Biochemical, cellular and structural characterization of novel and selective ERK3 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2020.127551

Triazolo[4,5-d]pyrimidin-5-amines were identified from kinase selectivity screening as novel ERK3 inhibitors with sub-100 nanomolar potencies in a biochemical assay using MK5 as substrate and with an attractive kinase selectivity profile. ERK3 crystal structures clarified the inhibitor binding mode in the ATP pocket with impact on A-loop, GC-loop and αC-helix conformations suggesting a potential structural link towards MK5 interaction via the FHIEDE motif. The inhibitors also showed sub-100 nM potencies in a cellular ERK3 NanoBRET assay and with excellent correlation to the biochemical IC₅₀s. This novel series provides valuable tool compounds to further investigate the biological function and activation mechanism of ERK3.

Full text of DOI:10.1016/j.bmcl.2020.127551

Bioorganic&MedicinalChemistryLetters30(2020)127551
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Biochemical, cellular and structural characterization of novel and selective
ERK3 inhibitors
⁎
Ulrich Grädlera, , Michael Buscha, Birgitta Leuthnera, Michael Rabab, Lars Burgdorfa,
Martin Lehmanna, Nina Lindea, Christina Esdara
a Merck KGaA, Frankfurter Str. 250, D-64293 Darmstadt, Germany
b Crelux GmbH – a WuXi AppTec Company, Am Klopferspitz 19 a, D-82152 München, Germany
A R T I C L E I N F O
A B S T R A C T
Keywords:
Triazolo[4,5-d]pyrimidin-5-amines were identified from kinase selectivity screening as novel ERK3 inhibitors
with sub-100 nanomolar potencies in a biochemical assay using MK5 as substrate and with an attractive kinase
selectivity profile. ERK3 crystal structures clarified the inhibitor binding mode in the ATP pocket with impact on
A-loop, GC-loop and αC-helix conformations suggesting a potential structural link towards MK5 interaction via
the FHIEDE motif. The inhibitors also showed sub-100 nM potencies in a cellular ERK3 NanoBRET assay and
with excellent correlation to the biochemical IC50s. This novel series provides valuable tool compounds to
further investigate the biological function and activation mechanism of ERK3.
ERK3
Mitogen activated protein kinase
Kinase inhibitor
Protein crystal structure
Cellular NanoBRET assay
The serine-threonine kinase ERK3 (MAPK6) belongs to the “aty-
pical” sub-family of mitogen activated protein kinases (MAPKs), which
also includes the paralogues ERK4, ERK7, ERK8 and Nemo Like Kinase
(NLK).1–4 ERK3 and ERK4 share 44% sequence identity and comprise a
specific S-E-G motif in the activation loop whereas classical MAP ki-
nases such as ERK1/2 contain a conserved T-X-Y motif.2 While dual
phosphorylation of the T-X-Y motif is essential for activation of classical
MAP kinases it is less clear if single phosphorylation within the S-E-G
motif of ERK3 (S189) is critical for its kinase activity or function.5 Class
I p21-activated kinases (PAKs) have been described as potential ERK3/4
activation loop kinases, but any upstream regulation by extracellular
stimuli is poorly understood.6,7 MAPK-activated protein kinase 5 (MK5
or PRAK) is the best characterized downstream substrate of ERK3/4. In
vitro phosphorylation of the S-E-G motifs in ERK3 (Ser189) and ERK4
(Ser186) by PAKs substantially increase MK5 binding.5 On the other
hand, ERK3/4 interaction with MK5 does not require MK5 kinase ac-
tivity, but is dependent on a specific FRIEDE motif (ERK3 FHIEDE
motif: 332-337) in the C-terminal regions of ERK3 and ERK4.8,9 Inter-
action with ERK3/4 promotes MK5 Thr182 autophosphorylation in its
activation loop in vitro and in vivo, but the catalytically inactive ERK3
S189A mutant is unable to activate MK5.8–10 The reason for this dis-
crepancy is unclear and might be due to ERK4 recruitment of ERK3
kinase-dead mutants to form ERK3/ERK4 heterodimers activating
MK5.11 The functional role of ERK3 in normal physiology and disease
context is only incompletely understood so far. A recently described
ERK3 knock-out phenotype is viable and fertile, while different studies
revealed a potential role of ERK3 signaling in cancer.12 Enhanced mi-
gration and invasion of lung cancer cells is promoted by ERK3 and
depends on phosphorylation of its activation loop.13,14 In contrast,
implication of ERK3 in regulation of breast cancer cell morphology and
migration is supposed to be at least partially independent of its kinase
activity.15 However, experimental validation of ERK3 as promising
target for cancer therapy is hampered by the lack of suitable potent and
selective inhibitors. Weak ERK3 inhibition was found for the multi-ki-
nase inhibitor Crizotinib (1), while ERK1/2 inhibitors like FR180204
(2), GDC-0994 (3) or SCH772984 (4) do not inhibit ERK3 (Fig. 1).20–22
The present study describes the identification and profiling of potent
and selective ERK3 inhibitors to enable pharmacological characteriza-
tion of ERK3 biology in disease.
