Identification and structure-activity relationship of 2-morpholino 6-(3-hydroxyphenyl) pyrimidines, a class of potent and selective PI3 kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2010.10.021

PI3 Kinases are a family of lipid kinases mediating numerous cell processes such as proliferation, migration, and differentiation. The PI3 kinase pathway is often de-regulated in cancer through PI3Kα overexpression, gene amplification, mutations, and PTEN

Full text of DOI:10.1016/j.bmcl.2010.10.021

Bioorganic & Medicinal Chemistry Letters 20 (2010) 6895–6898
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification and structure–activity relationship of 2-morpholino
6-(3-hydroxyphenyl) pyrimidines, a class of potent and selective PI3 kinase
inhibitors
Sabina Pecchi a, , Paul A. Renhowe a, Clarke Taylor c, , Susan Kaufman a, Hanne Merritt a, Marion Wiesmann b,
⇑
Kevin R. Shoemaker a, Mark S. Knapp a, Elizabeth Ornelas a, Thomas F. Hendrickson d, , Wendy Fantl e,
,
Charles F. Voliva a
a Novartis Institutes for Biomedical Research, Emeryville, CA, USA
b Novartis Institutes for Biomedical Research, Basel, Switzerland
c Lycera Corp., Plymouth, MI, USA
d Supergen Inc., Salt Lake City, UT, USA
e Nodality Inc., South San Francisco, CA, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
PI3 Kinases are a family of lipid kinases mediating numerous cell processes such as proliferation,
migration, and differentiation. The PI3 kinase pathway is often de-regulated in cancer through PI3K
Received 29 July 2010
Revised 2 October 2010
Accepted 5 October 2010
Available online 13 October 2010
a
overexpression, gene amplification, mutations, and PTEN phosphatase deletion. PI3K inhibitors represent
therefore an attractive therapeutic modality for cancer treatment. Herein we describe a novel series of
PI3K inhibitors sharing a pyrimidine core and showing significant potency against class I PI3 kinases in
the biochemical assay and in cells. The discovery, synthesis and SAR of this chemotype are described.
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
2-Morpholino pyrimidines
Phosphoinositol
PI3 Kinase
PI3 Kinase inhibitors
Cancer treatment
Phosphatidylinositol 3-kinases (PI3Ks) comprise a family of
lipid and serine/threonine kinases that catalyze the transfer of
phosphate to the D-30 position of inositol lipids to produce phos-
phoinositol-3-phosphate (PIP), phosphoinositol-3,4-diphosphate
(PIP2), and phosphoinositol-3,4,5-triphosphate (PIP3) that, in turn,
act as second messengers in numerous signaling cascades.1 Class
IA PI3Ks are heterodimers composed of a catalytic p110 subunit
nodal point for many intracellular signaling pathways important
for growth and survival.4,5 Aberrant regulation of PI3K, which often
increases survival through AKT activation (phosphorylation on
S473), is one of the most prevalent events in human cancer.6 This
can occur at multiple levels such as PTEN deletion7, gene amplifica-
tion (PIK3CA and Akt)8, p85
a mutations (and translocation) and
PIK3CA somatic missense mutations.9 These observations support
inhibition of Class IA PI-3 kinases as a potential treatment for a
variety of tumor types and other proliferative diseases10,11,12
In the course of our efforts toward the identification of PI3K
inhibitors we discovered that a pool from a solid phase combinato-
rial library of 2,4,6-trisubstituted pyrimidines showed sub
(
a
, b, and d isoforms) constitutively associated with a regulatory
subunit that can be p85 , p55 , p50 , p85b, or p55 . The Class
IB has one family member, a heterodimer composed of a catalytic
p110 subunit associated with one of two regulatory subunits,
a
a
a
c
c
p101 or p84. Class IA PI3Ks are activated by tyrosine kinases while
Class IB is activated directly by G protein-coupled receptors,
therefore linking upstream receptors with downstream cellular
activities including proliferation, survival, chemotaxis, cellular
trafficking, motility, metabolism, inflammatory and allergic
responses, transcription, and translation.2,3 In many cases, PIP2
and PIP3 recruit AKT to the plasma membrane where it acts as a
H
N
N
HN
4
N
HO
6
N 2 N
O
1
⇑
Corresponding author. Tel.: +1 510 923 7812; fax: +1 510 923 3360.
E-mail address: sabina.pecchi@novartis.com (S. Pecchi).
Present address.
