Discovery of a Novel Inhaled PI3Kδ Inhibitor for the Treatment of Respiratory Diseases

Article abstract of DOI:10.1021/acs.jmedchem.8b00873

Oral PI3Kδ inhibitors such as Idelalisib and Duvelisib have shown efficacy as anticancer agents and Idelalisib has been approved for the treatment of three B-cell cancers. However, Idelalisib has a black box warning on its product label regarding the risks of fatal and serious toxicities including hepatic toxicity, severe diarrhea, colitis, pneumonitis, infections, and intestinal perforation. Some of these side effects are mechanism-related and could hinder the development of Idelalisib for less severe conditions. For respiratory diseases, compounds administered by inhalation are delivered directly to the site of action and may improve the therapeutic index of a drug, minimizing undesired side effects. This work describes the discovery and optimization of inhaled PI3Kδ inhibitors intended for the treatment of severe asthma and COPD. Once the potency was in the desired range, efforts were focused on identifying the particular physicochemical properties that could translate into better lung retention. This medicinal chemistry exercise led to the identification of LAS195319 as a candidate for clinical development.

Full text of DOI:10.1021/acs.jmedchem.8b00873

Subscriber access provided by REGIS UNIV
Article
Discovery of a novel inhaled PI3K# inhibitor
for the treatment of respiratory diseases
Montse Erra, Joan Taltavull, Francisco Javier Bernal, Juan Francisco Caturla, Marta Carrascal, Lluís Pagès,
Marta Mir, Sonia Espinosa, Jordi Gracia, Maria Dominguez, Mar Sabaté, Stephane Paris, Monica Maldonado,
Begoña Hernández, Mònica Bravo, Elena Calama, Montserrat Miralpeix, Martin D. Lehner, and Marta Calbet
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00873 • Publication Date (Web): 15 Oct 2018
Downloaded from http://pubs.acs.org on October 16, 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 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of a novel inhaled PI3K inhibitor for the
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
treatment of respiratory diseases
*
£
£
£
£
Montse Erra , Joan Taltavull, Francisco Javier Bernal, Juan Francisco Caturla, Marta
£
£
£
£
£
Carrascal, Lluís Pagès, Marta Mir, Sònia Espinosa, Jordi Gràcia, María Domínguez,ˤ Mar
£ø
£
£□
∆
Sabaté,ˤ Stéphane Paris, Mónica Maldonado, Begoña Hernández, Mónica Bravo, Elena
Calama, Montserrat Miralpeix, Martin D. Lehner, and Marta Calbet∆
∆
∆θ
∆≠
£
ˤ
∆
Medicinal Chemistry and Screening, Pharmacokinetics and Metabolism, Respiratory
Therapeutic Area, Almirall R&D, 08980 Barcelona, Spain
ABSTRACT
Oral PI3K inhibitors such as Idelalisib and Duvelisib have shown efficacy as anticancer
agents and Idelalisib has been approved for the treatment of three B-cell cancers. However
Idelalisib has a black box warning on its product label regarding the risks of fatal and serious
toxicities including hepatic toxicity, severe diarrhoea, colitis, pneumonitis, infections and
intestinal perforation. Some of these side-effects are mechanism-related and could hinder the
development of Idelalisib for less severe conditions. For respiratory diseases, compounds
administered by inhalation are delivered directly to the site of action and may improve the
therapeutic index of a drug, minimizing undesired side effects. This paper describes the
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 55
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
discovery and optimization of inhaled PI3K inhibitors intended for the treatment of severe
asthma and COPD. Once the potency was in the desired range, efforts were focused on
identifying the particular physicochemical properties that could translate into better lung
retention. This medicinal chemistry exercise led to the identification of LAS195319 as a
candidate for clinical development.
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
INTRODUCTION
The PI3K (Phosphoinositide 3-kinase) family of lipid kinases comprises of three classes, PI3K
class I, class II, and class III. The PI3K class I is the most characterized, being activated by cell
surface receptors to produce phosphatidylinositol-3,4,5-trisphosphate (PIP3). PIP3 is a key
cellular mediator that activates downstream signaling pathways involved in relevant cellular
processes such as growth, proliferation, chemotaxis, metabolism and survival. The PI3K class I
is further subdivided into four different isoforms PI3K, PI3K, PI3K, PI3K the first two
isoforms being ubiquitously expressed and PI3K and PI3K showing a more restricted
expression pattern that seems to play a major role in leukocytes.1
PI3K is expressed, although not exclusively, in immune cells such as T-cells, B-cells,
neutrophils, mast cells and basophils modulating relevant functions that include cell migration,
cytokine production, antibody production and even steroid resistance.2,3 Such effects in immune
cells suggest that selective pharmacological inhibition of PI3K could be a promising approach
for treating respiratory diseases such as asthma and COPD (chronic obstructive pulmonary
disease).
Considering that oral PI3K inhibitors approved as anticancer agents, such as Idelalisib, have
reported serious adverse effects including hepatic toxicity, severe diarrhea, colitis, pneumonitis,
ACS Paragon Plus Environment
2