Triazolo[4,5-d]pyrimidin-5-amines 5–9 were identified from a ki-
nase selectivity screen (KINOMEscan™, DiscoverX, DX) as potent ERK3
binders (Table 1).23–25
Kinase activity profiling indicated promising kinase selectivity e.g.
for 5 (Kd = 0.98 µM) and 8 (Kd = 0.2 µM) hitting 4 out 389 kinases
above 80% inhibition at 1 µM (Fig. S1).26 Preparing for protein crys-
tallography, we further validated compounds 5–9 in comparison with
ERK1/2 reference inhibitors 1–4 in microscale thermophoresis (MST)
binding experiments using ERK3 (9–327) protein expressed from E. coli
(Table 1). As expected, ERK1/2 inhibitors 1–4 showed moderate (1:
MST Kd ~ 3.4 µM) or weak binding affinities (2 and 3: MST Kd = 26 µM
⁎ Corresponding author.
E-mail address: ulrich.graedler@merckgroup.com (U. Grädler).
https://doi.org/10.1016/j.bmcl.2020.127551
Received 8 June 2020; Received in revised form 4 September 2020; Accepted 7 September 2020
Availableonline11September2020
0960-894X/©2020ElsevierLtd.Allrightsreserved.

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
Fig. 1. Published ERK1/2 inhibitors (1–4).
and 14 µM) against ERK3 in line with inactivity in the DX binding assay
(Kd > 1 µM, Table 1). Compounds 5–9 demonstrated MST binding in
the Kd range of 120 nM up to 5.6 µM and in correlation with DX binding
data. ERK3 co-crystallization of inhibitors 7–9 failed, but we solved the
inhibitor binding, we conducted MST binding experiments with dif-
ferent wildtype and S189E mutant ERK3 proteins either produced from
E. coli or insect cells (Table 1). Inhibitors 5–9 demonstrated similar Kds
against different ERK3 proteins indicating no major impact on binding
mode by Ser189 phosphorylation. Based on the ERK3•7 structure, we
selected further analogs from our corporate compound library for
testing in a more functional biochemical ERK3 assay. We established an
ADP-Glo assay applying different ERK3 proteins (1-365, 1-471 and full-
length) together with MK5 kinase dead mutant (K51A) as substrate.31
Inhibitors 6–21 showed similar biochemical potencies using different
proteins in line with DX and MST binding data (Table 2). Compounds
5–12 and 17–21 were prepared according to Scheme 1 (preparation of
13–16 in Supplement Scheme 2–5).
structure with 7 by soaking into ERK3 (9–327, L290V) apo crystals
16,19
(Table S2).
