A hit to lead discovery of novel N-methylated imidazolo-, pyrrolo-, and pyrazolo-pyrimidines as potent and selective mTOR inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.07.027

A series of N-7-methyl-imidazolopyrimidine inhibitors of the mTOR kinase have been designed and prepared, based on the hypothesis that the N-7-methyl substituent on imidazolopyrimidine would impart selectivity for mTOR over the related PI3Kα and δ kinases. The corresponding N-Me substituted pyrrolo[3,2-d]pyrimidines and pyrazolo[4,3-d]pyrimidines also show potent mTOR inhibition with selectivity toward both PI3α and δ kinases. The most potent compound synthesized is pyrazolo[4,3-d]pyrimidine 21c. Compound 21c shows a K_i of 2 nM against mTOR inhibition, remarkable selectivity (>2900×) over PI3 kinases, and excellent potency in cell-based assays.

Full text of DOI:10.1016/j.bmcl.2013.07.027

Bioorganic & Medicinal Chemistry Letters xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A hit to lead discovery of novel N-methylated imidazolo-, pyrrolo-,
and pyrazolo-pyrimidines as potent and selective mTOR inhibitors
a,
a
a
c
b
Wendy Lee a, , Daniel F. Ortwine , Philippe Bergeron , Kevin Lau , Lichuan Lin , Shiva Malek ,
Jim Nonomiya b, Zhonghua Pei a, Kirk D. Robarge a, Stephen Schmidt b, Steve Sideris b, Joseph P.
Lyssikatos a
⇑
⇑
a Department of Discovery Chemistry, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
b Department of Biochemical Pharmacology, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
c Drug Metabolism and Pharmacokinetics, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of N-7-methyl-imidazolopyrimidine inhibitors of the mTOR kinase have been designed and
prepared, based on the hypothesis that the N-7-methyl substituent on imidazolopyrimidine would
Received 20 May 2013
Revised 10 July 2013
Accepted 16 July 2013
Available online xxxx
impart selectivity for mTOR over the related PI3K
pyrrolo[3,2-d]pyrimidines and pyrazolo[4,3-d]pyrimidines also show potent mTOR inhibition with selec-
tivity toward both PI3 and d kinases. The most potent compound synthesized is pyrazolo[4,3-d]pyrim-
a and d kinases. The corresponding N-Me substituted
a
idine 21c. Compound 21c shows a Ki of 2 nM against mTOR inhibition, remarkable selectivity (>2900ꢀ)
Keywords:
over PI3 kinases, and excellent potency in cell-based assays.
Mammalian target of rapamycin
mTOR kinase inhibitors
PI3K/Akt/mTOR pathway
Structure based design
Oncology
Ó 2013 Elsevier Ltd. All rights reserved.
The mammalian target of rapamycin (mTOR) is a serine/threo-
nine protein kinase of the phosphoinositide 3-kinase-related ki-
nase (PIKK) family and a key point for inhibition downstream in
the PI3K–Akt signaling pathway. mTOR plays a central role as a
regulator of cell growth, proliferation, metabolism, and survival.1
Inhibition of mTOR signaling can be used for treating solid tumors
with mutations or deletion of the tumor suppressor phosphatase
and tensin homolog (PTEN) gene.2 The mTOR signaling pathway
consists of two different multi-protein complexes: mTOR com-
plex-1 (mTORC1) and mTOR complex-2 (mTORC2). mTORC1 is
known for its critical role in protein synthesis and can phosphory-
late two key regulators of protein translation, 4EBP and S6K.
mTORC2 can phosphorylate Akt and functions as a key regulator
of cell growth, metabolism, and survival. In certain cancer cells,
Akt activity is upregulated and enhances cell survival due to a
feedback loop between mTORC1 and Akt.3 The natural product
rapamycin, and semi-synthetic analogs thereof, selectively inhibit
mTORC1 activity in an allosteric manner.4 These so-called
‘rapalogs’ have beenused clinically to validate mTOR as a cancer
target. Second generation mTOR inhibitors are small molecules
that bind competitively and reversibly to the mTOR–ATP binding
pocket. These ATP-competitive mTOR agents inhibit both mTORC1
and mTORC2, and have shown preclinically to block tumor cell
proliferation more effectively than rapalogs.5 Because of the highly
conserved ATP binding pockets of the PIKK family, the first set of
ATP-competitive mTOR molecules proved to be dual PI3K/mTOR
inhibitors. A concern for dual PI3K/mTOR inhibitors is the potential
effect of insulin resistance and glucose metabolism regulation
through the activation of Akt downstream from PI3K.6 Thus, there
have been intense efforts in finding selective mTOR inhibitors that
may offer increased tolerability relative to dual PI3K/mTOR inhib-
itors. Many researchgroups,7,8 including our own,9 have reported
studies toward identification of selective ATP-competitive mTOR
inhibitors.
A high-throughput screen of our corporate compound library
identified imidazolopyrimidine 110 as a modestly potent mTOR
inhibitor with a Ki of 72 nM (Fig. 1). Introduction of a substituted
morpholine at the hinge binding region, and replacement of the
aminopyrimidine with a phenyl urea on related pyrimidine scaf-
folds were reported to increase mTOR potency and selectivity over
PI3 kinases.7 The selectivities were due to the S-methyl morpholine
packing against a unique Trp2239 residue in mTOR and the ethyl
urea interacting hydrophobically with the Cb and C
c atoms of
the Glu2190 side chain. Thus, the morpholine/aminopyrimdine of
1 was replaced with (S)-3-methyl-morpholine/ethyl phenyl urea
to provide a more potent and selective mTOR inhibitor 2. Relative
⇑
Corresponding authors. Tel.: +1 650 225 7704; fax: +1 650 742 4943.
