Discovery of Potent, Efficient, and Selective Inhibitors of Phosphoinositide 3-Kinase δ through a Deconstruction and Regrowth Approach

Article abstract of DOI:10.1021/acs.jmedchem.8b01556

A deconstruction of previously reported phosphoinositide 3-kinase δ (PI3Kδ) inhibitors and subsequent regrowth led to the identification of a privileged fragment for PI3Kδ, which was exploited to deliver a potent, efficient, and selective lead series with a novel binding mode observed in the PI3Kδ crystal structure.

Full text of DOI:10.1021/acs.jmedchem.8b01556

Subscriber access provided by TULANE UNIVERSITY
Article
Discovery of Potent, Efficient and Selective Inhibitors of Phosphoinositide
3-Kinase # Through a Deconstruction and Regrowth Approach.
Nick Barton, Maire A. Convery, Anthony W.J. Cooper, Kenneth David Down, Nicole Hamblin, Graham
Inglis, Simon Peace, James E. Rowedder, Paul Rowland, Jonathan A Taylor, and Natalie Wellaway
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01556 • Publication Date (Web): 11 Dec 2018
Downloaded from http://pubs.acs.org on December 11, 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 30
Journal of Medicinal Chemistry
1
2
3
4
Discovery of Potent, Efficient and Selective Inhibitors of Phosphoinositide 3‑Kinase δ Through a
5
6
Deconstruction and Regrowth Approach.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Nick Barton, Maire Convery, Anthony W. J. Cooper, Kenneth Down*, J. Nicole Hamblin$, Graham Inglis,
Simon Peace, James Rowedder, Paul Rowland, Jonathan A. Taylor and Natalie Wellaway.
GlaxoSmithKline R&D, Medicines Research Centre, Gunnels Wood Road, SG1 2NY, Stevenage, UK
Abstract
A deconstruction of previously reported PI3Kδ inhibitors and subsequent regrowth led to the identification of a
privileged fragment for PI3Kδ which was exploited to deliver a potent, efficient and selective lead series with a
novel binding mode observed in the PI3Kδ crystal structure.
Introduction
Phosphoinositide 3‑kinase δ (PI3Kδ) and the closely related isoforms PI3Kα, β and γ are class 1 lipid kinases,
which catalyse the phosphorylation of phosphatidylinositol 4,5-bisphosphate to produce the key signalling
molecule phosphatidylinositol 3,4,5-triphosphate, which in turn triggers a series of downstream biological events
that affect cell growth, chemotaxis, differentiation, proliferation and survival1-5. The PI3Kα and β isoforms are
ubiquitously expressed in all tissues and knock-out mice are embryonically lethal, whereas PI3Kδ and γ are
predominantly expressed in leukocytes and mutagenic mice are viable with a phenotype consistent with
immunological effects on B-cells, T-cells, and neutrophils, making them a target for the treatment of inflammatory
conditions, such as asthma and COPD6-8. Many selective PI3Kδ inhibitors have been reported9-10, predominantly
for oncology indications (for example Idelalisib11 (fig 1.)), although AM-8508 (fig. 1) was reported by Amgen as
a potential treatment for inflammatory diseases12, leniolisib was progressed by Novartis into phase III trials for
the treatment of activated PI3K delta syndrome13 and UCB have progressed seletalisib into clinical trials for
patients with primary Sjogren’s syndrome14. We previously reported15 the discovery of highly potent PI3Kδ
inhibitors nemiralisib, 1 and GSK2292767, 2 (fig. 1) for the treatment of respiratory indications via inhalation.
Both these compounds are highly selective for PI3Kδ (≥ 1000-fold) with ideal properties for inhaled delivery (low
bioavailability and moderate to high clearance) and are progressing through clinical trials16-19. In this
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 30
1
2
3
4
5
6
communication, we describe the successful discovery of selective leads for PI3Kδ with different chemical starting
points compared to inhaled clinical compounds 1 and 2, achieved via a deconstruction and regrowth approach.
7
8
F
9
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
N
N
N
N
N
O
F
N
HN
N
NH
O
NH
N
N
F
N
N
N
N
N
F
N
H
N
NH2
Idelalisib
AM-8508
leniolisib
O
N
Cl
N
N
O
N
N
N
O
N+ F3C
O-
NH
N
H
O
S
N
N
N
N
N
N
HN
H
N
O
H
O
N
seletalisib
nemiralisib, 1
2
Figure 1. Selected PI3Kδ inhibitors.
Results and Discussion
We recently reported a lead series developed from an initial small, focussed array20 and herein we describe a
broader approach leading to the discovery of a further lead series. A key focus in this exploration was to identify
PI3K isoform selectivity differences such that subsequent lead optimisation studies would be better placed to
deliver highly selective compounds.
The binding modes of compounds 1 and 2, which had been determined based on homology modelling and later
confirmed by crystallography15, indicated that the indole/pyridine sulfonamide group occupied the affinity pocket
region of the kinase, the indazole made the key interaction with the “hinge” region and the oxazole/alkyl amine
group accessed a region of the kinase known to impart selectivity for PI3Kδ over the other isoforms21. A
deconstruction of compounds 1 and 2 to compounds 3-5 (Table 1) was performed to enable better understanding
of the minimum pharmacophore required for potency. The oxadiazole in compound 5 was chosen in place of the
oxazole primarily for ease of synthesis and, with a pendant morpholine substituent, was previously shown to have
no impact on PI3K potency15.
ACS Paragon Plus Environment

Page 3 of 30
Journal of Medicinal Chemistry
1
2
3
4
pIC50
5
6
7
Compound
Number
Structure
PI3Kδ (LE)
PI3Kα
PI3Kβ
PI3Kγ
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
N
1
9.9*
5.3ab
5.8a
5.2c
N
O
N
HN
N
H
O
N
N
O
2
10.1*
6.4
5.6
6.5
H
N
O
S
N
N
H
O
O
N
N
N
3
4
5.1 (0.39)
6.0a (0.37)
<3.8
5.1
4.6c
5.1
<3.8
5.6
H
HN
H
N
O
S
N
N
H
O
O
N
N
O
N
N
O
5
6
3.8d (0.25)
<4.3
5.2
<4.3
5.4
4.6a
N
N
H
H
N
O
S
N
N
H
6.4 (0.42)
5.6
O
Cl
N
Table 1. Inhaled clinical compounds 1 and 2 (data taken from reference15), deconstructed indazole compounds
3-5 and compound 6. n ≥ 3. *pKi reported from the standard assay ran at a 2 mM ATP concentration. a: Tested
pIC50 < 4.6 on one occasion. b: Tested <5.3 on one occasion. c: Tested pIC50 <4.6 on three occasions. d: n=1,
tested at high concentration to determine a pIC50.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 30
1
2
3
4
5
6
7
8
9
Compound 3 was an efficient fragment compound; however, it was approximately 10-fold less potent than
compound 4. We proceeded to synthesise a representative fragment that we anticipated would engage the PI3Kδ
selectivity region of the protein21, however compound 5 was inactive in our assay when tested at standard
concentrations, highlighting the importance of occupying the lipid kinase affinity pocket to enable measurement
of the activity of suitable fragments. This led us to also synthesise the 2-chloro substituent on the pyridine,
compound 6, due to our previous observation that a 2-chloro substituent displayed enhanced potency over the 2-
methoxypyridine substituent in the primary enzyme assay. This proved to be the case for the fragments herein
such that compound 6 was more potent and efficient than compound 4.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
With this data in hand we planned to construct a compound set (Fig. 2) that enabled exploration of the hinge
binder from either 2-chloro and 2-methoxy pyridines or an indole, with the aim of identifying novel, PI3Kδ-
selective hinge binders at R1.
H
N
O
S
R1
R1
HN
O
R2
N
R2 = Cl, OMe
Figure 2. Outline structures of planned array.
To design and select hinge binders a two-stepped approach was adopted. Firstly, an open-minded, inclusive
approach was taken to identify novel putative hinge binding groups, as our hypothesis was that PI3K isoforms
demonstrated a wider tolerance for hinge binding groups in the ATP binding site compared to many protein
kinases. In addition to accommodating hinge binding groups with only an H-bond acceptor, a tolerance for
saturated and therefore more bulky hinge binding groups (morpholine, for example) was reported22. This was
believed to be a critical advantage for maximising protein kinase selectivity of lipid kinase-targeted compounds.
Secondly, a modelling approach was applied to help evaluate and prioritise for synthesis the list of proposed hinge
binders, and in order to develop a fuller understanding of the potential interaction features of these moieties, the
Cresset field based representation was applied23. A 3D molecular interaction potential (MIP) built from an
extended atom forcefield was used to generate a more detailed model of the electrostatics that better described the
projected interactions of the template than traditional point charge models. In this case a crystal structure pose of
compound 3 in PI3Kδ was used as a template from a representative crystal structure and the designed hinge binder
ACS Paragon Plus Environment

