In vivo efficacy of natural product-inspired irreversible kinase inhibitors

Article abstract of DOI:10.1002/cbic.201000205

Hypothemycin and related resorcylic acid lactones (RAL) bearing a cis-enone moiety have emerged as an alternative pharmacophore to heterocyclic motifs for kinase inhibition, and are endowed with a unique selectivity filter based on the irreversible reacti

Full text of DOI:10.1002/cbic.201000205

DOI: 10.1002/cbic.201000205
In Vivo Efficacy of Natural Product-Inspired Irreversible
Kinase Inhibitors
Sofia Barluenga, Rajamalleswaramma Jogireddy, Girish K. Koripelly, and
Nicolas Winssinger*[a]
Hypothemycin and related resorcylic acid lactones (RAL) bear-
ing a cis-enone moiety have emerged as an alternative phar-
macophore to heterocyclic motifs for kinase inhibition, and are
endowed with a unique selectivity filter based on the irreversi-
ble reaction with a subset of the kinome bearing a suitably
positioned cysteine residue. Two prototypical examples of
“edited” RAL were evaluated for antitumoral, antimetastatic
and antiangiogenic efficacy in an orthotopic murine renal cell
carcinoma (RENCA) model. Both compounds (3 and 5) are
good inhibitors of VEGFRs in vitro, and inhibited tumor growth
in vivo with comparable efficacy to sunitinib, an FDA-approved
VEGFRs inhibitor. Compound 3 promoted lung metastasis to a
similar extent as sunitinib, while compound 5 strongly inhibit-
ed lung metastasis. This study attests to the potential of irre-
versible kinase inhibitors and molecular editing of natural
pharmacophores and provides encouraging results to a clini-
cally significant problem.
Introduction
Many natural products capitalize on a covalent bond with their
biological target to achieve irreversible inhibition of its func-
tion.[1] This terminal inhibition bypasses a number of pharma-
cokinetics and pharmacodynamics issues. As the target is ter-
minally inactivated, the efficacy does not depend on maintain-
ing a constant, effective drug concentration and is not jeopar-
dized by competing endogenous interactions, as is the case if
the drug interaction is transient and dynamic.[2,3] Furthermore,
as target selectivity will be driven by the rate of inactivation,
which depends on the optimal alignment of an electrophile
and a nucleophile, irreversible inhibitors might be endowed
with superior selectivity if they can capitalize on unique nucle-
ophilic residues. This has been elegantly demonstrated by
Taunton et al., who judiciously positioned an electrophilic
group on an otherwise promiscuous scaffold targeting the nu-
cleotide binding site of kinases to achieve exclusive inhibition
of RSK1 and 2.[4] Nearly half of the kinome has a cysteine resi-
due in the vicinity of the nucleotide binding pocket, which
could be harnessed for irreversible kinase inhibition. Kinase in-
hibition has become one of the most intensely pursued thera-
peutic target classes with nine small-molecule inhibitors ap-
proved by the FDA over the past decade.[5] Hypothemycin, a
member of the resorcylic acid lactone (RAL) family,[6] was
shown to capitalize on a different cysteine residue to irreversi-
bly inactivate a subset of the kinome bearing the adequately
positioned cysteine residue.[7,8] Other resorcylic acid lactones
bearing this cis-enone moiety have been shown to be potent
inhibitors of the same subset of kinases.[9–11] Inspired by this
natural pharmacophore, we extended the natural diversity
with a library designed to explore areas tolerant to modifica-
tions.[12] In biochemical assays against a panel of ten putative
targets of cis-enone RAL, the vascular endothelial growth
factor receptors (VEGFRs) and platelet-derived growth factor
receptors (PDGFRs) were predominantly the most inhibited
kinases. A broader profile against 359 kinases[13,14] (excluding
mutants) revealed a pattern similar to that of sunitinib
(SU11248), which is generally considered a multitarget receptor
kinase inhibitor of VEGFRs and PDGFRs. Renal cell carcinoma is
characterized by high expression of proangiogenic growth fac-
tors (VEGF and PDGF), and sunitinib has demonstrated antitu-
mor activity in renal cell carcinoma.[15]
Results
Cell proliferation assay and kinase selectivity profile
Following our biochemical screening,[12] five cis-enone RAL ana-
logues containing modifications, were found to be tolerated or
beneficial in a biochemical assay; these analogues were tested
[a] Dr. S. Barluenga, Dr. R. Jogireddy, Dr. G. K. Koripelly, Prof. N. Winssinger
Institut de Science et Ingꢀnierie Supramolꢀculaires (ISIS–UMR 7006)
Universitꢀ de Strasbourg–CNRS
8 allꢀe Gaspard Monge, 67000 Strasbourg (France)
Fax: (+33)368855112
E-mail: winssinger@unistra.fr
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201000205.