Page 3 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
infections and intestinal perforation, our internal approach was focused on reducing systemic
side effects by administering the compounds by the inhaled route. Inhaled drug delivery presents
a huge opportunity for the treatment of respiratory diseases, delivering drugs directly to the site
of action which may result in a rapid onset of desired effects and with a potential reduction or
avoidance of systemic side effects. Thus, despite challenges associated with designing drugs for
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
5
inhaled delivery, this route of administration represents an attractive alternative for the
treatment of these diseases compared to conventional routes of administration. In this respect,
GlaxoSmithKline and RespiVert have both reported inhaled PI3Kδ and dual PI3Kδ/PI3Kγ
inhibitors which have progressed to clinical studies.6
The aim of this paper is to provide an overview of physicochemical properties that may lead to
improved lung retention times of inhaled compounds with the intention of avoiding compounds
with extremely poor solubility. The low solubility approach can initially provide good duration
in the lung but persistence of undissolved material may be irrelevant as these solid particles are
unavailable to elicit a pharmacological effect and may provoke an irritant response if such
particles remain in the airways.
Herein is presented the medicinal chemistry program that allowed for the optimization of a
series of neutral compounds for inhaled administration and led to the discovery of LAS195319 as
an inhaled clinical candidate for the treatment of respiratory diseases such as asthma and COPD.
MEDICINAL CHEMISTRY STRATEGY
Initially the strategy focused on finding compounds with high potency, to support low inhaled
dose, and high plasma clearance to ensure low systemic exposure. Recently, we reported the
7
8
identification of a novel oral PI3Kδ inhibitor 1 (Scheme 1). Similar to IC87114, the co-crystal
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 55
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
structure of 1 and PI3Kδ shows that the ligand also adopts a propeller-shaped conformation in
which the pyrrolotriazinone moiety is sandwiched into an induced hydrophobic specificity
pocket between Trp760 and Met752. Increasing potency by interacting with the affinity pocket
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
9
has already been described by Shokat and Williams at University of California. This pocket is
flanked on the bottom by the side chains of Asp787 and Asp911 and on the top by Tyr813 as
shown in figure 1. Thus, efforts towards the identification of highly potent PI3Kδ inhibitors were
based on introducing functional groups that could interact with the amino acids present in this
region and also give rise to additional clearance routes (e.g. glucuronidation of phenols). Before
starting the SAR work around the affinity pocket, the methyl group present in the central core of
IC87114 was introduced onto the pyrrolotriazinone scaffold of 1 to give compound 2 which led
to both an increase in potency and metabolism (Scheme 1).
Scheme 1. Pyrrolotriazinone scaffold
O
O
O
N
N
N
N
N
N
N
N
HN
N
HN
N
N
N
N
N
N
N
N
N
NH2
NH2
H N
2
1
IC87114
PI3K (HTRF) IC50 = 130 nM
2
PI3K (HTRF) IC50 =3.9 nM
Metabolism (rat/human) = 15 / 24
PI3K (HTRF) IC50 = 1.0 nM
Metabolism (rat/human) = 44 / 80
%
%
Figure 1. X-ray co-crystal structure of the binding pocket of human PI3Kδ containing
compound 1 (PDB ID: 5M6U)
ACS Paragon Plus Environment
4

Page 5 of 55
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
In the figure hydrogen bonds are coloured in green dotted lines and the location of the affinity
pocket is shown by the blue circle.
RESULTS AND DISCUSSION
Table 1 outlines the SAR of the affinity pocket and the impact on both PI3Kδ enzymatic
potency and in vitro clearance for aminopyrimidine containing compounds (named series A).
Selectivity versus other class I PI3K family isoforms is also shown. The enzymatic potency of
each compound on the four class I PI3K recombinant human isoforms was determined by using a
biochemical HTRF assay with a compound pre-incubation time of 30 min. Assays were run at an
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 55
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
ATP concentration equal to the respective Km of every isoform and using a subnanomolar
concentration of the enzyme.
It is clear from the data outlined in Table 1, that incorporating a substituted aromatic or
heteroaromatic ring in the affinity pocket leads, in general, to high enzymatic potency for PI3Kδ
and high selectivity (>100-fold) over the related alpha and beta isoforms. The high enzymatic
potency observed could, in part, be due to the presence of heteroatoms in the aromatic ring that
may establish new interactions with the amino acids present in this pocket. These amino acids
are well conserved between isoforms and so the presence such interactions alone cannot explain
the good selectivity obtained. However, other known propeller-shaped inhibitors which are
described to stabilize a conformational change that opens the above-mentioned hydrophobic
specificity pocket in the active site have good selectivity and this process seems to occur more
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
8
readily in PI3Kδ than in the other isoforms. However, a propeller shape does not always
guarantee PI3Kδ specificity (vide infra).
In order to increase the systemic clearance of compounds, functional groups which could favor
additional routes of metabolism were introduced. Phase II metabolism involves reactions that
chemically change either the drug or oxidative metabolites into compounds (conjugates) that are
soluble enough to be excreted in urine. For this purpose the OH substituent on the aromatic ring
of phenols, well-known to form these conjugates, was introduced onto different positions of the
aromatic group which resided in the affinity pocket. Compound 14, with the OH in the para
position, showed high in vitro glucuronidation but moderate in vitro potency. Docking studies
suggested that in this position there was little potential for new interactions, but heteroatoms in
the meta position could potentially form hydrogen bonds either with Lys779 at one side of the
pocket or with Tyr813 at the opposite side. Indeed, introduction of a hydrogen-bond acceptor
ACS Paragon Plus Environment
6

