10304 Biochemistry 2002, 41, 10304-10314

Toward a PKB Inhibitor: Modification of a Selective PKA Inhibitor by Rational

Design†
Hadas Reuveni,‡,§ Nurit Livnah,*,‡,# Tamar Geiger,§ Shoshana Klein,§ Osnat Ohne,# Ilana Cohen,# Moran Benhar,§
Gary Gellerman,# and Alexander Levitzki*,§
Department of Biological Chemistry, The SilVerman Institute of Life Sciences, The Hebrew UniVersity of Jerusalem,
Jerusalem, Israel, and Peptor Ltd., RehoVot, Israel
ReceiVed March 29, 2002

ABSTRACT: Protein kinase B/Akt (PKB) is an anti-apoptotic protein kinase that has strongly elevated activity
in human malignancies. We therefore initiated a program to develop PKB inhibitors, “Aktstatins”. We
screened about 500 compounds for PKB inhibitors, using a radioactive assay and an ELISA assay that we
established for this purpose. These compounds were produced as combinatorial libraries, designed using
the structure of the selective PKA inhibitor H-89 as a starting point. We have identified a successful lead
compound, which inhibits PKB activity in vitro and in cells overexpressing active PKB. The new compound
shows reversed selectivity to H-89: In contrast to H-89, which inhibits PKA 70 times better than PKB,
the new compound, NL-71-101, inhibits PKB 2.4-fold better than PKA. The new compound, but not
H-89, induces apoptosis in tumor cells in which PKB is amplified. We have identified structural features
in NL-71-101 that are significant for the specificity and that can be used for future development and
optimization of PKB inhibitors.

Protein kinase B/Akt (PKB)1 is an anti-apoptotic protein PKBR was found to be amplified or overexpressed in gastric
kinase that has elevated activity in a number of human adenocarcinomas and in a breast cancer cell line (4, 5). PKBâ
malignancies. PKB was originally discovered as a viral is amplified or overexpressed in 3% of breast (6), 12% of
oncogene, v-Akt, in mouse T-cell lymphoma (1). It was later pancreatic (7), and 15% of ovarian cancers (6, 8). PKBγ is
established that v-Akt is the oncogenic version of the cellular overexpressed in estrogen receptor-deficient breast cancer
enzyme, PKBR/c-Akt1 (2). Sequencing of PKBR revealed and in androgen-independent prostate cell lines (9).
homology within the kinase domains to the PKA (∼68%) The mechanism of activation of PKB has been elucidated
and PKC isozymes (∼73%) (3), a fact that lent to its (reviewed in ref 10-12). The second messenger PIP3, which
renaming as PKB. There are three cellular isoforms of PKB. is produced by PI′-3-kinase, binds to the pleckstrin homology
(PH) domain of PKB and recruits the protein to the
† We would like to acknowledge CapCure for their partial support membrane, where it is phosphorylated on residues Thr308
of this study. and Ser473 and converted to its activated form. The enzyme
* To whom correspondence should be addressed: Fax: 972-2- PDK1 phosphorylates residue Thr308 of PKB in a PIP3-
651 2958. Telephone: 972-2-658 5404. E-mail: levitzki@vms.huji.ac.il. dependent manner (13). The mechanism of phosphorylation
Address: Prof. A. Levitzki, Department of Biological Chemistry, The
Silverman Institute of Life Sciences, The Hebrew University of on Ser473 is controversial, and several options have been
Jerusalem, Jerusalem, 91904 Israel. Additional addressee for cor- proposed. These include autophosphorylation (14), phos-
respondence: Fax: 972-2-940 7737. Telephone: 972-8-938 7774. phorylation by an as yet undescribed kinase dubbed PDK2
E-mail: nurit@peptor.co.il. Address: Dr. Nurit Livnah, Peptor Ltd.,
Kiryat Weizmann, Rehovot, 96326 Israel.
(15), phosphorylation by PDK1 converted to PDK2 by
‡ These authors contributed equally to this study. binding to a specific peptide (16), and phosphorylation by
§ The Hebrew University of Jerusalem.
ILK (17). Recently, it has been shown that phosphorylation
# Peptor Ltd.
1 Abbreviations: PKB (protein kinase B/Akt); PKA (protein kinase
of tyrosine residues, possibly by pp60c-Src, is also required
A/cAMP-dependent protein kinase); PKC (protein kinase C); PKG to activate PKB (18).
(protein kinase G/cGMP-dependent protein kinase); NMP (N-methyl Treatment of cells with many growth factors, such as
pyrollidine); PyBroP (bromo-tris-pyrrolydino-phosphonium hexafluoro IGF-1 (15, 19, 20) and PDGF (21), leads to activation of
phosphate); DIEA (diisopropylethylamine); DCM (dichloromethane);
TFA (trifluoroacetic acid); TIS (triisopropylsilane); DMF (dimethyl
PKB. Activation of pp60c-Src (22, 23) also leads to the
formamide); TLC (thin-layer chromatography); MPLC (medium- activation of PKB. Persistent activation of PKB also occurs
pressure liquid chromatography); DMSO (dimethyl sulfoxide); MS as a result of a deletion in the gene encoding the tumor
(mass spectrometry); HPLC (high-pressure liquid chromatography); HA suppressor PTEN, a negative regulator of PKB (24). PTEN
(hemaglutinin); KD-PKB((myristoylated kinase dead PKB); DMEM
(Dulbecco’s modified Eagle medium); FCS (fetal calf serum); IGF-1 deletion is the hallmark of many advanced metastatic cancers.
(insulin-like growth factor-1); TBST (10 mM Tris-HCl, pH 7.5, 170 The overexpression of PKB in ovarian, breast, and prostate
mM NaCl, and 0.2% Tween-20); ABTS (2,2′-azino-bis(3-ethylbenz- cancers is associated with a poor prognosis (6, 9). These
thiazoline-6-sulfonic acid); GSK (glycogen synthase kinase); ELISA findings establish the central role that PKB plays in onco-
(enzyme linked immunosorbence assay); ECL (enhanced chemilumi-
nescence); FACS (fluorescence activated cell sorter); PI (propidium genesis and imply that inhibitors of this protein kinase could
iodide); PBS (phosphate buffered saline). be powerful anticancer agents.
10.1021/bi0202530 CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/19/2002
Modification of a Selective PKA Inhibitor Biochemistry, Vol. 41, No. 32, 2002 10305