E-mail addresses: wendyle@gene.com (W. Lee), danielfo@gene.com
(D.F. Ortwine).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.07.027
Please cite this article in press as: Lee, W.; et al. Bioorg. Med. Chem. Lett. (2013), http://dx.doi.org/10.1016/j.bmcl.2013.07.027

2
W. Lee et al. / Bioorg. Med. Chem. Lett. xxx (2013) xxx–xxx
O
N
O
Phenyl urea addition
and
N
6N 1
N
(S)-3-Me-morpholine
7N
N
N
5
4
substitution
8
2
HO
HO
N
N
N
N
O
9
3
N
H
N
NH2
N
H
HO
HO
1
2
mTOR Ki = 72 nM
PI3Kα Ki= 2 nM, PI3Kα / mTOR = 0.03
PI3Kδ Ki = 4 nM, PI3Kδ / mTOR = 0.06
mTOR Ki = 9 nM
PI3Kα Ki = 190 nM, PI3Kα / mTOR = 21
PI3Kδ Ki = 160 nM, PI3Kδ / mTOR = 18
Figure 1. Improvement of mTOR potency and selectivities over P13K isoforms.
Table 1
Initial SAR of 2: mTOR, PI3Ka
, and PI3Kd potencies and selectivitiesa,b
O
N
N
N
N
R1
N
O
R2
N
H
N
H
Compd
R1
R2
mTOR Ki (nM)
PI3K
a
Ki (nM)
PI3K
a
/mTOR
PI3Kd Ki (nM)
PI3Kd/mTOR
2
3
4
–CH(CH3)2OH
–CH(CH3)2OH
H
–CH2CH2OH
Me
Me
9
7
12
190
390
370
21ꢀ
56ꢀ
31ꢀ
160
490
890
18ꢀ
70ꢀ
74ꢀ
a
Assays are described in Refs. 9b and 11.
Values are at least two separate determinations with typical variations of less than 30%.
b
to 1, 2 was eightfold more potent (mTOR Ki = 9.5 nM) and exhibited
suggested that this N-7-methyl could also pack against the
Trp2239 residue, occupying the predicted small hydrophobic pock-
et in this area. Additionally, the presence of this methyl was pre-
dicted to restrict the allowed conformations of the morpholine
ring such that the (S)-3-methyl group can swing toward Tyr2225
(Supplementary data Fig. S1b). We hoped this would further
increase mTOR potency whilemaintaining selectivity against the
21- and 17-fold selectivities for mTOR over PI3K
a and PI3Kd
respectively (Fig. 1). Initial SAR exploration by truncation of the
2-hydroxyethyl side chain led to methyl analog 3, which showed
comparable mTOR potency (Ki = 7 nM) with decreased PI3K
a and
d potencies (Table 1). Removal of the C-8 tertiary alcohol resulted
in 4, with similar potencies for mTOR (Ki = 12 nM) and
PI3Ka
(Ki = 370 nM), and a slight decrease in PI3Kd potency
PI3Ka and PI3Kd isoforms, without a substantial increase in molec-
(IC50 = 890 nM).
ular weight. In this Letter, we describe the synthesis and biological
evaluation of a series of mTOR selective N-7 substituted imidazol-
opyrimidines and expand the SAR to include N-Me-pyrrolopyrimi-
dine and N-Me-pyrazolopyrimidine scaffolds.
Using a homology model of mTOR built from an X-ray crystal
structure of PI3Kc, we initially hoped to improve selectivity against
PI3K by occupying a predicted small hydrophobic pocket adjacent
to the morpholine binding region formed by a rigidly held Trp2239
N-7-Substituted imidazolopyrimidines 11a–f (Scheme 1) were
prepared by alkylation of 4,6-dichloro-1H-benzo[d]imidazole with
either iodomethane or iodoethane under basic conditions to obtain
the desired N-7-alkylated 2,6-dichloro-imidazolopyrimidine 7a or
7b, respectively, in modest yield after separationfrom the major
N-9 alkylated products 6a or 6b. Displacement of 7 with various
morpholine derivatives selectively produced 6-morpholino inter-
mediates 9, which underwent standard Suzuki coupling with the
pinacol ester of N-ethyl-N0-phenyl urea 10 to give N-7-alkylated
imidazolopyrimidines 11a–f. Regiochemically controlled synthesis
of 7a was accomplished through treatment of theobromine 8 with
phosphoryl chloride in the presence of N,N-diethylaniline, but the
overall yield (10–15%) was not superior to N-alkylation.12 A high
yielding method for regiospecific controlled synthesis of 7a using-
residue (Ile 881 in PI3K
gen bond to the backbone NH of Val2240 in mTOR, in an equivalent
location to Val882 in PI3K
. Consistent with previous studies,7a,8e
c). The morpholine oxygenforms a hydro-
c
the (S)-3-methyl group on the morpholine of compounds 2–4
was predicted to fit into this hydrophobic pocket and pack against
Trp2239 without significantly disturbing its conformation (Fig. 2a).