1692
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1692 – 1699

In Vivo Efficacy of Irreversible Kinase Inhibitors
for their efficacy in inhibiting the growth of a renal carcinoma
cell line. As a bench mark, the natural product LL Z-1640-2 was
also included in this assay. As shown in Figure 1, compound 3,
which contains a benzylic ether, was as effective as the natural
typical examples of modified RALs for further investigations.
We next analyzed both compounds in a competition binding
assay[13,14] against a panel of 50 kinases, which contained 42
out of 48[7] of the kinases bearing the suitably positioned cys-
teine residue, as well as several
mutants of Flt3 and KIT (see
Figure 5 in the Discussion sec-
tion). The profile of both com-
pounds was consistent with our
previous profiles of cis-enone
RALs,[12] and showed comparable
inhibition for compounds 3 and
5 of VEGFRs, PDGFRs and down-
stream signaling kinases (MEKs).
Divergent synthesis of resorcyl-
ic acid lactones 3 and 5
Figure 1. Activity of different RALs in a RENCA cellular proliferation assay. The IC50 values were calculated from the
average of three experiments. The standard deviation at each concentration was between 0.5% and 16%. See fig-
ure S1 in the Supporting Information for graphs with error bars.
The gram-scale synthesis of com-
pounds 3 and 5 is shown in
Scheme 1. Addition of lithium
product, while the larger macrocycles were slightly less effec-
tive with the best (5) being nearly threefold less potent in
growth inhibition. Compounds 3 and 5 were selected as proto-
acetylide to propylene oxide followed by silylation of the
crude reaction product afforded compound 6. This was then
deprotonated (nBuLi) and treated with aldehyde 7, which was
Scheme 1. Divergent synthesis of 3 and 5.
ChemBioChem 2010, 11, 1692 – 1699
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1693

N. Winssinger et al.
readily obtained from d-2-deoxyribose in three steps (aceto-
nide protection,[16] olefination,[17] and oxidation). Benzoate pro-
tection of the product afforded the key common intermediate
8 in good overall yield (47% from propylene oxide). Ozonolysis
with a reductive workup afforded alcohol 9 of which the
alkene was reduced with Lindlar’s catalyst, and the alcohol was
converted to an iodide to obtain intermediate 10. Conversely,
the terminal alkene of compound 8 was hydroborated with an
oxidative work up to obtain alcohol 11, which was converted
to iodide 12 after a Lindlar reduction of the alkyne. Following
previously established chemistry, fragment 10 was coupled to
the aromatic moiety 13 using standard alkylation of the
phenol. Global silyl deprotection (TMSE and TBDPS) and Mitsu-
nobu cyclization afforded compound 14. Selective hydrolysis
of the benzoate, oxidation to reveal the cis-enone, and global
deprotection afforded macrocycle 3 in 30% yield from 13. The
larger macrocycle was prepared by alkylation of 12 with the
aryl moiety 15 bearing a benzylic selenide followed by oxida-
tion/elimination of the selenide.[11] By using the same proce-
dure as above with the exception of the final deprotection (HF
rather than sulfonic acid resin), compound 5 was obtained in
32% yield from 15 via 16.
In vivo efficacy and compound stability
We examined the effect of compounds 3 and 5 on the growth,
metastasis and angiogenic response of an orthotopic renal car-
cinoma (RENCA). Tumor growth was monitored noninvasively
by whole-body imaging during the course of treatment, and
tumor volumes and weights were measured at the end of the
study along with the extent of lung metastasis. A first-dose
study indicated systemic toxicity for both compounds 3 and 5
at high dose (75 mgkgÀ1), but good tolerability at 25 mgkgÀ1.