The ERK3‧7 crystal structure revealed a dimer in the
asymmetric unit (ASU) with clear electron density for the inhibitor in
the ATP site of chain A, while electron density for 7 was lacking in
chain B (Fig. S1). Inhibitor 7 comprises two classical and one non-
classical H-bonds of the aminopyrimidine ring towards Met111 and
Glu109 of the hinge region (Fig. 2). The triazolo[4,5-d]pyrimidin-5-
amine ring forms van-der-Waals (vdW) contacts with Met111 and Ala47
and π–π-interactions with Phe159. The 4-aminocyclohexyl ring of 7 is
sandwiched in vdW contacts between Leu26 and Thr113, whereas the
amino group is solvent exposed. The methoxyphenyl ring of 7 adopts
vdW interactions to Val34 of the GC loop and is in crystal contacts with
Glu163 and Asp164 of chain B (Fig. 2). We have no experimental evi-
dence, if the ERK3 dimer found in the ASU of the crystal structure is
also present in solution to draw any further conclusions on the potential
relevance of this structural arrangement (Fig. S2). The activation loop
in chain A is well defined except for residues 179–186 revealing Asp171
and Phe172 in DFG-in conformation. The orientations of the R/C-
spines, HRD motif and αC-helix indicate an active kinase conformation
(active CIDI: αC-helix-in/DFG-in) in the ERK3•7 structure.28–30
Synthesis of compounds 5–12 and 17–21: a) R1-NH2, NEt3 (1 eq.),
THF, 0–23 °C, o/n, b) H2, sponge-Ni, THF, RT, o/n, c) Fuming HCl,
Sodium Nitrite (2 eq.), 0 °C, 1 h, d) R2-NH2, methoxyethanol, 80 °C, 3 h.
All four stereoisomers of 1,3-diaminocyclopentyl derivatives 9–12
were synthesized and revealed the R,R-enantiomer 9 as most potent
analogue (IC50 range 157–274 nM). Core variations of Nitrogen and
Carbon altering in 1,3-diaminocyclopentyl derivatives 13–16 indicated
significant potency drops. Significant impact on ERK3 potency was
found for 17 and 18 with 1-pyridin-4-yl-pyrrolidin-3-yl-amine sub-
stituents as S-enantiomer (17: IC50s 1.8–4.5 µM) or R-enantiomer (18:
IC50s 73–93 nM). Analogs 19–21 bearing methylpiperidin-pyrazole
substituents in the amino position and variable moieties in the triazolo
position were also very potent (e.g. 20: IC50s 30–55 nM). ERK3 crystal
structures with 18 and 21 were solved from soaking into apo crystals
obtained from optimized crystallization conditions comprising a tet-
ramer in the ASU (Table S2). Inhibitors 18 and 21 were clearly defined
by electron densities in all ATP sites of the tetramer (Fig. S4) with
identical binding modes in comparison to 7 (Fig. 3).
The S-E-G motif is well defined by electron density and phosphor-
ylation of Ser189 can be ruled out for steric reasons (Fig. S3). Quali-
tative MS peptide mapping analysis of ERK3 wildtype protein from E.
coli showed Ser189 phosphorylation, which was not detectable by
Western blot analysis with pSer189 antibody indicating Ser189 phos-
phorylation levels below limit of detection (Fig. S6-S8). In contrast,
Ser189 phosphorylation was confirmed by Western Blot using the same
antibody in ERK3 full-length and ERK3 (9–327) wildtype proteins
produced from insect cells in line with previous reports (Table S1).13 To
further investigate the potential impact of Ser189 phosphorylation on
Inhibitor 18 gave the most reasonable fit as R-enantiomer to the
electron density revealing slightly shifted positions of the 4-pyridine
ring from superimposition of all chains of the tetramer. The same
2

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
Table 1
Overview of selected triazolo[4,5-d]pyrimidin-5-amines tested for binding to recombinant ERK3 kinase domain proteins from different expression systems either by
MST or fluorescence assays.
