Identification of a chromone-based retinoid containing a polyolefinic side chain via facile synthetic routes

Article abstract of DOI:10.1016/j.bmcl.2009.05.081

Attempts to prepare substituted chromones as novel retinoids revealed that some chromones were unstable under Wadsworth-Horner-Emmons reaction conditions. Hence, Wittig reactions were used to prepare chromone-based compounds as potential retinoids. Firstly, Wittig reagents prepared from 3-bromomethyl-chromen-4-one were reacted with olefinic-aldehydes to provide the target compounds with all-trans side chains in good yield. The approach supplies a useful general route to structurally diverse chromone-based compounds possessing a variety of side chains. Sequential Wittig reactions were used also to prepare a chromone-based retinoid. These novel compounds were evaluated in binding assays and a high affinity RAR ligand was identified. Crystal structures obtained for two key precursors aided the interpretation of binding data.

Full text of DOI:10.1016/j.bmcl.2009.05.081

Bioorganic & Medicinal Chemistry Letters 19 (2009) 4339–4342
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of a chromone-based retinoid containing a polyolefinic
side chain via facile synthetic routes
Weilin Sun a, Patrick J. Carroll c, Dianne R. Soprano b, Daniel J. Canney a,
*
a Department of Pharmaceutical Sciences, School of Pharmacy, Temple University, 3307 N. Broad Street, Philadelphia, PA 19140, USA
b Department of Biochemistry, School of Medicine, Temple University, 3307 N. Broad Street, Philadelphia, PA 19140, USA
c Department of Chemistry, University of Pennsylvania, 231 S. 34th Street, Philadelphia, PA 19104, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Attempts to prepare substituted chromones as novel retinoids revealed that some chromones were
unstable under Wadsworth–Horner–Emmons reaction conditions. Hence, Wittig reactions were used
to prepare chromone-based compounds as potential retinoids. Firstly, Wittig reagents prepared from
3-bromomethyl-chromen-4-one were reacted with olefinic-aldehydes to provide the target compounds
with all-trans side chains in good yield. The approach supplies a useful general route to structurally
diverse chromone-based compounds possessing a variety of side chains. Sequential Wittig reactions were
used also to prepare a chromone-based retinoid. These novel compounds were evaluated in binding
assays and a high affinity RAR ligand was identified. Crystal structures obtained for two key precursors
aided the interpretation of binding data.
Received 3 April 2009
Revised 16 May 2009
Accepted 20 May 2009
Available online 27 May 2009
Key words:
Substituted chromones
Retinoids
Polyolefinic chains
Wittig reactions
Rearrangements
Ó 2009 Elsevier Ltd. All rights reserved.
Vitamin A (retinol) and its biologically active derivatives, collec-
tively called retinoids, are important in the regulation of a wide
range of physiological and developmental processes. All-trans ret-
inoic acid (ATRA) and 9-cis retinoic acid (9-cis RA) (Fig. 1) are
endogenous ligands for the two families of retinoid nuclear recep-
tors: retinoic acid receptors (RARs) and retinoic X receptors
(RXRs).1 Each family of receptors is further divided into alpha
noids have been reported to bind and activate AhRs also.6,7 Hence,
the chromone nucleus was used to design synthetic retinoids
possessing polyolefinic side chains that mimic the endogenous
RAR ligand ATRA. Wittig and Wadsworth–Horner–Emmons
(WHE) reaction conditions were seen as efficient routes to the
target chromones. Initial attempts to prepare a retinoids by react-
ing commercially available 3-formylchromone or 6-i-propyl-3-for-
mylchromone with triethyl 3-methyl-4-phosphonocrotonate
under WHE reaction conditions led to a rearrangement/cyclization.
The reactions products were benzophenones as shown in Scheme
1. The structures of the rearranged products were confirmed by
NMR and X-ray crystal analysis.8 Interestingly, no rearrangement
was observed when 3-formylchromone was reacted with tri-
ethylphosphonoacetate (as seen in our hands and reported
previously9).