Page 5 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
6
fragments attached to the indole group were aligned to this template pose guided by the MIPs in the Cresset
software.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Cresset field point features of an indole affinity pocket group with either an isoquinoline hinge binder
(left) or a pyridine hinge binder (right) overlaid with template compound 3. Field points for the template and query
molecules are represented as cubes and spheres, respectively. Field point colors denote donor (red), acceptor
(blue), hydrophobic minima (gold) or shape (yellow) features and size indicates the magnitude of the feature.
As an example, an isoquinoline hinge binder (fig. 3, left) fitted the features of the indazole template very well,
with the field point features of the quinoline (spheres) closely matching the features of the template (cubes).
Moreover, it was apparent that the H-bond acceptor feature of the hinge binder fitted nicely against the Val828
backbone H-bond donor of the kinase hinge residue, although this was not used to drive the overlay. By contrast,
the overlay of the pyridyl hinge binder (fig. 3b) with the indazole template was very poor and the H-bond acceptor
feature of the pyridine nitrogen was not accessible to the kinase hinge region without disrupting the overlay of the
indole moiety. The same approach was taken for the pyridine sulfonamide and a similar outcome was observed.
To test the applicability of using the Cresset approach to rank compounds for synthesis, an initial set of compounds
was selected and synthesised, including hinge binders predicted to either fit well or fit poorly.
Data from this initial array demonstrated a large disconnect between the indole and the pyridine sulfonamide
groups such that the pyridine sulfonamide group was substantially more accommodating to different hinge binders
when compared to the indole, as exemplified by the compounds in Table 2. Compound 7 was approximately 400-
fold more potent at PI3Kδ than the indole equivalent compound 8 and substantially more ligand efficient (cf. a
25-fold change between compounds 9 and 10). Since compounds 7 and 9 were equipotent, this further
demonstrated the efficiency of the shorter hinge binder. In fact, as the PI3K enzyme assay floor was a pIC50 of
4.3 – 4.6 when operating at standard compound concentrations, the potency of certain indole compounds, for
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 30
1
2
3
4
5
6
7
8
9
example compound 8, had to be determined using higher compound concentrations. From these data, we
hypothesised that the pyridine sulfonamide was a privileged fragment specifically for PI3Kδ, as the potency
against PI3Kα, β and γ was consistently lower for compounds 4, 6, 7 and 9. Furthermore compound 7 was
essentially inactive (pIC50 <5) against a cross panel of 16 kinases screened (see supporting information).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound Number
pIC50 PI3Kδ, α, β and γ
PI3Kδ LE
Hinge Binder
N
N
7
9
H
N
O
S
6.6, 5.2, 6.1, 5.4
6.8, 5.6, 5.6, 6.1
O
Cl
N
0.5
0.42
8
4.1ab, <4.3, <4.3, <4.3
0.37
10
5.4, <4.5, 4.8, 4.8c
0.39
HN
Table 2. Selected compounds from the initial hinge binder array. n ≥ 3. a: Tested at high concentration to determine
a pIC50. b: n=2. c: Tested pIC50 < 4.6 on one occasion.
These data validated the Cresset model for the indole group, demonstrating that the isoquinoline was indeed a
more effective hinge binder than the pyridine but clearly demonstrated that the approach could not be applied to
prioritise hinge binders for the pyridine sulfonamide. Furthermore, the data for the pyridine sulfonamide built
confidence in the hypothesis that we had discovered a privileged fragment for PI3Kδ inhibition that not only gave
very potent and ligand efficient compounds, but also demonstrated reasonable selectivity over the other PI3K
isoforms. To explore the potency and selectivity of the template in a more robust manner, we expanded the planned
compound set to include variations to the ortho-pyridine and sulfonamide substituents. To enable synthesis of a
manageable compound set and consistency when analysing the data generated, we chose to fix specific R-groups,
when exploring the other R groups (fig. 4). The R1 group was fixed as dihydropyran or an N-methylpyrazole, with
R2 as either chloro or methoxy to explore the R3 sulfonamide substituents in more detail (fig. 4). Similarly, the R3
sulfonamide was fixed as methyl or phenyl when exploring R1 or R2 further. The majority of compounds were
synthesised from readily available sources via array chemistry, whilst the remainder were synthesised bespoke as
necessary.
ACS Paragon Plus Environment

Page 7 of 30
Journal of Medicinal Chemistry
1
2
3
4
H
N
O
S
R3
R1
5
6
O
R2
N
7
8
9
N
N
N
N
O
O
Or
Or
R2 =
R3 =
Cl
OCH3
Ph
R1 =
R3 =
Or
Or
R1 =
R2 =
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Me
Ph
OCH3
Or
Me
Or
R3 = varying
Cl
R1 = varying
R2 = varying
Figure 4. Compound set design to evaluate R1, R2 and R3.
Once all the data was in hand and after determining that the data set had reasonable cross coverage (see supporting
information, Fig. 1S), an R-group decomposition on the total 324 compounds synthesised was carried out (using
the R-groups as defined in fig. 4), followed by a Free-Wilson analysis24 applied to the potency data on each of the
PI3K isoforms. The Free-Wilson prediction demonstrated a good correlation for all PI3K isoforms (see supporting
information, Fig. 2S, 3S, 4S and 5S). This enabled evaluation of the contribution to PI3K isoform potency of each
of the R-groups independently and by considering the different impact across all four PI3K isoforms, allowed a
detailed assessment of the selectivity conferred by each R-group. The Free-Wilson analysis also provided a
potency contribution from the core. For this data set, compounds were generally more potent at PI3Kδ than the
other isoforms, and this was reflected in the contribution of the core to the potency for the four isoforms (pIC50
6.0, 5.2, 5.3 and 5.1 for PI3Kδ, α, β and γ, respectively), further supporting our hypothesis that the pyridine
sulfonamide was a privileged fragment for PI3Kδ.
In figures 5, 6 and 7, selected R-groups (for which the structure is indicated by the index-linked tables 3, 4 and 5)
are displayed on the x-axis and the average potency contributions to each isoform across the data set, as determined
by the Free-Wilson analysis, are displayed on the y-axis, indicated by colored diamonds (PI3Kδ in green, PI3Kα
in orange, PI3Kβ in red and PI3Kγ in yellow). All compounds were evaluated in the production PI3K assay format,
so the assay floor was pIC50 ~4.6 for all PI3K isoforms, and this was important to keep in mind when analysing
selectivity differences between compounds of low potency. For instance, the core potency contribution was higher
for PI3Kδ, therefore R-groups for PI3Kδ had a larger potential magnitude of negative potency contribution from
the Free-Wilson analysis than the same R-groups for other PI3K isoforms.
A wide range of potency contributions were observed for the different hinge binders (R1) across the different PI3K
isoforms, despite the very high protein homology of the amino acid sequence in this region of the PI3K family
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 30
1
2
3
4
5
6
7
8
9
(fig. 5, table 3). Dihydropyran (R1A) was moderately beneficial for all PI3K isoforms with some selectivity
preference for PI3Kδ. Tetrahydropyran (R1B) lowered potency against all PI3K isoforms, demonstrating that there
is a requirement for planarity in the immediate vicinity of the pyridine sulfonamide. N-linked morpholine (R1C)
demonstrated a small potency increase for PI3Kδ and β, but to a lower extent than dihydropyran (R1A), with some
selectivity over PI3Kα and γ. Homo-morpholine (R1D) was tolerated by all isoforms except PI3Kγ. Racemic 2-
methylmorpholine (R1E) increased potency and selectivity in a comparable manner to dihydropyran (R1A) and
was substantially better than racemic 3-methylmorpholine (R1F). 2-Oxa-6-azaspiro[3.3]heptane (R1G) and
racemic 2,6-dimethylmorpholine (R1H) were essentially inactive at all PI3K isoforms, further demonstrating the
requirement for some planarity in the region adjacent to the pyridine ring and the lack of space for di-substitution
on the morpholine. 2-Ethylmorpholine (R1I) was tolerated for all isoforms but preferentially increased potency at
PI3Kα and γ. Most of the more differentiated potential hinge binders, such as, thiomorpholine 1,1-dioxide (R1J)
and cyclohex-2-enone (R1K) were also essentially inactive against all PI3K isoforms. N-methylpyrazole (R1L)
and pyrazole (R1M) were both tolerated reasonably well by all isoforms when linked in the 4-position and N-
methylimidazole (R1N), linked in the 5-position, was tolerated by all PI3K isoforms except PI3Kγ. Pyrazole
(R1O), linked in the 3-position, was less potent against all PI3K isoforms. Surprisingly, incorporation of a phenyl
“hinge binder” (R1P) was tolerated with only a small reduction in potency at all PI3K isoforms. This observation
highlights the exceptional tolerance that the PI3K isoforms can have to different hinge binders if a preferred
affinity pocket group is appropriately situated. Pyridine (R1Q), substituted in the 4-position gave a small boost to
PI3Kδ and β over PI3Kα and γ, whereas 3-substituted pyridine (R1R) was tolerated equally by all isoforms with
only a small potency reduction. 4-Substituted 2-methylpyridine (R1S) was tolerated for all isoforms, with a slight
bias for PI3Kα. Thiazole (R1T) boosted PI3Kδ potency more than the other isoforms and isothiazole (R1U) showed
the same profile, albeit at a lower overall potency increase. 5-Substituted pyrimidine (R1V) increased PI3Kδ a
small amount whilst not affecting the other isoforms. 4-Substitued 2-aminopyrimidine (R1W) showed a large
preference for PI3Kγ over the other isoforms but one of the more dramatic discoveries was the 5-substituted 2-
aminopyrimidine (R1X) which, while demonstrating increased potency at all isoforms, had a very strong
preference for PI3Kγ. Removing a nitrogen from R1X to give 5-substituted 2-aminopyridine (R1Y) was tolerated
but much less beneficial for all the PI3K isoforms. Aryl ethers were also tolerated as hinge binders, for example
3-methoxyphenyl (R1Z) was tolerated by all PI3K isoforms and 4-methoxyphenyl (R1ZA) boosted PI3Kα and γ
much more than PI3Kδ and β. Aryl carboxamides and sulfonamides (such as R1ZB and R1ZC) were not well
tolerated by any PI3K isoform. Other hinge binders that were tolerated include 3-fluoro-4-chlorophenyl (R1ZD)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 9 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and 3-fluoro-4-hydroxyphenyl (R1ZE) although both had a slight preference for PI3Kγ. Following on from this
thorough exploration of R1 we identified the dihydropyran (R1A) and morpholine (R1C) groups as giving the best
balance of potency increase and selectivity for PI3Kδ over the PI3K isoforms, whilst keeping the aromatic ring
count low.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. A representative set of R1 group contribution scores to PI3Kδ, α, β and γ potency (green, orange, red
and yellow, respectively), normalised to R1L.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 30
1
2
3
4
5
6
7
8
9
A
G
B
H
C
I
D
J
E
F
L
O
O
N
O
O
O
N
O
N
*
N
*
K
O O
S
O
O
N
O
N
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
N N
N
N
M
S
N
T
O
U
P
Q
R
X
N
N
N
NH
N
N
N
*
HN N
N
S
V
W
NH2
N
S N
N
O
N
H2N
N
*
N
N
Y
Z
ZA
ZB
ZC
ZD
NH2
N
NH2
S
Cl
O
O
*
*
F
H2N
O
O
ZE
OH
F
Table 3: Structures of the representative set of R1 groups. R1 groups highlighted by * are based on one compound.
All other data was derived from two or more compounds.
Focussing on the R2 substituent (fig. 6, Table 4), the methoxy (R2A) and chloro (R2B) substituents confirmed our
original observations that the chloro was generally higher potency than methoxy across all PI3K isoforms, with
little preference for PI3Kδ. A trifluoromethyl substituent (R2C) boosted potency further at all PI3K isoforms other
than PI3Kγ, indicating the preference for an electron withdrawing group at this position. However, the cyano
substituent (R2D) only showed similar potency to the methoxy group at PI3Kδ and β and was less potent at PI3Kα
and γ indicating that some lipophilicity was required or that the H-bond acceptor of the cyano group is less
tolerated in those isoforms. The difluoromethoxy substituent (R2E) had similar potency to the cyano for PI3Kδ
and β but was less potent at PI3Kα and γ. This substituent, along with the cyano, therefore would provide a good
option to enhance selectivity of PI3Kδ over PI3Kα and γ if that were required during later lead optimisation
studies. Extending the methoxy group to ethoxy (R2F) reduced all PI3K isoform activity marginally and the
ACS Paragon Plus Environment

