Allosteric Akt (PKB) inhibitors: Discovery and SAR of isozyme selective inhibitors

Article abstract of DOI:10.1016/j.bmcl.2004.11.011

This letter describes the development of two series of potent and selective allosteric Akt kinase inhibitors that display an unprecedented level of selectivity for either Akt1, Akt2 or both Akt1/Akt2. An iterative analog library synthesis approach quickly provided a highly selective Akt1/Akt2 inhibitor that induces apoptosis in tumor cells and inhibits Akt phosphorylation in vivo.

Full text of DOI:10.1016/j.bmcl.2004.11.011

Bioorganic & Medicinal Chemistry Letters 15 (2005) 761–764
Allosteric Akt (PKB) inhibitors: discovery and SAR of
isozyme selective inhibitors
Craig W. Lindsley,a,* Zhijian Zhao,a William H. Leister,a Ronald G. Robinson,b
Stanley F. Barnett,b Deborah Defeo-Jones,b Raymond E. Jones,b George D. Hartman,a
Joel R. Huff,a Hans E. Huberb and Mark E. Duggana
aDepartment of Medicinal Chemistry, Technology Enabled Synthesis Group, Merck Research Laboratories,
Merck & Co., PO Box 4, West Point, PA 19486 USA
bDepartment of Cancer Research, Merck Research Laboratories, Merck & Co., PO Box 4, West Point, PA 19486 USA
Received 30 September 2004; revised 2 November 2004; accepted 2 November 2004
Available online 18 November 2004
Abstract—This letter describes the development of two series of potent and selective allosteric Akt kinase inhibitors that display an
unprecedented level of selectivity for either Akt1, Akt2 or both Akt1/Akt2. An iterative analog library synthesis approach quickly
provided a highly selective Akt1/Akt2 inhibitor that induces apoptosis in tumor cells and inhibits Akt phosphorylation in vivo.
Ó 2004 Elsevier Ltd. All rights reserved.
Akt (Protein kinase B/PKB) is a serine/threonine kinase
that has recently garnered a great deal of attention as a
promising molecular target for cancer therapy due to its
critical role as a regulator of the cellÕs apoptotic machin-
ery (Fig. 1).1 Akt belongs to the AGC family of kinases
and shares high homology with PKA and PKC. More-
over, the three isozymes of human Akt (Akt1, Akt2,
and Akt3) share >85% homology and all possess an
amino terminal pleckstrin homology (PH) domain and
a kinase domain separated by a 39-amino acid hinge
region.2
(versus other AGC family kinases) as well as isozyme
selective Akt inhibitors, since recent studies have
shown that the three isozymes of Akt are functionally
Akt is a critical downstream effector of multiple growth
factors and receptors involved in tumorigenesis. Akt is
also activated as a result of inactivation of tumor sup-
pressor PTEN, a lipid phosphatase mutated or deleted
in >50% of human cancers.3 Both Akt1 and Akt2 are
commonly overexpressed or constitutively active in a
large number of human cancers including brain, gastric,
colon, breast, lung, and prostate carcinomas and their
activation correlates to cancer progression.2,4
The study of Akt as a potential therapeutic target has
been hampered by a lack of Akt specific inhibitors
Figure 1. The Akt pathway. Akt can be activated by inactivation of
PTEN, stimulation of growth factors and GPCRs or amplification of
PI3K. Once activated, Akt increases phosphorylation of a number
of downstream substrates involved in cell survival, growth, and
proliferation.
Keywords: Akt; PKB; Kinase cancer.
*
Corresponding author. Tel.: +1 2156522265; fax: +1 2156526345;
e-mail: craig_lindsley@merck.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.11.011

762
C. W. Lindsley et al. / Bioorg. Med. Chem. Lett. 15 (2005) 761–764
distinct.1–8 In this letter, we disclose the SAR and opti-
mization of two series of allosteric Akt kinase inhibitors
that are specific for Akt versus the AGC family of kin-
ases (PKA, PKC, and SGK). More importantly, these
series provided Akt1 selective and Akt2 selective inhibi-
tors that defined the profile for optimal apoptotic effect
as well as a dual Akt1/Akt2 inhibitor that sensitized
tumor cells to apoptotic stimuli and inhibited Akt
phosphorylation in vivo.