Figure 1. Initial hit in the 2-morpholino-4-amino-6-phenylpyrimidine scaffold,
PI3K IC50 0.031 M.
a
l
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.10.021

6896
S. Pecchi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 6895–6898
Table 1
micromolar activity against PI3Ka. Pool deconvolution and screen-
Position 6- SAR in the 2-morpholino pyrimidine scaffold (IC50 in
l
M)
ing of the singletons led to the identification of compound 1
H
N
N
(Fig. 1), a hit with the remarkable potency of 0.031 M. The com-
pound was also quite selective against protein kinases. Out of 31
tyrosine and serine threonine kinases screened, 24 had IC50
l
HN
N
>10
M. In particular, PDK1 and AKT, important effectors in the PI3
kinase pathway, were not significantly inhibited, with IC50 >25
lM, 4 were between 5 and 10 lM and 3 were between 1 and
6
R
5
l
N
N
O
and >10 lM, respectively. Further analoging in the series was ini-
Ex #
R
PI3K
a
IC50 (n)
Ex #
R
PI3K a IC50 (n)
tially performed through solid phase synthesis (Scheme 1, top).13
1
12
13
14
15
m-OH
o-OH
p-OH
H
0.031 (11)
0.91 (6)
6.5 (2)
9.4 (3)
9.5 (3)
16
17
18
19
20
m-Cl
m-OMe
m-CN
m-COOMe
m-O(CH2)2OH
13 (2)
15 (6)
14 (3)
16 (2)
11 (3)
Resin bound methyl malonate 2 was reacted with aldehydes under
Knoevenagel conditions and the resulting a,b-unsaturated ester 3
was cyclized to the dihydropyrimidinone 4 with pyrazolo guani-
dine in basic conditions. Oxidation to the pyrimidinone 5 and reac-
tion with amines in the presence of PyBOP led to the formation of
2-pyrazolo-4-alkylamino-6-substituted pyrimidines 6. The 2-pyra-
zole group could then be displaced by a different set of amines
under SnAr conditions (7) and after cleavage from the resin and
decarboxylation the desired trisubstituted pyrimidines 8 were
obtained. Later on, a solution phase route was devised for 2-mor-
pholino substituted pyrimidines, allowing the synthesis of larger
amounts of material (Scheme 1, bottom).
m-F
Table 2
Position 4- SAR in the 2-morpholino-6-(30-hydroxyphenyl) pyrimidine series (IC50 in
l
M)
R4
N
Substituted ethyl 3-oxo-3-aryl propanoates 9, prepared
via a modification of a literature route14 were cyclized to the
pyrimidinones15 with commercially available morpholino carbox-
amidine hydrobromide. Conversion to the 4-position triflate 10
and reaction with a variety of primary or secondary amines yielded
the desired 2-morpholino 4-amino 6-substituted pyrimidines 11.16
We initially undertook the exploration of the substitution
around the phenyl group in the 6-position of the pyrimidine ring.
Some of the results are shown in Table 1. It was immediately
apparent that a phenol in the 6-position was a key binding feature.
m-Hydroxy seemed optimal, as shown by the o- and p-hydroxy-
phenyl analogs 12 and 13, with a decrease in potency of 30- and
200-fold, respectively, compared to 1. Removal of the hydroxyl
group (e.g., 14) or its replacement with F (e.g., 15) caused 300-fold
potency loss. A series of m-position substituents did not improve
potency with every compound (e.g., 16–20) being at least 300-fold
less potent than 1.
In the 4-position of the pyrimidine ring numerous substituents
were tolerated (Table 2), with the most favorable being heteroaro-
matic groups with hydrogen bond acceptors (e.g., 1, 25–27).
Benzylic (e.g., 28) or aliphatic amines, (e.g., 29–30) reduced
activity about 100-fold or more. The NH in the 4-position was
not necessary for binding, as apparent comparing example 25 with
its aryloxy analog 27. Data also suggested that most of the affinity
was achieved through position 2- and 6-substituents, as the
4-unsubstituted example 21, was only 10-fold less potent than
N
N
O
OH
Ex # R4
PI3K
a
Ex # R4
PI3Ka
IC50 (n)
IC50 (n)
H
N
OMe
N
H
1
0.031 (11) 26
0.017 (9)
N
N
H
N
N
21
22
23
H
0.26 (6)
1.1 (6)
1.3 (3)
27
28
29
0.05 (6)
2.0 (6)
O
N
H
N
NHMe
NHPh
O
H
0.34 (6)
H
N
OMe
NAc
O
N
24
25
0.19 (6)
30
0.57 (6)
0.4 (7)
H
N
N
N
0.056 (14) 31
H
N
Solid phase synthesis
O
O
O
O
e
a
c
b
O
O
O
NH
R1CHO
O
OMe
+
R1
N
N
N
O
OMe
R1
2
3
4
NR2R3
N
O
OH
N
O
NR2R3
N
O
NR2R3
N
d
f,g
O
O
O
R1
N
N
N
R1
N
N
N
R1
N
NR4R5 R1
N
NR4R5
5
6
7
8
Solution phase synthesis of 2- morpholino derivatives
OTf
NR2R3
N
O
O
NH2+Br-
h, i
j
R
+
N
OEt
R
R
H2N
N
O
N
10
N
N
N
9
11
O
O
Scheme 1. Reagents and conditions: (a) piperidine, AcOH, rt; (b) pyrazologuanidine, NaHCO3, NMP, 50 °C; (c) DDQ, toluene 50 °C; (d) R2R3NH, PyBOP, NMP, rt; (e) R4R5NH,
AcOH, NMP; (f) 95% TFA/H2O; (g) CH3CN/H2O 1:1, 60 °C; (h) Cs2CO3, DMF, 150 °C; (i) N-phenyl bis (trifluoromethane sulfonimide), Et3N, DMAP, CH2Cl2; (j) R2R3NH, Cs2CO3,
BINAP, THF, 60 °C.

