Adenosine-binding motif mimicry and cellular effects of a thieno[2,3-d]pyrimidine-based chemical inhibitor of atypical protein kinase C isoenzymes

Article abstract of DOI:10.1042/BJ20121871

The aPKC [atypical PKC (protein kinase C)] isoforms ? and ζ play crucial roles in the formation and maintenance of cell polarity and represent attractive anti-oncogenic drug targets in Ras-dependent tumours. To date, few isoform-specific chemical biology

Full text of DOI:10.1042/BJ20121871

Biochem. J. (2013) 451, 329–342 (Printed in Great Britain) doi:10.1042/BJ20121871
329
Adenosine-binding motif mimicry and cellular effects of a
thieno[2,3-d]pyrimidine-based chemical inhibitor of atypical protein
kinase C isoenzymes
1
Svend KJÆR*1, Mark LINCH† , Andrew PURKISS*, Brenda KOSTELECKY*†, Phillip P. KNOWLES*, Carine ROSSE†, Philippe
RIOU†, Christelle SOUDY‡, Sarah KAYE‡, Bhavisha PATEL‡, Erika SORIANO*, Judith MURRAY-RUST*, Caroline BARTON‡,
2
2
Christian DILLON‡, Jon ROFFEY‡, Peter J. PARKER†§ and Neil Q. MCDONALD*ꢀ
*Structural Biology, Cancer Research UK, 44 Lincoln’s Inn Fields, London WC2A 3LY, U.K., †Protein Phosphorylation Laboratories, Cancer Research UK, 44 Lincoln’s Inn Fields,
London WC2A 3LY, U.K., ‡Cancer Research Technology Discovery Laboratories, Wolfson Institute for Biomedical Research, University College London, Gower Street, London WC1E
6BT, U.K., §Division of Cancer Studies, King’s College London, New Hunts House, Guy’s Campus, London SE1 1UL, U.K., and ꢀInstitute of Structural and Molecular Biology,
Department of Biological Sciences, Birkbeck College, University of London, Malet Street, London WC1E 7HX, U.K.
The aPKC [atypical PKC (protein kinase C)] isoforms ι and ζ play
crucial roles in the formation and maintenance of cell polarity and
represent attractive anti-oncogenic drug targets in Ras-dependent
tumours. To date, few isoform-specific chemical biology tools are
available to inhibit aPKC catalytic activity. In the present paper,
we describe the identification and functional characterization
of potent and selective thieno[2,3-d]pyrimidine-based chemical
inhibitors of aPKCs. A crystal structure of human PKCι kinase
domain bound to a representative compound, CRT0066854,
reveals the basis for potent and selective chemical inhibition.
Furthermore, CRT0066854 displaces a crucial Asn-Phe-Asp
motif that is part of the adenosine-binding pocket and engages
an acidic patch used by arginine-rich PKC substrates. We show
that CRT0066854 inhibits the LLGL2 (lethal giant larvae 2) phos-
phorylation in cell lines and exhibits phenotypic effects in a range
of cell-based assays. We conclude that this compound can be used
as a chemical tool to modulate aPKC activity in vitro and in vivo
and may guide the search for further aPKC-selective inhibitors.
Key words: ATP pocket, atypical protein kinase C, cell-based
assay, cell migration, cell polarity, chemical inhibitor,
crystallography, Madin–Darby canine kidney (MDCK) cell,
protein kinase A/protein kinase G/protein kinase C family kinase
(AGC kinase).
The aPKC isoforms (PKCζ and PKCι) diverge from other PKC
family members as their regulatory domains are unresponsive
to diacylglycerol, phosphatidylserine and Ca2 + . Instead they
are regulated by protein interactions, for example with polarity
proteins Par-6, Par-3 and the Rho family GTPase cdc42 [5]. These
components assemble into an evolutionarily conserved Par-6–
aPKC–Par-3 complex present in both invertebrates and vertebrates
that controls apical–basal polarity in epithelial cells and anterior–
posterior positioning in asymmetric cell division [6,7]. In addition
to an established role in cell polarity, aPKC is required for normal
cell proliferation, mitotic spindle orientation [8,9] and migration
[10,11].
There has been considerable interest in aPKCs as drug targets
due to their role in cancer development and progression [12]. In
cancer-specific cell models, PKCζ is required for EGF (epidermal
growth factor)-induced migration of human breast and lung cancer
cells and CSF-1 (colony-stimulating factor 1) chemotaxis of
macrophages and acute monocytic leukaemic cells [13–15]. PKCι
promotes nicotine-induced migration and invasion of lung cancer
cells via phosphorylation of m- and μ-calpains [16] and is required
for non-adherent cell growth in a Rac-1-dependent manner [17].
The potential importance of aPKC in Her2-driven malignancies
INTRODUCTION
PKC (protein kinase C) isoenzymes are an important family
of serine/threonine protein kinases that contribute to many
diverse cellular and tissue functions, as well as human disease
pathologies including cancer development and progression
[1,2]. The molecular architecture of PKC family members is
conserved throughout the cPKC (classical PKC; cPKCα, cPKCβ
and cPKCγ ), nPKC (novel PKC; nPKCδ, nPKCε, nPKCη and
nPKCθ) and aPKC (atypical PKC; aPKCζ and aPKCι) isoforms.
They all comprise a C-terminal serine/threonine protein kinase
domain [belonging to the AGC kinase (protein kinase A/protein
kinase G/PKC family kinase) superfamily] and an N-terminal
regulatory domain [3]. The kinase domain has a C-terminal
extension unique to AGC kinases that contain essential ‘priming’
phosphorylation sites required for catalytic activation [4]; in
aPKCs, the most C-terminal of these is replaced by a phospho-
mimetic glutamate residue. The regulatory domain within all
PKC isoforms has an inhibitory region (pseudo-substrate motif)
and distinctive arrangements of C1, C2 and/or PB1 (Phox/Bem1)
domains (the last being exclusive to the aPKCs) that together are
crucial for conferring isoform-specific functions [3].
Abbreviations used: Ade, adenosine-binding; AGC kinase, protein kinase A/protein kinase G/protein kinase C family kinase; aPKC, atypical protein
kinase C; BIM-1, bisindolylmaleimide-1; CCD, charge-coupled-device; cPKC, classical protein kinase C; 2D, two-dimensional; 3D, three-dimensional;
DCM, dichloromethane; DMEM, Dulbecco’s modified Eagle’s medium; DTT, dithiothreitol; F-actin, filamentous actin; FBS, fetal bovine serum; GFP, green
fluorescent protein; GST, glutathione transferase; HEK, human embryonic kidney; HRP, horseradish peroxidase; HTS, high-throughput screen; IMAP,
immobilized metal ion-affinity fluorescence polarization; LC, liquid chromatography; LLGL, lethal giant larvae; MDCK, Madin–Darby canine kidney; MI,
mass ion; nPKC, novel protein kinase C; NRK, normal rat kidney cell; PKC, protein kinase C; PSAL, predominantly single apical lumen; RMSD, root mean
square deviation; RNAi, RNA interference; RT, retention time; SPR, surface plasmon resonance; TLS, translation libration screw-motion.
1
These authors contributed equally to this work.
Correspondence may be addressed to either of these authors (email neil.mcdonald@cancer.org.uk or peter.parker@cancer.org.uk).
2
The structural co-ordinates for PKCιk bound to CRT0066854 will appear in the PDB under the accession number 3ZH8.
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

330
S. Kjær and others
was demonstrated when the aberrant polarity induced by forced
Her2 dimerization in human mammary epithelial MCF-10A cells
embedded in MatrigelTM was partially reversed following genetic
disruption of aPKC binding to Par-6 [18]. In model organisms,
aPKC has been shown to co-operate with bona fide oncogenes,
including Ras and Notch in Drosophila melanogaster, to lead
to invasive and metastatic tumours [19,20]. Furthermore, in
transgenic mice K-Ras-induced lung cancer is reduced following
depletion of PKCι or its adaptor protein p62 [21,22].
To
a solution of 4-cyanopyridine (1.25 g, 12 mmol) in
dry dioxane (10 ml) was added 2-amino-4,5,6,7-tetrahydro-
benzo[b]thiophene-3-carboxylic acid ethyl ester (2.25 g,
10 mmol) followed by potassium-tert-pentylate 1.7 M in toluene
(12 ml, 20 mmol). The reaction mixture was stirred at room
temperature (20◦C) overnight. After completion the precipitate
formed was filtered and washed with diethyl ether. The residue
was used without any further purification in the next step.
LC (liquid chromatography)–MS method: RT (retention time):
1
In human malignancy, one of the most frequent amplicons is
3q26, the location of the PKCι gene (HGNC symbol PRKCI).
PKCι gene amplification leads to high protein levels and is cor-
related with poor clinical outcomes in squamous cell carcinomas
of the lung [17] and oesophagus [23]. High PKCι expression has
been correlated with poor prognosis and/or worse stage in an in-
creasing number of human cancers including breast [24], pancreas
[25], stomach [26], bile duct [27] and brain [28]. In addition to the
compelling evidence for a role of PKCι in human malignancy, the
interest in PKCι as a therapeutic target was further enhanced by
the finding that elevated PKCι protein is associated with chemo-
therapy resistance [29] and that depletion/inhibition of PKCι is
synergistic with some conventional chemotherapies [30–32].
Like many protein kinases, aPKC isoenzymes are amenable
to small-molecule chemical inhibition either through their ATP-
binding clefts or allosteric pockets [33]. However, there are
few selective chemical biology tools to inhibit aPKC catalytic
activity. Available tools include the non-selective and relatively
impotent BIM-1 (bisindolylmaleimide 1), Go¨6983 and Go¨6976
compounds or a PKCζ pseudo-substrate peptide containing a
membrane-targeting myristoylation site [34]. The development
of effective ATP-competitive inhibitors against aPKCs, similar
to other protein kinases, has been complicated by challenges
in both potency and selectivity [33]. In general, such inhibitors
acquire selectivity by associating with both the ATP-binding site
and with adjacent residues, which are less well conserved. Other
approaches include screening for allosteric inhibitors of PKCζ
that target the PIF pocket [35] or blocking PKCι interaction with
Par-6, for example with ATM (aurothiomalate), a quite different
class of PKCι inhibitor which exhibits potent antitumour activity
in lung cancer [36].
3.54 min, MI (mass ion): 284 [M + 1]. H NMR (300 MHz,
DMSO): 8.56 (d, 2 H), 8.12 (d, 2 H), 2.90 (m, 2 H), 2.67
(m, 2 H), 1.76 (m, 4 H).
Synthesis of 2,4,6-tri-isopropyl-benzenesulfonic acid 2-pyridin-4-
yl-5,6,7,8-tetrahydro-benzo[4,5]thieno[2,3-d]pyrimidin-4-yl es-
ter (A-2)
To
a solution of 2-pyridin-4-yl-5,6,7,8-tetrahydro-benzo[4,
5]thieno[2,3-d]pyrimidin-4-ol (1 g, 3.5 mmol) (A-1) in
DCM (dichloromethane, 10 ml) was added 2,4,6-tri-
isopropylbenzenesulfonyl
chloride
(1.3 g,
4.2 mmol),
triethylamine (Et3N) (1.5 ml, 10.5 mmol) and DMAP (4-
dimethylaminopyridine) (6 mg, 0.05 mmol). The mixture was
stirred at room temperature for 1 h. After completion the mixture
was diluted with water and the product was extracted into DCM
(2×2 ml). The combined extracts were dried with magnesium
sulfate, filtered and evaporated under reduced pressure to yield
the title compound as a brown solid, which was used without
further purification in the next step. LC–MS method: RT:
6.23 min, MI: 550 [M + 1].
In the present study, we report the identification of an aPKC
inhibitor CRT0066854 which is able to mimic both ATP and
a key side chain that forms part of the aPKC nucleotide-
binding cleft. We define the compound’s mechanism of action
by crystallographic and biochemical methods and demonstrate its
utility through the use of phenotypic cell-based assays. Together,
our data indicate that CRT0066854 is a potent and selective
chemical tool to modulate aPKC activity both in vitro and in
cells.
Synthesis of (S)-3-phenyl-propane-1,2-diamine (A-3)
To a suspension of (S)-2-amino-3-phenyl-propionamide hy-
drochloride (540 mg, 2.7 mmol) in DCM (5 ml) was added
triethylamine (Et3N) (380 μl, 2.7 mmol). The suspension was
stirred for 2 h at room temperature, the resulting solid was filtered
and the filtrate was concentrated under reduced pressure to yield a
white solid, to which was added dropwise a 1 M solution of BH3 in
THF (tetrahydrofuran, 20 ml, 20 mmol), the solution was stirred
overnight at reflux. After cooling the solution was hydrolysed by
slow addition of an excess of 10% acetic acid/methanol (30 ml)
and refluxed for a further 2 h. The solvent was removed under
reduced pressure, and the residue was dissolved in methanol and
passed through a SCX-2 cartridge and washed with methanol.
The product was released from the cartridge using a solution of
2 M ammonia/methanol. The solvent was evaporated to provide
the title compound as a white solid which was used in the next
step without further purification. LC–MS method: RT: 0.36 min,
MI: 151 [M + 1].
EXPERIMENTAL
Chemical synthesis of CRT0066854 and CRT0066390
Synthesis of (S)-N*1*-(2-pyridin-4-yl-5,6,7,8-tetrahydro-
benzo[4,5]thieno[2,3-d]pyrimidin-4-yl)-butane-1,2-diamine
(CRT0066854)
Synthesis of 2-pyridin-4-yl-5,6,7,8-tetrahydro-benzo[4,5]thieno
[2,3-d]pyrimidin-4-ol (A-1)
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