#
Structure
ERK3
ERK3 (9-327,
wildtype, E. coli,
untreated)
ERK3 (9-327, wildtype, ERK3 (9-327,
ERK3 (9-327,
wildtype, Sf21,
untreated)
bindinga Kd
[µM]
E. coli, LPP treated)d
MST assayb Kd [µM]
S189E, E. coli,
untreated)
MST assayb Kd
[µM]
MST assayb Kd
[µM]
MST assayb Kd
[µM]
1
NT
[3.4 ( 6.2)]e
NA
NA
NA
2
88% Ctrl @
1µM
NA
> 10
NA
NA
3
4
86% Ctrl @
1µM
26 ( 8)
32 ( 1.9)
8.2 ( 8.7)
25 ( 6.5)*
93% Ctrl @
1µM
14.3 ( 9.1)
NA
NA
NA
5
6
0.98
1.3 ( 0.4)*
1.5( 0.3)
0.26 ( 0.21)*
1.1 ( 0.3)
92% Ctrl @
1µM
5.6 ( 2.4)
10.4 ( 3.8)
6.3 ( 0.7)
5.6 ( 1.8)
(continued on next page)
3

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
Table 1 (continued)
#
Structure
ERK3
ERK3 (9-327,
wildtype, E. coli,
untreated)
ERK3 (9-327, wildtype, ERK3 (9-327,
ERK3 (9-327,
wildtype, Sf21,
untreated)
bindinga Kd
[µM]
E. coli, LPP treated)d
MST assayb Kd [µM]
S189E, E. coli,
untreated)
MST assayb Kd
[µM]
MST assayb Kd
[µM]
MST assayb Kd
[µM]
7
12% Ctrl @
1µM
0.13 ( 0.23)*
0.16 ( 0.21)/[1.1
0.27 ( 0.21)*
0.89 ( 0.49)*
(
0.6)]*,f
8
9
0.2
1.5 ( 0.7)*
0.83 ( 0.34)*
0.14 ( 0.8)*
2.2 ( 0.9)*
0.29
0.12 ( 0.1)*
0.360 ( 0.163)
0.42 ( 0.21)*
0.18 ( 0.13)
a
b
ERK3 DiscoverX binding assay.
Microscale Thermophoresis (MST) binding assay,
* ERK3 Fluorescence change observed in MST assay used for Kd determination, error bars in brackets.
d
e
f
Recombinant ERK3 (9-327) protein dephosphorylated with lambda protein phosphatase (LPP).
Compound showed autofluorescence at concentrations > 0.7 µM.
Compound indicated lower and higher affinity MST Kd values. NA = not applicable. ND = not determined.
Fig. 2. ERK3•7 X-ray structure (PDB-ID: 6YLL) at 2.9 Å resolution indicated inhibitor binding in chain A of the dimer (magenta: chain A, cyan: chain B). Protein
inhibitor interactions were calculated with ViewContacts32 and are displayed as dashed lines (black: H-bonds, yellow: vdW interactions, blue: π–π-interactions).
4

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
Table 2
Biochemical potencies of selected triazolo[4,5-d]pyrimidin-5-amines tested for binding to recombinant ERK3 proteins (kinase domain and full-length form) in
comparison with ERK3 NanoBret data.
