Benzoxazoles as transthyretin amyloid fibril inhibitors: Synthesis, evaluation, and mechanism of action

Article abstract of DOI:10.1002/anie.200351179

Benzoxazoles pevent misfolding: Benzoxazole-based inhibitors of transthyretin (TTR) amyloid fibril formation are among the most effective found to date. They stabilize TTR against both acid-mediated misfolding and urea denaturation by raising the activati

Full text of DOI:10.1002/anie.200351179

Communications
Small-Molecule Inhibitors
Benzoxazoles as Transthyretin Amyloid Fibril
Inhibitors: Synthesis, Evaluation, and Mechanism
of Action**
Hossein Razavi, Satheesh K. Palaninathan,
Evan T. Powers, R. Luke Wiseman, Hans E. Purkey,
Nilofar N. Mohamedmohaideen, Songpon Deechongkit,
Kyle P. Chiang, Maria T. A. Dendle,
James C. Sacchettini, andJeffery W. Kelly*
Transthyretin (TTR) is one of twenty nonhomologous pro-
teins, the misfolding, aggregation, and deposition (amyloido-
genesis) of which is linked to amyloid disease.[1,2] TTR is a
homotetramer composed of 127amino acid subunits [3] that
carries thyroxine[4] and holo-retinol binding protein[5] in the
blood. The TTR tetramer is non-amyloidogenic, but under-
goes dissociation, monomer misfolding, and misassembly into
numerous aggregated structures including amyloid under
partially denaturing conditions (for example, at low pH val-
ues).[6] Amyloidogenesis by wild type (wt) TTR appears to
cause senile systemic amyloidosis,[7] whereas amyloidogenesis
by one of over 80 TTR mutants results in either familial
amyloid polyneuropathy[8] or cardiomyopathy.[9] Incorpora-
tion of T119M trans-suppressor subunits into tetramers
otherwise composed of disease-associated subunits (such as
V30M or L55P) inhibits amyloidogenesis in vitro and ame-
loriates disease in V30M carriers by increasing the dissocia-
tive activation barrier and thereby kinetically stabilizing the
tetramer.[10,11] TTR has two largely unoccupied thyroxine
[*] Prof. J. W. Kelly, Dr. H. Razavi, Prof. E. T. Powers, R. L. Wiseman,
Dr. H. E. Purkey, S. Deechongkit, K. P. Chiang, M. T. A. Dendle
Department of Chemistry and
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
La Jolla, CA 92037 (USA)
Fax: (+1)858-784-9610
E-mail: jkelly@scripps.edu
Dr. S. K. Palaninathan, Dr. N. N. Mohamedmohaideen,
Prof. J. C. Sacchettini
Department of Biochemistry and Biophysics
Texas A & M University (USA)
[**] This work was supported by the NIH (DK 46335), the Skaggs
Institute for Chemical Biology, and the Lita Annenberg Hazen
Foundation. H.R. thanks the Skaggs Institute for Chemical Biology
for a postdoctoral fellowship. Additional support was provided by
the Robert A. Welch Foundation (J.C.S). Use of the Argonne
National Laboratory Structural Biology Center beam lines at the
Advanced Photon Source was supported by the United States
Department of Energy, Office of Energy Research, under contract
W-31-109-ENG-38. Use of the BioCARSSector 14 was supported by
the National Institutes of Health National Center for Research
Resources. We thank the General Clinical Research Center of The
Scripps Research Institute (supported by NIH grant RR00833) for
providing blood plasma through the normal donor blood drawing
program.
Supporting information for this article (experimental and crystallo-
graphic details) is available on the WWW under http://www.ange-
wandte.org or from the author.
2758
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200351179
Angew. Chem. Int. Ed. 2003, 42, 2758 – 2761

