2-Aminopyridine-Based Mitogen-Activated Protein Kinase Kinase Kinase Kinase 4 (MAP4K4) Inhibitors: Assessment of Mechanism-Based Safety

Article abstract of DOI:10.1021/acs.jmedchem.8b00152

Studies have linked the serine-threonine kinase MAP4K4 to the regulation of a number of biological processes and/or diseases, including diabetes, cancer, inflammation, and angiogenesis. With a majority of the members of our lead series (e.g., 1) suffering from time-dependent inhibition (TDI) of CYP3A4, we sought design avenues that would eliminate this risk. One such approach arose from the observation that carboxylic acid-based intermediates employed in our discovery efforts retained high MAP4K4 inhibitory potency and were devoid of the TDI risk. The medicinal chemistry effort that led to the discovery of this central nervous system-impaired inhibitor together with its preclinical safety profile is described.

Full text of DOI:10.1021/acs.jmedchem.8b00152

Subscriber access provided by - Access paid by the | UCSB Libraries
2-Aminopyridine-Based Mitogen-Activated Protein Kinase Kinase Kinase
Kinase 4 (MAP4K4) Inhibitors: Assessment of Mechanism-Based Safety
Robert L. Dow, Mark Ammirati, Scott W. Bagley, Samit K. Bhattacharya, Leonard
Buckbinder, Christian Cortes, Ayman F El-Kattan, Kristen Ford, Gary B. Freeman,
Cristiano R. W. Guimaraes, Shenping Liu, Mark Niosi, Athanasia Skoura, and David Tess
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00152 • Publication Date (Web): 23 Mar 2018
Downloaded from http://pubs.acs.org on March 24, 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 29
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2-Aminopyridine-Based Mitogen-Activated Protein Kinase Kinase Kinase Kinase 4
(MAP4K4) Inhibitors: Assessment of Mechanism-Based Safety
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
Robert L. Dow,* † Mark Ammirati, ‡ Scott W. Bagley, ‡ Samit K. Bhattacharya,† Leonard Buckbinder,† Christian Cortes,†
Ayman F. ElꢀKattan,† Kristen Ford,‡ Gary B. Freeman,† Cristiano R. W. Guimarães,† Shenping Liu,‡ Mark Niosi,‡ Athanasia
Skoura,† David Tess†
†Pfizer Worldwide Research & Development, Cambridge, Massachusetts 02139, United States
‡Pfizer Worldwide Research & Development, Groton, Connecticut 06340, United States
ABSTRACT: Studies have linked the serineꢀthreonine kinase MAP4K4 to the regulation of a number of biological processꢀ
es/diseases including diabetes, cancer, inflammation and angiogenesis. With a majority of the members of our lead series
(e.g. 1) suffering from timeꢀdependent inhibition of CYP3A4, we sought design avenues that would eliminate this risk. One
such approach arose from the observation that carboxylic acidꢀbased intermediates employed in our discovery efforts reꢀ
tained potent MAP4K4 inhibitory potency and were devoid of the TDI risk. The medicinal chemistry effort that led to the
discovery of this CNSꢀimpaired inhibitor together with its preclinical safety profile is described.
ꢀ
INTRODUCTION
Mitogenꢀactivated protein kinase kinase kinase kinase 4 (MAP4K4/HGK/ZCl) is a serineꢀthreonine protein kinase1 that has been
shown to be involved in the regulation of a number of biological processes including inflammation,2 cancer,3,4 diabetes,5 and angioꢀ
genesis.6 MAP4K4 has been shown to regulate a number of cellular processes including cell transformation, adhesion,3 and motiliꢀ
ty.6,7 An impediment to understanding the signaling pathways involved in these biological effects is the lack of well characterized
targets for MAP4K4. To date, NHE1,8 Arp2,9 ERM proteins10 and FARP111 have been identified as phosphorylation substrates for
this serineꢀthreonine kinase. Selective inhibitors of MAP4K4 have served an important role in the initial steps of deconvoluting the
signaling pathways of this kinase.12ꢀ14
Our group recently reported on a series of potent MAP4K4 inhibitors that arose from a virtual screening effort utilizing an inꢀ
house crystal structure of the kinase domain.13 Optimized analogs contain a 2ꢀaminopyridine core which establishes two points of
contact with the hinge region of MAP4K4. As part of their binding to MAP4K4, these compounds also make a πꢀπ interaction with
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 29
Tyr36, which is likely a driver in forming the unusual, folded Pꢀloop structure of this kinase.15 This feature is likely a contributor to
the high level of kinome selectivity observed for this series of inhibitors.
1
2
3
4
5
6
7
8
9
In analogy to the in vivo response observed following treatment with a MAP4K4ꢀsiRNA,16 inhibitor 1 reversed the inflammatory
response induced by endotoxin (LPS) in a rodent model.13 In a geneticallyꢀobese mouse model of insulin resistance, 1 performed
comparably to the insulin sensitizer rosiglitazone in lowering blood glucose levels. While 1 possesses a number of positive attribꢀ
utes, analysis of its safety profile reveals several liabilities, including a positive response in the AMES genotoxicity assay and
pharmacological promiscuity as demonstrated against a panel of nonꢀkinase targets. Timeꢀdependent inhibition (TDI) of CYP3A4
also led to drug accumulation following multiꢀday treatments with 1, resulting in doseꢀlimiting toxicities. Attempts to reduce the
propensity for TDI through closeꢀin structural modification approaches did not solve this problem.13 Broader profiling of structural
changes to the distal aminopyridine functionality, through incorporation of a range of neutral substituents to pyridyl/phenylꢀbased
tail groups also failed to solve the TDI issue.17 It is assumed that these modifications simply shifted the site of reactive species
formation to the amine functionality of the aminopyridine core involved in hinge binding interactions with MAP4K4.
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
Replacement of the distal aminopyridine ring of 1 with an aminopyrimidine moiety (2) suppressed timeꢀdependent inhibition of
CYP3A4, eliminating both the genotoxicity and TDIꢀmediated drug accumulation risks. This modification also provided an imꢀ
proved offꢀtarget pharmacology profile.13 Based on its in vitro safety and in vivo efficacy profiles, compound 2 was advanced to a
rodent in vivo safety evaluation. Cardiovascular assessment was prioritized given the high association of this endpoint toxicity with
the kinase inhibitor class.18 Oral dosing in telemetered rats for five days with 2 (25 mg/kg, bid) resulted in a number of adverse
effects. Substantial weight loss (7%) and body temperature increases (0.4oC) relative to controls were observed over the course of
the study. For cardiovascular endpoints, maximal heart rate increases of 25 bpm, relative to controls, were induced by treatment
with 2. Given that these effects were observed at free drug concentrations close to those required for inhibition of cellular kinase
activity and in vivo inflammatory response,13 further development of this compound was halted.
These safety findings also raised the question as to whether these toxicities were driven directly through inhibition of MAP4K4,
the inherent chemotype or due to some offꢀtarget liability(s) that had not been identified in safety pharmacology profiling. To date,
there have been no other reports of in vivo safety evaluation of a MAP4K4 inhibitor, making it difficult to place our own results in
context regarding this question. Thus identification of an inhibitor which possessed an orthogonal offꢀtarget and/or exposure proꢀ
file that could be advanced to an in vivo safety analysis provided the focus for the efforts disclosed here.
ꢀ
RESULTS AND DISCUSSION
With earlier efforts demonstrating that a wide range of structural modifications of either 1 or 2 led to compounds which retained
significant liabilities, most notably TDI, alternative design approaches were required. Most of this subsequent work focused on
ACS Paragon Plus Environment

