Semisynthesis and pharmacological activities of thyroxine analogs: Development of new angiogenesis modulators

Article abstract of DOI:10.1016/j.bmcl.2010.04.011

Novel thyroxine analogs with hindered phenol, amino and carboxylic acid groups have been synthesized and the effects of the synthesized compounds on angiogenesis using the chick chorioallantoic membrane and mouse matrigel models have been tested. Pharmaco

Full text of DOI:10.1016/j.bmcl.2010.04.011

Bioorganic & Medicinal Chemistry Letters 20 (2010) 3394–3398
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Semisynthesis and pharmacological activities of thyroxine analogs:
Development of new angiogenesis modulators
Alexandre Bridoux a, Huadong Cui a, Evgeny Dyskin a, Andreea-Ruxandra Schmitzer b, Murat Yalcin a,c
,
Shaker A. Mousa a,d,
*
a The Pharmaceutical Research Institute (PRI) at Albany College of Pharmacy and Health Sciences, One Discovery Drive, Rensselaer, NY 12144, USA
b Département de Chimie, Université de Montréal, C.P. 6128, Succursale Centre-ville, H3C 3J7, Montréal, Canada
c Uludag University, Veterinary Medicine Faculty, Department of Physiology, Bursa, Turkey
d King Saud University, Riyadh, Saudi Arabia
a r t i c l e i n f o
a b s t r a c t
Article history:
Novel thyroxine analogs with hindered phenol, amino and carboxylic acid groups have been synthesized
and the effects of the synthesized compounds on angiogenesis using the chick chorioallantoic membrane
and mouse matrigel models have been tested. Pharmacological profiles revealed that thyroxine tolerates
numerous modifications on the amino group and remains active. These results provide the rationale for
the selection of a novel thyroxine nanoparticle precursor.
Received 2 October 2009
Revised 3 April 2010
Accepted 7 April 2010
Available online 11 April 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Thyroxine
Esterification
Silylation
UV–vis
Angiogenesis
Nanopharmaceutical technology
Angiogenesis is as a vascular process that involves the induction
of growth of new blood vessels through cell adhesion to the extra
cellular matrix (ECM).1 For the treatment of aneurysm, heart attack
(e.g., acute myocardial infarction), stroke, or peripheral artery dis-
ease, there has been a critical need for an orally active compound
to boost angiogenesis with an optimal therapeutic index and min-
imal impact on other physiological processes.2 In the chick cho-
rioallantoic membrane (CAM) assay, the thyroid hormone T4 (2-
amino-3-[4-(4-hydroxy-3,5-diiodophenoxy)-3,5-diiodophe-
could be used in conjunction with targeted nanoparticle drug
delivery to stimulate angiogenesis. This novel nanoformulation
Table 1
Structures of T4 and T4 analogs 7–15
I
R3O
R2
I
R1
I
O
nyl]propionate, thyroxine) and the T4 analogs T3 (2-amino-3-[4-(4-
hydroxy-3-iodophenoxy)-3,5-diiodophenyl]propionate), and GC-1
([4-(4-hydroxy-3-isopropylphenoxy)-3,5-dimethylphenoxy]acetic
acid) stimulate angiogenesis as effectively as vascular endothelial
growth factor (VEGF) or fibroblast growth factor (FGF). T4 (Table
1, compound 1) and T4 analogs have also been identified as potent
I
Compounds
R1
R2
R3
ꢁ
Thyroxine 1
7
8
9
10
11
12
13
ꢀNH3
ꢀCOOꢂ
–COOH
–COOH
–COOH
–COOH
–COOH
–COOEt
–COOTiPS
–H
–H
–H
–Ac
–H
–H
–H
–H
–NHBoc
–NHBu
–NHAc
–NHMe
–NHEt
activators of integrin a
vb3, a cell adhesion molecule.3–5 The relative
pKa values (dissociation constants) of the functional groups of T4
have been estimated to be approximately 9.0 for the NH2 (amino)
group and 2.0 for the COOH (carboxylic acid) group. The PhOH
(phenol) group is a weakly basic site as determined by fluorescent
quenching (pKa = 6.2).6 The goal of the current study was to semi-
synthesize a new water-soluble formulation of thyroxine that
–NHEt
–NHBoc
-COO
14
15
–NHBoc
–NHBoc
O
O
O
–COOTiPS
* Corresponding author. Tel.: +1 518 694 7297; fax: +1 518 694 7567.