A novel small molecule aPKC inhibitor
331
C1 vector (Clontech) using 5ꢂ-SalI and 3ꢂ-BamHI restriction sites.
The cDNA contains two start codons at bp 1–3 and bp 28–30 with
the second methionine residue denoted as the first amino acid
of the protein. LLGL2 (lethal giant larvae 2) cDNA (GenBank®
accession number NM_004524.2), a gift from T. Pawson (Samuel
Lunenfeld Research Institute, ON, Canada), was subcloned into
pEGFP-C1 with and without a 5ꢂ FLAG-tag sequence using
5ꢂ-BamH1 and 3ꢂ-Sal1 restriction sites. For biochemical and
structural studies, the kinase domain of PKCι (residues 239–
587) was inserted into a pBacPAK-His3 vector (Clontech) using
BamHI and XhoI restriction sites. The vector was modified to
have a GST (glutathione transferase) followed by a 3C protease
cleavage site upstream of the kinase domain. A C586S mutation
was also introduced into the kinase domain using QuikChange®
(Stratagene) according to the manufacturer’s instructions in order
to avoid oxidative cross-linking during crystallization.
Synthesis of (S)-N*1*-(2-pyridin-4-yl-5,6,7,8-tetrahydro-benzo
[4,5]thieno[2,3-d]pyrimidin-4-yl)-butane-1,2-diamine
(Compound A, CRT0066854)
To
a solution of 2,4,6-tri-isopropyl-benzenesulfonic acid
2-pyridin-4-yl-5,6,7,8-tetrahydro-benzo[4,5]thieno[2,3-d]
pyrimidin-4-yl ester (A-2) (100 mg, 0.180 mmol) in DMA
(dimethylacetamide, 2 ml) was added (S)-3-phenyl-propane-1,2-
diamine (A-3) (30 mg, 0.180 mmol) followed by triethylamine
(Et3N) (50 μl, 0.36 mmol), the mixture was stirred at room
temperature for 2 h. After completion the mixture was loaded
on to a SCX-2 cartridge and washed with methanol. The
product was released from the cartridge using a solution of
2 M ammonia/methanol. The ammonia/methanol eluent was
concentrated under reduced pressure and the crude product
was purified by preparative HPLC to yield the title compound.
PKCιk and PKCι protein expression and purification
Recombinant baculoviruses used for infection were obtained
using standard protocols (Oxford Expression Technologies). High
Five cells were grown in shaker flasks in SFIII medium (Life
Technologies) and 10 μg/ml gentamycin (Life Technologies).
High Five cells were infected with recombinant baculoviruses at a
multiplicity of infection of 2. To increase the phosphorylation state
of PKCι to the fully active form, High Five cells were co-infected
with PDK-1 baculovirus at a multiplicity of infection of 1. High
Five cells were harvested 72 h post-infection and resuspended
in lysis buffer {20 mM Hepes (pH 7.4) (Sigma), 150 mM NaCl
(Sigma), 10 mM benzamidine (Sigma), 0.2 mM AEBSF [4-
(2-aminoethyl)benzenesulfonyl fluoride] (Melford Laboratories),
1 mM EDTA (Sigma) and 1 mM DTT (dithiothreitol; Melford
Laboratories)}. All purification steps were carried out at
4◦C or on ice. Cells were lysed by sonication and spun
down at 30000 g for 30 min. PKCιk or PKCι were purified
using glutathione–Sepharose 4B beads (Amersham Biosciences),
followed by removal of the GST affinity tag with GST-3C protease
(PreScission Protease, Amersham Bioscience) and ion-exchange
chromatography (Hi-Trap Q column, GE Healthcare). The protein
was dialysed into the final buffer [25 mM Tris (pH 8.0), 150 mM
NaCl, 1 mM EDTA and 1 mM DTT] and concentrated before use.
1
LC–MS method: 4, RT: 2.51 min, MI: 416 [M + 1]. H NMR
(300 MHz, DMSO): 8.64 (d, 2 H), 7.95 (d, 2 H), 7.36 (m, 5 H),
3.92 (m, 2 H), 3.46 (m, 2 H), 2.92 (m, 1 H), 2.91 (m, 2 H), 2.79
(m, 2 H), 1.83 (m, 4 H).
Synthesis
of
(R)-N*1*-(2-pyridin-4-yl-5,6,7,8-tetrahydro-
benzo[4,5]thieno[2,3-d]pyrimidin-4-yl)-butane-1,2-diamine
(Compound B, CRT0066390)
High-throughput screening and selectivity screens
We performed a small molecule HTS (high-throughput screen)
against recombinant PKCιk with confirmed hits profiled against
recombinant PKCζk to verify pan-aPKC inhibition. We
screened a 56000 CRT (Cancer Research Technology) chemical
library collection at a single concentration of 30 μM [IMAP
(immobilized metal ion-affinity fluorescence polarization) format,
Molecular Devices]. The 56000 compounds comprises 46000
from a diverse collection and 10000 from a kinase-focused
collection (including in-house compounds). The HTS assay
used 10 pM PKCι and 100 nM 5FAM (fluorescein-amidite)-
ERMRPRKRQGSVRRRV-NH2 peptide and 30 μM ATP in a
buffer consisting of 20 mM Hepes (pH 7.5), 10 mM MgCl2,
1 mM DTT and 0.25 mM EGTA. The PKCιk Km for ATP was
determined to be 30 μM using this IMAP assay. We used an
isoquinoline sulfonamide H-series compound H-9 as the assay
standard (IC50 = 2 μM). All final Z’-factors were greater than
0.5 and most were above 0.6. The H-9 assay standard inhibition
(R)-N*1*-(2-pyridin-4-yl-5,6,7,8-tetrahydro-
benzo[4,5]thieno[2,3-d]pyrimidin-4-yl)-butane-1,2-diamine
(Compound B, CRT0066390) was prepared from (R)-3-
phenyl-propane-1,2-diamine (A-4) and 2,4,6-tri-isopropyl-
benzenesulfonic
acid
2-pyridin-4-yl-5,6,7,8-tetrahydro-
benzo[4,5]thieno[2,3-d]pyrimidin-4-yl ester (A-2) to yield
the title compound. LC–MS method: RT: 2.54 min, MI: 416
[M + 1]. 1H NMR (300 MHz, DMSO): 8.74 (d, 2 H), 7.15 (d, 2
H), 7.34 (m, 5 H), 3.90 (m, 2 H), 3.52 (m, 2 H), 2.95 (m, 1 H),
2.94 (m, 2 H), 2.79 (m, 2 H), 1.83 (m, 4 H).
The tricyclic thienopyrimidine compound ASN2993118
(2-pyridin-4-yl-5,6,7,8-tetrahydro-benzo[4,5]thieno[2,3-
d]pyrimidin-4-ylamine) was purchased directly from Asinex.
Plasmids
+
at 2 μM gave 46.7 4.8% inhibition of PKCιk activity. Hits
−
For mammalian studies, human PKCι cDNA (GenBank® acces-
sion number NM_002740.5), a gift from T. Biden (Garvan Insti-
tute of Medical Research, Australia), was subcloned into pEGFP-
from the HTS screen gave IC50 values from 0.2 to 30 μM. The
initial hit rate was unusually low for a kinase HTS at 0.2%
with no confirmed hits arising from the kinase-focused subset
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

