Discovery and Preclinical Characterization of 5-[4,6-Bis({3-oxa-8-azabicyclo[3.2.1]octan-8-yl})-1,3,5-triazin-2-yl]-4-(difluoromethyl)pyridin-2-amine (PQR620), a Highly Potent and Selective mTORC1/2 Inhibitor for Cancer and Neurological Disorders

Article abstract of DOI:10.1021/acs.jmedchem.8b01262

Mechanistic target of rapamycin (mTOR) promotes cell proliferation, growth, and survival and is overactivated in many tumors and central nervous system disorders. PQR620 (3) is a novel, potent, selective, and brain penetrable inhibitor of mTORC1/2 kinase. PQR620 (3) showed excellent selectivity for mTOR over PI3K and protein kinases and efficiently prevented cancer cell growth in a 66 cancer cell line panel. In C57BL/6J and Sprague-Dawley mice, maximum concentration (C_max) in plasma and brain was reached after 30 min, with a half-life (t_1/2) > 5 h. In an ovarian carcinoma mouse xenograft model (OVCAR-3), daily dosing of PQR620 (3) inhibited tumor growth significantly. Moreover, PQR620 (3) attenuated epileptic seizures in a tuberous sclerosis complex (TSC) mouse model. In conclusion, PQR620 (3) inhibits mTOR kinase potently and selectively, shows antitumor effects in vitro and in vivo, and promises advantages in CNS indications due to its brain/plasma distribution ratio.

Full text of DOI:10.1021/acs.jmedchem.8b01262

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Article
Discovery and Preclinical Characterization of 5-[4,6-Bis({3-
oxa-8-azabicyclo[3.2.1]octan-8-yl})-1,3,5-triazin-2-yl]-4-
(difluoromethyl)pyridin-2-amine (PQR620), a Highly Potent and
Selective mTORC1/2 Inhibitor for Cancer and Neurological Disorders
Denise Rageot, Thomas Bohnacker, Anna Melone, Jean-Baptiste Langlois, Chiara
Borsari, Petra Hillmann, Alexander Markus Sele, Florent Beaufils, Marketa Zvelebil, Paul
Hebeisen, Wolfgang Loescher, John E. Burke, Doriano Fabbro, and Matthias P. Wymann
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01262 • Publication Date (Web): 25 Oct 2018
Downloaded from http://pubs.acs.org on October 25, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery and Preclinical Characterization of 5-[4,6-Bis({3-
oxa-8-azabicyclo[3.2.1]octan-8-yl})-1,3,5-triazin-2-yl]-4-
(difluoromethyl)pyridin-2-amine (PQR620), a Highly Potent and
Selective mTORC1/2 Inhibitor for Cancer and Neurological
Disorders
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
†
†
†
†
Denise Rageot* , Thomas Bohnacker* , Anna Melone* , Jean-Baptiste Langlois , Chiara Borsari , Petra
‡
†
†
†
‡
Hillmann , Alexander M. Sele , Florent Beaufils , Marketa Zvelebil , Paul Hebeisen , Wolfgang
§
£
‡
$,†
Löscher, John Burke , Doriano Fabbro , Matthias P. Wymann
†Department of Biomedicine, University of Basel, Mattenstrasse 28, 4058 Basel, Switzerland;
‡PIQUR Therapeutics AG, Hochbergerstrasse 60, 4057 Basel, Switzerland;
§Department of Pharmacology, Toxicology, and Pharmacy, University of Veterinary Medicine Hannover,
and Center for Systems Neuroscience, 30559 Hannover, Germany
£Department of Biochemistry and Microbiology, University of Victoria, Victoria, British Columbia BC
V8W 2Y2, Canada
*
Equal contribution: these authors have contributed equally.
$Correspondence to: matthias.wymann@unibas.ch; Dept. Biomedicine, University of Basel, Mattenstrasse
28; 4058 Basel; Tel. +41 61 207 5046
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
KEYWORDS
mammalian or mechanistic target of rapamycin (mTOR); mTOR kinase inhibitor; Tuberous
sclerosis complex (TSC); phosphoinositide 3-kinase (PI3K); brain penetrable; cancer; neurodegenerative
disorders; epilepsy, ATP-competitive; rapamycin, rapalogs.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ABSTRACT
Mechanistic target of rapamycin (mTOR) promotes cell proliferation, growth and survival, and is
overactivated in many tumors and central nervous system disorders. PQR620 (3) is a novel, potent,
selective, and brain penetrable inhibitor of mTORC1/2 kinase. PQR620 (3) showed excellent selectivity for
mTOR over PI3K and protein kinases, and efficiently prevented cancer cell growth in a 66 cancer cell line
panel. In C57BL/6J mice and Sprague Dawley maximum concentration (Cmax) in plasma and brain was
reached after 30 minutes, with a half-life (t ) > 5 hours. In an ovarian carcinoma mouse xenograft model
1
/2
(
OVCAR-3), daily dosing of PQR620 (3) inhibited tumor growth significantly. Moreover, PQR620 (3)
attenuated epileptic seizures in a Tuberous Sclerosis Complex (TSC) mouse model. In conclusion, PQR620
3) inhibits mTOR kinase potently and selectively, shows antitumor effects in vitro and in vivo, and
(
promises advantages in CNS indications due to its brain/plasma distribution ratio.
2
ACS Paragon Plus Environment

