4,6-Diaminopyrimidines as Highly Preferred Troponin I-Interacting Kinase (TNNI3K) Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.8b00125

Structure-guided progression of a purine-derived series of TNNI3K inhibitors directed design efforts that produced a novel series of 4,6-diaminopyrimidine inhibitors, an emerging kinase binding motif. Herein, we report a detailed understanding of the intrinsic conformational preferences of the scaffold, which impart high specificity for TNNI3K. Further manipulation of the template based on the conformational analysis and additional structure-activity relationship studies provided enhancements in kinase selectivity and pharmacokinetics that furnished an advanced series of potent inhibitors. The optimized compounds (e.g., GSK854) are suitable leads for identifying new cardiac medicines and have been employed as in vivo tools in investigational studies aimed at defining the role of TNNI3K within heart failure.

Full text of DOI:10.1021/acs.jmedchem.8b00125

Article
pubs.acs.org/jmc
Cite This: J. Med. Chem. XXXX, XXX, XXX−XXX
4,6-Diaminopyrimidines as Highly Preferred Troponin I‑Interacting
Kinase (TNNI3K) Inhibitors
Joanne Philp,† Brian G. Lawhorn,*,† Alan P. Graves,‡ Lisa Shewchuk,‡ Katrina L. Rivera,†
Larry J. Jolivette,† Dennis A. Holt,† Gregory J. Gatto, Jr.,† and Lara S. Kallander†
†Heart Failure Discovery Performance Unit and ‡Platform Technology and Sciences, GlaxoSmithKline, 709 Swedeland Road, King of
Prussia, Pennsylvania 19406, United States
S
* Supporting Information
ABSTRACT: Structure-guided progression of a purine-derived series of TNNI3K inhibitors directed design efforts that
produced a novel series of 4,6-diaminopyrimidine inhibitors, an emerging kinase binding motif. Herein, we report a detailed
understanding of the intrinsic conformational preferences of the scaffold, which impart high specificity for TNNI3K. Further
manipulation of the template based on the conformational analysis and additional structure−activity relationship studies provided
enhancements in kinase selectivity and pharmacokinetics that furnished an advanced series of potent inhibitors. The optimized
compounds (e.g., GSK854) are suitable leads for identifying new cardiac medicines and have been employed as in vivo tools in
investigational studies aimed at defining the role of TNNI3K within heart failure.
INTRODUCTION
RESULTS AND DISCUSSION
■
■
Troponin I-interacting kinase (TNNI3K) is a self-regulating,
cardiac specific kinase that currently lacks a clearly defined
biological function.1,2 Overexpression of Tnni3k has been
demonstrated to lead to impaired cardiac structure and function
in murine models of cardiomyopathy, pressure overload
induced heart failure, and in response to acute ischemic
injury.3,4 Conversely, both deletion of Tnni3k via transgenic
studies and pharmacological inhibition of TNNI3K have been
assigned a cardioprotective role within an ischemia/reperfusion
setting.4 Extrapolation of these findings suggests that inhibition
of TNNI3K represents a potential novel mechanism for the
treatment of heart failure. Accordingly, selective inhibitors of
TNNI3K would prove valuable tools in the characterization of
the biological role of TNNI3K and provide prospective leads
for drug development in an area of continuing unmet medical
need.5,6
Our efforts to identify potent and selective inhibitors of this
promising target uncovered a series of 4,6-diaminopyrimidines
that displayed excellent kinase selectivity. Although 2,4-
diaminopyrimidines have proven to be a clinically relevant7,8
and an extensively explored structural class of kinase inhibitors,
we noted a much rarer incidence of 4,6-diaminopyrimidines
reported in the literature.9−11 Consideration of the bioactive
conformation of the 4,6-diaminopyrimidine template identified
the origin of its innate selectivity, and application of these
learnings allowed further enhancement of potency, pharmaco-
kinetics, and the kinase selectivity profile of the series.
Conception and Design of 4,6-Diaminopyrimidines.
As previously disclosed, exploratory screening efforts identified
the purine motif, exemplified by compound 1, as providing
potent and tractable TNNI3K inhibitors with promising levels
of kinase selectivity.12 Initial investigation of this scaffold
demonstrated that the 7-position nitrogen was not a
prerequisite for activity as deazapurine 2 retained potency
and selectivity (Table 1).13 Based on this observation and with
the availability of TNNI3K cocrystal structures to direct the
design strategy,12 efforts to identify additional alternatives to
the purine core were initiated.
Examination of the crystal structure of 1 bound to TNNI3K
highlighted five key hydrogen bond interactions that anchor the
molecule within the ATP binding site (Figure 1A). The
sulfonamide moiety projects toward the backpocket and
contacts TNNI3K via interaction of its NH with Thr539, the
“gatekeeper” residue. The sulfonamide oxygens participate in
two additional hydrogen bonding interactions, one of which is
mediated by a resident water molecule. A bidentate hinge
binding interaction was also observed between amino acid
Ile542 and the purine core, mediated via the H-bond acceptor
capacity of the 3-position nitrogen and the donating ability of
the 9-position NH. Notably, in order to simultaneously
Received: January 23, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. Initial Purine and 4,6-Diaminopyrimidine SAR
Table 2. Initial Modifications of the C6-Aniline
TNNI3K IC50
(nM)
% kinases with >10-fold
a
cmpd
R
selectivity
TNNI3K IC50
(nM)
% kinases >10-fold
a
cmpd
A
R
selectivity
6
H
63
630
1300
3200
79
95% (36 of 38)
80% (30 of 38)
7
2-F
1
2
3
4
5
N
-
-
500
80
87% (62 of 71)
94% (61 of 65)
96% (69 of 72)
97% (34 of 35)
87% (21 of 24)
b
8
2-Me
nd
CH
9
2-OMe
3-Cl
nd
-
-
-
3-Me-Ph
4-CF3−Ph
CH3
32
10
3
95% (35 of 37)
96% (69 of 72)
92% (34 of 37)
88% (22 of 25)
84% (22 of 26)
95% (35 of 39)
97% (34 of 35)
92% (23 of 25)
96% (27 of 28)
79
3-Me
32
800
Percent of kinases where off-target kinase IC50/TNNI3K IC50 > 10.13
a
11
12
13
14
4
3-OMe
3-SO2Me
4-Cl
63
200
79
accommodate each of these key H-bond interactions, 1 must
adopt a coplanar binding conformation.
4-Me
79
4-CF3
4-OMe
4-SO2Me
79
In addition to guiding exploration of the deazapurine
scaffold12 study of the crystal structure of 1 led to the
conception of 4,6-diaminopyrimidines, such as 3, which were
anticipated to maintain the crucial network of H-bonding
interactions of 1 while allowing exploration of the previously
unoccupied front pocket from an alternative trajectory. This
proved a successful design concept as potency at TNNI3K was
improved (16-fold), and critically, the promising kinase
selectivity of starting point 1 was upheld with initial 4,6-
diaminopyrimidine 3 (Table 1).13 Indeed, broader evaluation of
3 against a range of kinases indicated an appreciable level of
inhibition against only 4% (8 of 189) of kinases when tested at
1 μM inhibitor concentration in the Millipore KinaseProfiler
panel (see Supporting Information). A cocrystal structure of 4,
a close analog of 3, bound to TNNI3K confirmed that the
intricate H-bond mediated recognition pattern of 1 and
consequent coplanar configuration were maintained (Figure
1B). The critical binding contribution of the aromatic ring of
the C6 aniline in the front pocket is highlighted by the fact that
truncated analog 5 displays over an order of magnitude
reduction in TNNI3K affinity, perhaps indicating a benefit
imparted by space filling or the influence of conformational
preferences, or a combination of both of these factors.
15
16
160
79
Percent of kinases where off-target kinase IC50/TNNI3K IC50 > 10.13
a
b
nd, not determined.
analogs 7−9. In contrast, incorporation of a variety of
functional groups at both the meta and para positions was
permitted by the target. At these positions, electron with-
drawing (10, 12, 13, 4, and 16) and electron donating (3, 11,
14, and 15) groups of differing sizes could be accommodated.
Favorable kinase selectivity13 was maintained across this range
of analogs, which encouraged us to further investigate the series
and the underlying origins of its specificity for TNNI3K.
General Kinase Selectivity. To decipher the features of 3
that contribute to its unique affinity and selectivity for
TNNI3K, a series of heterocyclic core analogs were evaluated.
Replacement of the pyrimidine N3 atom with CH to give
pyridine 17 resulted in a substantially less effective inhibitor
(>100-fold) of TNNI3K, illustrating a critical role for the N3
nitrogen (Figure 2). In some instances where 4,6-diaminopyr-
imidines exhibit preferential binding to a kinase, N3 has been
observed to participate in a specific interaction that facilitates
binding and imparts significant overall kinase selectivity.10
However, the cocrystal structure of 4 did not indicate an
interaction of this nature; instead, this dramatic effect was
attributed to the introduction of a potential severe steric
interaction (∼3 kcal/mol for planar vs skewed conformation for
related diphenylamines)14c between the pyridine C5−H and
Following the encouraging data observed with 3, an
exploration of alternative C6 substituents was initiated, and
evidence of a tractable structure−activity relationship (SAR)
against TNNI3K readily emerged (Table 2), with a clear
indication of a steric driven intolerance to substitution at the
C2 ring position, reflected in the reduced activity observed with
Figure 1. Cocrystal structures of 1 (A, PDB 4YFI)12 and 4 (B, PDB 6B5J) bound to TNNI3K and (C) a 2D representation of the binding mode of
4.
B
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 2. Two-dimensional representation of the preferred TNNI3K binding mode for various heterocyclic cores.
the benzenesulfonamide C2−H (Figure 2). As a consequence,
pyridine 17 is unable to adhere to the highly preferred coplanar
binding conformation (Figure 1) that is adopted by 3 leading to
reduced activity. Similarly, it is proposed that compound 18,
the 2,4-diaminopyrimidine counterpart of 3, exhibits a reduced
potency arising from this conformational bias (Figure 2).
However, this pyrimidine exhibits a heightened (∼8-fold)
affinity for TNNI3K compared to pyridine 17, in line with the
greater capacity of 18 to conform to the s-cis orientation
required for optimal hinge binding (Figure 1).10,14 Specifically,
the N3 of 18 alleviates the steric interaction (∼2 kcal/mol for s-
cis vs s-trans for related anilinopyrimidines)14b between the
C3−H of pyridine 17 and the C6 aryl resulting in an improved
bidentate hinge binding interaction with TNNI3K. In fact, the
reduced inhibition of TNNI3K observed for 2,4-diaminopyr-
imidine 18 compared to 4,6-diaminopyrimdine 3 represents a
departure from the norm, as most kinases prefer 2,4-
diaminopyrimidines because of this innate ability to adopt the
s-cis configuration.9,10,15 As a result, in contrast to 3, 2,4-
diaminopyrimdine 18 is a promiscuous kinase inhibitor that
exhibits substantial activity against many (15/21) of the kinases
examined.10 These observations also explain the reduction of
TNNI3K affinity associated with ortho-substituted C6 aryls
(Table 2, 7−9) wherein the preferred s-cis configuration is
further compromised. Incorporating these SAR and structure
informed conformational considerations into compound design
led to preparation of triazine 19, which simultaneously relieves
the steric penalties observed in the 2,4- and 4,6-diaminopyr-
imdine frameworks upon enzyme binding. Consequently, 19
exhibits comparable or increased affinity over 3 (1.6-fold) and
18 (32-fold), and the kinase selectivity associated with these
conformational preferences is restored (Figure 2).13
simultaneously serves to destabilize the alternative s-trans
orientation by introducing a lone pair−lone pair repulsion.14b
As such, enhanced potency (2.5-fold) is achieved within pyridyl
pyrimidine analog 21 compared to the analogous phenyl
triazine 20. Comparison of the potency of the phenyl vs 2-
pyridyl analog pairs 13/21 and 4/22 confirms the conforma-
tional hypothesis and defines the magnitude (∼4−5-fold) of
the conformational bias and resultant enhanced hinge binding
capacity imparted by the pyridyl nitrogen. Independent of our
efforts, an analogous 2-pyridyl unit has been employed in the
design of Lck inhibitors based on a similar conformational
rationale.14b However, as direct phenyl analogs of the 2-pyridyl
substituents were not reported in that study, the experiments
disclosed herein confirm the value of the 2-pyridyl design,
provide the magnitude of the effect experimentally (0.8 kcal/
mol using ΔG1 − ΔG2 = RT ln(K1/K2) where K is
approximated by IC50 values of 13 and 21 or 4 and 22), and
distinguish the contribution of the pyridyl N toward stabilizing
the s-cis conformation from other binding benefits arising from
the incorporation of C6-aryl amines (e.g., 3, IC50 = 32 nM)
instead of C6-alkyl amines (e.g., 5, IC50 = 800 nM). As
anticipated, pyridines 21 and 22 maintain the good overall
kinase selectivity profile of the parent system (>10-fold
selectivity vs >90% of kinases tested)13 and represent optimal
leads for further development.