An analysis of the morpholine-to-purine bond showed two possi-
ble low energy conformations when the ligand was analyzed in va-
cuo (Supplementary data Fig. S1) However, the same analysis
carried out in the presence of the protein showed only one con-
former predominated, which placed the (S)-3-methyl group adja-
cent to Tyr2225 (Fig. 2b). Additionally, the morpholino oxygen
could adopt an appropriate orientation to engage in an H-bonding
interaction with the backbone NH of Val2240 in mTOR.
trimethyloxonium
borofluorate
with
N-9-(4-methoxyben-
This proposed binding mode led to a hypothesis in the current
work that we may be able to further increase selectivity over
PI3K by incorporating a small lipophilic group such as a methyl
at the N-7 position on the imidazolopyrimidine scaffold. Modeling
zyl)purine in 2,2,2,-trifluoroethanol was later reported.13 To
obtain the C-8 tertiary alcohol 13 (Scheme 2), 9a was lithiated at
C-8 using lithium diisopropylamide and then quenched with
acetone, followed by Suzuki coupling, as described above.
Please cite this article in press as: Lee, W.; et al. Bioorg. Med. Chem. Lett. (2013), http://dx.doi.org/10.1016/j.bmcl.2013.07.027

W. Lee et al. / Bioorg. Med. Chem. Lett. xxx (2013) xxx–xxx
3
a
b
Figure 2. (a) An initial model of 4 docked into a homology model of the mTOR protein. Key residues predicted to contribute to binding and selectivity versus PI3K isoforms
are included, as are solvent accessible surfaces color coded by lipophilicity (green = lipophilic, pink = hydrophilic, grey–white = in between). Hydrogen bonds between ligand
and protein are shown as hashed lines with cylinders. The distance between the carbon on the (S)-3-methyl-morpholine and the closest Trp2239 carbon is shown in green. (b)
Model of an alternate conformation of compound 4 docked into a homology model of the mTOR protein. The distance between the carbon on the (S)-3-methyl-morpholine
and the closest Trp2225 carbon is shown in green.
O
R1
N
Cl
N
Cl
N
Cl
N
b
N
N
N
N
N
a
NH
N
N
N
+
N
H
N
N
O
Cl
Cl
Cl
R1
5
6a R1 = Me
6b
7a R1 = Me
7b
8
1 = Et
R
1 = Et
(theobromine)
R
O
B
O
O
N
H
N
Cl
R2
N
R2
N
H
R1
N
R1
N
R1
N
10
c
d
N
N
N
N
N
N
N
Cl
Cl
O
7a R1 = Me
7b
9a R1 = Me, R2 = (S)-3-methyl-morpholine
N
H
N
H
11a-f
R
1 = Et
R
R
R
1 = Et, R2 = (S)-3-methyl-morpholine
1 = Me, R2 = morpholine
9b
9c
9d
1 = Me, R2 = (R)-3-methyl-morpholine
9e R1 = Me, R2 = (S)-3-ethyl-morpholine
9f R1 = Me, R2 = homomorpholine
Scheme 1. Reagents and conditions: (a) K2CO3, R1I, acetone, 60 °C, 13% yield; (b) POCl3, N,N-diethylaniline, 130 °C, 18 h, 10–15% yield; (c) Morpholine derivatives, DIPEA,
DMF, 60 °C, 70–85% yield; (d) (PPh3)4Pd, KOAc, Na2CO3, ACN, H2O, 7–50% yield.
O
O
O
N
O
N
B
O
O
N
H
N
H
N
Me
N
Me
N
Me
N
10
a
b
N
N
N
HO
N
N
O
HO
N
N
N
Cl
N
Cl
N
H
N
H
9a
12
13
Scheme 2. Reagents and conditions: (a) (i) 5 equiv LDA, THF, ꢁ78 °C, 1 h, (ii) excess acetone. ꢁ78 °C to rt. Overall 60% yield; (b) (PPh3)4Pd, KOAc, Na2CO3, ACN, H2O, 45% yield.
Following the same methodology illustrated in Schemes 1 and
2, N-9-alkylated imidazolopyrimidines 2–4 and 14 were prepared
using the major alkylated products, 6a and 6b, as starting interme-
diates. The N-5-methyl-pyrrolo[3,2-d]pyrimidinyl analogs 15–16
Please cite this article in press as: Lee, W.; et al. Bioorg. Med. Chem. Lett. (2013), http://dx.doi.org/10.1016/j.bmcl.2013.07.027

4
W. Lee et al. / Bioorg. Med. Chem. Lett. xxx (2013) xxx–xxx
O
O
O
Me
Me
Me
N
N
N
a
b
N
OH
NH2
NH2
NH
N
N
N
H
O
R1
R1
R1
19a R1 = H
18a R1 = H
17a R1 = H
19b R1 = Me
18b R1 = Me
17b R1 = Me
Cl
N
R2
N
Me
Me
N
N
N
N
c
d
N
N
Cl
O
R1
R1
e
O
N
N
H
B
O
O
H
20a R1 = H
N
H
10
N
H
21a R1 = H, R2 = (S)-3-methylmorpholine
21b R1 = H, R2 = (S)-3-ethylmorpholine
21c R1 = Me, R2 = (S)-3-ethylmorpholine
20b R1 = Me
21d R1 = H, R2
=
O
N
Scheme 3. Reagents and conditions: (a) (i) fuming HNO3, H2SO4, 0 °C, (ii) SOCl2, NH4OH, 0°C to rt, (iii) 10% Pd/C, EtOH, 60 psi H2. Overall 46–51% yield; (b) CDI, ACN, reflux
98% yield; (c) POCl3, N,N-diethylaniline (3:2 v/v), 130 °C. 60–72% yield; (d) morpholine derivatives, DIPEA, DMF, 60 °C. 86–93% yield; (e) (PPh3)4Pd, KOAc, Na2CO3, ACN, H2O,
54–78% yield.