In this paper, we describe a rational approach to design MS m/z (%) ) 216 (M+, 100%), 1H NMR (CDCl3): δ
novel PKB kinase inhibitors, Aktstatins, based on combi- 2.76 (t, J ) 6 Hz, 2H, CH2), 3.36 (t, J ) 6 Hz, 2H, CH2),
natorial and parallel chemistry methodologies. Such meth- 7.57 (d, J ) 7 Hz, H-3), 7.68 (t, J ) 7 Hz, 1H), 7.88 (t, J
odologies have been extensively used in the past few years ) 7 Hz 1H), 8.05 (d, J ) 7 Hz, 1H, H-5), 8.19 (d, J ) 7
to derive structure-activity relationships (SAR) from series Hz, 1H-8), 8.78 (bs, NH), 8.93 (s, 1 H, H-2).
and libraries of compounds, and facilitate discovery and (iii) N-(cinnoline-4-carboxyamide)-ethane amine: yield
optimization of hits and leads (25, 26). These methods allow (98%).
the production of large numbers of analogues simultaneously, MS m/z (%) ) 217 (M+, 100%), 1H NMR (CDCl3): δ
with a variety of modifications resulting in a diverse set of 2.87 (t, J ) 6 Hz, 2H, CH2), 3.50 (t, J ) 6 Hz, 2H, CH2),
compounds based on a common structural motif. We 7.86 (t, J ) 7 Hz, 1H), 8.07 (t, J ) 7 Hz, 1H), 8.15 (d, J )
screened a series of commercially available serine/threonine 7 Hz, 1H, H-5), 8.31 (d, J ) 7 Hz, 1H-8), 9.02 (bs, NH),
kinase inhibitors for their effects on PKB activity. Of these 9.32 (s, 1 H, H-3).
compounds, H-89 was chosen as a starting point for (iv) N-(piperonylamide)-ethane amine: yield (92%).
development of a PKB inhibitor, for reasons of selectivity, MS m/z (%) ) 209 (M+, 100%), 1H NMR (CDCl3): δ
potency, and feasibility of chemical modification. We 2.99 (t, J ) 6 Hz, 2H, CH2), 3.48 (t, J ) 6 Hz, 2H, CH2),
employed parallel synthesis to examine the effects of 6.09 (s, 1 H, O-CH2-O). 7.02 (d, J ) 7 Hz, H-6), 7.40 (s,
chemical and structural modifications to the various diversity 1H-3), 7.48 (d, J ) 7 Hz, 1H-5), 8.45 (bs, NH).
zones of H-89, a commercially available PKA inhibitor, on (v) N-(isoquinoline-5-carboxyamide)-ethane (N′-methyl)-
the affinity and specificity of the compound analogues. To amine: yield (89%).
facilitate rapid screening for PKB inhibitors, we designed a MS m/z (%) ) 280 (M+, 100%), 1H NMR (CDCl3): δ
nonradioactive ELISA assay of PKB. We report here on our 2.97 (m, 6H, 2CH3), 3.42 (t, J ) 6 Hz, 2H, CH2), 3.58 (t, J
initial success in generating a prospective lead compound ) 6 Hz, 2H, CH2), 7.91 (t, J ) 7 Hz, H-8), 8.38 (d, J ) 7
for an effective PKB inhibitor. The new compound, NL-71- Hz, 1H, H-8), 8.47 (d, J ) 7 Hz, 1H, H-6), 8.55 (d, J ) 7
101, inhibited PKB activity in intact cells and led to apoptosis Hz, 1H, H-4), 8.75 (d, 1 H, H-3), 9.59 (s, 1 H, H-1).
of an ovarian cancer cell line in which PKB is amplified. (vi) N-(nicotinamide)-ethane amine: yield (91%).
MS m/z (%) ) 166 (M+, 100%), 1H NMR (CDCl3): δ
EXPERIMENTAL PROCEDURES 2.96 (t, J ) 6 Hz, 2H, CH2), 3.52 (t, J ) 6 Hz, 2H, CH2),
7.48 (t, J ) 7 Hz, H-5), 8.10 (d, J ) 7 Hz, 1H, H-4), 8.59,
General procedure for the synthesis of monosubstituted
(d, J ) 7 Hz, 1H, H-6), 8.77 (s, 1 H, H-1).
ethylenediamines on solid phase support:
General Procedure for the Synthesis of Monosubstituted
One gram of 1,2-diaminoethane trityl resin (0.35 mmol/ Ethylenediamines in Solution. Sulfonic acid was placed in a
g; NOVA Biochem Ltd.) was swollen for 18 h in a reaction round-bottomed flask and thionyl chloride (15 equiv) was
vessel equipped with a sintered glass bottom, and placed on added, followed by a drop of DMF. The mixture was refluxed
a shaker. for 2 h and then cooled and evaporated to dryness. The
After the sample was washed with NMP (2 times, 8 mL powdered residue was transferred in small portions to a
for 2 min), the appropriate acid (3 equiv) was preactivated cooled (0°) flask containing 10 equiv of amine in methylene
in NMP (8 mL) for 3 min using PyBroP (equiv) and DIEA chloride. The mixture was stirred at room temperature for 4
(6 equiv). The activated acid was transferred to a reaction h. The progress of the reaction was monitored by TLC using
vessel and shaken for 1.5 h at room temperature. Reaction 25% methanol in chloroform. The mixture was evaporated
completion was monitored by Kaiser test, and the solvent and purified either by flash chromatography on silica with a
was removed by filtration. The resin was washed with NMP gradient of 10-30% methanol in chloroform, or by MPLC
(5 times, 8 mL for 2 min) and DCM (2 times, 8 mL for 2 using an RP-C18 column.
min), dried, and submitted to cleavage. (i) N-(isoquinoline-5-sulfonamide)-ethaneamine: yield
Piperonyl chloride was reacted directly with the resin in 35%.
DCM (10 mL) in the presence of DIEA (6 equiv) for 1 h at MS m/z (%) ) 251 (M+, 100%), 1H NMR (MeOD): δ
room temperature. The workup was carried out by the same 2.64 (t, J ) 6 Hz, 2H, CH2), 2.92 (t, J ) 6 Hz, 2H, CH2),
mode as for acids. 7.81 (t, J ) 7 Hz, 1H, H-7), 8.36 (d, J ) 7 Hz, 1H), 8.43 (d,
CleaVage. The dry peptidyl-resin was treated with a cold J ) 7 Hz 1H), 8.53 (d, J ) 7 Hz, 1H), 8.60 (d, J ) 8 Hz,
mixture of 5% TFA and 1% TIS in 10 mL of DCM. The 1H), 9.36 (s, 1 H, H-1).
solvent was then removed by a stream of nitrogen, and the (ii) N-DANSYL-ethaneamine: yield 48%
residue was taken into diethyl ether. After centrifugation and MS m/z (%) ) 294.8 (M+, 100%), 1H NMR (MeOD): δ
removal of the solvent, the crude product was obtained. This 2.53 (t, J ) 6 Hz, 2H, CH2), 2.85 (t, J ) 6 Hz, 2H, CH2),
was submitted to the next step without any further purifica- 2.85 (s, 6H, N-dimethyl) 7.24 (t, J ) 8 Hz, 1H), 8.56 (m,
tion. 2H), 8.16 (d, J ) 8 Hz 1H), 8.53 (d, J ) 8 Hz, 1H).
(i) N-(isoquinoline-1-carboxyamide)-ethane amine: yield (iii) N-(isoquinoline-8-sulfonyl)piperazine: yield 52%.
(93%). MS m/z (%) ) 278.8 (M+, 100%), 1H NMR (MeOD): δ
MS m/z (%) ) 216 (M+, 100%), 1H NMR (CDCl3): δ 2.81 (m, 4H,), 3.11 (m, 4H), 7.87 (t, J ) 8 Hz, 1H), 8.46 (d,
3.27 (t, J ) 6 Hz, 2H, CH2), 3.84 (t, J ) 6 Hz, 2H, CH2), J ) 8H 2H), 8.62 (s, 2H), 9.40 (s, 1H).
7.18 (t, J ) 7 Hz, 1H), 8.13 (t, J ) 7 Hz, 1H), 8.23 (d, J ) (iv) N-(isoquinoline-8-sulfonyl)-1,2,diaminomethylcyclo-
7 Hz 1H, H-5), 8.45 (m, 3H-3,4,8). hexane: yield 55%.
(ii) N-(quinoline-4-carboxyamide)-ethane amine: yield MS m/z (%) ) 305.8 (M+, 100%), 1H NMR (MeOD): δ
(95%). 1.15 (m, 4H), 1.43 (m, 4H), 2.65 (m, 1H, CH), 3.32 (m, 1H,
10306 Biochemistry, Vol. 41, No. 32, 2002 Reuveni et al.

