Thyroid receptor ligands. Part 7: Indirect antagonists of the thyroid hormone receptor with improved affinity

Article abstract of DOI:10.1016/j.bmcl.2007.01.009

Based on the concept of 'indirect antagonism' of nuclear receptors, a series of thyroid hormone receptor (TR) antagonists were prepared with improved affinity compared with what was previously described. The results of a binding assay for the human TR and

Full text of DOI:10.1016/j.bmcl.2007.01.009

Bioorganic & Medicinal Chemistry Letters 17 (2007) 2018–2021
Thyroid receptor ligands. Part 7: Indirect antagonists of the
thyroid hormone receptor with improved affinity
Johan Malm,* Sandra Gordon, Peter Brandt, Bo Carlsson, Peter Agback,
Anna Ba¨ckbro Saeidi and Johnny Sandberg
Karo Bio AB, Novum, Huddinge S-141 57, Sweden
Received 6 December 2006; revised 4 January 2007; accepted 7 January 2007
Available online 13 January 2007
Abstract—Based on the concept of ‘indirect antagonism’ of nuclear receptors, a series of thyroid hormone receptor (TR) antagonists
were prepared with improved affinity compared with what was previously described. The results of a binding assay for the human
TR and reporter cell assay revealed, within this series, that an m-bromobenzoyl substituent (11f) was optimal in terms of affinity and
antagonist activity. Compared with already reported TR antagonists, their affinities are within the same range, thus potentially rep-
resenting useful approach to novel and high-affinity TR-antagonists.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The thyroid hormone receptor (TR) belongs to the ‘su-
per-family’ of nuclear hormone receptors (NRs), which
are intracellular ligand activated transcription factors.1
Thyroid hormones (THs) exert profound and central
effects on development and homeostasis of the body
and regulate a variety of important genes in the muscle,
liver and central nervous system. Endogenous levels of
TH control overall metabolic rate, lipid levels, heart rate
and mood.2 Endogenous thyroid hormones are current-
ly used primarily as replacement therapy for patients
with hypothyroidism. Treatment with 3,5,30-tri-iodo-L-
thyronine (L-T3, 1) and 3,5,30,50-tetra-iodo-L-thyronine
(L-T4, 2) (Fig. 1) usually returns metabolic functions
to the euthyroid state.
Treatment of hyperthyroidism is currently directed to-
wards reduction of the overproduction of THs by inhib-
iting their synthesis or release, or by ablating thyroid
tissue with surgery or radioiodine.5 However because
of the long half-life of thyroid hormones, these therapies
require several weeks before the euthyroid state is re-
stored. A TR-antagonist would quickly restore the
euthyroid state and consequently rapidly alleviate the
clinical manifestations mentioned above. Therefore, a
TR-antagonist could be used as a short-term supple-
mental therapy or even as a monotherapy if careful titra-
tion was applied. Ideally, such a ligand should be
non-selective for either TRa or TRb.
Rational ligand design of TR-antagonists has largely
While the potential use of TR-agonists is well-estab-
lished,3 the clinical applications of TR-antagonists
might be less obvious. A TR-antagonist may, however,
have significant clinical use for treating the thyroid hor-
mone excess states thyrotoxicosis and hyperthyroidism,
which arise when tissues are exposed to elevated levels
of circulating 1 and 2.4 Clinically, this state can manifest
itself in weight loss, hypermetabolism, lowering of lipid
levels, cardiac arrhythmias, heart failure, muscle weak-
ness, bone loss in postmenopausal women and anxiety.
been based on the attachment of a large extension group
0
at the R5 -position of 1 or other close analogues. This
extension is believed to displace the C-terminal helix (he-
lix 12) from a conformation that promotes co-activator
binding to a second conformation which promotes
co-repressor binding (direct antagonism). A number of
examples of synthetic TR-antagonists using this strategy
are known in the literature and is exemplified in Figure 2
Keywords: Thyroid hormone receptor; Thyroid hormone antagonist;
Indirect antagonism; Benzyloxyphenyl derivatives; Thyroid hormone
agonist; Thyrotoxicosis; Hyperthyroidism.
Figure 1. Chemical structures of L-T3 (1) and L-T4 (2), including ring-
numbering of 1.
*
Corresponding author. E-mail: johanm@excite.com
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.01.009