Table 2
mTOR potency and PI3K
a
and PI3K
c
selectivities of N-7-Me versus N-9-Me-imidazolopyrimidinesa,b
O
R3
N
7
N
N
R1
N
N
O
9
R2
N
H
N
H
Compd
R1
(N-9) R2
(N-7) R3
mTOR (nM)
PI3K
390
7000
370
>10,000
380
>10,000
a
(nM)
PI3K
a
/mTOR
PI3Kd (nM)
PI3Kd/mTOR
3
13
4
11a
14
11b
CH(CH3)2OH
CH(CH3)2OH
H
H
H
H
Me
7
30
12
30
8
55ꢀ
490
7300
890
>10,000
630
>10,000
68ꢀ
Me
Me
Et
230ꢀ
30ꢀ
240ꢀ
72ꢀ
Me
Et
>330ꢀ
46ꢀ
>330ꢀ
78ꢀ
40
>250ꢀ
>250ꢀ
a
Assays are described in Refs. 9b and 11.
Values are at least two separate determinations with typical variations of less than 30%.
b
were prepared from readily available 2,4-dichloro-5H-pyrrolo[3,2-
d]pyrimidine according to the synthetic route described in
Scheme 1.
polarization assay against PI3Ka and PI3Kd as previously
described.9b,11 Cellular potency was measured by the inhibition
of phosphorylation of two substrates of mTOR, Akt (serine 473)
and p70S6K, in a PTEN-null prostate carcinoma NCI-PC3 cell line.
Antiproliferative activity was determined in a 3-day viability assay
using NCI-PC3 and MCF7 neo/HER2 cell lines.11
In order to prepare pyrazolo[4,3-d]pyrimidine analogs of 21a–d,
we condensed pyrazolo intermediates 18 with CDI to give pyrimid-
inediones 19, which were treated with phosphoryl chloride in the
presence of N,N-diethylaniline to provide dichloropyrimidines 20
(Scheme 3). The pyrazolo intermediates 18a and 18b wereprepared
following a literature method from readily available 1-methyl-1H-
pyrazole-5-carboxylic acid and 1,3-dimethyl-1H-pyrazole-5-car-
boxylic acid, respectively.14 A typical SNAr reaction of 20 with
morpholine derivatives followed by Suzuki coupling with the
pinacol ester of N-ethyl-N0-phenyl urea 10 led to the desired
N-1-methyl-pyrazolo[4,3-d]pyrimidines.
A comparison of a small set of N-7 versus N-9 alkylated imidaz-
olopyrimidines is shown in Table 2. N-7 alkylated compounds 13,
11a, and 11b displayed a modest 2.5- to 5-fold decrease in mTOR
potency relative to their N-9 alkylated counterparts 3, 4, and 14,
respectively. The hoped-for increase in mTOR potency of 11a by
the presence of the N-7-imidazolopyrimidine methyl group forcing
the morpholine (S)-3-methyl toward Tyr2225 was not realized
(Supplementary data Fig. S1b). However, exquisite selectivities to-
All final compounds were tested for inhibition of mTOR, PI3K
a,
ward the PI3Ka and PI3Kd isoforms were maintained. The N-7-imi-
and PI3Kd kinases. In vitro mTOR potency was measured with a
LanthaScreen FRET assay against recombinant mTOR kinase do-
main; in vitro PI3K potency was determined using a fluorescence
dazolopyrimidine methyl group may have altered the allowed
torsions of the bond connecting the morpholine to the imidazolo-
pyrimidine, subtly affecting the ability of the morpholine oxygen
Please cite this article in press as: Lee, W.; et al. Bioorg. Med. Chem. Lett. (2013), http://dx.doi.org/10.1016/j.bmcl.2013.07.027

W. Lee et al. / Bioorg. Med. Chem. Lett. xxx (2013) xxx–xxx
5
Table 3
Effect of morpholine derivatives with N-7-methyl-imidazopyrimidinea,b
R1
N
Me
N
N
O
N
N
H
N
H
Potency/selectivity
PI3K (IC50 nM)
(fold)
Target modulationc
Pp70S6K IC50
Proliferation
Compd R1
mTor (Ki
nM)
a
PI3Kd (IC50 nM)
(fold)
p-Akt
NCI-PC3 EC50
(nM)
MCF7neo/Her2MDA EC50
(nM)
IC50(nM)
(nM)
O
(S)
N
11a
30
150
69
>10,000 (>330ꢀ)
>10,000 (>330ꢀ)
630
2200
290
1300
O
11c
N
O
(R)
N
11d
11e
11f
5800 (84ꢀ)
4000 (58ꢀ)
650
69
730
153
O
(S)
N
4
>10,000 (>2600ꢀ)
6400 (53ꢀ)
>10,000 (>2600ꢀ)
7300 (61ꢀ)
737
518
O
120
7200
8900
N
a
b
c
Assays are described in Refs. 9b and 11.
Values are at least two separate determinations with typical variations of less than 30%.