Sixty mice were divided into five groups and systemically treat-
ed with either vehicle (group 1), sunitinib (group 2, 40 mgkgÀ1
p.o., q1d), compound
3
(group 3, 20 mgkgÀ1 i.p., q1d;
group 4, 20 mgkgÀ1 i.p., q2d), or compound 5 (group 5,
20 mgkgÀ1 i.p., q1d). Animal weights decreased between day 2
and day 5 (8% decrease in the case of group 3 and 13% de-
crease for group 5), but remained stable afterwards (figure S2
in the Supporting Information). Due to partial weight loss in
groups 3 and 5, treatment was omitted twice for a two-day
period (days 11–12 and days 18–19). Two animals were found
dead in the vehicle group, and one animal in each of groups 3
and 5. Compounds 3 and 5 effectively suppressed tumor
growth (Figure 2) to a comparable extent as the positive con-
trol (sunitinib). Compound 5 effectively suppressed lung meta-
stasis, whereas groups treated with compound 3 and sunitinib
showed a higher level of lung metastasis compared with the
vehicle group (Figure 3). Next, we examined the level of angio-
genesis inhibition in the tumors (CD31 histology) and found
that while compound 3 effectively inhibited angiogenesis as
expected, compound 5 did not (Figure 4). Evaluation of the
stability of compounds 3 and 5 in blood revealed a half-life of
61 min and 96 min, respectively. Interestingly, compounds 3
and 5 were found to have significantly different rates of reac-
Figure 2. In vivo efficacy. Group 1: vehicle (n=12/10); group 2: sunitinib,
40 mgkgÀ1 po, q1d (n=12); group 3: 3, 20 mgkgÀ1 ip, q1d (n=12/11);
group 4: 3, 20 mgkgÀ1 ip, q2d (n=12); group 5: 5, 20 mgkgÀ1 ip, q1d
(n=12/11). A) Whole body in vivo luciferase activity (photons sÀ1); B) Primary
tumor volume (mm3); C) Primary tumor weight (g). Data are displayed as
means ÆSEM. The inset shows the individual data points together with
their corresponding mean values.
tion with dithiothreitol (DTT), with a half-life of 34 min and
3.6 min, respectively, (2 equivalents of DTT at 8 mm).
1694
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1692 – 1699

In Vivo Efficacy of Irreversible Kinase Inhibitors
However, enduring clinical responses are rare, and
there is mounting evidence that antiangiogenic ther-
apy, while delaying primary tumor growth, can con-
comitantly promote tumor adaptation and progres-
sion to stages of greater malignancy with heightened
invasiveness and distant metastasis.[24,25] It has recent-
ly been shown that short-term sunitinib treatment re-
sulted in accelerated metastasis in mice.[26] Substanti-
ating this pharmacological outcome, genetic ablation
of Vegf-A increased invasiveness.[27] Furthermore, pre-
treatment of healthy mice with these VEGFR inhibi-
tors prior to inoculation of tumor cells “conditioned”
the development of aggressive metastasis with short-
ened survival.[26] While several hypotheses have been
advanced to rationalize these observations, further
studies will be required to dissect the beneficial out-
come of tumor starvation from the undesired in-
creased invasiveness.[28] Encouraging results obtained
using a fusion protein that simultaneously binds
both VEGF-A and -C were recently reported;[29] this
concurs earlier findings regarding the relationship be-
tween lymphangiogenesis and metastasis.[30] In most
cancers, metastasis, not primary tumor growth, is ulti-
mately fatal.[31] Yet, most preclinical studies focus on
primary tumor growth with less attention to metasta-
sis. While sunitinib and sorafenib are very effective in-
hibitors of the angiogenic pathways and primary
tumor growth, we sought to evaluate these modified
resorcylic acid lactones, as the irreversible inhibition
of the target kinases might provide a different out-
come.[2,32,33] Several programs to develop irreversible
kinase inhibitors have already progressed to the
clinic,[34–36] however, none of which target the same
subset of kinases as the cis-enone RALs.
Compounds 3 and 5 are derived from a common
synthetic intermediate and can be readily prepared
in gram quantities. While compound 3 was notably
more soluble in aqueous buffers than 5, both com-
pounds were easily formulated in 10% DMSO, 40%
PEG, and 60% saline. Compound 3 was tested with
two different schedules (every day dosing and every
other day dosing), while compound 5 was tested at a
single schedule (every day dosing), and the results
were compared to a positive control, sunitinib, in an
orthotopic renal cancer model known to develop
metastasis. Both tested compounds 3 and 5 showed
a highly significant reduction of tumor growth as
measured by bioluminescence (Figure 2) and primary
tumor volume with P values >0.0010 comparable to
sunitinib (P=0.0004) for the daily dosing schedule.
