Identification and optimisation of novel and selective small molecular weight kinase inhibitors of mTOR

Article abstract of DOI:10.1016/j.bmcl.2009.08.069

A pharmacophore mapping approach, derived from previous experience of PIKK family enzymes, was used to identify a hit series of selective inhibitors of the mammalian target of rapamycin (mTOR). Subsequent refinement of the SAR around this hit series based

Full text of DOI:10.1016/j.bmcl.2009.08.069

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5898–5901
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification and optimisation of novel and selective small molecular
weight kinase inhibitors of mTOR
Keith A. Menear a, Sylvie Gomez a, Karine Malagu a, Christine Bailey a, Kristel Blackburn a,
Xiao-Ling Cockcroft a, Sally Ewen a, Alexandra Fundo a, Armelle Le Gall a, Gesine Hermann a,
Luisa Sebastian a, Mihiro Sunose a, Thomas Presnot a, Eleanor Torode a, Ian Hickson a,
Niall M. B. Martin a, Graeme C. M. Smith a,b, Kurt G. Pike b,
*
a KuDOS Pharmaceuticals Ltd, 410 Cambridge Science Park, Milton Road, Cambridge CB4 0PE, UK
b AstraZeneca, Alderley Park, Macclesfield, Cheshire SK10 4TG, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A pharmacophore mapping approach, derived from previous experience of PIKK family enzymes, was
used to identify a hit series of selective inhibitors of the mammalian target of rapamycin (mTOR). Subse-
quent refinement of the SAR around this hit series based on a tri-substituted triazine scaffold has led to
the discovery of potent and selective inhibitors of mTOR.
Received 14 July 2009
Revised 18 August 2009
Accepted 19 August 2009
Available online 21 August 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Mammalian target of rapamycin
mTOR
Kinase inhibitors
Phosphatidylinositol 3-kinase
PI3K
The mammalian target of rapamycin (mTOR) is a key target in
the development of antitumour therapies.1 Activated by growth
factor/mitogenic stimulation activation of the phosphatidylinositol
3-kinase (PI3K)/Akt signalling pathway mTOR is a central regulator
of cell growth and proliferation. This PI3K-Akt-mTOR pathway is
one of the most frequently dysregulated pathways in cancer.2
mTOR, a serine/threonine kinase of approximately 289 kDa in size,
is a member of the evolutionary conserved eukaryotic PI3K like ki-
nase (PIKK) family of proteins, for example, DNA-PK (DNA depen-
dent protein kinase) and ATM (Ataxia-telangiectasia mutated).3–5
The known mTOR inhibitor Rapamycin and its analogues
(RAD001, CCI-779, AP23573) bind to the FKBP12/rapamycin com-
plex binding domain (FRB), resulting in suppression of signalling
to the downstream targets p70S6K and 4E-BP1.6,7 The potent but
non-specific inhibitors of PI3K, LY294002 and wortmannin, have
also been shown to inhibit the kinase function of mTOR; however,
in this case the catalytic domain of the protein is targeted.8
Recently it has been shown that mTOR can exist in an alterna-
tive, rapamycin insensitive, complex that signals to Akt.9 The
existence of both a rapamycin sensitive complex (mTORC1) and a
rapamycin insensitive complex (mTORC2) may provide an
explanation for the differences observed in the earlier work of
Brunn et al.8 and Edinger et al.10 Therefore, it is proposed that a
compound directly targeting the kinase domain of mTOR would in-
hibit signalling through both mTORC1 and mTORC2 and as such the
compound would exhibit a different spectrum of pharmacology
compared with rapamycin. As such a number of groups, including
our own, have sought to identify selective inhibitors of mTORC1
and mTORC2.11,12
Herein, we report the discovery and optimisation of a series of
ATP-competitive morpholino-pyrimidines and triazines as potent
and selective inhibitors of mTOR kinase.
Our search for inhibitors of mTOR began with a pharmacophore
analysis derived from previous experience of another member of
the PIKK family; DNA-PK.13 Additional filtering constraints based
on physicochemical properties known to increase the chance of
drug-like properties were also considered14 and the resulting mod-
el applied to a database search of over 1 million commercially
available compounds. 108 candidates were subsequently nomi-
nated for purchase and screened for inhibitory activity against
mTOR kinase.15 13 compounds exhibited an IC50 <15
compounds exhibited an IC50 <0.5 M. In particular bis-morpho-
line-triazine (1) was considered an attractive start point for further
chemical optimisation with an mTOR IC50 of 0.27
M16 and com-
plete selectivity against the other PIKK family members tested
(PI3K and ATM IC50 >10 M), Figure 1.
lM and 2
l
l
* Corresponding author. Tel.: +44 1625 512733.
E-mail address: kurt.pike@astrazeneca.com (K.G. Pike).
a
l
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.08.069