(a), beta (b), and gamma subtypes (c). Numerous synthetic reti-
noids have been developed as pharmacological tools to dissect
the complex molecular mechanism of RARs, as well as therapeutic
agents for the treatment of cancer and hyperproliferative dis-
eases.2–4 A major concern regarding the use of currently available
nonselective retinoids is an adverse effect profile that includes
respiratory distress, fever, pulmonary edema, and pulmonary infil-
trates.5 In order to decrease the adverse effects of these agents and
to provide improved pharmacological tools, ligands exhibiting
more selectivity for retinoid receptor subtypes are required.2
Our early efforts in the development of subtype selective syn-
thetic retinoids involved heterocyclic compounds that proved too
bulky to bind to the target receptors. A search for more compact,
planar nuclei for ligand design led to the chromone system as a po-
tential scaffolding. Chromone-related flavonoids show activity at
the nuclear aryl hydrocarbon receptor (AhR) and synthetic reti-
9
COOH
ATRA
9-cis RA
COOH
* Corresponding author. Tel.: +1 215 707 6924; fax: +1 215 707 5620.
E-mail address: daniel.canney@temple.edu (D.J. Canney).
Figure 1. Structures of ATRA and 9-cis RA.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.05.081

4340
W. Sun et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4339–4342
O
O
O
nBuLi
H
R
CHO
R
CHO
H
P(O)(OEt)2
O
EtOOC
P(O)(OEt)2
EtOOC
OH
R
O
O
O
H
COOEt
R
COOEt
R = H, i-propyl
Scheme 1. Proposed mechanism for the formation of benzophenones via a rearrangement/cyclization.
In order to circumvent the rearrangement that leads to un-
tively (Scheme 2). The route to acid 6 involved the preparation of
the aldehydic ester 4 (Scheme 2). Dimethyl dioxirane (DMD) was
prepared by oxidizing acetone with Oxone in aqueous sodium
bicarbonate solution.10 Furan was oxidized in DMD solution to
form the malealdehyde intermediate which reacted with methyl
(triphenylphosphoranylidene) acetate to give 4 in the all-trans
conformation.11 The aldehyde 4 thus formed, was reacted with
ylide 3 to provide the corresponding ester 5 as the mixture of trans
and cis isomers in almost equal quantities which could not be sep-
arated by column chromatography. The trans isomer of 5 was iso-
lated from the mixture through recrystalization. The ester was
hydrolyzed to the corresponding acid 6 using LiOH. When com-
pound 6 did not bind to the target RAR receptors, an isopropyl
group was incorporated on the chromone nucleus in order to
explore the effect of aliphatic groups on binding. As shown in
Scheme 2, phosphorous ylide 9 was reacted with commercially
available fumaraldehyde mono(dimethyl) acetal to afford aldehyde
10 in situ. The all trans stereochemistry of 10 was verified using
wanted benzophenones, alternative routes to the target retinoids
were devised. Firstly, a generally applicable approach was devel-
oped that involved the preparation of chromone-based Wittig
reagents. Hence, 3-formylchromones were converted first to phos-
phorous ylides as described in Scheme 2. The ylide thus formed
was then reacted with the appropriate aldehyde to afford the chro-
mone-based precursor compounds (Scheme 2).