Page 11 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
trifluoroethoxy substituent (R2G) was significantly worse. Replacement of the substituent with hydrogen (R2H)
had a significant impact on PI3Kδ potency and lowered potency at PI3Kα, β and γ. Changing to a methyl
substituent (R2I) was also detrimental to potency for all PI3K isoforms albeit more tolerated in PI3Kδ. An
isopropyl substituent (R2J) reduced potency at all PI3K isoforms and the methylamino substituent (R2K) was
similarly detrimental. Finally, the hydroxyl substituent (R2L) was also lower potency for all PI3K isoforms,
potentially due to the propensity for hydroxy pyridines to prefer the pyridone tautomer25 which is not tolerated in
this region of the PI3K isoforms. After this exploration of R2 groups, the methoxy (R2A) and chlorine (R2B)
remained our preferred R2 groups for their PI3Kδ. We considered the trifluoromethyl (R1C) group as interesting
SAR but too lipophilic for further consideration.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Representative set of R2 group contribution scores to PI3Kδ, α, β and γ potency (green, orange, red and
yellow, respectively), normalised to R2A.
A
G
B
H
C
I
D
J
E
F
L
CF2H
O
Cl
H
CF3
CH3
C
N
O
O
K
NH2
CF3
OH
O
Table 4. Structures of the representative set of R2 groups. All R2 group data was derived from two or more
compounds.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 30
1
2
3
4
5
6
7
8
9
Finally, consideration of the sulfonamide substituent, R3, revealed some unexpected SAR (fig. 7, Table 5). Firstly,
alkyl groups, such as methyl (R3A), were generally less potent than phenyl (R3B) for all PI3K isoforms. Similarly,
a cyclohexyl group (R3C) was not well tolerated and potency was lower for all PI3K isoforms. 4-Cyanophenyl
(R3D) increased the potency for PI3Kα but not for the other isoforms, whereas 4-hydroxyphenyl (R3E) was more
potent against all isoforms. 4-Methoxyphenyl (R3F) was not substantially different to phenyl (R3B). The 3-
cyanophenyl group (R3G) was more potent for all PI3K isoforms and especially so for PI3Kδ and β. The 3-
methoxyphenyl (R3H), 2-methylphenyl (R3I) and 3-pyridyl (R3J) groups all increased potency for all PI3K
isoforms to a similar level above the phenyl (R3B). Different heterocycles were generally well tolerated, for
example an imidazole group (R3K) boosted PI3Kδ and β without increasing PI3Kα and γ. A thiazole group (R3L)
increased potency at all isoforms, with a preference for PI3Kβ, whereas a phenylthiazole group (R3M) increased
potency at all isoforms, with a dramatic preference for PI3Kδ, especially over PI3Kβ. A benzylimidazole group
(R3N) also selectively increased PI3Kδ potency whilst having much less impact on the other isoforms. A benzyl
group (R3O) showed similar potency to the phenyl (R3B). Finally, initial attempts to reach the selectivity pocket
using groups with a dimethyl morpholine (R3P or R3Q) were not successful and lowered potency against all PI3K
isoforms. After this exploration, we were left with a selection of interesting R3 groups and a clear indication that
this was a suitable position on the template to explore further, with scope to modulate potency, selectivity and
physicochemical properties.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 13 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
Figure 7. Representative set of R3 group contribution scores to PI3Kδ, α, β and γ potency (green, orange, red and
yellow, respectively), normalised to R3B.
6
7
8
9
A
G
B
H
C
I
D
J
E
F
L
CH3
OH
N
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
K
S
N
N
N
N
O
N
M
N
O
P
Q
N
N
S
N
O
O
N
N
*
Table 5. Structures of the representative set of R3 groups. R3 groups highlighted by * are based on one compound.
All others R3 groups were derived from two or more compounds.
The following compounds from the synthesised compound set (Table 6) were selected to highlight the trends
observed. Compound 11 was amongst the most efficient compounds with an LE for PI3Kδ of 0.53 and reasonable
PI3K isoform selectivity. For compound 12 the PI3Kγ activity increased more than 10-fold compared to
compound 11, whilst the other PI3K isoforms all increased less than 3-fold, demonstrating that even with a hinge
binder that demonstrated a clear selectivity preference for PI3Kγ, the pyridine sulfonamide resulted in a compound
that was equipotent at PI3Kδ and PI3Kγ. Compound 13 demonstrated the increase in PI3K enzyme potency
obtained when changing a methyl group to a phenyl across all PI3K isoforms, although the LE was slightly lower
than for compound 11. Compound 14 contained a phenyl hinge binder and, although the potency for PI3Kδ was
100-fold lower than compound 13, clearly demonstrated the broad tolerance for short, flat hinge binders attached
to the pyridine sulfonamide template. The difference between the methoxy and chlorine R2 group was exemplified
by compound 15 which is a less potent and efficient PI3Kδ inhibitor than compound 13, but maintains equivalent
PI3K isoform selectivity. Compound 16 was an example of the most PI3K isoform selective compound made in
this compound set, with ≥ 100-fold selectivity for PI3Kδ over the PI3K isoforms. This selectivity came at the
price of ligand efficiency (LE = 0.36), inferring that, although this R3 group may not offer much additional potency
for PI3Kδ over other R3 groups, it reduced relative potency at the other PI3K isoforms. Compound 16 was also ≥
100-fold selective against a cross-panel of 29 kinases (see supporting information) and inactive in a hERG binding
assay (pIC50 < 4.3). Compound 16 had a pIC50 of 7.3 in a peripheral blood mononuclear cell assay, using Cytostim
to stimulate interferon γ production from the T-lymphocyte compartment15. Passive permeability on MDCK cells
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 30
1
2
3
4
for compound 16 was high (466 nm/sec) and CLND solubility was 12 μg/mL. Protein binding (human serum
5
6
albumin) was 89.9% and microsomal clearance was high in rat and human, 17.4 and 7.9 mL/min/g, respectively.
7
8
pIC50
9
Compound
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Structure
PI3Kδ (LE) PI3Kα PI3Kβ PI3Kγ
Number
O
H
N
O
S
11
6.9 (0.53)
6.9 (0.5)
5.6
5.8
6.3
6.8
5.5
6.7
O
Cl
N
N
NH2
H
N
O
S
N
12
O
Cl
N
O
H
N
O
S
13
14
8.0 (0.48)
6.0 (0.36)
6.3
4.5
6.9
4.8
6.3
O
Cl
N
H
N
O
S
4.9a
O
Cl
N
O
H
N
O
S
15
16
7.4 (0.42)
5.7
6.2
6.4
6.0
5.9
6.0
O
O
N
N
O
H
O
8.2 (0.36)
N
N
S
O
O
N
Table 6: Specific compounds from the synthesised compound set. n ≥ 3. a: Tested pIC50 < 4.6 on two occasions.
To try and understand the origins of the selectivity and efficiency of the series, a crystal structure of compound
11 was obtained in PI3Kδ (as previously described26) (fig. 8). The dihydropyran oxygen made the anticipated
ACS Paragon Plus Environment