CH), 7.796 (t, J ) 8H 1H), 8.41 (d, J ) 8 Hz 1H), 8.49 (m, Tris-HCl, pH 7.5, 200 mM imidazole, 100 mM KCl, 10%
2H), 8.61 (s, 1H), 9.37 (s, 1H). glycerol). His-PKB activity was determined as described
(v) N-(isoquinoline-8-sulfonamide)-propanamine: yield below. One unit of PKB activity is the amount of enzyme
52%. that catalyzes the phosphorylation of 1 nmol of RPRTSSF
MS m/z (%) ) 266.8 (M+, 100%), 1H NMR (MeOD): δ peptide in 1 min in the presence of 100 µM ATP. The
1.57 (dt, J ) 7 Hz, 1H), 1.66 (dt, J ) 7 Hz, 1H), 2.59 (t, J specific activity of the enzyme produced was ∼35 units/mg
) 7H, 1H), 2.96 (m, 2H), 3.05 (t, J ) 7H, 1H), 7.81 (t, of His-PKB.
H)8H, 1H), 7.89 (s, 1H, H-amide), 8.38 (d, J ) 8 Hz 1H), In Vitro Kinase Assays. (a) RadioactiVe Assay of PKB
8.46 (d, J ) 8 Hz, 1H) 8.56 (m, 2H), 9.36 (s, 1H). ActiVity. PKB activity was assayed on a 7-mer peptide
General Procedure for ReductiVe Amination Using Paral- RPRTSSF, which is derived from the Crosstide peptide (30).
lel Synthesis in a 96-Well Format. Stock solutions of amines We verified that the peptide showed the same values of Km
and aldehydes were prepared as 0.5 M in methanol or and Vmax for PKB as the parent Crosstide (20). In our hands,
dimethyl sulfoxide (DMSO) depending on solubility. A total the Km for the 7-mer or for Crosstide was 18 µM, the Km
of 1.1 equiv of amine was pipetted into the wells, followed for ATP was 40 µM, and the reaction was linear at 15 min.
by 1 equiv of each of the various aldehydes, with accurate The reaction conditions for in vitro assay of PKB inhibition
documentation of aldehyde ID and well position in the plate. were adapted from ref 30. The reaction mix contained 50
The plate was shaken for 5 min. Then, 2 equiv of NaBH4- mM Tris-HCl, pH 7.5, 0.1 mM EGTA, 0.1% (v/v) 2-mer-
CN were added as a freshly made 0.5 M solution in captoethanol, 1 µM PKI, 10 mM magnesium acetate, the
methanol, immediately followed by 0.1 equiv of acetic acid. potential inhibitory compound, 20 µM RPRTSSF peptide,
The plate was shaken for 6 h. At the end of the reaction, the 10 µM [γ32P]ATP (1 µCi/assay), and 0.006 units of His-
solvent was evaporated to dryness using a speedvac, and PKB. The assay was carried out in a volume of 50 µL, for
reconstituted in neat DMSO, to a final concentration of 200 15 min at 30 °C, in 96-well plates. The reaction was
mM. Identity and homogeneity of the products were evalu- terminated by addition of EGTA to a final concentration of
ated by direct MS of all wells and HPLC of a representative 50 mM, and the reaction mixtures were spotted onto 6.25
sample of the wells. cm2 phosphocellulose strips (P81 Whatman). The strips were
Plasmids. (a) His6-HA-Tagged PKBR. The N-terminal washed six times with 75 mM phosphoric acid (10 mL/
segment (EcoRI to NotI) of hemaglutinin (HA)-tagged sample) and once with acetone, air-dried, and monitored
PKBR, in plasmid pCMV5-HA-PKBR (15), was replaced using a â-scintillation counter. The libraries of the crude
with a PCR fragment encoding a His6-tag before the HA inhibitory compounds were screened at a single concentration
tag. The 3′ primer for incorporation of the His6-tag was for each compound (20 µM), while purified compounds were
5′CAGAATTCGCCACCATGCATCATCATCATCATCAT- assayed at various concentrations, and their IC50 values were
GCAGCATACCCATATGATGTGCCAG. determined using the regression program.
(b) Myristoylated/Kinase Dead PKB (KD-PKB). The (b) ELISA Assay of PKB ActiVity. We established an
C-terminal segment (NotI to BamHI) of pCMV5-m/ ELISA assay of PKB activity for high throughput purposes.
p-HA-PKBR (27) was replaced with the corresponding 40 pmol/well of biotinylated Crosstide [K(biotin)GRPRTSS-
fragment from the kinase-dead construct, pCMV5-HA- FAEG; Sigma Israel Ltd.], dissolved in PBS, was incubated
PKBRK179A (15). for 30 min at room temperature in streptavidin-coated 96-
(c) GSK3. Plasmid pCMV5-EE-GSK3â was described by well plates (Steffens Biotechnische Analyzen GmbH, biotin
Shaw et al. (28). binding capacity 18 pmol/well). Following coupling, the
Cell Culture and Transfection. The human embryonic excess peptide was washed out. Reactions were carried out
kidney cell line, HEK293, the human ovarian epithelial in a final volume of 50 µL/well. First, 0.006 units of His-
cancer cell line, OVCAR-3 (8), and mouse NIH3T3 cells PKB, in 50 mM Tris-HCl, pH 7.5, 0.1 mM EGTA, 0.1%
were cultured at 37 °C in Dulbecco’s modified Eagle medium (v/v) 2-mercaptoethanol were placed in each well, except
(DMEM) supplemented with 10% FCS (Bet-Haemek, Israel). the blanks. Second, the inhibitory compounds were added
All treatments described in this report were performed in to a final concentration of 20 µM. Finally, the reaction was
DMEM containing 10% FCS. HEK293 and NIH3T3 cells started by the addition of the assay mixture (final concentra-
were transiently transfected using polyethyleneimine (29). tions: 50 mM Tris-HCl, pH 7.5, 0.1 mM EGTA, 0.1% (v/
PKB Preparation. HEK293 cells were grown on 10-cm v) 2-mercaptoethanol, 1 µM PKI, 10 mM magnesium acetate,
Petri plates and transfected with pCMV5-His6-HA-PKBR. and 10 µM ATP). Following agitation for 15 min at 30 °C,
Twenty-four hours after transfection, cells were starved in the reactions were stopped by addition of EDTA to a final
serum-free medium for 20 h, stimulated with 50 ng/mL concentration of 200 mM. Alternately, biotinylated Crosstide
IGF-1 (Sigma) for 10 min, and lysed in 0.7 mL of cold lysis (1 µM) was included in the reaction mix, the reactions were
buffer (50 mM Tris-HCl pH 7.5, 1% (v/v) Triton X-100, performed in 96-well plates with conical wells (Sterilin), and
0.1% (v/v) 2-mercaptoethanol, 1 mM sodium orthovanadate, after the reactions were terminated the contents of each well
20 mM sodium â-glycerophosphate, 50 mM sodium fluoride, were transferred to the 96-well streptavidin plates. The plates
5 mM sodium pyrophosphate, 1 µM microcystin-LR, 1 mM were incubated for 30 min at room temperature, to allow
benzamidine, 0.2 mM AEBSF, 10 µg/mL leupeptin, 1 µg/ coupling. The alternate protocol gave a higher signal/noise
mL aprotonin, and 10 µg/mL soybean trypsin inhibitor). The ratio. The plates were washed four times with TBST (10
lysate was centrifuged at 4 °C for 15 min at 19000g, and mM Tris-HCl, pH 7.5, 170 mM NaCl, and 0.2% Tween-20)
the supernatant was loaded onto nickel beads (Ni-NTA and blocked for 20 min at room temperature in blocking
agarose, Qiagen). The column was washed with buffer A solution (1:20 dilution of low-fat milk in TBST). Phospho-
(20 mM Tris-HCl, pH 7.5, 1 M KCl, 20 mM imidazole, 10% rylated Crosstide was detected by anti-phospho(Ser21)GSK3