Page 3 of 29
Journal of Medicinal Chemistry
incorporation of polar functionality at various positions on the lead structure in an attempt to reduce CYPꢀmediated activation/TDI.
An example of this was a set of amide analogs based on 3 (Table 1).17 It was hoped that incorporation of a modestly polar/electronꢀ
withdrawing group in this terminal ring would impact the TDI potential. However, neither 3 nor a set analogs varying the nitrogen
substituents of this lead provided a balanced potency/reduced TDI profile. In retrospect, the TDI associated with 3 is not too surꢀ
prising given that it still retains significant turnover in human liver microsomes (Clint,app = 22 uL/min/mg). Towards a goal of furꢀ
ther reductions in oxidative metabolic liability, carboxylic acid intermediates 4 and 5 were evaluated. Both of these acids possessed
potent MAP4K4 enzyme and modest cellular activity. The ability of this vector to tolerate an acid functionality is consistent with
previous SAR on polar substituents in this position and our knowledge from MAP4K4 crystal structures.13,15 More importantly,
these acids were devoid of a TDI liability, as measured by the change in midazolam metabolism by CYP3A4, when treated with test
compound in a dose responsive manner up to 60 ꢀM. One potential explanation for the reduction in CYP inhibition by these comꢀ
pounds could be ascribed to the electronꢀwithdrawing effect on the presumed reactive aminopyridine core by the carboxylic acid.
The relatively low TDI associated with Nꢀacetic acid analog 619 suggests that this is likely not the key driver and the reduced TDI
risk can be simply ascribed to low metabolic turnover for these carboxylic acids.20
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
Table 1. Enzyme, cell potencies, and timeꢀdependent inhibition of CYP3A4 of analogs of compound 1.
R
Enzyme IC50
Cell IC50
TDI (%)b
Permeability21
(nM)a
(nM)a
(x 10ꢀ6 cm/sec)
1
2
3
4
5
3.7
160
61
24
59
3
9
15
4.1
21
470
1182
3148
273
26
12
16
19
2.4
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 29
6
2.4
935
26
3
1
2
3
4
5
6
7
8
9
7
8
9
83
7.6
2.1
1121
1019
2018
ꢀꢀ
6
ꢀꢀ
18
5
ꢀꢀ
H
N
MeO2
S
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
10
11
5.6
4.0
1590
5388
5
4
7
57
aGeometric mean of ≥3 independent measurements run in triplicate Change in midazolam metabolism by CYP3A4 after
b
incubation with compound @ 60 uM @ 30 minutes.
In general, a right shift in cellular potency is observed for all members of this series, since they are ATPꢀcompetitive and the enzyꢀ
matic assay is run at Km ATP levels (10ꢀM) vs cellꢀbased concentrations which are in the millimolar range. This was confirmed
through a similar rightꢀshift in potency when the kinase assay was run in the presence of 1 mM ATP. Compound 5 inhibited
MAP4K4 in a cellꢀbased assay with an IC50 of 273 nM, which was in line with the potency in the enzyme assay run in the presence
of 1 mM ATP (IC50 = 93 nM). This acid is highly permeable (Papp = 19 x 10ꢀ6 cm/sec)22 and demonstrated no evidence of efflux
based on PgP ratios, suggesting these properties are not drivers in the rightꢀshifted cellular potency.
In order to make certain that incorporation of a carboxylic acid did not adversely affect the offꢀtarget profile of this series, 5 was
screened against kinome and pharmacology panels. Screening against twenty kinases at 10 ꢀM revealed that 5 inhibited only two
nonꢀMAP4K4 family kinases, Aurora and ABL, to a substantial extent (Supporting Information).23 Followꢀup screening in cellꢀ
based systems containing these two kinases revealed that 5 had an IC50 > 30 ꢀM against each. A clean profile (IC50 > 10ꢀM) was
also observed against a panel of ~20 common secondary pharmacologic targets (Supporting Information). In summary, the finding
that carboxylic acid substituents are compatible with potent, selective MAP4K4 inhibition provided a forward path for this chemoꢀ
type, ultimately leading to the identification of an orthogonal profile which could be advanced to a safety evaluation.
Given that the carboxylic acid had been incorporated at a site on the core structure that was expected to be near the solꢀ
vent/protein interface when bound to MAP4K4, it raised the question of whether this group was interacting with the protein and
contributing to the observed potency. To gain insight into this question unsubstituted congener 7 was prepared. Removal of the
acid led to ~30ꢀfold reduction in enzyme potency, suggesting the acid functionality has a direct interaction with MAP4K4. Since
ACS Paragon Plus Environment

Page 5 of 29
Journal of Medicinal Chemistry
attempts to generate a crystal structure of 5 bound to MAP4K4 were not successful it is not clear what interaction(s) might be ocꢀ
curring between the acid and protein. Moving the carboxylic acid into the 3ꢀposition (8) led to only a slight reduction in potency.
1
2
3
4
5
6
7
8
9
A limited set of carboxylic acid bioisosteres were explored (9ꢀ11) and found to be potent inhibitors of the MAP4K4 kinase doꢀ
main. A significant rightꢀshift in cellular activity was observed for all of them relative to the 5. This may in part have been driven
by the reduced passive permeability of these isosteric analogs compared to 5 (Papp = 2, 6 and 4 x 10ꢀ6 cm/s for 9ꢀ11, respectively).
Further efforts on bioisosteric replacements were not pursued based on this profile, as well as their relatively high polar surface
areas (93ꢀ110 Å2), since these properties would limit optimization efforts with regards to incorporation of additional polar substituꢀ
ents (e.g. modification of the chlorophenyl ring). Interestingly, tetrazole 11 also carries significant TDI risk even though it has low
turnover in human microsomes (Clint,app < 8 uL/min/mg).
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
Utilizing libraryꢀenabled chemistry the structureꢀactivityꢀrelationships for the pꢀchlorophenyl substituent were developed. Repreꢀ
sentative examples selected from the ~120 analogs prepared are shown in Table 2. The in vitro profiles of the desꢀchloro (12) and
mꢀchloro (13) analogs were slightly rightꢀshifted relative to 5. Ortho substitution on this ring with substituents possessing a range
of electronic/steric properties was not well tolerated (e.g. 14). This is consistent with MAP4K4 crystal structures bound to this
class of inhibitors which reveal an extended tunnelꢀlike region in the protein that surrounds the phenyl group with tight tolerances
for the ortho vectors.13 Consistent with these earlier SAR/structural studies, the 4ꢀposition of this series tolerated a wide range of
substituents (e.g. 15ꢀ19). MAP4K4 residues lining this pocket are hydrophobic in nature, suggesting that heteroaromatic replaceꢀ
ments for the chlorophenyl substituent would not be well tolerated. This was borne out by the finding that pyrimidine 20 was >50ꢀ
fold less active relative to 19. From a small set of heteroaromatic replacements, Nꢀmethylpyrazole 21 was identified as a potent
inhibitor of the catalytic domain of MAP4K4, but was inactive in the cell assay. Given concerns around the physical properties
associated with the triaryl series, replacements for the chlorophenyl ring were explored. Incorporation of a phenoxymethyl group
in the 3ꢀposition of the aminopyridine (22) led to ~10ꢀfold reduction in enzyme/cellular MAP4K4 inhibitory potency. Incorporaꢀ
tion of the regioisomeric ether linker (23) led to increased inhibitory potency relative to 22. Cyclobutylmethyl ether 24 had comꢀ
parable potency relative to the corresponding benzyl ether. This isosteric replacement of an aryl with a cyclobutyl ring is consistent
with the binding pocket in this region being formed by aliphaticꢀbased amino acid residues in MAP4K4.
Of the analogs in Table 2 tested for timeꢀdependent inhibition of CYP3A4, only ether 23 demonstrated significant activity at 60
ꢀM (41%). The highly electronꢀrich nature of this particular aminopyridine is likely the driving force behind the formation of reacꢀ
tive intermediates. While ethers 23 and 24 demonstrated good in vitro inhibitory profiles, further efforts around these
alkoxymethylpyridine cores were not pursued given their associated TDI risk.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Table 2. Enzyme and cell potencies of aryl analogs of 4.
Page 6 of 29
1
2
3
4
5
6
7
8
R
Enzyme IC50
Cell IC50
Permeability21
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
(nM)a
(nM)a
(x 10ꢀ6 cm/sec)
12
13
14
12
1692
17
11
62
1308
16
ꢀꢀ
>10000
15
16
17
18
19
20
21
7.6
3.5
7.2
6.0
3.5
238
13
1019
248
17
13
12
15
13
ꢀꢀ
3391
491
377
>10,000
>10,000
5
22
23
24
43
11
11
>8500
4006
1852
13
17
18
a Geometric mean of ≥3 independent measurements run in triplicate.
With these early studies suggesting that the benzoic acid moiety played a significant role in both MAP4K4 inhibitory activity and
suppression of TDI, our attention turned to development of the carboxylic acid SAR. Having failed to successfully obtain coꢀ
crystals of 5 or related analogs with the catalytic domain of MAP4K4, an empirical approach was employed. Homologation to
ACS Paragon Plus Environment