Page 15 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
acceptor interaction with Val828 in the hinge-region of the kinase and maintained planarity in the region of the
pyridine ring, providing further evidence for the low potency of those hinge binders with more 3-dimensional
structure in the region (e.g. tetrahydropyran, R1B). The pyridine nitrogen made an H-bond interaction to stabilise
a crystallographically observed water, bridging to Asp787 and Tyr813, and the sulfonamide group made a
bridging interaction between Lys779 and Trp760. This gave some insight into the preference for electron
withdrawing substituents at R2, which would increase the ionisation of the sulfonamide and thereby enhance the
strength of the interaction with Typ760 and Lys779. This also implied that the pyridine nitrogen H-bond to water
was not the key driver of potency. The crystal structure also indicated why bulkier R2 groups were less potent due
to the lack of space in the protein to accommodate them.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Compound 11 (cyan) crystal structure with PI3Kδ.
Comparison of the PI3Kδ crystal structure of compound 215 with compound 11 (fig. 9) showed distinct changes
in the PI3Kδ binding site. Firstly, the pyridine sulfonamide moiety of compound 11 was substantially closer to
the hinge region due to the shorter dihydropyran hinge binder (this compound movement led us to adopt the name
“Short hinge” for this series) and Lys779 moved to maintain the beneficial interaction with the sulfonamide group.
Secondly, Trp760, with which compounds 1 and 2 interacted to gain significant PI3Kδ selectivity, rotated through
180º in order to make the H-bond with an oxygen of the sulfonamide of compound 11. We believed this additional
interaction with Trp760 contributed to the enhanced potency and ligand efficiency observed for this short hinge
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 30
1
2
3
4
5
6
7
8
9
series when compared to the indazole series. The indole affinity pocket group of compound 8, for example, was
unable to make the same interactions and was therefore less potent and ligand efficient. The altered location of
Trp760 may also explain why our initial attempts to obtain similar interactions to those observed with compounds
1 and 2 with groups such as R3P and R3Q were not successful. The water-mediated H-bond interaction from the
nitrogen of the pyridine group to Asp787 and Tyr813 is observed for both compounds 2 and 11, despite the
significant ligand movement. Previously reported PI3Kγ crystal structures of non-specific inhibitors of lipid
kinases containing a pyridine sulfonamide, such as those reported by D’Angelo et al.27, Nishimura et al.28, and
Knight et al.29, demonstrated a similar binding mode to compound 2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. Compound 11 (cyan) PI3Kδ crystal structure overlaid with compound 2 (purple).
We also obtained a crystal structure of compound 14 in PI3Kδ which demonstrated an almost identical binding
mode to compound 11 (fig. 10) with only a small movement away from the hinge region observed, presumably
due to the lack of a strong H-bond between Val828 and the phenyl group ‘hinge binder’. Compound 14 maintained
the key H-bonds with Trp760 and Lys779, albeit with small positional differences and also maintained the
through-water interactions with Tyr813 and Asp787. This non-differentiated binding mode once again
demonstrated the broad tolerance of the privileged pyridine sulfonamide when attached to diverse hinge binders.
ACS Paragon Plus Environment

Page 17 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Compound 14 (maroon) PI3Kδ crystal structure overlaid with compound 11 (cyan).
Finally, a crystal structure of compound 16 in PI3Kδ was obtained and overlaid with compound 11 (fig. 11). This
demonstrated that compound 16 made the same interactions with Val828, Trp760 and Lys779 as compound 11,
albeit with a slight rotation of the pyridine group, potentially to accommodate the longer R2 methoxy group of
compound 16 over the chlorine in compound 11. The benzyl imidazole substituent of the R3 group occupied the
ribose pocket, but was not close enough to make the same selectivity-enhancing interaction with Trp760 as
compounds 1 and 2. Furthermore there were no isoform residue differences between the PI3K isoforms in the
immediate vicinity of this R3 group that could directly rationalise the observed selectivity. Other R3 groups that
were able to occupy this space in PI3Kδ (for example R3M) displayed enhanced selectivity over shorter
substituents, indicating it as an interesting area to explore during subsequent lead optimization studies.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 11. Compound 16 (gold) PI3Kδ crystal structure overlaid with compound 11 (cyan).
Conclusions
Through a deconstruction and regrowth of potent inhaled clinical compounds 1 and 2, we have discovered and
exploited a pyridine sulfonamide privileged fragment for PI3Kδ that tolerated attachment to a broad range of
hinge binders, including those that do not make a formal H-bond interaction with the hinge region of the PI3K
isoforms, due to a novel binding mode observed in the PI3Kδ crystal structure. Furthermore, we determined that
different hinge binders attached to the pyridine sulfonamide can have a substantial effect on PI3K isoform potency
and selectivity, even in a region of the protein known to be homologous across all PI3K isoforms. This moiety, in
combination with certain hinge binders that showed a relative preference for different PI3K isoform, resulted in
compounds equipotent at PI3Kδ, further demonstrating that the pyridine sulfonamide is a privileged fragment for
PI3Kδ; an observation that fitted well with known literature compounds containing a pyridine sulfonamide27-29
that were also equipotent across PI3K isoforms. Finally we reported compound 16, a potent and selective inhibitor
of PI3Kδ, and an exemplar of a series suitable for further lead optimisation studies.
ACS Paragon Plus Environment