E-mail address: Shaker.MOUSA@acphs.edu (S.A. Mousa).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.04.011

A. Bridoux et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3394–3398
3395
would be unable to gain access to the cell, and thus work only at
the cell surface. As part of this approach, suitable intermediates
were synthesized to determine the relative importance of the
NH2 group and the distance between the phenyl and COOH group
of T4 to pro-angiogenic activity.
epoxide-based linker.8 Thus, T4 was activated by alkylation after
the formation of carbon- or silicon-based esters and protection of
the NH2 group. The compounds were then tested for their ability
to stimulate angiogenesis. It is possible that the epoxide, when
linked to the OH (hydroxyl) of the PhOH, interacted with a serine
or tyrosine residue at or near the binding site of the integrin
receptor.9,10
A computer-aided chain ramification design in which the propi-
onic amino acid end of T4 had been modified was used to determine
the fit of derivatives into the integrin binding site.11 Several candi-
date modified analogs were identified that would be amenable to
a linear synthesis strategy of nanoparticle precursor (Table 1).12–14
The program was allowed to insert the designed T4 analogs into
A nanoscale particle consisting of poly-L-lactide-co-glycolide
(PLGA) and polyvinyl alcohol (PVA) saturated on its surface with
NH2 groups was selected as the nanoformulation for T4.7 N-ethyl-
N0-(3-dimethylaminopropyl)carbodiimide (EDC)-assisted amide
formation between the lyophilized polymer and ethylenediamine
was found to be the best pathway for obtaining NH2 groups on
the surface of the nanoparticle. Association of the activated
hormone with the nanoparticle would be mediated through an
Figure 1. (A) Ribbon diagram of the two protein subunits (yellow and orange) of integrin avb3. (B) Docking of the natural ligand (green), tetrac (white), and compound 7
(purple) between the two protein subunits.
Figure 2. (A) UV-vis spectra of compounds 12, 13, 14, 15 and 9 (from top to bottom). (B) UV-vis spectra of compounds 7, 11, 8 and 10 (from top to bottom).

3396
A. Bridoux et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3394–3398
Table 2
sium carbonate) (1.0 equiv). In method C, no bases were used and
Pro-angiogenic effects of T4 analogs in the CAM model
the reactions were allowed to proceed until thermodynamic equi-
librium was obtained. Cs2CO3 slightly increased the rate of the
reactions, but not as much as TEA (method A). The best yields were
obtained when T4 zwitterion was solely used (Method C), particu-
larly for compound 7.
Treatment
Mean branch pointsa
Mean % of stimulationb
(comparing to PBS)
PBS (control)
FGF (1.25
T4
8
9
10
11
FGF + 12
13
73
6
—
lg/mL)
130 11c
137 8c
98 11d
122 4c
105 7d
136 13c
46 2e
78 7c
88 7c
The COOH group was then modified into a carbon- or silicon-
based ester.15 Interestingly, the carboxylate group of analog 7 re-
acted better with a cluttered silicon-based esterifying agent. It re-
acted quantitatively with the TiPSCl group, but did not react with
the more electrophilic TMSCl group.16 Although recent data sug-
gested the existence of a putative siliconium ion, no further studies
were done to elucidate the mechanism of this reaction.17,18 T4 was
protected using the Boc reagent and then esterified with TiPSCl
(1.0 equiv) in situ after another round of deprotonation using
TEA (1.0 equiv) in anhydrous THF (tetrahydrofuran) at room tem-
perature to yield compound 13 (Supplementary data, Fig. S2). This
compound was isolated, epibromohydrin was added in excess, and
then anhydrous reflux conditions were applied to generate com-
pound 15, which was subsequently isolated after purification by
silica gel chromatography (40% overall yield). Interestingly, while
the general trends in the ultraviolet spectra of all pure products
were very similar, certain features were different. Monosubstituted
analogs of the COOH group had one characteristic absorption
wavelength of 304 nm in addition to one near 250 nm, whereas
any disubstituted analog had two maxima in the region at 290
and 299 nm (Fig. 2).
To investigate the structure–activity relationships of the deriv-
atives, the effect of each compound on angiogenesis was compared
to T4 using two model systems: the CAM model, a widely used
in vivo model in which the chicken embryonic membrane is ex-
posed to putative stimulatory and inhibitory compounds;15,19
and the matrigel assay, in which putative angiogenesis-inducing
and blocking compounds are incorporated into cold liquid Matri-
gel, which is then injected subcutaneously, serving as a scaffold
for the formation of new blood vessels.16,20
34 8d
67 5c
44 6d
86 9c
ꢀ38 4c (inhibition)
76
7
4
6
14
15
114 11c
98 8d
56 7c
34 7d
a
b
c
Data represent the means SEM (n = 8).
