Modular synthesis of radicicol A and related resorcylic acid lactones, potent kinase inhibitors

Article abstract of DOI:10.1002/anie.200702406

(Chemical Equation Presented) Short and sweet: A concise and modular synthesis of radicicol A and related resorcylic acid lactones using fluorous isolation technology and immobilized reagents is reported (see scheme, R ^F = C₃H_6<

Full text of DOI:10.1002/anie.200702406

Angewandte
Chemie
DOI: 10.1002/anie.200702406
Total Synthesis
Modular Synthesis of Radicicol A and Related Resorcylic Acid
Lactones, Potent Kinase Inhibitors**
Pierre-Yves Dakas, Sofia Barluenga, Frank Totzke, Ute Zirrgiebel, and Nicolas Winssinger*
Radicicol A (F87-2509.04, 1, Scheme 1) belongs to the family
target was not identified.[2,3] Subsequently, two other related
resorcylic acid lactones containing a cis-enone (3 and 5,
Scheme 1) were reported to be potent irreversible yet
selective kinase inhibitors.[4,5] More recently, Santi and co-
workers showed that hypothemycin, which also bears the cis-
enone moiety, covalently inactivates ERK2 by reacting with a
cysteine residue positioned in the active site (Cys166).[6]
Our interest in RALs stems from the observation that a
significant fraction of this family of natural products has been
shown to inhibit kinases and ATPases (ATP = adenosine 5’-
triphosphate). Despite the lack ofobvious similarities
between the RALs and ATP, these compounds have been
shown to bind to the ATP-binding pocket ofATPases. We
initiated a program to extend the diversity ofRALs and to
capitalize on this privileged structure.[1,7–9] A cornerstone of
these efforts is the development of chemistry that is compat-
ible with high-throughput synthesis. The disconnections that
were envisioned for radicicol A and related cis-enone RALs
are summarized in Scheme 2. A significant challenge is to
ofresorcylic acid lactones [1] (RAL) and was first reported by
Scheme 1. Structure of radicicol A and related resorcylides.
researchers from Sandoz who identified this fungal metabo-
lite from a screen for IL1b inhibition.[2] While it was observed
that 1 accelerated the degradation ofspeciicf mRNA
sequences including those ofIL1 b, its precise molecular
[*] P.-Y. Dakas, Dr. S. Barluenga, Prof. N. Winssinger
Institut de Science et IngØnierie SupramolØculaires
UniversitØ Louis Pasteur—CNRS UMR 7006
8 allØe Gaspard Monge, 67000 Strasbourg (France)
Fax: (+33)3-9024-5112
Scheme 2. Retrosynthetic disconnection of radicicol A. PG=protecting
group.
control the cis geometry ofthe enone as it can isomerizes to
the thermodynamically more favorable trans isomer as in
compound 4, which is known to be significantly less active.[5] It
was thus anticipated that the enone should be revealed only at
a late stage through a selective allylic oxidation and that the
cis-alkene would originate from a vinyl lithium addition onto
an aldehyde. As shown in Scheme 2, the molecule could thus
be disconnected into three fragments (7, 8, and 9) ofeven
E-mail: winssinger@isis.u-strasbg.fr
Homepage: www-isis.u-strasbg.fr/winssinger/
F. Totzke, U. Zirrgiebel
ProQinase GmbH
Breisacher Strasse 117, Freiburg (Germany)
Homepage: www.proqinase.com
[**] This project was partly supported by grants from the Agence
National de la Recherche (ANR) and Human Frontier Science
Program (HFSP). An MENRT fellowship (P.Y.D.) is also gratefully
acknowledged. The authors thank Kerstin Knepper for preliminary
work on this project.
complexity, in which the aldehyde or iodide ofrfagment
8
would need to be masked depending on the order ofassembly.
While the order ofcoupling ofthese three rfagments is
conceptually possible in all permutations, we reasoned that
starting the sequence with the coupling of 8 + 9 would allow
the use of a fluorinated protecting group for the alcohol 9,
Supporting information for this article, including experimental
procedures, physical characterization for the synthesis of 1, 3, and
31 as well as selectivity profile of compound 31 against a panel of
127 kinases, is available on the WWW under http://www.
angewandte.org or from the author.