J. Malm et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2018–2021
2019
more, in order to maximize affinity, and in accordance
with what has been reported previously, we decided to
keep the R1 side-chain substituted with propionic acid
and the R3, R5-positions with bromine.
Thus, a series of ligands where the outer ring of a thyr-
omimetic was replaced with a benzyl group containing
various functional group substitutions was prepared as
outlined in Scheme 1. Commercially available benzyl
bromides 9 and 12 were coupled with phenol 10 in the
presence of base, to give the appropriate benzyloxyphe-
nyl derivatives; which were subsequently deprotected
through saponification to give the end products 8,
11a–f and 13. Further manipulation of 13 gave the
monoethyl and diethyl-substituted anilines 14 and 15,
respectively.14
Figure 2. Chemical structures of direct antagonists 3–6.
by the extensively studied compound 36 and the more re-
cent compounds 47, 5a,8 5b9 and 6.10
The results of a radioligand human binding assay for the
human TRa1 and TRb1, and its calculated TR-isoform
selectivity is summarized in Table 1.
Based on the concept of ‘indirect antagonism’ of estro-
gen receptors (ERs),11,12 we previously showed that in-
stead of direct antagonism, antagonism for TR also
can be achieved through ligands that are smaller than di-
rect antagonists and even agonists (indirect antago-
nism).13 This work provided ligands where the ring
containing the phenol moiety in thyromimetics (outer
ring) was replaced with straight, branched or cyclic alkyl
entities. The best binder among these new ligands, a
cyclohexyl derivative (7), gave affinity comparable with
already reported direct TR-antagonists 3, 4 and 5a,
but less potent compared to the most recently reported
direct TR antagonists 5b and 6.
Affinity for TR-binding was marginally changed when
the methylcyclohexyl entity (7) was replaced with a ben-
zyl group (8), which encouraged us to start substituting
the outer ring with additional hydrophobic bulk (i.e.,
methyl). From these results, it is obvious that optimal
position of a methyl group is substitution at the R3-po-
sition (11b) of the benzyl group, which was superior for
TR-binding versus the other ligands with R2-methyl
(11a) or R4-methyl (11c) substitution. Compared with
the unsubstituted benzyl derivative (8), affinity for TR-
binding approximately doubled for 11b. Further substi-
tution of the R3-position with trifluoromethane (11d),
chlorine (11e) and bromine (11f) revealed that 11f was
optimal in terms of affinity as well as for TR-receptor
subtype selectivity. When the R3-position was substitut-
ed with a secondary amine (14) and a tertiary amine (15)
TR-affinity was even further improved, even though the
subtype selectivity with these ligand was somewhat
leaned towards TRb1. Ligands 11f and 14 had signifi-
cantly improved affinities (lower IC50’s) when compared
with the previously reported ‘indirect’ ligand 7, while
affinity was at least within the same range as for the ‘di-
rect’ antagonists 5a, 5b and 6.
Due to the greater ease of introducing a more variable
set of functional groups at different positions, we decid-
ed to utilize benzyl instead of methylcyclohexyl as the
outer ring. This strategy is outlined in Figure 3. Further-
Figure 3. Chemical structures of 7 and 8.
Scheme 1. Reagents and conditions: (a) K2CO3, acetone, D; (b) NaOH, MeOH, rt; (c) Na2S2O4, EtOH, 70 ꢁC; (d) NaBH4, CH3COOH, rt; (e)
NaBH4, CH3COOH, 60 ꢁC.