The phosphorous ylides 3 and 9 were prepared from the chro-
mone-aldehydes by first reducing the aldehydes to the alcohols
in moderate yield using 9-BBN. The alcohols were treated with
tetrabromomethane in the presence of PPh3 to afford the bromo-
methyl analogs 2 and 8. The bromo analogs were reacted with
PPh3 to provide the corresponding phosphorous ylides (3 and 9)
in good yield (Scheme 2). Ylides thus formed provide useful inter-
mediates en route to a variety of substituted chromones. Reaction
of ylides 3 and 9 with commercially available 4,4-dimethoxy-but-
2-enal or aldehydic ester 4 provided chromones 6 and 11, respec-
O
O
O
O
R
CHO
9-BBN
R
CH2OH
R
CBr4, Ph3P
Br
O
O
1 R = H, 60%
7 R = iPr, 54%
2 R = H, 87%
8 R = iPr, 82%
Br
O
O
PPh3
R
PPh3
3. R = H, 83% 9 R = iPr, 80%
R = iPr
R = H
COOMe
nBuLi
OMe
1) DMD
OHC
OMe
OHC
O
4
nBuLi
2) Ph3P
O
COOMe
O
O
COOR
R
5 R = Me,74%
6 R = H, 30%
LiOH
O
10 R = CHO, 48%
11 R = COOH, 63%
HClO2
Scheme 2. Facile synthetic routes to chromone-based ligands 6 and 11.

W. Sun et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4339–4342
4341
results obtained in binding studies of compounds 6 and 11 led to
a re-examination of the interaction between ATRA and retinoic
acid receptors. The co-crystal structure of ATRA in RAR
c
(PDB:2lbd) reveals that the b-ionone ring of ATRA is sandwiched
in a hydrophobic environment. Its dimethyl groups are reported
to be involved in hydrophobic interactions with lipophilic residues
of the protein (Fig. 2).13
Hence, the lack of affinity observed for 6 could be attributed, at
least in part, to the need for properly positioned aliphatic groups
on the aromatic ring. The isopropyl group of 11 was incorporated
to mimic the methyl groups on the b-ionone ring of ATRA but
did not lead to receptor affinity. In order to better understand these
observations, the crystal structure of 10 (precursor of target acid
11) was determined and is shown in Figure 3. The s-trans con-
former (two double bonds in red) in 10, rather than the s-cis found
in ATRA helps to explain the lack of affinity observed for 11.
Efforts to prepare ligands with s-cis geometry and appropriately
positioned aliphatic groups led to the design and synthesis of acid
15 (Scheme 3). Aldehyde 128 was prepared following a literature
procedure and reacted with 2-(triphenylphosphoranylidene)-pro-
pionaldehyde to form intermediate 13. Another Wittig reaction
involving 13 and (carbethoxyethylidene)triphenyl-phosphorane
provided 14 which was hydrolyzed under acidic conditions to fur-
nish acid 15. Compound 15 contains a tetramethyl alicyclic system
and an olefinic side chain with two methyl groups.
Figure 2. ATRA in the surface of RAR
Hydrophobic areas of the surface are tainted blue, hydrophilic areas are red and
neutral portions are white.
c binding pocket (adapted from 2lbd.pdb).
The crystal structure of precursor 14 reveals that this com-
pound has the s-cis conformation (Fig. 4; two double bonds in
red) found in ATRA. The affinity of acid 15 for RARs was evaluated
in preliminary radioligand binding assays using previously
described methods14 and found to have IC50 values of 15 nM,
40 nM, and 40 nM for RARa, RARb and RARc, respectively. These
data suggest that the chromone nucleus provides a useful scaffold
for RAR ligand design. Chromone-based retinoid 15 has been iden-
tified as a lead molecule for further ligand development.
In summary, initial efforts to prepare chromone-based retinoids
with polyolefinic side chains involved the use of formylchromones
and triethyl 3-methyl-4-phosphonocrotonate under WHE reaction
conditions. These efforts consistently led to benzophenones fol-
lowing a rearrangement and cyclization. In order to avoid this rear-
rangement, alternative routes to the target ligands were devised.
The first route involved conversion of 3-formylchromones to the
corresponding phosphorous ylides in three steps. The Wittig
reagents thus formed were reacted with olefinic aldehydes to yield
the target chromone-based products with the trans conforma-
tion.15 The approach supplies a useful general route to structurally
diverse chromone-based compounds possessing a variety of side
chains. A second route involving sequential Wittig reactions using
Figure 3. Ortep drawing of aldehyde 10.
