Rational discovery of a highly novel and selective mTOR inhibitor

Article abstract of DOI:10.1016/j.bmcl.2019.126659

Aided by Structure Based Drug Discovery (SBDD), we rapidly designed a highly novel and selective series of mTOR inhibitors. This chemotype conveys exquisite kinase selectivity, excellent in vitro and in vivo potencies and ADME safety profiles. These compo

Full text of DOI:10.1016/j.bmcl.2019.126659

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Rational discovery of a highly novel and selective mTOR inhibitor
⁎
Stanley Jina,b, , Satoshi Mikamia,1, Nick Scoraha, Young Chena,2, Petro Halkowycza, Lihong Shia,2
,
Jason Kahanaa,3, Patrick Vincenta, Ron de Jonga, Joy Atienzaa, Robyn Fabreya, Lilly Zhanga,4
,
Matthew Lardya,5
a Takeda California, 10410 Science Center Drive, San Diego, CA 92121, United States
b Fronthera US Pharmaceuticals, LLC, San Diego, CA, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
mTOR
Aided by Structure Based Drug Discovery (SBDD), we rapidly designed a highly novel and selective series of
mTOR inhibitors. This chemotype conveys exquisite kinase selectivity, excellent in vitro and in vivo potencies
and ADME safety profiles. These compounds could serve as good tools to explore the potential of TORC in-
hibition in various human diseases.
SBDD
Inhibitor
The mammalian target of rapamycin (mTOR), is a member of the
phosphatidylinositol 3-kinase-related kinase family of protein kinases,1
which includes two distinct complexes: mTOR complex 1 (mTORC1)
and mTOR complex 2 (mTORC2). As a serine/threonine protein kinase,
mTOR regulates protein synthesis, cell growth and proliferation. In
addition, mTOR has been shown to regulate the ability of tumors to
promote angiogenesis as well as the response of tumor cells to nutrients
and growth factors. Thus, inhibitors of mTOR are considered as a target
for immunosuppression, leukemia, and cancer treatment et al.2–5
As the first potent, synthetic mTOR inhibitor, PI-103 is a multi-
targeted PI3K inhibitor for p110α/β/δ/γ.6 It potently inhibits both the
rapamycin-sensitive (mTORC1) and rapamycin-insensitive (mTORC2)
complexes of the protein kinase mTOR.7 Also, PI-103 is a potent ATP-
competitive dual inhibitor of phosphatidylinositol 3-kinase (PI3K) and
mTOR.8 By inhibiting mTOR activity, PI-103 induces autophagy and
promotes cell apoptosis.9,10 In order to improve the properties of this
lead, structural modification of PI-103 led to the discovery of Com-
pound 1 (Fig. 1), a quaternary-substituted dihydrofuranpyrimidine,
which was found to inhibit human mTOR with a pIC50 of greater than
7.5.2,11 It was suggested that this quaternary structure would provide
additional selectivity over the PI3K, while blocking a benzylic position
was expected to reduce metabolism.12
binding mode consists of a pyrimidine core incorporating a morpholine
hinge-region binding group, and a back pocket. Most of those fragments
have been well studied and optimized in the past decades, and nu-
merous potent mTOR inhibitors were reported. Based on the SAR of this
series that morpholine moiety plays an important role in the mTOR
potency,14 we found that no related efforts have been reported to focus
on the conformation study between the morpholine moiety and the
pyrimidine core. This consideration brought us to hypothesize that this
circled region would be worth studying (Fig. 2). The conformation of
the previous mTOR inhibitors was uncertain due to its binding pattern
between the pyrimidine and the morpholine. Thus, we hypothesized
that the introduction of the ring fusion would keep the coplanarity and
fix the conformation of the morpholine moiety and pyrimidine core,
which would also occupy the corresponding region.
Molecular modeling by structure-based drug design (SBDD) tech-
niques suggested that the six-member ring-fused molecule best fits in
the pocket, comparing to the five- and seven-member fused-ring mo-
lecules. Thus, we designed and synthesized four targets (Compound 2,
Compound 3, Compound 4 and Compound 5, as shown in Fig. 3), and
ran the essay for the mTOR/PI3K inhibition to validate the molecular
modeling result. Consistent with modeling analyses, we observed that
analogues with six-member fused rings showed better mTOR inhibition
(pIC50 > 6) than five- and seven-member rings (pIC50 < 6). However,
the selectivity of mTOR inhibition over PI3Kα was low (e.g. pIC50 = 6.5
By studying the structure–activity relationship (SAR) of mTOR in-
hibitors in the literature,2,13–15 we found that the general mTOR
⁎ Corresponding author at: 11526 Sorrento Valley Road, STE D, San Diego, CA 92121, United States.
E-mail addresses: stanjin@frontherapharma.com (S. Jin), patrick.vincent313@att.net (P. Vincent), mlardy@predictivenumerics.com (M. Lardy).
1 Takeda Pharmaceutical Company Limited, Kanagawa, Japan.
2 Celgene, 10300 Campus Point Drive, Ste 100, San Diego, CA 92121, United States.
3 Integrity Biosolutions, 8724 Butano Ct, San Diego, CA 92129, United States.
4 Simcere Pharmaceutic Group.
5 Predictive Numerics.
https://doi.org/10.1016/j.bmcl.2019.126659
Received 9 July 2019; Received in revised form 28 August 2019; Accepted 31 August 2019
0960-894X/©2019ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:StanleyJin,etal.,Bioorganic&MedicinalChemistryLetters,https://doi.org/10.1016/j.bmcl.2019.126659