Overall, the TNNI3K selectivity of 4,6-diaminopyrimidines
(e.g., 3) can be primarily attributed to their unique ability to
adopt a coplanar binding conformation (Figure 2). This allows
the compounds to engage the hinge region H-bond interactions
while simultaneously positioning the sulfonamide to hydrogen
bond with the gatekeeper threonine (unique to 19% of the
kinome) and form polar contacts within the backpocket (Figure
1). As within other reports of favored binding by 4,6-
diaminopyrimidines the N3 nitrogen has a pivotal role in the
attainment of this preferential binding, and as such the 4,6-
diaminopyrimdine scaffold may be considered a privileged
structure for obtaining selectivity against the kinome.10,14b
EGFR Selectivity. Although parent compound 3 generally
demonstrated an excellent kinase selectivity profile, a potential
An alternative but equally effective approach to favor the
optimal TNNI3K binding mode within the 4,6-diaminopyr-
imidine system was realized in the design of compounds 21 and
22 (Figure 2), which incorporates a 2-pyridyl substituent in
place of the C6 aniline of 3. Inclusion of the pyridyl nitrogen
reinforces the requisite s-cis conformation by removing the
steric interaction with C5−H of the pyrimidine and
C
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
liability for the series is its potent inhibition of epidermal
growth factor receptor (EGFR) kinase. Although a valid
antioncogenic target, inhibition of EGFR has been associated
with cardiac dysfunction.16 The presence of such an off-target
activity would also confound interpretation of target validation
studies for TNNI3K. Therefore, it was deemed necessary to
attenuate this off-target activity to progress the 4,6-
diaminopyrimdines. This undesired activity is consistent with
findings that a related 4,6-diaminopyrimidine template can bind
preferentially to EGFR.10a Encouragingly, initial SAR efforts
demonstrated that replacement of the 3-methyl substituent of 3
with a larger group such as the isopropyl of 23 (Table 3)
as a result of two point changes between the active sites
(Ala605 to Thr854 and Leu513 to Met766). Thus, it is
hypothesized that there is insufficient space within EGFR to
tolerate the sulfonamide projecting into its back pocket. These
observations suggest that within EGFR the 4,6-diaminopyr-
imidines reside in a flipped binding mode wherein the smaller
C6-aniline group occupies the back pocket (Figure 3). In this
binding proposal, which mirrors that put forward by Gray for
related 4,6-diaminopyrimidine EGFR inhibitors, the pyrimidine
N1 participates in a hydrogen bond with the gatekeeper Thr790
leading to its uniquely potent activity at EGFR.10
Further evidence for this alternative binding mode is
exemplified by the relative insensitivity of EGFR activity to
the methylation of the C6 linked NH (27, Table 3). In contrast,
TNNI3K activity is vastly reduced by the removal of this hinge
binding NH. Conversely, alkylation of the C4 NH (28), which
represents the hinge binding NH within EGFR, results in
serious attenuation of activity at this kinase. Interestingly, this
methylation is also detrimental to TNNI3K affinity because 28
is prevented from assuming the crucial coplanar binding
conformation as this would introduce at least one severe H/H
overlapping interaction between the N-Me and its neighboring
aryl hydrogens.12
Table 3. Optimization of Selectivity for TNNI3K vs EGFR
TNNI3K
EGFR
selectivity
a
cmpd
R1
R2
R3
IC50 (nM) IC50 (nM)
(fold)
3
H
H
Me
32
50
3
25
0.1
0.5
1
As a consequence of this alternative binding mode in EGFR,
divergent SAR is also observed with C6-phenyl analog 13
(EGFR IC50 = 13 nM) and its C6-pyridyl derivative 21 (EGFR
IC50 = 630 nM, Figure 2), indicating the pyridyl nitrogen is not
well tolerated in the lipophilic back pocket of EGFR. As a
result, 21 exhibits significant selectivity (32-fold) for TNNI3K
over EGFR.
23
24
25
26
27
28
H
H
i-Pr
t-Bu
H
H
400
79
400
1600
4
H
H
SO2NHMe
NO2
20
b,c
H
H
nd
nd
H
Me
H
Me
2000
20
0.01
Me
Me
8000
2000
0.25
a
b
c
Selectivity = EGFR IC50/TNNI3K IC50. nd, not determined. Cmpd
Pharmacokinetic Properties. The initial 4,6-diaminopyr-
imidine inhibitor 3 showed a poor pharmacokinetic profile with
high clearance and low oral bioavailability in the rat (Table 4).
To reduce the metabolic liability of 3, the 3-methyl group was
replaced with a halogen to give analogs 10 and 29, which
offered a modest improvement in clearance. Relocating the
halogen substituent to the 4-position of the C6 aryl proved
beneficial with analog 13 providing a further 2-fold reduction in
clearance versus 10 and 29. These results suggest oxidation of
the aniline ring may be a primary metabolic pathway for 3 that
can be blocked through appropriate substitution. Compound
13 also offered enhanced oral bioavailability over 3, which may
result from reduced first-pass clearance. Importantly, TNNI3K
inhibitor 21, which had been optimized for potency by
inclusion of the 2-pyridyl moiety, maintained the attractive
PK profile of its phenyl analog 13, indicating the 4-Cl-2-pyridyl
aniline is an optimal C6-moiety.
26 showed modest activity (IC50 = 500 nM) in a TNNI3K cellular
assay.
resulted in a 7-fold reduction of EGFR inhibition with minimal
impact on TNNI3K affinity. Further increasing the steric
demand of this substituent, as highlighted by t-butyl analog 24,
provided an inhibitor that was equipotent at TNNI3K and
EGFR. Bis-sulphonamide compound 25 represented a key
advance, as it led to an inversion of TNNI3K-EGFR preference
compared to parent compound 3. SAR indicated that this
selectivity gain was not governed by electronics given that
analog 26, containing the strongly electron withdrawing nitro
group, displayed comparable EGFR inhibition to 3. To fully
interpret this steric-driven effect on EGFR activity, compound
3 was docked into a crystal structure (PDB 4I23) of EGFR
(Figure 3). Although both enzymes contain a threonine
gatekeeper, EGFR has a smaller back pocket than TNNI3K
Figure 3. Model of 3 bound to TNNI3K (A, PDB 6B5J) and EGFR (B, PDB 4I23).
D
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 4. Pharmacokinetics of Initial 4,6-Diaminopyrimidines
cmpd
X
R
rat Cl (mL/min/kg)
rat PO DNAUC (h·kg/L)
oral F (%)
TNNI3K IC50 (nM)
3
CH
CH
CH
CH
N
3-Me
3-Cl
3-Br
4-Cl
4-Cl
86
52
53
23
16
0.037
0.064
0.10
19
20
32
79
79
79
20
10
29
13
21
33
0.71
100
65
0.67
Table 5. Optimization of the C4-Substituent
cmpd
X
R
rat Cl (mL/min/kg)
25
rat PO DNAUC (h·kg/L)
oral F (%)
TNNI3K IC50 (nM)
13
30
31
32
33
34
35
36
37
38
39
40
41
CH
CH
CH
CH
CH
CH
CH
CH
N
H
0.71
1.1
100
nd
55
nd
100
52
nd
30
75
84
67
37
99
79
200
130
160
50
a
F
nd
Cl
Me
14
nd
26
4.5
nd
52
9.9
16
12
20
7.4
0.70
nd
OMe
0.69
2.1
SMe
400
2500
79
NHMe
NMe2
nd
0.096
1.3
b
OMe
≤10
CH
N
OCF3
0.86
0.91
0.32
2.3
32
c
OCF3
nd
CH
N
OCH2CF3
OCH2CF3
25
≤10
a
b
nd, not determined. Compound IC50 values less than 10 nM could not be accurately obtained due to titration of TNNI3K in the enzyme assay.
c
Cmpd 39 showed enhanced activity (IC50 = 63 nM) over 38 (IC50 = 320 nM) in a TNNI3K cellular assay.
Next, we turned our attention to exploiting other positions
within the molecule to further enhance the inhibitory activity
and PK profile of the series. Based on the X-ray crystal structure
of 4 bound to TNNI3K (Figure 1), we selected to probe the
C4 position of the benzenesulfonamide, as C4-substituents
were expected to project favorably into a relatively open area
within the TNNI3K binding pocket. The initial analogs (30−
36, Table 5) indicated that substitutions at the C4 position
greatly influenced TNNI3K activity, but modeling of the
substituents in the TNNI3K active site revealed that steric
effects could not account for the observed SAR. In our recent
publication, the varying effects of these C4-substituents on
TNNI3K binding were rationalized based on three separate
phenomena: (1) electronic modulation of the sulfonamide H-
bond donor interaction with TNNI3K (cf. 31, 35), (2)
intramolecular hydrogen bonding between C4-substituents and
the neighboring aniline NH (cf. 32, 33), and (3) stereo-
electronic effects of mono- versus disubstituted ortho-anilines
(cf. 35, 36).17 As a result of these effects, highly electron
donating substituents exhibit poor potency (e.g., 35) by
disrupting hydrogen bonding between the inhibitor and
TNNI3K, whereas groups that lack an ability to form an
internal hydrogen bond with the ortho-aniline NH are also
moderately potent (e.g., 30, 31, 32, 34). However, the mildly
electron donating MeO substituent (33), which can engage in
an internal H-bond with the ortho-aniline NH, exhibited good
TNNI3K activity (IC50 = 50 nM) and retained the favorable PK
properties of the unsubstituted parent 13.
This result prompted a survey of additional C4-ether analogs
(37−41), which revealed that compounds bearing ethers with
reduced electron-donating ability (e.g., 40−41) or electron-
withdrawing ability (e.g., 38-39) displayed enhanced activity at
TNNI3K while maintaining excellent PK properties. These
results also served to further verify the value of the C6 aryl
pyridyl nitrogen as this modification consistently provided
elevated inhibitory activity against TNNI3K in conjunction
with reduced clearance across each of the matched pairs (Table
5; 33/37, 38/39 and 40/41).