S. Pecchi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 6895–6898
6897
Table 3
Position 2- SAR in the 2-morpholino-6-(30-hydroxyphenyl) pyrimidine series (IC50 in
M)
The proposed binding mode of compound 25 in p110
a
is shown
in Figure 2. The model was built based on literature data and a co-
structure of the compound in p110c 17
Given the high homology
between the and isoforms a similar binding mode was ex-
l
.
H
N
a
c
N
pected and this assumption was supported by SAR data. In fact,
biochemical assay data showed only about a 10-fold potency dif-
ference of 25 against the two isoforms.18 The model of 25 in the
HN
N
N
R2
p110a ATP binding site provides a rationale for the observed SAR
trends and especially for the key role of the 2-position substituent.
The morpholine oxygen forms a key productive hydrogen bond
interaction with V851 in the hinge domain of the ATP binding site.
This interaction has been described previously.19,20 The pyrimidine
core provides the appropriate scaffolding to place the phenolic
moiety near the catalytic region, where the OH binds tightly be-
tween D810 and Y836. These interactions achieve maximum effi-
ciency through a nearly coplanar conformation of the central
pyrimidine ring and the position 4- and 6-substituents. Substitu-
ents in position 4- reach toward a partially solvent exposed area,
hence the higher tolerability.
OH
IC50 (n) Ex # R2
Ex # R2
PI3K
a
PI3Ka IC50 (n)
N
N
1
0.031 (11)
5.6 (5)
34
35
16 (4)
15 (7)
O
N
H
N
N
N
N
32
S
S
OMe
O
33
7.8 (3)
36
11 (10)
O
Compounds in this series inhibit class I PI3 kinases and are in
general approximately equipotent against the
a and d isoforms,
while they are about 10-fold less potent against the b and
c
isoforms.
Modulation of the PI3K pathway is confirmed by inhibition of
AKT phosphorylation on S473 in the A2780 cell line (ovarian carci-
noma, PTEN deleted) and by the resulting inhibition of cell prolif-
eration in the same cell line, as shown in Table 4.
In conclusion, we have identified a series of 2-morpholino-6-
hydroxyphenyl pyrimidines which are potent and selective class I
PI3K kinase inhibitors, showing cellular activity both as mecha-
nism modulation (inhibition of AKTS473 phosphorylation) and inhi-
bition of cell proliferation in a cell line with PI3K pathway
deregulation. Additional SAR, especially focused on identifying het-
erocyclic alternatives to the phenol in the 6-position will be re-
ported in due course.
Acknowledgments
The authors wish to acknowledge Yasuki Yamada for the syn-
thesis of the original solid phase library, John Nuss and Alex Harris
for guidance and advice during the early stage of the program.
Figure 2. The proposed binding mode of compound 25 in the class 1A PI3K-
isoform is shown. Key binding interactions with V851, D810, and Y836 are shown
by dotted red lines.
a
Supplementary data
Table 4
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2010.10.021.
Biochemical selectivity profile, cell mechanism modulation and functional effects of
selected PI3K inhibitors from the morpholino pyrimidine scaffold
Ex#
PI3K
IC50
a
PI3Kb
IC50
PI3Kd
IC50
PI3K
IC50
c
pAKT
EC50
Prolif.
EC50
References and notes
1. Vanhaesebroeck, B.; Leevers, S. J.; Ahmadi, K.; Timms, J.; Katso, R.; Driscoll, P.
C.; Woscholski, R.; Parker, P. J.; Waterfield, M. D. Annu. Rev. Biochem. 2001, 70,
535.