332
S. Kjær and others
of our library. As the HTS used a truncated form of PKCι,
we profiled our hits against the full-length human PKCι. The
tricyclic thienopyrimidine ASN2993118 demonstrated activity
against both full-length and the kinase domain of PKCι (as well
as the kinase domain of PKCζ). Further validation of PKCι
inhibitory activity was undertaken in orthogonal assay formats
with ASN2993118, and mechanistically it was shown to have
an ATP-competitive mode of inhibition. These studies confirmed
ASN2993118 as a bona fide hit, and it was progressed into lead
identification. Lead identification involved assaying modifications
to the C4 position of the central pyrimidine ring of ASN2993118.
The Millipore KinaseProfilerTM service (Millipore) was used to
determine the selectivity of CRT0066854 against a panel of
kinases. This service used a radiometric assay and 1 μM single-
point concentrations of CRT0066854. An ATP concentration of
30 μM (equivalent to the Km of PKCιk) was used.
were collected at Diamond [beamline IO3, at a wavelength of 0.92
Å (1 Å = 0.1 nm)] and processed with the CCP4 (collaborative
computational project 4) software [37]. Molecular replacement
was performed using the program Phaser [38], using the protein
structure of PKCιk bound to BIM-1 (PDB code 1ZRZ) as a
search model. Refinement was carried out with Phenix [39], using
torsion angle NCS (non-crystallographic symmetry) and TLS
(translation libration screw-motion; with three TLS groups per
protein chain) parameters. For a summary of the crystallographic
data, see Table 3. There are three molecules within the asymmetric
unit, denoted A, B and C, with RMSDs (root mean square
deviations) between Cα atoms A–B 0.13 Å for 321 pairs, A–
C 0.13 Å for 312 pairs and B–C 0.19 Å for 314 pairs. Missing
regions in the final model due to poor or non-existent electron
density included residues 448–455 and 543–551. These residues
belong to an aPKC-specific insert motif (residues 448–455) and
the Ade (adenosine-binding) motif respectively. Selected well-
defined water molecules were included together with two iodide
ions. The position of the iodide ions was confirmed by inspection
of an anomalous difference Fourier electron density map.
In vitro kinetic assays for PKCι and IC50 determinations
The ADP Quest kit (DiscoveRx) was used to determine the kcat and
Km values for fully primed phospho-species of PKCιk, PKCζk and
the nPKC isoform PKCεk as a positive control (Supplementary
Table S1 at http://www.biochemj.org/bj/451/bj4510329add.htm).
This assay uses a coupled reaction to convert ADP into a
product that has a fluorescence excitation at 530 nm and emission
at 590 nm. Two synthetic peptides, a model substrate peptide
(ERMRPRKQGSVRRRV) derived from PKCε and human
LLGL2 substrate peptide (LKKSLRQSFRRMRR). For the kcat
(app) and Km (app) reactions, the ATP concentration ranged from
0 to 200 μM while the synthetic peptide substrates were kept
constant at 200 μM to avoid product inhibition. PKC isoenzyme
concentrations were 5 nM for all three isoenzymes. The reactions
were measured every 2 min for 30 min in a 384-well plate
using a Safire2 plate reader (Tecan). For determination of IC50
values, reactions were prepared using 20 μM ATP and 20 μM
model substrate peptide. Protein concentrations were the same as
indicated in the previous paragraph. Serial dilutions of inhibitors
with concentrations ranging from 1 nM to 100 μM were added
and fluorescence intensity was measured in end point assay mode.
Cell culture
HEK (human embryonic kidney)-293, A549 and MDCK
(Madin–Darby canine kidney) cells were obtained from Cell
Production, Cancer Research UK (CRUK) and cultivated in
DMEM (Dulbecco’s modified Eagle’s medium), 10% FBS (fetal
bovine serum) and penicillin–streptomycin (Invitrogen). HCT-
116 cells were obtained from Cell Production, CRUK, and
grown in McCoy’s 5A medium with 10% FBS and penicillin–
streptomycin. NRK (normal rat kidney)-49F cells were a gift from
Dr J. Camonis (Institut Curie, Paris, France) and have previously
been characterized [10]. The cells were cultivated in DMEM and
10% FBS under 5% CO2 conditions.
Antibodies
The following antibodies were used for immunoblotting: mouse
monoclonal anti-PKCλ/ι antibody used at a 1:500 dilution
(BD Biosciences), mouse monoclonal anti-α-tubulin antibody at
a 1:10000 dilution (Sigma), rabbit polyclonal anti-pLLGL1/2
antibody at a 1:500 dilution (Abnova), rabbit polyclonal
anti-phospho-serine PKC substrate antibody used at a 1:500
dilution (Cell Signaling Technology; broad phospho-serine motif
recognition for AGC kinase substrates) and rabbit polyclonal
anti-GFP (green fluorescent protein) antibody used at a 1:1000
dilution (Santa Cruz Biotechnology). In ELISA, the anti-
pLLGL1/2 antibody (Abnova) was used at a 1:1000 dilution and
HRP (horseradish peroxidase)-conjugated goat anti-(rabbit IgG)
antibody (Perbio Science) was used at a 1:2000 dilution.
SPR (surface plasmon resonance) measurements for CRT0066854
binding to PKCιk
To conduct SPR measurements to obtain affinity and on and off
rates for the inhibitor, we immobilized an anti-GST antibody
(GEHC GST-capture kit) by amine coupling using standard
protocols in flow cells 1 and 2. We then captured GST-tagged
PKCιk in flow cell 2 using 25 μg/ml in running buffer, 180 s,
5 μl/min. The running buffer consisted of 50 mM Tris, 150 mM
NaCl, 10 mM MgCl2, 1 mM MnCl2, 1 mM DTT and 0.05% P20
at pH 7.5. A GST control was captured in flow cell 1 using
25 μg/ml in the same running buffer (180 s, 5 μl/min). Then
we ran a concentration series of inhibitor CRT0066854 over the
captured PKCιk. Concentrations used were 0, 16.125, 32.25, 62.5,
125, 250, 500, 1000, 1500 and 2000 nM.
Western blotting
For immunoblotting, lysates or immunoprecipitates were resolved
using precast NuPAGE 4–12% Bis-Tris gels (Invitrogen)
and transferred on to nitrocellulose membranes (PROTRAN,
Whatman). Immunoblots were blocked in 3% BSA–TBST
(Tris-buffered saline containing 0.1% Tween 20) and
probed with primary antibodies as indicated. Following
incubation with species-specific HRP-conjugated secondary
antibodies, bands were visualized radiographically.
Crystallography and structure determination
Crystals of PKCιk bound to CRT0066854 were obtained at 20◦C
using 200 nl drops formed by 100 nl of protein solution in 20 mM
Tris (pH 7.5), 150 mM NaCl and 2 mM DTT buffer (containing
a 5-fold molar excess of CRT0066854), mixed with 100 nl
of precipitant solution containing 0.2 M ammonium iodide and
20% (v/v) PEG3350 [poly(ethylene glycol) 3350]. Crystals were
cryoprotected using 20% ethylene glycol. Diffraction data
Immunoprecipitation
After 36 h, transfected HCT-116 cells were lysed in lysis buffer
[1% Triton X-100, 20 mM Tris/HCl (pH 8), 130 mM NaCl, 1 mM
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

A novel small molecule aPKC inhibitor
333
DTT, 10 mM sodium fluoride, complete EDTA-free protease
inhibitor cocktail (Roche) and phosphatase inhibitor cocktails
(set II + IV; Calbiochem)]. After centrifugation (13000 g, 4◦C,
10 min), soluble proteins were precleared then incubated with
anti-GFP magnetic beads (GFP–TRAP; Chromotek) for 90 min
at 4◦C. Beads were washed five times with lysis buffer
(containing 260 mM NaCl), then the bound proteins were eluted
in Laemmli sample buffer, resolved by SDS/PAGE and analysed
by immunoblotting. Where indicated, cells were treated for
2 h before lysis with DMSO (as a control), 10 μM BIM-1
(Calbiochem) or 5 μM Go¨6983 (Calbiochem) or CRT0066854.
3D (three-dimensional) cystogenesis assay
The 3D culture of MDCK cells in MatrigelTM was performed
as follows: MDCK cells in exponential-phase growth were
trypsinized and resuspended in standard medium, supplemented
with 2% low growth factor MatrigelTM (BD Biosciences) at
2×104 cells/ml. Each well of an eight-well chamber slide (BD
Biosciences) was precoated with 30 μl of 100% MatrigelTM
to which 400 μl of the cell suspension was added. Inhibitors
were added on day 1 and the medium of 2% MatrigelTM in the
presence or absence of inhibitor was changed on alternate days for
5 days.
Transfection
Wound migration assay
Transient reverse transfections of cDNA for HEK-293 cells
were performed on poly-L-lysine precoated plates using
LipofectamineTM 2000 as per the manufacturer’s instructions
(Invitrogen). Co-transfections of PKCι and LLGL2 were at a 4:1
DNA ratio. HCT-116 cells were transfected 24 h after plating
using FuGENE® HD as per the manufacturer’s instructions
(Roche). The cells were then grown in normal medium for
36 h before further manipulations. In ELISA, HEK-293 cells
were transfected with cDNA using JetPEI reagent as per the
manufacturer’s instructions (Autogen Bioclear).
Wounds were inflicted by scratching the cell surface with
a plastic pipette tip. Images were recorded using a Nikon
Eclipse Ti inverted videomicroscope and Coolsnap HQ2 camera
(Photometrics). All inhibitor treatments were performed with 1 h
pre-incubation. The migration speed was monitored by measuring
the change in the area surrounding the wound over time, using
Metamorph software. Statistical analysis (ANOVA 2 factors) were
performed using Prism Software.
Immunofluorescence microscopy
ELISA
MDCK cells grown in MatrigelTM on chamber slides were
fixed with 2% formaldehyde in PBS, washed in PBS and then
permeablized with 0.5% Triton X-100 in PBS. Following a
further wash, cells were directly stained with Phalloidin Alexa
Fluor® 546 (1:200 dilution; Invitrogen) and Hoechst dyes (1:5000
dilution; Invitrogen). After a quick rinse with PBS, the cultures
were mounted with Prolong Gold hard-set mounting medium
(Invitrogen).
MDCK cysts that had their F-actin (filamentous actin)
stained with Phalloidin Alexa Fluor® 546 were visualized
with a confocal microscope (Zeiss 510) and an apical lumen
assessment was carried out. The middle of the cyst in
the z-plane was identified and the following criteria were
observed to determine whether the cyst had a ‘predominantly
single lumen’: (i) a continuous F-actin-defined lumen; (ii)
the luminal actin staining being more intense than the
basal (outside) staining; (iii) a uni-dimensional measurement
of the lumen being at least one-third of the cyst diameter; and
(iv) the uni-dimensional measurement of the lumen being twice
the size of any other luminal structure.
HEK-293 cells were transfected in a 10-cm dish as per the
manufacturer’s instructions (Corning). After 16 h, the cells
were trypsinized and seeded into a 96-well plate at 1.5×104
cells/well and medium was replenished. After a further 24 h,
the medium was replaced by new medium and a range of
CRT0066854 inhibitor concentrations. After 1 h of inhibitor
treatment, lysates were prepared using ice-cold Tris lysis buffer
[150 mM NaCl, 20 mM Tris (pH 7.5), 1 mM EGTA, 1 mM EDTA
and 1% Triton X-100]. Lysates were transferred on to an anti-
FLAG-coated ELISA plate (Sigma) and incubated for 2 h with
gentle shaking, followed by an automated wash step (Tecan
plate washer) with wash buffer [50 mM Tris (pH 7.5), 0.15 M
NaCl and 0.02% Tween 20]. The immunocomplexed protein
was incubated with anti-pLLGL1/2 primary antibody overnight
at 4◦C, followed by an automated wash and then addition
of HRP-conjugated secondary antibody. After a further wash,
3,3ꢂ,5,5ꢂ-tetramethylbenzidine (Sigma) was added according to
the manufacturer’s instructions and attenuance was read at 450 nm
using an Ascent plate reader (Thermo Labsystems).
Cell viability
Statistical analysis
CellTitre-Glo® (Promega) cell viability assay was carried
out as per the manufacturer’s instructions. The luminescent
measurement of ATP was performed using the EnVision plate
reader (PerkinElmer).
Enzyme kinetic constants for PKCιk were determined by fitting
the data to the Michaelis–Menten equation using the graphics
program Prism (GraphPad Prism, version 5.0d, GraphPad
Software). Differences between the treatment groups were
assessed by ANOVA with a repeated measurement module using
Prism statistical software. P values of <0.05 were regarded
as significant. Dose–response curves and the calculation of
the Geometric Mean IC50 values were performed in SARview
software (IDPS, version 6.2).
Soft agar assay
Single-cell suspension of HeLa cells were resuspended in 0.35%
agarose in RPMI 1640 medium and 10% FBS at a density
of 1500 cells/ml. The agarose–cell mixture (3 ml) was added
to each 35-mm well that had been precoated with 1.5 ml of
0.5% agarose. Plates were incubated at 4◦C for 20 min to allow
the agarose to set before being placed in the 37◦C incubator
for 10–14 days. Colonies were stained with 0.005% Crystal
Violet for 1 h, after which images were taken using a CCD
(charge-coupled-device) camera and colony number was counted
using ImageJ software version 1.40g.
RESULTS
Discovery of potent and selective chemical inhibitors of aPKCs
To identify molecules that inhibit PKCι, a 56000-compound
diverse chemical library assembled by the CRT Discovery
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