glycerol) and His-PKB was eluted with buffer B (20 mM (obtained from New-England BioLabs; dilution 1:1000 in
Modification of a Selective PKA Inhibitor Biochemistry, Vol. 41, No. 32, 2002 10307
5% BSA in TBST, 45 min at RT). The wells were washed
four times with TBST and incubated with peroxidase-
conjugated anti-Rabbit IgG (from Jackson; dilution 1:10 000
in blocking solution) for 30 min at room temperature.
Following four washes with TBST and one wash with TBS
(10 mM Tris-HCl, pH 7.5, 170 mM NaCl), a color reaction
was initiated by addition of a solution containing 0.5 mg/
mL ABTS, 0.013% H2O2, 38 mM sodium hydrogen phos- FIGURE 1: Structure of H-89, showing the division into diversity
phate, and 31 mM citric acid. Optical densities were domains. Libraries were designed to explore the contribution of
determined, after about 20 min, using an ELISA-Reader. each domain to potency and specificity.
Inhibitor libraries were screened either by the radioactive
assay or by ELISA. H-89 was included in each assay as a reaction mixtures were separated by SDS-PAGE. The
positive control, at a concentration of 3 µM. amount of phosphorylated substrate was quantified using a
(c) RadioactiVe Assay of PKA ActiVity. PKA kinase phosphor-imager (Fujifilm, FLA 300).
inhibition was assayed as described by Roskoski (31). Immunoblotting. HEK293 cells were treated with inhibitors
Briefly, the assay contained 50 mM MOPS, pH 7.0, 10 mM 48 h following transfection with pCMV5-EE-GSK3. After
magnesium acetate, 0.2 mg/mL bovine serum albumin, 10 4 h incubation with the inhibitor, cells were lysed in boiling
µM [γ32P]ATP (1 µCi/assay), 20 µM LRRASLG (Kemptide), Laemmli sample buffer. Aliquots of cell extracts containing
and the inhibitory compound. The Km for ATP is 25 µM. equal amounts of protein were resolved by 10% SDS-PAGE
The reaction was started by the addition of 0.4 units/assay and electroblotted onto nitrocellulose filters. The membranes
of PKA (Promega Corp.). One unit of PKA was defined by were blocked with blocking solution, incubated with primary
the company as the amount of enzyme required to incorporate antibodies diluted in 5% BSA in TBST overnight at 4 °C,
1 pmol of phosphate into casein in 1 min. The reaction was and then with horseradish peroxidase-conjugated secondary
carried out in a volume of 50 µL/well in 96-well plates, at antibodies diluted in blocking buffer for 75 min at room
30 °C for 10 min, and stopped by addition of EGTA to a temperature. Immunoreactive bands were visualized using
final concentration of 100 mM. Twenty microliter aliquots ECL, and quantitated using the NIH-Image program.
of the assay mixture were spotted onto 2 cm2 phosphocel- Antibodies against phospho-Akt (Thr308), phospho-Akt
lulose strips. Washing and monitoring were conducted as (Ser473), and phospho-GSK3R (Ser21) were obtained from
described for PKB. The inhibitor libraries were screened in New England BioLabs. Antibodies against GSK3â were from
duplicate, using a single concentration of the inhibitor (1 or Transduction Laboratories. Anti-Akt1/2 (H-136) was from
5 µM, depending on the compound). Purified compounds Santa Cruz Biotechnology.
were assayed at various concentrations, and their IC50 values Apoptosis Assays. OVCAR-3 cells were seeded in 10-cm
were determined using the regression program. Petri plates (2 × 106 cells/plate) in DMEM containing 10%
(d) RadioactiVe Assay of PKC ActiVity. PKC was obtained FCS. Twenty-four hours later, NL-71-101, NL-71-101B, or
from Promega Corp. and assayed using a kit from the same H-89 were added to various concentrations. Following 40 h
manufacturer, in the presence and absence of phospholipids. of treatment, cells were harvested, washed twice with PBS,
The activity of PKC was determined by subtracting the and stained with Annexin-V-FLOUS + PI according to the
activity in the absence of phospholipids from that in the manufacturer’s instructions (Boehringer Mannheim). Cells
presence of phospholipids. The concentration of ATP in the were counted with a FACS.
assay was 10 µM (Km for ATP ) 50 µM). For DNA fragmentation analysis, cells were incubated with
(e) RadioactiVe Assay of PKG ActiVity. PKG was obtained PI solution (50 µg/mL PI, 0.1% sodium citrate, 0.1% Triton
from Promega Corp., and assayed according to the manu- X-100, 100 µg/mL RNAse, in PBS), and the proportion of
facturer’s instructions, in 40 mM Tris-HCl, pH 7.4, 20 mM the population in the sub-G1 phase was quantified by FACS
MgCl2, 200 µM [γ32P]ATP (1 µCi/assay) using 113 µg/mL analysis.
RKISASEF substrate, also from Promega Corp., in the
RESULTS
presence of 3 µM cGMP and inhibitor. PKG was assayed at
200 µM ATP, following the manufacturer’s instructions, H-89 as an Initial Lead Compound. Since there are no
whereas the other kinases were assayed at 10 µM ATP. Thus, known inhibitors of PKB, we screened a variety of com-
the experimental IC50 values obtained for the inhibition of mercial Ser/Thr protein kinase inhibitors, in an attempt to
PKG were corrected to values that correspond to those identify structural motifs that would assist in library design
expected in an assay that would contain 10 µM ATP (Km (data not shown). Several molecules inhibited PKB in the
for ATP ) 57 µM). low micromolar range. A combination of affinity and
(f) RadioactiVe Assay of p38 ActiVity. HA-tagged p38 specificity considerations, as well as analysis of the structural
protein was immunoprecipitated, with anti-HA antibodies, properties of these compounds, led us to select H-89 as the
from 200 µg of a lysate of UV-irradiated NIH3T3 cells that basis for library design. The structure of H-89 makes it a
had been transiently transfected a day previously with a suitable candidate for SAR study using combinatorial
plasmid encoding HA-p38. The enzyme was divided into chemistry, since it allows diversity in many regions of the
aliquots, which were incubated for 30 min at 30 °C, in the molecule (Figure 1). In the core (A, in Figure 1), the
presence of 10 µM GST-ATF2 as substrate, 20 µM [γ32P] 5-isoquinoline moiety can be replaced with various bicyclic
ATP (5 µCi/reaction), and inhibitor. The reactions were and aromatic residues, and the sulfonamide group can be
terminated by the addition of Laemmli sample buffer (50 replaced with carbonyl and other similar moieties. The
mM Tris-HCl, pH 6.8, 5% 2-mercaptoethanol, 3% sodium ethylenediamine bridge (B, in Figure 1) can vary in length,
dodecyl sulfate, and 0.5 mg/mL bromophenol-blue), and the hydrogen bonding properties, and substitution. Finally, the
10308 Biochemistry, Vol. 41, No. 32, 2002 Reuveni et al.