Page 7 of 29
Journal of Medicinal Chemistry
acetic acid 25 maintained MAP4K4 enzymatic and cellular potency (Table 3). While this compound had low projected human
clearance based on human liver microsomes (CLint,app,scaled < 8 mL/min/kg), analysis in rat liver hepatocytes suggested moderate to
high clearance (CLint,app = 62 uL/min/million). Intravenous dosing in rats confirmed this prediction (CL = 8 mL/min/kg, 1 mg/kg)
and suggested that further in vivo analyses in efficacy/safety models might be hampered by this profile, especially given the low
free drug fraction in this species (fu = 0.004). Regioisomeric analog 26 also potently inhibits MAP4K4, however, kinome profiling
suggested that it had reduced selectivity relative to the 25 (Supplementary Information). Incorporating the dimethyl substituents of
25 into carboꢀ/heterocycles (27ꢀ29) retained potent inhibitory activity. Unlike in the case of the benzoic acid leads, conversion of
the carboxylic acid in 27 to the corresponding tetrazole (29) did not lead to a significant reduction in cellular potency. Besides a
low predicted permeability (Papp = 1 x 10ꢀ6 cm/sec), 29 was also found to potently inhibit phosphodiesterase 5A1(IC50 = 27 nM).
Tying the benzylic position of the acetic acid pharmacophore back as part of a indane ring (30/31) also resulted in potent inhibition,
with 30 being 10ꢀfold more potent than its enantiomer. While 30 had an improved profile vs PDE 5A1 activity (IC50 = 3.4 uM), it
had moderate turnover in both human and rat hepatocytes (CLint,app = 12 and 67 uL/min/million, respectively).
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
In contrast to earlier carboxylic acidꢀbased analogs, a 2.5Å structure of 27 bound to the catalytic domain of MAP4K4 was sucꢀ
cessfully generated (Figure 1). In analogy to other crystal structures for this series, the aminopyridine makes acceptorꢀdonor interꢀ
actions with the hinge region of the enzyme.13 The structure also reveals that the acid is interacting with the backbone NꢀH of
Gly32, which is at the start of the Pꢀloop structure.15 This interaction assists in stabilizing this protein feature that forms the hydroꢀ
phobic tunnel accommodating this series of inhibitors. The carboxylic acid is sitting at the interface of the opening of this pocket
and solventꢀexposed space (Figure 2). Replacement of the carboxylic acid with a nitrile (32) results in a 300ꢀfold reduction in
MAP4K4 inhibitory potency, which is consistent with the nitrile being unable to make a productive interaction with the Pꢀloop reꢀ
gion of the enzyme.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 29
Figure 1. Crystal structure (2.5 Å resolution) of the complex between the catalytic domain of MAP4K4 and compound 27 (in yelꢀ
1
low) (PDB code: 5J95).
2
3
4
5
6
7
8
9
Based on the interactions identified from the crystal structure of 27 bound to MAP4K4, lengthening of the linker between the
phenyl ring and the carboxylic acid did not appear at first to be a fruitful avenue for pursuit. To assess the validity of this concluꢀ
sion, propionic acid derivative 33 was prepared and screened for MAP4K4 inhibitory activity. This homologated acid retains subꢀ
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
stantial enzymatic and cellular potency, suggesting the carboxyl acid in this case may be picking up productive interactions with the
protein further outside of the hydrophobic tunnel than Glyꢀ25.
Replacement of the propionic acid sidechain with the isosteric oxyacetic acid (34) led to a 4ꢀfold improvement in MAP4K4 enꢀ
zyme inhibitory activity, but a 10ꢀfold reduction in cellular activity relative to 33. This substantial shift in cellular activity is likely
not explained by differences in passive permeability (Papp = 7 and 12 x 10ꢀ6 cm/sec, respectively). To further address this question,
a series of analogs were prepared in which additional lipophilicity was introduced in the αꢀposition. Methyl enantiomers 35 and 36
retain potent inhibitory activity, with 35 showing a modest improvement in cellular potency compared to 34. Incorporation of a
dimethyloxyacetic acid tail (37) led to subꢀnanomolar enzymatic potency, as well as a further improvement in cell activity. This
modification also favorably impacted turnover in human and rat hepatocytes CLint,app < 6 and < 25 uL/min/million, respectively) in
addition to further separation from the offꢀtarget PDE5A1 activity (IC50 = 1.9 uM). To determine whether the carboxylic acid in 37
was a key driver of MAP4K4 potency, isopropyl ether 38 was prepared. In line with the earlier work around the benzoic acid anaꢀ
logs, removal of the carboxylate group led to ~3,000ꢀfold loss in enzymatic potency and ~25ꢀfold rightꢀshift in cellular potency.
The moderate permeability of 37 (Papp = 10 x 10ꢀ6 cm/sec) may be a contributing factor in the differential shift in cellular potency
relative to enzymatic potency for 38. An attempt to increase cell permeability through incorporation of a cyclopentyl ring (39) acꢀ
tually had the opposite effect (Papp = 3 x 10ꢀ6 cm/sec) and in fact resulted in an isolated enzyme to whole cell ratio consistent with
that seen for 37. These results further underscore the conclusion that multiple drivers are in play for this translation between these
in vitro activities. Additional closeꢀin structural modifications of 37, such as mꢀisomer 40 or incorporation of a pyridine ring (41),
in analogy to 1, did not provide an improved profile. Finally, extension of the linker length between the phenoxy ring and carboxꢀ
ylic acid (42) resulted in modest enzymatic/cellular potency.
ACS Paragon Plus Environment

Page 9 of 29
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Figure 2. Rendering of protein surface area of the catalytic domain of MAP4K4 bound to compound 27 (in yellow) (PDB code:
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
5J95).
Attempts to coꢀcrystallize 37 with the catalytic domain of MAP4K4 were not successful. Modeling of this ligand into the crystal
structure obtained with 27 suggests that it is unlikely the acid of 37 is interacting with the backbone NꢀH of Gly32.
Table 3. Enzyme and cell potencies of benzoic acid modifications of 5.
N
NH2
R
Cl
R
Enzyme IC50
Cell IC50
Permeability21
(nM)a
(nM)a
(x 10ꢀ6 cm/sec)
25
3.5
231
11
26
27
1.4
4.6
407
148
10
3
28
2
101
10
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 29
29
2.9
331
1
1
2
3
4
5
6
7
8
9
30b
31b
32
1.4
15
103
1052
ꢀꢀ
13
12
ꢀꢀ
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
1426
33
8.1
602
ꢀꢀ
34
2.0
1.0
5574
842
7
5
5
8
ꢀꢀ
3
35b
36b
37
7.9
8728
222
0.4
38
1421
0.6
4699
467
39
40
41
2.0
1.3
564
197
10
9
42
6.5
470
4
a Geometric mean of ≥3 independent measurements run in triplicate. bAbsolute stereochemistry arbitrarily assigned.
Based on its overall profile, oxyacetic acid analog 37 was selected for further evaluation. A summary of key in
vitro/ADME/safety properties of this acid relative to the aminopyridineꢀbased starting point 1 is detailed in Table 4. Incorꢀ
poration of the carboxylic acid resulted in an improved turnover in human microsomal/hepatocytes, with no detectable meꢀ
tabolism observed for 37 in either of these systems. As for the other carboxylic acidꢀbased analogs in this series, the risk of
TDI was also eliminated, which in turn is likely responsible for the differential mutagenic risk between these two comꢀ
pounds, as evidenced by the AMES results. Screening (10 ꢀM) against a panel of 58 offꢀtarget pharmacologies showed a
ACS Paragon Plus Environment

Page 11 of 29
Journal of Medicinal Chemistry
very clean profile, with only 5ꢀhydroxytryptamine receptor 1B and ACE showing >50% effect. Both of these targets were
evaluated in functional assays and 37 showed no significant activity for either at concentrations up to 30 ꢀM (Supporting
Information). Based on the selective kinome profile of related members of this series, 37 was screened in human peripheral
blood cell lysates for inhibitory activity against a panel of 232 kinases (ActivX screen). Other than MAP4K4/MINK/TNIK,
only STE kinome branch member germinal center kinase (GCK/MAP4K2) and the lipid kinase phosphatidylinositolꢀ5ꢀ
phosphate 4ꢀkinase, type II, gamma (PIP4K2C) were inhibited >70% by 37 (1 ꢀM). Comparable potencies for the closeꢀin
family members MINK and TNIK were confirmed by enzymatic screening at Km levels of ATP, with IC50 = 1.3 and <0.6
nM, respectively. Profiling against a set of 244 kinases screened in the presence of 1 mM ATP and 1 ꢀM 37 revealed less
than 50% inhibition by all these targets (Supporting Information). With an IC50 of 6 nM against MAP4K4 under physiologiꢀ
cal ATP concentrations, these results suggest that compound 37 possesses a high degree of kinome selectivity.
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
Table 4. MAP4K4/Physicochemical/ADME profiles for 37/1
Properties
37
1
MAP4K4 kinase domain IC50
(nM) @ [ATP] = 10 ꢀMa
MAP4K4 kinase domain IC50
(nM) @ [ATP] = 1 mMa
Cell IC50 (nM)a
0.4
3.7
6
134
222
383/1.2/85
<8
160
296/3.3/78
42
MW/eLogD/TPSAb
HLM Clapp,scaled (mL/min/kg)
hHepatocyte Clint,app
<4
31
(ꢀL/min/million)
Solubility, pH 1.2, mg/mL
1.4c
13d
0.14c
5.8d
pH 6.5, mg/mL
Permeability (x 10ꢀ6 cm/sec)
10
3
e
CYP inhibition IC50
>3 ꢀM
>3 ꢀM
TDI, %Inhibition (Dose)f
ꢀ4% (30ꢀM) 54%
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 29
(30ꢀM)
1
2
3
4
5
6
3% (60ꢀM) 61%
(60ꢀM)
7
8
9
g
h
hERG IC50
~100 ꢀM
>250 ꢀM
Negative
1.8 ꢀM
THLE IC50
>119 ꢀM
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
AMESi
Positive
aGeometric mean of ≥3 independent measurements run in triplicate; beLogD = measured logD at pH 7.4; cmonohydrate solvate ddiꢀ
hydrochloride salt; e<15% Inhibition at 3uM for isoforms (3A4, 2D6, 1A2, 2C8 and 2C9) fChange in midazolam metabolism after
incubation with compound @ 30 minutes23 gMeasurement of potassium current in human embryonic kidney 293 (HEK293) cells
stably transfected with the human etherꢀaꢀgoꢀgoꢀrelated gene (hERG) channel hTransformed human liver epithelial cell assay,
(THLE)24 iActivity in a bacterial mutagenesis assay.25
Compound 37 was advanced to pharmacokinetic profiling in anticipation of efficacy/safety studies. Consistent with their
respective microsomal turnover values, intravenous dosing in rat, dog and nonꢀhuman primate confirmed low clearance.
Moderate to long halfꢀlives were observed in all species, along with low volumes of distribution (Table 5). Predicted human
clearance based on either human microsomal/hepatocyte data or allometric scaling suggested low human clearance.
Table 5. Pharmacokinetic profile of 37.
Rata,b
0.76
Doga,c
1.5
NHPa,d
0.90
Clearance
(mL/min/kg)
T1/2 (hr)
3.7
2.0
10
Volume (L/kg)
0.25
0.42
0.52
aDosed at 1 mg/kg, i.v.; bFed male Wistar Han (n=2); cFed male beagle (n=2); dFasted male cynomolgus (n=2).
In preparation for in vivo safety evaluation of 37 a PK/PD analysis was carried out in the same model system utilized to evaluate
MAP4K4ꢀsiRNA inflammatory efficacy.2 Mice were dosed orally with inhibitor followed by an injection of lipopolysaccharides
(LPS) after 30 minutes. Plasma changes in TNFα were measured at 1.5 h and the corresponding free drug levels determined (Figꢀ
ACS Paragon Plus Environment