334
S. Kjær and others
Table 1 IC50 values for selected aPKC inhibitors
a 1 μM inhibitor concentration, screened at the Km for ATP
(Table 2). This is a radiometric assay format used with each kinase.
The selectivity score S(80) is the number of kinases inhibited by
greater than 80% at 1 μM compound/total number of kinases
tested. For CRT0066854, the S(80) score is close to 0.02, which
indicates a good kinase selectivity [40]. Moreover, the compound
showed striking selectivity for aPKC over other PKC isoenzymes
(Table 2).
To test whether there was selectivity of the tool compound
between the PKC subclasses in cells, GFP-tagged cPKCα,
nPKCε and aPKCι proteins were expressed in HCT-116 cells
and the cultures were treated with CRT0066854 for 30 or
90 min. As a control we used BIM-1 and Go¨6983 to potently
inhibit full-length novel and cPKC isoenzymes expressed in
this cell line (Supplementary Figure S1 at http://www.biochemj.
CRT0066854, CRT0066390 and ASN2993118 were compared against full-length (FL) aPKCι.
PKCζ and ROCK2 kinase domain (KD) were used for CRT0066854, and CRT0066390 was used
as a control. IC50 values are geometric mean values determined using five times the kinase Km
for ATP.
aPKC inhibitors
CRT0066854 (μM)
CRT0066390 (μM)
ASN2993118 (μM)
PKCι FL
PKCζ FL
ROCK2 KD
0.132
0.623
0.620
2.5
7.1
30
26
–
–
Laboratories was screened at 30 μM against a 10 pM
biphosphorylated (i.e. fully primed) form of the PKCι kinase
domain (residues 239–587, referred to hereafter as PKCιk). The
PKCι Km for ATP was determined to be 30 μM (Supplementary
Table S1). The screen employed an IMAP (Molecular Devices)
assay using a fluorescein-amidite (FAM)-tagged PKCε-derived
pseudo-substrate sequence containing an alanine-to-serine mutant
(PSS; pseudo-substrate substrate) that we found to be an excellent
PKCι substrate. The screen identified several compounds with
potent anti-PKCιk activity including a tricyclic thieno[2,3-
d]pyrimidine ASN2993118 (Figure 1A, upper panel). This hit
had an IC50 of 11 μM against PKCιk and 26.3 μM against
full-length PKCι. Counter screening against PKCζk showed
the compound inhibited with an IC50 of 36.5 μM. The low-
molecular mass and calculated lipophilicity of ASN2993118
made it an attractive starting point for optimization. Exploration
of the activity landscape around the central pyrimidine ring
of ASN2993118 led to the identification of compounds with a
diamine side chain at the pyrimidine C4 position that exhibited
increased inhibitory activity. This included CRT0066854 in which
the pyrimidine C4 position was substituted with a S-3-phenyl-
propane-1,2-diamine moiety (Figure 1A, lower panel).
org/bj/451/bj4510329add.htm).
Immuno-complexes
were
immune-isolated through the GFP-tag and immunoblotted with
an anti-phospho-serine-PKC substrate antibody. In CRT0066854-
treated cells, the number of phosphorylated substrate bands
apparent was significantly diminished for cells overexpressing
PKCι, but not PKCα or PKCε, demonstrating aPKC specificity
in cells (Figure 1C). Together, these data suggested that
CRT0066854 would be a potent and selective chemical tool to
modulate aPKC activity both in vitro and in cells.
Kinetics of CRT0066854 inhibitor binding to PKCιk
To characterize the interaction of CRT0066854 in solution with
purified recombinant PKCιk protein, we used SPR on a BIACORE
T200 to measure the kinetics and affinity of this inhibitor. GST-
tagged PKCιk was captured using anti-GST monoclonal antibody
immobilized via amine coupling. Sensorgrams were fitted to a 1:1
Langmuir binding model and kinetics constants were obtained
and an affinity constant of 340 nM was calculated from kd/ka
(Figure 1D, upper panel). A binding isotherm was also fitted to
a 1:1 steady-state binding model and an affinity of 870 nM was
obtained in good agreement with the value derived from the rate
constants (Figure 1D, lower panel).
To characterize CRT0066854 further, we used an ADP QuestTM
assay (Discoverx) to generate dose–response curves for fully
primed phosphospecies of PKCζk, PKCιk and the nPKC isoform
PKCεk as a control. The kcat and Km kinetic constants were
determined for PKCζk, PKCιk and the nPKC isoform PKCεk using
the ADP QuestTM assay (Supplementary Table S1). CRT0066854
was used at 10 μM. We used the structurally unrelated PKC
inhibitors BIM-1 (10 μM) and Go¨6983 (5 μM) from commercial
sources as positive controls. IC50 values for CRT0066854
against PKCζk and PKCιk were 450 nM and 86 nM respectively
(Figure 1B). Comparing the IC50 values measured for BIM-1
(PKCζk IC50 = 9.1 μM and PKCιk IC50 = 5.5 μM), indicates that
CRT0066854 is approximately 20-fold more potent for PKCζk
and over 60-fold more potent for PKCιk than BIM-1. A similar
increase in potency is observed for CRT0066854 compared with
Go¨6983 (PKCζk IC50 = 13.9 μM and PKCιk IC50 = 10.7 μM)
(Figure 1B). Inhibition of the nPKCεk by CRT0066854 is
significantly weaker, with a 16-fold higher IC50 (IC50 = 7.21 μM)
compared with PKCζk and more than 80-fold greater than the IC50
for PKCιk. We then measured IC50 values for full-length PKCζ and
PKCι, which gave 639 nM and 132 nM respectively, comparable
with those obtained from isolated aPKC kinase domains.
Since CRT0066854 has a chiral centre, we prepared its enantiomer
CRT0066390 and measured its IC50 values against full-length
PKCι and PKCζ. This showed only low levels of inhibition of
aPKC isoenzymes by CRT0066390 (Table 1).
Structural basis for CRT0066854 selectivity for the PKCι kinase
domain
To understand both the selectivity and binding mode of
CRT0066854 we determined the crystal structure of CRT0066854
bound to PKCιk (Table 3). Crystals of PKCιk bound to
CRT0066854 contain three very similar PKCιk molecules in the
asymmetric unit that each have a single inhibitor bound within
the ATP cleft. The overall PKCιk conformation bears many of the
hallmarks of an active AGC kinase conformation (Supplementary
Figure S2 at http://www.biochemj.org/bj/451/bj4510329add.htm)
except for the lack of bound magnesium and the absence of an
ordered Ade motif as discussed below [41]. The CRT0066854
compound makes predominantly hydrophobic interactions within
the PKCι nucleotide-binding cleft, as well as five hydrogen
bonds, including a water-mediated hydrogen bond (Figures 2A
and 2B). The fused tricyclic scaffold (cyclohexa[4,5]thieno[2,3-
d]pyrimidine) makes hydrophobic contacts with the PKCιk
glycine loop, stabilizing a ‘closed’ conformation. This tricyclic
ring system deviates significantly from planarity and is curved
with a puckered cyclohexane ring (Figure 2A). The thieno-1
nitrogen of the pyrimidine ring makes a crucial hydrogen bond
to the Thr386 side chain. The thieno[2,3-d]pyrimidine ring system
also has two important substituents: (i) a pyridine substituent
at position 2 and (ii) a (2-amino-3-phenylpropyl)(methyl)amine
To examine the selectivity profile of CRT0066854 in vitro,
the compound was tested at Millipore using the KinaseProfilerTM
screening service (Millipore) against 106 protein kinases at
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

A novel small molecule aPKC inhibitor
335
Figure 1 CRT0066854 selectively inhibits PKCζ and PKCι kinase activity in vitro and in cells
(A) Upper panel, chemical structure of the initial hit thieno[2,3-d]pyrimidine ASN2993118. Lower panel, chemical structure of compound CRT0066854. (B) Dose–response curves for purified
recombinant PKCζ and PKCι kinase domains comparing CRT0066854, BIM-1 and Go¨6983 chemical inhibitors. IC50 values are shown for the different inhibitors. The dots within the grey bar
indicate the number and position of phosphorylation sites in different PKC isoenzymes used in the assay, confirmed by Western blots using site-specific anti-phospho-threonine antibodies.
AL/TM/HM correspond to activation loop, turn motif and hydrophobic motif respectively. The precise phosphorylation site numbering for each PKC isoenzyme is given in Supplementary Table S1
at http://www.biochemj.org/bj/451/bj4510329add.htm. (C) CRT0066854 specifically inhibits aPKC in cells. Expression vectors encoding GFP–PKC isoforms were transfected into HCT-116 cells.
Cells were treated in the presence or absence of CRT0066854 chemical inhibitor for the indicated time points before lysis. Cell lysates were immunoprecipitated with GFP–TRAP and immunoblotted
(IB) with the indicated antibodies. Inhibitors were used at 10 μM (CRT0066854); 5 μM (Go¨6983) and 10 μM (BIM-1) concentrations. (D) SPR analysis of binding of CRT0066854 to immobilized
recombinant PKCιk. Sensorgrams obtained for various CRT0066854 concentrations were flowed across the sensor chip (upper panel). RU, resonance units. Lower panel shows the steady-state
binding isotherm of maximal response (RUmax) derived from fitting sensorgrams obtained over a range of CRT0066854 analyte concentrations to a one-site binding model. Affinity constant estimates
from both methods are in close agreement.
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