S. Jin, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
improved much, and the selectivity of mTOR inhibition over PI3Kαwas
still not ideal. It was reported that the replacement of morpholine by a
substituted morpholine with certain chirality significantly improved
mTOR potency and selectivity over PI3Kα.14 Thus, by studying the
stereochemistry of Compound 6, we found that the replacement of the
proton at the axial position with methyl group might reach our ex-
pectation. By synthesizing the target compound Compound 7 with
desired chirality, we observed that the mTOR inhibition was further
increased (pIC50 from 7.5 to 8.1), while its corresponding PI3Kα in-
hibition was decreased (pIC50 from 7.7 to 7.1). Further optimization on
the substitution group on the urea from methyl to cyclopropyl gave
much better mTOR inhibition selectivity over PI3Kα. Thus, the mTOR
inhibition selectivity over PI3Kα was increased by 1000-fold during the
Fig. 1. Structural modification of PI-103 led to the discovery of Compound 1.
whole optimization from Compound
3 to Compound 8 within
2 months. Synthetically, the methyl group could only be at the axial
position. Further optimization showed that Compound 9 exhibited
both slightly better mTOR inhibition activity and selectivity over PI3Kα
than Compound 8.16
One lead compound, Compound
9
(mTOR pIC50 = 7.9;
PI3KαpIC50 = 5.9) showed broad potencies in a variety of cancer cell
lines (Fig. 5a). Generally, cancer cell lines of blood, breast, colon,
kidney and ovary origins showed relatively increased sensitivity to
TORC inhibition by Compound 9. Combined treatments of Compound
9 with EGFRi (erlotinib) or gemcitabine caused synergistic effects in
LKB1/KRASmut NSCLC cell lines (Fig. 5b). No correlation between
combination effect and EGFR status was observed. Additive effects were
observed when Compound 9 was combined with PI3Kα inhibitor GDC-
0941, RAFi (sorafenib), and BCL2i (ABT-737).
Fig. 2. SAR study of mTOR inhibitors.
Compound 9 also proved highly potent in a number of xenograft
efficacy studies. For example, partial regressions were observed in a
Fig. 3. SBDD-aided molecular modeling and mTOR/PI3Kα inhibition results of
the designed targets.
for mTOR/7.5 for PI3Kα, Fig. 3).
In order to improve the selectivity of mTOR inhibition over PI3Kα,
further optimization of Compound 3 was carried out (Fig. 4). The
mTOR inhibition (pIC50 = 7.5) was increased by N-alkyation of the
fused amide. The corresponding PI3Kα inhibition (pIC50 = 7.7) was not
Fig. 5. a) Shows potent in vitro activity of Compound 9 against a broad range
of cancer cell lines; b) Effect of Erlotinib and Gemcitabine with LKB/KRASmut
cell lines. Effects above additive were considered synergistic.
Fig. 4. The discovery process of Compound 8.
2