A single tractable lead, 2,3-diphenylquinoxaline 1, re-
sulted from a high throughput screening effort to iden-
tify compounds capable of inhibiting the three Akt
isozymes (Fig. 2). Interestingly, 1 displayed a high de-
gree of isozyme selectivity (IC50s: Akt1 = 3400nM,
Akt2 = 23,000nM, and Akt3 >50,000nM) and was
found be specific for Akt (>50,000nM vs PKA, PKC,
and SGK). In assays with Akt mutants lacking the PH
domain, 1 displayed no inhibition; moreover, Akt inhi-
bition by 1 was not competitive with ATP, suggesting
an allosteric binding site. The selectivity observed has
been attributed to this allosteric mode of inhibition,
for which a model has been proposed.7 All of the Akt
inhibitors described herein are PH-domain dependent
and not competitive with ATP.
Scheme 1. Synthesis of amino-functionalized quinoxalines 4. Reagents
and conditions: (a) (i) HNR1R2, PS-DIEA, DCM, rt, (ii) Rf–SH, (iii)
Fluoroit FlashTM SPE, 62–95%; (b) 1,2-diaminobenzene, EtOH/HOAc
(9:1), 160°C, 10min, microwave, 89–99%. All compounds purified by
mass-guided HPLC.10
wave-assisted organic synthesis (MAOS), amenable to
an iterative analog library approach, to access a number
of diverse heterocyclic templates from common 1,2-
diketone intermediates (Fig. 3).11 While library work
continued around 5, new libraries of 5,6-diphenylpyr-
azin-2(1H)-ones 10 were prepared employing a vast
assortment of commercial available a-aminocarbox-
amide building blocks.
An iterative analog library synthesis approach was em-
ployed to rapidly develop SAR for 1. The initial strategy
to analog the 2,3-diphenylquinoxaline core focused on
replacing the primary amine in 1 with a diverse collec-
tion of functionalized amines, while at the same time
deleting the gem-dimethyl functionality. Commercially
available p-bromomethyl benzil 2 was treated with 200
different primary and secondary amines, followed by fluo-
rous scavenging to provide 200 amino-functionalized
benzils 3.9 Each benzil 3 was heated with 1,2-diamino
benzene under microwave irradiation (160°C, 10min)
to afford 200 amino-functionalized quinoxalines 4
(Scheme 1).11 Significantly, 5, possessing a GPCR priv-
ileged structure12 and lacking the gem-dimethyl moiety,
increased potency at Akt1 (IC50 = 290nM) and Akt2
(IC50 = 2090nM) by over 10-fold, relative to 1, while
maintaining selectivity versus Akt3 and the AGC family
of kinases (IC50 >50,000nM). Despite this advance, 5
displayed poor solubility and lacked cell activity.
Benzil 11, containing the key piperidinyl benzimidazo-
lone privileged structure,12 was treated with 24 function-
alized a-aminocarboxamides 12 under our microwave-
assisted protocol to deliver 48 regioisomeric 5,6-diphen-
ylpyrazin-2(1H)-ones 13/14 (Scheme 2) in good yields.
As highlighted in Table 1, the degree of inhibition of
the individual Akt isozymes was highly variable, provid-
ing Akt1 selective, Akt2 selective, and various Akt
combination inhibitors. Not only did each regioisomer,
13 and 14, display a distinct inhibition profile, but the
nature of the 3-position substituent also impacted the
isozyme selectivity. Of the two regioisomers derived
from H-Ala-NH2, 13b (IC50s: Akt1 = 760nM, Akt2 =
24,000nM, Akt3 >50,000nM) proved to be specific for
Akt1, whereas regioisomer 14b (IC50s: Akt1 = 1003nM,
Akt2 = 1179nM, Akt3 > 50,000nM) exhibited the
same degree of inhibition for both Akt1 and Akt2.