Page 3 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
The mechanistic target of rapamycin (mTOR, earlier mammalian TOR) operates downstream of
phosphoinositide 3-kinase (PI3K) and protein kinase B (PKB/Akt). The PI3K/mTOR pathway is a key
regulator of metabolism, cell growth, proliferation and survival, and is frequently over-activated in cancer
and neurodegenerative disease.1,2 The mTOR kinase is part of two main TOR complexes (TORC1 and
TORC2). TORC1 integrates signals from cell surface receptors, energy status, stress, availability of oxygen
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and amino acids. Cell surface receptors activate PI3K to produce PtdIns(3,4,5)P , which serves as a docking
3
site for PKB/Akt and 3-phosphoinositide-dependent protein kinase 1 (PDK1). PKB/Akt is phosphorylated
3
by PDK1 on Thr308 and by mTOR kinase integrated in TORC2 in the hydrophobic motif on Ser473. Fully
activated, PKB/Akt phosphorylates tuberous sclerosis 2 (TSC2, tuberin) in the tuberous sclerosis complex
(TSC), and blocks its GTPase-activating protein (GAP) activity – which leads to the accumulation of GTP-
loaded Rheb-GTPase (homolog enriched in brain), and assembly of active TORC1 on endomembranes.2
TORC2 is linked to PI3K activation by the Sin1 PH domain, as the Sin1-PtdIns(3,4,5)P interaction unlocks
3
4
TORC2 kinase activity after membrane translocation. The phosphorylation of S6 kinase (S6K) by TORC1
is key to control cellular protein and lipid synthesis. TORC1 activity also impacts lysosome dynamics and
autophagy, and TORC2 promotes metabolic activity, cell cycle and cytoskeletal rearrangements.2
Rapamycin (sirolimus), its more soluble rapamycin-derivative (rapalog) RAD001 (everolimus), as
5
well as CCI-779 (Temsirolimus) block TORC1 specifically by the formation of a TORC1/rapalog/FK506
6
binding protein 12 (FKBP12) complex. Rapamycin has a long history as immunosuppressive agent in
7
organ transplant recipients , and rapalogs have been exploited to prevent restenosis, based on their anti-
inflammatory and anti-proliferative effects.8
In 2007, Temsirolimus (CCI-779) was approved for the treatment of renal cell carcinoma (RCC),
9
followed by a phase III trial of everolimus in an advanced stage of the disease. Everolimus has been
approved for advanced pancreatic, lung and gastrointestinal neuroendocrine tumors, mantle cell lymphoma,
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
endometrial cancer, non-cancerous kidney tumors in patients with TSC, pediatric subependymal giant cell
astrocytoma (SEGA), and advanced ER+/HER2-negative breast cancer.10
In spite of the clinical success of rapalogs in the above indications, overall response rates to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
11
rapalogs as single-agent in major solid tumors have been modest. Reasons for this are i) adverse effects
of rapamycin due to its immune-suppressive action, limiting the maximal tolerated dose, but also ii) the
attenuation of drug action by a rapamycin-mediated suppression of a negative feedback loop, where TORC1
activates S6K and Grb10 to negatively control PI3K activation by receptors for insulin and growth factors.
In the presence of rapalogs, TORC1 is thus inhibited, but class IA PI3Ks are overactivated – resulting in a
TORC2-dependent increase in PKB/Akt activity.1,2 Compounds targeting both TORC1 and TORC2 are
therefore expected to provide a more efficient block of growth factor and PI3K downstream signaling.
ATP-competitive inhibitors have therefore been developed to target TORC1 and TORC2 protein
kinase activities.12 Sapanisertib (INK128 (73), MLN0128, TAK-228)13 is in multiple clinical trials
including advanced solid tumors, non-Hodgkin’s lymphoma, breast cancer, multiple myeloma, and
Waldenstrom’s macro-globulinaemia, while CC-223 (77)14 is explored in malignancies such as non-
Hodgkin’s lymphoma, multiple myeloma, diffuse large B-cell and follicular lymphoma.
15
Vistusertib[AZD2014 (78)] is currently evaluated in a range of advanced tumors, and has progressed to
phase II clinical trials as a single agent in non-responders to chemotherapy with mutated TSC (for more
studies see clinicaltrials.gov). Moreover, preclinically characterized chemical probes with different potency
and selectivity for TOR kinase, such as Torin1, PP242 (75), Ku-0063794, WAY-600, WYE-687, WYE-
54, Palomid 529 and more have been described (for a review see Ref. 5).
3
The over-activation of the mTOR pathway is not only important in malignant cancer, but also
promotes the development of hyperplastic lesions. In autosomal-dominant tuberous sclerosis complex
TSC), mutations in either TSC1 (hamartin) or TSC2 (tuberin) lead to a loss of control of mTOR signaling,
(
and the formation of initially benign hamartomas. These occur in multiple organs such as skin, heart,
4
ACS Paragon Plus Environment

Page 5 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
kidneys, lung, and brain. Some form of impaired CNS function is frequent in TSC, and includes epileptic
seizures in ~90% patients, which dominates TSC morbidity to a high degree.16
Rapalogs have recently been explored to alleviate epileptic seizures triggered by loss-of-function
mutations in the mTOR pathway. In a mouse model where TSC1 was selectively eliminated in neurons,
sirolimus (rapamycin) and everolimus (RAD001) significantly improved the survival of neuronal-specific
Tsc1 null mice and attenuated TORC1 downstream signaling.17 A similar action of rapalogs could be
achieved in TSC patients,18,19 finally culminating in the approval of everolimus for TSC-associated seizures
in April 2018. Rapalogs are, however, not generally effective in the whole TSC population, as resistance
and considerable adverse side effects lead to treatment termination and dose reduction.18,20
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1
Targeting mTOR has also been proposed to be beneficial in Alzheimer's disease , Huntington's
2
2
16
disease and Parkinson's disease, mostly based on the observation that inhibition of mTOR promoted
removal of toxic protein complexes by the induction of autophagy. As rapalogs display limited brain
permeability, there is a lack of molecules targeting mTOR in the CNS.
23
We have recently reported on PQR309 (1) , a pan-PI3K/mTOR inhibitor with excellent brain
penetration and oral bioavailability, which is currently in phase II clinical trials for the treatment of
lymphoma and solid tumors. This compound served as a starting point for the development and rational
design of selective and highly brain penetrable mTOR kinase inhibitors. By introducing substituents with
specific steric demands and defined electronic properties, we were able to derive a compound class with
increased affinity for mTOR and much reduced binding of PI3K (Figure 1). From this series we developed
the highly selective and potent ATP-competitive mTOR inhibitor PQR620 (3) targeting TORC1 and
TORC2.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O
N
O
N
O
N
mTOR
dial in
PI3K
F
F
F
F
F
F
F
N
N
N
N
N
N
dial out
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
N
N
N
N
N
O
O
O
N
NH2
N
NH2
N
NH2
PQR309 (1)
2
PQR620 (3)
K mTOR = 93.4 nM
K p110 = 15.6 nM
K mTOR = 6.9 nM
K p110 = 11.5 nM
K mTOR = 10.8 nM
K p110 = 4203 nM
i
i
i
i
i
i
Figure 1. Schematic design and optimization cascade leading to the mTOR kinase-selective ATP-binding site inhibitor PQR620
3). Binding affinity for mTOR kinase was increased by the substitution of the trifluoromethyl group in PQR309 (1) with a
(
difluoromethyl group in compound 2. The introduction of substituted morpholino groups was then explored to reduce PI3K binding,
resulting in PQR620 (3).
RESULTS AND DISCUSSION
The transition from a pan-PI3K to an mTOR inhibitor was initiated by the elucidation of the binding
mode of PQR309 (1) in the ATP-binding pocket of PI3K and mTOR kinases, where PQR309 (1) showed a
six-fold selectivity for PI3K over mTOR kinase (K for PI3Kα was 15.6 nM and K for mTOR was 93.4
i
i
nM). The binding mode of PQR309 (1) in PI3Kγ involves a hydrogen bond interaction in the hinge region
between the backbone amide of Val882 and the morpholine oxygen atom. Additional interactions occur
between Asp836 and Asp841 and the 2-aminopyridine moiety in the affinity pocket.23,24 An analogous
binding mode of PQR309 (1) is assumed in mTOR, involving residues Val2240 and Asp2190/2357.
Optimization of mTOR Affinity.
To assess the influence of the trifluoromethyl moiety at the C4-position of the aminopyridine (R-
substituent), towards selective mTOR inhibition, we prepared a series of compounds replacing the
6
ACS Paragon Plus Environment