K. A. Menear et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5898–5901
5899
tency against mTOR with an IC50 of 0.006
and high levels of selectivity against PI3K
Further exploration of heterocyclic replacements for the hydra-
zone motif was carried out with the monosubstituted pyrimidine
scaffold, Table 2.
Contrary to our initial expectations the switch to the monosub-
stituted pyrimidine scaffold resulted in a considerable loss in po-
tency (5 vs 6). The nature of the heterocycle appears to have a
significant impact on the mTOR potency, presumably due to the
subtle differences in the vectors of the two pendant rings as well
as the differing electronics. Switching to the pyrazole (7) or the thi-
ophene (8) resulted in a significant loss of potency whereas the
pyridine (9) and thiazole (10) examples showed similar levels of
potency. Interestingly the alternative thiazole isomer (11) and
the furan (12) showed significantly improved potency.
l
a
M in the enzyme assay
(IC50 >10 lM).
O
N
O
OH
O
N
N
N
N
N
N
H
O
Figure 1. Structure of initial hit compound (1).
The morpholine motif is a common pharmacophore present in
many inhibitors of the PIKK family17,18 and, at least for the case
of the prototypical PI3K inhibitor LY294002, the morpholine oxy-
gen has been shown to participate directly in a key hydrogen bond-
ing interaction in the ATP-binding site of PI3Kc 19
By analogy we
.
reasoned that one of the morpholine oxygens present in (1) may
also form a key interaction with residue Val 133 in the hinge region
of the mTOR kinase domain. Initial modifications of (1), outlined in
Table 1, were therefore designed to explore this hypothesis. Re-
moval of the oxygen from a single morpholine to give (2) resulted
in a moderate loss of activity; however, the removal of oxygen
from both morpholine rings (3) showed a significant loss of po-
tency. These findings are consistent with our view of one of the
morpholines being involved in a key hydrogen bonding interaction
and possibly suggests that the other morpholine motif could be
completely removed to maximise ligand efficiency; a tactic that
was employed later in our template optimisation. Interestingly
the trizaine ring could be replaced by pyrimidine with only a mod-
est effect on potency (4).
Clearly the pyrimidine scaffold exemplified in (6) can provide
alternative regioisomers. To determine the influence of this
arrangement the reference compound (13) was prepared, Figure 3.
A loss in potency with respect to compound (6) was noted which
encouraged the continued use of the 2-morpholine substitution
pattern for subsequent SAR investigation.
Table 2
Modifications to the central ring heterocycle
O
OH
O
N
Next our attention turned to replacement of the hydrazone
functionality in (1). Initial studies aimed at finding a more chemi-
cally stable hydazone replacement identified the cyclised analogue
(5), Figure 2. This compound displayed significantly improved po-
N
N
X
O
Compds
X
mTOR IC50
0.46
(lM)
N
N
6
N
Table 1
Modification of the trazine portion of the molecule
7
8
<50% at 10
<50% at 10
0.55
l
l
M
N
R2
O
OH
N
A
S
M
N
R1
O
N
N
H
O
N
R1
R2
A
mTOR IC50
0.75
(lM)
9
Compds
2
N
N
N
N
N
N
10
11
12
0.56
S
3
4
N
5.9
N
N
N
0.18
S
O
O
CH
0.66
O
0.052
O
N
O
O
N
N
N
OH
OH
N
N
N
N
N
N
O
N
N
O
O
O
Figure 3. Structure of compound (13), mTOR IC50 = 0.57 lM.
Figure 2. Structure of compound (5).