While optimizing the pharmacokinetic properties of the
TNNI3K inhibitors, we found TPSA to be an important
determinant of oral bioavailability in the series (Figure 4) as
compounds exhibiting TPSA ≥ 123 Å2 were consistently
observed to show low oral exposure (F < 30% for 28 of 28
tested) in rats. By contrast, compounds with TPSA < 123 Å2
showed a reasonable likelihood for high oral bioavailability (F >
30% for 69 of 128 tested), and those compounds displaying
E
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 4. Plot of oral bioavailability (F) vs TPSA for benzenesulfo-
namide TNNI3K inhibitors.
poor oral exposure within this class typically exhibited very low
aqueous solubility or contained a charged functional group,
which may negatively impact their passive permeability. This
finding is consistent with the historical use of TPSA and PSAd
in drug design, and it is noteworthy that TPSA was found to be
a superior predictor of oral bioavailability within this series
when compared to any other calculated property (e.g., cLogP,
%PSA, HBA, HBD), measured property (e.g., artificial
membrane permeability, ChromLogD), or a multiparameter
permeability model.18 This highlights the remarkable ability of
the simple TPSA calculation to capture the physical processes
(e.g., desolvation) involved in passive membrane permeability
and its utility for guiding lead optimization when competing
effects (e.g., internal H-bonds, varying charge states) do not
obscure its predictive relationship with oral bioavailability.18
B-Raf Selectivity. The 4,6-diaminopyrimidine 37 is a highly
potent analog that showed promising PK and notably also
maintains the broad spectrum kinase selectivity of the series,
with >50-fold selectivity against 27 of 29 (93%) kinases
assessed. However, 37 displays potent activity against B-Raf
(IC50 = 20 nM), a kinase that shares 82% sequence similarity
with TNNI3K among its ATP-binding site residues, and
elimination of this activity was considered critical since
inhibition of Raf kinases has been linked to effects in heart
failure models.19 A comparison of 37 modeled into X-ray
crystal structures of TNNI3K and B-Raf (Figure 5) reveals a
key residue difference between the two kinase active sites. In
TNNI3K, Leu595 resides below the 4,6-diaminopyrimidine
inhibitor, whereas B-Raf contains Phe582 in the corresponding
position. We have previously introduced selectivity for
TNNI3K over B-Raf by exploiting the electrostatic differences
between these two residues in a related series of TNNI3K
inhibitors.17,20 To optimize the 4,6-diaminopyrimidine inhib-
itors we sought a complementary approach that would
introduce selectivity for TNNI3K based on the observed steric
differences between TNNI3K and B-Raf. We noted the
proximity of the 4-OMe group of 37 to Leu595 of TNNI3K
and the Phe582 residue of B-Raf (Figure 5) and hypothesized
that larger 4-substituents might be accommodated by TNNI3K
while introducing a steric clash with the larger B-Raf Phe582
residue. Additional residues near the 4-OMe group of 37 are
conserved between TNNI3K (Gly470, Ser471, Asn593) and B-
Raf (Gly463, Ser464, Asn580), but Ser464 in B-Raf occupies an
alternate position relative to Ser471 in TNNI3K that further
constrains the size of the B-Raf pocket relative to TNNI3K
(Figure 5).
Figure 5. Models of 37 in TNNI3K (A, PDB 6B5J) and B-Raf (B,
PDB 4YHT) and 43 in TNNI3K (C, PDB 6B5J) and B-Raf (D, PDB
4YHT).
Larger 4-substituents that remain in the plane of the aryl ring
did not impart selectivity as compounds such as 41 exhibited
potent B-Raf activity (IC50 = 13 nM). Thus, we synthesized and
evaluated a series of 4-sulfonyl analogs (Table 6) of varying
sizes, anticipating that these groups would project out of the
plane of the aryl rings and project toward Leu595 and Ser471
of TNNI3K and clash with Phe582 and Ser464 of B-Raf
(Figure 5). Other critical features of the sulfone moiety that
pointed to its selection as an ideal 4-substituent included its
electron-withdrawing nature and ability to form an internal H-
bond with the neighboring aniline NH, both of which are
required for optimal TNNI3K binding (vide supra).17,21 An
initial methyl sulfone analog 42 retained good potency at
TNNI3K and high oral exposure in rats while reducing B-Raf
activity by 10-fold. Expanding the size of the sulfonyl group
further attenuated B-Raf binding in a stepwise fashion (43−45)
with no detrimental impact on TNNI3K affinity, such that the
tert-butyl sulfone (45) exhibited ≥200-fold selectivity for
TNNI3K over B-Raf. This significant size-dependent enhance-
ment in selectivity corroborates our model showing that
TNNI3K contains a larger pocket in this region compared to B-
Raf (Figure 5). In addition, the model of 43 bound to B-Raf
suggests that one of its sulfonyl oxygens is in close contact with
the backbone carbonyl of Ser464, likely creating an electrostatic
repulsion that further contributes to the loss in potency for B-
Raf. The oral exposure of the i-Pr (44) and t-Bu (45) analogs
was reduced compared to 42, and therefore, the ethyl sulfone
43 was deemed to have the optimal balance of selectivity and
PK properties for use as an in vivo tool in heart failure models.
Compound 43 (GSK854)4 exhibits substantial selectivity
against B-Raf (>25-fold) and EGFR (IC50 > 25 μM), excellent
broad spectrum kinase selectivity with >100-fold selectivity over
96% of kinases tested (≤50% inhibition @ 1 μM for 282 of 294
kinases, see Supporting Information), high activity in a
F
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 6. 4-Sulfonylbenzenesulfonamide SAR
a
b
cmpd
R
TNNI3K IC50 (nM)
B-Raf IC50 (nM)
selectivity (fold)
rat PO DNAUC (h·kg/L)
c
37
42
43
44
45
OMe
≤10
20
200
250
500
≥2
15
1.3
SO2Me
SO2Et
SO2iPr
SO2tBu
13
≤10
≤10
≤10
0.76
0.56
0.27
0.076
≥25
≥50
≥200
2000
a
b
c
Raf activity was evaluated using the B-Raf (V600E) mutant kinase. Selectivity = B-Raf IC50/TNNI3K IC50. Compound IC50 values less than 10
nM could not be accurately obtained due to titration of TNNI3K in the enzyme assay.
Scheme 1
TNNI3K cellular assay (IC50 = 8 nM), good pharmacokinetics
in rats (Cl = 14 mL/min/kg, Vdss = 1.9 L/kg, t1/2 = 2.3 h, F =
48%), and useful exposures in mice (100 mg/kg/d dietary dose
gives blood concentration = 0.4−1.6 μM total, 14−56 nM free
and heart concentration = 1.1 μM total, mouse PPB Fu = 3.5%)
and thus has been employed in rodent models and cellular
assays to delineate the cardiac biology of TNNI3K.4 Notably,
mice treated with 43 for 6 weeks following a myocardial
infarction exhibited reduced left ventricular dysfunction and
remodeling compared to vehicle controls, and these effects
were linked to reduced ROS and reduced p38 MAP kinase
activation.4
coupling with the monochloride intermediate 48 (Scheme 1,
Method B).10,11
CONCLUSIONS
■
Structure-based design led to the discovery of a series of 4,6-
diaminopyrimidines that function as highly specific TNNI3K
inhibitors that are valuable probes for understanding the
biology of TNNI3K. Further scrutiny of the binding mode of
these 4,6-diaminopyrimidines and the SAR of alternative
heterocycles demonstrated that the 4,6-diaminopyrimidine
scaffold uniquely produces potent and selective inhibitors as a
result of its intrinsic conformational preferences. Application of
this analysis to analog design produced further enhancements
in TNNI3K affinity. Concurrent interpretation of additional
SAR uncovered simple and predictable strategies to weaken
activity at EGFR and B-Raf, two homologous kinases for which
the 4,6-diaminopyrimidines also serve as a privileged structure,
albeit by virtue of a variant mechanism and binding mode for
EGFR. The pharmacokinetics of the 4,6-diaminopyrimidines
was also optimized, which culminated in the discovery of 43, a
preferred TNNI3K inhibitor tool compound and lead for
discovery of new medicines.
CHEMISTRY
■
4,6-Diaminopyrimidine TNNI3K inhibitors were synthesized
through sequential, acid-mediated (HCl or AgOTf) substitu-
tions10,11,17 of 4,6-dichloropyrimidine (46) with appropriately
adorned anilines (Scheme 1, Method A), which were either
commercially available or synthesized as described previ-
ously.12,22 Typically, microwave irradiation was employed to
accelerate reaction rates on small scale, while conventional
heating conditions were used for multigram preparations.23
Pyridine (17), 2,4-diaminopyrimidine (18), and triazine (19,
20) variants were synthesized in an analogous fashion from 2,4-
dichloropyridine, 2,4-dichloropyrimidine, or 2,4,6-trichloro-
1,3,5-triazine, respectively. Incorporation of 2-aminopyridine
substituents was accomplished through Pd-catalyzed cross
EXPERIMENTAL SECTION
■
General Experimental. Compounds were evaluated for activity
against TNNI3K and B-Raf (V600E) as previously described.12 For
the TNNI3K enzyme assay, compound IC50 values less than 10 nM
could not be accurately obtained due to titration of TNNI3K in the
G
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
assay. TNNI3K cellular assays, rat pharmacokinetics, and X-ray
crystallography studies were conducted as previously described.12 All
studies were conducted in accordance with the GSK Policy on the
Care, Welfare, and Treatment of Laboratory Animals and were
reviewed by the Institutional Animal Care and Use Committee either
at GSK or by the ethical review process at the institution where the
work was performed. The purity of each inhibitor was determined to
be >95% on an Agilent 1100 HPLC equipped with a Sunfire C18 5.0
μm column (3.0 mm × 50 mm) using a gradient of 10% to 100%
MeCN/H2O/0.05% TFA at 1 mL/min flow rate with detection at 220
and 254 nm. Mass determinations were conducted using an Agilent
6110 Quadrupole MS with positive ESI. Preparative HPLC was
conducted on a Waters 2525 system at a flow rate of 50 mL/min with
254 nm detection using either a Sunfire C18 OBD 5 μm column (30 ×
150 mm) with a gradient of MeCN/H2O/0.1% TFA or an Xbridge
C18 OBD 5 μm column (30 × 150 mm) with a gradient of MeCN/aq
NH4OH pH 10. Flash column chromatography was conducted on
silica gel eluting with EtOAc−hexanes or MeOH−CH2Cl2 mixtures or
with a sequence of CH2Cl2, Et2O, EtOAc, and acetone. Compounds
were synthesized using procedures that followed reported proto-
cols,17,22 and specific experimental details have been published
previously for compounds 1, 2, 13, 30−36, 38, and 40.12,17 Reagents
and building blocks were purchased and used as is or synthesized
following reported procedures.12,22
prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methylbenzene-
sulfonamide and aniline using Method A as a white solid in 35% yield
(84 mg): 1H NMR (400 MHz, DMSO-d6) δ 9.75 (br. s., 1H), 9.49 (br.
s., 1H), 8.38 (s, 1H), 8.07 (br. s., 1H), 7.87 (d, J = 8.03 Hz, 1H), 7.53
(d, J = 6.78 Hz, 3H), 7.46 (d, J = 4.27 Hz, 1H), 7.29−7.41 (m, 3H),
7.02−7.16 (m, 1H), 6.99 (s, 1H), 6.20 (s, 1H), 2.44 (d, J = 4.27 Hz,
3H). MS (m/z) 356.0 (M + H+).
3-({6-[(2-Fluorophenyl)amino]-4-pyrimidinyl}amino)-N-methyl-
benzenesulfonamide Trifluoroacetate (7). Compound 7 was
prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methylbenzene-
sulfonamide and 2-fluoroaniline using Method A as a yellow solid in
1
26% yield (33 mg): H NMR (400 MHz, DMSO-d6) δ 9.71 (s, 1H),
9.21 (s, 1H), 8.33 (s, 1H), 8.08 (s, 1H), 7.83−7.89 (m, 1H), 7.71−
7.78 (m, 1H), 7.49−7.56 (m, 1H), 7.45 (q, J = 4.85 Hz, 1H), 7.36 (d, J
= 7.78 Hz, 1H), 7.27−7.34 (m, 1H), 7.17−7.25 (m, 2H), 6.12 (s, 1H),
2.44 (d, J = 5.02 Hz, 3H). MS (m/z) 374.1 (M + H+).
N-Methyl-3-({6-[(2-methylphenyl)amino]-4-pyrimidinyl}amino)-
benzenesulfonamide (8). Compound 8 was prepared from 3-[(6-
chloro-4-pyrimidinyl)amino]-N-methylbenzenesulfonamide and 2-
methylaniline using Method A as a white solid in 19% yield (52
mg): 1H NMR (400 MHz, DMSO-d6) δ 9.85 (br. s., 1H), 9.21 (br. s.,
1H), 8.34 (s, 1H), 8.03 (br. s., 1H), 7.80 (d, J = 8.03 Hz, 1H), 7.50−
7.57 (m, 1H), 7.42−7.49 (m, 1H), 7.40 (d, J = 7.53 Hz, 1H), 7.19−
7.37 (m, 4H), 5.82 (s, 1H), 2.43 (d, J = 4.77 Hz, 3H), 2.23 (s, 3H).