Page 7 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
atom such as a fluorine (compound 15) in this position increased both the affinity for PI3K and
also the extent of glucuronidation of the parent compound. Aminopyrimidine 16 showed similar
potency to compound 14 but much lower in vitro metabolism. Then, a series of compounds was
prepared in which the phenol was fixed in the meta position and a hydrogen bond acceptor group
was introduced in the other meta position with the intention to interact with both Lys779 and
Tyr813. Such compounds (17-19) had excellent affinity for PI3Kδ and high in vitro clearance.
Compound 20 also showed good PI3Kδ potency, probably due to the capacity of the relatively
acidic sulphonamide NH group and/or the oxygens of the sulphone to form hydrogen bonds in
this pocket. Removing the OH group of compound 20 gave an equipotent compound 21,
indicating that double interaction might not be taking place. Other constrained NH derivatives
were also studied (compounds 22 and 23) leading to less potent compounds. Finally, introduction
of different linkers between the hinge binding motif and the phenol ring was also studied
(compounds 24-26), the carboxamide analogue 26 being the most potent.
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 1. SAR exploration of the affinity pocket (series A)
O
N
N
N
HN
R1
N
N
NH2
b
Co
R1
Enzymatic potency IC (nM)
Metabolism
Microsomes (%)
+NADPH +UDPGA
Liver
50
mpd.
PI3Kδ
PI3K
PI3K
PI3K
(rat/human)c (rat/human)d
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 55
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
1
2
1
-CN
-CN
3.9
1.0
22
12000
2300
8400
640
280
65
16 / 24
44 / 80
22 / 84
-
120
-
4
5
1700
160
21 / 76
HO
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
1
0.5
5.0
1100
87
11
44 / 71
96 / 97
ND / 6
HO
F
N
6
7
67%@
360
120
ND / 27
H2N
N
10M
HO
HO
HO
1
0.6
0.2
0.2
1900
560
110
57
40
-
75 / 84
54 / 55
-
26 / 74
83 / 97
-
F
18
Cl
1
2
9
0
310
10
6.8
CF3
NH
HO
0.6
0.3
360
21
37
15
-
-
O
O
S
2
1
730
1.7
32 / 97
47 / 50
O
NH
O
S
22
3.0
11
8500
140
420
430
ND / 9.6
NH
2
3
27%@
5M
1300
O
O
NH
2
2
2
4
5
22
6600
3000
950
-
HO
HO
S
1.8
8400
45
94 / 100
ACS Paragon Plus Environment
8

Page 9 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
2
6
0.2
1400
39
10
42 / 60
HO
NH
a
b
All assay results are reported as the geometric mean of at least two separate runs. PI3Kδ,  
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  activities were measured with an ATP concentration fixed to the Km of each isoform by
using a HTRF assay, where the PIP3 product formed from PIP2 is detected by displacement of
c
biotin-PIP3 from the complex resulting in a loss of energy transfer (HTRF). NADPH-dependent
metabolism (%) expressed as disappearance of parent compound after microsomal incubation for
d
30 min (1 mg/ml protein and 5 µM compound at 37 ºC). UDPGA-dependent metabolism (%)
expressed as disappearance of parent compound after microsomal incubation for 120 min (1
mg/ml protein and 5 µM compound at 37 ºC).
In parallel, another well-known kinase hinge binding motif, the pyrrolopyrimidine (which led
to series B), was explored as shown in Table 2. In this case ortho-substituted phenols were
tolerated (compounds 32-35) although potency was improved by introducing of additional
hydrogen-bond acceptor groups in the meta position (compound 33-35), as observed in the
previous SAR. The para-substituted phenol 36 was more potent than the corresponding analogue
1
4 and this could be explained by docking studies, which showed a potential new interaction
with Lys779 not observed with analogue 14. Potency could be slightly improved by adding a
hydrogen-bond acceptor atom such as fluorine in the meta position (compound 37). Docking
studies suggested that a new interaction with Asp787 could be gained from this meta position.
However, poor selectivity versus PI3K was observed for this compound, demonstrating that not
all propeller-shaped compounds have good PI3Kδ specificity. One could hypothesize that the
double interaction with the affinity pocket could compensate for the energy required to open the
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 55
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
above-mentioned specificity pocket in other PI3K-isoforms. This lack of selectivity was also
found in compounds 38 and 39, although the latter just fulfilled our selectivity criteria (≥100-fold
vs PI3Kδ). Removal of the phenolic OH-group of 39 gave rise to the equally potent and more
selective compound 40. In this case, the double interaction with Lys779 and Asp787 is not
possible and, although this did not have an impact on PI3K delta affinity, it could explain the
higher selectivity observed versus the other isoforms (vide supra). Phenol bioisosteres such as
indoles were then studied leading to potent compounds 41 and 42. In addition, compounds
containing small heterocycles such us pyrazoles were synthesized giving rise to the highly potent
compounds 43 and 44. Finally, addition of various linkers between the hinge binding motif and
the phenol ring was studied and led to potent compounds 45-47 with high in vitro clearance.
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
O
N
N
N
HN
R1
N
N
Table 2. SAR exploration of the affinity pocket (II) (series B)
NH
b
Com
pd
R1
Enzymatic potency IC (nM)
Metabolism
Liver
50
Microsomes (%)
+NADPH +UDPGA
PI3Kδ
PI3K
PI3K
PI3K
(
rat/human)c (rat/human)d
3
2
5.1
0.2
-
-
-
72 / 99
58 / 99
OH
OH
33
760
14
81
44 / 90
O
3
3
4
5
1.9
0.3
1000
250
52
12
27 / 41
21 / 58
OH
F
F3C
7.3
3.3
OH
ACS Paragon Plus Environment
1
0