[10,11]
thus enabling the use oflfuorous isolation technology
Angew. Chem. Int. Ed. 2007, 46, 6899 –6902
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6899

Communications
until the penultimate cyclization. Two features make this
strategy particularly attractive. First and foremost, it has been
shown that multiple components tagged with fluorinated tags
of different lengths can be carried through a synthesis as a
mixture and ultimately resolved using fluorous chromatog-
raphy.[12,13] Second, this technology has now been automated,
making it amenable to high-throughput synthesis.[14]
The aromatic fragment 7 was prepared with a silyl-based
protecting group (TMSE, 2-(trimethylsilyl)ethyl) for the
carboxylic acid. As shown in Scheme 3, this synthesis was
Scheme 4. Synthesis of key intermediate 17. a) LiAlH4 (1.4 equiv),
THF, from 08C to 238C, 2 h, 95%; b) TBDPSCl (1.0 equiv), imid.
(1.5 equiv), DMF, 238C, 2 h, 66%; c) PS-IBX (3.0 equiv), CH2Cl2, 238C,
2 h, 100%, DMF=dimethyl formamide, IBX=2-iodoxybenzoic acid,
TBDPS=tert-butyldiphenylsilyl.
ofPMB trichloroacetimidate 19.[12,17] Cross-metathesis with
the vinyl borolane 20 afforded the trans-vinyl borolane[18] 21
in excellent yield and stereoselectivity (greater than 20:1 E/Z)
using the second-generation Grubbs catalyst or even faster
and with equally good stereoselectivity using the Grela
modified catalyst.[19] Importantly, the trans-vinylborolane 21
could be stereospecifically converted to the cis-vinyl bromide
Scheme 3. Synthesis of aromatic moieties 13–15. a) (COCl)2,
(1.0 equiv), DMF (cat.), CH2Cl2, 08C, 1 h, then 2-(trimethylsilyl)ethanol
(1.0 equiv), Et3N (2.6 equiv), 4-DMAP (cat.), 238C, 1 h, 96–98%;
b) LDA (2.0 equiv), (PhSe)2 (1.0 equiv), THF, À788C, 1 h, 89–91%, 4-
DMAP=4-dimethylamino pyridine, LDA=lithium diisopropylamide,
TMS=trimethylsilyl.
[20]
22 in excellent yield using the protocol ofBrown et al.
Transmetalation of 22 with tBuLi and addition onto the crude
aldehyde 17 afforded, after EOM protection, product 23 as a
mixture ofdiastereoisomers (3:1). This lack ofselectivity is
inconsequential, as the stereogenic center in question will
ultimately be oxidized to the ketone. Conversion ofthe silyl-
protected hydroxyl group to the iodide (TBAF; Ph3P, I2)
afforded compound 24, which was alkylated in excellent yield
with the three different aromatic fragments 13–15 previously
deprotonated by LDA. The selenide was then oxidized and
eliminated to afford compounds 25–27. This reaction
sequence did not require a single traditional workup. The
crude reaction mixtures were simply loaded on fluorous-silica
columns and eluted first with 75% MeOH in water to remove
all non-fluorous-tagged components and then washed with
MeOH to recover the desired compounds. All the compounds
(21–27) came at the solvent front with the MeOH wash and
could be recovered in excellent yields. Reactions that were
performed under basic conditions (22 to 23 and alkylation
with 24) were quenched with benzoic acid resin prior to
loading on the column. The PMB group was removed under
the action ofDDQ, and again, the byproducts ofthese
reactions were easily sequestered using fluorous solid-phase
extraction, while the TMSE ester was remove with TBAF.
The hydroxyacids thus obtained were engaged in Mitsunobu
macrocyclisations using fluorous-tagged triphenyl phosphine
and diazo dicaboxylate, thus yielding the macrocycles 28–30
after a fluorous solid-phase extraction. The EOM and
acetonide groups could be quantitatively deprotected using
sulfonic acid resin in MeOH, while the ortho-phenol could be
selectively cleaved in the presence ofthe other methyl ethers
using BCl3, or more conveniently, all operations could be
achieved in one step with BCl3. The allylic alcohols were then
selectively oxidized with the polymer-bound version ofIBX
to afford radicicol A (1),[21] 5-(Z)-7-oxozeaenol (3),[22] and
radicicol analogue 31 in nearly quantitative yield following a
simple filtration.