MS (m/z) 370.1 (M + H+).
Method A. General Coupling Method for Anilines and Chloro-
pyrimidines. A mixture of a chloro-heterocycle and the required
aniline (1.3 equiv) in i-PrOH or NMP (2 mL) was treated with 1 M
aqueous HCl (1 equiv) and subjected to microwave irradiation (150
°C) for 15−60 min before being cooled to room temperature.
Analytically pure solids were obtained after purification by reverse-
phase HPLC (4, 6, 7, 8, 10, 11, 12, 16, 23, 24, 25, 26, 27, and 28),
flash column chromatography (3 and 9), or by collecting the
precipitate by filtration (5, 14, and 15).
N-Methyl-3-[(6-{[2-(methyloxy)phenyl]amino}-4-pyrimidinyl)-
amino]benzenesulfonamide (9). Compound 9 was prepared from 3-
[(6-chloro-4-pyrimidinyl)amino]-N-methylbenzenesulfonamide and 2-
1
methoxyaniline using Method A in 85% yield (345 mg): H NMR
(400 MHz, DMSO-d6) δ 9.46 (s, 1H), 8.45 (s, 1H), 8.26 (s, 1H), 8.10
(s, 1H), 7.88 (d, J = 8.28 Hz, 1H), 7.67 (d, J = 7.53 Hz, 1H), 7.49 (t, J
= 7.91 Hz, 1H), 7.38−7.44 (m, 1H), 7.30 (d, J = 7.78 Hz, 1H), 7.04−
7.16 (m, 2H), 6.91−7.00 (m, 1H), 6.07 (s, 1H), 3.83 (s, 3H), 2.44 (d,
J = 5.02 Hz, 3H). MS (m/z) 386.1 (M + H+).
Method B. General Coupling Method for 2-Aminopyridines and
Chloro-pyrimidines. A mixture of a chloro-heterocycle, the required
aniline (1.3 equiv), Pd2(dba)3 (0.02 equiv), Xantphos (0.04 equiv),
and K3PO4 in 1,4-dioxane (1.3 mL) was subjected to microwave
irradiation (150 °C) for 30−90 min before being cooled to room
temperature. Analytically pure solids were obtained after purification
by reverse-phase HPLC (21, 22, 37, 39, 41, 42, 44, and 45).
N-Methyl-3-({6-[(3-methylphenyl)amino]-4-pyrimidinyl}amino)-
benzenesulfonamide Trifluoroacetate (3). Compound 3 was
prepared from 6-chloro-N-(3-methylphenyl)-4-pyrimidinamine and
3-amino-N-methylbenzenesulfonamide using Method A as a cream
solid in 15% yield (89 mg): 1H NMR (400 MHz, DMSO-d6) δ 9.75 (s,
1H), 9.43 (br. s., 1H), 8.37 (s, 1H), 8.02−8.11 (m, 1H), 7.87 (dd, J =
1.51, 8.03 Hz, 1H), 7.54 (t, J = 7.91 Hz, 1H), 7.46 (q, J = 4.85 Hz,
1H), 7.38 (d, J = 7.78 Hz, 1H), 7.29−7.35 (m, 2H), 7.20−7.27 (m,
1H), 6.90 (d, J = 7.28 Hz, 1H), 6.18 (s, 1H), 2.44 (d, J = 4.77 Hz, 3H),
2.31 (s, 3H). MS (m/z) 370.1 (M + H+).
3-({6-[(3-Chlorophenyl)amino]-4-pyrimidinyl}amino)-N-methyl-
benzenesulfonamide Trifluoroacetate (10). Compound 10 was
prepared from 6-chloro-N-(3-chlorophenyl)-4-pyrimidinamine and 3-
amino-N-methylbenzenesulfonamide using Method A as a white solid
1
in 38% yield (126 mg): H NMR (400 MHz, DMSO-d6) δ 9.68 (s,
1H), 9.54 (s, 1H), 8.41 (s, 1H), 8.09 (t, J = 1.88 Hz, 1H), 7.90−7.94
(m, 1H), 7.88 (t, J = 2.01 Hz, 1H), 7.53 (t, J = 7.91 Hz, 1H), 7.41−
7.49 (m, 2H), 7.29−7.39 (m, 2H), 7.03 (dd, J = 1.25, 8.03 Hz, 1H),
6.21 (s, 1H), 2.45 (d, J = 4.77 Hz, 3H). MS (m/z) 390.1 (M + H+).
N-Methyl-3-[(6-{[3-(methyloxy)phenyl]amino}-4-pyrimidinyl)-
amino]benzenesulfonamide Trifluoroacetate (11). Compound 11
was prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methylben-
zenesulfonamide and 3-(methyloxy)aniline using Method A as a pale
yellow solid in 32% yield (70 mg): 1H NMR (400 MHz, DMSO-d6) δ
9.77 (s, 1H), 9.49 (br. s., 1H), 8.38 (s, 1H), 8.06 (s, 1H), 7.88 (d, J =
8.28 Hz, 1H), 7.54 (t, J = 7.91 Hz, 1H), 7.46 (q, J = 4.68 Hz, 1H), 7.38
(d, J = 7.78 Hz, 1H), 7.21−7.29 (m, 1H), 7.18 (s, 1H), 7.09 (d, J =
8.03 Hz, 1H), 6.65 (dd, J = 2.01, 8.03 Hz, 1H), 6.22 (s, 1H), 3.76 (s,
3H), 2.44 (d, J = 4.77 Hz, 3H). MS (m/z) 386.1 (M + H+).
N-Methyl-3-[(6-{[4-(trifluoromethyl)phenyl]amino}-4-
pyrimidinyl)amino]benzenesulfonamide Trifluoroacetate (4). Com-
pound 4 was prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-
methylbenzene-sulfonamide and 4-(trifluoromethyl)aniline using
N-Methyl-3-[(6-{[3-(methylsulfonyl)phenyl]amino}-4-
pyrimidinyl)amino]benzenesulfonamide Trifluoroacetate (12).
Compound 12 was prepared from 3-[(6-chloro-4-pyrimidinyl)-
amino]-N-methylbenzenesulfonamide and 3-(methylsulfonyl)aniline
1
Method A as a white solid in 21% yield (52 mg): H NMR (400
MHz, DMSO-d6) δ 9.71 (s, 1H), 9.67 (s, 1H), 8.42 (s, 1H), 8.11 (t, J
= 1.88 Hz, 1H), 7.93 (dd, J = 1.51, 8.28 Hz, 1H), 7.86 (d, J = 8.78 Hz,
2H), 7.65 (d, J = 8.78 Hz, 2H), 7.53 (t, J = 8.03 Hz, 1H), 7.45 (q, J =
5.02 Hz, 1H), 7.35 (d, J = 7.78 Hz, 1H), 6.27 (s, 1H), 2.45 (d, J = 5.02
Hz, 3H). MS (m/z) 423.9 (M + H+).
N-Methyl-3-{[6-(methylamino)-4-pyrimidinyl]amino}-
benzenesulfonamide Hydrochloride (5). Compound 5 was prepared
from 6-chloro-N-methyl-4-pyrimidinamine and 3-amino-N-methylben-
zenesulfonamide using Method A as a white solid in 57% yield (106
mg): 1H NMR (400 MHz, DMSO-d6) δ 10.61 (br. s., 1H), 8.82 (br. s.,
1H), 8.44 (s, 1H), 7.95 (br. s., 1H), 7.75 (br. s., 1H), 7.55−7.67 (m,
2H), 7.52 (d, J = 7.28 Hz, 1H), 6.08 (br. s., 1H), 2.88 (br. s., 3H), 2.45
(d, J = 4.77 Hz, 3H). MS (m/z) 294.0 (M + H+).
1
using Method A as a white solid in 27% yield (62 mg): H NMR
(400 MHz, DMSO-d6) δ 9.79 (s, 1H), 9.74 (s, 1H), 8.43 (s, 1H), 8.21
(s, 1H), 8.08 (s, 1H), 7.99 (d, J = 7.78 Hz, 1H), 7.93 (d, J = 8.03 Hz,
1H), 7.49−7.63 (m, 3H), 7.43−7.49 (m, 1H), 7.37 (d, J = 7.78 Hz,
1H), 6.24 (s, 1H), 3.22 (s, 3H), 2.45 (d, J = 4.52 Hz, 3H). MS (m/z)
434.1 (M + H+).
N-Methyl-3-({6-[(4-methylphenyl)amino]-4-pyrimidinyl}amino)-
benzenesulfonamide Hydrochloride (14). Compound 14 was
prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methylbenzene-
sulfonamide and 4-methylaniline using Method A as a white solid in
1
15% yield (7.8 mg): H NMR (400 MHz, DMSO-d6) δ 9.49 (s, 1H),
9.13 (s, 1H), 8.30 (s, 1H), 8.07−8.14 (m, 1H), 7.85−7.92 (m, 1H),
7.50 (t, J = 8.03 Hz, 1H), 7.37−7.46 (m, 3H), 7.31 (d, J = 7.78 Hz,
N-Methyl-3-{[6-(phenylamino)-4-pyrimidinyl]amino}-
benzenesulfonamide Trifluoroacetate (6). Compound 6 was
H
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1H), 7.13 (d, J = 8.28 Hz, 2H), 6.15 (s, 1H), 2.44 (d, J = 5.02 Hz, 3H),
2.27 (s, 3H). MS (m/z) 370.1 (M + H+).
with a solution of 3-amino-N-methylbenzenesulfonamide (808 mg, 4.3
mmol) in AcOH (1.6 mL) and H2O (6.8 mL) dropwise over 15 min.
The mixture was stirred an additional 10 min before being diluted with
saturated aqueous NaCl. The precipitate was collected by filtration,
washed with H2O and Et2O, and dried to give 3-[(4,6-dichloro-1,3,5-
triazin-2-yl)amino]-N-methylbenzenesulfonamide as an off-white solid
N-Methyl-3-[(6-{[4-(methyloxy)phenyl]amino}-4-pyrimidinyl)-
amino]benzenesulfonamide Hydrochloride (15). Compound 15 was
prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methylbenzene-
sulfonamide and 4-(methyloxy)aniline using Method A as an off-white
solid in 50% yield (100 mg): 1H NMR (400 MHz, DMSO-d6) δ 10.28
(br. s., 1H), 9.96 (br. s., 1H), 8.42 (s, 1H), 7.99 (s, 1H), 7.77 (d, J =
8.03 Hz, 1H), 7.51−7.62 (m, 2H), 7.47 (d, J = 7.78 Hz, 1H), 7.35 (d, J
= 8.78 Hz, 2H), 7.01 (d, J = 8.78 Hz, 2H), 6.12 (s, 1H), 2.43 (d, J =
4.77 Hz, 3H). MS (m/z) 386.1 (M + H+).
1
(1.05 g, 69% yield): H NMR (400 MHz, DMSO-d6) δ 11.35−11.47
(m, 1H), 8.07−8.15 (m, 1H), 7.84 (d, J = 7.53 Hz, 1H), 7.60−7.69
(m, 1H), 7.50−7.60 (m, 2H), 2.44−2.49 (m, 3H). MS (m/z) 334.0
(M + H+).
A mixture of 3-[(4,6-dichloro-1,3,5-triazin-2-yl)amino]-N-methyl-
benzenesulfonamide (160 mg, 0.48 mmol) in AcOH (4 mL) and H2O
(1 mL) was treated with 3-methylaniline (56 mg, 0.53 mmol) and
NaOAc (47 mg, 0.58 mmol) and stirred at room temperature
overnight. The mixture was diluted with saturated aqueous NaCl, and
the precipitate was collected by filtration and washed twice with H2O.