Page 11 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
3
6
7
0.4
0.2
110
1.2
4.1
1.3
0.7
0.3
HO
HO
F
O
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
8
0.2
0.2
9
2.3
2.5
0.6
0.7
OH
H
39
N
22
16 / 67
23 / 74
28 / 94
S
O
O
OH
H
N
40
41
0.3
1.3
190
-
29
79
3.1
-
S
O
O
NH
4
2
0.2
410
12
17
74 / 92
HN
43
44
N
HN
0.4
0.2
-
4.3
22
-
21 / 97
36 / 91
N
N
230
3.6
6 / 91
45
1.4
1.0
2.0
0.5
2900
770
210
170
230
10
64
-
72 / 100
64 / 99
HO
S
4
4
5
6
7
0
74 / 84
HO
HO
O
1000
1900
13
36
NH
H
H2N
N
32 / 68
10 / 30
S
O
O
HN
a
b
All assay results are reported as the geometric mean of at least two separate runs. PI3Kδ,  
and  activities were measured with an ATP concentration fixed to the Km of each isoform by
using a HTRF assay, where the PIP3 product formed from PIP2 is detected by displacement of
ACS Paragon Plus Environment
1
1

Journal of Medicinal Chemistry
Page 12 of 55
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
c
biotin-PIP3 from the complex resulting in a loss of energy transfer (HTRF). NADPH-dependent
metabolism (%) expressed as disappearance of parent compound after microsomal incubation for
d
30 min (1 mg/ml protein and 5 µM compound at 37 ºC). UDPGA-dependent metabolism (%)
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
expressed as disappearance of parent compound after microsomal incubation for 120 min (1
mg/ml protein and 5 µM compound at 37 ºC).
The X-ray co-crystal structure of 39 bound to PI3Kδ was resolved at 2.72 Å, revealing a
number of polar interactions in the affinity pocket as was predicted. The OH of the phenol group
forms hydrogen bonds with the side chain atoms of Asp787 and Tyr813, while one of the
oxygens of the sulfone forms a hydrogen bond with the side chain of Lys779. In addition, the
pyrrolopyrimidine forms hydrogen bonds to the hinge region with the main chain atoms of
Glu826 and Val828. Similar to the co-crystal structure obtained with compound 1, the molecule
adopts a propeller-shaped conformation where the pyrrolotriazinone moiety is sandwiched into
the induced hydrophobic specificity pocket between Trp760 and Met752.
Figure 2. X-ray co-crystal structure of the binding pocket of human PI3Kδ containing
compound 39 (PDB ID: 6G6W)
ACS Paragon Plus Environment
1
2

Page 13 of 55
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
In the figure hydrogen bonds are coloured in green dotted lines and distances are shown in blue
(Å).
Since many compounds with good in vitro profiles were found in these initial studies, before
embarking on further SAR work around the affinity pocket, an in vivo pharmacokinetic assay
was used to identify molecules with the highest lung retention. Briefly, five compounds were
pooled and administered in solution to rats via the intratracheal route using cassette dosing. Each
compound was administered at a dose of 0.015 mg/kg; a low enough dose to minimize any
potential solubility issue which could have an impact on the lung retention profile of the
ACS Paragon Plus Environment
1
3