Page 19 of 30
Journal of Medicinal Chemistry
1
2
3
4
Experimental Section
5
6
7
8
Chemistry. General. All solvents and reagents, unless otherwise stated, were commercially available and were
used as purchased without further purification. Proton and carbon nuclear magnetic resonance (1H NMR and 13C
NMR) spectra were recorded on a Bruker AVI (400 MHz), Bruker Nano (400 MHz), Bruker DPX300 (300 MHz)
or Bruker AVII+ (600 MHz) spectrometer (with cryoprobe) in the indicated solvent. Chemical shifts δ are reported
in parts per million (ppm) relative to tetramethylsilane and are internally referenced to the residual solvent peak.
HRMS data were recorded on a Waters XEVO G2-XS QTof. Coupling constants (J) are given in hertz (Hz) to the
nearest 0.1 Hz. LCMS and HPLC methods are detailed in the Supporting Information. The purity of all compounds
screened in the biological assays was examined by LCMS analysis and was found to be ≥95% unless otherwise
specified. Mass directed automated preparative HPLC (MDAP) methods are detailed in the Supporting
Information.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compounds were generally synthesised as described in Scheme 1. Reacting amine A with R3-sulfonyl chlorides
gave aryl bromides B, which could be subjected to Suzuki-Miyaura cross-coupling with a boronic acid or ester,
or alternatively subjected to Buchwald-Hartwig cross-coupling with an amine, to give target compounds C.
These steps could be reversed, giving intermediates amines D, to vary the sulfonamide group.
H
O
N
Br
S
R3
O
R2
N
b or c
a
H2N
R2
Br
H
O
B
N
R1
S
R3
b
a
O
N
A
R1
R2
N
C
H2N
R2
N
D
Scheme 1. Synthesis of array compounds. (a) R3-SO2Cl, in pyridine or THF with DIPEA; (b) R1-boronic acid or
ester, Pd catalyst (Pd(dppf)Cl2, Pd(PPh3)4, or chloro(di-2-norbornylphosphino)(2'-dimethylamino-1,1'-biphenyl-
2-yl)palladium(II)), base (Na2CO3, K2CO3 or K3PO4), solvent (1:4-dioxane or THF, in water); (c) amine,
Pd2(dba)3, Cs2CO3, RuPhos, DMF:1,4-dioxane.
Exemplified compounds were synthesised as follows:
6-(1H-Indol-4-yl)-1H-indazole, 3. 6-Bromo-1H-indazole (100 mg, 0.51 mmol), sodium carbonate (161 mg, 1.52
mmol) and PdCl2dppf (37.1 mg, 0.051 mmol) were added to a microwave vial. 1H-Indol-4-ylboronic acid (90 mg,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 30
1
2
3
4
5
6
7
8
9
0.558 mmol) was dissolved in 1,4-dioxane (8 mL) and added to the vial followed by water (2 mL). The reaction
was heated to 80 °C for a total of 3 h. The residue was separated between DCM (20 mL) and brine (20 mL), the
organic layer was evaporated to dryness and the mixture was purified using MDAP (method A). Pure fractions
were combined and evaporated under a stream of nitrogen to give the title compound, 29.5 mg (25% yield). LCMS
(method A): 100%, rt = 0.91, [M+H]+ = 234. 1H NMR (400 MHz, DMSO-d6) δH 13.08 (br s, 1H), 11.29 (br, s,
1H), 8.33 (HCOOH, 10% by weight), 8.11 (s, 1H), 7.86 (d, J=8.6 Hz, 1H), 7.75 (m, 1H), 7.44 (m, 3H), 7.21 (t,
J=7.8 Hz, 1H), 7.16 (dd, J=7.3, 0.8 Hz, 1H), 6.59 (br, s, 1H). HRMS (ESI) for C15H12N3 [M+H]+ calcd.: 234.1032,
found: 234.1039.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N-(5-(1H-Indazol-6-yl)-2-methoxypyridin-3-yl)methanesulfonamide, 4. See compound 4 in ref 20
.
4-[5-(4-Morpholinylmethyl)-1,3,4-oxadiazol-2-yl]-1H-indazole, 5. 6-Bromo-4-[5-(4-morpholinylmethyl)-
1,3,4-oxadiazol-2-yl]-1-(phenylsulfonyl)-1H-indazole (compound S4, 500 mg, 0.991 mmol) was dissolved in
ethyl acetate (60 mL) and added to Palladium on carbon (158 mg, 0.149 mmol) and left under hydrogen for 2
days. The reaction mixture was filtered through Celite and the solvent removed in vacuo. The residue was
dissolved in isopropanol (10 mL) and 2M sodium hydroxide (aq.) (0.992 mL) and stirred overnight. The reaction
mixture was diluted with water (10 mL) and extracted with DCM (2 x 10 mL). The organic layer was passed
through a hydrophobic frit and concentrated in vacuo. The residue was purified by silica chromatography (50 g
column) eluting with 0 - 100% ethyl acetate in cyclohexane over 40 min, followed by 100% methanol. The
methanol fractions were purified by silica chromatography (50 g) eluting with 0 - 15% methanol (containing 1%
triethylamine) in DCM over 40 min. Product containing fractions were purified by silica chromatography (50 g)
eluting with 0 - 25% methanol in DCM over 60 min. Pure fractions were combined and dried to give title
compound, 80 mg (28% yield). ). LCMS (method A): 93%, rt = 0.52, [M+H]+ = 286. 1H NMR (400 MHz, DMSO-
d6) δH 13.55 (br. s., 1H), 8.54 (s, 1H), 7.77-7.87 (m, 2H), 7.54 (t, J=7.9 Hz, 1H), 3.96 (s, 2H), 3.61 (t, J=4.8 Hz,
4H), 2.54 (t, J=4.8 Hz, 4H). HRMS (ESI) for C14H16N5O2 [M+H]+ calcd.: 286.1305, found: 286.1308.
N-[2-Chloro-5-(1H-indazol-6-yl)-3-pyridinyl]methanesulfonamide,
6.
6-(4,4,5,5-Tetramethyl-1,3,2-
dioxaborolan-2-yl)-1H-indazole (100 mg, 0.41 mmol), N-(5-bromo-2-chloropyridin-3-yl)methanesulfonamide15
(117 mg, 0.410 mmol), and tetrakis(triphenylphosphine) palladium (0) (23.7 mg, 0.02 mmol) were weighed to a
microwave vial. The vial was sealed and THF (3 mL) was added followed by potassium carbonate (56.6 mg, 0.41
mmol) in water (2 mL). The reaction was heated at 150 ºC for 2 h. The reaction was diluted with DCM (30 mL)
and washed with brine (30 mL) and water (2 x 30 mL). The organic layer was passed through a hydrophobic frit
ACS Paragon Plus Environment

Page 21 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and concentrated under vacuum. The aqueous layer was concentrated under vacuum. Both residues were purified
by MDAP (method A). Product fractions were combined and evaporated to dryness to give title compound, 25
mg (19% yield). LCMS (method A): 93%, rt = 0.73, [M+H]+ = 323/325. 1H NMR (400 MHz, DMSO-d6) δH 13.28
(br. s., 1H), 10.00 (br. s. 1H), 8.62 (d, J=2.5 Hz, 1H), 8.11-8.18 (m, 2H), 7.91 (d, J=8.6 Hz, 1H), 7.84 (s, 1H), 7.46
(dd, J=1.4, 8.4 Hz, 1H), 3.17 (s, 3H). HRMS (ESI) for C13H12ClN4O2S [M+H]+ calcd.: 323.0370, found: 323.0371.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N-(6-Chloro-3,4'-bipyridin-5-yl)methanesulfonamide, 7. A mixture of N-(5-bromo-2-chloropyridin-3-
yl)methanesulfonamide (286 mg, 1 mmol), pyridin-4-ylboronic acid (246 mg, 2.000 mmol), Na2CO3 (318 mg,
3.00 mmol) and PdCl2dppf.DCM (82 mg, 0.1 mmol) in 1,4-dioxane and water (10 mL, 4:1, v/v) under nitrogen
was heated at 70 ºC for 4 h, then filtered. The filtrate was concentrated and the residue was purified by silica
chromatography (methanol: DCM, 1:50) to obtain a crude product. The crude was washed with ethyl acetate (5
mL), filtered and dried in vacuo to give the title compound 56 mg (20% yield). HPLC (method A): 99.5%, rt =
4.23. 1H NMR (300 MHz, CDCl3) δH 8.78 (m, 2H), 8.50 (m, 1H), 8.22 (s, 1H), 7.52 (m, 2H), 6.96 (s, 1H), 3.09
(s, 3H). HRMS (ESI) for C11H11ClN3O2S [M+H]+ calcd.: 284.0261, found: 284.0264.
4-(Pyridin-4-yl)-1H-indole, 8. A mixture of 4-bromo-1H-indole (70 mg, 0.36 mmol), 4-(4,4,5,5-tetramethyl-
1,3,2-dioxaborolan-2-yl)pyridine (109 mg, 0.56 mmol), catalytic tetrakis(triphenylphosphine) palladium (0), sat.
sodium bicarbonate (aq.) (1 mL) and DMF (2 mL) were combined in a microwave vial and heated at 160 ºC for
30 min. The reaction was concentrated and separated between DCM and sat. sodium bicarbonate (aq.). The
material was purified by MDAP (method B), then freeze-dried from 1,4-dioxane and water to give the title
compound. LCMS (method B): 100%, rt = 2.08, [M+H]+ = 195. 1H NMR (600 MHz, DMSO-d6) δH 11.64 (br. s.,
1H), 8.87 (d, J=4.9 Hz, 2H), 8.16 (d, J=6.0 Hz, 2H), 7.64 (d, J=7.9 Hz, 1H), 7.59 (t, J=2.6 Hz, 1H), 7.40 (d, J=7.2
Hz, 1H), 7.25-7.33 (t, J=7.7 Hz, 1H), 6.73 (br. s., 1H). HRMS (ESI) for C13H11N2 [M+H]+ calcd.: 195.0923, found:
195.0930.
N-[2-Chloro-5-(7-isoquinolinyl)-3-pyridinyl]methanesulfonamide, 9. 7-Isoquinolinylboronic acid (100 mg,
0.578 mmol), N-(5-bromo-2-chloropyridin-3-yl)methanesulfonamide (165 mg, 0.578 mmol), Tripotassium
phosphate (368 mg, 1.734 mmol) and PdCl2dppf (32.4 mg, 0.058 mmol) were weighed to a microwave vial. The
vial was sealed and 1,4-dioxane (4 mL) and water (1 mL) were added. The vial was then flushed with nitrogen
and heated at 80 ºC for 3 h. The residue was diluted with DCM (30 mL) and washed with brine (30 mL), then
water (2 x 30 mL). The organic layer was then passed through a hydrophobic frit and evaporated to dryness. The
residue was purified by MDAP (method A). Pure fractions were evaporated and combined to give the title
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 30
1
2
3
4
5
6
7
compound 35.9 mg (19% yield). LCMS (method A): 100%, rt = 0.52, [M+H]+ = 334/336. 1H NMR (400 MHz,
DMSO-d6) δH 9.42 (s, 1H), 8.73 (d, J=2.5 Hz, 1H), 8.51-8.59 (m, 2H), 8.25 (d, J=2.3 Hz, 1H), 8.09-8.19 (m, 3H),
7.89 (d, J=5.8 Hz, 1H), 3.17 (s, 3H). HRMS (ESI) for C15H13ClN3O2S [M+H]+ calcd.: 334.0418, found: 334.0418.
8
9
7-(1H-Indol-4-yl)isoquinoline, 10. In a microwave vial, 7-bromoisoquinoline (95 mg, 0.46 mmol), tripotassium
phosphate (242 mg, 1.14 mmol) and indole-4-boronic acid (74 mg, 0.46 mmol) were partially dissolved into a
two-phase mixture of CPME (8 mL) and water (2 mL). This was purged by bubbling a stream of nitrogen through
it for 5min. PdCl2(dppf) (34mg, 0.046mmol) was added and the reaction was heated at 140 ºC for 30 min in a
microwave. The reaction was diluted with DCM (60 mL) and water (20 mL) then the phases separated using a
hydrophobic frit. The aqueous phase was extracted with further DCM (20 mL) then the combined organic phases
were concentrated in vacuo. The residue was purified by MDAP (method C). Product fractions were combined
and concentrated in vacuo to give the title compound, 56.6 mg (51% yield). LCMS (method C): 100%, rt = 1.04,
[M+H]+ = 245. 1H NMR (400 MHz, DMSO-d6) δH 11.37 (br. s., 1H), 9.42 (s, 1H), 8.53 (d, J=5.8 Hz, 1H), 8.40
(s, 1H), 8.07-8.16 (m, 2H), 7.89 (d, J=5.8 Hz, 1H), 7.46-7.52 (m, 2H), 7.22-7.29 (m, 2H), 6.68 (br. s., 1H). HRMS
(ESI) for C17H13N2 [M+H]+ calcd.: 245.1080, found: 245.1083.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N-(2-Chloro-5-(3,6-dihydro-2H-pyran-4-yl)pyridin-3-yl)methanesulfonamide, 11. A mixture of N-(2-chloro-
5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin-3-yl)methanesulfonamide (797 mg, 2.4 mmol), 3,6-
dihydro-2H-pyran-4-yl trifluoromethanesulfonate (464 mg, 2 mmol), PdCl2dppf.DCM (163 mg, 0.2 mmol),
sodium carbonate (530 mg, 5 mmol) in 1,4-dioxane (16 mL) and water (4 mL) was heated at 80 ºC for 2 h. The
mixture was filtered and the filtrate was concentrated. The residue was purified by silica chromatography, eluting
with ethyl acetate: petroleum ether (1:5, v/v). The material was washed with ethyl acetate (5 mL), filtered, washed
with further ethyl acetate, then dried in vacuo to give the title compound 245 mg (45% yield). HPLC (method B):
1
99.6%, rt = 6.25. H NMR (300 MHz, CDCl3) δH 8.26 (d, J=2.4 Hz, 1H), 7.98 (d, J=2.4 Hz, 1H), 6.78 (s, 1H),
6.28 (s, 1H), 4.33-4.36 (m, 2H), 3.95 (t, J=5.4 Hz, 2H), 3.07 (s, 3H), 2.50-2.53 (m, 2H). HRMS (ESI) for
C11H14ClN2O3S [M+H]+ calcd.: 289.0414, found: 289.0418.
N-(5-(2-Aminopyrimidin-5-yl)-2-chloropyridin-3-yl)methanesulfonamide, 12. A mixture of N-(2-chloro-5-
(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin-3-yl)methanesulfonamide (40 mg, 0.12 mmol), 5-
bromopyrimidin-2-amine (21 mg, 0.12 mmol), tripotassium phosphate (76 mg, 0.36 mmol), chloro(di-2-
norbornylphosphino)(2'-dimethylamino-1,1'-biphenyl-2-yl)palladium(II) (7 mg, 0.012 mmol) in 1,4-dioxane (0.8
mL) and water (0.2 mL) were heated in a sealed tube in a microwave at 120 ºC for 20 min. The mixture was
ACS Paragon Plus Environment