NMR and X-ray crystal data. Oxidation of this unsaturated alde-
hyde using sodium hypochlorite under mild acidic conditions
yielded 11.12 Binding studies performed on acid 11 revealed that
the compound had no affinity for RAR receptors. The negative
O
O
O
CHO
PPh3
COOEt
CHO
CHO
PPh3
O
13
72%
12
O
O
COOEt
COOH
H2SO4/HOAc
O
O
15
14
50%
27%
Scheme 3. Facile synthetic route to retinoid 15.

4342
W. Sun et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4339–4342
6. Lu, Y-F.; Santostefano, M.; Cunningham, B. D. M.; Threadgill, M. D.; Safe, S.
Biochem. Pharmacol. 1996, 51, 1077.
7. Gambone, C. J.; Hutcheson, J. M.; Gabriel, J. L.; Beard, R. L.; Chandraratna, R. A.
S.; Soprano, K. J.; Soprano, D. R. Mol. Pharmacol. 2002, 61, 334.
8. Sun, W.; Desai, S.; Piao, H.; Carroll, P.; Canney, D. J. Heterocycles 2007, 71, 557.
9. Bodwell, G. J.; Hawco, K. M.; da Silva, R. P. Synlett 2003, 2, 179.
10. Adam, W.; Bialas, J.; Hadjiarapoglou, L. Chem. Ber. 1991, 124, 2377.
11. Adger, B. J.; Barrett, C.; Brennan, J.; McGuigan, P.; McKervey, M. A.; Tarbit, B. J.
Chem. Soc., Chem. Commun. 1993, 1220.
12. Dalcanale, E.; Montanari, F. J. Org. Chem. 1986, 51, 567.
13. Renaud, J. P.; Rochel, N.; Ruff, M.; Vivat, V.; Chambon, P.; Gronemeyer, H.;
Moras, D. Nature 1995, 681.
14. Zhang, Z-P.; Shukri, M.; Bambone, C. J.; Gabriel, J. L.; Soprano, K. J.; Soprano, D.
R. Arch. Biochem. Biophys. 2000, 380, 339.
15. 7-(4-Oxo-4H-chromen-3-yl)-hepta-2,4,6-trienoic acid methyl ester (trans isomer)
(5). To a chilled (À78 °C) solution of compound 3 (0.5 g, 1 mmol) in dry THF
(15 mL) under N2 was added n-butyllithium (2 M in heptane, 0.75 mL,
1.5 mmol) with stirring. After 30 min, aldehyde 4 (0.12 g, 0.86 mmol) in dry
THF (5 mL) was added dropwise. After stirring at À78 °C for an additional
30 min, the resulting mixture was allowed to warm to room temperature and
stirred overnight. The reaction was quenched with saturated aqueous NH4Cl
and extracted with EtOAc (15 mL Â 3). The organic layer was washed with
brine, dried over MgSO4, concentrated and purified by flash chromatography
(silica gel; 3:1, hexane/EtOAc) to afford a mixture of trans and cis isomers (1:1
based on NMR) (0.18 g, 74%). A solution of trans and cis isomers in EtOAc and
hexane (15 mL; 5:1) was allowed to evaporate gradually at room temperature.
Compound 5 precipitated as pale yellow crystals: mp 115–117 °C; 1H NMR
(400 MHz, CDCl3): d 8.28 (d, J = 8.0 Hz, 1H), 8.04 (s, 1H), 7.68 (t, J = 8.0 Hz, 1H),
7.33–7.56 (m, 4H), 6.67 (dd, J = 14.8, 10.8 Hz, 1H), 6.55 (d, J = 15.4 Hz, 1H), 6.47
(dd, J = 14.8, 10.8 Hz, 1H), 5.93 (d, J = 15.4 Hz, 1H), 3.76 (s, 3H).