Quite unexpectedly, 13f was uniformly inactive
(IC50s: Akt1 > 20,000nM, Akt2 = 18,000nM, Akt3 >
20,000nM) while isomer 14f was a potent and selective
A key objective for this nascent program was to evaluate
the apoptotic response of Akt1, Akt2, and dual Akt1/
Akt2 inhibition. In an effort to provide small molecule
tools with these selectivity profiles and improved physi-
cal properties, subsequent libraries were designed to
maximize structural diversity. To enable these efforts,
our laboratory developed protocols employing micro-
Figure 2. HTS screening lead.
Figure 3. 1,2-Diketone approach to diverse heterocycles.

C. W. Lindsley et al. / Bioorg. Med. Chem. Lett. 15 (2005) 761–764
763
Figure 4. Caspase-3 assay: A2780 cells treated with 13b and 14f alone
or in combination with doxorubicin. Each experiment employed a
12lM total dose of inhibitor(s).
Scheme 2. Library synthesis of diverse pyrazinones 12. Reagents and
conditions: (a) 3.0equiv 12, 9:1 EtOH/HOAc, 160°C, 20min., micro-
wave, 45–73%. Compounds purified by mass-guided HPLC.10
Our attention now focused on the development of a po-
tent Akt1/Akt2 dual inhibitor by further optimization of
5. As shown in Scheme 3, benzil 11 was heated, under
microwave irradiation, with 50 different functionalized
aryl-1,2-diamines 15 to deliver C6/C7-functionalized
and tricyclic quinoxalines 16 in excellent yields (Scheme
3).
Table 1. Structures and activities for pyrazinones 13/14
Compd
R
Akt1 IC50 Akt2 IC50 Akt3 IC50
(nM)a
As highlighted in Table 2, the degree of inhibition of the
individual Akt isozymes was conserved in the quinoxa-
line series, providing a number of potent dual Akt1/
Akt2 inhibitors. Notably, the regioisomeric C6/C7
tetrazoles, 16c and 16d, were potent dual inhibitors,
but possibly due to zwitterionic character, neither were
cell permeable. N-alkylation of the tetrazole moiety re-
moved the zwitterionic character; however, 16e lost over
15-fold potency for Akt1/Akt2 and 16f displayed poor
physical properties. The tricyclic analog 16h was selected
for further evaluation as it displayed improved potency
(IC50s: Akt1 = 58nM, Akt2 = 210nM, and Akt3 = 2119
nM), solubility and cell permeability. In our cell-based
IPKA (C33A) assay, 16h was found to possess a similar
pattern of inhibition for Akt1 and Akt2 (IC50s:
Akt1 = 305nM, Akt2 = 2086nM) while having no inhi-
bition of Akt3 (IC50 > 25,000nM).
(nM)a
(nM)a
13a
14a
13b
14b
H
3029
1500
760
15,700
>50,000
24,000
1179
>50,000
>50,000
>50,000
33,100
CH3
1003
13c
14c
17,000
>50,000
>50,000
1755
>50,000
3973
13d
14d
21,670
>50,000
45,270
5407
>50,000
>50,000
13e
14e
17,000
>50,000
>50,000
4517
>50,000
>50,000
13f
14f
>50,000
21,200
18,000
325
>50,000
21,870
a Average of at least three measurements; enzyme protocol.7 All
compounds >50,000nM versus PKA, PKC, SGK Regioisomer
structures assigned by 2-D NMR and NOE.
Akt2 inhibitor (IC50s: Akt1 >20,000nM, Akt2 =
325nM, Akt3 >20,000nM). As with 5, Akt inhibition
was PH domain dependent and noncompetitive with
ATP, providing specificity against the AGC family of
kinases (>50,000nM vs PKA, PKC, SGK).
With 13b and 14f in hand, we were poised to determine
if either Akt1 inhibition or Akt2 inhibition alone was
sufficient to induce apoptosis. As shown in Figure 4,
when A2780 cells were pretreated with either 13b or
14f, and then incubated with doxorubicin (Dox), a mod-
est 3-fold increase (versus doxorubicin alone) in caspase-
3 activity was observed. However, a 1:1 mixture of
13b:14f elicited a significant 10-fold increase in cas-
pase-3 activity. Similar results were obtained in LnCaP
cells with a number of chemosensitizing agents, suggest-
ing that inhibition of both Akt1 and Akt2 were required
for a maximal apoptotic response.