Page 13 of 29
Journal of Medicinal Chemistry
ure 3). While TNFα levels were elevated at the lower drug doses in multiple experiments, it was assumed that it was not a drugꢀ
related effect given that the free concentrations at these doses were below the cellular IC50. Only the top two dose groups exhibited
substantial suppression of TNFα release upon stimulation with LPS, which is consistent with their corresponding free exposures
being greater than the MAP4K4 cellular potency of 37.
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 3. Doseꢀresponse effects of 37 on LPSꢀinduced TNFα levels in male C57ꢀBL/6J mice (p.o., 0.5% methylcellulose vehicle).
*p=0.0002, **p=0.0001. Average free concentrations of 37 are shown above each dose bar in the graph.
In considering additional MAP4K4 chemotypes to advance to safety studies, carboxylic acids afforded the potential to design in a
peripherallyꢀrestricted exposure profile. Exclusion from the brain would aid our understanding of whether safety signals, such as
the reduced food intake and body temperature, observed for 2 were centrallyꢀmediated.26 To assess whether 37 had the potential to
be excluded from the brain it was screened for efflux potential in MDCKꢀMDR1(NIH)27 and BCRP28 containing cell lines. In both
of these assay systems 37 exhibited a profile suggesting high efflux potential, with MDR1ꢀ and BCRPꢀbased BA/AB ratios of 3
and 116, respectively. These results were shown to be predictive, with an average free drug brain/plasma ratio of 0.14 over a 5ꢀhr
test period after an oral dose of 10 mg/kg of 37 in rats. The overall profile of this brainꢀimpaired inhibitor led us to advance it to
general toxicology and cardiovascular safety profiling.
No significant changes in clinical signs were observed following a single oral dose of 500 mg/kg of 37 in rats. The average free
drug concentration over this 24 hr period was 143 nM. Achieving high free exposure levels in rodent species was particularly chalꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 29
lenging given the extremely high protein binding (rat fuplasma = 0.0005). In this single dose rat study the Cmax,free was just 250 nM.
The relatively low peak concentrations for this compound help to minimize the chance of offꢀtarget effects skewing interpretation
1
2
3
4
5
6
7
8
9
of these safety evaluations. Based on its clean profile in this singleꢀdose study, 37 was advanced to a dose escalation study in dogs.
Exposure was limited in this study due to emesis occurring at all doses, with Cavg,free = 50, 85 and 293 nM at 20, 100, 500 mg/kg,
respectively. The only clinical sign observed in any of the treatment groups was mucoid feces and transient tremors in one of the
two dogs receiving 500 mg/kg. To determine whether 37 was differentiated relative to the negative cardiovascular effects observed
for 2, a slingꢀtrained dog cardiovascular evaluation was performed. No significant changes were observed during a stepwise escaꢀ
lating i.v. infusion of 37 for either blood pressure or heart rate at free drug concentrations up to 40ꢀfold the Ceff determined in the
LPS assay. Based on this encouraging set of singleꢀdose data, 37 was advanced to a 2ꢀweek rat safety study (20, 100 and 500
mg/kg/day, p.o., b.i.d. dosing). Average Cavg,free exposures at these doses on the first day of dosing were 15, 67 and 166 nM. Conꢀ
sistent with 37 showing no significant activity in the in vitro TDI assay, interꢀanimal day 1and 14 exposures did not vary substanꢀ
tially. While no effect on body weight or food intake were observed at the two lower doses, at the highest dose there was a reducꢀ
tion of body weight gain (0.66x) and food intake (0.8x) relative to control animals. By day 4, animals in the high dose group startꢀ
ed to exhibit a number of signs of moribundity and there was a death of one animal, leading to termination of the study. The effects
observed in this group mimicked very closely what was observed in a fiveꢀday evaluation carried out in rats with 2. As in this preꢀ
vious case, profound toxicity was observed at exposures close to efficacious concentrations.
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
The significant toxicology signals associated with 2 and 37 are suggestive of a MAP4K4ꢀdriven safety liability. This is supported
by their high MAP4K4ꢀspecificity, differential exposure profiles and the relatively low concentrations leading to these safety findꢀ
ings. It is important to note that these inhibitors also inhibit the other kinase family members, MINK and TNIK, which could also
be contributing to the observed effects. Addressing this issue will likely require development of an allosteric MAP4K4 inhibitor.
Such a compound would have the benefit of high kinome specificity, including MAP4K4ꢀfamily selectivity, and the potential to
achieve efficacy at much lower concentrations given its nonꢀATP competitive nature.
ꢀ
SYNTHESIS
Synthesis of the inhibitors described and characterized herein relied mainly upon a SuzukiꢀMiyaura cross coupling reaction to form
the bond between the aminopyridine core and the acidic tail aromatic ring.29 Representative synthetic approaches utilized in this
study can be found in Schemes 1ꢀ3. Reaction of aminopyridineꢀbased boronate 4313 with bromoester 44 under standard Suzukiꢀ
Miyaura coupling conditions provided ester 45. Cleavage of the tꢀbutyl ester with acid afforded acid 3.
____________________________________________________
ACS Paragon Plus Environment

Page 15 of 29
Scheme 1a
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
aReagents and conditions: (i) Pd(dppf)Cl2, Na2CO3, 1,4ꢀdioxane, H2O, 85 oC; (ii) HCl, EtOH, CH2Cl2, RT.
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
____________________________________________________
An alternative approach to the preparation of compounds possessing variations of the acidꢀcontaining aromatic tails is to reverse
the reactivity of the two partners in the SuzukiꢀMiyaura reaction. Scheme 2 details this approach utilized for the synthesis of 37.
Bromide 46 was converted to the corresponding boronate ester 47. Coupling of this boronate with bromoaminopyridine 48 under
SuzukiꢀMiyaura conditions provided ester 49. Base hydrolysis afforded the target oxyacetic acidꢀbased derivative 37.
____________________________________________________
Scheme 2a
aReagents and conditions: (i) 4,4,4’,4’,5,5,5’,5’ꢀoctamethylꢀ2,2’ꢀbiꢀ(1,3,2ꢀdioxaborolane), Pd(dppf)Cl2, KOAc, 1,4ꢀdioxane, 80 oC;
(ii) Pd(AmPhos)2Cl2, Na2CO3, EtOH, H2O, 60 oC; (iii) NaOH, H20, EtOH, reflux.
____________________________________________________
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 29
An approach for rapid development of structureꢀactivityꢀrelationships in the 3ꢀposition of the aminopyridine head group was develꢀ
oped in which a late stage 3ꢀbromo intermediate was utilized (Scheme 3). This intermediate was generated via a twoꢀstep sequence
involving a SuzukiꢀMiyaura coupling of boronic acid 50 and aminobromopyridine 51 to afford biaryl 52. Subsequent bromination
of the intermediate with Nꢀbromosuccinimide provided 53 in 81% overall yield for this sequence. Coupling of this bromide with
phenyboronic acid afforded ester 54. Acid catalyzed cleavage of the tꢀbutyl ester afforded acid 12 as the TFA salt.
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
____________________________________________________
Scheme 3a
aReagents and conditions: (i) Pd(dppf)Cl2, Na2CO3, 1,4ꢀdioxane, H20, 80 oC; (ii) Nꢀbromosuccinimide, CH2Cl2, RT; (iii) phenylꢀ
boronic, Pd(dppf)Cl2 acid, K2CO3, 1,4ꢀdioxane, H20, 80 oC; (iv) trifluoroacetic acid, CH2Cl2, 30 oC.
____________________________________________________
ꢀ
CONCLUSION
Incorporation of an oxyacetic acid moiety into a lead series of MAP4K4 inhibitors afforded high target potency and selectivity.
This modification also provided a significant degree of CNSꢀimpairment, which was viewed as a positive attribute for targeting
disease endpoints such as diabetes and other inflammationꢀdriven indications. While this profile was distinct when compared to
early members of this series, a similar preclinical safety profile was still observed. These results are suggestive that inhibition of
MAP4K4 and/or its closelyꢀrelated family members being responsible for eliciting these safety signals. Alternative approaches,
such as the identification of nonꢀATP competitive inhibitors, would be of value in assessing this hypothesis.
ꢀ
EXPERIMENTAL SECTION
ACS Paragon Plus Environment