Page 23 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
loaded onto a C18 SPE cartridge (pre-conditioned with acetonitrile containing 0.1% TFA), and washed through
with acetonitrile containing 0.1% TFA (3 mL). The solvent was removed under a flow of nitrogen. The residue
was purified by MDAP (method D). The solvent was dried under a stream of nitrogen to give the title compound
4.2 mg (9% yield). LCMS (method C): 100%, rt = 0.51, [M+H]+ = 300/302. 1H NMR (600 MHz, DMSO-d6) δH
8.62 (s, 2H), 8.45 (d, J=2.3 Hz, 1H), 7.99 (d, J=2.3 Hz, 1H), 6.98 (s, 2H), 3.12 (s, 3H). HRMS (ESI) for
C10H11ClN5O2S [M+H]+ calcd.: 300.0323, found: 300.0328.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N-(2-Chloro-5-(3,6-dihydro-2H-pyran-4-yl)pyridin-3-yl)benzenesulfonamide, 13. A mixture of N-(2-chloro-
5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin-3-yl)benzenesulfonamide (946 mg, 2.4 mmol), 3,6-
dihydro-2H-pyran-4-yl trifluoromethanesulfonate (464 mg, 2 mmol), PdCl2dppf.DCM (163 mg, 0.2 mmol),
sodium carbonate (530 mg, 5 mmol) in 1,4-dioxane (16 mL) and water (4 mL) was heated at 80 ºC for 2 h under
nitrogen. The mixture was filtered and the filtrate was concentrated. The residue was purified by silica
chromatography, eluting with ethyl acetate: petroleum ether (1:5, v/v). The material was washed with ethyl acetate
(5 mL), filtered, washed with further ethyl acetate, then dried in vacuo. The residue was further purified by
preparative-HPLC to give the title compound 287 mg (41% yield). HPLC (method C): 98.9%, rt = 5.97. 1H NMR
(300 MHz, DMSO-d6) δH 10.33 (s, 1H), 8.34 (d, J=2.4 Hz, 1H), 7.56-7.74 (m, 6H), 6.37 (s, 1H), 4.22 (m, 2H),
3.81 (t, J=5.4 Hz, 2H), 2.36-2.51 (m, 2H). HRMS (ESI) for C16H16ClN2O3S [M+H]+ calcd.: 351.0571, found:
351.0576.
N-(2-Chloro-5-phenylpyridin-3-yl)benzenesulfonamide, 14. To a solution of N-(5-bromo-2-chloropyridin-3-
yl)benzenesulfonamide (200 mg, 0.57 mmol) and phenylboronic acid (63 mg, 0.52 mmol) in 1,4-dioxane (10 mL)
and water (5 mL) was added potassium carbonate (216 mg, 1.56 mmol) and the mixture was degassed.
Tetrakis(triphenylphosphine) palladium (0) was added and the reaction was refluxed for 13 hours. The reaction
was extracted with ethyl acetate, the organic layer was washed with brine, dried over magnesium sulfate, filtered
and evaporated. The residue was purified by silica chromatography using 10% ethyl acetate in DCM, followed by
reverse-phase purification using acetonitrile: water (1:1, v/v) to give the title compound, 50 mg (28% yield).
LCMS (method B): 100%, rt = 1.15, [M+H]+ = 345/347. 1H NMR (400 MHz, Methanol-d4) δH 8.43 (d, J=2.3 Hz,
1H), 8.16 (d, J=2.3 Hz, 1H), 7.80-7.84 (m, 2H), 7.61-7.68 (m, 3H), 7.50-7.57 (m, 4H), 7.44-7.50 (m, 1H). HRMS
(ESI) for C17H14ClN2O2S [M+H]+ calcd.: 345.0465, found: 345.0467.
N-(5-(3,6-Dihydro-2H-pyran-4-yl)-2-methoxypyridin-3-yl)benzenesulfonamide, 15. 5-(3,6-Dihydro-2H-
pyran-4-yl)-2-methoxypyridin-3-amine (compound S7, 105 mg, 0.509 mmol) was dissolved in pyridine (2 mL)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 30
1
2
3
4
5
6
7
8
9
and benzenesulfonyl chloride (0.066 mL, 0.509 mmol) was added. The reaction was stirred at room temperature
overnight. The reaction was evaporated to dryness. Half the material was dissolved in DMSO: methanol (1 mL,
1:1, v/v) and purified by MDAP (method A). Product containing fraction was evaporated to dryness to give title
compound 61 mg (35% yield). LCMS (method A): 100%, rt = 0.94, [M+H]+ = 347. 1H NMR (400 MHz, DMSO-
d6) δH 9.85 (s, 1H), 8.00 (d, J=2.3 Hz, 1H), 7.71-7.76 (m, 2H), 7.60-7.66 (m, 2H), 7.53-7.59 (m, 2H), 6.14 (tt,
J=1.6, 3.1 Hz, 1H), 4.21 (q, J=2.8 Hz, 2H), 3.81 (t, J=5.4 Hz, 2H), 3.61 (s, 3H), 2.37 (dt, J=2.8, 5.0 Hz, 2H).
HRMS (ESI) for C17H19N2O4S [M+H]+ calcd.: 347.1066, found: 347.1068.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1-Benzyl-N-(5-(3,6-dihydro-2H-pyran-4-yl)-2-methoxypyridin-3-yl)-2-methyl-1H-imidazole-4-
sulfonamide, 16. (Bromomethyl)benzene (0.05 mL, 0.414 mmol) was added to N-(5-(3,6-dihydro-2H-pyran-4-
yl)-2-methoxypyridin-3-yl)-N-(4-methoxybenzyl)-2-methyl-1H-imidazole-5-sulfonamide (compound S9, 130
mg, 0.276 mmol) and potassium carbonate (57.3 mg, 0.414 mmol) in DMF (1 mL) and stirred at room temperature
over the weekend. The reaction was quenched with water (3 mL). Ethyl acetate (5 mL) was added and the aqueous
was separated and washed with ethyl acetate (3 mL). The combined organics were washed with 7.5% LiCl (aq.)
(3 mL x 2), water (3 mL), brine (3mL), then separated, passed through a hydrophobic frit and dried under a stream
of nitrogen. DCM (1 mL) and TFA (0.5 mL) were added to the residue and the reaction was stirred overnight at
room temperature, then transferred to a microwave vial and heated in the microwave at 80 °C for a total of 30
min. The reaction was added cautiously to sat. sodium bicarbonate (aq.) (4 mL) and diluted with DCM. The DCM
was passed through a hydrophobic frit then evaporated to dryness. The residue was dissolved in DMSO: methanol
(1 mL, 1:1, v/v), filtered and purified by MDAP (method A). Product containing fraction was evaporated under a
stream of nitrogen to give the title compound, 18 mg, (15% yield). LCMS (method A): 100%, rt = 0.91, [M+H]+
1
= 441. H NMR (600 MHz, DMSO-d6) δH ppm 9.46 (br. s, 1H), 7.96 (d, J=1.5 Hz, 1H), 7.80 (s, 1H), 7.76 (d,
J=2.2 Hz, 1H), 7.28 - 7.37 (m, 3H), 7.09 (d, J=7.3 Hz, 2H), 6.12 (br. s, 1H), 5.18 (s, 2H), 4.19 (q, J=2.8 Hz, 2H),
3.79 (t, J=5.5 Hz, 2H), 3.74 (s, 3H), 2.36 (td, J=5.2, 2.8 Hz, 2H), 2.23 (s, 3 H). 13C (151 MHz, DMSO-d6) δC ppm
155.8, 146.8, 136.7, 130.6, 129.32, 129.30, 129.28, 129.2, 128.4, 127.6, 125.6, 122.8, 114.6, 65.5, 64.0, 55.6,
53.8, 49.6, 26.7, 13.3. HRMS (ESI) for C22H25N4O4S [M+H]+ calcd.: 441.1597, found: 441.1596.
Biology. PI3Kα, -β, -γ, and -δ HTRF Assays. Inhibition of PI3Kinase enzymatic activity was determined using a
homogeneous time-resolved fluorescence (HTRF) kit assay format provided by Millipore. Reactions were
performed in assay buffer containing 50 mM HEPES, pH 7.0, 150 mM NaCl, 10 mM MgCl2, <1% cholate (w/v),
<1% CHAPS (w/v), 0.05% sodium azide (w/v), and 1 mM DTT. Enzymes were preincubated with compound,
ACS Paragon Plus Environment