Compound 3, when administered every other day,
Figure 3. Lung metastasis. Group 1: vehicle (n=12/10); group 2: sunitinib, 40 mgkgÀ1
po, q1d (n=12); group 3: 3, 20 mgkgÀ1 ip, q1d (n=12/11); group 4: 3, 20 mgkgÀ1 ip,
q2d (n=12); group 5: 5, 20 mgkgÀ1 ip, q1d (n=12/11). A) Number of lung metastases.
Data are displayed as means ÆSEM. The inset shows the individual data points together
with their corresponding mean values. B) Lung luciferase activity (LU per mg protein).
Data are displayed as means ÆSEM. The inset shows the individual data points together
with their corresponding mean values.
Discussion
showed a marginal decrease in activity (P=0.0011). Primary
tumor weight confirmed these observations (Figure 2, panel C).
It should be noted that we and others have found that cis-
enone RALs, such as LL Z-1640-2, generally have a short
plasma half-life and are inactivated by thiols, such as GST or
DTT.[37,38] Evaluation of the half-life of compounds 3 and 5 in
Multiple strategies for inhibiting the VEGF pathway have been
shown to hinder tumor growth, and the recent approval of
small-molecule inhibitors (sorafenib[18,19] and sunitinib[15,20]), as
well as neutralizing antibodies (bevacizumab[21,22]), targeting
the VEGF receptors have clinically validated this strategy.[23]
ChemBioChem 2010, 11, 1692 – 1699
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1695

N. Winssinger et al.
high count of nodules, whereas
the untreated group only had
two animals (Figure 3). The over-
all level of luciferase activity
from lung homogenates was
threefold higher in the sunitinib-
treated group than in the un-
treated group. A similar result
was obtained with compound 3
irrespective of the dosing sched-
ule. However, within the group
treated with compound 5, no
animal was found with a high
count of lung metastasis (P=
0.0445) nor significant luciferase
activity in lung homogenates
(P=0.0014); this indicates that
compound 5 effectively inhibited
both the tumor growth and
metastasis. Tumor vasculature
histology
(CD-31
staining)
showed a strong inhibition of
angiogenesis for animals treated
with compound 3 but not for
animals treated with com-
pound 5 (Figure 4). This observa-
tion was surprising considering
that both compounds have a
comparable profile of activity
against
a
panel of kinases
(Figure 5), with VEGFR2 and 3
being strongly inhibited in both
cases. However, considering that
both compounds 3 and 5 are in-
activated by thiols with com-
pound 5 being inactivated much
Figure 4. Tumor vasculature histology (CD-31 staining). For vehicle group (top), group 3 (3, 20 mgkgÀ1 ip, q1d;
middle); and group 5 (5, 20 mgkgÀ1 ip, q1d; bottom).
mice blood revealed similar findings with half-lives of 61 and
96 min, respectively. However, both compounds had dramati-
cally different susceptibility to thiols (8 mm), with compound 5
having a half-life of just 3.6 min versus 34 min for compound 3.
While these half-lives are rather short, the fact that these com-
pounds were highly efficient in suppressing primary tumor
growth does show that inactivation of their target is fast com-
pared to the inhibitor inactivation. Furthermore, the small dif-
ference in activity between the two dose schedules of com-
pound 3 (every day or every other day) suggests that target
turnover is slower than 12 h. Thus, while it has been shown
that the stability of cis-enone RALs to thiols might be extended
by introducing a fluorine at the a position of the enone,[37,38]
prolonged stability is not essential. A key objective of this
study was to measure the effect of these compounds on meta-
stasis. Upon termination of the study, the lungs were inspected
for metastatic nodules, and the ex vivo luciferase activity of
lung homogenates was assayed as a measure of the extent of
metastasis. In agreement with other studies and clinical obser-
vations, treatment with sunitinib resulted in a significant in-
crease in lung metastastatic nodules with five animals having a
faster (3.6 min versus 34 min in the presence of 8 mm DTT), dif-
ferences in the kinetics of kinase inactivation in vivo relative to
inhibitor inactivation might become significant. Indeed, it was
shown by Santi and co-workers that the rate of inactivation by
hypothemycin of 16 tested kinases varied significantly.[7] Thus,
a difference in the rate of kinase inhibition relative to a back-
ground drug inactivation could account for the difference in
selectivity observed. While kinetic selectivity is often exploited
in synthetic chemistry to favor a given reaction pathway, this
has not been extensively exploited pharmacologically to bias
target selectivity, and represents a unique feature of irreversi-
ble inhibitors. Although the precise difference between com-
pounds 3 and 5 in their target inactivation in vivo cannot be
readily assessed, the striking difference in angiogenesis inhibi-
tion between the two compounds, despite a fairly similar in-
hibition in biochemical assays, coupled with unequal suscepti-
bility to thiols suggests that such kinetic selectivity is operat-
ing. While compound 5 does not inhibit angiogenesis, the
tumor growth inhibition can be rationalized by activity on
other kinases such as MEKs. It should be noted that cis-enone
RALs were also shown to inhibit the inflammation path-
1696
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1692 – 1699

In Vivo Efficacy of Irreversible Kinase Inhibitors
Conclusions
The cis-enone resorcylic acid lac-
tones are highly promising
agents in oncology and are
worthy of further development.