Page 17 of 29
Journal of Medicinal Chemistry
All procedures performed on these animals were in accordance with regulations and established guidelines and were reꢀ
viewed and approved by Pfizer Institutional Animal Care and Use Committee.
1
2
3
4
5
6
7
8
9
MAP4K4 Z’ꢀLyte Assay. The in vitro kinase assay utilized human recombinant protein, GSTꢀtagged MAP4K4 catalytic
domain, residues 1ꢀ328 (Invitrogen #PV4013) and the Z’ꢀLyte Ser/Thr 7 FRET peptide substrate (Invitrogen #PV3208)
which is labeled with fluorescein and coumarin. This FRET assay is based on the differential sensitivity of phosphorylated
and nonꢀphosphorylated peptides to proteolytic cleavage using methods outlined by the manufacturer (Invitrogen). The assay
was optimized for a MAP4K4 concentration that phosphorylated 20ꢀ40% of the Z’ꢀLyte Ser/Thr 7 peptide at Km ATP (10
µM, Invitrogen #PV3227) in a 90 min reaction at room temperature (RT; 1.25 nM). The 10 ꢀL reactions were carried out in
384ꢀwell black Optiplate (Perkin Elmer #6007279) in assay buffer (50 mM HEPES pH 7.5, 1 mM EGTA, 10 mM MgCl2,
and 0.01% Brijꢀ35; Invitrogen #PV3189), containing 0.5% compound/DMSO, 2 µM Z’ꢀLyte Ser/Thr peptide, 1.25 nM
MAP4K4, and either 10 µM ATP (Km) or 1 mM ATP (“physiological levels”). The phosphorylation range was determined
using 0% and 100% phosphorylation controls. ATP is omitted in 0% phosphorylation controls. The 100% phosphorylation
controls used a phosphorylated standard of the Ser/Thr peptide (2 µM; Invitrogen # PV3223) in place of enzyme, peptide and
ATP. Enzyme and compound were incubated for 30 min at RT before ATP addition. Reactions were carried out for 25 min
in the presence of 1 mM ATP, or for 90 min in the presence of 10 µM (Km) ATP at RT (optimized to achieve 20%ꢀ40%
phosphorylation). At 60 min, 5 µL of Development Reagent A diluted in Development Buffer was added to each well
(1:65,536; Invitrogen #PV3295, PV3127) and incubated for 60 min at RT. The reaction is stopped with 5 µL Stop Buffer
(Invitrogen #PV3094), shaken for 30 sec and read immediately on a PerkinꢀElmer Envision in FRET mode (excitation filter,
Photometric 405 (barcode 302), emission filter 1, FITC 535 (barcode 206), emission filter 2, Umbelliferone 460 (barcode
207)). Percent inhibition was calculated based on sample wells containing either DMSO only (0% inhibition) or a standard
compound (100% inhibition). Compound concentrationꢀresponses were fit to a 4ꢀparameter equation to determine compound
IC50 values.
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
MAP4K4ꢀtraf2 Cell ELISA.
The assay was designed to identify compounds inhibiting MAP4K4’s phosphorylation of serine/threonine residues on traf2.
GripTite™ 293 MSR Cells (Life Technologies #R795ꢀ07 ) were maintained in complete media (DMEM #11965ꢀ092 with
10% FBS, 10 mM HEPES (#15630ꢀ080), 1x MEM NEAA (#11140ꢀ050), 600 ꢀg/mL Geneticin (#10131ꢀ027) and 1,000
U/mL Penicillin/Streptomycin (#15140ꢀ122, Life Technologies). For culturing and plating, cells were washed with cationꢀ
free DPBS (Lonza #17ꢀ512Q) and incubated 5 min at room temperature (RT) with 37°C Versene (Life Technologies
#15040ꢀ066). After decanting, 5 mL 0.25% TrypsinꢀEDTA (Life Technologies #25200ꢀ056) was added and the flask gently
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 29
rocked to loosen cells. Cells were passaged 1:15 in complete media. An inꢀwell transient transfection was performed. Cells
1
2
3
4
5
6
7
were seeded at 35,000 cells/well in 100 ꢀL complete media in 96ꢀwell polystyrene TCꢀcoated plates (Corning #3598). After
15 min at RT, plates were moved to a 37 °C, 5% CO2, humidified incubator for 6 h. To transfect one plate of cells, 25 ꢀL
FuGENE® HD (Promega #E2311) was added to 500 ꢀL OptiMEM prewarmed to 37 °C (Life Technologies #31985ꢀ070)
8
9
and incubated for 5 min at RT. 8.75 ꢀg of pcDNA3.1ꢀtraf2ꢀFLAG tag (NM_001242559.1) and 3.5 ꢀg of pcDNA3.1ꢀ
MAP4K4ꢀmycꢀHis tag (NM_021138.3) were added to the OptiMem mixture, gently mixed and incubated for 15 min at RT.
The DNAꢀFugene mix was diluted with 2.625 mL OptiMem, and 25 ꢀL was dispensed to each well of plated cells. The
plates were gently swirled before moving back to the 37 °C, 5% CO2, humidified incubator for 40 h. The cells are treated
with compound solubilized in DMSO, diluted in complete media, for a final DMSO concentration of 0.25%. The plates were
incubated for 1.5 hr in a 37°C, 5% CO2, humidified incubator. The plates were then decanted and washed with 150 µL/well
cationꢀfree PBS. Then 150 µL/well HNGT Buffer (50 mM HEPES pH 7.5, 150 mM NaCl, 10% glycerol, 1%TritonꢀX,) with
1x HALT Protease/Phosphatase Inhibitor Cocktail (Pierce #78447) was added. Plates were heat sealed and placed on a rotary
plate shaker for 15 min with vigorous shaking, then centrifuged at 3000 rpm for 5 min and placed in a ꢀ80°C freezer until the
ELISA was performed. ELISA plates were prepared the day before the assay by coating black highꢀbinding 96ꢀwell plates
(Greiner #675077) with 75 µL/well antiꢀFLAG antibody (Sigma #F3165) at 5 µg/mL in 0.5M carbonate/bicarbonate buffer,
pH 9.6 and incubated overnight at 4°C. The next day the antiꢀFLAGꢀcoated plates were washed of primary antibody using an
ELXꢀ405 PlateWasher and wash buffer (1x TBS/0.05% Tweenꢀ20 buffer; 5 x 100 ꢀL/well). Plates were then blocked with
75 µL/well 5% BSA in wash buffer for 1 hr at RT. Blocking buffer was removed and cell lysate added (50 µL/well) for 1 hr
at RT. Plates were washed 5 x 100 ꢀL/well and 50 µL/well of antiꢀPhosphoꢀThreonine antibody (Cell Signaling Technology
#9381s) diluted 1:1,500 in wash buffer was added and incubated for 1 hr at RT. Plates were washed as above and then and
then 50 µL/well Donkey antiꢀRabbit antibody (Jackson #711ꢀ036ꢀ152) diluted 1:4,000 in wash buffer was added to the wells
and incubate 1 hr at RT. After washing as above, 50 µL/well 1x SuperSignal ELISA Pico Substrate (Pierce #37069) was
added. Following a 5ꢀmin incubation, plates were read on a PerkinꢀElmer Envision using 96ꢀwell luminescence protocol.
Where not specified above, chemicals were purchased from Sigma. Percent inhibition was calculated based on sample wells
containing either DMSO only (0% inhibition) or a standard compound (100% inhibition). Compound concentrationꢀ
responses were fit using a 4ꢀparameter equation to calculate IC50 values.
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
Mouse LPS Model. Male mice (C57/Bl6) were obtained from Charles River Laboratories. Mice (n=6ꢀ8 per group) were
administered compound by oral gavage followed by intravenous injection (10,000 U/kg) of LPS (Invivogen, E. coli
O111:B4) 30 minutes later. Blood samples for TNFꢀα analysis were obtained by terminal cardiac bleeding 90 minutes post
ACS Paragon Plus Environment