Table 1: SAR from Screening Various C Moieties with 5-Isoquinoline-8-sulfonamide ethylenediamine

cinnamoyl moiety (C, in Figure 1) can be modified to a large Structure-ActiVity Study. We examined the effects of
variety of structures for the evaluation of the optimal structural and chemical modifications in the diversity regions
properties of this region. of H-89 defined above. In most libraries, starting materials
Modification of a Selective PKA Inhibitor Biochemistry, Vol. 41, No. 32, 2002 10309

FIGURE 2: A, B, and C moieties used in combinatorial arrays for the SAR studies of H-89 analogues.

were prepared from sulfonic or carboxylic acids reacted with compounds with similar structures were synthesized later to
diamines to produce sulfonamides or amides, respectively. confirm these observations, and these also proved to be very
These starting materials were subjected to parallel synthesis poor inhibitors (data not shown). At this stage, we concluded
using alkylation or reductive amination, with the appropriate that interaction with a highly hydrophobic feature in region
halides or aldehydes, respectively. The libraries were screened C is very important.
for both PKB and PKA inhibition, to assess the contribution Next, we synthesized a small library to evaluate the
of each structural modification to the inhibition of each chemistry of the bond connecting regions C and B. An amine
enzyme. A simultaneous SAR study for both PKA and PKB bond (as in H-89) and an amide bond were synthesized and
was necessary, since the initial lead compound, H-89, is a compared, for three different bicyclic A cores and three
PKA antagonist, and the SAR study was directed mainly at different C moieties (data not shown). This series clearly
identifying the specific enzyme-ligand interactions that are showed that an amide bond is less favorable for the inhibition
not common to both PKA and PKB. Such interactions could of PKB than an amine. We also synthesized a small library
be modulated by structural features that either adversely with a urea bond connecting these two regions. Again,
affect the interaction with PKA (“irritants”) without signifi- significantly lower inhibition was observed by these com-
cantly affecting the interaction with PKB, or selectively pounds than by their amine analogues.
improve binding to PKB. On the basis of these results, we designed several libraries
H-89 analogues for SAR study were synthesized in parallel with an amine connecting part C and B, and a bulky
in a 96-well format, using solution chemistry. In the initial hydrophobic residue in part C. The libraries were composed
tests, the crude compounds (∼80-85% pure) were tested at of combinatorial arrays of the A, B, and C moieties depicted
20 µM in the assays of PKB activity, and at 1 or 5 µM in Figure 2. SAR analysis of these libraries showed that
(depending on the compound; see Table 1) in the PKA changes in the structures of A or B adversely affected the
assays. In every assay, H-89 itself was included as a reference inhibition of PKB (data not shown). On domain A, any
for comparison. replacement of the 5-isoquinoline-sulfonamide moiety, in-
In preliminary studies, we built a small library based on cluding various other quinolines, abolished the inhibition.
aliphatic residues in C to examine the necessity for the These results imply that the nitrogen at position 5 is
aromatic moiety that H-89 bears in this region. Very low significant for the inhibition. Substitutions on the A core also
inhibition was observed by this library. Several single diminished the inhibition, possibly implying that the binding
10310 Biochemistry, Vol. 41, No. 32, 2002 Reuveni et al.