Scheme 3. Synthesis of functionalized quinoxalines 16. Reagents and
conditions: (a) EtOH/HOAc (9:1), 160°C, 10min, microwave, 85–99%.
All compounds purified by mass-guided HPLC.10

764
C. W. Lindsley et al. / Bioorg. Med. Chem. Lett. 15 (2005) 761–764
Table 2. Structures and activities for quinoxalines 16
Acknowledgements
The authors wish to thank Dr. Charles W. Ross III for
obtaining HRMS data and Dr. Sandor Varga for NMR
work to identify and characterize the regioisomeric quin-
oxalines and pyrazinones.
Compd
R
Akt1 IC50 Akt2 IC50 Akt3 IC50
(nM)a
(nM)a
(nM)a
References and notes
16a
16b
16c
16d
16e
16f
6-COOH
7-COOH
240
166
63
281
388
65
>50,000
3200
1228
1. For excellent reviews on Akt see: (a) Graff, J. R. Expert
Opin. Ther. Targets 2002, 6, 103; (b) Nicholson, K. M.;
Anderson, N. G. Cell. Signal. 2002, 14, 381; (c) Li, Q.;
Zhu, G.-D. Curr. Top. Med. Chem. 2002, 2, 939.
2. (a) Hanks, S.; Hunter, T. FASEB 1995, 9, 576; (b) Zinda,
M. J.; Johnson, M. A.; Paul, J. D.; Horn, C.; Konicek, B.
W.; Lu, Z. H.; Sandusky, G.; Thomas, J. E.; Neubauer, B.
L.; Lai, M. T.; Graff, J. R. Clin. Cancer Res. 2001, 7,
2475.
3. (a) Hsu, J. H.; Shi, Y.; Hu, L. P.; Fisher, M.; Franke, T.
F.; Lichtenstein, A. Oncogene 2002, 21, 1391; (b) Page, C.;
Lin, H.; Jin, Y.; Castle, V. P.; Nunez, G.; Huang, M.; Lin,
J. Anticancer Res. 2000, 20, 407.
4. (a) Cheng, J. Q.; Ruggeri, B.; Klein, W. M.; Sonoda, G.;
Altomare, D. A.; Watson, D. K.; Testa, J. R. Proc.
Natl. Acad. Sci. U.S.A. 1996, 93, 3636; (b) Haas-Kogan,
D.; Shalev, N.; Wong, M.; Mills, G.; Yount, G.;
Stokoe, D. Curr. Biol. 1998, 8, 1195; (c) Staal, S. P. Proc.
Natl. Acad. Sci. U.S.A. 1987, 84, 5034; (d) Brognard, J.;
Clark, A. S.; Ni, Y.; Dennis, P. A. Cancer Res. 2001, 61,
3986.
6-(2H-tetrazole)
7-(2H-tetrazole)
20
144
1877
332
1613
>50,000
>50,000
6-(2-Me-tetrazole) 1089
7(-2-Me-tetrazole) 55
16g
16h
85
58
300
210
2400
2119
a Average of at least three measurements; enzyme protocol.7 All
compounds >50,000nM versus PKA, PKC, SGK Regioisomers
assigned by 2-D NMR and NOE.