336
S. Kjær and others
Table 2 Effect of 1 μM CRT0066854 on the activity of 106 protein kinases
Table 2 Continued
A radiometric assay was carried out by the Millipore Kinase Profiling Service (Millipore). The
assay determined the selectivity of 1 μM single-point concentrations of CRT0066854 against
a panel of 106 protein kinases and 30 μM ATP (Km of PKCιk). The selectivity score S(80) is
the number of kinases inhibited by more than 80% at 1 μM compound/total number of kinases
tested. h, human; m, murine.
Percentage of activity
1 μM CRT0066854
Percentage of inhibition
1 μM CRT0066854
MAPKAP-K2(h)
MAPK2(h)
MEK1(h)
MELK(h)
Mer(h)
101
99
100
77
−1
1
64
65
66
67
68
−0.5
Percentage of activity
Percentage of inhibition
1 μM CRT0066854
23
1 μM CRT0066854
89
11
Met(h)
100
106
108
90
100
88
103
105
80
89
89
96
97
107
102
89
99
109
13
101
108
23
96
83
0
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
Abl(h)
Abl (M351T)(h)
ALK(h)
ARK5(h)
ASK1(h)
Aurora-A(h)
Axl(h)
Blk(m)
Bmx(h)
BRK(h)
CDK1/cyclinB(h)
CDK2/cyclinA(h)
CDK5/p35(h)
CHK1(h)
CHK2(h)
CK1δ(h)
CK1γ 1(h)
CK1γ 2(h)
CK1γ 3(h)
cKit(h)
cKit(D816H)(h)
cKit(V560G)(h)
CSK(h)
c-RAF(h)
cSRC(h)
DAPK1(h)
DDR2(h)
93
97
95
91
104
96
106
98
107
112
86
76
32
100
90
100
101
120
110
99
80
90
103
94
89
94
108
94
102
78
87
61
69
102
101
106
102
99
88
91
109
102
96
84
96
33
97
11
3
5
1
2
3
4
5
6
7
8
9
MSK1(h)
MST1(h)
MST3(h)
NEK2(h)
p70S6K(h)
PAK4(h)
PDGFRα(h)
PDGFRα(D842V)(h)
PDK1(h)
Pim-1(h)
PKBα(h)
PKCα(h)
PKCβI(h)
PKCγ (h)
PKCδ(h)
PKCε(h)
PKCη(h)
PKCι(h)
PKCμ(h)
PKCθ(h)
PKCζ(h)
PKD2(h)
Plk3(h)
PRK2(h)
Ret(h)
ROCK-I(h)
ROCK-II(h)
Ron(h)
−6
−8
10
0
12
−2
−5
20
15
11
9
−4
−8
−6
2
−7
−12
14
30
68
0
10
0
−1
−20
−10
1
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
61
62
63
−3.5
0
−7
−2
11
−1
−5
83.5
0
−1
20
10
−3
3
77
4
17
84
−6
−9
63
16
8
106
109
37
114
94
102
102
100
114
114
102
91
−3.5
−8
6
DYRK2(h)
EGFR(h)
−14
−4
22
13
39
31
−2
−1
−6
−2
1
12
9
−9
−2
−11
16
4
67
3
2
4
6
8
2
7
Ros(h)
SAPK2a(h)
Snk(h)
TAK1(h)
Tie2(h)
TrkA(h)
Yes(h)
ZAP-70(h)
ZIPK(h)
6
EGFR(L858R)(h)
EGFR(L861Q)(h)
EGFR(T790M)(h)
EGFR(T790M, L858R)(h)
EphA2(h)
EphA7(h)
EphB4(h)
ErbB4(h)
FAK(h)
Fer(h)
Fes(h)
FGFR1(h)
FGFR2(h)
FGFR3(h)
FGFR4(h)
Flt1(h)
Flt3(D835Y)(h)
Flt3(h)
−2
−2
0
−14
−14
−2
9
100
101
102
103
104
105
106
S(80)0.019
86
14
substituent including a benzyl group, a chiral carbon and a primary
amine. The pyridine substituent acts as a kinase domain hinge-
binding moiety by forming a hydrogen bond to the main-chain
amide of Val326, approximating to the position of the adenine ring
of ATP. The primary amine of the 3-phenyl-propane-1,2-diamino
headpiece of CRT0066854 makes a crucial charged interaction
with both the backbone carbonyl of Asp373 and the side chain of
Asp330. Finally, the benzyl group in the inhibitor headpiece binds
a small hydrophobic pocket formed by the kinase domain hinge,
usually occupied by Phe543 from the Ade motif [41] (Figure 2C;
compare left-hand and middle panel). The benzyl group also
makes an orthogonal π-stacking interaction with the pyridine
ring hinge-binder, thus locking the inhibitor into a distinctly non-
planar and U-shape conformation (Figure 2A).
Superimpositions of CRT0066854 and ATP-bound structures
of PKCιk indicate the general equivalence of the benzyl moiety
and Phe543 (Figure 2D). An important consequence of inhibitor
binding is that the entire Ade motif, including Phe543 from
the Phe-Xaa-Xaa-Phe motif, becomes disordered, whereas it is
an integral part of the ATP-binding cleft in the ATP–PKCιk
Flt4(h)
Fms(h)
98
96
94
92
98
93
103
88
29
85
117
94
102
96
42
GSK3α(h)
Hck(h)
HIPK1(h)
HIPK2(h)
HIPK3(h)
IGF-1R(h)
IKKα(h)
IR(h)
IRAK4(h)
JAK3(h)
JNK1α1(h)
KDR(h)
LIMK1(h)
LKB1(h)
−3
12
65
15
−17
6
−2
4
58
5
95
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

A novel small molecule aPKC inhibitor
337
Table 3 X-ray data collection and refinement statistics
Effect of CRT0066854 on A549 lung carcinoma cell viability
A549 cells are K-Ras mutant cells of lung origin, known to
be dependent on PKCι for survival [17]. The A549 cells were
treated with CRT0066854 for 48 h at various concentrations
(0.01–20 μM), and cell viability determined by the CellTitre-Glo
luminescent assay. An IC50 of 3.47 μM was obtained from three
independent determinations (Figure 3C).
Parameter
Values
P31
Data collection
Space group
Cell dimensions
a, b, c (A)
α, β, γ ( )
Resolution (A)*
Rmerge
˚
113.6, 113.6, 82.4
90, 90, 120
49.2–2.74 (2.89–2.74)
11.0% (71.4%)
5.8% (38.4%)
6.3 (1.1)
◦
˚
Decreased colony formation due to CRT0066854
Rpim
I/σ
Although a reduction in adherent cell growth has been reported
following the depletion of PKCι in selected cell lines [46],
the anti-proliferative effects appear more pronounced for non-
adherent cell growth [17,46]. As such, we cultured HeLa cells,
which reliably form colonies within 12 days when embedded in
soft agar, and replenished inhibitor (0.1–10 μM) on alternate
days. Two separate inhibitors were used for these experiments,
the aPKC-specific CRT0066854 and its enantiomeric equivalent
CRT0066390, which is essentially inactive towards PKCι in vitro
(Table 1). At 1 μM the CRT0066854 inhibitor led to a 65%
decrease in colony formation and, as expected, the same dose of
CRT0066390 resulted in considerably less suppression of colony
formation (25% decrease) (Figures 3D and 3E). This is consistent
with CRT0066854 having a specific cellular effect via PKCι
inhibition.
Completeness
Multiplicity
100% (100%)
5.6 (5.4)
Refinement
Resolution (A)
˚
2.74
31307 (4570)
21.5% / 25.6%
Number of unique reflections
Rwork/Rfree
Atoms
Protein
Ligand/ion
Water
B-factors
Protein
Ligand/ion
Water
RMSD
Bond lengths (A)
7468
180
66
49.8
42.6
31.9
˚
◦
0.003
0.726
Bond angles ( )
*Numbers in parentheses are values in the highest resolution shell.
Inhibition of polarized epithelial morphogenesis by CRT0066854
Depletion of PKCι in cells using RNAi (RNA interference) leads
to impaired luminal morphogenesis in epithelial cell 3D cysts as
a consequence of impaired spindle formation and apical–basal
polarity [9,47]. We therefore tested whether chemical inhibition
could induce the same phenotype in MDCK cells embedded
in MatrigelTM. MDCK cells are widely used as models for
studying epithelia, as they have clear apical–basolateral polarity
and polarize in 2D (two-dimensional) and 3D cell cultures. In
MatrigelTM, MDCK cells are known to form cysts with a single
lumen. We determined the number of MDCK cysts that had a
PSAL (predominantly single apical lumen) after growing in the
presence of 0, 0.33 or 1 μM concentrations of CRT0066854.
Surprisingly, inhibitor addition caused significant disruption to
lumen formation, resulting in a multiple luminal morphology
(Figure 4A). This disruption in lumen formation was dose-
dependent and quantification of this effect indicated that at 1 μM
concentrations of CRT0066854 there was a 50% reduction in
PSALs compared with a DMSO control (Figure 4B).
complex (PDB code 3A8W) (Figure 2C). Recent data have
shown for PKCβ2, an equivalent phenyl side chain to Phe543
that has been implicated in allosteric activation [42]. This
residue is held out of the ATP cleft by the C1 domain
preventing full enzyme activation until diacylglycerol binds,
providing an additional layer of PKC regulation. It seems
likely that CRT0066854 targets PKCιk for inhibition not only
by competitive binding to the nucleotide-binding cleft, but
also by preventing a catalytically competent conformation via
displacement of the Ade motif by the benzyl inhibitor substituent.
Both features combine to contribute towards the high specificity
of CRT0066854 for aPKCs. A detailed comparison of inhibitor
contents compared with other AGC kinases is given in the
Supplementary Online Data and Supplementary Table S2 (at
http://www.biochemj.org/bj/451/bj4510329add.htm).
CRT0066854 inhibition of LLGL2 phosphorylation in cells
LLGL is a recognized substrate of aPKC in D. melanogaster
[43] and in immunocomplexes from HEK-293 cells expressing
exogenous aPKC [44]. As LLGL2 is expressed at low levels
in HEK-293 cells and due to the lack of a sensitive phospho-
specific antibody against LLGL1/2, we used a co-expression
system of aPKC and LLGL2 to evaluate the cellular IC50 of
CRT0066854 in HEK-293 cells. Treatment of HEK-293 cells,
co-transfected with PKCι wild-type and FLAG-tagged LLGL2,
with CRT0066854 (0.125–8 μM) for 1 h led to a reproducible
dose-dependent reduction in LLGL2 phosphorylation as seen by
Western blotting (Figure 3A). In order to test this robustly in
a high-throughput format, the phosphorylation signal window
was optimized using the co-transfection of constitutively active
PKCι-A120E [45] and FLAG-tagged LLGL2 and an ELISA was
established. This demonstrated that CRT006854 has a cellular
phospho-LLGL2 IC50 of 0.87 μM comparable with the IC50
measured in vitro against PKCι (Figure 3B).
CRT0066854 impedes directed migration of NRK cells
Depletion of aPKC using RNAi decreases the rate of healing
in an NRK cell scratch-wound assay due to an impairment
of the polarized delivery of the exocyst to the cell’s leading
edge [10]. We used this model to test the effect of aPKC
inhibition by CRT0066854 on directional migration. Scratch
wounds were generated in wells of confluent NRK cells after
which DMSO, or 3, 6 or 12 μM CRT0066854 was added to
the wells and time-lapse imaging recorded the wound healing
over 10 h (Supplementary Movie S1 at http://www.biochemj.org/
bj/451/bj4510329add.htm). Both 6 μM and 12 μM CRT0066854
led to a significant reduction in wound healing, with 12 μM
halving the normalized migration speed (Figures 4C and 4D).
Note that optimum knockdown of aPKCs in this model reduces
migration by 60% [10].
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