Journal of Medicinal Chemistry
Page 14 of 55
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 assayed (the thermodynamic solubility of some compounds tested in water at pH=7.4
was <0.01 µg/mL) . A total of 64 compounds (20 from series A and 44 from series B) were
tested in this assay and the lung and plasma profiles of the compounds were evaluated. Figure 3
shows the correlation between the amount of compound recovered in the lung (expressed as % of
dose) at the first kinetic time point after administration versus the overall exposure (AUC) in the
lung. From this plot we can conclude that good initial lung recovery does not always translate
into good lung retention, but what seems clear is that compounds that show poor initial lung
recovery never translate into compounds with good lung retention.
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. Plot of the relationship between dose normalized lung exposure (AUC/Dose)
and the amount of compound recovered in the lung (expressed as % of dose) at the first
kinetic time point after administration.
ACS Paragon Plus Environment
1
4

Page 15 of 55
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
The figure was produced using Vortex. Colour indicates compound series (red for series A,
green for series B). Purple line corresponds with a dose normalized lung exposure of 100 µg*h/g
and the blue line corresponds with amount of compound recovered in the lung (expressed as %
of dose) of 50% at the first kinetic time point after administration.
The lung and plasma concentrations at different kinetic time points were studied for each
compound and a ratio between both was then calculated. For compounds intended to be
administered via the inhaled route, the plasma concentrations and overall systemic exposure
should be minimized as much as possible in order to avoid potential systemic side effects.
Figure 4 shows the correlation between the lung/plasma ratio at equilibrium (1 hour after
administration) versus the dose-normalized lung exposure (AUC/Dose) for each compound. In
ACS Paragon Plus Environment
1
5

Journal of Medicinal Chemistry
Page 16 of 55
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
this case a good correlation is observed between both parameters, allowing the selection of
compounds with good overall lung exposure and an optimal lung/plasma split. Many compounds
showed lung/plasma ratios of over a 100-fold and a few even showed ratios of over 1000-fold.
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 4. Plot of the dose normalized lung exposure (AUC/Dose) for each compound
versus the lung/plasma ratio at equilibrium (1 hour after administration).
The figure was produced using Vortex. Colour indicates compound series (red for series A,
green for series B). Blue lines correspond with lung/plasma ratio of 100 and 1000 at 1 hour after
compound administration.
ACS Paragon Plus Environment
1
6

Page 17 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To better understand the physicochemical properties that could influence lung retention,
several parameters were plotted against each other in order to study possible correlations.
First of all, permeability was studied and compared with dose normalized lung exposure as
previously described. As shown in figure 5, compounds with limited permeability: PAMPA < 1
show, in general, better lung retention than those with higher permeability.
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
Figure 5. Plot of the relationship between PAMPA (expressed as cms10 ) and dose
normalized lung exposure (AUC/Dose) for each compound
ACS Paragon Plus Environment
1
7

Journal of Medicinal Chemistry
Page 18 of 55
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 figure was produced using Vortex. Colour indicates compound series (red for series A,
green for series B). Blue line corresponds with a dose normalized lung exposure of 50 µg*h/g
-6
and the purple line corresponds with a PAMPA of 1 *10 cm/s.
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
Other physical properties such as hydrogen bond donor counts and polar surface area have been
1
0
described to be positively correlated with lung retention. Thus, when tPSA was plotted versus
MW (figure 6), it was observed that compounds with higher lung exposure (represented by the
bigger circles) were the ones with MW above 550 and with a tPSA higher than 150. However, as
polar surface area tends to increase with molecular weight, it is difficult to find examples in the
top left and bottom right quadrants and one cannot discard that only one and not both parameters
are required for higher lung retention times. An increase in the number of H-bond donors also
positively impacted on lung exposure as shown in figure 7. However, it should be taken into
account that increasing the number of H-bond donors is likely to compromise the cellular
permeability of compounds for an intracellular target. To this end, the ratio between cellular and
enzymatic potency was calculated for all compounds and a higher proportion of compounds with
ratios > 10 was found for those having 5 H-bond donors (data not shown). Therefore, examples
with 6 H-bond donors were not synthesized.
Figure 6. Relationship between theoretical Polar Surface Area (TPSA) and molecular
weight (MW) vs lung exposure
ACS Paragon Plus Environment
1
8

Page 19 of 55
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
The figure was produced using Vortex. Colour indicates compound series (red for series A,
green for series B). Blue lines corresponds with a Molecular Weight of 550 Da and a theoretical
Polar Surface Area of 150.
Figure 7. Plot of the relationship between number of hydrogen bond donors (HBD)
versus dose normalized lung exposure (AUC/Dose)
ACS Paragon Plus Environment
1
9

Journal of Medicinal Chemistry
Page 20 of 55
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
The figure was produced using Vortex. Colour indicates compound series (red for series A,
green for series B). Dose normalized lung exposure is represented in the y-axis and the number
of hydrogen bond donors in the x-axis.
The time course graphs of compound recovery from lung tissue after i.t. administration of a set
of compounds from both series at 0.015 mg/kg in solution were superimposed and coloured by
the number of H-bond donors present in each molecule (figure 8). The visual data suggests that
increasing the number of H-bond donors leads to a positive impact on the distribution phase,
resulting in a higher lung AUC.
ACS Paragon Plus Environment
2
0