Page 7 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
trifluoromethyl moiety with various substituents and tested them in cellular and enzymatic assays (Table
1
). Physicochemical properties, such as clogP and PSA, were calculated for all studied compounds (see
Table S9). Substitutions on the heteroaromatic ring (compounds 1, 2, 4-9) had a major impact on the balance
between hydrophilicity and lipophilicity (clogP variations from 1.08 to 2.72).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. SAR Study of Selected R-substituted Triazine Core Compounds.
O
N
M0 =
M1 =
M2 =
M3 =
N
N
R
N
O
N
O
N
O
N
O
N
Mn
N
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH2
IC50 [nM]a
[nM]b
K
i
Ratio K
i
Compound
R
M
n
pPKB/Akt
404
pS6
382
p110α
mTOR
46.1
p110α/mTOR
4*
H
M
M
M
M
M
M
M
M
M
M
M
94.0
2.0
0
0
0
0
0
0
0
0
1
2
3
5**
CH
401
178
102
397
341
1198
1848
147
144
156
939
194
118
218
528
2109
1918
138
112
113
61.2
15.6
11.5
68.6
81.3
129
609
93.4
6.9
0.10
0.17
1.7
3
PQR309
CF
3
(1)*
2
6
7
8
9
CHF
2
CH
CH
2
2
F
54.3
74.5
474
483
9.6
1.3
OH
1.1
CH(OCH
OCH
3
)
2
0.27
1.4
3
693
1
1
1
0
1
2
CHF
CHF
CHF
2
20.8
23.7
25.8
2.1
2
3.4
7.0
2
11.2
2.3
aPhosphorylation of PKB/Akt on Ser473 and ribosomal protein S6 on Ser235/236 were analyzed in A2058 cells exposed to the
23
b
indicated inhibitors and subsequent detection of phosphoproteins in an in-cell Western assay . Compounds were tested in vitro
for binding to the catalytic subunit of PI3Kα (p110) and mTOR using a commercially available time-resolved FRET (TR-FRET)
displacement assay (LanthaScreenTM). Means of K
SD) are depicted in Table S9.
s are shown; data sets *extended or **derived from Ref. 23. Standard deviations
i
(
Among this bismorpholino-substituted triazine series, compound 2 with a difluoromethyl
substituent (R = CHF ) was found to be the most potent compound in vitro (K for mTOR of 6.9 nM) and
2
i
8
ACS Paragon Plus Environment

Page 9 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
in cells (IC for phosphorylated PKB = 102 nM, for phosphorylated S6, IC = 118 nM). In general, the
5
0
50
affinity towards mTOR increased as a function of R in the order: CH ≈ CH(OCH ) ≈ OCH < CF ≈
3
3 2
3
3
CH OH ≈ CH F ≈ H < CHF . For compound 2 the cellular data displayed both efficient inhibition of TORC1
2
2
2
upstream of ribosomal protein S6 kinase (see IC value for pS6 in Table 1), and TORC2 activity upstream
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
of PKB phosphorylation (see IC for pPKB in Table 1). The low mTOR affinities of compounds 8 and 9
5
0
can only in part be attributed to bulky substituents as the affinity differences in CF -, CHF -, CH F- and
3
2
2
CH -substituted compounds point to an importance of the dipole moment of CHF -group in R in mTOR,
3
2
but less so in PI3K.
25
The CHF -group was recently proposed to act as a lipophilic hydrogen donor, and has as such
2
two possible conformational interaction modes in PI3K and mTOR: as depicted in Figure 2A, the proton of
the CHF -group could i) interact with a triazine core nitrogen in an intramolecular fashion, or ii) form
2
interactions with Asp836 in PI3K or Glu2190 in mTOR (Figure 2B). An intramolecular proton-nitrogen
interaction with the core does not disturb the overall orientation of the inhibitor, although in class I PI3K
the CHF -group is in close proximity to a methionine (Met804 in PI3K, distance ca. 3.3 Å), while mTOR
2
presents an isoleucine (Ile2163, > 7Å) at this position. Turned towards Glu2190 (mTOR) or Asp836
(
PI3K), the CHF -group appears to be closer to the negative side chain in mTOR than in PI3K. This, and
2
the slightly more hydrophobic nature of the pocket in mTOR (see Ile2163), is likely to enforce a CHF2-
proton - Glu2190 hydrogen bond, and thus explains the pronounced increase in mTOR affinity for CHF2-
substituted compounds.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A
B
L2185 I831
K833/2187
I831
L2185
P810/2169
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
K833/2187
I2163
I879/2237
I879/2237
E2190
F
S
F
N
Met804
F
H
S
S2165
S806
H
Met804
F
N
V882/2240
D836
Figure 2. Predicted interactions of compound 2 with the catalytic core of PI3K (grey) and mTOR kinase (teal). Displayed are
amino acids in close contact with the inhibitor; three-digit numbers refer to PI3K, four-digit codes to mTOR kinase. Labelled
spheres: N, nitrogen atom of the hinge region Val (882/2240); S, sulfur atom of Met804; F and H, difluoromethyl group. (A) One
morpholine group (left) of 2 interacts with the nitrogen atom of the hinge region Val. Here the hydrogen atom of the difluoromethyl
group is oriented to form an intramolecular interaction with the triazine core. (B) The hinge region valine interaction is oriented
towards the back, and the hydrogen atom of the difluoromethyl group is well positioned to form an interaction with Glu2190 of
mTOR. (A, B) Compound 2 in PI3K was modelled based on a PQR309 (1)–PI3K complex23 (PDB ID code: 5OQ4; resolution
of 2.7 Å); and the mTOR-compound 2 complex was derived from a 3.6 Å resolution structure of PI103 bound to mTOR (PDB
ID code: 4JT6) by docking as described in Ref. 23. A comparative listing of amino acid residues depicted here across PI3K
isoforms can be found in Table S1.
Increasing Selectivity for mTOR, Dialing out PI3K.
As many of the above R-substituents maintained low K values for PI3Kα (see PQR309 (1), 2, 5
i
and 6; Table 1), we introduced various morpholine derivatives as M -substituents, such as (R)-3-
n
methylmorpholine (M ), 8-oxa-3-azabicyclo[3.2.1]octane (M ) and 3-oxa-8-azabicyclo[3.2.1]octane (M ,
1
2
3
Table 1). Morpholine derivatives M , M and M with increased steric demands were examined to determine
1
2
3
whether the interaction in the hinge region of mTOR occurs with the oxygen-atom of the morpholine M0
or to the substituted morpholine (M ). The effect of various substitution of morpholines on mTOR versus
1
-3
1
0
ACS Paragon Plus Environment