achieved in two steps from the different acids 10–12 by
esterification with 2-(trimethylsilyl)ethanol and subsequent
formation of the selenide via LDA deprotonation of the
benzylic position and reaction with diphenyldiselenide. This
procedure was found to be effective for the trimethoxy-
substituted aryl ring corresponding to radicicol A (10) as well
as for the dimethoxy aryl ring corresponding to 5-(Z)-7-
oxozeaenol (11) and the chlorodimethoxy substitution pres-
ent in radicicol (12). While several protecting groups for the
ortho-phenol (MOM (methoxymethyl), EOM (ethoxy-
methyl), PMB (para-methoxybenzyl)) were found to be
suitable in the case ofthe 2,4-dimethoxy substitution, these
protecting groups were not stable for the 2,4,5-trimethoxy-
subtituted aryl ring. The presence ofthe 5-methoxy substitu-
ent renders any acid-labile protecting group at the 2-position
particularly susceptible to acids and, more generally, makes
this aryl moiety prone to oxidation. Nevertheless, it was found
that a methyl ether could be cleaved very selectively in the
presence ofthe other methyl ethers using boron trichloride
for all three aromatic systems.
Intermediate 8, with the iodide masked in the form of a
silyl-protected hydroxyl group, was conveniently obtained
from acetal-protected deoxyribose[15] in three steps
(Scheme 4). Protected deoxyribose (obtained in one step
from 2-deoxy-d-ribose) was reduced using LiAlH4, and the
crude reaction product was selectively silylated on the
sterically less encumbered alcohol using TBDPSCl (greater
than 15:1 selectivity) in the presence ofimidazole in DMF.
The free alcohol was then oxidized using an immobilized
version ofIBX, [16] thus affording the aldehyde, which could be
used directly in the subsequent reaction without workup or
further purification.
Radicicol A and the analogues bearing different substitu-
ents on the aromatic ring as well as their corresponding allylic
alcohols were tested in vitro against a panel of24 kinases to
assess their activity and specificity. In agreement with the
The synthesis offragment 9 began with the protection of
(R)-2-hydroxypentene (18, Scheme 5) with a fluorous version
6900 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 6899 –6902

Angewandte
Chemie
corresponding quinones. We have also
noted that compounds related to 3 could
be oxidized to the quinone, albeit with
stronger oxidizing agents, such as mCPBA
(m-chloroperbenzoic acid). A number of
RALs, such as pochonins C and D, are
substituted with a chlorine atom at the 5-
position rather than a methoxy group.
Inspired by these analogues, we tested
compound 31 in the same panel ofkinases.
While it was overall slightly less active than
1 and 3, it remained a potent inhibitor of
therapeutically
important
kinases
(Table 1). The activity ofthis compound
was further evaluated in a panel of 127
kinases, which showed that the only kinases
inhibited at nanomolar concentrations
were VEGF-R2 and 3, PDGFR-a and b,
as well as MEK1 (see the Supporting
Information), all of which have a cysteine
residue in the active site. Kinases inhibited
at low micromolar concentrations included
FLT3, KIT, and VEGF-R1, which also bear
a cysteine residue in the active site. The
discrimination by 31 among the subset of
kinases bearing an appropriately positioned
cysteine residue indicates that unique inter-
actions with its target are necessary for the
Michael addition to proceed and that its
selectivity could be honed or tuned to other
kinases. The potential efficacy of this com-
pound in vivo was then assessed by meas-
uring the level ofVEGF-R2 autophosphor-
ylation in the presence ofits ligand
(VEGF165). Immortalized human umbilical
vein endothelial cells (HUVECs) known to
express high levels ofVEGF-R2 were
incubated with inhibitor 31 for 90 min and
then stimulated with VEGF165 for 7 min.
Scheme 5. Synthesis of radicicol A (1), 5-(Z)-7-oxozeaenol (3), and radicicol A analogue 31.