Page 21 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 8. Time course of compound recovery from lung tissue after i.t. administration of
different compounds at 0.015 mg/kg in solution
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
4
HBD
HBD
5 HBD
The figure was produced using GraphPad Prism 7. Colours indicate the number of hydrogen
bonds donors present in each compound (blue for 3 HBD, red for 4 HBD and green for 5HBD).
With this information it can be concluded that, for this series of non-basic compounds, the
physicochemical properties (and the respective optimal values) contributing to higher lung
retention are; PAMPA <1, MW > 550, PSA >150 and H bond donors > 3. For this reason,
chemical efforts for subsequent compounds were focused on the synthesis of molecules
fulfilling the aforementioned parameters.
Following this strategy compound 50 was identified and compared with compound 15, an
equipotent inhibitor. As shown in Table 3, only compound 50 fulfills the criteria previously
described in terms of permeability and physicochemical properties. Such differences should
translate into better lung retention for compound 50 compared to compound 15.
ACS Paragon Plus Environment
2
1

Journal of Medicinal Chemistry
Page 22 of 55
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 3. In vitro profiles of compounds 15 and 50
O
O
N
N
N
F
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
N
N
H
HN
N
H2N
N
HN
N
S
O
N
O
N
NH2
NH
HN
OH
Compound 15
Compound 50
Compd Enzymatic HTRFb
PI3K
M-CSF
p-Aktc
IC (nM) cm/s)d
PAMPA
(
10-6
app *
MW
tPSA
HBD
P
IC (nM)
50
5
0
1
5
0.5/1100/87/11
0.5/1900/10/36
4.4
2.7
1.7
0.2
471
594
119
169
3
5
50
a
b
All assay results are reported as the geometric mean of at least two separate runs. PI3Kδ,  
and  activities were measured with an ATP concentration fixed to the Km of each isoform by
using a HTRF assay, where the PIP3 product formed from PIP2 is detected by displacement of
c
biotin-PIP3 from the complex resulting in a loss of energy transfer (HTRF). THP-1 cells were
treated with compounds for 30 min., stimulated with M-CSF at EC for 3 minutes and then
8
0
d
lysed to measure (by ELISA) pAkt (Thr308) produced through PI3Kδ. Permeability values at
RT in PBS (containing 2% DMSO), compound concentration in donor compartment: 20 µM.
Thus, both compounds were further profiled in rat in vivo PK experiments. As shown in Table
4
, both compounds showed high clearance and high volumes of distribution, resulting in short
terminal half-lives and poor oral bioavailability. These compounds were then administered via
the intratracheal route both in solution and in suspension (Table 5). As predicted, compound 50
ACS Paragon Plus Environment
2
2

Page 23 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
showed a much higher lung exposure compared to compound 15 in solution as well as in
suspension. In addition, a much higher lung/plasma partition ratio was obtained with compound
50, reinforcing our goal of reducing on-target systemic side effects. Lung half-lives were slightly
increased after administration of the compounds in suspension as compared to in solution,
probably due to the dissolution rate of these poorly soluble compounds (thermodynamic aqueous
solubility of 36 µg/mL for compound 15 and 3 µg/mL for compound 50). Compound 50 also
showed an improved lung profile versus other closely related compounds 41 and 42 (data not
shown)– neither of these examples fulfilled the above-defined criteria.
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. Rat PK profiles of compound 15 in comparison with 50
Compd
AUC
Cmax
Cl
Vss
t1/2
F
PPB (%
ng*h/mL)a
(ng/mL)a
(mL/min/Kg)a (L/Kg)a (h)a
(%)b
bound)c
(
1
5
149
171
236
455
112
96
5.2
5.2
1.0
1.7
13
94.6
98.9
50
1%
a
b
Mean values (n=2) in Wistar rat after an administration of 1mg/kg i.v. Mean values (n=2) in
c
Wistar rat after an administration of 1 mg/kg p.o. Mean rat PPB values determined by
equilibrium dialysis at a plasma concentration of 1 µM at 37 ºC.
Table 5. Rat i.t. PK profiles of compound 15 in comparison with 50
Compound
Dose
Route
AUCa
Plasma
Lung /
Plasma
Cmaxa
Plasma
t1/2a
Plasma
(h)
0.7
Lung
Lung
(µg/g)
0.7
Lung
(h)
4.7
(
µg*h/g) (ng*h/mL) ratio
(ng/mL)
36.3
15
0.2 6.4 31
0
.015mg/kg
(it, solution)
ACS Paragon Plus Environment
2
3