338
S. Kjær and others
Figure 2 Structural basis for CRT0066854 inhibition of PKCιk
(A) Close-up of the CRT0066854-binding site within the PKCι nucleotide-binding cleft and kinase domain hinge, highlighting the five hydrogen-bond contacts and the lack of planarity of CRT0066854.
Selected protein side chains and main chains are shown for clarity. A bound water molecule (red sphere) and an iodide ion (purple sphere) are also shown. (B) A 2D representation of CRT0066854
inhibitor–PKCι contacts produced by LIGPLOT + [51]. (C) Comparison of the nucleotide pockets between CRT0066854 and ATP complexes of PKCι. The left-hand panel shows the electrostatic
surface of PKCι, including the two acidic residues that interact with the inhibitor (yellow sticks). Phe543 from the Asn-Phe-Asp motif of PKCι (middle panel, green sticks and salmon-pink cartoon)
contacts ATP in a similar manner to Phe327 of protein kinase A (right-hand panel, green cartoon and sticks). Inhibitor binding leads to (i) displacement of the Asn-Phe-Asp motif (left-hand panel),
(ii) changes in the glycine loop, and (iii) changes in the Tyr256 (glycine loop)–Trp289 (Cα-helix) interaction. (D) Superimposition of the nucleotide-binding pockets of PKCι bound to ATP and to
CRT0066854. This shows that the phenyl group of CRT0066854 (yellow sticks) mimics the placement of Phe543 (green sticks) from the Ade motif observed in the ATP complex.
available inhibitors have greater efficacy for the cPKCs and
nPKCs [34]. The discovery of CRT0066854 and a knowledge
of the detailed contacts it makes within the aPKC kinase ATP-
binding cleft demonstrates that isoform specificity can be obtained
using the residues surrounding the nucleotide-binding site, and
provides a useful starting point for rational design of even more
potent and specific inhibitors.
DISCUSSION
In the present study, we report the identification and
characterization of a selective inhibitor of the aPKC isoenzymes.
Selectivity and potency issues have proved particularly
challenging in the search for inhibitors of individual PKC
isoforms, particularly aPKCζ and aPKCι, where all commercially
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

A novel small molecule aPKC inhibitor
339
Figure 3 CRT0066854 inhibits LLGL2 phosphorylation and cell viability
(A) HEK-293 cells were co-transfected with LLGL2 and PKCι-WT (wild-type). After 35 h, the cells were treated with CRT0066854 (0.125–8 μM) for 1 h, lysates were generated separated by gel
electrophoresis and immunoblotted with pLLGL1/2. (B) HEK-293 cells co-transfected with FLAG–LLGL2 and PKCι-A120E (constitutively active). After 40 h, cells were treated with CRT0066854
(0.001–12 μM) for 1 h and an ELISA was performed. (C) A549 cells in log-phase growth as a monolayer were treated with CRT0066854 at the indicated concentrations for 48 h and the cell viability
was estimated using CellTitreGlo. (D) HeLa cells from a single cell suspension were cultured in soft agar for 14 days, with the medium and inhibitor being replenished on alternate days. Colonies
+
were stained with Crystal Violet and images of the 35-mm well taken using a mounted CCD camera. (E) The effect of aPKC inhibitors on HeLa colony growth was quantified and the means S.E.M.
is presented for at least three separate experiments.
−
Recent studies have shown differential effects between PKCζ
and PKCι for compounds targeting other aPKC-specific pockets
outside of the ATP cleft [35,48]. The selectivity of CRT0066854
shown in the present study for both aPKCζ and aPKCι
could prove useful in both clinical and research settings. We
provide compelling evidence for both in vitro and phenotypic
effects of CRT0066854 using a range of cellular assays
relevant to cancer therapy. Non-adherent cell growth, migration
and aberrant polarity are all processes exploited by cancer and
were more affected by CRT0066854 than adherent cell
growth. This differential dependence on aPKC for 3D and 2D
cellular phenotypes has been demonstrated previously by genetic
depletion of aPKC [17], and together with the work described in
the present paper, supports the role of CRT0066854 as a selective
aPKC inhibitor in cells. PKCι overexpression has been linked to
Our crystallographic analysis has revealed that CRT0066854
has a bifunctional inhibitory effect. First, it is a competitive ATP
inhibitor, binding within the nucleotide-binding pocket of aPKC.
Secondly, it uses a benzyl group to displace the Ade motif from
a peripheral hinge pocket adjacent to the nucleotide cleft and
prevents it forming a functional ATP-binding pocket. These two
features ensure that CRT0066854 is highly selective for aPKC.
Some strain/induced fit is evident within the compound in view
of its lack of planarity when bound to PKCιk. Inhibitory effects
on other kinases such as ROCK2 and PRK2 may be due to the
presence of a similar phenyl group equivalent to Phe543, suggesting
that related scaffolds may permit a more generic approach towards
inhibiting AGC kinases bearing an Asn-Phe-Asp motif. Binding
isotherms show that the inhibitor has a low off-rate, generally
deemed to be an advantageous feature for a kinase inhibitor.
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

340
S. Kjær and others
Figure 4 Phenotypic effects of CRT0066854 on cell polarity and directed cell migration
(A) MDCK cells were cultured in MatrigelTM for 6 days. The cultures were treated with 0, 0.33 or 1 μM of the aPKCι inhibitor CRT0066854 which were replenished, along with growth medium,
on alternate days. Cultures were fixed and stained with Phalloidin (F-actin) and counterstained with Hoechst (DNA). Representative single mid-section confocal images are shown. The scale bars
+
represent 50 μM. (B) Quantification of the number of PSALs was made. At least 100 cysts were counted per condition in each experiment and the means S.E.M. are presented for at least three
−
separate experiments. (C) Confluent monolayers of NRK cells were wounded, treated with CRT0066854 at the indicated concentrations and healing was followed over time. The panels shown are
representative of at least three independent experiments. (D) The normalized mean migration speed is plotted for different CRT0066854 concentrations. The error bars represent the S.E.M. and
one-way ANOVA was used to determine statistical significance of difference: *P < 0.05, ***P < 0.001.
inhibitor efficacy against multiple PKC isoforms in cells; Christian Dillon and Jon Roffey
ledthechemistryproject team thatdiscovered CRT0066854; CarolineBartonand Bhavisha
Patel conducted the LLGL2 phosphorylation assays and cell viability assay; Christelle
Soudy synthesized the inhibitors used in these studies; Carine Rosse conducted the
directed cell migration assay; Sarah Kaye carried out IC50 assays on full-length aPKC
isoenzymes; Mark Linch carried out the LLGL2 phosphorylation assay, cell polarity and
colony formation assay; Peter Parker and Neil McDonald planned the project and designed
experiments;andNeilMcDonald, JudithMurray-RustandMarkLinchpreparedtheFigures
and wrote the paper.
several types of cancer, including non-small-cell lung cancer and
ovarian cancer [17,49,50] and further optimization of inhibitors on
the basis of the CRT0066854 structure could yield more specific
cancer therapeutics. CRT0066854 and its derivative compounds
will be useful tools to further dissect PKCζ and PKCι roles in
establishing cell polarity and growth-factor-stimulated signalling
pathways.
AUTHOR CONTRIBUTION
ACKNOWLEDGEMENTS
Svend Kjær purified, assembled and crystallized the CRT0066854 inhibitor–PKCιk
complex; Andrew Purkiss collected the data, and determined and refined the CRT0066854-
bound structure; Brenda Kostelecky and Erika Soriano carried out kinetic assays with
inhibitors; Phillip Knowles carried out the SPR experiments; Philippe Riou assessed
We thank members of the N.Q.M. and P.J.P. laboratories for helpful discussions and
comments on the paper before submission. We gratefully acknowledge the expert
assistance of Tamar Telem-Shafir (GE Healthcare) in performing and analysing the SPR
data.
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