Angewandte
Chemie
binding sites that are created by its quaternary
structural interface.[12] The tetramer can be
stabilized by small-molecule-binding to these
sites, potentially providing a means to treat TTR
amyloid disease with small-molecule drugs.[13]
Many families of compounds have been discov-
ered, the binding of which stabilizes the tetra-
meric ground state to a degree proportional to
the small-molecule dissociation constants Kd1
and Kd2.[11] This binding also effectively increases
the dissociative activation barrier and inhibits
amyloidosis by kinetic stabilization.[11] Such
inhibitors are typically composed of two aro-
matic rings, with one ring bearing halogen
substituents and the other bearing hydrophilic
substituents.[13–20] Benzoxazoles substituted with
a carboxylic acid at C4–C7and a halogenated
phenyl ring at C2 also appeared to complement
the TTR thyroxine binding site. A small library
of these compounds was therefore prepared by
dehydrocyclization of N-acyl amino hydroxy-
benzoic acids (Scheme 1).[21,22]
The benzoxazoles were evaluated by using a
series of analyses of increasing stringency. A
previously developed fibril-formation assay was
used as the first screen.[13,15] wtTTR (3.6 mm) was
incubated for 30 min (pH 7, 378C) with a test
compound (7.2 mm). Since at least one molecule
Figure 1. Suppression of TTR fibril formation by compounds 1–28. The position of the carboxy
group on the benzoxazole is shown on the left-hand side, while the C2 phenyl ring is shown at
the bottom. The bars indicate the percent fibril formation (ff), that is, the amount of fibrils
formed from TTR in the presence of 1–28 relative to the amount formed by TTR in the absence
of inhibitor (which is defined as 100%). The most effective compounds (lowest ff) are indicated
in red.
cipitated using a sepharose-bound polyclonal TTR antibody.
The TTR with or without bound inhibitor was liberated from
the resin at high pH value, and the inhibitor:TTR stoichiom-
etry was ascertained by HPLC analysis (Figure 2).[23] Benzox-
azoles with carboxylic acids in the 5- or 6-position, and 2,6-
dichlorophenyl (13, 20) or 2-trifluoromethylphenyl (11, 18)
substituents at the 2-position displayed the highest binding
stoichiometries. In particular, 20 exhibited excellent inhib-
itory activity and binding selectivity. Hence, its mechanism of
action was characterized further.
Scheme 1. General synthesis of benzoxazoles: a) ArCOCl, THF, pyri-
dine (Ar=phenyl, 3,5-difluorophenyl, 2,6-difluorophenyl, 3,5-dichloro-
phenyl, 2,6-dichlorophenyl, 2-(trifluoromethyl)phenyl, and 3-(trifluoro-
methyl)phenyl); b) TsOH·H2O (Ts=tosyl), refluxing xylenes;
c) TMSCHN2, benzene, MeOH; d) LiOH, THF, MeOH, H2O (8–27%
yield over four steps).
To confirm that 20 inhibits TTR fibril formation by
binding strongly to the tetramer, isothermal titration calo-
of the test compound must bind to each molecule of TTR
tetramer to be able to stabilize it, a test compound concen-
tration of 7.2 mm is only twice the minimum effective
concentration. The pH value was then adjusted to 4.4, the
optimal pH for fibrilization. The amount of amyloid formed
after 72 h (378C) in the presence of the test compound was
determined by turbidity at 400 nm and is expressed as % fibril
formation (ff), 100% being the amount formed by TTR alone.
Of the 28 compounds tested, 11 reduced fibril formation to
negligible levels (ff < 10%; Figure 1, red bars).
The 11 most-active compounds were then evaluated for
their ability to bind selectively to TTR over all other proteins
in the blood.[23] Human blood plasma (TTR conc. 3.6–5.4 mm)
was incubated for 24 h with the test compound (10.8 mm) at
378C. The TTR and any bound inhibitor were immunopre-
Figure 2. Stoichiometry (s) of benzoxazoles bound to TTR after incuba-
tion in human blood plasma; the maximum possible value of s is 2.
The thin vertical lines indicate the measurement error. The stoichiome-
tries shown are not corrected for wash-associated loss of small mole-
cules and therefore are lower limits for the true stoichiometries.
Angew. Chem. Int. Ed. 2003, 42, 2758 – 2761
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2759