Page 19 of 29
Journal of Medicinal Chemistry
LPS administration. Mouse plasma TNFꢀα concentrations were determined by ultraꢀsensitive MSD cytokine assay (MSD,
K153BHCꢀ1). The data are presented as the mean ± SE and statistical analysis was performed using GraphPad Prism 5. Oneꢀ
way ANOVA with Bonferroni’s posttests were used for statistical analysis and P <0.05 were considered statistically signifiꢀ
cant. All procedures performed on these animals were in accordance with regulations and established guidelines and were
reviewed and approved by Pfizer Institutional Animal Care and Use Committee.
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
Synthetic Materials and Methods. All chemicals, reagents and solvents were purchased from commercial sources and used
without further purification. All reactions were performed under an atmosphere of nitrogen unless otherwise noted. Proton (1H
NMR) nuclear magnetic spectra were recorded with 400 MHz and 500 MHz Varian spectrometers. Chemical shifts are reported in
parts per million relative to the residual solvent signals (i.e. for CDCl3 δH = 7.27 ppm, for DMSOꢀd6 δH = 2.50 ppm and for
CD3OD δH = 4.78 and 3.31 ppm). The peak shapes are denoted as follows: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet;
br s, broad singlet. Mass spectrometry (MS) was performed via atmospheric pressure chemical ionization (APCI) or electronspray
ionization (ESI) sources. Liquid chromatography mass spectrometry (LCMS) was performed on an Agilent 1100 Series (Waters
Atlantis C18 column, 4.6 x 50 mm, 5 ꢁm; 95% water/acetonitrile linear gradient to 5% water/acetonitrile over 4 min, hold at 5%
water/ 95% acetonitrile to 5.0 min, trifluoroacetic acid modifier (0.05%); flow rate of 2.0 mL/ min) and ultraꢀpressure liquid chroꢀ
matography mass spectrometry (UPLCꢀMS) was performed on a Waters Acquity System (Waters Acquity HSS T3 C18 column,
2.1 x 50 mm, 1.7 ꢁm; 95% water/acetonitrile linear gradient to 5% water/acetonitrile over 1.0 min, hold at 5% water/95% acetoniꢀ
trile to 1.5 min, formic acid modifier (0.1%) or ammonia modifier (0.1%); flow rate of 1.25 mL/min. Silica gel chromatography
was performed using a medium pressure Biotage or ISCO system using columns preꢀpackaged by various commercial vendors inꢀ
cluding Biotage and ISCO. Whatman preꢀcoated silica gel plates (250 ꢁm) were used for analytical TLC. Compound purity was
determined by HPLC, with all compounds determined to be >95% pure. HPLC purity determination methods are described in the
Supporting Information section. The syntheses of compounds 1, 2, 3 and 7 were previously reported.13 Compound 37 (catalog
#PZ0283) is commercially available via Sigma Aldrich.
6'ꢀAminoꢀ5'ꢀ(4ꢀchlorophenyl)ꢀ3,3'ꢀbipyridineꢀ6ꢀcarboxylic acid, hydrochloride salt (3). A 4 M solution of HCl in
EtOAc (0.31 mL, 1.26 mmol) was added to a mixture of 45 (120 mg, 0.31 mmol) in DCM (5 mL). After the reaction mixture
had been stirred at RT for 16 h, it was concentrated, and the resulting solids were triturated with DCM (10 mL) to provide
the title compound (37 mg, 0.11 mmol, 35%) as a white solid. 1H NMR (400 MHz, DMSOꢀd6) δ 9.2 (s, 1H), 8.65 (s, 1H),
8.4–8.3 (d, 1H), 8.35 (s, 1H), 8.2–8.1 (d, 1H), 8.2–7.9 (br s, 3H), 7.65 (br s, 4H); LCMS: [M+1] = 326.1; Purity: HPLC
Method C, 97.1%.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 29
4ꢀ(6ꢀAminoꢀ5ꢀphenylpyridinꢀ3ꢀyl)benzoic acid, TFA salt (12). To a solution of 54 (40 mg, 0.12 mmol) in DCM (2 mL)
was added TFA (0.4 mL), and the reaction mixture was stirred at 30 °C for 1 h. Concentration afforded the title compound
(30 mg, 74 µmol, 62%) as a yellow solid. 1H NMR (400 MHz, DMSOꢀd6) δ 8.45 (d, J = 2.0 Hz, 1H), 8.22 (d, J = 2.0 Hz,
1H), 7.98–8.05 (m, 2H), 7.91 (d, J = 8.5 Hz, 2H), 7.72 (br s, 2H), 7.48–7.61 (m, 5H); LCMS: [M+1] = 290.9; Purity: HPLC
Method A, 98.8%.
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
2ꢀ{4ꢀ[6ꢀAminoꢀ5ꢀ(4ꢀchlorophenyl)pyridinꢀ3ꢀyl]phenoxy}ꢀ2ꢀmethylpropanoic acid (37). A suspension of 49 (30.3 g, 73.7
mmol) in EtOH (180 mL) was treated with 4 M NaOH (55.3 mL), and the reaction mixture was heated at reflux for 45 min.
Upon reaching reflux temperature the solids dissolved, providing a pale yellow solution. The reaction was cooled to 60 °C
and titrated to pH 3.5 with 5 M HCl (approximately 45 mL). A very thick, pale yellow precipitate formed and the mixture
was allowed to stir for 4 h at 60 °C. The solids were collected by filtration and dried overnight in a vacuum oven to deliver
the hydrate of the title compound (26.34 g, 68.80 mmol, 93%) as a crystalline white solid. 1H NMR (400 MHz, DMSOꢀd6) δ
13.07 (br s, 1H), 8.24 (d, J = 2.3 Hz, 1H), 7.45–7.63 (m, 7H), 6.86 (d, J = 8.6 Hz, 2H), 5.75 (s, 2H), 3.32 (br s, 2H water
peak), 1.52 (s, 6H); 13C NMR (101 MHz, DMSOꢀd6) δ 175.0, 155.4, 154.3, 144.8, 136.9, 135.5, 132.0, 131.1, 130.5, 128.8,
126.5, 125.0, 119.1, 118.9, 78.5, 25.1; LCMS: [M+1] = 383.1; Purity: HPLC Method E, 98.1%. Anal. Calcd for
C21H19ClN2O3·H2O; C, 62.92; H, 5.28; Cl, 8.84; N, 6.99. Found C, 62.89; H, 5.26; N, 6.96.
tertꢀButyl 6'ꢀaminoꢀ5'ꢀ(4ꢀchlorophenyl)ꢀ3,3'ꢀbipyridineꢀ6ꢀcarboxylate (45). To a mixture of 3ꢀ(4ꢀchlorophenyl)ꢀ5ꢀ
(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)ꢀ2ꢀpyridinamine (43)13 (150 mg, 0.45 mmol) in 3:1 dioxane/H2O (4 mL) were
added tertꢀbutyl 5ꢀbromopyridineꢀ2ꢀcarboxylate (44) (117 mg, 0.45 mmol), Na2CO3 (96 mg, 0.91 mmol), and Pd(dppf)Cl2
(33 mg, 45 µmol). After the reaction mixture had been stirred at 85 °C for 6 h, it was cooled to RT, and extracted with DCM
(40 mL) {in some of the experiments referencing this procedure, extraction was carried out using EtOAc}; the organic layer
was washed with brine, dried over Na2SO4, filtered, and concentrated. The crude product was purified via silica gel chromaꢀ
tography (gradient: 0% to 80% EtOAc in petroleum ether) to deliver the title compound (140 mg, 0.367 mmol, 81%) as a
yellow solid. 1H NMR (400 MHz, DMSOꢀd6) δ 9.01 (s, 1H), 8.46 (s, 1H), 8.3–8.2 (d, 1H), 8.0–7.9 (d, 1H), 7.78 (s, 1H), 7.55
(br s, 4H), 6.11 (s, 2H), 1.56 (s, 9H).
Ethyl 2ꢀ(4ꢀbromophenoxy)ꢀ2ꢀmethylpropanoate (46). A mixture of 4ꢀbromophenol (200 g, 1.16 mol) and K2CO3 (191
g, 1.39 mol) in ethyl 2ꢀbromoꢀ2ꢀmethylpropanoate (400 mL) was heated at 110 °C for 48 h. The reaction mixture was cooled
to RT and filtered through diatomaceous earth; the filtrate was concentrated to provide the title compound (280 g, 0.975 mol,
84%) as a colorless oil, which was used for the next step without purification.
ACS Paragon Plus Environment