Journal of Medicinal Chemistry
Page 24 of 55
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
5
0
0
15.5
30.6
459.6
5.3
2925
207
3.4
1.4
2.8
5.7
4.5
3.2
1.1
3.2
.015mg/kg
(
it, solution)
15
.3 mg/kg
it, susp.)
147.6
96.5
18.6
73.3
2000.7
16.4
0
(
5
0
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
0
4763
.3 mg/kg
(it, susp.)
aMean values (n=2) in Wistar rat after an administration of 0.015mg/kg i.t. in solution or after
an administration of 0.3 mg/kg i.t. in suspension.
Considering the difference in terms of lung retention, a better in vivo efficacy was expected for
compound 50 compared to compound 15. The OVA-induced inflammation model in the brown
Norway rat was selected to assess the in vivo efficacy of both compounds. In this model, an
OVA challenge in sensitized rats causes an allergic inflammatory response that provokes an
influx of inflammatory cells to the animal airways and induces functional lung changes that
mimic certain features of human asthma. Intratracheal administration of compound 50
(
LAS195319) as a suspension at 0.3 and 0.1 mg/kg i.t. 1 h before OVA aerosol challenge caused
an inhibition of eosinophil and neutrophil infiltration in BAL fluid of 55% and 77% respectively
at the higher dose (Figure 9) and 21% and 42% at the lower dose. Fluticasone at 0.3 mg/kg used
as positive control, caused inhibition of neutrophil and eosinophil infiltration of 89% and 82%
respectively. Results obtained with compound 15 demonstrated that it was required to increase
the dose in order to see efficacy, showing no effect in eosinophil infiltration and a modest
inhibition of neutrophil infiltration in BAL fluid of 37% when administered at 1mg/kg. (data not
shown).
At the efficacious dose of 0.3mg/kg for LAS195319, the maximum free drug concentration in
plasma (free Cmax) after i.t. administration was 0.3 nM, not covering the cellular IC50
ACS Paragon Plus Environment
2
4

Page 25 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
determined for this compound and thus, minimizing the potential for side effects associated with
systemic drug exposure.
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 9. Inhibition of inflammatory cell infiltration in the OVA rat model after i.t.
administration of LAS195319.
A)
B)
6
104
105
105
105
3
2
1
4104
*
2
104
***
****
****
0
0
0.3
(mg/kg)
Fluticasone
vehicle
vehicle
0.1
LAS195319
0.3
vehicle
vehicle
0.1
0.3
LAS195319
0.3
(mg/kg)
Fluticasone
OVA challenge
OVA challenge
Inhibition of inflammatory cell infiltration: neutrophils (A) and eosinophils (B) in BALF
induced by OVA challenge in a Brown Norway rat model after i.t. administration of
LAS195319. Two independent experiments were performed n=11-16 per dose. One way
ANOVA, Dunnett's test *, ***, ****: P<0.05, 0,001 and 0,0001 vs vehicle + OVA group
As a consequence of these encouraging results, LAS195319 was further characterized.
Enzymatically, the compound showed higher selectivity versus PI3Kα than for PI3Kβ and PI3Kγ
(Table 2). The IC for cellular inhibition of PI3K delta using THP-1 cells stimulated with M-
5
0
CSF was measured at 2.7 nM for this compound (Table 3). To evaluate the cellular inhibition of
PI3Kβ, an assay based on stimulation of HUVEC cells with sphingosine-1-P was employed.11
The results, 61% inhibition at 10 µM, indicated that the poor enzymatic selectivity between δ
ACS Paragon Plus Environment
2
5

Journal of Medicinal Chemistry
Page 26 of 55
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
and β isoforms (20-fold) was not translated to the cellular level. In addition, LAS195319 was
highly selective against an extensive panel of protein and lipid kinases and GPCRs (see
supporting information). In terms of cardiac safety, LAS195319 did not show any activity on the
hERG channel up to a concentration of 10 M. Finally, LAS195319 confirmed very low oral
bioavailability in a second preclinical specie such as dog (F <1%) as was desired to avoid the
systemic side effects that have been described for oral PI3K inhibitors.
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
Taken together, these properties made LAS195319 an ideal inhaled candidate for clinical
development for the treatment of respiratory diseases. A full biological profile will be published
in due course.
CHEMISTRY
The synthesis of the key intermediate 9 started from the pyrrole carboxylic acid 3, which was
transformed into the amide 4 by formation of the acyl chloride and subsequent reaction with
aniline. N-Amination of the pyrrole 4 was achieved by a biphasic reaction with chloramine
generated in situ from the oxidation of ammonia with sodium hypochlorite using N-methyl-N,N-
dioctyloctan-1-aminium chloride as phase-transfer catalyst. Under these conditions the amination
occurs selectively on the NH of the pyrrole, giving 5 as the sole regioisomer. Condensation of 5
with Boc-protected alanine mediated by propylphosphonic anhydride gave amide 6. The
1
0
cyclization of 6 to the corresponding pyrrolotriazine 8 involved a two-step process. In the first
step, treatment of 6 with the complex formed from triphenylphosphine and bromine in methylene
chloride gave the O-cyclized intermediate 7, which, without isolation, was rearranged to the
more stable pyrrolotriazine 8 by treatment with a 7M solution of ammonia in methanol. Finally,
deprotection of the Boc group with trifluoroacetic acid gave amine 9.
ACS Paragon Plus Environment
2
6