Table 2: IC50 Values of Purified H-89 Analogues

pocket is very tight at this region. On the bridge B, elongation substitution on the chain. Placement of secondary amine
from two to three carbons diminished activity, as did derivatives in region B adversely affected the inhibition:
Modification of a Selective PKA Inhibitor Biochemistry, Vol. 41, No. 32, 2002 10311
Piperazine or N,N-dimethylethylamine led to significantly
reduced inhibition, whereas homopiperazine eliminated the
inhibitory effect. These results imply that the amino protons
are important for the interaction, possibly through hydrogen
bonding, and that the distance and relative positions of these
two basic centers are crucial. The possibility that only one
of the hydrogen bonds actively participates in the interaction
needs to be further elucidated, by eliminating only one of
the basic centers.
The most significant SAR was observed in the moieties
tested for part C, where for the same A and B, we observed
a significant effect of the nature of C on potency and
specificity (Table 1). The performance of compounds derived
from 5-isoquinoline-sulfonamide-ethylenediamine, in which
various moieties were placed in region C, ranged from almost
no PKB inhibition to inhibition of over 90% at 20 µM.
Significant variation was observed in the effects on PKA as
well. In several wells, the inhibition of PKA was significantly
decreased. Overall, these results indicated that the interaction
with region C is important for both enzymes, and further-
more, that an “irritant” interaction with PKA might exist in
region C, which does not significantly affect the interaction
with PKB. Specifically, the initial results suggested that an
additional substitution at the carbon bearing the aromatic FIGURE 3: Structures and inhibition potency of H-89 and NL-71-
residue in region C is unfavorable for PKA inhibition, while 101 in cell free assays. (A) Structures of H-89 and NL-71-101.
its effect on PKB inhibition is negligible. (B) In vitro inhibition of PKA activity by the inhibitors. (C) In
Several promising compounds were selected for resyn- vitro inhibition of PKB activity by the inhibitors.
thesis, purification, and further characterization, and accurate
IC50 values were evaluated for them. In addition, we Table 3: IC50 Values of NL-71-101 and H-89 for Several Related
Kinasesa
synthesized, purified, and characterized several other com-
pounds with an additional hydrophobic interaction in region IC50 values (µM)
C, to confirm that this position distinguishes the interaction enzyme NL-71-101 H-89
of the compounds between PKB and PKA. The IC50 values PKB 3.7 2.5
obtained in radioactive assays using the purified compounds PKA 9 0.035
are depicted in Table 2. These results confirm that this PKG 36 (10.8) 0.88 (0.29)
position is indeed a PKA “irritant”, shifting its IC50 value PKC 104 15
p38 .100 not determined
from less than 50 nM to several micromolar, while its effect
a
on PKB is minor. Values in parentheses represent calculated IC50 values corrected
to 10 µM ATP.
Of all the compounds examined, NL-71-101 showed the
best combination of selectivity and activity against PKB. As
shown in Figure 3, NL-71-101 was almost as potent as H-89 transiently transfected with a plasmid encoding the PKB
against PKB (IC50 of NL-71-101 ) 4 µM, of H-89 ) 2.5 substrate, GSK3. Forty-eight hours later, the transfected cells
µM), but had lost potency against PKA (IC50 of NL-71-101 were treated with NL-71-101 at various concentrations, for
) 9 µM, of H-89 ) 0.035 µM). Furthermore, NL-71-101 4 h. IGF-1 (10 ng/mL) was added and 10 min later lysates
had also lost potency against two other Ser/Thr kinases, PKC were prepared, separated by SDS-PAGE, and subjected to
and PKG as compared to H89 and was inactive toward Western analysis. Phosphorylation of GSK3 was detected
p38MAPK, even at concentrations of 100 µM (Table 3). Thus, using a specific antibody to GSK3 phosphorylated at residue
using SAR studies, we have converted H-89, which is a 70- Ser21, and the intensities of the bands corresponding to
fold more potent inhibitor of PKA than of PKB, into a phosphorylated GSK3â were normalized to the intensities
compound that is dramatically reduced in its potency against of the total (phosphorylated plus unphosphorylated) GSK3â
PKA, without affecting the degree of PKB inhibition. bands, detected using anti-GSK3â antibody (Figure 5). NL-
NL-71-101 is an ATP-competitiVe inhibitor. The kinetics 71-101 inhibited GSK3 phosphorylation in a dose-dependent
of inhibition of PKB and PKA by NL-71-101were examined manner with an IC50 ∼ 25 µM (Figure 5). The effect of NL-
using radioactive kinase assays. Enzyme activities were 71-101 treatment of HEK293 cells on GSK3 phosphorylation
assayed at several concentrations of NL-71-101, over a range levels was comparable to the specific inhibition of PKB
of ATP concentrations (5-50 µM). Reciprocal plots of activity by a dominant negative, myristoylated kinase dead
reaction velocity versus ATP concentration intersected on PKB mutant (KD-PKB; Figure 5). Note that the amount of
the Y-axis (Figure 4), showing that NL-71-101 is competitive GSK3 was low in the cells cotransfected with both GSK3
with ATP. and KD-PKB plasmids; this may be due to competition for
NL-71-101 Inhibits Phosphorylation of the PKB Substrate, transcription/translation factors when the two genes were
GSK3, in Intact Cells. To examine the effect of NL-71-101 overexpressed simultaneously. NL-71-101 also inhibited the
on the activity of PKB in intact cells, HEK293 cells were phosphorylation of residue Ser 473 of PKB, a site that is
10312 Biochemistry, Vol. 41, No. 32, 2002 Reuveni et al.