5900
K. A. Menear et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5898–5901
to show some promise. In the case of the benzyl alcohol (19) the
O
N
additional methylene spacer presumably allows the positioning
of the hydroxyl group to make the putative hydrogen bond with
the enzyme necessary for activity. Attempts to find phenolic bio-
isosteres, such as (20) and (21) proved unsuccessful.
O
O
O
N
N
N
N
N
N
H
A selection of the most potent mTOR inhibitors were further
O
profiled in
a cellular screen using the growth inhibition of
U87MG cells as the readout, Table 4. However, despite excellent
levels of enzyme potency and selectivity these compounds showed
disappointing levels of activity in cells. Although the reasons for
the disappointing cell activity are as yet unknown, it is tempting
to speculate that the presence of phenolic functionality in the mol-
ecules results in compromised cell permeability. Indeed a differ-
ence in apparent permeability (Papp) was observed between
compounds (17) and (19) in an MDCKII assay,20,21 which may be
regarded as supporting this hypothesis (Papp in the A-B direction
was 2.2 Â 10À6 cm/s and 33 Â 10À6 cm/s, respectively). Improving
the cellular potency remains an important goal for the further opti-
misation of this series as does more fully exploring the selectivity
profile of this series.
Figure 4. Structure of compound (14), mTOR IC50 = 30 lM.
To investigate the importance of the phenolic functionality
present in these molecules, the reference compound (14) was pre-
pared, Figure 4. This compound showed a dramatic loss of potency
compared to the initial hit (1) suggesting that the presence of this
hydrogen bond donor is critical for activity.
More detailed studies of the SAR in this portion of the molecule
were conducted on the optimised pyrimidinylfuranyl template,
Table 3.
In line with previous SAR the 4-hydroxy-3,5-dimethoxyphenyl
compound (15) displays excellent potency. Paring down the tri-
substituted system to give either compound (16) or (17) proved
detrimental to potency. Investigation into alternative functional
groups with the capacity to act as a hydrogen bond donor appeared
For the analogues described above synthetic access was vari-
ously through modification of mono, di or trichloropyrimidine or
triazine (27, 29, 32 or 35) as shown in Scheme 1.
Table 4
Table 3
Cellular activity with selected potent inhibitors
Modification of the phenolic functionality
O
N
Compds
mTOR IC50
(lM)
PI3K
a
IC50
(lM)
U87MG GI50 (lM)
5
11
12
15
0.021
0.18
0.052
0.023
>10
>10
>10
>10
24
21
3.7
16
N
O
R3
N
O
N
O
X
N
Compds
R3
mTOR IC50
0.023
(lM)
Cl
N
X, Y = O
X = O, Y = CH2
X, Y = CH2,
1, 14
2
3
ii)
i)
N
N
OH
O
N
N
15
Cl
Cl
N
N
22
Cl
Y
OH
iii)
iv)
N
16
17
18
19
20
21
0.55
1.5
N
N
6, 7
N
N
23
N
X
Cl
Cl
N
Cl
Br
O
Y
v)
iv)
N
N
OH
8 - 11
A
N
N
24
N
Br
O
X
Y
OH
O
5.5
Cl
N
N
vi)
vii)
N
15 - 21
Cl
Cl
O
N
N
25
0.51
0.52
>10
Br
OH
O
Scheme 1. Reagents and conditions: (i) morpholine or piperidine, acetone, 0 °C or
(a) morpholine (50% aqueous solution), MeOCH2CH2OMe, À15 °C; (b) piperidine,
K2CO3, DMF, À5 °C; (ii) (a) hydrazine hydrate, ethanol, reflux; (b) 4-hydroxy-3,5-
dimethoxybenzaldehyde, TsOH, ethanol, reflux; (iii) (a) 4-bromoimidazole or 4-
bromopyrazole, K2CO3, DMF, 0 °C; (b) morpholine, NaH, DMF, 0 °C then microwave
120 °C; (iv) 4-hydroxy-3,5-dimethoxyphenyl boronic acid, K2CO3, Pd(Ph3)4, diox-
ane, microwave 150 °C; (v) (a) dibromoheterocyle, nBuLi, DDQ, Et2O, À78 °C to rt;
(b) morpholine, NaH, DMF, 0 °C then microwave 120 °C; (vi) morpholine, ethanol,
Et3N, 0 °C to rt; (vii) (a) 2-furanbornic acid, K2CO3, Pd(Ph3)4, dioxane, microwave
150 °C; (b) NBS, DMF, 0 °C; (c) aryl boronic acid, K2CO3, Pd(Ph3)4, dioxane,
microwave 150 °C.
N
NH