#
Structure
ERK3 bindinga Kd ERK3 (9-327) MST
Biochemical ADP Glo Assayc
ERK3 NanoBret Assayd
IC50 [µM]
[µM]
assayb Kd [µM]
ERK3 (9-365)
IC50 [µM]
ERK3 (9-471)
ERK3 (1-721)
IC50 [µM]
IC50 [µM]
NT
3
86% Ctrl @1µM
26 ( 8)
NT
NT
21.8
26
6
92% Ctrl @1µM
5.6 ( 1.8)
> 30
5.4
> 30
7
8
12% Ctrl @1µM
0.89 ( 0.49)*
2.2 ( 0.9)*
0.504
0.227
0.3
0.3
0.549
0.368
0.2
0.138
0.143
9
0.29
0.18 ( 0.13)
0.274
1.57
2.2
0.158
1.33
1.0
0.157
1.97
0.86
0.736
2.0
10
11
1.4
NT
NT
66% Ctrl @1µM
3.4
12
100% Ctrl @1µM NT
8.2
4.6
5.0
9.36
(continued on next page)
5

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
Table 2 (continued)
#
Structure
ERK3 bindinga Kd ERK3 (9-327) MST
Biochemical ADP Glo Assayc
ERK3 NanoBret Assayd
IC50 [µM]
[µM]
assayb Kd [µM]
ERK3 (9-365)
IC50 [µM]
ERK3 (9-471)
IC50 [µM]
ERK3 (1-721)
IC50 [µM]
13
92% Ctrl @1µM
NT
> 30
> 30
> 30
> 50
7.16
14
15
78% Ctrl @1µM
87% Ctrl @1µM
NT
NT
7.2
5.5
6.0
> 30
9.4
> 30
> 50
16
17
76% Ctrl @1µM
NT
NT
11
8
18
13.5
8.6
NT
1.8
4.5
2.2
18
19
0.17
NT
NT
0.093
0.11
0.078
0.050
0.073
0.049
0.038
0.165
0.035
(continued on next page)
6

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
Table 2 (continued)
#
Structure
ERK3 bindinga Kd ERK3 (9-327) MST
Biochemical ADP Glo Assayc
ERK3 NanoBret Assayd
IC50 [µM]
[µM]
assayb Kd [µM]
ERK3 (9-365)
IC50 [µM]
ERK3 (9-471)
IC50 [µM]
ERK3 (1-721)
IC50 [µM]
20
0.023
NT
0.055
0.030
0.033
0.177
0.013
0.055
21
0.2
NT
0.244
0.120
a
b
ERK3 DiscoverX binding assay.
Microscale Thermophoresis (MST) binding assay using ERK3 (9-327, wildtype, Sf21, untreated),
* ERK3 Fluorescence change observed in MST assay used for Kd determination.
c
d
Enzyme coupled ADP Glo assay using ERK3 1-365 and 1-471 from Sf21 and ERK3 full-length together with MK5 full-length as kinase dead (K51A) mutant.
NanoBret assay in HEK293 cells transiently expressing NanoLuc®-MAPK6 fusion vector. NT = not tested, NA = not applicable, ND = not determined.
comparison to the other three chains. Superimposition of ERK3•18 with
non-phosphorylated and phosphorylated ERK2 X-ray structures suggest
direct vicinities of the ERK3 αC-helix and A-loop positions towards the
L16 region containing the FHIEDE motif (Fig. 5).10
Based on this structural comparison, we assume that inhibitor in-
duced A-loop and αC-helix conformations found in the ERK3 crystal
structures might also affect the conformation of the FHIEDE motif and
protein-protein interaction with MK5. Crystallization of ERK3 (9–363)
containing the FHIEDE motif failed so far and additional experiments
also regarding the structural impact of S189 phosphorylation on A-loop,
αC-helix and FHIEDE motif conformations are required to understand
Scheme 1.
the exact mode-of-action of our inhibitors in context with functional
biochemical assay using MK5. Finally, we tested the most potent
compounds in a NanoBRET assay for binding to transiently expressed
binding position towards the solvent region was also found for the 4-
full-length ERK3 in HEK293 cells. As negative controls, we used bio-
chemically inactive (IC50 > 30 µM) compounds 6 and 13 together
with ERK1/2 inhibitor GDC-0994 (3). NanoBRET IC50s overall corre-
lated very well with biochemical IC50s (Table 2, correlation plots in
Figs. S15-S17) and all negative controls showed reduced or no occu-
pancies (6: IC50 = 26 µM, 13: IC50s > 50 µM, 3: IC50s = 22 µM).
Derivative 18 was among the most potent compounds (NanoBRET
IC50 = 38 nM, biochemical IC50s 73–93 nM) and indicated an attractive
selectivity profile with only 2 out of 389 kinases inhibited above 80% at
1 µM (Fig. S1).