The molecular editing[42,43] of this
natural pharmacophore has re-
vealed important differences in
biological activity in vivo, which
likely stems from differences in
the kinetics of target inactivation
versus inhibitor inactivation. In
most cases, metastasis, rather
than primary tumor growths,
kills the cancer patient. The in-
hibition of lung metastasis in
this present study is thus highly
relevant as it overcomes a major
limitation of current therapies.
Experimental Section
Cell proliferation assay: Murine
RENCA cells were originally ob-
tained from a tumor that arose
spontaneously in the kidney of
BALB/c mice. Histologically, RENCA
consists of granular-cell type ade-
nocarcinoma, which is pleomor-
phic with large nuclei. To obtain
RENCA-LN, RENCA cells were trans-
duced using a retrovirus encoding
a luciferase-neomycin fusion pro-
tein and selected using 1 mgmLÀ1
G418. RENCA-LN cells were plated
with 1500 cells per 150 mL per well
in Dulbecco’s modified eagle
medium (DMEM) supplemented
with 10% fetal calf serum (FCS) in
96-well culture dishes. The follow-
ing day the serial diluted test sam-
ples (from 1 mm to 0.1 mm in
DMSO) were mixed with medium
(1:25) and 50 mL of the mix was
added to each well of the cultured
cells (1:4); this resulted in a final
DMSO concentration of 1%. Cells
were grown for 72 h at 10% CO2
and 378C. After 72 h, RENCA-LN
cells were lyzed in 100 mL of luci-
ferase assay buffer, and 10 mL of
the lysate was assayed for lucifer-
ase activity. The mean value of the
low controls of each 96-well cell-
culture plate represents the back-
ground that was subtracted from
Figure 5. Kinase affinity profile. Profile of binding affinity of compounds 3 (blue bars) and 5 (red bars) for 50 kinas-
es in a competition binding assay.[13,14] The data are presented as the percentage of residual binding of a competi-
tive binder to the kinase. Thus, the lowest values reflect the strongest inhibition of the tested compound.
ways,[9,39,40] which have been implicated in facilitating metasta-
all other data points of the respective plate. Raw data were con-
verted into percent proliferation relative to the high control, which
was set to 100%. IC50 value determination was done with Graph-
Pad Prism 5.01 software with constrain of bottom to zero, and top
sis.[26] Finally, while hypothemycin had already been tested in
mice bearing colon and lung tumors, only modest growth in-
hibition was observed.[41]
ChemBioChem 2010, 11, 1692 – 1699
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1697

N. Winssinger et al.
to 100 using a nonlinear regression curve fitted with a variable Hill
slope. The equation is a four-parameter logistic equation. Experi-
ments were run in triplicates (IC50 graphs are shown in figure S1 of
the Supporting Information).
and positive (known binder) control. Residual binding (%)=[(test
compoundÀpositive control)/(tested compoundÀnegative con-
trol)ꢁ100].
Stability of compounds 3 and 5 in blood and to DTT: Com-
pounds 3 and 5 (100 mm in DMSO) were diluted in mouse blood
to a concentration of 1 mm and incubated at 378C with gentle
shaking. Samples of 100 mL were taken at 0, 15, 30, 60, 120 and
180 min and treated with methanol (300 mL) followed by centrifu-
gation (15000g) for 10 min. The supernatant was directly analyzed
by LC-MS, and the level of compounds 3 and 5 was quantified. The
half-life of compounds 3 and 5 was 61 min and 96 min, respective-
ly. Compounds 3 and 5 (4 mm) were dissolved in 10% [D6]DMSO in
deuterated PBS (pH 7.4) and treated with two equivalents of DTT
(8 mm) in the same solvent system. The conversion was monitored
by NMR spectroscopy at 298 K. The half-life of compound 3 and 5
was 34 min and 3.6 min, respectively.