Assayed in NCI-PC3 cell line.
to form an optimal hydrogen bond to the backbone NH of Val2240
of mTOR. The modest 2.5- to 5-fold drop in potency for these com-
pounds were difficult to explain with a homology model. The ter-
tiary alcohol at C-8 of 11a was equipotent to N-9-Me analog 13
and had no effect on mTOR potency. Modeling predicted that these
moieties on C-8 were largely solvent exposed. For 11a and 11b, a
entropy of the system and reducing the potency. The (R)-3-methyl
enantiomer 11d showed a two fold drop in mTOR potency relative
to the S-enantiomer 11a. To accommodate the morpholine (R)-3-
methyl into the lipophilic pocket created by Tyr2225, the morpho-
line may have twisted slightly out-of-plane, altering the optimal
hydrogen bond vector from the ring oxygen to the backbone NH
of Val2240 of mTOR. Homomorpholine 11f showed more than a
four fold decrease in mTOR activity. An increase in steric clash be-
tween the larger homomorpholine and the N-7-methyl may have
resulted in the homomorpholine ring adopting an out-of-plane
conformation, which would be less tolerated by mTOR. Introduc-
tion of an (S)-3-ethyl substituent on the morpholine (11e) resulted
in a 7.5-fold gain in mTOR inhibition relative to the (S)-3-methyl
analog 11a, leading to a single digit nanomolar inhibitor against
mTOR (Ki = 4 nM) while keeping the IC50> 10,000 nM for both
PI3K isoforms. N-7-methyl substitution restricted the morpho-
line-to-imidazolopyrimidine torsion angle such that the (S)-3-
ethyl moiety could fully pack into the hydrophobic space adjacent
to Tyr2225 (Fig. 3). Compound 11e also showed greater than a nine
fold potency boost over 11a in the cellular p-AKT (IC50 = 69 nM)
and p-p70S6K (IC50 = 153 nM) assays, and acceptable anti-prolifer-
ative potency in the submicromolar range.
dramatic reduction in PI3Ka and PI3Kc potencies was observed,
with IC50> 10,000 nM for both isoforms. This was consistent with
the hypothesis from modeling that an additive effect between a
small lipophilic group at the N-7 position on the imidazolopyrim-
idine and a (S)-3-methyl on the morpholine could result in a selec-
tivity gain towards PI3 kinases. The N-7 methyl has packed tightly
against the face of Trp2239 in mTOR, whereas a non-optimal inter-
action with the edge of Trp812 (PI3K numbering) was seen in the
corresponding region in PI3K. Due to its favorable lead-like proper-
ties (clogP = 2.2, TPSA = 97, ligand efficiency 0.36), we focused on
11a as a starting point for evaluating the effects of morpholine
substitution.
The mTOR potencies and PI3K selectivities among N-7 methyl-
imidazolopyrimidine and substituted morpholine analogs are
shown in Table 3. Compound 11c, lacking a methyl group at the
3-position of the morpholine, was five fold less potent against
mTOR than 11a. As mentioned previously, the imidazolopyrimi-
dine N-7 methyl may have projected the morpholine 3-methyl
group into a space adjacent to Tyr2225 in the homology model
(Fig. 2b), partly filling a hydrophobic cavity in the area. Without
a 3-methyl substituent occupying the lipophilic pocket created
by Tyr2225, the unsubstituted morpholine in 11c could freely twist
to adopt other low energy conformations, effectively increasing the
We then extended our N-7-alkylated imidazolopyrimidine
hypothesis by replacing the heterocyclic core with either
pyrrolo[3,2-d]pyrimidine or pyrazolo[4,3-d]pyrimidine (Table 4).
These similar ring systems have been reported as potent PI3K
and/or mTOR inhibitors.8,16 Additionally, the pyrazolo ring
nitrogen off the pyrazolopyrimidine can orient more toward sol-
vent and may decrease a desolvation penalty on binding. In the
Please cite this article in press as: Lee, W.; et al. Bioorg. Med. Chem. Lett. (2013), http://dx.doi.org/10.1016/j.bmcl.2013.07.027

6
W. Lee et al. / Bioorg. Med. Chem. Lett. xxx (2013) xxx–xxx
Figure 3. Compound 11e docked into a homology model of mTOR. Predicted ligand–protein H-bonds are shown in green, with distances between heavy atoms included. A
solvent accessible surface of the active site is present, color coded by molecular electrostatic potential calculated using the protein active site atoms (red = negative;
blue = positive; white = neutral).15 Key residues hypothesized to contribute to mTOR potency and selectivity against PI3K isoforms have been included.
Table 4
mTOR, PI3K
a
, and PI3Kd potencies and selectivitites of N-methyl-pyrrolo[3,2-d]pyrimidines and N-methyl-pyrazolo[4,3-d]pyrimimdinesa,b
R1
Me
N
N
X
O
N
R2
N
H
N
H
Potency/selectivity
Target modulationc
Proliferation
Compd R1
R2
H
X
mTor(Ki
nM)
PI3Ka (Ki nM)
(fold)
PI3Kd (Ki nM)
(fold)
p-Akt IC50
(nM)
p-p70S6K
IC50(nM)
NCI-PC3 EC50
(nM)
MCF7neo/Her2MDA EC50
(nM)
O
>10,000
(>518ꢀ)
>10,000
(>518ꢀ)
(S)
15
CH 19
77
11
8
110
40
47
15
8
1900
280
420
130
86
9000
420
N
O
>10,000
(>1800ꢀ)
>10,000
(>1800ꢀ)
(S)
16
H
CH
N
6
N
O
>10,000
(>1290ꢀ)
>10,000
(>1290ꢀ)
(S)
21a
21b
21c
21d
H
8
760
N
O
>10,000
(>8300ꢀ)
(S)
H
N
1
3400 (>2800ꢀ)
5800 (2900ꢀ)
5135 (180 ꢀ)
8
260
N
O
(S)
Me
H
N
2
>10,000 (>5300)
4
160
N
O
N
28
5147 (180x)
340
270
2900
1700
N
a
b
c
Assays are described in Refs. 9b and 11.
Values are at least two separate determinations with typical variations of less than 30%.