(1H-pyrazol-1-yl)piperidine moiety of 21 with two alternative or-
ientations of the piperidine ring either towards Thr113 or Asp114
(Fig. 3). Interestingly, we observed significant A-loop flexibilities in
ERK3•18 and ERK3•21 structures revealing two chains of the tetramers
either in active CIDI (chain B and C) or inactive ωCD/DFG-intermediate
(chain A and D) conformations (Fig. 3). In addition, the nitrile group of
21 is oriented via different angles towards variable DFG-in and DFG-out
conformations. The A-loop of chain D is completely defined in both
structures, but electron density is partly lacking for the other chains
indicating flexibility. Again, Ser189 phosphorylation in the S-E-G motif
can be ruled out for steric reasons. Of note are unusual GC-loop con-
formations comprising short helices in two chains of the tetramers,
which might be due to the presence of four glycine residues (Fig. 4).
Most remarkable is the presence of two alternative αC-helix con-
formations within the tetramers of ERK3•18 (Fig. 4) and ERK3•21 (Fig.
We have identified a series of triazolo[4,5-d]pyrimidin-5-amines as
novel ERK3 inhibitors with sub-100 nM biochemical potency and at-
tractive kinase selectivity. Protein crystallography enlightened inhibitor
binding in the ATP site with impact on A-loop, GC-loop and αC-helix
conformations suggesting a potential structural link towards MK5 in-
teraction via the FHIEDE motif. The inhibitors were also potent in a
cellular ERK3 NanoBRET assay with sub-100 nM potencies and in
S5) structures indicating αC-helix out positions of chain
D in
7

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
Fig. 3. ERK3 X-ray structures in complex with 18 (PDB-ID: 6YKY, 2.5 Å resolution) and 21 (PDB-ID: 6YLC, 2.4 Å resolution) showed identical inhibitor binding
positions together with A-loop flexibility (DFG-in & DFG-intermediate) in different chains of the tetramer (green: chain A, cyan: chain B, magenta: chain C, yellow:
chain D).
excellent correlation with biochemical IC50s. This novel series provides
valuable tool compounds to further investigate the biological function
and activation mechanism of ERK3.
Declaration of Competing Interest
The authors declare the following financial interests/personal re-
lationships which may be considered as potential competing interests:
U.G., M.B., B.L., L.B., M.L., N.L. and C.E. are full-time employees of
Merck KGaA, Germany. M.R. is an employee of Crelux GmbH – a WuXi
AppTec company, Germany.
Accession codes
Coordinates and structure factors have been deposited to the protein
data bank with codes 6YLL, 6YKY and 6YLC.
8

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
Fig. 4. ERK3•18 X-ray structure (PDB-ID: 6YKY,
2.5 Å resolution) indicated variable αC-helix posi-
tions together with A-loop (DFG-in & DFG-inter-
mediate) and GC-loop flexibilities in different chains
of the tetramer (green: chain A, cyan: chain B, ma-
genta: chain C, yellow: chain D).
Fig. 5. Superimposition of two alternative αC-helix
positions (magenta: chain C, yellow: chain D) in
ERK3•18 X-ray structure (PDB-ID: 6YKY) with ERK2
non-phosphorylated (PDB-ID: 1ERK, black) and
phosphorylated forms (PDB-ID: 2ERK, grey) suggests
direct vicinity of the αC-helix, A-loop and FHIEDE
motif (green and orange).
Acknowledgement
References
We thank Prof. Dr. Krishnaraj Rajalingam (Cell Biology Unit
Universitymedical center, Mainz, Germany) for fruitful discussions and
Dr. Andreas Blum (Merck Darmstadt, Germany) for support in pre-
paring kinase selectivity data for illustration.
1. Boulton TG, Nye SH, Robbins DJ, et al. ERKs: A family of protein-serine/threonine
kinases that are activated and tyrosine phosphorylated in response to insulin and
NGF. Cell. 1991;65(4):663–675. https://doi.org/10.1016/0092-8674(91)90098-J.
2. Cargnello M, Roux PP. Activation and function of the MAPKs and their substrates, the
MAPK-activated protein kinases. Microbiol Mol Biol Rev. 2011;75(1):50–83. https://
doi.org/10.1128/mmbr.00031-10.