Page 25 of 30
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
serially diluted 4-fold in 100% DMSO, for 15 min prior to reaction initiation upon addition of substrate solution
containing ATP at Km for the specific isoform tested (α at 250 μM, β at 400 μM, δ at 80 μM, and γ at 15 μM),
PIP2 at either 5 μM (PI3Kδ) or 8 μM (PI3Kα, -β, and -γ) and 10 nM biotin-PIP3. Assays were quenched after 60
min by addition of a quench/detection solution prepared in 50 mM HEPES, pH 7.0, 150 mM NaCl, <1% cholate,
<1% Tween 20, 30 mM EDTA, 40 mM potassium fluoride, and 1 mM DTT containing 16.5 nM GRP-1 PH
domain, 8.3 nM streptavidin-APC, and 2 nM europium-anti-GST and were left for a further 60 min in the dark to
equilibrate prior to reading using a BMG RubyStar plate reader. Ratio data were normalized to high (no
compound) and low (no enzyme) controls prior to fitting using a logistical four-parameter equation to determine
IC50. Ligand efficiency was calculated according to30.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ancillary Information
Supporting Information Availability
The Supporting Information is available free of charge on the ACS Publications website.
Molecular formula strings
Crystallography methods
Chemistry methods
Assay methods for data on compound 16
Kinase screening data for compounds 7 and 16
Free-Wilson analysis, coverage and correlation plots
Abbreviations
CLND, chemiluminescent nitrogen detection; CPME, cyclopentyl methyl ether; HTRF, homogeneous time
resolved fluorescence; MDAP, mass directed automated preparative HPLC; MDCK, Madin-Darby canine kidney;
MIP, molecular interaction potential.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 30
1
2
3
4
5
6
Author Information
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Corresponding Author
*E-mail: Kenneth.d.down@gsk.com. Telephone: 00441438762042.
$Present address: Charles River, Chesterford Research Park, Saffron Walden, Essex, CB10 1XL, UK.
Notes
The authors declare the following competing financial interest(s): All authors were employees of
GlaxoSmithKline at the time the work was carried out.
Acknowledgements
The authors thank members of Protein & Cellular Sciences, PTS, GSK, for protein production; Damien Bonvalot,
Aurelie Champigny, Paul Cox, G. Huang, J. Kasparec, X. Lui, Susan Macintyre, Elvis Maduli, David Perez, Andy
Philp and Xiangsheng Li for their synthetic chemistry contributions; and Colin Edge and Zoë Henley for review
of the manuscript.
Accession Codes
Coordinates have been deposited with the Protein Data Bank with accession codes XXX, XXX & XXX.
References
1.
Cantley, L. C., The phosphoinositide 3-kinase pathway. Science 2002, 296 (5573), 1655-1657.
ACS Paragon Plus Environment