Assayed in NCI-PC3 cell line.
pyrrolo[3,2-d]pyrimidine series, (S)-3-Me morpholine 15 and(S)-3-
Et morpholine 16 had comparable biochemical potency to 11a and
11e, respectively. The pyrazolo[4,3-d]pyrimidine series displayed
increased mTOR inhibition, with 21a and 21b displaying single
digit nanomolar potency in the mTOR biochemical assay and
desirable selectivity against the PI3K
a
(>2800ꢀ) and PI3Kd
Please cite this article in press as: Lee, W.; et al. Bioorg. Med. Chem. Lett. (2013), http://dx.doi.org/10.1016/j.bmcl.2013.07.027

W. Lee et al. / Bioorg. Med. Chem. Lett. xxx (2013) xxx–xxx
7
2. Liu, Q.; Thoreen, C.; Wang, J.; Sabatini, D.; Gray, N. S. Drug Discovery Today: Ther.
Strategies 2009, 6, 47.
3. O’Reilly, K. E.; Rojo, F.; She, Q. B.; Solit, B.; Mills, G. B.; Smith, D.; Lane, H.;
Hofmann, F.; Hicklin, D. J.; Ludwig, D. L.; Baselga, J. B.; Rosen, N. Cancer Res.
2006, 66, 1500.
(>8300ꢀ) isoforms. We then explored the SAR of pyrazolo[4,3-
d]pyrimidines by preparing a bridged morpholine 21d (S,S isomer)
to evaluate the effect of a constrained morpholine packing against
the face of Tyr2225. An N-7-methyl substituent with a bridged
morpholine was tolerated (Ki = 28 nM) although mTOR potency
was reduced 3.7- and23-fold relative to 21a and 21b, respectively.
This reduction may be due a less tight packing of the extra
methylene unit against Tyr2225. This bridged substitution was also
reported in a potent morpholine derivative when combined with
N-9 alkylated pyrazolopyrimidines.7a Substitution of C-3 on the
pyrazolo[4,3-d]pyrimidine was also probed. The C-3 methyl
analog 21c was equipotent to 21b and displayed a 2900-fold
selectivity toward PI3K. Furthermore, 21c showed exceptional
cellular potencies in single digit nanomolar range [p-AKT IC50 = 4
4. Guertin, D. A.; Sabatini, D. M. Sci. Signaling 2009, 2, pe 24.
5. Fasolo, A.; Sessa, C. Expert Opin. Invest. Drugs 2011, 20, 381.
6. (a) Foukas, L. C.; Claret, M.; Pearce, W.; Okkenhaug, K.; Meek, S.; Peskett, E.;
Sancho, S.; Smith, A. J. H.; Withers, D. J.; Vanhaesebroeck, B. Nature 2006, 441,
366; (b) Knight, Z. A.; Gonzalez, B.; Feldman, M. E.; Zunder, E. R.; Goldenberg, D.
D.; Williams, O.; Loewith, R.; Stokoe, D.; Balla, A.; Toth, B.; Balla, T.; Weiss, W.
A.; Williams, R. L.; Shokat, K. M. A. Cell 2006, 125, 733.
7. (a) Zask, A.; Kaplan, J.; Verheijen, J. C.; Richard, D. J.; Curran, K.; Brooijmans, N.;
Bennett, E. M.; Toral-Barza, L.; Hollander, I.; Ayral-Kaloustian, S.; Yu, K. J.Med.
Chem. 2009, 52, 7942; (b) Nowak, P.; Cole, D. C.; Brooijmans, N.; Bursavich, M.
G.; Curran, K. J.; Ellingboe, J. W.; Gibbons, J. J.; Hollander, I.; Hu, Y.; Kaplan, J.;
Malwitz, D. J.; Toral-Barza, L.; Verheijen, J. C.; Zask, A.; Zhang, W.-G.; Yu, K.
J.Med. Chem. 2009, 52, 7081; (c) Zask, A.; Verheijen, J. C.; Curran, K.; Kaplan, J.;
Richard, J. L.; Nowak, P.; Malwitz, D. J.; Brooijmans, N.; Bard, J.; Svenson, K.;
Lucas, J.; Toral-Barza, L.; Zhang, W.-G.; Hollander, I.; Gibbons, J. J.; Abraham, R.
T.; Ayral-Kaloustian, S.; Mansour, T. S.; Yu, K. J.Med. Chem. 2009, 52, 5013; (d)
Verheijen, J. C.; Yu, K.; Toral-Barza, L.; Hollander, I.; Zask, A. Bioorg. Med.
Chem.Lett. 2010, 20, 375; (e) Verheijen, J. C.; Richard, D. J.; Curran, K.; Kaplan, J.;
Lefever, M.; Nowak, P.; Malwitz, D. J.; Brooijmans, N.; Toral-Barza, L.; Zhang,
W.-G.; Lucas, J.; Hollander, I.; Ayral-Kaloustian, S.; Mansour, T. S.; Yu, K.; Zask,
A. J.Med. Chem. 2009, 52, 8010; (f) Curran, K. J.; Verheijen, J. C.; Kaplan, J.;
Richard, D. J.; Toral-Barzab, L.; Hollander, I.; Lucas, J.; Ayral-Kaloustiana, S.; Yu,
K.; Zask, A. Bioorg.Med. Chem. Lett. 2010, 20, 1440.
nM and p-p70S6K IC50
potencies [NCI-PC3EC50 = 86 nM
EC50 = 160 nM].