Communications
rimetry (ITC) and sedimentation velocity experiments were
The role of the substituents in 20 is evident from its
cocrystal structure with TTR (Figure 4). Compound 20
orients its two chlorine atoms near halogen binding pockets
2 and 2’ (so-called because they are occupied by iodine atoms
when thyroxine binds to TTR). The 2,6 substitution pattern
conducted with wtTTR. ITC showed that two equivalents of
20 bind with average dissociation constants of Kd1 = Kd2 = 55
(Æ 10) nm under physiological conditions. These values are
comparable to the dissociation constants of many other highly
efficacious TTR amyloidogenesis inhibitors. For the sedimen-
tation velocity experiments, TTR (3.6 mm) was incubated with
20 (3.6 mm, 7 .2mm, 36 mm) under optimal fibrilization con-
ditions (72 h, pH 4.4, 378C). The tetramer (55 kDa) was the
only detectable species in solution with 20 at 7.2 or 36 mm.
Some large aggregates formed with 20 at 3.6 mm, but the TTR
remaining in solution was tetrameric.
T119M subunit inclusion and small-molecule binding both
prevent TTR amyloid formation by raising the activation
barrier for tetramer dissociation.[11] The ability of an inhibitor
to do this is most rigorously tested by measuring its efficacy at
slowing tetramer dissociation in 6m urea, a severe denatura-
tion stress.[11] Thus, the rates of TTR tetramer dissociation in
6m urea in the presence and absence of 20, 21, or 27 were
compared (Figure 3). TTR (1.8 mm) was completely dena-
Figure 4. X-ray cocrystal structure of 20 bound to TTR. The green and
yellow structures represent its two symmetry equivalent binding
modes. The surface of the binding site is shown in gray. Equivalent
residues in different subunits are distinguished with primed and
unprimed residue numbers, as are the pairs of halogen binding
pockets (HBPs). The inset shows an entire TTR tetramer (each subunit
colored differently) with the binding site highlighted. Details of the
crystal structure are given in the Supporting Information.
Figure 3. Dissociation as a function of time (t) for wtTTR (1.8 mm) in
6m urea without inhibitor (c), or in the presence of 3.6 mm of
20 (c), 21 (c), or 27 (c), or 1.8 mm 20 (c). The dimension-
less extent of tetramer dissociation (fu, or fraction unfolded) was deter-
mined from the intensity of the circular dichroism spectrum at 215–
218 nm. Although this is a measure of secondary structure, it corre-
sponds indirectly to tetramer dissociation because dissociation must
precede denaturation, and monomer denaturation is virtually irreversi-
ble and instantaneous in 6m urea (the half time for denaturation is
approximately 70 ms).
on the phenyl ring forces the benzoxazole and phenyl rings
out of planarity, optimally positioning the carboxylic acid on
+
the benzoxazole unit to hydrogen bond with the e-NH3
groups of Lys 15/15’. Hydrophobic and van der Waals inter-
actions between the aromatic rings of 20 and the side chains of
Leu17, Leu110, Ser117, and Val121 contribute additional
binding energy (the side chains of Leu17and leu110 obscure
the view of 20 in Figure 4 and were therefore omitted for
clarity).
In summary, 28 benzoxazole compounds were each pre-
pared in four steps. Many of these compounds effectively
inhibited TTR amyloid formation in vitro and bound selec-
tively to TTR in blood plasma. Urea denaturation experi-
ments demonstrated that 20 was particularly effective at
kinetic stabilization of the TTR tetramer. The X-ray crystal
structure of the TTR–20 complex provides a structural basis
for understanding its ability to stabilize the native state. The
benzoxazole compounds presented here represent a new class
of TTR inhibitors that may have better pharmacological
properties than those described previously.[13–20] Their efficacy
in vivo will be determined in animal models for TTR
amyloidosis.[24]
tured after 168 h in 6m urea. In contrast, 20 at 3.6 mm
prevented tetramer dissociation for at least 168 h (> 3 ” the
half-life of TTR in human plasma). With an equimolar
amount of 20, only 27% of TTR denatured in 168 h.
Compound 27 (3.6 mm) was much less able to prevent
tetramer dissociation (90% unfolding after 168 h), even
though it was active in the fibril formation assay. Compound
21 did not hinder the dissociation of TTR at all. These results
show that inhibitor binding to TTR is necessary but not
sufficient to kinetically stabilize the TTR tetramer under
strongly denaturing conditions; it is also important that the
dissociation constants are very low (or that the off rates are
very slow). Also, the arrangement of functional groups on 20 is
apparently optimal for stabilizing the TTR tetramer; moving
the carboxylic acid from C6 to C7, as in 27, or removing the
chlorines, as in 21, severely diminishes its activity.
2760
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 2758 – 2761