Page 21 of 29
1
2
3
4
5
6
7
Journal of Medicinal Chemistry
Ethyl 2ꢀmethylꢀ2ꢀ[4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)phenoxy]propanoate (47). This reaction was carried
out in two batches of the same scale. A suspension of 46 (100 g, 348 mmol), B2Pin2 (88.5 g, 348 mmol), and KOAc (85.4 g,
871 mmol) in dioxane (1.5 L) was degassed and refilled with N2 several times. Pd(dppf)Cl2 (14.2 g, 17.4 mmol) was added,
and the mixture was degassed and refilled with N2 three times. The reaction mixture was heated to 80 °C for 16 h, and then
cooled to RT. The two batches were combined at this point, then filtered and concentrated. After the residue had been partiꢀ
tioned between EtOAc (3 L) and water (1.5 L), the aqueous layer was extracted with EtOAc (1 L), and the combined organic
layers were washed with water (2 L) and with brine (2 L), dried over Na2SO4, filtered, and concentrated. The crude product
was subjected to chromatography on silica gel (Gradient: 0% to 10% EtOAc in petroleum ether), and the purified material
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
1
was triturated with petroleum ether (600 mL) to deliver the title compound (110 g, 329 mmol, 47%) as a white solid. H
NMR (400 MHz, CDCl3) δ 7.70 (d, J = 8.5 Hz, 2H), 6.81 (d, J = 8.5 Hz, 2H), 4.22 (q, J = 7.0 Hz, 2H), 1.62 (s, 6H), 1.33 (s,
12H), 1.23 (t, J = 7.3 Hz, 3H); LCMS: [M+1] = 335.2.
Ethyl 2ꢀ{4ꢀ[6ꢀaminoꢀ5ꢀ(4ꢀchlorophenyl)pyridinꢀ3ꢀyl]phenoxy}ꢀ2ꢀmethylpropanoate (49). A flask containing 48 (15.0
g, 52.9 mmol), 47 (19.4 g, 58.2 mmol), Na2CO3 (16.8 g, 159 mmol), and Pd(AmPhos)2Cl2 (375 mg, 0.53 mmol) was purged
with N2. Degassed 5:1 EtOH/H2O (180 mL) was added and the reaction mixture was heated at 60 °C for 3 h. It was then diꢀ
luted with EtOH (150 mL), followed by addition of H2O (360 mL), and the resulting mixture was allowed to stir while coolꢀ
ing to RT. After approximately 1 h of stirring, the precipitate was collected by filtration and the solids dried overnight under
vacuum to deliver the title compound (21.42 g, 52.13 mmol, 99%) as a tan solid. 1H NMR (400 MHz, DMSOꢀd6) δ 8.24 (d, J
= 2.7 Hz, 1H), 7.48–7.57 (m, 7H), 6.83 (d, J = 8.6 Hz, 2H), 5.75 (s, 2H), 4.18 (q, J = 7.0 Hz, 2H), 1.53 (s, 6H), 1.19 (t, J =
7.2 Hz, 3H); LCMS: [M+1] = 411.2
tertꢀButyl 4ꢀ(6ꢀaminopyridinꢀ3ꢀyl)benzoate (52). Dioxane (180 mL) and aq 2 M Na2CO3 (86.7 mL, 173 mmol) were
added to a mixture of 2ꢀaminoꢀ5ꢀbromopyridine (51) (10.0 g, 57.8 mmol) and [4ꢀ(tertꢀbutoxycarbonyl)phenyl]boronic acid
(50) (14.1 g, 63.6 mmol). After the solution had been degassed with N2 for 10 min, Pd(dppf)Cl2 (2.11 g, 2.89 mmol) was
added, and the reaction mixture was heated at 80 °C for 2 h, then allowed to cool to RT. Dioxane was removed via concenꢀ
tration, and the aqueous residue was partitioned between water (500 mL) and EtOAc (300 mL). The aqueous layer was exꢀ
tracted with additional EtOAc (2 x 300 mL), and the combined organic layers were dried over Na2SO4, filtered, and concenꢀ
trated to provide a grey solid, which was taken up in water (500 mL), acidified with 1 M HCl (100 mL), and washed with 1:1
EtOAc/heptane (2 x 300 mL). The organic washes were combined and extracted with 0.2 M HCl (250 mL), and all of the
acidic aqueous layers were combined, basified with 8 M aq NaOH, and extracted with EtOAc (3 x 250 mL). These organic
extracts were combined, dried over Na2SO4, and filtered through a pad of silica gel (250 g) in a 2 L fritted funnel, eluting
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 29
with additional EtOAc (2 L). Concentration of the filtrate afforded the title compound (13.87 g, 51.3 mmol, 89%) as a tan
1
2
3
4
solid. 1H NMR (400 MHz, CDCl3) δ 8.13 (d, J = 5.5 Hz, 1H), 8.07 (d, J = 8.6 Hz, 2H), 7.63 (d, J = 8.6 Hz, 2H), 6.91 (dd, J
= 5.3, 1.4 Hz, 1H), 6.74 (s, 1H), 4.74 (br s, 2H), 1.62 (s, 9H); LCMS: [M+1] = 272.1.
5
6
7
8
9
tertꢀButyl 4ꢀ(6ꢀaminoꢀ5ꢀbromopyridinꢀ3ꢀyl)benzoate (53). Freshly recrystallized NBS (8.78 g, 49.3 mmol) was added
portionꢀwise over 15 min to a solution of 52 (13.34 g, 49.35 mmol) in DCM (400 mL). After the reaction mixture had been
stirred for 1 h at RT, it was washed with water (2 x 400 mL) and with 10% aq Na2S2O3 (200 mL). The organic layer was then
dried over Na2SO4, filtered, concentrated, and purified using silica gel chromatography (Gradient: 0% to 50% ethyl acetate in
heptane). The resulting brown solid was triturated with 10% DCM in heptane (200 mL) to afford the title compound (15.74
g, 45.07 mmol, 91%) as a tan solid. 1H NMR (400 MHz, DMSOꢀd6) δ 8.11 (s, 1H), 7.98 (d, J = 8.6 Hz, 2H), 7.51 (d, J = 8.6
Hz, 2H), 6.46 (s, 1H), 6.29 (s, 2H), 1.57 (s, 9H); LCMS: [M+1] = 349.0.
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
tertꢀButyl 4ꢀ(6ꢀaminoꢀ5ꢀphenylꢀ3ꢀpyridinyl)benzoate (54). Prepared from 53 (100 mg, 0.29 mmol), phenylboronic acid
(22.1 mg, 0.286 mmol), K2CO3 (99 mg, 0.72 mmol), and Pd(dppf)Cl2 (21.0 mg, 29 µmol) in 3:1 dioxane/water (8 mL), acꢀ
cording to the method utilized for compound 45, to afford the title compound (45 mg, 0.13 mmol, 45%) as a yellow oil. 1H
NMR (400 MHz, CDCl3) δ 8.37 (d, J = 1.0 Hz, 1H), 8.04 (d, J = 8.0 Hz, 2H), 7.65 (d, J = 1.0 Hz, 1H), 7.60 (d, J = 8.0 Hz,
2H), 7.54–7.46 (m, 4H), 7.46–7.39 (m, 1H), 4.78 (br s, 2H), 1.62 (s, 9H).
ꢀ
ASSOCIATED CONTENT
Accession Code
Atomic coordinates MAP4K4 kinase domain bound to compound 27 (PDB code 5J95) have been deposited with the Protein
Data Bank.
AUTHOR INFORMATION
Corresponding Author
*Eꢀmail: robert.l.dow@pfizer.com. Phone:1ꢀ617ꢀ551ꢀ3254.
ACS Paragon Plus Environment

Page 23 of 29
Journal of Medicinal Chemistry
Notes
1
2
3
4
5
6
The authors declare the following competing financial interest(s): All authors are current or former Pfizer employees. The
design and financial support for this research were provided by Pfizer. Pfizer participated in the interpretation of data, reꢀ
view and approval of the publication.
7
8
9
‡Former Pfizer employee.
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
ACKNOWLEDGMENTS
We would like to thank Hui Zhang, Allysa Dantonio, Michael West for performing TDI evaluations; John Umland for
MDR1/BCRP data; Stephen Jenkinson, Lisa Nottebaum for functional ACE inhibition analysis; Peter Burch, Jennifer
Cheung, Declan Flynn, Ted Schmahai, Eric Wolfgang, for drug safety support, Kimberly Fennell for protein production and
Katherine Brighty for assistance in writing experimentals.
ABBREVIATIONS
ACE = angiotensin converting enzyme; ATP = adenosine triphosphate ; BCRP = breast cancer resistance protein; CYP3A4 = cytoꢀ
chrome P450 3A4; MW = molecular weight; TPSA = topological polar surface area; HLM=human liver microsomes; LPS = lipoꢀ
polysaccharide; bpm = beats per minute; MDCKꢀMDR1 = Madin Darby canine kidneyꢀmultiꢀdrug resistant;
PDE=phosphodiesterase; PK/PD = pharmacokinetic/pharmacodynamics; TNFα = tumor necrosis factor alpha.
PDB Code
Compound 27: 5J95. Authors will release the atomic coordinates upon article publication.
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI:
Additional information for the HPLC purity analysis conditions, molecular formula strings, experimental procedures for
compounds 4ꢀ11, 13ꢀ36 and 38ꢀ42, kinome profiles of compounds 25 and 26, offꢀtarget profiles for compounds 4 and 37,
ACE/5HT1B functional profiles of compound 37, ActivX kinome profile of compound 37, protein production, purification
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 29
and Xꢀray crystal structure data for compound 37 bound to the kinase domain of MAP4K4 are provided. Atomic coordinates
1
2
and experimental data have been released for PDB ID code 5J95.
3
4
5
6
7
8
9
REFERENCES
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) Yao, Z.; Zhou, G.; Wang, X. S.; Brown, A.; Diener, K.; Gan, H.; Tan, T. H. A novel human STE20ꢀrelated protein kinase,
HGK, that specifically activates the cꢀJun Nꢀterminal kinase signaling pathway. J. Biol. Chem. 1999, 274: 2118–2225.
(2) Aouadi, M.; Tesz, G. J.; Nicoloro, S. M.; Wang, M.; Chouinard, M.; Soto, E.; Ostroff, G. R.; Czech, M. P. Orally delivꢀ
ered siRNA targeting macrophage Map4k4 suppresses systemic inflammation. Nature 2009, 458, 1180ꢀ1184.
(3) Wright, J. H.; Wang, X.; Manning, G.; LaMere, B. J.; Le, P.; Zhu, S.; Khatry, D.; Flanagan, P. M.; Buckley, S. D.; Whyte,
D. B.; Howlett, A. R.; Bischoff, J. R.; Lipson, K. E.; Jallal, B. The STE20 kinse HGK is broadly expressed in human tumor
cells and can modulate cellular transformation, invasion, and adhesion. Mol. Cell. Biol. 2003, 23, 2068ꢀ2082.
(4) Liang, J. J.; Wang, H.; Rashid, A.; Tan, T.ꢀH.; Hwang, R. F.; Hamilton, S. R.; Abbruzzese, J. L.; Evans, D. B.; Wang, H.
Expression of MAP4K4 is associated with worse prognosis in patients with stage II pancreatic ductal adenocarcinoma. Clin.
Cancer Res. 2008, 14, 7043ꢀ7049.
(5) Bouzakri, K.; Zierath, J. R. MAP4K4 gene silencing in human skeletal muscle prevents tumor necrosis factorꢀαꢀinduced
insulin resistance. J. Biol. Chem. 2007, 282, 7783ꢀ7789.
(6) Vitorino, P.; Yeung, S.; Crow, A.; Bakke, J.; Smyczek, T.; West, K.; McNamara, E.; EasthamꢀAnderson, J.; Gould, S.;
Harris, S. F.; Ndubaku, C.; Ye, W. MAP4K4 regulates integrinꢀFERM binding to control endothelial cell motility. Nature
2015, 519, 425ꢀ430.
(7) Collins, C. S.; Hong, J.; Sapinoso, L.; Zhou, T.; Liu, Z.; Michklash, K.; Schultz, P. G.; Hampton, G. M. A small interferꢀ
ing RNA screen for modulators of tumor cell motility identifies MAP4K4 as a promigratory kinase. Proc. Natl. Acad. Sci.
USA 2006, 103, 3775ꢀ3780.
(8) Yan, W.; Nehrke, K.; Choik J.; Barber, D. L. The Nckꢀinteracting kinase (NIK) phosphorylate the Na+ꢀH+ exchanger
NHE1 and regulates NHE1 activation by plateletꢀderived growth factor. J. Biol. Chem. 2001, 276, 31349ꢀ31356.
ACS Paragon Plus Environment