As anticipated from data obtained in caspase-3 assays
with a 1:1 mixture of 13b:14f (Fig. 4), the dual Akt1/
Akt2 inhibitor 16h displayed a similar profile. When
LnCaP cells were pretreated with 16h and then incu-
bated with TRAIL, a dramatic increase in caspase-3
activity (6–10-fold relative to control or TRAIL alone)
was observed.8,13 This sensitization of tumor cell lines
with 16h was not limited to LnCaP cells as similar apop-
tosis induction was observed in HT29, MCF7, and
A2780 cells, among others, with chemosensitizers such
as camptothecin, herceptin, and doxorubicin.8 Based
on these encouraging data, a mouse PD study was
undertaken to determine if 16h could inhibit Akt phos-
phorylation in vivo. Mice were dosed with 16h (ip,
50mpk, 3 doses, every 90min) achieving plasma concen-
trations of 1.5–2.0lM, and then the animals were tail
vein injected with IGF to stimulate Akt phosphoryl-
ation. By IP Western, both basal and IGF stimulated
Akt1 and Akt2 phosphorylation were inhibited in
mouse lung, with no effect on Akt3 phosphorylation.8
5. Kozikowski, A. P.; Sun, H.; Brognard, J.; Dennis, P. A.
J. Am. Chem. Soc. 2003, 125, 1144.
6. Breitenlechner, C. B.; Wegge, T.; Berillon, L.; Graul, K.;
Marzenell, K.; Friebe, W.; Thomas, U.; Huber, R.; Engh,
R. A.; Masjost, B. J. Med. Chem. 2004, 47, 1375.
7. For Akt enzyme assay details and the proposed model for
the allosteric mode of inhibition of these inhibitors, see:
Barnett, S. F.; Defeo-Jones, D.; Fu, S.; Hancock, P. J.;
Haskell, K. M.; Jones, R. E.; Kahana, J. A.; Kral, A.;
Leander, K.; Lee, L. L.; Malinowski, J.; McAvoy, E. M.;
Nahas, D. D.; Robinson, R.; Huber, H. E. Biochemistry,
in press.
8. For details on the Akt IPKA assay, the caspase-3 assay
and the in vivo experiment with 16h, see: Jones, R. E.;
Defeo-Jones, D.; Fu, S.; Barnett, S. F.; Hancock, P. J.;
Haskell, K. M.; Jones, R. E.; Leander, K.; Lee,
L. L.; Malinowski, J.; McAvoy, E. M.; Nahas, D. D.;
Robinson, R.; Huber, H. E.; Lindsley, C. W.; Zhao, Z.;
Duggan, M. E. Mol. Cancer Ther., in press.
In summary, two novel series of selective, allosteric Akt
(PKB) kinase inhibitors have been developed based on
either a 2,3-diphenylquinoxaline core or a 5,6-diphenyl-
pyrazin-2(1H)-one core. These potent Akt1 selective and
Akt2 selective inhibitors demonstrated that both Akt1
and Akt2 must be inhibited for a maximal apoptotic
response. An optimized dual Akt1/Akt2 inhibitor was
shown to sensitize tumor cells to induce apoptosis when
combined with a number of chemotherapeutic agents
and biologicals, such as TRAIL, and inhibit the phos-
phorylation of Akt1 and Akt2 in vivo. Moreover, this
work has demonstrated that it is possible to develop spe-
cific Akt inhibitors (versus PKA, PKC, and SGK) as
well as isozyme selective Akt inhibitors.
9. (a) Lindsley, C. W.; Zhao, Z.; Leister, W. H. Tetrahedron
Lett. 2002, 43, 4225; (b) Zhang, W.; Curran, D. P.; Chen,
C. H.-T. Tetrahedron 2002, 58, 3871.
10. Leister, W.; Strauss, K.; Wisnoski, D.; Zhao, Z.; Lindsley,
C. W. J. Comb. Chem. 2003, 5, 322, Note, all compounds
purified to >98% by mass-guided prep HPLC and fully
characterized by LCMS, HRMS, and NMR.
11. Zhao, Z.; Wisnoski, D. D.; Wolkenberg, S. E.; Leister, W.
H.; Wang, Y.; Lindsley, C. W. Tetrahderon Lett. 2004, 45,
4873, and references cited therein.
12. Patchett, A. A.; Nargund, R. P. Annu. Rep. Med. Chem.
2000, 35, 289, and references cited therein.
13. Rokhlin, G. W.; Guseva, N. V.; Tagiyev, A. F.; Glover, R.
A.; Cohen, M. B. The Prostate 2002, 52, 1, and references
cited therein.