Angewandte
Chemie
Experimental Section
The general procedure for benzoxazole synthesis and complete
characterization of the products (1H and 13C NMR spectroscopy and
high-resolution mass spectra) are given in the Supporting Informa-
tion. The analytical ultracentrifugation and X-ray crystallography
experiments are also detailed in the Supporting Information. The
procedures for the fibril formation assay,[13,15] the immunoprecipita-
tion assay,[23] ITC[14,20] and urea-induced dissociation kinetics[11] have
been described elsewhere.
Received: February 13, 2003 [Z51179]
Keywords: aggregation · bioorganic chemistry · drug design ·
.
inhibitors · structure–activity relationships
[1] J. D. Sipe, Crit. Rev. Clin. Lab. Sci. 1994, 31, 325.
[2] J. W. Kelly, Curr. Opin. Struct. Biol. 1996, 6, 11.
[3] C. C. Blake, M. J. Geisow, S. J. Oatley, B. Rerat, C. Rerat, J. Mol.
Biol. 1978, 121, 339.
[4] L. Bartalena, J. Robbins, Clin. Lab. Med. 1993, 13, 583.
[5] H. L. Monaco, Biochim. Biophys. Acta 2000, 1482, 65.
[6] J. W. Kelly, W. Colon, Z. Lai, H. A. Lashuel, J. McCulloch, S. L.
McCutchen, G. J. Miroy, S. A. Peterson, Adv. Protein Chem.
1997, 50, 161.
[7] R. E. McCarthy III, E. K. Kasper, Clin. Cardiol. 1998, 21, 547.
[8] V. Plante-Bordeneuve, G. Said, Curr. Opin. Neurobiol. 2000, 13,
569.
[9] D. R. Jacobson, R. D. Pastore, R. Yaghoubian, I. Kane, G. Gallo,
F. S. Buck, J. N. Buxbaum, N. Engl. J. Med. 1997, 336, 466.
[10] P. Hammarstrom, F. Schneider, J. W. Kelly, Science 2001, 293,
2459.
[11] P. Hammarstrom, R. L. Wiseman, E. T. Powers, J. W. Kelly,
Science 2003, 299, 713.
[12] J. C. Sacchettini, J. W. Kelly, Nat. Rev. Drug Discov. 2002, 1, 267..
[13] G. J. Miroy, Z. Lai, H. A. Lashuel, S. A. Peterson, C. Strang, J. W.
Kelly, Proc. Natl. Acad. Sci. USA 1996, 93, 15051.
[14] S. A. Peterson, T. Klabunde, H. A. Lashuel, H. Purkey, J. C.
Sacchettini, J. W. Kelly, Proc. Natl. Acad. Sci. USA 1998, 95,
12956.
[15] P. W. Baures, S. A. Peterson, J. W. Kelly, Bioorg. Med. Chem.
1998, 6, 1389.
[16] P. W. Baures, V. B. Oza, S. A. Peterson, J. W. Kelly, Bioorg. Med.
Chem. 1999, 7, 1339.
[17] V. B. Oza, H. M. Petrassi, H. E. Purkey, J. W. Kelly, Bioorg. Med.
Chem. Lett. 1999, 9, 1.
[18] T. Klabunde, H. M. Petrassi, V. B. Oza, P. Raman, J. W. Kelly,
J. C. Sacchettini, Nat. Struct. Biol. 2000, 7, 312.
[19] H. M. Petrassi, T. Klabunde, J. Sacchettini, J. W. Kelly, J. Am.
Chem. Soc. 2000, 122, 2178.
[20] V. B. Oza, C. Smith, P. Raman, E. K. Koepf, H. A. Lashuel,
H. M. Petrassi, K. P. Chiang, E. T. Powers, J. Sacchettinni, J. W.
Kelly, J. Med. Chem. 2002, 45, 321.
[21] M. Terashima, M. Ishii, Synthesis 1982, 484.
[22] M. R. DeLuca, S. M. Kerwin, Tetrahedron 1997, 53, 457.
[23] H. E. Purkey, M. I. Dorrell, J. W. Kelly, Proc. Natl. Acad. Sci.
USA 2001, 98, 5566.
[24] M. H. Teng, J. Y. Yin, R. Vidal, J. Ghiso, A. Kumar, R. Rabenou,
A. Shah, D. R. Jacobson, C. Tagoe, G. Gallo, J. Buxbaum, Lab.
Invest. 2001, 81, 385.
Angew. Chem. Int. Ed. 2003, 42, 2758 – 2761
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2761