a) 19 (1.0 equiv), CSA (cat.), CH2Cl2, 238C, 12 h, 92%; b) 20 (2.0 equiv), Grubbs II
(2.5 mol%); toluene, 808C, 12 h,92%; c) Br2 (1.0 equiv, 1m in CH2Cl2), Et2O, À208C,
10 min, then NaOMe (2.2 equiv, 1m in MeOH), À208C, 30 min, 89%; d) tBuLi (2.0 equiv),
THF/Et2O, À1008C, 15 min, then 17 (1.0 equiv), À1008C, 15 min, 88%; e) EOMCl
(8.0 equiv), iPr2EtN (8.0 equiv), TBAI (cat.), DMF, 238C, 12 h, 96%; f) TBAF (2.5 equiv),
THF, 238C, 12 h, 92%; g) PPh3 (1.5 equiv), I2 (1.5 equiv), imidazole (2.5 equiv), THF, 08C,
1 h, 91%; h) 13–15 (1.0 equiv), LDA (2.0 equiv), THF/HMPA 10:1, À788C, 20 min, 88–
91%, i) H2O2 (2.0 equiv), THF, 238C, 2 h, 79–82%, j) DDQ (1.2 equiv), CH2Cl2/H2O 2:1,
238C, 2 h, 70–80%; k) TBAF (3.0 equiv), THF, 238C, 2 h, 87%; l) RFPh3P (2.0 equiv),
RFDEAD (2.0 equiv), toluene (10 mm), 238C, 2 h, 81%; m) BCl3 (3.0 equiv), CH2Cl2, 08C,
15 min, 86%; PS-IBX (3.0 equiv), CH2Cl2, 238C, 1 h, greater than 90%. CSA=camphor-10-
sulfonic acid, DEAD=ethoxycarbonylazocarboxymethyl, DDQ=dichlorodicyanoquinone,
Grubbs II=ruthenium[1,3-bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro(phenyl-
methylene)(tricyclohexylphosphine), HMPA=hexamethylphosphoramide, LDA=lithium di-
isopropylamide, RF =C3H6C6F13, TBAF=tetrabutylammonium fluoride, TBAI=tetrabutylam-
monium iodide, TBDPS=tert-butyldiphenylsilyl.
proposal that the enone moiety acts as a Michael acceptor,
none ofthe products lacking the cis-enone showed significant
activity (data not shown). Radicicol A, on the other hand,
exhibited potent inhibitory activity (low nanomolar range)
against VEGF-R2, VEGF-R3, FLT3, and PDGFR-b
(Table 1). From a therapeutic perspective, radicicol Aꢀs 2-
hydroxy-5-methoxy aryl substitution may present a liability,
as such electron-rich aryl moieties may be oxidized to the
The level ofautophosphorylation was measured by ELISA
using anti-VEGF-R2 as capture antibody and anti-phospho-
tyrosine as detection antibody. Results are expressed as a
percentage ofmaximal autophosphorylation in the absence of
inhibitor (Figure 1). Compound 31 was found to have a
cellular IC50 of440 n m, which is consistent with its inhibition
at the enzymatic level (90 nm). Consistently, PDGFR-b
autophosphorylation was mildly inhibited by 31 (IC50 =
Table 1: IC50 proflile (nm) against a panel of 24 therapeutically relevant kinases.[a]
[a] Empty fields: IC50 >100000 nm.
Angew. Chem. Int. Ed. 2007, 46, 6899 –6902
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 6901

Communications
oxozeaneol (also known as LL-Z1640-2), see: G. A. Ellestad,
F. M. Lovell, N. A. Perkinson, R. T. Hargreaves, W. J. McGah-
ren, J. Org. Chem. 1978, 43, 2339.
[5] A. Zhao, S. H. Lee, M. Mojena, R. G. Jenkins, D. R. Patrick,
H. E. Huber, M. A. Goetz, O. D. Hensens, D. L. Zink, D. Vilella,
A. W. Dombrowski, R. B. Lingham, L. Huang, J. Antibiot. 1999,
52, 1086.
[6] A. Schirmer, J. Kennedy, S. Murli, R. Reid, D. V. Santi, Proc.
Natl. Acad. Sci. USA 2006, 103, 4234.
[7] S. Barluenga, P. Y. Dakas, Y. Ferandin, L. Meijer, N. Winssinger,
Angew. Chem. 2006, 118, 4055; Angew. Chem. Int. Ed. 2006, 45,
3951; .
[8] E. Moulin, S. Barluenga, F. Totzke, N. Winssinger, Chem. Eur. J.
2006, 12, 8819.
Figure 1. Cellular inhibition of VEGF-R2 (left) and PDGFR-b (right)
autophosphorylation upon stimulation with their cognate ligand after
preincubation with compound 31 (the y axis represents the percentage
of phosphorylation compared to maximal phosphorylation).
[9] E. Moulin, V. Zoete, S. Barluenga, M. Karplus, N. Winssinger, J.
Am. Chem. Soc. 2005, 127, 6999.
[10] D. P. Curran, Handbook of Fluorous Chemistry, Wiley-VCH,
Weinheim, 2004, p. 101.
[11] W. Zhang, D. P. Curran, Tetrahedron 2006, 62, 11837.