The filtered solid was dissolved in Et2O and the solution concentrated
to give 3-({4-chloro-6-[(3-methylphenyl)amino]-1,3,5-triazin-2-yl}-
amino)-N-methylbenzenesulfonamide as a white solid (166 mg, 86%
yield): 1H NMR (400 MHz, DMSO-d6) δ 10.53 (br. s., 1H), 10.30 (br.
s., 1H), 8.12 (br. s., 1H), 7.96 (br. s., 1H), 7.78 (br. s., 1H), 7.52−7.65
(m, 1H), 7.48 (d, J = 7.53 Hz, 4H), 7.24 (t, J = 7.78 Hz, 1H), 6.92 (d, J
= 7.28 Hz, 1H), 2.44 (br. s., 3H), 2.29 (br. s., 3H). MS (m/z) 405.0
(M + H+).
N-Methyl-3-[(6-{[4-(methylsulfonyl)phenyl]amino}-4-
pyrimidinyl)amino]benzenesulfonamide Trifluoroacetate (16).
Compound 16 was prepared from 3-[(6-chloro-4-pyrimidinyl)-
amino]-N-methylbenzenesulfonamide and 4-(methylsulfonyl)aniline
1
using Method A as a white solid in 69% yield (143 mg): H NMR
(400 MHz, DMSO-d6) δ 9.85 (s, 1H), 9.72 (s, 1H), 8.45 (s, 1H),
8.09−8.12 (m, 1H), 7.93 (dd, J = 1.76, 8.03 Hz, 1H), 7.80−7.91 (m,
4H), 7.54 (t, J = 7.91 Hz, 1H), 7.45 (q, J = 5.02 Hz, 1H), 7.37 (d, J =
7.78 Hz, 1H), 6.30 (s, 1H), 3.16 (s, 3H), 2.45 (d, J = 5.02 Hz, 3H).
MS (m/z) 434.0 (M + H+).
N-Methyl-3-((2-(m-tolylamino)pyridin-4-yl)amino)-
benzenesulfonamide Trifluoroacetate (17). A mixture of 2,4-
dichloropyridine (50 mg, 0.34 mmol) and 3-amino-N-methylbenze-
nesulfonamide (63 mg, 0.34 mmol) in i-PrOH (2 mL) was treated
with 1 M aqueous HCl (0.34 mL, 0.34 mmol) and stirred at 80 °C for
60 h before being cooled to room temperature. The mixture was
subjected to reverse phase HPLC to give 3-((2-chloropyridin-4-
yl)amino)-N-methylbenzenesulfonamide as a white solid (59 mg, 42%
A mixture of 3-({4-chloro-6-[(3-methylphenyl)amino]-1,3,5-triazin-
2-yl}amino)-N-methylbenzenesulfonamide (166 mg, 0.41 mmol) in
EtOH (1 mL) was treated with ammonium formate (129 mg, 2.1
mmol) and 10% palladium on carbon (44 mg, 0,04 mmol) and stirred
at 80 °C for 1 h before being cooled and filtered through Celite and
washed with EtOH and then DMSO. The filtrate was subjected to
1
yield): H NMR (400 MHz, DMSO-d6) δ 9.47 (s, 1H), 8.08 (d, J =
5.52 Hz, 1H), 7.58−7.63 (m, 1H), 7.44−7.58 (m, 4H), 6.91−6.98 (m,
2H), 2.45 (d, J = 4.77 Hz, 3H). MS (m/z) 298.0 (M + H+).
A mixture of 3-((2-chloropyridin-4-yl)amino)-N-methylbenzenesul-
fonamide (30 mg, 0.07 mmol) and m-toluidine (0.03 mL, 0.28 mmol)
in i-PrOH (2 mL) was treated with 1 M aqueous HCl (0.1 mL, 0.1
mmol) and subjected to microwave irradiation (150 °C) for 60 min
before being cooled to room temperature and subjected to reverse
phase HPLC to give 17 as a white solid (16 mg, 45% yield): 1H NMR
(400 MHz, DMSO-d6) δ 12.40 (br. s., 1H), 10.10 (br. s., 1H), 9.85
(br. s., 1H), 7.77 (d, J = 7.28 Hz, 1H), 7.65−7.72 (m, 1H), 7.52−7.63
(m, 4H), 7.30−7.38 (m, 1H), 7.05−7.15 (m, 3H), 6.59 (dd, J = 2.26,
7.28 Hz, 1H), 6.40 (d, J = 2.26 Hz, 1H), 2.42 (d, J = 4.77 Hz, 3H),
2.34 (s, 3H). MS (m/z) 369.1 (M + H+).
1
reverse phase HPLC to give 19 as a white solid (60 mg, 38%): H
NMR (400 MHz, DMSO-d6) δ 10.07 (br. s., 1H), 9.80 (br. s., 1H),
8.40 (s, 1H), 8.18 (br. s. 1H), 7.50−7.61 (m, 3H), 7.38−7.50 (m, 2H),
7.22 (t, J = 7.91 Hz, 1H), 6.88 (d, J = 7.53 Hz, 1H), 2.44 (d, J = 5.02
Hz, 3H), 2.30 (s, 3H). MS (m/z) 405.0 (M + H+).
3-({4-[(4-Chlorophenyl)amino]-1,3,5-triazin-2-yl}amino)-N-meth-
ylbenzenesulfonamide (20). A mixture of 3-[(4,6-dichloro-1,3,5-
triazin-2-yl)amino]-N-methylbenzenesulfonamide (160 mg, 0.48
mmol) in AcOH (4 mL) and H2O (1 mL) was treated with 4-
chloroaniline (73 mg, 0.58 mmol) and NaOAc (47 mg, 0.58 mmol)
and stirred at room temperature overnight. The mixture was diluted
with saturated aqueous NaCl, and the precipitate was collected by
filtration and washed with water. The filtered solid was dissolved in
Et2O, and the solution was concentrated to give 3-({4-chloro-6-[(4-
chlorophenyl)amino]-1,3,5-triazin-2-yl}amino)-N-methylbenzenesul-
fonamide as a white solid (230 mg, 100% yield). MS (m/z) 425.0 (M
+ H+).
N-Methyl-3-((2-(m-tolylamino)pyrimidin-4-yl)amino)-
benzenesulfonamide Trifluoroacetate (18). A mixture of 2,4-
dichloropyrimidine (160 mg, 1.1 mmol) and 3-amino-N-methylbenze-
nesulfonamide (100 mg, 0.5 mmol) in i-PrOH (4 mL) was treated
with Na2CO3 (171 mg, 1.6 mmol) and stirred at 80 °C for 60 h before
being cooled to room temperature and filtered. The filtrate was
concentrated and subjected to flash column chromatography (50−75%
EtOAc-hexanes) to give 3-((2-chloropyrimidin-4-yl)amino)-N-meth-
A mixture of 3-({4-chloro-6-[(4-chlorophenyl)amino]-1,3,5-triazin-
2-yl}amino)-N-methylbenzenesulfonamide (230 mg, 0.49 mmol) in
EtOH (1.2 mL) was treated with ammonium formate (153 mg, 2.4
mmol) and 10% palladium on carbon (52 mg, 0.05 mmol) and stirred
at 60 °C for 30 min before being cooled to room temperature. The
mixture was diluted with DMSO (15 mL) and filtered through Celite,
and the filtrate was subjected to reverse HPLC to give 20 as a pale
1
ylbenzenesulfonamide as a pale yellow solid (133 mg, 60% yield): H
NMR (400 MHz, DMSO-d6) δ 9.47 10.36 (br. s., 1H), 8.24 (d, J =
5.77 Hz, 1H), 8.07 (s, 1H), 7.85−7.94 (m, 1H), 7.61 (t, J = 8.03 Hz,
1H), 7.47 (d, J = 8.28 Hz, 2H), 6.81 (d, J = 5.77 Hz, 1H), 2.48 (s, 3H).
MS (m/z) 299.0 (M + H+).
1
pink solid (39 mg, 19% yield): H NMR (400 MHz, DMSO-d6) δ
10.13 (br. s., 1H), 10.00 (br. s., 1H), 8.43 (br. s., 1H), 8.08 (br. s., 2H),
7.79 (br. s., 2H), 7.57 (t, J = 7.91 Hz, 1H), 7.41−7.52 (m, 2H), 7.38
(d, J = 8.53 Hz, 1H), 2.44 (d, J = 4.52, 3H). MS (m/z) 391.0 (M +
H+).
A mixture of 3-((2-chloropyrimidin-4-yl)amino)-N-methylbenzene-
sulfonamide (60 mg, 0.2 mmol) and m-toluidine (0.043 mL, 0.4
mmol) was treated with 1 M aqueous HCl (0.2 mL, 0.2 mmol) and
stirred at 80 °C for 60 min before being cooled to room temperature,
concentrated, and subjected to reverse phase HPLC to give 18 as a
3-({6-[(5-Chloro-2-pyridinyl)amino]-4-pyrimidinyl}amino)-N-
methylbenzenesulfonamide Trifluoroacetate (21). Compound 21
was prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methylben-
zenesulfonamide and 5-chloro-2-pyridinamine using Method B as a
1
white solid (70 mg, 71% yield): H NMR (400 MHz, DMSO-d6) δ
10.53 (br. s., 1H), 9.97 (br. s., 1H), 8.26 (d, J = 7.28 Hz, 1H), 8.02 (d,
J = 6.78 Hz, 1H), 7.74 (s, 1H), 7.48−7.58 (m, 3H), 7.31−7.42 (m,
2H), 7.22−7.30 (m, 1H), 6.97 (d, J = 7.28 Hz, 1H), 6.38 (d, J = 6.78
Hz, 1H), 2.43 (d, J = 5.02 Hz, 3H), 2.29 (s, 3H). MS (m/z) 370.1 (M
+ H+).
N-Methyl-3-({4-[(3-methylphenyl)amino]-1,3,5-triazin-2-yl}-
amino)benzenesulfonamide (19). A mixture of 2,4,6-trichloro-1,3,5-
triazine (800 mg, 4.3 mmol) in AcOH (13 mL) at 15 °C was treated
1
pale yellow solid in 6% yield (22 mg): H NMR (400 MHz, DMSO-
d6) δ 10.32 (br. s., 1H), 9.97 (br. s., 1H), 8.44 (s, 1H), 8.31 (d, J = 2.51
Hz, 1H), 8.18 (s, 1H), 7.88−7.94 (m, 1H), 7.85 (dd, J = 2.64, 8.91 Hz,
1H), 7.51−7.59 (m, 2H), 7.46 (q, J = 4.85 Hz, 1H), 7.38 (d, J = 7.78
Hz, 1H), 7.24 (s, 1H), 2.45 (d, J = 5.02 Hz, 3H). MS (m/z) 390.9 (M
+ H+).
I
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
N-Methyl-3-[(6-{[5-(trifluoromethyl)-2-pyridinyl]amino}-4-
pyrimidinyl)amino]benzenesulfonamide Trifluoroacetate (22).
Compound 22 was prepared from 3-[(6-chloro-4-pyrimidinyl)-
amino]-N-methylbenzenesulfonamide and 5-(trifluoromethyl)-2-pyr-
idinamine using Method B as a pale yellow solid in 17% yield (69 mg):
1H NMR (400 MHz, DMSO-d6) δ 10.57 (br. s., 1H), 9.98 (s, 1H),
8.62 (s, 1H), 8.48 (s, 1H), 8.20 (s, 1H), 8.05−8.12 (m, 1H), 7.93 (d, J
= 8.03 Hz, 1H), 7.72 (d, J = 8.78 Hz, 1H), 7.55 (t, J = 8.03 Hz, 1H),
7.46 (q, J = 4.60 Hz, 1H), 7.35−7.42 (m, 2H), 2.46 (d, J = 4.77 Hz,
3H). MS (m/z) 424.9 (M + H+).
°C for 6 h. The mixture was cooled and quenched with aqueous
NH4OH and stirred for 30 min. The precipitate that formed was
collected by filtration, washed with hexanes, and dried to give 29 as a
1
yellow solid (17.5 g, 93% yield): H NMR (400 MHz, DMSO-d6) δ
9.67 (br. s., 1H), 9.54 (br. s., 1H), 8.39 (s, 1H), 8.11 (br. s., 1H), 8.03
(br. s., 1H), 7.93 (d, J = 7.53 Hz, 1H), 7.41−7.58 (m, 3H), 7.34 (d, J =
7.53 Hz, 1H), 7.25 (t, J = 7.91 Hz, 1H), 7.13 (d, J = 7.28 Hz, 1H), 6.25
(s, 1H), 2.45 (d, J = 4.52 Hz, 3H), MS (m/z) 436.0 (M + H+).