Page 11 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
26
PI3K selectivity has been reported for pyrazolopyrimidines, showing a profound loss in PI3Kα activity
leading to a > 1000-fold selectivity for mTOR over PI3Kα when morpholine M was replaced with the 2,6-
0
or 3,5-ethylene bridged morpholine (M and M ). Likewise, the introduction of (R)-3-methylmorpholine
2
3
(
M ) improved the selectivity for mTOR.
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Consequently, we prepared analogs of compound 2 (with R = CHF ) with selected substituted
2
morpholines. Among M -substituents, morpholine M showed the highest affinity and selectivity for mTOR
n
2
in 11 (K for mTOR = 3.4 nM, and ratio of K of p110/mTOR ~ 7). Inhibitors 10, 11 and 12 with
i
i
substituents M = M , M and M , respectively, proved to be all excellent inhibitors for mTOR and PI3K in
n
1
2
3
vitro and in cells (Table 1). These results underline that the increase in mTOR activity is dominated by the
-(difluoromethyl)pyridin-2-amine group rather than from substitutions of one morpholine group.
Compounds 10-12 maintain most likely affinity for PI3K as they all bear one unsubstituted morpholine
4
(
M ), which is able to bind to the hinge region of the ATP-binding site of PI3K. The bulky morpholines
0
(
M ) would then be solvent exposed and have a negligible effect on PI3K or mTOR binding.
1
-3
These observations prompted us to focus our SAR study on symmetrically substituted triazines
Table 2). Introduction of methyl groups, methylene or ethylene bridges into morpholine groups (see 13-
7) did not significantly affect the clogP. Bulky alkylic substitutions on the morpholine yielded, however,
(
1
more lipophilic compounds (compounds 18-24 and 26; see Table S9). Many of these compounds showed
excellent mTOR activity in vitro (K ≤ 30 nM for PQR620 (3), 13-15, 18-19, 21, and 24-25). The affinity
i
for mTOR was reduced when both, the C2- and C6-morpholine positions were substituted with a methyl-
group (Ki of 650 nM for compound 22), whereas methyl-substitution at both the C3- and C5-position
showed only a minor effect on mTOR binding. Excellent in vitro (> 100-fold) selectivity for mTOR versus
PI3Kα were obtained for compounds PQR620 (3, M ), 15 (M ), 20 (M ), and 26 (M ); good selectivity
3
4
9
15
(
>70-fold) for compounds 14 (M ), 23 (M ) and 25 (M ).
2
12
14
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Compounds containing 3,5-dimethylmorpholines (M , M and M ) showed relevant differences in
8
9
10
cellular potency and in selectivity for mTOR versus PI3Kα. Whereas compound 20 (M ) showed good
9
selectivity (K of p110/mTOR > 250) and no inhibition of mTOR/PI3K in cells, its enantiomer 21 (M )
i
10
was much less selective (K of p110/mTOR ~ 3) and was slightly more active in cells. The enantiomeric
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
i
pair 15 and 16 (M and M ) showed poor potency in cellular assays (IC > 1000 nM) although, a remarkable
4
5
50
difference in affinity for mTOR between them (K = 18.5 nM for 15 vs. K = 308 nM for 16).
i
i
Generally, the symmetrically substituted derivatives in Table 2 displayed reduced cellular activities
as compared to unsymmetrically substituted analogues (Table 1, compounds 10-12), with the exception of
the selected compound PQR620 (3). Compound PQR620 (3) was the most potent in biochemical and
cellular assays (K for mTOR = 10.8 nM; IC for pPKB = 190 nM and IC for pS6 = 85.2 nM) and very
i
50
50
selective for mTOR versus PI3K (389-fold selectivity).
1
2
ACS Paragon Plus Environment

Page 13 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. SAR of Symmetrically Substituted 4-(difluoromethyl)-5-(1,3,5-triazin-2-yl)pyridine-2-amines.
M1 =
M6 =
M2 =
M7 =
M3 =
M8 =
M4 =
M9 =
M =
5
O
O
N
N
O
O
N
N
O
O
N
N
O
O
N
N
O
O
N
N
Mn
N
F
F
N
N
M10 =
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Mn
N
NH2
M11 =
M12 =
M13 =
M14
=
M15 =
O
N
O
N
O
N
O
N
N
O
O
IC50 [nM]a
pPKB pS6
364 310
K
[nM]b
Ratio K
i
i
Compound
13
M
n
p110α
mTOR
12.6
p110α/ mTOR
M
M
M
M
M
M
M
M
M
1
2
3
4
5
6
7
8
9
633
1297
50
14
630
190
301
13.9
10.8
18.5
308
93
389
>1000
15
PQR620 (3)
85.2
4203
15
16
4037
17685
17851
354
1565
>20000
4632
12572
5768
209
1
1
1
2
2
2
2
2
2
2
7
8
9
0
1
2
3
4
5
6
>20000
551
49
32.2
141
>20000
64.0
13.6
3.5
2.4
252
160
40
3570
379
1189
294
79.9
21.1
650
>250
3.0
M
M
M
M
M
M
10
11
12
13
14
15
20872
4815
942
10380
2329
526
15364
8787
24
73.4
32.8
31.0
60.2
120
19
632
357
275
2379
77
1212
1298
>20000
>330
a
b
PKB phosphorylation on Ser473 and S6 phosphorylation on Ser235/236 were analyzed as described in Table 1. Compounds were
tested for the in vitro binding to the ATP-binding site of p110α and mTOR as in Table 1. SDs are depicted in Table S9.
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Elucidation of Binding Modes to PI3K and mTOR.
Computational modeling studies were used to elucidate the binding mode of PQR620 (3), and
ultimately provided structural features defining the compounds high selectivity towards mTOR versus
PI3K. Exploiting the 2.7 Å resolution x-ray structure of PQR309 (1) in PI3Kγ (PDB code 5OQ4), we
substituted the inhibitor for PQR620 (3). Energy minimization calculations of the resulting PI3K-ligand
complex maintained important interactions such as hydrogen bonds between the aminopyridine and
Asp836/964 as well the one between one morpholine oxygen atom and the backbone amine of Val882 (as
present in the parental structure with PQR309 (1)). In PQR620 (3), the ethylene bridge of the morpholine
pointing towards Val882 can take two main conformations, with the bridge oriented towards Met953
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
("bridge-up", Figure 3A,B) or towards Ile881 ("bridge-down", Figure 3C). The bridge-up conformation
induces steric clashes within a region of the ATP-binding pocket that has previously been identified as very
rigid in PI3K, and is defined by residues Tyr867, F961 and I963.27
The “bridge-down” conformation of PQR620 (3) generates steric clashes with the backbone of
Ile881 and Glu880 and side chains of Ile831 and Ile881. All these steric clashes can only be minimized
when the distance between the Val882 and the morpholine O-atom is significantly increased to >3.5 Å, as
compared to a 2.5 Å distance in the PQR309 (1)-PI3K complex, and therefore no longer represents a strong
hydrogen bond. Steric clashes, and as a consequence the weakening of the essential hydrogen bond to the
PI3K hinge region explain the reduced affinity of PQR620 (3) for PI3K.
1
4
ACS Paragon Plus Environment

Page 15 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Steric clashes of PQR620 (3) docked into PI3K (light grey). Two opposite conformations of the ethylene
bridges of the morpholines of PQR620 (3) that allow the morpholine oxygen to form a hydrogen bond with the hinge
Val882 nitrogen (N) are shown. A) In the “bridge-up” conformation, the ethylene bridge of PQR620 (3) (yellow) points
27
towards Met953, and is flanked by a rigid part of PI3K , represented by Tyr867, F961 and I963 (colored in brown). B)
When the surface of PI3K is depicted as a mesh and PQR620 (3) as space filling model (atom size set to 0.9 times of
Van der Waals radii), steric clashes become visible as penetrations of the PI3K surface mesh. C) In the “bridge-down”
conformation (pink), steric clashes imposed by the ethylene bridges in PQR620 (3) occur with the backbone of Ile881 and
E880 and side chains of Ile831 and Ile881. As for Fig. 2, the models were derived from coordinates of a PQR309 (1)–
PI3K complex23 (PDB ID code: 5OQ4; resolution of 2.7 Å). A comparative listing of amino acid residues depicted here
across PI3K isoforms can be found in Table S2.
Analogous computational modeling studies were performed for a PQR620 (3)-mTOR kinase
complex starting with the 3.6 Å resolution structure of the PI103-mTOR complex (PDB ID 4JT6). After a
ligand exchange, the oxygen atom of one bridged morpholine of PQR620 (3) forms a hydrogen bond with
the hinge region Val2240 backbone nitrogen and the 2-amino moiety interacts with Asp2195/2357 side
chains in the affinity binding pocket of mTOR. The binding pocket of mTOR is deeper due to the presence
of the flexible Leu2354 replacing the rigid Phe961 in PI3Kγ (see Figure 2). As a result, an ethylene-bridged
morpholine (M ) can be easily accommodated in the pocket pointing towards Val2240. Although these
3
modeling efforts suggest some preference for the "bridge-up" conformation in mTOR, the low resolution
of the mTOR template structure prevents a precise placement of the ethylene bridge in the morpholine
pointed towards Val2240 (Figure S1).
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Glu2190 in mTOR can be superimposed with Asp836 of PI3K, but Glu2190 has a different vector
allowing it to interact with the 2-amino moiety (3.1 Å) and the dipole of the CHF -group (3.4 Å). This view
2
is supported by biological data pointing to an important role of the CHF -subtituent in the selectivity of
2
mTOR binding. The combination of the Glu2190 CHF -group dipole interaction in mTOR and the steric
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
clashes in PI3K explain the enhanced binding affinity for mTOR and the reduced affinity of PQR620 (3)
for PI3K as compared to PQR309 (1).
In order to ascertain the choice for a lead compound, the SAR study was extended to pyrimidine
24
core analogs of PQR620 (3) (Table 3). Therefore, we prepared BKM120 (27), its regioisomer 29 and the
corresponding 4-(difluoromethyl)pyridin-2-amine analogs 28 and 30. Replacement of trifluoromethyl by
difluoromethyl as R-substituent improved mTOR binding by a factor of 5-20 for pyrimidine analogs in
vitro (Table 3), validating the CHF -dipole mTOR interactions discussed above. Still, both pyrimidine core
2
regioisomers 28 and 30 showed reduced potency compared to their triazine analog 2 (compare Table 1).
Pyrimidine core derivatives of PQR620 (3), such as compounds 32 and 34, displayed higher
selectivity for mTOR as compared to their CF -substituted analogs (31 and 33). The pyrimidine core
3
compounds bearing substituted morpholines were inactive towards PI3K, but triazine-based PQR620 (3)
surpassed its pyrimidine analog 32, regarding selectivity and potency towards mTOR in vitro and in cells.
1
6
ACS Paragon Plus Environment