Page 27 of 30
Journal of Medicinal Chemistry
1
2
3
4
2.
Vanhaesebroeck, B.; Guillermet-Guibert, J.; Graupera, M.; Bilanges, B., The emerging mechanisms of
5
6
7
isoform-specific PI3K signalling. Nat Rev Mol Cell Biol 2010, 11 (5), 329-341.
3. Foster, J. G.; Blunt, M. D.; Carter, E.; Ward, S. G., Inhibition of PI3K signaling spurs new therapeutic
8
9
opportunities in inflammatory/autoimmune diseases and hematological malignancies. Pharmacol Rev 2012, 64
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(4), 1027-1054.
4.
of growth and metabolism. Nat Rev Genet 2006, 7 (8), 606-619.
5. McNamara, C. R.; Degterev, A., Small-Molecule Inhibitors of the PI3K Signaling Network. In Future
Medicinal Chemistry, Future Science: 2011; Vol. 3, pp 549-565.
6. Vanhaesebroeck, B.; Ali, K.; Bilancio, A.; Geering, B.; Foukas, L. C., Signalling by PI3K isoforms:
insights from gene-targeted mice. Trends Biochem Sci 2005, 30 (4), 194-204.
7. Rowan, W. C.; Smith, J. L.; Affleck, K.; Amour, A., Targeting phosphoinositide 3-kinase delta for
allergic asthma. Biochem Soc Trans 2012, 40 (1), 240-245.
8. Sriskantharajah, S.; Hamblin, N.; Worsley, S.; Calver, A. R.; Hessel, E. M.; Amour, A., Targeting
phosphoinositide 3-kinase delta for the treatment of respiratory diseases. Ann N Y Acad Sci 2013, 1280, 35-39.
9. Norman, P., Selective PI3Kδ inhibitors, a review of the patent literature. Expert Opinion on
Therapeutic Patents 2011, 21 (11), 1773-1790.
10. A. Sabbah, D.; Hu, J.; A. Zhong, H., Advances in the development of class I phosphoinositide 3-kinase
(PI3K) inhibitors. Current Topics in Medicinal Chemistry 2016, 16 (13), 1413-1426.
11. Dienstmann, R.; Rodon, J.; Serra, V.; Tabernero, J., Picking the point of inhibition: a comparative
review of PI3K/AKT/mTOR pathway inhibitors. Mol Cancer Ther 2014, 13 (5), 1021-1031.
12. Shin, Y.; Suchomel, J.; Cardozo, M.; Duquette, J.; He, X.; Henne, K.; Hu, Y. L.; Kelly, R. C.;
Engelman, J. A.; Luo, J.; Cantley, L. C., The evolution of phosphatidylinositol 3-kinases as regulators
McCarter, J.; McGee, L. R.; Medina, J. C.; Metz, D.; San Miguel, T.; Mohn, D.; Tran, T.; Vissinga, C.; Wong,
S.; Wannberg, S.; Whittington, D. A.; Whoriskey, J.; Yu, G.; Zalameda, L.; Zhang, X.; Cushing, T. D.,
Discovery, optimization, and in vivo evaluation of benzimidazole derivatives AM-8508 and AM-9635 as potent
and selective PI3Kdelta inhibitors. J Med Chem 2016, 59 (1), 431-447.
13.
Hoegenauer, K.; Soldermann, N.; Zecri, F.; Strang, R. S.; Graveleau, N.; Wolf, R. M.; Cooke, N. G.;
Smith, A. B.; Hollingworth, G. J.; Blanz, J.; Gutmann, S.; Rummel, G.; Littlewood-Evans, A.; Burkhart, C.,
Discovery of CDZ173 (leniolisib), representing a structurally novel class of PI3K delta-selective inhibitors. ACS
Med Chem Lett 2017, 8 (9), 975-980.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 30
1
2
3
4
14.
Allen, R. A.; Brookings, D. C.; Powell, M. J.; Delgado, J.; Shuttleworth, L. K.; Merriman, M.; Fahy, I.
5
6
7
8
J.; Tewari, R.; Silva, J. P.; Healy, L. J.; Davies, G. C. G.; Twomey, B.; Cutler, R. M.; Kotian, A.; Crosby, A.;
McCluskey, G.; Watt, G. F.; Payne, A., Seletalisib: Characterization of a novel, potent, and selective inhibitor of
PI3Kdelta. J Pharmacol Exp Ther 2017, 361 (3), 429-440.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15.
Down, K.; Amour, A.; Baldwin, I. R.; Cooper, A. W.; Deakin, A. M.; Felton, L. M.; Guntrip, S. B.;
Hardy, C.; Harrison, Z. A.; Jones, K. L.; Jones, P.; Keeling, S. E.; Le, J.; Livia, S.; Lucas, F.; Lunniss, C. J.;
Parr, N. J.; Robinson, E.; Rowland, P.; Smith, S.; Thomas, D. A.; Vitulli, G.; Washio, Y.; Hamblin, J. N.,
Optimization of novel indazoles as highly potent and selective inhibitors of phosphoinositide 3-kinase delta for
the treatment of respiratory disease. J Med Chem 2015, 58 (18), 7381-7399.
16.
Cahn, A.; Hamblin, J. N.; Begg, M.; Wilson, R.; Dunsire, L.; Sriskantharajah, S.; Montembault, M.;
Leemereise, C. N.; Galinanes-Garcia, L.; Watz, H.; Kirsten, A. M.; Fuhr, R.; Hessel, E. M., Safety,
pharmacokinetics and dose-response characteristics of GSK2269557, an inhaled PI3Kdelta inhibitor under
development for the treatment of COPD. Pulm Pharmacol Ther 2017, 46, 69-77.
17.
18.
19.
20.
https://clinicaltrials.gov/ct2/show/NCT02294734 (accessed July 25, 2018).
https://clinicaltrials.gov/ct2/show/NCT03345407 (accessed July 25, 2018).
https://clinicaltrials.gov/ct2/show/NCT03045887 (accessed July 25, 2018).
Amour, A.; Barton, N.; Cooper, A. W.; Inglis, G.; Jamieson, C.; Luscombe, C. N.; Morrell, J.; Peace,
S.; Perez, D.; Rowland, P.; Tame, C.; Uddin, S.; Vitulli, G.; Wellaway, N., Evolution of a novel, orally
bioavailable series of PI3Kdelta inhibitors from an inhaled lead for the treatment of respiratory disease. J Med
Chem 2016, 59 (15), 7239-7251.
21.
Sutherlin, D. P.; Baker, S.; Bisconte, A.; Blaney, P. M.; Brown, A.; Chan, B. K.; Chantry, D.;
Castanedo, G.; DePledge, P.; Goldsmith, P.; Goldstein, D. M.; Hancox, T.; Kaur, J.; Knowles, D.; Kondru, R.;
Lesnick, J.; Lucas, M. C.; Lewis, C.; Murray, J.; Nadin, A. J.; Nonomiya, J.; Pang, J.; Pegg, N.; Price, S.; Reif,
K.; Safina, B. S.; Salphati, L.; Staben, S.; Seward, E. M.; Shuttleworth, S.; Sohal, S.; Sweeney, Z. K.; Ultsch,
M.; Waszkowycz, B.; Wei, B., Potent and selective inhibitors of PI3Kδ: Obtaining isoform selectivity from the
affinity pocket and tryptophan shelf. Bioorganic & Medicinal Chemistry Letters 2012, 22 (13), 4296-4302.
22.
Andrs, M.; Korabecny, J.; Jun, D.; Hodny, Z.; Bartek, J.; Kuca, K., Phosphatidylinositol 3-Kinase
(PI3K) and phosphatidylinositol 3-kinase-related kinase (PIKK) inhibitors: importance of the morpholine ring. J
Med Chem 2015, 58 (1), 41-71.
ACS Paragon Plus Environment

Page 29 of 30
Journal of Medicinal Chemistry
1
2
3
4
23.
Cheeseright, T. J.; Mackey, M. D.; Melville, J. L.; Vinter, J. G., FieldScreen: virtual screening using
5
6
7
8
molecular fields. Application to the DUD data set. J Chem Inf Model 2008, 48 (11), 2108-2117.
24. Free, S. M.; Wilson, J. W., A mathematical contribution to structure-activity studies. Journal of
Medicinal Chemistry 1964, 7 (4), 395-399.
25. Yang, H. W.; Craven, B. M., Charge density study of 2-pyridone. Acta Crystallographica Section B
Structural Science 1998, 54 (6), 912-920.
26. Berndt, A.; Miller, S.; Williams, O.; Le, D. D.; Houseman, B. T.; Pacold, J. I.; Gorrec, F.; Hon, W.-C.;
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ren, P.; Liu, Y.; Rommel, C.; Gaillard, P.; Ruckle, T.; Schwarz, M. K.; Shokat, K. M.; Shaw, J. P.; Williams, R.
L., The p110[delta] structure: mechanisms for selectivity and potency of new PI(3)K inhibitors. Nat Chem Biol
2010, 6 (2), 117-124.
27.
D'Angelo, N. D.; Kim, T. S.; Andrews, K.; Booker, S. K.; Caenepeel, S.; Chen, K.; D'Amico, D.;
Freeman, D.; Jiang, J.; Liu, L.; McCarter, J. D.; San Miguel, T.; Mullady, E. L.; Schrag, M.; Subramanian, R.;
Tang, J.; Wahl, R. C.; Wang, L.; Whittington, D. A.; Wu, T.; Xi, N.; Xu, Y.; Yakowec, P.; Yang, K.; Zalameda,
L. P.; Zhang, N.; Hughes, P.; Norman, M. H., Discovery and optimization of a series of benzothiazole
phosphoinositide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) dual inhibitors. J Med Chem 2011,
54 (6), 1789–1811.
28.
Nishimura, N.; Siegmund, A.; Liu, L.; Yang, K.; Bryan, M. C.; Andrews, K. L.; Bo, Y.; Booker, S. K.;
Caenepeel, S.; Freeman, D.; Liao, H.; McCarter, J.; Mullady, E. L.; San Miguel, T.; Subramanian, R.; Tamayo,
N.; Wang, L.; Whittington, D. A.; Zalameda, L.; Zhang, N.; Hughes, P. E.; Norman, M. H., Phospshoinositide
3-kinase (PI3K)/mammalian target of rapamycin (mTOR) dual inhibitors: discovery and structure-activity
relationships of a series of quinoline and quinoxaline derivatives. J Med Chem 2011, 54 (13), 4735-4751.
29.
Knight, S. D.; Adams, N. D.; Burgess, J. L.; Chaudhari, A. M.; Darcy, M. G.; Donatelli, C. A.; Luengo,
J. I.; Newlander, K. A.; Parrish, C. A.; Ridgers, L. H.; Sarpong, M. A.; Schmidt, S. J.; Van Aller, G. S.; Carson,
J. D.; Diamond, M. A.; Elkins, P. A.; Gardiner, C. M.; Garver, E.; Gilbert, S. A.; Gontarek, R. R.; Jackson, J.
R.; Kershner, K. L.; Luo, L.; Raha, K.; Sherk, C. S.; Sung, C.-M.; Sutton, D.; Tummino, P. J.; Wegrzyn, R. J.;
Auger, K. R.; Dhanak, D., Discovery of GSK2126458, a highly potent inhibitor of PI3K and the mammalian
target of rapamycin. ACS Medicinal Chemistry Letters 2010, 1 (1), 39-43.
30.
Hopkins, A. L.; Groom, C. R.; Alex, A. Ligand efficiency: a useful metric for lead selection. Drug
Discovery Today 2004, 9, 430−431.
ACS Paragon Plus Environment