J. D.; Reddy, M. V.; Li, H.; Li,
W.; Godwin, J. L.; Schulz, W. G.; Cable, E. E.; Hou, J.; Wu, R.;
Fujitaki, J. M.; Hecker, S. J.; Erion, M. D. Synthesis and biological
evaluation of a series of liver-selective phosphonic acid thyroid
hormone receptor agonists and their prodrugs. J. Med. Chem. 2008, 51,
7075−7093.
(33) Nascimento, A. S.; Dias, S. M. G.; Nunes, F. M.; Aparicio, R.;
Ambrosio, A. L. B.; Bleicher, L.; Figueira, A. C. M.; Santos, M. A. M.;
de Oliveira Neto, M.; Fischer, H.; Togashi, M.; Craievich, A. F.;
Garratt, R. C.; Baxter, J. D.; Webb, P.; Polikarpov, I. Structural
rearrangements in the thyroid hormone receptor hinge domain and
their putative role in the receptor function. J. Mol. Biol. 2006, 360,
586−598.
(34) (a) Reynolds, C. H. Computer-aided drug design: a practical
guide to protein-structure based modeling. In Drug Design: Ligand and
Structure Based Approaches; Merz, K. M., Ringe, D., Reynolds, C. H.,
Eds.; Cambridge University Press: Cambridge, U.K., 2010; pp 181−
196. (b) Kangas, E.; Tidor, B. Optimizing electrostatic affinity in
ligand−receptor binding: theory, computation, and ligand properties. J.
Chem. Phys. 1998, 109, 7522−7545.
(35) Danzi, S.; Klein, I. Cardiac specific effects of thyroid hormone
analogues. Horm. Metab. Res. 2011, 43, 737−742.
(36) Taub, R.; Grinsby, J.; Larigan, J. D.; Pietranico, S.; Dvorozniak,
M.; Conde-Knape, K. Unpublished results.
(37) MacPymol; DeLano Scientific LLC: South San Francisco, CA,
2006.
3923
dx.doi.org/10.1021/jm4019299 | J. Med. Chem. 2014, 57, 3912−3923