Page 17 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 3. SAR Study of Selected Pyrimidine Core Compounds.
Mn
X
Y
R
N
M0 =
M3 =
O
N
O
N
Mn
N
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH2
aIC50 [nM]
bK
i
[nM]
mTOR
i
Ratio of K s
p110α/mTOR
Compound
R
M
n
X
Y
pPKB/Akt
pS6
p110α
BKM120
CF
3
M
0
0
0
0
3
3
3
3
N
CH
CH
N
416
207
729
243
520
287
3318
650
553
184
926
256
318
164
1944
395
20
18
199
35
0.10
0.51
0.045
0.69
73.0
174
(
27)
8
2
2
3
CHF
2
2
2
2
M
M
M
M
M
M
M
N
9*
0
CF
CHF
CF
CHF
CF
CHF
CH
CH
N
41
911
42
3
N
29
31
32
3
CH
CH
N
2190
2090
5668
2230
30
N
12
3
3
3
4
3
CH
CH
435
60
13.0
37.2
N
aPKB/Akt phosphorylation on Ser473, phosphorylation of S6 on Ser235/236 and in vitro binding to the ATP-binding site of p110α
and mTOR were analyzed as in Table 1. Compound is referred to as BKM120-R1 (regio-isomer of BKM120) in Ref. 24. SDs are
depicted in Table S9.
*
Chemical Synthesis.
Triazine analogs described above and substituted with two unsubstituted morpholines (M0)
(
PQR309 (1), 2, 4-9) or with one morpholine M and a M -morpholine (compounds 10-12) were prepared
0 1-3
according to the synthetic route A depicted in Scheme 1 starting from commercially available 4-(4,6-
dichloro-1,3,5-triazin-2-yl)morpholine (35). Morpholine derivatives were introduced as M -substituents by
n
nucleophilic aromatic substitution reaction (for intermediates 37-39, with M ≠ M ) and a subsequent
n
0
palladium catalyzed Suzuki coupling between the desired organoboronic ester and cholorotriazine
intermediate 36-39 gave the inhibitors PQR309 (1), 2 and 4-12. The synthetic strategy B was used to prepare
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
di-M -substituted triazines (PQR620 (3) and compounds 13-26) and route C to synthesize pyrimidine core
n
analogs (compound 28-34). All three synthetic routes only differ by the reaction conditions used. Overall,
this strategy represents a robust and modular synthetic route giving access to a broad variety of compounds.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. Synthesis of Tri-substituted Compounds with Triazine and Pyrimidine Corea
aReagents and conditions: (i) morpholine derivative (M
boronic acid pinacol ester or 2-amino-4-methyl-pyridine-5-boronic acid pinacol ester, Pd(dppf)Cl
dimethoxyethane / H O, 90 °C, 16 h (for 4, 5); (iii) 1) Boronic ester 72, Pd(OAc) / PPh (cat.), K CO , THF / H
HCl, H O, 55 °C, 16 h (for PQR309); (iv) Boronic ester 64, XPhosPdG2 (cat.), K PO , dioxane / H O, 95 °C, o/n (for 10, 11, 12,
4, 17, 18, 23, 24, PQR620); (v) Boronic ester generated in situ, XPhosPdG2 (cat.), K PO , dioxane / H O, 95 °C, 3.5 h; 2) TFA,
CH Cl , add. at 0 °C, rt, 6 h (for 8); (vi) 1) Boronic ester 64 or 72 or boronic ester generated in situ, XPhosPdG2 (cat.), K PO
dioxane / H O, 95 °C, 3-16 h; 2) HCl, dioxane / H O, 60 °C, 3-16 h (for 2, 6, 7, 13, 15, 16, 19-22, 25, 26, 28, 30-34; for 9 the
reaction was carried out in ethanol at 80 °C); (vii) M -H, DIPEA, CH Cl , 0 °C → rt, o/n (for 41-47, 52, 53, 55); (viii) M -H,
DIPEA, THF, add. at 0 °C, 70 °C, o/n (for 48-51, 54); (ix) morpholine derivative (M -H), DIPEA, EtOH, Δ, o/n (for 57-60); (x) 1)
Boronic ester 72, Pd(OAc) / PPh (cat.), K CO , dioxane / H O, 80 °C, 16 h; 2) HCl, H
-H), DIPEA, EtOH, 0 °C → rt, o/n (for 37-39); (ii) 2-aminopyridine-5-
(cat.), Na CO , 1,2-
O, 55 °C, 2 h; 2)
n
2
2
3
2
2
3
2
3
2
2
3
4
2
1
3
4
2
2
2
3
4
,
2
2
n
2
2
n
n
O, 60 °C, 18 h (for BKM120, 2924).
2
2
3
2
3
2
1
8
ACS Paragon Plus Environment