3. Zhu AX, Zhao Y, Moller DE, Flier JS. Cloning and characterization of p97(MAPK), a
novel human homolog of rat ERK-3. Mol Cell Biol. 1994;14(12):8202–8211.
4. Cheng M, Boulton TG, Cobb MH. ERK3 is a constitutively nuclear protein kinase. J
Biol Chem. 1996;271(15):8951–8958. https://doi.org/10.1074/jbc.271.15.8951.
5. Déléris P, Rousseau J, Coulombe P, Rodier G, Tanguay PL, Meloche S. Activation loop
phosphorylation of the atypical MAP kinases ERK3 and ERK4 is required for binding,
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2020.127551.
9

U. Grädler, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127551
activation and cytoplasmic relocalization of MK5. J Cell Physiol.
19. Coulombe P, Meloche S. Atypical mitogen-activated protein kinases: structure, reg-
ulation and functions. Biochim Biophys Acta - Mol Cell Res. 2007;1773(8):1376–1387.
https://doi.org/10.1016/j.bbamcr.2006.11.001.
2008;217(3):778–788. https://doi.org/10.1002/jcp.21560.
6. De La Mota-Peynado A, Chernoff J, Beeser A. Identification of the atypical MAPK
Erk3 as a novel substrate for p21-activated Kinase (Pak) activity. J Biol Chem.
2011;286(15):13603–13611. https://doi.org/10.1074/jbc.M110.181743.
7. Déléris P, Trost M, Topisirovic I, et al. Activation loop phosphorylation of ERK3/
ERK4 by group I p21-activated kinases (PAKs) defines a novel PAK-ERK3/4-MAPK-
activated protein kinase 5 signaling pathway. J Biol Chem. 2011;286(8):6470–6478.
https://doi.org/10.1074/jbc.M110.181529.
20. Ohori M, Kinoshita T, Okubo M, et al. Identification of a selective ERK inhibitor and
structural determination of the inhibitor-ERK2 complex. Biochem Biophys Res
Commun. 2005;336(1):357–363. https://doi.org/10.1016/j.bbrc.2005.08.082.
21. Blake JF, Burkard M, Chan J, et al. Discovery of (S)-1-(1-(4-Chloro-3-fluorophenyl)-
2-hydroxyethyl)-4-(2-((1-methyl-1H-pyrazol-5-yl)amino)pyrimidin-4-yl)pyridin-
2(1H)-one (GDC-0994), an extracellular signal-regulated kinase 1/2 (ERK1/2) in-
hibitor in early clinical development. J Med Chem. 2016;59(12):5650–5660. https://
doi.org/10.1021/acs.jmedchem.6b00389.
8. Åberg E, Perander M, Johansen B, et al. Regulation of MAPK-activated protein kinase
5 activity and subcellular localization by the atypical MAPK ERK4/MAPK4. J Biol
Chem. 2006;281(46):35499–35510. https://doi.org/10.1074/jbc.M606225200.
9. Schumacher S, Laaß K, Kant S, et al. Scaffolding by ERK3 regulates MK5 in devel-
opment. EMBO J. 2004;23(24):4770–4779. https://doi.org/10.1038/sj.emboj.
7600467.
22. Morris EJ, Jha S, Restaino CR, et al. Discovery of a novel ERK inhibitor with activity
in models of acquired resistance to BRAF and MEK inhibitors. Cancer Discov.
2013;3(7):742–750. https://doi.org/10.1158/2159-8290.CD-13-0070.
23. Fabian MA, Biggs 3rd WH, Treiber DK, et al. A small molecule-kinase interaction map
for clinical kinase inhibitors. Nat Biotechnol. 2005;23(3):329–336. https://doi.org/
10.1038/nbt1068.
10. Åberg E, Torgersen KM, Johansen B, Keyse SM, Perander M, Seternes O-M. Docking
of PRAK/MK5 to the atypical MAPKs ERK3 and ERK4 defines a novel MAPK inter-
action motif. J Biol Chem. 2009;284(29):19392–19401. https://doi.org/10.1074/jbc.
M109.023283.