2020
J. Malm et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2018–2021
Table 1. Thyroid hormone receptor-binding affinities (IC50 or KD’s) of
1 and synthetic thyromimetics 4, 5a, 5b, 6–8, 11a–f and 13–15a
b
Compound
hTRa1 (IC50
)
hTRb1 (IC50
)
hTRa1/hTRb1
1
0.24
1600
0.26
910
35 12
0.54
1.0
5.7
4.6
1.6
1.4
1.7
3.8
1.8
0.98
1.4
1.5
0.97
4.5
3.4
3.2
4
5a
5b
6
200 60
93 29
20
22
7
3
36
460
680
1300
360
600
370
500
99
3
7
190
230
200
120
360
160
190
60
8
11a
11b
11c
11d
11e
11f
13
14
15
2600
78
820
23
480
87
a IC50 values are expressed as nM and are calculated means of dupli-
cate runs for 1, 4, 7, 8, 11a–f and 13–15. Data for 1, 4, 5a 5b, 6 and 7
are taken from Refs. 7–10, 13 and 15, respectively. These data are
intended for comparison with the new ligands.
b Normalized selectivity: IC50 hTRa1/(IC50 hTRb1 · 1.7). For an
explanation, as well as experimental procedures, see Ref. 15 (and
references cited therein).
Figure 4. Transcriptional effects of 11f + 1 (L-T3) (red curve) and 11f
only (blue curve) on CHOK1-cells (Chinese hamster ovary cells),
stably transfected with hTRa1 and with an alkaline phosphate (ALP)
reporter gene downstream to a thyroid response element (TRAFa1).17
The y-axis is expressed as light units emitted from ALP and the x-axis
as log concentration of added ligand. The concentration of 11f
required for 50% inhibition of L-T3 (1) is approximately 90 nM. The
response value for each concentration of ligand is the mean of
triplicate determinations SD for each value indicated.
Table 2. Thyroid hormone receptor-binding affinities (IC50) of 1 and
synthetic thyromimetics 11f and 14a
Table 3. Various nuclear receptor-binding affinities (IC50) for 11fa
Compound
hTRAFa1
(IC50
hTRAFb1
(IC50
hTRAFa1
hTRAFa1
AR
ERa
GR
PR
MR
)
)
(EC50
)
(EC50
)
IC50
7700
29,000
14,000
2500
8700
11f
14
380 (80)b
NE
230 (63)
NE
NEc
23 (88)
NE
49 (76)
a IC50 values are expressed as nM and are calculated means of dupli-
cate runs. AR, androgen receptor; ERa, extrogen alpha receptor;
GR, glucocorticoid receptor; PR, progesterone receptor; MR, min-
eral corticoid receptor. The data are the result of a single measure-
ment and constructed and measured analogously to the TR-binding
assay described in Ref. 15.
a IC50 and EC50 values are expressed as nM and are calculated means
of duplicate runs for 11f and 14. The variability was approximately
25% between the runs. Experimental procedures for this assay can be
found in Ref. 15 (and references cited therein).
b Values within parentheses refer to % antagonism or agonism,
respectively.
c NE, no effect or very weak effect.
As 11f appeared to be the overall best ligand in terms of
affinity and antagonism action, it was further tested for
other nuclear receptors (NRs). This is shown in Table 3.
The result of this NR-receptor screen revealed that selec-
tivity for most receptors was almost >100 times, except
for the progesterone receptor (PR) where selectivity for
TRa1 over PR binding is only 25 times. This finding
clearly represents a potential problem and further struc-
ture–activity relation (SAR) work has to be conducted
in order to avoid this unwanted activity.
The two most promising ligands in terms of affinity and
selectivity in the TR-binding assay, 11f and 14, were
tested in a reporter cell assay employing Chinese ham-
ster ovary cells stably transfected with hTRa1 or hTRb1
and an alkaline phosphate reporter gene downstream
from a thyroid hormone receptor response element
(TRAFa1 and TRAFb1) as shown in Table 2.
Interestingly, 14 behaves like an agonist in the transacti-
vation assay, while 11f displays an expected TR-antago-
nism. This might be due to the H-binding capacity of 14
to His-435 (TRb1-numbering),16 which eventually by
virtue of its own can re-arrange the ligand binding pock-
et into an agonist conformation. As mentioned before in
the literature this property is likely to be important for
functional activity.3
In summary, we have shown that the concept of ‘indirect
antagonism’ among NRs might be further rationalized
also for TRs. Compared with our previous work, incre-
mental improvement for affinity was achieved for this
class of ligands. Compared with already reported ‘direct’
TR-antagonists, affinity for this ‘indirect’ TR-antagonist
11f was clearly within the same range. Optimal affinity,
selectivity, and antagonist activity were found with 11f,
but most probably there is a limitation to size of groups
that can be accommodated within this region by the
ligand-binding pocket of TR, as well as a dependence
on the exact orientation. We also revealed that the
A representative graph for TRAFa1 with 11f is shown in
Figure 4. The graphs clearly show that 11f is a compet-
itive antagonist for TRa1 in this cell system.