Page 27 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 2. Synthetic route to amino intermediate 9a
O
O
O
a, b
c
d
N
OH
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
H
H
N
NH
NH
NH2
3
4
5
O
O
N
O
N
N
N
N
H
e
f
N
g
N
N
NH
O
N
O
N
N
HN
HN
NH2
Boc
HN
Boc
Boc
6
7
8
9
a
Reaction conditions: (a) oxalyl chloride, DCM; (b) aniline, DIPEA, DCM; (c) NaClO, NaOH, NH
4 4
OH, NH Cl,
water/MTBE, N-methyl-N,N-dioctyloctan-1-aminium chloride; (d) N-Boc-Ala, 2,4,6-tripropyl-1,3,5,2,4,6-
trioxatriphosphinane 2,4,6-trioxide (PPACA), DIEA, EtOAc; (e) PPh
3 2 3
, Br , TEA, DCM; (f) NH /MeOH; (g) TFA,
DCM.
Aminopyrimidine-containing compounds 14 to 26 were prepared according to the synthetic
route described in Scheme 3. Nucleophilic attack of amine 9 on chloropyrimidines 10 in the
presence of cesium fluoride gave compounds of type 12. Reaction of compounds 12 (either
X=Br or the more reactive X=I) with the corresponding boronic acids or boronic esters under
Suzuki reaction conditions yielded final compounds 14-24. Alternatively, thioether 25 was
obtained by a copper-catalyzed arylation of the corresponding aromatic thiol with intermediate
1
2.
The preparation of compound 26, started with the nucleophilic attack of amine 9 on the
protected chloropyrimidine ester 11. In this case, milder conditions to those described above
ACS Paragon Plus Environment
2
7

Journal of Medicinal Chemistry
Page 28 of 55
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
were used - refluxing ethanol in the presence of DIEA was sufficient to complete the reaction.
Subsequent hydrolysis of the ester with lithium hydroxide gave carboxylic acid 13, which was
reacted with an aniline under standard HATU-mediated coupling conditions to give the
corresponding amide and, after a final deprotection of the p-methoxybenzyl group with
trifluoroacetic acid, compound 26 was isolated.
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 3. Synthetic route to compounds 14-26 (Table 1)a
O
N
Cl
N
b
N
N
N
X
O
N
HN
Ar
N
NH2
O
1
0
N
1
4-23
N
N
N
a
N
NH2
N
HN
N
O
N
NH2
9
1
2
N
N
X
X = Br or I
c or d
N
NH2
Cl
N
N
HN
Y
N
EtOOC
O
N
O
NHPMB
Ar
N
11
N
N
NH2
e, f
N
g, h
N
2
2
4 Y= CH2
N
5 Y= S
HN
N
HN
N
H
HO
N
N
N
Ar
O
NHPMB
O
NH2
13
26
a
Reaction conditions: (a) CsF, DIEA, tBuOH; (b) Ar-B(OH)
Ar-CH -Bpin, PdCl dppf·DCM, Na CO , dioxane; (d) Ar-SH, CuI, K
THF/EtOH/water; (g) ArNH , HATU, DIEA, DMF; (h) TFA.
2
or Ar-Bpin, PdCl
2 2 3
dppf·DCM, Na CO , dioxane; (c)
2
2
2
3
2
CO , DCM; (e) DIEA, EtOH; (f) LiOH,
3
2
For the preparation of the compounds with the pyrrolopyrimidine hinge binder 32-47, an
analogous synthetic route was followed (Scheme 4).
ACS Paragon Plus Environment
2
8

Page 29 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Nucleophilic attack of amine 9 on chloropyrrolopyrimidine 27 in the presence of cesium
fluoride gave intermediate 28. Reaction of bromoderivative 28 with the corresponding boronic
acids or boronic esters under Suzuki reaction conditions and followed by SEM deprotection
yielded compounds 32 to 45. The copper-catalyzed reaction of 28 with an aromatic thiol failed
but the desired target 46 was finally able to be prepared by nucleophilic substitution of amine 9
with the pre-functionalized chloropyrrolopyrimidine 29, followed by SEM deprotection. This
deprotection took place in two steps as mainly the corresponding hemiaminal intermediates were
obtained after subjecting the SEM-protected compounds to acidic conditions - subsequent
treatment with ammonia was required to yield the final compounds.
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
The nucleophilic substitution reaction of amine
9
with the unprotected
chloropyrrolopyrimidine ester 30 was conducted in tBuOH. Due to the enhanced reactivity of 30,
the addition of cesium fluoride was not needed in this case for the reaction to reach completion.
Subsequent removal of the benzyl group by Pd-catalyzed hydrogenation gave carboxylic acid 31,
which was reacted with the corresponding aniline after propylphosphonic anhydride activation to
yield amide 47.
Scheme 4. Synthetic route to compounds 32-47 (Table 2)a
ACS Paragon Plus Environment
2
9