A novel small molecule aPKC inhibitor
341
24 Kojima, Y., Akimoto, K., Nagashima, Y., Ishiguro, H., Shirai, S., Chishima, T., Ichikawa, Y.,
Ishikawa, T., Sasaki, T., Kubota, Y. et al. (2008) The overexpression and altered
localization of the atypical protein kinase C λ/ι in breast cancer correlates with the
pathologic type of these tumors. Hum. Pathol. 39, 824–831
25 Scotti, M. L., Bamlet, W. R., Smyrk, T. C., Fields, A. P. and Murray, N. R. (2010) Protein
kinase Cι is required for pancreatic cancer cell transformed growth and tumorigenesis.
Cancer Res. 70, 2064–2074
FUNDING
S. Kjær was funded by a Cancer Research UK fellowship and by a Marie Curie Fellowship
[grant number MEIF-CT-2005-010656]; M.L. was funded by a Cancer Research UK
Clinical Research Fellowship. Research in the N.Q.M. and P.J.P. laboratories is supported
by CRUK core funding to the London Research Institute.
26 Takagawa, R., Akimoto, K., Ichikawa, Y., Akiyama, H., Kojima, Y., Ishiguro, H., Inayama, Y.,
Aoki, I., Kunisaki, C., Endo, I. et al. (2010) High expression of atypical protein kinase C
λ/ι in gastric cancer as a prognostic factor for recurrence. Ann. Surg. Oncol. 17, 81–88
27 Li, Q., Wang, J. M., Liu, C., Xiao, B. L., Lu, J. X. and Zou, S. Q. (2008) Correlation of
aPKC-ι and E-cadherin expression with invasion and prognosis of cholangiocarcinoma.
Hepatobiliary Pancreatic Dis. Int. 7, 70–75
28 Patel, R., Win, H., Desai, S., Patel, K., Matthews, J. A. and Acevedo-Duncan, M. (2008)
Involvement of PKC-ι in glioma proliferation. Cell Proliferation 41, 122–135
29 Murray, N. R. and Fields, A. P. (1997) Atypical protein kinase C ι protects human
leukemia cells against drug-induced apoptosis. J. Biol. Chem. 272, 27521–27524
30 Baldwin, R. M., Garratt-Lalonde, M., Parolin, D. A., Krzyzanowski, P. M., Andrade, M. A.
and Lorimer, I. A. (2006) Protection of glioblastoma cells from cisplatin cytotoxicity via
protein kinase Cι -mediated attenuation of p38 MAP kinase signaling. Oncogene 25,
2909–2919
31 Jin, Y. T., Ying, X. X., Hu, Y. H., Zou, Q., Wang, H. Y. and Xu, Y. H. (2008) aPKC inhibitors
might be the sensitizer of chemotherapy and adoptive immunotherapy in the treatment of
hASIPa-overexpressed breast cancer. Oncol. Res. 17, 59–68
32 Kikuchi, K., Soundararajan, A., Zarzabal, L. A., Weems, C. R., Nelon, L. D., Hampton, S.
T., Michalek, J. E., Rubin, B. P., Fields, A. P. and Keller, C. (2012) Protein kinase C ι as a
therapeutic target in alveolar rhabdomyosarcoma. Oncogene 32, 286–95
33 Knight, Z. A. and Shokat, K. M. (2005) Features of selective kinase inhibitors. Chem. Biol.
12, 621–637
34 Roffey, J., Rosse, C., Linch, M., Hibbert, A., McDonald, N. Q. and Parker, P. J. (2009)
Protein kinase C intervention: the state of play. Curr. Opin. Cell Biol. 21, 268–279
35 Frohner, W., Lopez-Garcia, L. A., Neimanis, S., Weber, N., Navratil, J., Maurer, F., Stroba,
A., Zhang, H., Biondi, R. M. and Engel, M. (2011) 4-Benzimidazolyl-3-phenylbutanoic
acids as novel PIF-pocket-targeting allosteric inhibitors of protein kinase PKCζ. J. Med.
Chem. 54, 6714–6723
36 Erdogan, E., Lamark, T., Stallings-Mann, M., Lee, J., Pellecchia, M., Thompson, E. A.,
Johansen, T. and Fields, A. P. (2006) Aurothiomalate inhibits transformed growth by
targeting the PB1 domain of protein kinase Cι. J. Biol. Chem. 281, 28450–28459
37 Collaborative Computational Project-Number 4 (1994) The CCP-4 suite: programs for
protein crystallography. Acta Crystallogr., Sect. D: Biol. Crystallogr. 50, 760–763
38 McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C. and
Read, R. J. (2007) Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674
39 Adams, P. D., Grosse-Kunstleve, R. W., Hung, L. W., Ioerger, T. R., McCoy, A. J., Moriarty,
N. W., Read, R. J., Sacchettini, J. C., Sauter, N. K. and Terwilliger, T. C. (2002) PHENIX:
building new software for automated crystallographic structure determination. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 58, 1948–1954
REFERENCES
1
Bosco, R., Melloni, E., Celeghini, C., Rimondi, E., Vaccarezza, M. and Zauli, G. (2011)
Fine tuning of protein kinase C (PKC) isoforms in cancer: shortening the distance from
the laboratory to the bedside. Mini-Rev. Med. Chem. 11, 185–199
2
Rosse, C., Linch, M., Kermorgant, S., Cameron, A. J., Boeckeler, K. and Parker, P. J.
(2010) PKC and the control of localized signal dynamics. Nat. Rev. Mol. Cell Biol. 11,
103–112
3
4
Parker, P. J. and Murray-Rust, J. (2004) PKC at a glance. J. Cell Sci. 117, 131–132
Pearce, L. R., Komander, D. and Alessi, D. R. (2010) The nuts and bolts of AGC protein
kinases. Nat. Rev. Mol. Cell Biol. 11, 9–22
5
6
7
8
Suzuki, A., Akimoto, K. and Ohno, S. (2003) Protein kinase C λ/ι (PKCλ/ι): a PKC
isotype essential for the development of multicellular organisms. J. Biochem. 133, 9–16
Knoblich, J. A. (2010) Asymmetric cell division: recent developments and their
implications for tumour biology. Nat. Rev. Mol. Cell Biol. 11, 849–860
Suzuki, A. and Ohno, S. (2006) The PAR–aPKC system: lessons in polarity. J. Cell Sci.
119, 979–987
Hao, Y., Du, Q., Chen, X., Zheng, Z., Balsbaugh, J. L., Maitra, S., Shabanowitz, J., Hunt,
D. F. and Macara, I. G. (2010) Par3 controls epithelial spindle orientation by
aPKC-mediated phosphorylation of apical Pins. Curr. Biol. 20, 1809–1818
Durgan, J., Kaji, N., Jin, D. and Hall, A. (2011) Par6B and atypical PKC regulate mitotic
spindle orientation during epithelial morphogenesis. J. Biol. Chem. 286, 12461–12474
9
10 Rosse, C., Formstecher, E., Boeckeler, K., Zhao, Y., Kremerskothen, J., White, M. D.,
Camonis, J. H. and Parker, P. J. (2009) An aPKC–exocyst complex controls paxillin
phosphorylation and migration through localised JNK1 activation. PLoS Biol. 7,
e1000235
11 Etienne-Manneville, S. and Hall, A. (2003) Cell polarity: Par6, aPKC and cytoskeletal
crosstalk. Curr. Opin. Cell Biol. 15, 67–72
12 Fields, A. P. and Regala, R. P. (2007) Protein kinase C iota: human oncogene, prognostic
marker and therapeutic target. Pharmacol. Res. 55, 487–497
13 Guo, H., Gu, F., Li, W., Zhang, B., Niu, R., Fu, L., Zhang, N. and Ma, Y. (2009) Reduction
of protein kinase C ζ inhibits migration and invasion of human glioblastoma cells. J.
Neurochem. 109, 203–213
14 Liu, Y., Wang, B., Wang, J., Wan, W., Sun, R., Zhao, Y. and Zhang, N. (2009)
Down-regulation of PKCζ expression inhibits chemotaxis signal transduction in human
lung cancer cells. Lung Cancer 63, 210–218
15 Sun, R., Gao, P., Chen, L., Ma, D., Wang, J., Oppenheim, J. J. and Zhang, N. (2005)
Protein kinase C ζ is required for epidermal growth factor-induced chemotaxis of human
breast cancer cells. Cancer Res. 65, 1433–1441
16 Xu, L. and Deng, X. (2006) Protein kinase Cι promotes nicotine-induced migration and
invasion of cancer cells via phosphorylation of μ- and m-calpains. J. Biol. Chem. 281,
4457–4466
17 Regala, R. P., Weems, C., Jamieson, L., Copland, J. A., Thompson, E. A. and Fields, A. P.
(2005) Atypical protein kinase Cι plays a critical role in human lung cancer cell growth
and tumorigenicity. J. Biol. Chem. 280, 31109–31115
40 Bain, J., Plater, L., Elliott, M., Shpiro, N., Hastie, C. J., McLauchlan, H., Klevernic, I.,
Arthur, J. S., Alessi, D. R. and Cohen, P. (2007) The selectivity of protein kinase
inhibitors: a further update. Biochem. J. 408, 297–315
41 Romano, R. A., Kannan, N., Kornev, A. P., Allison, C. J. and Taylor, S. S. (2009) A
chimeric mechanism for polyvalent trans-phosphorylation of PKA by PDK1. Protein Sci.
18, 1486–1497
42 Leonard, T. A., Rozycki, B., Saidi, L. F., Hummer, G. and Hurley, J. H. (2011) Crystal
structure and allosteric activation of protein kinase C βII. Cell 144, 55–66
43 Betschinger, J., Mechtler, K. and Knoblich, J. A. (2003) The Par complex directs
asymmetric cell division by phosphorylating the cytoskeletal protein Lgl. Nature 422,
326–330
44 Plant, P. J., Fawcett, J. P., Lin, D. C., Holdorf, A. D., Binns, K., Kulkarni, S. and Pawson, T.
(2003) A polarity complex of mPar-6 and atypical PKC binds, phosphorylates and
regulates mammalian Lgl. Nat. Cell Biol. 5, 301–308
18 Aranda, V., Haire, T., Nolan, M. E., Calarco, J. P., Rosenberg, A. Z., Fawcett, J. P., Pawson,
T. and Muthuswamy, S. K. (2006) Par6-aPKC uncouples ErbB2 induced disruption of
polarized epithelial organization from proliferation control. Nat. Cell Biol. 8, 1235–1245
19 Brumby, A. M. and Richardson, H. E. (2003) scribble mutants cooperate with oncogenic
Ras or Notch to cause neoplastic overgrowth in Drosophila. EMBO J. 22, 5769–5779
20 Pagliarini, R. A. and Xu, T. (2003) A genetic screen in Drosophila for metastatic behavior.
Science 302, 1227–1231
21 Duran, A., Linares, J. F., Galvez, A. S., Wikenheiser, K., Flores, J. M., Diaz-Meco, M. T.
and Moscat, J. (2008) The signaling adaptor p62 is an important NF-κB mediator in
tumorigenesis. Cancer Cell 13, 343–354
22 Regala, R. P., Davis, R. K., Kunz, A., Khoor, A., Leitges, M. and Fields, A. P. (2009)
Atypical protein kinase C ι is required for bronchioalveolar stem cell expansion and lung
tumorigenesis. Cancer Res. 69, 7603–7611
45 Pears, C. J., Kour, G., House, C., Kemp, B. E. and Parker, P. J. (1990) Mutagenesis of the
pseudosubstrate site of protein kinase C leads to activation. Eur. J. Biochem. 194, 89–94
46 Ishiguro, H., Akimoto, K., Nagashima, Y., Kojima, Y., Sasaki, T., Ishiguro-Imagawa, Y.,
Nakaigawa, N., Ohno, S., Kubota, Y. and Uemura, H. (2009) aPKCλ/ι promotes growth of
prostate cancer cells in an autocrine manner through transcriptional activation of
interleukin-6. Proc. Natl. Acad. Sci. U.S.A. 106, 16369–16374
23 Yang, Y. L., Chu, J. Y., Luo, M. L., Wu, Y. P., Zhang, Y., Feng, Y. B., Shi, Z. Z., Xu, X., Han,
Y. L., Cai, Y. et al. (2008) Amplification of PRKCI, located in 3q26, is associated with
lymph node metastasis in esophageal squamous cell carcinoma. Genes, Chromosomes
Cancer 47, 127–136
47 Horikoshi, Y., Suzuki, A., Yamanaka, T., Sasaki, K., Mizuno, K., Sawada, H., Yonemura, S.
and Ohno, S. (2009) Interaction between PAR-3 and the aPKC–PAR-6 complex is
indispensable for apical domain development of epithelial cells. J. Cell Sci. 122,
1595–1606
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

342
S. Kjær and others
48 Pillai, P., Desai, S., Patel, R., Sajan, M., Farese, R., Ostrov, D. and Acevedo-Duncan, M.
(2011) A novel PKC-ι inhibitor abrogates cell proliferation and induces apoptosis in
neuroblastoma. Int. J. Biochem. Cell Biol. 43, 784–794
49 Eder, A. M., Sui, X., Rosen, D. G., Nolden, L. K., Cheng, K. W., Lahad, J. P., Kango-Singh,
M., Lu, K. H., Warneke, C. L., Atkinson, E. N. et al. (2005) Atypical PKCι contributes to
poor prognosis through loss of apical-basal polarity and cyclin E overexpression in
ovarian cancer. Proc. Natl. Acad. Sci. U.S.A. 102, 12519–12524
50 Zhang, L., Huang, J., Yang, N., Liang, S., Barchetti, A., Giannakakis, A., Cadungog,
M. G., O’Brien-Jenkins, A., Massobrio, M., Roby, K. F. et al. (2006) Integrative
genomic analysis of protein kinase C (PKC) family identifies PKCι as a biomarker
and potential oncogene in ovarian carcinoma. Cancer Res. 66, 4627–4635
51 Laskowski, R. A. and Swindells, M. B. (2011) LigPlot + : multiple ligand–protein
interaction diagrams for drug discovery. J. Chem. Inf. Model. 51,
2778–2786
Received 18 December 2012/12 February 2013; accepted 18 February 2013
Published as BJ Immediate Publication 18 February 2013, doi:10.1042/BJ20121871
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