8
8 nM] and good anti-proliferative
and MCF7neo/Her2MDA
A representative compound (21a) from the N-Me pyrazolo[4,3-
d]pyrimidine series was selected for a mouse pharmacokinetic
study. The mouse iv profile showed 21a possessed acceptably
low plasma clearance and
a low volume of distribution
(Cl = 13 mL/min/kg, Vss = 0.36 L/kg). After a 25 mg/kg oral dose,
21a was found to suffer from non-optimal bioavailability
(%F = 4.4; dosed orally as a suspension in MCT). The cause for
low bioavailability was not permeability or transporter mediated
efflux,since a MDCK (Madin–Darby canine kidney epithelial cells)
cellular assay of 21a showed good permeability with minimal ef-
flux potential (Papp B–A = 8.6 ꢀ 10ꢁ6 cm/s; net flux ratio = 0.55).
Limitation of oral absorption may be due to low aqueous solubility,
given that the thermodynamic aqueous solubility of 21a was mod-
8. (a) Mortensen, D. S.; Sapienza, J.; Lee, B. G. S.; Perrin-Ninkovic, S. M.; Harris,
R.; Shevlin, G.; Parnes, J. S.; Whitefield, B.; Hickman, M.; Khmbatta, G.;
Bisonette, R. R.; Peng, S.; Gamez, J. C.; Leistend, J.; Krishna, N.; Fultz, K. E.;
Sankar, S. Bioorg.Med. Chem. Lett. 2013, 23, 1588; (b) Takeuchi, C. S.; Kim, B. G.;
Blazey, C. M.; Ma, S.; Johnson, H. W. B.; Anand, N. K.; Arcalas, A.; Bail, T. G.;
Bur, C. A.; Cannoy, J.; Epshteyn, A. J.; Lara, K.; Lee, M. S.; Wang, L.; Leah, J. W.;
Nuss, J. M.; Aay, N.; Aoyama, R.; Foster, P.; Lee, J.; Lehoux, I.; Munagala, N.;
Plonowski, A.; Rajan, S.; Woolfrey, J.; Yamaguchi, K.; Lamb, P.; Miller, N. J.Med.
Chem. 2013, 56, 2218; (c) Liu, K. K. C.; Bailey, S.; Dinh, D. M.; Lam, H.; Li, C.;
Wells, P. A.; Yin, M.-J.; Zou, A. Bioorg. Med. Chem.Lett. 2012, 22, 5114; (d)
Peterson, E. A.; Boezio, A. A.; Andrews, P. S.; Boezio, C. M.; Bush, T. L.; Cheng,
A. C.; Choquette, D.; Coats, J. R.; Colletti, A. E.; Copeland, K. W.; DuPont, M.;
Graceffa, R.; Grubinska, B.; Kim, J. L.; Lewis, R. T.; Liu, J.; Mullady, E.;
Potashman, M. H.; Romero, K.; Shaffer, P. L.; Stanton, M. K.; Stellwagen, J. C.;
Teffera, Y.; Yi, S.; Cai, T.; La, D. S. Bioorg.Med. Chem. Lett. 2012, 22, 4967; (e)
Finlay, M. R. V.; Buttar, D.; Critchlow, S. E.; Dishington, A. P.; Fillery, S. M.;
Fisher, E.; Glossop, S. C.; Graham, M. A.; Johnson, T.; Lamont, G. M.; Mutton, S.;
Perkins, P.; Pike, K. G.; Slater, A. M. Bioorg.Med. Chem. Lett. 2012, 22, 4163; (f)
Liu, Q.; Wang, J.; Kang, S. A.; Thoreen, C. C.; Hur, W.; Ahmed, T.; Sabatini, D.
M.; Gray, N. S. J. Med.Chem. 2011, 54, 1473.
9. (a) Estrada, A. A.; Shore, D. J.; Blackwood, E.; Chen, Y.-H.; Deshmukh, G.; Ding,
X.; DiPasquale, A. G.; Epler, J. A.; Friedman, L. S.; Koehler, M. F. T.; Liu, L.; Malek,
S.; Nonomiya, J.; Ortwine, J. F.; Pei, Z.; Sideris, S.; St-Jean, F.; Trinh, L.; Truong,
T.; Lyssikatos, J. P. J.Med. Chem. 2013, 56, 3090; (b) Pei, Z.; Blackwood, E.; Liu, L.;
Malek, S.; Belvin, M.; Koehler, M. F. T.; Ortwine, D. F.; Chen, H.; Cohen, F.;
Kenny, J. R.; Bergeron, P.; Lau, K.; Ly, C.; Zhao, X.; Estrada, A. A.; Truong, T.;
Epler, J. A.; Nonomiya, J.; Trinh, L.; Sideris, S.; Lesnick, J.; Bao, L.; Vijapurkar, U.;
Mukadam, S.; Tay, S.; Deshmukh, G.; Chen, Y. H.; Ding, X.; Friedman, L. S.;
Lyssikatos, J. P. ACSMed. Chem. Lett. 2013, 4, 103; (c) Koehler, M. F. T.; Bergeron,
P.; Blackwood, E.; Bowman, K. K.; Chen, Y. H.; Deshmukh, G.; Ding, X.; Epler, J.;
Lau, K.; Lee, L.; Liu, L.; Ly, C.; Malek, S.; Nonomiya, J.; Oeh, J.; Ortwine, D. F.;
Sampath, D.; Sideris, S.; Trinh, L.; Truong, T.; Wu, J.; Pei, Z.; Lyssikatos, J. P.