Page 25 of 29
Journal of Medicinal Chemistry
(9) LeClaire, L. L.; Rana, M.; Baumgartner, M.; Barber, D. L. The Nckꢀinteracting kinase NIK increases ARP2/3 complex
1
2
activity by phosphorylating the Arp2 subunit. J. Cell Biol. 2015, 208, 161ꢀ170.
3
4
5
6
7
8
(10) Baurmgartner, M.; Sillman, A. L.; Backwood, E. M.; Srivastava, J.; Madson, N.; Schilling, J. W.; Wright, J. H.; Barber,
D. L. The Nckꢀinteracting kinase phosphorylates ERM proteins for formation of lamellipodium by growth factors. Proc.
Natl. Acad. Sci. USA 2006, 103, 13391ꢀ13396
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
(11) Schwaid, A. G.; Su, C.; Loos, P.; Wu, J.; Nguyen, C.; Stone, K. L.; Kanyo, J.; Geoghegan, K. F.; Bhattacharya, S. K.;
Dow, R. L.; Buckbinder, L.; Carpino, P. A. MAP4K4 is a threonine kinase that phosphorylates FARP1. ACS Chem. Biol.
2015, 10, 2667ꢀ2671.
(12) Ndubaku, C. O.; Crawford, T. D.; Chen, H.; Boggs, J. W.; Brobnick, J.; Harris, S. F.; Jesudason, R.; NcNamara, E.;
Nonomiya, J.; Sambrone, A.; Schmidt, S.; Smyczek, T.; Vitorino, P.; Wang, L.; Wu, P.; Yeung, S.; Chen, J.; Chen, K.; Ding,
C. Z.; Wang, T.; Xu, Z.; Gould, S. E.; Murray, L. J.; Ye, W. Structureꢀbased design of GNEꢀ495, a potent and selective
MAP4K4 inhibitor with efficacy in retinal angiogenesis. ACS Med. Chem. Lett. 2015, 6, 913ꢀ918.
(13) Ammirati, M.; Bagley, S. W.; Bhattacharya, S. K.; Buckbinder, L.; Carlo, A. A.; Conrad, R. Cortes, C.; Dow, R. L.;
Dowling, M. S.; ElꢀKattan, A.; Ford, K.; Guimaraes, C. R. W.; Hepworth, D. Jiao, W.; LaPerle, J.; Liu, S.; Londregan, A.;
Loria, P. M.; Mathiowetz, A. M.; Munchhof, M.; Orr, S. T. M.; Petersen, D. N.; Price, D. A.; Skoura, A. Smith, A. C.; Wang,
J. Discovery of an in vivo tool to establish proofꢀofꢀconcept for MAP4K4ꢀbased antidiabetic treatment. ACS Med. Chem.
Lett. 2015, 6, 1128ꢀ1133.
(14) Schröder, P.; Förster, T.; Kleine, S.; Becker, C.; Richters, A.; Ziegler, S.; Rauh, D.; Kumar, K.; Waldmann, H. Neuritoꢀ
genic militarinoneꢀinspired 4ꢀhydroxypyridones target the stress pathway kinase MAP4K4. Angew. Chem. Int. Ed. 2015, 54,
12398ꢀ12403.
(15) Guimarães, C. R. W.; Rai, B. K.; Munchhof, M. J.; Liu, S.; Wang, J.; Bhattacharya, S. K.; Buckbinder, L. Understandꢀ
ing the impact of the Pꢀloop conformation on kinase selectivity. J. Chem. Inf. Model. 2011, 51, 1199ꢀ1204.
(16) Aouadi, M.; Tesz, G. J.; Nicoloro, S. M.; Wang, M.; Chouinard, M.; Soto, E.; Ostroff, G. R.; Czech, M. P. Orally delivꢀ
ered siRNA targeting macrophage MAP4K4 suppresses systemic inflammation. Nature, 2009, 458, 1180ꢀ1184.
(17) Unpublished results on MAP4K4 enzymatic/TDI profiling of ~60 analogs from this chemical series.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 29
(18) Shah, R. R.; Morganroth, J.; Shah, D. R. Cardiovascular safety of tyrosine kinase inhibitors: with a special focus on carꢀ
1
2
diac repolarisation (QT interval). Drug Saf. 2013, 36, 295ꢀ316.
3
4
5
6
7
8
(19) Percent change in midazolam metabolism by CYP3A4 @ 60, 30, 10, 3 and 1 ꢀM test compound concentrations: 1: 61%,
54%, 27%, 14%; 5: 26%, 3%, 1%, ꢀ4%, respectively.
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
(20) Apparent, intrinsic human liver microsomal clearance for 1, 3, 4 and 5: 42, < 8, < 8 and 14 mL/min/kg, respectively.
(21) Di, L.; WhitneyꢀPickett, C.; Umland, J. P.; Zhang, H.; Zhang, X.; Gebhard, D. F.; Lai, Y.; Fedrico III, J. J.; Davidson, R.
F.; Smith, R.; Reyner, E. L.; Lee, C.; Feng, B.; Rotter, C.; Varma, M. V.; Kempshall, S.; Fenner, K.; ElꢀKattan, A. F.; Liston,
T. E.; Troutman, M. D. Development of a new permeability assay using lowꢀefflux MDCKII cells. J. Pharm. Sci. 2011, 100,
4974ꢀ4985.
(22) Given their high ATPꢀsite homology, members of this aminopyridine series also inhibit the other GCK family members
(ZC2, TNIK and ZC3, MINK) to a similar degree as MAP4K4.
(23) Orr, S. T.; Rip, S. L.; Ballard, T. E.; Henderson, J. L.; Scott, D. O.; Obach, R. S.; Sun, H.; Kalgutkar, A. S.. Mechanismꢀ
based inactivation (MBI) of cytochrome P450 enzymes: structure−activity relationships and discovery strategies to mitigate
drug−drug interaction risks J. Med. Chem. 2012; 55, 4896ꢀ4933.
(24) Benbow, J. W.; Aubrecht, J.; Banker, M. J.; Nettleton, D.; Aleo, M. D., Predicting safety toleration of pharmaceutical
chemical leads: cytotoxicity correlations to exploratory toxicity studies. Toxicol. Lett. 2010, 197, 175ꢀ182.
(25) Aubrecht, J.; Osowski, J. J.; Persaud, P.; Cheung, J. R.; Ackerman, J.; Lopes, S. H.; Ku, W. W. Bioluminescent Salmoꢀ
nella reverse mutation assay: a screen for detecting mutagenicity with high throughput attributes. Mutagenesis 2007, 22, 335ꢀ
342.
(26) Following completion of our work there was a report of an alternative approach to peripherallyꢀrestricted MAP4K4
inhibitors: Reference 12.
(27) Kikuchi, R.; de Morais, S. M.; Kalvass, J. C. In vitro Pꢀglycoprotein efflux ratio can predict the in vivo brain penetration
regardless of biopharmaceutics drug disposition classification system class. Drug Metab. Dispos. 2013, 41, 2012ꢀ2017.
ACS Paragon Plus Environment

Page 27 of 29
Journal of Medicinal Chemistry
(28) Zhou, L.; Schmidt, K.; Nelson, F. R.; Zelesky, V.; Troutman, M. D.; Feng, B. The effect of breast cancer resistance
protein and Pꢀglycoprotein on the brain penetration of flavopiridol, imatinib mesylate (gleevac), prazosin, and 2ꢀmethoxyꢀ3ꢀ
(4ꢀ(2ꢀ(5ꢀmethylꢀ2ꢀphenyloxazolꢀ4ꢀyl)ethoxy)phenyl)propanoic acid (PFꢀ0407288) in mice. 2009, 37, 946ꢀ955.
1
2
3
4
5
6
7
8
9
(29) Miyaura, N.; Suzuki, A. Palladiumꢀcatalyzed crossꢀcoupling reactions of organoboron compounds. Chem. Rev. 1995,
95(7), 2457ꢀ2483.
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

Journal of Medicinal Chemistry
Page 28 of 29
Table of Contents Graphic
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
28
ACS Paragon Plus Environment

Page 29 of 29
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
74x23mm (300 x 300 DPI)
ACS Paragon Plus Environment