J. Malm et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2018–2021
2021
Kollman, P. A.; Jorgensen, E. C. J. Med. Chem. 1977, 20,
863; (g) Koerner, D.; Schwartz, H. L.; Surks, M. L.;
Oppenheimer, J. H. J. Biol. Chem. 1975, 250, 6417.
presence of an H donator in the outer ring will turn this
class of ligands into agonists in a transactivation assay.
7. Baxter, J. D.; Goede, P.; Apriletti, J. W.; West, B. L.;
Feng, W.; Mellstro¨m, K.; Fletterick, R. J.; Wagner, R. L.;
Kushner, P. J.; Ribeiro, R. C. J.; Webb, P.; Scanlan, T. S.;
Nilsson, S. Endocrinology 2002, 143, 517.
8. Chiellini, G.; Nguyen, N. H.; Apriletti, J. W.; Baxter, J.
D.; Scanlan, T. S. Bioorg. Med. Chem. Lett. 2002, 10, 333.
9. (a) Nguyen, N. H.; Apriletti, J. W.; Lima, S. T. C.; Webb,
P.; Baxter, J. D.; Scanlan, T. S. J. Med. Chem. 2002, 45,
3310; (b) Lim, W.; Nguyen, N. H.; Yang, H. Y.; Scanlan,
T. S.; Furlow, J. D. J. Biol. Chem. 2002, 277, 35664.
10. Koehler, K.; Gordon, S.; Brandt, P.; Carlsson, B.;
Ba¨cksbro-Saeidi, A.; Apelqvist, T.; Agback, P.; Grover,
Acknowledgment
Beatriz Campos is acknowledged for her support in
compound handling and recording of MS-spectra.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.
2007.01.009.
˚
G. J.; Nelson, W.; Grynfarb, M.; Fa¨rnegardh, M.;
Rehnmark, S.; Malm, J. J. Med. Chem. 2006, 49, 6635.
11. Shiau, A. K.; Barstad, D.; Radek, J. T.; Meyers, M. J.;
Nettles, K. W.; Katzenellenbogen, B. S.; Katzenellenbo-
gen, J. A.; Agard, D. A.; Greene, G. L. Nat. Struct. Biol.
2002, 9, 359.
References and notes
1. (a) Evans, R. M. Science 1988, 240, 889; (b) Mangelsdorf,
D. J.; Thummel, C.; Beato, M.; Herrlich, P.; Schutz, G.;
Umesono, K.; Blumberg, B.; Kastner, P.; Mark, M.;
Chambon, P.; Evans, R. M. Cell 1995, 83, 835; (c) Ribeiro,
R. C.; Kushner, P. J.; Baxter, J. D. Annu. Rev. Med. 1995,
46, 443.
12. The distinction between ‘direct’, ‘indirect’ and ‘passive’
antagonists was outlined in Ref. 13 below (and therein
Ref. 35).
13. Hedfors, A.; Appelqvist, T.; Carlsson, B.; Bladh, L.-B.;
Litten, C.; Agback, P.; Grynfarb, M.; Koehler, K. F.;
Malm, J. J. Med. Chem. 2005, 48, 3114.
˚
2. (a) Lazar, M. A. Endocr. Rev. 1993, 14, 184; (b) Yen, P.
M. Phys. Rev. 2001, 81, 1097.
14. Experimental procedures and analytical data can be found
in the Supplemental material.
3. (a) Malm, J. Curr. Pharm. Des. 2004, 10, 3522; (b) Webb,
P. Exp. Opin. Invest. Drugs 2004, 13, 489.
4. Braverman, L. E. Clin. Chem. 1996, 42, 174.
15. Ye, L.; Li, Y.-L.; Mellstro¨m, K.; Bladh, L.-G.; Koehler,
K.; Garg, N.; Garcia Collazo, A. M.; Litten, C.; Husman,
B.; Persson, K.; Ljunggren, J.; Grover, G.; Sleph, P. G.;
George, R.; Malm, J. J. Med. Chem. 2003, 46, 1580.
16. TRs display unconserved N-terminal domains and in
humans and other species TRa1, TRb1 and TRb2 the
difference in length is due to the variable N-terminal
domain. Consequently, the amino acid sequence num-
bering will be different for isoforms and splice
variants.
5. Gittoes, N. J.; Franklyn, J. A. Drugs 1998, 55, 543.
6. (a) Reid, D. G.; Maclachlan, L. K.; Voyle, M.; Leeson, P. D.
J. Biol. Chem. 1989, 264, 2013; (b) Shulkin, B. L.; Bolger, M.
B.; Utiger, R. D. J. Endocrinol. Invest. 1988, 11, 657; (c)
Koehrle, J.; Auf’mkolk, M.; Rokos, H.; Hesch, R. D.;
Cody, V. J. Biol. Chem. 1986, 261, 11613; (d) Tropsha, A. E.;
Rakhmaninova, A. B.; Iaguzhinskii, L. S. Bioorg. Khim.
1984, 10, 483; (e) Bolger, M. B.; Jorgensen, E. C. J. Biol.
Chem. 1980, 255, 10271; (f) Dietrich, S. W.; Bolger, M. B.;
17. Carlsson, B.; Singh, B. N.; Temciuc, M.; Nilsson, S.; Li,
Y.-L.; Mellin, C.; Malm, J. J. Med. Chem. 2002, 45, 623.