FIGURE 5: Treatment with NL-71-101 inhibits GSK3 phosphory-

lation in HEK293 cells. HEK293 cells were transiently transfected
with a GSK3 encoding plasmid and treated with NL-71-101 at the
indicated concentrations for 4 h, or transiently cotransfected with
FIGURE 4: Kinetics of PKB and PKA kinase inhibition by NL-71- both the GSK3 plasmid and another plasmid encoding the dominant
101. The activities of PKB and PKA were determined using the negative PKB mutant, myr-KD-PKB. Cells were then exposed to
radioactive assays, as described in Materials and Methods. The IGF-1 for 10 minutes, to activate PKB. Lysates were prepared and
reactions were performed in the presence of a fixed concentration subjected to Western analysis, using anti-phospho(Thr308)Akt1-
of the 7-mer peptide substrate (20 µM), varying concentrations of (PKBR), anti-phospho(Ser473)Akt(PKB), or anti-phospho(Ser21)-
ATP (5, 7.5, 10, 20, 50 µM), and varying concentrations of NL- GSK3R antibodies. The latter antibody recognized both R and â
71-101 (0, 2, 4, 8, 12 µM). isoforms of GSK3, when phosphorylated. Following stripping, the
blots were treated with anti-Akt1/2(PKBR/â) or anti-GSK3â
antibodies. The lower panel shows the quantitation of phosphory-
believed to be subject to autophosphorylation (14). There lated GSK3â, normalized to total GSK3â levels.
was a slight effect on phosphorylation of residue Thr308.
This might be due to preclusion of enzyme activation due compound, NL-71-101B, which might have been expected
to binding of the inhibitor, or to nonspecific effects of the to have comparable overall toxic properties. NL-71-101B,
inhibitor. In summary, NL-71-101 affects the ability of although similar to NL-71-101 in structure, did not inhibit
endogenous PKB to phosphorylate its substrate, GSK3, in PKB in vitro, up to a concentration of 100 µM (Table 1).
intact cells. As can be seen in Figure 6B, neither did NL-71-101b lead
NL-71-101 Induces Apoptosis. Since PKB is an anti- to apoptosis of OVCAR-3 cells. Thus, there is a correlation
apoptotic kinase, we investigated the effects of NL-71-101 between the ability of a compound to inhibit PKB in vitro,
on apoptosis. We treated OVCAR-3 ovarian carcinoma cells, and its ability to induce apoptosis upon treatment of intact
in which PKB is strongly amplified (8), with NL-71-101 at cells.
various concentrations. Apoptosis was assessed using both Since NL-71-101 inhibits both PKA and PKB activities
annexin-V staining of phosphatidyl-serine residues on the (Figure 3), we asked whether the induction of apoptosis in
cell surface (an indicator of early apoptosis; Figure 6A) and OVCAR-3 cells was a consequence of PKA or PKB
propidium iodide staining of the sub-G1 cell population inhibition. To this end, we compared the effects of NL-71-
(which gains prominence later in apoptosis; Figure 6B). 101 with those of its parent, the much more selective PKA
Stained cells were counted by FACS. Treatment of OVCAR-3 inhibitor, H-89. We argued as follows: If inhibition of PKA
activity causes apoptosis of OVCAR-3 cells, H-89 should
cells with NL-71-101 led to a dose-dependent increase in
lead to apoptosis at concentrations at which it inhibits PKA.
apoptosis (Figure 6C). A significant increase in the propor-
If, on the other hand, inhibition of PKB activity causes
tion of apoptotic cells was evident after treatment with 50
apoptosis, H-89 should only lead to apoptosis at the much
µM NL-71-101, and more than half of the cells were
higher concentrations at which it also inhibits PKB. NL-71-
apoptotic after treatment with 100 µM NL-71-101 (Figure 101 inhibits PKA in vitro with an IC50 of 9 µM (Figure 3),
6). Thus, NL-71-101 caused intact OVCAR-3 cells to and leads to significant apoptosis of cells at 50 µM (Figure
undergo apoptosis. 6). Since H-89 inhibits PKA in vitro with an IC50 of 0.035
To assess whether the apoptosis of OVCAR-3 cells in µM, we might expect apoptosis to occur at concentrations
response to NL-71-101 was due to a possible general toxic of 0.2 µM (i.e., [50 × 0.035/9] µM), had PKA been involved.
effect of the compound rather than due to inhibition of PKB, In fact, up to concentrations of 20 µM, H-89 had no effect
we compared NL-71-101 treatment with that of a related on apoptosis of OVCAR-3 cells. Apoptosis was induced at
Modification of a Selective PKA Inhibitor Biochemistry, Vol. 41, No. 32, 2002 10313

FIGURE 6: Treatment with NL-71-101 induces apoptosis of OVCAR-3 cells. OVCAR-3 cells were exposed to NL-71-101 at the indicated
concentrations for 40 h in medium containing 10% FCS. (A) Cells were stained with annexin-V, and the proportion of the apoptotic
population was quantified by FACS analysis. (B) Cells were stained with propidium iodide and the proportion of the sub-G1 population
was determined by FACS analysis. (C) Graphic presentation of dose-dependent apoptosis of OVCAR-3 cells following 40 h of treatment
with NL-71-101 (summary of section A), NL-71-101B and H-89, using annexin-V staining.