Biochem. J. (2013) 451, 329–342 (Printed in Great Britain) doi:10.1042/BJ20121871
SUPPLEMENTARY ONLINE DATA
Adenosine-binding motif mimicry and cellular effects of a
thieno[2,3-d]pyrimidine-based chemical inhibitor of atypical protein
kinase C isoenzymes
1
Svend KJÆR*1, Mark LINCH† , Andrew PURKISS*, Brenda KOSTELECKY*†, Phillip P. KNOWLES*, Carine ROSSE†, Philippe
RIOU†, Christelle SOUDY‡, Sarah KAYE‡, Bhavisha PATEL‡, Erika SORIANO*, Judith MURRAY-RUST*, Caroline BARTON‡,
2
2
Christian DILLON‡, Jon ROFFEY‡, Peter J. PARKER†§ and Neil Q. MCDONALD*ꢀ
*Structural Biology, Cancer Research UK, 44 Lincoln’s Inn Fields, London WC2A 3LY, U.K., †Protein Phosphorylation Laboratories, Cancer Research UK, 44 Lincoln’s Inn Fields,
London WC2A 3LY, U.K., ‡Cancer Research Technology Discovery Laboratories, Wolfson Institute for Biomedical Research, University College London, Gower Street, London WC1E
6BT, U.K., §Division of Cancer Studies, King’s College London, New Hunts House, Guy’s Campus, London SE1 1UL, U.K., and ꢀInstitute of Structural and Molecular Biology,
Department of Biological Sciences, Birkbeck College, University of London, Malet Street, London WC1E 7HX, U.K.
d]pyrimidine scaffold of CRT0066854. The equivalent residue in
most other PKC isoforms is alanine, which is unable to form such
a side-chain hydrogen bond. ROCK2 and PRK2 also have alanine
at this position. The exception is PKCγ , which also contains
threonine at this position, as do PKB and PKA. However, PKB
and PKCγ are unresponsive to CRT0066854 at 1 μM inhibitor
concentration in in vitro assays, compared with 77% inhibition
for PKCζ (Table 2 in the main text).
CRT0066854-bound PKCιk adopts an active kinase conformation
Both PKCιk structures bound to ATP (PDB code 3A8W) or
bound to CRT0066854, have similar relative N-lobe and C-lobe
orientations consistent with a closed ‘active’ kinase conformation.
Active kinase conformational elements include (i) an occupied
nucleotide-binding pocket, (ii) an ordered activation loop and (iii)
a salt bridge between the Cα-helix Glu293 and N-lobe Lys274 that
co-ordinates ATP phosphate moieties adjacent to the predicted
divalent metal site. The glycine-rich loop of PKCιk bound to
CRT0066854 adopts a ‘closed’ conformation making multiple
van der Waals contacts with the inhibitor (Figure 2 in the main
Both aPKCs and nPKCs have a leucine residue at position
376, whereas cPKCs have methionine. The higher potency
of BIM-1 towards cPKC isoenzymes may be influenced by a
methionine residue at this position as has been suggested for
PKB [1]. However, the presence of Leu376 alone cannot account
for CRT0066854 potency towards aPKC over nPKCs. Val307 of
aPKCs aligns with threonine in the cPKC and nPKC isoforms
as well as in PKB. The side-chain oxygen of this threonine
residue is involved in binding BIM-1, whereas such an equivalent
interaction is not possible for the aPKC isoenzymes. Val307 may
thus select against BIM-1 binding [2]. The equivalent to the
gatekeeper residue Ile323 is methionine in all cPKC and nPKC
isoforms as well as in PKB, PKA, PRK2 and ROCK2. The Ile323
side chain forms hydrophobic contacts with CRT0066854 and,
although the hydrophobic methionine would add some additional
bulk, its presence in PRK2 and ROCK2 suggests that it does not
perturb CRT0066854 binding. Ile251 may also play a role in drug
selectivity. Isoleucine at this position is unique to the aPKCs,
and ROCK2 with the equivalent residue being leucine in all other
PKCs and many AGC kinases. The slight change induced by
the isoleucine to leucine switch may also help to determine the
difference in BIM-1 sensitivities for the respective isoforms. The
AGC kinases PRK2 and ROCK2 are both inhibited to a significant
extent by CRT0066854. Each has an Asn-Phe-Asp motif and
therefore may provide a positive selection for these kinases to
bind CRT0066854 in a similar manner to PKCι. We conclude
that rather than depending on a single individual side chain, a
combination of several residues is responsible for the isoform
specificity of CRT0066854 for aPKCs by both selecting for this
inhibitor, as well as selecting against CRT0066854 in other AGC
kinases.
text). The threonine residues of the activation loop (Thr403
)
and the turn motif (Thr555) are phosphorylated in both PKCιk
structures. Comparison with other PKC kinase domain structures
shows considerable variation in the precise position of the turn
motif, indicating some plasticity within this poorly conserved
region. The hydrophobic motif (residues 570–578) is well ordered
and are bound to the hydrophobic cleft of the N-lobe as
expected.
Structural comparisons of CRT0066854 and BIM-1 bound to the
PKCι kinase domain
We compared contact residues from the CRT0066854 and BIM-
1 complexes to further characterize the selectivity of the aPKC-
selectiveinhibitor. This analysis does notconsider thecontribution
of bound solvent between inhibitor and protein as few ordered
solvent molecules are observed for these structures given their
modest resolution. Supplementary Table S2 shows the contact
residues together with their structural equivalents in other classes
of PKC isoenzymes, the AGC kinases PKA (protein kinase A),
PKB (protein kinase B) as well as PRK2 and ROCK (both
inhibited by CRT0066854, Table 2 in the main text). Of the PKCι
side chains involved in CRT0066854 binding, BIM-1 binding
or both, five residues stand out: Ile251, Val307, Ile323, Leu376 and
Thr386 (Figure 2B of the main text). Thr386 is a good candidate
for conferring CRT0066854 sensitivity on aPKCs owing to its
side-chain hydrogen bond to a nitrogen atom of the thieno[2,3-
1
These authors contributed equally to this work.
Correspondence may be addressed to either of these authors (email neil.mcdonald@cancer.org.uk or peter.parker@cancer.org.uk).
2
The structural co-ordinates for PKCιk bound to CRT0066854 will appear in the PDB under the accession number 3ZH8.
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

S. Kjær and others
Table S1 Selected PKC kinase domain kinetic constants
AL/TM/HM, activation loop, turn motif and hydrophobic motif. Values are presented as
+
means S.E.M.
−
Kinase domain phosphorylation
kcat (min− 1
)
Km (μM)
+
+
PKCεk AL/TM/HM (pT566/pT710/pS729)
PKCιk AL/TM/HM* (pT403/pT555/E574)
PKCζk AL/TM/HM* (pT410/pT560/E579)
147.6 4.8
199.0 3.3
28.8 2.8
65.4 2.6
−
−
+
+
−
−
+
+
178.6 1.6
25.9 0.7
−
−
*The HM phospho-acceptor is a glutamate residue.
Table S2 Inhibitor contacts with residues within the PKCιk nucleotide-
binding cleft and comparisons with other AGC kinases
Figure S1 Inhibition of cPKCs and nPKCs in cells by use of cPKC inhibitors
can be detected biochemically
Structurally equivalent residues contacting inhibitors in the nucleotide-binding sites of PKCι,
PKCβII (a cPKC) and a nPKC are shown. Also included are the equivalent residues in four
other AGC kinases, two of which (ROCK2 and PRK2) are also inhibited by CRT0066854 (see
Table 2 in the main text). Residues in bold indicate the residue was involved in binding the
inhibitor indicated in the row header. Residues in italics indicate the equivalent residues to those
shown in bold for the indicated proteins. Asterisks indicate functional residues conserved in all
kinases. Underlined residues indicate those residues proposed to be important for BIM-1 and
CRT0066854 binding specificity. PKA, protein kinase A; PKB, protein kinase B.
Expression vectors encoding GFP–PKC isoforms were transfected into HCT-116 cells. Cells
were treated with chemical inhibitors (BIM-1 and Go¨6983) for 2 h before lysis. Cell lysates were
immunoprecipitated with GFP–TRAP and immunoblotted (IB) with the indicated antibodies.
PKCι
CRT0066854
PKCι
BIM-1
PKCβII
miBIM-1
nPKC
ROCK2
PRK2
PKA
PKB
Ile251
Ile251
Gly252
Gly254
Val259
Ala272
Lys274
Val307
Leu348
Gly349
Gly351
Leu
Gly
Gly
Val
Ala
Lys
Thr
Met
Tyr/Phe
Val
Gly
Asp
Asp
Leu
Ala
Thr
Ile
Leu
Gly
Gly
Val
Ala
Lys
Val
Met
Tyr
Ala
Gly
Asp
Asp
Leu
Ala
Asp
Leu
Gly
Gly
Val
Ala
Lys
Val
Met
Tyr
Val
Gly
Glu
Glu
Leu
Thr
Asp
Leu
Gly
Gly
Val
Ala
Lys
Thr
Met
Tyr
Ala
Gly
Glu
Glu
Met
Thr
Asp
Gly252
*
*
Gly
Gly
Val
Ala
Lys
Val
Met
Tyr
Met
Gly
Asp
Asp
Leu
Ala
Asp
Gly254
Val259
Ala272
Val356
Ala369
Lys371
Thr404
Met420
Tyr422
Lys274
Val307
*
Ile323
Tyr325
Val326
Gly329
Ile323
Tyr325
Val326
Gly329
Asp330
Asp373
Leu376
Thr386
Asp387
Val423
Gly426
Asp330
Asp373
Leu376
Thr386
Asp387
Asp427
Asp470
Met473
Ala483
Asp484
*
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society

A novel small molecule aPKC inhibitor
Figure S2 Structure of PKCι bound to CRT0066854 compared with PKCι bound to ATP and PKA bound to ATP
Overall fold of PKCιk in complex with CRT0066854 (left-hand panel, cyan cartoon and yellow sticks), and ATP (middle panel, PDB code 3A8W, salmon-pink cartoon and sticks), with the PKA complex
of ATP (right-hand panel, PDB code 1ATP, green cartoon and sticks) for comparison.
REFERENCES
1
Gassel, M., Breitenlechner, C. B., Konig, N., Huber, R., Engh, R. A. and Bossemeyer, D.
(2004) The protein kinase C inhibitor bisindolyl maleimide 2 binds with reversed
orientations to different conformations of protein kinase A. J. Biol. Chem. 279,
23679–23690
2
Messerschmidt, A., Macieira, S., Velarde, M., Badeker, M., Benda, C., Jestel, A.,
Brandstetter, H., Neuefeind, T. and Blaesse, M. (2005) Crystal structure of the catalytic
domain of human atypical protein kinase C-ι reveals interaction mode of phosphorylation
site in turn motif. J. Mol. Biol. 352, 918–931
Received 18 December 2012/12 February 2013; accepted 18 February 2013
Published as BJ Immediate Publication 18 February 2013, doi:10.1042/BJ20121871
ꢁ
c
ꢁ
c
The Authors Journal compilation 2013 Biochemical Society