24. Dorsch, D, Hoelzemann, G, Calderini, M, Wegener, A, Poeschke O. Preparation of
triazolo[4,5-d]pyrimidine derivatives for the treatment of diseases such as cancer.
2014;21(12).
11. Kant S, Schumacher S, Singh MK, Kispert A, Kotlyarov A, Gaestel M. Characterization
of the atypical MAPK ERK4 and its activation of the MAPK-activated protein kinase
MK5. J Biol Chem. 2006;281(46):35511–35519. https://doi.org/10.1074/jbc.
M606693200.
25. Dorsch, D, Hoelzemann, G, Calderini, M, Wegener, A, Poeschke, O, Busch, M,
Seenisamy J. Preparation of 9-(aryl or heteroaryl)-2-(pyrazolyl, pyrrolidinyl or cy-
clopentyl)aminopyrine derivatives as anticancer agents. 2014;560102(12).
26. Graczyk PP. Gini Coefficient: a new way to express selectivity of kinase inhibitors
against a family of kinases. J Med Chem. 2007;50(23):5773–5779. https://doi.org/
10.1021/jm070562u.
12. Ronkina N, Schuster-Gossler K, Hansmann F, et al. Germ line deletion reveals a
nonessential role of atypical mitogen-activated protein kinase 6/extracellular signal-
regulated kinase 3. Mol Cell Biol. 2019;39(6) https://doi.org/10.1128/MCB.
00516-18.
28. Kornev AP, Haste NM, Taylor SS, Ten Eyck LF. Surface comparison of active and
inactive protein kinases identifies a conserved activation mechanism. Proc Natl Acad
Sci U S A. 2006;103(47):17783–17788. https://doi.org/10.1073/pnas.0607656103.
29. Kornev AP, Taylor SS, Ten Eyck LF. A helix scaffold for the assembly of active protein
kinases. Proc Natl Acad Sci U S A. 2008;105(38):14377–14382. https://doi.org/10.
1073/pnas.0807988105.
13. Elkhadragy L, Alsaran H, Morel M, Long W. Activation loop phosphorylation of ERK3
is important for its kinase activity and ability to promote lung cancer cell invasive-
ness. J Biol Chem. 2018;293(42):16193–16205. https://doi.org/10.1074/jbc.RA118.
003699.
14. Long W, Foulds CE, Qin J, et al. ERK3 signals through SRC-3 coactivator to promote
human lung cancer cell invasion. J Clin Invest. 2012;122(5):1869–1880. https://doi.
org/10.1172/JCI61492.
30. Ung PMU, Rahman R, Schlessinger A. Redefining the protein kinase conformational
space with machine learning. Cell Chem Biol. 2018;25(7):916–924.e2. https://doi.
org/10.1016/j.chembiol.2018.05.002.
15. Al-Mahdi R, Babteen N, Thillai K, et al. A novel role for atypical MAPK kinase ERK3
in regulating breast cancer cell morphology and migration. Cell Adhes Migr.
2015;9(6):483–494. https://doi.org/10.1080/19336918.2015.1112485.
16. Filippakopoulos P, Turnbull AP, Savitsky P, Fedorov O, Berridge G, Das S, Elkins J,
Soundararajan M, Yang XW, Doyle D, Papagrigoriou E, Pike ACW, Bunkonczi G,
Ugochukwu E, Debreczeni J, von Delft F, Edwards A, Arrowsmith C, Sundstrom M,
Knapp S SGC (SGC). No Title. https://www.thesgc.org/structures/2i6l.
31. Coulombe P, Meloche S. Dual-tag prokaryotic vectors for enhanced expression of full-
length recombinant proteins. Anal Biochem. 2002;310(2):219–222. https://doi.org/
10.1016/S0003-2697(02)00319-6.
32. Kuhn B, Fuchs JE, Reutlinger M, Stahl M, Taylor NR. Rationalizing tight ligand
binding through cooperative interaction networks. J Chem Inf Model.
2011;51(12):3180–3198. https://doi.org/10.1021/ci200319e.
10