Page 19 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Enzymatic and Cellular Profiling, Determination of Selectivity, and in vitro ADME.
Due to its superior activity and mTOR selectivity PQR620 (3) was chosen for further
characterization. DiscoverX KINOMEscanTM assays confirmed PQR620 (3)'s excellent mTOR selectivity:
PQR620 (3) was ~3700-fold more potent for mTOR as compared to class I and class III PI3K isoforms, and
did not inhibit PI4K even at excessive concentrations (Table 4). The selectivity of PQR620 (3) for mTOR
over PI3K thus exceeds competitor compounds such as INK128 (73) (~40x), CC223 (77) (~80x) and
AZD2104 (78) (~230x), and is closely followed by SB2602 (74) with a selectivity of ~3100 fold.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 4. mTOR and Lipid Kinase Binding Constants of PQR620 (3) and Reference Compounds.
a
Inhibitor binding constants K
d
, [nM]
PI3K
23000 18000
Fold selectivity
*PI3Kx / mTOR
~3700x
Kinase 
mTOR
0.27
0.092
0.85
3
PI3K
1000
15
PI3K
22000
81
PI3K
PI4K
>30000
n.d.
VPS34
2750
8200
8300
n.d.
PQR620 (3)$
INK128 (73)
SB2602 (74)
PP242 (75)b
CC223 (77)
AZD2014 (78)
30
3.7
~40x
2700
150
20000
120
>30000 22000
>30000
940
~3100x
31
6200
1500
42
7100
8400
~10x
28
2300
33
18000
3300
39
2500
23000
~80x
0.14
>30000
~230x
a)Dissociation constants (K
indicated compounds. K is the mean value from experiments performed in duplicate and were calculated from standard dose
response curves using the Hill equation (n≥2). b)Dissociation constants (K ) of PP242 (75) are reprinted from Ref. 28 with the
) were determined using ScanMax technology (DiscoveRx) with 11 point 3-fold serial dilutions of the
d
d
d
consent of DiscoverX Corporation. n.d.: not determined. *Fold selectivity: ratio of K
d
of most sensitive class I PI3K isoform
s for
$
(displayed in bold type) over K
d
for mTOR. A preliminary dataset has been compared with Rapamycin and Everolimus (K
d
the FKBP12 site were 0.6 and 2 nM, respectively) in Ref. 29. Standard deviations are reported in Table S3, except for cited values
of PP242.
Moreover, compound PQR620 (3) showed negligible off-target effects when screened against a
DiscoverX scanMAXTM kinase assay panel containing a set of 456 diverse wild-type kinases (Figure S2).
At a concentration of 10 μM, PQR620 (3) and SB2602 (74) reached outstanding selectivity scores (S(10)
of 0.005 and 0.003, respectively; Table S4), while values of INK128 (73) and PP242 (75) exceeded 0.1 for
S(10) [S(10) = number of hits /number of tested kinases (at a threshold of 10% of control; see Table S4)].
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The in vitro pharmacology profile of PQR620 (3) at 10 µM was determined in a panel of assays
covering a broad range of targets including receptors, ion channels, transporters, enzymes and second
messengers. No off-target binding activities were identified and PQR620 (3) showed excellent selectivity
versus unrelated receptors and ion channels including hERG (Figure S3). Detectable signals for COX1 and
PDE4D2 inhibition remained below 50% (Figure S4).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Cellular Activities.
Compound PQR620 (3) was tested in vitro across a panel of tumor cell lines including glioblastoma
(A-172, T98G and “U-87 MG”), neuroblastoma (SK-N-AS and SK-N-FI) and malignant melanoma (A375,
COLO 829 and RPMI-7951) cancer cell lines (Figure 4 and Table S6). Growth inhibitory activity was
9
5
evaluated after an exposure time of 72 hours with inhibitor concentrations ranging from 10 to 10 M.
Compound PQR620 (3) showed potency across these cell lines with a mean IC value of 0.92 µM for
5
0
growth inhibition. An analysis of the most frequently mutated cancer genes in the cell panel did not yield a
significant and relevant association of mutations and susceptibility to the mTOR inhibitor (Figure S6).
1.0
0
0
.5
.0
IC50, mean = 919 nM
-
0.5
1.0
-
Figure 4. Waterfall plot representing the potency of PQR620 (3) in a full NTRC (Netherlands Translational Research
Center B. V.) Oncolines™ cell panel of 66 cancer cell lines. Concentrations of half-maximal growth inhibition (IC50) for
PQR620 (3) were obtained from dose−response growth curves, and individual IC50 values of a cell line was related to the
mean IC50 of all cells lines; cell lines were sorted by lowest to highest sensitivity for PQR620 (3) from left (SK-N-FI) to
right (A-498). Individual cell lines and values (IC50 and GI50) are given in Table S6.
2
0
ACS Paragon Plus Environment

Page 21 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
When compared with other mTOR kinase inhibitors such as PP242 (75), SB2602 (74) and the PI3K
inhibitor GDC0941 (76), PQR620 (3) shows a very similar potential to prevent cell growth and proliferation
in two reference cell lines A2058 and SKOV3 (Figure 5). INK128 (73) is more potent, but exceeds the
expected ca. three-fold difference defined by its in vitro potency for mTOR by far, which could indicate
that PI3K inhibition (Table 4) or other off target kinase actions (Figure S2) contribute to its anti-proliferative
potential.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Due to a loss of PTEN (Phosphatase and Tensin homolog) and an activating B-Raf mutation A2058
cells are rather resistant to PI3K inhibition23,30, while growth of SKOV3 cells is promoted by a
constitutively activated PI3K, rendering these cells more sensitive to PI3K and mTOR kinase inhibitors.
Even for excessive concentrations of Rapamycin (10 µM), no attenuation of cell proliferation could be
observed, indicating that the inhibition of mTORC1 is insufficient to block proliferation of A2058 and
SKOV3. Cell cycle analysis revealed that all tested PI3K and mTOR kinase inhibitors accumulated cells in
G1, diminishing cell numbers in S and G2/M cell cycle phases (Figure 5 and Figure S5). Moreover, at 10
µM PQR620 (3) did not show cytotoxicity in 65 of 66 cell lines of the NTCR panel, a typical feature for
the blockage of the PI3K/mTOR pathway.24
Growth inhibition goes hand in hand with the inhibition of mTORC1 and mTORC2 targets, where
PQR620 (3), INK128 (73), PP242 (75), SB2602 (74), and GDC0941 (76) efficiently block the
phosphorylation of S6, S6K, 4E-BP1 and PKB/Akt (Figure 5 and Figure S5). As observed earlier for PI3K
24
inhibitors, also the mTOR inhibitors yield half-maximal growth inhibition only when mTOR targets are
close to 90% inhibited. While Rapamycin is highly efficient to block phosphorylation of mTORC1 targets
such as S6K (and therefore S6) and 4E-BP1, it leads to an increase of pPKB/Akt in A2058 and SKOV3
cells. This corresponds to a previously reported feedback loop, where rapalog-dependent mTORC1 and
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
S6K inactivation enforces surface receptor to PI3K signaling,31 which explains the reduced anti-
proliferative activity of rapalogs as compared to mTOR kinase inhibitors.
A
B
Figure 5. Proliferation, cell
cycle and cellular signaling.
A2058 A) and SKOV3 B)
cells were exposed to
indicated inhibitors for 72h;
cell numbers are shown as %
of DMSO controls. The
Dashed line indicates cell
numbers at t = 0 h, as % of
cells of DMSO control at 72
h (n = 9, 3 independent
experiments (ie); Rapamycin
and GDC0941 (76) n = 6; 2
ie; each data point measured
at least in triplicate, mean ±
SEM. Cell cycle distribution
of C) A2058 and D) SKOV3
cells after 24 h exposure to C)
PQR620
INK-128
PP242
SB2602
SKOV3
Rapa
GDC0941
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A2058
1
00
50
0
100
50
0
-9
-8
-7
-6
-5
-9
-8
-7
-6
-5
Log Inhibitor, [M]
Log Inhibitor, [M]
C
D
DMSO
PQR620
A2058
INK-128
PP242
SB2602
SKOV3
Rapa
GDC0941
5
μM or D) 2 μM of drugs or
DMSO. After Hoechst33342
staining, cell cycle profiles
were determined by FACS.
Quantification and statistics
are provided in Supporting
Information in Figure S5. E)
A2058 and F) SKOV3 cells
were incubated for 1 h with
E
F
PQR620
INK-128
PP242
SB2602
Rapa
GDC0941
150 A2058
150 SKOV3
compounds
and
then
1
00
100
50
0
subjected to in cell western
detection of phosphorylation
of PKB/Akt and S6.
Phospho-protein signals are
presented as % of DMSO
controls (n=3, mean ± SEM);
Immunoblots of phospho-
PKB/Akt, S6, S6K and 4E-
BP1 are on Figure S5.
5
0
0
-10 -9 -8 -7 -6 -5
-10 -9 -8 -7 -6 -5
1
50 A2058
150 SKOV3
1
00
100
50
0
50
0
-10 -9 -8 -7 -6 -5
Log Inhibitor, [M]
-10 -9 -8 -7 -6 -5
Log Inhibitor, [M]
Metabolic Stability.
2
2
ACS Paragon Plus Environment