In vivo efficacy: Experimental protocols involving animals were ap-
proved by the Ethics Committee for Animal Experimentation ac-
cording to the United Kingdom Coordinating Committee on
Cancer Research guidelines. RENCA-LN cells were obtained and cul-
tured as described above. Sixty female BALB/c mice of 5–6 weeks
were orthotopically implanted with RENCA-LN cells into the left
kidney (4ꢁ105 cells in 25 mL aliquots into the subcapsular space
through a flank incision; day 0). The mice were randomly divided
into five groups containing 12 mice each. On day 2, therapy was
initiated. Animals in the vehicle control group (group 1) received
10 mLkgÀ1 vehicle (PBS without Ca/Mg, 10% DMSO, and 40% PEG-
300) via intraperitoneal (i.p.) administration once daily until day 19.
Animals in the positive control group (group 2) received
40 mgkgÀ1 sunitinib via oral (p.o.) administration once daily until
day 20. Animals of treatment groups 3 and 4 received 20 mgkgÀ1
of 3 via i.p. administration once daily and every other day, respec-
tively. Animals of treatment group 5 received 20 mgkgÀ1 of 5 via
i.p. administration once daily. Due to partial weight loss, therapy in
groups 3 and 5 was suspended on days 11 and 12 and stopped on
day 17. On day 20, the study was terminated and a necropsy was
performed. All animals were weighed, anaesthetized by isoflurane
and killed by exsanguinations. Two animals in group 1 were found
dead on day 19 and one animal in groups 3 and 5 on days 11
and 9, respectively. At necropsy, primary tumor weight and
volume, lung weight and the number of lung metastases were de-
termined. Primary tumor tissues were collected and divided into
two parts. One part of each tumor was snap frozen in liquid nitro-
gen and stored at À808C, whereas the other part was analyzed
using ex vivo bioluminescence imaging. During the course of this
study, the growth of the orthotopically implanted RENCA tumors
was monitored on days 7, 14 and 19 using in vivo bioluminescence
imaging. For this purpose, luciferin was injected (ip) into the mice,
and light emission was measured post injection (10 min) using a
NightOWL LB 981 bioluminescence imaging system (Berthold Tech-
nologies, Germany). The migration of RENCA tumor cells into the
lung (metastasis) was analyzed post necropsy by counting nodules
and using the luciferase assay. For this purpose, lungs were collect-
ed, homogenized, and assayed for luciferase activity using the kit
from Promega (catalogue number: E1501) according to the manu-
facturer’s instructions. Protein concentrations were determined
using a Bradford assay and used to normalize luciferase activities
(except for the primary tumor). A histological examination of the
tumor vasculature was performed. For this purpose, cryosections
of primary tumor tissues (thickness=5–10 mm) from the snap-
frozen portion of the organ and blood vessels were visualized by
immunohistochemical staining for CD 31 (PECAM-1; PharMingen,
San Diego, CA). Statistical analysis: All relevant data described are
graphically displayed as means ÆSEM. Statistical analysis of effica-
cy data was done using the Mann–Whitney test.
Acknowledgements
This work was funded by a grant from the Agence National de la
Recherche (ANR) and Conectus. The Institut Universitaire de
France (IUF) is gratefully acknowledged for its support. There are
no conflicts of interest.
Keywords: antitumor agents · inhibitors · kinetics · lactones ·
natural products
[1] C. Drahl, B. F. Cravatt, E. J. Sorensen, Angew. Chem. 2005, 117, 5936–
5958; Angew. Chem. Int. Ed. 2005, 44, 5788–5809.
[2] R. A. Copeland, D. L. Pompliano, T. D. Meek, Nat. Rev. Drug Discovery
2006, 5, 730–739.
[3] D. C. Swinney, Curr. Opin. Drug Discov. Devel. 2009, 12, 31–39.
[4] M. S. Cohen, C. Zhang, K. M. Shokat, J. Taunton, Science 2005, 308,
1318–1321.
[5] J. Zhang, P. L. Yang, N. S. Gray, Nat. Rev. Cancer 2009, 9, 28–39.
[6] N. Winssinger, S. Barluenga, Chem. Commun. 2007, 22–36; S. Barluenga,
P. Dakas, M. Boulifa, E. Moulin, N. Winssinger, C. R. Chim. 2008, 11,
1306–1317; T. Hofmann, K. Altmann, C. R. Chim. 2008, 11, 1318–1335.