1
26
0.031 (11)
0.017 (9)
0.24 (18)
0.24 (23)
0.029 (5)
0.006 (1)
0.57 (2)
nd
0.44 (1)
0.073 (1)
0.94 (4)
0.37 (2)
2. Katso, R.; Okkenhaug, K.; Ahmadi, K., et al Annu. Rev. Cell Dev. Biol. 2001, 17,
615.
IC50 and EC50 in lM. Cell line: A2780 ovarian carcinoma, PTEN deleted.
3. Wymann, M. P.; Zvelebil, M.; Laffargue, M. Trends Pharmacol. Sci. 2003, 24, 366.
4. Kohn, A. D.; Takeuchi, F.; Roth, R. A. J. Biol. Chem. 1996, 271, 21920.
5. Andjelkovic, M.; Alessi, D. R.; Meier, R., et al J. Biol. Chem. 1997, 272, 31515.
6. Liu, P.; Cheng, H.; Roberts, T. M.; Zhao, J. J. Nat. Rev. Drug Disc. 2009, 8, 627.
7. Ali, I. U.; Schriml, L. M.; Dean, M. J. Natl. Cancer Inst. 1999, 91, 1922.
8. Cross, D. A.; Alessi, D. R.; Cohen, P., et al Nature 1995, 378, 785.
9. Samuels, Y.; Wang, Z.; Bardelli, A., et al Science 2004, 304, 554.
10. Kong, D.; Yamori, T. Cancer Sci. 2008, 99, 1734.
11. Nuss, J. M.; Tsuhako, A. L.; Anand, N. K. In Annu. Rep. Med. Chem.; Macor, J. E.,
Ed.; Academic Press: Oxford, 2009; Vol. 44, pp 339–351.
12. Workman, P.; Clarke, P. A.; Raynaud, F. I.; Van Montfort, R. L. M. Cancer Res.
2010, 70, 2146.
13. Gordeev, M. F.; Patel, D. V.; Wu, J.; Gordon, E. M. Tetrahedron Lett. 1996, 37,
4643.
14. Kim, J.-W. PCT Int. Appl. WO04048369, 2004.
15. Gueremy, C.; Audiau, F.; Renault, C.; Benavides, J.; Uzan, A.; Le Fur, G. J. Med.
Chem. 1986, 29, 1394.
compound 1. Although, in general, this series suffered from poor
solubility, 4-position variations allowed some modulation of the
physicochemical properties. As an example, compared to com-
pound 1 (<0.1 lM), compounds 26, 29, and 31 were at least 50-fold
more soluble.
SAR around position 2- (Table 3) underlined the unique role of
morpholine in binding.
The replacement with S, SO, or NH (e.g., 32–34) or removal of
the cyclic constraint (e.g., 35) were, in fact, not tolerated and steric
bulk on both sides of the morpholine oxygen also caused a signif-
icant (>300-fold, e.g., 36) potency loss.

6898
S. Pecchi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 6895–6898
16. Spectral data for compound 1: 3-(6-(1H-indazol-5-ylamino)-2-morpholinopyr-
imidin-4-yl)phenol hydrochloride. 1H NMR (acetonitrile-d3/D2O, 300 MHz): d
ppm 8.09 (s, 1H), 8.03 (br s, 1H), 7.61 (d, 1H, J = 7.8 Hz), 7.55 (br m, 1H), 7.38
(app. t, 1H, J = 7.8 Hz), 7.17 (br d, 1H, J = 7.8 Hz), 7.10 (br s, 1H), 7.06 (d, 1H,
J = 8.7 Hz), 6.42 (br s, 1H), 3.75 (br s, 8H); LCMS m/z (%): 389.2 (M+H).
of the public domain p110
protein-compound interactions are retained between the p110
the p110 model and no new interactions are expected with the other PI3K
Class I isoforms.
18. PI3K IC50 is 0.55
a
apo-structure (PDB deposition code: 2RD0). All key
c
structure and
a
c
lM for 25.
17. The model for compound in p110
the compound 25 coordinates from a Novartis structure of the compound
bound to PI3K p110 (PDB accession code: 3P2B) with the aligned coordinates
a
was built in Maestro 8.0.308 after merging
19. Walker, E. H.; Pacold, M. E.; Perisic, O., et al Mol. Cell 2000, 6, 909.
20. Knight, Z. A.; Chiang, G. G.; Alaimo, P. G.; Kenski, D. M.; Ho, C. B.; Coan, K.;
Abraham, R. T.; Shokat, K. M. Bioorg. Med. Chem. 2004, 12, 4749.
c