Page 23 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To predict the metabolic stability of PQR620 (3), the compound was incubated in vitro with
microsomes and hepatocytes from different animal species. Compound PQR620 (3) was found to be very
stable in rat liver microsomes (no metabolites detected), whereas it was moderately metabolized in mouse
liver microsomes (~ 70% remaining PQR620 (3) after a 30 min incubation, Table 5). Similarly, compound
PQR620 (3) was more stable in rat hepatocytes (RH) than in mouse hepatocytes (MH) (CLhep = 1.850
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
6
μL/min/10 cells for RH and CL = 2.671 μL/min/10 cells for MH).
hep
Table 5. Stability of Compound PQR620 (3).
aMicrosome stability, [% remaining
after 30 min]
Hepatocyte
6
b
Stability t1/2 [min]b
Clearance [μL/min/10 cells]
Rat
08 ± 3.3
Mouse
70.2 ± 0.33
Rat
1.85
Mouse
2.67
Rat
468.4
Mouse
324.4
1
aExperiments were carried out by Aphad S.r.l. with rat and mouse liver microsomes (Xenotech). Each experiment performed n =
2
. bExperiments were carried out by Pharmacelsus GmbH using cryopreserved hepatocytes.
In addition, PQR620 (3) showed moderate passive permeability (52.23 ± 7.35 nm/s) in a PAMPA
assay (parallel artificial membrane permeability assay), suggesting non-limiting cellular permeability. As a
further indicator for bioavailability the thermodynamic solubility was assessed over a range of pH, as well
as in fasted and fed state simulated intestinal fluids (FaSSIF and FeSSIF). The solubility of compound
PQR620 (3) was 20.0 ± 6.5 mM at pH 1.2, 33.7 ± 4.3 μM at pH 6.8 and reached 335 ± 148 µM in FeSSIF
buffer (Table 6). Overall, PQR620 (3) displays good ADME properties and promising physicochemical
characteristics, such as the experimental logD value (3.47 in PBS buffer pH 7.4), the partition coefficient
clogP (3.06) and molecular weight (445.47). Moreover, the CNS MPO score, an algorithm used to estimate
blood-brain barrier permeability, was 3.8 for PQR620 (3), which suggests good penetration to the brain.
Table 6. Physicochemical Properties of Compound PQR620 (3).
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Thermodynamic solubility, [μM]
CNS
MPO
Score
Permeability
Papp [nm/s]
logD pH
7.4
Mol.wt.
clogPc [g/mol]
pH 1.2
pH 4.5
pH 6.8
FaSSIFa
FeSSIFb
5
±
2.23
7.35
19967
± 6487
78.2
33.7
66.0
335
± 148
3.47
3.06
445.47
3.8
± 15.5
± 4.3
± 3.2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
aFaSSIF: Fasted Simulated Intestinal Fluid. bFeSSIF: Fed Simulated Intestinal Fluid. clogP values were calculated using
c
MarvinSketchTM. Solubility assays were performed by Aphad Analytical Solutions.
In Vivo Pharmacokinetic and Pharmacodynamic Evaluation.
Compound PQR620 (3) was then profiled in vivo in rats and mice to evaluate PK and PD. After a
single oral dose of PQR620 (3) (10 mg/kg) given to healthy female Sprague Dawley (SD) rats, PQR620 (3)
concentrations in plasma and brain were monitored over time (Figure 6): the maximal concentration (Cmax
of PQR620 (3) was reached in plasma and brain after 30 min (1355 ng/g and 1310 ng/g, respectively). After
hours, the total exposure AUC0-8 was 409 µg*h/ml in plasma and 404 µg*h/ml in brain. A comparison
)
8
with SB2602 (74) showed that its plasma Cmax exceeded the one of PQR620 (3), while plasma AUC levels
were relatively comparable. Brain levels of PQR620 (3) were maintained at ca. 900 ng/g for > 8 h, for
SB2602 (74) they dropped to ca. 170 ng/g after 4 hours. PQR620 (3) displayed approximately a ~1:1
distribution in brain/plasma, while brain access of SB2602 (74) was limited (brain/plasma levels ~1:5.6;
Figure 6).
Dosed at 50 mg/kg in mice, a single oral application of PQR620 (3) yielded a Cmax of 4835 ng/g in
plasma and 7706 ng/g in brain (see Table S7), confirming the excellent brain penetration of PQR620 (3)
(brain/plasma levels ~1.6:1). A brain tissue binding assay revealed an fu (fraction unbound) value of 3%,
which matches the levels of compounds with established brain activity in the same assay (see diazepam; fu
~
4%, Table S8A). Brain penetration data are also in accordance with MDCK permeability assays, which
document a rapid, passive diffusion, independent of P-gp (P-glycoprotein 1, MDR1) transport (see Table
S8B).
2
4
ACS Paragon Plus Environment