[12] S. Dandapani, M. Jeske, D. P. Curran, Proc. Natl. Acad. Sci. USA
2004, 101, 12008.
[13] Z. Luo, Q. Zhang, Y. Oderaotoshi, D. P. Curran, Science 2001,
291, 1766.
13 mm), whereas autophosphorylation ofTIE2 was una-f
fected. The demonstration that radicicol A inhibits key
kinases involved in MAPK cascades can rationalize the
original observation that it accelerates the degradation of
Il1bꢀs RNA, as there is mounting evidence that MAP kinases
are involved in the regulation ofprotein translation by
[23,24]
controlling the stability ofspecific mRNAs.
[14] W. Zhang, Y. Lu, J. Comb. Chem. 2006, 8, 890.
[15] J. Barbat, J. Gelas, D. Horton, Carbohydr. Res. 1983, 116, 312.
[16] G. Sorg, A. Mengei, G. Jung, J. Rademann, Angew. Chem. 2001,
113, 4532; Angew. Chem. Int. Ed. 2001, 40, 4395.
[17] Q. Zhang, H. Lu, C. Richard, D. P. Curran, J. Am. Chem. Soc.
2004, 126, 36.
[18] C. Morrill, R. H. Grubbs, J. Org. Chem. 2003, 68, 6031.
[19] A. Michrowska, R. Bujok, S. Harutyunyan, V. Sashuk, G.
Dolgonos, K. Grela, J. Am. Chem. Soc. 2004, 126, 9318.
[20] H. C. Brown, C. Subrahmanyam, T. Hamaoka, N. Ravindran,
D. H. Bowman, S. Misumi, M. K. Unni, V. Somayaji, N. G. Bhat,
J. Org. Chem. 1989, 54, 6068.
[21] The synthetic product was found to be identical to a natural
sample kindly provide by Dr. Traber (Novartis).
[22] For previous syntheses, see: K. Tatsuta, S. Takano, T. Sato, S.
Nakano, Chem. Lett. 2001, 172; P. Selles, R. Lett, Tetrahedron
Lett. 2002, 43, 4627.
[23] M. Gaestel, Nat. Rev. Mol. Cell Biol. 2006, 7, 120.
[24] D. V. Henley, C. J. Bellone, D. A. Williams, M. F. Ruh, Arch.
Biochem. Biophys. 2004, 424, 112.
In conclusion, we have developed a concise synthesis of
radicicol A and related resorcylic acid lactones using a
combination oflfuorous tag-isolation and polymer-bound
reagents, thus making the chemistry amenable to high-
throughput synthesis. In addition to the therapeutic potential
ofthis afmily ofnatural products, the irreversible nature of
these inhibitors also offers a new lead structure for the design
ofselective inhibitors towards engineered kinases [25,26] as well
as for chemical proteomics.[27]
Received: June 3, 2007
Published online: August 3, 2007
Keywords: inhibitors · resorcylic acid lactones ·
.
solid-phase extraction · total synthesis
[1] N. Winssinger, S. Barluenga, Chem. Commun. 2007, 22.
[2] T. Kastelic, J. Schnyder, A. Leutwiler, R. Traber, B. Streit, H.
Niggli, A. Mackenzie, D. Cheneval, Cytokine 1996, 8, 751.
[3] D. Cheneval, P. Ramage, T. Kastelic, T. Szelestenyi, H. Niggli, R.
Hemmig, M. Bachmann, A. Mackenzie, J. Biol. Chem. 1998, 273,
17846.
[25] J. A. Blair, D. Rauh, C. Kung, C. H. Yun, Q. W. Fan, H. Rode, C.
Zhang, M. J. Eck, W. A. Weiss, K. M. Shokat, Nat. Chem. Biol.
2007, 3, 229.
[26] M. S. Cohen, C. Zhang, K. M. Shokat, J. Taunton, Science 2005,
308, 1318.
[27] C. Drahl, B. F. Cravatt, E. J. Sorensen, Angew. Chem. 2005, 117,
5936; Angew. Chem. Int. Ed. 2005, 44, 5788.
[4] J. Ninomiya-Tsuji, T. Kajino, K. Ono, T. Ohtomo, M. Matsu-
moto, M. Shiina, M. Mihara, M. Tsuchiya, K. Matsumoto, J. Biol.
Chem. 2003, 278, 18485; for the original isolation of 5-(Z)-7-
6902 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 6899 –6902