K. A. Menear et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5898–5901
5901
4. Sabatini, D. M.; Erdjument-Bromage, H.; Lui, M.; Tempst, P.; Snyder, S. H. Cell
1994, 78, 35.
5. Abraham, R. T. Curr. Opin. Immunol. 1996, 8, 412.
6. Fingar, D. C.; Salama, S.; Tsou, C.; Harlow, E.; Blenis, J. Gene Dev. 2002, 16, 1472.
7. Fingar, D. C.; Richardson, C. J.; Tee, A. R.; Cheatham, L.; Tsou, C.; Blenis, J. Mol.
Cell. Biol. 2004, 24, 200.
8. Brunn, G. J.; Williams, J.; Sabers, C.; Wiederrecht, G.; Lawrence, J. C., Jr.;
Abraham, R. T. EMBO J. 1996, 15, 5256.
9. Sarbassov, D. D.; Ali, S. M.; Sabatini, D. M. Curr. Opin. Cell Biol. 2005, 17, 596.
10. Edinger, A. L.; Linardic, C. M.; Chiang, G. G.; Thompson, C. B.; Abraham, R. T.
Cancer Res. 2003, 63, 8451.
11. Thoreen, C. C.; Kang, S. A.; Chang, J. W.; Liu, Q.; Zhang, J.; Gao, Y.; Reichling, L. J.;
Sim, T.; Sabatini, D. M.; Gray, N. S. J. Biol. Chem. 2009, 284, 8023.
12. Feldman, M. E.; Apsel, B.; Uotila, A.; Loewith, R.; Knight, Z.; Ruggero, D.; Shokat,
K. M. PLoS Biol. 2009, 7, 371.
13. Hollick, J. J.; Rigoreau, L. J. M.; Cano-Soumillac, C.; Cockcroft, X.; Curtin, N. J.;
Frigerio, M.; Golding, B. T.; Guiard, S.; Hardcastle, I. R.; Hickson, I.;
Hummersone, M. G.; Menear, K. A.; Martin, N. M. B.; Matthews, I.; Newell, D.
R.; Ord, R.; Richardson, C. J.; Smith, G. C. M.; Griffin, R. J. J. Med. Chem. 2007, 50,
1958.
Cyanuric chloride can be sequentially reacted with the appro-
priate amines to furnish intermediate (22). The remaining halogen
can be displaced by hydrazine and condensed with the appropriate
aldehyde to furnish the desired hydrazone containing compounds.
An analogous route could be used to obtain (4). Bromoheterocycles
could be introduced by either the nucleophilic attack of nitrogen
heterocycles on dichloropyrimidine or by deprotonation of 2-chlo-
ropyrimidine and subsequent reaction with dibromoheterocycles
in the presence of DDQ. Subsequent reaction with morpholine fur-
nished intermediates (23) and (24). Installation of the 4-hydroxy-
3,5-dimethoxyphenyl motif was achieved using Suzuki coupling
reactions. An analogous route was used to furnish (5) and (13) Stir-
ring 2,4,6-trichloropyrimidine with morpholine resulted in reac-
tion in both the 2- and 4-positions. The remaining halogen could
be reacted with 2-furanboronic acid in a Suzuki coupling reaction
and the subsequent material brominated to furnish intermediate
(25). A variety of aryl groups were subsequently introduced using
Suzuki couplings. An analogous route was used to furnish (12).
In conclusion, we have identified a novel series of PIKK family
selective inhibitors of mTOR based on a hit found through a phar-
macophore analysis. The systematic optimisation of the molecules
has resulted in the identification of a number of extremely potent
and selective inhibitors of the enzyme mTOR kinase.
14. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feency, P. J. Adv. Drug Delivery Rev.
1997, 23, 3.
15. Inhibition of mTOR kinase was assayed using full-length mTOR (extracted from
HeLa cytoplasmic extract by immunoprecipitation) with a recombinant 4E-BP1
protein substrate and an ATP concentration of 50
lM. Detection of
phosphorylated product through an ELISA format was essentially as
described in Ref. 8 Figures quoted are the mean of at least 2 determinations
and variability around the mean was <50%.
16. mTOR Kinase inhibition experiments were conducted at increasing
concentrations of inhibitor (1) (50, 100, 200, 350, 500 nM) in varying
concentrations of ATP (10–100 lM). The data was plotted using a
Lineweaver-Burke double reciprocal plot and revealed characteristic
behaviour of competitive inhibition.
17. Verheijen, J. C.; Zask, A. Drugs Future 2007, 32, 537.
Acknowledgements
18. Hickson, I.; Zhao, Y.; Richardson, C. J.; Green, S. J.; Martin, N. M.; Orr, A. I.;
Reaper, P. M.; Jackson, S. P.; Curtin, N. J.; Smith, G. C. Cancer Res. 2004, 64, 9152.
19. Walker, E. H.; Pacold, M. E.; Perisic, O.; Stephens, L.; Hawkins, P. T.; Wymann,
M. P.; Williams, R. L. Mol. Cell. 2000, 6, 909.
The authors would like to thank Dr Lisa Smith and Mrs Sharon
Maguire for their help in producing the biological data presented.
20. Monolayers of MDCKII (Madin Darby Canine Kidney Cells) cells transfected
with the human MDR1 (Multi-Drug Resistance 1) transporter protein were
cultured in transwells and used to study the permeability and efflux potential
References and notes
of compounds (17) and (19) at
a concentration of 10 lM. Since neither
1. Huang, S.; Houghton, P. J. Curr. Opin. Pharmacol. 2003, 3, 371.
2. Yap, T. A.; Garrett, M. D.; Walton, M. I.; Raynaud, F.; de Bono, J. S.; Workman, P.
Curr. Opin. Pharmacol. 2008, 8, 393.
compound showed significant levels of efflux in this assay, the Papp in the A to
B direction is considered to provide a measure of the passive permeability of
the compounds in these cells.
3. Brown, E. J.; Albers, M. W.; Shin, T. B.; Ichikawa, K.; Keith, C. T.; Lane, W. S.;
Schreiber, S. L. Nature 1994, 369, 756.
21. Keogh, J. P.; Kunta, J. R. Eur. J. Pharm. Sci. 2006, 27, 543.