S. Jin, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Vincent P. MET tyrosine kinase inhibition enhances the antitumor ef-
ficacy of an HGF antibody. Mol Cancer Ther. 2017 Jul;
16(7):1269–1278.
786-O (PTEN-null) RCC xenograft
1000
100
10
mTor assay
Compound 9, QD
Compound 9, QD
Compound 9, QD
Materials
Rx: 1-14
Re-growth
FRAP1(mTor) recombinant protein and GFP-4E-BP1 were pur-
chased from Invitrogen. (Carlsbad, CA, US) The kinase activity assay
was conducted in 50 mM Hepes buffer containing 10 mM MgCl2, 0.01%
Brij, 0.2 mM EDTA, 2 mM DTT and 1% DMSO (after compound addi-
tion) at pH 7.3.
1
0
10
20
30
40
50
Day
Fig. 6. Compound 9 exhibits excellent tumor growth inhibition.
Methods
IC50 determination – the IC50 value for each compound was de-
termined in the presence of compound (various concentrations, from 0
to 10 μM) and a fixed amount of ATP (50 μM, final concentration),
substrate GFP-4E-BP1 (400 nM, final concentration). The enzymatic
reaction was initiated by adding FRAP1(mTor) (2 nM, final concentra-
tion). The assay was conducted at room temperature (~22 °C). After
30 min, the enzymatic reaction was stopped using Tb-anti-p4E-BP1
(2 nM final concentration, Invitrogen) and EDTA (20 mM final con-
centration, Sigma) in TR-FRET dilution buffer provided by Invitrogen.
All the reagents were dispensed using a Multidrop Combi reagent dis-
penser (ThermoFisher Scientific, Waltham, MA US) into black 384 SV
Greiner plates. The release of product was detected using a BMG
PHERAstar plate reader (BMG LABTECH, Ortenberg, Germany) and the
Lantha Screen module (337 nm (excitation wavelength) and measuring
the ratio of fluorescence 520/490 nm (emission wavelengths).
Experimental data was fitted using Eq. (1):
Scheme 1. Synthesis of Compound 9.
786-O xenograft study at 30 mpk QD, (Fig. 6). Treatment was well
tolerated while maintaining a < 10% body weight.
V
V
o
100
i
=
n
I
50
1 +
(
)
The preparation of target compound Compound 9 was described in
Scheme 1. The target compound was synthesized from commercially
available 3-morpholinecarboxylic acid A via 6 steps. The initial cycli-
zation of piperazinone ring was resulted from 3-morpholinecarboxylic
acid A and 5-amino-2,4-dichloropyrimidine B in 70.3% yield. N-alky-
lation of compound C using chloroacetone afforded compound E in 73%
yield, which was subsequently fluorinated by DAST under mild condi-
tion to give di-fluoro intermediate F in 59% yield. After treating F with
strong base, the C-alkylation was realized by addition of methyl iodide
in 54% yield. Suzuki coupling reaction between G and para-urea sub-
stituted phenylboronic ester H afforded the racemic target compound I
in 44% yield. Final chiral separation afforded the desired target com-
pound 9.
IC
Vi and Vo are the rates of the enzyme activity in the presence and in the
absence of inhibitor; n is the Hill coefficient; I is the free inhibitor
concentration, respectively; IC50 is a measure of potency that is
equivalent to a concentration of inhibitor that leads to a 50% inhibition
of the enzyme activity.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2019.126659.
References
In summary, we first carefully studied the SAR of both literature and
in-house mTOR and PI3Kα inhibitors to gain knowledge on the subtle
differences of the mTOR and PI3Kαbinding pockets. This knowledge in
turn inspired us to come up with design hypotheses, which we meth-
odically interrogated and prioritized with the aid of SBDD. The initial
hit Compound 3 was actually 10-fold PI3Kαselective. However, a clear
understanding of the chemotype’s conformation resulted in the break-
through design element of a pseudo-axial methyl group, which ulti-
mately gave us excellent (> 100-fold) mTOR selectivity over PI3Kα.
This chemotype conveys exquisite kinase selectivity, excellent in-vitro
and in-vivo activity along with desired ADME safety profiles. These
compounds could serve as good tools to explore the potential of TORC
inhibition in various human diseases.
[1]. Mitra A, Luna JI, Marusina AI, et al. J Invest Dermatol. 2015;135:2873.
[2]. Dong, Q.; Jin, B.; Lardy, M.; Zhao, F. U.S. Pat. Appl. 2011021515, 2011.
[3]. Giles FJ, Albitar M. Curr Mol Med. 2005;5:653.
[4]. Zhang L, Bu T, Bao X, et al. Bioorg Med Chem Lett. 2017;27:3395.
[5]. Faivre S, Kroemer G, Raymond E. Nat Rev Drug Discovery. 2006;5:671.
[6]. Hayakawa M, Kaizawa H, Moritomo H, et al. Bioorg Med Chem Lett. 2007;17:2438.
[7]. Knight ZA, Gonzalez B, Feldman M, et al. Cell. 2006;125:733.
[8]. Schenone S, Brullo C, Musumeci F, Radi M, Botta M. Curr Med Chem. 2011;18:2995.
[9]. Fan Q, Weiss WA. Autophagy. 2011;7:536.
[10]. Park S, Chapuis N, Bardet V, et al. Leukemia. 2008;22:1698.
[11]. Jin, B.; Lardy, M.; Zhou, F.; Dong, Q. WO 2012099581 A1, 2012.
[12]. Cohen F, Bergeron P, Blackwood E, et al. J Med Chem. 2011;54:3426.
[13]. Sutherlin DP, Sampath D, Berry M, et al. J Med Chem. 2010;53:1086.
[14]. Verheijen JC, Yu K, Toral-Barza L, Hollander I, Zask A. Bioorg Med Chem Lett.
2010;20:375.
[15]. Kaplan J, Verheijen JC, Brooijmans N, et al. Bioorg Med Chem Lett. 2010;20:640.
[16]. Jin, B.; Scorah, N.; Dong, Q. U.S. Pat. Appl. 20110053921, 2011.
In vitro and in vivo assays were performed as outlined in: Farrell PJ,
Matuszkiewicz J, Balakrishna D, Pandya S, Hixon MS, Kamran R, Chu S,
Lawson JD, Okada K, Hori A, Mizutani A, Iwata H, de Jong R, Hibner B,
3