40 µM H-89, at which concentrations H-89 would be DISCUSSION

expected to inhibit PKB in intact cells, by a similar argument.
It thus appears that induction of apoptosis by NL-71-101 Recognizing the pivotal role PKB plays in oncogenesis,
correlates with inhibition of PKB, rather than with inhibition we have embarked on an attempt to generate PKB inhibitors
of PKA. as prospective antitumor agents. The strategy adopted in the
10314 Biochemistry, Vol. 41, No. 32, 2002 Reuveni et al.
present study was based on the assumption that the high 3. Jones, P. F., Jakubowicz, T., Pitossi, F. J., Maurer, F., and
degree of homology between PKB and PKA should enable Hemmings, B. A. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 4171-
5.
one to develop PKB inhibitors from selective PKA inhibitors.
4. Staal, S. P. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 5034-7.
Our initial screen of known Ser/Thr kinase inhibitors indeed
5. Jones, P. F., Jakubowicz, T., and Hemmings, B. A. (1991) Cell
showed that H-89 is a relatively potent inhibitor of PKB, Regul. 2, 1001-9.
albeit much more potent (70-fold) against PKA. H-89 is an 6. Bellacosa, A., de Feo, D., Godwin, A. K., Bell, D. W., Cheng, J.
attractive starting point for drug development: The modular Q., Altomare, D. A., Wan, M., Dubeau, L., Scambia, G.,
structure of H-89 allowed us to employ parallel chemistry Masciullo, V., et al. (1995) Int. J. Cancer 64, 280-5.
and SAR to identify a moiety, the cinnamoyl tail, which 7. Cheng, J. Q., Ruggeri, B., Klein, W. M., Sonoda, G., Altomare,
could be modified to change the selectivity of the compound D. A., Watson, D. K., and Testa, J. R. (1996) Proc. Natl. Acad.
Sci. U.S.A. 93, 3636-41.
while retaining its affinity for PKB. Furthermore, we
8. Cheng, J. Q., Godwin, A. K., Bellacosa, A., Taguchi, T., Franke,
developed an ELISA assay to allow high throughput testing T. F., Hamilton, T. C., Tsichlis, P. N., and Testa, J. R. (1992)
of potential inhibitors. We have succeeded in reversing the Proc. Natl. Acad. Sci. U.S.A. 89, 9267-71.
selectivity of H-89 toward PKA/PKB, arriving at a com- 9. Nakatani, K., Thompson, D. A., Barthel, A., Sakaue, H., Liu, W.,
pound, NL-71-101, which is slightly more potent against Weigel, R. J., and Roth, R. A. (1999) J. Biol. Chem. 274, 21528-
PKB than PKA. Interestingly, NL71-101 inhibits PKG with 32.
similar potency to its inhibition of PKA but is a poor inhibitor 10. Downward, J. (1998) Curr. Opin. Cell. Biol. 10, 262-7.
of PKC (Table 3), in contrast to H89 which inhibits PKG 11. Chan, T. O., Rittenhouse, S. E., and Tsichlis, P. N. (1999) Annu.
10-fold less well than PKA (Table 3). Also, NL 71-101 ReV. Biochem. 68, 965-1014.
inhibits PKB 26-fold better than its inhibitory effect on PKC, 12. Kandel, E. S., and Hay, N. (1999) Exp. Cell Res. 253, 210-29.
whereas H89 inhibits PKB only 6-fold better than its 13. Alessi, D. R., James, S. R., Downes, C. P., Holmes, A. B., Gaffney,
P. R., Reese, C. B., and Cohen, P. (1997) Curr. Biol. 7, 261-9.
inhibition of PKC (Table 3).
14. Toker, A., and Newton, A. C. (2000) J. Biol. Chem. 275, 8271-
The effects of NL-71-101 on the viability of intact cells 4.
are consistent with its function as a PKB inhibitor. We have 15. Alessi, D. R., Andjelkovic, M., Caudwell, B., Cron, P., Morrice,
shown, using the easily transfectable HEK293 cell line, that N., Cohen, P., and Hemmings, B. A. (1996) EMBO J. 15, 6541-
NL-71-101 inhibits endogenous PKB, leading to a reduction 51.
in phosphorylation of exogenous GSK3 (Figure 5). NL-71- 16. Balendran, A., Casamayor, A., Deak, M., Paterson, A., Gaffney,
P., Currie, R., Downes, C. P., and Alessi, D. R. (1999) Curr. Biol.
101 induced apoptosis in two carcinoma cell lines in which 9, 393-404.
PKB is amplified (7), namely, the ovarian carcinoma line 17. Delcommenne, M., Tan, C., Gray, V., Rue, L., Woodgett, J., and
OVCAR-3 (Figure 6) and the pancreatic carcinoma line Dedhar, S. (1998) Proc. Natl. Acad. Sci. U.S.A. 95, 11211-6.
PANC-1 (data not shown). These data validate our postula- 18. Chen, R., Kim, O., Yang, J., Sato, K., Eisenmann, K. M.,
tion that a suitable PKB inhibitor should be effective in McCarthy, J. B., Chen, H., and Qiu, Y. (2001) J. Biol. Chem.
treatment of cancers in which the activity of PKB is 276, 31858-62.
enhanced. 19. Dudek, H., Datta, S. R., Franke, T. F., Birnbaum, M. J., Yao, R.,
Cooper, G. M., Segal, R. A., Kaplan, D. R., and Greenberg, M.
PKB represses a number of apoptotic pathways (10-12). E. (1997) Science 275, 661-5.
A recent report has shown that PKB phosphorylates and 20. Cross, D. A., Alessi, D. R., Cohen, P., Andjelkovich, M., and
inactivates apoptosis signal-regulating kinase 1 (ASK1), an Hemmings, B. A. (1995) Nature 378, 785-9.
upstream activator of the stress-activated MAP kinase, p38 21. Franke, T. F., Yang, S. I., Chan, T. O., Datta, K., Kazlauskas, A.,
(32). We have preliminary evidence that this may be one Morrison, D. K., Kaplan, D. R., and Tsichlis, P. N. (1995) Cell
81, 727-36.
pathway whereby NL-71-101 induces apoptosis, as we saw
22. Datta, K., Bellacosa, A., Chan, T. O., and Tsichlis, P. N. (1996)
dose-dependent phosphorylation of p38 following treatment J. Biol. Chem. 271, 30835-9.
of OVCAR-3 cells with NL-71-101 (data not shown). We 23. Liu, A. X., Testa, J. R., Hamilton, T. C., Jove, R., Nicosia, S. V.,
cannot, however, rule out the possibility that NL-71-101 led and Cheng, J. Q. (1998) Cancer Res. 58, 2973-7.
to general toxic effects and hence to stress pathway activa- 24. Wu, X., Senechal, K., Neshat, M. S., Whang, Y. E., and Sawyers,
tion, particularly in light of the high concentrations of the C. L. (1998) Proc. Natl. Acad. Sci. U.S.A. 95, 15587-91.
compound to which the cells were exposed. NL-71-101 acts 25. Thompson, L. A., and Ellman, J. A. (1996) Chem ReV. 96, 555-
at high concentrations, but it is a promising lead compound 600.
for further development of high affinity, selective PKB 26. Dole, R. E., and Nelson, K. H. (1999) J. Comb. Chem. 1.
inhibitors. 27. Andjelkovic, M., Alessi, D. R., Meier, R., Fernandez, A., Lamb,
N. J., Frech, M., Cron, P., Cohen, P., Lucocq, J. M., and
The crystallographic structure of PKA with H89 was Hemmings, B. A. (1997) J. Biol. Chem. 272, 31515-24.
reported (33), showing that H89 occupies the ATP binding 28. Shaw, M., Cohen, P., and Alessi, D. R. (1997) FEBS Lett. 416,
site of the kinase. Based on this data and on our kinetic 307-11.
studies, showing that NL-71-101 functions as an ATP 29. Boussif, O., Lezoualc’h, F., Zanta, M. A., Mergny, M. D.,
competitive inhibitor (Figure 4), further studies are currently Scherman, D., Demeneix, B., and Behr, J. P. (1995) Proc. Natl.
Acad. Sci. U.S.A. 92, 7297-301.
being performed with the aim to improve the affinities of
30. Alessi, D. R., Caudwell, F. B., Andjelkovic, M., Hemmings, B.
NL-71-101 analogues towards the PKB kinase domain. A., and Cohen, P. (1996) FEBS Lett. 399, 333-8.
31. Roskoski, R., Jr. (1983) Methods Enzymol. 99, 3-6.
REFERENCES 32. Kim, A. H., Khursigara, G., Sun, X., Franke, T. F., and Chao, M.
V. (2001) Mol. Cell Biol. 21, 893-901.
1. Staal, S. P., and Hartley, J. W. (1988) J. Exp. Med. 167, 1259-
64. 33. Engh, R. A., Girod, A., Kinzel, V., Huber, R., and Bossemeyer,
D. (1996) J. Biol. Chem. 271, 26157-64.
2. Bellacosa, A., Testa, J. R., Staal, S. P., and Tsichlis, P. N. (1991)
Science 254, 274-7. BI0202530