Page 25 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
23
As reported earlier, inhibitors targeting PI3K trigger an acute peak of plasma insulin. PQR620
(
3) and SB2602 (74) did not lead to an increase of plasma insulin concentrations in rats and mice as
compared to background levels, while the pan-PI3K inhibitor PQR309 (1) at the same dose produced a
prominent insulin peak. In rats, glucose levels could not be distinguished from background after PQR620
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
3) administration, while PQR309 (1) lead to a slow glucose increase, which even extends beyond the 8 h
23
measurements presented here. In contrast, an elevated dose and exposure of PQR620 (3) in mice caused
an elevation of glucose plasma levels that were delayed in comparison to the increase triggered by PQR309
(
1). While PQR309 (1) response pattern with a glucose and insulin increase can be explained by the
induction of insulin resistance by suppression of GLUT4 in muscle and adipocytes, the glucose increase
triggered by mTOR kinase inhibitors is less well defined. It was proposed that mTOR kinase inhibitors
32
acutely cause hyperglycemia via TORC2-mediated regulation of glycolysis in skeletal muscle. PQR620
(
3) displayed a good exposure after oral dosing and excellent brain permeability, which makes this
compound suitable as potential agent for CNS cancer indications and neurological disease.
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A
Brain: PQR620, SB2602
Plasma: PQR620, SB2602
B
Brain: PQR620, PQR309
Plasma: PQR620, PQR309
Figure 6. PK and PD
in female Sprague
Dawley rats (left
panels) and male
1
1
1
04
104
03
103
C57BL/6JRj
mice
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(right panels). A,B)
plasma and brain
levels (broken lines) of
02
102
0
5
10
0
5
10
C
E
D
F
PQR620 SB2602
PQR620 PQR309
2
.0
.5
.0
.5
2.0
1.5
1.0
0.5
0.0
indicated
drugs
1
determined after
a
1
0
single dose p.o. of 10
mg/kg in rats, and 50
mg/kg in mice*. C,D)
0.0
0
5
10
0
5
10
brain/plasma
ratio
Insulin:PQR620, PQR309
Insulin: Vehicle1,2
over time calculated
from values above.
Plasma E,F) insulin
and G,H) glucose
levels measured in
animals above. All
experiments depicted
as mean ± SEM (n =
3); error bars are
omitted where smaller
PQR620 SB2602 PQR309
100
100
50
0
5
0
0
0
5
10
0
5
10
G
H
Glucose:PQR620, PQR309
Plasma: PQR620, PQR309
PQR620 SB2602 PQR309
4
0
0
40
30
20
10
0
8
6
4
2
0
3
than
symbols.
20
10
0
*PQR620 data in B)
was compared to
0
5
10
0
5
Time [h]
Time [h]
Rapamycin
and
Everolimus in Ref. 29.
Evaluation of in vivo Efficacy.
The in vivo efficacy of compound PQR620 (3) was pre-clinically evaluated in a xenograft tumor
model (Figure 7). Compound PQR620 (3) showed very good tolerability in mice (MTD ~ 150 mg/kg), and
was well tolerated in 14-day toxicological study in rats, where the MTD was defined to be ~ 30 mg/kg.
PQR620 (3) was efficacious in an OVCAR-3 human ovarian cancer xenograft model, where BALB/C nude
2
6
ACS Paragon Plus Environment

Page 27 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mice were inoculated subcutaneously with cis-platin resistant OVCAR-3 cells. Compound PQR620 (3) and
PQR309 (1) at 50 mg/kg were well tolerated in tumor bearing mice, and did not cause a significant body
weight loss, except for an outlier in the PQR620 (3) treatment group (1 out of 8 mice) where one mouse
lost >10% of its weight (Figure 7). Tumor growth and terminal tumor weight were attenuated by treatment
with compounds PQR620 (3) and PQR309 (1) by >50%. This demonstrates that mTOR kinase inhibition
can be equivalent to the inhibition of upstream PI3K in some settings.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A
B
C
Vehicle
PQR620, 50 mg/kg
PQR309, 50 mg/kg
1000
1
500
000
500
2
5
7
50
00
1
$$ $$
5
20
*
$
$
$
$
$
250
0
*
15
le
0
9
20
30
40
20
30
40
i
c
2
0
6
3
h
e
R
R
Time after tumor inoculation, [days]
Time after tumor inoculation, [days]
V
Q
Q
P
P
Figure 7. OVCAR-3 human ovarian cancer xenograft model in BALB/c nude mice: tumor cells were subcutaneously
inoculated at day 0, and daily oral application of the indicated agents was started at day 17 (28 x QD). (A) Body weight
was determined at the depicted time points. Body weight loss of >10% was observed in one mouse (of 8) in the PQR620
(3) treatment group. (B) The tumor size was measured and calculated as described in the experimental section. (C) The
tumor weight was determined at day 45 after inoculation. Statistics: (*p<0.05; p<0.001; $$p<0.0001; n=8; (A,B) Two-
Way ANOVA with Bonferroni or Dunnett's correction for multi-comparisons, mean ± SEM; (C) mean ± SD; One-Way
ANOVA with Tukey's correction for multi-comparisons.
$
Attenuation of Neurological Disease.
It has been previously reported that Tsc1GFAP conditional knockout mice (Tsc1flox/flox x GFAP-
Cre) with targeted disruption of the Tsc1 gene in glial fibrillary acidic protein containing cells (mostly
neurons) show a high incidence of spontaneous postnatal epilepsy. Treatment with rapamycin reduced the
number of seizures and mortality.33
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 81
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Here we evaluated the capacity of PQR620 (3) and PQR309 (1) to rescue the neuronal loss of
function Tsc1 phenotype: mice treated with vehicle (post-natal day 21 to 55) suffered robust electrographic
seizures, while a daily treatment of PQR620 (3) (100 mg/kg, p.o.) eliminated seizures almost completely
(Figure 8). The reduction of seizures by PQR309 (1) did not reach significance. It is tempting to speculate
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
PQR309’s (1) capacity to inhibit mTOR was insufficient, and that targeting of class I PI3Ks cannot remedy
the TSC phenotype. It is possible, however, that it is PQR309’s (1) short half-life in mice that contributes
to diminish its effectiveness. The protective effect of PQR620 (3) demonstrates clearly, that targeting
TORC1 and TORC2 mTOR kinase activity efficiently counteracts TSC-dependent over-activation of the
mTOR pathway and thereby caused epileptic seizures.
A
B
Vehicle
PQR620
**
PQR309
15
2.0
*
1.5 **
1.0
*
1
0
5
0
0.5
0.0
35-42
43-48
49-55
Postnatal day
Figure 8. Effects of mTOR and PI3K inhibition on frequency of electrographic seizures caused by
neuron-specific elimination of TSC1.33 A) Tsc1GFAP conditional knockout mice were treated from
postnatal day 21 to 53 with vehicle, PQR620 (3) (100mg/kg) or PQR309 (1) (50mg/kg). In three
observation intervals, EEG traces were recorded, and the number of seizures were determined (depicted
as seizure count/48h; mean ± SEM, n = 30-48. Two-way ANOVA test with Dunnet’s multiple
comparison related to vehicle control; *p = 0.032; **p = 0.0021; ***p = 0.0002). B) Body weights
averaged over the whole observation time in the three treatment groups (mean ± SEM, n  13).
2
8
ACS Paragon Plus Environment

Page 29 of 81
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CONCLUSIONS
In summary, we demonstrated that lead compound PQR620 (3) potently inhibits mTORC1/2 in
vitro and in vivo. A > 1000-fold selectivity towards mTOR over PI3Kα in binding assays is obtained by the
introduction of a difluoromethyl-substituent and sterically demanding 3,5-ethylene bridged morpholines.
PQR620 (3) exhibits excellent selectivity over a wide panel of protein and lipid kinases, as well as excellent
selectivity versus unrelated receptor enzymes and ion channels. PQR620 (3) demonstrates its potency to
inhibit cancer cell line proliferation with a potential comparable to pan-PI3K inhibition. The physico-
chemical properties (LogD, solubility, liver microsomal stability (t ), and apparent permeability) of
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
/2
compound PQR620 (3) resulted in a good exposure in both plasma and brain following oral administration.
PQR620 (3) shows significant anti-tumor effects in a xenograft mouse model, is well tolerated in rats and
mice, and lacks some of the adverse mechanism-based side effects of pan-PI3K inhibitors. The almost
complete elimination of seizures in the mouse TSC null model above, and a significantly increased seizure
29
threshold in a mouse model of chronic epilepsy demonstrate the efficiency of PQR620 (3) in the CNS.
The excellent brain/plasma partitioning promises a therapeutic potential in a wider range of
neurodegenerative diseases and epilepsy with a reduced impact on host defense responses as compared to
competitor compounds that require higher plasma levels to reach mTOR inhibition in brain tissue.
Altogether, the preclinical data support a further development of this compound.
2
9
ACS Paragon Plus Environment