[7] A. Schirmer, J. Kennedy, S. Murli, R. Reid, D. V. Santi, Proc. Natl. Acad. Sci.
USA 2006, 103, 4234–4239.
[8] G. Rastelli, R. Rosenfeld, R. Reid, D. V. Santi, J. Struct. Biol. 2008, 164, 18–
23.
[9] J. Ninomiya-Tsuji, T. Kajino, K. Ono, T. Ohtomo, M. Matsumoto, M. Shiina,
M. Mihara, M. Tsuchiya, K. Matsumoto, J. Biol. Chem. 2003, 278, 18485–
18490.
[10] 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, 1086–1094.
[11] P. Y. Dakas, S. Barluenga, F. Totzke, U. Zirrgiebel, N. Winssinger, Angew.
Chem. 2007, 119, 7023–7026; Angew. Chem. Int. Ed. 2007, 46, 6899–
6902.
[12] R. Jogireddy, P-Y. Dakas, G. Valot, S. Barluenga, N. Winssinger, Chem. Eur.
J. 2009, 15, 11498–11506.
[13] M. W. Karaman, S. Herrgard, D. K. Treiber, P. Gallant, C. E. Atteridge, B. T.
Campbell, K. W. Chan, P. Ciceri, M. I. Davis, P. T. Edeen, R. Faraoni, M.
Floyd, J. P. Hunt, D. J. Lockhart, Z. V. Milanov, M. J. Morrison, G. Pallares,
H. K. Patel, S. Pritchard, L. M. Wodicka, P. P. Zarrinkar, Nat. Biotechnol.
2008, 26, 127–132.
[14] M. A. Fabian, W. H. Biggs III, D. K. Treiber, C. E. Atteridge, M. D. Azi-
mioara, M. G. Benedetti, T. A. Carter, P. Ciceri, P. T. Edeen, M. Floyd, J. M.
Ford, M. Galvin, J. L. Gerlach, R. M. Grotzfeld, S. Herrgard, D. E. Insko,
M. A. Insko, A. G. Lai, J.-M. Lꢂlias, S. A. Mehta, Z. V. Milanov, A. M. Velas-
co, L. M. Wodicka, H. K. Patel, P. P. Zarrinkar, D. J. Lockhart, Nat. Biotech-
nol. 2005, 23, 329–336.
Kinase selectivity profile: The selectivity profile was measured by
using the KinomeScan technology (http://www.kinomescan.com/)
based on active-site-dependent competition binding assays .[13,14]
Compounds 3 and 5 were profiled against a panel of 50 kinases,
including 42 kinases from the group of 48 kinases predicted by
Santi et al[7] to be targeted by hypothemycin, and five mutants of
FLT3, as well as three mutants of KIT. Both compounds were as-
sayed at 1 mm. The results are shown in Figure 5. The results are
expressed as a percentage of signal between a negative (DMSO)
1698
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1692 – 1699

In Vivo Efficacy of Irreversible Kinase Inhibitors
[15] R. J. Motzer, M. D. Michaelson, B. G. Redman, G. R. Hudes, G. Wilding,
R. A. Figlin, M. S. Ginsberg, S. T. Kim, C. M. Baum, S. E. DePrimo, J. Z. Li,
C. L. Bello, C. P. Theuer, D. J. George, B. I. Rini, J. Clin. Oncol. 2006, 24,
16–24.
[16] J. Barbat, J. Gelas, D. Horton, Carbohydr. Res. 1983, 116, 312–316.
[17] X. Geng, S. J. Danishefsky, Org. Lett. 2004, 6, 413–416.
[18] B. Escudier et al., N. Engl. J. Med. 2007, 356, 125–134.
[19] G. K. Abou-Alfa, L. Schwartz, S. Ricci, D. Amadori, A. Santoro, A. Figer, J.
De Greve, J.-Y. Douillard, C. Lathia, B. Schwartz, I. Taylor, M. Moscovici,
L. B. Saltz, J. Clin. Oncol. 2006, 24, 4293–4300.
[30] J. Krishnan, V. Kirkin, A. Steffen, M. Hegen, D. Weih, S. Tomarev, J. Wilt-
ing, J. P. Sleeman, Cancer Res. 2003, 63, 713–722.
[31] G. P. Gupta, J. Massague, Cell 2006, 127, 679–695.
[32] D. C. Swinney, Nat. Rev. Drug Discovery 2004, 3, 801–808.
[33] M. H. Potashman, M. E. Duggan, J. Med. Chem. 2009, 52, 1231–1246.