All thyroid analogs were tested at a concentration of 10
P <0.01 as compared to PBS (control).
P <0.05.
lg/CAM.
d
e
P <0.01 as compared to FGF.
the integrin binding pocket with no constraints imposed for the site
of docking. In other words, the ‘ligands’ were not assigned a partic-
ular place to bind. Fifty conformers of the lowest energy were con-
sidered for each compound probed. For example, of the fifty
conformers for compound 7, 38 of the ‘ligands’ positioned them-
selves at the same place (Fig. 1). A choice of different anhydrides
[Boc2O (di-tert-butyl dicarbonate), Bu2O (butyric anhydride), Ac2O
(acetic anhydride)], silylating agents [TMSCl (trimethylsilyl chlo-
ride), TiPSCl (triisopropylsilyl chloride)], and electrophilic alkylating
agents [MeI (iodomethane), EtI (ethyl iodide), epibromohydrin] has
been shown to modulate the functional groups according to nucle-
ophilicity and steric hindrance.
For the synthesis of novel analogs, a comparative study of yield
and kinetics of acylation and alkylation of T4 was carried out using
three different methods (A, B and C; Supplementary data) in order
to find the most suitable and reproducible synthesis protocol. In
method A, TEA (triethylamine) (1.0 equiv) was used to activate
the NH2 group prior to the addition of functional agents. In method
B, the organic base was replaced by an inorganic base (Cs2CO3; ce-
Compounds were considered strong agonists if they induced the
same level of stimulation as the reference molecules (Table 2). T4
analogs substituted at the NH2 and OH ends had a stimulatory ef-
Figure 3. Representative images showing the effect on angiogenesis of b-FGF, T4 and 11 (pro-angiogenic); 12 (anti-angiogenic); and 15 (no activity).

A. Bridoux et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3394–3398
3397
fect on new blood vessel formation, but the effect was not compa-
rable to the reference compounds. The extent of angiogenesis in-
duced by compounds in which the NH2 group was protected
with small moieties such as Me (methyl) (10), or Et (ethyl) (11)
was similar (105 and 136 branch points, respectively) to that of
free T4 (137 branch points) and FGF (130 branch points) (Fig. 3
and Table 2). Similar activity was observed for analogs 8 and 9.
Conversely, disubstituted compounds with highly lipophilic groups
(12, 13, and 15) were less effective than the control. In fact, com-
pound 12 inhibited the effect of FGF. Compound 14, in which T4
was disubstituted with a hydrogen bond acceptor at the COOH
end, stimulated angiogenesis to a lesser extent than T4. Synthe-
sized analogs 8, 9, 10, 11, 12, 13, 14, and 15 were also tested for
their ability to stimulate angiogenesis, as measured by the produc-
tion of hemoglobin (Hb), in the matrigel assay. T4 significantly in-
duced angiogenesis (Table 3). T4 analogs 8, 9, and 10 also induced
angiogenesis at levels comparable to T4 (62, 95, and 87%, respec-
tively). T4 analog 11 induced angiogenesis to a lesser extent
(45%). There was no apparent stimulation of angiogenesis by ana-
logs 12, 13, 14, and 15. T4 analog 12 appeared to inhibit T4-induced
angiogenesis (Table 4).
The results of the CAM and matrigel assays (Tables 2–4) indi-
cated that the greater the extent of hindrance of the NH2 group,
the greater the inhibition of T4 pro-angiogenic activity. That is,
all mono-acylated and mono-alkylated analogs demonstrated
clearly diminished stimulation. A second substitution on the PhOH
end (9) did not significantly reduce the potency of the drug, which
suggested that this part of the molecule did not participate in bind-
ing. A second substitution at the COOH end, on the other hand, pre-
vented stimulation completely. One analog, compound 12, had the
surprising effect of entirely inhibiting FGF and T4 activity. Com-
pared to analog 11, the presence of an Et group at the acid end
of 12 appeared to disable the molecule in terms of stimulating
the production of new blood vessels and antagonized the effects
of T4 to the same extent as tetrac ([4-(4-hydroxy-3,5-diiodophen-
oxy)-3,5-diiodophenyl]acetic acid) (Table 4).
activity. A useful compound has been validated as a precursor for
a water-soluble nanoparticle formulation of the hormone. The
results suggested the OH of the PhOH group as the best site for
condensation with a nanoparticle.