J.Med. Chem. 2012, 55, 10958; (d) Cohen, F.; Bergeron, P.; Blackwood, E.;
Bowman, K. K.; Chen, Y. H.; DiPasquale, A. G.; Epler, J. A.; Koehler, M. F. T.; Lau,
K.; Lewis, C.; Liu, L.; Ly, C. Q.; Malek, S.; Nonomiya, J.; Ortwine, D. F.; Pei, Z.;
Robarge, K. D.; Sideris, S.; Trinh, L.; Truong, T.; Wu, J.; Zhao, X.; Lyssikatos, J. P.
J.Med. Chem. 2011, 54, 3426.
erate at 63.2 lM (pH 7.4). Solubility-related physicochemical fac-
tors17 such as compound dissolution in biorelevant media,
basicity (pKa), or particle size may also be contributing to its low
bioavailability. Compound 21a was observed to have a partial crys-
talline form, affecting its solubility due to different dissolution
rates and particle sizes (data notshown).
In conclusion, guided by structure-based design, we were able
to synthesize and evaluate a variety of N-7-Me-imidazolopyrimi-
dines, N-5-Me-pyrrolo[3,2-d]pyrimidines, and N-1-Me-pyrazol-
o[4,3-d]pyrimidines for mTOR inhibition and selectivity against
PI3K. We identified N-1-Me-pyrazolo[4,3-d]pyrimidine as a lead
series with exquisite mTOR potency, high selectivity versus PI3K
a
and PI3Kd, and excellent cellular potency. Unfortunately, poor oral
bioavailability precluded further development of this series. Subse-
quent to this work, Xray crystal structures of the mTOR kinase
domain have been reported.18 Our homology model of the protein
matched up quite well to the observed residue positions in the ac-
tive site, validating the useof the model in our structure-based de-
sign studies (Supplementary data Fig. S2).
10. Compound was disclosed in: Castanedo, G.; Chuckowree, I.; Folkes, A.;
Sutherlin, D. P.; Wan, N. C. WO 2009146406, 2009.
Acknowledgments
11. Sutherlin, D. P.; Sampath, D.; Berry, M.; Castanedo, G.; Chang, Z.; Chuckowree,
I.; Dotson, J.; Folkes, A.; Friedman, L.; Goldsmith, R.; Heffron, T.; Lee, L.; Lesnick,
J.; Lewis, C.; Mathieu, S.; Nonomiya, J.; Olivero, A.; Pang, J.; Prior, W.; Salphati,
L.; Sideris, S.; Tian, Q.; Tsui, V.; Wan, N. C.; Wang, S.; Wiesmann, C.; Wong, S.;
Zhu, B.-Y. J.Med. Chem. 2010, 53, 1086.
The authors gratefully thank the purification and analytical
groups within Genentech Small Molecule Drug Discovery for their
support. We also thank Dr. James Crawford for helpful discussions
and critical review of the manuscript.
12. Uretskaya, G. Ya.; Rybinka, E. I.; Men’shikov, G. P. Zh. Obshch. Khim. 1960, 30,
327.
13. Lebraud, H.; Cano, C.; Carbain, B.; Hardcastle, I. R.; Harringon, R. W.; Griffin, R.
J.; Golding, B. T. Org. Biomol.Chem. 1874, 2013, 11.
14. Terrett, N. K.; Bell, A. S.; Brown, D.; Ellis, P. Bioorg. Med. Chem. Lett. 1819, 1996,
6.
15. The surface is color coded by a screened (localized) electrostatic potential that
uses a charge density obtained by solving the Poisson–Boltzmann equation
using MMFF94 partial charges on the protein. Default values for the color scale
range were used, as provided in the MOE software (www.chemcomp.com).
16. (a) Chen, Z.; Venkatesan, A. M.; Dehnhardt, C. M.; Ayral-Kaloustian, S.;
Brooijmans, N.; Mallon, R.; Feldberg, L.; Hollander, I.; Lucas, J.; Yu, K.; Kong,
F.; Mansour, T. S. J.Med. Chem. 2010, 53, 3169; (b) Gilbert, A. M.; Nowak, P.;
Brooijmans, N.; Bursavich, M. G.; Dehnhardt, C.; Santos, E. D.; Feldberg, L. R.;
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.07.
027.
References and notes
1. Laplante, M.; Sabatini, D. M. Cell 2012, 149, 274.
Please cite this article in press as: Lee, W.; et al. Bioorg. Med. Chem. Lett. (2013), http://dx.doi.org/10.1016/j.bmcl.2013.07.027

8
W. Lee et al. / Bioorg. Med. Chem. Lett. xxx (2013) xxx–xxx
Hollander, I.; Kim, S.; Lombardi, S.; Park, K.; Venkatesan, A. M.; Mallon, R.
17. Thomas, V. H.; Bhattachar, S.; Hitchingham, L.; Zocharski, P.; Naath, M.
Surendran, N,; Stoner, C. L., El-Kattan, A Expert Opin. Drug Metabol. Toxicol.
2006, 2, 591.
18. Yang, H.; Rudge, D. G.; Koos, J. D.; Vaidialingam, B.; Yang, H. J.; Pavletich, N. P.
Nature 2013, 497, 217.
Bioorg.Med. Chem. Lett. 2010, 20, 636; (c) Hayakawa, M.; Kaizawa, H.;
Moritomo, H.; Kawaguchi, K.; Koizumi, T.; Yamano, M.; Matsuda, K.; Okada,
M.; Ohta, M. WO 2001083456, 2001.; (d) Zask, A.; Verheijen, J. C.; Richard, D. J.
Expert Opin. Ther.Patents 2011, 21, 1109.
Please cite this article in press as: Lee, W.; et al. Bioorg. Med. Chem. Lett. (2013), http://dx.doi.org/10.1016/j.bmcl.2013.07.027