Figure 4. Ortep drawing of compound 14.
a tetramethylcyclohexyl-chromone-based aldehyde yielded an ole-
finic ester with s-cis geometry similar to ATRA.16,17 Crystal struc-
tures obtained for two key precursors aided the interpretation of
binding data.18 A novel chromone-based ligand was identified as
a lead for the development of future RAR. Molecular modification
of chromone-based acid 15 is underway in an effort to develop
subtype selective RAR ligands.19
16. (E)-2-methyl-3-(6,6,9,9-tetramethyl-4-oxo-6,7,8,9-tetrahydro-4H-benzo[g]chro-
men-3-yl)acrylaldehyde (13). An oven-dried flask was flushed with N2 and
charged with 12 (0.134 g, 0.47 mmol) and 2-(triphenylphosphoranylidene)
propionaldehyde (0.605 g, 1.88 mmol) in 10 mL of CH2Cl2. The resulting
mixture was stirred at room temperature under N2 for 48 h. The mixture was
condensed under reduced pressure and the residue was purified by flash
chromatography to yield 13 (0.11 g, 72%) 1H NMR (CDCl3, 400 MHz) d: 9.65 (s,
1H), 8.21 (s, 1H), 8.19 (s, 1H), 7.52 (s, 1H), 7.41 (s, 1H), 2.01 (s, 2H), 1.75 (s,
2H), 1.37 (s, 6H), 1.35 (s, 6H).
17. Attempts to transform aldehyde 13 to a phosphorous ylide (see synthesis of 3
and 9 above) failed. Under Wittig reaction conditions, no reaction occurred
between 3-formylchromones and 2-(triphenylphosphoranylidene)propional-
dehyde.
18. Crystallographic data (excluding structure factors) for compounds 10 and 14 in
this Letter have been deposited with the Cambridge Crystallographic Data
Centre as supplementary publication numbers CCDC 283693 (for 10) and
722736 (for 14). Copies of the data can be obtained, free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax: +44 (0)1223-
336033 or e-mail: deposit@ccdc.cam.ac.uk].
Acknowledgements
Dr. Jerome Gabriel is acknowledged for molecular modeling
data utilized in the design of ligands. The authors thank Shyam De-
sai for providing precursor 12. D.J.C. and W.S. are grateful to Wyeth
Research, the Pennsylvania Department of Health, and Temple Uni-
versity, School of Pharmacy for generous financial support. WS
thanks Dr. Jeffrey Pelletier, Wyeth Research, for valuable remarks
regarding the manuscript and Yanlong Kang, Sloan-Kettering Can-
cer Center, for assistance with Figure 2.
References and notes
1. Mark, M.; Ghyselinck, N. B.; Chambon, P. Annu. Rev. Pharmacol. Toxicol. 2006, 46,
451.
19. The hydrolysis of chromone-based esters (5 and 14) under alkaline or acidic
conditions was challenging due to the instability of the chromone nucleus (see:
Kona, J.; Fabian, W.M.F.; Zahradnik, P. J. Chem. Soc., Perkin Trans. 2 2001, 422).
No effort was made to optimize yields here but methods are available that
report higher yields (for example see; He, X.; Li, Z.; You, Q. Synth. Commun.
2002, 32, 709).
2. Germain, P.; Chambon, P.; Eichele, G.; Evans, R. M.; Lazar, M. A.; Leid, M.; De
Lera, A. R.; Lotan, R.; Mangelsdorf, D. J.; Gronemeyer, H. Pharmacol. Rev. 2006,
58, 712.
3. Aucci, L.; Gronemeyer, H. Nat. Rev. Cancer 2001, 1, 181.
4. Degos, L.; Wang, Z. Y. Oncogene 2001, 20, 7140.
5. Larson, R. S.; Tallman, M. S. Best Practice Res. Clin. Haematol. 2003, 16, 453.