Acknowledgments
We thank the staffs at the Rensselaer Polytechnic Institute NMR
facility, University at Albany Proteomics Facility, and Center for
Functional Genomics Mass Spectrometry facility. We also thank
Ms. Susan Shipherd for reviewing this publication and Dr. Huaz-
dong He for helpful comments on the manuscript, Dr. Mary Rose
Burnham and Mr. Jason Fishel for editorial assistance, Dr. Christine
Bianchini, and Dr. Jay Jayaraman for his help and constant support.
This work was funded in part by the Charitable Leadership Founda-
tion, Medical Technology Acceleration Program and the Pharma-
ceutical Research Institute.
Supplementary data
Supplementary data (synthesis and analytical characterization
of all compounds) associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2010.04.011.
References and notes
1. Mousa, S. A. Mol. Biotechnol. 2008, 38, 33.
2. Mousa, S. A.; Davis F. B.; Davis P. J. Int. Patent WO/2008/051291, 2008.
3. Mousa, S. A.; Davis, F. B.; Mohamed, S.; Davis, P. J.; Feng, X. Int. Angiol. 2006, 25,
407.
4. Mousa, S. A.; O’Connor, L. J.; Bergh, J. J.; Davis, F. B.; Scanlan, T. S.; Davis, P. J. J.
Cardiovasc. Pharmacol. 2005, 46, 356.
5. Mousa, S. A.; Davis, F. B.; Davis, P. J. Immun., Endocr. & Metab. Agents Med. Chem.
2009, 9, 1.
6. Nilsson, S. F.; Peterson, P. A. J. Biol. Chem. 1971, 246, 6098.
7. Shakweh, M.; Besnard, M.; Nicolas, V.; Fattal, E. Eur. J. Pharm. Biopharm. 2005,
21, 1.
8. Nakajima, N.; Ikada, Y. Bioconjugate Chem. 1995, 6, 123.
9. Xiong, J.-P.; Stehle, T.; Diefenbach, B.; Zhang, R.; Dunker, R.; Scott, D. L.;
Joachimiak, A.; Goodman, S. L.; Arnaout, M. A. Science 2001, 294, 339.
10. Xiong, J. P.; Stehle, T.; Zhang, R.; Joachimiak, A.; Frech, M.; Goodman, S. L.;
Arnaout, M. A. Science 2002, 296, 151.
In summary, synthetic modifications of thyroid hormone led to a
significantly improved understanding of the functional importance
of the COOH and NH2 groups of the molecule for pro-angiogenic
11. Docking of 7 and tetrac to avb3 was performed using Autodock version 2.4. This
program allows flexibility in the ligand structure but uses a rigid body protein
approximation in order to speed up the calculation. The 1L5G.pdb file for the
protein used for docking was taken from the Brookhaven database. Partial
charges for protein atoms were taken from the AutoDock version of the AMBER
force field5. For the model ligands, flexibility was allowed for all torsion angles.
Partial atomic charges required for the docking calculation were obtained by
ab initio quantum chemistry calculations using the HyperChemÒ program
(HyperCube). Grids of probe atom interaction energies were computed first;
84 Å side grids with a spacing of 0.7 Å were used for all the modules. The ligand
probes were then docked by simulated annealing. Typically, the 50 lowest
energy coordinate sets were extracted for each ligand type and used for
visualization in the software WebLab Viewer (Molecular Simulation, Inc).
12. Weiner, S. J.; Kollman, P. A.; Nguyen, D. T.; Case, D. A. J. Comput. Chem. 1986, 7,
230.
13. Weiner, S. J.; Kollman, P. A.; Case, D. A.; Singh, U. C.; Ghio, C.; Alagona, G.;
Profeta, S., Jr.; Weiner, P. J. Am. Chem. Soc. 1984, 106, 765.
14. Morris, G. M.; Goodsell, D. S.; Huey, R.; Olson, A. J. J. Comput. Aided Mol. Des.
1996, 10, 293.
15. Bridoux, A.; Cui, H.; Dyskin, E.; Yalcin, M.; Mousa, S. A. Bioorg. Med. Chem. Lett.
2009, 19, 3259.
Table 3
Pro-angiogenic effects of T4 analogs on angiogenesis in the mouse matrigel model
Hemoglobin (mg/mL) SEMa
Mean stimulation of
angiogenesisb (%)
Control 0.3 0.1
T4
8
—
0.9 0.2
0.7 0.2
0.9 0.2
0.8 0.2
0.6 0.2
0.3 0.1
0.3 0.1
0.3 0.1
100 22
62 38
95 33
87 33
45 33
9
10
11
13
14
15
0
0
0
17
17
17
a
Data represents the means SEM (n = 6–7).