3-({6-[(5-Chloro-2-pyridinyl)amino]-4-pyrimidinyl}amino)-N-
methyl-4-(methyloxy)-benzenesulfonamide Trifluoroacetate (37).
Compound 37 was prepared from 3-[(6-chloro-4-pyrimidinyl)-
amino]-N-methyl-4-(methyloxy)benzenesulfonamide and 5-chloro-2-
pyridinamine using Method B as a pale yellow solid in 4% yield (22
mg): 1H NMR (400 MHz, DMSO-d6) δ 10.31 (br. s., 1H), 9.15 (br. s.,
1H), 8.38 (s, 1H), 8.29−8.31 (m, 1H), 8.26−8.28 (m, 1H), 7.85 (dd, J
= 9.03, 2.76 Hz, 1H), 7.49−7.56 (m, 2H), 7.30−7.34 (m, 1H), 7.27−
7.30 (m, 1H), 7.18 (br. s., 1H), 3.93 (s, 3H), 2.43 (d, J = 5.02 Hz, 3H).
MS (m/z) 421.0 (M + H+).
3-({6-[(5-Chloro-2-pyridinyl)amino]-4-pyrimidinyl}amino)-N-
methyl-4-[(trifluoromethyl)oxy]benzenesulfonamide Trifluoroace-
tate (39). Compound 39 was prepared from 3-[(6-chloro-4-
pyrimidinyl)amino]-N-methyl-4-[(trifluoromethyl)oxy]-
benzenesulfonamide and 5-chloro-2-pyridinamine using Method B as a
pale brown solid in 25% yield (123 mg): 1H NMR (400 MHz, DMSO-
d6) δ 10.31 (br. s., 1H), 9.62 (s, 1H), 8.44 (d, J = 2.01 Hz, 1H), 8.37
(s, 1H), 8.31 (d, J = 2.26 Hz, 1H), 7.85 (dd, J = 8.91, 2.64 Hz, 1H),
7.54−7.68 (m, 4H), 7.33 (s, 1 H), 2.49 (d, J = 4.77 Hz, 3H). MS (m/
z) 475.0 (M + H+).
3-({6-[(5-Chloro-2-pyridinyl)amino]-4-pyrimidinyl}amino)-N-
methyl-4-[(2,2,2-trifluoroethyl)oxy]benzenesulfonamide Trifluoroa-
cetate (41). Compound 41 was prepared from 3-[(6-chloro-4-
pyrimidinyl)amino]-N-methyl-4-[(2,2,2-trifluoroethyl)oxy]-
benzenesulfonamide and 5-chloro-2-pyridinamine using Method B as a
pale yellow solid in 12% yield (49 mg): 1H NMR (400 MHz, DMSO-
d6) δ 10.49 (br. s., 1H), 9.47 (br. s., 1H), 8.38 (s, 1H), 8.31 (d, J = 2.51
Hz, 1H), 8.01 (d, J = 2.01 Hz, 1H), 7.87 (dd, J = 8.91, 2.64 Hz, 1H),
7.62 (dd, J = 8.66, 2.13 Hz, 1H), 7.49 (d, J = 8.78 Hz, 1H), 7.42−7.47
(m, 2H), 7.05 (br. s., 1H), 4.92 (q, J = 8.78 Hz, 2H), 2.44 (d, J = 5.02
Hz, 3H). MS (m/z) 489.0 (M + H+).
3-({6-[(5-Chloro-2-pyridinyl)amino]-4-pyrimidinyl}amino)-N-
methyl-4-(methylsulfonyl)benzenesulfonamide (42). Compound 42
was prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methyl-4-
(methylsulfonyl)benzenesulfonamide and 5-chloro-2-pyridinamine
using Method B as an off-white solid in 11% yield (36 mg): 1H
NMR (400 MHz, DMSO-d6) δ 10.26 (s, 1H), 9.08−9.10 (m, 1H),
8.45−8.48 (m, 1H), 8.38−8.40 (m, 1H), 8.31−8.34 (m, 1H), 8.12 (d, J
= 8.28 Hz, 1H), 7.84 (dd, J = 2.76, 9.03 Hz, 1H), 7.80 (q, J = 4.52 Hz,
1H), 7.68 (dd, J = 1.76, 8.28 Hz, 1H), 7.61 (d, J = 9.03 Hz, 1H), 7.42
(s, 1H), 3.33 (s, 3H), 2.50 (d, 3H, obscured by solvent). MS (m/z)
469.0 (M + H+).
3-({6-[(5-chloro-2-pyridinyl)amino]-4-pyrimidinyl}amino)-4-(eth-
ylsulfonyl)-N-methylbenzenesulfonamide p-Toluenesulfonate
Monohydrate (43). A mixture of 3-[(6-chloro-4-pyrimidinyl)amino]-
4-(ethylsulfonyl)-N-methylbenzenesulfonamide (1.5 g, 3.84 mmol), 5-
chloro-2-pyridinamine (0.99 g, 7.68 mmol), Cs2CO3 (3.75 g, 11.5
mmol), BINAP (0.096 g, 0.154 mmol), and Pd(OAc)2 (0.034 g, 0.154
mmol) in 1,4-dioxane (6.4 mL) was treated with a few drops of DMF
(to solubilize the reagents) and subjected to microwave irradiation
(150 °C) for 30 min before being cooled to room temperature and
filtered. This procedure was repeated nine times, and the filtrates from
all reactions were combined and concentrated. The residue was
suspended in CH2Cl2, and the resulting solid was collected by
filtration, triturated in MeOH (3 × 50 mL), triturated in acetone (3 ×
50 mL), and then collected and dried to give 43 as a crystalline free
base (7.8 g, 42% yield). The resulting crystalline free base product was
converted to the p-TsOH salt as this produced an amorphous form
that maximized oral bioavailability for in vivo studies. Thus, 2 g of
crystalline free base was suspended in ∼50 mL of MeCN/MeOH
(2:1) and treated with p-TsOH-H2O (796 mg, 1 equiv), and the
resulting solution was heated at 60 °C for 15 min to give a clear
solution. The solution was concentrated, the residue was dissolved in
N-Methyl-3-[(6-{[3-(1-methylethyl)phenyl]amino}-4-pyrimidinyl)-
amino]benzenesulfonamide (23). Compound 23 was prepared from
3-[(6-chloro-4-pyrimidinyl)amino]-N-methylbenzenesulfonamide and
3-(1-methylethyl)aniline using Method A as an off-white solid in 12%
1
yield (25 mg): H NMR (400 MHz, DMSO-d6) δ 9.49 (br. s., 1H),
9.15 (br. s., 1H), 8.31 (s, 1H), 8.08 (s, 1H), 7.90 (d, J = 7.32 Hz, 1H),
7.46−7.53 (m, 1H), 7.44 (d, J = 7.81 Hz, 1H), 7.39 (q, J = 4.80 Hz,
1H), 7.29−7.34 (m, 2H), 7.22 (t, J = 7.81 Hz, 1H), 6.89 (d, J = 7.57
Hz, 1H), 6.18 (s, 1H), 2.86 (dt, J = 6.93, 13.73 Hz, 1H), 2.44 (d, J =
5.13 Hz, 3H), 1.22 (s, 3H), 1.20 (s, 3H). MS (m/z) 398.0 (M + H+).
3-[(6-{[3-(1,1-Dimethylethyl)phenyl]amino}-4-pyrimidinyl)-
amino]-N-methylbenzenesulfonamide Trifluoroacetate (24). Com-
pound 24 was prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-
methylbenzenesulfonamide and 3-(1,1-dimethylethyl)aniline using
1
Method A as a white solid in 53% yield (148 mg): H NMR (400
MHz, DMSO-d6) δ 9.72 (br. s., 1H), 9.42 (br. s., 1H), 8.36 (s, 1H),
8.03 (s, 1H), 7.86 (d, J = 8.06 Hz, 1H), 7.53 (t, J = 7.93 Hz, 1H),
7.35−7.46 (m, 4H), 7.28 (t, J = 7.81 Hz, 1H), 7.11 (d, J = 7.81 Hz,
1H), 6.16 (s, 1H), 2.44 (d, J = 4.88 Hz, 3H), 1.29 (s, 9H). MS (m/z)
412.2 (M + H+).
3,3′-(4,6-Pyrimidinediyldiimino)bis(N-methylbenzenesulfona-
mide) Trifluoroacetate (25). Compound 25 was prepared from 4,6-
dichloropyrimidine (1 equiv) and 3-amino-N-methylbenzenesulfona-
mide (2 equiv) using Method A as a white solid in 41% yield (81 mg):
1H NMR (400 MHz, DMSO-d6) δ 9.75 (s, 2H), 8.41 (s, 1H), 8.08 (s,
2H), 7.92 (d, J = 7.78 Hz, 2H), 7.54 (t, J = 7.91 Hz, 2H), 7.46 (q, J =
4.85 Hz, 2H), 7.37 (d, J = 7.78 Hz, 2H), 6.24 (s, 1H), 2.45 (d, J = 4.52
Hz, 6H). MS (m/z) 449.1 (M + H+).
N-Methyl-3-({6-[(3-nitrophenyl)amino]-4-pyrimidinyl}amino)-
benzenesulfonamide Trifluoroacetate (26). Compound 26 was
prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methylbenzene-
sulfonamide and 3-nitroaniline using Method A as a yellow solid in
1
40% yield (102 mg): H NMR (400 MHz, DMSO-d6) δ 9.87 (br. s.,
1H), 9.74 (br. s., 1H), 8.71 (br. s., 1H), 8.45 (br. s., 1H), 8.08 (br. s.,
1H), 7.99 (d, J = 7.53 Hz, 1H), 7.93 (d, J = 7.53 Hz, 1H), 7.81 (d, J =
7.78 Hz, 1H), 7.49−7.63 (m, 2H), 7.41−7.49 (m, 1H), 7.36 (d, J =
7.28 Hz, 1H), 6.25 (br. s., 1H), 2.44 (d, J = 2.51 Hz, 3H). MS (m/z)
401.0 (M + H+).
N-Methyl-3-({6-[methyl(3-methylphenyl)amino]-4-pyrimidinyl}-
amino)benzenesulfonamide (27). Compound 27 was prepared from
3-[(6-chloro-4-pyrimidinyl)amino]-N-methylbenzenesulfonamide and
methyl(3-methylphenyl)amine using Method A as a pale brown solid
in 48% yield (60 mg): 1H NMR (400 MHz, DMSO-d6) δ 9.58 (br. s.,
1H), 8.38 (s, 1H), 8.07 (br. s., 1H), 7.77 (d, J = 7.28 Hz, 1H), 7.36−
7.54 (m, 3H), 7.33 (d, J = 7.28 Hz, 1H), 7.10−7.24 (m, 3H), 5.74 (s,
1H), 3.41 (s, 3H), 2.42 (d, J = 4.77 Hz, 3H), 2.37 (s, 3H). MS (m/z)
384.1 (M + H+).
N-Methyl-3-(methyl{6-[(3-methylphenyl)amino]-4-pyrimidinyl}-
amino)benzenesulfonamide (28). Compound 28 was prepared from
3-[(6-chloro-4-pyrimidinyl)(methyl)amino]-N-methylbenzenesulfona-
mide and 3-methylaniline using Method A as a white solid in 23%
1
yield (37 mg): H NMR (400 MHz, DMSO-d6) δ 9.46 (br. s., 1H),
8.34 (s, 1H), 7.61−7.81 (m, 4H), 7.55 (q, J = 4.77 Hz, 1H), 7.12−7.34
(m, 3H), 6.85 (d, J = 7.03 Hz, 1H), 5.79 (s, 1H), 3.44 (s, 3H), 2.44 (d,
J = 5.02 Hz, 3H), 2.26 (s, 3H). MS (m/z) 384.1 (M + H+).