[34] C. Erlichman, M. Hidalgo, J. P. Boni, P. Martins, S. E. Quinn, C. Zacharch-
uk, P. Amorusi, A. A. Adjei, E. K. Rowinsky, J. Clin. Oncol. 2006, 24, 2252–
2260.
[35] P. A. Janne, J. von Pawel, R. B. Cohen, L. Crino, C. A. Butts, S. S. Olson,
I. A. Eiseman, A. A. Chiappori, B. Y. Yeap, P. F. Lenehan, K. Dasse, M.
Sheeran, P. D. Bonomi, J. Clin. Oncol. 2007, 25, 3936–3944.
[36] F. A. Eskens, C. H. Mom, A. S. Planting, J. A. Gietema, A. Amelsberg, H.
Huisman, L. van Doorn, H. Burger, P. Stopfer, J. Verweij, E. G. de Vries, Br.
J. Cancer 2008, 98, 80–85.
[20] G. D. Demetri, A. T. van Oosterom, C. R. Garrett, M. E. Blackstein, M. H.
Shah, J. Verweij, G. McArthur, I. R. Judson, M. C. Heinrich, J. A. Morgan, J.
Desai, C. D. Fletcher, S. George, C. L. Bello, X. Huang, C. M. Baum, P. G.
Casali, Lancet 2006, 368, 1329–1338.
[21] K. Miller, M. Wang, J. Gralow, M. Dickler, M. Cobleigh, E. A. Perez, T.
Shenkier, D. Cella, N. E. Davidson, N. Engl. J. Med. 2007, 357, 2666–
2676.
[37] H. Du et al., Bioorg. Med. Chem. Lett. 2009, 19, 6196–6199.
[38] R. Jogireddy, S. Barluenga, N. Winssinger, ChemMedChem 2010, 5, 670–
673.
[22] H. Hurwitz, L. Fehrenbacher, W. Novotny, T. Cartwright, J. Hainsworth,
W. Heim, J. Berlin, A. Baron, S. Griffing, E. Holmgren, N. Ferrara, G. Fyfe,
B. Rogers, R. Ross, F. Kabbinavar, N. Engl. J. Med. 2004, 350, 2335–2342.
[23] J. Folkman, Nat. Rev. Drug Discovery 2007, 6, 273–286.
[24] G. Bergers, D. Hanahan, Nat. Rev. Cancer 2008, 8, 592–603.
[25] R. S. Kerbel, N. Engl. J. Med. 2008, 358, 2039–2049.
[26] J. M. Ebos, C. R. Lee, W. Cruz-Munoz, G. A. Bjarnason, J. G. Christensen,
R. S. Kerbel, Cancer Cell 2009, 15, 232–239.
[39] T. Kastelic, J. Schnyder, A. Leutwiler, R. Traber, B. Streit, H. Niggli, A.
Mackenzie, D. Cheneval, Cytokine 1996, 8, 751–761.
[40] M. Goto, J. Chow, K. Muramoto, K. Chiba, S. Yamamoto, M. Fujita, H. Ob-
aishi, K. Tai, Y. Mizui, I. Tanaka, D. Young, H. Yang, Y. J. Wang, H. Shirota,
F. Gusovsky, J. Pharmacol. Exp. Ther. 2009, 331, 485–495.
[41] H. Tanaka, K. Nishida, K. Sugita, T. Yoshioka, Jpn. J. Cancer Res. 1999, 90,
1139–1145.
[42] R. M. Wilson, S. J. Danishefsky, J. Org. Chem. 2006, 71, 8329–8351.
[43] K. C. Nicolaou, J. S. Chen, S. M. Dalby, Bioorg. Med. Chem. 2009, 17,
2290–2303.
[27] M. Pꢃez-Ribes, E. Allen, J. Hudock, T. Takeda, H. Okuyama, F. Vinals, M.
Inoue, G. Bergers, D. Hanahan, O. Casanovas, Cancer Cell 2009, 15, 220–
231.
[28] S. Loges, M. Mazzone, P. Hohensinner, P. Carmeliet, Cancer Cell 2009, 15,
167–170.
Received: April 1, 2010
[29] D. Zhang, B. Li, J. Shi, L. Zhao, X. Zhang, C. Wang, S. Hou, W. Qian, G.
Kou, H. Wang, Y. Guo, Cancer Res. 2010, 70, 2495–2503.
Published online on July 7, 2010
ChemBioChem 2010, 11, 1692 – 1699
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1699