All thyroid analogs were tested at a concentration of 10 lg/matrigel plug.
b
16. Brook, M.; Chan, T. Synthesis 1983, 3, 201.
17. Lambert, J. B.; Kania, L.; Zhang, S. Chem. Rev. 1995, 95, 1191.
18. Reed, C. Acc. Chem. Res. 1998, 31, 325.
Table 4
Anti-angiogenic effects of T4 analog 12 on angiogenesis in the mouse matrigel model
19. To assess angiogenesis in the CAM model system, 10 day-old fertilized chicken
eggs (Sunrise Farms, Inc, Catskill, NY) were incubated at 37 °C in 55% relative
humidity. In the dark, with the help of a candling lamp, a small hole was
punctured in the area of the shell covering the air sac using a hypodermic
needle. A second hole was punctured on the wider side of the egg above an
avascular area of the embryonic membrane. An artificial air sac was created
below the second hole by gently applying a vacuum to the first hole using a
small rubber squeeze bulb. An approximately 1.0 cm2 window was cut in the
shell over the dropped CAM with the use of a mini drill. Filter disks (filter paper
#1, Whatman International, Ltd, United Kingdom) were soaked in 3 mg/mL
cortisone acetate solution (in 95% ethanol) and then air-dried under sterile
conditions. Sterile filter disks were saturated with L-T4 or FGF; control disks
Hemoglobin (mg/mL) SEMa
Mean inhibition of
angiogenesisb (%)
Control
T4
0.3 0.1
1.2 0.4
—
—
T4 + Tetrac 0.4 0.1
T4 + 12 0.4 0.1
89 11
89 11
a
Data represents the means SEM (n = 25).
T4, 12, and tetrac were tested at a concentration of 10 lg/matrigel plug.
b

3398
A. Bridoux et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3394–3398
were saturated with PBS without Ca and Mg. Using sterile forceps, one filter
was placed on the CAM and then the window was sealed with transparent
tape. After 1 hr, 10 L of inhibitor were applied topically to the L-T4 or FGF-
the start of treatment. Matrigel (BD Biosciences, San Jose, CA) was thawed
overnight at 40 °C and placed on ice. Aliquots of matrigel were placed into cold
polypropylene tubes to prevent the matrigel from solidifying, and compound
was added to the matrigel with or without an agonist. Matrigel plugs were
l
stimulated CAM. After 48 h, CAM tissue directly beneath the filter disk was
harvested and placed in a 35 mm petri dish. CAM sections were examined
using an SV6 stereomicroscope (Carl Zeiss MicroImaging, Inc) at 50ꢃ
subcutaneously injected at three sites in each animal (total of 100
lL/animal).
At day 14 after plug implant, all animals were sacrificed in a CO2 chamber, and
matrigel plugs were collected. Plug hemoglobin content was analyzed from
three implants per mouse (n = 6 per group) to measure angiogenesis. For
hemoglobin measurement, matrigel plugs were carefully stripped of any
remaining peritoneum. The plugs were placed into a 0.5 mL tube of ddH2O and
homogenized for 5–10 min. The samples were subjected to centrifugation at
magnification. Digital images were captured using
a 3-CCD color video
camera system (Toshiba America, Inc, New York, NY). The images were
analyzed using Image-Pro Plus (Media Cybernetics, Inc). The number of blood
vessel branch points within a circular region superimposed on the area of the
filter disk was counted for each section (10 eggs per condition).
20. Mouse matrigel model of angiogenesis: Normal male mice (C57BL/6NCr) 6–
8 weeks of age weighing approximately 20 g were purchased from Taconic
Farms, Inc. Animals were housed four per cage under controlled environmental
4000 rpm for 10 min, and then 50
50 L of Drabkin’s reagent and allowed to sit at room temperature for 15–
30 min, at which point 100 L of the mixture were placed in a 96-well plate.
lL of the supernatant were mixed with
l
l
conditions (20–24 °C, 60–70% humidity and
a
12 h light/dark cycle) and
Absorbance was measured with a Microplate Manager ELISA reader at 540 nm.
Hemoglobin concentration was determined by comparison with a standard
curve in mg/mL.
provided with food and water ad libitum. All experimental protocols were
approved by the Institutional Animal Care and Use Committee of the VA
Hospital of Albany, NY, USA. Mice were allowed to acclimate for 5 days prior to