3-({6-[(3-Bromophenyl)amino]-4-pyrimidinyl}amino)-N-methyl-
benzenesulfonamide (29). A mixture of 3-[(6-chloro-4-pyrimidinyl)-
amino]-N-methylbenzenesulfonamide (15 g, 50 mmol) and 3-
bromoaniline (7.8 g, 43 mmol) in isoamylalcohol (10 mL) was
treated with HCl (3 mL of a 2 M solution, 6 mmol) and stirred at 131
J
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
CH2Cl2, and the resulting solution was concentrated to give the
amorphous salt. This process was repeated in multiple 1−2 g batches
for the remaining material, and the batches were combined and mixed
by mortar and pestle to give the p-TsOH salt of 43 as a pale yellow
conjugated pi groups was enforced. The hydrogen bonding constraints
to the hinge residues described above were used to guide the
automatic selection of poses. Postdocking minimization was performed
with strain correction terms applied. Glide XP scores and visual
inspection were used to select reported poses for 3, 37, and 43 bound
to TNNI3K and 37 and 43 bound to B-raf.
1
solid (8.16 g, 77%): H NMR (500 MHz, DMSO-d6) δ 10.67 (br. s.,
1H, C6-aniline-NH), 9.58 (br. s., 1H, C4-aniline-NH), 8.47 (s, 1H,
pyrimidine C2−H), 8.29−8.38 (m, 2H, pyridine C6−H, benzene-
sulfonamide C2−H), 8.14 (d, J = 8.31 Hz, 1H, benzenesulfonamide
C5−H), 7.91 (dd, J = 2.32, 8.93 Hz, 1H, pyridine C4−H), 7.77−7.85
(m, 2H, sulfonamide-NH, benzenesulfonamide C6−H), 7.45−7.53
(m, 3H, pyridine C3−H, p-TsOH−C2-H/C6-H), 7.23 (br. s., 1H,
pyrimidine C5−H), 7.12 (d, J = 7.83 Hz, 2H. p-TsOH−C3-H/C5-H),
3.41 (q, J = 7.34 Hz, 2H, SO2CH2CH3), 2.49−2.51 (m, 3H,
SO2NHCH3), 2.30 (s, 3H, p-TsOH-CH3), 1.11 (t, J = 7.34 Hz, 3H,
SO2CH2CH3). 13C NMR (500 MHz, DMSO-d6) δ 155.5 (pyrimidine-
C2), 152.73 (pyridine-C2), 146.60 (p-TsOH-C1), 146.24 (pyridine-
C6), 145.89 (benzenesulfonamide-C3), 139.00 (benzenesulfonamide-
C1), 138.60 (pyridine-C4), 138.45 (p-TsOH-C4), 132.70 (benzene-
sulfonamide-C5), 128.92 (p-TsOH-C3/C5), 126.36 (p-TsOH-C2/
C6), 125.5 (benzenesulfonamide-C2), 124.67 (pyridine-C5), 124.0
(benzenesulfonamide-C6), 115.49 (pyridine-C3), 92.0 (pyrimidine-
C5), 49.77 (SO2CH2CH3), 29.57 (SO2NHCH3), 21.66 (p-TsOH-
CH3), 7.74 (SO2CH2CH3). 1H and 13C NMR resonances were
assigned based on gCOSY45, gHMQC, gHMBC, and 13C gaspe
spectra. Anal. Calcd (C18H19ClN6O4S2−C7H8O3S−H2O): C, 44.61;
H, 4.34; N, 12.48. Found: C, 44.75; H, 4.06; N. 12.35. HRMS
(C18H19ClN6O4S2) [M + H]+ found m/z 483.0676, calcd 483.0676.
3-[6-(5-Chloro-pyridin-2-ylamino)-pyrimidin-4-ylamino]-N-meth-
yl-4-(propane-2-sulfonyl)-benzenesulfonamide (44). Compound 44
was prepared from 3-[(6-chloro-4-pyrimidinyl)amino]-N-methyl-4-
[(1-methylethyl)sulfonyl]benzenesulfonamide and 5-chloro-2-pyridin-
Given that compound 3 is relatively different from the compound
crystallized to EGFR and those crystallized to TNNI3K and B-raf, the
resultant Glide XP poses of 3 bound to EGFR were further optimized
with Prime MM-GBSA using the VSGB solvation model and OPLS3
force field. Ligand poses from Glide XP and receptor residues within 5
Å of these poses were minimized to relieve steric clashes. Prime MM-
GBSA scores and visual inspection were used to select the final pose of
3 bound to EGFR.
All structure figures were generated with PyMOL version 1.8.6.0
(Schrodinger, LLC, New York, 2017).
̈
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.8b00125.
Kinase selectivity results for 3, 33, and 43, table of
statistics for X-ray crystal structures and tabulated data
used to generate Figure 4 (PDF)
Coordinates for docking models of compound 3 (PDB,
PDB), compound 37 (PDB, PDB), and compound 43
(PDB, PDB)
SMILES representation of compounds with key data
(CSV)
1
amine using Method B as a white solid in 40% yield (49 mg): H
NMR (400 MHz, DMSO-d6) δ 10.49 (br. s., 1H), 9.36 (br. s., 1H),
8.41 (s, 1H), 8.39 (d, J = 1.54 Hz, 1H), 8.30 (d, J = 2.65 Hz, 1H), 8.06
(d, J = 8.38 Hz, 1H), 7.84 (dd, J = 2.65, 8.82 Hz, 1H), 7.78 (q, 1H),
7.69 (dd, J = 1.65, 8.27 Hz, 1H), 7.52 (d, J = 8.82 Hz, 1H), 7.28 (s,
1H), 3.45−3.54 (m, 1H), 2.47 (d, 3H, obscured by solvent), 1.14 (d, J
= 6.62 Hz, 6H). MS (m/z) 496.9 (M + H+).
4-(tert-Butylsulfonyl)-3-(6-(5-chloropyridin-2-ylamino)pyrimidin-
4-ylamino)-N-methylbenzenesulfonamide Trifluoroacetate (45).
Compound 45 was prepared from 3-[(6-chloro-4-pyrimidinyl)-
amino]-4-[(1,1-dimethylethyl)sulfonyl]-N-methylbenzenesulfonamide
and 5-chloro-2-pyridinamine using Method B as a pale yellow solid in
50% yield (61 mg): 1H NMR (400 MHz, DMSO-d6) δ 10.27 (s, 1H),
9.28 (s, 1H), 8.61 (d, J = 1.54 Hz, 1H), 8.40 (s, 1H), 8.32 (d, J = 2.43
Hz, 1H), 7.98 (d, J = 8.38 Hz, 1H), 7.76−7.83 (m, 2H), 7.55−7.60
(m, 2H), 7.40 (s, 1H), 2.47 (d, 3H, obscured by solvent), 1.22 (s, 9H).
MS (m/z) 511.2 (M + H+).
Accession Codes
PDB code: 6B5J (TNNI3K-4).
AUTHOR INFORMATION
Corresponding Author
*E-mail: brian.2.lawhorn@gsk.com. Phone: 484-923-3513.
■
ORCID
Brian G. Lawhorn: 0000-0003-4618-5314
Notes
The authors declare the following competing financial
interest(s): Authors affiliated with GlaxoSmithKline have
received compensation in the form of salary and stock.
ACKNOWLEDGMENTS
■
Computational Modeling. The crystal structure of 4 bound to
TNNI3K (PDB 6B5J) reported in this article was used to model the
binding of compounds 3, 37, and 43 to TNNI3K. The crystal structure
PDB 4I23 was used to model the binding of 3 to EGFR, and the
crystal structure PDB 4YHT was used to model the binding of 37 and
43 to B-raf. All computational modeling was performed using Maestro
We gratefully acknowledge Steve Zhao and Claire Jarvis for
synthesis of 9, 31, and 32 and Nathan Gaul, Evanson Agyeman,
Peter Caprioli, Michael Shaber, Joi Brown, Brian Dombroski,
Mehul Patel, and Laurie Carson for conducting the TNNI3K
and B-Raf assays.
from Schrodinger Release 2017−1 (Schrodinger, LLC, New York,
̈
̈
2017). Protein structures were prepared using the Protein Preparation
Wizard using default parameters. Protons were added, water
orientations were sampled, and hydrogen bond assignments were
optimized. Final hydrogen orientations were optimized by performing
a restrained minimization with heavy atoms kept fixed.
Receptor grids to be used as inputs for Glide were generated around
the crystallized ligands in the respective protein structures. Hydrogen
bond constraints were defined for the backbone carbonyl and NH of
the hinge residue in TNNI3K (Ile542), EGFR (Met793), and B-raf
(Cys531). The crystallographic water molecule that is observed to
make hydrogen bond interactions with the sulfonyl oxygen of the
crystallized ligand and the protein in the back pocket of TNNI3K and
B-raf was included in the receptor grids for those targets.
ABBREVIATIONS USED
■
TNNI3K, troponin I-interacting kinase; EGFR, epidermal
growth factor receptor; DNAUC, dose-normalized area under
the curve; Cl, iv clearance; F, oral bioavailability; Vdss, volume
of distribution; μwave, microwave; TPSA, topological polar
surface area; Fu, fraction unbound as percent; Xantphos, 4,5-
Bis(diphenylphosphino)-9,9-dimethylxanthene; dba, dibenzyli-
deneacetone; BINAP, 2,2′-Bis(diphenylphosphino)-1,1′-bi-
naphthyl
REFERENCES
■
Glide XP was used to dock the ligands into the respective structures
using default parameters with the exception that enhanced planarity of
(1) Zhao, Y.; Meng, X. M.; Wei, Y. J.; Zhao, X. W.; Liu, D. Q.; Cao,
H. Q.; Liew, C. C.; Ding, J. F. Cloning and characterization of a novel
K
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
cardiac-specific kinase that interacts specifically with cardiac troponin I.
J. Mol. Med. (Heidelberg, Ger.) 2003, 81, 297−304.
enhance immunoregulatory functions of dendritic cells. Proc. Natl.
Acad. Sci. U. S. A. 2014, 111, 12468−12473. (b) Sundberg, T. B.;
Liang, Y.; Wu, H.; Choi, H. G.; Kim, N. D.; Sim, T.; Johannessen, L.;
Petrone, A.; Khor, B.; Graham, D. B.; Latorre, I. J.; Phillips, A. J.;
Schreiber, S. L.; Perez, J.; Shamji, A. F.; Gray, N. S.; Xavier, R. J.
Development of chemical probes for investigation of salt-inducible
kinase function in vivo. ACS Chem. Biol. 2016, 11, 2105−2111.
(12) Lawhorn, B. G.; Philp, J.; Zhao, S.; Louer, C.; Hammond, M.;
Cheung, M.; Fries, H.; Graves, A. P.; Shewchuk, L.; Wang, L.; Cottom,
J. E.; Qi, H.; Zhao, H.; Totoritis, R.; Zhang, G.; Schwartz, B.; Li, H.;
Sweitzer, S.; Holt, D. A.; Gatto, G. J., Jr.; Kallander, L. S. Identification
of purines and 7-deazapurines as potent and selective type I inhibitors
of troponin I-interacting kinase (TNNI3K). J. Med. Chem. 2015, 58,
7431−7448.
(13) Compound IC50 values were determined in a panel of internal
kinase assays at GSK at a highest tested concentration of 10 μM. The
composition of the kinase panel was fluid over the course of the
project so that compounds were evaluated against an overlapping, but
nonidentical number and set of kinases. This data provides a rough
approximation of general kinome selectivity for a given compound. For
early compounds, a 10-fold selectivity margin was chosen to allow an
equitable comparison of general kinase selectivity for weakly active
compounds (IC50 ≈ 1 μM) where no greater than 10-fold selectivity
could be assessed given the 10 μM highest test concentration. Selected
compounds (3, 33, and 43) were evaluated in a broader kinase panel
and show favorable selectivity using a more stringent 100-fold
selectivity measure (see Supporting Information).
(2) (a) Lal, H.; Ahmad, F.; Parikh, S.; Force, T. Troponin I-
interacting protein kinase: a novel cardiac-specific kinase, emerging as
a molecular target for the treatment of cardiac disease. Circ. J. 2014, 78,
1514−1519. (b) Milano, A.; Lodder, E. M.; Bezzina, C. R. TNNI3K in
cardiovascular disease and prospects for therapy. J. Mol. Cell. Cardiol.
2015, 82, 167−173.
(3) (a) Wheeler, F. C.; Tang, H.; Marks, O. A.; Hadnott, T. N.; Chu,
P. L.; Mao, L.; Rockman, H. A.; Marchuk, D. A. Tnni3k modifies
disease progression in murine models of cardiomyopathy. PLoS Genet.
2009, 5, e1000647. (b) Tang, H.; Xiao, K.; Mao, L.; Rockman, H. A.;
Marchuk, D. A. Overexpression of TNNI3K, a cardiac-specific
MAPKKK, promotes cardiac dysfunction. J. Mol. Cell. Cardiol. 2013,
54, 101−111. (c) Abraham, D. M.; Marchuk, D. A. Inhibition of the
cardiomyocyte-specific troponin I-interacting kinase limits oxidative
stress, injury, and adverse remodeling due to ischemic heart disease.
Circ. Res. 2014, 114, 938−940.
(4) Vagnozzi, R. J.; Gatto, G. J., Jr.; Kallander, L. S.; Hoffman, N. E.;
Mallilankaraman, K.; Ballard, V. L. T.; Lawhorn, B. G.; Stoy, P.; Philp,
J.; Graves, A. P.; Naito, Y.; Lepore, J. J.; Gao, E.; Madesh, M.; Force, T.
Inhibition of the cardiomyocyte-specific TNNI3K limits oxidative
stress, injury, and adverse remodeling in the ischemic heart. Sci. Transl.
Med. 2013, 5, 207ra141.
(5) Mudd, J. O.; Kass, D. A. Tackling heart failure in the twenty-first
century. Nature 2008, 451, 919−928.
(6) Kaye, D. M.; Krum, H. Drug discovery for heart failure: a new era
or the end of the pipeline? Nat. Rev. Drug Discovery 2007, 6, 127−139.
(7) Harris, P. A.; Boloor, A.; Cheung, M.; Kumar, R.; Crosby, R. M.;
Davis-Ward, R. G.; Epperly, A. H.; Hinkle, K. W.; Hunter, R. N., III;
Johnson, J. H.; Knick, V. B.; Laudeman, C. P.; Luttrell, D. K.; Mook, R.
A.; Nolte, R. T.; Rudolph, S. K.; Szewczyk, J. R.; Truesdale, A. T.; Veal,
J. M.; Wang, L.; Stafford, J. A. Discovery of 5-[[4-[(2,3-dimethyl-2H-
indazol-6-yl)methylamino]-2-pyrimidinyl]amino]-2-methyl-benzene-
sulfonamide (pazopanib), a novel and potent vascular endothelial
growth factor receptor inhibitor. J. Med. Chem. 2008, 51, 4632−4640.
(8) (a) Roberts, W. G.; Ung, E.; Whalen, P.; Cooper, B.; Hulford, C.;
Autry, C.; Richter, D.; Emerson, E.; Lin, J.; Kath, J.; Coleman, K.; Yao,
L.; Martinez-Alsina, L.; Lorenzen, M.; Berliner, M.; Luzzio, M.; Patel,
N.; Schmitt, E.; LaGreca, S.; Jani, J.; Wessel, M.; Marr, E.; Griffor, M.;
Vajdos, F. Antitumor activity and pharmacology of a selective focal
adhesion kinase inhibitor, PF-562,271. Cancer Res. 2008, 68, 1935−
1944. (b) Parsons, J. T.; Slack-Davis, J.; Tilghman, R.; Roberts, W. G.
Focal adhesion kinase: targeting adhesion signaling pathways for
therapeutic intervention. Clin. Cancer Res. 2008, 14, 627−232.
(9) (a) Beattie, J. F.; Breault, G. A.; Ellston, R. P. A.; Green, S.;
Jewsbury, P. J.; Midgley, C. J.; Naven, R. T.; Minshull, C. A.; Pauptit,
R. A.; Tucker, J. A.; Pease, J. E. Cyclin-dependent kinase 4 inhibitors as
a treatment for cancer. Part 1: identification and optimization of
substituted 4,6-bis anilino pyrimidines. Bioorg. Med. Chem. Lett. 2003,
13, 2955−2960. (b) Breault, G. A.; Ellston, R. P. A.; Green, S.; James,
S. R.; Jewsbury, P. J.; Midgley, C. J.; Pauptit, R. A.; Minshull, C. A.;
Tucker, J. A.; Pease, J. E. Cyclin-dependent kinase 4 inhibitors as a
treatment for cancer. Part 2: identification and optimization of
substituted 2,4-bis anilino pyrimidines. Bioorg. Med. Chem. Lett. 2003,
13, 2961−2966.
(10) (a) Zhang, Q.; Liu, Y.; Gao, F.; Ding, Q.; Cho, C.; Hur, W.; Jin,
Y.; Uno, T.; Joazeiro, C. A. P.; Gray, N. Discovery of EGFR selective
4,6-disubstituted pyrimidines from a combinatorial kinase-directed
heterocycle library. J. Am. Chem. Soc. 2006, 128, 2182−2183.
(b) Deng, X.; Okram, B.; Ding, Q.; Zhang, J.; Choi, Y.; Adrian, F.
J.; Wojciechowski, A.; Zhang, G.; Che, J.; Bursulaya, B.; Cowan-Jacob,
S. W.; Rummel, G.; Sim, T.; Gray, N. S. Expanding the diversity of
allosteric Bcr-Abl inhibitors. J. Med. Chem. 2010, 53, 6934−6946.
(11) (a) Sundberg, T. B.; Choi, H. W.; Song, J.-H.; Russell, C. N.;
Hussain, M. M.; Graham, D. B.; Khor, B.; Gagnon, J.; O’Connell, D. J.;
Narayan, K.; Dancik, V.; Perez, J. R.; Reinecker, H.-C.; Gray, N. S.;
Schreiber, S. L.; Xavier, R. J.; Shamji, A. F. Small-molecule screening
identifies inhibition of salt-inducible kinses as therapeutic strategy to
(14) (a) Tari, L. W.; Hoffman, I. D.; Bensen, D. C.; Hunter, M. J.;
Nix, J.; Nelson, K. J.; McRee, D. E.; Swanson, R. V. Structural basis for
the inhibition of Aurora A kinase by a novel class of high affinity
disubstituted pyrimidine inhibitors. Bioorg. Med. Chem. Lett. 2007, 17,
688−691. (b) Zhang, G.; Ren, P.; Gray, N. S.; Sim, T.; Liu, Y.; Wang,
X.; Che, J.; Tian, S.-S.; Sandberg, M. L.; Spalding, T. A.; Romeo, R.;
Iskandar, M.; Chow, D.; Seidel, H. M.; Karanewsky, D. S.; He, Y.
Discovery of pyrimidine benzimidazoles as Lck inhibitors: Part I.
Bioorg. Med. Chem. Lett. 2008, 18, 5618−5621. (c) Ito, A.; Ota, K.;
Yoshizawa, K.; Tanaka, K.; Yamabe, T. Ab initio MO study of diphenyl
derivatives: diphenylamine, diphenylaminium, diphenylaminyl and
diphenylnitroxide. Chem. Phys. Lett. 1994, 223, 27−34.
(15) (a) Manley, P. J.; Balitza, A. E.; Bilodeau, M. T.; Coll, K. E.;
Hartman, G. D.; McFall, R. C.; Rickert, K. W.; Rodman, L. D.;
Thomas, K. A. 2,4-disubstituted pyrimidines: A novel class of KDR
kinase inhibitors. Bioorg. Med. Chem. Lett. 2003, 13, 1673−1677.
(b) Bardelle, C.; Cross, D.; Davenport, S.; Kettle, J. G.; Ko, E. J.;
Leach, A. D.; Mortlock, A.; Read, J.; Roberts, N. J.; Robins, P.;
Williams, E. J. Inhibitors of the tyrosine kinase EphB4. Part 1:
structure-based design and optimization of a series of 2,4-bis-
anilinopyrimidines. Bioorg. Med. Chem. Lett. 2008, 18, 2776−2780.
(16) Force, T.; Krause, D. S.; Van Etten, R. A. Molecular mechanisms
of cardiotoxicity of tyrosine kinase inhibition. Nat. Rev. Cancer 2007, 7,
332−344.
(17) Lawhorn, B. G.; Philp, J.; Graves, A. P.; Holt, D. A.; Gatto, G. J.,
Jr.; Kallander, L. S. Substituent effects on drug-receptor H-bond
interactions: correlations useful for the design of kinase inhibitors. J.
Med. Chem. 2016, 59, 10629−10641.
(18) (a) Ertl, P.; Rohde, B.; Selzer, P. Fast calculation of molecular
polar surface area as a sum of fragment-based contributions and its
application to the prediction of drug transport properties. J. Med.
Chem. 2000, 43, 3714−3717. (b) Palm, K.; Stenberg, P.; Luthman, K.;
Artursson, P. Polar molecular surface properties predict the intestinal
absorption of drugs in humans. Pharm. Res. 1997, 14, 568−571.
(c) Kelder, J.; Grootenhuis, D. J.; Bayada, D. M.; Delbressine, L. P. C.;
Ploemen, J.-P. Polar molecular surface as a dominating determinant for
oral absorption and brain penetration of drugs. Pharm. Res. 1999, 16,
1514−1519.
(19) (a) Harris, L. S.; Zhang, S.; Treskov, I.; Kovacs, A.; Weinheimer,
C.; Muslin, A. J. Raf-1 kinase is required for cardiac hypertrophy and
cardiomyocyte survival in response to pressure overload. Circulation
2004, 110, 718−723. (b) Muslin, A. J. Role of Raf proteins in cardiac
L
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
hypertrophy and cardiomyocyte survival. Trends Cardiovasc. Med.
2005, 15, 225−229.
(20) Lawhorn, B. G.; Philp, J.; Graves, A. P.; Shewchuk, L.; Holt, D.
A.; Gatto, G. J., Jr.; Kallander, L. S. GSK114: a selective inhibitor for
identifying the biological role of TNNI3K. Bioorg. Med. Chem. Lett.
2016, 26, 3355−3358.
(21) The calculated TPSA (143 Å2) of the 4-sulfone analogs exceeds
the limit required for oral bioavailability within the series. However,
these analogs were expected to retain oral bioavailability (as observed)
based on their presumed internal H-bond with the neighboring aniline
NH, which is unaccounted for in the TPSA calculation, leading to an
overestimated value of PSA.
(22) (a) Hammond, M.; Kallander, L. S.; Lawhorn, B. G.; Philp, J.;
Sarpong, M. A.; Seefeld, M. A. Azolopyrimidine Compounds as Kinase
Inhibitors and Their Preparations and Use in the Treatment of
Diseases and Methods. PCT Int. Appl. WO2011149827, 2011.
(b) Kallander, L. S.; Lawhorn, B. G.; Philp, J.; Zhao, Y. 3-[(6-
Aminopyrimidin-4-yl)amino]benzenesulfonamide Compounds Useful
as TNNI3K Inhibitors, Their Preparation, and Methods of Use in the
Treatment of Congestive Heart Failure. PCT Int. Appl.
WO2011088027, 2011. (c) Kallander, L. S.; Philp, J. Compounds
for Treating Congestive Heart Failure. PCT Int. Appl.
WO2011088031, 2011.
(23) Hartung, C. G.; Backes, A. C.; Felber, B.; Missio, A.; Philipp, A.
Efficient microwave-assisted synthesis of highly functionalized
pyrimidine derivatives. Tetrahedron 2006, 62, 10055−10064.
M
DOI: 10.1021/acs.jmedchem.8b00125
J. Med. Chem. XXXX, XXX, XXX−XXX